Superfund Proposed Plan
             Hudson  River  PCBs  Superfund  Site
                                             New York
 Region 2
                                                                                       December 2000
Community Role In the
Selection Process

USEPA relies on public input
to ensure that the concerns of
the   community   are
considered in selecting an
effective remedy for each
Superfund site. To this end,
the Reassessment  reports,
including the Feasibility Study
and the Proposed Plan, have
been made available to the
public for a public comment
period  which   begins  on
December  12,  2000  and
concludes  on  February 16,
2001.

A series of public meetings
will be held during the public
comment period  to  further
elaborate on the reasons for
recommending the preferred
remedy and initiate the receipt
                 PURPOSE OF THE PROPOSED PLAN

"Ttits  Proposed Plan describes the remedial alternatives considered for the
 I Hudson River PCBs Superfund Site and identifies the preferred remedial
alternative with the rationale for this preference.

The Proposed Ran was developed by the U.S. Environmental Protection Agency
(USEPA) in consultation with the New York State Department of Environmental
Conservation (NYSDEC). USEPA is issuing the Proposed Plan as part of its public
participation responsibilities under Section 117(a) of the Comprehensive Environ-
mental Response, Compensation, and Liability Act (CERCLA)  of  1980, as
amended, and Section 300.430(f}(2) of the National Oil and Hazardous Substances
Pollution Contingency Plan(NCP).Thealtematwessur^
in the Feasibility Study and other documents contained In theAdmtotstratiye Record
file for this site. USEPA encourages Hie public to review these documents to gain
a more comprehensive understanding of the site and Superfund activities that have
been conducted at the site.

This Proposed Plan is being provided as a supplement to the Agency's Hudson
River PCBs Reassessment reports, including the Feasibility Study, to inform the
public of USEPA's preferred remedy, and to solicit public comments pertaining to
the preferred alternative, as well as all the remedial alternatives evaluated.

USEPA's preferred remedy Consists of removal (targeted dredging) of 2.65 million
cubic yards of contaminated sediment containing over 100,000 pounds of PCBs
from the Upper Hudson River using environmental dredging techniques which
minimize adverse environmental impacts, including the resuspension of sediments
during dredging. The preferred remedy also Includes Monitored Natural Attenuation
(MNA) of the residual PCS contamination that remains in the dredged areas and the
unremediated areas until the concentration of PCBs  in fish tissue are at an
acceptable level,  Institutional controls such as fish consumption advisories and
fishing restrictions will remain in place (although perhaps in a modified form) until
these acceptable levels are achieved. Some of the dredged areas will be backfilled
with approximately one foot of clean material to isolate residual PCB contamination
and to replace  habitat for biota.  Dredged sediments will be dewatered and
stabized at  treatment/transfer facilities and then transported via rail to off-site
permitted disposal facilities outside of the Hudson River valley; A new landfill will
not be constructed in the Hudson River Valley to receive the dredged sediments.
The dredging will occur in conjunction with a separate Non Time Critical Removal
Action (NTCRA) to be implemented to control upstream PCB sources in the vicinity
of the General Electric Hudson Falls plant.

The remedy described in this Proposed Plan is the preferred remedy for the site.
Changes to the preferred remedy or a change from the preferred remedy to another
remedy may be made if public comments or additional data Indicate that such a
change will result in a more  appropriate remedial action.  The final decision
regarding the selected remedy will be made after  USEPA has taken  into
consideration all significant public comments. USEPA is soliciting public comment
on all of the alternatives considered in the detailed analysis of the Feasibility Study
because USEPA may select a remedy other than the preferred remedy.
EPA Region 2
                                                                   Pagel

-------
 Superfund Proposed Plan
                  Hudson River PCBs Reassessment
  The administrative record file, which
  contains the information upon which
  the selection of the response action
  wilt be based, is available at the
  following locations:

  Crandall Library
  City Park
  Glens Falls. NY 12801
  Phone 518-792-6508
  Hours:  Monday. Tuesday, Wednesday - 9 a.m. to 9
         p.m.
         Thursday & Friday- 9 a.m. to 6 p.m.
         Saturday-9a.m. to 5 p.m.
         Sunday -1 p.m. to 5 p.m.

  Adrlance Memorial Library
  93 Market Street
  Poughkeepsie, NY 12601
  Phone 845-485-3445

  Hours:  Monday tttfu Thursday-9 a.m. to 9p.m.
          Friday 4 Saturday-9 a.m. to 5 pjru
          Sunday-1p.m. to 5p.m.

  USEPA-Kegtan2
  Supedund Records Center
  290 Broadway, 18th Ftoor
  New York, NY 10007-1866
  (212)637-4308

  Hoots:  Monday-Friday, 9 a-m. - 5 p.m.
of public comments. The first meeting will be at the Civic
Center, Saratoga Springs, NY on December 12,2000 at
7:00 P.M.  A second meeting will be held on Thursday,
December 14, 2000 at 7:00 p.m. at the Sheraton Civic
Center Hotel in Poughkeepsie, NY.  Additional  public
meetings/availability sessions will  be  held during the
public comment period,  after the  public has had  an
opportunity to  review and  consider  the  preferred
alternative and supporting documents.  These additional
public meetings will be announced by USEPA  in the
media to allow for participation by interested individuals.

Comments received at the public meetings, as well as
written comments, will be documented and responded to
in the Responsiveness Summary appended to the Record
of Decision, the document that formalizes the selection of
the remedy.

Written  comments  on this  Proposed  Plan should  be
addressed to:
            Alison Hess/Doug Tomchuk
        Hudson River PCBs Public Comment
        U. S. Environmental Protection Agency
              290 Broadway, 19* Floor
             New York. NY 10007-1866

USEPA will  accept comments postmarked by Friday,
February 16, 2001.
 SITE BACKGROUND

 Site Description

 The Hudson River flows in a generally southerly direction
 approximately 315 miles from its source at Lake Tear-of-
 the-Clouds on Mount Marcy in the Adirondack Mountains
 to the Battery in New York City.  The Hudson River PCBs
 Superfund Site extends nearly 200 river miles (320 km)
from the Fenimore Bridge in Hudson Falls (River Mile
[RM] 197.3) to the Battery in New York City ([RM 0) at the
tip of Manhattan Island. The Superfund site traditionally
has been divided into the Upper Hudson River and the
 Lower Hudson River, based on physical and chemical
characteristics. The  Upper Hudson River extends from
the Fenimore Bridge in Hudson Falls to the Federal Dam
at Green Island in Troy (RM 153.9), a distance of about 43
 river miles. The Lower Hudson River extends from the
 Federal Dam to the Battery (RM 0) (see Figure 1).

The Upper Hudson River was further divided into three
sections in the Feasibility Study to evaluate remedial
alternatives (see Figure 2). River Section 1 consists of the
Thompson Island Pool, and extends about 6.3 miles from
the former Fort Edward Dam (RM 194.8) to the Thompson
 Island Dam at RM 188.5.

The first 2.5 miles from the Fenimore Bridge to the former
 Fort Edward Dam are not a major focus of the Proposed
 Plan because the sediment PCB contamination has largely
 been addressed. This area consists primarily of rocky
 outcrops and little sediment, or areas of sediment that have
 already been  remediated (i.e.,  the Remnant Deposits,
which are discussed in greater detail in Site History). The
 area between the former Fort Edward Dam and  the
 northern/upstream end of the Thompson Island Pool, a
 distance of 0.2 miles, contains shallow, fast moving water
 and primarily course-grained sediments that are believed
 to have minimal PCB inventory.

 River Section 2 extends from the Thompson Island Dam to
 the Northumberland Dam (sometimes referred to as Lock
 5) near Schuylerville (RM 183.4) an extent of about 5.1
 river miles.  River Section 3 extends from below the
 Northumberland Dam to  the Federal Dam at Troy,  an
 extent of about 29.5 river miles.

 The mean annual flow of the Hudson River at Fort Edward
 is approximately 3,800  cubic  feet per second  (cfs).
 Downstream of Fort Edward, the river is joined by several
 tributaries; the largest in the Upper Hudson River is the
 Hoosic River near Schaghticoke. The combined total of
 the tributaries nearly doubles  the flow of  the  Upper
 Hudson by the time it reaches Waterford, where the mean
 annual flow of the river is approximately 7,100 cfs.

 Within the Superfund site, there are eight dams with locks
 that form a series of pools in the Upper Hudson. The flow
 in the Upper Hudson is controlled by these dams and, to
EPA Region 2
                                           Page 2

-------
Superfund Proposed Plan
                  Hudson River PCBs Reassessment
                          \ Glens Falls
       NEW YORK
                Kingston
               Newburgh
                           Poughkpepsie
                                          cr
        NEW
        JERSEY
Figure 1

a lesser  degree,  by  wetlands  and backwaters in the
vicinity of the river that act as buffers for high and low flow
conditions.   The  flow  in the  Upper Hudson is also
controlled by several  reservoirs above Glens Falls, the
most significant of which is Great Sacandaga Lake. The
mean gradient of  the Upper  Hudson River is about 3
feet/mile.  The  gradient within each pool is smaller than
the mean gradient for the Upper Hudson River overall,
with elevation drops between the pools at the dams. The
width of the Upper Hudson above Lock 4 in Stillwater is
approximately  400 feet.  The  Upper Hudson has  an
average depth  of less than 8 feet in the shoal areas and
approximately  18  feet in the channel, with  a maximum
depth of more than 45 feet in a section below Thompson
Island Dam. The total  surface  area of the Upper Hudson
River is approximately 3,900 acres.

The Champlain Canal is coincident with portions of the
Upper Hudson  River, extending from Waterford (RM 158)
on the  Hudson to Whitehall at the southern  end of Lake
Champlain.  The  Champlain Canal is 60  miles long,
including  37 miles of canalized Upper Hudson River from
Waterford to  Fort Edward and 23  miles  of land-cut
sections.  The  canal diverges from the river at Fort
Edward just below Lock 7 and proceeds in a northeasterly
direction to Lake Champlain.  Additional land cut areas
exist at Stillwater, Northumberland, and  Fort Miller. The
Hudson  River between the Federal Dam at Troy and Lock
1 at Waterford is part of the Champlain Canal and the Erie
Canal.

Bedrock, cut away to  form  the Champlain Canal, is
exposed in some areas  and lacustrine silts and clays of
glacial age are exposed  in other areas. Coarser-grained
sediments have been observed in the channel and finer-
grained  sediments are more commonly  seen in shallow
slower-moving water. Wood chips are present in surface
sediments in many locations as well as sediment mounds
likely created by historic  disposal of dredged spoils in the
river.

Both  federal  and state  freshwater  wetlands  exist
throughout the  Upper  Hudson  region. The  100-year
floodplain ranges from approximately 400 to over 5,000
feet wide at places in the Upper Hudson.

Primary  uses of the Hudson River include:

        Public water supplies: The  cities of Waterford,
        Poughkeepsie,  and Rhinebeck, as  well as the
        Highland and  Port Ewen Water Districts, obtain
        their water supplies  directly from the Hudson
        River.  In addition, a water intake near Chelsea,
        which is north  of Newburgh, may  be used to
        supplement New York City's  water supply during
        periods of drought.   Waterford  is  the  only
        municipal  water supply intake in  the Upper
        Hudson River. The treatability study at Waterford
        Water  Works,  which was  completed in  1990
        pursuant to  the  1984  Record  of  Decision,
        indicated that the treated water met standards
        applicable to public water supplies.

        Industrial and commercial  purposes:  Hudson
        River water is extensively used for hydroelectric
        and  thermal power generation, as  well as for
        manufacturing   processes,   cooling,  and  fire
        protection. A limited commercial fishery remains
        in the Lower Hudson River.

        Domestic  and  agricultural  purposes:  Hudson
        River water is also used for domestic  purposes
        (watering lawns and  gardens)  and agricultural
        purposes (irrigating crops).

        Transportation:  The  Champlain Canal was a
        major   transportation  route  in  the  past.
        Commercial traffic has declined significantly on
        the  Canal  over the  last 30  years,  but may
        experience some growth in the future through
        revitalization programs such as the American
        Heritage River initiative.
EPA Region 2
                                           Page 3

-------
Superfund Proposed Plan
                                        Hudson River PCBs Reassessment
                                                         HUDSON FALLS
                                                          GE Plans

                                                       Jbnrcr Fort Edwgd Pan
                                                          FORT EDWARD
                                                         U)dc7(RM193.7)
                                                 Oiffa Island'
                                    GANSEVOORT          ft Thomson Island Dam (RM !8aS)
                                              Iharpsan libnd
        (RM 186.2)



9CHUYLERVILLE

        luck 5(RM 1826)!
                                        STILLWATER
                                                                                  SECTION 1
                                                                                  SECTION 2
                             MECHANICVILLE

                                 Lock3(RMl«6.0)
                                            SECTIONS
                                              TROY
                                            lFedoalDanimM153.9)
                      Scale in Mies
                        4
W
                    Scale in KSonvters
                                              FIGURE 2

                                        UPPER HUDSON RIVER
EPA Region 2
                                                                   Page 4

-------
Superfund Proposed Plan
                  Hudson River PCBs Reassessment
       Recreation: The Hudson River supports a variety
       of water-based recreational  activities, which
       include   sport  fishing,   waterfowl  hunting,
       swimming, and boating.

•      Ecological  Resources:  The  Hudson  River
       supports  206 species of fish as  well as 143
       species of resident and migrating birds. Sixty-four
       (64)   species  are  listed  as  Threatened,
       Endangered,  Rare or of Special  Concern by
       federal and New York State authorities. There
       are 39 areas of significant habitat identified in the
       Lower Hudson River.

In addition, the Hudson River has been designated as an
American Heritage River and has played a  major role in
the history, culture, and economy of the area.

Site History

General Electric  Company (GE)  operated a plant in
Hudson Falls and still operates a plant in Fort Edward,
New York.  During an approximate 30-year period that
ended in 1977, GE used polychtorinated biphenyls (PCBs)
in the manufacture of electrical capacitors at both of these
facilities.  Excess PCB oils were discharged both directly
and indirectly (particularly at the GE Hudson Falls plant)
into the Hudson River.  Estimates of the total quantity of
PCBs discharged from the two plants into the river from
the 1940s to 1977 range  from 209.000 to 1,330,000
pounds (95,000 to 603,000 kg).

Many of the PCBs discharged to the river adhered to
sediments and accumulated with the sediments as they
settled in the impounded pool behind the  Fort Edward
Dam.  as well  as  in  other impoundments  farther
downstream. Because of its deteriorating condition, the
Fort  Edward Dam was removed in  1973.    During
subsequent spring floods, PCB-contaminated sediments
from this area were scoured and transported downstream.
These  areas, which were surveyed by the NYSDEC in
1976-1978 and 1984, typically had average total PCB
concentrations of 50 parts per million (ppm) or greater and
are known as the NYSDEC-defined PCB hot spots. There
were 40 NYSDEC-defined hot spots, located between RM
194 at Rogers Island and Lock 2 at RM 163. Hot spots 1
through 4 have  been dredged by New York  State for
navigation purposes.   In addition, not all the PCB-
contaminated sediments behind the former Fort Edward
Dam were scoured and transported downstream. Five
areas of PCB-contaminated sediments were exposed due
to lowering of the river water level when the Fort Edward
Dam was removed. These five areas are known as the
Remnant Deposits.

Legal action  brought against GE by NYSDEC in 1975
resulted in a $7 million program for the investigation of
PCBs and the development of methods to reduce or
remove the threat of PCB contamination.  In 1975, the
New York State Department of Health (NYSDOH) began
to issue health advisories recommending that people limit
consumption of fish from the Upper Hudson River.  In
1976, NYSDEC issued a ban on fishing in the Upper
Hudson River from Hudson Falls to the Federal Dam at
Troy, due to the potential risks from consumption of PCB-
contaminated fish, and a ban on commercial fishing of
striped bass, which  migrate upriver  into the  Lower
Hudson. NYSDEC replaced the ban against fishing in the
Upper  Hudson River with a catch-and-release  fishing
program in 1995. NYSDOH continues to recommend that
people eat none of the fish in the Upper Hudson and that
children under the age of 15 and women of child-bearing
age eat none of the fish in the river for the entire length of
the Superfund site.  In addition, the commercial  striped
bass fishery in the Lower Hudson is still closed.

In 1974, the New York State Department of Transportation
(NYSDOT) dredged approximately 250,000 cubic yards
from the  channels adjacent  to   Rogers Island for
navigational  purposes.  The dredged materials were
disposed of in Special Area 13, which is located along the
west bank  of  the  river just south of Rogers  Island.
Another 384,000 cubic yards of sediment were dredged
from the east and west channels in  1974 and 1975 and
disposed  of in the Old Moreau Landfill, located  on the
west shore of  the river opposite the southern  end of
Rogers Island (and just north of Special Area 13).

