United States Office of Science March 1995
Environmental Protection and Technology
Agency Washington, DC 20460
&EPA Regulatory Impact
Analysis Of The
Great Lakes Water Quality
Guidance
Final Report
Recycled/Recyclable • Printed with Vegetable Based Inks on Recycled Paper (20% Postconsumer)
-------
-------
REGULATORY IMPACT ANALYSIS
OF THE FINAL GREAT LAKES
WATER QUALITY GUIDANCE
Final Report
Prepared for:
Office of Science and Technology
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
March 1995
-------
-------
ACKNOWLEDGEMENTS AND DISCLAIMER
This report has been reviewed and approved for publication by the U.S. Environmental
Protection Agency, Office of Science and Technology. This report was prepared with the
support of RCG/Hagler Bailly (contract 68-C4-0060), under the direction and review of Mr.
Mark L. Morris of the Office of Science and Technology. Neither the United States
Government nor any of its employees, contractors, subcontractors, or their employees makes
any warranty, expressed or implied, or assumes any legal liability or responsibility for any
third party's use of or the results of such use of any information, apparatus, product, or
process discussed in this report, or represents that its use by such third party would not
infringe on privately owned rights.
-------
-------
CONTENTS
List of Tables
List of Figures
Executive Summary
I
Chapter 1 Introduction 1-1
Chapter 2 The Requirement for the Guidance
2.1 The Statutory Requirement for the Guidance ! 2-1
2.2 An Overview of Environmental Concerns in the Great Lakes Basin 2-2
2.2.1 Adverse Effects Caused by BCCs and Other Contaminants 2-2
2.2.2 The Effects of Contaminants in the Great Lakes 2-4
2.3 Conclusion 2-11
Chapter 3 Alternatives Considered in Developing the Final Guidance 3-1
Chapter 4 Revised Analysis of Costs and Cost-Effectiveness
4.1 Changes to the Final Guidance Impacting the Cost Analysis 4-1
4.2 Modifications to the Methodology for Estimating Costs 4-4
4.3 Results . . 4-10
Chapter 5 Analyses Related to the Attribution of Benefits to the Guidance
5.1 Introduction 5-1
5.2 Studies and Data Related to the Point Source Contribution to Loadings
of Guidance-Impacted Contaminants 5-2
5.2.1 Available Data on the Relative Contribution of Point Source
Loadings \ 5-2
5.2.2 Atmospheric Input of PCBs, t-DDT, Benzo(a)pyrene, Lead, and
Mirex to the Great Lakes I 5-4
5.2.3 The Green Bay Mass Balance Study 1 5-6
5.2.4 Conclusions Regarding the Relative Contribution of Point
Sources to Total Loadings 5-6
-------
page u
5.3 A Screening Analysis of Potential Nonpoint Sources of Contaminants in
the Basin ; 5"6
5.3.1 Air Emissions as a Potential Source of Contaminants in the
Basin 5'7
5.3.2 Agricultural Runoff as a Potential Nonpoint Source of Guidance-
Regulated Contaminants in the Basin 5-9
5.3.3 National Priorities List Sites as a Potential Nonpoint Source of
Guidance-Regulated Contaminants in the Basin 5-11
5.3.4 Summary of Potential Sources of Loadings 5-12
5.4 Modeling the Relative Contribution of Point Source Loadings to Fish
Tissue Concentrations • 5"13
5.4.1 Technical Approach 5-13
5.4.2 Results and Discussion 5-16
5.4.3 Summary of Model and Results 5-20
5.5 Cost-Effectiveness of Point Source Controls 5-22
Chapter 6 Updated Risk Assessments for Great Lakes Anglers
6.1 Great Lakes Sport Angler Risk Assessment 6-2
6.2 Risk Assessment for Native Americans 6'18
6.3 Summary of Risk Assessment Results 6"30
Chapter 7 Revised Case Study Benefit-Cost Analyses
7.1 Revisions to the Case Study Benefits Estimates 7-1
7.1.1 The Lower Fox River/Green Bay Case Study 7-2
7.1.2 The Saginaw River/Saginaw Bay Case Study 7-10
7.1.3 The Black River Case Study • • 7-16
7.2 Comparison of Benefits and Costs for the Case Studies 7-19
7.3 Case Study Representativeness 7-21
Chapter 8 References • 8-1
Appendices:
A Characteristics and Toxicity of Contaminants of Concern for the Guidance
B Characterization of the Positive Economic Impacts of the Guidance on the Great Lakes
Region
C Sources of Air Emissions
D Great Lakes Exposure and Bioaccumulation Model
E Supplemental Information to the Basinwide Risk Assessments
-------
TABLES
S-l Summary of Annualized Compliance Costs of the Final Guidance S-3
S-2 Estimated Share of Total Loadings Attributable to Point Sources for the Great
Lakes S-6
S-3 Model-Estimated Threshold Exceedences S-7
S-4 Summary of Potential Reduction in Angler Excess Cancer Cases Anticipated to
Result from the Guidance S-10
S-5 Comparison of the Potential Benefits to the Potential Costs of the Guidance for
the Case Study Areas S-14
2-1 Fish Tissue Concentrations from Various Sources 2-10
3-1 Issues Considered in Developing the Final Guidance 3-2
4-1 Pollutants Considered in the Analysis of Costs of the Final Guidance 4-6
4-2 Summary of Annualized Compliance Costs of the Final Guidance 4-11
4-3 Estimated Annualized Compliance Costs for the Final Guidance by Industry
and Cost Category Low Estimate 4-12
4-4 Estimated Annualized Compliance Costs for the Final Guidance by Industry
and Cost Category High Estimate 4-13
4-5 Estimated Unweighted Baseline Pollutant Loadings and Reductions in Loadings
Anticipated to Result from the Guidance Basinwide 4-15
4-6 Estimated Toxic Weighted Pounds Equivalent Pollutant Loadings and
Reductions in Loadings Anticipated to Result from the Guidance Basinwide ... 4-17
5-1 Relative Contribution of Point Source Loadings to Total Loadings in the Great
Lakes Basin 5-3
5-2 Relative Contribution of Direct Point Source Discharges to Total Inputs of
PCBs and Lead in Lakes Superior and Huron 5-4
5-3 Relative Contribution of Atmospheric Deposition to Total Inputs of Five
Contaminants in the Great Lakes 5-5
5-4 Estimated Share of Total Loadings Attributable to Point Sources for the Great
Lakes - 5-7
5-5 Air Emissions and Long-Range Transport as a Possible Source of Pollutants
Associated with Guidance-Related Reductions 5-8
5-6 Summary of Results of Screening Analysis for Air Emissions as a Potentially
Significant Source of Loadings in the Great Lakes Basin 5-9
5-7 Annual Quantities of Pesticides Used on Corn and Soybean Crops in U.S. and
Canadian Counties Bordering the Great Lakes : 5-10
5-8 Number of CERCLA Sites in the Great Lakes Basin •; 5-12
5-9 Model-Estimated Fish Tissue Concentrations 5-20
-------
page iv
5-10 Model-Estimated Threshold Exceedences 5-20
5-11 Model-Estimated Fish Tissue PCB Concentrations >-"
6-1 Great Lakes Basin Counties ; "
6-2 Characterization of Great Lakes Sport Angler Fish Consumption 6O
6-3 Chemical Toxicity Values and Fish Tissue Concentrations °-iu
6-4 Exposure Assumptions Used to Calculate Risks to Sport Fishermen 6-11
6-5 Baseline Potential Excess Cancer Cases for Sport Fishermen 6-12
6-6 Percentage of Each Compound's Contribution to the Total Carcinogenic Risk . . 6-13
6-7 Baseline Systemic Risks for Sport Fishermen • 6'14
6-8 Estimated Reduction in Fish Tissue Contaminant Concentrations Due to the
_ ., 6-15
Guidance ' ' .
6-9 Potential Reduction in Sport Angler Excess Cancer Cases Due to the Guidance . 6-16
6-10 Potential Impact of the Guidance on Systemic Risks to Sport Anglers ••••••• 6'17
6-11 Characterization of the Subsistence Fishing among Native American Tribes in
the Great Lakes Basin
6-12 Estimated Fish Consumption Values for Native Americans and Subsistence
. . 6-26
Anglers ' .
6-13 Exposure Assumptions Used to Calculate Health Risks to Native American
Tribes in the Great Lakes Basin ™
6-14 Baseline Potential Excess Cancer Cases for Native Americans 6-28
6-15 Percentage of Each Compound's Contribution to the Total Carcinogenic Risk . . 6-29
6-16 Potential Reduction in Native American Excess Cancer Cases due to the
_ .. 6-30
Guidance
6-17 Potential Impact of the Guidance on Average Systemic Risks to Native
. 6-30
Americans • • • • • • •;
6-18 Summary of Potential Reduction in Angler Excess Cancer Cases due to the
Guidance
7-1 Baseline Point Source Loadings and Reductions in Loadings Anticipated to
Result from the Guidance in the Fox River Case Study Area 7-3
7-2 Toxic-Weighted Baseline Point Source Loadings and Reductions in Loadings
Anticipated to Result from the Guidance in the Fox River Case Study Area 7-4
7-3 Estimated Reduction in Fish Tissue Contaminant Concentrations Anticipated to
Result from the Guidance for the Fox River Case Study Area . . , "-7
7-4 Baseline Sport Angler Excess Cancer Risks and Reductions in Risks
Attributable to the Guidance for the Fox River Case Study Area 7-7
7-5 Baseline Sport Angler Excess Cancer Cases and Reductions in Cancer Cases
Attributable to the Guidance for the Fox River Case Study Area 7-8
7-6 Potential Annual Benefits Attributable to the Guidance for the Fox River/Green
Bay Case Study Area .•; • • •
7-7 Baseline Point Source Loadings and Reductions in Loadings Anticipated to
Result from the Guidance in the Saginaw River Case Study Area 7-11
-------
.page v
7-8 Toxic-Weighted Baseline Point Source Loadings and Reductions in Loadings
Anticipated to Result from the Guidance in the Saginaw River Case Study Area 7-12
7-9 Estimated Reduction in Fish Tissue Contaminant Concentrations Anticipated to
Result from the Guidance for the Saginaw River/Bay Case Study Area 7-14
7-10 Baseline Sport Angler Excess Cancer Risks and Reductions in Risks
Attributable to the Guidance for the Saginaw River/Bay Case Study Area 7-15
7-11 Baseline Sport Angler Excess Cancer Cases and Reductions in Cancer Cases
Attributable to the Guidance for the Saginaw River/Bay Case Study Area 7-15
7-12 Potential Annual Benefits Attributable to the Guidance for the Saginaw River
Case Study Area 7-16
7-13 Baseline Point Source Loadings and Reductions in Loadings Anticipated to
Result from the Guidance in the Black River Case Study Area 7-17
7-14 Toxic-Weighted Baseline Point Source Loadings and Reductions in Loadings
Anticipated to Result from the Guidance in the Black River Case Study Area . . 7-18
7-15 Potential Annual Benefits Attributable to the Guidance for the Black River
Case Study Area 7-19
7-16 Comparison of the Potential Benefits to the Potential Costs of the Guidance for
the Case Study Areas 7-20
7-17 Case Study Representativeness 7-23
-------
-------
FIGURES
S-l Relationship Between Exceedences of Fish Consumption Advisories for PCBs
and Percent Contribution of Point Source Loadings to Total Loadings S-8
5-1 Schematic Representation of Great Lakes Exposure and Food Web Model .... 5-15
5-2 Model-Estimated Probability Distribution of Tissue PCB Concentrations ...... 5-17
5-3 Model-Estimated Probability Distribution of Baseline PCB Concentrations in
Omnivores ^-18
5-4 Model-Estimated Probability Distribution of Baseline PCB Concentrations in
Piscivore Fillets • 5'19
5-5 Model-Estimated Probability Distribution of PCB Concentrations in Piscivore
Fillets under a Scenario of a 50% Reduction in Point Source Loadings 5-21
5-6 Relationship Between Reductions and Exceedences of Fish Consumption Advisories
for PCBs and Percent Contribution of Point Source Loadings to Total Loadings 5-23
6-1 Density of Low-Income Minorities in Relation to the Location of Point Sources
in the Great Lakes Basin ...... 6-7
6-2 Density of Fishing Licenses by County in Relation to the Location of Point
Source Dischargers in the Great Lakes Basin 6-8
6-3 Chippewa Treaty Ceded Areas 6-19
6-4 Location of Native Americans Engaged in Subsistence Fishing in the Great
Lakes Basin in Relation to Point Source Dischargers 6-25
-------
-------
EXECUTIVE SUMMARY
This Regulatory Impact Analysis (RIA) updates the benefit-cost analysis presented in the RIA
for the proposed Great Lakes Water Quality Guidance (the Guidance) dated April 1993
(RCG/Hagler Bailly, 1993). The Guidance provides direction to the Great Lakes states and
tribes on minimum water quality standards, and contains numeric water quality criteria for 30
pollutants as well as methodologies for the development of water quality criteria for
additional pollutants discharged to these waters. It also provides guidance to Great Lakes
states and tribes on antidegradation policies and implementation procedures.
I
This RIA was prepared in compliance with Executive Order 12866, which requires federal
agencies to perform an analysis comparing the benefits and costs of the regulation, analyze
alternative approaches to the regulation, and identify the need for the regulation for each
major rule proposed or promulgated. The revisions and updates presented here have been
made primarily in response to comments received during the public comment period and from
the Office of Management and Budget (OMB). This revised RIA is provided as a supplement
to the April 1993 RIA and, as such, does not repeat information presented previously.
The Great Lakes Water Quality Guidance "was developed under the Great Lakes Critical
Programs Act (CPA) of 1990. The CPA codified the Great Lakes Water Quality Initiative
(GLWQI) effort, a joint endeavor of the Great Lakes states and EPA to address water quality
concerns in the Great Lakes system associated with toxic water pollutants. As part of the
basinwide effort to improve and protect water quality, the CPA (CWA Section 118(c)(2))
required EPA to publish proposed water quality guidance for the Great Lakes Basin which
conforms with the objectives and provisions of the Great Lakes Water; Quality Agreement
(under which the United States and Canadian governments established common water quality
objectives for the Great Lakes System) and is no less restrictive than provisions of the Clean
Water Act and national water quality criteria and guidance. The CPA required that the
guidance specify minimum water quality criteria protective of human health, aquatic life, and
wildlife; antidegradation policies; and implementation procedures. :
S.I ALTERNATIVES CONSIDERED IN DEVELOPING THE FINAL GUIDANCE
Comments on the proposed Guidance raised several issues that warranted further
consideration prior to finalizing the Guidance. In addition, there were issues that remained
unresolved at the time the proposed Guidance was formulated. Several options were
considered for issues with a potential significant impact on costs, issues related to regulatory
requirements, issues related to criteria and standards, and issues related to implementation of
the Guidance. The evaluation of alternative options for the final Guidance was based on
-------
EXECUTIVE SUMMARY >• S-2
_—. —
analyses of the options for those issues with significant impacts on costs ("cost drivers").
Numerous analyses were conducted to evaluate the impact of the cost drivers on the total
compliance costs of the Guidance. The major issues driving costs are related to detection
levels intake credits, Tier H criteria, wildlife criteria and criteria for mercury eliminating
mixing zones for BCCs, accounting for the additive effects of carcinogens, and the
antidegradation provisions of the Guidance.
S.2 REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS
A revised analysis of costs and cost-effectiveness was conducted for the final Guidance. The
revisions reflect changes to the proposed Guidance as well as modifications to the analytical
methodology used to estimate costs and cost-effectiveness. In general, the modeling
methodology described in the April 1993 RIA was employed for the cost analysis of the final
Guidance However, methodology revisions were made to more accurately project the costs to
the regulated community and to better account for the pollutant load reductions.
Changes to the Guidance which impacted costs include the promulgation of criteria for metals
in the dissolved form, intake credit provisions, and provisions for the consideration ot the
additivity of human carcinogenic effects when two or more carcinogens are discharged.
Modifications to the original methodology for estimating costs and pollutant load reduct.ons
include the consideration of Tier II pollutants and a revised compliance cost decision matrix,
and the use of updated data for sample facilities, indirect dischargers, and toxic weights.
Estimated Costs of the Guidance
As shown in Table S-l, the total annuaiized costs of implementing the final Guidance to
direct and indirect dischargers is estimated to range from $61 million to $376 million. This
reflects a downward revision from the estimated cost of the proposed Guidance. The
downward revision in costs is largely attributable to the intake credit provisions of the final
Guidance and the use of dissolved metals criteria. The lowering of the permit baseline since
the proposal of the Guidance also results in an overall decrease in compliance costs and load
reductions.
Under the low estimate of compliance cost, direct dischargers incur about 67% of total costs
while indirect dischargers incur about 33%. Under the high estimate of compliance cost,
direct dischargers account for about 98% of the total cost, and indirect dischargers account tor
2%.
Pollutant Reductions Anticipated to Result from the Guidance
Baseline point source pollutant loadings in the Great Lakes Basin are projected to be just over
35 million toxic pounds-equivalent per year (Ibs-eq/year). Under the low compliance cost
-------
EXECUTIVE SUMMARY > S-3
Table S-l i
Summary of Annualized Compliance Costs of the Finial Guidance
Cost Categories
Major Direct Dischargers — Industrial
Major Direct Dischargers — Municipal
Minor Direct Dischargers
Indirect Dischargers
Total
Number
of
Facilities
272
316
3,207
3,528
7,323
Estimated Costs
(Millions of First Quarter 1994 $)
Low
15.7
23.8 !
1.6
19.9
61.0
High
108.2
259.8
1.6
6.6
376.2
Source: SAIC, 1995.
estimate, pollutant loadings would be reduced by 5.8 million Ibs-eq/year basinwide, which
represents a 16% reduction from baseline levels. Under the high compliance cost estimate,
pollutant loadings would be reduced by 7.6 million Ibs-eq/year in the basin, which represents
a 22% reduction of the baseline loadings.
Under the low compliance cost estimate, the largest pollutant load reductions occur for
dieldrin and lead, which account for over 50% of the toxic-weighted load reduction.
Chlordane, heptachlor, and pentachlorobenzene are also expected to be reduced by significant
amounts from the baseline. Under the high compliance cost estimate, the largest pollutant load
reductions occur for heptachlor, dieldrin, and lead, which account for about 70% of the toxic-
weighted load reduction. Approximately 80% of the pollutant load reduction for the final
Guidance, regardless of the scenario, is attributable to reducing bioaccumulative pollutants of
concern (BCCs) as a result of pollution minimization plans and end-of-pipe treatment.
Baseline pollutant loadings and reductions in loadings were also estimated for the three
economic benefit case study sites. These results are discussed in Section S.5.
Cost-Effectiveness of Pollutant Reductions
Under the low compliance cost estimate, the cost per toxic-weighted pound of pollutant
reduced, or the cost-effectiveness of the final Guidance, is approximately $10.30/lbs-eq.
Under the high cost estimate, cost-effectiveness is estimated at just aver $49/lbs-eq. For
comparative purposes, cost-effectiveness values for effluent limitations guidelines and
standards range from just over $l/lb-eq to over $500/lbs-eq.
-------
EXECUTIVE SUMMARY > S-4
S.3 ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE
GUIDANCE
The revised analysis of benefits for the final Guidance utilized the revised pollutant loading
reductions presented in Section S.2 as well as new information on the role of point sources, m
the basin's toxic-related problems.
The April 1993 RIA described the basis by which a portion of the potential benefits of water
qualityTmprovements illustrated in the case study benefits analyses were attnbuted to he
proposed Guidance, and the numerous sources of uncertainty surrounding these estimates In
generat baseline resource values and the value of water quality toff^™*"* °n
ava\lab e data and applied research. However, data were not available on the potentia
contribution of the proposed Guidance toward attaining the larger water quality objectives.
Because of the lack of information on the attribution issue, benefits were attributed tc.the
Guidance in the case study analyses for illustrative purposes based on general information
atartite Ues. Although me lack of data was not refuted, public comments on the proposed
GuTdarfce L RIA focused on the uncertainty surrounding the attribution of benefits and the
Serf potentially "optimistic" attribution assumptions. OMB commented that EPA should try
to modertiTe expected reductions in fish tissue contaminant concentrations (and thus changes
n fish consumption advisories) resulting from the Guidance. Thus, subsequent research efforts
were directed towards better quantification of the potential impact of the Guidance in bringing
about future toxic-oriented benefits. Efforts were focused on determining the potential
contribution of point source loadings to the toxic-related problems in the basin.
Available Information on the Relative Contribution of Point Sources
One study investigating the relationships between various PCB sources and sinks in the Great
Lake Basics the" ongoing Green Bay/Fox River Mass Balance Study (GBMBS) Prehmmary
results from this modeling effort indicate that point sources contributeApproximately 9.4% of
the total PCB loadings to Green Bay at the mouth of the Fox River (Bierman et ^ 1992>
Table 9-8 p 172) These results imply that the effect of the Guidance on future PCB-onented
benefits in the Fox River/Green Bay case study area may be limited. However, given the
highly site- and contaminant-specific nature of toxic-related water quality impairments, the
results of the GBMBS may not be representative of the universe of sites or pollutants
impacted by the Guidance.
In general there is insufficient data available to estimate total basinwide loadings (and thus
calculate the relative point source contribution) for almost all of the contaminants addressed
by the Guidance, and results are likely to be highly site- and contaminant-specific, tor
example, using limited studies, EPA was able to estimate basinwide loadings only for
mercury, lead, cadmium, and PCBs (Warren, 1993). However, if these results are utilized to
provide a preliminary indication of the relative contribution of point sources (based on Permit
-------
EXECUTIVE SUMMARY > S-5
Compliance System data for point source loadings), the estimated relative contribution of
point sources ranges from approximately 2% to 40%, depending on the contaminant.
Strachan and Eisenreich (1988) estimated inputs of direct industrial and wastewater
dischargers of PCBs and lead to Lakes Superior and Huron, and compared them to total
loadings. Although the data probably underestimate point source loadings, the estimates
demonstrate that the lower Great Lakes receive a greater percentage of total loadings from
point sources than the upper lakes. The upper lakes are estimated to receive a greater share of
their total loadings of toxic contaminants from atmospheric sources, this is due to the relative
lack of local sources and the larger surface area of the upper lakes. Conversely, extensive
pollutant loadings from sources on the Detroit-St. Clair and Niagara leaver systems result in a
relatively higher share of total loadings from point sources to the lower lakes (Strachan and
Eisenreich, 1988). Strachan and Eisenreich estimate that the contribution of point sources
ranges from 0.7% to 1.5% for Lake Superior, and from 2.0% to 7.0% for Lake Huron.
Strachan and Eisenreich (1988) were able to analyze total pollutant loads and the relative
contribution of inputs from the atmosphere for PCBs, t-DDT, benzo(a)pyrene, lead, and mirex
(Lake Ontario only), although they found large uncertainties in the data. Information on the
relative contribution of atmospheric sources to total loadings provides an indication of the
potential contribution of point sources. Strachan and Eisenreich's work showed atmospheric
inputs varied greatly by contaminant and lake. For example, atmospheric inputs of PCBs
ranged from 90% for Lake Superior, to 7% for Lake Ontario.
I
Estimated Contribution of Point Sources to the Toxics-Related Problems in the Basin
Based on the information presented above, assumptions regarding the relative contribution of
point sources to total toxic loadings were developed for each lake (Table S-7). These
assumptions were then used in the benefits analysis to attribute an appropriate share of water
quality benefits to the Guidance. Typically, the share of benefits attributed to the Guidance
was estimated by multiplying the projected pollutant loading reductions resulting from the
point source controls of the Guidance by the estimated contribution of point sources to total
loadings for each lake (as shown in Table S-2). For example, the Guidance is projected to
reduce toxic-weighted loadings at the Saginaw River/Saginaw Bay case study site by 60.5%.
Multiplying this estimate by the estimated range for the relative contribution of point sources
to total Lake Huron loadings (5% to 10%) gives the expected reductions in total loadings
attributable to the Guidance (3.03% to 6.05%). ,
Modeled Reductions in Fish Tissue Contaminant Concentrations
A modeling effort was also designed to estimate the potential reductions in fish tissue
contaminant concentrations and changes in fish consumption advisories resulting from point
source loadings reductions. A generalized exposure and bioaccumulaition model was
developed and applied to PCBs in Lake Michigan's Green Bay. PCB contamination of Green
-------
EXECUTIVE SUMMARY »• S-6
Table S-2
Estimated Share of Total Loadings Attributable to Point Sources for the Great Lakes
Great Lake
Percentage of Total Loadings Attributable to Point Sources
Lake Superior
1-2%
Lake Michigan
5-10%
Lake Huron
5-10%
Lake Erie
10-15%
Lake Ontario
10-15%
Based on Strachan and Eisenreich (1988), Warren (1993), and Bierman et al., (1992).
Bay was modeled because PCB sources and food web parameters were available from the
GBMBS However, Green Bay represents a potential "worst-case" scenario in terms of
potential benefits from point source controls because current point source loadings contribute
a relatively small fraction of total PCB exposure to fish (9.4%) (Bierman et al., 1992).
The model was used to estimate changes in fish tissue concentrations and exceedences of
human and ecological health thresholds under different scenarios of point source loadings
reductions. Reductions in point source loadings in Green Bay were shown to have a modest
impact on fish tissue PCB concentrations and exceedences. For example, as shown in
Table S-3 a 50% reduction in point source loadings reduces baseline exceedences of the
health-based fish tissue concentration threshold from 10.6% to 8.9%. This result occurs
because for this contaminant and this site, existing sediment contamination is the dominant
source of PCB exposure to fish (point sources contribute only 9.4% to loadings), and baseline
exceedences are only 10.6%. However, loading reduction estimates for the Fox River/Green
Bay case study area indicate that the Guidance may reduce point source PCB loadings by
89.6%. As shown in Table S-3, a 90% reduction lowers exceedences of the human health
threshold from 10.6% to 6.8%.
Figure S-l shows the model-estimated change in exceedences of the human health threshold
of 2 ppm (relative percentage reductions) in response to point source loading reductions.
Relative percentage reductions in exceedences from baseline are 2%, 16%, and 36% under
scenarios of 10%, 50%, and 90% reductions in point source loadings, respectively
(represented in Figure S-l as points B, A, and C, respectively). Application to other sites and
conditions may show greater benefits from point source reductions. For example, sites where
point source loadings represent a greater percent of total loadings are expected to show a
greater change from baseline conditions. This is illustrated as moving to the right along the
horizontal axis in Figure S-l. For example, under a scenario of a 50% reduction in point
-------
EXECUTIVE SUMMARY •• S-7
Table S-3
Model-Estimated Threshold Exceedences1
Point Source Loading Scenario
Baseline
10% Reduction
50% Reduction
90% Reduction
Ecological Health2
99.9%
99.8%
99.8%
99.7%
Human Health3
10.6%
10.4%
8.9%
6.8%
1 Concentrations were rounded to the nearest 0.1%.
2 Threshold: 0.3 ppm in omnivore whole tissue.
3 Threshold: 2 ppm in piscivore fillet (e.g., walleye, brown trout).
source loadings, the estimated reduction in exceedences may increase from about 16% (point
A, with point sources contributing 9.4% of total loadings) to about 35% (a point between A
and D, where point sources contribute 20% of total loadings).
S.4 UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS
The April 1^93 RIA contained a preliminary assessment of health-related risks to Great Lakes
Basin sport anglers and potential risk reductions resulting from the proposed Guidance. In
accordance with Executive Order (EO) 12898, which directs federal agencies to incorporate
environmental justice considerations into their missions, additional data and information were
collected to refine the sport angler risk assessment to reflect minority and low-income
exposures, and to conduct a separate assessment for Native Americans engaged in subsistence
fishing in the basin. Since Native Americans fishing on reservations and treaty-ceded fishing
grounds are not required to purchase fishing licenses, they would not be accounted for in the
sport angler assessment.
Fish tissue concentrations for additional pollutants were also incorporated. Carcinogenic and
noncarcinogenic (systemic) risks due to PCB, DDT, mercury, and dieldrin, chlordane,
hexachlorobenzene, 2,3,7,8-TCDD, and toxaphene exposure were assessed. In addition, risks
were assessed on a lake-specific basis to more accurately match fish tissue contaminant
concentrations with the exposed populations. Finally, potential reductions in baseline risks
were updated to reflect the revised loadings reductions anticipated to result from the point
source controls of the Guidance and revised assumptions regarding the contribution of point
sources to total loadings in the basin. ;
-------
EXECUTIVE SUMMARY »• S-8
Figure S-l
Relationship Between Exceedences of Fish Consumption Advisories
for PCBs and Percent Contribution of Point Source Loadings to Total Loadings
100%
10% 20% 30% 40%
Percent Contribution of Point Sources to Total Loadings
50%
Note- Results are model-estimated reductions in exceedences of 2 ppm PCBs in fillets under 10/o, SO/.
or 90% reductions in point source loadings. Reductions increase as the percent of pomt source
loadings to total loadings increase. Green Bay point source loadings were modeled as 9.4/o of
total loadings.
-------
EXECUTIVE SUMMARY > S-9
Data on fishing license sales in the Great Lakes Basin were used to estimate the number of
potentially exposed sport anglers. For Native Americans, tribal representatives, EPA tribal
liaisons, and the Great Lakes Fish and Wildlife Commission were contacted to determine
whether a tribe was engaged in subsistence fishing. Approximately 2.7 million sport anglers
and 13,600 Native Americans in the basin are estimated to be exposed to contaminants in
Great Lakes fish. A literature review of the consumption of Great Lakes fish by sport anglers
and Native American and subsistence anglers was conducted to derive fish ingestion
scenarios. For sport anglers, consumption was found to vary by minority and income status.
For Native Americans, low, moderate, and high consumption scenarios were developed based
on the literature.
Baseline Risk Levels
Baseline lifetime cancer risks for low-income minorities ranged from 2.5 * 10' (Lake
Superior) to 1.2 * 10-2 (Lake Michigan); for other minorities, baseline cancer risks ranged
from 6.5 x 10'4 (Lake Superior) to 3.0 x 10"3 (Lake Michigan); and for all other sport
fishermen, these risks ranged from 9.7 x 10'4 (Lake Superior) to 4.5 x 10'3 (Lake Michigan).
For Native Americans, lifetime cancer risks ranged from 1.8 x 10'3 to 3.7 x 10'2 for the low
and high fish consumption scenarios, respectively. Baseline systemic (noncarcinogenic) risks
to both sport anglers and Native Americans are indicative of a high potential for systemic
injury. The estimated cancer and systemic risks are driven by PCB exposure.
Reductions in Risks due to the Guidance
Table S-4 summarizes the potential reduction of excess cancer cases for sport anglers and
Native Americans anticipated to result from the Guidance. For sport anglers, the Guidance
was estimated to result in a reduction of 24.5 to 46.8 excess lifetime cancer cases. For Native
American subsistence anglers, the Guidance was estimated to result in a reduction of 0.1 to
0.3 excess lifetime cancer cases under the low fish consumption scenario, a conservative
assumption. In total, these risk reductions represent between 0.35 and 0.67 excess cancer
cases per year, and potential monetized benefits of the Guidance of between $0.7 million and
$6.7 million per year (based on the estimated value of a statistical life of between
$2.0 million and $10.0 million (see Violette and Chestnut, 1983; 1986; U.S. EPA, 1989).
Since not all excess cancer cases will necessarily result in mortality, however, the monetized
benefits estimate may be overstated.
The estimated reductions in risks are small because PCB contamination comprises the
majority of both systemic and carcinogenic risks, and the modeled basinwide loadings
reductions show no reductions in PCBs resulting from the Guidance. However, the basinwide
results are based on a sample of 59 facilities, and may result in conservative estimates of
actual basinwide reductions. For example, the projected reductions in loadings for the Fox
River/Green Bay and Saginaw River/Bay case studies, which are based on a more detailed
examination of all facilities in the areas, show an 89% reduction in PCBs from baseline
-------
EXECUTIVE SUMMARY »• S-10
Table S-4
Summary of Potential Reduction in Angler Excess Cancer Cases
Anticipated to Result from the Guidance
Sport Angler: Low Income Minorities1
Sport Angler: Other Minorities1
Sport Angler: Other Sport Fishermen
Native American Subsistence Anglers2
•M^H*«M
Total
~\
2
Note:
Lakeshore counties only. Low income is defined as less than $25,000 per household per year.
Based on low fish ingestion scenario (32 gpd).
Detail may not add to total due to rounding.
levels. Thus, the potential reductions in risks due to the Guidance may be underestimated at
the basinwide level.
For example recalculating the above results using an average of the estimated loadings
SuSto^toe cL study areas instead of the modeled basinw.de: reduces results
n a greater estimate of benefits. PCB loadings are estimated to be reduced by 89.6/. in the
Fox River case study area, 89.4% in the Saginaw Bay/River case study area, and by 0.0/o in
fhe Black River case study area, giving an average PCB reduction of 59.7%^ Using this result,
a^d Ae resulting average "reductions for the additional contaminants included in the risk
SseoTncnt the Guidance would be attributed with a total reduction of between 3., and
eTexcess cancer cases per year, and potential benefits of between $6.6 million and
$601m 1 ion per year. Thus, the more detailed case study loadings reductions assessment
Lggesrhuman hedth benefits almost ten times greater than those derived on the basis of the
59 modeled facilities basinwide.
S.5 REVISED CASE STUDY BENEFIT-COST ANALYSES
A revised analysis of benefits for the case study analyses was conducted for the final
GridLe. The three case study areas,are the lower Fox River and Green Bay "«"«
Wisconsin (Lake Michigan); the Saginaw River and Saginaw Bay in Michigan (Lake Huron),
and the Black River in Ohio (Lake Erie). The case study benefits estimates presented in the
April 1993 RIA were updated to reflect the revised assumptions regarding the relative
contribution of point sources to the toxic-related problems in the basin (see Table S-2) and
revised estimates of the reduction in point source loadings anticipated to result from the
-------
EXECUTIVE SUMMARY »• S-ll
Guidance. In general, the revised results assume a linear relationship between water quality
improvements and benefits (e.g., if the effect of the Guidance is to move water quality
conditions 10% toward "toxic-free," benefits are estimated to be 10% of the total toxics-
oriented net benefits). Original baseline resource values were taken from the April 1993 RIA
and updated to 1994 (first quarter) dollars using the Consumer Price Index.
In addition, based on the additional data and information collected to update the risk
assessments for Great Lakes sport anglers basinwide, human health risk reduction benefits
were calculated for the Fox River/Green Bay and Saginaw River/Bay case study areas. These
benefits categories were discussed in the April 1993 RIA but were not monetized at that time
due to a lack of data.
i
Fox River and Green Bay Case Study
Many of the original benefits estimates for the lower Fox River/Green Bay case study area
were based on or related to the assumption that the Guidance could reasonably be attributed
with 50% of the credit for attaining fishable, "toxic-free" waters. Subsequent analyses
indicated that point sources are likely to constitute between 5% and 10% of total toxic
loadings'to the area. Thus, recreational fishing, commercial fishing, nonconsumptive use, and
nonuse benefits for the lower Fox River/Green Bay case study area were revised to reflect
this finding.
The Guidance is anticipated to reduce toxic-weighted loadings by an estimated 28.2% in the
case study area, including significant reductions in aluminum, benzo(a)pyrene, dieldrin,
hexachlorobenzene; mercury, PCBs, and other chemicals. Multiplying the expected reduction
in loadings (28.2%) by the estimated contribution of point sources to total loadings (5% to
10%) results in an expected reduction in total toxic loadings of 1.41% to 2.82%. The revised
benefits estimates are based on or related to this estimated range. ;
Revised.annual potential benefits of the Guidance are estimated to range from $27,000 to
$3.8 million for recreational fishing, $22,000 to $173,000 for nonconsumptive recreational
use, $19,000 to $120,000 for commercial fishing, and $32,000 to $1.9 million for nonuse
values. In addition, potential annual human health benefits of between $250,000 and
$2.48 million are estimated to result from a reduction of between 0.12 and 0.25 excess cancer
cases per year. Thus, the annual potential benefits of the Guidance associated with these
categories total $0.3 million to $8.5 million. Annualized costs are estimated at $3.6 million.
Saginaw River/Saginaw Bay Case Study
Many of the benefits estimates for the Saginaw River/Bay case study site were originally
based on assumptions about how the Guidance would contribute to improved water quality.
For this revised analysis, the contribution of the Guidance to future toxics-oriented water
quality improvements is based on the expected contribution of point sources to total loadings
-------
EXECUTIVE SUMMARY •• S-12
in the case study area (5% to 10%). Thus, recreational fishing, nonconsumptive recreation,
waterfowl and other hunting, commercial fishing, and nonuse benefits for the Sagmaw
River/Bay case study area were revised to reflect this finding.
The Guidance is anticipated to reduce toxic-weighted loadings by an estimated 60.5% in the
Saginaw River/Bay case study area, including significant reductions in aluminum arsenic
OT DDT lindane, mercury, PCBs, and other chemicals. Multiplying this estimate by the
estimated range for the relative contribution of point sources (5% to 10%) results m an
efp^ld reaction in total loadings of 3.0% to 6.1%. The revised benefits estimates are based
on or related to this estimated range.
Revised potential annual benefits of the .Guidance are estimated to range from $60 000 to
$4.7 million for recreational fishing, $8,000 to $66,000 ^.^"^^^^^^
Sll 000 for hunting, $7,000 to $72,000 for commercial fishing, and $30,000 to^$2.3 million
L nTusTvlTs. In addition, potential human health benefits of between $60,000 and
$580^0 are estimated to resuU from a reduction of between 0.03 and 0.06 excess cancer
cases oer year Thus, the annual potential benefits of the Guidance associated with these
categoriestoTal $168,000 to $7.7 million. Estimated annual compliance costs are $2.6 million.
Black River Case Study
The benefits estimates for the Black River case study area were originally based on the
assumption that the Guidance could be credited with 1% to 5% of the benefits that would
accrue from "total water quality improvement" In this revised analysis, the potential benefi s
attributable to the Guidance are based on the estimated contribution of point sources, tc> total
loading in the case study area (10% to 15%). Thus, recreational fishing; recreational boating,
waterskiing sailboarding! and swimming; and nonuse benefits for the Black River case study
area were revised to reflect this assumption.
The Guidance is anticipated to reduce toxic-weighted loadings by an estimated 36 6% for the
Black River case study area, including significant reductions in fluoride, lindane lead, and
mercury. Multiplying this estimate by the estimated range for the relative contribution of
point sources (10% to 15%) results in an expected reduction in total loadings of 3.7/o to
5.5%. The revised benefits estimates are based on this estimated range.
Revised annual benefits of the Guidance are estimated to range from $251,000 to $719,000
for recreational fishing, $33,000 to $67,000 for nonconsumptive water-based recreation, and
$126 000 to $667 000 million for nonuse values. Total annual potential benefits of the
Guidance associated with these categories range from $0.4 million to $1.5 million. The
estimated compliance cost is $2.1 million per year.
-------
EXECUTIVE SUMMARY •> S-13
Comparison of Benefits and Costs for the Case Studies
Two methods were used to compare the estimated potential case study benefits to estimated
compliance costs in the April 1993 RIA: 1) a direct comparison of annualized benefits and
costs, and 2) a comparison of present-value (i.e., discounted) benefits ,and costs. A comparison
of the revised potential case study benefits and costs using these methods is presented in
Table S-5. Streams of benefits and costs were discounted to incorporate a 10- and 20-year
phase-in of annual benefits and the present value of a stream of annual costs. Capital costs
were annualized using a-7 percent real interest rate. Annual costs and benefits were both
discounted by a 3 percent real rate each year.
Benefits ranges across case study areas are roughly comparable. Potential annual benefits
range from approximately $200,000 to several million dollars, reflecting the uncertainty in the
benefits estimates. Annualized costs are commensurate with annual benefits; costs are
approximately $2 million to $3 million per year for each of the case studies. The net present
values of streams of benefits and costs over 30 years are also generally similar for the Fox
River/Green Bay and Saginaw River/Bay case studies. Discounted costs are above the
discounted benefits ranges for the Black River case study.
Case Study Representativeness
The three case study areas were originally selected for the benefits analysis on the basis of
data availability, on the relative importance of point source discharges to the watersheds'
problems, and to portray spatial diversity throughout the Great Lakes region. There was no
reason to' anticipate, a priori, that the selected sites were atypical of the region. However,
public comments on the April 1993 RIA suggested that the benefits analysis was not
representative of basinwide benefits because it was based on case studies of three "hot spots."
Two additional case study sites were investigated for possible inclusion in the benefits
analysis. It was determined that adding case studies would offer only limited insights, because
sites with readily available data have profiles similar to the existing case studies (e.g., large
historical sediment loads). Therefore, instead of additional case study benefits analyses, an
analysis of the representativeness of the existing case study sites was conducted.
I
The representativeness of the case study sites was assessed by comparing the percentage of
total benefits estimated to accrue in the case study areas to percentage of basinwide costs they
will incur. Benefits-related measures (such as population, recreational angling days, and
nonconsumptive recreation days) were used in place of total benefits for this analysis because
there is no estimate of benefits for the entire Great Lakes Basin.
Overall, there is no evidence to suggest that the three case studies reflect an unrepresentative
level of benefits relative to costs. The three case studies combine to account for nearly 14%
of the Guidance total cost, nearly 17% of the loadings reductions, and between 4% and 10%
-------
EXECUTIVE SUMMARY * S-14
Table S-5
Comparison of the Potential Benefits to the Potential Costs
of the Guidance for the Case Study Areas
(millions of 1994 first quarter dollars)
Benefits
Range
Midpoint of
Benefits
Range
Direct Annualized Comparison1
Discounted Benefits and Costs2
Fox River and Green Bay Case Study
$0.3 - $8.5 $4.5
10-Year Phase-In of Benefits
20-Year Phase-In of Benefits
$5.4 - $133.9
—I, -
$4.1 - $101.4
Saginaw River/Bay Case Study
Direct Annualized Comparison1
•-———^ • '
Discounted Benefits and Costs2
10-Year Phase-In of Benefits
_^—-~~~~"
20-Year Phase-In of Benefits
$0.2 - $7.7
$2.6 - $120.9
$2.0 - $91.5
$4.0
«0^_>^—
$61.7
$46.8
Black River Case Study
^^^—
Direct Annualized Comparison1
_...
Discounted Benefits and Costs2
10-Year Phase-In of Benefits
20-Year Phase-In of Benefits
- $1.5
- $22.7
-$17.2
$0
$1'
$11
r.5
Costs
$3.6
$71.8
$71.8
$2.6
$53.0
$53.0
$2.1
—^^^—^—^—~ Mil —
$42.
$42.
Based on annualized costs assuming a 10-year capital life and reflecting a 7% real
SSS^OSS)1over 30 vears. Annualized costs (assummg 10-year capital life and
7% real interest rate on capital) and benefits are discounted at a 3% real discount rate.
-------
EXECUTIVE SUMMARY > S-15
of the benefits proxies (population, recreational angling, nonconsumptive recreation, and
commercial fishery harvest). Thus, the three case studies represent a reasonably proportionate
share of costs and benefits.
S.6 CONCLUSIONS
i
This RIA represents an updated version of the April 1993 RIA that accompanied the
Guidance at proposal. The RIA has been updated to reflect new cost and pollutant removal
estimates, additional data indicating the share of water quality improvements potentially
attributable to point source controls under the Guidance, more detailed human health risk
assessments, and an evaluation of the representativeness of the case study sites. The revised
analysis indicates that the benefits of the Guidance appear to be commensurate with its
estimated cost.
Among the key quantified benefit-cost findings for the Guidance are the following:
>• The annualized compliance costs are estimated to range from $61 million to
$376 million. Under the lower compliance cost scenario, loadings reductions amount
to an estimated 5.8 million Ibs-eq per year, with a cost-effectiveness of $10.30/lbs-eq.
•> Baseline human health risks faced by subsistence (Native American) and the
approximately 2.7 million sport anglers in the basin, due to the contaminants addressed
by the Guidance, are extremely high (on the order of 1 x 10'2 to 1 x 10'4).
+ Under the model plant-based loadings reductions estimated for the low cost scenario,
the annual value of monetizable risk reduction benefits (reduced incidence of cancer)
amount to between $0.7 million to $6.7 million per year. However, if the basinwide
baseline and post-Guidance loadings are based on the more detailed assessment of
actual facilities in case study areas, then monetized health benefits amount to between
$6.6 million and $60.1 million per year. Thus, health benefits to anglers alone may be
sufficient to justify the compliance costs.
>• In three representative case study areas of the basin, estimates of monetized annual
benefits are commensurate with the estimated annualized costs of the Guidance.
Average benefits across the three case studies range from $0.3 million to $5.9 million
per year, with a midpoint of $2.8 million. Average annual costs across the case studies
are also $2.8 million.
-------
-------
CHAPTER 1
INTRODUCTION
This Regulatory Impact Analysis (RIA) updates the benefit-cost analysis presented in the RIA
for the proposed Great Lakes Water Quality Guidance (the Guidance) dated April 1993
(RCG/Hagler Bailly, 1993). The Guidance provides direction to the Great Lakes states and
tribes on minimum water quality standards, and contains numeric water quality criteria for
30 pollutants as well as methodologies for the development of water quality criteria for
additional pollutants discharged to these waters. It also provides guidance to Great Lakes
states and tribes on antidegradation policies and implementation procedures.
This RIA was prepared in compliance with Executive Order 12866, which requires federal
agencies to perform an analysis comparing the benefits and costs of the regulation, analyze
alternative approaches to the regulation, and identify the need for the regulation for each
major rule proposed or promulgated. The revisions and updates presented here have been
made primarily in response to comments received during the public comment period and from
the Office of Management and Budget (OMB). This revised RIA is provided as a supplement
to the April 1993 RIA and, as such, does not repeat information presented previously.
This report is organized into seven chapters:
I- The Requirement for the Guidance
»• Alternatives Considered in Developing the Final Guidance
* Revised Analysis of Costs and Cost-Effectiveness
*• Analyses Related to the Attribution of Benefits to the Guidance'
> Updated Risk Assessments for Great Lakes Anglers
»• Revised Case Study Benefit-Cost Analyses.
Chapter 2 ("The Requirement for the Guidance") expands on the information presented in
Chapter 2 of the April 1993 RIA notably with respect to the discussion of the environmental
factors necessitating the development of the Guidance.
Chapter 3 ("Alternatives Considered in Developing the Guidance") presents the options EPA
considered in developing the final Guidance, including options for the most costly
1 Implementation of these provisions is dependent upon future promulgation of provisions consistent
with the final Guidance by state or tribal agencies, or, if necessary, EPA. Until actions are taken to
promulgate and implement these provisions (or equally protective provisions consistent with the final
Guidance), there will be no economic effect of this rule on any entities—government or private.
-------
INTRODUCTION »• 1-2
components of the rule (cost drivers), regulatory requirements, criteria/standards, and
implementation.
Chapter 4 ("Revised Analysis of Costs and Cost-Effectiveness") presents: (1) revised
basinwide estimates of the cost to comply with the final Guidance, (2) revised estimates of
the reduction in baseline pollutant loads resulting from implementation of the final Guidance,
and (3) a revised assessment of the cost-effectiveness of the Guidance based on the revised
cost and pollutant reduction estimates. A discussion of the changes to the final Guidance and
the methodology for estimating costs that impacted the cost analysis is also provided. .
Chapter 5 ("Analyses Related to the Attribution of Benefits to the Guidance") presents
information and analyses related to determining the relative contribution of point sources to
total pollutant loadings in the Great Lakes Basin, including (1) a review of studies and data
on the contribution of point sources to total loadings of the contaminants addressed by the
Guidance (2) a screening analysis of other potential sources of loadings for contaminants
addressed by the Guidance, and (3) modeled reductions in fish tissue contaminant
concentrations, and fish consumption advisories, resulting from point source loading reduction
scenarios. Assumptions regarding the contribution of point sources to total loadings m the
basin are made based on the research presented. In addition, information on the cost-
effectiveness of pursuing point source controls in the basin is provided.
Chapter 6 ("Updated Risk Assessments for Great Lakes Anglers") provides a revision to the
basinwide risk assessment presented in the April 1993 RIA. In accordance with Executive
Order 12898 which directs federal agencies to incorporate environmental justice
considerations into their missions, risk assessments for Great Lakes sport anglers (including
minority and low-income anglers) and Native Americans engaged in subsistence fishing in the
basin are provided.
Chapter 7 ("Revised Case Study Benefit-Cost Analyses") presents the revised case study
benefits estimates, compares the revised benefits and costs estimates, and discusses the
representativeness of the three case study sites. The revised case study benefits estimates
reflect the revised loadings reductions for the Guidance and the assumptions developed in
Chapter 5 regarding the relative contribution of point sources to total loadings in the basin
(and thus the percentage of future toxics-oriented benefits reasonably attributable to the
Guidance).
References are provided in Chapter 8.
-------
CHAPTER 2
THE REQUIREMENT FOR THE GUIDANCE
This chapter presents the statutory requirement and the environmental factors that support the
development of the Guidance. Section 2.1 presents the statutory requirement. Section 2.2
looks at environmental concerns in the Great Lakes Basin.
2.1 THE STATUTORY REQUIREMENT FOR THE GUIDANCE
The Great Lakes Water Quality Guidance was required by the Great Lakes Critical Programs
Act of November, 1990. The Act codified the Great Lakes Water Quality Initiative (GLWQI)
effort, which was a joint endeavor of the Great Lakes states and EPA to address water quality
concerns in the Great Lakes Basin that were associated with toxic water pollutants. The
GLWQI was intended to provide a forum for developing uniform water quality criteria and
implementation procedures for the Great Lakes Basin. This voluntary effort was designed to
develop guidance on minimum requirements for the Great Lakes states' water quality
programs The participants planned to use the results of this effort as a basis for revising state
water quality standards pursuant to Section 303(c) of the Clean Water Act (CWA). Section
303 of the CWA requires the eight Great Lakes states to review, revise, and adopt updated
water quality standards every three years as part of a continuing triennial review process.
As part of the basinwide effort to improve and protect water quality, the Great Lakes Critical
Programs Act (CWA Section 118(c)(2)) required EPA to publish proposed water quality
guidance for the Great Lakes Basin which conforms with the objectives and provisions of the
Great Lakes Water Quality Agreement (under which the United States and Canadian
governments established common water quality objectives for the Great Lakes) and is no less
restrictive than provisions of the Clean Water Act and national water quality criteria and
guidance. The Critical Programs Act required that the guidance specify the following for the
Great Lakes Basin:
»• minimum water quality criteria protective of human health, aquatic life, and
wildlife
> antidegradation policies
»• implementation procedures.
The Guidance fulfills these requirements.
-------
THE REQUIREMENT FOR THE GUIDANCE > 2-2
2.2 AN OVERVIEW OF ENVIRONMENTAL CONCERNS IN THE GREAT LAKES
BASIN
Despite significant efforts to reduce toxic loadings to the Great Lakes over the past two
decades, fish consumption advisories are in effect throughout the region, the effects of toxic
contaminants on fish and wildlife populations continue to be observed, and the potential for
human health effects remains. Once introduced into the Great Lakes—whether by water point
sources, atmospheric deposition, contaminated sediments, groundwater, or surface
runoff—some toxic substances have physical, chemical, or biological properties that allow
them to persist for extended periods in the aquatic environment and to biomagnify through the
food chain. While the concentrations of these chemicals in water may be so low as to be
undetectable by available analytical techniques, persistence and bioaccumulation can increase
the levels of these contaminants to toxic concentrations in sediment, fish and wildlife, and
other receptors.
