U.S. EPA j GE-PITTSF i E LD/H O U S ATO N I C RIVER SITE
LEARN MORE AT:www.epa.gov/region 1/ge
Statement of Basis for EPA's Proposed Remedial Action for
the Housatonic River "Rest of River
THE RIVER The Housatonic River is contaminated with
polychlorinated biphenyls (PCBs) released from the General
Electric Company (GE) facility in Pittsfield, MA. The entire site
consists of the 254-acre GE facility; the Housatonic River and its
banks and floodplains from Pittsfield, MA, to Long Island Sound;
and other contaminated areas. Under a federal Consent Decree,
GE is required to address contamination throughout the site,
including in the River.
SUMMARY:
After careful study of the impacts of PCBs released to the
Housatonic River from the GE-Pittsfield/Housatonic River site in
Pittsfield, MA, and in consideration of the contaminant reduction
accomplished by cleanup activities at other parts of the site, EPA
proposes the following cleanup actions, known as corrective mea-
sures, or remedial action, for the "Rest of River" component of
the GE-Pittsfield/Housatonic River site. EPAs Proposed Remedial
Action was developed after consultation with Massachusetts De-
partments of Environmental Protection (MassDEP) and Fish and
Game (MassDFG) and the Connecticut Department of Energy
and Environmental Protection (CT DEEP).This Statement of
Basis, in conjunction with the Draft Modification to the Reissued
R.CRA Permit, constitute EPA's "Proposed Plan" or "Proposed
Cleanup Plan," setting forth EPA's Proposed Remedial Action
for the Rest of River and Operation and Maintenance (O&M) as
prescribed by Paragraph 22. n. of the Consent Decree (termed
the "Proposed Remedial Action" or "Proposed Cleanup Plan"
throughout this document) to address polychlorinated biphenyl
(PCB) contamination in river sediment, banks and floodplain
soil, and biota which poses an unacceptable risk to human
health and the environment.
In addition to addressing risks in the areas slated for cleanup,
the Proposed Remedial Action also includes provisions to re-
duce downstream transport of PCBs, relax or remove fish con-
sumption advisories, and to avoid, minimize and/or mitigate ad-
verse impacts to state-listed species and their habitats regulated
under the Massachusetts Endangered Species Act (MESA), and
continued >
YOUR OPINION COUNTS:
OPPORTUNITIES TO COMMENT
EPA is accepting public comment on this proposal from
June 25, 2014 through August 8, 2014. EPA's Proposed
Remedial Action is based on current information and the
cleanup plan couid change in response to public comment
or new information. The following two public informa-
tional meetings will include a presentation describing the
Proposed Remedial Action, followed by a question and
answer session. EPA will begin a formal public comment
period on June 25, 2014. Near the end of the public
comment period, EPA will schedule a Public Hearing
where the public will have an opportunity to make oral
comments during this Hearing for EPA to consider. You
may also submit written comments - see page 43 to find
out how.
For further information about these meetings, call
Kelsey O'Neil of EPA's Community Affairs office at
617-918-1003, or toll-free at 1-888-372-7341.
Public Informational Meeting
Wednesday, June 18, 2014 at 6:00 pm at Lenox Memorial
Middle/High School, Lenox, MA
Public Informational Meeting
Tuesday, June 24, 2014 at 6:00 pm at Kent Town Hall,
Kent, CT
Public Hearing
date/time/location to be determined
SEPA Environmental Protection
United States
Environmental
I llllll lillllllli Kill Mil Hill lill lili June 2014
SDMS Doc ID 558621
-------
Table of Contents
Summary of the Proposed Cleanup Plan 1
Scope of This Document 2
Where Are We In The Cleanup Process? 3
A Closer Look at EPA's Proposed Cleanup Plan 3
Why EPA is Proposing This Cleanup Plan 11
Expected Outcome of the Proposed Remediation 11
Site Description 12
Why Cleanup Is Needed 14
Description of Cleanup Objectives and
Alternatives Considered 18
How Does EPA Choose A Final Cleanup Plan? 25
Comparative Analysis of Combined Sediment/Floodplain
Alternatives 28
Comparative Analysis of Treatment/Disposal
Alternatives 35
For More Detailed Information 42
Key Contacts 42
Public Comment Period 43
Send Us Your Comments 43
Acronyms 43
continued from front pg >
the ecosystem, a part of which has been designated as a state
Area of Critical Environmental Concern (ACEC). In addition,
an Adaptive Management approach is proposed to ensure that
the cleanup is performed using the best available technologies
and methods. EPA also proposes that contaminated material be
shipped off-site to existing licensed facilities for disposal.
Consistent with actions at other contaminated sediment sites,
this Proposed Remedial Action relies on a combination of
cleanup approaches that apply to specific "reaches" of the river,
as described below:
Removing and capping PCB-contaminated sediment in
some reaches in the Housatonic River.
Monitoring natural recovery in some reaches in the
Housatonic River.
Removing PCB-contaminated soil from some areas in the
10-year floodplain adjacent to the river, including vernal
pools, and restoring affected areas.
Stabilizing PCB-contaminated erodible river banks that
are a source of PCBs that could be transported down-
stream, focusing on the use of bioengineering techniques
in restoring any disturbed banks.
Transporting and disposing of all excavated contaminat-
ed soil and sediment off-site at existing licensed facilities
approved to receive such soil and sediment.
Placing restrictions (Institutional Controls) on eating fish,
waterfowl, and other biota where PCB tissue concen-
trations pose an unacceptable risk unless/until such
consumption advisories are no longer needed, as well as
restricting other activities that could potentially expose
remaining contamination.
Establishing procedures to address PCB contamination
associated with future work.
Maintaining remedy components and monitoring over
the long-term to assess the effectiveness of the cleanup
and recovery of the river and floodplain.
Establishing mechanisms for additional response actions if
land uses change (e.g. dam removal, changes in floodplain
land use)
Conducting periodic reviews following the cleanup to
evaluate the effectiveness and adequacy of the cleanup in
protecting human health and the environment.
The cost of the Proposed Remedial Action is estimated at
$613 million and will take approximately 13 years to implement.
A more detailed description of the Proposed Remedial Action
begins on page 3.
SCOPE OF THIS DOCUMENT
This document, in conjunction with the Draft Modification
to the Reissued RCRA Permit ("Draft Permit"), satisfies the
requirements set forth in the law, regulations, and Consent De-
cree governing this matter, United States, et. al., v General Elec-
tric Company, CA No. 99-30225 (D. Mass) (entered Oct. 27,
2000) ("CD" or "Decree") for a RCRA "Statement of Basis," 40
C.F.R. §124.7. Namely, this document, together with the Draft
Permit, describes the derivation of the Performance Standards
in the Draft Permit and the associated remedial actionor cor-
rective measures necessary to meet the Performance Standards
to address PCBs and any other hazardous waste, constituents
or substances that have migrated from the GE facility to surface
water, sediment, floodplain and bank soil, and biota in the Rest
of River. The Draft Permit also includes the identification of the
applicable or relevant and appropriate requirements (ARARs)
under federal or state law that must be met by such corrective
measures, and where EPA proposes to waive any such ARARs,
the basis for such waiver.
The Proposed Remedial Action is based upon the information
included in EPA's Administrative Record which can be reviewed
at the information repositories identified on page 42 of this
document. An index of the Administrative Record can be found
at www.epa.gov/region 1 /ge/proposedcleanupplan.html.
Using the information in the Administrative Record, EPA has
evaluated different combinations of cleanup alternatives for
river sediment/ banks and floodplain soil to remove, contain,
monitor and/or treat PCB contaminated material to protect
2
-------
Key:
I I EPA
I I GE
B Public
completed
in progress
Rest of River Investigation
T
ERA
HHRA
T
1
RFI Report
Modeling Study
Cleanup Goals (IMPGs)
Corrective Measures Proposal/Study (CMS)
Proposed Plan/
I
Final Cleanup Decision
Appeals Process
Remediation
ERA: Ecological Risk Assessment | HHRA: Human Health Risk Assessment | RFI: Resource Conservation and Recovery Act (RCRA) Facility Investigation
FIGURE 1 SUMMARY OF REST OF RIVER CONSENT DECREE PROCESS
human health and the environment from exposure to contami-
nated soil, sediment, surface water and biota, control sources of
releases, and attain (or waive) ARARs. The wide range of the
various combinations of cleanup alternatives that were evaluat-
ed for Rest of River are summarized in this document beginning
on page 18, as well as the cleanup or remedial action that EPA
is proposing for public comment ("Proposed Remedial Action").
WHERE WE ARE IN THE CLEANUP PROCESS
EPA proposed the site to the Superfund National Priorities List
(NPL) in September 1997. The federal and state government
agencies, the City of Pittsfield, the Pittsfield Economic Develop-
ment Authority, and GE entered into negotiations late in 1997
in an attempt to reach a comprehensive settlement to address
contamination at and from the GE facility. These negotiations
resulted in a CD approved by the court on October 27, 2000.
This CD governs the cleanup of the site,
In the CD, the river is to be addressed in three stages:
1 The first half-mile cleanup area adjacent to GE's former
Pittsfield plant, completed by GE in 2002;
2. The next one and one-half miles (1.5 mile), which was
cleaned up under EPA's direction with a cost-sharing
agreement with GE, completed in 2007;
3. The "Rest of River", extending from the end of the 1,5
mile cleanup, at the confluence of the East and West
Branches of the Housatonic River at Fred Garner Park
in Pittsfield, through Massachusetts and Connecticut. IT.is
third stage is the subject of this Proposed Cleanup Plan,
The CD did not include a specific cleanup plan for Rest of River
but rather identified a process for selecting a remedial action as
illustrated in Figure 1. As part of this process, this Statement of
Basis is to be issued along with a Draft Modification to the Reis-
sued RCRA Permit, and EPA's Proposed Remedial Action must
be issued for public comment. Following the public comment
process and other requirements outlined in the CD, EPA will
issue a response to public comments and a final Modification
of the Reissued RCRA Permit ("the Final Permit"). The Final
Permit could be different from the Proposed Remedial Action,
depending upon information that EPA considers as a result of
public comments.
A CLOSER LOOK AT EPA'S CLEANUP PLAN
The Rest of River at the GE-Pittsfield/Housatonic River Site
is broken into a series of reaches, designated as Reaches 5
through 16, that contain sediment and riverbank soil contami-
nated with PCBs. In addition, areas with PCB-contaminated soil
greater than 1 milligram per kilogram (mg/kg) in the floodplain
adjacent to these reaches are also included in the Rest of River
study area. Figures 2 and 3 show the location of the various
river reaches and the floodplain areas of the site.
Based on its careful evaluation of a range of alternatives using
the nine criteria specified in the RCRA Permit, EPA is proposing
3
-------
for public comment the Rest of River cleanup outlined below.
EPA's preferred alternative or Proposed Remedial Action is
Combination Alternative 9 (SED9/FP4 MOD with TD1).
Combination Alternative 9 requires excavation and capping/
restoration of sediment, river banks and floodplain soil in cer-
tain areas to protect human health and the environment while
seeking to avoid, minimize or mitigate unacceptable impacts to
state-listed species and their habitats and the Area of Critical
Environmental Concern ("ACEC"). The Proposed Remedial
Action also includes disposal of all excavated contaminated soil
and sediment off-site at existing licensed facilities approved to
receive such soil and sediment, with a preference to maximize
transport via rail. The proposed Performance Standards and
corrective measures required to implement this cleanup are
outlined in the Draft Permit. EPA's Proposed Remedial Action
was developed in consultation with MassDEP, MassDFG, and
CT DEEP.
River Sediment and Banks
The following corrective measures and performance standards
for river sediment and banks are being proposed by EPA to: re-
duce risks to humans from consumption of fish and waterfowl;
reduce risks from direct contact to sediments; reduce ecological
risks; and to control the sources of releases to reduce down-
stream transport of PCBs. Specific Performance Standards and
benchmarks for fish tissue and waterfowl concentrations, soil
and sediment concentrations, and downstream transport, and
the basis for these, have been included in the Draft Permit. EPA
is specifically seeking comment on the appropriateness of these
or alternative numerical standards. These Performance Stan-
dards and benchmarks apply throughout the Rest of River.
In this Proposed Remedial Action, removal of PCB-contaminat-
ed sediment is required in a number of areas followed by the
placement of a cap. Specifically, an engineered cap will be de-
signed to physically and chemically isolate the residual PCBs in
sediment and provide habitat for aquatic plants and animals and
reduce downstream transport of PCBs. A more detailed descrip-
tion of the design of the engineered caps is provided on page 8.
Reach 5A
In Reach 5A, the 5 miles from the confluence of the East and
West Branches of the Housatonic (at Fred Garner Park in
Pittsfield) to the Pittsfield wastewater treatment plant, the Pro-
posed Cleanup Plan requires the removal of river bed sediment
throughout Reach 5A and soil in eroding river banks contami-
nated with more than 5 mg/kg PCBs, capping of the river bed,
and stabilization of contaminated erodible river banks. Additional
data will be collected to better quantify the concentrations of
PCBs in river banks and locations of erodible river banks and to
determine the cap thickness and removal depth in the river,
A focus of the river bank work will be to reduce the mobiliza-
tion of PCBs into the river from the erosion of contaminated
banks while maintaining the dynamic nature of the River. For
banks that need to be addressed, reconstruction and stabili-
zation of remediated banks can be achieved in a number of
different ways, including using the principles of bio-engineering
and natural channel design. See the information outlined on
page 5 of this document regarding these concepts. Activities in
the banks will follow the hierarchy below of most preferred to
least preferred methods:
Pittsfield
Reach 10
Reach 11
Litchfield
Cornwall
j Bridge
Reach 12
CONNECTICUT
Reach 13
Fairfield
FIGURE 2 HOUSATONIC RIVER, REACHES 5
THROUGH 17
Confluence -
Reach 5A5
Reach 5B{ Reach 5
Reach 50
NEW YORK j Reach 6 (iWoods Pond)
Hampshire
S Reach 7
- Reach 8 (Rising Pond)
Sreat
Bariflgton- } Berkshire j Hampde
Reach 9
S
MASSACHUSETTS
- \
Hartford
\ Reach 14 (Lake Lillinonah)
New Haven
Reach 15 (Lake Zoar)
Reach 16
(Lake Housatonic)
Long Island Sound
4
-------
soil bioengineering
restored banks -
habitat layer
- proposed grade
-typical baseflow <
NATURAL CHANNEL DESIGN AND BIOENGINEERING
Over the past 200 years, the Housatonic River ecosystem has undergone a history of channel disturbances and channel relo-
cations, and as a result, there is evidence of bank erosion which contributes a significant amount of sediment and associated
PCBs to the river system. The Housatonic River is currently recovering from these past disturbances, and over time, the
river will approach sustainable dynamic equilibrium.
Natural channel design is a method of stream restoration that attempts to create a stable stream channel that is capable of
balancing flows and sediment loads by accelerating the trajectory towards a sustainable, dynamic equilibrium by working with
the stream processes. A naturally stable stream channel maintains its dimension, pattern and profile such that the stream
does not degrade (erode) nor aggrade (rise). One tool that is often used in natural channel design is the Bank Assessment
for Non-point source Consequences
of Sediment (BANCS) model which
assesses the erodibility of the river
banks, as further described in the
Draft Permit,
Another too! that is often used when
dealing with eroding banks is bioen-
gineering, which combines biological
and engineering concepts. Rather
than stabilizing an eroding or unstable
bank solely with rock, bioengineer-
ing techniques combine the natural
elements of the site, such as rock, soil,
trees, and other native vegetation, to
create a complex mix of material to
rebuild the banks. Once the vegeta-
tion has established itself, the need for
maintenance is reduced or completely
eliminated.
1. Where possible, the goal is to leave banks intact
with no disturbance or excavation, (i.e., where they
do not exceed the criteria outlined in the draft
Permit).
2. Reconstruct remediated banks with bio-engineering
"soft" techniques;
3. Reconstruct remediated banks with an engineered
cap extending into the river bank covered with a
bio-engineering/"soft" layer;
4. Piace rip-rap cap or hard armoring on surface of
banks (e.g. this approach may be necessary in areas
where infrastructure such as bridges and culverts
must be protected).
Reach 5B
In Reach 5B (the 2 miles of river from the Pittsfield wastewater
treatment plant to Roaring Brook in Lenox, NA), the Proposed
Cleanup Plan requires the excavation and restoration of areas
of river bed and banks that exceed the reach-specific Perfor-
mance Standard of 50 mg/kg PCBs. Additional data will be
collected to determine the location of PCB concentrations
in sediment and banks that exceed 50 mg/kg that would be
targeted for remediation. Any excavated Reach 5B riverbanks
would be restored using the hierarchy listed under Reach 5A.
In addition, this component of the remedy includes a provision
for a pilot study regarding Enhanced Nonitored Natural Recov-
ery (Enhanced NNR or ENNR) throughout Reach 5B. This pilot
study would be performed to evaluate the effectiveness of using
sediment amendments such as activated carbon to reduce the
bioavailability of PCBs. Following the review of the pilot study
results, and through an adaptive management framework, the
use of amendments will be applied to all of Reach 5B.
Reach 5C
In Reach 5C (the 3 miles between Roaring Brook and the
headwaters of Woods Pond), the Proposed Cleanup Plan
5
-------
Confluence of East and West Branches
>West Pittsfie.ld-""
DALTON
lapkgfvtttev..
Reach 5A
ip Marion White'
Russell
Steven § Corner
RICHMOND"
LENOX
iichmond]
Reach 5C
)Richmond Furnace
WASHINGTON
Lernox-Sta-tten j
Reach 6 (Woods Pond)
/est Stock bridge
Reach 7A
Lenox Dal
lahkeenac Heigf
Reach 7B
STOCKBRIDGE
Columbia Mill 'Pam
WEST STOCKBRIDGE
!)lnterlaken
Reach 7C
)West Stockbridge Center
)Larrywaug
Reach 7D
lendale Dam
/South Let
Willow Mill Dam
Reach 7E
NhfRINGrtAM
GREAT BARRINGTON
OTyringham
SOv
MONTEREY
LEGEND:
O Town/City
Roads
/v Reach Division Line
Housatonic River
State Park
Municipal Boundary
10-Year Floodplain
N
S
1 0.5 0
1 Miles
0.850.425 o 0.85 Kilometers
GE- Pittsfield/Housatonic River Site
Rest of River
FIGURE 3
HOUSATONIC RIVER,
PRIMARY STUDY AREA
(REACHES 5 AND 6) AND
REACHES 7 AND 8
File: Y:\gepitt\mxds\era_reaches_update_april2014.mxd, 4/11/2014 3:16:24 PM, ricksc
-------
requires removal of river bed sediment followed by capping
residual PCBs in sediment, Riverbanks in this reach generally
are not eroding and will be left intact, unless disturbed by other
remediation activities.
Backwaters Adjacent To Reaches 5 Through 7
There are a series of backwater areas adjacent to the river in
Reaches 5, 6, and 7, EPA's Proposed Cleanup Plan requires the
removal of contaminated sediment in Backwaters to allow a cap
to be placed over PCBs that exceed an average of 1 mg/kg.
Some portions of backwaters designated as having high-quality
habitat for state-listed species (known as "Core Area 1 habitat")
will not be remediated except in discrete areas with PCB con-
centrations greater than 50 mg/kg PCBs, Additional data will
be collected to assess PCB concentrations in the backwaters to
determine areas for cleanup, EPA will also require that a pilot
study be conducted to determine if other approaches, such as
the addition of activated carbon, would be useful in reducing
the bioavailability of PCBs in the backwater areas which are not
actively remediated (i.e. Core Area 1 habitat),
Woods Pond (Reach 6)
In Reach 6 (Woods Pond), the Proposed Cleanup Plan specifies
the removal of contaminated sediment and the placement of a
cap, with the design generally providing a minimum water depth
of six feet in the pond with shallower water depths in the near-
shore areas. In addition to reducing risks from fish (and other bi-
ota) consumption and ecological risks, this action in Woods Pond
will reduce risk to people from direct contact with the sediment.
It also will remove a significant mass of PCBs, reducing the po-
tential for release of PCB contaminated sediments in the case of
dam failure, and increasing the PCB-trapping efficiency of Woods
Pond, thus assisting in reducing downstream transport of PCBs.
