HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
CsSfe)
PRO^
For
U.S. Environmental Protection Agency
Region II
and
U.S. Army Corps of Engineers
Kansas City District
Book 1 of 3
TAMS Consultants, Inc.
Limno-Tech, Inc.
TetraTech, Inc.
Menzie-Cura & Associates, Inc.
-------
*1 PfiO,t
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION 2
290 BROADWAY
NEW YORK. NY 10007-1866
DEC 291998
To All Interested Parties:
Earlier this year, the U.S. Environmental Protection Agency (EPA) decided to issue
responsiveness summaries for the various Hudson River PCB Site Reassessment reports as the
study progressed, rather than issuing a single responsiveness summary after the Proposed Plan,
as had previously been scheduled. As such, this document is the first of the responsiveness
summaries to be prepared on the Phase 2 Reports.
This document contains written comments from various reviewers on the Database Report, the
Preliminary Model Calibration Report, and the Data Evaluation and Interpretation Report, and
the Agency's response to significant comments on those reports. In addition, the responsiveness
summary includes a "Review and Commentary" on a report submitted to EPA by General
Electric summarizing the company's analyses of Thompson Island Pool Sediment PCB sources.
A responsiveness summary for the Low Resolution Sediment Coring Report, the Human Health
Risk Assessment Scope of Work, the Ecological Risk Assessment Scope of Work, and the
Feasibility Study Scope of Work will be issued in Spring 1999.
Public involvement is important to the Agency. I am pleased to provide this response to your
concerns about both technical and policy issues.
Sincerely yours,
Richard L. Caspe, Director
Emergency and Remedial Response Division
Internet Address (URL) • http //www epa gov
R«cycl»d/R*cyclabl* •Printed wrth Vegetable Oil Based Inks on Recycled Paper (Minimum 25% Postconsumer)
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
S7^
ySfei
V
*<- PRCfl4"
For
U.S. Environmental Protection Agency
Region II
and
U.S. Army Corps of Engineers
Kansas City District
Book 1 of 3
TAMS Consultants, Inc.
Limno-Tech, Inc.
TetraTech, Inc.
Menzie-Cura & Associates, Inc.
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Page
Table of Contents i
List of Corrections xiv
List of Figures xiv
List of Tables xv
I. INTRODUCTION AND COMMENT DIRECTORY
1. INTRODUCTION CD-I
1.1 Recent Developments CD-2
2. REPORT COMMENTING PROCESS CD-2
2.1 Reports Distribution CD-2
2.2 Review Period and Informational Meetings CD-3
2.3 Receipt of Comments CD-3
2.3.1 Comments on the Database Report and Database CD-3
2.3.2 Comments on the Preliminary Model Calibration Report CD-3
2.3.3 Comments on the Data Evaluation and Interpretation Report CD-6
2.3.4 General Electric Report: Thompson Island Pool Sediment PCB Sources . . . CD-6
2.4 Distribution of Responsiveness Summary CD-6
3. ORGANIZATION OF COMMENTS AND RESPONSES TO REPORTS CD-7
3.1 Identification of Comments CD-7
3.2 Location of Responses to Comments CD-9
3.3 Types of Responses CD-10
4. COMMENT DIRECTORY CD-I 1
4.1 Guide to Comment Directory Responsiveness Summary CD-11
4.2 Comment Directory for the Database Report and Database CD-I3
4.3 Comment Directory for the PMCR CD-14
4.4 Comment Directory for the DEIR CD-22
December 22. 1998
TAMS/LTI/TctraTech/MCA
-------
HUDSON RIVER PCBs REASSESSMENT RJLFS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Eage
II. RESPONSES TO COMMENTS
A. DATABASE REPORT AND DATABASE DB-1
RESPONSES TO GENERAL COMMENTS ON DATABASE REPORT DB-1
B. PMCRREPORT PMCR-1
RESPONSES TO GENERAL COMMENTS ON PMCR REPORT PMCR-1
RESPONSES TO SPECIFIC COMMENTS ON PMCR REPORT PMCR-2
Executive Summary PMCR-2
1. Introduction PMCR-3
1.1 Background PMCR-3
1.2 Purpose of Report PMCR-3
1.3 Report Format and Organization PMCR-3
2. Summary and Preliminary Conclusions PMCR-4
2.1 Summary PMCR-4
2.1.1 Overall Approach PMCR-4
2.1.2 Water Column and Sediment Models PMCR-4
2.1.3 Fish Body Burden Models PMCR-4
2.2 Preliminary Conclusions PMCR-4
2.2.1 Upper Hudson River PCB Mass Balance PMCR-4
2.2.2 Thompson Island Pool Hydrodynamics and Sediment Erosion . PMCR-4
2.2.3 Upper Hudson River Fish Body Burdens PMCR-5
2.2.4 Lower Hudson PCB Mass Balance and Striped Bass
Bioaccumulation PMCR-5
3. Modeling Approach: Transport and Fate PMCR-6
3.1 Introduction PMCR-6
3.2 Modeling Goals and Objectives PMCR-6
December 22. 1998 T AMS/LTi/T ctra l'ech/MC A
ii
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Ease
3.3 Conceptual Approach PMCR-6
3.4 Hudson River Database PMCR-6
3.5 Upper Hudson River Mass Balance Model PMCR-6
3.5.1 Introduction PMCR-7
3.5.2 State Variables and Process Kinetics PMCR-8
3.5.3 Spatial-Temporal Scales PMCR-8
3.5.4 Application Framework PMCR-8
3.6 Thompson Island Pool Hydrodvnamic Model PMCR-8
3.6.1 Introduction PMCR-8
3.6.2 State Variables and Process Mechanisms PMCR-8
3.6.3 Spatial-Temporal Scales PMCR-8
3.6.4 Application Framework PMCR-8
3.7 Thompson Island Pool Depth of Scour Model PMCR-8
3.7.1 Introduction PMCR-9
3.7.2 Process Representation PMCR-9
3.7.3 Spatial Temporal Scales PMCR-9
3.7.4 Applications Framework PMCR-9
3.8 Lower Hudson River PCB Transport and Fate Model PMCR-9
3.8.1 Introduction PMCR-9
3.8.2 State Variables and Process Kinetics PMCR-10
3.8.3 Spatial-Temporal Scales PMCR-10
3.8.4 Applications Framework PMCR-10
4. Calibration of Upper Hudson River PCB Model PMCR-11
4.1 Introduction PMCR-11
4.2 Historical Trends in Water Quality Observations PMCR-12
4.3 Overview of Preliminary Calibration Data Set PMCR-12
4.4 Model Input Data PMCR-12
4.4.1 System-Specific Physical Data PMCR-13
4.4.2 l:\ternalLoadings PMCR-13
December 22. 1W8 TAMS/LTl/TetraTcch/MCA
III
-------
HUDSON RIVER PCBs REASSESSMENT Rl/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Ease
4.4.3 Forcing Functions PMCR-14
4.4.4 Boundary Conditions PMCR-15
4.4.5 Initial Conditions PMCR-15
4.5 Internal Model Parameters PMCR-15
4.5.1 Solids Model Parameters PMCR-16
4.5.2 PCB Model Parameters PMCR-16
4.6 Calibration Approach PMCR-16
4.6.1 Transport Model (Water Balance) Specification PMCR-16
4.6.2 Solids Model PMCR-16
4.6.3 PCB Model PMCR-16
4.7 Calibration Results PMCR-17
4.7.1 Solids Model PMCR-19
4.7.2 PCB Model PMCR-19
4.8 Mass Balance Component Analysis PMCR-20
4.9 PCB Model Calibration Sensitivity Analysis PMCR-20
5. Calibration of Thompson Island Pool Hydrodynamic Model PMCR-21
5.1 Introduction PMCR-21
5.2 Model Input Data PMCR-21
5.2.1 System-Specific Physical Data PMCR-21
5.2.2 Forcing Functions PMCR-21
5.2.3 Boundary Conditions PMCR-21
5.3 Internal Model Parameters PMCR-21
5.4 Calibration Approach PMCR-21
5.5 Calibration Results PMCR-21
5.6 Model Validation PMCR-21
5.6.1 Rating Curve Velocity Measurements PMCR-21
5.6.2 FEMA Flood Studies PMCR-21
5.7 100 Year Flood Model Results PMCR-21
5.8 Sensitivity Analyses PMCR-21
December 22 1998 TAMS/LTI/TciraTech/MCA
iv
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Page
5.8.1 Manning's 'n' PMCR-21
5.8.2 Turbulent Exchange Coefficient PMCR-21
5.9 Conversion of Flow Velocity to Shear Stress PMCR-21
5.9.1 Results PMCR-21
5.10 Discussion PMCR-21
6. Application of Thompson Island Pool Depth of Scour Model PMCR-22
6.1 Introduction PMCR-22
6.2 Available Data PMCR-22
6.2.1 Bottom Sediment Distribution PMCR-22
6.2.2 Resuspension Experiments PMCR-22
6.3 Model Parameterization and Uncertainty PMCR-22
6.3.1 Rearrangement of Erosion Equation PMCR-23
6.3.2 Parameter Estimation PMCR-23
6.3.3 Prediction Limits PMCR-23
6.4 Depth of Scour Predictions at Selected Locations in Cohesive
Sediment Areas PMCR-23
6.5 Global Results for Cohesive Sediment Areas PMCR-23
7. Application of Lower Hudson River PCB Transport and Fate Model PMCR-24
7.1 Introduction PMCR-24
7.2 Model Input Data PMCR-24
7.2.1 System-Specific Physical Data PMCR-24
7.2.2 External Loadings PMCR-24
7.2.3 Forcing Functions PMCR-24
7.2.4 Boundary7 Conditions PMCR-24
7.2.5 Initial Conditions PMCR-24
7.3 Internal Model Parameters PMCR-24
7.4 Application Approach PMCR-24
7.5 Application Results PMCR-25
7.6 Diagnostic Analyses PMCR-25
December 22. 1998 TAMS/LTI/TctraTech/MCA
V
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Page
7.6.1 Component Analysis .. PMCR-25
7.6.2 Sensitivity Analysis PMCR-25
7.7 Discussion PMCR-25
8. Modeling Approach: Fish Body Burdens PMCR-26
8.1 Modeling Goals and Objectives PMCR-26
8.2 Background PMCR-26
8.2.1 PCB Compounds PMCR-26
8.2.2 PCB Accumulation Routes PMCR-26
8.3 Theory for Models of PCB Bioaccumulation PMCR-27
8.4 Bivariate Statistical Model for Fish Body Burdens PMCR-27
8.4.1 Rationale and Limitations for Bivariate Statistical Model .... PMCR-27
8.4.2 Theory for Bivariate Statistical Models of PCB
Bioaccumulation PMCR-27
8.5 Probabilistic Bioaccumulation Food Chain Model PMCR-27
8.5.1 Rationale and Limitations PMCR-27
8.5.2 Model Structure PMCR-27
8.5.3 Spatial Scale for Model Application PMCR-27
8.5.4 Temporal Scales for Estimating Exposure to Fish PMCR-27
8.5.5 Characterizing Model Compartments PMCR-27
9. Calibration of Bivariate Statistical Model for Fish Body Burdens PMCR-28
9.1 Data Used for Development of Bivariate BAF Models PMCR-28
9.1.1 Fish Data PMCR-28
9.1.2 Standardization of PCB Results for NYSDEC Fish Analyses . PMCR-28
9.1.3 Water Column Data PMCR-28
9.1.4 Sediment Data PMCR-28
9.1.5 Functional Grouping of Sample Locations for Analysis PMCR-28
9.2 Results of Bivariate BAF Analysis PMCR-28
9.3 Discussion of Bivariate BAF Results PMCR-28
9.4 Summary PMCR-29
December 22. 1998 TAMS/LTI/TeiraTech/MCA
vi
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Eage
10. Calibration of Probabilistic Bioaccumulation Food Chain Model PMCR-30
10.1 Overview of Data Used to Derive BAFs PMCR-30
10.1.1 Benthic Invertebrates PMCR-30
10.1.2 Water Column Invertebrates PMCR-30
10.1.3 Fish PMCR-30
10.1.4 Literature Values PMCR-30
10.2 Benthic Invertebrate:Sediment Accumulation Factors (BSAF) PMCR-30
10.2.1 Sediment Concentrations PMCR-30
10.2.2 Approach PMCR-30
10.2.3 Calculations of BSAF Values for Benthic Invertebrates PMCR-31
10.3 Water Column Invertebrate:Water Accumulation Factors (BAFs) PMCR-32
10.3.1 Approach PMCR-32
10.3.2 Calculation of BAFwater for Water Column Invertebrates PMCR-32
10.3.3 Alternative Approaches PMCR-32
10.4 Forage Fish:Diet Accumulation Factors (FFBAFs) PMCR-32
10.4.1 Approach PMCR-32
10.4.2 Water Column Concentrations Used to Derive FFBAF Values PMCR-32
10.4.3 Forage Fish Body Burdens Used to Derive FFBAF Values . . . PMCR-32
10.4.4 Calculation of FFBAF Values for Forage Fish PMCR-33
10.4.5 Calculation of FFBAFs for Small Pumpkinseed Sunfish PMCR-33
10.5 Piscivorous Fish:Diet Accumulation Factors (PFBAF) PMCR-33
10.5.1 Approach Used for Yellow Perch PMCR-33
10.5.2 Approach Used for Largemouth Bass PMCR-33
10.5.3 Approach Used for White Perch PMCR-33
10.6 Demersal Fish:Sediment Relationships PMCR-33
10.6.1 Approach and Calculations of BAF Values PMCR-33
10.7 Summary of Probabilistic Food Chain Models PMCR-33
10.8 Illustration of Food Chain Model Application PMCR-33
10.9 Comparison of Bivariate Statistical and Food Chain Models PMCR-33
December 22. 1998 TAMS/LTI/TetraTech/MCA
vii
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Page
References , PMCR-33
Glossary PMCR-34
Appendix A: Fish Profiles PMCR-34
Appendix B: Mathematical Modeling, Technical Scope of Work PMCR-34
C: DEIR REPORT DEIR-1
RESPONSES TO GENERAL COMMENTS ON THE DEIR REPORT DEIR-1
RESPONSES TO SPECIFIC COMMENTS ON THE DEIR REPORT DEIR-5
Executive Summary DEIR-5
Chapter 1 - Introduction DEIR-9
1.1 Purpose of Report DEIR-9
1.2 Report Format and Organization DEIR-9
1.3 Technical Approach of the Data Evaluation and Interpretation Report DEIR-9
1.4 Review of the Phase 2 Investigations DEIR-9
1.4.1 Review of PCB Sources DEIR-9
1.4.2 Water Column Transport Investigation DEIR-9
1.4.3 Assessment of Sediment PCB Inventory and Fate DEIR-10
1.4.4 Analytical Chemistry Program DEIR-10
Chapter 2 - PCB Sources to the Upper and Lower Hudson River DEIR-12
2.1 Background DEIR-12
2.2 Upper Hudson River Sources DEIR-12
2.2.1 NYSDEC Registered Inactive Hazardous Waste Disposal
Sites DEIR-12
2.2.2 Remnant Deposits DEIR-12
2.2.3 Dredge Spoil Sites DEIR-13
2.2.4 Other Upper Hudson Sources DEIR-13
2.3 Lower Hudson River Sources DEIR-13
December 22. 1998
viii
TAMS/LTI/TetraTech/MCA
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Pass
2.3.1 Review of Phase 1 Analysis DEIR-15
2.3.2 Sampling of Point Sources in New York
New Jersey (NY/NJ) Harbor DEIR-15
2.3.3 Other Downstream External Sources DEIR-15
Chapter 3 - Water Column PCB Fate and Transport In the Hudson River .... DEIR-16
3.1 PCB Equilibrium Partitioning DEIR-16
3.1.1 Two-Phase Models of Equilibrium Partitioning DEIR-18
3.1.2 Three-Phase Models of Equilibrium Partitioning DEIR-18
3.1.3 Sediment Equilibrium Partition Coefficients DEIR-18
3.1.4 Summary DEIR-18
3.2 Water Column Mass Loading DEIR-18
3.2.1 Phase 2 Water and Sediment Characterization DEIR-19
3.2.2 Flow Estimation DEIR-20
3.2.3 Fate Mechanisms DEIR-23
3.2.4 Conceptual Model of PCB Transport in the Upper Hudson DEIR-25
3.2.5 River Characterization DEIR-25
3.2.6 Mass Load Assessment DEIR-25
3.2.7 Source Loading Quantitation DEIR-35
3.3 Historical Water Column Transport of PCBs DEIR-37
3.3.1 Establishing Sediment Core Chronologies DEIR-37
3.3.2 Surface Sediment Characterization DEIR-37
3.3.3 Water Column Transport of PCBs Shown by Sediment
Deposited After 1975 DEIR-38
3.3.4 Estimation of the PCB Load and Concentration across the
Thompson Island Pool based on GE Capillary Column Data DEIR-43
3.3.5 Estimated Historical Water Column Loadings Based on
USGS Measurements DEIR-44
3.3.6 Conclusions Concerning Historical Water Column Transport DEIR-44
3.4 Integration of Water Column Monitoring Results DEIR-44
December 22. 1998 TAMS/LTl/T etraTech/MCA
ix
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Page
3.4.1 Monitoring Techniques and PCB Equilibrium DEIR-44
3.4.2 Loadings Upstream of the Thompson Island Pool DEIR-44
3.4.3 Loading from the Thompson Island Pool during 1993 DEIR-45
3.4.4 Loading at the Thompson Island Dam - 1991 to 1996 DEIR-46
3.4.5 PCB Loadings to Waterford DEIR-46
3.4.6 PCB Loadings to the Lower Hudson DEIR-47
3.5 Integration of PCB Loadings to Lower Hudson River and
New York/New Jersey Harbor DEIR-47
3.5.1 Review of Lower Hudson PCB Mathematical Model DEIR-47
3.5.2 Estimate of 1993 PCB Loading from the
Upper Hudson River DEIR-47
3.5.3 Revised PCB Loading Estimates DEIR-47
3.6 Water Column Conclusion Summary DEIR-47
Chapter 4 - Inventory and Fate of PCBs in the Sediment of the Hudson River DEIR-49
4.1 Characterization of Upper Hudson Sediments by Acoustic
Techniques DEIR-51
4.1.1 Geophysical Data Collection and Interpretation Techniques DEIR-51
4.1.2 Correlation of Sonar Image Data and Sediment
Characteristics DEIR-51
4.1.3 Delineation of PCB-Bearing and Erodible Sediments DEIR-51
4.2 Geostatistical Analysis of PCB Mass in the Thompson Island Pool.
1984 DEIR-51
4.2.1 Data Preparation for PCB Mass Estimation DEIR-51
4.2.2 Geostatistical Techniques for PCB Mass Estimation DEIR-51
4.2.3 Polygonal Declustering Estimate of Total PCB Mass DEIR-51
4.2.4 Geostatistical Analysis of Total PCB Mass DEIR-51
4.2.5 Kriging Total PCB Mass DEIR-51
4.2.6 Kriged Total Mass Estimate DEIR-51
4.2.7 Surface Sediment PCB Concentrations DEIR-51
" '998
.0012
X
TAMS/LTI/TetraTech/MCA
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 1 OF 3 Ease
4.2.8 Summary' DEIR-51
4.3 PCB Fate in Sediments of the Hudson River DEIR-51
4.3.1 Anaerobic Dechlorination and Aerobic Degradation DEIR-57
4.3.2 Anaerobic Dechlorination as Documented in Phase 2
High-Resolution Sediment Cores DEIR-60
4.4 Implication of the PCB Fate in the Sediments for Water Column
Transport DEIR-68
4.5 Summary and Conclusions DEIR-73
References DEIR-75
Volume 2C (Book 2 of 3) Tables, Figures, and Plates DEIR-75
Volume 2C (Book 3 of 3)
Appendix A: Data Usability Report for PCB Congeners High Resolution
Sediment Coring Study DEIR-75
Volume 2C (Book 3 of 3)
AppendixB: Data Usability Report for PCB Congeners Water Column
Monitoring Program DEIR-98
Volume 2C (Book 3 of 3)
Appendix C: Data Usability Report for Non-PCB Chemical and Physical Data DEIR-99
D. ADDITIONAL REFERENCES FOR THE RESPONSIVENESS SUMMARY R-l
December 22. 1998
XI
T AMS/1-TI/T eira I ech/MCA
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B; PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 2 of 3
III. COMMENTS ON THE PHASE 2 REPORTS
A. COMMENTS ON THE DATABASE REPORT AND DATABASE
General Electric (DB-1)
B. COMMENTS ON THE PMCR
Federal (PF-1)
Local (PL-1)
Community Interaction Program (PC-1)
Public Interest Groups and Individuals (PF-1)
General Electric (PG-1)
C. COMMENTS ON THE DEIR
Federal (DF-2)
State (DS-2)
Local (DI.-1)
Community Interaction Program (DC-1 through DC-4)
Public Interest Groups and Individuals (DP-1 through DP-5)
General Electric (DG-1)
December 22. 1998
xti
TAMS/LTI/TetraTech/MC A
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
BOOK 3 OF 3
IV. USEPA REVIEW AND COMMENTARY ON THE GENERAL ELECTRIC/QEA
REPORT, MARCH 1998
A. REVIEW AND COMMENTARY ON THE GE/QEA REPORT
B. GE/QEA REPORT: THOMPSON ISLAND POOL SEDIMENT PCB SOURCES,
MARCH 1998
December 22. 1998
xiii
T AMS/L Tl/T etraTech/MC A
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
Page
LIST OF CORRECTIONS
Book I of3
Section 3.2 Water Column Mass Loading ....
Subsection 3.2.2 Correction to Flow Estimation
LIST OF FIGURES
Book 1 of3
Figure DC-4.6 Relative Percent Difference for Phase II Water Column Split
Samples (Total PCBs) DEIR-11
Figure DF-2.6 Relationship Between Total Suspended Solids and Total Organic
Carbon for 1993 Phase 2 Transect and Flow-Averaged Samples . . DEIR-50
Figure DF-2.7 General Electric Hudson Falls Source - Seepage Homologue
Distribution, May 1993 DEIR-14
Figure DG-1.15A Water Column PCB Concentrations Within the Vicinity of Fort
Edward from the 1995 River Monitoring Test DEIR-29
Figure DG-1.15B PCB and Solids Transport During 1992 Spring High Flow DEIR-31
Figure DG-1.15C Temporal Trends in TSS and PCB Concentration and Loading
During the 1997 Spring High Flow Period DEIR-32
Figure DG-1.15D Water Column PCB Concentrations at Bakers Falls Plunge Pool
and Fort Edward from Hydrofacility Monitoring Program DEIR-34
Figure DG-1.17 Trend of Various H.H' Markers in Recent Sediments (0-2 cm) as a
Function of River Mile DEIR-42
Figure DG-1.19A The Number of Chorines per Biphenyl vs. The GE/HydroQual
Dechlorination Ratios for the High Resolution Core Data DEIR-54
Figure DG-1.19B The Relationship Between the Number of Chorines per Biphenyl
and the Molar Dechlorination Product Ratio for the High Resolu-
tion Core Data DEIR-55
December 22. 1998 TAMS/LTI/TetraTech/MCA
\iv
DEIR-18
DEIR-19
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY FOR
VOLUME 2A: DATABASE REPORT
VOLUME 2B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME 2C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
TABLE OF CONTENTS
Page
Figure DG-1.20A Total PCBs vs. The GE/HydroQual Dechlorination Ratios for the
High Resolution Core Data (Upper Hudson) DEIR-66
Figure DG-1.20B Total PCBs vs. The GE/HydroQual Dechlorination Ratios for the
High Resolution Core Data (Lower Hudson Freshwater) DEIR-67
Figure DG-1.20C Relationship Between the Number of Chorines per Biphenyl, Molar
Dechlorination Product Ratio and Total PCBs for the High
Resolution Core Data DEIR-69
Figure DG-1.20D The Number of Chorines per Biphenyl vs. The GE/HydroQual
Dechlorination Ratios for the High Resolution Core Data (Lower
Hudson) DEIR-70
Figure DG-1.26R Histogram of the Change in Molecular Weight as a Function of
Time of Deposition in Post-1954 Dated Sediments from the Hudson
River DEIR-64
Figure 3.2.2A Fort Edward to Stillwater Incremental Summer Average Flow vs.
Total Precipitation for Glens Falls DEIR-21
Figure 3.2.2B Fort Edward to Stillwater Incremental Summer Average Flow vs.
Total Precipitation for NCDC-Division 5 (Hudson River Valley) . . DEIR-22
LIST OF TABLES
Book of 1 of 3
Table 1 Distribution of Reports CD-4
Table 2 Information Repositories CD-5
Table DF-2.2A Water Column Study - Dissolved PCBs DEIR-77
Table DF-2.2B Water Column Study - Particulate Data DEIR-80
Table DF-2.2C High Resolution Coring Study - Sediment Core Sample Data .... DEIR-83
Table DG-1.17 Suspended Solids Yields for the Hudson River to Albany DEIR-40
Notes: Figures and Tables for the USEPA commentary on the GE QEA report. March 1998 are
listed in the Table of Contents contained in Book 3 of this Responsiveness Summary.
December 22. 1998 TAMS/LTiyTetraTech/MCA
xv
-------
i
THIS PAGE LEFT BLANK INTENTIONALLY
-------
Introduction
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY
VOLUME A: DATABASE REPORT
VOLUME B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
I. Introduction and Comment Directory
1. Introduction
USEPA has prepared this Responsiveness Summary for the first three volumes of the Phase
2 Report, specifically, the Database Report, the Preliminary Model Calibration Report (PMCR), and
the Data Evaluation and Interpretation Report (DE1R) for the Hudson River PCB Reassessment
Remedial Investigation/Feasibility Study (RI/FS). It addresses comments received during the review
of these three reports. This Responsiveness Summary also presents and comments on a free-standing
report prepared by General Electric Company.
For the Reassessment, USEPA has established a Community Interaction Program (CIP) to
elicit on-going feedback through regular meetings and discussion and to facilitate review of and
comment upon work plans and reports prepared during all phases.
The three reports are incorporated by reference and are not reproduced herein. No revised
copy of the reports will be published as such. The comment responses and revisions noted herein
are considered to amend two of the three reports; i.e, the Database Report and the Data Evaluation
and Interpretation Report. Although the Preliminary Model Calibration Report (PMCR) will not be
republished as such, the PMCR forms the basis for the forthcoming Baseline Modeling Report
(BMR); comments on the PMCR will be incorporated as noted into the BMR. For complete
coverage, the reports and this Responsiveness Summary must be used together.
The first part of this three-part Responsiveness Summary is entitled "Introduction and
Comment Directory." It describes the Report review and commenting process, explains the
organization and format of comments and responses and contains a comment index or directory.
The second part, entitled "Responses", contains the USEPA responses to all comments. This
section is broken down into three parts; one for each of the three reassessment reports. Responses
are grouped first according to the report, and then according to the section number of the report to
which they refer, e g , responses to comments on Section 2.1 of the DEIR are found in DEIR Section
2.1 of the Responsiveness Summary. Additional information about how to locate responses to
comments is contained in the Comment Directory.
December 22 1998
CD-I
TAMS/LTI/TelraTech/MCA
-------
The third part, entitled "'Comments on the Phase 2 Report", contains the copies of the
comments submitted to the USEPA on Volumes A, B, and C of the Phase 2 report. The comments
are identified by commentor and comment number, as further explained in the Comment Directory.
In addition to specific comments, GE submitted a separate report on the Hudson River
("Thompson Island Pool Sediment PCB Sources: QEA, 1998). The USEPA has prepared a critique
of the GE report; both GE's report and the USEPA critique are contained in this responsiveness
summary as Book 3 of the Responsiveness Summary.
1.1 Recent Developments
Since the issuance of the DEIR, further review of the data and some of the models has
revealed errors and corrections that affect these reports, as noted below.
• The Hudson River flow estimate has been revised. The USGS data are now being used for
flow estimation. This requires revising the loads below the Thompson Island Dam. This
revision will be provided in the response to comments on the Low Resolution Sediment
Coring Report.
• General Electric has noted an error in their data, and is correcting their data to reflect a
greater presence of the lighter (lower molecular weight) PCB congeners; resulting in about
a 40 percent increase to the total PCB concentrations they report.
• Due to the location of the sampling station at the Thompson Island Dam, the PCB load at the
TI Dam may be overestimated. The degree of overestimation varies as a function of flow and
the PCB load at Rogers Island. The degree of overestimation during low-flow conditions was
estimated to be 20 percent (less than 4000 cfs) during the period 1991 - 1995. The degree of
overestimation is also less pronounced for trichlorinated and higher congeners relative to
total PCBs.
This section documents and explains the commenting process and the organization of
comments and responses in this document. Readers interested in finding responses to their
comments may skip this section and go directly to the Comment Guide to the Comment Directory
following page CD-10.
2. Reports Commenting Process
2.1 Reports Distribution
The Database Report, the PMCR. and the DEIR. were issued in November, 1995, October,
1996. and February, 1997, respectively, and were distributed to federal and state agencies and
officials, participants in the Community Interaction Program (CIP). and General Electric, as shown
in Table 1. Distribution was made to approximately 100 agencies, groups, and individuals. Copies
CD-2
T AMS/LTI/T etraTech/MCA
-------
of the reports were also made available tor public review in 17 information repositories, as shown
in Table 2.
2.2 Review Period and Informational Meetings
Official thirty-day comment periods were specifically associated with each report, but
comments on aJl reports, both current and prior, have been welcome throughout the process. USEPA
held three Joint Liaison Group meetings that were open to the public to present these reports. The
meetings were held in December 1995 (Latham, NY), October 1996 (Albany, NY), and February
1997 (Albany, NY).
Minutes for these meetings are contained in a binder entitled Project Documents Binder.
This binder is part of the project information available for public review at 11 of the 17 information
repositories (Table 2). Four of the six repositories that do not currently have a Project Documents
Binder (Marist Library, RPI Library, SUNY Albany Library, and USMA Library) are partial
repositories maintained primarily for their CD-ROM capability. The other two, Sojourner Truth
Library at SUNY New Paltz, and the Sea Grant office in Kingston, will have copies of Project
Documents Binders in the near future .
As stated in USEPA's letter transmitting the Reports, all citizens were urged to participate
in the Reassessment process and to join one of the Liaison Groups formed as part of the Community
Interaction Program. USEPA requested that all comments, including those of Liaison Groups, be
sent to USEPA.
2.3 Receipt of Comments
Comments on the reports were received in two ways: letters or other written submissions
to USEPA; and written statements as follow-up to statements made during the meetings.
Comments received on the Reports have been recorded and are addressed in this
Responsiveness Summary. Comments were received from approximately 17 commentors. for the
three reports. Total comments numbered over 320.
2.3.1 Comments on the Database Report and Database
One set of comments was received on the database report (from GE).
2.3.2 Comments on the Preliminary Model Calibration Report
Five sets of comments were received on the PMCR. These included one Federal comment
set (from the National Oceanic and Atmospheric Administration (NOAA); identified as PF-1); one
set of local government comments (from the Saratoga County Environmental Management Council
- PL-1 from Hodgson/Adams); two sets from members of the Science and Technical Committee, a
part of the community interaction program (PC-1. G. Putman from SUNY - Albany, and PP-1 from
J. Sanders); and one set from General Electric (PG-1).
December 22 1998
CD-3
TAMS/LTI/TetraTech/MCA
-------
TABLE 1
DISTRIBUTION OF REPORTS
HUDSON RIVER PCBs OVERSIGHT COMMITTEE MEMBERS
USEPA ERRD Deputy Division Director (Chair)
USEPA Project Manager
USEPA Community Relations Coordinator, Chair of the Steering Committee
NYSDEC Division of Hazardous Waste Management representative
NYSDEC Division of Construction Management representative
National Oceanic and Atmospheric Administration (NOAA) representative
Agency for Toxic Substances and Disease Registry (ATSDR) representative
US Army Corps of Engineers representative
New York State Thruway Authority (Department of Canals) representative
USDOI (USF&W) representative
NYSDOH representative
GE representative
Liaison Group Chairpeople
Scientific and Technical Committee representative
SCIENTIFIC AND TECHNICAL COMMITTEE MEMBERS
STEERING COMMITTEE MEMBERS
USEPA Community Relations Coordinator (Chair)
Governmental Liaison Group Chair and two Co-chairs
Citizen Liaison Group Chair and two Co-chairs
Agricultural Liaison Group Chair and two Co-chairs
Environmental Liaison Group Chair and two Co-chairs
USEPA Project Manager
NYSDEC Technical representative
NYSDEC Community Affairs representative
FEDERAL AND STATE REPRESENTATIVES
Copies of the reports were sent to relevant federal and state representatives who have been involved with
this project. These include, in part, the following:
The Hon. Daniel P. Moynihan - The Hon. Michael McNulty
The Hon. Alfonse M. D'Amato - The Hon. Sue Kelly
The Hon. Gerald Solomon - The Hon. Benjamin Gilman
The Hon. Nita Lowey - The Hon. Richard Brodsky
The Hon. Maurice Hinchev - The Hon. Bobby D'Andrea
The Hon. Ronald B. Stafford
17 INFORMATION REPOSITORIES (see I able 2)
December 22. 1998
(TM
T AMS/I .TI /TelraT ech/MC A
-------
TABLE 2
INFORMATION REPOSITORIES
Adriance Memorial Library
93 Market Street
Poughkeepsie. NY 12601
Catskill Public Library
1 Franklin Street
Catskill, NY 12414
A Cornell Cooperative Extension
Sea Grant Office
74 John Street
Kingston. NY 12401
Crandall Library
City Park
Glens Falls. NY 12801
County Clerk's Office
Washington County Office Building
Upper Broadway
Fort Edward. NY 12828
* A Marist College Library
Marist College
290 North Road
Poughkeepsie. NY 12601
* New York State Library
CEC Empire State Plaza
Albany. NY 12230
New York State Department
of Environmental Conservation
Division of Hazardous Waste Remediation
50 Wolf Road. Room 212
Albany. NY 12233
* A R. G. Folsom Library
Rensselaer Polytechnic Institute
Troy. NY 12180-3590
Saratoga County EMC
50 West High Street
Ballston Spa, NY 12020
* Saratoga Springs Public Library
49 Henry Street
Saratoga Springs, NY 12866
* A SUNY at Albany Library
1400 Washington Avenue
Albany, NY 12222
* A Sojourner Truth Library
SUNY at New Paltz
New Paltz, NY 12561
Troy Public Library
100 Second Street
Troy, NY 12180
United States Environmental Protection
Agency
290 Broadway
New York, NY 10007
* A United States Military Academy Library
Building 757
West Point. NY 10996
White Plains Public Library
100 Martine Avenue
White Plains. NY 12601
Repositories with Database Report
CD-ROM (as of 10/98)
Repositories without Project
Documents Binder (as of 10/98)
December 22. 1998
CD-5
TAMS/l.TI/Tetra Tech/MCA
-------
2.3.3 Comments on the Data Evaluation and Interpretation Report
The most extensive group of comments was received on the Review Copy of the DEIR. A
total of 14 comment sets were received, submitted by two federal agencies, one state agency, one
local government; four community interaction program participants; five public responders; and
General Electric.
Federal agency comments included one set from NOAA (DF-2, 6/3/97).
One set of NY State comments was received from New York State Department of
Environmental Conservation (DS-2; from Deputy Commissioner Sternman, 4/25/97).
Local government comments were submitted by the Saratoga Environmental Management
Council (DL-1).
Comments were submitted by four members of the community interaction program including
T. Borden (chairperson, Agricultural Liason Group; DC-1 4/11/97), M. Pulver (co-chair,
Agricultural Liason Group and Fort Edward Town Board; DC-2 4/11/97); S. Ruggi (member,
Environmetal Group; DC-3); and George Putman (member. Science and Technical Committee; DC-
4 4/11/97).
Five sets of comments from various groups and one individual were received. These
comments were submitted by Hudson River Sloop Clearwater (DP-1), Hudson Riverkeeper Fund
(on Pace Environmental Litigation Clinic letterhead; DP-4, dated 4/11/97), Scenic Hudson (DP-5),
and Sherwood Davies (resident; DP-2 and DP-3, letters dated 11/11/94 and 4/5/97).
General Electric (DG-1) comments constituted virtually a free-standing report, with 63 pages
of text plus 22 pages of tables and figures as well as five additional appendices.
2.3.4 General Electric Report: Thompson Island Pool Sediment PCB Sources
A consultant to General Electric, Quantitative Environmental Analysis (QEA), sumbitted a
separate, free-standing report entitled "Thompson Island Pool Sediment PCB Sources" dated March
19, 1998. This QEA report has been considered separately. The USEPA response to this report is
provided separately in Book 3 of this report. The full text of the QEA report is included in Book 3
as well.
2.4 Distribution of Responsiveness Summary
This Responsiveness Summary, like all other documents prepared for the Reassessment, has
been distributed to the members of the Steering Committee, the Hudson River PCB Oversight
Committee, the Scientific and Technical Committee. NYSDEC and General Electric. This
Responsiveness Summary has also been placed in the 17 Information Repositories and is part of the
Administrative Record.
December 22. 1998
CD-6
TAMS/LTI Aetra Tech/MCA
-------
3.
ORGANIZATION OF COMMENTS AND RESPONSES TO REPORTS
3,1 Identification of Comments
Each comment submitted for a Report was assigned a dual letter code. The first letter
references an individual Report (P for PMCR and D for DEIR) for which the comment was
addressed and the second letter was used to denote one of the following:
F - Federal agencies and officials;
S - State agencies and officials;
L - Local agencies and officials;
C - Community Interaction Program Committees and Liaison Groups;
P - Public Interest Groups and Individuals; and
G - General Electric.
It should be noted that the code for Database Report comments (DB) were numbered
sequentially, and did not use the second letter code defined above, due to the limited nature of the
comments received. The letter codes were assigned for the convenience of readers and to assist in
the organization of this document; priority or special treatment was neither intended nor given in the
responses to comments.
Once a letter code was assigned, each submission was then assigned a number, in the order
that it was received and processed, such as PF-1, PF-2 and so on. Each different comment within
a submission was assigned its separate sub-number. Thus, if a federal agency submitted three
different comments under the same cover on the DEIR, they are designated as DF-1,1, DF-1.2, DF-
1.3.
Written comment submissions have been reprinted following the third tab of this document.
In addition, a separate report was provided by a consultant (QEA) to General Electric (GE) entitled
"Thompson Island Pool Sediment PCB Sources, Final Report" and is provided in Section IV of this
report. USEPA's response to this GE/QEA report is also included in Section IV of the report, found
in Book 3.
The alphanumeric code associated with each reprinted written submission is marked at the
top right corner of the first page of the comment letter; the sub-numbers designating individual
comments are marked in the margin, as shown in the sample letter on the following page. Comment
submissions are reprinted in numerical order by letter code in the following order: F, S, L, C, P, and
G.
In a few instances, a commentor may have more than one submission listed in the Comment
Directory, because he/she made several submissions.
It was not always clear if a commentor intended to represent a CIP Committee or Liaison
Group, was representing an interest group or was commenting as an individual. The reader is
advised to examine both the C (CIP) category for the name of the CIP Committee or Liaison Group
December 22. 1998
rrv?
TAMS/LTI/Tctra Tech/MCA
-------
MPLE COMMENT LETTER
U.S. DEPARTMENT OF COMMERC?;
National Oceanic and Atmospnortc
Administration
National Ocaan Sarvica
Of flea at Ocaan flaaourcaa Conaarvation and Asaaaamant
Hazaraoua Matanaia Raaoonaa and Asaaaamant Division
Coastal Raaouicas Cooratnaaon Brancn
290 Broaaway, flm 1831
Naw York. New Ycric 10007
DF-
June 3,1997
)oug Tomctak
J.S. EPA
imageucy and Remedial Response Division
ledimMt Projects/Caribbean Team
:90 Broadway
'Jew York. NY 10007
^ear Doug:
fhanir you for the opportunity to review the February 1997 Phase 2 Report, Further Site
Zharaccnzarion and Analysis. Volume 2C - Data Evaluation and Interpretation Report (DE!R) far
he Hudson River PCB Reassessment Remedial Invcsrigarinn/Ffflsihility Study (RI/FSV Hie
olio wing nwimww are submitted by the National Oceanic and Atmospheric AriminiMrarim
NOAA).
Summary
rhe Phase 2 DEIR Report was prepared as part of the overall Phase 2 Reassessment RI/FS
icaviucs currendy ongoing to provide further characterizanon and analysis of the Hudson River
PCB Site which extends from Hudson Falls. NY to the Battery in New York Harbor. The
Reassessment BJJFS Work Plan, completed in September 1992. identified various data collection
ictivitics to support the reassessment effort. The February 1997 document presents gcochmrical
analyses of water column anH sediment data collected during the Phase 2 asseMinirui and data from
other sources including New York State Department of Environmental Conservation (NYSDEC),
United States Geological Survey (USGS") and General Electric (GE).
The Phase 2 objectives were as follows: 1) estimate the current and recent PCB source
contributions exclusive of the Upper Hudson River wtimpwrt 2) characterize-the sources,
movement and distribution of water column and wtirrn-m avjodnTrri PCBs. and 3) rxaminr PCB
distrihurinn anri invunmry within the ITpp-r Wnriwi w44rrn-nr»'
Gtnsra! rnmnKntt .
NOAA commends the authors of this report for a generally well thougnr-out site chaiaarnmnon
and analysis effort. Overall, the report co vexed appropriate subjects and addressed them in a
credible manner. The anthrax
-------
and the P (Public Interest Group or Individual) category for the specific name of an interest group
or his/her own name.
3.2 Location of Responses to Comments
The Comment Directory, following this text, contains a complete listing of all commentors
and comments. This directory allows readers to find responses to comments and provides several
items of information. In several cases, the name of the agency or organization of the commentors
has been abbreviated, as follows:
- NOAA National Oceanic and Atmospheric Adiministration
- USGS United States Geological Survey
- NYSDEC New York State Department of Environmental Conservation
- SC EMC/GLC Saratoga County Environmental Management Council/Govermental
Liaison Committee
- S&T Committee Science and Technology Committee
- ALG Agricultural Liaison Group
- ELG Environmental Liaison Group
- PELC Pace Environmental Litigation Clinic, Inc.
- GE General Electric Corporation
• The first column lists the names of commentors. Comments are grouped first by: F
(Federal). S (State). L (Local). C (CIP), P (Public Interest Group or Individual) or G
(General Electric) preceded by a P or a D for PMCR or DEIR Report, respectively.
Within each of these groups, commentors' names are listed alphabetically.
• The second column identifies the alphanumeric comment code, e g., DF.1-1. assigned
to each comment.
• The third column identifies the location of the response by Report section number. For
example, comments raised on Section 3.2 of a Report can be found in the corresponding
Section 3.2 of the Responses, following the second tab of this document.
The fourth, fifth, and sixth columns list key words that describe the subject matter of
each comment. Readers will find these key words helpful as a means to identify subjects
of interest and related comments.
Responses are grouped and consolidated by section number in order that all responses to
related comments appear together to help achieve consistency among the responses and for the
convenience of the reader interested in responses to related or similar comments.
In a few instances, several commentors commented on the same or very similar items. These
comments are answered by one common response that addresses the common issue being raised.
Thus, a comment is not necessarily answered by an individualized response.
December 22 1998
CD-9
TAMS/l.Tl /TetraTech/MCA
-------
In other cases, closely related but somewhat different comments pertaining to the same report
section are made. Thus, a section number may contain more than one response.
3.3 Types of Responses
Responses to comments include the types described below.
• General Responses
In some instances, comments were general and pertained to the Reassessment process
or the Report overall rather than to a specific section of it. Responses to these
comments are coded as General and appear at the very beginning of the Responses,
under the heading General.
• Specific Responses to Comments
These comments are answered in the Responses, grouped by section number of the
Report to which they refer. A common response is provided when commentors
question the same or very similar items. In some cases, commentors voiced opposite
opinions about the same point, typically a controversial one, but both comments took
issue with the same part of the Report. The rationale for the report's findings or
resolution of the issue may also be contained in a common response addressing the
conflicting nature of the comments and the controversy surrounding the issue.
No separate section is provided for Responses to Comments on Executive Summary.
These comments are answered in one of two ways. When a comment referencing the
Executive Summary was specific or technical or dealt with information contained in
a specific section of the Report, it is addressed in this document under the appropriate
technical section as identified in the Comment Directory. When the comment
concerned the wording of the Executive Summary or dealt with the overall nature of
either, it is addressed under the heading Executive Summary, as appropriate. In all
cases, the Comment Directory refers the reader to the location of the response.
Additional References
Full citations are provided only for new references not previously listed in the
References Sections of the Reports. These citations are provided at the end of the
responses for each report; e.g., new references provided in comment responses to the
DEIR appear following the responses to the DEIR. but within the DEIR tab section.
• Corrections
Corrections to the text are noted in the appropriate report section. No subsequent
action will be taken since the reports will not be reissued.
December 22. 1998
CD-10
T AMS/LTI/T etraT ech/MCA
-------
4. COMMENT DIRECTORY
A Comment Guide, a sample comment letter, a diagram illustrating how to find responses
to comments, and the Comment Directory follow.
As stated in the preface to this Responsiveness Summary, this document does not reproduce
the three reports. Readers are urged to utilize this Responsiveness Summary in conjunction with the
three reports for which comments were received.
4.1 GUIDE TO COMMENT DIRECTORY
RESPONSIVENESS SUMMARY
Step 1
Step 2
Step 3
Find the commentor or the key
words of interest in the
Comment Directory. Comments
are separated by report and
commentor group
Obtain Comment Codes and
Report Section. Find coded
comments following the
COMMENT tab.
Find the responses following the
Responses tab. See the table of
contents to locate the page of the
Responsiveness Summary for the
Report Section.