This section describes the levels and effects of bioaccumulative chemicals of concern (BCCs)
and other contaminants in the Great Lakes. Contamination effects in the Great Lakes Basin
became obvious in the 1960s and 1970s, when high levels of contaminants such as PCBs,
DDT, mercury, and mirex were detected in fish. While levels of these contaminants have
decreased since the early 1970s, high contaminant levels in fish still restrict the commercial
market and advisable consumption levels for a number of species (Environment Canada,
1991). In fact, trend data at some sites indicate no reduction in total PCB and total DDT
concentrations in spottail shiners, an indicator species for near shore areas, since the 1970s
(Rang, 1994). Additionally, consumption restrictions continue for important sport fish species
(e.g., lake trout, rainbow trout, coho and chinook salmon, walleye) in all of the Great Lakes
(Rang, 1994). These findings demonstrate the persistence of a number of BCCs found in the
Great Lakes environment, even at low levels. Persistence, toxicity and, ultimately, ecological
and health effects of BCCs and other contaminants are discussed below.
2.2.1 Adverse Effects Caused by BCCs and Other Contaminants
Toxic pollutants, including metals and synthetic organic chemicals, can be acutely poisonous
in relatively small amounts and can be injurious, through chronic exposure, at very low
concentrations. Many contaminants present in the Great Lakes Basin have the potential to
increase the risk of cancer, birth defects, genetic mutations, and reproductive impacts through
long-term exposure. As described in Section 2.2.2, adverse impacts on fish, bird, and mammal
populations in the Great Lakes caused by the effects of toxic chemicals include cancer,
premature death, eggshell thinning, population declines, reduced hatching success, abnormal
behavior (such as abandonment of nests), infertility, birth defects (such as crossed beaks and
club feet) and illnesses such as chick edema. They also include more subtle effects, including
abnormalities in the thyroid, liver, and endocrine systems.
-------
THE REQUIREMENT FOR THE'GUIDANCE > 2-3
BCCs are toxic compounds that become more concentrated as they are accumulated by
organisms at higher levels in the food chain. Chemicals such as dioxin, PCBs, and
organochlorine pesticides biomagnify from the lowest trophic levels (phytoplankton) to top
predators (fish-eating birds and humans) (Thomann et al., 1992; Gilbertson et al., 1991).
Because they are lipophilic (partition into fat) and are slowly metabolized, they are retained in
organisms once absorbed. Biomagnification of a variety of BCCs has been observed in Great
Lakes food webs (Environment Canada, 1991). Contaminants are also absorbed from the
sediments by bottom-dwelling animals such as insect larvae and molluscs, and then can
accumulate in terrestrial and aquatic organisms (Thomann et al., 1992; Gilbertson et al.,
1991). Additionally, aquatic organisms may absorb toxic chemicals directly from water
flowing across their gills during respiration (Barron, 1990).
Because of the tendency of certain toxic chemicals to bioaccumulate in fish and wildlife and
to biomagnify through the food chain, even low concentrations of these compounds in the
water column are cause for concern (Gilbertson et al., 1991). The effects of biomagnification
can be significant; while PCBs in open water in the Great Lakes are in the low parts per
trillion range, those in the eggs of eagles at the top of the Lake Erie food web have been
measured at greater than 25 parts per million as recently as 1988 (Environment Canada,
1991). This corresponds to a biomagnification factor of 25 million (Environment
Canada, 1991).
i
I
Once contaminants enter the system and are absorbed into the food chain, they can have
severe adverse effects on fish, wildlife, and humans. These effects include reproductive,
developmental, metabolic, hormonal, physiological, neurological, behavioral, and growth
effects (Giesy et al., 1994). Characteristics and toxicity of compounds of concern for the
Guidance are summarized in Appendix A.
Rationale for Establishing New Water Quality Criteria and Implementation Procedures
Two principal factors underlie the rationale for implementing the Guidance in order to
provide greater protection of natural systems and human populations from Great Lakes
contaminants. First, as noted in the Great Lakes Water Quality Agreement and supported by
the International Joint Commission (IJC), the persistence of toxic compounds suggests that a
virtual elimination strategy will be required to adequately control them. Second, evidence
suggests that existing water quality criteria are not fully protective of a number of wildlife
species (Ludwig et al., 1994; Environment Canada, 1991).
Persistent toxics require a virtual elimination strategy. In August 1993, a Task Force of the
IJC published a strategy for virtually eliminating the input of persistent toxic substances into
the Great Lakes Basin (IJC, 1993). This strategy was developed to support the amended Great
Lakes Water Quality Agreement, which contains such a requirement. In order to protect
human health, protect the health and productivity of aquatic resources, and ensure ecosystem
protection, the IJC believes that it is necessary to virtually eliminate present inputs of
-------
THE REQUIREMENT FOR THE GUIDANCE > 2-4
persistent toxic substances and to prevent future inputs (IJC, 1993). According to the IJC
statement, it is the persistence of these toxic substances in the environment, rather than simply
their toxicity, that provides a compelling case for their virtual elimination.
Existing water quality criteria have not fully protected all wildlife species. Due to several
factorsf existing water quality criteria are not fully protective of a number of wildlife species.
Most of the Great Lakes states do not have wildlife-based water quality criteria. In addition,
in the past wildlife-based criteria have been based primarily on acute toxic effects on aquatic
organisms (Ludwig et al., 1994). These calculations typically have not included higher
troDhic-level wildlife, such as mink, bald eagles, and colonial water birds (Ludwig et al.,
1994) Also current risk assessment methods do not account for wildlife exposed to complex
mixtures of hazardous compounds (Ludwig et al., 1994). Finally, because wildlife do not have
the option of limiting intake of contaminated species, certain wildlife species may be at
greater risk than human populations (Ludwig et al., 1994).
In a recent study, Ludwig et al. (1994) developed recommended water quality criteria that
would be protective of six sensitive Great Lakes species. Based on 1986 data Ludwig et al.
0994) found that the concentrations of PCBs in the water of all five of the Great Lakes
exceeded the recommended values for all six species studied. Ludwig et al. (1994) argued that
a general wildlife criteria of 0.1 Pg/l (ppq) for total concentrations of PCBs in waters of the
Great Lakes would be necessary to protect the most sensitive species.
Observed adverse effects on fish-eating birds indicate that contaminant levels continue to
threaten some species (Giesy et al., 1994). The presence of previous high levels of substances
such as dieldrin was likely responsible for preventing the restoration of some indigenous
species such as the double-crested cormorant to Saginaw Bay (Gilbertson et al., 1991).
Finally, there is evidence of increasing contaminants in some lakes over the past three years,
such as an increase in dioxin levels in Lake Ontario lake trout (Rang, 1994).
In summary, evidence suggests that the most sensitive species can tolerate only extremely low
doses of BCCs before demonstrating adverse effects. Additional evidence has demonstrated
continued effects on fish and wildlife species from Great Lakes BCCs despite the significant
reductions in contaminant loadings over the past two decades.
2.2.2 The Effects of Contaminants in the Great Lakes
The previous section summarized evidence linking the contaminants of concern for the
Guidance with adverse human health and environmental effects. This section presents
information about the presence of these contaminants in the Great Lakes, and observed
adverse effects on Great Lakes fish and wildlife that are linked to these contaminants.
-------
THE REQUIREMENT FOR THE GUIDANCE > 2-5
Sensitivity of the Great Lakes Basin
According to the 7992 National Water Quality Inventory Report to Congress (U.S. EPA,
1994a), the seven Great Lakes states assessed water quality at approximately 99% of the total
U.S. Great Lakes shoreline in 1990-1991. These assessments found that of a total of
5,319 miles, only 2% of shore miles fully support fishing, swimming, and other uses, and an
additional 1% currently support uses which are threatened. Uses in the remaining 96% of
shore miles are fully or partially impaired (U.S. EPA, 1994a). While water quality impairment
reflects a range of environmental problems, issues related to toxic contaminants are an
important factor (Environment Canada, 1991). Several characteristics of the Great Lakes
create a particular susceptibility in the system to relatively nondegradable, lipophilic
chemicals including: (1) long hydraulic retention times and (2) the presence offish and
wildlife populations confined to and solely dependent on the Great Lakes Basin for their food
and water supply (40 CFR Part 132, April 16, 1993).
In spite of their large size and volume of fresh water, the Great Lakes are extremely sensitive
to the impacts of a wide range of pollutants. Because outflows from the Great Lakes are
relatively small (less than 1% per year) in comparison with the total volume of water,
pollutants that enter the Lakes are not readily flushed from the Great Lakes Basin as in a
riverine system. Recycling within the system, such as nutrient cycling, adds to the overall
retention time of chemicals in the Great Lakes. It is because of these long retention times that
bioaccumulating substances become concentrated in organisms at levels which greatly exceed
the ambient concentrations in the open waters of the Great Lakes (40 CFR Part 132, April 16,
1993).
In addition to the sensitivity of the system due to its physical and hydrologic properties,
characteristics of certain species residing in the basin make them particularly vulnerable to the
effects of pollution. Several fish-eating bird species are at greater risk from exposure to
pollutants in the Great Lakes than in other aquatic systems because their foraging range is
entirely within the Great Lakes Basin for all or part of each year (Ludwig et al., 1994).
Species offish-eating birds known to be affected by pollutants in the Great Lakes include the
double-crested cormorant, black-crowned night heron, osprey, herring gull, common tern,
Forster's tern, and bald eagle (Ludwig et al., 1994). Researchers believe that some of these
species, such as herring gulls and double-crested cormorants, however, are relatively tolerant
to certain BCCs. This may explain why the populations of herring gulls and double-crested
cormorants are increasing in most areas of the Great Lakes after being decimated by toxic
chemicals including DDE and dieldrin (Fox and Gilbertson, 1991). PCB concentrations seem
to be approaching thresholds for subtle effects in these more tolerant species rather than
causing gross mortalities (Ludwig et al., 1994). According to one study, it is likely that the
competitive balance among Great Lakes predators 'has been "tipped by contaminants to favor
these more tolerant species over more sensitive competitors" (Ludwig et al., 1994).
-------
THE REQUIREMENT FOR THE GUIDANCE » 2-6
As a result of loadings of bioaccumulative chemicals, the Great Lakes states have issued
164 fish consumption advisories for the waters of the Great Lakes Basin. Pollutants for
which fish advisories currently exist include 8 of the 28 bioaccumulative chemicals of
concern identified in the proposed Guidance. Although concentrations of certain pollutants
with system-wide impacts have significantly declined in recent years, the rate of decline may
be slowing Fish tissue concentrations of some pollutants appear to be approaching
equilibrium at concentrations well above levels of concern as defined by water quality criteria
calculations (U.S. EPA, 1993 as cited in Ludwig et al, 1994).
Fish consumption advisories in all states recommend incremental reductions in fish
consumption based on several characteristics: sensitivity of the individuals consuming fish,
type offish consumed, and specific location of the fish. These advisories range from simple
advice to remove fat from fish to advisories to avoid consumption of certain species
altogether. For example, the State of Wisconsin's advisory for PCBs and pesticides divides
fish into three groups (Wisconsin Dept. of Natural Resources, 1994):
* fish posing the lowest risk, which should simply have fat and skin removed
before consumption
»• fish that should not be eaten by women or children
>• fish that should not be eaten by anyone.
Similaiiy the State of Wisconsin includes four incremental advisories on consumption of
merrury-contaminated fish (Wisconsin Dept. of Natural Resources, 1994). Fish in specific
water bodies are placed into four categories, with the following associated advisories:
»• Group 1: pregnant women should eat no more than one meal a month
v Group 2: pregnant or breastfeeding women, women who plan to have children,
and children under 15 should not eat these fish; others should eat no more than
26 meals per year
> Group 3: pregnant or breastfeeding women, women who plan to have children,
and children under 15 should not eat these fish; others should eat no more than
13 meals per year
*• Group 4: no one should eat these fish.
1 1994 fish consumption advisories from Illinois, Michigan, Minnesota, New York, Ohio, and
Wisconsin were reviewed.
-------
THE REQUIREMENT FOR THE GUIDANCE *• 2-7
Given the incremental nature of Great Lakes fish consumption advisories, even if the
Guidance does not enable certain advisories to be lifted entirely, water quality improvements
may result in incremental improvements in advisory levels. Thus, increased fish consumption,
such as an increase in the recommended number of fish meals per month for at-risk
subpopulations, may result from Guidance implementation well before fish consumption
advisories could be rescinded entirely.
Areas of Concern
It is important to review information on water quality in embayments, the lower portions of
tributaries, and near-shore areas, because these areas support more human uses and shore
aquatic and wildlife species than open water areas. Several such Areas of Concern (AOCs)
experiencing problems from contaminants that are in part contributed by point source
discharges include:
> Clinton River, Michigan. Problems include contamination by heavy metals, PCBs,
and pesticides, and low levels of dissolved oxygen. Suspected sources are municipal
discharges (7 plants), industrial sources (22 dischargers), nonpoint urban and
agricultural runoff, and contaminated sediments and groundwater. Fish and river
bottom aquatic communities have been affected by contaminants and low dissolved
oxygen (Center for the Great Lakes, 1989a).
»• Detroit River, International. Problems include contamination by heavy metals
(mercury, cadmium, chromium, copper, and zinc), and organic chemicals, and bacterial
contamination. Suspected sources are industrial discharges (primarily U.S.),
petrochemical industries, oil and chemical spills, sewage and combined sewer
overflows, industrial and municipal wastewater discharges, leachates from waste
disposal sites, runoff and leachates from industrial storage sites, and nonpoint source
runoff. Swimming, other water contact sports, fishing, and duck hunting have all been
adversely affected, as have fish, wildlife, and aquatic communities (Center for the
Great Lakes, 1989b).
» Manistique River, Michigan. Problems include contamination by PCBs and heavy
metals (zinc, lead, and cadmium). Suspected sources are contaminated sediments,
waste site runoff, a paper-products manufacturer, and a wastewater treatment plant. A
fish consumption advisory has been issued for carp from the river. Other aquatic life
has also been adversely affected (Center for the Great Lakes, 1989c).
»• Muskegon Lake, Michigan. Problems include contamination by PCBs. Suspected
sources are past discharges, groundwater contamination from landfills and industrial
activity, industrial and municipal discharges, atmospheric deposition, and storm-sewer
drains. Walleye and largemouth bass in this water body have slightly elevated mercury
-------
THE REQUIREMENT FOR THE GUIDANCE > 2-8
levels, and carp are contaminated with PCBs and chlordane (Center for the Great
Lakes' 1990).
Rouge River, Michigan. Problems include contamination by heavy metals (cadmium,
mercury, and lead), bacteria, organic chemicals (industrial solvents, cyanide oil, and
grease) PCBs, and PAHs. Suspected sources are municipal (4 plants) and industrial
(30 plants) dischargers, combined sewer overflows, separate sewer overflows, waste
site runoff, contaminated sediments, and other nonpoint sources. Impaired uses include
water-contact recreation, consumption offish, use of river water as an industrial and
agricultural water supply, and aesthetics (Center for the Great Lakes, 1989d).
> St. Clair River, International. Problems include contamination by conventional
contaminants heavy metals, and organic chemicals (mainly chlorinated organics and
volatile hydrocarbons). Suspected sources include 13 industrial dischargers, waste site
runoff municipal dischargers, and nonpoint sources. Swimming and sport fishing are
impaired or restricted in the area. Fish in the area have been found to contain elevated
levels of hexachlorobenzene, octachlorostyrene, polynuclear aromatic hydrocarbons,
PCBs, mirex, and pesticides (Center for the Great Lakes, 1989e).
> White Lake, Michigan. Problems include contamination by PCBs, heavy metals,
phosphorus, and the discharge of contaminated groundwater to the lake from a
chemical company. Suspected sources include wastewater discharges, industrial
discharges, urban runoff, combined sewer overflows, and contaminated sediment. Fish
consumption and drinking water have been impacted (Center for the Great
Lakes, 1989f).
The Great Lakes contain 39 designated AOCs. As demonstrated by the above descriptions of
selected AOCs, significant effects on aquatic communities and human uses of these waters
occur at these sites. Thus, despite low ambient concentrations of contaminants in Great Lakes
open waters, numerous near-shore areas continue to exhibit contaminant levels sufficient to
impair uses.
Fish, Birds, and Other Wildlife
Biodiversity of Great Lakes vegetation, fish, and wildlife is in part affected by.the presence of
toxic chemicals (The Nature Conservancy, 1994). Toxics can cause mortality or chronic
impairments of species, such as diminished reproductive success or inhibition of growth.
According to a recent Nature Conservancy study (1994), the installation of pollution controls
has greatly reduced inputs of these chemicals. Nevertheless, the study observes that top
predators in the aquatic system remain threatened by bioaccumulative substances previously
released into the environment.
-------
THE REQUIREMENT FOR THE GUIDANCE » 2-9
There is published scientific evidence that 14 wildlife species, all top predators, have
exhibited population declines, reproductive effects, and physiological problems related to
persistent toxic substances since the 1960s (Environment Canada, 1991). These species
include mink, otter, double-crested cormorant, black-crowned night-heron, bald eagle, herring
gull, ring-billed gull, Caspian tern, common tern, forester's tern, snapping turtle, lake trout,
brown bullhead, and white sucker. Effects include population decrease, effects on
reproduction, eggshell thinning, congenital malformations, behavioral changes, biochemical
changes, mortality, and alterations in recruitment (Environment Canada, 1991). Except for the
bald eagle and the common and Forster's tern, bird species have experienced significant-
recoveries in their reproductive success (Environment Canada, 1991).
Fish. Several contaminants of concern in the Great Lakes could potentially accumulate in fish
tissue and affect the health of humans consuming those fish. Data from the Great Lakes
National Program Office (GLNPO) (Giattina, 1993) include fish tissue concentrations for
chlordane, dioxin, and mercury. The IJC (1993) also collects data on concentrations of
mercury in fish tissue. Additionally, fish tissue data in the U.S. EPA National Study of
Chemical Residues in Fish (U.S. EPA, 1992a) provides information on concentrations of
chlordane, mercury, and dioxin in fish tissue for Lakes Ontario, Erie, Michigan, and Superior,
and for the Cuyahoga and Fox Rivers. Table 2-1 provides the fish tissue concentrations from
these reports.
Concentrations provided in Table 2-1 are for sport fish only; no bottom feeders (e.g., carp,
catfish) were included. Bottom feeders may bioaccumulate higher contaminant concentrations
than sport fish. Fish samples for the 1991 U.S. EPA study were collected primarily in 1987.
The majority of the GLNPO fish samples (Giattina, 1993) were collected in 1988. These data
indicate that concentrations of contaminants in fish tissue are highest in Lake Michigan.
Birds and Other Wildlife. Several species of birds, including the bald eagle, osprey, and most
colonial nesting birds, as well as aquatic mammals and turtles, rely on fish as a food source.
Through the processes of bioaccumulation and biomagnification, these top predators have very
high levels of organochlorine chemicals in their tissues (Giesy et al., 1994; Ludwig
et al., 1994).
i
Although concentrations of PCBs, polychlorinated dibenzo-p-dioxins (PCDD), and
polychlorinated dibenzofurans (PCDF) have decreased in Great Lakes; piscivorous birds since
the 1970s, populations of several species continue to decline (Giesy et al., 1994). These
include the common tern and Forster's tern. Other bird populations, such as double-crested
cormorants and herring gulls, have made significant recoveries. Concentrations of dioxin-like
contaminants in several species still appear to be greater than the threshold for discernable,
population-level effects at several locations around the Great Lakes (Giesy et al., 1994). For
example, double-crested cormorants and Caspian terns in Saginaw Bay and Green Bay display
excess rates of developmental deformities and embryo lethality (Giesy et al., 1994).
-------
THE REQUIREMENT FOR THE GUIDANCE » 2-10
Table 2-1
Fish Tissue Concentrations (ppm) from Various Sources
Source
IJC (1993)
Giattina (1993)
Giattina (1993)
Giattina (1993)
Giattina (1993)
Giattina (1993)
U.S. EPA
(1992a)
U.S. EPA
(1992a)
U.S. EPA
(1992a)
Location
Great Lakes
Lake Erie
Lake Huron
Lake Michigan
Lake Ontario
Lake Superior
Lake Ontario
Lake Michigan
Lake Erie
PCBs
1.32
1.35
1.50
2.72
2.18
0.45
Not
provided
Not
provided
Not
provided
— -
Dieldrin
0.09
0.03
0.08
0.18
0.08
0.04
Not
provided
Not
provided
Not
provided
Chlordane
Not provided
0.03
0.21
0.45
0.27
0.10
0.024
0.206
0.002
Mercury
0.17
0.10
0.09
Not
available
0.12
0.13
0.32
0.22
0.20
=====
Dioxins
Not
provided
Not
detected
0.00001975
0.00000283
0.0000221
Not
detected
0.00001289
0.00000374
Not
detected
=====
In a recently published study (Giesy et al., 1994), exposures to chlorinated chemicals were
still found to be causing lethality and deformities in embryos of all of the populations of
colonial, fish-eating water birds (e.g. double-crested cormorant) examined. The researchers
found that the effects observed in the Great Lakes populations were greater than those
observed in populations outside of the Great Lakes (Giesy et al., 1994). However, the effects
at the individual-level in the Great Lakes populations were only translated into biologically
significant population-level effects at the most contaminated sites (Giesy et al., 1994).
Some scientists warn that trace amounts of organic chemicals which act as synthetic estrogens
and endocrine disrupters can interfere with the action of reproductive hormones and damage
the fertility of exposed animal populations and humans (Stevens, 1994). Evidence to support
this conclusion includes studies conducted in Taiwan, Florida, and the Great Lakes region.
Concentrations of chemicals such as polychlorinated dioxin and furans, and PCBs have
declined in the Great Lakes Basin since the late 1970s, but the rate of decline appears to be
slowing; these contaminants still pose a threat to the reproductive success of several types of
waterfowl and other species (Stevens, 1994).
-------
THE REQUIREMENT FOR THE GUIDANCE » 2-11
Human Health Effects
A significant number of studies have been conducted which have demonstrated the effects of
toxic chemicals on individual species of fish, birds, and wildlife. Fewer authoritative studies
exist examining human health effects (Environment Canada, 1991). A 1985 report by the
Royal Society of Canada and the U.S. National Research Council found "substantial evidence
that the human population living in the Great Lakes Basin is exposed to and accumulates
appreciably more toxic chemical burden than people in other large regions of North America
for which data are available" (Flint and Vena, 1991). One study in Michigan (Humphrey,
1988, as cited in Fiore et al., 1989) demonstrated that sport anglers who ate Great Lakes fish
had higher blood and tissue levels of PCBs than individuals who seldom or never ate such
fish. Additionally, there is suggestive evidence from a study of women who ate several meals
of Lake Michigan fish per month for at least six years preceding their pregnancies that their
children had lower birth weights, shorter gestation periods, and smaller head circumferences
at birth when compared to children of women who had not consumed Lake Michigan fish
(International Joint Commission, 1993; Jacobson, 1990 and Jacobson, 1991 both as cited in
International Joint Commission, 1993). These children also showed discernible cognitive,
motor, and behavioral deficits when tested at seven months and four years.
It is estimated that fish consumption provides a significantly greater potential dose of PCBs
for the general population than from water and atmospheric exposures (Fiore et al., 1989).
Sport fishermen are likely to consume quantities of fish in excess of the national average and
are more likely to consume fish with contaminant levels above those found in commercial
fish species. A study of Wisconsin anglers (Fiore et al., 1989) found a statistically significant
positive correlation between total PCBs residues and the total number of sport-caught fish
meals in 1985. PCB body burdens of anglers in this study appear to be above those of the
general populace. The researchers found that the mean PCB level for the 192 study subjects
was 2.2 ]ig/i. More than 20% of the subjects of the Wisconsin study had PCB sera residues
above 3.0 |ig/l, the level defined by a Finnish study as the practical reference upper 99%
confidence limit for the PCB level in the general population of Finland2 (Fiore et al., 1989).
'I
i
2.3 CONCLUSION
i
Research conducted in the Great Lakes Basin demonstrates the negative health effects of toxic
chemicals, especially BCCs, on humans and on species of fish, birds, and wildlife. Many fish
consumption advisories are in effect throughout the region, despite efforts to reduce toxic
loadings to the Great Lakes. Existing water quality criteria have not fully protected all
wildlife species; there is published scientific evidence that several top predators in the region
2 The PCB level for the general population of Finland may be higher than that for the general
population of Wisconsin since fish is a larger portion of the Finnish diet (Fiore et al., 1989).
-------
THE REQUIREMENT FOR THE OUTDANCE »• 2-12
have exhibited populations declines and other negative effects. In addition, multiple AOCs
throughout the region pose continuing risk to human health.
Some sites have shown no reduction in the occurrence of toxic chemicals. Further, the Great
Lakes Basin is especially sensitive to toxics contamination, and contaminants can
bioaccumulate to high levels over long periods of time. The Great Lakes Critical Programs
Act addressed concerns over toxics-related problems with a requirement that EPA develop
and publish the Guidance.
-------
CHAPTER 3
ALTERNATIVES CONSIDERED IN DEVELOPING
THE FINAL GUIDANCE
Comments on the proposed Guidance raised several issues that warranted further
consideration prior to finalizing the Guidance. In addition, there were also some issues that
remained unresolved at the time the proposed Guidance was formulated (e.g., whether to
assume additive cancer risk from carcinogen mixtures). These issues were divided into four
categories:
+ Cost Drivers. Issues labeled "cost drivers"' due to their relatively large
potential impacts on the estimated costs of the Guidance (e.g., the treatment of
intake credits).
> Regulatory Requirements. Issues related to regulatory requirements include
adoption requirements and schedules for Great Lakes states.
+ Criteria and Standards. Issues related to the development of criteria and
standards, such as whether to base criteria for metals on dissolved
concentrations.
> Implementation. Issues related to the implementation of the Guidance,
including provisions for the development of total maximum daily loads
(TMDLs).
The evaluation of alternative options for the final Guidance was based on evaluation of the
options for the issues with significant impacts on costs ("cost drivers"). Numerous analyses
were conducted to evaluate the impact of the cost drivers on the total; compliance costs of the
Guidance. The major cost driver issues and analyses conducted are described below. These
analyses make comparisons to the lower end of the estimated range.1 A description of the
issues and the selected options for the final Guidance for all of the issues considered are
shown in Table 3-1.
1 There was little relative difference observed between cost driver impacts on the low and high
estimated costs.
-------
en
A
Q
S
P
c.
9
CO
^J
5
o2
ci:
=============
II
II
11 &)
11 e
IS C3
II ""
II '^
III ^J
11
|| p*
13
E
CU
II ^*
•*^
Table 3-1
msidered in Developing
i °
U
v>
S
l o
~
=
Cost Drivers
5
"5 .1
|l
3
e
c
O
c,
^
~
"•5
.2*
u
3 *
Q
3
en
en
=====
12
3
3
3
a
w"
u
o
u
o
T3
U
•w
J2
u
en
1
J3
"o
a.
o
~~* V.
0 0
~ C
"S S
0 £
1%.2
ill
2 a c.
:§•§ S
O CO 4)
& > • — • 'o
2 §
s -a
11.
~-
=====
b
0
u
u
q
~a
^j
•"-*
la
"x
o aj
G ha
**W 3 23
~ -° '>
J.2 *
C SeS
g'BI
u ._ .i
&••£ 3
j> •« S"
3 « c
eo JS —
4> e O
1 1 C
c2 S =:
§ -S o
-fi o P
en (2
IS -0 "e
Tier 11 methodology be used lo esla
Use of Tier II methodology require
parameters, including aquatic life ai
12
"3
.§
on A
—
VI
P
BO
.s
>
*c
u
«*N
SO
o
"3
•a
o
•s
u
•o
er~ ™
o a
a •§
rs "g
11
*° -s
|-cj
0 eo
*t- O
5 •=
•* *~ u
M "O O
«M QJ C
•== .-=: <3
u g "0
•5 ••='!>
sll
" o «
u ° -a
III
1*1 '»
o > (Jj
S" 2
u u >
•5 P -
en
*C3
^ A
===================
U
•"" "a °
•a 3 •«
1 ||l
"? 1 •" **>
i 2^- §
"o" "" ® <*J*
•M (Q ^^ O
_v* ^5 S * "
"en C 0 "3
•C o x >
~ o • —
D. U
^ P? M
Hi
T3 en -g
u '-3 o
•° -a «
we
j a ts
u ~ «
CQ S en
eid
!> - GQ
2^0
1.1^
« o
> j;
£.i A
Wl
H 2 ?
.52 •£ u-
«^ en CJ
CQ 4^
'S e u
0, S >>
0 §,<=>
^ "s "^
.2 o **•"
U CJ •— *
U we
•n 0-3
"S -si
S § o
X e —
« g -
background is calculated to
same body of water?
for simple pass allowed wto
set for facilities discharging
lished TMDL or at criteria.
hould WQBELs be developed when
intake water, and discharge is to the
Finding of no reasonable potential
waler and "no nel increase" limils
WQBELs set consistenl wilh estab
w u
Z •£
•S.s A
•o
-------
EX}
CJ
Q
O
M
O
cu
O
a
ta
Q
CJ
00
<
E
CO
CJ
CQ
."2
S
o
s
fa
Table 3-1 (cont.)
Issues Considered in Developing the
Cost Drivers
en
g!
"3
1- S
a *
3
S
O
c.
•s
B JJ«
!_
CJ
en A
CJ *
O
""
CO
1
u
1
SO
C co
jl
C3 O
*° ¥
|1
Is
.s §
*" -o
<2, o
1 |
•^ o
u -0
-° o
'o oo
fi
If
i^
~
^~
_o
e^. "3
£5 -£>
immendatic
QBELs are
the Pollution Minimization Plans be a regulatory requirement or recc
Pollution Minimization Plans required as permit condition where W
quanlification level.
-a
"3
o
£75 A
ON
tN
O
«
rs
'3
00
CS
t .
—
0
s
^ w
c »
1 §
3 '•?
ortions of Ihe anlidegradation proposal should be regulation versus g
Antidegradalion framework in established in regulation; implementa
Q.
'cs
^ A
O
„
1
CS
"cS '^
•— e
^~^ o ti
co ^ 3
S o "o
CS *u fi
? § §
1 " *
04 s S-
J3 es O
iould Tier 11 of antidegradation be implemented (i.e., definition of "h
Defined on a pollutant-by-pollulant basis; stales allowed to exclude
recreational, or aeslhelic significance from antidegradation review (
"co
;>
0
a: A
f^
r~*
nl
"O
1
^
"«
U
XI
liould trigger antidegradalion review for non-BCCst
New and increasing limits trigger review. De minimis increases wil
CO
CS
^
> A
r-i
""
•o
II
•g o
o -—
CO '5
£• c
a. u
CJ p,
,. *-
CJ X
IB 2
rz o
« M
||
-a ."S
liould trigger antidegradalion review for BCCs't
Baseline loading level is established as a trigger, which is calculate
monitoring data. The baseline loading level becomes Ihe permit lim
CO
cs
^
^ A
r-i
~~
-------
TT
A
a
1
5
|
Cd
O
Q
z
Q
33
8
1
i
1
i
1
able 3-1 (cont.)
n Developing the Final Guidance
H —
1
o>
T3
°tn
I
C/3
O
V)
6O
M4
latory Requirements
I
en
S
e
'S -9
0 §
*•• "s
o .-
en
u e
u e
3
z
s
o
0
u
u
"71
o M
'•S
.£•
2
tf) A
SJ *
a
4J
tn
Crt
VO
u
"3
o 'S
< S
OT
1 1
S u,
provision of the Great Lakes Critical P
iust adopt provisions that are equal to o
as technical or policy guidance.
•« 21
•5 o ra
£ J3 C
- "C .2P
ould EPA interpret the "consisten
vs. Mandatory Provisions)?
i portions identified where states/I
e final Guidance. Other parts des
~" U -5 *£
J— C3 O ^
"C "3
•a a
o u
ac d. A ,
2
n
0
01
ea
£
n
1
/*« Q)
o ^
5 A
rovisions consistent with the final Guidi
18 months, but extensions provided on i
o.
f i
"S *
2 "o
o '5 'of
? c 1
.« o *•
3 *« \O
O* crt [
^•il
Cu en .Ti
si™
2 0 ™
« •— o
JJ o a
3 Ti vE
1 0 ^
•* — • —
^ -2 rt
'3
J5
£
s
O A
W>
m
o
•«•»
•1
•Ss
en <:
o CO
toxicity?
burden of proof placed on discharger t
.0 those chemicals with field-measured
•a TS -a
s-sl
8-s.S
c S —
O Wl
•s ^H
S|^
£2 =
•a T3 — ;
3 C <
•3 « «•
SOS
•- o u
cj S .-3
o .3 e
ffl 2 S
^S.o
§^ s
'•3 e •—
;s -S i
13 .tS a
O £ 3
y JJ "C
j— Cj c
2
3
O
53 A
vd
1
m
.2 T3
Ui <1>
^ 'S '>
- oo
™ d^
C O ^^
oj S °
'5 2 |
a a §>
> ^S
u o "*
2> -o
'•3 S
i— o ea
c 'is
1 H*
! is
! 1
U- TO U-
1 |f
S «J CD
O oo
jz '5-S
**^ •"* cri
.s §3
"O — S
*Q O **
3 ll
o -a >
•Q O *^
2 "3 —
s «r e
O
n^ c3
^ « A
O-'
-------
A
U
Q
S
w
a
o
a,
O
«
td
Q
z
O
oo
ca
Final Guidance
tMfl
O
^ -4-
§S'
o e
u B
^^ C.
— o
^1
|^
E^.S
•o
eu
cu
^
"en
O
U
£
3
en
en
M
undards
!/3
1
_2
'u
u
«•*
i
en
0
?!
& *
z
e
"c
O
u
~
g V.
i^
_c.
LH
8 *
Q
3
ws
"•
CN
cn
rs?
ted bioaccumulation factoi
o a
o -3
<2 u
ll
tors or biocentra'
using measured
« T3
<2 g
al Guidance use bioaccumulation
n health and wildlife criteria deri
C K
=
J3 DS
•^
^
o
CO A
oo
•<*•
2
cn
£
rovision requiring states/tr:
Endangered Species Act.
Q. o
cs •£
"2 •«
•3 §
*^ **
cn cn
.22 .—
3 '3
Q. 0,
cn cn
= L.
1 5
I |
.i §•
1M !
"on 4-* S
111
ill
"cB i3 c
c-^ i
o «.!:
r* 3 rf
™ CT*
13
"3 S
0 5
C^l .2 A
OS
•' •
Tf
cn
S
S
^_»
ble or dissolved concentra
recoveral
latic life as total
entrations.
l«
cs o
al rule express metals criteria for
s criteria expressed as dissolved c
«!
.a 2
"T3
"3
o
CO A
o
CN
^
0-
1
u
u
1
eo
C
T3
IM
>,
inal methodolog
ly was retained.
«"2
'"* en
"O S
s? =
3 an
2 t"
O JS
J5 U
en cs
0 0
s i
§•£
11
S 8
S Q.
,£
— H
cn ^
i— •
V««l
CS
% A
^.^
CN
m
t*
2
•Q
S
3
e
i!
5|
rS °
13 —
Hi 1*
1!
g jo
— -a
.§ 2
es for deriving 1
were insufficieni
1|
CS ^
•a u
CN ^
"•C "2
!^ "
0. « '"
0 u u
e "i "i
3 *™^ ^^
00 u i-
u.H.Si
JSHH
"^
"3
o
CO A
CN
CN
VI
C J
U i>
S &
a. -a
— e
13 "
5!
w
s? OT
W «
3 •?
^ 1
cn U
n js
° ^ 3
"^1^2
n o ao
«2 -s e
S es O
cs 3 Q Q}
X „-« 3
O U £ ^
FT] — Q "T3
3 .^3
O t5 5 ^
||||
.0 §• u H
o *~ 2 ~
c_, _>.,co _a
2 >
"3 i-5
0 13
-C -a
CO CS A
m'
CN
rn
^
•a
u
_o
Q
CO
•B
•*- >
[^ cn
I> o
1-1 «
Is
"cs .-a
(U &
cn
cn ' U
U C
C g
CO C
.S'P
1 -g
•£
3 c
CS "^
"^
"3
o
CO A
^^
CN
n
larger variances.
_=
o
cn
"7
_u
.S-
i
o
T3
O
T3
8
o.
0)
0
body variances be allowed?
sal language retained, but guidan
u =
'cs £
S °"
•^
"3
o
CO A
tft
CN
^
g
1
'5
c
'£
co
CS ^
m cs
CJ S*
O ^
CS J3
'C n
CS o
> a,
.^•3
1<§
cr —
u, o
u **
cs g
*^ 'C
1) K?
^>
•^
"3
o
CO A
^o
CN
^,
•o
cs
2"
T3 I
i 5
=3 2
i *
•a
S CS
CS __„
c "**
!i
c «£
o 1=
1 •-
1 1
o. a-
— . cs
| ,2
= "S
« s
ss ^
ecific modifications that allow le
on factors be allowed?
or less stringent modifications al
under specific circumstances.
w 'is 8 £
2 "3 •§ **•
-as
3
2 o
= «
Q ™
c/5 IS A
t>.
CN
-------
VO
A
6
DEVELOPING THE F
RNATIVES CONS
i *
Table 3-1 (cont.)
Issues Considered in Developing the Final Guidanc
1
-•
Implementation
-
in
a
O %
•8 3
U)
^ *
S
3
)n
Selected Option
•^
CM
*2
U
S *
a
u
VI
VI
=====
(S
« 1
•S "^
o g
ra ",
u bO
w e
nk should be retained between the need for TMDL development and determination of
,1?
National approach for determining the need for a TMDL utilized (Section 303(d) hstn
setting process).
._ ra
*•• "5
2 w
> !. A
oo
CN
m
o
I
00
'I
'§
f\
chronic mixing zone requirements for non-BCCs be specified in the final regulation?
10:1 dilution factor for discharges to open waters of the Great Lakes retained; ckonii
specified for tributaries (variations allowed based on mixing zone studies).
2
o
00 A
o\
CN
n
£
"3
0
i
T3
£Q VS
U -
S °o
O QJ
CJ 1
CJ «
CO g
§ S
c so
«§•!
S S
II
so S3
II
2
"a
o
OQ A
0
rf
o»
w
«
Die should fish tissue data play in determining background concentration of ambient w;
Fish tissue data use retained; use of resident or caged fish allowed.
t-<
tS
i-+
{£> A
^_;
Tf
1
3 "ra
« '3
^ «
(M '^
W -03
§ to
s» «?
JS U
lould data points below detection be handled, for purposes of calculating background,
s values above and below detection?
Proposed method for designation of default values retained, but use of properly verifi
techniques also allowed.
•55 .S
> 2
S s
£ 8 A
S
en
irocedures should be established for developing interim minimum levels (MLs)?
Method detection level conversion factor included as guidance; regulatory definition
include concept of ML.
O,
2
^ A
m'
CO
{N
'& II
i
B. 1
£
W
bio-uptake studies for BCCs with WQBELs less than level of quantitation be a requir
on?
Bio-uptake monitoring recommended as guidance.
T3"-3
"5 T3
o e
v5 8 A
Tt
-------
9
O
O
Ci.
q
>
Q
Z
Q
§
O
CO
ca
' ' ' •• ••'..(•
.. " ' i «•
ty
w
cs
o
15
Table 3-1 (cont.)
Issues Considered in Developing the Fin
Implementation
<*>
s
o
11
"o 3
J3 1
z
s
^•o
M4
o
on
Selected
•^3
.£.
u
S *
a
u
3
V9
CO
CO
2
in
0 0
1 t
1 i
1 ^
3 .2
CB ^
S o
£ g
2 o
*"* w
o o
e *
§ |
1 ^
P. -O
CO CO
•S u
1 1
O
Q. C
s .S
~ 5
"^ *
.•i 1
§ "*
a. "o
•* e- 2
t*^ ^N ^3
u S K
"S ^ '•§
111
CO _ ""
"T3 o
"3 «3
o ^
c/3 "3 A
wi
"*
"3
es
a.
o
11
•2 M
2P .£
"u .S
"° —
O C3
JJ |u
.c "S,
.9 ^
IS *
eo o
:l §
'5 "3
^-j
"3
O
v3 A
VO
"*
g
CQ
U ^
u a
_D S
s S
^ «8
u fe
O W
•S 3
g §
J= •"
_. -o
^ 1
o 15
S s
^ 1
i 1
* 1
^ 5
0 3
^3 43
§ o"
S «
CS 3
^ 1
U U
3 S
'S o
T3 .S
U o
•^ .•£
S g
1 -§
1 1
2^ 1
Q. =3
0. cs
m 00
t. S
£ -a
cs a
for consideration as SBOW?
if other criteria are met — particularly if the conti
, groundwater contaminated by man-made source
J2-S |
r£ S £
_ap _O ^
"o "« *
jj g>|
Si *
groundwal
SBOW fi
receiving
-Q
"3
o
E75 A
00
m
e
.a
•S
t>~ iH
C/3 O
O —
gaa —
•5 u
C T3
1 1
Q} U
«S "°
u «
-° X
•^ 2
1 "S
« 1
co O
.
0
a: A
m
CN
'
00
C/l
'S
U
e§
u
i
w
O
1
£
en
Ui
rgers be allowed to have compliance schedules?
llowed for new dischargers; increasing discharge
casing discha
schedules a
S 1 E
,- a o
-;S cs oo
= 11
> O to
s u ^
;o
"3
o
C/3 A
p
-------
ALTERNATIVES CONSIDERED IN DEVELOPING THE FINAL GUIDANCE > 3-8
Cost Drivers
Detection Level. Many comments on the proposed Guidance stated that regular improvements
in analytical detection levels should be expected over time, thereby dramatically impacting the
costs to comply with the Guidance in the future. The potential impact of improvements in
analytical detection levels on compliance cost was evaluated under two scenanos: (1) method
detection levels (MDLs) improve 10-fold over time and (2) MDLs improve 100-fold over
time. The analysis indicated that the cost-effectiveness measures associated with these
changes were in the low-end of the range consistent with effluent guidelines.
Intake Credits. There were many comments on the estimated compliance costs for the
proposed Guidance associated with regulating discharge pollutants that originate in the
discharger's water supply. In order to evaluate the impact of intake credits on the estimated
cost of compliance, compliance cost estimates were developed for several intake credit
scenarios. For discharges to different bodies of water, relaxing (no net increase) or tightening
(no intake credits allowed) the intake credit provision were found to not significantly impact
either compliance costs or pollutant load reductions (less than a 0.5% difference from the
final compliance cost estimate). However, for discharges to the same body of water,
eliminating intake credits was found to increase costs by over 600% from the final
compliance cost estimate. As a result, intake credits are included as part of the final
Guidance.
Tier II Values. The cost estimate for the proposed Guidance was based only on the
compliance costs associated with pollutants for which numeric Tier I criteria were proposed.
The compliance cost estimate for the final Guidance is based on additional estimated Tier I
criteria for human health, aquatic life, and wildlife, and Tier II values for human health and
aquatic life criteria (see Chapter 4). Many of these criteria and values were established for the
purposes of estimating cost's and should not be construed to reflect the criteria and values that
would be derived using the final Guidance. If only Tier I criteria were used, the annual
compliance costs for direct dischargers would decrease by approximately $5.7 million (about
12% of the total compliance cost estimate). Adding Tier II wildlife values would not increase
annual compliance costs.
Wildlife Criteria/Mercury Criteria. A number of commenters stated that the wildlife criteria,
and the mercury criteria in particular, would result in significant compliance costs. For the
final Guidance, the wildlife criteria for mercury was increased from 180 pg/1 to 1000 pg/1.
The cost estimate for the final Guidance includes Tier I wildlife criteria for mercury, 2,3,7,8-
TCDD, PCBs, and DDTs. In addition, wildlife criteria were developed for 23 other pollutants
to assist in estimating costs.2 Including or excluding the additional wildlife criteria does not
2 Many simplifying assumptions underlie these values due to the lack of toxicity data. These values
should not be viewed as reflecting values that would be derived using the wildlife methodology in the final
Guidance.
-------
ALTERNATIVES CONSIDERED IN DEVELOPING THE FINAL GLFIDANCE » 3-9
significantly change the final compliance cost estimate, indicating that factors other than
wildlife criteria drive the costs of the final Guidance. While wildlife.criteria are generally
more stringent than human health criteria, both are below analytical detection levels and,
therefore, result in the same pollutant minimization program requirements and costs.
Elimination of Mixing Zones for BCCs. Numerous commenters suggested that the eventual
elimination of mixing zone provisions for BCCs would impose large costs without
commensurate benefits. While the final Guidance retained the requirement for the elimination
of mixing zones within 10 years, some flexibility was provided in the form of limited BCC
mixing zones for facilities that can show that all prudent and feasible treatment technologies
are being implemented to reduce the BCC discharge to the maximum possible extent. To
evaluate the sensitivity of the costs to the elimination of mixing zones for BCCs, a cost
estimate was developed that allowed the same mixing zone for BCCs as is allowed for non-
BCCs. This analysis showed that the elimination of mixing zones for BCCs adds less than 1%
to the estimated compliance cost and loadings reduction. This occurs because the criteria for
most BCCs are usually well below analytical detection levels even with the dilution afforded
by the mixing zones.
Additivity. Some commenters felt that the additivity provision would increase the compliance
cost of the Guidance. The impact of the additivity provision was evaluated under three
scenarios: (1) assumed that additivity was accounted for by basing individual criteria on a 10'5
risk level; (2) assumed total carcinogenic risk of the mixture would not exceed a 10"5 risk
level; and (3) assumed that additivity was accounted for by basing individual criteria on a 10'6
risk level. The first scenario and second scenarios were found to have fairly insignificant
impacts on the estimated compliance cost and loadings reduction (a decrease of approximately
0.5% and 1.0%, respectively). However, the third scenario was found to increase the annual
compliance cost for direct dischargers by $10.3 million while loadings would only be
decreased by an additional 16,000 toxic lbs-eq/yr.3 Therefore, the additivity provision in the
final Guidance could have a significant impact on compliance costs depending on how states
and tribes implement it. However, the implementation procedures are guidance to allow
flexibility and cost-effective implementation.
Antidegradation. Many commenters argued that the RIA for the proposed Guidance neglected
the costs related to the antidegradation provisions. A separate cost analysis was performed to
estimate the lost opportunity cost related to the implementation of the antidegradation
provision. The analysis was based on using existing effluent quality as the trigger for
antidegradation review. The analysis rests on the general premise that economic growth in the
3 The mismatch between costs and loadings is due to the increase in the estimated number of
facilities pursuing regulatory relief, for which no pollutant loading reductions credit was taken.
-------
ALTERNATIVES CONSIDERED IN DEVELOPING THE FINAL GUIDANCE » 3-10
region would continue at a pace equal to the average for the last eight years (1987 to 1994).
The analysis found that under the worst case (all facilities with BC.Cs in their discharge
request antidegradation review and are denied), 5% of annual growth would be lost
r$43 2 million) Using a more realistic assumption that 10% of the facilities discharging BCCs
request antidegradation review and half are denied, a loss'of approximately $2^2 million was
estimated The lower estimate was considered more realistic as 18 of the 28 BCCs are banned
or severely restricted, or are the by-product of banned or severely restricted BCCs.
4 Economic growth was measured by growth in the total value of shipments for six major categories
of direct dischargers in the basin.
-------
CHAPTER 4
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS
This chapter presents the revised analysis of costs and cost-effectiveness for the final
Guidance. The revisions reflect changes to the proposed Guidance made in response to
comments received on the April 1993 RIA. The revisions also reflect modifications to the
analytical methodology used to estimate costs and cost-effectiveness that were made to more
accurately project the costs to the regulated community and to better account for the pollutant
load reductions.
This chapter is organized as follows. The changes to the final Guidance that had an impact on
the cost analysis are presented in Section 4.1. Modifications to the methodology for
estimating costs and pollutant loading reductions are presented in Section 4.2. The results of
the cost analysis for the final Guidance, including revised costs, pollutant loadings reductions,
and cost-effectiveness estimates, are presented in Section 4.3. Finally, although an economic
impact analysis of the final Guidance was not conducted, a general characterization of the
potential positive economic impacts on the Great Lakes region is provided in Appendix B.
4.1 CHANGES TO THE FINAL GUIDANCE IMPACTING THE COST ANALYSIS'
In general, the modeling methodology for estimating compliance costs and pollutant load
reductions described in the April 1993 RIA was employed for the cost analysis of the final
Guidance. However, several revisions were made to the implementation procedures and
criteria in final Guidance which impacted the cost analysis. These changes, as described
1 Some changes to the final Guidance did not impact costs, including procedures to account for
equivalent toxicity in effluent, a concept discussed in the proposed Guidance. The procedures result in
stricter criteria for dioxin. For the cost analysis, the use of toxicity equivalent factors was considered for
establishing criteria for compounds similar to 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD).
However, for the sample facilities for which 2,3,7,8-TCDD water quality-based effluent levels (WQBELs)
were established, no concentration data existed for the chlorinated dibenzo-p-dioxins (CDDs) and
chlorinated dibenzofurans (CDFs) in an effluent, and thus equivalent concentrations for 2,3,7,8-TCDD-type
compounds were not developed. Thus, there were no impacts of this provision on the cost analysis.
A change was also made to the requirement that facilities comply with an acute whole effluent toxicity
(WET) criterion of 1.0 acute toxic unit (TUa) at the end-of-pipe (i.e., no mixing zone allowed). The final
Guidance sets the criterion at 0.3 TUa, but allows for dilution. This provision had an insignificant impact
on costs.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS •» 4-2
below include the promulgation of criteria for metals in the dissolved form, intake credit
provisions, and provisions for the consideration of the additivity of human carcinogenic
effects when two or more carcinogens are discharged.
EPA has also revised its assessment of compliance costs to reflect modifications made to the
final Guidance to provide increased flexibility for state and tribal implementation. For
example:
Site-Specific Modifications: Great Lakes states and tribes may adopt either more or
less stringent modifications to human health, wildlife, and aquatic life criteria based on
site-specific circumstances. All criteria, however, must be sufficient not to cause
jeopardy to threatened or endangered species listed or proposed to be listed under the
Federal Endangered Species Act.
Intake Credits: Great Lakes states and tribes may- consider the presence of intake water
pollutants in establishing water quality-based effluent limits.
Mixing Zones: Great Lakes states and tribes may authorize mixing zones for existing
discharges of BCCs after a 10-year phase-out period, if the permitting authority
determines, among other things, that the discharger has reduced its discharge of the
BCC for which a mixing zone is sought to the maximum extent possible. Water
conservation efforts that result in overall reductions of BCCs are also allowed even if
they result in higher effluent concentrations.
For purposes of estimating compliance costs, EPA assumed that permitting authorities would
use the more stringent provisions specified in the Guidance even when the Guidance provides
for less stringent alternatives. This was done to reduce uncertainty in the cost analysis and to
provide a conservative estimate of costs on which to base decisions.
Dissolved Metals Water Quality Criteria
In the final Guidance, criteria for metals were revised to reflect new and updated data, and
expressed for the dissolved form for aquatic life criteria (for the proposed Guidance, criteria
were based on total metals). This change generally results in less stringent criteria for metals
compared with criteria expressed in total form.
In order to apply metals criteria in the dissolved form for the cost analysis, conversion factors
based on toxicity testing results were used to revise criteria from the total form to the
dissolved form. The conversion factors were based on the ratio of dissolved to total metals
present in the laboratory toxicity tests performed to establish criteria for toxic metals.
