Hudson
PCBs SUPERFUND SITE
Engineering Performance Standards
Technical Basis and
Implementation of the
Resuspension Standard
-------
Engineering Performance Standards
Technical Basis and Implementation
of the Resuspension Standard
April 2004
Prepared for:
U.S. Army Corps of Engineers, Kansas City District
USAGE Contract No. DACW41-02-D-0003
On Behalf of: U.S. Environmental Protection Agency, Region 2
Prepared by:
Malcolm Pirnie, Inc.
104 Corporate Park Drive
White Plains, New York 10602
and
TAMS Consultants, Inc.
an Earth Tech Company
300 Broadacres Drive
Bioomfield, New Jersey 07003
Volume 2 of 5
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
Volume 2: Technical Basis of the Performance Standard for
Dredging Resuspension
Table of Contents
List of Acronyms
1.0 Technical Background and Approach 1
1.1 Summary Statement of the Standard 1
1.2 Record of Decision 2
1.3 Definitions 3
1.4 Contaminants of Concern in Addition to PCBs 5
1.5 Remedial Design Consideration 6
1.6 Case Studies 6
2.0 Supporting Analyses 7
2.1 Turbidity and Suspended Solids at Other Sites 8
2.1.1 Reported Levels of Turbidity and Suspended Solids 9
2.1.2 Correlations Among Turbidity, Suspended Solids and PCBs 10
2.1.3 Turbidity and Suspended Solids Monitoring 11
2.2 PCB Releases at Other Dredging Sites 12
2.3 Hudson River Water Column Concentration Analysis 13
2.4 Resuspension Sensitivity Analysis 15
2.5 Dissolved-Phase Releases 19
2.6 Far-Field Modeling 23
2.6.1 Human Health and Ecological Receptor Risks 27
2.6.2 Accidental Release Short-Term Impacts 30
2.7 Near-Field Modeling 31
2.7.1 CSTR-Chem and TSS-Chem 31
2.7.1.1 CSTR-Chem 32
2.7.1.2 TSS-Chem 32
2.7.1.3 Desorption Rate Input to the Models 33
2.7.1.4 Applicability of the Models 33
2.7.2 Near-field Model Results 34
2.7.2.1 Solids and PCB Load HUDTOX Inputs 34
2.7.2.2 Solids Transport Simulation 34
2.7.3 PCB Deposition Immediately Downstream at the Dredge Operations 36
2.8 Relationship Among the Resuspension Production, Release, and Export Rates
37
2.8.1 300 g/day Export Rate Scenario 37
2.8.2 600 g/day Export Rate Scenario 38
2.8.3 350 ng/L Total PCB Concentration Scenario 38
2.8.4 500 ng/L Total PCB Concentration Scenario 39
2.9 Review of Applicable or Relevant and Appropriate Requirements (ARARs) . 40
2.10 Summary of Supporting Analyses 41
Hudson River PCBs Superfund Site
Engineering Performance Standards
l
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
Volume 2: Technical Basis of the Performance Standard for
Dredging Resuspension
Table of Contents
2.10.1 Turbidity and Suspended Solids at Other Sites 41
2.10.2 PCB Releases at Other Sites 42
2.10.3 Hudson River Water Column Concentration Analysis 42
2.10.4 Resuspension Sensitivity Analysis 42
2.10.5 Dissolved-Phase Releases 42
2.10.6 Far-Field Modeling 43
2.10.7 Near-Field Modeling 43
2.10.8 Relationship Among the Resuspension Production, Release, and Export
Rates 43
2.10.9 Review of Applicable or Relevant and Appropriate Requirements
(ARARs) 44
3.0 Rationale for the Standard 45
3.1 Development of the Basic Goals and Resuspension Criteria 45
3.1.1 Development of Water Column Concentration Criteria for PCBs 46
3.1.1.1 Obj ective 1 Development of Primary Criteria for Drinking Water 47
Drinking Water: Maintain PCB concentrations in raw water at drinking water
intakes at levels less than the federal MCL of 500 ng/L 47
3.1.1.2 Objective 2 Development of Primary Criteria for PCB Loads 50
3.2 Rationale for a Tiered Approach 58
3.2.1 PCB Considerations 58
3.2.2 Suspended Solids Considerations 59
3.2.3 Near-field Suspended Solids Criteria 61
3.2.4 Far-Field Suspended Solids Criteria 62
3.3 Monitoring Rationale 63
3.3.1 Far-Field Concerns 64
3.3.2 Near-Field Concerns 65
3.4 Data Quality Objectives 68
3.4.1 Objectives for Far-Field Monitoring in the Upper Hudson 69
3.4.1.1 Objective 1 69
3.4.1.2 Objective II 70
3.4.1.2.1 Objective III 72
3.4.1.3 Objectively 73
3.4.1.4 Objective V 74
3.4.1.5 Objective VI 75
3.4.1.6 Objective VII 75
3.4.1.7 Objective VIII 76
3.4.1.8 Objective IX 77
3.4.1.9 Objective X 77
3.4.2 Objectives for Near-Field Monitoring in the Upper Hudson 78
3.4.2.1 Objective XI 78
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
Volume 2: Technical Basis of the Performance Standard for
Dredging Resuspension
Table of Contents
3.4.2.2 Objective XII 78
3.4.2.3 Objective XIII 79
3.4.3 Additional Monitoring Objectives 79
3.4.3.1 Objective XIV 79
3.4.3.2 Objective XV 80
3.4.3.3 Objective XVI 80
3.4.4 Statistical Justification of the Sampling Frequency 81
3.5 Summary of Rationale 83
4.0 Implementation of the Performance Standard for Dredging Resuspension 85
4.1 Resuspension Criteria 85
4.1.1 Evaluation Level 87
4.1.1.1 Far-Field Net Total PCB Load 87
4.1.1.2 Far-Field Net Tri PCB Load 89
4.1.1.3 Far-Field Average Net Suspended Solids Concentration 89
4.1.1.4 Near-Field Net Suspended Solids Concentration 300 m Downstream. 90
4.1.1.5 Near-Field Net Suspended Solids Concentration 100 m Downstream
and at the Side Channel Station Without Barriers 92
4.1.2 Control Level 93
4.1.2.1 Far-Field Total PCB Concentration 93
4.1.2.2 Far-Field Net Total PCB Load 93
4.1.2.3 Far-Field Net Tri PCB Load 93
4.1.2.4 Far-Field Average Net Suspended Solids Concentration 94
4.1.2.5 Near-Field Net Suspended Solids Concentration 300 m Downstream. 94
4.1.2.6 Far-Field Net PCB Seasonal Load 95
4.1.2.7 Adjustment to the Load-Based Thresholds 97
4.1.3 Resuspension Standard Threshold 98
4.1.4 Calculation Details 98
4.1.4.1 Calculation of Total and Tri+ PCBs from Congener Data 98
4.1.4.2 Non-Detect Values 99
4.1.4.3 Upstream Source Concentrations 99
4.2 Monitoring Plan for Compliance with the Standard 100
4.2.1 Measurement Technologies 101
4.2.2 Consistency with the Baseline Monitoring Program 101
4.2.3 Compliance Monitoring Programs 102
4.2.3.1 Far-Field Monitoring 103
4.2.3.2 Near-Field Monitoring 105
4.2.4 Monitoring Locations 106
4.2.4.1 Far-field Monitoring 106
4.2.4.2 Near-Field Monitoring Locations 107
4.2.5 Potential for Reduction in the Near-Field Monitoring Locations 108
Hudson River PCBs Superfund Site
Engineering Performance Standards
ill
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
Volume 2: Technical Basis of the Performance Standard for
Dredging Resuspension
Table of Contents
4.2.6 Frequency and Parameters 109
4.2.6.1 Analytical Methods for Suspended Solids 109
4.2.6.2 Sampling Methods for Suspended Solids 109
4.2.6.3 Far-field Monitoring Parameters and Frequency 110
4.2.6.4 Lower Hudson River and the Mohawk River at Cohoes 115
4.2.6.5 Near-field Monitoring 116
4.3 Reverting to Lower Action Levels 117
4.4 Special Studies 118
4.4.1 Near-Field PCB Concentrations 118
4.4.1.1 Duration 119
4.4.1.2 Sample Collection 119
4.4.1.3 Sample Handling 119
4.4.1.4 Analytical and Direct Reading Methods 120
4.4.1.5 Definition of the Study Areas 120
4.4.2 Development of a Semi-Quantitative Relationship between TSS and a
Surrogate Real-Time Measurement For the Near-Field and Far-Field
Stations (Bench Scale) 121
4.4.2.1 Near Field 121
4.4.2.2 Far-Field 121
4.4.2.3 Study Procedures 122
4.4.2.4 Selection of Sediment Characteristics for the Study 122
4.4.2.5 Duration 122
4.4.3 Develop and Maintain of a Semi-Quantitative Relationship between TSS
and a Surrogate Real-Time Measurement For the Near-Field and Far-Field
Stations (Full Scale) 122
4.4.3.1 Duration 127
4.4.4 Phase 2 Monitoring Plan 127
4.4.4.1 Definition of the Study Areas 127
4.4.4.2 Duration 127
4.4.4.3 Assessment of Data 128
4.4.4.4 Automatic Samplers for PCB Sample Collection 128
4.4.5 Non-Target, Downstream Area Contamination 129
4.4.5.1 Definition of the Study Areas 129
4.4.5.2 Duration 130
4.4.5.3 Sampler Deployment and Collection 130
4.4.5.4 Sample Handling 130
4.4.5.5 Analytical Methods 130
4.4.5.6 Definition of the Study Areas 130
4.4.6 Further Development of the Special Studies 131
4.5 Engineering Contingencies 131
Hudson River PCBs Superfund Site
Engineering Performance Standards
iv
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
Volume 2: Technical Basis of the Performance Standard for
Dredging Resuspension
Table of Contents
4.5.1 Timeframe for Implementing Engineering Evaluations and Engineering
Improvements 132
4.5.2 Engineering Evaluations 133
4.5.3 Implementation of Control Technologies 134
4.5.3.1 Remedial Operations 134
4.5.4 Requirements of the Standard 135
4.5.5 Settled Contaminated Material and the Need for Resuspension Barriersl35
5.0 References 137
LIST OF TABLES
Table 1-1 Resuspension Criteria
Table 1-2 Sampling Requirements on a Weekly Basis - Upper River Far-Field
Stations
Table 1-3 Sampling Requirements on a Weekly Basis - Lower River Far-Field
Stations
Table 1-4 Sampling Requirements on a Weekly Basis - Upper River Near-Field
Stations
Table 1-5 Case Study Resuspension Summary
Table 2-1 Summary of Case Studies for PCB Losses Due to Dredging
Table 2-2 Far-Field Forecast Model Runs Completed for the Performance Standard
Table 2-3 Upper Hudson Conceptual Dredging Schedule
Table 2-4 Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
Table 2-5 Modeled Year-of-Compliance with Human Health Risk Assessment-
Based Concentrations for various Resuspension Scenarios
Table 2-6 Estimated Non-cancer Indices via Long Term Fish Ingestion for Several
Resuspension scenarios-Adult Angler and Upper Hudson Fish
Table 2-7 Estimated Cancer Indices via Long Term Fish Ingestion for Several
Resuspension scenarios-Adult Angler and Upper Hudson Fish
Table 2-8 Upper Hudson River Average Largemouth Bass (Whole Fish) PCB
Concentration (in mg/kg)
Table 2-9 Modeled Year-of-Compliance for River Otter Risk-Based Fish
Concentrations Upper Hudson River
Table 2-10 Lower Hudson River Average Largemouth Bass (Whole Fish) PCB
Concentration (in mg/kg)
Table 2-11 Modeled Year-of-Compliance for River Otter Risk-Based Fish
Concentrations Lower Hudson River
Table 2-12 Results for Average Dredging-Related Source Strength Estimated Fluxes
Hudson River PCBs Superfund Site
Engineering Performance Standards
v
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
Volume 2: Technical Basis of the Performance Standard for
Dredging Resuspension
Table of Contents
Table 2-13 Resuspension Production, Release, and Export Rates from TSS-Chem and
HUDTOX Models
Table 2-14 Increase in PCB Mass from Settled Material 2-Acres Below the Target
Area Estimating Using the TSS-Chem Model Results
Table 3-1 Upper 95th Percentile Estimates of Total PCB Concentrations at TI Dam
and Schuylerville Under Baseline Conditions
Table 3-2 Summary of Sampling Frequency Requirements and Expected Error Rates
Table 3-3 Summary of Sampling Frequency Requirements and Expected Error Rates
for Automatic Sampler
Table 4-1 Estimated 7-Day Total PCB Concentrations Corresponding to the
Evaluation Level (300 g/day) at the Schuylerville Monitoring Station
Table 4-2 Estimated 7-Day Total PCB Concentrations Corresponding to the Control
Level (600 g/day) at the Schuylerville Monitoring Station
Table 4-3 Estimates of Baseline Concentrations at TI Dam, Schuylerville and
Waterford
Table 4-4 Far-Field Monitoring - Analytical Details
Table 4-5 Near-Field Monitoring - Analytical Details
Table 4-6 Possible Study Areas for Nature of Release of PCB
Table 4-7 Recommended Study Areas for Nature of Release of PCB
Table 4-8 Resuspension Criteria (alternate)
LIST OF FIGURES
Figure 1-1 Schematic of Near-field Monitoring Station Locations
Figure 1-2 Schematic of Far-field Water Column Monitoring Stations
Figure 2-1 Comparison Between Upper Hudson River Remediation Scenario
Forecasts for Thompson Island Dam
Figure 2-2 Comparison Between Upper Hudson River Remediation Scenario
Forecasts for Schuylerville
Figure 2-3 Comparison Between Upper Hudson River Remediation Scenario
Forecasts for Waterford
Figure 2-4 Cumulative PCB Loads at Waterford
Figure 2-5 HUDTOX Forecasts of Whole Water, Particulate, and Dissolved Total
PCB Concentrations for Evaluation Level - 300 g/day Scenario
Figure 2-6 HDUTOX Forecasts for Whole Water, Particulate and Dissolved Total
PCB Concentration for Control Level 600 g/day Scenario (srOl)
Figure 2-7 HUDTOX forecasts for Whole Water, Particulate, and Dissolved Total
PCB Concentrations for Control Level 350 ng/L Scenario (sr04)
Hudson River PCBs Superfund Site
Engineering Performance Standards
vi
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
Volume 2: Technical Basis of the Performance Standard for
Dredging Resuspension
Table of Contents
Figure 2-8 Composite Fish Tissue Concentrations for the Upper River
Figure 2-9 Composite Fish Tissue Concentrations for the Lower River
Figure 2-10 Estimated Total PCB Concentrations at Waterford for the Accidental
Release Scenario
Figure 2-11 PCB Concentrations Downstream of Dredge for 350 ng/L Scenario
Section 1 at 1 mile and 3 miles
Figure 3-1 Examination of Analytical Precision Based on Blind Duplicates
Figure 4-1 Simplified Flow Chart for Near-field SS
Figure 4-2 Simplified Flow Chart for Far-field PCB
Figure 4-3 Simplified Flow Chart for Far-field SS
Figure 4-4 Preliminary Study Areas for the Special Studies Showing LWA
Concentrations
Figure 4-5 Preliminary Study Areas for the Special Studies Showing Sediment Types
Figure 4-6 PCB Profile in the Cores Samples Collected Post- Non-Time Critical
Removal Action in the Grasse River
ATTACHMENTS
Attachment A
Hudson River Water Column Concentration Analysis
Attachment B
Resuspension Sensitivity
Attachment C
Examination of Mechanisms for High Dissolved Phase PCB
Concentrations
Attachment D
Modeling Analysis
Attachment E
Engineering Contingencies Considerations
Attachment F
Measurement Technologies
Attachment G
Statistical Justification of the Sampling Frequency
for Phase 1 Monitoring Program
Attachment H
Estimated Cost and Feasibility of the Phase 1 Monitoring Program
Hudson River PCBs Superfund Site
Engineering Performance Standards
vii
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
List of Acronyms
ADCP
Acoustic Doppler current profiler
AMN
Water treatment facility (formerly known as SRMT)
ARARs
Applicable or Relevant and Appropriate Requirements
ATL
Atlantic Testing Labs
CAB
Cellulose Acetate Butyrate
CAMU
Corrective Action Management Unit
Cat 350
Caterpillar Model 350
CDF
Confined Disposal Facility
CERCLA
Comprehensive Environmental Response, Compensation, and Liability
Act
CF
cubic feet
cfs
cubic feet per second
CLP
Contract Laboratory Program
cm
centimeter
CPR
Canadian Pacific Railroad
CSO
Combined Sewer Overflow
CU
certification unit
CWA
Clean Water Act
cy
cubic yard(s)
DDT
Di chl orodipheny ltri chol or ethane
DEFT
Decision Error Feasibility Trials
DGPS
Differential Global Positioning System
DMC
Dredging Management Cells
DNAPL
Dense Non-Aqueous Phase Liquid
DO
Dissolved Oxygen
DOC
Dissolved Organic Carbon
DQOs
Data Quality Objectives
DSI
Downstream of the dredge area inside the silt curtain
DSO
Downstream of the dredge area outside the silt curtain
EDI
Equal Discharge Interval
EMP
Environmental Monitoring Plan
EPS
Engineering Performance Standards
EQUIL
Software model used to determine chemical equilibrium between the
particle-bound solid and the water column or aqueous phase
ESG
ESG Manufacturing, LLC
EWI
Equal Width Interval
FIELDS
Field Environmental Decision Support
FISHRAND
USEPA's peer-reviewed bioaccumulation model
Hudson River PCBs Superfund Site
Engineering Performance Standards
ACR-1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
F JI F ort Jame s W ater Intake
fps feet per second
FRRAT Fox River Remediation Advisory Team
FS Feasibility Study
ft foot
ft2 square feet
GE General Electric Company
GEHR General Electric Hudson River
GCL Geosynthetic Clay Liner
g/cc grams per cubic centimeter
g/day grams per day
GIS Geographic Information Systems
GM General Motors
gpm gallons per minute
GPS Global Positioning System
HDPE High Density Polyethylene
HUDTOX USEPA's peer-reviewed fate and transport model
IDEM Indiana Department of Environmental Management
JMP a commercial software package for statistical analysis
kg/day kilograms per day
lbs pounds
LWA length-weighted average
MCL Maximum Contaminant Level
MCT Maximum Cumulative Transport
MDEQ Michigan Department of Environmental Quality
MDS ESG Manufacturing model #. For example, MDS-177-10
MFE Mark for Further Evaluation
MGD million gallons per day
ug/L micrograms per liter
mg/kg milligrams per kilogram (equivalent to ppm)
mg/L milligrams per liter
MPA Mass per Unit Area
MVUE minimum unbiased estimator of the mean
ng/L nanograms per liter
NBH New Bedford Harbor
NJDEP New Jersey Department of Environmental Protection
NPDES National Pollution Discharge Elimination System
NPL National Priorities List
Hudson River PCBs Superfund Site
Engineering Performance Standards
ACR-2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
NTCRA
Non-Time-Critical Removal Action
NTU(s)
Nephelometric Turbidity Units
NYSDEC
New York State Department of Environmental Conservation
NYSDOH
New York State Department of Health
OBS
Optical Backscatter Sensor
O&M
Operations and Maintenance
PAHs
Polycyclic Aromatic Hydrocarbons
PCBs
Polychlorinated Biphenyls
PCDFs
Polychlorinated Dibenzofurans
pcf
pounds per cubic foot
PL
Prediction Limit
ppm
part per million (equivalent to mg/kg)
PVC
Polyvinyl Chloride
Q-Q
Quantile-Quantile
QA/QC
Quality Assurance / Quality Control
QAPP
Quality Assurance Project Plan
QRT
Quality Review Team
RCRA
Resource Conservation and Recovery Act
RDP
Radial Dig Pattern
RI
Remedial Investigation
RI/FS
Remedial Investigation/Feasibility Study
RM
River Mile
RMC
Reynolds Metals Company
ROD
Record of Decision
RS
Responsiveness Summary
Site
Hudson River PCBs Superfund Site
SLRP
St. Lawrence Reduction Plant
SMU
Sediment Management Unit
SOP
Standard Operating Procedure
SPI
Sediment Profile Imaging
SQV
Sediment Quality Value
SRMT
St. Regis Mohawk Tribe Water treatment facility (former name for AMN)
SSAP
Sediment Sampling and Analysis Program
SSO
Side-stream of the dredge area outside of the silt curtain
SVOCs
Semi-Volatile Organic Compounds
TAT
Turn-around Time
TDBF
Total Dibenzofurans
TG
turbidity generating unit
TI
Thompson Island
TIP
Thompson Island Pool
Hudson River PCBs Superfund Site
Engineering Performance Standards
ACR-3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
TM
turbidity monitoring
TOC
Total Organic Carbon
Tri+
PCBs containing three or more chlorines
TSCA
Toxic Substances Control Act
TSS
Total Suspended Solids
UCL
Upper Confidence Limit
USACE
United States Army Corps of Engineers
USEPA
United States Environmental Protection Agency
USGS
United States Geological Survey
USI
Upstream of the dredge area outside the silt curtain
USO
Upstream of dredge area outside the silt curtain
uss
US Steel
voc
Volatile Organic Compound
WDNR
Wisconsin Department of Natural Resources
WINOPS
Dredge-positioning software system used to guide the removal of
contaminated sediment
WPDES
Wisconsin Pollutant Discharge Elimination System
WSU
Wright State University
WTP
Water Treatment Plant
Hudson River PCBs Superfund Site
Engineering Performance Standards
ACR-4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Engineering Performance Standards
Hudson River PCBs Superfund Site
Volume 2: Technical Basis of the Performance Standard for
Dredging Resuspension
1.0 Technical Background and Approach
This section provides a brief summary of the standard for reference throughout this
volume of the text (subsection 1.1), identifies the basis for the standard as specified in the
ROD, (subsection 1.2), defines terms used in this volume (subsection 1.3); identifies
additional contaminants (subsection 1.4); discusses remedial design considerations
(subsection 1.5); and provides an overview of case studies applicable to the approach
(subsection 1.6) that are presented in detail in Volume 5 - Appendix.
1.1 Summary Statement of the Standard
A brief summary of the Resuspension Standard is included in this volume for
convenience. A thorough statement of this standard is provided in Volume 1. In the
formulation of the performance standard, several action levels were established so that
remediation-related problems can be quickly identified and corrected before criteria are
exceeded that would require temporarily halting the dredging operations. The
resuspension criteria include Total PCB concentration, Total and Tri+ PCB1 load, and
suspended solids concentration thresholds. These criteria are defined in Table 1-1.
The Resuspension Standard includes criteria for both PCBs and suspended solids for both
near-field and far-field conditions, which are defined as follows:
Near-field conditions are those within a few hundred meters of the remedial
operation. Only suspended solids criteria are applicable to the near-field stations.
Far-field conditions are those at specific, permanent monitoring locations that are
located at least one mile downstream of the remedial operation. Both PCBs and
suspended solids criteria are applicable to the far-field stations.
Figures 1-1 and 1-2 depict the location of the near-field and far-field monitoring stations.
Compliance with the resuspension criteria is tested through monitoring. Tables 1-2, 1-3
and 1-4 contain the compliance monitoring requirements for this program. In addition to
compliance monitoring, there are sampling requirements in the form of special studies to
gather information that can be used to further refine elements of the standard. These
studies include:
1 Total PCBs refers to the sum of all measurable PCB congeners in a sample, while Tri+ PCBs refers to the
sum of PCB congeners containing three or more chlorine atoms.
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/Earth Tech
Volume 2: Resuspension - April 2004
-------
Near-field PCB Release Mechanism (Dissolved vs. Particulate)
Development of a Semi-Quantitative Relationship between TSS and a Surrogate
Real-Time Measurement
Development of Far-Field Real-Time Suspended Solids Surrogate Measure
Non-Target, Downstream Area Contamination
Implementation of the monitoring program is specified in Section 4, as are the required
engineering contingencies in the event of exceedance of the resuspension criteria.
1.2 Record of Decision
As part of USEPA's responsibilities during the remedial design for the Hudson River
PCBs site, the agency will develop an engineering performance standard that addresses
the release and downstream transport of PCBs due to dredging operations. As specified in
the Hudson River ROD (USEPA, 2002a):
Performance standards will address (but may not be limited to) resuspension rates
during dredging... These performance standards will be enforceable, and based
on objective environmental and scientific criteria. The standards will promote
accountability and ensure that the cleanup meets the human health and
environmental protection objectives of the ROD. (ROD, page 88)
This standard is to be applied during the Phase 1 dredging effort and revised as necessary
at the end of Phase 1 to reflect knowledge gained from the first year of dredging
activities, as stated in the ROD:
...The information and experience gained during the first phase of dredging
will be used to evaluate and determine compliance with the performance
standards. Further, the data gathered will enable EPA to determine if
adjustments are needed to operations in the succeeding phase of dredging, or
if performance standards need to be reevaluated. (ROD ง 13.1, page 97)
The need for a performance standard concerning the release and downstream transport of
PCBs was recognized in the ROD:
...Although precautions to minimize resuspension will be taken, it is likely
that there will be a localized temporary increase in suspended PCB
concentrations in the water column and possibly in fish PCB body burdens.
(ROD ง 11.5, page 85)
This Resuspension Standard provides criteria to minimize the release of PCBs that are
consistent with the rates of release anticipated in the ROD, while at the same time
facilitating the removal of PCB-contaminated sediments from the river bottom. Like the
residual and productivity performance standards, the ultimate goal of this standard is to:
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
...ensure that dredging operations are performed in the most efficacious
manner, consistent with the environmental and public health goals of the
project. (ROD ง 11.5, page 85)
The ROD also identifies several applicable or relevant and appropriate requirements
(ARARs), and recognizes the need to conform to these federal and state requirements for
water quality. These guidelines were considered, to the extent appropriate.
1.3 Definitions
Dredging is fundamentally a subaqueous earthmoving action. Just as ground-based
earthmoving operations generate dust, dredging generates sediment particles that are
released into the water column. Further, just as air currents spread dust from a
construction site, ambient water currents transport resuspended sediments downstream.
Resuspended sediments with particulate-associated PCBs increase water column PCB
concentration, just as contaminated dust particles contribute to the total concentration of
airborne contaminants.
In order to clearly describe the PCB release and downstream transport related to
dredging, the following terms have been defined relative to the operation and distance
downstream:
Resuspension production rate. Dredging-related disturbances suspend PCB-
bearing sediments in the water column. The rate at which this occurs is the
resuspension production rate.
Resuspension release rate. Since most of the sediments to be remediated in the
Upper Hudson are fine sands, a significant fraction and often the majority of the
small amount of material that escapes the dredge will settle in the immediate
vicinity of the dredge. Materials that remain in the water column are then
transported away by river currents. The rate of sediment transport in the
immediate vicinity of the dredge is defined as the resuspension release rate.
Dissolved-phase PCBs. As suspended solids are transported away from the
dredge, they will continue to settle, at the same time releasing a portion of their
PCB burden into the water column, where the PCB is no longer bound to a solid
particle. PCBs located within the water column but not bound to a solid particle
are defined as dissolved-phase PCBs (smaller than 0.7 microns).
Particulate PCBs. As suspended solids are transported away from the dredge,
they will continue to settle, while at the same time PCBs bound to the solid
particles will be released into the water column. PCBs that are not released into
the water column and continue to be bound with the suspended solids are defined
as particulate PCBs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Most of this settling takes place within a few hundred yards of the dredge. Given
the extensive area of remediation in the Upper Hudson and its focus on
depositional areas, it is expected that much of the material settling in the vicinity
of the dredge will be collected during subsequent dredging passes.
Resuspension export rate. Beyond roughly one mile, further PCB removal from
the water column by particle settling becomes small, and most of the PCBs in the
water column are likely to travel long distances before being removed or captured
by baseline geochemical processes such as volatilization or aerobic degradation.
The rate at which PCBs are transported beyond one mile is defined as the
resuspension export rate. It is this rate of PCB loss, with its potential for
downstream impacts, that is the focus of the resuspension discussion in the ROD.
PCB loss due to resuspension. For the
purposes of this performance standard, PCB
loss due to resuspension, as stated in the
ROD, is defined as the resuspension export
rate just described. The standard addresses the
net export of PCBs resulting from any activity related to the removal of PCB-
contaminated sediments from the river bottom. This definition includes PCB
export resulting from the dredging operation itself and from dredging-related boat
movements, materials handling, and other activities. This definition requires both
the disturbance and the downstream transport of PCBs from the source.
PCB loss due to resuspension
requires both disturbance and
downstream transport of PCBs
from the source.
The Resuspension Standard
does not regulate resuspension
within engineered control
barriers, except for unacceptable
downstream export.
An important point is that the standard does
not directly address the resuspension release
rate or the resuspension production rate.
These rates are considered only indirectly to
the extent that they produce an export of
PCBs beyond a distance of one mile
downstream of dredging activity. Similarly, The standard does not regulate
resuspension within engineered control barriers (e.g., silt curtains), other than the
extent to which resuspension within the barriers results in unacceptable export of
PCBs downstream.
Net export of PCBs to the Lower Hudson. The net export of PCBs to the Lower
Hudson is defined as the PCB resuspension export rate at the Waterford-Lock 1
Station, i.e., the load of PCBs at this location that is attributable to dredging-
related activities. The Waterford-Lock 1 station was selected because it is
downstream of the target areas identified in the feasibility study (FS) (USEPA,
2000b) but upstream of the Mohawk River, which was shown to be a minor but
measurable source of PCBs to the Lower Hudson River (USEPA, 1997). The
Federal Dam, which is the lower boundary of the Upper River, was not chosen
because this location is downstream of the Mohawk River.
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
It is important to note that resuspension of sediments also results from other natural
processes (e.g., bioturbation and high-flow events) and anthropogenic processes (e.g., the
movement and actions of other vessels in the river). For instance, sediments are
resuspended by propeller action during recreational boating activities or commercial
shipping. Resuspension and any ensuing PCB export by these processes are accounted for
by use of the baseline monitoring water column PCB concentrations in the development
of the action levels.
In recognition of the nature of PCB release via resuspension, the Resuspension Standard
addresses two areas with respect to dredging, the near-field area and the far-field area.
Near-field area The near-field area is defined as
the region in the immediate vicinity of the
remedial operation, nominally extending from
100 feet (ft) upstream to 1 mile downstream of
the remedial operation. This area represents the
region of the water column most directly
impacted by the remedial operation. The production of suspended solids by the
dredge yields a resuspension release rate that controls local PCB levels in the water
column. Resuspension and settling superimposed on the flowing river result in
heterogeneous water column conditions in this area, making monitoring difficult.
Each remedial operation has its own near-field area, although they can readily
overlap, if deployed in the same vicinity.
Far-field area The far-field area is the region
well downstream of the remedial operations,
beginning no less than 1 mile downstream of the
dredging operation. Typically, by this distance
downstream, the majority of particle settling
related to dredging-related activities is expected
to have occurred. Additionally, there has been sufficient travel time that water column
conditions can be expected to be relatively homogeneous and, therefore, can be
sampled in a representative manner with a manageable level of effort. At this point,
PCBs in the water column resulting from dredging constitute the resuspension export
rate and are considered to be available to contaminate downstream regions.
1.4 Contaminants of Concern in Addition to PCBs
Although much of the data collected for the Hudson River focuses on PCBs because
these were selected as the contaminants of concern during the RI/FS, other contaminants
(including dioxins and metals) may also be of concern in sections of the river. This
performance standard does not address these contaminants. New York State is developing
substantive water quality certification requirements for the environmental dredging
pursuant to the federal Clean Water Act. The water column concentrations of compounds
with certification requirements will be monitored during the remediation.
Near-field area: the region in the
immediate vicinity of the
dredging, from 100 ft upstream
to 1 mile downstream of the
operation.
Far-field area: the region well
downstream of the dredging, no
less than 1 mile downstream of
the operation.
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
1.5 Remedial Design Consideration
Development of the performance standard for PCB loss due to resuspension will be done
prior to the acquisition of the design support sampling, baseline monitoring sampling and
the remedial design. As such, some broad and basic assumptions about the remedial
design are required in order to construct the standard. The performance standard only
specifies that the design must be able to achieve the performance standard; the standard
does not dictate any other specifics of the remedial design. The equipment and
procedures selected by the design team will be constrained in no other way by this
standard.
As an example, the limits on the spread of resuspended sediments that may be afforded
by the use of silt curtains or other barriers will not be considered in the development of
the standard. The design team will determine whether these measures are required.
Technologies and procedures that may be utilized to control resuspension are described
and are based on an examination of the results from case studies and the analyses
prepared for the Hudson River FS.
1.6 Case Studies
Preparation of the Resuspension Performance Standard included a review of previous
monitoring programs associated with environmental dredging efforts. Review of
historical case studies was conducted for both PCBs and suspended solids (turbidity or
suspended solids). These studies provided a useful perspective on both the extent of
dredging-related releases and the techniques used to monitor the dredging operation.
While the Resuspension Standard was developed to be specific to the conditions of the
Hudson River, these historical studies provided useful data used to support the selected
criteria and requirements.
The PCB resuspension analysis that was completed for the Responsiveness Summary (RS)
of the Record of Decision (USEPA, 2002a) provides detailed information on specific
studies of PCB release. This work has been augmented here by the inclusion of a review
of dredging-related turbidity issues. The applicable information from the case studies is
summarized as appropriate under subsection 2.2, Supporting Analyses. A discussion of
the case studies can be found in Appendix A to the Draft Engineering Performance
Standard (provided under separate cover). Table 1-5 contains a brief summary of project
information for the case studies reviewed for this standard.
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.0 Supporting Analyses
Supporting analyses were conducted during preparation of the Resuspension Standard to
address and resolve issues pertaining to the impact of dredging and PCB transport from
the dredge area to downstream locations. These analyses were completed to gather
information and to gain an understanding on the following issues:
What levels of turbidity or suspended solids have been observed at other
environmental dredging sites? (Subsection 2.2.1)
Does a correlation exist between suspended solids, turbidity and PCBs, so one can
be a surrogate indicator of the other? (Subsection 2.1)
What levels of PCB release have been observed at other environmental dredging
sites? (Subsection 2.2)
What are the baseline levels and variability of suspended solids and Total PCBs in
the Hudson River water column? (Subsection 2.3)
What is the upper bound baseline contaminant concentration per month or per
season in the Hudson River? (Subsection 2.3)
How will releases due to dredging be quantified relative to the ongoing releases
from the sediments? (Subsection 2.4)
How does the anticipated solids release from dredging compare to the baseline
levels? (Subsection 2.4)
By what mechanisms will dissolved PCBs be released and how does this compare
with particulate PCB levels? (Subsection 2.5)
Does the release of dissolved PCBs represent a significant impact that may occur
from dredging? (Subsection 2.5)
What would be considered a significant release (i.e., resuspension export rate)
from the dredging operation? (Subsection 2.6)
How may potential releases affect human health and ecological risks? (Subsection
2.6)
How much PCB may be released during dredging (i.e., resuspension production
and release rates)? (Subsection 2.7)
At what rate will resuspended sediment settle out of the water column?
(Subsection 2.7)
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
How far downstream will the settling of resuspended material occur? (Subsection
2.7)
How much material will be deposited and what is impact on the deposition areas
outside of the targeted (dredged) areas? (Subsection 2.7)
Where should monitoring be conducted to measure PCB mass loss from
dredging? (Subsections 2.1 and 2.6)
How far from the dredge should water quality monitoring be conducted and what
parameters should be measured? (Subsections 2.1 and 2.7)
To address these issues, supporting analyses were completed to define a basis on which
the standard could be established. Several of these issues were addressed as part of the
analyses completed for the ROD. Other issues required further analysis. This section
briefly summarizes these analyses and the conclusions drawn. Extensive descriptions of
the analyses completed specifically for this standard can be found in the attachments to
this document (Attachments A to G).
2.1 Turbidity and Suspended Solids at Other Sites
An evaluation was conducted to gather data concerning turbidity and suspended solids
from completed dredging projects as well as current and design-phase dredging projects.
The review of the available sites is extensively documented in Appendix A (Volume 4 of
4). Dredge sites previously researched during preparation of the Hudson River FS report
and the Hudson River RS report were also included in this study. Among the issues
addressed by this evaluation are the following:
What levels of turbidity or suspended solids have been observed at other dredging
sites?
Does a correlation exist among suspended solids, turbidity, and PCBs, so that one
can be a surrogate indicator of the other?
How far from the dredge should water quality monitoring be conducted and what
parameters should be measured?
These issues are specifically addressed in subsections 2.2.1 to 2.2.3, respectively. Table
1-5 provides a brief summary of the various sites where dredging-related turbidity or
suspended solids data were available.
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.1.1 Reported Levels of Turbidity and Suspended Solids
In most dredging studies, turbidity was the main
monitoring parameter. In several instances, data were
also collected to correlate turbidity with suspended
solids, with varying degrees of success. As to the
absolute values of turbidity or suspended solids
reported, most studies only noted the instances where conditions exceeded the site-
specific criteria. This information is useful in that it can provide some examples of
turbidity extremes related to dredging.
In most instances, the main area of turbidity or suspended solids monitoring was
conducted in the near field, as defined previously. This is discussed further in subsection
2.2.3. In general, probe measurements or sample collection were most often performed
within 1,000 ft of the dredging operation, although data were also obtained further away.
With regard to turbidity criteria, the review of case studies indicated that typical turbidity
criteria were established at levels between 25 and 50 nephelometric turbidity units (NTU)
above background levels. However, although many studies noted that turbidity
monitoring was conducted during dredging operations, no turbidity threshold was
provided in the reports, nor were data available for review. Instead, the reports concluded
that turbidity never exceeded background levels. However, useful information on
turbidity levels was obtained from some sites, as discussed below.
For New Bedford Harbor remediation in the lower harbor area, the turbidity standard was
set at 50 NTU above background levels, 300 ft from the dredge. It was indicated that the
50 NTU standard was reached infrequently and further action was not needed since this
level was not detected 600 ft from the dredge.
At the General Motors (GM) Central Foundry Division site (St. Lawrence River,
Massena, New York), the turbidity threshold was set at 28 NTU. Turbidity measurements
were periodically taken upstream and downstream of the dredge. When the value
downstream exceeded the upstream value by 28 NTU, real-time turbidity measurements
continued until the exceedance ended. Prolonged exceedances required modifications to
the waterborne remediation activities until the problem was rectified.
During dredging at the GM Massena site, 18 of 923 turbidity samples exceeded the action
level of 28 NTU above background (ranging from 31 to 127 NTU). These exceedances
were observed at a depth of 1 ft below the water surface (except for one measurement at
9 ft). The duration of the exceedance was generally reported to be two to eight minutes,
with two exceedances that lasted for 15 minutes and 45 minutes, respectively.
Both the reported values and the near-field turbidity criteria suggest maximum turbidity
values around 25 to 50 NTU above baseline conditions. Few sites routinely reported all
of their data, making further conclusions as to turbidity levels difficult. Suspended solids
data were even more rare, and in most cases were assumed to correlate with turbidity.
In most dredging studies,
turbidity was the main monitoring
parameter.
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.1.2 Correlations Among Turbidity, Suspended Solids and PCBs
Information with regard to turbidity, suspended solids and Total PCB data and associated
correlations was examined where available. Little data were available for most sites.
However, for three dredging projects, an attempt was made to correlate collected data and
draw a conclusion. In all three instances, however, the correlations were between
turbidity and suspended solids. No correlations were reported between PCBs and either
the turbidity or suspended solids parameter.
At the GM Massena site, bench scale tests were conducted prior to dredging to develop a
relationship between suspended solids and turbidity. The following correlation was
developed for overall conditions, including elevated suspended solids results {i.e., >300
milligrams per liter [mg/L]):
Turbidity (NTU) = 7.3745+(0.611058 x SS) + (0.00094375 x SS2); r2=0.941
where: SS = the suspended solids concentration in mg/L.
Based on a regression analysis completed on the data set generated from the bench scale
tests to determine whether a relationship existed between suspended solids and turbidity
at lower concentrations {i.e., when suspended solids values are less than 60 mg/L and
turbidity values are less than 60 NTU), the foregoing equation was simplified to the
following relationship by applying a linear fit curve to the plotted data set at lower
concentrations, as indicated previously:
SS (mg/L) = [0.63 x (Turbidity)] + 6.8; r2 = 0.43
where: Turbidity = the turbidity reading in NTU
Using this relationship, it was concluded that a turbidity value of 28 NTU corresponded
to a suspended solids concentration of less than 25 mg/L. It should be noted that this
relationship was the basis of the turbidity standard of 28 NTUs set for the dredging
project. It can be concluded, in essence, that GM Massena's threshold was not only to
maintain a turbidity of less than 28 NTU, but it was also to maintain a suspended solids
concentration of 25 mg/L or less.
At the Cumberland Bay remediation site (Lake Champlain, New York), a technical
specification set for the contractor was the development of a site-specific correlation
between suspended solids and turbidity. This relationship was expected to yield action
levels for the more easily measured parameter, turbidity, which in turn could be readily
correlated to suspended solids action levels during the dredging operation. To accomplish
this, the contractor performed bench scale tests prior to initiating dredging. The end result
was that a reliable suspended solids-turbidity correlation could not be determined. This
was attributed to unforeseen factors related to algae blooms and light refraction, which
caused turbidity to vary in a way that precluded its direct correlation to suspended solids.
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
A similar series of bench scale tests was conducted prior to dredging at the Fox River
Deposit N dredging site (Kimberly, Wisconsin). In addition to the tests correlating
turbidity with suspended solids, studies were conducted to determine sediment
resuspension and settling rates. The first steps were to submerge a 1-ft-thick aliquot of
Deposit N sediment under 5 ft of river water and introduce forced air into the system to
agitate it. Water samples were collected for turbidity and suspended solids analyses, and
sediment settling rates were observed within the system.
The results of this study produced the following relationship between turbidity and
suspended solids:
SS = -1.27 + 1.313 x Turbidity; r2 =0.98
Where:
SS = suspended solids in mg/L
Turbidity = turbidity in NTU
As a result of this relationship, suspended solids were estimated in the field during
dredging based on real-time turbidity measurements.
Given the success observed for the two riverine sites, it may be possible to generate a
site-specific relationship between turbidity and suspended solids for the Hudson River
during Phase 1, or with a laboratory test prior to Phase 1.
2.1.3 Turbidity and Suspended Solids Monitoring
At the dredging projects examined, the locations of near-field monitoring generally
included water quality monitoring stations upstream of the dredge, downstream of the
dredge and to the side of the dredge (a side-stream station). At sites where containment
such as sheet piling or turbidity barriers was deployed, monitoring stations were placed at
the aforementioned locations outside of the containment area. Inside the containment area
there were generally no monitors. If there were monitors, they did not have a specific
threshold level to adhere to, but instead were used to evaluate the dredge operation itself.
At sites where dredging was not contained, the monitor was located an average of 300 ft
from the dredge. Monitoring locations for several of the larger sites examined are
described below.
At the New Bedford Harbor Hot Spot dredging site, water quality monitoring stations
were situated 300 ft from the dredge. This 300-ft radial area was referred to as the
"mixing zone," an area where environmental impacts were not directly monitored. There
were no set threshold levels within the 300-ft area surrounding the dredge, as it was
assumed that solids settling out within this radius from the dredge would not result in an
adverse impact to the harbor. However, beyond 300 ft, it was assumed that solids would
have the potential to impact downstream resources.
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Another project at New Bedford Harbor, the dredging of the lower harbor, utilized the
concept of the 300-ft mixing zone as well. For this project, a turbidity threshold of 50
NTU was set at the 300-ft distance from the dredge, as noted previously. In the event that
the 50 NTU threshold was exceeded at this location, additional turbidity monitoring was
required 300 ft from the location at which the exceedance was detected (i.e., 600 ft from
the dredge) to confirm the reading and assess the magnitude of the plume.
Many of the Commencement Bay dredging projects, located off the coast of Washington
State, also utilized the concept of the mixing zone. No containment was used due to the
tidally influenced waterways; however, monitoring was conducted at the limit of the
mixing zone, which was typically established 300 ft from the dredge to ensure
compliance with state and federal waterway regulations.
At the Grand Calumet River, Indiana, remediation site, monitoring is planned at locations
both upstream and downstream, 300 ft from the dredge.
During dredging operations at the GM Massena site, water quality monitoring stations
were positioned between 200 and 400 ft downstream of the sheet piling that enclosed the
remedial operations.
Much of the available data on turbidity and suspended solids monitoring is focused in the
near-field region, where turbidity measurement is the primary parameter. Monitoring
locations are typically located 300 ft from the operation, with additional monitoring
performed at greater distances on a less-frequent basis. These locations appear to have
been selected based on professional judgment. Monitoring at these locations appears to
have successfully measured the suspended solids transport from the vicinity of the
remedial operations.
2.2 PCB Releases at Other Dredging Sites
PCB releases at other dredging sites were extensively explored as part of the RS for the
ROD (White Paper - Re suspension ofPCBs During Dredging, USEPA, 2002a). As part
of this review, three sites were found to have sufficient PCB data to permit an
examination of the rate of PCB release (see Table 2-1). Since the completion of the RS,
no other sites have been found that have data to support a similar analysis. For two of
these sites, GE Hudson Falls and New Bedford Harbor Hot Spots, monitoring around the
location was sufficient to permit an estimate of the mass of PCB transported away from
the site (i.e., out of the near-field region).
This loading information was combined with information regarding the mass of PCBs
removed to provide an estimate of the fraction of PCB lost. As noted in the white paper,
the rates of loss observed for these sites (0.36% and 0.13%, respectively) are in close
agreement with the 0.13% estimate presented in the FS for the Hudson River based on a
dredging release model.
Hudson River PCBs Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
As discussed at length in the white paper, there were specific issues on sample collection
techniques and sampling locations that compromised the data from the Fox River study in
terms of developing a flux estimate. The PCB loss estimated for this site was 2.2%. In
particular, the close proximity of the monitoring location to the dredging operation during
portions of the study (less than 0.25 mile) was a significant factor impacting the data.
These results suggest that much greater separation between source and sampling location
is needed in order to correctly represent dredging-related losses. Nonetheless, although
the magnitude of loss estimated is considered to be an overestimate, the rate of loss
estimated by the US Geological Survey (USGS) for this site was considered in the
modeling analysis in the RS, as well as later in this document.
2.3 Hudson River Water Column Concentration Analysis
Extensive study of PCB levels in the Hudson River was conducted during the
Reassessment RI/FS; however, these analyses were focused on understanding the sources
of existing loads and concentrations within the river. For the purposes of establishing a
standard for PCB losses due to resuspension, it became necessary to develop a basis for
distinguishing between dredging-related and preexisting baseline conditions. To this end,
an analysis of the mean and variation of monthly conditions in the Upper Hudson was
conducted using data obtained primarily through the ongoing post-construction remnant
deposits monitoring program conducted by GE under a consent decree with USEPA.
These data were also combined with flow data routinely recorded by USGS to provide
estimates of PCB loads in the Upper Hudson.
The analyses, details of which are presented in Attachment A, were primarily intended to
address the following two issues:
What are the baseline levels and variability of suspended solids and Total PCBs in
the Hudson River water column?
What is the upper bound baseline contaminant concentration per month or per
season in the Hudson River?
By establishing baseline concentrations and loads as well as the inherent variability of
these parameters, it becomes possible to discern the additional contributions of PCBs
originating with the remedial operations. That is, by establishing baseline conditions,
deviations from these conditions can be identified and attributed to dredging-related
releases as appropriate.
The following section briefly summarizes Attachment A of this report. The quantitative
answers to the two issues cited above are found in the tables of the attachment and are not
repeated here.
Post-1996 data collected by GE in the ongoing weekly sampling program were used in
the baseline calculations, since they represent the most comprehensive water column
Hudson River PCBs Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
dataset and probably best reflect the current conditions in the Hudson River. Baseline
conditions for suspended solids and Total PCB data were analyzed from this data set.
Three of GE's monitoring stations were analyzed for these purposes:
Rogers Island (Fort Edward)
Thompson Island Dam (TI Dam)
Schuylerville
Results for both the PRW2 and the TID-West stations at TI Dam were examined
separately. The data from Rogers Island is considered characteristic of concentrations and
loads originating upstream of the remediation area. The TI Dam and Schuylerville
stations are representative of conditions within the remediation area and are therefore
important far-field monitoring locations. Although these data are extensive, however, the
data may not be completely representative of the river conditions because of the sampling
and analytical methods employed.
The examination was limited to the months of May through November, representing the
expected dredging season. The data were examined on a monthly basis, in recognition of
the significant month-to-month variation in conditions documented in the Reassessment
RI/FS (refer to Appendix D1 of the FS). The analysis included the statistical
characterization of each month for each station, establishing a basis for estimation of the
mean and the variance of the population as a whole. Correlations with flow were
examined as well and applied when significant and useful. Minor correlations with flow
were ignored if the magnitude of the change in concentration or load was small.
Using these statistics, the following values were established for each month and station
for both PCBs and suspended solids:
The arithmetic average for a particular month
The 95th%ile upper confidence limit (95% UCL) on the average value for the
month
Data for adjacent months were combined when no significant difference was found
between means and seasonal conditions were deemed similar (e.g., May and June,
October and November). The formula applied to estimate these factors was dependent on
the underlying distribution of the data (i.e., normal, lognormal, or non-parametric).
Attachment A, Table 2, of this document contains a summary of these results.
June yielded the maximum concentrations in suspended solids at all stations; maximum
PCB concentrations were observed in both May and June; and maximum upper
confidence limits for suspended solids also occurred exclusively in June. Maximum
upper confidence limits for PCBs proved to be less systematic.
Hudson River PCBs Superfund Site
Engineering Performance Standards
14
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The baseline concentrations and loads presented in Attachment A can be used as a basis
to evaluate dredging resuspension during the remedial operation. At a minimum, daily
Total and Tri+ PCB measurements will be obtained at the far-field stations. These results
will be compared to the baseline values to determine whether the dredging-related
releases are in excess of the load-based criteria. Similarly, suspended solids will also be
used to identify dredging-related releases. In this instance, continuous or multiple daily
measurements will be used to estimate the net suspended solids increase at the far-field
stations. Net suspended solids increases above baseline will be considered indicative of
dredging-related releases. See subsection 4.1 for implementation details.
Water column concentrations may on occasion be elevated above the upper confidence
limits due to baseline processes, but it is unlikely that the concentrations will be elevated
above these levels for sustained periods of time without an obvious cause (such as a flood
event).
Each far-field station specified in the standard will be monitored during the baseline
monitoring program. These baseline data will be used to revise the estimates of the
averages and 95% UCLs at all stations and will form the basis for identifying dredging-
related releases in Phase 1.
2.4 Resuspension Sensitivity Analysis
During the dredging operation, adequate monitoring will be essential to demonstrate that
the resuspension criteria are adhered to and to verify that minimal downstream transport
of PCBs occurs. An analysis was conducted to examine the impacts of plausible dredging
releases relative to the estimated monthly baseline concentrations. Ultimately, this
analysis was needed to address portions of the following issues:
How will releases due to dredging be quantified relative to the ongoing releases
from the sediments?
How does the anticipated solids release from dredging compare to the baseline
levels?
Two analyses are summarized in this section that
have a direct bearing on this analysis. In Attachment
A, baseline concentrations and variances were
examined for two of the main far-field monitoring
stations, the TI Dam and Schuylerville. This analysis
established an average monthly concentration and an
upper bound on monthly mean concentrations. These
data were then used in an analysis to estimate
monthly loads for PCBs. A second important piece of information with respect to the
estimated fractions of PCB mass that may be exported during dredging may be found in
subsection 2.2. Values in case studies listed in Table 2-1 correspond to 0.13%, 0.36%,
Baseline concentrations and
variances were examined for two
of the main far-field monitoring
stations that established an
average monthly concentration
and an upper bound on the
monthly mean concentration.
Hudson River PCBs Superfund Site
Engineering Performance Standards
15
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
and 2.2% of the PCB mass removed. These values can be translated into an absolute mass
export rate for the Upper Hudson remediation, as follows:
y L .AQQQg
dredge ^ ^mo^SOdays^^ hr dredge
yr \mo day &
where: Fdredge = dredging resuspension export rate (or flux) in g/hr
Muh = mass of PCBs in the sediments of the Upper
Hudson to be removed by dredging (69,800 kg or
150,000 lbs) in kg
5 yrs = period of remediation (half year production in first
and last dredging seasons with four full-production-
rate years in between)2
7 mo/yr = dredging season per year
30 days/mo = days per month
14 hr/day = expected mean dredging period per day
Ldredge = dredging resuspension export rate as a fraction of
removal (unitless)
By this formula, the three percentages given above (0.13%, 0.36%, and 2.2%) translate to
PCB export rates of 6, 17, and 104 grams per hour (g/hour) of dredge operation,
respectively. These values are comparable in magnitude to the nominal baseline daily
flux of PCBs during the dredging season, generally ranging from 20 to 80 g/hr.3 Thus the
lower end of the possible export rates will be difficult to observe relative to the
magnitude and variability of baseline fluxes as demonstrated in the variations discussed
in Attachment A. In light of this observation, three nominal resuspension export rates
were explored in this analysis: 0.5%, 1.0%, and 2.5%. These translate to 24, 47, and 119
g/hr respectively (or nominally 300, 600, and 1,600 g/day on a 14 hour/day basis).
Recognizing the anticipated range in river conditions over the dredging season, the
analysis was conducted for Total PCBs in the Upper Hudson River over a wide range of
river flow rates (2,000 to 10,000 cubic feet per second [cfs]). The suspended solids
increase in the water column was estimated based on:
Volume of sediment removed, the density of the sediment
Dredging-induced resuspension export rate
2 This removal rate represents the target removal schedule in the Productivity Performance Standard.
3 This range is based on a range of flows from 3,000 to 5,000 cfs and a water column concentration of 75 to
150 ng/L, typical of conditions in the TI Pool in June and July.
Hudson River PCBs Superfund Site 16 Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Resuspension - April 2004
-------
Flow rate
Length of the dredging season
Similarly, the Total PCB increase in the water column was computed as a function of:
Mass of Total PCBs to be removed
Dredging-induced resuspension export rate
River flow rate
Length of the dredging season
These results are presented in Attachment B of this performance standard. Because
dredging-related export is calculated as a net addition of PCB or suspended solids (mass
per unit time), the additional flux is independent of the river flow; however, the estimated
increase in water column concentration will vary inversely with flow. For these estimates,
dredging releases were not considered to be flow-dependent but rather to result from
spillage, equipment handling, etc., all of which are independent of flow.
These estimated increases in concentration were then translated into a dredging-induced
suspended solids and Total PCB concentrations in the river system. This was computed
by adding the system's baseline variation of suspended solids and Total PCB
concentrations (the estimated baseline concentrations) to the estimated increase in
concentration (loading) as a result of solids loss from the dredging operation. Comparison
of these potential in-river suspended solids and Total PCB concentrations were evaluated
against the estimated suspended solids and Total PCB monthly baseline concentrations to
determine the level of "significant" increase in the river system over baseline
concentrations that signals an unacceptable dredging-related impact.
This analysis was completed for both monitoring stations at the TI Dam and for the
Schuylerville monitoring station. Attachment B provides a detailed analysis for each
monitoring station. The analysis identified periods of the dredging season wherein 600
g/day PCB export rate loading from the dredging operation would increase the Total PCB
water column levels to a concentration just below 350 ng/L at the Schuylerville
monitoring station. These elevated Total PCB water column concentrations were
estimated for the months of May and June during low-flow conditions at the
Schuylerville monitoring stations. Similar values were estimated for the TID-PRW2
station. Concentrations exceeding 350 ng/L were calculated for the TID-West station at
low flow. In all three instances, however, the data may not be truly representative of the
river conditions at the location, in light of concerns over collection techniques. Thus, any
conclusions drawn from the data are tentative.
With the exception of estimated Total PCB concentrations during the months of May and
June during low-flow conditions, it was concluded that 300 g/day and 600 g/day releases
of Total PCBs due to dredging will correspond, overall, with a Total PCB concentration
in the water column of less than 300 ng/L Total PCBs on average. This indicates that
concentrations can be maintained below the 350 ng/L criterion of the Control Level.
Generally, this analysis identified problematic times of year during the dredging season
Hudson River PCBs Superfund Site
Engineering Performance Standards
17
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
wherein extra care will need to be taken to maintain minimal releases from the dredge to
avoid exceedance of the Total PCB concentration resuspension criteria. These results also
suggest that if low-flow conditions occur during the months of May and June, less-
contaminated areas might be chosen for remediation in favor of more highly
contaminated areas.
A sensitivity analysis was conducted on the annual
PCB loading baseline to evaluate the impact
associated with a dredging-induced PCB loading into
the water column. This analysis was completed to
evaluate whether the remediation of the Upper
Hudson via dredging will have a measurable impact
on the annual PCB loads. The baseline annual PCB loading was estimated for each of the
monitoring stations for the period of 1992 through 2000 and compared to the dredging-
induced PCB loading, assuming PCB export rates of 300 g/day, 600 g/day, and 2,300
g/day. The 2,300 g/day value corresponds to load conditions at the Resuspension
Standard threshold for Total PCBs of 500 ng/L.
Assuming that dredging work would occur seven days per week and that the increase in
concentrations would occur only during the 14-hour-per-day working period, the
dredging-induced PCB loading for each of these scenarios was computed as a function of
the following:
Volume of sediment removed
Total PCB concentration on the solids
Induced Total PCB flux
Section of the river being remediated
This analysis is presented in Attachment B of this document.
Comparison of the baseline annual PCB loading
to the dredging-induced PCB loading for the three
scenarios indicated that a well-controlled
dredging project at full production (export of 300
g/day Total PCBs from dredging) would release
less than 65 kg per year Total PCBs into the river
as a result of the remediation, and that a 600
g/day Total PCB export rate from dredging would
result in an annual loading of about 130 kg per
year Total PCBs.
The Resuspension Standard threshold would result in
an annual loading of 500 kg/year into the river. It can
be seen that these rates of mass loss begin to become
significant relative to the baseline annual loads. It
was concluded that an annual dredging-induced 65
A sensitivity analysis conducted
on the annual PCB loading
baseline evaluated the impact of
dredging-induced PCB loading
into the water column.
Analysis indicated:
A well-controlled dredging project
exporting 300 g/day Total PCBs
would release < 65 kg per year Total
PCBs into the river.
A 600 g/day Total PCB export rate
from dredging would result in approx.
130 kg/yr annual loading of Total
PCBs to the river.
The Resuspension Standard
threshold results in a 500
kg/year annual loading of Total
PCBs to the river.
Hudson River PCBs Superfund Site
Engineering Performance Standards
18
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
kg/year Total PCB loading is a relatively small fraction of the baseline load to the river in
most years, and that the Total PCB load induced by the Resuspension Standard threshold
is similar to PCB loadings that occurred in the early 1990s. This rate of export will be
controlled through limits on the seasonal and daily rates of dredging-induced PCB export
to prevent excessive PCB loss when the baseline PCB concentrations are low and the
concentration criteria would allow higher export rates.
It is concluded from this analysis that the
PCB concentration and load criteria
established for the Resuspension Standard
and action levels are protective of the river
system and would generate Total PCB
concentrations typically within the baseline
variability of the river system.
2.5 Dissolved-Phase Releases
Evidence has been reported from the Fox River study (USGS, 2000) to suggest that a
large dissolved-phase release of PCBs is possible in the absence of any apparent increase
in the water column loading of suspended solids. As a result, theoretical analyses were
conducted to assess the potential mechanisms by which dissolved PCBs could be released
into the water column. An attempt was then made to quantify the potential release of
PCBs in the dissolved phase. The following issues were explored through theoretical
analyses to estimate a quantity of PCBs that may be released into the river system in the
dissolved phase:
By what mechanisms will dissolved PCBs be released and how does this compare
with particulate PCB levels?
Does the release of dissolved PCBs represent a significant impact that may occur
from dredging?
To some degree, resuspended solids lost from the
dredge will release their PCB burden into the
dissolved phase as the solids concentrations attempt
to establish equilibrium. PCBs will continue to move
from the particulate phase on the resuspended solid to
the dissolved phase in the water column until a steady
state is reached, a process that is otherwise known as establishing equilibrium.
Once equilibrium is reached, the PCB concentration on the resuspended solid can be
estimated, as well as the concentration of PCBs in the dissolved phase. Impacts of
resuspension downstream of the dredging area can now be determined, since the PCB
flux from the dredging area has been quantified. In addition, the quantity of dissolved
phase PCBs released into the water column may have a significant impact on the water
Conclusions:
Annual dredging-induced 65 kg/yr Total PCB
loading is relatively small fraction of baseline
annual load to the river.
Resuspension Standard criteria and action
levels are protective of the river system.
PCBs move from the particulate
phase on the resuspended solids
to a dissolved state until a
steady state, or equilibrium, is
reached in the water column.
Hudson River PCBs Superfund Site
Engineering Performance Standards
19
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
column quality, depending on the concentration and quantity of the dissolved-phase
release.
There are two basic pathways by which dissolved-phase PCBs can be released into the
water column:
Through direct releases of pore water to the overlying water column as a result of
the dredge's making a cut into the sediment
Directly from a solids release/loss into the water column from dredging
Once solids are displaced into the water column, PCBs begin to partition from the
particulate phase to the dissolved phase in an attempt to reach equilibrium within the
system. In the event that the suspended solids added to the water column are of sufficient
mass and contamination level, the dissolved-phase concentration will rise markedly.
These analyses are described in detail in Attachment C to this document. A summary of
the analyses assumptions, methodology, and conclusions are presented below.
The first theoretical model analyzed was the three-phase partitioning model, which was
examined to evaluate conclusions drawn from PCB-loss calculations estimated for
dredging conducted at the Fox River dredging sites. Specifically, the reported fraction of
total mass loss as dissolved phase during dredging was approximately 1% of the total
mass removed (USGS, 2000).
The three-phase partitioning model presented here assumes that the contaminant, PCBs,
reaches equilibrium among particulate, truly dissolved, and dissolved organic carbon
(DOC)-bound phases. This model was employed on a mass of contaminant-per volume of
sediment basis. The three-phase partitioning model was evaluated using the Hudson
River data. Detailed analysis and parameters used for this model can be found in Section
2 of Attachment C.
It was determined, using the three-phase equilibrium model, that the Hudson River
sediment pore water contains very little of the in situ sediment PCB mass. More
specifically, the three-phase partitioning model indicated that the dissolved phase
represents 0.002% of the Tri+ fraction of PCBs relative to the sediment-bound PCB
fraction of 99.998%. For the mono- and di-homologue fractions, the dissolved phase
represents 0.004%, as compared to the sediment-bound PCB fraction of 99.996%.
These percentages of dissolved-phase PCBs per sediment-bound PCBs were then used to
estimate the number of pore water volumes that would need to be displaced to achieve a
P/o mass loss, as reported from the Fox River case study. The number of pore water
volumes is computed as the proportion of water-to-sediment volume or the desired mass
to be lost (1%>) over the mass available in a single pore water volume (either 0.002% for
Tri+ or 0.004% for mono- and di-chlorobiphenyls).
This computation estimated that 420 volumes of pore water would need to be released for
the Tri+ fraction, or 210 cubic yards (cy) of water per cy of sediment, assuming the
Hudson River PCBs Superfund Site
Engineering Performance Standards
20
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
sediment are half water and half sediment. For the
mono- and di-chlorobiphenyls, approximately 250
pore water volumes would need to be released, or 125
cy of water, assuming the sediment is half water and
half sediment. It was concluded from this analysis
that a direct loss of PCBs to the water column from
the dissolved phase through the pore water would be
highly unlikely, because such a large volume of water
must be displaced to result in a measurable release of dissolved PCBs.
Another analysis was the application of the two-phase partitioning model to estimate the
distribution of the dissolved-phase PCBs to the total concentration of PCBs in the water
column due to dredging. This analysis was conducted to evaluate whether it is sufficient
to simply measure whole-water PCBs during dredging or whether the dissolved phase
must also be measured if it is representative of a significant concentration. This model
assumes equilibrium exists between the dissolved-phase fraction and the suspended phase
fraction.
Data collected in the GE float surveys show that sediments continue to release PCBs to
the water column throughout the year even when low-flow conditions exist and no
observable resuspension is occurring in the system. Thus, for this analysis, a scenario was
assumed in which a suspended solids concentration of 1 mg/L would be temporarily
added to the system as a result of dredging. It was thought that evaluating the magnitude
of PCBs in the water column for this scenario would allow for a preliminary evaluation
as to whether the effects of dredging could be distinguished from the baseline river
conditions.
In fact, the estimated fraction of dissolved phase PCBs estimated for the dredging-
induced scenario in which suspended solids was released into the water column was
similar to background concentrations. The fraction of dissolved phase to total water
column PCB concentration for both background and after dredging is similar, on the
order of 0.9. It was concluded that it is not possible to distinguish the effect of dredging
by simply comparing the fraction of the dissolved phase increase in the water column.
Both of the foregoing analyses assume that the solids and dissolved phase PCBs reached
equilibrium. Recent studies have indicated, however, that full chemical equilibrium may
not be reached since the desorption rates of hydrophobic chemicals from sediment tends
to be slow. It is thought that the residence time of a resuspended particle in the water
column from dredging is relatively short {i.e., on the order of hours). Assuming a few
hours' residence time, it is not likely that the PCBs will reach equilibrium. In response to
this concern, a literature review was conducted to obtain data on desorption equilibrium
and kinetics of PCBs so this analysis could be carried out and evaluated.
The PCBs desorption rate constants reported in the literature are homologue-based,
except for those of Carrol, et al. (1994), who used an untreated PCB that was comprised
of 60% to 70% mono- and di-chlorinated biphenyls. The desorption rate constants were
The amount of displaced pore
water needed to achieve a
measurable release of
dissolved-phase PCBs is so
high that direct loss of PCBs to
the water column through pore
water is highly unlikely.
Hudson River PCBs Superfund Site
Engineering Performance Standards
21
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
determined to vary between 4.2 x 10"4 to 0.2 hr The reported rate constants correspond
to a half-life of approximately 3 to 1,700 hours and equilibrium times of 26 hours to 980
days. Given the length of time that it takes the PCBs to reach equilibrium, as described by
these rate constants, it was concluded that it is highly unlikely that there will be large
amounts of dissolved-phase PCBs released as a result of dredging. To validate this
statement, the reported desorption rate constants were applied to the Hudson River
sediment and dredging conditions. Applying these values into a kinetic rate equation, it
was estimated that the dissolved-phase PCB released due to dredging may range from 7.6
x 10"5 to 3.2 ng/L, which is approximately 0.04 to 18% of the whole-water PCB
concentration. These estimates indicate that the amount of dissolved-phase PCBs
introduced into the system will be relatively small and comparable to background
concentrations.
The theoretical analyses conclude that the release of a large amount of dissolved-phase
PCBs is unlikely to occur as a result of dredging. It is possible to assess these results
using field measurements of dissolved and suspended PCB concentrations in the water
column during dredging, using the case study data. Measurements of dissolved- and
particulate-phase PCBs were collected during the predesign field test conducted at the
New Bedford Harbor during August 2000 (USACE, 2001).
The particulate PCB and suspended solids measurements taken during the dredging at
New Bedford Harbor show patterns of concentrations similar to what would be expected
during the remediation. At the point of dredging, the particulate PCB concentrations are
elevated by a factor of about ten over the upstream conditions, but within 1,000 ft
downstream of the dredge, the concentrations were only slightly greater than the highest
measured upstream concentration. Turbidity levels drop off quickly with distance to
upstream monitoring point conditions.
The dissolved-phase PCB concentrations at the dredge are again about ten times larger
than the upstream concentrations but these concentrations drop off quickly into the range
of the upstream samples. The upstream PCBs concentrations are about 60% dissolved. At
the dredge this value drops to below 20%, indicating that PCBs released via dredging are
primarily solids-bound. Downstream of the dredge the% of dissolved-phase PCBs is
more variable but remain less than the 60% fraction at the upstream location. This
variability in the downstream samples is mirrored in the particulate PCB and suspended
solids measurements.
These results of this study are consistent with a mechanism of PCB release through the
suspension of contaminated solids. This conclusion is more convincing in light of the
high concentrations in New Bedford Harbor (860 ppm on average in the top 0 to 1 foot
segment) relative to the Hudson River (approximately 50 ppm on average in the
Thompson Island Pool [TI Pool]).
Hudson River PCBs Superfund Site
Engineering Performance Standards
22
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.6 Far-Field Modeling
To study the long-term impacts of dredging, far-field modeling was used to simulate
water column, sediment and fish Tri+ PCB concentrations in the Upper and Lower
Hudson River as a result of the dredging operation. The far-field model was applied to
determine the following:
What would be considered a significant release {i.e., resuspension export rate)
from the dredging operation?
How may potential releases affect long-term human health and ecological risks?
What would be the short-term impact of an accidental release on the public water
supply?
The modeling efforts were focused on examining the impact of running the dredging
operation at the specified action levels in the Resuspension Standard. The water column,
sediment, and fish Total PCB concentrations were forecasted using USEPA's peer-
reviewed, coupled, quantitative models for PCB fate, transport, and bioaccumulation in
the Upper Hudson River, HUDTOX and FISHRAND, which were developed for the
Reassessment RI/FS.
HUDTOX was developed to simulate PCB transport and fate for the 40 miles of the
Upper Hudson River from Fort Edward to Troy, New York. HUDTOX is a fate and
transport model that is based on the principle of conservation of mass. The fate and
transport model simulates PCBs in the water column and sediment bed, but not in fish.
For the prediction of the future fish PCB body burdens, the FISHRAND model was used.
FISHRAND is a mechanistic time-varying model incorporating probability distributions.
It predicts probability distributions of expected concentrations in fish based on
mechanistic mass-balance principles, an understanding of PCB uptake and elimination,
and information on the feeding preferences of the fish species of interest. Detailed
descriptions of HUDTOX and FISHRAND models can be found in the Revised Baseline
Modeling Report (USEPA, 2000c).
For the Lower Hudson River, the Farley et al. (1999) fate and transport model was used.
The water and sediment concentrations from the Farley fate and transport model were
used as input for FISHRAND to generate the PCB body burden estimates for fish species
examined in the Lower Hudson.
As part of the modeling effort for the Resuspension Standard, the following scenarios
were simulated using HUDTOX, FISHRAND, and Farley models:
Dredging scenario with no resuspension release rate (HUDTOX run number
d004).
Hudson River PCBs Superfund Site
Engineering Performance Standards
23
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Dredging scenario with a net increase in Total PCB mass export of 300 g/day at
the far-field monitoring stations (run number sr02). This essentially simulates the
Evaluation Level of the Resuspension Standard.
Dredging scenario with a net increase in Total PCB mass export of 600 g/day at
the far-field monitoring stations (run number srOl). This corresponds to the
Control Level of the Resuspension Standard.
Dredging scenario with a maximum Total PCB concentration of 350 ng/L at the
far-field monitoring stations (run number sr04). This corresponds to the Control
Level of the Resuspension Standard.
Dredging scenario with an accidental release during the 600 g/day dredging
operation conditions.
Table 2-2 contains a list of completed model runs used in this report. Unlike the previous
modeling efforts performed for the RS for the ROD (USEPA, 2002a), the model
simulations completed for the Resuspension Standard track the sediment being
resuspended as a result of dredging. The dredging scenarios with resuspension release
were simulated with additional solids and Tri+ PCB loading to the model segments. In
addition to simulating the "best estimate" of PCB resuspension release during dredging,
the dredging schedule was shifted from 2004 to 2006, as seen in the start years listed in
Table 2-3.
The resuspension scenarios in the foregoing bullets are specified as the PCB export rate
at the far-field monitoring stations. Due to the nature of the HUDTOX model structure,
PCB loads cannot be readily specified at far-field locations {i.e., specifying the
resuspension export rate). Rather, the input of PCBs is specified as an input load at a
location within the river, equivalent to a resuspension release rate. In order to create a
correctly loaded HUDTOX run, it is first necessary to estimate the local resuspension
release rate from the dredging operation; that is, the rate of Total PCB and solids
transport at the downstream end of the dredge plume. At this location most of the solids
that are going to settle out will have settled out and the suspended solids will more
closely resemble those simulated by HUDTOX.
Unfortunately, there is no direct way to establish the relationship between the
resuspension release and export rates prior to running the models. To estimate the
suspended solids flux input loading term for HUDTOX, a near-field model was
developed (TSS-Chem) based in part on the work by Kuo and Hayes (1991). The Total
PCB input loading term for HUDTOX (the resuspension release rate) was derived
iteratively so as to obtain the desired PCB export rate at the far-field monitoring location.
The resuspension release rate was obtained by checking the resuspension export rate
(output from HUDTOX) until the model output gave the desired Total PCB export rate.
Once the resuspension release rate that created the desired resuspension export rate was
obtained, the corresponding suspended solids flux associated with the Total PCB release
Hudson River PCBs Superfund Site
Engineering Performance Standards
24
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
rate was estimated using TSS-Chem model. Detailed descriptions of the TSS-Chem and
HUDTOX models and their use are provided in Attachment D.
Appendix D contains a complete discussion on the
effects of different formulations for suspended solids
flux input to the model. From this study, it was
concluded that the PCB export rate is not particularly
sensitive to the amount of solids (suspended solids
flux) loaded with the PCBs. A scenario with no solids
added to the model segments increases the Total PCB export rate minimally (less than
15%) compared to the scenario with the suspended solids flux added derived from the
one-mile plume scenario of the TSS-Chem model.
Figures 2-1 through 2-3 present comparisons of predicted HUDTOX Tri+ PCB
concentrations in the water column at various locations throughout the Upper Hudson
River for the monitored natural attenuation (MNA), no resuspension, and three action
level scenarios over a 70-year forecast period.
The effect of running the dredging operations at the Evaluation Level (300 g/day) and the
Control Level (600 g/day) on predicted water column Tri+ PCB concentrations is largely
confined to the six-year active dredging period (2006 through 2011). Outside of the
period of scheduled dredging (2012 and later), impacts on water column Tri+ PCB
concentrations are minimal. However, in River Section 3 only, running the dredging
operations at the Control Level or 350 ng/L (or 1,600 g/day) results in significantly
higher water column concentrations during the dredging period and slightly elevated
water column concentrations for approximately ten years after completion.
The cumulative Tri+ PCB load at Waterford as forecasted by HUDTOX was used to
determine what would be considered a significant release {i.e., resuspension export rate)
from the dredging operation. Figure 2-4 shows the Tri+ PCB load forecasts for several
load conditions. The lower bound will be the ideal conditions of dredging, where there
are no sediments being spilled (no resuspension) and the upper bound will be the MNA
scenario. The 300 g/day scenario was only simulated through 2020. From the figure, it
was shown that the Tri+ PCB load for this scenario crosses the MNA by the completion
of dredging (2011).
The HUDTOX forecast for the Tri+ PCB load from
the 600 g/day scenario remained higher than the
MNA for a little longer, approximately four years
after completion of dredging operations (2015).
However, HUDTOX forecasts showed that Tri+
PCB cumulative loads for both 300 g/day and
600g/day scenarios would be lower than the MNA.
This suggests that these two scenarios would yield acceptable loads to the Lower River.
HUDTOX results for the 350 ng/L scenario showed that cumulative Tri+ PCB loads will
go below the MNA cumulative loads for the 70-year forecast period. This suggests that
The PCB export rate is not
particularly sensitive to the
amount of solids (suspended
solids flux) loaded with the
PCBs.
HUDTOX forecasts showed that
Tri+ PCB cumulative loads for both
300 g/day and 600g/day scenarios
would be lower than the MNA,
suggesting acceptable loads to the
Lower River.
Hudson River PCBs Superfund Site
Engineering Performance Standards
25
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
by running the dredging operations at the Control Level of 350 ng/L for the entire
program, significantly more Tri+ PCB mass will be transported to the Lower River
relative to the MNA scenario, yielding an unacceptable amount of release.
Similar conclusions can be drawn for the Total PCB load estimates, although longer
periods are estimated until the 300 g/day and 600 g/day dredging scenarios cross the
MNA trajectory. These forecasts are considered less certain, however, since the models
do not directly simulate Total PCBs, but rather Tri+ PCBs. The Total PCB estimates are
based on estimates of Tri+ to Total PCBs in the resuspended sediments (refer to the
White Paper - Relationship Between Tri+ and Total PCBs in the RS for more details
[USEPA, 2002a]).
In addition to giving an indication of significant release, the results from HUDTOX runs
may also provide an indication of the water column concentrations for the different
dredging scenarios. Figures 2-5 through 2-7 show the whole water, dissolved phase, and
particulate phase Total PCB concentration for the 300 g/day, 600 g/day, and 350 ng/L
scenarios during the dredging period (2006 to 2011).
The HUDTOX model predicted that by running the dredging operations at the Evaluation
Level (Total PCB flux of 300 g/day), the mean whole water column Total PCB
concentrations at the TI Dam would be less than 160 ng/L. At Schuylerville and
Waterford, the HUDTOX model predicted that the whole water column concentrations
would average less than 120 and 80 ng/L, respectively (Figure 2-5). The water column
Total PCB concentrations as a result of running the dredging operations at 600 g/day
would be higher than those of the 300 g/day scenario, as expected. The mean whole water
Total PCB concentrations at the TI Dam during the dredging period (2006 to 2011) for
the 600 g/day scenario are predicted to be less than 250 ng/L except for few days in June
2008 (Figure 2-6). The whole water Total PCB concentrations at the Schuylerville and
Waterford monitoring stations are predicted to be lower than 200 and 150 ng/L,
respectively.
For the 350 ng/L scenario, as expected, the HUDTOX forecast shows that on average, the
whole water Total PCB concentrations will be approximately 350 ng/L (Figure 2-7). The
predicted Total PCB concentrations in the water column during River Section 2 dredging
are higher than 350 ng/L because the forecast flow used in the model during that
dredging period (August 16 to November 30, 2009) is about 15% lower than the
historical average flow based on the USGS data. Therefore, the higher concentrations are
expected. However, the average concentration during the entire dredging period for River
Section 2 (August 16 to November 30, 2009 and May 1 to August 15, 2010) is expected
to be around 380 ng/L.
Figure 2-8 depicts the annual species-weighted fish body burdens for human fish
consumption at RM 189, 184, and 154. The fish concentrations used are the species-
weighted averages, based on Connelly et al. (1992), and are considered to represent a
reasonable ingestion scenario among the three fish species consumed to any significant
extent by human receptors (anglers) (USEPA, 2000a):
Hudson River PCBs Superfund Site
Engineering Performance Standards
26
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Largemouth bass - 47%
Brown bullhead - 44 %
Yellow perch - 9%
FISHRAND fish body burdens forecasts for the MNA, no resuspension, 350 ng/L Total
PCBs, and 600 g/day Total PCBs scenarios were plotted in the figure. The 300 g/day
scenario was not simulated since the Tri+ PCB loads to the Lower River are lower than
both the 600 g/day and 350 ng/L scenarios. FISHRAND modeling results for the Upper
River show that the impact of the 600 g/day scenario on fish tissue concentrations is
largely confined to the dredging period in River Sections 1 and 2 (Figure 2-8), similar to
the water column results from the HUDTOX model. In River Section 3, the impact to the
fish tissue concentrations lasts about three years beyond the dredging period to
approximately 2014.
The forecast results from the different dredging scenarios indicated that the impacts to
fish tissue concentration would largely be short-term (i.e., confined to the years during
the dredging period) for River Section 1, even for the 350 ng/L scenario. The impact of
the 350 ng/L scenario is slightly longer lasting in River Section 2 compared to that for
River Section 1 (Figure 2-8).
Long-term human health and ecological risks are discussed in the following subsection.
2.6.1 Human Health and Ecological Receptor Risks
This subsection compares long-term risks (i.e., after completion of dredging) from
consumption of PCB-contaminated fish to anglers and ecological receptors (as
represented by the river otter [Lutra canadensis]) under the following scenarios:
No resuspension
350 ng/L Total PCBs
600 g/day Total PCBs
Monitored natural attenuation scenarios
Risks were calculated with exposure durations beginning one year after the year in which
dredging will be completed in the each section of the river and the average of the upper
river (Table 2-3). Exposure durations (e.g., 40 years for evaluating cancer risks to the
reasonably maximally exposed [RME] adult angler, 7 years for evaluating non-cancer
health hazards to the RME adult angler) and all other risk assumptions, locations, toxicity
values, receptors, and fate, transport, and bioaccumulation models used here are the same
as those used for baseline conditions throughout the Hudson River PCBs RI/FS in the
Revised Human Health Risk Assessment, the Revised Baseline Ecological Risk
Assessment, the FS, and the ROD Responsiveness Summary reports.
