PB97-963120
EPA/541/R-97/059
November 1997
EPA Superfund
Explanation of Significant Difference
for the Record of Decision:
Commencement Bay, Near Shore/Tide Flats,
Operable Unit 01 - Sediments &
Operable Unit 05 - Source,
Pierce County, WA
7/28/1997
-------
EXPLANATION OF SIGNIFICANT DIFFERENCES
COMMENCEMENT BAY NEARSHORE/TIDEFLATS SUPERFUND SITE
I. INTRODUCTION
Site Name and Location
Commencement Bay Nearshore/Tideflats Superfund Site
Tacoma, Pierce County, Washington
Operable Unit 01 - Sediments; and Operable Unit 05 - Sources
Lead and Support Agencies
U.S. Environmental Protection Agency (EPA) - Lead Agency for Sediment Remediation
Washington State Department of Ecology (Ecology) - Lead Agency for Source Control; Support
Agency for Sediment Remediation
Puyallup Tribe of Indians • Support Agency
Statutory Authority
Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), Section
117(c) and National Oil and Hazardous Substances Contingency Plan (NCP), Section
300.435(cX2XI).
Purpose
The purpose of this Explanation of Significant Differences (ESD) is to modify the cleanup level
for remediation of marine sediments contaminated with polychlorinated biphenyls (PCBs) at the
Commencement Bay Nearshore/Tideflats (CB/NT) Superfund site. EPA's September 30, 1989,
Record of Decision (ROD) for the CB/NT Site established cleanup levels, called Sediment
Quality Objectives (SQOs), for several problem chemicals found to be causing adverse effects to
human health and the environment at the CB/NT Site. The SQO for PCBs was set at 150 /zg/kg
(micrograms per kilogram or parts per billion) dry weight (DW). The ROD required that the
SQOs be met within ten years after completion of sediment remedial action. The ROD predicted
that, if sediments with PCB concentrations greater than a Sediment Remedial Action Level
(SRAL) of 240 - 300 Mg/kg PCBs were removed, the 150 /ig/kg PCB SQO would be met in 10
years through natural recovery processes. With this ESD, EPA is modifying the PCB SRAL to
450 A*g/kg, to be achieved during cleanup, and the PCB SQO to 300 /zg/kg, to be achieved within
10 years after cleanup.
EPA believes it is appropriate in some circumstances to make changes to existing decisions at
Superfund sites to enhance overall remedy effectiveness and cost-effectiveness, as long as it does
not compromise protectiveness or other objectives of the Superfund program. In this case, EPA
decided to re-evaluate the PCB cleanup level for the CB/NT Site for the following reasons.
-------
During pre-design sampling, new data were collected from the Hylebos Waterway showing that
approximately twice the amount of sediment originally estimated in the ROD would require
cleanup, and that cleanup costs would also be about twice the estimate in the ROD. In addition,
EPA has updated the toxicity information it uses to assess human cancer risks associated with
PCBs.
EPA's reevaluation of the PCB SQO at the CB/NT Site is described in a document entitled
"Reevaluation of Residual Risks Associated with a Range of Sediment PCB Cleanup Levels in
the Hylebos Waterway, Thea Foss Waterway, and Overall CB/NT Superfund Site" (Weston,
1997a) and in EPA's public review draft ESD (EPA, 1997). Based on EPA's reevaluation of the
human health and environmental risks associated with PCBs, and through our evaluation using
EPA's nine Superfund remedy selection criteria, EPA has determined that it is appropriate to
modify the PCB cleanup level to 450 Mg/kg, to be achieved during cleanup, and 300 ^g/kg, to be
achieved within 10 years after cleanup through natural recovery processes.
Cleanup to 450 /zg/kg is expected to result in a post-cleanup average PCB concentration of less
than 150 Mg/kg in all waterways at the CB/NT Site. EPA estimates that the post-cleanup
average PCB sediment concentration after cleanup to 450 A*g/kg will be 75 ^g/kg for the CB/NT
Site, 124 Mg/kg for the Hylebos Waterway, and 108 A*g/kg for the Thea Foss Waterway. PCB
sediment concentrations are expected to be further reduced over time due to natural recovery
processes to approximately 63 Mg/kg for the CB/NT Site, 80 ^g/kg for the Hylebos Waterway,
and 81 Mg/kg for the Thea Foss Waterway.
Administrative Record
This ESD and other documents that EPA relied upon to make this decision are part of the
Administrative Record for the CB/NT Superfund site, which is available to the public at the
following locations:
U.S. Environmental Protection Agency
1200 6th Avenue, Records Center (7th floor)
Seattle, Washington 98101
Tacoma Public Library
Northwest Room
1102 Tacoma Avenue
Tacoma, Washington 98402
II. BACKGROUND
CB/NT ROD
The CB/NT Superfund site is located iit Tacoma, Washington at the southern end of the main
2
-------
basin of Puget Sound. The site includes 10-12 square miles of shallow water, shoreline, and
adjacent land, most of which is highly developed and industrialized. The upland boundaries of
the site are defined according to the contours of localized drainage basins that flow into the
marine waters. The marine boundary of the site is limited to the shoreline, intertidal areas,
bottom sediments, and water of depths less than 60 feet below mean lower low water level. The
nearshore portion of the site is defined as the area along the Ruston shoreline from the mouth of
Thea Foss Waterway to Pt. Defiance. The tideflats portion of the site includes the Hylebos,
Blair, Sitcum, Milwaukee, St. Paul, Middle, Wheeler-Osgood, and Thea Foss Waterways; the
Puyallup River upstream to the Interstate-5 bridge; and the adjacent land areas (EPA, 1989).
(Although the Blair and St. Paul Waterways have been deleted from the CB/NT Site because
cleanups in those waterways are complete, they were included in the human health and
ecological risk revaluation of the PCB cleanup level.)
The Commencement Bay site has been divided into smaller project activities, called operable
units (OU), in order to more effectively manage the overall cleanup of the site. In a September
30, 1989 ROD, EPA designated two operable units for the cleanup of the nearshore/tideflats
portion of Commencement Bay: source control (OU 5), which focuses on efforts to control
upland discharges or releases to the Bay, and sediment remediation (OU 1), which addresses the
cleanup of the contaminated marine sediments in Commencement Bay. The Washington
Department of Ecology (Ecology) is the lead agency for source control and EPA is the lead
agency for sediment remediation.
EPA placed the CB/NT Site on the National Priorities List (NPL) of sites requiring investigation
and cleanup under EPA's Superfund Program on September 8, 1983. A remedial
investigation/feasibility study (RI/FS) was completed by Ecology in 1988. The RI/FS concluded
that sediments in the nearshore/tideflats area were contaminated with a large number of
hazardous substances at concentrations greatly exceeding those found in Puget Sound reference
areas. In the RI, a multi-step decision-making process was used to identify problem chemicals,
and to identify and prioritize problem areas where these chemicals were present at
concentrations which are harmful to humans and wildlife. Over 50 problem chemicals and nine
high priority problem areas were identified by this process.
In the 1989 ROD, EPA selected a remedial action for eight of the nine sediment problem areas.
These problem areas are: 1) Mouth of Hylebos Waterway, 2) Head of Hylebos Waterway, 3)
Sitcum Waterway, 4) St Paul Waterway, 5) Middle Waterway, 6) Head of Thea Foss (formerly
City) Waterway, 7) Mouth of Thea Foss Waterway, and 8) Wheeler-Osgood Waterway. The
ninth problem area off-shore of the Asarco smelter will be addressed in a separate ROD.
PCBs, along with a number of other chemicals, were identified as problem chemicals in the
Mouth and Head of Hylebos Waterway problem areas. PCBs have been detected at
concentrations as high as 24,000 Atg/kg in the Hylebos Waterway. PCBs are also present in the
Thea Foss Waterway problem areas, however, current information indicates that cleanup of other
chemicals such as polynuclear aromatic hydrdcarbons (PAHs) and metals will also encompass
-------
PCB-contaminated areas. PCBs are not widely distributed in elevated concentrations in other
CB/NT problem areas. -
The 1989 ROD describes a sediment remediation process which includes a combination of
natural recovery and active sediment cleanup. For those areas in which modeling indicates that
SQOs will not be achieved through natural recovery processes within ten years after sediment
remediation, the ROD provides for confining and isolating the contaminated sediments by using
one of four disposal options: in-place capping, dredging and confined aquatic disposal, dredging
and nearshore disposal, or dredging and upland disposal. Natural recovery is the process
whereby sediment concentrations in the upper sediment layers are reduced over time after source
control and cleanup of highly contaminated sediments through mixing with and burial by more
recently deposited clean sediments, as well as other natural processes such as biodegredation and
diffusive loss to the water column. Other components of the selected remedial action for the
eight CB/NT problem areas are: source control, site use restrictions, and long-term monitoring.
1989 ROD Cleanup Goab
The cleanup goal for the Commencement Bay problem areas is reduction of contaminant
concentrations in sediments to levels that will support a healthy marine environment and will
protect the health of people eating seafood from the Bay. The ROD designated biological test
requirements and associated sediment chemical concentrations referred to as Sediment Quality
Objectives (SQOs) in order to achieve this goal. The goal is established to allow a diverse range
of uses in the bay including industrial, commercial, navigation, fisheries, and recreation.
SQOs for all problem chemicals were set based on an evaluation of the ecological and human
health risks posed by these chemicals. The SQO for PCBs was based on the human health risk
assessment. SQOs for all other chemicals were based on the ecological risk assessment, because
the ecologically-based cleanup levels were determined to be also protective of human health.
Ecological Risk-Based Cleanup Goals
The chemical SQOs for protection of aquatic life were set using the Apparent Effects Threshold
(AET) method. An AET is the sediment concentration of a crtemical above which statistically
significant biological effects are always observed in the test organism used to generate AET
values. In other words, if any chemical exceeds its AET value for a particular biological
indicator, then an adverse biological effect is predicted for that indicator. The three biological
effects used to define the AET-derived SQOs were benthic infauna abundance, amphipod
mortality, and oyster larvae abnormality. This method has subsequently been used, with some
modifications, to develop the State of Washington's Sediment Management Standards (SMS -
Chapter 173-204 WAC). The AET method predicted that a sediment PCB concentration of
1,000 Mg/kg (dry weight) would be protective of aquatic life for the species tested. The AET
method does not address bioaccumulation, and thus may underestimate risks to organisms who
eat invertebrates or fish contaminated with bioaccumulative compounds like PCBs. It was
-------
determined that the SQO for PCBs should be set based on the risks to human health from eating
PCB-contaminated seafood, because a lower PCB cleanup level was necessary to protect human
health.
Human Health-Based Cleanup Goals
Human health risks from consumption of contaminated seafood at the CB/NT Site were
evaluated in the 1988 RI/FS using chemical analysis of English sole (a bottom-dwelling flatfish)
muscle tissue, English sole livers, and crab muscle tissue. Of the more than 100 chemicals
analyzed, only PCBs were measured in seafood at concentrations significantly greater than
background and sufficiently high to pose a potential threat to human health due to fish
consumption. The risk assessment estimated a lifetime excess cancer risk of 6 x 10'3, or 6 in
1 ,000, assuming that one pound of Commencement Bay fish are consumed each day. This
means that, at most, a person has a 6 in 1,000 chance of getting cancer over his or her lifetime
from eating one pound per day of fish with the PCB contaminant levels measured in
Commencement Bay. A lifetime excess cancer risk of 2 x 10"*, or 2 in 10,000, was estimated for
a person eating one pound of Commencement Bay fish per month. The analysis focused on
cancer risks as the most conservative estimate of risks to human health. The risk assessment
estimated cancer risks only, because a PCB cleanup level based on cancer risks was shown to be
protective of non-cancer related health risks as well.
These risks were estimated based on measured concentrations of total PCBs in fish. Maximum
PCB concentrations in English sole muscle tissue were found in the Hylebos Waterway at 1,300
Mg/kg (wet weight). English sole from the Hylebos Waterway had an average PCB concentration
of 332 Mg/kg, approximately ten times higher than the average concentration found in fish from
the reference area at Carr Inlet (36 Mg/kg). The CB/NT Site-wide average PCB concentration in
English sole muscle tissue was 210
Based on the human health risk assessment, a sediment SQO of 1 50 Mg/kg (dry weight) total
PCBs was set using the following method:
• An equilibrium partitioning approach was used to estimate that the PCB fish muscle
tissue concentration of 36 /ig/kg (wet weight) found in the Carr Inlet reference area
would be associated with a sediment concentration of 30 A*g/kg (dry weight) PCBs. A
post-cleanup geometric mean sediment concentration of 30 Atg/kg PCBs was set as a goal
for the CB/NT Site.
• The overall post-cleanup sediment concentration was calculated based on the geometric
mean of the post-cleanup data set assuming that PCB concentrations would be reduced to
20 Mg/kg in all areas where PCB-contaminated sediments were remediated. It was
determined that cleanup of areas with sediment PCB concentrations greater than 150
M g/kg would achieve an average post-cleanup residual PCB concentration of 30 Mg/kg in
sediments.
-------
By this method, EPA calculated that a PCB SQO of 150 Mg/kg would result in attainment of
PCB concentrations in fish tissue similar to those in Puget Sound reference areas (36 yug/kg).
The lifetime excess cancer risk associated with consumption of 12.3 grams offish per day
(approximately one pound per month) at the reference PCB concentration of 36 Mg/kg was
estimated to be 4 x 10'3. EPA's goal for Superfund cleanups is to reduce human health risks to
within a range of 10"4 to 10"* (or 1 in 10,000 to 1 in 1,000,000).
III. DESCRIPTION OF AND BASIS FOR THE SIGNIFICANT DIFFERENCES
EPA started its reevaluation of the PCB SQO in 1996, by calculating the human health risks
associated with PCBs remaining in sediments after cleanup to potential sediment cleanup levels
ranging from 50 jzg/kg to 900 Mg/kg PCBs. That analysis used the risk assessment assumptions
and equation used in the 1989 ROD, and updated it with current sediment quality data (Weston,
1996). This report was sent to several parties for review, and EPA received 20 comment letters.
Many of the commentors asked EPA to update the risk assessment using current methods and
assumptions. Several commentors also asked EPA to address ecological risks in its updated risk
evaluation. The National Oceanic and Atmospheric Administration (NOAA) and the U. S. Fish
and Wildlife Service (FWS) provided information on recent studies showing effects of PCBs on
wildlife, as well as site-specific studies in the Hylebos Waterway. Based on these comments,
EPA updated the ecological and human health risk evaluations (Weston 1997a), as discussed
below.
In its analysis, EPA evaluated human health and ecological risks remaining immediately after
cleanup, and in ten years after cleanup. EPA compared potential PCB cleanup levels to the PCB
SRAL in the 1989 ROD of approximately 300 Mg/kg. Reduction in risk over time due to natural
recovery is also discussed. Because PCBs remain in tissues for a long time after exposure, it is
important that exposure to PCBs in sediments is reduced immediately after cleanup, as well as in
the long term.
The area evaluated for both the human health and ecological risk evaluations is the marine
portions of the CB/NT Site as it is defined in the 1989 ROD. This includes the seven waterways
at the southeast corner of Commencement Bay, and the shoreline, intertidal areas, bottom
sediments, and water of depths less than 60 feet below mean lower low water level. The area
considered in the risk evaluations in shown in Figure 1. Because there are no known sources of
PCB contamination in Commencement Bay outside of the CB/NT Site, exposure to PCBs by
people who fish and marine organisms which utilize other areas of Commencement Bay will be
lower than estimated here.
Ecological Risk Evaluation
Several reviewers to the revised human health evaluation pointed out that EPA should reevaluate
the ecological effects associated with a variety of potential cfeanup levels using new information
-------
developed since the ROD was signed. To address ecological concerns, EPA used information
provided by NOAA, FWS, and other sources to evaluate potential threats to wildlife, including
invertebrates, fish, and piscivorous (fish-eating) birds, at a range of PCB sediment cleanup
levels. A summary of the analysis (Weston, 1997a) is provided below.
Invertebrates
The 1989 ROD indicated that a 1,000 ^g/kg PCB sediment cleanup level would be protective of
aquatic invertebrates based on the AET approach, using direct measurements of benthic
community abundance in Commencement Bay and in other areas in Puget Sound. Ecology
subsequently used the AET approach as part of its Sediment Management Standards (SMS) to
set a total organic carbon-normalized minor adverse effects level of 65 mg PCBs/kg carbon
based on benthic infauna abundance, equivalent to EPA's 1,000 Mg/kg (dry weight or DW).
Ecology also set a no adverse effects level for PCBs equivalent to 130 /zg/kg (DW), based on the
Microtox™ test. In addition, a variety of effects levels indices have been developed by other
agencies based on compilations of nationally collected data, for example, NOAA's effects range-
median (ER-M) of 180 Mg/kg (DW) PCBs. These indices, for the most part, fall within the range
set by Ecology's no adverse effects and minor adverse effects levels (130 to 1,000 /zg/kg PCBs).
Fishes
In 1994 and 1995, the NOAA National Marine Fisheries Service conducted a series of
investigations on behalf of Commencement Bay Natural Resource Trustees to determine
contaminant exposure and associated injuries to Hylebos Waterway fish. NOAA measured
juvenile salmonid and adult English sole exposure to toxic chemicals in Hylebos Waterway
sediments, and potentially associated effects on English sole, including impaired growth,
mortality, reduced disease resistance, and reproductive dysfunction.
These studies found that juvenile salmon have been exposed to a wide variety of contaminants,
including PCBs, as compared to fish from hatcheries or reference areas. The levels of PCBs in
juvenile salmonids from the Hylebos Waterway were found to be similar to levels shown in
previous studies to be associated with injuries (e.g., reduced growth, increased mortality).
English sole within Hylebos Waterway were also found to be exposed to PCBs and other
compounds and exhibited both liver lesions and reproductive dysfunction. There has been little
change over the last 10 years in indicators of fish health, such as liver lesions. NOAA's studies
showed precocious sexual maturation in adult females and inhibited gonadal development in
adult males. Increased incidence of early onset of sexual maturation appears to be most closely
associated with exposure to PCBs. The NOAA studies have not identified an effects threshold
concentration in either fish tissue or sediment which could be used to set a protective PCB
cleanup level.
-------
Figure 1
Commencement Bay
Areas Considered in Human Health and
Ecological Risk Evaluation
Waterways:
1 H'
2 Bfalr Waterway
3 Sitcum Waterway
4 Milwaukee Waterway
5 St. Paul Waterway
6 Middle Waterway
7 Thea Foss/Wheeler-Osgood
8 Nearshore Area
A EPA
-------
EPA calculated a no adverse effects level for salmon based on information showing that in the
Duwamish River, hatchery salmon had statistically significantly lower growth, higher mortality,
and lower body burdens of PCBs (about 137 /zg/kg) than fish exposed to sediments and prey
within the river. A protective sediment concentration can be calculated for Commencement Bay
to be about 83 ^zg/kg PCBs as a sediment concentration that would not result in adverse effects to
juvenile salmonids.
Because fish and birds obtain their prey from more than one location, it is appropriate to compare
effects levels to average sediment concentrations that will be achieved as a result of cleanup,
rather than the cleanup levels themselves. By dividing the estimated effects level by the post-
cleanup average PCB concentration (as shown on Table A-l), a hazard quotient (HQ) can be
calculated. An HQ which is less than 1 indicates a low potential for adverse effects. The
estimated HQ for Commencement Bay and the individual waterways is shown in Table 1 . As
shown in Table 1, at the 300 /zg/kg PCB SQO, the HQs for juvenile salmonids at the CB/NT
Site, and the Hylebos and Thea Foss Waterways are all less than 1 . At the 450 jzg/kg PCB
SRAL, the estimated HQ for juvenile salmonids are less than or equal to 1 for the CB/NT Site
and the Hylebos and Thea Foss Waterways.
Table 1 . Hazard Quotients for juvenile salmonids based on the 150
and the revised PCB SQO of 300 /zg/kg and SRAL of 450 /zg/kg.
SQO in the 1989 ROD
Organism
Juvenile
Salmonids
Cleanup
level •
(M9/kg)
150
300
450
Hazard Quotient
CB/NT Site
<1
< 1
<1
Hylebos Waterway
<1
< 1
1
Thea Foss Waterway
< 1
< 1
< 1
Note: HQ for the CB/NT Site assumes fish obtain 100% of their diet from the CB/NT Site. HQ for the
individual waterways assume fish obtain 100% of their diet from that waterway.
Bjids
For its analysis of potential effects of PCB-contaminated sediments on birds, EPA drew upon a
FWS assessment of the potential for injury based on calculation of a hazard quotient or ratio
between the predicted egg concentration and an egg concentration reported in the literature to be
associated with significant ecological impacts, such as embryonic deformity and egg lethality.
The amount of sediment transfer from sediments to fish, from fish to birds and subsequently to
bird eggs, was estimated using information from EPA and FWS studies on transfer of PCBs and
dioxins between predators and prey.
Hazard quotients were calculated for shorebirds and piscivorous (fish eating) birds by dividing
the estimated sediment concentrations above which adverse effects to bird eggs are predicted to
8
-------
occur under the assumptions in the risk assessment by the post-cleanup average sediment PCB
concentration. These results are shown in Table 2. Table 2 shows HQs of less than or equal to 1
for shorebirds and piscivorous birds at the 300 /^g/kg PCB SQO. At the 450 /zg/kg PCB SRAL,
HQs for shorebirds and piscivorous birds are at 2 or below.
Table 2. Hazard Quotients for shorebirds and piscivorous birds, based on the 150 Mg/kg SQO in
the 1989 ROD and the revised PCB SQO of 300 ^g/kg and SRAL of 450
Organism
Shorebirds
Piscivorous
Birds
Cleanup
level
(M9/kg)
150
300
450
150
300
450
CB/NT Site
< 1
<1
1
1
1
2
Hazard Quotient
Hytebos Waterway
<1
<1
1
<1
1
2
Thea Foss Waterway
< 1
<1
<1
< 1
< 1
1
Note: HQ for the CB/NT Site assumes birds obtain 100% of their diet from the CB/NT Site. HQ for the
individual waterways assume birds obtain 100% of their diet from that waterway.
Uncertainties
The estimates of potential risk to invertebrates, fish, and birds incorporate a number of
assumptions, all of which have uncertainties associated with them. Examples of uncertainties
and their effect on the resulting risk estimates are discussed below.
Although the AET database used to estimate risks to invertebrates was developed using
Commencement Bay data, along with data from other areas in Puget Sound, recent biological
data collected for the Hylebos Waterway indicate that the AET database may have overestimated
the chemica' concentration at which impacts were expected to occur.
For birds, the biomagnification factor (an estimate of contaminant transfer between predators and
prey) was developed based on empirical data on alewives and herring gull in the Great Lakes.
The accuracy of this estimate when applied to other species, especially species higher on the food
chain, is uncertain.
The calculation of a protective sediment concentration for juvenile salmonids is uncertain
because of the extrapolation of Duwamish estuary data to Commencement Bay and application
of a biota-sediment accumulation factor (an estimate of the transfer of contaminants from
sediments to organisms) developed for bottom fish to a water column species. Use of these data
-------
and associated assumptions may either over- or underestimate risks to juvenile salmomds, and
should not be extrapolated to other types of fish.
The estimated HQs for fish and birds assume that they obtain all of their food from within the
CB/NT Site. The actual foraging habits and foraging range offish and birds varies by species
and in many cases, by season. The assumption used in the ecological risk evaluation will
overestimate the exposure of species or individuals with large foraging ranges (such as migratory
birds) but the CB/NT Site-wide risk estimate may underestimate the exposure of resident species
that preferentially feed at a specific location.
Sumrrarv f Ecoloical Ris
In summary, the updated ecological risk analysis showed that the 300 izg/kg PCB SQO and 450
Aig/kg PCB SRAL is protective of the benthic community, juvenile salmonids, shorebirds and
piscivorous birds. Cleanup to the 300 Mg/kg PCB SQO will reduce all HQs estimated for these
species to 1 or below.
Human Health Risk Evaluation
Exposure Assumptions
EPA updated the human health risk evaluation and used it as a basis to evaluate the risks
associated with a variety of potential PCB cleanup levels in a February 1997 report (Weston,
1997a). Although EPA's risk assessment methodology has not been modified substantially since
the original risk assessment was performed in 1988, some of the exposure and toxicity
assumptions have been changed based on new information and new Superfund guidance.
As with the 1989 ROD, the updated risk evaluation focused on risks due to consumption of PCB-
contaminated seafood. The National Contingency Plan (40 CFR Part 300) calls for EPA to use a
reasonable maximum exposure (or "high-end") scenario for making Superfund cleanup
decisions. EPA also recommends calculating an average exposure scenario for comparison
purposes. Four scenarios were used in the updated risk evaluation: average recreational fishing,
"high-end" recreational fishing, average tribal fishing, and "high-end" tribal fishing.
Because the Puyallup Tribe of Indians has treaty rights to fish in Commencement Bay, high-end
tribal fishing was used as the reasonable maximum exposure scenario for EPA's decision-making
purposes. An average and high-end recreational fishing scenario and an average tribal fishing
scenario were also calculated for purposes of comparison. Fish consumption rates for the
recreational fishing scenario are the same as used in the 1989 ROD. Because no studies have
documented tribal fish consumption rates in Commencement Bay, they were estimated from
recently completed surveys offish consumption by members of two other Puget Sound Indian
tribes, the Tulalip and Squaxin Island Tribes (Toy et al., 1996). The high-end tribal scenario
10
-------
represents risks to a tribal fisherperson who consumes a relatively large amount (upper 90th
percentile) of fish compared to other tribal members.
A summary of the fish ingestion rates assumed for each of the four scenarios used in the updated
residual risk evaluation and the scenario used in the ROD risk evaluation is presented in Table 3.
Information about all of the assumptions used in the risk assessment, and modifications made to
update the original risk assessment, is provided in Weston (1997a) and Appendix A.
Table 3: Fish ingestion rates used in the updated residual risk evaluation, and in the CB/NT
ROD.
Scenario
High-end Tribal
Fishing
Average Tribal Fishing
High-end Recreational
Fishing
Average Recreational
Fishing
CB/NT ROD
Fish Ingestion Rate (1)
123 g/day
(approx. 20 meals/month)
41. 7 g/day
(approx. 7 meals/month)
95.1 g/day
(approx. 15 meals/month)
12. 3 g/day
(approx. 2 meals/month)
12.3 g/day
(approx. 2 meals/month)
Basis
Tulalip/Squaxin Island Fish Consumption
Survey (Toy et al., 1996)
Tulalip/Squaxin Island Fish Consumption
Survey (Toy et al., 1996)
CB/NT ROD (1989)
CB/NT ROD (1989)
A combination four recreational fishing
surveys, as summarized in Tetra Tech,
1988
(1) One half pound of fish per meal was used to calculate meals/month.
Cancer Risks
Post-cleanup residual cancer risks were calculated for a range of potential PCB sediment cleanup
levels from 50 A*g/kg to 900 Mg/kg. Results for the high-end tribal fishing scenario at the former
PCB SQO of 150 Atg/kg and the current 300 ^g/kg PCB SQO and 450 A*g/kg SRAL are shown in
Table 4. Cleanup to all of these cleanup levels will result in substantial risk reduction over
current conditions. The residual human health cancer risks for the high-end recreational fishing
scenario were similar to results for the high-end tribal fishing scenario, and results for the
average recreational and average tribal fishing scenarios are three to ten times lower than for
either of the high-end scenarios. For comparison, human health cancer risks from fish
consumption at Puget Sound background conditions in non-industrial areas under the high-end
tribal fishing scenario is 6 x 10'5 (Weston 1997c). Additional information about the results of
the risk evaluation is presented in Appendix A.
11
-------
Table 4. Post-cleanup residual cancer risk (using the high-end tribal fishing scenario) based on
the 150 Mg/kg PCB SQOjn the 1989 ROD and the revised PCB SQO of 300 Mg/kg and SRAL of
450
PCB Cleanup level
fag/Kg)
Current Conditions
(no cleanup)
150
300
450
Estimated post-cleanup residual cancer risk
under the high-end tribal fishing scenario
CB/NT Site
9.8 x 10"
9.4x10-*
1.2x10"
1.4x10"
Hylebos Waterway
1.1 x10°
4.9 x 10-*
1.1 x10"
1.6x10-*
Thea Foss Waterway
1.7x10"
4.6 x 10s
7.6 x 10-*
1.0x10"
As shown in Table 4, the estimated post-cleanup cancer risks at the 300 f/g/kg PCB SQO are
within EPA's acceptable risk range of 10"4 to lO"6. Cleanup to the 450 Atg/kg PCB SRAL will
result in interim risks that are also within EPA's acceptable risk range. Although the estimated
risk is 1.4 x 1V for the CB/NT Site and 1.6 x 10" for the Hylebos Waterway, EPA policy states
that the upper boundary of the risk range is not a discrete line at 1x10". Cleanups to levels
slightly greater than 1x10"* may be considered acceptable if justified based on site-specific
conditions. People are more likely to fish in more than one location in Commencement Bay than
in Hylebos Waterway alone, so the CB/NT Site-wide risk estimate is the best estimate of risks to
area fisherpersons.
These cancer risk estimates are based on a number of assumptions, all of which have
uncertainties associated with them. For example, although studies suggest that exposure to PCBs
increases cancer risk in humans, EPA's cancer slope factor (an estimate of the potency of PCBs
to cause cancer) is derived mainly from studies on laboratory animals. Similarly, estimates of
the transfer of PCBs from sediments to fish and from fish to humans, the amount and type of fish
consumed by individuals in the area, and many other factors make it difficult to estimate an
individual's exposure to PCBs from eating fish from Commencement Bay. Although the risk
numbers in Table 4 and Appendix A are estimates that have many uncertainties associated with
them, they were derived in a manner to ensure they are protective of health and are unlikely to
underestimate actual risks to fish consumers.
Non-cancer risks
In the human health risk assessment done for the CB/NT Site in 1985, the potential for both
cancer and non-cancer health effects from exposures to PCBs in fish from the CB/NT Site was
evaluated. Based on the information available on the toxicity of PCBs at that time, it was
concluded that the potential for non-cancer impacts was not of concern. Therefore, when the
12
-------
ROD was written in 1989, EPA based its cleanup level for PCBs on human health cancer risks
from ingestion of PCB-eontaminated fish caught in the Bay.
Similarly, in the updated risk evaluations (Weston 1996, 1997a) and in EPA's public review
draft ESD (EPA, 1997), EPA focused on cancer risks associated with PCBs, and updated non-
cancer risk information was not presented in these reports. In response to public comments,
EPA has prepared a technical memorandum which provides information regarding non-cancer
risks under current and post-cleanup conditions using the scenarios developed for the cancer
risk evaluation. This technical memorandum (Weston, 1997b) has been added to the
Administrative Record for our final decision. A summary is provided below.
Based upon an understanding of the development of non-cancer health effects, potential non-
cancer impacts are evaluated by EPA assuming that there is a level of exposure below which
health impacts are unlikely to occur. The estimate of this level of exposure is called the
reference dose, or RfD, and is defined as "an estimate (with uncertainty spanning perhaps an
order of magnitude) of a daily exposure to the human population (including sensitive
subgroups) that is likely to be without an appreciable risk of deleterious effects during a
lifetime." Exposures that are less than the RfD are not likely to be associated with adverse
health impacts.
In order to calculate non-cancer risks for site-specific risk assessments, EPA compares the
RfD to the exposures estimated for that site (e.g., from eating contaminated fish). This
comparison, called the Hazard Quotient or HQ, is the ratio between the estimated site
exposure and the RfD. As with cancer risk, the assumptions used in calculating the HQ are
conservative ones (health protective) to ensure that remedial decisions based upon them will
protect more sensitive individuals. Because of the way in which the HQ is derived, it should
not, however, be viewed as a strict demarcation between toxic and nontoxic. HQ values above
1 do not mean that non-cancer health impacts will occur, but rather that the potential for such
impacts increases as 1 is exceeded. The potential for impacts depends on a number of factors,
including the protectiveness of both the RfD and the exposure assumptions used to calculate
the HQ. The derivation of the RfD for PCBs is based upon a large body of experimental data
and incorporates a several hundred fold uncertainty factor ("safety" factor) to ensure
protection.