Key to Comment Codes:
Comment codes are in this format XY-a.b
X=Report (DB=Database Report. D=DEIR, P=PMCR)
Y=Commentor Group
(F=Federal, S=State, L=Local, C=Community Interaction Program, P=Public Interest Group or
Individuals, G=General Electric)
a=Letter or report containing comments
b=Numbered comment
Example:
Comment Response Assignment for the DEIR
AGENCY/
Comment
REPORT
KEY WORDS
Name
CODE
SECTION
1
2
J
NOAA. Rosman DF-2.1 Appendix A, 5.4 Data Quality
Find comment under tab "Federal (DF)".
Find response under tab "Response (DEIR)" page DEIR- 82 where comments relating to Appendix A are
discussed.
December 22. 1998
CD - 11
TAMS/l.TI/TetraTech/MCA
-------
THIS PAGE LEFT BLANK INTENTIONALLY
December 22, 1998 CD-12 TAMS/LTl.'TeiraTech/MCA
-------
Comment Directory
-------
4.2 Comment Directory for the Database Report and Database
AGENCY/
COMMENT
REPORT
KEY WORDS
Name
CODE
section
1
2
3
GE, Haggard
DB-1.1
General
Content
GE, Haggard
DB-1.2
General
Releases
GE, Haggard
DB-1.3
General
Administrative
Record
December 22. 1998
CD- 13
TAMS/LTI/TetraTech/MCA
-------
4.3 Comment Directory for the PMCR
! AGENCY/
Comment
CODE
REPORT
SECTION
KEY WORDS
Name
1
2
3
NOAA. Rosman
PF- 1.1
1.1
Fish
Bans
Advisories
NOAA, Rosman
PF- 12
2.2.1
Model
Segment average
Predicted
values
NOAA, Rosman
PF- 1.3
2.2.1
Data
Solids
Flow
NOAA. Rosman
PF- 1.4
2.2.1
Congeners
Values
Total PCBs
NOAA. Rosman
PF- 1.5
2.2.3
Model
Fish
Tissue
NOAA, Rosman
PF- 1.6
2.2.3
Model
Pathways
Bullhead
NOAA, Rosman
PF- 1.7
2.2.3
Data
External Loads
Lower Hudson
NOAA. Rosman
PF- 1.8
2.2.4
Model
Striped Bass
PCB Uptake
NOAA. Rosman
PF- 1.9
2.2.4
Revision
Editorial
NOAA. Rosman
PF- 1.10
3.7.1
Model
Sediment Cores
Scour
NOAA, Rosman
PF- 1.1 1
3.8.1
Thomann
- model
Food Chain
Revison
NOAA. Rosman
PF- 1.12
4.1
Congeners
Food Chain
Properties
NOAA. Rosman
PF- 1.13
4.1
Sediment
Water Column
Location
NOAA, Rosman
PF- 1.14
4.9. Figure 4.5
Revision
Editorial
NOAA, Rosman
PF- 1.15
4.3
Model
Data set
Values
NOAA. Rosman
PF- 1.16
4.4.2
Model
Data set
TSS, flow
NOAA, Rosman
PF- 1.17
4.4.2
Data
Loading
USGS
NOAA, Rosman
PF- 1.18
4.4.2
Model
Mass Balance
Averaging
NOAA, Rosman
PF- 1.19
4.4.2 par 1 & Fig
4.46(b)
Revision
Editorial
NOAA. Rosman
PF- 1.20
4.4.2 par 3
Revision
Editorial
NOAA. Rosman
PF- 1.21
4.4.2 par 4
Data
Correlation
Coefficients
NOAA, Rosman
PF- 1.22
4.4.2
Data
Outlier
Exclusion
NOAA. Rosman
PF- 1.23 | 4.4.2
Data
Concentration
Derivation
NOAA. Rosman
PF- 1 24 1 4.4.2
Load
Sources
Percentages
NOAA, Rosman
PF- 1 25
4.4.3
Model
Data
Atmosphere
NOAA. Rosman
PF- 1.26
4.4.3
Model
Water Column
Stratification
NOAA. Rosman
PF- 1.27 , 4.4.5
Model
Water Column
Stratification
NOAA. Rosman
PF- 1.28 | 4.4.5
GE 1991 Data
Sediment
Average
NOAA, Rosman
PF- 1 29
4.4.5
Density
Basis
Dry weight
NOAA. Rosman PF-1.30
4.4.5
Model
Sediment
Congener
NOAA. Rosman
PF-1.31 | 4.5.2
Model
Parameters
Values
NOAA. Rosman
. PF- 1.32 : 6.5
! Model
Scour
Depth
NOAA. Rosman PF- 1.33
i
1 !
7.1
Lower HRj Updated
Model | Thomann
December 22. 1998
I'D - 14
TAMS/LTl/TelraTech/MC A
-------
AGENCY/
Name
Comment
CODE
REPORT
SECTION
KEY WORDS
1
2
3
NOAA, Rosman
PF- 1.34
7.2.3
Migration
Striped Bass
References
NOAA, Rosman
PF- 1.35
7.6.2
Revision
Editorial
Sensitivity'
NOAA, Rosman
PF- 1.36
7.7
Model
Thomann
Marking
NOAA, Rosman
PF- 1.37
8.5.5
Model
Water Column
Carbon
NOAA, Rosman
PF- 1.38
9.1.5, Table 9-2
Model
Concentration
Weight Basis
NOAA, Rosman
PF- 1.39
9.1.5, Table 9-7
Data
Accuracy
R squared
NOAA, Rosman
PF- 1.40
9.2, Figures 9-11 and
9-12
Revision
Editorial
Axis label
NOAA, Rosman
PF- 1.41
9.3, Figures 9-8 thru
9-13
Revision
Editorial
Regression line
NOAA, Rosman
PF- 1.42
9.3
Model
Fish burden
Outputs
NOAA, Rosman
PF- 1.43
9.3
Revision
Data
NOAA, Rosman
PF- 1.44
9.4
Reference
Sex
Differences
NOAA, Rosman
PF- 1.45
10.1.3
Data
Water Column
Chironomid
NOAA, Rosman
PF- 1.46
10.2
Tissue
Sediment
Lipid
NOAA, Rosman
PF- 1.47
10.2.2
Data
BSAF
Calculations
NOAA, Rosman
PF- 1.48
10.2.2
Model
Fish
Diet
NOAA, Rosman
PF- 1.49
10.2.3
Model
Percentiles
Clarification
NOAA, Rosman
PF- 1.50
10.2.3
Model
Percentile
Interpretation
NOAA, Rosman
PF- 1.51
10.2.3
Revision
Data
Means error
NOAA, Rosman
PF- 1.52
10.2.3
Chronomids
Feeding
Habits
NOAA, Rosman
PF- 1.53
10.2.3
Data
Sediment
Biota, Format
NOAA, Rosman
PF- 1.54
10.2.3
Data
Organisms
Feeding
NOAA, Rosman
PF- 1.55
10.4.1
Data
Fish
Size
NOAA, Rosman
PF- 1.56
10.4.3
Data
Fish
Diet
NOAA. Rosman
PF- 1.57
10.4.3
Sampling
Fish
Composites
NOAA, Rosman
PF- 1.58
10.4.4
Revision
Editorial
NOAA, Rosman
PF- 1.59
A- 1.3.2
Revision
Editorial
Units
NOAA, Rosman
PF- 1.60
A- 1.10.4
Revision
Spottail Shiner
Diet
NOAA, Rosman
PF- 1.61
A-1.10.5
Data
Fish
Feeding
NOAA, Rosman
PF- 1.62
A- 1.10.5
Model
BAFs
Water Column
NOAA. Rosman
PF- 1.63
A- 1.10.5
Revision
Forage Fish
Diet
SC EMC/GLC,
Adams
PL- 1.1
General
Report
Format
Organization
SC EMC/GLC,
Adams
PL- 1.2
General
Report
Content
Concept
Summary
SC EMC/GLC. j PL-1.3
Adams 1
3.5.2
Model
HUDTOX
Relationships
SC EMC/GLC.
Adams
PL- 1.4 | 8.2.1
1
Model i Congeners
1
Fish
December 22. 1998 CD -15 TAMS/LTI/TetraTech/MCA
-------
AGENCY/
Name
Comment
CODE
REPORT
SECTION
KEY WORDS
1
2
3
SC EMC/GLC,
Adams
PL- 1.5
8.3
Model
Fish
Burden
SC EMC/GLC,
Adams
PL- 1.6
3.5
Model
Appropriateness
Upper
Hudson
SC EMC/GLC,
Adams
PL- 1.7
3.1, Figure 3-1
Symbols
Explanation
SC EMC/GLC,
Adams
PL- 1.8
3.5
Submodel
Rationale
Carbon
SC EMC/GLC,
Adams
PL- 1.9
3.5.2
Submodel
Database
Explanation
SC EMC/GLC,
Adams
PL- 1.10
3.5.2
Submodel
Access
Application
SC EMC/GLC,
Adams
PL- 1.11
3.71
Model
Scour
Reassessment
SC EMC/GLC,
Adams
PL- 1.12
4.4.2
Data
Air
Volatilization
SC EMC/GLC,
Adams
PL- 1.13
4.4.2
Tables
Sediment
Concentration
SC EMC/GLC,
Adams
PL- 1.14
4.7
Data
Explanation
Settling
velocity
SC EMC/GLC.
Adams
PL- 1.15
4.7
Data
Explanation
Average
velocity
SC EMC/GLC,
Adams
PL- 1.16
4.5, Table 4.10
Data
Explanation
Ux
SC EMC/GLC,
Adams
PL- 1.17
Chapter 4 Tables,
Tables 4-13 thru 4-17
Data
Variance
Statistics
SC EMC/GLC,
Adams
PL- 1.18
' 7.2
Data
Dispersion
Coefficient
SC EMC/GLC,
Adams
PL- 1.19
10.4.4
Data
FFBAF
Equations
STC/Putman
PC- 1.1
4.0
Solids
TSS
Resusupension
STC/Putman
PC- 1.2
4.7
Model
Loading
Resuspension
STC/Putman
PC-1.3 | 4.2
Model
Transport
Movement
STC/Putman
PC- 1.4
4.4
Model
Flow
Concentration
STC/Putman
PC- 1.5
4.4
Data
References
Mass
STC/Putman
PC- 1.6
6.2
Model
Scour
Resuspension
!
Sanders. John
PP-l.A | 4.0
Model
Time
Date
Sanders. John
PP- I B
4.0
Data
Time
Date
Sanders. John
PP- l .C
General
Data [Sediment
Sampling
Sanders. John
PP- 1.1
General Revision
Editorial
Hyphen
'998
CD- 16
T AMS/LTI/TetraT ech/MC A
-------
AGENCY/ ! Comment
Name j CODE
i |
REPORT
SECTION
KEY WORDS
1
2 3
Sanders. John 1 PP- 1.2
General
Revis
on
Editorial
Hyphen
Sanders. John PP- 1.3 General
Rev is
on
Editorial | Hyphen
Sanders. John
PP- 1.4 j General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.5
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.6
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.7
General
Revis
on
Editorial
Hyphen
Sanders, John
PP- 1.8
General
Revis
on
Editorial
Hyphen
Sanders, John
PP- 1.9
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.10
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.1 1
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.12
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.13
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.14
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.15
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.16
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.17
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.18
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 119
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.20
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.21
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.22
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.23
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.24
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.25
General
Revis
on
Editorial
Hyphen
Sanders. John | PP- 1.26
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.27
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.28
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.29
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.30
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.3 I
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.32
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.33
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.34 1 General
Revis
on
Editorial
Hyphen
Sanders. John ¦ PP- 1.35
General
Revis
on
Editorial
Hyphen
Sanders. John , PP- 1.36
General
Revis
on
Editorial
Hyphen
Sanders. John
PP- 1.37 General
Revis
on
Editorial
Hyphen
Sanders. John i PP- 1.38 ' General
i
Revis
on Editorial
Hyphen
Sanders. John ' PP- 1.39 i General
Revis
on I Editorial
Hyphen
Sanders. John ! PP- 1.40
General
Revis
on | Editorial | Hyphen
Sanders. John i PP-1.41
General
Revis
on Editorial Hyphen
December 22 1998 CD-17 TAMS/LTI/Tctrafech/MCA
-------
AGENCY/
Name
Comment
CODE
REPORT
SECTION
KEY WORDS j
1
2
3
Sanders, John
PP- 1.42
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.43
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.44
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.45
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.46
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.47
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.48
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.49
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.50
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.51
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.52
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.53
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.54
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.55
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.56
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.57
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.58
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.59
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.60
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.61
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.62
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.63
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.64
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.65
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.66
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.67
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.68
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.69
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.70
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.71
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.72 i General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.73 | General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.74
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.75
General
Revision
Editorial
Comma
Sanders. John
PP- 1.76
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.77
General
Revision
Editorial
Comma
Sanders. John
PP- 1.78
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.79
General : Revision j Editorial
Hyphen
Sanders. John i PP-1.80 | General jRevision
'Editorial
Hyphen
Sanders. John 1 PP- 1.81 ! General j Revision
Editorial
Comma
December 22. 1998 CD-18 TAMS/LTI/Tetra Tech/MCA
-------
AGENCY/
Name
Comment i REPORT
CODE j SECTION
KEY WORDS
1
1
2
3
Sanders, John
PP- 1.82 General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.83
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.84
General
Revision
Editorial
Comma
Sanders, John
PP- 1.85
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.86
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.87
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.88
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.89
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.90
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.91
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.92
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.93
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.94
General
Aroclors
Homologues
PCBs
Sanders. John
PP- 1.95
General
Revision
Editorial
Style
Sanders. John
PP- 1.96
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.97
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.97A
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.97B
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.97C
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.98
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.99
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.100
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.101
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.102
General
Revision
Editorial
Comma
Sanders. John
PP- 1.103
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.104
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.105
General
Revision
Editorial
Style
Sanders. John
PP- 1.106
General
Revision
Editorial
Style
Sanders, John
PP- 1.107
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.108
General
Revision
Editorial
Comma
Sanders. John
PP- 1.109
Executive Summary
Model
HUDTOX
Aliens Mill
Sanders. John
PP- 1.1 10
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.111 ' General
Revision
Editorial
Style
Sanders. John
PP- 1.112 Executive Summary
Sediment
Congener
Finger print
Sanders. John
PP-1.113 1 General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.1 14A | General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.1 14B1 General
Revision i Editorial
Hyphen
Sanders. John
PP- 1.115 : General
Revision 'Editorial
Hyphen
Sanders. John , PP- I.I 16 j General
Revision [Editorial
Hyphen
December 22. I
-------
AGENCY/ j Comment
REPORT
KEY WORDS |
Name j CODE
SECTION
1
2
3
Sanders, John
PP- 1.117
General
Revision
Editorial
Style
Sanders, John
PP- 1.118
General
Revision
Editorial
hyphen
Sanders. John
PP- 1.119
General
Revision
Editorial
hyphen
Sanders, John
PP- 1.120
General
Revision
Editorial
-lyphen
Sanders, John
PP- 1.121
General
Revision
Editorial
4yphen
Sanders, John
PP- 1.122
General
Revision
Editorial
Style
Sanders, John
PP- 1.123
General
Revision
Editorial
lyphen
Sanders, John
PP- 1.124
General
Revision
Editorial
lyphen
Sanders, John
PP- 1.125
General
Revision
Editorial
lyphen
Sanders, John
PP- 1.126
General
Revision
Editorial
typhen
Sanders, John
PP- 1.127
General
Revision
Editorial
-typhen
Sanders. John
PP- 1.128
General
Revision
Editorial
hyphen
Sanders, John
PP- 1.129
General
Revision
Editorial
hyphen
Sanders. John
PP- 1.130
General
Revision
Editorial
Comma
Sanders, John
PP- 1.131
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.132
General
Revision
Editorial
Hyphen
Sanders, John
PP- 1.133
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.134
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.135
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.136
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.137
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.138
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.139
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.140
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.141
General
Revision
Editorial
Comma
Sanders. John
PP- 1.142
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.143
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.144
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.145
1.1
Survey
NYSDEC
1984 FS
Sanders. John
PP- 1.146
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.147
General
PCBs
Homologues
Aroclors
Sanders. John
PP- 1.148
General
PCBs
Homologues
Aroclors
Sanders. John
PP- 1.149
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.150
General
Revision
Editorial
Style
Sanders. John
PP- 1.151
General
Revision
Editorial
Hyphen
Sanders. John
PP-1.152
General
Revision
Editorial
1
Style
Sanders. John
PP- 1.153
General
Revision Editorial
Style
Sanders. John j PP- 1.154
General j Revision
Editorial
Style
December 22. 1W8 CD - 20 TAMS/LTI/Tetralcch/MCA
-------
AGENCY/
1 Name
Comment
CODE
REPORT
SECTION
KEY WORDS
1
2
3
Sanders. John
PP- 1.155
General
Revision
Editorial
Style
Sanders. John
PP- 1.156
General
Revision
Editorial
Style
Sanders. John
PP- 1.157
General
Revision
Editorial
Hyphen
Sanders. John
PP- 1.158
General
Revision
Editorial
Spelling
Sanders, John
PP- 1.159
General
Revision
Editorial
Spelling
GE
PG- 1.1
4.6
Fate
Transport
Solids balance
GE
PG- 1.2
3.5
Mass Balance
Transfer
PCB balance
GE
PG- 1.3
3.5
PCB Balance
Conditions
Model
GE
PG- 1.4
4.7.2
Mechanisms
Groundwater
TIP
GE
PG- 1.5
3.1
Data
Calibration
Model
GE
PG- 1.6
8.1
Steady State
Predictions
Model
GE
PG- 1.7
3.7
Scour
100 year flood
Lick equation
GE
PG- 1.8
General
Data
Predictions
Testing
GE
PG- 1.9
7.1
Thomann
Update
Use
Additional GE Material posing no direct questions but for which responses
are appropriate.
GE
PG- 1.10
6.1
Model
Evaluation
Criteria
GE
PG- 1.1 1A
4.3
PCB
Mass
Balance
GE
PG-1.11B
4.7, 4.7.2
Model
Calibration
Groundwater
GE
PG- I.I 1C
4.7
Model
Calibration
Solids
GE
PG-1.11D
4.7
Model
Calibration
Sediment
GE
PG- 1.11E
4.7. 4.7.2
Model
Calibration
Resuspension
GE
PG- 1.11F
4.7
Model
Calibration
Variation
GE
PG- 1.11G
4.7
Model
Calibration
Data
GE
PG- 1.1 1H
4.7
Model
Dechlorination
Biodegradation
GE
PG- 1.111
4.7
Model
Calibration
Results
GE
PG- 1.12
8.1
GE
PG- 1.13B
5.9.1
Shear stress
Hydrodvnamic
RMA-2V
GE
PG- 1.13C
6.3
Lick Equation
Resuspension
Densities
GE
PG- 1.13D
6.3
Scour Model
Transport Model
Inconsistency
GE
PG- 1.13E
6.5
Scour Model
Sediment
Non-cohesive
GE
PG- 1.13F | 6.5
Scour Model
Non-cohesive
Erosion
GE
PG- 1.14 ! 7.7
Model
Predictive
Validation
GE
PG- 1.15 ; 7.0
Upper Hudson
Remediation ; Impacts
GF.
PG- 1.16 1 7.7
Model (Limitations j
December 22. 1998
CD - 21
TAMS/LTl/TetraTcch/MCA
-------
4.4 Comment Directory for the DEIR
AGENCY/
Name
Comment
CODE
REPORT
SECTION
KEY WORDS
1
2
3
NOAA. Rosman
DF-2.1
Appendix A, 5.4
Data
Quality
NOAA, Rosman
DF-2.2A
Appendix A
Data
Quality
Compositions
NOAA, Rosman
DF-2.2B
Appendix A
Data
Quality
BZ#44
NOAA, Rosman
DF-2.2C
Appendix A
Data
Quality
Risks
NOAA, Rosman
DF-2.3A
3
Discussion
Congener
Loading
NOAA, Rosman
DF-2.3B
3
Discussion
Congener
BZ#118
NOAA, Rosman
DF-2.4
3.3.3
Analyses
Congener
Concentration
NOAA, Rosman
DF-2.5
3.4.3
Discussion
Congener
Weathering
NOAA, Rosman
DF-2.6
4
Data
Interpretation
Use
NOAA, Rosman
DF-2.7
2.2.2
Analysis
Independence
Sources
NOAA, Rosman
DF-2.8
2.3
Patterns
Summer
Load
NOAA, Rosman
DF-2.9
2.3
Collections
Dynamics
Deposits
USGS, Pearsall
DF-3.1
3.3
Data
Bias
USGS, Pearsall
DF-3.2
3.4
Revision
Editorial
USGS, Pearsall
DF-3.3
3.3.5
Data
Homologue
Water column
USGS, Pearsall
DF-3.4
3.2.3
Definition
Transition
Flow
USGS, Pearsall
DF-3.5
3.2.3
Data
Source
USGS. Pearsall
DF-3.6
3.2.3
Data
Collection
Time
USGS, Pearsall
DF-3.7
3.2.3
Figures
Information
USGS, Pearsall
DF-3.8
3.4.3
Load
Time
USGS, Pearsall
DF-3.9
3.4.6
Model
Thomann
Water column
NYSDEC. Sterman
DS-2.1
2.2 1
Sediments
Fish
Site
NYSDEC. Sterman
DS-2.2
2.2.1
Seepage
Rate
Usage
NYSDEC, Sterman
DS-2.3
2.2.1
Dump
Listing
Existing
NYSDEC, Sterman
DS-2.4
2.2.2
Remediation
Consideration
Feasibility
Study
NYSDEC, Sterman
DS-2.5
3.2.6
Process
Load
Water column
NYSDEC. Sterman
DS-2.6
3.2.6
Loads
Water column
Flow
NYSDEC. Sterman
DS-2.7
3.2.6
Load
Prediction
Fate
NYSDEC. Sterman
DS-2.8
3.2.6
Processes
Exchange
Load
NYSDEC. Sterman
DS-2.9
3.2.6
Load
Sediments
Influence
NYSDEC. Sterman
DS-2.10
3.2.7
Sediments
Scour
Hoosic River
NYSDEC. Sterman j DS-2.11
3.2.7
Load
Increase
Spring
NYSDEC. Sterman ' DS-2.12 i 3.3.1 ISediments
Cores
Chronology
NYSDEC. Sterman 1 DS-2.13 ' 3.4.1 iData
Water column
i Representative
NYSDEC. Sterman DS-2.14 i 3.4.3 Load
Storage j TIP
December 22. 1998 CD - 22 TAMS/LTl/TetraTech/MCA
-------
AGENCY/ j Comment
REPORT
SECTION
KEY WORDS
Name
CODE
1
3
NYSDEC. Sterman
DS-2.15
3.4.4
Sewer
overflow
Problem
Allen Mill
NYSDEC, Sterman DS-2.16
3.4.4
Loading
Interpretation
Relation
NYSDEC, Sterman
DS-2.17
3.4.4
Sediments
Source
Water column
NYSDEC, Sterman
DS-2.18
3.4.4
Comparison
Load
Basis
NYSDEC. Sterman
DS-2.19
3.6
Sediments
Age
NYSDEC, Sterman
DS-2.20
3.6
Sediments
Load
Dechlorination
NYSDEC, Sterman
DS-2.21
4.4
Sediment
Water column
Exchange
NYSDEC, Sterman
DS-2.22
4.4
Loads
Quantify
NYSDEC, Sterman
DS-2.23
4.4
Loads
Sources
Alteration
NYSDEC, Sterman
DS-2.24
4.4
Load
Duration
Stability
NYSDEC, Sterman
DS-2.25
4.4
Load
Porewater
Exchange
NYSDEC, Sterman
DS-2.26
4.4
Remediation
Remedy
Controls
NYSDEC, Sterman
DS-2.27
General
Data
Model
Verify
NYSDEC, Sterman
DS-2.28
General
Data
Monitoring
Long term
NYSDEC, Sterman
DS-2.29
General
Data
Core
Incorporation
NYSDEC, Sterman
DS-2.30
General
Data
Sediments
Impact
SCEMC&GLC, Balet
DL-1.1
3.4.3
Data
Loading
Suspect
SCEMC&GLC. Balet
DL-1.2
3.4.5
Data
Sediment
Hot spots
SCEMC&GLC, Balet
DL-1.3
3.2.3
Sediments
Time
Hot spots
SCEMC&GLC, Balet
DL-1.4
4.5
Scenario
Sediment
Transport
SCEMC&GLC, Balet
DL-1.5
General
Data
Water column
Concentration!
SCEMC&GLC, Balet
DL-1.6
4.5
Evaluate
Inputs
Source
SCEMC&GLC. Balet
DL-1.7 4.3.1
Significance
Chlorine
Health Risk
SCEMC&GLC, Balet
DL-1.8 3.3.3
Input
Congener
Fingerprinting
i
|
Agriculture Liaison
Group, Borden
DC-1.1 General
Participation
Public meetings
Agriculture Liaison
Group, Borden
DC-1.2 General
Conclusions
Reports
Schedule
Agriculture Liaison
Group, Borden
DC-1.3 ¦ General
Comments
Opinions
Parties
Agriculture Liaison
Group, Borden
DC-1.4 ! 3.4.2 iSources
1
Ignored
i
Agriculture Liaison
Group, Borden
DC-1.5 1.3
1
Data
Reference
Time
.
Agriculture Liaison
Group. Pulver
DC-2.1 General IDredizine Objection
i ~ ~ i
i Dairy
;
December 22. 1998
CD - 23
TAMS/LTI/Telra lech/MC A
-------
AGENCY/
Name
Comment
CODE
REPORT
SECTION
KEY WORDS
1
2
3
Environmental Liason
Group, Ruggi
DC-3.1
4.5
Sediments
Dechlorination
Quantification
Environmental Liason
Group, Ruggi
DC-3.2
4.5
!
Theory
Flux
Groundwater
S&T Committee,
Putman
DC-4.1
3.3.2
Concentrations
Flow
Cores
S&T Committee,
Putman
DC-4.2
3.2.3
Sedimentation
Rate
Scour
S&T Committee,
Putman
DC-4.3
3.2.3
Loading
Groundwater
Discharge
S&T Committee,
Putman
DC-4.4
Executive Summary
Loading
Increase
Hot spots
S&T Committee,
Putman
DC-4.5
3.2.7
Data
Concentrations
Water samples
S&T Committee,
Putman
DC-4.6
1.4.2
Data
Sampling
Unreliable
S&T Committee,
Putman
DC-4.7
3.2.4
Discharge
Calibration
Error
Clearwater, Mele
DP-1.1
Executive Summary
Inventory
Depletion
Time
Davies, Sherwood
DP-2.1
General
Exposure
T ransport
Water source
Davies, Sherwood
DP-2.2
General
Ozone
By-products
Toxicity
Davies, Sherwood
DP-2.3
General
Risk
Infiltration
Water source
Davies, Sherwood
DP-3.I
Executive Summary
Sediments
Porewater
Migration
Pace Environ.
Litigation Clinic.
Boehlje
DP-4
General
j
i
Scenic Hudson, Lee j DP-5 General
! i
GE ' DG-1.1 Executive Summary.
! P-2
Water
Column
TI Dam
Pipeline
GE | DG-1.2
Executive Summary,
p. 3
Load
Homologue
Water column
GE j DG-1.3
j
1
Executive Summary,
j p. 4
Load j Source
Vague
GE DG-1.4
i
Executive Summary,
pp. 5-8
Sediment i Dechlorination
Toxicity
December 22. 1998
CD-24
TAMS/LTI/TctraTech/MCA
-------
10.0044
AGENCY/
Comment
: REPORT
KEY WORDS
Name
CODE
SECTION
1
2
3
GE
DG-1.4A
3.2.7
PCB
Fate
Low flow
GE
DG-1 4B
3.2.7
Fate
Volatilization
Deposition
GE
DG-I.4C
3.2.7
Inconsistency
Reaches
Treatment
GE
DG-1.4D
3.2.7
1993
Conditions
Aliens Mill
Loading
GE
DG-1.4E
General
PCB
Behavior
Upper Hudson
GE
DG-1.5
4.4
Diffusion
Load
Increase
GE
DG-1.6
4.4
Resuspension
Sediments
pre-1984
GE
DG-I.7
4.4
Resuspension
Dechlorination
Low flow
GE
DG-1.8
4.4
Fingerprint
Dechlorination
Source
GE
DG-1.9
3.2.6, p. 3-76
Discharge
Load
Undechlorinatio
n
GE
DG-1.10
3.2.6
Source
Reduction
Control
GE
DG-1.10A
3.2.6
Source
High Flow
GE
DG-1.11
3.2.6
Load
Balance
Calculation
GE
DG-1.12
3.3.4
Fingerprint
Load
Undechlorinatio
n
GE
DG-1.13
General
Fingerprint
Fish
Consistency
GE
DG-1.14
3.3.4
Load
Unaccounted
Allen Mill
GE
DG-I.I5
3.2.6
Load
Understated
GE
DG-1.15A
3.2.6
GE
DG-1.15B
3.2.6
GE
DG-1.15C
3.2.6
GE
DG-1.15D
3.2.6
GE
DG-1.16
3.2.6
Volatization
Deposition
Decrease
GE
DG-1.17
3.3.3
Load
Sources
External
GE
DG-1.18
4.3
Mechanism
Dechlorination
Reduction
GE
DG-1.19
4.3
Report
Dechlorination
Indices
GE
DG-1.20
4.3.2. p. 4-69
Dechlorination
Concentration
GE
DG-1.21
4.3
Dechlorination
Cessation
GE
DG-1.22
3.3.3. p. 3-1 19
Dechlorination
Pattern
H/H'
GE
DG-1.23
3.1
Partitioning
Kp
Temperature
GE
DG-1.24
3.2.3. p. 3-55
Volatilization
Low Flow
Seasonal
GE
DG-1.25A
Appendix A. 5.4
Analytical
Issues
Congeners
GE
DG-1.25B
Appendix A, 5.4
Analytical
Issues
Congeners
GE
DG-1.25C
Appendix A. 5 2
Analytical
Issues
Extraction
Gas flow
GE :
DG-I26A
4.3. p. 4-49
Remobilization 1 Resuspension
Sediment
December 22 1998
CD - 25
TAMS/LTI/TetraTech/MCA
-------
AGENCY/
Comment
REPORT
KEY WORDS
Name
CODE
SECTION
1
2
3
GE
DG-I.26B
4.3.1. p. 4-50
Microorganisms
Degradation
High mol wt
GE
DG-I.26C
4.3.1, p. 5-50
Anoxic
Biotransformation
Dechlorination
GE
DG-1.26D
4.3.1. p. 4-51
Dechlorination
ortho-chlorine
High mol wt
GE
DG-1.26E
4.3.1, p. 4-51
PCB molecule
Destruction
Total
GE
DG-1.26F
4.3.1, p. 4-51
Carcinogenic
Neurological
Congeners
GE
DG-1.26G
4.3.1, p. 4-51
Anerobic
Degradation
Sediments
GE
DG-I.26H
4.3.1, p. 4-52
Biodegradation
Prodcucts
Sediments
GE
DG-I.26I
4.5, p. 4-88
Concentration
Dependence
Dechlorination
GE
DG-I.26J
4.3.2, p. 4-52
Dechlorination
Sources
Basis
GE
DG-1.26J1
4.3.2, p.4-56
GE
DG-I.26K
4.3.2, p. 4-56
BZ#8
Indicator
MDPR
GE
DG-1.26L
4.3.2, p. 4-57
ortho-chlorine
Dechlorination
Limits
GE
DG-1 26M
4.3.2, p. 4-59
Dechlorination
Processes
H/H'
GE
DG-1.26N
4.3.2, p. 4-60
MDPR
Sensitivity
Congeners
GE
DG-1.260
4.3.2, p. 4-62
MDPR
Partitioning
Degradation
GE
DG-1 26P
4.3.2. p. 4-65
Dechlorination
Maximum
Decrease
GE
DG-1.26Q
4.3.2, p. 4-69
delta-MW
Concentration
Dependence
GE
DG-1.26R
4.3.2. p. 4-70
Sediments
Age
Dechlorination
December 22. 1998
C D • 26
TAMS/L Tl/TetraTcch/MC'A
-------
Responses (DATABASE)
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY
VOLUME A: DATABASE REPORT
DECEMBER 1998
A. DATABASE REPORT AND DATABASE:
Responses to Comments on the Database Report
No comments were received on the Database Report per se. However, General Electric did provide
comments on the database itself, as issued in the April 1996 CD-ROM version, in their letter dated
May 29, 1996.
Response to DB-1.1
GE identified the data sets identified below as missing from the database (copied from GE
Attachment 1).
LOCATION/
PROGRAM
ORIGIN
MEDIA
PARAMETER
CURRENT
FORM
STATUS
Lower River,
New York Harbor,
Long Island Sound
GE/Harza
1988 - 1991
Sediment
Biota
PCBs
Pesticides
lipids
dBase IV files
Partially
included
TIP TSS Survey
GE/OBG
1991
Water
TSS
report,
dBase IV files
Included
Polygon (3)
GE/OBG
1990
Sediment
PCB
dBase IV
Not
included
EPA Lower River
Helicopter Survey
EPA
1976
Sediment
PCB
EPA Report
Not
included
EPA Lower River
Survey
EPA
1981
Sediment
PCB
EPA Report
Not
included
NYU Lower River
Biota
NYU
' pre-1982
Biota
PCB
NYU Report
Not
included
General Electric data updates received from O'Brien and Gere (OBG) are regularly
incorporated into the Hudson River Database. The data structure is modified to conform to a
relational database structure, but data points are not omitted. The latest release of the database
(Release 4.1, dated August 1998) included the data received on 7/28/98 from O'Brien and Gere. As
stated in OBG's transmittal letter, GE/Harza data from 1988-1991 have only been partially provided
to USEPA. All GE/Harza data provided in the recent GE database transmittal have been included
in the USEPA database. Sediment and fish data from this sampling event are expected in upcoming
deliverables from OBG. Pesticide data have not been included in the data received from GE/OBG.
There are 651 TSS samples from the 1991 Hudson River PCB Water Survey. Although not
December 22. 1998
DB-1
TAMS/LTLTetraTech/MCA
-------
specifically labeled TIP TSS Survey data, this is believed to be the data referred to in the table.
Polygon (3) data were not specifically identified in the data files supplied by GE. However, if these
data were included in the recent GE database transmittal then they are included in the TAMS'
Hudson River database. The missing data including GE/Harza sediment and biota data, pesticide
data, and Polygon (3) sample data, will be added to the Hudson River Database once received from
General Electric.
The data from last three programs listed in the table are not available in electronic format.
These programs focus on the Lower Hudson River in 1976, 1981 and pre-1982. To date these data
have not been needed to perform the Hudson River PCBs Reassessment. Only if these data are
required for the Reassessment will they be added to the database.
Response to DB-1.2
The Hudson River Database is released whenever significant additions of data or
modifications to data are made. This includes addition of data from General Electric and NYSDEC
as well as other sources.
Response to DB-1.3
The status of each data set identified by GE as missing from the Hudson River Database is
listed in the table on page DB-1. Two of the six data sets have been added to the database in the form
provided by GE. An incomplete set of sediment PCB data from the lower river New York Harbor
and Long Island Sound data is included, but GE did not provided the pesticide data. The TIP TSS
survey data are included in the database and this data set appears to be complete. The Polygon (3)
program data was not provided by GE. Sediment data in the database from 1990 was collected by
Harza Engineering, but the Polygon (3) study was performed by O'Brien and Gere.
The EPA's lower river surveys (1976 and 1981) and NYU's lower river biota study have not
been included in the database. The data from these programs was not required to perform the Hudson
River PCBs Reassessment. These programs focus on the Lower Hudson River in 1976, 1981 and
pre-1982. Only if these data are required for the Reassessment will they be added to the database.
December 22. 1908
DB-2
TAMS/LTLTetraTech/MCA
-------
Responses (PMCR)
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY
VOLUME B: PRELIMINARY MODEL CALIBRATION REPORT
DECEMBER 1998
B: PMCR REPORT
Responses to General Comments on the PMCR Report
Response to PL-1.2:
The Executive Summary included at the front of each Reassessment report is intended to
summarize the information presented in the report in a manner that can be understood by less
technically-trained citizens. It is noted, however, that there is a limit to the extent to which this
rather complex topic can be made non-technical without oversimplifying, and perhaps distorting, the
issues.
Response to PI.-1.1:
The format of the report, which provides the tables and figures as a second book, allows the
reader to have both the text and tables/figures open at the same time, rather than continually having
to leaf back and forth w ithin a single volume. Such a format is standard for all Reassessment reports.
Response to PP-1 1 through PP-1.159:
PP-1 are comments from John E. Sanders. Mr. Sanders provided 159 numbered comments
in an attachment, plus three additional comments in the cover/transmittal letter. For convenience.
Mr. Sanders' numbering has been retained {e.g., his comment 08 is identified herein as "PP-1.8");
with comments in the cover letter identified by letter (e.g., PP-1.A). A large number of the
comments are editorial, reflecting use of hyphens, commas, or wording. These editorial comments
(for which no response is provided) are identified immediately below; technical comments are
provided with a specific response in the appropriate section of this Responsiveness Summary.
Comments on hyphen use: PP-1.1 - 1.63; 1.65 - 1.74; 1.76; 1.78 - 1.80; 1.85 - 1.93; 1.96- 1.101;
1.103; 1.104; 1.107; 1.113- 1.116; 1.118- 1.121; 1.123 - 1.140; 1.142 - 1.144; 1.146; 1.151;and
1.157.
Comments on comma use: PP-1.75. 1.77. 1.81, 1.84. 1.102, 1.1.108. and 1.141.
Comments on spelling, wording, or word use: PP-1.64; 1.95; 1.105: 1.106; 1.111; 1.117; 1.122:
1.150: 1.152 - 1.156: 1.158; and 1.159.
Response to PG-1.8:
The scope of work for the Baseline Modeling and Ecological Modeling efforts
w ill include consideration of the model validation tests suggested by the reviewer.
December 22. 1998
PMCR-1
T AMS/l.Tl/TetraTech/MCA
-------
Responses to Specific Comments on the PMCR Report
EXECUTIVE SUMMARY
Response to PP-1.109:
The additional PCB DNAPL seepage from the Allen Mill site was discovered after the
PMCR was completed; the HUDTOX simulation predates its cleanup.
The Allen Mill gate structure failed in September, 1991. The period of simulation for the
modeling results in the PMCR corresponded only to the 1993 EPA Phase 2 sampling period. It is
possible that loads from this- failure -event and from DNAPL seepage influenced PCB dynamics in
Thompson Island Pool during 1993. The Baseline Modeling effort will include a hindcasting
simulation for the period 1977-1997 to gain a better understanding of the impacts of long-term
changes in PCB loads.
Response to PP-1.112:
USEPA is continuing to evaluate existing water column and sediment data for potential
congener "fingerprints''' which may help constrain the importance of pore-water flux for PCB loading
from sediments to the Hudson River.
The PCB modeling effort includes several individual PCB congeners to assess differences
in physicochemical properties (e.g., partition coefficients) which can result in differential transport
and exchange between sediments and the water column. The Baseline Modeling effort will continue
to examine individual PCB congeners.
December 22. 1998
PMCR-2
TAMS/LTl/TetraTech/MCA
-------
1.
INTRODUCTION
1.1 Background
Response to PP-I.145:
The work described in the PMCR was not intended to assess the historical data; e.g., the 1984
NYSDEC Survey. The focus of the PMCR was on development and preliminary calibration of a
PCB transport and fate model for the 1993 EPA Phase 2 sampling period. Problems with
representation of PCBs among different historical data sets were still unresolved at the time of the
PMCR, More detailed descriptions and analyses of earlier NYSDEC investigations can be found
in the Phase I report (USEPA, 1991), the Data Evaluation and Interpretation Report (USEPA, 1997),
and the Low Resolution Sediment Coring Report (USEPA, 1998). The Baseline Modeling effort
will include historical data from 1977 through 1997, including data from the 1984 NYSDEC Survey.
Response to PF-1.1:
Although this information (e.g.. a more detailed listing of commercial fishing bans and
advisories for the Hudson River) would be useful, the topic is not germane to the specific subject
matter of the PMCR. These items were included in the Phase I report and will be updated and
revisited in the forthcoming Human Health Risk Assessment.
1.2 Purpose of Report
1.3 Report Format and Organization
No significant comments were received on Sections 1.2 and 1.3.
December 211998
PMCR-3
TAMS/LTI/T etraT ech/MCA
-------
2.
SUMMARY AND PRELIMINARY CONCLUSIONS
2.1 Summary
2.1.1 Overall Approach
2.1.2 Water Column and Sediment Models
2.1.3 Fish Body Burden Models
2.2 Preliminary Conclusions
No significant comments were received on Sections 2.1 through 2.2.
2.2.1 Upper Hudson River PCB Mass Balance
Response to PF-1.2:
The reviewer is referred to Table 4-15. The three cases where segment-average values for
the model output were significantly different than observed values represent conditions for model
segments downstream of the Thompson Island Pool and for three different PCB congeners. No
consistent pattern of significant under- or over-prediction is apparent. The t-tests were applied to
a relatively sparse set of matched predicted and observed conditions for the 1993 EPA Phase 2
sampling period; the reader is cautioned against over-interpretation of the preliminary results.
Response to PF-1.3:
The solids gain (across Thompson Island Pool) during the high flow period may have been
under-represented due to incomplete information on suspended solids loads for the Moses and Snook
Kill tributaries. The Spring 1994 high flow solids monitoring data were not available for the PMCR.
The Baseline Modeling effort will incorporate these data, as well as other estimation methods, to
more accurately determine flows and solids loads for the 1977-1997 hindcasting period.
Response to PF-1.4:
There is an error in the PMCR. The last sentence of "Item 13" in Section 2.2.1 (page 2-5)
should be re-stated as: "The principal factor responsible for differences between total PCBs and
lower-chlorinated congeners appears to be that sediments in Thompson Island Pool are relatively
more contaminated with lower-chlorinated congeners than with higher-chlorinated congeners."
This sentence should have appeared as a separate item, since it refers to findings presented
in several preceding "Items".
2.2.2 Thompson Island Pool Hydrodynamics and Sediment Erosion
No significant comments were received on Sections 2.2.2.
December 22. 1998
PMCR-4
TAMS/LTl/TetraTech/MCA
-------
2.2.3 Upper Hudson River Fish Body Burdens
Response to PF-1.5
Differences in PCB tissue residues by fish sex. age, and/or season have been qualitatively
evaluated, but it is not currently planned to conduct an extensive quantitative evaluation. There are
not enough data to provide a robust statistical evaluation of differences by age, sex. or season. The
1993 Phase II data set provided fish collections at one point in time (August), and the NYSDEC data
sets, for the most part, tend to collect species once per season. Insofar as differences in PCB uptake
by sex reflect endpoints of interest (e.g., females vs. males in terms of reproductive endpoints), these
will be explored.
Response to PF-1.6:
The model structure is dominated by the feeding preferences of the particular fish species.
These feeding preferences are based on life history studies of the species (preferentially from the
Hudson River), anecdotal evidence from analyses of fish stomach contents using preserved samples
provided by NYSDEC, and studies from the 1970s and 1980s by environmental engineers evaluating
the impacts of power plant effluent on fish populations. The model structure does allow feeding
preferences to be changed and/or expressed as distributions, and a sensitivity analysis provides
information on the relative impact of feeding preference assumptions on predicted body burdens.
Response to PF-1.7:
The statement that "...the estuarine portion of the Lower Hudson River is influenced
primarily by direct external loadings and loadings from the vicinity of NY City" (Preliminary
Conclusions, Section 2.2; pg. 2-9) was based on preliminary results from the Upper Hudson River
model and results from the original version of the Lower Hudson River model. A more accurate
assessment must await results from the Baseline Modeling effort and the updated Lower Hudson
River model.
2.2.4 Lower Hudson PCB Mass Balance and Striped Bass Bioaccumulation
Response to PF-1.8:
The point made by the reviewer (i.e., the conclusion that striped bass net PCB uptake occurs
primarily between RM 18.5 and 78.5 is an artifact of the model, since the model did not consider the
distribution of striped bass in the river above RM 80) is discussed in Section 7.7 of the PMCR. The
summary of findings from the Lower Hudson River bioaccumulation model results should have
qualified the statement regarding striped bass PCB uptake to reflect the discrepancy between the
assumed migration patterns in the existing model and more recent observations which indicate the
presence of both adults and juveniles above RM 80.
Response to PF-1.9
Editorial; correction noted.
rVrrmher ? I I WK
PMrR
T A VIS/1 TI/TetraTech/MC A
-------
3.
MODELING APPROACH: TRANSPORT AND FATE
3.1 Introduction
Response to PL-1.7:
Figure 3-1 was meant to provide a general overview of the principal components (water,
solids and PCBs) of the WASP model mass balance framework. Separate diagrams representing the
conceptual model framework for the solids and PCB mass balances were provided in Figures 3-2
and 3-4, respectively. The presentation of this material will be clarified (e.g, explanation of the
symbols used) in the Baseline Modeling Report.
Response to Pfi-1.5:
The inclusion of groundwater advection was simply a numerical experiment designed to
determine the degree to which this process might influence water column PCB concentrations in
Thompson Island Pool. USEPA does not advocate inclusion (or exclusion) of groundwater
advection within the HUDTOX model. This process will be further assessed in the Baseline
Modeling effort. A model hindcast calibration to historic data will help verify that the model does
not misrepresent either PCB flux from sediments or the rate of depletion of PCBs in sediment.
3.2 Modeling Goals and Objectives
3.3 Conceptual Approach
3.4 Hudson River Database
No significant comments were received on Sections 3.2 through 3.4.