Hudson River PCBs Superfund Site
Engineering Performance Standards
27
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The fate and transport and bioaccumulation of PCBs in the upper river were predicted as
Tri+ PCB concentrations using the HUDTOX and FISHRAND models. The Tri+ PCB
group includes the PCB compounds that are most toxic to fish, wildlife, and humans and
is considered to capture the majority of toxicity associated with PCB compounds. PCB
contamination in fish tissue from the Hudson River has been shown to consist almost
exclusively of Tri+ PCBs, with average values ranging from 98% to nearly 100%
(USEPA, 2002).
The Revised HHRA and ERA (USEPA, 2000a and 2000e, respectively) have shown
ingestion of fish to account for most of the risk to human and ecological receptors;
therefore, the use of Tri+ PCBs for risk assessment modeling requires no revisions for
comparison to available toxicological literature for PCB effects expressed as total PCBs
or Aroclors.
Table 2-4 presents annual species-weighted fish fillet
Tri+ PCB concentrations in the Upper Hudson River, as
compared to the risk-based remediation goal (RG) for the
protection of human health of 0.05 mg/kg PCBs in fish
fillet. That RG is based on non-cancer hazard indices for
the RME adult fish consumption rate of one half-pound
meal per week and is protective of cancer risks as well. Other target concentrations
presented are 0.2 mg/kg PCBs in fish fillet, protective at a fish consumption rate of one
half-pound meal per month, and 0.4 mg/kg PCBs in fish fillet, protective of the central
tendency (CT) or average angler who consumes one half-pound meal every two months.
The time to reach human health fish target concentrations of 0.2 mg/kg Tri+ PCB and 0.4
mg/kg Tri+ PCB in the Upper Hudson River was shorter for all resuspension scenarios as
compared to monitored natural attenuation in the upper river as a whole and in River
Sections 1 and 2 (Table 2-5). In River Section 3, all active remediation scenarios
achieved the RG of 0.05 mg/kg Tri+ PCB prior to MNA. The greatest differences seen in
the time to achieve fish target concentrations between the active remediation scenarios
and MNA were seen in River Section 1, where the MNA scenarios took up to 17 years
longer to achieve some target concentrations, while the smallest differences were seen
between scenarios in River Section 3.
Using fish fillet concentrations based upon the three resuspension scenarios {i.e., no
resuspension, 350 ng/L, and 600 g/day), human health fish consumption cancer risks and
non-cancer hazards show at least a 50% reduction in the upper river as a whole, Section 1
(River Mile 189), and Section 2 (River Mile 184) compared to monitored natural
attenuation for both RME and average exposures (Tables 2-6 and 2-7). Risk reductions in
Section 3 were seen for the no resuspension and 600 g/day scenarios as compared to
monitored natural attenuation, but not for the 350 ng/L Total PCB scenario.
Based on site-specific angler surveys, the Human Health Risk Assessment (USEPA,
2000a) determined that Mid-Hudson River anglers have a different diet than anglers in
the upper river:
The risk-based remediation
goal (RG) for the protection
of human health is 0.05
mg/kg PCBs in fish fillet.
Hudson River PCBs Superfund Site
Engineering Performance Standards
28
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Brown bullhead - 53%
Largemouth bass - 15%
Yellow perch - 1.4%
White perch - 7.6%
Striped bass - 23%
Striped bass concentrations were not modeled for resuspension scenarios and therefore
human health cancer risks and non-cancer hazards for Mid-Hudson River anglers could
not be calculated. To provide an estimate of relative risks amongst the resuspension
scenarios, angler intake was calculated using fish concentrations from the FISHRAND
model. Striped bass intake was proportionally divided between the remaining fish species
(i.e., 69% brown bullhead, 19% largemouth bass, 2.0% yellow perch, and 10% white
perch) and white perch concentrations from the FISHRAND Model were used in the
absence of Farley Model data.
Calculated fish exposure concentrations were used only for comparison between
alternatives and do not represent predicted intake concentrations based on mid-river
angler consumption patterns. As expected, fewer differences were seen between the
resuspension scenarios in the lower river than in the upper river. Long-term cancer risks
and non-cancer hazards differed by a maximum of 32%. The no resuspension and 600
g/day Total PCBs scenarios showed the greatest risk reductions as compared to
monitored natural attenuation scenario. The 350 ng/L Total PCBs showed lower and
sometimes no reductions in risk, owing to elevated concentrations of PCBs predicted in
fish tissues for several years following dredging operations (Figure 2-9).
Risks to ecological receptors, as represented by the river otter, were evaluated by
examining largemouth bass whole fish PCB concentrations. In the Upper Hudson River
the lowest-observed-adverse-effect-level (LOAEL) target levels were reached within the
modeling timeframe for the upper river as a whole and in Section 3 for all scenarios
(Table 2-8). In the upper river as a whole, all resuspension scenarios reached the LOAEL
target level of 0.3 PCBs mg/kg 17 years prior to the MNA scenario (Table 2-9).
Ecological target levels were not reached within the modeling timeframe for Sections 1
and 2 of the river. In Section 3, all scenarios reached the LOAEL target level within five
years of one another.
Largemouth bass PCB concentrations in the Lower Hudson River were lower under all
resuspension scenarios than under the MNA scenario (Table 2-10). The LOAEL PCB
target concentration in largemouth bass was reached 4 to 11 years sooner under the
various resuspension scenarios than under MNA in various sections of the lower river
(Table 2-11).
Resuspension may temporarily increase PCB
concentrations locally, resulting in slight increases
in fish PCB concentrations. However, human health
non-cancer hazards and cancer risks and ecological
Conclusion:
Human health and
environmental impacts from
dredging are predicted to be
Hudson River PCBs Superfund Site
Engineering Performance Standards
29
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
risks were calculated to be well below those under the MNA scenario. Minor differences
were seen between the various resuspension scenarios, indicating the human health and
environmental impacts from dredging are predicted to be minimal, particularly since
levels of resuspension approaching the performance criteria are expected to occur on an
intermittent, rather than continuing basis.
2.6.2 Accidental Release Short-Term Impacts
HUDTOX was used to model an accidental release scenario to demonstrate the short-
term and long-term impacts to the public water intakes downstream of the incident. The
following accidental release scenario was analyzed:
A hopper barge containing 870 tons of silty sand (barge capacity is 1000 tons,
with 87% sediment and 13% water) that has been removed by mechanical
dredging from River Section 2 is damaged and releases the entire load in the area
just above Lock 1.
The contents fall in a mound and no effort is made to remove or contain the
material.
Over a period of one week, the entire load is swept downstream.
The background concentrations are at the 600 g/day Total PCB flux at the River
Section 3 monitoring location.
For this scenario, an additional release of 113,000 kg/day suspended solids is
anticipated, with a baseline condition of 20,000 kg/day for a one-week period
(from July 1 through 7, 2011).
This scenario is quite conservative in that the average concentration from River Section 2
is higher than in the TI Pool. This is because areas with mass per unit area greater than 10
g/m2 are targeted in this river section, whereas in the TI Pool, areas greater than 3 g/m2
are targeted. The hopper barge was used because it has a larger capacity than the deck
barge (200 tons) that was also proposed in the FS. The location of the accident is just
above the public water intakes at Halfmoon and Waterford, minimizing the opportunity
for reductions to the water column concentration resulting from settling and dilution.
Because a mechanical dredge is assumed to have removed the sediment, nearly the entire
weight of the release would be attributed to sediment, with little dilution with water. The
already elevated water column concentrations result in water column concentrations at
the public water intakes greater than the MCL. This scenario is also conservative from
the realistic standpoint that a spill of this magnitude would almost certainly be contained
within hours of occurrence.
HUDTOX provided the whole water, particulate-bound, and dissolved-phase PCB
concentrations in the water column. The model predicted that the accidental release
scenario results in a short-term increase of the whole water Total PCBs above the MCL
in the water column at Waterford (Figure 2-10); however, the highest dissolved phase
Total PCB concentration was less than 350 ng/L (Figure 2-10). Because HUDTOX
Hudson River PCBs Superfund Site
Engineering Performance Standards
30
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
assumed instantaneous attainment of PCB equilibrium between the dissolved and
suspended phases, the dissolved-phase PCB concentrations are overestimated, providing
an additional conservative assumption.
While the Total PCB concentration entering the public water intake would be in excess of
the federal and state MCL, it is likely that the concentration in the influent would be
greatly reduced by minimal treatment because approximately 850 ng/L of the total 1,150
ng/L Total PCB peak concentration would be attributed to the suspended phase.
Assuming that the bulk of the contaminated suspended solids would be removed by
filtration, the delivered concentration without further treatment would be closer to the
dissolved-phase PCB concentration of 300 ng/L. Thus, the water output from the plant
would still meet the federal MCL of 500 ng/L.
As previously noted, the dissolved phase PCB concentrations estimated by HUDTOX are
already biased high. The dissolved phase PCB concentrations would probably be further
reduced by activated carbon treatment, which is currently implemented at the Waterford
public water intake. This analysis suggests that the concentration reaching the public
would be substantially less than the MCL even in the event of an accidental release in the
vicinity of the intakes as described in the hypothetical accidental release scenario.
While this analysis suggests that the planned operations are unlikely to impact the public
water supplies in the event of an accident, further consideration on the protection of
public water supplies and the requisite monitoring will be given in the development of a
community health and safety plan (CHSP).
2.7 Near-Field Modeling
Two models (CSTR-Chem and TSS-Chem) were developed to estimate the conditions
within 1 mile downstream of the dredge head. These near-field models were used to
estimate the suspended solids and Total PCB plumes resulting from resuspension of
solids. The models were useful in identifying the most appropriate location for the
placement of water column monitoring stations in the near-field and provided an estimate
of solids transported into the far-field. In addition, the TSS-Chem model was used to
estimate the effects of settled material on sediment concentrations within the near-field.
2.7.1 CSTR-Chem and TSS-Chem
CSTR-Chem and TSS-Chem models were developed and utilized for the near-field
modeling effort to estimate the transport and concentration of suspended solids and Total
PCBs from the dredge head to the far-field region (approximately one mile downstream
of the dredge head).
Hudson River PCBs Superfund Site
Engineering Performance Standards
31
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.7.1.1
CSTR-Chem
CSTR-Chem is used to model the area immediately around the dredge. The model is
based on an ideal reactor configuration consisting of a continuous stirred tank reactor
(hence CSTR). This construct represents a means to simplify the mathematical modeling
of constituent concentrations in the immediate vicinity of the dredge head. CSTR-Chem
assumes that a constant flow influent with a known constant concentration (i.e., upstream
river water) is instantaneously mixed as it enters a confined, well-mixed tank (e.g., the
region immediately around the dredge head). Physical and chemical reactions occur while
the water is within the ideal tank and the tank effluent is at the same flow as the influent
and at the uniform concentration within the tank.
The CSTR concept is most appropriate to the analysis of dredging operations because
turbulence in the area of the dredge, coupled with ambient flows, may be assumed to
produce mixed conditions similar to that in an ideal tank reactor. A complete discussion
of the CSTR-Chem and TSS-Chem model development is presented in Attachment D.
The input for the CSTR-Chem is model is the subsequent resuspension rate. Since solids
will settle within this area, the solids flux out will not be equal to the resuspension
production rate of solids. The rate at which solids exit the immediate dredge area is
termed the source strength. The source strength represents the solids available for
downstream transport and is the input for the TSS-Chem model. However, since the TSS-
Chem model simulates a point source and CSTR-Chem has a non-zero width, the two
models cannot be directly linked. Nevertheless, CSTR-Chem can still be used to provide
for input to TSS-Chem, particularly with regard to the dissolved PCB concentration and
the silt fraction.
2.7.1.2 TSS-Chem
The TSS-Chem model has two components:
A Gaussian plume transport model that describes the dispersion and settling of the
particles downstream
A geochemical component that uses two-phase partitioning of PCBs from solids
into the dissolved phase taking into account a kinetic desorption rate
TSS-Chem utilizes the same solids transport equations for a mechanical dredge as
DREDGE (Kuo and Hayes, 1991), outlined in Appendix E.6 of the FS and the White
Paper - Resuspension of PCBs During Dredging (USEPA, 2002a). The TSS-Chem
model was used to estimate PCB water column conditions downstream of the dredge
across the width of the river up to a distance of one mile. TSS-Chem is useful for the
near-field downstream transport of solids and PCBs but is inadequate in estimating the
net contribution of solids and dissolved and suspended phase PCBs to the water column
in the immediate vicinity of the dredging operations (i.e., relating the resuspension
production rate to the source strength). For this purpose, the CSTR-Chem model was
developed.
Hudson River PCBs Superfund Site
Engineering Performance Standards
32
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.7.1.3 Desorption Rate Input to the Models
One of the important input parameters in the CSTR-Chem and TSS-Chem models is the
desorption rate constant. The conclusions drawn from CSTR-Chem and TSS-Chem
models depend on an accurate desorption rate constant assumption. An extensive
literature review on the PCB desorption rate constant was conducted for the
Resuspension Standard and is presented in Attachment C. Due to lack of knowledge on
the amount of "labile" (fast) and "non-labile" (slow) fractions in the dredged material,
only fast desorption rate constants are considered in this study in order to provide a
conservative (upper bound) estimate of the amount of PCBs that partition into the
dissolved phase. The rate of desorption used for TSS-Chem and CSTR-Chem is 0.2 hr"1.
This desorption rate was applied to the difference between the PCB concentration of the
suspended sediments and the equilibrium concentration by allowing more PCBs to
remain in the water column with the existing soluble PCB concentration. Attachment D
contains additional detail on the two-phase partitioning equations.
2.7.1.4 Applicability of the Models
Applicability of the CSTR-Chem model depends upon the presence of near-field
conditions that can reasonably be represented as well mixed; it is important that the
diameter of the cylindrical area that is approximated as a CSTR should reflect the extent
to which well-mixed conditions exist. For the purposes of this analysis, a CSTR width of
10 meters (m) is used. Buckets that may be used in the Hudson River project are
generally 2 to 3 m in diameter closed and somewhat larger when open. It was assumed
that velocities induced by bucket movement could extend across most of a 10-m width
used in this analysis.
The CSTR-Chem results suggest that under transient partitioning conditions, which are
expected within the CSTR, the PCB releases from dredging operations will generally be
less than 1% dissolved. The model results also suggest there is no significant loss of silt
particles from the settling within the CSTR. The results of the CSTR-Chem model were
used to develop the assumptions made concerning the source strength of the TSS-Chem
model. The results indicated that:
When the dissolved fractions estimated by the CSTR-Chem were input into the
TSS-Chem, the results did not significantly vary from runs that had no initial
dissolved phase.
The silt fraction within the sediments is the only parameter that significantly
affected the TSS-Chem PCB flux at one mile.
Incorporating these model observations, the TSS-Chem model was used to simulate the
near-field dredging operations, from just beyond the dredge head to a one-mile distance
downstream. Attachment D contains a more detailed discussion on the relationship
between the TSS-Chem model assumptions and the CSTR-Chem.
Hudson River PCBs Superfund Site
Engineering Performance Standards
33
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.7.2 Near-field Model Results
Near-field modeling was performed to address the following issues:
How much PCBs may be released during dredging?
How far from the dredge should water quality monitoring be conducted?
At what rate will resuspended sediment settle out of the water column?
How far downstream will the settling occur?
How much material will be deposited and what is the impact on the deposition
areas outside of the targeted (dredged) areas?
2.7.2.1 Solids and PCB Load HUDTOX Inputs
TSS-Chem was used to estimate solids and PCB loads for input to the HUDTOX model.
Conditions at one mile were taken for input to the HUDTOX model, recognizing the
difference in model scales. As outlined in Appendix E.6 of the FS (USEPA, 2000b) and
White Paper - Resuspension of PCBs During Dredging (USEPA, 2002a), the average
resuspension rate is based on a combination of field data from other sites and a
resuspension model. The downstream transport rates (source strengths) only apply to silts
and finer particles within the sediment (65% of cohesive and 20% of non-cohesive
sediments for the Hudson River). The use of only silts does not significantly affect the
PCB flux estimates because the silt resuspension rate, essentially equal to the silt source
strength, is the driving source term for the PCB flux downstream
The production rates for the average source strength calculations were based on a total of
five full production dredging seasons, using the estimated amount of sediment removal
necessary and the time limitations involved. Each source strength estimate was run
through TSS-Chem to calculate the resulting flux and concentration increases at one mile.
Table 2-12 contains the production rates, source strengths, and results are shown in. The
average source strength was estimated at approximately 0.7 to 0.9 kg/s. For the various
river sections these source strengths corresponded to PCB fluxes of approximately 80 to
210 g/day at one mile. The variation in the PCB fluxes for the different river sections is
mainly caused by the different sediment concentrations. The highest flux is from
dredging activities in River Section 2, which has a sediment concentration roughly 2.2
times greater than River Section 1.
2.7.2.2 Solids Transport Simulation
The TSS-Chem model was used to simulate the solids transport in the water column due
to dredging operations up to one mile downstream. Simulations were performed for the
300 g/day, 600 g/day, 350 ng/L and 500 ng/L scenarios. The results suggest that the water
column at one mile downstream of the dredge head has a significant amount of dissolved
phase, but the suspended solids phase is still dominant (Figure 2-11). The fraction of
Hudson River PCBs Superfund Site
Engineering Performance Standards
34
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
dissolved phase Total PCB is greater for scenarios with lower amounts of solids
introduced to the water column (i.e., lower resuspension rates and source strengths)
(Table 2-13).
For example, for the 300 g/day scenario, which has the lowest SS flux range from 0.3 to
1.3 kg/s at the dredge head, the TSS-Chem predicted that the fraction of dissolved phase
Total PCBs one mile downstream of the dredge head ranges from 0.2 to 0.4 (Table 2-13).
The 500 ng/L scenario has the highest amount of solids introduced to the water column
(ranges from 3 to 9 kg/s at the dredge head). For this scenario, the TSS-Chem model
results showed that the fraction of dissolved phase Total PCB in the water column ranges
only from 0.05 to 0.1.
According to the TSS-Chem model results, the suspended solids concentration decreases
and the width of plume increases as the solids are transported downstream. The
suspended solids concentration at 300 m downstream is about one-quarter to one-third of
the concentration at 50 m downstream, while the width of the plume at 300 m
downstream is about twice the plume width at 50 m downstream. The greater width of the
plume at 300 m suggests that this location may be easier to monitor using a stationary,
continuous reading suspended solids sensor. It is also likely that by this distance
downstream, water column concentrations of suspended solids will be more
homogeneous. As a result, in an attempt to balance between the wider, more
homogeneous plume conditions farther downstream and the easier identification of the
center of the plume, 300 m downstream of the dredge head was chosen as the location of
a primary near-field monitoring station.
The time that the particles remain suspended is primarily a function of the sediment type.
Generally, silt particles will remain suspended longer than coarse particles. In the near-
field models, the rate at which particles fall through the water column is determined by
the particle settling velocity. Different settling velocities are defined for fine and coarse
particles in the models. Attachment D contains a summary of settling velocities from
various studies. For most of the studies, Stokes' Law was the theoretical basis for
estimating the settling velocity of sand particles. This approach is appropriate for discrete
particles that do not aggregate and was applied to the coarse material in the near-field
models.
Stokes' Law only applies to discrete particles settling and does not account for
flocculation during settling. Flocculation increases the rate at which silts settle from the
water column, but the rate of flocculation depends on site-specific conditions and
sediment properties. Therefore, silt settling velocities presented in QEA's report (1999)
for Hudson River sediments were used in the near-field models, since these values were
derived for Hudson River conditions and included the effects of flocculation.
The TSS-Chem results indicate that with a flow rate of 4,000 cfs, approximately 30 m
downstream from the dredge head most of the coarse material has settled to the bottom of
the river. At this distance, the coarse material is less than 0.1% of the net suspended
solids from dredging. Since the coarse material settles much faster than the silts, it does
Hudson River PCBs Superfund Site
Engineering Performance Standards
35
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
not contribute significantly to PCB loads and concentrations at one mile. The results also
suggest that there is a significant amount of settling within one mile downstream of the
dredge head. The amount of Total PCBs being introduced to the water column from the
dredge head is reduced by approximately 80% in River Section 1 and approximately 70%
for River Sections 2 and 3 at one mile downstream of the dredge head (Table 2-13). For
example, in River Section 1, when the amount of Total PCBs added to the water column
due to dredging is 1,700 g/day, the load at one mile is approximately 400 g/day.
2.7.3 PCB Deposition Immediately Downstream at the Dredge Operations
If the suspended solids that settle onto the riverbed
during transport downstream are contaminated,
PCB mass and concentration will be added to the
surrounding downstream areas. Using the modeled
suspended solids concentrations in the water
column downstream of the dredge, with the
associated PCB concentration on the suspended
solids, it is possible to estimate the increase in PCB mass in these areas. The increase in
mass per unit area and the length-weighted average concentration of the top 6-in
bioavailable layer were used to measure the effect of the settled material. Since these
areas are outside of the target areas, the settled particles are not scheduled for removal.
The spatial distribution of the settled contamination
will vary according to the shape of the target area and
the rate of dredging. For this estimate, the target area
is assumed to be 5 acres, 200 ft across, and
approximately 1,100 ft long, because the areas of
contamination are typically located in the shoals of
the river and are narrow. From the FS, the time needed to dredge a 5-acre area with 1-m
depth of contamination would take 15 days, operating 14 hours per day. It is assumed that
the dredge will move in 50 ft increments across and down the target area. With these
assumptions, the dredge will relocate approximately every two hours. To simulate the
deposition of settled material, the amount of PCB mass per unit area, the mass of the
settled material, and the thickness of the settled material that is deposited in two hours
downstream at each modeled location is added on a grid as the dredge moves across and
down the area.
The TSS-Chem results for each river section and action levels were used to estimate the
additional mass per unit area and length-weighted average concentration approximately
two acres downstream of the target area. The remediation could operate continuously at
the Evaluation Level of 300 g/day and the Control Level of 600 g/day, but not the
Control Level of 350 ng/L. The results are shown in Table 2-14.
The ROD defines 1 mg/kg as the acceptable residual
concentration; the length-weighted area concentrations
were calculated assuming that the PCB concentration in
If the suspended solids that settle
onto the riverbed during transport
downstream are contaminated,
PCB mass and concentration will
be added to the surrounding
downstream areas.
Spatial distribution of the
settled contamination will vary
according to the shape of the
target area and the rate of
dredging.
The ROD defines 1 mg/kg
as the acceptable residual
concentration.
Hudson River PCBs Superfund Site
Engineering Performance Standards
36
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
the sediment underlying the settled material is 1 mg/kg. In the two acres below the target
area in River Section 2, for example, the concentrations range from 2 to 9 mg/kg.
These increases suggest that dredging should proceed from upstream to downstream if no
silt barriers are in place, so that the dredge inside the target areas can capture settled
material. Also, silt barriers may be needed to prevent the spread of contamination to areas
downstream of the target areas have already been dredged or are not selected for
remediation, as this settled material is likely to be unconsolidated and may be easily
resuspended under higher flow conditions.
2.8 Relationship Among the Resuspension Production, Release, and
Export Rates
During dredging operations, it is necessary to specify the near-field load to the water
column that would yield the targeted export rates {i.e., resuspension criteria) at the far-
field stations. In order to estimate these loads, computer models were utilized to provide a
relationship between the far-field and the near-field dredging-induced PCB transport and
loss. The TSS-Chem and HUDTOX models were used to represent and link the
resuspension production (at the dredge-head), release and export rates. The resuspension
release rate (and source strength) in the region from the dredge to a distance of one mile
is represented by the TSS-Chem model. The resuspension export rate in the region
beyond one mile is represented by HUDTOX.
The TSS-Chem and HUDTOX models were used to examine the:
Amount of sediment being suspended in the water column at the dredge head.
Suspended solids and Total PCB flux at one mile downstream of the dredge head.
Total PCB flux at the far-field monitoring stations for the 300 g/day, 600 g/day,
and 350 ng/L scenarios.
Table 2-12 shows the resuspension production, release, and export rates for the
simulations. Because HUDTOX predicted different rates of export for different reaches
of the river given the same PCB release rate, the TSS-Chem model was run under
different conditions so as to yield a consistent output from HUDTOX (e.g., 600 g/day,
350 ng/L) for all river sections.
2.8.1 300 g/day Export Rate Scenario
From the results, it was predicted that in order to
create an export rate of 300 g/day of Total PCBs at
the TI Dam, the amount of Total PCBs in bulk
sediments that needs to be suspended is
approximately 900 to 1,700 g/day, depending on
the location of the dredge-head to the monitoring
To create a 300 g/day export rate
of Total PCBs at the TI Dam,
approx. 900-1,700 g/day Total
PCBs would need to be
suspended in bulk sediment,
depending on distance between
dredge head and monitoring
station.
Hudson River PCBs Superfund Site
Engineering Performance Standards
37
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
stations. The farther the dredge is from the far-field monitoring location, the greater the
amount of solids and PCBs that would need to be suspended into the water column
(Table 2-12).
Resuspension production rates that create an export rate of 300 g/day are on the order of
2% to 3% of the removal rate of Total PCBs via dredging. That means that in River
Sections 2 and 3, the following amounts of Total PCBs in bulk sediment would need to
be suspended from the water column are as follows:
River Section 2: 1,000 g/day Total PCBs
River Section 3: approximately 1300 g/day when the dredge head is farther away
from the far-field monitoring location; around 1,000 g/day when the dredge head
moves closer (downstream) to the monitoring station
Overall, the Total PCB resuspension export fraction relative to the PCB resuspension
production rate for the 300 g/day scenario is estimated to range from 0.17 to 0.34.
2.8.2 600 g/day Export Rate Scenario
To obtain an export rate of 600 g/day Total PCBs, the amounts of Total PCB mass that
would need to be suspended into the water column in the three river sections are as
follows:
River Section 1: from 3,000 to 4,000 g/day (on the order 5% to 6% of the Total
PCB removal rate via dredging)
River Section 2: approximately 2,000 g/day (approximately 2% of the Total PCB
removal rate via dredging)
River Section 3: approximately 2,000 to 3,000 g/day (on the order of 2% of the
Total PCB removal rate by dredging)
Overall, the Total PCB export fraction relative to the PCB resuspension production rate
for the 600 g/day scenario is estimated to range from 0.17 to 0.31, similar to that for the
300 g/day scenario.
2.8.3 350 ng/L Total PCB Concentration Scenario
The 350 ng/L Total PCB concentration at the far-field monitoring stations scenario was
also simulated. The Total PCB fluxes at the TI Dam, Schuylerville and Waterford that
would represent the 350 ng/L are 1,200, 2,000, and 2,300 g/day, respectively. The
resuspension production rates, i.e., the g/day volume of Total PCB mass that would need
to be suspended to the water column to create an export rate of 350 ng/L Total PCB
concentrations, are as follows:
Hudson River PCBs Superfund Site
Engineering Performance Standards
38
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
At the TI Dam: approximately 6,000 to 7,600 g/day (approximately 10% to 13%
of the Total PCB removal rate via dredging)
River Section 2: approximately 7,000 to 8,300 g/day (approximately 6% to 7% of
the Total PCB removal rate via dredging)
River Section 3: approximately 8,400 to 11,000 g/day (approximately 15% to
19% of Total PCB removal rate via dredging)
These resuspension production rates are approximately 19% to 24% of the Total PCB
removal rate via dredging. The Total PCB export fraction for this scenario ranges from
0.16 to 0.28.
2.8.4 500 ng/L Total PCB Concentration Scenario
The 500 ng/L Total PCB condition was only
simulated by TSS-Chem model, without a
subsequent HUDTOX model forecast. As a result,
the Total PCB fluxes at the far-field monitoring
stations were extrapolated based on the 500 ng/L
input conditions and the results of the previous
HUDTOX simulations. The TSS-Chem results for the 500 ng/L scenario suggest that the
Total PCB export fraction of the resuspension production rate ranges from 0.16 to 0.29
(i.e., 16% to 29%) of the PCB mass removed would have to be spilled to yield a 500 ng/L
condition in the river). To obtain 500 ng/L Total PCB concentration at the far-field
monitoring station, g/day Total PCB mass that would need to be suspended to the water
column would be as follows:
River Section 1: approximately 10,000 to 13,000 g/day (approximately 17% to
23%) of the Total PCBs removal rate via dredging).
River Section 2: approximately 9,300 to 11,000 g/day (approximately 8% to 9%
of the Total PCBs removal rate via dredging)
River Section 3, approximately 13,000 to 16,600 g/day (approximately 23% to
29%) of the Total PCBs removal rate via dredging)
These model calculations yield an important
conclusion concerning criteria developed for the
Resuspension Standard. While the model analysis
of the concentrations and loads that comprise the
standard show relatively little long-term impact on
downstream receptors and conditions, the amount
of sediment spillage required to attain these levels
is quite large. Spillage at these levels is unlikely and certainly well beyond what is
expected for standard environmental dredging practices. Based on these analyses,
compliance with the Resuspension Standard appears to be attainable, including the lowest
action criteria.
Modeling results suggest that
from 16% to 29% of the PCB
mass removed during dredging
would have to be spilled to yield
a 500 ng/L condition in the river.
Sediment spillage at levels that
would be required in order to have
long-term impact on downstream
receptors and conditions is unlikely
and well beyond what is expected
for standard environmental
dredging practices
Hudson River PCBs Superfund Site
Engineering Performance Standards
39
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.9 Review of Applicable or Relevant and Appropriate Requirements
(ARARs)
The evaluation of potentially applicable federal and state water quality standards for the
purpose of the performance standard development was based on work previously done
for the ROD) for the Hudson River PCBs Site (USEPA, 2001). In the ROD, seven
chemical-specific ARARs for PCBs were identified:
500 ng/L Federal MCL [40 CFR ง 141.61] and NYS MCL [10 NYCRR, Chapter I,
Part 5, Section 5.1.52, Table 3]
90 ng/L NYS standard for protection of human health and drinking water sources
[6 NYCRR Parts 700 through 706]
30 ng/L Federal Water Quality Criterion (FWQC) criteria continuous
concentration (CCC) for saltwater [Aroclor-specific 40 CFR ง 131.36]
14 ng/L Federal Water Quality Criterion (FWQC) criteria continuous
concentration (CCC) for freshwater [Aroclor-specific 40 CFR ง 131.36]
1 ng/L Federal Ambient Water Quality Criterion for Navigable Waters [40 CFR ง
129.105(a)(4)]
0.12 ng/L NYS standard for protection of wildlife [6 NYCRR Parts 700 through
706]
0.001 ng/L NYS standard for protection of human consumers of fish [6 NYCRR Parts
700 through 706]
Of these criteria, USEPA waived the three lowest
concentration standards (0.001 ng/L to 1 ng/L) due to lฐn9 as the water column
technical impracticality (USEPA, 2001), as it is are below
. . .F , , , the 500 ng/L federal and state
technically impractical to reach these concentration MCL protection ofhuman heam
levels in the Hudson River with the continuing input will be achieved.
from the upstream sources. As long as the water
column Total PCB concentrations are below the federal and state MCL (500 ng/L),
protection ofhuman health will be achieved. Only the 500 ng/L total PCB standard is not
regularly exceeded by the main stem Upper Hudson River stations downstream of Rogers
Island under existing (baseline) conditions; therefore, the other ARARs were not applied
in the development of the Resuspension Standard. No other chemical-specific criteria
were identified as ARARs or To-Be-Considered criteria (TBCs) in the ROD or the
RRI/FS Feasibility Study (USEPA, 2000b).
Additional surface water quality criteria were considered for parameters that may be
impacted by the remediation. These parameters are pH, dissolved oxygen (DO), and
turbidity. NYS guidelines [6 NYCRR Parts 700 through 706] set the following standards:
pH 6.5 to 8.5 for Class A surface water
DO Not less than a daily average of 6 mg/L for trout bearing waters; not less
than 5 mg/L for non-trout bearing waters; and
Hudson River PCBs Superfund Site
Engineering Performance Standards
40
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Turbidity No criteria for surface water
Specific resuspension criteria have not been established for these water quality
parameters. The water quality parameter data will be used for comparison to the
continuously monitored data at both the near-field and far-field stations. These standards
may be used as resuspension criteria in Phase 2, if appropriate.
2.10 Summary of Supporting Analyses
Numerous analyses were completed in support of this performance standard. Review of
case studies have provided examples for the way the issue of resuspension of
contaminated material has been handled at other sites leading to development of the
elements of this standard, including resuspension criteria and monitoring and engineering
contingencies. The calculations described suggest that the standard is achievable and, if
complied with, will be protective of the environment and human health.
The context for these analyses will be evident in Section 3, Discussion of Rationale. A
brief synopsis of the supporting analyses follows.
2.10.1 Turbidity and Suspended Solids at Other Sites
A surrogate measurement of suspended solids
concentrations such as turbidity may become an
important real-time indicator of PCB concentration
levels, if it is proven in Phase 1 that the primary
mechanism of contaminant release from the
remediation is resuspension of sediment. Turbidity
measurements are instantaneous, whereas analyses
for suspended solids or PCBs are more time-
consuming and limit the time available to warn downstream water supplies in the event of
an exceedance of the standard.
Case studies were reviewed to provide an indication of turbidity and suspended solids
concentrations in the water column and the thresholds that were established at these sites
to limit resuspension. Because suspended solids measurements are needed for
comparison to resuspension criteria, a correlation must be developed between suspended
solids and a surrogate before a surrogate measurement could be used for this purpose.
Review of case studies and literature indicates that such correlations are site-specific,
have been established at other sites, and could potentially be developed for the Hudson
River. The case studies described the configuration of monitors relative to the remedial
operations. This information was considered when specifying the near-field monitoring
locations required by the standard.
Turbidity may become an
important real-time indicator of
PCB concentration levels, if
Phase Iremediation indicate that
the primary mechanism of
contaminant release is
resuspension of sediment.
Hudson River PCBs Superfund Site
Engineering Performance Standards
41
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
2.10.2 PCB Releases at Other Sites
The case studies also provided information with which to calculate the amount of PCB
released from other dredging sites. The rate of loss provides another indication of what a
reasonable load-based resuspension criterion would be. These estimates of loss can also
be used to determine the average increase in water column concentration during the
remediation. Estimated rates of contaminant loss from other sites are 0.13%, 0.36%, and
2.2%.
2.10.3 Hudson River Water Column Concentration Analysis
Approximately five years of baseline water column PCB concentration data are available.
Although there are concerns over the quality of these data due to the sampling methods
and analytical methods used, estimates of the average expected water column PCB can be
made. These values can be compared directly to the PCB concentration-based
resuspension criteria to indicate whether, in some months, the PCB concentration may
routinely approach the standard, even without the added impact of the suspension. The
results indicate that the average PCB water column concentrations will be less than the
concentration-based resuspension criteria, although in some months it is expected that the
criteria would be exceeded on occasion.
2.10.4 Resuspension Sensitivity Analysis
The resuspension sensitivity analysis was built on the Hudson River water column
concentration analyses by adding the estimated increase in concentration for a given
increase in PCB load on to the estimated baseline PCB water column concentrations. This
analysis suggests that the load-based resuspension criteria will not routinely elevate the
water column concentration over the concentration-based criteria. The results indicate
that the average PCB water column concentrations during dredging will be less than the
concentration-based resuspension criteria, although in some months it is expected that the
criteria would be exceeded on occasion. Variability in the water column concentrations
may on occasion result in exceedance of the load-based criteria, although the true
dredging-related releases are below the 300 g/day and 600 g/day Total PCB limits.
2.10.5 Dissolved-Phase Releases
Concerns were raised during the public comment
period for the Hudson River ROD that dissolved-
phase PCB concentrations could be significant during
remediation of PCB-contaminated sediment, and that
a release of this kind could not be detected by a surrogate measure such as suspended
solids or turbidity. The calculations described in subsection 2.5 indicate that a release of
this kind would not be possible without an associated suspended solids release, because
A dissolved-phase PCB release
undetected by a surrogate
measure such as turbidity or
suspended solids is not possible.
Hudson River PCBs Superfund Site
Engineering Performance Standards
42
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
the bulk of the PCB contamination is bound to the sediment and there is not a sufficient
amount of PCBs dissolved in the pore water to cause a substantial release.
2.10.6 Far-Field Modeling
The impacts of allowing the remediation to continue at the levels indicated by the
resuspension criteria were determined through model simulation, using the fate, transport,
and bioaccumulation models developed during the Reassessment RI/FS phase for this
purpose. The results indicate that operation at the total PCB load-based resuspension
criteria, which are the only criteria at which the remediation could operate for extended
periods of time, will result in short-term impacts to the environment during the
remediation, but will have little impact on the fish tissue concentrations post-dredging.
Analysis of a hypothetical accidental release scenario in the vicinity of the Upper Hudson
River public water intakes (subsection 2.6.2) indicated that although the concentrations
entering the intake would be greater than the MCL, minimal water treatment would be
sufficient to reduce the concentrations below the MCL.
2.10.7 Near-Field Modeling
Models of surface water concentrations in the vicinity of the dredge were developed to:
Determine the amount of PCBs released from the dredging operation.
Predict the downstream water column concentrations.
Calculate the area in which the resuspended material would settle and the increase
in PCB concentration in that area.
Identify the appropriate locations for near-field monitoring.
The modeling indicated that the PCBs released by the dredge would be largely suspended
phase. The amount of dissolved PCBs increased to a limited extent as the plume traveled
downstream, but this process is slow because of the small coefficient of desorption. The
relative amount of dissolved-phase to suspended-phase PCBs increases as the solids
settle. Settling of contaminated material downstream of the dredge has the potential to
raise surface concentrations substantially. This would be of concern if the area were not
subsequently dredged, and may indicate the need for containment if this condition were
verified. The results of these models suggest both the locations of the far-field and near-
field monitoring points relative to the remedial operations and the suspended solids near-
field resuspension criteria.
2.10.8 Relationship Among the Resuspension Production, Release, and Export
Rates
The Total PCB load-based resuspension criteria were based on engineering judgment and
the balance of several factors, including the:
Hudson River PCBs Superfund Site
Engineering Performance Standards
43
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Best engineering estimate of resuspension production and export.
Minimum detectable PCB load increase.
Load defined by the water column concentration criteria.
Impact of load on fish tissue recovery.
Delivery of Total PCBs and Tri+ PCBs to the Lower Hudson.
Subsection 2.8 contains a detailed description of the selection process for the load-based
criteria. A series of models was used to examine the relationship among the resuspension
production, release, and export rates. The model calculations yield an important
conclusion concerning the relationship between the resuspension production rate and the
performance standard criteria. While the model analysis of the concentrations and loads
that comprise the standards show relatively little long-term impact on downstream
receptors and conditions, the amount of sediment spillage required to attain these levels is
quite large. Spillage at these levels is certainly well beyond what is likely, given standard
environmental dredging practices.
2.10.9 Review of Applicable or Relevant and Appropriate Requirements (ARARs)
Federal and state surface water quality guidelines were reviewed to determine if these
regulations would provide a concentration level that was achievable during the
remediation and protective of human health. The federal and state MCL of 500 ng/L total
PCBs met these criteria.
Hudson River PCBs Superfund Site
Engineering Performance Standards
44
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
3.0 Rationale for the Standard
3.1 Development of the Basic Goals and Resuspension Criteria
The performance standard for PCB losses due to resuspension is unique among the
engineering performance standards in that the basic criteria are not defined in the ROD.
Unlike the Production and Residuals Standards that have basic goals defined in the ROD
{i.e., approximately 2.65M cubic yards in six years and 1 mg/kg Tri+ PCB, respectively),
the performance standard for PCB losses due to resuspension must justify both the
ultimate numerical goals as well as the required implementation.
The remedial action objectives (RAOs) provide the ultimate basis for the development of
the Resuspension Standard. As discussed in the 2002 ROD (USEPA, 2002a):
[the] RAOs address the protection of human health and protection of the
environment. (ROD ง 9.1, page 50)
The RAO specifically addressed by this Resuspension Standard is the following:
Minimize the long-term downstream transport of PCBs in the river. (ROD
ง 9.1, page 51)
In the ROD, the goal of the Resuspension Standard for PCB losses is defined in the
following context:
...Analysis of yearly sediment resuspension rates, as well as resuspension
quantities during yearly high flow events, shows the expected
resuspension due to dredging to be well within the variability that
normally occurs on a yearly basis. The performance standards and
attendant monitoring program, that are developed and peer reviewed
during design, will ensure that dredging operations are performed in the
most efficacious manner, consistent with the environmental and public
health goals of the project. (ROD ง 11.5, page 85)
And again:
... Sampling and monitoring programs will be developed and implemented
during the design, construction and post-construction phases
to...determine releases during dredging.... These monitoring programs
will include sampling of biota, water and sediment such that both short-
and long-term impacts to the Upper and Lower Hudson River environs, as
a result of the remedial actions undertaken, can be determined and
evaluated. EPA will increase monitoring of water supply intakes during
each project construction phase to identify and address possible impacts
on water supplies drawn for drinking water. The locations, frequency and
other aspects of monitoring of the water supplies in the Upper and Lower
Hudson River PCBs Superfund Site
Engineering Performance Standards
45
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Hudson will be developed with public input and in consultation with New
York State during remedial design. (ROD ง 13.3, page 99)
Controlling the export of PCBs during the remediation will keep the water column
concentrations close to current baseline levels and, by extension, keep fish tissue
concentrations close to baseline levels during the remediation. In short, the goal of the
standard is to minimize PCB losses during dredging to reduce risks to human and
ecological health by controlling PCB exposure concentrations in drinking water and
fish tissue.
3.1.1 Development of Water Column Concentration Criteria for PCBs
As discussed in subsection 2.9, there are seven applicable chemical-specific ARARs for
PCBs. Of these, the three lowest concentration standards (0.001 ng/L to 1 ng/L Total
PCBs) were waived in the ROD, because it is technically impractical to reach theses
levels in the Hudson River with continuing input from the upstream sources. Three of the
remaining ARARs are concentrations that fall within the range of baseline conditions (14
ng/L to 90 ng/L Total PCBs) and cannot be considered for resuspension criteria during
the remediation.
Only the 500 ng/L Total PCB MCL is a practical
limit for use as a resuspension criterion, because this
concentration generally falls outside of the baseline
concentrations. The standard is written to permit a
short-term increase in water column concentrations as
long as the long-term goals of the remedy as defined
in the ROD are met.
The river sediments are currently the primary source of the contamination in the Upper
Hudson River and the removal of sediments is essential for the long-term benefit of the
river. Additionally, removal of PCB-contaminated sediments will provide benefit to the
remediated portions of the river during the remediation. As such, a limited amount of
resuspension will be permitted because the benefits to the river outweigh the short-term
impacts from dredging. This is consistent with the USEPA sediment principles recently
promulgated by the Office of Solid Waste and Emergency Response (USEPA, 2002c).
Best management practices (BMPs) were considered as a basis for the standard. BMPs
that could be implemented include structural and non-structural practices. Structural
practices include:
Containment.
Shoreline restoration.
Placement of backfill prior to removal of containment.
Non-structural practices include:
Only the 500 ng/L Total PCB
MCL is a practical limit for use as
a resuspension criterion,
because this concentration
generally falls outside of the
baseline concentrations.
Hudson River PCBs Superfund Site
Engineering Performance Standards
46
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Cessation of work at velocities above a set rate.
Minimization of the use of boats with the potential to produce significant prop
wash.
Structural practices are not required by this standard, because the locations where there is
need for these practices should be identified during the design, when all available data
can be fully analyzed. Similarly for non-structural practices, these requirements are
specific to the equipment chosen and specific local conditions and are better set as
requirements during the design. The cost, impact on productivity, and effectiveness of
these practices should be carefully weighed prior to setting these requirements. It is
expected that the design will include some best management practices to achieve
compliance with the standard, but these will not be specifically required by the standard.
Ulitimately, this standard is peformance-based and not prescriptive so as to encourage
engineering innovation to protect the environment, optimize operations and complete the
remediation as quickly as possible.
The most important ARAR for drinking water supplies is the federal maximum
contamination limit, or MCL, for drinking water supplies, 500 ng/L Total PCBs4. This
ARAR establishes the first of two objectives for the Resuspension Standard:
3.1.1.1 Objective 1 Development of Primary Criteria for Drinking Water
Drinking Water: Maintain PCB concentrations in raw water at drinking water
intakes at levels less than the federal MCL of 500 ng/L.
Objective 1 establishes a numerical limit on PCB
concentrations in the Upper Hudson. Adherence to
this level provides assurance that no public water
supplies will be adversely impacted by the
remediation, regardless of a given water treatment
plant's (WTP's) ability to treat PCB-bearing water.
Most of the WTPs potentially affected by the
remediation have treatment systems that can reduce the concentration of PCBs in the
finished water, although the current degree of reduction is unknown. For this reason, this
standard will take the more conservative approach and not rely on this capability. Instead,
this standard will be structured such that compliance with the standard achieves
acceptable water column concentrations in the raw water.
Based on this objective, PCB export must be sufficiently controlled so as to prevent
exceedance of the 500 ng/L Total PCBs level at the water supply intakes at Waterford
and Halfmoon, New York, the first public water supply intakes downstream of the
remedial areas. While dilution and degradation can be expected to reduce PCB
concentrations in the water column during transit from River Sections 1 and 2 to the
4 The New York State MCL is also 0.0005 mg/L Total PCBs (500 ng/L).
The Resuspension Standard
takes a conservative approach
and is structured to achieve
acceptable water column
concentrations in raw water,
regardless of WTP capability.
Hudson River PCBs Superfund Site
Engineering Performance Standards
47
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
public water intakes, these processes cannot be relied upon while dredging in River
Section 3. Thus, dredging in River Section 3 requires that PCB export due to dredging
not result in water column concentrations in excess of the federal MCL. As a
conservative approach for the protection of the water supplies, this same concentration
level (500 ng/L Total PCB) is applied at all far-field monitoring locations and is the
standard for water column concentrations {i.e., the Resuspension Standard threshold).
An action level criterion was also derived from
Objective 1. Although the 500 ng/L level represents acf,on 'eve' concentration
i r. limit of 350 ng/L, below the 500
a level not to be exceeded, there is need for an ng/L federal ^L, wiUsetveasa
action level below the MCL. Specifically, it is trigger for additional monitoring
desirable to keep water column concentrations and engineering controls.
below the federal MCL while still meeting the
productivity goals of the remedial operation. To this end, a second concentration limit of
350 ng/L Total PCBs was established. This value represents 70% of the MCL value and
serves as a trigger for additional monitoring. This limit can also be derived from
statistical considerations based on the variability of the water column concentrations and
the analytical uncertainty in the PCB measurements, as described below.
No estimate exists of the likely variability of water column PCB concentrations in the
Upper Hudson due to dredging. The variability of baseline conditions can be substituted
as an initial estimate, or surrogate, although it is likely that dredging-related variability
will be greater than the baseline variability. For the analysis that follows, the baseline
variability of the Schuylerville station will be used. In order to scale this variability, the
o_
ratio of the standard deviation to the mean, {i.e., the coefficient of variance LxJ) wiH be
used. For this location, based on the GE data set, the coefficient of variance (CV) is
approximately 0.35. The 95th percentile is approximately 2 CVs, or 0.70 of the value. For
the value of 500 ng/L Total PCB, this represents + 350 ng/L Total PCB with a lower 95th
percentile of 150 ng/L.
As can be seen in the table of baseline data in Table 3-1, this value is near or within the
range of baseline variability and is thus not useful as an action level threshold.
As an alternative, it is also possible to determine a value that has no more than a 5%
probability that the actual value is 500 ng/L Total PCB. That is, determine a threshold
value based on the same CV such that 500 ng/L is the 95th percentile upper bound.
This is given as:
7 + 0.7*7 = 500
where
Y = the threshold value
0.7 = 2*CV.
Hudson River PCBs Superfund Site
Engineering Performance Standards
48
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
In this case, Y has a value of 294 ng/L, lower than the selected value of 350 ng/L.
If the Control Level were to require a response based on a single value, then this value,
nominally 300 ng/L, might be a preferable choice over 350 ng//L. However, the Control
Level is based on a one-week average, representing the mean of seven measurements. For
an average, the upper and lower bounds are based on the standard error and not the
standard deviation.
The ratio of the standard error (SE) to the mean becomes
where 7 is the number of samples in the seven-day running average. In this instance, the
equation for the threshold value using 2*SE becomes:
This yields a threshold value of 397 ng/L. The selected value of 350 ng/L falls in the
center of the two threshold estimates and is considered a good initial value for the
program, given the unknown variance associated with dredging-related PCB
concentrations.
Analytical precision must also be considered as it pertains to water column
measurements. The precision of the historical analyses is quite good. At the Schuylerville
station, the historical blind duplicate pairs yielded a median relative percent difference
(RPD) of 8.1% and a mean RPD of 12.7% (see Figure 3-1). Ninety percent of all pairs
had an RPD less than 22%. For an actual concentration of 350 ng/L, the mean RPD
would suggest a possible analytical range of uncertainty of 328 ng/L to 372 ng/L (actual
value + RPD/2). On this basis, the analytical variability should not limit the applicability
of the 350 ng/L threshold value.
Engineering evaluations and improvements are required if the average concentration
increase is 350 ng/L or higher for a week. These activities are required to identify and
correct any potential problems that may cause a subsequent exceedance of the federal
MCL, thus causing a possible disruption in the operations and requiring contingency
actions on the part of the municipal water suppliers. This concentration threshold is
defined as a Control Level criterion.
Compliance with these resuspension criteria at the far-field stations attains the objective
and protects public water supplies during the remedial efforts. These criteria are designed
to limit short-term impacts, since the river will deliver any resuspended PCBs to the
downstream water supplies at Waterford and Halfmoon in a matter of days. However, the
ROD clearly is also concerned with the impacts to fish and downstream consumers of
Hudson River PCBs Superfund Site 49 Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Resuspension - April 2004
SE_
X
7 + 0.26*7 = 500
-------
fish. This concern requires a longer perspective, since fish integrate their exposure to
PCBs over both time and area. Thus, fish tissue concentrations are likely to be more
affected by a long, steady loss of PCBs than a single large release event. A second
objective can be defined specific to this issue, as discussed in the following section.
3.1.1.2 Objective 2 Development of Primary Criteria for PCB Loads
Fish Tissue: Minimize long-term net export of PCBs from dredged areas to control
temporary increases in fish tissue concentrations.
Objective 2 addresses the need to limit the impact of
the remediation on the anticipated recovery of river
after the remedial dredging is completed. This
objective recognizes that the export of PCBs during
dredging has the potential to slow the rate of recovery
for fish body burdens and related exposures if it is
sufficiently large. However, this objective also recognizes that it is primarily the long-
term release of PCBs that has the potential to create an adverse impact. Short-term
releases can be tolerated so long as the long-term average continues to satisfy the criteria.
In general, short-term releases are of the time scale of hours to days, while long-term
releases are considered in terms of several weeks to months or longer. Thus, from the
perspective of the ROD, the short-term releases are manageable so long as the eventual
recovery of the river is not compromised. As noted in the ROD (USEPA 2002a):
Although precautions to minimize resuspension will be taken, it is likely
that there will be localized temporary increases in suspended PCB
concentrations in the water column and possibly on fish PCB body
burdens. (ROD ง 11.5, page 85)
This objective can be approached from two perspectives:
Ideal rate of PCB export
Acceptable maximum rate of PCB export
The ideal rate is obviously no PCB release at all. However, this is also unattainable. The
case study analysis presented in subsection 2.2 and the resuspension analysis presented in
the RS (USEPA, 2002b) provide some useful target values, however. The two sites
examined in subsection 2.2, the GE Hudson Falls remediation and the New Bedford
Harbor Hot Spot remediation, achieved net PCB export rates of 0.36% and 0.13%,
respectively, relative to the mass of PCBs removed. These percentages translate to Total
PCB resuspension export rates of 240 and 86 g/day of operation, or 50 and 18 kg/yr on an
annual basis for the remediation of the Hudson, respectively. These annual values
represent only a small fraction of the annual baseline load of 260 to 400 kg/yr observed
for the period 1996-2002 (see Figure 7 of Attachment B). Export at this level is unlikely
Short-term release of PCBs can
be tolerated as long as the long-
term average continues to satisfy
the Resuspension Standard
criteria.
Hudson River PCBs Superfund Site
Engineering Performance Standards
50
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
to have any discernable impact on fish tissue concentrations, given the baseline
variability.
In developing the load criteria for the standard, several different perspectives were
examined to make the standard meaningful {i.e., not too high) and achievable {i.e., not
too low). These include the following:
Best engineering estimate of resuspension production and export
Minimum detectable PCB load increase
Loads defined by the water column concentration criteria of 350 and 500 ng/L
Total PCBs
Impact of load on fish tissue recovery
Delivery of Total PCBs and Tri+ PCBs to the Lower Hudson {i.e., Waterford
load)
Each of these perspectives has the potential to provide some level of constraint on the
selection of a PCB load criterion. Each is discussed below.
Best Engineering Estimate of Resuspension Production and Export
The analysis presented in Appendix E.6 of the
feasibility study (USEPA, 2000b) and in the
responsiveness summary (USEPA, 2002b)
provided an initial engineering estimate of the
rate of PCB release from the dredge operation.
The analysis estimated a resuspension production
rate and a resuspension release rate, yielding an
estimated Total PCB export rate of approximately I
the PCB mass to be removed from the river bottom
In the preparation of the Resuspension Standard,
the initial model analysis of suspended solids
transport has been expanded and improved to
more realistically represent conditions as well as
to account for the kinetics of PCB dissolution.
These results were discussed previously in
subsection 2.7, and a detailed discussion is
provided in Attachment D. These analyses confirm the results initially presented in the
FS (USEPA, 2000b). The current analysis estimates a PCB export rate only slightly
greater than the original estimate, at 90 g/day (19 kg annually5) or about 0.14% of the
PCB mass to be removed. Based on these results, a best engineering estimate of
approximately 20 kg per dredging season was selected as the target load value.
5 The target PCB export rate of 19 kg/year represents a daily resuspension export rate of 90 g/day,
assuming a 210-day dredging season (May through November) and seven days per week of operation. This
is conservative in that operations less than seven days per week would effectively result in lower average
daily export rates.
The best engineering estimate in the
FS and RS estimated a Total PCB
export rate of approximately 86
g/day (18 kg/yr), or 0.13% of the
PCB mass to be removed from the
river bottom.
g/day (18 kg annually), or 0.13% of
9,800 kg).
The best engineering estimate
analysis for the Resuspension
Standard estimates a Total PCB
export rate of approximately 90
g/day (19 kg/yr) or 0.14% of the PCB
mass to be removed from the river
bottom - only slightly higher.
Hudson River PCBs Superfund Site
Engineering Performance Standards
51
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Although a target level of 90 g/day Total PCB would appear a desirable target (the
analysis presented in the FS shows this loading rate to have a negligible6 impact on the
recovery of fish tissue concentrations throughout the river), this value does not account
for activities other than the dredge operation. Boat movements, debris removal, barrier
installation and removal, and other activities related to the dredging operation all have the
potential to release PCBs, but are difficult to quantitate. Hence, a set of criteria is needed
to define reasonable upper limits for dredging-related releases based on estimated
impacts to the river. Much of the analysis described in subsection 2.2 was completed with
the intention of providing input to the selection of these limits.
Minimum Detectable PCB Load Increase
An important limitation in selecting the PCB load criteria is the ability to measure the net
increase in load due to dredging activities. Several considerations must be addressed in
this regard. The selection of the far-field locations as the main PCB monitoring locations
is a direct result of this concern. Baseline loads of PCBs originating from the sediments
are similar in magnitude to those expected from dredging. Much of the sediment initially
added to the water column will rapidly settle, releasing little or no PCBs. Hence the
ability to detect a net PCB load increase in the poorly mixed region around the dredge
operation {i.e., at the near-field monitoring stations) is difficult and highly uncertain. For
this reason, PCB monitoring will be conducted well away from the dredging operation
{i.e., far-field monitoring), where the net PCB load should be more stable and can be
detected over baseline conditions.
As discussed in subsection 2.4 and Attachment B,
this approach does have a limit on the ability to
measure PCB export at a far-field station. Based
on the historical variability observed in the
available data, it is unlikely that PCB export
below 300 g/day (65 kg Total PCBs annually7)
can be differentiated from baseline conditions.
This value then provides a minimum observable PCB export rate or load. Notably, the
target load for PCB export due to dredging previously provided falls below the detectable
load rate. Thus, if the best engineering estimate of an approximate 20 kg/dredging season
export rate is achieved, there will be no measurable increase in PCB export. From a
monitoring perspective, the goal for dredging is no observable increase in PCB load
above baseline.
Because it is unlikely that PCB
export below 300 g/day (65 kg Total
PCBs per yr) can be differentiated
from baseline conditions, this value
represents the minimum observable
PCB export rate, or load.
6 A negligible impact in the Upper Hudson is defined as a forecast fish tissue concentration difference
relative to the no-resuspension dredging scenario of 0.5 mg/kg or less within 5 years after the cessation of
dredging.
7 This rate of PCB export corresponds to slightly less than 0.5% of the estimated mass of PCBs to be
removed.
Hudson River PCBs Superfund Site
Engineering Performance Standards
52
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Loads Defined by the Water Column Concentration Criteria of 350 and 500 ng/L
Total PCBs
The federal MCL provides a means to obtain an upper bound on the annual and daily load
rate. If daily Total PCB concentrations remain at a monthly average concentration of 500
ng/L throughout the dredging season, the PCB export load can be calculated from the
difference between 500 ng/L and the average baseline concentration for the month. This
o
calculation yields an export rate of about 2,300 g/day (500 kg annually ). The 350 ng/L
Total PCB resuspension criterion also provides a basis for a loading estimate.
To maintain a weekly average concentration of 350
ng/L Total PCBs, the resuspension export rate must
be approximately 1,600 g/day (340 kg annually9). For
the purposes of this standard, the Control Level is
expected to be the maximum operating condition,
since concentrations above this level will require engineering improvements to reduce the
releases. From this consideration, 1,600 g/day (340 kg annually) represents the likely
maximum annual load that can be derived from the water column concentration criteria.
This level cannot be maintained indefinitely, however, because the load-based limits are
set at lower values [600 g/day].
Impact of Load on Fish Tissue Recovery
The ability to measure a net increase in PCB export relative to baseline conditions and
the water concentration criteria provide potential bounding criteria for an acceptable
export rate. However, it is still necessary to demonstrate that export rates at these levels
do not substantively alter the recovery period of the river as measured by the decline in
PCB concentrations in fish tissue. The model simulation for the best engineering estimate
for resuspension presented in the responsiveness summary is the basis for comparison10.
To investigate this, a series of model forecasts were conducted at resuspension release
rates (near-field) and resuspension export rates (far-field) derived from the load
considerations discussed in the foregoing subsections. The model runs dealing with long-
range forecasts are summarized in subsection 2.6. The near-field model analysis is
summarized in subsection 2.7. A complete discussion of the supporting model analyses is
provided in Attachment D. Table 2-4 lists the completed model runs along with brief
descriptive information.
Due to the inherent nature of the HUDTOX model structures, PCB loads cannot be
readily specified at far-field locations. Rather, the input of PCBs is specified as an input
load at a location within the river, equivalent to a resuspension release rate. For the
8 This rate of PCB export corresponds to about 3.8% of the PCB mass to be removed.
9 This rate of PCB export corresponds to about 2.4% of the PCB mass to be removed.
10 Since the completion of the Feasibility Study, various factors and considerations have lead to a suggested
start date for the remediation of 2006, instead of 2004 as originally planned. Since the best estimate
simulation prepared for the Feasibility Study was barely discernable above the "no resuspension"
simulation, the simulations prepared here were simply compared against a revised "no resuspension" result,
reflecting the later start date. The 90 g/day best estimate condition was not rerun.
1,600 g/day (340 kg annually)
represents the likely maximum
load derived from water column
concentration criteria.
Hudson River PCBs Superfund Site
Engineering Performance Standards
53
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
supporting model runs, the resuspension release rate was derived iteratively, by
estimating the resuspension release rate (input to the model) and then checking the
resuspension export rate (the model output) until the model output met the desired
criteria. This process was necessary in order to make the model match the potential
control criteria at the planned monitoring locations. These iterations also took into
account the different river sections, with their corresponding target sediment properties
{i.e., silt fraction), PCB concentrations and hydrodynamics. The simulations also account
for the changes in dredging location as the remediation progresses.
For example, to simulate the 350 ng/L Total PCB condition {i.e., the Control Level
threshold for the entire dredging program), it was necessary to provide the following
loads in the three river sections:
River Section 1: approximately 1,550 g/day Total PCBs and 56,000 kg/day of
sediment
River Section 2: approximately 2,300 g/day Total PCBs and 35,000 kg/day of
sediment
River Section 3: approximately 2,800 g/day Total PCB and 94,500 kg/day of
sediments.11
These PCB and sediment loads reflect the differences in PCB concentration, river flow
and monitoring locations among the three river sections. PCB and sediment loads had to
be further varied to reflect the year-to-year movements of the dredges within each river
section. As would be expected, less resuspension was necessary to achieve a specified
PCB concentration or load at the far-field station the closer the dredge was to the station.
Model simulations for the 350 ng/L Total PCBs scenario were run to examine the impact
of this criterion on the recovery of the river, using the recovery of fish tissue
concentrations as the main measure (see Figures 2-8 and 2-9). This scenario showed
some fish body burden increases during dredging but negligible12 changes to fish tissue
trajectories during the post-dredging period. After noting the negligible impact of the 350
ng/L scenario, there was no need to run a 300 g/day scenario since its impact would
clearly be much less.
A 600 g/day Total PCBs scenario was run, based on its selection as a load criterion (see
below). As expected, the 350 ng/L scenario has a greater impact than the 600 g/day
scenario. However, both model runs indicate negligible13 changes in fish tissue
concentrations in regions downstream of the dredging. Within five years of the
11 To put the suspended solids values in perspective, at a nominal flow rate of 4,000 cfs and 2 to 4 mg/L of
suspended solids, the Hudson transports 20,000 to 40,000 kg of solids per day, respectively.
12 A negligible impact in the Upper Hudson is defined as a forecast fish tissue concentration difference
relative to the no-resuspension dredging scenario of 0.5 mg/kg or less within 5 years after the cessation of
dredging. In the Lower Hudson, it is defined as a forecast fish tissue concentration difference relative to the
no-resuspension dredging scenario of 0.05 mg/kg or less within 15 years after the cessation of dredging.
Note that in the Lower Hudson, fish tissue concentration forecasts always agree within 0.5 mg/kg except
for one year during the dredging period for the 350 ng/L scenario at River Mile 152.
13 See footnote 12.
Hudson River PCBs Superfund Site
Engineering Performance Standards
54
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
completion of dredging there is little discernable
impact from the dredging releases based on the fish
tissue forecasts. The model results suggest that
compliance with the water concentration criteria
previously developed {i.e., 350 ng/L and 500 ng/L)
will also minimize dredging impacts to the long-term
recovery of the river.
Delivery of Total PCBs and Tri+ PCBs to the Lower Hudson
In addition to recovery of the river as measured by fish tissue concentrations, impacts to
the river due to dredging can also be gauged by the absolute mass of PCBs released. For
this comparison, both Total PCBs and Tri+ PCBs are considered. The emphasis is placed
on the estimated Tri+ PCB releases, however, since this is the fraction of PCBs that is
bioaccumulative. This fraction is also far better understood from the perspective of
sediment inventory and geochemical processes (the USEPA models simulate Tri+ PCBs).
As noted previously, the main consideration in developing a load standard is to minimize
the release of PCBs. For this reason, the cumulative PCB load at Waterford, as forecast
by the HUDTOX model, provides a useful gauge of any suggested loading standard. In
this instance, the ideal condition is that given by the no resuspension scenario for the
selected remedy. The upper bound would be the load delivered by the original monitored
natural attenuation scenario (MNA). The forecast for acceptable load criteria would fall
between the MNA and the no resuspension scenario.
The Tri+ PCB load forecasts for several load conditions are presented in Figure 2-4. The
lowest curve, representing the least amount of PCBs transported downstream, represents
the no resuspension scenario. MNA is also indicated on the figure. Because of the
dredging-related PCB releases, all scenarios except no resuspension exceed the MNA
forecast during the dredging period. Unlike the lower PCB release scenarios (see the
upper diagram in Figure 2-4), the forecast curve corresponding to the 350 ng/L criteria
never crosses over the MNA curve, indicating that setting the loading standard on the
basis of this water concentration criterion would deliver significantly more Tri+ PCB
mass to the Lower Hudson than MNA.
The 300 g/day scenario, equivalent to 100 g/day Tri+ PCBs (run to 2020), crosses the
MNA curve just before the cessation of dredging. While this scenario was not run for the
full forecast period, it is evident that the Tri+ PCB load level for the 300 g/day scenario
would deliver much less Tri+ than the MNA. Also shown on the figure is a forecast curve
for a Tri+ PCB load for the 600 g/day scenario, equivalent to 200 g/day Tri+ PCBs14.
This curve also crosses the MNA forecast, just after the completion of dredging. On the
basis of this analysis, both the 300 and a 600 g/day load standards would yield acceptable
Tri+ PCB loads to the Lower Hudson.
Within five years of completion of
dredging, there is little
discernable impact from
dredging releases, based on the
fish tissue forecasts.
14 This load is equivalent to 130 kg/year of Total PCB and 44 kg/yr of Tri + PCBs, or slightly less than 1%
of the estimated mass of Total PCBs to be removed.
Hudson River PCBs Superfund Site
Engineering Performance Standards
55
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The impacts of the possible load criteria were also examined for Total PCBs, as
illustrated in the lower diagram of Figure 2-4. These Total PCB curves are considered
less certain, since the EPA models were developed to simulate Tri+ PCBs and not Total
PCBs. Nonetheless they provide some guidance. The results from this analysis also show
an unacceptably high Total PCB load to the Lower Hudson, based on the 350 ng/L
criterion. Both the 300 and the 600 g/day forecasts show less total load delivered to the
Lower Hudson than MNA, although the equivalence points occur later in time. The 600
g/day forecast crosses MNA about 20 years after the completion of dredging. The overall
load difference between the 600 g/day scenario and MNA is relatively small, such that an
increase in the daily load to 700 g/day would probably exceed the MNA curve. Given the
uncertainties in the Total PCB estimates, the Tri+ PCB forecasts are considered the more
reliable gauge among these scenarios.
Selection of a Load-Based Criterion
Taking into account the various considerations described above, it is clear that the target
load of 90 g/day is not measurable, and the load equivalent to 350 ng/L delivers an
unacceptably large mass of PCBs to the Lower Hudson. None of the load scenarios
chosen as criteria yield an unacceptable impact on fish tissue concentrations, so this
gauge is not useful here. The need to control PCB loads to the Lower Hudson provides
the strictest limitation in the selection of a load criterion. This criterion is primarily based
on Tri+ PCBs, the form of PCBs simulated by USEPA's models. Total PCB restrictions
are more uncertain in this regard since they were not the focus of USEPA's models.
While no exact value results from this analysis, it
is clear that the loading standard must fall
between the ability to measure it {i.e., 300 g Total
PCBs/day detection threshold) and the 350 ng/L-
based load of 1,600 g/day, which results in
unacceptable loads to the Lower Hudson.
A load of 300 g Total PCBs/day has been selected as a resuspension criterion, because it
represents a best management practices goal. A load of 600 g/day, representing 130 kg
annually, is the daily equivalent of the maximum allowable annual load and is also
selected as a load criterion. It is twice the load detection threshold and therefore
measurable. It is less than the 350 ng/L - 1,600 g/day condition and results in acceptable
Tri+ and Total PCB load increases to the Lower Hudson.15 In term of absolute loads, the
130 kg/year represents slightly more than a 40% increase in the mean annual load at
Schuylerville (300 kg/yr for 1998-2002). Added to this value, the load increase would
yield 430 kg/yr, which is just beyond the observed range at Schuylerville between 1998
and 2002 (180 to 410 kg/yr).
Relative to TI Dam loads, this 600 g/day load increase represents a 40% to 90% increase
in the observed loads (TID West and TID-PRW, respectively) for 1996 to 2002. More
importantly though, this load represents a nearly seven-fold increase relative to the best
15 As was noted previously, the Total PCB load is not considered a robust constraint due to its uncertainty.
The loading standard must fall
between the ability to measure -
300 g Total PCBs/day) and the
350 ng/L-based load of 1,600
g/day that results in unacceptable
loads to the Lower Hudson.
Hudson River PCBs Superfund Site
Engineering Performance Standards
56
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
engineering estimate of 90 g/day, thus providing a reasonable allowance for other
dredging-related releases (e.g., boat traffic and debris removal). Yet as noted above, this
load increment would have negligible16 impacts on the long-term river recovery,
generating only brief (one-to-two-year) increases in fish tissue concentrations relative to
the MNA scenario.
Based on these considerations, the value of 600 g/day
has been selected as the primary load criterion: 600
g/day is equivalent to 650 kg load loss over the entire
remediation and 65 kg/yr in Phase 1 assuming half
the targeted production rate will be achieved.
Long-term maximum load loss limits of 650 kg Total PCBs and 220 kg Tri+ PCBs for
the entire remediation have been derived from review of the model predictions.
Adherence to these limits is important for the recovery of the river and protection of the
Lower Hudson River. These limits have not been included as resuspension criteria
directly, because these are end-of-remediation goals that do not fit within the
performance standard structure. Indirectly these limits are implemented over shorter
times frames, with daily limits for Total PCBs and Tri+ PCBs at 600 g/day and 200
g/day, respectively, and Phase 1 dredge season and annual limits of 65 kg and 22 kg,
respectively, for the Phase 1 dredge season. As long as the load-based resuspension
criteria are adhered to, the long-term load loss limit will not be exceeded.
Because Tri+ PCBs are the most important
component of Total PCBs for the recovery of fish
tissue concentrations, a load criterion is desired for
this parameter as well. This criterion is simply
derived from the Total PCB load criterion and the observation that the Total PCB to Tri+
PCBs ratio in the sediments is approximately 3:1. Since sediments are the main form of
release of PCBs, it is expected that the net addition of Tri+ PCBs will be one-third that of
Total PCBs, yielding a primary criterion for Tri+ PCBs of 200 g/day.
The last consideration for selecting the load-based
criteria is the time frame over which these apply.
Taking into consideration the long-term nature of
the load impacts and the likely high degree of
short-term variability, the criteria should be based
on longer-term conditions in order to avoid major
disruptions to the operation due to short-term
exceedances. For this reason, the Evaluation Level and Control Level load criteria will be
measured over seven-day periods by constructing a running average of Tri+ and Total
PCB loads at all far-field stations for the entire dredging season.
16 A negligible effect in the Upper Hudson is defined as a forecast fish tissue concentration difference
relative to the no-resuspension dredging scenario of 0.5 mg/kg or less within 5 years after the cessation of
dredging.
600 g/day, the daily equivalent of
a 650 kg load loss over the
entire remediaion and 65 kg/yr in
Phase 1, has been selected as
the primary load criterion.
The primary load criterion for
Tri+ PCBs is 200 g/day, one-
third that of Total PCBs.
Evaluation and Control Level load
criteria will be measured over 7-
day periods by constructing a
running average of Tri+ and Total
PCB loads at all far-field stations
for the entire dredging season.
Hudson River PCBs Superfund Site
Engineering Performance Standards
57
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
3.2 Rationale for a Tiered Approach
The actions levels (Evaluation Level, and Control
Level) were developed to facilitate a steady level of
remedial activities while still providing
environmental protection. The tiered approach is
intended to require additional sampling and
engineering controls as PCB levels rise above those
predicted by the best engineering analysis. This tiered
approach provides action levels to trigger monitoring contingencies and implementation
of additional engineering controls and thereby avoid a complete cessation in the
operation. It is the intention of this standard to both minimize PCB losses and facilitate
uninterrupted remedial operations.
In this approach, monitoring requirements will increase as the action levels are exceeded
to provide data to clarify the nature of the PCB losses. These data can then be used to
direct engineering control improvements while dredging operations continue unabated.
The monitoring requirements will have no effect on dredging operations and productivity
since there is no affect on the equipment and crews involved.
A tiered approach provides for
additional sampling and
appropriate engineering controls
as PCB levels rise, thereby
avoiding the need to cease
dredging operations.
3.2.1 PCB Considerations
In developing the tiers of the standard, the need to control PCB export must be balanced
with the need to comply with the federal standard. As extensively discussed in
Attachments A and B, baseline water column PCB concentrations vary from month to
month, necessarily complicating the structure of the standard. Based on these concerns,
the PCB component of the Evaluation Level is a flux-based action level. The Control
Level has both flux-based and concentration-based PCB criteria. Exceedance of absolute
concentrations for the flux-based criteria at the Evaluation Level is not a concern in this
instance. The purpose of the Evaluation Level is to control PCB export and potential
long-term impacts to the recovery of the river.
The PCB concentration-based criterion of 350 ng/L is included in the Control Level to
address the concern over exposure to PCBs through public water supplies as the MCL is
approached. The duration for the exceedance is one week, based on a seven-day average
in acknowledgement of the anticipated variability in water column conditions. As
previously discussed, the federal MCL of 500 ng/L Total PCBs represents an absolute
maximum concentration, the exceedance of which will cause the temporary halting of the
remedial operations following confirmation of the concentration.
The Control Level at 350 ng/L Total PCBs will be
the effective maximum allowable level, since
exceedance of this level means that the absolute
Exceedance of the Control Level
350 ng/L Total PCBs means the
MCL is being approached and
serves as an effective trigger for
engineering controls.
Hudson River PCBs Superfund Site
Engineering Performance Standards
58
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
maximum is being approached and that extra efforts are required to control PCB export.
By requiring operations to maintain water column conditions below this value (350 ng/L
Total PCBs), the Control Level provides a relatively large window of protection,
decreasing the likelihood of a 500 ng/L Total PCB event.
When concentrations exceed 350 ng/L Total PCBs on average for one week or more,
engineering evaluations and engineering improvements become mandatory until riverine
conditions falling below the Control Level are achieved. Notably, months with high
baseline concentrations will have relatively little "room to spare" and may require tight
controls on the dredging operations to comply with this criterion. Exceedance of the
Control Level may prompt temporary cessation of operations as deemed necessary by
USEPA.
The monitoring and engineering requirements of the Control Level reflect the gravity of
the exceedances. The increased sampling frequency is needed to have sufficient
confidence in the results. These results may prompt costly engineering improvements if
exceedance of the criteria is demonstrated. Extensive monitoring requirements and
mandatory engineering controls are needed at this level to quickly identify the problems
and render a solution, thereby avoiding a cessation of the dredging operation.
Exceedance of the Resuspension Standard threshold
(500 ng/L Total PCBs) will require a cessation of
operations if the exceedance is confirmed by
samples collected the following day. If dredging-
related PCB concentrations and loads increase
gradually, there should have been at least two
attempts (one for each of the two lower action levels) to understand and control any
resuspension problem prior to the exceedance of the 500 ng/L threshold. Alternatively, a
rapid rise in PCB concentration from baseline to more than 500 ng/L represents an
unexpected and poorly understood event. In either case {i.e., exceedance of the
Resuspension Standard threshold), temporary halting of operations is required since
conditions are clearly not as anticipated and may have significant consequences.
3.2.2 Suspended Solids Considerations
While PCB concentrations and loads are clearly the most important focus of this
standard, determination of PCB conditions in the river is time-consuming, with a
significant lag between the collection of samples and the availability of preliminary
(draft) data. For this reason, it is desirable to measure and monitor parameters that
correlate with PCBs and can be determined readily. Suspended solids, in particular, fit
this requirement and have been selected for monitoring as well. Suspended solids
measurements are reflective of short-term conditions, since the concentrations will vary
rapidly in response to sediment disturbances. For this reason the suspended solids criteria
will be derived from the water column concentration criteria described in subsection
Exceedance of the 500 ng/L
Total PCBs Resuspension
Standard threshold requires
cessation of operations if
confirmed by sampling.
Hudson River PCBs Superfund Site
Engineering Performance Standards
59
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
3.1.1. Acceptable suspended solids concentrations were developed for both near-field and
far-field conditions.
To further support the development of the
suspended solids criteria, near-field conditions were
simulated using a Gaussian plume model (TSS-
Chem) to estimate the impact of various
resuspension release rates. This analysis,
summarized in subsection 2.7 and described in
Attachment D, indicates that resuspension release
rates corresponding to PCB loads of 300 to 2,000
g/day are rapidly reduced in the near-field region, with resulting PCB export rates at the
far-field stations two to six times less than the release rates.
This analysis included an estimation of kinetically controlled PCB desorption, suggesting
relatively minimal rates of dissolved-phase PCB release in the immediate vicinity of the
dredge. In the region between 10 and 1,000 m downstream of the dredge, PCB loads
steadily diminish while gradually decreasing the fraction borne by suspended matter
relative to the dissolved phase. At the point of departure from the near-field region, PCB
loads are primarily dissolved-phase, but overall the loads are substantively reduced
compared to the immediate dredge area.
It can be concluded that downstream export of PCBs
(at one mile beyond the dredge operation) is unlikely
to exceed the 300 g/day Total PCB Control Level on
a regular basis. Furthermore, the analysis of
suspended solids release and PCB desorption,
presented in subsection 2.5 and Attachment C,
indicates the resuspension process alone controls the
PCB release within the dredging region. The creation of dissolved-phase releases by
processes other than PCB desorption from suspended solids is highly unlikely, further
supporting the focus of this performance standard on solids-related release mechanisms.
This assumption will be tested by the separate phase PCB analyses to be completed as
part of a special study.
Suspended solids criteria were developed for the Evaluation and Control Levels to
provide a means to identify potentially significant PCB releases more rapidly. In most
instances, suspended solids exceedances will necessitate additional PCB monitoring,
which in turn should identify whether the PCB criteria are being exceeded. While these
suspended solids criteria will require additional monitoring, it is the PCB concentrations,
not the suspended solids concentrations, that will trigger the need for additional
engineering controls. The additional monitoring will be limited to the far-field monitoring
requirements for the nearest representative far-field station, with the sampling timed to
capture the plume causing the exceedance. Near-field suspended solids sampling
frequency will remain as a continuous surrogate measurement (turbidity) with an added
Resuspension release rates
corresponding to PCB loads of
300 to 2,000 g/day are rapidly
reduced in the near-field region,
with resulting PCB export rates
at the far-field stations 2 to 6
times less.
Analysis supports focusing the
Resuspension Standard on
solids-related release
mechanisms, as it is apparent
that the resuspension process
alone controls PCB release
within the dredging region.
Hudson River PCBs Superfund Site
Engineering Performance Standards
60
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
suspended solids measurement {i.e., 2 samples per day) to be obtained only at the
noncompliant nodes..
3.2.3 Near-field Suspended Solids Criteria
Derivation of the suspended solids action levels is
described in detail in Attachment D and briefly
summarized here. The near-field suspended solids
action levels were derived using the TSS-Chem
model to simulate suspended solids conditions
corresponding to the PCB concentration resuspension
criteria.
For the Evaluation and Control Levels, suspended solids thresholds represent average
suspended solids concentrations 300 m downstream of the dredge that would yield a
Total PCB concentration exceeding 350 ng/L at the far-field station. The same suspended
solids values are used for both action levels; only the duration of the exceedance varies
between the levels. This was done to simplify the monitoring while still maintaining the
ability to identify significant events.
A location of 300 m downstream was selected since the model suggests a plume width of
50 m and a relatively homogeneous water column at this distance. At this distance, it
should be easy to reliably maintain a sensor in the plume and also minimize moment-to-
moment variability in suspended solids measurements. If barriers are installed, this
station will be placed 150 m downstream of the barrier. At these locations, a sustained
concentration of 100 mg/L suspended solids in River Sections 1 or 3, and 60 mg/L
suspended solids in River Section 2 will trigger an exceedance of either the Evaluation
Level or the Control Level, depending on the duration of the exceedance.
Additional monitoring will be required at a location closer to the dredge to provide the
operator with real-time information on the effectiveness of the dredge operations and the
suspended solids controls. A distance of 100 m downstream of the dredge was selected as
sufficiently downstream to provide some level of mixing and smoothing of the suspended
solids signal while still being close enough to provide rapid feedback to the dredging
operation. Feedback may be crucial in identifying operations or actions that cause
excessive turbidity but can also be controlled to minimize water quality impacts.
Another station will be located 10 m to the side of the dredge nearest the channel. At
these locations, a sustained concentration of 700 mg/L suspended solids will trigger an
exceedance of the Evaluation Level. If barriers are in place, these stations will not have
an associated resuspension criterion. In all cases, adjustment of the monitoring locations
will be considered if alternate sites can be shown to be more effective to the monitoring
goals.
The same suspended solids
values are used for both the
Evaluation and Control Level;
only the duration of the
exceedance varies between
the levels.
Hudson River PCBs Superfund Site
Engineering Performance Standards
61
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Unlike the PCB criteria, the near-field suspended solids criteria should be prorated
among all the active dredge operations in a given area, but for Phase 1, the concentration
criteria for the suspended solids will apply to each operation individually.