Results of this non-cancer risk evaluation are shown in Tables 5 and 6. Both the cleanup
required under the 1989 ROD and in this ESD provide for substantial reduction in the non-
cancer risks associated with PCBs in sediments at the CB/NT Site. As with cancer risks, the
residual human health non-cancer risks for the high-end recreational fishing scenario were
similar to results for the high-end tribal fishing scenario, and results for the average recreational
and average tribal fishing scenarios are three to ten times lower than for either of the high-end
scenarios. For comparison, the human health hazard quotient for non-cancer risks due to PCBs
from fish consumption at Puget Sound background conditions in non-industrial areas under the
"high-end" tribal fishing scenario is 3 (Weston, 1997c). As with cancer risks, given the range
13
-------
of uncertainty in risk calculations, these post-cleanup HQs are not significantly different. Also,
the exposure assumptions used in the CB/NT risk evaluation were selected to be protective for
a consumer of large amounts of fish from the CB/NT Site over a 30-year period. Given these
conservative assumptions, the small increases above an HQ of 1 estimated for the various
target cleanup levels and for background suggest a low potential for non-cancer impacts for the
fish consumers considered in the calculations. Individuals who eat less fish from the CB/NT
Site will have exposures and HQs that are lower and, therefore, their potential for non-cancer
impacts will be less. Therefore, EPA does not believe that the 300 Mg/kg PCB SQO provides
significantly different non-cancer risks than the PCB cleanup level in the 1989 ROD.
Table 5. Post-cleanup residual non-cancer risk (using the high-end tribal fishing scenario) based
on the 150 Atg/kg PCB SQO in the 1989 ROD and the revised PCB SQO of 300 jzg/kg and
SRAL of 450
PCB Cleanup level
G*g/»
-------
SRAL to 450 Mg/kg, to be achieved during cleanup. EPA used the nine remedy selection criteria
contained in the NCP and listed below to evaluate and select the revised PCB cleanup level.
Based on our analysis, EPA has determined that a PCB SQO of 300 Mg/kg and SRAL of 450
/zg/kg achieves the best balance of the nine evaluation criteria. Table 7 compares the PCB
cleanup level in the 1989 ROD to the new cleanup levels.
Table 7. Comparison of post-cleanup maximum and average PCB concentration under the PCB
cleanup level in 1989 ROD to the PCB cleanup level in this ESD.
CB/NT Site
Hytebos Waterway
Thea Foss Waterway
1989 ROD
Immediately after
cleanup (^g/kg)
300 max/63 avg
300 max/80 avg
300 max/81 avg
Within 10 years
after cleanup
Ug/kg)
-------
contamination.
In the public review draft ESD, EPA's proposal was to require that PCB-contaminated sediments
be cleaned up to 450 ^g/kg PCBs, with no further natural recovery requirements. EPA's final
decision is to add a natural recovery component to achieve 300 Mg/kg PCBs in sediments within
10 years after remedial action. EPA believes this change is appropriate for two reasons:
1) EPA's original proposal relied on estimates of additional natural recovery to ensure
protectiveness. Setting a long-term SQO, however, ensures that natural recovery modeling and
monitoring will occur to verify these reductions in PCB concentrations over time, and that
additional cleanup work will be implemented to achieve 300 Mg/kg PCBs if sediments do not
naturally recover to this level within 10 years.
2) Although dredging to 300 Mg/kg PCBs would be unreasonably costly for the relatively small
risk reduction achieved, costs associated with natural recovery to 300 Mg/kg are relatively small
and do not outweigh the benefits of additional risk reduction.
Natural recovery models done for the 1989 ROD indicate that PCB concentrations should be
reduced to less than 300 ^tg/kg within 10 years after cleanup to 450 A*g/kg. The cost of long-term
monitoring in the natural recovery areas is estimated at $310,000. The cost of additional
modeling is negligible given that modeling work is already being done to predict natural
recovery rates of other chemicals. Even if the models have not accurately predicted the rate of
natural recovery, and additional cleanup is needed ten years after cleanup, it is anticipated that
cost-effective methods can be used to remediate sediments with contaminant concentrations
between 300 and 450 /zg/kg PCBs.
EPA's Nine Evaluation Criteria
Threshold Criteria
Overall protection of human health and the environment
Compliance with applicable, or relevant and appropriate, regulations (ARARs)
Primary Balancing Criteria
Long-term effectiveness and permanence
Reduction of toxicity, mobility and volume through treatment
Short-term effectiveness
I mplementability
Cost
Modifying Criteria
State and Tribal acceptance
Community acceptance
The 1989 ROD evaluated cleanup alternatives against the nine CERCLA evaluation criteria. In
this section, these criteria ate discussed only to the extent that they would be affected by the
16
-------
change in the PCB cleanup level. Because a modification to the PCB cleanup level will
significantly affect only the Hylebos Waterway cleanup, the analyses of cost and
implementability focus mainly on the Hylebos Waterway. In this analysis the 450 ^g/kg PCB
SRAL is compared to the 1989 ROD SRAL of approximately 300 Mg/kg, and the 300 Mg/kg PCB
SQO is compared to the 150 ^g/kg PCB SQO in the 1989 ROD. Other potential cleanup levels
are also discussed.
Overall Protection of Human Health and the Environment
The NCP states that for chemicals which are known to cause cancer, cleanup levels should be
selected within a cancer risk range of 10"4 to 10"6. Within that range, the NCP calls for EPA to
select a cleanup goal which achieves the best balance of the remaining remedy selection criteria,
with a preference for selecting cleanup goals that approach 10~* where possible.
As shown in Table A-4, the human health risk evaluation for the CB/NT Site showed that all of
the potential cleanup levels evaluated, from 50 Mg/kg PCBs to 900 Mg/kg PCBs, would result in
post-cleanup residual cancer risks within EPA's risk range of 10"4 to 10"6. Lower PCB cleanup
levels provide more protectiveness, and calculated residual risks are closer to EPA's "point of
departure" of 10"6. Higher cleanup levels are at the less protective end of the range. As shown in
Table A-4, there is very little difference in residual human health cancer risks between the ROD
PCB SQO of 150 Mg/kg and the revised SQO of 300 jig/kg set by this ESD. Both will result in
human health cancer risks for the CB/NT Site of approximately 1x10**. As shown in Table 7,
average sediment PCB concentrations after cleanup to the 450 Mg/kg SRAL will be reduced to
well below 150 /zg/kg and will be further reduced over time to approximately 63 Mg/kg within 10
years after cleanup.
The NCP does not set a numeric target range for non-cancer risks, but states that acceptable
exposure levels shall represent concentrations to which the human population, including
sensitive subgroups, may be exposed without adverse effect during a lifetime or part of a
lifetime, incorporating an adequate margin of safety. Cleanup to 300 Mg/kg PCBs will result in a
CB/NT Site-wide HQ of 7. The HQ of 7 is not appreciably different than the HQ of 6 estimated
for cleanup to 150 /zg/kg PCBs under the 1989 ROD. Although EPA generally attempts to
achieve a HQ of 1 or below in Superfund cleanups, in this case the HQ is greater than 1 even
under background conditions. Because of the conservative assumptions built into the risk
assessment and into the reference dose used to calculate the HQ, EPA believes that cleanup to
300 //g/kg PCBs is protective of non-cancer risks.
The ecological evaluation also shows that lower cleanup levels are associated with increased
protection to wildlife. EPA's analysis estimates that a PCB SQO of 300 /ig/kg and SRAL of 450
Mg/kg will be protective of ecological receptors. Both values are below the 1,000 Mg/kg PCB
AET calculated in the CB/NT ROD as being protective of benthic infauna. Both values also fall
between Ecology's minor adverse effects level (equivalent to 1,000 Mg/kg dry weight) and no
adverse effects level (equivalent to 130 Mg/kg dry weight) as promulgated under the State's
17
-------
Sediment Management Standards to protect aquatic life. The updated ecological risk analysis
for fish, shorebirds, and_piscivorous birds estimates a HQ of less than or equal to 2 immediately
after cleanup and less than or equal to 1 within 10 years after cleanup.
The risk evaluation shows that cleanup to EPA's revised SQO of 300 Mg/kg PCBs and SRAL of
450 Mg/kg PCBs will result in a substantial reduction in risk to human health and the
environment due to PCBs at the CB/NT Site. They will reduce cancer risks to within EPA's
acceptable risk range, and will be protective of non-cancer risks and wildlife.
Compliance with ARARs
There are no federal or state regulations that provide human-health based PCB cleanup levels for
sediments. Since publication of the ROD, the State of Washington has promulgated Sediment
Management Standards (SMS), which require that contaminant levels in sediments within the
State be protective of human health and aquatic life. The SMS contains numeric chemical
criteria for protection of aquatic life (Chapter 173-204 WAC). The SMS also requires that
sediment cleanups be protective of human health, and states that Ecology will determine
sediment criteria which are protective of human health on a case by case basis, but contains no
numeric standards for human health protection.
The CB/NT ROD and the Washington State SMS share the same narrative goal of the absence of
acute or chronic adverse effects on biological resources or significant human health risk. They
also share the use of the AET process (as described in the 1989 ROD, with some modifications
introduced into the State SMS) to select chemical sediment criteria for protection of aquatic life.
The NCP provides that a promulgated state environmental requirement is an ARAR if it is more
stringent than cleanup levels developed by EPA as a result of the human health and ecological
risk assessments. The NCP also states that state environmental standards promulgated after the
signing of a ROD must be attained only if they are applicable, or relevant and appropriate, and
necessary to ensure protectiveness. EPA evaluated the State SMS requirements for ecological
and human health-based cleanup levels and determined that it was not more stringent than the
evaluation process EPA used to select a PCB cleanup level, nor was application of the State SMS
requirements necessary to ensure protectiveness.
Long-term Effectiveness and Permanence
Under this criterion, EPA evaluates the magnitude of residual risk after cleanup, and the
adequacy and reliability of long-term controls. As discussed under the "protection of human
health and the environment" section, cleanup to the 450 A*g/kg SRAL and natural recovery to the
300 Mg/kg SQO, rather than the 1989 ROD SRAL of 300 /zg/kg and SQO of 150 Mg/kg, will
result in a slightly higher residual risk after cleanup. CB/NT Site-wide estimated residual cancer
risks due to PCBs for the 1989 SQO of 150 Mg/kg PCBs are 9.4 x 10'5 versus 1.2 x 10"1 for the
300 Afg/kg SQO under this ESD. For Hylebos Waterway, estimated residual cancer risks for the
\S
-------
former 150 /ig/kg PCB SQO are 4.9 x 10'5 versus 1.1 x lO"4 for the 300 ^g/kg SQO under this
ESD. Similarly, the non-cancer HQ for the CB/NT Site is 6 under the 1989 ROD SQO and 7
under the revised SQO.
Average post-cleanup PCB concentrations will be reduced to less than 150 ^g/kg immediately
after cleanup, and will be reduced further over time through natural recovery processes. As
shown in Table 7, PCB concentrations will be reduced to a maximum of 300 /zg/kg and an
average of 63 //g/kg PCBs in 10 years.
For cleanup to the PCB cleanup levels in the 1989 ROD and this ESD, tested and reliable
technologies exist for containing sediments and preventing contaminant migration in the long
term. As discussed below, a greatly increased volume of PCB-contaminated sediments would
require containment under the current ROD, as compared to the PCB cleanup required under this
ESD. By reducing the volume of sediments that require confinement, a smaller disposal site (or
fewer disposal sites) will be required. This will reduce the cost and complexity of long-term
monitoring and engineering controls to ensure the protectiveness of confined disposal.
Reduction of toxicity. mobility and volume through treatment
This criterion is unaffected by the range of alternatives evaluated for a modified PCB sediment
cleanup level because the cleanup plan for contaminated sediments under the 1989 ROD does not
involve treatment.
Short-term effectiveness
The analysis of short-term effectiveness remains the same as presented in the original ROD,
except to note that changing the SRAL from 300 to 450 ^g/kg PCBs decreases the volume of
sediments requiring remediation by approximately 400,000 cubic yards. This in turn decreases
the short term disruption to aquatic organisms living in the sediments during remediation.
Approximately twice as much of the surface area of the Hylebos Waterway would have been
disrupted through dredging or capping at the 300 Mg/kg PCB SRAL than at 450 Mg/kg. For all
cleanup volumes, a monitoring program will be implemented and measures will be taken to
control releases of contaminated sediments to the water column.
Implementabiljty
Cleanup of contaminated sediments under the PCB cleanup levels in the 1989 ROD and this ESD
is technically feasible. However, due to the limited number and capacity of disposal sites in
Commencement Bay, the larger the volume of sediments to be remediated, the greater the
difficulty in selecting a suitable disposal site or sites. Also, at higher cleanup volumes, multiple
disposal sites would be needed for the Hylebos Waterway cleanup, in addition to sites required
for other waterway cleanups, which will increase the cost, as well as technical and administrative
difficulties, associated with implementing the remedy. Because remediation to the 300 /ig/kg
19
-------
PCB SRAL under the 1989 ROD requires remediation of appproximately 400,000 cubic yards
more sediments than remediation to the 450 Mg/kg SRAL required under this ESD, EPA's
current proposal is more technically and administratively feasible than cleanup to the former
SRAL.
CflSl
Because the volume of sediments requiring remediation in CB/NT problem areas other than the
Hylebos Waterway is largely unaffected by the PCB cleanup level, the cost analysis presented
here is focused on the Hylebos Waterway. The ROD estimated that 1,167,000 cy of Hylebos
Waterway sediments are contaminated at concentrations exceeding the ROD SQOs. The ROD
estimated that a cleanup of 447,000 cy of contaminated sediments would allow the Waterway to
recover to meet SQOs for all chemical contaminants in 10 years. The ROD estimated a cost of
$13,850,000 for cleanup of Hylebos Waterway sediments using confined aquatic disposal, and
that costs would be higher for upland disposal and lower for other disposal options.
Table 8 shows the volume of sediments requiring remediation and estimated cleanup costs for the
Hylebos Waterway, based on current information. As shown in Table 8, the estimated cost to
achieve the 150 ^g/kg PCB cleanup level in 10 years (assuming an SRAL of 300 Mg/kg) has
more than doubled since the ROD estimated $13 million.
Table 8. Estimated sediment cleanup volumes for the Hylebos Waterway and associated cost of
remediation at a range of PCB cleanup levels.
PCB Sediment
Remedial Action Level
(M9/kg)
150
300
450
600
750
900
Cleanup Volume (PC8-
contaminated
sediments only) (cy)(1)
1.115,700
559,721
246.565
149,444
107,206
64,910
Cleanup Volume
(cleanup of all
contaminated
sediments) (cy) (1)
1.339,000
891.000
508,000
436.000
409.000
399.000
Estimated Cleanup Cost
(for all contaminated
sediments) (millions) (2)
$46.5 (3)
$31(3)
$18
$15.5
$14.5
$14
(1) Estimated volume of PCB-contaminated sediments was calculated in Fugtevand (1996). Estimated
volume of all contaminated sediments was estimated using Fuglevand (1996) and an estimate of 350,000
cy for cleanup of all contaminants other than PCBs. based on current unpublished estimates of the volume
of sediment requiring cleanup to ecologically-based cleanup levels of approximately 300,000 to 400,000
cy.
(2) Estimated cost of cleanup was derived using a $35/cy estimate for nearshore fill from Hartman (1996).
The estimated cost for confined aquatic disposal is approximately the same as for nearshore fill; upland
20
-------
disposal would increase the cleanup cost by approximately 46%.
(3) For cleanup volumes of_greatef than 700,000 cubic yards, multiple disposal sites may be needed, so
cleanup costs for these options may be underestimated.
Table 8 and Figure A-l shows the volume of sediments requiring cleanup at the range of PCB
cleanup levels considered. The cost of cleanup is directly related to the volume of contaminated
sediments requiring cleanup, and the post-cleanup average PCB concentration gives an idea of
the environmental benefit gained at each incremental increase in cleanup level. Table 9 shows
the cost of cleanup as compared to the average PCB sediment concentration achieved after
cleanup in Hylebos Waterway and in the CB/NT Site as a whole. Since fish and fisherpersons
are likely to move throughout a waterway or throughout the CB/NT Site, the post-cleanup
average gives an idea of the amount of PCB contamination these receptors would be exposed to.
At PCB cleanup concentrations between 900 and 450 A*g/kg, the environmental benefit in terms
of reduction in the post-cleanup average PCB concentration is approximately equal to or greater
than the percent increase in volume requiring remediation. Because of the large volume of
sediments with PCB concentrations of 450 iig/kg or less, the volume of sediments requiring
remediation becomes very high in comparison to the percent reduction in average PCB
concentrations.
Table 9. Post-cleanup average PCB concentrations and cleanup costs at a range of PCB sediment
cleanup levels.
PCB Sediment
Remedial Action Level
(M9/kg)
150
300
450
600
750
900
Post-cleanup average
concentration (Hylebos
Waterway only) Gug/kg)
37
80
124
143
156
161
Post-cleanup average
PCB concentration
(CB/NT Site) (^g/kg)
51
63
75
82
86
89
Cleanup cost (Hylebos
Waterway only)
(millions) (1)
$46.5
$31
$18
$15.5
$14.5
$14
(1) Costs presented are Hylebos Waterway cleanup costs only. Bay-wide cleanup costs can be estimated
at Hylebos Waterway cost + $18 million (ROD estimated cost for cleanup of all other problem areas).
Based on this analysis, EPA has determined that a PCB SRAL of 450 Mg/kg is cost-effective. A
remedial alternative is considered "cost effective" if its costs are proportional to its overall
effectiveness. Overall effectiveness is determined by evaluating its short-term and long-term
effectiveness (and reduction in toxicity, mobility, and volume, in decisions where treatment is
being considered), as discussed above. Cleanup to 450 Mg/kg PCBs would significantly reduce
the overall post-cleanup average PCB concentration, and subsequently, the risks to human health
and the environment, in the individual waterways and in Commencement Bay as a whole.
21
-------
Cleanup to the ROD SRAL of 300 Mg/kg PCBs would be significantly more costly (the cost of
cleanup to 300 Mg/kg is about $13 million higher than the cost of cleanup to 450 ,ug/kg), while
the difference in long-term and short-term protectiveness is not large.
Although the majority of the cost of the cleanup is associated with meeting the SRAL, there are
some costs associated with assuring the SQO is met within 10 years after remedial action through
natural recovery. These costs are expected to result in a very small increase to the cost of the
remedy, as discussed below.
• Natural recovery predictions in the 1989 ROD will be updated with new natural recovery
models using pre-design sampling data. This cost of modeling for natural recovery of
PCBs will be a negligible addition to the efforts already underway at most CB/NT
problem areas to predict natural recovery rates for other chemicals.
• Some additional remedial action may be required in areas where natural recovery models
predict that a lower PCB SRAL is needed to achieve PCB concentrations of 300 A*g/kg in
10 years after the remedial action. This may require additional remedial action at a few
stations, with little impact on the overall cost of the remedy.
• Implementation of this change will require natural recovery monitoring in areas where
post-cleanup PCB concentrations are between 450 A*g/kg and 300 Mg/kg. EPA estimates
the cost of monitoring during the 10-year natural recovery period at $310,000 (Weston,
1997d).
• If models incorrectly predict natural recovery rates, additional remedial action will be
needed in areas which do not naturally recover to 300 ^g/kg PCBs. However, EPA
believes that low cost remedial action alternatives can be used to remediate the remaining
areas where PCB concentrations fall between 450 Mg/kg and 300 jig/kg 10 years after
remedial action. Potential remedial action alternatives include but are not limited to
enhanced natural recovery, limited capping, or limited dredging, if a disposal site is
readily available. EPA's best estimate is that these actions will not be needed, so
contingent costs for these actions are not included in EPA's cost estimate. If additional
remedial actions do prove necessary, they are expected to cost much less than the current
estimated marginal cost of $13 million for full-scale cleanup of areas where sediment
PCB concentrations currently fall between 450 and 300 Mg/kg.
State and Tribal Acceptance,
The State of Washington Department of Ecology concurs with the selected PCB cleanup level.
Ecology's concurrence letter is included as Appendix B. The Puyallup Tribe of Indians does not
concur with the selected PCB cleanup level.
22
-------
Community Acceptance
Based on the public comments received during EPA's initial community relations activities
associated with the PCB cleanup level and during the public comment period held from March
10 through April 9, 1997, it is clear that the community is divided on this issue. Some members
of the community believe the cleanup level should be raised, while many others believe it should
remain as stated in the 1989 ROD, or that the PCB standard should be even more stringent than
stated in the ROD. Appendix C contains a summary of the comments received during the public
comment period and EPA responses to those comments.
Summaiv of the Comparative Analysis of Alternatives
Based on the human health and environmental risk evaluation, and the comparative analysis of
alternatives, EPA is modifying the PCB sediment SQO in the CB/NT ROD to 300 /ug/kg, to
replace the former SQO of 150 /zg/kg. The 300 /zg/kg PCB SQO must be achieved in all CB/NT
problem areas within 10 years after cleanup. Because of the long-term human health risks
associated with PCBs, EPA is also setting a maximum PCB SRAL of 450 ^g/kg for the CB/NT
Site. The SRAL must be achieved in all CB/NT problem areas during cleanup. EPA's
calculations show that an average post-cleanup PCB concentration of less than 150 Mg/kg will be
achieved in all waterways and in Commencement Bay as a whole after cleanup to the SRAL of
450 A*g/kg, and will be reduced to an average of 63 ^g/kg within 10 years through natural
recovery processes. EPA has selected 300 Mg/kg as the appropriate PCB SQO and 450 Mg/kg as
the maximum PCB SRAL for the CB/NT Site for the following reasons:
• It is within the EPA's acceptable risk range for Superfund cleanups and is protective of
human health cancer risks. EPA's human health risk calculations show that a PCB SQO
of 300 Mg/kg will result in post-cleanup residual risks for persons consuming fish from
the CB/NT Site of 1 x lO^for the Hylebos Waterway and the CB/NT Site as a whole, and
8 x 10'5 for the Thea Foss Waterway. Residual risks during the 10-year natural recovery
period will be only slightly higher, at 1 x 10"4 for the CB/NT Site and Thea Foss
Waterway, and 2 x 10"4 for the Hylebos Waterway.
• It meets the NCP standard for non-cancer risks of providing post-cleanup concentration
levels to which the human population, including sensitive subgroups, may be exposed
without adverse effect during a lifetime or part of a lifetime, incorporating an adequate
margin of safety. Although the non-cancer HQ for the high-end tribal fishing scenario is
greater than one, EPA believes that there is a sufficient margin of safety built into the
estimates of toxicity and exposure to provide for protection of human health.
The PCB SQO of 300 /zg/kg and SRAL of 450 ^g/kg are protective of ecological
receptors. They are below the 1,000 /zg/kg PCB AET calculated in the CB/NT ROD as
being protective of benthic infauna. They also falls between Ecology's no adverse effects
level and minor adverse effects level as promulgated under the State's SMS to protect
23
-------
aquatic life. The updated ecological risk analysis indicates that the SQO and SRAL are
protective of juvenile salmonids, shorebirds, and piscivorous birds.
• A 450 A*g/kg SRAL is the most cost-effective alternative. Remediation to a lower
cleanup level requires remediation of a substantially larger volume of sediments.
Cleanup to the former 300 ^g/kg PCB SRAL would result in a 70% increase in the
volume of sediments to be remediated in the Hylebos Waterway, and a 150 Mg/kg PCB
cleanup level would result in a 150% volume increase, when compared to the 450 Mg/kg
cleanup level. This increased volume would also result in an increase in the cost of
cleanup, the area needed for a disposal site, and the area of the Waterway where aquatic
organisms are disrupted during dredging.
• The 300 Mg/kg PCB SQO provides for additional protection at a small additional cost.
By requiring that modeling be done to confirm ROD predictions that PCB concentrations
will be reduced by natural recovery processes within 10 years after cleanup, and requiring
long-term monitoring and potential additional cleanup to ensure the SQO is achieved, the
300 A*g/kg PCB SQO provides additional assurance of the long-term protectiveness of the
remedy. Over the 10-year natural recovery period, human health cancer risks would be
reduced by an additional 14% for the CB/NT Site as a whole, 34% for the Hylebos
Waterway, and 31% for the Thea Foss Waterway. There will be similar reductions in
non-cancer risks and ecological risks. EPA believes that establishing a maximum SRAL
and a 300 Atg/kg SQO is cost-effective, especially since the natural recovery component
of the remedy increases the cost of the remedy by a very small amount Even if natural
recovery models have not accurately predicted the rate of natural recovery and additional
cleanup is needed to achieve 300 Atg/kg PCBs within 10 years after cleanup, low cost
solutions such as enhanced natural recovery can be used to remediate sediments with
contaminant concentrations between 450 /zg/kg and 300 Mg/kg PCBs.
IV. DOCUMENTATION OF CHANGES FROM THE PROPOSED ESD
In February 1997, EPA released its draft ESD (EPA, 1997) with a proposed change to the
PCB cleanup level. In the ESD, EPA proposed to modify the PCB sediment SQO from 150
A^g/kg to be achieved within 10 years after cleanup, to 450 Mg/kg, to be achieved during
cleanup. EPA received several comments on the proposed ESD, and made some modifications
to the proposed remedy in response to public comments and those of the support agencies, the
24
-------
Washington State Department of Ecology, and the Puyallup Tribe of Indians. The changes made
in response to those comments are summarized below.
Addition of a 300 /zg/kg PCB SQO
An SQO of 300 /zg/kg PCBs has been added as an enforceable standard for the CB/NT Site,
along with the 450 ^g/kg PCB SRAL proposed in the draft ESD. This means that CB/NT
sediments with PCB concentrations of 450 /zg/kg or higher will be remediated as part of any
CB/NT cleanup. In addition, natural recovery modeling will be performed to ensure that
remediation of PCB-contaminated sediments to 450 /zg/kg will result in PCB sediment
concentrations of 300 /zg/kg or below in all areas of the CB/NT Site. If natural recovery models
predict that in some areas, a PCB cleanup level lower than 450 /zg/kg is needed to achieve 300
/zg/kg PCBs within 10 years of sediment remedial action, a lower PCB SRAL will be required in
those areas.
The 450 /zg/kg SRAL is being set as the maximum PCB SRAL for any problem area at the
CB/NT Site. Even if natural recovery modeling shows that some areas with a higher PCB
sediment concentration will naturally recover to 300 /zg/kg in 10 years, areas with PCB sediment
concentrations greater than 450 /zg/kg PCBs must be remediated. EPA considers 450 /zg/kg
PCBs to be the maximum concentration of PCBs which should be allowed at the CB/NT Site
after cleanup, because of the long-term health human health and ecological effects associated
with PCBs.
Addition of an Analysis of Non-cancer Risks
In the draft ESD, EPA relied on the analysis in the 1989 ROD that a PCB cleanup level which is
protective of human health cancer risks is also protective of non-cancer risks. In response to
public concerns about non-cancer risks associated with PCBs, EPA performed an analysis of the
residual non-cancer risks associated with cleanup of PCBs to a range of potential PCB cleanup
levels (Weston, 1997b) and considered this information in its selection of the final PCB cleanup
level.
Clarification of CERCLA Requirements for Review of Protectiveness of the Remedy
Several commentors pointed out that a considerable amount of new research is being done on the
harmful effects of PCBs. In addition, NOAA continues to do research on the effects of
contaminants at the CB/NT Site to wildlife. These studies may, in the future, show that the
toxicity of PCBs is greater than currently believed. Section 121(c) of CERCLA states that if
EPA selects a remedial action that results in any hazardous substances, pollutants, or
contaminants remaining at the site, EPA shall review such remedial actions no less often than
each 5 years after the initiation of such remedial action to assure that human health and the
environment are being protected by the remedial action. Pursuant to this statutory requirement,
EPA will review the appropriateness of the remedy no less often than each 5 years after its
25
-------
initiation. If new studies indicate that PCBs are responsible for more toxic effects than was
known at the time of thisJESD, EPA will review such information to determine whether any
additional remedial action is necessary to ensure the protectiveness of the remedy.
V. AFFIRMATION OF THE STATUTORY DETERMINATION
Considering the new information that has been developed in this BSD and in the Administrative
Record, EPA believes that the modified PCB SQO and SRAL values for remediation of
contaminated sediments at the CB/NT Site remain protective of human health and the
environment, comply with Federal, State and tribal requirements that are applicable or relevant
and appropriate to this remedial action, is cost-effective, and otherwise meets the standards of
Section 121 of CERCLA.
VI. PUBLIC PARTICIPATION ACTIVITIES
EPA developed a report in October 1996 which provided human health risk calculations at a
range of potential PCB cleanup levels. That report was sent to several parties for review, and
EPA received 20 comment letters. EPA held a public meeting to discuss our proposed approach,
which 29 people attended. As noted in Section III of this ESD, EPA made several modifications
to its approach to evaluating potential cleanup levels based on these comments.
In February, EPA released its draft ESD (EPA, 1997) with a proposed change to the PCB
cleanup level. In the ESD, EPA proposed to modify the PCB sediment cleanup from 150
Mg/kg to be achieved within 10 years after cleanup, to 450 Mg/kg, to be achieved immediately
after cleanup. A fact sheet was sent in March 1997 to over 2000 people announcing EPA's
proposed change and inviting their comments. EPA placed a display ad in the Tacoma News
Tribune on March 10, 1997 announcing the public comment period and a public meeting. A
public meeting was held on March 26, 1997, to present EPA's proposal, to answer questions
about the proposal, and to take public comments. This meeting was attended by
approximately 50 people. The public comment period for the draft ESD and the
Administrative Record (AR) of documents supporting the proposal extended from March 10,
1997 to April 9, 1997. Thirty-six verbal and written comments were received by EPA, and
nine comment letters were received after the public comment period ended. All comments,
including those received before and after the public comment period, were considered in
EPA's decision. The AR for this decision contains all public comments EPA has received on
this proposal. Appendix C contains a summary of the comments received before and during the
public comment period and EPA responses to those comments.
Signed:
_
dall F. Smith, Director Date -/
Office of Environmental Cleanup
26
-------
REFERENCES
Fuglevand, Paul. 1996. FCB Residual Concentration, Arithmetic Mean. Memorandum to
Allison Hiltner, U. S. EPA, Region X. December 18.
Hartman Associates, Inc. 1996. Hylebos Waterway Pre-Remedial Design - Preliminary
Disposal Site Evaluation Report. Prepared for the Hylebos Cleanup Committee. June 26.
National Oceanic and Atmospheric Administration. 1996. Sediment PCB cleanup level;
Commencement Bay Nearshore/Tideflats Superfund Site. Letter to Allison Hiltner, U.S. EPA,
Region X. December 6.
Tetra Tech. 1988. Health Risk Assessment of Chemical Contamination in Puget Sound
Seafood. Prepared for U. S. EPA, Region X. September.
Toy, K.A., N.L. Polissar, S. Liao, and G.D. Mittelstaedt. 1996. A Fish Consumption
Survey of the Tulalip and Squaxin Island Tribes of the Puget Sound Region. Tulalip Tribes,
Department of Environment, 7615 Totem Beach Road, Marysville, WA.
U.S. Fish and Wildlife Service. 1996. Additional comments on sediment PCB cleanup level;
Commencement Bay Nearshore/Tideflats Superfund Site. Letter to Allison Hiltner, U.S. EPA,
Region X. December 2.
U.S. Environmental Protection Agency. 1989. Commencement Bay Nearshore/Tideflats
Record of Decision. September 30.
U.S. Environmental Protection Agency. 1997. Public Review Draft Explanation of
Significant Differences; Commencement Bay Nearshore/Tideflats Superfund Site. March 3.