Conversion factors ranged from 0.333 for trivalent chromium, to 1.0 for trivalent arsenic.
Where conversion factors were not available, EPA assumed the most protective conversion
factor of 1.0. Since most effluent data are reported in the total form, the criteria for the
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS •> 4-3
dissolved form of metals were then adjusted back to the total form using a theoretical
partitioning relationship (U.S. EPA, 1993).
Intake Credits
The proposed Guidance did not allow intake credits except for the exclusion of facilities that
do not alter chemical concentrations in their waste streams (i.e., simple pass through). In the
final Guidance, intake credits are allowed for discharges into nonattainment waters.
For the cost analysis, simple pass through and intake credits were evaluated as follows.
Outfalls were determined to have no reasonable potential for the disch«arge to cause or
contribute to an excursion above a narrative or numeric water quality criterion (simple pass
through) if they met the following criteria:2 ,i
> The facility withdraws 100% of the intake water containing the pollutant from,
the same body of water into which the discharge is made.
»• The facility does not contribute any additional mass of the identified intake
water pollutant to its wastewater.
»• The facility does not alter the identified intake water pollutant chemically or
physically in a manner that would cause adverse water quality impacts that
would not occur had it been left in-stream. i
>• The facility does not increase the identified intake water pollutant concentration
compared to the pollutant concentration in the intake water.
i
»• The timing and location of the discharge does not cause adverse water quality
impacts to occur that would not occur if the identified intake pollutants were
left in-stream. ;
Intake credits were granted to sample facilities when the level of the pollutant upstream of the
discharge exceeded the most stringent applicable water quality criterion for that pollutant.
When this situation occurred, relief was provided by making the water quality-based effluent
level (WQBEL) for the pollutant(s) equal to the most stringent Guidance criterion. This was
done for both discharges to different and same bodies of water.3
2 There were not adequate intake pollutant data for the sample facilities to determine whether all of
the criteria described above would be met. This limitation may underestimate compliance costs.
3 The final Guidance allows "no net increase" (i.e., discharge at background concentrations) for up
to 10 years or until a total maximum daily load (TMDL) is established for discharges to the same body of
water. It was conservatively assumed that TMDLs justifying loads greater than criteria would not be
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS •» 4-4
Additivity of Carcinogenic Effects
The proposed Guidance introduced the concept of accounting for the additivity of human
carcinogenic effects of carcinogens contained in a discharge. Under the final Guidance states
and tribes will have to consider additivity in the development of criteria or when establishing
water quality-based effluent limits. Criteria for individual pollutants are developed under the
Guidance to protect the population at a chosen risk level (based on an assumed fish
consumption level). The final Guidance allows use of an incremental cancer risk equal to one
in 10 000 (10-4) to account for the additivity of carcinogenic effects for a mixture as a whole.
Alternatively, additivity may be accounted for by establishing a risk-associated dose for ^
human health at a level corresponding to an incremental cancer risk of one in 1,000,000 (10 )
for individual pollutants in the mixture, or applying a scientifically defensible method to
account for the additive effects of carcinogens.
For the cost analysis, a one in 100,000 (10'5) risk level in a mixture was used to account for
additivity since some states may choose this level. The estimated cost attributable to additivity
was calculated based on the number of potential carcinogens discharged by a modeled facility,
the concentration of those pollutants in the discharge, and the background concentrations of
the pollutants. The human cancer value (HCV) associated with a lifetime incremental cancer
risk of one in 100,000 was determined for the pollutants identified in the discharge. Then, the
HCVs were divided by the total number of carcinogens in the discharge to determine the
allowable concentration for each carcinogen that could be discharged. These concentrations
were compared to the actual concentrations in the discharge to determine whether the facility
would be required to reduce pollutant concentration levels. This is a more costly approach
than would be allowed under the final Guidance.
4.2 MODIFICATIONS TO THE METHODOLOGY FOR ESTIMATING COSTS"
Modifications to the original methodology for estimating costs and pollutant load reductions
include the consideration of Tier II pollutants and a revised compliance cost decision matrix,
and the use of updated data for sample facilities, indirect dischargers, and toxic weights.
These modifications are described below.
developed after 10 years and dischargers to the same body of water would eventually need to comply with
the most stringent criteria at the end-of-pipe. This assumption may overestimate compliance costs.
4 An attempt was also made to verify or revise the original cost estimates analysis for waste
minimization/pollution prevention techniques. However, due to the general lack of additional information
available, no changes to the original cost estimates were made. Both of the EPA Pollution Prevention
Office and the American Institute of Pollution Prevention acknowledged the difficulty in developing
generic costs because of the site-specific nature of manufacturing processes and pollutants being removed.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS + 4-5
Criteria for Tier n Pollutants
Tier II pollutants (pollutants for which insufficient data are available to derive Tier I criteria)
were excluded from the cost analysis of the proposed Guidance. In response to comments
calling for EPA to address Tier II pollutants, compliance costs were estimated for these
pollutants as described below.
Three criteria were used to determine whether pollutants for which Tier I criteria were not
available should be included in the cost analysis of the final Guidance: (1) loadings,
(2) frequency of occurrence, and (3) toxicity. To determine which pollutants exhibited
significant loadings to the Great Lakes Basin, the loadings of the 138 pollutants of initial
focus for the Guidance at the 59 study facilities were calculated based on facility permit
limits or measured effluent concentrations. The loadings were multiplied by EPA-derived
toxic weights to normalize the toxicity of each pollutant to that of copper. Using the statistical
extrapolation factors developed for the costing analysis, the total toxic weighted loadings for
the 138 pollutants were extrapolated to the universe of major dischargers in the Great Lakes
Basin. Those pollutants that exhibited loadings under 10 pounds-equivalent per day were
omitted from the final costing analysis.5 In addition, any pollutant that: was limited, detected,
or required to be monitored at three or more of the 59 sample facilities was also included in
the final costing analyses. Since the loadings analysis should have captured the most
significant pollutants of concern, this frequency of occurrence evaluation was considered a
"safety net."
As a final "safety net," any pollutant limited, detected, or required to be monitored at any
facility that exhibited a high toxicity (high toxic weight) was included in the final costing
analysis. This evaluation was performed by multiplying the "frequency of occurrence" for a
given pollutant by its toxic weight. The resulting value was designated the "Occurrence Toxic
Weight" (OTE). This procedure captured those pollutants that might have not been detected,
and escaped the loadings evaluation, but that had monitoring requirements at one or more
facilities. The target value of 0.1 OTE was selected to ensure that any pollutant with a toxic
weight of 50 or greater and even a single monitoring requirement at one sample facility,
would be included in the final costing analysis.
The OTE evaluation identified an additional 76 pollutants that were limited, detected in the
effluent, or required to be monitored at one or more of the 59 sample facilities. From this list
of.76 pollutants, 37 were determined to be of consequence to the costing and loadings
analyses using the three criteria described above. Thus, a total of 69 pollutants were evaluated
in the cost analysis for the final Guidance. Table 4-1 lists the pollutants included in the
original cost analysis and those added for the cost analysis of the final Guidance, as well as
those found but not included. After establishing the list of pollutants for the final costing
5 This value corresponds to a total pollutant load from all major point source dischargers to the
Great Lakes Basin of 10 pounds of copper per day.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS * 4-6
Table 4-1
Pollutants Considered in the Analysis of Costs of the Final Guidance
Pollutants Included in the
Cost Analysis of the Proposed
Guidance
2,2,7,8-TCDD
2,4-Dimethylphenol
2,4-Dinitrophenol
Arsenic (III)
Benzene
Cadmium
Chlordane
Chlorobenzene
Chromium (III)
Chromium (IV)
Copper
Cyanide, free'
Cyanide, total
DDT
Dieldrin
Endrm
Heptachlor .
Hexachlorobenzene
Hexachloroethane
Lindane
Mercury
Methylene Chloride
Nickel
Parathion
PCBs
Pentachlorophenol
Phenol
Toluene
Total Selenium
Toxaphene
Trichloroethylene
Zinc
Pollutants Added for Analysis
of Costs of the Final Guidance
1,1 -Dichloroethane
1,1 -Dichloroethylene
1,1,1 -Trichloroethane
1,2-Dichloroethane
1,2-Dichloropropane
1,2-trans-Dichloroethylene
1,2,4,5-Tetrachlorobenzene
2,4,6-Trichlorophenol
3,3-Dichlorobenzidine
4,4-DDD
4,4-DDE
Acrylonitrile
Aldrin
alpha-Endosulfan
alpha-Hexachlorocyclohexane
Aluminum
Antimony
Benzidine
Benzo(a)pyrene
Beryllium
beta-Endosulfan
beta-Hexachlorocyclohexane
Carbon tetrachloride
Chloroform
Chlorpyrifos
Chrysene
Endosulfan
Fluoranthene
Hexachlorocyclohexane
Iron
Lead
Pentachlorobenzene
Phenanthrene
Silver
Tetrachloroethylene
Thallium
Pollutants Found but Not
Included in Cost Analysis
2-Nitrophenol
4-Nitrophenol
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
3,4-Benzofluoranthene
11,12-Benzofluoranthene
1,1,2-Trichloroethane
1,2,4-Trichlorobenzene
2-Chlorophenol
2,4-Dichlorophenol
2,4-Dichlorophenoxyacetic acid
Acenaphthalene
Acrolein
Anthracene
Bis(2-chloroethyl)ether
Bis(2-chloroisopropyl)ether
Bromoform
Butylbenzylphthalate
Chlorodibromomethane
Diethyl phthalate
Di-n-butylphthalate
Dimethylphthalate
Ethylbenzene
Fluorine
Hexachlorobutadiene
Indeno(l,2,3 cd)pyrene
Isophorone
Methyl bromide
Methylchloride
Naphthalene
Octachlorostyrene
Pyrene
Vinyl Chloride
Source: SAIC, 1995.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS » 4-7
analyses, criteria for these pollutants were developed utilizing the Tier I and Tier II
procedures outlined in Appendices A, B, C, and D of the final Guidance.
Compliance Cost Decision Matrix
The cost analysis of the final Guidance reflects modifications to the compliance decisions of
the modeled facilities. In the cost analysis of the proposed Guidance, it was assumed that
when treatment costs became excessive, or if information regarding the existing treatment
system was lacking, that waste minimization/pollution prevention techniques would be the
preferred control approach selected by the regulated community. Alternatives available to
facilities through regulatory relief mechanisms such as variances, mixing zone studies,
phased-TMDLs, site-specific criteria, etc., were not considered.
In an effort to ensure consistency in estimating the general types of controls that would be
necessary for a sample facility to comply with the final Guidance, as well as to integrate the
other alternatives available through the final Guidance into the cost analysis, a costing
decision matrix was used for each sample facility. The underlying principle of the decision
matrix is that a facility will examine least-cost alternatives prior to incurring the expense and
potential liabilities associated with constructing end-of-pipe treatment facilities.
Under the decision matrix, costs for small treatment plant operating and facility changes were
considered firsl Where it was not technically feasible to simply adjust existing operations,
waste minimization/pollution prevention controls were considered; however, these controls
were costed only where they were considered feasible based on EPA's understanding of the
process employed at a facility.6 If waste minimization was deemed unfeasible to reduce
pollutant levels to those needed to comply with the final Guidance criteria, a combination of
waste minimization/pollution prevention and simple treatment was considered. If these
relatively low-cost controls could not achieve the Guidance-based WQBELs then, finally, end-
of-pipe treatment was considered.
However, before assuming that end-of-pipe treatment would be installed by the facility, the
relationship between the cost of adding the treatment and other types of remedies or controls
was considered. If it was concluded that other remedies or controls would be more feasible
than installing end-of-pipe treatment (i.e., estimated annualized cost for removal of a pollutant
exceeded a specified dollar amount per toxic pounds-equivalent, for example, $200 for the
low end cost scenario), then it was assumed that a facility would pursue some type of
regulatory relief from the WQBEL. When it was assumed that facilities would pursue
alternative relief, no treatment cost was estimated for a facility; however, a nominal cost for
some efforts to reduce the pollutant until the relief is granted was included. No credit was
6 In general, detailed treatment and manufacturing process information is not available in NPDES
permit files; therefore the assessment of feasibility was primarily based upon best professional judgement
using general knowledge of industrial and municipal operations.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS »• 4-8
taken for load reductions for any pollutant for which alternative relief was assumed. Through
discussions with EPA regional and state permitting agencies and outside experts, it was
estimated that the typical costs for facilities pursuing the relief mechanisms could range from
a high of $1,000,000 per pollutant for phased-TMDLs to a low of $20,000 for criteria
modifications. For the compliance cost estimate, a mid range cost value of $200,000 per
pollutant each time a relief mechanism was assumed necessary was utilized.
In developing and using the cost decision matrix, it was acknowledged that granting relief
from WQBELs is dependent upon the specific circumstances at a facility, as well as the
judgement and implementing procedures of the state and tribal permitting authority. It was
also acknowledged that opportunities for waste minimization are dependent upon the specific
circumstances at a facility. The use of a $200 per toxic pounds-equivalent trigger for a
"facility" assumes that the regulatory flexibility in the Guidance would be available and
granted to all facilities that exceed the cost trigger. The cost estimate based on the $200 per
toxic pounds-equivalent trigger for a "facility" was considered representative of the low-end
of compliance costs attributable to the Guidance.
Acknowledging that the use of regulatory relief may be limited depending upon the particular
circumstances for a "facility," costs were also estimated under a higher cost scenario that
assumes regulatory relief would be granted only when the cost for the particular "category of
dischargers" exceeds a cost trigger. Particularly, if the estimated annualized cost for a
"category of dischargers" exceeded $500 per toxic pounds-equivalent then it was assumed that
dischargers within the "category" would be granted regulatory relief. This cost trigger was
based on the highest costs to comply with promulgated effluent guideline limitations and
standards for direct discharger industrial categories which ranged from $1 - $500 per toxic
pounds equivalent per industrial category.
Updated Data for Sample Facilities
Updated and more comprehensive data for the sample facilities was collected to revise the
cost analysis for the final Guidance. The cost analysis of the proposed Guidance reflected data
from EPA Region 5 and state permitting authorities. Discharge data were based on 1990
Permit Compliance System (PCS) data; facility-specific permit file information was generally
from 1992. For the cost analysis of the final Guidance, the most current information and data
(including permits, fact sheets, permit applications, and other relevant discharge information)
were used as the basis for comparison to Guidance requirements. In addition, state permitting
authorities were requested to review the original cost estimate for each sample facility and
provide comments and additional information as necessary to ensure the accurate reflection of
current permit requirements and discharge conditions.
The costs and loadings estimates for the final Guidance are based on 1993 PCS discharge
data, and permit file information and data provided by state permitting authorities (generally
representing permits issued as late as mid-1994). As a result of using more recent data, the
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS » 4-9
baseline permit requirements for the sample facilities were lowered due to more stringent
NPDES permit requirements being applied by permitting authorities.. The overall effect of this
lower baseline is that the estimated compliance costs and pollutant load reductions are not as
substantial as those projected for the proposed Guidance in the April 1993 RIA.
The original compliance cost analysis was limited by a lack of upstream receiving water
concentration data for the sample facilities. To fully evaluate the provisions of the final
Guidance for intake credits (i.e., determining whether discharges are to same or different
bodies of water and for identifying nonattainment waters) as well as to ensure that all
available data were used for the cost analysis, additional background concentration data for
each of the sample facilities was gathered. Data submitted as a part of the public comments
as well as information in the water quality files of the STORET data base were examined.
State permitting authorities were involved in the collection of all applicable data.
Consistent with Procedure 3 of Appendix F to the final Guidance, fish tissue data (either
caged or resident fish tissue data) was collected to represent ambient water column
background concentrations. When fish tissue data were available for the pollutants being
evaluated at a sample facility, the tissue data was converted to ambient water column
concentrations by dividing fish tissue data (in mg/kg wet weight) by the pollutant-specific
bioaccumulation factor (BAF) (in I/kg) used to derive Tier I Criteria and then multiplying the
result by 1,000 to give the result as concentration of pollutant (ug/1). When data for more
than one species was available, the geometric mean for all species was calculated and used.
Indirect Dischargers
For the cost, analysis of the proposed Guidance, it was assumed that the number of indirect
discharges that could be affected ranged from 10% to 30% based on analysis of one sample
municipal facility (publicly owned treatment works, or POTW). In order to verify this range,
data for an additional eight POTWs in Wisconsin and Michigan was collected and analyzed.
Additionally, the original sample POTW was re-evaluated based on changes to the final
Guidance (as reflected in estimated WQBELs for the POTW).
Since not all of the eight POTWs were selected as sample facilities and analyzed under the
study, it was assumed that the pollutants limited by each POTWs existing NPDES permit
would be the same as those that would require regulation under the Guidance. For each
POTW, the potential indirect dischargers of each regulated pollutant were identified from
among the POTWs list of indirect dischargers and the number of industrial users found to be
violating the POTWs permit limits for any of the pollutants of concern over a 1-year period.
These data suggested that the range of potentially affected indirect users is between 8% and
44% of the total number of the indirect dischargers to a POTW. The results showed that the
assumed range of 10% to 30% for the number of indirect dischargers, affected had a
reasonable basis.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS » 4-10
For purposes of developing costs for the final Guidance, it was assumed that 30 /o of all
indirect dischargers in the Great Lakes Basin would be impacted by. source control efforts by
POTWs as a result of more restrictive Guidance-based WQBELs under the low cost scenario.
For the high scenario in which a greater share of municipalities were assumed to install end-
of-pipe treatment, 10% of indirect dischargers were assumed to be impacted. The average
compliance cost per direct discharger facility was updated to reflect revisions made to the
sample facilities as a result of the final Guidance and was used as the compliance cost for
indirect dischargers impacted under the relevant scenario.
Revised Toxic Weights
Toxic weights were used to derive cost-effectiveness estimates for the proposed Guidance, as
well as to compare the relative loadings of the 138 pollutants of concern analyzed for the cost
study Toxic weights are used by EPA as normalizing factors that relate the toxicity of any
pollutant to that of copper. The factor considers the aquatic toxicity and the human health
effects of a pollutant and is calculated with the following formula:
Toxic Weight - 5.6/[fresh water chronic criteria (ug/1)] + 5.6/[human health criteria (ug/1)]
The value of 5 6 ug/1 was the original national chronic water quality criterion for copper
(making the toxic weight for copper equal to one). The national chronic water quality
criterion for copper has been revised to 12 ug/1. However, the 5.6 value for computing toxic
weights has been retained by EPA for consistency. This results in copper currently having a
toxic weight of 5.6/12, or 0.47.
Toxic weights from 1988 were used in calculating baseline pollutant loads and load
reductions for the proposed Guidance. In analyzing the impact of the final Guidance toxic
weights were developed or recalculated for all 69 pollutants included in the cost study, using
the most recent criteria and toxicity information available to EPA. These updates raised some
of the 1988 toxic weights and lowered others, depending on the toxicity data available tor a
specific pollutant.
4.3 RESULTS
As shown in Table 4-2, the total annualized costs of implementing the final Guidance to
direct and indirect dischargers is estimated to range from $61 million to $376 million.
Tables 4-3, and 4-4 show the breakdown of costs by industry and cost category. Under the
7 This represents a downward revision in costs from the proposed Guidance. The downward revision
in costs is largely attributable to the intake credit provisions of the final Guidance and the use of dissolved
metals criteria. The lowering of the permit baseline also results in an overall decrease in compliance costs
and pollutant load reductions.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS >• 4-ll
Table 4-2
Summary of Annualized Compliance Costs of the Final Guidance
Cost Categories
Major Direct Dischargers — Industrial
Major Direct Dischargers — Municipal
Minor Direct Dischargers
Indirect Dischargers
Total
Number
of
Facilities
272
316
3,207
3,528
7,323
Low
Estimated Costs
(First Quarter 1994 S,
Millions)
15.7
23.8
1.6
19.9
61.4
High
Estimated Costs
(First Quarter 1994 S,
Millions)
108.2
259.8
1.6
6.6
376.2
Source: SAIC, 1995.
low estimate, direct dischargers incur about 67% of total estimated compliance costs while
indirect dischargers incur about 33%. Under the high estimate, direct dischargers account for
about 98% of the total estimated cost, and indirect dischargers account for 2%. This shift m
proportion of costs between direct and indirect dischargers between the high and the low
estimates is due to the increased use of end-of-pipe treatment for direct dischargers under the
high estimate. In addition, it was assumed that a smaller proportion of indirect dischargers
(10%) would be impacted under the high estimate, since municipalities are adding end-of-pipe
treatment, which should reduce the need for source controls (i.e., reduce the need for increased
pretreatment program efforts). The low and high cost scenarios are discussed in greater detail
below.
Low Estimate of Compliance Cost
Under the low estimate of compliance cost, for the direct dischargers, municipal majors are
expected to incur 58% of total costs and industrial majors about 38% of total costs. Minor
direct dischargers are estimated to incur 4% of the total costs. The two major industrial
categories with the largest total annualized cost are pulp and paper (2:0%) and miscellaneous
(11%). The food and food products and metal finishing categories are estimated to incur less
than 1% of the total annualized cost.
Although the municipal major category accounts for over 58% of the total estimated cost, the
average annual cost is just over $75,000 per facility. Average annualized costs for industrial
majors vary widely by industry, with the highest average cost estimated for miscellaneous
($168,000 per plant) and pulp and paper ($151,000 per plant) facilities. For minor facilities,
average costs are negligible at an estimated $500 per facility.
-------
ts
1
A
CO
CO
§
E
I
w
t—
8
Q
^
?2
CO
8
u.
o
co
c73
3
S
Q
tu
CO
I
•
fc
o
DC
S
rj
^^
2
^^
•o
B
••*
M
|
1 ^
S
8
|
1
3! J
3 w J3
« 5 i.
H v. ^
i3
8
Q
V
u
.S
"a.
IB
O
Q
^
"Si
H "5
•3 o
S IB
Is
o
i
3
^
£,
1 «
99 J£J
9 jj
§"*
tf
"a
_0
lU ^*
-2: «
3-1
^1
ii
a
•^ iS
§ «
^ ••*]
12 1
2
o
M
B
*c
1
^
'•••1
'E
U
Industry
813638383
o — o o o' o oo — *»•'
oo o co TT
fsj OO *"* *-^
0 — ' 0 0*
— . «« — a\ r~
r*, ir> m oe O
o o o e>i o
vo ^O f*1 ^^ *o
oo v~> m "< Q
^j. o 0 — 0.
o o o. m o
o o *-« --
o o o" . .4. (—j ^^ — *• Q\ rt\ ^«
^^ r^ ^M ^3 ON ^f 2^ CN t"*
ooooo»— «^-*oo
o o o o d o o o o*
ON t**^
CO OC
0 — '
"a
fi
CA
t3 J2
^ § CO
1 1 -3 f js
fi S o S
S>'S|s «a'§S-^S
g§J'J 1 J 8- § S
jg te. U .« §go,g8
•S 1 -2 (S ^ o 1 S J
lllllllfll
.^
f
3s
1O
f*
f^
EN
S
^^
S
ro
m
S
-o
CO
»0
o
1
2
S
Dischargers —
ts
.S
Q
t>
o
'o1
S
[^
S
*•>*
o
^1
2
»«^
^M
m
o
Dischargers
I
Ut
2
.5
vt
o
PM
E2
u
I
H
S
•5
u
.5
•o
^
5
sO
•»
"O
in
ts
r^
^4
Tt
vo
f*^
m
es
•n
•o
VO
^B
vo
2
VO
9
"^
^
^
s>
1
0
^J
c/a
• "8
en 2
.2 M
i1i.t
>-, a <" T3 H
0 1 -&<>**
0 *. 0 «
S? g S-§
^ S S "«
^ cs &0 "X
"8 1 J o
H 2 o *»
-.£ 0..2 ^3
111 1
- 1 2 £
'§••1 s ^
U £ C3 Q
o
0
-------
C/3
00
w
U.
tL.
U
I
U
O
U
5
00
5
Q
a
00
^
o
WO
"3
ta>
WJ
O
U
•3
S
•M
in
•a
B
£?
>
U
•«^
V3
•e
ts
llsSSSiSS
o «s o o o" NO y-i — —
oo •—
2
*o
o
I/-1 u~l w~l CN
O O O 0
o o o o
oc v~i m — ro
•<*• O o — O
o' o" o' p-i o'
v-i w-i r
« C3
0 g>
• 'S 00 .
J2 > oo oo
re W C3 s
QJ U. Crt • —
"^ S ^° "e
o S3 •—• s
— i- O O
00 C ^
S 2 u o
> (-* L^ -"^
O 0 u
xp C * 3
iitl
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS » 4-14
Costs to direct dischargers for developing and implementing pollutant minimization plans
(required when WQBELs are below detection levels) account for most of the costs (58% of
total annual costs). Annualized capital and operation and maintenance (O&M) costs make up
just over 21% of total annual costs; waste minimization costs account for almost 11%.
By pollutant, controls for mercury account for over 20% of annual costs (attributable
primarily to costs related to pollutant minimization plans). Other pollutants that account for
significant costs include methylene chloride, aluminum, benzene, and copper.
High Estimate of Compliance Cost
Under the high estimate of compliance cost, for the direct dischargers, municipal majors are
expected to incur about 70% of total costs and industrial majors about 29% of total costs.
Minor direct dischargers are estimated to incur less than 1% of total costs. The two major
industrial categories with the largest total annualized cost are pulp and paper (23%) and
miscellaneous (3%). Even under the high estimate, the food and food products and metal
finishing industries are estimated to incur less than 1% of the total annualized cost.
The municipal major category accounts for almost 70% of the total estimated annual cost; the
average annual cost is just over $822,000 per facility. Average annualized costs for industrial
majors vary widely across categories, with the highest average cost estimated for pulp and
paper ($1,583,000 per plant) and miscellaneous ($433,700 per plant) categories. For minor
facilities, average costs are negligible at an estimated $500 per facility.
For the high estimate, costs to direct dischargers shifted away from developing and
implementing pollutant minimization plans and waste minimization to capital and O&M costs
(over 52% of total annual costs) associated with construction and application of end-of-pipe
treatment. Annualized costs for developing and implementing pollutant minimization plans
make up about 6% of total annual costs; waste minimization costs account for less than 1%.
By pollutant, controls for lead account for over 60% of annual costs (attributable primarily to
costs related to end-of-pipe treatment). Other pollutants that account for significant costs
include heptachlor, pentachlorophenol, lindane, and mercury.
Estimated Pollutant Reductions
Tables 4-5 and 4-6 present the estimated unweighted and toxicity-weighted pollutant baseline
and reductions for the final Guidance. As shown in Table 4-6, baseline pollutant loadings are
projected to be just over 35 million toxic pounds-equivalent per year (lbs-eq/year).8 It should
8 The baseline pollutant loading represents a downward revision from the baseline estimated for the
proposed Guidance.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS » 4-15
Table 4-5
Estimated Unweighted Baseline Pollutant Loadings and Reductions in Loadings
Anticipated to Result from the Guidance Basinwide (Ibs/year)
Pollutant
Acrylonitrile
Aldrin
Aluminum
Antimony
Arsenic (III)
Benzene
Benzidine
Benzo(a)pyrene
Beryllium
Cadmium
Carbon tetracbloride
Chlordane
Chlorobenzene
Chloroform
Chlorpyrifos
Chromium (III)
Chromium (VI)
Chrysene
Copper
Cyanide, free
Cyanide, tool
4,4-DDD
4,4-DDE
DDT
3,3-Dichlorobenzidine
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethylene
1 ,2-trans-Dichloroethylene
1 ,2-Dichloropropane
Dieldrin
2,4-Dimethylphenol
2,4-Dinitrophenol
Endosulfan
alpha-Endosulfan
beta-Endosulfan
Endrin
Fluoranthene
Baseline
Loadings1
37,185,781
5,494
9,111
66,313
4,985
424
61,429
5,111
87,218
0
0
10
15,132
4,677
344
333
56
1,934
deduction2
(Low)
397,172
5,389
56
0
4,054
289
3,333
1,583
9,657
0
0
0
9,400
3,065
0
333
37
1,875
Percent
Change
-1.1%
-98.1%
-0.6%
0.0%
-81.3%
-68.1%
-5.4%
-31.0%
-11.1%
-51.1%
-47.6%
-0.2%
-62.1%
-65.5%
0.0%
-100.0%
-65.6%
-97.0%
Reduction2
(High)
397,172
5,389
56
0
4,323
289
3,333
1,583
9,657
0
0
0
12,392
: 3,065
0
; 333
37
1,875
Percent
Change
-1.1%
-98.1%
-0.6%
0.0%
-86.7%
-68.1%
-5.4%
-31.0%
-11.1%
-51.1%
-47.6%
-0.2%
-81.9%
-65.5%
Of\n/
.0%
-100.0%
-65.6%
-97.0%
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS > 4-16
Table 4-5 (cont.)
Estimated Unweighted Baseline Pollutant Loadings and Reductions in Loadings
Anticipated to Result from the Guidance Basinwide (Ibs/year)
Pollutant
Fluoride
iHeptachlor
Hexachlorobenzene
Hexachlorocyclohexane
alpha-Hexachlorocyclohexane
beta-Hexachlorocyclohexane
Hexachloroethane
Iron
JilUll
Lead
T inriane
JL«111IJI(U1W
Mercury
Methylene Chloride
Nickel
Parathion
prB<;
JT ^-ji"*o
Pentachlorobenzene
Pentachlorqphenol
Phenanthrene
1 Phenol
Selenium, total
Silver
2,3,7,8- fCDD
1 ,2,4,5 -Tetrachlorobenzene
Tetrachloroethylene
Thallium
Toluene
Toxaphene
11,1,1 -Trichloroethane
Trichloroethylene
2,4,6-Trichlorophenol
Zinc
Totals
Baseline
Loadings1
2,102,400
567
754
1,926
1,929
1,926
3,166,429
997,118
77
1,039
11,905
2,444
61
192,974
13,484
9,078
0
194,448
2,143
580
96
34,020
44,183,751
Reduction2
(Low)
0
106
272
1,843
1,900
1,869
0
600,078
0
133
4,762
2,111
0
191,967
0
0
0
188,401
0
1
72
1,490
1,431,248
Percent
Change
0.0%
-18.7%
-36.1%
-95.7%
-98.5%
-97.0%
0.0%
-60.2%
0.0%
-12.8%
-40.0%
-86.4%
0.0%
-99.5%
0.0%
0.0%
0.0%
-96.9%
0.0%
-0.2%
-75.0%
-4.4%
-3.2%
Reduction2
(High)
0
537
272
1,843
1,902
1,869
0
666,078
76
136
4,762
2,111
0
191,969
11,782
0
0
193,268
357
1
72
1,490
1,452,029
Percent
Change
0.0%
-94.7%
-36.1%
-95.7%
-98.6%
-97.0%
OAfl/
.0%
60.2%
-98.6%
-13.1%
-40.0%
-86.4%
Or\nj
.0%
-99.5%
-87.4%
0.0%
0,0%
-99.4%
-16.7%
-0.2%
-75.0%
-4.4%
-3.3%
1 Based on current permit limits, not actual loadings.
2 Based on difference between current permit limits and projected Guidance-based limits.
Source: SAIC, 1995.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS » 4-17
Table 4-6
Estimated Toxic Weighted Pounds Equivalent* Pollutant Loadings and Reductions
in Loadings Anticipated to Result from the Guidance Basinwide
Pollutant
Acrylonitrile
Aldrin
Aluminum
Antimony
Arsenic (III)
Benzene
Benzidine
Benzo(a)pyrene
Beryllium
Cadmium
Carbon tetrachloride
Chlordane
Chlorobenzene
Chloroform
Chlorpyrifos
Chromium (III)
Chromium (VI)
Chrysene
Copper
r r
Cyanide, free
Cyanirif , total
4,4-DDD
4,4-DDE
DDT
3,3-Dichlorobenzidine
1,1-Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethylene
1 ,2-trans-Dichloroethylene
1 ,2-Dichloropropane
Dieldrin
2,4-Dimethylphenol
2,4-Dinitrophenol
Endosulfan
alpha-Endosulfan
beta-Endosulfan
Endrin
Fluoranthene
Baseline
Loadings1
2,379,890
21,975
164
344,827
648
975,523
129
2,401
95,940
45
21
88,152
110,466
29
62
5
3,190,719
189,557
Reduction2
(Low)
25,419
21,556
1
0
527
664,604
7
744
10,623
23
10
212
68,619
19
0
5
2,092,368
183,778
Percent
Change
-1.1%
-98.1%
-0.6%
0.0%
-81.3%
-68.1%
-5.4%
-31.0%
-11.1%
-51.1%
-47.6%
-0.2%
-62.1%
-65.5%
0.0%
-100.0%
-65.6%
-97.0%
Reduction2
(High)
I
25,419
21,556
1
0
562
664,604
: 7
744
10,623
23
10
212
90,465
19
0
5
2,092,368
\
,'
183,778
Percent
Change
-1.1%
-98.1%
-0.6%
0.0%
-86.7%
-68.1%
-5.4%
-31.0%
-11.1%
-51.1%
-47.6%
-0.2%
-81.9%
-65.5%
0.0%
-100.0%
-65.6%
-97.0%
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS > 4-18
Table 4-6 (cont.)
Estimated Toxic Weighted Pounds Equivalent* Pollutant Loadings and Reductions
in Loadings Anticipated to Result from the Guidance Basinwide
Pollutant
Fluoride
Heptachlor
Hexachlorobenzene
Hexachlorocyclohexane
alpha-Hexachlorocyclohexane
beta-Hexachlorocyclohexane
Hexachloroethane
Iron
Lead
Lindane
Mercury
Methylene Chloride
Nickel
Parathion
PCBs
Pentachlorobenzene
Pentachlorophenol
Phenanthrene
Phenol
Selenium, total
Silver
2,3,7,8-TCDD
1,2,4,5-Tetrachlorobenzene
Tetrachloroethylene
Thallium
Toluene
Toxaphene
1.1.1 -Trichloroethane
77
Trichloroethylene
2,4,6-Trichlorophenol
Zinc
Totals
Baseline
Loadings1
73,584
2,324,390
542,816
34,675
82,945
23,117
17,732
1,794,813
5,366
519,286
5
88
454,908
443,840
6,742
426,685
3,989,245
388,895
12
16,833,496
8
1,735
35,364,937
Reduction2
(Low)
0
434,659
195,908
33,172
81,721
22,423
0
1,080,141
0
66,304
2
76
0
441,523
0
0
0
376,802
0
36,956
6
76
5,838,284
Percent
Change
0.0%
-18.7%
-36.1%
-95.7%
-98.5%
-97.0%
0.0%
-60.2%
0.0%
-12.8%
-40.0%
-86.4%
0.0%
-99.5%
0.0%
0.0%
0.0%
-96.9%
0.0%
-0.2%
-75.0%
-4.4%
-16.5%
Reduction2
(High)
0
2,201,441
195,908
33,172
81,788
22,423
0
1,080,141
5,289
67,878
2
76
0
441,528
5,891
0
.0
386,536
2
36,956
6
76
7,649,509
Percent
Change
0.0%
-94.7%
-36.1%
-95.7%
-98.6%
-97.0%
II
0.0%
-60.2%
-98.6%
-13.1%
-40.0%
-86.4%
0.0%
-99.5%
-87.4%
0.0%
0.0%
-99.4%
-16.7%
-0.2%
-75.0%
-4.4%
-21.6%
* Copper-based.
1 Based on current permit limits, not actual loadings.
2 Based on difference between current permit limits and projected Guidance-based limits.
Source: SAIC, 1995.
-------
REVISED ANALYSIS OF COSTS AND COST-EFFECTIVENESS > 4-19
be noted that the modeled baseline pollutant loadings represent allowable loadings as reflected
by current permits, not actual loadings based on discharge concentration data.
Reductions in loadings are modeled as the change from current permit, limits to Guidance-
based permit limits. Under the low cost estimate, pollutant loadings would be reduced by
5 8 million Ibs-eq/year, which represents a 16% reduction of the baseline pollutant loadings.
Under the high cost estimate, pollutant loadings would be reduced by 7.6 million Ibs-eq/year,
which represents a 22% reduction of the baseline pollutant loadings.
Under the low cost estimate, the largest pollutant load reductions occur for dieldrin and lead,
which account for over 50% of the toxic-weighted load reduction. Chlordane, heptachlor, and
pentachlorobenzene were also reduced by significant amounts from the baseline. Under the
high cost estimate, the largest pollutant load reductions occur for heptachlor, dieldrin, and
lead, which account for about 70% of the toxic-weighted load reduction.
Approximately 80%. of the pollutant load reduction for the final Guidiance, regardless of the
scenario, is attributable to reducing bioaccumulative pollutants of concern (BCCs) as a result
of pollution minimization plans and end-of-pipe treatment. However, it should be noted that
for several BCCs (e.g., PCBs, 2,3,7,8-TCDD, mercury, toxaphene), little or no reduction from
the baseline is estimated. This phenomenon occurs because of the method used to derive load
reductions. When an existing permit limit or a Guidance-based WQBEL is below the
analytical detection level, one-half of the method detection level is used for each. The result
of this approach is that no pollutant reduction is estimated regardless of whether the
Guidance-based WQBEL is further below detection levels than the existing permit limit. In
addition, for several of the toxic pollutants of concern, the lack of estimated reduction is due
to the downward shift in the permit baseline (i.e., more stringent existing permit limits) as a
result of updating the database for the final RIA.
Cost-Effectiveness
Under the low cost estimate, the cost per toxic-weighted pound of pollutant reduced, or the
cost-effectiveness of the final Guidance, is approximately $10.30/lbs-eq. Under the high cost
estimate, cost effectiveness is estimated at just over $49.00/lbs-eq. Cost-effectiveness values
for effluent limitations guidelines and standards range from just over $1.00/lb-eq/year to over
$500/lbs-eq/year.
9 Thus, baseline loadings may be projected where none are present due to the content of permits.
The state of Wisconsin, for example, puts limits for BCCs in the permits for all facilities with significant
discharges.
-------
-------
CHAPTER 5
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS
TO THE GUIDANCE
5.1 INTRODUCTION
The April 1993 RIA described the basis by which a portion of the potential benefits of water
quality improvements illustrated in the case study analyses were attributed to the proposed
Guidance, and the numerous sources of uncertainty surrounding these estimates. In general,
baseline resource values and the value of water quality improvements were based on available
data and applied research. However, data were not available on the potential contribution of
the proposed Guidance toward such changes.
One uncertainty in the attribution of benefits related to the water quality baseline relevant to
implementation of the proposed Guidance. This baseline is expected to be some point beyond
(cleaner than) current conditions as a result of a number of regulatory actions already in place
(e.g., efforts to comply with 303(c)(2)(B) of the Clean Water Act, Guidance for nonpoint
source controls in coastal areas). Similarly, at the other end of the water quality spectrum, it
was not clear what water quality benchmark the proposed Guidance would attain with respect
to removing fish consumption advisories and moving toward a "toxic free" status (i.e., the
basis by which much of the benefits were estimated based on research by Lyke, 1992).
Another important uncertainty in the attribution issue was the relative contribution of point
sources to the toxics-related water quality problems in the Great Lakes Basin. That is,
although the reduction in point source loadings expected to result from the proposed Guidance
was estimated, the contribution of point sources to total loadings of the relevant pollutants in
the basin was not known.
Because of the lack of information on the attribution issue, benefits v/ere attributed to the
Guidance in the case study analyses for illustrative purposes based on general information
about the sites. For example, for the Fox River case study, the propoised Guidance was
expected to reduce loadings of several chemicals of primary concern in the watershed,
including PCBs, dioxin, and mercury. Thus, for the most significant benefits categories, 50%
of future toxics-oriented benefits were attributed to the proposed Guidance. In addition, a
sensitivity analysis was performed to show the impact of alternative attribution assumptions
on the benefits estimates.
Although the lack of data was not refuted, public comments on the proposed Guidance and
RIA focused on the uncertainty surrounding the attribution of benefits and the use of
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE'GUIDANCE »5-2
potentially "optimistic" attribution assumptions. OMB commented that EPA should try to
model the expected reductions in fish tissue contaminant concentrations (and thus changes in
fish consumption advisories) resulting from the Guidance. In addition, numerous comments
suggested that EPA should be addressing nonpoint source controls in addition to or instead ot
more stringent point source controls. Thus, subsequent research efforts were directed toward
better quantification of the potential impact of the Guidance in bringing about future toxic-
oriented benefits. Efforts were focused on determining the potential contribution of point
source loadings to the toxic-related problems in the basin; the results of this research are
presented in the following sections. The cost-effectiveness of point source controls was also
examined.
This chapter is organized as follows: Section 5.2 summarizes studies and data related to the
point source contribution of toxic contaminants to the Great Lakes. Section 5.3 presents a
screening analysis of other potential sources of toxic contaminants to the Great Lakes. Section
5 4 presents a modeling effort to estimate changes in fish tissue concentrations resulting from
reductions of point source loadings. Section 5.5 presents information on the cost-effectiveness
of point source controls.
5.2 STUDIES AND DATA RELATED TO THE POINT SOURCE CONTRIBUTION
TO LOADINGS OF GUIDANCE-IMPACTED CONTAMINANTS
Inferences regarding the point source contribution to loadings of the contaminants of concern
for the Guidance can be made from studies and data related to the inputs of contaminants in
the Great Lakes Basin. A significant amount of research has been conducted on the mass
balancing of contaminants in the Great Lakes; however, appreciable data gaps remain. A
summary of the available information and the inferences that can be drawn for the attribution
issue is provided below. In general, this research indicates that there is insufficient data
available to estimate total loadings (and thus calculate the point source contribution) for
almost all of the contaminants addressed by the Guidance, and that results are likely to be
highly site- and contaminant-specific.
5.2.1 Available Data on the Relative Contribution of Point Source Loadings
Assessing the contribution of point sources to toxic-related problems in the basin requires an
estimate of total loadings in the basin for the relevant contaminants. Of the 138 contaminants
of concern for the Guidance, EPA had sufficient information to attempt to estimate basinwide
loadings for only four of these: mercury, lead, cadmium, and PCBs (Warren, 1993). Even for
these contaminants for which information is available, the limitations of the studies used to
develop the estimates are significant (Warren, 1993).
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE » 5-3
Based on available information, EPA's estimates reflect atmospheric and tributary loadings
(i.e., total loadings) to the Great Lakes Basin. However, these estimates do not include the
entire load entering a lake from a connecting channel (e.g., Niagara Fiver) but only the
portion of that load that enters it from point sources and tributaries between lakes.
Atmospheric and tributary loadings were taken from published sources to the extent possible,
and these loadings were often based on one or more studies not necessarily in the Great
Lakes (Warren, 1993). For one lake, the tributary load was based on average water
concentrations and annual flow values, which likely resulted in a significant underestimate of
the load (Warren, 1993). Nonpoint source loadings (e.g., from sedime:nt, runoff) also may not
be included in the loadings for connecting channels.
Despite these limitations, the total loadings estimates can be compared to point source
loadings for a preliminary indication of the relative contribution of point sources. Total point
source estimates were also developed for the four chemicals based on the Permit Compliance
System (PCS); the PCS is currently the best source for estimating point source loads (U.S.
EPA, 1992b). Point source loadings and the estimated total loadings are reported in Table 5-1.
The relative contributions of point sources range from approximately 2% to 40%, depending
on the chemical.
Table 5-1
Relative Contribution of Point Source Loadings to Total (Atmospheric and Tributary)
Loadings in the Great Lakes Basin (kilograms pur year)
Pollutant
Mercury
Lead
Cadmium
PCBs
Point Sources
Loadings1
347
53.322
14,646
54
Total (Atmospheric +
Tributary) Loadings2
3,970 - 13,906
425,955 - 858,538
36,615 - 96,618
1,411-3,269
Point Source Loadings
Contribution
2.5% - 8.7%
6.2% - 12.5%
15.2% - 40.0%
1.7%- 3.8%
1 U.S. EPA, 1992b.
2 Warren, 1993.
Strachan and Eisenreich (1988) also estimated inputs of direct industrial and wastewater
dischargers of PCBs and lead to Lakes Superior and Huron (this does not include all point
sources because dischargers to tributaries are rolled into tributary flux), and compared them to
total loadings. These estimates are shown in Table 5-2. Although the data are expected to
underestimate point source loadings, the estimates indicate that the lower Great Lakes receive
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE
5-4
Table 5-2
Relative Contribution of Direct Point Source Discharges to Total Inputs
of PCBs and Lead in Lakes Superior and Huron
PCBs
Lake Superior
Lake Huron
Lake Superior
Lake Huron
Total
Input
(kg/yr)
606.0
636.0
Direct Wastewater
Discharge (kg/yr)
2.2 (0.4%)
38.0 (6.0%)
Direct Industrial
Discharge (kg/yr)
1.8 (0.3%)
8.0.(1-0%)
Lead
Total
Input
(103 kg/vr)
241.0
430.0
Direct Wastewater
Discharge
(103 kg/yr)
0.4 (0.2%)
3.3 (0.8%)
Direct Industrial .
Discharge
(103 kg/yr)
3.3 (1.4%)
5.1 (1.2%)
Total Direct Point
Source Discharge
(kg/yr)
4.0 (0.7%)
46.0 (7.0%)
Total Direct Point
Source Discharge
(103 kg/yr)
3.7 (1.5%)
8.4 (2.0%)
Note: Detail may not add to total due to rounding.
Source: Strachan and Eisenreich, 1988.
a greater percentage of total loadings from point sources than the upper lakes. The upper
lakes are estimated to receive a much greater fraction of their total loadings of toxic
contaminants from atmospheric sources (Table 5-3) than the lower lakes due to the relative
lack of local sources and the larger surface area of the upper lakes, and to the extensive
loadings to the lower lakes from sources on the Detroit-St. Clair and Niagara River systems
(Strachan and Eisenreich, 1988). The relative contribution of point sources in this study
ranges from 0.7% to. 1.5% for Lake Superior, and from 2.0% to 7.0% for Lake Huron.
5.2.2 Atmospheric Input of PCBs, t-DDT, Benzo(a)pyrene, Lead, and Mirex to the
Great Lakes
Strachan and Eisenreich (1988) found that the organic and inorganic contaminants .of concern
in the Great Lakes Basin are "present in rain and snow, atmospheric aerosols, and in the
vapor state in the Great Lakes Basin." Data were'not sufficient to determine reliably the
relative importance of atmospheric deposition to the total input of contaminants to the lakes,
or to perform mass balance studies for many toxic contaminants. However, sufficient
information was available to analyze total inputs and the relative contribution of inputs from
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE;GUIDANCE >• 5-5
Table 5-3
Relative Contribution of Atmospheric Deposition to Total Inputs
of Five Contaminants in the Great Lakes
Total Input
(kg/yr)
Percent of Total Inputs Attributable to
Atmospheric Deposition1
PCBs
Lake Superior
Lake Michigan
Lake Huron
Lake Erie
Lake Ontario
606
685
636
2520
2540
90%
58%
78%
13%
7%
t-DDT
Lake Superior
Lake Michigan '
Lake Huron
Lake Erie
Lake Ontario
92
65
92
319
111
98%
98%
97%
22%
31%
Benzo(a)pyrene
Lake Superior
Lake Michigan
Lake Huron
Lake Erie
Lake Ontario
72
208
290
122
155
96%
86%
80%
79%
72%
Lead2
Lake Superior
Lake Michigan
Lake Huron
Lake Erie
Lake Ontario
241
543
430
567
426
97%
99%
98%
46%
73%
Mirex
Lake Ontario
69
4.5%
1 Strachan and Eisenreich, 1988.
2 Total input is in 103 kg/yr.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-6
the atmosphere for PCBs, t-DDT, benzo(a)pyrene, lead, and mirex (Lake; Ontarici only)
(although the authors note that large uncertainties accompany the data). Table 5-3 provides a
summary of these results by lake.
5.2.3 The Green Bay Mass Balance Study
The Green Bay/Fox River Mass Balance Study (GBMBS) is an ongoing project investigating
the relationships between various PCB sources and sinks in the Fox River and Lake
MlchTgS?s Green Bay (Beltran, 1992a). Preliminary estimates (1989 data) show that point
sources contribute an estimated 9.4% of total whole bay PCB loadings. (Bierman et.al, 1992
TaWe 9-8, p. 172). PCB loadings in Green Bay are mostly attributable to the resuspension of
contaminated Fox River sediments (Beltran, 1992a).
5.2.4 Conclusions Regarding the Relative Contribution of Point Sources to Total
Loadings
Based on the information presented above, assumptions regarding the relative contribution of
poS sources to total loadings were developed for each lake (Table 5-4). These assumptions
reflect the findings of Strachan and Eisenreich (1988) that the lower Great Lakes receive a
greater percentage of total loadings from point sources than do the upper Great Lakes.
Strachar, and Eisenreich (1988) estimate the relative contribution of point sources to total
loadings of lead and PCBs in Lake Superior to range from 0.7% to 1.5% and 2.0/„ to l.Q/o
for Lake Huron.1 The assumptions also reflect the GBMBS findings and data developed by
EPA. Bierman et al. (1992) estimated that point sources contributed 9.4/o of the PCB
loadings to Lake Michigan's Green Bay. Estimates of total loadings developed by EPA imply
point sources contribute between 2% and 40% of basinwide loadings of mercury lead,
cadmium, and PCBs. The assumptions shown in Table 5-4 are used to attribute future toxics-
oriented water quality benefits to the Guidance in Chapters 6 and 7.
5.3 A SCREENING ANALYSIS OF POTENTIAL NONPOINT SOURCES OF
CONTAMINANTS IN THE BASIN
In addition to the point source discharges to the waters of the Great Lakes, there are other
potentially significant sources of the contaminants of concern m the basin. Although little
information is available on the relative contribution of different sources to total loadings of
relevant contaminants, it may be possible to rule out potential sources based on the
physical/chemical properties of the compounds and known means by which they are released
The authors suspect that their data underestimates point source loadings, however.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE, GUIDANCE > 5-7
Table 5-4
Estimated Share of Total Loadings Attributable to Poiint Sources
for the Great Lakes1
Great Lake
Lake Superior
Lake Michigan
Lake Huron
Lake Erie
Lake Ontario
Percentage of Total Loadings Attributable to
Point Sources
1-2%
5-10%
5-10%
10-15%
10-15%
1 Based on Strachan and Eisenreich (1988), Warren (1993),
and Beirman et al,
(1992).
into the environment. In this section, screening analyses are conducted in order to investigate
potential sources of contaminants in the basin (primarily air emissions) to shed light on the
potential role of the Guidance in reducing total loadings. Nonpoint sources of contaminants
including agricultural runoff and Superfund (National Priorities List) sites are also discussed.
5.3.1 Air Emissions as a Potential Source of Contaminants in the Basin
Seven contaminants associated with potentially significant reductions due to the Guidance
(chlordane, copper, dieldrin, heptachlor, mercury, and trichloroethylene (TCE)) were screened
to determine whether air emissions have the potential for rainout and deposition (i.e., whether
air emissions have the potential to be a source of the pollutant in the Great Lakes waters) or
whether degradation occurs in the atmosphere. PCBs were also investigated since they are
associated with significant problems in the basin.2 The results are summarized in Table 5-5,
and a summary of potential sources of release to air for each contaminant is provided in
Appendix C.
As shown in Table 5-5, air emissions within the Great Lakes Basin or long-range transport of
air emissions from elsewhere in the world to the basin can be ruled out as potential sources
for chlordane, dieldrin, heptachlor, and TCE. Therefore, Guidance-relaited reductions may
These seven contaminants were selected based on loadings reductions projected for the proposed
Guidance (U.S. EPA 1993).