3.2.4 Far-Field Suspended Solids Criteria
Far-field suspended solids criteria were developed for the Evaluation and Control Levels,
reflecting the decreased sensitivity of suspended solids measurements at the far-field
monitoring station. The suspended solids at the far-field stations are derived from the far-
field PCB resuspension standard. The far-field suspended solids criterion was developed
by simply calculating the amount of suspended solids that can result in a net increase of
PCB concentration above the primary PCB criterion, assuming that the PCB
concentration on the suspended solids is the same as on the dredged sediment The 500
ng/L far-field Total PCBs standard was used as a basis to calulate the suspended solids
criteria for the far-field stations.
Assuming the baseline level of PCB concentration is approximately 100 ng/L Total
PCBs, the net PCB concentration increase will be 400 ng/L Total PCBs. As stated in the
responsiveness summary, the average Total PCB concentration on the dredged sediment
across the three river sections is about 34 ppm. Based on these values, the increase in
suspended solids concentration above baseline is calculated to be about 12 mg/L. This
increase in suspended solids concentration must occur across the entire river and not just
within the dredge plume for the associated PCB concentration increase to occur. This
level (12 mg/L suspended solids increase) is close to baseline variability, however.
Considering the uncertainty in the calculation assumptions as well as the baseline
variability in suspended solids concentration, a value twice 12 mg/L, i.e., 24 mg/L, was
also selected. As a result, the Control Level uses 24 mg/L suspended solids as the far-
field suspended solids criterion. The Evaluation Level uses approximately half of this
value (12 mg/L suspended solids), with a shorter duration. The periods of exceedance are
the same as those for the near-field suspended solids action levels. The increased
monitoring requirements will be limited to the nearest downstream far-field station, with
the sample collection timed in order to capture the plume.
Due to the variable conditions within the river over time, some action levels may conflict
with one another, particularly in May and June when baseline concentrations are
relatively high. In these instances, the Control Level criteria for Total PCB concentration
may be exceeded even though the Total PCB load does not exceed the Control Level
criteria. The concentration-based action levels will govern the response, since these levels
are intended to provide protection for the downstream public water supplies and therefore
represent the more protective criteria in these instances.
Similarly, the suspended solids criteria may identify potentially important PCB
concentration or load conditions that are not verified by subsequent PCB sampling and
analysis. Exceedance of the suspended solids criteria prompts limited additional far-field
Hudson River PCBs Superfund Site
Engineering Performance Standards
62
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
sampling to determine the PCB concentration in the plume as it reaches the far-field
station. These additional samples are incorporated in the equations used to determine
whether the water quality levels are in compliance with the standard (subsection 4.1). In
all cases, exceedances of the action level criteria by any parameter {i.e., Total PCBs, Tri+
PCBs, or suspended matter) will spur additional monitoring requirements in order to have
a sufficient number of samples from which decisions can be made from the data with
confidence.
3.3 Monitoring Rationale
The rational for the standards describe above supports the framework and criteria which
form the Resuspension Performance Standard. Monitoring to verfy comliance or
excedance of the standard criteria is an integral part of the framework. This section
presents the overall rationale for the monitoring program as it is currently configured.
Significant adjustments to the monitoring program can only be made after the impacts of
the adjustments are evaluated in one or more special studies. Adjustments to certain
portions of the monitoring program may prompt evaluation of other aspects of the
sampling and possible revision of the resuspension criteria. For example, an alternate
monitoring program using automatic samplers to collect the PCB samples is presented in
Section 4.
As noted in the ROD (USEPA, 2002a), the export of PCBs from the dredging area to
regions downstream is the ultimate concern of this performance standard, since it affects
both fish and public water supplies. Thus, the most important monitoring stations are
those that monitor the rate of PCB export downstream. This increase in PCB export can
be best and most easily measured at sufficient distance downstream of the dredging
operation so that the river can homogenize the water column inputs from dredging. This
distance should also be sufficient to avoid the inclusion of solids suspended during
dredging that will settle in close proximity to the dredging operation and thus not
represent a source to regions downstream.
Based on historical evidence as well as concerns highlighted by the Fox River study
(USGS, 2000), these stations will be used for direct comparison with the Resuspension
Standard criteria only when the stations are at least one mile downstream of the dredging
operations. Baseline PCB conditions will be well characterized at these locations,
allowing the load increase due to dredging-related operations to be measured. In the near
field, the baseline is not characterized and may be highly variable.
Since the dredging program extends over nearly 30 miles, with potentially impacted
downstream water supplies as far away as 100 miles from the TI Dam, the far-field
monitoring program will consist of several major monitoring locations that can be readily
and regularly occupied to obtain water column samples for PCB analysis. It is important
to measure the PCB concentrations and the PCB mass loading from each of the river
sections. In addition to showing how much mass is exported from each of the river
Hudson River PCBs Superfund Site
Engineering Performance Standards
63
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
sections, the size of the region subjected to the PCB export can be determined.
Additionally, water treatment plants downstream can be notified in the event of a large
release.
3.3.1 Far-Field Concerns
Because of the importance of the Hudson River as a public water supply and the need to
assure public safety, daily samples will be collected at all far-field monitoring stations.
Discrete samples will be collected from each station to represent the entire river cross
section (e.g., an equal-discharge representation of the river's cross section). The samples
must be collected to represent the dredging period. That is, samples from an affected
water parcel at each far-field station must be collected. Without consideration for time-
of-travel between the remedial operations and the representative far-field station, false
low values may be obtained and potentially large releases may go unidentified, even
though samples will be collected daily under routine monitoring. (Note that this does not
imply the requirement of strict time-of-travel sampling, only that the samples should be
collected when it can be reasonably expected that dredging-related water quality impacts
can be captured by sampling at each downstream far-field station.) The daily discrete
routine monitoring will include the following variables:
17
Total and Tri+ PCBs (whole water , congener-specific, low detection limits)
Suspended solids
Dissolved organic carbon (DOC)
Organic carbon on suspended solids (weight loss on ignition on suspended solids,
or similar measurement)
Temperature
pH
Dissolved oxygen (DO)
Conductivity
In situ probes are required for the following:
Turbidity (continuous)
Suspended solids size distribution via a particle counter (continuous at nearest far-
field station only)
The discrete samples for PCBs are clearly required to document compliance with the far-
field action level criteria and the Resuspension Standard threshold. The suspended solids,
DOC, and organic carbon on suspended solids are all needed to support the interpretation
of the PCB data, particularly when action levels are exceeded. The continuous reading
parameters are needed as supporting information to confirm a minimal impact of
17 Whole water samples require separation of dissolved and suspended matter fractions for separate
extraction. Extracts may be combined into a single analysis.
Hudson River PCBs Superfund Site
Engineering Performance Standards
64
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
dredging on water quality as well as to prompt additional PCB sampling in the event of a
substantive suspended solids release.
The daily discrete monitoring parameter analytical methods must be sufficiently sensitive
to avoid non-detect values at most stations and provide data that can characterize PCB
concentrations during both routine and unusual conditions. In general, the analytical
methods chosen for the monitoring program must meet or exceed the specifications of the
methods used to develop the baseline water column concentration data. As discussed in
further detail in the next section, the frequency and type of samples will be adjusted as
action levels are exceeded. For example, the frequency of PCB sampling will be
increased to as often as four times per day.
In addition to the daily discrete sample collection, two other forms of sampling will be
included at these stations. Specifically, continuous suspended solids monitoring by means
of turbidity and particle counters and the use of an integrating PCB sample (e.g., an Isco
sampler) will also be required. A surrogate relationship must be developed for suspended
solids using a real-time measurement (turbidity or particle counter). These measurements
will be conducted continuously and recorded on a regular basis for use within the same
day. The surrogate relationship must be developed prior to Phase 1 and maintained
throughout the program.
An integrating PCB sampler will be required as well to provide an alternate measurement
basis for water column PCB concentrations. These sampling techniques provide a useful
integration of water column loads over time and can be compared to historical
measurements (to be collected during the remedial design) or simply to the prior months'
data. The data from the integrating PCB sample can be used to document changes in PCB
export from the dredging operations to the extent the changes occur in between daily
discrete samples. The results can be compared to the more quantitative but instantaneous
daily measurements of PCB concentration to generate a rough estimate of PCB transport.
More importantly, these samplers provide a long-term integration of PCB load,
monitoring the relatively long periods of time between the daily sampling events. This
information serves to confirm that river conditions as captured by the daily discrete
samples are representative of general river conditions. These samplers do not provide
real-time data but rather confirm that the discrete samples are providing a useful measure
of average conditions. These samplers will be deployed in a manner similar to the regular
water column points, (i.e., multiple points in the river cross section will be sampled to
obtain a representative sample where possible). These samples will be collected biweekly
at the five Upper River main stem stations from Rogers Island to Waterford.
3.3.2 Near-Field Concerns
Local variation prevents useful monitoring of PCBs in the immediate vicinity (near-field)
of the dredging operation. From the float studies conducted by GE in the late 1990s, it is
clear that the PCB concentrations in the water column can increase greatly over relatively
short distances from exposure to the contaminated sediments. Near-field downstream
Hudson River PCBs Superfund Site
Engineering Performance Standards
65
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
monitoring of the PCB concentrations cannot distinguish between the contribution
resulting from resuspension during dredging and the contribution from the sediments.
Additionally, the time lag between sample collection and the availability of PCB data
(normally at least 24 hours, even with an accelerated turn-around time) preclude the use
of PCB measurement as a real-time monitor of dredging operations. The ROD
acknowledges that the water quality may be reduced in the short term, to allow the
remediation to result in attainment of the long-term goals. Therefore, it is not useful to
implementconcentration criteria in the near field, given the high spatial and temporal
variability as well as the delayed receipt of information.
The near-field monitoring program is designed to
provide a real-time measure of conditions around the
dredging operation. It is designed recognizing that the
far-field monitoring program cannot provide direct
feedback to the dredge operators concerning the day-
to-day operation of the equipment and engineering
controls. For this reason, near-field monitoring will
entail continuous measurement of turbidity through the use of electronic sensors (see
Attachment F) to allow real-time response to changing conditions and dredge operator
activities.
A surrogate relationship between turbidity and suspended solids must be developed and
maintained throughout the program. Suspended solids samples will be collected daily to
assess the predictive capability of the surrogate relationship. Suspended solids sampling
only increases to once per three hours if the surrogate relationship fails to provide a
sufficiently conservative estimate of the TSS concentrations. The criteria for the
surrogate relationship are provided in Section 4.
The near-field monitoring program is not intended to provide quantitative measures of
PCB loss from the dredging operations but rather to provide a more sensitive qualitative
measure of the possible impacts of various dredging activities. These results will be used
in coordination with far-field turbidity, suspended solids, and PCB monitoring so that
acceptable levels of near-field turbidity can be developed from the net effects observed
downstream.
The near-field monitoring program will include suspended solids and turbidity
monitoring both upstream and downstream of the dredging operation, so that dredging-
related turbidity and associated suspended solids can be identified. Sensors will be
deployed at specific distances downstream of the dredging operation that have been
determined based on information available in the literature as well as on results of the
near-field modeling analysis described in Attachment D. In addition to direct sensor
measurements, daily discrete particle counter suspended solids measurement will also be
collected in situ to provide analytical confirmation of the sensors.
Using electronic sensors for
continuous measuring of turbidity
the near-field monitoring
program will provide a real-time
gauge of conditions around the
dredging operation.
Hudson River PCBs Superfund Site
Engineering Performance Standards
66
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The near-field monitoring program provides the best
opportunity to obtain real-time results that can be used to
guide the dredging operations and to identify activities that
may result in unacceptable releases of PCBs from the
sediments. While PCB monitoring is the ultimate measure of
downstream impacts, the real-time turbidity and suspended
solids monitoring provides the best means of minimizing
suspended solids and PCB release.
While the use of turbidity or suspended solids monitoring provides valuable real-time
data, there are some issues that need to be considered in the design of the monitoring
program and interpretation of the data. Besides the straightforward issues of sample
accuracy and representativeness, the installation of backfill concurrent with the dredging
operation may serve to confound the turbidity and suspended solids signals. To the extent
that backfill creates large amounts of turbidity, it is possible that the contribution of
dredging-related turbidity or suspended solids may be indiscernible. The expected close
proximity of dredging and backfill operations will make it difficult to estimate the
suspended solids load upstream of dredging but downstream of the backfilling. Thus,
measurement of the local impact of dredging by suspended solids monitoring may be
compromised. This is addressed to the extent possible by placing a suspended solids and
turbidity monitoring station just upstream of each dredging operation. It is, however,
expected that backfilling operations will not always coincide with dredging, which would
simplify the suspended solids monitoring during these intervals.
Further refinement of the near-field and far-field suspended solids criteria is anticipated
at the completion of Phase 1, and possibly during Phase 1 if appropriate. Pending the
results of Phase 1, suspended solids criteria may be developed that may require
engineering evaluations or improvements. (As currently constructed, the Resuspension
Performance Standard only requires an engineering action in response to PCB-based
exceedances.)
In summary, both PCB and suspended solids monitoring have limitations that affect the
usefulness of the data. For PCBs, the time lag between sampling and the availability of
the data as well as the baseline release of PCBs limit the measurement sensitivity. For
suspended solids, the near-field heterogeneity, the sensitivity of the surrogate
measurement and the impact of backfilling resuspension potentially confound the
measurement. Nonetheless, these measures taken together can provide a rigorous basis on
which to monitor downstream transport and compliance with the Resuspension Standard.
Real-time turbidity and
suspended solids
monitoring provides the
best means of minimizing
suspended solids and
PCB release.
Hudson River PCBs Superfund Site
Engineering Performance Standards
67
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
3.4 Data Quality Objectives
The monitoring plan for the Resuspension Performance Standard is summarized in
Tables 1-2, 1-3 and 1-4. The main objectives of the monitoring plan are described in the
following subsections, along with the techniques intended to satisfy these objectives. This
analysis represents an initial analysis of the DQOs that will undergo subsequent
refinement during preparation of the quality assurance plans for dredging-related
monitoring. As such, it is expected that the monitoring requirements developed for the
standard represent a minimum level of monitoring and that additional sampling beyond
these requirements will be needed to completely understand the nature of any dredging-
related release.
These monitoring requirements, therefore, are primarily intended to document
compliance with the various criteria of the Resuspension Standard. Special studies, as
outlined in Section 4.0 will provide information to verify assumptions made about the
behavior of the contaminant releases due to dredging (e.g., PCB dissolution, suspended
solids settling and dissipation). Information collected to verify these assumptions during
the Phase 1 period should serve to improve the monitoring program during Phase 2 in
several ways. The Phase 1 data should permit identification of the most effective
monitoring locations and monitoring techniques as well as those that are not useful. This
information may also permit a reduction in the frequency and complexity of monitoring
during Phase 2.
Subsections 3.4.1 to 3.4.3 contain a discussion of the main DQOs and a discussion of the
sampling approaches needed to satisfy each objective. Subsection 3.4.4 provides a
summary of the analyses that confirm the adequacy of the sampling frequencies required
as part of the routine and non-routine monitoring programs. More detail is provided in
Attachment G. These analyses, which conform to USEPA methods for assessing
statistical uncertainty (USEPA, 2000f), cover only routine monitoring and the minimum
levels of contingency monitoring as defined in the Resuspension Standard. Additional
monitoring related to the required engineering studies at the Concern and Control Levels
(as well as exceedance of the standard threshold) may be required, depending on the
anticipated cause of the exceedance. These may be conducted as a part of the engineering
evaluations. The design of these additional monitoring programs may occur during the
remedial design period. It is also likely that monitoring plans will need to be developed
by the contractor during the dredging operation in response to observations made at the
time.
A particular limitation to the analysis summarized in subsection 3.4.4 is lack of
information on the variance of river conditions in response to dredging-related releases.
Little data exist on which to develop the estimate of variance. As a result, the variation of
baseline conditions was used as a means to estimate the variance for dredging operations.
These estimates for sampling requirements and the associated error rates will require
review once additional data become available during Phase 1.
Hudson River PCBs Superfund Site
Engineering Performance Standards
68
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
3.4.1 Objectives for Far-Field Monitoring in the Upper Hudson
The far-field monitoring program in the Upper Hudson River addresses several DQOs.
This program is the primary monitoring effort for the protection of public water supplies
and for determining the magnitude of long-term PCB releases. Following the statement of
each data quality objective is a discussion of the sampling techniques to be used to satisfy
the objective.
3.4.1.1 Objective I
Provide a set of data to demonstrate compliance with the Total PCB concentration
components of the Resuspension Standard (Le., the 350 ng/L criterion for the
Control Levels and the 500 ng/L criterion for the standard threshold).
Dredging-related operations are expected to occur throughout the Upper Hudson
between Fort Edward and Waterford. Hence, dredging-related PCB releases may
occur over the entire region as well. In particular, while the majority of dredging
is focused north of Schuylerville, boat traffic and other operations are expected to
occur downstream of Schuylerville. Thus, PCB concentrations must be monitored
throughout the Upper Hudson River. Additionally, PCB release due to dredging is
not expected to be constant with time but is expected to vary substantively over
time. Thus, discrete grab samples collected at one station at one point in the day
may miss more substantial release events occurring at other times. As the river
carries these releases, natural mixing and dispersion will serve to homogenize
PCB concentrations to some degree, spreading them out and making it easier to
collect representative samples at locations farther downstream. Multiple stations,
therefore, provide the ability to capture conditions representing a longer period of
time.
Note that the desire to obtain many samples from the river to characterize
conditions must be tempered by the availability of laboratories to analyze the
samples. For this reason, sampling under routine conditions (expected to be the
majority of the conditions while dredging) will only require daily samples from
the far-field stations plus a limited number of longer-term integrated samples (see
Table 1-2). This consideration also recognizes the need to obtain and analyze
samples sufficiently rapidly to address Objective II below. An alternative to these
discrete samples is the collection of daily composite samples, integrated over a
24-hour period at each station. These samples still require the collection of a cross
section composite for each day. Additional sampling will be required if 24-hour
composites are collected when certain resuspension criteria are exceeded.
It is necessary to correctly characterize the PCB
concentration throughout the river cross section,
recognizing that both baseline and dredging-
related releases create heterogeneous PCB
concentrations. This has been extensively
demonstrated by the paired sample data
Because both baseline and
dredging-related releases
create heterogeneous PCB
concentrations, It is necessary
to correctly characterize the
PCB concentration throughout
the river cross section.
Hudson River PCBs Superfund Site
Engineering Performance Standards
69
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
collected at TID-West and TID-PRW2. For this reason, at least five points are
required at each sampling station, based on equal width or equal discharge
considerations as given by USGS guidance (USGS, 2002). Multiple points are
required for discrete samples as well as the alternative daily composite samples.
To support the use of discrete samples as representative of mean river conditions,
it is also necessary to obtain integrated samples. These samples will serve to
demonstrate compliance with the standard during periods between discrete
samples. Integrated samples will cover two-week intervals during routine
monitoring, providing a longer perspective on PCB transport and concentration
with relatively little increase in the total number of PCB analyses. Rapid
turnaround of results is not needed for the integrated samples because these
samples take longer to collect. The resulting PCB average concentrations provide
confirmation of data obtained from daily discrete samples. As such, these results
are needed during Phase 1 to provide supporting data for the discrete samples. If
the viability of the discrete sampling program is confirmed, these samples may be
dropped or greatly reduced during Phase 2.
Samples must be collected at sufficient frequency to provide a reasonable
statistical certainty that conditions are in compliance with the Resuspension
Standard criteria. Higher statistical uncertainty is acceptable when concentrations
are well below the standard criteria. As the action levels and the standard
threshold are approached, sampling frequency must be increased to provide
greater certainty that conditions are still in compliance. In particular, it is
important to minimize the false negative error, the error of accepting conditions to
be in compliance when in reality they are not. The issue of sampling frequency is
discussed in subsection 3.4.4 and Attachment G more extensively.
Analytical methods for the monitoring program must meet or exceed the
specifications for the baseline monitoring program to provide sufficient precision,
sensitivity, accuracy, and representativeness. The monitoring results from the
baseline program are a basis of comparison for the Resuspension Standard and
must be consistent.
3.4.1.2 Objective II
Provide a means to rapidly assess water column Total PCB levels so that the USEPA
can advise public water suppliers when water column concentrations are expected
to approach or exceed the federal MCL (ie., 500 ng/L) during the remediation.
In this manner, public water suppliers can take contingency actions, if needed, to
maintain safe water for their users. Appurtenant to this objective, determine the
relationship of dredging-related PCB contamination at the upstream far-field stations (TI
Dam and Schuylerville) to that at the downstream far-field stations (Stillwater and
Waterford) in order to use the far-field stations near the remediation as predictors of
downstream concentrations entering the public water intakes.
Hudson River PCBs Superfund Site
Engineering Performance Standards
70
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
There are several aspects of the monitoring plan that are required to achieve these closely
related objectives. These are described below.
Measurements of PCB concentrations at all Upper Hudson far-field stations are
needed on a daily basis to identify possible exceedances of the standard threshold
and any action level criteria. These data satisfy both components of this objective,
since the data will document the PCB concentrations and also serve as a database
to resolve the relationship between upstream and downstream PCB concentration
increases related to dredging.
Reduced turnaround time for PCB samples from the two far-field stations nearest
to the dredging operations is required. During Phase 1 these stations will probably
be TI Dam and Schuylerville, although the Phase 1 dredging area has not yet been
defined. The results from sampling at these stations will be used to assess the
need to warn the public water supplies that the concentrations entering the intakes
may be elevated. The travel time between remediation activities in River Sections
1 and 2 and the Waterford public water supply intakes is generally at least two
days, although during high flow events, the travel time is shorter. Thus, in order to
have information from the primary dredging areas in time to provide a warning to
the downstream intakes, a turn-around time of 24 hours or less is required for the
samples obtained from the two nearest downstream far-field stations. (Note that
because the turn-around time for PCB analysis is 24-hours, it is also important to
develop a real-time indicator of elevated contaminant levels.)
While actual PCB measurements provide the most certain basis for assessing PCB
loads and concentrations, these cannot be obtained in real-time. Resuspension of
contaminated sediment is thought to be the primary mechanism of dredging-
related contamination release. When verified, suspended solids monitoring
provides one of the best means of warning the public water supplies of potential
exceedances, since it provides the longest lead time between knowledge of the
release and its arrival at the downstream intakes.
Additionally, as the dredging operations move farther downstream, suspended
solids monitoring will provide the only real-time data for the protection of
downstream impacts. Specifically in River Section 3, there will be insufficient
time to collect, analyze, and evaluate a PCB sample and still warn the
downstream intakes. As a result, the standard requires that a surrogate measure of
suspended solids concentrations (turbidity or laser particle counter) be developed
and maintained throughout the remediation. Samples will be collected once a day
for suspended solids analysis to provide confirmation of the surrogate results.
Each week, the measured suspended solids results will be compared to the
predicted values to determine if the surrogate is providing sufficiently accurate
results, based on a statistical analysis. (See Section 4 for details of these special
studies.)
Hudson River PCBs Superfund Site
Engineering Performance Standards
71
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Frequent suspended solids measurements (every three hours) will only be required
when the surrogate measurements for suspended solids are not providing a
conservative measurement. A modified method for suspended solids analysis will
be specified to permit a short turn-around time (three hours). Co-located samples
will be collected for both the modified method for suspended solids and the
unmodified method that is based on ASTM 3977-97 once a day as to assess the
accuracy of the modified method.
At the far-field stations, monitoring for suspended solids via a surrogate is
conducted on a 24-hour-per-day basis.
3.4.1.2.1 Objective III
Provide a set of data to demonstrate compliance with the Total PCB load
components of the Resuspension Standard (ie., 300 g/day and 600 g/day).
As stated in subsection 3.4.1.1 Objective I, dredging-related operations are
expected to occur throughout the Upper Hudson between Fort Edward and
Waterford, increasing PCB loads as well as concentrations. PCB loads, however,
represent a longer-term concern since the associated impacts will take longer to
occur and thus require a sustained level of loading in order to occur. A high
frequency of monitoring in Phase 1 can provide an opportunity to identify
substantive increases in load soon after occurring so that the root cause can be
identified and long term impacts avoided. To this end, the monitoring frequency
required to satisfy the concentration criteria is expected to also satisfy this
objective.
Since PCB loads over time are the primary concern of this objective, it is
desirable to obtain integrative samples for this objective as well. For this reason,
integrative samples will be obtained at the four main far-field stations during
Phase 1, as discussed under Objective I. These will provide confirmation of the
initial conclusions drawn regarding PCB loads based on the more frequent
discrete samples.
Data on river discharge is also needed to address load considerations. Data from
the USGS stations at Fort Edward and Waterford will be used to this end. In the
event that the USGS discontinues these stations, data on flow must be obtained by
an alternate means. Additional data on meteorological conditions must be
obtained to supplement the USGS data and permit an accurate representation of
flows at the stations not monitored by the USGS.
Sample collection must be timed to capture the
impacted water column. If samples were collected each
day from the nearest far-field station at the onset of the
operations, it is unlikely that the water collected would
show the dredging-related impacts. The plume will
Sample collection
must be timed to
capture the impacted
water column.
Hudson River PCBs Superfund Site
Engineering Performance Standards
72
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
widen and lengthen as it travels downstream, making it more likely that the
downstream stations will capture dredging-related impacts. (This is not, however,
a time-of-travel sampling. Although the parcel of water sampled must be
impacted, the same parcel of water need not be tracked as it passes down the
river.)
Equal discharge increment (EDI) or equal width increment (EWI) sampling as
defined by USGS will be required. This type of sampling method is required to
capture a representative cross-sectional sample. A single center channel station
will not be sufficient, because extensive natural mixing across the channel is
unlikely in most of the Upper Hudson and plumes confined to the shoreline by
river hydrodynamics will not be accurately represented, resulting in low-biased
results.
3.4.1.3 Objective IV
Determine the primary means of PCB release via dredging-related activities. (Verify
that dissolved phase releases are minimal as estimated by modeling and that the
primary mechanism of release is suspension of sediment.)
During the public comment period on the Hudson River ROD, concerns were
raised that dredging of PCB contaminated sediment could release a substantial
amount of dissolved-phase PCBs. Calculations to determine whether and how
such a release could occur (Attachments C and D) have indicated that this
scenario cannot occur and that the primary release mechanism would be
resuspension of contaminated sediment. This mechanism would be accompanied
by an increase in suspended solids concentration and could be tracked in the near
field.
Though convincing, the calculations done to
determine the primary mechanism of release
need to be verified in order to be certain that the
goals of the ROD can be achieved (long-term
recovery of the river, protection of the
environment and human heath). This will be
accomplished by a special study, which will
characterize dissolved-phase and suspended-phase Total PCB concentrations in
the vicinity of dredging operations. Several of these studies will be conducted to
characterize the releases for various concentration ranges, sediment types and
dredging equipment. Samples will be collected daily for a week at each selected
location to provide a sufficient number of samples given the high degree of
variability in the near-field conditions. More details of this special study are
provided in Section 4.
The objective of the special study is to determine whether there is a substantial
dissolved-phase release from the remedial operations consistent with that
Hudson River PCBs Superfund Site 73 Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Resuspension - April 2004
Dredging is not expected to
release substantial amounts of
dissolved-phase PCBs directly.
A special study will assess the
primary release mechanism in
the vicinity of the dredging
operation.
-------
estimated by the USGS at the Fox River site. The study will not be designed to
quantify a low-level dissolved-phase release; hence, it will not be necessary to
extensively characterize the baseline conditions in these areas. A station upstream
from the remedial operations will be monitored for comparison.
Additional parameters will be required to aid in the interpretation of the split
phase data. Dissolved organic carbon, suspended organic carbon, suspended
solids, turbidity, and temperature provide an indication of the distribution of
dissolved-phase and suspended-phase PCBs. These parameters will also be
measured for the discrete samples collected during routine monitoring and
contingency monitoring. In this manner, changes in these supplemental
parameters may help identify the nature of the mechanism responsible for the
PCB release throughout Phase 1, assuming equilibrium has been reached.
3.4.1.4 Objective V
Determine the baseline Total PCB levels entering River Section 1 from upstream
sources.
PCBs entering River Section 1 should be identified so as to differentiate these
additional concentrations from the releases occurring downstream. Based on
monitoring data from the past five years, PCBs have been at non-detect or low
concentrations entering River Section 1. However, changes in upstream
conditions such as construction at the source areas could result in higher PCB
concentrations entering the TI Pool. Monitoring at Bakers Falls and Rogers Island
for PCBs will be required to identify such situations. If the contribution from
upstream sources were to increase, the Bakers Falls and Rogers Island results
should document this occurrence and provide a basis to adjust the dredging-
related load contribution.
This information will help to avoid an unnecessary enforcement of the
engineering or monitoring contingencies of the standard and should be done on a
case-by-case basis. The sampling frequency will be once per week at Bakers Falls
and once per day at Rogers Island. With USEPA's approval, the frequency at
Rogers Island may be further reduced if these concentrations are shown to be
consistently low relative to dredging-related releases.
Both Bakers Falls and Rogers Island stations are needed for this purpose. An
important assumption in the ROD was the continued reduction of the releases
from the GE Hudson Falls facility. Differences in PCB concentration and load
between these two stations will be used to document this process. In the event that
these data are collected as part of other remedial activities upstream of Rogers
Island, these data do not have to be duplicated by the dredging-related monitoring.
However, these data must meet the data quality objectives defined here and in the
subsequent quality assurance plans issued for the Resuspension Standard.
Hudson River PCBs Superfund Site
Engineering Performance Standards
74
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Detection limits for Total PCBs for these data must achieve equal or better
reporting limits as those achieved for the remedial design baseline monitoring
program, approximately 4.0 ng/L for an eight-liter sample. Lower reporting limits
{i.e., less than 4 ng/L) will be required if sample results at Rogers Island routinely
fall below the reporting limit since accurate quantitation of this load is an integral
part of the long-term monitoring program.
Additional data will be required to aid in the interpretation of downstream data.
Baseline levels of DOC, suspended organic carbon, suspended solids, and
temperature are needed to characterize the changes in these parameters that may
be caused by dredging-related activities. Dissolved oxygen measurements will be
taken at Rogers Island for the same purpose.
Since baseline conditions should not
change in response to dredging-related
releases, the frequency of baseline
monitoring does not increase in response to
action level or threshold exceedances.
3.4.1.5 Objective VI
Determine ancillary remediation-related effects on the river {e.g., barge traffic-
related resuspension, spillage during transit or off-loading of sediment) that may
occur in areas that are not captured by the nearest representative far-field station.
During Phase 1, the remediation will probably be limited to the TI Pool. Once the
material is dredged it will be conveyed to another location for further processing and
shipping to a landfill. This destination may not be in the TI Pool, resulting in transport of
contaminated material throughout stretches of the Hudson River by barge or pipeline. To
verify that the transport of material is not causing the release of PCB contamination to an
extent that would cause exceedance of the resuspension criteria, sampling will be
required at each Upper Hudson River far-field station (except Bakers Falls) at least once
per day.
3.4.1.6 Objective VII
Verify that the water column PCB concentrations developed from the grab samples
adequately characterize the average concentration.
Discrete grab samples will be used for comparison to the PCB flux and
concentration resuspension criteria. The Resuspension Standard requires that
samples must be timed to capture the impacted water column, increasing the
likelihood that the samples will be representative of the dredging-related impacts.
As described in subsection 3.4.4, the sampling frequency is sufficient to compare
the results of the analyses to the resuspension criteria with confidence, but this
analysis is based on an assumption of the variability of the water column
concentrations. This estimate of variability is derived from the baseline
Baseline conditions should not
change in response to dredging-
related releases; thus baseline
monitoring frequency does not
increase in response to action
level or threshold exceedances.
Hudson River PCBs Superfund Site
Engineering Performance Standards
75
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
conditions, which do not include the added variability of the dredging-related
releases. This added variability could change within a day as different operations
are completed and different dredge operators are employed.
To verify that the grab samples are sufficiently indicative of average river
conditions, integrating samplers are required for deployment periods ranging from
two weeks under routine monitoring to one day under Control Level monitoring.
Integrating samplers cannot entirely replace the required grab samples at TI Dam
and Schuylerville, even if all other DQOs are met by this sampling method,
because it will be important to have some measure of the upper and lower bound
concentrations that are occurring in the river as well as the average condition near
the remedial operations.
Integrating samplers are required for daily measurements in place of discrete grab
sampling at Stillwater and Waterford at the Concern and Control Level
monitoring as well. This sampling method is used because of the concern that the
water column concentrations are approaching the MCL. Integrating samples are
used here instead of multiple grab samples to reduce the overall number of PCB
analyses while still obtaining data on PCB concentrations over a 24-hour period.
3.4.1.7 Objective VIII
Confirm the exceedance of the action level criteria as well as the standard criterion.
Sampling frequency must be increased to verify exceedances of the resuspension
criteria. At lower levels of exceedance, the consequences of error in deciding
whether the resuspension criteria have been exceeded are less serious than at
higher levels of exceedance. Hence, a higher level of decision uncertainty is
acceptable at exceedances involving the lower action levels. At the Evaluation
Level, the concern is adherence to best practices and long-term PCB release
impacts, concerns that do not require a rapid {i.e., 24-hour) response. At PCB
concentrations close to or above the Resuspension Standard, public water supplies
could be impacted and a shutdown of the dredging operations may be required.
Thus, a greater level of certainty is required when the consequences are greater.
This is the primary reason for requiring increased frequency of sampling in the
standard. The development and level of certainty provided by the various
sampling regimes are further discussed in subsection 3.4.4.
An increase in monitoring frequency will be required as a contingency at the two
representative far-field stations during Phase 1. These stations provide the best
opportunity to document river conditions in response to dredging-related releases
and also provide a warning to downstream public water supply intakes. With the
uncertainty related to dredging-related releases, the second station will confirm
the observations of the nearest far-field station and thus provide a sound basis for
whatever response actions are required.
Hudson River PCBs Superfund Site
Engineering Performance Standards
76
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Monitoring of the downstream far-field stations (Stillwater and Waterford) for
PCBs will be changed to daily integrated sampling to capture the average
concentrations that would be entering the public water supply, while PCB
concentrations collected from stations nearer to the remediation may be
approaching the resuspension standard threshold. Data from the integrated far-
field samples provide further subsequent confirmation of the estimated
concentrations based on conditions closer to the dredging operations.
Results from these downstream stations can be used to refine the means of
predicting the PCB concentrations that will enter the public water supplies based
on the concentrations measured nearer to the remediation. These results will
indicate the degree to which the PCB concentrations dissipate as the water column
passes downstream. The switch from a daily discrete sample to an integrated
sample reflects the need to characterize the entire day's water conditions while
minimizing the number of samples collected, so that results can be made rapidly
available and interpreted.
3.4.1.8 Objective IX
Confirm Alternate Monitoring Programs.
The monitoring program outlined in Tables 1-2, 1-3 and 1-4 has been constructed around
the standard. It may be possible to employ alternate monitoring techniques. However, the
ability of alternate monitoring programs to achieve the data quality objectives must be
demonstrated. Modifications to the resuspension criteria and required actions if exceeded
may be required as well in response to the changes. This will be the subject of a special
study. Details are provided in Section 4.
3.4.1.9 Objective X
Measure the parameters with precision, accuracy, representativeness,
comparability, completeness and sensitivity that is equivalent to the baseline
monitoring program specifications.
Analytical methods for the monitoring program must meet or exceed the
specifications for the baseline monitoring program to provide sufficient precision,
sensitivity, accuracy, representativeness, comparability, completeness and
sensitivity. The monitoring results from the baseline program are a basis of
comparison for the resuspension standard and must be consistent.
Sample collection and sample handling must be consistent with the approach
taken during baseline.
As verification of these methods it will be necessary to have performance
evaluation PE samples. The purposes of these samples will be to determine that
the results for multiple laboratories are consistent in terms of both accuracy and
Hudson River PCBs Superfund Site
Engineering Performance Standards
77
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
precision. The PE samples will be used to verify that the congener distribution
identified among the laboratories is consistent.
An exception to this objective will be the specification of a modified analytical method
for suspended solids that will permit the results to be available in three hours.
3.4.2 Objectives for Near-Field Monitoring in the Upper Hudson
3.4.2.1 Objective XI
Provide a real-time indication of suspended solids release in the near field.
A real-time indication of the amount of suspended solids in the water column in the near
field will aid the dredge operators in minimizing the release of suspended solids and
associated PCBs during the remediation. This monitoring will also provide the earliest
evidence for a substantive PCB release and allow further response by direct PCB
measurements downstream. To this end, turbidity monitors will be placed around each
dredging or debris area undergoing remediation. Information from these monitors will
provide continuous feedback to the operators, allowing real-time adjustments to be made
to the operations as needed.
3.4.2.2 Objective XII
Determine the amount of suspended solids released by the remedial operations to
provide an indication of PCB export.
Calculations presented in Attachment C indicate that the primary release
mechanism of dredging-related contamination is resuspension of contaminated
sediment. Thus, an increase in suspended solids should correlate with an increase
in PCB contamination. The standard requires that a surrogate relationship be
developed for suspended solids concentrations in the near field and maintained
throughout Phase 1. Samples will be collected daily for suspended solids analysis
as a means of assessing the surrogate relationship. Samples will be collected twice
daily for suspended solids analysis if there is an exceedance of the suspended
solids criteria. This increase in susppended solids sampling is limited to the
noncompliant nodes. If the continuous reading surrogate (e.g., turbidity) fails to
adequately predict suspended solids concentrations, samples will be collected
every three hours for suspended solids analysis until an adequate surrogate
relationship is developed. More details are provided in Section 4 on this special
study.
Near-field sampling is limited to the hours of operation, with some pre- and post-
dredging sampling.
Exceedance of the near-field criteria prompts limited far-field sampling for PCB
analysis (and supporting analyses) at the nearest downstream representative far-
Hudson River PCBs Superfund Site
Engineering Performance Standards
78
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
field station. These data, combined with the results of the far-field PCB analytical
results, can be used to develop a relationship between suspended solids and PCB
concentrations, and also provide a means of adjusting the suspended solids-based
resuspension criteria, although a predictive correlation is not expected due to the
heterogeneity of the sediment concentrations.
3.4.2.3 Objective XIII
Verify that the NYSDEC surface water quality regulations are not violated during
the remediation.
NYSDEC has water quality standards for pH and dissolved oxygen (DO). At both the
near-field and far-field stations, pH and DO will be monitored discretely each time a
sample is collected. These parameters, plus conductivity, will also provide a measure of
quality assurance for the data collected.
3.4.3 Additional Monitoring Objectives
3.4.3.1 Objective XIV
Monitoring in the Lower Hudson: Determine the extent of short-term impacts to the
Lower Hudson River and examine the effect of Upper Hudson dredging activities on
Lower Hudson PCB concentrations.
The monitoring program for the Lower Hudson is designed to measure the short-
term impacts to the freshwater portion of the river (previously referred to as the
Mid-Hudson River during the Reassessment) resulting from the remediation. The
sampling requirements in the Lower Hudson are not designed for comparison to
the resuspension criteria. This is addressed by the frequent sampling at Waterford,
which will be extrapolated to conditions downstream.
Requirements for additional monitoring at the public water supply intakes will be
prepared as part of the community health and safety plan (CHASP) currently
under public review. The Lower Hudson stations are intended to characterize
general water column conditions in response to elevated PCB concentrations and
loads originating from dredging. These stations consist of a single center channel
location that can be readily reoccupied. Cross sectional sampling is not required,
since flow is not unidirectional and thus flux cannot easily be estimated.
The frequency of sampling is increased in the
Lower Hudson in response to greater loads
and concentrations in the Upper Hudson, e.g.,
when the concentration at Troy is expected to
exceed 350 ng/L Total PCBs. This is done to
examine Lower Hudson conditions in
response to these loads, part of documenting recovery of the river.
The monitoring program for the
Lower Hudson will measure
short-term impacts to the
freshwater portion of the river
(referred to as the Mid-Hudson
River during the Reassessment)
resulting from the remediation.
Hudson River PCBs Superfund Site
Engineering Performance Standards
79
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
3.4.3.2
Objective XV
Verify the selection of the monitoring locations.
The locations of the far-field and near-field monitoring stations were selected based on
several considerations, including near-field and far-field monitoring, ease of access, and
level of planned dredging activities. The suspended solids and PCB analyses will be used
to verify that these locations are appropriate. Monitoring of the far-field station less than
one mile from the remediation will be required even though the PCB measurements will
not be used for comparison to resuspension criteria during Phase 1. These results will
determine whether the station is heavily impacted by the nearby remediation and will
validate the requirement that far-field stations be more than one mile from the
remediation. (Monitoring for compliance with the far-field net suspended solids
resuspension criteria will be required each day, no matter how close the remediation is to
the far-field stations.)
3.4.3.3 Objective XVI
Non-Target Area Monitoring: Determine the degree and extent of contamination
resulting from the remedial operations downstream from the target areas.
A special study will be conducted to measure the amount of resuspended material
that has settled in the immediate downstream areas and is a potential source of
future contamination of the water column and downstream surficial sediment. The
primary DQO for this study is to determine the extent of contamination in terms
of spatial extent, concentration and mass of Tri+ PCB contamination deposited in
non-target near-field areas downstream from the dredged target areas.
This study is needed because contaminated material may be disturbed by the
remedial operations and move downstream along the bottom of the river, only to
be identified by the water column monitoring during the next high flow event of
sufficient force to transport the material. The near-field suspended solids
monitoring is not conservative with regard to this issue because these criteria are
based on the assumption that a single dredge meets the full production, when it is
likely that several dredges will be required. Resuspension due to several dredges
can theoretically create more local deposition because of settling between dredge
operations. The near-field suspended solids criteria were established based on a
single large plume since this approach is conservative for PCB dissolution and
thus maximum PCB transport. Therefore, it is not sufficient to assume that
compliance with the resuspension criteria means that the loss from the remedial
operations will not create an unacceptable degree of contamination downstream.
To address this objective, a special study will be conducted to measure the
amount of resuspended material that settles in the downstream areas and that may
act as a potential source of future contamination to the water column and
downstream surficial sediment. Each study areas will be located downstream of a
Hudson River PCBs Superfund Site
Engineering Performance Standards
80
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
dredging area and will be approximately five acres in size. Samplers (e.g.,
sediment traps) will be installed at multiple locations prior to the start of the
dredging in the area under study. The exact number of locations per study area
will be determined as part of the sampling plan development. At each sample
location two or more co-located sediment traps will be deployed. Sediment
accumulated in one of the samplers at each location will be collected and sent for
analysis once the dredging in the area is completed, including any redredging
attempts. Sediment accumulated in the second set of samplers will be collected at
a higher frequency (perhaps weekly) to examine the relationship between various
dredging operations and sediment accumulation.
The study will be conducted at several target areas to determine the degree and
extent of contamination over a range of conditions. The selected areas must
represent a range of sediment textures, contamination levels and remedial
equipment.
Measurement techniques will include suspended solids mass, PCBs, and other
pertinent variables. The techniques employed will meet or exceed the
specification for the analytical and sampling methods with the SSAP.
3.4.4 Statistical Justification of the Sampling Frequency
The adequacy of the sampling frequencies required as part of the routine monitoring
programs was examined using the USEPA defined methods for assessing statistical
uncertainty (USEPA, 2000). The analyses cover only routine monitoring and the
minimum levels of contingency monitoring as defined in the Resuspension Standard.
The final sampling requirements for the standard were developed using USEPA's
Decision Error Feasibility Trials Software (DEFT) (USEPA, 2001), a program to
estimate sampling requirements based on a project-specific error rate. The results of this
analysis are presented in Table 3-2.
As defined in DEFT:
A false acceptance decision error occurs when the sample data lead to a
decision that the baseline condition is probably true when it is really false.
A false rejection decision error occurs when the limited amount of sample
data lead to a decision that the baseline condition is probably false when it
is really true.
The gray region is a range of true parameter values within the alternative
condition near the action level where it is "too close to call."
Hudson River PCBs Superfund Site
Engineering Performance Standards
81
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The analysis of the various criteria and acceptable gray region around each criterion
yielded the results shown in Table 3-2. The table is organized by measurement type {i.e.,
PCB and suspended solids). False acceptances were minimized because this is the more
serious error. For all criteria except the confirmation of the 500 ng/L exceedance, the null
hypothesis assumed that river conditions were in compliance.
In general, decisions that are more critical {e.g.,
confirmation of exceedance of the Resuspension
Standard which requires shut down, or exceedance of
the Control Level, which requires intense monitoring
and engineering responses) require a large number of
samples and have greater certainty than the less critical
decisions. For the suspended solids measurements, it is
clear from this analysis that the implementation of a continuous monitor capable of
estimating suspended solids concentrations will be needed to have a reasonable amount
of certainty in these decisions. The low level of certainty is tolerable only because any
decisions made as a result of exceedance of the suspended solids will be confirmed by
measurements of PCB concentrations in the impacted water column.
Table 3-2 shows that the higher level of sampling associated with the higher action levels
and the Resuspension Standard yield low false error rates, as expected, reflecting the
need to be accurate before taking costly actions or improvements. In some instances, the
false rejection rate is fairly high, indicating that additional sampling may be
unnecessarily triggered. However, this represents a protective approach from the
perspective of the safety of the public water supplies. Additionally, the higher monitoring
rates will quickly confirm the need to remain at the action level thought to be exceeded.
Higher error rates are estimated in transitions from routine conditions to the Evaluation
and Control Levels, reflecting the relative low sampling rate required for routine
sampling. Also shown in the table is the one-week confirmation result {i.e., the error rate
for the combination of one week of routine monitoring and one week at the action level).
In each instance the false acceptance error is brought below 5%, thereby confirming the
need to sample at the higher rate or indicating that sampling at the routine rate may be
resumed.
The results for monitoring requirements for exceedance of the standard demonstrate the
need for the intensive sampling specified. In this instance, the river is assumed be in
exceedance of the standard. Four additional discrete samples (Table 3-2) do not provide
sufficient certainty given that the next day's decision will involve the temporary halting
of the dredging operations, a costly choice. However, by collecting hourly composites,
the power of the same four analyses is greatly improved and the 5% false acceptance rate
is attained. Table 3-2 also shows the results for the long-term integrative samples. These
samples will serve to confirm the results of daily routine monitoring or to demonstrate the
need for more frequent sampling. The results assume the automated collection of eight
samples per day over a one- to two-week period.
Decisions that are more
critical generally require a
large number of samples
and have greater certainty
than the less critical
decisions.
Hudson River PCBs Superfund Site
Engineering Performance Standards
82
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The results for suspended solids illustrate the need to
use a continuous sampling system such as a turbidity
probe. In the lower portion of the table, results for the
discrete sampling program are compared with those
that can be achieved with a continuous probe recorded
once every 15 minutes. In almost all cases, the
continuous reading probe provides more than an order
of magnitude of improvement in the expected error rate. Better rates can be achieved with
the continuous probes by simply recording more frequently.
Note that this analysis does not consider any uncertainty introduced by use of a probe
over discrete samples. Nonetheless, given a semi-quantitatve relationship between the
probe and actual suspended solids levels, it is highly likely that the probes will provide a
substantial reduction in the expected error rates for suspended solids monitoring, thereby
reducing unnecessary additional PCB sampling prompted by a false indication.
Table 3-3 contains the following information related to use of the automatic sampler:
Summary of the various criteria
Associated gray region
Sampling frequency required by the resuspension standard
False acceptance and false rejection levels for Total PCB sampling requirements
when the automatic sampler is used
Using the automatic sampler, the error rates for most of the sampling requirements are
less than 1%. The highest error rate was about 2% for the false rejection of the sampling
requirement from evaluation to control level. However, this value is still below 5% error
rate. This analysis shows that, theoretically, the power of the sampling program for Total
PCBs using automatic sampler is greatly improved. In actuality, an alternate monitoring
program that is primarily based on sample collection via automatic samplers will only be
as good as the implementation. There are numerous challenges associated with such a
program that must be carefully worked through during a special study. See Section 4 for
more information.
It has been demonstrated that
continuous reading turbidity
probes provide more than an
order of magnitude of
improvement in the expected
error rate.
3.5 Summary of Rationale
The rationale for the performance standard for PCB loss due
to resuspension has its basis in the goals outlined in the ROD
(USEPA, 2002a). The need to protect downstream fish and
fish consumers and the need to protect public water supply
intakes define the objectives for the standard. Action levels
were derived from consideration of ARARs for the site and
RAOs from the ROD, as well as the ability to detect a net
increase in PCB loads. These criteria were shown by modeling analysis to produce little
change in downstream fish tissue recovery, further supporting their use as action levels.
The rationale for the
performance standard
for PCB loss due to
resuspension has its
basis in the goals
outlined in the ROD.
Hudson River PCBs Superfund Site
Engineering Performance Standards
83
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Specifically, PCB releases commensurate with 500 ng/L Total PCB had no substantive
impact on the fish recovery once dredging operations were completed.
Ultimately the RAO concerning the transport of PCBs to the Lower Hudson provided the
lowest upper bound on the acceptable amount of PCB loss {i.e., 600 g/day Total PCB or
650 kg Total PCB over the entire period of dredging). Additional action levels were
needed to provide a tiered series of action levels with an increasing amount of
contingencies as the action levels are exceeded. The criteria, monitoring requirements,
and engineering contingencies are all designed with the intention of identifying and
correcting minor problems in the dredging operation while keeping the dredging
operation functioning smoothly and steadily.
Due to the variable conditions within the river over time, the Total PCB concentrations
may be greater than 350 ng/L Total PCBs, even though the load-based criteria are not
exceeded. This results from elevated baseline conditions and is most likely to occur in
May and June. The concentration-based action levels will govern, since these action
levels are intended to provide short-term protection for the downstream public water
supplies and therefore represent the more protective criteria in these instances. It is also
possible that the suspended solids criteria may indicate elevated PCB concentrations that
are not verified by subsequent PCB sampling and analysis. This is recognized in the
standard by requiring only additional sampling of the impacted water column at the
nearest representative far-field station for comparison against the resuspension criteria as
outlined in subsection 4.1.
Hudson River PCBs Superfund Site
Engineering Performance Standards
84
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
4.0 Implementation of the Performance Standard for Dredging
Resuspension
The Resuspension Performance Standard consists of the standard threshold and
associated action levels, monitoring requirements and engineering requirements. The
implementation of the action levels is described in subsection 4.1. Compliance
monitoring requirements including measurement techniques, monitoring locations and
other specifics are described in subsection 4.2. The procedures to revert to lower action
levels or routine monitoring are presented in subsection 4.3 The requirements for the
special studies are defined in subsection 4.4. For engineering contingencies, the
engineering evaluations, technologies for controlling releases that may be implemented,
and the requirements of the standard regarding engineering contingencies are described in
subsection 4.5. Reporting requirements are described in subsection 4.6.
Flowcharts depicting the implementation of the Resuspension Standard are provided in
Figures 4-1, 4-2 and 4-3 for the near-field suspended solids criteria, far-field Total PCBs
and far-field suspended solids. These flowcharts are a simplified depiction of the
interaction between the three aspects of the standard: action levels, monitoring and
engineering controls. To fully implement the Resuspension Standard the specifications
provided throughout this document must be upheld.
4.1 Resuspension Criteria
This subsection contains details of the implementation of the standard. Table 1-1
contains the requirements and criteria of the standard in tabular form. Implementation of
the Resuspension Standard will necessarily require monitoring for the parameters of
concern. Daily measurements of suspended solids and PCB concentrations can then be
compared with the appropriate action level or the Resuspension Standard threshold.
Load-based criteria require more than a simple measure of concentration, since flow must
be incorporated into the load estimate. Comparisons to the resuspension criteria must be
made on a daily basis for each of the Upper Hudson far-field stations. This will include
assessment of the load and concentration seven-day averages and the total load loss for
the season vs. the productivity rate.
Note that in the event that dredging occurs in more than
one river section, effectively creating two "nearest" far-
field stations, this standard is applied in the same manner
to both stations. That is, the near-field concentration
criteria apply to both stations equally. Given the various
uncertainties in load estimation, no "pro-rating" of the
standard for the upper station will be required, although
the dredge operators should consider doing so, as needed. This means that any of the far-
field stations can dictate response actions.
If dredging occurs in more
than one river section,
effectively creating two
"nearest" far-field stations,
this standard applies in the
same manner to both
stations.
Hudson River PCBs Superfund Site
Engineering Performance Standards
85
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The Total PCB load-based criteria will be assessed using the results of the baseline
monitoring program, which is scheduled to begin in 2004. Historical data collected prior
to the baseline period will be incorporated into the analysis of the baseline data if a
relationship between the historical and current baseline data can be developed. Estimates
of flow will be derived from USGS gauging stations currently operating in the Upper
Hudson, along with data from additional stations developed for this monitoring program
(e.g., Schuylerville). As noted previously, the load-based criteria will also be adjusted to
reflect the anticipated dredging period length with the maximum allowable net release of
650 kg Total PCBs18 or 220 kg Tri+ PCB over baseline for the duration of the
remediation.
Both of the action levels have associated load-based criteria. To simplify review of the
monitoring results and avoid additional computations during the remediation, the load-
based criteria will be converted to look-up tables of concentrations that correspond to
various load-based levels as a function of river flow and month. Examples of these tables
for Total PCBs at the Schuylerville station are included as Tables 4-1 and 4-2 for the
Evaluation Level and Control Level far-field net Total PCB load, respectively. The tables
were developed using the existing GE data for this location. However, as mentioned
previously, the existing water column data from the Upper Hudson are limited in
applicability,19 and were used to provide a preliminary set of values for these tables. Final
values for these tables for both Total PCBs and Tri+ PCBs will be developed from the
baseline monitoring program. Exceedance of the final values to be developed for these
tables from the baseline monitoring program for a given month and given flow will
constitute exceedance of the corresponding action level.
Both the Evaluation Level and Control Level contain far-field criteria based on 7-day
running averages. These averages are to be calculated daily for comparison against the
appropriate criteria on a daily basis. Similarly, both action levels contain near-field
suspended solids criteria based on 3-hour, 6-hour or daily running averages. These
averages are to be calculated throughout the day on a three hour basis to determine
compliance.
For all flux estimates, the load calculation may be corrected for contributions originating
upstream of the remediation area (i.e., above Rogers Island) in the event that loads from
this region are above the levels typically observed. See Section 4.1.2.7 for the means of
adjusting for a significant difference in the upstream loads.
In the event that dredging operations move to a location less than one mile upstream of a
far-field monitoring point, the next downstream far-field station becomes the
18 A daily rate of PCB loading can be derived consistent with the attainment of the recommended Target
Cumulative Volume as specified in the Productivity Standard and the cumulative PCB mass delivery to the
Lower Hudson. The daily load figure as well as an annual load goal should be prorated according to the
production rate planned in the Production Schedule to be submitted annually to USEPA.
19 Single point monitoring locations at Thompson Island Dam and Schuylerville or any of the far-field
stations are not adequate (i.e., not sufficiently representative of river conditions) for the purposes of
estimating recent baseline load conditions. A cross-sectional composite sample is required, as will be
obtained during both the baseline monitoring and the remedial monitoring programs for this purpose.
Hudson River PCBs Superfund Site
Engineering Performance Standards
86
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
representative far-field station for the operation. The nearer
far-field station will continue to be monitored, not to judge
compliance with the standard, but rather to evaluate the
requirement that the far-field station be more than one mile
from the remedial operations for the monitoring data to be
comparable to the resuspension criteria.
For exceedance of suspended solids criteria at either near-
field or far-field locations, the impacted water column must be sampled at the far-field
station to determine the concentration of PCBs in the plume. Suspended solids and
turbidity measurements collected from the representative far-field station will document
20
that the sample has been collected from the plume.
In the subsections that follow, the text describes the details of each of the action levels
and the threshold. Equations provided in the sections below are the basis for comparing
the monitoring results to the resuspension criteria.
4.1.1 Evaluation Level
4.1.1.1 Far-Field Net Total PCB Load
If dredging operations
move to less than one mile
upstream of a far-field
monitoring point, the next
downstream far-field
station becomes the
representative far-field
station for the operation.
I Ik- net increase in loial /'('/> ///
-------
F7 = Seven-day average load of Total PCBs at the far-field station due to
dredging-related activities in g/day
C= Flow-weighted average concentration of Total PCBs at the far-field
station as measured during the prior seven-day routine discrete
sampling in ng/L.
This is given as:
Cffi ~
7
J= 1 1
(4-2)
where:
C
ffi
The Total PCB concentration at the far-field station for day j. If
more than one sample is collected in a day due to exceedance of
the near-field or far-field criteria, the arithmetic average of all the
measurements will be used.
Qj
Q>/
The daily average flow at the far-field station for day j,
Estimated 95% upper confidence limit (UCL) of the arithmetic
mean baseline concentration of Total PCBs at the far-field station
for the month in which the sample was collected, in ng/L. Initial
estimates for these values are given in Table 4-3.
Q7
This value is determined from baseline monitoring data and
represents the upper bound for the average concentration at the far-
field station in the absence of dredging. Where the 95% UCL
varies within the 7-day period of interest (e.g., at the end of a
month), time-weighted average 95% UCL is calculated as the sum
of each day's 95% UCL dividing by the number of days.
Seven-day average flow at the far-field station, determined either
by direct measurement or estimated from USGS gauging stations,
in cfs
T d7
Average period of dredging operations per day for the seven-day
period, in hours/day, as follows:
2 Td
Tdi '
(4-3)
where:
The period of dredging operations for day j in
hours.
Hudson River PCBs Superfund Site
Engineering Performance Standards
88
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
If F7 is 300 g/day Total PCBs or greater, this is considered to be an exceedance of the
Evaluation Level. This formula is intended to identify a mean loading of Total PCBs due
to dredging in excess of the action level. The 95% UCL of the water column PCB
concentrations at each station and month is chosen to represent baseline concentrations
(Cw ), because this is a comparison to the average condition for a short duration. The
confidence limit indicates the probability or likelihood that the interval contains the true
population value.
Because the seven-day average value will be compared to the monthly mean, it is
appropriate to estimate the range of values that may contain the mean. Values that fall
outside this range are unlikely to be part of the original population of baseline values;
therefore, these PCB export levels are likely to represent a dredging-related release of
PCBs. Note that this and all PCB load standards may be adjusted for the production rate
as described in subsection 4.1.2.7.
4.1.1.2 Far-Field Net Tri+ PCB Load
The net increase in Tri+ PCB mass transport due to
dredging-related activities at any downstream far-field
monitoring station exceeds 100 g day day for a seven-
day running average.
Equations 4-1, 4-2, and 4-3 will be used to calculate the far-field net Tri+ PCB load at
each Upper River mainstem station on a daily basis by substituting the daily Tri+ PCB
concentrations and baseline Tri+ PCB 95% UCL values for the Total PCB
concentrations. Baseline Tri+ PCB concentrations have not been calculated for this
report, but the Tri+ PCB 95% UCLs will be calculated using the data collected during the
Baseline Monitoring Program. An F7 value of 100 g/day Tri+ PCBs or greater constitutes
an exceedance of the Evaluation Level.
4.1.1.3 Far-Field Average Net Suspended Solids Concentration
The sustained suspended solids concentration above ambient conditions at a
far field station exceeds 12 mg /.. To exceed this criterion, this condition
must exist on average for 6 hoars or a period corresponding to the daily
dredging period (whichever is shorter). Suspended solids are measured
continuously by turbidity (or an alternate surrogate) or every three hoars by
discrete samples.
The net increase in suspended solids concentration over baseline levels will be calculated
during the daily dredging period for each main stem Upper River far-field station. If the
suspended solids concentration is estimated continuously using a surrogate, the six-hour
running average net increase will be calculated throughout the daily dredging period. If
Hudson River PCBs Superfund Site
Engineering Performance Standards
89
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
the suspended solids concentration is measured by discrete samples at three-hour
intervals, the net increase will be calculated throughout the day for each six-hour interval
as the data become available from the laboratory. The suspended solids data must be
available within three hours of sample collection (three-hour turnaround time using a
modified method for suspended solids analysis). The net increase in suspended solids is
calculated as follows:
Net Increase in Suspended Solids (mg/L) = Cavg CbaseUne (4-4)
where:
CaVg = Arithmetic average suspended solids concentration for the time
interval at the far-field station
Cbaseiin e Arithmetic average baseline suspended solids concentration for the
same far-field station and month (based on the baseline monitoring
program results)
Suspended solids contributions from the tributaries will appear to be dredging-related
increases in suspended solids. This criterion may be waived with USEPA review if the
increase in suspended solids can be traced to meteorological events. The baseline
concentrations at each station will be developed from the results of the baseline
monitoring program.
The Evaluation Level is exceeded if the net increase in
suspended solids concentration is 12 mg/L or greater.
Exceedance of this criterion prompts Evaluation Level
sampling at one far-field station. The station will be
chosen to measure the Total PCB concentration in the
suspended solids plume in order to determine whether
additional actions need to be taken. Sample collection will be timed to measure the
concentration of PCBs in the impacted water column. The frequency of this sampling will
be equivalent to that defined in Table 1-2 for the representative stations (TI Dam and
Schuylerville). Only the grab sample will be collected for this purpose.
4.1.1.4 Near-Field Net Suspended Solids Concentration 300 m Downstream
the sustained suspended solids coiiccniralioii a how ambient conditions at a
location 300 hi downstream (i.e.. near-field monitoring) of the dredging
operation or /j<> ill downstream from any suspended solids control measure
(e.g.. silt curtain) exceeds KM) mg f for River Sections / and 3 and Cid mg /
for River Section 2. lo exceed this criterion, this condition miisi exist on
average jor six hours or for the daily dredging period (whichever is shorter).
Suspended solids are measured continuously by surrogate or every three
hours by discrete samples.
The Evaluation Level is
exceeded if the net increase in
suspended solids
concentration is 12 mg/L or
greater at any far-field station.
Hudson River PCBs Superfund Site
Engineering Performance Standards
90
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The net increase in suspended solids concentration between the upstream near-field
station and the downstream near-field stations will be calculated during the daily
dredging period for each remedial operation. Without barriers, these near-field stations
will be located approximately 300 m downstream of the dredge. With barriers, these
stations will be located approximately 150 m downstream of the barrier. If the suspended
solids concentration is estimated continuously using turbidity or any other surrogate, the
six-hour running average net increase will be calculated throughout the daily dredging
period. If the suspended solids concentration is measured by discrete samples at three-
hour intervals, the net increase will be calculated throughout the day for each six-hour
interval as the data become available from the laboratory. The suspended solids analysis
will require a three-hour turnaround time (using a modified method for suspended solids).
The net increase in suspended solids is calculated as follows:
Net Increase In SSnear_field = Cavg - Cup (4-5)
where:
Cup = The arithmetic average upstream near-field station concentration
over the time interval
Cavg = The arithmetic average downstream concentration over the time
interval. Samples will be collected from two stations located 300 m
downstream. The average concentration from each location over
the time period will be calculated separately and the higher average
concentration will be chosen for use in this equation
Evaluation Level exceedances are as follows:
River Sections 1 and 3: at a net increase in suspended solids concentration of 100
mg/L or higher
River Section 2: at a net increase in suspended solids concentration 60 mg/L or
higher
Exceedance of this criterion prompts Evaluation Level sampling at the nearest
representative far-field station. Sample collection will be timed to measure the
concentration of PCBs in the impacted water column.
Each near-field 300 m station (150 m station with barriers) will be compared to either
100 mg/L or 60 mg/L, depending on the location of the remediation during Phase 1,
while the behavior of the system is tested. In Phase 2, when multiple dredging operations
are conducted simultaneously within the same section of the river, the sum of the
concentrations measured at the near-field station may be compared to the criteria,
because this approach is in keeping with the development of the criteria. This criterion
may be waived with USEPA review if the increase in suspended solids can be traced to
meteorological events.
Hudson River PCBs Superfund Site
Engineering Performance Standards
91
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
4.1.1.5 Near-Field Net Suspended Solids Concentration 100 m Downstream and at
the Side Channel Station Without Barriers
The sustained suspended solids concentration above ambient conditions
at the near-field side channel station or the 100 m downstream station
exceeds 700 mg /.. To exceed this criterion, this condition must exist for
more than three hoars on average measured continuously or a confirmed
occurrence of a concentration greater than 700 mg/L when suspended
solids are measured every three hoars by discrete samples.
Without barriers, the average suspended solids concentration over the time period at the
upstream near-field stations for a remedial operation will be subtracted from the average
suspended solids concentration over the same time period at the 100 m downstream
station to get the net suspended solids concentration. Also, the average suspended solids
concentration over the time period at the upstream near-field stations for a remedial
operation will be subtracted from the average suspended solids concentration over the
same time period at the side channel station to get the net suspended solids
concentration.21
If the suspended solids concentration is estimated
continuously using turbidity as a surrogate, a three-hour
average net suspended concentration of 700 mg/L or
higher is an exceedance. If the suspended solids
concentration is measured by discrete samples at three-
hour intervals, two consecutive samples of 700 mg/L or
higher is an exceedance. Exceedance of this criterion prompts Evaluation Level sampling
at the nearest representative far-field station. Sample collection will be timed to measure
the concentration of PCBs in the impacted water column.
The net suspended solids concentration at each near-field 100 m station or side channel
station will be compared to 700 mg/L while the remediation is in Phase 1. In Phase 2,
when multiple dredging operations are conducted simultaneously within the same section
of the river, the sum of the concentrations measured at the near-field 100 m stations (or
side channel station) may be compared to 700 mg/L, because this approach is more in
conformance with the development of the criterion. This criterion may be waived with
USEPA review if the increase in suspended solids can be traced to meteorological events.
Exceedance of this criterion
prompts Evaluation Level
sampling at the nearest
representative far-field
station.
21 Note that this standard also applies to the 300 m station in the unlikely event that a 700 mg/L event
occurs at that location, but does not affect the 100 m and side channel stations.
92
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
4.1.2 Control Level
4.1.2.1 Far-Field Total PCB Concentration
I he / oial /'('/> concentration during dredgi tig-related
activities at any downstream Jar-Jicid monitoring sialion
exceeds Hi: / for a se veil-day running average.
The arithmetic average of the past seven days' monitoring will be calculated on a daily
basis for each of the Upper River mainstem far-field stations. For each station, a day will
be represented by a single value. If more than one sample is collected in a day for a
station, the arithmetic average of the Total PCB measurements for a station will be used
as a single day's concentration in the seven-day average. If the arithmetic average of the
Total PCB concentration is 350 ng/L or higher at a far-field station, this is considered to
be an exceedance of the Control Level.
4.1.2.2 Far-Field Net Total PCB Load
I he uel increase in loial /'('/> mass transport due la dredgmg-
relaied activities at any downstream Jar-held monitoring station
exceeds f>()l) g day on average over a seven-day period.
The far-field net Total PCB load will be calculated using Equations 4-1, 4-2, and 4-3 on a
daily basis. A seven-day Total PCB load of 600 g/day or higher constitutes an
exceedance of the Control Level.
4.1.2.3 Far-Field Net Tri+ PCB Load
I he net increase in In /'( H mass transport due to dredgmg-
relaied activities at any downstream Jar-field monitoring
station exceeds 201) g day on average over a seven-ilay period.
Equations 4-1, 4-2 and 4-3 will be used to calculate the far-field net Tri+ PCB load at
each Upper River mainstem station on a daily basis by substituting the daily Tri+ PCB
concentrations and baseline Tri+ PCB 95% UCL values for the Total PCB
concentrations. Baseline Tri+ PCB concentrations have not been calculated for this
report, but the Tri+ PCB 95% UCLs will be calculated using the data collected during the
baseline monitoring Program. An F7 value of 200 g/day Tri+ PCBs or greater constitutes
an exceedance of the Control Level.
Hudson River PCBs Superfund Site
Engineering Performance Standards
93
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
4.1.2.4
Far-Field Average Net Suspended Solids Concentration
The sustained suspended solids concentration above ambient conditions at
a far-field station exceeds 24 mg/L. To exceed this criterion, this condition
must exist for a period corresponding to the daily dredging period (six
hoars or longer) or 24 hoars if the operation runs continuously (whichever
is shorter) on average. Suspended solids are measured continuously by
suuroeate or everv three hoars bv discrete samvles.
The net increase in suspended solids concentration between far-field stations will be
calculated on a daily basis for each mainstem Upper River far-field station as soon as the
data become available (within 3 hours of sample collection). The net increase in
suspended solids concentration will be estimated for the daily dredging period (longer
than 6 hours) or for 24 hours if dredging is continuous. Equation 4-4 can be used to
calculate the net increase in suspended solids for the time period of concern.
Suspended solids contributions from the tributaries will appear to be dredging-related
increases in suspended solids. This criterion may be waived if the increase in suspended
solids can be traced to meteorological events.
The Control Level is exceeded if the net increase in
suspended solids concentration is 24 mg/L or
greater. Exceedance of this criterion prompts
Control Level sampling at one far-field station. The
station will be chosen to measure the Total PCB
concentration in the suspended solids plume in order to determine if additional actions
need to be taken. Sample collection will be timed to measure the concentration of PCBs
in the impacted water column. The frequency of this sampling will be equivalent to that
defined in Table 1-2 for the representative stations (TI Dam and Schuylerville). Only the
grab sample will be collected for this purpose.
4.1.2.5 Near-Field Net Suspended Solids Concentration 300 m Downstream
The sustained suspended solids concentration above ambient conditions at
a location 300 meters downstream (i.e., near-field monitoring) of the
dredging operation or 150 meters downstream from any suspended solids
control measure (e.g., silt curtain) exceeds 100 mg I. for River Sections 1
and 3 and 60 mg I. for River Section 2. To exceed this criterion, this
condition mast exist for a period corresponding to the daily dredging
period (6 hoars or longer) or 24 hoars if the operation runs continuously
(whichever is shorter) on average. Suspended solids are measured
continuously by surrogate or every three hoars by discrete samples.
The Control Level is exceeded if
the net increase in suspended
solids concentration is 24 mg/L
or greater at any far-field station.
Hudson River PCBs Superfund Site
Engineering Performance Standards
94
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The net increase in suspended solids concentration between the upstream near-field
station and the downstream near-field stations will be calculated during the daily
dredging period for each remedial operation. Without barriers, these near-field stations
will be located approximately 300 m downstream of the dredge. With barriers, these
stations will be located approximately 150 m downstream of the barrier. The net increase
in suspended solids concentration will be estimated for the daily dredging period (longer
than 6 hours) or 24 hours if dredging is continuous. Equation 4-5 can be used to calculate
the net increase in suspended solids for the time interval of concern.
Control Level exceedances are as follows:
River Sections 1 and 3: at a net increase of 100 mg/L or higher in suspended
solids concentration
River Section 2: at a net increase of 60 mg/L or higher in suspended solids
concentration
Exceedance of this criterion prompts Control Level sampling at the nearest representative
far-field station. Sample collection will be timed to measure the concentration of PCBs in
the impacted water column.
Each near-field 300 m station (150 m station without barriers) will be compared to either
100 mg/L or 60 mg/L, depending on the location of the remediation during Phase 1 while
the behavior of the system is tested. In Phase 2, when multiple dredging operations are
conducted simultaneously within the same section of the river, the sum of the
concentrations measured at the near-field stations may be compared to the criterion,
because this approach is in conformance with the development of the criterion. This
criterion may be waived if the increase in suspended solids can be traced to
meteorological events.
4.1.2.6 Far-Field Net PCB Seasonal Load
I he net increase /// /'('/> ///
-------
is met. Assuming the target productivity schedule is followed, this value rises to 130
kg/yr Total PCBs or 44 kg/yr Tri+ PCBs during Phase 2.
The formula to estimate the dredging-related release to date is:
^todate (Cffst )X Qtodate X Tti
0.02832 m 3600s 1 g 1000L
todate X ~ X X~ X
where:
Ftodate
c
ffst
where:
ft'
hr
Wng
m
(4-6)
load loss of Total PCBs at the far-field station for the dredging period to
date due to dredging-related activities in g/day
flow-weighted average concentration of Total PCBs at the far-field station
as measured from the start of the dredging period to date in ng/L. For once
per day sampling, this is given as:
C
ffit
^ c'ffitj xQj
/ = !
n
j=i 1
(4-7)
c
ffsti
Qj
n
c
bit
Qtodate
Ttodate
Total PCB concentration at the far-field station for day j. If more than one
sample is collected in a day, the arithmetic average of all the
measurements will be used
daily average flow at the far-field station for day j
number of days from the start of dredging period
Estimated arithmetic mean baseline concentration of Total PCBs at the
far-field station for the period in which the sample was collected, in ng/L.
This value is determined from baseline monitoring data. Time-weighted
averages are calculated as the sum of the arithmetic average for each day
dividing by the number of days
average flow at the far-field station, determined either by direct
measurement or estimated from USGS gauging stations, in cfs
average period of dredging operations per day for the time period, in
hours/day, as follows:
Hudson River PCBs Superfund Site
Engineering Performance Standards
96
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
where:
The period of dredging operations for day j in hours.
The allowable Ftodate in Phase I is 65 kg of Total PCBs if the total PCB mass anticipated
to be dredged is 10 percent of the total PCB mass remediated as estimated in the FS
(USEPA, 2001). If the production rate is different than 10 percent, the PCB load loss may
be adjusted for production rate as described in subsection 4.1.2.7. This formula is
intended to identify the amount of loading of Total PCBs due to dredging from the start
of the dredging period to the day of measurement. It is based on the collection of grab
samples, hence the correction for the daily period of operation. If sampling is conducted
on an alternate basis {i.e., composites), this formula will require adjustment to reflect this.
The load loss of the Total PCB at the far-field stations will be compared to the allowable
load loss for the dredging season.
4.1.2.7 Adjustment to the Load-Based Thresholds
The production rate will be reviewed on a weekly basis. The allowable Total PCB load
loss for the season will be adjusted if this target rate is not met using the following
equation:
The allowable seven-day Total PCB load loss thresholds will be revised if the production
rate varies from the anticipated value or the operation schedule differs from that assumed
for this report. This revision is to be calculated once per dredging season {i.e., the 7-day
running average criterion is set once per season). The equation for estimating the
allowable Total PCB load loss is as follows:
650(kg) (4-9)
where:
m
Total PCB mass anticipated to be dredged in a season (kg)
M
Total PCB mass to be dredged in the remediation (kg), 70,000 kg
as estimated in the FS (USEPA, 2001)
LoadTPCBauowabk
TV)
Load,hM (4-10)
target
where:
m dredge = Total PCB mass dredged within a period, kg
Hudson River PCBs Superfund Site
Engineering Performance Standards
97
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
^target = Targeted production rate, kg/hour. This is given as:
M
target J1 * D * Y ^ ^ ^
d year
where:
M = Total PCB mass targeted to be dredged in the remediation (kg),
70,000 kg as estimated in FS (USEPA, 2001)
Td = assumed average period of dredging operations per day, 14
hours/day
Dyear = assumed number of days in one dredging season, 210 days/season
Y = number of dredging seasons during the remediation
Loadthreshold = Total PCB load thresholds specified in action levels, such as 300
g/day and 600 g/day
The load calculation may be corrected for contributions originating upstream of the
remediation {i.e., above Rogers Island) in the event that loads from this region fall above
levels typically observed. See subsection 4.1.4.3.
4.1.3 Resuspension Standard Threshold
Resuspension Standard threshold is a confirmed occurrence of 500 ng/L Total PCBs,
measured at any main stem far-field station. To exceed the standard threshold, an initial
result greater than or equal to 500 ng/L Total PCBs must be confirmed by the average
concentration of four samples collected within 48 hours of the first sample. The standard
threshold does not apply to far-field station measurements if the station is within one mile
of the remediation.
4.1.4 Calculation Details
4.1.4.1 Calculation of Total and Tri+ PCBs from Congener Data
To estimate the Tri+ PCB and Total PCB concentrations from congener data the
following equations will be used:
n
Tri+PCBs = E Congeners Tri+i (4-12)
i=\
where:
Hudson River PCBs Superfund Site
Engineering Performance Standards
98
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
CongenerTri+ ,= Concentrations for PCB congeners with three or more chlorine
atoms measured
n
Total PCBs = E Congeners all t (4-13)
i=\
where:
Congeneraii, = Concentrations for PCB congeners measured
4.1.4.2 Non-Detect Values
Half the detection limit will be substituted for nondetect values in the formulas.
4.1.4.3 Upstream Source Concentrations
To identify the load contributions originating upstream of the remediation area (i.e.,
above Rogers Island), the 7-day running average of the Total PCBs should be used and
compared to the monthly baseline average obtained from the baseline monitoring
program. Appropriate means test should be used when comparing the 7-day running
average to the baseline monthly average. Prior to performing the means test, the
following should be considered:
Normality - test for normality of the data, either using W-Test for n<=50 or
Kolmogorov-Smirnov Test for n>50.
Data transformation - repeat the test for normality on transformed data for
parameters that are no normally distributed.
Homogeneity of variance - test for homogeneity of variance using Levene's test.
After considering the above criteria, perform the appropriate one-tailed means test
comparison:
For normally distributed data, t-test should be used if the variance is
homogeneous, otherwise approximate t-test should be used.
For data that are not normally distributed, the non-parametric Mann-Whitney U
test should be used.
After the means test is performed, the downstream load calculations may be corrected by
subtracting the load obtained from the difference between the average concentrations.
The additional load originating from upstream source can be calculated as follows:
Fri - (Cri ~ Cftjbi) X Qff X Tdl X
Hudson River PCBs Superfund Site
Engineering Performance Standards
99
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
where: Fri = Average additional load of Total PCBs originating from upstream
source in g/day,
Cm = Seven-day average concentration of Total PCBs from upstream
source (above Rogers Island) in ng/L.
Cmbi = Baseline monthly average concentration of Total PCBs from
Baseline Monitoring Program.
Of, = Average flow at the far-field station, determined either by direct
measurement or estimated from USGS gauging stations, in cfs, and
Tdl = Average period of dredging operations per day for the seven-day
period, in hours/day, as defined in equation 3.3.
The corrected downstream load is then:
Flcorr = ^7 ~ (4" 15)
where: F7Corr = Corrected load at the far-field station in g/day.
F7 = Seven-day average load of Total PCBs at the far-field station due
to dredging-related activities in g/day.
Fri = Average additional load of Total PCBs originating from upstream
source in g/day.
4.2 Monitoring Plan for Compliance with the Standard
Implementation of the monitoring program for compliance is provided in the subsections
4.2 and 4.3. Measurement techniques, monitoring locations, parameters, sampling
frequency and requirements of the standard are provided. Attachment F provides a
description of measurement techniques for the continuous monitoring requirements.
Some of the more stringent aspects of this monitoring program, such as the need to have
a real-time surrogate measurement of suspended solids to provide a warning of elevated
contaminant levels, may be relaxed after Phase 1. A clear rationale for each element of
the monitoring plan is presented in Section 3. Additional monitoring in the form of
special studies is required to gather information that can be used to refine the standard.
This is described in subsection 4.4.
Hudson River PCBs Superfund Site
Engineering Performance Standards
100
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
4.2.1 Measurement Technologies
Sampling techniques and technologies have been reviewed to select the appropriate
means of obtaining the monitoring data needed to confirm adherence to the standard. The
far-field monitoring will be similar to the baseline monitoring program implemented
during the remedial design period (2003 - 2005). The near-field monitoring will have a
reduced set of parameters and has little relation to previous sampling efforts. Some
additional components will be required to give a full picture of the conditions during
dredging (e.g., continuous monitoring for PCBs), but will not be assessed against the
action levels in Phase 1.
Instruments that provide an instantaneous measure of water column conditions will be
used for the following parameters:
Turbidity
Dissolved oxygen
Temperature
pH
Conductivity
Laser particle counters
Continuous measurement of water column conditions will be made for:
Turbidity
Laser particle counters
Integrating sampler for PCBs (continuous sampler)
The analytical methods will need to be sensitive enough to measure water column
concentrations at each station. This is most important for PCBs. For Total and Tri+
PCBs, a congener-specific method with a detection limit low enough to detect expected
PCB congener concentrations at Bakers Falls, Rogers Island, and Waterford is required.
4.2.2 Consistency with the Baseline Monitoring Program
There will be several monitoring programs developed throughout the remediation. The
primary programs are:
Baseline
Remedial
Long-term
To capture a consistent picture of the site conditions, there
must be consistency throughout the programs on key issues.
Two items that must be carefully chosen are the measurement
techniques (analytical or direct reading) and the locations of
Measurement techniques
and stations or substation
locations must be carefully
selected.
Hudson River PCBs Superfund Site
Engineering Performance Standards
101
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
the stations or substations.
The analytical methods chosen for this program must meet or exceed the specifications of
the methods used in the baseline monitoring program in terms of precision, sensitivity,
accuracy, representativeness, comparability, completeness and sensitivity. The only
exception to this requirement will be the modified method specified for TSS to allow a
reduced turn-around time. The same analytical methods chosen for each station will be
maintained at each station throughout the program for consistency.