In 1977, the manufacture and sale  of PCBs within the
United  States were generally prohibited under provisions
of the Toxic Substances Control Act (TSCA). Although
commercial uses of PCBs ceased in 1977, PCBs from
GE's Fort Edward and Hudson Falls plants continued to
contaminate the Hudson River, due primarily to erosion of
the contaminated Remnant Deposits, discharges of PCBs
via bedrock fractures from the GE Hudson Falls plant, and
erosion from contaminated deposits above the water line
near the GE Fort Edward plant outfall.

About  14,000  cubic yards  of  highly  contaminated
sediments were removed by NYSDEC from  Remnant
Deposit 3A  in  1978 and  were placed in a  secure
encapsulation site in Moreau, along with some 215,000
cubic yards of sediment that had been dredged  by
NYSDOT from the east channel of Rogers Island  to clear
the navigation  channel just  below  the location of the
former Fort Edward Dam. Unstable river banks of two of
the Remnant Deposits were reinforced at that time. Three
remnant sites were revegetated to prevent public  contact
with the sediments and to minimize erosion and release of
PCBs into the environment.

No dredging in the Upper Hudson  River has occurred
since 1979, except for removal of coarse, uncontaminated
sediments that periodically accumulate at the mouth of the
Hoosic River, a tributary  that empties into the Hudson
River at RM 167.5 near Schaghticoke.

The site was proposed to the National Priorities List in
1983 and formally listed in 1984. The two GE capacitor
manufacturing plants in Hudson Falls and Fort Edward are
listed under the New York  State Inactive Hazardous
EPA Region 2
                                          Pages

-------
 Superfund Proposed Plan
                  Hudson River PCBs Reassessment
Waste Sites program.

In 1984, USEPA completed a Feasibility Study and issued
a Record of Decision for the site. The Record of Decision
did not address the PCBs-contaminated oil that is leaking
through bedrock in the vicinity of the GE Hudson Falls
plant, which  was  unknown  at  the time.   USEPA
recognized that PCB contamination in the Upper Hudson
River sediments was a problem, but selected an interim
No  Action  remedy  for  the contaminated  sediments
because, in  the  Agency's view, the  reliability  and
effectiveness of remedial technologies available at that
time were uncertain. The Record of Decision contained
the following decisions:

       An  interim  No Action decision with regard to
       PCBs in  the sediments  of the Upper Hudson
       River;

       In-place capping, containment and monitoring of
       exposed Remnant Deposits (in the area of RM
       195 to 196), stabilization of the associated river
       banks and revegetation of the areas; and

       A detailed evaluation  of the Waterfbrd Water
       Works treatment facilities, including sampling and
       analysis of treatment  operations to see  if an
       upgrade  or alterations of the  facilities  were
       needed.

GE,  under a 1990 Consent Decree  with  USEPA,
conducted the in-place capping of the Remnant Deposits
on the river bank (in the area of RM 195 to RM 196) from
the  former impoundment behind the Fort Edward Dam.
The in-place capping of these Remnant Deposits included
placement of a two-foot layer of soil and a manufactured
geosynthetic  clay   liner, followed  by  grading  and
revegetating to minimize erosion. The river banks were
stabilized with rock to prevent scouring. Cap construction
and  the erection  of gates  to limit site access  were
completed in 1991.

NYSDEC, with funding provided by USEPA, conducted a
treatability study  at the Waterfbrd Water Works.   The
study was released in   1990 and  found that  PCB
concentrations were below analytical detection limits after
treatment and met standards applicable to public water
supplies.

In December 1989, USEPA announced a reassessment
of the interim No Action decision for the  Upper Hudson
River sediments as part of the five-year review required by
CERCLA,   and  in  consideration   of  advances  in
methodologies for sediment dredging technologies for
PCB treatment/destruction,  as well as a request for a
reevaluation  from  NYSDEC.    The  Reassessment
Remedial Investigation/Feasibility Study has been divided
into three phases.   Phase  1,  consisting primarily of a
review of existing data, was completed in August  1991.
Phase 2, which included the collection and analysis  of
new data as well as modeling studies and human health
and ecological risk assessments, began in  December
1991 and concluded in November 2000.  Phase 3, also
known as the Feasibility Study, began in September 1998
with release of the Feasibility Study Scope of Work. The
Feasibility Study is being released concurrently with this
Proposed Plan.

As USEPA was beginning Phase 2 of the Reassessment
in September 1991,  GE detected an increase in PCB
concentrations at the Upper Hudson River water sampling
stations being monitored as part of the Remnant Deposits
capping.  GE ultimately attributed the higher levels to the
collapse of a wooden gate structure within the abandoned
Allen Mill located adjacent to the river bank at the GE
Hudson Falls plant site.  As reported by GE,  the gate
structure had diverted water from flowing through a tunnel
cut into bedrock, thereby preventing oil-phase PCBs that
had  migrated to the tunnel via  subsurface  bedrock
fractures from flowing into the river.  From 1993 to 1995,
GE removed approximately 45 tons of PCB-bearing oils
and  sediments  from  the  tunnel  under   NYSDEC
jurisdiction.

In   1994,  GE documented the  presence  of PCB-
contaminated oils in bedrock  seeps at Bakers  Falls
adjacent to its Hudson Falls  plant.  GE has instituted a
number of mitigation efforts that have  resulted in  a
decline, but not total cessation, of PCBs entering the river
through the seeps.

The GE Fort Edward plant 004 Outfall has also been a
source of PCBs to the river.  In January 2000, NYSDEC
signed a  Record of Decision that called for removal of
PCB-contaminated soils and sediments near the 004
Outfall. NYSDEC is currently undertaking the Remedial
Design of that remedy.

In 1998, USEPA conducted an evaluation of whether an
early action would be warranted prior to completion of the
Reassessment This evaluation was prompted by findings
of the Low Resolution Sediment Coring Report, in which
USEPA determined that there were statistically significant
losses of PCBs from the sediment to the water column.
USEPA decided in December 1998 that no feasible and
appropriate interim action was available, and the USEPA
would complete the Reassessment as planned.

Extensive public involvement occurred during all phases of
the  Reassessment.    In  the  early  stages  of  the
Reassessment, USEPA initiated the Community Interaction
Program  to  involve interested citizens,  government
officials, environmental groups, and private interest groups
in  a unique effort  to include  their interests in  the
Reassessment.   The  foundation  of  the  Community
Interaction Program consists of four Joint Liaison Groups:
Agricultural,  Citizen, Environmental and Governmental.
The Community Interaction Program also established the
Steering Committee and the Oversight Committee, as well
as  the Scientific and Technical Committee, which is a
group of  researchers and scientists familiar with the site,
EPA Region 2
                                           Page6

-------
Superfund Proposed Plan
                  Hudson River PCBs Reassessment
PCBs, modeling, toxicity, and other relevant disciplines.
All Reassessment reports have been reviewed by these
groups  as well  as any  individuals  interested  in the
Reassessment.

USEPA has established and maintained 16 Information
Repositories located in public buildings from Fort Edward
to New York  City and  has placed  copies of the
Reassessment reports into these repositories. USEPA has
held more than 65 public meetings during the course of the
Reassessment. USEPA has responded to public comment
on the  Reassessment reports and  has  placed these
Responsiveness  Summaries   in   the   Information
Repositories as well.

Peer Review

In accordance with USEPA guidance and the Peer Review
Handbook,  the  scientific  work  conducted  for  the
Reassessment that is the  basis for this proposed action
has undergone external peer review.  USEPA's major
Phase 2 Reports have undergone external peer review by
five panels of independent experts. These reports were
the October 1996 Preliminary Model Calibration Report,
the geochemistry  reports  (the February  1997  Data
Evaluation and Interpretation Report  and the July 1998
 What am PCBs?

 The contaminant of concern at the Hudson River PCBs site
 is polychtorinated biphenyfs (PCBs).

 PCBs were widely used as a fire preventatjve and insulator
 in the manufacture of transformers and capacitors because
 of their abffity to withstand exceptionally high temperatures.

 PCBs are considered probable human carcinogens and are
 linked to other adverse health effects such as developmental
 effects, reduced birth weights and reduced ability to fight
 infection.

 PCBs are a group of chemicals consisting of 209 individual
 compounds, known as congeners. The congeners can have
 from one to fen chlorine atoms per molecule, each with its
 own set of chemical properties.  When grouped by the
 numberof chlorine atoms per molecute. the term homologue
 is used. PC8s were sold in mixtures containing dozens of
 congeners. These commercial mixtures were known tn flie
 U.S. as Aroclors.

 When released into the environment various processes can
 alter the pattern of PCBs from the original Aroclors.
 Analytical techniques vary and have improved over time.
 Congener-specific analyses  were  conducted for the
 Reassessment,  but most of  the  older data  was  an
 interpretation of Aroclors.  Therefore, a translation method
 was developed for the Reassessment to allow use of historic
 and recent datasets on a common basis.  The parameter
 common to ail data sets is known as 7W+  PCBs, and
 represents the sum of PCBs with 3 to 10 chlorine atoms per
 molecule.
Low Resolution Sediment Coring Report), the August
1999 Human Health Risk Assessment, the August 1999
Ecological Risk Assessment, and  the  January 2000
Revised Baseline Modeling Report.   Each peer review
panel was asked to address specific questions, together
called the "charge," regarding the methods USEPA used,
the findings and conclusions of the report being reviewed,
and controversial issues that were identified by the public
prior to and during the peer review meeting.  In addition,
the panels were invited to address any other issues that
were not specifically identified in the charge.

The peer reviewers generally agreed with the findings and
conclusions of the reports, although they  also requested
revisions, including extensive revisions to the Ecological
Risk Assessment.  USEPA issued Responses to  Peer
Review Comments for each of the peer reviews as well as
a Revised Human Health Risk Assessment and a Revised
Ecological Risk Assessment,  which  include all changes
made to address the peer review comments on those
reports. Revisions were incorporated, as appropriate, into
the Feasibility Study.

In addition,  the   Scientific  and Technical Committee
described previously, has provided  peer input into the
various documents USEPA  prepared as part of the
Reassessment.

RESULTS OF REMEDIAL INVESTIGATION

Summary of Sampling Results

For its Reassessment Remedial Investigationl/Feasibility
Study, USEPA used data collected during  its own sampling
investigations, as well as data collected by many other
agencies, institutions, and GE. The investigations include
sediment  surveys,  river   flow  and   water  quality
investigations, fish and biota sampling, air monitoring, and
plant and crop uptake studies. USEPA's data collection for
the Reassessment focused on the Upper Hudson River
because that portion of the site is under consideration for
possible remedial action. It was also the focus for the 1984
Record of Decision. PCB concentrations remain elevated
in the Hudson River in the sediments, in the water column,
and in the fish. Concentrations associated with the site
generally decrease with distance down river.

Sediment

Areas of elevated concentrations of PCBs in sediment,
i.e., hot spots, are found in depositional areas throughout
the Upper Hudson (River Sections 1,2, and 3).

River Section 1 (Thompson Island Pool) contains 20 of the
40 hot spots identified  by NYSDEC in 1977 and 1984.
The sediments exhibit a high degree of heterogeneity with
respect to the distribution of PCBs.  Historically, the
highest concentrations of PCBs in sediments have been
observed within the cohesive sediments of River Section
1, and  generally lower PCB concentrations are found
 EPA Region 2
                                            Page 7

-------
 Stipe/fund Proposed Plan
                                                                          Hudson River PCBs Reassessment
within the  non-cohesive  sediments.  The maximum
concentration measured was approximately 2000 mg/kg
PCBs (or  parts per  million  (ppm)).   The  average
concentration in sediments (0 - 25 cm deep) in 1991 was
approximately 42 mg/kg. It is estimated that there are
approximately 34,000 Ibs (15,400 kg) of PCB mass in the
sediments in River Section 1.

River  Section  2  (Thompson   Island  Dam  to
Northumberland Dam near Lock 5) contains 15 of the 40
NYSDEC-defined hot spots.  The average concentration
of PCBs in  surface sediment (0 - 25 cm) in 1991 was
approximately 26 mg/kg. The maximum concentration of
PCBs in the Hudson, 4000 mg/kg, was found in this river
section, in Hot Spot 28 in a thin slice of a sediment core.

River Section 3 (Northumberland Dam to Federal Dam at
Troy) contains 5 of the 40 NYSDEC-defined hot spots.
The average concentration in surface sediment (0 - 25
cm) in 1991  was approximately 9 mg/kg PCBs.

An assessment of concentrations of PCBs in sediments
below Federal Dam is limited by the availabile data for this
region (approximately 153 miles of the Lower Hudson
River). An assessment of the Lower Hudson performed in
the  1980s indicated that the average concentration of
PCBs in sediment in New York Harbor was 0.8 mg/kg in
1970 and 0.5 to 0.7 mg/kg in the 1980s. USEPA estimates
that approximately 50% of this contamination is attributable
to the releases from the GE Hudson Falls and Fort Edward
plants to the Upper Hudson.

Water Column

The dominant sources of PCBs to the water column of the
Upper Hudson River may be separated into two groups:
(1) PCB-contaminated sediments on the river bottom; and
(2) PCB-contaminated oil from bedrock seeps from the
GE Hudson Falls plant There are other lesser discharges
upstream of Rogers Island.

U.S. Geological Survey monitoring of PCBs in the water
of the Upper Hudson River began in 1977.  GE began
monitoring of the Upper Hudson River in 1991.  In River
Section  1,  PCB concentrations in  the water  column
indicate that the sediments of the Thompson Island Pool
are the major source of PCBs to the water column during
low flow conditions,  which are important as they coincide
with the period of greatest biological activity.

During the  summer of 1998  (June-September), the
average concentration at the Thompson Island Dam-West
station was 134 nanograms per liter (ng/L or parts per
trillion). Concentrations from January 1996 through March
2000 averaged 90 ng/L. Five observations in excess of
300 ng/L were noted during the winter of 1999-2000.
 Fish
PCB concentrations observed in fish are a result of
exposure to PCBs in water and surface sediment, through
either an aquatic food chain or  a benthic food  chain,
respectively.

NYSDEC continues to collect and analyze fish tissue data
from many  locations  in the Upper  Hudson  River.
Converted to a Tri+ PCB basis, the concentrations in River
Section 1 (Thompson Island Pool) in 1999 averaged about
21 mg/kg (wet weight) in largemouth bass and 13 mg/kg in
brown bullhead.  The maximum PCB concentrations
measured were 114 mg/kg (wet weight) in largemouth
bass and 31 mg/kg in brown bullhead.

Concentrations in River Section  3 (Stillwater) in 1999
averaged about 6 mg/kg (wet weight) in largemouth bass
and 6 mg/kg in brown bullhead.  The maximum PCB
concentrations measured were 23 mg/kg (wet weight) in
largemouth bass and 15 mg/kg in brown bullhead.

For comparison  purposes, USEPA has determined that
0.05 mg/kg (wet weight in fish fillets) is an acceptable PCB
concentration for Hudson River fish based on an annual
consumption of 51 half-pound meals per year by an adult.

Because PCBs tend to accumulate in fatty tissues, it is also
useful to examine PCB concentrations in fish on a lipid (fat)
basis for analysis of trends. The lipid-based Tri+ PCB
concentrations for 1998 are generally similar to those
observed from 1995 to 1997 in both River Section 1 and
River Section 3, with little evidence of a consistent decline.

Time trends of lipid-based Tri+ PCB concentrations for
brown bullhead, largemouth  bass, and pumpkinseed in
River Section 3 for the Stillwater reach  (RM 168.1) show
that PCB concentrations in the fish appear to have been
nearly stable in recent years.

USEPA's analysis of all the data indicate that the  spatial
variability of PCB concentrations in fish is  determined
primarily by distance downstream of the Thompson Island
Pool.

Geochemistry and Modeling Conclusions

The Reassessment Remedial Investigation/Feasibility
Study has evaluated PCB contamination at the site using
a  number of tools.   These tools include geochemical
analyses of the water and sediment,  analyses of the
biological monitoring data, and synthesis of the data in a
complex mathematical (computer) model. The model was
calibrated to a 21-year historical data set.  The model is
particularly useful in understanding general trends in PCB
fate, transport, and bioaccumulation, and is the primary
tool available to forecast future concentrations of PCBs in
sediment, water, and fish. However, the other tools are
sometimes more  appropriate for understanding  the
system on a finer scale, as well as for providing insight
into the model projections. The following summarize the
 EPA Region 2
                                           PageS

-------
Superfund Proposed Plan
                  Hudson River PCBs Reassessment
key  conclusions  of  the  Reassessment  Remedial
Investigation.

