Regional Response to the
National Remedy Review Board Comments
on the Site Information Package for
the General Electric (GE)-Pittsfield/
Housatonic River Project, Rest of River
DCN HR-080212-AARX
SDMS 518898
August 3, 2012
Prepared by
^DST«\
v,
iSEh
PRO^
U.S. Environmental
Protection Agency
New England Region
Boston, Massachusetts
11P-0432-3
-------
TABLE OF CONTENTS
REGIONAL RESPONSE TO THE NATIONAL REMEDY REVIEW BOARD
COMMENTS ON THE SITE INFORMATION PACKAGE FOR THE GENERAL
ELECTRIC (GE)-PITTSFIELD/HOUSATONIC RIVER PROJECT, REST OF RIVER
REGIONAL RESPONSE TO NATIONAL REMEDY REVIEW BOARD
RECOMMENDATIONS FOR THE GE-PITTSFIELD/HOUSATONIC RIVER,
REST OF RIVER
APPENDIX A PRELIMINARY DRAFT OF DRAFT OUTLINE OF
POTENTIAL PERFORMANCE STANDARDS FOR
ALTERNATIVE SED 9/FP 4 MOD
Attachment A-l Massachusetts Division of Fisheries and Wildlife Core Habitat Area
Maps
APPENDIX B REVISED COMPARATIVE ANALYSIS OF ALTERNATIVES
Attachment B-l Use of Channel Realignment Along the Housatonic River for
Restoration and Remediation of PCB Contamination
Attachment B-2 Channel Dynamics and Ecological Conditions in the Housatonic
River Primary Study Area
Attachment B-3 Activated Carbon Summary
Attachment B-4 Massachusetts Division of Fisheries and Wildlife Core Habitat
Area Maps
Attachment B-5 Cap Cross Section Refinement - Layer Sizing, Rest of River -
Reach 5 A
Attachment B-6 Comparison Metrics
Attachment B-7 Post-East Branch Remediation Boundary Conditions
Attachment B-8 Food Chain Model Output
Attachment B-9 Preliminary Draft ARAR Tables for SED9/FP 4 MOD
Attachment B-10 Cost Assumptions Memorandum for SED9/FP 4 MOD
APPENDIX C GENERAL ATTACHMENTS
Attachment C-l National Remedy Review Board Recommendations for the
Housatonic River, Rest of River Site
Attachment C-2 Assessment of Recent Tissue PCB Chemistry and Lipid Data
Attachment C-3 Screening Evaluation for Potential Improvement in Sediment
Trapping in Woods Pond
Attachment C-4 Housatonic River Status Report: Potential Remediation
Approaches to the GE-Pittsfield/Housatonic River Site "Rest of
River" PCB Contamination
Attachment C-5 Bank Erosion/Restoration, Housatonic River, Massachusetts
L:\20502169.095\NRRB_RESPONSBMASTERTOC_REV.DOCX
111
8/1/2012
-------
#1«» w4%.
.* _ * UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
5 POST OFFICE SQUARE, SUITE 100
BOSTON, MA 02109-3912
MEMORANDUM
DATE: August 3,2012
SUBJECT: Regional Response to National Remedy Review Board Recommendations
for the GE-Pittsfield/1 lousatonic River. Rest of River
)5§&———
FROM: lames T vaosJli-D
Office aroite Remediation and Restoration
U.S. eTm New England Region 1
TO; Amy R. Legare. Chair
National Remedy Review Board
Stephen J. Ells. Chair
Contaminated Sediments Technical Advisory Group
The National Remedy Review Board and the Contaminated Sediments Technical
Advisory Group (the Boards) completed their review of the proposed cleanup acli in for
the I lousatonic River, Rest cf River site, in Pittsfield. MA., as documented in its
memorandum of October 20. 201 1. The Region appreciates the Boards' input anc
recommendations. Subsequent to die Boards" review, the Region has made signifcani
progress in addressing many of the issues raised by the Boards, coordinating w ith our
state partners in Massachusetts and Connecticut, and moving toward a potential remedy
for the Rest of River. We hf.vc summarized these efforts in a series of technical
documents that arc being released to the public in advance of a formal remedy proposal.
I he Region has incorporated the Boards" recommendations, as appropriate, into these
technical documents, which serve to supplement the Site Information Package submitted
to the Boards in June 201 1: a draft Outline of Potential Performance Standards foi
Alternative SET) 9/FP 4 MOD (included as Appendix A), a Revised Comparative
Analysis of Alternatives (included as Appendix B), and General Attachments (included
as Appendix C). or the Region has otherwise addressed the Boards' rccommenda ions as
described below. The Boarcs" recommendations are listed below in italics follow «d by
the Region's response (see Attachment C-i of Appendix C for the complete text of the
Boards" comments and recommendations).
Recommendation No. t - Site Characterization
In the package presented to '.he Boards, modeling results played an important role in
evaluating MNR as a remedial option. The Boards recommend that additional adult
1
-------
largemouth bass fish tissue data he collected and analyzed in the context of historical
data and model output. If (he apparent discrepancy between the 20QH data (mean of
about 5 ppm PCB in filled and model output tabout I#ppm) remains, the modeling
should jc updated to provide risk projections that more appropriately reflect current
conditions. In addition, the updated sampling results may he used to evaluate the
effectiveness and benefits of the upstream remediation.
Related to the above recommendation, but from broader perspective, the Boards
recommend that the Region expand the adult fish tissue collection efforts to provide an
adequate baseline database far evaluating the effectiveness of completed, ongoing and
planned remedial actions.
Based cm the model predictions described in Appendix F of the package, the Region
concluded that Woods Pond, even if modified by deepening and changing the flow
direction of the input channel could not be an effective sediment trap. Based upon a brief
analysis of the empirical data far the site, however, it appears to the Boards that the
model predictions far trapping efficiency may not be consistent with some of the
historical sedimentation data far the site. The Boards believe that a modified Woods
Pond, acting as a sediment trap, could reduce the amount of PCBs released over the dam
in addition to the reductions that would result from other proposed active remedial
measures. Therefore, the Boards recommend that the Region further evaluate the
potential incremental improvement in sediment trapping of a modified Wood Ponds and
recommends that the Region ask engineers from the US Army Corps of Engineers to
assist in this evaluation.
Region's Response
The Region docs not believe that there is a true discrepancy between the measured fish
tissue concentrations and the modeled concentrations, as the epparent discrepancy lies in
the abnormally low lipid content of the fish that were analy?cd. This issue is explored
and ciscussed more fully in Attachment C-2 of Appendix C.
Based on the Boards input, the Region requested that Of-, conduct additional adult fish
tissue sampling in September 2011. GE submitted the data to the Region in January
2012. 'J"he concentrations are largely the same as measured ir 2008. The Region has
reviewed those data, and Gh's analysis and those reports are available on the website at
http://www.epa.gov/regionl /ge/thesile/restofriver''rcports/497987.pdf.
As a fol ow-up lo the Boards' recommendation, the Region worked with the U.S. Annv
Corps o Engineers and the Commonwealth of Massachusetts to further evaluate the
potential for incremental improvement in sediment trapping in Woods Pond, The Region
agrees t lat additional deepening or 'Other measures could enhance the Woods Pond
component of a cleanup plan. The results of this evaluation are included in Attachment
C-3 el Appendix C. 1'his information was used to inform the Revised Comparative
Analysis of Alternatives (Appendix B).
Recomr.lendation No. 2 - Human I lealth/hcoloaica 1 Risk
-------
Ow ing the presentation, the Region slated it is conducting a risk-based PCB clear-up as
described in 40 Code of Federal Regulations (CFRj '61.61(c). The Boards recommend
that, since, for example, the region plans to leave soih with PCB contamination u<
excess of 50 parts per million (ppmi. die Superjundprogram closely coordinate with the
Region's Toxic Substances Control Act program to ensure the remedy meets the
requirements of 40 CFR "61.61(c).
From the presentations by the Commonwealth and the Region to the Board, it appears
that there is a fundamental disagreement concerning the interpretation and applic.ition
of some of the criteria for remedy selection. Particularly noteworthy are the differences
in perspective on the balancing of short-term and potential long-term environmental
impacts from remedy implementation and the reduction of long-term risks predicted to be
achieved by a protective remedy. The presentation by the Commonwealth indicated that it
sees the impacts to Commonwealth-listed species resulting from the need to control
stream meandering as a long-term impact whereas the Region contends that habitat
restoration and other impact reduction measures will he effective in meeting the
requirements of the Commonwealth's endangered species law and therefore any impacts
will be only short-term The Commonwealth's presentation also indicated that it believes
the long-term ecological risks (e.g. adverse effects to mink and wood duck) were
acceptable when balanced against the impacts of remediation on habitat loss.
Alternately. EPA sees these long-term ecological risks as requiring remediation rc meet
the threshold criteria for selecting a remedy that is protective. The Boards recommend
that the Region consolidate the discussion on the documented ecological impacts ar the
site and compare them to the Agency s requirements under CERCLA and the RCR.4
Permit to select a remedy protective of all identified receptors (assessment eridpointsi.
This consolidated presentation wilt allow for a direct comparison of short term and
long-term risks and impacts and how these risks are balanced, justified and consistent
with remedy selection criteria in any decision documents.
The Boards note thai CF.RCi.A and the RCRA Permit identify protectiveness of human
health and the environment as a threshold criterion that all remedies must achieve.
Furthermore, the NCP states that the use of institutional controls should supplement (not
substitute for) active response measures (e.g., fCs should not substitute for active
response measures as the sole remedy unless such active measures are determined not to
be practicable). The remedy supported by the Commonwealth appears to rely solely on
institutional controls (ICs) to protect hitman health through consumption offish by
restricting all consumption, whereas the remedy preferred by the Region would achieve a
measure of risk reduction that results in risks from fish consumption within the
acceptable risk range and at a hazard quotient of I under a central tendency exposure
scenario in virtually all reaches. The Board recommends that the Region emphas ze in
the decision document
-------
The Region is coordinating internally with the loxic Substances Control Act (TSC'A)
program to ensure the remedy meets the requirements of 40 CFR 761.61(c).
Since l> e Board meeting, (he Region and representatives from 1 !Q and the states of
Massac lusctts and Connecticut have been working cooperatively for the last several
months to discuss poicnlial approaches to clean up the Rest of River. These discussions
focused, in part, oti the need to address the risks from polvchlorinated biphcnyls (PCBs)
to hi miiis. fish, wildlife, and other organisms while avoiding, mitigating, or minimizing
the impacts of the cleanup on the unique ecological character of the Housatonic River, in
Maj' 2012. KPA released a fact sheet summarizing many of the discussions among HP A
and the states. This fact sheet, entitled "Potential Remediation Approaches to the CiF.-
Pitts icld/l lousatonic River Site 'Rest of River' PCB Contamination*' has been attached
in Attachment C-4 of Appendix C for your information. The Revised Comparative
Analysis of Alternatives (Appendix B) reflects the current thinking from these
disci.ss;ons.
In these discussions, it was agreed that the protection of human health, including 1 he
consumption of lish. was a high priori!). The draft Outline o 'Potential Performance
Standards (Appendix A) reflects that thinking.
Recommendation No. 3 - Principal I hrcat Waste
/ he package presented to the Boards included a discussion of principal threat waste
(PFly'). White the discussion addressed contaminant mobility, if did not specifically
address toxicity and why the high concentrations of PCBs (some locations at greater than
800 ppm) in floodplain soils would not be considered PTW materials subject to
Comprthensive Environmental Response, i Compensation and Liability Act's iCERCl.A \st
and the NCP's preference for treatment to the maximum extent practicable. Consistent
with A Guide to Principal Threat and Low Level Threat Wastes (OSWER Directive No.
9380J-06FS) which addresses the preference for treatment of highly toxic materials, and
in light of A Guide on Remedial Actions at Superfund Sites with PCB Contamination
(OSWER Directive No. 9355.4-01 FS) which states that PTW will generally include soils
contaminated at concentrations greater than 100 ppm PCBs, the Boards recommend that
in its decision documents, the Region more thoroughly explain how its reading of Agency
guidance and its approach to treatment at this site are consistent with the statute and
NCP.
Regional Response
A Guide to Principal Threat and Low Level Threat Wastes (OS WFR Directive No.
9380.3-06FS), Highlight 2. lists contaminated sediment and contaminated soil as
examples of "source material," The description of a source material as a principal threat
waste is based on whether the material is considered to be highly loxic or highly mobile
and generally cannot be reliably contained or poses a significant risk to human health or
the em ironmcnt if exposure were to occur. ! his directive also states a preference for
4
-------
treatment of highly toxic materials and. as the Boards note, in A Guide on Remedial
Actions at Superfund Sites with PCB Contamination (OSWHR Directive No. 9355.4-
01FS), it states that principal threat waste will generally include soil contaminated at
concentrations greater than 100 ppm PCBs in residential areas.
At the Rest of River site, contaminated sediment and bank soil in many reaches of the
rher have been demonstrated to be highly mobile, resulting in downstream transport and
unacceptable risks (e.g.. greater than 10"1 for human fish consumption) to human hea:th
and the environment and arc considered to be principal threat wastes. 1 lowever. there are
no locations at which concentrations greater than 100 ppm occur on residential
properties,
With respect to contaminated sediments, FPA's Contaminated Sediment Remediation
Guidance for Hazardous Waste Sires (KPA, 2005) states that although the NC'P provides
a preference lor treatment for "principal threat waste," treatment has frequently not been
selected for contaminated sediment. I Ugh costs, uncertain effectiveness, and/or
community preferences (for on-site operations) are factors lhat lead to treatment b :ing
selected infrequently at sediment sites. The contaminated sediment guidance goe- on to
state that "... the practicability of treatment, and whether a treatment alternative should
be selected, should be evaluated against the NCR's nine remedy selection criteria. Based
on available technology, treatment is not considered practicable at most sediment sites."
Also. "fi]t should be recognized that in-situ containment can also be effective for
principal threat wastes, where that approach represents the best balance of the NC-;J> nine
remed; selection criteria."
Recommendation No. 4 - Remedial Action Objective
The review package slates that R.-i Os w ill address human and ecological risks as veil as
downstream migration of PCBs. The Boards recommend that any decision documents for
an engineering performance-based (dredging lo a depth to allow placement of a 2-2.5
foot cap) remedy that isolates PC Bs in ihe sediments through a hank-to-bank design
should clearly explain why a numeric remediation goal (known as interim media
protection goals [IMPGs] in the review package.) for sediments that is protective of
human health will not he developed. The decision documents should also better explain
where the IMPGs/cleanup standards will be applied (i.e., in which exposure area) in the
floodplain and how meeting these levels will be met and how the RAO will be achieved.
The current thinking on how the RAOs for the remedy will be achieved is retleete 1 in the
draft. Outline of Potential Pe Tormance Standards in Appendix A and in the Revised
Comparative Analysis of Alternatives (Appendix B).
Recommendation No. 5 - Rcmeclv Performance
Based on she information presented, the Boards believe that the proposed cleanup at this
site would leave large quantities of PCBs infloodplain soils. In the future, EPA n ay
determine thai leaving this remaining waste on site is not protective of human health and
5
-------
the env ran mail. Therefore, the Boards recommend that the Region consider including a
contingency remedy {e.g. pursuing other response actions in an adaptive framework) in
the decision documents that would describe a cleanup approach resulting in more risk
reduction through additional floodplain soil source removal or other active remediation
alternatives.
The Region s presentation included a discussion on implementing an adaptive
manage menl approach to the remedial action. The Board and CSTAG recommend that
the decision document better describe that the selected remedy is based on the current
understanding and knowledge of the site and that its implementation will be phased and
conducted within the adaptive management framework. For example, the first phase of
implementation could begin with remediation (or a demonstration project) of Reach 5/1
and Woods Pond (pending the results of further analysis of Woods Pond being a potential
sediment trap) that includes habitat replacement and reconstruction. Additionally, the
Region should describe the various implementation contingency approaches (e.g.,
remediation and habitat mitigation/'replacement/reconstruction methods) that will be
developed to provide implementation options within the adaptive framework. This
description should also include provisions to pilot test amendments to the cap, such as
active amendments and/or granular activated carbon, to reduce the bioavailability of
PCBs. Recent pilot projects for in-sifu amendment.-, at Hunter "v Point (('A) and Grasse
River C'Y) have demonstrated reduction in PCB bioavailability
The Re}, ion stated that there are a number of dams (including the ones at Woods Pond
and Ris'ng Fondj that must he maintained in order for the remedy to be protective. The
Boards note that dams are being removed in a number of places across the country to
improve the environmental conditions of rivers Therefore, the Boards recommend that
the remedy include requirements for addressing contaminated sediments stored behind
the dams as part of any future dam maintenance and/or dam removal activities. Costs for
dam maintenance (to the extent necessary to ensure that sediments remain contained)
and/or sediment removal activities should be included in the cost estimates.
A critical component potentially affecting the success of the Region 's preferred remedy is
the prevention of the future releases of PCBs from the eroding banks in the upper seven
miles ot so of the river. The ('ommonwealth and many of the stakeholders acknowledge
that the banks are eroding significant amounts of PCBs but are strongly opposed to the
type of hard bank stabilization techniques that were used in the upper two miles. The
Boards recommend that the Region provide additional informat ion in the decision
documents supporting the effectiveness of softer bioengineering techniques in this part of
the river with its low gradient, locations with sleep banks, and high flow rates during
storm events. The Region also should explain the key uncertainties that were considered
in evaluating the long-term effectiveness of these bioengineering techniques. In its
presentation to the Boards, the Commonwealth was confident that the extensive bank
stabilize lion proposed in the preferred remedy Mould prevent the river from meandering
and the subsequent formation of new ox how 'lakes. The Commonweal! h believes that
containment of the river within its current banks would have Tjng-lasting detrimental and
irrevoccble impacts on the floodplain wetlands, vernal pools, and many of the
-------
Commonwealth-listed wildlife andplant species that depend on these habitats. 'The
Boards recommend that in the decision documents the Region expand its rationale on
why hank stabilization will not result in the long-term adverse impacts to the ecosystem
suggested by the Commonwealth. I he rationale should address the relative importance or
oxbow lake formation versus periodic flooding on the long-term continued existen "e of
wetlands, vernal pools, and the Commonwealth-listed species that relv on a wetland
ecosystem. The Boards also recommend that in the decision documents, the Region
directly address the Commonwealth's position that channel migration is critical to
" maintaining] a diverse mosaic of wetlands and habitats that support species diversity
over time, " The Boards believe it would he useful for purposes of evaluating alternatives
and ensuring meaningful public participation for the Region to estimate how many of the
66 vernal pools and how many acres of wetlands would disappear or he ecologically
nonfunctional if the river stops meandering.
.Regional Response
The Region and representatives from 1IQ and the states of Massachusetts and
Connecticut have been working cooperatively for the last several months to discuss
potential approaches to clean up the Rest of River. These discussions ibcused, in part, on
the need to address the risks from polychlorinaicd biphcnyls (PC'Bs) to humans, fish,
wildlife, and other organisms while avoiding, mitigating, or minimizing the impacts of
the cleanup on the unique ecological character of the Housatomc River, including the
meandering nature of the river and contaminated eroding banks, and habitat areas ror
state-listed species of concern in lloodplain areas. The draft Outline of Potential
Performance Standards (Appendix A) and the Revised Comparative Analysis of
Alternatives (Appendix B) reflect the current thinking from these discussions. In these
documents, there are provisions that if a future change in land use in the lloodplain
occurs, then performance standards for the new use would apply and could require
additional removal; and provisions for the evaluation of residual levels of contan ination
in the lloodplain that could impact the compliance with the biota and downstream
transport performance standards. The Region believes that the potential approach
outlined in these documents strikes the appropriate balance between these priorities. The
response to the bank restoration questions posed by the Boards is included in Attachment
C-5 of Appendix C.
Adaptive management is included in many facets of the current thinking in the dn?ft
Outline of Potential Performance Standards (Appendix A) and the draft cleanup plan
summarized in Appendix B. The implementation of adaptive management ranges from
conducting, the river cleanup and restoration in a phased approach to piloting the
inclusion of an additive such as organic carbon.
The Region believes that the potential for dam removal and/or maintenance can bt dealt
with in two ways. The first is a contingency remedy providing for cleanup of
contaminated sediment behind the dams to dovetail with a dam removal action: the
second is through a combination of institutional controls on dam monitoring and
7
-------
maintenance, and by having Gh remain responsible for incremental increases in costs
assoeia.ed with the PCBs encountered in normal dam maintenance by a third party.
Recommendation No. 6 - Stakeholders
The Boards appreciate all of the time and effort taken by the stakeholders to provide their
thoughts on the future actions to he taken at this portion of the site.
The package provided to the Board nullities the complexity of the remedy components as
selected through the RCRA permit process yet implemented as a Superftmd remedial
action. It may be challenging to stakeholders to understand the logic/basis of the remedy
option components, how they fit into the overall remedy, and how the remedy as a whole
meets and is consistent with Superfund remedy selection criteria and guidance. The
Board recommends that the Region develop a communication plan for the stakeholders to
concisely and clearly convey how the individual components of the remedy fit together to
achieve the remedial action objectives and meet the criteria for remedy selection.
Regional Response
ihe Re,z;on agrees with the Boards that the role of the regulatory programs at the site is
complex and can be confusing. [ he Region has worked with the stakeholders to clearh
comniu iicate the nine criteria for remedy selection specified n the Reissued RCRA
Permit. The Region sponsored a series of workshops and a charrcttc that, among other
things, did just that. The Region will continue to communicate with stakeholders through
an outre ach program as we go forward.
Recommendation No. 7 - Early Action
In the presentation the Region identified three residential areas above Superfund
residential PCB action levels (i.e. 1 ppmper OSWER Directive No. 9355.4-01 FS, A
Guide on Remedial Actions at Superfund Sites With PCB Contamination) and high use
recreational areas (river access, camping, etc) above PCB action levels. Since the Rest
of River will he implemented as a Superfund remedial action, the Board recommends that
the Region consider conducting an early action (e.g., removal or early interim action) in
parallel with the other Rest of River activity to address the exposure as soon as practical.
Regional Response
The Region has initiated action with GE for the Removal Act'on Area outside the Rest of
River ir which GE has to sample and remediate ""Actual and Potential Lawns" on the
residential properties within the floodplain of Rest of River that exceed the 2 mg/kg
residential cleanup level (based on the Massachusetts residential cleanup standards). The
Region vviil consider other early actions during the remedial design phase of the Rest of
River cleanup.
8
-------
1 be Region appreciates the 3oards" assistance on this complex project. In the months
ahead, we will continue to work on a potential approach to cleanup to release for public
comment. It vou have additional questions regarding the responses in this memoranda!
or any ol the information presented m the appendices, please feel free to contact me or
Susan Svirsky.. Remedial Project Manager, at 671-918-1434.
Cc: James Wootford
David W. Charters
Susan Svirsky
Dean Tagliaferro
Boh Cianciaruio
Tim Conway
-------
APPENDIX A
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL
PERFORMANCE STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
-------
PRELIMINARY DRAFT
GE-PITTSFIELD/HOUSATONIC RIVER SITE
REST OF RIVER
DRAFT OUTLINE OF POTENTIAL PERFORMANCE STANDARDS
FOR ALTERNATIVE SED 9/FP 4 MOD
AUGUST 2012
-------
TABLE OF CONTENTS
I. INTRODUCTION 1
II. POTENTIAL PERFORMANCE STANDARDS 1
A. INTRODUCTION 1
1. General 1
B. DESCRIPTION OF POTENTIAL PERFORMANCE STANDARDS
AM) CORRECTIVE MEASURES 2
1. River Sediment and Banks 2
2. Floodplain 8
3. Restoration of Impacted Areas 12
4. Long-Term Operations, Monitoring, Maintenance 13
5. Sequencing Implementation of Corrective Measures 13
6. Off-Site Disposal of Contaminated Sediment and Soil 13
7. Institutional Controls 14
8. Review of Response Actions 16
9. General Corrective Measure Provisions 16
10. Scope of Work 17
LIST OF TABLES
Table 1 Potential FMPLs for PCBs for Floodplain Soil by EA - Current Use 21
Table 2 Potential FMPLs for PCBs for Floodplain Soil Frequently Used Subareas -
Current Use 22
Table 3 Potential FMPLs for Unrestricted Use - Floodplain and Riverbank Soil 22
Table 4 Potential FMPLs for Agricultural Uses in Floodplain Soil 22
Table 5 Potential FMPLs for PCBs for Floodplain Soil - Future Use 23
LIST OF FIGURES
Figure 1 Housatonic River, Rest of River 24
Figure 2 Housatonic River, Primary Study Area (Reaches 5 and 6) and Reaches 7 and 8 25
Figure 3 Exposure Area Index Map for Reaches 5 and 6 26
Figure 4 Exposure Area Index Map for Reaches 7 and 8 27
Figure 5 Frequently Used Subareas 28
Figure 6 Potential Implementation of Cleanup GE - Pittsfield/Housatonic River Site Rest
of River 29
iii
8/1/2012
-------
LIST OF ATTACHMENTS
ATTACHMENT A-l MASSACHUSETTS DIVISION OF FISHERIES AND WILDLIFE
CORE HABITAT AREA MAPS
IV
8/1/2012
-------
LIST OF ACRONYMS
ARAR
applicable or relevant and appropriate requirement
BEHI
Bank Erosion Hazard Index
CD
Consent Decree
CERCLA
Comprehensive Environmental Response, Compensation, and Liability Act
CFR
Code of Federal Regulations
CSTAG
Contaminated Sediments Technical Advisory Group
CT DEEP
Connecticut Department of Energy and Environmental Protection
EA
Exposure Area
EMNR
Enhanced Monitored Natural Recovery
EPA
U.S. Environmental Protection Agency
EPC
exposure point concentration
ERE
Environmental Restrictions and Easement
FPML
Final Media Protection Level
GE
General Electric Company
HASP
Health and Safety Plan
LTM
Long-Term Monitoring
MA
Commonwealth of Massachusetts
MESA
Massachusetts Endangered Species Act
MNR
Monitored Natural Recovery
NBS
Near Bank Stress
OMM
operation, monitoring, and maintenance
PCB
polychlorinated biphenyl
PSA
Primary Study Area
RCRA
Resource Conservation and Recovery Act
RD/RA
Remedial Design/Remedial Action
SIP
Site Information Package
SOW
Scope of Work
tPCB
total polychlorinated biphenyl
TSDF
transportation, storage, or disposal facility
v
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
I. INTRODUCTION
In July 2011, the United States Environmental Protection Agency (EPA)'s New England
regional office presented site information and potential cleanup strategies to the National
Remedy Review Board. Representatives of EPA's Contaminated Sediments Technical Advisory
Group (CSTAG) also coordinated with and participated in the Board review for this site. In
preparation for that July 2011 meeting, the regional office prepared and submitted a
comprehensive Site Information Package (SIP) summarizing the site history, the nature and
extent of contamination, risks to human health and the environment posed by that contamination,
potential cleanup options under consideration, and stakeholder views on the project.
After the review meeting, the Board issued a set of recommendations to EPA New England,
dated October 20, 2011. Many of the Board's questions and comments concerned how EPA New
England might draft various performance standards and establish cleanup goals. In order to
address many of the recommendations put forth by the Board and to further develop a potential
cleanup strategy for the Rest of River, EPA conducted additional technical evaluations and
worked closely with co-regulators from the Commonwealth of Massachusetts and the State of
Connecticut in a series of facilitated technical discussions, which began in October 2011. In
May 2012, EPA published a status report entitled "Potential Remediation Approaches to the GE-
Pittsfield/Housatonic River Site 'Rest of River' PCB Contamination." This status report
provided an update to the public on the discussions among the agencies and outlined potential
remediation approaches for the Rest of River. EPA New England has also prepared a
supplement to the SIP which, among other things, provides a more detailed description of the
potential approach outlined in the May 2012 Status Report, an approach identified as Alternative
SED 9/FP 4 MOD in the supplement to the SIP.
In an effort to answer many of the Board's questions regarding potential performance standards
and cleanup goals, this preliminary draft document has been prepared. This document outlines
potential performance standards and corrective measures for a potential cleanup similar to the
remedy outlined in the May 2012 Status Report. It is important to note that no formal remedy
proposal decisions have been made and this supplemental information is intended to provide a
more detailed summary of EPA New England's considerations regarding the potential
approaches to cleanup.
II. POTENTIAL PERFORMANCE STANDARDS
A. Introduction
1. General
This document describes a draft set of potential Performance Standards,
and the corrective measures necessary to attain such Performance
Standards, that the Permittee the General Electric Company [GE]) would
perform and submit, pursuant to the Consent Decree (CD), for the
Housatonic River Rest of River if these conditions were included in a
Reissued Resource Conservation and Recovery Act (RCRA) Permit for
Rest of River. The Rest of River area is generally depicted in Figures 1
1
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
and 2. The potential Performance Standards described in this draft outline
address polychlorinated biphenyls (PCBs) that are present in river water,
sediment, floodplain and bank soil, and biota in Rest of River.
As described in the CD and this draft outline, all response activities
associated with Rest of River would be performed by the Permittee under
the oversight and with the approval of the EPA, after reasonable
opportunity for review and comment by the Commonwealth of
Massachusetts (MA) and the Connecticut Department of Energy and
Environmental Protection (CT DEEP). All Permittee activities would be
conducted pursuant to such Permit and the CD. All EPA approvals of
plans and other submittals under such Permit would be pursuant to Section
XV of the CD.
In order to provide a better understanding to reviewers of the approach
EPA might use to establish various Permit criteria for a potential remedy,
this document has been drafted to closely approximate what might
ultimately be included in a draft modification to the Reissued RCRA
Permit if Alternative SED 9/FP 4 MOD were proposed as EPA's preferred
cleanup plan. Thus, language such as "Permittee shall ..as well as use
of the term "shall" in general, has been included in the draft outline.
Notwithstanding, this document represents only one potential approach
and no formal decisions have been made regarding the appropriate scope
of cleanup, specific performance standards, or any specific Permit
requirements.
B. Description of Potential Performance Standards and Corrective Measures
The Permittee shall conduct all necessary corrective measures to meet the
Performance Standards outlined below:
1. River Sediment and Banks
a. General Performance Standards
(1) The Permittee shall implement corrective measures to
achieve and maintain a PCB flux rate over Woods Pond
Dam and Rising Pond Dam, respectively, of 2.0 kg/year or
less, which shall be achieved within 5 years of completion
of construction-related activities outlined herein.
Compliance will be determined based upon the measured
PCB flux over each dam over a range of flow conditions on
an annual basis.
(2) The Permittee shall implement corrective measures such
that the Near-Term Biota Performance Standards are met in
each reach of the river and backwaters within 15 years of
2
8/2/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
completion of construction-related activities for that reach
(or if the reach is subject to Monitored Natural Recovery
(MNR), upon completion of the closest upstream reach
subject to cleanup) under this Permit. Monitoring shall also
be conducted to demonstrate continued progress towards
achieving the Long-Term Biota Performance Standards.
Compliance with the Near-Term Performance Standard will
be determined through the collection of PCB fish and
waterfowl tissue data pursuant to the Long-Term
Monitoring (LTM) Plan. The Near-Term and Long-Term
Biota Performance Standards, or Final Media Protection
Levels (FPMLs) are as follows:
Biota
PCB Concentration (in mg/kg)
Tissue Type
Near-Term Standard
(CTE l() '/HI= l)
Massachusetts Long-Term
Standard (RME 10 5/HI=1)
Connecticut
Long-Term
Standard
Fish tissue - bass fillet
1.5
0.064
0.00018
Duck Breast
1.4
0.075
TBD
b. Reach 5 A
(1) River bed sediment shall be removed and an Engineered
Cap shall be placed throughout Reach 5 A. Sediment
removal and subsequent capping shall result in a final grade
consistent with the original grade or with modifications as
needed considering the principles of Natural Channel
Design, generally using engineering methods employed
from within the river channel or other methods approved by
EPA.
(2) Contaminated soil from eroding riverbanks in Reach 5 A
shall be removed, generally using engineering methods
employed from within the river channel or other methods
approved by EPA.
The locations of contaminated eroding riverbanks shall be
determined using a Bank Assessment for Nonpoint
Consequences of Sediment (BANCS) model1 calibrated for
the Housatonic River and the collection of additional
1 A description of the BANCS model can be found at http://water.epa. gov/scitech/datait/tools/warsss/pla box08.cfm
3
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
riverbank soil PCB data. A bank shall be considered
contaminated if it contains 5 mg/kg or greater of tPCBs
(tPCBs) measured in the surficial foot as the average of
three 12-inch cores taken at the toe, midpoint, and top of
the bank at a spacing of every 25 feet of linear bank. A
bank shall be considered to be erodible if its Bank Erosion
Hazard Index (BEHI) and Near Bank Stress (NBS) rating is
classified as "Moderate-High" or greater, using the
methodology outlined in the BANCS model. The
Permittee shall complete bank excavation for the Thiessen
polygon represented by the sample that is considered to be
contaminated and eroding.
(3) Excavated riverbanks shall be reconstructed to minimize
erosion considering the principles of Natural Channel
Design to result in a channel that is in dynamic equilibrium
and balances flow and sediment loads and reduces erosive
forces. This will allow the maximum use of bioengineering
methods in restoring riverbanks. Riverbank reconstruction
shall follow a hierarchy of approaches as follows:
(a) Reconstruct disturbed banks with bioengineering
restoration techniques;
(b) Reconstruct disturbed banks with a cap layer
extending into the riverbank placed under a
bioengineering layer; or,
(c) Place riprap cap or hard armoring on surface of
banks (e.g., where needed for protection of adjacent
infrastructure).
c. Reach 5B
Three composite samples (center, left, right) shall be collected at
transects placed every 25 feet along the river channel. River bed
sediment exceeding 50 mg/kg tPCBs in any discrete composite
sample collected within the top foot of the sediment bed, shall be
removed and backfilled. Sediment excavation will be conducted
for the Thiessen polygon which is represented by that sample. The
backfill will consist of material with characteristics similar to
existing sediment.
Subsequent to excavation and backfill, Enhanced Monitored
Natural Recovery (EMNR) shall be implemented throughout
Reach 5B. An additive, such as activated carbon or organoclay,
shall be added to Reach 5B sediment to reduce the bioavailability
of the remaining PCBs in the sediment bed. The use of such an
4
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
1
additive in Reach 5B shall be the subject of a Pilot Study to
2
determine the most effective approach that results in compliance
3
with the near-term biota and flux performance standards.
4
[Placeholder for potential additional EMNR Performance
5
Standards /Pilot Study Objectives]
6
Riverbank soil exceeding 50 mg/kg tPCBs in any of three samples
7
(bottom, midpoint or top of the riverbank) collected from the
8
surficial foot of the riverbank collected at an interval of 25 feet of
9
linear bank shall be removed from the Thiessen polygon
10
represented by that sample, and disturbed banks shall be
11
reconstructed using bioengineering methods, similar to those
12
specified in Section II.B.l.b.(3)(a) above, to minimize erosion and
13
reduce downstream transport of the residual PCBs in bank soil.
14
The location of soil and sediment to be excavated per this
15
subsection shall be determined based on the collection of the
16
additional bank soil and sediment PCB data.
17
d.
Reach 5C
18
River bed sediment shall be removed and an Engineered Cap shall
19
be placed throughout Reach 5C.
20
River bed sediment shall be removed, generally using engineering
21
methods employed from within the river channel, with either
22
dredging or wet excavation techniques.
23
e.
Backwaters
24
Contaminated sediment in portions of the backwater areas located
25
outside of Core Area 1 Habitat (shown in Attachment A.l)
26
exceeding 1 mg/kg tPCBs shall be removed to a depth of 1 foot
27
followed by placement of an Engineered Cap.
28
Contaminated sediment in portions of backwater areas located
29
within Core Area 1 Habitats exceeding 50 mg/kg tPCBs shall be
30
removed to a depth of 1 foot followed by placement of an
31
Engineered Cap.
32
The Permittee shall evaluate the placement of an additive such as
33
activated carbon and/or other methods to reduce bioavailability of
34
contamination in areas defined as Core Area 1 Habitat where tPCB
35
concentrations are between 1 and 50 mg/kg in the top foot of
36
sediment. The evaluation may include a Pilot Study. The
37
Permittee shall submit this evaluation, along with any
5
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
recommended actions, to EPA for review and approval. Upon
approval from EPA, the Permittee shall implement such actions.
f. Woods Pond (Reach 6)
Contaminated sediment shall be removed and an Engineered Cap
shall be placed over residual PCBs to result in 1) a minimum of 1
foot sediment removal, and 2) a post-capping minimum water
depth of 6 feet, except in near-shore areas where an appropriate
bathymetry that replicates the existing littoral habitat shall be
constructed.
The Permittee shall evaluate additional deepening, reconfiguration
and/or placement of structures within Woods Pond to increase the
rate of sedimentation to ensure that the PCB flux Performance
Standard, in Section II.B.l.a.(l) above, is met. The Permittee shall
submit this evaluation, along with any recommended actions, to
EPA for review and approval. Upon approval from EPA, the
Permittee shall implement such actions.
The Permittee shall include the placement of an additive such as
activated carbon in Woods Pond after excavation and prior to and
after capping to sequester the residual PCBs and PCBs that migrate
into Woods Pond.
If, during LTM, EPA determines that significant concentrations
and depth of PCB-contaminated sediment has accumulated in
Woods Pond, the Permittee shall remove such sediment and
dispose of off-site.
g. Columbia Mill Impoundment (Reach 7B), Eagle Mill
Impoundment (Reach 7C), Willow Mill Impoundment (Reach 7E),
Glendale Impoundment (Reach 7G), and Rising Pond (Reach 8)
Prior to design and implementation, additional sediment sampling
shall be conducted to determine tPCB concentrations to define the
extent of the corrective measures.
In these areas, the Permittee shall:
(1) Remove sediment, design and install an Engineered Cap in
areas where sediment tPCB concentrations exceed 1 mg/kg
at any depth and repair dams as necessary to allow for the
caps to function properly; or
(2) Remove all sediment greater than 1 mg/kg tPCBs, and
place backfill, if necessary. Sediment removal activities
6
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
shall be coordinated with dam removal activities that may
be performed by others; or
(3) a combination of (1) and (2).
h. Flowing Subreaches in Reach 7 and Reaches 9 Through 16
MNR shall be implemented in these reaches. LTM of PCB
concentrations in affected media, including sediment, surface
water, and biota shall be conducted to evaluate the progress of
MNR in achieving the Near-Term Biota Performance Standard and
toward meeting the Long-Term Biota Performance Standard.
i. Engineered Cap Design
All Engineered Caps constructed as part of the corrective measures
outlined herein shall conform to the following Performance
Standards:
(1) Engineered Caps shall include a sacrificial mixing layer to
minimize the potential for the isolation layer to become
contaminated upon placement on the existing sediment.
(2) Engineered Caps shall include an isolation layer that will
minimize the PCB flux up through the cap and into the
surface water.
(a) In Reaches 5A and 5C, the default standard shall be
a 6-inch isolation layer with an additive such as
activated carbon to attenuate the flux of PCBs.
However, the performance of such a layer
(including the composition and thickness) shall be
demonstrated through the use of a model within the
peer-reviewed literature, such as the model used in
Technical Attachment K of Appendix E of the CD.
(b) In the Backwaters, Woods Pond, and the Reach 7
impoundments, the default standard for the isolation
layer shall be a 6-inch layer that has adequate
resistance to the bed sheer stress and contains or is
comprised of an "active layer," including an
organoclay and/or organic carbon. This layer shall
meet the function requirements of the isolation layer
described above. This may eliminate the need for
an additional erosion/ bioturbation protection layer
(described below).
7
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
(3) Engineered Caps shall include an erosion and bioturbation
protection layer that prevents exposure to and erosion of
the underlying isolation layer. The design criteria shall be
a 100-year flow event.
(4) Engineered Caps shall include a habitat layer that provides
functions and values equivalent to the pre-existing surficial
sediment substrate.
(5) The Permittee shall evaluate, and submit for EPA approval,
the need for additional layers (such as a filter layers) or
other cap configurations necessary for the proper function
of the cap and include any such additions in its cap
design(s).
(6) Installation of the cap shall not adversely affect flood
storage capacity.
(7) Engineered Caps shall be inspected, monitored, and
maintained to ensure long-term protectiveness and to
ensure that they continue to function as designed.
2. Floodplain
a. Floodplain Soil Adjacent to Reaches 5 through 8
For each Exposure Area (EA) (see Figures 3 and 4), excavate and
replace the top 1 foot of soil to achieve either the Primary or
Secondary Performance Standards (or Final Media Protection
Levels (FMPLs)) listed in Table 1. In addition, for each
Frequently Used Subarea (shown in Figure 5), excavate and
replace the top 3 feet of soil to achieve the Performance Standards
(or FMPLs) listed in Table 2. The Permittee shall achieve these
Performance Standards [or FMPLs] in accordance with the
following procedures:
(1) The Permittee shall conduct additional sampling of
floodplain soil (as needed) to confirm the tPCB
concentration and the exposure point concentration (EPC)
for each EA using a Thiessen polygon approach.
(2) The Permittee shall design a remediation plan based on
meeting Primary Performance Standards (or FMPLs) for
each EA and each Frequently Used Subarea.
(3) The Permittee shall conduct additional field reconnaissance
as needed to identify and quantify the potential impacts of
proposed remediation on state-listed species.
8
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
(4) The Permittee shall evaluate the potential impacts on state-
listed species and their habitats, and shall formulate
procedures to avoid, minimize, or mitigate any such
adverse impacts in accordance with the substantive
requirements of a "Conservation and Management Plan"
under the Massachusetts Endangered Species Act (MESA),
including providing a long-term Net Benefit for the
conservation of the affected state-listed species. All
evaluations shall be conducted following the above
referenced MESA regulatory standards. In conducting this
evaluation, the following approach shall be used:
(a) Remediation in Frequently Used Subareas to attain
Primary Performance Standards (FPMLs) in these
subareas;
(b) Avoidance of Core Area 1 Habitats for EAs other
than Frequently Used Subareas, except in limited
areas to meet Secondary Performance Standards
(FPMLs)2;
(c) Minimization or mitigation for Core Area 2 (as
shown in Attachment A. 1); and
(d) Case-by-case determination for Core Area 3 (as
shown in Attachment A. 1).
The Permittee shall submit this evaluation and procedures to EPA
for review and approval.
(5) Based on the outcome of Section (4) above, EPA shall
identify any areas to be avoided and the Permittee shall
recalculate the EPC to ensure that the resultant excavation
plan meets, at a minimum, Secondary Performance
Standards (FPMLs).
(6) To the extent that Secondary Performance Standards are
not met, the Permittee shall propose additional areas to be
excavated in order to meet Secondary Performance
Standards (FPMLs), repeating Steps (4) and (5) as needed.
(7) The Permittee shall also evaluate the presence of any areas
of remaining PCB concentrations in floodplain soil for
2 For example, the Commonwealth has identified a portion of Exposure Area (EA) 10 as a Core 1 area that should be
avoided, while portions of Core 1 areas in EAs 19, 37a, and 62 could be excavated to meet Performance
Standards. Case-by-case determinations will be made by EPA in consultation with the Commonwealth.
9
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
erosion potential and the likelihood of future downstream
transport at levels that could result in not meeting the biota
and PCB flux Performance Standards in Section II.B.l.a.
The Permittee shall reevaluate, as needed, any area of
proposed floodplain soil remediation, considering the
erosion potential and Steps (3) through (6) above, and shall
propose further action as necessary to ensure compliance
with the biota and PCB flux Performance Standards
(FMPLs).
The Permittee shall submit the revised evaluation to EPA. Upon
approval by EPA, the Permittee shall implement the required
actions.
b. Future Use Scenarios for All Floodplain Areas (Adjacent to
Reaches 5 through 16)
The Permittee shall implement institutional controls at all
properties where floodplain soil exceeds 2 mg/kg tPCBs in
Massachusetts or 1 mg/kg PCBs in Connecticut (see Table 3) to a)
restrict future uses, or b) require additional evaluations and, if
necessary, additional response actions, to be protective of such
future uses. Institutional Controls shall be implemented consistent
with Section 7 below.
For floodplain areas where the use changes from current use, the
Permittee shall:
(1) Determine the appropriate exposure scenario from Tables 4
and 5.
(2) Determine the EPC for the EA.
(3) Evaluate whether or not the EPC meets the Performance
Standards in Table 4, where applicable, or the Primary
Performance Standard/FMPL listed in Table 5. For non-
agricultural future uses, if the EPC exceeds the Primary
Performance Standard/FMPL, then the Permittee shall
follow the procedures outlined in Section II.B.2.a. above to
determine whether additional response actions are required.
The Permittee shall submit this evaluation to EPA. Upon approval
by EPA, the Permittee shall implement the required actions.
c. Vernal Pools
In addition to any remediation conducted in Vernal Pools to meet
Floodplain FMPLs, the Permittee shall implement the following:
10
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
(1) The Permittee shall conduct one or more site visits to
confirm or revise the delineation of the vernal pools. The
Permittee shall submit this evaluation to EPA for approval.
EPA, in consultation with the Commonwealth, will make
the determination as to what constitutes a vernal pool.
(2) The Permittee shall conduct additional sampling and
characterization of vernal pools, to generate baseline data
on the concentrations of tPCBs and health/abundance of
animal species, including but not limited to, state-listed
species. The Permittee shall also conduct additional field
reconnaissance as needed to quantify the potential effects
of remediation of the vernal pools on any state-listed
species.
(3) The Permittee shall identify vernal pools that exceed the
Vernal Pool-Specific Performance Standard FMPL of 3.3
mg/kg tPCBs (based upon risk to amphibians).
(4) For vernal pools identified as requiring cleanup solely to
meet the Vernal Pool-specific Performance Standard
FMPL, EPA will make case-specific remedial decisions
(including traditional excavation/restoration, alternative
remedial strategies, deferment of remediation, and
preservation of existing conditions) weighing field
evidence of species health/abundance with the Vernal Pool-
Specific Performance Standard FMPL using the following
approach:
(a) EPA will select an initial number of approximately
8 to 10 pools for remediation by traditional means
(excavation and reconstruction); pools within Core
Area 1 will be excluded from consideration from
this initial number of pools.
(b) EPA will select an initial number of additional
pools for pilot testing of an additive such as
activated carbon in lieu of excavation.
(c) EPA will select an initial number of additional
pools for pilot testing by a third remediation method
to be proposed by the Permittee for EPA approval
and/or additional pools to be monitored
concurrently with remediated pools as a "reference"
group for comparison purposes.
11
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
(d) EPA will identify performance metrics and
evaluation criteria for comparison of the various
remediation approaches.
(5) The Permittee shall evaluate potential impacts on state-
listed species and their vernal pool habitat and shall
formulate procedures to minimize or mitigate any such
adverse impacts in accordance with the substantive
requirements of a "Conservation and Management Plan"
under MESA, including providing a long-term Net Benefit
for the conservation of the affected state-listed species. All
evaluations shall be conducted following the above
referenced MESA regulatory standards. The Permittee
shall submit the evaluation and proposed remediation plan
to EPA for approval.
(6) The Permittee shall complete the initial round of vernal
pool remediation referenced above in (4)(a) through (4)(d)
and submit to EPA an evaluation of the success of
remediation, mitigation and restoration of each method.
(7) Based on the evaluation of the initial round of vernal pool
remediation and mitigation and taking into the
consideration the Core Areas, EPA will determine the
preferred method/approach to remediation of each
subsequent vernal pool and the Permittee shall implement
this approach. In considering the Core Areas, the following
approach will generally be used:
(a) Avoidance of excavation in Vernal Pools within
Core Area 1 Habitats;
(b) Minimization or mitigation for excavation in Vernal
Pools within Core Area 2 (as shown in Attachment
A.l); and
(c) Case-by-case determination for excavation in
Vernal Pools within Core Area 3 (as shown in
Attachment A-l).
3. Restoration of Impacted Areas
A habitat restoration program shall be designed as a component of the
corrective measures to restore the functions and values of areas impacted
through implementation of corrective measures. Particular focus shall be
placed on complying with the substantive requirements of pertinent
statutes and regulations, including the MESA.
12
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
4. Long-Term Operations, Monitoring, Maintenance
A long-term operation, monitoring, and maintenance program (OMM)
shall be implemented to evaluate the effectiveness of the corrective
measures in achieving Performance Standards and to determine whether
maintenance or other response actions are necessary to achieve and
maintain compliance with Performance Standards, including ensuring that
continued progress is being made toward achieving the Long-Term Biota
Performance Standards. This program shall be designed to be consistent
with an Adaptive Management framework, as outlined in Section II.B.9.a.
below.
5. Sequencing Implementation of Corrective Measures
Implementation of the corrective measures shall begin concurrently in
Reach 5A (sediment and floodplain) and Woods Pond. Corrective
Measures shall proceed downstream from Reach 5 A and Woods Pond on a
parallel track as shown in Figure 6. The final sediment caps in the
impoundments shall not be placed, however, until all remediation of
upstream Reaches has been completed. Following the placement of the
cap in Reach 7G, sediment removal and subsequent capping shall take
place in Rising Pond (Reach 8). This approach shall be subject to review
under an Adaptive Management framework to evaluate the effectiveness
of sequencing.
If the Permittee opts to remediate any part of Reaches 7 or 8 to meet the
1 mg/kg Performance Standard in lieu of capping, EPA may require an
alternative sequencing to that described above.
The corrective measures in the floodplain shall be performed by the
Permittee while the adjacent sediment cleanup activities are taking place
and shall share construction infrastructure to the maximum extent
practicable to minimize the corrective measures footprint.
6. Off-Site Disposal of Contaminated Sediment and Soil
All remediation waste material, including contaminated sediment and
soil, shall be disposed off-site at permitted treatment, storage, or disposal
facilities (TSDFs) that are in compliance with EPA's off-site rule (40
Code of Federal Regulations [CFR] 300.440.] The Permittee shall
maximize the transport of contaminated material to off-site permitted
facilities via rail.
The Permittee shall comply with Paragraph 41 of the CD.
During the implementation of the corrective measures, the Permittee may
propose to EPA for approval, innovative treatment technologies, as part of
the Adaptive Management as outlined in Section II.B.9.a. below.
13
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
[Placeholder for Standards for rail facilities]
7. Institutional Controls
a. The Permittee shall post biota consumption advisories and shall
maintain such advisories until the Long-Term Biota FMPL is met.
The Permittee shall cooperate with EPA and the states in the
preparation and distribution of educational and outreach activities
to improve public awareness of the advisories, including the
provision to hunting and fishing license distributors of appropriate
written notices regarding such advisories to be included with
licenses.
b. Sediment/Dams—The Permittee shall:
(1) Operate, inspect, monitor and maintain Woods Pond and
Rising Pond dams unless sediment concentrations behind
the dams are 1 mg/kg tPCBs or less, even if the Permittee
transfers ownership interest in the dams. Operation,
inspection, monitoring and maintenance activities shall be
designed and implemented to prevent dam failure and
unpermitted releases of sediment.
(2) With respect to Columbia Mill, Eagle Mill, Willow Mill
and Glendale dams, conduct activities to prevent dam
failure and unpermitted releases of sediment unless a) the
Permittee demonstrates that these activities are being
performed by another party, b) the dam is removed, or c)
the sediment concentrations behind the dams are 1 mg/kg
tPCBs or less.
(3) Pay for all costs associated with and attributable to the
presence of PCBs for any legally permissible use that
requires sampling, handling or off-site disposal of sediment
with tPCB concentrations greater than 1 mg/kg, including,
but not limited to, activities related to dam maintenance or
removal, road, infrastructure projects, and activities such as
installation of canoe and boat launches, docks, etc. with
respect to Reaches 5 through 16 in Rest of River.
(4) To the extent that any dam failure and/or unpermitted
release occurs with respect to a dam, pay for the costs
associated with PCBs with respect to Reaches 5 through 16
in Rest of River.
c. Floodplain Soils—For floodplain soils, the Permittee shall, for all
properties with data confirming that soil concentrations exceed 2
mg/kg in Massachusetts (based upon the Massachusetts
14
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
Contingency Plan SI soil standards) or 1 mg/kg in Connecticut
(based upon the Connecticut Remediation Standard Regulations
Direct Exposure Criteria):
(1) Prepare and record Environmental Restrictions and
Easements (EREs) for properties owned by GE and the
Commonwealth that restrict soil handling and changes in
current uses (e.g., prohibits conversion to agricultural or
residential use). These activities shall be conducted in
accordance with Section XIII of the CD.