The same stations must be occupied during the remediation
as during the baseline monitoring program. Any change to
the location or method of collection will result in changes to
the resulting data that cannot be easily accounted for,
confounding estimates of PCB conditions. The data collected
during baseline monitoring will be the means of
differentiating dredging-related loads from baseline loads. A correction would need to be
applied to the baseline data to properly determine compliance with the load-based
resuspension criteria, but there is no basis for developing this correction factor.
Therefore, it is essential to maintain the same monitoring locations and sampling
methods.
Two important aspects of the baseline monitoring program are the equal discharge
interval sampling method and the requirement that samples collected from the water
column cannot be split among multiple sample jars. These requirements must be
maintained for the resuspension standard monitoring program.
4.2.3 Compliance Monitoring Programs
Monitoring will be required for at least the remedial operations listed below. Other
operations related to dredging may be included as well:
Dredging
Debris removal
Resuspension control equipment removal
Off loading to the processing facility
Cap placement
Backfill placement
Installation of containment devices other than silt curtains (sheet piling and other
structural devices requiring heavy equipment operation and disturbance of the
river bottom)
Shoreline excavation and restoration
The following remedial operation will not require near-field monitoring:
Data collected during
baseline monitoring will
be the means of
differentiating dredging-
related loads from
baseline loads.
Hudson River PCBs Superfund Site
Engineering Performance Standards
102
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Silt curtain placement
4.2.3.1 Far-Field Monitoring
The far-field stations will be used to monitor water column conditions in the Upper and
Lower Hudson River. These results are needed for comparison to the baseline water
column concentrations to estimate the magnitude of any dredging-related release. Due to
the anticipated extent of remediation and associated barge traffic, dredging-related
releases may not be limited to a single area; thus, monitoring of multiple stations is
anticipated throughout the dredging period.
The parameters of primary interest are PCBs and
related parameters including turbidity, suspended
solids, DOC, and suspended OC. A surrogate real-
time measure of suspended solids will be used as an
indicator of dredging-related releases, assuming the
mechanism for increased PCB concentrations from
dredging operations will involve the resuspension of
contaminated sediment. DOC and suspended OC describe the dissolved and suspended
matter distribution of PCBs in the water column. These parameters also may be useful in
determining the source of elevated concentrations.
Discrete Samples
The far-field Upper Hudson River sampling will require the measurement of PCB
congeners, suspended solids, and OC by taking discrete, cross-sectional grab samples.
These measurements are necessary to assess the impacts of the dredging operations and
to provide a basis for a warning system for the downstream water intakes. The required
sampling in the Lower Hudson River is similar to the far-field Upper Hudson River
sampling, but is more limited in the extent and frequency of sampling. Data from these
samples will identify increased impacts to the Lower Hudson River from dredging and be
compared to resuspension criteria.
Unless stated otherwise, the monitoring and sampling at each station will be performed
using equal discharge increment (EDI) sampling for the Phase 1 monitoring program.
Equal width increment (EWI) sampling techniques may be considered for an alternate
Phase 2 monitoring program. The EDI or EWI methods usually result in a composite
sample that represents the discharge-weighted concentrations of the stream cross-section
for the parameter that is being monitored or sampled. The EDI and EWI methods are
used to divide a selected cross section of a stream into increments having a specified
volume of flow or width.
The samples will be integrated both vertically and horizontally. The term vertical refers
to that location within the increment at which the sampler or the measurement probe is
lowered and raised through the water column. EWI verticals are located at the midpoint
of each width increment. EDI verticals are located at the centroid, which is a point within
each increment at which stream discharge is equal on either side of the vertical. If
A surrogate real-time measure of
suspended solids will be used as an
indicator of dredging-related
releases, assuming the mechanism
for increased PCB concentrations
associated with dredging will be
resuspension of contaminated
sediment.
Hudson River PCBs Superfund Site
Engineering Performance Standards
103
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
properly implemented, EDI and EWI methods should yield identical results. These
sampling methods will be applied for all parameters measured in the water column.
Daily average flow rates at each far-field station will be recorded for comparison of the
discrete sample measurements to the load-based criteria.
Continuous Integrating Samplers for PCBs
Continuous integrating samplers will be set up at the far-field stations between Fort
Edward and Waterford. These samplers will be used throughout the dredging program to
integrate PCB loads and concentrations over time, providing a measure of PCBs releases
between the discrete samples. Integrating data over time intervals in the periods between
the discrete water column samples will enable identification of dredging-related releases,
including dissolved-phase PCBs that cannot otherwise be identified by examining
surrogate measurements such as suspended solids. The Phase 1 results may be used to
develop resuspension criteria for Phase 2.
Continuous Monitoring for Suspended Solids Surrogate at the Representative Far-
Field Stations
The suspended solids will be continuously monitored via surrogate direct reading
monitors (e.g., laser diffraction-based particle counters and turbidity monitors). A special
study will be conducted to determine an initial surrogate relationship (see subsection 4.4)
that will allow the suspended solids concentrations to be estimated in real time, which
provides a warning system for downstream water intakes in the Hudson River. The real-
time estimates of suspended solids will be compared to measured values from samples
collected once per day at each station. At least three substations must be monitored (one
in the channel, one on each shoal), but preferably, five or six substations will be occupied
in the same manner as the Baseline Monitoring Program sampling.
If the relationship between surrogate and TSS does not provide sufficiently accurate
estimates of TSS, samples will be collected for suspended solids analysis every three
hours with a three-hour turn-around (using a modified TSS method) for compliance to the
standard until an appropriate surrogate relationship is developed and implemented. In the
event of an exceedance of the suspended solids resuspension criteria based on the
surrogate measurement, TSS samples will be collected for confirmation twice a day at the
station with the exceedance.
All continuous monitors will be checked daily for problems such as bio-fouling and
damage.
Suspended Solids Collection at Other Downstream Far-Field Stations
The standard requires that samples be collected for suspended solids analysis every 3
hours on a 24-hour basis at the downstream far-field stations, excluding the
representative far-field station. These samples will be collected by means of automatic
samplers placed in the cross-section of the river. At a minimum, there will be one center
Hudson River PCBs Superfund Site
Engineering Performance Standards
104
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
channel station and two shoal stations, one on each side of the river, but preferably these
samplers will be deployed in a manner that is consistent with EDI. The samplers must be
capable of collecting and storing a series of three hour composite samples. The samples
will be collected twice a day. The turnaround time for these results will be 12-hours.
Decontamination procedures must be developed for these samplers that meet with
USEPA approval.
Monitoring Parameters Without Resuspension Criteria
Monitoring parameters required by the performance standard {i.e., discrete measures
taken whenever samples are collected for PCB or suspended solids analysis), but not
compared to resuspension criteria, are:
Temperature - because the distribution of PCB concentrations between the
dissolved and suspended phases is partially dependent on water column
temperature.
pH - to provide a measure of quality assurance by comparing values to the New
York State surface water standard (6.5 to 8.5 [6 NYCRR part 703.3])
Conductivity - to provide a measure of quality assurance
DO - because high suspended solids could exert a demand on oxygen levels,
which is potentially damaging to biota in the region of the dredge.
4.2.3.2 Near-Field Monitoring
Monitoring in the near field will be performed to determine the suspended solids releases.
This will include both continuous measurements of surrogates and discrete samples.
Continuous monitoring for a suspended solids surrogate is required to address two goals
of the Phase 1 standard:
To provide a real-time measure of conditions surrounding the dredging operation
To provide feedback to the dredge operator
The real-time measure provides an immediate signal to the dredge operator in the event
of an unexpected release. It also provides the dredge operator with feedback in the form
of information on the amount of resuspension resulting from various dredge
manipulations. Using this information, the dredge operator is expected to optimize the
manipulations of the dredge to avoid unnecessary resuspension. Based on this need,
continuous reading probes must be deployed in the near field even if their output does not
yield a sufficiently useful estimate of TSS.
The continuous suspended solids monitoring consists of monitoring suspended solids via
surrogate direct reading monitors {e.g., turbidity monitors) at each near-field station. A
special study will be conducted to determine an initial surrogate relationship (see
subsection 4.4). This relationship will allow the suspended solids concentrations to be
estimated in real time based on the continuous reading probes. The real-time estimates of
Hudson River PCBs Superfund Site
Engineering Performance Standards
105
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
suspended solids will be compared to measured values from samples collected once per
day at each station.
If the relationship between surrogate and suspended solids is not sufficiently protective of
the various action level criteria, samples will be collected for suspended solids analysis
every three hours with a three-hour turnaround (using a modified TSS method) for
compliance to the standard until an appropriate surrogate relationship is developed and
implemented. In the event of an exceedance of the suspended solids resuspension criteria
based on surrogate readings, samples will be collected for confirmation twice a day at the
station with the exceedance.
Continuous monitors will be deployed such that the measurements are made from the
middle of the water column (halfway between the river bottom and the water surface).
Continuous monitors will be checked daily for problems such as bio-fouling and damage.
Daily particle counter measurements will be required at each near-field monitoring
location. This will provide an additional means of relating a real-time measurement to
suspended solids.
4.2.4 Monitoring Locations
The monitoring plan has two baseline stations (Bakers Falls and Rogers Island), four
Upper Hudson far-field stations, and two Lower Hudson far-field stations. In addition,
each dredging operation has 5 near-field stations.
4.2.4.1 Far-field Monitoring
The following stations, locations of which are shown in Figure 1-2, comprise the far-field
monitoring stations for the Upper Hudson River:
TI Dam22
Schuylerville
Stillwater
Waterford
Two upstream baseline stations will be monitored in the Upper Hudson River:
Bakers Falls
Rogers Island
The Bakers Falls and Rogers Island stations represent baseline conditions for the
remediation area and thus need to be monitored regularly to avoid confusion between
22 The Thompson Island Dam station will be a true cross-sectional station, as opposed to the historical TID
West or PRW2 stations.
Hudson River PCBs Superfund Site 106 Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Resuspension - April 2004
-------
dredging-related releases and those that may have occurred upstream. The frequency of
monitoring at Bakers Falls will be less than that at Rogers Island, if the Bakers Falls
station continues to exhibit low baseline levels of PCBs and suspended solids relative to
Rogers Island conditions.
In the Lower Hudson River, the following stations will be monitored:
Albany
Poughkeepsie
In addition to these Lower Hudson River stations, a monitoring station will also be
required on the Mohawk River at Cohoes to estimate PCB loads from the Mohawk
watershed. This station will be used in conjunction with the measurements at the Lower
Hudson River monitoring locations to aid in identifying the fraction of any PCB load
increase from the Mohawk River, as opposed to the Upper Hudson River remedial
activities.
The daily (and any continuous) measurements at the far-field stations must reflect the
river cross section at the monitoring location by using EDI (USGS, 2002). At least five
locations will be monitored in each cross section. Discrete samples in the cross section
may be composited, but continuous reading devices (i.e., turbidity) are required at
multiple locations in the cross section.
4.2.4.2 Near-Field Monitoring Locations
Near-field monitoring locations are associated with individual remedial operations and
move as the operation moves. The data from these locations have two principal
objectives: provide feedback to the dredge operator and, provide a measure of suspended
solids release from the operation. Each remedial operation requires five monitoring
locations, which are arranged as shown in Figure 1-1 and described as follows:
One station located approximately 100 m upstream of the dredging operation will
monitor water quality conditions entering the dredging area to establish ambient
background conditions.
One station located 10 m to the channel side of the dredging operation will
monitor local boat traffic impacts.
One station located 100 m downstream of the dredging operation or 50 m
downstream of the most exterior silt control barrier will monitor the dredge
plume.
Two stations located 300 m downstream of the dredging operation or 150 m
downstream of the most exterior silt control barrier will monitor the dredge
plume.
Hudson River PCBs Superfund Site
Engineering Performance Standards
107
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The locations and associated criteria were chosen using the TSS-Chem model assuming
that a single dredging operation was achieving full production (refer to Attachment D of
this report). If control barriers are installed, the five stations will be placed outside of the
barrier. A sixth location within the barrier is required in the controlled area downstream
of the dredge. While there is no standard for this inner station, it is needed to develop a
relationship between conditions inside the silt barriers and the near-field monitoring
stations downstream. The distances from the remedial operations are approximate and the
location of the near-field stations may be changed in the field to better capture the plume,
if USEPA approves the change.
It is acknowledged that the location of remedial
operations and control barriers will be determined
during the design. As a result, the location of the near-
field monitoring stations can only be anticipated in this
standard, but will be reviewed as a part of the design.
Work plans developed for the remediation must specify
a means of verifying that the downstream monitors are placed to capture the plume. The
acoustic doppler current profiler (ADCP) may be useful in this regard. With changing
river conditions and movement of the dredge, periodic adjustment of the monitoring
locations will be required.
4.2.5 Potential for Reduction in the Near-Field Monitoring Locations
In order to provide an accurate representation of the suspended solids around the
dredging operations, all five monitoring (or six with containment barriers) are necessary.
However, if remedial operations are located in close proximity to one another, it may not
be feasible to maintain all of the locations since they may pose a safety concern to the
technicians or they will be within the working area for a downstream operation. In this
case, monitoring locations may be dropped at the discretion of the construction manager
for as long as this condition exists.
Such decisions must be documented in the weekly reports. At this time, it is anticipated
that stations will be dropped only if the remedial operations are located on the same side
of the river. A more prescriptive definition of the conditions when dropping a station
would be appropriate cannot be developed at this time, because this is contingent on
design specifications, including equipment types and schedule, that are not presently
available.
A possible example of conditions in which the number of stations can be reduced is when
remedial operations are conducted within 600 m (0.25 mile) of each other on the same
side of the river. This is the distance initially prescribed between the upstream station and
the farthest downstream stations, assuming no containment. In this situation, the
monitoring locations of the upstream operation would fall within the work zone of the
downstream operation. To remedy this, one or more of the downstream monitoring
locations for the upstream operation may be dropped. Additionally, the remaining
Work plans developed for the
remediation must specify a
means of verifying that the
downstream monitors are
placed to capture the plume.
Hudson River PCBs Superfund Site
Engineering Performance Standards
108
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
stations may serve as both downstream monitors for the upstream operation as well as
baseline monitors for the downstream operation.
If the operations are sufficiently close {i.e., within 0.25 miles and on the same side of the
river), the USEPA field coordinator may approve the monitoring of the two operations as
a single unit, thus halving the monitoring requirements. A similar configuration may
occur for contained areas, but the revisions to the monitoring program cannot be
specified, without further information to be developed during the design.
4.2.6 Frequency and Parameters
Tables 1-2, 1-3, and 1-4 contain the parameters and frequency of sampling required by
the Resuspension Standard for routine monitoring and each action level. The parameters
required are constant throughout, but the sampling method or analytical technique may
differ in some instances. The sampling frequency varies by station and action level.
4.2.6.1 Analytical Methods for Suspended Solids
Suspended solids measurements are required at both near-field and far-field stations.
While a surrogate measurement of suspended solids concentrations is in use for
compliance with the standard, a method equivalent to ASTM method 3977-97 will be
used with a turnaround time of 12 hours. This method will be equivalent to the suspended
solids analysis specified for the baseline monitoring program. A second modified method
will be specified that will allow an estimate of suspended solids concentrations to be
made with a three-hour turnaround time. Modifications to the standard method to permit
a reduced turnaround time may include:
Collection of a larger sample volume when suspended solids are visibly low
Reduction in drying time
Higher drying temperature
Field filtration
Co-located samples for both the standard and modified suspended solids methods will be
collected at a frequency of once per day for the first month of operation. The samples
should be collected from a range of concentrations to permit a full comparison of the
methods. If the methods are in good agreement (relative percent difference is less than
30%), the sampling frequency for co-located samples by the full ASTM method may be
reduced.
4.2.6.2 Sampling Methods for Suspended Solids
Suspended soilids samples will be collected for confirmation of the surrogate
measurements and compliance monitoring (in place of the surrogate measurements), and
in support of the PCB analyses. The collection method for confirmation of the surrogate
measurements will differ in that the sample must collected at the location of the turbidity
Hudson River PCBs Superfund Site
Engineering Performance Standards
109
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
sensor. For the far-field stations, the volume that is equivalent to the percentage of
discharge that a continuous monitor represents must be acquired for each substation. The
collection method for compliance monitoring will be vertically integrated samples at each
near-field station or compliance with the EDI method at the far-field station. The
sampling for supporting information for PCB analyses will be consistent with the PCB
sample collection process.
No splitting of water samples is permissible for any
measurements that must accurately reflect the suspended
solids content. If duplicate samples are required, the
sample bottles for the duplicate and sample analysis can
be deployed at once or in series to generate co-located
samples. Sample bottles for PCB and suspended solids
analysis should be deployed simultaneously if possible.
4.2.6.3 Far-field Monitoring Parameters and Frequency
Table 4-4 presents the relevant information for each parameter that will be monitored as
part of the far-field Upper Hudson River program. PCB congeners will be analyzed using
the Green Bay method or an equivalent method. Attachment F-2 provides a synopsis of
PCB analytical methods and associated detection limits. As stated above, the analysis for
suspended solids will be conducted using a method equivalent to ASTM method 3977-
97. The entire sample collected will be used for the suspended solids and PCB analyses.
All measurement techniques require sufficient sensitivity to avoid non-detect values at
most stations. For PCB congeners, low detection limits will be required at Bakers Falls,
Rogers Islands, and Waterford. Discrete samples must be collected from a potentially
impacted water parcel as it passes the station, although samples from different stations do
not need to be timed to correspond to the same water parcel.
The type of integrating sampler will be determined during design. Analysis for DOC,
suspended OC, and suspended solids will be required in addition to PCB congeners for
these samples, if this is appropriate for the type of sampler chosen.
The standard requires that samples for suspended solids be collected every three hours
continuously at each of the far-field stations, but that at the near representative far-field
station, a surrogate relationship will be developed to have a real-time indication of the
suspended solids concentrations. If suspended solids analyses for compliance have a turn-
around time of 12 hours at all other far-field stations, but if samples are collected for
compliance at the representative near-field station (e.g., TI Dam if dredging is limited to
the TI Pool), the turn-around time is three hours. It will be permissible to use an
integrated sampler to collect the eight samples per day for suspended solids (if the
sampler is capable of collecting eight separate samples over time) and sending all eight
samples to the laboratory once per day. This will greatly reduce the labor requirements
for the monitoring program.
No splitting of water samples
is permissible for any
measurements that must
accurately reflect the
suspended solids content.
Hudson River PCBs Superfund Site
Engineering Performance Standards
110
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Whole water samples for PCB analysis must include a process to extract PCBs from the
dissolved and suspended phases separately, using matrix-specific extraction and cleanup
methods used for the Reassessment RI/FS or similar methods demonstrated to be capable
of achieving equivalent extraction efficiencies. Justification for this approach is provided
in Attachment F-3. Analyses may be done on the combined extracts.
Routine monitoring of the six Upper River
mainstem stations will consist of grab samples and
continuous monitoring. Non-routine monitoring
will require the same analyses, but the sampling
method and frequency will vary with the station and
action level. Grab samples will be composited from
five or six samples in the cross section using the EDI sample collection method and
consistent with the approach taken during the baseline monitoring program. Continuous
monitors will be located in at least three locations (on channel station and two shoal
stations), although it would be preferable to have the stations deployed consistent with
EDI or EWI locations.
At Bakers Falls, one whole water PCB sample will be collected per week. DOC,
suspended OC, and suspended solids will be measured for these samples. The surface
water quality parameters to be measured are as follows:
Turbidity
Temperature
pH
Conductivity
DO
Routine and non-routine monitoring are the same for this station. Laboratory results must
be available within 72 hours of the collection of the sample. This station will be sampled
from only one location in the cross section.
At Rogers Island, one whole water PCB sample will be collected per day. DOC,
suspended OC, and suspended solids will be measured for these samples. Surface water
quality parameters to be measured continuously are as follows:
Turbidity
Temperature
pH
Conductivity
Dissolved oxygen measurements will be made along with each grab sample collected for
suspended solids. Samples will be collected for suspended solids every 3 hours, 24 hours
per day. An integrating sampler will be deployed continuously for a two-week period
throughout the construction season. The turn-around time for the PCB analysis is 72
Routine and non-routine
monitoring of the 6 Upper River
mainstem stations will both require
the same analysis, but sampling
method and frequency will vary
with the station and action level.
Hudson River PCBs Superfund Site
Engineering Performance Standards
111
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
hours from the collection of the sample. Routine and non-routine monitoring are the same
for this station. The monitoring frequency at Rogers Island may be reduced to weekly for
all parameters except suspended solids, if the data will not be used to monitor for releases
from the upstream sources that could be interpreted as releases from the remediation.
Reduction in frequency at this station will require approval from USEPA.
USEPA has not yet identified the location of the Phase 1 dredging. Assuming that the
remediation will be limited to the northern end of the TI Dam during Phase 1, there will
be two representative stations that are sampled with a shorter turn-around and a higher
frequency for monitoring contingencies: the TI Dam and Schuylerville stations.
During Phase 1, the Stillwater and Waterford stations will be monitored to measure the
PCB concentrations entering the Upper Hudson River public water treatment plants in
Halfmoon and Waterford, and to confirm or adjust the means of by which Total PCB
concentrations for the Waterford station have been estimated based on the concentrations
at the upstream stations. This information will be important during Phase 1 to understand
the behavior of the system, but the frequency of sampling at these downstream locations
will most likely be reduced in Phase 2.
Routine monitoring for the four Upper River far-field stations from the TI Dam to
Waterford will be identical to the monitoring at Rogers Island, with some exceptions:
Suspended solids will be continuously monitored with a particle counter at these
stations.
Grab sample laboratory results for parameters other than suspended solids must
be available within 24 hours of the collection of the sample for the TI Dam and
Schuylerville.
The nearest representative station, which would be the TI Dam station if dredging is
conducted in the TI Pool throughout Phase 1, will be required to have a surrogate
relationship for suspended solids concentrations in place of the suspended solids
sampling.
Non-routine monitoring at the two representative stations (TI Dam and Schuylerville)
will increase in frequency for the PCB, DOC, suspended OC, and suspended solids
samples, and the PCB analyses will be on the dissolved and suspended phases instead of
whole water. For the Evaluation Level, the samples will be collected twice a day. For the
Control Level samples will be collected three times a day. For the Resuspension Standard
threshold, the samples will be collected four times a day, but will be composited from
samples collected hourly over one six-hour period.
The deployment period for the integrating sampler will also vary. For the Evaluation
Level, the deployment period is the same as for routine monitoring. For the Control
Level, the integrating sampler will be deployed for periods of one week. For the
Resuspension Standard threshold, the integrating sampler will be deployed for one-day
periods.
Hudson River PCBs Superfund Site
Engineering Performance Standards
112
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The sampling frequency and turn-around time is
unchanged from routine monitoring for the
Evaluation Level for Stillwater and Waterford,
the farthest downstream stations. The sampling
method changes for the Control Level from
discrete grab samples to daily integrating samples
to capture the average concentration in what could be a rapidly changing environment.
The analytical results will be required within 24 hours for the Control Level. This shorter
turn-around time requirement is warranted for this action level because the Total PCB
concentration could be approaching the Resuspension Standard threshold, or because the
PCB load loss to the Lower Hudson River has exceeded the allowable rate for an
extended period of time. For the Resuspension Standard threshold, these stations will be
sampled four times a day for:
Whole water PCBs
DOC
Suspended OC
Suspended solids
Surface water quality
In addition, an integrating sampler will be deployed for one-day periods. The turn-around
time for PCB analyses from the integrating sampler will only be specified where the
information is needed quickly for comparison to the resuspension criteria. For the
Resuspension Standard, the turn-round times will be 24 hours for the two representative
far-field stations (TI Dam and Schuylerville stations) and the stations farther downstream
(Stillwater and Waterford stations). For the Concern and Control Levels at Stillwater and
Waterford, the turn-around times will be 72 hours and 24 hours, respectively.
These monitoring contingencies are for remediation of River Section 1 more than one
mile upstream from the TI Dam monitoring location. If dredging were conducted in River
Sections 2 and 3, the two stations downstream of the dredging will have the parameters,
frequency, sampling methods, and turn-around times associated with the TI Dam and
Schuylerville as described above, and stations below these stations will have the
parameters, frequency, sampling methods and turn-around times associated with
Stillwater and Waterford, also as described above.
If the remediation is conducted in more than one river
section, more than two stations are representative. If there
were an accidental release in a section that was not
undergoing remediation at that time, the two stations at least
one mile downstream of the accidental release would be
representative until the situation was resolved.
Representative stations must always be more than one mile downstream from the source
of the resuspended material. In the event that a far-field suspended solids resuspension
criterion is exceeded, the far-field station would be monitored for PCBs.
Sampling frequency and turn-around
time for Stillwater and Waterford is the
same as routine monitoring for the
Evaluation Level but changes from
discrete grab samples to daily integrating
samples for the Control Level.
In the event that a far-
field suspended solids
resuspension criterion is
exceeded, the far-field
station would be
monitored for PCBs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
113
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Exceedance of Evaluation Level criteria will prompt far-field Evaluation Level discrete
sample monitoring requirements. Exceedance of Control Level criteria will prompt far-
field Control Level monitoring discrete sample monitoring requirements. This additional
far-field sampling will be limited to the nearest downstream representative far-field
station or the next downstream station, depending on the location of the plume causing
the exceedance. Sample collection will be timed to capture the plume. The frequency,
parameters and sampling methods will be the same as those defined for the TI Dam and
Schuylerville in Table 1-2.
If the monitoring requirements change because of exceedance of a resuspension criterion
or reverting to lower action levels, the deployment period of the continuous integrating
samplers may change before completion of the period. If the deployment period is
reduced, the sample already collected will be sent for analysis. If the deployment period
is extended, the sampling period can be extended to match the new requirements.
Affirmation Sampling
Integrating PCB samplers are required to verify whether the grab samples are sufficiently
indicative of average river conditions. The deployment for the integrating sampler varies
from routine monitoring to different action levels. For routine monitoring and evaluation
level, the deployment periods are once every two weeks. At the control level, the
integrating sampler deployment periods at TID and Schuylerville are increased to once a
week. For Stillwater and Waterford far-field stations, the deployment periods are
increased to once a day at the control level. Similarly, at the resuspension standard
threshold, the deployment periods are once a day for all the far-field stations.
To ensure that the grab samples represent the average river conditions, the appropriate
means test comparison of the grab samples to the integrated samples need to be
performed. To perform the means test comparison, the following should be considered:
Normality - test for normality of the data, either using the W-Test for n<=50 or
the D'Agostino Test for n>50.
Data transformation - repeat the test for normality on transformed data for
parameters that are not normally distributed.
After considering the above criteria, perform the appropriate one-tailed means test
comparison:
For normally distributed data, t-test should be used if the variance is
homogeneous, otherwise approximate t-test should be used.
For not normally distributed data, the non-parametric Mann-Whitney U test
should be used.
If the means test results indicate that the mean of the grab samples is not statistically
different from the corresponding integrating samples, the sampling frequencies and
Hudson River PCBs Superfund Site
Engineering Performance Standards
114
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
approach for both the grab and integrating samplers are appropriate. On the other hand, if
the means test indicate that the mean of the grab samples is statistically different from the
mean of the integrating samples, additional study for both integrating and grab samples
needs to be performed to assess the adequacy of the grab and integrating samples.
4.2.6.4 Lower Hudson River and the Mohawk River at Cohoes
Far-field stations in the Lower Hudson River and at one location in the Mohawk River
will require routine monitoring. Sampling at these stations will include the analysis of
PCBs congeners, DOC, suspended OC, and suspended solids. The samples will be whole
water, not split phase. Discrete measurements will be made for the following:
Surface water quality measurements for turbidity
Temperature
pH
Conductivity
The results of the analyses will be required within 72 hours. Samples will be collected
every four weeks under routine monitoring. (This low frequency is contingent on the
results of the baseline monitoring program showing Total PCB concentrations less than
100 ng/L on average to allow a margin of safety for the public water supplies.) The
Mohawk River station will be sampled using EDI, but only a single center-channel
station is required for the Lower Hudson River stations.
Non-routine monitoring at these locations will be triggered by an estimated Total PCB
concentration of 350 ng/L or higher at Waterford or Troy. The first round of non-routine
monitoring will be timed to capture the parcel of water that triggered the non-routine
Lower Hudson River and Mohawk River monitoring.
The concentration is estimated using the following equation:
DO
( ' y
Lower Hudson ^ Farfield
where:
Ciroy = Estimated water column concentration Troy
CFar-fieid = Measured water column concentration at the far-field station
QFar-fieid = Instantaneous flow at the far-field station (cfs) at the time of
sample collection
Hudson River PCBs Superfund Site
Engineering Performance Standards
115
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Qiroy = Instantaneous flow over Federal Dam at Troy
4.2.6.5 Near-field Monitoring
Routine Sampling for Compliance
The parameters that are monitored in the near field are summarized in Table 4-5 along
with the relevant information for each parameter. The standard requires that a surrogate
real time measurement for suspended solids be developed and maintained throughout the
program for compliance with the near-field resuspension criteria. It is expected that
turbidity will be the surrogate measure chosen.
Each near-field station will have continuous monitoring for turbidity, temperature, and
conductivity for one hour prior to beginning remedial operations and for at least two
hours after the operation ceases. This applies to the five stations required if there are no
barriers installed, and to all six stations if barriers are installed. The information from
these monitors will provide immediate feedback to the dredge operator.
Confirmation Sampling of the Surrogate
Samples will be collected daily from each near-field monitoring location for confirmation
of the surrogate relationship. The ability of the surrogate to adequately predict the
suspended solids concentrations will be assessed on a daily basis. The criteria and method
for assessing the surrogate relationship is provided in Section 4.4. If the resuspension
criteria are exceeded at a near-field monitoring station, two samples will be collected per
day for confirmation of the surrogate.
In the event that the surrogate fails to adequately predict the suspended solids
concentrations, samples will be collected every three hours and analyzed for suspended
solids using the modified method with a three-hour turn-around. Vertically integrated
samples will be collected from each near-field station every three hours with the results
of the analysis available within three hours. These results will be compared to the
resuspension criteria. One sample from each near-field station will be collected one-hour
prior to beginning the remedial operations at a location.
After completing the remedial operation, at least two
samples collected one hour apart will be used to confirm
that the suspended solids concentrations have stabilized.
This will require the sampling to continue for at least
another four to five hours because of the three-hour turn-
around time on the analyses. More samples will be
required if the suspended solids concentrations have not
stabilized two hours after completing the remedial
operation. If the remediation is halted due to hazardous conditions such as thunderstorm,
the near-field monitoring to show that the suspended solids concentrations have stabilized
will not be required.
After completing the remedial
operation, at least two samples
collected one hour apart for
four to five hours will be used
to confirm that the suspended
solids concentrations have
stabilized.
Hudson River PCBs Superfund Site
Engineering Performance Standards
116
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Other Parameters
Discrete laser particle counter measurements will be made on any samples collected for
suspended solids analysis.
At both the near-field and far-field stations, pH and DO will be monitored discretely each
time a sample is collected.
Exceedance of the Near-Field Resuspension Criteria
Exceedance of near-field Evaluation Level suspended solids criteria will prompt far-field
Evaluation Level monitoring. Similarly, exceedance of near-field Control Level
suspended solids criteria will prompt far-field Control Level monitoring. This additional
sampling will be limited to the nearest downstream representative far-field station and
timed to capture the plume from the remedial operation. The frequency, parameters and
sampling methods will be the same as those defined for the TI Dam and Schuylerville, as
shown in Table 1-2.
Engineering Evaluations
Additional sampling in the near field may be conducted as a part of the engineering
evaluations. Samples for PCB analysis may be collected in the vicinity of the dredges or
in other areas affected by the remediation. The same sampling and analytical methods
will be used for comparison to the near-field and far-field data.
4.3 Reverting to Lower Action Levels
Any reduction in monitoring requires approval from USEPA before the changes are
made. USEPA may approve a reduction in the level of monitoring when the following
occurs for Total PCB criteria:
For the exceedance of a Control Level concentration criterion, the running
average concentration must fall below the action level for one week before the
contingencies can be relaxed.
For the exceedance of a Evaluation or Control Level seven-day running average
load-based criterion, the running average load level must fall below the action
level for one week before the contingencies can be relaxed.
Following exceedance of Resuspension Standard threshold, temporary halting of
in-river operations, and modification of the remedial operation, Control Level
monitoring requirements will commence unless otherwise instructed by USEPA.
Hudson River PCBs Superfund Site
Engineering Performance Standards
117
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Routine monitoring will resume in the Lower Hudson after non-routine
monitoring has confirmed that the concentrations in the Lower Hudson are below
350 ng/L Total PCBs and the estimated concentration at Waterford and Troy have
fallen below 350 ng/L Total PCBs for at least two days.
When suspended solids criteria are exceeded, the suspended solids concentrations
must fall below the action level for one day before USEPA may approve a
reduction in the level of monitoring and the contingencies can be relaxed.
During temporary halting of in-river remedial operations, routine monitoring of the
Upper River far-field stations will continue. If the operations are temporarily halted,
monitoring in the Lower Hudson will continue at non-routine frequency until the
requirements listed above are met.
4.4 Special Studies
The monitoring programs for the resuspension and residual standards are organized to
separate sampling necessary to measure compliance with the standard from sampling
efforts needed to evaluate and refine the implementation of the standard. This has been
accomplished by designating the second category of sampling efforts as "special studies."
The special studies will be conducted for limited periods of time to gather information for
specific conditions that may be encountered during the remediation or to develop an
alternate strategy for monitoring. Specific conditions may include different dredge types,
contaminant concentration ranges, and varying sediment textures. Each of these studies is
integral to the Phase 1 evaluation, the development of Phase 2, and is also tied to
compliance issues.
There are a total of five special studies for the resuspension standard. These are as
follows:
Near-field PCB Release Mechanism (Near-field PCB Concentrations)
Development of a Semi-Quantitative Relationship between TSS and a Surrogate
Real-Time Measurement for the Near-field and Far-field Stations (Bench Scale)
Development of a Semi-Quantitative Relationship between TSS and a Surrogate
Real-Time Measurement for the Near-field and Far-field Stations (Full Scale)
Non-Target, Downstream Area Contamination
Phase 2 Monitoring Plan
The main components of each of these studies is described below.
4.4.1 Near-Field PCB Concentrations
A special study will be conducted in the near field to characterize the nature of PCB
release due to dredging-related activities, specifically to evaluate whether the PCB
Hudson River PCBs Superfund Site
Engineering Performance Standards
118
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
release due to these activities occurs as the result of dissolved-phase or suspended-matter
phase releases. Data from this study will be used to evaluate the use of suspended solids
as a useful surrogate to identify PCB releases. Suspended solids will be a useful predictor
of PCB exceedances if the nature of the release is primarily resuspension of suspended-
phase PCBs. Following are some of the specifics that pertain to the Near-Field PCB
Concentrations Study.
4.4.1.1 Duration
Each study will last for a full work week (six or seven days) in each selected area. The
duration for the study of debris removal may be reduced if the debris removal is
completed in less than a week.
4.4.1.2 Sample Collection
The study will entail daily sample collection for each study area during the week of
investigation. This should allow for the collection of a sufficient number of samples to
distinguish dredging-related conditions from variations in the water column due to
baseline conditions.
The sampling locations will be arrayed in two transects located 100 ft and 200 ft
downstream of the dredge and will also include one upstream location. If there is
containment around the dredge, one composite sample consisting of three discrete
locations will be collected from within the containment and the transects will be located
just downstream of the containment and 100 ft downstream.
Each transect will contain five sample locations. If the water depth is greater than 10 ft,
two samples will be collected from each location (0 to 10 ft and deeper than 10 ft). A
sample will also be collected from a station 50 ft upstream from the dredge. Figure 1-1
depicts the layout of the monitoring stations. The location of the sampling stations may
be adjusted with the approval of the USEPA's field coordinator.
The plume will be identified at the transect locations using ADCP. This will be done with
a second boat that will continuously monitor for the location of the plume during the
sample collection. The signal from the ADCP increases markedly once the edge of the
plume is encountered.
4.4.1.3 Sample Handling
Vertically integrated samples will be collected following EDI techniques to represent the
area around the dredge (not the entire river width) and composited. Each sample
(comprised of several vertically integrated sampling nodes) will be filtered in the field as
soon as possible from the time of collection. Filtering of the sample must be completed
within two hours of collection. Samples will be collected in separate bottles at each
substation for each parameter measured. No samples will be split.
Hudson River PCBs Superfund Site
Engineering Performance Standards
119
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
4.4.1.4
Analytical and Direct Reading Methods
The following parameters will be measured on each sample:
PCB congener analysis (dissolved and suspended phases)
Suspended organic carbon
Dissolved organic carbon
Suspended Solids
Turnaround times must be assigned to allow sufficient amount of time to meet the
reporting requirements.
Measurements with a probe will be made at each substation for:
Turbidity
Temperature
pH
Conductivity
Laser-based particle size distribution
All measurements will be analytically consistent with the far-field monitoring program.
4.4.1.5 Definition of the Study Areas
Near-field total PCBs will be measured at several locations to determine the nature of
PCB releases for different sediment types (cohesive and non-cohesive), concentration
ranges, and dredge types. A near-field study will also be conducted during at least one
debris removal event.
Table 4-6 summarizes the possible areas for special study in the near field to characterize
the nature of PCB release due to dredging-related activities. The areas were chosen based
on:
Type of sediment as classified by the side scan sonar.
Type of sediment as classified by ASTM Method D422.
Range of Tri+ PCB entire core length weighted averages (LWAs) concentration.
Draft dredge area boundaries were used to guide the selection of the possible study areas.
(Note that these dredge area boundaries have not been approved by the USEPA; however,
while the boundaries have not been approved, the identified locations are expected to be
included in the final delineation of dredge areas and so were identified for this special
study.) Figure 4-4 shows the possible study areas, sediment type as classified by side scan
sonar, and the Tri+ PCB LWA range. Figure 4-5 shows the possible study areas and
different types of sediment as classified by ASTM Method D422. Of the 13 possible
study areas depicted, 5 areas are recommended for the special study (Table 4-7).
Selection of these 5 study areas did not take into consideration other engineering factors
Hudson River PCBs Superfund Site
Engineering Performance Standards
120
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
and the type of equipment that will be used for dredging; therefore, the final selection of
study areas may be different. The final selection of the study areas will be determined
upon USEPA approval of the Phase I Intermediate Design Report.
4.4.2 Development of a Semi-Quantitative Relationship between TSS and a
Surrogate Real-Time Measurement For the Near-Field and Far-Field
Stations (Bench Scale)
Laboratory studies correlating the direct measurement of suspended solids (i.e., TSS
analyses) and turbidity-based field measurements (or another surrogate real-time measure
of suspended solids) are required such that the near-field and the far-field suspended
solids analyses can be replaced with a surrogate real-time measure of suspended solids.
The need for a real-time measurement is evident from the sample frequency analysis,
which demonstrates that given the variability in baseline suspended solids concentrations,
samples will be collected every 15 minutes to monitor a suspended solids release with
sufficient confidence. This can only be achieved with a direct reading field measurement
device. These analyses will provide a link between the direct but time-consuming
measurement of suspended solids and surrogate suspended solids measurements, which
can be performed continuously and remotely with the use of a buoyed monitoring station
(or another equivalent method for the far-field stations).
4.4.2.1 Near Field
The relationship between suspended solids concentrations and turbidity for the Hudson
River Remediation is expected to be an evolving one, with the relationship potentially
changing over time as different sediments and hydrodynamic conditions are encountered.
Additionally, near-field requirements will be different due to the stronger suspended
solids and turbidity signals near the dredge operation. The concerns dictate the need for
separate study goals appropriate to near-field and far field conditions. It also necessitates
the need to review and revise the relationships as new field data are obtained.
For these reasons, the initial near-field suspended solids bench scale study must focus on
the sediments of the Phase 1 target areas. Subsequently, the daily sampling of near-field
TSS along with turbidity must be used to verify the initial relationship or slowly modify
the relationship.
4.4.2.2 Far-Field
The development of a surrogate for suspended solids in the far-field must also be
included in this special study. At a distance of 1 mile from the dredge, it will be difficult
to discern a simple increase in suspended solids concentration due to dredging given the
baseline variability and the small increase of concern (12 to 24 mg/L). To this end, the
far-field monitoring will include laser-based particle counters or equivalent to provide
data on the distribution of particle sizes in the water column in addition to the turbidity
monitors. The distribution of particle sizes due to dredging is expected to be quite
different from baseline, due in part to the different fractions of organic matter in the
Hudson River PCBs Superfund Site
Engineering Performance Standards
121
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
sediments vs. normal water column conditions. Based on these observations, it should be
possible to discern a rise in TSS approaching the threshold due to baseline variability
from a rise due to dredging resuspension. The combination of increased suspended solids
concentrations and turbidity along with a change in particle size distribution should
provide the most accurate signal of dredging-related releases and the need to sample.
Given this approach, it will also be necessary to collect data on the natural range of
particle size distributions under baseline prior to dredging (as part of baseline
monitoring).
4.4.2.3 Study Procedures
The procedures to do this study are described in guidance from the US Army Corps of
Engineers (Thackston and Palermo, 2000). Both the USACE Long Tube Settling Test and
batch tests as per Earhart (1984) will be conducted. However, the procedures involving
long tube settling tests for compression are not needed, which should reduce the time
required for the study.
4.4.2.4 Selection of Sediment Characteristics for the Study
Hudson River sediments will be collected from a number of locations in each river
section to encompass the range of sediment types that will be encountered while
dredging. This range of samples should provide a basis to examine the relationship
between direct measurement of suspended solids and turbidity measurements and permit
turbidity to serve as a surrogate of suspended solids measurement for a broad range of
sediment types.
This study will characterize the response for a minimum of three sediment types (silt, fine
sand and medium sand) by collecting at least 8 separate samples of each sediment class.
Samples must have median diameters consistent with their intended class (e.g., silt must
fall between 5 and 75 um median diameter) and have that class as the major fraction in
the sample.
4.4.2.5 Duration
A typical bench scale test can be conducted within a week. The initial study will be
conducted prior to the beginning of Phase 1. Subsequent bench scale tests may be
conducted if a surrogate measurement fails to predict suspended solids concentrations
with sufficient accuracy. See Section 4.4.3 for more information.
4.4.3 Develop and Maintain of a Semi-Quantitative Relationship between TSS and
a Surrogate Real-Time Measurement For the Near-Field and Far-Field
Stations (Full Scale)
This special study addresses the means by which the surrogate relationship for suspended
solids will be evaluated and updated using the confirmatory sample data. Surrogate
relationships for the near-field and far-field monitoring stations will be developed
Hudson River PCBs Superfund Site
Engineering Performance Standards
122
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
initially using only laboratory methods as described above. It is expected, however, that
field samples of TSS and estimates based on the surrogate relationship will deviate
somewhat from the laboratory-based relationships. Thus, it is necessary to continually
review and revise the relationships as new field data are obtained. At first, the evolution
will transition the relationship to represent field conditions. However, these relationships
are expected to evolve throughout the program as different sediment and hydrodynamic
conditions are encountered.
Daily confirmatory samples will be collected at near-field and far-field stations under
normal conditions. If there is an exceedance of the suspended solids-based resuspension
criteria, the rate of confirmatory sampling increases to two per day at the station with the
exceedance. These daily samples will be used to verify the initial relationship and
eventually modify it. Over time, the daily monitoring requirements should provide a large
data set with which to examine and establish a field-specific suspended solids-turbidity
relationship.
Statistical approaches will be used to evaluate data as it is collected, determine if the
TSS-Turbidity relationship should be modified, and refine the relationship based upon the
new data. This assessment will be conducted separately for the near-field and far-field
surrogate relationships.
Statistical Assessment
To verify that the surrogate relationship from the field data does not deviate significantly
from the initial relationship developed in the laboratory, statistical tests need to be
performed. Additionally, as the data set of field measurements grows, the combined field
and laboratory data can be combined into a single data for the purposes of defining the
relationship. The following statistical tests may be used:
Examine the proportion of the field data that falls within the 95 percent
confidence bounds of the predictive relationship . The confidence bounds are
those for the prediction interval from the regression. The confidence interval for
an individual point prediction, yo, is given by:
v ^ fn-2 \n + \ + n(x-ฅf
y
-------
= variance of predicted TSS concentration estimated
from the regression,
= approximately 1.96 for 95 percent confidence
intervals and large sample size (Normal
approximation),
= standard deviation of the TSS, given by
V n
where:
TSS predicted = predicted TSS concentration
estimated from the regression,
TSSfleid = measured TSS concentrations.
The above equation will give the fraction of measured suspended solids
concentrations that fall within the 95 percent confidence limits of the regression.
If more than 10 percent of the measured suspended solids concentrations data fall
outside the 95 percent confidence limits, it is considered to be a poor fit.
Chow's F test (Fisher, 1970) can test whether the parameters for two data sets
(e.g., the initial laboratory data versus a collection of field measurements) are
significant. It requires calculating the error sum of squares or sum of squared
residuals (SSEs) for regression models on each of the data sets individually and
an SSE for a regression on the pooled data. The comparison is made by forming
an F statistic with k and (ti+t2-2k) degrees of freedom, formed as (Kennedy,
1979):
\SSE(constrained) - SSE(unconstrained)\/ k
r
SSE (unconstrained) !{tx +t2 -2k)
where
SSE(unconstrained)
SSE(constrained)
in-
n-2
^y.x
= the sum of the SSEs from the two separate
regressions,
= the SSE from the regression of the pooled data,
sse= ^(r-r)2
Where: Y= measured TSS concentrations and
Y= predicted TSS concentrations.
11 = the number of observations in the first sample set,
t2 = the number of observations in the second sample set,
and
k = the number of parameters in the model, including the
intercept term.
Hudson River PCBs Superfund Site
Engineering Performance Standards
124
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The resulting statistic can then be compared to a tabulation of the F distribution
with k and (tl+ t2 - 2k) degrees of freedom to test the hypothesis that parameters
have changed significantly between data sets 1 and 2. If the calculated F statistic
exceeds the critical value F, the null hypothesis (no change in the regression lines)
can be rejected. An F statistic with a 95 percent probability of occurring can be
considered indicative of a significant difference in the parameters, and by
inference, a difference between the laboratory and field relationships.
Theil's U statistic that gives a measure of the consistency between the forecasts
(e.g., field data predictions using the initial surrogate relationship model) and the
data used to develop the forecasts. It ranges from 0 to 1, with 0 indicating perfect
predictions. The variance of the U statistics can be approximated (for U less than
0.3) as U2/T, where T is the number of samples in the "forecast." The U statistic is
defined as (Pindyck and Rubinfeld, 1981):
The numerator of U is simply the root mean square simulation error, but the
scaling of the denominator is such that U always falls between 0 and 1. The U
statistics may also be decomposed into portions attributed to bias or systematic
error (Um), variance or ability of the model to replicate the degree of variability in
the variable of interest (Us), and covariance or unsystematic error (Uc). These
proportions of inequality, which sum to 1, are defined as:
where
Yts = simulated TSS value for observation t,
Yta = actual TSS value of the observation t, and
T = total number of observations.
Bias Error
jjm Vฑ 1 )
"(i/7-)^(7;-y;)2
(ys Yay^
Variance Error
JJS V ^ v ^gj
"(i/t^y'-y;:)2
(g.-g,,)2
Covariance Error
JjC
~ (Ulj^WrY??
2(1 -p)oson
Hudson River PCBs Superfund Site
Engineering Performance Standards
125
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
where:
Ys = the mean of the series of the simulated TSS Yts,
Ya = the mean of the series of actual TSS Y",
<7S = the standard deviation of the series Yts,
<7U = the standard deviation of the series Yta,
p = the correlation coefficient of the two series.
When U is non-zero, a desirable evaluation of a model will show that the non-
zero component is dominantly attributable to the covariance or unsystematic
component, which represents non-controllable random variability. Weight on the
bias component indicates that the linear relationship differs between the two data
sets. Weight on the variance component indicates that the difference is
attributable primarily to differing variances between the two data sets.
For the purposes of the TSS-turbidity relationship, consistency in the relationship
would be exhibited by a high Uc component and low values for Um and Us. Values
of Um and Us over 0.2 are indicative of a significant difference between the
laboratory and field relationships.
Low Bias Assessment
In addition to the statistical tests, the measured suspended solids concentration data need
to be checked for low bias compared to the surrogate regression. If 75 percent of the
measured suspended solids data falls under the regression for 4 days out of 7 days, it is
recommended that the surrogate relationship be reassessed.
Evaluation
The statistical tests and the comparison of the field data to the current surrogate
relationship need to be performed daily. The frequency of assessing the data may be
lowered in Phase 2, if appropriate. Data from confirmation suspended solids sampling
collected during the previous seven days (if applicable to current operating conditions)
will be compared to the data used to develop and maintain the surrogate relationship.
This data will initially be composed initially of the bench scale test results. When Phase 1
begins and confirmatory samples for suspended solids are collected, these results will be
compared to the bench scale results in the manner described below.
If Chow's F, Theil's U and the low bias assessments, show the surrogate relationship to
be in compliance, continue use of the surrogate for evaluation of the suspended solids
based resuspension standard.
If Chow's F or Theil's U statistics fail, and there is no low bias, the surrogate relationship
is in compliance, but the data from the previous day should not be used to reassess the
current regression. It is recommended that the regression be reassessed.
Hudson River PCBs Superfund Site
Engineering Performance Standards
126
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
If Chow's F or Theil's U statistics fail, and there is no low bias, the surrogate relationship
is not in compliance. It is required that the regression be reassessed.
The regression will be re-evaluated weekly to capture the information from the field
results and adjust the means of calculating the suspended solids concentrations from the
surrogate. Daily measurements will be evaluated in terms of the existing relationship.
Reassessment of the Surrogate Relationship
In the event that the reassessment of the surrogate relationship is needed, there are two
options. Sediment in the current area could be collected and a bench scale study that
conforms to the special study described in Section 4.4.2 could be conducted. This
method is preferred. Alternatively, the confirmatory samples for suspended solids can be
assessed to determine if a revised surrogate measurement can be derived from the
available data. Until a revised surrogate regression can be derived and approved by
USEPA, samples will be collected every three hours for suspended solids analysis with
three hour turnaround (using the modified method for suspended solids) and used for
compliance with the standard. This sampling will apply to either the near-field or the far-
field, depending on which surrogate relationship needs reevaluation.
4.4.3.1 Duration
This study will be conducted throughout Phase 1. It is likely that this study will be
maintained in some form throughout the remediation, because the surrogate relationships
are likely to require adjustment as the remediation moves throughout the river.
4.4.4 Phase 2 Monitoring Plan
This study will be conducted to demonstrate the feasibility and implementability of
alternated monitoring programs that are proposed for Phase 2 of the remediation. The
study will determine if the alternate program fully meets the data quality objectives
defined for the Resuspension Standard monitoring program. The results of the study will
be used to adjust the resuspension criteria, monitoring program and engineering
contingencies for the Phase 2 standard.
4.4.4.1 Definition of the Study Areas
The Phase 2 Monitoring Program would need to be implemented at all stations where
changes to the Phase 1 Montioring Program are proposed.
4.4.4.2 Duration
The Phase 2 monitoring plan must be implemented for enough time to allow potential
problems with the alternate sampling methods to be identified. The program must be in
use during the month of full production, but the extent to which the duration of the study
will extend beyond that period will depend on the details of the Phase 2 monitoring plan.
Hudson River PCBs Superfund Site
Engineering Performance Standards
127
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Alternate monitoring programs with more challenging aspects may require longer periods
of implemention.
4.4.4.3 Assessment of Data
The data acquired during this study will be compare to the results of the Phase 1
monitoring program to determine if the alternate program succeeds in achieving the data
quality objectives defined for the Phase 1 program. The study will be reviewed to
determine if there are implementation issues that require alteration. The reliability of the
alternate program will be assessed.
4.4.4.4 Automatic Samplers for PCB Sample Collection
An alterative to the Phase 1 monitoring plan that may be contemplated would be use of
automatic sampling devices to collect the PCB samples under routine conditions. Once a
fuller understanding of the nature of contaminant release is acquired through the
monitoring program as written is acquired, a well designed monitoring that included the
use of automatic samplers for collection of the PCBs could conceivably be of benefit
providing more temporal coverage and may reduce costs.
Specific Requirements
While conceptually reasonable, there are aspects associated with the use of automatic
samplers that may make implementation difficult. For instance:
How will these samplers be maintained to ensure that samples are always being
collected and the instruments have not clogged?
If piping is needed, how will the integrity of the pipes be maintained?
If piping is needed, how will the system be designed to avoid settling of
suspended mater in the pipe?
How will the samplers be decontaminated between samples?
How will the samplers be protected from boat traffic and still collect
representative samples from the cross-section?
Some specific requirements of an alternate monitoring program that includes automatic
samplers are:
The stations must be in the same location as the baseline monitoring program,.
Samples must be collected in a manner that is compliant with EDI or EWI.
The reliability of the system must be demonstrated.
Decontamination procedures must be demonstrated.
A comprehensive maintenance plan must be developed.
Hudson River PCBs Superfund Site
Engineering Performance Standards
128
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Resultant Changes to the Standard
Use of automatic samplers to collect PCBs may prompt changes to other aspects of the
standard. The resuspension criteria, aspects of the near-field and far-field monitoring
program and engineering contingencies would need to be evaluated.
The higher sampling frequency that can be achieved with composite sampling would
provide a more reliable measurement of the water column concentration. Assuming that
the issues identified above can be overcome, there would be more certainty in these
measurements and the period in which an exceedance of the resupsension criteria can be
known would be reduced. Table 4-8 shows a possible revision to the resuspension criteria
should PCB samples be collected with automatic samplers. The time period for each
PCB-based resuspension criteria has been reduced from seven days to time periods of
two to four days, which are derived from considerations of statistical certainty. The
engineering contingencies and monitoring contingencies associated with these
exceedances would need to be re-evaluated. The time frames for implementation of
engineering contingencies would also need to be re-evaluated.
If the Phase 2 Monitoring Program demonstrated that this means of sampling were
acceptable, alterations would be made to the Phase 2 Resuspension Standard criteria in
light of the information acquired during Phase 1.
4.4.5 Non-Target, Downstream Area Contamination
This study will examine the amount of resuspended material that has settled in the local
downstream areas of the dredging operation and could act as a potential source of future
contamination of the water column and downstream surficial sediment. The primary data
quality objective for this study is to determine the extent of contamination in terms of
spatial extent, concentration and mass of Tri+PCB contamination deposited downstream
from the dredged target areas in non-target areas.
The data acquired from this study will be used to determine if the resuspension controls
are adequately limiting downstream transport of contamination. A basis for this
determination may be a comparison to the thresholds for MPA and surface concentrations
provided in the ROD. If the local downstream areas are exceeding these criteria, the
resuspension controls will require evaluation. Another consideration will be the amount
of mass that is transported downstream near the bottom of the river.
4.4.5.1 Definition of the Study Areas
Study areas will be identified in the same manner as the Near-Field PCBs special study
(Section 4.4.1.5). The study area will cover approximately five acres. In addition to these
specifications, the area downstream area will not be rock or gravel as defined by the side
scan sonar. Because these areas will be located in the Phase 1 dredge zones, the areas that
are sampled may not be non-target areas as defined by the dredge line delineations, but
will be studied to have this information early on in the project.
Hudson River PCBs Superfund Site
Engineering Performance Standards
129
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
4.4.5.2
Duration
The studies will be conducted throughout Phase 1.
4.4.5.3 Sampler Deployment and Collection
Sediment traps or equivalent equipment will be deployed in the study area. Sediment
traps will be deployed at the rate of eight per five-acre area. The sediment traps will be
laid out on a triangular grid. The sediment traps will be co-located (approximately 10 ft
apart). This will allow one of the co-located sediment traps to be sampled each week,
while the other remains in place for the duration of the study and is sampled at the end.
The sediment traps will be installed at the start of the dredging in the area under study.
4.4.5.4 Sample Handling
Suspended sediments collected in the trap will be weighed to determine mass collected
and then homogenized for subsequent organic carbon and PCB analysis. PCB analysis for
the short deployment traps may not be possible if a limited mass of sediment is obtained.
4.4.5.5 Analytical Methods
The following parameters will be measured on each sample:
Sediment mass collected
Organic carbon content
PCB congener analysis
The following field measurements will be recorded:
Date and time of deployment
Date and time of sample collection
Depth of sediment in the sediment trap
Approximate distance from the dredge operation.
All measurements will meet or exceed the analytical specifications for the SSAP
program.
4.4.5.6 Definition of the Study Areas
The areas to be studied will be identified in a similar manner to the Near-Field PCBs
special study (Section 4.4.1.). Each study area will cover approximately five acres
downstream from an area undergoing remediation. In addition to these specifications, the
downstream study area will not include rock or gravel as defined by the side scan sonar,
since these are generally poor depositional zones and unlikely to accumulate sediment
from the dredge. Because these areas will be located downstream of the Phase 1 dredge
zones, the areas that are sampled may not be non-target areas as defined by the dredge
Hudson River PCBs Superfund Site
Engineering Performance Standards
130
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
line delineations. This is not a concern since the Phase 1 downstream areas are typically
depositional and should provide a conservative estimate of the amount of deposition that
can occur over non-target areas.
4.4.6 Further Development of the Special Studies
The special studies will be further developed and specific implementation details
documented in work plans and quality assurance project plans developed during the
design phase. Modification of some aspects of the special studies as outlined may be
permissible as long as the objectives of the studies can be achieved. All modifications to
the programs as outlined in this document will require USEPA review and approval.
4.5 Engineering Contingencies
For the Hudson River remediation, engineering
contingencies must be considered for the dredging
operation if the action levels are exceeded:
Engineering contingencies will be
recommended for consideration when the Evaluation Level is exceeded by any
measure {i.e., suspended solids or PCBs, near-field or far-field).
Engineering contingencies will be required and implemented if the Total PCB or
Tri+ PCB concentrations exceed the Control Level or the Resuspension Standard
(500 ng/L Total PCBs), based on monitoring results at the far-field stations for
PCB load- or concentration-based criteria, not suspended solids criteria.
If the Control Level or the Resuspension Standard threshold is exceeded, an
adjustment to the remedial operation is mandatory.
If the Evaluation Level, the lower tier action level, is exceeded, an adjustment to
the operation is optional.
Additional monitoring is mandatory when any of the action levels criteria parameter {i.e.,
Total PCBs, Tri+ PCBs, or suspended solids) is exceeded. Engineering evaluations of the
source of the exceedance are also required when the Control Level or the Resuspension
Standard threshold is exceeded.
The performance standard requires increased monitoring contingencies, engineering
evaluations, and modification of remedial operations for exceedance of the action levels.
Subsections 4.2 and 4.3 describe the monitoring contingencies. This section describes the
engineering evaluations, suggested technologies to control resuspension, and the
requirements of the standard in this regard. These engineering evaluations and
Engineering contingencies
must be considered for the
dredging operation if the action
levels are exceeded.
Hudson River PCBs Superfund Site
Engineering Performance Standards
131
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
technologies are described in general terms here, but will be specified during the remedial
design and possibly modified during the remedial operation.
Recommended and required engineering contingencies are listed below for each action
level and the Resuspension Standard threshold.
Evaluation Level
Evaluate and identify any problems.
Examine boat traffic patterns near the dredges.
Examine sediment transfer pipelines for leaks.
Recommend engineering evaluations near the dredges and barges.
Perform other such engineering evaluations as appropriate.
Recommend PCB sample collection in the near-field or other areas of
the operation as a part of an engineering evaluation.
Control Level
Initiate mandatory engineering evaluation and continual adjustments to
dredging operations until the Evaluation Level or better is attained.
Evaluate and identify any problems.
Consider change in resupension controls, dredge operation, or dredge
type.
Consider implementing additional resuspension controls.
Consider changing location and rescheduling more highly
contaminated areas for later in the year (applies to May and June
only), if all other options are not effective.
Temporarily cease operations if required.
Resuspension Standard
Mandatory cessation of all operations in the river is required if Total
PCB concentration levels in excess of 500 ng/L Total PCBs are
confirmed by next day's samples.
Restart requires engineering evaluation and USEPA approval.
4.5.1 Timeframe for Implementing Engineering Evaluations and Engineering
Improvements
The time frame for the initiation and completion of engineering evaluations and
implementation of the engineering solutions must be specified as part of the remedial
design. The actual implementation schedule in the field is subject to USEPA review and
oversight. It is anticipated that engineering evaluations will begin immediately upon
receipt of data indicating the exceedance of a criterion. It is similarly anticipated that the
Hudson River PCBs Superfund Site
Engineering Performance Standards
132
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
required engineering contingencies should begin as soon as possible so as to minimize
PCB releases. At a minimum, engineering contingency actions should begin within a
week of an exceedance, assuming conditions remain in exceedance. In the case of a
temporary halt of the operations, an evaluation should be completed with five days. In the
event of a temporary cessation, every effort should be made to correct the problem and
minimize the length of time of the stoppage.
4.5.2 Engineering Evaluations
The engineering evaluation includes the study of
all dredge-related operations and supporting
components, including review of the dredging
operation, barrier installation, and sediment
transportation system. Engineering evaluations
are required for exceedance of the Control Level
and Resuspension Standard and recommended but not required for exceedance of the
Evaluation Level.
Exceedance of the suspended solids criteria must be confirmed by PCB measurements
before actions other than increased monitoring are required. The evaluation and review of
the dredging operation should include additional turbidity measurements in the vicinity of
the dredge, barge, pipeline, etc., and will be conducted to evaluate the possible source(s)
and mechanism(s) causing the exceedance. An engineering evaluation will include the
following as needed:
ฆ Examination of the containment barrier, if it is in use, for leaks and stability
ฆ Examination of the sediment transport pipeline, if a hydraulic dredge is used
ฆ Examination of the barge loading system and barge integrity, if barges are used
ฆ Examination of the turbidity associated with the sediment transport barges and
other support vehicles
ฆ Analysis of near-field water column samples for Total PCBs, as well as analysis
of samples from other locations such as along the sediment transport pipeline, the
channel, etc.
The evaluation will be briefly documented in a report with approach, results, and
conclusions for submittal to USEPA. Submittal of a report is mandatory in cases where
USEPA must approve modifications to the remediation or give approval to resume
operations following temporary halting of remedial operations.
Engineering evaluations are required for
exceedance of the Control Level and
Resuspension Standard and
recommended but not required for
exceedance of the Evaluation Level.
Hudson River PCBs Superfund Site
Engineering Performance Standards
133
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
4.5.3 Implementation of Control Technologies
This subsection discusses engineering contingencies recommended for consideration in
the event of an exceedance of the Control Level or Resuspension Standard. The
contingencies consist of implementation of specific control technologies and apply to
remedial operations. A more detailed description of these technologies is provided in
Attachment E to the Resuspension Standard. Use of these contingencies resulted
primarily from the review of relevant case studies (See Volume 5) and from research
done during preparation of the Hudson River FS Report (USEPA, 2000b).
4.5.3.1 Remedial Operations
Barriers and modifications to operations and equipment are the principal and most useful
methods for reducing the suspended solids and PCB concentrations downstream of the
dredging operation.
Barriers
Barrier types reviewed in Attachment E include:
Fixed structural barriers such as sheet piling.
Non-structural barriers such as silt curtains and silt screens.
Portable barriers systems such as the Portadam and Aqua-Barrier systems.
Air bates.
Control zone technology.
If a barrier system has been implemented, but action levels are still exceeded, further
steps that can be considered include the following:
Monitor or inspect the barrier for leaks
Identify and correct problems with the installation
Change the barrier material to a more effective material such as high density
polyethylene (HDPE)
Install multiple layers of barriers
Fasten the barrier to the river bottom
Operation and Equipment Modifications
Operation and equipment modifications that may reduce the generation of suspended
sediments include:
Limiting/reducing boat speeds to reduce prop wash.
Restricting the size of boats that can be used in certain areas.
Loading barges to less than capacity where necessary to reduce draft.
Use of smaller, shallow draft boats to transport crew members and inspection
personnel to and from the dredges.
Hudson River PCBs Superfund Site
Engineering Performance Standards
134
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Selection of an alternate dredge with a lower resuspension rate.
Selection of another means of placing backfill/capping materials.
Scheduling changes to the dredge plan/pattern to avoid remediation of highly
contaminated areas during times of year when background PCB concentrations
are high (applies to May and June only).
4.5.4 Requirements of the Standard
The standard provides a series of action levels by which the severity of the dredging-
related release can be measured and quantified. As an action level is exceeded,
engineering evaluations and engineering solutions will be suggested or required, based on
the level of the exceedance. This tiered level of enforcement is set up to allow for the
remediation to be conducted continuously without operation near the Resuspension
Standard threshold, thus avoiding subsequent temporary halting of remedial operations
due to a confirmed exceedance.
In summary, the Resuspension Standard requires the following:
Action l.i'U'l
Monitoring
(onliniii'iH'ics
Ki'(| ii i mi-
Mn^iiH'i'i'in^
r.\iiliiiilion
Ki'(|iiiiT(l
r.njiiiu'CTinii
( (uilinjicnck's Ki-(|iiiiT(l
Evaluation
Yes
Recommended
No
Control
Yes
Yes
Yes
Resuspension Standard
Threshold
Yes
Yes
Yes
* Monitoring requirements for suspended solids exceedances limited for the far-field monitoring to only
one or two stations, in order to capture the PCB concentrations in the impacted water column.
4.5.5 Settled Contaminated Material and the Need for Resuspension Barriers
The near-field modeling results presented in subection 2.6 and Attachment D indicate
that a substantial amount of the suspended solids will settle in the immediate vicinity of
the dredge. In particular, coarse-grained sediments settle very rapidly and so will most
likely be captured by a subsequent dredging pass. However, fine-grained sediments may
remain in the water column sufficiently long to settle beyond the areas selected to be
dredged.
While modeling analysis does not show these additions to be significant in terms of long-
distance transport, the redeposited sediments may potentially create small regions of
elevated contamination just outside the remedial areas. This could elevate the PCB
concentration of the river bed surficial sediments downstream of the remediation to
concentration levels that are unacceptable even for the least stringent PCB load-based
action level (300 g/day).
Hudson River PCBs Superfund Site
Engineering Performance Standards
135
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
The potential for redeposition leads to the conclusion
that, where appropriate, resuspension barriers of some
type should be considered to contain the resuspended
material within the target areas and control the spread
of contaminated material.
The need for these controls is suggested by evidence obtained from the dredging on the
Grasse River. Rising concentrations of cesium-137 and PCB in the surface layer sediment
downstream were observed as part of the post-dredge sampling of the Grasse River non-
time critical removal action (NTCRA). As shown in Figure 4-6, cesium-137 increases in
the uppermost layers of all four cores collected downstream of the dredging operation.
The surface layer sediment represents the most recently deposited material. In term of
natural variation, the concentration for cesium-137 is not expected to increase since its
source (atmospheric weapons testing) no longer exists. This significant increase is
consistent with the release and redeposition of older sediments containing high levels of
cesium-137 as a result of dredging operations. The relatively thin layer suggests this is
not a significant redeposition on the scale of miles (the distance among the cores) but
does demonstrate its occurrence.
PCBs do not show as much response as Cesium 137 in the Grasse River sediment, but
evidence of a recent PCB release is clear in one core (18M). This core shows
significantly elevated PCB concentrations at the surface, also consistent with a suspended
solids release. The elevated PCB levels associated with this core may also reflect the
generally higher PCB levels in recently deposited sediments, suggesting that the location
may collect more of the fine-grained, PCB-contaminated sediments than the other coring
locations. Notably, the triple silt barriers used at this site were not fastened to the river
bottom, potentially permitting resuspended material to travel beneath them and move
downstream. While these data cannot be construed as proof, this does do suggest that
suspended solids settling estimates warrant further consideration. Some form of sediment
monitoring outside the target areas will be required. Sediment monitoring for this purpose
is required in one of the special studies discussed previously, the Non-Target Area
Contamination study.
These data also suggest that dredging should
generally proceed from upstream to downstream, or
the associated redeposition will recontaminate
remediated areas. Where resuspension barriers are
used, the water flow rate within the barriers is
expected to be greatly reduced, thereby significantly
reducing this problem.
The potential for redeposition
indicates that, where
appropriate, resuspension
barriers should be considered.
Dredging should generally
proceed from upstream to
downstream, or the associated
redeposition will recontaminate
remediated areas.
Hudson River PCBs Superfund Site
Engineering Performance Standards
136
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
5.0 References
Ault, J. S., K.C. Lindeman, and D.G. Clarke. 1998. "FISHFATE: Population dynamics
models to assess risks of hydraulic entrainment by dredges," DOER Technical Notes
Collection (TN DOER-E4), U.S. Army Engineer Research and Development Center,
Vicksburg, MS. www.wes.army.mil/el/dots/doer
Burton, W. H., S.B. Weisberg, and P. Jacobson. 1982. "Entrainment effects of
maintenance hydraulic dredging in the Delaware River estuary on striped bass
ichthyoplankton," Report to Delaware Basin Fish and Wildlife Management Cooperative,
West Trenton, NJ.
Carroll, K.M., M.R. Harkness, A.A. Bracco, and R.R. Balcarcel. 1994. "Application of a
Permeant/Polymer Diffusional Model to the Desorption of Polychlorinated Biphenyls
from Hudson River Sediment." Environ. Sci. Technol. Vol 28, pp. 253-258. 1994.
Connelly, N.A., B.A. Knuth, and C.A. Bisogni. 1992. Effects of the Health Advisory
Changes on Fishing Habits and Fish Consumption in New York Sport Fisheries. Human
Dimension Research Unit, Department of Natural Resources, New York State, College of
Agriculture and Life Sciences, Fernow Hall, Cornell University, Ithaca, New York.
Report for the New York Sea Grant Institute Project No. R/FHD-2-PD, September. (Raw
survey data also received electronically from study authors.)
Farley, K.J., R.V. Thomann, T.F. Cooney, D.R. Damiani, and J.R. Wand. 1999. An
Integrated Model of Organic Chemical Fate and Bioaccumulation in the Hudson River
Estuary. Prepared for the Hudson River Foundation. Manhattan College, Riverdale, NY.
Fisher, F., 1970. "Test of equality between sets of coefficients in two linear regressions:
an expository note." Econometrica, Vol. 38, No. 2, pp. 361-366, March 1970.
General Electric (GE). 2003. Habitat Delineation and Assessment Work Plan, Hudson
River PCBs Superfund Site. Prepared by Blasland, Bouck, and Lee. May.
Kennedy, P. 1979. A Guide to Econometrics, The MIT Press, Cambridge, MA.
Kuo, A. Y. and D. F. Hayes. 1991. Model for Turbidity Plume Induced by Bucket
Dredge, Journal of Waterway, Port, Coastal, and Ocean Engineering, Vol. 117, No. 6,
Nov/Dec.
Pindyck, R.S. and D.L. Rubinfeld. 1981. Econometric Models and Economic Forecasts
(2nd edition). McGraw-Hill, NewYork.
Quantitative Environmental Analysis, LLC (QEA) 1998. Supplementary Remedial
Studies Data Report, Grasse River Study Area, Massena, New York, Aluminum
Company of America, Massena, New York. October 1998.
Hudson River PCBs Superfund Site
Engineering Performance Standards
137
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Rice, C.P. and White, D.S. (1987) PCB availability assessment of river dredging using
caged clams and fish, Environmental Toxicology and Chemistry. 6:259-274.
Sloan, R. 1999. NYSDEC. Briefing on 1997 striped bass PCB results. Sent to J.
Colquhoun, NYSDEC. February 11.
Thackston, E.L., and Palermo, M. R., 2000. "Improved Methods for correlating turbidity
and suspended solids monitoring," DOER Technical Note E8, U.S. Army Research and
Development Center, Vicksburg, MS.
Theil, H. 1961. Economic Forecasts and Policy. North-Holland, Amsterdam.
U.S. Army Corps of Engineers (USACE). 2001. Final Pre-Design Field Test Dredge
Technology Evaluation Report, New Bedford Harbor Superfund Site, New Bedford,
Massachusetts. Prepared by Foster Wheeler Environmental Corporation, Boston,
Massachusetts. August 2001.
United States Environmental Protection Agency (USEPA). 1997. Phase 2 Report, Further
Site Characterization and Analysis, Volume 2C -Data Evaluation and Interpretation
Report (DEIR), Hudson River PCBs RI/FS. Prepared for USEPA Region 2 and USACE
by TAMS Consultants, Inc., the Cadmus Group, Inc., and Gradient Corporation. February
1997.
USEPA 2000a. Phase 2 Report: Further Site Characterization And Analysis. Volume 2F
Revised Human Health Risk Assessment, Hudson River PCBs Reassessment RI/FS.
Prepared for USEPA Region 2 and USACE, Kansas City District by TAMS Consultants,
Inc. and Gradient Corporation. November 2000.
USEPA 2000b. Phase 3 Report: Feasibility Study, Hudson River PCBs Reassessment
RI/FS. Prepared for EPA Region 2 and the US Army Corps of Engineers (USACE),
Kansas City District by TAMS Consultants, Inc. December 2000.
USEPA 2000c. Phase 2 Report: Revised Baseline Modeling Report, Hudson River PCBs
Reassessment RI/FS. Prepared for EPA Region 2 and the US Army Corps of Engineers
(USACE), Kansas City District by TAMS Consultants, Inc. January 2000.
USEPA 2000d. Guidance for the Data Quality Objectives Process. EPA/600/R-96/055.
August 2000.
USEPA 2000e. Phase 2 Report Further Site Characterization and Analysis, Volume 2F -
Revised Baseline Ecological Risk Assessment Hudson River PCBs Reassessment RI/FS.
Prepared TAMS Consultants, Inc. and Menzie-Cura & Associates, Inc. November 2000.
USEPA. 2002a. Record of Decision, Hudson River PCBs Reassessment RI/FS, USEPA
Region 2, New York, http://www.epa.gov/hudson/rod.htm
Hudson River PCBs Superfund Site
Engineering Performance Standards
138
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
USEPA. 2002b. Responsiveness Summary for the Hudson River PCBs Site Record of
Decision. Prepared by TAMS Consultants, Inc. January 2000.
USEPA. 2002c. Principles for Managing Contaminated Sediment Risks at Hazardous
Waste Sites. OSWER Directive 9285.6-08, 12 February 2002, Office of Solid Waste and
Emergency Response, Washington, D.C.
http://www.epa.gov/superfund/resources/remedy/pdf/92-85608-s.pdf
USGS. 2000. A Mass Balance Approach for Assessing PCN Movement during
Remediation of a PCB-Contaminated deposit on the Fox River, Wisconsin. USGS Water-
resources Investigations Report 00-4245, December.
USGS, 2002. National Field Manual for the Collection of Water-Quality Data,
Techniques of Water-Resources Investigations, Book 9, Handbooks for Water-Resources
Investigations, Section 4.1.1, http://water.usgs.gov/owq/FieldManual/. Visited 11/2002.
Last updated 9/1999.
Additional references are provided in the attachments.
Hudson River PCBs Superfund Site
Engineering Performance Standards
139
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Tables
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Resuspension - April 2004
-------
Table 1-1
Resuspension Criteria
Parameter
Resuspension Standard
Threshold
Control Level2
Evaluation Level
Limit
Duration
Limit
Duration
Limit | Duration
Far-Field PCB
Concentration
Total PCBs
500 ng/L
Confirmed
Occurrence 8
350 ng/L
7-day running average
Far-Field Net PCB
Load3
Total PCBs
65 kg/year4
Dredging Season
Tri+ PCBs
22 kg/year4
Total PCBs
600 g/day
7-day running average
300 g/day
7-day running average
Tri+ PCBs
200 g/day
100 g/day
Far-Field Net Suspended
Solids Concentration5'6
All Sections
24 mg/L
Daily dredging period
(> 6 hrs.)
OR
24 hrs. on average
12 mg/L
6-hour running average net
increase
OR
average net increase in the
daily dredging period if the
dredging period is less than 6
Near-Field (300 m) Net
Suspended Solids
Concentration7
Sections 1 & 3
100 mg/L
Daily dredging period
(> 6 hrs.)
OR
24 hrs. on average
100 mg/L
6-hour running average net
increase
UK
average net increase in the
daily dredging period if the
dredging period is less than 6
U^r.
Sections 2
60 mg/L
60 mg/L
Near-Field (100 m and
Channel-Side) Net
Suspended Solids
Concentration7
All Sections
700 mg/L
3 continuous hrs. running
average.
Notes:
1. Implemention of the criteria is described in Section 3.
2. Engineering contingencies for the Control Level will include temporary cessation of the operation.
3. Net increases in PCB load or suspended solids concentration refers to dredging related releases over baseline as defined in the text.
4. During Phase 1, half of the anticipated average production rate will be achieved. As a result, the total allowable export for Phase 1 is half of the fullscale value
of 130 kg/year for a total of 650 kg for the entire program. This is equivalent to the 600 g/day Total PCB release at the target productivity schedule, during the
dredging season from May to November. The Tri+ PCB values are 22 kg/year for Phase 1, 44 kg/year for full scale production and 220 kg for the entire program.
5. The increased far-field monitoring required for exceedance of suspended solids criteria must include a sample timed so as to capture the suspended solids plume's arrival at the far-field station.
6. The monitoring requirements for exceedance of the suspended solids action levels are increased frequency sampling at the nearest far field station. The
increased frequency at this station will be the same as the frequency required for the PCB action levels.
7. All remedial operations will be monitored in the near-field during Phase 1, including backfilling.
8. Exceedance of the Resuspension Standard must be confirmed by the 4 samples that are collected once a concentration greater than 500 ng/L Total PCB is
detected. The average of the 5 sample concentrations is compared to the Resuspension Standard. The Resuspension Standard is exceeded if the average
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 1-2
Sampling Requirements on a Weekly Basis - Upper River Far-Field Stations
Routine Monitoring
PCB
Laboratory Analyses
Probe
Number of Samples per
Week
Lab
Turn-
TSS
Integrating
Do,
Temp.,
Laser
Around
Congener-Specific
DOC &
(1/3-
Sampler
ph,
Particle
Time (hr.)
PCBs Whole Water
Susp. OC
SS
hours)
for PCBs
Turbidity
Cond.
Counter
RM 197.0 - Bakers Falls Bridge
72
1
Discrete
RM 194.2 - Fort Edward
72
7
7.5
7.5
56
0.5
Continous
Discrete
Discrete5
RM 188.5- TIDam
24
7
7.5
7.5
56
0.5
Continous
Discrete
None
RM 181.4- Schuylerville
24
7
7.5
7.5
56
0.5
Continous
Discrete
None
RM 163.5 - Stillwater
72
7
7.5
7.5
56
0.5
Continous
Discrete
None
RM 156.5 - Waterford
72
7
7.5
7.5
56
0.5
Continous
Discrete
None
Samples/Week
36
38.5
38.5
280
2.5
PCB analyses/week
38.5 or 5.5 /day |
Evaluation Level
PCB
Laboratory Analyses
Probe
Number of Samples per
Week
Lab
Turn-
Integrating
Do,
Temp.,
Laser
Around
Congener-Specific
DOC &
SS (1/3-
Sampler
ph,
Particle
Time (hr.)
PCBs Whole Water
Susp. OC
SS
hours)
for PCBs
Turbidity
Cond.
Counter
RM 197.0 - Bakers Falls Bridge
72
1
Discrete
RM 194.2 - Fort Edward
72
7
7.5
7.5
56
0.5
Continous
Discrete
Discrete5
RM 188.5- TIDam
24
14
14.5
14.5
56
0.5
Continous
Discrete
None
RM 181.4- Schuylerville
24
14
14.5
14.5
56
0.5
Continous
Discrete
None
RM 163.5 - Stillwater
72
7
7.5
7.5
56
0.5
Continous
Discrete
None
RM 156.5 - Waterford
72
7
7.5
7.5
56
0.5
Continous
Discrete
None
Samples/Week
50
52.5
52.5
280
2.5
PCB analyses/week
52.5 or 7.5 /day |
Control Level
PCB
Laboratory Analyses
Probe
Number of Samples per
Lab
Week
Do,
Turn-
TSS
Integrating
Temp.,
Laser
Around
Congener-Specific
DOC &
(1/3-
Sampler
ph,
Particle
Time (hr.)
PCBs Whole Water
Susp. OC
SS
hours)
for PCBs
Turbidity
Cond.
Counter
RM 197.0 - Bakers Falls Bridge
72
1
Discrete
RM 194.2 - Fort Edward
72
7
7.5
7.5
56
0.5
Continous
Discrete
Discrete5
RM 188.5- TIDam
24
21
22
22
56
1
Continous
Discrete
None
RM 181.4- Schuylerville
24
21
22
22
56
1
Continous
Discrete
None
RM 163.5 - Stillwater
24
7
7
56
7
Continous
Discrete
None
RM 156.5-Waterford
24
7
7
56
7
Continous
Discrete
None
Samples/Week
50
66.5
66.5
280
16.5
PCB analyses/week
66.5 or 9.5 /day |
Threshold4
Number of Samples per
Day Only
PCB
Lab
Turn-
Around
Time (hr.)