Reach 6 will be monitored over the long term following the
cleanup and, if substantial PCBs accumulate in the pond, removal
of the accumulated sediment will be required. In addition to solic-
iting comments on the overall Proposed Cleanup Plan, EPA is also
soliciting comments on other options for Woods Pond that could
accomplish similar reductions in risk and downstream transport
and could also be suitable under the nine criteria,
Columbia Mill Impoundment (Reach 7B), Eagle Mill Impound-
ment (Reach 7C), Willow Mill Impoundment (Reach 7E), and
Glendaie Impoundment (Reach 7G)
This component of EPA's Proposed Cleanup Plan addresses
the impoundments behind the four dams in Reach 7 (Columbia
Mill Dam, Eagle Mill Dam, Willow Mill Dam and Glendaie Dam
(Reaches 7B, 7C, 7E, and 7G, respectively). EPA is proposing a
number of potential approaches to better integrate the clean-
up with potential dam or impoundment use, maintenance, or
removal. First, if dam maintenance or removal is planned, the
Proposed Cleanup Plan provides for GE to work with those
planning work on these dams, to fund sampling and analysis,
and take responsibility for the incremental costs associated with
the assessment, removal, management, and disposal of PCBs,
Dam removal itself is not a component of this cleanup plan and
would be conducted by others in coordination with GE and
appropriate State and Federal agencies. If no dam removal plans
have materialized by the time that GE would otherwise be
required to move forward with remediation of these impound-
ments, sediment would be removed from the river bed prior
to placement of a cap to sequester remaining contamination
exceeding an average of 1 mg/kg PCBs, In such a case, as part
of Institutional Controls, GE would remain responsible for incre-
mental costs due to PCBs for future dam work, Final removal
depths, locations, and engineered cap configurations will be
determined during remedial design. An additional option, in lieu
of capping, would allow GE to excavate the sediment in each
impoundment to meet an average of 1 mg/kg PCBs through-
out the impoundments.
Rising Pond (Reach 8)
In Reach 8 (Rising Pond), removal of contaminated sediment
is required prior to placement of a cap to sequester remaining
contamination in areas that exceed 1 mg/kg PCBs. Additional
data will be collected to assess PCB concentrations in Rising
Pond to determine areas for cleanup, In lieu of capping, GE
could excavate the sediment in Rising Pond to meet an average
of 1 mg/kg PCBs throughout the pond.
Flowing Sub- reaches in Reach 7 and Reaches 9 through 16
Monitored Natural Recovery (MNR) will be implemented in
the flowing sub-reaches in Reach 7 between Woods Pond and
Rising Pond (7A, 7D, 7F, and 7H) as well as Reaches 9 through
16 (from Rising Pond Dam through Connecticut). Rather than
requiring active measures such as excavation or capping, MNR
typically relies on physical, chemical, and biological processes to
isolate, destroy, or otherwise reduce exposure to, or toxicity
of, contaminants in sediment and to achieve Performance Stan-
dards, For this site, MNR is generally occurring by the physical
processes of sedimentation and dilution of upstream sources,
In addition, the progressive increase in river flow and associat-
ed solids from tributaries located downstream of Rising Pond
naturally attenuate PCB concentrations in sediments as they
combine with PCB-impacted upstream water and solids, The
effects of MNR are exhibited in decreasing trends in fish and
benthic invertebrate PCB levels that have been observed in in
reaches 9-16 during the last 25 years. Long-term monitoring in
both Massachusetts and Connecticut is a necessary component
of MNR to ensure that risk reduction and ecological recovery
by natural processes are continuing to occur as expected and
downstream transport and biota Performance Standards are
met; and there is progress towards the long-term biota bench-
marks outlined in the Draft Permit.
7
-------
Engineered Cap Design
Several components of EPA's Proposed Cleanup Plan require
construction of an engineered cap. In each area to be capped,
sediment would be removed to allow the placement of an en-
gineered cap to the final grades determined to be appropriate
during design of the remedy, generally to the pre-existing grade.
Each cap will likely consist of a mixing layer, a chemical isolation
layer to minimize PCB migration up through the cap, a filter
layer (if necessary), a protective layer (to prevent disruption
and erosion of the chemical isolation layer and exposure of the
underlying contaminated sediment), a bioturbation layer, and
a habitat layer. See Panel 1 of Figure 4 for an example of the
composition of an engineered cap. During remedial design it will
be determined if certain cap component layers can be com-
bined, additional cap components are necessary, or other cap
configurations are appropriate.
Specific engineered cap Performance Standards and design
principles have been included in the Draft Permit. Cap thickness
affects the dredge/excavation depth required (and thus, the
volume of contaminated sediment required to be transported
and disposed of off-site). There is value in minimizing the total
cap thickness while ensuring that the cap will, over the long-
term, be successful in physically isolating residual PCBs from
human and environmental receptors and protecting against
future downstream transport. In that context, EPA has pro-
posed Engineered Cap Performance Standards that do not
specify particular thicknesses and are flexible enough to allow
for construction of caps that are protective, permanent, and
implementable and are suitable under conditions that may be
associated with climate change, while still being designed to
minimize cap thickness. Cap thickness may also be minimized
to the extent that one layer of material can satisfy more than
one functional requirement such as mixing, chemical isolation,
erosion protection, bioturbation, or habitat functions, as well as
through other means.
The thickness of the engineered cap will determine the depth
of sediment removal in most reaches and the final grade of the
sediment bed determined to be appropriate during design of the
remedy. Generally, it is expected that in most areas the sediment
bed elevation will be returned to that which was present prior to
removal. This concept is illustrated in Panels 2 and 3 of Figure 4.
Floodplains And Vernal Pools
These components of the Proposed Cleanup Plan would be
performed in the floodplain in Reaches 5 through 8 concurrent
with the nearby sediment cleanup activities described above.
These cleanup actions will reduce ecological risk and direct
contact risk to humans. The specific risk-based cleanup stan-
dards that apply to each exposure area within the floodplain
are identified in the Draft Permit (see Figures 3 through 5 and
Tables 1 through 5 in the Draft Permit). In general, the process
to implement the cleanup in the floodplain includes:
Gathering additional information, including further
delineation of PCB concentrations, to support the final
cleanup design.
Avoiding, minimizing or mitigating impacts to state-listed
species and their habitats, as identified by the Common-
wealth in Core Areas.
Removing floodplain soil with PCB concentrations
exceeding the cleanup standards to a depth of one foot,
except in frequently used subareas, which will be exca-
vated to three feet.
Restoring the excavated floodplain areas, including
removing access roads, staging areas and other areas
affected by the cleanup.
During the Human Health Risk Assessment, EPA determined
that certain areas of the floodplain constituted "frequently
used subareas" that were subject to more intense use patterns
than other areas and thus are proposed to undergo additional
cleanup beyond that required for other direct contact exposure
pathways (see the third bullet above). EPA is also soliciting com-
ment from the public on the areas identified in Figure 5 of the
Draft Permit as those that are frequently or heavily used.
This Proposed Cleanup Plan includes an adaptive management
framework for the cleanup of vernal pools. Three different
approaches would be implemented concurrently in an initial
subset of vernal pools:
Conventional cleanup methods (e.g., excavation and
reconstruction) in a group of pools (8 to 10) would
be used to achieve the vernal pool soil/sediment
cleanup level of 3.3 mg/kg followed by active resto-
ration. Pools which occur within Core Area 1 habitat
will be excluded from consideration in this initial set
of vernal pools.
A pilot study would be conducted in a second group
of vernal pools to evaluate the effectiveness of a
sediment amendment such as activated carbon in
reducing the bioavailability of PCBs to biota and the
impacts of the amendment on these pools.
A pilot study using an innovative method would also
be conducted in a third group of vernal pools.
Based on the outcome of the first phase of vernal pool reme-
diation and restoration, EPA will determine how and where
8
-------
Habitat layer
-&{¦:Protective layer(s)
(erosion/bioturbation layers)
i>yynv!:-'ib.-''».V'.
Cap Layers
Not to Scale
Filter layer (if necessary)
Chemical isolation layer
Mixing layer
'A <2o . a' ^ (III- i*
Underlying sediment inventory
Panel 1 Example of Engineered Cap Components
F]oodplain
Panel 2
\ ijfel Floodplain
Cap from Panel 1
being installed to
grade from above cut.
Panel 3
14P-0094
FIGURE 4 EXAMPLE OF ENGINEERED CAP COMPONENTS AND PLACEMENT
9
-------
additional vernal pool remediation will occur to meet the vernal
pool-specific cleanup level of 3.3 mg/kg PCBs in soil/sediment.
For remediation of vernal pools in Core Areas, the approach
that will be generally used is to avoid excavation in vernal pools
within Core Area 1 and to minimize impacts of remediation, on
a case-by-case basis, of vernal pools within Core Areas 2 and 3.
Restoration
A restoration program will be required to address the impacts of
the cleanup on state-listed species and their habitats and on the
floodplain, river bottom and banks, impoundments, and vernal
pools with the broad objective to return, to the extent feasible
and consistent with the remediation requirements, the functions,
values, characteristics, species use, and other ecological attributes
existing prior to remediation. This program will include surveying
pre-remediation conditions, establishing restoration objectives
and evaluation criteria to measure success, and requiring coordi-
nation of restoration with remedial activities, post-remediation
restoration actions, monitoring and maintenance.
Off-Site Disposal of Contaminated
Sediment and Soil
All contaminated soil and sediment will be disposed of off-site
at existing licensed facilities approved to receive such soil and
sediment. EPA's Proposed Cleanup Plan includes maximizing
the use of rail to transport contaminated material to existing
off-site licensed facilities approved to receive such soil and
sediment.
Monitoring, Maintenance, Inspections,
Periodic Reviews, and Institutional
Controls
The baseline and long-term monitoring programs are major
components of the proposed action in both the areas with
sediment and soil removal and areas subject to MNR. Robust
monitoring programs will be implemented in Massachusetts
and Connecticut to evaluate the effectiveness of the remedial
actions in achieving Performance Standards and reducing the
risks posed by PCBs. Maintenance of the remedy in the river
bed and banks and floodplain will be required to ensure that
the Performance Standards continue to be met in the future.
Monitoring and maintenance activities will also be performed
to ensure that residual PCB-contaminated soil does not cause
unacceptable levels of PCB transport downstream.
Institutional controls, such as biota consumption advisories,
will be implemented to advise people against eating fish or
waterfowl or other biota where PCB concentrations pose an
unacceptable risk for consumption, as well as restricting other
activities that could potentially expose remaining contamination.
In all areas where unrestricted use is not achieved, institutional
controls will be put in place to restrict or place conditions on
activities that would cause unacceptable risks, such as distur-
10
bance of caps, excavation in floodplain areas, or future mainte-
nance or removal of dams. In addition, GE will be responsible
for (1) the cost of any investigation or materials handling and
disposal necessary in connection with maintenance work or to
implement a change in use (e.g. removal or maintenance of dams
or change in property use including agricultural uses) that could
otherwise lead to unacceptable risks or (2) further cleanup to
meet Performance Standards associated with the new use.
ADAPTIVE MANAGEMENT
Adaptive management is a process that allows a proj-
ect management team to adapt and optimize project
activities as they are implemented to account for
new information, changing conditions, and additional
opportunities such as innovative technologies. Adap-
tive management is intended to facilitate a process
that endeavors to minimize cost and maximize the
environmental benefits achieved by the actions taken.
EPA envisions that the corrective measures identi-
fied in the Proposed Remedial Action will be imple-
mented in a phased manner using such an adaptive
management approach. This approach will be
administered during design and construction activities
(including restoration), to adapt and optimize project
activities to account for "lessons learned," new infor-
mation and data, changing conditions, pilot studies,
and additional opportunities that may present them-
selves over the duration of the project.
In accordance with CERCLA and the CD, periodic reviews
(every five years) would also be conducted to evaluate the
effectiveness and adequacy of the remedial measures in protect-
ing human health and the environment.
Potential Community Impacts
While many of the cleanup areas in Rest of River are located in
relatively undeveloped areas, there are some that are in close
proximity to residential neighborhoods where the proposed
work would temporarily impact the surrounding communities.
Potential impacts during construction could include air quality,
odors, noise, lighting, traffic, impacts on local infrastructure
(such as roads and bridges), impacts on cultural resources, and
restrictions on use of the river for recreation or other uses. The
relatively undeveloped areas are, in large part, owned by the
Commonwealth of Massachusetts for their biological diversity
and wildlife-dependent outdoor recreation, as well as an adjoining
wildlife area owned by the Massachusetts Audubon Society.
-------
During remedial design, various measures will be evaluated and
subsequently implemented to reduce these potential impacts
on surrounding neighborhoods and communities. For example,
instead of having all neighborhoods affected at once, the work
would be done in phases working generally north to south, and
temporary haul roads would be built to limit use of local roads
and reduce construction traffic as much as possible.
To ensure careful coordination and enhanced safety for residents,
GE will be required to work closely with EPA, and in consultation
with the appropriate city and town officials, in developing man-
agement strategies and plans to guide the cleanup work.
' 1 1 i i i ' i e n ' ' . ¦ i , ¦ ' . i ' a - 1 "
In order to expeditiously and efficiently complete the proposed
remediation, EPA expects that several phases of the remedy will
be conducted concurrently to speed the overall completion of
construction. Sediment and floodplain work, including vernal pools,
will begin in Reach 5A and proceed downstream. Concurrently to
starting work at Reach 5A, work will also begin at Woods Pond
and proceed downstream to Rising Pond, however placement of
the engineered caps (if necessary) in the downstream impound-
ments will not occur until all remediation has been completed
upstream. See Figure 5. Additional data collection, baseline
assessments, and pilot studies will begin as early in the process as
practicable. It is also expected that, using an adaptive management
approach, the work will be phased, with each phase designed and
implemented individually. Under this approach, while construction
work is proceeding in one stretch of the river, planning and design
work, as well as review of activities conducted to date will be ongo-
ing for subsequent phases of work.
Note, the provisions to coordinate any required cleanup work
in the Reach 7 Impoundments with plans for dam use, removal,
or maintenance activities could lead to a change in the timing of
work in any of those impoundments.
Using the assumptions established in the Corrective Measures
Study (CMS), construction is expected to take 13 years to com-
plete. The estimated total cost for the preferred cleanup plan
including sediment and floodplain remediation, off-site transpor-
tation and disposal at facilities approved to receive such soil and
sediment, as well as operation, maintenance, and monitoring is
approximately $613 million.
WHY EPA IS PROPOSING THIS CLEANUP
PLAN
Based on the information in the Administrative Record,
including the RCRA Facility Investigation (RFI) and Corrective
Measures Study (CMS), EPA believes that the Proposed Reme-
dial Action or the Proposed Cleanup Plan best suits the Permit
evaluation criteria. The Draft Permit includes the Performance
Standards and corrective measures necessary to meet the
Performance Standards to address unacceptable risks to human
health and the environment, and reduce the potential for down-
stream transport of PCBs, while minimizing adverse impacts
to state-listed species and their habitats and being sensitive to
the characteristics of the Rest of River and related biodiversity
which formed the basis of the ACEC designation in a portion
of the study area. Also based on this analysis, certain areas
in the river and floodplain will be left undisturbed, including a
large part of Reach 5B. The Proposed Remedial Action also
removes and disposes off-site of large volumes of PCB-contam-
inated sediment and soil, from both the River itself, and the
associated floodplain. The Proposed Cleanup Plan provides for
the isolation of PCB contaminated sediments to reduce the risk
to human health and the environment. Any remaining contam-
ination will be monitored over the long term to evaluate the
continued effectiveness of the remedy.
Based on information currently available, EPA believes the
Proposed Remedial Action meets the General Standards for
Corrective Measures and provides the best balance of tradeoffs
among the other alternatives with respect to the relevant crite-
ria. EPA also expects the Proposed Remedial Action to (1) con-
trol the sources of releases so as to reduce or eliminate, to the
maximum extent practicable, further releases that may pose
a threat to human health and the environment; (2) attain the
Performance Standards; (3) comply with applicable standards
for management of wastes; and (4) be protective of human
health and the environment; (5) comply with ARARs (or justify
a waiver); (6) be cost-effective; (7) utilize permanent solutions
and alternative treatment technologies or resource recovery
technologies to the maximum extent practicable; and (8) satisfy
the preference for treatment as a principal element, or explain
why the preference for treatment will not be met.
EXPECTED OUTCOME OF THE PROPOSED
REMEDIATION
The cleanup reduces unacceptable human health risks from
direct contact with sediment and floodplain soil. In addition, the
cleanup is expected to result in reductions in biota concentra-
tions to allow increased human consumption of fish and other
biota taken from the river within a short time after remediation
is completed, and to greatly reduce the downstream transport
of PCBs. This should result in further reductions in PCB levels
in fish in both Massachusetts and Connecticut, which, over
time, should allow the consumption of additional fish meals or
increased consumption of other biota.
The sediment and river bank cleanup will reduce risk to ecolog-
ical receptors from exposure to PCBs by capping and removal
11
-------
Reach 5A
Reach 5B
Reach 5C
Backwaters
Reach 6
(Woods Pond)
Reach 7B*
Reach 7C*
Reach 7E
Reach 7G
Reach 8
(Rising Pond)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Year of Implementation
'Response actions, including sampling and sediment removal in coordination with dam
removal activities, could occur prior to completion of sediment removal at Woods Pond.
' 1 ,dP-nn<
ZB>
Removal and
Immediate Capping
Removal
Capping
IZH>
D^>
B>
¦*
~*>
FIGURE 5 ESTIMATED TIMELINE TO IMPLEMENTATION OF CLEANUP
GE-PITTSFIELD/HOUSATONIC RIVER SITE
REST OF RIVER
of PCBs and thereby reducing the exposure of ecological
receptors to PCBs. The floodpiain and vernal pool remediation
components of the Proposed Cleanup Plan are designed to
reduce risk for ecological receptors while being sensitive to
adverse impacts to state-listed species and their habitat.
EPA's Proposed Cleanup Plan will require restoration to
address the adverse impacts to the river bed, banks, wetlands
and floodpiain caused by the remediation. Engineered Caps in
the river bed will be required to include a habitat layer to assist
the recovery of aquatic organisms. Impacted river banks will be
restored using bioengineering wherever possible and appro-
priate. Areas of the floodpiain that are cleaned up or that are
disturbed for temporary remedial infrastructure (such as access
12
roads) will be restored. Vernal pools will be carefully monitored
prior to and following cleanup to assess recovery and the need
for further work. Restoration of impacted wetland habitat
is expected to be effective and reliable. Specific restoration
techniques will be implemented, evaluated and assessed and
modified as necessary using an adaptive management approach.
SITE DESCRIPTION
The Rest of River includes approximately 125 miles of river
over 12 river reaches (Reaches 5 through 16, as shown in
Figures 2 and 3) in Massachusetts and Connecticut and asso-
ciated floodpiain within the 1 mg/kg isopleth (approximated
by the 10-year floodpiain within the first 10 Vi miles). Reaches
5 through 8 flow through the towns of Pittsfield, Lenox, Lee,
-------
Stockbridge, and Great Barrington, Massachusetts (upstream to
downstream). Reach 9 flows through Sheffield, Massachusetts
and Reaches 10 through 16 are in Connecticut, from Canaan
downstream to Derby.
The first 10 1/2-mile stretch starting at the confluence of the
East and West Branches to Woods Pond Dam is referred to
as Reaches 5 and 6. This stretch of the River (Reaches 5 and
6) is the most contaminated portion of river addressed in this
Proposed Remedial Action and is estimated to contain approx-
imately 90% of the mass of PCBs that remain in the river sys-
tem (river and floodplains). The channel in this area is typically
60 to 90 feet (ft) wide (and occasionally as narrow as 40 ft or
as wide as 125 ft), bordered by an extensive flood plain, and has
a meandering pattern with numerous oxbows and backwaters.
This dynamic characteristic of the River, combined with the in-
tact, undeveloped floodplains, generates and maintains a diverse
mosaic of natural communities, including wetlands and fisheries
and wildlife species and habitats.
Eroding contaminated riverbanks are a significant source of
PCBs in Reach 5, currently contributing an estimated 45%
of the PCB load to the river and therefore are an important
consideration in evaluating remedial alternatives. The floodplain
is primarily associated with Reach 5, is up to 3,600 feet wide
and encompasses approximately 1,000 acres. Almost all of the
Primary Study Area (PSA), including Reach 5, is mapped by the
Commonwealth as priority habitat for state-listed species pro-
tected under MESA, including areas with dense concentrations
of overlapping habitat for eight (8) or more state-listed species.