(2) Offer compensation for an ERE (or equivalent land use
restriction) to non-residential property owners where there
is reasonable potential for changes in future use to activities
such as residential or agricultural. These activities shall be
conducted in accordance with Section XIII of the CD.
(3) If the owner declines the ERE offer, implement a formal
Conditional Solution whereby the Permittee commits to
perform additional response actions in the future, should
the property owner commit to change the current use of the
property. In addition, the Permittee shall pay all testing,
soil handling and disposal costs associated with PCBs for
property owners who elect to remove soil from their
property for a legally permissible use.
(4) For current residential properties, excluding the actual or
potential lawn portion of properties that are addressed in
the Residential Floodplain Properties located Downstream
of the Confluence, implement a Conditional Solution that
obligates the Permittee to perform additional response
activities should the type of use change to exposures
consistent with residential/agricultural use and to require
the Permittee to pay all testing, soil handling and disposal
costs associated with PCBs for property owners who elect
to remove soil from their property for a legally permissible
use. Alternatively, the Permittee may purchase portions of
the property and place EREs on the GE-owned property.
(5) For non-residential properties where there is not a
reasonable potential for a change in future use, Permittee
shall pay all testing, soil handling and disposal costs
associated with PCBs for property owners who elect to
remove soil from their property for a legally permissible
use.
(6) The Permittee shall conduct inspections every 5 years to
determine whether or not property owners have changed
15
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
the use of a property such that a re-evaluation of
protectiveness is required. If so, the Permittee shall inform
EPA and EPA will determine whether additional response
actions are necessary. In addition, if the EPA or the State
notifies the Permittee of such conditions outside of the
Permittee's 5-year review process and determines that
additional response actions are necessary, the Permittee
shall conduct such response actions.
8. Review of Response Actions
In accordance with paragraph 43 of the CD, the Permittee shall conduct
studies and investigations as requested by EPA to permit EPA to conduct
periodic reviews, consistent with Section 121(c) of Comprehensive
Environmental Response, Compensation, and Liability Act (CERCLA)
and any applicable regulations, of whether Remedial Actions are
protective of human health and the environment. The Permittee shall also
comply with any additional requirements pursuant to Section X of the CD
with respect to periodic reviews.
9. General Corrective Measure Provisions
The Permittee shall design, construct, operate, inspect, monitor, and
maintain the corrective measures in compliance with all applicable
provisions of the CD and Permit (including all technical attachments and
submittals thereunder) and in compliance with the Performance Standards
identified therein. This section sets forth a number of general requirements
associated with the corrective measures for Rest of River:
a. Adaptive Management
An adaptive management approach shall be implemented between
the Permittee and EPA in the conduct of the corrective measures to
adapt and optimize project activities to account for "lessons
learned," new information, changing conditions, innovative
technologies, results from pilot studies, and additional
opportunities that may present themselves over the duration of the
project. The Permittee shall modify the implementation of the
corrective measures, with EPA approval, through this process to
minimize the adverse impacts of the response action, expedite the
response, improve the restoration activities, and to ensure
compliance with, or continued progress toward, achieving
Performance Standards.
The Permittee shall perform the corrective measures in accordance
with any modifications that are identified by the Permittee (with
EPA's approval), or are required by EPA, through adaptive
16
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
1
management, to the Scope of Work (SOW), or any other plans,
2
specifications, schedules, or other documents.
3
b.
Applicable or Relevant and Appropriate Requirements (ARARs)
4
The federal and state laws and regulations that constitute ARARs
5
for the response actions for Rest of River will be identified in
6
tables contained in an attachment to the Permit.
7
The tables included in such an attachment would present a
8
description of the listed ARARs and a determination by EPA as to
9
whether and how the listed ARARs should be met, or whether a
10
waiver is proposed. Unless EPA determines that a waiver is
11
appropriate, the Permittee shall comply with and attain the ARARs
12
listed in those tables. To the extent that EPA determines that
13
waiver of an individual ARAR is appropriate, the Permittee shall
14
comply with any modified performance requirement based on
15
EPA's waiver determination.
16
In addition, the technical Remedial Design/Remedial Action
17
(RD/RA) submittals for response actions for the Rest of River shall
18
specify additional ARARs (not listed in such attachment), if any,
19
for such response actions and shall contain a proposal as to how
20
the response action will comply with such additional ARARs or if
21
a waiver is appropriate. The Permittee shall comply with and attain
22
all such additional ARARs that EPA determines should be met by
23
such response action.
24
c.
Coordination of Corrective Measures
25
Corrective measures associated with the Rest of River will require
26
a significant level of project scheduling, coordination, and
27
sequencing which shall be addressed by the Permittee in the Scope
28
of Work (SOW). As the corrective measures are expected to be
29
implemented in a phased approach, it is expected that the work to
30
be implemented in each phase will have its own set of deliverables,
31
including several of the deliverables identified in Section 7.
32
10. Scope of Work
33
a.
As required in Paragraph 22.x of the CD, the Permittee shall
34
submit an SOW for the RD/RA process for implementation of the
35
corrective measures, including pre-design activities and the
36
subsequent performance of corrective measures. The SOW shall
37
include a description of, and a submittal schedule for, at a
38
minimum, the following documents:
17
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
(1) Overall Strategy and Schedule for Implementation of the
Corrective Measures
(a) Coordination of floodplain and sediment
remediation
(b) Sequence of remediation
(c) Methods to minimize impact to neighborhoods and
general public and to limit use of certain roads
(d) Project management structure
(2) Pre-Design Investigation Work Plans
(3) Pre-Design Investigation Summary Reports
(4) Conceptual RD/RA Work Plans
(5) Final RD/RA Work Plans
(6) Supplemental Implementation Plans (contractor Health and
Safety Plans [HASPs], operations plan)
(7) Updated Project Operations Plan and Field Sampling
Plan/Quality Assurance Project Plan for Rest of River-
Specific Changes
(8) Sediment Processing/Off-Site Transfer Facility Work Plan.
(9) Quality of Life Compliance Plan
(a) Noise, dust, odor, light standards
(b) Continued recreational activities
(c) Road use, including restrictions on transport of
waste material through residential areas
(10) Plan for Compliance with MESA
(11) Pilot Study Proposals
(12) Restoration Plan
(13) Adaptive Management Plan
(14) Community Engagement Plan
(15) Cultural Resource Plan
18
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
(16) Model Reevaluati on PI an
(17) Dam Operation, Inspection and Maintenance Plan
(18) Completion of Work Reports
(19) Long-Term Operation, Monitoring, and Maintenance Plan
(20) Institutional Control Plan
(21) Plan for Further Response Actions, and any implementation
of further response actions, in accordance with Section X of
the Decree (Review of Response Actions)
b. Adherence to the procedures and protocols presented in the above
plans will provide a level of consistency and comparability for the
evaluations and response actions conducted for each corrective
measure, and will also establish minimum requirements concerning
analytical and construction quality assurance, site security, and
health and safety and compliance with ARARs.
The Overall Strategy will present the Permittee's overall strategy
for implementing the corrective measures that have been selected
by EPA, including the preparation of work plans, designs, and
reports, completion of pre-design investigations, and construction
and implementation of long-term OM&M. In addition, the Overall
Strategy will describe the Permittee's project organizational
structure, and roles and responsibilities, and lines of
communication between the Permittee and EPA, and will include
the project organization and a project implementation schedule.
In addition, the contents of the documents required by the SOW
are subject to modification or adjustment based on specific
activities for a given corrective measure and any site- or activity-
specific considerations using an adaptive management approach. If
deviations to such documents are proposed for a specific corrective
measure, such proposals will be presented in the technical
deliverables specific to that corrective measure.
c. Community Engagement
The Permittee shall, consistent with the CD, prepare a Community
Engagement Plan that builds upon the current stakeholder and
community outreach program to include any material changes in
the level of concern or information needs of the community that
are anticipated to occur during design and construction activities.
The following activities, at a minimum, shall be specified in the
Community Engagement Plan, for conduct by the Permittee:
19
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
1
2
3
4
5
(1)
Prepare and distribute progress fact sheets to inform the
public of work completed and notify the public of planned
activities at the initiation of the major phases of work,
including major pre-design investigations, final designs,
and major construction activities.
6
7
8
9
10
(2)
Maintain a public information repository and website
(similar to that of the Hudson River project) to provide
community access to work products, including technical
reports, work plans and project fact sheets as well as
current project activities.
11
12
13
(3)
Designate a point of contact to provide the community with
a source for information regarding the status of the
corrective measures and related activities;
14
15
16
(4)
Conduct regular meetings, information, and listening
sessions to obtain input and convey updates on the design
and construction of the remedial measures; and
17
18
19
(5)
Continue regular correspondence and/or meetings with
local community groups.
20
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
1 Table 1 Potential FMPLs for PCBs for Floodplain Soil by EA - Current Use
Exposure Areas
FMPL (in mg/kg)
Primary
(RME 10 s/
HI=1)
Secondary
(RME 10
4/ HI=1)
Exposure Scenario Basis
Exposure
Scenario/Receptor
Assumed
Frequency of
Use
10, 10a, 70, 87
4.6
4.6
General Recreation, young
child (high use)
90d/yr
2b, 25, 78, 85b
27
27
General Recreation, older
child (high use)
90d/yr
3, 11, 13-17, 19, 20, 24,32,
33, 38, 44-46, 48, 54, 58,
67-69, 73-77, 79, 89
14
38
General Recreation, adult
(high use)
90d/yr
2, 4, 5, 7, 12, 21, 22, 26a,
26F, 27, 28, 30, 31,31a, 35,
35a, 37, 37b, 40, 40b, 55,
57, 59, 60, 90
14
27
General Recreation,
adult/older child (high use)
90d/yr
1, 56
21
40
General Recreation,
adult/older child (medium
use)
60d/yr
23, 88
40
40
General Recreation, older
child (medium use)
60d/yr
18, 34, 41, 42, 43
21
58
General Recreation, adult
(medium use)
60d/yr
6, 49, 50, 51, 80a, 81, 82, 84
43
115
General Recreation, adult
(low use)
30d/yr
2a, 9
80
80
General Recreation, older
child (low use)
30d/yr
29
43
80
General Recreation,
adult/older child (low use)
30d/yr
37a, 38a, 40a, 41a, 42a, 43a,
59a, 70a, 71,72, 87a
26
42
Bank Fishing
adult/older child
30d/yr
22a, 27a, 28a
14
14
Dirt Biking/ATVing (older
Child)
90 d/yr
8,47, 47F, 52, 53, 60a, 85a
12
28
Recreational Canoeist
Adult 60 d/yr
Older child
30 d/yr
39
7.8
13
Marathon Canoeist
150d/yr
26b, 36b, 80b
12
43
Agricultural Use (farmer)
40d/yr
36a
89
126
Low-Use Commercial
(groundskeeper)
30d/yr
83, 86
18
25
High-Use Commercial
(groundskeeper)
150 d/yr
61-66
169
242
Utility Worker
5 d/yr
50a, 51a, 55a, 56a
90
140
Waterfowl Hunting
14 d/yr
Notes:
ATV = All Terrain Vehicle
21
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
Table 2 Potential FMPLs for PCBs for Floodplain Soil Frequently Used
Subareas - Current Use
Exposure Area
Primary
(in mg/kg)
Exposure Scenario Basis
Exposure
Scenario/Receptor
Assumed Frequency
of Use
4, 12, 26a, 37b, 40, 58, 59
14
General Recreation,
adult/older child (high use)
90d/yr
39
7.8
Marathon Canoeist
150d/yr
47, 52, 53, 60a
12
Recreational Canoeist
Adult 60 d/yr
Older child 30 d/yr
Table 3 Potential FMPLs for Unrestricted Use - Floodplain and Riverbank
Soil
State
IMP I
(PCBs in mg/kg)
Basis
Massachusetts
2
MCP S-l Standard, 301 CMR 40.0000
Connecticut
1
RSR Soil Standards, Conn. Gen. Stat. 22a-
133k-l through K-3 Appendix A
Notes:
MCP = Massachusetts Contingency Plan
RSR = Remediation Standard Regulations
Table 4 Potential FMPLs for Agricultural Uses in Floodplain Soil
Because the Interim Media Protection Goals (IMPGs) for agricultural uses are expressed
as diet, a formula back-calculating from the dietary concentrations to concentration of
PCBs in floodplain soil is necessary for evaluating risk associated with future use of
floodplain exposure areas. The calculation has to take into account the risk level for an
exposure scenario, the exposure point concentration, and the fraction of the use
conducted in the floodplain. The calculation below encompasses these factors.
/"* /"* Rt Fea
Lt — Lea r, F >
Kea ?t
where:
Ct = the target contaminant concentration,
Cea = the known or nominal concentration associated with a calculated risk,
Rt = the target risk level (HI = 1 or 10~5 probabilistic risk),
Rea = the risk calculated for the known or nominal concentration,
Ft = fraction in the floodplain for the target EPC, and
Fea = fraction in the floodplain for the known or nominal EPC.
22
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
1 Table 5 Potential FMPLs for PCBs for Floodplain Soil - Future Use
Type of
Area/Exposure
Scenario
Receptor
Assumed
Frequency
of Use
FMPL (in mg/kg)
Primary
(RME 10 s/
HI=1)
Secondary
(RME 104/
HI=1)
Residential
All
All
2
2
General Recreation
Young child
90 d/yr
4.6
4.6
15 d/yr
27
27
Older child
90 d/yr
27
27
60 d/yr
40
40
30 d/yr
80
80
Adult
90 d/yr
14
38
60 d/yr
21
58
30 d/yr
43
115
Bank fishing
Older child
30 d/yr
42
42
Adult
30 d/yr
26
56
Dirt biking/ATVing
Older child
90 d/yr
14
14
Marathon canoeist
Adult
150 d/yr
7.8
13
Recreational canoeist
Older child
30 d/yr
42
42
Adult
60 d/yr
12
28
Waterfowl hunting
Older child
14 d/yr
140
140
Adult
14 d/yr
90
196
Agricultural use
(farmer)
Adult
40 d/yr
12
43
Commercial
(groundskeeper)
Adult
150 d/yr
18
25
30 d/yr
89
126
Utility worker
Adult
5 d/yr
169
242
2
23
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
Figure 1 Housatoriic River, Rest of River
24
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
Confluence of East and West Branches
HTTBREIG
Reach 5A
ach 5D
>k wate rs
(fifchncrt
Reach 7 A
Reach 7B
BTCCKBBOGB
Columbia Milli
LeejEagle Mills Dar
Reach 7C
Reach 7D
Willow Mil ID
GREAT iBARR N OT Off
Fife: vr*g6pltftnxds'fira_reacbef_»pd3t jnxd. 6/19fiU12 10:15:15*11. rlcftsc
LEGEND:
O Towi/City
Roads
Housatonic River
State P ark
Municipal Boundary-
Reach Division Line
10-Vfear Floodplain
~s n.+s ~ ds Kilo metera
GE- P ittsti eld JHousatonic River Site
Rest of River
HOUSATONIC RIVER,
PRIMARY STUDYAREA
(REACHES SAND 6)AND
REACHES 7 AND 8
Reach 5C
/oidW* P,S3o) Reach 6 (Woods Pond)
Reach 7E
OTjTtTJl-tSTl
Figure 2 Housatonic River, Primary Study Area (Reaches 5 and 6)
and Reaches 7 and 8
25
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
10 CEpiitWi_e®a(I.^dc..MR_SepL20«li56„MBJ ap> | Imta7-11 o:igoti»W. .MPL2KWHV-1 8f» | ItkO* AIA 21/2005 1
Figure 3 Exposure Area Index Map for Reaches 5 and 6
26
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE STANDARDS FOR ALTERNATIVE
SED 9/FP 4 MOD
10 «E(*l»_e«^m.ifc_Nw_8epL20Wf78Jnda apr | lnda._7-3 | afcepMh_ei^_±JtajepL20WW-Zep* I «W AM. 2/1/30061
Figure 4 Exposure Area Index Map for Reaches 7 and 8
27
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
LOCATOR MAP
Frequent Use Subareas
Figure 5 Frequently Used Subareas
28
8/1/2012
-------
PRELIMINARY DRAFT OF DRAFT OUTLINE OF POTENTIAL PERFORMANCE
STANDARDS FOR ALTERNATIVE SED 9/FP 4 MOD
Reach 5A
Reach 5B
Reach 5C
Backwaters
Reach 6
(Woods Pond)
Reach 7B
Reach 7C
Reach 7E
Reach 7G
Reach 8
(Rising Pond)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Year of Implementation
m
I Removal and
Immediate Capping
I Removal
l Capping
r
w>
1?P-flfiA5-1R
Figure 6 Potential Implementation of Cleanup
GE - Pittsfield/Housatonic River Site
Rest of River
29
8/1/2012
-------
ATTACHMENT A-1 MASSACHUSETTS DIVISION OF FISHERIES AND
WILDLIFE CORE HABITAT AREA MAPS
-------
Commonwealth of Massachusetts
Division off
Fisheries & Wildlife
MassWildlife
Wayne F. MacCallum, Director
July 31, 2012
Robert G. Cianciarulo, Chief
Massachusetts Superfund Section
Office of Site Remediation and Restoration
EPA New England (OSRR-07-01)
5 Post Office Square
Boston, MA 02109-3912
Re: Housatonic River, Core Habitat Areas in the Primary Study Area
Dear Mr. Cianciarulo:
As you are aware, the states of Massachusetts and Connecticut have been working cooperatively
for the last several months to discuss potential approaches to clean up the Rest of River portion of
the GE Housatonic site. These discussions have focused, in part, on the need to address the risks
from polychlorinated biphenyls (PCBs) to humans, fish, and wildlife while avoiding, mitigating
or minimizing the impacts of the cleanup on the unique ecological character of the Housatonic
River. Minimizing impacts to habitat and, in particular, species listed pursuant to the
Massachusetts Endangered Species Act, M.G.L. c. 131A ("MESA"), and 321 CMR 10.00 (the
"MESA Regulations") presents unique challenges as almost the entire Primary Study Area (PSA)
is mapped as Priority Habitat for state-listed species (for a description of Priority Habitat and its
regulatory function please see:
http://www.mass.gov/dfwele/dfw/nhesp/regulatory re view/priority habitat/priority habita
t home .htm. Therefore, in order to help identify the most important areas for habitat protection,
as well as habitats and species that might be particularly sensitive to impacts from PCB
remediation activities, the Massachusetts Division of Fisheries and Wildlife ("DFW") developed
maps of "Core Habitat Areas." The purpose of this letter is to provide an overview of the
approach we used to identify the Core Areas.
As part of our Priority Habitat mapping process, taxonomic experts from DFW's Natural
Heritage & Endangered Species Program ("NHESP") routinely delineate habitat for each state-
listed species, based on actual field-documented records, or "occurrences." There are four types
of Housatonic Core Areas. Core Areas 1, 2, and 3 represent subsets of the delineated state-listed
species habitat found in the PSA. Core Area 4 represents a subset of the documented and
potential vernal pool habitat in the PSA. Please refer to the enclosed maps dated May 21, 2012
which depict the locations of these Core Areas, entitled "Core Habitat Areas, Housatonic River
Primary Study Area (PSA)", "Core Habitat Areas (Core Area 2), Housatonic River Primary Study
Area (PSA)", and "Part of the Housatonic River Showing Primary Study Area, High Species
Richness, and Vernal Pools".
Core Area 1 includes the highest quality habitat for species that are most likely to be adversely
impacted by PCB remediation activities (Table 1). As can be seen in Table 1, most of these species
are plants that are not mobile, arid are very sensitive to the expected effects of soil remediation
www. massw ildlife. org
Division of Fisheries and Wildlife
Field Headquarters, North Drive, Westborough, MA 01581 (508) 389-6300 Fax (508) 389-7891
An Agency of the Department ofFish and Game
-------
Housatonic Core Habitat Areas, 7/31/2012, page 2 of 3
activities. Core Area 1 also includes habitat for one state-listed moth species that inhabits mature
floodplain forest, one habitat area for the Jefferson's Salamander, and Triangle Floater mussel
beds. Some of the plant species found in Core Area 1 are located in floodplain forest, which is
not readily restorable and would take decades to return to its current state, if ever. Finally, Core
1 includes areas that are excellent examples of two rare natural communities—High Terrace
Floodplain Forest and Black Ash Bur Oak Hemlock Swamp.
Core Area 2 includes the highest quality habitat for more mobile species that may be less
vulnerable to remediation impacts, species where the habitat is likely to be somewhat more easily
restored, and listed species that may be of a somewhat lower conservation concern, given their
state-wide distribution (e.g. American Bittern; see Table 2). For example, the Mustard White is a
Threatened butterfly species of significant conservation concern that uses a mix of natural areas
along the river and old field habitat. It may be possible to remediate its habitat in phases,
restoring and replacing host plants as the work is completed.
Core Area 3 includes those areas with dense concentrations of state-listed species. Specifically,
Core Area 3 includes areas where Division biologists have delineated overlapping habitat for
eight (8) or more state-listed species.
Core Area 4 includes all certified vernal pools in the PSA as well as additional potential vernal
pool habitat areas which, based on information provided by GE and EPA, are likely to meet the
Massachusetts criteria for vernal pool certification based on the presence of "obligate" vernal
pool breeding amphibians see:
http://www.mass.gov/dfwele/dfw/nhesp/vernal pools/vernal pool cert.htm.
These Core Areas played an important role during recent discussions between the EPA and the
states of Massachusetts and Connecticut regarding potential remediation approaches to Rest of
River. Consistent with the requirements of MESA and the MESA Regulations, the Core Areas are
helping to guide efforts to avoid, minimize and mitigate impacts to state-listed species. Although
a final MESA evaluation will not be completed until the remedy design phase, by focusing on the
Core Areas, EPA and the Commonwealth believe that a framework has been established to
achieve MESA permitting standards of assessing alternatives to both temporary and permanent
impacts to state-listed species, and of limiting the impact to an insignificant portion of the local
populations of affected species. See 321 CMR 10.23. For example, the parties focused on
avoidance of some of the most important and sensitive rare species habitats in Core Area 1.
Similarly, in Core Areas 2 and 3, avoidance of impacts when practical, careful consideration of
PCB remediation methods and the sequence and timing of remediation activities, as well as after-
the-fact habitat mitigation are all approaches that will assist in achieving the substantive
requirements of MESA. Although the Core Areas play an important role in guiding avoidance
and minimization of impacts to state-listed species, in some cases the "take" of state-listed species
is likely to be unavoidable. In those cases, consistent with MESA's status as a location-specific
applicable or relevant and appropriate requirement ("ARAR"), the Commonwealth will work
with GE and the EPA to minimize impacts and to ensure that an adequate long-term net-benefit
mitigation plan for the affected state-listed species is designed and implemented, as required by
321 CMR 10.23(2)(c).
If you have any questions about this letter, please don't hesitate to contact me.
-------
Housatonic Core Habitat Areas, 7/31/2012, page 3 of 3
Sincerely,
^ v. k-
Jon Regosin, Ph.D.
Chief of Conservation Science
Natural Heritage & Endangered Species Program
End.: Table 1. Species and Natural Communities Included in Core Area 1 Delineation
Table 2. Species and Natural Communities Included in Core Area 2 Delineation
cc: Mark Tisa, MA Division of Fisheries & Wildlife
Richard Lehan, MA Department of Fish & Game
Mike Gorski, MA Dept. of Environmental Protection
Eva Tor, MA Dept. of Environmental Protection
Trad Iott, CT Dept. of Energy & Environmental Protection
-------
Jefferson Salamander
Reach 5A
jf 'Ostri£h Fern Borer
Narrow Leaved Spring Beauty
[CROF.UTiSTREETj
[dENOa
jMsfllElGfllQNI
Legend
Primary Study Area (PSA)
Core Area 1
These areas show the most
im porta nt/d istu rba nce-sensitive
habitat areas for state-listed species
I W00DS)R0ND'\
Date of Aerial Photo- 2009
Map produced by DFW GIS, 5/21/2012
Core Habitat Areas
Housatonic River Primary Study Area (PSA)
-------
Reach 5A
Mustard White
iMOREW.O&DlJAKEi
PiTTSFiELO
I
American Bittern
iD'ALfllO'Nl
mm
LWA'S HHI N GJiONl
WAIlEWAYjlidNDl
Legend
Primary Study Area (PSA)
Core Area 2-
These areas show the most
important/disturbance-sensitive
habitat areas for the more mobile
state-listed species
M&QD'S'iiOND]
Date of Aerial Photo- 2009
Map produced by DFW GIS, 5/21/2012
-------
Reach 5A
Wood Turtle
1t/IQ8EW.OOD]tf^KE 1
leiimsEiEiM
iD'AfrTQM
LENOX
1 W.Q'ODSJ^QNDl
Mustard White
Legend
C3
Primary Study Area (PSA)
M
Core Areas- High Species Richness
( 8 or more state-listed species)
Vernal Pool Core Areas
Date of Aerial Photo- 2009
Map produced by DFW GIS, 5/21/2012 |
Part of the Housatonic River
Showing Primary Study Area, High Species Richness, and Vernal Pools
-------
APPENDIX B
REVISED COMPARATIVE ANALYSIS OF ALTERNATIVES
-------
REVISED COMPARATIVE ANALYSIS OF ALTERNATIVES
TABLE OF CONTENTS
Section Page
1 INTRODUCTION TO EPA'S COMPARATIVE ANALYSIS OF
ALTERNATIVES 1
2 OVERALL PROTECTION OF HUMAN HEALTH AND THE ENVIRONMENT 6
3 CONTROL OF SOURCES OF RELEASES 12
4 COMPLIANCE WITH FEDERAL AND STATE ARARS 15
4.1 CHEMICAL-SPECIFIC ARARS 15
4.1.1 Chemical-Specific ARARs 15
4.1.2 Location-Specific and Action-Specific ARARs 15
5 LONG-TERM RELIABILITY AND EFFECTIVENESS 16
5.1 MAGNITUDE 01 RESIDUAL RISK 16
5.1.1 Potential Residual Risks Associated with River Sediment, Water, and
Fish 16
5.1.2 Potential Residual Risks Associated with Floodplain Soil 19
5.2 ADEQUACY AM) RELIABILITY 21
5.2.1 Use of Technologies Under Similar Conditions 21
5.2.2 General Reliability and Effectiveness 22
5.2.3 Reliability of Operation, Monitoring, and Maintenance Requirements
and Technical Component Replacement Requirements 23
5.3 POTENTIAL LONG-TERM IMPACTS ON HUMAN HEALTH AND THE
ENVIRONMENT 23
5.3.1 Potentially Affected Populations 23
5.3.2 Long-Term Impacts on Habitats and Biota 24
5.3.3 Long-Term Impacts on State-Listed Species 29
5.3.4 Long-Term Impacts on Aesthetics and Recreational Use 30
5.3.5 Long-Term Impacts on Fluvial Geomorphic Processes 30
6 ATTAINMENT OF IMPGS 31
6.1 COMPARISON TO HUMAN HEALTH IMPGS 31
6.1.1 Human Direct Contact with Floodplain Soil and Sediment 31
6.1.2 Human Consumption of Floodplain Agricultural Products 32
6.1.3 Human Consumption of Fish 32
6.2 COMPARISON TO ECOLOGICAL IMPGS 32
6.2.1 Benthic Invertebrates 33
6.2.2 Amphibians 33
6.2.3 W arm water and C ol dwater Fi sh 34
6.2.4 Insectivorous Birds 35
6.2.5 Piscivorous Birds 36
HI 8/1/2012
-------
TABLE OF CONTENTS (Continued)
6.2.6 Piscivorous Mammals 36
6.2.7 Omnivorous/Carnivorous Mammals 37
6.2.8 Threatened and Endangered Species 38
6.3 SUMMARY 38
7 REDUCTION OF TOXICITY, MOBILITY, OR VOLUME OF WASTES 40
8 SHORT-TERM EFFECTIVENESS 41
8.1 IMPACTS ON THE ENVIRONMENT - EFFECTS WITHIN THE REST OF
RIVER AREA 42
8.2 CARBON FOOTPRINT - GHG EMISSIONS 45
8.3 IMPACTS ON LOCAL COMMUNITIES AND COMMUNITIES ALONG
TRUCK TRANSPORT ROUTES 46
8.4 POTENTIAL MEASURES TO AVOID, MINIMIZE, OR MITIGATE SHORT-
TERM COMMUNITY IMPACTS 49
8.5 RISKS TO REMEDIATION WORKERS 49
9 IMPLEMENT ABILITY 50
9.1 TECHNICAL IMPLEMENT ABILITY 50
9.2 ADMINISTRATIVE IMPLEMENTABILITY 52
10 COST 52
11 COMPARATIVE ANALYSIS OF TREATMENT/DISPOSITION
ALTERNATIVES 53
11.1 OVERVIEW 01 ALTERNATIVES 53
11.2 OVERALL PROTECTION OF HUMAN HEALTH AND THE
ENVIRONMENT 54
11.3 CONTROL OF SOURCES OF RELEASES 56
11.4 COMPLIANCE WITH FEDERAL AND STATE ARARS 56
11.5 LONG-TERM RELIABILITY AND EFFECTIVENESS 57
11.5.1 Magnitude of Residual Risk 57
11.5.2 Adequacy and Reliability of Alternatives 57
11.5.3 Potential Long-Term Adverse Impacts on Human Health or the
Environment 58
11.5.4 Summary of Long-Term Reliability and Effectiveness 60
11.6 ATTAINMENT OF IY1PGS 60
11.7 REDUCTION OF TOXICITY, MOBILITY, OR VOLUME 60
11.8 SHORT-TERM EFFECTIVENESS 61
11.8.1 Impacts on the Environment 61
11.8.2 Carbon Footprint - GHG Emissions 61
11.8.3 Impacts on Local Communities 62
11.8.4 Potential Measures to Avoid, Minimize, or Mitigate Short-Term
Environmental and Community Impacts 65
IV 8/1/2012
-------
TABLE OF CONTENTS (Continued)
11.8.5 Risks to Remediation Workers 65
11.8.6 Summary of Short-Term Effectiveness 66
11.9 IV1PI.I:V1I:M ABILITY 66
11.9.1 Technical Implementability 66
11.9.2 Administrative Implementability 67
11.10 COST 67
11.11 OVERALL CONCLUSION FOR TREATMENT/DISPOSITION
ALTERNATIVES 68
LIST OF ATTACHMENTS
Attachment B-l Use of Channel Realignment along the Housatonic River for Restoration
and Remediation of PCB Contamination
Attachment B-2 Channel Dynamics and Ecological Conditions in the Housatonic River
Primary Study Area
Attachment B-3 Activated Carbon Summary
Attachment B-4 Massachusetts Division of Fisheries and Wildlife Core Habitat Area
Maps
Attachment B-5 Cap Cross Section Refinement - Layer Sizing, Rest of River - Reach 5 A
Attachment B-6 Comparison Metrics
Attachment B-7 Post-East Branch Remediation Boundary Conditions
Attachment B-8 Food Chain Model Output
Attachment B-9 Preliminary Draft ARAR Tables for SED 9/FP 4 MOD
Attachment B-10 Cost Assumptions Memorandum for SED 9/FP 4 MOD
v
8/1/2012
-------
LIST OF TABLES
Table 1 Evaluation of IMPG Attainment for Human Consumption of Fish for
Combined SED/FP Scenarios 9
Table 2 Percent Reduction in Annual PCB Mass Passing Woods Pond and Rising
Pond Dams and Transported to the Reach 5/6 Floodplain for Combinations
of Alternatives (relative to current conditions) 13
Table 3 Modeled Subreach Average Fish (Fillet) PCB Concentrations at End of
Project Period and Percent Reductions for Combinations of Alternatives 18
Table 4 Summary of Percent of Floodplain and Sediment Exposure Areas Achieving
IMPGs for Direct Human Contact 21
Table 5 Habitat Areas in Primary Study Area Affected by Combinations of
Sediment and Floodplain Alternatives 25
Table 6 Overlap of the 57.8 Acres of Floodplain Soil Requiring Remediation with
Core Areas 1 through 3 29
Table 7 Summary of Percent Benthic Invertebrate Averaging Areas Achieving
IMPGs for Benthic Invertebrates 33
Table 8 Summary of Percent of Amphibian Averaging Areas Achieving IMPGs for
Amphibians 34
Table 9 Summary of Percent of Averaging Areas Achieving Warmwater and
Coldwater Fish Protection IMPGs 35
Table 10 Summary of Percent of Averaging Areas Achieving IMPGs for
Insectivorous Birds 35
Table 11 Summary of Percent of Averaging Areas Achieving Piscivorous Bird
IMPGs 36
Table 12 Summary of Percent of Averaging Areas Achieving IMPGs for Piscivorous
Mammals 37
Table 13 Summary of Percent of Averaging Areas Achieving IMPGs for
Omnivorous/Carnivorous Mammal s 38
Table 14 Removal Volume and Corresponding PCB Mass for Combinations of
Alternatives 41
Table 15 Construction Duration for Alternative Combinations 41
Table 16 Calculated GHG Emissions Anticipated to Result from Combinations of
Sediment and Floodplain Alternatives 46
VI 8/1/2012
-------
LIST OF TABLES (Continued)
Table 17 Estimated Truck Trips for Removal of Excavated Material and Delivery of
Capping/Backfill Material for Combinations of Sediment and Floodplain
Alternatives 48
Table 18 Incidence of Accident-Related Injuries/Fatalities Due to Increased Truck
Traffic 49
Table 19 Incidence of Accident-Related Injuries/Fatalities Due to Implementation of
Sediment-Floodplain Alternative Combinations 50
Table 20 Required Capping/Backfill/Stabilization Material Volumes for
Combinations of Sediment and Floodplain Alternatives 51
Table 21 Cost Summary for Combinations of Sediment and Floodplain Alternatives 53
Table 22 Calculated GHG Emissions Anticipated to Result from
Treatment/Disposition Alternatives 62
Table 23 Estimated Off-Site Truck Trips for Treatment/Disposition Alternatives 63
Table 24 Incidence of Accident-Related Injuries/Fatalities Due to Increased Off-Site
Truck Traffic 64
Table 25 Incidence of Potential Accidents/Injuries Due to Implementation of
Alternatives TD 2 through TD 5 65
Table 26 Cost Summary for Treatment/Disposition Alternatives 69
Vll 8/1/2012
-------
LIST OF FIGURES
Figure 1 Number of Fish Meals per Year Acceptable to be Eaten by an Individual for
Noncancer HI of 1 for a CTE Adult at the End of the Modeling Period 11
Figure 2 Average Fillet PCB Concentrations in Largemouth Bass from Reach 6 20
Vlll 8/1/2012
-------
LIST OF ACRONYMS
ARAR applicable or relevant and appropriate requirement
BMP best management practice
Board National Remedy Review Board
CDF combined disposal facility
CERCLA Comprehensive Environmental Response, Compensation, and Liability Act
CFR Code of Federal Regulations
CSTAG Contaminated Sediments Technical Advisory Group
cy cubic yard
DFW Division of Fisheries and Wildlife
EA exposure area
EMNR Enhanced Monitored Natural Recovery
EPA U.S. Environmental Protection Agency
EPC exposure point concentration
ERE environmental restrictions and easements
GE General Electric Company
GWTP groundwater treatment plant
HI hazard index
IC institutional control
IMPG Interim Media Protection Goal
lb pound
MATC maximum acceptable threshold concentration
MESA Massachusetts Endangered Species Act
mg/kg milligrams per kilograms
MNR monitored natural recovery
NCP National Contingency Plan
NHESP Natural Heritage and Endangered Species Program
NRWQC National Recommended Water Quality Criteria
O&M operation and maintenance
OMM operation, monitoring, and maintenance
PCB polychlorinated biphenyl
PSA Primary Study Area
RCMS Revised Corrective Measures Study
RCRA Resource Conservation and Recovery Act
RME reasonable maximum exposure
SIP Site Information Package
tPCB total polychlorinated biphenyl
TMV toxicity, mobility, or volume
WWTP Wastewater Treatment Plant
IX
8/1/2012
-------
1
REVISED COMPARATIVE ANALYSIS OF ALTERNATIVES -
2 SUPPLEMENT TO THE NATIONAL REMEDY REVIEW BOARD
3 SITE INFORMATION PACKAGE FOR THE HOUSATONIC RIVER,
4 REST OF RIVER (JUNE 2011)
5 In July 2011, the U.S. Environmental Protection Agency (EPA) New England regional office
6 presented site information and potential cleanup strategies to the National Remedy Review Board
7 (the Board). Representatives of EPA's Contaminated Sediments Technical Advisory Group
8 (CSTAG) also coordinated with and participated in the Board review for this site. In preparation
9 for that July 2011 meeting, the regional office prepared and submitted a comprehensive Site
10 Information Package summarizing the site history, the nature and extent of contamination, risks to
11 human health and the environment posed by that contamination, potential cleanup options under
12 consideration, a proposed remediation approach, and stakeholder views on the project.
13 After the review meeting, the Board issued a set of recommendations to EPA New England,
14 dated October 20, 2011. In order to address many of the recommendations put forth by the
15 Board and to further develop a potential cleanup strategy for the Rest of River, EPA conducted
16 additional technical evaluations and worked closely with co-regulators from the Commonwealth
17 of Massachusetts and the State of Connecticut in a series of facilitated technical discussions that
18 began in October 2011. In May 2012, EPA published a status report entitled "Potential
19 Remediation Approaches to the GE-Pittsfield/Housatonic River Site 'Rest of River' PCB
20 Contamination." This status report provided an update to the public on the discussions among
21 the agencies and outlined potential remediation approaches for the Rest of River.
22 While considering the input from the Board and the States during these technical discussions,
23 EPA compiled additional technical information, conducted additional modeling work to refine
24 potential remediation approaches, and evaluated these approaches in light of the criteria outlined
25 in the Permit. All of this work has led EPA to supplement the analysis originally presented to
26 the Board. Specifically, EPA has prepared a revised Comparative Analysis of Alternatives
27 outlined below. This revised information is intended to supplement and replace the information
28 contained in Sections 9 and 11 of the June 2011 Site Information Package as well as certain
29 additional supporting documentation referenced below.
30 While no formal remedy proposal decisions have been made, this supplemental information is
31 intended to provide a more detailed summary of EPA New England's considerations regarding
32 the potential approaches to cleanup.
33 1 INTRODUCTION TO EPA'S COMPARATIVE ANALYSIS OF
34 ALTERNATIVES
35 In this section, the seven combinations of remedial alternatives for river sediment (SED) and
36 floodplain soil (FP) that were described in Section 8 of the General Electric Company (GE)
37 Revised Corrective Measures Study (RCMS), and an additional combined alternative
38 (hereinafter referred to as "SED 9/FP 4 MOD") developed by EPA in consultation with the states
39 of Massachusetts and Connecticut following their review of the RCMS, are evaluated relative to
40 each other using the evaluation criteria specified in the Reissued Resource Conservation and
41 Recovery Act (RCRA) Permit for the GE-Pittsfield/Housatonic River Rest of River Site.
1
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
SED 9/FP 4 MOD consists of the following components:
River Sediment and Banks
¦ Reach 5 A
For Reach 5A, the approximately 5-mile stretch of the Housatonic River from the
confluence of the East and West Branches of the Housatonic (at Fred Garner Park in
Pittsfield) to the Pittsfield Wastewater Treatment Plant (WWTP), SED 9/FP 4 MOD
requires the removal of river bed sediment throughout the reach and soil in
contaminated eroding riverbanks, and stabilization of contaminated erodible
riverbanks to meet cleanup levels in fish tissue, and reduce ecological risk and
downstream transport. The residual polychlorinated biphenyls (PCBs) in the riverbed
would subsequently be capped, as discussed further below. Additional data will need
to be collected before and during the implementation of the work to better quantify
the concentrations of PCBs in riverbanks and locations of erodible riverbanks and to
determine actual riverbed removal depth and cap thickness.
An important focus of the riverbank work will be to reduce bank erosion to
acceptable levels while maintaining the dynamic nature of the Housatonic River.
Wherever possible, the goal would be to leave banks intact, with no disturbance or
excavation. For banks that do need to be addressed, reconstruction and stabilization
of disturbed banks can be achieved in a number of different ways. One method is to
disturb only the more highly erodible banks and reconstruct those banks with a stone
or hard cap layer extending into the riverbank, covered with a bio-engineering "soft"
layer. A second potential approach, designed to minimize the need for the hard cap
layer extending into the riverbank, would address more riverbanks than the first
instance while reconstruction would be performed consistent with the principles of
"natural channel design." See Attachment B-l, Use of Channel Realignment Along
the Housatonic River for Restoration and Remediation of PCB Contamination, and
Attachment B-2, Channel Dynamics and Ecological Conditions in the Housatonic
River Primary Study Area, regarding this concept. Any bank restoration approach
will follow the hierarchy below of most preferred to least preferred actions:
1. Reconstruct disturbed banks with bio-engineering "soft" restoration techniques.
2. Reconstruct disturbed banks with a cap layer extending into the riverbank covered
with a bio-engineering "soft" layer.
3. Place riprap cap or hard armoring on surface of banks (for example, for protection
of adjacent infrastructure).
¦ Reach 5B
For Reach 5B, the approximately 2-mile stretch of the river from the Pittsfield
WWTP to Roaring Brook in Lenox, MA, SED 9/FP 4 MOD requires the excavation
and restoration of areas of river bed and banks that exceed the reach-specific cleanup
level of 50 milligrams per kilograms (mg/kg) and use of Enhanced Monitored Natural
Recovery (EMNR) throughout Reach 5B. EMNR involves the pilot study and review
of the effectiveness of an additive such as activated carbon or organoclay (see
2
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Attachment B-3) to sediment to reduce the bioavailability of PCBs in the sediment to
meet cleanup levels in fish tissue and reduce ecological risk and downstream
transport. Following the review of the Pilot Study results and through an adaptive
management framework, the use of additives could be applied to all of Reach 5B.
Additional data will be collected to determine PCB concentrations that exceed reach-
specific cleanup standards in the bed and banks to determine areas for targeted
removal. Any excavated Reach 5B riverbanks would be restored using bio-
engineering restoration techniques. Backfill, including a suitable habitat layer, will
be used to restore the riverbed.
¦ Reach 5C
For Reach 5C, the approximate 3-mile stretch of Housatonic River between Roaring
Brook and the headwaters of Woods Pond, SED 9/FP 4 MOD requires removal of
river bed sediment throughout the reach to meet fish tissue cleanup levels and reduce
ecological risk and downstream transport. The residual PCBs in the riverbed below
the depth of excavation would subsequently be capped, as discussed further below.
There are few eroding riverbanks in this reach. Any riverbanks in this reach will be
left intact, unless disturbed by other remediation activities.
¦ Backwaters
SED 9/FP 4 MOD requires that contaminated sediment in the majority of backwater
areas adjacent to Reaches 5 and 6 be removed to allow a cap to be placed over the
residual PCBs. Backwaters in certain areas designated as having high-quality habitat
for state-listed species (known as "Core Area 1", see Attachment B-4) will generally
not be remediated, except in discrete areas with PCB concentrations greater than 50
mg/kg. Additional data will be collected to determine PCB concentrations in the
backwaters to determine areas for targeted removal. As part of the remedy, an
evaluation will also be conducted to determine if other approaches, such as the
addition of activated carbon, would be useful in reducing the bioavailability of PCBs
in the backwaters which are not actively remediated (i.e., in those high-quality habitat
areas for state-listed species that would otherwise be avoided).
¦ Reach 6 (Woods Pond)
For Reach 6 (Woods Pond), SED 9/FP 4 MOD requires that contaminated sediment
be removed throughout the reach, which would result in meeting cleanup levels in
fish tissue and for direct contact and would reduce ecological risk and downstream
transport. An engineered multi-layered cap will be placed over residual PCBs, with
the design generally providing a minimum water depth of 6 feet and including an
appropriate configuration in the near-shore areas to replace the existing habitat. An
evaluation will be conducted to determine the appropriateness of additional deepening
and/or placement of structures within the pond to further promote sedimentation as
well as the use of an additive such as activated carbon before and/or after final cap
placement to sequester residual PCBs and minimize the potential for downstream
migration of PCBs. Reach 6 will be monitored over the long term and if substantial
PCBs accumulate in the pond, removal of the residual PCBs will be required. Final
3
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
removal depths, structures, use of additives, and engineered cap configurations will
be determined during remedial design.
¦ Columbia Mill Impoundment (Reach 7B). Eagle Mill Impoundment (Reach 7C).
Willow Mill Impoundment (Reach 7E). Glendale Impoundment (Reach 7G). and
Rising Pond (Reach 8)
This component of SED 9/FP 4 MOD addresses the impounded areas behind several
dams located in the Housatonic River south of Woods Pond. These include four
dams in Reach 7 (Columbia Mill Dam, Eagle Mill Dam, Willow Mill Dam, and
Glendale Dam (Reaches 7B, 7C, 7E, and 7G, respectively) and Rising Pond (Reach
8). To meet cleanup levels in fish tissue and reduce direct contact risk, ecological
risk, and downstream transport, contaminated sediment would be removed from the
riverbed prior to placement of a multi-layered cap to sequester remaining
contamination. Final removal depths and engineered cap configurations will be
determined during remedial design. In addition, to the extent that dam removal is
considered in the future, SED 9/FP 4 MOD includes a secondary or "contingency"
remedy allowing for the excavation of the sediment in the impoundment behind
Columbia Mill Dam or other Reach 7 impoundments to a level of 1 mg/kg PCBs to
better dovetail the remedy with dam removal projects. Dam removal itself is not a
component of SED 9/FP 4 MOD and would be conducted by others in coordination
with GE and appropriate agencies.
¦ Flowing Subreaches in Reach 7 (Reaches 7A. 7D. 7F. 7H) and Reaches 9 through 16
Monitored natural recovery (MNR) would be implemented in the flowing sub-reaches
in Reach 7 (between Woods Pond and Rising Pond) as well as Reaches 9 through 16
(from Rising Pond Dam through Connecticut). MNR would include monitoring to
confirm progress toward achieving cleanup levels in fish tissue and reduce ecological
risk and downstream transport, compliance with state and National Recommended
Water Quality Criteria (NRWQC) (to the extent not waived), and to support
modifications to fish consumption advisories.
Engineered Cap Design
Several components of SED 9/FP 4 MOD require construction of an engineered cap
following sediment removal. In each area to be capped, sediment would be removed to
allow the placement of an engineered cap to the final grades determined to be appropriate
during design of the remedy and to result in no net loss of flood storage capacity. Each
cap will likely consist of an isolation layer to minimize PCB migration up through the
cap, a protective layer (to prevent disruption and erosion of the isolation layer and
exposure of the underlying contaminated sediment), and a habitat layer. During remedial
design it will be determined if additional cap components are necessary (e.g., a filter
layer or a mixing layer) or other cap configurations are appropriate (see Attachment B-5).
In Reaches 6, 7, and 8, the cap will likely consist of an additive such as activated carbon
or organoclay and will be designed to withstand scouring and include a habitat layer.
During remedial design, it will be determined if additional cap components are necessary
or other cap configurations are appropriate. As outlined above, if dam removal activities
take place in the Reach 7 impoundments, sediment contaminated with PCBs at levels
4
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
greater than 1 mg/kg could be removed as another approach to remediation, thus making
the installation of a cap in those areas unnecessary.
Floodplain Adjacent to Reaches 5 through 8
This part of SED 9/FP 4 MOD would be performed in the floodplain while the adjacent
sediment cleanup activities (described above) are taking place and includes:
¦ Gathering additional information to support the final cleanup design and to achieve
cleanup levels.
¦ Avoiding, minimizing, or mitigating impacts to state-listed species and habitats, as
identified by the state. These areas are referred to as "Core Areas" as designated by
the Massachusetts Natural Heritage and Endangered Species Program (see
Attachment B-4).
¦ Removing floodplain soil contaminated above cleanup levels to a depth of 1 foot,
except in frequently used subareas, which will be excavated to 3 feet.
¦ Remediating certain vernal pools, based on whether or not they are in Core Areas, to
meet the ecological-based amphibian cleanup level. This work will be implemented
using an adaptive management framework, beginning with a sub-set of vernal pools.
Concurrently, other means to reduce the bioavailability of PCBs in vernal pools will
be investigated and pilot tested. Based on the outcome of the remediation of the
initial set of vernal pools, other investigations and pilot testing, the location of the
vernal pools and associated habitat, determinations will be made about how and
where additional vernal pool remediation will occur.
¦ Restoring the excavated floodplain areas, access roads, and staging areas.
As noted in bullet 3 above, during the human health risk assessment, EPA determined that
certain areas of the floodplain constituted "frequently used subareas" that were subject to
more intense use patterns than other areas and thus are proposed to undergo more cleanup
than required for other direct contact exposure pathways.
Additional SED 9/FP 4 MOD Remedy Components
The SED 9/FP 4 MOD alternative would also include long-term monitoring, maintenance,
inspection, periodic reviews, and institutional controls (ICs).
Evaluation of Sediment/Floodplain Alternatives
These combined alternatives (from Section 8 in the RCMS and SED 9/FP 4 MOD) were
selected to represent the full range of potential approaches to address contamination at the
GE-Pittsfield/Housatonic River Site. The alternatives were compared using a variety of
quantitative, semi-quantitative, and qualitative metrics (see Attachment B-6) so that the
principal advantages and disadvantages of each combination of alternatives are identified.
This approach to analyzing alternatives is designed to provide sufficient information to
adequately compare the alternatives, and select an appropriate remedy for the site, which is
best suited to meet the general standards in consideration of the decision factors (specified in
the Reissued RCRA Permit), including a balancing of those factors against one another.
5
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
The criteria for evaluation of remedial alternatives for the Rest of River are specified in Part
II, Section G, of the Reissued RCRA Permit for the GE-Pittsfield/Housatonic River Site
(Appendix G to the Consent Decree) and are similar, but not identical to, evaluation criteria
delineated in the National Contingency Plan (NCP), 40 Code of Federal Regulations (CFR)
Section 300.430(e)(9)(iii). The nine evaluation criteria include three general standards, and
six selection decision factors:
¦ General standards:
- Overall protection of human health and the environment.
- Control of sources of releases.
- Compliance with federal and state applicable or relevant and appropriate
requirements (ARARs).
¦ Selection decision factors:
- Long-term reliability and effectiveness.
- Attainment of Interim Media Protection Goals (IMPGs).
- Reduction of toxicity, mobility, or volume (TMV) of wastes.
- Short-term effectiveness.
- Implementability.
- Cost.
Each of these nine criteria is evaluated with respect to the degree to which it is achieved by the
eight selected combinations of SED and FP alternatives in Sections 2 through 10 below.
An overview and a comparative analysis of treatment/disposition alternatives are presented in
Section 11. The nine criteria for the treatment/disposal alternative analysis are the same as
described above for river sediment and floodplain soil.
2 OVERALL PROTECTION OF HUMAN HEALTH AND THE
ENVIRONMENT
The evaluation of whether a particular combination of sediment and floodplain alternatives
would provide overall human health and environmental protection relies heavily on the
evaluations under several other permit criteria, notably the following: (1) comparison to IMPGs,
(2) compliance with ARARs, (3) long-term reliability and effectiveness (including potential
long-term adverse impacts), and (4) short-term effectiveness. A summary of the comparative
evaluation of the alternative combinations considering these factors is presented below.
SED 2/FP 1 (MNR in all reaches of the river and no action in the floodplain) is the least
protective alternative, relying on natural recovery processes to achieve reductions in PCB
concentrations in sediment, surface water, and fish tissue, and a reduction in PCB loading to the
river and PCB transport to the floodplain. This alternative would also include monitoring and
ICs throughout the river in both Massachusetts and Connecticut. Given the persistence and
unsafe concentrations of PCBs in floodplain soil, riverbanks, sediment, and biota in many
reaches, the continual input of PCBs from eroding banks and channel incision into the
floodplain, and the uncertainties surrounding the effective implementation of ICs over miles of
river for centuries, this alternative is not protective.