Laboratory Analyses
Probe
Congener-Specific
PCBs Whole Water
TSS
DOC & (1/3-
Susp. OC SS hours)
Integrating
Sampler
for PCBs
Do,
Temp., Laser
Ph, Particle
Turbidity Cond. Counter
RM 197.0 - Bakers Falls Bridge
RM 194.2 - Fort Edward
RM 188.5- TI Dam_,_
RM 181.4 - Schuylerville ,
RM 163.5 - Stillwater
RM 156.5-Waterford
72
72
24
24
24
24
1
1
4
4
4
4
1 1
1 1 8
5 5 8
5 5 8
5 5 8
5 5 8
1/2-weeks
1
1
1
1
Discrete
Continous Discrete Discrete5
Continous Discrete None
Continous Discrete None
Continous Discrete None
Continous Discrete None
Samples/day
18
22 22 40
4
PCB analyses/day 22 /day |
Note:
1. TI Dam and Schuylerville will be representative stations while the dredging is ongoing in the Phase 1 areas and will be sampled more intensely.
Samples will be composited from hourly grab samples for theResuspension Standard threshold at these two stations.
2. TSS sampling every 3- hours will be required for compliance at the nearest representative far-field stations only if the semi-quantative relationship
between TSS and a surrogate is not sufficiently conservative (See Section 4). Samples collected at the other stations will have 12-hour turnaround.
3. The turnaround time for PCB analyses from the integrating sampler will only be specified when the information is needed quickly for comparison to
the resuspension criteria. For the Resuspension Standard the integrating sample turnaround times will be 24-hours for the two representative far-field
stations (TI Dam and Schuylerville stations) and 72-hours for the stations farther downstream (Stillwater and Waterford stations). For the Control Level
at Stillwater and Waterford, the turnaround times will be 72-hours and 24-hours, respectively.
4. The monitoring for the Resuspension Standard threshold is required for one day only for verification of the elevated concentration.
5. Continuous laser particle analysis is required only ar the nearest far-field station to the dredge operation. For the purpose of this table, the Phase-1
area was assumed to occur in the TIP
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 1-3
Sampling Requirements on a Weekly Basis - Lower River Far-Field Stations
Lower River Sampling Requirements on a Weekly Basis
Routine Monitoring1
Laboratory Analyses
Probe
Lab
Congener-
Turn-
specific
Turbidity,
Around
PCBs Whole
DOC &
Temp., pH,
Dissolved
Time (hr.)
Water
Susp. OC
SS
Cond.
Oxygen
Mohawk River at Cohoes
72
0.25
0.25
0.25
0.25
0.25
RM 140 - Albany
72
0.25
0.25
0.25
0.25
0.25
RM 77 - Highland
72
0.25
0.25
0.25
0.25
0.25
Samples/Week
0.75
0.75
0.75
0.75
0.75
Non-Routine Monitoring2
Laboratory Analyses
Probe
Lab
Congener-
Turn-
specific
Turbidity,
Around
PCBs Whole
DOC &
Temp., pH, Dissolved
Time (hr.)
Water
Susp. OC SS
Cond. Oxygen
Mohawk River at Cohoes
24
1
1 1
1 1
RM 140 - Albany
24
1
1 1
1 1
RM 77 - Highland
24
1
1 1
1 1
Samples/Week
3
3 3
3 3
Notes:
1. Routine monitoring samples for the Lower Hudson stations are collected once per month.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 1-4
Sampling Requirements on a Weekly Basis - Upper River Near-Field Stations
Near-Field Sampling Requirements on a Weekly Basis 1,2,3,4
Routine Monitoring (Use of continuous reading probe to indicate suspended solids concentrations.)
No. of SS
No. of Discrete Measurements
No. of
No. of
Laboratory
by Laser Particle Counter
Continuous
Analyses J
per week
Monitors
Operations
per week
1
35
35
5
2
70
70
10
3
105
105
15
4
140
140
20
5
175
175
25
6
210
210
30
7
245
245
35
8
280
280
40
9
315
315
45
10
350
350
50
Non-Routine Monitoring (If the surrogate analysis fails to predict TSS concentrations adequately.)1'5'6
Number of SS Laboratory Samples with 3-Hour Turn-Around, per Week
Discrete Probe Measurements
No. of
Number of Stations (where surrogate is out of compliance)
All Stations7
for Turbidity & Laser
Particle Counter (No. per
Operations
1
2 3
4
5
week)
1
49
98 147
196
245
35
2
98
196 294
392
490
70
3
147
294 441
588
735
105
4
196
392 588
784
980
140
5
245
490 735
980
1,225
175
6
294
588 882
1,176
1,470
210
7
343
686 1,029
1,372
1,715
245
8
392
784 1,176
1,568
1,960
280
9
441
882 1,323
1,764
2,205
315
10
490
980 1,470
1,960
2,450
350
Notes:
1. A surrogate must be established to determine compliance with the TSS based resuspension criteria. Only if this surrogate relationship fails
to adequately predict TSS concentrations will sampling for TSS concentrations every 3-hours with a 3-hour turnaround be required. If
compliance is based on TSS samples, 1 sample will be collected an hour prior to beginning of the operation and at least 3 samples will be
collected at 1-hour intervals after completing for the day.
2. One TSS samples will be collected per day per station to confirm the surrogate semi-quantitative relationship.
3. If a TSS resuspension criteria is exceeded at a monitoring station, two TSS samples will be collected per day at that station to confirm the
surrogate semi-quantitative relationship.
4. Turbidity, temperature, pH, conductivity and dissolved oxygen will be monitored continuously at each of the five near-field stations.
5. Assumed hours of operation: 14 hours of dredging per 24 hours of operation per day for the quantities above.
6. Exceedence of a suspended solids criteria will also prompt monitoring at the representative far-field station nearest to the location of the
exceedence at the frequency of sampling indicated for the action level.
7. If containment is used in an area, 6 stations will be required, increasing the total
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 1 - 5
Case Study Resuspension Summary Table
Project/Site Name
Dates of
Operation
Project Setting
Water Quality Parameters Monitored
Water Quality Standard
Water Quality Monitoring Stations
Water Quality Measurements Reported During Dredging
Fox River:
Kimberly, Wisconsii
DepositN
November 1998
to December
1998 (Phase I);
August 1999 to
November 1999
(Phase II)
Riverine
Turbidity, TSS, andPCBs
Turbidity - Threshold limit based on hourly
average value; Specific threshold not stated ii
materials reviewed; PCBs- water column
concentrations compared to pre-dredge
concentrations and upstream samples versus
downstream samples compared-specific
threshold not indicated
Real time turbidity monitoring at 6 stations: (1)
upstream, (1) side gradient, (1) downstream, (1) atlLP
water intake, (1) at the ILP effluent discharge, and (1)
within the contained dredge area. Measured turbidity at
50% total water depth
Average PCB water column concentration during Phase I (1998)
downstream of dredging was 11 ng/L compared to an average
upstream measured concentration of 3.2 ng/L during dredging.
Baseline concentration before Phase I was 5.0 ng/L. Average
downstream PCB concentration during Phase II (1999) was 24
ng/L compared to an average upstream PCB concentration of 14
ng/L. Minor differences between upstream and downstream
turbidity during dredging. No apparent difference in TSS data
collected upstream and downstream of the dredge was noted fron
measurements collected during dredging.
Fox River:
Green Bay,
Wisconsin
SMU 56/57
Phase I
August to
December 1999
(Phase I)
Riverine
Turbidity, TSS, andPCBs
Not indicated in documents reviewed
Real time turbidity monitoring at 6 locations: (1)
upstream dredge outside turbidity barrier,(1) upstream
dredge inside turbidity bairier;(l) side stream dredge
outside turbidity barrier;(l) downstream dredge outside
turbidity bairier;(l) downstream dredge inside turbidity
barrier; (1) at Fort James water intake - Each meter
located in water column at 50-60% of the water depth
for location
Average PCB water column concentration downstream of the
dredge was 90 ng/L compared to an upstream concentration of 51
ng/L during dredging and a baseline concentration prior to
dredging of 52 ng/L. Turbidity monitors downstream of the
dredge, outside the silt curtain were indicative of periodic
turbidity increases. TSS samples only showed minor differences
between the upstream and downstream locations. Monthly
averaged turb idity data indicated that a high turb idity of 41 NTU
occurred during the first month of dredging (August) downstrean
of the dredge, outside the silt curtain.
Fox River:
Green Bay,
Wisconsin
SMU 56/57
Phase II
August 2000 to
November 2000
(Phase II)
Riverine
Turbidity, TSS, andPCBs
Turbidity - Reached threshold if downstream
turbidity reading was two or more times
higher than the upstream reading and cause
was related to dredging; Specific PCB
threshold not indicated in documents
reviewed
Real time turbidity monitoring at 3 locations: (1)
upstream of silt curtain at the Fort James water intake;
(1) 10-ft downstream of the silt curtain; and (1) 50-ft
downstream of the silt curtain
Upstream and downstream turbidity values never varied by more
than a factor of two during dredging. Contractor didnotperfoim
PCB water column monitoring since turbidity threshold was nevt
exceeded however PCB water column sampling was performed
by the USGS.
Manistique River,
Michigan
1995 - 1999
Riverine
Turbidity, TSS, andPCBs
TSS concentration less than 2X the
background turbidity within 50-feet of the
dredge head; Literature reviewed stated that
this level was achieved within 10-feet of the
dredge head. PCB water quality threshold not
stated in literature reviewed. It was noted tha
PCB concentrations were compared to pre-
dredge water column PCB concentrations
For 1997 Dredging: seven samples from one station
near dredge; one sample from upstream; six samples
from a station downstream; and two samples from a
station outside of the dredge area. For 1998: 9 samples
from station upstream of dredge; 8 samples from
locations downstream of dredge- distance and exact
location not specified.
In 1997: avg. PCB water column concentrations outside dredge
area was 0.37mg/L and avg. [PCB] downstream of dredge was
0.23 mg/1 compared to pre-dredge concentration of 0.001 mg/L.
The background sample collected during this event was 0.062
mg/L PCBs. In 1998: Avg. upstream [PCB] was 0.093 mg/L anc
the average [PCB] downstream was 0.066 mg/L.
Reynolds Metals: St.
Lawrence River,
Massena. NY
April 2001
through
November 2001
Riverine
Turbidity and water column samples
(PCBs, PAHs, and PCDFs); TSS was not
measured in this project
Turbidity action level of 25 NTU above the
background level, which was derived based
on 28 NTU action level used at GM Massena
The action levels for water column samples
were 2 ug/L of PCBs, 0.2 ug/L for PAHs and
detectable PCDFs above the practical
quantitation limit (PQL).
Monitoring was conducted at different locations for
each project phase (sheetpile installation, dredging,
capping, and sheet pile removal); All locations
identified in Final Case Study Table (Appendix A of th
Resuspension standard). For dredging: (4) stations
outside the sheet piling- oneupcuirent (100ft from the
active dredge) and 3 down current stations (10, 150 an(
300 ft from the sheetpile wall closest to the dredge
being monitored). Within the sheetpiling-Water Qualitj
was monitored at 12 to 19 different stations based on
dredge location.
Outside the sheet piling : Turbidity during dredging ranged
between 0.5 to 1.5 NTUs. During dredging, water column PCB
concentrations ranged between 0.05 to 0.53 ug/L. and PAH and
PCDF were non-detect in samples analyzed
GM Massena: St
Lawrence River,
Massena, NY
May 1995
through
December 1995
Riverine
Turbidity, PCBs, PAHs
Action level was selected based on a 1994
site-specific bench-scale laboratory
correlation between TSS and turbidity, and
experience in previous dredging projects.
Downstream turbidity 28NTUs above
background corresponded to a downstream
TSS of 25 mg/L above background. For
PCBS: 2 ug/L (at downstream monitoring
locations)
Visual observations and real-time turbidity monitoring
at 3 locations: 50 feet upstream of western extent of
control system, two between 200 feet and 400 feet
downstream of easternmost active installations.
Measurements collected from 50% water depth.
Water column sampling at the same two downstream
locations as the turbidity measurements.
In 18 out of 923 turbidity samples, the 28 NTU action level was
exceeded (31-127 NTU) at 1-ft below the water surface for a
duration of 2-8 minutes, on average, however 2 exceedances
lasted for 15 minutes and 45 minutes respectively. Exceedance
deteimined to be a result of water overflow from the dredge area
over the sheet piling due to inadequate height/installation. PCBs
monitored at same station as turbidity. High PCB concentrations
correlated with times where high turbidity (> 28 NTU) measured
Filtered [PCB] ranged between 0.94-2.4 ug/L and unfiltered
ranged between 4.51 to 9.84 ug/L. These PCB measurements
collected at end of Phase I after sheet piling removed.
Cumberland Bay:
New York
April 1999 to
May 2000
Western side ol
Lake
Champlain
TSS, turb idity andPCB
Turbidity was used only to alert the operators
of a potential re-suspension problem-not
associated with an action level.Operational
Monitoring: TSS 25 mg/L above
background. Compliance Monitoring
(outside turbidity barrier): TSS 4 mg/L above
background. When TSS action level was
exceeded, dredging was suspended or
modified.
Operational Monitoring: Real-time turbidity
monitoring in 2 locations: on dredge head and using a
float that trailed behind the dredge.
Compliance Monitoring: FourOBS-3 sensor stations
which changed for each active work zone: one sensor ii
a background location (near breakwater) and three
sensors outside the perimeter of the woik zone silt
curtain (an additional temporary sensor was located
near Georgia-Pacific's industrial water intake).
Documentation Monitoring: Six fixed turbidity
monitoring (TM) buoys (in 1999 outside perimeter
turbidity curtain; 2000 locations different).
Documentation reviewed indicated that the TSS levels were not
exceeded and dredging was never suspended.
United Heckathorn:
Parr Canal and
Lauritzen Channel
on the San Francisco
Bay
August 1996
through March
1997
Bay area -
shipping
inlet/slip
TSS and Contaminants of Concern: DDT
and Dieldrin
Surface water: Dieldrin 0.14ng/L and DDT
0.59ng/L both based on EPA AWQ (Ambien
water Quality criteria) and also based on
human health standards (risk)
Four water quality sampling stations- Locations were
established both upstream and downstream of area
being dredged and downstream/outside channel/ship
inlet/slip in the harbor and bay at both ends
Data not available in documents reviewed for water qualify data
during dredging however it was noted that the area is extremely
turbid naturally due to ship traffic; Post-dredge water quality dat<
collected 4-months after dredging indicated concentrations equal
to or greater than predredge conditions. This was a result of
incomplete dredging near banks and around structures. Dredging
not a success at this site and further action to be taken.
Hudson RiverPCBs Superfund Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2 Dredging Resuspension - April 2004
-------
Table 1 - 5
Case Study Resuspension Summary Table
Project/Site Name
Dates of
Operation
Project Setting
Water Quality Parameters Monitored
Water Quality Standard
Water Quality Monitoring Stations
Water Quality Measurements Reported During Dredging
Grand Calumet
River, Indiana
Dredging Began
November 2002
(currently in
progress)
Riverine
Level 1: Flow, total ammonia, specific
conductance, DO, pH, sulfides, temp., am
turbidity monitored daily by multi-
parameter automatic data logger system;
Level 2: microtox chemical testing for
acute and chronic toxicity; Level 3:
chemical monitoring for total ammonia,
pH, sulfides, temp, free cyanide, hardness
oil and grease, TSS, dissolved aluminum,
dissolved copper, dissolved lead, total
mercury, dissolved zinc, select VOCs, an<
total PCBs; Each Level Monitoring is
conducted concurrently at a pre-set
frequency. A contingency plan exists for
each Level monitoring in the event that a
high measurement is recorded.
IDEM (Indiana Department of Environmenta
Management) chronic and acute state surface
water criteria
(1) upstream background sampling location; (1) locate(
near mid-channel 200-yd downstream from open water
dredge; (1) downstream sampling site below 5-mile
dredge area; (1) proposed sample location for
verification analysis located 200-yd upstream of open
water dredging in cell c
Data Not yet available; dredging cuirently underway
New Bedford Harbot
(Hot Spots), New
Bedford,
Massachusetts
April 1994 to
September 1995
Estuaiy/Bay
PCBs (24-hr turn-around) and metals.
PCBs (Total PCBs: dissolved and
particulate tested separately and
summed).
PCBs: 1.3 mg/L determined by a pilot study
and a Maximum cumulative transport (MCT)
of PCBs during the entire operation of 240
Kg PCBs.
Down current locations: 50 ft, 300 ft, 700 ft, and 1,000
ft. from dredging area. Background measurements: ~
1,000 ft up-current of dredging operations. Sampling
depth: ~ mid-depth in the water column.
By the end of project, a total PCB transport of 57 kg was
reported. Thus, the MCL was not exceeded. Toxicity tests
completed during dredging were not indicative of acute toxicity
and PCB accumulation in mussels was not significantly greater
then predredge measurements.
New Bedford Harbot
(Pre-Design Field
Test), New Bedford,
Massachusetts
Demonstration
Project in August
2000
Estuaiy/Bay
TSS, turbidity and PCBs (dissolved and
particulate, PCB congeners)
PCBs: No set limit since background
concentrations exceeded Federal criteria
however did set the maximum Cumulative
Transport (MCT) for PCB loss from dredgini
at the limit of mixing zone (300 ft from the
dredge) of 400 kg PCBs throughout entire
dredging project. Turbidity: 50 NTU
above background at limit of mixing zone
(300 ft from the dredge)
2 Monitoring stations 300 ft away from dredge;
additional sampling as required 600 ft from dredge.
Background measurements ~ 1,000 ft up-current of
dredging operations.
Turbidity measurements exceeded the 50 NTU threshold
infrequently at the 300-ft limit of the mixing zone and no further
action was taken. Bioassay tests completed when turbidity
exceeded 50 NTU were not indicative of an ecological impact.
Commencement
Bay: Hylebos
Waterway
Small hot spot
dredging October
2002; Full-scale
dredging begun
2003
Tidal Waterway
Turbidity and dissolved oxygen (system
currently exhibits a low dissolved oxygen
level and managers do not want dredging
to deplete it any further)
It is anticipated that the turb idity standard wi
be set at either 20 NTU or 50 NTU over
background.
2 anticipated monitoring stations; one near dredge heac
and one at the limit of the mixing zone (300-ft from the
dredge)
Data not yet available.
Commencement
Bay: Thea, Foss,
Wheeler, Osgood
Waterway
Full-scale
dredging begun
2003
Tidal Waterway
Turbidity however water quality
monitoring plan still in design
It is anticipated that the turb idity standard wi
be set at either 20 NTU or 50 NTU over
background.
2 anticipated monitoring stations; one near dredge heac
and one at the limit of the mixing zone (300-ft from the
dredge)
Data not yet available.
Hudson RiverPCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2 Dredging Resuspension - April 2004
-------
Table 2-1
Summary of Case Studies for PCB Losses Due to Dredging
Total PCBs
Total PCBs
Resuspension
Percentage
Removed
Loss
Lost
Project
Period of Dredging
(kg)
(kg)
(%)
GE Hudson Falls Dredging
3,890
14
0.36%
Oct.-Dec. 1997, Aug.-Nov. 1998
New Bedford Harbor Hot Spots
1994-1995
43,700
57
0.13%
Fox River Deposit N
Nov. - Dec. 1998 (Phase I)
Aug. -Dec. 1999 (Phase II)
111
4.20
3.5% - 14% (1)
Fox River SMU 56/57
Aug. - Nov. 1999 (Phase I)
1,490
22
2.2% (2)
Notes:
(1) Average Daily Percentage Loss varied over dredge season based on dredge location and
uncertainty associated with PCB removal estimation.
(2) PCB Percentage Loss based on USGS study while other values taken from the
SMU 56/57 Final Summary report (September 2001).
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-2
Far-Field Forecast Model Runs Completed for the Performance Standard
Completed Sim
ulations 4
Rate of PCB
Release 1
Period of
Upper Hudson
Lower Hudson
Scenario 5
Description
g/day (kg/yr)3
Dredging
Start Year
HUDTOX
FISHRAND
Farley
FISHRAND
-
MNA
NA
-
-
X
X
X
X
-
No resuspension
0(0)
6
2004
X
X
X
X
d004
No resuspension
0(0)
6
2006
X
X
X
X
-
2.5% Export2
1,700 (350)
6
2004
X
X
X
X
srOl
300 g/day
300 (70)
6
2006
X
X
X
X
sr02
600 g/day
600 (130)
2006
X
X
X
X
sr04
350 ng/L
1,600 (340)
6
2006
X
X
X
X
-
Accidental Release
600 (130)
6
2006
X
Notes:
1. All PCB resuspension scenarios were based on a resuspension release rate (near-field release)
at the specified percentage of dredging loss unless noted otherwise.
2. The model run included with the Responsiveness Summary for the ROD is effectively a
2.5 percent export scenario since all PCBs were loaded as dissolved phase. See text
for further discussion.
3. The rates are based on 7 months of operation, 7 days per week at 14 hours per day.
4. x = completed for ROD
X = completed for this report
5. The d004 and srOl and sr04 and srOx series of scenarios were created during the development of the performance standards.
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Dredging Resuspension - April 2004
-------
Table 2-3
Upper Hudson Conceptual Dredging Schedule
Sediment removal season
Dredging
Location
Speed of
operation
May 1 - Nov. 1,2006
Sec. 1
half
May 1 - Nov. 30, 2007
Sec. 1
full
May 1 - Nov. 30, 2008
Sec. 1
full
May 1 - Aug. 15,2009
Sec. 1
full
Aug. 16-Nov. 30,2009
Sec. 2
full
May 1 - Aug. 15,2010
Sec. 2
full
Aug. 16-Nov. 30,2010
Sec. 3
full
May 1 - Aug. 15,2011
Sec. 3
full
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-4
Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
No Resuspension (d004)
350 ng/L (sr04)
600 g/day (srOl)
Monitored Natural Attenuation
Upper River
River Section
River Section
River Section
Upper River
River Section
River Section
River Section
Upper River
River Section
River Section
River Section
Upper River
River Section
River Section
River Section
Year
Average
1 (RM 189)
2 (RM 184)
3 (RM 154)
Average
1 (RM 189)
2 (RM 184)
3 (RM 154)
Average
1 (RM 189)
2 (RM 184)
3 (RM 154)
Average
1 (RM 189)
2 (RM 184)
3 (RM 154)
1998
3.317
6.813
9.271
1.537
3.316
6.807
9.276
1.537
3.316
6.807
9.276
1.537
3.353
6.774
9.659
1.529
1999
3.328
6.908
9.406
1.510
3.328
6.909
9.410
1.509
3.328
6.909
9.410
1.509
3.212
6.621
8.877
1.501
2000
2.866
5.747
8.346
1.300
2.865
5.751
8.338
1.300
2.865
5.751
8.338
1.300
2.791
5.563
8.028
1.292
2001
2.582
5.098
7.588
1.177
2.583
5.104
7.585
1.177
2.583
5.104
7.585
1.177
2.504
4.924
7.210
1.171
2002
2.370
4.841
6.925
1.053
2.372
4.848
6.924
1.054
2.372
4.848
6.924
1.054
2.301
4.705
6.571
1.047
2003
2.182
4.340
6.471
0.978
2.182
4.338
6.474
0.978
2.182
4.338
6.474
0.978
2.129
4.290
6.090
0.980
2004
2.290
5.285
6.356
0.946
2.290
5.286
6.354
0.946
2.290
5.286
6.354
0.946
2.204
5.084
5.934
0.942
2005
1.905
3.912
5.712
0.816
1.911
3.910
5.740
0.821
1.908
3.909
5.726
0.819
1.852
3.739
5.523
0.812
2006
1.617
2.996
5.119
0.716
1.703
3.111
5.350
0.770
1.666
3.076
5.237
0.746
1.574
2.890
4.904
0.716
2007
1.487
2.838
4.669
0.647
1.709
3.461
5.141
0.739
1.614
3.225
4.920
0.697
1.474
2.862
4.489
0.654
2008
1.297
2.318
4.226
0.571
1.673
3.762
4.743
0.694
1.525
3.216
4.582
0.634
1.371
2.774
4.168
0.586
2009
0.964
1.573
2.949
0.489
1.323
2.317
3.769
0.687
1.106
1.907
3.140
0.583
1.262
2.616
3.877
0.519
2010
0.595
0.899
1.355
0.398
0.928
1.012
1.835
0.753
0.707
0.943
1.411
0.535
1.116
2.321
3.533
0.440
2011
0.447
0.661
0.847
0.332
0.817
0.736
1.122
0.781
0.568
0.697
0.901
0.483
0.971
1.921
3.164
0.388
2012
0.404
0.723
0.786
0.269
0.631
0.774
0.999
0.537
0.469
0.747
0.818
0.350
0.878
1.851
2.879
0.324
2013
0.342
0.568
0.717
0.229
0.515
0.600
0.883
0.433
0.389
0.572
0.734
0.291
0.791
1.682
2.601
0.287
2014
0.318
0.593
0.669
0.199
0.453
0.602
0.803
0.361
0.353
0.582
0.675
0.248
0.742
1.666
2.396
0.258
2015
0.289
0.520
0.638
0.178
0.400
0.524
0.751
0.312
0.316
0.506
0.638
0.219
0.686
1.535
2.229
0.237
2016
0.294
0.586
0.651
0.170
0.391
0.589
0.750
0.287
0.317
0.573
0.648
0.205
0.680
1.610
2.126
0.231
2017
0.296
0.671
0.612
0.161
0.379
0.672
0.704
0.260
0.315
0.660
0.610
0.190
0.649
1.573
1.978
0.221
2018
0.272
0.606
0.574
0.149
0.344
0.605
0.665
0.233
0.289
0.595
0.577
0.173
0.593
1.437
1.765
0.210
2019
0.281
0.710
0.567
0.140
0.341
0.702
0.656
0.210
0.295
0.694
0.572
0.161
0.577
1.497
1.619
0.200
2020
0.243
0.584
0.502
0.125
0.292
0.579
0.584
0.180
0.253
0.571
0.507
0.142
0.512
1.270
1.480
0.182
2021
0.217
0.471
0.482
0.117
0.260
0.468
0.557
0.164
0.226
0.459
0.486
0.131
0.460
1.080
1.365
0.171
2022
0.215
0.476
0.477
0.114
0.253
0.473
0.548
0.155
0.222
0.464
0.482
0.126
0.450
1.093
1.296
0.166
2023
0.216
0.529
0.454
0.108
0.247
0.524
0.514
0.142
0.222
0.517
0.461
0.118
0.435
1.088
1.225
0.158
2024
0.195
0.484
0.417
0.094
0.219
0.480
0.463
0.122
0.200
0.474
0.427
0.102
0.385
0.939
1.123
0.139
2025
0.176
0.415
0.391
0.088
0.196
0.413
0.426
0.110
0.181
0.406
0.402
0.094
0.350
0.842
1.019
0.129
2026
0.163
0.357
0.377
0.084
0.180
0.355
0.405
0.103
0.166
0.347
0.388
0.089
0.325
0.757
0.952
0.124
2027
0.183
0.490
0.380
0.083
0.197
0.488
0.403
0.100
0.186
0.483
0.387
0.088
0.339
0.888
0.920
0.121
2028
0.177
0.509
0.353
0.076
0.189
0.508
0.371
0.090
0.179
0.504
0.353
0.080
0.322
0.863
0.875
0.111
2029
0.158
0.414
0.337
0.072
0.168
0.412
0.351
0.084
0.159
0.407
0.332
0.076
0.287
0.720
0.801
0.105
2030
0.143
0.326
0.326
0.072
0.152
0.325
0.342
0.082
0.143
0.320
0.322
0.075
0.261
0.620
0.735
0.103
2031
0.151
0.422
0.303
0.067
0.159
0.421
0.320
0.075
0.152
0.418
0.302
0.069
0.257
0.679
0.675
0.095
2032
0.138
0.362
0.288
0.064
0.145
0.362
0.305
0.071
0.139
0.357
0.289
0.066
0.234
0.602
0.610
0.091
2033
0.133
0.349
0.277
0.061
0.138
0.349
0.295
0.066
0.133
0.343
0.279
0.063
0.219
0.560
0.564
0.086
2034
0.132
0.368
0.259
0.060
0.134
0.368
0.276
0.060
0.132
0.366
0.261
0.059
0.208
0.545
0.521
0.082
2035
0.123
0.279
0.249
0.068
0.116
0.279
0.266
0.056
0.114
0.275
0.251
0.055
0.191
0.443
0.475
0.089
2036
0.148
0.356
0.242
0.087
0.124
0.356
0.258
0.051
0.125
0.352
0.244
0.055
0.209
0.504
0.446
0.104
2037
0.137
0.297
0.234
0.086
0.115
0.298
0.250
0.053
0.125
0.295
0.237
0.070
0.190
0.427
0.410
0.101
2038
0.140
0.337
0.221
0.083
0.130
0.337
0.235
0.068
0.140
0.335
0.224
0.083
0.189
0.456
0.386
0.098
2039
0.128
0.270
0.214
0.083
0.132
0.271
0.227
0.087
0.131
0.268
0.218
0.087
0.173
0.382
0.363
0.096
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-4
Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
2040
0.124
0.262
0.214
0.079
0.132
0.262
0.225
0.087
0.128
0.260
0.217
0.085
0.164
0.352
0.346
0.092
2041
0.140
0.359
0.219
0.079
0.150
0.360
0.228
0.091
0.146
0.358
0.222
0.087
0.180
0.461
0.347
0.092
2042
0.143
0.400
0.223
0.074
0.153
0.401
0.229
0.087
0.148
0.399
0.225
0.081
0.178
0.486
0.337
0.084
2043
0.123
0.318
0.202
0.068
0.132
0.318
0.206
0.080
0.129
0.320
0.205
0.075
0.155
0.386
0.316
0.078
2044
0.108
0.245
0.191
0.064
0.114
0.246
0.193
0.073
0.114
0.256
0.195
0.069
0.136
0.301
0.289
0.074
2045
0.112
0.282
0.190
0.063
0.118
0.283
0.191
0.070
0.118
0.301
0.194
0.066
0.137
0.329
0.278
0.071
2046
0.105
0.258
0.184
0.058
0.109
0.256
0.184
0.064
0.110
0.273
0.187
0.062
0.131
0.319
0.269
0.067
2047
0.109
0.284
0.187
0.058
0.112
0.271
0.187
0.065
0.112
0.285
0.190
0.062
0.153
0.474
0.261
0.066
2048
0.115
0.329
0.188
0.057
0.118
0.318
0.187
0.064
0.116
0.316
0.190
0.061
0.175
0.612
0.263
0.066
2049
0.116
0.339
0.190
0.055
0.120
0.340
0.189
0.062
0.117
0.328
0.192
0.059
0.166
0.574
0.259
0.063
2050
0.105
0.289
0.183
0.052
0.109
0.290
0.182
0.057
0.106
0.283
0.185
0.055
0.151
0.498
0.251
0.060
2051
0.101
0.286
0.180
0.047
0.104
0.287
0.178
0.052
0.104
0.294
0.182
0.050
0.140
0.457
0.242
0.055
2052
0.094
0.244
0.181
0.047
0.097
0.246
0.180
0.051
0.099
0.263
0.184
0.049
0.130
0.402
0.236
0.054
2053
0.113
0.359
0.187
0.048
0.116
0.359
0.185
0.052
0.118
0.379
0.189
0.050
0.146
0.494
0.244
0.055
2054
0.105
0.311
0.185
0.047
0.107
0.311
0.184
0.050
0.109
0.327
0.187
0.049
0.134
0.430
0.235
0.053
2055
0.098
0.274
0.182
0.045
0.100
0.274
0.180
0.048
0.101
0.287
0.183
0.047
0.125
0.383
0.231
0.052
2056
0.105
0.307
0.195
0.046
0.106
0.307
0.193
0.048
0.108
0.322
0.195
0.047
0.129
0.407
0.233
0.051
2057
0.105
0.323
0.185
0.045
0.107
0.324
0.183
0.047
0.108
0.337
0.186
0.046
0.126
0.397
0.231
0.050
2058
0.095
0.253
0.188
0.045
0.096
0.253
0.186
0.047
0.097
0.264
0.188
0.046
0.116
0.337
0.226
0.050
2059
0.109
0.356
0.181
0.043
0.110
0.356
0.181
0.045
0.111
0.366
0.182
0.044
0.127
0.422
0.228
0.047
2060
0.091
0.256
0.175
0.040
0.092
0.256
0.175
0.042
0.093
0.266
0.175
0.041
0.106
0.316
0.209
0.044
2061
0.086
0.234
0.169
0.040
0.087
0.233
0.169
0.042
0.087
0.241
0.169
0.041
0.100
0.286
0.200
0.043
2062
0.091
0.261
0.171
0.040
0.091
0.261
0.170
0.042
0.092
0.268
0.170
0.041
0.102
0.297
0.197
0.043
2063
0.091
0.261
0.172
0.041
0.091
0.260
0.171
0.041
0.092
0.266
0.171
0.041
0.101
0.296
0.196
0.043
2064
0.093
0.268
0.175
0.041
0.093
0.268
0.174
0.042
0.094
0.273
0.175
0.042
0.103
0.306
0.196
0.044
2065
0.092
0.255
0.178
0.043
0.093
0.255
0.177
0.043
0.093
0.260
0.177
0.043
0.100
0.283
0.195
0.045
2066
0.105
0.353
0.172
0.041
0.105
0.353
0.171
0.041
0.106
0.358
0.171
0.041
0.113
0.377
0.195
0.043
2067
0.095
0.275
0.180
0.042
0.095
0.275
0.179
0.042
0.096
0.279
0.179
0.043
0.101
0.301
0.183
0.044
BOLD-ITALICIZED - First occurrence of species-weighted fish fillet average PCB concentration below risk-based remediation goal of 0.05 mg/kg. Target concentrations of 0.2 mg/kg PCBs (protective at a fish consumption rate
of 0.5 lbs/month) and 0.4 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/ 2 months) are also italicized.
Upper Hudson River average is weighted by river section length. River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-5
Modeled Year-of-Compliance with Human Health Risk Assessment-Based Concentrations for various Resu:
Resuspension Scenarios
No Resuspension
(d004)
350 ng/L (sr04)
600 g/day (srOl)
MNA
Upper River Average
Human Health risk-based RG 0.05 mg/kg
>2067
>2067
>2067
>2067
Fish Target Concentration 0.2 mg/kg
2024
2025
2024
2035
Fish Target Concentration 0.4 mg/kg
2013
2015
2013
2024
River Section 1- RM 189
Human Health risk-based RG 0.05 mg/kg
>2067
>2067
>2067
>2067
Fish Target Concentration 0.2 mg/kg
>2067
>2067
>2067
>2067
Fish Target Concentration 0.4 mg/kg
2026
2030
2026
2043
River Section 2- RM 184
Human Health risk-based RG 0.05 mg/kg
>2067
>2067
>2067
>2067
Fish Target Concentration 0.2 mg/kg
2044
2044
2044
2061
Fish Target Concentration 0.4 mg/kg
2025
2028
2026
2038
River Section 3- RM 154
Human Health RG 0.05 mg/kg
2051
2055
2051
2059
Fish Target Concentration 0.2 mg/kg
2014
2020
2017
2019
Fish Target Concentration 0.4 mg/kg
2010
2014
2012
2011
Hudson River PCBs Superfund Site
Engineering Performance Standards
Note: RG = risk-based remediation goal
Upper Hudson River average is weighted by river section length. River Section 1: 6.3 miles = 15.4%;
River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-6
Estimated Non-cancer Indices via Long Term Fish Ingestion for Several Resuspension scenarios-
Adult Angler and Upper Hudson Fish
Remedial
PCB Cone.
Intake
Reference
Hazard
Alternative
in Fish
(Non-Cancer)
Dose
Index
(mg/kg ww)
(mg/kg-day)
(mg/kg-day)
Reasonable Maximum Exposure
Upper Hudson Average
No Resuspension d004
0.30
1.4E-04
2.0E-05
6.9
350 ng/L sr04
0.58
2.6E-04
2.0E-05
13
600 g/day srO 1
0.50
2.3E-04
2.0E-05
11
MNA
1.4
6.4E-04
2.0E-05
32
River Section 1 (RM 189)
No Resuspension d004
0.62
2.8E-04
2.0E-05
14
350 ng/L sr04
0.64
2.9E-04
2.0E-05
15
600 g/day srO 1
0.62
2.8E-04
2.0E-05
14
MNA
1.7
7.7E-04
2.0E-05
39
River Section 2 (RM 184)
No Resuspension d004
0.66
3.0E-04
2.0E-05
15
350 ng/L sr04
0.79
3.6E-04
2.0E-05
18
600 g/day srO 1
0.67
3.1E-04
2.0E-05
15
MNA
2.3
1.0E-03
2.0E-05
52
River Section 3 (RM 154)
No Resuspension d004
0.18
8.0E-05
2.0E-05
4.0
350 ng/L sr04
0.30
1.4E-04
2.0E-05
6.8
600 g/day srO 1
0.21
9.7E-05
2.0E-05
4.8
MNA
0.23
1.1E-04
2.0E-05
5.4
Central Tendency
Upper Hudson Average
No Resuspension d004
0.27
1.2E-05
2.0E-05
0.6
350 ng/L sr04
0.52
2.4E-05
2.0E-05
1.2
600 g/day srO 1
0.46
2.1E-05
2.0E-05
1.0
MNA
1.2
5.5E-05
2.0E-05
2.8
River Section 1 (RM 189)
No Resuspension d004
0.60
2.7E-05
2.0E-05
1.4
350 ng/L sr04
0.61
2.8E-05
2.0E-05
1.4
600 g/day srO 1
0.59
2.7E-05
2.0E-05
1.4
MNA
1.50
6.9E-05
2.0E-05
3.5
River Section 2 (RM 184)
No Resuspension d004
0.59
2.7E-05
2.0E-05
1.4
350 ng/L sr04
0.70
3.2E-05
2.0E-05
1.6
600 g/day srO 1
0.60
2.7E-05
2.0E-05
1.4
MNA
1.9
8.7E-05
2.0E-05
4.4
River Section 3 (RM 154)
No Resuspension d004
0.15
6.8E-06
2.0E-05
0.3
350 ng/L sr04
0.24
1.1E-05
2.0E-05
0.5
600 g/day srO 1
0.18
8.0E-06
2.0E-05
0.4
MNA
0.21
9.4E-06
2.0E-05
0.5
Notes: The RME non-cancer exposure time frame is seven years, while the CT time frame is 12 years.
Upper Hudson River average is weighted by river section length. River Section 1: 6.3 miles = 15.4%;
River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Dredging Resuspension - April 2004
-------
Table 2-7
Estimated cancer Indices via Long Term Fish Ingestion for Several Resuspension scenarios-
Adult Angler and Upper Hudson Fish
Remedial
PCB Cone.
Intake
Cancer Slope
Cancer
Alternative
in Fish
(Cancer)
Factor
Risk
(mg/kg ww)
(mg/kg-day)
(mg/kg-day)
Reasonable Maximum Exposure
Upper Hudson Average
No Resuspension d004
0.18
4.6E-05
2
9.3E-05
350 ng/L sr04
0.32
8.3E-05
2
1.7E-04
600 g/day srO 1
0.30
7.7E-05
2
1.5E-04
MNA
0.60
1.7E-04
2
3.3E-04
River Section 1 (RM 189)
No Resuspension d004
0.43
1.1E-04
2
2.2E-04
350 ng/L sr04
0.43
1.1E-04
2
2.2E-04
600 g/day srO 1
0.42
1.1E-04
2
2.2E-04
MNA
0.86
2.2E-04
2
4.5E-04
River Section 2 (RM 184)
No Resuspension d004
0.36
9.3E-05
2
1.9E-04
350 ng/L sr04
0.40
1.0E-04
2
2.1E-04
600 g/day srO 1
0.36
9.4E-05
2
1.9E-04
MNA
0.90
2.4E-04
2
4.9E-04
River Section 3 (RM 154)
No Resuspension d004
0.09
2.4E-05
2
4.8E-05
350 ng/L sr04
0.12
3.2E-05
2
6.4E-05
600 g/day srO 1
0.10
2.7E-05
2
5.3E-05
MNA
0.12
3.2E-05
2
6.4E-05
Central Tendency
Upper Hudson Average
No Resuspension d004
0.27
2.1E-06
1
2.1E-06
350 ng/L sr04
0.52
4.0E-06
1
4.0E-06
600 g/day srO 1
0.46
3.6E-06
1
3.6E-06
MNA
1.2
9.5E-06
1
9.5E-06
River Section 1 (RM 189)
No Resuspension d004
0.60
4.7E-06
1
4.7E-06
350 ng/L sr04
0.61
4.8E-06
1
4.8E-06
600 g/day srO 1
0.59
4.7E-06
1
4.7E-06
MNA
1.5
1.2E-05
1
1.2E-05
River Section 2 (RM 184)
No Resuspension d004
0.59
4.7E-06
1
4.7E-06
350 ng/L sr04
0.70
5.5E-06
1
5.5E-06
600 g/day srO 1
0.60
4.7E-06
1
4.7E-06
MNA
1.9
1.5E-05
1
1.5E-05
River Section 3 (RM 154)
No Resuspension d004
0.15
1.2E-06
1
1.2E-06
350 ng/L sr04
0.24
1.9E-06
1
1.9E-06
600 g/day srO 1
0.18
1.4E-06
1
1.4E-06
MNA
0.21
1.6E-06
1
1.6E-06
Notes: The RME cancer exposure time frame is 40 years, while the CT time frame is 12 years.
Upper Hudson River average is weighted by river section length. River Section 1: 6.3 miles = 15.4%;
River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Dredging Resuspension - April 2004
-------
Table 2-8
Upper Hudson River Average Largemouth Bass (Whole Fish) PCB Concentration (in mg/kg)
Year
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
Total PCB 600 g/day (srOl)
Monitored Natural Attenuation
Upper
River
Average
Section 1
(RM 189)
Section 2
(RM 184)
Section 3
(RM 154)
Upper
River
Average
Section 1
(RM 189)
Section 2
(RM 184)
Section 3
(RM 154)
Upper
River
Average
Section 1
(RM 189)
Section 2
(RM 184)
Section 3
(RM 154)
Upper
River
Average
Section 1
(RM 189)
Section 2
(RM 184)
Section 3
(RM 154)
1998
7.13
16.73
17.22
3.33
7.13
16.70
17.24
3.33
7.13
16.70
17.24
3.33
7.19
16.61
18.04
3.29
1999
7.04
17.11
16.80
3.20
7.04
17.12
16.83
3.20
7.04
17.12
16.83
3.20
6.76
16.16
15.91
3.17
2000
5.84
13.71
14.51
2.66
5.84
13.74
14.47
2.66
5.84
13.74
14.47
2.66
5.74
13.09
14.57
2.64
2001
5.29
12.01
13.33
2.47
5.30
12.04
13.32
2.47
5.30
12.04
13.32
2.47
5.13
11.34
12.94
2.45
2002
4.91
11.63
12.30
2.20
4.92
11.66
12.29
2.20
4.92
11.66
12.29
2.20
4.76
11.11
11.84
2.18
2003
4.43
10.12
11.39
2.01
4.43
10.11
11.40
2.01
4.43
10.11
11.40
2.01
4.33
9.92
10.73
2.03
2004
5.12
14.37
11.49
2.04
5.12
14.38
11.48
2.04
5.12
14.38
11.48
2.04
4.88
13.63
10.57
2.02
2005
3.94
9.68
9.91
1.67
3.95
9.67
9.97
1.68
3.94
9.67
9.95
1.68
3.85
9.04
10.09
1.66
2006
3.14
6.44
8.80
1.45
3.38
6.61
9.48
1.63
3.28
6.57
9.17
1.55
3.06
5.97
8.70
1.46
2007
2.96
6.45
8.04
1.33
3.63
8.59
9.25
1.59
3.35
7.78
8.73
1.47
2.96
6.39
7.95
1.36
2008
2.59
5.37
7.38
1.17
3.88
11.02
8.77
1.51
3.40
9.02
8.30
1.36
2.78
6.45
7.30
1.21
2009
2.00
4.08
5.15
1.02
3.06
6.90
7.31
1.50
2.49
5.39
5.93
1.27
2.60
6.16
6.88
1.10
2010
1.35
2.88
2.56
0.81
2.14
3.17
3.68
1.66
1.65
3.00
2.76
1.17
2.31
5.51
6.40
0.92
2011
1.00
2.02
1.57
0.68
1.94
2.18
2.05
1.86
1.34
2.12
1.67
1.11
1.95
4.24
5.61
0.83
2012
0.94
2.35
1.48
0.55
1.38
2.45
1.85
1.07
1.07
2.41
1.54
0.70
1.78
4.21
5.16
0.68
2013
0.76
1.69
1.30
0.47
1.08
1.75
1.59
0.85
0.85
1.71
1.34
0.59
1.55
3.47
4.60
0.61
2014
0.72
1.80
1.22
0.41
0.97
1.81
1.44
0.71
0.79
1.80
1.23
0.50
1.46
3.49
4.23
0.55
2015
0.64
1.52
1.16
0.37
0.85
1.53
1.35
0.62
0.70
1.51
1.16
0.44
1.33
3.13
3.87
0.50
2016
0.68
1.72
1.26
0.36
0.87
1.72
1.43
0.59
0.73
1.71
1.26
0.43
1.36
3.53
3.65
0.50
2017
0.73
2.17
1.18
0.35
0.89
2.16
1.34
0.54
0.77
2.16
1.18
0.40
1.38
3.73
3.60
0.49
2018
0.66
1.93
1.09
0.32
0.79
1.91
1.24
0.48
0.70
1.92
1.10
0.37
1.24
3.29
3.21
0.46
2019
0.72
2.34
1.13
0.30
0.83
2.32
1.28
0.43
0.75
2.33
1.14
0.34
1.25
3.68
2.94
0.43
2020
0.59
1.89
0.92
0.26
0.68
1.86
1.06
0.36
0.61
1.87
0.93
0.29
1.08
3.02
2.71
0.38
2021
0.51
1.44
0.90
0.25
0.59
1.43
1.03
0.33
0.53
1.42
0.91
0.27
0.93
2.43
2.40
0.36
2022
0.51
1.43
0.92
0.24
0.58
1.43
1.04
0.33
0.53
1.42
0.93
0.27
0.93
2.51
2.26
0.36
2023
0.54
1.69
0.88
0.24
0.60
1.67
0.98
0.30
0.55
1.68
0.89
0.25
0.94
2.67
2.21
0.35
2024
0.49
1.58
0.79
0.20
0.53
1.57
0.87
0.25
0.50
1.57
0.81
0.21
0.82
2.26
2.05
0.29
2025
0.43
1.29
0.74
0.19
0.46
1.29
0.80
0.23
0.44
1.28
0.76
0.20
0.73
1.98
1.82
0.28
2026
0.38
1.08
0.71
0.18
0.41
1.07
0.75
0.21
0.39
1.06
0.72
0.19
0.66
1.69
1.68
0.26
2027
0.47
1.60
0.74
0.18
0.50
1.59
0.78
0.21
0.48
1.59
0.75
0.19
0.75
2.29
1.66
0.27
2028
0.46
1.69
0.65
0.16
0.48
1.69
0.68
0.18
0.46
1.69
0.66
0.17
0.73
2.33
1.61
0.23
2029
0.39
1.34
0.63
0.15
0.41
1.33
0.65
0.17
0.40
1.33
0.62
0.16
0.62
1.83
1.44
0.22
2030
0.35
0.99
0.63
0.16
0.36
0.98
0.65
0.18
0.35
0.98
0.62
0.17
0.55
1.45
1.33
0.23
2031
0.40
1.42
0.58
0.15
0.41
1.41
0.61
0.16
0.40
1.41
0.58
0.15
0.59
1.86
1.27
0.21
2032
0.35
1.18
0.55
0.14
0.36
1.18
0.58
0.15
0.35
1.18
0.55
0.14
0.53
1.59
1.13
0.20
2033
0.34
1.14
0.53
0.13
0.35
1.13
0.56
0.14
0.34
1.13
0.53
0.13
0.49
1.47
1.04
0.18
2034
0.34
1.23
0.49
0.13
0.35
1.23
0.52
0.13
0.34
1.23
0.49
0.13
0.48
1.50
0.98
0.17
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-8
Upper Hudson River Average Largemouth Bass (Whole Fish) PCB Concentration (in mg/kg)
2035
0.29
0.88
0.47
0.14
0.28
0.87
0.50
0.12
0.28
0.87
0.48
0.12
0.41
1.12
0.87
0.18
2036
0.40
1.21
0.48
0.22
0.33
1.21
0.50
0.11
0.33
1.20
0.48
0.12
0.51
1.43
0.85
0.26
2037
0.36
0.98
0.46
0.21
0.29
0.98
0.49
0.11
0.32
0.98
0.47
0.15
0.45
1.19
0.75
0.24
2038
0.36
1.13
0.43
0.19
0.33
1.13
0.45
0.14
0.37
1.13
0.43
0.20
0.45
1.32
0.72
0.22
2039
0.33
0.89
0.42
0.19
0.34
0.89
0.44
0.21
0.34
0.89
0.42
0.21
0.41
1.09
0.68
0.22
2040
0.31
0.86
0.42
0.17
0.33
0.86
0.44
0.20
0.32
0.86
0.42
0.19
0.38
0.98
0.63
0.20
2041
0.37
1.23
0.44
0.18
0.40
1.23
0.45
0.22
0.39
1.23
0.44
0.20
0.45
1.42
0.66
0.21
2042
0.39
1.40
0.46
0.16
0.42
1.40
0.47
0.20
0.41
1.40
0.46
0.18
0.46
1.56
0.65
0.19
2043
0.33
1.10
0.39
0.15
0.35
1.10
0.40
0.18
0.34
1.10
0.40
0.16
0.39
1.22
0.62
0.17
2044
0.28
0.82
0.37
0.14
0.29
0.82
0.37
0.16
0.28
0.83
0.37
0.15
0.32
0.88
0.55
0.16
2045
0.30
0.97
0.38
0.14
0.31
0.97
0.38
0.16
0.31
1.00
0.38
0.15
0.34
1.04
0.52
0.16
2046
0.27
0.86
0.36
0.13
0.28
0.86
0.36
0.14
0.28
0.88
0.36
0.14
0.32
0.95
0.51
0.15
2047
0.28
0.93
0.37
0.13
0.29
0.91
0.37
0.14
0.29
0.93
0.37
0.14
0.35
1.17
0.49
0.15
2048
0.30
1.08
0.37
0.13
0.31
1.07
0.37
0.14
0.31
1.07
0.37
0.13
0.39
1.42
0.50
0.15
2049
0.31
1.14
0.39
0.12
0.33
1.15
0.39
0.14
0.32
1.13
0.39
0.13
0.38
1.39
0.50
0.14
2050
0.28
0.96
0.36
0.12
0.29
0.96
0.36
0.13
0.28
0.95
0.37
0.12
0.34
1.21
0.49
0.13
2051
0.27
0.96
0.36
0.10
0.28
0.96
0.36
0.11
0.27
0.96
0.37
0.11
0.32
1.12
0.47
0.12
2052
0.24
0.80
0.36
0.10
0.25
0.80
0.36
0.11
0.25
0.82
0.36
0.11
0.29
0.98
0.44
0.12
2053
0.32
1.26
0.38
0.11
0.32
1.26
0.38
0.12
0.33
1.28
0.38
0.11
0.37
1.41
0.49
0.12
2054
0.29
1.08
0.38
0.11
0.29
1.08
0.38
0.11
0.30
1.10
0.38
0.11
0.32
1.18
0.46
0.12
2055
0.26
0.93
0.36
0.10
0.26
0.93
0.36
0.11
0.27
0.95
0.36
0.10
0.30
1.06
0.44
0.11
2056
0.28
1.03
0.41
0.10
0.29
1.02
0.40
0.11
0.29
1.04
0.41
0.10
0.32
1.16
0.45
0.11
2057
0.29
1.14
0.37
0.10
0.30
1.14
0.37
0.10
0.30
1.15
0.37
0.10
0.32
1.17
0.46
0.11
2058
0.25
0.85
0.37
0.10
0.25
0.85
0.37
0.10
0.25
0.87
0.38
0.10
0.27
0.91
0.43
0.11
2059
0.31
1.27
0.36
0.10
0.31
1.26
0.36
0.10
0.31
1.28
0.36
0.10
0.33
1.31
0.46
0.10
2060
0.24
0.88
0.35
0.09
0.25
0.87
0.35
0.09
0.25
0.89
0.35
0.09
0.26
0.93
0.40
0.10
2061
0.23
0.79
0.33
0.09
0.23
0.79
0.33
0.09
0.23
0.80
0.33
0.09
0.25
0.84
0.38
0.09
2062
0.25
0.89
0.34
0.09
0.25
0.89
0.34
0.09
0.25
0.90
0.34
0.09
0.26
0.91
0.38
0.10
2063
0.24
0.89
0.35
0.09
0.25
0.89
0.34
0.09
0.25
0.89
0.35
0.09
0.26
0.91
0.37
0.10
2064
0.25
0.92
0.36
0.09
0.25
0.92
0.36
0.09
0.25
0.92
0.36
0.09
0.27
0.97
0.38
0.10
2065
0.25
0.88
0.36
0.10
0.25
0.87
0.36
0.10
0.25
0.88
0.36
0.10
0.25
0.87
0.38
0.10
2066
0.30
1.25
0.34
0.09
0.30
1.25
0.34
0.09
0.30
1.25
0.34
0.09
0.31
1.26
0.40
0.09
2067
0.26
0.95
0.37
0.09
0.26
0.95
0.37
0.09
0.26
0.95
0.37
0.09
0.27
0.97
0.37
0.10
Notes:
Fish fillets multiplied by 2.5 to obtain whole fish concentrations.
Upper Hudson River average is weighted by river section length. River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
All whole fish PCB concentrations are above target fish concentration of 0.3 mg/kg and/or 0.03 mg/kg based on the river otter lowest-observed-adverse-effects-level (LOAEL) and no-observed-adverse-effects-level (NOAEL), respectively.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-9
Modeled Year-of-Compliance for River Otter
Risk-Based Fish Concentrations
Upper Hudson River
Modeled Year of Compliance
LOAEL 0.3 PCBs
NOAEL 0.03 PCBs
River Otter - RI/FS TRVs (whole fish tissue)
mg/kg
mg/kg
Upper Hudson River Average
No Resuspension (d004)
2035
>2067
Total PCB 350 ng/L (sr04)
2035
>2067
Total PCB 600 g/day (srOl)
2035
>2067
Monitored Natural Attenuation
2052
>2067
Upper Hudson River Section 1
No Resuspension (d004)
>2067
>2067
Total PCB 350 ng/L (sr04)
>2067
>2067
Total PCB 600 g/day (srOl)
>2067
>2067
Monitored Natural Attenuation
>2067
>2067
Upper Hudson River Section 2
No Resuspension (d004)
>2067
>2067
Total PCB 350 ng/L (sr04)
>2067
>2067
Total PCB 600 g/day (srOl)
>2067
>2067
Monitored Natural Attenuation
>2067
>2067
Upper Hudson River Section 3
No Resuspension (d004)
2019
>2067
Total PCB 350 ng/L (sr04)
2024
>2067
Total PCB 600 g/day (srOl)
2020
>2067
Monitored Natural Attenuation
2024
>2067
Hudson River PCBs Superfiind Site
Engineering Performance Standards
Notes:
First year in which fish target concentrations are achieved are provided.
Upper Hudson River average is weighted by river section length. River Section 1: 6.3 miles = 15.4C
River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-10
Lower Hudson River Average Largemouth Bass (Whole Fish) PCB Concentration (in mg/kg)
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
Total PCB 600 g/day (srOl)
Monitored Natural Attenuation
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
River Mile
Year
152
113
90
50
152
113
90
50
152
113
90
50
152
113
90
50
1998
7.15
5.21
3.55
3.26
7.15
5.21
3.55
3.26
7.15
5.21
3.55
3.26
7.54
5.30
3.55
3.24
1999
4.53
4.12
3.30
3.01
4.53
4.12
3.30
3.01
4.53
4.12
3.30
3.01
4.37
4.06
3.28
2.99
2000
3.81
3.56
2.93
2.73
3.81
3.56
2.93
2.73
3.81
3.56
2.93
2.73
4.01
3.56
2.91
2.71
2001
4.50
3.54
2.66
2.49
4.50
3.54
2.66
2.49
4.50
3.54
2.66
2.49
4.51
3.54
2.65
2.47
2002
3.97
3.19
2.49
2.31
3.97
3.19
2.49
2.31
3.97
3.19
2.49
2.31
3.91
3.17
2.47
2.28
2003
3.42
2.82
2.26
2.10
3.42
2.82
2.26
2.10
3.42
2.82
2.26
2.10
3.39
2.82
2.25
2.08
2004
2.42
2.26
1.97
1.89
2.42
2.26
1.97
1.89
2.42
2.26
1.97
1.89
2.39
2.23
1.96
1.88
2005
2.27
1.95
1.69
1.67
2.27
1.95
1.69
1.67
2.27
1.95
1.69
1.67
2.25
1.94
1.68
1.66
2006
2.37
1.85
1.49
1.48
2.53
1.89
1.49
1.49
2.49
1.86
1.49
1.49
2.34
1.86
1.49
1.47
2007
1.93
1.71
1.35
1.34
2.37
1.86
1.40
1.36
2.20
1.79
1.38
1.34
1.89
1.70
1.35
1.32
2008
1.54
1.41
1.22
1.20
2.33
1.77
1.33
1.25
1.97
1.60
1.27
1.23
1.57
1.42
1.21
1.20
2009
1.21
1.15
1.06
1.05
2.03
1.53
1.18
1.12
1.62
1.34
1.12
1.08
1.27
1.16
1.06
1.05
2010
1.10
1.02
0.92
0.94
2.55
1.71
1.16
1.06
1.73
1.30
1.02
1.00
1.36
1.13
0.94
0.95
2011
1.25
1.01
0.84
0.86
5.16
2.57
1.35
1.10
2.43
1.49
1.01
0.96
1.63
1.22
0.91
0.89
2012
0.92
0.86
0.75
0.77
2.17
2.06
1.38
1.13
1.32
1.20
0.96
0.90
1.30
1.11
0.86
0.83
2013
1.02
0.82
0.68
0.71
1.78
1.63
1.28
1.11
1.27
1.08
0.88
0.84
1.48
1.13
0.83
0.79
2014
0.86
0.74
0.62
0.64
1.33
1.29
1.12
1.04
1.01
0.92
0.78
0.77
1.27
1.03
0.79
0.74
2015
0.72
0.65
0.56
0.59
1.04
1.04
0.96
0.94
0.82
0.78
0.69
0.70
1.00
0.90
0.73
0.70
2016
0.55
0.53
0.50
0.53
0.76
0.78
0.79
0.83
0.61
0.61
0.60
0.63
0.76
0.72
0.65
0.64
2017
0.46
0.45
0.44
0.48
0.54
0.60
0.65
0.73
0.51
0.51
0.51
0.56
0.68
0.62
0.57
0.59
2018
0.43
0.41
0.39
0.44
0.45
0.50
0.54
0.63
0.47
0.45
0.45
0.50
0.65
0.58
0.51
0.53
2019
0.34
0.35
0.35
0.40
0.35
0.39
0.44
0.54
0.37
0.38
0.39
0.45
0.52
0.50
0.46
0.49
2020
0.42
0.35
0.32
0.36
0.42
0.37
0.38
0.46
0.45
0.37
0.35
0.40
0.68
0.51
0.42
0.44
2021
0.41
0.34
0.30
0.34
0.41
0.36
0.34
0.41
0.44
0.36
0.32
0.36
0.63
0.49
0.40
0.41
2022
0.35
0.32
0.29
0.32
0.35
0.33
0.31
0.37
0.37
0.34
0.30
0.34
0.51
0.45
0.38
0.39
2023
0.30
0.29
0.27
0.30
0.30
0.29
0.2S
0.33
0.32
0.30
0.2S
0.32
0.46
0.41
0.35
0.37
2024
0.32
0.28
0.25
0.28
0.32
0.28
0.26
0.31
0.33
0.29
0.26
0.30
0.48
0.40
0.34
0.35
2025
0.35
0.30
0.25
0.27
0.35
0.30
0.26
0.29
0.37
0.31
0.26
0.29
0.53
0.43
0.34
0.34
2026
0.33
0.29
0.25
0.27
0.33
0.29
0.25
0.28
0.34
0.29
0.25
0.28
0.48
0.40
0.33
0.33
2027
0.26
0.26
0.24
0.26
0.26
0.26
0.24
0.27
0.26
0.27
0.24
0.27
0.37
0.36
0.32
0.32
2028
0.24
0.24
0.23
0.25
0.24
0.25
0.23
0.26
0.25
0.25
0.23
0.26
0.35
0.34
0.30
0.31
2029
0.29
0.25
0.22
0.24
0.29
0.25
0.22
0.25
0.30
0.25
0.22
0.25
0.42
0.34
0.28
0.30
2030
0.29
0.25
0.22
0.24
0.29
0.25
0.22
0.24
0.30
0.25
0.22
0.24
0.40
0.34
0.28
0.29
2031
0.25
0.24
0.21
0.23
0.25
0.24
0.21
0.23
0.25
0.24
0.22
0.24
0.34
0.32
0.27
0.28
2032
0.25
0.24
0.21
0.23
0.25
0.24
0.21
0.23
0.26
0.24
0.22
0.24
0.35
0.32
0.27
0.28
2033
0.23
0.23
0.21
0.23
0.23
0.23
0.21
0.23
0.23
0.23
0.21
0.23
0.31
0.30
0.26
0.28
2034
0.22
0.22
0.20
0.22
0.22
0.22
0.20
0.22
0.22
0.22
0.20
0.23
0.29
0.29
0.25
0.27
2035
0.35
0.25
0.21
0.22
0.35
0.25
0.21
0.22
0.27
0.23
0.20
0.22
0.42
0.33
0.25
0.26
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-10
Lower Hudson River Average Largemouth Bass (Whole Fish) PCB Concentration (in mg/kg)
2036
0.48
0.32
0.23
0.23
0.48
0.32
0.23
0.23
0.23
0.22
0.20
0.22
0.54
0.38
0.27
0.26
2037
0.57
0.39
0.26
0.24
0.57
0.39
0.26
0.24
0.40
0.28
0.21
0.22
0.69
0.46
0.30
0.28
2038
0.58
0.40
0.28
0.26
0.58
0.40
0.28
0.26
0.65
0.38
0.24
0.23
0.65
0.47
0.32
0.29
2039
0.48
0.39
0.29
0.27
0.48
0.39
0.29
0.27
0.56
0.41
0.27
0.25
0.55
0.44
0.33
0.30
2040
0.43
0.37
0.29
0.27
0.43
0.37
0.29
0.27
0.51
0.40
0.29
0.26
0.48
0.42
0.33
0.31
2041
0.30
0.32
0.28
0.27
0.30
0.32
0.28
0.27
0.35
0.35
0.28
0.27
0.35
0.36
0.31
0.30
2042
0.25
0.27
0.26
0.26
0.25
0.27
0.26
0.26
0.29
0.30
0.27
0.27
0.28
0.30
0.28
0.29
2043
0.29
0.26
0.24
0.25
0.29
0.26
0.24
0.25
0.33
0.29
0.25
0.26
0.35
0.30
0.26
0.28
2044
0.35
0.28
0.23
0.24
0.35
0.28
0.23
0.25
0.38
0.31
0.25
0.25
0.42
0.32
0.26
0.27
2045
0.33
0.28
0.23
0.24
0.33
0.28
0.23
0.24
0.34
0.30
0.24
0.25
0.38
0.32
0.26
0.26
2046
0.29
0.26
0.22
0.24
0.29
0.26
0.22
0.24
0.30
0.27
0.23
0.24
0.33
0.30
0.25
0.26
Notes:
Fish fillets multiplied by 2.5 to obtain whole fish concentrations.
All whole fish PCB concentrations are above target fish concentration of 0.3 mg/kg and/or 0.03 mg/kg based on the river otter lowest-observed-adverse-effects-level (LOAEL) and no-observed-adverse-effects-level (NOAEL), respectively.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-11
Modeled Year-of-Compliance for River Otter
Risk-Based Fish Concentrations
Lower Hudson River
River Otter - RI/FS TRVs (whole fish
tissue)
LOAEL 0.3 PCBs
mg/kg
NOAEL 0.03 PCBs
mg/kg
Lower Hudson River RM 152
2027
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2027
>2067
Total PCB 600 g/day (srOl)
2027
>2067
Monitored Natural Attenuation
2034
>2067
Lower Hudson River RM 113
2023
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2023
>2067
Total PCB 600 g/day (srOl)
2024
>2067
Monitored Natural Attenuation
2034
>2067
Lower Hudson River RM 90
2021
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2023
>2067
Total PCB 600 g/day (srOl)
2023
>2067
Monitored Natural Attenuation
2028
>2067
Lower Hudson River RM 50
2023
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2025
>2067
Total PCB 600 g/day (srOl)
2024
>2067
Monitored Natural Attenuation
2029
>2067
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
Notes:
Hudson River PCBs Superfund Site ^'rS* ^ear 'n ฎs'1 target concentrations are achieved are provided.
Engineering Performance Standards
-------
Table 2-12
Results for Average Dredging-Related Source Strength Estimated Fluxes
INPUT
TSS-Chem RESULTS
PERCENT LOSS
Net Total PCB
Net Fraction
Concentration
PCB Production
Sediment
TSS Silt Source
Net TSS Flux at
Flux at 1 mile
Dissolved PCBs
increase at 1
TSS Loss
PCB Loss
rate
production rate
Silt Fraction
Strength (1,2)
1 mile (2)
(2)
at 1 mile
mile
at 1 mile
at 1 mile
kg PCB/day
kg solids/day
unitless
(kg/s)
(kg/day)
(VdaO
unitless
1 IIU 1)
%
%
Kin or Seel ion
Section 1
57
2,099,921
u.37
u.u77
2,303
78
0.35
14
0.11
0.14
Section 2
116
1,857,493
0.48
0.088
2,642
209
0.39
37
0.14
0.18
Section 3
45
1,563,927
0.48
0.074
2,225
81
0.40
14
0.14
0.18
Notes:
1. Source strengths apply to silt and finer particles only
2. Production rates are based on 7 days/week, 14 hours per day, 630 days in Section 1 and 210 days each in River Sections 2 & 3.
3. Values are based on river-wide volumetric flow of 4000 cfs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 2-13
Resuspension Production, Release, and Export Rates from TSS-Chem and HUDTOX Models
TSS-Chem and HUDTOX Simulations
Total PCB Flux at
Fraction-
Resuspension
Production
Resuspension
Production
Silt Fraction in
Net SS Flux at
Total PCB flux at
1 mile3 from TSS-
Chem
Fraction
Dissolved total
Far-field Monitoring
Stations from
HUDTOX4
Removal Rate
Removal
Rate of
Source Strength
as Percentage of
Resuspension
Export Rate as
Percentage of
(Resuspension
Export
Rate/Resuspensio
Dredging Location and
Rate of
Rate of Total
Dredged
1 mile from SS-
(Resuspension
PCB from TSS-
(Resuspension Expon
of total PCB
Solids via
total PCB
total PCB
n Production
Scenario
Sediment Removal Period
Monitoring Station
Sediment1
PCB2
Material
Chem
Release Rate)
Chem
Rate)
via Dredging
Dredging7
Removed
Removed
Rate)
(kg/s)
(g/day)
(kg/s)
(g/day)
(g/day)
(g/day)
(kg/s)
<%)
(%)
1,700
Evaluation
Level - 300
g/day total
PCB Flux at
the Far-Field
May 1 - November 30,2007
Section 1, T1D
1.3
0.37
0.27
410
0.22
320
5.7.E+04
42
3%
0.56%
0.19
May 1 - November 30,2008
Section 1, T1D
1.1
1,500
0.37
0.24
360
0.23
300
5.7.E+04
42
3%
0.53%
0.20
Mav 1 - August 15,2009
Section 1, T1D
0.9
1,300
0.37
0.20
310
0.25
310
5.7.E+04
42
2%
0.54%
0.24
August 16 - November 30,2009
Mav 1 - August 15,2010
Section 2, Schuylerville
Section 2, Schuvlerville
0.3
0.3
1,100
900
0.48
0.48
0.10
0.08
360
310
0.35
0.37
330
300
1.2.E+05
1.2.E+05
37
37
1%
0.29%
0.26%
0.30
0.33
Stations
August 16 - November 30,2010
Section 3, Waterford
0.9
1,300
0.48
0.25
400
0.25
340
4.5.E+04
31
3%
0.75%
0.26
Mav 1 - August 15,2011
Section 3, Waterford
0.7
1,000
0.48
0.19
310
0.28
340
4.5.E+04
31
2%
0.75%
0.34
Control Level
- 600 g/day
total PCB
Flux at the
Far-Field
May 1 - November 30,2006
Section 1, T1D
2.6
3,600
0.37
0.57
820
0.15
620
5.7.E+04
42
6%
1.1%
0.17
May 1 - November 30,2007
Section 1, T1D
2.6
3,600
0.37
0.57
820
0.15
630
5.7.E+04
42
6%
1.1%
0.18
May 1 - November 30,2008
Section 1, T1D
2.3
3,100
0.37
0.50
720
0.16
620
5.7.E+04
42
6%
1.1%
0.20
Mav 1 - August 15,2009
Section 1, T1D
2.0
2,700
0.37
0.43
620
0.18
590
5.7.E+04
42
5%
1.0%
0.22
August 16 - November 30,2009
Mav 1 - August 15,2010
Section 2, Schuylerville
Section 2, Schuvlerville
0.7
0.6
2,300
1,900
0.48
0.48
0.21
0.17
730
630
0.29
0.30
620
590
1.2.E+05
1.2.E+05
37
37
2%
2%
0.5%
0.5%
0.27
0.31
Stations
August 16 - November 30,2010
Section 3, Waterford
1.9
2,700
0.48
0.52
810
0.17
660
4.5.E+04
31
6%
1.5%
0.24
Mav 1 - August 15,2011
Section 3, Waterford
1.4
2,100
0.48
0.40
630
0.20
650
4.5.E+04
31
5%
1.4%
0.31
Control Level
May 1 - November 30,2006
Section 1, T1D
5.6
7,600
0.37
1.2
1,700
0.09
1,200
5.7.E+04
42
13%
2.1%
0.16
- 350 ng/L
May 1 - November 30,2007
Section 1, T1D
5.6
7,600
0.37
1.2
1,700
0.09
1,200
5.7.E+04
42
13%
2.1%
0.16
total PCB
May 1 - November 30,2008
Section 1, T1D
4.9
6,700
0.37
1.1
1,500
0.10
1,300
5.7.E+04
42
12%
2.3%
0.19
Coneentratio
Mav 1 - August 15,2009
Section 1, T1D
4.2
5,700
0.37
0.91
1,300
0.11
1,200
5.7.E+04
42
10%
2.1%
0.21
ns at the Far
August 16 - November 30,2009
Section 2, Schuylerville
2.7
8,300
0.48
0.75
2,500
0.14
2,000
1.2.E+05
37
7%
1.7%
0.24
Field
Mav 1 - August 15,2010
Section 2, Schuvlerville
2.3
7,100
0.48
0.64
2,100
0.16
2,000
1.2.E+05
37
6%
1.7%
0.28
Monitoring
August 16 - November 30,2010
Section 3, Waterford
7.5
10,900
0.48
2.1
3,100
0.06
2,200
4.5.E+04
31
24%
4.9%
0.20
Stations
May 1 - August 15,2011
Section 3, Waterford
5.8
8,400
0.48
1.6
2,400
0.07
2,300
4.5.E+04
31
19%
5.1%
0.27
TSS-Chem Simulations Only
Scenario
Sediment Removal Period
Dredging Location and
Monitoring Station
Resuspension
Production
Rate of
Sediment1
(kg/s)
Resuspension
Production
Rate of Total
PCB2
(g/day)
Silt Fraction in
Dredged
Material
Net SS Flux at
1 mile from SS-
Chem
(kg/s)
Total PCB flux at
1 mile3 from TSS-
Chem
(Resuspension
Release Rate)
(g/day)
Fraction
Dissolved total
PCB from TSS-
Chem
Total PCB Flux at
Monitoring Stations10
(Resuspension Export
Rate)
(g/day)
Removal Rate
of total PCB
via Dredging
(g/day)
Removal
Rate of
Solids via
Dredging7
(kg/s)
Source Strength
as Percentage of
total PCB
Removed
<%)
Resuspension
Export Rate as
Percentage of
total PCB
Removed
(%)
Total PCB Export
Fraction-
(Resuspension
Export
Rate/Resuspensio
n Production
Rate)
Kesuspensio
n Standard -
500 ng/L
total PCB
Coneentratio
ns at the Far
Field
Monitoring
May 1 - November 30,2006
May 1 - November 30,2007
May 1 - November 30,2008
Mav 1 - August 15,2009
Section 1, T1D
Section 1, T1D
Section 1, T1D
Section 1, T1D
9.4
9.3
8.2
7.1
12,800
12,700
11,200
9,600
0.37
0.37
0.37
0.37
2.0
2.0
1.53
2,800
2,800
2,500
2,100
0.06
0.06
0.06
0.07
2,100
2,100
2,100
2,100
5.7.E+04
5.7.E+04
5.7.E+04
5.7.E+04
42
42
42
42
23%
22%
20%
17%
3.7%
3.7%
3.7%
3.7%
0.16
0.17
0.19
0.22
August 16 - November 30,2009
Mav 1 - August 15,2010
Section 2, Schuylerville
Section 2, Schuvlerville
3.5
3.0
10,900
9,300
0.48
0.48
0.99
0.84
3,200
2,800
0.12
0.13
2,700
2,700
1.2.E+05
1.2.E+05
37
37
9%
2.3%
2.3%
0.25
0.29
August 16 - November 30,2010
May 1 - August 15,2011
Section 3, Waterford
Section 3, Waterford
11
16,600
12,800
0.48
0.48
3.2
2.5
4,800
3,700
0.04
0.05
3,500
3,500
4.5.E+04
4.5.E+04
31
31
37%
28%
7.7%
7.7%
0.21
0.27
Hudson River PCBsS
le 2 Dredging Resuspenacฎ - April 2004
-------
Table 2-14
Increase in PCB Mass from Settled Material 2-Acres Below the Target Area
Estimated Using the TSS-Chem Model Results
Management
Condition at Far Field Station
River
Total PCBs Length-
Level
Section
Weighted Average
Concentration (mg/kg) (0-6
inches)
Evaluation
300 g/day PCB Mass Loss
1
2.6
Control
600 g/day PCB Mass Loss
1
4.2
Control
350 ng/L
1
6.6
Evaluation
300 g/day PCB Mass Loss
2
2.0
Control
600 g/day PCB Mass Loss
2
3.3
Control
350 ng/L
2
9.1
Evaluation
300 g/day PCB Mass Loss
3
2.2
Control
600 g/day PCB Mass Loss
3
3.5
Control
350 ng/L
3
8.6
1. Mass/Area used to define the lateral extent of dredging in River Sections 1 and 2 is
approximately 6.6 g/sq. m and 34 g/sq. m, respectively. In River Section 3, a
mass/area was not used to select the areas in this way.
2. The length weighted average concentration was calculated assuming the
concentration below the deposited PCBs is 1 mg/kg Total PCBs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 3-1
Upper 95th Percentile Estimates of Total PCB Concentrations at TI Dam and
Schuylerville Under Baseline Conditions
Units: ng/L
TTD-West
Mav
.Tune
Julv
Aimust
Sept.
Oct. &
Nov.
Prediction
interval
368
212
149
1 19
297
TID-PRW2
May&June
Low Flow
(<5000 cfs)
May&June
High Flow
(>5000 cfs)
July and
August
Sept.
Oct.
Nov.
Prediction
interval
161
68
106
70
92
55
Schuylerville
May and
June
July
August
Sept.
Oct.
Nov.
Prediction
interval
1"^
99
107
85
1 18
107
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 3-2
Summary of Sampling Frequency Requirements and Expected Error Rates
Analysis Transition
Detail
Sampling Time Period
Action Level
Number of Samples1
Grey Region
Limit
False
Rejection
Error Limit -
a (%)
False
Acceptance
Error Limit -
P(%)
Figure
Number
Total PCB Sampling Requirements (25% CV)
Far Field
Routine to Evaluation Level
Routine to Control Level
Confirmation of the Control Level
Routine to Control Level
Confirmation of the Control Level
Routine to > 300 g/day
Routine to >600 g/day
Confirmation of > 600 g/day
Routine to > 350 ng/L
Confirmation of > 350 ng/L
1 week
1 week
1 week routine + 1 week
1 week
1 week routine + 1 week
300 g/day
600 g/day
600 g/day
350 ng/L
350 ng/L
7 (1 sample/day for 1 week)
7 (1 sample/day for 1 week)
28 (7 samples routine + 21 samples
control level)
7 (1 sample/day for 1 week)
28 (7 samples routine + 21 samples
control level)
400 g/day
700 g/day
700 g/day
400 ng/L
400 ng/L
7.5
25
5
27.5
10
5
15
4
20
5
1
2
3
4
5
Evaluation to Control Level
300 g/day to > 600 g/day
1 week evaluation + 1 week
600 g/day
35 (14 samples evaluation level + 21
samples control level)
700 g/day
4
2
6
Resuspension Standard Threshold
Confirmation of > 500 ng/L2
1 day routine + 1 day
500 ng/L
5 (1 sample routine + 4 samples
confirmation)
400 ng/L
15
30
7
Confirmation of > 500 ng/L (24 hours)2
1 day
500 ng/L
4 composites of 6 aliquots each
400 ng/L
5
7
8
Routine to Control Level
Continuous Total PCB 1-week or 2-week
deployment
1 week or 2 weeks
350 ng/L
2 composites of 56 aliquots each
400 ng/L
6.5
5
9
Suspended Solids Sampling Requirements (75% CV)
Far Field
Routine to Evaluation Level
Far-field - Baseline to > 12 mg/L
1 day (3 hrs for 24 hrs)
1 day (15 min for 24 hrs)
14 mg/L
14 mg/L
8 (discrete)
96 (continuous)
21 mg/L
21 mg/L
27.5
0.1
12.5
0.1
10
11
Routine to Control Level
Far-field - Baseline to > 24 mg/L
1 day (3 hrs for 24 hrs)
1 day (15 min for 24 hrs)
26 mg/L
26 mg/L
8 (discrete)
96 (continuous)
39 mg/L
39 mg/L
27.5
0.1
12.5
0.1
12
13
Evaluation to Control Level
Far-field -12 mg/L to > 24 mg/L
1 day evaluation + 1 day
1 day evaluation + 1 day
26 mg/L
26 mg/L
16 (discrete)
192 (continuous)
39 mg/L
39 mg/L
15
0.5
5
<0.5
14
15
Near Field
Routine to Control Level
Near Field - River Sections 1 and 3
Baseline to > 100 mg/L
6 hours (1 sample per 3 hours)
6 hours (1 sample per 15 min)
100 mg/L
100 mg/L
3 (discrete)
24 (continuous)
150 mg/L
150 mg/L
35
6.6
25
5
16
17
Routine to Control Level
Near Field - River Section 2
Baseline to > 60 mg/L
6 hours (1 sample per 3 hours)
6 hours (1 sample per 15 min)
60 mg/L
60 mg/L
3 (discrete)
24 (continuous)
90 mg/L
90 mg/L
35
6.6
25
5
18
19
Evaluation to Control Level
Near Field - River Sections 1 and 3
Baseline to > 100 mg/L
1 day (3 hrs for 15 hrs)
1 day (15 min for 15 hrs)
100 mg/L
100 mg/L
5 (discrete)
60 (continuous)
150 mg/L
150 mg/L
27.5
0.7
20
0.5
20
21
Evaluation to Control Level
Near Field - River Section 2
Baseline to > 60 mg/L
1 day (3 hrs for 15 hrs)
1 day (15 min for 15 hrs)
60 mg/L
60 mg/L
5 (discrete)
60 (continuous)
90 mg/L
90 mg/L
27.5
0.7
20
0.5
22
23
Routine to Evaluation Level
Near Field
Baseline to > 700 mg/L
3 hours (1 sample per 3 hours)
3 hours (1 sample per 5 min)
700 mg/L
700 mg/L
2 (discrete)
36 (continuous)
1000 mg/L
1000 mg/L
40
16.5
30
5
24
25
Note
1 Sampling frequency at the different action level can be found in Table 1-2 of Volume 1 of the document
2 Null hypothesis for the 500 ng/L assumed that river conditions were not in compliance, for all other action levels, the null hypohesis assumed that river conditions were in compliance. See text for discussions.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension-April 2004
-------
Table 3-3
Summary of Sampling Frequency Requirements and Expected Error Rates for Automatic Sampler
An alysis T ransition
Detail
Sampling Time Period
Action Level
Number of Samples
Grey Region
Limit
False
Rejection
Error Limit -
a (%)
False
Acceptance
Error Limit -
P(%)
Figure
Number
Total PCB Sampling Requirements (25% CV)
Far Field
Routine to Evaluation Level
Routine to > 300 g/day
1 week
300 g/day
7 composites of 24 aliquots each (1
sample/day for 1 week)
400 g/day
0.1
<0.1
29
Routine to Control Level
Routine to > 600 g/day
1 week
600 g/day
7 composites of 24 aliquots each (1
sample/day for 1 week)
700 g/day
0.5
0.1
30
Confirmation of the Control Level
Confirmation of > 600 g/day
1 week routine + 3 day
600 g/day
10 (7 samples routine + 3 samples
control level)
700 g/day
0.5
<0.5
31
Routine to Control Level
Routine to > 350 ng/L
1 week
350 ng/L
7 composites of 24 aliquots each (1
sample/day for 1 week)
400 ng/L
1
1
32
Confirmation of the Control Level
Confirmation of > 350 ng/L
1 week routine + 3 day
350 ng/L
10 (7 samples routine + 3 samples
control level)
400 ng/L
0.5
<0.5
33
Evaluation to Control Level
300 g/day to > 600 g/day
2 day evaluation + 3 day
600 g/day
5 (composite sampling every 1 hour, 1
sample/day)
700 g/day
2
1
34
Hudson River PCBs Superfiind Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension-April 2004
-------
Table 4-1
Estimated 7-Day Total PCB Concentrations1 Corresponding to the Evaluation Level
(300 g/day) at the Schuylerville Monitoring Station
Total PCB (ng/L)- Schuylerville Station2
Flow (cfs)
Flow (nrrVs)
Total PCB increase
(ng/L)
May &
June
July
August
Sept.