Woods Pond (Reach 6), the first impoundment downstream
of the GE facility, is a shallow 60-acre impoundment that was
formed by the construction of a dam in the late 1800s. Down-
stream of Reach 6, the impoundments along Reach 7 and 8
also continue to be important sources of PCBs to downstream
transport and are contributing sources to biota consumption
advisories in those impoundments and in areas further down-
stream. These impoundments include Columbia Mill Impound-
ment (Reach 7B), Eagle Mill Impoundment (Reach 7C), Willow
Mill Impoundment (Reach 7E), Glendale Impoundment (Reach
7G), and Rising Pond (Reach 8).
Below Reach 8, PCB concentrations in all media drop off significant-
ly in the remaining downstream reaches. However, fish consump-
tion advisories and risks to wildlife that eat fish remain in place
along the Housatonic River below Reach 8 into Connecticut.
Site History
The industrial history of the Housatonic River floodplain dates
from the late 1700s and includes the development of paper
mills, blast furnaces, wool factories, and grist mills, along with
agriculture and home construction to accommodate both
permanent and seasonal population growth. This industrial and
agricultural history was accompanied by substantial changes
to the Housatonic River. Much of the river was modified and
realigned and the floodplain cleared. Several impoundments
were created both upstream of and within the site, altering the
natural flow regime of the river.
The Housatonic River, its sediment, floodplain, and biota are
contaminated with PCBs released from the GE facility located
in Pittsfield, MA. The 254-acre facility is the only major source
of the PCBs found in the Housatonic River sediment through
Reach 16 and in floodplain soil. Although GE conducted various
activities at the Pittsfield facility, the activities conducted by
the Transformer Division (construction and repair of electrical
transformers using dielectric fluids, some of which contained
PCBs) were the primary source of PCB contamination. Accord-
ing to GE reports, from 1932 through 1977, releases of PCBs
reached the wastewater and stormwater systems associated
with the facility and then were discharged/released to the East
Branch of the Housatonic River and to Silver Lake, a 26-acre
lake adjacent to the GE facility, as well as other environmental
media, including soil and groundwater. PCBs are presently
discharged into the Housatonic River from GE's Pittsfield facility
and are regulated under a National Pollution Discharge Elimina-
tion System (NPDES) discharge permit.
PCBs were initially discovered in sediment and fish in im-
poundments along the Housatonic River in Connecticut in the
mid-1970s. Since that time, numerous investigations have been
conducted by EPA, GE, and others to assess the presence and
extent of PCBs and other hazardous substances in various
media. PCBs detected in Housatonic River floodplain soil, sedi-
ment, and biota show little degradation over time in any media.
The State of Connecticut posted a fish consumption advisory
for most of the Connecticut section of the river in 1977 as a
result of the PCB contamination. In 1982, the Massachusetts
Department of Public Health (MassDPH) issued a consumption
advisory for fish, frogs, and turtles. In 1999, MassDPH issued a
waterfowl consumption advisory from Pittsfield to Great Bar-
rington due to PCB concentrations in wood ducks and mallards
collected by EPA above Woods Pond (Reach 6).
Approximately 818 acres of riparian area and floodplain within
Reaches 5 and 6 are owned by the Commonwealth, acting
through the Department of Fish and Game and its Division of
Fisheries and Wildlife, for its biological diversity and wildlife-de-
pendent outdoor recreation. In addition, a large portion of the
PSA was designated an Area of Critical Environmental Concern
(ACEC) by Massachusetts in 2009.
13
-------
The GE-Pittsfield/Housatonic River site has been subject to reg-
ulatory investigations dating back to the early 1980s, including
the issuance by EPA of a Resource Conservation and Recovery
Act (RCRA) Corrective Action Permit, which became effec-
tive in 1994 and two Administrative Consent Orders (ACOs)
entered into between GE and the Massachusetts Department
of Environmental Protection in 1990. In 1997, EPA proposed
listing the Site on the National Priorities List (NPL) of uncon-
trolled hazardous waste sites. In 2000, a settlement (memorial-
ized in a Consent Decree (CD)) was finalized with GE that in-
cluded an extensive plan to address contamination at and from
the GE facility. Pursuant to this CD, significant work has been
done at and near the GE facility as well as in the Reaches of the
River closest to the GE plant. Twenty separate removal action
areas were identified in the CD as areas requiring cleanup as
well as five groundwater management areas, and the first two
miles of the Housatonic River. To date, cleanup construction
has been completed at 18 of the 20 removal action areas and in
the first two miles of the River, while investigation and cleanup
work continues at the groundwater management areas and the
remaining two removal action areas.
i rn ti ri i t y Involveme ti t
Throughout the duration of the Rest of River project, EPA has
kept the local community and other interested stakeholders up
to date on various project investigations and activities. In May
2012, EPA published a status report, entitled Potential Reme-
diation Approaches to the GE-Pittsfield-Housatonic River Site
"Rest of River" PCB Contamination, summarizing the results of
EPA's technical discussions with the states of Connecticut and
Massachusetts and EPA's thinking regarding cleanup strategies.
A series of public meetings were held in 2012 to outline the
information contained in that document.
Prior to issuing the Status Report, EPA issued Fact Sheets
regarding the following topics:
The human health and ecological risks (June 2003,
summarized again in August 2009)
The Corrective Measures Study (October 2007,
March 2008, and September 2008)
An overview of PCBs, their properties, effects, and
fate and transport (January 2011)
EPA's Cleanup Decision Process (April 2011)
A description of the alternatives being evaluated in
the Revised CMS (April 2011)
In addition, in April/May 2011, EPA held a series of workshops
for the public on Rest of River topics, culminating in an all-day
"charrette" to discuss different considerations related to the
Rest of River.
EPA holds regular meetings with the Citizens Coordinating
Council to update them on the Rest of River as well as the oth-
er activities at the GE site. EPA has held an informal public input
period for deliverables generated for Rest of River and contin-
ues to place documents for the entire site on its website and to
maintain repositories throughout the affected communities.
In addition, between August 2012 and December 2013, EPA
held extensive discussions with GE regarding potential remedial
approaches for the Rest of River. The EPA/GE discussions,
which are described further in the Administrative Record,
included discussion of potential remedial components linked
with potential forbearance by GE of its ability to challenge a
proposed or final remedy. The EPA/GE discussions concluded
in December 2013 without reaching agreement.
: I l: i ;¦ .-""r. ¦-P
Past practices at the GE facility resulted in contamination of
Housatonic River sediment, floodplain soil, and biota including
fish and other animals. PCBs contaminated the floodplain by
the movement of contaminated sediment onto the floodplain
during times of high water. Based upon risk assessments con-
ducted by EPA, PCBs in the Housatonic River sediment, flood-
plain soil, and biota pose unacceptable risks to both human and
ecological populations.
: ; , r i ¦ ' .i
PCBs have been demonstrated to cause a wide variety of
adverse health effects. PCBs have been shown to cause cancer
in animals and are classified as probable human carcinogens.
Studies in humans provide evidence for potential carcinogenic
and non-carcinogenic effects of PCBs. PCBs have also been
shown to cause a number of serious non-cancer health effects in
animals, including effects on the immune system, reproductive
system, nervous system, endocrine system and other organs.
. , i in a n s a n d A n i i i
Just because contamination exists does not mean the environ-
ment or people are at risk. One has to have exposure to the
contaminant to have a potential risk. Exposure occurs when
people or other living organisms eat, drink, breathe or have
direct skin contact with a contaminant. Based on existing or
reasonably anticipated future land use at a site, EPA develops
different possible exposure scenarios to determine potential
risk, appropriate cleanup levels for contaminants, and potential
cleanup approaches.
Human health and ecological risk assessments have been pre-
pared by EPA for the site. These risk assessments use a number
of contamination exposure scenarios to determine if and where
there are current or potential future unacceptable risks. A com-
plete discussion of the risks posed at the site can be found in
14
-------
the final Human Health Risk Assessment (HHRA) and Ecolog-
ical Risk Assessment (ERA), EPA has also developed a number
of fact sheets which outline the findings of the risk assessments,
released in June 2003 (Human Health Risk Assessment), July
2003 (Ecological Risk Assessment) and August 2009 (summa-
rizing both risk assessments). These and other fact sheets can
all be found at EPA's website at www.epa.gov/ne/ge.
Human Health
People in Massachusetts have the potential for exposure to the
site's contaminants through eating fish from the river and hav-
ing contact with site sediment and floodplain soil, and through
potential future agricultural land use. In Connecticut, the risk is
mainly centered on eating fish from the river, Overall, the risk
assessment determined that the exposure pathways outlined
below pose an unacceptable risk: Figures 6 and 7 also provide
a summary of cancer risks and non-cancer risks for the various
pathways evaluated in the HHRA,
In the HHRA, risks are presented as numbers. Cancer Risk
is the increased probability, or chance, of additional cases of
cancer in an exposed population as a result of exposure to
chemicals at a site, In the reports for this site, a 1 in 1,000,000
chance is written as 1E-06 or 1 x 10"6, Non-Cancer Risk is a
comparison of an allowable exposure to the amount of exposure
estimated at a site. The comparison is called the Hazard Index (HI),
(site exposure)
HI =
(allowable exposure)
An HI greater than 1 indicates that the site exposure exceeds
the allowable exposure, Acceptable Risks for cancer are con-
sidered by EPA to be less than 1 in 1,000,000, Between a 1 in
1,000,000 (1 x 10'4) and a 1 in 10,000 (1 x 10"4) chance, EPA
looks at the site-specific factors affecting risk and the uncertain-
ties with the estimate, For non-cancer health effects, an HI less
than 1 means people are unlikely to be harmed.
Unacceptable Risk from Consumption of Fish and Waterfowl
The Housatonic River in Massachusetts and Connecticut is
currently under various state restrictions regarding the con-
sumption of fish and other animals from the river due to the
PCB contamination, Although current advisories are assumed
to reduce the amount of fish and other biota that some people
eat, there may be others who do not follow the advisories and
consume fish from the Housatonic River, The fish and water-
fowl consumption portion of the HHRA evaluated cancer and
non-cancer risks to individuals consuming quantities of these
foods that would be anticipated in the absence of restrictions,
as required by EPA guidance, In calculating risks, values were
established for the various factors, such as fish meals per year,
amount of fish per meal, cooking method, etc., that determine
the amount of an individuals exposure to PCBs from consum-
ing fish caught in the River, Some of these values were estab-
lished using site-specific data and others were assigned based
on national EPA risk assessment guidance. This evaluation was
done for both average and "maximally exposed" individuals, and
a similar procedure was followed for waterfowl consumption,
For fish, it was assumed that the maximally exposed individ-
ual consumes approximately 50 fish meals per year from the
Housatonic River, and for the average person, it was assumed
that they consume 7 meals per year from the Housatonic River.
The unacceptable risks identified in the HHRA for the maximally
exposed individual are summarized below:
Cancer risks from consuming PCB-contaminated fish and
waterfowl greatly exceed EPA's risk range in Massachu-
setts (River Reaches 5 through 8). The cancer risks in
Massachusetts are greater than those in Connecticut; with
risk estimates in Massachusetts reaches as high as two in
1,000 (2 x 10"3) for consumption of fish and one in 1,000
(1 x 10"") for consumption of waterfowl,
Non-cancer risks from consuming fish and waterfowl
greatly exceed EPA's threshold of a Hazard Index (HI) of
1 in River Reaches 5 through 16. The non-cancer risks in
Massachusetts are greater than those in Connecticut, with
His in Massachusetts reaches as high as 120 for consump-
tion of fish and 76 for consumption of waterfowl.
Although the amount of frogs and turtles consumed was
assumed to be much less than fish and waterfowl, con-
centrations of PCBs in these species would also result in
unacceptable risk if consumed in large quantities,
Unacceptable Risks from Direct Contact with River Sediment
Eight areas along the Housatonic River between Reaches 5 and
8 (in Massachusetts) were evaluated for risk due to exposure
to PCBs in sediment, Risks were based on recreational activities
that had contact with the sediment, such as wading, swimming,
fishing, waterfowl hunting, canoeing, natural history classes, and
other activities, The unacceptable risks from direct contact to
river sediment for the maximally exposed individual are summa-
rized below:
Non-cancer His exceeded EPA's threshold of 1 in Woods
Pond (Reach 6) and in portions of Reach 5C and Reach 5D
backwaters, and Glendale Dam impoundment (Reach 7G),
with His as high as 3.5,
15
-------
Mid 1970s PCBs detected in sediment in the Connecticut section of the Housatonic River
1977 State of Connecticut issues a report summarizing PCB contamination in fish in Connecticut
1980-1982 Massachusetts and GE agree on an Administrative Consent Order directing GE to investigate the Housatonic River; GE
conducts initial sampling and characterization of PCBs in the Housatonic River in Massachusetts
1983 GE submits report on investigation of contamination in fish in Connecticut impoundments
1985 GE submits a report on potential remedial actions and potential disposal sites for PCB-contaminated sediment located in the
River between the GE facility and Woods Pond
198 5 GE submits a report on screening of remedial alternatives for Connecticut impoundments
198 6 GE submits a report with additional information on wet dredging techniques, a five year plan for biodegradation and a
sediment sampling plan
1988-1989 GE submits reports on investigations into the potential effectiveness of velocity and sedimentation control methods
199 0 GE undertakes a Comprehensive Site Investigation for the Housatonic River
199 1 GE submits a Comprehensive Site Investigation Report
EPA issues GE a RCRA Permit, which includes the Housatonic River. Several parties appeal the Permit
1994 RCRA Permit is reissued and becomes effective
GE initiates additional investigations of the Housatonic River under the State ACO and EPA RCRA Permit
199 6 GE submits first RCRA Facility Investigation Report for the Housatonic River
1997 EPA proposes site for inclusion on the Superfund National Priorities List (NPL)
1997-1999 GE conducts remediation and restoration activities of sediment and riverbank soil in a 550-foot section of the East Branch
of the River adjacent to GE facility
199 8 GE and EPA initiate further studies of the Rest of River
1999 GE initiates riverbank soil and sediment remediation activities in a 1/2-mile stretch of the East Branch Housatonic
River adjacent to their facility
200 0 Court Approves Consent Decree governing Site cleanup and establishing process for Rest of River Study
200 2 GE completes the riverbank soil and sediment remediation and restoration activities in the 1/2-mile stretch of the Housatonic
River adjacent to their facility; EPA initiates riverbank soil and sediment remediation in the next 1.5 Mile stretch of the River
200 3 GE completes RCRA Facility Investigation Report for the Rest of River
200 5 EPA completes Peer-Reviewed Human Health & Ecological Risk Assessments for the Rest of River
200 6 EPA completes peer reviewed Watershed, Fate & Transport, and Food Chain Model Framework GE completes Interim
Media Protection Goals (IMPG) Document
EPA completes remediation and restoration activities in the 1.5 mile stretch of the East Branch of the Housatonic
River from the GE facility to the confluence of the East and West Branches of the Housatonic River
200 7 GE's submits Corrective Measures Study Proposal for the Rest of River
200 8 GE submits First Draft Corrective Measures Study (CMS) for EPA, State and Public Review
EPA Provides Comment to GE on Draft CMS
200 9 GE responds, in part, to EPA's comments; GE's response presented to Public for comment
GE submits Revised CMS Proposal to evaluate additional alternatives, after EPA, State, and Public Review, EPA approves
200 9 GE conducts remediation and restoration activities of sediment and riverbank soil in a 600-foot section of the
West Branch of the Housatonic River
201 0 GE submits Revised/Supplemental CMS for EPA, State, and Public Review
201 1 EPA holds series of workshops and "Charette" to engage public in remedy selection process
EPA presents potential proposed remedy to EPA's National Remedy Review Board (NRRB) and Contaminated Sediments
Technical Advisory Group (CSTAG)
Facilitated technical discussions among EPA and States initiated
201 2 Facilitated technical discussions among EPA and States concluded; Status Report entitled "Potential Remediation
Approaches to the GE-Pittsfield-Housatonic River Site 'Rest of River' PCB Contamination" released
Technical discussions with GE initiated
201 3 Technical discussions with GE conclude
201 4 EPA conditionally approves GE's Revised/Supplemental CMS
EPA Proposes Cleanup Plan, Statement of Basis, and Draft Modification to the Reissued RCRA Permit for Public Comment
-------
SUMMARY OF THE
RANGE OF NON CANCER RISKS
FROM TOTAL PCBs
JDL
~
A//,/
r~
Jt
/V /v yV ^
Hazard Indices from Direct Contact Exposure
Hazard Indices from
Fish and Waterfowl
Consumption
Note: The range of hazard indices consists of the lowest central tendency exposure hazard index
to the highest reasonable maximum exposure hazard index.
FIGURE 7 SUMMARY OF THE RANGE OF
NON-CAIMCER RISKS FROM TOTAL PCBs
o
o
oL
Based on the data collected, unacceptable risks from direct
contact with river sediment in Connecticut are not expected.
Unacceptable Risks from Direct Contact with Floodplain Soil
The floodplain along the Housatonic River in Massachusetts was
divided into 90 separate "exposure areas", two-thirds of them
between Reaches 5 and 6, Each area was evaluated for risk
due to exposure to PCBs in floodplain soil. Specific exposure
scenarios were evaluated for each area (e.g., recreational uses,
farming). The amount of exposure depended upon the accessi-
bility of a particular area, if people spend more time in a more
contaminated part of an exposure area, the risks will be higher.
The unacceptable risks from direct contact to floodplain soils
for the maximally exposed individual are summarized below:
Non-cancer His from exposure to PCBs in soil exceeded
the EPA threshold of 1 in 13 of the 90 exposure areas as
well as in 11 sub-areas, and in 5 of the subareas which are
expected to be more heavily used, with His as high as 16.
Of the 90 exposure areas, 41 areas had cancer risks
greater than 1 in 100,000, as well as in 13 subareas and
in 10 frequently used subareas, which are subject to more
intense use patterns than other areas.
Risk from Consumption of Agricultural Products Grown in the
Floodplain
The agricultural portion of the HHRA evaluated risks from
consuming commercial (from a facility in the Rest of River area)
and backyard (home grown) meat, dairy, and produce as well
as risks associated with home gardens. It also provided an esti-
mate of the risk associated with consuming other food sources
such as deer and wild edible plants.
The unacceptable risks identified in the HHRA for the maximally
exposed individual are summarized below:
For commercial farm families who consume their
products or crops and for backyard agricultural uses
(assuming that all feed and crops and grazing are in soil
with average PCB soil concentrations of less than 2 mg/
kg), cancer risks from PCBs are within EPA's Risk Range,
and non-cancer His are below EPA's threshold, However,
if average soil concentrations are higher in areas used by
animals or in areas where feed or crops are grown, both
cancer and non-cancer risks are likely.
Based on current land use, no remediation is required, Howev-
er, should additional areas of the floodplain be used for agricul-
ture in the future, the risk will need to be reevaluated.
Unacceptable Environmental Risks
Fish and wildlife are exposed to PCBs in sediment on the
bottom of the river and backwaters, or floodplain and vernal
pool soil or within the water column. PCBs are also in the tissue
and organs of animals living in the contaminated river and the
floodplain. Predatory fish and wildlife feed on contaminated an-
imals or organisms such as forage fish, crayfish or larval stages of
aquatic insects that live in the sediment (benthic invertebrates) or
animals in the floodplain are at risk from their foraging activities.
Risks from PCB exposure in the soil, sediment, and diet were
SUMMARY OF THE
RANGE OF CANCER RISKS
FROM TOTAL PCBs
IK-03 -
EG 1E-04
U 1L-05
IE-06
T"
~r
jk J' A J? a? Ji
-r
vv
Or
. '// ///' '
Risks from
Fish and Waterfowl
Consumption
Risks from Direct Contact Exposure
PxjEPA risk range (IE-06 to 1E-04)
Note: The range of cancer risks consists of the lowest central tendency exposure risk
to the highest reasonable maximum exposure risk.
FIGURE 6 SUMMARY OF THE RANGE OF CANCER
RISKS FROM TOTAL PCBs
17
-------
evaluated in the ERA for eight different groups of organisms
that reside in the Housatonic River and its floodplain; three
of these were aquatic (benthic invertebrates, amphibians, and
fish) and five were wildlife (insect-eating birds, fish-eating birds,
fish-eating mammals, other mammals, and Special Status Species
(e.g., endangered or threatened). Whenever possible, three
distinct lines of evidence were evaluated to best assess risk
(site-specific field studies, site-specific toxicity studies, and ad-
verse effects reported in the literature). Based on the weight of
evidence in this evaluation, the following unacceptable ecological
risks were identified in Massachusetts in Reaches 5 and 6 and
are described as high or intermediate:
PCBs in sediment and prey, as well as in the flood-
plain and vernal pools adjacent to those areas, posed
high risk to amphibians and piscivorous (fish-eating)
mammals. Risk was also high for some insectivorous
birds, such as wood duck;
Risk was intermediate to high for benthic inverte-
brates, organisms that live in and on river sediment
and form the base of the food chain;
Risk was high for exposure to prey for bald eagle
and American bittern, two birds selected to rep-
resent the Threatened & Endangered (T&E) spe-
cies, and intermediate for a T&E mammal species
(small-footed myotis, a bat); and
Risk was intermediate for piscivorous birds (osprey
and belted kingfisher), and for omnivorous and car-
nivorous mammals (red fox and short-tailed shrew).