3.5 Upper Hudson River Mass Balance Model
Response to PG-1.2:
The HUDTOX model as presented in the PMCR was not intended to be used as a predictive
tool to assess remedial action scenarios. The technical concerns raised by the reviewer {i.e., by
overestimating resuspension and deposition, underestimating tributary solids loadings, and
decoupling sedimentation from the other solids parameters, the model overstates the transfer of
PCBs from sediment to water) are recognized and are being addressed in the ongoing Baseline
Modeling effort, as outlined in Appendix B, (revised July 1998).
Response to PG-1.3:
The Baseline Modeling effort will include simulation of the historical period from 1977
through 1997. The GE database was acknowledged, subsequent to preparation of the PMCR, to
contain significant analytical biases. USEPA's previous efforts to "correct'' the GE PCB data have
been superseded by GE's own efforts to make appropriate corrections to these data. These newly-
corrected data have been provided to USEPA and are being used in the ongoing Baseline Modeling
December 21. 1998
PMCR-6
TA MS/LTI/T etraTech/MCA
-------
effort. It is not true that the USEPA model fails to consider the effects of PCB dechlorination; it is
precisely to capture dechlorination effects that selected individual congeners, as well as total PCBs.
will he modeled.
Response to PL-1.8:
A constant is assigned to the fraction of TSS consisting of organic carbon (Section 3.5.2,
Solids Submodel, last paragraph). Representation of spatial-temporal variability in fraction organic
carbon (flx) requires either direct field measurements or a separate model of organic carbon
dynamics. The available data are not sufficient to support development and application of a separate
mass balance model for organic carbon dynamics. Furthermore, in USEPA's judgment, the organic
carbon dynamics in the Upper Hudson River do not warrant the level of process resolution that
would be included in such a model. In the Baseline Modeling effort, spatial-temporal variability in
foe will be examined using available direct field measurements and included in the model if
appropriate.
Tables 4-10 and 4-14 contain as a PCB model process-related parameter.
The rationale for estimating biotic solids loads due to primary production is presented in
Section 4.4.2. The data used represent the only available information on primary production within
the freshwater portion of the Hudson River.
Response to PL-1.6:
A discussion of why it was felt necessary to develop a new model for the Upper Hudson
instead of applying the existing Lower Hudson model will be included in the Baseline Modeling
Report.
The available version of the Lower Hudson River model was not directly applicable to the
study questions in the Upper Hudson River. The Lower Hudson model represented PCB
homologues and was intended to represent temporal scales on the order of years to decades. The
Upper Hudson River model was designed to represent individual PCB congeners and event-scale
processes.
Work on the Lower Hudson River model has been suspended, because the Lower Hudson
River model is currently being updated by Drs. Robert Thomann and Kevin Farley as part of a
separate project. Upon completion of this work, USEPA will review the updated model and decide
how it should be used to assess the impacts of PCB loads from the Upper Hudson River on the lower
river and estuary.
3.5.1 Introduction
So significant comments were received on Section 3 J I.
December 21 1098
PMCR-7
TAMS/1 .Ti/TctraTcch/MCA
-------
3.5.2 State Variables and Process Kinetics
Response to PI.-1.3:
Additional documentation of the HUDTOX mathematical modeling framework should have
been incorporated within the report for completeness. Relevant equations, expressing the
relationships between various model state variables, will be included in the Baseline Modeling
Report. The reader will still be referred to official USEPA documentation and other sources (e.g.,
Ambrose, et al.) for specific details insofar as HUDTOX and WASP5 share common features.
Response to PL-1.9:
The inclusion of both truly dissolved phase PCBs and DOC-bound PCBs is important and
thus both of these PCB forms have been included in the HUDTOX model. Unfortunately, it is very
difficult to achieve accurate separation of the DOC-bound phase in environmental samples, for
which reason the DOC-bound phase was not directly quantified in either the EPA Phase 2 work or
GE's sampling effort. The Data Evaluation and Interpretation Report (1997) provides an extensive
discussion of PCB partitioning to DOC. together with an analysis of available evidence on partition
coefficients to DOC for individual PCB congeners. The lack of inclusion of DOC-bound PCBs in
the database precludes direct calibration of DOC-bound PCB concentrations computed by the
HUDTOX model. Both the estimated values developed in the Data Evaluation and Interpretation
Report and literature values for partition coefficients between PCBs and DOC will be used in the
model to verify that results for DOC-bound PCB concentrations are reasonable.
Response to PI.-1I0:
More complete information on model enhancements (referenced in the last paragraph of
section 3.5.3) will be provided in the Baseline Modeling Report.
3.5.3 Spatial-Temporal Scales
3.5.4 Application Framework
3.6 Thompson Island Pool Hydrodvnamic Model
3.6.1 Introduction
3.6.2 State Variables and Process Mechanisms
3.6.3 Spatial-Temporal Scales
3.6.4 Application Framework
So significant comments were received on Sections 3.5.3 through 3 6 4
3.7 Thompson Island Pool Depth of Scour Model
December 21. 1998
PMCR-8
TAMS/LTI/TetraTech/MCA
-------
3.7.1 Introduction
Response to PF-1.10:
The Depth of Scour Model in the Baseline Modeling effort has been expanded to include
representation of both cohesive and non-cohesive sediment types.
Response to PI-1.11:
The different models in this Reassessment RI/FS were designed to answer different
questions. The Depth of Scour Model was designed to estimate masses of solids and PCBs
resuspended. and corresponding depth of sediment scour, for large flood events at fine spatial scales.
The HUDTOX model was designed to estimate transport, fate and redistribution of solids and PCBs
over a range of river flows at coarser spatial scales. When used in a complementary fashion, these
two models will adequately address the transport and redistribution of eroded sediments.
Response to PG 1.7:
The application of the Lick equation to the dynamics of cohesive sediment resuspension in
the Depth of Scour Model will be reviewed. If any errors are discovered they will be corrected. The
USEPA will consider use of a modified van Rijn model for resuspension properties of non-cohesive
sediments.
3.7.2 Process Representation
3.7.3 Spatial-Temporal Scales
3.7.4 Applications Framework
No significant comments were received on Sections 3 7.2 through 3. ~ 4.
3.8 Lower Hudson River PCB Transport and Fate Model
3.8.1 Introduction
Response to PF-1.11:
The Baseline Modeling Report will state that the Lower Hudson River model is currently
being updated by Drs. Robert Thomann and Kevin Farley as part of a separate project. Upon
completion of this work. USHPA will review the updated model and decide how it should be used
to assess the impacts of PCB loads from the Upper Hudson River on the lower river and estuary.
At the present time it is not known how results from the updated model may differ from those in the
original model.
December 21 1998
PMCR-9
TAMS/l Tl/TetraTech/MCA
-------
3.8.2 State Variables and Process Kinetics
3.8.3 Spatial-Temporal Scales
3.8.4 Applications Framework
No significant comments were received on Sections 3.8.2 through 3.8.4.
December 21. 1998
PMCR-10
TAMS/LTl/TetraTech/MCA
-------
4.
CALIBRATION OF UPPER HUDSON RIVER PCB MODEL
Response to PP-1 A:
There are some inconsistencies between displayed results and corresponding labels (dates).
Results and labels will be consistent in the Baseline Modeling Report.
Response to PP-1. B:
The USGS data for flow at Green Island has been obtained in electronic form from the
USGS, therefore the offer of the electronic file is appreciated but not needed. Since the Lower
Hudson will not be modeled by the USEPA as part of the Reassessment, the suspended matter data
are not required for this effort. Again the offer is appreciated by the USEPA.
Response to PC-1.1:
The Baseline Modeling effort will include investigation of more comprehensive data sets for
solids (TSS) and river discharge. For example, the Spring 1994 high flow solids survey conducted
by Dr. Richard Bopp included daily measurements in the main stem and principal tributaries. After
completion of the PMCR. 5000 additional USGS TSS measurements were discovered and
incorporated within the USEPA/TAMS database. The solids mass balance in the PMCR will be
extended to include the historical period from 1977 to 1997.
4.1 Introduction
Response to PF-1.12:
This information (a description of the physico-chemical properties for the five selected PCB
congeners and the relevancy of the five selected congeners to the food chain model and ecological
risk assessment) will be presented in the Baseline Modeling Report.
Response to PF-1.13:
Matched pairs of sediment and water column PCBs are not available to directly assess
whether there is such a concomitant increase. There is not yet a full understanding of processes
controlling PCB dynamics between Fort Edward and Thompson Island Dam. The Data Evaluation
and Interpretation Report (USEPA. 1997) provides a detailed discussion and analyses of available
data from all sources on changes in water column load across the Thompson Island Pool. The
recently released Low Resolution Coring Report provides an analysis of changes in sediment
inventory of PCBs over time. General Electric has recently (OBG. 1997) conducted float studies
in the Thompson Island Pool designed to sample PCB concentrations in discrete parcels of water as
they traverse the pool. Results from these studies will be evaluated as part of the Baseline Modeling
effort.
December 21 IW8
PMCR-11
TAMS/l. TI/TetraTech/MCA
-------
4.2 Historical Trends in Water Quality Observations
Response to PC-1.3:
There are uncertainties and temporal variability in PCB loadings to the Upper Hudson River
from the vicinity of Hudson Falls due to failure of the Allen Mill gate structure in September 1991,
and to the recently-documented seepage of PCB in DNAPL form. These phenomena complicate
efforts to develop an accurate picture of historical PCB loading. Analysis of the available data on
water column loads of PCBs is contained in the Data Evaluation and Interpretation Report (USEPA,
1997).
4.3 Overview of Preliminary Calibration Data Set
Response to PF-1.15:
The impact of these zero values for BZ#138 in water and sediment will be investigated more
thoroughly in the Baseline Modeling effort. It should be noted that the GE PCB data set has also
been revised. Consequently, many of the data values used in the PMCR will be revised in the
Baseline Modeling Report.
Response to PG-1.11 A:
The 1991 GE sediment data were the most appropriate available data for specification of
sediment initial conditions for the modeling effort in the PMCR. The USEPA/TAMS Phase 2 low
resolution sediment coring data (1994) were unavailable when the PMCR modeling effort was
conducted. The ongoing Baseline Modeling effort will include all available sediment data for the
historical period from 1977 through 1997.
4.4 Model Input Data .
Response to PC-1.4:
Please refer to the following comment (PC 1.5) and response. They are both a continuation
of the same technical issue.
Response to PC-1.4 and PC-1.5:
The development of PCB loads at Fort Edward is necessary for applying the PCB model
within Thompson Island Pool. The uncertainty in these loads is acknowledged, but applying PCB
loads developed using the data at the downstream end of the pool to the upstream boundary at Fort
Edward would be inappropriate for simulating PCB transport through the pool. It should be noted
that GE has collected new and relevant information during 1997 that pertains to representativeness
of the Fort Edward PCB data. This information will be assessed in the Baseline Modeling effort.
With regard to the PCB "spikes" during high flow events, please refer to the response to PC-
1.3. No mass balance model can correct bias that is inherent in a field sampling program. Such a
model can only reconcile external loadings and system responses within the uncertainties in the
December 21. 1998
PMCR-12
T AM S/LTl/TetraT ech/MC A
-------
available field data. In response to concerns that potential bias may exist in the TIP data. GE has
recently (1996 and 1997) conducted additional water column sampling to better represent the PCB
concentrations in the vicinities of Rogers Island and the Thompson Island Dam. These data will be
utilized in the Baseline Modeling effort. Finally. PCB loadings in the Baseline Modeling effort will
be investigated using flow-weighted average, regression, and ratio estimator methods in order to best
represent the influence of PCB concentration "spikes" during high flow events.
4.4.1 System-Specific Physical Data
No significant comments were received on Section 4.4.1.
4.4.2 External Loadings
Response to PL-1.12:
The specification of atmospheric PCB concentrations (e.g., use of the Green Bay data) will
be re-assessed in the Baseline Modeling effort.
Response to PL-1.13:
USEPA elected not to present the sediment PCB initial concentrations in these deeper layers
(5-10 and 10-25 cm) in the PMCR because they have no influence on model results for the nine-
month simulation period. Furthermore, the 1994 low resolution sediment coring data were
unavailable for the PMCR and there were numerous unresolved technical issues with representation
of PCBs in the historical data.
Response to PF-1.16 and PF-1.17:
Solids loads for the modeling effort in the PMCR were required for only a nine-month period
in 1993. These loads were estimated using data from only 1993 to avoid complications due to
historical differences in sampling protocols and sampling locations.
After completion of the PMCR. 5000 additional USGS TSS measurements were discovered
and incorporated within the USEPA/TAMS database. In the Baseline Modeling effort, solids loads
will be extended to include measurements for the historical period from 1977 to 1997. Any
differences in sampling protocols or sampling locations will be addressed and resolved.
Response to PF-1.18:
Preliminary modeling efforts were conducted with Batten Kill and Fish Creek solids being
treated in similar fashion to the direct drainage (ungaged) flows. The text (p. 4-7. par. 4) should
have made this clear. In the Baseline Modeling effort, solids loads from these and other tributaries
will assessed in greater detail, especially since more extensive solids data are now available.
Response to PF-1.19:
Text should state "as can be inferred from Figure 4-6(b). the principal external . .
December 21 1998
PMCR-13
T AMS/LTI/TctraTcch/MCA
-------
Response to PF-1.20:
The "other tributaries" are the Mohawk and Hoosic Rivers.
Response to PF-1.21:
These correlations (PCBs vs TSS and flow, and difference in correlation between higher and
lower chlorinated congeners) will be reviewed as part of the Baseline Modeling effort and results
will be included in the Baseline Modeling Report.
Response to PF-1.22:
The reviewer is referred to the Data Evaluation and-Interpretation Report (DEIR) which
became available in February, 1997 with regard to the exclusion of a high PCB measurement in
January 1993 . Page 3-127 of the DEIR states with regard to the GE PCB data that "a high estimated
load at RM 194,6 (Rogers Island) in January 1993 is due to a single observation of 1086 ng/L total
PCBs on January 15. Sampling did not capture this concentration slug at RM 188.5, and it is
possible that this sample could also represent disturbance of contaminated sediment in the sampling
procedure, although it is not annotated as having been collected from shore." This data point may
not be an analytical outlier; however, it does not appear to represent upstream PCB loads at Rogers
Island.
This value (10 ng/L PCBs) was estimated based on an assessment of the lowest reported
tributary PCB concentrations. The Baseline Modeling Report will contain more specific
information. It should be noted that PCB loads from these tributaries are relatively insignificant.
Response to PF-1.24:
The reviewer is correct. The PMCR (pg 4-9) should state that BZ#4 from upstream sources
accounts for 49% (not 27%) of the load during spring high flow, and that 76% (not 68%) of the total
PCB external load to the Upper Hudson occurs during spring high flow.
4.4.3 Forcing Functions
Response to PF-1.25:
The need to better support specification of atmospheric PCB inputs to the Upper Hudson
River (pg 4-10) is acknowledged. USEPA thanks the reviewer for offering to provide information
that was unavailable for the modeling effort in the PMCR.
Response to PF-1.26:
The exact depth for the temperature measurements is not available in the database, but will
be investigated as part of the Baseline Modeling effort. Although some temperature variation with
depth is possible within deeper reaches of the Upper Hudson River (especially behind dams) it is
December 21. 1998
PMCR-S4
T AMS/LT1/T etraT ech/MCA
-------
unlikely that these conditions are significant. The HUDTOX model is vertically averaged and does
not explicitly represent water column stratification.
4.4.4 Boundary Conditions
No significant comments were received on Section 4.4.4.
4.4.5 Initial Condition
Response to PF-1.27:
Please refer to the response to comment PF-1.25, above (section 4.4.4). USEPA
acknowledges the potential utility of these data for the Baseline Modeling effort.
Response to PF-1.28:
The GE 1991 sediment data were composited horizontally by groupings of individual coring
locations, but an attempt to reflect vertical PCB gradients in the sediments was made by forming
these composite samples within distinct layers. These data cannot be characterized simply as being
vertically averaged, except perhaps within the context of each vertical layer that was sampled.
Response to PF-1.29:
The sediment bulk densities quoted in the report are dry weight-based concentrations.
Response to PF-1.30:
The GE PCB data have been substantially revised and incorporated into the USEPA/TAMS
Phase 2 database (Release 4.1, August 1998). These revisions will improve the comparability
between the GE and USEPA/TAMS data sets. No correction factors should be necessary in order
to use the GE data in the Baseline Modeling effort.
Results for sensitivity analyses of model outputs to changes of plus and minus 30 percent in
initial sediment PCB concentrations are presented in the PMCR (Section 4.9).
Specification of deep sediment layer PCB concentrations was not important for the nine
month simulation period in the PMCR. It will be important for the longer-term simulations in the
Baseline Modeling effort and will receive due attention using all of the available field data.
4.5 Internal Model Parameters
Response to PL-1.16:
The reviewer is referred to the Data Evaluation and Interpretation Report (DEIR) of
February. 1997. An analysis of sources of variability in PCB partition coefficients provided in the
DEIR indicated that the particle concentration effect, which would be represented by Ux. is not
significant for Upper Hudson River PCBs. The DEIR was in draft form at the time the PMCR was
December 21. 1998
PMCR-15
TAMS/LTl/TetraTech/MCA
-------
produced and therefore was not referenced. Such inconsistencies will be eliminated in the Baseline
Modeling Report.
4.5.1 Solids Model Parameters
No significant comments were received on Section 4.5.1.
4.5.2 PCB Model Parameters
Response to PF-1.31:
This question (the effect of median [vs. mean vs. maximum] parameter values for total PCBs
has on the output) will be addressed in more-detail in the Baseline Modeling Report.
4.6 Calibration Approach
Response to PO-1.1:
The three "significant shortcomings" listed in the comment are addressed below.
1. The PMCR solids loads from Snook and Moses Kills are likely underestimated for the
1993 high-flow period, but not necessarily for the lower flow periods. Flow and solids data
were unavailable for Snook and Moses Kills at the time of the modeling work in the PMCR.
Spring 1994 high flow solids data are now available and will be used, along with other new
solids data, in the Baseline Modeling effort. These additional data will improve estimates
of solids loads for the main stem and tributaries in the Upper Hudson River.
2. Concerns regarding solids deposition and resuspension rates will be addressed in the
Baseline Modeling effort through calibration to high flow events and a long-term (1977-
1997) historical calibration. Solids resuspension rates should be reduced during low flow
periods.
3. Sedimentation and water column solids dynamics were de-coupled in the preliminary
model because of the short (nine months) calibration period and the lack of solids data with
which to test the model during high-flow events. The Baseline Modeling effort will
determine sedimentation rates as the net of deposition and resuspension. This is more
appropriate for long-term simulation of both historical conditions and remedial scenarios.
4.6.1 Transport Model (Water Balance) Specification
4.6.2 Solids Model
4.6.3 PCB Model
No significant comments were received on Sections 4.6.1 through 4.6.3.
December 21. 1998
PMCR-16
T AMS/LT1/T etraTech/MC A
-------
4.7 Calibration Results
Response to PC-1.2:
The concerns expressed in this comment are addressed in the response to comments PG-
1.1 IC. PG-1.1 ID. PG-1.1 IE and PG-1.11F.
Response to PG-1.1 IF.:
Gross hydrodynamic resuspension of solids is significantly reduced and possibly negligible
at low flows. However, this rate should not be arbitrarily set to zero at low flows. Even a small
gross resuspension of solids (whether from hydrodynamic or other causes) is likely to have an impact
on sediment-water column PCB exchange in areas with high levels of sediment PCB contamination.
USEPA also believes it is appropriate in a "box model" to allow simultaneous gross resuspension
and settling. USEPA does agree that there is no net resuspension. but most likely net settling, of
solids during low flow conditions.
Response to PG-1.1 IF:
Although incorporation of the variation in solids load and deposition velocity with river flow
would be useful, currently available site-specific data are insufficient to accurately characterize
changes in the solids load composition with respect to particle sizes as a function of flow. If
additional information becomes available during the Baseline Modeling effort, then this suggestion
will be considered.
Response to PG-1.11G:
The 1991 GE sediment data were the most appropriate available data for specification of
sediment initial conditions for the modeling effort in the PMCR. The USEPA/TAMS Phase 2 low
resolution sediment coring data (1994) were unavailable when the PMCR modeling effort was
conducted. The ongoing Baseline Modeling effort will include all available sediment data for the
historical period from 1977 through 1997.
The relationship between fish body burden and its food web components will be evaluated
in the ecological components of the modeling effort.
Response to PG-1.11H:
The conceptual framework (Figure 3-4 in the PMCR) does explicitly acknowledge possible
degradation/dechlorination of PCBs in the sediments. This potential loss mechanism should be
addressed in long-term forecasting simulations. This technical issue {i.e.. the rate of dechlorination
and its variation over time) is unresolved at the present time.
Most of the references suggested by General Electric for estimating degradation rates post-
date publication of the PMCR. Furthermore. General Electric states (page 24 of their 11/21/96
comment) "Although recently conducted laboratory experiments (Fish. 1996) indicate that
December 21 1098
PMCR-17
TAMS/lTI/TetraFech/MCA
-------
dechlorination can occur at a rate sufficient to contribute to the excess loading as described above,
extrapolation of these results to the field is not yet complete."
Accordingly, US EPA plans to address these loss mechanisms using sensitivity analyses for
the calibration and forecasting simulations unless sufficient information to estimate in situ
degradation rates becomes available within a usable time frame.
Response to PL-1.14:
The specified settling velocity (Vs of 2.0 m/day) was a model calibration parameter. The
typical ranges provided in the PMCR (Vs ranging from 0.25 to 3.05) represent calibrated settling
rates from model applications to other similar water bodies. A value of 2.0 m/day is consistent with
settling velocities used for PCBs in the Upper Hudson River (Hydroscience, 1978) and in Green Bay
(Bierman el al.. 1992).
There is no single correct, constant value for gross settling velocity in the Hudson River. A
more refined approach to specifying gross settling as a function of variations in flow and solids
concentrations is being developed as part of the Baseline Modeling effort.
Response to PL-1.1S:
A reference should have been included to a General Electric report (prepared by HydroQual)
examining sedimentation in the Upper Hudson River. The report entitled "Quantification of
Sedimentation in the Upper Hudson River", October 31, 1995, examined several methods of
estimating long-term sedimentation rates. The analysis in this report suggested that sedimentation
rates (Vb) in Thompson Island Pool ranged between 0.24 to 0.57 cm/year, and averaged
approximately 0.21 cm/year between Fort Edward and Stillwater.
Sedimentation rate will be fully coupled with solids settling and resuspension dynamics in
the Baseline Modeling effort. Consequently, sedimentation rate will no longer be an independent
model parameter, but will instead be determined by the balance between dynamic settling and
resuspension processes.
Response to PG-l.llB:
The inclusion of groundwater influx was simply a numerical experiment designed to
determine the degree to which this process might influence water column PCB concentrations in
Thompson Island Pool. The experiment was also conducted to demonstrate the significance of this
potential influx on the lower-chlorinated, less hydrophobic PCB congeners that constitute most of
the observed gain in PCB mass across Thompson Island Pool. The specified groundwater inflow
was based on an assessment of USGS flow data and a simplified application of Darcy's Law. The
PMCR did not advocate inclusion (or exclusion) of groundwater influx within the HUDTQX model.
This process will be further assessed in the Baseline Modeling effort.
The possible influx of groundwater to sections of the river downstream of Thompson Island
Pool was not evaluated in the PMCR. This or any other hypothesis employed to explain the
December 21. 1998
PMCR-18
TAMS/LTI/TetraTech/MCA
-------
observed increase in PCB concentrations across Thompson Island Pool should be validated to the
extent possible through the evaluation of appropriate field data.
Response to PG-1.11C:
The preliminary model calibration was for only a nine-month period (January through
September 1993) and excluded three months during 1993 that may have been net depositional for
solids. Over an annual cycle the preliminary model may still have predicted net erosion, but to a
significantly lesser degree than suggested by the results from the limited nine-month simulation
period.
It should be noted that a prediction of net solids erosion, by itself, does not imply that
external solids loads were underestimated. External solids loads and sediment-water exchanges can
only be accurately constrained by conducting a long-term hindcasting simulation in which solids and
PCB mass are balanced simultaneously.
Response to PG-1.1 ID:
The bulleted actions suggested by General Electric to "improve the accuracy of its PCB fate
model" are being addressed as part of the Baseline Modeling effort. Both the long-term historical
record and high-flow event periods will be simulated as part of the PCB model validation process.
Response to PG-1.111:
Results from the preliminary model can not be used to judge its utility for conducting long-
term simulations. The preliminary PCB model was intended to represent conditions during the 1993
USEPA Phase 2 sampling period. Due to its preliminary nature and acknowledged data limitations,
this model was not intended to simulate long-term changes in transport and fate of sediment PCBs.
The Baseline Modeling effort will include a long-term hindcasting calibration for the period 1977
through 1997.
4.7.1 Solids Model
No significant comments were received on Section 4. 7, /.
4.7.2 PCB Model
Response to PG-1.4. PG-1.11B and PG-1.11E:
The governing equations in the HUDTOX model are inherently mass-conserving. The model
does balance PCB mass across Thompson Island Pool. The model was not able to exactly match
increases in PCB concentrations across the pool, especially for lower-chlorinated congeners. This
problem will receive continued emphasis during the Baseline Modeling effort.
GE's interpretations of a potential mass imbalance across the Thompson Island Pool were
rendered invalid by subsequent corrections for analytical bias in the GE data set. A "mass
imbalance" is not evident in the USEPA Phase 2 data. Since release of the PMCR, both USEPA and
December 21. 1998
PVfCR-19
TAMS'LTI/Tctra Tech/MCA
-------
GE have been pursuing a variety of analyses on mechanisms and explanations for observed PCB
mass transport across the Thompson Island Pool, including examination of the various hypotheses
proposed by GE in comment PG 1.11. The results of these additional analyses will be incorporated
into the Baseline Modeling Report.
4.8 Mass Balance Component Analysis
4.9 PCB Model Calibration Sensitivity Analysis
No significant comments were received on Sections 4.8 and 4.9.
CHAPTER 4 TABLES
Response to PL-1.17:
As part of the Baseline Modeling effort, both parametric and non-parametric statistical
methods will be used to compare model output with observed data (e.g., Tables 4-13 through 4-17).
This will be done in recognition of the large variances that sometimes occur in both model output
and field observations.
CHAPTER 4 FIGURES
Response to PF-1.14:
Suggested text should have been added to Figure 4-5 for clarity.
December 21. 1998
PMCR-20
T A M S/LT t/TetraT ech/MC A
-------
5. CALIBRATION OF THOMPSON ISLAND POOL HYDRODYNAMIC MODEL
5.1 Introduction
5.2 Model Input Data
5.2.1 System-Specific Physical Data
5.2.2 Forcing Functions
5.2.3 Boundary Conditions
5.3 Internal Model Parameters
5.4 Calibration Approach
5.5 Calibration Results
5.6 Model Validation
5.6.1 Rating Curve Velocity Measurements
5.6.2 FEMA Flood Studies
5.7 100 Year Flood Model Results
5.8 Sensitivity Analyses
5.8.1 Manning's 'n"
5.8.2 Turbulent Exchange Coefficient
No significant comments were received on Sections 5.1 through 5.8.2.
5.9 Conversion of Flow Velocity to Shear Stress
5.9.1 Results
Response to PG-1.13B:
The Depth of Scour Model for the Baseline Modeling effort has been modified to use the
predicted RMA-2V bottom shear stress.
5.10 Discussion
No significant comments were received on Section 5,10.
December 21. 1998 PMCR-21 TAMS/LTl/TetraTech/MCA
-------
6. APPLICATION OF THOMPSON ISLAND POOL DEPTH OF SCOUR MODEL
6.1 Introduction
Response to PG-1.10:
Neither the PMCR nor the Baseline Modeling Report were intended to include forecasting
simulations to compare system responses to different potential remedial scenarios. Forecasting
simulations are expected to be conducted as part of the Feasibility Study.
6.2 Available Data
Response to PC-1.6:
Data from the TIP Depth of Scour Model (Section 6.2.2) Resuspension Experiments must
be interpreted within the context of the assumptions and uncertainties inherent in collecting and
testing sediment samples. These assumptions and uncertainties are discussed in the primary
references cited in Section 6 of the PMCR.
6.2.1 Bottom Sediment Distribution
6.2.2 Resuspension Experiments
No significant comments were received on Sections 6,2.1 or 6.2.2
6.3 Model Parameterization and Uncertainty
Response to PG-1.13C:
The application of the Depth of Scour Model in the Baseline Modeling effort has been
modified to use localized data representing dry bulk density.
Response to PG-1.13D:
The Depth of Scour Model in the Baseline Modeling effort will incorporate armoring in the
cohesive sediment areas. The HUDTOX fate and transport model will be consistent with the Depth
of Scour Model to the maximum extent possible.
In the Baseline Modeling effort, both settling and resuspension will be flow-dependent in two
dimensions within Thompson Island Pool. The RMA-2V hydrodynamic model will be used to
generate 2-D velocity distributions for both the Depth of Scour and HUDTOX models.
Resuspension results from HUDTOX will be compared to results from the Depth of Scour Model
for high flow events.
December 21. 1998
PMCR-22
TAMS/LTl/TetraTech/MCA
-------
6.3.1 Rearrangement of Erosion Equation
6.3.2 Parameter Estimation
6.3.3 Prediction Limits
6.4 Depth of Scour Predictions at Selected Locations in Cohesive Sediment Areas
No significant comments were received on Sections 6.3.1 through 6.4.
6.5 Global Results for Cohesive Sediment Areas
Response to PG-1.13F:
An approach for applying the Depth of Scour Model to non-cohesive sediments has been
developed and was presented at a meeting between USEPA/TAMS and General Electric in March.
1997.
Response to PG-M3F:
The limitations of the Depth of Scour Model with regard to non-cohesive sediments are
acknowledged. The application of the Depth of Scour Model to non-cohesive sediments produces
only an "upper bound" estimate for depth of scour.
Response to PF-1.32:
The conclusion (that flood events will not erode PCB contaminated cohesive sediments to
any large degree) is valid and reasonable for the cohesive sediment areas characterized by the high
resolution sediment cores and the resuspension experiments. This conclusion will be independently
assessed by comparing sediment PCB distributions from the 1984 NYSDEC survey with those from
the 1994 Phase 2 low resolution sediment cores. It should also be noted that the HUDTOX mass
balance model represents the entire sediment area of Thompson Island Pool. Results from the Depth
of Scour Model and HUDTOX models will be compared to verify' reasonableness and consistency.
December 21. 1998
PMCR-23
TAMS/l.TITeiraTech/MCA
-------
7. APPLICATION OF LOWER HUDSON RIVER PCB TRANSPORT AND FATE
MODEL
Response to PG-1.15:
USEPA has suspended work relating to the Lower Hudson River model, due to the fact is
that the Lower Hudson River model is currently being updated by Drs. Robert Thomann and Kevin
Farley as part of a separate project. Upon completion of their work, USEPA will review the updated
model and decide how it should be used to assess the impacts of PCB loads from the Upper Hudson
River on the lower river and estuary.
7.1 Introduction
Response to PF-1.33 and PG-1.9
See the above response to comment PG-1.15.
7.2 Model Input Data
Response to PL-1.18:
The rationale and justification for these decisions is presented in Thomann et al. (1989;
1991).
7.2.1 System-Specific Physical Data
7.2.2 External Loadings
\'o significant comments were received on Sections 7 2.1 or 7. 2.2.
7.2.3 Forcing Functions
Response to PF-1.34.
It is expected that the most current knowledge about striped bass migration patterns will be
incorporated into the updated Lower Hudson River model currently being developed by Drs. Robert
Thomann and Kevin Farley. It is not known whether the reference cited is the most current
information.
7.2.4
Boundary Conditions
7.2.5
Initial Conditions
7.3
Internal Model Parameters
7.4
Application Approach
December 21.1998
PMCR-24
TAMS/l.Tl/TclraTech/MCA
-------
7.5
Application Results
7.6 Diagnostic Analyses
7.6.1 Component Analysis
No significant comments were received on Sections 7.5 through 7.6.1.
7.6.2 Sensitivity Analysis
Response to PF-1.3 5
The reader is correct in noting that model segment 2 is not sensitive to loading or
volatilization. The reader is also correct in noting that PCB concentrations were quite or very
sensitive to the various processes listed.
7.7 Discussion
Response to PF-1.36:
Figure 7-7 was not meant to contain the original Thomann model. It was meant to contain
comparative results with the original Thomann model for two different assumptions on PCB loads.
Figure 7-7 does contain this information. The labeling should have been more clear, and will be
corrected in the BMR.
Response to PG-1.14 and PG-1.16:
The limitations of the preliminary Hudson River PCB model calibration were acknowledged
within the PMCR. The PMCR represents the status of the preliminary PCB modeling effort as of
Fall, 1995. Datasets, database corrections and other pertinent information which became available
after October 1995 were not incorporated within the fate and transport modeling presented in the
PMCR. The ongoing PCB modeling effort will result in a model that is both scientifically credible
and useful to EPA for evaluating remedial assessment alternatives. Results from this ongoing effort
will be presented in the BMR.
December 2). 1998
PMCR-25
TAMS/LTI/TelraTech/MCA
-------
8.
MODELING APPROACH: FISH BODY BURDENS
8.1 Modeling Goals and Objectives
Response to PG- \ .6 and PG-1.12
General Electric contends that neither the Bivariate BAF model nor the Probabilistic Food
Chain Model should be used to make predictions because they ignore the short and long-term
variability in the relationships among PCB levels in the water column, sediment, and fish and do not
attempt to describe or respond to the mechanisms by which fish bioaccumulate PCBs. Instead, GE
recommends using a time-variable, mechanistic food web model, such as the Gobas model.
In response, it should first be noted that USEPA has already proposed use of the Gobas
model (PMCR, p. 8-2), although results were not ready for the PMCR, and has actively worked with
GE/QEA since release of the PMCR on the development of modeling approaches to address PCB
bioaccumulation in fish. In addition, since the completion of the Modeling Approach Peer Review,
USEPA has decided to also use the mechanistic Gobas model, as per the recommendation of that
panel. The PMCR provides only preliminary results on bioaccumulation modeling for the Hudson
River, which will be expanded and further developed in the Baseline Modeling Report.
It is also not the intention of USEPA to use any single bioaccumulation modeling approach
to make predictions; rather, using a weight-of-evidence approach which draws upon all three
modeling tools is proposed. The state-of-the-science is not sufficiently advanced that a mechanistic
food web model alone can be relied upon for predictions, and that sufficient data are not available
for all food chain compartments to fully constrain a mechanistic food web model for the Hudson.
8.2 Background
No significant comments were received on Sections 8.2.
8.2.1 PCB Compounds
Response to PL-1.4:
Fish PCB body burdens will mirror sediment and water concentrations. In terms of the
management objectives for the Hudson River, the endpoint of interest is total PCBs. To the extent
that the bioaccumulation models adequately and accurately predict the uptake of total PCBs, this will
achieve the stated goal. As to the potential risks of individual PCB congeners, data are inadequate
to determine human effects while ecological toxicity benchmarks are available for several PCB
congeners.
8.2.2 PCB Accumulation Routes
No significant comments were received on Section 8.2.2,
December 22. 1998
PMCR-26
TAMS/LTl/TetraTech/MCA
-------
8.3 Theory for Models of PCB Bioaccumulation
Response to PL-1.5
Under the steady-state assumption incorporated in the BAF or transfer factor" approach,
PCB removal processes are implicitly included by representing net accumulation (i.e.. steady-state
balance between uptake and depuration). This approach does not assume a "one-to-one" relationship,
but rather a liner equilibrium relationship. Use of Aroclors to some extent accounts for differential
accumulation with chlorination level. In addition, the use of time-varying or bioenergetic models is
being explored.
8.4 Bivariate Statistical Model for Fish Body Burdens
8.4.1 Rationale and Limitations for Bivariate Statistical Model
8.4.2 Theory for Bivariate Statistical Models of PCB Bioaccumulation
8.5 Probabilistic Bioaccumulation Food Chain Model
8.5.1 Rationale and Limitations
8.5.2 Model Structure
8.5.3 Spatial Scale for Model Application
8.5.4 Temporal Scales for Estimating Exposure to Fish
-\'o significant comments were received on Sections 8.4 through 8.5.4.
8.5.5 Characterizing Model Compartments
Response to PF-1.37:
In the discussion presented on page 8-18, the discussion is focused on the more chlorinated
congeners, as opposed to the total PCB concentrations which is the focus of the DEIR. In viewing
total PCBs, the data presented in the DEIR demonstrate the considerably higher dissolved phase
concentrations relative to those of the suspended matter phase. However, the importance of the
dissolved phase decreases with increasing degree of chlorination. The assumption that most of the
higher chlorinated congeners are present as suspended matter is not a perfect one since the
importance of the dissolved phase varies across the suite of congeners found in fish. Nonetheless,
this assumption is a useful place to begin the modeling analysis. Subsequent modeling analysis will
examine this assumption in greater detail. It is anticipated that subsequent bioaccumulation modeling
will be driven with estimated concentrations of both dissolved and suspended PCB concentrations.
December 2 1. 1998
PMCR-2?
TAMS/LTl/TetraTech/MCA
-------
9.
CALIBRATION OF BIVARIATE STATISTICAL MODEL FOR FISH BODY
BURDENS
9.1 Data used for Development of Bivariate BAF Models
9.1.1 Fish Data
9.1.2 Standardization of PCB Results for NYSDEC Fish Analyses
9.1.3 Water Column Data
9.1.4 Sediment Data
No significant comments were received on Sections 9.1 through 9.1.4.
9.1.5 Functional Grouping of Sample Locations for Analyses
Response to PF-1.38:
Table 9-2 lipid and PCB concentration values are on a wet-weight basis.
Response to PF-1.39:
Table 9-7 gave R2 values as percentages, as is commonly done. The column label should
have included "%".
9.2 Results of Bivariate BAF Analysis
Response to PF-1.40:
Figures 9-11 and 9-12 would be improved by the inclusion of zero on the y axis, and this
change will be made in the Baseline Modeling Report.
9.3 Discussion of Bivariate BAF Results
Response to PF-1.41:
Figures 9-8 through 9-13 would be improved by the inclusion of a 1:1 line of perfect
agreement, and this change will be made in the Baseline Modeling Report.
Response to PF-1.42:
The reviewer provided a variety of suggestions for improving the interpretation of the
bivariate regression model results for predicted fish body burden. The bivariate regression analysis
will be re-done for the Baseline Modeling Report, and these suggestions will be taken into
advisement for that effort. Results of the bivariate BAF model approach presented in the PMCR will
December 21. 1998
PMCR-28
TAMS/LTl/TetraTech/MCA
-------
undergo revisions to reflect a variety of modifications to the data on which the regressions are based.
These data revisions include:
1. Updating of the NYSDEC database and additional research on interpretation of historic
NYSDEC PCB quantitations.
2. Thorough review, reconciliation, and quality assurance checks of the USGS water column
PCB data, completed in April 1997.
3. New information and analyses on the interpretation of historic USGS water column PCB
quantitations.
4. Availability of low resolution coring results.
The data revisions can be expected to result in some modifications to the regression results.
In addition, the planned work for updating the bivariate BAF work will include investigation of
using fate and transport model-generated time series of surface sediment concentrations as an
independent variable.
Response to PF-1.43:
The statement in question refers to "about 42 percent" and "about 58 percent". The value
of 58 percent was chosen, rather than the exact rounded value of 59 percent, so that the two values
would sum to 100 percent.
9.4 Summary
Response to PF-1.44:
Potential differences by sex should have been mentioned in the summary on pg. 9-16.
Potential differences by sex were discussed earlier in this chapter.
December 21. 1998
PMCR-29
T AMS/LT1/T etraT ech/MC A
-------
10.
CALIBRATION OF PROBABILISTIC BIOACCUMULATION FOOD CHAIN
MODEL
10.1 Overview of Data Used to Derive BAFs
10.1.1 Benthic Invertebrates
10.1.2 Water Column Invertebrates
No significant comments were received on Sections 10.1 through 10.1.2.
10.1.3 Fish
Response to PF-1.45:
No, the data from the chironomid short-term study is not presented in this document. The
short-term chironomid study provides information on short-term uptake kinetics of several PCB
congeners to chironomid. These data have been evaluated in terms of the information provided
relative to the time it takes to achieve steady-state concentrations in water-column invertebrates.
These data cannot be directly used in the modeling but have been and continue to be used to provide
information generally on the kinetics of PCB uptake of specific congeners to water-column
invertebrates.
10.1.4 Literature Values
No significant comments were received on Section 10.1.4.
10.2 Benthic Invertebrate: Sediment Accumulation Factors (BSAF)
Response to PF-1.46:
Tissue is lipid normalized and sediment is organic carbon normalized.
10.2.1 Sediment Concentrations
No significant comments were received on Section 10.2.1
10.2.2 Approach
Response to PF-1.47:
The objective of the modeling exercise is to estimate the population distribution of
concentrations for a given fish species following changes in the PCB concentrations of its prey,
which result from changes in the sediment and/or water PCB levels. Because the available data on
predator concentrations are approximately lognormallv distributed, the parameters of interest are the
geometric mean concentration and the geometric standard deviation. From these parameters, the
distribution of expected PCB concentrations in a particular species can be estimated. However, in
December 21.1998
PMCR-30
T AMS/LTI/TetraT ech/MC A
-------
response to this comment, the analysis is being redone using arithmetic means instead of geometric-
means for the denominator(s) of the BAF estimates. Preliminary results suggest that there is no
difference in the estimated geometric standard deviation, and that there is a non-significant
difference in the mean accumulation factor(s).
Response to PF-1.48:
Benthic invertebrate samples were identified down to the class level (in terms of PCB
analyses) and no biomass information was available. There are not enough data to evaluate
differences in a statistically robust manner. For example, although the mean chironomid BSAF is
higher than for other benthic species, and the observed variance far greater, there are only three
samples. Significant effort has been spent-on determining fish consumption preferences at the lowest
practicable taxonomic level specifically to evaluate this question. It is difficult to evaluate species
differences given that data are only available to a class level.
10.2.3 Calculations of BSAF Values for Benthic Invertebrates
Response to PF-1.49:
The figures presenting goodness-of-fit results do differ by the individual congeners. From
a management perspective, the endpoint of interest is total PCBs. This comment will be evaluated
further in the context of the Baseline Modeling Report.
Response to PF-1.50:
The 75th percentile will be included in the figure in the BMR.
Response to PF-1.51:
Figure 10-13 shows the means do range from 0.2 to about 1.6; with four of the means
between 0.2 and 0.5; two means approximately equal to 0.5; and four means between about 0.7 and
1.6.
Response to PF-1.52:
There are not enough data to adequately characterize congener profiles in chironomid
species. First, chironomid were not identified to the species level; and second, there are only three
samples identified as chironomid (at three different stations). The short-term uptake studies may be
of some help in this regard, but used the same chironomid species for all studies and only focused
on a few PCB congeners.
Response to PF-1.53
Due to a document assembly error, not all goodness-of-fit plots were presented in the same
format. The BMR will address consistency of format; however, it is noted that the referenced figures
(10-16 and 10-20) will not be present in the PMCR form in the BMR. as a more formal algorithm
for goodness-of-fit will be used.
I I oos
1 1
T A &4C/1 TI/T»»nT^h/MrA
-------
Response to PF-1.54:
The data are not detailed enough to be able to evaluate this question (i.e.. if the greater
differences in BSAF at certain locations due to the differences in benthic organisms and their feeding
habits). However, this issue is currently being explored and results presented in the context of the
baseline modeling report.
10.3 Water Column Invertebrate: Water Accumulation Factors (BAFs)
10.3.1 Approach
10.3 .2 Calculation of BAFwlltr for Water Column Invertebrates
10.3.3 Alternative Approaches
S'o significant comments were received on Sections 10,3 through 10.3.3.
10.4 Forage Fish: Diet Accumulation Factors (FFBAFs)
10.4.1 Approach
Response to PF-1.55:
At most stations, forage fish are typically represented by spottail shiners less than 10 cm.
For those stations at which a composite forage fish concentration was estimated, the feeding
preference leading to that concentration was accordingly adjusted by the number of fish of each
species represented by the composite. Considerable effort is being expended on refining the feeding
preferences for particular species and sizes of fish. It is acknowledged that the feeding preferences
of the young-of-year of any species will differ from adult fish, hence using only those fish less than
10 cm.
10.4.2 Water Column Concentrations Used to Derive FFBAF Values
So significant comments were received on Section 10.4.2.
10.4.3 Forage Fish Body Burdens Used to Derive FFBAF Values
Response to PF-1.56:
The largemouth bass and white and yellow perch collected in the 1993 Phase II effort were
young-of-year. Consequently, these data were not used in modeling adult body burdens. Sediment
and water concentrations at collocated tessellated darter and spottail shiner stations show high
variances: in other words, any given fish could have been exposed to a higher concentration than any
other fish. The observed variability in all media (more so in the upper Hudson) is very high, so an
observed individual high body burden in any one fish may or may not be attributable to a
correspondingly high concentration in either sediment or water.
December 22, WS
PMCR-'*-
TAMS'l. flTcira lech/MCA
-------
Response to PF-1.57:
Most of the fish samples collected in 1993 represented composites of multiple (typically five
to ten) individual fish.
10.4.4 Calculation of FFBAF Values for Forage Fish
Response to PL-1.19:
The equations defining FFBAF, calculations of FFBAF, and FFBAF tables should have been
included; this will be corrected in the BMR.
Response to PF-1.58:
Editorial; acknowledged.