Oct.
Nov.
95% UCL Baseline Total PCB Concentration
121
103
81
60
84
75
2,000
57
105
226
208
186
165
189
180
2,500
71
84
205
187
165
144
168
159
3,000
85
70
191
173
151
130
154
145
3,500
99
60
181
163
141
120
144
135
4,000
113
53
174
155
133
113
136
128
4,500
127
47
168
149
127
107
131
122
5,000
142
42
163
145
123
102
126
117
5,500
156
38
160
141
119
98
122
113
6,000
170
35
156
138
116
95
119
110
6,500
184
32
154
135
113
92
116
108
7,000
198
30
151
133
111
90
114
105
7,500
212
28
149
131
109
88
112
103
8,000
227
26
148
129
107
86
110
101
8,500
241
25
146
127
105
85
109
100
9,000
255
23
145
126
104
83
107
99
9,500
269
22
143
125
103
82
106
97
10,000
283
21
142
124
102
81
105
96
Notes:
1. Total PCB concentrations are estimated based on the assumption of a 7-day per week
operation, 14 hours per day for May to November (210 days). This is conservative since
operating less than 7-days per week would increase the daily allowable Total PCB load.
These values will be adjusted to reflect the planned period of operation once it is defined
as part of the remedial design.
2. Shaded areas are the concentration at the mean flow for the month, based on flow
estimates derived from the USGS flow data (1977-present).
3. Condition for June.
4. Condition for May.
5. The values provided in this table are based on historical data. These values will be
revised prior to Phase 1 when baseline monitoring data are available and more is known
about the operating schedule and production rate.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 4-2
Estimated 7-Day Total PCB Concentrations1 Corresponding to the Control Level
(600 g/day) at the Schuylerville Monitoring Station
Total PCB (ng/L) - Schuylerville Station2
Flow (cfs)
Flow (nrrVs)
TPCB increase (ng/L)
May &
June
July
August
Sept.
Oct.
Nov.
95% UCL Baseline Total PCB Concentration
121
103
81
60
84
75
2,000
57
210
331
313
291
270
294
285
2,500
71
168
289
271
249
228
252
243
3,000
85
140
261
243
221
200
224
215
3,500
99
120
241
223
201
180
204
195
4,000
113
105
226
208
186
165
189
180
4,500
127
93
215
196
174
154
177
169
5,000
142
84
205ฐ
187
165
144
168
159
5,500
156
76
198
179
157
137
160
152
6,000
170
70
191
173
151
130
154
145
6,500
184
65
186
167
145
125
149
140
7,000
198
60
181
163
141
120
144
135
7,500
212
56
177
159
137
116
140
131
8,000
227
53
174
155
133
113
136
128
8,500
241
49
171
152
130
110
133
125
9,000
255
47
168
149
127
107
131
122
9,500
269
44
166
147
125
104
128
119
10,000
283
42
163
145
123
102
126
117
Notes:
1. Total PCB concentrations are estimated based on the assumption of a 7-day per week
operation, 14 hours per day for May to November (210 days). This is conservative since
operating less than 7-days per week would increase the daily allowable PCB load. These
values will be adjusted to reflect the planned period of operation once it is defined as part
of the remedia 1 design.
2. Shaded areas are the concentration at the mean flow for the month, based on flow
estimates derived from the USGS flow data (1977-present).
3. Condition for June.
4. Condition for May.
5. The values provided in this table are based on historic al data. These values will be
revised prior to Phase 1 when baseline monitoring data are available and more is known
about the operating schedule and production rate.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 4-3
Estimates of Baseline Concentrations at IT Dam, Schuylerville and Waterford1
Preliminary estimate of 95% UCL (Q,/) for use in the equations presented in Section 4.11.
Station
Total PCB Concentrations (ng/L)
May
June
July
August
September
October
November
TID
181
205
151
106
83
241
241
West2
TID
1113
1113
71
71
50
64
45
PRW22
47 4
47 4
Schuyler
ville
121
121
103
81
60
84
75
Waterfor
d5
90
90
76
60
44
62
56
Notes:
1 These values will be revised using the data collected during the baseline monitoring program. Similar
values will be determined for Stillwater and Waterford from the baseline monitoring as well.
2 The actual TID values are expected to fall between those obtain for TID West and TID PRW2.
3 For flow < 5000 cfs.
4 For flow > 5000 cfs.
5 These values were estimated by multiplying the Schuylerville Total PCB concentrations by a dilution
factor of 0.74 to account for additional tributary flow to Waterford.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 4-4
Far-Field Monitoring - Analytical Details
Parameter
Analytical Method/
Instrument
Detection Limit Goal
Method Range
Accuracy
Precision
Sample Si/e
I lolding l ime
Sample Container
Preservation
Congener-specific
PCBs (Total)
Green Bay or
equivalent
0.05 ng/L/congener
Lab-specific and
congener-specific
60-150%
40% RPD1
1 Liters
5/402 days
1 Liter amber glass
Maintain at 4ฐ C(ฑ 2ฐ
C)
Congener-specific
PCBs (Water)
Green Bay or
equivalent
0.05 ng/L/congener
Lab-specific and
congener-specific
60-150%
40% RPD
20 Liters
5/402 days
4 Liter amber glass
Maintain at 4ฐ C(ฑ 2ฐ
C)
Congener-specific
PCBs (Particle)
Green Bay or
equivalent
1 ng/kg
Lab-specific and
congener-specific
60-150%
40% RPD
200-800 mg
5/402 days
Amber glass
Maintain at 4ฐ C (ฑ 2ฐ
C)
DOC (TOC on filtered
water)
Persulfate Digestion
(415.2)
0.025 mg/L
50 (xg/L to 10 mg/L
90-110%
20% RPD
2 x 40 mL (25 mL
minimum)
28 days
VOA vial
Maintain at 4ฐC
H2S04 pH =2
TSS
ASTMD 3977-97
0.5 mg/L (on 1 L
sample)
0.5 to 2000 mg/L on 1
L sample
90-110%
20% RPD
1 Liter
7 days
4 Liter plastic
Maintain at 4ฐ C(ฑ 2ฐ
C)
TSS (using particle
counter)
LISST Series
TBD
1.2 to 250 |_im
TBD
TBD
25-50 mL
Field
Per instrument
requirement
NA
TSS (fast turnaround)
Modified
1.0 mg/L (on 1 L
sample)
0.5 to 2000 mg/L on 1
L sample
80 - 120 %
35% RPD
1 Liter
N/A
1 Liter plastic
None
Turbidity
YSI 6-Series
2 NTU
Oto 1000 NTU
ฑ 5% or 3 NTU3
5%
25-50 mL
Field
Per instrument
requirement
NA
Temperature
YSI 6-Series
0.15ฐC
-5 to +45 ฐC
ฑ0.15ฐC
ฑ0.15ฐC
25-50 mL
Field
Per instrument
requirement
NA
PH
YSI 6-Series
0.2 pH unit
0 to 14 pH units
ฑ 0.2 pH unit
ฑ0.2 pHunit
25-50 mL
Field
Per instrument
requirement
NA
Dissolved Oxygen
YSI 6-Series
0.2 mg/L
0 to 50 mg/L
0-20 mg/L: ฑ2% or 0.2
mg/L3
15%
25-50 mL
Field
Per instrument
requirement
NA
Conductivity
YSI 6-Series
0.001 mS/cm
0 to 100 mS/cm
ฑ0.5% or 0.001
mS/cm3
10%
25-50 mL
Field
Per instrument
requirement
NA
TOC on SS - routine
EPA 160.4
Volatile solids on SS as
surrogate for TOC.
0.5% dry wt based on
SS
ฑ 0.3 mg assuming 0.1
mg sensitivity
ฑ 10%) orฑ 0. 2 mg
ฑ 0.4mg or 10%
100 mg solids based on
0.1 mg sensitivity
Lab
Glass only
NA
TOC for SS - periodic
confirm
L Kahn - EPA Region
II
0.5 % dry wt basis on
SS
100 mg/kg
80 - 120%
RSD <10 percent on
quadruplicate
20 g filtered matter at
0.5%
Lab
Glass only
NA
Notes:
1 RPD = Relative Percent Difference; RPD criteria applicable only where sample concentrations = 5 x the sample reporting limit.
2 Holding times for extraction/analysis from time/date of sample collection.
3 Whichever is greater
NA Not applicable
TBD To Be determined
TOC Total Organic Carbon
ICP Inductively Coupled Plasma - atomic emission spectrometry
CV Cold Vapor atomic absorption
SS Suspended solids (i.e., particulate matter on filter)
mS milli-siemen
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension-April 2004
-------
Table 4-5
Near-Field Monitoring - Analytical Details
Parameter
Method
Analytical/Direct
Reading
Detection
Limit
Range
Accuracy
Precision
Sample
Size
Holding
Time
Sample
Container
Preser-
vative
Turbidity Continuous
YSI 6-Series
2 NTU
0-1000 NTU
+/- 5% or 3 NTU
5%
NA
Field
NA
NA
TSS using particle Discrete
counter
LISST Series
TBD
1.2-250 um
TBD
TBD
25-50 mL
Field
NA
NA
TSS Laboratory Discrete
ASTM D3977-97
0.01 mg/L
20%
LCS 90-110%
NA
TBD
7 days
plastic
bottle
4 liter
TSS (fast turnaround)
Modified
1.0 mg/L (on 1 L
sample)
0.5 to 2000
mg/L on 1
L sample
80- 120 %
35% RPD
1 Liter
N/A
1 Liter
plastic
None
1 liter
Dissolved Oxygen Discrete
YSI 6-Series
TBD
0 to 500% air
saturation
0-200 % : ฑ2% air sat. or ฑ2%
of reading, whichever is
greater; 200-500%
0.1% air saturation or
1% selectable
NA
Field
NA
NA
Conductivity Discrete
YSI 6-Series
0.001
mS/cm
0 to 100
mS/cm
ฑ 0.5% or 0.001 mS/cm3
0.1
25-50 mL
Field
NA
NA
Temperature Discrete
YSI 6-Series
0.15o C
-5 to +45 oC
ฑ0.15oC
ฑ0.15oC
25-50 mL
Field
NA
NA
Notes:
1. Analytical Method ASTM D3 977-97 Standard test method for determining sediment concentration in water samples.
2. TBD - to be determined
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 4-6
Possible Study Areas for Nature of Release of PCB
Sediment Type
Sediment Type
(ASTM Method
Mean Tri+ PCB
Recommended
(Side Scan
D422
Concentrations1
Study Area
Sonar)
Classification)
(mg/kg)
1
IV
CL, SI, FS, MS
10
2
IV
FS, MS
30
3
II
MS
11
4
IV
FS
15
5
IV
CL, SI, FS, MS
39
6
I
SI, FS
15
7
II
FS, MS
14
8
I
SI, FS, MS
8
9
II
FS
13
10
I
CL, SI, FS
14
11
I
FS
12
12
I
CL, SI, FS
15
13
I
CL, SI, FS
28
Note:
1 Mean Tri+ concentrations are based on the length weighted averages of the
entire core at location. Concentration represents the mean of draft dredge
areas. Note that the draft dredge area boundaries have not yet been approved
by the USEPA.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 4-7
Recommended Study Areas for Nature of Release of PCB
Sediment Type
Tri+ PCB Entire
Sediment Type
(ASTM Method
Core LWA
Recommended
(Side Scan
D422
Concentrations
Study Area
Sonar)
Classification)
(mg/kg)
1
IV
CL, SI, FS, MS
10
2
IV
FS, MS
30
3
II
MS
11
6
I
SI, FS
15
10
I
CL, SI, FS
14
Note:
1 Mean Tri+ concentrations are based on the length weighted averages of the
entire core at location. Concentration represents the mean of draft dredge
areas. Note that the draft dredge area boundaries have not yet been approved
by the USEPA.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Table 4-8
Resuspension Criteria (alternate)1
Parameter
Resuspension Standard
Threshold
Control Level
Evaluation Level
Limit
Duration
Limit
Duration
Limit | Duration
Far-Field PCB
Concentration
Total PCBs
500 ng/L
Confirmed
Occurrence
350 ng/L
4-day running average
(composite sampling every 1
hour, 1 sample/day)
Far-Field Net PCB Load3
Total PCBs
65 kg/year4
Dredging Season
Total PCBs
600 g/day
3-day running average
(composite sampling every 1
hour, 1 sample/day)
300 g/day
2-day running average
(composite sampling every 1
hour, 1 sample/day)
Tri+ PCBs
200 g/day
100 g/day
Far-Field Net Suspended
Solids Concentration5'6
All Sections
24 mg/L
Daily dredging period
(> 6 hrs.)
OR
24 hrs. on average
12 mg/L
6-hour running average net
increase
OR
average net increase in the daily
dredging period if the dredging
period is less than 6 hrs.
Near-Field (300 m) Net
Suspended Solids
Concentration7
Sections 1 & 3
100 mg/L
Daily dredging period
(> 6 hrs.)
OR
24 hrs. on average
100 mg/L
6-hour running average net
increase
OR
average net increase in the daily
dredging period if the dredging
period is less than 6 hrs.
Sections 2
60 mg/L
60 mg/L
Near-Field (100 m and
Channel-Side) Net
Suspended Solids
Concentration7
All Sections
700 mg/L
3 continuous hrs. running
average.
Notes:
1. Implemention of the criteria is described in Section 3.
2. Engineering contingencies for the Control Level will include temporary cessation of the operation.
3. Net increases in PCB load or suspended solids concentration refers to dredging related releases over baseline as defined in the text.
4. During Phase 1, half of the anticipated average production rate will be achieved. As a result, the total allowable export for Phase 1 is half of the fullscale value of 130
kg/year for a total of 650 kg for the entire program. This is equivalent to the 600 g/day Total PCB release at the target productivity schedule, during the dredging season from
5. The increased far-field monitoring required for exceedance of suspended solids criteria must include a sample timed so as to capture the suspended solids plume's arrival at the far-field station.
6. The monitoring requirements for exceedance of the suspended solids action levels are increased frequency sampling at the nearest far field station. The increased frequency
at this station will be the same as the frequency required for the PCB action levels.
7. All remedial operations will be monitored in the near-field during Phase 1, including backfilling.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figures
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Resuspension - April 2004
-------
Figure 1-1
Schematic of Near-field Monitoring Station Locations
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension-April 2004
-------
Bakers Falls
Inset of Lower Hudson
ฆ
Mohawk!
Rogers Island
'Waterford
Thompson Island Dam
Schuylerville
Poughkeepsie
Stillwater
15 Miles
LEGEND
Far-field station
Dam and Lock
River Mile Marker
Shoreline at 8,471 cfs
Major Road
County Limits
Waterford
Mohawk
Hudson
River
Scale in Miles
Far-field Water Column Monitoring Stations
Scale in Kilometers
TAMS Consultants, Inc.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-1
Comparison Between Upper Hudson River Remediation Scenario Forecasts for Thompson Island Dam
45 i
MNA (p3nas2)
No-Resuspension (d004)
Total PCB 300 g/day (sr02)
Total PCB 600 g/day (srOl)
Total PCB 350 ng/L(sr04)
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-2
Comparison Between Upper Hudson River Remediation Scenario Forecasts for Schuylerville
MNA (p3nas2)
No-Resuspension (d004)
Total PCB 300 g/day (sr02)
Total PCB 600 g/day (srOl)
Total PCB 350 ng/L(sr04)
Hudson River PCB s Superfund Site
Engineering Perfonnance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-3
Comparison Between Upper Hudson River Remediation Scenario Forecasts for Waterford
MNA (p3nas2)
No-Resuspension (d004)
Total PCB 300 g/day (sr02)
Total PCB 600 g/day (srOl)
Total PCB 350 ng/L(sr04)
Hudson River PCB s Superfund Site
Engineering Perfonnance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-4
Cumulative PCB Loads at Waterford
1800
1600
1400
'3d
* 1200
a
es
O
-J
CO
u
a.
+
H
Dredging
Period
Tri+ PCB Cumulative Load at Waterford
1000
800
600
400
200
0
^
MNA (p3nas2)
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
Total PCB 600g/day (srOl)
Total PCB 300 g/day (sr02)
Accidental Release (srAl)
1
1
2005
2015
2025
2035
2045
2055
2065
2075
Year
Dredging
OX)
e3
o
-1
CO
U
Ph
o
H
3500
3000
2500
2000
1500
1000
500
Total PCB Cumulative Load at Waterford
NV
2005
2025
2045
Year
2065
2085
'MNA (P3NAS2)
No Resuspension (d004)
Total PCB 350 ng/L
fraction remaining
.adjusted (sr04)
Total PCB 600g/day
(srOl)
"Total PCB 300 g/day
(sr02)
Accidental Release (srAl)
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-5
HUDTOX Forecasts of Whole Water, Particulate, and Dissolved Total PCB Concentrations
for Evaluation Level - 300 g/day Scenario
Thompson Island Dam
15 per. Mov. Avg. (Whole Water)
15 per. Mov. Avg. (Particulate)
15 per. Mov. Avg. (Dissolved)
Note:
Lines represent 15 day moving
averages.
Date
Schuylerville
Date
Waterford
180
Date
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-6
HDUTOX Forecasts for Whole Water, Particulate and Dissolved Total PCB Concentration
for Control Level - 600 g/day Scenario (srOl)
Date
V oil mrlainrilla
Date
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-7
HUDTOX forecasts for Whole Water, Particulate, and Dissolved Total PCB
Concentrations for ControlLevel 350 ng/L Scenario (sr04)
Thompson Island Dam
600
^o
o
^o
o
^o
o
o
o
o
o
00
o
00
o
00
o
00
o
OS
o
-------
Figure 2-8
MNA (p3nas2)
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
Total PCB 600 g/day (srOl)
Composite Fish - River Section 2 (RM 184)
6.0
Composite Fish - River Section 1 (RM 189)
6.0
0.0
2005 2010 2015 2020 2025
Year
ox
a
~Sd 4.0
ฃ
35
U
-
+
0.0
2005 2010 2015 2020 2025
Year
Note:
Fish composite is 47% largemouthbass + 44% brown bullhead + 9% yellow perch
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Page 1 of 2 Volume 2: Dredging Resuspension - April 2004
-------
ox
a
~03D
ฃ
35
U
-
6.0
5.0
4.0
3.0
2.0
1.0
0.0
Figure 2-8 (Cont.)
Composite Fish Tissue Concentrations for the Upper River
Composite Fish - River Section 3 (RM 154)
2005
MNA (p3nas2)
-No Resuspension (d004)
-Total PCB 350 ng/L (sr04)
Total PCB 600 g/day (srOl)
2010
2015
2020
2025
Year
Notes:
Fish composite is 47% largemouthbass + 44% brown bullhead + 9% yellow perch
The bottom figure is portion of the top figure.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Page 2 of 2 Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-9
Composite Fish Tissue Concentrations for the Lower River
Composite Fish - RM 152
Note:
Fish composite is 47% largemouthbass + 44% brown bullhead + 9% yellow perch
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Page 1 of 2 Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-9 (Cont.)
Composite Fish Tissue Concentrations for the Lower River
1.6
1.4 ฆ
1.2
"St
S 1.0
m
v
a.
H
a 0.6
Composite Fish - RM 90
0.8
2005
MNA (p3nas2)
-No Resuspension (d004)
-Total PCB 350 ng/L (sr04)
Total PCB 600 g/day (srOl)
2010
2015
2020
2025
Year
2030
2035
2040
2045
Composite Fish - RM 50
Year
Note:
Fish composite is 47% largemouthbass + 44% brown bullhead + 9% yellow perch
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Page 2 of 2 Volume 2: Dredging Resuspension - April 2004
-------
Figure 2-10
Estimated Total PCB Concentrations at Waterford for the Accidental Release Scenario
Date
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
600
30
ฆEquilibrium (Dissolved)
-Dissolved Concentration
Equilibrium (Particulate)
Particulate Concentration
0
11000
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Distance Downstream (m)
Figure 2-11
PCB Concentrations Downstream of Dredge for 350 ng/L Scenario
Section 1 at 1 mile and 3 miles
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Figure 3-1
Examination of Analytical Precision Based on Blind Duplicates
Data from the Schuylerville Station,
GE Data
Distributions
RPD (%)
"i 1 r
10 15 20
Count
Quantiles
100.0%
maximum
130.64
99.5%
130.64
97.5%
102.24
90.0%
21.82
75.0%
quartile
15.51
50.0%
median
8.12
25.0%
quartile
2.89
10.0%
1.27
2.5%
0.36
0.5%
0.21
0.0%
minimum
0.21
Moments
Mean
Std Dev
Std Err Mean
upper 95% Mean
lower 95% Mean
N
12.7
19.2
2.6
17.9
7.4
54
Values represent relative percent difference calculated between blind duplicate pairs. See
text for additional discussion.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirme fTAMS-Earth Tech
Volume 2: Dredging Resuspension-April 2004
-------
Collect water column samples at far-
field monitoring stations per the
monitoring plan for routine sampling
Figure 4-2
Simplified Flow Chart for Far-field PCB
Temporary halting of all
operations in the river if
Total PCB concentration
levels in excess of 500
ng/L are confirmed by
next day' s samples.
Restart requires
engineering evaluation
and USEPA approval
and routine monitoring
will be resumed.
No
Continue routine monitoring
Conduct Evaluation Level non-
routine monitoring per Monitoring
Plan i
Evaluate and identify any problems.
Examine boat traffic patterns near the
dredges. Examine sediment transfer
pipelines for leaks. Recommended
engineering evaluations near the dredge
and barges. Other engineering
evaluation recommended as well.
Recommend PCB sample collection in
the near-field or other areas of the
operation as a part of an engineering
study.
Note:
1. Non-routine monitoring will be required continuously for the period of time as specified in Section 3.3.5.
Mandatory engineering evaluation and
continual adjustments to dredging operations
until Concern Level or better is attained.
Evaluate and identify any problem. Consider
change in silt barriers or dredge type.
Consider implementing silt barriers, if not
already in use. Consider changing location
and rescheduling more highly contaminated
areas for later in the year (applies to May and
June only), if all other options are not ^
effective. Temporary cessation of operation
may be required.
Hudson River PCBs Superfund Site
Engineering Performance Standards
I
Malcolm Pirme /TAMS- Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirme fTAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
-------
LEGEND
A/ Matchline
180 Rivermile
Tri+ PCB Entire Core
LWA Concentrations (mg/kg)
< 1
o 1-3
o 3-10
o 10-30
o 30 - 100
> 100
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Study N
Area 1
\ Study
\ Area 2
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
Match Line 1
Hudson
River
Figure 4-4
Preliminary Study Areas for the Special
Studies Showing LWA Concentrations
Sheet 1 of 7
Study
Area 4
/
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standard Volume 2: Resuspension - April 2004
-------
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
LEGEND
A/ Matchline
180 Rivermile
Tri+ PCB Entire Core
LWA Concentrations (mg/kg)
< 1
o 1-3
o 3-10
o 10-30
o 30 - 100
> 100
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
200 400 600 Feet
Hudson
River
Figure 4-4
Preliminary Study Areas for the Special
Studies Showing LWA Concentrations
Sheet 2 of 7
Study /N
Area 5 7
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
Match Line 1
Study
Area 4 /
/
Match Line 2
-------
N
LEGEND
A/ Matchline
180 Rivermile
Tri+ PCB Entire Core
LWA Concentrations (mg/kg)
< 1
o
1-3
o
3-10
o
10-30
o
30 - 100
o
> 100
Study
Area 7 \
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
200 0 200 400 600 Feet
Hudson
River
Figure 4-4
Preliminary Study Areas for the Special
Studies Showing LWA Concentrations
Sheet 3 of 7
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Resuspenion - April 2004
-------
LEGEND
/V Matchline
180 Rivermile
Tri+ PCB Entire Core
LWA Concentrations (mg/kg)
< 1
o 1-3
o 3-10
o 10-30
o 30 - 100
> 100
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Match Line 3
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
200 0 200 400 600 Feet
Hudson
River
Figure 4-4
Preliminary Study Areas for the Special
Studies Showing LWA Concentrations
Sheet 4 of 7
Match Line 4
Study
Area 8
Study
Area 9
Study
^ Area 10
Hudson River PCBs Superfund Site Malcolm Pimie/TAMS-Eartiti Tech
Engineering Performance Standard Volume 2: Resuspension - April 2004
-------
LEGEND
A/ Matchline
180 Rivermile
Tri+ PCB Entire Core
LWA Concentrations (mg/kg)
< 1
o 1-3
o 3-10
o 10-30
o 30 - 100
> 100
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
Match Line 5
Hudson
River
Figure 4-4
Preliminary Study Areas for the Special
Studies Showing LWA Concentrations
Sheet 5 of 7
1 Study
\ Area 9
Study
Area 10
Match Line 4
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standard Volume 2: Resuspension - April 2004
-------
Study
Area
LEGEND
A/ Matchline
180 Rivermile
Tri+ PCB Entire Core
LWA Concentrations (mg/kg)
< 1
o
1-3
o
3-10
o
10-30
o
30 - 100
o
> 100
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
N
200 0 200 400 600 Feet
Hudson
River
Figure 4-4
Preliminary Study Areas for the Special
Studies Showing LWA Concentrations
Sheet 6 of 7
Hudson River PCBs Superfund Site Malcolm Pimie/TAMS-Eartiti Tech
Engineering Performance Standard Volume 2: Resuspension - April 2004
-------
184
/
1 Study
/ Area 11
/ /
N
LEGEND
/V Matchline
180 Rivermile
Tri+ PCB Entire Core
LWA Concentrations (mg/kg)
< 1
o 1-3
o 3-10
o 10-30
o 30-100
> 100
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
Study
Area 13
300 0 300 600 900 Feet
Figure 4-4
Preliminary Study Areas for the Special
Studies Showing LWA Concentrations
Sheet 7 of 7
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pirnie/T AMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
LEGEND
A/ Matchline
180 Rivermile
Major Sediment Types (ASTM D422)
ฆ CL
~ SI
~ FS
~ MS
~ CS
ฆ GR
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
Study N
Area 1
\ Study
\ Area 2
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
Match Line 1
Hudson
River
Figure 4-5
Preliminary Study Areas for the Special
Studies Showing Sediment Types
Sheet 1 of 7
Study
Area 4
/
-------
LEGEND
A/ Matchline
180 Rivermile
Major Sediment Types (ASTM D422)
ฆ CL
~ SI
ฐ FS
~ MS
~ CS
ฆ GR
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
200 400 600 Feet
Hudson
River
Figure 4-5
Preliminary Study Areas for the Special
Studies Showing Sediment Types
Sheet 2 of 7
Study /N
Area 5 7
Study /
Area 6 /
/
/
/
/
/
Match Line 2
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
Match Line 1
Study
Area 4 /
/
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pimie/T AMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
N
+'
Area 7 \
LEGEND
A/ Matchline
180 Rivermile
Major Sediment Types (ASTM D422)
ฆ CL
~ SI
ฐ FS
~ MS
~ CS
ฆ GR
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
200 0 200 400 600 Feet
Hudson
River
Figure 4-5
Preliminary Study Areas for the Special
Studies Showing Sediment Types
Sheet 3 of 7
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Match Line 3
LEGEND
A/ Matchline
180 Rivermile
Major Sediment Types (ASTM D422)
ฆ CL
~ SI
ฐ FS
~ MS
~ CS
ฆ GR
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
200 0 200 400 600 Feet
Hudson
River
Figure 4-5
Preliminary Study Areas for the Special
Studies Showing Sediment Types
Sheet 4 of 7
Study J
Area 8 /
Match Line 4
Study
Area 9
Study
^ Area 10
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pimie/T AMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
W"
T
Match
200 0 200 400 600 Feet
Hudson
River
Figure 4-5
Preliminary Study Areas for the Special
Studies Showing Sediment Types
Sheet 5 of 7
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
Match
LEGEND
A/ Matchline
180 Rivermile
Major Sediment Types (ASTM D422)
ฆ CL
~ SI
~ FS
~ MS
~ CS
ฆ GR
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Study \
Area 8 \
LEGEND
A/ Matchline
180 Rivermile
Major Sediment Types (ASTM D422)
ฆ CL
~ SI
~ FS
~ MS
~ CS
ฆ GR
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
N
200 0 200 400 600 Feet
Hudson
River
Figure 4-5
Preliminary Study Areas for the Special
Studies Showing Sediment Types
Sheet 6 of 7
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
184
Study
Area 11
N
LEGEND
A/ Matchline
180 Rivermile
Major Sediment Types (ASTM D422)
ฆ CL
~ SI
ฐ FS
~ MS
~ CS
ฆ GR
Sediment types (Side Scan Sonar)
Type I
Type II
Type III
Type IV
Type V
Draft dredge area boundaries
~ Shoreline
Note: Draft dredge area boundaries are for
illustration purposes only. These boundaries
have not been approved by the USEPA.
} Study
Area 13
300 0 300 600 900 Feet
Figure 4-5
Preliminary Study Areas for the Special
Studies Showing Sediment Types
Sheet 7 of 7
Hudson River PCBs Superfund Site
Engineering Performance Standard
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Resuspension - April 2004
-------
Core23N
Cs-137
Core7M
NTCRA
Area V
Index Map
Figure 4 -6
PCB Profile in the Cores Samples Collected Post- Non-Time Critical Removal Action in theGrasse River
CorelSM
Cs-137 (pCi/g diy)
1000 2000 Feet
Core30S
Cs-137 (pCS/g dy)
2 3 4
ฆ ฆ ฆ ฆ ฆ - ฆ ฆ ฆ ฆ ฆ ฆ
Hudson River PCBs Superfund Site
Engineering Performance Standards
300 400 500
Total PCB (ppm)
Gs-137 (pCi/g diy)
0 1 2 3 4 5 6 7
120 I ' ' ' I 1
0 200 400 600 800 1000 1200 1400
Total PCB (ppm)
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Dredging Resuspension - April 2004
Total PCB (ppm)
70 i i i | i i i i | i i i i | i
0 50 100 150
Total PCB (ppm)
Cs-137 (pCi/g diy)
0 1 2 3 4 5 6
Li i i i Lui i i I 11 i i I 11 i i I i i 11 I i i 11 I
-------
Attachment A
Hudson River Water Column Concentration Analysis
Table of Contents
1.0 Introduction 1
2.0 Estimation of Hudson River Flow Rates at Stations Within the Upper Hudson
River 3
3.0 Baseline TSS and Total PCB Analysis 7
3.1 Methodology 7
3.2 Results and Discussion 16
4.0 References 21
LIST OF TABLES
Table 1 PCB versus Flow Correlation Analysis Based on the Fit Curve
Generated from Plot
Table 2 Statistic Results and Estimated Baseline Level of TSS and PCB
Concentration at Upper Hudson River Monitoring Stations
LIST OF FIGURES
Figure 1 Upper Hudson River Basin USGS Flow Gage Stations
Figure 2 Stillwater versus Ft. Edward Daily Runoff Yield 1998-2001
Figure 3 Ft. Edward Station Monthly TSS Concentration Variation
Figure 4 Ft. Edward Station Monthly PCB Concentration Variation
Figure 5 TID-West Station Monthly TSS Concentration Variation
Figure 6 TID-West Station Monthly Total PCB Concentration Variation
Figure 7 TID-PRW2 Station Monthly TSS Concentration Variation
Figure 8 TID-PRW2 Station Monthly Total PCB Concentration Variation
Figure 9 Schuylerville Station Monthly TSS Concentration Variation
Figure 10 Schuylerville Station Monthly Total PCB Concentration Variation
Figure 11 Schuylerville Monitoring Station Monthly TSS and PCB Concentrations
Plotted Against the Monthly Mean
Figure 12 Schuylerville Station Box Plots TSS Concentration vs. Month (Top
Diagram) Total PCB Concentration vs. Monthly (Bottom Diagram)
Figure 13 TID-West Monitoring Station Flow versus Total PCB Concentration
Months of May and June
Figure 14 TID-PRW2 Monitoring Station Flow versus Total PCB Concentration
Months of May and June
Hudson River PCBs Superfund Site
Engineering Performance Standards
l
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Attachment A
Hudson River Water Column Concentration Analysis
Table of Contents
LIST OF FIGURES (continued)
Figure 15 Schuylerville Monitoring Station Flow versus Total PCB Concentration
Months of May and June
Figure 16 Ft. Edward Monitoring Station Monthly TSS Data versus Estimated TSS
Baselines
Figure 17 Ft. Edward Monitoring Station Monthly Total PCB Data versus Estimated
Total PCB Baselines
Figure 18 TID-West Monitoring Station Monthly TSS Data versus Estimated TSS
Baselines
Figure 19 TID-West Monitoring Station Monthly Total PCB Data versus Estimated
Total PCB Baselines
Figure 20 TID-PRW2 Monitoring Station Monthly TSS Data versus Estimated TSS
Baselines
Figure 21 TID-PRW2 Monitoring Station Monthly Total PCB Data versus Estimated
Total PCB Baselines
Figure 22 Schuylerville Monitoring Station Monthly TSS Data versus Estimated
TSS Baselines
Figure 23 Schuylerville Monitoring Station Monthly Total PCB Data versus
Estimated Total PCB Baselines
Hudson River PCBs Superfund Site 11 Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
Attachment A
Hudson River Water Column Concentration Analysis
1.0 Introduction
An analysis of the existing total suspended solids (TSS) and PCB concentrations in the
water column of the Hudson River was conducted to determine baseline concentrations of
TSS and PCBs in the river. These baseline concentrations will be used as a comparison to
TSS and PCB concentrations measured during dredging to evaluate the impact o dredging
on the water column. However, it should be noted that the baseline and sensitivity
calculations provided here will be revised based on the results of the Baseline Monitoring
Program. To estimate the baseline concentrations of TSS and total PCBs in the water
column, the following tasks were performed:
Evaluation of the monthly flow rate over the dredging season.
Review and analysis of existing TSS and PCB data collected by General Electric
(GE) since 1996 were.
Estimation of the baseline PCB and TSS concentrations.
Limitations of the Existing Data
Much of the data analysis planned for the development of the resuspension performance
standard focuses on determining the pre-construction variability of contaminant
concentrations, or loads, in the water column. Previous studies, notably the Data
Evaluation and Interpretation Report (DEIR, USEPA, 1997), have shown that the
variability of contaminants in the water column changes throughout the year. The
variability of contaminants in the water column is greatest during the spring, and it
gradually decreases through the summer and into the fall.
For PCBs, the amount of available data is much greater, since nearly weekly sampling
was conducted from the early 1990s to the present. But only limited locations were
monitored, with the southernmost station located at Lock 5 in Schuylerville. Because the
amount of data from stations close to the Mid-Hudson portion of the river is limited, the
variability of contaminants in the water column at Waterford (sampled at the Troy Dam)
will be inferred from the Upper River stations. This approach is reasonable, but not
perfect. The contaminant concentrations at the TI Dam are much more variable than those
at the downstream stations because the dam is closer to the contaminant sources. As the
contaminant load travels downstream, the "signal" is dampened by dilution from tributary
inputs, homogenization, and settling of the contaminants. Thus, if the TI Dam variability
is assumed to apply to the Waterford area, the variability will be too high, leading to a
performance standard that is less conservative than it should be. Direct measurements of
the water column, expected to be provided by future GE sampling, will give a more
accurate representation of conditions at the Troy Dam.
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Although the dataset for PCBs encompasses most of the 1990s through the present, the
data sampled prior to 1996 may not be useable for performance standard development
due to the lack of source control at the GE facilities prior to that year. This leaves
approximately five years of data at the TI Dam, and less at the other water column
stations, for use in the planned evaluation. While this dataset would seem to be sufficient
to examine the variability of contaminant concentrations, there are concerns regarding the
location of the monitoring stations within the river at the TI Dam and Schuylerville.
At Schuylerville, the station is located near the Battenkill, but not at a point where
contaminant concentrations would be influenced by this tributary's input (the
station was not situated where complete mixing would be expected to occur).
Because of this, the Schuylerville (Lock 5) station may not fully represent the
Hudson River water column concentrations under all conditions. It is hoped that
future Schuylerville (Lock 5) samples will be collected from locations in the river
where the flows from the Hudson River and the Battenkill are sufficiently
homogenized, adopting a standard USGS sampling approach.
At the TI Dam, both a west wing station and a central channel station are
frequently sampled. Both stations have limitations. An analysis performed for the
Responsiveness Summary for the Data Evaluation and Interpretation Report
(USEPA, 1998) on the results in the west wing indicated that the concentrations
from this station may be strongly influenced by the nearby sediments, particularly
during times of low flow. The center channel station is north (upstream) of the
west wing station, and thus does not measure the impact from the side channel
sediments near the dam. Also, the center channel is inaccessible during the winter
months due to ice cover, so the dataset is limited to the warmer months.
Subsequent analysis indicated that the downstream concentrations (Schuylerville)
are unlike either station taken separately, but resemble a mix of the concentrations
measured at the two stations.
These concerns regarding the existing water column dataset have an impact on the
evaluation of water column contaminant variability. It is unclear whether the estimated
variability derived only from historic data will be more or less conservative than the
actual conditions in the river. If GE adjusts the locations of the monitoring stations during
future sampling events, a better measure of variability will be obtained.
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
2.0 Estimation of Hudson River Flow Rates at Stations Within the
Upper Hudson River
Monitoring of resuspension in the water column of the Hudson River during dredging
will include the measurement of PCBs at the far-field monitoring locations and
measurement of turbidity and TSS at locations near the dredging operation, to ensure that
the loss of PCBs from dredging is not occurring at a level greater than the baseline
variability of PCBs already present in the water column.
Based on this need, it has been concluded that the far-field monitoring stations will be
situated at the downstream limit of each of the three pools during dredging. Of these
locations, only three have a long history of water column measurements: the TI Dam,
Schuylerville (Lock 5), and Waterford (Troy Dam). For each of these locations, the
baseline variability of TSS and PCB loading to the water column must be computed to
establish a baseline for monitoring during implementation of the remedy. To determine
the baseline variability of PCBs and TSS concentrations at the monitoring locations, the
flow rates at these locations are needed.
The USGS monitors the flow rate of the Hudson River at gauges in the following
locations:
At Ft. Edward, along the Hoosic River
On the Batten Kill before it converges with the Hudson River at Schuylerville,
On the Hudson River just north of Waterford,
Within the drainage areas surrounding the Hudson River.
In addition, the flow rate at Stillwater is estimated by the USGS. The flow rates at TI
Dam and Schuylerville are not readily available.
Flow rates at the TI Dam and Schuylerville were computed using the drainage-area ratio
method and known flow rates from existing USGS gauge stations. Flows were
determined for the period 1977 to 2001 to incorporate all flow rate data available at the
gauged stations.
Schuylerville Flow Rate Calculation
As shown in Figure 1, the flow rate of the Hudson River as it passes through
Schuylerville is equal to the sum of the following:
The flow rate of the Hudson River measured at the USGS gauge station at Ft.
Edward.
The flow rate measured by USGS at the gauge station along the Batten Kill.
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
The flow contribution from this gauged station along the Batten Kill westward to
its confluence with the Hudson River.
The flow rate between Ft. Edward and Schuylerville.
This relationship is described by the following equation:
Flow rate at Schuylerville = Fschuy = FFtEd + FBKg + FBKung + Afung.schuy ...Equation 1
where FFt.Ed = Flow at Ft. Edward
FBKg = Flow at the Batten Kill gauge station
FBkung = Flow within the ungauged section of the Batten Kill
Afung-schuy= Change in flow rate of the ungauged section of the Hudson
River between Ft Edward and Schuylerville
Using the drainage-area ratio method, the relationship of watershed yield times the
drainage area of the watershed was used to compute the corresponding flow rate of the
watershed. In the foregoing equation, the flow rate within the ungauged section of the
Batten Kill (FBKung) was computed by multiplying the yield of the Batten Kill by the
change in watershed area over the ungauged section of the Batten Kill (the difference of
the total area of the Batten Kill minus the gauged area along the Batten Kill) before it has
its confluence with the Hudson River. This relationship is expressed in Equations 2 and 3,
shown below.
FBKg = yBKg * AsKg Equation 2
where FBKg = Flow rate at the Batten Kill USGS gauge station
yBKg = Yield for the Batten Kill gauged section of the River
Auivg = Drainage area for the Batten Kill gauged section of the river
FBKung - yBKg * AflKung - (FBKg/ABKg)* AflKung Equation 3
where FBKung = Flow rate for the ungauged section of the Batten Kill
AsKung = Drainage area for the ungauged section of the Batten Kill
=Abk - AflKg
Abk = Total drainage area of the Batten Kill
The flow rate contributed by the section of the Hudson River between Ft. Edward and
Schuylerville was computed as the change in flow rate between the flow rates measured
at Ft. Edward and Stillwater by USGS and both the gauged and ungauged sections of the
Batten Kill.
Afung-schuy Aaung-schuy * Yung Equation 4
where
Yung (Fstwtr ฆ Fpt.Ed " FBKg " FBKung)/( Astwtr" Apt.Ed " AsKg " AgKung) Equation 5
Hudson River PCBs Superfund Site 4 Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
and
A&ung-schuy Asdnjy " Apt.Ed " AgKg ~ AjBkung Equation 6
and
Afung-schuy = Change in flow rate of the ungauged section of the Hudson River between
Ft. Edward and Schuylerville
Aaung-schuy = Change in the drainage area of the ungauged section of the Hudson River
between Ft. Edward and Schuylerville
yung= Yield for the ungauged section of the Hudson River between Ft.
Edward and Stillwater
Fstwtr = USGS estimated flow rate of the Hudson River at Stillwater
Astwt = Drainage Area that enters the Hudson River at Stillwater
AFt.Ed = Drainage area that enters the Hudson River at Ft. Edward
Aschuy = Drainage area that enters the Hudson River at Schuylerville
For select days over the period 1977 through 2001, the estimated flow rates at Stillwater
were less than that of Fort Edward. In these instances, the following relationship was
used to estimate the flow rate at Schuylerville:
Fschuy Fpt.Ed Fukg-'- Fgkung Aaung-Schuy * YBKg Equation 7
Thompson Island Dam Flow Rate Calculation
The flow rate at the TI Dam was computed similarly to the flow rate at Schuylerville; the
drainage-area ratio method and the measured flow at the Ft. Edward gauge were used to
estimate the flow at the dam. The following equations, Equations 8, 9, and 10, depict the
relationships used to predict the flow rate at the TI Dam (FTid):
Ftid = FFt.Ed + Afxio Equation 8
where
AfnD = Aaxio * yUng Equation 9
and
Aaxro = Atid - AFt.Ed Equation 10
and
Ftid = Flow rate of the Hudson River at the Thompson Island Dam
AfxiD = Change in flow rate along the Hudson River between Ft. Edward and the
Thompson Island Dam
Aaxio = Change in the drainage area into the Hudson River between Ft. Edward and the
Thompson Island Dam
Atid = Drainage area into the Hudson River at the Thompson Island Dam
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
For days where data gaps existed at the Ft. Edward USGS gauge station, the flow at Ft.
Edward was estimated from the regression equation generated from the plot of the daily
runoff yield at Stillwater versus the daily runoff yield at Ft. Edward (Figure 2). This plot
generated the following equation that was used to estimate the flow rate at Ft. Edward:
FFt.Ed= 1.05*ystwh-*AFt.Ed Equation 11
where ystwtr = Yield for the Hudson River drainage area at Stillwater
and other parameters as defined above
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
3.0 Baseline TSS and Total PCB Analysis
The major concern associated with the dredging operation is the resuspension of sediment
particles that may introduce additional PCB contamination into the water column. The
water column PCB concentration will be monitored during dredging operations, and
actions will be taken to minimize the impact of resuspension on the river system when
the PCB concentration exceeds a certain level/threshold. Previous sampling results
indicate that the variability of water column contaminant concentrations in the Hudson
River can, to some extent, be attributed to the uncertainty of laboratory analyses.
However, the variability in contaminant concentration in the water column is actually
primarily the result of variability of the river system. In order to measure the "net" effect
of the dredging operation, it is necessary to distinguish the dredging-related contribution
of PCB contamination to the water column from the flux of PCBs currently present in the
water column. If the new measurements collected during dredging are within the
variability determined by the samples collected prior to the onset of dredging activities, it
will be assumed that there is no impact from dredging. This poses the question of whether
each new observation/sample collected belongs to the populations created from the
baseline data and if the new observations generate the same central tendency as the
baseline data. To evaluate this question, a statistical analysis was performed over the
multiple-year baseline water column data set to investigate the typical condition of the
river and to estimate the upper bound and typical PCB contaminant levels representative
of the river system.
3.1 Methodology
Samples collected by GE during their ongoing weekly sampling program were used to
estimate the current PCB water column contamination conditions in the Hudson River.
The GE sample results were used because they provide a long record of PCB and TSS
concentrations in the Hudson River, have measured PCB concentrations using a congener
method, represent the most comprehensive dataset of water column PCB results, and
probably best reflect the current situation in the Hudson River. There are some problems
with the data collection method that make this data less than representative; the samples
were collected from a single centroid sample to represent the cross-section, and the
detection limits are not low enough to detect concentrations at all stations throughout the
year. Only post-1996 water column samples were used in this analysis (due to the lack of
source control at the GE facilities prior to that year) to estimate the baseline conditions in
the Hudson River prior to any impact that may result from the dredging operation.
GE has been monitoring the water column situation in the Upper Hudson River at four
stations since the early 1990s. These four stations are located at Fort Edward, at the west
side of the TI Dam near the shore (TID-West), in the channel section above the TI Dam
(TID-PRW2), and at Schuylerville (Lock 5). Data collected at the above-listed stations
were investigated in this study to estimate the natural variability of TSS and PCB
concentration in the river system at different locations. Daily average flow measured and
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
reported by the USGS was used for PCB and TSS analysis at the Fort Edward station.
The flow rate at the other three stations was estimated based on the flow rate at Fort
Edward, as described in detail in Section 2.0 of this attachment.
Since the proposed dredging season spans the months of May through November, only
data associated with these seven months were examined for distribution and variability
herein. As stated in the Hudson River Feasibility Study Report (USEPA, 2000), dredging
is not expected to be performed during high flow conditions. Therefore, samples with
flows greater than 10,000 cfs were excluded from this analysis in order to provide the
best estimate of what conditions will be during dredging activities. Field duplicate
samples were collected for 1 percent of the total samples taken, and an average
concentration was calculated to represent the results of all duplicates. In addition, for
cases where multiple samples were collected at different times in the same day, a daily
average concentration was calculated and used in this analysis in order to evenly weight
each sampling day.
Non-detected values for both TSS and PCBs exist in the GE data set. Typically, when
these results are used in a calculation, a value is substituted for the detection limit to
estimate the concentration in the sample. Usually, either zero or one-half the detection
limit is used in the substitution. In the data reviewed, GE did not provide a detection limit
for TSS, and, in some instances, for PCBs. To determine the best estimate of the
concentration in the non-detect TSS samples, a concentration of 0.5 mg/L TSS, one-half
of the lowest detected TSS concentration, was assigned to the non-detect samples. To
determine the best estimate of the concentration in the non-detect PCB samples, half of
the reported detection limit for PCBs (5.5 ng/L) was assigned to PCB samples reported as
non-detect from the laboratory.
The impact of resuspension on water column PCB concentrations is the focus of concern
during the dredging operation. Some PCBs stored in the sediment will be introduced into
the water column via resuspended particles. As a result, a change in the TSS
concentration can be used as an indicator of a possible increase in the PCB concentration
in the water column. There are currently no instruments capable of making reliable
measurements of PCBs in-situ. Measurements of PCB concentration must be performed
through laboratory analysis and measurement, which can take hours to perform. Due to
the inability to obtain real-time PCB concentrations in the water column during dredging,
TSS will be used as a surrogate indicator of dreging related releases and thereby PCB
release also. Therefore, baseline conditions for both PCB and TSS concentrations were
analyzed herein.
Review of the PCB and TSS data collected by GE since 1996 at the Ft. Edward, TID-
West, TID-PRW2, and Schuylerville monitoring stations indicated the following:
Variation exists among different months' data, and
A single concentration could not be computed for TSS or PCB to represent the
background concentration over the seven-month dredging period.
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Based on the above, PCB and TSS data were analyzed starting on a monthly basis at each
of the four monitoring stations. This monthly variation can be seen in Figures 3 and 4 for
the Ft. Edward station; Figures 5 and 6 for the TID-West station; Figures 7 and 8 for the
TID-PRW2 station; and in Figures 9 and 10 for the Schuylerville monitoring station.
An example of the data analysis performed for the monthly TSS and PCB data at the
above-listed stations follows, using the results from the Schuylerville station.
Figure 11 depicts results from the weekly PCB and TSS water column samples collected
at the Schuylerville monitoring station from 1996 through 2001, grouped by month (May
through November). The mean TSS and PCB concentrations for each month were
calculated and plotted to show the trend of the monthly concentrations. The data indicate
that relatively high TSS and PCB concentrations were detected more frequently in May
and June as compared to the rest of the study period. PCB data in May and June are
distributed over a broader range than the data in the other dredging months. The mean
TSS concentration fluctuates for the period of July through September, while the mean
PCB concentration declines over that same period. In addition, October's mean PCB
concentration is greater than the mean PCB concentration for September and November.
The data strongly suggests that a single uniform TSS or PCB baseline value cannot be
applied to every month. Similar analyses were performed for each of the data sets
representative of the other three monitoring stations, and the same conclusion was drawn:
that significant difference exists between the data collected at different times of the year,
(for example, data collected during a spring month differ significantly from data collected
during a summer month), and a uniform baseline value would not be representative of the
range of conditions expected to be encountered during the dredging period. The baseline
variability of the Hudson River should be addressed by a set of time-specific groupings of
the available data in a reasonable way.
There are approximately 20 to 25 data points available for each month. A data group of a
smaller size will not permit a reliable statistic analysis result, so one month is the smallest
unit to group the data into for this analysis. In addition, it is physically meaningful to
generate a baseline number for each month. Statistical analysis was conducted on each of
the monthly datasets to determine whether or not it would be appropriate to group data
for some months together. JMP (SAS, 1997), a statistical program, was used to perform
the statistical analysis. This study included the following:
Calculation of the minimum, mean, and maximum concentrations for each
month
Calculation of the 10 percent, 25 percent, 75 percent, and 90 percent quantiles
Use of the Tukey-Kramer Honestly Significant Difference (HSD) to
determine whether or not two sets of data are significantly different.
Hudson River PCB s Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
A sample plot for this study of TSS and PCB concentrations measured at the
Schuylerville Station over the seven months of interest is shown in Figure 12. Generally
speaking, this statistical study allowed months exhibiting insignificantly different means
to be grouped. Circles for means that are significantly different either do not intersect or
intersect slightly so that the outside angle of intersection is less than 90 degrees. If the
circles intersect by an angle of more than 90 degrees, or if they are nested, the means are
not significantly different. Figure 12 shows that TSS data for the period of July through
November at the Schuylerville station are similar. Thus, data for these "similar" months
can be consolidated into one dataset for further analysis to determine a baseline TSS
concentration. Figure 12 also indicates that PCB data for the months of May and June are
similar and can be consolidated into one dataset.
The studies performed on TSS and PCB data for the Ft. Edward, TID-West, and TID-
PRW2 stations allowed the consolidation of several months of data into one dataset in the
following cases:
At the Ft. Edward station: consolidation of TSS data for September through
November, and PCB data for July through September and October and
November.
At the TID-West station: consolidation of TSS data for July through October and
PCB data for October and November.
For the TID-PRW2 station: consolidation of TSS data for July through November
and PCB data for the months of July and August.
The variability of monthly and consolidated monthly TSS and PCB data was analyzed
based on interval estimates. Interval estimates are intervals that have a stated probability
of containing the true population value. The intervals are wider for datasets having
greater variability. There are two types of interval estimates: the prediction interval (PI)
and the confidence interval. The prediction interval indicates the likelihood that a single
data point with a specific magnitude comes from the population under study, while the
confidence interval indicates the probability or likelihood that the interval contains the
true population value. For each of the four monitoring stations, the prediction interval and
the 95 percent confidence interval were estimated for each month and consolidated month
dataset over the dredging period, since previous analysis of the data indicated that PCB
and TSS concentration data varied.
Prediction intervals are computed for a different purpose than confidence intervals. The
prediction interval deals with the individual data values as compared to a summary
statistic such as the mean. A prediction interval is wider than the corresponding
confidence interval because an individual observation is more variable than a summary
statistic computed from several observations. Unlike a confidence interval, a prediction
interval takes into account the variability of single data points around the median and
mean, in addition to the error in estimating the center of the distribution.
Hudson River PCB s Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
In order to judge whether a new observation is likely to have come from the same
distribution as previously collected data or, alternatively, from a different distribution, the
prediction interval needs to be computed from the existing data and compared to the new
observation. Prediction intervals contain 100*(l-a) percent of the data distribution, while
100*a percent are outside of the interval. If a new observation comes from the same
distribution as previously collected data, there is a 100*a percent chance that it will lie
outside the prediction level. Therefore, being outside of the interval does not "prove" that
the new observation is different, just that it is likely to be so. Prediction intervals are used
in this study as the upper bound limit for a single incident, and will be used as a baseline
for comparison for a single sample result collected during the dredging operation. Sample
results obtained during dredging falling above this upper bound limit (the prediction
interval) will be assumed to represent some dredging-related release.
In addition to providing the prediction limit which provides an upper bound limit for
individual samples, the confidence limit on the average was used as the second layer
criterion to control the average of new observations. Therefore, if a group of samples are
each below the prediction limit, but the average is above the upper confidence limit, it is
likely that the group of samples belong to a different population than the baseline {i.e.
indicative of dredging related releases).
Considering the possible impact of flow rate on PCB and TSS concentrations,
correlations between PCB concentration and flow and between TSS concentration and
flow were examined for the dredge season, either monthly or per consolidated set of
dredging months, at each station. For each monitoring station, flow was plotted against
PCB and TSS water column concentrations. Overall, no correlation was observed
between TSS and flow at any of the four monitoring stations.
No correlation between PCB and flow was observed at the Ft. Edward monitoring station,
but data indicated that correlations existed between PCB concentration and flow rate
during the months of May and June at the TID-West and TID-PRW2 stations. Data for
the Schuylerville station also indicated a correlation between PCB and flow for the
months of May and June. Statistical data were indicative of these correlations based on a
high r-squared value and an observed significant probability that was less than 0.05. The
above-described correlations are presented in the following figures: TID-west station
(Figure 13), TID-PRW2 station (Figure 14), and Schuylerville station (Figure 15).
For months where PCB data appeared to be correlated with the flow rate, JMP was used
to estimate the center confidence and individual confidence of the data corresponding to
different flows. The center confidence puts a confidence limit on the predicted central
tendency, and the individual confidence interval includes both the variability of the
estimates and the variability of the observation itself and is thus appropriate for a
prediction interval. The JMP program was able to compute these values while performing
a regression analysis between two correlated variables. The lower 95 percent confidence
interval is not presented in these plots, since only the upper bound estimates were of
interest in this study.
Hudson River PCB s Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Fit curves estimating the center confidence and individual confidence of the data were
generated for the PCB monthly data at the TID-West, TID-PRW2, and Schuylerville
monitoring stations for months in which the data indicated a correlation between PCB
concentration and flow rate. These fit curves are shown in Table 1. For stations with a
defined timeframe where PCBs are correlated with flow, the flow rate was applied to the
listed formulas and fit curves to determine the baseline PCB concentration, the prediction
interval, and the UCL at different flows. Velocities of 2000 cfs, 4000 cfs, and 8000 cfs
were used to calculate the baseline levels, representing the lower bound flow, the average
flow, and the upper bound flow, respectively, during dredging operations.
For the monthly and consolidated monthly datasets where a correlation between flow and
concentration was not observed, the prediction interval and UCL were estimated solely
based on the concentration data.
The upper bound prediction interval was estimated using methods provided by Helsel and
Hirsch (Helsel and Hirsch, 2002). Three methods were used to calculate the upper 95th
prediction interval on each of the datasets. These methods were the parametric symmetric
prediction interval, the parametric asymmetric prediction interval, and the nonparametric
prediction interval. Because the goal of this study was to determine the upper bound level
of existing data, a one-side prediction interval was applied in all three methods. The
nonparametric prediction interval does not require the data to follow any particular
distribution shape, while the symmetric prediction interval is calculated based on the
assumption that the data follow a normal distribution. The following formula, Equation
12, is used to compute the symmetric prediction interval:
PI = X + t(0.05,n- 1) -y/52 + (V / n) Equation 12
where PI = the upper bound of the prediction interval
X = the mean value of the data set (mean concentration for the TSS and
PCB data sets)
t = the student's t for alpha equal to 0.05 and n-1 degrees of freedom
s = the variance of the data set
n = number of data points
The parametric asymmetric prediction interval assumes that the data follows a lognormal
distribution, and the prediction interval is computed using the formula shown in Equation
13.
PI = exp(j + t(0.05,n - 1) + s2y / n Equation 13
where y = ln(x), y is the mean and s2y is the variance of the logarithms
y = the mean logarithm
s y =the variance of the logarithms
n = number of data points
Hudson River PCBs Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
t = the student's t for alpha equal to 0.05 and n-1 degrees of freedom
The non-parametric prediction interval is computed from statistical analysis of the data
and is representative of the 95th percentile of the dataset.
Similarly, three methods were used to estimate the upper bound confidence interval for
each dataset based on the distribution of the data. The following formula, Equation 14,
was used to compute the 95 percent UCL on datasets exhibiting a normal distribution:
UCL = X + t (sHn) Equation 14
where X = arithmetic mean of the sample data set for the compound of
concern,
s = sample standard deviation of the sample data set for the compound
of concern,
t = the student's t statistic for the 95 percent confidence interval for a one
tailed distribution. The t-statistic is a function of the number of
samples collected, and;
n = number of samples in the data set
For data sets that exhibited a lognormal distribution, the 95 percent UCL was computed
using Equation 15, shown below.
UCL = EXP [ X + 0.50s2 + Hs Vn -1 ] Equation 15
where X = arithmetic average of the natural log-transformed data;
s2 = variance of the log-transformed data;
s = sample standard deviation of the log-transformed data;
H = H statistic. The H value differs from the t-values because the
formula is designed to estimate the UCL on the basis of the log-
transformed data. H is a function of the standard deviation of the
log-transformed data and the number of samples in the data set. H
was taken from a standard table of calculated values (Gilbert,
1987) or linearly interpolated between values given in the table
where necessary; and
n = the number of samples in the data set.
For non-parametric data sets, the 95 percent UCL was calculated using ProUCL (USEPA,
2001). ProUCL does provide several types of non-parametric UCLs. As recommended in
the User's Guide for ProUCL, the 95 percent Chebyshev UCL was selected for this
analysis since all of the datasets that were neither normally distributed nor lognormally
distributed had a standard deviation (a) less than 1.
The Shapiro-Wilk test (W-test) and D'Agostino's test were used to determine the best data
relationship among each of the monthly data sets for all four stations so the prediction
Hudson River PCBs Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
interval and the 95 percent UCL could be calculated, based on the determined distribution
of the data, using one of the above-listed equations. For months in which the number of
samples was less than 50 (n<50), the W-test was used to evaluate the distribution of the
dataset. For months in which the number of samples was greater than 50 (n>50),
D'Agostino's Test was used to evaluate the distribution of the dataset.
The W-test generates a W-value and an ln-W value, which are then compared to the 5
percent W critical value. If the calculated W-value is greater than this critical value, the
distribution is determined to be normal at the 5 percent confidence level. Similarly, if the
computed ln-W value is greater than the critical value, then the data distribution is
determined to be lognormal. In the event that the computed W-value and ln-W value are
both greater than the critical value, the larger computed value (i.e., the W-value or the ln-
W value) will determine the data distribution. If both of the computed values (i.e., the W-
value and the ln-W value) are less than the critical W value, then the distribution is
determined to be non-parametric.
For monthly and consolidated monthly datasets with more than 50 samples, D'Agostino's
test was used to compute a Y-value and an ln-Y value, which are then compared to a
range of set critical values. The distribution is considered to be normal when the
calculated Y-value is within the range of critical Y-values. The data set is determined to
be lognormal when the ln-Y value is within the range of critical ln-Y values. If the
computed Y-value and ln-Y value satisfy both the normal distribution and lognormal
distribution requirement, then the value representing the smallest absolute value of Y
dictates the data distribution. Lastly, if the Y-value and ln-Y-value do not meet the
criteria that are indicative of normal or lognormal distribution, then the data set is
determined to be non-parametric.
For monthly and consolidated monthly datasets determined to have a normal distribution
of data, the prediction interval and the 95 percent UCL were computed from Equations
12 and 14, respectively, to determine the baseline concentrations for TSS and PCB at
each station. Similarly, for monthly and consolidated monthly datasets determined to
have a lognormal distribution of data, the prediction interval and the 95 percent UCL
were computed from Equations 13 and 15, respectively, to determine the baseline
concentrations for TSS and PCB at each station. Lastly, as described above, the 95th
percentile of the dataset was computed to determine the prediction interval baseline, and
ProUCL was used to determine the 95 percent UCL baseline for months and consolidated
months where the data were distributed in a non-parametric relationship.
These statistical tests were performed for each of the seven dredging months and
consolidated dredging months at each of the four monitoring stations. The results are
presented in Table 2, and were indicative of the following at each of the monitoring
stations:
A prediction interval baseline for PCB and TSS per month and consolidated
months
Hudson River PCB s Superfund Site
Engineering Performance Standards
14
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
A 95 percent UCL baseline for PCB and TSS per month and consolidated
months
The results for each monitoring station are included below, along with a discussion of the
estimated baseline concentrations for the dredging season. Ultimately, these baselines
will be compared against PCB and TSS measurements made during dredging operations
to assess potential dredging-related impacts.
Note that only the samples associated with high flow events were excluded during the
data analysis procedure. No data were excluded as outliers. Some elevated values found
in the dataset are representative of values that could occur during the remediation,
thereby making it inappropriate to treat them as outliers, although in a strict mathematical
sense the values might fall into that category. This analysis is also intended to show the
approach used to estimate the baseline. The final baseline values will be calculated using
Baseline Monitoring Program data, which is scheduled for collection between 2004 and
2005. When the baseline data is available, some outlier analysis methods, such as Dicson
analysis and Mahanalobis Distance, may be used to identify the outliers based on
engineering judgment in order to provide a baseline level for addressing the Hudson
River condition prior to dredging.
For the datasets in which PCBs were determined to be correlated with flow, the
prediction interval and UCL of the PCB concentration were estimated using the same
method that was used for datasets where concentration is not correlated with flow. The
prediction interval and UCL values generated by this method are similar to the results
obtained assuming a flow of 4000 cfs and using the equations listed in Table 1. A flow
rate of 4000 cfs is assumed to be the average velocity that will be observed during the
dredging period. Therefore, the values generated by this simple (no flow involvement)
method adequately reflect the PCB concentration under the average river flow conditions.
It was also found that the estimated prediction interval and UCL values calculated for
velocities of 2000 cfs and 8000 cfs were approximately within 20 percent of the values
calculated for a velocity of 4000 cfs. The 20 percent variance is not a pronounced
difference when considering other uncertainties involved in the analysis.
Lastly, it was thought that the measurement of the flow rate and application of the above
formulas may be impractical tasks for the dredging operator to perform in the field in
order to determine the PCB concentration. A developed baseline with PCB
concentrations defined for each month and set of months over the dredging season would
be the easiest and the most practical method for field application. It was concluded that
the baseline levels (prediction interval and UCL) are all estimated based on the
assumption that there is no correlation between flow and concentrations. The flow-
independent prediction interval and UCL values are calculated and summarized in Table
2 for each month and consolidated months at each station.
It should be noted that all the analyses listed above are intended to demonstrate the
approach used to estimate the baseline. When the new baseline data is available, the same
Hudson River PCBs Superfund Site
Engineering Performance Standards
15
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
type of analyses will be conducted, and the results may suggest some ways to simplify
the process. The baseline level will be finalized based on both the new baseline level data
and historic data.
3.2 Results and Discussion
Ft. Edward Monitoring Station
Water quality data for TSS were analyzed individually for May, June, July, and August,
and jointly over the period of September through November. PCB data were analyzed
individually for May and June, jointly over the period of July through September, and
jointly over the period of October and November. These results are shown on Table 2.
As shown in Table 2, data collected for TSS during the months of May, July, and
September through November have a normal distribution. In contrast, the data collected
for TSS during the month of June has a non-parametric distribution and that collected for
August has a lognormal distribution.
Figure 16 indicates that the prediction interval baseline generally tends to correspond to
the maximum measured TSS concentration for a particular month, with the exception of
months where elevated TSS data points exist. June and August each have one TSS data
point that contains one TSS sample result that is more than twice the concentration of all
other TSS results obtained for these months. For these two instances, the prediction
interval baseline and the 95 percent UCL are representative of the majority of the data. It
should also be noted that the 95 percent UCL is greater than the prediction interval for the
month of June. However, for all other months, the prediction interval represents the upper
limit TSS baseline concentration.
The prediction interval baseline is highest in August, with a concentration of 5.5 mg/L. In
the months prior to August, the prediction interval is approximately 4.0 mg/L, on
average, while for the remainder of the dredging season, in the months of September
through November, the prediction interval decreases to 3.0 mg/L and levels out. The 95
percent UCL baseline follows the same seasonal distribution as the prediction interval,
but reaches a maximum concentration of 5.7 mg/L in June. This baseline then decreases
by 3 mg/L and fluctuates through July and August, eventually leveling out at 1.8 mg/L
during the period of September through November.
The estimated 95 percent UCL baseline for TSS appears to be consistent with the mean
TSS data concentration for each month, and the estimated prediction interval appears to
be consistent with the upper bound measured TSS concentration for each month, with the
exception of June and August where two outlying TSS concentrations exist (as previously
discussed). It can be concluded that if a single TSS measurement made during dredging is
greater than the prediction interval concentrations, or if the average of a set quantity of
measured samples are greater than the 95 percent UCL baseline, the measured TSS
concentration is most likely a result of the dredging operation.
Hudson River PCB s Superfund Site
Engineering Performance Standards
16
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
An analysis of total PCB data collected during the proposed dredging season at the Ft.
Edward monitoring station indicated that all data were representative of a non-parametric
distribution. The results are presented in Table 2. The estimated baselines were plotted
against the total PCB monthly datasets. These relationships are presented in Figure 17.
Figure 17 indicates that total PCB concentrations measured for this station were greatest
in the months of July through August, and that the lowest concentrations were measured
during the month of May. Data indicate that the estimated prediction interval baseline
corresponds to the upper bound total PCB concentrations measured each month. The
prediction interval baseline is the highest for total PCBs during the months of July
through September, and lowest total PCB concentration during the month of May. The
prediction interval baseline decreases by 15 ng/L from September to October and levels
out at 19 ng/L for the period of October through November. It can be concluded that any
PCB measurements with a concentration greater than the prediction interval can most
likely be attributed to dredging.
The 95 percent UCL baseline result per month is always less than the prediction interval
baseline result, and tends to correspond to the mean total PCB concentration per month,
as shown in Figure 17. This lowest baseline concentration on the curve occurs during the
month of May and the maximum concentration occurs during the month of June. Baseline
values occur during the months of July through September, and are lower in
concentration than the maximum estimated concentration by approximately 0.4 ng/L. The
95 percent UCL baseline concentration decreases to 10.4 ng/L in October, a result that is
8 ng/L less than the September level. This concentration remains constant during the
months of October and November. It can be concluded that if the average of the PCB
measurements reported during dredging activities exceeds the 95 percent UCL, it is most
likely attributable to the dredging operation.
Thompson Island Dam (TID) Monitoring Stations
There are two GE monitoring stations located at the TI Dam: TID-West, located on the
west side of the TI Dam near the shore, and TID-PRW2, located in the channel section of
the river near the dam. TSS and total PCB monthly data and consolidated monthly data
were analyzed for each of these stations. Subsequently, the prediction interval and the 95
percent UCL baseline were determined for each station's monthly and monthly
consolidated TSS and total PCB data.
TID- West Monitoring Station
As shown in Table 2, TSS data analyzed at the TID-West station exhibited a non-
parametric relationship for May and June. A lognormal relationship was determined for
consolidated monthly data representing the period July through October and also for the
month of November. The estimated prediction interval and 95 percent UCL are shown in
Figures 18 and 19.
Hudson River PCBs Superfund Site
Engineering Performance Standards
17
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Figure 18 compares the monthly TSS data at the TID-West station with the estimated
prediction interval baseline and the estimated 95 percent UCL baseline. This figure
depicts that the prediction interval baseline is always greater than the 95 percent UCL
baseline and tends to follow the maximum measured TSS concentration reported for each
dredging month. Exceptions to this conclusion exist during months where elevated TSS
concentrations exist, in this case May, June, July, and August. In these instances, the
prediction interval baseline tends to correspond to a data point midway between the
majority of the sample results and the elevated data point (i.e., the prediction interval
tends to fall at a data point consistent with the maximum concentration of samples,
excluding the outlier for these months). The maximum TSS prediction interval baseline
value occurs during the month of May. This baseline decreases through June to
approximately 5 mg/L during the month of July. The baseline remains level until
October, and then increases slightly to 6.4 mg/L during the month of November.
The 95 percent UCL baseline shown in Figure 18 tends to follow the mean TSS
concentration in each dredging month, with a maximum estimated concentration
occurring in May and June and a minimum concentration occurring during the months of
July through October.
The total PCB data reported for this station follow a lognormal distribution for May,
June, August, and September. Total PCB data reported for July were determined to
follow a normal distribution, and total PCB data for the period of October through
November were determined to represent a non-parametric relationship.
As shown in Figure 19, the estimated prediction interval baseline consists of total PCB
concentrations greater than those estimated for the 95 percent UCL baseline. The
prediction interval maximum total PCB result occurs during the months of May and June,
with a total PCB concentration of approximately 370 ng/L. The prediction interval
baseline then decreases through July (211 ng/L) and August (150 ng/L), and reaches a
minimum value of 120 ng/L during the month of September. During the months of
October and November, the prediction interval baseline total PCB concentration
increases to 300 ng/L. It was also noted that the prediction interval tends to be consistent
with the maximum total PCB data concentration reported for each dredging month, on
average.
The estimated 95 percent UCL baseline for total PCBs at the TID-West station tends to
correspond with the mean total PCB concentration for most dredging months, on average.
This can be seen in Figure 19. This baseline concentration is approximately 200 ng/L
from May to June, and decreases through July (150 ng/L) and August (106 ng/L). The
baseline reaches a minimum concentration of 83 ng/L in September, and then increases to
a maximum concentration of 241 ng/L during the period of October and November. It is
noted that the 95 percent UCL baseline follows the same seasonal variation as the
estimated prediction interval baseline.
Hudson River PCBs Superfund Site
Engineering Performance Standards
18
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
TID-PRW2 Monitoring Station
TSS data collected at this station exhibited a lognormal distribution for the month of May
and for the period of July through November. Data indicated a non-parametric
distribution for the month of June.
Figure 20 shows that the estimated prediction interval baseline tends to generally
correspond with the maximum monthly TSS concentration for all months, with the
exception of May, June, July, and August, where elevated TSS data exist. In these
instances, the estimated prediction interval tends to represent the maximum TSS
concentration associated with the majority of the data points. The prediction interval
baseline concentration reaches a maximum during the month of June (15 mg/L) and
decreases to 5 mg/L for the months of July through November.
The estimated 95 percent UCL baseline for TSS, shown in Figure 20, tends to correspond
with the monthly mean TSS concentration for all months, with the exception of May and
June. The baseline reaches a maximum during June (14 mg/L TSS), and decreases to a
concentration of 2 mg/L for the months of July through November.
The total PCB data indicated that the months of May, June, October, and November all
exhibited a normal data distribution, and that the datasets for the consolidated months of
July and August and the month of September each exhibited a lognormal data
distribution.
Figure 21 indicates that the estimated prediction interval fluctuates throughout the
proposed dredge season, with a minimum concentration in May and June and a maximum
concentration through the period of July and August. The estimated total PCB
concentration in September and November are just above the minimum estimated
concentration in May and June, but less than the estimated baseline value for the month
of October. For most months, with the exception of May and June, the estimated
prediction interval baseline tends to correspond with the maximum monthly total PCB
concentration. This relationship is not observed during May and June because the total
PCB concentration tends to vary with the flow rate. The prediction interval was estimated
for a low flow condition of less than 5,000cfs and for a high flow condition greater than
5,000cfs. A greater range of PCB concentrations is evident during May and June.
Additionally, Figure 21 indicates that the prediction interval baseline varies during May
and June, and that low flow conditions result in a 100-ng/L PCB increase in the water
column. It was noted that while the estimated prediction interval value for May and June
shown is representative of a flow rate greater than 5,000 cfs, the prediction interval
baseline data point is representative for a flow rate less than 5,000 cfs. This is also
indicated in Table 2.
The estimated total PCB 95 percent UCL baseline follows the same seasonal trend as the
estimated prediction interval baseline. This relationship is presented in Figure 21. The
minimum estimated 95 percent UCL baseline concentration of approximately 45 ng/L
occurs during May and June. However, under low flow conditions, this value could
Hudson River PCB s Superfund Site
Engineering Performance Standards
19
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
increase by almost 60 ng/L. This data point is shown on Figure 21. The maximum total
PCB 95 percent UCL baseline value of 70 ng/L occurs during July and August. The 95
percent UCL baseline for total PCBs then decreases to 50 ng/L in September, increases to
65 ng/L in October, and decreases during the month of November to a total PCB
concentration of 45 ng/L. Generally, the total PCB 95 percent estimated UCL baseline
tends to correspond with the mean total PCB concentration for each month.
Schuylerville Monitoring Station
Monthly TSS data for the Schuylerville monitoring station was determined to have a
lognormal distribution for May and for the period July through November. As indicated
in Figure 22, the prediction interval TSS baseline concentration in May is approximately
7 mg/L, and increases to its maximum value of 11 mg/L during the month of June. The
estimated prediction interval baseline then decreases to a TSS concentration of
approximately 5 mg/L, where it remains for the period of July through November.
The estimated TSS 95 percent UCL baseline for Schuylerville follows the same seasonal
trend as the estimated prediction interval, as shown in Figure 22. The estimated 95
percent UCL baseline reaches a maximum TSS concentration of approximately 10 mg/L
during the month of June, and then decreases to a constant TSS concentration of 2 mg/L
for the period July through November, representative of the minimum estimated 95
percent UCL baseline TSS concentration.
Total PCB results indicate that data collected for May, June, August, September, and
November exhibit a lognormal distribution, and that the total PCBs dataset for the month
of July exhibits a non-parametric distribution. Data for the month of October exhibit a
normal data distribution.
As shown in Figure 23, both the estimated prediction interval and the 95 percent UCL
baseline for total PCBs have a maximum concentration during the months of May and
June. Both estimated total PCB baselines then fluctuate through the remainder of the
proposed dredge season, with a minimum baseline value for both baseline curves
occurring during the month of September and corresponding to a total PCB concentration
of 85 ng/ L total PCBs (prediction interval) and 60 ng/L total PCBs (95% UCL baseline).
As noted previously at other monitoring stations, the prediction interval baseline tends to
be consistent with the maximum monthly total PCB concentration. Except for the months
of May and June, the 95 percent UCL baseline tends to be consistent with the mean
monthly total PCB concentration.
Hudson River PCBs Superfund Site
Engineering Performance Standards
20
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
4.0 References
Helsel and Hirsch, 2002. Draft Techniques of Water-Resources Investigations of the
United States Geological Survey. Chapter A3 Statistical Methods in Water Resources.
Prepared by D.R. Helsel and R.M. Hirsch.
SAS, 1997. JMP Software and Reference Manual.
US Environmental Protection Agency (USEPA). 1997. Phase 2 Report, Further Site
Characterization and Analysis, Volume 2C - Data Evaluation and Interpretation Report
(DEIR), Hudson River PCBs RI/FS. Prepared for USEPA Region 2 and USACE by
TAMS Consultants, Inc., the Cadmus Group, Inc., and Gradient Corporation. February
1997.
USEPA. 1998. 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. Prepared for USEPA Region 2
and USACE, Kansas City District by TAMS Consultants, Limno-Tech, Inc, Menzie-Cura
& Associates, Inc., and TetraTech, Inc. December 1998.
USEPA, 2000. Phase 3 Report: Feasibility Study, Hudson River PCBs Reassessment
RI/FS. Prepared for EPA Region 2 and the US Army Corps of Engineers (USACE),
Kansas City District by TAMS Consultants, Inc. December 2000.
USEPA, 2001. Pro-UCL User's Manual
Hudson River PCBs Superfund Site
Engineering Performance Standards
21
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Tables
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
Table 1. PCB versus Flow Correlation Analysis Based on the Fit Curve Generated from Plot
May and June at TID W
May and June Low Flow (<5000 cfs) at
TID PRW2
May and June at Schuylerville
Fit curve
Y = 283.23 - 0.026946X
Y= 186.82-0.030192X
Y= 176.19-0.012506X
Lower 95% Confidence Limit
Y = 246.5 - 0.015*x - 1.51 E-6*xA2
Y = 144 - 8.73E-3*x - 3.56E-6*xA2
Y= 151.16-6.97E-3*x-4.93E-7*xA2
Upper 95% Confidence Limit
Y = 386.95 - 0.0474*x + 1.51 E-6*xA2
Y = 229.64 - 5.17E-2*x + 3.56E-6*xA2
Y = 201.22 - 1.80E-2*x +4.93E-7*xA2
Upper 95% Individual Limit
Y = 522.19 - 0.0342*x + 2.85E-7*xA2
Y = 242.14 - 3.72E-2*x + 1,18E-6*xA2
Y = 234-0.0138*x+ 1.16E-7*xA2
Notes:
Y = PCB concentration
X = Flow (cfs)
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Table 2
Statistics Results and Baseline Level of TSS and PCB Concentration
at Upper Hudson River Monitoring Stations
Parameter
Fort Edward
TSS (mg/L)
PCB (ng/L)
May
June
July
August
Sept thru Nov
May
June
July thru
Sept.