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE
5-8
Table 5-5
Air Emissions and Long-Range Transport as a Possible Source of Pollutants
Associated with Guidance-Related Reductions
Chemical
Chlordane*
Copper
•I I Illl ••••
Dieldrin*
Heptaohlor*
Ml^_H|1i____i^a-i_'
Mercury*
issions as Source?
no
yes
no
no
yes
yes
no
Long-I
—
no
no
no
no
yes
yes
no
potentially have a significant impact on total loadings of these compounds (i.e., point sources
may be a significant percentage of total loadings).
Air emissions could be a potential source of copper, mercury, and PCBs in the basin. Mercury
and PCBs can be transported over large distances and may also be coming from outside the
basin Even if the Guidance significantly reduces point source discharges of these
contaminants local and long-range air emissions could be contributing substantially to the
total loadings in the basin. Copper, however, can only be transported over short distances.
Therefore, air emissions as a nonpoint source of copper in the basin can be evaluated based
on actual releases to air in the region. Based on 1992 Toxic Resource Inventory (TRI) data
maintained by state agencies in the eight Great Lakes states, air emissions of copper and
copper-related compounds total approximately 5 million pounds per year. This total is
approximately 1,000 times larger than the estimated loadings of copper attributable to point
sources. Thus, an appreciable share of copper loadings may result from the deposition of air
emissions.
As a result of this analysis, the local and long-range transport of air emissions of mercury and
PCBs and local air emissions of copper cannot be omitted as potentially significant
contributors to total basin loadings. Air emissions within the Great Lakes Basin or long-range
transport of air emissions into the basin can be ruled out as potential sources for chlordane,
dieldrin heptachlor, and TCE; thus, the Guidance is expected to be most effective at reducing
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-9
total loadings for these four chemicals. These results are summarized in Table 5-6. Of the
seven chemicals discussed in this section, only copper and TCE are not bioaccumulative.
Table 5-6
Summary of Results of Screening Analysis for Air Emissions
as a Potentially Significant Source of Loadings in the Great Lakes Basin
Chemical
Copper
Mercury
PCBs
Chlordane
Dieldrin
Heptachlor
TCE
Ruled Out?
no
no
no
yes
yes
yes
yes
Possibly Significant Air Source
Local1
X2
X
X
Long-range
X
X
|!
RULED OUT
1 The significance of local air emissions for mercury and PCBs relative to water point source
discharges was not investigated because long-range transport is a possibility for these
chemicals.
2 X indicates potentially significant air source of loadings to the Great Lakes.
5.3.2 Agricultural Runoff as a Potential Nonpoint Source of Guidance-Regulated
Contaminants in the Basin
Agricultural runoff might also be a significant nonpoint source of loadings in the Great Lakes
Basin because it contains pesticide residues. At least 56 million pounds of pesticides are used
annually in the United States and Canada in the Great Lakes watershed, most of which are
used on agricultural crops (U.S. GAO, 1993).3 Approximately 80% of the total amount of
agricultural pesticide is used on corn and soybean crops, and data on quantities for these
crops are available by county (U.S. GAO, 1993). Amounts of pesticides used on these crops
in counties bordering the Great Lakes are reported in Table 5-7; Canadian border counties (in
Quantities in this section are reported in pounds of active ingredients.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE » 5-10
Table 5-7
Annual Quantities of Pesticides Used on Corn and Soybean Crops
U.S. and Canadian Counties Bordering the Great Lakes
in
State/Province
Indiana
Illinois
Michigan
Minnesota
I New York
Ohio
Pennsylvania
Wisconsin
Counties Bordering Great Lakes
Lake, Laporte
Cook, Lake
Allegan, Alcona, Alger, Alpena, Antrim, Arenac, Baraga,
Bay, Benzie, Berrien, Charlevoix, Cheboygan, Chippewa,
Delta, Emmet, Gogebic, Grand Traverse, Houghton, Huron,
losco, Keweenaw, Leelanau, Luce, Mackinac, Macomb,
Manistee, Marquette, Mason, Menominee, Monroe,
Muskegon, Oceana, Ontonagon, Ottawa, Presque Isle,
Schoolcraft, St. Clair, Tuscola, Van Buren, Wayne
Cook, Lake, St. Louis
Cattaragus, Cayuga, Chautauqua, Erie, Jefferson, Monroe,
Niagara, Orleans, Oswego, Wayne
Ashtabula, Cuyahoga, Erie, Lake, Lorain, Lucas, Ottawa,
Sandusky
Erie
Ashland, Bayfield, Brown, Door, Douglas, Iron, Kenosha,
Kewaunee, Manitowoc, Marinette, Milwaukee, Oconto,
Ozaukee, Racine, Sheboygan
Total for U.S. Counties
Ontario
Durham, Elgin, Essex, Frontenac, Grey, Haldimand-Norfolk,
Huron, Lambton, Lennox & Addington, Muskoka, Niagara,
Northumberland, Parry Sound, Prince Edward, Simcoe,
Victoria, York _
Total for U.S. and Canadian Counties
Source: U.S. GAO, 1993. Data are from 1991.
Quantity of
Pesticide (Ibs/yr)
216,786
11,216
3,032,517
2,575
29,717
1,056.347
6,333,703
6,881,886
13,215:589
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-11
the province of Ontario) are also included.4 Pesticide residues from these counties are most
likely to affect Great Lakes water quality because of proximity. A total of 13 million pounds
of pesticides are used on corn and soybean crops in bordering counties, annually.
Herbicides used on corn and soybean crops account for three-quarters of the total amount of
agricultural pesticide used in the basin (U.S. GAO, 1993). The majority of herbicide used in
the basin is one of five types—atrazine, metolachlor, alachlor, cyanizirie, and pendimethalin—
none of which are of concern for the Guidance. After a preliminary review of the literature, it
appears that none of these five major herbicides metabolize into chemicals of concern for the
Guidance. Many insecticides and fungicides used in the basin also do not degrade into
chemicals of concern for the Guidance. Thus, although agriculture runoff may present site-
specific problems within the basin, the data suggest that agricultural runoff is not a significant
source of loadings of chemicals of concern for the Guidance.
5.3.3 National Priorities List Sites as a Potential Nonpoint Source of Guidance-
Regulated Contaminants in the Basin
In 1980, the Comprehensive Environmental Response, Compensation, and Liability Act
(CERCLA), also known as Superfund, was passed to identify and clean up existing toxic
waste sites. The most threatening sites, which are chosen based on extent of contamination
and scope of human health risks, are placed on the National Priorities List (NPL); there are
84 such sites in the Great Lakes Basin as shown in Table 5-8. ;
Release of hazardous substances from NPL sites could cause the contamination of resources
in the Great Lakes Basin. This contamination may be a significant nonpoint source of
chemical loadings in the rivers and lakes of the Great Lakes surface water network. However,
state agencies have virtually no data on nonpoint source loadings that originate at NPL sites.5
States may have some limited information on large accidental spills at.hazardous waste sites.
These data may be incomplete in many cases because of the underreporting of spills (Joe
Goodner, Illinois EPA, personal communication), and can only be acquired by making site-
specific requests.
Nonpoint source monitoring is not typical for NPL sites, although there may be some data on
the most important sites (Bill Bowl en, Illinois NPL site project manager, personal
communication). While analysts often attempt to attribute loadings to their sources, less
rigorous and less costly analyses (e.g., soil or well monitoring) are often performed, rather
than the actual quantification of nonpoint source discharges. Some sites are monitored only
4 These annual amounts are based on 1991 data.
5 Nonpoint source discharges are not included in TRI data maintained by state environmental
agencies.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-12
Table 5-8
Number of CERCLA Sites in the Great Lakes Basin
Ohio
•••—^—i
Pennsylvania
• —
Wisconsin
Total
Source: EPA Industrial Facilities Database
every few years, and very little seepage may be expected (Brad Bradley Illinois NPL site
project manager, personal communication). Further, no work has been done to synthesize
^formation on nonpoint source discharges in RAP areas (Greg Goudy, Sagmaw River RAP
coordinator; Calvin Rogers, Black River RAP coordinator, personal communications).
Based on the examination of available information, the assessment of nonP°|n\ so"r'e .
loadings from NPL sites appears to be very difficult. It would not be possible ^ determine
the extent of this source of loadings or its relationship to the success of the Guidance without
substantially more research. However, as these sites are currently being monitored and
remediated only limited releases of Guidance-regulated chemicals may be expected.
5.3.4 Summary of Potential Sources of Loadings
Air emissions and, to a lesser extent NPL sites, have the potential to be significant nonpoint
contributors to specific contaminant loadings, which could constrain the effectiveness of
Guidance-related point source controls for reducing total loadings of certain chemicals to the
basin Urban runoff and sediments are two other potentially significant nonpoint sources that
were not examined in this analysis due to lack of necessary data. While specific nonpoint
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-13
sources may not be very important contributors to total loadings, the combined contribution
from all nonpoint sources may be substantial in some sites, for some contaminants.
5.4 MODELING THE RELATIVE CONTRIBUTION OF POINT SOURCE
LOADINGS TO FISH TISSUE CONCENTRATIONS
This section presents the results of a modeling effort to estimate the potential reductions in
fish tissue contaminant concentrations and changes in fish consumption advisories resulting
from point source loadings reductions. The modeling was applied to PCBs in Lake Michigan's
Green Bay. A generalized Great Lakes exposure and bioaccumulation model for PCB
contamination in Green Bay was developed to:
1. Estimate changes in fish tissue concentrations resulting from reductions of
point source loadings in Green Bay for comparison with human and ecological
health thresholds.
2. Estimate the relative importance of point source loadings and sediment sources
of contaminants in determining fish tissue concentrations under model
scenarios.
PCB contamination of Green Bay was selected to evaluate the modeling approach because (1)
PCB sources (including loadings) and food web parameters were available from the GBMBS
(Bierman et al., 1992; Connolly et al., 1992); and (2) PCBs are a bioaccumulative
contaminant of concern resulting in fish consumption advisories and known toxicity to
wildlife. .Also, Green Bay may represent a "worst-case" scenario in terms of potential benefits
from point source controls because current point source loadings are estimated to contribute a
relatively small fraction of total PCB exposure to fish (< 10%).
Section 5.4.1 summarizes the technical approach used, and Section 5.4.2 presents and
interprets the results of model scenarios for Green Bay. Technical details of the model (e.g.,
model-based equations, parameters, assumptions) are presented in Appendix D.
5.4.1 Technical Approach
The model incorporated food chain and sediment pathways as well as water column exposures
because inclusion of only water column bioconcentration factors will underestimate fish tissue
concentrations (Thomann et al., 1992; Barren et al., 1994). Site-specific and literature data
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE "5-14
were used to estimate average fish tissue concentrations of PCBs for the whole bay 'The
model's performance (calibration) was evaluated by comparing model estimates ot FLBs in
fish and plankton with concentrations measured in Green Bay biota.
Modeling Framework
A simplified bioenergetics-based food web and bioconcentration model (Thomann et al,
1992- Gobas 1993) was coupled to sediment and water source compartments as shown in
Figure 5-1 7 PCB loadings to Green Bay (e.g., atmospheric deposition, point sources) were
assumed to be distributed between sediments and the water column, based on equilibrium
partitioning (see Appendix D). PCBs in fish tissue were assumed to be in equilibrium with
PCBs in water and sediment.
Application
The model was used to evaluate site-specific point source loadings and reduction scenarios,
with model parameters obtained from literature sources and available data for PCBs m oreen
Bav fe g Connolly et al., 1992). Sediment concentrations were obtained from whole bay
averages estimated in the GBMBS (Bierman et al, 1992; DePinto, 1994). The uncertainty
(specifically, variability) associated with estimates of fish tissue concentrations and he
associated probabilities of exceeding human and ecological health thresholds were also
quantified.
Human and Ecological Thresholds
The model-estimated fish tissue concentrations were compared to human health and ecological
thresholds for PCBs. The human health threshold selected was the fish consumption advisory
for PCBs of 2 ppm (2000 ug/kg) in muscle fillets. The ecological health threshold selected
was 0.3 ppm, which was a modification of the International Joint Commission (1989)
threshold value of 0.1 ppm.8
6 Measured fish tissue concentrations in Green Bay generally decrease between the inner and outer
bay (Connolly et al., 1992).
7 The model is an equilibrium model, and thus assumes a steady-state distribution of contaminants
in the system. The model does not include dynamic, spatial, or temporal heterogeneities in contaminant
concentrations. Dynamic models such as the Green Bay mass balance model are required to characterize
contaminant dynamics under nonequilibrium conditions.
8 Our threshold assumes that one-third of the diet of wildlife is contaminated fish; the IJC threshold
is protective if 100% of the diet of wildlife is contaminated fish.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE »5-15
Figure 5-1
Schematic Representation of Great Lakes Exposure and Food Web Model
Piscivores
(e.g., walleye, brown trout)
I
0.35
Omnivores
(e.g., perch, smelt)
A A A
).5o| 0.15 0.35
Benthivores
(e.g., carp, catfish)
0.85
Benthic Invertebrate
. *
0.05 10.1
Sediment
0.65
Planktivores
(e.g., alewife, gizzard shad
i.b
Plankton
Water
Note: Numbers are dietary fractions consumed by biota. Arrows without numbers designate equilibrium
partitioning.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE » 5-16
Limitations and Caveats
The modeling effort was not intended to duplicate the mass balance or food chain modeling
performed for the Fox River/Green Bay system as part of the GBMBS (Bierman et al, 1992;
Connolly et al 1992) The GBMBS was a comprehensive evaluation of sources, storage,
transport, and bioaccumulation in Green Bay, and cost an estimated $12 million (DePmto,
1994) Costs associated with mass balance modeling for the GBMBS exceeded an estimated
$2 million (DePinto, 1994).9 Rather, the objective was to develop a generalized Great Lakes
model that relied on equilibrium (steady-state) assumptions, which could be used to estimate
changes in fish tissue concentrations in response to a reduction in loadings, using the
available data for Green Bay in an initial application of the model. In addition, the model
incorporates the inherent variability of estimated fish tissue concentrations, which allowed a
quantitative assessment of the probability of exceeding human and ecological health
thresholds.
5.4.2 Results and Discussion
Available measured concentrations of PCBs in Green Bay biota (Beltran, 1992b) were
compared to model-estimated tissue concentrations. Measured and predicted concentrations
overlapped for each of the biota categories of the Green Bay food web (see Figure 5-2). Thus,
the model successfully predicted measured PCB concentrations in Green Bay.
Point Source Loadings Reductions
An estimated 9.4% of the 1989 PCB loadings to Green Bay were from point source loadings
(Bierman et al, 1992; Table 9-8; p. 172). For modeling scenarios it was assumed that, at
baseline, point source loadings contributed 9.4% of the PCB exposure to biota in Green Bay
(no loadings reductions). Model-estimated PCB concentrations (baseline conditions) in
omnivores (consumed by wildlife) and in fillets of piscivores (consumed by humans) are
presented in Figure 5-3 and Figure 5-4, respectively.
To estimate changes in fish tissue concentrations in response to a reduction in loadings,
scenarios of a 10%, 50%, and 90% reduction in point source loadings were modeled.
Table 5-9 presents the mean estimated tissue concentrations of PCBs. Model-estimated fish
tissue concentrations of PCBs exhibited only limited decreases in response to reductions of
loadings in Green Bay.
9 One of the principal advantages of a mass balance model is the capability to define source
attribution and evaluate changes in individual source loadings (Beltran, 1992). Disadvantages of mass
balance models include extensive data requirements (Beltran, 1992).
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-17
Figure 5-2
Model-Estimated Probability Distribution of Tissue PCB Concentrations
Probability
Probability
Probability
307.
Walleve. Brown Trout
24% Green Bay measured
185S values: 0.5 to 6.5
12* "" Model Expected
Value: 4.9
65! "'
|
0 3.125 6.25 9.375 12.5 15.625 18.75 21.875 25
Tissue Concentration (ppm)
2TO
Gizzard Shad. Alewife
Rainbow Smelt
Green Bay measured
12% • m values: 0.1 to 2.1
I
• ••• Model Expected
I |||H 1 : Value: 2.4
Illlliil-
0 1.25 2.5 3.75 5 6.25 7.5 S.75 10
Tissue Concentration (ppm)
2or.J ] "
Plankton
16r° "• Green Bay measured
„_, ___ -, values: 0.07 to 0.6
' HHH
Model Expected
Value: 0.27
or.
.125 .25 .375 .5 .625 .75 .875 1
Tissue Concentration (ppm)
Note: Distributions represent probability of a specific tissue concentration; probabilities sum to 100%.
Green Bay measured values are from Beltran (1992b).
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-18
Figure 5-3
Model-Estimated Probability Distribution of Baseline PCB Concentrations in Omnivores
(No point source reductions)
Probability
20%
16%-
Ecological Health Threshold = 0.3 ppm
Fish Tissue Concentration (ppm)
Note: Distribution represents probability of a specific fish tissue concentration; probabilities sum to
100%. Concentrations in mg/kg (whole body). Omnivores are consumed by wildlife and include
rainbow smelt and alewives.
The model-estimated fish tissue concentrations were compared to selected human and
ecological health thresholds for PCBs. As shown in Table 5-10, reductions of 10%, 50%, and
90% in point source loadings had no effect on exceedences of the ecological threshold
because of the very low threshold value relative to fish tissue PCB concentrations. Reductions
of 10% 50% and 90% in point source loadings reduced exceedences of the human health
threshold but the effect was minimal partly because the baseline exceedences are minimal
(i e 10 6%) (Table 5-10). Because of the limited reductions in fish tissue concentrations only
the'data for the 50% reduction in point source loadings are presented for fillets (Figure 5-5).
Comparing Figure 5-4 (baseline) to Figure 5-5 (50% point source reduction), and as
summarized in Table 5-10, a 50% reduction of PCBs from point sources decreases the
probability offish tissue exceeding the human health criteria from 10.6% to 8.9%. Therefore,
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE >• 5-19
Figure 5-4
Model-Estimated Probability Distribution of Baseline PCB Concentrations
in Piscivore Fillets (No point source reductions)
Probability
20%"
16%—
12%
8% -
4% —
0%
Human Health Thrsshold = 2.0 ppm
0
.625 1.25 1.875 2.5 3.125 3.75 4.375 5
Fish Fillet Concentration (ppm)
Note: Distribution represents probability of a specific fish fillet concentration; probabilities sum to 100%.
• Concentrations in mg/kg (fillet). Piscivores are consumed by humans and include walleye and
brown trout.
the model shows PCB concentrations in Green Bay are not particularly responsive to
reductions in point source loadings. This result for Green Bay arises because existing
sediment contamination is the dominant source of exposure to biota. Historical point source
loadings likely contributed to current sediment contamination. However, at other sites, or for
other contaminants, point source reductions could have a more appreciable affect.
Sediment Versus Water Column Exposures
In the environment, PCBs in sediment partition into water, resulting in both water column and
food chain exposures to biota. To evaluate the importance of sediment contamination,
sediment-plus-water and sediment-only exposures were simulated. Sediment contributed an
estimated 0% to 100% of the PCBs bioaccumulated in Green Bay biota, depending on biota
feeding habits (e.g., benthivore versus planktivore) (see Table 5-11).
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-20
Table 5-9
Model-Estimated Fish Tissue Concentrations1
Point Source Loading Scenario
.«•••••—•"•••
Baseline
50% Reduction
————^——
90% Reduction
Omnivores2
(ppm in whole body)
Piscivores3
(ppm in fillet)
Concentrations were rounded to the nearest 0.1 ppm.
Omnivores (e.g., alewife, rainbow smelt) represent the fish consumed by wildlife.
Piscivores (e.g., walleye, brown trout) represent the fish consumed by humans.
Table 5-10
Model-Estimated Threshold Exceedences1
Point Source Loading Scenario
Baseline
,«••••—^—^—
10% Reduction
-
50% Reduction
• -i •— ' '"
Ecological Health2
99.9%
99.8%
Human Health3
10.6%
10.4%
1 Concentrations were rounded to the nearest 0.1%.
2 Threshold: 0.3 ppm in omnivore whole tissue.
3 Threshold: 2 ppm in piscivore fillet (e.g., walleye, brown trout).
—^ i.»^^^— i • —-"^~"^^^^^^^si^^s ••!
5.4.3 Summary of Model and Results
The generalized Great Lakes exposure and bioaccumulation model successfully predicted PCB
concentrations in the Green Bay food web. The model incorporated the inherent variability of
estimated fish tissue concentrations, which allowed visualization of the uncertainty in the
model estimates. The advantage of this approach over the development of single point
estimates is that it also provides a quantitative assessment of the probability of exceeding
human health and ecological thresholds.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE »• 5-21
Figure 5-5
Model-Estimated Probability Distribution of PCB Concentrations in Piscivore Fillets
under a Scenario of a 50% Reduction in Point Source Loadings
20%
Human Health Threshold = 2.0 pom
0%
.625 1.25 1.875 2.5 3.125 ;3.75 4.375 5
Fish Fillet Concentration (ppm)
Note: Distribution represents probability of a specific fish fillet concentration; probabilities sum to 100%.
Concentrations in mg/kg (fillet). Piscivores are consumed by humans and include walleye and
brown trout.
The model was used to estimate changes in fish tissue concentrations and exceedences of
human and ecological health thresholds under different scenarios of point source loadings
reductions. Reductions in point source loadings in Green Bay were shown to have a modest
impact on fish tissue PCB concentrations and exceedences. For example, as shown in
Table 5-10, a 50% reduction in point source loadings reduces baseline exceedences of the
human health threshold from 10.6% to 8.9%. This result occurs because for this contaminant
and this site, existing sediment contamination is the dominant source of PCB exposure to fish
(point sources contribute only 9.4% to loadings), and baseline exceedences are only 10.6%.
However, loading reduction estimates for the Fox River/Green Bay case study area indicate
that the Guidance may reduce point source PCB loadings by 89.6% (see Chapter 7). As
shown in Table 5-10, a 90% reduction lowers exceedences of the human health threshold
from 10.6% to 6.8%.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE
5-22
Table 5-11
Model-Estimated Fish Tissue PCB Concentrations (ppm)
% from Sediments1
100
Sediment and Water
* ' 4
Percentage of ti sue PCBs from sediment (excludes PCBs in sediment which have partitioned
Figure 5-6 shows the model-estimated change in exceedences of the human health threshold
of 2 com (relative percentage reductions) in response to pomt source loading reductions^
pSSe leductiL in exceedences from baseline are 2%, 16%, and 6% under
os of 10% 50%, and 90% reductions in point source loadings, respective y (in
5-6 point B, A and C, respectively). Application to other sites and conditions my
greater benefits from point source reductions (i.e., Green Bay is considered a wont
a™ fc nario). For example, sites where point source loadings represent a greater percent of
total loadings are expected to show a greater change from baseline conditions. This s
SStSS^movuJto the right along the horizontal axis in Figure 5-6/or example, under
a scenario of a 50% reduction in point source loadings, the estimated reduction in
exceedences may increase from about 16% (point A, with point sources contributing 9 A /. of
total loadings) to about 35% (a point between A and D, where point sources contribute 20/o
of total loadings).
5.5 COST-EFFECTIVENESS OF POINT SOURCE CONTROLS
Determining whether point source controls are a "cost-effective-approach for reducing BCC
concentrations in the Great Lakes depends in part on the cost and **{*
alternative methods for reducing contaminant concentrations. The method^1J"
sediment sources of PCBs in the Great Lakes were investigated (RCG/Hagler
1995) because of the environmental persistence and widespread contamination of PCBs in
sediment (U.S. EPA, 1994b), and because releases of PCBs from sediments te^he Great
Lakes are a significant continuing source of exposure to Great Lakes biota (RCG/Hagler
Bailly, Inc., 1995).
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE •• 5-23
Figure 5-6
Relationship Between Reductions in Exceedences of Fish Consumption Advisories
for PCBs and Percent Contribution of Point Source Loadings to Total Loadings
100%
in point Source Loadings
10% 20% 30% 40%
Percent Contribution of Point Sources to Total Loadings
50%
Note: Results are model-estimated changes (reductions) in exceedences of 2 ppm PCBs in fillets under
10%, 50% or 90% reductions in point source loadings. Reductions increase as the percent of point
source loadings to total loadings increase. Green Bay point source loadings were modeled as 9.4%
of total loadings.
Numerous uncertainties surround the estimation of costs and effectiveness of options for
reducing sediment-related exposures to PCBs, including uncertainty in 'the quantity of PCB-
contaminated sediment; the contribution of site-specific contamination to human and wildlife
PCB exposures; and site- and technology-specific costs (and effectiveness) of sediment
remediation options. For example, estimating PCB reduction costs in sediments is uncertain
because there are only limited data available describing the total quantity of contaminated
sediments within the Great Lakes. Even some Areas of Concern (AOC) within the basin
(areas identified as having impaired water and sediment quality) lack estimates of the total
volume of contaminated sediments (U.S. EPA, 1994b). i
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE > 5-24
Sediment Reduction Methods
PCB concentrations in sediments may be reduced by in situ or ex situ remediation processes.
In situ treatments include no action and capping (U.S. EPA, 1993). Ex situ treatments involve
removal and subsequent treatment of contaminated sediments. The major advantage of in situ
treatment is that there is no need to remove contaminated sediments. The only sediment
removal option currently in use is dredging (U.S. EPA, 1993; 1994c). Dredged sediments
contaminated with PCBs must be treated or contained. Current available processes include
confined disposal facility (CDF) storage, incineration, biological treatment KPEG
dechlorination (a dehalogenation process), solvent extraction, and soil washing (Davila, 1993;
U S EPA, 1993; 1994c). Water removed from contaminated sediments may also require
treatment (U.S. EPA, 1993; Davila, 1993).
Selection of the remediation scenario for contaminated sediments is dependent on site-specific
conditions such as the extent of contamination (i.e. depth, area, concentration) and the
potential for disturbance or re-release of PCBs. Costs vary substantially by treatment option,
• and effectiveness will depend on the site-specific contribution of sediments to human and
wildlife exposures. Thus, calculation of cost-effectiveness is not possible on a lake or basin-
wide basis Calculation of cost-effectiveness for specific case studies will require apnori
knowledge of the appropriate PCB reduction technique and site-specific exposure information
(e.g. contribution of site-specific sediment contamination to total risk).
Comparison of Sediment Removal Costs with Costs of Point Source Controls
The estimated costs of reducing PCBs in sediment range from a low of $65,000 to $290,000
per hecrare (dredging plus CDF disposal), to a moderate range of $430,000 to $650,000 per
hectare (in situ sediment capping), to a high of $1 million to $6.5 million per hectare
(dredging plus ex situ treatment.) These costs may be conservative because dredging to a
depth of 0.5 meters is assumed, whereas contaminated sediments may be present at greater
depths. For example, various sediment remediation strategies are being considered for the
Buffalo River AOC, including dredging to a 3 meter depth (U.S.EPA, 1994c).
In comparison, the annual cost of the final Guidance to control a range of toxic contaminants
is estimated to be between $61 million to $376 million (see Chapter 4). Redirecting the
estimated costs of the Guidance for sediment clean-up might remediate from less than 10
hectares (using the highest cleanup cost and the lowest Guidance cost) to up to as many as
5 800 hectares (using the lowest cleanup cost and the highest Guidance cost). To place this in
perspective, consider that one AOC, the Saginaw River/Saginaw Bay AOC, has a surface area
of approximately 300,000 hectares (Bolattino, 1993).
The actual surface area of PCB contaminated sediments in the Great Lakes is unknown. And,
although contaminated sediments contribute substantially to the PCB exposure to humans and
Great Lakes biota, reduction of sediment PCBs may not be entirely feasible or cost-effective.
-------
ANALYSES RELATED TO THE ATTRIBUTION OF BENEFITS TO THE GUIDANCE »• 5-25
For example, estimated remediation costs for a portion (1 million cubic yards) of the Grand
Calumet River/Indiana Harbor Canal AOC is estimated to be approximately $40 million
(Bolattino, 1993). Therefore, although sediment may contribute a larger share of current
contamination to Great Lakes biota than do point sources, sediment remediation alone may
not be a cost-effective approach to address the problem. In addition, if point sources are not
controlled, remediated sediments may be recontaminated by point source discharges.
-------
-------
CHAPTER 6
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS
The April 1993 RIA contained a preliminary assessment of health-related risks to Great Lakes
Basin sport anglers and potential risk reductions resulting from the proposed Guidance, based
largely on U.S. EPA's 1991 review draft Great Lakes Basin Risk Characterization Study. The
preliminary assessment evaluated carcinogenic and noncarcinogenic (systemic) risks due to
PCB, DDT, mercury, and dieldrin exposures.
In 1994, Executive Order (EO) 12898 was issued. EO 12898 established a presidential policy
for incorporating environmental justice into Federal agency missions by directing agencies to
identify and address, as appropriate, disproportionately high and adverse human health or
environmental effects of its programs, policies, and activities on minority populations and
low-income populations. To assist in identifying the need for ensuring protection of
populations who principally rely on fish and/or wildlife for subsistence, the EO directs
agencies, whenever practicable and appropriate, to collect, maintain, and analyze information
on the consumption patterns of those populations and to communicate to the public the risks
of those consumption patterns.
In accordance with EO 12898, additional data and information were collected on the
consumption of Great Lakes Basin fish to refine the sport angler risk assessment to reflect
minority and low-income exposures and to conduct a separate assessment foi Native
Americans engaged in subsistence fishing in the basin. Since Native Americans fishing on
reservations and treaty-ceded fishing grounds are not required to purchase fishing licenses,
they would not be accounted for in the sport angler assessment. Fish tissue concentrations for
additional pollutants were also incorporated. In addition to the chemicals listed above, risks
were addressed for chlordane, hexachlorobenzene, 2,3,7,8-TCDD, and toxaphene. Finally,
potential reductions in baseline risks were revised to reflect the revised estimates of loading
reductions due to the Guidance.
This chapter is organized as follows. Section 6.1 presents the sport angler risk assessment.
Section 6.2 presents the risk assessment for Native Americans engaged in subsistence fishing
in the basin. Section 6.3 provides a summary of the combined risk assessment results.
Appendix E provides background information on the calculation of cancer and systemic risks,
including a description of the carcinogenic and systemic effects associated with the relevant
contaminants.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS » 6-2
6.1 GREAT LAKES SPORT ANGLER RISK ASSESSMENT
Data on fishing license sales in the Great Lakes Basin were used to estimate the number of
potentially exposed recreational anglers. Data from 1991 through 1993 (the most recent sale
S available for each state) indicate approximately 2.69 million fishing licenses were sold
o residents in Great Lakes Basin counties.1 A list of counties in tile basin is provided m
Table 6-1 To the extent that anglers purchase their licenses outside the basin or share their
catch with unlicensed family members, the use of fishing license data may result m an
underestimate of the potentially exposed population.
Sport Angler Fish Consumption
Several studies of licensed angler consumption patterns have been conducted in the Great
Lakes Basin. These studies, described in Table 6-2, provide results for a range of fishing
locations and reflect water quality conditions that often include the presence of fish
consumption advisories. Nonetheless, fish consumption is shown to be significant.
The study results also indicate a relationship between consumption levels and socioeconomic
characteristics. As shown in Table 6-2, minorities are estimated to have higher fish
consumption than whites in the basin. Work by West et*l. (1993) shows that the combination
rfSrity status and relatively low income (annual income less than $25,000) results m
higher consumption levels-the highest level estimated for sport anglers m the basin.
Based on the research by West et al. (1993), consumption scenarios were developed for
various subgroups of sport anglers in the basin. In counties in closest proximity to the Great
Lakes (the contiguous lakeshore counties), licensed anglers were divided into minorities with
relatively low income (those earning less than $25,000 per year), other minorities, and
everybody else This division was made using county-level census data and applying the
percentage of each group's population in the county to the total number of licensed anglers m
L county Minorities with an income of less than $25,000 per year in the lakeshore counties
were assumed to have a fish intake of 43.1 gpd; the remainder of licensed minorities in these
counties were assumed to consume 11.1 gpd. For the non-minority angler population m the
lakeshore counties and in the remaining counties of the basin, fish intake was assumed to be
16.7 gpd.
EPA mapped the location of low-income minorities and licensed angler populations in relation
to the location of point source dischargers in the basin. As shown in Figures 6-1 and 6-2, the
greatest density of low-income minorities and licensed anglers are located near the large
population centers (and highest density of point sources) bordering the Great Lakes.
1 Two states sell resident husband and wife fishing licenses. These licenses were counted twice in
estimating the exposed population.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES .ANGLERS >• 6-3
Illinois:
Cook
Indiana:
Adams
Allen
DeKalb
Elkhart
Michigan:
Alcona
Alger
Allegan
Alpena
Antrim
Arenac
Baraga
Barry
Bay
Benzie
Berrien
Branch
Calhoun
Cass
Charlevoix
Cheboygan
Chippewa
Clare
Clinton
Crawford
Delta
Dickinson
Eaton
Emmet
Genesee
Gladwin
Gogebic
Grand Traverse
Minnesota:
Carlton
Cook
Table 6-1
Great Lakes Basin Counties
Lake
Kosciusko
La Porte
Lagrange
Lake
Gratiot
Hillsdale
Houghton
Huron
Ingham
Ionia
losco
Iron
Isabella
Jackson
Kalamazoo
Kalkaska
Kent
Keweenaw
Lake
Lapeer
Leelanau
Lenawee
Livingston
Luce
Mackinac
Macomb
Manistee
Marquette
Mason
Mecosta
Menominee
Midland
Lake
•
'
Noble
Porter
St. Josieph
Steuben
Missaukee
Monroe
Montcalm
Montmorency
Muskegon
Newaygo
Oakland
Oceana
Ogemaw
Ontonagon
Osceola
Oscodia
Otsego
Ottawa
Presque Isle
Roscommon
Saginaw
St. Clair
St. Joseph
Sanilao
Schoolcraft
Shiawassee
Tuscola
Van Buren
Washtenaw
Wayne
Wexford
>'
St. Louis
-------
New York:
Allegany
Cattaraugus
Cayuga
Chautauqua
Clinton
Erie
Franklin
1 Genesee
Herkimer
Jefferson
Ohio*
Allen
Ashland
1 Ashtabula
11 Auglaize
1 Cra\vford
Cuyahoga
Defiance
II Erie
Fulton
1 Geauga
1 Hancock
1 Pennsylvania:
| Crawford
1 Wisconsin:
II Adams
Ashland
Bayfield
11 Brown
I Calumet
Columbia
Door
Douglas
Florence
Fond du Lac
Forest
Green Lake
Lewis
Livingston
Madison
Monroe
Niagara
Oneida
Onondaga
Ontario
Orleans
Hardin
Henry
Huron
Lake
Lorain
Lucas
Marion
Medina
Mercer
Ottawa
Paulding
Erie
Iron
Kenosha
Kewaunee
Langlade
Manitowoc
Marathon
Marinette
Marquette
Menominee
Milwaukee
Oconto
Oneida
. •
Oswego
St. Lawrence
Schuyler
Seneca
Steuben
Tompkins
Wayne
Wyoming
Yates
Portage
Putnam
Richland
Sandusky
Seneca
Summit
Trumbull
Van Wert
Williams
Wood
Wvandot
Potter
Outagamie
Ozaukee
Portage
Racine
Shawano
Sheboygan
Vilas
Washington
Waukesha
Waupaca
Waushara
Winnebago
Sources: USGS, 1974.
USGS, 1988.
Great Lakes National Program Office, 1994.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS » 6-5
Table 6-2
Characterization of Great Lakes Sport Angler Fish Consumption
Study
Description
Findings
1991-92 Michigan Sport
Anglers Fish Consumption
Study (West et aL, 1993)1
Year-round mail survey of
randomly selected Michigan
licensed anglers (sample size
7,000; response rate 46.3%) using
a 7-day recall period. Minorities
were not disaggregated because
of small sample sizes, but the
two main groups were blacks and
non-reservation Native
Americans.
Average fish consumption
(grams/da]/):2
Licensed anglers
16.7 (sport-caught)
26.5 (sport and commercial)
Minorities; earning S25,000/year or
more
11.1 (sport-caught)
22.9 (sport and commercial)
Minorities! earning less than
S25,000/year
43.1 (sport-caught)
57.9 (sport and commercial)
1990-91 New York State
Angler Cohort Study
(Vena, 1992).
Survey of NY resident fishing
license holders in their
reproductive years (ages 18-40)
from 16 counties bordering Lake
Ontario, the Niagara River, and
the St. Lawrence River. Sample
was constructed to identify
anglers with a high probability of
consuming fish from Lake
Ontario; 11,717 surveys were
returned (a response rate of
40%).
Race was the strongest demographic
determinant of consumption of Lake
Ontario fish:
Frequency of consuming at least 1
meal per week
Whites 8%
African Americans 26%
Native Americans 19%
Frequency of consuming 10 or
more meals per month
Whites 1%
African Anericans 11%
Native Americans 5%
Anglers with higher frequency of
consuming sport-caught fish also
consumed larger portions.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS » 6-6
Table 6-2 (cont.)
Characterization of Great Lakes Sport Angler Fish Consumption
Study
Description
Findings
Sport Fish Consumption
and Body Burden Levels
of Chlorinated
Hydrocarbons: A Study of
Wisconsin Anglers (Fiore
et al., 1989)
Survey of Wisconsin residents
from 10 counties in proximity to
waters included in the Wisconsin
fish consumption advisory
holding fishing licenses in 1984.
801 surveys were returned
(sample size 1,600; 50% response
rate).
Average number of meals in 1984
18 meals (sport-caught)
41 meals (sport and commercial)
Average daily consumption
(assuming average meal is 8 oz.
and removing those who
consumed no sport-caught fish)
12.3 grams/day (sport-caught)
26.1 grams/day (sport and
commercial)
Minority Anglers and
Toxic Fish Consumption:
Evidence from a
Statewide Survey of
Michigan (West et al.,
1989)
Mail survey of a stratified sample
of Michigan fishing license
holders in 1988 (sample size
2,600; response rate 47.3%).
Winter-spring consumption:
»• Minorities had higher
consumption than whites
(differences termed marginally
insignificant).
> Older anglers had higher
consumption. This was the
highest single variable result and
differences by age were
statistically significant.
*• Higher consumption by blacks
was in cities; high consumption
by Native Americans occurred in
small towns and rural areas.
»• Income did not have a
statistically significant bivariate
relationship with consumption.
Results were verified and new analyses performed using imputed income values to recapture
individuals (Jacobs, 1994). Estimates using the imputed income values were similar to the 1993
results, but not computed for all groups (minorities earning $25,000 per year or more). Thus, we
used the 1993 estimates.
Results not adjusted for nonresponse bias. An earlier survey by West et al. (1989) showed
nonrespondents consumed less fish, but only slightly less (2.2 grams/person/day for the means);
nonresponse bias was not tested for different socio-economic groups.
-------
_e
*S?
ffl
o>
ty
O
3
O
GO
_o
O
Crt
CJ
O
c
4)
s
O
I
O
a
5
oc
-------
-------
-------
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS >• 6-9
Fish Tissue Contaminant Levels
1
Health risks were calculated based on exposure to chlordane, DDT, dieldrin,
hexachlorobenzene, mercury, PCBs, 2,3,7,8-TCDD, and toxaphene. These chemicals were
chosen based on their potential to cause adverse human health effects (i.e., cancer or disease)
and the availability of information on fish tissue contaminant concentrations. Chemicals were
included in the risk assessment only if fish tissue concentration data v/ere available. Table 6-3
shows the assumptions regarding chemical toxicity and concentrations in fish tissue.
Fish tissue contaminant levels were estimated by lake, based on data from several sources
(Devault, 1983; International Joint Commission (IJC), 1989; Amrhein., 1990; Gooch et al.,
1990; U.S. EPA, 1992a; Miller et al., 1992; D. DeVault, Great Lakes National Program
Office, U.S. EPA, personal communication, 1994; and data provided by the Great Lakes
Indian Fish and Wildlife Commission (GLIFWC)). Data were lake-specific except for those
from the IJC.
Exposure Assumptions
Exposure was calculated based on the assumption that each fish contained all contaminants
listed at the concentrations reported in Table 6-3. Risks were estimated separately for each
lake to more accurately match fish tissue concentrations with the exposed population. To the
extent that not all fish contain all contaminants at the assumed concentrations, exposures and
risks may be overestimated. However, data were only available for a small portion of the
contaminants covered by the Guidance, which could result in the underestimation of risks.
Human health risks were calculated based on the consumption scenarios for low-income
minority sport fishermen residing in lake-shore counties, other minority sport fishermen in
lake-shore counties, and all other sport fishermen. These fish consumption values were given
as daily intake estimates, although in actuality, fewer, larger meals may be consumed.
Standard EPA assumptions were used regarding length of residence (Le., 70 years; by
convention) and body weight (70 kilograms) (U.S. EPA, 1989). Table 6-4 reports exposure
assumptions used to determine human health risks.
Baseline Human Health Risks
Potential adverse human health effects resulting from the consumption of contaminated fish
include both the increased risk of cancer and the potential for systemic or noncancer risks
such as kidney damage. As shown in Table 6-5, baseline lifetime cancer risks for low-income
minorities ranged from 2.5 x 10'3 (Lake Superior) to 1.2 x 10'2 (Lake Michigan); for other
minorities, baseline cancer risks ranged from 6.5 x iQ"4 (Lake Superior) to 3.0 x 10"3 (Lake
Michigan); and for all other sport fishermen, these risks ranged from 9.7 x IQ'4 (Lake
Superior) to 4.5 x 10'3 (Lake Michigan). For all lakes combined, at baseline fish contaminant
-------
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-11
Table 6-4
Exposure Assumptions Used to Calculate Risks to Sport Fishermen
Exposed Population
Lake Erie
Lake Huron
Lake Michigan
Lake Ontario
Lake Superior
Fish Intake (grams per day)
Exposure Duration (years)
Exposure Frequency (days/year)
Body Weight (kg)
Low-Income
Minority, Lakeshore
39,610
2,035
42,853
4,691
2,630
43.1
70
365
70
Other Minority,
Lakeshore
25,197
1,831
33,900
3,307
1,262
11.1
70
365
70
All Other Sport
Fishermen
693,023
436,784
980,011
164,580
258,343
16.7
70
365
70
1 Represents average fish consumption based on West et al., 1993.
concentrations, the population of low-income minorities would have an estimated
10.1 potential cancer cases per year, other populations of minorities would have an estimated
1.9 potential cancer cases per year, and for other sport fishermen, 100.5 potential cancer cases
per year are estimated. As shown in Table 6-6, baseline cancer risks are driven by fish tissue
PCB concentrations. PCB concentrations in fish tissue are highest in Lake Michigan and
lowest in Lake Superior.
Systemic (noncancer) risks such as kidney damage are assumed to be additive, and are
assessed by means of a Hazard Index (HI). The HI is a compilation of Hazard Quotients (HQ)
for each contaminant, and is described in Appendix E. The HQ is calculated by dividing the
expected exposure level (dose) by the EPA oral reference dose (RfDs), where the RfD
indicates the level of chronic exposure below which no adverse health effects are expected.
Therefore, a HI of 1.0 or more, implies that chemical exposures exceed EPA-established
"thresholds" of toxicity, and is indicative of the potential for adverse health effects to occur.
The potential for detrimental health effects increases as the HI increases above 1.0. As shown
in Table 6-7, baseline HI values for low-income minorities living in lakeshore counties range
from 32.0 (Lake Superior) to 183.0 (Lake Michigan). For other lakeshore county minorities,
baseline HI values range from 8.3 (Lake Superior) to 47.0 (Lake Michigan). Baseline HI
values for all other sport fishermen range from 12.4 (Lake Superior) to 70.7 (Lake Michigan).
As all of these baseline values exceed 1.0, they are all indicative of a. high potential for
systemic injury. The greatest risk of systemic injury also results from PCB exposure. The
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS » 6-12
Table 6-5
Baseline Potential Excess Cancer Cases for Sport Fishermen
Individual Excess
Lifetime Risk Level
Yearly Cases
Lake Erie
Low Income Minorities1
Other Minorities1
Other Sportfishermen
Total
3.7 x 1Q-3
9.6 x 1Q-4
1.4 x lO'3
Lake Huron
Low Income Minorities1
Other Minorities1
Other Sportfishermen
Total
5.5 x lO'3
1.4 x lO'3
2.1 x TO'3
2.1
0.3
14.0
16.4
0.2
0.0
13.0
13.2
Lake Michigan
Low Income Minorities1
Other Minorities1
Other Sportfishermen
Total
1.2 x 10'2
3.0 x lO'3
4.5 x 10'3
7.2
1.5
63.0
71.7
Lake Ontario
Low Income Minorities1
Other Minorities1
Other Sportfishermen
Total
7.5 x 10'3
1.9 x lO'3
2.9 x 10'3
0.5
0.1
6.9
7.5
Lake Superior
Low Income Minorities1
Other Minorities1
Other Sportfishennen
Total
2.5 x 10'3
6.5 x 1Q-4
9.7 x 1Q-4
0.1
0.0
3.6
3.7
Lifetime
Cases
147
24
996
1,167
11
3
936
950
500
102
4,432
5,034
35
6
479
520
7
1
251
259
All Lakes
Total Low Income Minorities
Total Other Minorities
Total Other Sportfishermen
Grand Total All Lakes
10.1
1.9
100.5
112.5
700
136
7,094
7,930
1 Lakeshore counties only.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS + 6-13
1
Table 6-6
Percentage of Each Compound's Contribution to the Total Carcinogenic Risk
Compound
Chlordane
DDT
Dieldrin
HCB
Mercury
PCBs
TCDD
Toxaphene
Total
Lake Michigan
1.8
0.8
' 8.3
0.0
0.0
81.2
2.2
5.5
100.0
Lake Superior
0.6
2.2
7.9
0.0
0.0
64.3
7.5
17.6
100.0
Lake Huron
1.6
1.0
7.1
0.0
0.0
64.3
22.0
4.1
100.0
Lake Ontario
0.6
1.4
6.6
0.1
0.0
69.4
22.0
0.0
100.0
Lake Erie
0.2
0.5
2.6
0.0
0.0
87.0
9.8
0.0
100.0
systemic adverse health effects associated with the assessed contaminants are described in
Appendix E. i
Potential Risk Reductions Due to the Guidance
To determine the potential reduction in human health risks that may result from the Guidance,
it is necessary to know the reduction in loadings due to the Guidance and the effect this
loadings reduction will have on fish tissue contamination. The reduction offish tissue
contamination attributable to the Guidance will depend on the percentage of fish tissue
contamination due to point source loadings versus other sources such as atmospheric
deposition or sediment resuspension.
As described in Chapter 5, assumptions regarding the relative contribution of point sources to
total loadings were developed for each lake. These percentages were multiplied by the
estimated basinwide reduction in point source loadings anticipated to result from the Guidance
for each contaminant to estimate potential reductions in fish tissue concentrations for each
lake (Table 6-8).2 The estimated reductions in fish tissue contaminant concentrations were
then used to determine the potential reductions in human health risks. Reductions in potential
Loadings reductions reflect the low cost (and loadings reduction) scenario. However, for the
relevant pollutants, increased reductions were-projected under the high end scenario only for mercury.
Mercury loadings are estimated to be reduced by 12.1% under the low end scenario and by 13.8% under
the high end scenario. This change was considered insignificant for the assessment of risk reductions under
the Guidance.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-14
Table 6-7
Baseline Systemic Risks for Sport Fishermen
Hazard Index (HI)
Lake Erie
Low Income Minorities1
62.6
Other Minorities1
16.1
Other Sportfishermen
24.2
Lake Huron
Low Income Minorities1
75.7
Other Minorities1
19.5
Other Sportfishermen
29.3
Lake Michigan
Low Income Minorities1
183.0
Other Minorities1
47.0
Other Sportfishermen
70.7
Lake Ontario
Low Income Minorities1
110.0
Other Minorities1
28.2
Other Sportfishermen
42.5
Lake Superior
Low Income Minorities1
32.0
Other Minorities1
8.3
Other Sportfishermen
12.4
Lakeshore counties only.
excess cancer cases per year are reported in Table 6-9; Table 6-10 reports potential reductions
in systemic risks.
The estimated reductions in human health risks due to the Guidance are small. This is
especially true for systemic risk reductions, which are negligible. These modest risk
reductions are due to the fact that PCB contamination comprises the majority of both systemic
and carcinogenic risks, and the modeled basinwide loadings presented in Chapter 4 do not
indicate reductions in PCBs as a result of the Guidance. However, the basinwide results are
based on a sample of 59 facilities in the basin, and may result in conservative estimates of
actual basinwide reductions. Indeed, estimated loadings reductions for the Fox River and
Green Bay and Saginaw River/Bay case studies, which are based on results for all facilities in
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-15
Table 6-8
Estimated Reduction in Fish Tissue Contaminant Concentrations Due to the
Guidance
Chemical
Chlordane
DDT
Dieldrin
HCB
Mercury
PCBs
TCDD
Toxaphene
Basinwide Reductions
in Loadings1 (%)
68.1
0.2
65.6
36.1
12.8
0.0
0.0
0.2
Reduction in Fish Tissue Concentration (%)
Lake
Superior2
0.68 - 1.36
0.002 - 0.004
0.66 - 1.32
0.36 - 0.72
0.13 - 0.26
0-0
0-0
0.002 - 0.004
Lakes Michigan
and Huron3
3.4 - 6.8
0.01 - 0.02
3.3 - 6.6
1.8 -3.6
0.65 - 1.3
0 -0
0-0
0.01 - 0.02
Lakes Erie
and Ontario4
6.8 - 10.2
0.02 - 0.03
6.6 - 9.9
3.6 - 5.4
1.3 - 1.95
0 -0
0-0
0.02 - 0.03
1 SAIC (1995). Reflects low end cost scenario. However, under the high end scenario, increased
reductions were projected only for mercury (13.1%, or a 0.3% change). This change was
considered insignificant for the assessment of risk reductions under the Guidance.
2 Calculated by multiplying column 1 by 1-2%.
3 Calculated by multiplying column 1 by 5-10%.
4 Calculated by multiplying column 1 by 10-15%.
the areas, show an 89% reduction in PCBs from baseline levels (see Chapter 7).3 Thus, the
potential risk reduction benefits of the Guidance may be underestimated.
For example, recalculating the above results using an average of the estimated loadings
reductions for the three cast study areas instead of the modeled basinwide reductions results
in a greater estimate of benefits. PCS loadings are estimated to be reduced by 89.6% in the
Fox River case study area, 89.4% in the Saginaw Bay/River case study area, and by 0.0% in
the Black River case study area, giving an average PCB reduction of 59.7%. Using this result,
and the resulting average reductions for the additional contaminants included in the risk
assessment, sport anglers are estimated to have a potential reduction of 3.3 to 6.0 excess
cancer cases per year.