6
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
The other combinations of sediment and floodplain alternatives would result in reductions in
PCB concentrations and potential exposures by permanently removing PCB-containing
sediment, removing and stabilizing riverbank soil, capping certain areas of the river, and
removing PCB-containing floodplain soil. These alternatives offer varying degrees of protection
as well as short-term disturbance. All alternatives include MNR and ICs for the flowing
subreaches in Reach 7 and in Reaches 9 through 16.
SED 10/FP 9 would result in a slight improvement over MNR with selective removal of
sediment in Reach 5A and some bank stabilization and limited floodplain soil removal. This
results in some reduction in the mass of PCBs transported through the system and a marginal
improvement in fish tissue concentrations. In the floodplain, the soil removal results in the
reasonable maximum exposure (RME) human health risks below a hazard index (HI) of 1 and a
10"4 cancer risk. Some ecological IMPGs are achieved in some areas of the floodplain and river.
This alternative has reduced short-term impacts but is questionable in its long-term effectiveness.
SED 3/FP 3 includes remediation of all of Reach 5 A, but relies on MNR and ICs in Reach 5B, a
portion of Reaches 5C, 5D, and 7 impoundments, and on thin-layer capping in a portion of
Reach 5C and in Reach 6. This alternative offers marginal improvements in PCB mass
transported through the system and in fish tissue concentrations when compared to SED 10/FP 9,
and achieves the RME 10"6 risk for one sediment exposure area (EA). The upper-bound
ecological IMPGs are achieved. Human health risks for direct contact in the floodplain are
below an HI of 1 and achieve 10"4 for the RME individual. In addition, the RME 10"5 risk level
is achieved in the frequently used subareas. This alternative also has reduced short-term impacts
and uncertain long-term effectiveness.
The combination alternatives, SED 5/FP 4, SED 6/FP4, SED 8/FP 7, SED 9/FP 8, and SED 9/FP
4 MOD, allow an evaluation of varying techniques and amounts of removal and capping. SED
5/FP 4 and SED 6/FP 4 include some components of capping without removal and thin-layer
capping. Capping without removal will impact the bathymetry and hydrodynamics of the river,
and thin-layer capping is not a suitable alternative with the mass and high concentrations of
PCBs in the sediment and is not expected to have a significant long-lasting impact in the reaches
for which it is considered. The model predictions for the annual mass of PCBs transported
through the system is similar for all of these alternatives, as are the predicted fish tissue
concentrations. While SED 8/FP 7 removes the vast majority of the PCBs from the river and a
significant amount from the floodplain, it is projected to take approximately 50 years to
implement, thus the improvements are not realized at the same time scale as the other
alternatives.
For the floodplain, these combinations would involve removal of progressively more PCB-
containing soil, in increasing order of removal, SED 9/FP 4 MOD, SED 5/FP 4, SED 6/FP 4,
SED 9/FP 8, and finally, SED 8/FP 7. Consequently, there would be progressively greater
reduction in exposure and risk to human health and ecological receptors, yet with associated
increasing impacts to floodplain habitat and potential adverse impacts to habitat supporting state-
listed species. The floodplain component of SED 9/FP 4 MOD was developed specifically with
these adverse impacts in mind and represents a balance between reducing risks to humans and
animals by removing PCBs from the floodplain, and the considerable unavoidable and
potentially long-lasting impacts to Core Area habitats that would result from remediation
contemplated with the more extensive alternatives. This alternative will achieve a human health
7
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
direct contact level of 10"5 or an HI of 1 in many areas, yet avoids conducting remediation in
Core Area 1 unless necessary to achieve an HI of 1 or 10"4.
To evaluate the PCB concentrations in fish tissue and consumption of fish, a computer modeling
framework was used to predict fish tissue concentrations during and following the
implementation of an alternative. The boundary conditions used for this model framework
reflect the cleanup that occurred in the upstream reaches (see Attachment B-7). The output from
the model is included in Attachment B-8.
The fish tissue PCB concentrations predicted to result from all sediment-floodplain combinations
at the end of the model simulation period (52 to -80 years) would not achieve concentrations that
meet RME IMPGs in all reaches (Table 1). As a result, under all combinations, ICs (fish
consumption advisories) would likely be needed for a period of time following remediation to
provide human health protection from fish consumption. However, a number of alternatives
result in allowable fish consumption at different risk levels (e.g., the probabilistic central
tendency exposure [CTE] noncancer risk HI concentrations are achieved in all reaches within the
modeling period). The time to achieve various risk levels is also of concern when evaluating the
alternatives. Performance of the alternatives for all risk levels is shown in Attachment B-8.
Note that for many alternatives shown in these figures, upon completion of the remediation, the
lines converge at a particular concentration (which varies by reach) and flattens out with a slight
downward trajectory over time. This is, in the most part, driven by the non-zero PCB boundary
conditions specified in the model, and therefore, is uncertain. If the boundary PCB loads are less
than were assumed, the fish tissue concentrations would continue to decline; however, if the
boundary PCB loads are greater than assumed, the point of convergence would be at a higher
tissue concentration.
As another way of assessing relative performance, the number of fish meals per year that could
be eaten by an individual (based on the assumptions in the risk assessment) and still remain
within selected risk levels was calculated for each alternative and reach. As an example, the
results for the noncancer HI of 1 for a CTE adult using the probabilistic risk calculations are
shown in Figure 1. This could be regarded as an interim goal for results following the
conclusion of remediation, with an increase in allowable meals to be achieved over time.
Although the number of meals varies for other risk levels, the overall pattern of the alternatives
relative to each other remains generally consistent. These results indicate that under SED 2/FP 1
(MNR) and SED 10/FP 9, even at the conclusion of remediation, fewer than three fish meals
would result in an exceedance of an HI of 1 for the adult CTE receptor in any reach under
consideration for remediation. For SED 3, the number of allowable meals increases to more than
10 per year in Reach 5A, but remains less than 10 for all other reaches. The remaining
alternatives, which involve more extensive sediment remediation, achieve much better
performance, with SED 8/FP 7 allowing over 25 meals per year in some reaches1.
1 As a point of reference, the assumption in the Human Health Risk Assessment was that the RME individual
consumes 50 fish meals per year, and the CTE individual consumes 7 fish meals per year from the Housatonic
River.
8
8/1/2012
-------
Table 1
Evaluation of IMPG Attainment for Human Consumption of Fish for Combined SED/FP Scenarios
Average Fish (fillet) Concentrations (mg/kg)1,!
10"6 Cancer Risk
10"6 Cancer Risk
104 Cancer Risk
Non-Cancer: Child
Non-Cancer: Adult
00
ON
00
On
00
ON
00
ON
00
ON
00
ON
River
£
s
s
&
s
S
1
| P
» o
hi
s
s
s
e
s
S
1
| P
s o
8!
s
s
s
e
s
s
1
| P
s o
8!
s
s
s
e
s
S
1
| P
s o
8!
s
s
s
s
S
1
| P
s o
81
s
s
s
e
s
S
1
| P
s o
Reach
%
p
a
p
« i,
a
P
a £
a
P
« i,
$
P
«m,
a
P
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
Human Consumption of Fish (Bass Fillets, Deterministic RME)
5A
7.3
0.25
0.26
0.26
0.17
0.31
4.2
0.26
>250
237
249
230
>250
234
>250
>52
>250
149
156
14
188
15
>250
>52
>250
62
64
62
74
68
>250
>52
>250
137
14
134
17
140
>250
>52
>250
105
109
103
129
109
>250
>52
5B
9.3
3
0.23
0.22
0.15
0.27
6.6
3.48
>250
>250
>250
235
>250
232
>250
>52
>250
>250
159
145
186
148
>250
>52
>250
>250
59
56
70
63
>250
>52
>250
>250
146
133
170
136
>250
>52
>250
>250
108
99
125
104
>250
>52
5 C
7.4
1.8
0.17
0.16
0.11
0.18
5.8
0.82
>250
>250
>250
24.
>250
229
>250
>52
>250
>250
159
14:
179
139
>250
>52
>250
207
44
44
48
>250
>52
>250
>250
14:
129
16
127
>250
>52
>250
>250
100
92
11.
93
>250
>52
5D
9.5
6.3
0.36
0.35
0.29
0.41
11
1.1
>250
>250
>250
>250
>250
IT
>250
>52
>250
195
>250
>250
>250
IT
>250
>52
>250
138
>250
>250
117
IT
>250
>52
>250
187
>250
>250
>250
IT
>250
>52
>250
165
>250
>250
>250
IT
>250
>52
6 (WP)
8.6
0.71
0.18
0.17
0.13
0.16
3.7
0.74
>250
>250
>250
>250
>250
231
>250
>52
>250
>250
187
17
193
13
>250
>52
>250
>250
50
48
1 -
>250
>52
>250
>250
168
153
174
125
>250
>52
>250
>250
11
106
12.
89
>250
>52
7A
6.4
1.3
0.42
0.4
0.34
0.42
4.2
1.12
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
233
138
112
166
120
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
207
>250
219
>250
>52
7B
5.7
2.1
1.6
0.41
0.1
0.21
4.2
0.67
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
205
>250
>250
>52
>250
>250
>250
>250
46
60
>250
>52
>250
>250
>250
>250
18.
245
>250
>52
>250
>250
>250
>250
11
164
>250
>52
7C
6.3
1.8
1
0.2
0.12
0.2
4.4
0.81
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
181
200
171
>250
>52
>250
>250
>250
53
52
52
>250
>52
>250
>250
>250
164
180
155
>250
>52
>250
>250
>250
116
123
110
>250
>52
ID
5.5
1.4
0.79
0.7
0.63
0.75
3.7
1.37
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
210
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
IE
4.1
1
0.57
0.34
0.18
0.22
2.8
0.64
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
213
>250
209
>250
>52
>250
154
173
83
64
61
>250
>52
>250
>250
>250
195
>250
189
>250
>52
>250
224
>250
146
174
133
>250
>52
7F
3.2
0.82
0.49
0.45
0.38
0.45
2.2
0.82
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
195
165
128
182
140
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
228
>250
>250
>250
>52
7G
3.5
1.3
1
0.4
0.15
0.22
2.6
0.38
0.0019
>250
>250
>250
>250
>250
>250
>250
>52
0.019
>250
>250
>250
>250
>250
>250
>250
>52
0.19
>250
>250
>250
154
52
63
>250
>52
0.026
>250
>250
>250
>250
>250
232
>250
>52
0.062
>250
>250
>250
>250
176
158
>250
>52
7H
2.8
0.72
0.43
0.39
0.35
0.39
1.9
0.69
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
219
174
139
226
147
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
8 (RP)
3.6
1.6
0.34
0.22
0.17
0.24
2.7
0.37
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
65
63
72
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
177
204
182
>250
>52
BBD
0.16
0.04
0.01
0.009
0.006
0.009
0.1
0.022
>250
244
126
91
116
101
>250
>52
230
94
40
36
60
34
246
>52
31
11
11
18
15
13
17
10
203
74
27
28
56
25
210
37
128
22
21
22
34
16
111
19
LL
0.11
0.03
0.009
0.006
0.005
0.006
0.08
0.015
>250
222
113
82
106
90
>250
>52
200
72
33
31
57
26
207
36
26
9
8
9
11
11
9
8
173
52
24
25
55
-
28
98
17
19
20
31
15
72
16
LZ
0.08
0.02
0.006
0.005
0.004
0.004
0.05
0.011
>250
199
99
7:
96
7
>250
>52
170
49
25
26
56
23
16
27
6
6
4
6
7
9
6
0
14:
34
22
23
54
18
13
21
68
12
15
19
17
13
27
12
LH
0.08
0.02
0.006
0.004
0.003
0.004
0.05
0.010
>250
197
97
72
94
77
>250
>52
167
46
25
26
56
22
162
26
5
5
4
5
6
8
4
0
140
27
22
23
41
18
126
20
65
12
11
19
17
13
26
12
Human Consumption of Fish (Bass Fillets, Probabilistic RME (5th percentile))
5A
7.3
0.25
0.26
0.26
0.17
0.31
4.2
0.26
>250
19
200
18
242
190
>250
>52
>250
103
108
102
127
108
>250
>52
240
15
15
15
17
19
18
17
>250
106
11
105
13
111
>250
>52
>250
80
82
79
9c
85
>250
>52
5B
9.3
3
0.23
0.22
0.15
0.27
6.6
3.48
>250
>250
207
188
242
188
>250
>52
>250
>250
106
98
123
103
>250
>52
>250
213
16
16
20
15
>250
>52
>250
>250
110
101
128
106
>250
>52
>250
>250
79
74
91
80
>250
>52
5 C
7.4
1.8
0.17
0.16
0.11
0.18
5.8
0.82
>250
>250
213
190
241
181
>250
>52
>250
>250
98
91
110
91
>250
>52
>250
123
19
20
31
14
>250
>52
>250
>250
102
94
114
94
>250
>52
>250
239
67
63
76
67
500
>52
5D
9.5
6.3
0.36
0.35
0.29
0.41
11
1.1
>250
221
>250
>250
>250
IT
>250
>52
>250
16.
>250
>250
>250
IT
>250
>52
>250
108
21
21
31
15
239
>52
>250
167
>250
>250
>250
IT
>250
>52
>250
149
>250
>250
17:
IT
>250
>52
6 (WP)
8.6
0.71
0.18
0.17
0.13
0.16
3.7
0.74
>250
>250
>250
229
>250
182
>250
>52
>250
>250
114
105
120
88
>250
>52
>250
79
22
23
41
16
189
>52
>250
>250
119
109
125
91
>250
>52
>250
>250
75
71
82
62
>250
>52
7A
6.4
1.3
0.42
0.4
0.34
0.42
4.2
1.12
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
205
>250
216
>250
>52
>250
117
24
24
43
120
235
>52
>250
>250
>250
211
>250
223
>250
>52
>250
>250
188
151
234
161
>250
>52
7B
5.7
2.1
1.6
0.41
0.1
0.21
4.2
0.67
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
114
162
>250
>52
>250
232
>250
23
43
15
>250
>52
>250
>250
>250
>250
120
169
>250
>52
>250
>250
>250
>250
70
103
>250
>52
7C
6.3
1.8
1
0.2
0.12
0.2
4.4
0.81
>250
>250
>250
24.
>250
228
>250
>52
>250
>250
>250
11
121
108
>250
>52
>250
197
16
23
44
16
>250
>52
>250
>250
>250
118
12
112
>250
>52
>250
>250
>250
79
82
76
>250
>52
7D
5.5
1.4
0.79
0.7
0.63
0.75
3.7
1.37
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
142
83
62
76
74
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
7E
4.1
1
0.57
0.34
0.18
0.22
2.8
0.64
>250
>250
>250
>250
>250
>250
>250
>52
>250
222
>250
14
171
13
>250
>52
232
79
38
23
44
16
224
>52
>250
227
>250
149
179
136
>250
>52
>250
183
223
109
11C
91
>250
>52
7F
3.2
0.82
0.49
0.45
0.38
0.45
2.2
0.82
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
225
>250
251
>250
>52
205
75
25
26
45
-
>250
>52
>250
>250
>250
232
>250
>250
>250
>52
>250
240
219
169
>250
187
>250
>52
7G
3.5
1.3
1
0.4
0.15
0.22
2.6
0.38
0.0064
>250
>250
>250
>250
>250
>250
>250
>52
0.064
>250
>250
>250
>250
172
155
>250
>52
0.64
243
156
142
23
45
16
>250
16
0.059
>250
>250
>250
>250
182
162
>250
>52
0.12
>250
>250
>250
218
101
102
>250
>52
7H
2.8
0.72
0.43
0.39
0.35
0.39
1.9
0.69
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
188
66
23
23
46
16
>250
>52
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
242
196
>250
210
>250
>52
8 (RP)
3.6
1.6
0.34
0.22
0.17
0.24
2.7
0.37
>250
>250
>250
>250
>250
>250
>250
>52
>250
>250
>250
173
200
179
>250
>52
233
190
22
24
53
18
>250
19
>250
>250
>250
181
210
187
>250
>52
>250
>250
>250
111
120
117
>250
>52
BBD
0.16
0.04
0.01
0.009
0.006
0.009
0.1
0.022
>250
165
>250
59
79
60
>250
>52
125
22
21
22
34
16
107
18
0
0
0
0
0
0
0
0
132
26
21
22
39
17
116
19
71
11
11
18
15
13
17
13
LL
0.11
0.03
0.009
0.006
0.005
0.006
0.08
0.015
>250
14:
64
51 | 68
50
>250
>52
96
17
19
20
31
15
68
15
0
0
0
0
0
0
0
0
103
19
20
21
32
16
7
16
26
9
8
9
11
11
9
"\
LZ
0.08
0.02
0.006
0.005
0.004
0.004
0.05
0.011
>250
120
51
38
62
36
>250
>52
66
12
15
19
17
13
27
12
0
0
0
0
0
0
0
0
73
14
18
20
26
14
41
13
6
6
4
6
7
9
6
9
LH
0.08
0.02
0.006
0.004
0.003
0.004
0.05
0.010
>250
117
50
37
62
35
>250
>52
62
12
11
19
17
13
26
12
0
0
0
0
0
0
0
0
69
13
18
19
22
14
37
12
5
5
4
5
6
8
4
9
Human Consumption of Fish (Bass Fillets, Deterministic CTE)
5A
7.3
0.25
0.26
0.26
0.17
0.31
4.2
0.26
>250
11:
118
11
141
11
>250
>52
>250
22
22
22
23
35
205
26
82
8
8
8
10
8
36
9
>250
62
64
62
74
68
>250
>52
>250
26
26
26
39
38
21
33
5B
9.3
3
0.23
0.22
0.15
0.27
6.6
3.48
>250
>250
118
109
137
113
>250
>52
>250
241
18
18
21
22
>250
>52
123
12
10
10
14
9
81
16
>250
>250
59
56
70
63
>250
>52
>250
>250
21
20
23
34
>250
>52
5 C
7.4
1.8
0.17
0.16
0.11
0.18
5.8
0.82
>250
>250
111
102
125
102
>250
>52
>250
142
20
20
32
15
>250
>52
98
10
14
14
17
10
69
11
>250
207
44
44
48
>250
>52
>250
151
20
21
32
16
>250
>52
5D
9.5
6.3
0.36
0.35
0.29
0.41
11
1.1
>250
171
>250
>250
>250
IT
>250
>52
>250
115
21
21
31
16
>250
>52
136
58
17
17
27
12
108
12
>250
138
>250
>250
117
IT
>250
>52
>250
118
22
22
32
24
>250
>52
6 (WP)
8.6
0.71
0.18
0.17
0.13
0.16
3.7
0.74
>250
>250
130
119
136
99
>250
>52
>250
13
22
23
42
16
209
>52
132
11
18
19
37
12
25
12
>250
>250
50
48
51
44
>250
>52
>250
161
23
24
42
17
219
>52
7A
6.4
1.3
0.42
0.4
0.34
0.42
4.2
1.12
>250
>250
>250
227
>250
240
>250
>52
>250
142
36
33
44
37
>250
>52
78
9
10
10
12
11
26
12
>250
233
138
112
166
120
>250
>52
>250
155
48
41
48
48
>250
>52
7B
5.7
2.1
1.6
0.41
0.1
0.21
4.2
0.67
>250
>250
>250
>250
134
186
>250
>52
>250
>250
>250
23
43
16
>250
>52
69
9
10
10
12
11
26
11
>250
>250
>250
>250
46
60
>250
>52
>250
>250
>250
23
43
16
>250
>52
7C
6.3
1.8
1
0.2
0.12
0.2
4.4
0.81
>250
>250
>250
129
138
122
>250
>52
>250
234
227
24
45
17
>250
>52
78
10
10
10
13
12
36
12
>250
>250
>250
53
52
52
>250
>52
>250
>250
>250
24
45
18
>250
>52
7D
5.5
1.4
0.79
0.7
0.63
0.75
3.7
1.37
>250
>250
>250
>250
>250
>250
>250
>52
>250
174
124
94
127
114
>250
>52
64
9
9
10
12
11
11
12
>250
>250
>250
210
>250
>250
>250
>52
>250
189
144
110
153
134
>250
>52
7E
4.1
1
0.57
0.34
0.18
0.22
2.8
0.64
>250
239
>250
159
197
148
>250
>52
>250
96
69
24
45
17
>250
>52
34
9
7
9
11
11
9
11
>250
154
173
83
64
61
>250
>52
>250
104
84
24
45
17
>250
>52
7F
3.2
0.82
0.49
0.45
0.38
0.45
2.2
0.82
>250
>250
>250
249
>250
>250
>250
>52
231
102
51
41
48
39
>250
>52
9
8
6
8
10
10
8
11
>250
195
165
128
182
140
>250
>52
244
114
68
55
61
56
>250
>52
7G
3.5
1.3
1
0.4
0.15
0.22
2.6
0.38
0.049
>250
>250
>250
>250
203
178
>250
>52
0.49
>250
196
193
24
46
17
>250
31
4.9
10
8
6
8
11
10
8
9
0.19
>250
>250
>250
154
52
63
>250
>52
0.43
>250
216
218
24
47
18
>250
35
7H
2.8
0.72
0.43
0.39
0.35
0.39
1.9
0.69
>250
>250
>250
>250
>250
>250
>250
>52
214
99
34
26
47
22
>250
>52
7
7
5
7
8
9
6
10
>250
219
174
139
226
147
>250
>52
226
116
51
37
48
35
>250
>52
8 (RP)
3.6
1.6
0.34
0.22
0.17
0.24
2.7
0.37
>250
>250
>250
200
234
205
>250
>52
>250
231
23
25
53
19
>250
30
10
8
6
8
11
11
8
9
>250
>250
>250
65
63
72
>250
>52
>250
>250
24
25
54
19
>250
33
BBD
0.16
0.04
0.01
0.009
0.006
0.009
0.1
0.022
148
37
22
23
54
19
13
22
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
31
11
"
,s
15
13
17
10
0
0
0
0
0
0
0
0
LL
0.11
0.03
0.009
0.006
0.005
0.006
0.08
0.015
119
23
21
22
36
17
99
18
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
26
9
8
9
11
11
9
8
0
0
0
0
0
0
0
0
LZ
0.08
0.02
0.006
0.005
0.004
0.004
0.05
0.011
89
17
19
20
31
15
.
15
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
6
6
4
6
7
9
6
0
0
0
0
0
0
0
0
0
LH
0.08
0.02
0.006
0.004
0.003
0.004
0.05
0.010
85
17
19
20
31
15
54
14
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
5
5
6
8
*
0
0
0
0
0
0
0
0
0
9
-------
Table 1
Evaluation of IMPG Attainment for Human Consumption of Fish for Combined SED/FP Scenarios
Average Fish (fillet) Concentrations (mg/kg)1,!
10"6 Cancer Risk
10"6 Cancer Risk
104 Cancer Risk
Non-Cancer: Child
Non-Cancer: Adult
00
ON
00
On
00
ON
00
ON
00
ON
00
ON
River
£
s
s
&
s
S
1
| P
s o
hi
s
s
s
e
s
S
1
| P
s o
8!
s
s
s
e
s
s
1
| P
s o
8!
s
s
s
e
s
S
1
| P
s o
8!
s
s
s
e
s
S
1
| P
s o
81
s
s
s
e
s
S
1
| P
s o
Reach
%
p
6
a
6
6
p
« i,
a
P
a &
a
P
« i,
a
P
a
P
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
VI
Human Consumption of Fish (Bass Fillets, Probabilistic CTE (50th percentile))
5A
7.3
0.25
0.26
0.26
0.17
0.31
4.2
0.26
>250
108
112
106
133
11
>250
>52
249
18
18
18
18
23
19
21
71
7
7
7
9
7
26
9
232
14
14
14
16
16
179
15
174
11
11
11
13
lo\ 125
12
5B
9.3
3
0.23
0.22
0.15
0.27
6.6
3.48
>250
>250
111
103
129
107
>250
>52
>250
225
17
17
21
18
>250
>52
107
11
10
10
13
9
65
13
>250
202
16
16
20
14
>250
>52
>250
124
14
14
18
11 1 203
>52
5 C
7.4
1.8
0.17
0.16
0.11
0.18
5.8
0.82
>250
>250
104
96
116
96
>250
>52
>250
131
19
20
31
14
>250
>52
81
10
11
11
14
9
54
10
>250
116
19
19
31
14
>250
>52
226
65
18
18
28
12
193
14
5D
9.5
6.3
0.36
0.35
0.29
0.41
11
1.1
>250
167
>250
>250
>250
IT
>250
>52
>250
111
21
21
31
15
341
>52
122
54
17
17
27
"
12
>250
105
20
21
31
15
320
>52
249
87
19
19
29
13
249
35
6 (WP)
8.6
0.71
0.18
0.17
0.13
0.16
3.7
0.74
>250
>250
121
11
127
9:
198
>52
>250
103
22
23
41
16
5.
>52
113
11
17
19
37
11
—.
10
12
>250
53
22
23
41
15
180
>52
>250
14
20
21
40
14
12
15
7A
6.4
1.3
0.42
0.4
0.34
0.42
4.2
1.12
>250
>250
>250
214
>250
226
>250
>52
>250
128
25
25
43
25
246
>52
63
9
9
9
11
11
11
12
>250
107
24
24
42
17
225
>52
192
26
21
22
39
14
152
18
7B
5.7
2.1
1.6
0.41
0.1
0.21
4.2
0.67
>250
>250
>250
>250
122
172
>250
>52
>250
250
>250
23
43
16
>250
>52
52
9
9
9
11
11
11
10
>250
217
238
22
42
15
>250
35
201
103
63
21
41
14
193
14
7C
6.3
1.8
1
0.2
0.12
0.2
4.4
0.81
>250
>250
>250
120
128
114
>250
>52
>250
213
192
24
44
17
>250
>52
62
9
9
9
12
11
12
12
>250
182
141
23
44
16
>250
>52
202
76
23
22
42
14
177
16
ID
5.5
1.4
0.79
0.7
0.63
0.75
3.7
1.37
>250
>250
>250
>250
>250
>250
>250
>52
>250
155
101
76
98
91
>250
>52
38
9
8
9
11
11
9
11
>250
129
68
51
60
58
>250
>52
206
38
22
22
42
15
180
33
IE
4.1
1
0.57
0.34
0.18
0.22
2.8
0.64
>250
229
>250
151
182
138
>250
>52
243
86
52
23
44
16
>250
>52
11
8
6
8
10
10
8
10
222
73
27
23
44
15
212
35
149
23
21
21
40
14
124
14
7F
3.2
0.82
0.49
0.45
0.38
0.45
2.2
0.82
>250
>250
>250
235
>250
>250
>250
>52
216
87
36
32
46
30
>250
>52
7
7
5
7
8
9
7
10
195
65
24
24
44
17
236
>52
122
19
20
21
33
14
113
16
7G
3.5
1.3
1
0.4
0.15
0.22
2.6
0.38
0.057
>250
>250
>250
>250
186
165
>250
>52
0.57
>250
173
164
24
46
17
>250
28
5.7
8
7
5
7
9
9
7
8
0.71
231
140
122
23
45
16
>250
15
1.5
145
34
21
21
42
14
158
13
7H
2.8
0.72
0.43
0.39
0.35
0.39
1.9
0.69
>250
>250
>250
>250
>250
>250
>250
>52
199
80
24
24
46
17
>250
>52
4
5
3
5
7
8
3
9
178
52
22
23
45
16
>250
38
106
15
19
20
32
13
99
14
8 (RP)
3.6
1.6
0.34
0.22
0.17
0.24
2.7
0.37
>250
>250
>250
185
215
190
>250
>52
246
208
23
24
53
18
>250
23
7
7
5
7
9
10
6
8
222
174
22
24
52
18
>250
18
141
58
20
22
50
16
182
15
BBD
0.16
0.04
0.01
0.009
0.006
0.009
0.1
0.022
135
26
21
22
39
17
12
20
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
LL
0.11
0.03
0.009
0.006
0.005
0.006
0.08
0.015
106
19
20
21
32
16
81
16
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
LZ
0.08
0.02
0.006
0.005
0.004
0.004
0.05
0.011
75
14
18
20
26
14
41
14
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
LH
0.08
0.02
0.006
0.004
0.003
0.004
0.05
0.010
72
13
18
19
22
14
37
13
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Key:
Notes:
= post-remediation EPC is higher than IMPG
= post-remediation EPC is lower than IMPG
= time to achieve predicted by the model
= time to achieve based on highly uncertain extrapolation of the model results as described in Section 3.2.1 of the Revised CMS Report
Model endpoint concentrations after projection (autumn average); whole body concentrations divided by a factor of 5.0 to convert to fillet basis
2 Results for CT impoundments are highly uncertain as they were estimated from the CT 1-D Analysis.
CTE = central tendency exposure
RME = reasonable maximum exposure
BBD: Bulls Bridge Dam Impoundment; LL: Lake Lillinonah; LZ: Lake Zoar; LH: Lake Housatonic
10
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
_Q
O
13
T3
<
rH
_h
I
at
CL
SED 1/2 SED 3 SED 4
SED 5 SED6 SED7
Alternative
SED 8
SED 9
SED 10
Figure 1 Number of Fish Meals per Year Acceptable to be Eaten by an Individual
for Noncancer HI of 1 for a CTE Adult at the End of the Modeling Period2
Estimates from the Connecticut one-dimensional (1-D) analysis indicate that RME 10""/HI = 1
deterministic IMPGs are achieved in two of the four impoundments modeled in Connecticut
under SED 2/FP 1 (MNR), with SED 10/FP 9 achieving it in three, and all other alternatives
achieving it in all four impoundments at the end of the modeling period (see Table 1).
Notwithstanding, the State of Connecticut has calculated more stringent criteria for unlimited
fish consumption that may not be met in any of these impoundments at the end of the modeling
period.
In addition, alternatives SED 2/FP 1 and SED 10/FP 9 would not meet the federal and state water
quality criterion for freshwater aquatic life and therefore, would not be protective of the
environment; however, the other alternatives do meet this criterion in all reaches by the end of
the modeling period. None of the alternatives analyzed would achieve the federal and state water
quality criterion for human consumption of organisms in any of the Massachu setts reaches, while
SED 2/FP 1, SED 3/FP 3, and SED 10/FP 9 would not achieve this criterion in any Connecticut
impoundments. SED5/FP 4, SED 6/FP 4, SED 8/FP 7, SED 9/FP 8, and SED 9/FP 4 MOD
: The modeling period is 52 years for most alternatives.
11
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
would restore water quality consistent with this criterion in significant segments of the river in
Connecticut consistent with these requirements, based on estimates of meeting this criterion in
the future in 50% or more of the Connecticut impoundments.
All alternatives rely to varying degrees on ICs throughout the river in both Massachusetts and
Connecticut to be protective of human health in the long term. Those alternatives that rely more
extensively on these controls (SED 2/FP 1 and SED 10/FP 9) over longer timeframes and larger
areas have more uncertainty that they will protect human health in the long term, and such
controls provide no protection for ecological risks. Those alternatives that rely less on these
controls (SED 8/FP 7, SED 9/FP 8, and SED 9/FP4 MOD) over shorter timeframes or smaller
areas have higher overall protection of human health.
In summary, the standard of overall protection of human health and the environment includes a
balancing of the short-term and long-term adverse impacts of the alternatives with the residual
risks. Given that restoration of the riverbed, riverbanks, and floodplain can be achieved and
maintained with properly implemented restoration techniques and a long-term monitoring and
maintenance program (see Attachment C-5, Bank Erosion/Restoration, and Appendix D, River
and Floodplain Restoration, in the 2011 Site Information Package), the short-term impacts to the
environment can be successfully mitigated, and will be outweighed by the long-term reduction of
risk from PCBs in the environment, and can be done in a manner that avoids, minimizes, and/or
mitigates impacts to state-listed species. Therefore, implementation of SED 9/FP 4 MOD would
provide overall protection of human health and the environment because it achieves this
important balance between short-term and long-term risks.
3 CONTROL OF SOURCES OF RELEASES
The extent to which each of the combinations of sediment and floodplain alternatives reduce or
minimize possible further PCB releases was evaluated. This evaluation is driven by a comparison
of the sediment and riverbank components of the sediment-floodplain alternative combinations
because the floodplain soil is not a significant source of PCB releases to the river except in the
situation of the river channel relocating into contaminated floodplain.
The model simulation predicts that, in 52 years, the reductions from upstream source control and
other upstream and facility remediation, along with natural recovery processes within the Rest of
River (as reflected in SED 2), would result in reductions of 37% and 41% in the annual mass of
PCBs passing Woods Pond and Rising Pond Dams, respectively, and a reduction of 50% in the
annual mass of PCBs transported from the river to the floodplain in Reaches 5 and 6.3
The reductions relative to current conditions in the annual PCB mass transported within the river
(as represented by the predicted PCB mass) passing Woods Pond and Rising Pond Dams) and to
the floodplain within Reaches 5 and 6 at the end of the model projection period for the various
alternatives are summarized in Table 2.
3 The initial (i.e., current) annual PCB mass values used in the model are 20 kilograms per year (kg/yr) passing
Woods Pond Dam, 19 kg/yr passing Rising Pond Dam, and 12 kg/yr transported from the river to the floodplain in
Reaches 5 and 6.
12
8/1/2012
-------
1 Table 2
2
3 Percent Reduction in Annual PCB Mass Passing Woods Pond
4 and Rising Pond Dams and Transported to the Reach 5/6
5 Floodplain for Combinations of Alternatives (relative to current conditions)
l.nciilion
Sl l) 21
IP 1
Sl l) 3/
IP 3
SI. 1)5/
IP 4
SI.IM./
IP 4
Sl l) X/
IP 7
Sl l) <)/
IP X
Sll) 10/
I P ')
Sl l) <)/ I P
4 MOD
Woods Pond Dam
37%
94%
97%
97%
98%
97%
62%
89%
Rising Pond Dam
41%
87%
93%
95%
96%
96%
62%
89%
Reach 5/6 Floodplain
50%
97%
98%
98%
99%
98%
68%
92%
Solids Trapping
Efficiency in Woods Pond
15%
13%
15%
15%
15%
26%
24%
30%
6
7 The model results show that the percentage decrease of the mass of PCBs passing Woods Pond
8 and Rising Pond Dams, respectively, ranges from 37% and 41% for SED 2 to 98% and 96% for
9 SED 8. All alternatives that provide some active remediation would achieve a minimum decrease
10 of at least 87% for all three compliance points, except for SED 10, which provides for PCB mass
11 reductions in the 60 to 70% range.
12 The comparison of the alternatives against MNR is also presented in Figure 1 in Attachment B-6.
13 This figure shows that SED 2 and SED 10 perform poorly relative to the other alternatives.
14 As additional sources are controlled by permanently removing and/or capping PCB-containing
15 sediment and reducing the contribution of PCBs from the contaminated eroding banks,
16 significant additional reductions in PCB mass transport in the river and transport to the
17 floodplain occurs. As a result, SED 2/FP 1 and SED 10/FP 9 do the least to control releases.
18 While SED 8/FP 7 and SED 9/FP 8 do the most to control releases, SED 3/FP 3, SED 5/FP 4,
19 SED 6/FP 4, and SED 9/FP 4 MOD also provide significant control of releases.
20 SED 9/FP 8, SED 9/FP 4 MOD, and SED 10/FP 9 nearly double the solids trapping efficiency of
21 Woods Pond when compared to the other alternatives. PCBs are attached to solids that move
22 through the river system. Therefore, the increase in trapping of solids in Woods Pond is a
23 mechanism to reduce downstream migration of PCBs. Alternatives SED 9/FP 8 and SED 9/FP 4
24 MOD, and to a lesser extent, SED 10/FP 9, also control sources of releases by removing a
25 significant mass of PCBs from behind Woods Pond Dam. In the event of a serious breach or
26 failure of the dam, the release of PCBs downstream would be less for these alternatives than for
27 the other alternatives that rely primarily on capping or MNR.
28 The different combinations are expected to have different responses in the occurrence of an
29 extreme flood event. SED 2/FP 1 will have no different response than what would be expected
30 to occur under current conditions because there is no active remediation. In this case, PCB-
31 contaminated sediment and soil from eroding banks are expected to be released and mobilized
32 downstream. SED 10/FP 9 is expected to result in similar, but slightly less, downstream
33 transport because it has only a small area in Reach 5A that is addressed by an engineering
34 approach, and residual PCBs in Woods Pond are not capped. SED 3/FP 3 will result in slightly
13
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
less transport than the previous alternatives; however, the use of a thin-layer cap in Reach 5C
and Woods Pond, and MNR in Reach 5B, the Backwaters, and Reach 7 impoundments is not
expected to adequately control sources of releases in an extreme event. Alternatives SED 5/FP 4
and SED 6/FP 4 are expected to provide adequate protection in an extreme event in Reaches 5
and 6, but the use of thin-layer capping and backfill in the downstream reaches provides a high
level of uncertainty in performance during such an event. Alternative SED 8/FP 7 followed by
SED 9/FP 8 are expected to provide the highest level of protection of all the combinations during
an extreme event because they provide the greatest amount of remediation with corresponding
engineering controls. SED 9/FP 4 MOD is expected to provide adequate protection in an extreme
storm event in all reaches, with the exception of Reach 5B, which is subject to MNR and
therefore, bed sediment and bank soil may erode and be transported downstream. However, the
areas of the highest concentrations in Reach 5B will be removed, and the remaining
concentrations are low enough that the impacts are not expected to be unacceptable.
In addition, the results for SED2/FP 1 and SED 10/FP 9 have increased uncertainty because
while the model does include processes associated with bank erosion, it cannot represent changes
in the planform of the river channel, which could result in large contributions of soil (and
associated PCBs) from the erosion into the floodplain over time. The results for the remaining
alternatives are less uncertain than those associated with SED 2/FP 1 and SED 10/FP 9 because
they are based on the assumption that the riverbanks would be stabilized and subject to
operation, monitoring, and maintenance (OMM); therefore, the potential for large contributions
of soil (and associated PCBs) from the banks and floodplain would not exist. SED 9/FP 4 MOD
addresses all eroding contaminated banks in Reach 5A and targets only banks in Reach 5B that
have PCB concentrations exceeding 50 mg/kg and specifies bioengineering techniques wherever
possible.
To assess the extent to which the sediment components of these combinations of alternatives
would mitigate the potential effects of an extreme high-flow event that could cause buried
sediment to be exposed, model predictions of erosion and reach-average PCB concentrations in
surface sediment following an extreme high-flow event were compared. Although the model
simulation predicts varying responses to high-flow events, including the extreme event (50- to
100-year flood) simulated in Year 26 of the projection, the results generally show that buried
sediment containing PCBs would not be exposed to any significant extent during high-flow
events under any remediation alternative. However, this conclusion has some uncertainty
because survey transects, Acoustic Doppler Current Profiler measurements, and deep sediment
cores collected in the river indicate that high-flow events have the potential to remobilize the
sediment column to considerable depths that are not reflected in the two-dimensional averaged
model grid cells. Therefore, the alternatives that include thin-layer capping or backfill are not
likely to perform as well as the model predicts.
Finally, there are differences among the combinations of sediment and floodplain alternatives in
terms of the potential for releases during implementation, including both resuspension-related
releases during sediment removal and potential releases from open excavations in the floodplain
during an extreme weather event. Although engineering controls and/or best management
practices (BMPs) would be applied to minimize such releases, they could not entirely prevent
such releases. The potential for such short-term releases would be a function of the duration of
the remedy and the overall extent of open excavation/dredging areas. For alternatives involving
14
8/1/2012
-------
1 active remediation, durations range from 5 to 52 years and areas of excavation and dredging
2 range from 76 acres to over 700 acres. The effects of such releases are reflected in the model
3 output.
4 4 COMPLIANCE WITH FEDERAL AND STATE ARARs
5 Review of potential chemical-, location-, and action-specific ARARs indicates the following
6 regarding the extent to which the combinations of sediment and floodplain alternatives would
7 meet the ARARs, or could potentially require waiver of some ARARs under the Comprehensive
8 Environmental Response, Compensation, and Liability Act (CERCLA) and the NCP. A chart
9 summarizing the determination of ARARs for SED 9/FP 4 MOD is provided in Attachment B-9.
10 4.1 CHEMICAL-SPECIFIC ARARs
11 A summary of some of the more significant chemical-, location-, and action-specific ARARs is
12 included below.
13 4.1.1 Chemical-Specific ARARs
14 Chemical-specific ARARs include federal and state water quality criteria for PCBs (such as
15 NRWQCs). These criteria consist of freshwater aquatic life and human health criterion (based on
16 consumption of water and/or organisms).
17 Alternatives SED 2/FP 1 and SED 10/FP 9 would not achieve the federal and state water quality
18 criteria for freshwater aquatic life in Massachusetts (but would in Connecticut). All other
19 alternatives would achieve these criteria in all reaches of the river.
20 None of the alternatives would achieve the federal and state water quality criteria for human
21 consumption of organisms in any of the Massachusetts reaches or in all Connecticut reaches.
22 However, alternatives SED 5/FP 4, SED 6/FP 4, SED 8/FP 7, SED 9/FP 8, and SED 9/FP 4
23 MOD would restore water quality in significant segments of the river in Connecticut consistent
24 with these requirements, based on estimates of meeting this criterion in the future in 50% or
25 more of the Connecticut impoundments. Because the water quality criteria for human
26 consumption of organisms (0.000064 micrograms per liter) is not expected to be met in the
27 Housatonic River in Massachusetts and only in portions of Connecticut under any of the
28 alternatives evaluated, and because it is below the reliable limits of analytical detection, EPA is
29 proposing to waive this criterion as technically impracticable.
30 4.1.2 Location-Specific and Action-Specific ARARs
31 All alternatives meet action-specific ARARs.
32 All active alternatives would involve temporary destruction of wetlands and a discharge of
33 dredged or fill material into waters of the state and/or the United States. SED 9/ FP 4 MOD is
34 the least damaging practicable alternative under the Clean Water Act and state and other federal
35 wetlands requirements.
36 In addition, the Massachusetts Endangered Species Act (MESA) is applicable to all active
37 alternatives (SED 3/FP 3, SED 5/FP 4, SED 6/FP 4, SED 8/FP 7, SED 9/FP 8, SED 9/FP 4
38 MOD, and SED 10/FP 9). MESA and its regulations were promulgated to protect state-listed
15
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
species and their habitats. Unacceptable levels of PCBs are present in such habitat areas in the
Rest of River. During the implementation of the preferred alternative, the removal of PCBs from
the Rest of River is anticipated to provide a benefit to state-listed species inhabiting the area due
to the reduction in adverse effects to ecological receptors from the PCBs. In overseeing the
response actions, EPA, in coordination with the Massachusetts Department of Fish and
Game/Division of Fisheries and Wildlife (DFW), consistent with the requirements of MESA
(Massachusetts General Laws [MGL] c. 131 A) and its implementing regulations (321 Code of
Massachusetts Regulations [CMR] 10.00; MESA), will guide efforts to avoid, minimize, and
mitigate impacts to state-listed species. Although a final MESA evaluation will not be
completed until the remedy design phase, by focusing on the Core Areas (Attachment B-4), EPA
and the Commonwealth believe that a framework has been established to achieve MESA
permitting standards of assessing alternatives to both temporary and permanent impacts to state-
listed species, and of limiting the impact to an insignificant portion of the local populations of
affected species (see 321 CMR 10.23). For example, the parties focused on avoidance of some
of the most important and sensitive rare species habitats in Core Area 1. Similarly, in Core
Areas 2 and 3, avoidance of impacts when practical, careful consideration of PCB remediation
methods, and the sequence and timing of remediation activities, as well as after-the-fact habitat
mitigation, are all approaches that will assist in achieving the substantive requirements of
MESA. Although the Core Areas play an important role in guiding avoidance and minimization
of impacts to state-listed species, in some cases the "take" of state-listed species is likely to be
unavoidable. In those cases, consistent with MESA's status as a location-specific ARAR, the
Commonwealth will work with GE and EPA to minimize impacts and to ensure that an adequate
long-term net-benefit mitigation plan for the affected state-listed species is designed and
implemented, as required by 321 CMR 10.23(2)(c).
5 LONG-TERM RELIABILITY AND EFFECTIVENESS
The assessment of long-term reliability and effectiveness for the combinations of sediment and
floodplain alternatives included an evaluation of the magnitude of residual risk, the adequacy and
reliability of the alternatives, and the potential long-term impacts on human health or the
environment.
5.1 MAGNITUDE OF RESIDUAL RISK
The magnitude of residual risk for each of the sediment-floodplain alternative combinations is
evaluated in this subsection considering the individual sediment and floodplain components
separately, primarily because residual risks differ between the in-river and floodplain
environments.
5.1.1 Potential Residual Risks Associated with River Sediment, Water, and Fish
SED 2/FP 1 would rely on natural processes to reduce PCB concentrations and would include
monitoring the effectiveness of these processes. Implementation of the sediment component of
the other combined alternatives being evaluated would further reduce the potential for exposure
to PCBs for humans and ecological receptors through various combinations of removal, capping,
thin-layer capping, and/or natural recovery processes. The model was used to predict the extent
to which each sediment alternative would reduce PCBs in surficial sediment, surface water, and
fish tissue. For purposes of comparison, fish tissue PCB concentrations are presented here
16
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
because fish tissue concentrations integrate the effects of changes in surface sediment and water
column concentrations and therefore, are representative of the trends and relative success of each
alternative in reducing the potential for PCB exposure. Figures 2 and 3 in Attachment B-6 show
the residual surface sediment concentrations and surface water concentrations.
Table 3 presents the subreach-average largemouth bass fillet4 PCB concentrations at the start of
the model projection period and those at the end of the projection period (52-year simulation5 for
all alternatives except SED 8/FP 7, which was 81 years due to the longer construction time for
sediment remediation), and shows the percent reduction in tissue PCB concentrations for each of
the sediment alternatives included in the combinations under evaluation. These results are also
presented graphically for Reaches 5 through 8 and for the Connecticut impoundments in
Attachment B-8.
Based on the above comparisons, other than SED 2/FP 1 (MNR), SED 10/FP 9 provides the least
long-term reductions in fish PCB concentrations. All of the remaining alternatives produce a
reduction of approximately 99% in Reach 5A. For the other reaches, SED 3/FP 3 results in
markedly less reduction in comparison to the more active alternatives (SED 5/FP 4 through SED
9/FP 4 MOD), which are effective in achieving large reductions in fish tissue PCB
concentrations over all reaches of the river. The sole exception is Reach 5B for the SED 9/FP
MOD alternative. This alternative would reduce bioavailability of PCBs via amendment of the
sediment with an additive such as activated carbon. The Housatonic River model, upon which
these results are based, is not able to simulate this process and therefore, fish tissue
concentrations are likely over-estimated in Reach 5B. The resulting reduction in concentrations
from the amendment is anticipated to be similar to that achieved by the removal alternatives in
Reach 5B.
Although some level of fish consumption advisory would need to be maintained at the
conclusion of remediation for many of the alternatives, an additional measure of long-term
reliability and effectiveness that can be used to distinguish among the alternatives is the time
required to achieve a certain IMPG. Plots of fish tissue concentrations by reach in Attachment
B-8 (average fillet PCB concentrations) show that although SED 10 would have the shortest
implementation schedule and would achieve some reductions quickly relative to other removal
alternatives, SED 9 has improved performance relative to all other alternatives, balancing the
magnitude of the reductions with the time required to achieve them.
4 The fillet concentrations are derived by dividing the whole-body tissue concentrations output from the food-chain
model by a factor of 5.
5 The simulation period is 52 years for all alternatives except SED 8/FP 7, which is 81 years due to the longer
construction time for SED 8/FP 7 and the requirement for 30-year projections post-remediation.
17
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
Table 3
Modeled Subreach Average Fish (Fillet) PCB Concentrations at End of Project
Period and Percent Reductions for Combinations of Alternatives
Reach
Initial
( one.
Sl l) 2/
IP 1
Sl l) 3/
IP 3
sr. D5/
IP 4
si:i)(./
IP 4
Sl l) X/
I P ¦>
Sll)')/
IP X
Sll) 10/
I P ')
Sll)')/
IP 4
MOD
Fish PCB Concentration (mg/kg wet weight)
Reach 5A
18
7.3
0.3
0.3
0.3
0.2
0.3
4.2
0.3
Reach 5B
17
9.3
3.0
0.2
0.2
0.2
0.3
6.6
3.5
Reach 5C
14
7.4
1.8
0.2
0.2
0.1
0.2
5.8
0.8
Reach 5D (Backwaters)
22
9.5
6.3
0.4
0.4
0.3
0.4
11
1.1
Reach 6
15
8.6
0.7
0.2
0.2
0.1
0.2
3.7
0.7
Reach 7
6.4-13
2.8-6.4
0.7-2.1
0.4-1.6
0.2-0.7
0.1-0.6
0.2-0.7
1.9-4.4
0.4-1.4
Reach 8
6.3
3.6
1.6
0.3
0.2
0.2
0.2
2.7
0.4
Connecticut (Bulls
Bridge Dam
Impoundment)
0.4
0.2
0.04
0.01
0.009
0.007
0.009
0.1
0.022
Percent Reduction in Fish PCB Concentration Relative to Initial Conditions
Reach 5A
60%
99%
99%
99%
99%
98%
77%
99%
Reach 5B
47%
83%
99%
99%
99%
98%
62%
80%
Reach 5C
48%
87%
99%
99%
99%
99%
59%
94%
Reach 5D (Backwaters)
57%
72%
98%
98%
99%
98%
51%
95%
Reach 6
44%
95%
99%
99%
99%
99%
76%
95%
Reach 7
45 -63%
80 -91%
84 -97%
94 -98%
94 -99%
93 -
98%
59 -75%
86-
95%4
Reach 8
43%
75%
95%
97%
97%
96%
57%
94%
Connecticut (Bulls
Bridge Dam
Impoundment)
60%
91%
97%
98%
98%
98%
73%
95%
Percent Reduction in Fish PCB Concentration Relative to SED 2 (MNR)
Reach 5A
96%
96%
96%
97%
96%
42%
96%
Reach 5B
68%
98%
98%
98%
97%
29%
61%
Reach 5C
76%
97%
97%
99%
97%
22%
89%
Reach 5D (Backwaters)
34%
96%
96%
97%
96%
-16%
89%
Reach 6
92%
98%
98%
99%
98%
57%
91%
Reach 7
67-75%
75-86%
89-93%
91-96%
89-93%
31-32%
75 - 88%4
Reach 8
56%
92%
94%
94%
94%
25%
87%
Connecticut (Bulls
Bridge Dam
Impoundment)
80%
95%
96%
97%
96%
50%
81%
Notes:
1.
2.
3.
4.
PCB concentrations shown (except for the initial concentrations) represent subreach-average values predicted by EPA's
model at the end of the model projection period (52 years for SEDs 2, 3, 5, 6, 9, and 10, and 81 years for SED 8).
Values shown as ranges in Reach 7 represent the range of modeled PCB concentrations at the end of the projection within
each of the Reach 7 subreaches.
Percent reduction represents the change in annual average PCB concentrations predicted by EPA's model between the
beginning and the end of the projection period.
Reach 7 reductions were calculated separately by subreach. Individual subreach initial and SED 2 concentrations were not
provided by GE in the CMS, so reductions shown for SED 9/FP 4 MOD were calculated from EPA model results.
18
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
For example, in Reach 6 (Woods Pond) (see Figure 2), reduction in fillet tissue PCB
concentrations corresponding to the CTE 10"5 cancer risk would not be achieved by SED 2/FP 1
and SED 10/FP 9 during the 52-year simulation period and, based on the trajectories, for many
years thereafter. SED 3/FP 3 and SED 9/FP 4 MOD similarly do not achieve the CTE 10"5
cancer risk concentration during the simulation period but have significantly better performance
than SED 2/FP 1 and SED 10/FP 9, achieving the Massachusetts consumption advisory
concentration and a trajectory that will reach the CTE 10"5 cancer risk concentration many
decades earlier than SED 2/FP 1 and SED 10/FP 9.