Oct. & Nov.
n
17
22
21
20
60
25
30
79
48
Minimum Detected
0.5
0.5
0.5
0.5
0.5
5.50
5.50
5.50
5.50
Maximum Detected
4.1
16
4.1
9.3
3.3
18.31
31.41
55.51
21.88
Arithmetic Mean
2
3
2
2
2
9
13
13
8
Standard Deviation
1
3
1
2
1
5
8
11
4
Median
1.9
2.2
2.2
1.95
1.6
5.5
14
12
6
W-Test (n<=50)
W
0.920
0.429
0.936
0.648
0.657
0.862
0.531
W-LN
0.872
0.783
0.825
0.927
0.641
0.829
0.535
Critical W
0.892
0.911
0.908
0.905
0.918
0.927
0.947
D'Agostino's Test (n>50)
Y
-1.79
-19.20
-0.20
-10.67
-0.25
-7.33
-1.49
-12.20
-19.66
Yin
-2.19
-8.70
-2.69
-3.16
-1.91
-7.33
-1.15
-3.82
-18.28
UCL95%
y y
5.7
2.4
3.1
1.8
12.7
19.7
18.6
10.4
UCL 95% Lognormal
2.6
3.6
3.0
3.1
1.9
10.3
17.3
15.5
8.3
UCL 95% Normal
2.2
4.0
2.4
3.1
1.8
10.2
15.8
15.4
8.6
LCL 95%
1.4
2.2
1.7
1.9
1.4
6.9
10.9
11.6
6.7
LCL 95% Log normal
1.4
2.2
1.7
1.9
1.4
7.2
11.2
11.6
6.7
LCL 95% Normal
1.4
1.8
1.7
1.7
1.4
6.9
10.9
11.5
6.5
Data Distribution (Normal,
Lognormal or non-parametric)
Normal
non-
parametric
Normal
Lognormal
Normal
non-
parametric
non-
parametric
non-
parametric
non-
parametric
95th percentile
0.5
1.6
3.2
3.7
3.1
16.9
27.7
34.3
19.1
Prediction Interval (Normal)
3.4
8.2
3.9
5.6
3.0
16.8
27.1
31.1
15.1
Prediction Interval (LogNormal)
4.6
6.5
5.8
5.6
3.9
17.5
33.1
32.9
14.0
Prediction interval
3.4
4.2
3.9
5.6
3.0
16.9
O "7 "7
Z. ( . 1
34.3
19.1
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Table 2 (cont'd)
Statistics Results and Baseline Level of TSS and PCB Concentration
at Upper Hudson River Monitoring Stations
Parameter
TID West
TSS (mg/L)
PCB (ng/L)
May
June
July thru Oct.
Nov.
May
June
July
August
Sept.
Oct. & Nov.
n
17
24
90
22
24
32
30
29
27
54
Minimum Detected
1.20
1.40
0.50
0.50
24.5
60.1
65.52
49.02
40.00
25.82
Maximum Detected
26.00
36.00
6.50
6.70
813.6
413.4
219.45
164.00
126.25
1424.00
Arithmetic Mean
4
5
2
2
127.6
169.1
138
96
75
127
Standard Deviation
7
7
1
1
160.3
85.8
43
27
22
193
Median
2
3
1
2
81.0
156.5
135
92
73
88
W-Test (n<=50)
W
0.514
0.454
0.892
0.6
0.9
0.961
0.931
0.962
W-LN
0.780
0.823
0.930
1.0
0.9
0.943
0.973
0.980
Critical W
0.892
0.916
0.911
0.927
0.926
0.923
D'Agostino's Test (n>50)
Y
-13.07
-18.63
-11.94
-2.89
-14.2
-0.8
0.76
-2.38
-0.42
-34.51
Yin
-4.84
-5.48
-2.12
-1.37
-0.7
0.8
0.10
-1.35
0.18
-8.09
UCL95%
11.5
11.5
1.9
3.3
181.3
205.3
150.9
105.8
83.1
241.4
UCL 95% Lognormal
6.6
6.2
1.9
3.3
181.3
205.3
154.9
105.8
83.1
134.8
UCL 95% Normal
7.2
7.5
1.8
2.9
183.6
194.8
150.9
104.9
81.9
170.9
LCL 95%
2.6
3.4
1.5
1.9
124.3
88.8
68.2
97.7
LCL 95% Lognormal
2.6
3.4
1.5
1.9
90.5
146.0
124.9
88.8
68.2
97.7
LCL 95% Normal
1.6
2.4
1.4
1.8
71.5
143.4
124.3
88.0
67.6
82.8
Data Distribution (Normal,
Lognormal or non-parametric)
non-
parametric
non-
parametric
Lognormal
Lognormal
Normal
Lognormal
Lognormal
non-
parametric
95th percentile
18.8
15.5
3.6
4.3
264.1
280.6
202.2
151.1
113.7
297.4
Prediction Interval (Normal)
16.4
17.8
3.5
4.9
407.9
316.8
211.6
142.7
112.3
453.4
Prediction Interval (LogNormal)
12.6
12.2
3.9
6.4
367.8
367.8
368.3
368.3
233.3
148.7
119.2
272.1
Prediction interval
18.8
15.5
3.9
6.4
211.6
148.7
119.2
297.4
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Table 2 (cont'd)
Statistics Results and Baseline Level of TSS and PCB Concentration
at Upper Hudson River Monitoring Stations
Parameter
TID PRW
TSS (mg/L)
PCB (ng
/L)
May
June
July thru
Nov.
May&June
Low Flow
(<5000 cfs)
May&June
High Flow
(>5000 cfs)
July and
August
Sept.
Oct.
Nov.
n
14
13
75
19.0
21
40
19
23
20
Minimum Detected
0.50
1.80
0.50
32.0
15.58
28.30
26.20
23.24
20.00
Maximum Detected
24.80
29.50
6.60
166.4
67.05
141.76
65.44
93.26
64.28
Arithmetic Mean
4
5
2
96.8
42
65
44
57
40
Standard Deviation
6
7
1
35.8
15
21
13
20
14
Median
2
3
2
107.1
41
62
44
55
39
W-Test (n<=50)
3.707150762
W
0.468
0.434
1.0
0.968
0.936
0.929
0.970
0.943
W-LN
0.896
0.729
0.9
0.914
0.992
0.934
0.937
0.924
Critical W
0.874
0.866
0.908
0.940
0.901
0.914
0.905
D'Agostino's Test (n>50)
Y
-13.66
-13.99
-10.21
0.2
0.14
-2.85
0.43
0.32
0.50
Yin
-3.26
-6.17
-1.73
-0.9
-1.50
-0.41
0.58
-0.79
0.12
UCL 95%
6.5
14.0
2.2
111.1
47.1
70.9
50.1
64.2
45.4
UCL 95% Lognormal
6.5
7.4
2.2
118.9
50.2
70.9
50.1
67.3
47.5
UCL 95% Normal
6.7
8.7
2.1
111.1
47.1
70.3
48.9
64.2
45.4
LCL 95%
2.2
3.2
1.6
36.0
59.4
39.1
50.2
34.4
LCL 95% Log normal
2.2
3.2
1.6
83.5
36.2
59.4
39.1
50.6
34.9
LCL 95% Normal
0.9
1.4
1.6
82.6
36.0
58.8
38.6
50.2
34.4
Data Distribution (Normal,
Lognormal or non-parametric)
Lognormal
non-parametric
Lognormal
Normal
Lognormal
Lognormal
Normal
Normal
95th percentile
12.0
15.0
4.5
148.1
64.0
93.5
64.0
86.3
61.4
Prediction Interval (Normal)
15.1
18.8
4.1
160.5
67.6
101.2
66.7
91.6
65.0
Prediction Interval (LogNormal)
11.7
13.1
4.6
189.6
80.2
106.4
71.8
104.9
73.5
Prediction interval
11.7
15.0
4.6
160.5
67.6
106.4
71.8
91.6
65.0
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 3 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Table 2 (cont'd)
Statistics Results and Baseline Level of TSS and PCB Concentration
at Upper Hudson River Monitoring Stations
Parameter
Schuylerville
TSS (mg/L)
PCB (ng/L)
May
June
July thru
Nov.
May and
June
July
August
Sept.
Oct.
Nov.
n
10
12
74
34.0
19
21
17
23
22
Minimum Detected
1.60
2.00
0.50
43.0
61.00
50.18
26.30
34.94
38.94
Maximum Detected
8.00
17.50
7.80
211.3
157.18
107.00
78.22
111.64
105.25
Arithmetic Mean
3
5
2
106.5
82
74
52
75
67
Standard Deviation
2
4
1
41.7
20
17
15
24
20
Median
3
3
2
94.9
81
71
49
75
63
W-Test (n<=50)
W
0.739
0.548
0.9
0.694
0.953
0.948
0.936
0.933
W-LN
0.909
0.813
1.0
0.830
0.971
0.955
0.881
0.965
Critical W
0.842
0.859
0.901
0.908
0.892
0.914
0.911
D'Agostino's Test (n>50)
Y
-5.08
-10.41
-12.01
-0.5
-9.00
0.04
0.10
0.13
-0.56
Yin
-1.63
-4.31
-1.52
0.6
-5.09
0.59
-0.10
-1.48
0.24
UCL 95%
4.4
9.9
2.2
121.3
102.7
80.6
60.1
83.8
75.2
UCL 95% Lognormal
4.4
6.5
2.2
121.3
89.5
80.6
60.1
88.0
75.2
UCL 95% Normal
4.3
6.8
2.1
118.6
90.3
79.9
58.5
83.8
74.1
LCL 95%
2.5
3.4
1.6
75.8
67.9
46.4
66.5
60.4
LCL 95% Log normal
2.5
3.4
1.6
95.5
75.8
67.9
46.4
66.8
60.4
LCL 95% Normal
2.1
2.5
1.6
94.4
74.0
67.4
45.8
66.5
59.6
Data Distribution (Normal,
Lognormal or non-parametric)
Lognormal
non-parametric
Lognormal
Lognormal
non-parametric
Lognormal
Lognormal
Normal
Lognormal
95th percentile
6.1
10.8
4.4
175.9
98.7
105.0
73.7
108.2
40.0
Prediction Interval (Normal)
6.7
12.4
4.2
178.1
118.7
103.1
79.1
117.6
101.6
Prediction Interval (LogNormal)
7.0
10.8
4.7
194.6
115.9
106.7
85.5
135.7
107.2
Prediction interval
7.0
10.8
4.7
194.6
98.7
106.7
85.5
117.6
107.2
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 4 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Figures
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
Federal
Dam
1
- 2
3
4
5
6
7
8
Hudson River at Fort Edward
Hudson River at Stillwater
Hudson River above Lock 1 near Waterford
Glowegee Creek at West Milton
Kayaderosseras Creek near West Milton
Hoosic River near Eagle Bridge
Mohawk River Diversion at Crescent Dam
Mohawk River at Cohoes
/ 7^ ^
10 km
Figure 1. Upper Hudson River Basin USGS Flow Gage Stations
Used in HUDTOX Modeling.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Daily Runoff Yield
1998-2001
Stillwater
(cfs/mi2)
Figure 2. Stillwater versus Ft. Edward Daily Runoff Yield 1998-2001
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Figure 3. Fort Edward Station Monthly TSS Concentration Variation
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
Figure 4. Fort Edward Station Monthly PCB Concentration Variation
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
Hudson River PCBs Superfund Site
Engineering Performance Standards
Figure 5. TTD-West Station Monthly TSS Concentration Variation
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
1600
1400 -
1200 -
e/j
ฃ
B 1000
"cs
-
800 -
=
u
=
o
U
pa
600
o
H
400 H
200 -
~
~
I i
t i
~~r~
6
r~
9
I
10
12
Month
Figure 6. TID-West Station Monthly Total PCB Concentration Variation
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
Hudson River PCBs Superfund Site
Engineering Performance Standards
Figure 7. TID-PRW Station Monthly TSS Concentration Variation
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Figure 8. TID-PRW Station Monthly Total PCB Concentration Variation
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
Hudson River PCBs Superfund Site
Engineering Performance Standards
Figure 9. Schuylerville Station Monthly TSS Concentration Variation
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Figure 10. Schuylerville Station Monthly Total PCB Concentration Variation
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment A - April 2004
-------
20 -i
18 -
16 -
14
12 -
10
2 -
0
Monthly TSS Concentration at the Ncliin ler\ ille Station
~ TSS Monthly Data
-ฆMean Monthly TSS Concentration
10
11
12
Month
250
J 200
2 150 -
s 100 -\
0
U
n
01 50
Monthly PCB Concentration at the Schuylerville Station
~ Monthly PCB Data
Mean Monthly PCB Concentration
10
11
12
Month
Figure 11. Schuylerville Monitoring Station Monthly TSS and PCB Concentrations Plotted Against the Monthly Mean
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
18
o
CN
CO
CO
ฆ J,
I. g
= Similar Data
10
Month
Student's t
0.05
Tukey-Kramer
0.05
110 -
m
o
Q.
J ฆ
: o
T a g
= Similar Data
g
ฆ ฆ
i u 3 a i
+ LjJ R
ง
8
10
11
Month
Student's t
0.05
Tukey-Kramer
0.05
Figure 12. Schuylerville Station Box Plots
TSS Concentration vs. Month (Top Diagram)
Total PCB Concentration vs. Month (Bottom Diagram)
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Figure 13: TID-West Monitoring Station
Flow verus Total PCB Concentration
Months of May and June
Figure 14: TID-PRW Monitoring Station
Flow versus Total PCB Concentration
Months of May and June
Flow_schuy
Figure 15. Schuylerville Monitoring Station
Flow versus Total PCB Concentration
Months of May and June
Units: Flow-cfs, PCB-ng.L
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
18
16 -
~ Monthly TSS Data
"^H95% UCL Baseline
-A Prediction Interval Baseline
14
J
~Hii
12
10 -
U
0J
U
C
o
U
ifi
in
H
2 -
~
t
10
11
Month
Figure 16. Fort Edward Monitoring Station Monthly TSS Data versus Estimated TSS Baselines
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
60
55 -
50 -
~ Monthly PCB Data
*^H95% UCL Baseline
Prediction Interval Baseline
45 -
40 -
3
~6d
"a 35
o
=
o
y
CO
30 -
25
o
H
20
15
10
I
t
i
~
~
~
10
11
12
Month
Figure 17. Fort Edward Monitoring Station Monthly Total PCB Data versus Estimated Total PCB Baselines
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
4 5 6 7 8 9 10 11 12
Month
Figure 18. TID-West Monitoring Station Monthly TSS Data versus Estimated TSS Baselines
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
~Hii
o>
u
1500
1400 -
1300 -
1200 -
1100 -
1000 -
900 -
800 -
0 700
PQ
U 600
CLh
1 500
H
~ Monthly Total PCB Data
95% UCL Baseline
-A Prediction Interval
Baseline
Month
Figure 19. TID-West Monitoring Station Monthly Total PCB Data versus Estimated Total PCB Baselines
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
35
30 -
~ Monthly TSS Data
95% UCL Baseline
-A Prediction Interval Baseline
25 -
J
~Hii
20
ea
o>
u
e
o
U
(Z3
(Z3
H
15 -
10
~
~
~
-A
T
t
t
1
10
11
12
Month
Figure 20. TID-PRW Monitoring Station Monthly TSS Data versus Estimated TSS Baselines
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
180
160 -
140
~
~
~
~
Total [PCB] varies with flow during
May and June; These estimated
baseline data points represent low flow
conditions. ~
~ Monthly Total PCB Data
-*95% UCL Baseline
""^-Prediction Interval
mi
es
o>
o
e
o
U
CO
u
a.
o
H
120
100 -
80 -
60
40
20 -
r~
9
10
11
12
Month
Figure 21. TID-PRW Monitoring Station Monthly Total PCB Data versus Estimated Total PCB Baselines
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
20
18 -
16
14
J
~Hii
S 12
~ Monthly TSS Data
ฆ^H95% UCL Baseline
~ Prediction Interval Baseline
CG
h 10
o>
u
o
U
(Z3
(Z3
H
2 -
~
~
~
~
I
!
~
I
T
4
I
10
11
12
Month
Figure 22. Schuylerville Monitoring Station Monthly TSS Data versus Estimated TSS Baselines
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
250
200
mi
_c,
=
"-C
-
=
w
C
o
U
CQ
U
0.
o
H
150
100
50
~
t
~
~
~
*
~
~ Total PCB Monthly Data
95% UCL Baseline
-A Prediction Interval Baseline
10
11
12
Month
Figure 23. Schuylerville Monitoring Station Monthly Total PCB Data versus Estimated Total PCB Baselines
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment A - April 2004
-------
Attachment B
Resuspension Sensitivity
Table of Contents
1.0 Objective 1
2.0 Methodol ogy 2
3.0 Discussion 4
4.0 Results 6
4.1 TID Monitoring Locations 6
4.2 Increases In Total PCBs Average Concentrations Due to Dredging 6
4.3 Increases In Total PCBs Single Sample Concentrations Due to Dredging 7
5.0 Comparison of the Annual Dredging Induced PCB Load to the Baseline PCB
Load 9
6.0 References 11
LIST OF TABLES
Table 1 Volume of Sediment Removed by Dredging Season
Table 2 Estimated Tri+ and Total PCB Mass to be Remediated
Table 3 Suspended Solids Estimated Increase to the Water Column
Table 4 Total PCBs Estimated Increase to the Water Column
Table 5 Estimated Total PCB Concentrations Compared to the 95 Percent UCL
Baseline Data at the TID-West Monitoring Station Assuming a 300
g/day Total PCB Release Rate
Table 6 Estimated Total PCB Concentrations Compared to the 95 Percent UCL
Baseline Data at the TID-PRW2 Monitoring Station Assuming a 300
g/day Total PCB Release Rate
Table 7 Estimated Total PCB Concentrations Compared to the 95 Percent UCL
Baseline Data at the Schuylerville Monitoring Station Assuming a 300
g/day Total PCB Release Rate
Table 8 Estimated Total PCB Concentrations Compared to the 95 Percent UCL
Baseline Data at the TID-West Monitoring Station Assuming a 600
g/day Total PCB Release Rate
Table 9 Estimated Total PCB Concentrations Compared to the 95 Percent UCL
Baseline Data at the TID-PRW2 Monitoring Station Assuming a 600
g/day Total PCB Release Rate
Table 10 Estimated Total PCB Concentrations Compared to the 95 Percent UCL
Baseline Data at the Schuylerville Monitoring Station Assuming a 600
g/day Total PCB Release Rate
Hudson River PCBs Superfund Site
Engineering Performance Standards
l
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Attachment B
Resuspension Sensitivity
Table of Contents
LIST OF TABLES (continued)
Table 11 Estimated Total PCB Concentrations Compared to the Prediction
Interval Baseline Data at the TID-West Monitoring Station Assuming a
300 g/day Total PCB Release Rate
Table 12 Estimated Total PCB Concentrations Compared to the Prediction
Interval Baseline Data at the TID-PRW2 Monitoring Station Assuming a
300 g/day Total PCB Release Rate
Table 13 Estimated Total PCB Concentrations Compared to the Prediction
Interval Baseline Data at the Schuylerville Monitoring Station Assuming
a 300 g/day Total PCB Release Rate
Table 14 Estimated Total PCB Concentrations Compared to the Prediction
Interval Baseline Data at the TID-West Monitoring Station Assuming a
600 g/day Total PCB Release Rate
Table 15 Estimated Total PCB Concentrations Compared to the Prediction
Interval Baseline Data at the TID-PRW2 Monitoring Station Assuming a
600 g/day Total PCB Release Rate
Table 16 Estimated Total PCB Concentrations Compared to the Prediction
Interval Baseline Data at the Schuylerville Monitoring Station Assuming
a 600 g/day Total PCB Release Rate
Table 17 Calculation of the Annual Dredging Induced PCB Load for the Fully
Exhausted Standard (500 ng/L)
Table 18 Calculation of the Annual Dredging Induced PCB Load for the 300 and
600 g/day Total PCB Mass Loss Control Limits
Table 19 Dredging Induced Loss - Percent of the Baseline Annual Load
LIST OF FIGURES
Figure 1 TID-West Monitoring Station - 95 Percent UCL - Total PCB
Figure 2 TID-PRW2 Monitoring Station - 95 Percent UCL - Total PCB
Figure 3 Schuylerville Monitoring Station - 95 Percent UCL - Total PCB
Figure 4 TID-West Monitoring Station - Single Incident - Total PCB
Figure 5 TID-PRW2 Monitoring Station - Single Incident - Total PCB
Figure 6 Schuylerville Monitoring Station - Single Incident - Total PCB
Figure 7 Water Column Total PCB Load at Fort Edward, TID West and
Schuylerville Compared to Estimated Dredging Induced Total PCB
Load
Hudson River PCB s Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
1.0 Objective
Attachment B
Resuspension Sensitivity
Baseline levels of PCBs in the water column fluctuate due to seasonal variables and
heterogeneous sources. Therefore it is essential determine the dredging-related PCB
releases as a function of time and flow that are detectable above the baseline variations.
Furthermore, if data from water samples collected during dredge operations indicate that
the PCB concentration transported downstream is within the baseline variation, then it is
unlikely that the downstream concentrations will be noticeably impacted from dredging.
Furthermore, the resuspension criteria must be set above the baseline variation in order to
avoid false exceedances and unnecessary encumbrances to the dredging operations. This
monitoring analysis involves the statistical range of baseline variations in total PCB water
column concentrations (formulated in Attachment A) and the ability to identify a
"significant increase" in the running averages that would signal an unacceptable
dredging-related release {i.e., exceedance of resuspension criterion) and require
engineering contingencies. Historic data from the Thompson Island Dam (TID) and
Schuylerville were used in this analysis, however the baseline and sensitivity calculations
should be revised based on the results of the Baseline Monitoring Program. The 95
percent UCL calculations were analyzed for the all the resuspension criteria since they
are based on running averages. The prediction limits are also provided, however, the
prediction limit analyses indicate the likelihood that any given sample may exceed the
criteria and does not apply to running averages. Assuming operations continued at the
various criteria, the overall increases in loads within a dredging season were also
examined.
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
2.0 Methodology
During remediation, water column monitoring will be implemented at far-field stations
down-gradient of the work areas. Since the river system has baseline PCB levels, it is
necessary to confirm that exceedances of the resuspension criteria are recognizable above
the inherent variations around the baseline. If exceedances of the criteria were not
discernible from the baseline variations, then either PCB levels of concern would not be
detected or false exceedances could occur. To this end, an analysis was performed over a
wide range of river flow rates (2,000 through 10,000 cfs) and dredging-induced
resuspension PCB release rates (300 and 600 g/day), taking into account the variations in
the baseline water column concentration (discussed in Attachment A of this report).
The total PCB increases due to dredging activities are based on the volume of sediment
removed during each dredging season, the percent solids loss to the water column due to
dredging activities, and the river discharge rate. These components are described as
follows:
ASS=F-"XPX'ฐSS
Qxtd
x 9.07 x10s
(1)
where:
ASS
Vsed
P
loss
0
td
9.07xl08
SS increase in water column (mg/L)
volume of sediment to be removed (cy)
density of the sediment (tons/cy)
dredging-induced resuspension loss rate (%
flow rate (L/s)
length of dredging season (s)
conversion factor from tons to mg
The estimated volume of sediment to be removed with overcut, as estimated in the
Feasibility Study (USEPA, 2000), is 2.6xl06 cy. The dredging season is scheduled to
occur from May 1 through November 30. Table 1 summarizes the estimated volume of
sediment removal for each dredging season and the density of the sediment for each river
section.
The total PCB increase in the water column due to dredging was calculated as follows:
(2)
atpcb = Mjpcj?x1qssxio12
Qxtd
where: ATPCB = TPCB increase in water column (ng/L)
= mass of total PCB remediated (kg)
= factor to convert kilograms to nanograms
and other parameters are defined above.
Mjpcb
1012
Hudson River PCB s Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
The estimated mass of Tri+ and total PCBs to be remediated are summarized in Table 2.
The total PCB concentrations calculated for velocities of 2,000 cfs and 8,000 cfs,
assuming 300 g/day and 600 g/day release rates and the 95 percent UCL and prediction
interval baseline conditions, are presented in this analysis. These flow rates were selected
based on historical flow data recorded during months in which dredging is antivipaed to
occur {i.e. the dredging season months). Thus, at these two flow rates, the range of SS
and total PCB conditions that will exist in the Hudson River during dredging operations
were estimated. It should be noted that dredging activities are not expected to occur at
Fort Edward flow rates as high as 8,000 cfs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
3.0 Discussion
As shown in the relationships demonstrated by Equations 1 and 2, the estimated total
PCB concentration increase in the water column is a function of two things: the river
flow rate and the solids loss rate from dredging. The estimated SS and total PCB
increases as a result of 0.5 percent and 1 percent solids releases are shown in Tables 3
and 4. The 0.5 and 1 percent solids releases are equivalent to loss rates of 0.21 and 0.42
kg/s of solids, and correspond to 300 and 600 g/day total PCB releases, respectively. Data
indicate that the increase in SS and PCB concentrations for a given loss rate is greatest
under low flow conditions.
In order to ensure that the resuspension criteria are discernible from the baseline
variations, a sensitivity analysis was performed. The sensitivity analysis was performed
for the following:
The baseline total PCB concentrations were compared with the estimated
increases from dredging for total PCB release rates of 300 and 600 g/day and
varying flow rates.
The estimated total PCB water column concentrations during dredging operations
associated with these release rates were computed by adding the estimated
concentration increases (shown in Table 4) to the 95 percent upper confidence
limit (UCL) baseline concentrations and the 95th percentile prediction interval
baseline concentrations.
The dredging related releases were superimposed onto the 95th percent UCL
baseline to provide a table of conditions (dependent on flow and season), which
can be compared to the running averages in order to discern if an exceedance is
due to dredging operations.
The 95 percent UCL baseline data approximates the baseline variability of the total
PCBs, and can be compared with resuspension criteria based on running averages. The
prediction interval baseline data approximates the upper bound baseline concentration for
one sampling incident, and can be compared with total PCB data collected from a single
sample or incident during dredging activities to allow for the detection of a sudden
increase or a change in river conditions. This method is only applicable to criteria that do
not involve multiple samples, so it is not directly relevant to the current resuspension
criteria.
This analysis was completed for three far field monitoring stations (Thompson Island
Dam-West (TID-West), TID-PRW2, and Schuylerville) over the proposed dredging
period (May through November) using historic data. New data collected during the
Baseline Monitoring Program will provide a better estimate of the baseline level at the
far-field monitoring stations.
The total PCB release rate of 300 g/day represents the lowest significantly detectable
PCB concentration increase when added to the monthly baseline conditions. An analysis
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
(based on the GE dataset for 1996-2000) of the annual PCB loading and 600 g/day total
PCB release rate in the water column indicated the following:
That a 600 g/day total PCB release rate due to dredging corresponds to
approximately two standard deviations of the annual PCB loading of the river.
That a 600 g/day total PCB release rate due to dredging corresponds to a
dredging-induced PCB loading of approximately 130 kg per year.
It was also determined that the standard deviation for the annual PCB loading, based on
existing GE water column data for the period 1996 to 2000, is approximately 70 kg total
PCBs per year. Thus, a total PCB release rate greater than 600 g/day is likely to exceed
the river system's annual baseline PCB loading, supporting the use of the 600 g/day
release rate as an upper bound for PCB loading.
As a result, it was recommended that engineering evaluations and solutions be
implemented when dredging releases approach 300 g/day total PCBs and it is mandatory
that engineering evaluations and solutions be implemented for instances when dredging
releases are greater than the river's baseline variation {i.e. 600 g/day total PCB).
Ultimately, PCB loading corresponding to 300 and 600 g/day, combined with the results
of this sensitivity analysis (described herein) were utilized to design a tiered,
resuspension monitoring plan comprised of different action levels and monitoring
requirements. These levels of monitoring will be implemented based on measured PCB
concentrations and corresponding PCB loading estimates.
Additional criteria are based on SS, but the goal of the SS-based criteria is determine net
dredging contributions, rendering baseline sensitivity analyses unnecessary. The
monitoring programs for SS are described in Chapter 3 and Attachment F of this report.
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
4.0 Results
The following sections present the results of the sensitivity analysis and a discussion of
estimated total PCB concentrations. The results presented assume the following:
Variable flow rates
Estimated baseline concentrations
Total PCB release rates of 300 and 600 g/day.
The baseline conditions are examined at three monitoring stations, two at the TID (TID-
West and TID-PRW2) and one at Schuylerville.
4.1 TID Monitoring Locations
Both TID-West and TID-PRW2 are located at the TID. As explained in Attachment A of
this report, both of these stations have limitations associated with their data. The total
PCB concentrations for TID-West were examined in the Responsiveness Summary for the
Data Evaluation and Interpretation Report (DEIR) (USEPA, 1998). This analysis
concluded that samples collected at the TID-West station are influenced by nearby
sediment during low flows. It was also noted in the DEIR that samples collected at TID-
PRW2 tend to be limited to the warmer months due to inaccessibility in the winter. Thus,
it is thought that the results presented herein may not represent actual water column
background conditions, and that adjustments to the location of the sampling station and
sample collection in the years prior to dredging will provide a new baseline that is more
appropriate. The following data, therefore, are representative of the best data that exist to
date, though limitations and concerns with the data are apparent.
4.2 Increases in Total PCBs Average Concentrations Due to Dredging
As stated above, the PCB increases from dredging were estimated for PCB release rates
of 300 and 600 g/day for flow rates ranging from 2,000 to 10,000 cfs. The 95 percent
UCL baseline results for a total PCB release rate of 300 g/day are shown in Tables 5
through 7, and the results for a release rate of 600 g/day in Tables 8 through 10. Data for
both release rates at all three monitoring stations are included. The estimated PCB
concentration increases at 2,000 cfs and 8,000 cfs were added to the 95 percent UCL
baseline conditions and shown in Figures 1 to 3 for TID-West, TID-PRW2 and
Schuylerville respectively.
As depicted in Figures 1 through 3, the PCB concentrations are generally highest during
the months of May and June, except for TID-PRW2, which also has high concentrations
in October and November. The increases from dredging are more difficult to discern from
baseline levels at higher flows, since the concentration increases are less than those at
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
lower flows. In general the concentrations for these release rates are sufficiently above
baseline to be discernable (at 8,000 cfs a release rate of 300 g/day increases the baseline
concentration by more than 20 ng/L). In particular, TID-PRW2 and Schuylerville have
fairly consistent total PCB concentrations from these releases at any given flow. However
concentrations associated with these total PCB loads will have large variations with flow,
making accurate flow rate measurements a necessity.
Due to the dependence of the load criteria on flow rate measurements, a second criterion
for total PCBs of 350 ng/L is applied to same action level as the 600 g/day (the Control
level). For TID-PRW2 and Schuylerville, this concentration is slightly higher than the
600 g/day PCB release rate and 95 percent UCL baseline concentration estimates. For
TID-West, the concentrations for the 600 g/day release rates in May, June, October, and
November and the 300 g/day release rate for October and November are estimated to be
above the 350 ng/L criteria, assuming the 95 percent UCL baseline. This indicates that at
low flows during these months, dredging in areas with high concentrations may require
additional precautions to prevent dredging-related PCB releases from causing
exceedances of the 350 ng/L criterion.
None of the concentrations estimated using the 300 g/day or 600 g/day loads at the 95th
percentile UCL baselines are greater than the Resuspension Standard of 500 ng/L.
However, since an exceedance of the Resuspension Standard only requires a confirmed
occurrence, it is useful to compare the standard to the 95th prediction limits for the
baseline with the 300 g/day and 600 g/day total PCB loads superimposed.
4.3 Increases in Total PCBs Single Sample Concentrations Due to
Dredging
In order to examine the sensitivity of a single sampling incident, the prediction interval
baseline results were applied for total PCB release rates of 300 g/day (Tables 11 to 13)
and 600 g/day (Tables 14 through 16) for TID-West, TID-PRW2 and Schuylerville
respectively. The estimated PCB concentration increases at 2,000 cfs and 8,000 cfs were
added to the prediction interval baseline conditions and shown in Figures 4 through 6 for
TID-West, TID-PRW2 and Schuylerville, respectively.
The PCB increases and prediction level baseline conditions for the 600 g/day total PCB
release rate at 2,000 cfs shown in Figures 5 and 6 are below the USEPA Safe Drinking
Water Act Maximum Contaminant Level (MCL) of 500 ng/L for TID-PRW2 and
Schuylerville. However, for the analysis at TID-West, this 600 g/day total PCB release
rate at 2,000 cfs exceeds 500 ng/L when added to the prediction level baseline for May,
June, October, and November. However, the final monitoring station at the TID is
expected have baseline conditions that are similar to a combination of those at TID-West
and TID-PRW2. Therefore, the results from TID-West station alone are not expected to
be truly representative of the PCB concentrations at the TID. Furthermore, an exceedance
of the Resuspension Standard threshold requires the collection of four additional samples
(in one day) to be analyzed with expedited turn-around times. Therefore, the final
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
decision to cease operations will be based on at least 5 samples. Since the prediction limit
shown represents a 5 percent chance of having one sample exceed the 500 ng/L criterion,
the likelihood of 5 samples exceeding the 500 ng/L criterion will be lower. However,
these results imply that in order to be conservative, dredging operations during these
months at low flows may require additional precautions to prevent dredging-related PCBs
from causing exceedances of the Resuspension Standard.
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
5.0 Comparison of the Annual Dredging Induced PCB Load to the
Baseline PCB Load
Further analyses were performed to compare the annual baseline total PCB loads with
the average annual total PCB loads resulting from solids releases of 0.21 kg/s and 0.42
kg/s, which are associated with the resuspension release criteria of 300 g/day and 600
g/day. The analysis assumed that these solids releases were consistently maintained
throughout the dredging period. In addition, the annual loads associated with the
Resuspension Standard of 500 ng/L were also examined.
Results and Discussion
The annual load, assuming that dredging operations continued with a far-field
concentration of 500 ng/L throughout the dredging season (though it should be noted
operations would not continue at this level), was calculated using the United States
Geological Survey (USGS) daily discharge rates averaged by month at Fort Edward. The
estimated loads are shown in Table 17. For these loads, it was assumed that the work will
occur six days per week and that the increase in concentration occurs only during the 14-
hour-a-day working period. The 0.5 and 1 percent solids releases are equivalent to loss
rates of 0.21 and 0.42 kg/s of solids, and correspond to the 300 and 600 g/day total PCB
release rates, respectively. The annual total PCB loads associated with these release rates
were calculated, taking into account the dredging schedule proposed in the FS (USEPA,
2000) and the average concentration in each river section. The estimated loads are shown
in Table 18.
The annual total PCB loads for 1992 through 2000 were calculated using the GE water
column monitoring data and the USGS daily discharge estimates. The TID total PCB
concentrations were adjusted for the TID-West bias according to the method described in
the Responsiveness Summary to the DEIR (USEPA, 1998). At each station the daily load
was calculated and the values were averaged within their respective months to get a
monthly average. This average, along with the number of days within the each month,
provided the monthly load. The monthly loads were then summed to determine the
annual loads at each station. The average annual total PCB loads from 1992 to 2000 are
shown in Table 19.
The annual loads from 1992-2000 from above Rogers Island, the TI Pool, and the stretch
of river between the TID and the Schuylerville station are presented in Figure 7. The high
concentrations detected in 1992 (which gradually declined) were the result of the Allen
Mills failure. Controls put in place by the end of 1996 have reduced the seepage of dense
non-aqueous phase liquid (DNAPL) into the Hudson River at the GE Hudson River Falls
site. The DNAPL leakage is shown as the load at Fort Edward. The load for the
Thompson Island (TI) Pool (Rogers Island to the TID) also decreased from the levels
detected in 1992 - 1994, with the loads varying year to year between 1995 and 2000. The
loads at Schuylerville are substantially less than the upstream loads, though data were
available only for the years spanning 1998 - 2000.
Hudson River PCB s Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Calculations presented in Attachment D of this Report, indicate that the best engineering
estimate of the TSS fraction released from dredging would not exceed 0.13 percent. This
loss rate represents approximately 110 kg of Total PCBs released throughout the entire
dredging project. Assuming the same schedule presented in the FS, this amounts to an
average of approximately 105 g/day (ranging from 78 to 209 g/day for the various river
sections). This loss rate is less than half of that estimated using the lower resuspension
criteria of 300 g/day total PCBs (i.e. the 300 g/day total PCB loss rate is over twice what
is anticipated under normal dredging conditions), allowing for additional resuspension
and mass loss resulting from the other components of the remediation, such as vehicle
traffic, without exceeding the criteria. A well-controlled remediation of the Hudson River
should not result in a mass loss in excess of the lower resuspension PCB load criteria;
specifically, that less than 65 kg per year will be released to the river as a result of the
remediation. The 65 kg/year of total PCBs is a small fraction of the baseline load to the
river in most years, as shown in Table 19. A loss of 65 kg/yr represents less that 20
percent of the annual load for six of the nine years with load estimates.
A continued solids release of 0.42 kg/s would represent a release of approximately 130
kg/year total PCBs to the river. This rate of loss is approximately two standard deviations
of the baseline annual loads from 1996-2000. A total PCB load of 130 kg/year within a
dredging season with full production is similar to a load of 65 kg/year within a dredging
season with half production (e.g., the Phase 1 resuspension criteria). Since this annual
load represents continual releases that are considerably greater than the best engineer
estimate resuspension rates in the FS, the dredging operations should not exceed these
criteria unless excess resuspension is occurring. Continued operation at the 500 ng/L
MCL would result in 500 kg/year of total PCBs being released to the river, a load similar
to those found in the early 1990s. This loss is above the current baseline conditions and
therefore operations cannot be maintained at this level and will be temporarily halted.
The baseline annual loads are highly variable and unpredictable. In earlier years, the
annual loading was dominated by DNAPL releases from the GE Hudson Falls Plants.
Since the controls have been installed, DNAPL releases have been greatly reduced and
the annual loads are dominated by the release of PCBs from the sediments of the TI Pool.
The annual loadings remain highly variable and significant. These calculations show that
if the remediation is controlled such that the rate of mass loss is below the action levels,
the increase in the annual loading will not be detectable.
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
6.0 References
USEPA. 1998. 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. Prepared for USEPA Region 2
and USACE, Kansas City District by TAMS Consultants, Limno-Tech, Inc, Menzie-Cura
& Associates, Inc., and TetraTech, Inc. December 1998.
USEPA, 2000. Hudson River PCBs Reassessment RI/FS; Feasibility Study. Prepared for
USEPA Region 2 and USACE, Kansas City District by TAMS Consultants, Inc.
December 2000.
USEPA, 2002. Hudson River Reassessment Record of Decision (ROD). Prepared for
USEPA Region 2 and USACE, Kansas City District by TAMS Consultants, Inc. January
2002.
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Tables
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment B - April 2004
-------
Table 1
Volume of Sediment Removed by Dredging Season
Sediment Removal Season, td
Dredging
Location
Dredging
speed
Volume of sediment
removed \ Vsed, (cy)
Sediment
density, p,
(tons/cy)
May 1 - Nov. 30, 2006
Sec. 1
half
260,000
0.94 2
May 1 - Nov. 30, 2007
Sec. 1
full
520,000
0.94 2
May 1 - Nov. 30, 2008
Sec. 1
full
520,000
0.94 2
May 1 - Aug. 15, 2009
Sec. 1 &
full
260,000
0.94 2
Aug. 16 - Nov. 30, 2009
Sec. 2
full
290,000
0.74 3
May 1 - Aug. 15,2010
Sec. 2 &
full
290,000
0.74 3
Aug. 16-Nov. 30, 2010
Sec. 3
full
255,000
0.71 4
May 1 - Aug. 15,2011
Sec. 3
full
255,000
0.71 4
Notes:
1. Calculations of volume sediment removed were presented in the FS, Table 8-9.
2. Based on the calculations in the FS, sediment removed consists of 50% cohesive (p =
0.71 tons/cy) and 50% non-cohesive (p = 1.16 tons/cy).
3. Based on the calculations in the FS, sediment removed consists of 93% cohesive (p =
0.71 tons/cy) and 7% non-cohesive (p = 1.16 tons/cy).
4. Based on the calculations in the FS, sediment removed consists of cohesive sediment
only (p = 0.71 tons/cy).
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 2
Estimated Tri+ and Total PCB Mass to be Remediated
River Section
Length of time for
remediation, td, (year)
Mass of Tri+ PCB
remediated2, A .
(kg)
Mass of TPCB
remediated2, MjpCB,
(kg)
River Section 1 (> 3 g/m2)
3.5
11,100
36,000
River Section 2 (> 10 g/m2)
1
7,100
24,300
River Section 3 (Select)
1
3,500
9,500
Total
5.5 1
21,700
69,800
Notes:
1. Dredging is scheduled to finish half way through the sixth year.
2. Mass of Tri+ and TPCB removed were calculated in the Responsiveness Summary,
Sediment PCB Inventory Estimates White Paper (USEPA, 2002).
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 3
Suspended Solids Estimated Increase to the Water Column
Sediment Removal Season
SS Increase @
SS Increase @
SS Increase @
2,000 cfs (mg/L)
5,000 cfs (mg/L)
8,000 cfs (mg/L)
Assuming a 0.21 kg/s Solids Loss Rate from Dredging
May 1 - Nov. 30, 2006
1.8
0.7
0.5
May 1 - Nov. 30, 2007
3.7
1.5
0.9
May 1 - Nov. 30, 2008
3.7
1.5
0.9
May 1 - Aug. 15, 2009
3.7
1.5
0.9
Aug. 16 - Nov. 30, 2009
3.2
1.3
0.8
May 1 - Aug. 15,2010
3.2
1.3
0.8
Aug. 16-Nov. 30, 2010
2.8
1.1
0.7
May 1 - Aug. 15,2011
2.8
1.1
0.7
Assuming a 0.42 kg/s Solids Loss Rate from Dredging
May 1 - Nov. 30, 2006
3.7
1.5
0.9
May 1 - Nov. 30, 2007
7.3
2.9
1.8
May 1 - Nov. 30, 2008
7.3
2.9
1.8
May 1 - Aug. 15, 2009
7.3
2.9
1.8
Aug. 16 - Nov. 30, 2009
6.5
2.6
1.6
May 1 - Aug. 15,2010
6.5
2.6
1.6
Aug. 16-Nov. 30, 2010
5.6
2.2
1.4
May 1 - Aug. 15,2011
5.6
2.2
1.4
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 4
Total PCBs Estimated Increase to the Water Column
Sediment Removal Season
Total PCB Increase
Total PCB Increase
Total PCB
Increase @ 8,000
cfs (mg/L)
@ 2,000 cfs (mg/L)
@ 5,000 cfs (mg/L)
Assuming a 300 g/day total PCB Loss Rate from Dredging
May 1 - Nov. 30, 2006
49
20
12
May 1 - Nov. 30, 2007
101
41
25
May 1 - Nov. 30, 2008
101
41
25
May 1 - Aug. 15, 2009
101
41
25
Aug. 16 - Nov. 30, 2009
202
81
51
May 1 - Aug. 15,2010
202
81
51
Aug. 16-Nov. 30, 2010
80
32
20
May 1 - Aug. 15,2011
80
32
20
Assuming a 600 g/day total PCB Loss Rate from Dredging
May 1 - Nov. 30, 2006
101
41
25
May 1 - Nov. 30, 2007
198
80
50
May 1 - Nov. 30, 2008
198
80
50
May 1 - Aug. 15, 2009
198
80
50
Aug. 16 - Nov. 30, 2009
418
168
105
May 1 - Aug. 15,2010
418
168
105
Aug. 16-Nov. 30, 2010
157
63
39
May 1 - Aug. 15,2011
157
63
39
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 5
Estimated Total PCB Concentrations Compared to the 95 Percent UCL Baseline
Data at the TID-West Monitoring Station Assuming a 300 g/day Total PCB Release Rate
Total PCB Release Rate of 300 g/day
Total PCB - TID-West Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May
June
July
August
Sept.
Oct. & Nov.
95% UCL Baseline TPCB Concentration
181
205
151
106
83
241
57
It >5
:s(.
" Id
:5(.
:i i
ISS
U(.
2,500
71
84
265
289
235
190
167
325
3,000
85
70
251
275
221
176
153
311
3,500
99
60
241
265
211
166
143
301
4,000
113
53
234
258
203
158
136
294
4,500
127
47
228
252
198
153
130
288
5,000
142
42
223
247
193
148
125
283
5,500
156
38
220
244
189
144
121
280
6,000
170
35
216
240
186
141
118
276
6,500
184
32
214
238
183
138
115
274
7,000
198
30
211
235
181
136
113
271
7,500
212
28
:u<>
233
179
134
111
269
X.000
::_
:<>
:os
1"
i
III')
:<.x
8,500
241
25
200
230
176
131
108
266
9,000
255
23
205
229
174
129
106
265
9,500
269
22
203
227
173
128
105
264
10,000
283
21
202
226
172
127
104
262
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 6
Estimated Total PCB Concentrations Compared to the 95 Percent UCL Baseline
Data at the TID-PRW2 Monitoring Station Assuming a 300 g/day Total PCB Release Rate
Total PCB Release Rate of 300 g/day
Total PCB - TID-PRW2 Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May & June (1)
July and
August
Sept.
Oct.
Nov.
95% UCL Baseline TPCB Concentration
(Data representative of flow Rates>5,000 cfs)
47
71
50
64
45
57
It >5
:i(.
n.
155
li.'J
I5(i
2,500
71
84
195
155
134
148
129
3,000
85
70
181
141
120
134
115
3,500
99
60
171
131
110
124
105
4,000
113
53
164
123
103
117
98
4,500
127
47
158
118
97
111
92
5,000
142
42
153
113
92
106
87
5,500
156
38
85
109
88
102
84
6,000
170
35
82
106
85
99
80
6,500
184
32
79
103
82
97
78
7,000
198
30
77
101
80
94
75
7,500
212
28
75
99
78
92
73
::_
:<
73
-------
Table 7
Estimated Total PCB Concentrations Compared to the 95 Percent UCL Baseline
Data at the Schuylerville Monitoring Station Assuming a 300 g/day Total PCB Release Rate
Total PCB Release Rate of 300g/day
Total PCB (ng/L)- Schuylerville Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May &
June
July
August
Sept.
Oct.
Nov.
95% UCL Baseline Total PCB Concentration
121
103
81
60
84
75
57
105
::<ฆ
:o"
IS5
l<>5
IX<>
ISO
2,500
71
84
205
186
164
144
168
159
3,000
85
70
191
172
150
130
154
145
3,500
99
60
181
162
140
120
144
135
4,000
113
53
174
155
133
113
136
128
4,500
127
47
168
149
127
107
131
122
5,000
142
42
163
144
122
102
126
117
5,500
156
38
160
140
118
98
122
113
6,000
170
35
156
137
115
95
119
110
6,500
184
32
154
134
112
92
116
107
7,000
198
30
151
132
110
90
114
105
7,500
212
28
149
130
108
88
112
103
X .000
::_
:<ฆ
I4S
128
106
86
110
101
8,500
241
25
146
127
105
85
109
100
9,000
255
23
145
125
103
83
107
98
9,500
269
22
143
124
102
82
106
97
10,000
283
21
142
123
101
81
105
96
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 8
Estimated Total PCB Concentrations Compared to the 95 Percent UCL Baseline
Data at the TID-West Monitoring Station Assuming a 600 g/day Total PCB Release Rate
Total PCB Release Rate of 600 g/day
Total PCB - TID-West Station
Flow (cfs)
Flow (m3/s)
TPCB
increase
(ng/L)
May
June
July
August
Sept.
Oct. & Nov.
95% UCL Baseline TPCB Concentration
181
205
151
106
83
241
57
:iu
^1
415
M,\
'l(.
:ซr.
45:
2,500
71
168
349
373
319
274
251
410
3,000
85
140
321
345
291
246
223
382
3,500
99
120
301
325
271
226
203
361
4,000
113
105
286
310
256
211
188
346
4,500
127
93
275
299
244
199
176
335
5,000
142
84
265
289
235
190
167
325
5,500
156
76
258
282
227
182
159
318
6,000
170
70
251
275
221
176
153
311
6,500
184
65
246
270
216
170
148
306
7,000
198
60
241
265
211
166
143
301
7,500
212
56
237
261
207
162
139
297
::_
53
25S
:
-------
Table 9
Estimated Total PCB Concentrations Compared to the 95 Percent UCL Baseline
Data at the TID-PRW2 Monitoring Station Assuming a 600 g/day Total PCB Release Rate
Total PCB Release Rate of 600 g/day
Total PCB - TID-PRW2 Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May & June (1)
July and
August
Sept.
Oct.
Nov.
95% UCL Baseline TPCB Concentration
(Data representative of flow Rates>5,000 cfs)
47
71
50
64
45
57
:iu
^:i
:xi
:<.o
:_4
:5(.
2,500
71
168
279
239
218
232
214
3,000
85
140
251
211
190
204
186
3,500
99
120
231
191
170
184
165
4,000
113
105
216
176
155
169
150
4,500
127
93
204
164
143
158
139
5,000
142
84
195
155
134
148
129
5,500
156
76
124
147
126
141
122
6,000
170
70
117
141
120
134
115
6,500
184
65
112
136
115
129
110
7,000
198
60
107
131
110
124
105
7,500
212
56
103
127
106
120
101
::_
53
loo
i:^
|0 i
1 r
-------
Table 10
Estimated Total PCB Concentrations Compared to the 95 Percent UCL Baseline
Data at the Schuylerville Monitoring Station Assuming 600 g/day Total PCB Release Rate
Total PCB Release Rate of 600 g/day
Total PCB (ng/L)- Schuylerville Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May &
June
July
August
Sept.
Oct.
Nov.
95% UCL Baseline Total PCB Concentration
121
103
81
60
84
75
2.000
57
210
331
313
291
270
294
285
2,500
71
168
289
271
249
228
252
243
3,000
85
140
261
243
221
200
224
215
3,500
99
120
241
223
201
180
204
195
4,000
113
105
226
208
186
165
189
180
4,500
127
93
215
196
174
154
177
169
5,000
142
84
205
187
165
144
168
159
5,500
156
76
198
179
157
137
160
152
6,000
170
70
191
173
151
130
154
145
6,500
184
65
186
167
145
125
149
140
7,000
198
60
181
163
141
120
144
135
7,500
212
56
177
159
137
116
140
131
8.000
227
53
174
155
133
113
136
128
8,500
241
49
171
152
130
110
133
125
9,000
255
47
168
149
127
107
131
122
9,500
269
44
166
147
125
104
128
119
10,000
283
42
163
145
123
102
126
117
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 11
Estimated Total PCB Concentrations Compared to the Prediction Interval Baseline
Data at the TID-West Monitoring Station Assuming 300 g/day Total PCB Release Rate
Total PCB Release Rate of 300 g/day
Total PCB- TID-West Station
Prediction Interval Baseline Total PCB
Concentrations
May
June
July
August
Sept.
Oct. &
Nov.
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
368
368
212
149
119
297
57
It >5
4"'
4"'
^r
254
::4
4t >:
2,500
71
84
452
452
296
233
203
381
3,000
85
70
438
438
282
219
189
367
3,500
99
60
428
428
272
209
179
357
4,000
113
53
420
421
264
201
172
350
4,500
127
47
415
415
258
195
166
344
5,000
142
42
410
410
254
191
161
339
5,500
156
38
406
406
250
187
157
336
6,000
170
35
403
403
247
184
154
332
6,500
184
32
400
401
244
181
151
330
7,000
198
30
398
398
242
179
149
327
7,500
212
28
396
396
240
177
147
325
::_
:<>
'M
1 "5
145
^24
8,500
241
25
393
393
236
173
144
322
9,000
255
23
391
392
235
172
143
321
9,500
269
22
390
390
234
171
141
319
10,000
283
21
389
389
233
170
140
318
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 12
Estimated Total PCB Concentrations Compared to the Prediction Interval Baseline
Data at the TID-PRW2 Monitoring Station Assuming a 300 g/day Total PCB Release Rate
Total PCB Release Rate of 300 g/day
Total PCB- TID-PRW2 Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May &
June (1)
July and
August
Sept.
Oct.
Nov.
Prediction Limit Baseline TPCB Concentration
68
106
72
92
65
(Data representative of flow Rates>5,000 cfs)
5"
It >5
:<ฆ<ฆ
:i i
1"
IT
m
2,500
71
84
245
190
156
176
149
3,000
85
70
231
176
142
162
135
3,500
99
60
221
166
132
152
125
4,000
113
53
213
159
124
144
118
4,500
127
47
207
153
118
138
112
5,000
142
42
203
148
114
134
107
5,500
156
38
106
145
110
130
103
6,000
170
35
103
141
107
127
100
6,500
184
32
100
139
104
124
97
7,000
198
30
98
136
102
122
95
7,500
212
28
96
134
100
120
93
8.000
227
26
94
133
98
118
91
8,500
241
25
92
131
97
116
90
9,000
255
23
91
130
95
115
88
9,500
269
22
90
128
94
114
87
10,000
283
21
89
127
93
113
86
Notes: (1) The 95percent UCL baseline varies as a function of flow rate for the months of May and June. It was
estimated that prediction interval baseline concentration is approximately 160 ng/L for flow rates less than 5,000 cfs.
This value was applied when estimating the total PCB concentration shown in the above table for all flow rates less
than 5,000 cfs.
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 13
Estimated Total PCB Concentrations Compared to the Prediction Interval Baseline Data at
the Schuylerville Monitoring Station Assuming a 300 g/day Total PCB Release Rate
Total PCB Release Rate of 300 g/day
Total PCB (ng/L) - Schuylerville Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May &
June
July
August
Sept.
Oct.
Nov.
Prediction Interval Baseline Total PCB
Concentrations
195
99
107
85
118
107
57
It >5
-Ot)
:o4
:i:
M
:i:
2,500
71
84
279
183
191
170
202
191
3,000
85
70
265
169
177
156
188
177
3,500
99
60
255
159
167
146
178
167
4,000
113
53
247
151
159
138
170
160
4,500
127
47
241
145
153
132
164
154
5,000
142
42
237
141
149
127
160
149
5,500
156
38
233
137
145
124
156
145
6,000
170
35
230
134
142
120
153
142
6,500
184
32
227
131
139
118
150
139
7,000
198
30
225
129
137
115
148
137
7,500
212
28
223
127
135
113
146
135
8.000
227
26
221
125
133
112
144
133
8,500
241
25
219
123
131
110
142
132
9,000
255
23
218
122
130
109
141
131
9,500
269
22
217
121
129
108
140
129
10,000
283
21
216
120
128
106
139
128
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 14
Estimated Total PCB Concentrations Compared to the Prediction Interval Baseline
Data at the TID-West Monitoring Station Assuming 600 g/day Total PCB Release Rate
Total PCB Release Rate of 600 g/day
Total PCB- TID-West Station
Prediction Interval Baseline Total PCB
Concentrations
May
June
July
August
Sept.
Oct. &
Nov.
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
368
368
212
149
119
297
57
:iu
5~X
5~X
4::
5( IS
2,500
71
168
536
536
380
317
287
466
3,000
85
140
508
508
352
289
259
437
3,500
99
120
488
488
332
269
239
417
4,000
113
105
473
473
317
254
224
402
4,500
127
93
461
462
305
242
213
391
5,000
142
84
452
452
296
233
203
381
5,500
156
76
444
445
288
225
196
374
6,000
170
70
438
438
282
219
189
367
6,500
184
65
432
433
276
213
184
362
7,000
198
60
428
428
272
209
179
357
7,500
212
56
424
424
268
205
175
353
::_
53
42n
421
2(4
:ui
r:
"ou
8,500
241
49
417
418
261
198
169
347
9,000
255
47
415
415
258
195
166
344
9,500
269
44
412
413
256
193
163
342
10,000
283
42
410
410
254
191
161
339
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 15
Estimated Total PCB Concentrations Compared to the Prediction Interval Baseline
Data at the TID-PRW2 Monitoring Station Assuming a 600 g/day Total PCB Release Rate
Total PCB Release Rate of 600 g/day
Total PCB- TID-PRW2 Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May &
June (1)
July and
August
Sept.
Oct.
Nov.
Prediction Limit Baseline TPCB Concentration
68
106
72
92
65
(Data representative of flow Rates>5,000 cfs)
5"
:iu
^"1
^r
:x:
'i>:
:_5
2,500
71
168
329
275
240
260
233
3,000
85
140
301
246
212
232
205
3,500
99
120
281
226
192
212
185
4,000
113
105
266
211
177
197
170
4,500
127
93
254
200
165
185
158
5,000
142
84
245
190
156
176
149
5,500
156
76
144
183
148
168
141
6,000
170
70
138
176
142
162
135
6,500
184
65
132
171
136
156
130
7,000
198
60
128
166
132
152
125
7,500
212
56
124
162
128
148
121
8.000
227
53
120
159
124
144
118
8,500
241
49
117
156
121
141
114
9,000
255
47
114
153
118
138
112
9,500
269
44
112
151
116
136
109
10,000
283
42
110
148
114
134
107
Notes: (1) The 95percent UCL baseline varies as a function of flow rate for the months of May and June. It was
estimated that prediction interval baseline concentration is approximately 160 ng/L for flow rates less than 5,000 cfs.
This value was applied when estimating the total PCB concentration shown in the above table for all flow rates less
than 5,000 cfs.
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 16
Estimated Total PCB Concentrations Compared to the Prediction Interval Baseline Data at
the Schuylerville Monitoring Station Assuming a 600 g/day Total PCB Release Rate
Total PCB Release Rate of 600 g/day
Total PCB (ng/L) - Schuylerville Station
Flow (cfs)
Flow (m3/s)
TPCB increase
(ng/L)
May &
June
July
August
Sept.
Oct.
Nov.
Prediction Interval Baseline Total PCB
Concentrations
195
99
107
85
118
107
57
:iu
405
^r
:<;<ฆ
^:x
i\~
2,500
71
168
363
267
275
254
286
275
3,000
85
140
335
239
247
226
258
247
3,500
99
120
315
219
227
206
238
227
4,000
113
105
300
204
212
191
223
212
4,500
127
93
288
192
200
179
211
201
5,000
142
84
279
183
191
170
202
191
5,500
156
76
271
175
183
162
194
184
6,000
170
70
265
169
177
156
188
177
6,500
184
65
259
163
171
150
182
172
7,000
198
60
255
159
167
146
178
167
7,500
212
56
251
155
163
142
174
163
8.000
227
53
247
151
159
138
170
160
8,500
241
49
244
148
156
135
167
157
9,000
255
47
241
145
153
132
164
154
9,500
269
44
239
143
151
130
162
151
10,000
283
42
237
141
149
127
160
149
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 17
Calculation of the Annual Dredging Induced PCB Load for the Fully Exhausted Standard
(500 ng/L)
Month
Average Fort
Edward Flow from
1976-1999
No. of
Work
Days/Mo.
Mass Loss (a) 500 ng/L
Daily Mass
Loss (kg)
Monthly
Mass Loss
(kg)
5
7,300
26
5
135
6
3,800
26
3
71
7
2,800
26
2
52
8
2,800
27
2
54
9
3,100
26
2
58
10
4,300
26
3
80
11
5,600
26
4
104
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 18
Calculation of the Annual Dredging Induced PCB Load for the 300 and
600 g/day Total PCB Mass Loss Control Limits
0.5% loss rate (Average of 300 g/day Total PCB Mass Loss)
Sediment Removal Season
Dredging
Location
speed
Cubic
yards of
sediment
removed
Total PCB
conc. on
solids
(mg/kg)
Total
PCB
flux
(g/day)
Total
PCB flux
(kg/day)
Total PCB
flux (kg/wk)
Total PCB
flux
(kg/year)
May 1 - Nov. 30, 2004
Sec. 1
half
260,000
27
140
0.14
0.84
25
May 1 - Nov. 30, 2005
Sec. 1
full
520,000
27
290
0.29
1.74
52
May 1 - Nov. 30, 2006
Sec. 1
full
520,000
27
290
0.29
1.74
52
May 1 - Aug. 15, 2007
Sec. 1 &
full
260,000
27
290
0.29
1.74
26
Aug. 16 - Nov. 30, 2007
Sec. 2
full
290,000
62
580
0.58
3.48
52
May 1 - Aug. 15, 2008
Sec. 2 &
full
290,000
62
580
0.58
3.48
52
Aug. 16 - Nov. 30, 2008
Sec. 3
full
255,000
28
230
0.23
1.38
21
May 1 - Aug. 15, 2009
Sec. 3
full
255,000
28
230
0.23
1.38
21
Total PCB flux (kg/project)
302
1% loss rate (Average of 600 g/day Total PCB Mass Loss)
Sediment Removal Season
Dredging
Location
speed
Cubic
yards of
sediment
removed
Total PCB
conc. on
solids
(mg/kg)
Total
PCB
flux
(g/day)
Total
PCB flux
(kg/day)
Total PCB
flux (kg/wk)
Total PCB
flux
(kg/year)
May 1 - Nov. 30, 2004
Sec. 1
half
260,000
27
290
0.29
1.74
52
May 1 - Nov. 30, 2005
Sec. 1
full
520,000
27
600
0.57
3.42
103
May 1 - Nov. 30, 2006
Sec. 1
full
520,000
27
600
0.57
3.42
103
May 1 - Aug. 15, 2007
Sec. 1 &
full
260,000
27
600
0.57
3.42
51
Aug. 16 - Nov. 30, 2007
Sec. 2
full
290,000
62
1200
1.2
7.2
108
May 1 - Aug. 15, 2008
Sec. 2 &
full
290,000
62
1200
1.2
7.2
108
Aug. 16 - Nov. 30, 2008
Sec. 3
full
255,000
28
450
0.45
2.7
41
May 1 - Aug. 15, 2009
Sec. 3
full
255,000
28
450
0.45
2.7
41
Total PCB flux (kg/project)
606
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Table 19
Dredging Induced Loss - Percent of the Baseline Annual Load
Year
Annual
Load to the
Water
Column
300 g/day
Loss (65
kg)
600 g/day
Loss (130
kg)
Fully
Exhausted
Standard
(500 kg)
1992
1,017
6%
13%
49%
1993
610
11%
21%
82%
1994
499
13%
26%
100%
1995
302
22%
43%
166%
1996
391
17%
33%
128%
1997
258
25%
50%
194%
1998
410
16%
32%
122%
1999
293
22%
44%
171%
2000
384
17%
34%
130%
Standard
Deviation
70 kg/yr for the years 1996-2000
220 kg/yr for the years 1992-2000
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Figures
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment B - April 2004
-------
500
450
400
3 350
&, 300
250
200
150
100
50
0
Figure 1
TID-West Monitoring Station - 95% UCL - Total PCB
600 g/day
~
300 g/day
0 cfs
0 cfs
May
June
July
August
Month
September
October
November
500
450
400
350 -
300 -
250 -
200 -
150
100
50
0
Figure 2
TID-PRW2 Monitoring Station - 95% UCL - Total PCB
600 g/day
~
300 g/day
95% UCL; Flow Rate < 5000 cfs
-t t
~ ~
95% UCL; Flow Rate > 5000 cfs
May
June
July
August
September
October
Q 2000 cfs
November
Month
Figure 3
Schuylerville Monitoring Station - 95% UCL - Total PCB
~J
~Bjd
ฃ
ซ
u
Ph
3
o
H
500
450
400 -
350 -
300
250
200 |
150
100
50 |
0
600 g/day
"iOT^jfiay
~
A
May
June
July
August September
Month
October
November
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
600 g/day
~
300c/day
Interval
~ 2000 cfs
8000 cfs
8000 cfs
Figure 4
TID-West Monitoring Station - Single Incident - Total PCB
^ 500 -
Vjd
400 -
09
U
^ 300 -
H 200 -
May June July August September October November
Month
Figure 5
TID-PRW2 Monitoring Station - Single Incident - Total PCB
May June July August September October November
Month
Figure 6
Month
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment B - April 2004
-------
Attachment C
Examination of Mechanisms
For High Dissolved Phase PCB Concentrations
Table of Contents
1.0 Introduction 1
2.0 Possible Release Mechanisms 4
2.1 Estimates of the Effects of Dredging on the Dissolved-Phase PCB
Concentration Using a Three-Phase Partitioning Model 4
2.1.1 Theoretical Estimation of the Mass of PCBs Available in the
Dissolved Phase 4
2.2 Analysis of Dissolved-Phase PCB Increase as a Result of Solids
Resuspension Using a Two-Phase Partitioning Model 9
3.0 Estimate of the Rate of PCB Desorption 14
3.1 Literature Revi ew 14
3.2 Dissolved Phase, Suspended Solids, and Whole Water PCB Concentration
Estimates using Desorption Rate Constants 16
4.0 Results from Field Studies with Dissolved and Suspended Phase PCB
Measurements 18
4.1 New Bedford Harbors 18
4.2 PCB Load Calculation 19
5.0 Conclusions 21
6.0 References 22
LIST OF TABLES
Table 1 Three-Phase Partition Coefficient Estimates for PCBs in Sediments of the
Freshwater Portion of the Hudson River
Table 2 Mean Length Weighted Average Concentration Estimate Using 1984
Thiessen Polygons, 1994 LRC and GE 1991 Composite Samples (was
Table 363334-2 of White Paper - Sediment PCB Inventory Estimates)
Table 3 Three-Phase Equilibrium Partitioning Model Results
Table 4 Water-Column Instantaneous PCB Loading at TI Dam
Table 5 Desorption Rate Constants from Literature
Table 6 PCBs Desorption Rate Constants and Partitioning Coefficients
Table 7 Background and Dredging Induced PCB Concentrations
Table 8 Dissolved Phase PCB Concentration Estimates
Hudson River PCBs Superfund Site
Engineering Performance Standards
l
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Attachment C
Examination of Mechanisms
For High Dissolved Phase PCB Concentrations
Table of Contents
LIST OF TABLES (continued)
Table 9 Summary of Field Samples and Analytical Data from the Pre-Design Field
Test - Dredge Technology Evaluation Report (8/6/2001)
Table 10 Dissolved and Particulate Percent PCB Mass Loss
LIST OF FIGURES
Figure 1 PCB, TSS and Turbidity vs. Distance from the Dredge
APPENDIX
Attachment C-l Literature Reviews
Hudson River PCB s Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Attachment C
Examination of Mechanisms for High Dissolved Phase PCB
Concentrations
1.0 Introduction
A United States Geological Survey (USGS) study of the Fox River SMU 56/57
demonstration projects (USGS, 2000) concluded that a large dissolved phase release of
PCBs had occurred in the absence of any apparent increase in the water column load of
suspended solids. Although there are some aspects of this study that suggest the
conclusions regarding dissolved phase release may be incorrect, the United States
Environmental Protection Agency (USEPA) has conducted several theoretical
assessments of possible mechanisms to determine if, in fact, such a release is a realistic
possibility. In order to address the issue of dissolved phase release, the proposed
monitoring program specifies the collection of whole water PCB data under normal
operating conditions (where water column concentrations are below a control limit that
varies by month and flow rate outlined in Attachment B). If the water column
concentrations are above a control limit, separate dissolved and particulate phase PCB
concentration analyses will be required. Other indicators of the total PCB concentration
in the water column will be measured, including total suspended solids, dissolved organic
carbon, and a qualitative measurement of dissolved phase PCB concentrations using
semipermeable membrane devices (SPMDs).
The Fox River dredging demonstration studies were examined in the White Paper -
Resuspension of PCBs During Dredging (USEPA, 2002). However, several significant
concerns were raised regarding the occurrence of a dissolved phase release during the
review of this study. To summarize the white paper: although a substantial amount of
data were collected from the Fox River dredging demonstration projects, the sampling
approach and compositing strategy mask the results. A close review shows that the study
results can only be considered inconclusive and should not be used as the basis for
estimating resuspension from any future dredging operations. The limitations in the Fox
River studies were discussed at length in the white paper, and are repeated here for the
convenience of the reader:
The load-gain estimate is based on a cross-section that is located too close to
the dredging area. The cross-section is also located in an area that is a likely
backwater (it is in a turning basin, with a nearby coal boat canal). It should be
noted that sampling activities during boat activity showed higher PCB
concentrations and were included in estimates of releases. Thus, flows through
the cross-section are unlikely to be consistent and the estimation of load from
concentration using these flows is suspect. The proximity of the cross-section
to the dredging area also increases the likelihood that the sampling will not be
representative of the total load, since the input from dredging will be poorly
mixed.
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
The sample compositing strategy, designed to reduce the number and cost of
PCB analyses, was not appropriate to support the mass flux analysis that was
attempted. The equal volume composites do not allow consideration of flow
variation across the cross-section. USGS (2000) states that stagnant areas and
even reversed flows were observed during sampling operations, confirming
the errors associated with the composite PCB samples. The TSS sample
composites induce less error and provide a more accurate estimate of
downstream TSS flux, yet they showed an unexplained decrease in suspended
sediment across the dredging operation. The decrease is almost certainly an
artifact associated with compositing equal volume samples from 20 percent
and 80 percent depth. Even though it has long been established that velocity
measurements from these depths represent the average velocity in an open
channel, there is no justification for suggesting that a composite sample from
these depths represents the average concentration along the profile. This is
particularly true in deeper water where the two samples represent 25 feet or
more of water depth.
The method of PCB collection was not documented, but it appears that the
method represents the dissolved and suspended matter fractions inaccurately,
based on the lack of change in PCB pattern across the dredging area. The load
gain is attributed to a large gain in dissolved PCBs, but this is inconsistent
with the PCB congener pattern. A large dissolved phase PCB contribution
from the sediments, either by porewater displacement or sediment-water
exchange, should yield a gain whose pattern is similar to the filter supernatant
(see Figure 336740-6 in the Responsiveness Summary to the ROD [USEPA,
2002]). The fact that the congener pattern is unchanged across the study area
would suggest a direct sediment addition, yet the suspended solids data
document no increase in suspended sediments.
Similarly, the total PCB concentration of the suspended matter doubles, yet
there is no change in the suspended matter loading. Given the proximity of the
downstream sampling cross-section to the source area, it is unlikely that the
majority of the TSS in the river could be directly affected by dredging induced
resuspension.
A review of the PCB loading over the dredging period shows that PCB loads
were relatively low for the first 2.5 months of operation, when dredging took
place at the more upstream end of the targeted area. During this period, the
estimated release was only 3 kg, or about 1.2 kg/month. This changed
dramatically during the last month of operation, when the loading rate
increased to about 13.5 kg/month. During this latter period, the dredging took
place at the downstream end of the targeted area, very close (the closest
station less than 80 feet) to the sampling cross-section, near areas with higher
PCB concentrations. As discussed in the USGS paper, another significant
factor that may have caused elevated PCB concentrations in the downstream
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
profile was increased water flow velocities. Proximity of dredging operations
to the deposit or water flow could have been significant contributing factors to
the increased PCB concentrations observed in the downstream profile. To
conclude that observed increases are only related to dredging fails to consider
these and other potential influences. Additionally, a lack of comparable
transect data for PCB water column concentrations pre-dredging (i.e.,
baseline) and during dredging also contributes to the uncertainty in evaluating
dredging surface water contributions.
The fact that significant loss of PCBs only occurred when the dredging area
was close to the sampling cross-section suggests that the settling of any
resuspended matter occurs within a short distance of the dredging operation.
Only when the monitoring location was close to the dredging could this signal
be found. This suggests that the loads obtained by this study do not represent
PCBs released for long-distance transport. Rather, the PCBs appear to be
quickly removed from the water column a short distance downstream. As
such, it is inappropriate to use these results to estimate downstream transport
from a dredging site.
There is much debate over the possibility of a dissolved phase PCB release during
dredging. In the following discussion, theoretical arguments are presented as to
mechanisms of release and a quantitative analysis of the magnitude of these releases. The
results of the New Bedford Harbor Pre-Design Test, where both dissolved and particulate
phase PCB concentrations were measured during dredging, are examined and compared
to the results of the theoretical analyses. A literature review of this issue is appended to
this Attachment (Attachment C-l).
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
2.0 Possible Release Mechanisms
In order to monitor PCBs correctly and minimize the impacts of dredging activities on
water quality, the nature of PCB releases due to dredging must be understood.
Specifically, the possibility that dredging will release dissolved phase PCBs must be
considered. There are two basic pathways through which dredging activities may cause
significant releases of dissolved phase PCBs:
The first mechanism requires a direct release of water containing dissolved phase
PCBs. Such water would most likely originate as porewater, since porewater is in
direct contact with the contaminated sediments and typically contains high
dissolved organic carbon concentrations, which can enhance the apparent
dissolved phase PCB concentration. The possibility of such a release mechanism
and the required water volumes are examined extensively from a theoretical
approach in Section 2.1. The analysis presented suggests that this pathway is
highly unlikely to result in significant releases.
The second mechanism of dissolved phase releases into the water column from
dredging is by desorption of PCBs from resuspended sediments. If the suspended
solids added are of sufficient mass and contamination level, the dissolved phase
concentration could rise markedly. It is worthy to note that the process of
equilibration will not be undone by adsorption if, as a result of downstream
transport, a large fraction of the suspended sediments are lost to settling. Since
equilibrium between solid and dissolved phase is concentration-driven and not
mass-driven, if a large mass of sediments is added to the water column, allowed to
equilibrate, and lost via settling, the water column will be left with a large
dissolved phase burden. This scenario is addressed in Section 2.2.
Although dissolved phase releases have historically been noted (USEPA, 1997; 2000)
under baseline conditions in the TI Pool, these releases occurred during summer low flow
periods without any significant resuspension of sediments. The conditions of these
releases suggest that a significant portion of the dissolved phase flux may be biologically
mediated. Due to the nature of dredging, it is unlikely that the same mechanism
underlying these releases will cause dredging-related dissolved phase releases.
2.1 Estimates of the Effects of Dredging on the Dissolved phase PCB
Concentration Using a Three-Phase Partitioning Model
2.1.1 Theoretical Estimation of the Mass of PCBs Available in the Dissolved Phase
During the Fox River PCB dredging project demonstration studies, the Water Resources
Institute of the University of Wisconsin reported that 25 percent of the PCB load released
from the Deposit N dredging demonstration project was in the dissolved phase (FRRAT,
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
2000). The United States Geological Survey (USGS) concluded in the paper A mass-
balance approach for assessing PCB movement during remediation of a PCB-
contaminated deposit on the Fox River, Wisconsin, "if chemical transport is to be
quantified during a PCB remediation, then monitoring of TSS and turbidity alone is not
adequate" (USGS, 2000). The study appeared to indicate that approximately 35 percent
of the PCB load from dredging Sediment Management Unit 56/57 was in the dissolved
phase (USGS, 2000). Due to this seemly high dissolved phase release it was concluded
that a concentration-based approach to assessing remediation could be misleading unless
the concentrations are converted into masses. Based on this, the PCB load into the water
column mass represented less than 2.5 percent of what was dredged from the deposit.
Since 35 percent of the PCB water column concentration increase was in the dissolved
phase, the fraction of total mass lost as dissolved phase PCBs during dredging was 0.9
percent (2.5 percent total loss x 35 percent as dissolved) or nearly one percent of the total
mass removed. Three phase partitioning models were used to estimate the volume of
Hudson River porewater required for a 1 percent release of dissolved phase PCBs into the
water column.
To evaluate the plausibility of the dissolved phase-based release mechanism, the
estimation of dissolved and DOC-bound PCB concentrations using a three-phase
equilibrium partitioning model was explored. Partitioning of organic chemicals between
sediment and porewater can be approached on either a mass concentration basis {i.e.,
mass of contaminant per dry weight of sediment), or a volumetric concentration basis
{i.e., mass of contaminant per volume of sediment). In this discussion, partitioning in the
sediments will be analyzed on a volumetric basis. The equilibrium partitioning model
assumes that the contaminant reaches equilibration among the different phases. On a
volumetric basis, one volume of sediment contains PCBs sorbed to the particulate phase
(solids) fraction, PCBs in the dissolved phase, and PCBs sorbed to the dissolved organic
carbon. The derivation of the following equations is based on the Data Evaluation and
Interpretation Report (DEIR) and Karickhoff (USEPA, 1997; Karickhoff, 1981). The
mass of PCBs in particulate phase is described as:
MP=CsolldxMsolidxlO-6 (EQ1)
where: Mp = mass of PCBs in particulate phase (mg)
CSoiid = concentration of PCBs on the suspended matter (mg/kg)
Msond = mass of sediments contained in the example volume (mg)
10"6 = factor to convert milligrams to kilograms
The mass of PCBs in the truly dissolved phase is described as:
= _CtJ M
Koc-foc Pซ
Md= solid x w_ x J q-6 (EQ2)
where: Md = mass of PCBs in the truly dissolved phase (mg)
Csolid = concentration of PCBs on the suspended matter (mg/kg)
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Koc = partition coefficient between water and organic carbon
(L/kg)
foe = fraction of organic carbon in the solid phase (unitless)
Mw = mass of water in example volume (mg)
pw = density of water (g/cc)
10"6 = factor to convert liters to cubic centimeters and grams to
milligrams
The mass of PCBs in the DOC-bound phase is described as:
Mdc = vCsoM, x Kdoc X Mdoc X 10_6 (EQ 3)
Koc ฆ f()C
where: Mdc = mass of PCBs in DOC-bound phase (mg)
Kdoc = partition coefficient between water and dissolved
organic carbon (L/kg)
10"6 = factor to convert kilograms to milligrams
Mdoc = Mass of dissolved organic carbon (mg), defined as
DOC X Vwater, where:
Vwater = Volume of water in example (L)
DOC = Dissolved organic carbon concentration (mg/L)
and other parameters are defined above.
The total concentration in the sample is given as the total mass of PCBs over the total
sample mass:
cSoi,M;
c.
solid
solid
CT =-
Koc-foe
- xM
C.
solid
K0C ' foe
xKD0C xMD0C
MSoM +MW +Mdoc
(EQ 4)
where: Ct = total concentration of PCBs
and other parameters are defined above.
The United States Army Corp of Engineers (USACE) Waterways Experiment Station
(WES) studied the partitioning of PCBs to organic carbon for differing degrees of
aromaticity (USACE, 1997). WES reported studies showing that the partitioning of
nonpolar organic compounds is strongly related to the octanol-water partitioning
coefficient of the compound (Karickhoff, 1981). The K0c values for a particular
compound have been reported to vary widely between sediments (Schrap and
Opperhuizen, 1989; Brannon et al., 1993, 1995a). Similarly, wide variations in Kdoc for
sediment porewater from different sediments have been observed (Chin and Gschwend,
1992, Brannon et al., 1995b). During their study, WES found that the measured values of
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Kdoc were consistently lower than the estimated Kdoc calculated using the method of
DiToro and others (1991) or Karickhoff (1981).
The USEPA estimated PCB partitioning coefficients using water column transect data
and the two-phase and three-phase sediment-water partition models during the Phase 2
reassessment. The results are summarized in the DEIR, Table 3-10a (USEPA, 1997). For
the purpose of evaluating the DOC-bound PCB fraction for the Hudson River, BZ#4 was
used to represent the mono- and di-chlorinated homologues fraction and BZ#28 and
BZ#31 to represent Tri+ PCBs. The partitioning coefficient for these congeners can be
found in Table 1.
The estimate of porewater DOC was obtained from the sediment sampling program
conducted by the General Electric Company (GE) in 1991 (O'Brien and Gere, 1993). The
median of composited porewater DOC was 37 mg/L (range of 10 to 212 mg/L), (USEPA,
1997).
The concentration of PCBs sorbed to solids in the sediment, Csoud, was obtained from the
length-weighted average PCB concentrations reported in the White Paper - Sediment
PCB Inventory Estimates. The average PCB concentration for River Section 1 was
calculated using data from the 1984 New York State Department of Environmental
Conservation (NYSDEC) survey, while concentrations in River Sections 2 and 3 were
computed using the 1994 low resolution coring data. Table 2 presents the in situ
remediated, non-remediated, and reach-wide length-weighted averages of Tri+ and Total
PCBs (without any overcut). In the calculations, the average concentration of 50 mg/kg
for the remediated sediment of Tri+ PCBs was used in the three-phase equilibrium
calculations. This average concentration serves as an upper bound value since the
remediated sediment average Tri+ concentrations for all three river sections are less than
50 mg/kg (Table 2).
To simplify the calculation, the entire Tri+ mass was assumed to act as BZ#28, which is
among the more soluble of the Tri+ congeners and thereby provides an upper bound on
the mass of Tri+ dissolved. Using this concentration, the mass of BZ#28 in the particulate
phase was 5x10" mg, while the mass of BZ#28 in the truly dissolved and DOC-bound
7 7
dissolved phases was estimated at 8.2x10" and 4.4x10" mg, respectively. The calculation
was repeated for BZ#31, another common constituent of the Tri+ congeners. The BZ#31
partitioning coefficients resulted in slightly higher truly dissolved and DOC-bound
7 7
phases; the values were 9.0x10" and 8.4x10" mg for the truly dissolved and DOC-bound
dissolved phases, respectively. Table 3 summarizes the results of the three-phase
equilibrium partitioning for BZ#4, BZ#28, and BZ#31.