•      The  PCBs were  released from  the two GE
       capacitor manufacturing plants in Hudson Falls
       and Fort Edward into the Hudson River. Once in
       the river, the PCBs adhered to sediments or were
       carried in the water column.

       PCBs  in  the  fine-grained sediments are  a
       continuing^source of contamination to the water
       column and biota, through aquatic and benthic
       food chains and through  processes  that have
       been empirically measured but are not easily
       modeled.

•      Because the river is a dynamic system, the PCB-
       contaminated sediments are not stable.  Some
       PCB-contaminated sediment may  be buried by
       deposition of cleaner sediments at some times,
       but in other places and at other times they may be
       redistributed by scouring.  There is little evidence
       of widespread  burial  of PCB-contaminated
       sediment by cleaner sediment in the Thompson
       Island  Pool sufficient to mitigate exposure to
       biota.

       As  of  1994, there  has been  a  statistically
       significant loss of PCB  inventory from  highly
       contaminated sediments in the Thompson Island
       Pool and a net loss of inventory from hot spot
       sediments between the Thompson Island Dam
       and the Federal Dam at Troy.

•      High flow events  (e.g.,  spring  floods) may
       increase  the bioavailabiltty of  contaminants to
       organisms in the water column. Water column
       sampling from a high flow event in January 1998
       showed elevated PCB concentrations.

       PCBs  in sediments  will not   be  naturally
       'remediated* via dechtorination.  The extent of
       dechlorination is limited, resulting in probably less
       than ten percent mass loss of PCBs.

•      The area of the site upstream of the Thompson
       Island  Dam  represents the primary  source of
       PCBs to fish within the freshwater Hudson. This
       includes the GE Hudson Falls and Fort Edward
       plants, the Remnant Deposits, and the sediments
       of the Thompson Island Pool.

•      The  modeling  showed   that alleviating  the
       upstream source is  important  to the long-term
       recovery of the river.  The upstream source is
       expected to be the dominant source of PCBs in
       fish within several decades. Source control alone
       will not, however, reduce PCB concentrations to
       acceptable levels in a reasonable time frame, nor
       reduce the downstream  transport of PCBs to
       acceptable levels  unless source control  is
       implemented   along  with  remediation  of
       contaminated sediments. In order to address the
       upstream source, USEPA has authorized an
       Engineering  Evaluation  and   Cost  Analysis
       (EE/CA) for a Non-Time Critical Removal Action
       (NTCRA) to address the on-going PCB source(s)
       in the bedrock that are still being released in the
       river near the GE Hudson Falls plant site. GE has
       discussed  an  approach  with  NYSDEC  and
       USEPA to  cut  off these PCB  releases by
       excavating a tunnel between the plant and the
       river and installing an oil collection system within
       it.  Assuming that such a tunnel, or equivalent
       containment system, is a viable response action
       to address the GE Hudson Falls plant site source,
       USEPA believes,  based upon these discussions,
       that a source control system can reasonably be
       expected to be in place and operating by January
       1.2005.

•      PCBs are transported from the Upper Hudson
       River to the Lower Hudson River (/.e., south ofthe
       Federal Dam at Troy).   The mass of PCBs
       transported over  the Federal Dam to the Lower
       Hudson declined from about  3,000 to 4,000
       kg/year (6,600 to 8,800 Ibs/year) in the late 1970s
       to  about 150  to 500 kg/year  (330 to 1,100
       Ibs/year) by the late 1980s to early 1990s. Based
       on 1998 data reported by GE, from a monitoring
       station at Schuylerville, 214 kg/year (471 Ibs/year)
       of PCBs are being transported over the Federal
       Dam at Troy..

•      PCB concentrations in fish, the primary pathway
       of concern, are still well above acceptable risk-
       based and advisory levels.

In sum, the  PCB-contaminated  sediments  of the
Thompson Island Pool strongly impact the water column,
generating a  significant  water column  PCB load and
exposure concentration whose congener pattern can be
seen   throughout  the  Upper   Hudson.  Burial of
contaminated  sediment  by  cleaner material is not
occurring universally,  and the stability of the sediment
deposits is not assured.

SUMMARY OF SITE RISKS

Based upon the results of the Reassessment, a baseline
risk assessment was conducted for the site to estimate the
risks associated with current and future site conditions. A
baseline risk assessment is an analysis of the potential
adverse human health and ecological effects caused by
hazardous substance releases from a site in the absence
of any  actions to control or mitigate exposure to these
hazardous substances.
EPA Region 2
                                           Page9

-------
  Superfund Proposed Plan
                                        Hudson R/ver PCBs Reassessment
  Human Health Risk Assessment

 A site-specific  Human Health Risk Assessment for the
 Hudson River PCBs site was developed to quantitatively
 evaluate both cancer risks and non-cancer health hazards
 from exposure  to  PCBs.    The  Human  Health  Risk
 Assessment considers exposure to PCBs in the  Upper
 and Mid-Hudson River beginning in 1999. assuming no
 remediation and no institutional controls, such as the fish
    WHAT IS RISK AMD HOW 13 IT CALCULATED?

    A Superfund human health risk assessment is an analysis of
    the potential adverse  health effects caused by hazardous
    substances releases ftom a s*e m the absent* of any actions
    to control or mftfcate these releases; it estimates the "basefine
    risk" In the absence of any remedial actions at H» site under
    current and future tend uses.  To estimate this baseline risk at
    a Superfund site, a four-step process is utfflred for assessing
    site-retaJed human health risks for reasonable maximum
    exposure scenarios.
    Huautf/cfenttfcattowThe hazard JdentSfc
BtepWenfifie
    the contaminants of concern at *» site fci vtrious media
    sol. 0rotmd*wer, surface water, air,ete.}based on such factors
    as toxfcfcy, frequency of occurrence, trie «*t transport of the
    contaminants in (he  environment,  concentration* of  the
    contaminants to specific media, mobility, persistence,  and
    btoaccumuJation.

    Biposie expected to occur, is calculated.
                              asseasmentdetermlnes Oie
   type* of adverse health effects associated wtft chemical
   exposures, and the relationship between the magraude  of
   exposure (dose) and severity of adverse effects (response).
   Poter^rwaJth effects are chen*^*peeifc^
   health dEfecte such as charges in the normal functions of
                         changes in trweffe^
   immunesystem), Some cherrtcate are capable of causing both
   cancer and non-cancer heath effects.

   ftftftOMacaartMlteiyThfe steps
   outputs of the exposure and toxicfty assessments to provide a
   quantitative assessmentofsto risks.  Exposures areevahiated
   based on 9m potential risk for developing cancer and  the
   potential for nornancer health hazards. The IkeShood of an
   individual developing cancer Is expressed as a probability.  For
   example, a 1O* cancer ask means a 'one-in-ten-thousand
   excess cancer risk'; or one addWonal cancer may be seen in a
   population of 10,000 people as a result of exposure to site
   contaminants under the conditions explained in the Exposure
   Assessment.   Current  federal Superfund  guidelines  for
   acceptable exposures are an individual lifetime excess cancer
   risk in  the range of 10^  to 10*  (corresponding to a
   one-in-terHhousand to a one4n>a-miKon excess cancer risk).
   For non-cancer health  effects, a  "hazard  index*  (HI) is
   calculated. An Hi represents the sum of the individual exposure
   levels compared to their corresponding reference doses (RfDs).
   The key concept for a non-cancer hazard index is that a
   "threshold level" (measured as an HI of 1) exists betow which
   non-cancer health effects are not expected to occur.
 consumption advisories and fishing restrictions currently in
 place.

 This section summarizes the results of the Human Health
 Risk Assessment and is based on the November 2000
 Revised Human Health Risk Assessment. The November
 2000 report combines into a single report the August 1999
 report for the Upper Hudson River, the December 1999
 report  for   the  Mid-Hudson River,  their  respective
 Responsiveness Summaries  issued in March 2000 and
 August 2000, and changes made  to address  the  peer
 review comments, which are documented in the November
 2000 Response to Peer Review Comments on the Human
 Health Risk Assessment for the Upper Hudson River.

 Hazard Identification

 PCBs, including dioxin-like PCBs, were identified as the
 contaminants of concern based on previous investigations
 and the  site definition.  The media  evaluated were fish,
 sediment, surface  water, and air.   Current  and future
 concentrations of PCBs in fish, sediments, river water and
 air are derived from USEPA's Reassessment database and
 Revised  Baseline Modeling Report (RBMR).

 Exposure Assessment

 The cancer risks and non-cancer health  hazards were
 evaluated for young children (aged 1  to 6 years old),
 adolescents (aged 7 to 18 years old); and adults (over 18).
 These individuals include anglers who  eat fish  from the
 Hudson River, recreators who swim, wade, or boat in the
 Hudson,  and residents along the Hudson who may inhale
 volatilized PCBs or use the river as their source of drinking
water.

 Consistent with USEPA policy and guidance, cancer risks
and non-cancer health hazards were evaluated for  the
 reasonably maximally exposed (RME) individual and the
central tendency  individual. The RME is considered the
 maximum exposure that is reasonably estimated to occur
at the site and is not a worst-case scenario.  The central
tendency exposure  is the  average  exposure  to  an
individual.

Fish Ingestion

 For the fish ingestion pathway, USEPA developed a site-
specific  exposure  duration  for the angler  based  on
population mobility for the counties surrounding the Upper
and Mid-Hudson River from the 1990 Census and fishing
durations reported in the state-wide 1991 New York Angler
survey. The concentrations of PCBs in fish were calculated
from the  forecasts in the RBMR by considering the fish
species preferentially eaten, accounting  for the change in
concentrations  of  PCBs  in  fish  with   river  mile,  and
averaging over the total exposure duration.

Fish ingestion rates were based on ingestion  rates for
types of fish found in the Hudson River, as reported in the
 1991  New York Angler survey. The rate derived for the
EPA Region 2
                                                                  Page 10

-------
Super-fund Proposed Plan
                   Hudson River PCBs Reassessment
RME adult is about one half-pound fish meal per week.
USEPA believes that this ingestion rate addresses fish
consumption of subsistence anglers in the Upper and Mid-
Hudson River, based on a  1986 study of low-income
families' consumption of freshwater fish in New York
State.  In addition,  this ingestion rate is essentially the
same as the rate used by the New York State Department
of Health in establishing the fish consumption advisories
for the Hudson River. USEPA's fish ingestion rate for the
average, or central tendancy adult angler, is about one
half-pound fish meal every two months.

USEPA also quantitatively evaluated cancer risks and
non-cancer health hazards to recreators who swim, wade,
or boat in the Hudson and residents along the Hudson
who may inhale volatilized PCBs or use the river as their
source of drinking water.  The cancer risks and non-
cancer health hazards from these exposure pathways
were determined to be within or below acceptable levels
under the federal Superfund program.  Other pathways
were evaluated qualitatively  (e.g., exposure to PCBs in
home-grown crops or dairy products) and determined to
be below levels of concern.

ToxicltyofPCBs

USEPA has determined that  PCBs cause cancer in
animals and probably cause cancer in humans. Serious
non-cancer health effects have been observed in animals
exposed to PCBs.  Studies of Rhesus monkeys exposed
to PCBs indicate a reduced  ability to fight infection and
reduced birth weight in offspring exposed in utero. In the
Human Health Risk Assessment, USEPA used the current
Agency consensus toxicity values for PCBs in determining
cancer and non-cancer health effects.
Risk Characterization

The Human Health Risk Assessment shows that cancer
risks and non-cancer health hazards to the RME individual
associated with ingestion of PCBs in fish from the Upper
and Mid-Hudson River are above levels of concern.  Fish
ingestion is the primary pathway for PCB exposure and for
potential adverse health effects.  Cancer risks and non-
cancer health hazards from other exposure pathways in the
Upper and Mid-Hudson River are generally within or below
levels of concern. The table below shows the cancer risks
and non-cancer health hazards for consumption of fish in
the Upper and Mid-Hudson  River beginning  in  1999
(rounded to the nearest whole number) in the absence of
remediation.   Cancer risks from exposure  to dioxin-like
PCBs were comparable to the cancer risks presented in the
table below.

In addition to these point estimate calculations, USEPA
calculated the cancer risks and non-cancer health hazards
for ingestion of fish in the Upper Hudson River using a
probabilistic risk assessment analysis called a Monte Carlo
analysis. The results of this approach support the results
of the point estimate calculations.

The Human Health Risk Assessment shows that, under the
baseline conditions, the cancer risks and the non-cancer
health hazards from ingestion of fish  from  the Upper
Hudson River are expected  to  be above  USEPA's
generally acceptable levels for the 40 year exposure
duration beginning  in 1999. The total cancer risk for the
RME individual is 1,000  times higher than the goal for
protection and 10 times higher than the highest risk level
generally allowed under the federal Superfund law.  Non-
cancer  health  hazards  for  the  RME  young  child,
adolescent, and adult are 104,  71, and 65 times higher
Summary of Cancer Rj$ks and Non-Cancer Health Hazards from Ingestton of Fish
Upper Hudson River Central Tendency RME
Fish Ingestion - cancer
Total (young child,
adolescent, and adult exposure)
Fish Ingestion - non-cancer
Adult
Adolescent
Young Child
Mid-Hudson River
Fish Ingestion - cancer
Total (young child,
adolescent, and adult exposure)
Fish Ingestion - non-cancer
Adult
Adolescent
Young Child
axioms in 100,000)
7
8
12
Central Tendency
'(x\Q*(\ in 100,000)
3
4
5
1 x10'3(1 in 1,000)
65
71
104
RME
7x10^(71010,000)
32
35
49
EPA Region 2
                                            Page 11

-------
 Superfund Proposed Plan
                   Hudson River PCBs Reassessment
 than the level considered to be protective of public health.
 Ingestion  of just one half-pound fish  meal every two
 months, the average ingestion rate, results in cancer risks
 to the central tendency individual that are above the goal
 for protection and results in non-cancer health hazards
 that are 7 to 12 times higher than the level considered to
 be protective.  The cancer risks and non-cancer health
 hazards from ingestion offish from the Mid-Hudson River
 are about half as high as those in the Upper Hudson, due
 to lower concentrations of PCBs in fish, but are also above
 levels of concern.

 Ecological Risk Assessment

 This section summarizes the results of the Ecological Risk
 Assessment and is based on the November 2000 Revised
 Ecological Risk Assessment.  The November 2000 report
 combines into a single report the August 1999 Ecological
 Risk Assessment, the December 1999 Ecological Risk
 Assessment for Future Risks in the Lower Hudson River,
 their respective Responsiveness Summaries issued in
 March 2000 and August 2000, and changes made to
 address  the  peer  review comments,  which  are
 documented in the November 2000  Response to Peer
 Review Comments on the Ecological  Risk Assessment

 The process used for assessing site-related ecological
 risks  for a  reasonable maximum  exposure scenario
 includes: Problem Formulation—a qualitative evaluation
 of contaminant release,  migration, and fate; identification
 of contaminants of concern (COCs), receptors, exposure
 pathways, and known ecological effects of the contami-
 nants;  and  selection of endpoints  for further  study;
 Exposure Assessment—a quantitative evaluation of con-
taminant release, migration, and fate; characterization of
exposure pathways and receptors; and  measurement or
 estimation of exposure point concentrations; Ecological
 Effecfs Assessment—literature reviews, field studies, and
toxicity tests, linking contaminant concentrations to effects
on  ecological  receptors;  and   Risk  Character-
ization—measurement or estimation of  both current and
future adverse effects.

 PCBs, including dioxin-like PCBs, are the contaminants of
 concern for the Ecological Risk Assessment based on the
 results of  earlier investigations and  the site  definition.
 Ecological exposure  to PCBs is primarily  an issue of
 bioaccumulation through the food chain rather than direct
 toxicity.  PCBs bioaccumulate in the  environment by
 bbconcentrating  (being  absorbed  from  water  and
 accumulated in tissue to levels greater than those found
 in surrounding water) and biomagnifying (increasing in
 tissue concentrations as they go up the food chain through
 two or more trophic levels).

 The Hudson River PCBs site is home to a wide variety of
 ecosystems.   The Upper Hudson  River is a nontidal
 freshwater system, while the Lower Hudson River (i.e.,
 south  of the  Federal  Dam) is tidal  with freshwater,
 brackish, and increasingly more saline water towards the
 Battery.
The Ecological Risk Assessment evaluates direct exposure
to PCBs in Hudson River sediments and river  water
through ingestion and indirect exposure to PCBs via the
food chain. Because PCBs are known to biomagnify, an
emphasis was placed on indirect exposure at various levels
of the food chain to  address PCB-related risks at higher
trophic levels.  The assessment endpoints  that were
selected  for the Hudson  River are benthic community
structure, which is a food source for local fish and wildlife,
and sustainability (survival, growth, and  reproduction) of
local fish populations,  insectivorous  bird  populations,
waterfowl  populations,  piscivorous  (fish-eating)  bird
populations,   insectivorous   mammal  populations,
omnivorous  mammal  populations,  and  piscivorous
mammal populations. The bald eagle, a federally-listed
and  New  York State-listed  threatened species, was
evaluated under the assessment endpoint for piscivorous
birds.