In addition, in limited areas downstream of Woods Pond to Ris-
ing Pond in Reaches 7 and 8, exposure to PCBs leads to poten-
tial risks to benthic invertebrates, amphibians, trout, piscivorous
mammals, and bald eagles. In Connecticut, exposures to PCBs
cause potential risks to wildlife that eat fish.
DESCRIPTION OF CLEANUP OBJECTIVES
AND ALTERNATIVES CONSIDERED
The cleanup alternatives were developed to address the follow-
ing cleanup objectives:
Reduce the cancer risk and non-cancer risk for humans
(defined as achieving concentrations that do not pose
unacceptable risks using EPA's cancer risk range of
one in one million to one in 10,000 (10"6 to 10"4) and a
non-cancer HI of 1) from exposure to PCBs in dietary
items (primarily fish and duck), floodplain soil, and/or
sediment in the Rest of River.
Reduce the risk to ecological receptors from exposure to PCBs.
Reduce the long-term downstream transport of PCBs
18
in the Rest of River. This objective also includes the
control of sources of releases to the river.
To meet these objectives, EPA has proposed Performance Stan-
dards, corrective measures, and identified ARARs for the Rest
of River which are outlined in more detail in the Draft Permit.
Cleanup alternatives were developed and evaluated by GE in
the Corrective Measures Study (CMS). EPA has supplemented
the analysis conducted by GE with additional supporting docu-
mentation. The cleanup options, or remedial alternatives, that
were evaluated in detail and were considered for the Rest of
River are summarized below.
i rn e ri t 1 . - ' ' r . . : ' s
Eleven alternatives were developed for addressing contamina-
tion in sediment and riverbanks. The 11 alternatives are termed
SED 1 through SED 9, SED 9 MOD, and SED 10. These
alternatives encompass a broad range of options from no action
to the removal of over 2 million cubic yards of sediment and up
to 35,000 cubic yards of riverbank soil. Ten alternatives (FP 1
through FP 4, FP 4 MOD, and FP 5 through FP 9) addressing
PCB contamination in floodplain soil in the Rest of River were
also developed. All of the floodplain alternatives involve re-
moval of different volumes of contaminated floodplain soil and
placement of backfill except FP 1, the no action alternative.
As part of the site study, a range of potential cleanup goals,
known as Interim Media Protection Goals (IMPGs) were devel-
oped as one of the factors to use in the comparison of remedial
alternatives. In addition to the IMPGs, it is important to note
that certain specific numerical Performance Standards, which
may differ from the IMPGs, are being proposed in the Draft
Modification to the Reissued RCRA Permit to be met as part
of the remedy. To develop a range of cleanup alternatives, dif-
ferent options for cleanup goals were used to address potential
cancer risk to human health. The cleanup goal options for hu-
man health used by EPA to develop alternatives are within the
range of what EPA considers to be protective. Human health
cleanup goals are based upon reducing risk to within acceptable
levels (to within EPA's 10"6 to 104 cancer risk range and/or
non-cancer Hazard Index of one). Similarly, a range of IMPGs
for ecological receptors were also developed.
The Performance Standards and corrective measures for EPA's
Proposed Cleanup Plan are discussed generally in the section
entitled "A Closer Look at EPA's Proposed Cleanup Plan" and
outlined in specific terms in the Draft Permit.
More detail on these individual options to address sediment, river-
banks, and floodplain soil can be found in the Administrative Record.
-------
nation Alternative
2
SED 2/FP 1
nation Alternative
3
SED 3/FP3
nation Alternative
4
SED 5/FP4
nation Alternative
5
SED 6/FP4
nation Alternative
6
SED 8/FP7
nation Alternative
7
SED 9/FP 8
nation Alternative
8
SED 10/FP9
nation Alternative
9
SED 9 MOD/FP4 MOD
Combined Sediment and Floodplain Soil Alternatives
The remedy for the Rest of River will necessarily involve both
sediment and floodplain components. In order to more easily
explain and compare the alternatives, the individual sediment
and floodplain alternatives have been combined into nine com-
prehensive alternatives for all contaminated materia! (floodplain
soil/sediment). The Combination Alternatives (or Combina-
tions), listed below, were designed to span the full range of
remedial actions in terms of removal volumes, methods, and
affected areas:
Combination Alternative 1: SED1/FP 1
(the "no action* alternative)
Combinatic
Combinatic
Combinatic
Combinatic
Combinatic
Combinatic
Combinatic
Combinatic
(EPA's Proposed Remedial Action)
A matrix showing each combination alternative broken down
by river reach and floodplain is shown in Table 1, Table 2
outlines estimated volumes, timeframe, and acres addressed for
each of the combinations, Please note that the terms "Com-
bination Alternative 1" through "Combination Alternative 9"
are used to simplify the discussion and analysis for the reader
of this document. In other technical documents that are part
of the Administrative Record, the various individual sediment
and floodplain alternatives are typically referred to using their
corresponding "SED" and "FP" designations,
The evaluation of cleanup alternatives for Rest of River was
based on eleven sediment alternatives, ten floodplain alter-
natives, and five treatment/disposition alternatives, In the
proposed RCRA Permit modification and in GE's Revised CMS
submittal, alternatives have been analyzed through the use of
combination alternatives for sediment and floodplain, These
combined alternatives recognize the interrelated nature of the
sediment and floodplain cleanup, infrastructure, and thus the in-
terrelated nature of decision-making for the proposed remedy.
As such, the combination alternatives are designed to make re-
view of the many possible combinations of different approaches
more manageable for the public. Nonetheless, EPA is soliciting
public input on each component of the Proposed Cleanup Plan,
and reviewers may comment on individual sediment or flood-
plain components, or on different potential combinations of sed-
iment and floodplain remediation that are not part of the nine
Combinations discussed below, Note, for all of the alternatives
presented below the values for the areas affected by remedia-
tion, amount of sediment or soil to be removed, durations, and
costs are estimates for comparison purposes only.
Combination Alternative 1
Combination Alternative 1 is a combination of Sediment Alter-
native SED 1 and Floodplain Alternative FP 1. This alternative
involves no action in either the river or the floodplain, Combi-
nation Alternative 1 does not involve the excavation or capping
of any contaminated soil and sediment. Since there is no active
remedy construction, this alternative does not take any time
to implement. Contamination remains in the River above safe
levels for human health and ecological receptors and is expect-
ed to remain that way for over 250 years and there are no
measures to prevent exposure, There is no cost associated with
this alternative.
Combination Alternative 2
Combination Alternative 2 is a combination of Sediment Alter-
native SED 2 and Floodplain Alternative FP 1, This alternative
involves monitored natural recovery (MNR) in all River reaches
(Reaches 5 through 16) and no action in the floodplain, Combi-
nation Alternative 2 does not involve the excavation or capping
of any contaminated soil and sediment. Since there is no active
remedy construction, this alternative does not take any time to
implement (not including the duration of monitoring). Contam-
ination remains in the River and floodplain above risk-based
levels (IMPGs) for human health and ecological receptors and
is expected to remain that way for over 250 years. Human
exposure in the interim is addressed by Institutional Controls,
The cost for this alternative is estimated at $5 million : .
Combination Alternative 3
Combination Alternative 3 is a combination of Sediment Alter-
native SED 3 and Floodplain Alternative FP 3. This alternative
involves removal of approximately 2 feet of river bed sediment
followed by capping in Reach 5A; bank soil removal and stabi-
lization of Reach 5A and 5B river banks; a combination of thin
layer capping (often referred to as enhanced MNR or EMNR)
and MNR in Reach 5C; thin layer capping/EMNR in Reach 6
(Woods Pond); and, MNR in all other River reaches (Reach 5B,
Backwaters, and Reaches 7 through 16),
For the floodplain, Combination Alternative 3 involves the
removal of one foot of contaminated soil with subsequent
backfilling to meet a human-health based cleanup target based
on 104 cancer risk or non-cancer HI = 1 (whichever is lower)
plus additional cleanup to a depth of 3 feet in certain frequent-
ly used areas to achieve a human-health based cleanup target
based on 10"- cancer risk or non-cancer HI = 1 (whichever
is lower). This alternative also includes additional floodplain
19
-------
Table 1
Combination Alternatives Matrix
Combination
Alternative
Reach 5A
Reach 5B
Reach 5 Erodible
Banks
Reach 5C
Reach 5
Backwaters
Reach 6
Woods Pond
Reach 7
Impoundments
Reach 7
Channel
Reach 8
Rising Pond
Reaches
9-16
Floodplain
1
(SED1/FP1)
No Action
No Action
No Action
No Action
No Action
No Action
No Action
No
Action
No Action
No Action
No Action
2
(SED 2/ FP 1)
MNR
MNR
MNR
MNR
MNR
MNR
MNR
MNR
MNR
MNR
No Action
3
(SED 3/FP 3)
2 ft removal
with
capping
MNR
Removal/
stabilization
Combination of
TLC and MNR
MNR
TLC
MNR
MNR
MNR
MNR
Remove/replace top 12 inches to 10-4
ICR or HI = 1;
In frequently used areas
remove/replace top 3 feet to 10-5;
Additional floodplain excavation to
achieve the less strict ecological risk-
based IMPGs;
Remove/replace vernal pool soils > 5.6
mg/kg
4
(SED 5/FP 4)
2 ft removal
with
capping
2 ft removal
with
capping
Removal/
stabilization
Combination of
2 ft removal
with capping
(in shallow
areas) and
capping (in
deeper areas)
Combination of
TLC and MNR
Combination of
1.5 ft removal
with capping in
shallow areas
and capping in
deep area
MNR
MNR
TLC
MNR
Remove/replace top 12 inches to 10-5
ICR or HI = 1;
In frequently used areas
remove/replace top 3 feet to 10-5;
Additional floodplain excavation to
achieve the less strict ecological
risk-based IMPGs;
Remove/replace vernal pool soils > 5.6
mg/kg
5
(SED 6/ FP4)
2 ft removal
with
capping
2 ft removal
with
capping
Removal/
stabilization
2 ft removal
with capping
Removal of
sediments in
>50 mg/kg in
top 1 ft (with
capping/
backfill); TLC
for remainder
>1 mg/kg
Combination of
1.5 ft removal
with capping in
shallow areas
and capping in
deep area
TLC
MNR
Combination
of TLC in
shallow areas
and capping
in deep areas
MNR
Remove/replace top 12 inches to 10-5
ICR or HI = 1;
In frequently used areas
remove/replace top 3 feet to 10-5;
Additional floodplain excavation to
achieve the less strict ecological
risk-based IMPGs;
Remove/replace vernal pool soils > 5.6
mg/kg
6
(SED 8/ FP 7)
Removal to
1 mg/kg
depth
horizon with
backfill
Removal to
1 mg/kg
depth
horizon with
backfill
Removal/
stabilization
Removal to 1
mg/kg depth
horizon with
backfill
Removal to 1
mg/kg depth
horizon with
backfill
Removal to 1
mg/kg depth
horizon with
backfill
Removal to 1
mg/kg depth
horizon with
backfill
MNR
Removal to 1
mg/kg depth
horizon with
backfill
MNR
Remove/replace top 12 inches to 10-6
ICR but not <2 ppm;
In frequently used areas
remove/replace top 3 feet to 10-6;
Additional floodplain excavation to
achieve the more strict ecological
risk-based IMPGs;
Remove/replace vernal pool soils > 3.3
mg/kg
7
(SED 9/ FP 8)
2 ft removal
with
capping
2 ft removal
with
capping
Removal/
stabilization
2 ft removal
with capping in
upper reach
and 1.5 ft
removal with
capping in
lower reach
Combination of
sediment
removal with
capping and
capping
without
removal
3.5 ft removal
and capping in
shallow areas
and 1 ft removal
and capping in
deep areas
Removal depths
of 1 to 1.5 ft with
capping
MNR
Removal
depths from 1
to 1.5 ft with
capping
MNR
Remove/replace top 12 inches to 10-5
ICR or HI = 1;
In frequently used areas
remove/replace top 3 feet to 10-5;
Remove/replace vernal pool soils > 3.3
mg/kg;
Remove/replace any additional soils in
top 12 inches > 50 mg/kg
8
(SED 10/
FP 9)
2 ft removal
capping in
selected
areas
MNR
Removal/stabilization
in selected areas
MNR
MNR
Removal of 2.5
ft in areas >13
mg/kg in top 6
inches
MNR
MNR
MNR
MNR
Remove/replace top 12 inches to 10-4
ICR or HI = 1;
In frequently used areas
remove/replace top 3 feet to 10-4
9
(SED 9/
FP4 MOD)
2.5 ft
removal
and capping
Removal
and backfill
of areas >
50 mg/kg
and EMNR
in
remainder
of reach
Removal/
stabilization of
erodible river banks
in Reach 5A and
banks in reach 5B
w/PCBs > 50mg/kg
2 ft removal
with capping
Combination of
1 ft removal
and capping in
areas > 1
mg/kg,
excluding
certain high
priority habitat
Combination of
removal with
capping ranging
from 4 to 7 ft of
removal based
on water depth
Coordinate w/
dam removal;
Removal depths
of 1 to 1.5 ft with
capping; or
cleanup to 1
mg/kg
MNR
Removal
depths of 1 to
1.5 ft with
capping or
cleanup to 1
mg/kg
MNR
Remove/replace top 12 inches to 10-5
ICR or HI = 1;
Except in in high priority habitat areas,
then remove/replace top 12 inches to
10-4 ICR or HI = 1;
In frequently used areas
remove/replace top 3 feet to 10-5;
Remove/replace vernal pool soils > 3.3
mg/kg
Note: Sediment removal depths specified in this table are approximate and are for volume/cost estimation and for comparison purposes only. Actual removal depths would be determined in accordance with the
Modification of the Reissued RCRA Permit.
MNR - Monitored Natural Recovery
EMNR- Enhanced Monitored Natural Recovery
ICR - Incremental Cancer Risk
IMPGs - Interim Media Protection Goals
TLC - Thin-Layer Capping
o
rs|
-------
Table 2
Comparison of Combination Alternatives
Combination:
1
2
3
4
5
6
7
8
9
SED 1/
FP 1
SED 21
FP 1
SED 3/
FP 3
SED 5/
FP 4
SED 6/
FP 4
SED 8/
FP 7
SED 9/ FP
8
SED 10/
FP 9
SED 9
MOD /
FP 4 MOD1
Sediment
Removal
Volume (cubic
yards (cy))
0
0
134,000
377,000
521,000
2,252,000
886,000
235,000
890,000
Bank Soil
Removal
Volume (cy)
0
0
35,000
35,000
35,000
35,000
35,000
6,700
25,000
Sediment
Capping after
Removal
(acres)
0
0
42
126
178
0
333
20
298
Sediment
Backfill after
Removal
(acres)
0
0
0
0
0
351
0
0
0
Sediment
Capping without
Removal
(acres)
0
0
0
60
45
0
3
0
0
Thin Layer
Capping (acres)
0
0
97
102
112
0
0
0
0
Floodplain Soil
Removal
Volume (cy)
0
0
74,000
121,000
121,000
615,000
177,000
26,000
75,000
Floodplain
Acres
Excavated
(acres)
0
0
44
72
72
377
108
14
45
Total
Soil/Sediment
Volume
Removal (cy)
0
0
243,000
533,000
677,000
2,902,000
1,098,000
267,700
990,000
Estimated PCB
Mass Removed
(pounds)
0
0
21,700
33,300
37,300
94,100
53,100
13,900
46,970
Estimated Time
to Implement
(years)
0
0
10
18
21
52
14
5
13
Notes: Monitored Natural Recovery (MNR) is a component of all Combinations except Combination Alternative 1.
Volumes and areas specified in this table are approximate and are for volume/cost estimation and for comparison purposes
only. Actual volumes and areas will be determined in accordance with the Modification of the Reissued RCRA Permit.
1 Combination 9 sediment removal and capping estimates based upon capping of four Reach 7 impoundments, which is one
possible outcome of the cleanup approach proposed for these impoundments.
21
-------
excavation to achieve the less stringent ecological risk-based
numerical values (IMPGs).
Combination Alternative 3 involves the excavation of approxi-
mately 134,000 cubic yards of sediment, 35,000 cubic yards of
bank soil and 74,000 cubic yards of floodplain soil. This alterna-
tive involves the excavation of approximately 44 acres of flood-
plain area and also includes the capping of 42 acres of river bed
after excavation, and 97 acres of thin-layer capping of sediment,
Institutional Controls, long-term operation, monitoring, and
maintenance are also components of this alternative, This alter-
native is estimated to take 10 years to implement, The cost for
this alternative is estimated at $177 million, excluding costs for
transportation or disposal of excavated soil or sediment.
Combination Alternative 4
Combination Alternative 4 is a combination of Sediment Alter-
native SED 5 and Floodplain Alternative FP 4. This alternative
involves removal of approximately 2 feet of river bed sediment
followed by capping in Reaches 5A and 5B; bank soil removal
and stabilization of Reach 5A and 5B river banks; a combina-
tion of 2 foot removal followed by capping (in shallower areas)
and capping (in deeper areas) in Reach 5C; a combination
of thin layer capping/EMNR and MNR in the Backwaters; a
combination of 1.5 foot removal with capping in shallow areas
and capping (without sediment removal) in deeper areas of
Reach 6 (Woods Pond); thin layer capping/EMNR in Reach 8
(Rising Pond) and MNR in all other River reaches (Reach 7 and
Reaches 9 through 16),
For the floodplain, Combination Alternative 4 involves the
removal of one foot of contaminated soil with subsequent
backfilling to meet a human-health based cleanup target based
on 10° cancer risk or non-cancer HI = 1 (whichever is lower).
This alternative also includes additional floodplain excavation to
achieve the less stringent ecological risk-based numerical values.
Combination Alternative 4 involves the excavation of approx-
imately 377,000 cubic yards of sediment, 35,000 cubic yards
of bank soil and 121,000 cubic yards of floodplain soil. This
alternative involves the excavation of approximately 72 acres
of floodplain area and also includes the capping of 126 acres
of river bed after excavation, 60 additional acres of river bed
capping in areas not slated for excavation, and 102 acres of
thin-layer capping of sediment. Institutional Controls, long-term
operation, monitoring, and maintenance are also components
of this alternative. This alternative is estimated to take 18 years
to implement. The cost for this alternative is estimated at $319
million, excluding costs for transportation or disposal of exca-
vated soil or sediment
1 All cost estimates referenced in this document are in total 2010
dollars, for present worth values, see Table 6,
22
Combination Alternative 5
Combination Alternative 5 is a combination of Sediment
Alternative SED 6 and Floodplain Alternative FP 4, This alter-
native involves removal of approximately 2 feet of river bed
sediment followed by capping in Reaches 5A, 5B, and 5C; bank
soil removal and stabilization of Reach 5A and 5B river banks;
one foot removal followed by capping in areas of Backwaters
exceeding 50 mg/kg PCBs; 1.5 foot removal with capping
in shallow areas and capping (without sediment removal) in
deeper areas of Reach 6 (Woods Pond); thin layer capping/
EMNR in the Reach 7 impoundments; a combination of thin
layer capping/EMNR in shallow areas and capping in deep areas
of Rising Pond (Reach 8); and, MNR in all other River reaches
(Reach 7 channel and Reaches 9 through 16).
For the floodplain, Combination Alternative 5 involves the
removal of one foot of contaminated soil with subsequent
backfilling to meet a human-health based cleanup target based
on 10° cancer risk or non-cancer HI = 1 (whichever is lower).
This alternative also includes floodplain excavation to achieve
the less stringent ecological risk-based numerical values,
Combination Alternative 5 involves the excavation of approxi-
mately 521,000 cubic yards of sediment, 35,000 cubic yards of
bank soil and 121,000 cubic yards of floodplain soil. This alterna-
tive involves the excavation of approximately 72 acres of flood-
plain area and also includes the capping of 178 acres of river bed
after excavation, 45 additional acres of river bed capping in areas
not slated for excavation, and 112 acres of thin-layer capping of
sediment. Institutional Controls, long-term operation, monitoring,
and maintenance are also components of this alternative. This
alternative is estimated to take 21 years to implement, The cost
for this alternative is estimated at $397 million, excluding costs
for transportation or disposal of excavated soil or sediment.