That is, modeled results for the basin suggested no reductions in PCBs would result from the
Guidance whereas detailed investigation of two case study sites indicated otherwise.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS » 6-16
Table 6-9
Potential Reduction in Sport Angler Excess Cancer Cases Due to the Guidance
1
Post-Guidance Individual
Expected Risk Level1
Reduction in Cancer Cases
Yearly Cases
Lake Erie
Low Income Minorities
1 Other Minorities3
| Other Sportfishermen
3.7 x lO'3
9.5 x 10"4
1.4 x 10'3
Total
0.01
0.00
0.03-0.04
0.04-0.05
j Lake Huron
i Low Income Minorities3
J Other Minorities3
1 Other Sportfishermen
5.5 x lO'3
1.4 x 10'3
2.1 x lO'3
^ — — • — •
Total
0.00
0.00
0.04-0.08
0.04-0.08
1 Lake Michigan
j Low Income Minorities3
|| Other Minorities3
| Other Sportfishermen
1.2 x 10'2
3.0 x ID'3
45 x 10'3
1 Total .
0.03-0.05
0.01
0.20-0.40
0.24-0.46
Lifetime Cases
0.27-0.41
0.04-0.07
1.80-2.80
2.11-3.28
0.03-0.07
0.01-0.02
2.70-5.40
2.74-5.49
1.70-3.40
0.34-0.69
15.00-30.00
17.04-34.09
Lake Ontario
Low Income Minorities3
Other Minorities3
1 Other Sportfishermen
7.5 >
1.9 >
2.9 >
1 Total
10'3
< lO'3
« 10°
0.00
0.00
0.03-0.05
0.03-0.05
0.17-0.25
0.03-0.05
2.30-3.40
2.50-3.70
Lake Superior
Low Income Minorities3
Other Minorities3
| Other Sportfishermen
2.5 x
6.5 x
9.7 x
(Total
io-3
io-4
10"4
0.00
0.00
0.01
0.01
0.00-0.01
0.00
0.14-0.28
0.14-0.29
All Lakes
Total Low Income Minorities
Total Other Minorities
Total Other Sportfishermen
Grand Total All Lakes
0.04-0.07
0.01
0.31-0.58
0.36-0.66
2.17-4.14
0.42-0.83
21.90-41.90
24.50-46.90
Based on upper estimate of reduction in fish tissue concentrations.
Range over estimated range of reduction in fish tissue concentrations.
Lakeshore counties only.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-17
Table 6-10
Potential Impact of the Guidance on Systemic Risks to Sport Anglers
Hazard Index (HI)
Baseline
Post-Guidance
Lake Erie
Low Income Minorities1
Other Minorities1
Other Sportfishermen
62.6
16.1
24.2
62.6
16.1
24.2
Lake Huron
Low Income Minorities1
Other Minorities1
Other Sportfishermen
75.7
19.5
29.3
75.6
19.5
29.3
Lake Michigan
Low Income Minorities1
Other Minorities1
Other Sportfishermen
183.0
47.0
70.7
182.0
46.9
70.6
Lake Ontario
Low Income Minorities1
Other Minorities1
Other Sportfishermen
110.0
28.2
42.5
109.0
• 28.2
42.4
Lake Superior
Low Income Minorities1
Other Minorities1
Other Sportfishermen
32.0
8.3
12.4
32.0
8.3
12.4
1 Lakeshore counties only.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-18
6.2 RISK ASSESSMENT FOR NATIVE AMERICANS
When Europeans arrived in North America, Native American tribes living in the Great Lakes
Basin included the Chippewa,4 Mohawk, and Ottawa. Fishing was a central feature of the
Hfestyle of many of the Native American tribes in the basin. As the United States expanded,
the interests of the Native Americans and settlers were often at odds. The U.S government
entered into many treaties with the tribes in the Great Lakes Basin. For example, 42 treaties
between the U S. and Chippewa people were signed between 1785 and 1867 (St. Arnold,
1992) The treaties of 1836, 1837, 1842, and 1854 were particularly important to the
Chippewa living in what is now Minnesota, Wisconsin, and Michigan because the Chippewa
reserved hunting, fishing, and gathering rights in the areas ceded to the U.S government
Today, the hunting, fishing, and gathering rights reserved by the ninetee*& century treaties
are known as treaty rights. Figure 6-3 shows the areas involved in the 1836, 1837, 1842, and
1854 treaties.
As early as the late 1800s, the Great Lakes states began attempting to regulate Native
American fishing (Doherty, 1990). In the 1970s, following arrests of Native American
fishermen for fishing with illegal devices and without licenses, the Chippewa and Ottawa
began fighting for their treaty rights in the courts. The major issues in the treaty rights cases
can be summarized by the following two questions: (1) did the modern descendants of the
treaty signatories retain the rights reserved by the treaties, and (2) if the rights were retained,
how are they to be used today? In the first of many federal court decisions throughout
Michigan Wisconsin, and Minnesota addressing treaty rights, the Michigan Supreme Court
affirmed the rights of the Keweenaw Bay Indian Community fishermen (L'Anse Chippewa) to
fish Keweenaw Bay without regard to state fishing regulations in 1971 (Cleland, 1992). Ine
Voieht decision in 1983 upheld the hunting, fishing, and gathering rights of the six Wisconsin
Chippewa bands under the treaties of 1837 and 1842. Disputes in the courts over treaty rights
continue in basin states today.
There are six Chippewa or Ottawa reservations in Michigan. The 1981 Fox Decision affirmed
the treaty rights of the Bay Mills, Grand Traverse, and Sault Ste. Marie tribes to fish in Lakes
Huron Michigan and Superior based on the 1836 treaty. Today, members of those tribes as
well as the Lac Vieux Desert and Keweenaw Bay Indian Community engage in subsistence
fishing activities in the Michigan portion of the Great Lakes Basin. Three tribes in Michigan
Bay Mills Indian Community, Sault Ste. Marie Band of Chippewa, and Grand Traverse Band
of Ottawa and Chippewa, have formed the Chippewa-Ottawa Treaty Fishery Management
Authority to regulate Native American commercial fisheries in the waters of Michigan's
GLIFWC (GLIFWC, 1993a).
4 The Chippewa are also known as Ojibwa and Anishinabe.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-19
Huron
Figure 6-3 Chippewa Treaty Ceded Areas.
Source: GLIFWC, 1993.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-20
Today there are six Chippewa reservations in Wisconsin. The Wisconsin Chippewa hold
treaty rights to areas ceded under the treaties of 1837 and 1842, including most of northern
Wisconsin as well as part of Lake Superior. The off-reservation fishery seasons include spring
spearing, hook and line, and netting. Lake Superior provides an important source of food and
income for the Chippewa. In Wisconsin, two bands, the Red Cliff and the Bad River, have
Lake Superior treaty commercial fisheries. The Wisconsin Chippewa also have off-reservation
seasons for harvesting deer, waterfowl, small game, and wild rice. Off-reservation treaty
activities by Wisconsin Chippewa are regulated by the individual bands as well as the
GLIFWC (GLIFWC, 1993 a).
There are six Chippewa Bands in Minnesota today (nine reservations). Off-reservation
hunting fishing, and gathering rights reserved under the 1837 and 1854 treaties are held by
the Fond du Lac, Grand Portage, and Bois Forte Bands. The Mille Lacs Band is currently
involved in litigation concerning their rights under the 1837 treaty. As the Mille Lacs await
the outcome of the litigation, members of the band exercise their rights in Wisconsin^ The
Grand Portage and Bois Forte Bands have entered into an agreement with the state of
Minnesota in which the band limit their activities under off-reservation rights in return for an
annual payment from the state. There are Grand Portage and Bois Forte off-reservation
seasons regulated by the 1854 Authority for large game hunting, trapping, and fishing. The
Fond du Lac Band regulates off-reservation seasons for its members, including seasons for
deer, bear, moose, wild rice, and fishing. The Grand Portage Band is the only^ Mnmesota
Chippewa band exercising commercial fishing rights on Lake Superior (GLIFWC, 1993D).
Estimating the Potentially Exposed Population
Table 6-11 summarizes the data and information used to determine the potentially exposed
Native American population in the basin. Data from the 1990 census (the most recent
available) were used to estimate the on-reservation population. In addition to the research
discussed above, tribal representatives, EPA tribal liaisons, and the GLIFWC were contacted
to determine whether a tribe was engaged in subsistence fishing. Of the 38 tribes in the basin,
24 are currently engaged in subsistence fishing. The tribes engaged in subsistence fishing are
located in Minnesota, Michigan, and Wisconsin; tribes located in New York are not
subsistence fishing. For at least one tribe in the basin, the traditional subsistence fishing
lifestyle has been curtailed by fish consumption advisories. As shown in Table 6-11, the
potentially exposed population is estimated to be 13,648. Figure 6-4 shows the location of the
potentially exposed population in relation to the point source dischargers in the basin.
Fish Consumption
Table 6-12 provides a summary of information on the consumption of fish by Native
Americans and subsistence anglers. Based on these studies, low, moderate, and high estimates
-------
VO
A
O
O
«;
o
co
00
w
00
CO
1
S
JS
"5
es
03
V)
es
-
•M
es
•_
O
£
ts
Linerican Tribes
^1
es
DJD
C
la
en
fa
U
a
*c«
3
CU
^M
•**
t*m
O
e
cs
'•Z
£
es
U
"S
•4^
cs
e
'5 22
is 2
s.
0
B.
«
? *fe*>
ll
•§ £
e -
o e
*s °
cs '-3
^ ^M
h s
»: C.
W O
oi &<
CJ
Reservation/Tr
e
feC
^
§
JS
02
if
5 3
r near Sault
ity commen
lay, Lake Superioi
Lake Superior tre;
is located on Whitefish E
under 1836 treaty and a 1
Reservation
treaty rights
0
S
ro
o
^*
CS
|| Bay Mills/Chippew
\O
i
0
T3
e
S
f
•P
1
1
m
S
.M
ll
o u
2 £
is located on Grand Travi
Lake Superior treaty com
Reservation
treaty and a
oc
0
CN
S
oo
o
CN
|
CS
Grand Traverse/Ott
Chippewa
TJ
Jn
,£•
6
u
u
in Menomin
upper peninsula, i
1.
H--
§i
o CO
3 c
.S a
Reservation
mouth of Gr
o
fi
r^
i— i
^
£ *
'S
| Hannallville Comm
1 WI Potawatomi
1
1
cu
I
S3
is in central Michigan ne
Reservation
o
u~>
ON
r/*»
«Isabella/Saginaw
Chippewa
_c
'«
c
o
>
«
ula near t
S
; upper pent
end of Michigan's
[836 treaty.
is located at the western <
1 has treaty rights under !
Reservation
border. Band
O\
~*
S
o\
I-H
W
•a
(J CS
W
to O,
Q> Q«
X U
J §
s 1
J VI
a
"3
S
c
u
o.
I
§•
W
c
1
1
.£
i
Bay, Lake Superi
t 1854 treaties.
is located on Keweenaw
aty rights under 1836 ani
Reservation
Band has tre
CN
^
s
_).
CN
C"^
| L'Anse/Chippewa
c
0
a •
*° «
§ -g
fc '—
o „
11
3 i
s §
•i °
el between
iperior treat;
1 S
0 "1
on »-5
1 §
8 t
is on Sugar Island in the
aty rights under 1836 tre
Reservation
Band has tre
VI
in
S
,3-
>o
VI
« Sault Ste.
Marie/Chippewa
^^
^^
r^
t?
ON
^
*N
*•*•*
1
p
!=
§
",
m
oo
u
-o
i
>>
t«
1
•o
I
e
o
is in northern Minnesota,
:aties.
Reservation
and 1854 tre
vo
^»
m
S
vO
n-
rn
«Bois Forte (Nett
Lake)/Chippewa
•S
.c
u
1
O
I
.2
nnesota lake distri
S
E
f
£
.a
c
o
The reservat
reservation.
0
S
VO
CS
|| Deer Creek/Chippe1
'rt
t|
S «
1 1
•a .c
I i
Ji
(M .•*
Is
55 '§
ll
Minnesota, along
es. Consumption ;
f 5
^ OO
C C
— rt
T3 f^
"3 oo
u —
2. u
Reservation
rights under
waters.
VO
^^
(«4
S
VO
**—?
i__-
CS
(D
1 Fond du Lac/Chipp
-------
^
R
en
"S
o
u ,,
There are four portions of the reservation, one area is on trie snore 01 Mine La
another is near the St. Clair River. The tribe is currently involved in litigation i
rights under the 1837 treaty.
oo
cs
S
CO
CS
Miile Lacs/Chippewa
f
c
Location of these lands has not been determined.
0
00
CS
Minnesota Chippewa
Trust Lands4
id
3
— v
S
h-
This reservation has 3 parcels: 1) main portion around upper and Lower n.eu
bordering Canada on Lake of the Woods, and 3) area in Pine Island State Fori
0
S
cs
0
en
Red Lake/Chippewa
u
^
3
Reservation is located in central Minnesota, west ot Uulutn. it is pan 01 Leecr
reservation.
o
S
en
Sandy Lake/Chippewa
;
c
Reservation is located in Scott Co., south-central Minnesota on rnor Laiw.
o
£
^
Jshakopee/Mdewakanto:
Sioux
g
3
2 C
° 3 .
Reservation is located in northern Minnesota, on Vermillion LaKe. it is pan 01
{Nett Lake) reservation. Band has treaty rights under the 1837 and 1854 treati
00
S
oo
U Vermillion
Lake/Chiooewa
3
L.
U
^
| Reservation is located in north-western Minnesota in Mahnomen co. ana Dec
O
S
CS
li
S
1
•j
>
I
j
^9> ^
r>
^
j^^
1
i
c
J3
u.
U
1
Reservation is in south-western New York near the New York-Pennsylvania t
Allegheny River. It is part of Seneca Nation.
to
»
)
'
o
cs
vo
o
Allegany/Seneca4
3
a.
.
£ '.
^
*
u
-
Reservation is in western New York along the Cattaraugus Kiver ana Lane ci
Seneca Nation.
O
cs
Cattaraugus/Seneca4
5
3
3
It
J
] Reservation is in south-western New York near Cuba Lake, u is pan 01 aeno
0
u
J
o
—
Oil Springs/Seneca4
-------
m
CN
vo
s
O
O
oi
O
00
H
00
00
<
g
I
D
^e
VI
&
«
"«
V
O
JS
1
American Tribes in t
*• — ^ ^<
C ^
u '-3
\o s
o
Q 5
R
c-i on
gj
•••
JS
fa
eu
en
_Q
3
t/5
V
^
O
c
•2
'C
Charactel
"o
«
s
•^ B«
"a *
c.
o
a,
g
.2 la
5« ^CW
CQ
C —
o e
"*3 ®
ss "-3
<« C,
U O
K P-
Reservation/Tribe
a
o
"S
•o
§
m
1
-D
1
U
13
1
.z
e
o
Reservati
O
p.
JOneida (East)/0neida
Nation
1
00
o
5
o
n
is located in central New York,
e
o
Reservati
O
cs
UOnondaga/Onondaga
Nation4
5-3
g A
8 *
•5 w
C tn
C •*•
's g
li
t. Lawrence '.
i fishing opp
CD u
3i
!l
u »
1:1
is located along the U.S.-Canad;
Consumption advisories severely
Reservati
New Yor
members.
O
S
m
cs
—
ISt. Regis
Mohawk/Mohawk
I
60
.2
ri
\j
i
i
(2
6
i
13
J3
u
S
is in upstate New York between
}ws through the reservation.
Reservati
tributary)
O
m
i
^a
c
.Ms
^
5
«j
13
e
^s
^:
"E ^
li
Band has tr
ommercial fi
ba U
if
CO I.
it
e w
located in northern Wisconsin o
837 and 1842. Band has a Lake
e «
O '• •
Reservati
treaties o
oo
oo
s
oo
oo
o
«Bad River/Lake Super!
Chippewa
_00
1
I
1
n
c
1
U
W
5
u
is located in Sawyer County, no
eaties of 1837 and 1842.
C J3
O
Reservati
under the
FT
^~*
S
r-
—
-i
Lac Courte Oreilles/La
Superior Chippevva
"93
S
|
S
.SJ
u
ichigan bord
IS
I
!a
41
C
c
'i
is located in north-central Wiscc
the treaties of 1837 and 1842.
i Keservaii
rights unc
IN
'^
'"•
'S
IN
4_k
^^
«Lac du
Flambeau/Chippewa
c
c
o
u
1
I
o
c
.s
I
•S
is west of Green Bay (Lake Mic
e
o
Reservati
CN
00
CO
s
oo
ro
u
Menominee/Menomine
(Algonquin)
>>
«t
ca
i
est of Or
S
I
1
located in eastern Wisconsin, 7
c
o
Reservati
0
£
cs
ra
T3
°5
O
CO
1
CS
"2
'S
O
-------
-------
tj~)
t—
<
nr
<
ID
cx
CO
2
c
v?
to
C
o
ian Reservatii
•a
a
1
•o
»
•c
1
4J
I
Engaged in
enoe Fishing J
u
m co
II
eg
+
a
a
2s
CO
„ •§
•§ a
e o
1 I
3
f.« "
• !••§ a
"s 2
1
i *
a
•o ao
S|
ll
w.
s J
n n
||
O
£3
"ra
d
a
V
eg
0
1 '
V
ao
•a
ia
V
j Jiis6oricaMy
1
ff
'I
1
5
3
•5
1
1
^
. +• * "
iJ^'M^I
f-':- '•,-• r4w* . ••?., ' *•/
ts
u>-
^1
!-C
o>'
^-
\*A/ -S-«,
:'.'5*^'i
^
•A:
Vo:
Z^
^
'7
^
3
O
§.s
-------
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS »• 6-26
Table 6-12
Estimated Fish Consumption Values for Native Americans and Subsistence Anglers
•I
Study
Columbia River
Intertribal Fish
Commission (CRJTFC)
Wisconsin Tribes
Comparative Risk Project
(U.S. EPA, 1992c)
Environmental Defense
Fund (Bailey et aL, 1993)
1993 GLIFWG Survey of
Tribal Spearers
Description
Study of four tribes on the
Columbia River in Washington.
Analysis used dietary profile based
on USDA information (not tribal
survey), and estimates of local
harvest based on tribal information.
Review of adequacy of current EPA
fish consumption assumptions.
Recommends values to use in
absence of site-specific data.
Mail survey of 245 of the 360
persons who speared during 1993 to
investigate concerns about mercury
in fish. 69 persons responded, some
of whom may have responded to
both of two mailings, although no
double responses were noted.
Fish Consumption (grams/day)
57.8*
35.0 (total)
31.5 (local harvest)
Subsistence Anglers
90-165
i
142.6 - 295.0
(Averages across all seasons derived
from survey results showing
average meal of 13-27 oz. and
average number of meals per week
of 3.6 in spring and 2.2-2.7 the rest
of the year).
* Preliminary results.
offish intake (31.5, 57.8, and 140.0 grams/person/day, respectively) were used to estimate
risks. The low estimate is. based on the Wisconsin Tribes Comparative Risk Project. This
study reflects the geographic region of relevance to the Guidance, but does not reflect tribal
consumption data. The moderate consumption estimate is based on the Columbia River study
(CRITFC). The CRITFC results reflect recent survey data from several tribes in the Columbia
River Basin; however, they may not be representative of consumption patterns in the Great
Lakes region. The high estimate of consumption lies within the range suggested by the
Environmental Defense Fund report, but below the values suggested by the GLIFWC spearing
survey. To the extent that the GLIFWC survey is representative of tribal consumption in the
basin, the high estimate may underestimate fish intake. '
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-27
Exposure Assumptions
For this risk assessment, Native Americans were assumed to reside permanently in the area.
Exposure to fish tissue contaminants was assumed to occur daily; fish intake values were
converted to a daily value, although consumption probably consisted of fewer, larger meals.
Standard EPA assumptions were used regarding body weight. Risks were estimated by lake to
more accurately match fish tissue concentrations with the exposed population. Based on the
location of the Native Americans engaged in subsistence fishing in the basin, risks were
assessed for Lake Superior and Lake Michigan. Table 6-13 shows the non-chemical
assumptions used to determine human health risks.
Table 6-13
Exposure Assumptions Used to Calculate Health
Risks to Native American Tribes in the Great Lakes Basin
Exposure Assumptions
Lake Superior
Lake Michigan
Exposed Population
9,811
3,837
Intake (grams per day)
High
Moderate
Low
140.0
58.0
32.0
140.0
58.0
32.0
Exposure Duration
70 vears
70 years
Exposure Frequency
365 davs/vear
365 davs/vear
Bodv Weight
70kg
70kg
Fish Tissue Contaminant Levels
Fish tissue contaminant levels were estimated by lake, based on data from several sources
(UC, 1989; Amrhein, 1990; Gooch et al., 1990; Miller et al., 1992; U.S. EPA, 1992a;
D. DeVault, Great Lakes National Program Office, U.S. EPA, personal communication, 1994;
and Great Lakes Indian Fish and Wildlife Commission, 1994). Data were lake-specific except
for those from the UC.
The chemicals used to evaluate risks were chlordane, DDT, dieldrin, hexachlorobenzene,
mercury, PCBs, 2,3,7,8-TCDD, and toxaphene. These chemicals were chosen based on their
potential to cause adverse human health effects (i.e., cancer or disease) and the availability of
information on fish tissue contaminant concentrations. Chemicals were not included in the
risk assessment if fish tissue concentration data were not available. Table 6-3 shows the
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-28
assumptions regarding chemical toxicity and concentrations in fish tissue for Lakes Superior
and Michigan.
Baseline Human Health Risks
The lifetime cancer risk due to the ingestion of contaminated fish at baseline conditions was
calculated for the exposed population, using the low, moderate, and high fish ingestion
scenarios. Baseline cancer risks range from 1.8 * 10'3 to 3.7 * 10'2, as shown in Table 6-14.
For the exposed population, these cancer risks translate into potential cancer cases over, a
lifetime of 51, 93, and 226, for the low, moderate, and high fish ingestion scenarios. This is
equivalent to 0.8, 1.4, and 3.1 excess deaths per year, respectively. As shown by Table 6-15,
risks are driven by PCB exposure.
Table 6-14
Baseline Potential Excess Cancer Cases for Native Americans
Lake Michigan
Lake Superior
Total
Low Fish Ingestion (32 gpd)
Lifetime Cancer Risk
Yearly Cancer Cases
Lifetime Cancer Cases
8.6 x lO'3
0.5
33.3
1.8 x IQ-3
0.3
18
0.8
51
Moderate Fish Ingestion (58 gpd)
Lifetime Cancer Risk
Yearly Cancer Cases
Lifetime Cancer Cases
1.6 x lO'2
0.9
60
3.3 x lO'3
0.5
33
1.4
' 93
High Fish Ingestion (140 gpd)
Lifetime Cancer Risk
Yearly Cancer Cases
Lifetime Cancer Cases
3.7 x lO'2
2.0
146
8.1 x lO'3
1.1
80
3.1
226
Note: Detail may not add to total due to rounding.
Systemic (noncancer) risks are assessed in the same manner as for sport anglers. The average
baseline HI values for all tribes in the Great Lakes Basin, using the low, moderate, and high
fish ingestion scenarios, are 79.7, 144.4, and 348.5, respectively. All of these values are
indicative of a high potential for systemic injury.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS » 6-29
Table 6-15
Percentage of Each Compound's Contribution to the Total Carcinogenic Risk
Compound
Chlordane
DDT
Dieldrin
HCB
Mercury
PCBs
TCDD
Toxaphene
Total
Lake Michigan
1.8
0.8
8.3
0.0
0.0
81.2
2.2
5.5 '
100.0%
Lake Superior
0.6
2.2
7.9
0.0
0.0
64.3
7.5
17.6
100.0%
Risk Reductions Due to the Guidance
To determine the potential reduction in human health risks to Native Americans, the
percentage reduction in fish tissue contaminant concentrations expected to result from the
Guidance was estimated. These reductions are discussed in Section 6.2 and are shown in
Table 6-8. The reduced fish tissue contaminant concentrations result in a reduction in excess
lifetime cancer cases of between 0.1 and 0.3 for the low fish ingestion scenario; 0.2 and 0.5
for the moderate fish ingestion scenario; and 0.5 to 1.1 for the high fish ingestion scenario.
The expected reductions in cancer cases are shown in Table 6-16. The potential reduction in
average (across lakes) systemic risks are shown in Table 6-17. Systemic risks remain above
one, indicating the potential for adverse systemic health effects.
Similar to the sport angler risk assessment, the estimated reductions in human health risks to
Native Americans anticipated to result from the Guidance are small since the modeled
basinwide loadings do not indicate reductions in PCBs. However, since the Fox River and
Green Bay and Saginaw River/Bay case studies indicate PCB loadings may be reduced by
89% (based on the estimated loadings reductions resulting from examination of all facilities in
the case study areas), the potential risk reductions to Native Americans may be
underestimated. Recalculating the above results using an average of the estimated loadings
reductions for the three cast study areas instead of the modeled basinwide reductions (as was
done for sport anglers) would result in a greater estimate of benefits. Using and average PCB
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-30
Table 6-16
Potential Reduction in Native American Excess Cancer Cases due to the Guidance
Lake Michigan
Lake Superior
Total
Low Fish Ingestion (32 gpd)
Yearly Cancer Cases
Lifetime Cancer Cases
0.00
0.11-0.23
0.00
0.01-0.02
0.00
0.12-0.25
Moderate Fish Ingestion (58 gpd)
Yearly Cancer Cases
Lifetime Cancer Cases
0.00-0.01
0.20-0.41
0.00
0.02-0.04
0.00-0.01
0.22-0.45
High Fish Ingestion (140 gpd)
Yearly Cancer Cases
Lifetime Cancer Cases
0.01
0.49-0.98
0.00
0.05-0.09
0.01 •
0.54-1.10
Note: Detail may not add to total due to rounding.
Table 6-17
Potential Impact of the Guidance on Average Systemic Risks to Native Americans
Fish Ingestion Scenario
Low
Moderate
High
Hazard Index
Baseline
79.7
144.4
348.5
Post-Guidance
79.6
144.2-144.3
348.0-348.2
reduction of 59.7% and the resulting average reductions for the additional contaminants
included in the risk assessment, Native Americans are estimated to have a potential reduction
of excess annual cancer cases ranging from 0.01 to 0.03 for the low fish ingestion scenario
and from 0.06 to 0.11 for the high fish ingestion scenario.
6.3 SUMMARY OF RISK ASSESSMENT RESULTS
Table 6-18 summarizes the potential reduction of excess cancer cases; for sport anglers and
Native American subsistence anglers anticipated to result from the Guidance. For sport
anglers, the Guidance is estimated to result in a reduction of 24.5 to 46.8 annual excess
lifetime cancer cases. For Native American subsistence anglers, the Guidance is estimated to
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-31
Table 6-18
Summary of Potential Reduction in Angler Excess Cancer Cases due to the Guidance
Sport Angler: Low Income Minorities1
Sport Angler: Other Minorities1
Sport Angler: Other Sport Fishermen
Native American Subsistence Anglers2
Total
Lifetime Cases
2.2 -4.1
0.4 - 0.8
21.9 -41.9
0.1 - 0.3
24.6 - 47.1
1 Lakeshore counties only. Low income is defined as less than $25,000 per year.
2 Based on low fish ingestion scenario (32 gpd).
Note: Detail may not add to total due to rounding.
result in a reduction of 0.1 to 0.3 excess lifetime cancer cases under the low fish consumption
scenario, a conservative assumption. In total, these risk reductions represent between 0.35 and
0.67 excess cancer cases per year, and potential benefits of the Guidance of between
S0.7 million and $6.7 million per year (based on the estimated value of a statistical life of
between $2.0 million and $10.0 million (see Violette and Chestnut, 1983; 1986; U.S. EPA,
1989). However, not all excess cancer cases will necessarily result in mortality; therefore the
monetized benefits estimate may be overstated. However, no monetary value is estimated here
for reductions in exposures to noncarcinogens (systemic) and associated reductions in HI
values.
As discussed above, the estimated reductions in risks due to the Guidance are small because
PCB contamination comprises the majority of both systemic and carcinogenic risks, and the
modeled basinwide loadings reductions do not indicate reductions in PCBs resulting from the
Guidance. However, the basinwide results are based on a sample of 59 facilities in the basin,
and may result in conservative estimates of actual basinwide reductions. The estimated
loadings reductions for the Fox River and Green Bay and Saginaw River/Bay case studies,
which are based on examination of all facilities in the case study areas, show an 89%
reduction in PCBs from baseline levels (see Chapter 7). Thus, the potential reductions in risk
due to the Guidance may be underestimated. If, as shown above, the average reduction in
loadings for the three case study areas was used instead of the modeled basinwide reductions,
the Guidance would be attributed with a total reduction of between 3.3 and 6.0 excess cancer
cases per year, and potential benefits of between $6.6 million and $60.1 million per year.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS > 6-32
Interpretation of Results
To interpret the significance of the excess cancer risk estimates for Native Americans and
sport anglers, EPA policy must be considered. In the National Oil and Hazardous Substances
Pollution Contingency Plan (NCP) [40 CFR Part 300], EPA states that acceptable exposure
levels are generally concentration levels that represent an excess upper-bound lifetime cancer
risk to an individual of between 10"4 and 10"6. EPA also states in the preamble to the NCP
that the 10"6 risk level be used as a point of departure for establishing remediation goals for
the risks from chemicals at Superfund sites. Therefore, as shown by 'the risk assessments
presented above, the estimated cancer risks to Great Lakes anglers exceed NCP-based
acceptable levels both before and after implementation of the Guidan.ce.
Sources of Uncertainty
There are numerous sources of uncertainty in the estimate of baseline risks to Native
Americans, and in the estimated reduction in risks from implementation of the Guidance.
Variations in health risks may occur depending upon the following conditions:
* Type and size of fish consumed. Bottom-feeding fish or older, larger fish are generally
more contaminated than sport fish or young, small fish.
*• The type of fish cleaning and cooking techniques used. These may result in differences
in contaminant concentration.
* Accuracy of exposure assumptions used in risk calculations.
»• The congener of PCB present in fish tissue. PCB congeners vary in toxicity, and
congener type was not specified in the concentrations used.
* Accuracy of fish tissue contaminant concentrations used to determine risks. The fish
tissue concentrations used to assess risks were average concentrations for each lake,
and as such, may not represent site conditions.
> Likelihood of all contaminants being present in fish fillet consumed. When calculating
risks, it was assumed that every meal contained a mixture of all contaminants for
which data was available. Although data from U.S. EPA (1992a) indicate that a wide
variety of contaminants can be present in the same fish tissue sample, it is possible
that not all contaminants will be present in every fish fillet consumed. Absence of
some contaminants (such as PCBs) could potentially result in lower risks.
> Inclusion of other fish tissue contaminants. Risks were based on available data of fish
tissue contamination; it is possible that other contaminants are present in fish tissue.
This would cause risks and risk reductions to be underestimated.
-------
UPDATED RISK ASSESSMENTS FOR GREAT LAKES ANGLERS "6-33
Accuracy of estimates of reductions in contamination. Upper and lower estimates of
reductions in fish tissue contamination may not represent actual conditions. It might
take years for fish tissue contaminant reductions to occur, with some reductions lower
than anticipated because of continual release of contamination from sediments into the
water column. Additionally, risk reductions will vary depending upon the amount of
contamination contributed from point sources and nonpoint sources. ,
Risks to Native Americans were calculated using total reservation populations;
however, risks to children and elderly persons may not be accurately represented by
these estimates.
Availability of health care. It is possible that quality health care is not as readily
available for minority or lower income populations, which might result in increased
health risks from exposure to contaminants.
-------
CHAPTER 7
REVISED CASE STUDY BENEFIT-COST ANALYSES
This chapter presents revised quantitative benefits estimates for the case study analyses
presented in the April 1993 RIA. The three case study areas are the lower Fox River and
Green Bay in northeastern Wisconsin (Lake Michigan); the Saginaw River and Saginaw Bay
in Michigan (Lake Huron); and the Black River in Ohio (Lake Erie). The revised benefits
estimates reflect new or more complete information on the relative contribution of point
sources to total loadings at each of the sites (the attribution issue), and revised estimates of
the point source loadings reductions anticipated as a result of the Guidance. In general, the
revised results assume a linear relationship between water quality improvements and benefits
(e.g., if the effect of the Guidance is to move water quality conditions 10% toward "toxic-
free," benefits are estimated to be 10% of the total toxics-oriented net benefits).
In addition, based on the data and information collected to conduct the basinwide risk
assessments presented in Chapter 6, human health risk reduction benefits were calculated for
the Fox River and Saginaw Bay case study areas. These benefits categories were discussed in
the April 1993 RIA but not monetized due to a lack of data.1 Also, in response to comments
suggesting that the case studies are not representative of basinwide benefits of the Guidance,
an analysis of the representativeness of the case studies is presented.
This chapter is organized as follows. Section 7.1 reports revised annual case study benefits
estimates. Section 7.2 provides a comparison of the benefits to estimated costs for the case
studies. Section 7.3 discusses the representativeness of the case study sites in relation to the
universe of sites in the basin.
7.1 REVISIONS TO THE CASE STUDY BENEFITS ESTIMATES
The revised benefits estimates are reported below for each site. Potential benefits are reported
in 1994 (first quarter) dollars. Except for human health benefits, original baseline values were
taken from the April 1993 RIA and updated to 1994 (first quarter) dollars using the Consumer
Price Index.
1 Human health benefits were not calculated for the Black River case study. As discussed in the
April 1993 RIA, the primary pollutants of concern for the Black River site are PAHs. There was not
sufficient fish tissue concentration data available to conduct a risk assessment for these pollutants.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-2
7.1.1 The Lower Fox River/Green Bay Case Study
Table 7-1 reports the estimated baseline point source pollutant loadings and reductions in
loadings due to the Guidance; Table 7-2 reports these results on a toxics-weighted basis. The
Guidance is anticipated to result in a 78.2% reduction in total point source loadings in the
case study area. On a toxic-weighted basis, this represents a 28.2% reduction in loadings.
As described in Chapter 5, several analyses were conducted which focused on better
quantifying the potential impact of the Guidance in bringing about future toxics-oriented
benefits. Many of the original benefits estimates for the lower Fox River/Green Bay case
study area were based on or related to the assumption that the proposed Guidance could
reasonably be attributed with 50% of the credit for attaining fishable, "toxic-free" waters.
Subsequent analyses indicated that point sources are likely to constitute between 5% and 10%
of total loadings to the area (see Chapter 5). Thus, recreational fishing, commercial fishing,
nonconsumptive use, and nonuse benefits for the lower Fox River/Green Bay case study area
were revised to reflect this assumption.
The Guidance is expected to reduce toxic-weighted loadings by 28.2%. Multiplying this
estimate by the estimated contribution of point sources to total loadings in the case study area
(5% and 10%) results in an expected reduction in total toxic loadings of 1.41% and 2.82%.
Most of the revised benefits estimates that follow are based on or related to this estimated
range.
Recreational Fishing
Two approaches were used in the April 1993 RIA to calculate Guidance-related benefits to
recreational fishing. These general methods are also applied to estimate the revised benefits.
Method 1 calculated benefits to recreational fishing as the sum of benefits accrued to
salmonid (trout and salmon) and yellow perch fishing, which were calculated separately.
Method 2 calculated total benefits to recreational fishing.
Method 1: Recreational Fishing Benefits as the Sum of Benefits to Salmonid and Yellow
Perch Fishing
Salmonid Fishery. Total consumer surplus under current conditions was estimated to range
from $77 to $97 million per year for salmonid fishing in the Wisconsin Great Lakes fishery,
30% of which was assigned to the case study area. Based on the results of a CVM survey,
Lyke (1992) estimated that a "toxic-free" fishery would be worth from 11% to 31% more
than the current fishery. Using the revised assumption that 1.41% to 2.82% of net benefits are
attributable to the Guidance, recreational salmonid fishing benefits are estimated to range
from $36,000 to $256,000 per year ($77 million x 0.30 x o.ll x 0.0141; $97 million x 0.30
x 0.31 * 0.0282).
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES > 7-3
Table 7-1
Baseline Point Source Loadings and Reductions in Loadings
Anticipated to Result from the Guidance in the Fox River Casie Study Area (Ib/yr)1
Contaminant
Aldrin
Aluminum
Arsenic (III)
Benzo(a)pyrene
Beryllium
Chlordane
Chrysene
Cyanide, free
DDT
3,3 - Dichlorobenzidine
Dieldrin
Endrin
Heptachlor
Hexachlorobenzene
Hexachlorocyclohexane
alpha - Hexachlorocyclohexane
beta - Hexachlorocyclohexane
Lindane
Mercury
PCBs
Pentachlorobenzene
Pentachlorophenol
Silver
2,3,7,8 - TCDD
1,2,4,5 - Tetrachlorobenzene
Toxaphene
Total
Baseline
0.10
1,152,781
109
18
124
14
97
3,522
0.7
520
2.2
79
3.8
55.7
3.1
75.7
75.2
72.6
294.8
51.3
7,706
422
13,348
0.002
7,706
21.2
1,187,101
Reduction
0.06
909,875
82
17
53
6
95
1,807
0.3
451
1.0
75
2.0
41.7
2.9
73.9
71.1
70.3
230.7
46.0
7,647
392
0
0
7,446
2.4
928,487
% Reduction
60.0%
78.9%
75.0%
95.0%
42.4%
39.8%
97.0%
51.3%
48.5%
86.8%
45.8%
95.6%
53.5%
75.0%
92.9%
97.7%
94.6%
96.9%
78.3%
89.6%
99.2%
92.9%
0.0%
0.0%
96.6%
11.1%
78.2%
Source: SAIC (1995); contaminants for which loadings are not reported were not analyzed or
not detected.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES > 7-4
Table 7-2
Toxic-Weighted Baseline Point Source Loadings and Reductions in Loadings
Anticipated to Result from the Guidance in the Fox River Case Study Area (Ib/yr) -*
Contaminant
Aldrin
Aluminum
I Arsenic (III)
1 Benzo(a)pyrene
| Beryllium
1 Chlordane
[I Chrysene
II Cyanide, free
DDT
1 3,3 - Dichlorobenzidine
|| Dieldrin
|| Endrin
1 Heptachlor
1 Hexachlorobenzene
Hexachlorocyclohexane
alpha - Hexachlorocyclohexane
beta - Hexachlorocvclohexane
Lindane
Mercury
PCBs
i Pentachlorobenzene
Pentachlorophenol
Silver
2,3,7,8 - TCDD
1,2,4,5 - Tetrachlorobenzene
Toxaphene
Total
Baseline
5
73,778
436
76,064
655
32,156
1,753
3,804
4,599
3,797
122,630
7,690
15,410
40.072
56
3,253
902
5,080
147,402
384,142
17,723
211
627,345
724,288
15,412
614,443
2,923,106
Reduction
3
58,232
327
72,231
278
12,799
1,701
1,952
2,232
3,294
56,192
7,348
8,242
30,041
52
3,178
853
4.923
115,360
344,095
17,589
196
0
0
14,892
68.207
824,217
% Reduction
60.0% II
78.9% ||
95.0% 11
42.4%
39.8%
97.0%
51.3%
48.5%
86.8%
45.8%
95.6%
53.5%
75.0%
92.9%
97.7%
94.6%
96.9%
78.3%
89.6%
99.2%
92.9%
0.0%
0.0%
96.6%
11.1%
28.2%
1 Copper-based toxic weights were used.
2 Source: SAIC (1995); contaminants for which loadings are not reported were not analyzed or
not detected.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES > 7-5
Yellow Perch.Fishery. Rehabilitation of the yellow perch fishery in the case study area was
estimated to be worth $1.9 million per year. In this revised analysis, it is assumed that the
Guidance might contribute between 1.41% and 2.82% of the benefits of rehabilitation. This
results in estimated benefits of $27,000 to $54,000 per year.
Thus, under Method 1, total recreational fishing benefits associated with the Guidance range
from $63,000 to $310,000 per year.
Method 2: Total Recreational Fishing Benefits
Based on benefits transfer, the total value of recreational fishing in the case study area under
current conditions was estimated to range from $17 million to $34 million per year. Lyke's
estimate of an increase in value of 11% to 31% for a "toxic-free" fishery was also applied in
this approach. Using the revised assumption that 1.41% to 2.82% of total benefits are
attributable to the Guidance, revised benefits for all recreational fishing range from $27,000
to $150,000 per year. This estimate is comparable to the range estimated using Method 1.
A third, alternative method was also used to estimate recreational fishing benefits for this
revised analysis using the results presented in Chapter 5 for the fish tissue modeling analysis.
In Tables 7-1 and 7-2, it is reported that the Guidance is expected to result in a 90%
reduction in PCB loadings. This reduction may have a positive effect on fish consumption
advisories in the case study area. Using the relationships in Figure 5-6 between exceedences
of PCB-related fish consumption advisories and the reduction in PCB point source loadings, a
90% reduction in point source loadings of PCBs is expected to reduce exceedences by
approximately 35% for the case study area. Assuming that 35% of the benefits of a toxic-free
Fox River/Green Bay fishery are attributable to the Guidance, recreational fishing benefits
range from $902,000 to $3.8 million per year.2
Nonconsumptive Recreation (Nature Viewing)
Using benefits transfer, the current value of nonconsumptive recreation in the case study area
was estimated to range from $28 to $40 million per year. It is assumed that the percentage
increase in nonconsumptive recreation values for "toxic-free" waters is one-half of the
increase for recreational fishing because the Guidance is expected to have a greater relative
impact on fishing. The Guidance may result in the reduction of the severity of some fish
consumption advisories. The estimated range is 6% to 16% (one-half of the 11% to 31%
range). Again, it is assumed that 1.41% to 2.82% of the net benefits of toxic-free waters are
attributable to the Guidance. This results in revised estimated benefits for this category of
$22,000 to $173,000 per year.
/
This estimate is derived by multiplying the sum of toxic-free benefits for the salmonid fishery
($2.6 million-$9.1 million) plus the yellow perch fishery ($1.9 million) by 35%.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-6
Commercial Fishing
If the Guidance, coupled with other efforts to improve water quality (e.g., compliance with
303(c)(2)(B) removal of contaminated sediments), has a strong enough effect to reopen the
commercial fishery in the case study area, it was estimated that total consumer and producer
surplus would range from $190,000 to $342,000 per year. In this revised analysis the
Guidance's impact is associated with 10% to 35% of that value,3 and Guidance-related
benefits are estimated to range from $19,000 to $120,000 per year.
Human Health Benefits
Human health benefits to sport anglers, in the Fox River case study area were calculated by
the method described in Chapter 6. A risk assessment was not conducted for Native
Americans since there are no reservations engaged in subsistence fishing in the case study
area, and the area is not part of treaty-ceded fishing grounds (see Figure 6-4).
The exposed population for the case study area was estimated based on fishing license sales
in Brown and Outagamie Counties. Fishing licenses totaled 48,251. In the lakeshore county
(Brown) licenses were divided into low-income minorities, other minorities, and all other
sport anglers using census data as described in Chapter 6. Consumption and baseline fish
tissue contaminant concentration assumptions also match those presented in Chapter 6 (fish
tissue concentration data for Lake Michigan were used).
The reduction in fish tissue contaminant concentrations anticipated to result from the
Guidance was calculated by multiplying the case study-specific loadings reductions by the
assumed contribution of point sources to total loadings for Lake Michigan (5%-10%). The
resulting reductions are shown in Table 7-3. In comparison with the modeled basmwide
loadings reductions, significant reductions in the contaminants driving baseline risks (i.e.,
PCBs) are indicated for the case study area. As described in Chapter 6, this result is due to
the fact that the case study loadings and reductions were based on a 100% sample of facilities
in the area. Basinwide loadings and reductions were modeled based on a sample of
59 dischargers.
Baseline excess lifetime cancer risks and the potential reductions in risks attributable to the
Guidance are shown in Table 7-4. Since the fish tissue contaminant concentrations for Lake
Michigan were used to assess risks, baseline risks match those for Lake Michigan sport
3 A larger range is attributable to the Guidance for the value of reopening the commercial fishery
than for achieving "toxic-free" status since reopening the fishery will not require removal of all
contaminants (the 1.41% to 2.82% range applies to all contaminants). If, for example, reopening the
fishery requires removal of 10% of contaminants, the Guidance might contribute 10%-20% (1.41%-2.82%
divided by 10%); 35% was used as the upper end of the range based on the results from the fish tissue
modeling presented in Chapter 6.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES »• 7-7
Table 7-3
Estimated Reduction in Fish Tissue Contaminant Concentrations Anticipated to
Result from the Guidance for the Fox River Case Study Area
Chemical
Chlordane
DDT
Dieldrin
HCB
Mercury
PCBs
TCDD
Toxaphene
Reduction in Loadings1
(%)
39.8
48.5
45.8
75.0
78.3
89.6
0.0
11.1
Reduction in Fish Tissue
Concentration (%)
1.99 - 3.98
2.43 - 4.85
2.29 - 4.58
3.75 - 7.50
3.92 - 7.83
4.48 - 8.96
0.00
0.56 - 0.01
1 SAIC (1995).
2 Calculated by multiplying column 1 by 5% to 10%.
Table 7-4
Baseline Sport Angler Excess Cancer Risks and Reductions in Risks Attributable to
the Guidance for the Fox River Case Study Area
•
Low Income Minorities, Lakeshore Counties
Other Minorities, Lakeshore Counties
All Other Sport Anglers
Baseline
Excess Lifetime
Cancer Risk
1.17 x lO'2
3.01 x lO'3
4.52 x lO'3
Post-Guidance
Excess Lifetime
Cancer Risk
Low
Reduction
1.12 x lO'2
2.89 x IQ-3
4.35 x lO'3
High
Reduction
1.08 x lO'2
2.77 x IQ-3
4.17 x lO'3
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES > 7-8
anglers shown in Chapter 6. Post-Guidance risk levels are lower than those shown for Lake
Michigan in Chapter 6, however, since the case study loadings show an 89.6% reduction in
PCBs.
The Guidance is anticipated to result in a reduction in total excess lifetime cancer cases of
between 8.7 and 17.4 (Table 7-5). On an annualized basis, this represents a reduction of
between 0.12 and 0.25 cases, and potential benefits attributable to the Guidance of between
$250,000 and $2.48 million per year (based on the value of a statistical life of between
$2.0 million and $10.0 million (see U.S. EPA, 1989; Violette and Chestnut, 1983; 1986).
However, since not all potential cancer cases will result in mortality, the monetized benefits
estimate may be overstated.
Table 7-5
Baseline Sport Angler Excess Cancer Cases and Reductions in Cancer Cases
Attributable to the Guidance for the Fox River Case Study Area
Low Income Minorities,
Lakeshore Counties
Lifetime
Per Year
Other Minorities,
Lakeshore Counties
Lifetime
Per Year
All Other Sport Anglers
Lifetime
Per Year
Total
Lifetime
Per Year
Exposed
Population
522
243
47,486
48,251
Excess Cancer Cases
Baseline
6.09
0.09
0.73
0.01
214.76
3.07
221.58
3.17
Post-Guidance
5.62 - 5.85
0.08
0.67 - 0.70
0.01
197.92 - 206.34
2.83 - 2.95
204.21 - 212.89
2.92 - 3.04
Reduction
0.24 - 0.48
0.00 - 0.01
0.03 - 0.06
0.00
8.42 - 16.83
0.12 - 0.24
8.68 - 17.37
0.12 - 0.25
Note: Detail may not add to total due to rounding.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES »> 7-9
Nonuse (Passive Use)
In the April 1993 RIA, two methods were employed to estimate nonu;;e benefits. The first
method assumed that nonuse benefits attributable to the Guidance are one-half of recreational
fishing benefits. This yields revised nonuse benefits attributable to the Guidance ranging from
$32,000 to $1.9 million per year (based on the range of estimated benefits over all methods
used to estimate recreational fishing benefits). The second method used benefits transfer to
estimate that total nonuse benefits associated with "toxic-free" waters are $8 million.
Applying the proportion of total benefits estimated to be attributable to the Guidance (1.41%
to 2.82%), revised nonuse values attributable to the Guidance range from $110,000 to
$219,000 per year. The ranges estimated using the two different methods for nonuse values
overlap. A third estimate assumes nonuse benefits are one-half of recreational fishing benefits
derived from the Chapter 5 fish tissue analysis. This estimate ranges from $784,000 to
$1.9 million per year.
Total Benefits Attributable to the Guidance
The revised benefits ranges for each category are summarized in Table 7-6. Total annual
benefits attributable to the Guidance range from $349,000 to $8.58 million per year.
Table 7-6
Potential Annual Benefits Attributable to the Guidance
for the Fox River/Green Bay Case Study Area1
(1994 first quarter dollars)
Benefits Category
Recreational Fishing
Nonconsumptive Recreation
Commercial Fishing
Human Health
Nonuse
Total Benefits
Low Estimate
$27,000
$22,000
$19,000
$250,000
$32,000
$349,000
High Estimate
$3,848,000
$173,000
$120,000
$2,480,000
$1,924,000
$8,544,000
1 Benefits are rounded to the nearest $1,000.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-10
7.1.2 The Saginaw River/Saginaw Bay Case Study
Table 7-7 reports the estimated baseline point source pollutant loadings and reductions in
loadings anticipated to result from the Guidance; Table 7-8 reports these results on a toxic-
weighted basis. The Guidance is expected to result in a 57.4% reduction in total point source
loadings in the case study area. On a toxic-weighted basis, this represents a 60.5% reduction
in loadings.
Many of the benefits estimates for the Saginaw River/Saginaw Bay case study site were
originally based on or related to arbitrary assumptions about how the Guidance would
improve water quality. For this revised analysis, the contribution of the Guidance to future
toxics-oriented water quality improvements is based on the expected contribution of point
sources to total loadings in the case study area (5% to 10%, as shown in Chapter 5). Thus,
recreational fishing, nonconsumptive recreation, waterfowl and other hunting, commercial .
fishing, and nonuse benefits for the Saginaw River/Saginaw Bay case study area were revised
to reflect this assumption.
The Guidance is anticipated to reduce toxic-weighted loadings by 60.5%. Multiplying this
estimate by the estimated range for the relative contribution of point sources (5% to 10%)
results in an expected reduction in total loadings of 3.03 to 6.05%. Most of the revised
benefits estimates that follow are based on or related to this estimated range.
Recreational Fishing
In the April 1993 RIA, benefits transfer was used to estimate that the current social benefits
of all recreational fishing in the case study area are $18 to $43 million per year. Lyke's
increase in value for a "toxic-free" fishery ranging from 11% to 31% was applied to estimate
the potential benefits attributable to the Guidance. Revised benefits attributable to the
Guidance are assumed to range from 3.03% to 6.05% of net benefits for a toxic-free fishery,
or $60,000 to $809,000 per year.
A second estimate was developed using the results from the fish tissue modeling analysis in
Chapter 5. The Guidance is expected to reduce the point source loadings of PCBs to the
Saginaw River case study area also by approximately 90%, which is in turn expected to
reduce exceedences of PCB-related fish consumption advisories by approximately 35%
Assuming that 35% of the benefits of a toxic-free Fox River/Green Bay fishery are
attributable to the Guidance, recreational fishing benefits range from $697,000 to $4.7 million
per year.4
4 This estimate is derived by multiplying the toxic-free benefits for recreational fishing ($2.0
million-S13.4 million) by 35%.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-11
Table 7-7
Baseline Point Source Loadings and Reductions ini Loadings
Anticipated to Result from the Guidance in the Saginaw River Case Study Area
(Ib/yr)1
Contaminant
Acrylonitrile
Aldrin
Aluminum
Arsenic (III)
Benzene
Cadmium
Carbon Tetrachloride
Chloroform
Chromium (VI)
Copper
Cyanide, free
DDT
1,2 - Dichloroethane
1,1 - Dichloroethylene
Iron
Lead
Lindane
Mercury
Nickel
PCBs
Pentachlorophenol
Phenol
Silver
2,3,7,8 - TCDD
1,2,4,5 - Tetrachlorobenzene
Total
Baseline
1,059
2.2
31,969
413
278
1,735
769
3,333
10.1
6,483
2,818
0.5
645
728
115,000
11,193
28
105
5,111
7
1,460
12,036
1,547
5.9E-07
219
196,947
Reduction
1,015
2.2
28,250
363
0
10.6
623
0.