Risks to the environment were evaluated for individual
receptors   of  concern  that  were  selected   to   be
representative of various feeding preferences, predatory
levels,   and  habitats  (aquatic,   wetland,  shoreline).
Receptors  of concern selected for the  Ecological Risk
Assessment  included the benthic macroinvertebrate
community, seven species of fish as represented by the
pumpkinseed (Lepomisgibbosus), spottail shiner (Notropis
hudsonius), brown  bullhead  (Ictalums,  now Ameiurus
nebulosus), white perch (Morone americana), yellow perch
(Perca  flavescens),  largemouth  bass  (Micropterus
salmoides), and striped bass (Morone saxatilis). Five bird
receptors were  selected: the tree swallow (Tachycineta
bicolor), mallard (Anas platyrhychos), belted kingfisher
(Ceryle alcyon), Great blue heron (Ardea herodias),  and
bald eagle (Haliaeetus leucocephalus).   Four mammal
receptors  were selected:  the  little brown bat (Myotis
lucifugus), raccoon (Procyon lotor), mink (Mustela vison),
and river otter (Lutra canadensis).

Complete exposure  pathways and exposure parameters
(e.g., body weight, prey ingestion rate, home range) used
to calculate the  concentrations or dietary doses to which
the receptors of concern may be exposed were obtained
from USEPA references,  the  scientific  literature,  and
directly from researchers.  Site^specific fish, invertebrate,
sediment, and surface water data and model forecasts
were used to estimate PCB concentrations.

Measures of lexicological effects were selected based on
Lowest Observed Adverse Effects Levels (LOAELs) and/or
No Observed Adverse Effects Levels  (NOAELs) from
laboratory and/or field-based studies as reported  in the
scientific literature.   Reproductive effects (e.g.,  egg
maturation, egg hatchability, and survival of juveniles) were
generally the most sensitive endpoints for animals exposed
to PCBs.

The Ecological Risk Assessment indicates that piscivorous
receptors are at risk from adverse reproductive, growth, or
survival effects from exposure to PCBs in prey.  The major
findings of the report include:
EPA Region 2
                                            Page 12

-------
Superfund Proposed Plan
                   Hudson River PCBs Reassessment
•      Birds and mammals that eat PCB-contaminated
       fish from the Hudson River,  such as the bald
       eagle, belted kingfisher, great blue heron, mink,
       and river otter, are at risk. Piscivorous birds are at
       risk at least at the individual level and piscivorous
       mammals  are at risk  at the population level.
       PCBs may adversely affect the survival, growth,
       and reproduction of these species.

•      Piscivorous fish, such  as the largemouth bass
       and striped bass, in the Hudson River are at risk
       at the individual level. Population level effects are
       unlikely to be seen.

•      Fragile   populations   of   threatened  and
       endangered  species,  represented by the bald
       eagle, are particularly susceptible to adverse
       effects from PCB exposure.

•      PCB concentrations in water and sediments in the
       Upper Hudson River generally exceed standards,
       criteria,  and  guidelines   established  to  be
       protective of the environment.

•      Piscivorous birds and mammals are expected to
       be at considerable risk through 2018 (the entire
       forecast period).

•      Ecological receptors were found to be at risk on
       both a total PCB and dbxin-like PCB basis. Risks
       are  greatest  in  the  Upper Hudson  River,
       particularly in River Section 1 (Thompson Island
       Pool), and decrease in relation to decreasing PCB
       concentrations down river.

Based upon the results of the Remedial Investigation and
the risk assessments, USEPA has determined that actual
or threatened releases of hazardous substances from the
site, if not addressed by the preferred alternative or one of
the other active measures considered, may present  a
current or  potential threat to human  health  and  the
environment.

SCOPE AND ROLE OF ACTION

The primary objective of this action is to address the PCB-
contaminated sediments in the Upper Hudson River.
Removal of the PCB-contaminated sediments will reduce
PCB concentrations in  fish tissue, thereby minimizing
potential future human health and  ecological risks.  In
addition, remediation will control a source of PCBs to the
water  column   which  contributes   to   fish  tissue
concentrations, and transports PCBs downstream.

The Reassessment Feasibility  Study  is focused on the
approximately 40  river miles from  the northern end of
Rogers Island to the Federal Dam at Troy.  While the
Superfund site covers both the Upper and the Lower
Hudson River,  the  Reassessment  Feasibility Study
evaluates  options to address the PCB-contaminated
sediments of the Upper Hudson River only, because this
portion contains all of the historical PCB hot spots.

In conjunction with the remedy for the sediments of the
Upper  Hudson  River, as previously described in  this
Proposed Plan, USEPA will evaluate and either implement
or require  implementation of source control measures to
reduce PCBs that continue to be released into the Hudson
River at Bakers Falls through bedrock underlying the GE
Hudson Falls plant.

REMEDIAL ACTION OBJECTIVES

Remedial action objectives are specific goals to protect
human health and the environment. These objectives are
based  on  available information and standards  such as
applicable or  relevant and  appropriate  requirements
(ARARs) and risk-based levels established using the risk
assessments.  There are no federal or New York State
cleanup standards for PCB-contamination in sediment.
The  following  remedial  action  objectives  have been
established for the site:

•       Reduce the cancer risks and non-cancer health
        hazards for people eating fish from the Hudson
        River by reducing the  concentration of PCBs in
        fish. The risk-based Preliminary Remediation Goal
        for the protection of human health is 0.05 mg/kg
        PCBs  in  fish fillet  based on the reasonable
        maximum exposure adult fish consumption rate of
        one half-pound meal  per week.   Other target
        concentrations are 0.2 mg/kg PCBs in  fish fillet,
        which is protective at a fish consumption rate of
        one half-pound meal per month, and 0.4 mg/kg
        PCBs  in fish  fillet  which is  protective  of the
        average angler who consumes one half-pound
        meal every two months.  These targets  of higher
        concentrations in fish represent points at which fish
        consumption advisories and  fishing  restrictions
        might become less stringent (e.g.  the "eat none"
        advisory for the Upper Hudson could be relaxed as
        conditions improve).

•       Reduce the  risks  to  ecological  receptors by
        reducing the concentration of PCBs in fish.  The
        Preliminary Remediation Goal for the ecological
        exposure pathway  is a range from 0.3 to  0.03
        mg/kg  total PCBs  in  fish (whole body), which
        corresponds tp PCB concentrations of 0.12 to
        0.012  mg/kg  in fish  fillets.   The ecological
        Preliminary Remediation  Goal is  based on the
        Lowest Observed Adverse Effect Level and the No
        Observed Adverse Effect Level for consumption of
        whole fish by the river otter, an upper trophic level
        piscivorous mammal;

•       Reduce PCB levels in sediments in order to reduce
        PCB concentrations in river (surface) water that
        are above surface water ARARs. The ARARs for
        surface water are:

        1x10* ug/l (one part per quadrillion) total PCBs,
EPA Region 2
                                            Page 13

-------
 Superfund Proposed Plan
                   Hudson River PCBs Reassessment
               the New York State ambient water quality
               standard for the protection of health of
               human consumers offish.

       1.2 x 10"4 ug/l, the New York State standard for
       protection of wildlife,

       1 x 10*3 ug/l, the federal ambient water quality
       criterion for navigable waters,

       0.09  ug/l, the  New York State  standard for
       protection of human health and drinking water
       sources, and
       0.0005  mg/l  (0.5 ug/l), the  federal  maximum
       contaminant level for PCBs in drinking water;

       Reduce the   inventory  (mass)  of  PCBs  in
       sediments that are or may be bioavailabte; and

       Minimize the long-term downstream transport of
       PCBs in the river.
SUMMARY OF REMEDIAL ALTERNATIVES

CERCLA§ 121(bX1), 42 U.S.C. § 9621(b)(1), mandates
that remedial actions must be protective of human health
and the environment, cost-effective, and utilize permanent
solutions and  alternative treatment technologies and
resource recovery alternatives to the maximum extent
practicable.    Section  121(b)(1)  also  establishes a
preference for remedial actions which  employ,  as a
principal   element,  treatment to  permanently  and
significantly reduce the volume, toxicity, or mobility of the
hazardous substances, pollutants and contaminants at a
site.  CERCLA § 121(d), 42 U.S.C. § 9621(d), further
specifies that a remedial action must attain a level or
standard  of  control of  the  hazardous  substances,
pollutants, and contaminants, which at least attains
ARARs under federal and state laws, unless a waiver can
be justified pursuant to CERCLA § 121(d)(4), 42 U.S.C.
§9621(d)(4).

Sediments which may contribute to the PCB levels in fish,
both now and in the future,  are considered  principal
threats.   The determination of the significance of  the
 Principal threat wastes am trx>se source
 materials that act as a reservoir far flw migration
 of contamination. Principal threat wastes are
 those source materials considered to be highly
 toxic and present a significant risk to human
 health or the environment should exposure
 occur, or are highly mobile such that they,
 generally, cannot be reliably contained. The
 decision to treat these wastes is made on a site-
 specific basis through a detailed analysis of
 alternatives, using the remedy selection criteria
 mat are described below in the EVALUATION OF
 ALTERNATIVES section of this Proposed Plan.
 This analysis provides a basis for making a
 statutory finding as to whether the remedy
 employs treatment as a principal element.
sediment contribution to fish is based primarily on model
projections, in conjunction with geochemical analyses. The
model  projections indicate that the significance of the
sediment contribution varies by river section, therefore the
sediment levels that are considered principal threats will
correspondingly  vary  by  river section.   The  PCB-
contaminated sediment concentrations considered  to be
principal threats,  as represented by mass per unit area
measurements, are 3 g/m2 in River Section 1  and 10 g/m2
in River Section 2.  The mass per unit area  approach is
explained in  the  Screening section of the Summary of
Remedial Alternatives, below.

Screening

The  process used to  develop and screen  appropriate
technologies and  alternatives to address the  PCB-
contaminated sediments in the Upper Hudson River can be
found in the Feasibility Study. The technologies that were
carried forward after the initial screening are:

•      No Action (without upstream source control)

•      Monitored Natural Attenuation (MNA) (assuming
       upstream source control)

•      Capping  (assuming upstream  source control)
       followed by MNA

•      Removal  (assuming  upstream source control)
       followed by MNA

Capping  alternatives  considered  an  engineered  cap
(including a bentonite layer) of approximately 1-1/2 foot total
thickness. However, because the addition of this material
would greatly alter the geometry of the river  (shoreline) in
shallow areas, areas with less than 6 feet average depth
would first require dredging to compensate for the 1 -1/2 foot
raising of the  river bottom.   The terms "dredging" or
"removal" mean environmental dredging in the remainder of
this Proposed Plan.  In addition, because the river is used
for navigational purposes, it is impractical to cap the channel
(which  later may require navigational dredging). Removal
is the only active remediation that would be performed in the
channel.

Due to the high variability of PCB sediment concentrations,
mass per unit area (MPA), rather than concentration, was
identified as the  most useful  measure of the  potential
contribution of an area to PCBs in surface water and fish.
PCB inventory  in sediment  is represented  by  MPA
measurements (i.e., grams of PCBs  per square meter),
which indicate the total mass of PCBs within the sediment.
MPA was plotted against area of non-cohesive sediment
for the Thompson Island  Pool (and against PCB  mass
remediated)  to determine breakpoints where a  small
change in MPA would mean a large increase in sediment
area or mass to  be  remediated.   This   provides  an
evaluation of the  efficiency of remediation by comparing
the mass of PCBs remediated to the  amount of  the
sediment  surface  that  would  require   remediation.
Breakpoints were found at approximately 3  g/m2 and 10
g/m2.  For a core with a depth of one foot, 3 g/m2 is
equivalent to a concentration of 10 mg/kg and 10 g/m2 is
EPA Region 2
                                            Page 14

-------
Superfund Proposed Plan
                   Hudson River PCBs Reassessment
equivalent to approximately 30 mg/kg.  Therefore, the
screening  of alternatives evaluated Monitored Natural
Attenuation (no  sediment  remediation)  plus  source
control, 10 g/m2, 3 g/m2, and 0 g/m2(full section), for River
Sections 1 and 2. In River Section 3, the 0 g/m2 scenario
was excluded because it would have required remediation
of an  unreasonably  large  area  (over 2,800  acres).
Similarly, a cleanup level such as 1 mg/kg (as sometimes
used at other sites) would have targeted unreasonably
large areas in Section 3. The target levels are defined as:

       0 g/m2         Full Section Remediation

       3 g/m2          Expanded Hot Spot Remediation

       10 g/m2        Hot Spot Remediation

Modeling  was conducted to  evaluate the  impact of
remediation for combinations of the target levels for each
river section.   It was found  that remediation in River
Section 1, the Thompson Island Pool, had the greatest
benefit with respect to lowering PCB levels in fish  and
surface water.   The model  did  not show substantial
benefits from remediation in River Section 3.  However,
data show increased water column concentrations in this
reach resulting from tributary high flow events that caused
scour in the main part of the Hudson, thereby elevating
the water-column PCB concentrations. Therefore, certain
areas in Section 3, i.e., NYSDEC hot spots 36,37, and the
southern  portion  of 39, were  selected for remediation
based on  PCB inventory  and  signs of potential loss of
PCB inventory. For example, a comparison of 1977 and
1994 sediment data showed that  over two thirds of the
PCB inventory was lost from Hot Spot 37.

During the screening analysis, it was also determined that
if a remedy that included dredging were to be selected, it
would not be administratively feasible to dispose of that
material in a locally-sited landfill. Therefore, only off-site
disposal options were carried  through into the Detailed
Analysis.

Treatment technologies, such as thermal desorption, are
technically feasible; however, the associated costs would
be significantly greater than off-site landfill disposal, and
a  locally-sited  thermal treatment facility  would  not be
expected to be administratively feasible.

Detailed Analysis

Detailed descriptions of  the  remedial alternatives for
addressing the contamination associated with the site can
be found in the Feasibility Study.  The construction  time
for each  alternative  reflects only  the time required to
construct or implement the remedy and does not include
the time  required to design the remedy,  negotiate
performance of the remedy with the responsible parties,
or procure contracts  for design and construction.  The
present-worth costs for the alternatives discussed below
are calculated using a discount rate of seven percent and
a 30-year time interval.
The remedial alternatives are:

Alternative 1:  No Action (no Upstream Source Control)

The No Action alternative consists of refraining from the
active  application  of  any remediation  technology  to
sediments in all three sections of the Upper Hudson River.
The No Action alternative also does not assume any source
control removal action (i.e., the  NTCRA) near the GE
Hudson Falls plant, any administrative actions (including
institutional controls, such as fish  consumption advisories
and fishing restrictions, which are  considered to be limited
action under the NCP), and any monitoring. A review of site
conditions would be conducted at five-year intervals, as
required by CERCLA and the costs of monitoring sediment,
water and fish which is necessary to support the five-year
reviews is included in the table below.
 Capital Cost

 Operation and Maintenance Cost:
 (present worth)

 Present-Worth Cost

 Construction Time:
      $0

$140,000


$140,000

  0 years
Alternative 2: Monitored Natural Attenuation (NINA) with
Upstream Source Control

The Monitored Natural Attenuation (MNA) alternative relies
on naturally occurring attenuation processes to reduce the
toxicity, mobility and volume of the contaminants in the
Upper Hudson River sediments and assumes a separate
source  control  removal action (NTCRA) near the GE
Hudson Falls plant. Natural attenuation processes may
include  biodegradation,  biotransformation, bioturbation,
diffusion,  dilution,  adsorption,  volatilization,  chemical
reaction or  destruction,  resuspension,  downstream
transport,  and burial  by  clean  material.      Long-term
monitoring would be conducted in sediments, in the water
column, and in fish to confirm that contaminant reduction is
occurring and that  the reduction is achieving Remedial
Action Objectives. Monitoring will include measurements of
PCB concentrations in river water, dated sediment cores,
PCB inventory in  sediment, sediment physical properties
(geophysics),  and   bioaccumutation  in  resident  fish.
Reductions in PCB concentrations and the PCB inventory
could  be  documented  by historical  trends or  PCB
concentration distributions that show a reduction in the total
mass of PCBs in sediments, water, and/or biota, or by the
presence  of  degradation products  in  sediments.   The
monitoring  data would also be used as input parameters in
the mathematical models to evaluate progress of the natural
attenuation processes against the original predictions.
EPA Region 2
                                             Page 15

-------
 Superfund Proposed Plan
                   Hudson River PCBs Reassessment
  Capital Cost                            $417,000

  Operation and Maintenance Cost::      $38,000,000
  (present worth)

  Present-Worth Cost:                   $39,000,000

  Construction Time:                        0 years
  Capital Cost                        $344,000,000

  Operation and Maintenance Cost::      $24,000,000
  (present worth)

  Present-Worth Cost:                 $370,000,000

  Construction Time:                        5 years
These costs  do not include  the  capital  cost  nor  the
operation and maintenance costs  of the NTCRA.  The
capital cost associated with MNA  includes the costs of
developing and running the mathematical models; this cost
is included with the following alternatives as well because
they contain MNA as a necessary component.