10.4.5 Calculation of FFBAFs for Small Pumpkinseed Sunfish
10.5 Piscivorous Fish; Diet Accumulation Factors (PFBAF)
10.5.1 Approach Used for Yellow Perch
10.5.2 Approach Used for Largemouth Bass
10.5.3 Approach Used for White Perch
10.6 Demersal Fish: Sediment Relationships
10.6.1 Approach and Calculations of BAF Values
10.7 Summary of Probabilistic Food Chain Models
10.8 Illustration of Food Chain Model Application
10.9 Comparison of Bivariate Statistical and Chain Models
No significant comments were received on Sections 10.5.4 through 10.9.
REFERENCES
No significant comments were received on the References section.
December 22. 1998
PMCR-33
TAMS/LTlTetra Tech/MCA
-------
GLOSSARY
No significant comments were received on the Glossary section.
APPENDIX A - FISH PROFILES
A-1.3.2 [White Perch] Range, Movement and Habitat within the Hudson River
Response to PF-1.59:
In the discussion about water depths, the units should be feet instead of meters.
A-l.10.4 Estimating Feeding Habits of Forage Fish
Response to PF-1.60:
The spottail shiner diet in this size range is 50% water column sources and 50% benthic
sources. Table A-15 will be corrected.
A-1.10.5 Estimating Composite Fish Feeding Habits
Response to PF-1.61:
The data available are in fact only reported for the area between Lock 7 and Troy Dam.
These data were extrapolated to the remainder of the river based on the stomach contents analyses
from other reports. This will be clarified in the Baseline Modeling Report.
Response to PF-1.62:
The preliminary modeling calibration report presents several alternative approaches to
evaluating the water column invertebrate box. In addition, efforts are currently underway to evaluate
the role of direct water uptake. This question (i.e.. what can be done to reduce the uncertainty in the
water column invertebrate compartment, which impacts all subsequent compartments) is currently
being evaluated.
Response to PF-1.63:
Table A-16 will be corrected to show that the forage fish diet is 33% benthic to 67% water
column invertebrates.
APPENDIX B - MATHEMATICAL MODELING, TECHNICAL SCOPE OF WORK
No significant comments were received on Appendix B.
December 22. 1998
PMCR-34
TAMS/LTl/TetraTech/MCA
-------
Responses (DEIR)
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY
VOLUME C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
C. DEIR REPORT
Responses to General Comments on the DEIR Report
Response to PG-1.4E
The writer describes some of the features of the modeling efforts being undertaken by GE
and its consultants. Since the submittal of the DEIR and the receipt of responses from GE, GE has
prepared a second modeling report (QEA, 1998) with somewhat different interpretations of the data,
including newer results obtained in the vicinity of the TI Dam. Comments on the GE report are
included in a separate section of this responsiveness summary. However, it should be noted that the
DEIR is an interpretation of the Phase 2 and other data sets, an interpretation which will be
subsequently analyzed in a more quantitative fashion in the Baseline Modeling Report. Thus, the
interpretation provided in the DEIR is meant to integrate and describe the net effect of all inputs,
while noting the most important components, such as the TI Pool load, which are readily discemable
from the data.
The writer's contention that a portion of the TI Pool load cannot be explained by the sources
cited is based on a model of the Upper Hudson developed by GE and its consultants. For the reasons
stated in USEPA's critique of the QEA report (Book 3 of this Responsiveness Summary), however,
USEPA does not accept the results of the QEA model. Subsequent modeling efforts by the USEPA
will examine this load in detail.
Response to DO-1.11
The DEIR is focused exclusively on the geochemical interpretation of the PCB contamination
of the Hudson River. Discussions on the uptake of PCBs by fish and the congener pattern of their
subsequent body burdens will be included as part of the ecological assessment and ecological
modeling work to be completed in 1999. The information presented by the writer will be considered
during these efforts. Results of any food web modeling efforts (the writer's figures 18 through 21)
cannot be reviewed thoroughly without a complete report explaining those efforts in detail. The
discussions presented on congener ratios in fish may have some potential in tracing the pathways
of fish exposure. The USEPA anticipates examining these ratios in detail in the subsequent
ecological assessment and modeling efforts.
Response to DP-4:
This letter "supports the major conclusions in the DEIR" and presents no critiques of the
report requiring a response.
December 22. 1993
DEIR-1
TAMS/LTlTetraT«:h/MCA
-------
Response to DP-5:
This letter (comment DP-5.1) acknowledges the depth, scope, and scientific rigor of the
DEIR. No critiques of the DEIR requiring a response were presented in this letter.
Response to DS-2.27
Although additional data is always of some value, it would be difficult to collect and analyze
additional data without delaying the program schedule. In addition, the type of data required to
clarify the mechanisms for release is not easily obtained and may require an extended study period
to obtain sufficient information to significantly improve the understanding of these mechanisms.
Given these constraints, USEPA believes that the available data, in conjunction with that regularly
acquired by GE, will provide an adequate basis for the ultimate conclusions which will need to be
made for the Hudson River.
It should be noted that USEPA will not delay the issuance of any Reassessment reports in
order to accommodate data collected after December 1997. However, USEPA will address changes
in Hudson River conditions that are reflected in data collected between December 1997 and when
USEPA issues a Record of Decision (ROD) for the site in 2001 in the Final Responsiveness
Summary to be issued with the ROD.
Response to DS-2.28
USEPA concurs that long-term monitoring will be required vyithin the river regardless of the
alternative selected. USEPA will evaluate whether there is any type of monitoring program that
should begin prior to issuing the Record of Decision.
Response to DS-2.29
Interpretation of the low resolution coring data has been included in the July 1998 Phase 2
Low Resolution Coring Report (USEPA, 1998). The ecological data will be included in the
forthcoming Ecological Risk Assessment Report, currently scheduled for August 1999.
Response to DS-2.30
Bioturbation is considered a potentially important mechanism explaining P|CB measurements
in the TT Pool. See, for example, response to comment DG-1.3, below.
Since issuance of this letter, NYSDEC and USEPA have met on a regular basis to discuss
elements of the Hudson River PCBs RRI/FS.
Response to DC-1.1
Public participation has always been an important part of the Hudson River PCBs
Reassessment RI/FS. In fact, the Community Interaction Program (CIP) developed and implemented
in 1990 is entirely unique to USEPA, and was designed specifically to meet the particular needs of
this project.
December 22. 199$
10.0085
DEIR-2
TAMS/LTl/TetraTecWMCA
-------
Major reports are distributed according to a pre-determined schedule. A meeting is scheduled
at the time of the report's publication to provide an overview of what is in the document so that
interested parties who read it are acquainted with the contents and aware of the major points. After
a period of a few weeks, follow-up meetings or availability sessions are held so that people who have
read the report may meet with USEPA representatives to discuss it, ask questions, or offer
comments.
All members of the public receive the reports at the same time; therefore, the initial public
presentation would be confined to USEPA's introduction of the material. Different perspectives
would logically only be available for presentation after other parties have the opportunity to read the
report.
Response to DC-1.2
Each of the various reports presents findings and conclusions. The DEIR, for example,
presents data and conclusions relating to the fate and transport of PCBs in the Hudson River. Human
health and ecological risks will be addressed in forthcoming Phase 2 reports scheduled for 1999, and
remedial options for the Hudson will be addressed in the Phase 3 Feasibility Study Report, scheduled
for 2000. These results are typical of a Supefund remedial investigation and feasibility study. Only
here they have been broken up into many individual components because of the size of the effort
involved on this project.
While the analysis detailed in the DEIR found that the PCB water column load at Rogers
Island had dropped to undetectable levels as a result of the remedial actions performed at Allen Mills
Site, the PCB loading between the Rogers Island and the TI Dam has remained relatively constant
from 1993 to 1996. This indicates that the TI Pool is a constant source of PCBs to the water column.
The source appears to be PCBs stored within the sediments. There also are significant gains to the
water column between the TI Dam and Waterford, NY, again with PCBs stored within the sediments
being the main contributor.
Response to DC-1.3
Since the beginning of the Reassessment, EPA has made a particular point of soliciting input
and opinions from any and all interested parties, and will continue to do so until the project is
complete. All input provided is reviewed and considered. The CIP especially was designed to
provide opportunities for direct participation on multiple levels. Further, in addition to running
multiple meetings over the last eight years where comment is always solicited, EPA has hosted
Availability Sessions on particular reports to answer questions and receive comments. Finally, on
a number of occasions EPA has extended standard comment periods to accommodate the public.
The official comment period for the DEIR Report was through April 11,1997. The Scientific
and Technical Committee (S.T.C.) meeting for the DEIR was on March 25, 1997, within the
comment period for that report. USEPA assumes therefore that the writer's concern about the
comment period's being closed prior to the Scientific and Technical Committee meeting referred not
to the DEIR but to the Preliminary Model Calibration Report (PMCR), released in October of 1996,
The S.T.C. meeting for that was held on March 26. 1997. To address the writer's concern, we refer
first to the cover letter of the DEIR. which states:
Dccanbn 32. I«I
DEIR-3
T AMS/LTO etraTeth/MCA
-------
"EPA will be accepting comments on the Data Evaluation and Interpretation Report until
April 11, 1997. The comment period is longer than we have provided for previous
Reassessment reports because of the extensive analyses that are included and the complexity
of those analyses. In addition, there are several findings discussed in this report which were
utilized in the Preliminary Model Calibration Report (released in October 1996). Therefore,
during this comment period, EPA will also accept comments on the Preliminary Model
Calibration Report as they pertain to findings from the Data Evaluation and Interpretation
Report."
Further, because these reports are a series and in essence build upon each other, USEPA
remains willing to entertain questions from members of the public on prior reports.
Response to DC-2.1
This report does not present or evaluate remedial options. A variety of options, including "No
Action" and Natural Attenuation as well as dredging, will be considered during the Feasibility Study
(FS) process.
Response to PL-1.5
This comment explicitly repeats comments made on the Hudson River models. As such, it
is addressed as part of the PMCR. This issue will be further addressed in the Baseline Modeling
Report (to be released in May 1999).
Response to DP-2.1 and DP-2.3
The safety of the drinking water supplied from an infiltration gallery located near the Hudson
River is under the jurisdiction of local and/or State Departments of Health. Risk from such intakes
depend on a variety of design specifications that need to be reviewed on a location-specific basis.
Risk from direct intakes of Hudson River water (/.e.,not from an infiltration gallery) will be included
in the Human Health Risk Assessment.
Response to DP-2.2
This comment cannot be addressed by the Hudson River PCBs Reassessment. This question
should be addressed by the designers of the water supply system.
December 22 I'm
DE1R-4
TAMS/LTI/TetraTtth/MCA
-------
Chapter 1 - Introduction
1.1 Purpose of Report
1.2 Report Format and Organization
No significant comments were received on Sections 1.1 and 1.2.
1.3 Technical Approach of the Data Evaluation and Interpretation Report
Response to DC-1.5
Water column data spanning over two decades, not just 1993 data, were examined in the
DEIR in various analyses. The 1993 sediment and water column data collected for the Phase 2
reassessment is used in the DEIR to analyze the geochemical principals of the system (such as
gaining an understanding of the mechanisms of PCB transport). In addition. General Elec trie's water
column monitoring program data, which began in 1990 and continues to date on a bi-monthly basis,
are used to quantify the PCB loadings above the TI Pool and in the TI Pool over time. USGS water
column data were also used for this purpose in the river below the TI Dam. Data from the 1984
NYSDEC Sediment Survey are used to estimate the total mass contained in the TI Pool by using
polygonal declustering and geostatistical techniques. An emphasis may have been placed on the
1993 data collected by USEPA for the detailed understanding of the system this data provided, but
these are clearly not the only data explored in the DEIR.
1.4 Review of the Phase 2 Investigations
No significant comments were received on Section 1.4.
1.4.1 Review of PCB Sources
No significant comments were received on Section 1.4.1.
1.4.2 Water Column Transport Investigation
Response to DC-4.6
Each of the three sources of variance mentioned (variance shown in analyses from the site;
variance caused by sampling methods: and known physical sources of variance including cross
channel and vertical inhomogeneity in the PCB distribution in the river) are examined separately
below. The Phase 2 data were generated in such a way as to minimize unwanted sources of
variability so that the actual trends in the data would not be obscured.
For variability shown in analyses from Rogers Island, Phase 2 data are compared to General
Electric data for the same period. Figure 3-105 of the DEIR shows monthly PCB loads from above
Bakers Falls, Bakers Falls to Rogers Island, and Rogers Island to the TI Dam. The Phase 2 flow-
averaged estimates agree well with the General Electric estimates in both total load and distribution
for July through September 1993 (post construction at the Bakers Falls area).
December 22
DEIR-9
TAMS/LTL/TelraTech/MCA
-------
The precision in the Phase 2 sampling methods is determined by comparing the split sample
data. Figure DC-4.6 shows the relative fractional differences of the ten split samples analyzed for
total PCBs taken during the flow-averaged and transect events. Although the distributions are right
skewed, the median values are low at 0.10, 0.13, and 0.13 for the dissolved, suspended and whole
water samples, respectively. Each sample required a large volume of water in order to achieve the
low quantitation limits for PCBs, which in turn necessitated a long sampling period. This may be
the cause of the occasional high relative fractional difference.
The impact of physical sources of variance due to cross channel and vertical inhomogeneity
in PCB distribution in the Hudson River is shown in Figure 4-22 of QEA's March 1998 report,
Thompson Island Pool Sediment PCB Sources. General Electric's routine composite sample of east
and west channels at Rogers Island is compared to shallow and deep samples taken at six stations
in a river cross section just upstream of Rogers Island. This was performed on 2 separate occasions.
While there are differences among the samples, it is clear that the routine sample provides a
reasonable estimate of the Hudson River PCB concentration at this station. The routine sample is
comparable to the Phase 2 sample at Rogers Island which was stationed before the river splits into
east and west channels.
1.4.3 Assessment of Sediment PCB Inventory and Fate
No significant comments were received on Section 1.4.3.
1.4.4 Analytical Chemistry Program
No significant comments were received on Section 1.4.4.
December 22. 1998
DEIR-10
TAMS/LTLTetra Tech/MCA
-------
Responses to Specific Comments on the Deir
Executive Summary
Response to DP-1.1
The time of depletion of the PCB inventory from the Thompson Island Pool will be more
rigorously examined in the forthcoming Baseline Modeling Report.
However, the last sentence of the Executive Summary, "[t]he time for depletion appears to
be on the scale of a decade or more and will be investigated further through the planned computer
simulations," is not in contradiction to the suggestion that the residence time of PCBs in the
Thompson Island Pool could be a several decades in duration.
Response to DP-3.1
This comment addresses potential contamination of the Town of Bethlehem water supply
drawn from a horizontal well or infiltration gallery on the west side of the Hudson River opposite
the Village of Castleton and determined to be ground water under the influence of surface water.
This well is south of Albany, and no direct evidence on pore water concentrations in this area has
been collected. However, total PCB concentrations in sediment in the Albany Turning Basin do
suggest the possibility that pore water concentrations in this area might exceed MCLs.
While the issue raised is a valid one, it is outside the scope of the DEIR. This issue will be
addressed in the Human Health Risk Assessment to the extent possible although without specific
monitoring data any risk analysis is likely to be highly uncertain. The commenter should also
consider raising these issues to the designers of the of the water supply system.
Response to DC-4.4 and DG-1.1
The analogy of a pipeline transport process mentioned in the executive summary has been
withdrawn in light of the revision to Upper Hudson flows noted elsewhere in this report. (See
corrections to DEIR Section 3.2.2, below.) However, the results as they are currently understood do
not suggest extensive loadings entering the water column in the region below the TI Dam, with the
possible exception of the region between the TI Dam and Schuylerville. To place the regions in
perspective, the TI Pool, a distance of only six miles, produces a water column inventory gain of 1
to 2 pounds per day during low periods. The 25-mile stretch between Schuylerville and Waterford
showed no measurable net load gain and is expected to show a substantive PCB loss when the load
calculations are revised. Thus, while downstream sediments undoubtedly make some contribution
to the water column PCB inventory, it is clear that they do not contribute at the same level as those
of the TI Pool. Data recently collected by GE appears to confirm the suggestion made in the DEIR
that additional loads may be generated between the TI Dam and Schuylerville.
However, this does not change the fact that the region above Schuylerville has been and
continues to be the primary source of PCBs to the fresh water Hudson. Water column congener
patterns support this contention as does the sediment PCB inventory which is concentrated above
December 22 l»S
DEIR-5
TAMS/LTLTelraTech/MCA
-------
Schuylerville (35 of the 40 hot spots are found in this region representing roughly 85 percent of the
fine-grained sediment PCB inventory). Essentially all of the sediment PCB contamination in the
Upper Hudson is attributable to GE so that if some of the water column load is generated below the
TI Dam, it is merely the re-release of PCBs which originated above the TI Dam. The possibility that
a small portion of the sediment PCB inventory at Albany (reported as 22 percent on page 3-138 of
the DEIR) may be generated external to the GE-related loads still does riot change the fact that the
GE-related loads to Albany represent 78 percent of the load at Albany, which can arguably be
considered the "primary" source. It is also important to note that the congener pattern analysis in
the cores did not consider weathering during transport, a process which would serve to shift the
congener pattern toward heavier mixtures and incorrectly suggest the input of another source. Note
that the water column load analysis as well as the PCB/137Cs analysis were characterized as having
an uncertainty of 25 percent. This does not implicitly mean that local additions at 25 percent of the
Upper Hudson load were present. Thus the conclusion is correct as written.
Response to DG-1.2
The data analyzed in the report covers nearly three years (1993 - 1996) after the initial
controls at the Allen Mills were in place. These data, as well as subsequent data collected, show that
water column loads during low flow periods, particularly those from May through November, are
produced within the TI Pool as a result of sediment release. The latter years of sampling have shown
further reduction in the load originating above Rogers Island, to the point where its annual
contribution has become negligible relative to that of the TI Pool. Also notable in the GE data is the
consistency of summertime loads produced by the TI Pool despite the abatement of the loads
produced upstream. Thus the conclusion correctly characterizes conditions in the river. In addition,
the statement correctly places the emphasis on the TI Pool and recognizes that river PCB loads will
be controlled almost exclusively by the sediments in the foreseeable future. Thus, while
understanding the impacts of GE's ongoing remedial activities is certainly useful, their importance
has waned relative to the demonstrably larger loads now originating from the TI Pool.
Response to DG-1.3
The conclusion (i.e., the PCB load from the Thompson Island Pool originates from the
sediments within the Pool) is distinctly important in that it places the responsibility for the TI Pool
load squarely on the sediments of the Pool and not on any other load source. While the analysis did
not permit the resolution of the exact mechanism or sediment type responsible for the load, the
analysis very clearly showed that the loads from the sediments were distinct from those originating
upstream of the Pool. While the GE releases are undoubtedly the original source of the PCBs in the
TI Pool sediments, it is now these sediments therhselves, and not the upstream inputs, which are
responsible for the water column loads throughout much of the year.
As far as the mechanism responsible for the release of PCBs from the sediments, there are
several possibilities, including resuspension, porewater diffusion, groundwater movement, biological
activity (see. for example, Templer et al., 1997) as well as any combination of the above. The
purpose of the DEIR was to provide an interpretation of the data, in part, as a guide for the
subsequent modeling analysis. GE's contention that no realistic mechanism exists to resuspend fine-
grained sediments is based largely on modeling arguments, not measurements, and it ignores
processes such as biologically-driven resuspension and recreational water craft use. Recent water
December 22. 1998
DEIR-6
TAMS/LTI/TetraTech/MCA
-------
column data show the TI Pool load to be largely constrained to the period May to November, the
principal period of both biological activity and recreational boat use. Additionally, given that the TI
Pool sediments represent a broad range of sediment contamination with a similarly wide range of
dechlorination, it is most likely that a combination of mechanisms and sediment types are
responsible for this load.
GE's contention that a portion of the TI Pool load originates with undetected oil droplets is
unfounded. Despite GE's many attempts to find such droplets, none have been detected. In fact,
some of their most recent results from a sampling cross section of the river just above Roger Island
(QEA, March 1998) shows the water column PCB concentration to be relatively homogeneous,
suggesting the absence of oil droplets. (Presumably, oil droplets near the bottom would cause the
deeper samples to yield markedly higher PCB levels.) Even if such droplets were to exist, it is
unclear how long it would take for these PCBs to leave the sediments and re-enter the water column.
Clearly, nearly all the PCBs present in the bottom sediments were once released as oil droplets in
GE's discharges. As discussed in the Low Resolution Sediment Coring Report (USEPA, 1998),
much of the sediment burden, though clearly not all of it is still in place, some at depth, most within
9 inches of the surface. The sediments of the TI Pool are clearly not stagnant, lake-like deposits but
rather a dynamic environment subject to resuspension and burial as well as diffusive and biological
processes. Thus, the simple addition of more PCBs during the period from September 1991 to 1996
serves to worsen the pre-existing problem but certainly does not define it. The strongest evidence
for this fact comes from the GE data itself, which demonstrates a measurable TI Pool input prior to
the September 1991 event as well as the consistency of the size of the TI Pool load each year from
1993 to the present despite the major reductions in the loads from upstream of the pool. Lastly, the
congener patterns of the TI Pool load are not consistent over time, as might be expected from a
single source type (i.e., oil droplets on sediments) but rather they vary as might be expected given
a variable mixture of sediments and processes. Thus the need for this conclusion, i.e., that the
sediments, and not any other phenomenon, are responsible for the TI Pool load.
Response to DG-1.4
The issue of the effect of dechlorination on PCB toxicity will be discussed in the Ecological
and Human Health Risk Assessments and is not appropriate for this report as it deals with PCB
geochemistry. As to the effect of dechlorination on PCB bioaccumulation, it is true that the
dechlorination products tend to bioaccumulate less, consistent with the decrease in the partition
coefficient. However, the lower partition coefficient also yields greater mobility for the partly
dechlorinated molecule, thereby increasing the potential for migration from the sediments to the
water column and possible biological exposure. Thus it is unclear as to the net effect of
dechlorination on bioaccumulation. This issue will be examined in greater detail in the ecological
risk assessment an8 associated modeling. The writer is reminded, however, that most sediment PCBs
mixtures do not experience extensive degrees of dechlorination due to their relatively low
concentration.
General Electric has consistently argued that the Upper Hudson acts essentially like a lake,
gradually and steadily burying older sediments with newer ones. This is simply not the case. The
Upper Hudson, while not a free-flowing river, is still a dynamic system, with sediment scour and
resuspension occurring on a regular basis. Thus there is no guarantee of burial as a means to
December 22. 1998
DEIR-7
T AMS/LTl/TrtraTech/MCA
-------
sequester contaminated sediments. This was demonstrated in the Low Resolution Coring Report,
which showed sediment inventory losses in a large number of historical study areas.
The following are responses to the writer's interpretations of the four 'central positions' of
the DEIR:
1. During periods of low flow, a portion of the PCBs entering the TIP from above
Rogers Island may be stored in the sediments of the TI Pool. This is based on the apparent
loss of several congeners from the Rogers Island load relative to the load at the TI Dam. In
light of current results obtained from the TI Dam, this phenomenon may also be caused by
incomplete mixing of the main channel and near-shore flows within the river, wherein the
upstream load is not completely detected at the TI Dam.
2. As stated in the DEIR on page 3-148, the TI Pool load may be produced by a number
of sediment sources, including relatively-less dechlorinated, low PCB level sediments
(regardless of age) as well as highly dechlorinated sediments, with concentrations of roughly
at least 30 mg/kg or higher, typical of sediments deposited prior to 1984. Some combination
of these sources is also plausible.
3. PCBs from the Upper Hudson clearly represent the major source of PCBs to the
entire freshwater Hudson. During low flow conditions, PCBs derived from Upper Hudson
sediments, particularly those of the TI Pool, represent the dominant source of PCBs to the
Upper Hudson and therefore to the Lower Hudson as well. Lower Hudson sediments (much
of whose contamination is also derived from the Upper Hudson) may add to the water
column PCB load borne by the water column but the Upper Hudson load still represents the
main external load to the Lower Hudson, regardless of flow conditions.
4. The writer has correctly summarized the USEPA's interpretation.
5. The writer has correctly summarized the USEPA's interpretation.
December 22, 1998
DEIR-8
TAMS/LTI/TelraTech/MCA
-------
c
3
O
4 -r
3
2
1 -t
0
0
Dissolved Fraction
0.5
Range
Minimum
0 0093
Maximum
0 4321
Mean
0.1574
Median
0.(003
1.5
c
3
o
U
c
3
o
U
5
4
3 -I-
2
5 -
4
3
2
1 4
0
0
Suspended Fraction
0.5
Range
0.5
4^
Range
Minimum
00436
Maximum
1.0462
Mean
0 2996
Median
0 1286
1.5
Whole Water
Minimum
0.0295
Maximum
0 9527
Mean
0.2112
Median
0.1290
1.5
Hudson River Database Release 4 I
Figure DC-4.6
Relative Fractional Difference for Phase II
Water Column Split Samples (Total PCBs)
TAMS/LTI/TctraTcch/M(
-------
Chapter 2- PCB Sources to The Upper and Lower Hudson River
2.1 Background
No significant comments were received on Section 2.1.
2.2 Upper Hudson River Sources
No significant comments were received on Section 2.2.
2.2.1 NYSDEC Registered Inactive Hazardous Waste Disposal Sites
Response to DS-2.1
The second to last sentence on page 2-3 should be changed to: "The significance of the
contaminated site is clearly evident in the NMPC Phase II fish tissue data..The commentor noted
that "remediation of the shoreline has been completed and preliminary fish monitoring data indicates
improvement."
Response to DS-2.2
For the first paragraph on page 2-14, the commentor suggested that the mean monthly
maximum flow (174 gpm, December 1991 through April 1994) cited on page 2-15 should be used
to estimate the seepage loading. Thus, at a total PCB concentration of 20 |ig/L, this flow (174 gpm
or 0.4 cfs) would result in a loading of 0.02 kg/day (0.04 lb/day) or 7 kg/year (15 lb/year).
Response to DS-2.3
The commentor noted that although the GE Fort Edward Plant Site is listed as a "dump" in
NYSDEC's Registry of Inactive Hazardous Waste Sites, the plant is currently an operational
capacitor manufacturing facility.
2.2.2 Remnant Deposits
Response to DS-2.4
As indicated in the 1984 Record of Decision for the site, USEPA will evaluate whether
further remedial action is appropriate for Remnant Deposits 2-5 if the Agency decides to take
additional remedial action with respect to sediments in the river.
Response to DF-2.7
Oversight of the activities and documentation associated with the GE facilities has been
performed by NYSDEC. Chapter 2 of the DEIR provides a summary of those activities and
documents. USEPA's assessment of the significance of the GE facilities and the Remnant Deposits
in relation to the sediments of the Upper Hudson River is presented in Chapter 3 of the DEIR. As
part of the Phase 2 geochemical and modeling efforts, the data have been and will continue to be
December 22 1991
DEIR-12
T AMS/LTLT etiaT ech/MC A
-------
critically reviewed in determining the current and projected PCB loading from all of the sources
above the Thompson Island Pool (i.e., above Rogers Island).
Figure DF-2.7 provides a comparison of the PCB homologue distribution in both the non-
aqueous phase liquid and water phases of seepage samples collected in May 1993 from the eastern
raceway area adjacent to the GE Hudson Falls facility with the USEPA Phase 2 Aroclor 1242
standard as well as the whole water sample collected at Rogers Island during the spring runoff event
of 1993. As can be seen, the sample patterns closely resemble the standard, suggesting that the
mixture is primarily Aroclor 1242. However, as shown by the H, H' congeners in Figure DG-1.20B
(in Section 4.3.2), there are still significant differences between the Rogers Island station and pure
Aroclor 1242. The results shown in Figure DG-1.20B show substantially higher levels of three of
the four congeners presented, indicating the presence of a heavier Aroclor mixture (e.g., Aroclor
1254) in the GE releases.
2.2.3 Dredge Spoil Sites
No significant comments were received on Section 2.2.3.
2.2.4 Other Upper Hudson Sources
.Vo significant comments were received on Section 2.2.4.
2.3 Lower Hudson River Sources
Response to DF-2.8
Sediments downstream of the Thompson Island Pool (TI Pool) represent a potentially
important source for PCB release to the water column. In fact, based on additional data collected by
GE after the release of the DEIR. it appears that the sediments between the TI Dam and Schuylerville
produce an additional PCB load roughly equal to about one half of that generated by the TI Pool
(QEA. 1998). This is based on the apparent load gain between the TI Dam and Schuylerville
documented by the GE samples.
The importance of the sediments below the TI Dam will be further evaluated as part of the
fate and transport modeling analysis currently ongoing. The model has been configured to represent
geochemical PCB transport from Rogers Island in the TI Pool to Waterford.
Response to DF-2.9
Ongoing sample collection efforts by GE and NYSDEC serve to provide much of the
information requested by the writer, specifically water column and fish data, respectively. Data to
clarify the nature of the PCB release processes are more difficult to obtain. Specifically, it is
difficult to collect samples which can integrate the individual micro- to macro-scale processes
currently responsible for the release of PCBs from the sediments. Without data to individually
constrain these processes, it is difficult to estimate their magnitudes. In their stead, the low resolution
coring program examined the integration of all these processes on the sediment inventory and
demonstrated a statistically significant loss from the most contaminated sediments. Additional data
December 22 1998
DE1R-13
TAMS/LTI/TetraTech/MCA
-------
(I 50
(l -15
li -ill
11 (5
O (I Ml
u
U.
(I 21
S <> 2"
II 15
li III
(I (15
(I IJIl
M>>iu>
In Iclr.i
liomolo);uc
IVnt.i
Nil I' ll(()ll .). |tl»CH| 941,976 m^l.
Si l l' 10(011 ). |llfH| - 94,040 my/I.
SI I I' 7(1111 ). |llt H| - 9K.220 mg/1.
SKI !> 7(1121 >), |il«C»| = 45 ntfl.
Ariiclor 1242. Itvisc 2 Standard
Rogers Island Transect 4 Whole Water
Souac: tiLi liaktf* Kails Investigation D»u in '1'AMJMrradtcnl Database TAMS/LTl/TetraTech/MCA
Figure DF-2.7
General Electric Hudson Falls Source - Seepage Homologue Distribution, May 1993
-------
may be helpful but studies of the individual processes have the potential of requiring very long
investigations and monitoring periods before useful results are obtained.
2.3.1 Review of Phase 1 Analysis
2.3.2 Sampling of Point Sources in New York/New Jersey (NY/NJ) Harbor
2.3.3 Other Downstream External Sources
No significant comments were received on Sections 2.3.1 through 2.3.3.
December 22 1998
DEIR-15
TAMS/LTl/TetraTech/MCA
-------
Chapter 3 - Water Column PCB Fate And Transport in The Hudson River
Response to Comment DF-2.3A
Further resolution of the sample data into congener-specific loads would provide greater
clarification on some aspects of PCB transport in the Upper Hudson. However, USEPA recognizes
that for the purposes of the DEIR, the homologue patterns presented in the report are sufficient to
support the conclusions drawn. Subsequent analysis anticipated as part of the fate and transport
modeling effort will also advance this issue farther since it will deal specifically with five individual
congeners, representing a range of PCB characteristics. Additional congener-specific interpretation
may also be completed during the Ecological and Human Health Risk Assessments if deemed
necessary.
Response to Comment DF-2.3B
The writer is correct in noting that BZ#118 alone was not used to identify the presence of
Aroclor 1254. Although not explicitly stated in the text, the presence of Aroclor 1254 was inferred
based on the ratio of BZ#118/BZ#52. The presence of Aroclor 1254 was indicated when this ratio
was well beyond that found in Aroclor 1242. Thus, unlike the eight unique congeners identified for
Aroclor 1260, no unique congeners specific to Aroclor 1254 could be discerned which could indicate
the presence of this Aroclor unequivocally. The amount of Aroclor 1254 in a sample was estimated
based on the degree to which this ratio exceeded that found in Aroclor 1242. The writer is also
correct in noting that for a mixture of 85 percent Aroclor 1242 and 10 to 15 percent Aroclor 1254,
each Aroclor would contribute roughly equally to the total amount of BZ#118. Nevertheless, the fact
that the BZ#118/BZ#52 ratio was well beyond that which could be attributed to Aroclor 1242 is
fairly clear proof of the presence of Aroclor 1254 in the sample. This indication is further supported
by the reported history of GE's PCB usage (Brown el a!,, 1984). GE usage of Aroclor 1254 began
in 1945 at Ft. Edward and continued at both plants until 1971. (This information is summarized in
Figure B.2-1 of the Phase 1 Report (USEPA, 1991).) Thus the presence of Aroclor 1254 in the PCB
contamination of the Hudson is highly likely.
3.1 PCB Equilibrium Partitioning
Response to DO-1.23
The comment noted that there appeared to be a difference in partitioning behavior for
samples in the Remnant Deposit and Rogers Island data compared to stations downstream and stated,
"it is incorrect to include Remnant and Rogers Island data in the determination of equilibrium
partition coefficients. Estimates including this data [sic] will yield partition coefficients well above
equilibrium".
USEPA is in full agreement that some samples from the Rogers Island station show higher
estimated partition coefficients than samples collected downstream, but disagrees with the
conclusions stated in this comment. While it now appears likely that some partition coefficient
estimates at Rogers Island are biased high, this fact was not known a priori. Indeed, because field
partitioning is often observed to differ systematically from laboratory KqW estimates, there is no clear
baseline against which to determine if a given estimate is biased high or low. It was USEPA's
Decanter 22. 1991
DEIR-16
TAMS/LTl/TetraTech/MCA
-------
judgement that any pre-selection of the data to eliminate undesirable sample observations would
subject the work to a criticism of bias. Therefore, the full set of data was used. However, because
it was judged likely that some samples were out of equilibrium, central tendency of the estimates
was summarized by the median (50th percentile), rather than the mean. The median is relatively
insensitive to the presence of biased outlying values, particularly in this case where only one out of
eight stations used for the analysis deviates strongly from the central tendency. The approach
USEPA has taken therefore does not yield partition coefficients well above equilibrium, because the
median from samples at eight stations is used, yet avoids an arbitrary prejudgement of the data.
GE's discussion also neglects the effect of changing POC concentrations between stations.
GE's Figure E-l displays variation in KP with station, implying a large upward bias at all stations
upstream of Thompson Island Dam. In fact, a part of the variability is due to variations in POC
concentrations, and corrections must be made for POC concentration. Figure 3-16 in the DEIR
shows that it is only samples from the Rogers Island station which exhibit a strong upward bias
relative to mean Kpo^ estimates for higher-chlorinated homologues. GE's Figure E-l lumps partition
coefficient estimates across all congeners. As there is significant variability in partitioning among
individual congeners, the position of the bars in this figure is determined as much by what congeners
happen to have been quantitated in both dissolved and particulate phases at a given station as by
variability between stations. Finally. GE's Figure E-l also misleadingly includes estimates from
Stations 1 and 2 (upstream of the PCB source area, with concentrations generally too low for
accurate determination of partition coefficients) and from Station 10 (Lock 7), which does not reflect
conditions within the river.
GE (p. E-3) raised several issues regarding temperature correction of partition coefficients.
GE first questioned why the data of Warren et al. (1987) was used to determine the temperature
dependence of partition coefficients when the Phase 2 data collected covers a sufficient range in
temperature to determine temperature dependence directly. The reasons for this choice should be
evident from the discussion in the DEIR: Because some samples were believed to be out of
equilibrium (including, but not necessarily limited to, samples from the Rogers Island station), while
estimates from other samples may be affected by analytical uncertainty, it was highly desirable to
use estimates of the temperature correction factor derived from the independent controlled laboratory
experiments (using Hudson River sediment) conducted by Warren et al. If only the Phase 2 data
were used for this calculation, it would not be possible to isolate the effect of temperature from other
sources of variability in partition coefficient estimates.
GE also stated that the temperatures used by EPA are not ambient, and over estimate actual
in situ temperatures (Figure E-5). In fact, the Phase 2 temperatures plotted by GE in Figure E-5 are
not those used for establishing partition coefficient temperature correction factors. Temperatures
which were used in the DEIR also appear to be higher than in situ temperatures, but were selected
intentionally, as described below.
True in situ water temperatures were not recorded during the Phase 2 data collection effort,
and water temperatures reported in connection with the various physical parameter measurements
had adjusted to some degree toward ambient air temperature. Release 4.1 of the Hudson River
RRI/FS Database now contains best estimates of in situ water temperatures based on examination
of sampling logs and determination of which of the reported temperatures were recorded closest to
time of sampling, which were not available in time for the DEIR. Filtration of samples in the field
December 22 1998
DEIR-17
T AMS/LT1/T etraT ech/MC A
-------
occurred up to four hours after sample collection, however, allowing some potential re-equilibration
between phases in response to temperature changes in the sample. The temperatures used for
calculation of partition coefficients in the DEIR were an estimate of temperature at time of filtration,
derived from temperatures recorded during measurement of conductivity. Some refinements to these
temperature estimates may be possible based on review of the sampling logs. Additional
investigations of this issue are ongoing.
3.1.1 Two-Phase Models of Equilibrium Partitioning
3.1.2 Three-Phase Models of Equilibrium Partitioning
3.1.3 Sediment Equilibrium Partition Coefficients
3.1.4 Summary
No significant comments were received on Sections 3.1.1 through 3.1.4.
3.2 Water Column Mass Loading
Correction to Section 3.2 - Water Column Mass Transport
Since the completion of the DEIR, several additional analyses as well as new or revised data
sets have been reviewed which indicate the need to revisit the water column PCB mass loading
estimates derived in the DEIR. Specifically, additional data analysis on Upper Hudson flows along
with two reports produced by GE have led to the following observations:
° Upper Hudson flows at Stillwater and Waterford for low flow conditions estimated
for the Phase 2 analysis may be 15 to 40 percent too high.
° GE PCB data obtained on water column concentrations through 1996 under reported
PCB concentrations by roughly 40 percent. Revised data are now available.
° GE sampling in the vicinity of the TI Dam suggests that some TI Dam samples
collected at low flow conditions in the absence of loadings above Rogers Island may
overestimate the water column load at the dam. For the five-year period of GE data
collection prior to 1996, the results suggest the values may be too high by 20 percent.
During 1996 and 1997, the low-flow estimates may be 36 percent too high. No
corrections are required for flows higher than 4000 cfs prior to 1996. These
corrections account for both flow and Rogers Island load which are shown to affect
the sampling bias. (See the discussion in Section 1 of the USEPA review of the
GE/QEA model in book 3 of this responsiveness summary.)
As a result of these observations, the conclusions concerning the water column transport need
to be reviewed for their accuracy. The main conclusion for the water column transport were given
in the executive summary of the DEIR as follows:
December 22. 1998
DEIR-18
TAMS/LTI/TetraTech/MCA
-------
1. The area of the site upstream of the Thompson Island Dam represents the primary
source of PCBs to the freshwater Hudson. This includes the GE Hudson Falls and Ft.
Edward facilities, the Remnant Deposit area and the sediments of the Thompson
Island Pool. Analysis of the water column data showed no substantive water column
load increases (i.e.. load changes were less than ten percent) from the Thompson
Island Dam to the Federal Dam at Troy during ten out of twelve monitoring events.
These results indicate the absence of substantive external (e.g., tributary) loads
downstream of the Thompson Island Dam as well as minimal losses from the water
column in this portion of the Upper Hudson. These results also indicate that PCB
transport can be considered conservative over this area, with the river acting basically
as a pipeline (i.e., most of the PCBs generated upstream are delivered to the Lower
Hudson). Some PCB load gains were noted during spring runoff and summer
conditions, which were readily attributed to Hudson River sediment resuspension or
exchange by the nature of their homologue patterns. These load gains were notable
in that they represent sediment-derived loads which originate outside the Thompson
Island Pool, indicating the presence of substantive sediment inventories outside the
Pool. The Mohawk and Hoosic Rivers were each found to contribute to the total PCB
load measured at Troy. The loading from each of these rivers during the 1993 Spring
runoff event could be calculated to be as high as 20 percent of the total load at Troy.
However, these loads represent unusually large sediment transport events by these
tributaries since both rivers were near or at 100-year flood conditions.
This conclusion is based on a number of lines of reasoning but it is clear that the downward-revised
flows at Stillwater and Waterford will yield lower PCB fluxes at these locations. Thus the statement
above that "PCB transport can be considered conservative over this area" does not apply at low flow
conditions in the sense that a net loss of PCB transport will be apparent during these conditions. The
finding that the PCB load generated above the TI Dam is essentially equal to that delivered by the
Upper Hudson to Waterford will only apply at high flow conditions. Thus, while it is likely that most
of the PCBs which cross the dam at Waterford at low flow will have been released from the area
above the dam. a substantial portion of the PCBs released will not be transported to Waterford.
Presumably this loss is attributable to processes like settling, gas exchange and aerobic degradation.
This result will not change the finding that the region above the TI Dam is the primary PCB source
to the fresh water Hudson since, even with these reduced loads, this region will still easily present
the largest load. In addition, the dated core evidence shows this region to be the dominant PCB
source to the fresh water Hudson as well.
The revised GE data more than compensate for the potential overestimation of PCB loading
by the TI Dam station since the corrections to the TI Dam loads are dependent on other factors such
as the total flow and the Rogers Island load. Thus the overall result will be to yield higher loads from
the TI Pool based on the GE data than previously estimated. An important finding becomes evident
from analysis of the GE data. That is, the sediments between TI Dam and Schuylerville may also
contribute significantly to the water column PCB load. The revised quantitation of these loads will
be completed in the near future.
December 22 1998
DEIR-19
TAMS/LTl/TetraTech/MCA
-------
3.2.1 Phase 2 Water and Sediment Characterization
No significant comments were received on Section 3.2.1.
3.2.2 Flow Estimation
Correction to Section 3.2.2 - Flow Estimation
As part of the original data analysis, estimates of flow in the Upper Hudson between
Schuylerville and Waterford were required due to the loss of USGS discharge stations at Stillwater
and Waterford. These stations were lost due to construction activities in the corresponding areas of
the river. In order to obtain flow estimates, staff gauge data from the NYS Champlain Barge Canal
system were analyzed and used to estimate flow, as reported in the DEIR.
Since the original preparation of the DEIR, additional information concerning Hudson River
flows between Schuylerville and Waterford during 1993 have become available. Specifically, USGS
estimates of water flow based on other data for the region have been published as well as additional
information on the nature of the modifications made to the dams in this region of the river.
Additional information concerning the barge canal staff gauge data was also obtained which
suggested that several of the staff gauges used in the flow analysis may have been affected by the
construction as well.
A thorough review of this information has shown that some of the original Phase 2 flow
estimates reported in the DEIR for the Hudson between Schuylerville and Waterford are probably
incorrect, representing overestimates of the actual flow. In general, it was found that the USGS and
Phase 2 flow estimates agreed at high flow conditions but that the Phase 2 results were 15 to 40
percent higher than those derived by the USGS at low flow conditions.
The subsequent choice of the "correct" flow records was based on the precipitation records
for the Upper Hudson area. A comparison of the flow and precipitation data was prepared which
established the relationship between flow and precipitation for the historical data prior to the
construction begun in 1993. This result is shown in DEIR Figures 3.2.2 A and B. These figures
compare the incremental flow gain between the USGS measurement stations at Fort Edward and
Stillwater with two different records of precipitation in the Upper Hudson area. The vertical axis in
each graph represents the mean June-to-September flow gain between Fort Edward and Stillwater.
Also shown on these graphs are the estimates for 1993 based on the Phase 2 and USGS analyses.
It is clear from these diagrams that the USGS estimates are in closer agreement with the historical
relationship than is the Phase 2 estimate. In light of this finding as well as the information
concerning the dam construction work in the region, the USGS estimates of flow at Stillwater and
Waterford will be used in subsequent Reassessment analyses in place of the Phase 2 flow estimates
reported in the DEIR.
As a result of this finding, the PCB fluxes calculated in the DEIR must be revised to reflect
the lower flow rates in the lower portion of the Upper Hudson. These calculations will be presented
as an appendix to the Responsiveness Summary for the Low Resolution Coring Report, as mentioned
previously.
Decanter 22. 1999
DEIR-20
TAMS/LTVTetr* Tech/MCA
-------
o
V
jo
"H
«
£
•o
w
as
£
C/D
1800
1600
1400"
1200'
1000
800
600
400
200 ^ r
y = 92.1 + 74.3x R = 0.223
USGS 1993
¦ ¦
t—i—|—i—i—r
i I i i
"i—i—r
I I
6 8 10 12 14 16 18
Cumulative Precipitation at Glens Falls, NY (inches)
Legend:
¦ USGS Stillwater - Ft. Edward. June - August Average
# TAMS Stillwater - Ft. Edward, June - August Average
X USGS Stillwater - Ft. Edward. 1993-1995
Regression for 1984-1992
Note:
Data retrieved from NOAA's National Climatic Data Center (NCDC)
web site (http://www.ncdc.noaa.gov/ncdc.html).
Hudson River Database Release 4.1 TAMS/LTI/TetraTech/MC
Figure 3.2.2A
Fort Edward to Stillwater Incremental Summer Average Flow vs.
Total Precipitation for Glens Falls
-------
u-
"H
c3
£
T3
W
l-i
V
£
cn
1800
1600
1400
1200
1000
800
600
400
200
y = -422 + lOlx R = 0.29
USGS 1993
M
i—rn—i—i—|—i—r
i i I i i i i i i i | i i i
6 8 10 12 14 16 18
Cumulative Precipitation for NY-Division 5 (inches)
Legend:
¦ USGS Stillwater - Ft. Edward. June - August Average
# TAMS Stillwater - Ft. Edward, June - August Average
X USGS Stillwater - Ft. Edward. 1993-1995
Regression or 1984-1992
Note:
Data retrieved from NOAA's National Climatic Data Center (NCDC)
web site (http://www.ncdc.noaa.gov/ncdc.html).