2.9
98
60
0.4
484
728
72,500
1,948
28
102
0
6
1,354 :
4,643
533
0
206
112,956
% Reduction
95.9%
99.1%
88.4%
87.9%
0.0%
0.6%
81.0%
0.0%
29.1%
1.5%
2.1%
86.4%
75.0%
100.0%
63.0%
17.4%
98.6%
97.7%
0.0%
89.4%
92.7%
38.6%
34.4%
0.0%
94.1%
57.4%
1 Source: SAIC (1995); contaminants for which loadings are not reported were not analyzed or
not detected.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-12
Table 7-8
Toxic-Weighted Baseline Point Source Loadings and Reductions in Loadings
Anticipated to Result from the Guidance in the Saginaw River Case Study Area
(Ib/yr)1
Contaminant
Acrylonitrile
Aldrin
Aluminum
Arsenic (III)
Benzene
Cadmium
Carbon Tetrachloride
Chloroform
Chromium (VI)
Copper
Cyanide, free
DDT
1,2 - Dichloroethane
1,1 - Dichloroethylene
Iron
Lead
Lindane
Mercury
Nickel
PCBs
Pentachlorophenol
Phenol
Silver
2,3,7,8 - TCDD
1,2,4,5 - Tetrachlorobenzene
Total
Baseline
900
110
2,046
1,651
5
9,021
100
7
357
. 3,047
3,043
2,969
4
131
644
20,148
1,970
52,345
184
49,525
730
337
72,711
249
438
222,672
Reduction
863
109
1,808
1,452
0
53
81
0
104
46
65
2,564
3
131
406
3,506 __,
1,943
51,141
0
44,278
677
130
25,043
0
412
134,815
% Reduction
95.9%
99.1%
88.4%
87.9%
0.0%
0.6%
81.0%
0.0%
29.1%
1.5%
2.1%
86.4%
75.0%
100.0%
63.0%
17.4%
98.6%
97.7%
0.0%
89.4%
92.7%
38.6%
34.4%
0.0%
94.1%
60.5%
1 Copper-based toxic weights were used.
2 Source: SAIC (1995); contaminants for which loadings are not reported were not analyzed or
not detected.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES »• 7-13
Nonconsumptive Recreation (Nature Viewing)
Benefits transfer was also used to estimate that current benefits for nonconsumptive recreation
in the case study area range from $5 to $7 million per year. Similar to the Fox River case
study analysis, it is assumed for this revised analysis that the percentage increase in the
values for "toxic-free" waters is one-half of the percentage increase in recreational fishing
values, and ranges from 6% to 16%. Applying the range of 3.03% to 6.05% for benefits
attributable to the Guidance, estimated benefits range from $8,000 to $66,000 per year.
i
Waterfowl and Other Hunting
The current benefits associated with waterfowl and other hunting were estimated at $1 million
per year, based on benefits transfer. Applying the same factors that v/ere applied for
nonconsumptive recreation, revised benefits range from $2,000 to $11,000 per year. ,
Commercial Fishing
It was estimated that the current consumer and producer surplus for commercial fishing in the
case study area ranges from $2 to $4 million per year. Assuming that the value of the
commercial fishery would increase by 11% to 31% if it were "toxic-free" (the same increase
that was used for recreational fishing), and applying 3.03% to 6.05% for the potential benefits
attributable to the Guidance, results in benefits ranging from $7,000 to $72,000 per year.
Human Health Benefits
Human health benefits to sport anglers in the Saginaw River/Bay case study area were
calculated by the method described in Chapter 6. A risk assessment was not conducted for
Native Americans since there are no reservations engaged in subsistence fishing in the case
study area, and the area is not part of treaty-ceded fishing grounds (see Figure 6-4).
The exposed population for the case study area was estimated based on fishing license sales
in Saginaw and Bay Counties. Fishing licenses totaled 31,875. In the lakeshore county (Bay),
licenses were divided into low-income minorities, other minorities, and all other sport anglers
using census data as described in Chapter 6. Consumption and baseline fish tissue
contaminant concentration assumptions also match those presented in Chapter 6 (fish tissue
concentration data for Lake Huron were used).
The reduction in fish tissue contaminant concentrations anticipated to result from the
Guidance was calculated by multiplying the case study-specific loadings reductions by the
assumed contribution of point sources to total loadings for Lake Huron (5%-10%). The
resulting reductions are shown in Table 7-9. In comparison with the modeled basinwide
loadings reductions, significant reductions in the contaminants driving baseline risks (i.e.,
PCBs) are indicated for the case study area. As described in Chapter 6, this result is due to
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-14
Table 7-9
Estimated Reduction in Fish Tissue Contaminant Concentrations Anticipated to
Result from the Guidance for the Saginaw River/Bay Case Study Area
Chemical
Chlordanc
Reduction in Loadings1 (%
0.0
Reduction in Fish Tissue
Concentration (%)
0.00
DDT
86.4
4.32 - 8.64
Dieldrin
0.0
0.00
HCB
0.0
0.00
Mercury
97.7
4.89 - 9.77
PCBs
89.4
4.47 - 8.94
TCDD
•^•(•^••"•m""
Toxaphene
0.0
0.0
0.00
0.00
SAIC (1995).
Calculated by multiplying column 1 by 5% to 10%.
the fact that the case study loadings and reductions were based on a 100% sample of facilities
in the area. Basinwide loadings and reductions were based on a sample of 59 dischargers.
Baseline excess lifetime cancer risks and the potential reductions in risks attributable to the
Guidance are shown in Table 7-10. Since the fish tissue contaminant concentrations for Lake
Huron were used to assess risks, baseline risks match those for Lake Huron sport anglers
shown in Chapter 6. Post-Guidance risks levels are lower than those shown for Lake Huron in
Chapter 6, however, since the case study loadings show an 89.4% reduction in PCBs.
The Guidance is anticipated to result in a reduction in total excess lifetime cancer cases of
between 2.02 and 4.03 (Table 7-11). On an annualized basis, this represents a reduction of
between 0.03 and 0.06 cases, and potential benefits attributable to the Guidance of between
860,000 and $580,000 per year (based on the value of a statistical life of between
$2.0 million and $10.0 million (see U.S. EPA, 1989; Violette and Chestnut, 1983; 1986).
However, since not all potential cancer cases will result in mortality, the monetized benefits
estimate may be overstated.
Nonuse (Passive Use)
Two methods were used in the April 1993 RIA to estimate nonuse benefits. The first method
assumed that nonuse benefits attributable to the Guidance were one-half of recreational
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES •> 7-15
Table 7-10
Baseline Sport Angler Excess Cancer Risks and Reductions in Risks Attributable to
the Guidance for the Saginaw River/Bay Case Study Area
Low Income Minorities, Lakeshore Counties
Other Minorities, Lakeshore Counties
All Other Sport Anglers
Baseline
Excess Lifetime
Cancer Risk
5.53 x 1Q-3
1.43 x ID'3
2.14 x 1Q-3
Post-Guidance
Excess Lifetime
Cancer Risk
Low
Reduction
5.37 x ID'3
1.38 x lO'3
2.08 x lO'3
High
Reduction
5.21 x lO'3
1.34 x lO'3
2.02 x lO'3
Table 7-11
Baseline Sport Angler Excess Cancer Cases and Reductions in Cancer Cases
Attributable to the Guidance for the Saginaw River/Bay Case Study Area
Low Income Minorities,
Lakeshore Counties
Lifetime
Per Year
Other Minorities,
Lakeshore Counties
Lifetime
Per Year
All Other Sport Anglers
Lifetime
Per Year
Total
Lifetime
Per Year
Exposed
Population
283
214
31,378
31,875
Excess Camcer Cases
Baseline
1.57
0.02
0.30
0.00
67.28
0.96
69.15
0.99
Post-Guidance
1.47 - 1.52
0.02
0.29 - 0.30
0.00
63.35 -65.31
0.91 - 0.93
65.11 - 67.13
0.93 - 0.96
Reduction
0.05 - 0.09
0.00
0.01 - 0.02
' 0.00
1.96 - 3.92
0.03 - 0.06
2.02 - 4.03
0.03 - 0.06
Note: Detail may not add to total due to rounding.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES > 7-16
fishing benefits. This yields revised nonuse Guidance-related benefits ranging from $30,000 to
$404,000 per year. The second method used benefits transfer to estimate that total nonuse
benefits associated with "toxic-free" waters are $38 million per year. Applying the 3.03% to
6.05% range to this estimate, revised nonuse values attributable to the Guidance range from
$1.1 million to $2.3 million per year. The range estimated using the second method results
in significantly higher estimated benefits than the first method. A third estimate of nonuse
benefits assumes nonuse benefits are one-half of recreational fishing benefits estimated using
the results of the fish tissue modeling analysis in Chapter 5. This estimate ranges from
$349,000 to $2.3 million per year, which is similar to the estimate using the second method.
Total Guidance-Related Benefits
The revised benefits ranges for each category are summarized in Table 7-12. Total annual
benefits attributable to the Guidance range from $168,000 to $7.7 million per year.
Table 7-12
Potential Annual Benefits Attributable to the Guidance
for the Saginaw River Case Study Area1
(1994 first quarter dollars)
Benefits Category
Low Estimate
High Estimate
Recreational Fishing
$60:000
$4,679:000
Nonconsumptive Recreation
$8,000
$66,000
Waterfowl and Other Hunting
$2,000
$11,000
Commercial Fishing
$7,000
$72,000
Human Health
$60,000
$580,000
Nonuse
$30,000
$2,340,000
Total Benefits
$168,000
$7,747,000
Benefits are rounded to the nearest $1,000.
7.1.3 The Black River Case Study
Table 7-13 reports the estimated baseline point source pollutant loadings and reductions in
loadings anticipated to result from the Guidance; Table 7-14 reports these results on a toxics-
weighted basis. The Guidance is anticipated to result in a 25.9% reduction in total point
source loadings in the case study area. On a toxic-weighted basis, this represents a 36.6%
reduction in loadings.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES »> 7-17
Table 7-13
Baseline Point Source Loadings and Reductions in Loadings
Anticipated to Result from the Guidance in the Black River Case Study Area (Ib/yr)1
Contaminant
Aluminum
Arsenic (III)
Cadmium
Copper
Cyanide, free
Fluoride
Iron
Lead
Lindane
Mercury
Nickel
Selenium, total
Zinc
Total
Baseline
23,344
217
264
2,551
284
134,486
24,464
4,284
1.4
23
12,139
138
22,608
224,803
Reduction
0
12
0
0
104
52,629
0
1,176
1.0
14
4,083
131
0
58,148
% Reduction
0.0%
5.3%
0.0%
0.0%
36.5%
39.1%
0.0%
27.4%
72.9%
60.2%
33.6%
94.7%
0.0%
25.9%
1 Source: SAIC (1995); contaminants for which loadings are not reported were not analyzed or
not detected.
The benefits estimates for the Black River case study area were originally based on the
assumption that the Guidance could be credited with 1% to 5% of the benefits that would
accrue from "total water quality improvement." In this revised analysis, the potential benefits
attributable to the Guidance are based on the estimated contribution of point sources to total
loadings in the case study area (10% to 15%, see Chapter 5). Thus, recreational fishing;
recreational boating, waterskiing, sailboarding, and swimming; and nonuse benefits for the
Black River case study area were revised to reflect this assumption.
The Guidance is expected to reduce toxic-weighted loadings by 36.6% in the case study area.
Multiplying this estimate by the estimated range for the relative contribution of point sources
(10% to 15%) results in an expected reduction in total loadings of 3.66% to 5.49%. The
revised benefits that follow are based on this estimated range.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-18
Table 7-14
Toxic-Weighted Baseline Point Source Loadings and Reductions in Loadings
Anticipated to Result from the Guidance in the Black River Case Study Area
(Ib/yr)u
Contaminant
Aluminum
Arsenic (III)
Cadmium
Copper
Cyanide, free
Fluoride
Iron
Lead
Lindane
Mercurv
Nickel
Selenium, total
Zinc
Total
Baseline
1,494
869
1,370
1,199
307
4,707
137
7,711
96
11,497
437
152
1,153
31,129
Reduction
0
46
0
0
112
1,842
0
2,116
70
6,923
147
144
0
11,400
% Reduction
0.0%
5.3%
0.0%
0.0%
36.5%
39.1%
0.0%
27.4%
72.9%
60.2%
33.6%
94.7%
0.0%
36.6%
1 Copper-based toxic weights were used.
2 Source: SAIC (1995); contaminants for which loadings are not reported were not analyzed or
not detected.
Recreational Fishing
Based on a benefits transfer, recreational angling benefits of a "toxic-free" lower Black River
were estimated to range from $7 million to $13 million annually. Using the revised
assumption that 3.66% to 5.49% of total benefits are attributable to the Guidance, revised
recreational angling benefits are estimated to be $251,000 to $719,000 per year.
Recreational Boating, Waterskiing, Sailboarding, and Swimming
The annual benefits to recreational boating, waterskiing, sailboarding, and swimming of a
"toxic-free" Black River were estimated to range from $0.9 to $1.2 million, based on a
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES > 7-19
benefits transfer. Applying the range of 3.66% to 5.49% to compute the potential benefits
attributable to the Guidance results in estimated benefits of $33,000 to $67,000 per year.
Nonuse (Passive Use)
Two methods were used in the April 1993 RIA to estimate nonuse benefits. The first method
assumed that nonuse benefits attributable to the Guidance are one-half of recreational fishing
benefits. This yields revised nonuse Guidance-related benefits ranging from $126,000 to
$360,000 per year. The second method used benefits transfer to estimate that total nonuse
benefits associated with "toxic-free" waters are $12 million. Applying the revised assumption
that 3.66% to 5.49% of total benefits are attributable to the Guidance;, nonuse benefits range
from $445,000 to $667,000 per year. The range estimated using the second method results in
somewhat higher estimated benefits, but the ranges are comparable.
Total Guidance-Related Benefits
The revised benefits ranges for each category are summarized in Table 7-15. Total annual
benefits attributable to the Guidance range from $0.4 to 1.5 million per year.
Table 7-15
Potential Annual Benefits Attributable to the Guidance for the Black River
Case Study Area1
(1994 first quarter dollars)
Benefits Category
Recreational Fishing
Recreational Boating
. Waterskiing. Sailboarding. and Swimming
Nonuse
Total Benefits
1 Benefits are
Low Estimate
$251.000
$33.000
$126.000
$410.000
High Estimate
$719.000
$67.000
$667,000
$1.454.000
rounded to the nearest $1,000.
7.2 COMPARISON OF BENEFITS AND COSTS FOR THE CASE STUDIES
Two methods were used to compare the estimated case study benefits to estimated compliance
costs in the April 1993 RIA: 1) a direct comparison of annualized benefits and costs, and 2) a
comparison of discounted benefits and costs. A comparison of the revised case study benefits
and costs using these methods is presented in Table 7-16.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-20
Table 7-16
Comparison of the Potential Benefits to the Potential Costs
of the Guidance for the Case Study Areas
(millions of 1994 first quarter dollars)
Benefits Range
Midpoint of
Benefits Range
Direct Annualized Comparison1
Discounted Benefits and Costs2
10-Year Phase-In of Benefits
—•i i
20-Year Phase-In of Benefits
$0.2 - $7.6
$2.6 - $120.9
$2.0 - $91.5
$4.0
$61.7
$46.8
Black River Case Study
Direct Annualized Comparison
Discounted-' Benefits and Costs2
10-Year Phase-In of Benefits
$0.4 - $1.5
$6.4 - $22.7
$0.9
$14.5
Costs
$2.6
$53.0
$53.0
Fox FiY«- and Green Bay Case Study
Comparison1
i and Costs2
ase-In of Benefits
ase-In of Benefits
$0.3 - $8.5
$5.4 - $133.9
$4.1 - $101.4
$4.5
$69.7
$52.7
Saginaw River/Bay Case Study
.
$3.6
$71.8
$71.8
20-Year Phase-In of Benefits
$4.8 - $17.2.
$11.0
$42.7
Based on annualized costs assuming a 10-year capital life and reflecting a 7% real interest rate
Present values (1994) over 30 years. Annualized costs (assuming 10-year capital life and 7%
real interest rate on capital) and benefits are discounted at a 3% real discount rate.
Benefits ranges across case study areas are roughly comparable. Annual benefits range from
approximately $200,000 to several million dollars, reflecting the uncertainty in the benefits
estimates. Annualized costs are commensurate with annual benefits; costs are approximately
$2 to $3 million per year for each of the case studies. The net present values of streams of
benefits and costs over 30 years are also generally similar for the Fox River/Green Bay and
Saginaw River/Bay case studies. Discounted costs are above the discounted benefits ranges
for the Black River case study.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES •• 7-21
7.3 CASE STUDY REPRESENTATIVENESS
The three case study areas were originally selected for the benefits analysis on the basis of
data availability, on the relative importance of point source discharges to the watersheds'
problems, and to portray spatial diversity throughout the Great Lakes region. There was no
reason to anticipate, a priori, that the selected sites were atypical of the region; however, an
inherent limitation of the case study approach is the inability to extrapolate to the universe of
sites. Indeed, commenters on the April 1993 RIA asserted that the benefits analysis was not
representative of basiriwide benefits because it was based on case studies of three "hot spots."
The choice of three of the basin's RAP areas as case study sites was; motivated by data
availability. RAP areas are typically well studied; thus, a wealth of relevant data are available.
Data limitations usually preclude conducting case studies of less well known sites. While
RAP areas are hot spots and can be expected to have a higher proportion of potential benefits
associated with total cleanup, they also are expected to have a higher than average proportion
of the costs. Further, other sites may have a greater share of benefits attributable to the
Guidance because contamination at hot spots typically results from historical problems that
are not addressed by the Guidance (e.g., hot spots often have highly contaminated sediments,
which would not be remediated by the Guidance alone).
Two additional case study sites were investigated for possible inclusion in the benefits
analysis: (1) the Ashtabula River in Ohio and (2) the St. Louis River in Minnesota. However,
it was determined that adding case studies would offer only limited insights, because sites
with readily available data have profiles similar to the existing case studies (e.g., large
historical sediment loads). Thus, instead of additional case study benefits analyses, an analysis
of the representativeness of the existing case study sites was conducted.
One means of assessing whether the case study benefits and costs are representative of the
universe of sites is to evaluate the percentage of total benefits represented by the case studies
in comparison to the percentage of basinwide costs these sites represent. Here,
representativeness is evaluated by comparing the proportion of estimated costs, baseline
loadings and expected reductions in loadings, and benefits-related measures represented by
each site. Benefits-related measures (such as population, recreational angling days, and
nonconsumptive recreation days) are used in place of total benefits for the analysis because
there is no estimate of benefits for the entire Great Lakes Basin. A description of the data
used to estimate case study representativeness and the results follows.
Description of Data
Costs. Basinwide and case study cost estimates are the estimated total annualized
undiscounted compliance costs for each site, as shown in Section 7,2. Site-specific cost
estimates for the three case study areas combined amount to 13.6% of the model-based cost
estimates for the basin as a whole.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES * 7-22
Loadings. Baseline loadings and expected reductions in loadings for the case study sites are
described in Section 7.1. The percentage of basinwide loadings is based on the total
basinwide estimates, as derived from the model plant analysis and presented in Chapter 4.
Loadings are shown in toxic pounds equivalent per year. The case study sites combined
account for approximately 9% of the baseline loadings, and 16.6% of the load reductions,
attributed the Guidance in the basin as a whole (in toxic-weighted pound equivalents).
Population. The total population located in the Great Lakes Basin is drawn from Thorp and
Allardice (1994). This estimate was calculated by overlaying a map of the basin boundaries
on 1990 Census data. For the case study sites, the data generally reflect the estimates
presented in the April 1993 RIA. For the Fox River/Green Bay, population was calculated by
multiplying the number of households in the watershed counties (122,800) by the number of
persons per household (2.5). The population of the Black River site was calculated by
multiplying the number of households in the watershed (96,000) by 2.5. The Saginaw Bay
population estimate is the sum of the population in the six counties adjacent to or including
the river or bay. This was determined to be a more reasonable estimate for use in the analysis
than the estimate presented-in the April 1993 RIA (calculated by Michigan DNR). The
combined population for the three case studies approximates 4% of the basin's total
population.
Recreational Angling Days. The total number of recreational angling days basinwide is
drawn from the 1991 US FWS Survey (Table 2) and represents Great Lakes fishing days for
Illinois Indiana, Michigan, Minnesota, New York, Ohio, Pennsylvania, and Wisconsin. For
the case studies, the data are from the April 1993 RIA. Recreational angling days for Fox
River/Green Bay are calculated by dividing 2.2 million recreational angler hours in Green Bay
(for 1990) by 4 4 hours per outing/day. For Saginaw Bay, recreational angling days for
represent the low end of .the range shown in the RIA (527,000 to 631,000 days/year in 1988).
Recreational angler days were not estimated for the Black River site.
Nonconsumptive Recreation Days. Estimated basinwide nonconsumptive recreation days
were also based on the 1991 US FWS Survey data. This estimate was obtained by multiplying
the total nonconsumptive use days for state residents from Table 23 by the Great Lakes
percentage of total fishing days, since the survey report does not include Great Lakes
nonconsumptive use days. The Fox River/Green Bay nonconsumptive use estimate is an
annual average use estimate for Bay Beach Park from park staff. The nonconsumptive use
days for Saginaw River/Saginaw Bay are also as presented in the RIA. No estimate was
calculated for the Black River.
Commercial Catch. The commercial catch estimate for the Great Lakes Basin is from the
Statistical Abstract of the United States (Table No. 1169), and represents the 1985 Great
Lakes catch. Data for the Saginaw River/Bay case study are for 1986, as shown in the April
1993 RIA. No estimates of commercial catch were made for the Fox River/Green Bay or
Black River sites.
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES » 7-23
Results
The results of the analysis are presented in Table 7-17.
Table 7-17
Case Study Representativeness
Measure (Share of Total)
Fox River
Saginaw
River/Bay
Black River
Great Lakes
Basin
Costs
Total Annualized Costs1
(millions of 1994 first quarter dollars)
$3.6
5.9%
$2.6
4.3%
$2.1
3.4%
$61.0
100.0%
Toxic-Weighted Loadings
Baseline
(toxic pounds equivalent/year)
Reduction1
(toxic pounds equivalent/year)
2,923,106
8.3%
824,217
14.1%
222,674
0.6%
134,815
2.3%
31,129
0.1%
11,400
0.2%
35,365,306
100.0%
5,838,289
100.0%
Benefits Measures
Population (thousands)
Recreational Angling Days
(thousands)
Nonconsumptive Recreation Days
(thousands)
Commercial Catch
(millions of pounds)
307
1.2%
500
2.1%
1,000
7.9%
—
459
1.9%
527
2.2%
250
2.0%
2
4.1%
240
1.0%
—
—
24,707
100.0%
24,193
100.0%
12,608
100.0%
54
100.0%
1 The low estimate of compliance cost and load reductions is used in the analysis because EPA
feels the low estimate is more representative of the likely impact of the final Guidance.
The Saginaw River/Bay site, which accounts for 0.6% of the loadings problem
addressed by the Guidance (measured in toxic-weighted baseline loadings), is expected
to incur 4.3% of the costs of the rule. The Saginaw River/Bay site will also account
for 2.3% of the reduction in loadings anticipated to result from the Guidance. Benefits
measures reflect similar shares of 1.9% based on population, 2.0% to 2.2% for
recreation levels, and 4.1% based on commercial catch. Thus, the Saginaw River/Bay
site represents a higher proportionate share of the total Guidance costs (over 4%) than
it does of the proxy measures for benefits (loading reductions, population, etc. are
generally around 2% of the basin.)
-------
REVISED CASE STUDY BENEFIT-COST ANALYSES > 7-24
The Fox River/Green Bay site presents a more ambiguous case. The Fox River/Green
Bav site accounts for a larger share of the total reductions in loadings (14/o) than it
represents in terms of baseline loadings (8%), or costs (6%). However the benefit
proxies of population and recreational angling are a far smaller share of the basmwide
levels (1% to 2%) than are the costs (6%).
The Black River site shows a clearly disproportionate level of costs (3.4%) relative to
benefit proxies (loading reductions of 0.2%, and population of 1.0%).
Overall there is no evidence to suggest that the three case studies reflect an unrepresentative
level of benefits relative to costs. Indeed, as the three case studies combine to account for
nearly 14% of the Guidance total cost, nearly 17% of the loadings reductions, and between
4% and 10% of the benefits proxies (basinwide population, recreational angling,
nonconsumptive recreation, and commercial fishery harvest), it may be that the three, case
studies represent a reasonably proportionate share of costs and benefits.
-------
CHAPTER 8
REFERENCES
Amrhein, J. 1990. Wisconsin Dioxin Data Summary - Fish and Animal Tissue. Prepared for
the Bureau of Water Resource Management, Surface Water Monitoring and Standards, State
of Wisconsin.
Bailey, David S. 1993. The Protection of Sport and Subsistence Fishing Populations in the
United States. Draft document circulated for peer review and comment. 21 p.
Barron, Mace G. 1990. Bioconcentration. Environmental Science and Technology. 24(11)
pp. 1612-1618.
Barren, M.J., JJ. Yurk, and D.B. Crothers. 1994. Assessment of Potential Cancer Risk from
PCBs Bioaccumulated in Fish and Shellfish. Environmental Health Perspectives. 102: 562-
567.
Beltran, R.F. 1992a. Green Bay/Fox River Mass Balance Study: Preliminary Management
Summary. U.S. EPA, Great Lakes National Program Office. Chicago, Illinois.
Beltran, R.F. 1992b. Green Bay/Fox River Mass Balance Study. Preliminary Management
Summary. Addendum. Great Lakes National Program Office, U.S. EPA.
Bierman V.J., J.V. DePinto, T.C. Young, P.W. Rodgers, S.C. Martin., and R. Raghunathan.
1992. Development and Validation of an Integrated Exposure Model for Toxic Chemicals in
Green Bay, Lake Michigan. U.S. EPA, Grosse He, MI.
Center for the Great Lakes. 1989a. Fact Sheet on Great Lakes Areas of Concern: Clinton
River, Michigan. February 22.
Center for the Great Lakes. 1989b. Fact Sheet on Great Lakes Areas of Concern: Detroit
River, International. March 16.
'I
Center for the Great Lakes. 1989c. Fact Sheet on Great Lakes Areas of Concern: Manistique
River, Michigan. June 26.
Center for the Great Lakes. 1989d. Fact Sheet on Great Lakes Areas of Concern: Rouge
River, Michigan. July 5.
-------
REFERENCES > 8-2
Center for the Great Lakes. 1989e. Fact Sheet on Great Lakes Areas of Concern: St. Clair
River. June 28.
Center for the Great Lakes. 1989f. Fact Sheet on Great Lakes Areas of Concern: White Lake,
Michigan. January 25.
Center for the Great Lakes. 1990. Fact Sheet on Great Lakes Areas of Concern: Muskegon
Lake, Michigan. February 2.
Cleland, C.E. 1992. Rites of Conquest: The History and Culture of Michigan's Native
Americans. Ann Arbor, MI: The University of Michigan Press.
Connolly J.P., T.F. Parkerton, J.D. Quadrini, S.T. Taylor, A.J. Thumann. 1992. Development
and Application of a Model of PCBs in the Green Bay, Lake Michigan Walleye and Brown
Trout and Their Food Webs. Prepared by Manhattan College, Riverdale, NY for the U.S.
EPA, Gross He, MI. October 2.
Davila, B., K.W. Wbitford and E.S. Saylor. 1993. Reading Notes: GLI PCB Remediation
Costs. 12 p.
DePinto, J.V. 1994. Role of Mass Balance Modeling in Research and Management of Toxic
Chemicals in the Great Lakes: The Green Bay Mass Balance Study. Great Lakes Research
Review. 1: 1-8.
DeVault, D. 1983. Polychlorinated Dibenzo Dioxins in Fish from the Great Lakes and Region
V Rivers. Prepared for the Great Lakes National Program Office. May.
Doherty, R. 1990. Disputed Waters: Native Americans and the Great Lakes Fishery.
Lexington, KY: The University Press of Kentucky.
Environment Canada. 1991. Toxic Chemicals in the Great Lakes and Associated Effects:
Synopsis. Prepared by the Department of Fisheries and Oceans, Health and Welfare, Canada.
March.
Fiore, B., H. Anderson, L. Hanrahan, L. Olson, W. Sonzogni. 1989. Sport Fish Consumption
and Body Burden Levels of Chlorinated Hydrocarbons: A Study of Wisconsin Anglers.
Archives of Environmental Health. 40(2) 82-88.
Flint, R.W. and J. Vena. 1991. Human Health Risks from Chemical Exposure: The Great
Lakes Ecosystem. Chelsea, MI: Lewis Publishers.
-------
REFERENCES »• 8-3
Fox, G.A. and M. Gilbertson. 1991. Proceedings of the Expert Consultation Meeting on Mink
and'otter. Report to the International Joint Commission, Great Lakes Science Advisory
Board's Ecological Committee. Windsor, Ontario. March 5-6. 30 pp.
Giattina, J. 1993. Responses to OMB's Concerns: Need for the Regulation. Draft
Memorandum from J. Giattina, U.S. EPA, GLNPO. February.
Giesy, IP, J.P. Ludwig, and D.E. Tillitt. 1994. Deformities in Birds of the Great Lakes
Region: Assigning Causality. Environmental Science and Technology. 28(3): 129-135.
Gilbertson, M., T. Kubiak, J. Ludwig, and G. Fox. 1991. Great Lakes Embryo Mortality,
Edema, and Deformities Syndrome (GLEMEDS) in Colonial Fish-Eating Birds: Similarity to
Chick-Edema Disease. Journal of Toxicology and Environmental Health. 33: 455-520.
GLIFWC (Great Lakes Indian Fish and Wildlife Commission). 1993a. A Guide to
Understanding Chippewa Treaty Rights, Wisconsin Edition. Great Lakes Indian Fish and
Wildlife Commission. Odanah, Wisconsin. June.
GLEFWC (Great Lakes Indian Fish and Wildlife Commission). 1993b. A Guide to
Understanding Chippewa Treaty Rights, Minnesota Edition. Great Lakes Indian Fish and
Wildlife Commission. Odanah, Wisconsin. September.
!
Gobas F.A.P.C. 1993. A Model Predicting the Bioaccumulation of Hydrophobic Organic
Chemicals in Aquatic Food-webs: Application to Lake Ontario. Ecological Modeling 69: 1-17.
Gooch, J.W., F. Matsumura, and MJ. Zabik. 1990. Chlordane Residues in Great Lakes Lake
Trout: Acute Toxicity and Interaction at the GAB A Receptor of Rat and Lake Trout Brain.
Chemosphere. 21 (3): 393-406.
Great Lakes National Program Office. 1994. Great Lakes Basin Discharges (NY).
International Joint Commission. 1989. Great Lakes Water Quality Agreement of 1978,
Revised. An International Joint Commission Agreement with Annexes and Terms of
Reference, between the U.S. and Canada signed at Ottawa, November 22, and Phosphorus
load reduction supplement signed October 16, 1983, as amended by protocol, signed
November 18, 1987.
International Joint Commission. 1993. A Strategy for Virtual Elimination of Persistent Toxic
Substances: Seven Reports to the Virtual Elimination Task Force. Windsor, Ontario.
Jacobs, Helen. 1994. Memorandum to Robert Cantilli, Leader, Human Health Mini-Work
Group. June.
-------
REFERENCES *• 8-4
Ludwig, J., J. Giesy, C. Summer, W. Bowerman, R. Aulerich, S. Bursian, H. Auman, P.
Jones, L. Williams, D. Tillitt, and M. Gilbertson. 1994. A Comparison of Water Quality
Criteria for the Great Lakes Based on Human and Wildlife Health. Journal of Great Lakes
Research. 19(4): 789-807.
Luotamo, M., J. Jarvisaio, A. Aito. 1985. Analysis of Polychlorinated Biphenyls (PCBs) in
Human Serum. Environmental Health Perspectives. 60: 327-332.
Lyke, A.J. 1992. Multiple Site Trip Generation and Allocation: A Travel Cost Model for
Wisconsin Great Lakes Sport Fishing. Draft. Ph.D. Thesis presented at the University of
Wisconsin, Madison.
Miller, M.A., C.P. Madenjian, and R.G. Masnado. 1992. Patterns of Organochlorine
Contamination in Lake Trout from Wisconsin Waters of the Great Lakes. Journal of Great
Lakes Research. 18(4): 742-754.
NAS (National Academy of Sciences). 1977. Medical and Biologic Effects of Environmental
Pollutants: Copper. Committee on Biologic Effects of Atmospheric Pollutants, Division of
Medical Sciences, National Research Council.
Nature Conservancy. 1994. The Conservation of Biological Diversity in the Great Lakes
Ecosystem: Issues and Opportunities. January.
Rang, S. 1994. The State of Toxic Contaminants in the Great Lakes. Draft. Prepared by
Environmental Economics International for Great Lakes Laboratory for Fisheries and Aquatic
Sciences, U.S. EPA Great Lakes National Program Office, Environment Canada, and Ontario
Ministry of Environment and Energy.
RCG/Hagler Bailly. 1993. Regulatory Impact Analysis of the Proposed Great Lakes Water
Quality Guidance - Final Report. Prepared by R.S. Raucher, E. Trabka, and A. Dixon for the
U.S. EPA. April 15.
RCG/Hagler Bailly. 1995. Cost-Effectiveness of Reducing PCB Inputs to the Great Lakes
from Point Sources. Memorandum from Mace Barren and Angela Patterson to Mark Morris,
USEPA/OW. January.
SAIC, 1995. Revised Assessment of Compliance Costs Resulting form Implementation of the
Final Great Lakes Water Quality Guidance.
St. Arnold, H. James. 1992. Chippewa Treaties:-'Understanding and Impact. Great Lakes
Indian Fish and Wildlife Commission.--Odanah, Wisconsin.
-------
REFERENCES > 8-5
Stevens, W.K. 1994. Pesticides May Leave a Legacy of Hormonal Chaos. The New York
Times. August 23.
Strachan, W.M.J. and S.J. Eisenreich. 1988. Mass Balancing of Toxic Chemicals in the Great
Lakes: The Role of Atmospheric Deposition. Prepared for the International Joint Commission.
May. 113 pp.
Thomann, R.V., J.P. Connolly, and T.F. Parkerton. 1992. An Equilibrium Model of Organic
Chemical Accumulation in Aquatic Food Webs with Sediment Interaction. Environmental
Toxicology and Chemistry. 11: 615-629.
Thorp, Steve and David R. Allardice. A Changing Great Lakes Economy: Economic and
Environmental Linkages. State of the Lakes Ecosystem Conference. [February 1994. Draft.
U.S. EPA. 1989. Monetized Health Benefits of Regulating Sewage Sludge Use and Disposal.
Final Report prepared by Abt Associates, Inc., for the Office of Policy Analysis and Office of
Water Regulations and Standards, U.S. EPA, Washington, DC.
U.S. EPA. 1992a. National Study of Chemical Residues in Fish - Volumes 1 and 2. Prepared
by U.S. EPA, Office of Science and Technology. EPA 823-R-92-008a/b. September
U.S. EPA. 1992b. National Pollutant Discharge Elimination System (Great Lakes Enforcement
Report for Annual Period Ending September 30, 1992. Prepared by U.S. Environmental
Protection Agency, FY 1992.
U.S. EPA. 1992c. Tribes at Risk: The Wisconsin Tribes Comparative Risk Project. Prepared
by the Office of Policy, Planning, and Evaluation. October.
U.S. EPA Great Lakes National Program Office (U.S. EPA GLNPO). 1993. Addendum to the
Fox River/Green Bay Mass Balance Study: Preliminary Management Summary. U.S. EPA
Great Lakes National Program Office. Chicago, Illinois.
U.S. EPA. 1994a. National Water Quality Inventory: 1992 Report to Congress. Prepared by
U.S. EPA, Office of Water, Washington, DC. EPA 841-R-94-001. March.
U.S. EPA. 1994b. A Summary of Contaminated Sediment Activities Within the United States
Great Lakes Areas of Concern. Prepared by Callie Bolattino, Intern National Network for
Environmental Management Studies Program, Indiana University, Bloomington, Indiana.
August, 1993. 78 p.
U.S. EPA. 1994c. Integrated Risk Information System (IRIS): On-line database of
chemical-specific toxicity factors. Prepared for U.S. EPA, Office of Research and
Development, Washington, DC.
-------
REFERENCES > 8-6
U.S. Fish and Wildlife Service. 1991. National Survey of Fishing, Hunting and Wildlife-
Associated Recreation.
U.S. General Accounting Office. 1993. Issues Concerning Pesticides Used in the Great Lakes
Watershed. Prepared for the Chairman, Subcommittee on Oversight of Government
Management, Committee on Governmental Affairs, U.S. Senate. June. 39 pp.
USGS. 1974. Hydrologic Unit Maps (Michigan, Minnisota, Wisconsin, Pennsylvania, Illinois,
Indiana).
USGS. 1988. Hydrologic Unit Map, State of Ohio.
Vena, John E. 1992. Risk Perception, Reproductive Health Risk and Consumption of
Contaminated Fish in a Cohort of New York State Anglers. Research Program in
Occupational and Environmental Health, State University of New York at Buffalo,
February 14. 67 p.
Violette, D.M. and L.G. Chestnut. 1983. Valuing Reduction in Risks: A Review of The
Empirical Estimates. Environmental Benefits Analysis Series. Washington, DC: U.S. EPA.
EPA-230-05-83-003.
Violette, D.M. and L.G. Chestnut. 1986. Valuing Risks: New Information on the Willingness
to Pay for Changes in Fatal Risks. Report to the U.S. EPA, Washington, DC. Contract
#68-01-7047.
Warren, G. 1993. Revised Whole-Basin Load Estimates for Critical Pollutants. Memorandum
from Glenn Warren to Arnie Leder. March 4.
West, P.C., J.M. Fly, F. Larkin, and R.W. Marans. 1989. Minority Anglers and Toxic Fish
Consumption: Evidence from a Statewide Survey of Michigan. In B. Bryant and P. Mohan,
eds. Race and the Incidence of Environmental Hazards: A Time for Discourse, pp.100-113.
West, P.C., J.M. Fly, R. Marans, and D. Rosenblatt. 1993. Minorities and Toxic Fish
Consumption: Implications for Point Discharge Policy in Michigan. Prepared for the Michigan
Great Lakes Protection Fund.
Wisconsin Department of Natural Resources, and Wisconsin Division of Health. 1994. Health
Guide for People Who Eat Sport Fish from Wisconsin Waters. Publ-IE-0194194 Rev.
Madison, Wisconsin: Wisconsin Department of Natural Resources.
-------
APPENDIX A
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN
FOR THE GUIDANCE
A.1 CADMIUM
A.1.1 Physical and Chemical Characteristics
Cadmium is a silver-white malleable metal that exists in nature in the 0 and +2 valence states
(Friberg et al., 1986). It is insoluble in water, although its chloride and sulphate salts are
soluble (Eisler, 1985). Because of cadmium's ability to bind to paniculate matter and
dissolved organic material (U.S. EPA, 1985a), cadmium has relatively low geochemical
mobility (tendency of an element to flow in natural water and remain there in stable dissolved
form) compared to other trace elements (Baudo, 1987). Yet on the basis of the toxicity and
availability, Wood (1974, cited in Baudo, 1987) classified cadmium as "very toxic and
relatively accessible" compared to other known elements. Contrary to some other trace metals,
cadmium toxicity has been suggested to not change or may increase in the presence of
organic matter (Ravera, 1991). Also unlike other metals, cadmium has an extensive biological
half-life (of about 30 years in humans) and is difficult to eliminate from its soft-tissue storage
sites (e.g., kidney and liver) (Jones and Cherian, 1990). As a nondegradable cumulative
pollutant with a long half-life, cadmium is considered capable of altering aquatic trophic
levels for centuries (Sorensen, 1991).
,
The median dissolved concentrations of cadmium in surface waters (U.S. lakes and rivers)
was 1.0 parts per billion (ppb) (Hem, 1972). Because cadmium is so prevalent in zinc
minerals (Hem 1972), much of cadmium's occurrences in the environment can be associated
with zinc minerals refining and processing activities. Cadmium cycling was studied in
Palestine Lake in Indiana, a long-term recipient of cadmium from an electroplating facility.
Cadmium concentrations in sediments ranged from 1.5 parts per million (ppm) in an
uncontaminated area to 805 ppm near the outlet of a metal-bearing ditch.
A.I.2 Ecotoxicological Characteristics
Cadmium is slow to reach steady-state conditions in fish and other aquatic organisms (i.e.,
caddisfly larvae), often taking much more than 2'8 days (Eisler, 1985, Timmermans et al.,
1992). A study with aquatic invertebrate larvae has shown that uptake of cadmium via food is
the dominant route, with certain caddisfly larvae never achieving steady-state conditions
(Timmermans et al., 1992). Steady-state conditions for cadmium in aquatic organisms may be
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-2
slow to achieve, or may not be achieved, in part, because of particularly efficient cadmium
uptake (e g., in mites, Timmermans et al., 1992) via diet or water exposure and because of
slow excretion or elimination of cadmium (e.g., especially in fish, Eisler, 1985). For instance,
studies have shown that the majority of cadmium originating at the fish gill is transported
transcellularly and accumulates in the liver and kidney, and that the retained-tissue cadmium
is intracellularly bound and unregulated. The intracellularly bound cadmium is likely to have
low turnover because cadmium is a non-regulated element and binds readily and repeatedly to
low molecular weight proteins (e.g., metallothionein) (Wicklund, 1990).
For invertebrate taxa, in general, Crustacea are the most sensitive and the insect larvae are the
least sensitive (Mance, 1987; Attar and Maly, 1982). Zooplankton appear to be extremely
sensitive to cadmium, for example, cadmium concentrations greater than or equal to
0.0002 mg/L are toxic to zooplankton populations (Marshall and Mellinger, 1980).
Fresh-water fish are especially vulnerable to cadmium exposure. Bioconcentration factors in
fish vary greatly, ranging from 33 to 7,440. Behavioral differences such as sediment
association play a large role-in cadmium accumulation. Cadmium accumulation is greater in
fish associated with sediments, such as the bottom-dwelling blacknose dace and fantail
darters These species accumulate 0.60 to 0.76 ppm, compared with free-swimming fish such
as redbreast sunfish and rock bass, which accumulate 0.38 to 0.40 ppm (Sorensen, 1991).
Cadmium accumulation is also influenced by diet. Omnivorous fish such as the warmouth and
bluegill tend to have higher concentrations of cadmium than carnivorous fish such as the
largemouth bass and black crappie (Sorensen, 1991). There is evidence that only lower
trophic levels biomagnify cadmium (Eisler, 1985).
Cadmium concentrations in fish in the United States vary greatly. Fish in the Great Lakes
have cadmium concentrations ranging from 0.1-1.4 mg/kg fresh weight (FW) (liver,
10 species). Cadmium concentrations offish in the Upper Clark Fork River, western Montana,
range from 0.3 to 0.8 mg/kg FW (liver, 7 species). Rainbow trout in Alaska have
concentrations of less than 0.07 mg/kg FW (whole fish) (Eisler, 1985).
The LC50 for Daphnia magna is 0.15 ug/1 and for rainbow trout it is 1.0 ug/1 (Eisler, 1985).
Cadmium exposure results in plasma changes in fish, including hypermagnesemia and/or
hypokalemia, hypocalcemia, and hyperphosphatemia. Hematological changes including anemia
and erythrocyte abnormalities are observed following cadmium exposure (Sorensen. 1991).
Among all species of freshwater biota studied, exposure to cadmium levels of 0.47 to 5.0 ppb
was associated with decreased growth, inhibited reproduction, immobilization, and population
alterations. Sublethal effects in birds are similar to those observed in other species, and
include growth retardation, anemia, and testicular damage (Eisler, 1985).
-------
'"'
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN > A-3
A. 1.3 Human Health Toxicity Characteristics
I
Pharmacokinetics
Cadmium is poorly absorbed by humans after exposure via ingestion. Percentages of
absorption of orally administered doses in humans have been reported to range between 4.6-
6%. Absorption can be altered by the presence or absence of other minerals and dietary
parameters. For example, deficiencies in calcium, vitamin D, protein, zinc, iron, and copper,
lead to increased cadmium absorption. Conversely, the presence of cadmium may inhibit
absorption of copper and calcium, leading to deficiencies in these substances. Inhaled
cadmium is more extensively absorbed, with estimates ranging up to 25% when the inhaled
cadmium compound is relatively soluble. Once absorbed, cadmium has a strong affinity for
both the liver and kidney. No homeostatic mechanism is known for cadmium, and the
removal of absorbed cadmium from the body is a slow process (U.S. EPA, 1985a).
Toxicity
The major adverse health effect associated with long-term cadmium exposure is renal
dysfunction, which can lead to disturbances in mineral metabolism and ultimately to the
formation of kidney stones. Proteinuria, excretion of unmetabolized proteins in the urine, is
one of the most sensitive indicators of renal damage. Hypertension has also been associated
with cadmium exposures (ATSDR, 1989a). ;
Basis for Criteria
The oral reference dose (RfD) for cadmium of 5E-04 mg/kg/day (water), is based on human
studies involving chronic exposure. A concentration of 200 ug/gm wet human renal cortex is
the highest level of cadmium in the human kidney cortex not associated with significant
proteinuria (critical effect). A toxicokinetic model was used to determine the level of chronic
human oral exposure which results in this concentration (200 ug/gm). The model assumes
0.01% elimination of cadmium body burden per day, and 5% absorption of cadmium from
water. The model predicts a NOAEL of 0.005 mg/kg/day for water ingestion. The RfD is
calculated using this NOAEL and an uncertainty factor (UF) of 10 (U.S. EPA, 1992).
Cadmium is listed as a Bl or probable human carcinogen following inhalation exposure.
There is no evidence that cadmium is carcinogenic following ingestion exposure. Seven
studies in which rats and mice were administered cadmium orally have shown no evidence of
carcinogenic response (U.S. EPA, 1992).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN > A-4
A.2 CHROMIUM III
A.2.1 Physical and Chemical Characteristics
The three most stable forms of chromium are the 0, +3 and +6 valence states. Most trivalent
chromium compounds (with the exception of the acetate and nitrate salts) are insoluble in
water (ATSDR, 1989b) Both chromium m and chromium VI exist in nature, and under the
appropriate natural conditions either can be converted to the other. As their toxicities are not
shown to be additive, they are treated as separate entities (U.S. EPA, 1985b).
A.2.2 Ecotoxicological Characteristics
Data are not available regarding the bioaccumulation of chromium IE by freshwater.
organisms, however, it does not appear to bioaccumulate to a great extent. Marine organisms
such as the blue mussel, soft-shell clam, and oyster have BCFs ranging from 86 to 153 (U.&.
EPA, 1985b) The LC50 for Daphnia magna is 1,200 ug/1, and for rainbow trout it is 4,400.
A chronic value of 68.63 ug/1 is available for the rainbow trout (U.S. EPA, 1985b). In
general, invertebrates are more sensitive than fish to chromium IE. (Thurston et al., 1978).
A.2.3 Human Health Toxicity Characteristics
Pharmacokinetics
Absorption of ingested trivalent chromium has been estimated to be less than 1% in humans.
Absorption of chromium following inhalation exposure is slow. The lungs are the only tissues
which appear to accumulate chr-omium with age. Inhalation, not oral, exposure appears to be
the primary source of exposure to this compound (U.S. EPA, 1985b).
Toxicity
Most of the toxic effects associated with chromium compounds are attributed to the more
highly soluble hexavalent form. Trivalent chromium is considered one of the least toxic of the
trace metals. The few available experimental studies of ingestion and inhalation exposures
using a variety of animal species have shown no adverse effects specifically attributable to
trivalent chromium (U.S. EPA, 1985b).
Basis for Criteria
The oral RfD of 1 mg/kg-day is based on a chronic feeding study with rats (U.S. EPA, 1992).
Groups of 60 male and female rats were fed chromic oxide (Cr2O3) at doses of 360, 720, and
1800 g/kg body weight. No effects were observed at any dose level, therefore the NOAEL is
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN * A-5
1800 g/kg body weight (converted by percentage of chromium in chromic oxide and number
of feeding days). An uncertainty factor of 100 accounts for the expected interhuman and
interspecies variability to the toxicity of the chemical, and an additional modifying factor of
10 reflects uncertainty (U.S. EPA, 1992).
A.3 CHROMIUM VI
A.3.1 Physical and Chemical Characteristics
The three most stable forms of chromium are the 0, +3 and +6 valence states. Reducing
processes and materials, such as the acid pH of the stomach, transform hexavalent chromium
into its trivalent form. Hexavalent chromium may exist in aquatic media as water soluble
complexes and may persist in water for long periods of time (U.S. EPA, 1985b).
A.3.2 Ecotoxicological Characteristics
BCFs for chromium VI are quite low. The rainbow trout has been determined to have a
bioconcentration factor of less than three (U.S. EPA, 1985b).
For brook trout and rainbow trout, respective chronic limits of 200 and 350 ug/1 have been
determined, with a resulting chronic value of 264.6 ug/1 for both. The most sensitive effect
for both species was mortality (U.S. EPA, 1985b). A chronic value of 1,987 ug/1 was set for
the fathead minnow (U.S. EPA, 1985b). Reduced growth of chinook salmon was observed
following exposure to concentrations of 16 ug/1 (U.S. EPA, 1985b). In general, invertebrates
are more sensitive than fish to chromium VI. (Thurston et al., 1978).
A.3.3 Human Health Toxicity Characteristics
Pharmacokinetics
Absorption of ingested hexavalent chromium is quite low. It has been estimated to be
approximately 2% in humans. Comparable levels have been observed in experimental animals.
Once absorbed, the acid environment of the stomach reduces much of the ingested hexavalent
chromium to trivalent chromium. Animal studies have indicated rapid absorption of
hexavalent chromium following inhalation exposure. The lungs are the only tissue which
appear to accumulate chromium with age (U.S. EPA, 1985b).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN > A-6
Toxicity
Chromium is an essential element in human nutrition. The daily requirement for chromium
has been estimated to be approximately 50 ^g/day (valence state not specified). There a« no
studies which determine toxicity in humans following oral exposure to chromium VI (U.S>.
EPA, 1985b).
Basis for Criteria
The oral RfD for chromium VI of 5E-03 mg/kg/day is based on a one year study of rats
exposed to chromium VI id'drinking water. A NOAEL of 2.4 mg/kg-day was determined. An
uncertainty factor of 500 represents two 10-fold factors to account for the expected
interhuman and interspecies variability, and an additional factor of 5 to compensate for the
less-than-lifetime exposure duration of the study. The RfD is limited to metallic chromium
(VI) of soluble salts (U.S. EPA, 1992).
Chromium VI is classified as a Group A human carcinogen by inhalation, based on sufficient
evidence of human carcinogenicity (U.S. EPA, 1992). There is no evidence that chromium VI
is a carcinogen via ingestion exposure.
A.4 COPPER
A.4.1 Physical and Chemical Characteristics
Copper exists in surface water in many different chemical forms, or "species," for instance,
copper can be in a freely dissolved state, can dissolve as complexes with inorganic or organic
material can form relatively large colloids of semidissolved complexes, or can adhere to
paniculate matter. The different copper species and forms have different chemical properties
and <5an vary greatly in their toxicity.