Institutional controls would be implemented as long-term
control measures as part of the MNA alternative, including
continuation of fish  consumption advisories and fishing
restrictions, which are currently in place. A review of site
conditions would be conducted at five-year intervals, as
required by CERCLA.
Alternative  3:   CAP  3/10/Select - Capping, with
Removal to Accommodate Cap, followed by MNA, with
Upstream Source Control

This alternative includes remediation by capping (after
removal of more than 1.73 million cubic yards, in areas that
either cannot be capped (navigation channels) or required
sediment removal to allow for placement of the cap) of
sediments with an MPA of 3 g/m2 PCBs or greater in River
Section 1, sediments with an MPA of 10 g/m2  PCBs or
greater in River Section 2, and select sediments within high
concentration  PCB target  areas in River Section 3
(NYSOEC hot spots 36, 37 and the southern portion of
39). This alternative also includes sediment removal in the
navigation channel as necessary  to  implement  the
remediation   and  allow normal  boat  traffic  during
remediation. The total area of sediments to be remediated
is 493 acres, of which approximately 207 acres would be
capped.   The estimated  volume of sediments to be
removed is 1.73 million cubic yards, which is estimated to
contain 33,100 kg (73,000 tbs) of PCBs.  It would take
approximately 3 years to design and 5 years to implement
this remedy. This alternative assumes a separate source
control removal action (/.a, NTCRA) near the GE Hudson
Falls plant and also relies on naturally occurring attenuation
processes to reduce the toxicity, mobility, and volume of
the remaining PCBs in the Upper Hudson River sediments
after the construction is completed.  A review of  site
conditions would be conducted at five-year intervals, as
required by CERCLA.
As with Alternative 2. these costs do not include the capital
cost or the operation and maintenance  costs of  the
NTCRA.
Capping  involves  placement of an engineered  cap
consisting of low permeability material on top of the PCB-
contaminated sediment, including a top layer of fill. The
low permeability material prevents or retards the movement
of contaminated pore water into the water column and
minimizes exposure of benthic organisms to the PCB-
contaminated sediments. The selected process option for
containment  is AquaBtok"" (or a similar material), a
manufactured product consisting of a composite of gravel
particles  encapsulated with bentonite.   Once deployed
through the  water column,  the  heavy center of  the
composite material carries the bentonite bearing particles
to the bottom where the bentonite absorbs water and
expands to form a continuous impermeable mat
A 12-inch  layer of AquaBlok"" was selected  for  the
conceptual approach in the Feasibility Study for several
reasons.  The proposed thickness would have a higher
probability of withstanding damage  from ice scour and
navigational incidents, as well as erosion due to normal or
storm-induced  flows,   without  exposing  the  high
concentrations of PCBs that currently exist  in the surface
sediments at some locations.
A 6-inch benthic substrate layer would be placed on top of
the AquaBlok1"1 layer to  prevent bioturbation of the  cap
material and to serve as a clean  habitat for  the benthic
organisms to repopulate. This material would also serve
as a sacrificial layer in the event of erosion or damage,
possibly allowing repairs to be conducted  before further
damage occurs.
Placement of 18 inches (1.5 feet) of capping material over
the existing surface, especially in shallower areas,  could
affect the hydraulics of the river, as well as  actually move
the shoreline toward the channel by as much as 25 to 50
feet in some areas.   Therefore, in order to prevent
changing the configuration of the river, 1.5 feet of sediment
would be removed prior to the placement of the cap in
shallow areas.
Sediment removal  would be accomplished with similar
equipment described for the removal alternatives below.
Production rates and sediment processing facilities would
be similar, but with appropriate quantity changes.
After construction is completed, this alternative relies on
institutional controls,  such  as the fish  consumption
advisories and fishing restrictions, and perhaps restrictions
 EPA Region 2
                                            Page 16

-------
 Superfund Proposed Plan
                    Hudson River PCBs Reassessment
 on activities that could compromise the integrity of the
 cap, and MNA, in areas not remediated until the Remedial
 Action Objectivess are achieved. A long-term monitoring
 program would be required to verify the integrity of the cap
 and to assess the effectiveness of the cap and natural
 attenuation processes in achieving the Remedial Action
 Objectives. If any portion of the cap has been eroded, it
 would require replacement. A review of site conditions
 would be conducted at five-year intervals, as required by
 CERCLA.
 Alternative 4: REM-3/10/Select - Removal followed by
 MNA, with Upstream Source Control
 This alternative includes remediation by removal of all
 sediments with an MPA of 3 g/m2 PCBs or greater in River
 Section 1, removal of all sediments with an MPA of 10 g/m2
 PCBs or greater in River Section 2. and removal of select
 sediments  with high concentrations  of PCBs  in River
 Section 3 (NYSDEC hot spots 36, 37, and the southern
 portion of 39).  This alternative also includes sediment
 removal  in  the  navigation  channel as  necessary  to
 implement the remediation. The total area of sediments
 targeted for removal is approximately 493 acres.  The
 estimated  volume of sediments to be removed is 2.65
 million cubic yards which is estimated to contain 45,600 kg
 (100,600 Ibs) of PCBs. It would take approximately 3 years
 to design and 5 years to implement this remedy.   This
 alternative assumes a separate source control  removal
 action (i.e., NTCRA) near the GE Hudson Falls plant. After
 construction  is  completed,  this  alternative  relies on
 institutional  controls,  such  as  the  fish  consumption
 advisories and fishing restrictions  (although perhaps in a
 modified form), and MNA in areas not remediated until the
 Remedial Action Objectives are achieved. A review of site
 conditions would  be conducted at five-year intervals, as
 required by CERCLA.

  Capital Cost                        $448,000,000

  Operation and Maintenance Cost::      $13,000,000
  (present worth)

  Present-Worth Cost:                 $460,000,000

  Construction Time:                         5 years

 As with Alternatives 2 and 3, these costs do not  include
 the capital cost or the operation and maintenance costs of
 the NTCRA.
Alternative 5: REM-0/0/3 - Removal followed by MNA
with Upstream Source Control

This alternative includes Full Section remediation by
removal in  River  Sections  1  and  2,  and removal of
sediments with an MPA of 3 g/m2 PCBs or greater in River
Section 3. This alternative also includes sediment removal
in the navigation channel as necessary to implement the

EPA Region 2
 remediation.  The  total area  of sediments targeted for
 removal is approximately  964 acres.  The  volume of
 sediments to be removed is estimated to be 3.82 million
 cubic yards which is estimated to contain more than 70,150
 kg (155,000 Ibs) of PCBs.  It would take approximately 3
 years to design and 7 years to implement this remedy.  This
 alternative assumes a separate source control removal
 action (i.e.. NTCRA) near the GE Hudson Falls plant. After
 construction is  completed,  this  alternative  relies  on
 institutional controls,  such  as  the  fish  consumption
 advisories and fishing restrictions, and MNA in areas not
 remediated until  the  Remedial Action Objectives are
 achieved.  A review of site conditions would be conducted
 at five-year intervals, as required by CERCLA.

  Capital Cost                         $556,000,000

  Operation and Maintenance Cost::      $13,000,000
  (present worth)

  Present-Worth Cost:                 $570,000,000

  Construction Time:                         7 years

 As with Alternatives 2 through 4, these costs do not include
 the capital cost or the operation and maintenance costs of
 the NTCRA.
General Removal Information (applicable to Alternatives
3 (in part), 4 and 5)

Removal by targeted dredging is the principal component of
the two REM alternatives and a major component of the
CAP alternative.  The criteria for selection of targeted areas
are based primarily on mass per unit area (e.g., 3 g/m2,10
g/m2) and PCB concentrations in surface sediment, as well
as engineering considerations, such as  minimum areas
targeted (50,000  square feet).

Of the  various dredging technologies reviewed  for the
Feasibility Study, both mechanical systems and hydraulic
systems appear to be applicable to conditions found in the
Upper Hudson River.  Dredging productivity, sediment in-
river transport/conveyance, and sediment processing would
vary between mechanical and hydraulic  systems.  Both
methods have been considered in the development and
evaluation of  alternatives to  preserve  options  in the
remedial design.
Within the target areas, the goal is to remove all of the PCB-
contaminated sediment, leaving a residual of approximately
1 mg/kgorless. Subsequent to removal, approximately one
foot of backfill would be placed where appropriate (excluding
the navigation channels) over the residual layer, which
would further reduce the available PCB concentration at the
surface and provide an appropriate substrate for biota.  In
addition, the backfill will  help stabilize  bank areas after
dredging and minimize hydraulic changes  to the river.
The  dredged sediments would be transported to land-

                                            Page 17

-------
 Superfund Proposed Plan
                   Hudson River PCBs Reassessment
 based sediment processing facilities. At these facilities
 the  sediment  would  be  dewatered  to  the  extent
 practicable. Portland cement would be added to the solids
 portion to stabilize it before loading onto rail cars.  The
 sediments would be disposed of at an existing licensed
 TSCA or solid waste landfill outside of the Hudson Valley.
 Siting  of  a local  landfill was screened  out due  to
 community objection.  Another solids  disposal option
 involves beneficial use of non-TSCA dredged material.

The water that is separated will undergo treatment to
 remove fine sediment particles and dissolved  PCBs.
 Ultimately, the water will be discharged back into the
 Hudson River in compliance with substantive New York
State  Pollutant   Discharge  Elimination  System
requirements, which are ARARs for this site.


EVALUATION OF ALTERNATIVES

In selecting a remedy for a site, USEPA considers the
factors set forth in CERCLA § 121,42 U.S.C. § 9621,  by
conducting a detailed analysis of the  viable remedial
alternatives pursuant to the NCP, 40 CFR § 300.430(e)(9),
USEPA's  Guidance  for  Conducting  Remedial
Investigations and Feasibility Studies, OSWER Directive
9355.3-01, and USEPA's A Guide to Preparing Superfund
Proposed Plans, Records of Decision, and Other Remedy
Selection  Decision  Documents, OSWER 9200.1-23.P.
The detailed analysis consists of an  assessment of the
individual  alternatives against each  of  nine evaluation
criteria and a comparative analysis  focusing upon the
relative performance of each alternative against those
criteria.

A comparative analysis of these alternatives based upon
the evaluation criteria (see box below) follows.
The comparison of the effectiveness of alternatives  is
based on the results of modeling each remedial alternative
as well as data projections.  Comparisons of the model
outputs to recent data trends suggest that the model may
be overly optimistic with regard to the rate of PCB decline
in fish predicted for the No Action (no source control) and
Monitored Natural Attenuation (assuming source control)
alternatives. This occurs because the model predictions
are averaged over larger spatial scales than  the foraging
range of many resident fish species.  Under the modeled
remedial  alternatives, this over-optimism is eliminated
wherever  PCBs  are  removed or capped,   because
projected rates of decline are replaced  by  specified
concentrations in the remediated areas. Consequently, the
benefits of remediation based on comparisons of the active
remediation alternatives to the No  Action and Monitored
Natural Attenuation alternatives are likely underestimated
by the models.

In order to bound this uncertainty in the No Action and MNA
alternatives, an  estimated   upper  bound  was also
calculated. Assuming that the over-optimism in the model
projections  stems from the uncertainty  surrounding the
PCB concentration in surface sediment calculated by the
model, an alternative method was used to calculate surface
sediment values  based  on  certain  fish data.   PCB
concentrations  in brown bullhead, which  are  affected
primarily by concentrations of PCBs in surface sediment,
were  used to back-calculate concentrations of  PCBs  in
surface sediment that would produce the decline seen  in
the data.  The newly calculated concentrations of PCBs in
surface sediment were then used in the model as an upper
bound estimate instead of the model-calculated surface
sediment values.   Therefore,  both the upper  bound
estimates and  the model calculated values for No Action
and MNA were used as  points of comparison  in the
                      NINE EVALUATION CRITERIA FOR SUPERFUND REMEDIAL ALTERNATIVES
        Overall Protection of Human Health and the Environment determines whether an alternative eliminates,
        reduces, or controls threats to public health and the environment through institutional controls, engineering controls,
        or treatment.
        Compliance wfth ARARs evaluates whether the alternative meets Federal and State environmental statutes,
        regulations, and other requirements that pertain to the site, or whether a waiver is justified.
        Long-term Effectiveness and Permanence considers the ability of an alternative to maintain protection of human
        health and the environment overtime.
        Reduction of Toxicity, Mobility, or Volume of Contaminants through Treatment evaluates an alternative's use
        of treatment to reduce the harmful effects of principal contaminants, their ability to move in the environment, and the
        amount of contamination present
        Short-term Effectiveness considers the length of time needed to implement an alternative and the risks the
        alternative poses to workers, residents, and the environment during implementation.	
        Implementability considers the technical and administrative feasibility of implementing the alternative, including
        factors such as the relative availability of goods and services.
        Cost includes estimated capital and annual operation and maintenance costs, as well as present worth cost.
        Present worth cost is the total cost of an alternative over time in terms of today's dollar value. Cost estimates are
        expected to be accurate within a range of +50 to -30 percent.
        Slate Acceptance considers whether the State agrees with the EPA's analyses and recommendations, as
        described in the RI/FS and Proposed Plan.
        Community Acceptance considers whether the local community agrees with EPA's analyses and preferred
        alternative. Comments received on the Proposed Plan are an important indicator of community acceptance.
EPA Region 2
                                             Page 18

-------
Superfund Proposed Plan
                   Hudson River PCBs Reassessment
evaluation of effectiveness of the capping and removal
alternatives.
1.   Overall Protection of Human Health  and the
Environment
Protection of Human Health

Alternative 1 (No Action) would not be protective of human
health and the environment because it would not address
the PCBs in the sediment, or from the upstream source,
which result in risks to humans, birds, fish, and mammals
that  are above  levels of concern  under the federal
Superfund program.

USEPA evaluated overall protection of human health in
two primary ways.  The first is the time that it would take
under each of the alternatives to reach the fish Preliminary
Remediation Goal  and the other target concentrations.
The second is the  relative reduction in cancer risks and
non-cancer health  hazards under the  five  remedial
alternatives.  Each  approach is discussed below.

Time to Reach Fish Target Levels

The fish Preliminary Remediation Goal is 0.05 ppm (or
mg/kg) PCBs (wet weight) in fillet  In addition, USEPA
considered a target concentration of 0.2 ppm PCBs (wet
weight) in fillet based on one half-pound meal per month,
and  a target concentration of 0.4  ppm  based on the
average (central tendency) consumption rate of one half-
pound meal every 2 months. The target concentrations
correspond  to points at which the fish consumption
advisories could be relaxed from the current "eat none"
recommendation in the Upper Hudson River to one of
limited fish ingestion.

In River Sections 1 and 2, and for the Upper Hudson River
as a whole, none of the alternatives meets the human
health Preliminary Remediation Goal of 0.05 ppm PCBs
within the modeling  time frame (to 2067),  unless the
upstream source is virtually eliminated.  However, even if
source control measures cannot completely eliminate the
upstream source, significant reductions in  risk can still be
achieved by implementation of one of the active remedies.
In River Section  3,  all of the  active  remediation
alternatives meet the Preliminary Remediation Goal of
0.05 ppm PCBs.  The MNA alternative also reaches it,
although it  takes  longer to  achieve.  The No Action
alternative does not meet the Preliminary Remediation
Goal within the modeling time frame.
Years to Reach Target Concentration in Fish
Averaged Over Entire Upper Hudson River

No Action
MNA
CAP-
3/10/Select
REM-
3/10/Select
REM-0/0/3
0.05 ppm
>67
>67
>67
>67
>67
0.2 ppm
>67
60 to
>67
35
35
26
0.4 ppm
>67
34 to
>67
21
20
11
 Note. ">" = greater than

As  can  be seen  in the table above,  the three active
remediation alternatives are projected to significantly
reduce the  time necessary to reach the 0.2 ppm and 0.4
ppm fish target levels.  The most aggressive alternative
analyzed, REM-0/0/3, would reduce the time to reach fish
target levels by the greatest amount.  These projected
improvement differentials are believed to be  on the
conservative side due to the previously discussed model
over-optimism related to No Action and MNA.