Hudson River Database Release 4.1 TAMS/LTLTetraTech/fy
Figure 3.2.2B
Fort Edward to Stillwater Incremental Summer Average Flow vs.
^ ^ Total Precipitation for NCDC-Division 5 (Hudson River Valley)
-------
3.2.3 Fate Mechanisms
Response to DL-1.3
The source of the TI Pool load appears to be PCBs stored within the sediments. This source
may consist of partially dechlorinated sediments which release PCBs via porewater. These sediments
are probably older but need not be buried. As shown in the Low Resolution Sediment Coring Report
(USEPA., 1998), burial occurs, but is not continuous throughout the river bottom. Also in this
subsequent report, PCB mass loss in fine-grained sediments was identified based on collocated
sampling points taken in 1984 and 1994. In gross terms, this mass loss from the sediments agrees
with the estimated water column loading across the TI Pool for that time period. Irregularly lineated
zones, often occurring in regions of fine grained sediment, were identified in the side scan sonar
analysis (Flood, 1993). These areas appear to have a large potential for erosion. Wood chips which
may have eroded out of these areas are found in some areas. These findings demonstrate that burial
is not occurring universally in the river, but that regions in the river are eroding, making more
contaminated and dechlorinated sediments available to the water column.
Response to DC-4.2
Two dated cores showing an annual sedimentation rate of about 1 cm/yr do not establish that
the entire TI Pool river bed is continuously undergoing burial. The river appears to be in a dynamic
state with some areas exhibiting scour and other areas exhibiting burial. This is extensively discussed
in the Low Resolution Sediment Coring Report. (See the response to comment DL-1.3.) As shown
in this report, a median value of PCB mass loss has occurred in many fine-grained areas of the TI
Pool of which only a small portion can be explained by dechlorination. Given that there is evidence
of scour and PCBs have been lost to the water column, this transfer of mass may be partially
explained by resuspension of contaminated sediments. Plausible mechanisms which are consistent
with the data are discussed in the DEIR, but no attempt was made to establish the mechanism or set
of mechanisms under every condition.
Response to DC-4.3
The presence of PCB contaminated porewater in the water column is suggested by the
distribution of congeners found in the water column samples. The high levels of mono- and di-
chlorobiphenyls can be generated by partitioning the PCBs between dechlorinated PCB
contaminated sediments and porewater. The mechanism of transport is not clearly defined but there
are several possibilities including bioturbation, diffusion, groundwater flux and sediment scour. One
study estimates that certain rooted macrophytes turn over the porewater from three to eight times per
growing season (Templer, 1997). The contribution from exposed high concentration sediments might
be significant, with the exposure caused by scour or water craft activities. Finally, there is evidence
that the TI Pool fine-grained sediments have lost 40 percent of the PCB mass between 1984 and
1994 (USEPA, 1998). Only 11 percent of this mass lost is explainable by dechlorination. The
remainder is lost to the water column by some combination of mechanisms.
December 22. 1998
DEIR-23
TAMS/LTIyTclraTech/MCA
-------
Response to DG-1.24
The writer asserts that a wind-driven gas exchange equation is not appropriate for use in the
Upper Hudson during low flow conditions. This is based on the assertion that the energy for water
surface renewal stems from the downstream flow of the river and not wind. Alternatively, the writer
contends that O'Connor and Dobbins (1958), which is based on water flow, is a better model of the
gas exchange phenomena. The writer also points out that for gases such as PCBs, gas-phase diffusion
as well as liquid-phase diffusion can be rate limiting for gas exchange.
The purpose of the discussion in Section 3.2.3 was to briefly examine the various processes
affecting PCB transport in the Hudson River. Among these was the process of gas exchange,
potentially an important means for PCB loss from the water column. The prediction of gas exchange
rates has been extensively studied via a large number of techniques (Hartmond and Hammond, 1984,
Clark et ai, 1994; Clark et ai, 1996; O'Connor and Dobbins, 1958). In the Upper Hudson, gas
exchange processes have not been explicitly studied. However, it is still possible to surmise some
of the importance of the three models described in the DEIR.
In the context of the O'Connor and Dobbins model, gas exchange is roughly proportional to
the square root of the ratio of the bulk water velocity to the mean water depth. At low flow this ratio
can range from 0.015 to 0.081 sec 1 at flows of 1,000 to 8,000 cfs. However, as noted in O'Connor
and Dobbins, this model typically does not apply where water velocities are controlled by dams and
not by the simple process of flowing downhill. Thus the applicability of this model to the Upper
Hudson with its eight dams between Fort Edward and Troy is uncertain at best. Alternatively, in the
Lower Hudson, the flow to depth ratio is comparable to that of the Upper Hudson at low flow (0.05
sec ') based on information from Deck, 1981, and Garvey, 1990. Yet in this area of the river the gas
exchange rate has been extensively documented to be well predicted by wind-driven gas exchange
models (Clark et ai, 1994, Clark et ai, 1996). Although the fetch of the Upper Hudson is
substantially less than that of the Lower Hudson, the similarity of the flow-to-depth ratios and the
predominance of wind-driven gas exchange in the Lower Hudson as well as the observation that the
Upper Hudson at low flow more closely resembles a series of dammed lakes than a river would
together suggest that wind-driven gas exchange may be an important factor under low flow
conditions. At this point in the investigation, it would appear inappropriate to simply rule out either
mechanism for gas exchange in the large, relatively quiescent pools of the Upper Hudson during
low-flow conditions.
Regardless of which of these mechanisms govern the quiescent areas of the Upper Hudson,
they pale in comparison to the gas exchange which occurs at each of the dams. The eight dams of
the Upper Hudson represent a total vertical drop of about 122 feet. At each of the dams, energy is
rapidly dissipated as the river flow cascades down the downstream surface of the dam. This process
serves to generate a great deal of turbulence, incorporating fine air bubbles. This serves to greatly
enhance the air-to-water exchange of gas since it will occur across the individual bubbles as well as
the river's upper surface. Simulation of this process will be included in the modeling analysis.
The USEPA agrees with the concern over the gas-phase diffusion raised by the writer. This
will be examined for its importance during the modeling analysis.
Dccembcf 22. \99t
DEIR-24
TAMS/LTI/TetraTech/MCA
-------
3.2.4 Conceptual Model of PCB Transport in the Upper Hudson
Response to DC-4.7
This comment (problems with relating TSS to discharge) will be addressed after further
analysis in the Baseline Modeling Report. Data relating TSS and flow was explicitly obtained for
this purpose and will be incorporated in the Baseline Modeling Report. Additionally, GE collected
similar TSS and flow data which will be examined in this context as well. Nonetheless, the USEPA
disagrees with the notion that sediment resuspension and PCB load cannot be inferred in some
instances, since subsequent changes in congener pattern and PCB load would be readily recognizable
as was seen at Waterford in Transect 3 (which occurred in March of 1993).
3.2.5 River Characterization
No significant comments were received on Section 2.3.5.
3.2.6 Mass Load Assessment
Response to DS-2.5
The statement should read as follows: "Water column processes such as gas exchange and
in situ degradation may also serve to remove a portion of the PCB load originating above Rogers
Island. However, this appears unlikely in view of the apparent absence of their effects below TI Dam
and the near-conservative PCB transport observed from the TI Dam to Waterford during these
transects."
Response to DS-2.6
USEPA agrees with the writer's observation. Subsequent to the release of the DEIR,
additional data have been obtained by GE which support the suggestion that the region from the TI
Dam to Schuylerville also yields a substantive PCB load. These results also suggest that the TI Dam
stations used by both the USEPA and GE may overestimate the total load over the TI Dam during
warm weather, low flow conditions when loads at Rogers Island are largely nondetect (see Section
1 of USEPA comments on the GE/QEA model in book 3 of this responsiveness summary). The end
result of the analysis of these data indicates that the TI Dam station yielded may occasionally
overestimate PCB loads depending upon other conditions The degree of overestimation varies from
about 0 to 36 percent but it is negligible at flows over 4000 cfs under most conditions seen during
the past 8 years. Only when loads at Rogers Island drop to negligible levels does the overestimation
approach 36 percent. The writer is referred to Section 1 of the USEPA review of the GE/QEA model
presented in book 3 of this responsiveness summary. The implications of these observations will be
fully explored as part of the modeling effort. Regardless of the outcome of this analysis, the four
major findings of the report will still apply. If, after a more thorough analysis of the new data, the
areas below the TI Dam are shown to yield a significant load, the indications are that these loads still
represent only about 20 percent of the total load during low flow conditions. Even with new
information, the conclusions of the DEIR concerning the overall importance of the TI Pool
sediments to PCB water column loads in the freshwater Hudson remains valid.
December 22. 1998
DEIR-25
TAMS/LTT/TetraTech/MCA
-------
Response to DS-2.7
To date, there are no data which are capable of specifically identifying the fate of the newest
PCB contributions to the TI Pool from above Rogers Island. Undoubtedly, these newer contributions
are subject to the same sorts of processes which affect the older PCBs but it is difficult to trace the
newest contributions short of an extensive program of high resolution cores. Even then there would
be no guarantee that the most recent releases would be readily discernable from reworked PCBs.
Response to DS-2.8
USEPA agrees that the processes affecting PCB transport in the TI Pool are unlikely to be
unique to the Pool. As noted, the scale of these processes would be expected to vary from pool to
pool in the Upper Hudson, dependent on the physical characteristics of each pool.
Response to DS-2.9
The phenomena noted, i.e., the consistency of the total load but the change in congener
distribution, is observed only during warm water, low flow conditions. Evidence for consistency in
total load and in congener pattern is noted for cold water and high flow conditions, suggesting that
the degree of "conservative" transport is probably dependent oh several factors including
temperature as well as water residence time in the Upper Hudson. This issue of quasi-conservative
transport is less important than assessing the overall magnitude of the load emanating from the
Upper Hudson sediments. Indeed it is quite conceivable that sediments downstream of the TI Dam
(or Schuylerville) may be in a quasi-equilibrium or steady-state with the historical water column
loads and concentrations produced in the upstream areas. This issue remains to be explored further
during the modeling effort. It should be noted, however, that the calculations demonstrating
"conservative" behavior using summer months are being revised. "Conservative" behavior is not
expected during summer months, as noted in the corrections in Section 3.2.
Response to DG-1.9
The concerns over the accurate representation of the PCB loads during the April 1993 high
flow event are valid. However, the transect 4 sample was collected at Rogers Island as the flow
peaked in the river and not after the peak as contended. It does not appear that the loading of PCBs
during this transect was directly attributable to scour since no change in suspended matter
concentration (and therefore load) occurred between Bakers Falls and Rogers Island (18.4 vs 17.5
mg/L, respectively). Thus, the concern over sampling during the period of rising river flow may not
be valid. Specifically, it is unclear as to the nature of the PCB release process in the Bakers Falls
area. Given that the PCB oils apparently enter the river via rock fissures and man-made pipes, it
would appear likely that the greater displacement of water through these conduits during high flow
is a more likely mechanism rather than a scouring of the river bottom. The displacing water could
originate as groundwater or as a result of higher water levels in the river channel. On this basis it is
unclear when the peak flushing of these PCB oils would occur.
Nonetheless, it is important to note that the majority of the load delivered to Rogers Island
during this transect is carried through the Upper Hudson to the Lower River. This is not to suggest
that the PCB inputs upstream of Rogers Island do not contaminate the sediments of the Upper
December 22, 1998
DEIR-26
TAMS/LTI/TetraTech/MCA
-------
Hudson, but simply that the Upper Hudson is not an efficient filter for these loads and that the
majority of their mass is delivered downstream.
Lastly, as made evident by the GE monitoring data, the importance of the Bakers Falls
releases during the spring, as well as for the rest of the year, has greatly decreased. This is attributed
to the remedial work conducted by GE at the Hudson Falls facility. Thus the issue of how and how
much of the water column load is generated above Rogers Island has become largely moot. The
sediments of the Upper Hudson, particularly those of the TI Pool, have become the major source of
PCBs to the Hudson. While it is undoubtedly true that the sediment contamination of the Upper
Hudson originated above Rogers Island, it is also true that the current surficial sediment
contamination is the integration of resuspension, settling, and biological activity as well as fresh
introductions of PCBs from the GE facilities.
Response to DG-1 10
The remedial efforts undertaken by GE have produced a marked decline in the loads
originating above Rogers Island since September 1991. In particular, the peak loads of the
September 1991-June 1993 period were substantially reduced in the ensuing months as recorded by
the GE monitoring program. Typical water column concentrations at Rogers Island were reduced
from levels as high as 1.000 ng/L during high flow levels to nondetect levels throughout much of
the year. This is shown in Figure 3-82 of the DEIR as well as in subsequent GE reports. The GE
contention that PCB transport into the TI Pool occurs as oil droplets has never been substantiated
despite sampling by GE to demonstrate its existence. Given the absence of evidence for this
phenomenon and the markedly reduced PCB loads originating above Rogers Island, it is unlikely
that fresh releases remain a major issue to PCB transport. GE's remedial efforts have sufficiently
reduced the loads from the Hudson Falls facilities so as to permit the unobstructed observation of
the TI Pool sediment load during both 1997 and 1998. As demonstrated in subsequent reports by GE
(e.g., QEA. March 1998), the loads upstream have become minor relative to that of the TI Pool on
an annual basis. Thus, while the mechanisms for PCB release from the Hudson Falls facilities remain
unknown, their importance to the Hudson River PCB problem has greatly diminished. In this
context, it appears that the degree of control attained by GE readily permits the study and
interpretation of the loads produced by downstream sediments.
Response to DG-1.1 OA
The contention that higher loads of unaltered PCBs appear to enter the TI Pool at higher
flows is supported by the USEPA and GE data. However, the comment fails to note that these loads
appear to be swept unmitigated to the Lower Hudson, with relatively little loss as a percentage of
total transport to the sediments of the Upper Hudson. While it is undoubtedly true that freshly
contaminated sediments are produced and deposited during these events, it is apparent that these
sediments represent just a small fraction of the PCBs released. These same conditions presumably
serve to resuspend. deposit, uncover and bury older and newer contaminated sediments, yielding a
surface sediment patchwork of PCB contamination. The contention that large portions of the PCB
load originating upstream of Rogers Island enter the TI Pool undetected and remain undetected
despite nearly eight years of monitoring as well as specific sampling techniques employed by GE
to demonstrate their occurrence is unsubstantiated.
December 22. 19<>g
DEIR-27
TAMS/LTl/TetraTech/MCA
-------
Response to DG-1.11
The writer contends that the PCB loads which occur during the passage of the river through
the TI Pool cannot be explained by the combination of the load at Rogers Island and a diffusive flux
from the river sediments. This calculation is based in part on the very limited data set obtained prior
to the September 1991 failure of the Allen Mill structure and the large ensuing PCB releases related
to the GE Hudson Falls facilities. The calculation is of questionable value for a number of reasons.
First, it is not clear that porewater exchange in the Tl Pool would be exclusively limited to diffusive
exchange. Groundwater migration and biological processes may serve to greatly enhance the
exchange of porewater with the overlying river. Second, the partition coefficients developed by the
USEPA may not have the precision assigned to them by the writer. As a result, there may be greater
uncertainty in the congener pattern of the truly dissolved PCBs in porewater than ascribed by the
writer. Thirdly, the assignment of a partition coefficient for DOC-bound PCBs at one-tenth of the
value developed by the USEPA is not defensible and appears without support. The values developed
in this manner are quite different from those developed from three-phase calculations in the DEIR.
Additionally, the DOC content of the porewater is not well defined and its concentration may vary
significantly. This variation would directly affect the congener pattern carried by the porewater. The
DOC measurements themselves are based on frozen samples collected in 1991. The effect of the
freezing process is unclear, adding further uncertainty to the calculation presented. Based on these
issues, the USEPA does not accept the calculation nor the conclusions based on it.
Response to DG-1.15
The writer raises four issues:
1. Transient PCB transport events are not captured by GE's monitoring efforts. (Responded to
as DG-1.15A)
2. The USEPA sampling program missed the peak loading event in 1993. (Responded to as
DG-1.15B)
3. PCB oils are the likely source of the spring peak PCB transport. (Responded to as DG-1.15C)
4. Flushing activities conducted by the hydroelectric plant at Bakers Falls serve to flush PCBs
out of the plunge pool on an irregular basis. (Responded to as DG1.15D)
Response to DG-1.15A
The capture of transient transport events will always remain a concern to some degree.
However, the length of weekly to biweekly sampling at the GE monitoring points, as required by
the USEPA, now covers more than eight years or over 400 samples at Rogers Island alone. The size
of this data set is sufficient to characterize the general rate of PCB loads originating above Rogers
Island. In fact, the data set clearly documents the decline in PCB loads at Rogers Island, largely
attributed to the remedial efforts at Hudson Falls conducted by GE. Current levels are frequently
nondetect. indicating reductions of more than two orders of magnitude in the Rogers Island load
since 1993.
In a subsequent report to USEPA (QEA, March 1998), GE demonstrated the homogeneity
of the water column load above Rogers Island on two separate occasions by collecting shallow and
deep samples at several locations in a river cross-section (see Figure DG-1.15A). Water column
concentrations both horizontally and vertically rarely varied more than 25 percent and typically
December 22. 199*
DEIR-28
TAMS/LTl/Tetra Tech/MCA
-------
100 -
90 -
80 -
70 -
60 -
60 -
40 -
30 -
20 -
10 -
0 -
~ SHALLOW
HOEEP
¦ ROUTINE
1
1
Wtst
9/17/95
I
3 4 5
HRM 194.9 Transect
East
HRM 194.4
100 .
~ SHALLOW
90 - ODEEP
ROUTINE
10/3/95
80
70 -
I 60
C
g 50 "
| 40
30 -
20 -
10 -
0 -
I
1
I
i
|
1
n.
Wait
East
HRM 194.9 Transect
HRM 194.4
From QEA. March 1998
T AMS/LTI/T etraT ech/
Figure DG-1.15A
Water Cnlnmn PPR Cnnrentrj»tinn<; Within thp Virinitv r>f Fnrt FHwarH
-------
agreed to within 10 percent. The data also showed that the normal Rogers Island monitoring station
agreed with these cross sections within reasonable bound, falling 25 percent lower and 58 percent
higher on the two sampling events. The close vertical agreement of the water column samples is
strong evidence for the absence of the oil droplets at the river bottom, since they would presumably
raise the concentration of the deeper water relative to the shallower water. Additionally, the data
show that the water column load originating in the Hudson Falls area has been thoroughly
homogenized in the water column by the time the load reaches Rogers Island. This is further
evidence of the accuracy of the Rogers Island monitoring station.
In view of this, it appears likely that the Rogers Island monitoring results will on average
accurately represent the loads entering the TI Pool.
Response to DG-1.1 SB
The writer also contends that the 1993 high flow sampling event missed the peak flow and
therefore did not capture the maximum PCB load. This is incorrect. The sampling was conducted
on the day of peak flow and not after the peak. The previous day's flow was 17,200 cfs and the
subsequent day's flow was 18,100 cfs. Subsequent flows later in the month exceeded the levels seen
on 4/12/93 but the 4/12/93 event represented the first high flow event that year. The
representativeness of this event was further supported by sampling performed in 1992 and 1997 by
GE. In both events, sampling was performed during the rising and falling flow periods of the spring
runoff event. During the 1992 event the peak PCB load of 35 kg/day was produced at the maximum
flow for the period, representing the first major flow event of the year. The coincidence of the peak
flow and peak load are shown in Figure DG-1.15B, a modified version of Figure 26 supplied by GE.
During the 1997 sampling, the peak load generated was only 2 lb/day in 1997 vs 18 in 1993.
Nonetheless the data clearly show the coincidence of the peak flow (ca. 17,000 cfs) and peak PCB
load (see Figure DG-1.15C). Subsequent high flows later in the month in 1992 and 1997 did not
yield such a substantive PCB load. While this does not constitute proof that the 1993 USEPA
sampling event represents the peak load for the year, it does suggest that the sampling event did
capture the maximum load carried by the river up to that point in the year and probably for the whole
year. Thus it is unlikely that the load underestimates the true PCB loading.
Response to DG-1.15C
The USEPA agrees with the assessment that high flow events are generally associated with
high PCB transport events, although the timing of these events seems to affect the amount of PCBs
they deliver. Since little or no increase in TSS occurs during these events between the upstream
station at Bakers Falls and the downstream station at Rogers Island, it would appear that either
highly contaminated sediments (percent levels) or PCB oils are mobilized during these events.
Either source would have no effect on the TSS but would greatly increase the PCB load. However,
the writer contends that this process would serve to deliver oil droplets undetected by the monitoring
station at Rogers Island. As stated previously, the USEPA rejects this contention for the following
reasons. First, no evidence has ever been found for their transport past Rogers Island despite GE's
efforts to discover them. Second, the region of the river between Bakers Falls and Rogers Island
probably represents one of the best areas for the homogenization of PCBs introduced to the river.
Specifically, the very large quantity of energy added to the river as turbulence from the 70 foot
Bakers Falls plus the large section of rapids at Remnant Deposit 1 and the former Fort Edward Dam
December 22. 1998
DE1R-30
T AMS/LTL'TctraTech/MC A
-------
20
25
as
O)
B
•
«
e
*6
o
o
fiL
30 35 4
Peak Flow and PCB Load
Julian Day (Starting 4/1/92)
PCB data are corrected for analytical bias
From GE Comments. April 1997
TAMS/LTI/TetraTech/!
Figure DG-1.15B
PfR anH Transnnrt Durina 1Q0? Snrino Hioh Flrvw
-------
30000
24000
•og
Sc
?"• 18000
12000
3*
k 0000
0
20
I "
<8 12
1 «
2L
3 4
Oft
3
o
H
3
4
3
2
1
0
April
Peak Flow. TSS and PCB Load
Note: Daily averages plotted: Non-detect PCBs plotted as open symbols at one-half MDL.
From QEA. March 1998 TAMS/LTI/TetraTech/M
Figure DG-1.15C
Temporal Trends in TSS and PCB Concentration and Loading
-------
site (now essentially a small falls) plus the two sharp river bends all serve to blend the water column
load both horizontally and vertically (see Figure DG-1.I5A). This energetic system would be
expected to dissolve any oil droplets as well. Thus, as stated above, it is highly unlikely that any
significant PCB load passes Rogers Island that is not reflected in the regular monitoring done by GE.
A second issue needs to be raised here as well. During the high transport event monitored
by USEPA, there was little evidence for decreases in PCB load downstream of Rogers Island,
implying that the vast majority of the PCB load released upstream was translated down to the Lower
Hudson, with relatively little storage in the sediments as compared to the amount translated. Other
sampling events clearly showed a net gain across the TI Pool. Thus even if an individual large PCB
transport event is missed by the monitoring station, its impact on the overall inventory of the TI Pool
and other sediments of the Upper Hudson is likely to be relatively small. Rather, the sediment
inventory of the Upper Hudson has been built up over many years, with the largest gains occurring
during the mid-1970s. Events from 1991 to 1993 undoubtedly added to this inventory. Resuspension,
groundwater migration, diffusion and biological activity all serve to rework the contaminated
sediments, replenishing the upper sediment layer. As a result, the release of PCBs from the
sediments has probably been occurring for many years and will continue to do so for many more to
come. There is no evidence in the data collected to date that this source is diminishing.
Response to DQ-1.I5P
The hydro plant flushing activities were studied by GE and reported to USEPA (QEA, March
1998). These studies show small increases in the PCB loads delivered to Rogers Island as a result
of these activities. Some results are shown in Figure DG-1.15D. These events serve to create highly
concentrations near the plunge pool which are subsequently homogenized by the transit to Rogers
Island. Thus, although these events do add to the PCB loads of the Upper Hudson, they do not
represent major loads unaccounted for by the Rogers Island monitoring.
Response to PG-1.16
As noted in the corrections to Chapter 3.2, the PCB transport estimates presented in the DEIR
will require revision. Based on the knowledge that the flow estimates at Stillwater and Waterford
used in the DEIR were probably 15 to 40 percent too high, it is expected that the low flow sampling
events completed by the USEPA will not show constant PCB loads from Schuylerville to Waterford.
In addition, evidence collected by GE at the TI Dam monitoring station suggest that this station may
overestimate loads at low flow, summer time conditions. Rather, the reanalysis using revised flows
will likely yield declining PCB load downstream of Schuylerville. This reanalysis will be submitted
as an appendix to the responsiveness summary for the Low Resolution Sediment Coring Report.
However, it is likely that the analogy of a pipeline to describe PCB transport will no longer be
appropriate.
The writer indicates that the DEIR did not consider the effects of various geochemical
processes below the TI Dam. In fact, these are mentioned and discussed at length in the report.
However, evidence for their occurrence and magnitude is far less obvious in the region below the
TI Dam than within the Pool itself. Downstream of the TI Dam loads change gradually under low
flow conditions (perhaps 40 percent in light of the revised flow data) while the load gain across the
Dcccmbw 21 mt
DEIR-33
TAMS/LTl/TttraTech/MCA
-------
Hydrofacility Monitoring
Plunge Pool
o>
c
2
C
Q>
O
C
o
o
CD
O
Q.
460
400
350
300
250
200
150 •
100
50
0
¦Before
j
~ During/After!
9/4/96
6/9/97
6/23/97
' N on-Detects Plotted at Detection Limit
Hydrofacility Monitoring
Fort Edward
D)
C
o
n
©
o
c
o
o
03
O
a.
100
90
80
70
60
50
40
30
20
10
0
¦ Before
~ During/After
9/4/96
6/9/97
6/23/97
' Non-Detects Plotted at Detection Limit
FromQEA. March 1998
TAMS/LTI/TctraTech/Mv.
Figure DG-1.15D
Vater Column PCB Concentrations at Bakers Falls Plunge Pool and
T?r»rt PrUiror/-! from J-JvHrrvF»pi1itv A/frmitnrino Prnarsm
-------
TI Pool typically increased by 300 to 500 percent. Thus the focus on the TI Pool was not
inappropriate in the context of the DEIR. Further analysis in the context of a mass balance model
was anticipated prior to the preparation of the DEIR. The mass balance model results will be
described in the upcoming Baseline Modeling Report.
The estimates of gas exchange and settling losses described by the writer cannot be examined
without the support of a complete report. A report describing part of this work was submitted to
USEPA and is critiqued elsewhere in this responsiveness summary.
3.2.7 Source Loading Quantitation
Response to DS-2.10
The data recently collected by the NYS Canal Corporation should provide insight into the
scour events that occurred in the area of the Hoosic River confluence. USEPA agrees that Transect
3, like Transect 6, provides clear evidence for PCB contributions from the sediments below the TI
Dam. Nonetheless, it is still concluded that the sediments of the TI Pool still represent the major
sediment source of PCBs to the water column of the Upper Hudson.
Response to DS-2.11
The TI Pool load of 0.65 kg/day measured during low flow conditions is small relative to the
spring runoff event load of 20 kg/day generated above Rogers Island. The ability to measure the TI
Pool load is limited by the relative difference. Estimated flux errors were of the scale of 20 to 25
percent, thus precluding resolution of a TI Pool flux of this magnitude. It would also be difficult to
resolve the presence of a TI Pool load based on congeners alone as was done for some lower Hudson
high resolution core samples in Chapter 3 of the DEIR. Both of the proposed conditions would yield
the measured results for the spring high flow transect and thus neither hypothesis can be ruled out
at this time. Modeling analysis may clarify this issue to some degree.
Response to DC-4.5
USEPA agrees that there are many challenges in estimating PCB loads, and trends in loads,
across the Thompson Island Pool and between Thompson Island Dam and Waterford. Comparisons
on the basis of concentration alone must be pursued with considerable caution, due to the presence
of short-term variability in concentrations and the fact that few samples, with the exception of the
Phase 2 transect samples and a few of the USGS samples, have been timed to attempt to sample the
same "parcel" of water at consecutive downstream stations. Although there are shortcomings in
terms of sample representativeness relative to flow rate, the best load estimates possible from
available data have been produced.
USEPA concurs that load across the Thompson Island Pool is most evident at low flow
conditions, and not obvious during high flow conditions. Indeed, during spring high-flow sampling
(Transect 4 and Flow-Averaged Event 1), a net loss of total PCBs was estimated across the
Thompson Island Pool, although this is believed to be due. at least in part, to a different protocol for
sampling and compositing during this sampling. These results are discussed in Section 3.2.6 of the
DEIR. Recent GE data suggest that the greatest load gain across the Thompson Island Pool occurs
December 22. 1998
DEIR-35
TAMS/LTl/TetraTech/MCA
-------
during spring-summer low flow conditions, which is consistent with biologically-mediated
mobilization of PCBs from the sediment, but may also reflect temperature effects on PCB
partitioning and seasonal changes in pore water movement.
The commentor states, as an example of "a basic weakness in estimating annual loadings",
the following characterization of USGS load estimates: "For example in Table 3-23, note that
estimated annual PCB loadings at Schuylerville (below TID) are greater than Fort Edward 1977-85,
and then abruptly reverse 1986-1993 (1990-93 relative to Waterford)." It should first be pointed out
that the change is not necessarily abrupt, as the load estimates at Fort Edward do not show a
statistically significant difference from those at Schuylerville for the period 1984-1988. Further, a
shift in the relative importance of sources above and below Fort Edward is fully consistent with our
understanding of historical PCB transport processes in the Upper Hudson River. During the earlier
years, large, unstable deposits of contaminated sediment derived from the former Fort Edward Dam
pool were present within the Thompson Island Pool, and can be expected to have contributed a
significant, but generally declining load of PCBs to the water column. In contrast, loading above Fort
Edward appears to have always contained a component derived from PCB DNAPL flux in the
Bakers Falls area, which has remained more constant over time. Thus, a gradual shift in importance
from sources below Fort Edward to sources above Fort Edward is expected over time as readily
mobilizable contaminated material in the Thompson Island Pool was buried or depleted. (It should
also be recalled that the USGS quantitation methods significantly under-represent the mono- and
dichiorobiphenyl fractions of total PCB load, and thus do not show much of the dechlorination
product flux from the Thompson Island Pool.)
USEPA disagrees with the characterization that "all too often high values have been
discarded as anomalous". In fact, only two data points were rejected for the calculation of historical
loads: one from GE data, and one from USGS data. The GE observation at Rogers Island on January
19, 1994 was not included due to concerns over sampling protocol and abnormally high TSS
concentrations. This sample was flagged by GE as having been collected from shore due to ice cover,
and the high TSS concentrations suggest that the sample was contaminated by disturbance of near-
shore sediment during sample collection. For the analysis of the USGS data, one anomalous
observation was omitted. This sample was taken on December 15, 1983 in the navigation canal (east
channel) at Rogers Island, and showed a reported PCB concentration of 77 fig/1, far greater than any
other reported water column concentration. On the same date the concentration in the main channel
at Rogers Island was only 0.01 jig/1.
Response to DG - 1.4A
The DEIR suggests that a portion of this load may be stored in the sediments of the TI Pool.
However, although this is a possibility, nearly all other analyses concerning the TI Pool load
examine net gain, effectively assuming that the entire Rogers Island load is passed through the pool
at all flow conditions. This provides a minimum estimate of the amount of PCBs produced by the
sediments of the TI Pool.
Response to DG-1,4B and DG-1,4C
The report does not ignore these mechanisms at all and in fact notes them directly (see DEIR
page 3-60, for example). However, the existence of these mechanisms does not change the nature
December 22, 1998
DEIR-36
TAMS/LTl/TetraTcch/MCA
-------
of the loading in the Upper Hudson, i.e.. that most of the Upper Hudson's PCB load is generated
above the TI Dam and that this load is similar in size and congener makeup to that delivered to
Waterford. Given the larger inventories of PCBs within the TI Pool as compared to those below
Schuylerville, it is reasonable to conclude that most of the load delivered to Waterford originated
above the TI Dam.
Response to DG-1.4D
The report examines data obtained by the USEPA sampling program as well as that collected
by the USGS and GE, representing more than 18 years of data collection. Thus it is not solely
focused on the conditions of 1993. However, the detailed sampling and analysis performed during
1993 can be used to increase the understanding of PCB geochemistry throughout the entire data
collection period as was done in the report. The conclusions made on the basis of the 1993 studies
were compared with the results obtained by other studies so as to confirm their applicability. For
example, note that the conclusion concerning the importance of the TI Pool loads was based not only
on the Phase 2 results but also on the three years of sampling by GE completed after the Phase 2
water column program.
3.3 Historical Water Column Transport of PCBs
No significant comments were received on Section 3. J.
3.3.1 Establishing Sediment Core Chronologies
Response to DS-2.12
The difficulty in obtaining a datable core from the TI Pool as well as elsewhere in the river
indicates that the majority of locations do not yield useful core chronologies. USEPA agrees with
the statement that "there are times when either there is no sedimentation, or events that result in the
removal of part of the sediment column." The Low Resolution Sediment Coring Report provides
further ev idence for the transient nature of sediment deposition and removal.
3.3.2 Surface Sediment Characterization
Response to DC-4.1
The water column transect suspended solids total PCB concentrations at Rogers Island
(median value 17.3 ppm) is higher than at the TI Dam (median value 5.3 ppm). As shown in DEIR
Figure 3-8. the partition coefficients derived for the Rogers Island station vary greatly from the
coefficients derived for the other stations. This indicates that the PCBs in the water column are not
in equilibrium at this station, which may be a result of the fresh source of PCBs input from above
Fort Edward. Thus, a disproportionate amount of PCBs is detected on the suspended fraction as
compared to the dissolved fraction in the water column at Rogers Island. This difference may also
result from sediment/water exchange during transit through the pool wherein suspended matter
originating above Rogers Island may be partially exchanged or removed from the water column.
December 22 1998
DEIR-37
TAMS/L TUTctra Tcck'MC A
-------
3.3.3 Water Column Transport of PCBs Shown by Sediment Deposited After 1975
Response to Comment DF-2.4
Further congener-specific load analysis will be completed for five congeners as part of the
fate and transport modeling effort. To the extent necessary, other congeners may be examined
individually as well. Nonetheless, the level of analysis in the DEIR provides a sufficient basis for
the conclusions drawn in the DEIR. Further analysis always has the potential to yield greater insights
and the option to continue the analysis of data may be pursued.
The writer is correct in noting the inconsistency on p. 3-122. The intention of the original
statement was to note the greater proportion of more chlorinated congeners in the sediments at RM
-1.9 relative to RM 177.8, and not an absolute increase inlhe concentration. In fact, as noted by the
writer, sediment concentrations for nearly all congeners are lower at RM -1.9 vs RM 177.8. The
trend to lower PCB concentrations downstream from RM 177.8 is the premise behind the PCB/l37Cs
analysis presented in Section 3.3.3 of the report.
Response to DL-l .8
Fingerprinting of congener patterns is discussed extensively in Section 3.3.3 of the DEIR.
This analysis compares water column and sediment samples from both the Lower and Upper Hudson
River and establishes a strong link between the sediments of the lower river and input from the
Upper Hudson. Congener patterns in the fish samples will be examined in the Ecological Risk
Assessment. However, as noted in the text, congener fingerprinting is not the only basis available
to establish the contribution of PCBs from the Upper Hudson River to the Lower Hudson River and
New York/New Jersey harbor. Other evidence is considered as well. The data presented in the DEIR
are sufficient to attribute freshwater Hudson River PCB contamination to the Upper Hudson River.
Response to DG-1.17
In this response the writer asserts that the PCB/l37Cs analysis presented in the DEIR is invalid
for the reasons listed below.
1. The simple dilution model used to estimate the PCB loading with respect to l37Cs fails to
account for changes in the suspended solids yield downstream of Ft. Edward.
2. Different deposition rates yield different l37Cs levels in the 0-2 cm layer.
3. The addition of Aroclor 1242 by other external sources is not accounted for by this analysis
and the Albany core top is biased toward the Upper River because of the high loads produced
by the 1991-1992 events at Allen Mills.
I. The writer cites the work of Phillips and Hanchar (1996) to contend that the suspended solids
load greatly increases between Fort Edward and Waterford and therefore the assumption that all
tributaries contribute suspended solids in proportion to their drainage areas is invalid. Furthermore,
the writer contends that the fact that the simple dilution model could not be applied to the area
between the TI Dam and Stillwater is further evidence for its rejection.
December 22. 1998 DEIR-38 TAMS/LTI/TetraTech/MCA
-------
The data provided by Phillips and Hanchar actually support the assumption of constant
suspended solids yield per unit area of basin as well as demonstrating why the model could not be
applied above Stillwater. Table DG-1.17 summarizes the information provided by Phillips and
Hanchar. Also shown in the table is the incremental increase in yield between the three USGS
suspended solids monitoring stations of the Upper Hudson.
Of particular note are the four yield values shown in bold. The yield of suspended solids for
the Upper Hudson increases steadily from 0.044 tons/day/sq.mi. at Fort Edward to 0.097
tons/day/sq.mi. at Waterford. To accomplish this, the tributaries downstream of Fort Edward must
yield substantially higher suspended solids loads relative to the region above Fort Edward. This is
reflected in the incremental yield column. Note that the incremental yield between Stillwater and
Waterford (0.235 tons/day/sq.mi.), principally reflecting the Hoosic River, is much higher than the
Upper Hudson. Thus, the model fails to explain the change in the PCB/137Cs ratio to this point in the
river, as was noted in the DEIR. because the suspended solids yield is changing so rapidly. However,
the suspended solids yield of the Hoosic watershed is essentially equal to that of the Mohawk (0.25
tons/day/sq.mi.), the next major watershed along the Hudson. Thus, the dilution model would be
expected to work in this region as in fact it was shown to. PCB concentrations in the Albany core
are predicted by those in the Stillwater core because the two intervening rivers introduce suspended
solids at essentially the same rate, thus achieving a dilution of the PCB/,37Cs ratio between Stillwater
and Albany which is proportional to the increase in drainage area.
The introduction of the Hoosic and Mohawk Rivers serves to increase the drainage area from
3773 sq.mi. at Stillwater to 8.090 at Albany, more than doubling the total area and halving the
concentration as predicted by the model. This result matches the measured trend quite well as shown
in Figures 3-66 to 3-68 of the DEIR Report. From Albany to Kingston the drainage area increases
by another 30 percent to 11.300 sq.mi. and the PCB/l37Cs ratio decreases according for both the
model and the measurements. Unfortunately, there are no suspended solids data to further support
the model but its success to Kingston and to Lents Cove would suggest that the suspended solids
yields of the Lower Hudson are comparable to those of the Mohawk and Hoosic Rivers. Thus the
data provided by Phillips and Hanchar provide the explanation as to why the model does not work
above Stillwater relative to the region between Stillwater and Lents Cove.
2. The writer's premise that the 0-2 cm layer does not consistently reflect the same time
horizon among the coring locations and therefore the l37Cs levels do not represent the same years of
deposition is incorrect. First, nearly all of the cores represented in DEIR Figure 3-63, which displays
the 137Cs levels in 0-2 cm slices in the Hudson, were shown to contain 7Be and therefore represented
recent deposition, even when the core sequence could not be dated. Secondly, as shown in Figures
3-53 to 3-54. l37Cs levels have not varied substantially during the most recent five years. Thus, as
long as the 0-2 cm core slice represented materials deposited sometime over the last five years, the
l37Cs levels could be expected to reflect differences in local deposition rates and not differences in
l37Cs deposition over time. On this basis, the information presented in Figure 3-63 reflects the
consistency of recent l37Cs deposition in the Hudson and shows the assumptions used by the simple
dilution model to be appropriate.
3. The addition of any Aroclor to the Hudson River by an external source will be reflected in
the PCB/l37Cs ratio by an increase in the ratio relative to that predicted by the dilution model. In this
regard, it does not matter which Aroclor is added to the mixture. The fact that the model is predictive
December 22. 1998
DEIR-39
TAMS/LTl/TetiaTech/MCA
-------
Table DG-1.17
Suspended Solids Yields for the Hudson River to Albany 1
USGS Monitoring
Location
River Mile
Suspended Solids
Yield
(tons/day/sq.mi.)
Drainage
Area
(sq.mi.)
Incremental
Drainage Area
(sq.mi.)
Incremental
Suspended Solids
Yield
(tons/day/sq.mi.)
Ft Ed
194.6
0.044
2,817
956
0.131
(Ft. Edward to Stillwater)
Stillwater
168.3
0066
3,773
847
0.235
(Stillwater to Waterford)
Waterford
156.5
0097
4,620
Mohawk
156.2 2
(1.250
3,456
Green Island
151.7
0 220
8,090
Waterford t Mohawk 5
151.7
0.163
8,076
Notes:
1. From Phillips and Hanchar, 1996
2. River mile at confluence with Hudson
3. Area-weighted average yield of the two watersheds
TAMS/LTI/TetraTech/MCA
-------
over a 17-year period indicates how long the GE releases have dominated the Hudson's PCB
contamination problem.
In considering the congener patterns of the sediments, it is true that a mixture identical to that
of GE's input will not be discernable. However, in order to reflect the congener patterns identified
in the sediments, PCBs from other sources would need to contain the exact proportions of 1242,
1016 and 1254 seen in the GE releases, an unlikely occurrence. As further evidence of the absence
of significant external inputs downstream of Stillwater, Figure DG-1.17 shows the decline in four
congener ratios in high resolution core tops (i.e., recently deposited sediment) as a function of river
mile. In each instance the ratio declines steadily, without any indication of additional input, which
would increase the ratio back to conditions seen at River Mile 195. These ratios serve to track the
absence of a GE-like input which might not be discerned using the overall congener pattern analysis
presented in the DEIR. In this manner, the combination of the PCB/l37Cs ratio, the congener pattern
analysis, and the H, H' congener ratios all serve to identify the region above the TI Dam as the
principal source of PCBs to the freshwater Hudson.
The concern that the core top comparison is biased because of the large GE-related release
events of the 1991-1992 period is unfounded. The fact that additional PCBs are transported
downstream from the Upper Hudson simply adds to the PCB inventory at each coring site. The cores
simply integrate the total loadings. If the Upper Hudson PCB load was greater during 1991-1992,
then this load constituted an even greater proportion of the total sediment concentration as would
be expected.
Response to DG-1.22
The writer contends that the H. H* dechlorination process occurs extensively throughout the
Hudson and that this represents a significant reduction in the PCB exposure, toxicity and
carcinogenicity.
The issues of toxicity and carcinogenicity are left for subsequent discussion in the upcoming
Human Health and Ecological Risk Assessment Reports. However, it should be noted that it is not
clear that the dechlorination processes render the congeners less toxic. As to exposure, the result of
dechlorination is less clear. Although the tendency for bioaccumulation decreases as a molecule is
dechlorinated, the mobility of the congener tends to increase since the sediment-water partition
coefficient also decreases as the molecule is dechlorinated.
Finally, although the H. H' patterns may be common in the Lower Hudson, their occurrence
does not indicate that substantive dechlorination of the PCB contamination will be found there.
Therefore, in the context of examining the dechlorination as a potential means to ameliorate the PCB
contamination of the Hudson, it is misleading to focus on the few congeners which may be affected
when the vast majority of the congeners in the Lower Hudson are unaltered by this process.
December 22 199S
DEIR-41
TAMS/LTLTctraTech/MCA
-------
BZ#74:49
BZ#66:49
RIVER MILE
BZ#60:49
BZ#56:49
RIVER MILE
100
RIVER MILE
" Recent Sediment
¦ Rogers island Suspended Matter - April, 1993
Note:
Trend lines are for visual reference only.
Hudson River Database Release 4 1 TAMS/LTI/TetraTech/MCA
Figure DG-1.17
Trend of Various H, H' Markers in Recent Sediments (0-2 cm) as a Function of River Mile
-------
3.3.4 Estimation of the PCB Load and Concentration across the Thompson Island Pool based on
GE Capillary Column Data
Response to DG-1.12
As discussed above, the analysis of congener patterns developed by the writer is not based
on defensible assumptions and as such cannot be used to develop a fingerprint for the "missing"
source. Given the range of possible release mechanisms and the uncertainties in their possible impact
on the water column load, it is not possible to definitively describe the PCB sources from the TI
Pool. Undoubtedly, the TI Pool load is generated by a blend of older, relatively unaltered PCBs,
older, highly altered PCBs, and recently released, unaltered PCBs.
Response tp DO-1.14
The analysis presented in the comment is used to suggest that the load differential between
Rogers Island and the TI Dam (i.e.. the TI Pool load) has greatly increased as a result of the 1991
Allen Mills event. The writer ascribes an increase in the TI Pool load from 1 to 2 lb/day as the result
of temporary storage of PCBs in the TI Pool sediments stemming from the Allen Mills event as well
as recently discovered PCB oil seeps. This storage of PCB oils is part of the writer's assertion of
undocumented oil droplets transferring PCBs to the TI Pool while undetected at the Rogers Island
monitoring station, a theory rejected by USEPA (see response to DG-1.3 and DG-1.10). This
analysis cites the USGS data as a further basis for defining this differential load gain.
This analysis suffers from several flaws. First, as discussed extensively in Section 3.3.5 of
the DEIR, the USGS PCB data does not reflect the lightest congeners in the water column due the
nature of the measurement scheme used. These congeners represent a large portion of the water
column load generated from the TI Pool; thus, the USGS data cannot be used to accurately
characterize the entire flux, as noted in the DEIR. Most likely, the USGS data can be used to reflect
the transport of trichlorinated and higher molecular weight congeners (Tri+) although its application
here is still limited due to detection limit and sample timing issues. Conversely, the GE data reflect
the total PCB load gain although the data presented with these comments are not accurate and were
subsequently revised and resubmitted by GE. The revised data, typically yielding higher PCB
concentrations at the TI Dam than those used for the estimates shown by the writer, reflect a large
increase in the water column load as the result of passage through the TI Pool during the period 1991
to 1998, based on the most current GE data submittals. (The USEPA does acknowledge that the TI
Dam monitoring station used for both GE and USEPA sampling programs may overestimate the
actual load gain across the TI Pool under some conditions, based on recent work performed by GE.