Weston. 1996. Evaluation of Residual Risks Associated with a Range of Sediment PCB
Cleanup Levels in the Hylebos Waterway, Thea Foss Waterway, and Overall Commencement
Bay Nearshore/Tideflats Superfund Site. Prepared for U.S. EPA, Region X. October.
Weston. 1997a. Evaluation of Residual Risks Associated with a Range of Sediment PCB
Cleanup Levels in the Hylebos Waterway, Thea Foss Waterway, and Overall Commencement
Bay Nearshore/Tideflats Superfund Site - Addendum. Prepared for U.S. EPA, Region X.
February.
Weston. 1997b. Potential for Non-cancer Health Impacts from PCBs in Sediments at the
Commencement Bay Nearshore/Tideflats Superfund Site. Prepared for U. S. EPA, Region X.
July.
27
-------
Weston 1997c. Technical Memorandum - Risks from Human Consumption of PCB-
Contaminated Fish under.Background Conditions in Puget Sound. Prepared for U. S. EPA
Region 10. July.
Weston. 19976. Cost Estimate for Natural Recovery Monitoring. Letter to Allison Kilmer, U.
S. EPA Region 10. July 23.
28
-------
APPENDIX A
Table A-1—Arithmetic Mean Residual PCB Sediment Concentrations (ug/kg-OW) over a Range of
PCB Cleanup Levels
Target PCB Cleanup
Level (/jg/kg-DW)
50
100
150
300
450
600
750
900
CB/NT Site-Wide
45
47
51
63
75
82
86
89
Hytebos Waterway
20
26
37
80
124
143
156
161
Thea Foss Waterway
22
34
49
81
108
136
150
168
Table A-2—Assumptions used in Human Health Residual Risk Evaluation
Parameter
PCB oral Cancer Slope Factor
(CSF0)
Representative PCB Residual
Sediment Concentration Statistic
PCB Biota Sediment
Accumulation Factor (BSAF)
Fraction Lipid in Fish Tissue
(U)
Exposure Duration (ED)
Exposure Frequency (EF)
Averaging Time (AT)
Body Weight (BW.)
Value1
2.0 (mg/kg-day)-1
arithmetic mean (see values in
Table A-1)
1.72
2.6%
30 yr
350 days/yr
70 yr
70kg
Source
Updated toxicrty value from
EPA's Integrated Risk
Information System (IRIS)
As recommended by EPA
Site-specific, bay-wide value
Site-specific, bay-wide value
EPA default upper end
residency value
Standard default exposure
factor, allows for 2 weeks away
per year
Standard default human lifetime,
used for averaging all cancer
risks
Average adult human body
weight
'Shaded values indicate those values that have been revised since the 1989 ROD.
-------
Table A-3—Flsii Ingestlon Rates used in Human Hearth Residual Risk Evaluation
Exposure Scenario
RME-High-end Tribal
Average Tribal
High-end Recreational
Average Recreational
Ingestion Rate (g/day)
123'
41.72
95. 13
12.34
Fraction Consumed Fish from the CB/NT Site (%)
695
69s
100
100
1 Upper 90th UCL of the mean ingestion rate of all finfish from Toy et al.. 1996
2 Mean ingestion rate of all finfish from Toy et al.. 1996
1 Upper 95th UCL of the mean fish ingestion rate from Tetra Tech composite study utilized in the ROD
4 Mean fish ingestion rate from Tetra Tech composite study used in the ROD
5 Weighted average (based on mean consumption rates by each tribe of each type offish) of percentages of finfish
consumed from Puget Sound from Toy et al., 1996
Table A-4-Estimated Residual Cancer Risks after cleanup to a range of PCB cleanup levels
Target PCB Cleanup
Level (Mg/kg-DW)
50
100
150
300
450
600
750
900
CB/NT Site-Wide
8.3E-5
8.7E-5
9.4E-5
1.2E-4
1.4E4
1.5E-4
1.6E-4
1.6E-4
Hylebos Waterway
2.7E-5
3.4E-5
4.9E-5
1.1 E-4
1.6E-4
1.9E-4
2.1 E-4
2.1 E-4
Thea Foss Waterway
2.1E-5
3.3E-5
4.6E-5
7.6E-5
1.0E-4
1.3E-4
1.4E-4
1.6E-4
-------
Figure A-1-Hylebos Waterway Target PCB Cleanup Level vs Sediment Removal Volume
2.000.000
1,800,000
1,600,000 -
3. 1.400,000 -
0)
E
-5 1,200,000 •
16
S 1,000.000 -
E
CD
£ 800,000 -
Q)
E
CO
600,000 •
400.000 •
200,000 •
* 0
200
400 600 800 1000
Target PCB Cleanup Level (ppb-DW)
1200
1400
1600
JRCE. Memorandum from Paul Fugelvand (Dalton. Olmsted Fuglevand, Inc.) to Allison Hiltner (US EPA)
larding PCB Residual Concentration. Arithmetic Mean. December 18. 1996
2/26/97 3:55 PM
-------
APPENDIX B
WASHINGTON DEPARTMENT OF ECOLOGY LETTER
SUPPORTING EPA'S EXPLANATION OF SIGNIFICANT DIFFERENCES
-------
V
X» <>j,
*t,
-------
Mr. Randy Smith
June 13, 1997
Page 2
Ecology's has determined that this cleanup scenario ensures significant additional risk reduction
with minimal additional costs expected. Over the 10 year period of natural recovery, mean
concentration for PCBs would decrease from about 124 to 80 ppb for Hylebos, from 1 08 to 8 1
for Thea Foss and from 75 to 63 Bay-wide. Concurrently, human health cancer risks would be
reduced by about 34% ,31%, and 14% (Hylebos, Thea Foss and Bay-wide, respectively) for the
scenario that is protective of the above average Tribal subsistence fisher. Non-cancer human
health hazard quotients would decrease from 9 to 6 for Hylebos, from 6 to 4 for Thea Foss and
from 8 to 7 for Bay-wide and, hazard indices for 3 groups of wildlife (juvenile salmonids,
shorebirds and piscivorous birds) would be reduced below one.
We believe our concepts could achieve a cleanup that makes great progress in meeting state
requirements without substantial cost increases. We also recognize the difficult job the
Environmental Protection Agency has in attempting to reconcile the many views presented to
you and we hope that our ideas might help bring the project to a successful conclusion for all of
the parties involved.
Sincerely,
Mary Burg
Program Manager
Toxics Cleanup Program
MB:jr
cc: Dave Jansen, Ecology
Russ McMillan, Ecology
-------
APPENDIX C
PUBLIC COMMENT RESPONSIVENESS SUMMARY
July 28, 1997
-------
Cfl/W PCB Responsiveness Summary
TABLE OF CONTENTS
Section Page
A. OVERVIEW OF THE REEVALUATION OF THE PCB CLEANUP LEVEL AND
RELATED COMMUNITY INVOLVEMENT 1
B. OVERVIEW OF COMMUNITY CONCERNS AND EPA RESPONSE 2
C. RESPONSES TO SPECIFIC PUBLIC CONCERNS 4
Part I—Responses to Comments Received Prior to Issuance of the Draft ESD 4
1 Comments Regarding the Risk Evaluation Approach 4
2 Comments Regarding Risk Management Issues 23
Part II—Responses to Comments Received Subsequent to Issuance of the Draft ESD . 27
3 Comments Related to the Proposed PCB Cleanup Level 28
4 Comments Related to the Risk Evaluations 32
5 Comments Related to Remedial Decision-Making 53
Part III—List of Commentors 69
D. REFERENCES 71
E. GLOSSARY OF TERMS 75
-------
PCB CLEANUP LEVEL REEVALUATION
COMMENCEMENT BAY NEARSHORE/TIDEFLATS SUPERFUND SITE
PUBLIC COMMENT RESPONSIVENESS SUMMARY
A. OVERVIEW OF THE REEVALUATION OF THE PCB CLEANUP LEVEL AND
RELATED COMMUNITY INVOLVEMENT
In 1983, the Commencement Bay Nearshore/Tideflats (CB/NT) Superfund Site was placed on
EPA's National Priorities List (NPL) of sites requiring investigation and cleanup. In 1989,
the Record of Decision (ROD; EPA, 1989a) addressing the marine sediment problem areas
within the CB/NT Site was released. The ROD addressed the seven waterways (i.e., the
Hylebos, Blair, Sitcom, Milwaukee, Saint Paul, Middle, and Thea Foss/Wheeler-Osgood
waterways) extending southeast from Commencement Bay, and the Nearshore area (i.e., area
from the Commencement Bay shoreline to the 60-foot bathymetric contour) of the bay. (See
Figure 1 in the Explanation of Significant Differences [EPA, 1997a].) It did not include the
Ruston shoreline, which will be addressed in a separate ROD. The CB/NT ROD established a
sediment quality objective (SQO) for polychlorinated biphenyl compounds (PCBs) of 150
micrograms per kilogram Cug/kg, dry weight) based on protection of human health from
ingestion of contaminated fish. The 150 Mg/kg SQO was to be met within 10 years of active
sediment cleanup.
During pre-design sampling, new data were collected from the Hylebos Waterway showing
that approximately twice the amount of sediment originally estimated in the ROD would
require cleanup; and EPA has updated the toxicity information it used to assess human cancer
risks associated with PCBs. In response to concerns about these issues, EPA decided to
reevaluate the PCB sediment cleanup level for the CB/NT Site. The Evaluation of Residual
Risks Associated with a Range of Sediment PCB Cleanup Levels in the Hylebos Waterway, the
Thea Foss Waterway, and the Overall Commencement Bay Nearshore/Tideflats Superfund Site
(referred to as the Evaluation of Residual Risks) was prepared in October 1996 to form a
technical basis for the revaluation of the PCB cleanup level (Weston, 1996). The report
quantitatively evaluated residual risks to human health at a variety of potential PCB cleanup
levels using updated sediment data, and the risk calculation and exposure assumptions in the
1989 ROD. It also contained an uncertainty analysis which qualitatively evaluated the impact
of varying the assumptions used in the risk estimates. EPA distributed the Evaluation of
Residual Risks report (Weston, 1996) to 31 parties and held a comment period in November
1996. EPA held an informational meeting attended by approximately 29 people on November
4, 1996, at the World Trade Center in Tacoma, Washington. EPA received approximately 20
comment letters during this informal comment period.
In response to comments on the Evaluation of Residual Risks report, EPA updated the risk
equations and several of the input parameter values. This analysis was released in February
1997 as the Addendum to the Evaluation of Residual Risks (Weston, 1997a). As with the
Evaluation of Residual Risks, the Addendum presents residual human cancer risks at potential
target PCB sediment cleanup levels from 50 to 900 Mg/kg. In addition, the Addendum
-------
CB/NT PCB Responsiveness Summary
includes a brief evaluation of ecological risks over a range of PCB cleanup levels. EPA
considered the information presented in both the Evaluation of Residual Risks and the
Addendum in its revaluation of the PCB cleanup level.
In February, EPA released its draft Explanation of Significant Differences (ESD) (EPA,
1997a) with a proposed change to the PCB cleanup level. In the ESD, EPA proposed to
modify the PCB sediment cleanup from 150 Mg/kg to be achieved within 10 years after
cleanup, to 450 Mg/kg. to be achieved during cleanup. A fact sheet was sent in March 1997 to
over 2,000 people announcing EPA's proposed change and inviting their comments. EPA
placed a display ad in the Tacoma News Tribune on March 10, 1997, announcing the public
comment period and a public meeting. A public meeting was held on March 26, 1997, to
present EPA's proposal, to answer questions about the proposal, and to take public comments.
This meeting was attended by approximately 50 people. The public comment period for the
draft ESD and the Administrative Record of documents supporting the proposal extended from
March 10, 1997, to April 9, 1997. Thirty-six verbal and written comments were received by
EPA.
This Responsiveness Summary addresses comments on the October 1996 Evaluation of
Residual Risks report (Weston, 1996) as well as comments shared with EPA during the public
comment period following release of the draft ESD.
B. OVERVIEW OF COMMUNITY CONCERNS AND EPA RESPONSE
Comments received by EPA on the October 1996 Evaluation of Residual Risks report
(Weston, 1996) ranged from concerns that the risk calculations were overly conservative and
would result in an unnecessary degree of cleanup to concerns that any increase to the existing
PCB cleanup level (150 A*g/kg) would not be adequately protective of people and wildlife in
the vicinity of the CB/NT Site. Commentors requested the use of more up-to-date values to
calculate site risks. Concerns were expressed regarding the value chosen to represent the
residual PCB sediment concentrations after cleanup at the selected target cleanup levels.
Commentors also requested that ecological risks associated with PCB exposure be more
directly addressed.
Many of the comments requesting that EPA use updated values in its risk calculations were
addressed in the February 1997 Addendum to the Evaluation of Residual Risks (Weston,
1997a). A new cancer slope factor, exposure duration, and fish consumption rate, along with
updated sediment chemistry data (for PCB concentrations and total fraction of organic carbon).
were used in risk calculations. A brief evaluation of risks to selected ecological receptors was
performed using data provided by the natural resource agencies.
tf
Many of the concerns expressed in November 1996 were repeated during the March 1997
public comment period, along with several additional concerns. Although some comrnentors
agreed with EPA's proposed PCB cleanup level of 450 Mg/kg, many comrnentors expressed
-------
CB/NT PCB Responsiveness Summary
•wncerns that the proposed cleanup level was too high or too low. Some felt that it was not
adequately protective of people and ecological receptors in the vicinity of the site, while others
felt that it mandates a higher cost than other potential cleanup levels (e.g., 600 jug/kg) without
providing a significant increase in protect!veness. Commentors also questioned EPA's
application of the nine Superfund evaluation criteria discussed in the BSD in relation to
establishing the proposed new PCB cleanup level. Additionally, commentors cited a need to
account for multiple chemical effects and global distributions of PCBs. Several commentors
expressed concerns about non-cancer risks associated with PCBs. Some commentors were
also concerned about the feasibility of cleaning up sediments to a 450 Mg/kg PCB cleanup level
and the potential for increased ecological risks during the remedial efforts.
In making its final decision on the PCB cleanup level for Commencement Bay sediments, EPA
carefully examined data regarding cancer and non-cancer risks to both people and the
environment to ensure that the proposed PCB cleanup level would be protective of both human
health and the environment. In response to concerns about non-cancer human health risks,
EPA evaluated potential non-cancer risks in a technical memorandum entitled "Potential for
Noncancer Health Impacts from PCBs in Commencement Bay Sediments" (Weston, 1997b).
This technical memorandum has been added to the Administrative Record for this ESD. EPA
also added to the Administrative Record a technical memorandum on PCB concentrations in
non-urban areas of Puget Sound (Weston, 1997c) to address questions about background
iitions. To address concerns that EPA should provide some assurance PCB concentrations
11 be reduced over time, EPA added a requirement that PCB concentrations in sediments
must be reduced to 300 Mg/kg within 10 years after completion of the cleanup. If the 300
Mg/kg standard is not achieved through natural recovery, additional cleanup work will be done
to meet that standard.
Specific responses to the concerns expressed by commentors are provided in the following
subsections within Section C:
Part I — Responses to Comments Received Prior to Issuance of the Draft ESD
1. Comments Regarding the Risk Evaluation Approach
2. Comments Regarding Risk Management Issues
Part II — Responses to Comments Received Subsequent to Issuance of the Draft ESD
3. Comments Related to the Proposed PCB Cleanup Level
4. Comments Related to the Risk Evaluations
5. Comments Related to Remedial Decision-Making
Part III - List of Commentors
Part I, which comprises sections 1 and 2, focuses on comments received during the 1996
comment period following release of\he Evaluation of Residual Risks (Weston, 1996).
Part II, which comprises sections 3 through 5, focuses on comments received during the 1997
public comment period following release of the draft ESD and supporting information.
-------
CB/NT PCB Responsiveness Summary
including the AddendunT (Western, 1997a). The responses to comments are followed in
Pan in by a list of the commentors. References are listed in Section D. Several technical
terms are used in the comments and responses. To assist the non-technical reader, technical
terms are defined in the glossary provided in Section E.
C. RESPONSES TO SPECIFIC PUBLIC CONCERNS
Part I—Responses to Comments Received Prior to Issuance of the Draft ESD
1 Comments Regarding the Risk Evaluation Approach
1.1 Comments related to use of up-to-date values for toricity and exposure parameters
and an up-to-date equation for calculating residual risks
1.1.1 The cancer slope factor in EPA's Integrated Risk Information System has been reduced
from 7.7 (mg/kg-day)'1 to a range of 0.04-2.0 (mg/kg-day)'1. EPA's risk evaluation should
use the more updated cancer slope factor to evaluate residual risk. (17, 14. 15, 6, 32)
Response: EPA agrees and has applied the updated oral cancer slope (toxicity) factor
of 2.0 (mg/kg-day),'1 as recommended for food chain exposures such as fish
consumption, in the assessment of residual risks presented in the Addendum to the
Evaluation of Residual Risks Report (Addendum; Weston, 1997a). The Addendum
contains the residual risk calculations used by EPA in its revaluation of the PCB
cleanup level for the CB/NT Site.
1.1.2 EPA could simply update the PCB cancer slope factor (CSF), leave all other exposure
assumptions unchanged, and the PCB cleanup criterion for the Hylebos would become
consistent with other RODs in the nation. Updating only the cancer slope factor in the ROD
risk assessment, without changing any of the other out-dated exposure assumptions, would
change the PCB cleanup criteria from 150 ng/kg to 1,500 ng/kg. If the cancer slope factor
decreases from 7.7 to 2 (a decrease by a factor of 3.85) and all other parameter values remain
the same (including risk), the allowable PCB concentration in fish will increase by a factor of
3.85 to 139 mg/kg. The corresponding PCB concentration in sediment will then increase to
116 mg/kg. Using the tables from Appendix B of EPA Region X's October 1996 report
(Weston, 1996), a geometric mean concentration (the parameter used by EPA) of 116 mg/kg
corresponds to a cleanup level of 1,500 mg/kg. By making only this one change, a roadblock
to cleanup of the Hylebos would be removed without having to conduct any additional studies
or risk assessments. (17, 22)
Response: As addressed in Comment 1.1.1, the updated cancer slope factor was
applied to the calculation of residual risks presented in the Addendum (Weston, 1997a).
However, while reducing the cancer slope factor would decrease residual risk
estimates, that alone would not reflect other key changes EPA has made to exposure
-------
CB/NT PCS Responsiveness Summary
values used to calculate human health risks. In the eight years since the 1989 CB/NT
ROD, EPA made changes in several exposure assumptions based on new research and
new EPA policies. For this reason it would have been inappropriate to update one
parameter value without examining all parameter values. Furthermore, as discussed in
response to Comment 1.2.2, it is more appropriate to use the arithmetic mean, not the
geometric mean, to represent residual sediment concentrations of PCBs at the CB/NT
Site. The Addendum includes an analysis of residual risks utilizing both updated
toxicity information and updated exposure scenarios relevant to the CB/NT Site.
1.1.3 Updated data sets should be used where practical (e.g.. PCB cancer slope factors,
total organic carbon in sediment, biota-sediment accumulation factors (BSAFs)for
PCBs from other analyses, seafood consumption rates). (2, 33. 4. 31, 21, 8, 32, 6)
Response: As noted in the responses to Comments 1.1.1 and 1.1.2, updated
information concerning the cancer slope factor, as well as the exposure factors (e.g.,
the ingestion rate), were applied to the calculation of residual risks presented in the
Addendum (Weston, 1997a). Updated parameter values were used for the cancer slope
factor, the fish ingestion rate, the exposure duration, exposure frequency, and the total
organic carbon fraction in the sediment. The biota-sediment accumulation factor
(BSAF) value for PCBs used in the ROD was retained because it was based on site-
specific data.
1.1.4 The result of incorporating more recent values for risk parameters would undoubtedly
decrease overall risk. Because EPA's report used outdated assumptions and generic values
instead of current science and site-specific data, it is clear that the risk assessment should have
been updated. (17)
Response: As discussed in the response to Comment 1.1.3, the risk equation input
parameter values were updated for all relevant parameters. However, all parameter
value changes did not result in a lowering of the risk estimates. For example, while
the lower toxicity factor decreased risk estimates, the higher fish ingestion rate
increased the risk estimates.
1.1.5 Some of the exposure parameter values used by EPA in their risk assessment are in
excess of those appropriate for estimating a reasonable maximum exposure (RME) such as
exposure duration and exposure frequency. EPA research has already established that the
average exposure duration is nine years, with 30 years as the reasonable maximum. (17)
Response: A 70-year exposure duration and an exposure frequency of 365 days per
year were applied to risk calculations in the 1989 ROD and the October 1996
Evaluation of Residual Risks (Weston, 1996). To be consistent with updated EPA
policy, a 30-year exposure duration and an exposure frequency of 350 days per year
were applied to risk calculations in the Addendum (Weston, 1997a). The 350 days per
year exposure frequency assumes two weeks away from home per year. EPA guidance
-------
CB/NT PCB Responsiveness Summary
calls for a nine-year average and a 30-year high-end exposure duration based on fixed
locations for exposure such as living next to a Superfund site and assumes that people
change residences every nine years (on average). This assumption does not account for
the fact that many people, when they move, stay within the region and likely retain use
of the same recreational areas (e.g., Commencement Bay). Since data were not
available to describe residence time in larger areas, EPA's high-end default value of 30
years was used to represent exposure duration for average and high-end exposure
scenarios for PCBs at the CB/NT Site.
1.1.6 EPA justifies an exposure duration of 70 years because PCBs bioaccumulate and have
an extended half-life in the body. Using bioaccumulation to justify an extended exposure
duration is inconsistent with EPA Headquarters' policy, as documented in EPA's current
Integrated Risk Information System (IRIS) file for the PCB cancer slope factor. (17)
As noted in the response to Comment 1.1.5, an exposure duration of 70 years was used
in the October 19% Evaluation of Residual Risks (Weston, 19%) to be consistent with
the 1989 CB/NT ROD. Exposure duration was reduced to 30 years in the Addendum
(Weston, 1997a). In EPA's most recent evaluation of PCB toxicity, as documented in
IRIS (EPA, 1997b), no quantitative factor was provided to account for continued
exposure to PCBs from the slow release of PCBs from fatty tissue within a person's
body. In neither the Evaluation of Residual Risks (Weston, 19%) nor the Addendum
(Weston, 1997a), was this potential exposure factored into the risk estimates.
However, in both documents, this potential impact to PCB toxicity was discussed in the
uncertainty analysis so that EPA could consider this factor, at least qualitatively, in its
reevaluation of the PCB cleanup level.
/. 1.7 National guidance calls for risk assessments to evaluate risk for both the average
exposure and the reasonable maximum exposure (RME). Based on their choice of parameter
values in the October 1996 report, EPA Region X has assessed the RME. (17)
Response: The PCB cleanup level in the 1989 ROD was based on a risk analysis using
fish consumption rates for an average recreational fisher. The Evaluation of Residual
Risks (Weston, 1996) applied that scenario as a reasonable exposure scenario for the
Site. As noted in responses to several of the previous comments, several adjustments
to the risk assessment input parameters were needed to update the risk assessment to
develop what would be considered a reasonable maximum exposure (RME) under
current EPA guidance. The analysis of residual risks presented in the Addendum
(Weston, 1997a) included evaluations of average and high-end recreational fisher
scenarios, as well as average and high-end tribal fisher scenarios. EPA chose the high-
end tribal fisher scenario to represent the RME for the CB/NT Site.
1.1.8 EPA did not reduce the concentration of PCBs in fish as a result of cooking in spite of
the fact that the decrease has been documented in the scientific literature. (17)
-------
CB/NT PCB Responsiveness Summary
Response: As discussed in the uncertainty analysis in the Evaluation of Residual Risks
(Western, 19%), reduction of PCB levels in fish following cooking was considered. It
is recognized that the various methods of preparing fish for consumption may affect
concentrations of PCBs in tissue consumed. Although some studies report that cooking
can substantially reduce PCB concentrations in fish tissue, other studies have shown
that PCB loss during cooking may be as little as two percent. Some cooking methods
also activate or create other carcinogenic chemicals. Because of the uncertainties about
the net effects of cooking on PCB concentrations in fish tissue, quantitative corrections
for the effects of cooking in the risk assessment are not possible at this time.
Furthermore, even if the PCB content in fish tissue decreased, content of the "pan
juices" may include lost PCB amounts from the fish tissue; should these "pan juices"
be consumed (e.g., used to make a sauce, or directly poured over the fish), the
decrease in PCB fish tissue levels will be balanced, and individuals may still be
exposed to nearly one hundred percent of PCBs present in the fish tissue. Also, some
cooking methods remove far less PCBs than others; therefore, EPA's assumption of
100 percent of PCBs remaining after cooking was considered appropriate to ensure that
derived cleanup levels would be protective of individuals who prepare their fish any
number of ways.
1.2 Comments related to the calculation of residual PCB sediment concentrations
7.2.7 The statistical assumptions and methods used for the calculation of the residual
sediment PCB concentrations (such as the use of half the sample's reported detection limit for
a concentration value) do not appear to be completely supported try the data and could result
in a bias (lowering) of the risk estimate. (8)
Response: Use of half the sample detection limit as the concentration for an
undetected analyte is consistent with EPA risk assessment methodology. It assumes
that the actual PCB concentrations are evenly distributed between zero and the
detection limit, and may either over- or underestimate the true concentration (resulting
in either an elevated or lowered estimate of risk). These possibilities are acknowledged
in the uncertainty analysis presented in the Evaluation of Residual Risks (Western,
1996) and the Addendum (Weston, 1997a) reports. EPA considered these uncertainties
in its revaluation of the PCB cleanup level.
7.2.2 The assumption of a lognormal distribution for the data set does not appear to be
completely supported by the data and could result in a bias (lowering) of the risk estimate.
The uncertainty associated with use of the geometric mean should be discussed. (8, 2)
Response: In preparation of the Evaluation of Residual Risks report (Weston, 1996),
the distribution of the PCB concentration data was generally found to be lognormal for
the individual waterways and for the overall CB/NT Site. This follows the theoretical
distribution of most environmental chemical data. Based on this analysis, a decision
-------
CB/NT PCB Responsiveness Summary
was made to assume a lognormal distribution; therefore, a geometric mean
concentration was used to represent the average residual PCB sediment concentration in
the Evaluation of Residual Risks report (Weston, 1996).
However, because use of the geometric mean concentration is not consistent with
current EPA risk assessment policy (EPA, 1992), the analysis presented in the
Addendum (Weston, 1997a) applied the arithmetic mean residual PCB sediment
concentration. Use of the arithmetic mean provided a more conservative representation
of the residual PCB concentration to which fish would be exposed and, therefore,
allowed for the possibility that some fish may be exposed to higher than average
concentrations of PCBs in the sediments due to a limited home range or a preference
for feeding in the more contaminated areas. See also EPA's response to Comment
4.1.2.
7.2.3 The arithmetic mean is appropriate regardless of the pattern of daily exposure over
time or the type of statistical distribution that might best describe the sampling data. The
geometric mean of a set of sampling results, however, bears no logical connection to the
cumulative intake that would result from long-term contact with site contaminants, and it may
differ appreciably from - and be much lower than - the arithmetic mean...Thus, preferential
feeding by smaller forage fish in the more contaminated intertidal areas, ultimately results in a
greater body burden in the larger fish that feed on these fish, than would be estimated by
either an arithmetic mean, or an area-weighted average. (24)
Response: As discussed in the response to Comment 1.2.2, the arithmetic mean
residual PCB sediment concentration was used in the risk evaluations presented in the
Addendum (Weston, 1997a).
1.2.4 The original EPA assessment considered a population that is a hybrid between the RME
and the average exposure. For example, the fish consumption value is for average exposure;
the concentration value is the geometric mean of PCB concentration in fish which is lower
than both the arithmetic mean (which according to EPA guidance, should be used to assess
average exposure) and the 95th percent confidence interval on the mean (which according to
EPA guidance, should be used to assess the RME). (17)
Response: As noted in previous comments, EPA agrees that the evaluation of risks in
the 1989 ROD is not consistent with current EPA guidance for estimating either an
average or a reasonable maximum exposure (RME). Both the fish consumption value
and the residual sediment concentrations were modified in the risk evaluation presented
in the Addendum (Weston, 1997a). As discussed above in response to Comment 1.1.7,
four exposure scenarios were evaluated, the RME scenario being represented by an
high-end tribal fisher. As discussed in response to Comment 1.2.2, the arithmetic
mean was used to represent the residual PCB sediment concentration usedvtn risk
calculations in the Addendum (Weston, 1997a). See the response to Comment 4.1.2
for a discussion of use of the 95th percent upper confidence level.
8
-------
CB/NTPCB Responsiveness Summary
1.2.5 On the basis of demographic and fish consumption information for this region, it is
unlikely that a large number of individuals would obtain a large percentage of the fish they
consume from either the Hylebos or Thea Foss/Wheeler Osgood Waterways within the CB/NT
site. Therefore, the range in risk represented by the entire Site should be used as the basis for
decisions. (2)
Response: EPA agrees.
7.2.6 The analysis as presented does not provide enough information to evaluate the effects
on risk estimates of raising SQOsfor CB/NT. Information that would allow such an
evaluation include: analysis of the adequacy of the available sediment samples for determining
a geometric mean of sediment for the CB/NT Site. (2)
Response: Analyses of sample adequacy and the uncertainty associated with using
sediment samples to estimate fish tissue concentrations are qualitative analyses.
However, over the past three years, more than 200 sediment samples have been
collected from each of the Hylebos and the Thea Foss waterways, waterways in which
PCBs were identified as a major contaminant of concern. These samples were
incorporated into the residual risks estimated in both the Evaluation of Residual Risks
(Western, 19%) and the Addendum (Western, 1997a). EPA considers this an adequate
number of samples to characterize PCB levels. Use of the geometric mean is further
discussed in responses to Comments 1.2.2 and 4.1.2.
1.2.7 The analysis as presented does not provide enough information to evaluate the effects
on risk estimates of raising SQOsfor CB/NT. Information that would allow such an
evaluation include presentation of the pre-cleanup geometric mean sediment level and
discussion of the drop in geometric mean sediment levels under each SQO in relation to the
desired drop in fish PCB levels from 330 ng/kg (i.e., ratio of pre-cleanup sediment and fish
tissue level should be equivalent to ratio of post-cleanup sediment and fish tissue level). (2)
Response: The information requested in this comment is presented in Tables 3-2, 3-3,
and 3-4 of the Evaluation of Residual Risks (Weston, 1996). As stated in this report, it
is assumed that there is a linear relationship between the residual sediment PCB
concentration and the fish tissue PCB concentration.
1.3 Comments related to uncertainties associated with the residual risk estimates
1.3.1 EPA claimed that uncertainty related to the several factors in the risk equation would
balance out because of perceived changes in several of them, affecting the risk estimates in
both directions. This rationale does not address the question of magnitude, where one or a
few factors carry significant weight or value in the equation. The analysis as presented does
not provide enough information to evaluate the effects on risk estimates of raising SQOsfor
CB/NT. Quantitative description of uncertainty and variability in the data should be
incorporated into the risk estimates. (2, 33, 8)
-------
CB/NT PCB Responsiveness Summary
Response: Qualitative and semi-quantitative uncertainty analyses were conducted for
the Evaluation of Residual Risks (Weston, 1996) and are summarized in the Addendum
(Weston, 1997a). EPA did not attempt a quantitative uncertainty analysis because it
would have been an extremely difficult undertaking, and would have been only
marginally more useful than the qualitative uncertainty analysis for the purposes of
decision-making. Uncertainty is inherent in most risk calculations due to natural
variability and uncertainty in various input parameters, some of which can, and some
of which cannot, be quantified. EPA considered all uncertainties linked with residual
risk estimates presented in both the Evaluation of Residual Risks (Weston, 19%) and
the Addendum (Weston, 1997a) in its reevaluation of the PCB cleanup level for the
CB/NT Site.