Relative  Reductions  in Cancer Risks  and Non-Cancer
Health Hazards

The years  included  in the exposure  calculation were
calculated on a river section-specific basis using different
long-term  period   starting   dates,  depending  on  the
construction schedule for each remedial alternative. The
long-term exposure  period  was  considered  to  start
immediately after a one-year equilibration period beyond
the completion of work in a given section. For example, if
the construction schedule for an alternative requires three
years to complete in River Section 1, given a start date in
2004, the construction would be complete at the end of
2006, equilibration would occur during the year 2007, and
the long-term period for calculation of cancer risks and
non-cancer health hazards would start on January 1,2008.

Cancer risks and non-cancer health hazards for the entire
Upper Hudson River under the active remedial alternatives
were then compared  separately (to the appropriate time
frame) to the No Action and MNA  alternatives, including
their estimated upper bounds, to estimate the reduction in
cancer risks and non-cancer health hazards achieved by
each alternative.

The fish concentrations used are  the  species-weighted
average, based on relative species consumption reported
in the 1991 state-wide New York Angler survey.  The
species-weighted average represents the relative intake of
the three modeled fish species consumed by anglers and
EPA Region 2
                                            Page 19

-------
 Superfund Proposed Plan
                  Hudson River PCBs Reassessment
 their families:  largemouth bass (47 percent);  brown
 bullhead  (44 percent); and yellow perch (9 percent).
 Exposure durations were the same as those presented in
 the Human Health Risk Assessment (cancer 40 years
 RME and 12 years central tendency; non-cancer, 7 years
 RME and 12 years central tendency). The 7 year time
 frame for non-cancer health hazards reflects a chronic
 dose for non-cancer health effects as well as the declining
 PCB concentrations with time.

 Upper Hudson River (RMs 194.5 -154)

 Non-Cancer Health Hazards

 Reductions of from 61% to 90% in the RME adult non-
 cancer health hazard indices are achieved by all active
 alternatives, compared  to  the  No Action  and MNA
 alternatives (including estimated upper bounds).  The
 MNA  alternative  achieves  a 25% to  35%  reduction
 compared to the No Action alternative, but only an 11 % to
 14% reduction compared to the estimated upper bound for
 the No Action alternative. The CAP-3/10/Setect alternative
 achieves a 71% to 81% RME risk reduction compared to
 No Action  and a 61% to 78% RME  risk  reduction
 compared to  MNA.   The  REM-3/10/Setect alternative
 achieves a 75% to 84% RME risk reduction compared to
 No Action  and a 67% to 82% RME  risk  reduction
 compared to MNA. The REM-0/0/3 alternative achieves
 a 84% to 90% RME risk reduction compared to No Action
 and a 77% to 88% RME risk reduction compared to MNA

 Cancer Risks

 Substantial (54% to 91%) reductions in RME adult cancer
 risk are achieved by  all active remediation alternatives,
 compared to the No Action  and MNA alternatives
 (including estimated upper bounds). The MNA alternative
 achieves  a  31% to 52% reduction compared to the No
Action alternative, but only a  13% to 18%  reduction
compared to the estimated upper bound for the No Action
 alternative. The CAP-3/10/Setect alternative achieves a
 76% to 87% RME risk reduction compared to No Action
 and a 54% to 84% RME risk reduction compared to MNA.
 The REM-3/10/Select alternative achieves a 79% to 88%
 RME risk reduction compared to No Action and a 58% to
 86% RME risk reduction compared to MNA.  The REM-
 0/0/3  alternative  achieves a  84% to 91% RME  risk
 reduction compared to No Action and a 66% to 89% RME
 risk reduction compared to MNA.

 Overall Protection of the Environment

 Ecological risks were calculated for the river otter and the
 mink.   The river otter is  a fish-eating  (piscivorous)
 mammal and was the receptor found to be at greatest risk
 at the site due to the high proportion offish in its diet. The
 mink is a  piscivorous mammal and is  known to  be
 sensitive to PCBs. Other species, such as the bald eagle,
 were also considered but are at less risk  than the river
 otter.  Similar to the overall protection of human health,
 the long-term exposure period for the river otter and mink
is  considered to start  immediately after a  one-year
equilibration period beyond the completion of work in a
given section.

River Otter

The  implementation of active remediation alternatives
results in 63% to 87% reduction in risk to the river otter
compared to the modeled No Action alternative and an
82% to 93% reduction in risk compared to the upper bound
estimate for the No Action alternative, except in  River
Section 3, which does not show a difference between the
active alternatives and the MNA alternative.

Mink

The implementation of the active remediation alternatives
results  in 64% to  87%  reduction in risk to the mink
compared to the No Action alternative and a 81% to 93%
reduction in risk compared to the estimated upper bound
for the No Action alternative,  excluding River Section 3,
which does not show a difference between the No Action
and MNA alternatives.

Transport of PCBs over Federal Dam

Reduction of the PCB load transported over the Federal
Dam and into the Lower Hudson is also a gauge of the
overall protectiveness of a remedy. Reduced PCB loading
from the Upper Hudson into the  Lower Hudson will
ultimately reduce concentrations of PCBs in sediment,
water, and fish and resulting human health and ecological
risks within the Lower Hudson. The REM-0/0/3 alternative
results in a 53% reduction of PCB  loading in the year
following completion of remediation (2011) in comparison
with  the  MNA alternative.   Both of these alternatives
assume source control upstream. For the REM-3/10/Setect
alternative, the PCB loading to the Lower Hudson in 2011
would be reduced by 42% as compared to MNA. For the
CAP-3/10/Setect alternative, the PCB loading to the Lower
Hudson in 2011 would be reduced by 38% as compared to
MNA. The comparison of the active remedies to the No
Action alternative would exhibit even greater PCB loading
reductions, since the  No  Action alternative does  not
assume source control.

2.  Compliance with ARARs

The  chemical-specific ARARs for PCBs in the water-
column are 0.5 ug/L (500 ng/L) federal Safe Drinking Water
Act MCL; 0.09 ug/L (90 ng/L) NYS standard for protection
of human health and drinking water sources; 1 ng/L federal
Ambient Water Quality Criterion; 0.12 ng/L NYS standard
for protection of wildlife; and 0.001  ng/L NYS standard for
protection of human consumers of fish.

Two chemical-specific ARARs for the surface water would
be met by all five remedial alternatives, while the remaining
three chemical-specific ARARs for the surface water are
not expected to be met by any of the five alternatives for
the 70-year forecast period.  Evaluation of the projected
EPA Region 2
                                          Page 20

-------
 Superfund Proposed Plan
                   Hudson River PCBs Reassessment
 PCB concentration in the water column by river section
 shows that the source control action near the GE Hudson
 Falls plant affects the difference (separation) between the
 rate of decline for the No Action and MNA alternatives.
 However,  the benefits of active  remediation of the
 sediments are readily apparent in the differences in the
 rate of decline for the MNA alternative and those for the
 active remediation alternatives.  As expected, the water
 quality is best  for  the  REM-0/0/3 alternative  and
 substantially improved for the CAP and REM-3/10/Select
 alternatives, compared to MNA.  These differences are
 most apparent for the first 20 years (between 2005 and
 2024) of the forecast period. However, even towards the
 end of the forecast period (in 2067), there is a  very
 substantial difference between the water quality for the No
 Action alternative (approximately 30 ng/L at Thompson
 Island Dam and Schuylerville and 10 ng/L at Federal
 Dam) and the other four alternatives (approximately 5
 ng/L at Thompson Island Dam and Schuylerville and 1.7
 ng/L at Federal Dam).

 Because there is no active remediation associated  with
 the sediments for the No Action and MNA alternatives,
 action-specific and location-specific ARARs do not apply.
 The three active remedial alternatives would comply with
 action-specific ARARs (e.g., Clean Water Act  Sections
 401 and 404; Toxic Substances Control Act; Section 3004
 of the Resource Conservation and Recovery Act; Section
 10 of the Rivers and Harbors  Act; New York State ECL
 Article 3, Title 3 and Article  27, Titles 7  and 9)  and
 location-specific ARARs (e.g., Endangered Species  Act;
 Fish and Wildlife Coordination Act; Farmland Protection
 Policy Act; National Historic Preservation Act; and New
 York State Freshwater Wetlands  Law.   Analysis of
 potential effects on wetlands and floodplains associated
 with the preferred remedial alternative will be performed
 during  remedial   design,  as  necessary,  to ensure
 compliance with Executive Orders 11990 and 11988 for
 wetlands and floodplains, respectively.

 3. Long-Term Effectiveness and Permanence

 Reduction of Residual Risk

 The  No  Action  and  MNA  alternatives  result in a
 continuation of the degraded  condition  of the surficial
 sediments and surface water quality of the Upper Hudson
 River, especially in the Thompson Island Pool, for several
 decades,  regardless of any reduction in the upstream
 water quality PCB loadings. These  alternatives remove
 no PCBs from the River and effect no active reduction in
 PCB levels in fish, other than through naturally occurring
 processes.  The MNA alternative,  for purposes of the
 Feasibility Study, does include upstream source control,
 and therefore,  will show reduced risks when compared to
 the No Action alternative.

 For the  CAP-3/10/Select  alternative,  residual risk is
 reduced  through  the capping of 207 acres of  PCB-
 contaminated sediments and removal of 1.73 million cubic
yards of sediments containing 73,000 Ibs (33,100  kg)

EPA Region 2
 PCBs. The total area remediated (capped plus removed)
 via this alternative encompasses 493 acres, and the total
 quantity of PCBs remediated totals 100,600 Ibs (45,600
 kg).   The  reduction in  cancer risks  through fish
 consumption ranges from 76% to 87% compared to the No
 Action alternative and  from 54% to 84% compared to the
 MNA alternative.   The reduction  in non-cancer health
 hazards ranges from 71% to 81% compared to the No
 Action alternative and from 61% to 78% compared to the
 MNA alternative.

 For the  REM-3/10/Select  alternative, residual  risk  is
 reduced through the removal of 2.65 million cubic yards of
 sediments containing 100,600 Ibs (45,600 kg) PCBs over
 an area of 493 acres.  The  reduction in  cancer  risks
 through fish  consumption  ranges from 79% to  88%
 compared to the No Action alternative and from 58% to
 86% compared to the MNA alternative.  The reduction in
 non-cancer health  hazards ranges from 75% to  84%
 compared to the No Action alternative and from 67% to
 82% compared to the MNA alternative.

 For the REM-0/0/3 alternative, residual risk is reduced
 through the  removal of 3.82  million cubic yards of
 sediments containing more than 145,000 Ibs (63,500 kg)
 PCBs over an area  of 964 acres. The reduction in cancer
 risks through fish consumption ranges from 84% to 91%
 compared to the No Action alternative and from 66% to
 89% compared to the MNA alternative.  The reduction in
 non-cancer health  hazards ranges from 84% to  90%
 compared to the No Action alternative and from 77% to
 88% compared to the MNA alternative.

 Based on the above analysis, the  three active  remedial
 alternatives are far superior to  the No Action and MNA
 alternatives in terms of this criterion due to the significant
 differences in  risk reduction and mass of PCBs removed
from the river.  The  three action alternatives are similar in
terms  of  risk  reduction;  however, the two  removal
 alternatives rank higher than the capping alternative due to
the quantities of PCBs removed from the river and the
 permanence  of  such  removal versus the  long-term
operation and maintenance required by capping of a
 portion of the PCB-contaminated sediments.

Adequacy of Controls

The No Action and MNA alternatives do not provide for
engineering controls on the river sediments. The MNA
alternative  does assume source control near the GE
 Hudson Falls plant and institutional controls. NYSDOH's
 1996  study of anglers in the Upper and Lower Hudson
 found that about 18% of the Upper Hudson respondents
 had fish in their possession when interviewed and 11 % had
 more than one fish. Most of the fish were largemouth bass,
 smallmouth bass, and bluegill, species that are often eaten;
 in the Mid-Hudson region, about 8% actually had fish in
 their possession when interviewed. Therefore, the existing
 institutional controls,  which rely heavily on voluntary
 compliance, are not adequate in reducing exposure to
 PCBs  due  to consumption of contaminated fish.  In

                                           Page 21

-------
 Superfund Proposed Plan
                   Hudson River PCBs Reassessment
 addition,  institutional  controls  are  inadequate  for
 protection of the environment.

 The CAP-3/10/Select alternative  provides  for select
 removal of some PCB-contaminated sediments in target
 areas and placement of an engineered cap over the
 remaining target areas.  Like the MNA alternative,  this
 alternative also provides for institutional controls, such as
 the fish consumption advisories and fishing restrictions
 (although perhaps in a modified form), and other site use
 restrictions in capped areas (e.g., sediment disturbance
 activities  such as waterfront improvements  by  private
 residences or commercial/industrial establishments along
 the shoreline).
 The REM-3/10/Select and REM-0/0/3 alternatives provide
 for removal of PCB-contaminated sediments in target
 areas. These two alternatives also provide for institutional
 controls, such as the fish consumption advisories and
 fishing restrictions (although perhaps in a modified form),
 but they are  unlikely to require additional site  use
 restrictions after removal activities are completed.

 Reliability of Controls

 Sediment  capping,  sediment removal  (dredging  and
 excavation), habitat replacement/backfilling, and off-site
 disposal/treatment of removed sediments are all reliable
 and proven  technologies.   However,  for  the  CAP-
 3/10/Select alternative, proper design,  placement, and
 maintenance of the cap in perpetuity are required for its
 effectiveness, continued performance, and reliability. The
 cap integrity  monitoring and maintenance program
 planned for the CAP-3/10/Setect alternative provides for
 reasonable  reliability.   Also, the  fish  consumption
 advisories and fishing restrictions will continue to provide
 some measure of protection of human health until PCB
 concentrations in fish are reduced to the point where the
fish consumption advisories and fishing restrictions can be
 relaxed or lifted.

 In general, the REM-3/1 OVSetect and REM-0/0/3 are the
 most  reliable,  as  there is  little  or no  long-term
 maintenance or residual risk associated with the remedial
 work. Of the removal alternatives, REM-0/0/3 is the most
 reliable, as it permanently removes the greatest amount
 of sediment (leaving the least amount of PCBs in the river)
 and achieves the greatest reduction of the potential scour-
 driven  resuspenskxi of  PCB-contaminated  sediments
 south of the confluence with the Hoosic River. The CAP-
 3/10/Select alternative does not achieve the same degree
 of reliability due to the potential for damage to the cap,
 thereby reducing its effectiveness, and would still require
 all of the sediment handling, processing,  and disposal
 required for the  removal alternatives.   The  No Action
 alternative is  the  least  reliable.  Although the MNA
 alternative is better than the No Action alternative, the
 institutional controls associated with this alternative do not
 protect ecological receptors, and human risk reduction
 relies on knowledge of and voluntary compliance with the
 fish consumption advisories and fishing restrictions.
4. Reduction in Toxicitv. Mobility, or Volume Through
Treatment
The No Action and MNA alternatives do not involve any
containment or removal of contaminants from the Upper
Hudson River sediments.  Because the MNA alternative
assumes a separate source control (NTCRA) near the GE
Hudson Falls plant, the PCB load to the water  column
upstream of the Thompson Island Pool is expected to be
reduced from 0.16 kg/day to 0.0256 kg/day by January 1,
2005. The No Action and MNA alternatives rely on natural
attenuation processes such as burial by cleaner sediments,
biodegradation,  bioturbation,  and  dilution  to  reduce
concentrations of PCBs in sediments and surface water.
Biodegradation processes may partially convert some of the
more highly chlorinated PCB congeners to less-chlorinated
congeners and biphenyls, and thereby reduce their toxicity.
Concentrations of PCBs in fish will respond slowly over time
to slow natural decreases  in concentrations  in sediments
and surface water.

For the CAP-3/10/Select alternative, the mobility of the
PCBs in capped areas (approximately 207 acres) is reduced
because these PCBs are sequestered under  the bentonite
cap.  However, capping does not satisfy the CERCLA
statutory preference for treatment. In addition, there is no
reduction in the toxicity or volume of the PCBs under the
cap. Under this alternative, the mass of PCBs and the
volume of contaminated sediments within the Upper Hudson
River are permanently reduced because approximately 1.73
million cubic yards of sediment, which contain an estimated
73,000 Ibs (33,100 kg) of  PCBs, are removed. A total of
100,600 Ibs (45,600 kg) would be removed or isolated from
the ecosystem by this alternative.   Because the CAP-
3/10/Select  alternative also assumes  source  control
(NTCRA) near the GE Hudson Falls plant, the PCB load to
the water column  is expected to be reduced from 0.16
kg/day to 0.0256 kg/day by January 1, 2005.  In addition,
after construction  of  the  remedy is completed, natural
attenuation processes will provide further  (but  slower)
reductions  in the  toxicity of PCBs in  the remaining
sediments and surface water.