Nonetheless, the degree of uncertainty in these estimates does not diminsh the importance of the TI
Pool as a source to the water column.)
On this basis, it is inappropriate to compare the estimated gains of the tri+ congeners based
on the USGS data with those of the entire PCB spectrum reflected in the GE data. The result of this
exclusion is that the writer's presentation now has but a single data point prior to the Allen Mills
event. This point, suggesting a TI Pool loading of 1 lb/day is less than that seen in some, but not all,
of the subsequent years of monitoring. This suggests that the annual loads from the TI Pool have an
intrinsic variability of about +1 lb/day. Thus it is unclear whether 1991 represented a truly different
condition relative to later years. While the USEPA does not dispute that the 1991 Allen Mills event
Deconbet 22. IMS
DEIR-43
T AMS/LTl/T etraT cch/MC A
-------
added to the existing sediment inventory, it is not clear that this additional contamination had a
major impact on the rate of sediment loading to the water column. It is not inconceivable that
similar load gains of about 2 lb/day were seen throughout the 10-year period prior to 1991. Thus,
the additional 1 lb/day load gains seen in the 1992 to 1996 period cannot directly ascribed to the
Allen Mills failure. The contention that the 1995 to 1996 rise in the TI Pool load is ascribed to
additional, undetected loads from upstream of Rogers Island is rejected by USEPA as well since it
is again ascribed to an undemonstrated phenomenon (undetected oil droplets) and because it is not
clear that the TI Pool load changed as a result of the Allen Mills event and subsequent leakages.
3.3.5 Estimated Historical Water Column Loadings Based on USGS Measurements
No significant comments were received on Section 3.3.5.
3.3.6 Conclusions Concerning Historical Water Column Transport
No significant comments were received on Section 3.3.6.
3.4 Integration of Water Column Monitoring Results
3.4.1 Monitoring Techniques and PCB Equilibrium
Response to DS-2.13
The USEPA data alone are not sufficient to rule out the possibility of occasional PCB oil
droplet transport. However, when the more-than-seven years of biweekly GE monitoring data and
the unsuccessful attempts by GE to measure an oil droplet flux are considered, it appears very
unlikely that PCB oil transport represents a substantive portion of the Rogers Island load. It is our
opinion that during the transit of water from Bakers Falls to Rogers Island, there is sufficient
turbulence (due to the falls itself as well as the rapids area near the former dam site) to disperse any
oil droplets into a relatively homogeneous water column concentration which is subsequently
monitored at Rogers Island. The only possible exception to this scenario may have occurred during
the failure of the Allen Mills structure when much greater quantities of PCB oils and contaminated
sediments were released.
3.4.2 Loadings Upstream of the Thompson Island Pool
Response to DC-1.4
Fresh sources of PCB contamination were extensively examined in the DEIR by using
General Eilectrie's water column monitoring data collected at Rogers Island from 1990 through 1996.
These data provided a means of quantitating the load and fingerprinting the contaminant as unaltered
PCBs from above Rogers Island. Since the inception of the Phase 2 water column program, data
collected by the USEPA and GE have documented the significant decline in fresh loadings to the
Hudson originating above Rogers Island. This has left the sediments of the Upper Hudson,
particularly those of the TI Pool, as the major PCB source.
December 22, 1998
DEIR-44
TAMS/lTl/TetraTech/MCA
-------
3.4.3 Loading from the Thompson Island Pool During 1993
Response to Comment DF-2.5
The writer is correct in noting that extensive weathering of the congener pattern seen at RM
177.8 would serve to yield an underestimate of Upper Hudson PCB load to NYC harbor. However,
the degree of weathering is difficult to assess independent of the other processes affecting PCBs and
so the degree of underestimation is difficult to determine. The possibility that the PCB contribution
from the Upper Hudson could exceed 50 percent of the total PCB load to the harbor is considered
to be well within the uncertainty of the estimate contained in the report.
Response to DS-2.14
The fate of recent PCB releases from the GE facilities is expected to be the same as the
historical releases from the facilities, i.e., processes such as sediment-to-water partitioning, gas
exchange, dechlorination, biological uptake, deposition and scour will slowly disperse the recently
released PCBs throughout the Hudson. The more difficult question is "What is the average residence
time for PCBs in the biogeochemically active portion of the sediments and how long will it take
before the PCBs in these sediments are purged from the sediments or deposited in areas of long term
burial?" The issues pertaining to the resolution of this question continue to be examined as part of
the Phase 2 modeling effort.
Response to DL-1.1
The contention, also claimed by GE, that a portion of the TI Pool load originates with
undetected oil droplets is unfounded. Despite GE's many attempts to find such droplets, none have
been detected. In fact, some of their most recent results from a sampling cross section of the river
just above Roger Island (QEA, 1998) shows the water column PCB concentration to be relatively
homogeneous, suggesting the absence of oil droplets. (Presumably, oil droplets near the bottom
would cause the deeper samples to yield markedly higher PCB levels.) Even if such droplets were
to exist, it is unclear how long it would take for these PCBs to leave the sediments and re-enter the
water column. Nearly all the PCBs present in the bottom sediments were once released as oil
droplets in GE's discharges. Much of the sediment burden, though clearly not all of it (USEPA,
1998) is still in place, some at depth, most within 9 inches of the surface. As discussed in USEPA,
1998, the sediments of the TI Pool are clearly not static, lake-like deposits but rather a dynamic
environment subject to resuspension and burial as well as diffusive and biological processes. Thus,
the simple addition of more PCBs during the period from September 1991 to 1996 serves to worsen
the pre-existing problem but certainly does not define it. The strongest evidence for this fact comes
from the GE data, which demonstrates a measurable TI Pool input prior to the September 1991 event
as well as the consistency of the size of the TI Pool load each year from 1993 to the present despite
the major reductions in the loads from upstream of the Pool. Lastly, the congener patterns of the TI
Pool load are consistent with a sediment-derived source which has been subject to a moderate
degree of dechlorination and not a fresh Aroclor mixture as might be derived from oil droplets. Thus
the need for this conclusion, i.e., that the sediments, and not any other phenomenon, are responsible
for the TI Pool load.
December 22. 1998
DEIR-45
TAMS/LTT/TelraTech/MCA
-------
3.4.4 Loading at the Thompson Island Dam - 1991 to 1996
Response to DS-2.15
Acknowledged. "The CSO was repaired in May 1993; water from the CSO still flowing
through the Allen Mill and discharging out the Tailrace tunnel until it was fixed"; this helps explain
the observed extensive PCB loadings which continued until June, 1993.
Response to DS-2.16
USEPA believes that the loads produced from the GE facilities from Sepember 1991 to June
1994 impacted the sediment-related PCB loads originating within the TI Pool. The data set collected
by GE prior to the Allen Mills failure represents only a few months of sample collection and is not
a sufficient basis on which to assess the change in the TI Pool sediment load to the water column.
The fact that sediment-related loads are still in evidence in 1998 and have not yet appeared to have
changed over time (GE post construction monitoring, 1998), despite the substantial reduction in
loads originating above Rogers Island, suggests that the TI Pool load has existed prior to September
1991 and will continue to exist for some time.
Response to DS-2.17
Yes (i.e., the TIP sediments, both those contaminated years ago combined with those recently
contaminated, are the primary source of PCBs to the river), with a lesser fraction of additional PCB
load from sediments downstream of the TI Dam.
Response to DS-2.18
This issue has been addressed in Section 3.4.4 and in Figure 3-106. In particular, this figure
shows that the PCB loads in the Upper Hudson were originally dominated by sediment-related
releases prior to 1991. Subsequently, the PCB releases from the Allen Mills dominated the Upper
Hudson and the total PCB loading to the water column increased five-fold. Nonetheless, sediment-
related releases were still in evidence. In the period June 1993 to June 1994, the sediments again
became the dominant source to the water column, perhaps at twice the loading rate seen in 1991.
At the same time the overall loading rate declined to one third of that observed for Sept. 1991 to May
1993. In later years, the sediment-related loads and the overall loading rate appear to have returned
to the levels seen in 1991. Data collected subsequently by GE shows that further reduction in the
load from the GE facilities has further declined while that from the TI Pool has remained relatively
constant. The change in these loads will be examined in detail as part of the "hindcasting" analysis
to be completed during the modeling analysis.
3.4.5 PCB Loadings to Waterford
Response to PL-1.2
The original presentation in the DEIR indicated that PCB loads at the TI Dam were
essentially equal to those at Waterford under many different conditions, despite the 30 miles of river
separating the two monitoring stations. As discussed in the corrections to Section 3.2, the USEPA
December 22. 1998
DEIR-46
TAMS/LTI/TetraTech/MCA
-------
is currently revising its assessment of PCB loads in the Upper Hudson. This analysis is expected to
show a substantial decline in the loads at Waterford and Stillwater relative to those at the TI Dam
under low flow cionditions based on revisions to the water flow rates at the time of sampling. The
analysis is also expected to show that the sediments between the TI Dam and Schuylerville may also
contribute to the water column load as was noted in the DEIR. Nonetheless, the analysis presented
in the DEIR did not indicate that the sediments below the TI Dam were subject to a different set of
processes than those of the TI Pool, as purported by the writer. Rather the text states that the PCB
losses and gains between the TI Dam and Waterford must balance each other such that the total loads
delivered to Waterford were the same as those generated from the Pool. The anticipated results from
the re-analysis of the Phase 2 transect data will not change the finding that the TI Pool is the
dominant source of PCBs to the freshwater Hudson under low flow. In fact, since PCB loads will
be shown to decrease below Schuylerville, the results will indicate an even smaller contribution from
the river sediments between Schuylerville and Waterford to the water column load at Waterford.
While it may be interesting to speculate why the region below Schuylerville contributes little to the
water column load (for example, 20 of the 40 hot spots are found in the TI Pool and 35 of the 40 hot
spots are located upstream of Schuylerville), the data show no significant increase in PCB load to
the water column from sediments below Schylerville.
3.4.6 PCB Loadings to the Lower Hudson
No significant comments were received on Section 3.4.6.
3.5 Integration of PCB Loadings to Lower Hudson River and New York/New Jersey Harbor
No significant comments were received on Section 3.5.
3.5.1 Review of Lower Hudson PCB Mathematical Model
3.5.2 Estimate of 1993 PCB Loading from the Upper Hudson River
3.5.3 Revised PCB Loading Estimates
No significant comments were received on Sections 3.5.1 through 3.5.3.
3.6 Water Column Conclusion Summary
Response to DS-2.19
It is highly unlikely that either PCB type (i.e., old or recently-contaminated sediments) is
solely responsible for the water column load generated by the sediments. Most likely, the PCB load
is a combination of both recently deposited and older PCBs. Further resolution of this issue will be
completed during the modeling analysis.
Response to DS-2.20
The importance of PCB loads from the sediments below the TI Dam will be further examined
during the modeling effort. As noted in the corrections to Section 3.2, total loads downstream of
December 22. 1998
DEIR-47
TAMS/LTl/TelraTech/MCA
-------
Schuylerville are expected to be markedly lower relative to the original DEIR estimates. See the
corrections to Section 3.2 for more detail.
December 22. i»8
DEIR-48
TAMS/LTI/TetiaTech/MCA
-------
Chapter 4 - Inventory And Fate of PCBs in The Sediment of The Hudson River
Response to PF-2.6
Although the text does not discuss the analyses, the physical data collected as a part of the
sediment coring efforts were examined in detail. Little correlation among PCB mass, sediment
texture, and other physical parameters was found. This is believed to stem from the fact that the high
resolution cores themselves showed relatively little variation in their physical properties, both
vertically within a core as well as among cores. The lack of variability is attributed to the careful
coring site selection process, which focused on finding depositional environments consisting
exclusively of fine sands and silts where little evidence of episodic sediment emplacement could be
seen. As a result, the high resolution cores are physically very similar and thus the correlations
between the PCB inventory and the physical sediment parameters was relatively weak. As was
demonstrated in the Low Resolution Sediment Coring Report (USEPA, 1998). when samples are
collected from a range of environments, yielding a range of physical properties, correlations among
PCBs and other sample characteristics can be seen.
The POC values developed for the water column samples were in the range of 0.3 to 3.4
mg-organic carbon/L, excluding the high flow conditions occurring around Transect 3. As to the
correctness of this data, a graph of total organic carbon (TOC) vs total suspended solids (TSS) is
provided in this response summary. TOC represents the fraction of organic matter in the suspended
matter. TOC is calculated from the weight-loss-on-ignition (WLOI) results for the suspended matter
and the relationship between TOC and WLOI developed for the high resolution sediment cores, as
described in the DE1R. The fraction of organic matter would be expected to vary inversely with
suspended solids in this system, since higher suspended solids loads are associated with high flows
and subsequently little in situ production of organic carbon. Conversely, at the low flow conditions
typical of summer, TOC would be expected to be high relative to TSS due to in situ organic carbon
production via photosynthesis coupled with little resuspension or erosion.
Figure DF-2.6 shows the relationship between TOC and TSS, supporting this general
relationship and indicating the validity of the data. This figure represents both transect and
flow-averaged samples for all mainstem Hudson stations as well as the tributary data collected. The
high TOC values are generally attributed to water column photosynthesis, which will yield
substantially higher TOC fractions as compared to high flow conditions or the sediments of the river
itself. The high flow/high TSS results are more commensurate with the TOC levels found in the high
resolution core sediments (0.3 - 12% TOC) as would be expected from scour of river sediment as
well as erosion of soils, both of which will contribute suspended matter with this level of TOC. The
high TOC levels seen in low flow conditions would not be expected in the sediments of the river
since biological reworking of suspended matter and surficial sediment efficiently extracts much of
the TOC from the particles. It should be noted that the calculated POC values discussed in the report
and shown on Figure 3-2 are simply the product of the TSS and TOC results for each of the
mainstem transect samples.
The writer is referred to the Low Resolution Sediment Coring Report (USEPA, 1998) for
further information concerning the relationships among PCBs and the physical parameters.
December 22 1998
DEIR-49
TAMS/LTl/TetraTech/MCA
-------
e
o
.o
u
«
U
B
a
w>
u
©
bp
'3
i
-a
J <8
^ -O
-O
S TD
T3 c
e
fl> Q.
9- 3
l/) 3
3 C/3
CO ^
0-H
o o/)
1 £-2
H £ *
r* cj
OA
u
o
o
o
60 -
50
40 -
30
20 -
10 -
_i i i i i i
-J L. 1 I I 1 l_
o
o
o
© e
o
o
o
°
-------
4.1 Characterization of Upper Hudson Sediments by Acoustic Techniques
4.1.1 Geophysical Data Collection and Interpretation Techniques
4.1.2 Correlation of Sonar Image Data and Sediment Characteristics
4.1.3 Delineation of PCB-Bearing and Erodible Sediments
4.2 Geostatistical Analysis of PCB Mass in the Thompson Island Pool, 1984
4.2.1 Data Preparation for PCB Mass Estimation
4.2.2 Geostatistical Techniques for PCB Mass Estimation
4.2.3 Polygonal Declustering Estimate of Total PCB Mass
4.2.4 Geostatistical Analysis of Total PCB Mass
4.2.5 Kriging Total PCB Mass
4.2.6 Kriged Total Mass Estimate
4.2.7 Surface Sediment PCB Concentrations
4.2.8 Summary
No significant comments were received on Sections 4.1 through 4.2.8.
4.3 PCB Fate in Sediments of the Hudson River
Response to DG-1.18
The writer's concerns focus on the effects of anaerobic dechlorination on PCB toxicity and
bioaccumulation. Issues concerning PCB toxicity will be covered in the Ecological and Human
Health Risk Assessments. The USEPA will consider comments related to these issues at the time
of their preparation. The issue of bioaccumulation will examined at that time as well. However, two
considerations can be discussed here in this context. First, while dechlorination does reduce the
tendency for bioaccumulation by converting heavier molecules to lighter ones, typically with lower
octanol-water partition coefficients (Kou), it has a second, related effect. In decreasing the partition
coefficient, the process also increases the tendency of the PCB molecule to dissolve in the porewater
and subsequently migrate, thus increasing the probability for biological exposure. In this manner,
greater exposure occurs to congeners of lower molecular weight. This result may or may not be
offset by the decrease in the tendency for bioaccumulation.
The second and more important issue concerns the lack of extensive dechlorination in most
sediments of the Hudson. Most sediments have not seen extensive dechlorination simply because
they lack the concentrations sufficient to support extensive dechlorination (hence the relationship
December 22.
DEIR-51
TAMS/LTl/TelraTech/MCA
-------
between the MDPR or AMW and PCB concentration). Because extensive dechlorination of PCBs
in the Hudson River is largely limited to the TI Pool and the areas immediately downstream, the
impact of dechlorination on bioaccumulation and toxicity is not as important as the writer contends.
Response to DG-1.19
The writer contends that the measures used to identify and quantify the dechlorination
process are insensitive and inappropriate to characterize the conditions in the Hudson River. In
particular the writer focuses on the extraction of a pure PCB oil as a possible means to enhance the
concentration of the congeners included in the molar dechlorination product ratio (MDPR). Also the
writer contends that by focusing on the final or near final end products, the MDPR does not reflect
the intermediate steps. Also the measure does not account for the final transformation from BZ#8
to BZ#1.
There are any number of possible ways to reflect the degree of dechlorination in a sample.
The DEIR presents two, the MDPR and the change in molecular weight with respect to Aroclor 1242
(AMW). GE has offered several additional suggestions in its comments, e.g., the ratios of 56:49,
60:49, 66:49, 74:49 and the number of chlorines per biphenyl. Since the dechlorination process is
the process of cleaving chlorines from a mixture, it is then appropriate to examine the average
number of chlorine atoms per molecule as a measure of the dechlorination. As it turns out, the term
AMW is algebraically linked to the number of chlorines per biphenyl (Cl/BP) as shown below. Thus
the two measures are equivalent and directly reflect the chlorine content of the congener mixture.
AMW is defined as:
A MW
MW
A J 242
MW
sample
MW
AI242
Solving for the molecular weight
of the sample:
MW
sample
= (1 - LMW)MW
A1242
Number of chlorine atoms per
biphenyl (Cl/BP) is defined as:
Mass Cl
No of CI atoms _ molecule
Biphenyl
Atomic WeightCl
where the mass of chlorine per
biphenyl molecule is given as:
Mass Cl
molecule
MW
sample
MWB,PHtnyt ~ ~ * At0miC WeiShtH
Substituting:
Cl
MW. - MWBp * —
a 'amP>< bp Bp
BP
Atomic Weight
ci
Decemfeet 21 I99»
DEIR-S2
TAMS/LTI/TetraTech/MCA
-------
(1 - AMW)MWAI242 - mwbp +
Cl
Substituting for the molecular
weight of the sample:
Cl
BP
Atomic Weightcl
BP
Solving for Cl
BP
(1 - AW) (265.7) - 154 35.45 '
35.45 35.45-1,
Yields a simple algebraic
relationship between Cl/BP and AMW:
— = 3.24 - 7.70 AMW
BP
where:
mwa1242 = molecular weight of Aroclor 1242 (265.7 g/mole) based on the Phase 2 analysis
Atomic Weight^-, = atomic weight of chlorine (35.5 g/mole)
Atomic WeightH = atomic weight of hydrogen (1 g/mole)
The MDPR is simply the proportion of congeners which have been converted to final
dechlorination products in a given congener mixture. As it turns out, the MDPR correlates very
strongly with AMW as shown in Figure 4-21 of the report. This is no coincidence, since in order for
the two parameters to correlate, the change in AMW must be accomplished by the conversion to the
final dechlorination products. This does not mean that every molecule converted must go to
completion immediately once it begins to dechlorinate but rather that the amount of conversion to
the final products is simply proportional to the overall degree of dechlorination of the entire suite
of molecules. Ultimately it is the conversion of the molecules to lighter congeners that is the focus
of the dechlorination process. That some minor components rapidly convert relative to the vast bulk
of the sediment PCBs is of only marginal interest and certainly does not substantively modify the
nature of the mixture.
GE proposes several ratios as other measures of various processes occurring in the sediments.
The ratios given above (i.e.. 56:49. 60:49, 66:49. and 74:49) are said to be related to H, H'
dechlorination as defined by GE. These ratios can be compared to the Cl/BP ratio as is shown in
Figure DG-1.19A. These ratios are plotted using a linear scale as opposed to the log scale used by
GE to better clarify the degree of variability; also, all data are shown. (The GE representation shows
only the mean and two standard errors on the mean.) While each ratio declines with decreasing
Cl/BP. they are clearly not linear relationships as reflected by the poor R2 values. Even if the data
are fit with a logarithmic curve, the R2 improves only marginally. These curves can be compared
with the relationship between MDPR and Cl/BP shown in Figure DG-1.19B. In this instance, the R2
is markedly better at 0.94 (the same as that for AMW as would be expected). Thus the MDPR is a
very strong predictor of the degree of dechlorination as reflected in the number of chlorines per
sample
molecular weight of sample
December 2: 1W8
DEIR-53
TAMS/LTfTetraTech/MCA
-------
Upper Hudson Samples
Upper Hudson Samples
-e— 74:49
1.2
1
0.8
0.6
0.4
0.2
0
' > = -0.59: + 0.367x R*= 0.3W
*¦
O
^scc
V>
os
Q'
•5
1.2
1
0.8
0.6
0.4
0.2
0
' y = -0.822 + 0.447x R-= (J.408
..C.
: V-O ^
O \ pCg*
\
.CgP~P
tc
VO X
"c \
c „ ®
•• Gfi
~ e.c\
c o
Cl/BP
3 2
Cl/BP
66:49
"T
Upper Hudson Samples
¦o
>*
X
' v= O.I37 r()()X36x R"= 0 13
0.5
0.4
0.3 "
~ 0.2
0.1 -
L'
x Cv 3& -
xx'~ o
x^cX;'
Gj£? ^ ~ n
; s&w,
s— 60:49
Upper Hudson Samples
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
y = -0.41W ~ (>.2h2x R'= D.2VS
vs.
o%-
0&> \C
b
-e~ 56:49
Cl/BP
Cl/BP
• Rogers Island Water Column
(Suspended Matter)
o Upper Hudson River
Sediment Samples
Hudson River Database Release 4.1 TAMS/LTl/TetraTcch/M
Figure DG-1.19A
The Number of Chlorines per Biphenyl vs. the GE/HydroQual Dechlorination Ratios
for the High Resolution Core Data
-------
y = 1.62 - 0.453x R = 0.94
0.8
= s
.2 o-
11 °-6
-2 o
JC \S
O KS
v ck
Q TZ 0.4
S 3
eu
0.2
0
—1—1—1—1—
—i r—H—| "T—i—i—r—j—i—i—i— T11 ¦
—1—1—1—I—
-
_
]Kr
mP
-
Ogp c
-
-
-
O :
1 ! 1 1 Y 1 -V 1 > 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
3 2
Cl/BP
0
Note:
Data from Phase II Upper Hudson samples shown. 10.0138
Hudson River Database Release 4 I TAMS/LTI/TetraTech/N,
Figure DG-1.19B
The Relationship Between the Number of Chlorines per Biphenyl
flnH flip N4r>lflr npp}ilnrinntir\n Prn^urt I?otirx fr\r flip UinK U<»cr»lntiAn (~*r\re* Hotfl
-------
biphenyl and is far better than the individual congener ratios given above. While there may be better
ratios to be discovered within the data, the MDPR is certainly a useful measure of the degree of
dechlorination seen by a mixture.
In particular, for the sediments of the high resolution cores, where isolation from active
exchange with the water column is likely, the MDPR presents an excellent measure of dechlorination
on what is effectively a closed system. As to the enhancement of ratios by extraction of PCB oils,
the USEPA does not accept the premise that extensive amounts of PCB oils pass the Rogers Island
monitoring station. Rather, by the time PCBs released from the Hudson Falls area reach Rogers
Island they have been largely homogenized in the water column (see response to DG-1.15) and thus
there is no issue of relative extraction. The whole of any PCB oils have been incorporated in the
dissolved and suspended matter phases of the water column. Thus the effect of any oil extraction
process on the MDPR is moot. As further proof of this, note how the Rogers Island water column
stations fall along the MDPR/AMW relationship, quite close to the point representing Aroclor 1242
(see Figure 4-29 of the DEIR). Thus, in the only environment where PCB oils have been shown to
exist, the water column mixture has already incorporated the entire congener mixture present in the
oil. If preferential extraction were occurring, the mixture would have a higher MDPR. Clearly the
values at Rogers Island tend to fall to lower, not higher values of MDPR relative to Aroclor 1242,
perhaps indicative of small amounts of Aroclor 1254 or 1260 in the congener mixture entering the
river. No preferential dissolution of the mixtures entering the river at Hudson Falls is evident. Based
on these observations, it is clear that the measures used by the USEPA in the DEIR are appropriate
and accurate reflectors of dechlorination.
Response to DG-1.21
The writer contends that dechlorination is still occurring in Hudson sediments. Specifically,
sediments less than 30 mg/kg still exhibit changes in several indices used by GE to identify
dechlorination processes. The writer also contends that recently deposited sediments will undergo
dechlorination as well.
The USEPA's assessment of dechlorination has focused on the "big picture" in terms of
dechlorination, i.e., can dechlorination be expected to substantively reduce the inventories of PCBs
in the sediments of the Hudson. After review of the writer's contentions, the USEPA's conclusion
remains the same. Dechlorination represents only a minor process in reducing PCB mass and then
only in the most contaminated sediments. Dechlorination can affect a large number of molecules in
these instances but for the vast majority of sediment contamination in the Upper Hudson, with
anticipated concentrations less than 100 mg/kg, dechlorination is largely unimportant.
The USEPA acknowledges the writer's observation that minor components in the congener
mixture may continue to dechlorinate over time, and these are in fact noted in the DEIR in both
sections 3.3.3 and 4.3. However, as shown in Figure DG-1.19A, these ratios are not good predictors
for the degree of dechlorination. In preparing Figure 32 of the GE comments, the writer has selected
sediments less than 30 mg/kg in showing the relationship of dechlorination with depth. These results
show that dechlorination of these individual congener pairs does not approach a final value until
perhaps 25 cm of depth. At deposition rates typical of the high resolution cores depicted in this
figure, this would suggest roughly 25 years before these ratios had reached their final values. Thus
for these congeners at least, the dechlorination process is a very slow one. These results need to be
December 22. 1998
DE1R-56
TAMS/LTT/TctraTech/MCA
-------
compared with Figure 4-28b of the DEIR. This figure depicts the absence of a time dependence for
dechlorination when all sediments are examined. From this it can be concluded that more massive
I
samples reach their final or near final level of dechlorination quite rapidly while sediments at low
levels of contamination may take decades to reach their final level of dechlorination. Regardless of
the time taken to reach these final levels, it is also clear that for sediments of low contamination
(<100 mg/kg), the final level of dechlorination represents only a minor change in sediment PCB
mass and modifications to a limited number of congeners.
The USEPA also agrees that recent deposition will be subject to some degree of
dechlorination. However, since recent deposition is substantially lower in PCB contamination
relative to the peak historical levels (35 mg/kg vs 2,000 mg/kg in the TI Pool), the dechlorination
process will probably take over 25 years to complete and therefore will have only a minor impact
on the recent sediments and can largely be ignored as an important mechanism to reduce or modify
the PCB mass in these sediments.
Thus, while dechlorination may still occur in the sediments of the Upper Hudson, it remains
only a minor process for sediments of relatively low (<100 mg/kg) contamination and will not have
a major impact on the PCB inventories beyond that already measured. In this context, it is clear that
dechlorination in the sediments of the lower Hudson is negligible.
Response to DG-1.26A
Resuspension of sediment and porewater displacement are mechanisms suggested for
transport of PCB contamination from sediments at depth to the water column in the DEIR.
Additional transport mechanisms include sediment displacement due to water craft activities and
bioturbation. These mechanisms correspond well with the seasonal nature of the TI Pool loading.
It is clear that PCBs are transported from TI Pool sediments to the water column based on the PCB
mass loss between 1984 and 1994 documented in the Low Resolution Sediment Coring Report
(USEPA, 1998) of which only 11 percent can be attributed to dechlorination.
USEPA intends to examine the information available to estimate the depth of the sediment
inventory responsible for the TI Pool load. However, given the dynamic nature of the sediments of
the Pool, it is clear that no single depth can accurately describe the exchange process.
4.3.1 Anaerobic Dechlorination and Aerobic Degradation
Response to PI .-1.7
Changes in PCB toxicity due to dechlorination are unclear. While heavier congeners are
associated with carcinogenicity, the lighter ones are associated with neurological impacts. Thus it
is unclear as to net effect of dechlorination on toxicity. The amount of bioaccumulation by fish and
the carcinogenic and noncarcinogenic risks to human health associated with the levels of PCB
contamination detected in the Hudson River will be studied in the Ecological and Human Health
Risk Assessments (ERA and HHRA. expected in 1999). The health effects of these contaminants
cannot be dismissed in light of the scientific literature. PCBs have been shown to cause metabolic,
neurological, developmental or reproductive affects on various species. A more complete review of
the toxicological literature will be presented in the HHRA.
December 2-. 1^8
DEIR-57
TAMS/LTLTetraTech/MCA
-------
Response to DG-1.26B
The statement is made in the DEIR (page 4-50) that, 'in general, these aerobic processes
affect only the lightest congeners, monochloro- to trichlorobiphenyls, and are ineffective at altering
heavy congeners (those with four or more chlorine atoms) under environmental conditions." This
statement was based upon review of literature which reported the occurrence and mechanisms of
PCB destruction under aerobic conditions (e.g., Furukawa, 1982; Bedard etal., 1987; Bedard, 1990;
Abramowicz and Brennan, 1991). While the presence of microorganisms which are capable of
degrading PCBs containing as many as six chlorines may have been established, the effectiveness
of these organisms in degrading significant amounts of PCB contamination has not been established
particularly in sediments under anaerobic conditions. The high resolution core results of the DEIR
as well as the results of the Low Resolution Sediment Coring Report demonstrate the continued
presence of high PCB concentrations in many areas of the Hudson. The fact that these areas of PCB
contamination still exist decades after their deposition indicates that aerobic degradation has not
reduced PCB contamination in sediments of the Hudson River to a significant degree.
Aerobic degradation and loss due to gas exchange are mechanisms proposed to explain the
loss of mono- and di-chlorobiphenyls from the water column found in lower river samples.
Response to DG-1.26C
On page 4-50 of the DEIR the statement is made that, "Reductive dechlorination of PCBs
by microorganisms from the Hudson River occurs anaerobically, i.e., in the absence of oxygen,
primarily through the removal of the meta-chlorines and, to a lesser extent, the para-chlorines in
PCBs bound to subsurface sediments (Rhee et al., 1993b; Quensen et al., 1990)." This sentence
explains that reductive dechlorination is an anaerobic process involving the removal of meta- and
para-chlorines which acts on subsurface sediment contamination. It is implied that reductive
dechlorination is the primary biotransformation acting on the subsurface anaerobic sediments.
Biodegradation of PCBs may occur in anaerobic sediments, but this primarily results from complete
dechlorination of the approximately three mole percent of congeners without ortho-chlorines which
can be transformed to biphenyls.
This statement does not rule out the possibility that aerobic and anaerobic biotransformations
may occur in surface sediments, but the impact of these processes is unlikely to be substantial, given
the close agreement of the PCB concentrations of recent sediment and Hudson River suspended
matter. It is incorrect to simply state that the source of PCB contamination to fish is PCBs in the
surface sediment and only the newest releases of PCBs (presumably from the Bakers Falls area).
The source of PCB contamination to fish will be examined in the Ecological Risk Assessment to be
released in 1999.
Response to DG-1.26D
As stated in the DEIR (page 4-51), "[t]he dechlorination process is more effective on the
heavier PCB homologues, in which meta- and para-chlorines occur frequently." The statement
indicates the PCB congeners with more meta- and para- chlorines can be altered more by reductive
dechlorination than congeners with less meta- and para- chlorines. For example, a congener with
nine chlorines might lose at least five meta- and para- chlorines through dechlorination while a
December 22. 1998
DEIR-58
T AMS/LTiyT etraTech/MC A
-------
congener with three chlorines could lose at most three meta- and para- chlorines. No mechanistic or
rate of reaction concepts are implied by this statement.
Response to DG-1.26F.
"The process of dechlorination has the net effect of reducing the mass of PCBs within the
sediments without reducing the total molecular PCB concentration unless the process removes all
chlorine atoms." This sentence from the DEIR implies that the molar PCB concentration will
decrease by the amount of congeners with only meta- or para- chlorines which can be removed by
reductive dechlorination. As stated in comment DG-1.26E, there are only approximately three mole
percent of congeners which can be degraded by reductive dechlorination, so the molar concentration
is not greatly lowered by this process, even if all congeners with meta- and para-chlorines are
degraded. Table 4-8 of the DEIR lists all congeners used in the report with final dechlorination
products.
The toxicity of the sediments contaminated with PCBs will be quantitated in the Ecological
and Human Health Risk Assessments due in 1999. The level of risk associated with any congener
cannot be inferred a priori by deductive reasoning.
Response to DG-1.26F
The toxicological affects of PCB congeners will be assessed in accordance with the USEPA
guidelines in the Ecological and Human Health Risk Assessments due in 1999.
Response to DG-1.26G
The primary biotransformation process affecting PCB contamination in anaerobic sediments
is reductive dechlorination. Aerobic degradation may occur in low oxygen levels, but the importance
of this process relative to dechlorination is small if only some low. unquantified level of degradation
occurs. Ultimately, it is the continued presence of identifiable Aroclor 1242-like mixtures in
sediments as old as 40 years which indicates that these degradative processes are largely unimportant
in substantively reducing the sediment PCB inventory.
Response to DG-1 26H
The Low Resolution Sediment Coring Report (USEPA, 1998) contains the analysis of
anaerobic dechlorination in the high and low resolution cores. The dechlorination indices used in this
report (molar dechlorination product ratio and fractional change in mean molecular weight relative
to Aroclor 1242) are sufficient for the purposes of the report. As noted in the responses to DG-1.19
and DG-1.20, other indices do not appear to be as sensitive or usefiil in assessing the overall degree
of dechlorination.
Decemberms
DE1R-S9
TAMS/LWT«raTech/MCA
-------
4.3,2 Anaerobic Dechlorination as Documented in Phase 2 High-Resolution Sediment Cores
Response to DG-1.26J
The MDPR is an estimate of the amount of dechlorination found in the Hudson River
sediments. It is an estimate based the assumptions outlined in the text, such as the PCB
contamination in the river is primarily Aroclor 1242. In regions were these assumptions do not hold,
the text discusses the consequences. From the DEIR, page 4-62, "The Lower Hudson sample
MDPRs tend to cluster just below the Aroclor 1242 value of 0.14. The mean MDPR for the Lower
Hudson is 0.11, suggesting the presence of a minor contribution by heavier Aroclors or, more likely,
possible loss of BZ#1,4, 8,10, and 19 prior to deposition due to their generally greater solubility
and degradability. The congener pattern comparisons made in Chapter 3 (Subsection 3.3.3), suggest
that both processes probably occur to some degree."
The power of the relationship between MDPR and total PCB is evident in the correlation and
range of values spanned. This relationship cannot be dismissed by argument, particularly when
alternative dechlorination ratios developed by GE/QEA are weakly predictive at best (see response
to comment DG-1.20).
Response to DG-1.26J1
The congener BZ#8 was included in the MDPR as a dechlorination product even though it
is the third most abundant congener in Aroclor 1242 by mass (approximately 7.3% of the mixture,
according to the Phase II analysis of standards). Inclusion of this congener is justified by the fact that
it is a significant intermediate dechlorination product with 14% of the high resolution core samples
containing more than 7.3% of BZ#8 and up to 25% by mass. The correlation between MDPR and
total PCBs was improved by including this congener. Finally, it is unimportant that the MDPR is
unchanged whether BZ#8 is dechlorinated to BZ#1 or not, because the MDPR is a measure of
dechlorination as shown by end products and a significant intermediate. The MDPR is not a direct
measure of all the chemical activity undergone by the sample as a result of dechlorination. However,
as discussed in DG-1.19, the MDPR does correlate strongly with the number of chlorines per
biphenyl, indicating singly that the MDPR changes in direct relation to the extent of dechlorination.
Response to DG-1.26K
The MDPR is an estimate of the number of affected PCB molecules. This estimate allows
insights on the nature and degree of dechlorination in the Hudson River sediments to be gained. As
discussed in the text, the MDPR is an underestimate of the number of affected PCB molecules
because the intermediates are not accounted for and lighter molecules being more soluble and more
susceptible to degradation processes may be lost from the sediments. Nevertheless, this measure
provides a relationship between the degree of dechlorination and the PCB concentration which is far
stronger than the GE/QEA dechlorination products which are designed to account for dechlorination
of specific congeners (see the Response to DG-1.20). Since the MDPR increases directly with the
decline in Cl/BP, and represents a sum of dechlorinated molecules, it is in fact a measure of the
molecules affected although not all affected molecules are directly represented in its sum.
December 22.
DEIR-60
T AMS/LTIT etraTcch/MC A
-------
Response to DG-1.26L
The text of the DEIR on page 4-57 is a discussion of reductive dechlorination in sediments
subject to an anaerobic environment and the MDPR as a measure of this process. Other processes,
such as aerobic degradation and photo-destruction in the water column are discussed in other
portions of the document, where appropriate. On page 4-61, the fact that the MDPR and AMW
cannot account for mass loss by degradation is noted. Although degradative loss may occur in the
sediments, these losses have not be demonstrated by be significant in the Hudson River sediments
(page 4-65). The potential to bioaccumulate and the risks associated with the contaminants detected
in the sediments and water column of the Hudson River will be examined in the Ecological and
Human Health Risk Assessments.
Response t
-------
Hudson have traveled far more miles and have descended over one hundred feet at the various dam
spillways of the Upper Hudson. There is then more time for re-partitioning from suspended matter
into the water column and exposure to gas-exchange aerobic degradation and photo-degradation.
Response to DG-1.26P
From the DEIR, page 4-65:
The fact that meta- and para-dechlorination can only decrease the PCB mass by a maximum
of 26 percent implies that sediment PCB inventories sequestered over the previous 40 years
(1954 to 1994) will remain in place unless subject to scour, extensive purging by porewater,
or remediation. Although some authors have demonstrated degradative PCB losses from the
sediments (e.g., Rhee et al, 1993b), these losses do not appear significant for the mix of
PCB congeners present in the Hudson.
As stated in the text, the theoretical maximum PCB mass loss due to dechlorination is 26
percent as measured by the described endpoints. Although, the existence of degradative processes
have been demonstrated, the amount of mass loss due to these processes has not been quantified and
appears to be relatively insignificant given the large quantity of PCBs still stored in the Hudson
River sediments (USEPA, 1998).
Response to DG-1.26Q
From the DEIR, page 4-69:
Given the correspondence of AMW and mass loss, it is important to note that sediments with
PCB concentrations less than the 30,000 jag/kg threshold, on average, experienced little, if
any, mass loss via dechlorination. Mass loss greater than 10 percent is generally restricted
to sediments of 30,000 fig/kg or higher, which in turn are still subject to the maximum loss
of 26 percent.
The value of 30 ppm occurs at the intersection of the initial AMW value of Aroclor 1242 and
the lower 95 percent confidence interval about the data. Below 30 ppm, the occurrence of
dechlorination is not predictable using AMW as a measure, because data fall above and below the
initial AMW value of Aroclor 1242. It is possible that samples with AMW values less than that of
an Aroclor 1242 have undergone dechlorination and preferentially lost the mono- and
dichlorobiphenyls.
While it is possible that some combination of dechlorination plus porewater migration could
yield the MDPR and AMW results seen for the sediments of the Lower Hudson, this would
subsequently require that this not occur in the sediments of the Upper Hudson where dechlorination
end products clearly build up in the sediments, yielding high MDPR values for the most concentrated
samples. If porewater migration or some other removal or degradative process were occurring, these
sediments would be the most likely candidates since they are so concentrated. The possibility
suggested by the writer is also contradicted by the congener evidence, which shows strong
similarities in congener patterns among sediments throughout the Hudson. Presumably if the
processes suggested by the writer were occurring, these similarities would not exist. The more
December 22, 1998
DEIR-62
TAMS/LTI/TetraTech/MCA
-------
dechlorinated but depleted sediments would no longer resemble the original mixture since the more
dechlorination sensitive congeners would be depleted relative to other mixtures. Furthermore, the
PCB/l37Cs relationship demonstrated for the Hudson from Stillwater to Lents Cove would no longer
apply since the downstream dechlorinated plus purged sediments would exhibit lower concentrations
than those predicted from upstream.
Response to DG-1.26R
From the DEIR at page 4-70:
The absence of a correlation with time can be seen more clearly in a subset of the samples
represented above. Using only dated sediment cores from the freshwater main stem Hudson,
the degree of dechlorination as measured by the AMW is plotted in a histogram as a function
of time of deposition (see Figure 4-28a). The data are grouped into approximately ten year
intervals beginning in 1954, Evident in the diagram is the clustering of the values around a
AMW value of 0, More important for this discussion, however, is the fact that a wide range
of mass loss (i.e., AMW) can be seen for any time period, indicating the lack of time
dependence for this process. The samples with the greatest degree of dechlorination are from
the period 1965 to 1974. However, this period also contains samples with essentially no
measurable dechlorination. Similar ranges can be seen for the periods 1955 to 1964 and 1975
to 1984. Only the most recent period presented, 1985 to 1992, has a lesser range. However,
its results are still consistent with those of the previous time periods. The limitation here is
simply that water column PCB loads during this most recent period were less than those of
the earlier periods, resulting in lower sediment PCB levels. (Historical water column
transport is discussed in detail in Section 3.3.)
As stated in the text, all datable cores from the freshwater mainstem Hudson are represented
in Figure 4-28a. The upstream source and low-level recent contamination are represented by the
1985-1992 period (more recent deposition). Because the top segments of the cores were more finely
sliced (2 cm), the more recent deposition is overweighted. In Figure DG-1.26R, this is corrected by
double counting the deeper samples which were sliced at 4 cm intervals. This correction does not
affect the main conclusion drawn from this analysis, namely that all time periods display a wide
range of mass loss. Additionally, the statement in the comment response that "there are >2x the
number of "new sediment samples" than ''old sediments" used to construct Figure 4-28a is not
factual because there are 44 samples from the 1985-1992 period and 68 from the 1954-1984 period.
Approximately 40% of the samples shown are from the 1985-1992 period.
The discussion of concentration effects on dechlorination center on the extent of dechlorination not
the type of dechlorination. As discussed in the response to comment DG-1.20, the indices of
dechlorination developed by GE/HydroQual are much less sensitive than either the MDPR or AMW.
See also the response to DG-1,20Q.
Response to DG-1.20
The MDPR is used as an estimate of the extent of dechlorination found in the Hudson River
sediment samples contaminated with PCBs. As discussed in Section 4.3.2. this ratio is strongly
December 22. !<»»
DE1R-63
TAMS/UTTetraTech/MCA
-------
~
1985-1992
~
1975-1984
El
1965-1974
¦
1954-1964
50"
40 ~
30 ~
-0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25
Change in Molecular Weight Relative to Aroclor 1242
Notes:
Data represented is the same as in Figure 4-28a of the DEIR. except that samples which are 4 cm
in length are represented twice and samples which are 2 cm in length are counted once.
This step elimates the overweighting of the more finely sliced surface samples.
Hudson River Database Release 4 I TAMS/LTL/TetraTech/MC
Figure DG-1.26R
Histogram of the Change in Molecular Weight as a Function of Time of Deposition
t •-» D I C rt/J < tUa I—11 i/^or\n Dm'AP
-------
correlated with the total PCBs mass detected in the mainstem Hudson samples (R2 = 0.75). This
demonstrates the strong relationship between PCB concentration and the degree of dechlorination.
The value of 30,000 ppb total PCBs is the intersection between MDPR equal to the initial
MDPR of Aroclor 1242 (0.14) and the lower 95% confidence limit of the data. For samples with
total PCB mass less than 30,000 ppb the occurrence of dechlorination from the initial Aroclor 1242
mixture is not predictable. Some samples have MDPRs greater than 0.14 and some have MDPRS
less than 0.14. This does not imply that dechlorination has not taken place, merely that the
occurrence of dechlorination is not predictable using this measure.
It should also be noted that for each of the congener ratios shown in Figure DG-1.20A, the
sediments typically fall well below the ratios seen in the Rogers Island water column station,
suggesting that much of the alteration occurs prior to deposition or just after deposition. That is the
ratios are reduced during water column transport or very soon after deposition. Additionally, this
phenomenon is shown in Figure DG-1.17, which shows these same ratios to decline in surface
sediment as a function of river mile. This shows that the further the PCBs have been transported, the
greater the change in ratio. Again this suggests that much of this alteration takes place during
transport, most likely in aerobic conditions, representing a degradation process which is unrelated
to anaerobic dechlorination.
There are several possible explanations for the low level MDPR values found in samples with
total PCB mass less than 30.000 ppb of which the majority of the samples are from the Lower
Hudson. From the DEIR. page 4-62. "The Lower Hudson sample MDPRs tend to cluster just below
the Aroclor 1242 value of 0.14. The mean MDPR for the Lower Hudson is 0.11. suggesting the
presence of a minor contribution by heavier Aroclors or, more likely, possible loss of BZ#1, 4, 8,
10, and 19 prior to deposition due to their generally greater solubility and degradability. The
congener pattern comparisons made in Chapter 3 (Subsection 3.3.3), suggest that both processes
probably occur to some degree."