Combination Alternative 6
Combination Alternative 6 is a combination of Sediment Alter-
native SED 8 and Floodplain Alternative FP 7. This alternative
involves removal of river bed sediment in Reaches 5A, 5B, and
5C, Backwaters, Woods Pond, the Reach 7 impoundments, and
Rising Pond to meet a PCB concentration of 1 mg/kg followed
by backfill; bank soil removal and stabilization of Reach 5A and
5B river banks; and, MNR in all other River reaches (Reach 7
channel and Reaches 9 through 16).
For the floodplain, Combination Alternative 6 involves the
removal of one foot of contaminated soil with subsequent back-
filling to meet a human-health based cleanup target based on a
10"6 cancer risk or non-cancer HI = 1 (whichever is lower). This
alternative also includes floodplain excavation to achieve the
more stringent ecological risk-based numerical values.
-------
Combination Alternative 6 involves the excavation of approxi-
mately 2,252,000 cubic yards of sediment, 35,000 cubic yards
of bank soil and 121,000 cubic yards of floodplain soil. This
alternative involves the excavation of approximately 387 acres
of floodplain area and also includes the backfill of 351 acres
of river bed after excavation. Institutional Controls, long-term
operation, monitoring, and maintenance are also components
of this alternative. This alternative is estimated to take 52 years
to implement. The cost for this alternative is estimated at $917
million, excluding costs for transportation or disposal of exca-
vated soil or sediment.
Combination Alternative 7
Combination Alternative 7 is a combination of Sediment
Alternative SED 9 and Floodplain Alternative FP 8, This alter-
native involves removal of approximately 2 feet of river bed
sediment followed by capping in Reaches 5A, 5B, and 5C; bank
soil removal and stabilization of Reach 5A and 5B river banks;
a combination of one foot removal followed by capping or
capping without removal in areas of the Backwaters exceeding
1 mg/kg PCBs; one to 3.5 foot removal followed by capping in
Reach 6 (Woods Pond); one to 1,5 foot removal followed by
capping in the Reach 7 impoundments and Rising Pond (Reach
8); and, MNR in all other River reaches (Reach 7 channel and
Reaches 9 through 16), This alternative differs from the other
sediment removal alternatives in that: (1) all sediment removal
and capping work, including in Reaches 5A and 5B, would be
performed in the "wet" by equipment operating in the river
(either on the river bottom or on barges), and (2) removal of
the sediment in Backwaters and Reaches 6, 7, and 8 would be
performed concurrently with removal activities in the Reach 5
channel, However, capping in those reaches would be delayed,
where necessary, until after all the removal/capping activities in
Reach 5 have been completed,
For the floodplain, Combination Alternative 7 involves the
removal of one foot of contaminated soil with subsequent
backfilling to meet a human-health based cleanup target based
on 10° cancer risk or non-cancer HI = 1 (whichever is lower)
and additional removal of soils exceeding 50 mg/kg PCBs, This
alternative also includes floodplain and vernal pool excavation to
achieve the more stringent ecological risk-based numerical values,
Combination Alternative 7 involves the excavation of approx-
imately 886,000 cubic yards of sediment, 35,000 cubic yards
of bank soil and 177,000 cubic yards of floodplain soil. This
alternative involves the excavation of approximately 108 acres
of floodplain area and also includes the capping of 333 acres of
river bed after excavation, and 3 additional acres of river bed
capping in areas not slated for excavation. Institutional Controls,
long-term operation, monitoring, and maintenance are also
components of this alternative, This alternative is estimated
to take 14 years to implement, The cost for this alternative is
estimated at $394 million, excluding costs for transportation or
disposal of excavated soil or sediment,
Combination Alternative 3
Combination Alternative 8 is a combination of Sediment Alter-
native SED 10 and Floodplain Alternative FP 9, This alternative
involves removal of approximately 2 feet of river bed sediment
followed by capping in select areas of Reach 5A and MNR in the
remainder of Reach 5A; bank soil removal and stabilization of
Reach 5A and 5B river banks; a combination of 2,5 foot removal
in areas with PCB concentrations greater than 13 mg/kg in the
top 6 inches, without subsequent capping or backfilling, and MNR
in other areas of Woods Pond; and MNR in all other River reach-
es (Reach 5B, Reach 5C, Backwaters, and Reaches 7 through 16).
For the floodplain, Combination Alternative 8 involves the
removal of one foot of contaminated soil with subsequent
backfilling to meet a human-health based cleanup target based
on 104 cancer risk or non-cancer HI = 1 (whichever is lower)
plus additional cleanup to a depth of 3 feet in certain frequently
used areas to achieve a human-health based cleanup target based
on 10° cancer risk or non-cancer HI = 1 (whichever is lower).
Combination Alternative 8 involves the excavation of approx-
imately 236,000 cubic yards of sediment, 35,000 cubic yards
of bank soil and 26,000 cubic yards of floodplain soil, This
alternative involves the excavation of approximately 14 acres
of floodplain area and also includes the capping of 20 acres of
river bed after excavation. Institutional Controls, long-term
operation, monitoring, and maintenance are also components
of this alternative, This alternative is estimated to take 5 years
to implement. The cost for this alternative is estimated at $94
million, excluding costs for transportation or disposal of exca-
vated soil or sediment.
Combination Alternative 9
EPA's Preferred Alternative (Proposed Remedial Action)
Combination Alternative 9 is a combination of Sediment Alter-
native SED 9 MOD and Floodplain Alternative FP 4 MOD. This
alternative involves removal of river bed sediment followed by
capping in Reaches 5A and 5C; bank soil removal and stabili-
zation of PCB-contaminated erodible Reach 5A river banks;
excavation of Reach 5B river bed and bank areas exceeding 50
mg/kg PCBs with EMNR (using activated carbon or other sed-
iment amendment) for remaining areas of Reach 5B sediment;
a combination of one foot removal followed by capping of the
Backwaters exceeding 1 mg/kg PCBs, excluding certain high
priority habitat areas; one to seven foot removal followed by
capping in Reach 6 (Woods Pond); excavation and/or capping
23
-------
to address Reach 7 impoundments and Rising Pond (Reach 8),
as discussed above; and, MNR in alI other River reaches (Reach
7 channel and Reaches 9 through 16).
EPA's May 2012 status report entitled "Potential Remediation
Approaches to the GE-Pittsfie!d/Housatonic River Site 'Rest of
River' PCB Contamination" (the Status Report) highlighted the
objectives of addressing the unacceptable risks posed by PCBs
and of minimizing the amount of bank excavation to preserve
the dynamic character and related biodiversity and habitats of
the river, To that end, the Status Report proposed a remedial
approach that, based on data collected prior to the issuance
of the permit, would result in an amount of bank excavation
in Reach 5A of 3,5 miles, and an amount of bank excavation
in Reach 5B of 0.2 miles. Under any alternative, the actual
remediation amounts would be determined during remedial
design, If the new data to be collected identified the need for
greater bank excavation, then the foregoing amounts of bank
excavation would change based on new data. Under Combina-
tion Alternative 9, the corrective measures for the river banks
would be designed and implemented to achieve Performance
Standards while minimizing impacts on river dynamics and
other ecological processes, and on the abundance of state-listed
and other wildlife species and the diversity of their habitats that
are supported by the existing river ecosystem,
This alternative is similar to Combination Alternative 7 and
differs from the other sediment removal alternatives in that: (1)
all sediment removal and capping work, including in Reaches
5A and 5B, would be performed in the "wet" by equipment op-
erating in the river (either on the river bottom or on barges);
and (2) removal of the sediment in the Backwaters and Reaches
6, 7, and 8 would be performed concurrently with removal
activities in the Reach 5 channel. However, capping in those
reaches would be delayed, where necessary, until after all the
removal/capping activities in Reach 5 have been completed, It
is important to note that the sediment removal depths outlined
above, for the most part, were derived based upon certain
assumptions on the estimated cap thicknesses in the various
reaches of the river, As outlined in the section entitled "Engi-
neered Cap Design" above, specific cap designs and thicknesses
will be determined based upon additional evaluations in the
future. Thus, the volume and cost estimates for this alterna-
tive outlined below could be reduced should a thinner cap be
deemed appropriate.
For the floodplain, Combination Alternative 9 involves the
removal of one foot of contaminated soil with subsequent
backfilling to meet a human-health based cleanup target based
on 10"s cancer risk or non-cancer HI - 1 (whichever is lower)
while providing for avoidance, minimization, or mitigation
of impacts in priority habitat areas for state-listed species of
concern by establishing a secondary remediation target to
meet a human-health based cleanup target based on 10"4 cancer
risk or non-cancer HI = 1 (whichever is Sower) in high priority
habitat areas, This alternative also includes additional cleanup
to a depth of 3 feet in certain frequently used areas to achieve
a human-health based cleanup target based on 10° cancer risk
or non-cancer HI = 1 (whichever is lower). This alternative also
includes vernal poo! excavation to achieve the more stringent
ecological risk-based cleanup target for amphibians,
This alternative also provides for a phased, adaptive manage-
ment approach to all remediation activities, For vernal poo!
remediation, this also includes the pilot testing of non-excava-
tion cleanup methods described previously.
Combination Alternative 9 involves the excavation of approx-
imately 890,000 cubic yards of sediment, 25,000 cubic yards
of bank soil and 75,000 cubic yards of floodplain soil, This
alternative involves the excavation of approximately 45 acres of
floodplain area and also includes the capping of approximately
298 acres of river bed after excavation to reduce the amount
of PCBs transported downstream. Pilot studies, Institutional
Controls, long-term operation, monitoring, and maintenance
are also components of this alternative. Additionally, this
alternative includes provisions for GE to maintain responsibility
for the incremental costs incurred due to the potential impacts
of PCBs on authorized activities within the Massachusetts and
Connecticut portions of the river. This alternative is estimated
to take 13 years to implement, The cost for this alternative is
estimated at $326 million, excluding costs for transportation or
disposal of excavated soil or sediment,
Treatment/Disposition Alternatives
Five alternatives were developed for treatment and/or dispo-
sition (TD) of removed sediment, riverbank soil, and floodplain
soil from the Rest of River, These alternatives are as follows:
TD 1: Off-Site Disposal in Existing Licensed Landfill(s)
(EPA's Preferred Alternative)
TD 2: Local Disposal in Confined Disposal Facility (CDF)
TD 3: Local Disposal in an On-Site Upland Disposal Facility
TD 4: Chemical Extraction
TD 5: Thermal Desorption
Alternative TD 1, disposal in an existing off-site licensed landfill
or landfills, would involve the transportation of removed
sediment and floodplain soil to commercial solid waste and/or
TSCA-licensed landfill(s) for disposal. In the CMS, GE evaluated
transport of contaminated material by trucks. In its comments,
EPA required that GE provide an evaluation of rail transport in
24
-------
the Revised CMS. GE provided a qualitative evaluation and con-
cluded that rail transport would be technically feasible; there-
fore transportation could be conducted either by trucks or by
rail. However GE did not provide cost information. EPA further
evaluated the feasibility of rail and developed a cost estimate.
This modification is also referred to in this document as TD 1
RR. The estimated cost for this alternative ranges from $55 to
$832 million for disposal via truck and $52 to $787 million for
disposal via rail, depending on which Combination Alternative
it is paired with. For the preferred sediment/floodplain alterna-
tive, the estimated cost of disposal via truck is $308 million and
via rail is $287 million.
Massachusetts' requirements regarding the disposal of contam-
inated soil and sediment have not been included as ARARs for
Alternative TD 1 since ARARs apply only to on-site activities
and the Proposed Remedial Action requires that all contaminat-
ed soil and sediment be disposed of off-site at existing licensed
facilities approved to receive such soil and sediment.
Alternative TD 2, disposition in a local in-water Confined Dis-
posal Facility/Facilities (CDF), would involve the placement of
dredged sediments in a CDF or CDFs located within the river or
backwater area. A CDF is an engineered structure consisting of
dikes or other structures that extend above an adjacent water
surface and enclose a disposal area for containment of dredged
sediments. Disposal of material that exceeds the capacity of the
CDFs would be disposed of in existing off-site licensed landfills.
The potential locations evaluated as part of this alternative are
shown in Figure 8. The estimated cost for this alternative ranges
from $100 to $510 million, depending on which Combination
Alternative it is paired with; with EPA's preferred Combination,
this alternative is estimated to cost $317 million.
Alternative TD 3, disposition in a local on-site Upland Disposal
Facility or Facilities, would involve the permanent disposition of
removed sediment/soil at an Upland Disposal Facility construct-
ed in close proximity to the River, but outside the 500-year
floodplain. The removed sediment and soil would be loaded
into trucks at the staging areas, covered, and transported over
on-site and local roadways to a nearby Upland Disposal Facility.
Three potential locations for an Upland Disposal Facility were
identified and evaluated by GE in the CMS. These sites are lo-
cated near Woods Pond, Forest Street in Lee, and Rising Pond
(referred to, respectively, as the Woods Pond, Forest Street,
and Rising Pond Sites). The potential locations evaluated as part
of this alternative are shown in Figure 8. The estimated cost
for this alternative ranges from $36 to $201 million, depending
on which Combination Alternative it is paired with; with EPA's
preferred Combination, this alternative is estimated to cost
$100 million.
Alternative TD 4, chemical extraction of PCBs from removed
sediment/soil, involves treatment of the removed sediments
and soils by a technology known as chemical extraction. In
general terms, chemical extraction is the process of mixing an
extraction fluid/solvent with removed sediment and soil, so
that PCBs in the sediment or soil are preferentially transferred
into the extraction fluid. The resulting PCB-contaminated fluid
is then treated or disposed of off-site along with treated sedi-
ments. The estimated cost for this alternative ranges from $89
to $999 million, depending on which Combination Alternative it
is paired with; with EPA's preferred Combination, this alterna-
tive is estimated to cost $399 million.
Alternative TD 5, thermal desorption of PCBs from removed
sediment/soil, would involve treatment of the removed sedi-
ments and soils by a technology known as thermal desorption.
Thermal desorption removes contaminants by raising the tem-
perature of the contaminated material to transfer the contami-
nants from the sediment or soil to a gas stream. The gas stream
is then treated to remove particulates and the organic contam-
inants. The material that remains is then sent to an appropriate
treatment/disposal facility. Treated sediments or soils may then
be disposed of in an appropriate disposal facility or potentially
reused, depending on its chemical concentrations and physical
characteristics. The estimated cost for this alternative ranges
from $103 million to $1.53 billion, depending on which Com-
bination Alternative it is paired with and how much material
is reused; with EPA's preferred Combination Alternative, this
alternative is estimated to cost between $515 and $540 million.
HOW DOES EPA CHOOSE A FINAL
CLEANUP PLAN?
Before making its recommendation, EPA coordinated with the
Commonwealth of Massachusetts and the State of Connecticut
regarding potential cleanup approaches. EPA worked closely
with the States on the development of the Performance Stan-
dards, corrective measures and identification of ARARs prior to
the issuance of this plan to the public.
EPA also held extensive discussions with GE, and solicited input
from the community through workshops and public meetings.
The timeline of these events is summarized elsewhere in this
document and information exchanged in these discussions is
also contained in the Administrative Record. The States, GE,
and the public also have the opportunity to comment on the
Proposed Remedial Action during the public comment period.
EPA used nine criteria that were established in the Permit to
compare alternatives, and propose and select a final cleanup
plan. Of the nine criteria, Overall Protection of Human Health
25
-------
J J
I :
Wwt P.th1»M
PI H SFIE l" DT*.d NdlM4(1!SAM
-------
and the Environment, Compliance with Applicable or Relevant
and Appropriate Federal and State Requirements (known as
"ARARs"), and Control of Sources of Releases are the three
General Standards for Corrective Measures. In addition, EPA
considered six other Selection Decision Factors; those factors are
as follows: Long-Term Reliability and Effectiveness, Attainment of
Interim Media Protection Goals; Reduction of Toxicity, Mobility or
Volume; Short-Term Effectiveness; Implementability; and Cost.
Following are definitions of the nine criteria from the Permit.
.¦I'.. -J _ _ . I : ¦ _ i :es
1. Overall Protection of Human Health and the Environ-
ment: How each alternative or combination of alterna-
tives would provide human health and environmental
protection, taking into account EPA's Human Health
and Ecological Risk Assessments.
2. Control of Sources of Releases: How each alterna-
tive or combination of alternatives would reduce or
minimize possible further releases, including (but not
limited to) the extent to which each alternative would
mitigate the effects of a flood that could cause contam-
inated sediments to become available for human or
ecological exposure.
3. Compliance with Applicable or Relevant and Appro-
priate Federal and State Requirements (ARARs): How
each alternative or combination of alternatives would
meet such requirements or, when such a requirement
would not be met, the basis for a waiver under CER-
CLA and the National Contingency Plan ("NCP"), per
the Consent Decree.
Selection Decision Factors
c. Any potential long-term adverse impacts of each alterna-
tive or combination of alternatives on human health or
the environment, including (but not limited to) potential
exposure routes and potentially affected populations,
any impacts of dewatering and disposal facilities on
human health or the environment, any impacts on
wetlands or other environmentally sensitive areas, and
any measures that may be employed to mitigate such
impacts.
5. Attainment of Interim Media Protection Goals (IMPGs):
The ability of each alternative or combination of alter-
natives to achieve the Interim Media Protection Goals,
including (if applicable) the time period in which each
alternative would result in the attainment of the IMPGs
and an evaluation of whether and the extent to which each
alternative would accelerate such attainment compared to
natural processes. Note that these IMPGs were used in the
comparison of remedial alternatives and are not necessarily
the same as the Performance Standards or Cleanup Stan-
dards proposed in the Draft Modification to the Reissued
RCRA Permit required to be met as part of the remedy.
6. Reduction of Toxicity, Mobility, or Volume of Wastes:
a. If applicable, treatment process used and materials treated;
b. If applicable, amount of hazardous materials destroyed or
treated;
c. If applicable, degree of expected reductions in toxicity,
mobility, or volume;
d. If applicable, degree to which treatment is irreversible; and
e. If applicable, type and quantity of residuals remaining after
treatment.
4. Long-Term Reliability and Effectiveness:
a. Magnitude of residual risk, including (but not limited
to) the extent to which each alternative would mitigate
long-term potential exposure to residual contamination,
and the extent to which and time over which each alter-
native would reduce the level of exposure to contami-
nants;
b. Adequacy and reliability of each alternative or combina-
tion of alternatives, including (i) operation, monitoring,
and maintenance requirements; (ii) availability of labor
and materials needed for operation, monitoring, and
maintenance; (iii) whether the technologies have been
used under analogous conditions; and (iv) whether the
combination of technologies (if any) have been used
together effectively; and
7. Short-Term Effectiveness: Impacts to nearby communities,
workers, or the environment during implementation of
each alternative, including (but not limited to) risks associat-
ed with excavation, transportation, dewatering, disposal, or
containment of sediments, soils, or other materials contain-
ing hazardous constituents.
8. Implementability:
a. Ability to construct and operate the technology, taking
into account any relevant site characteristics;
b. Reliability of the technology;
c. Regulatory and zoning restrictions;
d. Ease of undertaking additional corrective measures if
necessary;
e. Ability to monitor effectiveness of remedy;
27
-------
f. Coordination with other agencies;
g. Availability of suitable on-site or off-site treatment, stor-
age and disposal facilities and specialists; and,
h. Availability of prospective technologies.
9. Cost:
a. Capital costs;
b. Operating and maintenance costs; and,
c. Present worth costs.
Personnel from the Massachusetts Department of Environ-
mental Protection, the Massachusetts Department of Fish and
Game, and the Connecticut Department of Energy and Environ-
mental Protection have been consulted extensively as EPA was
preparing this cleanup proposal. Formal state and community
input on the Proposed Cleanup Plan received during the public
comment period will be considered prior to EPA issuing a final
cleanup plan.
COMPARATIVE ANALYSIS OF COMBINED
SEDIMENT/FLOODPLAIN ALTERNATIVES
This section presents a summary of a comparative evaluation of
the nine combination alternatives for river sediment and flood-
plain soil using the Permit criteria. A more detailed evaluation
of the criteria is in the Administrative Record.