For the REM-3/10/Select and REM-0/0/3 alternatives, the
mass of PCBs and volume of contaminated sediments in the
Upper Hudson River  are  permanently reduced because
sediment volumes from 2.65 to 3.82 million cubic yards,
respectively, containing a mass of PCBs from 100,600 Ibs
(45,600 kg) (REM-3/10/Select) to an estimated mass of
greater  than  140,000  Ibs (63,500 kg) (REM-0/0/3) are
removed from the  ecosystem.   Because these removal
alternatives  also assume source control  near the GE
Hudson Falls plant, the PCB load to the water column is
expected to be reduced from 0.16 kg/day to 0.0256 kg/day
by January 1,2005. Also, as stated for the CAP-3/10/Select
alternative, after construction of the remedy  is completed,
natural  attenuation processes will  provide further  (but
slower) reductions in the toxicity of PCBs in the remaining
sediments and surface water.
 EPA Region 2
                                            Page 22

-------
Superfund Proposed Plan
                   Hudson River PCBs Reassessment
in all three active remediation alternatives, the sediments
that are removed undergo limited treatment (stabilization
with Portland cement) prior to landfill disposal. A different
treatment process may be employed for the beneficial use
option. However, the action alternatives do remove large
volumes (and therefore significantly reduce mobility and
toxicity) of PCBs from the river,

5.  Short-Term Effectiveness

Protection of the Community During Remedial Actions

No construction  activities  are associated  with  the
remediation of sediments for the No  Action and MNA
alternatives, so neither alternative increases the potential
for direct contact with or ingestion and inhalation of PCBs
from the surface water and sediments.

Transfer facilities and treatment areas present potential
short-term risks to the  community  under the  active
alternatives.  Therefore, access to these areas will be
restricted to authorized personnel. In addition, monitoring
and engineering controls will be employed to minimize
short-term effects due to material processing  activities.
Increased traffic will also present an incremental risk to the
community.   The  potential for traffic accidents may
increase marginally as additional vehicles are on the road.
These effects  are  likely  to be minimal because most
transportation  of  sediments   for  disposal   will   be
accomplished by rail.  In addition to vehicular traffic, there
will be increased river traffic.  Work areas in the river will be
isolated (access-restricted), with an adequate buffer zone
so that pleasure craft and commercial shipping can safely
avoid such areas.  Finally,  the increased in-river barge
traffic will be monitored and controlled to minimize, to the
extent possible, adverse  effects on the commercial or
recreational use of the Upper Hudson River.  Targeted
dredging will be sequenced and directed to ensure that the
navigation channel  is not  closed  due to construction
activities.

Protection of Workers During Remedial Actions

For the No Action alternative, occupational risks to persons
performing the sampling activities (for the 5-year reviews)
will be unchanged  from current  levels.   There is an
increase  in occupational  risk associated with the MNA
alternative due to the greater degree of sampling involved
in  the river (and for the source control activities near the
GE Hudson Falls plant).  For the three active remediation
alternatives (CAP-3/10/Select. REM-3/10/Selectand REM-
0/0/3), potential occupational  risks to site workers from
direct contact, ingestion, and inhalation of PCBs from the
surface water and sediments, and routine physical hazards
associated with construction work and working on water,
are significantly higher than for the No Action  and MNA
alternatives.  For these alternatives, as well as the No
Action and MNA alternatives, personnel will follow a site-
specific health and safety plan, OSHA health and safety

EPA Region 2
procedures, and wear the necessary personal protective
equipment.

Potential  Adverse Environmental   Impacts  during
Construction

No  construction  activities  associated with  the  river
sediments  are conducted for the No  Action  and MNA
alternatives.  Neither continuation of the existing limited
sampling activities for the No Action alternative nor the
increased monitoring program for the MNA alternative is
anticipated to have any adverse effect on the environment,
beyond that already caused by the PCB contamination of
the sediments in the Upper Hudson River.

For  the  three  active  remediation  alternatives  (CAP-
3/10/Select, REM-3/10/Select and REM-0/0/3), the release
of PCBs from the contaminated sediments into the surface
water during construction (dredging and cap placement), as
well as the resuspension of sediment, will be controlled by
operational practices (e.g., control of sediment removal
rates, use of environmental dredges, and use of sediment
barriers).  Although precautions to minimize resuspension
will be taken, it is  likely that  there will be a  temporary
increase  in suspended PCB concentrations in the water
column, and possibly in fish PCB body burdens. Studies at
other sites, such as the  Fox River Demonstration Project,
have shown that such effects are controllable, small, and
transient, and that longer-term improvement is seen.

Remedial activities may also result in short-term temporary
impacts to aquatic and wildlife habitat of the Upper Hudson.
Habitat  replacement/backfilling  measures  will  be
implemented to mitigate these  impacts.  A  monitoring
program will be established to verify the attainment of the
habitat replacement objectives.  Although the degree of
impact will  be  directly related to the area remediated and
volume dredged, these differences among the alternatives
are not considered to be significant due to their temporary
nature and the mitigation measures which will be utilized.

For the CAP-3/10/Select alternative, there is the additional
potential transient impact from the temporary exposure of
deeper, potentially  highly PCB-contaminated  sediments
during the interval between excavation and cap placement.
This impact will be minimized by placement of the cap as
soon as practicable after the  removal operations are
complete, assumed to be no more than 30 days. Therefore,
this  impact is not considered  a significant difference
between CAP-3/10/Select, REM-3/10/Selectand REM-0/0/3
alternatives.

The magnitude of the short-term impacts discussed above
varies with the overall scope of the alternative, in terms of
volume of material excavated and area remediated.  The
implementation times for the active alternatives are 5 years
for CAP-3/10/Select and REM-3/10/Select, and 7 years for
REM-0/0/3.
                                              Page 23

-------
 Superfund Proposed Plan
                   Hudson River PCBs Reassessment
 6. Implementabilitv

 Technical Feasibility

 Both the No Action and MNA alternatives are technically
 feasible as no active measures are being taken for the
 PCB-contaminated sediments. The implementability of the
 source control measures will be evaluated as part of the
 EE/CA for the NTCRA near the GE Hudson Falls plant.

Technical feasibility for the active remediation alternatives,
which are all technically feasible, is discussed below in
terms of the main components of the alternatives.

Transfer facilities - It is expected that transfer facilities will
be established at two locations along the river to process
the  sediments  generated   by  removal  operations.
Development of two transfer facilities, a northern facility
adjacent to River Section 1 (Thompson Island Pod) and a
southern facility near Albany, both with river frontage and
rail access,  is considered technically  feasible. These
transfer facilities will be temporary in that they are expected
to be removed after completion of  the active remedial
activities.

Removal - The bulk of removal work  under  the CAP-
3/10/Select  and  REM-3/10/SeJect  and   REM-0/0/3
alternatives will be performed by mechanical or hydraulic
dredges. In difficult-to-access areas, there may be a need
to modify the selected equipment or, alternatively, employ
different equipment.

Capping  - An  evaluation  of the AquaBlok"" system is
currently in progress at several remedial sites (e.g., Ottawa
River. Ontario; Fort Richardson, AK). The implementability
and long-term  performance of the system have not yet
been  established, but it is expected that considerable
performance data will become available in the near future.
However,  the  principal  component of this system is
bentonfte, which is considered  a very  stable, low-
permeability  barrier.   Bentonite  has been  used  in
multimedia and day capping systems for many years and
has  demonstrated  effectiveness   for the  long-term
encapsulation of contaminants.

In-river  and  rail   transportation  -  Development  of
transportation systems to implement the active remedial
alternatives is  considered  feasible.  While the volume of
material that must be handled is large, it is well within the
capabilities of the waterbome and rail systems to handle
the stabilized dredged sediments.

REM-3/10/Select is more implementable from a technical
feasibility perspective than REM-0/0/3 due to the smaller
volume of material to be dredged and handled, as well as
the accessibility of the areas to be dredged.  Both removal
alternatives are more technically implementable than the
CAP  3/10/Select alternative due to the combination of
capping and dredging issues associated with the capping
alternative.

Administrative Feasibility

Local Landfill - Establishment of a permanent local landfill
was eliminated in consideration of the opposition from local
citizens, which would likely make the siting of a local landfill
administratively infeasible.

Both  No Action and MNA require  no active measures,
therefore  they are the most implementable  from an
administrative  feasibility perspective.   Active  remedial
measures are slightly more difficult to implement from an
administrative  feasibility  perspective  due to  access
agreements and making necessary arrangements to utilize
the river with minimal interruption of boat traffic.

For the active remediation alternatives (CAP-3/10/Select,
REM-3/10/Select, and REM-0/0/3). it is expected that the
two transfer facilities, constructed on land adjacent to the
river will be considered "on-site" for the purposes  of the
permit exemption under CERCLA Section 121(e), although
any  such  facilities will  comply  with  the substantive
requirements  of   any  otherwise  necessary   permits.
Operations under  these  alternatives will have  to be
performed in conformance with substantive requirements of
regulatory programs implemented by the U.S. Army Corps
of Engineers under Section 10 of the Rivers and Harbors
Act and Sections 401 and 404 of the Clean Water Act. In
addition, discharges during remediation will conform to NYS
regulations related to maintenance of Hudson River water
quality. Habitat replacement/backfilling will be implemented
in accordance with federal  and State requirements.  In
addition, it is expected that contract documents for  any of
the active remediation alternatives will contain substantial
restrictions on construction activity induding controls on the
types of dredging  and capping equipment  to be used,
restrictions on the speed of operations, constraints on barge
filling practices, and controls on temporary storage of
contaminated dredge spoils. Construction activities will also
be coordinated with the Canal Corporation, which operates
the Locks on the Upper Hudson River from May through
November. Finally, requirements of any other regulatory
programs will be incorporated as necessary on the basis of
design information developed during subsequent phases of
the project.

Availability of Services and Materials

For the No Action and MNA alternatives, all needed services
and materials are available. For the CAP-3/10/Select, REM-
3/10/Select, and REM-0/0/3 alternatives, the principal deficit
in services and materials relates to barges and towboats.
Because commercial operations on the Champlain Canal
system have largely ceased, this equipment may no longer
EPA Region 2
                                             Page 24

-------
Superfund Proposed Plan
                   Hudson River PCBs Reassessment
be available in the project vicinity. However, it is expected
that the contractors will obtain the needed equipment for a
project of the scale envisioned under these alternatives.
The discussions of the alternatives below do not include
any costs for source controls measures that will be taken
at the GE Hudson Falls plant as part of the NTCRA. See
table below of comparison of alternatives.

Net Present Worth.

The net present worth (year 2000 dollars) of the remedial
alternatives  ranges  from  $140,000 for No  Action  to
$570,000,000 for REM-0/0/3.  The net present worth of
REM/3/10/Select is $460,000,000, which is $110,000,000
less than REM-0/0/3. For the active remedial alternatives
(CAP-3/10/Select, REM-3/10/Select and REM-0/0/3), these
costs are based  on the use of mechanical  dredging
techniques to remove PCB-contaminated sediments from
the Upper Hudson River, and assume the disposal of all
dredged  materials at  licensed  TSCA  and  non-TSCA
landfills located outside of the Hudson River Valley. For
the option where  the non-TSCA  material is utilized for
beneficial  uses,  the  net present worth of the active
remedial alternatives ranges from $338,000.000 for CAP-
3/10/Select  to $496,000.000  for REM-0/0/3.   These
beneficial use option costs are also based on the use of
mechanical dredging techniques.  There is no significant
difference in the net present worth costs for the option
where hydraulic dredging techniques are utilized to remove
PCB-contaminated sediments.

Capital Cost

The No Action alternative has no capital cost.  The MNA
alternative has a present worth capital cost of $417,000 for
further refining the mathematical model for the Upper
Hudson River. The present worth of the capital costs for
the active remedial alternatives ranges from $344.000,000
for CAP-3/10/Select to $556,000,000 for REM-0/0/3. The
net present worth of the capital costs for REM-3/10/Select
is  $448,000,000. some $108,000,000  less than the net
present worth of the capital costs for REM-0/0/3. For these
active remediation alternatives, the present worth of the
capital costs includes the disposal of the stabilized dredged
materials at licensed  TSCA and non-TSCA landfills, and
assumes the use of  mechanical dredging techniques to
remove PCB-contaminated sediments from the river.

For the option where the non-TSCA material is utilized for
beneficial uses, the present worth of the capital costs for the
active remedial alternatives ranges from $314,000.000 for
CAP-3/10/Select to $483.000.000 for REM-0/0/3. The net
present worth of the capital costs of REM-3/10/Select under
the beneficial use option is $399,000,000. These beneficial
use option costs are also  based on the use of mechanical
dredging techniques.  There is no significant difference in
the present worth of capital costs for the option where
hydraulic dredging techniques are utilized to remove PCB-
contaminated sediments.

O & M Cost

Due to the varying frequency of different elements of the
monitoring program, and the five-year reviews required by
the NCP, O&M  costs will vary  on  an annual  basis.  The
present worth of the O&M costs for the No Action alternative
is $140,000 and for the MNA alternative is $38,000,000.
The  net present worth  of  the  O&M costs for CAP-
3/10/Select  is  $24,000,000,  for REM-3/10/Select  is
$13,000,000 and for REM-0/0/3 is $13,000,000. For the
active remediation alternatives, this present worth of the
O&M  costs assumes the use of mechanical  dredging
techniques to remove PCB-contaminated sediments from
the Upper  Hudson River, and  disposal of the  stabilized
dredged materials at remote TSCA and non-TSCA landfills.

For the option in which the non-TSCA material is utilized for
beneficial uses,  the present worth of the O&M costs for the
active remedial  alternatives ranges from $13,000,000 for
REM-0/0/3 to $24,000,000 for CAP-3/10/Select.  These
beneficial  use  costs  are also based on   the  use of
Alternative
No Action
Monitored Natural
Attenuation
CAP3/10/Select
REM 3/10/Select
REM 0/0/3
Area
Remediated
(Acres)
-
-
493
493
964
Area
Capped
(Acres)
-
-
207
-
-
Volume
Removed
(CY)
-
-
1,732,800
2,651,700
3,823,100
Estimated
PCB Mass
Remediated
(kg)
-
-
45,600
45,600
>63,500
Estimated
PCB Mass
Removed
(kg)
-
-
33,100
45,600
>63,500
Cost
(SMillion -
present
worth)
0.14
39
370
460
570
 EPA Region 2
                                             Page 25

-------
  Superfund Proposed Plan
            Hudson River PCBs Reassessment
  mechanical dredging techniques.  There is no significant
  difference in the present worth of the O&M costs for the
  option in which hydraulic dredging techniques are utilized
  to remove PCB-contaminated sediments.

 8. State Acceptance

 The NYSDEC has not yet submitted its determination on
 the preferred alternative but has indicated that it is in favor
 of an active remedial alternative for the Hudson River.

 9. Community Acceptance

 While there has been significant controversy concerning
 dredging, community acceptance of the preferred remedy
 will be assessed in the  Record of Decision  following
 review of the public comments received during the public
 comment  period  on  the  Reassessment  Remedial
 Investigation/Feasibility Study reports and this Proposed
 Plan. It should be noted that the use of a local landfill to
 dispose of dredged material was screened out based on
 community opposition.    The  Upper  Hudson  River
 Communities, as well as environmental groups, previously
 expressed opposition to such a landfill.