The GE/HydroQual dechlorination ratios do show that the majority of the Hudson River
sediment samples have undergone some degree of dechlorination, although in most instances there
has been little reduction in mass or the number of chlorines per biphenyl. These ratios are shown in
Figures 1.20A and 1.20B. The majority of Upper and Lower Hudson samples fall below the values
of the ratios for unaltered Aroclor 1242. However, the correlations for these relations range from
0.0557 to 0.465, far weaker than the relationship between the MDPR and total PCB mass. There is
much less ability to predict the degree of dechlorination at a given concentration due to this scatter.
For the BZ#74:49 ratio applied to the Upper Hudson samples for a concentration of 10,000 ppb, the
dechlorination ratio ranges from 0.1 to 0.9. Samples with BZ#74:49 ratios greater than 0.65 (the
unaltered Aroclor 1242 BZ#74:49 value) do not appear to have undergone dechlorination. These
ratios serve to gain further insights into the nature and extent of dechlorination, but are not free of
limitations.
The fact that these ratios change within the sediment does not necessarily imply that the
sediment was subject to substantive levels of dechlorination (i.e.. a large decline in the number of
chlorines per biphenyl). The diagrams shown in Figures DG-1.20A and B demonstrate the weak
correspondence between the dechlorination ratios proposed by the writer and the PCB mass of the
sediments. This is in sharp contrast to the relationship between MDPR and Cl/BP vs. total PCB mass
December IV98
DEIR-65
TAMS/LTl/TetraTech/MCA
-------
Upper Hudson Samples
Upper Hudson Samples
cc
c
o
&
fS
3
?
¦S
z
1.2
I
08
06
04
02
0
. = i :4.o lining*) R"=n:i:
-€— 74:49
'
%°c
. „ <S
c "
c o o d"
. ~c
100 1000 10 10
Total PCBs (ppb)
10
10
V = I IS - 0 :i3lngm R = 02M
o
3
o£
a
"j?
3
?
-a
x
66:49
Total PCBs (ppb)
Upper Hudson Samples
60:49
^ '*jAr
100 1000 10 10 10
Total PCBs (ppb)
10
Upper Hudson Samples
y = 0 756 ¦ 0 I lyl.mixi K'= 0 146
0.8
0.7
06
05
04
0 3
0.2
0 1
0
:>.c c<
X
56:49
.0'$f
"QsSq
100 1000 I04 10" 10" I07
Total PCBs (ppb)
Upper Hudson River
Sediment Samples
Aroclor 1242
Rogers Island Suspended Fraction
Rogers Island Whole Water
Hudson River Database Release 4.1 TAMS/LTl/Tetr«Tech/M(
Figure DG-1.20A
Total PCBs vs. the GE/HydroQual Dechlorination Ratios
for the High Resolution Core Data (Upper Hudson)
-------
Lower Hudson Freshwater Samples
at
c
o
. -1. i: ¦ o k =ow
1.2
I
08
06
04
02
0
74:49
o
"
¦
&
100 1000 10 10 10
Total PCBs (ppb)
10
Lower Hudson Freshwater Samples
y = I (i ¦ 0 3WIO|!(*) R" - »¦>!''
u
c
C
Q
3
C
66:49
100 1000 10 10 10
Total PCBs (ppb)
10
Lower Hudson Freshwater Samples
\ = O 24S R:= II 417
60:49
05
04
0 }
~ 0 2
0 1
0
100 1000 10 10"
Total PCBs (ppb)
10
10'
Lower Hudson Freshwater Samples
>=0itfiA-01?:i»;i>l R; = I) 4>X
08
0.7
0.6
05
04
0 1
0 2
0.1
0
-e— 56 49
100 1000 10 10
Total PCBs (ppb)
10°
10
- Upper Hudson River
Sediment Samples
* Aroclor 1242
Rogers Island Suspended Fraction
— — Rogers Island Whole Water
Hudson River Database Release 4 I TAMS/LTl/TetraTech
Figure DG-1.20B
Total PCBs vs. the GE/HydroQual Dechlorination Ratios
for the High Resolution Core Data (Lower Hudson Freshwater)
-------
as shown in Figure DG-1.20C. In this figure, both MDPR and Cl/BP correlate strongly with total
PCB mass (R: = 0.635 and 0.597. respectively). Note that this figure uses only the Upper Hudson
sediment results. If Lower Hudson sediment results are included (as in Figure 4-22 of the DEIR),
then the R: for MDPR vs. total PCB mass improves to 0.75, indicating that the Lower Hudson
sediments belong in this analysis. This information, coupled with the congener pattern analysis and
the PCB/l37Cs analysis, indicates the dominance of freshwater PCB contamination by the GE
facilities and therefore supports their analysis as a single data set in this context. The USEPA rejects
the premise posed by the writer that the Lower Hudson samples do not belong in this analysis.
The H, H' ratios of the Lower Hudson sediments show a weak decline with total PCB mass.
However, this decline does not correspond to a substantive change in the number of chlorines per
biphenyl, as shown in Figure DG-1.20D. This is also evident in Figure DG-1.19A which shows only
a weak correlation between Cl/BP and these ratios for the Upper Hudson
Ultimately, it appears that while these ratios weakly correlate with the degree of
dechlorination, they are not useful in predicting the overall level of modification to the sample and
instead simply track the changes in the congeners themselves (i.e., BZ#56,60,66 and 74). They may
offer some insights to some limited processes perhaps involving dechlorination but they do not
reflect the overall alteration of the mixture. The fact that the ratio changes occur at low
concentrations is indicative of the instability of these particular congeners. However, the
demonstrated presence of Aroclor 1242-like mixtures in sediments 30 to 40 years old shows that the
dechlorination process cannot be relied on to substantively reduce the PCB mass or toxicity of low
level sediment contamination.
4.4 Implication of the PCB Fate in the Sediments for Water Column Transport
Response to DS-2.21
The USEPA acknowledges the importance of the exchange between the water column and
the sediments throughout the river, not just in the TIP.
Response to DS-2.22
As part of the modeling analysis, congener patterns of the various potential PCB sources to
the TI Pool water column will be compared with the patterns measured within the water column.
This analysis should provide further clarification as to the nature of the sources responsible for the
load. It is unclear what sort of study could precisely define the PCB sources and release mechanisms
short of a very extensive, open-ended study where many water column and sediment samples would
be required over long periods of time.
Response to DS-2.23
The degree of dechlorination required to match the water column pattern is fairly high.
Recently deposited PCBs would not be expected to attain the same degree of dechlorination as older
sediments due to their generally lower concentrations. Thus some contribution by older sediments
December 22 199S
DEIR-68
T AMS/LTl/TctraTcch/MC A
-------
y = -0.597 + 0.226Iog(x) R~= 0.635
c 06
.2 0-
« Q
•S S
i->
o o
jz -a
o co
M OS
Q ~
u- o
A 3
11
^ ft-
TV' ' ¦ 1
i_UJ
10 10J
Total PCBs (ppb)
y = 4.76 - 0.468log(x) R'= 0.597
0 —i—' i 111hi
a.
§
U
1
2
3
4
--op
i
10'
10J 10 10*
Total PCBs (ppb)
10°
10'
~ Upper Hudson River
Sediment Samples
-Aroclor 1242
Rogers Island Suspended Fraction
Rogers Island Whole Water
Hudson River Database Release 4 1 TAMS/LTI/TctraTcch/"
Figure DG-1.20C 0152
Relationships Between the Number of Chlorines per Biphenyl, 10.0
Mnlflr nprKlnrtnntinn PrnHnrt Ratin *at\A Tntal PPRc fnr tliA W!nK Pacnlutmn C*r\rt* Hftta
-------
Lower Hudson Freshwaler Samples
Lower Hudson Freshwaler Samples
--— 74:49
y = n i h.i . n 1*1 R=nis
I
CI/BP
Lower Hudson Freshwaler Samples
a
3
?
-o
X
. = .DI1.W • DIIJVKt R' = IMISM6
-C— M)49
Cl/BP
Oi
c
o
8
a
?
¦5
z
e— 66:49
v = U H2t> •IHISi R'= I) >1*
ri/BP
Lower Hudson Freshwaler Samples
> = (i u-i ~ o i x i x k" = (i:iw
56:49
4 .1
CI/BP
• Rogers Island Water Column
(Suspended Matter I
~ Upper Hudson River
Sediment Samples
Hudson River Database Release 4.1
TAMS/LTl/TetraTech/MC
Figure DG-1.20D
The Number of Chlorines per Biphenyl vs. the GE/HydroQual Dechlorination Ratios
for the High Resolution Core Data (Lower Hudson)
-------
would be needed to replicate the congener pattern seen in the water column. Rapid dechlorination
of recently deposited PCBs would serve only to decrease the degree of dechlorination required for
the older sediments contribution.
Response to DS-2.24
The intent of this statement was to confer the notion that simply because the upstream loads
have been greatly reduced does not mean that the sediment loads will now quickly dissipate. USEPA
agrees with the writer in that the sediment-derived loads are unlikely to disappear for a long period
of time without intervention.
Response to DS-2.25
USEPA agrees that the exchange mechanisms operating within the TI Pool should also occur
downstream.
Response to DS-2.26
Although the remedial measures at the GE facilities have not yet been completed, it is still
clearly evident in the water column monitoring data collected by USEPA and GE that the PCB
releases from the GE facilities have been greatly reduced since 1991. Data collected subsequent to
the completion shows still further reduction in the loads originating around the GE facilities, thus
leaving the TI Pool sediments as the major source to the water column. USEPA acknowledges the
fact that the efforts toward remediation at the GE facilities are not final and that the possibility of
further releases from the GE facilities remains.
Response to DG-1.5
The contention that the sediments cannot provide a source for a diffusive flux of PCBs to the
water column is incorrectly based on an assumption of continuous deposition throughout the TI Pool.
The TI Pool represents a dynamic system characterized by periodic sediment resuspension and
settling. Thus the surficial sediments of the pool at any given moment in time represent an
integration of all the PCBs deposited there with subsequent reworking by various hydrological and
biological processes. These processes effectively serve to renew the active sediment layer so that
sediment-to-water fluxes are not limited by burial. Deeper, more contaminated sediments can be
brought to the surface and redistributed throughout the pool by sediment resuspension and settling.
Evidence for this kind of reworking was demonstrated by the Low Resolution Sediment Coring
Program which showed extensive loss of PCBs from the sediments at many locations throughout the
Upper Hudson. Sediment mass losses based on this investigation are commensurate with mass loads
based on the water column monitoring data obtained by USEPA, USGS and GE.
The writer incorrectly uses the high resolution cores as a basis to infer the fate of sediment
PCBs with respect to resuspension or scouring. High Resolution core locations for which the
sediment deposition can be determined have not had any significant resuspension or scour. These
locations are difficult to find, as evidenced by the number of high resolution cores which did not
provide date-able sequences. Thus, these locations do not provide a basis to assess the extent of
sediment resuspension. Because the writer's mass balance is based on the incorrect assumption of
December 1998
DE1R-7I
T AMS/LTLT etraT ech/MC A
-------
consistent burial throughout the TI Pool, citing the relatively unique high resolution cores as support,
the USEPA rejects the mass balance calculation presented in the comment.
Response to DG-1.6
USEPA does not accept the premise that the resuspension of sediments containing
dechlorinated PCBs is implausible. The analysis of the homologue pattern of the water column and
sediments suggests a strong similarity. As support for its position, GE contends that the congener
patterns of the water column PCBs and dechlorinated sediment PCBs do not match. A similar
analysis conducted by USEPA is found in Book 3 of 3. However, it should be noted even if the
congener patterns do not match exactly, this does not rule out this mechanism or source since, as
noted in the DEIR, the water column load may be the combination of several PCB release processes,
so that the resulting pattern is the integration of the various processes. In addition, the congener
pattern of the water column load is itself variable, suggesting multiple sources and/or mechanisms.
As mentioned in response DG-1.5, the sediment inventories of PCBs are dynamically renewed by
the processes which affect the river bottom. For this reason, the writer's contention that the reservoir
of PCBs for this release process is inherently limited is incorrect. In fact, as described in the Low
Resolution Sediment Coring Report (USEPA, 1998), the sediment PCB inventories have seen a
substantial reduction that would be consistent with a substantial release to the water column as well
as a resuspension mechanism.
The analysis presented by the writer concerning the areas of the TI Pool with PCB
concentrations greater than 100 mg/kg is based on the surface sediment kriging analysis presented
in the DEIR. However, the samples represented in this analysis constitute roughly 12 inch intervals
whose mean concentration is 100 mg/kg. Thus, the selection presented by the writer does not find
all areas with sediments greater than 100 mg/kg. Indeed, it likely that sediments greater than 100
mg/kg can be found throughout much of the TI Pool, given that the peak concentration in some cores
from the TI Pool is over 1,500 mg/kg. On this basis, the depth-of-scour analysis presented by the
writer does not apply since the areas selected by the writer are unlikely to be the only areas
responsible for the TI Pool load if resuspension is the main release mechanism. (Note that the
USEPA expects its modeling analysis, which is currently underway, to better resolve the actual
mechanism(s) responsible for PCB release from the sediments.) It should be noted as well that the
low resolution coring program found most of the sediment PCB inventory to occur in the top 9
inches (23 cm). Scouring to this depth in the sediments would be largely undetectable given the
current and historical bathymetric data. Thus the argument that the scour necessary to yield the
measured loads would generate measurable changes in bathymetry is incorrect. Higher PCB
concentrations are more pervasive and the depth of contamination is typically shallower than
suggested by the writer.
Response to DG-1.7
The writer's concern that the degree of resuspension and settling of sediments in the TI Pool
may not be sufficient to yield the water column loads emanating from the Pool has also concerned
the USEPA. Currently the USEPA is investigating the scale of these fluxes with its own modeling
efforts. However, there are two issues concerning this flux. First, as indicated in the report, it may
be that resuspension is responsible for only a portion of the TI Pool load; thus, the rate of sediment
resuspension and settling may not be limiting. Second, other processes unrelated to river flow, such
December 22. 1998
DEIR-72
TAMS/LTl/TetraTcch/MCA
-------
as biological activity and recreational boat use. may be responsible for sediment resuspension. The
seasonal variability of the last three to four years of monitoring data collected by GE is in fact
strongly indicative of the absence of a flow dependence in the TI Pool's PCB loads. The absence of
a flow dependence would suggest that resuspension resulting from flow is unlikely to be the cause
of the PCB loading from the Tl Pool. Nonetheless, this does not rule out resuspension via biological
activity and recreational boat use.
Response to DG-1.8
The writer contends that the congener pattern of the water column load at the TI Dam is very
different from that of dechlorinated sediments as characterized by samples greater than 100 mg/kg
in the high resolution coring data set. The analysis considers "averaged" PCB samples from the high
resolution core data set, although the method of averaging is not explained. Using the "average"
congener pattern from these cores, the writer then presents a series of graphs comparing the percent
mass (mass fraction x 100) of the six transects to the "average" dechlorinated sediment. There are
several concerns in this presentation which undermine its conclusions. First, the mass percent
comparisons are made on log scale plots, thus the degree of disagreement among the congener pairs
is partially masked. In addition, undue emphasis is placed on the congeners present at low to trace
levels, where presence and quantitation are least certain. In the comparisons themselves, the data
as presented show a better regression to the surface sediments then to the dechlorinated sediments.
However, as mentioned above, it is unclear how the "average" congener patterns were obtained. It
is also uncertain whether the dechlorination pattern obtained from the average of the high resolution
cores would be the same as that of the average dechlorinated sediment source. Accepting the
presentation at face value, it is also clear that the surface sediments fail to predict the correct mass
percent for the two most prevalent congeners and consistently underestimate their percentages by
a large margin. Thus, although the surface sediments provide better regressions, they cannot
represent the most massive congeners in the mixture. In light of this evidence, the more
dechlorinated sediments cannot be ruled out as a major source at this time.
4.5 Summary and Conclusions
Response to DC-3.1
As stated in the DEIR. the data gathered for the Phase 2 Hudson River PCBs RI/FS (high
resolution cores and water column samples) are not sufficient to determine the exact mechanisms
for sediment to water column transfer, but are sufficient to both suggest and rule out mechanisms,
qualitatively. Water column data collected by USEPA and General Electric at Rogers Island are a
monitor of the affect of the Allen Mill input in two ways. First, PCB concentrations in the water
column have dropped to undetectable levels at the Rogers Island station since remedial actions were
taken in 1993 and after peaking in 1991. Second, the unaltered contribution is easily distinguished
from altered (dechlorinated) contribution by the absence of high levels of mono- and di-chlorinated
congeners. The amount of PCBs released above Rogers Island has been quantified using General
Electric s Water Column monitoring data which began in 1990 and continues to date. See Section
3.4.2 of the DEIR. Loadings Upstream of the TI Pool.
December 22 1*58
DEIR-73
TAMS/LTlTetra Tech/MCA
-------
Response to DC-3.2
Groundwater flux is a well defined phenomenon. A thorough treatment of the topic can be
found in the following reference: Freeze, R. Allan, and John A, Cherry. Groundwater. Prentice-Hall
Inc., 1979.
Diffusion of PCBs from the sediment to the water column via porewater is a mechanism
suggested by comparing the water column homologue patterns with the porewater homologue
patterns derived from the sediment concentrations. These patterns closely match in some, but not all,
instances.
Respwse to PO-1.20I
From page 4-70 of the DEIR:
The absence of a correlation with time can be seen more clearly in a subset of the samples
represented above. Using only dated sediment cores from the freshwater main stem Hudson,
the degree of dechlorination as measured by the AMW is plotted in a histogram as a
function of time of deposition (see Figure 4-28a). The data are grouped into approximately
ten year intervals beginning in 1954. Evident in the diagram is the clustering of the values
around a A MW value of 0. More important for this discussion, however, is the fact that a
wide range of mass loss (i.e., AMW) can be seen for any time period, indicating the lack of
time dependence for this process. The samples with the greatest degree of dechlorination are
from the period 1965 to 1974. However, this period also contains samples with essentially
no measurable dechlorination. Similar ranges can be seen for the periods 1955 to 1964 and
1975 to 1984. Only the most recent period presented, 1985 to 1992, has a lesser range.
However, its results are still consistent with those of the previous time periods.
The conclusion that the degree of in situ PCB dechlorination is dependent on total PCB
concentration and not dependent on time is based on the comparison of dated core samples and the
AMW. This relationship is apparent from the field results, not the tenets of dechlorination
biochemistry. It is not disputed that dechlorination reactions are time dependent, but as discussed
in the comment, the time scale is on the order of months, since many of these samples have been in
situ for more than 20 years and still do not show extensive dechlorination, it is clear that time is not
a major component in determining the ultimate degree of dechlorination. See also the responses to
comments DG-1.19 and DG-1.20.
Response to PL-1.4
Due to the remedial actions performed by General Electric in the Bakers Falls area, the
concentration of PCBs has dropped to undetectable levels at the Rogers Island station. Still, PCBs
are detected at the TI Dam. This indicates that the TI Pool sediments are a source of PCBs to the
water column. The presence of oil droplets in the TI Pool is unproven (see the response to comment
DL-1,1), but even so, the levels of mono- and di-chlorobiphenvls in the water column could not be
generated by the partitioning of an unaltered Aroclor 1242 mixture. (See Sections 4.3 and 4.4 of the
DEIR.) The concentrations of lighter congeners found in the water column can be generated by
partitioning of a dechlorinated Aroclor mixture such as is found in the sediments of the TI Pool.
December 22. 1998
DE1R-74
TAMS/LTl/TetraTech/MCA
-------
Further, substantial mass has been lost from the hot spots of the Tl Pool between 1984 and 1994
(USEPA, 1998), Thus, the sediment of the TI Pool are not acting as a sink for PCBs, but rather as
a source of PCBs to the water column. The USEPA does not contend that sediments below the TI
Dam do not contribute PCBs to the water column. Rather, it is apparent from the data that the
contributions below the dam are substantially less than those from the TI Pool. This is particularly
evident for the area of the Upper Hudson below Schuylerville.
Response to DL-1.6
PCBs are currently detectable in water column samples taken from the Thompson Island
Dam stations (O'Brien & Gere, 1998) while samples from Rogers Island have dropped to
undetectable levels. Additionally, peak summer time concentrations have shown no sign of
decreasing, despite major reductions in the loads at Rogers Island since 1993. Thus the source of the
contamination must originate in the TI Pool. Examination of co-located sample pairs from 1984 and
1994 indicates loss of PCB mass in fine-grained TI Pool sediments over these ten years. At most
only 11% can be accounted for by dechlorination (USEPA, 1998). According to this evidence, the
source of PCBs to the water column at the TI Dam is the PCBs resident in the sediments of the TI
Pool. Various remediation strategies, including "No Action" and Natural Attenuation, as well as
dredging and capping of affected areas, and the accompanying ramifications, will be explored in the
Feasibility Study.
References
No significant comments were received on References.
Volume 2C (Book 2 of 3)
Figures, Tables, and Plates
Comments mentioning figures and tables were addressed in the Book I text section in which they
were referenced.
Volume 2C (Book 3 of 3)
Appendix A: DATA USABILITY REPORT FOR PCB CONGENERS HIGH RESOLUTION
SEDIMENT CORING STUDY
Response to DF-2.2A
Data quality issues relating to total PCB and homologue sums for the data used in the DEIR
were treated as follows:
A. In the creation of total PCB and homologue sums, non-detect and rejected congener
results were treated as null (or zero) values, i.e, they did not contribute to the sum in any
way. In this manner, each sum represents only what was measured and is therefore a
minimum estimate of the true PCB mass in the sample since presumably some congeners are
present at below the detection limit. This summation technique is different from that usually
employed to deal with nondetect values wherein nondetects are assigned a value at one half
December 22. 1998
DEIR-75
TAMS/LTI/TetraTech/MCA
-------
the detection limit. In general, this approach is believed to yield little difference from the true
value since most samples were quantitated over two orders of magnitude and therefore the
PCB mass present below the detection limit would represent less than a few percent of the
total mass of the sample.
B. In the examination of congener patterns and congener ratios, the interpretation followed
the same convention as described immediately above in A. Thus, nondetect values appear
as zero values on the diagrams showing ratios with respect to BZ#52. In this fashion,
matching points in a comparison of samples represent measured values only. Following this
convention, it is clear that all samples represented in the congener pattern plots had
measurable levels of BZ#52.
C. No sample was excluded from analysis based solely on the occurrence of non-detect
results for individual congeners. In most instances, non-detects were constrained to minor
components of the PCB levels found in most samples and therefore did not greatly affect the
homologue patterns seen. The individual non-detect congener results were examined only
when a sample deviated from an expected geochemical trend or congener pattern. In these
instances, a sample might be excluded from the analysis if the non-detect issues warranted
it. Several samples in Transect 2 in particular suggested this kind of concern.
Overall this treatment minimized the impact of non-detect results on the analyses presented.
Total PCB and homologue sums were not greatly impacted unless a major congener was affected.
These instances were typically obvious when the data were examined in a geochemical context.
Individual congeners were only compared to other measured values, thus avoiding the issue of a
possible but unconfirmed match between a measured level and a high detection limit non-detect
result. The following discussion provides a detailed discussion of the effect of the converted results.
A summary of the detected results negated (i.e., changed to undetected or "U" values)
because of blank contamination is provided in the Appendices A and B of the DEIR. Table A-6
summarizes the instances for the High Resolution Sediment Cores; Tables B-3 and B-4 provide a
similar summary for the Water Column Particulate Data and Dissolved PCBs, respectively. Based
on the reviewer's selection of congeners, the concern was focussed on the dissolved data (review of
Table B-4 shows that 22 to 62 percent of all sample results for the four cited congeners were negated
to non-detects based on blank contamination).
In order to assess, and enable others to assess, the effect and relative significance of the
negated congeners in the context of the total PCBs detected in any particular sample, USEPA has
prepared three tables, one for each matrix (Table DF-2.2A - Water Column-Dissolved PCBs; Table
DF-2.2B - Water Column-Particulate Data; and Table DF-2.2C - High Resolution Coring Study -
Sediment Core Sample Data). These tables list each sample for the respective matrix in which one
or more of 15 different congeners was negated (the 15 congeners consist of the 12 "principal"
congeners along with three other congeners - BZ #44, 77, and 153 - which were explicitly mentioned
in this or other review comments); the congener-specific negated concentration; the total (valid) PCB
concentration in the sample (column header "Total PCBs"); the total (sum) negated concentration
of the 15 congeners (header "Sum of Negated Values"); and the fraction of these negated values
relative to the total valid concentration (header "Fraction"; consisting of "Sum of Negated Values"
divided by "Total PCBs", expressed in percent).
December 22. 1991
DEIR-76
TAMS/LTl/Tetn Tech/MCA
-------
Table Dl( .A
Water Column Study - Dissolved PCBs
Negated Values (Changed to Undeteet Because of Plank Contamination)
Units: ng/L
mple II)
IY/.H 1
HZ#4 UZ»8 HZ# 10 »Z#I8
Li/.« 1 'J HZ#28
BZS44 BZ#52
BZ.#77
HZ 101/
90
n/ti 118
HZ# 138
HZ#153
BZ#180
Total PCBs
Sum of
Negated
Values
Iraction
V-409-0005
0 022
205 407
0.022
0%
V-005-0005
0.198
0.072
226.1 13
0 270
0%
V-309-0005
0.155
0.058
176.659
0.213
0%
V-3 09-0004
0 100
0.034
91.721
0.134
0%
V-209-0008
0.105
0.059
86.515
0.164
0%
V-004-0006
0 159
0.020
93.602
0 179
0%
V-005-0008M
0.051
0.067
0.018
68.995
0.136
0%
V-209-0005M
0 101
0.108
0 081
0029
156.769
0 318
0%
V-005-0004
0 194
0.166
0.099
198.438
0.459
0%
V-006-0005
0.132
0.154
0.018
107 524
0 304
0%
V-001-0006
0056
0043
0.007
36.349
0.105
0%
V-004-0007
0.089
0.079
0.032
65.365
0200
0%
V-003-0006
0 038
0 094
0.052
0.033
0.008
62.826
0.225
0%
V-004-0004
0 084
0.048
0.037
0.012
48.061
0.180
0%
V-309-0008M
0 172
0.105
0.037
77.349
0 314
0%
V-003-0005
0 050
0 314
0.063
0.014
98.444
0.441
0%
V-001-0005
0.102
0 075
0.040
0.006
46.565
0.223
0%
V-003-0008
0.080
15 114
0.080
1%
V-006-0008
0.153
0.089
0.019
48.523
0.261
1%
(M09-0004M
0.098
0.209
0.021
53.653
0.328
1%
V-109-0005
0.077
0.188
0.086
0.060
0.021
65.885
0.432
1%
V-003-0007M
0.028
0.082
0.060
0.046
0.010
33.346
0.226
1%
/-109-0008A1
0.116
0245
0.128
0.104
0.042
75.235
0.635
1%
V-l 09-0008A2
0.097
0.209
0.106
0.086
0.026
53.938
0 525
1%
V-008-0005
0.500
0.220
0.110
0.084
0.040
84.991
0954
1%
V-004-00I5M
0.022
0248
0 071
25.285
0 341
1%
V-004-0012
0027
1.963
0.027
1%
/-409-0008
0 656
0.172
0.094
0.008
68.056
0 929
1%
V-004-0008
0.106
0.497
0.236
0.099
0.077
0.019
74.987
1.034
1%
V-004-0014
0.065
0.379
0.163
0.119
0.065
0.018
53.466
0.809
2%
V-008-0004
0 172
0.733
0.317
0.150
0 111
0.043
99.280
1.526
2%
V-006-0014
0.216
0 054
0.114
0.093
0.024
28.920
0.500
2%
V-001-0003
2.670
1.950
0.141
252.231
4.761
2%
V-006-0004
0.065
0.320
0.085
0.071
0.012
27.694
0.553
2%
V-006-0015
0.287
0.266
0.096
0083
0.075
0.020
27.706
0.826
3%
V-001-0008
0.833
0.072
0.079
0.046
0.030
31.159
1.059
3%
V-006-0017
0.137
0.271
0.085
0.096
0088
0.020
19.347
0.697
4%
V-409-0004M
0.849
0.093
0.107
0.059
0.053
29.145
1.162
4%
12/22/98
10.0160
DF.IR- 77
TAMS/l.Tl/TcIra Iceh/MCA
-------
10.0161
Table DF-2.2A
Water Column Study - Dissolved PCBs
Negated Values (Changed to Undetect Because ol' Blank Contamination)
unple ID
BZ#l
BZ#4
B/.# 8
BZ#10
BZ# 18
BZ# 19
H/.#28
BZ#44
B/J52
BZS77
BZiOl /
90
ll/tt 118
BZ#I38
BZS153
BZ#180
Total PCBs
Sum of
Negated
Values
Fraction
tV-609-0008M
0.747
0 158
0466
0 188
0 118
0.094
0,027
39 889
1.797
5%
vV-008-0008
2 320
0 391
0.196
0 124
0.097
0.064
46 617
3 192
7%
#-609-0004
0.688
0.040
0,200
0 100
0064
0.046
0.019
16.575
1 156
7%
W-004-0011
0.082
0 241
0 121
0.112
0.039
7.163
0594
8%
W-003-0003
1.000
0.071
0033
0 139
0,036
0.021
0 005
14,960
1.305
9%
W-002-0004
0.297
0032
3.724
0.329
9%
W-003-0004
1.240
0.099
0.039
0.133
0 029
0.017
15.545
1.556
10%
W-001-0004
0.050
0.247
0.307
0.201
0 051
0.018
7.024
0.874
12%
W-004-0003
1.600
0.049
0 091
0.038
0.031
0.022
14.641
1.830
13%
W-002-0005M
1.540
1.690
2.250
1 480
0030
0.077
0.036
0.017
44.084
7.120
16%
W-005-0013
0.066
0 066
0,033
0 019
0.011
1 071
0.195
18%
W-002-0006
0.407
0.869
1.120
1.330
0.047
0.029
20.227
3.802
19%
W-004-0017
1.250
0.121
0.031
0.037
0.035
0.011
7.217
1.484
21%
W-005-0003
0.191
0.805
0.343
0.040
0.114
0.073
0.042
0.034
0.011
7.861
1 653
21%
W-003-0012
0.791
0.058
0.212
0.019
0.022
4.844
1.103
23%
W-001-0002
0.025
0.011
0 011
0.007
0.005
0.251
0.059
23%
W-006-0003M
0.293
0.573
0.577
0.573
0.049
0.197
0.081
0.040
0.028
10.109
2 411
24%
W-002-0007
1.020
0.825
1 140
1.310
0.040
0084
0.056
0.030
17.824
4.506
25%
W-005-0010
0.679
0.135
0.107
0.054
0.054
0 048
0.015
4 319
1.092
25%
W-004-0001
0.011
0.009
0,017
0.042
0.077
0 028
0.016
0.008
0.781
0.208
27%
W-006-001Q
0.264
0.264
0.271
0.087
0040
0.035
0,023
2.776
0,986
36%
W-005-0012
0.477
0.030
0.047
0.077
0.054
0.050
0.038
0.020
1.978
0.792
40%
W-002-0011
0.037
0.049
0.030
0.013
0.027
0.040
0.020
0.011
0,556
0,228
41%
W-001 -0011
0.104
0 044
0007
0.006
0.324
0,160
49%
W-003-0011
0,017
0.007
0.020
0.048
0024
0042
0,048
0.040
0.032
0 506
0.277
55%
W-005-0011
0.186
0.010
0.011
0.017
0,024
0.026
0.018
0023
0.020
0.014
0.549
0.350
64%
W-006-0012
0.092
0.092
0.117
0263
0.126
0.081
0.086
0.064
0.018
1.448
0.940
65%
W-002-0013
0.069
0.034
0.039
0 122
0.018
0.069
0.033
0.026
0.008
0.596
0 417
70%
W-003-0013
0.109
0.160
0 114
0,137
0.252
0 219
0.087
0.094
0.023
1.464
1.195
82%
W-004-0013
0.033
0.135
0.078
0.157
0.358
0.167
0.043
0.072
0.063
0.017
1 359
1.123
83%
A?-109-0002
0.013
0,012
0.021
0,021
0.047
0.104
0.087
0.033
0.027
0,010
0.446
0.376
84%
W-001-000!
0.020
0.044
0 021
0.012
0.093
0.011
0.224
0.199
89%
W-002-0001
0.103
0,036
0.007
0.008
0,019
0 005
0.013
0.210
0,191
91%
W-003-0001
0.550
0.007
0.006
0.027
0.085
0.013
0,032
0.020
0.010
0.812
0.751
92%
W-001-0013
0.024
0.077
0.033
0038
0 071
0.029
0.039
0.019
0.355
0 330
93%
W-001-0012
0.186
0.343
0.259
0.126
0.126
0.049
0.029
0.027
1.182
1.144
97%
SV-001 -0016
4.490
0.455
0.857
0.870
0978
0 831
0.044
0.040
0.007
8623
8.573
99%
W-006-0013
0.063
0.133
0.096
0.028
0.047
0.039
0.019
0.413
0.425
103%
Units: ng/L
12/2
Dl-
-------
Table d( 2A
Water Column Study - Dissolved PCBs
Negated Values (Changed to IJndeteet Because of Blank Contamination)
Units: ng/1.
implc 11)
W/.U 1
BZ#4
BZ#8
BZ# 10
BZ# 18
BZ" 19
B/.#28
BZ#44
BZ#52
W/M77
X/. 101/
90 BZ# 118 BZ#I38 BZ#I53 BZ#I80
Total I'CBs
Sum of
Negated
Values
Traction
W-005-0002
0 173
0.096
0033
0.047
0.024
0.030
0 015
0.012
0.012
0.009
0.007
0 420
0.458
109%
W-()()2-0()03
0 174
1 060
0.326
0.527
0 496
0 020
0 047
0032
0.016
2.377
2.699
1 14%
W-006-0001
0 014
0.012
0.017
0 011
0.011
0.017
0.073
0.083
1 14%
W-004-0002
0 010
0.008
0 019
0016
0.029
0.024
0.023
0.015
0.019
0.015
0.011
0 163
0 188
1 1 5%
W-002-0012
1.050
0.265
0203
0 399
0 168
0 171
0 014
0 068
0.038
0.041
0.008
1.979
2 426
123%
W-005-0001
0.094
0 021
0 037
0.019
0 021
0.018
0015
0.013
0.007
0 196
0.245
125%
W-409-0002
0 019
0.054
0.072
0 071
0.033
0.094
0 111
0.095
0.043
0.465
0.592
127%
W-3 09-0002
0.349
0.064
0 041
0.066
0 028
0.052
0.025
0.031
0.025
0.016
0.492
0.696
141%
W-003-0002
0 461
0.009
0.007
0.01 1
0 028
0.071
0 013
0.029
0.420
0 628
149%
W-002-0002
0 170
0.033
0.005
0.1 15
0.208
182%
W-209-0002
0 167
0 013
0.023
0.031
0.022
0.032
0020
0.030
0.020
0.021
0.015
0.016
0.195
0.409
210%
W-002-0008
2.910
2 620
0.816
0.924
1 560
1.050
1 620
0.620
0888
0035
0.153
0.085
0.046
0.028
0.007
4.451
13 363
300%
W-006-0002
0 031
0 042
0052
0.101
0034
0 053
0.053
0.036
0.116
0.401
345%
W-609-0002
0.036
0.089
0.134
0.040
0 046
0.040
0.047
0.084
0 431
516%
W-006-001 1
0.051
0.051
0.102
-
12/22/98
Ut-IR- 79
TAMS/1.11A1 clra l ech/MCA
-------
10.0163
Table DF-2.2B
Water Column Study - Particulate Data
Negated Values (Changed to Undetect Because of Blank Contamination)
Units: ng/kg
Sum of
BZIOI/
Negated
: IL)
BZ» 1
BZ#4
BZ#8
li'/M 10 BZ# 18 H/#I9 BZ#28 BZ»44 BZ#52 BZ#77 90 BZ#II8
BZ#I38
BZ#I53
BZ#180
Total PCBs
Values
Traction
'-0005
4.97
5765.15
4.97
0°/
'-0004M
12.16
12198.10
12 16
0°/
'-0008A1
2.80
2037 70
2.80
0°/
1-0005
12.32
522840
12.32
0°/
1-0013
0 89
379.12
0 89
0°/
1-0008
10.29
2628.96
10.29
0°/.
-0004
150 70 110.78
29.19
16.57
5345623
307 24
I0/
1-0011
9.57
1449.71
9 57
1°/.
1-0004
41.78
48.69
5900.82
90 47
2°/
1-0017
4.48
2.76
44662
7.25
2%
1-0006
119 58
730025
119.58
2%
1-0014
62.02
3627 42
62.02
2°/
-0003
751.02
66.91 860 26
445.15
118720 65
2123.35
2°/
1-0008
85.18
4698.15
85.18
2°/
•-0008
71.19
3843.47
71 19
2°/
'-0008A2
61.70
11.34
3816.72
73.04
2°;
1-0004
223.25
139.61
17853.09
362.86
2°/
'-0008M
31.82
24.41
16.56
3159.98
72.79
2°/
i-0008M
47.49
II 41
235800
58.90
2°/,
1-0007
127.33
4709.38
127.33
1°/,
1-0013
0 45
1.14
53.50
1 59
3°/
1-0004
514.83
16676.60
514.83
3°/
1-0003
49.58
3.50
1605 18
53.08
3°/,
1-0012
1.06
1 58
78.76
2.64
3°/
1-0004M
149.04
4310.93
149.04
3°.