SED 9/FP 8 achieves significant reductions in a shorter period of time than comparable
alternatives. SED 8/FP 7, while achieving the largest overall reductions, has a long
implementation period, such that the time to achieve risk reduction is extended beyond that of
other alternatives. For SED 9/FP 4 MOD, these results are also affected by the inability of the
model to simulate the amendment of Reach 5B sediment and therefore, inaccurately under-
predict the performance of SED 9/FP 4 MOD, which is anticipated to approach the effectiveness
of SED 9/FP 8.
Because SED 10/FP 9 specifies only partial remediation in Reach 5 A and therefore, would allow
unremediated sediment to remain exposed, and does not include remediation in the other reaches
upstream of Woods Pond, potential recontamination of the remediated areas due to transport of
PCBs from unremediated areas is a concern for this alternative.
5.1.2 Potential Residual Risks Associated with Floodplain Soil
Under SED 2/FP 1, floodplain soil PCB concentrations, as well as any potential risks, are
assumed to remain generally similar to current conditions. Implementation of the floodplain
component of the other combined alternatives (FP 3, FP 4, FP 4 MOD, FP 7, FP 8, and FP 9)
would reduce the potential risks to humans and ecological receptors from exposure to PCBs in
the floodplain by removing PCB-containing soil and backfilling those excavations with clean
material. The reduction in potential exposure and associated risks would occur upon completion
of remediation in a given area. As the removal volume and area affected among the alternatives
increase, the reduction in exposure also increases. Among the alternatives evaluated, SED 8/FP
7 would provide the greatest reduction in potential exposures, removing the largest volume of
PCB-containing soil over the greatest area of the floodplain (377 acres), and over the longest
period (52 years) (see Table 4).
19
8/1/2012
-------
Average Fillet PCB concentrations in Largemouth Bass (Average for fish ages 5 to 9)
SED 1/2 SED 3 SED 4 SED 5 SED 6 SED 7 SED 8 SED 9 SED 9 MOD SED10
Notes: Average calculated for days fi~om Aug. 28th through Oct. 26th of each year; average calculated for fish ages 5 to 9.
Fillet- based concentrations were calculated as whole body concentrations divided by 5.0.
Horizontal lines represent fish consumption (deterministic) IMPGs.
(Figures for other reaches are presented at the end of Attachment B-8.)
Figure 2 Average Fillet PCB Concentrations in Largemouth Bass from Reach 6
L:\20502169.095\NRRB RESPONSBAPPENDIX B\APPB ROR RRB SIP ALT AN ALYSIS. DOCX
20
-------
1 Table 4 Summary of Percent of Floodplain and Sediment Exposure Areas
2 Achieving IMPGs for Direct Human Contact
l'l\|)OMIIV
Assumptions
Risk l.e\cl
Pcrccnl ol' I2X Hoodplain iiiul Sediment r.xposiire A reus Achie\ing IMPGs
sr. 1)2/ sr. »j/ sr. 1)5/ sin (,i sr.nx/ sr.ny/ si:n in/ s!')'J/
I P 1 IP 3 I P 4 I P 4 I P 7 1.1> s 1 i> y
Cancer l(b
100
100
100
100
100
100
100
100
RME
Cancer 105
56
71
100
100
100
100
61
71-100
Cancer 106
7
9
13
14
100
15
7
9-13
Non-cancer
81
100
100
100
100
100
100
100
Cancer 10^
100
100
100
100
100
100
100
100
CTE
Cancer 105
100
100
100
100
100
100
100
100
Cancer 106
88
98
99
99
100
99
97
98-99
Noncancer
99
100
100
100
100
100
100
100
Percent of 12 Floodplain Frequently Used Subareas Achieving IMPGs
Cancer 10^
92
100
100
100
100
100
100
100
RME
Cancer 105
42
100
100
100
100
100
67
100
Cancer 106
17
42
42
42
100
42
17
42
Noncancer
58
100
100
100
100
100
100
100
Cancer 10^
100
100
100
100
100
100
100
100
CTE
Cancer 105
92
100
100
100
100
100
100
100
Cancer 106
67
100
100
100
100
100
92
100
Noncancer
67
100
100
100
100
100
100
100
3
4 Because different parts of the floodplain are used by human and ecological receptors in different
5 ways and with varying degrees of frequency and intensity, the extent to which each of the
6 combinations evaluated in this section would reduce potential residual risks from PCB exposure
7 in the floodplain has been evaluated in terms of the extent to which they would achieve the
8 IMPGs. The comparative evaluation of the alternative combinations based on achievement of
9 IMPGs is presented in Section 6. An evaluation of the achievement of the IMPGs and the time
10 relative to no action is provided in Section 6.3.
11 PCBs remain in soil below the depths designated for removal (1 ft except in the frequently used
12 subareas where the removal is to 3 ft). Exposure to this deeper soil is not anticipated under
13 current uses. In the event that future exposures to such deeper soil may be reasonably anticipated
14 in particular areas, it would be addressed, under all alternatives except SED 2/FP 1, by ICs.
15 Additionally, under those alternatives, ICs would be implemented where necessary to address
16 potential risks from reasonably anticipated future uses.
17 5.2 ADEQUACY AND RELIABILITY
18 5.2.1 Use of Technologies Under Similar Conditions
19 SED 2/FP 1 involves MNR with ICs in the river and no action in the floodplain. MNR has been
20 selected at other contaminated sediment sites as part of the overall remedy, and no action has
21
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
been adopted as a remedy component at other sites. The other six combinations involve different
combinations of remedial technologies and processes.
For the sediment alternatives, the selected approaches include removal in the dry and/or wet
(followed by capping or backfilling in most cases), capping without prior removal, thin-layer
capping, riverbank stabilization (using a combination of bioengineering and hard stabilization
techniques), and MNR. All of the remedial technologies included in the sediment alternatives
under evaluation have been used at other sites.
The floodplain components of the combinations involving remediation would rely primarily on
removing floodplain soil from areas of various types of habitats and backfilling the excavations,
and implementation of ICs. These technologies and combinations of technologies have been
implemented at other sites. (Restoration is discussed separately below.)
5.2.2 General Reliability and Effectiveness
The alternatives under evaluation generally use technologies that have been shown to be reliable
and effective at other sites. However, as noted in Section 13 of the June 2011 Site Information
Package, thin-layer capping is not expected to be a reliable or effective component of any
alternative for this site, and backfill may not be suitable for reaches with higher bed shear
stresses.
For all of the active remedial combinations except SED 10/FP 9, the eroding riverbanks in
Reaches 5 A and 5B would be stabilized using a combination of bioengineering techniques and, if
necessary, hard engineering techniques. SED 9/FP 4 MOD would be designed to specifically
target those sections of riverbank that are highly erodible and also contain the highest
concentrations of PCBs. The stabilization techniques would be similar for all of the sediment
components of these remedial combinations, except that they would potentially be modified in
alternatives that would require construction in the wet. Such combinations of techniques are
expected to be reliable and effective in stabilizing the banks and controlling erosion. Any
potential for long-term impacts would be mitigated through proper construction, monitoring, and
operation and maintenance (O&M) practices. Using this approach, riverbank stabilization would
not exacerbate erosion in other areas, and would not result in ecological impacts.
Finally, the areas remediated under the combinations of alternatives would require restoration.
Implementation of restoration methods would return the habitat functions and values over the
timeframes required to complete remediation for the alternatives (i.e., over 5 to 52 years).
Remediation would progress from upstream to downstream, impacting small stretches of the
river and floodplain at any given time; likewise the restoration would also proceed incrementally.
In addition, monitoring and OMM programs, including invasive species control, would ensure
proper re-establishment of vegetation for a period of time following remediation. There is a
significant body of knowledge with respect to ecosystem restoration that documents the ability to
re-establish the preremediation conditions and functions of the affected habitats (see Appendix D
of the 2011 Site Information Package). As such, these restoration techniques are expected to be
fully effective and reliable in returning these habitats, including vernal pool habitat, to their pre-
remediation state. As a result, the likelihood of effective restoration is equal under any of the
alternatives.
22
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
5.2.3 Reliability of Operation, Monitoring, and Maintenance Requirements and
Technical Component Replacement Requirements
All alternative combinations would incorporate reliable long-term monitoring and/or
maintenance techniques. For example, for the sediment alternatives, inspection and repair or
replacement of the caps or bank stabilization measures would be required. However, as the area
to be capped increases (progressively more from SED 10/FP 9 to SED 9/FP 8), there would be a
greater need for more extensive monitoring and maintenance.
Similarly, the backfilled/restored areas of the floodplain would be monitored through periodic
inspections to verify that the planted vegetation is surviving and growing, and to identify areas
(if any) where the backfill is eroding or in need of repair. This is a reliable means of assessing
the need for maintenance. However, monitoring and maintenance could be difficult to implement
in certain areas of the floodplain due to remoteness, the extent of standing water, and the extent
of vegetation both in and around the remediated areas. Depending on the timing, location, and
scale of any repairs, access roads and staging areas may need to be temporarily constructed in the
floodplain. These difficulties can be overcome to a great extent through proper planning,
selection of experienced contractors, and effective oversight of activities.
5.3 POTENTIAL LONG-TERM IMPACTS ON HUMAN HEALTH AND THE
ENVIRONMENT
The evaluation of potential long-term impacts on human health or the environment includes
evaluation of potentially affected populations, long-term impacts on the various habitats that
would be affected by the combinations of sediment and floodplain alternatives, and the biota that
inhabit those habitats (including impacts on state-listed species), impacts on the aesthetics and
recreational use of the river and floodplain, impacts on banks and bedload movement (i.e., fluvial
geomorphic processes), and potentially available measures that may be employed to mitigate
these impacts. The long-term impacts of exposure to PCBs left in place are not evaluated in this
section.
5.3.1 Potentially Affected Populations
Implementation of all of the alternatives except SED 2/FP 1 (which would not involve remedial
construction activities) would result in some level of short- and long-term impacts on floodplain
habitats, with the impacts being spread over longer periods of time as the combinations of
alternatives grow more comprehensive and the duration for implementation increases. These
habitat alterations would affect the people, animals, and plants that use these areas. In the case of
SED 9/FP 4 MOD, impacts to habitats supporting state-listed species would be limited due to the
design of the alternative, which includes specific protocols for addressing Core Areas. For all
alternatives, however, implementation of remediation would proceed from upstream to
downstream, affecting short stretches of river and associated floodplain at any given time. The
long-term impacts of the combinations of alternatives on the affected habitats and the plants and
animals that inhabit or use those habitats, as well as the long-term impacts on the aesthetics and
recreational use of the affected habitats by people, are discussed and compared below.
23
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
5.3.2 Long-Term Impacts on Habitats and Biota
The extent and severity of long-term impacts from remedial construction activities are dependent
on the types of habitat affected, the size of the affected areas, the success of the restoration
approach(es), and length of time needed for restoration. Table 5 from GE's RCMS identifies the
habitat types and summarizes the areas of each habitat affected by the combinations of sediment
and floodplain alternatives.6 As discussed above, long-term impacts would be mitigated through
proper restoration measures. Because restoration of affected habitats is dependent on several
factors and processes, the length of time necessary to restore a habitat is variable.
Aquatic Riverine Habitat: The potential post-restoration impacts of sediment removal/capping,
as well as capping or thin-layer capping without removal, on aquatic riverine habitat include the
following:
¦ The caps would cause a change in surface substrate type from its current condition
(sand, sand and gravel, or silt) to armor stone, lasting until deposition of natural
sediment from upstream changes the substrate surface back to a condition similar to its
prior condition. To the extent that a habitat layer is specified as the part of any cap,
this impact would be reduced or eliminated.
¦ There may be a temporary loss of woody debris and shade in Reaches 5A and 5B of
the Housatonic River, depending on the removal areas, bank stabilization techniques,
and restoration techniques. This could alter the riverine habitat because woody debris
provides structure that is important to many aquatic and semi-aquatic species, and
shading limits temperature increases in the river water. The addition of woody debris
would be a component of the restoration plan.
¦ Sediment removal and/or capping would remove or bury the existing aquatic
vegetation and benthic invertebrates, and temporarily displace the fish. Recolonization
would occur and the vegetation and invertebrates that would recolonize these areas are
not expected to differ substantially from the pre-existing species if a habitat layer is
included in the cap design. In addition after the removal of the negative effect of
PCBs on the benthic community it is expected that overall improvements to the
community would be realized.
¦ There is the potential that the disturbed areas could be colonized by invasive species.
This impact may be mitigated via active control of invasive species.
¦ In shallow areas subject to capping or thin-layer capping without removal, the increase
in substrate elevation due to the cap could change the hydrodynamics and vegetative
characteristics of the areas and the biota dependent on them.
6 EPA does not believe that the infrastructure included in these estimates provided by GE has been optimized and
expects that, for the selected remedy, the staging areas and roads will be designed to minimize the footprint and
adverse impacts to the floodplain, neighborhoods, and local roads while allowing for the remediation to proceed
in a timely and effective manner.
24
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Table 5
Habitat Areas in Primary Study Area Affected by Combinations
a
of Sediment and Floodplain Alternatives
llnhilnl
Sl.l) 21
IP 1
Sl.l) 3/
IP 3
Sl.l) 5/
IP 4
sr. ih./
IP 4
Sl.l) N/
IP 7
SI-11) *>/
IP s
Sl.l) 10/
IP')
sr.i) '>/
IP 4
MOD
Aquatic ki\ eniie llabilal (acrcij
-
"9
12"
12"
12"
12"
20
99
Riverbank (linear miles)
--
14
14
14
14
14
1.6
6.6
Impoundment Habitat (acres)
--
60
101
139
139
139
42
139
Backwater (acres)
--
0
61
70
86
66
0
61
Floodplain Wetland Forest (acres)
-
38
60
60
178
56
14
TBDd
Shrub and Shallow Emergent
Wetlands (acres)
-
19
22
22
70
31
3.7
TBDd
Deep Marshes (acres)
-
1.9
0.3
0.3
4.7
3.1
0
TBDd
Vernal Pools (acres) b
-
15 (58)
15 (58)
15 (58)
17(61)
18(61)
0
TBDd
Disturbed Upland Habitats (acres)
-
14
15
15
25
11
7.5
TBDd
Upland Forested Habitats (acres)
-
4.2
4.9
4.6
6.4
2.8
0.7
TBDd
Total (acres)0
--
231
406
453
653
454
88
TBDd
a Includes habitat areas within the boundaries of the Woodlot (2002) natural community mapping; includes
remediation areas as well as areas impacted by access roads and staging areas.
b Number of vernal pools affected is shown in parentheses.
0 Total habitat area affected does not include riverbanks, and can differ from total surface area affected since the
total shown includes all habitats within the boundaries of the Woodlot (2002) mapping (see note a).
d EPA estimates that the total area of floodplain to be affected equals 45 acres. Specific locations and habitat types
to be determined based on habitats and occurrences of state-listed species as defined by the Core Areas. These
estimates do not include supporting infrastructure.
In summary, in the aquatic riverine habitat subject to remediation, it is expected that over time
the physical substrate type in the river would approximate its prior condition and a biotic
community consistent with that substrate type would be present. The inclusion of a habitat layer
in any cap design and implementation of an appropriate restoration plan is expected to accelerate
the recovery of the aquatic biota. The amount of area affected in each alternative is summarized
in Table 5. For all alternatives, areas either upstream or downstream of the immediate
remediation at any given time would act as sources and refuge for aquatic species both during
and after remediation of an area is completed.
Riverbank Habitat. The potential impacts of bank stabilization on riverbank habitat include the
following:
¦ The implementation of stabilization measures that eliminate vertical and/or undercut
banks would result in a loss of habitat for birds and other animals that depend on such
banks (e.g., kingfisher, bank swallow, and the state-listed wood turtle). However,
25
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
proven techniques are available to provide adequate bank stabilization with minimal
loss of potential habitat.
¦ The removal of any mature trees overhanging the river as part of bank
stabilization/remediation would result in a temporary change in the vegetative
character of the banks from their current condition. Although this impact may be
mitigated to some extent by planting of trees following remediation, it is not practical
to replace large trees that are currently found along the banks if removal is necessary.
However, in the long term, the restoration will result in mature trees that overhang the
river and essentially restore the vegetative character to its preremediation conditions.
¦ The use of bank stabilization measures could potentially result in a temporary
reduction in slides and burrows of muskrat and beaver, and could potentially also
reduce access routes and movement of reptiles, amphibians, and smaller and less
mobile mammals between the river and the wetland habitats they use. These potential
impacts can be taken into account and mitigated in the design of bank stabilization.
¦ Any colonization by invasive plant species would require active control measures.
As a result of these impacts, the stabilized riverbanks would not immediately return to their
current condition or level of function; however, over time they are expected to do so. Because
all of the alternative combinations except SED 2/FP 1 and SED 10/FP 9 would involve
stabilization of the eroding banks in Reaches 5A and/or 5B, they would produce temporary
impacts along those banks. SED 2/FP 1 would have no such impacts. SED 10/FP 9 would
involve remediation and stabilization of only a small portion of the banks in Reaches 5A and 5B,
totaling approximately 1.6 linear miles. SED 9/FP 4 MOD would limit removal/stabilization of
banks in Reach 5A to only those areas with high erosion potential and PCB concentrations and
would specify a decision-tree approach to bank stabilization with soft restoration techniques
favored over hard armoring. Under this alternative, in Reach 5B, only a very small percentage of
riverbanks will be affected because only those areas greater than 50 mg/kg would be remediated.
Preliminary estimates are that SED 9/FP 4 MOD would entail disturbance of approximately 3.5
linear miles of Reach 5 A riverbank and less than 0.2 linear miles of Reach 5B riverbank.
Impoundment Habitat. The potential impacts from removal and/or capping or thin-layer capping
on the habitat of impoundments are similar to the impacts on aquatic riverine habitat discussed
above. In general, they would include a temporary or longer-term change in the surface substrate,
and an alteration in the biological community in the affected impoundment. It is anticipated that
as sand and organic sediment are deposited over time from upstream, a biological community
typical of such impoundments would re-establish itself. The alternatives that involve capping or
thin-layer capping without removal in the impoundments would change the bottom elevation,
potentially changing the vegetative characteristics, and the biota dependent on them, in the
shallow portions of the impoundments. By contrast, the placement of a cap or a thin-layer cap in
deeper areas of the impoundments, including the "deep hole" portion of Woods Pond, is not
expected to have any significant long-term ecological impacts. The inclusion of a habitat layer
in a cap would accelerate the recovery. The amount of acreage affected in each alternative is
summarized in Table 5.
26
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Backwater Habitat: The potential impacts of thin-layer capping or sediment removal/capping in
backwaters include the following:
¦ Change in surface substrate type from organic silty material to sand, which would
continue until enough silt and organic material have been deposited to approximate
prior conditions.
¦ Change in vegetative characteristics corresponding to the change in substrate type and
elevation (including, in shallower areas where the thin-layer cap exceeds the depth of
water, a potential change from emergent wetlands vegetation to species more tolerant
of less frequently inundated or drier conditions).
¦ Change in the wildlife communities using the backwaters until such time as the soil,
hydrological, and vegetative conditions of the backwaters return to conditions
comparable to preremediation conditions.
The area disturbed in each alternative is summarized in Table 5. All of these combinations
would have the potential impacts described above, which would be mitigated through the
inclusion of a habitat layer and using proper restoration techniques.
Floodplain Wetland Forest Habitat: The potential post-restoration impacts of floodplain soil
removal, as well as the construction of access roads and staging areas, on floodplain wetland
forest habitat include the following:
¦ The removal of mature trees from the forested floodplain areas subject to soil removal
or the construction of access roads and staging areas would result in a loss of mature
forested habitat in those areas. Following replanting, the plant community succession
in these areas would progress as a maturing forest for a period of years.
¦ Tree removal would cause a temporary loss of the coarse woody debris that is used as
structural wildlife habitat and, for a short period of time, the annual leaf litter that
provides habitat for numerous woodland species.
¦ There would be a temporary relocation or loss of the forest wildlife species that
currently utilize the mature forested habitats that would be removed, and the return of
those species, including sensitive species, would be encouraged through proper
restoration that reestablishes the functions of the ecosystem.
The area impacted by each alternative is summarized in Table 5.
Shrub and Shallow Emergent Wetlands and Deep Marshes: The potential post-restoration
impacts of floodplain soil removal include:
¦ Changes in soil composition and chemistry due to the replacement of existing wetland
soil.
¦ Changes in the hydrology of these wetlands due to impacts on the swales, drainage
features, and microtopography that influence the hydrology.
¦ Changes in vegetative characteristics due to the changes in soil and hydrological
conditions.
27
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
These potential impacts would be mitigated through proper restoration to ensure that soil and
hydrological conditions similar to preremediation conditions are reestablished. Table 5 shows
the area impacted by each alternative.
Vernal Pools and Surrounding Habitat: The potential impacts of floodplain soil removal and
associated facilities on vernal pools, as well as the surrounding non-breeding habitat for vernal
pool amphibians, include the following:
¦ The excavation and replacement of the surface soil and vegetation within and around
vernal pools could potentially change the sediment types and stratigraphy,
microtopography, and foliage cover of these pools, as well as the surface flow patterns
into and out of the pools. These changes could alter the hydrology of the pools.
However, these conditions would not occur if proper restoration techniques are
implemented.
¦ There is also the potential for temporary changes in the vegetative characteristics of
the vernal pool because the complex and mature organic vegetative composition (alive
and dead) of these pools would take some time to reestablish following remediation.
In addition, mature trees around the periphery of the pools, if removed, would take
time to become reestablished.
¦ Changes in soil composition in the vernal pools are possible; however, replacement
soil would be designed to match as closely as possible the characteristics of the
existing vernal pool soil.
¦ Habitats immediately adjacent to vernal pools are important for maintaining water
quality and providing shade and litter for the pool. The proximate non-breeding
terrestrial habitats, with features such as coarse woody debris and the burrows of small
mammals, provide a variety of protective cover, temperature and moisture regulation,
and overwintering habitat functions for vernal pool amphibians. Any impacts to these
adjacent areas will be restored using supplemental plantings to reestablish the native
plant community and habitat.
¦ Implementation of effective restoration techniques would reestablish vernal pool
functions that would allow sensitive vernal pool species (including wood frogs, spotted
salamanders, and the state-listed Jefferson salamander) to return to the vernal pools
following completion of remediation.
The area affected by each alternative is listed in Table 5. Due to the iterative decision-tree
approach to vernal pools included in SED 9/FP 4 MOD, it is not possible to calculate comparable
acreage for that alternative. The floodplain component of SED 9/FP 4 MOD would be
specifically designed to generally avoid, minimize, or mitigate excavation and infrastructure in
Core Area habitats and/or known occurrences of state-listed species and therefore, would have
more limited impacts on these resources than the other alternatives specifying remediation in the
floodplain.
Uyland Habitats: Most of the affected upland areas consist of disturbed upland habitats, which
include agricultural fields and cultural grasslands. Because these areas support altered or early
successional plant communities that have limited ecological value, no long-term impacts would
be expected from the remediation in these areas under any of the remedial combinations.
28
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Where the remediation or supporting activities would affect upland forested habitats, they would
have similar potential impacts as discussed for floodplain forests. As shown in Table 5, apart
from SED 2/FP 1, all of the combinations of sediment and floodplain alternatives would have
some, although relatively limited, impacts on these habitats.
5.3.3 Long-Term Impacts on State-Listed Species
All of the sediment-floodplain alternative combinations, except SED 2/FP 1, would affect the
priority habitats of some state-listed species of concern regulated under MESA. As discussed
previously, GE conducted an evaluation for each potentially affected state-listed species to assess
whether each of the remedial combinations would result in a "take" of that species under MESA
and, where there would be a take, to assess whether the combination would impact a significant
portion of the local population(s) of the species.
The SED 9/FP 4 MOD alternative differs from the other alternative combinations in that that
alternative provides more specificity on options for avoiding, minimizing, or mitigating impacts
to state-listed species. As part of their Priority Habitat mapping process, taxonomic experts from
DFW's Natural Heritage and Endangered Species Program (NHESP) routinely delineate habitat
for each state-listed species based on actual field-documented records or "occurrences." NHESP
has outlined four types of Housatonic Core Areas for this project (see Attachment B-4). Core
Areas 1, 2, and 3 (species richness) represent subsets of the delineated state-listed species habitat
found in the Primary Study Area (PSA). Core Area 4 represents a subset of the documented and
potential vernal pool habitat in the PSA. While an estimate for the number of species affected
cannot be summarized in a manner similar to that of other alternatives, the SED 9/FP 4 MOD
approach is intended to target cleanup depending on the location of these Core Areas.
For SED 5/FP 4, there are an estimated 57.8 acres of floodplain soil that would require
remediation to address the direct contact pathway. The overlap of these 57.8 acres with Core
Areas 1 through 3 is shown in Table 6.
Table 6
Overlap of the 57.8 Acres of Floodplain Soil Requiring Remediation with Core
Areas 1 through 3
Tolsil Acivsijic
Om'i I;i|) OiiI\
willi Coiv Aivsi 1
0\crhi|) with
Co iv Aivsi 3
(l-ACllKlillli CoIV
Aivsi 1)
(hcrlsip with
Co iv Aivsi 2
(l.\clii(lin;i ( oiv
Aivsis 1 siikI 3)
No ()\crlsi|>«illi
Co iv Aivsis 1. 2.
snul 3
5".S acres>
ll.o ;idcs
13 acres>
1. jcrcb
10.2 ;icics
Implementation of FP 4 currently specifies that sufficient remediation needs to occur to meet
HI = 1/10"5, regardless of the presence of Core Areas, whereas in SED 9/FP 4 MOD, the areas to
be remediated could be reduced or minimized in certain Core Areas, as long as the remediation
will meet a minimum of HI = 1/10"4 For example, remediation could be avoided in Core Area 1,
except where necessary to meet HI = 1/10"4 This could result in reducing the area to be
remediated by approximately 11 acres. A reduction of remediation in 50% of the overlap of
Core Area 3, along with mitigation/restoration for remediation in these areas, could reduce the
29
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
area to be remediated by an additional 6.5 acres, thus reducing the total estimated acreage of
floodplain remediation to approximately 40 acres under SED 9/FP 4 MOD.
In addition, based on the iterative approach for vernal pools called for in SED 9/FP 4 MOD, 5
acres of vernal pool are currently estimated to require active remediation. Thus, the total acreage
of floodplain excavation for SED 9/FP 4 MOD, including vernal pools, is estimated to be
approximately 45 acres. Therefore, this approach is expected to have less of a long-term impact
on state-listed species than other alternatives such as SED 5/FP 4.
5.3.4 Long-Term Impacts on Aesthetics and Recreational Use
All combinations of sediment and floodplain alternatives, except SED 2/FP 1, would have some
short-term impacts on the aesthetic features of the Rest of River. Floodplain soil removal
activities, as well as the construction of access roads and staging areas necessary to support
sediment and soil removal, would require removal of trees and vegetation, which would detract
from the natural pre-remediation appearance of those areas until such time as restoration
plantings have matured. The alternatives would have impacts on aesthetics corresponding to the
amount of area impacted (see Table 5) and the duration of the implementation of the remedy.
Similarly, all of the alternative combinations, except SED 2/FP 1, would disrupt, to some extent,
recreational use of the river and floodplain during the remediation period. These affected uses
include canoeing, fishing, waterfowl and other game hunting, hiking, dirt biking, and general
recreation. However, these impacts would not occur over the entire area to be remediated at a
single point in time because the remediation would proceed from upstream to downstream,
affecting small areas at a given time. It is expected that any alternative will include a component
to manage and maintain recreational opportunities safely during remediation.
5.3.5 Long-Term Impacts on Fluvial Geomorphic Processes
All of the combinations of sediment and floodplain alternatives involving active remediation,
except SED 10/FP 9 and SED 9/FP 4 MOD, would rely on stabilization of eroding riverbanks in
Reach 5 A and high PCB concentration riverbanks in Reach 5B. In SED 10/FP 9 and SED 9/FP
4 MOD, only select areas of the banks are proposed for stabilization. These bank stabilization
activities, which are intended to prevent bank erosion and channel migration from exposing new
areas of PCB-contaminated soil, would prevent or permanently curtail the current processes of
bank erosion and lateral channel migration. As discussed in Attachment B-l, the river was
altered substantially by human activities over the past centuries. These alterations have resulted
in an unstable river channel, which over time is acting to regain a state of dynamic equilibrium
that includes changes in the planform of the river channel. During remedial design, natural
channel design techniques can be implemented to reduce the instability of the river channel and
banks. Natural channel design, coupled with thoughtful bank stabilization and restoration
techniques, will provide for a mix of riverbank types, including vertical and undercut banks, and
less near-bank sheer stress.
The stabilization of the banks, as well as the capping of the riverbed, would reduce the supply of
sediment to the river from these sources. This reduction could affect such in-river processes
such as sediment transport (as bed load or suspended load), point bar development, and changes
in channel dimension (i.e., width and/or depth), as determined by sediment deposition/erosion
patterns. Based on geomorphological considerations and modeling results, the reduction in
sediment load associated with riverbank stabilization and riverbed armoring under any of the
30
8/1/2012
-------
1 alternative combinations would not be expected to result in a large-scale, long-term impact on
2 these river morphologic processes or on in-river hydrologic characteristics such as water depth
3 and current velocity.
4 5.3.5.1 Potential Measures to Mitigate Long-Term Impacts
5 For all of the combinations of sediment and floodplain alternatives that involve active
6 remediation, a variety of restoration measures are available to mitigate long-term impacts
7 resulting from their implementation. As summarized above, these methods, when implemented
8 properly, will successfully reestablish functions and values and minimize the potential for long-
9 term negative impacts from the remediation.
10 6 ATTAINMENT OF IMPGs
11 In the assessment of IMPG attainment for the combinations of sediment and floodplain
12 alternatives, the post-remediation average PCB concentrations in an exposure area, as defined in
13 the Human Health Risk Assessment (WESTON, 2005), were compared to the relevant IMPGs
14 for both the sediment and floodplain components of the combinations. Further, the whole-body
15 fish tissue PCB concentrations predicted by the model (or estimated by the Connecticut 1-D
16 analysis) at the end of the model projection period were converted to fillet concentrations and
17 compared to the fish consumption IMPGs (Attachment B-8).
18 For ecological receptors, the modeled sediment or prey tissue concentrations at the end of the
19 projection period and/or the estimated floodplain soil concentrations for the appropriate
20 averaging areas, were compared to the relevant IMPGs. For insectivorous birds and piscivorous
21 mammals, these comparisons used procedures that consider both the sediment and the floodplain
22 components of the alternative combinations.
23 This comparative evaluation focused on a comparison of the total number of averaging areas
24 with predicted PCB concentrations that achieve the applicable IMPG(s). In addition, for the
25 sediment component of each combination, as required by the Permit, the time that it would take
26 to achieve the IMPGs was estimated. For the floodplain component of each combination, the
27 timeframe to achieve IMPGs is assumed to be the same as that required to complete the
28 remediation in a particular area (i.e., the reduction in soil concentrations would occur upon
29 completion of backfill placement). IMPG attainment for each of these human exposure
30 pathways and ecological receptor groups is described in the following subsections.
31 6.1 COMPARISON TO HUMAN HEALTH IMPGs
32 6.1.1 Human Direct Contact with Floodplain Soil and Sediment
33 For all of the combinations of sediment and floodplain alternatives under evaluation, a detailed
n
34 comparison of human direct contact IMPG attainment (RME and CTE IMPGs, respectively ) for
35 the floodplain soil and sediment exposure areas (EAs) was conducted and is summarized in
36 Table 4 taken from GE's RCMS. These comparisons indicate the following regarding IMPG
37 attainment in the floodplain and sediment EAs:
7 The RME IMPGs are those based on RME assumptions (representing more highly exposed individuals), and the
CTE IMPGs are those based on CTE assumptions (representing individuals with average exposure).
31
8/1/2012
-------
1 Floodplain Direct Contact EAs: The floodplain alternatives included in the combinations were
2 by design established to achieve designated risk levels for the RME cancer risk or HI of 1, with
3 the exception of SED 2/FP 1. For direct contact with floodplain soil, the floodplain soil PCB
4 concentrations under SED 2/FP 1 (which were assumed to be the same as current levels) are
5 within or below the range of the RME and CTE IMPGs associated with the cancer risk of 10"4 in
6 all 120 floodplain EAs. However, the PCB concentrations exceed the noncancer-based RME
7 IMPG (HI = 1) in 24 of the EAs. Further, 5 of the 12 frequently used subareas do not achieve the
8 noncancer RME IMPG (and 1 does not achieve the RME IMPG associated with a cancer risk of
9 10"4. The risk levels achieved by the SED 9/FP 4 MOD alternative, which was not evaluated in
10 GE's RCMS, are also shown in Table 4. This alternative achieves the human health risk target
11 of 10"5 or 10"4 (depending on the impact to core habitat areas and following the process outlined
12 above), or an HI of 1, for RME receptors while avoiding Core Area 1 habitat areas whenever
13 possible (except as needed to achieve a minimum risk level of 10"4 or an HI of 1).
14 Sediment Direct Contact EAs: For direct contact with sediment, for SA 3 (Woods Pond) and SA
o
15 7 (Glendale impoundment) , which are the sediment EAs that do not currently achieve
16 acceptable risk levels due to RME noncancer risk exceeding an HI of 1, model projections
17 indicate that during the modeling period, the RME noncancer risk level (HI =1) would be
18 achieved with no action. The remaining alternatives all involve active remediation in Woods
19 Pond, and all achieve an HI of 1 in shorter periods of time, ranging from 21 years for SED 8/FP
20 7, to approximately 15 years for SED 5/FP 4 and SED 6/FP 4, and less than 10 years for SED
21 3/FP 3, SED 9/FP 8, SED 10/FP 9, and SED 9/FP 4 MOD.
22 6.1.2 Human Consumption of Floodplain Agricultural Products
23 As there are no current EAs in the floodplain being used for agricultural production, this pathway
24 does not pose current risks. However, there is the potential for future risk if land uses change
25 and in the implementation of any alternative ICs would need to be established for this
26 circumstance.
27 6.1.3 Human Consumption of Fish
28 Table 1 from GE's RCMS presents a detailed evaluation, for all of the combinations of sediment
29 and floodplain alternatives, of whether the fish PCB concentrations predicted by the model for
30 each river reach or subreach at the end of the modeled period (when converted to fillet
31 concentrations) would achieve the various RME and CTE IMPGs for human consumption of
32 fish. The risk levels for fish consumption for the SED 9/FP 4 MOD alternative, which was not
33 evaluated in GE's RCMS, have been included in this table. Attachment B-8 provides a graphical
34 representation of how the alternatives perform when compared to the various risk levels.
35 6.2 COMPARISON TO ECOLOGICAL IMPGs
36 This section compares the extent to which the combinations of sediment and floodplain
37 alternatives under evaluation would achieve the IMPGs for the various ecological receptors. The
38 tables included in this section are taken from GE's RCMS.
8 It appears that due to rounding issues GE does not recognize that SA 7 exceeds the RME HI of 1.
32
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
6.2.1 Benthic Invertebrates
The IMPGs for benthic invertebrates apply to the sediment in 32 averaging areas in Reaches 5
through 8 and how the IMPGs are achieved is summarized in Table 7 and shown graphically in
Attachment B-6, Figure 4. The table shows, for each combination, the percentage of those
averaging areas where the model-predicted sediment concentrations would achieve the upper-
bound and lower-bound IMPGs. The figure presents the same data in terms of total area over
which the benthic invertebrate IMPGs are achieved.
All alternative combinations evaluated, with the exception of SED 2/FP 1 and SED 10/FP 9,
achieve the upper-bound IMPG for benthic invertebrates of 10 mg/kg total PCBs (tPCBs) in
sediment in all areas. SED 6, 7, and 8 also achieve the lower-bound IMPG of 3 mg/kg tPCBs in
all averaging areas. SED 2, 3, 4, and 10 achieve the lower bound in a range from 22 to 91% of
the averaging areas. SED 9/FP 4 MOD achieves the lower-bound IMPG in 93% of the averaging
areas, but is anticipated to have better performance due to the amendment of Reach 5B sediment
with activated carbon, which will protect benthic invertebrates by reducing the bioavailability of
PCBs, a process that cannot be simulated by the model.
Table 7
Summary of Percent Benthic Invertebrate
Averaging Areas Achieving IMPGs for Benthic Invertebrates
Pm'i'iil ill" A\er;iiiiiiii Aivsis Achie\ inji
IMP(.s in Suiiiicc Sediments
r
*
5*
-t
IMPCs
CL.
CL.
CL.
CL.
CL.
D.
CL.
a. _
u.
Li.
u.
Li.
Li.
u. fl
?i
If.
*
—
? C
£
£
£
£
£
£
£ ~
LJ
LJ
/.
/.
/.
/.
/.
/.
/.
/.
Upper Bound (10 mg/kg in sediment)
72
100
100
100
100
100
84
100
Lower Bound (3 mg/kg in sediment)
22
63
91
100
100
100
34
931
1 Addition of activated carbon to Reach 5B sediment may achieve protection equivalent to 3 mg/kg at current total
organic carbon (TOC).
6.2.2 Amphibians
The IMPGs for amphibians apply both to the 66 vernal pools designated by EPA in the Reach 5
floodplain identified by Woodlot (2002) and to 29 separate backwater areas. Table 8 provides a
summary of the percent of the averaging areas achieving the lower-bound and upper-bound
amphibian IMPGs in the 66 vernal pools (based on the floodplain component of the
combinations) and in the 29 backwater areas (based on the sediment component of the
combinations). Attachment B-6, Figure 5, presents the same data graphically in terms of the
actual area in acres achieving the IMPGs.
SED 8/FP 7 and SED 9/FP 8 would achieve both the upper-bound (5.6 mg/kg tPCBs) and lower-
bound (3.27 mg/kg tPCBs) amphibian IMPGs in all areas, while SED 10/FP 9, the lowest
33
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
performing alternative, would provide only marginal improvement over MNR (SED 2/FP 1).
Although SED 3/FP 3 achieves the upper-bound IMPG in 85% of the averaging areas, as shown
in Attachment B-6, Figure 5, these represent only 51% of the total acreage. SED 9/FP 4 MOD
will achieve protection of amphibians through an iterative decision-tree process that will be
followed after extensive data collection to select a subset of vernal pools for remediation and
restoration using traditional techniques, and pilot testing of remediation technology options,
followed by implementation of concepts proven in this process. This approach will ensure that
remediation of vernal pools will not result in more harmful impacts than the current exposure to
PCBs. SED 9/FP 4 MOD will achieve the upper and lower bound IMPGs in all backwaters,
except potentially in backwaters, or portions thereof, that coincide with Core Area 1 habitats. In
these areas, an additive such as activated carbon may be used to further reduce bioavailability of
any residual contamination.
Table 8
Summary of Percent of Amphibian
Averaging Areas Achieving IMPGs for Amphibians
Percent «if A\oriiiiiiiii \iv;is Achk'\ inii
IMPCs in Surliicc Soil/ScdiiiK'iil
-i-
r-
zc
IMPCs
a.
Q-
a.
Om
Li.
Om
Li.
Om
Ui
Cl
Li.
r I
•r.
£
£
£
£
£
a
£
/.
/.
/.
/.
/.
/.
cn
Upper Bound (5.6 mg/kg in soil/sediment)
18
85
98
100
100
100
21
Lower Bound (3.27 mg/kg in soil/sediment)
13
27
40
48
100
100
14
6.2.3 Warmwater and Coldwater Fish
The IMPGs for fish protection apply to whole-body fish tissue PCB concentrations; the IMPG
for warmwater fish is 55 mg/kg and the IMPG for coldwater fish is 14 mg/kg. Table 9 is a
summary presentation of IMPG attainment for warmwater fish within the 14 subreaches of
Reaches 5 through 8 and coldwater fish within the 8 subreaches of Reach 7. Attachment B-6,
Figure 6, presents the projected warmwater fish tissue PCB concentrations by reach for the
alternatives evaluated; Attachment B-6, Figure 7, presents the projected fish tissue PCB
concentrations for coldwater fish for the Reach 7 subreaches.
All alternatives would achieve warmwater fish protection in 100% of the areas. SED 5/FP 4,
SED 6/FP 4, SED 8/FP 7, SED 9/FP 8, and SED 9/FP 4 MOD would achieve coldwater fish
protection in all areas. SED 3/FP 3 would achieve the coldwater fish IMPG in all except one of
the Reach 7 subreaches, while SED 10/FP 9 would not achieve the coldwater fish IMPG in any
reach and, in effect, would provide no improvement over MNR (SED 2/FP 1).
34
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Table 9
Summary of Percent of Averaging Areas
Achieving Warmwater and Coldwater Fish Protection IMPGs
IMPCs
1
7
Achic\ inti
r
Om
Ui
5c
y.
IMPC
zc
C/3
CL
7
Insectivorous Birds (4.4 mg/kg
in prey)
33
83
100
100
100
100
58
35
8/1/2012
-------
1 All alternative combinations evaluated, with the exception of SED 2/FP 1, SED 3/FP 3, and SED
2 10/FP 9, would achieve the wood duck IMPGs at the end of the model simulation period in
3 100% of the areas. Under MNR (SED 2/FP 1) the IMPG is achieved in 33% of the averaging
4 areas, representing 265 acres of the total 720 acres. SED 10/FP 9, would achieve the IMPG in
5 58%) of the areas (381 acres), while SED 3/FP 3 would achieve the IMPG in 83%> of the
6 averaging area (573 acres). SED 9/FP 4 MOD will protect insectivorous birds by substantially
7 reducing sediment PCB concentrations that drive contaminant concentrations in the aquatic
8 portion of the diet while simultaneously reducing floodplain soil PCB concentrations that lead to
9 elevated PCBs in the terrestrial portion of the diet.
10 6.2.5 Piscivorous Birds
11 The IMPG for piscivorous birds (represented by osprey) applies to whole-body fish tissue
12 concentrations in the 14 subreaches in Reaches 5 through 8.
13 Table 11 summarizes, for each combination, the percentage of the 14 subreaches (considered the
14 averaging areas) in which the model-predicted fish concentrations would achieve the piscivorous
15 bird IMPG. SED 6/FP 4, SED 8/FP 7, and SED 9/FP 8 would achieve the osprey IMPG in 100%
16 of the areas, while SED 9/FP 4 MOD would achieve the IMPG in 10 (71%>) of the averaging
17 areas and SED 5/FP 4 would achieve the IMPG in 13 (93%>) of the 14 averaging areas. SED
18 10/FP 9, active alternative evaluated, would provide protection in none of the areas, which
19 represents no improvement over MNR, while SED 3/FP 3 would achieve the IMPG in only 6
20 (43%) of the 14 averaging areas; Attachment B-6, Figure 9, shows the same data in terms of the
21 acreage achieving the IMPG.
22 Table 11
23
24 Summary of Percent of Averaging Areas Achieving Piscivorous Bird IMPGs
impc;
1 <11/t (I IS
IVrcciK
7
of A\cr;itii
•ri
7
»•* Amis
-i-
Q-
7
u*hic\ inii 1
r-
Om
7.
\1P(; in 1 is
zc
Om
7
li Tissue
Om
A
7
Sl.D'VI l>4 MOD
Piscivorous Birds (3.2 mg/kg
in fish)
0
43
93
100
100
100
0
71
25
26 6.2.6 Piscivorous Mammals
27 Similar to insectivorous birds, the IMPGs for piscivorous mammals (represented by mink for the
28 IMPGs, although otter would likely be more affected with a less-comprehensive sediment
29 remedy because their diet is assumed to be 100%> aquatic) apply to PCB concentrations in their
30 prey, which consist of both aquatic and terrestrial animals and thus, they depend on both
31 sediment and floodplain PCB concentrations in the two designated averaging areas (Reaches
32 5A/5B and Reaches 5C/5D/6). Because each remedial combination involves a specific sediment
36
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
component and a specific floodplain component, an assessment of the achievement of the
piscivorous mammal IMPGs was made by using the model-predicted sediment endpoint
concentration in each averaging area to determine the corresponding target floodplain soil
concentration in that area that would result in achievement of the upper- and lower-bound
IMPGs, and then comparing the estimated floodplain soil EPC in that area to those target levels.
Table 12 summarizes the comparison of the post-remediation floodplain EPC in each averaging
area to the target floodplain soil level in that area, presenting the percentage of the two averaging
areas that would achieve the upper-bound and lower-bound IMPGs for piscivorous mammals.
Attachment B-6, Figure 10, presents the same data in terms of acreage achieving the two IMPGs
under each alternative. Only SED 8/FP 7 would achieve both the upper-bound and lower-bound
IMPGs in both averaging areas. SED 5/FP 4, SED 6/FP 4, and SED 9/FP 8 would all achieve the
upper-bound IMPG only in both averaging areas. SED 10/FP 9 and SED 3/FP 3 would not
achieve either IMPG in either of the areas, and therefore, provide no improvement over MNR
(SED 2/FP 1). As discussed earlier with reference to insectivorous birds, SED 9/FP 4 MOD will
achieve protection of piscivorous mammals by simultaneously reducing PCB concentrations in
both the aquatic and terrestrial dietary components. However, as remediation areas have not yet
been determined it is unknown in which averaging areas IMPGs will be achieved.
Table 12
Summary of Percent of Averaging Areas
Achieving IMPGs for Piscivorous Mammals
IMPCs
1
-------
1 omnivorous/carnivorous mammal IMPGs in 100% of the areas. Both SED 3/FP 3 and SED
2 10/FP 9 would achieve the upper-bound IMPG only in 100% of the areas, which is only a slight
3 improvement over SED 2/FP 1 (MNR), which achieves the upper-bound IMPG in 86% of the
4 averaging areas. SED 3/FP 3 would achieve the lower bound in 71% of the areas, while both
5 SED 10/FP 9 and SED 2/FP 1 would achieve the lower bound in 57% of the areas. The targeted
6 remediation of floodplain soil included in alternative SED 9/FP 4 MOD will provide some
7 protection of omnivorous mammals; however, as remediation areas have not yet been determined
8 it is unknown in which averaging areas IMPGs will be achieved.
9
10 Table 13
11
12 Summary of Percent of Averaging Areas Achieving IMPGs for
13 Omnivorous/Carnivorous Mammals
IMPCs
IV i
ri
7
itiiI of A\
Q.
7
I'mgiiig Ai
-t
•r)
7
Oils Achio in
-t
y.
» IMPGs in
r-
Om
X
A
7
l-'loodpliiin
zc
Om
7
Soil
Q.
7
Upper Bound (34.3 mg/kg in
floodplain soil)
86
100
100
100
100
100
100
Lower Bound (21.1 mg/kg in
floodplain soil)
57
71
100
100
100
100
57
14 6.2.8 Threatened and Endangered Species
15 The IMPG for threatened and endangered species (represented by the bald eagle) applies to
16 whole-body fish PCB concentrations in the 14 subreaches in Reaches 5 through 8. All
17 alternatives would achieve the threatened and endangered species IMPG in all areas.
18 6.3 SUMMARY
19 In summary, an evaluation of whether, and to what extent, each alternative would accelerate
20 attainment of the IMPGs when compared to natural processes, or in this case SED 2/FP 1, is
21 provided below.
22 For human health direct contact with floodplain soil and agricultural use, all alternatives, with
23 the exception of SED 2/FP 1, achieve the risk level by which they are defined upon completion
24 of remediation. It would not be expected under SED 2/FP 1 that any reduction in risk would
25 occur over a reasonable timeframe.
26 For human health direct contact with sediment, for SA 3 (Woods Pond) and SA 7 (Glendale
27 impoundment), which are the sediment EAs that do not currently achieve acceptable risk levels
28 due to RME noncancer risk exceeding an HI of 1, model projections indicate that within 22 years
29 the RME noncancer risk level (HI = 1) would be achieved with no action. The remaining
30 alternatives all involve active remediation in Woods Pond and all achieve an HI of 1 in shorter
31 periods of time, ranging from 21 years for SED 8/FP 7, to approximately 15 years for SED 5/FP
38
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
4 and SED 6/FP 4, and less than 10 years for SED 3/FP 3, SED 9/FP 8, SED 10/FP 9, and SED
9/FP 4 MOD.
For human fish consumption, no remediation (SED 2) would result in the HI of 1 and the RME
10"4 level being exceeded for the RME and CTE adult and child for greater than 250 years. The
same is the case with SED 10/FP 9 for the HI of 1 and the RME 10"4 level; however, the
CTE 10"4 risk level is achieved in some reaches. All other alternatives achieve varying risk
levels far sooner than no action (see Table 1).
For benthic invertebrates, numerous EAs meet the upper-bound IMPG with no action; however,
very few EAs attain the lower-bound IMPG maximum acceptable threshold concentration
(MATC) within less than 200 years, with a similar pattern for SED 10/FP 9. SED 6/FP 4, SED
8/FP 7, SED 9/FP 8 and SED 9/FP 4 MOD all achieve the lower-bound IMPG, or its equivalent
in the case of SED 9/FP 4 MOD in Reach 5B, in all EAs within 20 years (with the exception of
some EAs, which take longer due to the duration for implementation of SED 8).
The attainment of IMPGs for amphibians is evaluated for both vernal pools in the floodplain and
also for EAs in the backwaters. Neither SED 2/FP 1 nor SED 10/FP 9 achieve either the upper
bound or lower bound in the majority of areas or pools within a long timeframe (likely more than
100 years). The active alternatives achieve either the upper bound or lower bound in many or all
areas or pools in a much more rapid timeframe, and for alternatives SED 6/FP 4 and SED 9/FP 8,
typically in less than 20 years. SED 9/FP 4 MOD provides protection to amphibians by reducing
exposure concentrations and through an iterative decision-tree approach to remediating vernal
pools.
Warmwater fish IMPGs are attained for all alternatives as well as for no action. However, the
coldwater fish IMPGs are not attained in the subreaches of Reach 7 either with no action or with
SED 10/FP 9, typically within a timeframe greater than 100 years. Alternatives that include
active remediation attain this IMPG in all but one subreach (in SED 3/FP 3) within a range of
timeframes dependent on the implementation schedule for the alternative.
The IMPG for insectivorous birds is not attained in 8 of 12 EAs with no action (represented by
SED 2/FP 1), and are not attained in 5 of 12 areas with SED 10/FP 9. For other alternatives,
most achieve the IMPG in all areas. It is not expected that there are processes occurring that will
result in changes in concentration in these areas in a reasonable timeframe because these
concentrations are derived from a combination of SED and FP concentrations.
Piscivorous bird IMPGs are not met for no action or for SED 10/FP 9 for any reach in most cases
in 100 to over 200 years, depending on the reach. SED 6/FP 4, SED 8/FP 7, and SED 9/FP 8 all
achieve the IMPG in all reaches in a much reduced timeframe, typically less than 20 years with
the exception of SED 8/FP 7, for which timeframes are controlled by the longer duration of
implementation.
The lower-bound IMPG for piscivorous mammals is achieved only in SED 8/FP 79. However,
SED 5/FP 4, SED 6/FP 4, SED 8/FP 7, and SED 9/FP 8 achieve the upper-bound IMPG. The
other alternatives do not achieve either IMPG. No action (SED 2/FP 1) would result in this
IMPG not being achieved for over 250 years.
9 This IMPG is based on both the sediment and floodplain prey concentration for mink because it is based on a
blended diet. For otter, it would be based solely on the sediment prey concentration.
39
8/1/2012
-------
1 With no action, the omnivorous/carnivorous mammal upper-bound IMPG is not met in three of
2 the seven EAs, with two achieving the lower-bound IMPG. All other alternatives achieve either
3 the upper-bound or lower-bound IMPG, with SED 5/FP 4, SED 6/FP 4, SED 8/FP 7, and SED
4 9/FP 8 all achieving the lower-bound IMPG.
5 The threatened and endangered species IMPG (based on the bald eagle) is achieved with no
6 action and therefore, for all other alternatives. However, it should be noted that for many guilds,
7 the IMPG for another receptor group may be more applicable (e.g., salamanders).
8 7 REDUCTION OF TOXICITY, MOBILITY, OR VOLUME OF WASTES
9 The degree to which the combinations of sediment and floodplain alternatives under evaluation
10 would reduce the toxicity, mobility, and volume (TMV) of PCBs is discussed below.