O . n ' n of Hu m*' i ' ! .« : .
a nent
This criterion was evaluated taking into account the HHRA
and ERA. Combination Alternative 1 provides no protection of
human health and the environment. Combination Alternatives
2 and 8 do not adequately meet IMPGs for humans or ecologi-
cal receptors and are, therefore not protective of human health
and the environment in the long term.
In addition, Combination Alternatives 1, 2, and 8 would not
meet the federal and state water quality criterion for fresh-
water aquatic life and therefore would not be protective of
the environment. None of the alternatives analyzed would
achieve the federal and state water quality criterion for human
consumption of organisms in any of the Massachusetts reaches
while Combinations 1,2, 3, and 8 would not achieve this criteri-
on in any Connecticut impoundments. Combinations 4, 5, 6, 7,
and 9 would restore water quality consistent with this criterion
in significant segments of the river in Connecticut, based on
estimates of meeting this criterion in the future in 50% or more
of the Connecticut impoundments. See "Compliance with Fed-
eral and State ARARs" for further discussion regarding water
quality criteria.
28
Combination Alternatives 6, 7, and 9 would provide the highest
level of protection to human health and the environment be-
cause the largest volume of sediment and floodplain soil would
be addressed (by a combination of removal and capping in
place, or amended with activated carbon to reduce the bioavail-
ability of PCBs) and downstream transport would be reduced
to the greatest extent. Combination Alternatives 3, 4 and 5
would also provide protection. However, more contaminated
sediment would remain in place in the river under these alterna-
tives under thin layer caps or subject to MNR. As a result, there
is a greater chance additional releases of contaminants could
occur in the future under these Alternatives. While thin layer
capping has been used successfully at other sites across the
nation, site-specific conditions (e.g., higher PCB concentrations
and higher flows) have raised concerns about its suitability for
the Housatonic River. In addition, Combinations 1, 2, 3, 4, 5,
and 8 leave more contaminated floodplain soil in place thereby
decreasing the overall protectiveness of these Alternatives.
Unless measures are undertaken to preserve the dynamic, me-
andering character of the river and avoid, minimize and mitigate
impacts to state-listed species habitat, Combination Alternatives
that require extensive excavation in these ecological resources,
including state-listed habitats (such as Combination 6) may re-
sult in less overall protection of the environment. By employing
a more targeted remediation approach, Combination 9 pro-
vides the best balance between addressing human health risks
and ecological risks and negative impacts of remedial work on
the river's ecosystem, including its array of state-listed species
habitats. Those Combination Alternatives that have minimal
or no impact to state-listed species (Combinations 1, 2, or 8)
have much less cleanup than Combination 9 and thus provide
reduced overall protection for risks to human health and the
environment.
Combination Alternatives 2 through 9 rely to varying degrees
on Institutional Controls throughout the river in both Massa-
chusetts and Connecticut to be protective of human health in
the long term. Those alternatives that rely more extensively on
these controls (Combinations 2 and 8) over longer timeframes
and larger areas have more uncertainty that they will protect
human health in the long term, and such controls provide no
protection for ecological risks. Those alternatives (Combi-
nations 6, 7, and 9) that rely on these controls over shorter
timeframes or smaller areas have higher overall protection of
human health.
Control of So u rces of Releases
A computer model was used to predict the reductions in the
mass of PCBs passing Woods Pond and Rising Pond Dams,
respectively, and the reductions in the mass of PCBs transport-
ed from the river to the floodplain versus today's conditions in
Reaches 5 and 62. These results are summarized in Table 3 for
-------
each Combination. Table 3 also shows trapping efficiency for
solids in Woods Pond for each Combination.
As additional sources are controlled by permanently removing
and/or capping PCB-contaminated sediment and reducing the
contribution of PCBs from the contaminated eroding banks,
significant additional reductions in PCB mass transport in
the river and transport to the floodplain occurs. As a result,
Combination Alternatives 1, 2 and 8 do the least to control
releases. While Combination Alternatives 6 and 7 do the most
to control releases, Combination Alternatives 3, 4, 5 and 9 also
provide significant control of releases.
Combinations 7, 8, and 9 nearly double the solids trapping effi-
ciency of Woods Pond when compared to the other Combina-
tions. PCBs are attached to solids that move through the river
system. Therefore, the increase in trapping of solids in Woods
Pond is a mechanism to reduce downstream migration of PCBs.
It is estimated that 25% of the mass of PCBs in the river sedi-
ment are within Woods Pond. Combinations 7 and 9, and to a
lesser extent, Combination 8, also control sources of releases
by removing a significant mass of PCBs from behind the Woods
Pond dam. In the event of a serious breach or failure of the
dam, the release of PCBs downstream would be less for these
alternatives (7 through 9) than for Combinations 1 through 6
that rely primarily on capping or MNR.
The different combinations are expected to have different re-
sponses in the occurrence of an extreme flood event. Combina-
tions 1 and 2 will have no different response than what would
be expected to occur under current conditions as there is no
active remediation. In this case, PCB-contaminated sediment
and soil from eroding banks are expected to be released and
mobilized downstream. Combination 8 is expected to result in
similar, but slightly less downstream transport as it has only a
small area in Reach 5A which is addressed with an engineering
approach, and residual PCBs in Woods Pond are not capped.
Combination 3 will result in slightly less transport than the pre-
vious alternatives, however the use of a thin-layer cap in Reach
5C and Woods Pond, and MNR in Reach 5B, the Backwaters
and Reach 7 impoundments is not expected to adequately
control sources of releases in an extreme event. Combinations
4 and 5 are expected to provide adequate protection in an
extreme event in Reaches 5 and 6 but the use of thin-layer
capping and backfill in the downstream reaches provides a high
level of uncertainty in performance during such an event. Com-
bination 6 followed by Combination 7 are expected to provide
the highest level of protection of all the combinations during an
extreme event as they provide the greatest amount of remedia-
tion with corresponding engineering controls. Combination 9 is
expected to provide adequate protection in an extreme storm
event in all reaches, with the exception of Reach 5B which is
subject to MNR and therefore bed sediment and bank soil may
erode and be transported downstream. However, the areas of
the highest PCB concentrations in Reach 5B will be removed.
' .,i cs With ' ¦ 1 « 1 a..
ARARs
A summary of some of the more significant chemical-, location-,
and action-specific ARARs is included below.
Chemical-Specific ARARs
Chemical-specific ARARs include federal and state water qual-
ity criteria for PCBs. These criteria are the freshwater chronic
aquatic life criterion of 0.014 micrograms per liter (ug/L) and the
human health criterion (based on consumption of water and/or
organisms) of 0.000064 ug/L (or 0.064 parts per trillion).
Combination Alternatives 1, 2, and 8 would not achieve the
federal and state water quality criteria for freshwater aquatic
life in Massachusetts (but would in Connecticut). Combination
Alternatives 3-7 and 9 would achieve these criteria in all
reaches of the river.
None of the alternatives would achieve the federal and state
water quality criteria for human consumption of water and
organisms in the any of the Massachusetts reaches. Combi-
nations 1,2, 3, and 8 would not achieve this criterion in any
Connecticut impoundments. Based on modeling, Combination
Alternatives 4, 6, 7, and 9 would restore water quality con-
sistent with this criterion in 50% or more of the Connecticut
impoundments. Because the water quality criteria for human
consumption of organisms (0.000064 ug/L) is not expected to
be met in the River in Massachusetts under any of the alterna-
tives evaluated, EPA is proposing to waive this criterion under
both Federal and State ARARs as technically impracticable in
Reaches 5 through 9. As a modified Performance Standard for
this waived criterion, the project will be required to meet the
Biota Performance Standard and the Downstream Transport
Performance Standard in the Permit.
Current modeling shows Combination Alternatives 7 and 9
will achieve the 0.000064 ug/L criterion in at least 3 of the
4 Connecticut impoundments. However, the results from the
Connecticut model are very uncertain due to the empirical,
semi-quantitative nature of the analyses. As such, it is not
possible to predict with certainty attainment or nonattainment
of the human health criterion based on human consumption
2 The initial (i.e., current) annual PCB mass values used in the model
are 20 kg/yr passing Woods Pond Dam, 19 kg/yr passing Rising
Pond Dam, and 12 kg/yr transported from the river to the floodplain
in Reaches 5 and 6.
-------
Table 3
Percent Reduction in Annual PCB Mass Passing Woods Pond
and Rising Pond Dams and Transported to the Reach 5/6
Floodplain and Solids Trapping Efficiency for Woods Pond for Combinations of
Alternatives (relative to current conditions)
Combination:
1
2
3
4
5
6
7
8
9
Location
SED 1/
FP 1
SED 2/
FP 1
SED 3/
FP 3
SED 5/
FP 4
SED 6/
FP 4
SED 8/
FP 7
SED 9/
FP 8
SED 10/
FP 9
SED 9 MOD/
FP 4 MOD
Percent Reduction in Annual PCB Mass
Woods Pond Dam
37%
37%
94%
97%
97%
98%
97%
62%
89%
Rising Pond Dam
41%
41%
87%
93%
95%
96%
96%
62%
89%
Reach 5/6 Floodplain
50%
50%
97%
98%
98%
99%
98%
68%
92%
Solids Trapping Efficiency for Woods Pond
Solids Trapping Efficiency of
Woods Pond
15%
15%
13%
15%
15%
15%
26%
24%
30%
include a set of protocols to heip evaluate how best to avoid,
minimize and mitigate impacts as part of floodplain soil/vernal
pool remediation.
Long-Term Reliability and
Effectiveness
Combination Alternatives 1 and 2 would provide no or little
long-term reliability and effectiveness as no actions or few
actions would be taken to mitigate long-term exposure to
contamination or reduce the level of exposure to contaminants,
All other Combination Alternatives provide varying degrees of
long-term reliability and effectiveness through active cleanup
and Institutional Controls to mitigate longterm exposure to
contamination and reduce the level of exposure to contami-
nants. Of these cleanup alternatives, those Combination Alter-
natives that remove the most contaminated soil and sediment
(Combination Alternative 6, followed by Combination Alterna-
tives 7 and 9) provide the best long-term reliability and effec-
tiveness because the magnitude of the residual risk that remains
is much lower than those alternatives that leave significantly
more contaminated material in place (Combination Alternatives
3, 4, and 8, and to a certain extent, Combination 5).
However, Combination Alternatives that fundamentally impact
the dynamic, meandering character of the river or require
extensive excavation in habitats supporting state-listed species
(such as Combinations 6 and 7) may result in reduced long-
term effectiveness because of potential long-term adverse effect
on the environment. As a result, Combination 9, which includes
more excavation than most alternatives, but also provides the
most measures and procedures to preserve and protect the
30
of water and organisms of 0.000064 ug/L in Connecticut
(Reaches 10 through 16). Thus, no waiver is being proposed for
Connecticut at this time,
Location-Specific and Action-Specific ARARs
All Combination Alternatives meet action-specific ARARs.
Combination Alternatives 3 - 9 would involve temporary
destruction of wetlands and a discharge of dredged or fill ma-
terial into waters of the state and/or U.S. Of the alternatives
providing adequate risk reduction (Combinations 6, 7, and 9),
Combination Alternative 9, is the least damaging practicable
alternative under the Clean Water Act and State and other fed-
eral wetlands requirements. See additional information under
Wetland and Floodplain Impacts elsewhere in this document.
The Massachusetts Endangered Species Act (MESA), M.G.L,
c. 131 A, is applicable to all active alternatives (Combination
Alternatives 3-9). MESA and its regulations at 321 CMR 10.00
were promulgated to conserve and protect state-listed species
and their habitats. Unacceptable levels of PCBs are present in
such habitat areas in the Rest of River. During the implemen-
tation of the Proposed Remedial Action, the removal of PCBs
from the Rest of River is anticipated to provide a benefit to
state-listed species inhabiting the area due to the reduction
in adverse effects to ecological receptors from the PCBs. In
overseeing the response actions, EPA, in coordination with the
Division of Fisheries and Wildlife in the Massachusetts Depart-
ment of Fish and Game, which administers MESA, will require
that implementation of the corrective measures avoid, minimize
and mitigate impacts to state-listed species and their habitats,
as required by MESA. In particular, the proposed corrective
measures for backwaters, floodplain soils and vernal pools each
-------
river's sensitive ecosystem, including its array of state-listed
species habitats, provides the best balance in terms of reducing
residual risk and minimizing long-term ecological impacts. All
active alternatives would require restoration and compliance
with relevant ARARs to mitigate the impacts of the reme-
diation. Restoration is expected to be effective and reliable,
returning habitats to their pre-remediation state for all active
alternatives on a timeframe appropriate for the type of habitat
being restored (e.g. a floodplain forest will take longer than an
emergent wetland). Where a considerable amount of soil or
sediment remains unaddressed or under a thin-layer cap (Alter-
natives 1,2, 3, 4, 5, and 8), there would be a greater potential
for contaminated material to move downstream. As a result,
the long-term reliability and effectiveness of these alternatives
is based significantly on long-term maintenance, monitoring and
Institutional Controls. Institutional Controls in this situation (for
large areas and long time frames) are difficult to monitor and
enforce and are not appropriate in managing ecological risks. As
a result, those alternatives that rely more heavily on these con-
trols and on monitoring and maintenance (Combinations 1, 2,
3, 4, 5, and 8) may not be adequate and would be less reliable
in the long-term compared to other, more active alternatives
(such as Combinations 6, 7, and 9). Combinations 6, 7, and 9
are also more reliable in the long-term based on their removal
of a large mass of PCBs from behind Woods Pond dam.
Finally, because all active alternatives (Combinations 3-9) rely
on essentially the same components, there is no significant
difference between these alternatives in terms of availability
of labor and materials needed for operation, monitoring, and
maintenance. In addition, the components in all active alterna-
tives have been used effectively together under comparable
conditions.
With regard to timeframes to reduce exposure to contami-
nants, see the discussion under "Attainment of IMPGs" below.
¦ Lmen' . " - " s
As part of the Corrective Measures Study process, human
health Interim Media Protection Goals (IMPGs) were developed
to address cancer risk and non-cancer risk for the following
three major routes of exposure:
Direct contact with sediment and floodplain soil.
Consumption offish and waterfowl.
Consumption of agricultural products.
Current land use in the floodplain no longer includes any agri-
cultural exposures; these IMPGs would be considered if future
uses were to change to agriculture.
Two sets of ecological IMPGs were also developed: more strin-
gent "lower-bound" IMPGs and less stringent "upper-bound"
IMPGs.
An evaluation of whether, and to what extent, each alternative
would achieve IMPGs or whether an alternative would accelerate
attainment of the cleanup levels when compared to natural process-
es, or in this case Combinations 1 and 2, was conducted.
For human health direct contact risk, Combinations 3-9 meet
many more IMPGs in more floodplain and sediment areas than do
Combinations 1 and 2.
For human fish consumption, most IMPGs would continue to
be exceeded for greater than 250 years under Combinations 1,
2, 3, and 8 in Massachusetts. All other alternatives meet some
of the IMPGs far sooner than these Combinations in many
reaches, including downstream in CT, within a relatively short
time after completion of work in a particular river reach. A full
evaluation of each alternative's performance regarding fish con-
sumption based IMPGs can be found in the Administrative Re-
cord, see Figure 9 for a representative example. Table 4 shows
the modeled average fish fillet PCB concentrations at the end
of the 52-year modeling period, and Table 5 provides estimated
reductions (by percentage) for the Combination Alternatives.
For ecological receptors, some of the upper- or lower-bound
IMPGs are attained in the some of the exposure areas for
Combinations 1,2, 3, and 8. By definition, Combinations 4
and 5 are designed to meet the upper-bound ecological IMPGs
(with some lower-bound IMPGs being achieved for some
receptors) and Combinations 6 and 7 are designed to meet
the lower-bound (more stringent) ecological IMPGs. While
each alternative represents a different balance between risk
reduction and habitat protection, EPA has determined that
Combination 9 provides the best balance between meeting the
ecological IMPGs while minimizing and mitigating the impact of
the remedy on the river's ecosystem and its array of state-listed
species and habitats.
" u ction ' i 1 i * ' or
Y o 1 u m e
Treatment is not part of any of the major components (removal
and capping) of the active Combination Alternatives, except
to the extent that use of activated carbon or other sediment
amendment is used to reduce toxicity in soils or sediment.
The degree to which the Combination Alternatives would
reduce the toxicity, mobility, or volume (TMV) of PCBs is
discussed below.
31
-------
Reduction of Toxicity: None of the Combination Alternatives
with the exception of Combination 9 includes any treatment
processes that would reduce the toxicity of PCBs in the
sediment or soil. Combination 9 requires the addition of an
amendment such as activated carbon in certain components
of the remedy, including vernal pools, Reach 5B sediment, and
Backwaters. The addition of such an amendment is expected to
reduce toxicity. Since none of the other Combinations provide
for this treatment, Combination 9 surpasses all other alter-
natives in the amount of materials treated and the degree of
reductions in toxicity due to treatment.
Reduction of Mobility:_Combination Alternatives 1 and 2 do
not reduce the mobility of PCBs in the river. Combination Al-
ternatives 3-9 reduce mobility through removal, capping, back-
filling, thin-layer capping, and/or bank stabilization activities. Of
those active remedies, Combination Alternative 6 provides the
greatest reductions in mobility followed by alternatives 7 and
9. Alternatives 3 and 8 provide the least reduction in mobility of
contaminants, while Alternatives 4 and 5 provide more reduction
than Alternatives 3 and 8, but less than Alternatives 6, 7, and 9.
Reduction of Volume: Combination Alternatives 1 and 2 do not
reduce the volume of PCBs in the river and flood plain. Combi-
nation Alternatives 3-9 reduce the volume of PCB-contaminat-
ed sediment, bank soil, and floodplain soil in the Rest of River
through permanent removal of the material. Table 2 includes a
summary of the approximate removal volume and correspond-
ing PCB mass that would be removed under each alternative.
r rn - - n e s s
As no active remediation is proposed for Combinations 1 or 2,
these would not result in any short-term risks to on-site work-
ers or adverse effects to the environment or community during
implementation. For the alternatives involving construction
work (Combinations 3-9), the estimated durations of con-
struction for the alternatives evaluated range from five years
(Combination 8) to 52 years (Combination 7). Because any re-
mediation would be conducted using a phased approach, these
impacts would be spread out over the remedial action period
and area, and thus, would not last for the entire duration of the
project in all affected areas. Combinations 3-9 all have potential
short-term impacts such as truck traffic, dust, and noise. Combi-
nations 7 and 9 also have the potential for short-term increases
in PCB concentrations in fish and/or surface water during and
immediately after construction in Reach 5A. Phased construc-
tion, dust suppression techniques, and perimeter air monitoring,
and other engineering controls would be undertaken to address
potential risks from construction to the community. Standard
safety measures would be taken to protect workers as part of
any cleanup work. The alternatives that limit active remediation
(Combinations 3 and 8) would have fewer short-term impacts
than the alternatives that propose remediation across several
reaches (Combinations 4, 5, 7, and 9), while more significant
impacts would be likely with Combination 6 due to the amount
of material being removed and the duration of the work, as this
is the most extensive and lengthy alternative evaluated. Short
term impacts to the environment would be expected to be
commensurate with the areal extent and volume of soil/sedi-
ment addressed. Thus, Combinations 3, 4, 5, 8, and 9 would be
expected to have fewer adverse short-term impacts than Com-
binations 6 or 7 based on their extent of river and floodplain
remediation. Estimated construction durations for the various
alternatives are included in Table 2.
1 I , ' '!l
Combination 1 does not present any implementability issues
since no action is being taken. Combination 2, which relies on
monitoring and institutional controls, has no construction-related
implementability issues. Otherwise, the implementability of Com-
binations 3 through 9 includes the following considerations:
Combinations 3 through 9 are readily able to be constructed
and operated, relying on established technologies, though the
larger the scope of the remedy, the greater the effort required
to construct. The equipment, materials, procedures, personnel
and technologies anticipated for Combinations 3 through 9 are
all readily available. Combination 9 relies on the use of activated
carbon or other sediment amendment in certain portions of
the river, backwaters, and vernal pools, which should be readily
available. Combination 9 also provides for an adaptive man-
agement approach which includes evaluating the use of new or
emerging innovative technology during the phased implementa-
tion of the remedy.
Those alternatives that rely to a greater extent on capping,
MNR and Institutional Controls (such as Combinations 2 and
8) are less reliable than those alternatives that rely more on
removal of contamination. While the scale of necessary tem-
porary staging areas or access roads varies depending on the
extent of remediation within each Combination, no Combina-
tion would involve complications that would serve to make it
less desirable under this criterion.