1-0005
275.20
6989.50
275.20
¦I0',
'-0005
293 35
67.23
36.18
8659.63
396 76
5°/i
I-0008M
114 54
47.43
27.51
3543.88
189 47
5°/,
1-0006
202.09 95.52
33.87
5853 73
331 48
6°/,
i-0010
4.27
3.08
0.89
138 51
8.23
6"/
1-0004
169.29 135.62
76.57
61.35
33.67
7609.70
476.50
6°/,
>-0017
16.53
59.33
16.53
1468.79
92.38
6°/,
1-0003
110.60 129.80 42.07
6.77
4562.59
289.24
6°/,
>-0002
1.65 2 44 5.06
1.48
166.23
10.62
6°/,
1-0005
169 17 72.25 88.99
19.12
5432.34
349.54
6"/,
1-0003
11162 94.94
48.54
24.16
4324.15
279.27
6°/
1-0001
0.50 1 07
2.37
58.04
3.95
7°/i
1-0002
4.53
65.44
4.53
TA
12/r
DKli
A^1S/l.T,rT><
rjC ^
-------
( Tabled 2B " (
Water Column Study - Paniculate Data
Negated Valu
us ((.'hanged to IJndeieet Heeause of Blank Contamination)
Units: ng/kg
Sum of
BZI01/
Negated
ID
HZ# 1
BZ.#4
HZ# 8
HZ# 10 l*Z# 18 HZ# 19
IJZ#28
BZ#44 HZ#52
HZ# 77
90 HZ# 118
HZ#138
HZ# 153
HZ#180
Total I'CBs
Values
l-'raetion
-0006
8 31
24 79
17 53
9.80
5.36
927 14
65 79
7'
-0007
55.72
82 33
29 91
15.03
2355.97
182.99
8'
-0003
31 26
16.95
36.86
29.71
14.42
1637 25
129.20
8'
-001 1
2.02
1.22
3.15
71 45
6.38
9'
-0006
152.63
203.93
73.29
99.59
22.64
6116.59
552.08
9'
-0016
16.65
18.50
18 16
13.92
8 64
820 15
75.88
91
-0008
71 34
94.07
71.34
2538.33
236.76
9'
-0015
38 26
71 20
38.26
24.86
1781 22
172 59
10'
-0005M
231 17
332 58
54.47
23.87
6484 95
642.09
10'
-0005
•153.75
177 14
44 50
68.64
31.12
7141 24
775.16
I 1'
-0010
10.13
3.70
1.17
137.20
14.99
1 P
-0002
59 49
135 39
53.33
17.59
2179.30
265 80
12'
-0005M
367.28
492.75
63.01
33.44
7530.30
956 48
13'
-0004
19 96
221 08
2504 31
18781.22
2745.35
15'
-0013
11 22
20 59
24.35
10.56
434.27
66.72
15'
-0002
32.37
210 66
32.37
15'
-0011
3.66
12.10
7.51
6.57
3.14
207.84
32 98
16'
-0005
360.15
73 10
22.94
42 59
14.08
3216.97
512.86
16
-0008
2 26
39.74
88.93
2 79
821.54
133.72
16'
-0001
13.11
9.52
134.99
22.63
17'
-0013
9.45
603
2.54
8 37
5.91
189.16
32 29
17'
-0011
3.18
47.73
156.27
4.43
1220.29
211 62
17'
-0013
14.61
37 41
8 94
24.63
486.33
85.60
18
-0012
6.28
6 19 8 14
293
9.35
4.01
208.48
36.90
18'
-0007 M
42.34
4 78
3 42
9.02
6.60
3.49
1.62
402.71
71.28
18
-0002
1.56
2.83
5.92
56.01
.10 31
18
-0003 M
96 19
77 86
42.47
112.84
144.90
2394.95
474.27
20
-0005
194 23
221.33
194.23
242 91
49 59
20.38
4393.62
922 66
21'
-0008
47.52
74.49
120 78
34.95
46.60
10.30
1575.77
334.63
21'
-0004
43.72
71.47
43.72
183.63
42.96
27.56
8.33
1852.12
421.39
23
-0002
20.38
18.73 35.81
21.23
35 33
19.05
24.41
20.87
21.19
913 75
217.01
24
-0011
0.89
1.13
2.40 4.29
2.37
4.33
2.52
2.16
1.85
91.10
21.94
24'
-0013
9.48
12.10
14.92
14.07
6.55
235.91
57.11
24'
1-0014
26.24
60.31
26.24
111.22
16.06
41.71
31.53
1241.21
313.31
25'
-0012
17.49
5.39
980
12.31
2.78
177.98
47.77
27'
-0001
10.57
11 59 25 32
12.55
18.92
9.15
13.64
10.84
12.33
45045
124.90
28
-0012
6.40
5.88
6.70
9.67
12.75
3.29
152.35
44.67
29'
-0001
5.04
53.69
7.73
5.14
212.65
71 60
34'
12/22/98
DlilR- 81
TAMS/l.TI/TetraTcch/MCA
-------
Table DF-2.2B
Water Column Study - Particulate Data
Negated Values (Changed to llndetect Because of Blank Contamination)
Units: ng/kg
Sum of
HZ 101/
Negated
: ID
BZ# 1
BZ#4 BZ#8
HZ# 10 BZ# 18 BZ# 19 BZ#28 BZ#44 BZ#52 BZ#77
90 BZ# 118
BZ#I38
BZ#153
BZ#180
Total I'CBs
Values
Fraction
5-0012
5 78
5 22
8 25
945
6.72
2.76
111.52
38 18
34°
)-0008
70 87
195 92 108.10 146.28 26.49
72 31
63.48
45.10
31.98
14.77
2040 01
775.31
38"
MKH)2
16 19
11.93
73 76
28 12
38°
3-0012
699
6 40
8.35
6.31
39.17
28.05
72°
S-OOOI
19 71
19.71 30.26
31.84
33.72
25.21
163.74
160 44
981
)-0002
11 49
5.43
1003
12 70
6.88
8.82
49.53
55.35
112°
3-0011
4.57
2.70
5 91
7.27
123°
1-0001
29.87 120.49
13.80
8.42
137.02
172.58
126°
5-(KK)2
68 87
390
8.88
4.96
7.48
66.08
94 08
142°
S-0002
40.57
40.57 15.44
3.22
9.24
8.49
2.76
120.30
12/22/98
DEIi
TAMS/l.TI/Tctnv' h/MCA
-------
Table D
High Resolution Coring Study - seaimeni core sample uaia
Negated Values (Changed to Undetect Because of Blank Contamination)
Units, ng/kg
Sum of
BZ.101/
Negated
niple ID
U/.SI BZ#4 BZ#8 BZ#I0 BZ# 18 BZ#I9 BZ#28
BZ#44 BZ#52 BZ#77 90 BZ# 1 8
BZ#138
BZ#I53
BZ#180
Total PCBs
Values
Fraction
*-005-44491'
2
0.262
7456
2
0%
*-020-20241'
175
548656
175
0%
*-023-40451'
77
170587
77
0%
*-018-l216P
1270
2483363
1270
0%
*-038-0002A
0.484
687
0.484
0%
i*-036-0002A
0.467
632
0 467
0%
*-018-0816G
1820
2350128
1820
0%
*-026-32401'
60
73202
60
0%
*-0l8-1620P
1370
1653208
1370
0%
*-026-04081'
143
158247
143
0%
<-019-0812P
105
75
197452
180
0%
*-019-0816(3
273
173
437362
446
0%
*-037-0002A
1
974
1
0%
*-026-0816G
132
112588
132
0%
*-026-0002P
120
99474
120
0%
*-022-4852P
628
515651
628
0%
*-039-0002A
1
1074
1
0%
*-023-4549P
0.318
0.519 1
1625
2
0%
*-002-36401'
9
5356
9
0%
*-002-44481'
6
3574
6
0%
*-030-0002A
2
1273
2
0%
1-020-I620P
166
88844
166
0%
1-020-0816G
329
657
462795
986
0%
1-026-1624 P
648 199
595
368
840055
1810
0%
1-026-2432P
318 119
282
330862
719
0%
(-019-0608P
109
67
69211
176
0%
*.-012-4852P
4
1685
4
0%
*-002-48511'
8
3122
8
0%
*.-019-16201'
2800
382
1130
559
1719639
4871
0%
*-002-28321'
5
1612
5
0%
*-0!9-2832P
170 601
380
397
500219
1548
0%
*-019-0002P
78
25128
78
0%
*-026-1216P
252 233 183
548
370
195
515745
1781
0%
*-003-4043P
5
1323
5
0%
*-009-1214P
8
2388
8
0%
*-026-0204 P
274
135
115690
409
0%
*-003-2024P
3
913
3
0%
*-031-0002A
4
1188
4
0%
I2/22/9X
DEIR- 83
TAMS/1 .TI/TctraTcch/Ml.'A
-------
Table DF-2.2C
High Resolution Coring Study - Sediment Core Sample Data
Negated Values (Changed to IJndeleei Because of Blank Contamination)
Units: ng/kg
Sum of
BZIOI /
Negated
ample ID
B/.#l BX#4 B/.#8 BZ# 10 BZ#I8 l)Z# 19 UZ#28 BZ#44 BZ#52 BZ#77 90 BZ/M18 BZ#I38 BZ#153 BZ#I80
Total PCBs
Values
Fraction
IR-023-2832P
6930
1830882
6930
0%
IR-003-2832P
4
1103
4
0%
IR-023-3236P
2830 440
838036
3270
0%
1R-003-08I6G
4
101 1
4
0%
IR-0I2-4448P
3
801
3
0%
IR-008-5660P
15
3624
15
0%
IR-007-0608P
7
1632
7
0%
IR-028-0608P
177 131 45
83601
353
0%
IK-007-0406P
6
1359
6
0%
IR-013-3236P
23
5202
23
0%
IR-007-0204P
5
1226
5
0%
IR-008-2024P
13
2906
13
0%
IR-001-1216P
5
1082
5
0%
IR-007-0002P
6
1146
6
0%
IR-035-0002A
21
4284
21
0%
1R-006-08I6G
7
1502
7
0%
1R-007-I2I6P
8
1626
8
1%
1R-008-2428P
25
4861
25
1%
IR-028-2024P
78 65 53 23
43134
220
1%
1R-00I-I620P
6
1265
6
1%
1R-025-0002P
33
6481
33
1%
1R-007-08I2P
7
1357
7
1%
1R-008-4448P
31
5957
31
1%
1R-025-0408P
32
6028
32
1%
IR-007-0816(i
9
1689
9
1%
1R-007-1620P
13
2377
13
1%
IR-028-0204P
405 459 290
212267
1154
1%
IR-003-12I6P
5
996
5
1%
1R-004-2428P
5
897
5
1%
1R-025-08I2P
26
4764
26
1%
W-007-2024P
29
5321
29
1%
1R-003-0812P
5
864
5
1%
1R-007-2428P
36
6497
36
1%
1R-003-0406P
6
1079
6
1%
IR-0I3-2428P
20 12
5732
32
1%
1R-008-08I6G
7 8
2681
15
1%
1R-028-08I2P
229 274 274 207 122
189840
1106
1%
1R-003-0204P
6
1045
6
1%
/
84 "rAMS/I.T,rr—f
-------
Table ~( 2C
High Resolution Coring Study - Sediment Core Sample Data
Negated Values (Changed to Undetecl Because ol" Blank Contamination)
Units: |ig/kg
Sum of
HZ 101/
Negated
linplc II)
BZ# 1 BZ.#4
bz#8
HZ# 10
HZ# 18 HZ# 19 BZJ28
HZ.#44 HZ#52 BZ#77 90 UZ#II8
BZ#138
BZ#153
BZ#180
Total PCBs
Values
Fraction
R-003-3236P
7
1223
7
\%
R-007-2832P
44
7292
44
1%
R-OI6-4044P
5
82
13783
87
1%
R-008-I620P
7
12
2919
18
1%
R-014-0812P
8
1337
8
1%
R-0I2-08I6G
i
167
1
1%
R-0I2-2024P
0964
153
0.964
1%
R-0I4-I216P
53
8357
53
1%
R-023-3640P
832
74
141060
906
1%
R-003-2428P
6
926
6
1%
R-028-0406P
110
78
49
27
39458
264
1%
R-0I0-I216P
9
1302
9
1%
R-003-3640P
8
1148
8
1%
R-013-28321'
70
14
12107
84
1%
R-010-0002P
2
5
958
7
1%
R-004-08I2P
7
3
1450
10
1%
R-0I4-0608P
10
1439
10
1%
R-003-I620P
6
832
6
1%
R-010-0812P
7
948
7
1%
R-0I0-0608P
9
1161
9
1%
R-006-3236P
52
6692
52
1%
R-0I7-2428P
1
3
531
4
1%
R-010-04061'
2
6
933
7
1%
R-004-0002P
1 1
1389
11
1%
R-006-2832P
17
54
8590
71
1%
R-003-0608P
7
888
7
1%
R-006-2428P
12
45
6918
57
1%
R-005-36401'
0.321
0.270
71
0 591
1%
R-006-3640P
41
4806
41
1%
R-008-6064P
14
34
5399
48
1%
R-016-4448P
7
3
26
4017
36
1%
R-0I0-I620P
16
1806
16
1%
R-005-0816G
7
0.659
3
1189
11
1%
R-028-3236P
119
161
140
157
47
68800
624
1%
R-013-0608P
13
5
12
6
4019
37
1%
R-010-0204P
3
II
1492
14
1%
R-0I0-4448P
98
10511
98
1%
R-022-2832P
171
18226
171
1%
12/22/98
10.0168
DEIR- 85
TAMS/l.TI/Tetra lech/MCA
-------
Table DF-2.2C
High Resolution Coring Study - Sediment Core Sample Data
Negated Values (Changed to IJndelcct Because of Blank Contamination)
Units: ng/kg
Sum of
BZ 101/
Negated
Sample ID
BZ#1 BZ#4
»Z#8
BZ# 10
BZ# i 8
BZ#19 BZJ28 BZ#44 BZ#52 UZS77 90 BZ#
18 BZ#138
BZ# 153 BZ#180
Total PCBs
Values
I raetion
HR-010-28321'
17
1828
17
1%
1IR-0I5-0204P
17
1777
17
1%
HR-018-2832P
526
54742
526
1%
HR-006-2024P
13
53
6795
65
1%
HR-034-0002A
63
6520
63
1%
HR-0I0-2428P
17
1730
17
1%
HR-005-I620P
10
984
10
1%
HR-OI5-4853P
5
61
6791
66
1%
IIR-008-6468P
8
18
2632
26
1%
HR-004-0608P
9
3
1222
12
1%
HR-004-0204P
13
1316
13
1%
HR-015-0002P
12
1129
12
1%
HR-008-I2I6P
10
11
2031
21
1%
HR-004-0406P
14
1340
14
1%
HR-013-4044P
13
10
2109
22
1%
HR-023-0002P
34
3139
34
1%
HR-0I7-3640P
5
6
12
7 4 5 12 4
4978
54
1%
HR-0I3-3640P
42
14
5202
57
1%
HR-028-2428P
134
652
72074
786
1%
HR-008-0002P
6
8
1241
14
1%
HR-008-4852P
22
28
4397
50
1%
HR-011-4852P
60
5164
60
1%
HR-0I7-3236P
4
20
8 5 6 9 3
4
5105
59
1%
HR-005-I2I6P
8
7
1221
15
1%
HR-007-5962P
6
448
6
1%
HR-015-4448P
23
256
22454
279
1%
HR-003-0002P
6
3
717
9
1%
HR-0I6-4852P
5
8
1037
13
l»'o
HR-012-3640P
3
3
409
5
1%
HR-001-08I6G
27
2088
27
1%
HR-O05-O204P
6
0.818
6
959
J2
1%
HR-015-0406P
42
3234
42
1%
HR-011-40441"
45
3403
45
1%
HR-008-5256P
30
3
27
4432
60
1%
HR-009-0506P
26
46
5282
72
1%
HR-0I2-6064P
2
1
2
397
5
1%
HR-028-0002P
2270
164451
2270
1%
IIR-017-2832P
1
5
6
6
1364
19
1%
1 86 < MS/IT'~ ^ + =
-------
Table c( 2C (
High Resolution Coring Study - Sediment Core Sample Data
Negated Values (Changed to IJndetect Because of Blank Contamination)
Units: ng/kg
Sum of
11/101/
Negated
illiplc II)
B/w 1
1)/HA
B/J8
BZMIO
li/tf 18
D/#I9 B/#28 B/#44 B/#52 B/.#77
90
li/.H 118 BZ# 138
BX# 153
UZ#180
Total PC.'Bs
Values
Traction
R-008-0204P
17
6
8
2143
30
1%
R-0I3-2024P
10
11
1506
22
1%
R-015-06081'
26
1764
26
1%
R-001-00021'
5
367
5
1%
R-010-36401'
48
3306
48
1%
R-0I2-3236P
2
2
281
4
1%
R-008-0406P
22
6
8
2378
35
1%
R-013-16201'
10
11
1407
21
1%
R-005-00021'
y
11
1349
20
2%
R-002-3236P
69
25
6178
94
2%
R-012-24281'
2
3
334
5
2%
R-OI1-20241'
50
3244
50
2°b
R-010-32361'
63
4105
63
2%
R-009-0607P
29
54
5372
83
2%
R-022-4044P
4290
352
295629
4642
2%
R-009-0405P
23
39
3961
63
2%
R-013-08121'
12
9
1325
21
2%
R-005-04061'
6
0.913
7
823
13
2%
R-014-20241'
5 1
361
6
2%
R-010-4044P
221
13654
221
2%
R-012-56601'
2
2
2
400
6
2%
R-OI 5-12161'
80
4928
80
2%
R-004-40431'
7
426
7
2%
R-017-4852P
0206
0 116
0.088
25
0 410
2%
R-007-44481'
55
24
10
5285
88
2%
R-008-06081'
23
6
8
2157
37
2%
R-008-0812P
26
6
9
2412
41
2%
R-015-20241'
93
5425
93
2%
R-022-12I6P
132
7717
132
2%
R-013-0002P
10
13
1354
23
2%
R-004-2024P
9
0.142
547
9
2%
R-018-3236P
402
107
29589
509
2%
R-004-0816G
16
1
1022
18
2%
R-004-2832P
8
0 259
465
8
2%
R-004-I216P
II
0.345
3
784
14
2%
R-004-I620P
12
680
12
2%
R-0I6-7680P
3
0.639
0.603
242
4
2%
R-006-0608P
10
7
8
1342
25
2%
12/22/98
Df-:iR-87
TAMS/l.TI/l c-lra I cth/MCA
-------
Table DF-2.2C
High Resolution Coring Study - Sediment Core Sample Data
Negated Values (Changed to IJndcteet Because of Blank Contamination)
Units: ng/kg
Sum of
HZ 101/
Negated
ample ID
BZ#I BZ#4
B/J8 UZ« 10
BZ.SI8
HZ# 19
U/J28
B/J44 BZ#52 B/#77 90 B/»l 18
B/J138
B/J 153
li/Jt 180
Total PCBs
Values
I faction
R-0I6-1620P
70
3677
70
2%
R-O04-364OP
11
571
11
2%
R-016-8084P
1
0.425
1 0.453 0.502
206
4
2%
R-0I9-3640P
181
12
7
10290
199
2%
R-0I6-00O2P
18
925
18
2%
R-OI 1*0812P
37
1850
37
2%
R-004-3236P
14
0.273
695
14
2%
R-OIO-5256P
106
39
7270
145
2%
R-013-0816G
12
II
1127
23
2%
R-005-2832P
2
5
0.265
347
7
2%
R-015-24281'
157
7686
157
2%
R-011-0608P
31
1522
31
2%
R-009-0304P
13
7
972
20
2%
R-009-081 OP
16
64
3699
79
2%
R-0I3-0406P
12
8
903
19
2%
R-006-0002P
16
4
948
21
2%
R-006-08I2P
10
5
702
15
2%
R-007-3236P
41
18
9
3041
68
2%
R-014-1620P
275
50
14429
325
2%
R-009-0102P
13
7
885
20
2%
R-010-5659P
113
44
6822
157
2%
R-009-0001P
13
6
819
19
2%
R-006-I216P
12
4
699
16
2%
R-009-0708P
28
69
4128
97
2%
R-002-2024P
21
869
21
2%
R-006-1620P
27
9
7
1820
43
2%
R-0I0-4852P
119
42
6733
161
2%
R-028-0002A
3880
160763
3880
2%
R-015-08121'
94
3856
94
2%
R-001-08I2P
4
16
780
20
3%
R-001-0406P
4
16
769
20
3%
R-007-4044P
23
11
12
S
1990
51
3%
R-0I8-4852P
0.766
1
0.484
88
2
3%
R-005-2428P
0.340
5
198
5
3%
R-00I-0204P
S
17
792
21
3%
R-016-5256P
1
2 4
4
402
11
3%
R-0I5-3236P
390
14206
390
3%
R-006-0204P
19
5
7
1097
30
3%
r^, '88 t TAMS/LTIT-ir/
( i I* I
-------
(. Table d( 2C
High Resolulion Coring Study - Sediment Core Sample Data
Negated Values (Changed to IJndetecl Because of Blank Contamination)
Units: ng/kg
Sum of
I i/101/
Negated
lliplc 11)
BZ.#I
BZM
BZ#8
B/.aiO
BZ#I8 B/.*I9
BZ#28
BZ#44 HZ#52 BZ&77 90 BZ#II8
BZ#138
BZ# 153
BZ#180
Total I'CBs
Values
Traction
•<-012-08121'
0.989
3
161
4
3%
<-006-0406P
16
7
7
1052
30
3%
<-016-72761'
2
1 0 349 0 529 0 535
173
5
3%
<-001-0608P
7
16
790
23
3%
<-012-06081'
0.962
3
147
4
3%
<-028-76791'
17700
907 1430
657
551
242
740021
21487
3%
<-008-36401'
226
124
11824
350
3%
<-01 1-5660C
146
70
7083
216
3%
<-018-06081'
1300
42515
1300
3%
<-011-52561'
102
44
4731
146
3%
<-015-40441'
3640
65 257 203
133477
4165
3%
<-008-28321'
152
4
98
8035
253
3%
<-020-32361'
1660
63 94
56967
1817
3%
<-008-32361'
212
106
9509
318
3%
<-009-16181'
II
1
0 912
0.829
4
4
636
21
3%
<-021-24281'
2270
105 119
63
74743
2557
3%
<-022-44481'
15600
424
449287
16024
4%
<-011-00021'
80
28
3017
108
4%
<-009-02031'
13
1
15
7
963
36
4%
<-009-0816G
10
6
432
16
4%
<-013-02041'
8
18
9
4
1022
38
4%
<-017-40441'
3
3 3
231
9
4%
<-019-04061'
1580
84
42
42557
1706
4%
<-013-44471'
7
5
1
328
13
4%
<-028-44481'
8210
674
386
117
232042
9387
4%
<-023-02041'
27
22
1210
49
4%
<-005-06081'
25
10
5
5
1100
45
4%
<-011-76801'
274
6330
274
4%
<-016-84881'
2
2
1
0.520
135
6
4%
<-006-40441'
2
0.797 0.846
0.765
0.508
0.178
120
5
4%
<-01 1-44481'
115
49
22
4221
186
4%
<-022-02041'
340
133
10678
473
4%
<-028-1216P
3700
436
230
193
80
104499
4639
4%
<-011-6872P
263
5914
263
4%
<-028-36401'
3800
545 671 221 132 147
122199
5516
5%
<-015-3640P
4100
45 167
92070
4312
5%
<-028-28321'
1720
98
74
53
22
39373
1967
5%
<-017-08121'
1
1
2
3
2 2 0.640
235
12
5%
12/22/98 1 0 . 0 1 72
DEIR- 89
TAMS/l.TI/TotraTcch/MCA
-------
Table DF-2.2C
High Resolution Coring Study - Sediment Core Sample Data
Negated Val
ucs (Changed to Undeleet Because of Blank Conlaminulion)
Units jag/kg
Sum of
BZ 101/
Negated
ample ID
BZ# 1
HZ#4
HZ# 8
BZ# 10
HZ# 18
HZ# 19
HZ# 28
BZ#44 HZ#52
HZ# 77 90 HZ#118
BZ#I38
BZ#153
BZ#180
Total PCBs
Values
Fraction
R-012-6870P
0 601
3
71
4
5%
R-OI8-3640P
6
7
3
313
16
5%
R-02I-I2I6I'
115
229
39
7411
383
5%
R-0II-8488P
212
86
5712
298
5%
IR-OI1-961 OP
270
5114
270
5%
IR-022-08I6G
305
119
7943
424
5%
R-0I8-4448P
25
2
491
27
5%
IR-024-I224P
1
5
1
124
7
6%
iR-028-6064P
9760
41700
5820 7810
1810 1710
1060
704
501
1234175
70875
6%
IR-OI I-7276P
226
91
5494
317
6%
IR-OI 7-0608P
0932
1
2
1
0.729 1
0.226
121
7
6%
IR-OI I-6468P
219
88
5209
307
6%
IR-022-I620P
354
141
8284
495
6%
IR-022-0406P
318
133
7547
451
6%
IR-022-0812P
250
97
5750
347
6%
IR-OI 9-2428P
2630
18000
2990
918
206
409447
24744
6%
IR-OI 2-0002P
0.762
3
3
108
7
6%
IR-012-0406P
0904
4
4
143
9
6%
IK-022-0608P
266
88
5794
354
6%
IR-018-4044P
2
4
101
6
6%
IR-028-5256P
15100
74000
12200 14600
3030 7530
3910
3130
1310
2159623
134810
6%
IR-020-2428P
42400
4290
471
981
735977
48142
7%
IR-OI 2-0204 P
1
4
5
146
10
7%
IR-021-2832P
11200
182 4740
299 284 55
220
73
69
259705
17122
7%
IR-OI 6-5660P
3
7
4
1
232
15
7%
IR-009-I4I6P
10
1
0.889
2
7
5
386
26
7%
IR-028-6872P
5610
21500
3960 5120
1040 962
111
415
575881
39384
7%
IR-021-3236P
3860
7050
103
361 181 33
124
• 101
171137
11813
7%
IR-OI 7-0406P
2
0.886
0.538 0.643
0.457
1
77
5
7%
IR-026-4048P
4
16
2
319
22
7%
IR-019-1216P
4930
23800
3650
373
468857
32753
7%
IR-016-6064P
4
7
3
1
227
16
7%
IR-017-2024P
3
10
6
4
6
3 2
467
35
7%
IR-0I6-6468P
2
5
1
0.489
1
129
10
7%
IR-001-2024P
0.755
5
1
2
112
9
8%
IR-028-08I6G
1540
39400
2650
1750 2350
1290 991
1070
837
484
643552
52362
00
IR-0I4-2428P
3
2
0.539 0.383
0.262
73
6
8%
IR-02 7-0002 P
1
2 1
49
4
8%
12/2
T"1S/I T"' -4
,/fc 4f ¦ A
-------
Table d( 2C
High Resolution Coring Study - Sediment Core Sample Data
Negated Values (Changed lo llndeleet Because of Blank Contamination)
llliplc II)
BZ#I
n/MA
BZ.#X
HZ# 10
HZ# IX
BZ u 19
nz#2x
BZ.#44
HZ# 5 2
HZ# 77
HZ 101/
90
HZ#118
HZ#138
HZ# 153
BZ#180
Total I'CBs
Sum of
Negated
Values
fraction
=1-017-02041'
0.799
0 779
0.360
0.265
0.574
0 311
0.665
42
4
9%
<-014-0816(1
2
13
3
7
3
0 735
0 950
308
29
9%
<-022-00021'
429
660
143
124X3
1232
10%
R-0I1-0XI6G
120
35
1504
155
10%
<-014-04061'
20
44
7
54
9
1 154
133
12%
<-021-00021'
207
258
45
79
4998
589
12%
<-021-06081'
174
221
40
80
4272
514
12%
<-021-04061'
291
387
63
147
7323
888
12%
<-017-44481'
1
3
1
2
1
1
76
10
12%
<-017-00021'
3
5
1
0.504
2
0.507
86
12
13%
<-021 -4044 P
118
184
77
15
32
3109
426
14%
<-009-18201'
12
0.897
0897
0 897
0.897
0 897
0.314
0.471
124
17
14%
<-0) 7-0816Ci
0.778
6
3
2
2
1
0933
0.846
120
17
14%
<-017-16201'
2
3
6
4
5
3
1
163
24
15%
<-021-08121'
256
404
60
128
5751
848
15%
<-021-02041'
218
302
239
64
131
6114
954
16%
<-026-4856P
1
1
2
0 981
1
43
7
16%
<-016-88901'
1
0.223
0.747
13
2
16%
<-014-36401'
0,249
1
0074
0.315
0 358
0.081
15
2
16%
<-009-22241'
3
0.988
0988
0 9X8
0.988
0988
0.988
0988
61
10
17%
<-024-04121'
9
18
16
3
2
287
48
17%
<-024-0816(i
3
II
7
2
7
2
179
33
18%
<-OI9-3236P
6130
1 140
8220
1900
140
95446
17530
18%
<-021 -16201'
615
839
121
312
10188
1887
19%
<-024-02041'
6
9
4
2
7
2
151
29
19%
<-021 -0816G
498
306
143
26
4555
973
21%
<-024-00021'
1
9
3
2
6
5
6
1
148
34
23%
(-033-0002A
101
52
638
153
24%
<-020-5660P
8
2
9
2
0.214
2
0.731
0486
90
24
27%
<-001-40441'
0 370
0.222
0.088
3
0680
27%
<-009-2022P
9
0.979
0.979
0 979
0.979
0.979
0.211
0.107
48
15
31%
<-021-3640P
16200
1020
2650
4190
238
246
103
211
110
78762
24968
32%
<-019-566 IP
0.611
0.841
2
0.451
0.268
0.835
15
5
34%
<-009-26281'
0 323
2
6
2
37%
<-006-48521'
0.563
0.193
2
0.756
41%
<-OI9-4448P
7
4
2
2
1
1
0.247
0.180
38
18
46%
<-019-4044P
6
4
1
4
2
0.852
0.243
0.101
40
18
46%
<-024-2434P
3
2
2
1
2
1
1
0.234
25
12
46%
I ¦'nits: (ig/kg
12/22/98
DKIR-91
TAMS/l.TI/TetraTech/MCA
-------
Table DF-2.2C
High Resolution Coring Study - Sediment Core Sample Data
Negated Values (Changed to (Jndeleci Hecause of Blank Contamination)
ample II)
HZ# 1
117*4
HZ# 8 HZ# 10
HZ# 18
HZ#
V
HZ#28
HZ.#44
HZ# 52
HZ# 77
HZ 101/
90
HZ#118
HZ#138
HZ# 153
BZ#180
Total l'CHs
Sum of
Negated
Values
fraction
IR-021-4449P
54
51
4 9
13
8
2
2
3
269
145
54%
IR-0I9-5256P
6
2
0 3
18
0 954
1
0.694
0.270
21
12
54%
IR-009-3032P
0.402
0 130
0.103
0.136
1
0.771
69%
IR-019-4852P
4
2
3
0 593
0.872
1
0.603
0.118
0.075
13
II
89%
IR-017-52561'
2
0569
3
3
92%
IR-0I7-6063P
2
0468
0.186
0.144
0.123
0.104
3
3
99%
IR-001-2428P
0.177
0.190
0.361
0 367
102%
IR-026-5664P
13
12
1
0 896
0 368
0.154
0 112
0.276
15
28
186%
IR-009-2426P
3
0.357
1
1
0.540
1
0.150
0.266
1
3
8
266%
IR-006-52561'
0 606
0 465
0.121
0.443
I
269%
IR-0I4-2832I1
0.246
4
0.679
0 140
0.354
2
6
274%
IR-026-6468P
4
13
2
0.755
5
0423
0.322
0.131
0.186
0212
0.160
9
26
295%
IR-020-4852P
2
0.456
0.438
0 190
1
3
320%
IR-020-4448P
2
0.359
0.377
0.097
0.172
0.783
3
361%
IR-020-4044P
1
0 251
0 356
0.098
0.515
2
392%
IR-006-4448P
0.627
0.980
0.339
2
474%
IR-020-3640P
3
0.362
0.446
0.102
0.658
4
578%
IR-020-5256P
10
2
0.974
4
0.276
0.203
2
17
972%
Units: ng/kg
12/2?"
12
TAMS/l.TI/Tetr^ /MCA i
-------
Due to the differences among matrices, they will be examined individually, beginning with
the water column study dissolved PCB data, as summarized in Table DF-2.2A. Subsequently the
particulate data (Table DF-2.2B) and high resolution sediment core data (Table DF-2.2C) will be
reviewed.
Of the 102 samples analyzed (5 blanks taken from Saratoga Springs groundwater are not
included in this total), 96 had at least one congener (of the 15) negated due to blank contamination
(see Table DF-2.2A). For the most part, the effect of the negated congeners on the total PCB
concentration was low. Only one mainstem sample downstream of the GE facility exceeded a
fraction of 25 percent. The other samples exceeding 25 percent were considered background stations
and unaffected tributaries whose absolute concentrations are quite low relative to the Hudson River.
As described previously, the fraction on Tables DF-2.2A , DF-2.2B, and DF-2.2C is defined as the
ratio of the sum of negated values to Total PCBs. Serious understatement of the total PCBs, defined
as those samples where the mass of negated congeners were greater than or equal to the reported
PCBs, are shown in the "Fraction" column as values of 100 percent or more. This occurred in 16
samples, including one in which the total PCBs were reported as not detected but two congeners
were detected but negated due to presence in blanks.
Even in these 16 samples, the overall effect on total PCBs is not great in most of them, as the
total of detected and negated PCBs was less than, or just slightly greater than, 1.0 ng/L, representing
background levels for the river. (The one sample in which PCBs were reported as not detected,
TW-006-0011, had a total of only about 0.1 ng/L negated congeners.) The four samples which the
effect was greatest were TW-002-0008, in which reported total PCBs were about 4.5 ng/L but
negated PCB congeners totaled over 13 ng/L (including over 1 ng/L each of BZ#1, #4, #18, #19, and
#28); TW-002-0012 (reported PCBs about 2.0 ng/L, negated congeners about 2.4 ng/L);
TW-002-0003 (reported PCBs about 2.4 ng/L, negated PCB congeners about 2.7 ng/L); and
TW-001-0016 (reported and negated congeners each about 8.6 ng/L). Most of these samples are from
water column Stations 0001, 0002, 0011, or 0012, which were collected from background Hudson
(stations 0001 and 0002) and tributaries free of GE contamination (Batten Kill, station 0011, and
Hoosic River, station 0012). or from Transect 2 (TW-002) which was problematic for other reasons.
(In fact. Transect 2 was not analyzed in detail because of the inconsistencies in congener patterns
among samples.) Thus, even large errors in the samples from background or tributary locations
represent small total mass errors as compared to the main stem Hudson River stations 0003 through
0008.
In the dissolved PCB data, the effect tended to be greatest in an absolute sense in the lower
molecular weight congeners. Although some of the higher molecular weight congeners such as
BZ#77, BZ#101/90, BZ#138, BZ#153, and BZ#138 were negated frequently (each in more than half
of the 102 samples), the absolute concentration of these congeners negated was always less than 1
ng/L, and with a few isolated exceptions (primarily BZ#101/90) consistently less than 0.1 ng/L.
Therefore, despite the high number of data points assigned non-detected values due to blank
contamination, the effect in general is not great. The effect on sample data was greater for some of
the lower molecular weight congeners such as BZ#4 and BZ#19. Although these congeners were
negated less often, the negated concentrations were greater in both an absolute sense (e.g., negations
of concentrations of 1 ng/L or higher) and in terms of the fraction of these congeners negated relative
to the total valid PCB concentration reported.
December 22. [998
DEIR-93
TAMS/LTI/Tecra Tech/MCA
-------
Of the 101 suspended solids samples analyzed, 86 (85 percent) had at least one congener (of
the 15) negated due to blank contamination (see Table DF-2.2B). For the most part, the effect of the
negated congeners on the total PCB concentration was low. Again, only one mainstem sample
downstream of the GE facility exceeded a fraction of 25 percent, with the remainder over 25 percent
representing background and unaffected tributaries. Serious understatement of the total PCBs,
defined as above (i.e., the sum of negated values were greater than or equal to the reported PCBs),
are shown in the "Fraction'" column as values of 100 percent or more. This occurred in five samples,
including one (TS-006-0002) in which the total PCBs were reported as not detected but seven
congeners were detected but negated due to presence in blanks.
In these five samples, the overall effect on total PCBs varies. These samples, and one other
sample potentially affected, are listed below. All of these samples are from water column Stations
0001,0002,0011, or 0012, which were collected from background Hudson (stations 0001 and 0002)
and tributaries free of GE contamination (Batten Kill, station 0011, and Hoosic River, station 0012)
and therefore likely to be quite low in concentration. Thus, even large errors in these samples
represent small total mass errors as compared to the main stem Hudson River stations 0003 through
0008. The Maximum Low Bias (MLB) shown is the percent of the worst-case maximum value of
total PCBs, assuming the that all the negated PCBs were actually present, represented by the
reported total PCB concentration. The MLB is calculated as:
MLB = [(total PCBs)/(total PCBs + sum UB)] * 100%
Sample ID Total PCBs Negated PCBs Maximum Low Bias
(ug/kg) (ug/kg)
TS-006-0001 163-74 160.44 50 5o/o
FS-209-0002 4953 55 35 47.2%
TS-006-0011 4953 121 44.8%
TS-001-0001 137-02 172*58 44.3%
TS-005-0002 6608 9408 41.3%
TS-006-0002 0 00 120-30 Not Defined
In the water column particulate PCB data, the effect tended to be greatest in an absolute
sense in the lower molecular weight congeners, although the effect is not as pronounced as for
the dissolved PCBs. Only BZ#180 was negated in more than half the samples (62 of 101, or 61
percent); however, the absolute concentrations were relatively low (less than 50 fig/kg [ppb] in
all cases), and never representing more than about 15 percent of the total valid PCBs (with the
exception of TS-006-0011, in which the negated concentration of BZ#180 [2.7 [ig/kg] was about
45 percent of the total valid PCB concentration [about 5.9 ug/kg]).
Due to the relatively high concentrations of PCBs in the particulate matter, the absolute
concentrations negated were higher than for the dissolved PCBs. However, as a fraction of the
total PCBs, the negated concentrations in the particulate matter were lower than in the dissolved
phase.
For example, the highest negated congener concentration in the particulate matter, about
2500 |ig/kg BZ#28 in TS-008-0004, represented less than 15% of the total valid PCB
concentration (about 18,800 |ig/kg) in that sample. The 2500 (ig/kg negation was by far the
December 22. 1998
DE1R-94
TAMS/LTI/TetraTech/MCA
-------
highest negated congener concentration; although congener concentrations 100 [ig/kg or greater
were negated in 45 other instances, only two of these negated values were greater than 500 [ig/kg,
with the next highest negated value being about 860 (ig/kg (BZ#19 in TS-001-0003).
There were 467 high resolution sediment core samples. Of these, 360 (77 percent) had at
least one congener (of the 15) negated due to blank contamination (see Table DF-2.2C). For the
most part, the effect of the negated congeners on the total PCB concentration was low. Only 28
of the 467 samples had fraction values greater than 25 percent. This represents less than 6 percent
of the total samples analyzed. Serious understatement of the total PCBs, defined as above,
occurred in only 12 samples (2.6 percent of the total high resolution sediment core samples
analyzed), including no samples in which the total PCBs were reported as not detected but
congeners were detected but negated due to presence in blanks.
As with the particulate phase samples, due to the relatively high concentrations of PCBs
in the sediment samples, the absolute concentrations negated were higher than for the dissolved
PCBs. However, as a fraction of the total PCBs (and considering the nearly five-fold greater
number of samples), the negated concentrations in the sediment samples overall were lower than
both the particulate matter and the dissolved phase. Even in the 12 samples in which the
"Fraction" exceeded 100 percent, the overall effect on total PCBs is not great in most of them,
as the total of detected and negated PCBs was less than 35 (ig/kg. There is only one sample
(HR-021-4449P) in which the fraction negated was greater than 50 percent of the valid detection
and the sum of the negated concentrations was greater than 100 jag/kg; and only one sample
(HR-021-3640P) in which the fraction negated was greater than 25 percent and the sum of
negated congener concentrations was greater than 1000 fig/kg.
With the exception of sample HR-21-3640P, the high resolution core samples with
fractions greater than 25 percent represent the deepest slices within the cores, with concentrations
less than 270 fig/kg. That is, they represent older sediments, which in most cases reflect
conditions near the onset or prior to the GE discharges to the Hudson River. As such their levels
are expected to be quite low. In these cases, the presence of a large proportion of blank
contamination has little impact on their interpretation.
Response to DF-2.2B
BZ#44 was similar to BZ#52 with respect to data quality issues. In fact, BZ#44 had a lower
frequency of blank contamination than BZ#52 for the water column samples and a slightly higher
frequency for the sediment samples (see Tables A-6, B-3 and B-4). Although it was not reviewed
at the same quality control level as BZ#52, it is expected that the data quality conclusions drawn
for BZ#52 would also apply to BZ#44.
Response to DF-2.2C
To the extent that these congeners become important in the assessment of human health
and ecological risks, USEPA agrees that it will be important to consider their suggested
concentration as inferred from the presence of other similar congeners. At this point in the
geochemical analysis, further interpretation of these specific congeners was not expected to
December 22. 1998
DEIR-95
TAMS/LTI/TetraTech/MCA
-------
greatly benefit the overall understanding of PCB fate and transport in the Hudson and so was not
pursued.
A. 1 Introduction
A.2 Field Sampling Program
A.3 Analytical Chemistry Program
A.3.1 Laboratory Selection and Oversight
A.3.2 Analytical Protocols for PCB Congeners
A.4 Data Validation
A.5 Data Usability
A. 5.1 Approach
No significant comments were received on Sections A. 1 through A.5.1.
A.5.2 Usability - General Issues
Response to DG-1.2SC
Internal consistency in the analytical procedure was maintained as much as possible and
was kept within specifications. USEPA data validation considers such issues to a limited degree
so that this concern was monitored. However, the issue of response factor changes in response
to changes in operating conditions only becomes important if one congener is more significantly
affected relative to another and not on an absolute basis. This comes about because both the
original analyses and the subsequent congener standard based correction were performed on a
ratio basis, that with respect to BZ#52. Thus as long as the response factor of the congener in
question and that of BZ#52 responded in the same manner to operational changes, then there was
no impact on the calibration correction. Undoubtedly some congeners were more sensitive and
some less sensitive to these factors, but it is deemed unlikely that the differences in relative
sensitivity would be very large. Since the primary purpose of the nontarget congener analysis was
to improve the overall estimate of total PCBs by simply including these additional congeners in
the sum. it is unlikely that this sensitivity would introduce a substantive error.
Sample drying was conducted in low temperature incubators at roughly 36°C. Samples
pairs of wet and dried analyses did not show substantive differences in congener pattern and
quantitation.
Samples were extracted using a hexane-acetone mixture. The statement on page A-4 is in
error.
A.5.3 Usability - Accuracy. Precision, Representativeness and Sensitivity
No significant comments were received on Section A. 5.3.
A.5.4 Usability - Principal Congeners
Response to DF-2.I
December 22. 199t
DElR-96
TAMS/LTT/TctraTech/MCA
-------
In the report, PCB data are examined in several ways, including total PCBs, homologue
sums, individual congener concentrations and congener ratios. No one way of examining the data
provided the key to all understandings gained during the preparation of the report. In each
analysis, the level of PCB data resolution (i.e., total PCBs, homologues or congeners) was
advanced as needed to address the specific question. While further resolution of the PCB data
could provide additional information, it was not deemed necessary for the purposes of the
questions being addressed.
In preparing the report, the focus was inherently limited to the fate of PCBs in the Hudson
River without considering the importance of a particular congener to fish or human health. This
focus was necessary since in this phase of the investigation it is unclear which congeners will be
of greatest importance in the subsequent analyses relating to fish or human health. It was also
important to understand the variability of PCB transport across the spectrum of congeners, since
many of the factors affecting transport will vary with increasing number of chlorines. The most
concentrated congeners provide the best basis to estimate this variability since these congeners
were most likely to be present in all media measured (i.e., sediments, water, and biota) and thus
provide information on the transport or exchange among media. This information can then be
applied to less concentrated congeners which may be of equal or even greater interest from a
toxicological perspective.
It should be noted here that although Appendices A and B focus on only 12 congeners, all
congener data was submitted to a rigorous, USEPA Region 2-approved, data validation procedure.
The discussion of 12 specific congeners in the appendices was focused on those congeners
expected to be of greatest importance to the ensuing data interpretation. This portion of the report
was prepared prior to the completion of the interpretive sections of the report when it was unclear
exactly which congeners would be incorporated in the report discussion. As a result, the
appendices discuss several congeners which are not specifically discussed in the report. Several
of them were discussed at length in the DEIR (e.g., BZ#1, 4, 8, 10, 19, and 52) while others will
be utilized in the upcoming modeling work (e.g., BZ#4,28,52,101, and 138). The latter set was
reviewed in the DEIR in anticipation of the data quality needs of the fate and transport models
to be discussed in the Baseline Modeling Report. The remaining three (BZ.#18, 118, and 180)
were selected because they were selected as representative of Aroclors 1242, 1254, and 1260,
respectively.
Ultimately, further congener analysis would almost certainly provide additional insights
to PCB fate and transport. However, for the purposes of the DEIR, the analyses presented provide
a sufficient level of understanding of PCB transport. Further interpretation may be completed
during the fate and transport modeling or during the ecological and human health assessments as
needed. As far as identification of problem samples is concerned, this was largely done during
the data interpretation process by examining the congener spectra associated with outliers.
Response to DG-1.25A
This congener ("PCB [sic] 12") was detected in the four lower molecular weight Aroclor
standards (1016, 1221. 1232, and 1242) at concentrations ranging from 0.083% to 0.436% of the
total mass of the Aroclor. BZ#126 was detected only in the standard for 1248 as 0.053% of that
Aroclor's mass.
December 22. 1998
DEIR-97
TAMS/LTl/Tctra 1 ech/MCA
-------
USEPA concurs that the congeners listed by GE as "trace" are generally such; analysis of
Aroclor standards (Aquatec, 1992) showed that BZ#58, 69, 140, 143, and 169 were not detected
in any of the standards analyzed for this project. BZ#184 was detected only in the Aroclor 1260
standard (0.087% of its mass); BZ#23 was detected in all Aroclors except 1260 at concentrations
consistently of about 0.06%; and BZ#96 was detected in five of seven Aroclors at concentrations
of 0.043% to 0.136%. The USEPA will take the writer's interpretation into consideration if any
of these congeners are considered individually.
Response to DG-1.25B
Congener-specific comments are presented which amplify some of the concerns and
analytical problems encountered in some of the analyses. These comments are for the most part
acknowledged and do not significantly impact the quality or usability of the data. Certain
congener-specific comments are addressed below however.
BZ#18. Data were not rejected for discrepancies up to a factor of five between the ECD and ITD
results because the ITD data were never intended to be quantitative. The purpose of the ITD
analysis was to generate confirmational data for the congener identification, not quantitation.
PCB #138. While USEPA has no knowledge that BZ#138 coelutes with #163, a fact which "was
not recognized" in the DEIR, this fact is not considered significant (regardless of whether or not
it is true). BZ#163 is not reported to be present in Aroclors 1242 or 1016. BZ#138 is present, but
at low concentrations (about 0.5% or less) in these Aroclors (Frame, 1997).
BZ#46 - The congeners reported represented the "state-of-the-art" at the time the analyses were
conducted; coupled with the necessity that the congener could be unequivocally and accurately
identified. It is noted that BZ#46 is reported to occur in only low concentrations (less than 1% of
the total mass) in Aroclors 1242 and 1016 (Frame, 1997), and only a very few of the limited
number of pentachlorobiphenyls in these Aroclors could even theoretically generate BZ#46 as a
degradation byproduct. Although consideration will be given to adding BZ#46 to any future
analysis performed, its omission from the Phase 2 analyses is not significant.
Volume 2C (Book 3 of 3)
Appendix B: DATA USABILITY REPORT FOR PCB CONGENERS WATER
COLUMN MONITORING PROGRAM
B.l Introduction
B.2 Field Sampling Program
B.3 Analytical Chemistry Program
B.3.1 Laboratory Selection and Oversight
B.3.2 Analytical Protocols for PCB Congeners
B.4 Data Validation
B.5 Data Usability
B.5.1 Approach
B.5.2 Usability - General Issues
B.5.3 Usability - Accuracy, Precision, Representativeness and Sensitivity
B.5.4 Usability - Principal Congeners
December 22. 1998
DEIR-98
TAMS/LTl/TetraTech/MCA
-------
B.6 Conclusions
No significant comments were received on Appendix B (note that some Appendix A comments
and responses also apply to Appendix B),
Volume 2C (Book 3 of 3)
Appendix C: DATA USABILITY REPORT FOR NON-PCB CHEMICAL AND
PHYSICAL DATA
C.l Introduction
C.2 High Resolution Coring Study and Confirmatory Sediment Sample Data
C.2.1 Grain Size Distribution Data
C.2.2 Total Organic Nitrogen (TON) Data
C.2.3 Total Carbon/Total Nitrogen (TC/TN) Data
C.2.4 Total Inorganic Carbon (TIC) Data
C.2.5 Calculated Total Organic Carbon (TOC) Data
C.2.6 Weight-Loss-on-Ignition Data
C.2.7 Radionuclide Data
C.2,8 Percent Solids
C.2.9 Field Measurements
C.3 Water Column Monitoring Program and Flow-Averaged Sampling Programs
C.3.1 Dissolved Organic Carbon (DOC) Data
C.3.2 Total Suspended Solids and Weight-Loss-on-lgnition (TSS/WLOI) Data
C.3.3 Chlorophyll-a
No significant comments were received on Appendix C.
rwwiw !
-------
THIS PAGE LEFT BLANK INTENTIONALLY
December 22. 19n
DEIR-100
TAMS/LTTretraTech/MCA
-------
References
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY
VOLUME A: DATABASE REPORT
VOLUME B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
D. ADDITIONAL REFERENCES FOR THE RESPONSIVENESS SUMMARY
Clark, J.F., R. Wanninkhof, P. Schlosser and H.J. Simpson, 1994. Exchanges rates in the tidal
Hudson River using a dual tracer technique. Tellus, Vol. 46B, pp. 274-285.
Clark, J.F., P. Schlosser, M. Stute and HJ. Simpson, 1996. SF6-3He Tracer Release Experiment: A
New Method of Determining Longitudinal Dispersion Coefficients in Large Rivers. Environmental
Science & Technology, Vol. 30, No. 5, pp. 1527-1532.
Deck, Bruce. 1981. Nutrient-Element Distributions in the Hudson Estuary. Doctoral Thesis, Faculty
of Pure Science, Columbia University, New York.
Flood, Roger D. Analysis of the Side-Scan Sonar, Bathymetric, Subbottom, and Sediment Data
from the Upper Hudson River Between Bakers Falls and Lock 5. Research Foundation of State
University of New York, 1993.
Frame, George M. 1997. A collaborative study of209 PCB congeners and 6 Aroclors on 20 different
HRGC columns - Part 2. Semi-quantitative Aroclor congener distributions. Fresenius J. Anal.
Chem. 357:714-722.
Freeze, R. Allan, and John A. Cherry. Groundwater. Prentice-Hall Inc., 1979.
O'Brien & Gere. 1993. Hudson River Project, 1991-1992 Sampling and Analysis Program,
Temporal Water Column Monitoring Program. Report to General Electric Company, Albany, NY.
O'Brien & Gere Engineers, Inc., Syracuse. NY.
O'Brien & Gere, 1998. Hudson River Project 1996 - 1997 Thompson Island Pool Studies, Data
Summary Report. Prepared for General Electric Company Corporate Environmental Programs,
Albany, NY by O'Brien & Gere Engineers, Inc. February.
QEA, 1998. Thompson Island Pool Sediment PCB Sources. Prepared for General Electric Company.
Albany, NY. Quantitative Environmental Analysis, LLC. Ramapo, NJ.
Rhea, J. 1998. Memorandum Re: Hudson River Project: Evaluation of Analytical Bias in the USGS
Water Column Database. Memorandum to General Electric Company, forwarded to USEPA Region
II. HydroQual, Inc., Mahwah, NJ.
Safe, S. 1990. Polychlorinated biphenyls (PCBs), dibenzo-p-dioxins (PCDDs), dibenzofurans
(PCDFs) and related compounds: environmental and mechanistic considerations which support the
development of Toxic Equivalency Factors (TEFs). Critical Reviews in Toxicol. 21:51-88.
December 21. 1998
R-l
TAMS/LTl/Tetr»Tech/MCA
-------
HUDSON RIVER PCBs REASSESSMENT RI/FS
RESPONSIVENESS SUMMARY
VOLUME A: DATABASE REPORT
VOLUME B: PRELIMINARY MODEL CALIBRATION REPORT
VOLUME C: DATA EVALUATION AND INTERPRETATION REPORT
DECEMBER 1998
Schroeder, R.A. and C.R. Barnes. 1983. Trends in Polychlorinated Biphenyl Concentration in
Hudson River Water Five Years after Elimination of Point Sources. Water-Resources Investigations
Report 83-4206. US Geological Survey, Albany, NY.
Templer, P., S. Findley, and C.Wigand, 1997. Sediment Nutrient Chemistry Associated with Native
and Non-native Emergent Macrophytes of a Hudson River Marsh Ecosystem. Institute of Ecosystem
Studies, Millbrook, NY.
USEPA, 1998. Phase 2 Report - Review Copy. Further Site Characterization and Analysis. Volume
2C-A: Low Resolution Coring Report, July.
Wershaw, R.L., M.J. Fishman, R.R. Grabbe, and L.E. Lowe, eds. 1983. Methods for the
Determination of Organic Substances in Water and Fluvial Sediments. USGS Techniques of Water-
Resources Investigations, Book 5: Laboratory Analysis, Chapter A3. Open-File Report 82-1004.
US Geological Survey, Denver. CO.
December 21. 1998
R-2
TAMS/LTl/TetriTech/MCA
------- |