11 Reduction of Toxicity. None of the sediment-floodplain alternative combinations includes any
12 treatment processes that would reduce the toxicity of PCBs in the sediment or soil. Accordingly,
13 this factor does not provide a basis for distinguishing among the alternatives.
14 Reduction of Mobility. Under SED 2/FP 1, no reduction of mobility of PCBs in the river would
15 be achieved through remedial action, and only past and ongoing upstream source
16 control/remediation and naturally occurring processes would provide for a reduction of PCB
17 mobility. Under all other combinations, reductions would be achieved through sediment removal,
18 capping, backfilling, thin-layer capping, and/or bank stabilization activities. Reduction in PCB
19 mobility can be viewed in terms of reduction in the annual mass of PCBs passing Woods Pond
20 and Rising Pond Dams, and the solids/PCB trapping efficiency of Woods Pond shown in
21 Attachment B-6, Figures 1 and 12.
22 The percent reduction in PCB mass passing over Woods Pond and Rising Pond Dams at the
23 conclusion of the 52-year (81-year in the case of SED 8/FP 7) model simulation period for each
24 of the alternative combinations evaluated is shown in Table 2 and discussed with reference to the
25 General Standard "Control of Sources of Releases" in Section 11.3.
26 Attachment B-6, Figure 12, shows the solids trapping efficiency of Woods Pond at the
27 conclusion of each of the alternative combinations evaluated. As indicated in this figure,
28 alternative combinations that include deepening of Woods Pond (SED 9/FP 8, SED 9/FP 4
29 MOD, and SED 10/FP 9) achieve modest, and nearly equivalent, increases in solids trapping in
30 the pond, increasing the trapping of solids from approximately 15% for MNR and alternatives
31 that do not include the deepening of Woods Pond, to approximately 25% in the case of SED 9/FP
32 8 and SED 10/FP 9, and to approximately 30% in the case of SED 9/FP 4 MOD. It is important
33 to note, however, that because of continuing release of PCBs from the trapped sediment, the PCB
34 trapping efficiency will be less than that for solids, although this effect will be similar for all
35 alternatives and therefore, does not distinguish among them.
36 Reduction of Volume: Implementation of each of the sediment-floodplain alternative
37 combinations except SED 2/FP 1 would reduce the volume of PCB-containing sediment, bank
38 soil, and floodplain soil in the Rest of River through permanent removal of the material. Table
39 14 from GE's RCMS and Attachment B-6, Figure 13, summarize the approximate removal
40 volume and corresponding PCB mass that would be removed under each such combination. The
41 volume and mass removed under the SED 9/FP 4 MOD alternative, which was not evaluated in
42 GE's RCMS, are also shown in this table.
40
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Table 14
Removal Volume and Corresponding PCB Mass for Combinations of Alternatives
Alli-rnali\c
Kcmn\;il Volume-
Sodiiiicnl/Soil
(c\)
Islimaled PCIJ Miiss
(III)
SED 2/FP 1
—
—
SED 3/FP 3
243,000
21,700
SED 5/FP 4
533,000
33,300
SED 6/FP 4
677,000
37,300
SED 8/FP 7
2,902,000
94,100
SED 9/FP 8
1,098,000
53,100
SED 10/FP 9
267,700
13,900
SED 9/FP 4 MOD
990,000
46,970
8 SHORT-TERM EFFECTIVENESS
Evaluation of the short-term effectiveness of the sediment-floodplain alternative combinations
includes consideration of the short-term impacts of implementing these combinations on the
environment (considering both ecological effects and increases in greenhouse gas [GHG]
emissions), on local communities (as well as communities along transport routes), and on the
workers involved in the remedial activities. Short-term impacts are those that would occur during
and immediately after the performance of the remedial activities in a given area. Because SED
2/FP 1 would involve no remedial construction activities, its implementation would not produce
any short-term impacts; all of the other combinations would have short-term impacts. Because
any remediation would be conducted using a phased approach, these impacts would be spread
out over the remedial action period and area, and thus, would not last for the entire duration of
the project in all affected areas. The tables shown in this section were taken from GE's RCMS
and modified where possible to include SED 9/FP 4 MOD. The estimated durations of the
combinations of alternatives evaluated are summarized in Table 15. As the table shows, the
durations range from 5 years for SED 10/FP 9 to over 50 years for SED 8/FP 7.
Table 15
Construction Duration for Alternative Combinations
Sl l) 21
IP 1
Sl l) 3/
IP 3
SI. 1)5/
IP 4
SKI) ۥ/
IP 4
Sl l) HI
IP 7
Sll)')/
IP S
Sll) 10/
I P ')
Sll)')/
IP 4
MOD
Construction
Duration
(years)
10
18
21
52
14
5
13
41
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
8.1 IMPACTS ON THE ENVIRONMENT - EFFECTS WITHIN THE REST OF RIVER
AREA
Short-term impacts on the Rest of River environment from remedial construction activities
would include PCB releases to the water column and air during sediment removal and other in-
river activities, as well as alteration of the various habitats where remediation work would be
conducted or support facilities would be built, with the attendant impacts on the plants and
animals that use those habitats. These impacts are described and compared among the
combinations of alternatives in the following subsections.
PCB Releases: Sediment removal activities would result in some resuspension of PCB-
containing sediment into the water column. This could potentially result in a transient increases
in PCB levels in aquatic biota downstream of the removal operations. Under all of the active
remediation combinations except SED 9/FP 8 and SED 9/FP 4 MOD, sediment removal in
Reach 5A and, where applicable, Reach 5B, would be conducted in the dry using sheetpile
containment, which would allow the greatest control of resuspension. However, the potential still
exists for suspended or residual sediment containing PCBs to be released from the work area
both during sheetpile installation and removal, and during a high-flow event when overtopping
of the sheeting could occur. Under SED 9/FP 8 and SED 9/FP 4 MOD, sediment removal in
those subreaches would be conducted in the wet, which would have the potential for causing
resuspension of PCB-containing sediment. In addition, under combinations of remedial
alternatives that would involve sediment remediation in other reaches, removal activities would
be conducted in the wet from barges. These activities, as well as boat and barge traffic, would
result in some resuspension of sediment containing PCBs, which would be minimized through
the use of engineering controls, such as silt curtains and/or BMPs.
Other than SED 2/FP 1, which does not involve sediment removal, SED 3/FP 3 has the lowest
potential for PCB resuspension because it would involve the smallest area of sediment removal
(42 acres in Reach 5A), and that removal would be conducted in the dry. SED 10/FP 9 would
involve a smaller area of dry removal (20 acres in Reach 5A), but would also involve the
removal of sediment in the wet from 42 acres in Woods Pond. The other alternatives would
involve substantially more sediment removal, with much of it conducted in the wet, which would
result in more resuspension than SED 3/FP 3 and SED 10/FP 9 over a longer time period.
Similarly, sediment and soil removal and related processing activities have the potential to
produce airborne PCB emissions that could impact downwind communities. This potential also
increases with the scope and duration of the removal activities, which increase from SED 3/FP 3
and SED 10/FP 9 through SED 8/FP 7. Monitoring, combined with the implementation of
BMPs, are expected to result in minimal releases.
Impacts on Aquatic Riverine Habitat: The potential short-term impacts of sediment remediation
activities, including removal with capping or backfilling and capping or thin-layer capping
without removal, on aquatic riverine habitat include: removal of the habitat used by aquatic
plants, benthic invertebrates, and fish; change in surface substrate from its current condition
(sand, sand and gravel, or silt) to armor stone or backfill material; removal or burial of most, if
not all, vegetation, benthic invertebrates, and other organisms present in the sediment; disruption
and displacement of fish; alteration of habitat for birds and mammals living adjacent to the river
that feed in areas subject to remediation; and possible colonization by invasive species. In
42
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
addition, capping or thin-layer capping without removal would raise the elevation of the river
bottom, which, in shallower areas, could change the vegetative characteristics of those areas and
the biota dependent on them.
Under SED 3/FP 3, these types of potential short-term impacts would occur over 42 acres of
aquatic riverine habitat, all in Reach 5A. Under SED 9/FP 4 MOD, remediation would be 42
acres in Reach 5A and 57 acres in Reach 5C, for a total of 99 acres of riverine habitat. Under
SED 5/FP 4, SED 6/FP 4, SED 8/FP 7, and SED 9/FP 8, these impacts would occur over
approximately 127 acres of aquatic riverine habitat. Under SED 10/FP 9, which involves the
smallest amount of removal of contaminated sediment, these impacts would occur in only 20
acres of such habitat (in Reach 5A).
Incorporation of a habitat layer in the cap design would mitigate some of these impacts. In
addition, implementation of the remediation in a phased approach affecting a small area at any
given time would also minimize some of these impacts.
Impacts on Riverbank Habitat: The potential short-term impacts of bank stabilization activities
in Reaches 5A and 5B on the riverbanks include removal of trees, other vegetation, and woody
debris from the riverbanks, with the resulting temporary loss of shading for the river and the loss
of the wildlife that use those features; short-term elimination of vertical and undercut banks used
by various species for nesting; short-term loss of slide and burrow habitat for muskrats and
beavers; potential short-term reduction in wildlife access routes and movement of various species
between their aquatic and terrestrial habitats; and the possible colonization by invasive species.
All of the sediment-floodplain alternative combinations except SED 2/FP 1 (MNR) and SED
10/FP 9 would result in such impacts on the eroding riverbanks subject to stabilization. SED
2/FP 1 would not have any such impacts, and SED 10/FP 9 would limit these impacts to a small
portion of riverbank in Reaches 5A and 5B. The approach to bank remediation in SED 9/FP 4
MOD is based on consideration of both the erosion potential of areas of bank as well as the PCB
concentrations in bank soil, reducing the amount of bank remediation by focusing only on those
portions of the banks in Reach 5A that have both high erosion potential and high PCB
concentration, and in Reach 5B on a limited amount of bank soil with the highest PCB
concentrations.
Impacts on Impoundment Habitat. The potential short-term impacts of sediment remediation
activities, including removal with capping (or backfilling), capping or thin-layer capping without
removal, and removal without capping, on impoundment habitat are similar to the short-term
impacts on aquatic riverine habitat, as described above, except that placement of a cap or thin-
layer cap in the deep hole portion of Woods Pond would not be expected to have any significant
short-term ecological impacts.
Apart from SED 2/FP 1, all of the sediment-floodplain alternative combinations under evaluation
would have some impacts on impoundment habitat. Table 5 shows the amount of area affected
by each combination of alternatives.
Impacts on Backwater Habitat: The potential short-term impacts of sediment remediation
activities, including thin-layer capping and sediment removal with capping (or backfilling), on
backwater habitat include: burial or removal of most, if not all, vegetation, benthic invertebrates,
and other organisms in the sediment.
43
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Because SED 2/FP 1, SED 3/FP 3, and SED 10/FP 9 would not involve any remediation in the
backwaters, they would have no short-term impacts to backwater habitat. The other alternative
combinations would all have such impacts because they would affect 61 to 86 acres of such
habitat (see Table 5).
Impacts on Floodylain Habitats: The potential short-term impacts on the various floodplain
habitats resulting from floodplain soil removal and the construction and use of access roads and
staging areas include the following:
¦ For floodplain wetland forest habitats, the short-term impacts could potentially
include: (1) removal of living trees, shrubs, and other vegetation, as well as dead tree
snags and downed woody debris, which would result in a temporary loss of cover,
nesting, and feeding habitat for wildlife species that rely on forested floodplains;
(2) possible colonization by invasive plant species; and (3) increase in construction
and equipment traffic, which could disrupt some forest animals or result in mortality to
certain slow-moving smaller animals. Many of these short-term impacts can be
mitigated by appropriate restoration activities, including replacement of existing soil
and leaf litter with backfill soil designed to function similarly to existing native soil, to
provide the best opportunity for plant growth and hydraulic conductivity, and
implementing an invasive species management program.
¦ For shrub and emergent wetlands (both shallow and deep), the short-term impacts
could potentially include: (1) clearing of vegetation, with consequent impacts on
nesting, burrowing, and/or escape habitat and food for birds, amphibians, reptiles,
mammals, and invertebrates that use these wetland areas; (2) alteration of the
hydrology of the wetlands; (3) possible colonization by invasive species; and
(4) increase in construction and equipment traffic, with the resulting potential for
disruption or mortality to slow-moving animals. Many of these short-term impacts can
be mitigated by appropriate restoration activities, including replacement of existing
soil with soil designed to function similarly to existing native soil, to provide the best
opportunity for plant growth and hydraulic conductivity and implementing an invasive
species management program.
¦ For vernal pools and the biota that use them, the short-term impacts could potentially
include: (1) removal of amphibian and invertebrate eggs, larvae, or adults in the
affected portions of the pools; (2) removal of physical components of the pools
(organic surface soil, vegetation, and other organic materials) and their replacement;
(3) alteration of the hydrology of the pools; (4) tree clearing within and adjacent to the
pools, temporarily reducing the shade and infusion of biomass provided to the pools;
(5) temporary loss of obligate vernal pool breeding species from all or parts of these
pools; (6) possible colonization by invasive species; (7) impacts on the non-breeding
terrestrial habitats surrounding the vernal pools; and (8) loss or fragmentation of
landscape connectivity among networks of vernal pools and between vernal pools and
non-breeding habitats. Many of these short-term impacts can be mitigated by
appropriate restoration activities, including replacement of preexisting physical
components such as woody debris, implementing an invasive species management
program, and conducting remediation in a phased approach.
44
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
¦ For upland habitats, the short-term impacts would potentially include temporary loss
of trees and associated vegetation and impacts to the wildlife that use such areas.
¦ In all of these habitats, and in the absence of any mitigation, the short-term impacts
would potentially include the direct removal or disruption of any state-listed species
present in the affected areas, as well as alteration of their habitat.
¦ The short-term impacts could potentially also include impairment of a number of other
functions provided by the floodplain, which would be mitigated through proper
restoration. For example, by removing woody debris and vegetation and altering
microtopography in disturbed areas, the floodplain remedial construction activities
would reduce the floodplain roughness that produces flow resistance and thus,
contributes to the important flood flow alteration function of the floodplain. In
addition, the construction activities could alter the floodplain's groundwater
recharge/discharge function and its functions of water quality maintenance, nutrient
process, and production export.
All of the combinations of sediment and floodplain alternatives involving removal would have
these potential short-term impacts on the habitats outside the river. Table 5 shows the amount of
each habitat type potentially impacted by each combination of alternatives.
With specific reference to vernal pools, SED 2/FP 1 (MNR) and SED 10/FP 9 (which does not
include remediation of contaminated soil in vernal pools) would have no direct impact on any of
the vernal pools. All of the other alternative combinations, with the exception of SED 9/FP 4
MOD, would impact those vernal pools to a generally similar extent. Because of the iterative
pilot-study-based approach to remediation/restoration of vernal pools included in the SED 9/FP 4
MOD alternative, it is not possible to provide similar quantitative information relative to these
two zones. However, the vernal pool component of SED 9/FP 4 MOD was designed specifically
to provide superior performance with regard to vernal pools, comprehensively considering both
the positive and negative impacts of active remediation.
8.2 CARBON FOOTPRINT - GHG EMISSIONS
Estimates have been developed of the GHG emissions (i.e., carbon footprint) anticipated to occur
through sediment removal/capping, floodplain soil and tree removal, and related ancillary
activities during the implementation of the sediment-floodplain alternative combinations under
evaluation. Table 16 summarizes the total carbon footprint associated with each combination,
including a breakdown of direct, indirect, and off-site emission sources. To provide context
regarding the emissions reported below, the number of passenger vehicles that would emit an
equivalent quantity of C02-eq in 1 year is also presented in the table. A graphical comparison of
the total GHG emissions for the alternatives evaluated is shown in Attachment B-6, Figure 14.
SED 10/FP 9 would have the lowest amount of total GHG emissions (40,000 tonnes); SED 3/FP
3 would have the next lowest amount (47,000 tonnes); SED 5/FP 4, SED 6/FP 4, and SED 9/FP
45
8/1/2012
-------
1 8 would have between 100,000 and 190,000 tonnes of such emissions; and SED 8/FP 7 would
2 have by far the greatest amount of GHG emissions (520,000 tonnes).10
3 Table 16
4
5 Calculated GHG Emissions Anticipated to Result from Combinations of Sediment
6 and Floodplain Alternatives
Allcriiiili\c
l oliil (.IK.
l-lmissions
(tonnes)
Dirccl
Emissions
(lOIIIICS)
Indirect
Emissions
((OIIIK'N)
orr-siic
r.inissions
(lOIIIICS)
No. of Vehicles
with l.(|iii\;ik'ii(
Anniiiil
1-'. missions
SED 2/FP 1
...
...
...
...
—
SED 3/FP 3
47,000
26,000
1,200
20,000
9,000
SED 5/FP 4
100,000
46,000
2,300
53,000
19,100
SED 6/FP 4
140,000
65,000
3,500
72,000
28,800
SED 8/FP 7
520,000
220,000
10,300
290,000
99,400
SED 9/FP 8
190,000
79,000
3,800
110,000
36,300
SED 10/FP 9
40,000
12,000
900
27,000
7,600
SED 9/FP 4
MOD
171,000
70,000
3,400
98,000
32,200
7
8 8.3 IMPACTS ON LOCAL COMMUNITIES AND COMMUNITIES ALONG TRUCK
9 TRANSPORT ROUTES
10 Implementation of all combinations of sediment and floodplain alternatives (except MNR, SED
11 2/FP 1) would result in some short-term impacts to the local communities along the Housatonic
12 River. These short-term effects would include changes to the visual appearance of the river,
13 riverbanks, and affected areas of the floodplain, as well as disruption of recreational activities in
14 those areas due to the remediation as well as the construction of access roads and staging areas.
15 They would also include increased construction traffic, noise, and nuisance dust in those areas.
16 Construction activities would affect some recreational activities along the river and in the
17 floodplain. Depending on the particular combination of alternatives, these potentially would
18 include fishing, canoeing (including canoe launches), hiking, dirt biking, general recreation, and
19 both waterfowl and other game hunting. During the period of active construction, restrictions on
20 recreational uses of the river and floodplain would be imposed in the areas where remediation-
21 related activities are taking place. Due to safety considerations, boaters, anglers, hikers, hunters,
22 and other recreational users would not be able to use the river, floodplain, or riverbank in the
10 Comparison among the three emission categories indicates that, on average, off-site emissions account for more
than half of the GHG emissions for each combination (the most significant off-site sources being steel sheeting
manufacture [with the exception of SED 9] and production of cement to be used in sediment stabilization). Direct
emissions sources (including those associated with construction and transportation activities) generally account
for 40 to 50% of the total GHG emissions.
46
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
construction and support areas. However, due to the phased nature of any remediation, only a
small portion of the total recreational acreage would be affected at any one time, and active
measures to decrease impacts to recreation (e.g., providing for portaging canoes around the area
being impacted) will be considered.
The extent of these impacts on the Housatonic River and floodplain use would vary depending
on the overall area affected by remediation and support facility construction, as well as the length
of time required to complete the remediation. (As noted above, although these impacts would not
last for the entire duration of the project in all affected areas, the total implementation duration
represents the overall time period over which short-term impacts would occur in some portion of
the Rest of River area.) These impacts would be least for SED 10/FP 9 (91 acres, 5 years). They
would be more extensive for SED 3/FP 3 (237 acres, 10 years), SED 9/FP MOD (300 to 400
acres, 13.4 years), SED 5/FP 4 (410 acres, 18 years), SED 6/FP 4 (447 acres, 21 years), and SED
9/FP 8 (469 acres, 14 years). The combination with the greatest potential impact on these uses of
the river and floodplain is SED 8/FP 7 (774 acres, 52 years)11.
In addition, due to the need to deliver equipment to the work areas, remove excavated materials,
and deliver capping, backfill, and bank stabilization materials to the site, both on-site and local
(off-site) truck traffic would increase over current conditions. This additional traffic could
increase the likelihood of accidents, noise levels, emissions of vehicle/equipment exhaust, and
nuisance dust to the air, and would persist over the duration of remedial activities. Table 17
summarizes the number of truck trips associated with transporting excavated materials from the
staging areas to the disposal or treatment facilities and delivering capping/backfill and bank
stabilization materials to the remediation areas. The total annual truck trips and total years of
truck traffic for each alternative are show graphically in Attachment B-6, Figure 15.
As shown in this table, apart from SED 2/FP 1, SED 10/FP 9 would involve the fewest number
of total truck trips (31,600) and SED 3/FP 3 would involve the next fewest (49,700). SED 5/FP
4, SED 6/FP 4, SED 9/FP 4 MOD, and SED 9/FP 8 would involve between 115,500 and 188,400
truck trips; and SED 8/FP 7 would require by far the most total truck trips (approximately
515,000). However, on an annual basis, SED 9/FP 8 would involve the greatest number of truck
trips per year (13,500) based on its accelerated schedule with work occurring in more than one
reach at a time.
11 EPA does not believe that the infrastructure included in these estimates by GE has been optimized and expects
that, for the selected remedy, the staging areas and roads will be designed to minimize the footprint and adverse
impacts to the floodplain, neighborhoods, and local roads while allowing for the remediation to proceed in a
timely and effective manner.
47
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Table 17
Estimated Truck Trips for Removal of Excavated Material and
Delivery of Capping/Backfill Material for Combinations
of Sediment and Floodplain Alternatives
\IUTii;ili\c
Truck Trips lor
l.\c;i\;ik'd Miild'hiT
Truck Trips lor
Clipping/liiickfill
M iiloriiil1'
1 oliil Truck Trips'
SED 2/FP 1
...
—
...
SED 3/FP 3
20,100 (2,000)
29,600 (3,000)
49,700 (5,000)
SED 5/FP 4
44,300 (2,500)
71,200 (4,000)
115,500 (6,500)
SED 6/FP 4
56,100 (2,700)
80,500 (3,800)
136,600 (6,500)
SED 8/FP 7
242,000 (4,700)
273,300 (5,300)
515,300 (10,000)
SED 9/FP 8
90,800 (6,500)
97,600 (7,000)
188,400 (13,500)
SED 10/FP 9
22,200 (4,400)
9,400 (1,900)
31,600 (6,300)
SED 9/FP 4 MOD
81,700 (6,100)
68,800 (5,100)
150,500(11,200)
a Truck trips estimated assuming 20-ton capacity trucks for hauling excavated material and 16-ton trucks for local
hauling of capping/backfill material. Note that many of these truck trips will not take place on public roads, and
will be on a network of on-site roads constructed specifically for the purposes of remediation.
b Capping material includes cap, thin-layer cap, backfill, and bank stabilization materials.
0 The number in parentheses represents average annual truck trips.
The additional truck traffic would also increase the risk of traffic accidents along transport
routes. The number of injuries or fatalities from the increased off-site truck traffic that would be
12
associated with the combinations of sediment and floodplain alternatives under evaluation is
summarized in Table 18, with the annual incidence of injuries and fatalities.
The incidence of potential injuries from accidents associated with increased truck traffic would
be lowest for SED 10/FP 9 (1.09 injuries), with estimated injuries for the other alternatives
ranging from 1.98 (SED 3/FP 3) to 11.0 (SED 8/FP 7). Similarly, estimated fatalities due to
increased truck traffic are lowest for SED 10/FP 9 (0.05), with estimated fatalities for the other
alternatives ranging from 0.09 (SED 3/FP 3) to 0.51 (SED 8/FP 7).
12 This analysis quantified transport-related risks only for trucks used to import capping, backfill, and bank
stabilization materials to the site over public roads, as well as to dispose of materials used for the staging areas
and access roads following completion of remediation. The risks from transporting excavated materials to the
staging areas are evaluated as part of risks to workers, discussed below; and the risks from transporting such
materials from the staging areas to local or off-site disposal or treatment facilities are evaluated as either worker
risks or traffic accident risks under the relevant treatment/disposition alternatives.
48
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Table 18
Incidence of Accident-Related Injuries/Fatalities Due to Increased Truck Traffic
Impacts
Sl l) 2/
1 PI
Sl l) 3/
IP 3
SI. 1)5/
IP 4
SKI) f./
IP 4
Sl l) X/
IP
Sll)')/
IP s
Sll) 10/
IP <)
Sll)')/
IP 4
MOD
Non-Fatal Injuries
Number
—
1.98
3.29
4.03
11.0
5.43
1.09
5.36
Average
Annual
Number
—
0.21
0.18
0.19
0.21
0.40
0.21
0.40
Probability*
—
0.86
0.96
0.98
1.00
1.00
0.66
1.00
Fatalities
Number
—
0.09
0.15
0.19
0.51
0.25
0.05
0.25
Average
Annual
Number
—
0.010
0.008
0.009
0.010
0.019
0.010
0.019
Probability*
—
0.09
0.14
0.17
0.40
0.22
0.05
0.22
Probability indicates the probability of at least one injury/fatality.
8.4 POTENTIAL MEASURES TO AVOID, MINIMIZE, OR MITIGATE SHORT-TERM
COMMUNITY IMPACTS
A number of measures would be employed in an effort to avoid, minimize, and mitigate potential
detrimental effects of construction activities on the affected communities (e.g., minimize truck
travel on local roads). As would be expected, the level of impact, and therefore, the extent of the
necessary mitigation, is related to the scale/scope of the alternative and the time period of
construction. Therefore, SED 8/FP 7 would have the most significant effect on local
communities and would require the greatest degree of mitigation. SED 10/FP 9 would have the
least such effect.
8.5 RISKS TO REMEDIATION WORKERS
There would be health and safety risks to site workers implementing each of these combinations
of sediment and floodplain alternatives. An estimate of the injuries or fatalities to workers from
implementation of the alternative combinations is summarized in Table 19.
Risks to site workers would be lowest with SED 10/FP 9 (2.6 injuries), with the estimated
injuries for all other alternatives at least twice that of SED 10/FP 9, ranging from 5.5 (SED 3/FP
3) to 30.2 (SED 8/FP 7). Similarly, estimated fatalities for site workers are lowest for SED 10/FP
9 (0.03), with estimated fatalities for the other alternatives ranging from 0.05 (SED 3/FP 3) to
0.34 (SED 8/FP 7).
49
8/1/2012
-------
1 Table 19
2
3 Incidence of Accident-Related Injuries/Fatalities Due to
4 Implementation of Sediment-Floodplain Alternative Combinations
Impacts
Sl l) 2/
IP 1'
Sl l) 3/
IP 3
SI. 1)5/
IP 4
si:i)(./
IP 4
Sl l) X/
IP 7
Sll)')/
IP s
Sll) 10/
ip y
Sll)')/
IP 4
MOD
Lahni'-lkiui's
(hours)
504
1 ,<.()
Vxi,-^x
:x5,i<)(,
1, ,
Duration (yrs)
-
10
18
21
52
14
5
13
Non-Fatal Injuries
Number
-
5.5
9.9
10.7
30.2
10.9
2.6
9.2
Average
Annual
Number
—
0.55
0.55
0.51
0.58
0.78
0.53
0.69
Probability13
-
1.00
1.00
1.00
1.00
1.00
0.93
1.00
Fatalities
Number
-
0.05
0.11
0.11
0.34
0.13
0.03
0.10
Average
Annual
Number
—
0.005
0.006
0.005
0.007
0.009
0.005
0.007
Probability13
-
0.05
0.10
0.11
0.29
0.12
0.03
0.10
5 a While the monitoring activities under SED 2 would involve the potential for accidents to site workers involved in
6 those activities, these risks would be minimal, and would be mitigated through implementation of health and
7 safety measures similar to those successfully applied during such activities on the river in the past.
8 b Probability indicates the probability of at least one injury/fatality.
9 9 IMPLEMENTABILITY
10 9.1 TECHNICAL IMPLEMENTABILITY
11 The equipment, materials, technology, procedures, and personnel necessary to implement and
12 monitor the effectiveness of the combinations of sediment and floodplain alternatives are all
13 readily available.
14 All of these combinations would be implemented using well-established and available in-river
15 remediation and floodplain soil removal methods and equipment, available construction
16 technologies to build land-based support facilities, and readily available methods to implement
17 monitoring and ICs. The remedial components selected (i.e., sediment removal in the dry or wet
18 via mechanical or hydraulic methods, sediment capping and thin-layer capping, floodplain soil
19 removal and backfilling, and MNR) have been used in similar applications as part of previous
20 work at the GE-Pittsfield/Housatonic River Site and at many other sites.
21 The individual components of these combinations are considered technically implementable, as
22 shown by previous work conducted at the site, including the ^-Mile and 1 '/2-Mile removal
50
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
actions, which included many of the components of the combination alternatives. Although
information regarding remedies at other sediment sites indicates that there have been a limited
number of dredging/removal projects of the magnitude of the largest of the combinations being
considered here (i.e., SED 8/FP 7), the techniques being used are considered readily scalable and
adaptable to the size and setting of the Rest of River. In addition, the Rest of River, in some
cases, provides much more land area for staging and access roads, and generally does not pass
through highly urbanized areas, which was an important factor in implementing the ^-Mile and
lV2-Mile removal actions. As a result, although implementation of these combinations would
likely involve different complications and uncertainties as compared with those from the
previous removal actions, these complications are not expected to be insurmountable.
Some of these potential uncertainties include: difficulties associated with contracting over long
time periods; uncertainties in obtaining the large quantities of capping and backfill materials
(which would range from approximately 308,000 cubic yards (cy) to approximately 2.9 million
cy, as shown in Table 20 from GE's RCMS below); greater potential for impacts from releases
during implementation given the long timeframes of some alternatives; and uncertainties in the
availability of landfill capacity or treatment capabilities (depending on the treatment/disposition
alternative selected). However, many of these challenges have been overcome at other sites, and,
in addition, the concept of adaptive management would be used to address these uncertainties by
reassessing the effectiveness of the selected option at regular intervals as it is implemented.
Additional corrective actions, such as repairs, if necessary, should provide the same
implementation challenges for all active alternatives.
In addition, habitat restoration techniques are available and have been used successfully at other
sites. Restoration can reliably reestablish pre-remediation conditions for these habitats over the
timeframes of the various alternatives, which range from 5 to 52 years, using a phased approach.
Post-remediation monitoring and maintenance will ensure that the selected restoration techniques
reestablish the prior conditions and functions of the affected habitats.
Table 20
Required Capping/Backfill/Stabilization Material Volumes
for Combinations of Sediment and Floodplain Alternatives
Comhiiiiilinn
Siind
(CJ )
Armor Mom*/
Kipnip
(O )
Soil liiickrill
(CJ )
l oliil Miilcriiil
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
9.2 ADMINISTRATIVE IMPLEMENTABILITY
Implementation of all combinations, except SED 2/FP 1, would require GE to obtain permission
for access to the properties where the work would be conducted or where the support facilities
would be located. Although many of these properties are owned by the Commonwealth or the
City of Pittsfield (which have agreed to allow access in the Consent Decree), it is anticipated that
access agreements would be required from numerous other property owners - up to
approximately 35 such landowners for SED 10/FP 9, 35 to 45 for SED 3/FP 3, 35 to 50 for SED
9/FP 4 MOD, 40 to 50 for SED 5/FP 4, 50 to 60 for SED 6/FP 4 and SED 9/FP 8, and 80 to 95
for SED 8/FP 7. Obtaining access to all these properties for the type of work and length of time
that may be needed would require negotiations with land owners; however, this is feasible given
the timeframe over which the work would be accomplished (5 to 52 years). In contrast to other
more extensive alternatives, SED 9/FP 8 and SED 9/FP 4 MOD may have an advantage in this
respect due to the remediation method (no sheetpile, no large cranes, less clearing, and smaller
access roads), requiring less extensive agreements with land owners in Reaches 5 A and 5B.
Finally, while all of the combinations would include coordination with EPA and state agencies in
implementation of biota consumption advisories and other ICs (e.g., environmental restrictions
and easements [EREs] and conditional solutions), obtaining access to state-owned lands, and
public/community outreach programs, the alternatives with a greater extent of remediation and a
longer implementation time would likely require more extensive and prolonged coordination
activities.
10 COST
The estimated costs for each of the combinations of sediment and floodplain alternatives
evaluated, including total capital costs, estimated annual OMM costs, and total estimated present
worth costs, are summarized in Table 21. The total costs for these combinations of sediment and
floodplain alternatives (without considering treatment/disposition costs) range from $5 million
(for MNR, SED 2/FP 1) to $917 million (most extensive remediation option, SED 8/FP 7). The
costs are based on information available at the time of the estimate and are based on GE's cost
estimates provided in GE's RCMS and are summarized in Attachment B-10. EPA generally
believes that GE may have under-estimated all costs. However, because all costs were estimated
by the same methodology, they are acceptable for comparing costs relative to each alterative,
including the proposed alternative. In addition, the actual costs of remediation depend on many
variables, including the quantity of material removed, disposal fees, health and safety
regulations, ARARs, actual labor, equipment, fuel and material costs, and the final project scope.
52
8/1/2012
-------
1 Table 21
2
3 Cost Summary for Combinations of Sediment and Floodplain Alternatives
l olill ( osl
Sl l) 2/
IP 1
Sl l) 3/
IP 3
sr. D5/
IP 4
SF.I)(./
IP 4
Sl l) X/
I P "•
Sll)')/
IP X
Sll) 10/
I P ')
SKI)')/
IP 4
MOD
Capital ($ M)
0
166.4
307.5
384.6
899.7
380.8
83.5
316
OMM ($ M)
5.0
10.2
11.9
12.8
17.0
12.9
10.0
12.0
Total ($ M)
5.0
177
319
397
917
394
93.5
328
Present Worth
($M)
1.8
133
193
219
300
251
78
229
4 Notes:
5 1. All costs are in 2010 dollars. $ M = million dollars.
6 2. Total capital costs are for engineering, labor, equipment, and materials associated with implementation.
7 3. Total OMM costs include costs for post-construction inspections and repair activities (if necessary) and for
8 the maintenance of institutional controls.
9 4. Total present worth cost is based on using a discount factor of 7%, considering the length of the
10 construction period and an OMM period of 100 years on a reach-specific basis.
11 5. Estimates do not include costs for treatment or disposition of any soil/sediment removed; those costs are
12 outlined in Section 11.10.
13 11 COMPARATIVE ANALYSIS OF TREATMENT/DISPOSITION
14 ALTERNATIVES
15 This section presents a comparative evaluation of the five alternatives for treatment and/or
16 disposition of excavated contaminated river sediment and floodplain soil that were presented in
17 GE's RCMS, plus an additional alternative that was developed jointly by EPA in consultation
18 with the states of Massachusetts and Connecticut subsequent to the RCMS. The alternatives
19 were evaluated using the same criteria that were used for the sediment/floodplain combination
20 alternatives.
21 This comparative analysis evaluates the relative performance of the various treatment/disposition
22 alternatives under the permit criteria to identify potential advantages and disadvantages of each
23 alternative relative to the others. The tables present information from GE's RCMS for the five
24 alternatives included in that document; information for the new alternative was developed by
25 EPA using, where possible, GE's underlying cost assumptions.
26 11.1 OVERVIEW OF ALTERNATIVES
27 All five alternatives would involve some disposition of the sediment and floodplain soil in a
28 disposal facility, either directly or after treatment. The three alternatives involving disposal only
29 are: (1) disposal in off-site permitted landfills (TD 1); (2) disposition in an on-site combined
30 disposal facility (CDF) in a local waterbody, e.g., Woods Pond or one or more backwaters (TD
31 2); and (3) disposal in an on-site upland disposal facility, for which three potential locations have
32 been identified by GE (TD 3). The other two alternatives would involve treatment, either by a
33 chemical extraction process (TD 4) or by thermal desorption (TD 5). EPA also evaluated certain
34 variations to TD 1, specifying transport of excavated material by rail to the maximum extent
53
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
practical; this variation on alternative TD 1 when referred to specifically below, is termed TD 1
RR.
The results of a bench-scale test of a representative chemical extraction process (TD 4) indicate
that PCB concentrations in the treated sediment and soil would not be sufficiently low to allow
reuse on-site; therefore, the treated sediment and soil would have to be transported to a landfill
for disposal. For TD 5, it is assumed that the thermal desorption process would reduce the
concentrations of PCBs in the treated solid materials to levels (around 1 to 2 mg/kg) that could
13
allow reuse in the floodplain and that it would not increase the metals leachability of those
materials so as to preclude such use. However, due to uncertainties regarding the ultimate
effectiveness of the treatment process (as well as issues relating to the reuse of the treated soil),
TD 5 has also been evaluated based on the alternate assumption that all the treated material
would be transported to an off-site landfill for disposal.
All of the treatment/disposition alternatives except TD 2 were evaluated considering the same
range of sediment and soil volumes that could be removed under the sediment and floodplain
alternatives. This range extends from 191,000 cy, based on a combination of SED 3 and FP 2, to
2.9 million cy, based on a combination of SED 8 and FP 7. Under TD 2, however, the in-water
CDF(s) would be used only for the disposition of hydraulically dredged sediment from Reaches
5C and 6, which would be generated only under SED 6, SED 7, SED 8, or SED 9. Thus, TD 2
was evaluated for a range of hydraulically dredged sediment volumes from 300,000 cy for SED 6
to 1,240,000 cy for SED 8. For cost comparison purposes, however, the TD 2 analysis assumes
that the sediment and soil not placed in the CDF(s) would be transported off-site for disposal.
Under this assumption, the lower-bound costs for TD 2 are based on the combined volumes from
SED 6 and FP 2 and the upper-bound costs are based on the combined volumes from SED 8 and
FP 7.
All five alternatives were evaluated against the nine criteria discussed on page 6 of this section.
ARARs are discussed in Section 4. There is no comparison or evaluation of attainment of
IMPGs because this is not applicable to material treatment/disposition.
11.2 OVERALL PROTECTION OF HUMAN HEALTH AND THE ENVIRONMENT
As was the case for the SED and FP alternatives, the evaluation of whether the
treatment/disposition alternatives would provide overall human health and environmental
protection draws on the evaluations under several other permit criteria, notably long-term
effectiveness and permanence (including long-term adverse impacts), and short-term
effectiveness.
TD 1 (off-site disposal) would provide protection of human health and the environment by
providing for permanent disposal of the PCB-containing sediment and soil in permitted off-site
landfills. Relative to other alternatives, only relatively minor on-site short-term impacts would
occur under TD 1.
13 For reuse as backfill in the floodplain, only 50% of the volume is assumed to be the treated material because
following thermal treatment the material would be sterile, requiring amendments to be suitable for floodplain
restoration.
54
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
TD 1 RR (off-site disposal with rail transport) would provide protection of human health and the
environment equivalent to TD 1 with respect to PCB-containing sediment and soil, with some
additional protection afforded by the rail transport component, which would reduce the effects on
surrounding neighborhoods from truck traffic. There would be somewhat greater on-site short-
term impacts due to the need to construct a small railyard and loading facility at some point
along the rail right-of-way.
TD 2 (disposition in on-site CDF[s]) would provide protection of human health by permanently
isolating the hydraulically dredged sediment from Reaches 5C and 6 in covered in-water CDF(s),
which would be subject to monitoring and maintenance to verify their long-term integrity.
However, this alternative would not provide for disposition of any remaining sediment or the
excavated floodplain soil, which would need to be disposed of elsewhere. Moreover,
implementation of TD 2 would cause significant long-term environmental impacts because the
CDF(s) would result in a permanent loss of the aquatic habitat in a large portion of Woods Pond
and/or one or more of the backwaters where the CDF(s) would be constructed, and potentially
could be breached in the future should a catastrophic event occur. In adjacent areas, TD 2 would
result in a permanent "take" of several state-listed species, would alter the natural appearance of
the areas containing the CDF(s), and would result in a permanent loss of flood storage capacity
in those areas (assuming that sufficient compensatory flood storage could not be provided).
TD 3 (on-site upland disposal) would provide protection of human health and the environment
by permanently isolating the PCB-containing sediment and soil in an upland disposal facility,
which would be constructed with an appropriate double liner, cover, and double leachate
collection system. While this alternative would cause a change in existing habitat within the
operational footprint of the upland disposal facility, the capped landfill area would be replanted
with grass, and the support areas that are no longer needed after closure would be restored. The
significance of the long-term or permanent change in habitat would depend on the existing
habitat at the selected location and the size of the facility. This alternative would have additional
short-term impacts such as truck transport of landfill leachate over public roads to GE's
groundwater treatment plant (GWTP) located in Pittsfield, and the operation of the landfill for
the duration of the remedy.
TD 4 (chemical extraction) would provide protection of human health and the environment by
reducing the PCB concentrations in the sediment and soil, followed by off-site disposal of the
treated material. However, the long-term reliability and effectiveness of the chemical extraction
process have not been demonstrated for Housatonic River sediment at the bench scale and at full
scale for PCBs in sediment and soil representative of those from the Rest of River. Specifically,
the bench-scale study for this technology failed to demonstrate that site sediment and soil can be
effectively treated, in part due to a failure to achieve reasonable mass balance calculations as
well as acceptable residual concentrations.
TD 5 (thermal desorption) would provide human health protection by reducing the PCB
concentrations in the sediment and soil, followed by on-site reuse and/or off-site disposal of
those treated materials and off-site disposal/destruction of the liquids containing the condensed
PCBs. On-site reuse of a portion of the treated soil would be protective of human health because
the treated solids would be sufficiently characterized to ensure that they would not cause adverse
human health effects. However, if a portion of the treated soil is reused as backfill in the
floodplain, that reuse would potentially result in long-term adverse environmental impacts in the
55
8/1/2012
-------
1 forested floodplain and other wetland areas due to the differences in soil characteristics between
2 those materials and the existing natural soil in those wetland areas unless the treated soil is
3 properly amended. In addition, regardless of whether treated soil is reused in the floodplain, TD
4 5 would produce the greatest amount of GHG emissions of any of the alternatives.
5 11.3 CONTROL OF SOURCES OF RELEASES
6 All of the treatment/disposition alternatives would control the potential for PCB-containing
7 sediment and soil to be released and transported within the river or onto the floodplain, although
8 some alternatives would provide more effective control of such releases than others. TD 1 (or
9 TD 1RR) best meet this criterion followed by TD 3.
10 Under TD 1, placement of the removed PCB-containing sediment and soil in a permitted off-site
11 landfill or landfills would effectively isolate those materials from being released into the
12 environment.
13 TD 1 RR would control sources of releases in a manner identical to that described above for
14 TD 1.
15 TD 2 would minimize the potential for releases through placement of some of the removed
16 materials into CDF(s). Under TD 2, there is a potential for releases of sediment into the river
17 during the CDF construction process and through releases of PCBs through the berms. It is also
18 possible that releases from the CDF(s) could occur after CDF closure through migration to
19 groundwater or due to damage caused by ice or floods.
20 TD 3 would address future releases through the placement of the materials in an upland disposal
21 facility and the implementation of a long-term monitoring and maintenance program. Placement
22 of the PCB-containing sediment and soil into an upland disposal facility would most likely
23 effectively isolate the removed materials from being released into the environment. However, the
24 potential remains for releases to occur to the Housatonic River watershed both during operations
25 and in the long term if not properly operated and maintained.
26 Under TD 4 and TD 5, the potential for the PCB-containing sediment and soil to be released
27 within the river or onto the floodplain during treatment operations would be minimal. However,
28 the potential remains for releases to occur to the Housatonic River watershed both during
29 operations and in the long term if not properly operated and maintained. Under TD 4, the treated
30 solid materials would be transported to an off-site landfill for disposal, the wastewater would be
31 subject to treatment prior to discharge to the river, and the water treatment sludge would also be
32 transported to an off-site landfill for disposal. Under TD 5, to the extent that some of the treated
33 solids are used as backfill in the floodplain, chemical characterization sampling would be
34 performed to verify that those materials would not present concerns regarding future releases or
35 exposure. The remainder of the treated solids, or all such solids if none are reused as floodplain
36 backfill, would be transported to an off-site landfill for disposal, and the concentrated PCB-
37 containing liquid condensate from the thermal desorption process would be sent off-site for
38 incineration.
39 11.4 COMPLIANCE WITH FEDERAL AND STATE ARARs
40 The potential ARARs identified for the treatment/disposition alternatives are discussed in more
41 detail in Section 12 and Table 12-1 of the June 2011 Site Information Package, and the ARARs
56
8/1/2012
-------
1 for SED 9/FP 4 MOD are included in Attachment B-9. Each of the TD alternatives would
2 involve moving the sediment, bank soil, and floodplain soil from the point of excavation to the
3 treatment/disposition point, and each TD alternative would attain the ARARs associated with
4 such movement, except for TD 2, which would not meet wetland and floodplain requirements.
5 TD 1, with disposal occurring off-site at one or more permitted disposal sites, would have fewer
6 additional ARARs than the other treatment/disposition alternatives, and would attain the
7 requirements. TD 1 RR would have all the same ARARs as TD 1 plus additional ARARs that
8 apply to the rail switch yard and loading facility. TD 2, an in-water CDF, would have ARARs
9 associated with its location in the river, it being a solid waste disposal site, and being in an area
10 of potential habitat of state-listed species. TD 2 would not meet wetland and floodplain
11 requirements. TD 3, on-site landfilling, has ARARs associated with being a solid waste disposal
12 site, and possibly impacts on wetland areas. Use of one of the potential locations for the TD 3
13 upland disposal facility, which is in a state-designated Area of Critical Environmental Concern,
14 may also be subject to ARARs. TD 4 and TD 5 have ARARs related to the treatment of toxic
15 substances/hazardous waste, and depending on their location, would have wetland, floodplain,
16 and/or species habitat ARARs to attain.
17 11.5 LONG-TERM RELIABILITY AND EFFECTIVENESS
18 The assessment of long-term reliability and effectiveness for the treatment/disposition
19 alternatives included an evaluation of the magnitude of residual risk, the adequacy and reliability
20 of the alternatives, and potential long-term adverse impacts on human health or the environment.
21 11.5.1 Magnitude of Residual Risk
22 Placement of PCB-containing sediment/soil in off-site permitted landfills (TD 1 and TD 1 RR),
23 in one or more CDF(s) (TD 2), or in an upland disposal facility (TD 3) would permanently
24 isolate those materials from direct contact with human and ecological receptors. Under TD 2, as
25 noted above, there is a greater potential for releases and resulting risk than under TD 1 and TD 3.
26 Under TD 4 and TD 5, it is not expected that there would be any significant residual risks,
27 because: (1) all treatment operations would be performed within secured areas, and residual
28 PCBs associated with the operations would be removed following completion of the treatment
29 operations; (2) all treated materials would be subject to verification sampling and successfully
30 and unsuccessfully treated material would be transported off-site for disposal, except for any
31 such material reused on-site under TD 5; and (3) any such treated materials reused on-site under
32 TD 5 would be sampled to verify that the material to be reused would not pose a residual risk.
33 In summary, all of the treatment/disposition alternatives would minimize any future residual risk
34 from exposure to the PCB-containing materials, although there would be a greater potential for
35 such exposure under TD 2 than under the other alternatives, for the reasons noted above.
36 11.5.2 Adequacy and Reliability of Alternatives
37 There are considerable differences in the adequacy and reliability of the five
38 treatment/disposition alternatives. Based on these differences, the adequacy and reliability
39 criterion favors either TD 1, TD 1 RR, or TD 3 for disposal of the excavated materials under all
40 alternatives.
57
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Use of off-site disposal facilities (TD 1 and TD 1 RR) is a common and effective means for
permanent disposition of PCB-containing material. As the volume of materials requiring disposal
increases, multiple facilities may be required, but that is not expected to be a major
consideration.
In-water CDFs (TD 2) have been used to dispose of dredged PCB-contaminated sediment at
some sites. In this case, as discussed above, there is a somewhat greater potential for releases
from the CDF(s) than from off-site or local upland disposal facilities.
On-site disposal of PCB-containing materials in an upland facility (TD 3) has been used as part
of a final remedy at a number of sites and is an effective and reliable means for permanently
isolating such materials, provided the facility is properly constructed, monitored, and maintained.
However, the potential extended duration of the operation of such a facility for the range of
volumes of sediment and soil and length of remedy implementation could necessitate that the
facility operate for an extended period of time. In addition, GE proposes to truck the leachate
generated under TD 3 to its water treatment facility located in Pittsfield. This involves a one-
way trip of between 10 and 20 miles along public roads for the foreseeable future. The proposed
facility at Woods Pond could generate as much as 600,000 gallons of leachate per month (based
on its maximum size of 18 acres for 2,000,000 cy) for 10 to 20 years, requiring over 1,000 truck
trips per year (120 per month) while the facility is still receiving material. Based on SED 8/FP 7,
which has a volume of 2,900,000 cy, the amount of leachate could be as high as 1,000,000
gallons per month (based on the maximum landfill footprint at the Rising Pond site). This
volume could occur for up to 52 years and would require 200 truck trips per month or 2,400 per
year.
The use of chemical extraction (TD 4) has not been demonstrated at full scale on sediment and
soil representative of those in the Rest of River. The bench-scale results using site sediment and
soil were not promising. As a result, there are uncertainties about the long-term reliability and
effectiveness of operating such a system for a project of the size and duration, and with the range
of PCB concentrations, that would be involved at the Rest of River. The bench-scale results did
not demonstrate that this technology would be effective at treating the PCB-contaminated soil
and sediment at this site. These and other factors create uncertainties regarding the effectiveness
and reliability of using the chemical extraction process in a full-scale application.
Thermal desorption (TD 5) has been used at several sites to treat PCB-containing soil; however,
there is only limited precedent for use of this technology on sediment due in part to the time and
cost of removing moisture from the sediment prior to treatment. At the sites identified where
thermal desorption has been used, the volumes of materials that were treated were substantially
smaller and the duration of the treatment operations was substantially shorter than the volumes
and duration that could be required at the Rest of River. Further, when on-site reuse of treated
materials has occurred, the materials have typically been placed in a small area and covered with
clean backfill. For these reasons, the adequacy and reliability of this process for a long-term
treatment operation with a large volume of materials such as sediment/soil from the Rest of
River is uncertain.
11.5.3 Potential Long-Term Adverse Impacts on Human Health or the Environment
Implementation of TD 1, TD 1 RR, TD 2, and TD 3 would isolate the removed sediment/soil
from potential human and ecological exposure because the material would be contained in
58
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
structures designed specifically for that purpose. Under TD 4, removed material would first be
treated, and then disposed of off-site. For TD 5, materials would be treated, and then a portion
might be reused in the floodplain, assuming that it has acceptable residual levels of
contaminants, with the remainder disposed of off-site. Thus, under all the treatment/disposition
alternatives, no long-term adverse impacts on humans or ecological receptors from exposure to
the PCB-containing materials are expected, with the potential exception of TD 2 if a release were
to occur (e.g., during an extreme storm event).
TD 1 would not cause any adverse long-term environmental impacts in the Rest of River area
because it would involve off-site transport and disposal of the PCB-containing materials.
TD 1 RR would also not result in adverse long-term environmental impacts in the Rest of River
area. The railyard and loading facility would be demobilized following completion of the
remedy and the area restored to its former condition.
For TD 2, the placement of an in-water CDF in Woods Pond and/or one of the two identified
backwaters would have the most significant long-term adverse environmental impacts, including
a permanent loss of the aquatic habitat in those areas. Depending on the location and size of the
CDF(s), TD 2 would adversely affect the priority habitat of up to nine state-listed species. In
addition, the CDF(s) would raise the topography of the CDF area(s), reduce available
shoreline/wetland habitat, and produce a loss of the existing flood storage capacity.
For TD 3, the construction of the upland disposal facility, which, for the Woods Pond site, is
located within an Area of Critical Environmental Concern, would result in the alteration of
existing habitat within the operational footprint of that facility. In the landfill area itself, as well
as any support areas (e.g., access roads) that would remain after closure, the habitat alteration
would be permanent, although the landfill would be capped and planted. The significance of the
change in habitat would depend on the existing habitat at the location of the facility, as well as
the size of the facility.
Under TD 4 and TD 5, the construction and operation of a 5-acre treatment facility at the former
DeVos property would result in some loss of the relatively low quality habitat within that area (a
former agricultural area that is now open grassland with scattered shrubs) during the period of
treatment operations and for a few years thereafter. That loss, as well as increased noise and
human presence in the area, would affect the wildlife in the area (which includes the priority
habitat for certain state-listed species) during that period. However, given the relatively small
size of the facility, the previously altered nature of the habitat, and the planned reseeding of the
area with a grassland mix following removal of the facility, long-term ecological impacts
associated with construction and operation of the facility, if any, would be minimal.