 PREFERRED REMEDY

 The  preferred  alternative  is the  removal  (targeted
 dredging) Alternative REM-3/10/Setect in conjunction with
 source control  at  the  GE  Hudson Falls  plant  to  be
 accomplished via a separate Non-Time Critical Removal
Action.    This  alternative  includes the  dredging  of
approximately  2.65 million  cubic yards  of  PCB-
contaminated sediment from the Upper Hudson  River.
The associated present worth costs are  approximately
$460 million.  The REM-3/10/Select alternative includes
the following components:

•      Removal of all sediments based primarily on an
       MPA of 3 g/m2 PCBs or greater (approximately
       1.56 million cubic yards of sediments) from River
       Section 1;

•      Removal of all sediments based primarily on an
       MPA of 10 g/m2 PCBs or greater (approximately
       0.58 million cubic yards of sediments) from River
       Section 2;

•      Removal   of   select sediments  with   high
       concentrations of PCBs (NYSDEC hot spots 36,
       37, and the southern portion of 39)(approximately
       0.51 million cubic yards) from River Section 3;

•      Dredging of the navigation channel as necessary
       to implement the remedy and to avoid hindering
       existing canal  traffic during implementation.
       Approximately 341,000 cubic yards of sediments
       will be removed  from the navigation  channel
       (included in  values  in the first  three bullets,
 above), but some portion of this is associated with
 the areas targeted for remediation;

 Within the areas targeted for remediation, the goal
 is  to  remove  all  of  the  PCB-contaminated
 sediments within these areas, leaving a residual of
 approximately 1 mg/kg;

 Monitored    Natural   Attenuation   of   PCB
 contamination that remains  in the  dredging
 residual and in unremediated areas;

 Use of environmental dredging techniques that will
 minimize and control resuspension of sediments
 during dredging;  transport of dredged sediments
 via barge or pipeline to treatment/transfer facilities
 for dewatering and stabilization;

 Backfill of dredged areas with approximately up to
 one foot of clean material to isolate residual PCB
 contamination  and   replace  habitat,   where
 appropriate;

 Rail transportation of dewatered,  stabilized
 sediments to  the  appropriate licensed off-site
 landfill(s). If a beneficial use of some portion of the
 dredged material is arranged, then an appropriate
 transportation  method will be determined (rail,
 truck, or barge);

 Monitoring  of fish, water,  and  sediment to
 determine when  Preliminary Remediation Goals
 are reached and implementation (or modification)
 of appropriate institutional controls such as fish
 consumption advisories and fishing restrictions,
 until goals are met; and,

 In order to preserve flexibility during the remedial
 design, the preferred alternative does not specify
the type of  dredge.    The  Feasibility Study
 considered   both  mechanical  and  hydraulic
environmental  dredges   with  respect  to
 implementability and cost   The design for the
 project will plan for a construction period of five
years. It is anticipated that it will take three years
for  remedial  design and mobilization,   so  that
dredging will begin in 2004.

 In addition,  during remedial  design USEPA will
consider whether there are  any new treatment
options for the  dredged sediment that would
improve the cost effectiveness of the remedy.

Because contamination will remain on-site above
health-based levels, five-year reviews are required.
EPA Region 2
                                     Page 26

-------
Superfund Proposed Plan
                   Hudson River PCBs Reassessment
RATIONALE  FOR  SELECTION  OF  PREFERRED
ALTERNATIVE

The selection of the preferred alternative is accomplished
through the evaluation of the nine criteria as specified in
the NCR. USEPA has evaluated the alternatives against
the first seven criteria. State and community acceptance
will be evaluated after the release of the Proposed Plan
and  associated   public comment  period,  although
opposition to a local landfill has already been factored into
the preferred  alternative.   The  NYSDEC has not yet
submitted its determination on the preferred alternative
but has indicated that it is in favor of an active remedial
alternative for the Hudson River.

The preferred alternative is protective of human health
and the environment. Risk is reduced through removal of
PCB-contaminated sediment,  followed by  Monitored
Natural  Attenuation.    The  preferred alternative  is
protective of human health and the environment because
it will significantly reduce the risks from consumption of
fish in River Sections 1,2, and 3, as well as in the Lower
Hudson.    The  modeling  projects  that  the   target
concentration  of 0.4 mg/kg, which is  protective of the
average adult who consumes one fish meal every two
months, is attained in River Sections 1 and 2 within 20
years of active remediation.  The target of 0.2 mg/kg,
protective of an adult who consumes one fish meal per
month,  is attained in River Section 2 within 32 years of
active remediation. These time periods are significantly
shorter than the time periods projected for attaining the
0.4 mg/kg and 0.2 mg/kg targets under either the No
Action alternative or the MNA alternative.

According to USEPA's model projections for the Upper
Hudson River, the preferred alternative will meet the
Preliminary Remediation Goal for human consumption of
fish, which is 0.05 mg/kg,  in River Section 3 within 43
years after completion of the active remediation.  As a
result, the Preliminary Remediation Goal of 0.05 mg/kg
also is expected to be attained in the majority of the Lower
Hudson River within this time  frame, due  to the  lower
initial concentration of PCBs in  the  Lower Hudson
compared to the Upper Hudson.  Due to the continuing
PCB toad of 2 ng/L assumed after implementation of the
source control action in the vicinity of the GE Hudson Falls
plant, the PCB concentration in fish averaged over the
Upper Hudson is expected  to be reduced to a range of
0.09 to 0.14 mg/kg, within the 70-year modeled time
period,   which  is  slightly  above   the   Preliminary
Remediation  Goal of  0.05  mg/kg.    However, the
protectiveness of the preferred  alternative is further
enhanced through implementation of institutional controls,
such as the fish  consumption advisories and fishing
restrictions.

The  preferred alternative  is  also protective of the
environment, because the preferred alternative will reduce
PCB concentrations in fish averaged overthe entire Upper
Hudson, and in the Lower Hudson, to levels that are at or
within the  range of 0.3  to 0.03 mg/kg  in whole  fish
(equivalent to 0.12 to 0.012 mg/kg in fish fillet) within the
70-year modeled time period, which is the Preliminary
Remediation Goal for ecological exposure.  Thus, the
preferred alternative is protective of the birds, fish  and
mammals that live in and near the Hudson River.


The preferred alternative, REM-3/10/Select, is more cost-
effective than the REM-0/0/3 alternative.  The preferred
alternative is $110 million less expensive than REM-0/0/3,
without substantial differences in the amount of ecological
or human health risk reduction. This is supported by the
following tables, which show the projected ecological and
human health risks for each of the alternatives.

The Reasonable Maximum Exposure cancer risks and non-
cancer health hazards for adult anglers for each alternative
averaged over the entire Upper Hudson River are shown
in the top table on the following page.

Ecological risks for the mink and river otter for each of the
three river sections are shown in the bottom table on the
following page.

Moreover, as noted previously, USEPA's comparison of the
relative  effectiveness of the alternatives is based on the
results of modeling each remedial alternative as well as data
projections.  Comparisons  of the model outputs to recent
data trends suggest that the model may be overly optimistic
with regard to the rate of PCB decline in fish predicted for
the No Action (no source control) and Monitored Natural
Attenuation  (assuming   source  control)  alternatives.
Consequently, the relative benefits of remediation are likely
to be greater than suggested by the models.

The preferred  alternative  will comply with the  location-
specific and action-specific ARARs identified, as well as two
of the five chemical-specific ARARs for the site. However,
although the preferred alternative will approach  these
numbers, three of the chemical-specific ARARs are not
expected to be met  because the PCB contamination
entering the Upper Hudson  River from above Rogers Island
(even after source control at the GE Hudson Falls plant) will
likely exceed those ARARs. Therefore, it is expected that
technical impracticability ARAR waivers will be required for
three chemical-specific ARARs (1  ng/L federal  Ambient
Water Quality Criterion; 0.12 ng/L New York State standard
for  protection of wildlife;  and 0.001 ng/L New York State
standard for protection of human consumers of fish). Even
the most aggressive removal alternative, REM-0/0/3, would
require these same waivers.

The preferred alternative is effective in the long term and
in  the  short  term.   Consideration  of  the statutory
requirement for  permanent remedies, to  the maximum
extent practicable, favors the removal alternatives (which
are permanent) over the capping alternative (which also
has significant long-term maintenance concerns).
EPA Region 2
                                             Page 27

-------
 Superfund Proposed Plan
Hudson River PCBs Reassessment
Cancer Risks and Non-Cancer Health Hazards for Adults from Fish Ingestion
Averaged over the Entire Upper Hudson River
Risk or Hazard
HI-RME (2009-2015)
HI-RME (2011-2017)
HI-CT (2009-2020)
HI-CT (2011-2022)
Cancer risk -RME
(2009-2048)
Cancer risk - RME
(2011-2050)
Cancer risk - CT
(2009-2020)
Cancer risk - CT
(2011-2022)
No Action
53-80
48-75
5.0-7.7
4.5-7.3
7.8x10* to 1.4x10°
7.3x10* to 1.3x10°
1.7x10*102.6x10*
1.5x10* to 2.5X1O*
MNA
40-71
34-66
3.4-6.7
2.9-6.3
4.0 x 1O* to
1.2x10°
3.5x10*10
1.1x10°
1.2x10* to
2.3 x10*
1.0x10-*to
2.1 10"*
CAP-3/10/Select
15

1.3

1.8x10*

4.5x10*

REM-3/10/Select
13

1.2

1.7x10*

4.0x10*

REM-0/0/3

8

0.7

1.2 x 1O*

2.4x10*
Eeotogteai Toxteity Quotient* -River Otter and iHnjt
(Average of 25-Year Time Frame)

No Action
start year
2008
No Action
start year
2009
MNA
start year
2008
MNA
start year
2009
CAP-
3/10/Select
REM-
3/10/Select
REM-0/0/3
River Section 1 (RM 189) Modeling Time frame is 2008-2032 for CAP-3/10/Setect and REM-3/10/Select and 2009-2033 for REM-0/0/3
Mink
River Otter
LOAEL
NOAEL
LOAEL
NOAEL
4.6-5.3
46-53
24-30
240-300
4.5-5.2
45-52
23-29
230-290
1.7-2.6
17-26
9.7-15
97-150
1.6-2.5
16-25
9.1-14
91-140
0.94
9.4
5.3
53
0.95
9.5
5.2
52
0.70
7.0
3.7
37
River Section 2 (RM 184) Modeling Time frame is 2009-2033 for CAP-3/10/Setect and REM-3/10/Setect and 201 1-2035 for REM-0/0/3
Mink
River Otter
LOAEL
NOAEL
LOAEL
NOAEL
1.5-2.7
15-27
14-27
140-270
1.3-2.6
13-26
12-26
120-260
0.94-2.5
9.4-25
9.2-24
92-240
0.79-2.4
7.9-24
7.8-23
78-230
0.36
3.6
3.5
35
' 0.31
3.1
2.9
29
0.19
1.9
1.8
18
River Section 3 (RM 154) Modeling Time frame is 2010-2034 for CAP-3/10/Setect and REM-3/10/Select and 2012-2036 for REM-0/0/3
Mink
River Otter
LOAEL
NOAEL
LOAEL
NOAEL
0.21
2.1
2.4
24
0.20
2.0
2.3
23
0.11
1.1
1.2
12
0.09
0.9
1.1
11
0.07
0.75
0.87
8.7
0.08
0.75
0.86
8.6
0.06
0.55
0.62
6.2
Notes: Toxicity Quotient above 1.0 indicate the potential for ecological risk.
NOAELs and LOAEL are discussed in section on the Ecological Risk Assessment.
Range of years calculated using bounding estimates are presented for the No Action and MNA alternatives.
EPA Region 2
                      Page 28

-------
Superfund Proposed Plan
                  Hudson River PCBs Reassessment
Implementation of the preferred alternative will greatly
reduce the amounts of PCBs in the sediments and lower
the average PCB concentration in surface sediments,
which in turn will reduce PCB levels in the water column
and fish, and the corresponding levels of risk.  Reduced
amounts of PCBs and surface sediment concentrations
will also reduce the long-term transport of PCBs from
each river section to the next and from the Upper Hudson
River to the Lower Hudson River.  For example, there is
at least a  40 percent reduction of the PCB load that is
transported into the Lower Hudson River in the 10 years
following remediation as compared to MNA alone (with
upstream source control).

The preferred alternative results in the targeted removal
of 2.65 million cubic yards of contaminated sediments
containing approximately 100,600 Ibs (45,000 kg) of PCBs
from the river environment. This results in a long-term
reduction in the toxicity, mobility, and volume of PCBs in
the river, even though treatment is not a principal element
of the remedy. However, after the sediments are dredged
from the river,  these sediments will be stabilized by
blending them with Portland cement, which is a form of
limited treatment.
natural attenuation of PCBs in sediment over the past five
years.  Geochemical analyses tell us that, while the river
is netdepositional, contaminated sediments are not being
significantly buried  or sequestered in  the river.  Water-
column sampling at both ends,  and throughout the
Thompson Island Pool, shows a sharp increase (over
three times)  in PCB water-column concentration as the
river flows through this 6.3 mile stretch.  PCB congener
fingerprinting demonstrates  that the  sediment  is the
source. Projections from the models developed for the
river, which have undergone external peer review, show
that fish within the Upper Hudson will be substantially less
contaminated sooner through active remediation.  As a
result,  human and ecological risks will be substantially
reduced below current unacceptable levels to a degree
that will be real and measurable.
The  preferred  alternative   is   technically   and
administratively feasible and implementable. All of the
necessary personnel, equipment, and services required
are expected  to  be readily available  or reasonably
arranged.

The preferred alternative, REM-3/10/Select, is similar to
the REM-0/0/3 alternative in terms of reduction of risks to
human health and the environment

In summary, the REM-3/10/Select alternative was chosen
as USEPA's preferred alternative based on the need for
active remediation in order to protect  human health and
the environment. The REM-3/10/Select alternative fulfills
the statutory requirement for permanent remedies, to the
maximum extent practicable, whereas capping does not,
and the REM-3/10/Select alternative is more cost-effective
than the REM-0/0/3 alternative.

The need to remediate a substantial  portion of  the
contamination in the river sediments  is not based on a
single analytical tool.  Instead, it is drawn from a weight of
evidence approach in which several analytical tools and
factual databases all point to the same conclusion, each
in its own way strengthening the others.

Historical information collected over the past 20 years
indicates that large quantities of PCBs are present within
the sediments in relatively high concentrations. Within the
Upper Hudson River, fish tissue sampling indicates that
PCB levels in brown bullhead in the Thompson Island
Pool do not show any significant reduction attributable to

EPA Region 2
                                           Page 29

-------
 Superfund Proposed Plan
                                                                         Hudson River PCBs Reassessment
 GLOSSARY OF TERMS
 Specialized terms used in this Proposed Plan are
 defined below:

 Applicable or relevant and appropriate requirements
 (ARARs) - the Federal and State environmental laws that
 a selected remedy will meet. These requirements may
 vary among sites and alternatives.

 Aroclor - a term used in commerce to denote a PCB
 type; typically followed by a 4-digit number, the last 2
 designating percent chlorine weight.

 Central Tendency (CT) - the average exposure expected
 to occur at a site.

 Congener - one of the 209 different configurations of a
 PCB molecule resulting from multiple combinations of
 hydrogen and chlorine positions on the PCB molecule.

 Consent Decree - a legal document, approved by a judge,
 that formalizes an agreement between USEPA and one or
 more potentially responsible parties (PRPs) outlining the
 terms by which the response action will take place.  A
 Consent Decree is subject to a public comment period prior
 to its approval by a judge and is enforceable as a final
judgment by a court.

 Ex situ - the removal of a medium (for example, water or
 soil) from its original place, as through excavation, in order
 to perform the remedial action.

 Lipids - a biochemical term for fat; most commonly used
 with respect to the percent contained in fish.

 LOAEL - Lowest observed adverse effect level. The lowest
 concentration at which an adverse effect is seen in field or
 laboratory toxicity studies.

 MPA  (mass per unit area) - a representation of the
 total mass of PCBs within a sediment core or within an
 area represented by sediment cores.
NOAEL - No observed adverse effect level. The highest
concentration  in a field or laboratory toxicity study at
which no adverse effect is seen.

PCB Inventory - the total mass of PCBs contained within
a area of sediment; relates to the mass per unit area
(MPA) value.

Present Worth Analysis - a method of evaluation of
expenditures that occur over different time periods.  By
discounting all costs to a common base year, the costs for
different remedial action alternatives can be compared on
the basis of a single figure for each alternative.  When
calculating present worth costs for Superfund sites, capital
and operation  and maintenance costs are included.

Safe Drinking Water Act Maximum Contaminant
Level (SDWA MCL) • the maximum permissible level of
a contaminant in water that is delivered to any user of a
public water system.

Reasonable Maximum Exposure (RME) - the highest
exposure that  is reasonably expected to occur at a site.

Reassessment - The evaluation conducted by USEPA to
reconsider the interim No Action alternative selected in
USEPA's 1984 Record of Decision. The Reassessment
is also referred  to  as the Reassessment Remedial
Investigation and Feasibility Study (Reassessment RI/FS).

TEQ - a subset of PCB congeners that are structurally
similar to dibenzo-p-dioxin and cause dioxin-specific
biochemical and toxic responses.

7n+ PCBs • a representation of the sum of PCBs with
3 to 10 chlorine atoms per molecule.

TSCA - the Toxics Substances Control Act.  This law
regulated the handling and disposal of PCBs.
                                                                                   printed on recycled paper
EPA Region 2
                                          Page 30

-------