1.3.2 This proposed SQO represents a potential concern because nonionic organics
(including PCBs) require normalization to total organic carbon (TOC) to reduce a major
source of variability which are attributed to the differences in the amount of carbon.... given
that the organic carbon data from Hylebos Waterway and Thea Foss/Wheeler-Osgood
Waterway range from approximately 0.4 to 13.5 percent (Table A-l) and from 0.3 to 16
percent (Table A-7) respectively; this could result in an error between one to two orders of
magnitude (ten to one hundredfold) based upon sampling alone. (12)
Response: As described in the first equation in Section 2.2 of the Evaluation of
Residual Risks (Weston, 1996), the fraction of total organic carbon (TOC) in sediment
is accounted for in the calculation of the residual PCB fish tissue concentration from
PCB sediment concentration. In the Addendum (Weston, 1997a), the TOC values were
updated to reflect current sampling data. The fraction of total organic carbon was
calculated for the overall CB/NT Site, for the Hylebos Waterway, and for the Thea
Foss Waterway, using data collected over the past three years. EPA considers this
application of average organic carbon fractions representative of the range of organic
carbon fractions throughout the CB/NT Site, including its waterways, and considers
this appropriate for use in the evaluation of residual risks.
1.3.3 Fish tissue concentrations should always be expressed in terms of a lipid-normalized
value when dealing with nonionic organic chemicals. This reduces the variability that lipid
(fat) has on sample concentrations. (12)
Response: EPA recognizes the need to account for lipid normalization when
evaluating the bioaccumulation of PCBs from sediment into fish tissue. Lipid
normalization is addressed via the first equation in Section 2.2 of the Evaluation of
Residual Risks (Weston, 1996) and via equation 2 in the Addendum (Weston, 1997a).
The fraction of fish lipid content is included in the calculation of the residual PCB fish
tissue concentration. The fraction used in this calculation was a site-specific value
calculated in support of the ROD (EPA, I989a). Additionally, this same fraction of
fish lipid was applied to residual risk calculations presented in the Addendum (Weston,
I997a).
10
-------
CB/NT PCB Responsiveness Summary
1.3.4 When PCBs are not present, it cannot be concluded that if other chemicals are cleaned
up to levels which are protective of ecological endpoints, that these levels would also be
considered protective of human health. All chemicals with a log K^ value of greater than 5
should be examined to ensure that sediment criteria that are based on human health do not
suggest a lower cleanup level than ecological criteria values. (12)
Response: In developing the 1989 ROD, EPA used a human health risk assessment to
determine which chemicals may cause a human health threat to consumers of fish and
shellfish. It was determined that only PCBs were present in sufficiently high
concentrations in fish and shellfish to pose a threat to humans. Therefore, the
ecologically based cleanup criteria for all contaminants except PCBs will result in
residual contaminant concentrations that are protective of human health.
1.4 Comments related to exposure scenarios evaluated
/. 4.1 While the 1981 Pierce survey did not address waterway by waterway use of the bay, 25
percent of respondents reported fishing or harvesting shellfish in the Commencement Bay area.
This summer, a CUB volunteer pollution monitor in the Hylebos Waterway witnessed a
commercial trawler harvesting fish from the waterway. CHB 's survey and monitoring of the
bay and its waterways clearly illustrate that people are using the area for fishing. Protection
of the community's health must remain a priority when considering any change in the PCB
cleanup level. (8)
Response: EPA agrees that protection of human health must be a primary
consideration in its reassessment of the PCB cleanup level for the CB/NT Site. EPA's
policy is to base its risk assessments on a "reasonable maximum exposure" scenario.
Because the CB/NT Site is part of the usual and accustomed fishing grounds for the
Puyallup Tribe, EPA modified its assumptions for fish consumption rates from those
presented in the 1996 Weston report to reflect a tribal fishing scenario. The 1997
Weston Addendum evaluates risks to a tribal member who consumes a higher than
average amount of fish from the CB/NT Site. By protecting tribal members, who
generally consume more fish than recreational fishermen, EPA believes it is setting
cleanup levels which are protective for all members of the community.
1.4.2 Subsistence fishing is no longer done by the Tribes, not just because of contamination,
but because times have changed. (21)
Response: EPA policy and regulations require that Superfund sites be cleaned up to
concentrations low enough to be protective under both current and reasonably likely
future uses for a site. EPA guidance requires that we evaluate risks associated with the
"highest level of exposure and risk that can be reasonably expected to occur" for
anticipated future uses of the site (EPA, 1991). Members of the Puyallup Tribe have
treaty rights to fish in Commencement Bay. In addition, members of the Tribe have
indicated to EPA that they would fish more in Commencement Bay if the contaminated
11
-------
CB/NT PCB Responsiveness Summary
sediments were cleaned up and they were not concerned about eating contaminated fish
(see Comment 1.4.3). For these reasons, EPA believes that a tribal fishing scenario is
an appropriate scenario to consider in the reevaluation of PCB sediment cleanup levels.
1.4.3 Many activities including recreation, fishing, and shellfishing, are no longer being
pursued by the Tribe due to the negligence of industry in the Hylebos Waterway. (29)
Response: As indicated in response to Comment 1.4.2, EPA requires that both current
and reasonably likely future uses for a site be evaluated. EPA acknowledged potential
use of the CB/NT waters by the Tribe and, therefore, considered risks to tribal fishers
in the Addendum (Weston, 1997a).
1.4.4 There is no documented subsistence fishing in the Hylebos Waterway. As noted by
EPA, the Pierce et al. study (1981) did not document subsistence fishing activities in the
Commencement Bay area. Although we do not dispute that more intensive fishing can be
associated with Native American culture or subsistence factors, the cited studies are not
relevant to Hylebos for the following reasons..Although Native Americans fish in
Commencement Bay and in the Puyallup River, there is no evidence of Native American fishing
in the Hylebos. In fact, site-specific data (Pierce et al., 1981) found only one Native American
catching fish in the entire summer survey and none in the fall survey out of the about 500 total
individuals surveyed for the entire area. (17)
Response: See response to Comment 1.4.2. Data on current practices alone are not
always sufficient to evaluate potential future exposures, such as that resulting from an
increase in Native American or subsistence fishing. Additionally, it should be noted
that the design of the Pierce et al. study may have underestimated fish catch by the
subsistence fishing population. Subsistence fishers often harvest at different times than
recreational fishers, so they may have not been present at times when the surveyors
were observing. Also, subsistence fishers may be less likely to share information with
surveyors. The RME scenario evaluated by EPA was based on a fish consumption
survey for Native Americans in the Puget Sound area. This scenario, with its relatively
high consumption rate, is expected to be protective of non-tribal subsistence fishers as
well.
7.4.5 Prior to changing the existing cleanup level for PCBs, it would be beneficial to assess
current fish consumption rates (the old data were collected prior to the installation of the Les
Davis Pier and the posting of the waterways with fish consumption warning signs). (34)
Response: As noted in responses to Comments 1.4.1 and 1.4.2, EPA requires that
both current and reasonably likely future use of a site be evaluated. Fish consumption
rates used in the ROD for the recreational fisher were determined by examination of
four different consumption studies, including the Pierce et al. study (1981). To
calculate risks to the reasonably maximally exposed individual, EPA used high-end
tribal fish ingestion rates as reported in a 1996 study representing two tribes from the
12
-------
CB/NT PCB Responsiveness Summary
Puget Sound area-fToy et at., 1996). EPA considered these fish ingestion data as the
most representative data to account for current and potential future human exposures.
A study conducted to assess current use of the CB/NT waters would be confounded by
the fact that several measures, including the posting of warning signs by the Tacoma-
Pierce County Health Department (TPCHD), have been taken to discourage fishing in
CB/NT waters. Therefore, a study of current conditions would not necessarily address
potential future uses. Furthermore, fish consumption studies require a considerable
amount of time. Conducting additional consumption studies at this time would involve
a large time delay in cleanup. Because available data are considered of adequate
quality, EPA does not believe that additional studies are warranted at this time.
1.4.6 Documentation should be provided to substantiate EPA's assumption that although the
number of people may be small, some individuals gather 100 percent of their fish from the
Hylebos Waterway. (17)
Response: EPA is unaware of any documentation on whether or not some individuals
currently collect fish exclusively from the Hylebos Waterway. However, EPA must
make cleanup decisions based on current and potential future exposure scenarios. For
the CB/NT Site, EPA is basing its PCB cleanup decision primarily on overall CB/NT
Site-wide risks. Waterway-specific risk estimates are presented to provide information
on the possible extreme exposures and to help EPA put CB/NT Site-wide risks into
perspective.
1.4.7 Although we do not dispute that more intensive fishing can be associated with Native
American culture or subsistence factors, the cited studies are not relevant to Hylebos for the
following reasons. Toy et al. (1994) is not site-specific. Site-specific data indicate that Native
American fishing in the area is primarily focused on salmon, which are not caught
productively in the Hylebos. Furthermore, since salmon spend little of their life in the
waterways, the PCB concentrations in salmon would have little, if any. relationship to the
PCB concentrations in the Hylebos. (17)
Response: First, whether or not subsistence fishing currently occurs at the CB/NT
Site, it is a reasonable future use scenario and must be considered when evaluating
residual risks present at the Site (see responses to Comments 1.4.1 and 1.4.2). While
the Toy et al. (19%) study was not conducted for the tribes in the immediate area of
the CB/NT Site, it was conducted for two tribes that each gather fish from the Puget
Sound area. In this case, the collection of current site-specific data on fish
consumption would reveal only current fish consumption rates and an approximation
would still have to be used for the case of future exposure scenarios.
Additionally, while salmon may be the favored species, it is not the only species of fish
caught from the CB/NT Site. English sole were evaluated in the risk assessment as a
surrogate species to represent fish consumed from the CB/NT Site. Although salmon
spend less time in the CB/NT area, they have a higher lipid content than English sole,
13
-------
CB/NT PCB Responsiveness Summary
and they tend to bioaccumulate PCBs at a higher rate. Therefore, despite the salmon's
more transient residence in the CB/NT area, it is not certain that it will have
accumulated significantly lower levels of PCBs from the CB/NT sediment than the
English sole (see response to Comment 1.4.13).
Finally, although EPA used a reasonable maximum exposure scenario based on
potential exposures to Native Americans, this scenario is also meant to be protective of
subsistence fishers who are not Native American and who may be consuming fish other
than salmon.
1.4.8 The report fails to address subsistence consumption. Much of EPA Region X's position
to retain the previous risk assessment centered on protecting subsistence fishermen and their
families. However, EPA did not use an ingestion rate representative of subsistence fishing in
the ROD. (17, 29)
Response: The commentors are correct that the 1996 Weston report did not address
subsistence fishing. The report has been revised to include residual risk estimates for
Tribal fishers (one example of a population of subsistence fishers in the
Commencement Bay area) (Weston, 1997a). See also the response to Comment 1.4.4.
1.4.9 Use of an industrial future use scenario for the Hylebos Waterway is consistent in spirit
with the USEPA Memorandum entitled 'Land Use in the CERCLA Remedy Selection Process,"
by Elliott Laws (1995) which calls for realistic future land use assumptions and recognizes
industrial land use as a "reasonable assumption where a site is currently used for industrial
purposes." (17)
Response: Elliott Laws' May 25, 1995, memorandum entitled "Land Use in the
CERCLA Remedy Selection Process" states that EPA will use reasonably anticipated
future use scenarios in developing remedial action objectives for Superfund cleanups.
Seeking a way to apply this memorandum to the CB/NT Site, the commentor proposes
that the Hylebos Waterway should be viewed like an industrial upland site. Comparing
an upland to an aquatic site is inappropriate for several reasons, as discussed below.
First, EPA's mandate for protection of aquatic resources is entirely different than for
protection of upland properties. The Clean Water Act sets forth a national goal for
water quality which provides for the protection and propagation of fish, shellfish, and
wildlife and provides for recreation in and on the water (CWA §101 [a]). The Clean
Water Act, under Sections 302 and 303, also charges the States with responsibility for
establishing water quality standards for all waters of the state, including designation of
the uses for which a water body is to be protected. The Hylebos Waterway has been
designated in Washington's Water Quality Standards (Chapter 173-201A WAC) as a
Class B water body. Class B water bodies shall be protected for the following uses:
water supply (industrial and agricultural),
14
-------
CB/NT PCB Responsiveness Summary
stock watering,
fish and shellfish (salmonid migration, rearing and harvesting; other fish
migration, rearing, spawning, and harvesting; clam, oyster, and mussel rearing
and spawning),
wildlife habitat,
recreation (secondary contact recreation, sport fishing, boating, and aesthetic
enjoyment), and
commerce and navigation.
The State of Washington has also identified goals for marine sediments in the 1989
Puget Sound Water Quality Management Plan. This plan sets forth a conceptual
sediment quality goal of the absence of acute or chronic adverse effects on biological
resources or significant human health risk. This goal was incorporated into the
Sediment Quality Objectives in the CB/NT ROD and subsequently led to the
development of the State Sediment Management Standards.
The Hylebos Waterway serves several functions other than industrial use. Most
importantly, it serves as habitat for a variety of marine organisms. The habitat
functions of the CB/NT Site are a critical component of the CB/NT ROD and habitat
considerations must be incorporated into CB/NT cleanup plans.
Second, EPA can take measures to restrict access and exposure to contaminants at an
upland site, and can work with local zoning authorities to ensure that future land use
will remain industrial. EPA cannot restrict access or exposure to contaminants at the
Hylebos Waterway by marine organisms or humans. Recent development trends in the
Hylebos Waterway, including construction of the Chinook Marina by the Puyallup
Tribe (which is close to an area with high levels of PCBs), indicate that the public may
have increased access to the Hylebos Waterway in the future.
1.4.10 The analysis of human health impacts from PCBs is inadequate to protect tribe
members who consume large quantities offish, crab, and shellfish. Consumption rates of
shellfish should also be included with consumption rates offinfish species to reflect the total
consumption rate of seafood from the study area. (29, 12)
Response: As indicated above, risks for tribal fishers (one example of a population of
subsistence fishers in the Commencement Bay area) are presented in the Addendum
(Weston, 1997a).
Consumption of crab and shellfish, grouped into the shellfish category, was considered
in a previous risk assessment (Versar, 1985) conducted as part of the remedial
investigation for the CB/NT Site. This risk assessment estimated the risks associated
with contaminated shellfish to be approximately the same as the risks associated with
eating contaminated finfish. In the risk assessment presented in the CB/NT ROD,
consumption of English sole was used as a surrogate for consumption of other finfish
15
-------
CB/NT PCB Responsiveness Summary
and shellfish at-the CB/NT site. It was determined in the ROD that an evaluation based
on finfish consumption would be adequately protective of human health concerns
related to seafood consumption from the CB/NT site. This assumption was also used in
the reevaluation of residual risks associated with PCBs in the Addendum (Western,
1997a).
1.4.11 Consumption rates need to address exposure rates of sensitive populations which may
be consuming seafood from the Commencement Bay area. These populations are not well
served by the moderate recreational, seafood consumption rate of 12.3 grams per day. (12)
Response: EPA agrees. As discussed in responses to Comments 1.1.6 and 1.4.1, the
risk evaluation presented in the Addendum (Weston, 1997a) was based on a revised set
of exposure scenarios, including an upper-end tribal fishing scenario that was selected
to better represent a reasonable maximum exposure to PCBs at the CB/NT Site. A
consumption rate of 123 g/day, 69 percent of which was expected to come from the
CB/NT Site, was used to represent a high-end tribal fisher.
1.4.12 There are no data to indicate whether shellfish bioaccumulate more PCBs than finfish
and EPA's own risk assessments assumed the same level between the two. Thus, there will be
no impact on the risk assessment results from considering shellfish. (17)
Response: As indicated in the response to Comment 1.4.10, shellfish consumption
was not included in the risk assessment used to support the PCB SQO in the CB/NT
ROD (EPA, 1989a), nor in the subsequent evaluations conducted for the reevaluation
of the PCB cleanup level.
1.4.13 In using PCB concentrations in English sole to represent PCB concentrations in all
fish, EPA assumes that the ingestion offish is 100 percent English sole. Using PCB
concentrations in English sole to evaluate PCB concentrations in salmon and other fishes is
inappropriate. The Washington State Department of Health determined that English sole is
not an accurate indicator of human health risks. English sole account for less than one
percent of the fish consumed. Furthermore, EPA notes that using the predicted concentration
of PCBs in English sole to represent the concentration of PCBs in all fish is conservative
because: 1) English sole may not be representative of the majority offish caught and eaten;
and 2) English sole may bioaccumulate PCBs to a greater degree (a factor of 3 by EPA's own
estimate) than the more typically eaten fish. EPA's report did not consider the site-specific
data available that would portray a more realistic cross section offish species consumed. (17)
Response: Each individual who consumes seafood from the CB/NT Site likely eats a
slightly different type and number of the various types of seafood found in the area.
To estimate the contaminant concentrations in seafood, EPA sampled and analyzed
English sole, which occur in relatively large numbers in Commencement Bay. The use
of surrogate or representative species, such as English sole, to represent the various
16
-------
CD/NT PCB Responsiveness Summary
types of seafood that might be eaten is a common practice in estimating risk from fish
consumption at Superfund sites and other potentially contaminated areas.
English sole were chosen because they were cited in the remedial investigation report
(Tetra Tech, 1985) as a conservative indicator of the contaminant levels that would be
expected to occur in edible tissue of harvested fish species. This is in large part
because they live in close association with the sediments and would be expected to
accumulate contaminants from sediments. Recent reports by the Puget Sound Ambient
Monitoring Program (PSAMP) on PCB concentrations in fish caught throughout Puget
Sound also show that the average PCB concentrations found in English sole fillets are
at the high end of the PCB concentrations found in most fish species in the Sound
(O'Neill el a/., 1995). The PSAMP data also show that contaminant concentrations in
English sole muscle tissue also correlate well with contaminant concentrations in
nearby sediment samples.
Chinook and coho salmon had higher Puget Sound-wide average PCB concentrations
than English sole, most likely due to their higher lipid content. Although salmon is an
important food item for people in the Commencement Bay area, it was not considered
to be a good species to use for the Superfund risk assessment because PCB
concentrations in salmon tissue in general do not correlate well with site location, and
salmon have a relatively short residence time at the CB/NT site. Therefore, EPA
considers use of English sole as a surrogate species an appropriate approach for its
reassessment of residual risks associated with PCBs in sediments.
1.4.14 The most important fish species consumed in the area (by weight) are salmon and squid
- both of which do not spend considerable time in the Hylebos. (2)
Response: While these fish do not spend ail of their time at the CB/NT Site, they do
spend some of their time there. In that time, these fish are exposed to CB/NT
contaminants, including PCBs. The risk estimates presented in the Evaluation of
Residual Risks (Weston, 1996) and the Addendum (Weston, 1997a) accounted for
consumption of all finfish, including salmon. Salmon, with their higher lipid content,
(see response to Comment 1.4.13) likely accumulate some PCBs from the CB/NT Site
even with a low residence time. Therefore, it is reasonable to consider consumption of
these fish from the CB/NT vicinity as contributing to human health risks attributable to
CB/NT contaminants.
1.4.15 It is not reasonable to assume that people consume fish livers from CB/NT fish on a
regular basis. (22, 17)
Response: Although ingestion of fish livers was considered in the original CB/NT risk
assessment, it was not a scenario retained in the risk evaluation supporting the PCB
SQO presented in the 1989 ROD, nor was it considered in the Evaluation of Residual
Risks (Weston, 1996) or the Addendum (Weston, 1997a).
17
-------
CS/AT PCB Responsiveness Summary
1.4.16 EPA's PCB Report (Weston, 1996) does not dearly identify the population being
assessed. As a result, a hybrid of assumptions are used that apply to no particular population.
If cleanup decisions are to be based on risk assessment, site-specific exposure assumptions
need to be used that account for the type and amount offish actually consumed by the real
population for a reasonable duration. While subsistence anglers might indeed consume more
than the amounts listed in the EPA report, the anglers are not fishing in the Hylebos, they are
not consuming English sole, and they are not consuming them for more than 30 years. By
calculating one risk number for a "hybrid" exposure scenario, it is impossible to know who is
being assessed in EPA's assessments. (17)
Response: The risk scenario that was established in the ROD (EPA, 1989a) and
applied in the Evaluation of Residual Risks (Weston, 1996) was not a hybrid scenario;
it comprised a recreational fisher who consumed average amounts of fish. The
exposure assumptions used were consistent with site-specific information and risk
assessment guidelines and policies available at the time. Residual risks presented in the
Addendum (Weston, 1997a) were updated to reflect current EPA risk assessment
guidelines (including the values applied to exposure duration), and were calculated for
four different exposure scenarios (i.e., both average and high-end recreational and
tribal fishers).
1.4.17 Current EPA policy, as articulated by EPA Administrator Carol Browner and others
calls for risk assessments to present and discuss a range of values, to discuss the impact of
parameter choices on the risk numbers, to discuss the populations being assessed, and to the
extent possible, make reasonable assumptions for each population being assessed. (17)
Response: A range of parameter values and exposure scenarios was considered in the
risk assessment presented in the ROD (EPA, 1989a). The Evaluation of Residual Risks
(Weston, 1996) presented only one of the scenarios from the ROD (EPA, 1989a)
because that scenario was used to calculate residual risks associated with the PCB
cleanup level developed in the ROD. The uncertainty analysis presented in the
Evaluation of Residual Risks (Weston, 19%) explored the impacts of different
parameter choices on the residual risk estimates. Additionally, the Addendum
(Weston, 1997a) presents residual risk estimates for four different exposure scenarios:
average and high-end recreational fishers, and average and high-end tribal fishers. The
high-end tribal fisher was chosen to represent the reasonable maximum exposure
scenario; other risk estimates were provided to show the range of risks at the Site. As
discussed in response to comments in Section 1.1, up-to-date parameters were applied
to risk calculations for the Addendum (Weston, 1997a). Responses to previous
comments in Section 1.4 discuss the reasonableness of parameter values and exposure
scenarios evaluated in the risk evaluations (Weston, 1996, 1997a) conducted in support
of the reevaluation of the PCB cleanup level.
1.4.18 The report fails to identify cumulative health impacts from a variety of chemicals, and
fails to address impacts via a number of pathways. (29)
18
-------
CB/NT PCB Responsiveness Summary
Response: See_response to Comment 4.6.1.
1.5 Comments related to fish/sediment bioaccumulation data
1.5.1 Use of an empirical Biota-Sediment Accumulation Factor (BSAF) of 4. Ofor English
sole is inconsistent both with the site-specific results of 1.4 and with EPA risk assessment
guidance. (17)
Response: EPA did not use a BSAF of 4.0 in any of the risk calculations for the
CB/NT Site. The BSAF applied to risk calculations in the ROD, the Evaluation of
Residual Risks (Weston, 1996) and the Addendum (Western, 1997a) was 1.72, which is
a site-specific value based on empirical data gathered from Commencement Bay. EPA
believes that it is more appropriate to use this site-specific value rather than a
literature-based value of 4.0 or the value of 1.4 calculated by Gradient Corporation
(Gradient, 1995).
In 1995, Gradient Corporation calculated a site-specific BSAF of 1.4 based only on
data from the Hylebos Waterway. The value of 1.4 was based on 1984 fish tissue data
and 1994 sediment data. Gradient back-calculated PCB sediment concentrations using
the BSAF of 1.72 and the calculated fish value. EPA does not believe Gradient's value
of 1.4 is appropriate because: (1) it was calculated specifically for the Hylebos
Waterway, while EPA believes it is more appropriate to assess risks to people fishing
over the entire CB/NT Site, and (2) it is not appropriate to compare two data sets
collected 10 years apart to calculate a BSAF. Also, as shown by Gradient's own
calculations, the use of their BSAF would have very little impact on estimated PCB
concentrations in fish and estimated residual risks after cleanup.
7.5.2 Prior to changing the existing cleanup level for PCBs, it would be beneficial to collect
current fish tissue data and develop a more accurate sediment concentration-fish tissue
concentration model. (34, 2)
Response: EPA agrees that it would be beneficial to collect current fish tissue data, but
we believe that the environmental and economic costs of a six-month to one-year delay
in the sediment cleanup process that would be associated with implementing such a
program outweigh the benefits. EPA believes that the site-specific BSAF of 1.72
calculated for the CB/NT ROD (EPA, 1989a) provides adequate information about fish
accumulation of PCBs from sediment and that application of this value to residual risk
calculations provides adequate information for revaluation of the PCB cleanup level
for the CB/NT Site.
7.5.5 The Biota-Sediment Accumulation Factor (BASF) value of 1.72 should be rounded to
1.7. (12)
19
-------
CD/NT PCB Responsiveness Summary
Response: The BSAF of 1.72 is the actual value used in the ROD (EPA, 1989a) as
reported from site-specific studies. Rounding of this value to 1.7 would have minimal,
if any, impact on calculated fish tissue concentrations and resulting residual risk
estimates.
1.5.4 Identify what current science predicts to be the trigger level for bioaccumulation of
PCBs. (8)
Response: There is no federal or state standard that sets a "trigger level" for
bioaccumulation of PCBs, nor is there one PCB concentration that is commonly
accepted in the scientific community as a threshold for bioaccumulative effects. The
only PCB trigger level used in the Puget Sound area is the Puget Sound Dredged
Disposal Analysis (PSDDA) program's PCB bioaccumulation trigger level of 38 mg/kg
organic carbon (OC) (equivalent to approximately 900 Mg/kg dry weight at the CB/NT
Site). Dredgers with sediments containing 38 mg/kg OC must perform and pass
bioaccumulation tests if they wish to dispose of their dredged sediments a the PSDDA
open-water disposal site. Because there is no "trigger level" promulgated in federal or
state laws for use in sediment cleanups, EPA must use site-specific evaluations such as
the one contained in this Explanation of Significant Differences to determine an
appropriate site-specific standard for PCBs that is protective of human health.
1.6 Comments related to ecological risks
1.6.1 The SQO should remain at 150 ng/kg total PCBs, based on the following natural
resource information for sediment dwelling invertebrates and fishes. A Sediment Effects
Concentration (SEC) of 303.8 ng/kg (dry weight at 1 percent organic carbon) identified by
MacDonald (1994) represents concentrations of PCBs at which, more often than not, adverse
biological effects were demonstrate infield studies. In a similar evaluation of a variety of data
sets from marine and estuarine areas of North America, Long et al. (1995) calculated an
Effects Range—Median (ER—M) of 180 ng/kg (dry weight at 1 percent organic carbon). This
ER-M represented concentrations above which more than half of the reviewed studies indicated
adverse biological effects. (24)
Response: EPA believes that it is more appropriate to set a PCB cleanup level based on
a site-specific risk assessment, rather than relying on a database of studies which may
or may not have relevance to the conditions at the CB/NT site. As noted in the January
1996 Eco Update (EPA, 1996b), PCBs are one of four chemicals or chemical groups
that have a relatively low correlation between chemical concentration and effects level
and thus have low accuracy with respect to predicting effects. Moreover, the PCB
SQO of 300 Mg/kg is equivalent to the sediment effects concentration of 303 Mg/kg
calculated by MacDonald (1994) and is actually fairly close to the revised ER-M (i.e.,
there are few points in the effects database separating the values of 180 and 300
Mg/kg). The 300 Mg/kg PCB SQO value falls within a range identified by Long et al.
(1995) as a probable effects range.
20
-------
CD/NT PCB Responsiveness Summary
1.6.2 U.S. Fish and Wildlife (FWS) has assembled a series of predictive models to illustrate
the relationship between the PCB concentration in sediment and the PCB concentration that
would result in the eggs offish-eating birds (FWS, 1996a,b). These models are based on
biota-sediment accumulation factors to determine the concentration in fish that would result
from a given sediment concentration, and on biomagnification factors to translate this into an
egg concentration. These models predict that a cleanup level of 30 ng/kg would be protective
of a piscivorous bird since the predictive level ofPCBs in the egg is generally below the lowest
observable adverse effect level (LOAEL)for injury as determined by the Great Lakes data.
However, at 150 ng/kg (the current SQO level for PCBs in the ROD), the predictive levels of
PCBs in the bird egg increases and exceeds the injury threshold for the LOAEL. At 450 t*g/kg
the predicted concentration level of PCBs in the egg increases anywhere from 2 to 7 times
higher than the threshold level. In addition, the model also illustrates the potential for greater
impact to trophic levels higher in the food web, such as the federally listed peregrine falcon
and bald eagle. (35)
Response: EPA appreciates the information provided in the U.S. Fish & Wildlife
Service's (FWS's) letter regarding impacts of PCBs to birds, and has used it to the
extent possible in its revised evaluation of residual risks (see Appendix A to the Weston
Addendum [Weston, 1997a]). However, EPA modified the approach proposed by
FWS, to make it consistent with EPA risk assessment policies and guidelines.
For example, FWS assumed in their analysis that fish and birds would be exposed only
to the highest PCB concentration present in sediments after cleanup (450 Mg/kg).
EPA's view is that this approach tends to overestimate risks and believes it is more
appropriate to evaluate risks to wildlife by using the average post-cleanup sediment
PCB concentration (75 Mg/kg Site-wide or 124 Mg/kg in Hylebos Waterway). This
assumes that the fish and birds feed in more than one location at the site or a waterway
and are exposed to a range of PCB concentrations. Even in the case where piscivorous
birds may exhibit a high degree of site fidelity, their prey are potentially exposed to
contaminants over a broader area. Thus, average concentrations in sediment are an
appropriate estimate for fish and bird exposure. In addition, use of an arithmetic
average is protective because contaminants in sediment have a statistical distribution
such that an average value tends to overestimate the true midpoint of the
concentrations. EPA also believes that it is important to consider that PCB
concentrations will be reduced over time through natural recovery.
In addition, the assumptions used to develop the biomagnification factors (BMFs) for
eagles and peregrine falcons are highly conservative and assume that biomagnification
is multiplicative between trophic levels. While the BMF for fish-eating birds was
derived from empirical studies of herring gulls, the raptor BMF was extrapolated from
other species. Currently, the FWS is conducting studies in the Columbia River to
evaluate the health of bald eagle populations in this watershed. As part of this study,
selected PCB congeners were measured in eagle eggs and several species. Data
presented in an interim report (FWS, 1996c) suggest that the BMF may be within the
21
-------
CB/NT PCB Responsiveness Summary
same order of magnitude as calculated for the herring gull. (BMFs for selected
congeners ranged from about 10 to 90. When weighted based on the fraction
represented by the individual congener, the average BMP for both prey species
combined was about 40, as opposed to 265 to 845 used in FWS's model. [This also
assumes that 100 percent of their diet was derived from fish]).
In summary, EPA's analysis shows that cleanup to 450 A*g/kg PCBs and subsequent
natural recovery to at least 300 //g/kg PCBs is protective of fish-eating birds. EPA
does not believe there is sufficient information at this time to assess risks to eagles and
peregrine falcons.
1.6.3 It appears as though the proposed SQO value for PCBs is based on ecological
endpoints rather than human health endpoints. (12)
Response: The PCB sediment cleanup level is based on protection of human health
from ingestion of fish. Ecological endpoints were presented and discussed for
comparative purposes because of the concern raised by reviewers that wildlife may be
more sensitive to PCB sediment contamination than people.
1.6.4 The SQO should remain at 150 ng/kg total PCBs, based on the following natural
resource information for sediment dwelling invertebrates and fishes. Further information that
suggests concern over raising the SQO comes from the Hylebos Waterway fish injury studies
currently being conducted by the National Marine Fisheries Service's Northwest Fisheries
Science Center for the Natural Resource Trustees. This study shows that under current
conditions concentrations of contaminants in juvenile Chinook and chum salmon are
comparable to levels previously shown to be associated with biological injury and nearly one
third of adult English sole showed inhibited gonadal reproductive impairment, and up to half
of juveniles displayed precocious sexual maturation. (24)
Response: The recent work by the National Marine Fisheries Service (NMFS)
demonstrates that deleterious impacts to sediment-dwelling invertebrates and fishes are
occurring under current conditions (i.e., pre-cleanup) and further emphasizes the need
for cleanup to occur as quickly as possible. The NMFS report does not identify a
protective PCB sediment cleanup level. Therefore, EPA relied on risk assessment
methods to evaluate potential PCB cleanup levels for the CB/NT site.