Thus, of the treatment/disposition alternatives, TD 2, and to a lesser extent TD 3 (depending on
the actual landfill location selected), would have the greatest long-term adverse environmental
impacts. TD 4 and TD 5 would have similar environmental impacts, but less than TD 3 because
they would be in place for the duration of the remedy. TD 1 would have the least long-term
impacts.
59
8/1/2012
-------
1 11.5.4 Summary of Long-Term Reliability and Effectiveness
2 In summary, TD 1 RR is considered the most reliable and effective of all the TD alternatives,
3 followed by TD 1, TD 3, and TD 5. TD 1 RR is considered more effective and reliable than TD
4 1 because the rail transportation component (vs. over the road truck transport) reduces potential
5 impacts on the local community. This reduction would more than offset the slight additional
6 short-term impact due to the need to construct a small railyard and loading facility. TD 3 is
7 evaluated similarly to TD 1 RR and TD 1; however, it would have the additional impacts
8 resulting from the long-term transport of leachate and the permanent alteration of habitat. TD 5
9 (thermal desorption) is a reliable and effective technology; however, the difficulties associated
10 with treating wet materials make its reliability somewhat questionable. TD 2 is viewed as less
11 favorable due to the potential for releases and the permanent loss of habitat. TD 4 is considered
12 the least reliable and effective due to the lack of proof that the technology can effectively treat
13 PCBs in sediment.
14 11.6 ATTAINMENT OF IMPGs
15 Attainment of IMPGs is not applicable to treatment and disposition alternatives.
16 11.7 REDUCTION OF TOXICITY, MOBILITY, OR VOLUME
17 The degree to which the treatment/disposition alternatives would reduce the TMV of PCBs is
18 discussed below.
19 Reduction of Toxicity. TD 1 through TD 3 (including TD 1 RR) would not include any treatment
20 processes that would reduce the toxicity of, or directly affect, PCB concentrations in the removed
21 sediment and soil. TD 4 and TD 5 would incorporate treatment processes that can, to varying
22 degrees, reduce concentrations of PCBs. Under TD 4, the chemical treatment process would
23 reduce the toxicity of the sediment and soil by permanently removing some PCBs from these
24 materials, although the effectiveness of this technology is questionable. Under TD 5, the indirect-
25 fired thermal desorption system would reduce the toxicity of the PCB-containing sediment and
26 soil by permanently removing PCBs from these materials, and the PCBs in the liquid stream
27 would be sent to a permitted off-site disposal facility for destruction.
28 Reduction of Mobility: All of the alternatives would reduce the mobility of PCBs in the sediment
29 and soil. In TD 1, TD 1 RR, TD 2, and TD 3, these materials would be removed and disposed of
30 in off-site permitted landfill(s) (TD 1 and TD 1 RR) or permanently contained within on-site
31 CDF(s) (TD 2) or an upland disposal facility (TD 3). TD 4 and TD 5 would reduce the mobility
32 of PCBs present in the sediment/soil via chemical extraction or thermal desorption.
33 Reduction of Volume: TD 1, TD 1 RR, TD 2, and TD 3 would not reduce the volume of PCB-
34 containing material. For TD 4, treatment of sediment/soil would reduce the volume of PCBs
35 present in those materials by transferring some of the PCBs to an aqueous waste stream for
36 subsequent treatment. PCB-containing sludge would be generated from the wastewater treatment
37 system and would be sent to a permitted off-site facility for disposal. For TD 5, treatment of
38 sediment/soil in the thermal desorption system would reduce the volume of PCBs present in
39 those materials, with the liquid condensate transported to an off-site facility for destruction.
60
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
11.8 SHORT-TERM EFFECTIVENESS
Evaluation of the short-term effectiveness of the treatment/disposition alternatives includes
consideration of the short-term impacts of implementing these alternatives on the environment
(considering both ecological effects and increases in GHG emissions), on the local communities
(as well as communities along truck transportation corridors), and on the workers involved in the
treatment and disposition activities.
11.8.1 Impacts on the Environment
All the treatment/disposition alternatives would produce some short-term adverse impacts on the
environment, but to varying degrees depending on the duration and scope of the alternative. TD
1 would have the least impacts of all the TD alternatives, requiring only access roads and staging
areas for loading of vehicles for off-site transport. TD 1 RR would require the construction of a
railyard and loading facility at some point along the existing rail right-of-way and would require
approximately the same amount of access roads and staging areas as TD 1. The short-term
impacts of TD 2 through TD 5 would include loss of habitat and loss or displacement of aquatic
biota and other wildlife in the areas where the disposition or treatment facilities are located, as
well as in adjacent areas, during construction and operations. TD 2 would affect a portion of
Woods Pond and/or one of the two backwaters identified for a CDF, as well as the adjacent
floodplain. Specific short-term impacts associated with TD 3 would depend on the habitat at the
selected location and the operational footprint of the facility. Construction of a treatment facility
for TD 4 or TD 5 on the former DeVos property would result in the temporary reduction of open
field habitat on that property.
All of the treatment/disposition alternatives could also have short-term effects on the
environment due to the potential for accidental releases of PCB-containing materials. In
particular, TD 3 has the risk of the release of leachate during its transport from the upland
disposal facility(s) to the GE GWTP in Pittsfield. In addition, TD 4 and TD 5 have the potential
for failure of process and control equipment during operations, which could result in a release of
PCB-containing materials. The potential for these types of effects would increase with the
length of the implementation period.
11.8.2 Carbon Footprint - GHG Emissions
GHG emission estimates were developed based on the ranges of the potential volumes of
sediment and soil that would require disposal or treatment. Table 22 summarizes the resulting
ranges of total GHG emissions associated with each TD alternative. To provide context
regarding the emissions reported, the number of passenger vehicles that would emit an
equivalent quantity of C02-eq in 1 year is also presented in the table.
As shown in Table 22 for the TD alternatives evaluated in the RCMS (excluding TD 2, which is
not comparable, and TD 1 RR for which estimates were not available), TD 5 would have the
greatest amount of total GHG emissions for the range of volumes; TD 4 would have the next
largest amount; followed by TD 1. TD 3 would have lowest amount of total GHG emissions for
the range of volumes, approximately 3 to 5 times less than the next lowest alternative (TD 1).
TD 1 RR would have significantly lower GHG emissions than TD 1 because the emissions due
to off-site truck transport would be replaced by the much lower emissions resulting from off-site
transport via rail. It should be noted, however, that the magnitude of the differences among
61
8/1/2012
-------
1 alternatives varies with the removal volume. For example, the lower-bound estimates for TD 1
2 and TD 3 are 19,000 and 5,500 tonnes, respectively, a difference of 13,500 tonnes. However, the
3 upper-bound estimates are 290,000 tonnes for TD 1 and 61,000 tonnes for TD 3, a difference of
4 229,000 tonnes (17 times more than the difference at the lower bound). The differences in GHG
5 emissions between TD 1 and TD 3 are due to the distance that materials need to be trucked
6 before ultimate disposition. Such differences are even more pronounced when comparing TD 3
7 with TD 4 and TD 5.
8 Table 22
9 Calculated GHG Emissions Anticipated to Result from
10 Treatment/Disposition Alternatives
Allcriiiili\c
loliil (.ll(. Kmissioiis
(lOIIIIOS)
No. Vehicles willi
r.(|iii\iilenl Kmissioiis
TD 1
19,000-290,000
3,600 - 55,400
TD 2
See Note 1
See Note 1
TD 3 (see Note 2)
5,500-61,000
1,100- 11,700
TD 4
27,000 - 370,000
5,200 - 70,700
TD 5 (with reuse)
66,000 - 1,000,000
12,600- 191,200
TD 5 (without reuse)
66,000- 1,100,000
12,600-210,300
11 Notes:
12 1. Emissions estimated for TD 2 range from 2,700 to 8,800 tonnes and do not include the emissions that would be necessary for
13 off-site transport and disposal of materials that are not placed in the CDF(s). As such, these estimates are not comparable to
14 the emissions listed for the other alternatives.
15 2. The lower bound of this range for TD 3 is based on disposal of the minimum potential removal volume at the Woods Pond site
16 (which would have the lowest GHG emissions of the identified sites) and the upper bound is based on disposal of the
17 maximum potential removal volume at the Rising Pond site, which is the only one of the identified local disposal sites that
18 could accommodate that maximum volume.
19 11.8.3 Impacts on Local Communities
20 All the alternatives would also result in short-term impacts to the local communities in the Rest
21 of River area. These impacts would include disruption, noise, and other impacts resulting from
22 the increased truck traffic and from the construction and operation of the on-site disposition or
23 treatment facilities. TD 1 RR, due to its use of rail transport, would result in a significant
24 decrease in impacts to local communities due to reduced off-site truck traffic. In addition,
25 unique to TD 3, leachate being transported via truck from the upland disposal facility(s) could be
26 released en route due to malfunctioning equipment or an accident, creating impacts to the local
27 communities, and the operation of the landfill for the duration of the remedy.
62
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
The estimated numbers of off-site truck trips for each alternative, based on the estimated range of
volumes that could be involved, are shown in Table 23.14
Table 23
Estimated Off-Site Truck Trips for Treatment/Disposition Alternatives
\lliTii;ili\i'
Olf-She Truck Trips for
Lowit-KoiiihI Volume
OIT-Siii' Truck Trips for
I ppcr-liouml Volume
TD 1
15,900 (2,000)
243,000 (6,100)
TD 2
See Note 3
See Note 3
TD 3 (see Note 4)
1,450 (180)
68,000 (3,600)
TD 4
15,900 (2,000)
243,000 (6,100)
TD 5 (with reuse)
13,300 (1,700)
190,500 (4,800)
TD 5 (without reuse)
14,300 (1,800)
218,900 (5,500)
TD 1 RR
0 (0) Note 7
0(0)
Notes
1. Truck trips estimated assuming 16-ton capacity trucks for importing material and equipment to the site, 20-ton capacity
trucks for transporting excavated materials, and 20% bulking factor in the trucks.
2. The number in parentheses represents average annual truck trips.
3. Truck trips estimated for TD 2 range from 5,600 to 19,500 and do not include the truck trips that would be necessary for off-
site transport and disposal of materials that are not placed in the CDF(s). As such, these estimates are not comparable to the
numbers of truck trips listed for the other alternatives.
4. The lower bound of this range for TD 3 is based on construction of an upland disposal facility at the Woods Pond site and
the upper bound is based on construction of such a facility at the Forest Street site.
5. A 10% volume reduction of sediment/soil after treatment has been assumed for thermal desorption treatment (TD 5).
6. For TD 5 with reuse, it is assumed that approximately 50% of the floodplain soil treated by thermal desorption would be
reused on-site and that all remaining materials would be transported off-site for disposal.
7. It was assumed for the purpose of this analysis that there would be zero off-site truck trips; however, there may be a
necessity to utilize trucks instead of rail under certain conditions.
As shown in this table, excluding TD 2, which is not comparable, TD 3 would involve the fewest
off-site truck trips for the range of volumes, while those for the other alternatives are roughly
comparable, with somewhat more for TD 1 and TD 4 than for TD 5. TD 1 RR will maximize the
transport of the contaminated soil via rail; therefore off-site truck traffic will be minimized.
Again, however, the magnitude of the differences among alternatives varies with the removal
volume. The additional truck traffic would also increase the risk of traffic accidents along
transport routes. An analysis of potential risks from the increased off-site truck traffic that would
be associated with the treatment/disposition alternatives in terms of potential fatalities and non-
fatal injuries is presented in Table 24.
14 For comparability among alternatives, this table shows only off-site truck trips, i.e., those for importation of
construction materials and equipment to the site over public roads for construction and closure of a local disposal
or treatment facility, as well as those for transport of excavated or treated soil/sediment to off-site disposal
facilities. It does not include transport of excavated materials from the staging areas to the local disposal or
treatment facility.
63
8/1/2012
-------
1 The incidence of potential injuries and fatalities resulting from accidents associated with
2 increased off-site truck traffic would be the greatest for TD 1 and TD 4, followed closely by TD
3 5, and would be far lower for TD 3. As with the number of off-site truck trips, the differences in
4 estimated injuries and fatalities resulting from such traffic become more pronounced as the
5 removal volumes increase. Because TD 1 RR would require no off-site truck traffic, no injuries
6 or fatalities are associated with this alternative because it was assumed for the purpose of this
7 analysis that there would be zero off-site truck trips; however, it may be necessary to use trucks
8 instead of rail under certain conditions.
9 Table 24
10
11 Incidence of Accident-Related Injuries/Fatalities
12 Due to Increased Off-Site Truck Traffic
llll|)iicls
II) 1
I I) 2
I I) 3 *
11)4
I I) 5
(willi
Reuse)
I I) 5
(without
Reuse)
II) 1 RR
Non-Fatal Injuries
Number
4.34-
67.03
See Note
2
0.03 - 1.60
4.11-
62.87
3.44-
49.24
3.70-56.59
Note 4
Average
Annual
Number
0.45-
1.28
See Note
2
0.0002 -
0.084
0.51-1.57
0.43 - 1.23
0.46-1.41
0
Probability1
99 - 100%
See Note
2
3 - 80%
98 - 100%
97 - 100%
98 - 100%
-
Fatalities
Number
0.20-
3.14
See Note
2
0.002 -
0.07
0.19-2.94
0.16-2.31
0.17-2.65
0
Average
Annual
Number
0.02-
0.06
See Note
2
0.0002 -
0.004
0.02 - 0.07
0.02 - 0.06
0.02-0.07
0
Probability1
18 - 96%
See Note
2
0.2 - 7%
18 - 95%
15 - 90%
16 - 93%
-
13 Notes:
14 1. Probability indicates the probability of at least one injury/fatality.
15 2. The estimated risks of accidents for TD 2 are based only on the truck trips necessary to transport materials to the site for the
16 construction of the CDF(s) and do not consider the truck trips for off-site transport of the materials that would not be placed
17 in the CDF(s). As such, those risks are not comparable to the estimated risks for the other treatment/disposition alternatives
18 (which consider all removed materials). Under the scenario evaluated, the risks estimated for TD 2 are 0.01 to 0.02 fatalities
19 (with a 1% to 2% probability of at least one fatality) and 0.13 to 0.46 non-fatal injuries (with a 12% to 37% probability of at
20 least one injury).
21 3. The lower bound of this range for TD 3 is based on construction of an upland disposal facility at the Woods Pond site and
22 the upper bound is based on construction of such a facility at the Forest Street site.
23 4. It was assumed for the purpose of this analysis that there would be zero off-site truck trips, however there may be a
24 necessity to utilize trucks instead of rail under certain conditions
64
8/1/2012
-------
1 11.8.4 Potential Measures to Avoid, Minimize, or Mitigate Short-Term
2 Environmental and Community Impacts
3 A number of measures would be employed in an effort to avoid, minimize, or mitigate the short-term
4 impacts of the treatment/disposition alternatives on the environment and the affected communities.
5 As would be expected, the level of impact and thus, the scope and duration of mitigation measures
6 are related to the scale/scope of the alternative and the duration of implementing the alternative. For
7 TD 1, the mitigation measures would relate to the increased truck traffic, while for the other TD
8 alternatives, mitigation measures would address the increase in truck traffic as well as the impacts
9 associated with construction and operation of the different facilities.
10 11.8.5 Risks to Remediation Workers
11 There would also be health and safety risks to site workers implementing each of these
12 alternatives. For TD 1 and TD 1 RR, these risks would consist of risks to the truck drivers and,
13 in the case of TD 1 RR, railroad employees, and to the employees of the off-site disposal
14 facilities, rather than to on-site remediation workers, and thus, were not quantified. For TD 2
15 through TD 5, an analysis of estimated risks to site workers is summarized in Table 25.
16 Estimated risks to site workers for the range of volumes would be lowest for TD 2 (due to its
17 fewer years of operation) and higher for the other alternatives, with TD 3 slightly higher than TD
18 4 and TD 5. In this case, there are no substantial differences among TD 3, TD 4, and TD 5 at the
19 same volumes, but there are significant differences between the lower and upper bounds.
20 Table 25
21
22 Incidence of Potential Accidents/Injuries Due to
23 Implementation of Alternatives TD 2 through TD 5
llll|)iicls
I I) 2
id r
11)4
I I) 5
Labor-hours (hours)
73,000-259,000
306,000 -
1,836,000
160,600 -
1,673,600
160,600 - 1,673,600
Years of Operation
6-20
8-40
8-40
8-40
Non-Fatal Injuries
Number
0.70-2.50
2.69-16.4
1.27-13.1
1.27-13.1
Average Annual Number
0.12-0.13
0.34-0.41
0.16-0.33
0.16-0.33
Probability13
50 - 92%
93 - 100%
72 - 100%
72 - 100%
Fatalities
Number
0.01-0.03
0.02-0.11
0.007-0.08
0.007-0.08
Average Annual Number
0.0012-0.0013
0.002-0.003
0.0009 - 0.002
0.0009-0.002
Probability13
1 - 3%
2-11%
0s-
00
1
r-
©
0s-
00
1
r-
©
24 a The lower bound of this range for TD 3 is based on disposal of the minimum potential removal volume at the Woods Pond
25 site, and the upper bound is based on disposal of the maximum potential removal volume at the Rising Pond site, which is the
26 only one of the identified local disposal sites that could accommodate that maximum volume and thus, has the longest period
27 of operations.
28 b Probability indicates the probability of at least one injury/fatality.
65
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
11.8.6 Summary of Short-Term Effectiveness
All of the treatment/disposition alternatives would have some short-term negative impacts on the
environment, local communities, and communities along transport routes. TD 2 through TD 5
would cause a loss of habitat and loss or displacement of wildlife in the area where the disposal
or treatment facility is located, as well as in adjacent areas, during construction and operation of
the facility. In addition, all alternatives would involve the potential for accidental releases of
various PCB-containing materials during transportation to off-site or local disposal or treatment
facilities. This potential would increase with TD 2, TD 3, TD 4, and TD 5 because those
alternatives would pose additional risks associated with the potential for failure of process and
control equipment during operations, and releases of process byproducts/chemicals/leachate to
the environment. Although all alternatives would generate GHG emissions, for the range of
volumes (excluding TD 2, which is not comparable), TD 5 would produce the most such
emissions and TD 3 would produce the least.
Estimates of off-site truck trips and traffic accident risks from that truck traffic indicate that, for
the range of volumes (excluding TD 2), TD 1 and TD 4 would involve the most off-site truck
trips and cause the most injuries related to such transport, followed closely by TD 5, with far
fewer off-site truck trips and transport-related injuries for TD 1 RR and TD 3. In terms of risks
to on-site workers, excluding TD 1 (which would not affect site workers) and TD 2 (which is not
comparable), the estimated injuries for the other three TD alternatives are roughly comparable
for the same volumes.
11.9 IMPLEMENTABILITY
11.9.1 Technical Implementability
All of the treatment/disposition alternatives are considered technically implementable, subject to
certain qualifications:
¦ For TD 1, there are uncertainties regarding the future availability of the necessary
capacity in off-site landfills for the alternatives that have the larger volumes and longer
durations.
¦ TD 1 RR has the same uncertainties regarding capacity in off-site landfills, plus some
additional uncertainty related to the timing and availability of rail transport capacity.
¦ For TD 2, it would likely not be feasible to obtain sufficient flood storage compensation
at the appropriate elevations/areas to provide for construction of a CDF(s) large enough
to hold the necessary sediment disposal volumes.
¦ For TD 3, construction and use of an upland disposal facility would be technically
implementable. Three potential locations for such a facility, with varying maximum
capacities (ranging from 1.0 to 2.9 million cy) have been identified.
¦ TD 4 and TD 5 would be implementable provided that vendors are available to operate
the treatment process. The former DeVos property could be used as a potential area to
locate a treatment facility. However, there are several uncertainties regarding full-scale
66
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
application of both chemical and thermal processes to sediment (e.g., moisture content),
particularly with some of the volumes associated with the sediment alternatives.
11.9.2 Administrative Implementability
Administrative implementability has been evaluated considering regulatory requirements, the
need for access agreements, and coordination with governmental agencies.
TD 1 and TD 1 RR would be conducted in accordance with the requirements of applicable
federal, state, and local regulations relating to the off-site transport and disposal. The four other
alternatives would be "on-site" activities for the purposes of the permit exemption set forth in
Section 121(e) of CERCLA and Paragraph 9.a of the Consent Decree. As such, no federal, state,
or local permits or approvals would be required. However, implementation of these alternatives
would need to comply with the substantive requirements of applicable and appropriate
regulations (i.e., ARARs) (unless waived).
Implementation of TD 1 would not require access agreements. Implementation of TD 2 and TD
3 would require permanent access to the location(s) selected for the disposal facility(ies).
Implementation of TD 4 and TD 5 would require access to the location selected for the treatment
facility; GE is the current owner of the potential location identified for TD 4 and TD 5, as well as
one potential location for TD 3. GE has the right to acquire the other two sites identified as
potential locations for TD 3. Therefore, assuming use of one or more of these locations, no site
access agreements would be required for implementation of TD 3 through TD 5, but such
agreements may be required for TD 2. TD 1 RR would require an access agreement for the rail
siding and loading facility, which would be assumed to be temporary. Finally, all alternatives
would require coordination with EPA, as well as state and local agencies. TD 2 and TD 3 would
require extensive coordination with local government and the public. It could require an
extended period of time for these options, and based on past public input received, these options
could encounter substantial local and state opposition, likely rendering these alternatives difficult
to implement. TD 4 and TD 5 would require similar coordination; however, the level of
coordination would likely be less than that for TD 2 and TD 3.
In conclusion, there is a clear distinction among the alternatives with respect to this criterion: TD
1 would be easiest to implement, followed closely by TD 1 RR, with TD 2 and TD 3 being the
most difficult and time consuming to implement from an administrative perspective, while TD 4
and TD 5 would experience similar difficulties from a technical perspective.
11.10 COST
The estimated cost ranges for each treatment/disposition alternative, including total capital cost,
estimated annual OMM cost, and total estimated present worth are summarized in Table 26 and
are taken from GE's RCMS, except for TD 1 RR, which is summarized in Attachment B-10.
Note that, in this case, the costs presented for TD 2 include not only the costs for disposition in
the CDF(s) of the hydraulically dredged sediment from Reaches 5C and 6 under SED 6 through
SED 9, but also the estimated costs for off-site transport and disposal of the remaining sediment
removed under those alternatives, as well as excavated floodplain soil (lower-bound costs
consider SED 6 and FP 2 and upper-bound costs consider SED 8 and FP 7). In addition, for TD
3, the range of costs presented are for an upland disposal facility constructed at the Rising Pond
site because that is the only single location with the capability to hold the maximum potential
67
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
volume of 2.9 million cy. As described above, two smaller landfills at different locations could
be constructed and used if necessary to handle that maximum removal volume, but specific costs
for this approach have not been estimated.
As shown in Table 26, TD 3 is the least costly alternative. At the low end of the volume range, it
would cost about 2 to 4 times less than the other alternatives; and at the high end of the range, it
would cost about 2 to 10 times less.
11.11 OVERALL CONCLUSION FOR TREATMENT/DISPOSITION ALTERNATIVES
For the reasons discussed above, it is concluded that of all the treatment/disposition alternatives,
TD 1 RR is best suited to meet the General Standards in consideration of the Selection Decision
Factors. This conclusion has been reached because it would permanently isolate the PCB-
containing sediment and soil from human and ecological receptors, would reduce impacts of
truck traffic on local communities, would have a high degree of reliability, would not cause
widespread long-term adverse environmental impacts in the Rest of River, would comply with
ARARs, and would be readily implementable from an administrative and technical feasibility
perspective. Although costs would be higher for TD 1 RR than for TD 3, the benefits of being
readily implementable, compliance with ARARs, and having a lower impact on the local
community outweigh the cost advantages of TD 3 and other advantages such as lower GHG
emissions and lower traffic accident risks.
68
8/1/2012
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Table 26
Cost Summary for Treatment/Disposition Alternatives
II) 1
I I) 2
I I) 3
11)4
I I) 5
(with reuse)
I I) 5
(without reuse)
II) 1 RR
Total Capital Costs
0
$6.5 - 20 M
$9.5-67 M
$17 - 20 M
$20 - 232 M
$20 - 232 M
$0.5- 1.5 M
Total Operations
Cost
0
$6.8-25 M
$11 - 110 M
$32-365 M
$47 - 698 M
$47 - 698 M
0
Total Off-Site
Disposal Costs
$55 -832 M
$75 - 445 M
0
$40-614 M
$36- 518 M
$39 -595 M
$52 - 786 M
Total Monitoring
and Maintenance
Costs
0
$12 - 20 M
$15 - 24 M
$0,125 M
$0,125 M
$0,125 M
0
Total Cost for
Alternative
$55 - 832 M
$100- 510 M
$36 - 201M
$89 - 999 M
$103 -1,450 M
$106 -1,530 M
$53 - 788 M
Total Present
Worth
$40 - 220 M
$46-131 M
$17 - 49 M
$70 - 286 M
$81 -569 M
$83 - 590 M
$38 -210M
Notes:
1. All costs are in 2010 dollars. $ M = million dollars.
2. With the exception of TD 2, the ranges of costs presented are the minimum and maximum anticipated costs based on the potential range of volumes that would be potentially
removed under the sediment and floodplain soil alternatives (191,000 cy to 2.9 million cy). For TD 2, the lower-bound costs are based on the combined volume of SED 6 and
FP 2 and the upper-bound costs are based on the combined volume of SED 8 and FP 7, with material not placed in the CDF(s) assumed to be transported off-site for non-
TSCA disposal. Thus, the upper-bound costs, but not the lower-bound costs, for TD 2 are comparable to the costs for the other alternatives.
3. Total capital costs are for engineering, labor, equipment, and materials associated with implementation.
4. Total operations costs consist of the total of the average annual costs for operation, placement, and/or treatment of sediment and/or soil, estimated for the range of durations
for implementing the alternatives.
5. Total monitoring and maintenance costs are for performance of post-closure monitoring and maintenance programs of 100 years for TD 2 and TD 3 and 5 years for TD 4 and
TD 5.
6. Total present worth cost is based on using a discount factor of 7%, considering the range of total potential durations for the alternative, and post-closure monitoring and
maintenance periods of 100 years for TD 2 and TD 3 and 5 years for TD 4 and TD 5.
7. For TD 5 with reuse, it is assumed that approximately 50% of the floodplain soil treated by thermal desorption would be reused on-site and that all remaining materials would
be transported off-site for disposal.
8. Costs for TD 3 do not include the very likely extensive costs associated with the approval process required for an on-site landfill.
9. Costs for TD 1 are based on transport via truck; costs for TD 1 RR are for transport via rail.
69
8/1/2012
-------
ATTACHMENT B-1 USE OF CHANNEL REALIGNMENT ALONG THE
HOUSATONIC RIVER FOR RESTORATION AND REMEDIATION OF
PCB CONTAMINATION
-------
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
1 USE OF CHANNEL REALIGNMENT ALONG THE HOUSATONIC RIVER FOR
2 RESTORATION AND REMEDIATION OF PCB CONTAMINATION
3 TABLE OF CONTENTS
4
5 1. INTRODUCTION 1
6 2. OVERVIEW OF GEOMORPHIC CONDITIONS ALONG THE HOUSATONIC RIVER 1
7 3. GEOMORPHIC LIMITATIONS OF BANK STABILIZATION AND RESTORATION
8 APPROACHES 3
9 4. THEORY OF CHANNEL REALIGNMENT TO PROVIDE GEOMORPHIC RESTORATION ... 4
10 5. POTENTIAL APPLICATION OF REALIGNMENT ON THE HOUSATONIC RIVER 6
11 5.1 BENEFITS OF CHANNEL REALIGNMENT FOR THE HOUSATONIC RIVER 6
12 5.1.1 Creation of New Oxbows and Wetlands 6
13 5.1.2 Floodplain Fill and Capping of the Existing Channel 7
14 5.1.3 Construction in the Dry Off-Line 8
15 5.1.4 Sediment and Erosion Control During Construction 8
16 5.2 DISADVANTAGES OF CHANNEL REALIGNMENT FOR THE HOUSATONIC RIVER... 8
17 5.2.1 Vegetation Disturbance and Impact 8
18 5.2.2 Change in the Character of the Housatonic 8
19 5.2.3 Additional Testing and Sampling During Construction 8
20 5.2.4 Change in Property Lines due to Channel Realignment 8
21 6. SUMMARY OF POTENTIAL CHANNEL REALIGNMENT ON THE HOUSATONIC RIVER.... 9
22 7. LITERATURE CITED 9
23 LIST OF FIGURES
24 Figure 1: Housatonic River Reach 5B Meandering Cutoffs and Oxbows 2
25 Figure 2: Reach 5B Comparison of 1886 Plan Form and Current Plan Form 3
26 Figure 3: Housatonic River Reach 5B-Unstable Plan Form with Tight Radii of Curvature
27 and a High Sinuosity 4
28 Figure 4: Example of Potential Oxbows and Wetlands in Reach 5A 7
29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
1. INTRODUCTION
This paper provides an overview of the use of channel realignment for geomorphic restoration of
disturbed river systems, and how that information may relate to the remediation of PCB
contamination in the Housatonic River. There are four main goals of this effort:
1. Document the current understanding of channel realignment as a stream restoration tool.
2. Document the limitations of bank stabilization alone as a restoration alternative following
remediation.
3. Describe the additional restoration opportunities provided by channel realignment.
4. Present the advantages and disadvantages of channel realignment as a tool for
geomorphic and ecological restoration.
The information in this paper integrates sciences such as fluvial geomorphology, engineering,
toxicology, hydraulics, sedimentology, biology, and ecology, as well as applied construction and
remediation technology. The ultimate goal for this paper is to build an understanding of
restoration processes using principles of geomorphology to help the assessment of potential
strategies for the remediation and ecological restoration of the Housatonic River.
2. OVERVIEW OF GEOMORPHIC CONDITIONS ALONG THE
HOUSATONIC RIVER
Over the past few hundred years, the Housatonic River ecosystem has undergone a long history
of channel disturbances and channel relocations, and in some cases has adapted to these channel
and watershed disturbances through changes to plan form and dimension. Evidence of past plan
form adjustments along the Housatonic River is displayed in Figure 1, where a chute cutoff
formed along the River, and created an oxbow wetland. As a result of this chute cutoff, the
sinuosity (ratio of channel length to the valley length) of this particular reach decreased from a
value of 2.6 to 1.7. Often chute cut-offs and other plan form channel adjustments occur at
locations where the ratio of the radius of curvature of the channel relative to the bankfull width
of the channel is less than 2. There are many locations on the Housatonic River where tight
bends exhibit low radius of curvature ratios (Rc), as exhibited in Figure 1.
Cutoffs are an inherent part of meandering behavior (Hooke 2004) and help streams maintain a
stable state by preventing the length and sinuosity of the channel from becoming too great and
developing an unstable plan form configuration (Camporeale et al. 2008). Empirical studies of
meander geometry show the radius of curvature of meander bends trends toward 2.4 times the
bankfull channel width (Garcia 2008), implying there is an equilibrium dimension to which
meandering rivers evolve (Lagasse 2004).
1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
Relic Channel
Figure 1: Housatonic River Reach 5B Meandering Cutoffs and Oxbows
Additional evidence of an accelerated plan form evolution is seen on the Housatonic River
upstream and downstream of New Lenox Road (Figure 2). An analysis of historical topographic
maps and aerial photographs reveals that the majority of the Housatonic River from the
confluence of the East and West Branches downstream to Woods Pond was artificially
straightened in the past. Most of this straightening likely occurred in association with railroad
construction and agricultural practices in the Housatonic River valley completed in the 1850s,
and was often accompanied by removal of the woody vegetation along the stream banks. Even
before the railroad was built, the majority of the floodplain had been deforested during
colonization in the 17d' and 18"' centuries. The river, which appears to be a naturally sinuous
system, has been undergoing a process of active readjustment to these historical channel-altering
disturbances.
2
-------
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
2 Figure 2: Reach 5B Comparison of 1886 Plan Form and Current Plan Form
3 3. GEOMORPHIC LIMITATIONS OF BANK STABILIZATION AND
4 RESTORATION APPROACHES
5 As described above, the Housatonic River corridor has been highly disturbed, which has caused
6 active channel adjustments resulting in very sinuous and tight meander bends. Figure 3 displays
7 a section of Reach 5A that has a sinuosity of 2.0 and radii of curvature as tight as 1.4 times the
8 width of the channel. In stream reaches such as this, the plan form has areas of greatly
9 accelerated bank erosion that will lead to future chute cut-offs, oxbows and other channel
10 avulsions. Such geomorphic changes and erosion would likely result in future releases of PCB
11 contaminated soils that are not removed during remediation.
3
-------
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
1
2 Figure 3: Housatonic River Reach 5B - Unstable Plan Form with Tight Radii of
3 Curvature and a High Sinuosity
4 Understanding fluvial processes is imperative for successful river management and channel
5 restoration design (Holmes, 1993). Without an understanding of fluvial geomorphology, and the
6 impacts of catchment developments, it is likely that inappropriate enhancements will be
7 proposed and executed only to be destroyed by natural river processes. Specifically, the
8 application of bank stabilization techniques without this geomorphic context is unlikely to be
9 successful. In areas of tight meander bends, chute cutoffs will be able to circumvent bank
10 treatments.
11 If the existing channel is in an unstable morphology, then there will be areas of increased near-
12 bank shear forces (the force applied to the bank by flowing water) due to the tight radii of
13 curvature of the current plan form. Realignment is one restoration approach that can reduce the
14 potential for bank erosion to the lowest possible value for long term stability.
15 4. THEORY OF CHANNEL REALIGNMENT TO PROVIDE
16 GEOMORPHIC RESTORATION
17 Geomorphic restoration of rivers requires an understanding of both hydraulic engineering and
18 geomorphology. By bringing together geomorphic principles and engineering methods, river
19 restoration can be completed while following a geomorphic-engineering framework for channel
20 restoration design in meandering rivers. By accounting for natural systems variability, the
Area of Very High Bank
Erosion and Tight
Radius of Curvature
Potential Chute Cutoff
Sinuosity of the Displayed
Sub-Reach of 5A = 2.0
Area of High
Bank Erosion
Potential
Chute Cutoff
Erosion Prediction
Moderate
Very High
Extreme
4
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
design framework is an appropriate methodology for generating restoration design solutions that
mimic the natural channel morphologies and environmental attributes in undisturbed systems,
while meeting multifunctional goals of channel stability and enhanced ecologic function. There
are various methodologies and approaches that will be briefly referred to in this paper. In
finding a solution for restoration of a natural meandering river, the practitioner is forced to use
empirical-statistical, analytical and analogue methods. Both the empirical and analogue methods
use reference reach data from other similar watercourses as a basis for design parameters on an
impaired reach. It is not currently in the limits of understanding of geomorphic river engineering
to use only analytical equations to solve the complex problems of a disturbed natural river.
Natural rivers in dynamic equilibrium possess a high degree of morphological diversity, in terms
of cross-sectional, longitudinal, and plan form variability. The physical characteristics of a river
channel include the shape and size of the channel cross section, the configuration of the bed
along the path of the channel, sediment deposits and other in-stream features, the longitudinal
profile, and the channel pattern (Simons, 1979). In addition, biological factors, such as riparian
vegetation, have significant influence on the system. For straight alluvial streams, Lane (1937)
identified a list of factors that may enter into a determination of stable channel shapes:
i) Hydraulic Factors - slope, roughness, hydraulic radius, mean velocity, velocity
distribution, and temperature
ii) Channel Shape - width, depth, and chemical and physical side factors
iii) Nature of Material Transported - size, shape, specific gravity, dispersion,
quantity, and bank and sub-grade material
iv) Miscellaneous Factors - alignment, uniformity of flow, and aging
The number of morphological equations required to obtain a determinate solution of the fluvial
system is controlled by the number of dependent variables that define the hydraulic geometry of
the system. Hey (1978, 1988) identified nine degrees of freedom for natural channels with
sinuous plan forms based on:
i) Cross-Sectional Shape - wetted perimeter, hydraulic radius, maximum depth
ii) Slope
iii) Plan Shape - sinuosity, meander arc length
iv) Velocity
v) Bed Forms - bankfull dune wavelength, bankfull dune height
The controlling, or independent, variables are discharge, input sediment load, bed and bank
sediment size, bank vegetation, and valley slope. With only three established process equations
available (continuity, flow resistance, and sediment transport), the system is indeterminate unless
empirical methods and other assumptions that include analog relationships are applied.
Analogue solutions for channel restoration can include a historical reconstruction of the
disturbed reach or the use of a stable geomorphic reference reach for a natural analogue.
5
-------
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
1 There are numerous practitioners and educators that use various combinations of analytical,
2 empirical and analogue relationships to define a slightly different approach to natural river
3 restoration design for meandering alluvial channels. It is not the purpose of this paper to
4 prescribe one approach or a specific suite of tools for the Housatonic River.
5 Channel realignment can be a valuable design approach when morphological parameters of a
6 disturbed reach fall outside a range likely to provide long-term stability in the context of a
7 particular project. There are numerous interrelated morphological parameters, including width,
8 depth, radius of curvature, sinuosity, riffle slope, and pool-to-pool spacing. Effective channel
9 restoration would consider channel realignment in concert with cross-sectional, plan form, and
10 longitudinal variables. Further, the placement of bank stabilization and habitat structures would
11 be developed after establishing the channel morphology design. Such structures could be used to
12 add additional bank and/or bed protection at critical design locations, redirect flow away from
13 vulnerable banks, and enhance terrestrial and aquatic habitat.
14 5. POTENTIAL APPLICATION OF REALIGNMENT ON THE
15 HOUSATONIC RIVER
16 There are many areas of the current alignment of the Housatonic River in reaches 5 A and 5B that
17 are not within the stable range of pattern and profile morphology compared to other stable
18 reference channels in New York and Massachusetts. For a simplification of this application of
19 channel realignment, three of the numerous morphological parameters were evaluated. The three
20 morphological parameters are listed with an appropriate range for a stable reference reach in
21 Table 1. The morphological parameters and associated ranges are for stable reference reaches of
22 similar channel type, valley type, channel size, and sediment supply. These morphological
23 parameters in Table 1 are dimensionless and normalized by a unit length of the Bankfull Width
24 or Valley Length.
25 Table 1: Reference Dimensionless Morphological Parameters
Morphological Parameter
Minimum Value
Maximum Value
Ratio of Radius of Curvature to Bankfull Width
2.0
4.0
Ratio of Pool to Pool Spacing to Bankfull Width
4.5
6.0
Sinuosity Ratio of Channel Length to Valley Length
1.2
2.0
27 Channel realignment was considered for all areas of the Housatonic River in Reaches 5 A and 5B
28 that did not meet these three morphological parameters. It was assumed that the remainder of
29 reach that met these three parameters could have the banks restored or stabilized with bank
30 stabilization techniques. All of the numerous morphological parameters should be compared to
31 the existing conditions as well as a reference condition to decide the appropriate length of the
32 Housatonic that should be realigned to achieve geomorphic restoration. This task cannot be done
33 until a geomorphic assessment and river survey is completed on Reaches 5A and 5B.
34 5.1 BENEFITS OF CHANNEL REALIGNMENT FOR THE HOUSATONIC RIVER
35 5.1.1 Creation of New Oxbows and Wetlands
36 The current rate of oxbow formation is likely greater than under natural conditions given the past
37 disturbance described above. The river exhibits characteristics that indicate it is not in an
38 equilibrium condition after such disturbance, so the rates of change as the river returns to an
6
-------
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
1 equilibrium condition are greater than a channel that is near equilibrium. Remediation and
2 restoration will result in a slower rate of oxbow creation. However, as part of the channel
3 realignment, there can be additional oxbows and wetlands created in the existing channel that
4 will become abandoned after realignment and geomorphic restoration. Figure 4 displays the
5 horizontal position of potential oxbow wetlands that could be created with the process of channel
6 realignment. The off-channel oxbows and wetlands would be depositional zones and may not
7 require removal of sediments. The transport of the PCBs in the oxbows and wetlands could be
8 reduced by capping the contaminated bed material with clean fill material.
9
10 Figure 4: Example of Potential Oxbows and Wetlands in Reach 5A
11 5.1.2 Floodplain Fill and Capping of the Existing Channel
12 With a realignment approach to restoration, the existing channel location could potentially be
13 used as a disposal location for contaminated sediments. In addition, the realigned channel can be
14 graded through some amount of clean material that could be used as capping material. The
15 channel could then be restored without armoring if the excavation is through clean fill. The
16 combination of placement of floodplain fill from on-site as well as realignment through areas of
17 clean sediments could result in less material taken off-site to achieve the same desired cleanup
18 level. Less material transported off-site would reduce the disturbance to existing wetlands for
19 access roads and the effort that would be needed to transport the contaminated sediments off-site.
Potential
Created
Wetlands
Existing
Channel
to be
Filled
Geomorphic Re-aligned
Channel with Optional
Restored Wetland Areas
Potential Created
Oxbow Wetlands
7
-------
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
1 5.1.3 Construction in the Dry Off-Line
2 The construction of the realigned restored channel can be done off-line in the dry. This
3 construction technique would allow for a more surgical approach to remediation and restoration.
4 An entire sub-reach of the realigned channel could be constructed in the dry, then the water
5 would be diverted to the realigned channel and using common engineering methods the existing
6 channel would be dry for cleanup and remediation. The existing channel could become a created
7 wetland feature or an on-site disposal location for sediments. The offline construction in the dry
8 may reduce construction duration and cost.
9 5.1.4 Sediment and Erosion Control During Construction
10 Construction of an off-line realigned channel will allow for easier management and control of
11 sediment and erosion during construction.
12 5.2 DISADVANTAGES OF CHANNEL REALIGNMENT FOR THE HOUSATONIC
13 RIVER
14 5.2.1 Vegetation Disturbance and Impact
15 Channel realignment will require the removal of a significant amount of vegetation on the
16 floodplain; however, efforts could be made to incorporate these activities into floodplain
17 remediation. It would be possible to incorporate some of the removed vegetation into in-stream
18 grade control, bank stabilization, and habitat structures. It is also possible that the most
19 ecologically or aesthetically important trees or stands of trees or habitats of concern can be
20 mapped and avoided during the design and construction. It should be noted that the majority of
21 the Housatonic River floodplain has been deforested in the past since colonization in the 17th and
22 18th centuries and have returned without the benefit of active restoration. Trees and other
23 vegetation are very important to any restoration project and will need to be re-established as part
24 of the channel restoration.
25 5.2.2 Change in the Character of the Housatonic
26 The Housatonic River currently has a character that will be changed with channel realignment.
27 The realignment would result in a less sinuous channel and a shorter reach length. The character
28 of the River is a subjective parameter that would most likely change with each person's opinion.
29 5.2.3 Additional Testing and Sampling During Construction
30 Channel realignment and geomorphic restoration to achieve the desired risk reduction would
31 require additional excavation that would need additional testing of sediments in the channel and
32 floodplain soils in areas of interest.
33 5.2.4 Change in Property Lines due to Channel Realignment
34 Property lines that were defined by the River centerline would need to be reestablished and
35 surveyed if the River was realigned.
8
-------
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
1 6. SUMMARY OF POTENTIAL CHANNEL REALIGNMENT ON THE
2 HOUSATONIC RIVER
3 Based on this assessment, 9,500 linear feet of the Housatonic River in Reaches 5A and 5B fall
4 outside of the range of the three reference ratios of radius of curvature, pool-to-pool spacing, and
5 sinuosity. This length equals 20 to25% of the total 41,000 linear feet that were assessed from the
6 confluence of the East and West Branches to below New Lenox Road into Reach 5B. The 9,500
7 linear feet should be considered as a minimum length for consideration of channel realignment.
8 A more-detailed geomorphic assessment would need to be carried out to determine if
9 realignment is appropriate for sections of stream outside of this minimum estimate of 9,500
10 linear feet.
11 7. LITERATURE CITED
12 Braden, Gladys B. 1959. Why streams meander. Proceedings of the Oklahoma Academy of
13 Sciences, pp. 104-107.
14 Camporeale, C., Perucca, E., Ridolfi, L., 2008, Significance of cutoff in meandering river
15 dynamics: Journal of Geophysical Research, v. 113, 11 p.
16 Florsheim, Joan L., Jeffrey F. Mount and Anne Chin. 2008. BioScience, vol. 58, no. 6, pp.:519-
17 529.
18 Garcia, M.H., 2008, Sedimentation engineering: processes, measurements, modeling, and
19 practice: Environmental and Water Resources Institute, American Society of Civil Engineers
20 Task Committee to Expand and Update Manual 54, 1132 p.
21 Henry, C.P., and C. Amoros. 1995. Restoration ecology of riverine wetlands: I. A scientific base.
22 Environmental Management. 19(6): 891-902.
23 Hey, R. D. 1978. Determinate hydraulic geometry of river channels. Journal of the
24 Hydraulics Division, Proceedings of the American Society of Civil Engineers,
25 104(HY6), 869-885.
26 Hey, R. D. 1988. Mathematical models of channel morphology. In: Anderson, M. G. (Ed.),
27 Modelling Geomorphological Systems. Wiley, Chichester, 99-126.
28 Holmes, N. T. H. 1993. River restoration/enhancement as an integral part of river management in
29 England and Wales. European Water Pollution Control, 3(3), 27-34.
30 Hooke, J.M., 2004, Cutoffs galore!: occurrence and causes of multiple cutoffs on a meandering
31 river: Geomorphology, v. 61, p. 225-238.
32 Howard, Alan D. 2009. How to make a meandering river. Proc. Natl. Acad. Sci., USA., vol. 106,
33 no.41, pp. 17245-17246.
34 Hupp, C. R., and W. R. Osterkamp. 1996. Riparian vegetation and fluvial geomorphic processes.
35 Geomorphology 14:277-295
36 Kitanidis, Peter K. and John F. Kennedy. 1984. Secondary current and meander formation. J.
37 Fluid Mechanics, vol. 144, pp. 217-229.
9
-------
Use of Channel Realignment along the Housatonic River for Restoration and Remediation of PCB Contamination
1 Lagasse, P.F., 2004, Handbook for predicting stream meander migration, Issue 553, Part 1:
2 National Cooperative Highway Research Program, National Research Council (U.S.),
3 Transportation Research Board, 97 p.
4 Lane. E. W. 1937. Stable channels in erodible material. Transactions of the American
5 Society of Civil Engineers, 102, Paper 1957, 123-142.
6 Langbein, Walter B. and Luna B. Leopold. 1966 .Theory of Minimum Variance. Physiographic
7 and Hydraulic Studies of Rivers, Geological Survey Professional Paper 422-H. United States
8 Government Printing Office, Washington, D.C.
9 Leopold, Luna B., M. Gordon Wolman, and John P. Miller. 1992. Fluvial Processes in
10 Geomorphology. Dover Publications, New York, New York.
11 Nanson, G. 1979. Point bar and floodplain formation of the meandering Beatton River,
12 northeastern British Columbia, Canada. Sedimentology, v.27, no. 1., pp. 3-29.
13 Simons, D. B. 1979. River and canal morphology. In: Shen (Ed.), Modeling of Rivers.
14 Wiley, New York, Chapter 5.
15 Ward, J.V., K. Tockner, D. B. Arscott and C. Claret. 2002. Riverine landscape diversity.
16 Freshwater Biology, vol. 47, pp. 517-539.
10
-------
ATTACHMENT B-2 CHANNEL DYNAMICS AND ECOLOGICAL
CONDITIONS IN THE HOUSATONIC RIVER PRIMARY STUDY AREA
-------
/
Biohabitats
The Stables Building
2081 Clipper Park Road.
Baltimore, MD 21211
tel 410 554 0156
fax 410 554 0168
www.biohabitats.com
February 2, 2012
Scott Campbell
Weston Solutions
10 Lyman Street
Pittsfield, MA 01201
Re: Channel Dynamics and Ecological Conditions White Paper
Dear Mr. Campbell,
Biohabitats is pleased to submit the attached white paper titled "Channel Dynamics and
Ecological Conditions in the Housatonic River Primary Study Area." The paper presents
principles of stream meandering, reviews data related to the Housatonic River in the Primary
Study Area (PSA), and describes wetlands in the PSA and associated Massachusetts
Endangered Species Act (MESA) species.
The white paper reflects the joint efforts of Biohabitats, Stantec, and Field Geology Services, as
well as input from the EPA and its wider technical consulting team members.
We appreciate the opportunity to work as a member of your team on this project.
Sincerely,
BIOHABITATS, INC.
Ellen M. McClure, C.E.
Senior Fluvial Geomorphologist
cc: Susan Svirsky, EPA
-------
TABLE OF CONTENTS
1. Introduction l
2. Meander Theory and Meander Formation Literature Review l
3. Role of Meander Migration in Creating and Maintaining Biological Diversity 5
4. Comparison of Reach 5 and 6 River Geometry with Other Regional Rivers 6
5. History of the Housatonic River Channel in Reaches 5 and 6 8
6. Rate of Oxbow Creation and Infilling 11
7. Existing Ecological Conditions 15
7.1 Overview of Natural Communities in Reach 5 and 6 16
7.2 Overview of Wetland Function and Values in Reach 5 and 6 19
7.3 Overview of Massachusetts Endangered Species Act and Priority Habitat in Reach 5 and 6 of
Rest of River 21
7.4 State-Listed Species Habitat Use In Reach 5 And 6 25
8. Summary 30
9. Literature Cited 32
APPENDIX A 36
APPENDIX B 47
LIST OF TABLES
Table 1 Meandering Characteristics of the Housatonic River and Other Rivers in New England and New
York 6
Table 2 Lengths of Housatonic River Oxbows Formed in Different Time Periods 12
Table 3 Elevations of Oxbows, Channel, and Adjacent Floodplain 15
Table 4 Acreage Calculations for the Wetland Communities Within Reaches 5 and 6 17
Table 5 Sizes of Oxbow Wetlands by Natural Community Type Within Reach 5 and 6 17
Table 6 Priority Habitat Acreage Calculations for the 25 State-Listed Species, as Identified by the NHESP,
Within the PSA 23
Table 7 Habitat Descriptions for the 25 State-Listed Species of Concern, as Identified by the NHESP, in
Reach 5 and 6 26
LIST OF FIGURES
Figure 1 Definition of Meandering Planform Characteristics 2
Figure 2 Magnitude of Channel Adjustments Through Time Following a Disturbance Relative to a
Quasi-Equilibrium Condition 4
i
-------
Figure 3 Meandering Characteristics of the Housatonic River and Other Rivers in New England and
New York 7
Figure 4 Topographic Maps from a) 1886 and b) 1988 Showing Reestablishment of Meanders along
an Artificially Straightened Reach of the Housatonic River just Upstream of Woods Pond 9
Figure 5 Topographic Maps from a) 1886 and b) 1988 Showing the Reformation of Meanders Along
an Artificially Straightened Section of the Housatonic River just Downstream of the Confluence of the
East and West Branches 10
Figure 6 Estimates of Time to Infill Oxbows Based on a) Linear, or Constant, Rate of Infilling and B)
Logarithmic, or Reduced, Rate of Infilling 14
-------
1. INTRODUCTION
This paper summarizes the current understanding of stream meander migration theory. There are
three main objectives of this effort:
1. Document the current understanding of stream meanders and the role they play in the
formation of floodplain geomorphological features such as point bars, scroll bars, cutoffs,
floodplain/oxbow wetlands, backwaters, sloughs, and vernal pools.
2. Present the current data on how the Housatonic River channel in the Primary Study Area
(PSA) relates to other river systems in the region, the history of channel migration in the
PSA, and the rates of channel change.
3. Describe, to the extent possible, the wetlands in the PSA and how those wetlands relate to
the Massachusetts Endangered Species Act (MESA) species in the PSA.