Although copper is an essential element at low concentrations for both plants and animals, at
relatively low concentrations copper is toxic to aquatic life (U.S. EPA, 1985c, Sorensen,
1991) Copper occurs in unpolluted surface waters at concentrations of 1 to 10 ug/1, but
concentrations may be much higher in the vicinity of municipal and industrial effluents. Rapid
removal of copper from the water column may occur due to the settling of solids (U.S. EPA,
1985c).
A.4.2 Ecotoxicological Characteristics
There are pronounced species-specific differences in copper accumulation. Examination of
livers of related species of fish collected from uncontaminated areas reveal differences of over
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-7
one-thousand-fold. Levels of 2,795 ppm copper have been shown to accumulate in the liver of
white perch, as compared to the normal level of approximately 3 ppm found in striped bass
(Sorensen, 1991). Fish death is not related to a specific level of copper in tissue. This could
be due to'the differential accumulation of different species of copper (Sorensen, 1991).
It is difficult to determine which species of copper is toxic. Exposure to Cu+2 definitely
induces toxicity, and CuOH+ appears to have toxic effects. Marine species are more sensitive
to copper toxicity than are fresh-water species (Sorensen, 1991). There are many factors
which control the toxicity of copper. Generally, an increase in water hardness, alkalinity,
salinity, organic levels, pH, and fish size results in a decrease in copper toxicity (Sorensen,
1991). Due to the differences in accumulation and toxicity, and the lack of knowledge
concerning copper species toxicity, copper should be treated with great caution.
A.4.3 Human Health Toxicity Characteristics
The U.S. EPA has not established an oral RfD for copper. Copper has been listed by the EPA
as a Group D carcinogen, or not classified as to human carcinogenicity (U.S. EPA, 1992).
A.5 CYANIDE
A.5.1 Physical and Chemical Characteristics
In water, cyanide occurs most commonly in the form of hydrogen cyanide (HCN), which is
soluble in water and dilute acid (including the stomach). It can also exist as the cyanide ion,
alkali metal cyanides, (e.g., KCN and NaCN); relatively stable metallocyanide complexes
(e.g., ferric cyanide); moderately stable complexes (e.g., copper cyanide); or easily
decomposable metallocyanide complexes (e.g., zinc cyanide). The fate of these compounds
can vary widely (U.S. EPA, 1992).
A.5.2 Ecotoxicological Characteristics
As of 1980, the criteria to protect freshwater aquatic life are 3.5 ug/l as a 24-hour average,
never to exceed 52 ug/l. To protect saltwater aquatic life, 30.0 ug/l on an acute toxicity basis;
2.0 ug/l on a chronic toxicity basis (Sax and Lewis, 1989). There is no available information
on cyanide BCFs or LCSOs.
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-8
A.5.3 Human Health Toxicity Characteristics
Pharmacokinetics
Cyanides are readily absorbed following inhalation, oral and dermal exposure. Hydrochloric
acid in the stomach can cause the release of HCN from ingested salts, which is rapidly
absorbed as the cyanide ion (Ellenhorn and Barceloux, 1988). Eighty percent of cyanide,
regardless of the route of entry to the body, is metabolized into a less toxic substance and
excreted in the urine (ATSDR, 1988a).
Toxicity
The toxicity of cyanide will vary with the form of cyanide. Free cyanide in the form of HCN
is the most toxic; many metallo-cyanide complexes are far less toxic. Those forms that
dissociate readily to release free cyanide, such as zinc and cadmium cyanide complexes, are
highly toxic. Others that exhibit moderate dissociation are less toxic; these include copper and
nickel cyanide complexes (ATSDR, 1988a).
The adult lethal dose of HCN is 50 mg. The lethal dose of cyanide salts, such as KCN, is 200
to 300 mg, with a calculated lethal dose in children of 1.2 - 5 mg/kg (Ellenhorn and
Barceloux,' 1988). Cyanides such as KCN or HCN are considered very poisonous. Signs of
acute poisoning include rapid breathing, gasping, tremors, convulsions and death. The severity
and rapidity of the onset of effects depends on the route, dose, duration of exposure, and the
type of compound administered (ATSDR, 1988a).
Basis for Criteria
The oral RfD for cyanide of 2E-02 mg/kg/day was based on a study in which no effects were
observed in rats fed hydrogen cyanide (HCN) in the diet for 2 years. The NOAEL was
10.8 mg/kg body weight. Rat subchronic oral bioassay studies revealed a LOAEL of
30 mg/kg/day CN-. The critical effects were decreased weight gain, thyroid effects
characterized by a decrease of thyroxin levels and myelin degeneration (U.S. EPA, 1992).
An uncertainty factor of 100 was applied, 10 for interspecies extrapolation, and 10 to protect
sensitive individuals. A modifying factor of 5 was also applied to account for the apparent
tolerance to cyanide when ingested with food rather than administration by gavage or in
drinking water (U.S. EPA, 1992).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-9
A.6 MERCURY
A.6.1 Physical and Chemical Characteristics
Mercury is a silver-white, naturally occurring metal which is primarily obtained from mining.
It exists in three valence states, 0, +1, and +2. Metallic mercury is volatile. Mercury can form
a number of organic and inorganic complexes (Friberg et al., 1986, ATSDR, 1988b). Elevated
levels of mercury in living organisms in mercury-contaminated areas may persist for as long
as 100 years after the source of pollution has been discontinued (Eisler, 1987a).
i
A.6.2 Ecotoxicological Characteristics
Mercury is a mutagen, teratogen, and carcinogen, and causes embryocidal, cytochemical, and
histopathological effects in fish and wildlife (Eisler, 1987a). Levels of mercury in the skeletal
muscle of rock bass from Lake Ontario have been measured to range between 100 and
2,700 ppb. Coho salmon in Lake Erie have mercury concentrations in skeletal muscle of
480 ppb (Sorensen, 1991). Mercury accumulation in fish muscle is extremely variable and can
range over two orders of magnitude (Sorensen, 1991). Mercury is biomagnified in the food
chain. Predators accumulate mercury from the food chain and from aqueous uptake. Bottom
feeders accumulate 2.9 ppm from water and 0.07 to 0.11 ppm through the food chain. Pike
accumulate 1.5 to 3.0 ppm from water, and 3.0 to 4.5 ppm through the food chain (Sorensen,
1991). BCFs for mercury range from 1,800 to 85,700 (U.S. EPA, 1985d).
The Great Lakes - Canada region is considered a "hot spot" for mercury contamination. Data
from 3,300 fish taken in 150 samples from locations such as Lakes Erie, Ontario, and
Champlain, as well as the St. Lawrence River, reveal 8 species of fish with average mercury
levels over 0.5 ppm (Sorensen, 1991). A safety level of 0.5 ppm has been established for
mercury in sea food, for the protection of humans against mercury toxicity.
A.6.3 Human Health Toxicity Characteristics
Pharmacokinetics
Inorganic mercury salts and metallic mercury appear to be poorly absiorbed from the
gastrointestinal tract following ingestion, with both animal and human studies suggesting
absorption levels of less than 15%. Absorption following inhalation of mercury vapors is
extensive and thought to occur primarily in the lung alveoli. Human studies have indicated
approximately 80% absorption of metallic mercury compounds following inhalation. Dermal
absorption of mercury salts is also thought to be relatively extensive. Absorbed mercury is
widely distributed and retained in numerous soft tissues, with the kidney generally attaining
the highest concentrations (ATSDR, 1988b).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-io
Toxicity
Major outbreaks of poisoning due to mercury ingestion took place in Japan in the 1950s and
1960s. Neurological symptoms resulted from the consumption of fish contaminated by methyl
mercury from Minimata Bay. Similar cases appeared in Niigata, Japan in 1965 (ATSDR,
1988b). The neurological syndrome observed was characterized by a prickling sensation in the
extremities, impaired peripheral vision, slurred speech, incoordination, irritability, memory
loss, depression, and difficulty sleeping. A recent analysis of causes of death for these
patients found that Minimata disease and noninflammatory diseases of the central nervous
system (CNS) were the most prevalent underlying causes of death. However, in later periods,
the proportion of individuals with noninflammatory diseases of the CNS decreased (ATSDR,
1988b). The mortality rate for all causes of death was significantly higher in Minimata
patients compared to the general population. Rates of liver diseases and nephrosis were 2.9
and 4.7 times higher in these patients than in the general population of Japan (Tamashiro et..
al., 1985). Elevated concentrations of methyl mercury were detected in the hair and brains of
victims. An epidemic of similar neurologic disorders was observed in Iraq, where the
ingestion of mercury-contaminated bread resulted in 500 deatiis (U.S. EPA, 1984).
Basis for Criteria
The EPA is currently assessing the oral RfD for mercury, and a new RfD is pending. The oral
RfD for mercury of 3 x 10'4 mg/kg-day, is the old mercury RfD. It is based on data from the
Niigata, Japan outbreak of mercury poisonings. The estimated threshold blood level was
200 ng Hg/ml blood for the development of neurological symptoms. Extrapolating the long-
term oral dose required to reach this blood level resulted in an estimated intake of Hg of
0.003 mg/kg-day. A safety factor of 10 was applied to determine an interim maximum safe
daily intake of 30 ng Hg/day. The EPA applied an additional oral uncertainty factor of 100
(U.S. EPA, 1984).
A.7 SELENIUM
A.7.1 Physical and Chemical Characteristics
Elemental selenium has four valence states, -2, 0, +4, and +6. The selenites, +4, are the salts
of selenium dioxide. Selenium dioxide is a crystalline-white powder at room temperature, and
forms selenous acid when it comes in contact with water. The selenates, +6, are the salts of
selenium trioxide, a yellowish-white powder, which forms selenic acid in water. Both the
selenites and selenates are soluble in water. The selenides, -2, are insoluble in water (Friberg
et al., 1986). Inorganic selenium may be converted to organic forms by biological action.
Biological systems may convert non-volatile selenium compounds to volatile ones which
might escape to air (U.S. EPA, 1980). Inorganic forms (i.e., selenates, selenites, and selenic
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-i 1
acid) exist in the aquatic environment. Due to microbial transformation, organic forms are
also present (Sorensen, 1991).
A.7.2 Ecotoxicological Characteristics
BCFs for fish muscle and whole fish range from 8 to 78. Exposure concentrations determine
whether selenium is toxic. Selenium is an essential nutrient for living organisms. Deprivation
results in increased mortality, reduced growth, hepatic injury, muscle lesions, and depressed
glutathione peroxidase activity. However, excess selenium concentrations are more toxic than
mercury and represent a threat to the survival of aquatic organisms. A 20 week exposure of
rainbow trout to a diet containing less than 3 ppm selenium resulted in higher mortality,
reduced growth rate, and poor weight gain (Sorensen, 1991).
Selenite is a cumulative toxicant to both fish and invertebrates, with mortality occurring well
beyond the usual four days of testing (U.S. EPA, 1980). Fish accumulating high quantities of
selenium have low hemoglobin levels, low erythrocyte numbers, small and irregularly-shaped
erythrocytes, and reductions in mean corpuscular volumes (Sorensen, 1991).
A.7.3 Human Health Toxicity Characteristics
Pharmacokinetics
Selenium is poorly absorbed following dermal absorption, and well absorbed following both
inhalation and ingestion exposure, with the rate of absorption depending on the chemical
form. Absorption of most selenium compounds is rapid, while elemental selenium is absorbed
more slowly. The degree of gastrointestinal absorption of selenium in humans appears not to
depend on the size of the dose and nutritional status of the exposed individual, indicating a
lack of homeostatic control. Following absorption, selenium is widely distributed to organs
and tissues, with the liver and the kidneys as the principal sites of deposition. In the liver,
many selenium compounds are biotransformed to excretable metabolites, such as dimethyl-
and trimethylselenide (ATSDR, 1989c).
Toxicity
Selenium is an essential trace nutrient for many species, including humans (Friberg et al.,
1986). The normal, background intake of selenium from food, about 80 ug/day, is enough to
meet the daily need for this nutrient (Ellenhorn and Barceloux, 1988). Selenium is an
essential component of the enzymes glutathione peroxidase and heme oxidase, which are
found in humans and animals. This enzyme protects the organism against oxidative damage
by hydrogen peroxide (Sax and Lewis, 1989).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-12
Elemental selenium has low toxicity. Although all selenium salts may produce toxicity
following ingesuon, inhalation, and dermal exposure, severe human toxicity is rare (Ell enhorn
and Barceloux, 1988). The magnitude of the toxic effects following excess selenium intake
depends on how much is ingested, how often it is ingested, and the chemical form. Typical
symptoms of significant overexposure include nausea, vomiting, abdominal pain, fatigue,
changes in finger and toe nails, hair loss, garlic odor on breath, metallic taste in mouth,
irritability and possibly death. Inhaled selenium dusts can produce respiratory tract irritation,
with symptoms consisting of nasal discharge, loss of smell, wheezing, and cough (Clayton
and Clayton, 1982).
Basis for Criteria
The oral RfD for selenium of 5E-03 mg/kg/day is based on a human epidemiological study
involving a population of about 400 people living in an area of China with high
environmental concentrations of selenium. Individuals were evaluated for clinical and
biochemical signs of selenium intoxication. A NOAEL of 0.015 mg/kg/day and a LOAEL of
0 023 mg/kg/day was determined. Clinical signs of selenium intoxication included the typical
"garlic odor" of excess selenium excretion in the breath and urine, hair and nail loss
morphological changes in the nails and teeth, skin lesions, and CNS abnormalities (U.S. EPA,
1992).
A.8 DIOXINS
A.8.1 Physical and Chemical Characteristics
The term polychlorinated dibenzo-p-dioxins (PCDD) or dioxins refers to a class of chemical
compounds consisting of a dibenzo-p-dioxin base structure which is chlorinated to varying
degrees These compounds are called dioxins. Dioxins are chemically and environmentally
stable and persist in the environment (Eisler, 1986a). Seventy-five different PCDD congeners,
which differ in the number and position of chlorine ions, are possible. The most toxic and
best characterized PCDD congener is 2,3,7,8-tetrachlorodibenzo-dioxin (TCDD).TCDD is
typically used as a benchmark against which the toxicity of other PCDDs are ranked (U.S.
EPA, 1985e).
TCDD is a colorless solid at room temperature, with no odor. It has very low solubility in
water (7 91 ng/1 at 20-22°C) and is soluble to some degree in most organic solvents including
benzene, chloroform, and acetone (ATSDR, 1989d). TCDD is lipophilic, exhibiting a higher
degree of solubility in fats and oils than in water (U.S. EPA, 1985e).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN •• A-13
A.8.2 Ecotoxicological Characteristics
TCDD tends to concentrate in fat. A BCF of 66,000 has been noted for dioxins. TCDD is
highly toxic and may result in acute and delayed mortality. Reproductive, carcinogenic,
teratogenic, mutagenic, immunotoxic, and histopathologic effects have been noted following
exposure. Concentrations of over 0.01 ppt in aquatic systems may prove harmful to aquatic
organisms (Eisler, 1986a; Peterle, 1991). High residues of TCDD have been reported in fish
(Eisler, 1986a).
A.8.3 Human Health Toxicity Characteristics
Pharmacokinetics
TCDD is readily absorbed following ingestion and dermal exposure. Animal studies have
shown that the absorption rate is greatly influenced by the medium in which the TCDD is
administered, with adsorption to soil or activated carbon greatly reducing TCDD absorption
(as low as 2%. by dermal and gastrointestinal routes) and oil administration increasing
absorption (75-85% in oral preparations). No information is available; on absorption following
inhalation, or on absorption of TCDD when it is in combination with other compounds, as it
is typically found in the environment (ATSDR, 1989d).
After being absorbed, TCDD is rapidly distributed to tissues with high lipid content (e.g. fat,
skin, and adrenals). Studies with rats have demonstrated that TCDD can cross the placenta
and reach fetal tissues. Exposure to TCDD results in induction of mixed function oxidase
activity and a proliferation of smooth endoplasmic reticulum, the major subcellular storage
site for TCDD. The ability of TCDD to produce this effect has been correlated with the
sensitivity of various strains of mice to TCDD toxicity (Van Miller et al., 1976, Poland and
Glover, 1980).
TCDD appears to be distributed throughout the body and stored primarily as the parent
compound (Olson et al., 1980). For TCDD to be excreted, generally metabolism to more polar
compounds must occur (Olson et al., 1980).
TCDD is excreted primarily in the feces, although polar metabolites are excreted in the urine.
Metabolized TCDD is excreted in both feces and milk. PCDDs have been detected in human
milk samples from Swedish and German mothers (U.S. EPA, 1985e). The biological half-life
in most experimental animal species ranges from 10 - 43 days (ATSDR, 1989d).
Toxicity
A variety of signs and symptoms have been reported in humans exposed to PCDDs and
related compounds following accidental exposure. In most cases, these exposures have
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-14
occurred in the workplace and usually involved a mixture of chemicals. Therefore it has been
difficult to determine whether the signs and symptoms observed were due to PCDDs or to
other chemicals. In addition, the actual level of PCDD exposure can only be roughly
estimated and the establishment of any dose-response relationship is extremely tenuous^ The
signs and symptoms observed from human exposure include: chloracne (skin lesions), hair
loss and hirsutism, porphyria (increased formation of compounds called porphyrms),
hematological abnormalities, central and peripheral nervous systems dysfunctions liver,
kidney and gastrointestinal disturbances. Behavioral changes, respiratory and cardiac disorders
and hypothyroidism have also been reported (ATSDR, 1989d).
Basis for Criteria
The oral reference dose for TCDD is based on a three generation reproductive study in
Sprague-Dawley rats receiving 0.001, 0.1 or 0.01 ug of TCDD/kg. The lowest adverse effect
level (LOAEL) was observed to be 0.001 ug of TCDD/kg based on a reduction in gestation
index decreased fetal weight, increased liver to body weight ratio, and increased incidence of
dilated renal pelvis. An uncertainty factor of 1000 was applied to account for mterspecies
variation and for the development of a LOAEL (U.S. EPA, 1985e).
The oral and inhalation cancer slope factors for TCDD of 1.56E+05 are derived from a
feeding study in female rats that exhibited a statistically significant increased incidence of
tumors in the liver, lungs, hard palate and nasal rurbinates. The assumptions used in the risk
estimate assume that human absorption by oral exposure is equal to that of the rat.
Information regarding absorption by inhalation is limited. However, based on an International
Commission of Radiological Protection (ICRP) lung uptake model, 75% of a dose is assumed
to be absorbed via inhalation exposure. The upper limit unit risks were calculated using a
multistage extrapolation model. The oral cancer slope factor is the same as the inhalation
cancer slope factor, with the implicit assumption being that TCDD is as potent by inhalation
exposure as it is by ingestion exposure (U.S. EPA, 1985e).
A.9 PAHS
A.9.1 Physical and Chemical Characteristics
PAHs, or polycyclic aromatic hydrocarbons, are composed of interconnected benzene rings.
Physical and chemical characteristics tend to vary with molecular weight, (e.g., number of
benzene rings), thus, PAHs differ in their behavior and distribution depending upon their size
(Eisler, 1987b). PAHs are ubiquitous in nature and have been detected in sediments, soils, air,
surface waters, and plant and animal tissues. The'concentration of PAHs is increasing steadily
in the environment (Eisler, 1987b).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN »• A-15
A.9.2 Ecotoxicological Characteristics
PAHs vary widely in toxicity and in chemical properties. PAHs which are unsubstituted and
those of lower molecular weights exhibit acute toxicity, but are noncarcinogenic; these
include benzo(k)fluoranthene. The higher molecular weight PAHs have been shown to be
carcinogenic, mutagenic, or teratogenic to a wide variety of organisms in natural and
laboratory studies; these include benzo(a)anthracene and benzo(b)fluoranthene. Increasing
frequencies of liver tumors in fish populations have been attributed to exposure to PAHs
(Baumann and Harshbarger, 1985).
lexicological and physiological responses to PAHs are variable among species and also may
be modified by ambient inorganic or organic compounds, including other PAHs. Crustaceans
and fish biotransform and excrete PAHs, whereas molluscs, polychaete annelids, and algae
have limited ability to metabolize these compounds. Thus, the potential for bioaccumulation
of PAHs is higher in these species. BCFs for PAHs range from 200 to 1,800.
Several generalizations about PAHs are discussed in the Fish and Wildlife Service's
Biological Report on PAHs (Eisler, 1987b). A limited number of criteria have been
promulgated for PAHs, primarily due to a lack of information on the effect of modifiers and
interactions. Many PAHs are acutely toxic at concentrations between 50 and 1,000 jig/1, and
sublethal responses are observed in the 0.1 to 5 |ig/l range.
A.9.3 Human Health Toxicity Characteristics
The U.S. EPA has not established an oral RfD or CSF for PAHs at this time.
I
A.10 PCBS
A. 10.1 Physical and Chemical Characteristics
I
PCBs, or polychlorinated biphenyls, are synthetic organic compounds characterized by two
joined phenyl rings that are chlorinated to different degrees. There are ten possible sites for
chlorination on this 12-carbon ring system, giving rise to ten PCB isomer groups (homologs):
monochlorobiphenyl through decachlorobiphenyl. Within each isomer group, the PCB
compounds contain the same number of chlorine atoms in each molecule, but differ in
properties and in structural arrangement on the phenyl rings. A number of positional isomers
(cogeners) are possible, depending on the number of chlorines in the molecule. There are a
total of 209 theoretically possible cogeners, however, only 120 cogeners have been identified
in the environment (Eisler, 1986b; Limburg et al., 1986).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN > A-16
m«-«™-»"—«™^
Commercial PCBs consist of mixtures of chlorinated biphenyl cogeners. Various trade names
have been given to industrial mixtures of PCBs, such as Aroclor (Monsanto, U.S.), Clophen
(Bayer West Germany), Kanechlor (Kanegafuchi, Japan), Phenochlor (Caffaro, Italy),
Pyralene (Prodelec, France) and Sovol (U.S.S.R.). The various commercial Aroclor mixtures
were named a four-digit number; the first two numbers indicating the number of carbon
atoms and the last two, the approximate weight percent of chlorine in the mixture. For
example, Aroclor 1242 is a U.S. produced PCB mixture which has 12 carbon atoms and has
approximately 42% chlorine content by weight. One important exception is Aroclor 1016,
which has 12 carbon atoms and contains approximately 41% chlorinate by weight. A major
difference between Aroclor 1016 and Aroclor 1242 is the relative amounts of contamination
by polychlorinated dibenzofurans (PCDFs); Aroclor 1016 is almost completely without
contamination, while Aroclor 1242 contains approximately 1.5-2.0 ug PCDFs per gram PCBs
(Clayton and Clayton, 1982). The fact that PCBs typically exist as mixtures complicates
assessment of their physical, chemical, and biological properties, since individual cogeners
comprising a given mixture may vary in these properties. Another complicating factor is the
presence in PCB mixtures of varying degrees of toxic contaminants (i.e. PCDFs) (ATSDR,
1987).
The physical properties of individual congeners range from liquids to waxes to solids.
Mixtures of congeners have different properties than those of their individual components.
PCBs are thermally stable and resistant to degradation via oxidation, acids, bases, and other
chemical agents. Volatilization is less likely to occur in the higher chlorinated PCBs.
Microbial degradation of PCBs depends upon the degree of chlorination and the position of
the chlorine atom on the biphenyl molecule. In general, the lower chlorinated PCBs are
readily transformed by bacteria. Higher chlorobiphenyls, i.e., those with five or more chlorine
atoms, have proven to be more persistent in the environment (Eisler, 1986b). Mono-, di-, and
trichlorinated biphenyls biodegrade fairly quickly, tetrachlorinated biphenyls biodegrade more
slowly, and the more chlorinated biphenyls appear relatively resistant to biodegradation.
However, anerobic biodegradation may degrade higher chlorinated cogeners more readily than
lower chlorinated cogeners. In addition to degree of chlorination, biodegradation rate is
determined by chlorine positions on the biphenyl ring. PCBs containing chlorines in the ortho
positions are more resistant, and PCBs containing chlorines in the para positions are
preferentially biodegraded as compared to other ring positions. PCBs containing all of the
chlorines on one ring are degraded faster than PCBs containing the chlorines distributed
between both rings (ATSDR, 1987). PCBs are soluble in biological lipids and most common
organic solvents, but only slightly soluble in water, glycerol, and glycols. Water solubility of
PCBs generally decreases with increasing chlorination (Clayton and Clayton, 1982; Eisler,
1986b). PCBs show a tendency to partition onto particulate matter, such as soils and
sediments. Levels are usually highest in aquatic sediments containing microparticulates and
high organic or clay content. PCBs may remain available for resuspension in the aquatic
sediment for up to 15 years (Eisler, 1986b).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN »• A-17
A. 10.2 Ecotoxicological Characteristics
There are at least two processes of PCB accumulation in organisms: bioconcentration from
water and bioaccumulation from sediments and the food chain. PCB homologs having greater
than five chlorine atoms tend to bioaccumulate, while the less-chlorinated ones are
metabolized and excreted (Limburg et al., 1986). This tendency of the more highly
chlorinated PCBs to bioaccumulate is due in part to the increased hydrophobicity and lipid
solubility of the more chlorinated PCBs. Bioaccumulation is also dependent upon chemical
molecular size and shape, and animal species and size (Barron, 1990). BCFs for PCBs range
from 1,700 (crayfish) to 47,000 (rainbow trout).
Water concentrations of PCBs of 0.006 ug/1 resulted in accumulation by filter-feeding
shellfish. Toxicity was demonstrated in sensitive species of teleosts with PCB residues in
excess of 500 ug/kg in the diet and 400 ug/kg in the whole body (Eisler, 1986b). PCBs cause
a variety of toxic responses, including death, birth defects, reproductive failure, liver damage,
tumors, and a wasting syndrome in aquatic and terrestrial organisms (Eisler, 1986b).
A.10.3 Human Health Toxicity Characteristics
Pharmacokinetics
Humans are exposed to PCBs predominately via ingestion of contaminated fish. Elevated
levels of PCBs were found in the serum and breast milk of women who ate PCB-
contaminated fish from Lake Michigan. Blood levels of PCBs in people who ingested
contaminated sport fish from Lake Michigan in 1973 varied with the amount of fish ingested.
Annual fish consumption of > 24 Ib resulted in a mean PCB blood level of 0.073 ppm, while
annual consumption of < 6 Ib fish resulted in a mean blood PCB level of 0.020 ppm. People
who ate no fish had an average PCB blood level of 0.017 ppm (ATSDR, 1987). While these
studies indicate human absorption of PCBs, they do not reveal the degree of absorption.
Studies of both individual cpgeners and Aroclor mixtures indicate that gastrointestinal
absorption by animals of these compounds following oral administration exceeds 90% in most
cases. The efficiency of PCB absorption in the rat was not affected by the degree of
chlorination. Rats treated by gavage with 19 PCB congeners and unchlorinated biphenyls at
doses of 5, 50, or 100 mg/kg revealed that absorption of all congeners was greater than 90%
(Albro and Fishbein, 1972). In rhesus monkeys, 90% absorption was observed with a single
dose of 1.5 or 3.0 g/kg Aroclor 1248 (Allen et al., 1974).
Following exposure, humans distribute PCBs to fatty tissue and breast milk. The half-life of
PCBs in breast milk has been reported to be 5 to 8 months, and the concentration of PCBs in
breast milk was reported to be 4 to 10 times that in maternal blood. Analysis of PCB
concentration in six Japanese women revealed that as the chlorine content of the PCB
-------
CHARACTERISTICS AND TOXICUY OF CONTAMINANTS OF CONCERN » A-18
congeners increased, the correlation between the placental content of congeners and maternal
blood and milk also increased (ATSDR, 1987).
Metabolism of PCBs depends on their chlorine content and on the site of chlorination. The
presence of two adjacent un-substituted carbon atoms that are not sterically hindered facilitate
metabolism and excretion of PCBs (Matthews and Kato, 1979). The capacity to metabolize
PCBs declines from mammals to birds to fish (Hutzinger et al., 1972). The majority of
identified PCB metabolites are phenolic products, in addition to sulfur-containing metabolites,
trans-dihydrodiols, polyhydroxylated PCBs, and methyl ether derivatives (U.S. EPA, 1987).
Phenol is an extremely poisonous crystalline compound (Taylor, 1988). The metabolism of
PCBs includes the intermediate formation of arene oxides. Arene oxides have been implicated
in carcinogenicity, mutagenicity, and cell death. Studies suggest most of the toxic responses
due to PCB exposure are initiated by the parent hydrocarbon first binding to a cytosolic
receptor protein.
PCB excretion following ingestion depends in large part on the metabolism of PCB cogeners
to more polar compounds (ATSDR, 1987). Studies of the blood of humans who had ingested
rice-bran oil contaminated with Kanechlor 500 and PCDFs revealed that the tetra- and some
penta- isomers are eliminated faster than other penta-, hexa-, and hepta- isomers. Half-lives in
blood were determined to be 9.8 and 8.7 months for the 2,4,5,2',4'- and 2,3,4,3',4'-penta-
isomers, respectively. It was also indicated that two adjacent unsubstituted carbon atoms at
the meta-para positions facilitated elimination from the blood (ATSDR, 1987).
Toxicity
The major target organs of PCB exposure are the liver and cutaneous tissue. Occupational
exposures to PCBs have resulted in inconsistent alterations in serum levels of liver enzymes
and dermatological effects including chloracne. However, monitoring data do not adequately
characterize exposures or exposure levels in many cases (ATSDR, 1987).
Basis for Criteria
The EPA has classified PCBs as a group B2, or probable human carcinogen, and has
determined an oral slope factor of 7.7 (mg/kg/day)-1, based on inadequate evidence of excess
risk of cancer in humans and sufficient evidence of increased cancer risk in tests with three
strains of rats and two strains of mice (U.S. EPA, 1992).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN * A-19
A. 11 PHENOLS
A.11.1 Physical and Chemical Characteristics
Phenol is an aromatic white crystalline mass with a sickly sweet, acrid odor. Phenol will
decompose slowly on contact with air and it is recommended to avoid contact of phenol with
strong oxidizing agents (Clayton and Clayton, 1982).
A. 11.2 Ecotoxicological Characteristics
The permissible concentration of phenol in water to protect freshwater aquatic life is
10,200 ug/1, based on acute toxicity data and 2,560 ug/1 based on chronic toxicity data. To
protect saltwater aquatic life, a limit of 5,800 ug/l has been set based on acute toxicity data.
There is a reported BCF of 1.4 for phenols (Sax and Lewis, 1989).
A.11.3 Human Health Toxicity Characteristics
Pharmacokinetics
Phenol is metabolized and excreted principally by the kidneys as a sulfate or glucuronide
conjugate (Goodman and Oilman, 1980). Some phenol may be excreted unchanged; this can
occur at especially high doses. Other reported metabolites include hyciroquinone, and various
other quinones and catechols. After absorption into the body, traces of "free" phenol are
eliminated with feces and expired air (Clayton and Clayton, 1982).
Toxicity ,
Phenol is toxic, with a probable oral lethal dose to humans of 50-500 mg/kg. Some
individuals may be hypersensitive with lethality or serious effects at very low exposures
(Clayton and Clayton, 1982). Death and severe toxicity are usually due to effects on the
central nervous system, heart, blood vessels, lungs, and kidneys. However, toxic
manifestations may vary somewhat with the route. Protracted or chronic exposure usually
results in major damage to the liver, kidneys and eyes. Pigmentary changes of the skin have
been noted. Consumption of water contaminated with phenol has resulted in diarrhea, mouth
sores, burning of the mouth, and dark urine (Clayton and Clayton, 1982).
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN » A-20
A. 12 DEFINITIONS
In order to enhance understanding of this appendix, the following definitions are provided.
Bioaccumulation: Bioaccumulation is the tendency for organisms to extract a compound
present in low concentrations and to store it cumulatively in their bodies.
Bioconcentration factor: A bioconcentration factor (BCF) relates the concentration of a
chemical in aquatic animals to the concentration in the water in which they live.
Biomagnification: Biomagnification is the tendency of a compound to be passed along a
trophic chain in a cumulative manner. In this way, each successive consumer may accumulate
a larger dose than the previous one.
Cancer Slope Factor (CSF): An index of the cancer-causing ability of a compound, based on
the slope of the dose-response curve. The potency factor is used to estimate an upper-bound
probability-of an individual developing cancer as a result of a lifetime of exposure to a
particular level of a potential carcinogen.
EPA classification system for carcinogenicity: The carcinogenicity of compounds is
designated by various letters, which are assigned the following meanings: A - Human
carcinogen; Bl or B2 - Probable human carcinogen (Bl means there is limited human data
available, B2 shows that there is sufficient evidence in animals and inadequate or no evidence
in humans); C - Possible human carcinogen; D - Not classifiable as to human carcinogenicity;
E - Evidence of noncarcinogenicity for humans.
IRIS: Integrated Risk Information System. IRIS is an on-line database created by the EPA
and available from the National Library of Medicine's Toxicology Data Network system. The
database contains EPA health risk and regulatory information on about 400 chemicals, with
both carcinogenic and non-carcinogenic risk assessment data for oral and inhalation routes of
exposure. These data include Reference Doses (RfD) (indicators of non-carcinogenic risks)
and unit risks, (indicators of carcinogenic risks). The regulatory information relates to
environmental statutes such as the Clean Air Act, Clean Water Act, and SUPERFUND
legislation. IRIS is further supplemented with EPA Drinking Water Health Advisories,
substance identification, chemical and physical properties, acute toxicity, and aquatic toxicity.
Lowest-Observed-Adverse-Effect-Level (LOAEL): In dose-response experiments, this is the
lowest exposure level at which there are statistically or biologically significant increases in
frequency or severity of adverse effects in an exposed population.
Maximum Contaminant Level (MCL): MCLs are the maximum allowable concentrations of
contaminants in drinking water as established by federal and state agencies responsible for
regulating public water systems. These are legally enforceable standards. MCLs are set to
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN > A-21
protect the public from acute or chronic health effects or from an "unacceptable" cancer risk.
They must also take into account technological feasibility and economic impact.
mg/kg-d: a dose rate of exposure to contaminants expressed in milligrams of contaminant per
kilogram of body weight per day.
No-Observed-Adverse-Effect-Level (NOAEL): In dose-response experiments, an exposure
level at which there are no statistically or biologically significant increases in the frequency or
severity of adverse effects between the exposed population and its appropriate control; -some
effects may be produced at this level, but they are not considered to be adverse, nor
precursors to specific adverse effects.
PCB Metabolite: Metabolites are formed by enzymes systems within an organism, generally
by introduction of an oxygen into the molecule. Metabolism of PCBs is dependent upon on
the number and position of chlorine atoms, with lesser chlorinated isomers metabolized more
readily than more chlorinated isomers. PCB metabolites are the products of this metabolism,
and tend to be more water soluble (and thus more easily excreted) than the parent compound.
Reference Dose (RfD): A benchmark for the daily dose to which humans, including sensitive
populations (such as children or pregnant women), may be exposed to without an appreciable
risk of adverse health effects during a lifetime of exposure. RfDs are: a dose, and are therefore
expressed in the typical units of dose of milligrams of contaminant per kilogram of body
weight per day (mg/kg-d).
A. 13 REFERENCES
Agency for Toxic Substances and Disease Registry (ATSDR). 1987. [Profile for Toxicological
Selected PCBs. U.S. Public Health Service: Atlanta, Georgia. Draft for Public Comment.
Agency for Toxic Substances and Disease Registry (ATSDR). 1988a. Toxicological Profile
for Cyanide. Published by Oak Ridge National Laboratory. Draft Profile.
Agency for Toxic Substances and Disease Registry (ATSDR). 1988b. Toxicological Profile
for Mercury. U.S. Public Health Service: Atlanta, GA. Draft for Public Comment.
Agency for Toxic Substances and Disease Registry (ATSDR). 1989a. Toxicological Profile
for Cadmium. U.S. Public Health Service: Atlanta, GA.
Agency for Toxic Substances and Disease Registry (ATSDR). 1989b. Toxicological Profile
for Chromium. U.S. Public Health Service: Atlanta, GA.
-------
CHARACTERISTICS AND TOXICUY OF CONTAMINANTS OF CONCERN * A-22
Agency for Toxic Substances and Disease Registry (ATSDR). December 1989c. Toxicological
Profile for Selenium. U.S. Public Health Service. Atlanta, GA.
Agency for Toxic Substances and Disease Registry (ATSDR). 1989d. Toxicological Profile
for Dioxin. Draft Profile. Published by Oak Ridge National Laboratory.
Albro P.W., and L. Fishbein. 1972. "Intestinal Absorption of Polychlonnated Biphenyls in
Rats." Bulletin of Environmental Contain Toxicology, 8:26.
Allen JR. DH Norback, and I.C. Hsu. 1974. "Tissue Modifications in Monkeys are Related
to Absorption Distribution and Excretion of Polychlonnated Biphenyls." Archives of
Environmental Contamination and Toxicology, 2(l):86-95.
Attar, E.N., and EJ. Maly. 1982. Acute toxicity of cadmium, zinc, and cadmium-zinc
mixtures to Daphnia magna. Arch. Environ. Contam. Toxicol. 11: 291-296.
Barren, M.G. 1990. "Bioconcentration." Environ. Sci. Technol. Volume 24, Number 11.
Baudo R. 1987 Heavy metal pollution and ecosystem recovery, pp. 325-352, In O. Ravera's
(ed.) Ecological Assessment of Environmental Degradation, Pollution and Recovery. Elsevier
Science Publishers, The Netherlands.
Bauman, P.C., and J.C. Harshbarger. 1985. "Frequencies of Liver Neoplasia in a Feral Fish
Population and Associated Carcinogens. Marine Env. Research 17(2-4):324-27.
Clayton, G.D. and F.E. Clayton, (eds). 1982. Patty's Industrial Hygiene and Toxicology.
Volume 2A. Third Revised Edition. John Wiley and Sons: New York, NY.
Eisler, R. 1985. Cadmium hazards to fish, wildlife, and invertebrates: A synoptic review. U.S.
Fish and Wildlife Service Biological Report.
Eisler, R. 1986a. Dioxin hazards to fish, wildlife, and invertebrates: A synoptic review. U.S.
Fish and Wildlife Service Biological Report.
Eisler, R. 1986b. Polychlonnated biphenyl hazards to fish, wildlife, and invertebrates: A
synoptic review. U.S. Fish and Wildlife Service Biological Report.
Eisler, R. 1987a. Mercury hazards to fish, wildlife, and invertebrates: A synoptic review. U.S.
Fish and Wildlife Service Biological Report.
Eisler, R. 1987b. Poly cyclic Aromatic Hydrocarbons hazards to fish, wildlife, and
invertebrates: A synoptic review. U.S. Fish and Wildlife Service Biological Report.
-------
CHARACTERISTICS AND Toxicnr OF CONTAMINANTS OF CONCERN > A-23
Ellenhora, M.J. and D.G. Barceloux. 1988. Medical Toxicology Diagnosis and Treatment of
Human Poisoning. Elsevier Science Publishing Company, Inc.: New York, NY.
Friberg, L., G.F. Nordberg, and V.B. Vouk. 1986. Handbook on the Toxicology of Metals,
Volume II. Elsevier Science Publishers: Amsterdam.
Goodman and Oilman. 1980. Pharmacological Basis of Therapeutics, 7th Edition.
Hem, J.D. 1972. Chemistry and occurrence of cadmium and zinc in surface water and
groundwater. Water Resources Research 8:661-679.
Hutzinger O., et al. 1972. "Polychlorinated Biphenyls: Metabolic Behavior of Pure Isomers in
Pigeons, Rats and Brook Trout." Science, 178:312 (as cited in U.S. EPA, 1980).
Jones, M.M., and M.G. Cherian. 1990. Review paper: The search for chelate antagonists for
chronic cadmium intoxication. Toxicology 62:1-25.
Limburg, K:E., Moran, M.A., and McDowell, W.H. 1986. The Hudson River Ecosystem.
Springer-Verlag New York Inc.: New York, NY.
Mance, G. 1987. Pollution Threats of Heavy Metals in Aquatic Environments. Elsevier
Science Publishing Co., Inc., New York, NY. 362 p.
Marshall, J.S., and D.L. Mellinger. 1980. Dynamics of cadmium-stressed plankton
communities. Can. J. Fish. Aquat. Sci. 403-414.
Matthews, H.B. and Kato, S. 1979. "The Metabolism and Disposition of Halogenated
Aromatics." Ann. N.Y. Acad. Sci. 320:131-138.
Olson, J.R.; T.A. Gasiewicz and R.A. Neal. 1980. "Tissue Distribution, Excretion, and
Metabolism, of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) in the Golden Syrian Hamster."
Toxicology and Applied Pharmacology. 56: 78 -85. ;
Peterle, T.J. 1991. Wildlife Toxicology. Van Nostrand Reinhold: New York, NY.
Poland, A. and E. Glover. 1980. "2,3,7,8-Tetrachlorodibenzo-p-Dioxin: Segregation of
Toxicity with the Ah Locus." Mol Pharmacol. 17(1): 86-94.
Ravera, O. Influence of heavy metals on the reproduction and embryonic development of
freshwater pulmonates (Gastropoda; Mollusca) and cladocerans (Crustacea; Arthropoda).
Comparative Biochemistry and Physiology 100C(l/2): 215-219.
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN •> A-24
Sax, N.I. and RJ. Lewis. 1989. Dangerous Properties of Industrial Materials. Seventh
Edition. Van Nostrand Reinhold Co.: New York.
Sorensen, E.M. 1991. Metal Poisoning in Fish. CRC Press: Boca Raton, FL.
Tamashiro, H., M. Arakaki, H. Akagi, M. Futatsuka, and L.H. Roht. 1985. "Mortality and
Survival for Minimata Disease." International Journal of Epidemiology, 14(4):582-588.
Taylor, E.J. (editor). 1988. Borland's Illustrated Medical Dictionary. W.B. Saunders Co.
Harcourt Brace Jovanovich, Inc. Philadelphia, PA.
Thurston, R.V., R.C. Russo, C.M. Ferterolf, T.A. Edsall, and Y.M. Barber Jr. 1978. Review of
the EPA Red Book: Quality criteria for water. Water Quality Section, American Fisheries
Society, Bethesda, MD.
Tiramermans, K., E. Spijkerman, M. Tonkes, and H. Covers. 1992. "Cadmium and Zinc
Uptake by Two Species of Aquatic Invertebrate Predators from Dietary and Aqueous
Sources." In: Can. J. Fish. Aquat. Sci., Vol. 49.
U.S. Environmental Protection Agency (Office of Water Regulations and Standards). October
1980. Ambient Water Quality Criteria for Selenium. Washington, DC: EPA 440/5-80-070.
U.S. Environmental Protection Agency. 1984. Mercury Health Effects Update, Health Issue
Assessment. Office of Health and Environmental Assessment. Research Triangle Park, NC.
EPA-600/8-84-019F.
U.S. Environmental Protection Agency (Office of Water Regulations and Standards). January
1985a. Ambient Water Quality Criteria for Cadmium - 1984. Washington, DC: EPA 440/5-84-
032.
U.S. Environmental Protection Agency (Office of Water Regulations and Standards). January
1985b. Ambient Water Quality Criteria for Chromium - 1984. Washington, DC: EPA 440/5-
84-029.
U.S. Environmental Protection Agency (Office of Water Regulations and Standards). January
1985c. Ambient Water Quality Criteria for Copper - 1984. Washington, DC: EPA 440/5-84-
031.
U.S. Environmental Protection Agency (Office of Water Regulations and Standards). January
1985d. Ambient Water Quality Criteria for Mercury - 1984. Washington, DC: EPA 440/5-84-
026.
-------
CHARACTERISTICS AND TOXICITY OF CONTAMINANTS OF CONCERN * A-25
U.S. Environmental Protection Agency. September 1985e. Health Assessment Document for
Poly chlorinated Dibenzo-p-Dioxins. Office of Health and Environmental Assessment.
Washington, D.C. EPA/600/8-84/014F. Final Report.
U.S. Environmental Protection Agency (Environmental Criteria and Assessment Office). 1987.
Drinking Water Criteria Document for Poly chlorinated Biphenyls (PCBs). Cincinnati, OH.
ECAO-CIN-414.
U.S. Environmental Protection Agency (Office of Health and Environmental Assessment).
April 1992. Integrated Risk Information System (IRIS). Washington, DC.
Van Miller, J.P.; R.J. Marlar and J.R. Allen. 1976. "Tissue Distribution and Excretion of
Tritiated Tetrachloro-dibenzo-p-Dioxin in Non-Human Primates and Rats." Food Cosmet.
Toxicol. 14(1): 31-34.
Wicklund, A. 1990. Metabolism of Cadmium and Zinc in Fish. Doctoral dissertation at
Uppsala University.
-------
-------
APPENDIX B
CHARACTERIZATION OF THE POSITIVE ECONOMIC
IMPACTS OF THE GUIDANCE ON THE GREAT LAKES REGION
Public comments on the proposed Guidance asserted that the Guidance would impose
substantial expenditures on the Great Lakes Region, and that these costs would have drastic
negative economic impacts such as decreased competitiveness; inhibited economic growth,
including the discouragement of efforts to expand production to prerecession levels; a loss of
markets and jobs; increased manufacturing costs; and limits to existing businesses. Generally,
the comments suggested that the Guidance would impose enormous compliance costs on
businesses, municipalities, and taxpayers, which will have a negative impact on the quality of
life in the Great Lakes Basin.
Estimating the impact of the Guidance on the economy of the Great Lakes region requires a
detailed econometric model of the region's economy. EPA did not perform an economic
impact analysis of the Guidance, however, an econometric analysis was performed
independent of the RIA for the Council of Great Lakes Governors (DRI/McGraw-Hill, 1993).
This analysis found that the Guidance would have a nearly imperceptible impact on the
region's economy in a worst case scenario. The estimated costs for tide worst case scenario
exceeded those estimated by EPA for the proposed Guidance by a lairge margin.
Manufacturing output was estimated to fall by 0.008% to 0.337% over a range of four
scenarios evaluated, while personal income loss was estimated at between 0.002% and
0.094% for these scenarios. As a result, the study concluded that the region was able to
"afford" the Guidance.
The estimated costs of the final Guidance presented in Chapter 4 of this report reflect a
downward revision from the estimated costs for the proposed Guidance. As the revised costs
are considerably less than the costs used in the impact analysis described above, the
conclusion that the impact will be minimal is likely to hold. In addition, some sectors of the
economy may benefit from Guidance-related expenditures. This section provides a
characterization of the potential positive economic impacts of the Guidance for one scenario
that uses the low cost estimate. First, the estimated costs are broken down by component to
determine which industries are likely to benefit from Guidance-related expenditures. Then, the
presence of these industries in the Great Lakes region is described to provide an indication of
the region's ability to benefit from compliance expenditures.
-------
POSITIVE ECONOMIC IMPACTS OF THE GUIDANCE > B-2
B.1 ESTIMATED COSTS OF THE GUIDANCE
The low end cost estimate indicates that the Guidance will cost direct and indirect dischargers
in the Great Lakes basin $61.0 million per year. Table B-l shows the allocation of costs
across direct and indirect dischargers and across major cost categories. According to these
estimates, the implementation of waste/chemical minimization techniques are the largest cost
component for direct and indirect dischargers; roughly 70.7% of the estimated total
annualized costs are for minimization studies. The second largest cost category is operating
and maintenance costs, comprising 16.2% of costs for direct and indirect dischargers. By
comparison, initial capital costs are rather small (5.6% for direct and indirect dischargers).
Table B-l
Estimated Annualized Costs of the Final Guidance to Direct and Indirect Dischargers
Low Estimate
(Thousands of 1994 Dollars)
Cost Category
Capital
Annual Monitoring
O&M
Special Monitoring
Minimization Studies
Unspecified Regulatory Relief Mechanisms
Total
Direct
Discharger
Costs
$2,226
$287
$6,515
$1,077
$28,495
$2,534
$41,134
Indirect*
Discharger
Costs
$1,154
$159
$3,343
$557
$14,686
$19,900
Total
Costs
$3,380
$446
$9,858
$1,634
$43,181
$2,534
$61,034
Percent
of Total
5.6%
0.7%
16.2%
2.7%
70.7%
4.2%
100%
Source: SAIC, 1995. .
* Indirect discharger costs are allocated across the cost categories in the same proportion as the
direct discharger costs.
Note: Detail may not add to total due to rounding.
B.2 DESCRDTTION OF COST COMPONENTS AND ASSOCIATED INDUSTRIES
Table B-2 provides a description of the components for each cost category shown in
Table B-l. In Table B-3, these components are associated with standard industrial
classification (SIC) codes for industries likely to provide these goods and services. Table B-3
may not include all of the potentially impacted industries, however, it includes the industries
-------
POSITIVE ECONOMIC IMPACTS OF THE GUIDANCE > B-3
Table B-2
Description of Components by Cost Category
Category
Capital
Annual Monitoring
O&M
Special Monitoring
Minimization Studies
Components
Administration and laboratory facilities1
Garage and shop facilities2
Line segregation (plant modifications to segregate wastewater
streams)
Yardwork
Piping
Instrumentation
Land
Engineering
Legal, fiscal, and administrative
Interest during construction
Contingency
Wastewater sampling
Labor and materials (e.g., wastewater treatment chemicals)
Wastewater sampling
Research and analysis of waste stream (including sampling)
pollutant minimization
for
1 No investment cost was included for this item: it was assumed that there was already an
existing building and space for administration and laboratory functions.
2 No investment cost was included for this item: costs were assumed to be a part of normal plant
costs and were not allocated to the wastewater treatment system.
Source: Personal communication with J. Parker and D. Hair, SAIC, 1995.
that are expected to receive the largest shares of compliance expenditure. For example,
engineering services and testing laboratories (SIC codes 8711 and 8734, respectively), are
likely to provide the research and testing services required for the minimization studies and
monitoring activities.1 Minimization studies and various monitoring activities account for
almost three-fourths of total Guidance expenditures. In comparison, construction and
structures are expected to comprise a relatively small share of expenditures.
Businesses not involved in the water pollution abatement industry will also be included in these
SIC codes, however, there is no greater level of industrial detail available.
-------
POSITIVE ECONOMIC IMPACTS OF THE GUIDANCE » B-4
Table B-3
Likely Industries Providing Goods and Services Included in Cost Components
O&M
Capital
....-•
st
n Studies
litoring
litoring
.•
Percent
of Costs
70.7%
2.7%
0.7%
16.2%
• 5.6%
SICC
871
873
873
873
289
327
154
Parts of Maj
17
28
32
33
Description
Engineering Services
Testing Laboratories
Testing Laboratories
Testing Laboratories
Chemicals and Chemical
Preparations, n.e.c.
Lime
General Contractors-Industrial
Buildings and Warehouses
Construction-Special Trade
Contractors
Chemicals and Allied Products
Stone, Clay, Glass, and
Concrete
Primary Metal Industries
n.e.c. = not elsewhere classified
B.3 GREAT LAKES SHARE OF ENGINEERING SERVICES INDUSTRY AND
TESTING LABORATORIES
The Great Lakes region is likely to receive a sizable portion of Guidance-related expenditures
because it is positioned to fill a large percent of the demand for research and testing resulting
from the Guidance. Fairly large shares of the related industries are located in the region. In
addition, factors such as time response and level of control make local firms more desirable
than distant providers (D. Hair, SAIC, personal communication, June, 1994). For the same
reasons, it is also likely that much of the growth in these industries resulting from increased
demand would occur in the region.