2 Comments Regarding Risk Management Issues
2.1 Comments related to consideration of additional criteria to develop a PCB cleanup
level
2.1.1 Explain why, if current Washington State Sediment Management Standards have
already been used as a basis for cleanups at Eagle Harbor, Harbor Island and Ruston Way.
22
-------
CB/NT PCB Responsiveness Summary
those same standards are not being applied to the Hylebos Waterway cleanup. We request that
the EPA alter its course and accept the current Sediment Management Standards using
reasonable assumptions and up-to-date science in setting cleanup criteria for Commencement
Bay and the Hylebos Waterway. (15, 7)
Response: The Washington State Sediment Management Standards (SMS) are not
being used as an applicable, or relevant and appropriate requirement (ARAR) for the
CB/NT Site because the SMS standards were promulgated after the CB/NT ROD was
signed in 1989. After a ROD has been signed by EPA, the federal National
Contingency Plan (NCP) (40 CFR Part 300) provides that a State or Federal
environmental regulation is an ARAR only if it is necessary to ensure that the remedy
is protective of human health and the environment.
EPA evaluated the Washington State SMS requirements and determined that application
of the State SMS requirements was not necessary to ensure protectiveness. In addition,
it was not more stringent than the evaluation process EPA used to select a PCB cleanup
level. EPA's CB/NT PCB SQO of 300 Mg/kg and sediment remedial action level
(SRAL) of 450 Mg/kg fall within the range of numeric PCB standards which have been
set under the SMS to protect aquatic life (equivalent to 130 Mg/kg dry weight to 1,000
A*g/kg dry weight). Although the State SMS contains a narrative human health
standard, the State has not yet set a numeric standard for PCB concentrations in
sediments for protection of human health. Therefore, EPA has determined that it is not
necessary to consider the State SMS an ARAR to ensure the protectiveness of the
remedy.
2.1.2 Review of all EPA RODs nationwide involving PCBs reveals that the Commencement
Bay ROD PCB criterion of 150 ng/kg is among the lowest in the nation. Out of 53 sites
nationwide, the Hylebos is held to the lowest range of PCB criteria; for similar Superfund
sites, Commencement Bay ROD PCB criteria of 150 ng/kg are also the lowest. RODs for
comparable sites provide for a more realistic PCB cleanup criterion of 1,000 jj.g/kg. even
without consideration of the 1996 update of the PCB cancer slope factor. (17)
Response: Cleanup levels presented in Superfund RODs are site-specific and should
not be taken out of context from the other information presented in the ROD. Although
EPA strives for national consistency on the process used for determining cleanup levels
for Superfund sites, site-specific circumstances often lead to selection of different
cleanup standards. Some of the PCB cleanup levels at other Superfund sites may be
higher than at the CB/NT Site for several reasons: (1) different exposure scenarios
may be of concern at those sites, or (2) technological or feasibility issues may limit the
possible extent of cleanup.
One of the things that makes the CB/NT Site different than many other Superfund sites
is the relatively high potential for exposure to site contaminants by humans and
wildlife. The CB/NT Site is part of the usual and accustomed fishing grounds of the
23
-------
CB/NT PCB Responsiveness Summary
Puyallup Tribe,- It is therefore important that the cleanup level take into account
protection of Tribal consumers of fish, who generally consume more fish than
recreational fishermen. Protection of wildlife must also be considered. The CB/NT
area is used for foraging by shorebirds and other wildlife. Consideration of these
factors may lead to selection of a lower sediment PCB cleanup level at the CB/NT Site
than at other Superfund sites.
In addition, cleanup at several Superfund sites with PCB contamination in sediments is
limited by practical constraints that are not present at the CB/NT Site. For example, at
many areas in the East Coast and Great Lakes, PCB contamination is so widespread
that 1,000 Mg/kg is the lowest practical PCB cleanup level that can be achieved. At
some sites, PCB concentrations are so high (thousands of parts per million [ppm]) that
extensive engineering controls are necessary to control migration of PCBs into the
water and air during dredging. Because the highest PCB concentrations at the CB/NT
Site are on the order of 25 mg/kg, potential releases of PCBs during dredging can be
controlled much more easily.
2.1.3 EPA should draw on other information such as cost-benefit evaluations to support the
selection of a realistic cleanup level for PCBs that will allow for the cleanup to get started.
(17)
Response: EPA considers cost-effectiveness as one of the nine evaluation criteria used
to make Superfund cleanup decisions.
2.1.4 EPA Region X should increase the cleanup goal for PCBs in sediments on the Hylebos
Waterway from the current SQO to 150 ng/kg. dry weight to at least 600 pg/kg, dry weight.
This change will decrease the cost of cleanup by approximately 35 million dollars based on
volumes and cost presented in Tables ES-1 of the Evaluation of Residual Risks (Weston, 1996),
while causing an insignificant increase in the potential risk to a very limited, hypothetical
population. Using the numbers from your own report, the risk to a potential fisherman who
spends his whole lifetime eating bottom fish daily from Commencement Bay would increase
from a 5.3 chance in 100,000 to a 7.3 chance in 100,000 of contracting cancer (Table 3.4 of
the Evaluation of Residual Risks.) (14)
Response: EPA considered each of the nine evaluation criteria, including cost, in its
revaluation of the PCB cleanup level for the CB/NT Site. See responses to comments
in Section 3.2 for discussion of particular cleanup levels.
2.2 Other Comments
2.2.7 Have the sources of PCB contamination been eliminated? (8)
Response: PCBs have not been manufactured since 1977. Although surplus stocks of
PCBs were allowed to be used in the short-term, PCBs have been replaced in all their
24
-------
CB/NTPCB Responsiveness Summary
applications over the last 20 years. Even so, PCBs may be present at facilities in the
CB/NT area if, for example, PCB-containing materials were spilled or buried in the
ground.
In order to control ongoing sources of contamination to the CB/NT Site, the
Washington State Department of Ecology (Ecology) has inspected or investigated
virtually every industry in the Commencement Bay tideflats area. They have identified
facilities they believe to be ongoing sources of contamination to the CB/NT Superfund
Site, and have initiated cleanups, required permits or required implementation of best
management practices to control pollution at most of them. PCBs have been found in
upland soils at a few facilities, primarily along the Hylebos Waterway. Ecology has
required cleanups at these facilities. Ecology hopes to complete all necessary source
control activities by 1998. See response to Comment 5.4.1 for further discussion of
source control activities at the CB/NT Site.
2.2.2 To address questions regarding the risk evaluation and revaluation of the SQOfor
PCBs, CHB recommends putting together a discussion forum... .A neutral facilitator should
moderate the discussion. (8)
Response: Subsequent to receipt of this letter, EPA and Citizens for a Healthy Bay
(CHB) discussed the possibility of a discussion forum with several potential
participants. EPA and CHB decided that there was not sufficient interest among
potential participants to hold a discussion forum. Instead, EPA initiated several
activities to solicit public input and to keep the public informed about the PCB
revaluation. These efforts are described in the introduction to this responsiveness
summary.
2.2.5 This evaluation is incomplete because it analyzes potential recommendation scenarios
as if PCBs were the only contaminant of concern in Commencement Bay. (24)
Response: As stated in the CB/NT ROD, PCBs were the only contaminant of concern
at the CB/NT Site for which ecological cleanup goals would not be sufficiently
protective of human health. For this reason, PCBs have been segregated from other
contaminants of concern. (See Section 7.1 of the CB/NT ROD, where discussion of
human health risks is presented.) Further discussion of this issue is presented in
response to comments in Section 4.6.
2.2.4 The PCB SQO was established by a human health risk assessment, which was thought
to be a more sensitive endpoint than an ecological endpoint at the time of the Record of
Decision. While the dose response curve for human cancer posed by PCB's cancer risk posed
by PCB to humans may be lower than previously thought, new studies are showing greater
effects to fish and wildlife than was previously known. (24)
25
-------
CB/NT PCB Responsiveness Summary
Response: EPA agrees that it would be beneficial to check the assumption in the ROD
that a PCB sediment cleanup level based on human health would also protect wildlife.
An evaluation of the threats to wildlife from exposure to contaminated prey or sediment
was performed by EPA as part of the Addendum (Weston, 1997a). The results were
used to ensure that the selection of a PCB cleanup level based on protection of human
health would also be protective of wildlife.
2.2.5 It has recently been discovered that the toxicity ofPCBs is not related to the degree of
chlorination. (17)
Response: EPA disagrees. In the support document that summarizes the calculation of
the new cancer slope factors for PCBs (EPA, 1996a), different slope factors are to be
used depending on both route of exposure and percent of chlorination. For example,
the lowest upper-bound slope factor of 0.07 per mg/kg-day is to be used only when
analyses verify that congeners with more than four chlorines comprise less than
1/2 percent of total PCBs. Route of exposure is considered because of the expected
partitioning of more highly chlorinated congeners in certain media (e.g., fish,
sediment, soils). As recommended by this document, the highest slope factor of 2.0
was used in the Addendum (Weston, 1997a) since the route of exposure assumed for all
exposure scenarios was fish consumption.
2.2.6 The Puyallup Tribe of Indians has not been recognized as a sovereign Indian nation
with governmental and proprietary interests in this matter, not limited to its treaty resources,
which have been seriously impacted by the release of hazardous substances including PCBs
into Commencement Bay. EPA's risk assessment fails to acknowledge the well documented
tribe's concerns; does not meet the legal requirements that provide for permanent cleanup of
the Hylebos Waterway; ignores EPA's obligations under its Environmental Justice mandate
and the protection of human health and the environment as demanded by Superfund. (29)
Response: EPA agrees that the Evaluation of Residual Risks (Weston, 1996) did not
fully address some of the Tribe's concerns. EPA's Environmental Justice guidelines
require consideration of populations who may have disproportionately higher exposures
to contaminants in Superfund risk assessments and in Superfund cleanup decisions.
For these reasons, and because EPA believes it represents a realistic future use scenario
for the site, EPA updated the human health risk assessment in the Addendum (Weston,
1997a) to evaluate residual risks after cleanup using high-end tribal fishing as the
reasonable maximum exposure scenario.
As to legal requirements for permanent cleanup, the Superfund law does state a
preference for treatment to permanently and significantly reduce the volume, toxicity,
and mobility of hazardous substances, pollutants, and contaminants. However, EPA
determined in the 1989 CB/NT ROD that the nature of the contamination at the CB/NT
site (i.e., widespread, low-level contamination) limits the feasibility of treatment
options. Confinement was therefore selected as the appropriate remedial action for
26
-------
CB/NT PCS Responsiveness Summary
CB/NT sediments with higher contaminant concentrations, and natural recovery for
sediments with moderate to low contaminant concentrations. EPA's reevaluation of the
PCB cleanup level does not change this fundamental aspect of the remedy EPA selected
in 1989.
Part II—Responses to Comments Received Subsequent to Issuance of the Draft ESD
NOTE: These comments relate primarily to two documents: the Public Review Draft
Explanation of Significant Differences (EPA, 1997a), and the Addendum to the Evaluation of
Residual Risks (Weston, 1997a), which included an ecological evaluation as art appendix.
3 Comments Related to the Proposed PCB Cleanup Level
3.1 Comments in support of the proposed PCB cleanup level
3.1.1 We support the EPA's proposal to modify the PCB cleanup level from 150 i*g/kg to be
achieved within ten years to 450 ng/kg to be achieved immediately after cleanup. (5, 10, 19,
22. 23. 20)
Response: Comment acknowledged.
3.1.2 While we believe that the 450 /*g/kg level proposed by EPA is significantly below PCB
cleanup standards set at other sediment Superfund sites, the level is more realistic and
reflective of current science than the 150 ^g/kg level contained in the 1989 Record of
Decision. We also believe that any level lower than 450 f*g/kg would seriously compromise
the ability of the parties at the various waterways to take advantage of combined disposal
opportunities. The volume of sediment which may have to be removed from the larger
waterways in order to meet an unrealistic level of 150 t*g/kg could far exceed the available
capacity at the various disposal sites currently being considered by the ponies in conjunction
with the Sediment Disposal Site Forum, thus eliminating the possibility of a combined disposal
site and/or necessitating additional sites to accommodate excess volumes. (23)
Response: Foremost among all considerations in making a cleanup decision, EPA
focused on establishing a PCB cleanup level that would be protective of human health
and the environment. As discussed in response to comments in Section 5.2, EPA
establishes cleanup levels on a site-specific basis. As discussed in the ESD, in
assessing the implementability, short-term effectiveness, and cost of various alternative
cleanup levels, EPA considered the volumes of sediment to be removed to achieve the
potential cleanup levels. EPA also considered the disposal options associated with
containing these volumes of sediment in its reevaluation of PCB cleanup levels at the
CB/NT Site.
3.2 Comments recommending a higher PCB cleanup level
27
-------
CB/NT PCB Responsiveness Summary
3.2.1 The 150 ng/kg TCB number should be modified to be consistent with all other similar
Superfund sediment sites—recent PCB cleanup standards have been set at 1,000 ng/kg. The
450 tig/kg PCB cleanup standard would be the most rigorous cleanup standard ever required
of any similar Superfund site in the U.S. Even at 600 fj.g/kg, the standard would be the
toughest in the nation for such sites. (6, 18)
Response: As discussed in response to Comment 2.1.2, EPA's cleanup decisions are
made on a site-specific basis and must be protective of human health and the
environment at the given site. EPA selected the 450 Mg/kg PCB cleanup standard
because it is protective of human health and the environment, and provides the best
balance of cost, long-term effectiveness and short-term effectiveness. Cleaning up to
600 Mg/kg PCBs or higher will be less protective of human health and the environment,
will provide less long-term effectiveness, and will result in only a small decrease in the
cost of the cleanup. For further discussion, see response to comment 3.2.2.
3.2.2 The human health risk estimates based on a PCB cleanup level of 600 ng/kg range
from 2.75 x 1O> to2.1x 104 (Weston, 1997a). Table 2 in the Addendum (Weston, 1997a)
shows that there is little difference in the 450 ng/kg and 600 ng/kg cleanup levels based on
residual risk estimates. EPA's explanation for the acceptability of 450 ng/kg as a cleanup
level (Weston, 1997a, p. 11) also applies to higher cleanup levels provided that the residual
risks do not substantially exceed 1 x 104. The highest estimated residual risk (High-End
Recreational Fisher), based on a PCBs cleanup level of 600 ng/kg, is 2.1 x 1O4. This
estimated residual risk is not substantially greater than 1 x 1O4. Given EPA's rationale for the
acceptability of 450 pg/kg, a PCBs cleanup level of 600 pg/kg would cut down on cost from
the proposed 450 fig/kg cleanup level, and would still meet the criterion for protectiveness.
(I, 6, 18, 9, 17, 19, 23, 25, 30. 13, 17)
Response: Because EPA assumes that there is no threshold below which carcinogenic
effects of PCBs will not occur, EPA strives to minimize this risk to the extent possible.
In selecting the 450 /zg/kg PCB SRAL, EPA considered other potential cleanup levels,
including a 600 Mg/kg cleanup level. Potential cleanup levels from 300 A*g/kg to 600
Mg/kg cleanup level fall at the high end of a range of values which EPA considers
protective of human health and the environment, based on a tribal fishing scenario, and
which offer similar benefits in protectiveness.
EPA's goal in the Superfund program is to select remedies which reduce human health
cancer risks associated with exposure to site contaminants to 10^* to 10"6 or below.
EPA chooses cleanup levels that will achieve reduction of risks to within this range,
accounting for the uncertainties associated with calculated risk estimates. EPA selected
the 450 Mg/kg PCB SRAL because it is protective of human health and the
environment, and provides the best balance of cost, long-term effectiveness and short-
term effectiveness. Because the cost difference between the 450 Mg/kg cleanup level
and the 300 Mg/kg cleanup level is very large (about $13 million, a 72 percent increase
in cost), EPA determined that the small environmental benefits to be achieved by going
28
-------
CB/NT PCB Responsiveness Summary
to the lower cleanup level were outweighed by the cost. However, EPA has added a
requirement that PCB concentrations be reduced to 300 ,ug/kg within 10 years of
completion of the remedy. Adding the 300 Atg/kg PCB SQO provides additional
protectiveness and adds only a small cost to the remedy (for natural recovery modeling
and monitoring, and potentially for additional remedial action if models have
incorrectly predicted natural recovery rates).
EPA also considered whether a 600 Mg/kg cleanup level was appropriate. The cost
difference between achieving a 450 jig/kg cleanup level versus a 600 Mg/kg cleanup
level is about $2.5 million, or a 16 percent increase in the cost of the Hylebos
Waterway cleanup. EPA determined that, as with the 300 Mg/kg SQO, the relatively
small increase in cost was justified to achieve an incremental increase in environmental
benefit, especially considering the uncertainties associated with estimates of the toxicity
of PCBs (see response to Comment 4.3.3).
3.2. 3 Reducing the volume to be dredged not only is more cost effective, but also is less
disruptive to existing aquatic life. Also, this reduces the volume of the disposal site material
which must be managed long term. Therefore, since both the 450 ng/kg and the 600 pg/kg
cleanup levels are protective of human health and the environment, 600 t*g/kg is a better
choice. (9)
Response: As part of its revaluation of the PCB cleanup level, EPA considered cost,
short-term effectiveness, and long-term effectiveness. Under the short-term
effectiveness criterion, EPA considered the potential detrimental effects to aquatic life
associated with cleanup to a range of different potential cleanup levels. EPA concluded
that cleanup of sediments to 450 Mg/kg provides sufficient additional protectiveness and
long-term effectiveness over cleanup to 600 Mg/kg to justify a small increase in cost
and disruption to aquatic life during dredging.
3.2.4 There is a tremendous cost savings if the cleanup level is set at 450 to 600
because there will be a reduction in the volume of sediment that must be remediated. While
cost savings alone should not determine human health risks, where the scientific data supports
a modification which will result in a savings of millions of dollars, we believe a compelling
case is made for the modification. In fact, the arithmetic mean baywide residual concentration
of PCBs, following a 600 ng/kg cleanup, is 82 ng/kg compared to 75 ng/kgfor a PCB
cleanup of 450 ng/kg, an insignificant difference. That's only a seven pan per billion cfiange
in PCB concentration for a cost of about 3 million dollars. (19, 17, 18)
Response: EPA considered the difference in cleanup volumes necessary to achieve
each of the potential PCB cleanup levels in its reevaluation. As discussed in response
to Comment 3.2.2, EPA considered the costs and benefits of selecting a 450 Mg/kg
versus a 600 Mg/kg PCB cleanup level for the CB/NT site and determined that
450 Mg/kg offered better protection of human health and the environment with only a
small incremental increase in cost.
29
-------
CB/NT PCB Responsiveness Summary
3.2.5 The difference in residual PCB sediment concentrations between cleaning up to 450
vg/kg and cleanup up to 1.000 ^g/kg is insignificant, 1,000 f^g/kg is a protective level, and it
would be more cost-effective to cleanup to l,000^g/kg. (6, 18, 7, 27)
Response: While EPA did not include a potential 1,000 Mg/kg PCB cleanup level in
its analysis, EPA did consider a 900 Mg/kg PCB cleanup level, which would give
roughly similar results as a 1,000 Mg/kg cleanup level. Because there are few areas of
Hylebos Waterway sediments contaminated with PCBs at concentrations between 450
and 900 Mg/kg, the cost difference to go from a 450 to a 900 Mg/kg PCB cleanup level
is not large, about $4 million. This represents a 29 percent cost increase to achieve a
450 Mg/kg cleanup level as opposed to a 900 Mg/kg cleanup level. EPA determined
that in this case the relatively small increase in cost was justified to achieve an
incremental increase in environmental benefit, especially considering the uncertainties
associated with risk estimates for PCBs. See EPA's response to Comment 3.2.2 for
further discussion.
3.3 Comments recommending a lower PCB cleanup level
3.3.1 The proposed 450 fj.g/kg cleanup level overprioriliz.es cost issues and would not be
protective of human health and the environment. (28, 11, 29, 35, 36, 9, 9)
Response: EPA is required under the Comprehensive Environmental Response,
Compensation, and Liability Act (CERCLA; Superfund) to select remedies for
Superfund cleanups that are protective of human health and the environment. EPA
must also select cost-effective response actions for Superfund cleanups. A remedial
alternative is considered "cost-effective" if its costs are proportional to its overall
effectiveness. Overall effectiveness is determined by evaluating its short- and long-
term effectiveness. In this case, the alternative PCB cleanup levels evaluated showed
only a small difference in short-term and long-term effectiveness and environmental
benefit, but a large difference in cost. Therefore, cost became a significant factor in
EPA's decision. EPA believes that its decision to require a 450 Mg/kg SRAL, to be
achieved at the time of dredging, and a 300 Mg/kg SQO, to be achieved within 10 years
of completion of the cleanup, is protective of human health and the environment, and
meets the CERCLA criterion for cost-effectiveness.
3.3.2 EPA should be more protective, not less protective, of human health and the
environment, and therefore, should not raise the cleanup level. (9)
Response: EPA agrees that it should be as protective as possible in selecting cleanup
levels for Superfund response actions. However, EPA is also required under CERCLA
to select cost-effective cleanups, and to ensure that significant environmental protection
is achieved for every dollar spent for cleanup. In this case, new information since the
1989 ROD indicates that remediation to the original cleanup level will involve a
30
-------
CB/NT PCB Responsiveness Summary
substantially larger amount of sediments than originally anticipated, along with
significantly increased costs. In addition, new information about the risks associated
with PCBs prompted EPA to reevaluate the original PCB cleanup level. After
considering this new information, EPA determined that remediation to an alternative
cleanup level would be considerably less expensive, with only a small increase in risk
when compared to the cleanup level in the 1989 ROD. The new cleanup levels are,
however, still within EPA's acceptable risk range for Superfund cleanups.
5.3.3 EPA should cleanup PCBs in Commencement Bay sediment to the area background
PCB sediment level. (9)
Response: CERCLA requires that Superfund cleanups be protective of human health
and the environment. In some circumstances, it is necessary to achieve background
concentrations to be protective. In this case, EPA believes protectiveness can be
achieved at concentrations above background. Estimates of area background PCB
concentrations in sediments range from 20 to 50 Mg/kg in Puget Sound. The average
PCB concentration in CB/NT sediments after cleanup will be 75 Mg/kg immediately
after cleanup and will be reduced to 63 Mg/kg or lower within 10 years after the
cleanup. PCB concentrations at the CB/NT site will be somewhat higher than
background after cleanup, but will still be protective.
3.3.4 EPA's Fact Sheet states that the proposed change in the PCB cleanup levels would have
a significant effect onfy in the Hylebos Waterway cleanup. The Hylebos Waterway completes
the drainage for Hylebos Creek, which connects the Bay with sensitive wetlands and uplands
throughout the watershed. As such the Hylebos Waterway should be treated with greater
concern than the transportation channels formed by other waterways, not sacrificed to the
"average" health of the Bay. (28)
Response: EPA agrees that the ecological health of the Hylebos Waterway should be
preserved and protected. By stating that the proposed change in the PCB cleanup level
would have a significant effect only on the Hylebos Waterway, EPA meant that it
focused the cost analysis on the Hylebos Waterway, because the PCB cleanup level has
little effect on the cost of cleanup of other Commencement Bay waterways. EPA's risk
analysis focused on CB/NT Site as a whole, because EPA felt it was reasonable to
assume that most fish, birds, and humans would utilize more than one area of the Site.
However, EPA includes in its analysis an assessment of the risks to humans and
wildlife that reside or consume fish only from Hylebos Waterway, to ensure that by
selecting a cleanup level that was protecting the Site as a whole, a cleanup level was
not selected that would have a disproportionately large impact to individuals who
preferentially utilize the Hylebos Waterway.
31
-------
CB/NT PCB Responsiveness Summary
4 Comments Related to the Risk Evaluations
4.1 Comments relating to the residual PCB sediment concentration
4. J.I Since dredging for cleanup would likely reach native sediments, using the background
value as the replacement value in the calculation of residual sediment concentrations is likely
an overestimate because native sediments should have a PCB concentration of zero. It would
be more appropriate to use zero, or half the PCB detection limit, or even the PCB detection
limit as the replacement value. (9)
Response: Because PCBs have been in use for so long, and because they are so
pervasive in the environment, native sediment PCB concentrations at the CB/NT Site
are not expected to be zero. EPA used a 20 Mg/kg PCB concentration for native
sediments, which is reflective of background concentrations in areas with no local
sources of PCBs. It is also in the range of method detection limits for PCBs in
sediments (20 to 40 Mg/kg) under the analytical methods used for pre-design sampling.
4.1.2 The use of the arithmetic mean simply because it is EPA policy is not appropriate if the
distribution of PCB concentrations is lognormal. Furthermore, using the arithmetic mean
PCB concentration to account for the fact that the more contaminated sediments tend to be
located near shore in the shallow subtidal and intertidal areas where many fish species or life
stages tend to be during flood tides, is flawed. If this is where the most contaminated
sediments are located, then this is where the remediation will occur. Therefore, the residual
concentrations in this area should be 0 (zero), or at least background. The remaining
concentrations will still be lognormalfy distributed and therefore it is appropriate to use a
geometric mean as the basis of the residual concentration. (Whether you use the geometric
mean or upper 90 or 95 percentile of the geometric mean is another issue.) (1, 9)
Response: The arithmetic mean was chosen to represent residual sediment
concentrations because it was consistent with current EPA policy and because it was
thought to be appropriately protective. EPA guidance (EPA, 1992) states that the
arithmetic mean is appropriate because it best represents the cumulative intake that
would result from long-term contact with site contaminants regardless of the pattern of
daily exposures over time or the type of statistical distribution that might best describe
the sampling data. In fact, EPA guidance recommends that the 95 percent upper
confidence limit of the arithmetic mean be used in risk assessments, because in most
environmental sampling, the sample size is too small to accurately reflect the central
tendency of the data. In this case, because of the large sample size (over 200 samples
for the Hylebos Waterway and over 400 samples for the CB/NT Site as a whole), EPA
determined that an arithmetic mean, rather than a 95 percent upper confidence level, is
an appropriate concentration term to use in the residual risk assessment.
At the CB/NT Site, while the distribution of PCBs within the sediment of individual
waterways tends to form a lognormal distribution, as noted by the commentor,
32
-------
CB/NT PCB Responsiveness Summary
sediments with higher PCB concentrations tend to be located along the banks of the
waterways where the fish are likely to spend a disproportionate amount of time. For
this reason, the arithmetic mean was chosen as a more protective value. While it is
true that the more contaminated sediment will be remediated, sediments with PCB
concentrations below the selected cleanup level will remain after cleanup. Because
there are relatively higher concentrations of PCBs in the bank areas, residual PCB
concentrations after cleanup will be somewhat higher nearer to the banks and in habitat
in which the fish may spend more time. Therefore, use of the geometric mean may
underestimate actual PCB concentrations to which fish may be exposed. See also
EPA's response to Comment 1.2.2.
4.1.3 It is not clear why EPA sometimes averages residual PCB sediment concentrations
based on waterways and sometimes across the entire bay. (8)
Response: Risks were calculated for the overall CB/NT Site and for two individual
waterways: the Hylebos and the Thea Foss waterways. While EPA did consider both
the CB/NT-wide and the waterway-specific residual risk estimates in its revaluation of
the PCB cleanup level for the CB/NT Site, the CB/NT-wide estimates were used as the
primary factor in making risk-based decisions.
4.1.4 Risks are not associated only with average PCB sediment concentrations, but also with
concentrated "hot spots" throughout the bay. Averaging residual sediment concentrations
underestimates the impact of localized or hot spot exposures. Exposure point concentrations
should not be averaged on a bay-wide or even waterway-wide basis. There is preferential use
of habitat and varying site fidelity among species. Averaging residual sediment concentrations
underestimates the potential for adverse impacts to fish and wildlife species that exhibit some
degree of site fidelity in feeding behavior. (35, 28, 8, 24)
Response: Risks were evaluated for potential exposures following sediment cleanup,
not for current conditions. The purpose of the cleanup is to remove PCB "hot spots."
After the hot spot cleanup, the remaining PCBs will be at lower concentrations and
more evenly distributed in the bay. For this and the reasons discussed in responses to
Comments 1.2.2 and 4.1.2, a representative average PCB sediment concentration is
appropriate for estimating concentrations of PCBs in fish to which people will be
exposed after cleanup. While some fish may be exposed to higher concentrations of
PCBs, other fish may be exposed to lower concentrations. Furthermore, fish are
mobile, and are exposed to varying concentrations of PCBs as they move throughout or
within different areas of the waterways.
The arithmetic mean is also an appropriate representation of residual sediment
concentrations with respect to evaluating risks from ecological exposures. In
ecological risk assessments, risks to populations and communities of receptors are
evaluated, not risks to individual receptors (except in the case of threatened or
endangered species). Therefore, while some fish may be exposed to slightly higher
33
-------
CD/NT PCB Responsiveness Summary
concentrations o_f PCBs due to habitat or site fidelity, other fish will be exposed to
lower than average concentrations. As stated above, the arithmetic mean was chosen
over the geometric mean to ensure protectiveness while allowing for the possibility that
fish might be exposed to elevated concentrations. The arithmetic mean is also
appropriate for higher order receptors (e.g., fish-eating birds) for the same reasons that
apply to it being protective of human health.
Furthermore, EPA does not believe that using a maximum single-point measurement
(as represented by a grab sample) is appropriate to evaluate the risks to wildlife,
because very few species (with the exception of sessile benthos) will feed repeatedly at
a single point. Most species that exhibit a high degree of site fidelity prey on species
that feed over an area.
4.1.5 The averaging of the cleanup levels in sediment should take into account habitat-specific
or habitat-weighted assessment in denving a protective sediment concentration. (24)
Response: Area-weighting by habitat type is an approach that could be used to
calculate residual PCB sediment concentrations. The highest sediment PCB
concentrations tend to occur in the intertidal banks and shallow nearshore areas because
of proximity to source and sediment transport mechanisms. These areas are also most
often used by sensitive taxa or life stages of species of concern. Most of these areas
with elevated concentrations are already identified as requiring remediation in both the
Thea Foss and Hylebos waterways. The remaining concentration of PCBs has been
represented in the reevaluation of risks by the arithmetic mean, which tends to
overestimate the actual concentration of these chemicals and is thus conservative with
respect to protection of natural resources. See also EPA's response to Comments 1.2.2
and 4.1.2.
4.1.6 The average sediment PCB concentrations used in the risk calculations over-predict the
post-cleanup concentration of PCBs because cleanup of other chemicals will remove additional
PCBs (ielow the target cleanup level) from the Hylebos Waterway. In addition, ongoing
natural recovery will further reduce the PCB concentration in sediment. (17)
Response: EPA agrees thai cleanup for other chemicals and natural recovery will
reduce contaminant concentrations beyond the averages presented in the 1997 Weston
report. In the final decision, EPA has incorporated natural recovery by adding a 10-
year PCB SQO of 300 ^g/kg. EPA agrees that there will be additional reductions of
PCB concentrations through incidental cleanup of PCBs present in cleanup areas for
other chemicals. This factor was not included in the 1997 Weston risk assessment
because it cannot be quantified until cleanup areas for other chemicals are finalized.
Omitting this factor from the quantitative analysis adds an element of conservatism in
the calculation of residual risks, which EPA believes is appropriate.
34
-------
CB/NT PCB Responsiveness Summary
4.1.7 How can EPA use a "background" concentration of 30 ng/kg PCBs in this process ?
PCBs are man-made compounds. By definition, there is not a natural background
concentration for PCBs and an area background of 30 ng/kgfor (southern?) Puget Sound is
outrageous! (It is unclear what region of the Sound this background level is defined for since
Commencement Bay is in the Central Basin and Can Inlet, whereas the 'background'' samples
were collected south of the Narrows and therefore in South Sound.) By accepting a
background concentration this high it appears that a risk to human health of approximately
1&S exists in Puget Sound from PCBs alone! Is this acceptable to EPA? Is EPA planning to
remediate the background sample collection site at Carr Inlet or the entire Puget Sound?