2. MEANDER THEORY AND MEANDER FORMATION LITERATURE
REVIEW
Alluvial rivers, rivers with the freedom to migrate across a self-formed floodplain, have a
propensity for developing a meandering planform (Leopold, 1994). A meandering river can
ultimately develop from a straight channel alignment as small perturbations, such as sediment
input from a tributary, result in large-scale meanders (Xu et al., 2011). As meanders develop,
they eventually reach a quasi-equilibrium state with the meander wavelength, amplitude, corridor
width, and radius of curvature staying relatively stable despite continuing meander growth and
channel migration (Xu et al., 2011; Figure 1). Although the time needed to reach an equilibrium
state will vary by river, the river tends towards a condition where a minimum rate of energy
dissipation occurs along its length (Xu et al., 2011). As such, sharp right angle bends along
alluvial rivers do not persist, because energy expenditure along the length of the channel is
focused at one point - the sharp bend. Ultimately, given such a sharp bend, a river adjusts
through erosion and deposition to form a smooth meander where the amount of turning, and
therefore energy expenditure, at any given point in the bend is equal to all other points.
Empirical studies of meander dimensions document several relationships that hold for rivers of
all sizes, such as a value of 11 for the ratio between meander wavelength and channel width
(Leopold, 2004). The consistency of meandering relationships suggests meandering is a
transient process that tends, as sinuosity increases, toward a planform equilibrium (Garcia,
2008).
A common process on meandering rivers is the cutting off of meander bends and creation of
oxbow ponds (Figure 1). Cutoffs are an inherent part of meandering behavior (Hooke, 2004) and
help a river maintain a stable state by preventing the channel length and sinuosity from becoming
too great and developing an unstable planform configuration (Camporeale et al., 2008). Two
l
-------
cutoff mechanisms are widely recognized: neck cutoffs and chute cutoffs (Constantine et al.,
2010). Neck cutoffs occur through bank collapse when the banks of two adjacent meanders
erode towards each other and eventually meet. A chute cutoff results when a new channel carves
across the inside bend of a meander and becomes the dominant conveyor of river discharge. The
processes of meander evolution ultimately enhance the likelihood of cutoffs developing by these
two mechanisms. Empirical studies of meander geometry show the radius of curvature of
meander bends generally falls within a range near 2.4 times the bankfull channel width (Garcia,
2008), implying this is an equilibrium dimension to which meandering rivers evolve (Lagasse,
2004).
5im»sity = Channel length / valley length
Lc = Channel length
W = Bankfull channel width
Rc = Radius of curvature
Figure 1 Definition of Meandering Planform Characteristics
Numerous studies also demonstrate that accelerated bank erosion rates on meandering rivers
occur in meander bends with a radius of curvature between 2 to 3 times the bankfull channel
width (Begin, 1981; Nanson and Hickin, 1986; Hooke, 1997). Consequently, as a meandering
river approaches an equilibrium condition, the rate of bank erosion, a process that ultimately
results in neck cutoffs, increases. This relationship also illustrates that geomorphic stability does
not imply channel position is static through time, because a river in dynamic equilibrium can
migrate across its floodplain through bank erosion on the outside bend of meanders while
2
-------
maintaining the same channel dimensions due to an equivalent amount of deposition on point
bars on the inside bend of meanders. A similar inherent tendency in meander development also
increases the probability of chute cutoffs developing. As channel sinuosity increases through
meander growth, the corresponding decrease in channel slope leads to channel aggradation (i.e.,
deposition) (Knighton, 1998) and ultimately increases the amount of overbank flow available to
carve a chute cutoff channel across the inside of a meander bend (Thompson, 2003).
Numerical modeling indicates channel sinuosity will increase along a meandering river until a
critical sinuosity of 3.14 (pi) is reached (Stolum, 1996). Once this critical value is reached or
exceeded, a cluster of cutoffs, both in space and time, is likely to occur. While the idealized
sinuosity value of 3.14 is rarely reached on real rivers, empirical evidence does suggest clusters
of cutoffs do occur once a critical sinuosity value is reached (Hooke, 2003). Channel
adjustments resulting from the decrease in channel length associated with a single cutoff tend to
promote the development of additional nearby cutoffs shortly after the initial one occurs (Stolum,
1996). As a result of multiple cutoffs, the channel sinuosity will fall below the critical sinuosity
and a period of meander growth will subsequently ensue, so the channel can once again tend
towards the critical sinuosity (Hooke, 2003). Consequently, meanders oscillate in sinuosity,
through alternating periods dominated by meander growth and cutoffs, respectively, to maintain
a critical sinuosity or equilibrium condition.
While intrinsic meandering dynamics control cutoffs and the formation of oxbows, the location
and frequency of cutoffs are also controlled by external conditions. Vegetated floodplains are
less likely to experience cutoffs because of the added floodplain resistance that slows the rate of
erosion (Constantine et al., 2010). Similarly, floodplain stratigraphy also controls cutoff
development. Clay plugs resulting from the infilling of older oxbows are more resistant to
erosion than the surrounding floodplain deposits, potentially reducing the rate of meander
migration and frequency of cutoffs. Maximum meander migration on the lower Mississippi
River, where clay plugs are common, occurs on meander bends with a radius of curvature of less
than 2 rather than between 2 and 3 times the channel width as observed on other rivers with
homogeneous floodplain stratigraphy (Hudson and Kesel, 2000). Fine-grained sediment loads
that might result from erosion of cohesive bank materials, such as clay, favor the development of
meandering channels with a higher sinuosity (Schumm and Khan, 1972). Channel sinuosity, and
in turn the likelihood of oxbow formation, is also controlled by valley gradient, with higher
sinuosities associated with lower gradients (Schumm, 1979). High discharges accompanying
floods are often the triggering event that causes cutoffs, although intrinsic meandering dynamics
ultimately control the location and number of such cutoffs (Hooke, 2004). Finally, human
activities can alter the rate of meander migration and oxbow development. Watersheds with
dams and channels that have been armored show decreased rates of channel migration, although
the planform dimensions (e.g., wavelength, amplitude, and radius of curvature) may remain
unchanged (Ollero, 2010; Magdaleno and Fernandez-Yuste, 2011).
3
-------
Natural events (e.g., large floods) and human activities (e.g., channel straightening) can
sometimes greatly alter the channel configuration such that the river is temporarily removed
from an equilibrium condition. Following the channel-altering disturbance, the river will
undergo a series of adjustments that will bring the channel back into equilibrium (Petts, 1994).
The adjustments will initially be very rapid but the magnitude and rate of change will decline as
the river once again approaches equilibrium (Figure 2). In some instances when the disturbance
permanently alters the watershed conditions, such as through urbanization, the river will not
return to the former equilibrium state but will achieve a new equilibrium condition with channel
dimensions different from those associated with the earlier equilibrium state.
t- quasi-equilibrium
Time
Figure 2 Magnitude of Channel Adjustments Through Time Following a Disturbance
Relative to a Quasi-Equilibrium Condition
Source: From Petts (1994)
A meander cutoff represents a small channel disturbance that results in local channel
adjustments. Following a cutoff, the newly created channel undergoes incision that migrates
upstream as a headcut due to the shortened stream length and increased slope (Hooke, 1995).
Vertical accretion of sediment occurs most dramatically at the upstream end of the oxbow
created by the cutoff as overbank flows enter the now-abandoned channel (Hooke, 1995;
Lagasse, 2004). The changes following a cutoff are initially very rapid, but the rate of change
declines with time (Hooke, 1995) such that oxbows can persist for hundreds of years before
becoming completely infilled with sediment (Lagasse, 2004).
Many of the basic principles of meandering rivers, point bar development, and floodplain
formation come from studies of rivers in the mid-Atlantic region of the United States (Wolman,
1955; Wolman and Leopold, 1957). Recent studies show that these rivers actually flow through
impoundment sediments deposited behind old mill dams and do not represent meandering rivers
4
-------
flowing through self-formed floodplains (Walter and Merritts, 2008). Consequently, the
presence of a meandering planform cannot immediately be assumed to develop from a standard
sequence of processes or events. Alternate models of meander formation must be considered
when studying any given river system and a full understanding of the history of the river
considered in the development of such models.
3. ROLE OF MEANDER MIGRATION IN CREATING AND
MAINTAINING BIOLOGICAL DIVERSITY
Meander migration, and the resulting channel cutoffs and oxbow lake/wetland formation are
well-documented and reasonably understood phenomena (Howard, 2009). Meander migration is
a complex interaction of many variables. Channel geometry, high-stage flows, sediment load and
transport, bank material resistance and type of vegetation on the banks all factor into the process,
making it very dynamic across the range of variables involved. The meander migration process
generally erodes material from one area of the channel and deposits material in another through
the erosion, sediment transport and sediment deposition processes. Meander migration is often
most pronounced during high flow or flooding events.
As the erosion/deposition of material in meander migration takes place, the geometry and
planform of the channel changes. The channel moves, and changes shape. Areas that once were
channel banks and associated riverine floodplain are removed, and new areas, where eroded
material is deposited, are created, often in predictable locations such as oxbow lakes, point bars
and scroll bars.
The movement of the planform of the channel and associated formation of channel cutoffs,
oxbow lakes, point bars and scroll bars creates topographic variations within the active
floodplain. As variations in topography in the floodplain become more pronounced, they in turn
create a wider range of hydrologic regimes within the floodplain, depending on the elevation,
proximity to the active channel, and depth to groundwater in any given location.
The combination of the range of floodplain elevations and hydro-periods created by meander
migration, and the resultant oxbow lake and bar formations, have been shown to dictate
vegetation patterns and potentially provide diverse floral and faunal communities in an array of
wetland and upland habitats (Nanson, 1979; Hupp and Osterkamp, 1996; Ward et al., 2002).
Henry and Amoros (1995) state that the "most widely valued function of wetlands, particularly
for riverine wetlands, is their contribution to the maintenance of regional biodiversity."
Florsheim et al. (2008) state that bank erosion and the consequent meandering of rivers is
beneficial, because it "is a geomorphic process that promotes riparian vegetation succession and
creates dynamic habitats crucial for aquatic and riparian plants and animals."
5
-------
4. COMPARISON OF REACH 5 AND 6 RIVER GEOMETRY WITH
OTHER REGIONAL RIVERS
The meandering characteristics of the Housatonic River are not unique in the northeastern United
States; other rivers in the region share similar meandering characteristics (Table 1; Figure 3;
Appendix A). For this study, portions of nine meandering rivers displaying the most well-
developed meandering conditions on those rivers were compared to determine if the meandering
traits of the Housatonic River were more strongly developed than elsewhere. The measured
meandering parameters were normalized to channel width in order make meaningful
comparisons between the rivers of varying size (Figure 3). Of the nine meandering rivers
selected, the sinuosity (as defined in Figure 1) was greatest on the Housatonic River, but only
slightly higher than the Saco River and Fort River. When comparing other meandering traits,
such as radius of curvature and meander amplitude, values for the Housatonic River fall within
the range of values measured for the nine northeastern rivers (Table 1; Figure 3). For the two
parameters best reflecting the amount and lateral extent of floodplain wetlands associated with
meandering rivers (i.e., length of oxbow per valley mile and the width of the meander corridor),
the values for the Housatonic River fall below the trend line for the nine rivers (Figure 3).
Consequently, habitat features, such as oxbows, should be considered better developed on other
rivers in the northeastern United States than on the Housatonic River.
Table 1 Meandering Characteristics of the Housatonic River and Other Rivers in New
England and New York
Meander
Radius of
Meander
Length of
wavelength
curvature
amplitude
oxbows
Corridor width
Drainage
Channel
(ftychannel width (ft)/channel width (ftychannel width
(fty valley
(ftychannel width
V\tatershed
Area (mi2)*
Location
Sinuosity width (ft)
(ft)
(ft)
(ft)
mile
(ft)
Say/will River
32
Montague, MA
1.83
40
7.5
2.5
3.8
1320
13.8
Fort River
56
Hadley, MA
2.13
50
10
2.4
4
2560
15
Baker River
143
Rumey, NH
1.54
100
9
2.7
4
4197
15
Housatonic River
148
Pittsfield, MA
2.27
90
5.6
1.9
3.1
3761
11.1
Batten Kill
149
Arlington, VT
1.36
70
10.7
1.9
3.2
3210
10
Poultney River
175
Fair Haven, VT
2
60
6.7
1.8
2.9
7329
16.7
Contoocook River
221
Eteerirg, NH
1.74
80
8.1
2
4.7
5659
20
Kinderhook Geek
316
Kinderhook, NY
1.45
140
7.1
2.1
2.7
6102
15
Saco River
444
Fryebirg, ME
2.25
300
13.3
3
9.2
8865
21.7
* Drainage area at site of interest.
6
-------
2.50
2.00
1.00
0.50
0.00
Sinuosity
£ 14.0
I 12.0
c
| 10.0
0
1 8.0
O)
1 6.0
>
CO
J 4.0
®
i20
2 0.0
50 100 150 200 250 300 350 400 450 500
Drainage area (mi2)
Meander wavelength
50 100 150 200 250 300 350 400 450 500
Drainage area (mi2)
£ 3.5
J 3.0
s
J 2.5
"5
© 2.0
5
ro . _
> 1.5
3
O
"S 1.0
CO
12 0-5
re
CC
0.0
Radius of curvature
50 1 00 150 200 250 300 350 400 450 500
Drainage area (mi2)
^ 10
| 9
¦$ 8
I 7
f 6
CD
| 5
Q- 4
E
re 3
CD
¦o 2
re
I 1
0
Meander amplitude
50 1 00 150 200 250 300 350 400 450 500
Drainage area (mi2)
10000
£ 9000
£
^ 8000
re 7000
6000
5000
4000
3000
2000
1000
0
1
Length of oxbows
-V
0
50 100 150 200 250 300 350 400 450 500
Drainage area (mi2)
25
20
j2 15
"C
V. 10
5
o
Corridor width
50 1 00 1 50 200 250 300 350 400 450 500
Drainage area (mi2)
~ Housatonic River
~ Other selected rivers in New England and New York
Figure 3 Meandering Characteristics of the Housatonic River and Other Rivers in
New England and New York
7
-------
5. HISTORY OF THE HOUSATONIC RIVER CHANNEL IN REACHES
5 AND 6
As was the case for rivers throughout New England, much of the Housatonic River was
artificially straightened in the past (WESTON, 2011). Topographic maps and aerial
photographs, both recent and historic, can be used to identify where channel straightening has
occurred by looking for three hallmarks of artificial straightening: 1) straight segments longer
than the wavelength of adjacent meanders; 2) straight segments that "hug" the valley sides
despite an adjoining wide floodplain on which meanders could form; and 3) the presence of
former meanders adjacent to straight channel segments (WESTON, 2011). Such evidence is
seen on the Housatonic River (Figure 4) and an analysis of topographic maps and aerial
photographs reveals at least 92 percent of the Housatonic River from the confluence of the East
and West Branches downstream to Woods Pond was artificially straightened. Much of this
straightening may have occurred in association with railroad construction completed in the
1850s, but agricultural practices and other land uses from perhaps even earlier likely were
important contributors as well. Whatever the exact date of channel manipulations, the
straightening was certainly completed prior to the time of the 1886 survey used to complete the
historic topographic map of 1893 (Web citation 1). Whatever the exact reason for and timing of
the straightening, the large-scale manipulation of the river channel represents a major
disturbance that would have shifted the channel away from the quasi-equilibrium condition
existing at the time of straightening (Figure 2).
In response to the artificial channel straightening and removal from a quasi-equilibrium state, the
Housatonic River has undergone a period of channel adjustment that has resulted in the natural
reformation of meanders along much of its length. By 1945, at least 55 percent of the
straightened channel had redeveloped a meandering planform (Web citations 2 and 3). The
redevelopment of meanders along artificially straightened channels has been documented
elsewhere in New England (Field, 2007) and other parts of the world (Ollero, 2010). Meanders
have reformed along artificially straightened channels by two primary mechanisms: 1) breakouts
and 2) build outs (WESTON, 2011). First, sediment, ice, or wood can clog the channel, allowing
flows to breakout rapidly across the floodplain and carve a new meander. Second, sediment
building out into the channel at the mouths of tributaries can force the river flow into the
opposite bank, with the ensuing bank erosion leading to the formation of a new meander. The
creation of single simple meanders through these processes has occurred on the Housatonic
River as documented through comparisons of historic and recent topographic maps (Figure 5).
Complexes of multiple meanders have also evolved from straightened reaches (Figure 4b) and
may have grown from a single breakout or build-out meander that served as the minor initial
perturbation required to develop the more complex meandering planform (Xu et al., 2011). In
some instances, the reformed meanders may be simply reoccupying old meanders abandoned
during the channel straightening.
8
-------
Figure 4 Topographic Maps from a) 1886 and b) 1988 Showing Reestablishment of
Meanders along an Artificially Straightened Reach of the Housatonic River just Upstream
of Woods Pond
Note: Evidence for straightening includes channel "hugging" the valley side, straight reaches longer than the natural
meander wavelength seen at upper end of the 1886 map, and presence of abandoned meanders adjacent to the
straightened channel seen on the 1988 map.
Long reaches of the river remain in a straightened configuration and are presumably not sensitive
to the breakout or build-out processes of meander reformation. The areas most sensitive to the
development of meanders are upstream of valley constrictions where floodwaters are impounded
and flows more easily escape the channel, breakout across the floodplain, and carve a new
meander. On the Housatonic River between the confluence of the East and West Branches and
Woods Pond, several natural and artificial valley constrictions are present. A natural valley
constriction is formed by a high ridge of glacial deposits that extends across most of the valley
bottom just upstream of the Pittsfield Wastewater Treatment Plant (WWTP). Among other
locations, artificial valley constrictions have been created where: 1) the sewer pipe crosses the
9
-------
Figure 5 Topographic Maps from a) 1886 and b) 1988 Showing the Reformation of
Meanders Along an Artificially Straightened Section of the Housatonic River just
Downstream of the Confluence of the East and West Branches
Note: Red boxes are in the same location on both maps.
river 0.7 mi upstream of the WW TP. 2) the elevated road grade crosses the floodplain at New
Lenox Road, 3) the railroad grade narrows the floodplain quite significantly across from and
continuing downstream of the Roaring Branch confluence, and 4) flow is backwatered behind the
Woods Pond Dam. Some of the most dramatic meandering on the Housatonic River has formed
immediately upstream of these constrictions, but unless the river "hugs" the valley sides (Figure
4), the meandering is too well developed to definitively determine if these meanders have
redeveloped from an artificially straightened channel.
Since 1945, no new meanders have formed along the Housatonic River between the confluence
and Woods Pond (although oxbows have been created, see below). The meanders have
continued to grow, with migration occurring through erosion of the outside bends at a rate no
greater than 0.9 ft/yr since 1952 (WESTON, 2006). It should be noted that the average erosion
10
-------
rate for the entire channel in Reaches 5A and 5B is on the order of 0.3 ft/yr (Stantec, 2009). The
lack of new meander formation along reaches that remain artificially straightened might be the
result of meanders having already reformed in the most sensitive areas (e.g., upstream of valley
constrictions, near the confluences of larger tributaries) and a lack of sufficient wood and
sediment to clog the channel and force the creation of new meanders in less-sensitive areas.
Sediment loads in New England rivers were high following widespread land clearance in the
1800s and generated considerable channel change (Brakenridge et al., 1988; Bierman et al.,
1997; Bierman et al., 2005), but sediment loads were greatly reduced by the 1940s as forests
redeveloped in the upper watersheds. Despite the regrowth of the forests, wood loads have
remained low in most New England rivers because of the periodic removal of wood from river
channels, a management practice that continued on the Housatonic River until at least the 1970s
(WESTON, 2011). The lack of further meander formation may also indicate the Housatonic
River is more closely approaching a quasi-equilibrium state where the frequency of significant
channel adjustments would be expected to decline (Figure 2; Petts, 1994).
The history of channel straightening and manipulation clearly demonstrates that the Housatonic
River is not a pristine fluvial system that has naturally meandered, undisturbed, across its
floodplain for thousands of years. Rather than resulting in permanent change, the river has
adjusted to the channel straightening in order to restore a quasi-equilibrium. The recreation of
well-developed meanders in less than a century along much of the straightened river channel
indicates the capacity of the river to recover from large-scale perturbations in a relatively short
time frame.
6. RATE OF OXBOW CREATION AND INFILLING
Although many of the oxbows present along the Housatonic River today were likely created
when long sections of the river channel were abandoned during artificial straightening, oxbow
formation also occurs naturally and eight oxbows have formed in the last 70 years, as determined
from historical aerial photographs, between the confluence of the East and West Branches and
Woods Pond (Appendix B). A similar number of oxbows occurred in the same time period
elsewhere on the Housatonic River (Pierce, 2006), perhaps resulting from continuity in the
setting or some autogenic equilibrium tendency. Because the oxbows remain as a low spot on
the floodplain after they are cut off from the main channel, they represent an important
component of the floodplain wetland complex. Oxbows also serve as sinks for sediment carried
by floodwaters, so they eventually fill in to the level of the surrounding floodplain. Along a
naturally meandering river in a state of quasi-equilibrium, new oxbows are created by cutoffs at a
sufficient rate to replace the wetland habitat lost through the infilling of older oxbows. At what
rate the oxbows fill is an important determinant of the location, distribution, and heterogeneity of
wetland habitats on the floodplain. While the literature suggests hundreds of years are needed to
fill in oxbows (Lagasse, 2004), an analysis of aerial photographs and topographic surveys was
11
-------
completed, and presented below, to ascertain how long oxbows persist on the Housatonic River
floodplain.
The locations of all oxbows between the Confluence and Woods Pond were mapped through
aerial imagery interpretation (Appendix B). Oxbows were identified by their arcuate form and/or
their continuing connection to the current river channel. Other wetlands without these
characteristics may also be oxbows but, being less definitive, were not included in the mapping.
An analysis of historical aerial photographs extending back to 1942 allowed oxbows created in
the last 70 years to be dated by bracketing their time of formation between two sets of aerial
photographs (Appendix B). Once mapped, the length of each oxbow was measured and the total
length of oxbows created in any given time period compiled. Since 1942, nearly one mile of
oxbow has been created (Table 2), representing 16.0 percent of the total length of oxbows
present. If a constant rate of oxbow formation is assumed through time, then all of the oxbows
on the floodplain would have taken more than 435 years (=70 years/0.16) to form. However, a
constant rate of oxbow formation cannot necessarily be assumed as many, if not most, of the
oxbows were created as the channel was artificially straightened. After channel straightening, a
long period of channel adjustment and meander reformation would need to occur before the
channel sinuosity was, once again, high enough to promote cutoffs and oxbow formation.
Consequently, over the long term, the total length of oxbows seen on the Housatonic floodplain,
although not formed at a constant rate, may roughly equate to what would have formed under
natural conditions.
Table 2 Lengths of Housatonic River Oxbows Formed in Different Time Periods
Date of Oxbow Formation
Length (mi)
Pre-1942
5.13
1942-1952
0.12
1952-1972
0.21
1972-1990
0.13
1990-2000
0.49
2000-2011
0.00
Total Length of River Segment
11.39
Total Length of Oxbows
6.08
Length of Oxbows Formed in 70 Years
0.95
The estimated 435 years required to form the total length of oxbows present must be considered
a minimum value. The rate of cutoffs and oxbow formation in the last 70 years was likely much
more rapid than on a natural undisturbed river for at least two reasons. First, in response to
channel straightening, the river has likely been undergoing a period of accelerated adjustment as
is typical of rivers following a major disruption (Figure 2; Petts, 1994). The Housatonic River
first experienced a period of rapid meander reformation prior to 1945, but since that time may be
experiencing a period of accelerated oxbow formation. Clusters of cutoffs are known to occur
12
-------
when a critical sinuosity value is reached or exceeded (Hooke, 2003), a condition that may have
been attained on the Housatonic River once sufficient sinuosity had been regained through
meander reformation. Second, many of the oxbows have formed in the last 70 years in proximity
to recent human alterations in the channel (e.g., treatment plant outfall) or artificial valley
constrictions (e.g., sewer pipe crossing). The disruption of hydraulic and geomorphic processes
around these human impacts likely increased the probability of cutoffs occurring. For both
reasons, the amount of oxbow formation in the last 70 years should be considered artificially
high in response to both past and recent human activities. Consequently, the rate of oxbow
formation before the 1940s was likely slower, so more than the estimated 435 years would have
been necessary to form all of the oxbows observed in the PSA.
The time required to form all of the oxbows on the floodplain can also be equated with the time
required to completely infill any given oxbow because earlier oxbows that have infilled would no
longer be visible on the floodplain and, therefore, not counted in the total length of oxbows
present. To independently corroborate that the oxbows are infilling over time periods in excess
of 400 years, topographic cross sections were analyzed to establish to what extent existing
oxbows of different ages have infilled. If the assumption is made that the bottom of the oxbow
at the time of its formation was at the same elevation as the present channel bottom, the current
difference in elevation of the oxbow bottom relative to the channel bottom represents the amount
of infilling that has occurred since the formation of the oxbow. Fourteen sites were identified
where cross section surveys had been conducted across oxbows and the adjacent active channel,
so bottom elevations of both could be compared (Table 3).
Three of the surveyed oxbows formed in the last 70 years, so the date of the first aerial
photograph showing the oxbow cutoff from the active channel provides the latest possible date
for its formation. The remaining 11 surveyed oxbows formed prior to the earliest aerial
photographs and are assumed to have been created 150 years ago when railroad construction and
other activities in the valley likely led to channel straightening and abandonment of the former
meandering channel. With the ages of each oxbow either directly established or assumed, an
estimate can be made of the time required to completely infill the oxbows based on the
percentage of infilling that has already occurred with the difference in elevation between the
current channel bottom and the adjacent floodplain surface representing the baseline for
comparison. Assuming a constant rate of infilling, or a linear relationship, the time required to
completely infill the oxbows is approximately 350 years (Figure 6a).
However, the rate of change following a cutoff decreases with time (Hooke, 1995), so a
logarithmic relationship better describes changes in the rate of infilling through time. When a
logarithmic relationship is applied to the oxbow infilling data on the Housatonic River, more
than 500 years would be expected to pass before 70 percent of any given oxbow is filled with
floodplain sediment and organic matter (Figure 6b). A decrease in the rate of infilling through
time can also be expected from changes in land use through time.
13
-------
I
e
5
r
±
100
90
8a
JO
SO
so
do
30
20
10
0
RJ = 0.47559
-X-
b( lLIO
yo
a so ioo iso 200 230 jud iso sao 45D sot?
nme since cwboiv formation (yrs)
SO
70
I
§ 60
? 50
R*=: 0.56266
E so *-
I
£ 30 4
20
10
30 100 150 200 230 300 350
Time since cxbcr* formation Ivrsl
suti
.'.50
5UU
Figure 6 Estimates of Time to Infill Oxbows Based on a) Linear, or Constant, Rate of
Infilling and B) Logarithmic, or Reduced, Rate of Infilling
14
-------
Sediment loads have decreased over time due to a transition from agricultural to forested land
use, so less sediment is available over time to contribute to the infilling of the oxbows. Although
the data are limited and assumptions regarding the age of older oxbows uncertain, the
topographic cross sections corroborate the findings of the floodplain mapping (Appendix B) that
oxbows on the Housatonic River floodplain are infilling over several centuries and not decades.
Table 3 Elevations of Oxbows, Channel, and Adjacent Floodplain
Years since
Oxbow
Channel
Cross
oxbow
bottom
Channel bottom
Floodplain
Oxbow elevation
elevation below
Percent
Section #
formation
elevation (ft)
elevation (ft)
elevation (ft) above channel (ft)
floodplain (ft)
infilling
167 and 111
2
954.8
952.2
958.7
2.6
6.5
40
175 and 121
27
952.5
949.8
955.1
2.7
5.3
51
222 and 218
2
944.2
944.5
954.7
-0.3
10.2
0
274 and 100
150
948.8
937.2
953.9
11.6
16.7
69
278 and 74
150
947.1
940.3
949.6
6.8
9.3
73
284 and 58
150
945.4
939.6
950.5
5.8
10.9
53
286 and 42
150
942.9
938.2
950.4
4.7
12.2
39
292 and 29
150
944.5
935.4
949.2
9.1
13.8
66
R5- 1
150
956.4
952.8
959.9
3.6
7.1
51
R5-2
150
956.3
950.8
959.8
5.5
9
61
R5-3
150
954.0
948.8
958.8
5.2
10
52
R5-4
150
949.5
943.6
954.5
5.9
10.9
54
R5-5
150
948.9
944.9
950.5
4
5.6
71
R5-6
150
946.1
940.8
949.8
5.3
9
59
Note: The R5 cross sections are numbered sequentially from upstream to downstream.
7. EXISTING ECOLOGICAL CONDITIONS
The entire Massachusetts length of the Housatonic River flows through the Western New
England Marble Valleys ecoregion (NHESP, 2010). The calcium-rich marine sediments of the
ancient seafloor were transformed to marble, and it is the underlying marble that makes the
Housatonic watershed unique (Woodlot Alternatives, 2002).
The approximately 10 miles of river in Reaches 5 and 6 generally ranges from 45 to 100 feet in
width, is bordered by extensive floodplains (up to 3,000 feet wide), and has a meandering pattern
with point bars, cut banks, and the persistence of backwater sloughs, abandoned channels, and
alluvial bars, as well as oxbows and backwaters throughout. Portions of the lower reaches of the
floodplain are inundated by water impounded by Woods Pond Dam. Channel widths range from
approximately 40 to 60 feet in the upper reaches near the confluence due to topography and
development of the historic floodplain, and approximately 60 to 120 feet in the lower reaches
that are not influenced significantly by Woods Pond Dam (i.e., flooding in the main channel and
backwater wetlands increases width significantly near Woods Pond). Stream depths range from
approximately 1.5 ft in urbanized areas to more than 8 ft under baseflow conditions, where
natural meanders, cut banks, and point bars have developed. The substrate of the upper channel
contains coarse gravels, cobbles, and small boulders, with occasional mid-stream bars of coarse
sands. Downstream from the confluence, there are larger sand deposits in point bars.
15
-------
Consideration of anthropogenic activities is of particular interest because of the land use history
and the effect that past human activities have had in shaping ecological conditions and processes
in the PSA (including meandering processes, downstream transport of contaminated sediment via
accelerated bank erosion, and settlement in the floodplain during flooding events).
The Pittsfield Wastewater Treatment Plant (WWTP), which discharges its effluent near the
midpoint of Reach 5, contributes an average flow of 0.5 cubic meters per second to the
Housatonic River (Harrington Engineering and Construction, Inc., 1996).
Clearing of riparian areas for development purposes, including urban development in the upper
3.1 miles, has occurred in the floodplain throughout Reaches 5 and 6. Portions of the floodplain
have been cleared for various purposes, primarily agriculture, residences, and various rights-of-
way (e.g., roads, railroads, power lines). Agricultural disturbances are the major source of forest
clearing within the riparian zone of the upper Housatonic River. Agricultural fields, including
corn and hay fields, are a predominant land use within Reach 5 and have affected the size of the
natural riparian habitats in the middle section of the Reach 5 and downstream sections near New
Lenox Road. Much of the upper two-thirds of Reach 5 appears to have been cleared for
agriculture at one time (Woodlot Alternatives, 2002).
Near and within areas that were previously disturbed, portions of the floodplain are inhabited by
multiple non-native and invasive shrub, herb, grass, sedge, and aquatic species (Woodlot
Alternatives, 2002). In Reach 5A, invasive species are prevalent in the floodplain where
ornamental shrubs [Morrow's honeysuckle (Lonicera morrowii), common privet (Ligustrum
vulgare), Chinese spindle-tree (Euonymus fortunei), European spindle-tree (Euonymus
europaea), and winged burning bush (Euonymus alatus)] have escaped from adjacent urban
areas. Other common invasive non-native herbs have also colonized floodplain forests [e.g.,
dame's rocket (Hesperis matronalis), goutweed (Aegopodium podagraria), garlic mustard
(Alliaria petiolata) and moneywort (Lysimachia nummularia)]. Invasive species such as purple
loosestrife (Lythrum salicaria), common reed (Phragmites australis), yellow iris (Iris
pseudacorus), and reed canary grass (Phalaris arundinacea) occur, and in some areas, dominate
the shoreline and marsh communities, including wet meadows and shrub swamps. Eurasian
water-milfoil (Myriophylum spicatum), water chestnut (Trapa natans), and crisped pondweed
(Potamogeton crispus) are also documented in Woods Pond and adjacent wetlands flooded by
the impoundment created by the dam. Water chestnut has increased in abundance from a few
plants in 1998-2000, and is now a dominant species within Woods Pond (John Lortie, personal
communication).
7.1 OVERVIEW OF NATURAL COMMUNITIES IN REACH 5 AND 6
Significant portions of Reach 5 and 6 (sometimes referred to as the Primary Study Area or PSA)
are open palustrine wetlands and riverine systems dominated by submersed, floating-leaved, and
emergent herbaceous vegetation. Table 4 provides an acreage summary by type for each of the
wetland communities in Reaches 5 and 6 based on natural community characterization in the
16
-------
Ecological Characterization Report (Woodlot Alternatives, 2002). For the purpose of this
analysis, the acreage of oxbow wetlands, which are present as part of several communities, albeit
at a relatively small scale, were separated from the community types where they occur and
treated as a separate calculation. The sizes of oxbow wetlands were subsequently calculated by
natural community type (Table 5).
Table 4 Acreage Calculations for the Wetland Communities Within Reaches 5 and 6
Wetland Natural Community Type1
Acreage
% Wetlands
Black ash-red maple-tamarack calcareous seepage swamp
79.0
7.4
Deep emergent marsh
28.5
2.7
High-gradient stream
0.1
<0.1
High-terrace floodplain forest
10.7
1.0
Low gradient stream
250.2
23.6
Medium-gradient stream
8.4
0.8
Moderately alkaline lake/pond
22.1
2.1
Red maple swamp
102.7
9.7
Rich mesic forest
4.5
0.4
Riverine point bar and beach
1.0
<0.1
Shallow emergent marsh
59.1
5.6
Shrub swamp
157.8
14.7
Transitional floodplain forest
193.8
18.3
Wet meadow
41.6
3.9
Oxbows (present in multiple wetland communities)2
102.0
9.63
Total
1061.5
100%
Notes:
1. Wetland natural community classification based on Swain and Kearsley (2000) as characterized in the Ecological
Characterization. Oxbow wetlands are not a distinct community
2. Oxbow wetlands occur within multiple wetland communities, including upland fringes and are not a distinct wetland
community recognized in Swain and Kearsley (2000).
3. Small inclusions of upland natural communities were mapped along the margin of some oxbow wetlands and are
included in the overall area calculation, thus are represented in the percentage.
With the exception of Woods Pond, most of the river in Pittsfield, Lenox, and Lee is classified as
a low-gradient stream (approximately 250.2 acres). A short section of Reach 5 A (approximately
8.4 acres) and sections of the river downstream of the Woods Pond impoundment are considered
medium-gradient streams. High-gradient streams (approximately 0.1 acre) flow off the west
slope of October Mountain and enter Reach 5 as they cross Woodland Road near Woods Pond.
Deep emergent marshes (approximately 28.5 acres), which are usually inundated through the
growing season and vegetated by robust herbs, are frequent along the river channel and
backwater edges of the floodplain. Shallow emergent marshes (approximately 59.1 acres),
17
-------
which are areas with saturated soil or shallow water and lower herbs, are less frequent in the
floodplain and most commonly observed within the vernal pools with longer hydroperiods.
Riverine point bars and beaches (approximately 1.0 acre) occur occasionally along the
Housatonic River, primarily near bends in the channel. Mud flats of limited size begin to appear
typically later in the season as the water levels decline and expose previously flooded substrate.
Several large wet meadows (approximately 41.6 acres) can be found in the floodplain; in these
areas the species composition is often influenced by past farming practices. Woods Pond, a
relatively shallow impoundment, and some of the larger backwater areas to the immediate north
are considered to be a moderately alkaline lake/pond habitat (approximately 22.1 acres).
Table 5 Sizes of Oxbow Wetlands by Natural Community Type Within Reach 5 and 6
Oxbow Wetland Community Type1
Acreage
Agricultural field
0.9
Black ash-red maple-tamarack calcareous seepage swamp
0.4
Cultural grasslands
0.7
Deep emergent marsh
12.4
High-terrace floodplain forest
0.1
Low-gradient stream
17.0
Moderately alkaline lake/pond
9.3
Northern hardwoods-hemlock-white pine forest
1.2
Red maple swamp
1.1
Red oak-sugar maple transition forest
0.8
Riverine point bar and beach
0.3
Shallow emergent marsh
14.1
Shrub swamp
17.9
Transitional floodplain forest
21.3
Wet meadow
4.5
Total
102.0
Notes:
1. Small inclusions of upland natural communities were mapped along the margin of some oxbow
wetlands and are included in the overall area calculation.
Within the floodplain, the structure of the palustrine communities is heavily influenced by
wetland hydrology and river flooding (Woodlot Alternatives, 2002). Most of the existing
landscape is forested, except where disturbance (i.e., forest clearing) or permanent flooding (i.e.,
river channel and backwater slough) prohibit tree growth. The forests can be categorized
generally as one of two types—those areas that receive groundwater discharge and those that do
not. Most of the floodplain forests do not receive groundwater discharge and are largely
classified as transitional floodplain forests (approximately 193.8 acres). These forests are within
the riparian corridor of the river and are subject to inundation during spring flooding and other
high water events. Vernal pool habitat and oxbows are present throughout the transitional
floodplain forest community. Vernal pools are relatively common in small seasonally inundated
18
-------
Ashless depressions, whereas oxbow wetlands are created by chute or neck cutoffs, as described
in Section 2.2.
In a few locations, the floodplain forests are situated on elevated berms and are referred to as
high-terrace floodplain forests (approximately 10.7 acres). In the lower portion of the
floodplain, the floodplain forests give way to black ash-red maple-tamarack calcareous seepage
swamps (approximately 78.9 acres). These forested communities are low-lying wetlands that are
enriched by high-pH groundwater discharge and the backwater effect of Woods Pond Dam.
Red maple swamps (approximately 102.7 acres), another type of forested wetland in the
floodplain, are primarily found in the transition between the floodplain forests and calcareous
seepage swamps. Shrub swamps (approximately 157.7 acres) are common along pool and river
channel borders, but they are especially frequent as an intermediate successional stage in areas
where pasture is reverting to forested floodplain.
Oxbow wetlands (approximately 102 acres) within the floodplain represent approximately 10
percent of the wetland mosaic. These oxbow wetlands differ spatially and temporally due to the
degree of succession of the associated plant community since the date of the initial oxbow
creation. Oxbow wetlands are most common in transitional floodplain forest (approximately
21.3 acres), shrub swamp (approximately 17.9 acres), and low-gradient stream (approximately
17.0 acres) communities (Table 5); however, they represent a relatively small area within these
communities. Oxbow wetlands are additionally represented within the shallow emergent marsh
(approximately 14.0 acres), deep emergent marsh (approximately 12.3 acres), moderately
alkaline lake/pond (approximately 9.2 acres), and wet meadow (approximately 4.4 acres)
communities. Oxbow wetlands represent approximately one percent or less of several additional
communities; most likely along the edge of these communities.
Very little terrestrial or upland habitat is found in within the 10-year floodplain. Red oak-sugar
maple transition forests (approximately 16.3 acres) are located in a few widely scattered
locations. Cultural grasslands (approximately 54.3 acres), which are open, upland habitats
periodically disturbed by mowing or grazing, occur near New Lenox Road. A few upland
inclusions of northern hardwoods-hemlock-white pine forest (approximately 60.0 acres) also
occur north of Yokun Brook. Most of the upland habitats occur adjacent to the floodplain as
cultural grassland, northern hardwoods-hemlock-white pine forest, and rich mesic forest
(approximately 4.9 acres).
7.2 OVERVIEW OF WETLAND FUNCTION AND VALUES IN REACH 5 AND 6
A wetland function-value assessment was performed for the PSA using the U.S. Army Corps of
Engineers 1995 Highway Methodology for Wetland Function-Value Evaluations manual
(TechLaw, 1998). Due to the underlying marble in the Housatonic River Valley, many of the
wetlands in the valley provide high-level functions and values.
Upper Section of the Floodplain
19
-------
The section from the Confluence to the farm fields and wet meadows south of New Lenox Road
is generally characterized as suburban or agricultural with some floodplain wetlands (primarily
Reaches 5A and 5B). The floodplain here is narrower than the lower half of the area and has
undergone a higher degree of floodplain development. The surrounding land contains a
combination of suburban development, agriculture, the Pittsfield Wastewater Treatment Plant,
and conservation land (the Audubon Society's Canoe Meadows Wildlife Sanctuary and sections
of the Commonwealth of Massachusetts' George L. Darey Wildlife Management Area).
The combination of land use changes over the years and a dynamic floodplain surficial geology
creates a mosaic of wetland types that include open water river, palustrine open water, wet
meadow, and emergent, scrub-shrub and forested wetlands. Features such as seasonal pools and
abandoned sloughs or oxbows contain open water with unconsolidated substrates, as well as
emergent vegetation, shrubs, and trees. Few macrophyte beds are found in this section due to
lack of suitable still or slow-moving, open-water habitat.
There is dense vegetation along most of the wetland-river edges. The steeper stream banks have
trees and shrubs rooted into the banks, while more low-angle bank areas tend to be densely
vegetated with floodwater-resilient herbs, as well as shrubs and trees. Most of the agricultural
land that slopes down to the floodplain has either been abandoned or has some naturally
vegetated buffer to protect it from soil erosion. Erosion due to the migration of meanders is
apparent in several areas of this section; but the narrower floodplain width, topography, and
development are limiting factors for meandering processes. Vegetation may thwart this type of
erosion and may influence the exact location of stream bank changes, but only temporarily as the
river attempts to move toward recovery and re-establishment of equilibrium following from
severe anthropogenic influences.
In addition to the abandoned channels, oxbows, and backwaters within the river, there is
evidence of recent channel overwash, erosion of cut banks, and accretion and erosion of point
bars. The floodplain wetlands also have a micro-topography of alluvial rills, mounds, and small
plateaus created by historic and recent flood events.
The principal functions and values provided by wetlands from the Confluence to the farm fields
and wet meadows on the south side of New Lenox Road include Floodflow Alteration, Fish and
Shellfish Habitat, Sediment/Toxicant Retention, Production Export, Sediment/Shoreline
Stabilization, Wildlife Habitat, Recreation, Educational/Scientific Value, Uniqueness/Heritage,
Visual Quality/Aesthetics, and Endangered Species Habitat. The wetlands herein provide 11 of
the 13 functions and values, and each is present at relatively high levels.
Lower Section of the PSA
This section extends from the farm fields and wet meadows south of New Lenox Road to Woods
Pond Dam (Reaches 5B (in part), 5C and 6). Vegetation cover types surrounding this area
include primarily forest land. A railroad right-of-way (ROW) runs along portions of the western
edge of the area, but several tributary wetlands west of the railroad are included. There is also an
20
-------
abandoned trolley line that bisects the southwestern tributary wetlands from the floodplain.
Much of the flashiness exhibited in upstream sections of the river has been dissipated by
upstream floodplain wetlands. River bed sediments in this section primarily include silt and fine
organic particles.
Most of the floodplain wetlands are within one foot of the typical Spring water surface elevation
of the river. This section has the most complex micro-topography in the floodplain. From the
fields at New Lenox Road, south to the backwaters of Woods Pond, there are abandoned oxbows
interlaced with floodplain pools and backwater sloughs. Based on aerial photography
interpretation and field observations, there appears to have been historical logging and ditching
in these wetlands. Discrete zones of forest, shrub, and emergent vegetation are indicative of past
flooding patterns and meander pathways. The Woods Pond impoundment has created
backwaters upstream and laterally at this location, negating natural meandering processes in the
inundated areas in contrast to the upper section.
Five major backwater ponds, and over a dozen smaller ones, provide still-water habitats for the
development of macrophyte beds. These are all located in the southern half of this section, and
include those connected directly to Woods Pond.
Emergent fringe wetlands are scattered around the edges of the main channel and along
transitional land located between backwaters and the river channel or between meanders where
frequent overwash occurs. These wetlands are primarily dominated by purple loosestrife. Some
cattail-dominated areas surround the larger backwaters, but even these have a significant purple
loosestrife component. The large wetland areas west of the railroad berm, in the southern part of
this section, have a more diverse vegetative community.
The principal functions provided by wetlands located in the south end of farm fields along New
Lenox Road to Wood Pond Dam include Groundwater Interchange, Floodflow Alteration, Fish
and Shellfish Habitat, Sediment/Toxicant Retention, Nutrient Removal, Production Export,
Sediment/Shoreline Stabilization, Wildlife Habitat, Recreation, Education/Scientific Value,
Uniqueness/Heritage, Visual Quality/Aesthetics, and Endangered Species Habitat. Each of the
13 evaluated functions and values were found to be significant.
7.3 OVERVIEW OF MASSACHUSETTS ENDANGERED SPECIES ACT AND
PRIORITY HABITAT IN REACH 5 AND 6 OF REST OF RIVER
The Massachusetts Endangered Species Act (MESA) regulations (321 CMR 10.00), promulgated
by the Division of Fisheries and Wildlife and administered by the Natural Heritage and
Endangered Species Program (NHESP), were designed to implement the MESA statute. Under
MESA, state-listed species are listed as classified as Endangered, Threatened, and Special
Concern. The regulations establish a process for determining whether there will be a "take" of
state-listed species protected under MESA. In reference to animals, take means "to harass, harm,
pursue, hunt, shoot, hound, kill, trap, capture, collect, process, disrupt the nesting, breeding,
21
-------
feeding or migratory activity or attempt to engage in any such conduct, or to assist such
conduct." Disruption of nesting, breeding, feeding or migratory activity may result from, but is
not limited to, the modification, degradation or destruction of Habitat. In reference to plants, take
means "to collect, pick, kill, transplant, cut or process or attempt to engage or to assist in any
such conduct."
The NHESP has mapped approximately 1,367.5 acres of Priority Habitatl for 25 identified state-
listed rare species in Reaches 5 and 6, which are protected from a "take" under MESA.
Approximately 98.4 % of the Rest of River is mapped as Priority Habitat. This Priority Habitat
is an amalgamation of smaller Priority Habitats mapped for individual state-listed species.
Approximately 547.1 acres of Priority Habitat is located within the upper section of the area
(from the Confluence to New Lenox Road in Reach 5B); while 820.4 acres is mapped in the
lower section (from New Lenox Road to Woods Pond Dam).
Table 6 provides a summary of individual species' Priority Habitat mapped by the NHESP. Note
that individual species' mapping overlaps due to shared habitat use patterns. Per the definition of
Priority Habitat, the NHESP uses "best scientific evidence available" and includes "necessary
supporting habitat." As a result, Priority Habitat mapping can overestimate actual habitat used
by state-listed species because the extent is often not verified on the ground and inferred from a
variety of available sources of information. In addition, some state-listed species are lesser
studied and exact management requirements are not fully understood.
On February 3, 2011, the NHESP proposed changes to the MESA list that may change the
number of state-listed species about which the NHESP is concerned during remediation activities
(NHESP 2011b). Triangle floater (Alasmidonta undulata), arrow clubtail {Stylurus spiniceps),
and zebra clubtail {Stylurus scudderi) are proposed for delisting. Spine-crowned clubtail
('Gomphus abbreviatus) is proposed to be downgraded from Endangered to Special Concern,
crooked-stem aster (Symphyotrichum prenanthoides) downgraded from Threatened to Special
Concern, and rapids clubtail (Gomphus quadricolor) upgraded from Threatened to Endangered.
The proposed addition of new species to the MESA list is not anticipated. If the triangle floater,
arrow clubtail, and zebra clubtail are delisted in the future, the Priority Habitat will not change
measurably since their mapping largely overlaps with other state-listed species.
1 Priority Habitat is defined as the geographic extent of Habitat for state-listed species as delineated by the Division
pursuant to the MESA regulations (321 CMR 10.12). Priority Habitat is delineated based on confirmed observations
of state-listed species within the last 25 years and the best scientific evidence available, defined as "species
occurrence records, population estimates, habitat descriptions, assessments, peer-reviewed scientific literature,
documented consultation with experts and information contained in records from the NHESP or other credible
scientific reports or species sighting information reasonably available to the Director." Priority Habitat mapping is
based on examination of individual species occurrence records in the context of the following criteria: "the nature
and/or significance of the occurrence as it relates to the conservation and protection of the species, including but not
limited to evidence of breeding, persistence, life stages present, number of individuals, extent of necessary
supporting habitat, and proximity to other occurrences." Priority Habitat mapping is updated on a four-year cycle.
22
-------
Table 6 Priority Habitat Acreage Calculations for the 25 State-Listed Species, as
Identified by the NHESP, Within the PSA
Scientific Name
Common Name
PSA Priority Habitat12 3
PSA
Reach(es)
Upper4
Lower
Total
Sagittaria cuneata
Wapato
172.2
209.8
382.0
5A, 5B, 5C, 6
Carex tuckermanii
Tuckerman's sedge
0.9
0.0
0.9
5A
Quercus macrocarpa
Bur oak
0.0
369.2
369.2
5B, 5C, 6
Elymus villosus
Hairy wild rye
19.3
0.0
19.3
5A
Ranunculus pensylvanicus
Bristly buttercup
29.4
0.8
30.1
5A, 5C
Carex alopecoidea
Foxtail sedge
5.6
68.8
74.3
5B, 5C
Carex grayi
Gray's sedge
0.0
143.9
143.9
5C, 6
Symphyotrich um
prenanthoides
Crooked-stem aster
0.0
12.6
12.6
5B, 5C
Claytonia virginica
Narrow-leaved spring
beauty
17.6
1.7
19.2
5B, 5C
Eleocharis intermedia
Intermediate spike-sedge
158.7
109.5
268.3
5A, 5B, 5C
Pieris oleracea
Mustard white
410.7
699.2
1109.9
5A, 5B, 5C, 6
Papaipema sp. 2 nr. pterisii
Ostrich fern borer moth
178.0
0.0
178.0
5A
Gomphus quadricolor
Rapids clubtail
77.8
95.2
173.0
5B, 5C
Ophiogomphus carolus
Riffle snaketail
106.0
0.0
106.0
5A
Gomphus abbreviatus
Spine-crowned clubtail
250.0
0.0
250.0
5A, 5B
Stylurus spiniceps
Arrow clubtail
328.9
404.7
733.6
5A, 5B, 5C, 6
Ophiogomphus aspersus
Brook snaketail
152.8
0.0
152.8
5A
Stylurus scudderi
Zebra clubtail
327.8
396.8
724.6
5A, 5B, 5C, 6
Alasmidonta undulata
Triangle floater
19.4
0.0
19.4
5A
Glyptemys insculpta
Wood turtle
491.6
338.4
830.0
5A, 5B, 5C
Ambystoma jeffersonianum
Jefferson salamander
0.0
78.1
78.1
5B, 5C
Sorex palustris
Water shrew
0.0
38.9
38.9
5C
Botaurus lentiginosus
American bittern
162.6
202.4
365.0
5A, 5B, 5C
Gallinula chloropus
Common moorhen
16.6
390.6
407.1
5C, 6
Haliaeetus leucocephalus
Bald eagle
0.0
186.5
186.5
5C
Notes:
1. Individual species' Priority Habitat mapping data was provided through a data sharing agreement with NHESP
dated February 24, 2010.
2. The current Priority Habitat mapping will remain effective through December 31, 2011, until the 14th Edition of
the Natural Heritage Atlas (Atlas) is published. Changes to existing Priority Habitat in the PSA are possible
based on new information available to the NHESP since the last Atlas.
3. On February 3, 2011, the NHESP proposed changes to the MESA list, in part, including delisting triangle floater,
arrow clubtail, and zebra clubtail.
4. The PSA was divided into upper and lower sections at New Lenox Road (southern portion of Reach 5B).
23
-------
The Priority Habitat is an amalgamation of individual habitats for multiple state-listed species,
occurring at various spatial and temporal scales. Priority Habitats mapped for highly mobile
animal species such as the wood turtle (Glyptemys insculpta), two dragonfly species, arrow
clubtail, zebra clubtail, and mustard white (Pieris oleracea) encompass a large percentage of the
PSA based on habitat usage patterns, whereas Priority Habitat for many state-listed plant species
occurs in specific reaches at smaller spatial scales as localized occurrences (Stantec, 2010).