As an indicator of industry share, regional employment and payroll statistics for selected
industries with were compared with national statistics. Table B-4 shows that the eight Great
Lakes states account for 27.1% of total national employment and 27% of the national payroll
-------
POSITIVE ECONOMIC IMPACTS OF THE GUIDANCE > B-5
Table B-4
Great Lakes Basin Share of Engineering Services Industry in 1991 (SIC Code 8711)
State
Wisconsin
Illinois
Pennsylvania
Michigan
Ohio
Indiana
New York
Minnesota
Basin
U.S.
Number of
Employees
8,744
20,849
43,283
28,306
21,375
7,527
36,673
9,723
176,480
652,056
% of ILS. Total
1.3%
3.2%
6.6%
4.3%
3.3%
1.2%
5.6%
1.5%
27.1%
100.0%
Annual Payroll
(Thousands)
307,432
865,190
1,641,,705
1,106,,576
769,761
241,376
1,581,,820
361,481
6,875,,341
25,485,311
% of U.S.
Total
1.2%
3.4%
6.4%
4.3%
3.0%
0.9%
6.2%
1.4%
27.0%
100.0%
Source: U.S. Department of Commerce, 1991.
in the engineering services industry (SIC 8711). These states comprise an even greater share,
approximately 31% (Table B-5), of the national employment and payroll for testing
laboratories (SIC 8734). In terms of taxable receipts, data from 19SI7 indicate the eight states
account for 26.3% of nationwide receipts in engineering services and 23.4% in testing
laboratories (Table B-6).
Total annualized expenditures for minimization studies and monitoring (about $45.3 million)
are a significant portion of total Guidance-related expenditures. Thus, even using conservative
estimates of the Great Lakes region's share of these expenditure categories results in a
significant amount of regional expenditures. For example, share estimates of 25% to 50% of
these expenditures remaining in basin states imply annual revenue to these regional industries
of about $11.3 million to $22.7 million, which is between 19% and 37% of the total annual
compliance cost.
-------
POSITIVE ECONOMIC IMPACTS OF THE GUIDANCE > B-6
Table B-5
Great Lakes Basin Share of Testing Laboratories Industry in 1991 (SIC Code 8734)
State
Wisconsin
Illinois
Pennsylvania
Michigan
Ohio
Indiana
New York
Minnesota
Basin
U.S.
Number of
Employees
975
3,832
3,294
2,385
3,674
635
4,101
667
19,563
63,978
% of U.S. Total
1.5%
6.0%
5.1%
3.7%
5.7%
1.0%
6.4%
1.0%
30.6%
100.0%
Annual Payroll
(Thousands)
21,242
108,433
83,811
88,043
106,176
15,221
117,036
17,821
557,783
1,830,521
% of U.S.
Total
1.2%
5.9%
4.6%
4.8%
5.8%
0.8%
6.4%
1.0%
30.5%
100.0%
Source: U.S. Department of Commerce, 1991.
Table B-6
Engineering Services Taxable Receipts (1987)
(Dollars in Thousands)
State
New York
Ohio
Pennsylvania
Michigan
Minnesota
Indiana
Illinois
Wisconsin
Basin
U.S.
Testing
Laboratories
(SIC 8711)
$2,817,451
$980,388
$3,421,516
$1,407,709
$382,504
$305,774
$1,249,670
$374,657
$10,939,669
$41,614,602
% of US Total
6.8%
2.4%
8.2%
3.4%
0.7%
0.7%
3.0%
0.9%
26.3%
100.0%
Testing
Laboratories
(SIC 8734)
$133,880
$117,860
$104,007
$63,683
$26,770
$18,028
$78,816
$17,766
$560,810
$2,394,674
% of US Total
5.6%
4.9%
4.3%
2.7%
0.8%
0.8%
3.3%
0.7%
23.4%
100.0%
Source: U.S. Department of Commerce, 1987.
-------
POSITIVE ECONOMIC IMPACTS OF THE GUIDANCE »• B-7
B.4 REFERENCES
DRI/McGraw-Hill. 1993. The Great Lakes Water Quality Initiative: Cost Effective Measures
to Enhance Environmental Quality and Regional Competitiveness. Prepared for the Council of
Great Lakes Governors, Chicago, IL. July.
SAIC, 1995. Revised Assessment of Compliance Costs Resulting form Implementation of the
Final Great Lakes Water Quality Guidance.
U.S. Department of Commerce. 1987. Census of Service Industries. Prepared by the Bureau
of the Census, Washington, DC.
U.S. Department of Commerce. 1991. County Business Patterns. Prepared by Economics and
Statistics Administration, Bureau of the Census, Washington, DC.
-------
-------
APPENDIX C
SOURCES OF AIR EMISSIONS
Chlordane
Large amounts of chlordane are not released directly to air from either use or manufacturing
(Syracuse Research Corporation, 1989a). However, chlordane may be present in the
atmosphere due to volatilization from soils and water, and from wind, erosion. Once in the
atmosphere, chlordane will degrade by photolysis and oxidation; rainout and dry deposition
are not expected to be significant (Syracuse Research Corporation, 1989a). Although
registration of chlordane for all uses above the ground was stopped in 1983, any chlordane
that had already been bought or was still on the store shelves could be used above the ground
until April 1988. However, starting in the fall of 1987, chlordane could only be used outside
buildings to kill termites. All other uses were stopped (Syracuse Research Corporation,
1989a). Chlordane will bioaccumulate in aquatic organisms (Syracuse Research Corporation,
1989a).
Copper
Sources of atmospheric copper emission include iron and steel production, coal and oil
combustion, copper smelting, zinc smelting, copper sulfate production, and the incineration of
municipal refuse and sewage sludge (U.S. EPA, 1992). Other sources of copper in air include
stack emissions of coal burning power plants and welding (U.S. EPA, 1992).
Dieldrin
In the United States there is currently no production or importation of dieldrin. It is, therefore,
difficult to quantify the current use patterns of this insecticide. Due to the cutoff in supplies
and its regulatory status, the use of dieldrin is believed to be minimal. Possible new releases
of dieldrin may come from the use of individually owned stockpiles and the improper
disposal of old stocks at landfill sites. Although a considerable proportion of the dieldrin used
in agriculture may reach the atmosphere, it is probable that atmospheric degradation prevents
accumulation in the air or deposition to soil or water. Air is not a significant contributor to
dieldrin exposure (Dynamac Corporation, 1989). Experimental evidence indicates the potential
for high bioaccumulation and for biomagnification (Dynamac Corporation, 1989).
-------
SOURCES OF AIR EMISSIONS * C-2
Heptachlor
Heptachlor enters the atmosphere primarily through volatilization from treated areas (Sittig,
1991) Persons living or working on or near heptachlor treated areas have a particularly high
inhalation exposure potential (Sittig, 1991). Heptachlor bioconcentrates in organisms at
various trophic levels (Sittig, 1991).
Mercury
Mercury may enter the atmosphere from anthropogenic sources; the major sources of
atmospheric anthropogenic mercury are mining and smelting operations and the combustion of
fossil fuels (Eisler, 1987). The major features of the biogeochemical cycle of mercury include
long-range transport in the atmosphere, wet and dry deposition to land and surface water>
sorption to soil and sediment, and bioaccumulation in terrestrial and aquatic food chains. The
general population is exposed to mercury primarily through ingestion of contaminated
foodstuffs, mainly fish (Clement Associates, 1989).
The residence time of mercury in the atmosphere has been estimated to range from 6 to 90
days to 0.3 to 2.0 years. The atmosphere is the smallest environmental reservoir for mercury,
containing only about 1,000 metric tons (Clement Associates, 1989).
PCBs
At present, the major source of PCB exposure in the general environment appears to be
environmental cycling of PCBs previously released into the environment. This cycling process
involves volatilization from ground surfaces into the atmosphere with subsequent removal
from the atmosphere by wet or dry deposition, and then revolatilization (Syracuse Research
Corporation, 1989b). PCBs may be transported long distances while in the atmosphere and
PCBs are known to bioaccumulate in aquatic organisms (Syracuse Research Corporation,
1989».
Trichloroethylene (TCE)
Most of the TCE used in the United States is released into the atmosphere by evaporative
losses. Vapor degreasing operations are the major source of these evaporative losses. TCE is
not a persistent atmospheric compound. TCE volatilizes rapidly from surface water and
surface soil. Experimental studies have shown that TCE has a low potential for
bioaccumulation (Syracuse Research Corporation, 1989c).
-------
SOURCES OF Am EMISSIONS •> C-3
C.I REFERENCES
Clement Associates. 1989. lexicological Profile for Mercury. Prepared for Agency for Toxic
Substances and Disease Registry (ATSDR), U.S. Public Health Service. December. 169 pp.
Dynamac Corporation. 1989. lexicological Profile for Aldrin/Dieldrin. Prepared for Agency
for Toxic Substances and Disease Registry (ATSDR), U.S. Public Health Service. May. 109
PP-
Eisler, R. 1987. Mercury Hazards to Fish, Wildlife, and Invertebrates: A Synoptic Review.
Prepared for U.S. Department of Interior, Fish and Wildlife Service. April. 90 pp.
Sittig, M. 1991. Handbook of Toxic and Hazardous Chemicals and Carcinogens. Park Ridge,
NJ: Noyes Publications.
Syracuse Research Corporation. 1989a. Toxicological Profile for Chlordane. Prepared for
Agency for Toxic Substances and Disease Registry (ATSDR), U.S. Public Health Service.
December. 141 pp.
Syracuse Research Corporation. 1989b. Toxicological Profile for Selected PCBs
(Aroclor-1260, -1254, -1248, -1242, -1232, -1221, and -1016). Prepared for Agency for Toxic
Substances and Disease Registry (ATSDR), U.S. Public Health Service. June. 135 pp.
Syracuse Research Corporation. 1989c. Toxicological Profile for Trichloroethylene. Prepared
for Agency for Toxic Substances and Disease Registry (ATSDR), U.S. Public Health Service.
October. 139 pp.
U.S. EPA. 1992. Hazardous Substances Database: Copper. October 29.
-------
-------
APPENDIX D
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL
This appendix describes the exposure and food web model used to esitimate fish tissue
concentrations of PCB for Green Bay. Figure 5-1 (in Chapter 5) provides a schematic
representation of the model.
D.I FOOD CHAIN MODEL
i
A bioenergetics-based food web model was utilized to estimate PCB concentrations in fish
tissue for Green Bay. Bioenergetics-based models are composed of a series of equations that
calculate PCB concentrations from partitioning (e.g., benthosisediments; plankton:water) or
intake of contaminated food (e.g., fish consuming contaminated plankton). The specific
equations of the model are presented below for each compartment of the model. Parameters
of the model (e.g., assimilation efficiency) were obtained from the GiBMBS food chain model
(Connolly et al., 1992), or published Great Lakes food web models (e.g., Ram, 1990;
Thomann et al., 1992; Gobas, 1993).
The bioenergetics-based model can be viewed conceptually as a series of compartments linked
by transport pathways (see Figure 5-1). Compartments of the model were defined based on
functional characteristics (e.g., feeding habits), in addition to source compartments (i.e., water,
sediment). In addition to the identification of pathways shown in Figure 5-1, the model also
calculated fish tissue concentrations resulting from exposure to PCBs dissolved in the water
column (e.g., PCB bioconcentration by benthivores, planktivores, omnivores, piscivores). A
functional based model was chosen because an organism's life stage, size, or age may
determine feeding habits of a particular fish species. For example, Lake Michigan salmonids
less than 30 cm are omnivores consuming primarily benthos, while larger salmonids are
piscivores consuming alewives, rainbow smelt, and gizzard shad (Jude et al., 1987).
Model Equations
Model-based equations are presented below for each compartment derived from Thomann et
al. (1992) and Gobas (1993). Parameters are defined below and in Table D-l.
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL >• D-2
Table D-l
Model Parameters (values were varied + 40% in simulations)
Parameter
i
*~s
WCC
Qv
*
f*oc
SQC
Q)W
•»
M
••I
BSF
^uw
^UF
P|
^EX
KEI
Ko
Wv
Qw
a.
FR
Definition
PCB Concentration in Sediment
Water Calibration Coefficient
Water Concentration
Organic Carbon Partition
Coefficient
Sediment Organic Carbon Fraction
Octanol Water Partition
Coefficient
PCB Concentration in Species I
Lipid Fraction in Species I
Biota:Sediment Accumulation
Factor
Uptake Rate Coefficient from
Water
Uptake Rate Coefficient from
Food
Proportion of a Species Consumed
Excretion Rate Coefficient
Egestion Rate Coefficient
Growth Rate Coefficient
Weight (wet) of Predator
Fish Aqueous Phase Transport
Parameter
Assimilation Efficiency
Feeding Rate
Units
mg/kg
unitless
mg/L
L/kg
unitless
unitless
mg/kg (wet)
unitless
unitless
L/(d x kg)
kg food/
(d x kg)
unitless
d'1
d'1
d-1
kg
L/day
unitless
kg food/day
Value
scenario specific
(see text)
scenario specific
(see text)
scenario specific
(see text)
1,585,000
0.03
10,000,000
model estimated
species specific
(see text)
4
calculated
(see text)
calculated
(see text)
see Figure 5-1
species specific
(see text)
species specific
(see text)
species specific
(see text)
species specific
(see text)
calculated
(see text)
0.7
calculated
(see text)
Reference
Bierman et al., 1992 11
; —
Bierman et al., 1992
Mackay, 1992
Studies reviewed by II
Connolly et al., 1992
Jude et al., 1987;
Connolly et al., 1992;
Thomann et al., 1992;
Landry and Stewart,
1993
Thomann et al., 1992
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION. MODEL > D-3
Sediment
The sediment concentration (Cs) was a direct input into the model, and was obtained from
average whole bay model estimates for the upper 4 cm sediments at equilibrium from the
GBMBS (Bierman et al., 1992; DePinto, 1994). The upper 4 cm represents sediments in
intimate contact with the water column and also sediments most likely in contact with
sediment-associated biota. Sediment concentrations used as input are shown in Table D-2.
Table D-2
Model Input Sediment PCBs and Resulting Water PCBs
Model Scenario
1989 Conditions
Baseline
10% Reduction
50% Reduction
Input Sediment PCBs (mg/kg)
0.160
0.131
0.130
0.125
Resulting Water PCBs (ng/L)
1.8
1.0
0.99
0.96
Water
Water concentrations (Cw; dissolved PCBs) were calculated from equilibrium partitioning of
PCBs between water and sediment, adjusted by WCC:
Cw = (Q x WCQI(KOC x Soc)
(1)
The Water Calibration Coefficient (WCC) used to calibrate predicted water concentrations
with expected concentrations reported by Bierman et al. (1992) for Green Bay. For example,
WCC equals 0.364 under steady-state conditions, and WCC equals 0.535 under 1989 non
steady-state conditions. A WCC less than one indicates that PCB concentrations in water are
less than predicted from equilibrium partitioning. Values less than one have been reported in
a variety of Great Lakes systems, indicating this condition is prevalent.
Plankton
Plankton are the micro and macroscopic plants (phytoplankton) and animals (zooplankton)
inhabiting the water column, and are the base of pelagic food webs. Contaminant
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL > D-4
concentrations in phytoplankton and zooplankton are closely coupled because of density
dependent population regulation, and rapid partitioning with PCBs dissolved in water
rConnoliv et al 1992). Phytoplankton and zooplankton were assumed to have equivalent
concentrations (weight.weight) of PCBs because there appears to be limited biomagmfication
Seen these two compacts in Green Bay (Connolly et al, 1992£ For example,
Connolly et al. (1992; p. 4-29) reported that the ratio of zooplankton PCBs: phytoplankton
PCBs ranged from 0.61 to 2.39 for various PCB congeners. Concentrations of PCBs in
plankton (Cm; zooplankton plus phytoplankton) were calculated from equilibrium partitioning
of PCBs between the lipids of plankton (LPN) and PCBs dissolved in water.
Equilibrium concentrations appeared to be a reasonable assumption because plankton appear
to equilibrate rapidly with water (Connolly et al, 1992). A mean hpid content. (LPN) of 1.35/o
for plankton was used in the model. These data were calculated by assuming 80% moisture
(Thomann et al, 1992) and using the reported mean dry weight fraction of hpid in Green Bay
zooplankton (Connolly et al, 1992; Table 4-4; p. 4-23).
Benthos
Benthos are the organisms associated with the sediment. Macroinvertebrate benthosi (e g.,
oligochaetes, crustaceans, insect larvae) generally comprise the largest fraction of the diet of
animals feeding on benthos, and thus may be the most important component of benthic-based
food webs Concentrations of PCBs in benthos were calculated from equilibrium partitioning
of PCBs between the lipids of benthos (3%) and the organic carbon of sediment in Green Bay
(3%), as modified by a biota:sediment accumulation factor (BSF):
r< T v (C /C1 ~\ x RVF (3)
CPW = LBV x (LS/ZOC) x azr
'BN
A BSF of 4 was used to adjust PCB concentrations estimated in benthos because benthic
invertebrates appear to accumulate PCBs at concentrations greater than predicted from
equilibrium partitioning (Connolly et al, 1992).
Benthivores
Benthivores are organisms that consume primarily benthos. For the model, benthivores
represented bottom feeding carp and catfish. Concentrations of PCBs in benthivores (Cv) were
calculated from the ratio of PCB intake (uptake from water, food consumption) to elimination
plus growth dilution as shown below (equation 4):
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL »• D-5
t x Q]
where dietary intake is the sum of the proportion of each ingested dietary item (Pj) (see
Figure 5-1) multiplied by the concentration of PCBs in the dietary item (CT). Equation (4) is a
generalized bioenergics relationship, and is also used to estimate PCB concentrations in
planktivores, omnivores, and piscivores (see below).
The uptake rate (K^) from water is dependent on initial transfer of PCBs from water to the
aqueous phase of the fish, then to the lipid phase of the fish (Gobas, 1993). Kuw was
calculated by:
UW
ir) + (WVI(KOW
where the weight of the benthivore (Wv) was 160 g, and Qw (species specific transport
parameter; equation (4)) was calculated by:
Qw = 88.3 x w°y6 (6)
The uptake rate from food (KUF; equation (4)) is dependent on the assimilation efficiency (a;
fraction of dietary PCBs absorbed), the feeding rate (FR) and organism weight. KUF was
calculated by:
KUF = a x FRIWV ; (7)
FR was weight and temperature dependent, and was calculated by:
FR = 0.022 x #£85 x eOM f T ; (8)
where temperature (T) was 8°C (annual average).
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL > D-6
The excretion rate of PCBs across the gills (KK; equation(4)) was calculated by:
K»V * L, x KOWIQW} + (Wv x I,
(9)
KEX is dependent on initial transfer of PCBs from the lipid to the aqueous phase of the fish,
then to the ambient water (Gobas, 1993).
The fecal egestion rate (KEG; equation (4)) was calculated as 0.25 of the uptake rate from
food:
r =ir_/4 0°)
PCB concentrations in fish are "diluted" by growth; i.e., increasing tissue mass of a fish
results in a decrease in PCB concentration. Growth dilution was estimated by calculating KG
(equation (4)) as a weight dependent coefficient:
KG = 0.000502
(11)
Planktivores
Planktivores are organisms that primarily consume plankton. For the model, planktivores
represented life stages of alewife, rainbow smelt (Landry and Stewart, 1993), and gizzard
shad feeding on plankton (e.g., age 0 fish). Concentrations of PCBs in planktivores were
calculated using equations (4) through (11) above, using planktivore specific values for food
consumption (see Figure 5-1), lipid (8%), and weight (5.4 g).
Omnivores
Omnivores are organisms that consume plankton, benthos, and smaller fish. For the model,
omnivores represented life stages of alewife, rainbow smelt (Landry and Stewart, 1993),
gizzard shad, yellow perch, and other fish feeding on a diversity of prey. Concentrations of
PCBs in omnivores were calculated using equations (4) through (11) above, using omnivore
specific values for food consumption (see Figure'5-1), lipid (7%), and weight (32 g).
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL *• D-7
Piscivores
Piscivores are organisms that primarily consume fish. For the model, piscivores represented
large walleye and brown trout. Concentrations of PCBs in piscivores were calculated using
equations (4) through (11) above, using piscivore specific values for food consumption (see
Figure 5-1), lipid (7%), and weight (2 kg).
D.2 MODELING SCENARIOS
Overview
'j
The modeling framework used a generalized equilibrium model incorporating sediment and
water source compartments. PCB loadings to Green Bay (e.g., point source, atmospheric
deposition) were assumed to distribute between water and sediment, based on equilibrium
partitioning (see equation (1) above). As Green Bay sediment and water concentrations do not
appear to be in steady-state equilibrium with current loadings (Bierman et al., 1992), two
model scenarios were developed to estimate fish tissue concentrations: 1989 conditions and
baseline conditions (see below).
Point Source Loadings to Green Bay
Point source loadings were assumed to contribute 9.4% of the concentrations of PCBs in
Green Bay sediment and water. This value was calculated from the fraction of point sources
of PCBs to total loadings to Green Bay in 1989 [point source/(tributaries + point source +
atmospheric deposition)] as reported by Bierman et al. (1992; Table 9-8; p. 172). The
contribution of PCBs from Lake Michigan is not an explicit component of our model
scenarios. Under current conditions, a net export of PCBs from Green Bay to Lake Michigan
is expected; under steady-state conditions a net import of PCBs from Lake Michigan is
expected in Green Bay (Bierman et al., 1992). The following scenarios were modeled:1
1989 Conditions
1989 conditions were used to calibrate the model because whole bay average PCB
concentrations in sediment and water, and biota (see Figure 5-2) were available for the 1989
year (Beltran, 1992; DePinto, 1994). Concentrations of PCBs in water and sediment were not
at steady-state in 1989; thus 1989 conditions were only used for model calibration.
Sediment and water concentrations are shown in Table D-2; results are presented in Chapter 5.
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL » D-8
Baseline Conditions
Baseline conditions were defined as PCB concentrations in biota, sediment, and water which
are predicted to occur if 1989 conditions were held constant until steady-state occurs.
Expected PCB concentrations in sediment and water under steady-state conditions were
approximations of the values reported by Bierman et al. (1992) for Green Bay. Approximate
values were used because original data were reported graphically, and sediment concentration
units required conversion from ng/L to ng/kg. The conversion factor (4.1 L/kg) was
calculated from the relationship: 160 ^g/kg - 39 ng/L (DePinto 1994; p. 4). The equilibrium
sediment and water concentrations of PCBs from the '1989 whole bay base run' from the
GBMBS (Bierman et al., 1992; Figure 11-6; p. 242) were estimated as 131 ng/kg sediment
and 1 ng/L water (dissolved). These data were defined as baseline for use in assessing
reductions in point source loadings (see Table D-2).
Loadings Reductions
Reductions in point source loadings (10% or 50%) were modeled by calculating the resulting
reduction in sediment and water concentrations, assuming PCBs distribute according to
equation (1) (see Table D-2). For example, a 10% reduction in loadings lowered sediment
concentrations by approximately 1%: 0.1 (fractional load reduction) * 0.094 (fraction of
sediment PCBs from point sources) * 0.131 ng/kg (baseline sediment concentration).
D.3 MODEL CALIBRATION
The data used to calibrate the food web model were PCB concentrations in Green Bay biota,
sediment, and water measured in 1989. A sediment concentration of 160 ng/kg was reported
by DePinto (1994; p. 4) as the approximate whole bay average sediment (upper 4 cm)
concentration of PCBs in 1989. A water concentration of 1.8 ng/L was reported by DePinto
(1994- p 4) as the calculated whole bay average dissolved water concentration of PCBs in
1989.'PCB concentrations in biota (see Figure 5-2) were estimated by using the 1989 whole
bay average sediment and water concentrations of PCBs as input parameters.
Model-estimated PCB concentrations in biota (mg/kg; whole body wet weight) were
compared to 1989 measurements of PCBs (mg/kg) in zooplankton and fish presented by
Beltran (1992). PCB concentrations measured in the biota of Green Bay (Beltran, 1992)
generally decreased between the inner and outer bay; thus the range of reported values were
used for comparison with our model estimates, which are average whole bay estimates.
Concentrations of PCBs measured in alewife, gizzard shad, and rainbow smelt were similar
within the same zone of Green Bay (Beltran, 1992), and were considered to represent
omnivores for our model comparisons. PCB concentrations measured in walleye and brown
trout were similar within the same zone of Green Bay (Beltran, 1992), and were considered to
represent piscivores (top predators; e.g., sport fish) for our model comparisons. Measurements
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL »• D-9
of PCBs in benthivores (e.g., carp, catfish) were not reported for 1939 in the GBMBS
(Beltran, 1992; Connolly et al., 1992), and thus were not compared, to model estimates.
Model-estimated concentrations of PCBs in biota were similar to measured values for Green
Bay biota (see Figure 5-2). Thus the performance of the model was judged to be adequate,
and calibration was not required.
D.4 UNCERTAINTY ANALYSIS
The uncertainty (specifically, variability) in estimates of tissue concentrations were modeled
by incorporating a uniform distribution of parameter values between 60% and 140% of the
mean (i.e. mean ± 40% of model input parameters, excluding Cs). This variability was
selected as an approximation of the variability reported in the scientific literature for food
web parameters (e.g., Connolly et al., 1992; Thomann et al., 1992). The software program
@Risk was used to sample the uniform distribution with 5,000 iterations (sampling events)
and to generate statistical distributions of model parameters. The results were distributions of
fish concentrations, in addition to single point estimates. This facilitated visualization of the
uncertainty associated with estimates of fish tissue concentrations, arid provided quantitative
estimates of the probability of exceeding health thresholds.
D.5 REFERENCES
Beltran, R.F. 1992. Green Bay/Fox River Mass Balance Study: Preliminary Management
Summary. U.S. EPA, Great Lakes National Program Office. Chicago, Illinois.
Bierman V.J., J.V. DePinto, T.C. Young, P.W. Rodgers, S.C. Martin, and R. Raghunathan.
1992. Development and Validation of an Integrated Exposure Model for Toxic Chemicals in
Green Bay, Lake Michigan. U.S. EPA, Grosse He, MI.
Connolly, J.P., T.F. Parkerton, J.D. Quadrini, S.T. Taylor, A.J. Thurnann. 1992. Development
and Application of a Model of PCBs in the Green Bay, Lake Michigan Walleye and Brown
Trout and Their Food Webs. Prepared by Manhattan College, Riverdale, NY for the U.S.
EPA, Gross lie, MI. October 2.
DePinto, J.V. 1994. Role of Mass Balance Modeling in Research and Management of Toxic
Chemicals in the Great Lakes: The Green Bay Mass Balance Study. Great Lakes Research
Review. 1: 1-8.
Gobas F.A.P.C. 1993. A Model Predicting the Bioaccumulation of Hydrophobic Organic
Chemicals in Aquatic Food-webs: Application to Lake Ontario. Ecological Modeling 69: 1-17.
-------
GREAT LAKES EXPOSURE AND BIOACCUMULATION MODEL » D-IO
Jude D J F J Tesar S.F. Deboe, and TJ. Miller. 1987. Diet and Selection of Major Prey
Species by Lake Michigan Salmonids, 1973-1982. Transactions of the American Fisheries
Society 116: 677-691.
Landry B F and D J Stewart. 1993. Ecological Energetics of Rainbow Smelt in the
Laurentian Great Lakes: An Interlake Comparison. Transactions of the American Fisheries
Society 122: 951-976.
Mackay D W Y. Shiu, and K.C. Ma. 1992. Illustrated Handbook of Physical-Chemical
Properties 'and Environmental Fate for Organic Chemicals: Volume I, Monoaromatic
Compounds, Chlorobenzenes, andPCBs. Ann Arbor, MI: Lewis Publishers.
Ram, R.N. 1990. An Exposure Assessment Model for Poly chlorinated Biphenyls in
Food'webs. Dissertation presented to Cornell University.
Thomann R.V JP. Connolly, and T.F. Parkerton. 1992. An Equilibrium Model of Organic
' Chemical'Accumulation in Aquatic Food Webs with Sediment Interaction. Environmental
Toxicology and Chemistry. 11: 615-629.
-------
APPENDIX E
SUPPLEMENTAL INFORMATION TO THE BASINWIDE
RISK ASSESSMENTS
E.I CARCINOGENIC EFFECTS: CALCULATING A UNIT RISK FACTOR
Part of EPA's assessment of carcinogenic substances is a rating of the "weight of evidence"
concerning carcinogenicity. The evidence from human and animal studies is cross-tabulated
into one of five categories: (i) sufficient evidence, (ii) limited evidence, (iii) inadequate
evidence, (iv) no data concerning carcinogenicity, and (v) no evidence of carcinogenicity
(U.S. EPA, 1989). The distinction between categories of'sufficient and limited evidence
requires a qualitative judgement with some guidance published by EPA (1989). Based on the
cross-tabulation of the human and animal data, each agent is classified in terms of
carcinogenicity. The classification scheme is displayed in Table E-l.
Table E-l
EPA Classification Scheme for Possible Carcinogens
Group
A
Bl
B2
C
D
E
Description
Human carcinogen; sufficient human evidence.
Probable human carcinogen; limited human evidence.
Probable human carcinogen; sufficient animal evidence,
inadequate or no human evidence, or no human data.
Possible human carcinogen; limited animal evidence,
inadequate or no human evidence, or no human data.
Not classifiable as to human carcinogenicity; inadequate human and animal evidence or
no data available.
Evidence of noncarcinogenicity for humans.
Source: U.S. EPA, 1986.
The weight of evidence classifications give a qualitative rating to potentially carcinogenic
substances. The quantitative analysis performed by EPA is summarized by the "unit risk
factors." To use the unit risk factors, it is important to understand their definition.
-------
SUPPLEMENTAL INFORMATION TO THE BASINWIDE RISK ASSESSMENTS »» E-2
+ A unit risk factor gives the increase in lifetime cancer risk that occurs to an
individual exposed continuously, from birth to death, to a one unit increase in
the concentration of the agent (U.S. EPA, 1994).
This appendix provides background information for the calculation of cancer (Section E.I)
and systemic risks (Section E.2) to humans. The carcinogenic and systemic effects of the
relevant contaminants are also described.
The unit risk factors provide estimates of low-dose exposure to suspected carcinogens. The
unit risk factors are estimated from data on human exposure to high doses and/or laboratory
high dose exposure to animals. The risk factor for an agent with an "A" rating will be
based—at least in part—on human data, while one with a "B2" rating will be based on animal
data.
Generating a Unit Risk Factor
In deriving .a unit risk factor, the available information about a chemical is evaluated and an
appropriate data set is selected (U.S. EPA, 1989). In choosing appropriate data sets, human
data of high quality are preferred to animal data. If animal data are used, the species that
responds most similarly to humans is preferable. If no clear choice is possible, the most
sensitive species is given the greatest emphasis. Occasionally, when no single study is judged
to be most appropriate, the geometric mean from several studies that collectively support the
estimate may be adopted as the unit risk factor (U.S. EPA, 1989).
Concept of Nonthreshold Effects
Carcinogens are assumed to have no threshold; i.e., there is no level of exposure that does not
pose some probability, however small, of generating a carcinogenic response. Therefore, the
development of a unit risk factor usually involves applying a model to the data set and
extrapolating from the high doses administered to experimental animals to the lower exposure
levels expected for human exposure in the environment (U.S. EPA, 1989). Because the
extrapolation is to a range dose-response function where data are not available, the actual unit
risk factors are uncertain. Hence, a large margin of safety is built into the risk factors.
Generally, the greatest risk that can be justified with a 95 percent confidence interval is
presented. This means that, given the data, it is unlikely that the risk factor is greater than
that reported.
Assessing the Carcinogenic Risks
The chemical-specific unit risks are multiplied by the chemical-specific daily intake to give
the incremental probability of a person developing cancer over a lifetime (70 years) as a
result of exposure to the potential carcinogen. A cancer risk of 1E-06 implies a probability of
one in a million of developing cancer.
-------
SUPPLEMENTAL INFORMATION TO THE BASINWIDE RISK ASSESSMENTS *• E-3
Combined Risks for Multiple Substances
To determine the overall potential for carcinogenic effects from exposure to more than one
chemical, the sum of chemical-specific cancer risks is calculated. This risk summation
assumes that there are no antagonistic or synergistic chemical interactions and that all
chemicals produce the same effect (i.e., cancer). If these assumptions are incorrect, over- or
under-estimation of the actual multiple-substance risk may result (U.S. EPA, 1989).
Carcinogenic Effects Associated with Fish Tissue Contaminants
A summary of the carcinogenic effects of each contaminant assessed is provided below.
Chlordane. Chlordane is classified as a B2 probable human carcinogen. Although human
carcinogenicity data is inadequate, there is sufficient evidence of carcinogenicity in studies in
which benign and malignant liver tumors were induced in four strains of mice of both sexes
and in male rats. This compound is structurally related to other known liver carcinogens. The
quantitative estimate of unit-risk, 1.3E+00 (mg/kg-day)"1 is based on the geometric mean of
the four mouse data sets; mice were the more sensitive species tested and risk estimates for a
similar compound (heptachlor) were similarly derived from mouse tumor data
(U.S. EPA, 1994).
DDT. DDT is classified as a B2 probable human carcinogen. No case studies or
epidemiological investigations are available concerning the carcinogenic effects of DDT in
humans following oral exposure. However, DDT is one of the most widely studied pesticides
in animals, and numerous carcinogenicity studies are available for several animal species. The
unit risk factor of 3.4E-01 (mg/kg-day)"1 is derived from the geometric mean of several data
sets, based on the incidence of liver tumors in mice and rats.
Dieldrin. Dieldrin is a B2 probable human carcinogen with a unit risk factor of 1.6E+01
(mg/kg-day)"1. There is inadequate evidence of human carcinogenicity. Two studies of
workers exposed to aldrin and dieldrin reported no increased incidence of cancer. However,
both studies were limited in their ability to detect an excess of cancer deaths, and exposure
was not quantified. The number of workers studied was small, the mean age of the cohort
was young, the number of expected deaths was not calculated, and the duration of exposure
and of latency was relatively short. The basis for this classification is> the carcinogenicity of
dieldrin in seven strains of mice when dieldrin is administered orally, Dieldrin is also
structurally related to other compounds (aldrin, chlordane, heptachlor, heptachlor epoxide, and
chlorendic acid) which produce tumors in rodents. The animal carcinogenicity data is deemed
sufficient. At different dose levels, the effects range from benign liver tumors to liver
carcinomas with transplantation confirmation, to pulmonary metastases (U.S. EPA, 1994).
Hexachlorobenzene. This compound has been determined to be a B2 probable human
carcinogen. Although human data are inadequate to prove carcinogeniicity, hexachlorobenzene
-------
SUPPLEMENTAL INFORMATION TO THE BASINWIDE RISK ASSESSMENTS » E-4
has been shown to induce liver, thyroid, and kidney tumors in three rodent species when
given orally. The liver appears to be the primary target organ for HCB-mduced cancer (U.S.
EPA, 1994).
Mercury. This compound is classified as a D carcinogen, i.e., it is not classifiable as to
human carcinogenicity. No human data are available, and animal and supporting data are
inadequate (U.S. EPA, 1994).
PCBs PCBs have been classified as B2 probable human carcinogens, based on liver ••
carcinomas in three strains of rats and two strains of mice, with inadequate yet suggestive
evidence of excess risk of liver cancer in humans by ingestion, inhalation, or dermal exposure
(U.S. EPA, 1994).
The data regarding occupational and accidental exposures of humans to PCBs are
inconclusive, due to confounding exposures or lack of exposure quantification. In one such
report where exposure was not quantified, an increased incidence of malignant melanomas
was observed in employees-of a petrochemical plant where Aroclor 1254 was used for 9 years
(Syracuse Research Corporation, 1989). A cohort study of 142 male Swedish capacitor-plant
workers who had been exposed to PCBs for an average of 6.5 years did not indicate any
excess mortality of cancer incidence, however the follow-up time was limited and the cohort
was small (Syracuse Research Corporation, 1989).
Statistically significant excess risk of liver cancer has been reported in a 16 year follow-up of
Yusho patients in Japan. However, this excess risk occurred in only one of the areas where
people had been exposed. Additionally there were confounding exposures to PCDFs and
PCQs; thus the findings are considered tentative (U.S. EPA, 1994). In general, although
studies of humans exposed to PCBs are suggestive of excess risk of liver cancer, the available
epidemiological data do not indicate a consistent tumorigenic effect (Syracuse Research
Corporation, 1989).
Several animal studies report that PCBs produce a carcinogenic response, and that they may
enhance the carcinogenic activities of other substances. Rats fed diets containing 100 ppm
Aroclor 1260 exhibited a higher rate of hepatocellular cancer than controls (Syracuse
Research Corporation, 1989). A group of rats were fed a diet containing Aroclor 1260 at a
concentration of 100 ppm for 16 months, followed by 50 ppm for an additional 8 months,
followed by a control diet for 5 months. In the treated rats examined after 18 months, 95% of
the females and 15% of the males had hepatocellular neoplasms (Syracuse Research
Corporation, 1989).
Hepatocellular carcinomas and liver nodules were observed in mice fed 500 ppm of a PCB
mixture for 32 weeks (Syracuse Research Corporation, 1989). A long-term bioassay of
Aroclor 1260 with rats produced hepatocellular carcinomas when 100 ppm was administered
for 630 days to 200 animals. In another study with 140 rats, females surviving at least
-------
SUPPLEMENTAL INFORMATION TO THE BASINWIDE RISK ASSESSMENTS »• E-5
18 months had a 91% incidence of hepatocellular carcinomas; however, males had only 4%
incidence rate (U.S. EPA, 1994). These results, in conjunction with shorter studies which
resulted in preneoplastic effects, suggest that hepatocellular carcinomas caused by PCBs can
be detected only in long-term experiments at doses low enough to prevent interfering toxicity,
and in studies using large numbers of animals (Syracuse Research Corporation, 1989).
TCDD. The EPA has classified 2,3,7,8-TCDD as a Group B2 probable human carcinogen
when considered alone, and in Group Bl (limited human evidence in addition to sufficient
animal evidence of carcinogenicity) when considered in association with phenoxyherbicides
and/or chlorophenols. The unit risk estimate for TCDD is based on a study showing increased
incidence of tumors of the lungs, liver, hard palate, and nasal turbinates in female rats
maintained on diets containing 2,3,7,8-TCDD for two years. The unit risk is based on two
different pathologic examinations that produced differences in tumor incidence, and hence,
slightly different values. The final unit risk, 1.6E+05 (mg/kg-day)'1, Is an average based on
these separate calculations (Syracuse Research Corporation, 1989).
Toxaphene. Toxaphene is classified as a B2 probable human carcinogen based on inadequate
human evidence and sufficient animal evidence. Oral exposure to this chemical has resulted in
dose-related increases in liver carcinomas in mice and thyroid tumors in rats.
E.2 SYSTEMIC EFFECTS: THRESHOLD CALCULATIONS AND CHEMICAL-
SPECIFIC HEALTH EFFECTS
Animals, including humans, are assumed to have protective mechanisms and function reserve
capacities that must be exceeded before a systemic (noncarcinogenic) effect occurs. Therefore,
a range of exposures exists from zero to some finite value that can be tolerated by the
individual without the manifestation of an adverse effect. A reference dose (RfD) is an
estimate of a daily exposure level for humans, including sensitive subgroups, that is not likely
to cause adverse effects during a lifetime. To develop an RfD, it is necessary to identify the
upper bound of a tolerance range for the most sensitive populations.
Derivation of an Oral RfD
To develop an oral RfD, EPA examines all available studies examining the toxicity of a
chemical following oral exposure. If adequate human data are available, this information is
used as the basis of the RfD. Otherwise, animal study data are used. In the case of using
animal study data, it is necessary to make judgements regarding the relevance and quality of
the experimental studies. The effect characterized by the "lowest observed adverse effect
level" (LOAEL) is referred to as the critical toxic effect.
-------
SUPPLEMENTAL INFORMATION TO THE BASINWIDE RISK ASSESSMENTS * E-6
After the study and critical toxic effect have been selected, EPA identifies the exposure level
that represents the highest level tested at which no adverse effects were demonstrated. This
highest "no observed adverse effect level" (NOAEL) is used to develop the RfD.
Applying Uncertainty Factors
The RfD is derived from the NOAEL by consistent application of uncertainty factors (UF)
and a modifying factor (MF), described as follows:
>• A UF of 10 is used to account for variation in the general population and is
intended to protect sensitive subgroups such as the elderly.
»• A UF of 10 is used when extrapolating from animals to humans, to account for
interspecies variability.
> A UF of 10 is used when a NOAEL derived from a subchronic rather than a
chronic study is used as the basis for a chronic RfD.
> A UF of 10 is used when a LOAEL is used instead of a NOAEL.
In addition to the UFs listed above, a modifying factor (MF) is used.
»• A MF ranging from >0 to 10 is used to reflect a qualitative professional assessment of
additional uncertainties in the critical study and in the entire data base for the chemical
not specifically addressed by the preceding UFs. The default value for the MF is 1.
Calculation of the RfD
To calculate the RfD, the appropriate NOAEL (or LOAEL if a suitable NOAEL is not
available) is divided by the product of all of the applicable uncertainty factors and the
modifying factor. The calculation would be as follows:
RfD = NOAEL or LOAEL/(UFl x UF2 x MF)
Oral RfDs are expressed in units of mg/kg-day.
Assessing the Potential for Systemic Effects
The RfD assumes that there is an exposure threshold below which adverse effects will not
occur. If the exposure level (intake) exceeds this level, adverse effects could potentially occur.
-------
SUPPLEMENTAL INFORMATION TO THE BASINWIDE RISK ASSESSMENTS * E-7
*
This potential for adverse systemic effects is called the hazard quotient (HQ). The calculation
to determine the HQ is:
t
HQ = Intake / RfD
The greater the value of the HQ above one, the greater the potential for adverse effects.
However, the HQ is not a statistical probability; a ratio of 0.001 does not mean that there is a
one in one thousand chance of the effect occurring. Furthermore, the potential for adverse
effects does not increase linearly as the RfD is approached or exceeded.
Combined Risks for Multiple Substances
i
To determine the overall potential for noncarcinogenic effects from exposure to more than
one chemical, a hazard index (HI) approach has been developed. This approach assumes that
simultaneous exposure to several chemicals below their threshold level could result in an
adverse health effect. The HI is equal to the sum of the HQs.
Systemic Effects Associated with Fish Tissue Contaminants
A summary of the critical systemic effects of each contaminant is given below.
Chlordane. The oral RfD for chlordane is based on a chronic study in which rats were fed
chlordane at various dietary levels. The critical systemic effect observed was liver lesions in
females, at 0.27 mg/kg-day. The RfD was based on a NOEL of 0.06 mg/kg-day. An
uncertainty factor of 100 was used to account for the inter- and intrasipecies differences. An
additional UF of 10 was used to account for the lack of an adequate 'reproduction study and
adequate chronic study in a second mammalian species, and the generally inadequate sensitive
endpoints studied in .the existing studies; UFs totaled 1,000. No modifying factor was used. A
medium confidence level was given to the critical study, although the database and the RfD
were given a low level of confidence.
DDT. The oral RfD for DDT was determined from a chronic study in which rats were fed 1
ppm (0.05 mg/kg-day) DDT in their diet for 27 weeks. The critical effect was liver lesions.
An uncertainty factor of 100 was used in this study to account for uncertainty in the
extrapolation of dose levels from laboratory animals to humans and uncertainty in the
* threshold for sensitive humans. No modifying factor was used (U.S. EPA, 1993a).
Dieldrin. The oral RfD for dieldrin is based on a two-year (chronic) study in which rats were
* administered dieldrin at dietary concentrations of 0, 0.1, 1.0, or 10.0 ppm. Based on the intake
assumptions, these dietary levels were presumed to be approximately equal to 0, 0.005, 0.05,
and 0.5 mg/kg-day. Body weight, food intake, and general health remained unaffected
throughout the two year period, although at 10.0 ppm (0.5 mg/kg-day) all animals became
irritable and exhibited tremors and occasional convulsions. No effects were seen in various
-------
SUPPLEMENTAL INFORMATION TO THE BASINWIDE RISK ASSESSMENTS > E-8
heraatological and clinical chemistry parameters. At the end of two years, females fed 1.0 and
10.0 ppm (0.05 and 0.5 mg/kg-day) had increased liver weights and liver-to-body weight
ratios (p < 0.05). Histopathological examinations revealed liver parenchymal cell changes.
These liver lesions were considered characteristic of exposure to an organochlorine
insecticide. The LOAEL was identified as 1.0 ppm (0.05 mg/kg-day) and the NOAEL as
0 1 ppm (0.005 mg/kg-day). An UF of 100 was used to account for uncertainty in the
extrapolation of dose levels from laboratory animals to humans and uncertainty in the
threshold for sensitive humans. No modifying factor was used (U.S. EPA, 1993b).
EPA has placed a low level of confidence in the study used to set the RfD, because it is an
older study for which detailed data are not available. The chronic toxicity evaluation was
relatively complete and supported the critical effect, if not the magnitude of effects.
Reproductive studies were lacking. The RfD and the database of health effects are given a
medium confidence rating, because of the support for the critical effect from other dieldnn
studies and from studies on organochlorine insecticides in general (U.S. EPA, 1993b).
Hexachlorobenzene. The derivation of an oral RfD is based on a chronic feeding study in
which male and female rats (the FO generation) were fed diets containing 0, 0.32, 1.6, 8.0, or
40 ppm of hexachlorobenzene for 90 days prior to mating and until 21 days after parturition
(at weaning). The number of offspring from these matings (Fl generation) was reduced to
50 males and 50 females per dose group at 28 days of age and fed their respective parents
diets. Thus the Fl animals were exposed to hexachlorobenzene and metabolites in utero, from
maternal nursing and from their diets for the remainder of their lifetime (130 weeks). No
hexachlorobenzene-induced adverse effects were reported in the 0.32 and 1.6 ppm HCB Fl
groups, indicating that these levels are NOAELS. The 8.0 ppm Fl groups were reported to
have an increase in liver tissue alterations (p < 0.05). The 40 ppm Fl groups showed
increases (p < 0.05) in pup mortality, liver tissue alterations, and sever chronic nephrosis
(males only) (U.S. EPA, 1993b). The critical effect was liver alterations; a NOAEL of
1.6 ppm (diet) (0.08 mg/kg-day) was used to set the RfD. An uncertainty factor of 100 was
used,--10 for interspecies variability and 10 for intraspecies variability. No modifying factor
was used.
A medium confidence rating was given to the study and to the RfD. The dosing scheme used
in the principle chronic study caused difficulties in determining the true doses received by
each experimental group. The study included extensive evaluation of systemic and neoplastic
pathological endpoints and was critically reviewed before release and publication. The
sensitive endpoint of porphyria was not evaluated in this study, otherwise a high confidence
in the RfD would be assigned. The database for hexachlorobenzene is given a high confidence
rating due to the extensive number of quality research studies available (U.S. EPA, 1994):
Mercury. The basis for the RfD for mercury is data from the Niigata, Japan incident of
mercury poisonings. The estimated threshold blood level was 200 ng Hg/ml blood for the
development of neurological symptoms. Extrapolating the long-term oral dose required to
-------
SUPPLEMENTAL INFORMATION TO THE BASINWTOE RISK ASSESSMENTS *• E-9
applied an uncertainty factor of 10 to estimate an oral RfD of 3.OHIO'4 mg/kg-day (U.S. EPA,
1984).
PCBs. The liver is a principal site of systemic PCB toxicity in animals. Liver effects have
been observed in numerous studies with exposed rats, mice, guinea pigs, rabbits, dogs, and
monkeys, but rats have been tested most extensively. The effects appear to be reversible at
low doses and are similar among species (Syracuse Research Corporation, 1989).
Nonmalignant proliferative lesions were observed in the liver at high frequencies in PCB-
treated rats and mice (Syracuse Research Corporation, 1989). Liver alterations were observed
in monkeys fed diets containing 2.5 and 5.0 ppm Aroclor 1248. However, these effects were
examined by autopsy in only one animal per dose level (Syracuse Research Corporation,
1989). In addition, rats exposed to Aroclor 1254 for 4 to 12 weeks experienced thyroid
alterations. Thyroid effects also appeared to be reversible after cessation of exposure
(Syracuse Research Corporation, 1989).
Studies of PCB-exposed workers provide inconsistent but suggestive evidence for subclinical
increases in serum enzymes that are indicators of possible liver damage. Liver dysfunction
has not been demonstrated in PCB-exposed workers.
TCDD. The oral reference dose for 2,3,7,8-TCDD is based on a three generation reproductive
study in Sprague-Dawley rats receiving 0.001, 0.1 or 0.01 ug of 2,3,7,8-TCDD/kg. The lowest
adverse effect level (LOAEL) was observed to be 0.001 ug 2,3,7,8-TCDD/kg based on a
reduction in gestation index, decreased fetal weight, increased liver to body weight ratio, and
increased incidence of dilated renal pelvis. An uncertainty factor of 1000 was applied to
account for interspecies variation and for the use of a LOAEL (U.S. EPA, 1985).
Toxaphene. The oral LD50 values in the rat range from 60 to 120 mg/kg body weight.
Salivation, vomiting, hyperexcitability, convulsions, and death may occur with acute oral
exposure. In acute and chronic studies, liver damage occurs at high dose levels (1,000 mg/kg)
depending on the species tested. Microsomal enzyme activity has been induced in the rat at
concentrations of 5 mg/kg-day (International Program on Chemical Safety, 1990). Chronic,
high-dose studies show liver and kidney damage, as well as central nervous system effects
(U.S. EPA, 1993c).
E.3 REFERENCES
International Program on Chemical Safety. 1990. Camphechlor Health and Safety Guide.
Prepared by the World Health Organization. 34 pp.
Syracuse Research Corporation. 1989. Toxicological Profile for Selected PCBs (Aroclor-1260,
-1254, -1248, -1242, -1232, -1221, and -1016). Prepared for Agency for Toxic Substances and
Disease Registry (ATSDR), U.S. Public Health Service. June. 135 pp^
-------
SUPPLEMENTAL INFORMATION TO THE BASINWIDE RISK ASSESSMENTS > E-10
U S EPA 1984 Mercury Health Effects Update, Health Issue Assessment. Office of Health
and Environmental Assessment. Research Triangle Park, NC. EPA-600/8-84-019F.
U S EPA. 1985 Drinking Water Criteria Document for 2,3,7,8-Tetrachlorodibenzo-p-Dioxin.
Environmental Criteria and Assessment Office. Cincinnati, OH. EPA Report No. 600/X-84-
194-1.
U S EPA 1989. Risk Assessment Guidance for Superfund, Volume I: Human Health
Evaluation Manual. Prepared by the Office of Emergency and Remedial Response,
Washington, DC.
U.S. EPA. 1993a. Health Effects Assessment Summary Tables. Prepared by Office of Solid
Waste and Emergency Response, Washington, DC.
U S EPA 1993b Integrated Risk Information System (IRIS): On-line database of
chemical-specific toxicity factors. Prepared for U.S. EPA, Office of Research and
Development, Washington, DC.
U.S. EPA. 1993c. Fact Sheet on Drinking Water Chemical Contaminants. Prepared by the
U.S. EPA, Office of Drinking Water.
U S EPA 1994 Integrated Risk Information System (IRIS): On-line database of
chemical-specific toxicity factors. Prepared for U.S. EPA, Office of Research and
Development, Washington, DC.
-------
-------
------- |