Please explain the logic that went into accepting this background concentration for PCBs. (9)
Response: While PCBs do not naturally occur in the environment, they are widely
distributed throughout world, including Puget Sound. PCBs are no longer produced in
the United States, but they are persistent chemicals that have built up in the
environment over time. PCBs are widely distributed because of sediment and airborne
transport. For this reason, PCBs often appear in non-industrial areas without a direct
point source of contamination. These PCBs form what is referred to as "background
concentrations." EPA used background concentrations as a basis for comparison to
potential PCB concentrations after cleanup, and did not base its cleanup levels on
background.
4.1.8 For what surficial area and depth of sediment are the arithmetic means in Tables 4, 5,
andA-I representative? (I)
Response: The arithmetic means presented in the Addendum (Weston, 1997a) and the
ESD (EPA, 1997a) are based on the surface areas of the entire Hylebos Waterway, the
entire Thea Foss/Wheeler-Osgood waterways, and the overall CB/NT Site, which
includes all waterways and the adjacent area of Commencement Bay out to the 60-foot
bathymetric contour. (See Figure 1 in the ESD.) Surface samples represent depths up
to 0.3 foot. Surface sample concentrations will be used to evaluate the need for
remediation of a given area. Remediation depths are anticipated to be determined by
the depth to reach native sediment. Based on existing data, this depth is anticipated to
average approximately 7 feet in the Hylebos and roughly 6 feet in the Thea Foss
Waterway. (See also response to Comment 4.1.10.)
4.1.9 Is not the great majority of the PCB exposure to fish from the surface (top layer) of the
sediments and the resident benthic organisms? If yes, what is the purpose of targeting the
cleanup level "at all points," i.e., at all depths? (I)
Response: EPA agrees that the primary exposure to contaminated sediments by aquatic
organisms occurs at the surface layer. Cleanup decisions will be based on PCB
concentrations in surface sediment, not concentrations at all depths. At the CB/NT
Site, it has been found that contaminants generally reside in the unconsolidated
sediments which lie above the native sediments (i.e., sediments which have
35
-------
CB/NT PCB Responsiveness Summary
accumulated since the waterway was last dredged), and that native sediments contain
only low concentrations of contaminants. Therefore, dredging depths were determined
based on the depths necessary to reach native sediment. Average residual PCB
sediment concentrations were calculated from surface PCB sediment concentrations
expected to remain at each sampling station after cleanup.
4.1.10 What PCB Arodors are present? Nowhere in the ROD or ESD are PCB Aroclors
identified. (1)
Response: The most prevalent aroclors present at the CB/NT Site are Aroclor 1254
and Aroclor 1260. Aroclor 1242 was also detected.
4.1.11 Many of the people who fish and harvest in the bay consume greater than average
quantities offish and shellfish than the general population. Therefore, use of average residual
sediment concentrations may underestimate the exposure to PCBsfrom the bay. (35, 28, 8)
Response: EPA acknowledges that some individuals who fish at the CB/NT Site
consume greater than average amounts of fish. Therefore, EPA is using a high-end
tribal fishing scenario for decision-making purposes. Residual sediment concentration
is only one parameter used to calculate residual risks. Examples of other parameters
include the amount of fish consumed, the fraction of fish consumed that comes from
the CB/NT area, and the duration of exposure. As discussed in response to Comment
4.2.2, EPA develops a reasonable maximum exposure scenarios for Superfund risk
assessments as a mix of average and upper end parameter values that best represents
some of the most highly exposed individuals at the Site. For this reason, EPA felt that
use of an average residual sediment concentration was appropriate and protective.
(Also, see responses to Comments 1.2.2 and 4.1.2.)
4.2 Comments Relating to human exposure factors
4.2.1 Subsistence fishing is a main source of food for many residents in the area and can be
witnessed on any weekend at the floating dock in the Thea Foss Waterway. Many of the
fishing community are low income and rety on fish and shellfish for a substantial part of their
diet. Those for whom subsistence fishing is a main source of food are much more vulnerable.
(3. 8).
Response: EPA recognizes that fish from the CB/NT Site may currently or in the
future provide a substantial part of some individuals' diets, even though local health
authorities have posted signs warning against consumption of fish from the CB/NT
Site. EPA has no information on the number of subsistence fishers currently utilizing
the CB/NT Site or on their fish consumption rates. Instead, EPA used information
from recent studies (Toy et al., 1996) of fish consumption by Puget Sound Native
American tribes to develop a "high-end tribal fisher" (i.e., a tribal fisher who
consumes a higher than average amount of fish) fish consumption rate of 123
36
-------
CB/NT PCB Responsiveness Summary
grams/day. EPA believes that in the absence of subsistence fishing data, the high-end
tribal fisher estimate provides a reasonable representation of the type of consumption
rates that might be seen in a subsistence population. For this reason, the fish
consumption exposure scenario has formed the basis of EPA's PCB risk revaluations
for the CB/NT Site. In the Addendum (Weston, 1997a), risks are estimated for both
average and high-end recreational and average and high-end tribal fishers.
4.2.2 The 95* percentile ingestion rate reported the Toy et at. (1996) should have been used
to represent the fish ingestion rate of the high-end tribal fisher. Use of the 90* percentile is
inconsistent with other Region X risk assessments (e.g., EPA, Ecological Risk Assessment and
Seafood Screening Risk Assessment Asarco Sediment Site, October 1996) and will likely result
in an under estimation of subsistence angler's fish consumption. A high-end exposure rate of
292 grams per day (about 10 ounces per day) should be used to calculate the predicted cancer
risk rates. EPA used a fish consumption rate of 123 grams per day. (8, 26, 36)
Response: The choice of the 95th or the 90th percentile fish consumption rate must be
made on a site-by-site basis in order to best represent site-specific exposures. EPA
guidance directs that to calculate risks based on a reasonable maximum exposure
(RME), a combination of average and high-end values for exposure parameters be used
to estimate exposures. Such a combination is expected to result in an RME that is a
realistic representation of risks to individuals with some of the highest exposures (90th
to 95th percentile) at the Site. Therefore, it is not necessary to use the highest values
available for all parameters to establish the RME scenario. For the CB/NT Site, it was
determined that in combination with other parameters, the 90th percentile ingestion rate
would be a realistic and protective value to represent some of the most highly exposed
tribal fishers at the Site.
In the Screening Risk Assessment done for fish consumption at the Asarco site, a range
of fish consumption values (from 1 gram of fish per day to 292 grams of fish per day)
was used. The highest value (292 grams per day) was based on the 95th percentile
consumption rates for the Squaxin Island Tribe in the draft version of the Toy et al.
study. The 95th percentile ingestion rates in the final Toy study are lower than those
in the draft, but somewhat higher than those used in the CB/NT PCB residual risk
evaluation. EPA chose to use the 90th raiher than the 95th percentile ingestion rate for
the reasons discussed above.
4.2.3 In the calculation of risk to people consuming fish from the Hylebos Waterway, the
assumption that 100 percent offish consumed from Commencement Bay came from the
Hylebos Waterway is an overestimate. The Hylebos Waterway is not a prime fishing area and
studies indicate that the fraction offish caught in the Hylebos is no more than 5 percent of the
total caught in Commencement Bay (Pierce et al., 1981). (17)
Response: EPA is primarily using the CB/NT-wide human health risk calculations, not
the Hylebos Waterway scenario, for decision-making purposes. The Hylebos
37
-------
CB/NT PCB Responsiveness Summary
Waterway scenario was developed only as a check to ensure that a cleanup level based
on the CB/NT risk evaluation would not result in an unacceptable risk to someone who
did obtain a large amount of fish from the Hylebos Waterway. Furthermore, EPA
evaluates potential cleanup levels using both current and potential future use scenarios
for a site. For this reason it is not sufficient to decide whether or not to evaluate a
scenario based solely on data collected in 1981.
4.2.4 Data demonstrate that there can be up to 90 percent PCB losses during frying offish
(Landolt et al., 1987) and frying is the most common preparation method. Yet, in the risk
calculations, it was assumed that 100 percent of PCBs remained in the fish. (17)
Response: See response to Comment 1.1.8.
4.3 Comments related to the toxicity of PCBs to people
4.3.1 As a practical matter, it is incongruous to accept a decrease in the PCB toxicity factor
from 7.7 to 2.0 -without making a corresponding change in magnitude to the cleanup level.
(25, 30, 13)
Response: As noted in response to Comment 1.1.2, in order to best represent the most
current EPA policies and most current scientific data, all parameter values used to
calculate risks were reevaluated. The PCB toxicity factor is not the only input
parameter value to be updated since the ROD (see parameter values used in the
Addendum [Weston, 1997a]). Therefore, the resulting change to risk estimates, and
consequently, to the proposed PCB cleanup level, is not solely proportional to the
change in the PCB toxicity factor.
4.3.2 The cancer-based human health risk assessment does not acknowledge or address the
wide range of non-cancer health effects associated with PCBs. (3, 36)
Response: In the human health risk assessment done for Commencement Bay in 1985,
the potential for both cancer and non-cancer health effects from exposures to PCBs in
fish from the Bay was evaluated. Based on the information available on the toxicity of
PCBs at that time, it was concluded that the potential for non-cancer impacts was not of
concern. Therefore, when the ROD was written in 1989, EPA based its cleanup level
for PCBs on human health cancer risks from ingestion of PCB-contaminated fish caught
in the Bay.
Similarly, in the updated risk evaluations (Weston 1996, 1997a,b), EPA focused on
cancer risks associated with PCBs, and updated non-cancer risk information was not
presented in these reports. In response to comments EPA received on the Weston
report, EPA has prepared a technical memorandum which provides information
regarding non-cancer risks under current and post-cleanup conditions using the
scenarios developed for the cancer risk evaluation. This technical memorandum has
38
-------
CB/NT PCB Responsiveness Summary
been added to the Administrative Record for EPA's final decision (Weston, 1997b). A
summary is provided below.
Based upon an understanding of the development of non-cancer health effects, potential
non-cancer impacts are evaluated by EPA assuming that there is a level of exposure
below which health impacts are unlikely to occur. The estimate of this level of
exposure is called the reference dose, or RfO, and is defined as "an estimate (with
uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
population (including sensitive subgroups) that is likely to be without an appreciable
risk of deleterious effects during a lifetime." Exposures that are less than the RfD are
not likely to be associated with adverse health impacts. (See the response to
Comment 4.3.3 for more information on the RfO for PCBs that was used to estimate
non-cancer health impacts.)
In order to calculate non-cancer risks for site-specific risk assessments, EPA compares
the RfD to the exposures estimated for that site (e.g., from eating contaminated fish).
This comparison, called the Hazard Quotient or HQ, is the ratio between the estimated
site exposure and the RfD. As with cancer risk, the assumptions used in calculating
the HQ are conservative ones (health protective) to ensure that remedial decisions based
upon them will protect more sensitive individuals. Because of the way in which the
HQ is derived, it should not, however, be viewed as a strict demarcation between toxic
and nontoxic.
Results of this non-cancer risk evaluation show that, as for cancer risks, current
conditions in the CB/NT Site are of concern for people who may be eating large
amounts of fish from the Site. Under current conditions for the high-end tribal fisher,
the HQ is 60. Under the 450 ^g/kg PCB cleanup level, the HQ would be reduced to 8
immediately after cleanup and to 7 in 10 years after cleanup. Under the 150 ^g/kg
PCB cleanup level in the 1989 ROD, the HQ would be reduced to 7 immediately after
cleanup and to 6 in 10 years after cleanup. For an average tribal fisher, these HQs
would be reduced by a factor of 3. Both the cleanup required under the 1989 ROD and
in the ESD (EPA, 1997a) provide for substantial reduction in the non-cancer risks
associated with PCBs in sediments at the CB/NT Site. As with cancer risks, given the
range of uncertainty in risk calculations, these post-cleanup HQs are not significantly
different. Therefore, as with cancer risks, EPA does not believe that the 450 ^g/kg
PCB cleanup level provides significantly different non-cancer risks than the PCB
cleanup level presented in the 1989 ROD.
It should be noted that HQ values above 1 do not mean that non-cancer health impacts
will occur, but rather that the potential for such impacts increases as 1 is exceeded. The
potential for impacts depends on a number of factors, including the protectiveness of
both the RfD and the exposure assumptions used to calculate the HQ. As described in
the response to Comment 4.3.3, the derivation of the RfD for PCBs is based upon a
large body of experimental data and incorporates a several-hundred-fold uncertainty
39
-------
CB/NT PCB Responsiveness Summary
factor ("safety"-factor) to ensure protection. Also, the exposure assumptions used in
the CB/NT risk evaluation were selected to be protective for a consumer of large
amounts of fish from the CB/NT Site over a 30-year period. Given these conservative
assumptions, the small increases above an HQ of 1 estimated for the various target
cleanup levels and for background suggest a low potential for non-cancer impacts for
the fish consumers considered in the calculations. Individuals who eat less fish from the
CB/NT Site will have exposures and HQs that are lower and, therefore, their potential
for non-cancer impacts will be less.
4.3.3 EPA's Reference Dose (RfD) does not acknowledge or address the wide range of non-
cancer health effects associated with PCBs including impacts on the immune system;
reproductive and developmental system (miscarriages, impaired testicular descent, genital
deformities, and other general impairments in reproductive and sexual development);
neurological system (lowered intelligence, behavioral problems); and endocrine system
(including hormone mediated cancers and the human breast cancer connection) and the fact
that embryos, fetuses and young children are at much greater risk than adults particularly with
respect to the hormone disruption impacts of PCBs. (8, 26, 36)
Response: In developing the RfD for a specific chemical, EPA scientists review all of
the non-cancer studies on that chemical, including those done on experimental animals
and those done on humans that have been exposed to the chemical. Experimental
animal data are often from experiments performed on different species and at different
exposures. From these data, the study showing the toxic effect at the lowest level of
exposure is chosen as the "critical study." The level in this study at which no effect
was seen (or the lowest level, if effects were seen at all exposure levels) is then
selected and uncertainty factors are applied to this level to account for uncertainties in
the database (e.g., in using animal data) and to protect sensitive individuals (e.g., the
fetus, children, asthmatics).
The EPA RfD used for calculating the HQs (Hazard Quotients) discussed in Comment
4.3.2, is the RfD for Aroclor 1254 (an industrial mixture of PCBs) (EPA, 1997b) since
this was the mixture that most closely approximated the PCBs present in CB/NT
sediments. EPA published this RfD in 1994. In developing this RfD, scientists from
several EPA offices reviewed all of the published (and some unpublished) literature
(over 90 studies) on the non-cancer effects of PCBs. These included studies that were
conducted on several species of experimental animals as well as studies on human
populations that had been exposed to PCBs. The types of effects studied in
experimental animals ranged from very obvious toxicity (e.g., death, reduced survival,
obvious reproductive effects like gross abnormalities and reduced liner size, and organ
damage) seen at higher exposures to PCBs to those that are more subtle (e.g.,
immunological effects and impacts on blood chemistry) and which can occur at lower
exposures. The exposure levels in the animal studies ranged from those as low as 5
Mg/kg-day (micrograms per kilogram per day) to those that were thousands of times
higher. The number of human studies available and the types of effects studied were
40
-------
CB/NT PCS Responsiveness Summary
less than that for experimental animals, but both obvious and more subtle toxicological
effects resulting from occupational and accidental exposures were reviewed.
The data chosen to develop the RfD (the "critical study") was a series of studies on
rhesus monkeys that had ingested PCBs for five years at levels as low as 5 /zg/kg-day
(the lowest level of exposure studied in animals) (Arnold et at., 1993a, 1993b;
Tryphonas et al., 1989, 1991a, 1991b). A battery of effects was studied including
impacts on general health, blood chemistry, the immune system, reproductive
endocrinology, and hormone levels. This study was chosen for the RfD because it was
technically sound and it showed effects at the lowest dose of PCBs tested. These
effects included inflammation of the eyelid glands and eye discharge; changes in the
fingers and toenails such as elevated nail beds and abnormal nail foldings; and an
inability of the immune system to respond.. A "safety" or uncertainty factor of 300
was applied to the lowest effect level seen in this study (5 ^g/kg-day) to account for
several uncertainties, including extrapolating from monkeys to humans, to protect
sensitive individuals in the human population, and to account for the use of data where
the lowest effect (rather than no effect) had occurred. Therefore, the final RfD of
0.02 ^g/kg-day is at a level that is 300 times lower than the level at which an effect
was seen.
At the lowest levels tested, this study did not show effects on the menstrual cycle or on
estrogen levels. Effects on reproduction and on the offspring of these monkeys has not
yet been published but a preliminary review of the unpublished data by EPA indicated
that there may be reproductive effects at the lowest level tested. This is in contrast to
another study done on rhesus monkeys (Levinskas et al., 1984) which showed no
reproductive effect at the 5 /zg/kg-day level. The Levinskas study included evaluations
of the exposed adults as well as the offspring of these adults. In both the adults and
their offspring, in addition to clinical analyses of toxicity (e.g., skin and eye
problems), studies were done on blood and urine chemistry. The adults were also
studied for male reproductive effects and female conception rates. This study found no
effect at the 5 /zg/kg-day exposure level.
It should be noted that potential adverse impacts on several of the organ systems
mentioned by the cotnmentors as being of concern for PCBs were looked for in the
animal studies done on PCBs and were considered in the development of the 1994 RfD.
Effects on some of these systems were not observed seen until exposures exceeded that
used in developing the RfD (the lowest effect level); therefore, the RfD is expected to
be protective for these effects that are occurring at higher exposure levels.
Some commentors stated that EPA was ignoring several studies of non-cancer health
impacts in humans. In developing the RfD, EPA reviewed the available human data on
the non-cancer effects of PCBs, but concluded that these data were only useful in a
qualitative manner. Studies have been done on the general population who were
exposed to PCBs via consumption of contaminated food (e.g., fish). Infants have been
41
-------
CB/NT PCB Responsiveness Summary
evaluated for impacts on the nervous system and on behavior after being exposed in the
uterus or through breast feeding. However, the types of PCBs to which these
populations were exposed, the levels of PCB exposure, the levels of exposure to other
contaminants (e.g., in fish and breast milk), and other details of exposure are not
known, making it very difficult to use these data to calculate an RfD. This is also true
for studies of workers exposed to PCBs. Although the majority of the human studies
of toxic effects of PCBs is from occupational exposures, these data are insufficient to
develop an RfD due to lack of information on the levels of PCB exposure and
concurrent exposures to other chemicals.
Some commentors criticized EPA for not considering endocrine-disrupting effects in
the development of its RfD. An endocrine disrupter is an agent that interferes in some
way with the natural hormones in the body (e.g., those secreted by the pituitary,
thyroid, pancreas, adrenal, testes and ovaries). An agent might disrupt the endocrine
system by affecting any of the various stages of hormone production and activity. A
variety of chemicals, including PCBs, have been found to cause endocrine disruption in
laboratory studies, including disruption of female and male reproductive function (such
as sperm production, ability to conceive) and effects on the thyroid gland, which helps
maintain normal metabolism. It should be noted that several of the experimental animal
studies reviewed in developing the RfD for PCBs looked for these types of effects.
Also, some of the human endocrine disrupter data cited by the commentors as not being
considered by EPA in its RfD development (e.g., neurobehavioral effects in children)
were reviewed for the RfD. As discussed above, these studies could only be used in a
qualitative manner.
Since the development of the RfD in 1994, there has been additional research done on
endocrine disrupters. A review of all of the research on endocrine disrupters, including
this newest data, was completed by EPA's Risk Assessment Forum in February of
1997 (EPA, 1997c). In its report on this review, EPA concluded that, with few
exceptions, a causal relationship between exposure to a specific environmental agent
and an adverse health effect in humans operating via endocrine disruption has not been
established. Exceptions are incidents of high chemical exposures in the workplace and
exposure to the drug, DES. For example, conclusive evidence linking environmental
exposure to endocrine disrupters with infertility or cancers of the breast or prostrate
(all mentioned by the commentors) is not available at this time.
The 1997 report also recognized that there is concern about the possibilities of impacts
to human health due to exposure to endocrine disrupters, including the potential risk to
the developing young who may be at more risk than adults (as mentioned by the
commentors), and that more research into this area and into other areas mentioned by
the commentors needs to be done. EPA is preparing a draft research policy on
endocrine disrupters that will be released later this year. EPA has also established a
task force to develop screening and testing methods for use in evaluating chemicals for
42
-------
CD/NT PCB Responsiveness Summary
endocrine effects. Efforts are underway to coordinate endocrine disrupter research
throughout the federal government.
Finally, EPA will soon complete a reassessment on the toxicity of and exposures to
chlorinated dioxins and furans, as well as dioxin-like PCBs. Dioxin-like PCBs are
those PCBs having a chemical configuration similar to chlorinated dioxins and furans,
and which are, therefore, thought to act in a lexicologically similar fashion to dioxins
and furans. The results of the review of existing data and the new data generated by
this reassessment could have an impact on the way in which non-cancer impacts from
this specific group of PCBs are evaluated.
The results of the research efforts described above will be used by EPA scientists to
determine whether a modification to the PCB RfD is necessary. If EPA decides to
significantly modify the RfD for PCBs, it may be necessary for EPA to reevaluate
cleanup decisions made at Superfund sites where PCBs are a contaminant of concern to
determine whether the cleanups at those sites are still protective of human health and
the environment.
4.3.4 Because of the limited numbers of hormone receptors in developing human embryos and
fetuses, the fact that hormones are already operative, and the essential role hormones play in
survival and proper development, a strong case can be made that there is no safe level for
PCBs. (36)
Response: If by a "safe level," the commentor is referring to a "no adverse effects"
level, EPA agrees that this may be the case for many cancer-causing chemicals.
Because of the way in which chemicals are thought to cause cancer, EPA assumes that
it is not possible to define a level for exposure to a cancer-causing chemical which it is
certain will be associated with no effects. Therefore, all exposures are assumed to
potentially have some risk associated with them with the risk of developing cancer
increasing as the exposure increases. EPA regulations require that Superfund cleanups
achieve a reduction in contaminant concentrations such that the chance of developing
cancer under a "reasonable maximum exposure" scenario is within the range of lO'* to
For the non-cancer effects of a chemical, it is assumed that a "no adverse effects" level
of exposure can be determined— this level is defined as the RfD (Reference Dose) by
EPA. In developing the RfD for PCBs (discussed in detail in the response to
Comment 4.3.3), all of the available non-cancer data on PCBs were considered,
including those relating to adverse hormonal impacts.
However, as discussed in its recently released review on endocrine disrupters (see
Comment 4.3.3), EPA recognized the need for further research on exposure of
neonates and the human fetus to environmental levels of endocrine disrupters and
discussed several ongoing efforts to expand and improve upon the research in this area.
43
-------
CB/NT PCB Responsiveness Summary
The results of tfiis research should provide further information on the question of a safe
threshold for non-cancer impacts due to PCBs.
4.4 Comments related to ecological exposures to PCBs at the CB/NT Site
4.4.1 The bioaccumulcaion model assumes that piscivorous birds {fish-eating birds) obtain all
of their prey (forage fish) within the Hylebos Waterway. This assumption overestimates the
exposure of migratory birds or birds with large foraging ranges, like the eagle. (17)
Response: EPA primarily used the CB/NT-wide risk estimates, not the Hylebos
Waterway risk estimates, for decision-making purposes. The waterway-specific
information was used, however, as a check to ensure that a cleanup level was not
selected that would have a disproportionately large impact to individuals who
preferentially utilize one waterway. EPA believes that this is a reasonable approach
based on FWS information indicating that individual birds can demonstrate a high
degree of site fidelity in their foraging/hunting activities (Krausmann, 1996). See also
EPA's response to Comment 3.3.4.
It should be noted that the primary basis for EPA's decision on the PCB cleanup level
was the human health risk evaluation. The wildlife risk evaluations provided in FWS
comments (FWS, 1996a,b) and the 1997 Weston report were used only to confirm that
the selected PCB cleanup level for protection of human health would also be protective
of ecological receptors. It should also be noted that EPA did not use the eagle data
provided by FWS for decision-making purposes, due to concerns about the uncertainty
associated with the biomagnification factor and the lack of data on foraging ranges.
4.4.2 The use of BSAF of 1.7 should be reevaluated. Current models and studies in the
scientific literature recommend a BSAF of 4. (24)
Response: See response to Comment 1.5.1.
4.4.3 The use of an average total organic carbon (TOC) value for the various waterways may
not be appropriate. The lowest TOC value representative of an area in the waterway should
be used to more accurately reflect what a species with site fidelity would be exposed to. (24)
Response: The distribution of total organic carbon represents a gradient of low TOC
at the mouth and high TOC at the head in both the Thea Foss and Hylebos waterways.
Use of the lowest TOC would reflect only a small portion of the waterway. It is more
accurate to represent exposure based on prevailing conditions, which are better
represented by the mean TOC for the waterway.
4.4.4 Fish in Commencement Bay contain high levels of PCBs in their bodies. A recent study-
by the National Marine Fisheries Service (NMFS) in the Hylebos Waterway of Commencemeni
Bay links the presence of PCBs to altered sexual reproduction of flatfish. In (he studies, half
44
-------
CB/NT PCB Responsiveness Summary
of all juvenile female flatfish showed signs of premature sexual development. The same study
also revealed levels ofPCBs in young salmon found migrating through the Hylebos Waterway
were comparable to levels found in other young salmon in the Duwamish River. The
Duwamish River study documented impaired growth, suppression of immune function, and
increased mortality rates after exposure to pathogens. English and rock sole in the Hylebos
Waterway have high rates of liver lesions which are associated with PCBs in their bodies.
Despite the passage of 10 to 15 years and some remediation, there has been no real
improvement in terms of exposures and injuries for these fish. (8, 26, 36)
Response: EPA agrees that current conditions are adversely impacting fish in the
Hylebos Waterway, which is one of the reasons EPA wants to complete the sediment
cleanup as soon as possible. The NMFS studies, however, do not provide information
about a PCB cleanup level that would prevent such effects from continuing to occur.
4.4.5 In the last two years, we have seen an increase in predation on the great blue heron
colony by bald eagles. We believe these eagles are at further risk of adverse impacts due to
the nature of PCBs bioaccumulating through the food chain. (35)
Response: Although raptors may represent sensitive receptors, EPA does not have
sufficient data to evaluate the risks to these receptors at this time. It is EPA's
understanding that FWS is conducting work that would support this evaluation, but that
these data are not yet available.
4.4.6 Fish-eating birds, such as bald eagles and blue herons, in the Commencement Bay area
have high levels of PCBs in their embryos. PCB levels in blue heron eggs from Dumas Bay,
located near Commencement Bay, are 5 to 10 times higher than the observed level at which
negative effects occur. The herons from Dumas Bay use Commencement Bay as their feeding
grounds. (8, 26, 36)
Response: It is EPA's understanding that FWS is conducting work that would support
the evaluation of threats to heron and bald eagle, but these data are not yet available.
EPA agrees that at current conditions, PCBs in sediments may be adversely affecting
piscivorous birds, and may present an imminent and substantial endangerment to
wildlife until a cleanup is implemented. EPA has included an evaluation of the
potential impacts to piscivorous birds from ingestion of contaminated prey, based on
the herring gull data from the Great Lakes, to ensure that the selected PCB cleanup
level will be protective of wildlife.
45
-------
CB/XT PCB Responsiveness Summary
4.5 Comments related to the ecological toxicity of PCBs
4.5.1 We disagree that the range ofLOAELs (Lowest Observable Adverse Effects Levels)
presented represents a variability of risk, and therefore uncertainty, in the risk assessment.
For example, the adverse effects associated with levels of PCBs in the egg of a bald eagle do
not increase or decrease the adverse effects observed and associated with the levels of PCBs in
the egg of a Caspian tern. The risk assessment calculated hazard quotients for birds based on
the "average LOAEL presented by the FWS." (emphasis added). However, the tables
presented by the FWS on LOAELs were intended to show the levels of total PCBs associated
with adverse effects to specific bird and wildlife species from various published literature
sources. Data from this table was not averaged in its original presentation nor was the intent
to imply that averaging effects to individual species would be appropriate in an ecological risk
assessment designed to determine overall protectiveness for fish and wildlife. Risk should be
calculated for the most sensitive of the species of concern if the intent is to be protective of that
species. (35)
Response: EPA agrees that it is important to protect for sensitive species. There was
little difference in the lowest observable adverse effects levels (LOAELs) reported for
piscivorous bird egg effects (3.5 to 5.0 mg/kg). Use of the LOAEL for the most
sensitive species (double-crested cormorant; 3.5 mg/kg) compared to the average
(4.2 mg/kg; equivalent to the Caspian tern LOAEL) would have resulted in only a
small change in the hazard quotient and would not have changed the overall
interpretation of the estimated risk.
4.5.2 // is well established that PCBs can have significant genetic and reproductive effects to
fish and wildlife at low concentrations. Because of the ecological significance of these effects,
specifically to the viability offish and wildlife populations which utilize Commencement Bay, a
conservative approach to the ecological evaluation is warranted. (24, 36)
Response: As noted in response to Comment 4.4.1, EPA's primary basis for decision-
making on the PCB cleanup level was the human health risk evaluation. EPA used
information about ecological risks presented in the FWS comment letters (FWS,
1996a,b) and the Addendum (Weston, 1997a) to confirm that the PCB level considered
protective of human health would also be protective of ecological receptors. As
discussed in the response to Comment 1.6.2, EPA has used the information provided
by the FWS and NOAA to evaluate the risks to wildlife. EPA used the PCB
bioaccumulation models provided by the FWS to address impacts to higher order
receptors, such as fish-eating birds that may bioaccumulate PCBs through the food
chain, in the evaluation of the potential for impacts to ecological receptors. EPA
assumed piscivorous birds were obtaining 100 percent of their diet from the CB/NT
Site. EPA also used the arithmetic mean to represent the residual PCB concentration,
which, as discussed in responses to Comments 1.2.2, 4.1.2, and 4.1.5, is a
conservative representation.
46
-------
CB/NTPCB Responsiveness Summary
4.5.3 Using cancer as an end point in the assessment process for PCB 's does not account for
the true injury to the biota of Commencement Bay and will not be protective of these biota.
(29, 36)
Response: EPA agrees. Cancer was not used as the endpoint for evaluation of the
risks to wildlife. Rather, EPA evaluated reproductive effects in piscivorous birds,
growth and survival in juvenile salmonids, and abundance and diversity of benthic
infauna to determine the potential effects to wildlife, under various cleanup scenarios
proposed for the protection of human health.
4.5.4 According to the United States Fish and Wildlife Service and the National Marine
Fisheries Service studies, PCBs cause not only increased mortality rates, but several other
short-of-fatal but serious problems. These include impaired growth, suppression of immune
function and impaired or premature sexual development. The results of the FWS model (FWS,
1996) indicate that by leaving 450 ng/kg of PCBs in the Commencement Bay environment,
fish-eating birds could accumulate levels of PCBs 5 to 10 times higher than levels where we
would first expect to see adverse reproductive impacts such as embryonic deformities and
death. Not only is it unconscionable to knowingly permit such harmful levels to continue, it is
simply wrong-headed to believe that humans will escape the effects of such toxic chemical
substances. Thirty ^gfkg will protect fish-eating birds, so we are assured that even 150 ng/kg
is high for fish. (3, 35)
Response: EPA agrees that PCBs in CB/NT sediments at current concentrations may
present an imminent and substantial endangennent to a variety of wildlife species and
people. EPA agrees that PCBs should not be allowed to remain in CB/NT sediments at
current concentrations.
The commentor is correct that FWS has recommended a 30 ^g/kg PCB cleanup level to
protect fish-eating birds. FWS assumed that after cleanup, fish would reside in areas
where they would be exposed to only the maximum PCB concentration remaining in
sediments, and that birds would only eat fish that had these high exposures.