In addition, habitat restoration techniques that would be a
component of Combinations 3 through 9 are available and have
been used successfully at other sites. Restoration can reliably re-
establish pre-remediation conditions for these habitats over the
timeframes of the various alternatives, which range from five
(Combination 8) to 52 years (Combination 6), using a phased
and adaptive management approach. Post-remediation monitor-
ing and maintenance will ensure that the selected restoration
32
-------
Table 4 Modeled Subreach Average Fish (Fillet) PCB Concentrations at
End of Model Projection Period
Combination:
1
2
3
4
5
6
7
8
9
SED 9
Initial
SED 1/
SED 2/
SED 3/
SED 51
SED 6/
SED 8/
SED 9/
SED 10/
MOD/ FP
Reach
Cone.
FP 1
FP 1
FP 3
FP 4
FP 4
FP 7
FP 8
FP 9
4 MOD
Fish PCB Concentration (mg/kg wet weight)
Reach 5A
18
7.3
7.3
0.3
0.3
0.3
0.2
0.3
4.2
0.3
Reach 5B
17
9.3
9.3
3.0
0.2
0.2
0.2
0.3
6.6
3.5J
Reach 5C
14
7.4
7.4
1.8
0.2
0.2
0.1
0.2
5.8
0.8
Reach 5D (Backwaters)
22
9.5
9.5
6.3
0.4
0.4
0.3
0.4
11
1.1
Reach 6
15
8.6
8.6
0.7
0.2
0.2
0.1
0.2
3.7
0.7
Reach 7
6.4 -13
2.8-6.4
2.8-6.4
0.7-2.1
0.4 - 1.6
o
k>
o
--J
0.1 -0.6
0.2-0.7
1.9-4.4
0.4 - 1.4
Reach 8
6.3
3.6
3.6
1.6
0.3
0.2
0.2
0.2
2.7
0.4
Connecticut (Bulls Bridge
Dam Impoundment)
0.4
0.2
0.2
0.04
0.01
0.009
0.007
0.009
0.1
0.02
Notes:
1. PCB concentrations shown (except for the initial concentrations) represent subreach-average values predicted by EPA's model
at the end of the model projection period (81 years for Combination 6, 52 years for all other combinations).
2. Values shown as ranges in Reach 7 represent the range of modeled PCB concentrations at the end of the projection within
each of the Reach 7 subreaches. For Combination 9, the Reach 7 reductions were calculated separately by subreach.
Individual subreach initial concentrations were not provided by GE in the CMS, so reductions shown for Combination 9 were
calculated from EPA model results.
3. For Combination 9, the Reach 5B PCB concentrations do not factor in the use of an amendment, such as activated carbon.
The use of this amendment is expected to reduce fillet PCB concentrations to less than the 3.5 mg/kg predicted by the
modeling; the modeling does not factor in the effects of the amendment.
4. The results from the Connecticut model are very uncertain due to the empirical, semi-quantitative nature of the analysis.
Table 5 Percent Reductions in Fish PCB Concentrations for Combinations of Alternatives
Combination:
1
2
3
4
5
6
7
8
9
Reach
SED 1/
FP 1
SED 2/
FP 1
SED 31
FP 3
SED 5/
FP 4
SED 6/
FP 4
SED 8/
FP 7
SED 91
FP 8
SED 10/
FP 9
SED 9
MOD/ FP 4
MOD
Percent Reduction in Fish PCB Concentration Relative to Initial Conditions
Reach 5A
60%
60%
99%
99%
99%
99%
98%
77%
99%
Reach 5B
47%
47%
83%
99%
99%
99%
98%
62%
80%
Reach 5C
48%
48%
87%
99%
99%
99%
99%
59%
94%
Reach 5D (Backwaters)
57%
57%
72%
98%
98%
99%
98%
51%
95%
Reach 6
44%
44%
95%
99%
99%
99%
99%
76%
95%
Reach 7
45 - 63%
45 - 63%
80-91%
84 - 97%
94 - 98%
94 - 99%
93 - 98%
59 - 75%
86 - 95%
Reach 8
43%
43%
75%
95%
97%
97%
96%
57%
94%
Connecticut (Bulls Bridge
Dam Impoundment)
60%
60%
91%
97%
98%
98%
98%
73%
95%
Percent Reduction in Fish PCB Concentration Relative to Combination 1 or Combination 2 (MNR)
Reach 5A
96%
96%
96%
97%
96%
42%
96%
Reach 5B
68%
98%
98%
98%
97%
29%
61%
Reach 5C
76%
97%
97%
99%
97%
22%
89%
Reach 5D (Backwaters)
34%
96%
96%
97%
96%
-16%
89%
Reach 6
92%
98%
98%
99%
98%
57%
91%
Reach 7
67 - 75%
75 - 86%
89 - 93%
91 - 96%
89 - 93%
31 - 32%
75 - 88%
Reach 8
56%
92%
94%
94%
94%
25%
87%
Connecticut (Bulls Bridge
Dam Impoundment)
80%
95%
96%
97%
96%
50%
81%
Notes:
1. Percent reduction represents the change in annual average PCB concentrations predicted by EPA's model between the
beginning and the end of the projection period,
2. The results from the Connecticut model are very uncertain due to the empirical, semi-quantitative nature of the analysis.
33
-------
WHAT'S THE DIFFERENCE BETWEEN IMPGS AND PERFORMANCE STANDARDS?
This Statement of Basis and the Draft Modification of the RCRA Permit include discussion of two related measures for the
Rest of River remedy - the Interim Media Protection Goals (IMPGs), and the Performance Standards.
In the investigation of Rest of River, EPA completed a Human Health Risk Assessment (HHRA) and an Ecological Risk
Assessment. Taking into account the conclusions of the risk assessments, GE was required to propose iMPGs, which consist
of preliminary goals that are shown to be protective of human health and the environment, and which served as points of
departure in evaluating potential corrective measures in the Corrective Measures Study. Most of these IMPGs were identi-
fied as residual PCB concentrations in sediment, soil, or environmental media (like fish fillet tissue) across numerous risk-
based benchmarks, including cancer risk (at 10'5, 10s, and 1 (>4 risk levels) across a number of exposure scenarios (residential,
recreational, etc.), non-cancer risks, and ecological risks calculated at an "upper bound" (less stringent) and "lower bound"
(more stringent) risk level. The discussion in the "Comparative Analysis of Combined Sediment/Floodplain Alternatives" in
this document includes a discussion of how each alternative performs in attaining these various IMPGs.
In the Draft Permit, EPA adopts certain of these IMPGs as Performance Standards. GE will be required to meet these and
other Performance Standards as part of the remedy, as outlined in more detail in the Draft Permit. See Section II as weii as
Tables 1 through 4 of the Draft Permit for specific details.
One example of the relationship of the IMPGs and the Performance Standards is the following, in the HHRA, EPA evalu-
ated risks to humans from consuming PCB-contaminated fish tissue. GE used the information from the HHRA to develop
the IMPGs for fish consumption, which are presented as a range of concentrations associated with different risk levels that
correspond to different consumers and to different points on the EPA risk range. IMPGs were developed for both determin-
istic and probabilistic risk analyses. The range of concentrations for probabilistic IMPGs is shown on Figure 9. EPA selected
one point in this range of concentrations to serve as the Performance Standard for fish consumption, the PCB concentration
of 1.5 mg/kg in fish fillet tissue which is associated with the non-cancer probabilistic risk for the average adult fish consumer
who is assumed to consume 14 fish meals per year, haif of those from the Housatonic River. This Performance Standard is
met when fish fillet concentrations are less than 1.5 mg/kg in all Reaches. Other fish tissue IMPGs were retained as bench-
marks in the Draft Permit, whereas other IMPGs for fish tissue were not carried over into the Permit.
O)
E.
¦S2
Jy
iZ
c
c
o
c
0)
o
c
o
O
CD
o
a
100
10
0.1
0.01
0.001
0.01
Cancer 1 (VMS RME
0.001
30
Years of Simulation
Combination 1
-Combination 6
-Combination 2
-Combination 7
- Combination 3
- Combination 8
Combination 4
¦ Combination 9
Combination 5
Figure 9
Average Fillet PCB Concentrations in Largemouth Bass (Average for Fish Ages 5 to 9)
Compared to Probabilistic IMPGs
34
Probabilistic IMPGs
: 10
-j Cancer 10A-4 CTE
Non-Cancer Adult CTE
1 MA Advisory
Non-Cancer Child CTE
Cancer 10A-4 RME
Cancer 10"-5 CTE
Non-Cancer Adult RME
31 Cancer 10^5 RME
Non-CancerChild RME
Cancer 10A-6 CTE
CT Advisory
-------
techniques reestablish the prior conditions and functions of the
affected habitats.
None of the Combinations preclude the implementation of
additional corrective measures if deemed necessary. Additional
corrective actions, such as cap or bank repairs, if necessary,
should provide the same implementation challenges for all
active alternatives.
EPA anticipates a robust monitoring program to monitor the ef-
fectiveness of the remedy. Each of the components of the active
remedy combinations (Combinations 3-9) can be monitored
effectively. However, alternatives that have little or no active re-
mediation are less reliable, therefore, they would require more
extensive monitoring.
No regulatory and/or zoning restrictions are known that
would affect the implementability of the sediment/floodplain
Combinations. Implementation of all alternatives (except
alternatives 1 and 2) would require GE to obtain access from all
property owners. Issues associated with obtaining access would
be similar for alternatives 3-9, except that alternative 9 avoids
the large-scale use of sheet pile and large cranes, which may
facilitate access negotiations.
-All of the combinations would require coordination with EPA
and state agencies to ensure compliance with state ARARs. In
addition, implementation of Institutional Controls, obtaining
access to State and municipally-owned properties, conduct-
ing public/community outreach programs and implementing
biota consumption advisories will require both state and local
coordination. The alternatives that require a greater extent
of remediation and a longer implementation time would likely
require more extensive and prolonged coordination activities.
However, implementation of Institutional Controls where less
remediation is performed would require more extensive Institu-
tional Controls.
Lastly, regulatory and zoning restrictions, state and local coordi-
nation related to treatment, storage and disposal facilities, and the
availability of suitable of such facilities and specialists is discussed
below in the evaluation of Treatment/Disposition alternatives.
Cost
Estimated total and present worth for all of the Combination
Alternatives are presented in Table 6. In addition, costs asso-
ciated with these Combinations coupled with the Treatment/
Disposal Alternatives can be found in Table 7. The costs are
based primarily on information available at the time of the
estimate and are based on GE's unit cost estimates provided in
GE's Revised CMS. As shown in Table 6, Combination 1 is the
least costly alternative while Combination 6 is the most costly.
For purposes of direct comparison of treatment and disposal
costs associated with EPA's preferred sediment and flood plain
alternative, total treatment/disposal costs for Combination
Alternative 9 have also been included in Table 7.
COMPARATIVE ANALYSIS OF TREATMENT/
DISPOSITION ALTERNATIVES
This section presents a summary of a comparative evaluation of
the five alternatives for treatment and/or disposal of excavated
contaminated river sediment and floodplain soil using the same
criteria that were used for the sediment/floodplain combina-
tion alternatives. All five alternatives would involve disposition
of the sediment, riverbank soil, and floodplain soil in a disposal
facility, either directly or after treatment. The three alternatives
involving disposal only are TD 1 /TD 1 RR (off-site disposal
in permitted landfill(s)), TD 2 (on-site in a Confined Disposal
Facility (CDF)), and TD 3 (on-site in upland disposal facility or
facilities). The other two alternatives would involve treatment,
either by a chemical extraction process (TD 4) or by thermal
desorption (TD 5), followed by disposition of the byproducts of
the treatment and the treated soil/sediment.
. - ¦ . . . ' , i >, ma
a e Environment
TD 1, 3 and 5 would provide high levels of protection to human
health and the environment because all excavated contaminat-
ed material would either be removed from the site (TD 1),
contained in an upland disposal facility (TD 3), or treated to
levels safe for off-site disposal or potential reuse (TD5). TD 2
could also provide human health protection as long as monitor-
ing, maintenance and/or Institutional Controls are effective in
the long term, in order to avoid negative impacts to the river
system. Alternative TD 4 (chemical extraction) may not be able
to effectively treat PCB contamination from the site, calling into
question the protectiveness of this alternative.
t i 1 , II r
All the treatment/disposal alternatives would control the po-
tential for PCB-contaminated sediment and soil to be released
and transported within the river or onto the floodplain, al-
though some alternatives would provide more effective control
of such releases than others. TD 1 best meets this criterion,
followed by TD 3.
Under TD 1, placement of the removed PCB-contaminated sed-
iment and soil into a licensed off-site landfill or landfills would
effectively isolate those materials from being released into the
Housatonic River and associated floodplain. Under TD 2, there
is a potential for releases of sediment into the river during the
CDF construction process. TD 3 would address future releases
through the placement of the materials in an upland disposal
facility that will have a double liner and the implementation of a
long-term monitoring and maintenance program. Placement of
35
-------
the PCB-contaminated sediment and soil into an upland dispos-
al facility could effectively isolate the removed materials from
being released into the environment. However, there is the
potential for PCB releases to the Housatonic watershed If the
landfills are not properly operated, monitored and maintained.
Under TD 4 and TD 5, the potential for the PCB-contaminated
sediment and soil to be released within the river or onto the
floodplain during treatment operations would be minimal as
long as these facilities are properly operated and maintained.
Compliance with Federal and State
ARARs
The ARARs identified for the treatment/disposal alternatives
are discussed in more detail in the Administrative Record, Each
of the TD alternatives would involve moving the sediment,
bank soil, and floodplain soii from the point of excavation to
the treatment/disposition point. Of all the disposal alternatives
(TD 1, TD 2, TD 3), only TD 1 complies with all State ARARs.
TD 4 and TD 5 could potentially meet all ARARs. TD 2 will not
meet, without limitation, wetland and floodplain requirements;
and not all potential locations of TD 2 or TD 3 will meet the re-
quirements of 310 CMR 30.700, 310 CMR 16.40(3)(4), and/or
990 CMR 5.04, which prohibit, without limitation, hazardous
waste and solid waste facilities in an Area of Critical Environ-
mental Concern ('ACEC") or adjacent to or in close proximity
to an ACEC such that it would fail to protect the outstanding
resources of an ACEC.
Long-Term Reliability and Effectiveness
TD 1, 4, and 5 result in the greatest reductions in residual risk.
With TD 1, all material is removed from the site and sent to
an offsite disposal facility; with TD 4 and TD 5, all material
that was treated but did not reach safe PCB levels would be
removed from the site and sent to an offsite disposal facility.
Contamination remains on-site untreated under TD 2 and
TD 3 and therefore the residual risk is greater under these
alternatives. However, TD 3 would permanently isolate those
materials from direct contact with human and ecological
receptors in a secure location outside the floodplain. Under
1 D 4 and TD 5, residual risk is decreased because treatment
reduces the levels of contaminants, however the reductions
may not be to levels allowing for unrestricted reuse.
There are considerable differences in the adequacy and
reliability of the five treatment/disposal alternatives. TD 1 is
adequate and reliable because it does not rely on operation,
monitoring, and maintenance requirements (except at the re-
ceiving facility) to adequately and reliably address the contam-
ination. The other alternatives rely on operation, monitoring,
and maintenance requirements to address the contamination
remaining onsite to be effective in the long-term. Both TD 4
and TD 5 rely on these requirements to ensure that material
is safely treated to acceptable concentrations. TD 2 and TD
3 rely particularly on monitoring and maintenance in the long
Table 6 Cost Summary for Combinations of Sediment and Floodplain Alternatives
Combination;
1
2
3
4
5
6
7
8
9
SED
1/FP 1
SED 21
FP 1
SED 31
FP 3
SED 5/
FP 4
SED 6/
FP 4
SED 8/
FP 7
SED 9/
FP 8
SED 10/
FP 9
SED 9
MOD/FP 4
MOD
Total Capital Costs
0
0
S167 M
$307 M
$384 M
S900 M
$381 M
$84 M
$314 M
Total Operations
Monitoring and
Maintenance Costs
0
$5 M
$10 M
$12 M
$13 M
$17 M
$13 M
$10 M
$12 M
Total Cost for
Alternative
(excluding
Transportation
and Disposal)
0
$5 M
$177 M
$319 M
$397 M
$917 M
$394 M
$94 M
$326 M
Total Present
Worth
0
$1.8 M
$133 M
$193 M
$219 M
$300 M
$251 M
$78 M
$228 M
Notes:
1 All costs are in 2010 dollars. $ M = million dollars.
2. Total capital costs are for engineering, labor, equipment, and materials associated with implementation,
3. Total OMM costs include cost for monitoring, post-construction inspections and repair activities (if necessary), long-term
monitoring {fish, sediment, water column, visual), and for the maintenance of institutional controls and EREs.
4. Total present worth cost is based on using a discount factor of 7%, considering the length of the construction period and an
OMM period of 100 years on a reach-specific basis.
5. Estimates do not include costs for treatment or disposition of any soil/sediment removed; those costs are outlined below (see
Table 7),
-------
term to ensure that material remains adequately contained, and
TD3 may require long-term transport of leachate to the GE
facility in Pittsfield or construction of a separate facility to treat
leachate.
Labor and materials are available for operation, monitoring, and
maintenance for all of these alternatives. While TD 1, 2, 3 have
been used under similar conditions, TD 4 has not been demon-
strated at full scale on sediment and soil representative of those
in the Rest of River. TD 5 has been used to treat PCB-contam-
inated soil but only in limited cases for treatment of sediment,
thereby creating some uncertainty regarding the adequacy and
reliability of this alternative.
None of the alternatives are expected to have long-term
adverse impacts on human health, however TD 2 will have sig-
nificant long-term impacts on wetlands and floodplain areas. TD
3 may have long-term environmental impacts depending upon
where the upland facility is located.
¦ ¦ . men I , " "
Attainment of IMPGs is directly applicable to the sediment and
floodplain remediation approaches outlined and evaluated for the
various Combination Alternatives discussed earlier in this doc-
ument. IMPG attainment is not directly applicable to the trans-
portation and disposal alternatives, thus EPA did not conduct a
comparative analysis for these alternatives for this criterion.
Reduction of Toxicity, Mobility, or Volume
Reduction of Toxicity: TD 1 through TD 3 would not include
any treatment processes that would reduce the toxicity of, or
directly affect, PCB concentrations in the removed sediment
and soil. TD 4 and TD 5 would incorporate treatment process-
es that can, to varying degrees, reduce concentrations of PCBs.
Under TD 4, the chemical treatment process would reduce
the toxicity of the sediment and soil by permanently removing
some PCBs from these materials but likely will not reduce con-
centrations to levels allowing reuse of the material, and as such
would still require landfilling. Under TD 5, the thermal desorp-
tion system would reduce the toxicity of the PCB-contaminated
sediment and soil by permanently removing PCBs from these
materials. The PCBs in the liquid stream would be sent to a
licensed off-site disposal facility for additional treatment. The
degree of expected reduction in toxicity, and the amount of
hazardous materials to be destroyed or treated are dependent
on the sediment/floodplain alternatives selected, with Combi-
nations 3 through 9 providing varying levels of expected remov-
al of PCBs from the River and floodplain. For TD4 and TD5,
the treatment process would be irreversible and the reduction
in toxicity would be permanent.
Reduction of Mobility:_AII of the alternatives would reduce the
mobility of PCBs in the sediment and soil. In TD 1, TD 2, and TD
3, these materials would be removed and disposed of in off-site
permitted landfill(s) (TD 1) or contained within on-site CDF(s)
(TD 2) or an on-site upland disposal facility (TD 3). TD 4 and TD
5 would reduce the mobility of PCBs present in the sediment/
soil via ex-situ chemical extraction or thermal desorption.
Reduction of Volume: TD 1, TD 2, and TD 3 would not reduce
the volume of PCB-contaminated material, although, TD 1
would reduce the volume of material that remains at the Site.
For TD 4, treatment of sediment/soil would reduce the volume
of PCBs present in those materials by transferring some of the
PCBs to an aqueous waste stream for subsequent treatment.
PCB-contaminated sludge would be generated from the waste-
water treatment system and would be sent to a permitted
off-site facility for disposal. For TD 5, treatment of sediment/
soil in the thermal desorption system would reduce the volume
of PCBs present in those materials, with the liquid condensate
transported to an off-site facility for destruction.
i r m ' ' . 1
Each of the alternatives has the potential for short-term impacts
to the community. Alternatives that require on-site treatment
(TD 4 and TD 5) require operation of a treatment facility, which
would have air emissions albeit at very low levels, which could be
treated prior to discharge if needed to meet regulatory levels.