Triangle floater, narrow-leaved spring beauty (Claytonia virginica), crooked-stem aster
(Symphyotrichum prenanthoides), and hairy wild rye (Elymus villosus) each have Priority
Habitat areas less than 20 acres within the PSA, while Tuckerman's sedge (Carex tuckermanii)
has the lowest, with approximately 1 acre.
The existing railroad right-of-way (ROW) running parallel with and west of the Housatonic
River within Reaches 5C and 6 has altered wetland hydrology within Reaches 5C and 6 and to
some extent Reach 5B. The past disturbance resulting from ROW construction has influenced
hydrology on both sides, and occasionally the ROW is used as the boundary of Priority Habitat
mapping for several species. Species such as crooked stem aster and foxtail sedge {Carex
alopecoidea), which are sedentary, have a Priority Habitat boundary terminating along the
eastern limit of the ROW. Several more mobile species, including mustard white, bald eagle
(Haliaeetus leucocephalus), Jefferson salamander (Ambystoma jeffersonianum), and wood turtle,
have Priority Habitat mapped on both sides. Two of the five areas mapped as Priority Habitat for
American bittern (Botaurus lentiginosus) and the largest of the three Priority Habitat areas for
common moorhen (Gallinula chloropus) occur west of the ROW.
Conceivably, portions of, or most of these individual species' Priority Habitat areas, have been
influenced and are likely maintained by the ROW. The habitats on the west side of the ROW are
likely hydrologically disconnected from the river, and may have been created in part during
excavation associated with ROW construction.
In addition, Priority Habitat mapped within the floodplain for several species extends beyond the
10-year floodplain defined as Rest of River [e.g., ostrich fern borer moth (Papaipema sp. 2 nr.
pterisii), mustard white, hairy wild rye, rapids clubtail (Gomphus abbreviates), Jefferson
salamander, and American bittern] and is unlikely to be directly altered by remediation activities.
In the case of foxtail sedge, approximately 45 percent of the Priority Habitat occurs outside the
Rest of River.
While 25 identified state-listed species are found in the Rest of River, many of these species
occur in other locations in the state. Actual state-listed species occurrence data in the NHESP
database is exempt from disclosure as a public record (Massachusetts General Law c.66 S.17D of
the Public Record Law), so state-listed species distribution ranges have been inferred from
general occurrence descriptions by county in the publically available factsheets maintained by
the NHESP (NHESP, 2011a) and prior General Electric (GE) reports (ARCADIS, Anchor QEA,
and AECOM, 2010). Per the data publically available, two of the 25 state-listed species (1
24
-------
butterfly and 1 plant) have known occurrences restricted to Berkshire County [mustard white and
bur oak (Quercus macrocarpa)\. Three of the 25 state-listed species [3 plants: wapato
(Sagittaria cuneata), foxtail sedge, and crooked-stem aster] have occurrences restricted to
Berkshire and one additional western county (i.e., Hampshire, Franklin, or Hampden). One moth
[ostrich fern-borer], two plants [Gray's sedge (Carex grayi) and intermediate spike-sedge
(Eleocharis intermedia)], and two dragonflies [riffle snaketail (Ophiogomphus carolus) and
rapids clubtail] are considered to have distribution in western Massachusetts (three or more
western counties). The remaining 15 species have occurrences in western and central
Massachusetts (3 plants, 2 dragonflies, 1 amphibian) or have a scattered statewide distribution (1
plant, 2 dragonflies, 1 mussel, and 1 reptile, 1 mammal and 3 birds). Overall, the anticipated
number of individuals for each mapped occurrence is probably relatively low and may include a
single or a few individuals. Plant occurrences are likely to contain more than a single individual.
7.4 STATE-LISTED SPECIES HABITAT USE IN REACH 5 AND 6
The 25 identified state-listed plant and animal species in the Rest of River are highly dependent
upon and require habitat conditions within the current wetland mosaic in order to complete their
life cycle (Table 7). Many are capable of inhabiting varying types of floodplain wetlands, river
banks, pond shores and/or wetlands influenced by impoundment of Woods Pond. The river
meandering process may create oxbow wetlands, but with similar wetland function and values to
other habitat types. These other similar habitat conditions for many species are present within
the wetland mosaic. None of the state-listed species are dependent upon or restricted to oxbow
wetlands but utilize these habitats as part of their overall habitat use patterns within the wetland
mosaic. Species occurring in backwater areas further from the river channel and/or lower section
of the floodplain influenced by Woods Pond Dam are not dependent on the meandering process
as it is no longer a dominant process under current conditions with the dam. As described in
Section 7.1, the Rest of River provides a wetland mosaic with variable wetland communities, of
which oxbow wetlands represent approximately 10 % of the habitat. Bank habitat conditions in
the Rest of River are variable as result of seasonal overbank flooding and bank erosion due to
river meandering. Many state-listed species are not strictly limited to bank habitats in the Rest of
River, but species will utilize these areas as part of their overall habitat use patterns or banks are
within the range of suitable habitat types. For example, wood turtles will shelter and may
overwinter in the river under overhanging banks, whereas several state-listed plants may inhabit
disturbed banks. In the case of the water shrew, its principal habitat use occurs along banks
close to the river in subterranean burrows. In addition to wetland habitat use patterns, several of
the more mobile animal species also require upland habitat as part of their life history strategy
(e.g., adult dragonflies and wood turtles).
For the 10 state-listed plant species identified in the floodplain, all are affiliated with riverine
systems that are prone to natural disturbance (e.g., seasonal flooding or low water conditions)
and/or open areas/edges within floodplain wetlands. In the case of wapato (Sagittaria cuneata),
which has the largest Priority Habitat for a plant species within the PSA, occurrences of this
25
-------
perennial may be emergent, floating, or fully submerged when present in floodplain wetlands,
demonstrating its tolerance of a hydrological gradient. Intermediate spike-sedge (Eleocharis
intermedia) has more specific habitat requirements, inhabiting muddy alkaline river banks and
ponds during periods of low water when muddy substrates are exposed. In addition to the
habitats describe above, crooked-stem aster (Symphyotrichum prenanthoides) and bristly
buttercup (Ranunculus pensylvanicus) also inhabit managed wetlands (e.g., utility corridors and
roadsides). Overall, these wetlands and their edges are indicative of early successional
vegetative communities as a result of the natural and/or anthropogenic disturbance regimes.
Such disturbances also create susceptibility to colonization by invasive plant species, which as
noted in Section 7.1 are present throughout and dominate some PSA wetland communities. Rare
species observation forms completed during the Ecological Characterization noted the threat of
invasive plant species to state-listed species occurrences in multiple locations (Woodlot
Alternatives, 2002).
Of the 15 identified state-listed wildlife species, many are wide-ranging and occupy different
seasonal habitats (e.g., breeding and nesting) within the PSA. Water birds migrate seasonally to
complete their life cycle (e.g., returning in spring/summer for breeding) and are not present year
round in the floodplain, while others may travel outside the Rest of River. Dragonfly species
may travel between or be present in adjacent watershed(s)s or up/downstream of Reach 5 and 6.
The wood turtle, Jefferson salamander, and water shrew occur year round in the Rest of River,
undergoing subterranean hibernation during winter months.
As part of the evaluation of habitat use patterns of state-listed species in Reach 5 and 6, it is also
relevant to evaluate available data sources. The Ecological Characterization was conducted by
skilled wildlife biologists and botanists during thousands of field survey hours from 1998-2000
to evaluate ecological resources within the Rest of River, particularly Reach 5 and 6. The 2008-
2009 NHESP rare species surveys completed over thousands of hours of field surveys within the
entire Housatonic River watershed (NHESP, 2010). It is possible that additional occurrences are
represented in the current individual state-listed species' Priority Habitats to supplement those
from the Ecological Characterization and other recent NHESP surveys (NHESP, 2010). As a
note, surveys during the Ecological Characterization for state-listed plant species at historic or
other previously known sites did not document species presence (Woodlot Alternatives, 2002).
Table 7 Habitat Descriptions for the 25 State-Listed Species of Concern, as Identified by
the NHESP, in Reach 5 and 6
Scientific Name
Common
Name
Habitat Description
Sagittaria cuneata
Wapato
Riverine floodplain habitats on muddy substrates along the
shores of rivers, ponds, oxbows, and marshes, preferring
shallow and very slow-moving alkaline waters.
Carex tuckermanii
Tuckerman's
sedge
Deciduous forest swamps, stream borders, pond margins,
oxbows, vernal pools, and wet meadows.
26
-------
Table 7 Habitat Descriptions for the 25 State-Listed Species of Concern, as Identified by
the NHESP, in Reach 5 and 6 (Continued)
Scientific Name
Common
Name
Habitat Description
Quercus
macrocarpa
Bur oak
Forested fens, forested swamps, floodplain forests influenced
by calcareous (alkaline or basic) seepage water, and in mesic to
wet sites in shady areas subject to seasonal flooding
Elymus villosus
Hairy wild rye
Floodplain forests (high terrace floodplain forests in particular),
rich moist thickets, and rocky woodlands. Stream banks,
marshes, and moist woods also provide suitable habitat.
Ranunculus
pensylvanicus
Bristly
buttercup
Colonizes variety of habitats via seed dispersal by water and
wildlife. Suitable habitats include marshes, bogs, moist
clearings, wet woods, stream banks, and ditches under open to
filtered sunlight. Frequently inhabits disturbed river banks and
managed wetland communities in utility corridors
Carex alopecoidea
Foxtail sedge
Floodplain meadows and thickets, generally in alkaline alluvial
soils. Typically found in open swales within floodplain forests.
Carex grayi
Gray's sedge
Preferred habitat is floodplain forest along major rivers where
the floodplain forest is subject to spring flooding, wet deciduous
forests on alluvial soils, swampy woods, calcareous meadows,
and remnant floodplain forests bordered by open pastures.
Symphyotrichum
prenanthoides
Crooked-stem
aster
Occurs in a variety of habitats, including exposed gravel and
cobble substrates, rich alluvial soils in floodplain forests,
thickets, and meadows, riverbanks and streamside seeps,
partially wooded swamps, and roadside habitats under open to
semi-open conditions.
Claytonia virginica
Narrow-leaved
spring beauty
Inhabits rich, damp to moist deciduous woods, thickets,
floodplain forests, and open clearings on alluvial soils seasonally
flooded.
Eleocharis
intermedia
Intermediate
spike-sedge
Typically found on muddy, alkaline river banks and pond shores,
usually during periods of low water that expose muddy shores.
Pieris oleracea
Mustard white
Typically found in understory and along edges of moist, rich,
openings in deciduous woodlands including riparian floodplains.
Nearby open areas including streamsides, shallow marshes, wet
meadows, open fields and pastures also utilized. Two-leaved
toothwort, cuckoo-flower, and other mustard family plants are
essential larval host plants. Adults attracted to garlic mustard,
common winter cress, and field pennycress as potential host
plants.
Papaipema sp. 2
nr. pterisii
Ostrich fern
borer
Primarily associated with mature floodplain forests and wooded
swamps with moderate to dense stands of ostrich fern. Adults
likely to be found in shaded to partially shaded forested
floodplain habitats or red maple swamps containing the larval
host plant species
27
-------
Table 7 Habitat Descriptions for the 25 State-Listed Species of Concern, as Identified by
the NHESP, in Reach 5 and 6 (Continued)
Scientific Name
Common
Name
Habitat Description
Gomphus
quadricolor
Rapids clubtail
Occurs in or near clear, cold streams and rivers that have
intermittent segments of rocks and rapids. Larvae found in
shallow pools (just below sediment surface) located
downstream of rapids that often contain cattail or other
emergent plants. Adults may travel far from waterway to feed,
before returning to mate.
Ophiogomphus
carolus
Riffle snaketail
Larvae prefer sandy substrates (and reside close to surface) in
clear running water, and have a relatively high oxygen
requirement among this family. Upon emergence, flies into
adjacent woodland or shrubland to hide among vegetation and
continue to develop. Adults may live rest of the summer far
from the stream, often in dense woodland or shrubland.
Gomphus
abbreviatus
Spine-crowned
clubtail
Typically in or near medium to large rivers with sandy or rocky
bottoms and silt deposits. Upon emergence, flies into adjacent
woodland to hide in the tree tops. Adult males return to
waterway to feed and mate. Adult males prefer sandy stretches
of the shoreline or overhanging vegetation as perching sites.
Adult females spend a majority of their lives in the forested
areas away from the river, returning for a brief period to mate.
Stylurus spiniceps
Arrow clubtail
Larvae prefer silty to sandy substrates (near surface) in running
water, with a moderate oxygen requirement. Upon emergence,
flies to adjacent woodland to continue developing. After one to
several weeks, adults return to waterway to feed and mate.
Adults may live rest of the summer away from the waterway,
often in dense woodland. Adults believed to spend most of
time in treetops.
Ophiogomphus
aspersus
Brook
snaketail
Larvae prefer sandy substrates (near surface) in clear running
water, with relatively high oxygen requirement. Upon
emergence, flies to adjacent woodland or shrubland to continue
developing. After one to several weeks, adults return to the
waterway to feed and mate. Adults may live rest of the summer
far from the waterway, often in dense woodland or shrubland.
Stylurus scudderi
Zebra clubtail
Larvae prefer silty to sandy substrates (near surface) in running
water, with a moderate oxygen requirement. Upon emergence,
flies into adjacent woodland to hide in the trees and continue to
develop. After one to several weeks, adults return to the
waterway to feed and mate. Adults may live rest of the summer
away from waterway, often in dense woodland.
Alasmidonta
undulata
Triangle floater
Prefers low gradient rivers with flowing water and sand and
gravel substrate; may be found in lake habitats, and can survive
in a wide variety of substrate types. Glochidia attach to multiple
common fish species, where they grow and eventually fall to
develop into adults on the bottom.
28
-------
Table 7 Habitat Descriptions for the 25 State-Listed Species of Concern, as Identified by
the NHESP, in Reach 5 and 6 (Continued)
Scientific Name
Common
Name
Habitat Description
Glyptemys
insculpta
Wood turtle
Requires clear, moving water, such as rivers, streams and
creeks, also utilize variety of shallow wetlands, such as swamps,
bogs, oxbows, and seasonal pools. Use variety of upland
habitats and generally prefer mosaic of communities near
water. Require wide range of habitats for food availability,
thermoregulation, nesting and overwintering. They also use
emergent logs or grassy, sandy, and muddy banks for basking.
Ambystoma
jeffersonianum
Jefferson
salamander
Primarily upland species, prefers well-drained deciduous or
mixed forests in proximity to small shallow vernal pools or
fishless ponds surrounded by vegetation. Adults hide beneath
leaf litter, loose soil, stones, and rotting logs, or in subterranean
burrows. Vernal pool habitat, full of detritus to conceal larvae,
is necessary for reproduction, and submerged woody shrubs or
grasses needed for egg mass attachment.
Sorex palustris
Water shrew
Found near rivers and streams with exposed banks, rocks, and
downed logs along the waterways. Lives on river banks where
moss-lined burrows are hidden between tangles of roots along
undercut banks or boulders. Seldom found more than a few
yards from the nearest water. Prefers forested habitat proximal
to water.
Botaurus
lentiginosus
American
bittern
Inhabits freshwater and brackish wetlands, including marshes,
meadows, bogs, and fens, where occurs in emergent vegetation
like cattails, sedges, and rushes. Occasionally utilizes upland
grasslands for foraging and nesting. Prefers wet meadows for
nesting sites, but known to construct platforms of vegetation a
foot above water or nest in uplands adjacent to wetlands. Also
occasionally nest in upland fields next to water.
Gall inula
chloropus
Common
moorhen
Inhabits large freshwater marshes and ponds with cattails and
other emergent vegetation. Generally takes cover in dense
vegetation and feeds by wading or diving at the edges of open
water. Preferred habitat is waterbodies at least one foot deep,
with dense cattails and occasionally shrub swamps adjacent to
open water with aquatic vegetation bed.
Haliaeetus
leucocephalus
Bald eagle
Inhabits coastal areas, estuaries, and larger inland waters.
Requires high amount of water-to-land edge with forest stands
to nest and trees above the canopy for perching, an adequate
supply of moderate-sized to large fish, an unimpeded view, and
minimal human disturbance.
29
-------
8. SUMMARY
This paper presents the current understanding of stream meandering processes along the
Housatonic River, and the connection between associated habitat and species assemblages. The
findings illustrate that the Housatonic has a complex history of planform change associated with
anthropogenic influences and meandering processes. Meandering processes along the
Housatonic tend to progress slowly in the riverine environment (e.g., 0.9 ft/yr along outside
meander bends) and generate floodplain features (such as cutoffs and oxbows) over an extended
time scale (i.e., centuries). There are 25 state-listed species in the PSA that use habitat
throughout the PSA wetland mosaic.
Additional major points discussed in this paper are summarized below:
1. River meandering is a process by which some floodplain features are created, including
channel cutoffs, oxbow lakes, point bars, and scroll bars, which are utilized by aquatic
and riparian plants and animals.
2. The meandering character and planform pattern of the Housatonic River is not unique to
rivers of the northeastern United States.
3. Anthropogenic activities along the Housatonic River have affected riverine processes,
including meandering, and associated aquatic and terrestrial habitats over the past several
centuries.
4. In Reach 5, channel straightening and other channel manipulations have dramatically
altered the channel and accelerated meander development and chute/cutoff formation
during this time, as the river has attempted to re-equilibrate to these disturbances. Rather
than resulting in permanent change, the recreation of well-developed meanders in less
than a century along much of the straightened river channel indicates the capacity of the
river to develop a new quasi-equilibrium and recover from large-scale perturbations in a
relatively short time frame.
5. Since 1952, meanders in Reach 5 have continued to migrate slowly (with erosion along
the outside bends at a rate no greater than 0.9 ft/yr). It should be noted that the average
erosion rate for the entire channel in Reaches 5A and 5B is on the order of 0.3 ft/yr based
on a study completed in 2009. Since 1945, no new meanders have formed along the
Housatonic River between the Confluence of the East and West Branch and Woods Pond.
6. Assuming similar hydrology and sediment supply and the absence of significant channel
straightening in the future, future meandering rates are likely to be less than those
observed historically.
7. Sedimentation within existing oxbows occurs on a time scale of centuries. Based on
available historical records and assuming a constant rate of oxbow formation and infilling
through time, it is estimated that the existing oxbows would become 70% infilled over
30
-------
the next 500 years with perhaps several hundred more years beyond that necessary to
completely infill individual oxbows and begin to lose associated wetland habitats.
8. Extensive field studies of the channel and floodplain environments along the Housatonic
have characterized the presence and relative abundance of state-listed species. The
existing state-listed species data collected by Woodlot Alternatives (2002) in Reach 5 and
6 and the NHESP (2010) from the watershed provide reliable documentation of actual
habitat and species occurrences that are available for evaluating potential impacts during
remediation of the Housatonic River.
9. The floodplain is a wetland mosaic with variable wetland communities, of which oxbow
wetlands represent approximately 10% of community structure.
10. The state-listed species are not strictly dependent upon or restricted to oxbow wetlands or
other geomorphic features associated with meandering, but utilize these habitats as part
of their overall habitat use patterns within the wetland mosaic.
31
-------
9. LITERATURE CITED
ARCADIS, Anchor QEA, and AECOM, 2010. Housatonic River, Rest of River, Revised
Corrective Measures Study Report. Prepared for General Electric Company, Pittsfield,
MA. October.
Begin, Z.B., 1981. Stream curvature and bank erosion: A model based on the momentum
equation: Journal of Geology, v. 89, p. 497-504.
Bierman, P.R., Howe, J., Stanley-Mann, E., Peabody, M., Hilke, J., and Massey, C.A., 2005.
Old images record landscape change through time: GSA Today, v. 15, p. 1-6.
Bierman, P., Lini A., Zehfuss, P., Church A., Davis, P.T., Southon, J., and Baldwin, L., 1997.
Postglacial ponds and alluvial fans: recorders of Holocene landscape history: GSA
Today, v. 7, p. 1-8.
Brakenridge, G.B., Thomas, P.A., Conkey, L.E., and Schiferle, J.C., 1988. Fluvial sedimentation
in response to postglacial uplift and environmental change, Missisquoi River, Vermont:
Quaternary Research, v. 30, p. 190-203.
Camporeale, C., Perucca, E., and Ridolfi, L., 2008. Significance of cutoff in meandering river
dynamics: Journal of Geophysical Research, v. 113, 11 p.
Constantine, J. A., McLean, S.R., and Dunne, T., 2010. A mechanism of chute cutoff along large
meandering rivers with uniform floodplain topography: Geological Society of America
Bulletin, v. 122, p. 855-869.
Field, J.J., 2007. The recreation of meanders along artificially straightened stream channels in
New England: Geological Society of America Abstracts with Programs, v. 39, p. 243.
Florsheim, J. L., Mount, J.F., and Chin, A., 2008. BioScience, v. 58, no. 6 , p. 19-52.
Garcia, M.H., 2008. Sedimentation engineering: processes, measurements, modeling, and
practice: Environmental and Water Resources Institute, American Society of Civil
Engineers Task Committee to Expand and Update Manual 54, 1132 p.
Harrington Engineering and Construction Inc., 1996. Report on the Preliminary Investigation of
Corrective Measures for Housatonic River and Silver Lake Sediment. Report prepared
for General Electric Company, Pittsfield, Massachusetts.
Henry, C.P., and Amoros, C., 1995. Restoration ecology of riverine wetlands: I. A scientific
base. Environmental Management, v. 19, no. 6, p. 891-902.
Hooke, J.M., 1995. River channel adjustment to meander cutoffs on the River Bollin and River
Dane, northwest England: Geomorphology, v. 14, p. 235-253.
32
-------
Hooke, J.M., 1997. Styles of channel change, in Thorne, C.R., et al., eds., Applied Fluvial
Geomorphology for River Engineering and Management: New York, John Wiley and
Sons, p. 237-268.
Hooke, J., 2003. River meander behavior and instability: a framework of analysis: Transactions
of the Institute of British Geographers, v. 28, p. 238-253.
Hooke, J.M., 2004. Cutoffs galore!: occurrence and causes of multiple cutoffs on a meandering
river: Geomorphology, v. 61, p. 225-238.
Hudson, P.F., and Kesel, R.H., 2000. Channel migration and meander-bend curvature in the
lower Mississippi River prior to major human modification: Geology, v. 28, p. 531-534.
Howard, Alan D., 2009. How to make a meandering river. Proc. Natl. Acad. Sci., USA., v. 106,
no. 41, p. 17245-17246.
Hupp, C. R., and Osterkamp, W. R., 1996. Riparian vegetation and fluvial geomorphic
processes. Geomorphology v. 14, p. 277-295
Knighton, A.D., 1998. Fluvial Forms and Processes: A New Perspective: London, Arnold, 383
P-
Lagasse, P.F., 2004. Handbook for predicting stream meander migration, Issue 553, Part 1:
National Cooperative Highway Research Program, National Research Council (U.S.),
Transportation Research Board, 97 p.
Leopold, L.B., 1994. A View of the River: Harvard University Press: Cambridge, 298 p.
Lortie, J., 2011. Personal Communication September 27, 2011. Stantec Consulting Services,
Topsham, Maine.
Knighton, A.D., 1998. Fluvial Forms and Processes: A New Perspective: London, Arnold, 383
P-
Magdaleno, F., and Fernandez-Yuste, J.A., 2011. Meander dynamics in a changing river
corridor: Geomorphology, v. 130, p. 197-207.
Nanson, G.C., and Hickin, E.J., 1986. A statistical analysis of bank erosion and channel
migration in western Canada: Geological Society of America Bulletin, v. 97, p. 497-
504.
Nanson, G., 1979. Point bar and floodplain formation of the meandering Beatton River,
northeastern British Columbia, Canada. Sedimentology, v. 27, no. 1, p. 3-29.
Natural Heritage and Endangered Species Program (NHESP), 2010. Rare Species and Natural
Community Surveys in the Housatonic River Watershed of Western Massachusetts.
33
-------
Massachusetts Division of Fisheries and Wildlife. Westborough, Massachusetts.
Natural Heritage and Endangered Species Program (NHESP), 2011a. List of Endangered,
Threatened, and Special Concern Species (including links to rare species fact sheets).
Available at
http://www.mass.gov/dfwele/dfw/nhesp/species info/mesa list/mesa list.htm. Website
viewed September 9, 2011.
Natural Heritage and Endangered Species Program (NHESP), 201 lb. Recommended Changes to
the Massachusetts List of Endangered (E), Threatened (T) & Special Concern (SC)
Species. Available at
http://www.mass.gov/dfwele/dfw/nhesp/species info/mesa list/mesa list changes.htm.
Website viewed September 2, 2011.
Ollero, A., 2010. Channel changes and floodplain management in the meandering middle Ebro
River, Spain: Geomorphology, v. 117, p. 247-260.
Petts, G.E., 1994. Rivers: dynamic components of catchment ecosystems, in Calow, P., and
Petts, G.E., eds., River Handbook, Volume 2: Oxford, Blackwell Scientific, p. 3-22.
Pierce, H.A., 2006. A Planform Analysis of Channel Change on the Housatonic River in
Southwestern Massachusetts: University of Connecticut, Master's Thesis, 112 p.
Schumm, S.A., 1979. Geomorphic thresholds: the concept and its applications: Transactions of
the Institute of British Geographers, v. 4, p. 485-515.
Schumm, S.A., and Khan, H.R., 1972. Experimental study of channel patterns: Geological
Society of America Bulletin, v. 83, p. 1755-1770.
Stantec Consulting Services Inc. (Stantec), 2009. Housatonic River BEHI Report.
Stantec Consulting Services (Stantec), 2010. State-Listed Rare Species MESA Assessment for
Remedial Alternatives Housatonic River - Rest of River. Topsham, Maine.
Stolum, H.H., 1996. River meandering as a self-organization process: Science, v. 271, p. 1710-
1713.
Swain, P. and Kearsley, J., 2000. Classification of the Natural Communities of Massachusetts.
Massachusetts Natural Heritage and Endangered Species Program, Westborough, MA,
USA.
TechLaw, Inc., 1998. Preliminary Report: Wetland Characterization and Function-Value
Assessment, Housatonic River from Newell Street to Woods Pond. TechLaw, Inc.,
Boston, MA, USA.
Thompson, D.M., 2003. A geomorphic explanation for a meander cutoff following channel
34
-------
relocation of a coarse-bedded river: Environmental Management, v. 31, p. 385-400.
Walter, R.C., and Merritts, D.J., 2008. Natural streams and the legacy of water-powered mills:
Science, v. 319, p. 299-304.
Ward, J.V., K. Tockner, D. B. Arscott and Claret, C., 2002. Riverine landscape diversity.
Freshwater Biology, v. 47, p. 517-539.
WESTON (Weston Solutions, Inc), 2006. Final Model Document Report: Modeling Study of
PCB Contamination in the Housatonic River, Appendix A.4: Erosion and Accretion
Mapping Meander Study. Prepared for the U.S. Army Corps of Engineers and the U.S.
Environmental Protection Agency. DCN GE-111006-ADII.
WESTON (Weston Solutions, Inc), 2011. National Remedy Review Board Site Information
Package for the Housatonic River, Rest of River, Appendix A -Historical Changes in
Housatonic River Morphology: SDMS 487318, June 2011. Privileged, Attorney-Client
Work Product, Deliberative Document, Do Not Release.
Wolman, M.G., 1955. The natural channel of Brandywine Creek, Pennsylvania: U.S. Geological
Survey Professional Paper 271, 56 p.
Wolman, M.G., and Leopold, L.B., 1957. River flood plains: some observations on their
formation: U.S. Geological Survey Professional Paper 282-C, 30 p.
Woodlot Alternatives, Inc., 2002. Ecological Characterization of the Housatonic River. Report
prepared for U.S. Environmental Protection Agency Region 1, Boston, Massachusetts.
Xu, D., Bai, Y., Ma, J., and Tan, Y., 2011. Numerical investigation of long-term planform
dynamics and stability of river meandering on fluvial floodplains: Geomorphology, v.
132, p. 195-207.
Web citations
Web citation 1: http://docs.unh.edu/MA/beck93nw.ipg
Web citation 2: http://docs.unh.edu/MA/pite47sw.ipg
Web citation 3: http://docs.unh.edu/MA/estl48nw.ipg
35
-------
APPENDIX A
Location Maps and Meandering Characteristics of Selected Rivers in the
Northeastern United States
36
-------
6=4
0 Miles 50
37
-------
Sawmill River (Montague, MA)
Drainage area: 32 sq mi
Sinuosity: 1.83
Channel width: 40 ft
Meander wavelength: 300 ft
Radius of curvature: 100 ft
Meander amplitude: 150 ft
Length of oxbows / valley mile: 1,320 ft
Corridor width: 550 ft
r—i—
Note: Red box highlights area of detail shown in aerial photograph
38
-------
Fort River (Hadley, MA)
Drainage area: 56 sq mi
Sinuosity: 2.13
Channel width: 50 ft
Meander wavelength: 500 ft
Radius of curvature: 120 ft
Meander amplitude: 200 ft
Length of oxbows / valley mile: 2,560 ft
Corridor width: 750 ft
0 600 1,200 1,800
Note: Red box highlights area of detail shown in aerial photograph
39
-------
Baker River (Rumney,NH)
Drainage area: 143sqmi Radius of curvature: 270 ft
Sinuosity: 1.54 Meander amplitude: 400 ft
Channel width: 100 ft Length of oxbows / valley mile: 4,197 ft
Meander wavelength: 900 ft Corridor width: 1,500 ft
Runint;y Dejx
Note: Red box highlights area of detail shown in aerial photograph
40
-------
cr
LO
00
<3-
vD
rv
on
tu
o
&
s ™ -
V rsi 2
ra .. 5
QJ ^ _
cn ±i cu
ra J5 c
•- § c
ra c ^
ir .E -C
O 1/1 U
O ,£
° g
cn 2!
C =
CD ji
(U 5
> c
> 3
¦> U
aj o
"O u"i
C =3
Oj
ra
CC
rC ^
^ = 4±
a ^ 9
"O un
2 5 ^
D- -§ "£
£ g 2
Z ° 5
53 ° o
£ -a
« ?"E
U c r>
5 ii °
/'¦ ¦¦* >\
41
-------
Batten Kill (Arlington, VT)
Drainage area: 149sqmi Radius of curvature: 130 ft
Sinuosity: 1.36 Meander amplitude: 225 ft
Channel width: 70 ft Length of oxbows / valley mile: 3,210 ft
Meander wavelength: 750 ft Corridor width: 700 ft
Note: Red box highlights area of detail shown in aerial photograph
42
-------
Poultney River (Fair Haveri, VT)
Drainage area: 175sqmi
Sinuosity: 2.00
Channel width: 60 ft
Meander wavelength: 400 ft
i
Radius of curvature: 11 Oft
Meander amplitude: 175 ft
Length of oxbows / valley mile: 7,329 ft
Corridor width: 1,000 ft
xi / /i I.--.
Note: Red box highlights area of detail shown in aerial photograph
43
-------
Contoocook River (Peering, NH)
Note: Red box highlights area of detail shown in aerial photograph
Drainage area: 221 sq mi
Sinuosity: 1.74
Channel width: 80 ft
Meander wavelength: 650 ft
Radius of curvature: 160 ft
Meander amplitude: 375 ft
Length of oxbows / valley mile: 5,659 ft
Corridor width: 1,600 ft
-------
Kinderhook Creek (Kinderhook, NY)
Drainage area: 316 sq mi
Sinuosity: 1.45
Channel width: 140 ft
Meander wavelength: 1000 ft
Radius of curvature: 290 ft
Meander amplitude: 375 ft
Length of oxbows / valley mile: 6,102 ft
Corridor width: 2,100 ft
Note: Red box highlights area of detail shown in aerial photograph
45
-------
Saco River (Fryeburg, ME)
Note: Red box highlights area of detail shown in aerial photograph
Drainage area: 444 sq mi
Sinuosity: 2.25
Channel width: 300 ft
Meander wavelength: 4,000 ft
Radius of curvature: 900 ft
Meander amplitude: 2,750 ft
Length of oxbows / valley mile: 8,865 ft
Corridor width: 6,500 ft
-------
APPENDIX B
Location of All Oxbows Between the East and West Branch Confluence and Woods Pond
47
-------
-------
-------
-------
-------
ATTACHMENT B-3 ACTIVATED CARBON SUMMARY
-------
The Isosceles Group technical memorandum
DATE: July 27, 2012
TO: Scott Campbell, Weston Solutions, Inc.
FROM: Dick McGrath
SUBJECT: Review of recent studies of activated carbon amendment of contaminated
sediment, with reference to bioavailability, ecotoxicology, and engineering
considerations for activated carbon application
Introduction
One important aspect of the well-established equilibrium-partitioning behavior of non-polar organic
contaminants in sediment is the relationship between the organic carbon content of the sediment and
bioavailability of the contaminant. Bioavailability, in turn, is a major factor controlling the toxicity of the
sediment or, conversely, the maximum concentration of the contaminant that can be tolerated by
exposed biota without adverse impact. In general, increasing the sediment organic carbon content
decreases contaminant bioavailability, which has the practical application for management of
contaminated sediment sites of allowing higher concentrations of contaminants to remain in situ
without adverse biological effects. Such an approach may have significant advantages at sites where
active remediation of contaminated sediment via more traditional removal techniques such as dredging
would result in undesirable harm to the habitat.
Project managers and investigators have become increasingly interested over the last several years in
amending natural sediments with activated carbon (AC) to increase the total organic carbon content,
and also with technologies to apply AC directly to sediments or to incorporate AC into sediment caps
that can be used either in conjunction with, or as a substitute for, sediment removal. A growing body of
research on AC amendment of contaminated sediments, both laboratory studies and in situ, has been
conducted to demonstrate the applicability of this approach. The majority of these studies address
three important questions regarding the applicability of AC amendment to contaminated sediment sites:
1. Reduction in bioavailability - Although reduction in contaminant bioavailability has been amply
demonstrated in concept through research studying equilibrium-partitioning behavior, will
similar effects be seen in real-world situations with the elevated levels of contaminants typical
of hazardous waste sites?
2. Potential toxicity of AC - does the amendment of contaminated sediments with AC in itself
present a risk to plants or animals inhabiting the sediment?
3. Engineering considerations - can AC be effectively incorporated into sediments and sediment
cap designs and can AC or AC-amended caps be placed at contaminated sediment sites in a
manner that meets the long-term project objectives?
The Isosceles Group, Inc. Page 1 of 7
A Sleeman Hanley & DiNitto Company
50 Congress Street, Suite 840
Boston, Massachusetts 02109
www.theisogroup.com
-------
In February of 2012, U.S. EPA Region 10 (Seattle) sponsored a Technical Workshop entitled "Use of
Activated Carbon Amendment as an In-situ Sediment Remedy at the Lower Duwamish Waterway
Superfund Site." The Workshop brought EPA regulators and site managers together with many of the
leading investigators in the field and, although focused on the Lower Duwamish Waterway, the
Workshop presentations covered a wide range of other contaminated sediment sites around the world.
Accordingly, the proceedings of the Workshop represent a review of the current state-of-the-art with
respect to AC amendment at contaminated sediment sites and provide direct insight into the three
questions referenced above. This memorandum is intended to address each of the questions by
summarizing material presented at the Workshop as well as other studies from the scientific literature.
Reduction in Bioavailability
As a result of extensive research conducted over many years, and the now widespread acceptance of
equilibrium-partitioning theory to explain the behavior and toxicity of many organic contaminants in
aquatic systems, there is little question that the organic carbon fraction of sediment is a major
determinant of bioavailability and that, for the range of organic fractions commonly seen in the
environment, bioavailability is strongly inversely correlated with the total organic content of the
sediment. Several of the presentations at the Workshop addressed, either directly or indirectly, the
question of whether the well-understood concepts of equilibrium partitioning behavior are applicable to
natural sediment and/or sediment caps that have been artificially amended with AC.
There was a strong consensus from the presentations at the Workshop that AC amendment is effective
in reducing bioavailability of contaminants in sediment and that the reduced bioavailability remains in
effect for several years. Ghosh (2012a) reviewed a number of studies that clearly demonstrated large
reductions in bioavailability of organic contaminants with AC amendment in the range of 5% by weight,
and also that AC amendment reduced both contaminant concentrations in pore water and contaminant
flux to overlying water. Greenberg (2012a), reviewing an AC amendment pilot study conducted at the
Grasse River, concluded that AC amendment in the range of 4 to 5% by weight reduced bioavailability of
PCBs by over 95%. Cho et al. (2012a) also reported a decrease in PCB bioavailability with AC
amendment from a pilot study conducted at the Hunters Point site in San Francisco Bay, indicating that
the effectiveness of the AC persisted for at least 5 years. Reible (2012), in summarizing results from a
number of sites, concluded that AC amendment of sediment provides substantial reduction in
contaminant bioavailability and mobility, even as the AC becomes fouled over time. He also found that
incorporating AC into a sediment cap is particularly effective in reducing contaminant exposure and flux.
In related studies, Cornelisson et al. (2006) found that reduction in bioavailability from AC amendment
might be species-specific. In a study that measures biota-sediment bioaccumulation factors (BSAFs) for
PAHs with and without AC amendment, they reported reductions in the range of a factor of six to seven
for a marine polychaete worm, but relatively little change for a marine gastropod. Fagervold et al.
(2010) found that bioavailability of dioxins and furans from floodplain soils was naturally low in
floodplain soil with high natural organic content, but could be reduced still further (up to 91% less) with
AC amendment. For naturally low organic content soils, reductions of over 99% were possible with as
little as 2% AC amendment by weight. Janssen et al. (2010) showed that bioaccumulation of PCBs by a
marine worm was decreased by 95% in laboratory experiments, with no adverse effects on the
organisms.
The Isosceles Group, Inc. Page 2 of 7
A Sleeman Hanley & DiNitto Company
50 Congress Street, Suite 840
Boston, Massachusetts 02109
www, theisogroup.com
-------
McLeod et al. (2004) demonstrated the type of carbon to be important in determining overall
effectiveness, finding that activated carbon was considerably more effective than other types of carbon
in reducing bioavailability of benzo[o]pyrene and PCB-52 to a marine clam. The effectiveness of AC
amendment in reducing bioavailability was confirmed in a second study with the same species of marine
clam (McLeod et al., 2007), and in a similar study using freshwater clams in the Grasse River (McLeod et
al., 2008).
Numerous additional studies (e.g., Millward et al, 2005; Sun and Ghosh, 2007, 2008; Sun et al, 2008;
Tomaszewski et al, 2008) report substantially the same findings for studies conducted both in the
laboratory and in situ, using various organisms, in estuarine and freshwater environments, and with
different organic contaminants. Similar results have also been reported from studies conducted using
passive sampling devices (PSDs) as surrogates for living organisms.
Potential Toxicity of Activated Carbon
Although virtually all sediments contain varying amounts of organic carbon, and in some cases include
carboniferous materials similar to AC, amendment of sediment with up to 5% AC represents the
introduction of a comparatively large amount of a foreign substance into the natural sediment.
Accordingly, for AC amendment to be a viable alternative technology for management of contaminated
sediment sites, it must be demonstrated that addition of activated carbon in the quantities necessary to
achieve a reduction in bioavailability of the contaminant does not itself have a detrimental effect on
resident organisms. The relative lack of adverse impact resulting from AC amendment may be inferred
from the numerous bioavailability studies using living organisms, which make no mention of any
significantly increased mortality observed in the course of the study. Indeed, it would not have been
possible to complete such studies if the adverse effects of AC amendment were substantial. In addition,
there are a smaller number of studies that have been conducted specifically to address the question of
potential toxicity.
Ghosh (2012b) summarized in situ work conducted at the Grasse River site to investigate potential
harmful effects of AC amendment on benthic animals and plants. He reported that benthic
macroinvertebrate community parameters (e.g., numbers, diversity, biomass) were similar between
locations that had received AC amendment and upstream reference locations. It was noted, however,
that decreased plant (Elodea canadensis) growth was correlated with increasing amounts of AC
amendment, with the highest treatment by weight (7.5%) decreasing plant growth by 25%. Subsequent
experiments demonstrated that this decrease in growth appeared due to simple dilution of natural
sediment and the effect decreased over time as the AC aged following application.
Greenberg (2012b), summarizing work conducted by Kupryianchyk et al. (2011), reported that AC
amendment had limited effects on an amphipod and isopod at low concentrations but that AC
amendment could lead to increased mortality in amphipods, perhaps due to sequestration of necessary
nutrients. The study was carried out with sediments that were sufficiently contaminated to cause 100%
mortality in both species without treatment, however, so it was concluded that AC amendment resulted
in substantial benefits that outweighed any deleterious effects. He also examined the Grasse River data
and reached conclusions consistent with those of Ghosh (2012b).
In a study of the effects of AC amendment on the benthic community in San Francisco Bay, Cho et al.
(2012) concluded that the amendment had no significant effect on the benthic community. The study
also examined the secondary effect of the AC on deposit feeders, determining that any effects were
The Isosceles Group, Inc. Page 3 of 7
A Sleeman Hanley & DiNitto Company
50 Congress Street, Suite 840
Boston, Massachusetts 02109
www, theisogroup.com
-------
minor. Menzie (2012) reviewed a range of studies, concluding that any effects of AC amendment on a
range of species were minor. He also examined the potential effect of AC on fish, concluding that AC
would not be toxic and exposure for fish would be low.
In a study focused primarily on powdered, as opposed to granulated, AC, Jonker et al. (2009) found that
powdered AC was directly toxic to several organisms investigated and recommended that such
powdered material be washed out of AC prior to application. They also found that marine crustaceans
avoided AC-amended sediment and that AC in sediment apparently disrupts the feeding behavior of
oligochaete worms. These effects appeared to be most directly associated with powdered AC, and the
authors acknowledged that application of granular AC may not lead to similar issues.
Taken together, these studies tend to indicate that although there is some potential for ecotoxicological
effects resulting from AC amendment of contaminated sediment, such risks can be managed by
controlling the type of AC and the details of its application. The differing results reported for different
target species, AC type, and application method underscore a need for well-designed pilot studies
before widespread use of AC amendment at a particular site.
Engineering Considerations for AC Application
AC amendment has been conducted successfully using a variety of engineering techniques. These range
from simple broadcast application in the field, to mixing with other material which is then carefully
placed on the sediment surface, to incorporation as a layer in engineered caps. Several of the
presentations at the EPA Workshop described the methods that have been used to date and reviewed
their effectiveness.
Carscadden (2012) reviewed the use of AC in the Slip 4 Early Action Site, part of the Lower Duwamish
Estuary. Granulated AC was applied as part of a chemical isolation layer 12" in thickness over the
previously dredged area. Blending of the AC was conducted onshore using standard equipment. The
material was then applied and spread using a typical bucket dredge. Eek et al. (2012) presented the
results of a thin-layer capping application in Norway at locations of 30 m and 100 m depth. For this
work, materials were mixed in a hopper dredge and applied by pumping to the bottom. This study
found the AC layer to be effective but also underscored the importance of adapting methods to local
conditions.
McDonough et al. (2007), in a study conducted in the Anacostia River, demonstrated that it is possible to
incorporate a thin layer of AC into reactive core mat (RCM) as part of a sediment cap which includes an
overlying habitat layer. The RCM/AC layer was effective in sequestering underlying contaminants and
was not disturbed by the subsequent development of a benthic community in the habitat layer.
Melton (2012) reviewed a number of technologies that can be used to apply AC to meet project
objectives at various types of sites, noting that the particular form of AC and also the method used to
apply AC-amended sediment caps can, and should, be tailored to the site to ensure long-term physical
stability. Engineering techniques are available that protect the integrity and function of AC amendment
in both low and high scour areas. A variety of site-specific conditions such as natural waterway
dynamics, vessel traffic, infrastructure, and human activities must be considered in selecting the best
application method but that there are a wide range of available methods that can successfully account
for these factors.
The Isosceles Group, Inc. Page 4 of 7
A Sleeman Hanley & DiNitto Company
50 Congress Street, Suite 840
Boston, Massachusetts 02109
www, theisogroup.com
-------
Overall, the presenters emphasized the need to select the form of AC as well as the application method
to site-specific conditions. Pilot studies, such as those conducted in the Grasse River, are particularly
necessary to ensure that the proposed approach will be effective.
References
Carscadden, R. 2012. Carbon amendment at Slip 4 Early Action Site. Presentation at EPA Region 10
Sponsored Technical Workshop: Use of Activated Carbon Amendment as an In-situ Sediment Remedy
at the Lower Duwamish Waterway Superfund Site. February 14 & 15, 2012.
Cho,Y., Y. Choi, and R. Luthy. 2012. Hunters Point Pilot Study Experiences[l]: Long-term effectiveness.
Presentation at EPA Region 10 Sponsored Technical Workshop: Use of Activated Carbon Amendment
as an In-situ Sediment Remedy at the Lower Duwamish Waterway Superfund Site. February 14 & 15,
2012.
Cho,Y., E. Janssen, and R. Luthy. 2012. Hunters Point Pilot Study Experiences[ll]: Ecological effect.
Presentation at EPA Region 10 Sponsored Technical Workshop: Use of Activated Carbon Amendment
as an In-situ Sediment Remedy at the Lower Duwamish Waterway Superfund Site. February 14 & 15,
2012.
Cornelisson, G., G. Breedveld, K. Naes, A. Oen, and A. Ruus. 2006. Bioaccumulation of native polycyclic
aromatic hydrocarbons from sediment by a polychaete and a gastropod: freely dissolved
concentrations and activated carbon amendment. Environ. Toxicol. Chem. 25: 2349-2355.
Eek, E., A. Oen, G. Breedveld, M. Schaanning, B. Beylich, K. Naes, and G. Cornelisson. 2012. Field testing
of thin layer capping with AC and passive materials in Norway. Presentation at EPA Region 10
Sponsored Technical Workshop: Use of Activated Carbon Amendment as an In-situ Sediment Remedy
at the Lower Duwamish Waterway Superfund Site. February 14 & 15, 2012.
Fagervold, S., Y. Chai, J. Davis, M. Wilken, G. Cornelisson, and U. Ghosh. 2010. Bioaccumulation of
polychlorinated dibenzo-p-dioxins/dibenzofurans in E. fetida from floodplain soils and the effect of
activated carbon amendment. Environ. Sci. Technol. 44: 5546-5552.
Ghosh, U. 2012a. New advances in contaminated sediment remediation by controlling bioavailability.
Presentation at EPA Region 10 Sponsored Technical Workshop: Use of Activated Carbon Amendment
as an In-situ Sediment Remedy at the Lower Duwamish Waterway Superfund Site. February 14 & 15,
2012.
Ghosh, U. 2012b. Potential effects of activated carbon on benthic animals and plants: Grasse River
Study. Presentation at EPA Region 10 Sponsored Technical Workshop: Use of Activated Carbon
Amendment as an In-situ Sediment Remedy at the Lower Duwamish Waterway Superfund Site.
February 14 & 15, 2012.
Greenberg, M. 2012a. Grasse River NY activated carbon study. Presentation at EPA Region 10 Sponsored
Technical Workshop: Use of Activated Carbon Amendment as an In-situ Sediment Remedy at the
Lower Duwamish Waterway Superfund Site. February 14 & 15, 2012.
The Isosceles Group, Inc. Page 5 of 7
A Sleeman Hanley & DiNitto Company
50 Congress Street, Suite 840
Boston, Massachusetts 02109
www, theisogroup.com
-------
Greenberg, M. 2012b. Ecological effects considerations. Presentation at EPA Region 10 Sponsored
Technical Workshop: Use of Activated Carbon Amendment as an In-situ Sediment Remedy at the
Lower Duwamish Waterway Superfund Site. February 14 & 15, 2012.
Janssen, E., M.-N. Croteau, S. Luoma, and R. Luthy. 2010. Measurement and modeling of polychlorinated
biphenyl bioaccumulation from sediment for the marine polychaete Neanthes arenaceodentata and
response to sorbent amendment. Environ. Sci. Technol. 44: 2857-2863.
Jonker, M., M. Suijkerbuijk, H. Schmitt, and T. Sinnige. 2009. Ecotoxicological effects of activated carbon
addition to sediments. Environ. Sci. Technol. 43: 5959-5966.
Kupryianchyk, D., E. Reichman, M. Rakowska, E. Peeters, J. Grotenhuis, and A. Koelmans. 2011.
Ecotoxicological effects of activated carbon amendments on macroinvertebrates in nonpolluted and
polluted sediments. Environ. Sci. Technol. 45: 8567-8574.
McDonough, K., P. Murphy, J. Olsta, Zhu, Y., D. Reible, and J. Lowry. 2007. Development and placement
of a sorbent amended thin layer sediment cap in the Anacostia River. Soil. Sedimen. Contam. 16: 313-
322.
McLeod, P., M. Van Den Heuvel-Greve, R. Allen-King, S. Luoma, and R. Luthy. 2004. Effects of particulate
carbonaceous matter on the bioavailability of benzo[o]pyrene and 2,2',5,5'-tetrachlorobiphenyl to the
clam, Macoma balthica. Environ. Sci. Technol. 38: 4549-4566.
McLeod, P., M. Van Den Heuvel-Greve, S. Luoma, and R. Luthy. 2007. Biological uptake of
polychlorinated biphenyls by Macoma balthica from sediment amended with activated carbon.
Environ. Toxicol. Chem. 26:980-987.
McLeod, P., S. Luoma, and R. Luthy. 2008. Biodynamic modeling of PCB uptake by Macoma balthica and
Corbicula fluminea from sediment amended with activated carbon. Environ. Sci. Technol. 42: 484-490.
Melton, J. 2012. Engineering considerations for activated carbon placement and stability. Presentation
at EPA Region 10 Sponsored Technical Workshop: Use of Activated Carbon Amendment as an In-situ
Sediment Remedy at the Lower Duwamish Waterway Superfund Site. February 14 & 15, 2012.
Menzie, C. 2012. Evaluating the potential for adverse effects of activated carbon on aquatic and marine
animals. Presentation at EPA Region 10 Sponsored Technical Workshop: Use of Activated Carbon
Amendment as an In-situ Sediment Remedy at the Lower Duwamish Waterway Superfund Site.
February 14 & 15, 2012.
Millward, R., T. Bridges, U. Ghosh, J. Zimmerman, and R. Luthy. 2005. Addition of activated carbon to
sediments to reduce PCB bioaccumulation by a polychaete (Neanthes arenaceodentata) and an
amphipod (Leptocheirus plumulosus). Environ. Sci. Technol. 39:2880-2887.
Reible, D. 2012. Contaminant partitioning and bioavailability. Presentation at EPA Region 10 Sponsored
Technical Workshop: Use of Activated Carbon Amendment as an In-situ Sediment Remedy at the
Lower Duwamish Waterway Superfund Site. February 14 & 15, 2012.
Sun, X. and U. Ghosh. 2007. PCB bioavailability control in Lumbriculus variegatus through different
modes of activated carbon addition to sediments. Environ. Sci. Technol. 41: 4774-4780.
The Isosceles Group, Inc. Page 6 of 7
A Sleeman Hanley & DiNitto Company
50 Congress Street, Suite 840
Boston, Massachusetts 02109
www, theisogroup.com
-------
Sun, X. and U. Ghosh. 2008. The effect of activated carbon on partitioning, desorption, and biouptake of
native polychlorinated biphenyls in four freshwater sediments. Environ. Toxicol. Chem. 11:2887-2895.
Sun, X., D. Werner, and U. Ghosh. 2008. Modeling PCB mass transfer and bioaccumulation in a
freshwater oligochaete before and after amendment of sediment with activated carbon. Environ. Sci.
Technol. 43: 1115-1121.
Tomaszewski, J., P. McLeod, and R. Luthy. 2008. Measuring and modeling reduction of DDT availability
to the water column and mussels following activated carbon amendment of contaminated sediment.
Water Research 42: 4348-4356.
The Isosceles Group, Inc. Page 7 of 7
A Sleeman Hanley & DiNitto Company
50 Congress Street, Suite 840
Boston, Massachusetts 02109
www, theisogroup.com
------- |