EPA's policy is to use a "reasonable maximum exposure" scenario in developing
cleanup levels for Superfund sites, which is defined as the highest exposure that is
reasonably expected to occur at a site. The area within the CB/NT Site where
maximum PCB concentrations (i.e., PCB concentrations at the cleanup level) will occur
post-cleanup will be quite small. In the evaluation of the impacts to wildlife at various
PCB cleanup levels, EPA assumed that fish would move about within the waterways
and be exposed to average PCB concentrations, rather than maximum concentrations.
EPA used FWS's estimates to predict the impact to birds eating these fish. Based on
EPA's analysis, the impacts to birds and fish at the 450 ^g/kg PCB SRAL and 300
PCB SQO are not appreciably different than the impacts at the original 150
PCB SQO.
47
-------
CB/NT PCB Responsiveness Summary
4.5.5 PCBs are one ef several chemicals known to disrupt the endocrine systems of birds and
mammals. Other studies of PCBs show impaired sexual development in marine life. Small
penises in young Columbia River otters have been linked to PCBs in their bodies and these
young otters do not appear to be able to produce sperm. (8, 26, 36, 29)
Response: EPA is also concerned about the potential for PCBs to disrupt endocrine
systems, as discussed in the response to Comment 4.3.3.
4.5.6 The new wildlife criterion for aqueous concentrations of PCBs in the Great Lakes
(EPA, 1997) is 1.2* 104 ng/mL. This is based on an average log Kw of 6.59. PCBs are
expected to partition similarly in freshwater and marine systems. Equilibrium partitioning can
be used to predict the sediment concentration that should produce this aqueous concentration.
The equation: fsedj/f^ = K^ * [water], is used to predict sediment concentration that will
produce the criterion concentration of 1.2 x 104 ng/mL. Setting K^ = K^ the predicted
sediment concentration is 466 ng/g organic carbon. For a TOC of 2.4 percent, the sediment
concentration would equal 11.2 ng/g. (24)
Response: While some PCBs may partition to the water column, chemical properties
of PCBs will result in a significantly larger amount remaining in the sediments.
Therefore, EPA does not believe it is appropriate to increase the uncertainty associated
with the risk estimates by extrapolating to another medium (i.e., water). Furthermore,
the approach suggested by the commentor proposes to make additional extrapolations
and assumptions (e.g., K^ = K^.) that add more layers of uncertainty to the risk
estimates. EPA believes the approach utilized in the human health and ecological risk
evaluations presented in the Addendum (Weston, 1997a) is a more appropriate
evaluation for the conditions present at the CB/NT Site.
4.6 Comments Relating to additional exposure to contaminants
4.6.1 Combinations of chemicals can be far more deadly than PCBs acting alone.
Combinations of PCBs and other pollutants can also have much greater effects than the
chemicals considered one at a time. PCBs are not the only pollutants in sediment, in animals,
in people's daily environments, and in people's fat, food and breast milk. At a minimum, each
person carries 250 synthetic pollutants in his or her body. More and more studies have
documented that combinations of chemicals can have far worse effects than chemicals in
isolation. This proposal does not consider the overall risks due to the synergistic or additive
nature of these chemicals with PCBs and their impacts on the natural resources which inhabit
this embayment. (36, 29)
Response: The available scientific data on mixtures (including those of endocrine
disrupters) are primarily from laboratory experiments and some occupational studies.
These data suggest that exposures to multiple chemicals may result in the following
toxicological effects : (1) additivity such that the resulting toxicity is equivalent to the
sum of the toxicity of each chemical alone; (2) antagonism such that the resulting
48
-------
CB/NTPCB Responsiveness Summary
toxicity is less than the sum of each chemical added together; or (3) synergism such
that the resulting toxicity is greater than the sum of the each chemical added together.
However, the current database is not complete enough to determine which of these
effects are occurring from a particular environmental exposure or to quantify these
effects. Therefore, EPA's current approach is to assume additivity of effects (EPA,
1986). For this assumption, the combined effects of antagonism and additivity are
assumed to balance the possible effects of synergism.
For carcinogenic effects, additivity is assessed in the Superfund program by combining
the cancer risks resulting from chemical exposures from the site contaminants. For
non-cancer effects, additivity is applied to site-related contaminants that produce their
toxicological effect by similar types of actions (e.g., specific impacts on the immune
system). As a first step, all of the HQs (see Comment 4.3.2 for a definition of the HQ)
are added, and the combined HQ for all contaminants is evaluated.
In the human health risk assessment done for the CB/NT Site in 1985, the potential for
both cancer and non-cancer health effects from exposures to contaminants in fish from
the Site was evaluated. This evaluation was done using data on contaminant levels in
fish tissue collected from several parts of the Site, as well as from Carr Inlet, which
was selected as the reference or background site. EPA toxicity values were used with
these observed fish contaminant levels and a range of fish ingestion rates (up to 1
pound or 454 grams per day) to estimate potential cancer risks and non-cancer hazards.
The mixtures of PCBs present at the CB/NT Site were accounted for in this risk
assessment because the toxicity data used and the chemical contaminant data from fish
were those for the mixture of PCBs at the Site.
The risk assessment concluded that, for chemicals that are potentially carcinogenic,
three of these are present in fish tissue at levels different from the levels in background
(Carr Inlet) fish—PCBs, tetrachloroethene, and bis(2-ethylhexyl)phthalate (BEHP).
The risks for tetrachloroethene and BEHP were more than an order of magnitude lower
than for PCBs. For chemicals with non-cancer effects, it was concluded that only three
chemicals were at levels that are slightly above (less than 2 times higher) the
Acceptable Daily Intake (ADI) (now called the RfD or Reference Dose by EPA), but
that there is essentially no difference in the tissue levels of these three contaminants at
the CB/NT Site versus Carr Inlet.
Although additivity of toxicity for these three site-related carcinogens was not explicitly
addressed in the risk assessment, the predicted risk values for tetrachloroethene and
BEHP are so much lower than those for PCBs, they would not significantly add to the
CB/NT Site risk due to PCBs.
As discussed in the response to Comment 4.3.3, new research is being done on
endocrine disrupters and dioxin-like PCBs. It is possible that the results of this new
49
-------
CB/NT PCB Responsiveness Summary
research could change the methods EPA currently uses to assess mixtures of PCBs and
mixtures of endocrine-disrupters like PCBs.
4.6.2 PCBs are but one of several classes of chemicals known to disrupt the endocrine
systems offish, wildlife and humans (Colborn, 1993). Other "endocrine-disrupters" such as
certain pesticides, heavy metals, and other industrial chemicals are known to exist in various
quantities in the CB/NT Site. The current proposal makes no reference to how the proposed
changes in the PCB cleanup level will affect overall risk due to the synergistic or additive
nature of these other chemicals remaining in the Commencement Bay environment. To date,
the FWS has provided the EPA with information concerning the adverse impacts of elevated
PCBs on fish and wildlife in Commencement Bay. However, literature suggests that, at the
very least, adverse impacts to fish, wildlife and humans may be additive in the presence of
other endocrine-disrupting chemicals (Colborn, 1993). (35)
Response: Please see responses to Comments 4.3.3 and 4.6.1.
4.6.3 I urge EPA to think about the concept of global loads of PCBs in the Commencement
Bay decision. Your narrow analysis misses the big picture of a planet which desperately needs
your leadership in many different sites concurrently to reduce the overall PCB loadings in the
global environment. Many studies have documented that people who live far away from any
sources of pollutants have some of the highest levels of these pollutants. Scientists now
understand that this is the result of pollutants volatilizing in warmer regions, traveling by air
to colder regions, and condensing there. Vie request that EPA investigate and factor in
evaporation and deposition routes of exposure for people living in colder regions where PCBs
have been documented to collect via transfer from industrial areas. (36)
Response: EPA agrees studies have shown that PCBs have been transported globally
and can be found in remote and non-industrial regions. However, EPA's Superufund
regulations require that cleanup decisions be based on potential exposures and health
impacts from contaminants from a specific Superfund site. The concept of global
leadings is not considered in the Superfund law or regulations. For the CB/NT Site,
decisions have been based on protection of those individuals who are expected to have
the highest exposure to site contaminants. EPA believes that by selecting cleanup
levels that are protective of those living close by the site and who will have the highest
potential exposure to site contaminants, (in this case, a tribal fisher who eats a higher
than average amount of fish from the site), the selected PCI! cleanup level will also be
protective of those who live further away and receive a smaller dose through air
transport. Because of the low solubility of PCBs, the potential dose of PCBs one might
receive from their partitioning from sediments to the water column and then entering
the atmosphere from the CB/NT Site is quite small, especially when compared to the
dose received by eating PCB-contaminated fish.
4.6.4 Even if the concept of safe thresholds of PCBs and other pollutants is still valid—a
matter in great dispute—fish, people and other animals are so contaminated with PCBs and
50
-------
CB/NT PCB Responsiveness Summary
other pollutants already that we have passed those thresholds. Study after study have shown
that fish, people and other animals are carrying very significant quantities of these PCBs in
our bodies. A risk assessment cannot accurately depict risk if it does not take these levels into
account. (36)
Response: EPA recognizes that people may be exposed to a variety of chemicals in
everyday life, in addition to potential exposures from a Superfund site. Because each
individual will have different exposures based on their occupation, place of residence,
and lifestyle, it is difficult to develop a method which accounts for these varied
exposures. Instead, EPA's Superfund regulations and guidance focus on the lifetime
excess cancer and non-cancer risks from exposure to Superfund site contaminants and
bases cleanup decisions on those risks. EPA assumes that receptors do have other
exposures than those from the Superfund site, and strives to set Superfund cleanup
levels that do not pose an unacceptable additional risk.
PCBs are carcinogenic chemicals, for which EPA does not recognize a "safe"
threshold. However, in assessing carcinogenicity to humans, EPA has set a standard
that estimated lifetime excess cancer risks on the order of 10"* to 10* fall within an
acceptable range. EPA evaluates all risks to ecological receptors based on a variety of
effects, for which a threshold value is recognized. Risks to ecological receptors are
calculated using lowest observed adverse effects levels (LOAELs) or no observed
adverse effects levels (NOAELs). Exposures that exceed these thresholds are identified
as potential risks.
4.6.5 With thousands of pollutants in the environment and a minimum of hundreds in each
person, how will you account for the myriad of potential combinations of exposure? Will you
test each combination? Will you have side calculations in your risk assessments for each
possible combination? (36)
Response: See responses to Comments 4.6.1 and 4.6.4.
5 Comments Related to Remedial Decision-Making
5.1 Comments related to the protectiveness of the proposed PCB cleanup level
5.1.1 The risk assessment is based on sediment concentrations and, as such, has a higher
degree of uncertainty than an assessment based on fish tissue concentrations. Because of this
higher degree of uncertainty, it is not sufficiently protective to utilize a cleanup level that has a
low margin of safety, such as the proposed 450 ng/kg PCB cleanup level. At the 450 ^.g/kg
cleanup level, both the estimated incidence of cancer and the Hazard Indices are at the high
end of the acceptable range. A more restrictive (lower) cleanup level would be more
protective of ecological health. (34, 11)
51
-------
CB/NT PCB Responsiveness Summary
Response: EPA agrees that the 450 ^g/kg PCB SRAL falls at the high end of a range
of cancer and non-cancer risks which EPA has determined to be acceptable for
Superfund cleanups. EPA has now added a requirement that PCB concentrations in
sediments must be reduced to 300 Mg/kg within 10 years after remedial action. This
will ensure that human health cancer risks are reduced by an additional 14 percent in
the 10 years following remedial action. It should be noted, however, that all PCB
cleanup levels examined from 450 to 150 ^g/kg result in residual risks at the high end
of EPA's risk range, due to the conservative scenario used for risk calculations (a high-
end tribal fisher who fishes exclusively in the CB/NT area).
If EPA's analysis had shown that cleanup to the 1989 PCB SQO of 150 Mg/kg would
result in a significantly smaller risks to humans or wildlife, and that such a cleanup was
cost-effective, EPA would agree that lower PCB cleanup level in the 1989 ROD should
be retained. However, EPA's analysis shows that residual ecological risks at the
revised SQO of 300 /zg/kg PCB are not significantly higher than the risks associated
with cleanup to a 150 pgfkg PCB SQO under the 1989 ROD. See Tables 1 through 6
of the ESD (EPA, 1997a) for a comparison of the estimated risk values at current
conditions, at the ROD PCB cleanup level, and at the new PCB SQO of 300 Mg/kg and
SRALof450Mg/kg.
5.1.2 Although below the 2LAET (second lowest Apparent Effects Threshold) of 1000 fig/kg,
the proposed cleanup level of 450 ng/kg is significantly higher than the LAET of 130 ng/kg,
which is the target sediment quality value set by the State of Washington Sediment
Management Standards. As stated in previous comments, the proposed cleanup level of
450 pg/kg is also significantly higher than the median effects range (ER-M) concentration of
180 ng/kg. The ER-M is calculated from a national database ofbenthic invertebrate studies
where more than half of the studies showed adverse effects to benthic invertebrates. (24)
Response: Even if this decision were being made under the State of Washington
Sediment Management Standards (SMS), meeting the lowest apparent effects threshold
(LAET) of 130 /ig/kg PCBs is not an absolute requirement under the State law. EPA
and the State have the discretion to chose cleanup values that fall between the range of
cleanup values set in the SMS, based on several factors, including cost-effectiveness.
The Washington Department of Ecology concurs with EPA's modification of the PCB
cleanup level. See EPA's response to Comment 1.6.1 for a discussion of the ER-M,
and to 2.1.1 for further discussion of the SMS.
5.1.3 EPA's risk assessment presents contrary views in the risk assessment uncertainties
section in the ESD, which states: "The assumption used in the ecological risk evaluation will
overestimate the exposure of species or individuals with large foraging ranges (such as
migratory birds) and may underestimate the exposure of resident species that preferentially
feed at a specific location. " How was it determined that this methodology is therefore both
"appropriate" and protective of human health and the environment when in the uncertainties
52
-------
CB/NT PCB Responsiveness Summary
section it clearly states-the proposed action may not be protective for resident fish and wildlife
species. (35)
Response: EPA believes that its ecological risk evaluation (Western, 1997a) is
conservative with respect to protection of ecological receptors, including resident
species. While the CB/NT Site-wide risk estimate may underestimate risks to resident
species because PCB concentrations are averaged over the entire Site, the Waterway-
specific risk assessments for the Hylebos and Thea Foss waterways provide
conservative estimates of risks to resident species which preferentially feed in a
particular Waterway. The referenced statement has been clarified in the final ESD.
See responses to Comments 4.5.2 and 4.5.4 for further discussion.
5.1.4 The ecological risk evaluation presented in the Addendum (Weston. 1997a) states that
there will be an increased risk to fish and birds as a result of weakening the cleanup
standards. The hazard index used to determine risk to these creatures shows that young
salmon and fish-eating birds will be exposed to an unacceptable level ofPCBs at the proposed
450 ng/kg cleanup level. (8)
Response: The analysis in the 1997 Weston report shows that there is little difference
in the calculated risks at the three cleanup levels examined (150 to 450 Mg/kg PCBs).
All of the cleanup levels examined represented a safe level for fish and birds (i.e.,
hazard quotients are at or near 1.) (It should be noted that EPA did not use the eagle
analysis in the Weston report or the FWS letter for decision-making purposes due to
uncertainties associated with the biomagnification factor and the lack of data on the
foraging range of eagles in the area.) In addition, cleanup of other contaminants of
concern and natural recovery processes will further reduce the risks to these receptors.
5.7.5 Even under this averaging approach, the 450 ng/kg cleanup level will not be protective
offish and birds. The Record of Decision states that a 10-year goal of 150 t*g/kg will result in
an average concentration of 30 t*g/kg in the bay and in the Hylebos Waterway (ROD, Section
7.1.4.). In a recent public testimony, the US Fish and Wildlife Service stated that a
strengthening of the PCB cleanup standard is what is needed to protect fish-eating birds.
However, under EPA's 450 (j.g/kg proposal, in ten years the average for the bay will be 63
ng/kg (no information was given about the Hylebos Waterway). The 450 fj.g/kg proposal will
result in over twice the amount of PCBs in the bay. (8, 26, 36, 3)
Response: EPA based its evaluation of ecological risks presented in the Addendum
(Weston, 1997a) on reasonable maximum exposures to ecological receptors. As
discussed in response to Comment 4.5.2, EPA believes an adequate degree of
conservatism was applied and that the proposed PCB SRAL of 450 Mg/kg is protective
of ecological receptors at the CB/NT Site.
Additionally, the residual sediment PCB concentration after cleanup to 450 ,ug/kg will
be further reduced over time through natural recovery. EPA has added a requirement
53
-------
CB/NT PCB Responsiveness Summary
in the final BSD that PCB concentrations must be reduced to a minimum of 300 ,ug/kg
within 10 years after remedial action, to ensure that PCB concentrations will be
reduced over time. Additional cleanup will be done if PCB concentrations are not
reduced to 300 ^g/kg after 10 years.
5. 1. 6 EPA 's own guidance calls for a cancer risk rate of no more than 100 in 1, 000. 000 yet
EPA 's cancer risk assessment predicts a cancer risk rate of 140 in 1,000,000 after the
proposed cleanup at 450 parts per billion. EPA has the authority to require a cleanup level
that will at least provide a cancer risk rate of 1 in 1,000,000. This decision barely met the 1 in
10,000 criteria, but EPA is willing to go ahead and judge it acceptable. You have hazard
indices that do not meet the ratio of one, but EPA chooses to ignore those and decided the
ecological effects were acceptable. (8, 26, 36, 3)
Response: To meet EPA's criterion for protectiveness of human health, cleanup
actions must result in risks on the order of 10"4 to 10"6 or lower. EPA policy states that
the upper boundary of the risk range is not a discrete line at 1 x 10"*, and cleanups to
levels slightly greater than 1 x 10"* may be considered acceptable if justified based on
site-specific conditions. For non-cancer health impacts, EPA's Superfund regulations
do not contain a numeric standard, but state that EPA must achieve cleanup levels to
which humans may be exposed without adverse effects during a lifetime or part of a
lifetime, incorporating an adequate margin of safety. Cleanups must also be protective
of ecological receptors. EPA believes this decision meets these standards, especially
due to the conservative assumptions built into the human health and ecological risk
assessments. Furthermore, risks will be further reduced over time due to natural
recovery processes.
5. 1. 7 The Environmental Protection Agency is supposed to do just that: protect the
environment. EPA 's current proposal will not provide the best protection of the environment
or to the people of the Commencement Bay community. EPA 's position is unacceptable
because the stakes are simply too high. The viability of wildlife populations, the survival of
economies and cultures which refy upon them, and the health and intelligence of our children
are all at stake. (36, 8)
Response: EPA's primary criterion in any Superfund cleanup decision is protection of
human health and the environment. EPA's 300 /zg/kg PCB SQO meets that standard.
as did the 150 ^g/kg PCB SQO in the 1989 ROD. Either level results in a major
decrease in the human health and ecological risks currently present at the site
immediately after cleanup, with further reductions over time due to natural recovery.
In fact, EPA's decision here will result in the most stringent PCB sediment cleanup of
any appreciable size anywhere in the United States. Most Superfund sediment cleanups
elsewhere have set PCB cleanup standards of 1 ,000 Mg/kg or higher. Commencement
Bay will be protected to a greater degree than any other major Superfund site in the
country.
54
-------
CD/NT PCB Responsiveness Summary
5.1.8 The agency fails to recognize that cleanup issues are matters of civil rights. EPA
weighs the costs of cleanup born by those who made the mess to begin with against the costs to
be born by others. Instead of standing up for a clean environment for all, it treats polluters
and their victims as equal "stakeholders," and it reopens a remedial action plan at the request
of polluters to save them money. Trampled in the process is a simple justice—the right each
person has to a clean environment. The civil rights issue is heightened by the fact that many
Native Americans, Asian Americans and low-income people eat comparatively large quantities
offish each year, thereby suffering disproportionate exposures. EPA's oft stated commitment
to environmental justice rings hollow when it ignores the environmental racism and injustice
inherent in a recommendation to weaken cleanup standards. (36)
Response: EPA must make cleanup decisions that are protective of human health and
the environment. EPA strives to protect human health by identifying those individuals
among all affected people at the Site who have some of the highest current or potential
future exposures to contaminants at a site. In the vicinity of the CB/NT Site, EPA
acknowledges that there are some populations who are likely to be exposed to greater
than average amounts of CB/NT PCBs because they consume greater than average
amounts of CB/NT fish. In the Addendum (Weston, 1997a), EPA considers risks to
high-end tribal fishers to represent a reasonable maximum exposure scenario on which
to base its revaluation of the PCB cleanup level.
EPA does not believe environmental justice is incompatible with its commitment to
selecting cost-effective cleanup solutions for Superfund sites. EPA's 450 Mg/kg PCB
SRAL was selected to be protective of who may rely on Commencement Bay fish for
an important portion of their diet, and also allows for a cost-effective cleanup.
5.2 Comments related to the application of balancing criteria
5.2.1 Six Superfund projects have completed remedial dredging to-date (Waukegan Harbor,
Sheboygan River, GM Central Foundry, Bayou Bonfouca, New Bedford Harbor, and
Marathon Battery.) Volumes of sediment removed ranged from 3000 cy to 159,000 cy.
Overall costs ranged from $140 to $1430 per cubic yard, with an average of about $700 per
cy. For three of the six projects, the cost does not include final disposal, which hasn 't
occurred at the Sheboygan River, GM Central Foundry, or New Bedford Harbor. How does
EPA justify estimating an average cost as low as $73 per cy (Table 6, $18 million) ? (1)
Response: EPA used a cost estimate of $35 per cubic yard for dredging and disposal
of contaminated sediments. This estimate is based on a June 26, 1996, report prepared
by Hartman and Associates and other consultants to the Hylebos Cleanup Committee
entitled "Hylebos Waterway Pre-Remedial Design Preliminary Disposal Site
Evaluation." The report was reviewed by EPA, the U.S. Army Corps of Engineers,
and EPA's contractor, Roy F. Weston, Inc. This estimate is also based on experience
with the Sitcum Waterway sediment remediation project, which was completed in
1994.
55
-------
CB/NT PCB Responsiveness Summary
The six projects cited have significantly different situations and features than the
Hylebos Waterway, or other problem areas within the CB/NT Site. Costs associated
with dredging projects are extremely equipment- and location-specific, making it very
difficult to reasonably compare costs among different projects unless the specific
project requirements and features are compared. PCB contamination levels for the
referenced projects are significantly higher than at the CB/NT Site. Many of the
projects listed in the comment have proposed some sort of sediment treatment option,
either incineration, low temperature thermal desorption, or fixation/stabilization; and
confinement of some type, generally using upland disposal methods. All of these
remedial options require operational, equipment and handling methods which add
significant costs to these projects, compared to the CB/NT Site. In addition, the New
Bedford project included items such as water treatment, and the Sheboygan project
included upland sediment storage, armoring and stabilization, none of which is
included in the CB/NT cleanup plan. The CB/NT cleanup plan includes capping or
dredging and disposal of contaminated sediment in an upland or aquatic disposal
facility. Because contaminant concentrations are low compared to other contaminated
sites, it is not anticipated that specialized equipment, sediment treatment, water
treatment, or special handling will be needed to protect water quality during cleanup.
These items justify a significantly reduced unit price estimated for the CB/NT project,
compared to the other projects.
5.2.2 EPA fs Fact Sheet states that long-term effectiveness of the cleanup, reduction of
toxicity, and short-term effectiveness would be affected by the proposed modification. The
modification would increase risks to the marine environment, wildlife, and public health,
especially through exposure pathways associated with remaining concentrated "hot spots" of
450 ngfkg PCB. In addition, Tribal and community opposition has been documented. In
exchange for these impacts, cost would be reduced. We believe that this is an unbalanced and
unacceptable exchange, and that consideration of all the remedy selection criteria do not
support the proposed modification and that cost considerations discussed in the Fact Sheet are
inappropriate in this case, given the delays in implementing the 1989 ROD cleanup plan. (28,
3, 8, 9. 36)
Response: EPA agrees that if increasing the PCB cleanup level had a significant
impact on toxicity and short- and long-term effectiveness, a modification solely based
on cost would not be justified. The PCB cleanup level was changed only because
EPA's analysis showed that the reduction in toxicity and long-term effectiveness and
any increase in risks to the marine environment, wildlife, and the public health will be
small. In the analysis of short-term effectiveness, the 450 Mg/kg SRAL ranks higher
than the original cleanup standard because short-term impacts to the waterway during
dredging are reduced as the volume of sediments to be remediated are reduced.
5.2.3 We are concerned with how the nine different and often conflicting criteria the EPA
uses were each weighed and considered during the decision-making process. Attempts to avoid
costs at this stage of the cleanup by implementing a less stringent cleanup effort may cost the
56
-------
CB/NT PCB Responsiveness Summary
public a greater expense at a later date. Specifically, these long term costs to our natural
resources from chemical contamination would likety include: diminished returns of healthy
salmon and other fish species, and decreases in breeding bird populations in the Bay. (35)
Response: The nine criteria EPA uses to evaluate and select Superfund cleanup actions
are not weighted equally. The first two criteria, protection of human health and the
environment and compliance with applicable, or relevant and appropriate requirements,
are threshold criteria which must be met under all circumstances. CERCLA has some
provisions for waivers of other laws, but includes no waivers for the requirement of
protectiveness. The next five criteria: long-term effectiveness and permanence;
reduction of toxicity, mobility, or volume through treatment; short-term effectiveness;
implementability; and cost, are considered balancing criteria. EPA strives to achieve
the best balance of these criteria in remedy selection. The last two criteria, state and
community acceptance, are modifying criteria. EPA may modify its decision based on
these criteria.
With regard to the ecological concerns mentioned in the comment, EPA must meet the
threshold criterion of protection of humans and wildlife in selection of the PCB cleanup
level. When EPA's analysis shows, as it did in this case, that a range of PCB cleanup
levels would meet the criterion for protectiveness, we use the other criteria to help
select the cleanup level that provides the best balance of the remedy selection criteria.
5.3 Comments related to natural recovery
5.5. / What is the mechanism that will cause the natural recovery? Is it due to gradual burial
with clean sediments, in which case onfy the PCB availability at the sediment surface is being
reduced? Is it due to dechlorination? If yes, what rate of dechlorination has been assumed?
Is it due to a continuing PCB flux from the sediment into the water column and subsequent
transport out of the waterway? (1)
Response: Natural recovery of PCBs at the CB/NT Site '."ill occur primarily because
of burial by clean sediments (sources include the Puyallup River sediment load and
small streams entering the waterways). Dechlorination processes are extremely slow
and sediment tlux to water is negligible because of the binding of PCBs to sediment
carbon. In assessing the potential for natural recovery of PCBs at the CB/NT Site,
EPA assumed there would be no biological or chemical degradation of PCBs, and that
all natural recovery would be due to burial and mixing with clean sediments.
Natural recovery is determined for surface sediment as represented by the biologically
active zone (top 10 cm). This is the stratum where most of the sediment-dwelling
organisms that serve as prey to fish live. Receptors of concern are not exposed to
deeper sediment.
57
-------
CB/NT PCS Responsiveness Summary
5.3.2 Use of the 198$ standards will allow for corrective measures and, based on monitoring
to ensure that natural recovery is occurring, will make sure we collect all of the PCBs still
trickling into the sound. (3)
Response: Regardless of which PCB cleanup level is chosen, long-monitoring of
sediments will occur following cleanup to ensure that contaminant concentrations
remain at acceptable levels. In addition, EPA has added a requirement to the BSD that
PCB concentrations must be reduced to 300 ^g/kg within ten years of the remedial
action. This means that long-term monitoring of PCB-contaminated sediments will be
required to ensure that this standard is met. Finally, control of ongoing sources of
contamination is required under the 1989 ROD, and will continue to be an integral pan
of the CB/NT cleanup.
5.3.3 Since the proposed 450 t*g/kg cleanup level is based on no natural recovery, EPA
should not use natural recovery to assess the level of protectiveness that will be achieved by
that cleanup level. (8)
Response: EPA believes it is appropriate to use natural recovery to assess the
protectiveness of the remedy, because we have added natural recovery as a required
element of the remedy in the BSD. In response to public comments, and those of the
State, EPA has modified the remedy as proposed in the March, 1997 draft ESD (EPA,
1997a) to include a PCB cleanup standard for natural recovery. If PCB concentrations
in sediments do not naturally recover to at least 300 Atg/kg within 10 years after the
remedial action, EPA will take additional measures to achieve this concentration.
Natural recovery modeling done to date indicates that, after source control is complete
and sediments are remediated to 450 ptg/kg PCBs, sediment PCB concentrations will be
reduced to even lower than 300 /zg/kg within 10 years.
5.3.4 The ROD provides natural recovery rate estimates for the Hylebos. The HCC is
required by the AOC to develop an updated natural recovery rate analysis for the Hylebos.
The natural recovery rate can be used to show areas of the Hylebos that will recover to below
the 450 ng/kg PCB cleanup level in ten years. EPA should not conclude that natural recovery
to 450 ng/kg is not appropriate for PCBs prior to reviewing the natural recovery analysis that
EPA required the HCC to conduct under the AOC. (17)
Response: EPA agrees that natural recovery should be included as part of the remedy,
but disagrees that 450 ^g/kg should be the 10-year standard. EPA's SRAL of
450 Mg/kg PCBs is at the high end of the risk range considered acceptable to EPA. If
EPA allowed for natural recovery to 450 Mg/kg, PCB concentrations in the interim 10-
year period would be unacceptably high. EPA did consider natural recovery in its
original proposal, in the sense that we acknowledged that one of the reasons it was
acceptable to select a cleanup level at the high end of EPA's risk range, was that risks
would be reduced over time through natural recovery. In the final ESD, EPA has
added a requirement that sediments must naturally recover to 300 Mg/kg PCBs within
58
-------
CB/NT PCB Responsiveness Summary
10 years after cleanup. This will ensure that if sediments do not naturally recover,
additional cleanup work will be done to ensure that a 300 Mg/kg PCB cleanup level is
achieved.
5.3.5 Weston states that natural recovery will farther reduce PCB concentrations in the
sediment, but does not address the uncertainty associated with natural recovery itself. (35)
Response: EPA agrees that there is uncertainty associated with natural recovery
estimates. A complete analysis of natural recovery potential at the CB/NT Site was not
conducted for the revaluation of the PCB cleanup level. Instead, EPA reiied upon the
natural recovery estimates in the 1989 CB/NT ROD.
In the revised ESD, EPA has added a requirement that sediments must naturally
recover to at least 300 Mg/kg PCBs. Even though EPA has not attempted to quantify
uncertainties associated with natural recovery estimates, we believe an estimate of
natural recovery to 300 Mg/kg is conservative because it falls at the high end of the
range of estimated natural recovery rates for PCBs in the Hylebos Waterway. Using
natural recovery rates in the ROD, sediments are predicted to naturally recover to 280
to 225 A*g/kg PCBs following cleanup to 450 Mg/kg. Additional natural recovery
modeling will be required as part of pre-design work to verify the estimates in the 1989
ROD.
5.4 Comments related to current contamination and source control
5.4.1 EPA has failed to produce results on controlling ongoing sources of PCBs in
Commencement Bay showing an even more feeble approach to PCB cleanup. The National
Marine Fisheries Service study states that the health of the bay has not improved in the last ten
years. Can EPA demonstrate that the next ten years will be any better? (8)
Response: Source control activities have been ongoing at the CB/NT site since the
Record of Decision was signed in 1989. Ecology is responsible for working with
individual facilities to achieve source control. For the Hylebos Waterway, Ecology has
inspected or investigated 141 potential sources of contamination and have identified 26
ongoing sources. Ecology has worked with each of these facilities to control their
sources, either through upland cleanups or implementation of best management
practices. Of the 26 ongoing sources identified by Ecology, 21 have been controlled as
of July 1997. Ecology plans to have all ongoing sources of contamination controlled
(except intertidal areas, as discussed below), by mid-1998.