Alternatives that require on-site containment (TD 2 and TD 3)
would also have additional short-term impacts to the areas and
community surrounding the disposal sites. Construction of such
facilities will temporarily increase community impacts during the
time work is done in these areas. The alternative with off-site
disposal (TD 1/TD 1 RR) will have short-term impacts during
transport of the waste material; however, the impacts of truck
traffic may be greatly reduced by reliance on rail transporta-
tion. The short-term impacts to workers are all relatively the
same under all alternatives. All alternatives have the potential
for accidental releases of various PCB-contaminated materials
during transportation to off-site or local disposal or treatment
facilities. However, actions will be taken to prevent these poten-
tial releases. All alternatives would require truck traffic. TD 1
and TD 4 require transportation of the most material, followed
closely by TD 5, then TD 3. Depending on the location of the
upland disposal facility under TD 3, TD 3 may have truck traffic
comparable to TD 1. The impacts of truck traffic may be greatly
reduced by reliance on rail transportation, consistent with EPA's
intention to maximize use of rail.
There are also some differences in impacts to the environment
under the different alternatives. TD 2 through TD 5 could
37
-------
cause permanent loss of habitat and loss or displacement of
wildlife in the area depending upon where the disposal or treat-
ment facility is located. TD 1 would have fewer impacts on the
environment than the other alternatives.
1 i , i * ; n a fi : , ' r . j
The implementability of TD1 through TD5 includes the follow-
ing considerations:
All of the alternatives are readily able to be constructed and
operated, with the acknowledgement that for off-site disposal
via rail, some of the rail lines will need to be upgraded.
The reliability of technologies depends on the specific alterna-
tive. TD 1 and TD 3 are both reliable landfilling technologies.
CDFs (TD 2) have been implemented at many locations and
have been shown to be reliable when constructed and operated
properly. For both TD 4 and TD 5, there are several uncer-
tainties regarding the reliability of full-scale application of both
chemical and thermal processes to sediment (e.g., moisture
content), particularly with some of the volumes associated with
the sediment alternatives.
Regarding regulatory and zoning requirements and coordina-
tion with other agencies, the existing licensed off-site facility in
TD 1 would already have satisfied regulatory requirements.
Coordination with state and local agencies would be required
to site the rail loading facility. Both state and local communi-
ties have expressed a strong preference for rail, which should
facilitate resolution of any remaining regulatory, zoning, access
or facility siting issues.
TD 2 could raise issues in accounting for sufficient flood storage
compensation at the appropriate elevations/areas to provide
for construction of a CDF(s) large enough to hold the neces-
sary sediment disposal volumes, and permanent access to the
CDF(s) would be required for inspections and maintenance.
As discussed in the Compliance with Federal and State ARARs
section above, TD 3 would have significant issues with the
ACEC regulations, the Massachusetts Hazardous Waste Facility
Site Safety Council Regulations, and the site suitability criteria
in the Commonwealth's Site Assignment Regulations for Solid
Waste Facilities. In addition, TD 2 and TD 3 would both require
extensive coordination with state and local officials, increasing
the period of time before these could be implemented, and
both TD 2 and TD 3 would likely encounter significant local
and state opposition that may render these alternatives more
difficult, and potentially not feasible, to implement.
TD 4 and 5 would require access to large areas for the con-
struction and operation of a treatment facility. Locating such a
facility would require coordination would state and local agencies.
Other access and zoning issues may also be present. Since state
and local officials have expressed a strong preference for off-site
disposal, these alternatives may encounter significant opposition,
thus rendering these alternatives difficult to implement.
Regarding the availability of licensed off-site disposal facilities
(TD 1) while the current universe of facilities is sufficient, there
are uncertainties regarding the future availability of the neces-
sary capacity in off-site landfills for the alternatives that have
larger volumes and longer durations.
For TD 2 and TD 3, the availability of on-site disposal facilities
may be limited by opposition from state and local officials and
regulatory issues, as discussed above. However, if these ob-
stacles are overcome, there is sufficient availability of facilities
for TD 3. There may be limitations on the capacity of CDF(s)
depending on the combination alternative selected.
Regarding the ease of undertaking additional corrective
measures, if necessary, if additional wastes were generated as
part of future actions, it is likely that the facilities construct-
ed under TD 2 through TD 5 would no longer be available
for additional treatment and/or disposal. While it may be
technically feasible to expand an upland disposal facility after
closure (TD 3), it would likely be administratively difficult and
not cost-effective to implement this option. Thus, TD 1 is the
most implementable in this regard.
TD 1 through TD 5 all can be monitored effectively. TD 1
would require the least amount of monitoring. TD 2 and TD 3
would require extensive long-term monitoring to ensure the in-
tegrity and effectiveness of the disposal facility(s). TD 4 and TD
5 would require extensive monitoring of the treatment facilities
during treatment operations.
Cost
The estimated cost ranges for each treatment/disposal alterna-
tive, including total capital cost, estimated annual maintenance
and monitoring cost, and total estimated present worth are
summarized in Table 7. These costs are expressed as ranges
since they account for treatment or disposal of a wide range
of volumes depending on the sediment and flood plain remedi-
ation approach selected. As shown in Table 7. TD 3 is the least
costly alternative while TD 4 and TD 5 are the most costly.
For purposes of direct comparison of treatment and disposal
costs associated with EPA's preferred sediment and floodplain
alternative, total and present worth treatment/disposal costs
for Combination Alternative 9 are also included in Table 7.
38
-------
Table 7 Cost Summary for Treatment/Disposition Alternatives
TD 1
TD 1 RR
TD 2
TD 3
TD 4
TD 5
(with reuse)
TD 5
(without reuse)
Total Capital Costs
0
$300,000
$6-20 M
$10-67 M
$17-20 M
$20 - 232 M
$20 - 232 M
Total Disposal,
Operations,
Monitoring, and
Maintenance Costs
$55 - 832 M
S52 - 787 M
$94-490 M
$26-134 M
$72 - 979 M
S83- 1,216 M
$86- 1,293 M
Total Cost for
Alternative
$55 - 832 M
$52-787 M
$100-510 M
$36 - 201 M
$89 - 999 M
$103-1,450 M
$106-1,530 M
Total Present
Worth
$40 - 220 M
$38- 210 M
$46-131 M
$17-49 M
370-286 M
$81 - 569 M
$83 - 590 M
Total TD Cost for
Combination 9
$308 M
$287 M
$317 M
$100 M
$399 M
$515 M
$540 M
Total Present
Worth for
Combination 9 TD
Cost
$196 M
$183 M
$85 M
$33 M
$170 M
$280 M
S295M
Notes:
1. All costs are in 2010 dollars, except total present worth values. S M = million dollars,
2. The fraction of TSCA material has been assumed to be 35%. A density of 1.62 tons per cubic yard was assumed.
3. The Massachusetts hazardous waste transport fee is not included in these estimates. The fee would potentially apply to TSCA
material transported off-site via truck. This fee would potentially apply to TD-1, and portions of TD-2, TD-4 and TD-5. The fee is
currently $56.25 per ton, including a vehicle identification fee. For TD 1 for Combination 9, the total fee is estimated to be $31.3
million. The fee is not applicable to off-site disposal via rail {TD 1 RR).
4. With the exception of TD 2. the ranges of costs presented are the minimum and maximum anticipated costs based on the
potential range of volumes that would be potentially removed under the sediment and floodplain soil alternatives (191,000 cubic
yards to 2.9 million cubic yards). For TD 2, the lower-bound costs are based on the combined volume of SED 6 and FP 2 and the
upper-bound costs are based on the combined volume of SED 8 and FP 7, with material not placed in the CDF{s) assumed to be
transported off-site for non-TSCA disposal. Thus, the upper-bound costs, but not the lower-bound costs, for TD 2 are comparable
to the costs for the other alternatives.
5. Total capital costs are for engineering, labor, equipment, and materials associated with implementation,
6. Total operations costs consist of the total of the average annual costs for operation, placement, and/or treatment of sediment
and/or soil, estimated for the range of durations for implementing the alternatives.
7. Total monitoring and maintenance costs are for performance of post-closure monitoring and maintenance programs of 100 years
for TD 2 and TD 3 and 5 years for TD 4 and TD 5.
8. Total present worth cost is based on using a discount factor of 7%, considering the range of total potential durations for the
alternative, and post-closure monitoring and maintenance periods of 100 years for TD 2 and TD 3 and 5 years for TD 4 and TD 5.
9. For TD 5 with reuse, it is assumed that approximately 50% of the floodplain soil treated by thermal desorption would be reused
on-site and that all remaining materials would be transported off-site for disposal.
10. Costs for TD 3 do not include the very likely extensive costs associated with the approval process required for an on-site landfill.
39
-------
EPA IS ASKING FOR PUBLIC COMMENT ON THE FOLLOWING PROPOSED REGULATORY DETERMINATIONS
Wetland Impacts
The cleanup plan proposed by EPA includes
activities that impact wetlands, including vernal
pools. Before EPA can select a cleanup plan
that will impact wetlands, Section 404 of
the Clean Water Act and Executive Order
11990 (Protection of Wetlands) require that
EPA make a determination that there is no
practicable alternative to conducting work that
will impact wetlands. EPA has determined that
because significant levels of contamination exist
in wetlands within the site's cleanup areas,
there is no practicable alternative to conduct-
ing work in these wetlands.
For those wetland areas that would be impact-
ed by cleanup activities, EPA is also required
to make a determination that the cleanup
alternatives that are conducted are the least
environmentally damaging practicable alterna-
tives. EPA has determined that the proposed
cleanup alternative (Combination Alternative
9) is the least environmentally damaging practi-
cable alternative.
EPA would minimize potential harm and avoid
adverse impacts on wetland resources, to the
extent practical, by using best management
practices to minimize harmful impacts on the
wetlands, wildlife or habitat. Wetlands would
be restored and/or replicated nearby consis-
tent with the requirements of federal and state
wetlands protection laws.
A more detailed discussion of this determina-
tion is in the Wetlands/Floodplain Analysis
that is part of the Administrative Record.
Floodplain Impacts
For any cleanup alternative with impacts on
the floodplain, EPA is required to determine
whether the activities proposed will result in
occupancy or modification of the floodplain,
and if so, before EPA can select such a cleanup
alternative, Executive Order 11988 (Floodplain
Management) requires EPA to make a determi-
nation that there is no practicable alternative
to doing so.
Portions ofthe Proposed Remedial Action will
take place in floodplains. As described in the
Proposed Remedial Action, the Rest of River
area includes many contaminated floodplains.
In addition, for other components ofthe
Proposed Remedial Action, such as excava-
tion and capping ofthe riverbed sediments,
support activities are proposed to take place
in the floodplain, such as use of temporary
access roads to access the contaminated
riverbed and to transport contaminated soil
and sediment away from the project areas, as
well as establishing staging areas for contam-
inated material. These features would not
be permanent, and would be restored after
remediation.
EPA has considered alternatives to avoid po-
tential adverse effects in the floodplains. Based
on its evaluation ofthe alternatives identified
in the revised Corrective Measures Study, EPA
has determined that there is no practicable
alternative to the activities that take place in
the floodplain in the proposed alternative.
To reduce the levels of PCB contamination
in floodplain to acceptable levels, there is no
practicable alternative to excavation ofthe
limited amount of floodplain soil as identified in
the Proposed Cleanup Plan. In addition, with
respect to road construction and other ancil-
lary support areas such as staging areas for ex-
cavated soils and sediments, such construction
of roads and ancillary support areas would not
be permanent, and would be restored after
remediation. To the extent that the limited ac-
tivities to remove PCB contamination from the
floodplain, and the support activities for other
proposed Rest of River remediation activities
are considered occupancy and modification
ofthe floodplains, EPA has determined there
is no practicable alternative to occupancy and
modification.
The activities in the proposed Cleanup Plan
that affect the floodplain are not permanent,
and would be subject to mitigation following
remediation. Section II.B.3 ofthe draft Permit
Modification, Restoration of Impacted Areas,
provides specific requirements for GE to
address areas such as floodplain areas that
have been impacted by remediation activities.
In addition, the Proposed Cleanup Plan is de-
signed to minimize impacts on flood storage
capacity from cleanup activities within the
100-year floodplain. For example, the engi-
neered cap proposed in Reach 5 ofthe River
will be designed and placed so that it will not
decrease flood storage capacity. In addition,
among the applicable or relevant and appro-
priate requirements ("ARARs") that must be
complied with for the project is compliance
with standards on floodplain management,
to ensure, among other requirements, that
activities will not cause a loss of flood storage
capacity or increase in water surface elevation
or velocity.
A more detailed discussion ofthis determina-
tion is in the Wetlands/Floodplain Analysis
that is part ofthe Administrative Record.
Technical Impracticability Determination
of Certain Water Quality Criteria
The federal and state water quality criteria
standard for PCBs for human consumption of
organisms is 0.000064 micrograms per liter
(ug/L). Under any ofthe alternatives evaluat-
ed, these criteria are not expected to be met
in the River in Massachusetts.
Therefore, EPA, with concurrence from
MassDEP, proposes waiving these criteria
within Reaches 5 through 9 (the portion ofthe
site in Massachusetts) because achievement of
the federal and state water quality criteria for
human consumption of organisms is technically
impracticable from an engineering perspective
under Section 121(d)(4)(C) ofthe Compre-
hensive Environmental Response, Compensa-
tion, and Liability Act. Meeting these criteria
is not feasible from an engineering perspective
because current engineering methods cannot
be reasonably implemented to achieve these
levels in surface water in Massachusetts. As a
modified Performance Standard for this waived
criterion, the project will be required to meet
the Biota Performance Standard and the
Downstream Transport Performance Standard
in the Permit.
These criteria are not being waived in Con-
necticut because they can potentially be met in
the future. Such a waiver may be considered
in the future should it become apparent that
these criteria cannot be met based on technical
impracticability. The federal and state water
quality criteria based on freshwater aquatic
life, 0.014 ug/L, is believed to be achievable
throughout the river and will apply to this
cleanup action.
Proposed Finding: PCB Cleanup is
Protective
Pursuant to regulations implementing the
Toxic Substances Control Act at 40 C.F.R.
Section 761.61, EPA has made a draft determi-
nation that the proposed manner of sampling,
storage, cleanup and disposal of PCBs outlined
in the Draft Permit, including the Performance
Standards and associated corrective measures
to meet the Performance Standards outlined
therein, will not pose an unreasonable risk of
injury to health or the environment. See At-
tachment D to the Draft Permit for the draft
determination.
40
-------
41
-------
KEY CONTACTS
The Administrative Record, which includes all documents that
EPA has considered or relied upon in proposing this cleanup
plan, is available for public review and comment at the following
locations:
EPA Records and Information Center
5 Post Office Square, First Floor
Boston, MA 02109-3912
617-918-1440
EPA c/o Weston Solutions
10 Lyman Street
Pittsfield, MA 01201
413-442-4224
The Administrative Record will also be available
at the following locations:
Berkshire Athenaeum Public Library
Reference Department
Pittsfield, MA 01201
413-499-9480
Jim Murphy
EPA New England
Superfund Community Involvement
617-918-1028
murphy.jim@epa.gov
Kelsey O'Neil
EPA New England
Superfund Community Involvement
617-918-1003
oneil.kelsey@epa.gov
Michael Gorski
Massachusetts Department of Environmental Protection
413-755-2213
Michael.Gorski@state.ma.us
Mark Tisa
Massachusetts Department of Fish and Game,
Division of Fisheries and Wildlife
508-389-6363
Mark.Tisa@state.ma.us
Cornwall Public Library
Cornwall, CT 06796
(860) 672-6874
Housatonic Valley Association
Cornwall Bridge, CT 06754
(860) 672-6678
Dennis Schain, Director of Communications
Connecticut Department of Energy and
Environmental Protection
860-424-3110
Dennis.schain@ct.gov
Massachusetts Department of Environmental Protection
436 Dwight Street
Springfield, MA 01103
413-784-1100
Connecticut Department of Energy and
Environmental Protection
79 Elm Street
Hartford, CT 06106
860-424-3854
Information is also available for review on-line
at www.epa.gov/region 1 /ge
42
-------
P D
ACRONYMS
EPA will accept public comments during a formal comment pe-
riod on the Proposed Remedial Action. This comment period
is currently slated to run from June 25, 2014 through August
8, 2014. EPA considers these comments to improve its clean-
up approach. During the formal comment period, EPA will
accept written comments via mail, email, and fax. Addition-
ally, verbal comments may be made during the formal Public
Hearing, where all offered comments will be recorded. EPA
will hold a brief informational meeting prior to the start of
the formal Public Hearing. EPA will not respond to comments
during the formal Public Hearing. See the cover page of this
document for information regarding informational meetings
and the Public Hearing.
EPA will review the transcript of formal comments received
during the Public Hearing, and written comments received
during the formal comment period, before making a final clean-
up decision. EPA will then prepare a written response to the
formal written and oral comments received. Any formal com-
ments will become part of the official public record. The tran-
script of comments and EPA's written responses will be issued
in a document called a Responsiveness Summary when EPA
releases the Final Modification to the Reissued RCRA Permit.
The Responsiveness Summary and Final Permit will be made
available to the public on-line, at the Information Repositories,
and at the EPA Records Center (see addresses below). EPA will
announce the final decision on the cleanup plan through the
local media and via EPA's website.
ACEC Area of Critical Environmental
Concern
ARARs Applicable or Relevant and
Appropriate Requirements
CERCLA Comprehensive Environmental
Response, Compensation, and
Liability Act
CMS Corrective Measures Study
CTDEEP Connecticut Dept. of Energy and
Environmental Protection
CWA Clean Water Act
EMNR Enhanced Monitored Natural
Recovery
IMPGs Interim Media Protection Goals
MassDEP Massachusetts Department of
Environmental Protection
MassDFG Massachusetts Department of
Fish & Game
NCP National Contingency Plan
MNR Monitored Natural Recovery
PCBs Polychlorinated Biphenyls
RCRA Resource Conservation and
Recovery Act
SEND US YOUR COMMENTS
Provide EPA with your written comments about the Proposed
Cleanup Plan.
Please email (r1housatonic@epa.gov), fax (617-918-0028) or
mail comments, postmarked no later than
Friday, August 8, 2014 to:
Dean Tagliaferro
EPA New England
c/o Weston Solutions
10 Lyman Street
Pittsfield, MA 01201
43
-------
U.S. EPA I GE-PITTS Fl E LD/H O U S ATO N I C RIVER SITE
LEARN MORE AT:www.epa.gov/region 1/ge
EPA's Draft Modification to the Reissued RCRA Permit as well as other supporting documentation are available at
www.epa.gov/ region 1 /ge/ proposedcleanupplan.html
HIGHLIGHTS OF PROPOSED CLEANUP PLAN
EPA is proposing a cleanup plan consisting of a combination of targeted soil and sediment removal, riverbed capping, and monitored
natural recovery to address risks posed by PCBs. This Proposed Cleanup Plan is expected to achieve the following outcomes:
Reduce risks to children and adults from direct contact with soil and sediment;
Reduce soil contamination in the floodplain to levels which allow continued recreational use without unacceptable risk;
Reduce PCB concentrations in fish to levels that allow increased consumption of fish caught from the River in
Massachusetts and Connecticut, and reduce impact to affected communities relying on the fish for economic
considerations or cultural practices;
Reduce the potential movement of PCBs from the river onto the floodplain, from the banks into the River, and
from upstream to downstream locations, including the downstream transport into Connecticut;
Reduce contamination and risk for ecological receptors (fish, wildlife, and other organisms) in the river, backwa-
ters, floodplain, and vernal pools;
Reduce PCB surface water and sediment concentrations by addressing PCB sources in sediment and soil to
advance future compliance with water quality standards in Massachusetts and Connecticut and attainment of the
highest possible use of the River consistent with the Clean Water Act;
Reduce the mass of PCBs in Housatonic River sediment and floodplain soil available for exposure and down-
stream transport;
Provide flexibility to address potential removal of dams in the river, changes in floodplain use over time, and
consideration of new technologies that may be useful in the future;
Transport of all contaminated soil and sediments off-site to existing licensed facilities approved to receive such
soil and sediment, thereby avoiding the need to construct new landfills in the watershed;
Protect and preserve the unique ecological characteristics of the Upper Housatonic Watershed in conduct-
ing remedial efforts; and,
Establish procedures to address PCB contamination associated with future work.
United States
Environmental Protection
M %Agency
June 2014
------- |