One of the significant sources of contamination remaining in the Hylebos Waterway is
the intertidal sediments. These sediments contain the highest concentrations of PCBs
and other chemicals found in the-waterway. These intenidal sediments will act as
sources of contamination to the rest of the waterway until they are remediated.
Remediation of intertidal areas will occur either as part of Ecology cleanups or as part
59
-------
CB/NTPCB Responsiveness Summary
of the overall waterway cleanup. Additionally, subtidal sediments with high
concentrations of PCBs act as continuing sources of PCBs through resuspension of
sediments and movement to other parts of the waterways. However, after cleanup of
areas where sediments exceed EPA's cleanup levels for PCBs and other contaminants,
these sources of sediment recontamination will be removed, allowing for natural
recovery to occur.
The NMFS study cited in the comment addresses current contamination in the Hylebos
Waterway. See response to Comment 5.4 2 for a discussion of the results of this
study.
5.4.2 When recent toxicopathic results are compared to previous work done as part of the
RI/FS, it is evident that conditions for flatfish (based on lesion prevalence and biochemical
measures of early response to contaminant exposure) in the waterway have not measurably
improved. This fact is significant as EPA is espousing a cleanup level three times higher than
that originally stated in the ROD with the supporting rationale that natural recovery would be
expected to occur (allegedly decreasing the concentration of PCS's in half over a 10 year
period). The effects of natural recovery, since the RI data was collected (more than 10 years
ago) have not been observed. (24, 8, 36)
Response: The NMFS study cited in the comment evaluates current conditions in the
Hylebos Waterway. EPA agrees that those and other studies show that natural
recovery has not significantly reduced PCB concentrations or toxic effects in the
Hylebos Waterway in the last 10 years. However, these studies are not a good
predictor of conditions in the Waterway in the 10 years after completion of the cleanup.
Many of the source control activities required by the ROD have been completed in the
last three to four years, and some will be completed over the next year. As noted in
the response to Comment 5.4.1, PCB-contaminated intertidal sediments, one of the
major sources of PCB contamination remaining in the waterway, have not yet been
remediated. Because source control is not complete and sediments containing high
concentrations of PCBs have not been remediated, no appreciable natural recovery of
PCBs has occurred to date in the Hylebos Waterway.
Once source control and sediment remediation is complete, conditions in the Hylebos
Waterway will change substantially. The highly contaminated sediments which are
currently subject to erosion and movement to other areas of the waterway will be
removed and replaced by clean sediments. These clean sediments will continue to
move about the waterway and further reduce contaminant concentrations in areas where
low levels of PCBs were left in place, leading to an overall reduction of contaminant
concentrations in the waterway. In addition, clean sediments from other sources such
as Hylebos Creek and the Puyallup River will continue to enter the waterway and
further reduce contaminant concentrations through burial and mixing. For these
reasons, EPA believes there will be natural recovery of PCB-contaminated sediments
60
-------
CB/NT PCB Responsiveness Summary
after completion of the cleanup, even though it has not occurred in the 10 years since
the ROD was signed.
5.4.3 A recent study by the National Marine Fisheries Service in the Hylebos Waterway of
Commencement Bay revealed elevated levels ofPCBs in fish and links the presence ofPCBs to
altered sexual reproduction of flatfish. Fish-eating birds, such as bald eagles and blue herons,
in the Commencement Bay area have high levels ofPCBs in their embryos. The herons from
Dumas Bay use Commencement Bay as their feeding grounds. Seven eggs were collected from
a colony of great blue herons in Commencement Bay in 1988. Chemical analysis revealed that
PCB concentrations surpassed the threshold level for adverse impacts. Last year, 5 more eggs
were collected from the Dumas Bay great blue heron colony, and the levels ofPCBs in some of
the eggs were still at levels where FWS would anticipate adverse affects. Recently, we have
also determined that a significant portion of birds from this colony feed in Commencement Bay
and its waterways. (35, 8, 26, 36)
Response: Please see responses to Comments 5.4.2 and 4.4.6.
5.4.4 Hasn 't scientific testing shown this area to be polluted and one of the most toxic spots
for the last 15 years? (9)
Response: EPA agrees that concentrations of chemical contaminants are unacceptably
high at the CB/NT Site and present unacceptable risks to human health and the
environment. This assessment is the basis of EPA's 1989 ROD which requires source
control and sediment cleanups at eight contaminated sediment problem areas in the
Commencement Bay tideflats.
5.5 Comments related to changing the ROD
5.5.1 In other EPA regions, federal courts, at the behest of EPA, give great credence to ROD
numbers and the idea they should be frozen in time, otherwise a plethora of other values could
be proposed and the remedial process could be thwarted. It is interesting that when
potentially responsible party (PRP)-sponsored cleanup numbers come along later which are
less restrictive, this theory seems to be quickly abandoned. (11)
Response: EPA has a policy of not reopening Records of Decision to incorporate
regulations passed after the ROD is signed, unless necessary to ensure protectiveness.
because of the potential for long delays in implementing cleanups if the basis of a
remedy had to be continually reevaluated during the design and implementation phases.
However, this does not mean that EPA will ignore new information which indicates
that the remedy may not be cost-effective when coupled with evidence that it may be
possible to dredge a smaller volume of sediments with no appreciable change in the
protectiveness of the remedy.
61
-------
CB/NT PCB Responsiveness Summary
EPA has also championed the policy of reevaluating old remedies when new
information indicates that there may be a more cost-effective way to implement the
cleanup. EPA will also reevaluate remedies when new information indicates that a
remedy is not sufficiently protective. This is true for Superfund-financed as well as
potentially-responsible-party-financed cleanups.
In addition, the new PCB cleanup number is not a PRP-sponsored number. See
comments in Section 3.2 for information about PCB cleanup levels preferred by PRPs.
5.5.2 The proposed modification, requested by several potentially responsible ponies, sends
the wrong message — if you delay cleanup long enough, you can weaken standards and cut
your liability. (28)
Response: EPA agrees that Superfund cleanups should be implemented as quickly as
possible for a variety of reasons, the most important of which is to remove
contamination from the environment as quickly as possible. Regardless of the length of
time between development and implementation of a cleanup plan, EPA believes that it
is appropriate to reevaluate a Superfund remedy if new information indicates that the
original decision may be significantly flawed. This is true whether the remedy is being
paid for through Superfund monies or if it is privately funded.
5.6 Comments related to the role of parties outside EPA
5.6. / Ecology must abide by the 1989 ROD which established 150 ng/kg as the allowable
level ofPCBs in sediment. Will this mean a dual standard for cleanup of toxic sites in
Washington State? Will a bill be submitted to the legislature demanding that state standards
cannot exceed federal standards in any case? (3)
Response: The 1989 ROD for the CB/NT Site is an EPA document, with concurrence
by Ecology. Ecology has also concurred with EPA's decision to modify the ROD by
issuing this ESD. The PCB cleanup standards set by this ESD of 450 /zg/kg
immediately after cleanup and 300 Mg/kg within 10 years after cleanup were developed
based on the site-specific circumstances at the CB/NT Superfund site, and applies only
to the cleanup of that site. Ecology is not required to follow the CB/NT ROD or this
ESD for setting sediment cleanup levels elsewhere in the state. Ecology has not yet set
numeric standards for PCBs in sediments for protection of human health and may set
standards that are more or less stringent that the standard set for the CB/NT Superfund
site.
5.6.2 Outside of the ecological concerns I also wondered if the state health department
played any role in the PCB decisions? I am unable to find any record of their involvement or
concurrence. (11)
62
-------
Cfl/AT PCB Responsiveness Summary
Response: The Washington State Department of Health (WDOH) was notified that
EPA was considering modifying the PCB cleanup level for the CB/NT Superfund site
in February 1996, and was given the opportunity to comment on Weston's first draft
evaluation of potential alternative PCB cleanup levels. WDOH's comment letter is in
the Administrative Record for this ESD.
5.6.3 The businesses and the Port of Tacoma which stand to save $13 million dollars from
EPA's proposal underestimate the long term impacts a decision in their favor would cause.
(8)
Response: See response to Comment 5.2.2.
5.7 Comments related to the remedial action
5.7.1 We have serious reservations as to whether a remedial program for achieving the
targeted cleanup level is either (I) feasible or (2) is at all predictable with regard to ultimate
schedule and cost. No dredge or combination of dredges can consistently achieve a cleanup
level of even several ppm. Please clarify EPA's position in this regard. (1)
Response: In general, the achievement of the cleanup level via dredging is not
predominantly a function of dredging technology itself but more a function of how
much contaminated material exists and how much requires removal to achieve the
cleanup objective. In most areas of Commencement Bay, sediment contamination is
associated with overlying unconsolidated sediments. The underlying native sediments
are, with a few exceptions, relatively free of contamination. Sediment cores have been
used to identify the depth of contaminated material, which in many areas is only a few
feet deep. EPA's experience with the St. Paul and Sitcum Waterway cleanups in
Commencement Bay shows that cleanup to the Sediment Quality Objectives in the ROD
is feasible, and with sufficient sampling, is predictable with regard to schedule and
cost.
5.7.2 The more than half century of industrial use of the Hylebos Waterway must have
resulted in wide-spread and sizable quantities of debris disturbed and embedded on the bottom.
Debris is detrimental to the efficient and continuous operation of a dredge. Debris, or rocks,
was a serious problem at least four of the six Superfund sites mentioned in Comment 5.2.1.
Has EPA assessed the debris problem ? How would this problem be managed and overcome
during the proposed dredging process? What would the disposition of the debris be? Have the
substantial inefficiencies due to debris interferences, and management and disposal of debris.
been included in the cost estimate? (1)
Response: During the initial investigations of the Hylebos Waterway, a side-scan
sonar survey was conducted to identify the quantity and location of subsurface debris.
This survey showed that there is some metal debris and some sunken logs in the
Hylebos Waterway, but the extent of debris was not as great as has been seen at other
63
-------
CB/NT PCB Responsiveness Summary
Puget Sound dredging projects. A similar survey has not yet been conducted for the
Thea Foss Waterway, but we have no reason to believe there would be more debris at
Thea Foss than at Hylebos Waterway.
Dredging plans have not yet been developed for either waterway. These plans will
include plans for handling and disposition of debris. The presence of debris is one of
the factors to be used in determining the dredging method, as use of a clamshell dredge
rather than a hydraulic dredge will minimize difficulties in handling debris. Regardless
of the selected dredging method, the debris will most likely be handled and disposed of
separately. It is not anticipated that debris handling will significantly impact the cost of
the remedy.
5.7.3 Is capping a candidate remedial technology for the Hylebos Waterway, or is the
reference to capping inadvertent? Please explain the status of a capping remedy. (1)
Response: Capping is one of the sediment confinement options selected in the 1989
ROD. However, capping is only appropriate in areas where there are no or limited
navigational constraints and where the cap is not likely to be disturbed through erosion,
scour, or future dredging. Capping is precluded in much of the Hylebos Waterway due
to these constraints. However, it is being considered for some of the intertidal areas in
the Hylebos Waterway and portions of the Thea Foss Waterway.
5.7.4 The ESD acknowledges the benefit of having only one disposal site, rather than
multiple disposal sites. One disposal site decreases the amount of area initially disturbed,
decreases monitoring efforts, and also may decrease the likelihood of breach of containment.
As pointed out in the "Explanation of Significant Differences...", pages 15 and 17, a cleanup
volume greater than 700,000 cubic yards (cy) may require multiple disposal sites. Since the
proposed cleanup level of 450 ng/kg would result in the removal of approximately 508,000 cy
of contaminated sediments and the original cleanup level (300 pg/kg) would result in
approximately 891,000 cy, it may be best to extrapolate the cleanup level based on 700,000 cy
of contaminated sediment. This approach would provide a cleanup level that is more
protective than the proposed 450 ng/kg concentration and yet should not increase the technical
feasibility of the cleanup effort. (34)
Response: Backcalculating a cleanup level based on volume limits is an interesting
idea; however, it would require a fixed volume above which we will need multiple
disposal sites. Since a disposal site has neither been selected nor designed for the
remaining cleanups, at this time we only have estimates of the capacities of various
disposal sites which are under consideration. The figure of 700,000 cubic yards
mentioned in the ESD is merely an estimate of a volume over which multiple disposal
sites are likely to be needed, because most of the sites under consideration have
capacities of less than 700,000 cubic yards. Waiting for selection and design of a
disposal site, then calculating a PCB cleanup level based on the capacity of that site.
would cause a long delay in the cleanup.
64
-------
CB/NT PCB Responsiveness Summary
5.8 Other Comments
5.8.1 As a general concern, the documents released for public review did not address the
EPA's authority to issue an explanation of significant difference. We would appreciate
information on the regulatory framework for which an explanation of significance is allowed
and appropriate. (8)
Response: The implementing regulations for the Superfund law are contained in the
NCP, which is codified at 40 CFR Part 300. Section 300.435(c) of the NCP sets out
the procedures EPA must follow if, after the ROD is adopted, new information
developed during the remedial design or remedial action, or an enforcement action, or
consent decree, results in the remedial action differing significantly from the remedy
selected in the ROD with respect to scope, performance, or cost. The NCP requires
that a summary of the explanation of significant differences be published in a
newspaper of general circulation, and that the explanation of significant differences and
supporting documentation be placed in the Administrative Record for public review.
Significant changes to a remedy are generally changes to a component of the remedy
that do not fundamentally alter the overall remedial approach. Fundamental alterations
to the ROD are considered changes which make the remedy no longer reflective of the
selected remedy and, therefore, require an amendment to the ROD.
In this case, information collected during pre-design sampling indicated that the volume
of sediments required to cleanup up to the PCB cleanup level in the ROD was
significantly larger than originally anticipated. The ROD estimated that remediation of
448,000 cubic yards of contaminated sediments would be required in the Hylebos
Waterway at an estimated cost of $13.8 million. Current estimates place this volume at
891,000 cubic yards and the cost at $31 million, mainly due to new estimates of the
extent of PCB contamination. This, in itself, is a significant change in the scope,
performance, and cost of the ROD. With this ESD, the amount of sediment requiring
remediation will change to 508,000 cubic yards, at a cost of $18 million.
The revised PCB cleanup level significantly changes a performance standard
established in the ROD, but does not significantly change the volume and cost of the
cleanup as anticipated in the 1989 ROD. The remedial approach remains the same as
outlined in the ROD (i.e., confinement of contaminated sediments that will not
naturally attenuate below the cleanup levels within 10 years). Additionally, in this
case, EPA chose to go beyond what the regulations require for public review of an
ESD by informing the public of our proposed PCB cleanup level and taking public
comment on that proposal before making a final decision.
5.5.2 There is a connection between what you require in cleanup and pollution prevention. If
you send the message with this cleanup that they can leave more there, you are sending a
message to everyone that continued pollution — you may not be held accountable for continued
pollution. (36)
65
-------
CB/NT PCB Responsiveness Summary
Response: EPA agrees that the Superfund law and its requirements for cleanup of
contamination send a strong message that pollution carries with it a high cost. With the
PCB SRAL set at 450 Mg/kg, PRPs will be required to dredge approximately 500,000
cubic yards of sediments at a cost of $18 million for the Hylebos Waterway. Polluters
are still getting a message that there is a significant cost associated with pollution of the
CB/NT Superfund site.
5.8.3 lam also concerned as to how these levels may be used in negotiations with potentially
responsible parties to lessen the footprint of contaminated sites, thereby reducing damage
assessment calculations. (11)
Response: Natural Resource Trustees assess damage to natural resources independently
of EPA's efforts to define a cleanup area. EPA's selection of a PCB cleanup level and
associated footprint of a cleanup area should not affect the Trustee's ability to recover
costs for damage to natural resources.
5.8.4 This action by EPA appears to be cost driven, with little publicity, an abbreviated
response time, and limited participation. Tahoma Audubon Society did not receive a notice
and there was no published notice in the News Tribune, leading me to wonder how wide the
contact was. I believe that one newspaper article appeared on the subject. Finally two weeks
from the public hearing to the response deadline is a very short time for such a complex
subject. (3)
Response: EPA's efforts to inform and involve the public in this decision are described
on page 2 of this responsiveness summary. EPA sent fact sheets to over 2,000 people
to announce a 30-day public comment period on the proposed ESD, which ran from
March 10 through April 9, 1997. On March 10th, the beginning of the comment
period, a display advertisement was placed in the Tacoma News Tribune announcing
EPA's proposal, the comment period, and the date of the public meeting, which was
held on March 26. These efforts are consistent with or greater than EPA's practice at
other Superfund sites.
EPA has added this commentor to our Commencement Bay mailing list. Others who
would like to be on EPA's mailing list to receive fact sheets on the CB/NT Site, are
welcome to contact Jeannie O'Dell at (206) 553-6919 to ask to be added to the mailing
list.
66
-------
CB/NT PCB Responsiveness Summary
Part HI—List of Commentors
HYLEBOS PCB COMMENT SUMMARY RESPONDENTS
Abbreviated
Reference
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Full Reference
Applied Environmental Management. Inc.
Agency for Toxic Substances and Disease Registry
Tahoma Audubon Society
Bay Zinc Company, Inc.
Bonneville Power Administration
Graham & James LLP/Riddell Williams P.S. representing the Buffelen Woodworking
Company
Byrd Real Estate Services, Inc.
Citizens for a Healthy Bay
Individual Citizens
City of Tacoma
Washington Department of Ecology (as Natural Resource Trustee)
Washington State Department of Health
Dunlap Towing Company
Floyde & Snider, Inc.
F.O.F.. Inc.
Gradient Corporation
Hylebos Cleanup Committee
Hylebos Marina
Manke Lumber Company
Mintercreek Development
Modutech Marine. Inc.
Murray Pacific Corporation
Middle Waterway Action Committee
National Oceanic and Atmospheric Administration
PALS Investments
People for Pugct Sound
Petroleum Reclaiming Service, Inc.
Puget Soundkecper Alliance
Puyallup Tribe of Indians
67
-------
CB/NT PCB Responsiveness Summary
HYLEBOS PCB COMMENT SUMMARY RESPONDENTS
Abbreviated
Reference
30
31
32
33
34
35
36
Full Reference
Joseph Simon & Sons
Steich Brothers Machine Works
Bucknell Stenlik representing Taylor Way Properties, Inc.
Tacoma Pierce County Chamber of Commerce
Tacoma-Pierce County Health Department
U.S Fish and Wildlife Service
Washington Toxics Coalition
68
-------
CB/NT PCB Responsiveness Summary
D. REFERENCES
Arnold. D.L., F. Bryce, R. Stapley et al. 1993a. lexicological Consequences of Aroclor
1254 Ingestion by Female Rhesus (Macaco mulatto) Monkeys, Part 1A: Prebreeding
Phase—Clinical Health Findings. Food Chem. Toxicol. 31:799-810.
Arnold. D.L., F. Bryce, K. Karpinski et al. 1993b. Toxicological Consequences of Aroclor
1254 Ingestion by Female Rhesus (Macaco mulatto) Monkeys, Part IB: Prebreeding
Phase—Clinical and Analytical Laboratory Findings. Food Chem. Toxicol. 31:811-824.
EPA (U.S. Environmental Protection Agency). 1997a. Explanation of Significant
Differences—Commencement Bay Nearshore/Tideflats Superfund Site: Public Review Draft.
3 March 1997.
EPA. 1997b. Integrated Risk Information System (IRIS). Office of Research and
Development, National Center for Environmental Assessment. May.
EPA. 1997c. Special Report on the Environmental Endocrine Disruption: An Effects
Assessment and Analysis. Risk Assessment Forum. U.S. EPA, Washington, D.C. February.
EPA/630/R-96/012.
EPA. 1996a. Cancer Dose-Response Assessment and Application to Environmental
Mixtures. National Center for Environmental Assessment. Office of Research and
Development. September. EPA/600/P-96/001F.
EPA. 1996b. EcoUpdate: Ecotox Thresholds. January. EPA/540/F-95/038.
EPA. 1992. Supplemental Guidance to RAGS: Calculating the Concentration Term. Office
of Solid Waste and Emergency Response. 9285.7-081 Washington, D.C. May.
EPA. 1991. Role of the Baseline Risk Assessment in Superfund Remedy Selection Decisions.
Memorandum from D.R. Clay, Assistant Administrator, U.S. EPA, Washington, D.C. Office
of Solid Waste and Emergency Response. Directive 9355.0-30.
EPA. 1989a. Record of Decision—Commencement Bay Nearshore Tideflats. Prepared by
EPA Region 10, Seattle, Washington. September.
EPA. 1989b. Risk Assessment Guidance for Superfund Volume I—Human Health Evaluation
Manual (Pan A)—Interim Final. EPA/540/1-89-002.
EPA. 1986. Guidelines for Health Risk Assessments of Chemical Mixtures.
69
-------
CB/NT PCB Responsiveness Summary
FWS (U.S. Fish and Wildlife Service). 1996a. Correspondence to Allison Hiltner, Remedial
Project Manager, U.S. EPA, Region X, Seattle, WA, regarding sediment PCB cleanup levels;
Commencement Bay Nearshore/Tideflats Superfund site. 19 November.
FWS. 1996b. Correspondence to Allison Hiltner, Remedial Project Manager, U.S. EPA,
Region X, Seattle, WA, regarding additional comments on sediment PCB cleanup levels;
Commencement Bay Nearshore/Tideflats Superfund site. 2 December.
FWS. 1996c. Interim Report. Environmental Contaminants in Bald Eagles Nesting along the
Lower Columbia River. Prepared for the Lower Columbia River Bi-State Water Quality
Program.
FWS. 1995. Biological Opinion Regarding the Effects of the Great Lakes Water Quality
Guidance on the Clubshell (Pleurobema clava), Northern Riffleshell (Epioblasma torulos
rangiana), White Cat's Paw Pearly Mussel (Epioblasma obliquata perbliqud), Bald Eagle
(Haliaeetus leucocephalus), Peregrine Falcon (Falcoperegrinus), and Piping Plover
(Charadrius melodus). Chicago Field Office, U.S. Fish and Wildlife Service.
FWS. 1994. Biological Opinion on the Effects of Concentrations of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin, to be Attained Through Implementation of a Total Maximum
Daily Load, on Bald Eagles along the Columbia River. Letter to Charles Findley, U.S.
Environmental Protection Agency, Region X. 6 January.
Gahler, A.R., J.M. Cummins, J.N. Blazevich, R.H. Rieck, R.L. Arp, C.E. Gangmark,
S.V.W. Pope, and S. Filip. 1982. "Chemical Contaminants in Edible, Non-Salmonid Fsh and
Crabs from Commencement Bay, Washington." Environmental Services Division Laboratory,
Region 10, Seattle, WA. EPA-910/9-82-093. December.
Gradient Corporation. 1995. Risk Assessment for Polychlorinated Biphenyls (PCBs) for the
Hylebos Waterway, Tacoma, Washington. Prepared for the Hylebos Cleanup Committee,
Dalton, Olmsted & Fuglevand, Inc, Bothell, WA, by Gradient Corporation, Cambridge, MA.
September.
Hartman and Associates, et al. 1996. Hylebos Waterway pre-Remedial Design Preliminary
Disposal Site Evaluation. Prepared for the Hylebos Cleanup Committee. 26 June 1996.
Krausmann, J. 1996. U.S. Fish and Wildlife Services. Personal communication with
N. Musgrove. Roy F. Weston, Inc.
Landolt, M.L., F;R. Hafer, A. Nevissi, G. van Belle, K. Van Ness and C. Rockwell. 1985.
Potential Toxicant Exposure Among Consumers of Recreationally Caught Fish from Urban
Embayments of Puget Sound. NOAA Technical Memorandum NOS OMA 23. November.
70
-------
CB/NT PCB Responsiveness Summary
Laws, E. 1995. Land Use in the CERCLA Remedy Selection Process. U.S. EPA Office of
Solid Waste and Emergency Response. 25 May.
Levinskas, G.J., D.P. Martin, H.R. Seibold and J.L. Cicmanec. 1984. Aroclor 1254:
Reproduction Study with Rehesus Monkeys (Macaco mulatto). Unpublished study by
Monsanto.
Long, E.R., D.D. MacDonald, S.L. Smith, and F.D. Calder. 1995. Incidence of Adverse
Biological Effects Within Ranges of Chemical Concentrations in Marine and Estuarine
Sediments. Environmental Management Vol. 19(1): 81-97.
MacDonald, D.D. 1994. Sediment Injury in the Southern California Bight: Review of the
Toxic Effects of DDTs and PCBs in Sediments. In: Southern California Bight Natural
Resource Damage Assessment Experts Report. 220 pp.
O'Neill, S.M, I.E. West, and S. Quinell. 1995. Contaminant Monitoring in Fish: Overview
of the Puget Sound Ambient Montioring Program Fish Task. Puget Sound Research '95
Proceedings. Vol 1. pp. 35-50.
Pierce, D., et al. 1981. Commencement Bay Seafood Consumption Study. Tacoma-Pierce
County Health Department. December.
Tetra Tech, Inc. 1988. Health Risk Assessment of Chemical Contamination in Puget Sound
Seafood. Final Report. TC-3338-28. Prepared for Region X EPA. September.
Tetra Tech, Inc. 1985. Commencement Bay Nearshore/Tideflats Remedial Investigation,
Final Report. TC-3752. August.
Toy, K.A., N.L. Polissar, S. Liao, and G.D. Mittelstaedt. 1996. A Fish Consumption
Survey of the Tulalip and Squaxin Island Tribes of the Puget Sound Region. Tulalip Tribes,
Department of Environment, 7615 Totem Beach Road, Marysville, WA 98271.
Toy, K.A., G.D. Gawne-Mittelstaedt, N.L. Polissar and S. Liao. 1994. A Fish Consumption
Survey of the Tulalip and Squaxin Island Tribes of Puget Sound, Draft. Seattle, WA.
November.
Tryphonas, H., M.I. Luster, G. Schiffman et al. 1991a. Effect of Chronic Exposure of PCB
(Aroclor 1254) on Specific and Nonspecific Immune Parameters in the Rhesus (Macaca
mulatto) Monkey. Fund. Appl. Toxicol. 16(4):773-786.
Tryphonas, H., M.I. Luster, K.L. White et al. 1991b. Effects of PCB (Aroclor 1254) on
Non-Specifc Immune Parameters in Rhesus (Macaca mulatto) Monkeys. Int. J.
Immunopharmacol. 13 -.639-648.
71
-------
CB/NT PCB Responsiveness Summary
Tryphonas, H., S. Hayward, L. O'Grady et al. 1989. Immunotoxicity Studies of PCB
(Aroclor 1254) in the Adult Rhesus (Macaca mulatto) Monlcey—Preliminary Report. Int. J.
Immunopharmacol. 11:199-206.
Versar. 1985. Assessment of Human Health Risk from Ingesting Fish and Crabs from
Commencement Bay. Prepared for Washington State Department of Ecology and U.S.
Environmental Protection Agency Office of Solid Waste and Emergency Response.
EPA/910/9-85/129. April.
WESTON (Roy F. Weston, Inc.). 1997a. Addendum-Evaluation of Residual Risks
Associated with a Range of Sediment PCB Cleanup Levels in the Hylebos Waterway, the Thea
Foss Waterway, and the Overall Commencement Bay Nearshore/Tideflats Superfund Site.
Prepared for U.S. EPA, Region X, Seattle, WA, by Roy F. Weston, Inc. February.
WESTON. 1997b. Potential for Noncancer Health Impacts from PCBs in Sediments at the
Commencement Bay Nearshore/Tideflats Superfund Site. Technical Memorandum prepared
for U.S. EPA, Region X, Seattle, WA, by Roy F. Weston, Inc., Seattle, WA. July.
WESTON (Roy F. Weston, Inc.). 1997c. Risks from Human Consumption fo PCB-
Contaminated Fish under Background Conditions in Puget Sound. Technical Memorandum
prepared for U.S. EPA, Region X, Seattle, WA, by Roy F. Weston, Inc. July.
WESTON. 1996. Evaluation of Residual Risks Associated with a Range of Sediment PCB
Cleanup Levels in the Hylebos Waterway, the Thea Foss Waterway, and the Overall
Commencement Bay Nearshore/Tideflats Superfund Site. Prepared for the U.S. EPA, Region
X, Seattle, WA, by Roy F. Weston, Inc. October.
72
-------
CB/NT PCB Responsiveness Summary
E. GLOSSARY OF TERMS
Apparent Effects Threshold: A chemical concentration above which deleterious biological
effects are predicted to always occur.
Arithmetic mean: In testing, the same as ordinary average determined by adding all the
scores and dividing the sum by the number of scores in the group. The resulting quotient is
the average or mean.
Aroclors: Industrial mixtures of polychlorinated biphenyl compounds.
Bathymetry: A topographic map of the bed of the ocean.
Biomagnification factor (BMP): A factor describing the degree of accumulation of an
organic chemical taken up from an environmental medium in a living organism.
Biota-sediment accumulation factor (BSAF): A factor describing the rate of transfer of an
organic chemical from sediment to an organism.
Cancer slope factor (CSF): A term used to describe the adverse cancer effects of a
contaminant. CSFs are measured by the probability of a person developing cancer over a
lifetime.
Endocrine disrupter: An agent that interferes in some way with the natural hormones in the
b°dy (e.g., those secreted by the pituitary, thyroid, pancreas, adrenal, testes and ovaries). An
agent might disrupt the endocrine system by affecting any of the various stages of hormone
production and activity.
Geometric mean: That value obtained by multiplying all the items of a series together and
extracting the nth root of this product, where n is the number of items.
Hazard quotient: A ratio comparing a chemical concentration in an environmental medium
(e.g., sediment) to a toxicity threshold value or criterion. A result greater than 1 indicates an
exceedance of the threshold or criterion.
K,w: A laboratory measure of the partitioning of a chemical between n-octanol and water. It
provides a measure of the water solubility of a chemical.
Lognormal: A statistical distribution of an environmental variable (e.g., chemical
concentration) in which lower values are more frequently encountered. The central tendency
(or middle) of the distribution tends to be lower than that estimated by an arithmetic mean and
is often estimated by the geometric mean.
73
-------
CB/NT PCB Responsiveness Summary
Microtox: A commercial bioassay test measuring changes in bioluminescence in a bacterium
in response to contaminant exposure.
Organic carbon normalized: Expression of a dry-weight chemical concentration in terms of
parts of organic carbon. Derived by dividing a chemical concentration by the decimal fraction
of organic carbon in the sediment.
Parts per Billion (ppb): Parts of a chemical (e.g. , PCBs) per billion parts of an
environmental medium (e.g.. fish tissue). May be expressed as
Parts per Million (ppm): Parts of a chemical (e.g. , PCBs) per million parts of an
environmental medium (e.g., fish tissue). May be expressed as mg/kg.
Pathogens: Microorganisms that can cause disease in other organisms or in humans, animals
and plants (e.g., bacteria, viruses, or parasites) found in sewage, in runoff from farms or rural
areas populated with domestic and wild animals, and in water used for swimming. Fish and
shellfish contaminated by pathogens, or the contaminated water itself, can cause serious
illness.
Potentially responsible parties (PRPs): Any individual or company— including owners,
operators, transporters or generators-potentially responsible for, or contributing to, a spill or
other contamination at a Superfund site.
Reasonable maximum exposure (RME): The highest exposure reasonably expected to occur
in a population.
Reference dose (RiD): The concentration of a chemical known to cause health problems; also
can be referred to as the ADI, or acceptable daily intake.
Total organic carbon: Carbon derived from a living or decomposing material that has been
incorporated into a sediment matrix.
Trophic level: A step along the food chain from numerous small organisms to decreased
numbers of large organisms.
: Micrograms of a chemical (e.g., PCBs) per kilogram of an environmental medium
(e.g., fish tissue); also expressed as parts per billion or ppb.
mg/kg: Milligrams of a chemical (e.g., PCBs) per kilogram of an environmental medium
(e.g., fish tissue); also expressed as parts per million or ppm.
74
------- |