To simulate the mono- and di-homologue fraction, BZ#4, the principal di-homologue
found in the sediment was used in the calculation. The concentration on the solid phase
for this calculation was obtained from River Section 2 (see Table 2). The Total PCB
average concentration of in situ sediment (without any overcut) targeted for remediation
in the FS for River Section 2 was 147 mg/kg, while the Tri+ average concentration for
this section of the river was only 44 mg/kg. This indicates that the mono- and di-
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
chlorinated homologues represent the majority of PCB mass in the sediments that may be
selected for remediated in River Section 2. Based on this information, an average
concentration of approximately 100 mg/kg was selected for the combined mono- and di-
chlorinated homologue concentration. Using BZ#4 as a surrogate for this group, the mass
of BZ#4 in the particulate phase is l.OxlO"1 mg and the mass of BZ#4 in the truly
dissolved and DOC-bound dissolved phases is 3.5xl0"7 and 3.5xl0"6 mg, respectively.
Assuming equilibrium conditions, it is clear that the sediment porewater contains very
little of the in situ sediment PCB mass. For the Tri+ fraction, the ratio of combined
dissolved and DOC-sorbed phases to the sediment-bound PCB fraction is given by:
(8.2xl0~7 +4.4xl0"7) ,
- = 2.4x10
5.2x10
or0.002per cent
Similarly for the mono- and di-homologue fractions:
(3.5 x 10"7 + 3.5 xlQ-6) _39x|0-5
lx 10"1
or 0.004 percent
A simple calculation can be used to estimate the number of porewater volumes that
would have to be displaced to achieve the roughly 1 percent of mass reportedly lost for
the Fox River study. This calculation assumes that each porewater volume would be
mixed with the sediments and brought to equilibrium before being released to the river.
Thus, to remove 1 percent of the mass via a dissolved phase displacement (without
resuspension), the proportion of water to sediment volume is given by the ratio of the
desired mass to be lost (1 percent) over the mass available in a single porewater volume
(0.0024 for Tri+ and 0.004 for mono- and di-homologues). Using the higher fraction to
yield the minimum number of volumes gives:
- = 250
0.004
or 250 porewater volumes. Since the sediments are roughly half water by volume, to
achieve the 1 percent loss without resuspension would require that each cubic yard of
sediment be washed with 250 porewater volumes, or about 125 cubic yards of water. For
the Tri+ fraction, with a lower percentage in the dissolved phase, this proportion would
nearly double to 420 volumes, or 210 cubic yards of water. It is important to note that this
mixing volume would have to be achieved for each yard of sediment removed and not for
the much smaller fraction of sediment that is lost or spilled.
In conclusion, assuming an equilibrium-based porewater concentration, a direct loss of
dissolved phase PCBs to the water column from porewater is highly unlikely. The
required mixing volumes of sediment to water are unlikely to be attained under any
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
reasonably well-operated dredging program. In fact, the mixing ratios suggested are
much more akin to a resuspension flux where the volume of water to the mass of solids
can easily achieve this, or even a much higher1, proportion. Thus, if a large mass of
dissolved PCBs is present in the water column downstream of the dredging operation, it
is more likely to be the result of the resuspension of sediment accompanied by PCB
equilibration between dissolved and suspended matter.
2.2 Analysis of Dissolved Phase PCB Increase as a Result of Solids
Resuspension Using a Two-Phase Partitioning Model
Section 2.1 demonstrates that it is highly unlikely that the increases in dissolved phase
concentration reported for the Fox River resulted from a direct release of dissolved or
"apparently dissolved" DOC-bound PCBs from the sediments. An alternative explanation
for the increase in reported dissolved concentrations is that it is due to desorption from
temporarily resuspended contaminated sediments. This section examines the mechanisms
for dissolved phase increase as a result of solids resuspension. The analysis also
examines the related question of whether the dissolved fraction of PCBs present in the
water column can be used as an indicator of dredging-related PCB releases.
A primary objective of the resuspension monitoring is to distinguish the dredging-related
contribution of PCB contamination to the water column from the baseline flux of PCBs
from the contaminated sediments. To meet this objective, it is important to determine
whether or not measurement of the whole-water PCB concentration is sufficient to
characterize an increase in the water column PCB concentration resulting from dredging,
or if the measurement of the dissolved phase PCB concentration is also necessary.
One way to distinguish a dredging-related PCB release from the baseline PCB
concentration is to compare the concentration of PCBs in the dissolved phase to the total
concentration of PCBs in the water column due to dredging activities. The next step
would be to compare these values to those of the baseline PCB concentrations in the TI
Pool. If the ratio of the concentrations detected during dredging operations differs from
the baseline ratio, then it is possible to distinguish dredging-induced inputs from the
baseline.
As evidenced by the GE float survey, USEPA Phase 2 inventory assessment, and GE
water column monitoring program data, Hudson River sediments continue to release
PCBs to the water column throughout the year. The data analyzed during the Phase 2
reassessment and subsequent data collected by GE show that PCBs are released to the
water column during low flow periods without resuspension of sediment, particularly
from May through November. During low flow periods, the observed suspended phase
concentration in the water column was low.
1 The addition of solids to achieve a concentration of 10 mg/L (a nominal value from Section 3 of this
attachment) represents a liquid to solids ratio of roughly a million to one.
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Recognizing the fact that there is PCB release to the water column without any
corresponding increase in total suspended solids (TSS), a scenario where dredging
operations cause the TSS levels to increase temporarily is considered. The affect of the
TSS increase to the water column is examined using a two-phase partitioning model. This
model provides a preliminary evaluation as to whether the effects of dredging activities
could be distinguished from baseline river conditions by examining the relative
magnitude of dissolved phase to total PCB releases to the water column.
As in the sediments, PCBs in the water column behave as a three-phase system, with
components of a dissolved phase, a phase sorbed to sediment, and a phase sorbed to
DOC. However, as discussed in the DEIR, the DOC-sorbed phase is of relatively minor
importance in the water column of the Hudson River. In addition, because DOC
concentrations are relatively constant, the system can be analyzed as an equivalent two-
phase system consisting of a sediment-sorbed fraction and an "apparent" (or unfilterable)
dissolved fraction that consists of truly dissolved and DOC-sorbed PCBs. Therefore, the
analysis that follows is presented in terms of a two-phase partitioning model.
The two-phase partitioning model assumes that the water column and the sediments are
in equilibrium. In a two-phase system, the PCB concentration in the water column is
equal to the sum of the dissolved phase fraction and the suspended solids fraction, such
that:
^^Total ^dissolved^ ^suspended ^ 'dissolved^" TSSx ( 'dissolved^ ^ii ^ ^ ^
(EQ 5)
where: C Total = total water column PCB concentration (ng/L)
Cdissolved = PCB concentration of apparent (non-filterable)
dissolved fraction (ng/L)
Csuspended = PCB concentration of suspended solids fraction
(ng/L)
Kd = soil-water partition coefficient (L/kg)
TSS = total suspended solids concentration (ng/L)
The whole water background concentration of the water column in the northern portion
of the TI Pool is nominally 50 ng/L. The background TSS value of 1 mg/L is assumed.
The concentration of the PCBs on the suspended matter, obtained from the instantaneous
total PCB water column loading for Transect 6 (USEPA, 1999), is approximately 5
mg/kg. Using these values and the equation above, the suspended solids concentration of
PCBs is estimated as:
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
CpCB -susp x Ctss - Cpcb as susp
(EQ 6)
(5 ng/mg) x (1 mg/L) = 5 ng/L
where: Cpcb-susP = concentration of PCBs on the suspended solids in
ng/mg (same as mg/kg)
Ctss = concentration of suspended solids in the water
(mg/L)
Cpcb as susp = concentration of PCBs on suspended solids per unit
volume of water (ng/L)
and the dissolved phase concentrations is estimated at:
(50 ng/L) - (5 ng/L) = 45 ng/L
The sediment-water partition coefficient for this example can be checked against the
values determined in the DEIR (nominally 105) by dividing the concentration in the
sediment by the concentration in the dissolved phase. The estimated Rvalue is:
(5 mg/kg) / (45 x 10"6 mg/L) = 1.1 x 105
which agrees well with the more rigorous calculation done in the DEIR. For this
calculation, the dredging operation is assumed to take place midway through the TI Pool.
For dredging scenarios with 1 percent loss rate at full production and flow between 2000
"3
to 5000 cfs (57 to 142 m /s), the additional TSS value to the water column due to
dredging is approximately 7 to 3 mg/L. Assuming the sediment concentration of 50
mg/kg (which is an upper bound for remediated sediment average concentrations for all
three river sections, USEPA, 2002), and the median TSS concentration (5 mg/L), the
additional PCB concentration associated with the suspended solids becomes:
(50 ng/mg) x (5 mg/L) = 250 ng/L
Therefore, the total concentration of PCBs in the water column accounting for the
additional TSS releases from dredging becomes:
(250 ng/L) + (45 ng/L) + (5 ng/L) = 300 ng/L
The dissolved phase fraction of PCBs added due to the TSS increase in the water column
can be calculated using equation 5 as:
(300 ng/L) = Cdissolved + [(5 mg/L + 1 mg/L) x Cdissolved* 1.1x10s L/kg x 10"6 kg/mg],
which gives: Cdissolved = 180 ng/L.
The sediment concentration (Csed) becomes:
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Csed ~ Cdissolved X ^li X ' ^ (EQ 7)
Csed = (180 ng/L) x (1.1x10s L/kg) x (1 kg/106 mg) = 20 ng/mg
or 20 mg/kg.
Assuming, at the subsequent monitoring station, that all the dredging-related TSS has
resettled and equilibrium is achieved before the sediment settles, the TSS inventory goes
from:
(6 mg/L) x (20 mg/kg) =120 ng/L
to
(1 mg/L) x (20 mg/kg) = 20 ng/L.
The loss in the inventory is approximately 100 ng/L, which means the total water column
concentration decreases from 300 ng/L to 200 ng/L during transport from the dredging
location to the downstream monitoring station.
The fraction of the dissolved phase to the total concentration of PCBs in the water
column due to dredging is:
(180 ng/L)/(200 ng/L) = 0.9.
Thus, resuspension of contaminated sediment and re-equilibration in the water column
provides a plausible explanation for the observation of an increased dissolved phase
concentration downstream of a dredging site.
As shown in the DEIR and FS, the sediments in the TI Pool continue to release PCBs to
the water column. Additionally, the seasonal variability of the last three to four years of
monitoring data collected by GE is strongly indicative of the absence of flow dependence
in the TI Pool's PCB loads. The absence of flow dependence would suggest that
resuspension resulting from flow is unlikely to be the cause of the PCB loading from the
TI Pool.
PCB loadings in the TI Pool were extensively quantified during the Phase 2 reassessment.
The Phase 2 water column monitoring program presents estimates of water column fluxes
for the period January to September 1993 (USEPA, 1997). Based on both instantaneous
and 15-day mean measurements, the TI Pool sediment was shown to be the dominant
source of PCBs to the water column in eight out of nine months of monitoring. This
source released less chlorinated PCB congeners that were predominantly found in the
dissolved phase in the water column (USEPA, 1997). In addition, GE and USGS water
column monitoring data support the findings based on Phase 2 data. In particular, the GE
data show the importance of the TI Pool sediment source for the period of 1991 to 1995.
Hudson River PCBs Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
These observations can be seen in Transects 5 and 6 during low flow conditions (Figure
C-3 and Figure 3-47 [corrected] of Appendix C of the Low Resolution Coring [LRC]
Responsiveness Summary, respectively [USEPA, 1999]). The values of whole (total)
water column, dissolved phase, and suspended solids concentrations at TI Dam and
Schuylerville are summarized in Table 4. These data showed that the baseline flux of
PCBs to the water column have a relative magnitude of dissolved phase to total
concentration on the order of 0.9.
Since the fraction of the dissolved phase to the total water column PCB concentration for
both background and after dredging is similar (on the order of 0.9), it is not possible to
distinguish the effect of dredging by examining the fraction of the dissolved phase
increase in the water column.
Hudson River PCBs Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
3.0 Estimate of the Rate of PCB Desorption
3.1 Literature Review
The theoretical assessments presented above are based on the three-phase and two-phase
partitioning models. Both theoretical arguments assume that the solid and dissolved phase
PCBs reach equilibrium. Recent studies have demonstrated that desorption of
hydrophobic chemicals from sediments can be quite slow, and that chemical equilibrium
may not be a good approximation in many real situations. In a dredging scenario, the
residence time (contact time) of the resuspended sediment in the water column is
relatively short, on the order of hours. It is unlikely that PCBs reach equilibrium in this
period of time. Desorption rates and the relative fractional amounts of hydrophobic
organic chemicals, including PCBs, released from sediment have been studied (Carroll et
al., 1994, Borglin et al., 1996; Cornelissen et al., 1997; ten Hulscher et al., 1999, 2002;
and Ghosh et al., 2000). Such kinetic rates could be used as an alternative to provide
estimates of the dissolved phase PCBs resulting from dredging activities. Literature on
the desorption rates of PCBs was reviewed to obtain desorption equilibrium and kinetics
rates for PCBs.
Many researchers showed evidence that desorption of contaminants takes place in at least
two steps: a fast and slow step. The desorption of PCBs from Hudson River sediments
was studied by Brown (1981) and Carroll and associates (Carroll et al., 1994). Brown
developed and tested a method for the analysis of rates of PCB desorption from sediment
suspended by dredging activities. The data used were taken from dredging operations in
the Hudson River at the town of Fort Edward in 1977. The monitoring stations were
placed in the east channel of Rogers Island. Brown used the Freundlich isotherms model
to obtain the sinking and sorption-desorption rate constants of Aroclor 1016. In the
report, the author used the term "sinking" to refer to the rate constant for the first order
settling coefficient. The sinking and sorption-desorption rates were chosen using trial and
error methodology to fit the measured concentration of Aroclor 1016 under low and high
flow conditions. For low flow conditions, it was found that a sinking rate of-0.08 hr"1
and desorption rate constants ranging from 0.025 hr"1 to 0.05 hr"1 fit the measured data
well. Under high flow conditions, a reasonable fit was obtained using a sinking rate of-
0.4 hr"1 and desorption rate constants on the order of 1.0 hr"1. Brown concluded that the
rate of PCB desorption from solids is proportional to the difference between the PCB
burden of the suspended sediments and the burden that would be in equilibrium with the
existing soluble concentration.
Carroll and associates studied the desorption of PCBs from Hudson River sediment using
XAD-4 resin as a PCB adsorbent. They used sediments contaminated with high, medium,
and low levels of PCBs from the Hudson River near Moreau, New York. The three
Hudson River sediments used in their study contained 25, 64, and 205 mg/kg (dry
weight) PCBs with total organic carbon contents of 0.96, 3.43, and 4.59 percent,
respectively. They reported that the PCBs present in the sediments consisted primarily of
Hudson River PCBs Superfund Site
Engineering Performance Standards
14
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
mono- and di-chlorinated biphenyls (60-70 percent of total). Both a rapidly desorbing
labile component and a more slowly desorbing resistant component were observed. Rate
constants for the labile (fast) and resistant (slow) fractions were obtained using a model
developed by Berens and Huvard (1981). For the purpose of this study, the desorption
rate constant of the untreated moderately (64 mg/kg dry weight PCB) PCB-contaminated
Hudson River sediment is considered. The desorption rate constant obtained from the
Carroll and associates study was approximately 0.018 hr"1 (Table 5).
Borglin and associates studied parameters affecting the desorption of hydrophobic
organic chemicals from suspended sediments (Borglin et al., 1996). In their paper,
Borglin and associates presented the results from long-term experiments performed for
three hydrophobic organic chemicals (hexachlorobenzenes and two polychlorinated
biphenyls). They concluded that the desorption times are on the order of a month to
several years, and observed that the desorption rates are dependent on the:
Particle/floc size and density distributions.
Type of water.
Amount of organic carbon in the sediments.
Time of adsorption before desorption.
Chemical partition coefficient.
Borglin and associates presented results describing the amount of PCBs
(monochlorobiphenyl and hexachlorobiphenyl) desorbed over time. The rate constants
calculated are on the order of 0.0049 hr"1 and 0.00042 hr"1 for monochlorobiphenyl and
hexachlorobiphenyl, respectively.
Cornelissen and associates studied desorption kinetics for chlorobenzenes, PAH, and
PCBs for different contact times and solute hydrophobicity (Cornelissen et al., 1997).
They used a technique employing Tenax TAฎ beads as a "sink" for desorbed solute to
measure the kinetics of desorption of the compounds mentioned above. For PCBs, they
studied PCB-65 (2,3,5,6-tetrachlorobiphenyl) and PCB-118 (2,3',4,4',5-
pentachlorobiphenyl). The sediment used was taken from Lake Oostvaardersplassen,
located in the Netherlands. They observed two stages of desorption rates: the rapid
release of the "labile" sorbed fraction, and slow release of the "non-labile" fraction. Two
different contact times were considered in this study: 2 days and 34 days. The desorption
rate constants were varied for the different contact times for both the rapid and slow
release. The values are summarized in Table 5.
In 1999, ten Hulscher and associates studied desorption kinetics and partitioning of
chlorobenzenes, PCBs, and PAHs in long-term field contaminated sediment cores and top
layer sediment (ten Hulscher et al., 1999). They concluded that the desorption from
sediment was triphasic: fast, slow, and very slow. In this study, they used the sediment
from Lake Ketelmeer, located in The Netherlands. Only core results were presented for
3 1
PCB-28. They reported desorption rate constants with values of 0.21x10" hr" and
3 1
0.19x10" hr" for a very slow fraction.
Hudson River PCBs Superfund Site
Engineering Performance Standards
15
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Ghosh and associates studied the relationship between PCB desorption equilibrium,
kinetics, and availability during land biotreatment (Ghosh et al., 2000). For this purpose,
they conducted a study of the equilibrium partitioning and desorption kinetics using
industrial lagoon sediments containing 0.91 percent oil and grease as a function of
biotreatment duration. A two compartment model was used to model the desorption of
PCBs from sediment. Desorption rate constants were reported for tri-, tetra-, penta-, and
hexa-chlorobiphenyls. Values for the untreated sediment are summarized in Table 5.
Recently, ten Hulschler and associates studied the desorption kinetics of in-situ
chlorobenzenes and 2,4,4'-trichlorobiphenyl (PCB-28) from River Rhine suspended
matter in Lobith, located in The Netherlands (ten Huschler et al., 2002). They observed
fast, slow, and very slow desorption rates for PCB-28. Rate constants observed were on
an average of 0.2 hr"1 for fast, 0.0004 hr"1 for slow, and 0.00022 hr"1 for very slow
desorption rates.
Table 5 summarizes the PCB desorption rate constants from different literature. From this
table it can be seen that there is a high degree of variation in the magnitude of PCB
desorption rate constants.
3.2 Dissolved Phase, Suspended Solids, and Whole Water PCB
Concentration Estimates using Desorption Rate Constants
Most of the reported values of desorption rate constants for PCBs are homologue-based,
except for Carroll, et al. who used an untreated PCB consisting of 60-70 percent mono-
and di-chlorinated biphenyls. The desorption rate constants from literature vary from
4.2x10"4 to 0.2 hr"1 (Table 6). The highest desorption rate constant reported is within the
range of those reported by Brown in 1981 for the Hudson River sediment (0.025 to 1.0 hr"
J). The reported rate constants correspond to a half-life range of approximately 3 to 1,700
hours and equilibrium range of 26 hours to 980 days (Table 6).
Given the length of time required for PCBs to reach equilibrium for desorption, it is
unlikely that there will be large release of dissolved phase PCBs as a result of dredging
activities. To demonstrate this hypothesis, the amount of dissolved phase PCBs within
one hour of dredging was estimated using the two-phase partitioning model, as was
described in Section 2 of this attachment. The desorption rate constants were used to
estimate what level of equilibrium was achieved in one hour. Due to lack of knowledge
on the amount of "labile" (fast) and "non-labile" (slow) fractions in the dredged material,
only fast desorption rate constants (ranging from 4.2x10"4 to 0.2 hr"1) are considered in
this study in order to be conservative. Since the reported desorption rate constants were
homologue-based, the ratios of the homologue to total PCBs are needed. The ratio of the
homologue to total PCBs for the sediment was taken from the low resolution coring data
(USEPA, 1998), while the ratio for the suspended solids and dissolved phase were taken
from Transect 6 water column PCB homologue composition for the TI Pool reported in
the DEIR (USEPA, 1997).
Hudson River PCBs Superfund Site
Engineering Performance Standards
16
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
The background and additional concentrations and TSS values used in this analysis were
the same as the values used in Section 2 of this attachment. The whole water background
concentration is 50 ng/L and the corresponding TSS value is 1 mg/L (Table 7). The
additional TSS value is 5 mg/L and sediment concentration is 50 mg/kg (Table 7).
Assuming a residence time of 1 hour, the dissolved phase PCB released due to dredging
ranges from 7.6xl0"5 ng/L to 3.23 ng/L (Table 8). The percentage of the dissolved phase
to the total concentration of PCB in the water column due to dredging ranges from 0.042
to 11 percent. From this analysis, it appears that the amount of dissolved phase in the
water column as a result of dredging is relatively small.
Hudson River PCB s Superfund Site
Engineering Performance Standards
17
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
4.0 Results from Field Studies with Dissolved and Suspended Phase
PCB Measurements
4.1 New Bedford Harbor
The analyses presented in Sections 2 and 3 of this attachment conclude that a significant
release of dissolved phase PCBs is unlikely to occur as a result of dredging activities. It is
possible to assess these results using case study field measurements of dissolved and
suspended PCB concentrations data in the water column during dredging. Measurements
of dissolved and particulate phase PCBs were collected during the pre-design field test
conducted at the New Bedford Harbor during August 2000.
A hybrid environmental mechanical/hydraulic excavator dredge was delivered and
demonstrated by Bean Environmental LLC. The system included a portable, shallow
draft barge platform, a horizontal profiling grab bucket (HPG), a crane monitoring system
(CM), the Bean-patented slurry processing unit (SPU), and a water recirculation system.
The average production rate for the dredge was 80 cubic yards per hour. An estimated
optimal rate for the system is 95 cubic yards per hour.
A summary of field samples and analytical data is presented in Table 9. TSS and turbidity
were measured along with dissolved and suspended phase PCBs. 18 National Oceanic
and Atmospheric Administration (NOAA) congeners were measured and an equation
developed during a previous study was used to calculate the total PCB concentration. The
following information was available:
Two pre-dredging measurements
Data from upstream and downstream monitoring points during dredging activities
Two measurements at the point of dredging.
The pre-dredging samples were collected 1000 feet to the north and south of the dredging
location. The harbor is tidal, so the upstream/downstream locations reverse periodically.
That is, the stations are located either north or south of the dredge, depending on the tide.
Sampling locations were placed as follows:
Location
Initially
Adjusted in
Field
Upstream
1000'
1000'
Downstream
50'
50'
Downstream
100'
300'
Downstream
500'
700'
Downstream
NA
1000'
Graphs of PCBs, TSS, and turbidity vs. distance from the dredge are shown in Figure 1.
The results for the pre-dredging samples are shown at +/-1500 feet on Figure 1 for
Hudson River PCBs Superfund Site
Engineering Performance Standards
18
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
comparison. The particulate PCB and TSS measurement give similar patterns of
concentrations as would be expected. At the point of dredging, the particulate PCB
concentrations are elevated about ten times over the upstream conditions, but 1000 feet
downstream the concentrations are just above the highest measured upstream
concentration. Turbidity measurements drop off quickly with distance to a level similar to
the upstream monitoring point conditions.
The dissolved phase PCB concentrations at the dredge are about ten times larger than the
upstream concentrations, but these concentrations drop off quickly into the range of the
upstream samples. Looking at the fraction of dissolved phase PCBs in the water column,
the upstream PCBs are about 60 percent dissolved. At the dredge, this percentage drops
to below 20 percent. Downstream of the dredge, the percent of dissolved phase is more
variable but still less than the 60 percent fraction detected at the upstream location. This
variability in the downstream samples is mirrored in the particulate PCB and TSS
measurements.
These results are consistent with a mechanism of PCB release through the suspension of
contaminated solids, not a significant dissolved phase release mechanism. This
conclusion is more convincing in light of the high concentrations at this location (857
ppm on average in the top 0- to 1-foot segment) relative to the Hudson River
(approximately 50 ppm on average in the TI Pool) and the nearly full production rate.
4.2 PCB Load Calculation
Dissolved and particulate phase PCB loads can be calculated using PCB concentrations
and estimates of the flow rate. Linear velocity was measured at one location 1500 feet
downstream of the dredging area. The estimate is quite crude because the volumetric flow
rate is not known, but can only be calculated by using a rough estimate of the cross-
sectional area at the point of the linear velocity measurement and by making the
assumption that the linear velocity measurement represents the entire cross-section. This
calculation further assumes that the PCB concentrations are a measure of concentration in
the entire cross-section, not a portion of the harbor that has been influenced by the plume.
The linear velocity was measured at a reference station 1500 feet south of the dredge
area. This section of the harbor is approximately 800 feet wide and varies from 7 to 10.5
feet in depth, depending on the tide. The velocity was measured every 10 minutes. The
northern velocities peaked at 14 cm/s. A velocity 10 cm/s will be used as an average flow
rate for the calculation. A limited southern component of flow was detected, indicating a
stratified system.
Several measurements of the PCB concentrations were made at locations from 50 to 1000
feet downstream from the dredge area. For this estimate of load, the maximum
concentration detected at the 100- to 1000-foot stations was selected to represent the
mass that would remain in the water column outside of the influence of the dredge. Both
the maximum dissolved and particulate concentrations were measured on the same day at
Hudson River PCBs Superfund Site
Engineering Performance Standards
19
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
700' from the dredge. A maximum dissolved phase PCB concentration of 0.95 ug/L was
detected. A maximum particulate phase PCB concentration of 2.6 ug/L was detected.
Two background measurements were made. The dissolved and particulate phase
background concentrations will be subtracted. The duration of the dredging operation in
hours was estimated from the time of the turbidity measurements.
Using these measurements of flow, concentration, and dredging operation duration, the
maximum likely PCB loads are 1.8 kg in the dissolved phase and 7.0 kg in the particulate
phase. The calculation is shown in Table 10. Twenty percent of the load is in the
dissolved phase, and 80 percent in the particulate phase. It was estimated that 1,495 kg of
PCBs were removed from the evaluation area. The dissolved phase load translates into
0.1 percent of the total mass removed, and the particulate phase load translates into 0.5
percent of the total mass removed.
Hudson River PCBs Superfund Site
Engineering Performance Standards
20
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
5.0 Conclusions
The release of a significant amount of dissolved phase PCBs as a result of dredging but
independent of the process of sediment resuspension would appear highly unlikely in
light of the discussion presented above. There is not a sufficient reservoir of dissolved
phase PCBs available to be the sole cause of a large increase in water column
concentrations. That noted, the process of suspended sediment-water contact could result
in a large inventory of dissolved phase PCBs if sufficient time is available to permit
exchange between suspended sediments and water. It is this latter process that may be of
concern during the Hudson River remediation.
Two important issues arise from this process, however. If the equilibration of dissolved
and suspended matter PCBs occurred sufficiently fast, the original nature of the source
{i.e., the suspended solids-borne PCBs) could be masked by the changes that occur. For
this reason, whole water PCB concentrations will be the main measure of PCB transport,
capturing all forms of PCBs present. Measurement of suspended matter PCBs alone may
under-represent the total level of PCB release.
The second issue relates to the usefulness of suspended solids as a surrogate and real-time
monitoring parameter. Near-field monitoring of suspended solids can probably be relied
upon to provide a useful indication of the amount of resuspension, although it will not be
quantitative for several reasons, including the issue discussed above. The monitoring of
suspended solids at the main downstream stations will be less sensitive to resuspension
inputs, but will still provide a useful measure of conditions in general. Given the typically
low suspended solids load of the Hudson during the dredging season, it is likely that
major suspended solids releases will still be discernable at these stations. To account for
this, whole water PCB samples will suffice when both suspended solids and PCB
concentrations fall below the lowest control limit. In the event that concentrations of
either parameter exceed this control limit, a second level of sampling will be required,
with more frequent sampling and separate analysis of both dissolved phase and
suspended matter PCBs. In addition, SPMDs will be deployed on a continuous basis to
give an indication of the dissolved phase concentrations between the water column
sampling events.
Hudson River PCBs Superfund Site
Engineering Performance Standards
21
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
6.0 References
Brannon, J.M., C.B. Price, F.J. Reilly, J.C. Pennington, and V.A. McFarland. 1993.
"Effects of sediment organic carbon on distribution of radiolabeled and PCBs among
sediment interstitial water and biota." Bulletin of Environmental Contamination and
Toxicology, Vol. 51, pp. 873-879. As cited in USACE/USAEWES Environmental
Effects of Dredging Technical Notes, EEDP-02-22.
Brannon, J.M., J.C. Pennington, W.M. Davis, and C. Hayes. 1995a. "Flouranthene
KDOC in sediment pore waters." Chemosphere, Vol. 30, pp. 419-428. As cited in
USACE/USAEWES Environmental Effects of Dredging Technical Notes, EEDP-02-22.
Brannon, J.M., J.C. Pennington, W.M. Davis, and C. Hayes. 1995b. "The effects of
sediment contact time on Koc of nonpolar organic contaminants." Chemosphere, Vol. 31,
pp. 3465-3473. As cited in USACE/USAEWES Environmental Effects of Dredging
Technical Notes, EEDP-02-22.
Berens, A.R. and G.S. Huvard. 1981. Particle Size Distribution of Polymers by Analysis
of Sorption Kinetics, Journal of Dispersion Science and Technology, Vol 2, pp. 359-387,
1981. As cited in Carroll, et al. (1994).
Borglin, S., A. Wilke, R. Jepsen, and W. Lick. 1996. "Parameters Affecting the
Desorption of Hydrophobic Organic Chemicals from Suspended Sediments." Env. Tox.
Chem. Vol. 15, No. 10, pp. 2254-2262.
Brown, M. 1981. "PCB Desorption from River Sediments Suspended During Dredging:
An Analytical Framework." New York State Department of Environmental Conservation,
Technical Paper No. 65. April 1981.
Carroll, K.M., M.R. Harkness, A.A. Bracco, and R.R. Balcarcel. 1994. "Application of a
Permeant/Polymer Diffusional Model to the Desorption of Polychlorinated Biphenyls
from Hudson River Sediment." Environ. Sci. Technol. Vol 28, pp. 253-258. 1994.
Chin, Y. and P.M. Gschwend. 1992. Partitioning of poly cyclic aromatic hydrocarbons to
porewater organic colloids. Environmental Science and Technology, Vol. 26, pp. 1621-
1626. As cited in USACE/USAEWES Environmental Effects of Dredging Technical
Notes, EEDP-02-22.
Cornelissen, G., P.C.M. Van Noort, and A. J. Govers. 1997. "Desorption kinetics of
chlorobenzenes, polycyclic aromatic hydrocarbons, and polychlorinated biphenyls:
sediment extraction with Tenaxฎ and effects of contact time and solute hydrophobicity "
Environ. Toxicol. Chem. Vol 16, No. 7, pp. 1351-1357, 1997
Hudson River PCBs Superfund Site
Engineering Performance Standards
22
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
DiToro, DM., C. S. Zarba, D. J. Hansen, W.J. Berry, R.C. Swartz, C.E. Cowan, S.P.
Pavlou, H.E. Allen, N.A. Thomas, and P.R. Paquin. 1991. "Technical basis for
establishing sediment quality criteria for nonionic organic chemicals using equilibrium
partitioning." Environmental Toxicology and Chemistry 10:1541-83. As cited in
USACE/USAEWES Environmental Effects of Dredging Technical Notes, EEDP-02-22.
Fox River Remediation Advisory Team (FRRAT). 2000. Evaluation of effectiveness of
remediation dredging - The Fox River deposit N demonstration project, November 1998-
January 1999: University of Wisconsin-Madison, Water Resources Institute Special
Report, WRI SR00-01, http://www.dnr.state.wi.us/org/water/wm/lowerfox/sediment/
frratdepositnreport.pdf. June 2000.
Ghosh, U., A.S. Weber, J.N. Jensen, and J.R. Smith. 2000. "Relationship between PCB
desorption equilibrium, kinetics, and availability during land biotreatment." Environ. Sci.
Technol. Vol. 34, No. 12, pp. 2542-2548, 2000.
Karickhoff, S. W. 1981. "Semi-empirical estimation of sorption of hydrophobic
pollutants on natural sediments and soils" Chemosphere 10:833-46.
O'Brien and Gere. 1993. Data Summary Report, Hudson River Sampling and Analysis
Program, 1991 Sediment Sampling and Analysis Program. Prepared for General Electric
Company Corporate Environmental Programs. O'Brien and Gere Engineers, Inc.,
Syracuse, New York. As cited in USACE/USAEWES Environmental Effects of Dredging
Technical Notes, EEDP-02-22.
ten Hulscher, Th.E.M., B.A. Vrind, H. Van den Heuvel, L.E. Van der Velde, P.C.M. Van
Noort, J.E.M. Beurskens, and H.A.J. Govers. 1999. "Triphasic desorption of highly
resistant chlorbenzenes, polychlorinated biphenyls, and polycyclic aromatic
hydrocarbons in Field Contaminated Sediment." Environ. Sci. Technol. Vol. 33, No. 1,
pp. 126-132, 1999.
ten Hulscher, Th.E.M., B.A. Vrind, P.C.M. van Noort, and H.A.J. Grovers. 2002.
"Resistant sorption of in situ chlorobenzenes and a polychlorinated biphenyl in river
Rhine suspended matter." Chemosphere, Vol 49, pp. 1231-1238, 2002.
U.S. Army Corps of Engineers (USACE). 1997. Environmental Effects of Dredging
Technical Notes, EEDP-02-22. U.S. Army Engineers Waterways Experiment Station
(USAEWES). August 1997.
USACE, 2001. Final Pre-Design Field Test Dredge Technology Evaluation Report New
Bedford Harbor Superfund Site. Prepared by Foster Wheeler Environmental Corporation.
August 2001.
U.S. Environmental Protection Agency (USEPA). 1997. Phase 2 Report, Further Site
Characterization and Analysis, Volume 2C - Data Evaluation and Interpretation Report
(DEIR), Hudson River PCBs RI/FS. Prepared for USEPA Region 2 and USACE by
Hudson River PCB s Superfund Site
Engineering Performance Standards
23
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
TAMS Consultants, Inc., the Cadmus Group, Inc., and Gradient Corporation. February
1997.
USEPA. 1998. Further Site Characterization and Analysis. Volume 2C-A Low
Resolution Sediment Coring Report (LRC), Addendum to the Data Evaluation and
Interpretation Report, Hudson River PCBs Reassessment RI/FS. Prepared for USEPA
Region 2, New York by TAMS Consultants, Inc., Gradient Corporation, and TetraTech,
Inc. July 1998.
USEPA, 2000. Phase 3 Report: Feasibility Study, Hudson River PCBs Reassessment
RI/FS. Prepared for EPA Region 2 and the US Army Corps of Engineers (USACE),
Kansas City District by TAMS Consultants, Inc. December 2000.
USEPA. 1999. Responsiveness Summary for Volume 2C-A Low Resolution Sediment
Coring Report, Addendum to the Data Evaluation and Interpretation Report. Prepared for
USEPA Region 2 and the USACE, Kansas City District by TAMS and TetraTech, Inc.
February 1999.
USEPA, 2002. Responsiveness Summary Hudson River PCBs Site Record Of Decision.
Prepared for USEPA Region 2 and USACE by TAMS Consultants, Inc. January 2002.
U.S. Geological Survey (USGS). 2000. A mass-balance approach for assessing PCB
movement during remediation of a PCB-contaminated deposit on the Fox River,
Wisconsin. USGS Water Resources Investigations Report 00-4245. December 2000.
Hudson River PCBs Superfund Site
Engineering Performance Standards
24
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Attachment C-l
Literature Review
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Attachment C-l
Literature Reviews
1.0 Literature Search For the Impact of Dissolved Phase
Contaminants During Sediment Removal Operations
Evidence has been reported that suggests that a significant dissolved phase release of
PCBs is possible without any apparent increase in the suspended solids load in the water
column. Because of this, several theoretical assessments of the possible mechanisms
behind such an increase have been performed by the USEPA.
Two basic pathways exist that can result in high dissolved phase PCB concentrations due
to dredging. The first is the direct release of water with a high dissolved phase PCB
concentration. This water would most likely originate as contaminated porewater within
the sediment. Porewater can be highly contaminated for two primary reasons: it is in
direct contact with contaminated sediments, and it typically contains a high concentration
of dissolved organic carbon, a medium that can enhance the apparent dissolved phase
concentration. In addition to porewater, water that comes in contact with the sediments
during the dredging process may also contain relatively high concentrations.
The second mechanism with the potential to create a high dissolved phase concentration
is an event that suspends a large mass of contaminated sediments in the water column.
PCBs will tend to equilibrate between solid and dissolved phases, effectively removing
PCBs from the suspended sediments to the water column. If the suspended solids added
are of sufficient mass and contamination level, the dissolved phase concentration can rise
markedly. It can be noted that the process of equilibration will not be undone if a large
fraction of the suspended sediments is lost to settling as the plume is transported
downstream. Because the equilibrium between the solid and dissolved phases is
concentration-driven and not mass-driven, the water column will be left with a large
dissolved phase burden if a significant mass of sediments is added to the water column,
allowed to equilibrate, and lost via settling.
To try to predict the changes in the water column dissolved PCB concentration during an
intrusive activity like dredging, it is important to have a basic understanding of the
possible mechanisms that could result in the dissolution of sorbed PCBs. The scientific
papers below were reviewed towards that end.
1. Rapidly Desorbing Fractions of PAHs in Contaminated Sediments as a Predictor
of the Extent of Bioremediation (Cornelissen et al., 1998)
Desorption kinetics of PAHs from contaminated sediments before and after
bioremediation are discussed in this study. The rapid desorption rate constant was
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
approximately 100-3000 times larger than the slow desorption rate constant. It is
concluded that the rapidly desorbing PAHs are primarily degraded during bioremediation
and the slowly desorbing amounts remain unchanged.
Reference:
Cornelissen, G.; Rigterink, H.; Ferdinandy, M. M. A.; Van Noort, P. C. M. "Rapidly
Desorbing Fractions of PAHs in Contaminated Sediments as a Predictor of the Extent of
Bioremediation," Environmental Science and. Technology, Vol. 32, pp. 966-970, 1998.
2. A Simple Tenax Extraction Method to Determine the Availability of Sediment-
Sorbed Organic Compounds (Cornelissen et al., 2001)
Fractions of PAHs, PCBs and chlorobenzenes that can be removed from contaminated
sediments by means of a single Tenax extraction are evaluated in this study. Two
extraction times (6 and 30 hours) in six different contaminated sediments collected from
various locations in The Netherlands were used to determine the fractions of PAHs,
PCBs, and chlorobenzenes that could be removed using the Tenax Extraction Method.
Results of the experiment indicated that extraction by Tenax for 30 hours completely
removed the rapidly desorbing fractions, plus some part of the slowly desorbing fraction,
whereas the fraction extracted by Tenax for 6 hours removed about half of the rapidly
desorbing fraction for chlorobenzenes, PCBs, and PAHs.
This study concluded that the concentration in sediment of rapidly desorbing, linearly
sorbed fractions can be determined by the amount desorbed to Tenax. For PCBs, the
amount linearly sorbed is about two times the amount desorbed to Tenax after a six-hour
contact time.
Reference:
Cornelissen, G.; Rigterink, H.; Ten Hulscher, D. E. M.; Vrind, B. A.; Van Noort, P. C.
M. "A Simple Tenax Extraction Method to Determine the Availability of Sediment-
Sorbed Organic Compounds;" Environmental Toxicology and Chemistry, Vol. 4, pp. 706-
711, 2001.
3. Fate and Transport of PCBs at the New Bedford Harbor Superfund Site
(Garton, et al., 1996)
This study presents a modeling approach, combining the theoretical, deterministic, and
empirical elements that were used to predict the fate and transport of PCBs at the
estuarine New Bedford Harbor Superfund Site. The theoretical approach was used to
characterize volatilization and sorption. Sediment processes including settling,
flocculation, resuspension, advection, and dispersion were characterized empirically and
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
sediment settling velocity deterministically from experimental data. The following
observations were reported from the model:
Sorption to sediments was reported to be the preferred state of PCBs in water
23 0 4 3
environments, with sorption coefficients ranging from 10" to 10" m /g for
Aroclor 1242 and Aroclor 1260. Affinity to sediments reportedly increased with
an increase in the percent chlorine.
Sediments in the harbor were undergoing continuous resuspension to the water
column and corresponding deposition. Resuspension and deposition were driven
by the change in the suspended solids concentration and tides. Deposition was
found to be greater during flood, while resuspension was greater during ebb.
Fluid shear was the most significant flocculation mass removal mechanism
contributing to the settling velocity calculation. It was observed by means of
visual observation that differential settling accounted for 30 percent of the mass
removal and fluid shear for 90 percent of the mass removal. Both these
mechanisms accounted for 100 percent mass removal and particle removal via
fluid shear occurred before differential settling.
It was concluded that the PCBs at the New Bedford Harbor are not very soluble and that
they volatilize or sorb to sediment rather than staying in solution. This allows PCB
transport from the harbor, either sorbed to sediments, transferred to mobile sediments
during resuspension activity, or by volatilization, thus leading to PCB contamination of
the water column, downstream areas, or atmosphere.
Reference:
Garton, L.S.; Bonner, J. S.; Ernest, A.N.; Autenrieth, R. L. "Fate and Transport of PCBs
at the New Bedford Harbor Superfund Site," Environmental Toxicology and Chemistry,
Vol. 15, pp. 736-745, 1996.
4. PCB Availability Assessment of River Dredging Using Caged Clams and Fish
(Rice et al., 1987)
The effects of dredging to remove PCB-contaminated sediments in the South Branch of
the Shiawassee River in south-central Michigan are presented in this study. The
bioavailability of PCBs was monitored using caged fingernail clams and fathead
minnows. Changes in water column concentrations of PCBs before dredging, during
dredging, and up to six months after dredging was completed were monitored and
compared to PCB bioavailabity data.
Monitoring of water, clams, and fish during dredging indicated that significant amounts
of PCBs were released from the sediments during dredging, which declined quickly
farther downstream. There were increases in the availability of PCBs for at least six
months at all locations downstream and in the area of dredging. However, there was no
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
noticeable change in the total PCB concentration in the water column after dredging.
Post-dredge uptake was also higher downstream. Overall, clams showed less uptake than
fish. It was concluded that dredging worsened the problem of bioavailability, at least over
the short term.
The researchers noted several important site-specific features of the south branch of the
Shiawassee River:
Large PCB deposits were found to occur along with fine, erodable, and
distinctly organic silt.
The sediment of the river was essentially lacking in clay.
The researchers stated that these factors might tend to make PCBs more available than
would be the case in the well mixed, sand-silt-clay type typically found on larger rivers.
Overall, it was concluded that among water, clams, and fish, there was no one ideal
monitor for the true bioavailability of PCBs in the South Branch of the Shiawassee River.
The fish were sensitive indicators of changes in PCB availability more than six miles
downstream of the dredging site. Uptake by fingernail clams appeared to reflect local
conditions at the sediment-water interface, but was not a sensitive indicator more than
one mile downstream.
Reference:
Rice, C. P.; White, D. S. "PCB Availability Assessment of River Dredging Using Caged
Clams and Fish," Environmental Toxicology and Chemistry, Vol. 6, pp. 259-274, 1987.
5. PCB Removal from the Duwamish River Estuary: Implications to the
Management Alternative for the Hudson River PCB Cleanup (Pavlou et al.,
1979)
This study presents the cleanup of the Duwamish River, Washington, and uses it as a test
case to compare it to the Hudson River problem. A transformer handling accident
resulted in a spill of transformer fluid, containing PCBs, into the river.
The initial cleanup was staged by divers using a hand dredge to recover submerged pools
of the liquid. This dredging ended within 20 days of the spill occurrence. The second,
more extensive cleanup that took place approximately 17 months later used a hydraulic
dredge and lasted approximately 24 days. Suspended particulate matter (SPM) and water
column concentrations were monitored during this second cleanup phase. The results of
monitoring reportedly revealed the following:
No change in the SPM concentration was observed throughout the dredging
operation.
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Water column PCB concentrations were observed to be constant throughout the
dredging operation.
Greater than 90 percent of PCBs were recovered in 44 days of dredging.
This study concluded that the dredging operations did not significantly alter the PCB
characteristics of the river.
Using the performance results of dredging in the Duwamish River as the basis, four
management alternatives for cleanup of the Hudson River were proposed. The
management alternatives included:
No Management Action Further Study
Stabilization and / or Removal of Remnant Deposits
Removal of Remnant Deposits and Sediments > 50 ppm
Removal of all River Sediments > 1 ppm.
The researchers stated that the best alternative for cleanup of the Hudson River would be
"Removal of Remnant Deposits and Sediments > 50 ppm," as this alternative was similar
to what was done in the Duwamish River, where no changes in the PCB levels of SPM
and water were observed. The paper also concluded that this alternative would also
remove 90 percent of the toxicant load, as was done in the Duwamish River, within
reasonable economic limits.
Reference:
Pavlou, S.P; Horn, W. "PCB Removal from the Duwamish River Estuary: Implications to
the Management Alternative for the Hudson River PCB Cleanup," ANNALS N.Y.
ACAD. SCI., Vol. 320, pp. 651-672, 1979.
6. Predicting Effluent PCBs from Superfund Site Dredged Material (Thackston et
al., 1992)
This paper discusses a feasibility study of dredge use to remove PCBs from sediments in
New Bedford Harbor, Massachusetts. Part of the study evaluated the usage of an onshore
confined disposal facility (CDF) to contain dredged material. A CDF is commonly used
in the disposal of dredged material that contains a wide range of contaminants.
The researchers also evaluate the validity of results generated by the modified elutriate
test to determine dissolved contaminant concentration and the concentrations associated
with suspended solids in the effluent generated from a CDF.
The modified elutriate test simulates the expected chemical and physical conditions
present in the CDF, and is based on both the dissolved and total concentrations of each
contaminant in the elutriate. The test is used to predict the contaminant concentrations in
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
the dissolved phase and also the concentrations associated with suspended solids present
in the elutriate.
The paper concludes that the elutriate test is a useful, accurate, and conservative predictor
of the concentrations of contaminants in the effluent from a CDF receiving highly
contaminated sediments.
Reference:
Thackston, Edward L; Palermo, Michael R. "Predicting Effluent PCBs from Superfund
Site Dredged Material," Journal of Environmental Engineering, Vol. 118, no. 5, 657-665,
1992.
7. Predicting Release of PCBs at Point of Dredging (DiGiano et al., 1993)
A dredging elutriate test (DRET) was used to predict the concentration of contaminants
(dissolved and suspended PCBs) as a function of initial concentration of sediment,
aeration time, and settling time in the water column at the point of dredging. Results from
the DRET were compared to field data from a pilot dredging operation at New Bedford
Harbor, Massachusetts.
The total PCB concentrations were proportional to the final TSS, while the soluble PCB
concentrations are nearly independent of the final TSS. The DRET tests also found that
aeration time had little effect on final TSS concentration. Settling times greater than six
hours produced little further removal of TSS, regardless of the initial TSS concentrations
or aeration time.
This study found that while small particles dominate the particle distribution with
increasing settling time, the PCB concentration per unit mass is not any greater than for
larger particles, thus the fraction of organic carbon, which determines the extent of
partitioning in the sediment, is not a function of particle size.
The New Bedford Harbor Field Data used three different dredge heads (cutter head,
horizontal auger, and matchbox), and samples taken directly from the ports of the dredge
head and from within 30m of the dredging area (plume samples). Sorbed and dissolved
PCB concentrations for the field plume samples were similar to the DRET data. The data
indicate that the horizontal auger causes the largest concentration of PCBs in the water
column of the three methods used.
All results suggest TSS is the most important factor in determining the PCB released into
the water at the point of dredging. The relationship between aqueous TSS concentration
and aqueous Total PCB concentration is directly proportional. The researchers proved
that the DRET could describe partitioning. The flocculent nature of particle settling
implies that far less efficient settling and thus higher total PCB concentrations may be
expected in freshwater dredging operations where destabilization of particles is less
effective.
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Reference:
DiGiano, F. A.; Miller, C. T.; Yoon, J. "Predicting Release of PCBs at Point of
Dredging," Journal of Environmental Engineering, Vol. 119, No. 1, pp. 72-89,
January/February, 1993
8. The Effect of Sediment Dredging on the Distribution of Organochlorine
Residues in a Lake Ecosystem (Sodergren et al., 1984)
Redistribution and deeper penetration of remaining residues of DDT compounds and
PCBs were observed in a Swedish lake after dredging. Water, sediment, and fish samples
were analyzed. Dredging was carried out in the summers of 1970 and 1971, and removed
300,000m3 of contaminated sediment.
Ten years after dredging, the level of PCBs in the upper 5 cm of sediment was about
twice as high as it had been immediately after the operations. The researchers believe that
the dredging operations apparently caused mixing and internal circulation of sediment
particles.
Levels of PCBs in sediment from an area of the lake that were not dredged were about
ten times higher than those in the central part of the lake before dredging.. Relatively
high PCB concentrations in this undredged area may be due to the historic contamination
of the area as an industrial dump for drainage water.
Reference:
Sodergren, Anders. "The Effect of Sediment Dredging on the Distribution of
Organochlorine Residues in a Lake Ecosystem," Ambio. Stockholm [AMBIO.], Vol. 13,
no.3, pp. 206-210, 1984.
9. Slowly Reversible Sorption of Aliphatic Halocarbons in Soils I. Formation of
Residual Fractions (Pignatello et al., 1990)
This study describes the formation (thermodynamics and kinetics) of slowly reversible
sorbed fractions of various halogenated aliphatic hydrocarbons (HHCs) (halogenated
solvents CT, TCA, TCE, TeCE, and soil fumigants 1,3-D, 1,2-DCP, EDB, and DBCP) in
two surface soils (Cheshire fine sandy loam, and an Agawam fine sandy loam). Soils
were allowed to sorb the compounds under two conditions: unsaturated soil (10 percent
moisture by weight), and soil suspended in an aqueous solution of HHC.
Desorption experiments using batch extraction of the HHCs from the soils with water
showed that the apparent soil-water distribution coefficients increased progressively to as
much as 200 times greater than equilibrium sorption coefficients Kd, obtained separately
from sorption isotherms. In each desorption case, the apparent distribution coefficient
(Kd,aPP) increased with each extraction from a value after the first extraction that was
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
comparable to Ka, to a value after the 16th extraction that was 1 to 2 orders of magnitude
greater than Kj. Thus, after repeated extraction, the soil retained significant quantities of
HHC, releasing it only slowly to the aqueous phase. Desorption experiments of HHCs on
soil using a continuous removal of Tenax CC polymeric absorbent beads yielded slowly
reversible residual fractions in the soil.
Desorption experiments using Tenax in an aqueous suspension showed that desorption
from the soil was rate-limiting. The researchers note that it is possible that uptake by
Tenax actually occurred from the vapor phase, although distribution of the HHCs from
the aqueous phase into Tenax is highly favorable; because Tenax is poorly wetted by
water and is known from extensive use in GC applications to be an efficient absorbent of
organic vapors.
The results of these experiments show that even compounds normally regarded as labile
in the environment by their volatility and weak equilibrium sorption tendencies can
generate kinetically slow sorbed residues.
Reference:
Pignatello, J.J. "Slowly Reversible Sorption of Aliphatic Halocarbons in Soils. I.
Formation of Residual Fractions," Environmental Toxicology and Chemistry, Vol. 9, pp.
1107-1115, 1990.
10. Why biota still accumulate high levels of PCB after removal of PCB ontaminated
sediments in a Norwegian fjord (Voie et al., 2002)
This study focused on a marine fjord located outside of Haakonsvern, a naval base in
Norway. Sediments contained in the fjord were found to be highly contaminated with
PCBs, and were removed via dredging in 1998. The objective of this study was to
determine which of the following hypotheses best corresponds to the reality of
bioavailability:
That contaminated food is the most important source accumulation due to the
low concentration of PCBs in water (estimated using the octanol-water partition
coefficient).
That the PCBs in the dissolved phase are the most important source of exposure.
Accumulation of low chlorinated PCB congeners with a low Kow in blue mussels and
SPMDs was higher than for the highly chlorinated congeners with a high Kow-
Bioaccumulation concentrations of PCBs before, during, and after dredging did not
change. Suspended matter/solids concentrations were not addressed. Water column
concentrations were not reported.
Related experiments indicated that PCBs are accumulated from the water column, and
that bioaccumulation in blue mussels and SPMDs occurs mostly from PCBs dissolved in
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
the water column. After dredging, more coarse materials were exposed to the seabed. The
coarse material has less ability to bind PCBs. Also, fine contaminated particles might
settle after dredging, leaving a thin contaminated layer of material.
Lower chlorinated PCBs are transported a longer distance than the higher chlorinated
congeners, thus accumulation of low chlorinated PCBs was higher in less contaminated
areas (4 km away).
If PCBs accumulate in blue mussel and SPMDs due to presence in the water column, the
bioaccumulation amounts in the biota may not have varied as significantly, as the water
concentrations of PCBs remained unchanged after dredging due to the low solubility of
PCBs.
Reference:
Voie, O. A.; Johnsen, A.; Rossland, H. K. "Why biota still accumulate high levels of
PCB after removal of PCB contaminated sediments in a Norwegian fjord," Chemosphere,
Vol. 46, pp. 1367-1372, 2002.
11. Desorption Kinetics of Chlorobenzenes, Polycyclic Aromatic Hydrocarbons, and
Polychlorinated Biphenyls: Sediment Extraction with Tenaxฎ and Effects of
Contact time and Solute Hydrophobicity (Cornelissen et al., 1997)
The kinetics of desorption of chlorobenzenes, polychlorinated biphenyls, and
polyaromatic hydrocarbons using Tenax beads from contaminated sediment (Lake
Oostvaardersplassen, Netherlands) was studied.
The sediment was dried to remove remaining organic contaminants as well as a number
of non-identified components that disturb chromatographic analyses. Contaminated lake
sediments and contaminated water spiked with concentrations ranging from 1 to 100 |Lxg/l
were allowed to equilibrate for 2 days and 34 days. After the equilibration time, sediment
and supernatant were separated by centrifugation, extracted with hexane, and analyzed
for contaminants and dissolved organic carbon.
Kinetics of desorption were determined by the Tenax extraction method. Rates of
extraction from the aqueous phase were also measured separately without any sediment.
The added amount of Tenax in this experiment was rendered insufficient due to the
amount of organic carbon present in the samples.
DOC data indicate that DOC is slowly released from the sediment during equilibration.
The fractions of contaminant present in the slowly desorbing sediment compartment,
Fsiow, are observed to increase with increasing test compound hydrophobicity. The rate
constants of slow desorption, ksiow, are observed to decrease with increasing equilibration
time, while Fsiow slightly increased with equilibration time. This phenomenon can be
explained by proceeding diffusion into the slowly exchanging sediment part (higher Fsiow)
and by the presence of the solute at more remote locations from which desorption is
slower (lower ksiow).
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
First order rate constants of rapid desorption were in the order of 10~Vh. First order rate
"3
constants of slow desorption were in the order of 10" /h. These correlate well with the
molecular volumes of the compounds used and decrease between 2 and 34 days of
equilibration. Slowly desorbing fractions increase with both increasing solute
hydrophobicity and increasing equilibration time.
Reference:
Cornelissen, G.; Van Noort, P. C. M.; Govers, H. "Desorption Kinetics of
Chlorobenzenes, Polycyclic Aromatic Hydrocarbons, and Polychlorinated Biphenyls:
Sediment Extraction with Tenaxฎ and Effects of Contact time and Solute
Hydrophobicity," Environmental Toxicology and Chemistry, Vol. 16, No. 7, pp. 1351-
1357, 1997.
12. Comparing Polychlorinated Biphenyl Concentrations and Patterns in the
Saginaw River Using Sediment, Caged Fish, and Semipermeable Membrane
Devices (Echols et al., 2000)
This experiment compared three possible techniques to assess the amount of bioavailable
polychlorinated biphenyls (PCBs) in the Saginaw River, Michigan:
Measurement of PCB concentrations in sediments.
Measurement of PCB concentrations in caged channel catfish.
Measurement of PCB concentrations in SPMDs.
The caged fish and SPMDs were placed in the river for 28 days at five sites where
sediments were sampled. Rates of PCB accumulation by SPMDs that have been reported
previously were used to estimate the aqueous concentrations from the PCB
concentrations found in the SPMDs, sediment-water partition coefficients were used to
estimate the dissolved PCB concentration from the sediment, and steady-state
bioaccumulation factors and depuration rate constants were used to estimate the aqueous
PCB concentration from the caged fish. The relative PCB patterns from the three
techniques were compared using principal components analysis.
The study found that SPMD and sediment results were complementary; the sediment
concentrations represent long-term accumulation and weathered components, while the
SPMDs show accumulations only from the sampling period. The lower chlorinated PCBs
predominate in the SPMDs as compared with the distribution in the fish and the
sediments, likely due to the higher solubilities of the lower chlorinated PCBs. The
distribution differences between the fish and the SPMDs are likely the result of
metabolism and depuration of certain congeners by the fish.
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Results from the water pattern modeling did not cluster on the principal component
analysis plot, co-varying positively and negatively on different axes. The sediment and
SPMD modeled data had similar patterns in the principal component analysis, but the
water concentrations derived from the sediment model were three to nine times higher
than those calculated from the SPMD model. The fish model results were closer to those
obtained from the SPMD model, but the patterns were different, likely due to the use of
alternate fish constraints (due to the lack of species-specific constraints available on then
model) or congener metabolism and depuration.
Reference:
Echols, K. R.; Gale, R.W.; Schwartz, T. R.; Huckins, J. N.; Williams, L. L.; Meadows, J.
C.; Morse, D.; Petty, J. D.; Orazio, C. E.; Tillitt, D. E. "Comparing Polychlorinated
Biphenyl Concentrations and Patterns in the Saginaw River Using Sediment, Caged Fish,
and Semipermeable Membrane Devices," Environmental Science and Technology, Vol.
34, pp. 4095-4102, 2000.
13. Mobilization of PAHs and PCBs from In-Place Contaminated Marine
Sediments During Simulated Resuspension Events (Latimer et al., 1999)
This study used a particle entrainment simulator (PES) to investigate the resuspension
transport of hydrophobic organic compounds, specifically PCBs and polycyclic aromatic
hydrocarbons (PAHs), to the overlying water column through the experimental
production of representative estuarine resuspension events. During the experiment, the
contaminants were evaluated in bulk sediments, size-fractioned sediments, resuspended
particulate material, and, in some cases, dissolved phases. Two types of sediment,
dredged material and bedded estuarine sediment, were used in this study, and they
represented gradients in contaminant loadings and textual characteristics. The sediments
were collected from Black Rock Harbor, Connecticut, and Narragansett Bay, Rhode
Island. The objectives of the study were to evaluate the chemistry and dynamics of the
contaminants as a function of the magnitude of resuspension.
Several conclusions regarding the resuspension chemistry and dynamics of hydrophobic
organic compounds were drawn:
The size of the particles entrained from the bedded sediments changed as the
resuspension magnitude increased. This can be attributed to the non-uniform
characteristics of the sediment with depth in the resuspension zone (up to 1
mm). In a case of more highly contaminated sediments, the mean particle size
was relatively constant under varying conditions of resuspension. The mean
particle size was also similar to that of the bulk sediment characteristics. In
contrast, for the less contaminated bedded sediment, the particle sizes decreased
over the same applied shear range. Also, the particle size distribution exhibited
by the bedded sediments during resuspension was more skewed toward smaller
particles than the bulk sediments.
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
On the basis of mass loading and an organic carbon loading weight, the amount
of PCBs and PAHs with a log Kow< 6 in the entrained particulate material was
somewhat depleted as the applied shear increased and the amount of material
resuspended in the water column was increased. Alternately, some higher
molecular weight PAH (log Kow < 6) showed slightly enriched loadings under
the same conditions. On a volume-weighted basis, the concentration of organic
contaminants increased in the water column as more material was resuspended.
In the case of the bedded sediments, accurate predictions of the entrained PAH
and PCB loadings on resuspended material were made using the resuspended
particle sizes and the concentration of the PAHs and PCBs in the particle size
pools of the bulk sediment. This prediction could not be made for the dredged
material, possibly due to contributions from the colloidal particles not
specifically measured in the study.
During the resuspension events, the distribution of PAHs between the dissolved
and particulate phases (Kds) showed relatively minor decreases with increased
applied shear and TSS levels. It was possible to calculate within a factor of 2 the
fraction with which the PAHs were associated based on the amount of organic
carbon in each of the resuspended samples. In order to obtain more accurate
predictions, however, kinetic factors and the role of other unmeasured substrates
would need to be taken into consideration.
The research suggests that resuspension, while periodic in nature, is likely an important
process affecting the fate and effects of contaminants in the coastal and marine
environment. Further study is needed to address the roles played by different sized
particles in this contamination contribution to shallower water systems and the conditions
under which theses contributions occur.
Reference:
Latimer, J.S.; Davis, W.R.; Keith, D.J. "Mobilization of PAHs and PCBs from In-Place
Contaminated Marine Sediments During Simulated Resuspension Events." Estuarine,
Coastal, and Shelf Science, Vol. 49, pp. 577-595, 1999.
14. Distribution of Organic Carbon and Organic Xenobiotics Among Different
Particle-Sized Fractions in Sediments (Kukkonen et al., 1996)
The distributions of benzo[a]pyrene, hexachlorobiphenyl, and total organic carbon in
sediment samples taken from Lake Michigan and Florissant, Missouri, were determined
and compared to the known bioavailability of the compounds. The goals of the study
were to demonstrate that the settling velocity method can be used for measuring the
xenobiotic distribution among sediment particles; to measure the effect of water quality
(lake water vs. distilled water) on the distribution of particles, organic carbon, and
xenobiotics in two different sediments; and to examine the sorption behavior of two
Hudson River PCBs Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
different xenobiotics (one PAH and one PCB) of similar hydrophobicity to try to account
for previously observed differences in bioavailability.
The distribution of the organic compounds among particles < 63 |j,m in diameter differed
from that of the total organic carbon;, however, the organic matter remained the major
sorbent for most of these compounds. Altering the fractionation conditions by performing
the procedure in distilled water rather than natural lake water changed the particle
distributions for both the organic carbon and the xenobiotics.
In addition, the contaminant distribution relative to the organic carbon content differed
between particle-size fractions and between contaminants of different compound classes,
e.g., PAHs and PCBs. The different distributions of the contaminants in the particle
fractions likely contributed to the observed differences in the bioavailability of the
organic contaminants to benthic organisms and may be exacerbated by selective feeding.
Reference:
Kukkonen, J.; Landrum, P.F.; "Distribution of Organic Carbon Xenobiotics Among
Different Particle-Size Fractions in Sediments," Chemospehere, Vol. 32, no. 6, pp. 1063-
1076, 1996.
Hudson River PCBs Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
2.0 Literature Review for PCB Desorption Rates
Resistant Sorption of In Situ Chlorobenzenes and a Polychlorinated Biphenyl in
River Rhine Suspended Matter
In this study, desorption kinetics of in situ chlorobenzenes (dichlorobenzenes,
pentachlorobenzene and hexachlorobenzene) and 2,4,4'-trichlorobiphenyl (PCB-28) were
measured for River Rhine suspended matter in Lobith, The Netherlands. The desorption
behavior of these pollutants (chlorobenzenes and PCB-28) in the suspended matter was
compared to their desorption behavior in the top layer (5-10 cm) of sediment in Lake
Ketelmeer, as this suspended matter was reported to be the main source of sediment
accumulation in Lake Ketelmeer.
Results of this study showed similarity of desorption profiles between River Rhine
suspended matter and the top layer of sediment from Lake Ketelmeer. Rate constants
observed were on an average 0.2 h"1 for fast desorption, 0.004 h"1 for slow desorption and
0.00022 h"1 for very slow desorption, which were in agreement to the values reported in
the literature. Fast desorbing fractions were not detected for any of the compounds other
than PCB-28 (1.6 percent of fast desorbing fractions were detected). The results of this
study concluded the following:
Slow and very slow desorbing fractions were already present in the material forming
the top layer of Lake Ketelmeer and were not formed after deposition of this material
in the lake.
The absence of recent pollution of the suspended matter could have caused the
absence of detectable fast fractions for most compounds in the suspended matter.
Rapid disappearance of compounds from the fast fraction could also be due to a
combination of a high affinity of very slow sites for these compounds and their
relatively high volatility.
The presumed differences in desorption patterns between a sediment top layer (5-10
cm) and the deepest layers (>10 cm) did not always exist.
Reference:
ten Hulscher, T. E. M.; Vrind, B. A.; van Noort, P. C. M.; Govers, H. A. J. "Resistant
Sorption of In Situ Chlorobenzenes and a Polychlorinated Biphenyl in River Rhine
Suspended Matter," Chemosphere, Vol. 49, pp. 1231-1238, 2002.
Hudson River PCB s Superfund Site
Engineering Performance Standards
14
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Desorption Rates of Two PCB Congeners from Suspended Sediments - I.
Experimental Results
Desorption of 2,5,2', 5'-tetrachlorobiphenyl (PCB-52) and 2,4,5,2', 4', 5'-
hexachlorobiphenyl (PCB-153) from suspended particles in a gas stripping reactor were
studied in this paper and experimental results reported. The objectives of the research
were to study the effects of particle size, congener properties, and equilibration time on
PCB desorption rates during resuspension events, and to develop a kinetic model to
simulate such a desorption process.
The experimental results indicated that PCB desorption was characterized by a two-stage
behavior - an initial rapid desorption followed by a prolonged slower desorption. PCB
desorption was found to be dependent on octanol-water partition coefficient (Koc),
independent of particle size during the initial rapid desorption stage and dependent on
particle size during the second desorption stage. Inverse relationship (decrease in overall
desorption as the equilibration time increased from 20 days to 3 years) between
desorption rate and equilibration time (aging effect) was observed and was reported to be
consistent with previous results reported in the literature.
The aging effect observed reportedly suggested that the release rates of PCBs in natural
systems were likely much lower than those observed in short-term laboratory
experiments, indicating that not only a kinetic model should be used in many aquatic
system models, but also that kinetic constants obtained in short-term laboratory
experiments may not be directly applicable to the desorption process in natural systems.
Reference:
Gong, Y.; Depinto, J. V.; Rhee, G. Y.; Liu, X. "Desorption Rates of Two PCB Congeners
from Suspended Sediments - I. Experimental Results," Water Resources, Vol. 32, No. 8,
pp. 2507-2517, 1998.
Desorption Rates of Two PCB Congeners from Suspended Sediments - II. Model
Simulation
Development of a two-compartment diffusion model and its application to simulate the
desorption kinetics of two PCB congeners 2,5,2', 5'-tetrachlorobiphenyl (PCB-52) and
2,4,5,2', 4', 5'-hexachlorobiphenyl (PCB-153) from suspended aquatic sediments are
discussed in this paper. The primary objectives of this paper were:
To explore other potential mechanisms (in addition to the retarded pore diffusion) that
may contribute to the two-distinct-rate behavior of PCB desorption.
Hudson River PCBs Superfund Site
Engineering Performance Standards
15
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
To develop a sorption kinetics submodel that was consistent with the majority of
mechanistic models and was practicable for system-level modeling of PCB transport
and fate.
To apply the developed model to simulate the experimental results presented in the
preceding paper (Paper 2 above).
The simulation results of this model concluded the following:
Both non-equilibrium sorption and non-uniform particle size distribution of the
natural sediments may contribute to the two-distinct-rate desorption behavior of the
PCBs that has been observed.
Compared to the single retarded pore diffusion model, the two-compartment
diffusion model, which assumed that one fraction of PCBs in solid phase reached an
instantaneous equilibrium with the surrounding aqueous phase while the other
fraction followed intra-particle diffusion, fit the data far better than the single
retarded pore diffusion model.
Increased adsorption time (aging) would in general decrease the instantaneous
equilibrium fraction and the effective pore diffusion coefficient.
Reference:
Gong, Y.; Depinto, J. V. "Desorption Rates of Two PCB Congeners from Suspended
Sediments - II. Model Simulation," Water Resources, Vol. 32, No. 8, pp. 2518-2532,
1998.
Polychlorinated Biphenyl Desorption from Low Organic Carbon Soils:
Measurement of Rates in Soil-Water Suspensions
Desorption-release rates of 13 individual PCB congeners from four contaminated soils
suspended in water were investigated using the gas purge technique. The soil samples
used for this investigation were obtained from PCB spill sites and had been in contact
with Aroclor 1242/1254 mixtures for 3 or more years, therefore it was assumed that
sorption equilibrium was obtained in these soil samples. Soils analyzed were
"engineered" ground cover materials used at utility industry substations and consisted of
fine rock chips and sand-silt-clay fractions with organic carbon < 0.2 percent. The PCB
congeners in the soils contained three to five chlorine
atoms. Proper functioning of the gas purge technique for measurement of congener
release rates was confirmed by measuring the Henry's law constants for 14C-labeled
congeners 24', 22'55' and 22'44'55' and comparing the results obtained with the values
reported in the literature.
Hudson River PCBs Superfund Site
Engineering Performance Standards
16
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
For all 13 congeners and all soil samples analyzed the following results were reported:
The labile fraction was found to be 80 to 90 percent of the total congener
concentration.
Majority of the labile fraction was desorbed or released within 48 hours of contact
with water.
Release of the remaining non-labile fraction persisted for over six months with
complete release estimated to be one to two years.
Release rate constants, Kd were found to decrease with increase in the number of
chlorines. The typical Kd values for labile and non-labile fractions were found to
range from 1.4 to 0.5 d"1 and 0.008 to 0.0006 d"1, respectively.
Reference:
Girvin, D. C.; Sklarew, D. S.; Scott, A. J; Zipperer, J. P. "Polychlorinated Biphenyl
Desorption from Low Organic Carbon Soils: Measurement of Rates in Soil-Water
Suspensions," Chemosphere, Vol. 35, No. 9, pp. 1987-2005, 1997.
A Simple Tenax Extraction Method to Determine the Availability of Sediment-
Sorbed Organic Compounds
Fractions of PAHs, PCBs and chlorobenzenes that can be removed from contaminated
sediments by means of a single Tenax extraction are evaluated in this study. Two
extraction times (6 and 30 hours) in six different contaminated sediments from various
locations in the Netherlands were used to determine the fractions of PAHs, PCBs, and
chlorobenzenes that could be removed using the Tenax Extraction Method. Results of the
experiment indicated that extraction by Tenax for 30 hours completely removed the
rapidly desorbing fractions plus some part of the slowly desorbing fraction, whereas the
fraction extracted by Tenax after 6 hours was about 0.5 times the rapidly desorbing
fraction for chlorobenzenes, PCBs an PAHs.
Reference:
Cornelissen, G.; Rigterink, H.; Ten Hulscher, D. E. M.; Vrind, B. A.; Van Noort, P. C.
M. "A Simple Tenax Extraction Method to Determine the Availability of Sediment-
Sorbed Organic Compounds;" Environmental Toxicology and Chemistry, Vol. 4, pp. 706-
711, 2001.
Hudson River PCBs Superfund Site
Engineering Performance Standards
17
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
PCB in the Upper Hudson River: Sediment Distributions, Water Interactions, and
Dredging
This paper is a summary of a number of studies performed by the DEC and various
consultants dealing with the PCB sediment distribution, water interaction, and dredging
for the Upper Hudson River. The studies were grouped by type and presented together.
The following conclusions were reached in the area of sediment distributions:
Over the course of mapping the sediment distributions in the Upper Hudson, it
was found that sampling on transects across the river and obtaining precise
locations for those samples was essential. The variation of PCB concentrations
across the river was extreme, while the concentration variation was more
gradual down the river.
The distribution of PCBs in the sediments can be classified as lognormal.
The PCB concentration was generally highest in silty sediments, next highest in
coarse sands containing wood chips, and lowest in the sands and gravels that do
not contain any woodchips or organics. The same trend held in sieved samples
composed of sand, wood chips, and silt.
PCB hot spots that contained concentrations above 50 (j,g/g were typically found
in low velocity and near bank areas. In the Upper Hudson, about 68 percent of
the total mass of PCBs is contained in hot spots that cover only 8 percent of the
river area.
PCB concentration was positively correlated with Cs-137, specific heavy metals,
and volatile solids. PCB concentration was negatively correlated with total
solids.
Sediment cores indicated that the maximum PCB concentrations were normally
found 8-30 cm below the top of the core. Dating using Cs-137 techniques placed
the peak discharge of PCBs in the 1960s. PCB contamination was rarely found
below 90 cm in the first 10 km from the contamination source, and rarely below
60 cm for the rest of the Upper Hudson.
The following conclusions were drawn from the water interaction studies:
The bulk of PCBs were adsorbed on solids in a concentrated sediment-water
mixture. When moving from a 10/1 elutriate test to a more dilute river system,
the sediment-water coefficient increased, and a higher percentage of the PCBs in
the mixture became soluble in the water. Given that Aroclor 1221 has a lower
sediment-water partition coefficient than Aroclor 1254, this finding is significant
to groundwater attenuation, river transport, and dredging systems.
Hudson River PCBs Superfund Site
Engineering Performance Standards
18
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Cationic polymers and chitosan were helpful in rapidly removing the suspended
solids in a sediment-water mixtures and reducing the concentration of PCBs in
the water.
High PCB concentrations occurred at low flow in the river, a phenomenon
possibly explained by desorption of PCBs from bottom sediments. The highest
concentrations of PCBs occurred during very high flows that eroded and
suspended bed material. The water PCB concentrations were lowest under
intermediate flow conditions.
The projected loss of PCBs to the Lower Hudson river over 20 years averaged
3,630 kg/yr, and modeled results indicated that this would occur for decades if
no action was taken.
The rate of PCB volatilization from the Upper Hudson varies with temperature,
wind speed, and turbulence conditions. The volatilization rate is projected to be
0.45-4.5 kg/day. This is in the range of the total river water transport of PCBs
under low flow conditions of 3-5 kg/day.
The examination of dredging projects yielded the following conclusions:
20 mg/1 of cationic polymer was found to be effective in boosting PCB and
suspended solids removals in spoils lagoons for three full-scale hydraulic
dredging projects on the Hudson. The best results were achieved when the
polymer was fed at an intermediate box between the two lagoons.
A minimum of one-hour retention time is recommended in the spoils lagoon
system for a hydraulic dredging project in the Hudson.
Scum removal in the hydraulic spoils lagoons and in the river downstream of a
dragline dredge was found to be essential in the Hudson due to the high
concentration of PCBs in the scum.
Hydraulic and mechanical dredging losses to the water column for the hot spot
dredging were projected to be about 2 percent of the PCB and 1 percent of the
solids, based on the monitoring data. The contaminated solids not picked up by
the dredge were projected to be 5 percent or greater. If the dredge operation is
not precisely controlled, the loss could potentially be greater than 5 percent.
Over 60 percent of the total mass of 200,000 kg of PCBs in the upper river is
expected to be removed via dredging of the hot spots and routine maintenance
dredging in 8 percent of the Upper Hudson.
Reference:
Toffelmire, T. J.; Hetling, L. J., Quinn, S.O. "PCB in the Upper Hudson River: Sediment
Distributions, Water Interactions, and Dredging," DEC Technical Paper No. 55, January
1979.
Hudson River PCBs Superfund Site
Engineering Performance Standards
19
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Volatilization of PCB from Sediment and Water: Experimental and Field Data
Studies done on the Hudson River PCB issue have suggested that the loss of PCBs
through the process of volatilization is substantial despite the fact that the contaminant
has a low vapor pressure. This report summarizes initial data and studies done to examine
PCB loss from the Hudson River through volatilization at the water-air and solid-air
interfaces.
Experimental data suggested that the volatilization of PCBs can be an important source of
air pollution under certain environmental conditions. The results of field monitoring have
shown that that PCB concentrations are fairly high in the ambient air and in vegetation
growing near PCB dump sites or certain contaminated dredge sites.
PCBs volatilized from contaminated water and sediment at substantial rates. For a
number of open PCB disposal and dredge spoil sites along the Upper Hudson River it
was observed that volatilization of PCBs was a worse problem than groundwater
contamination, although traditional control programs have been aimed at preventing
groundwater pollution.
Improved methods to prevent and control losses due to volatilization are needed, and
their long-term costs and consequences need to be considered. The comparison of some
exposure routes for PCBs indicate that intake from air exposure is greater than intake
from drinking water.
Reference:
Toffelmire, T. J.; Shen, T. T.; Buckley, E. H. "Volatilization of PCB from Sediment and
Water: Experimental and Field Data." Technical Paper # 63, December 1981.
Parameters Affecting Desorption of Hydrophobic Organic Chemicals from
Suspended Sediments
This study used long-term batch experiments to address the issue of chemical equilibrium
and its applicability as an approximation of the adsorption and desorption of hydrophobic
organic chemicals to soils and sediments. The experiments examined the behavior of
three hydrophobic organics: hexachlorobenzene, a monochlorobiphenyl, and a
hexachlorobiphenyl in Detroit River sediments suspended in pure water and/or filtered
tap water.
The experiments performed using hexachlorobenzene were extensive and demonstrated
the dependence of desorption rates on the particle/floc size and density distributions, the
type of water, and the organic content of the sediments. It was also demonstrated that
desorption was more rapid for sediments that were only partially equilibrated with the
chemical after a short-term adsorption period.
Hudson River PCBs Superfund Site
Engineering Performance Standards
20
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
The studies done on HCB also indicated that the rate of desorption was greatest initially
and decreased as the compound was desorbed, suggesting that the rates are also
dependent on the sediment concentration.
The experiments performed using PCBs demonstrated that desorption rates were also
dependent on the equilibrium coefficient partition coefficient of the chemical. For
example, the larger the partition coefficient, the slower desorption occurred. For more
highly chlorinated PCBs and other hydrophobic chemicals with high partition
coefficients, the desorption process is relatively slow, with desorption times on the order
of years. For areas where the effective particle sizes are or can potentially be much larger
(for example, bottom sediments and soils), the desorption times would be proportionately
greater.
It was also demonstrated that a chemical diffusion model with a diffusion coefficient that
is dependent on the porosity of the particle/floc, the organic content of the sediments, the
chemical partition coefficient, and also the distribution of the particle/floc size and
density distributions, was sufficient to explain the experimental results.
Reference:
Borglin, S.; Wilke, A.; Jepsen, R.; Lick, W. "Parameters Affecting the Desorption of
Hydrophobic Organic Chemicals from Suspended Sediments," Environmental Toxicology
and Chemistry, Vol. 15, No. 10, pp. 2254-2262, 1996.
PCB Desorption from River Sediments Suspended During Dredging: An Analytical
Framework
The purpose of this paper was to develop and test a method to analyze the rates of PCB
desorption from sediment that has been suspended by dredging activity. The data used
were taken from the monitoring of a dredging operation in the Hudson River at Fort
Edward in 1977. The monitoring activities took place in the east channel of Roger's
Island.
A system of PCB sorption-desorption kinetics that was developed to describe food chain
sorbents was used in the framework of a one-dimensional advective transport model and
solved at steady state conditions. The partition coefficient for Aroclor 1016 was chosen
for use in the model due to the prevalence of that particular PCB in the system. Due to
this, only Aroclor 1016 data will be included in the study. The sinking rate coefficient
was calculated using data from one of the monitoring stations, and the boundary
conditions were estimated using the partition coefficient and the total water column PCB
concentration.
hr 1
The application of a sinking rate of -0.08 " and sorption-desorption rate constants
ranging from 0.025 hr~' to 0.05 hr~' fitted the low flow average water column concentration
of Aroclor 1016 (Ct) reasonably well. However, applying a significantly slower rate
indicates that if no PCBs moved from the sorbed phase to the dissolved phase, the model
Hudson River PCBs Superfund Site
Engineering Performance Standards
21
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
results would not differ significantly from what was observed. A mechanistic fit of the
data using a higher sinking rate requires the utilization of a higher desorption rate
constant.
In the natural system, the results indicate that if the sinking rates are very large compared
to the rate of desorption, then a very low concentration of PCBs would be lost during
suspension. Conversely, if the desorption rates were high relative to the sinking rates,
then a substantially higher concentration of PCBs would be lost during suspension.
The best fits during model runs attempting to simulate high flow average monitoring
hr 1
results for suspended solids were produced sinking rates between -0.4 and -0.5 " and
desorption rate constants on the order of 1.0 hr_1. Rate constants that produced reasonable
hr 1
fits for either high or low flow data ranged from 0.025 to 1.0
Reference:
Brown, M. "PCB Desorption from River Sediments Suspended During Dredging: An
Analytical Framework," DEC Technical Paper No. 65, April 1981.
Hudson River PCBs Superfund Site
Engineering Performance Standards
22
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Tables
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment C - April 2004
-------
Table 1
Three-Phase Partition Coefficient Estimates for PCBs in Sediments of the Freshwater Portion of
the Hudson River
PCB Congener (BZ#)
Water Column Partition
Coefficient Estimates3
log Koc
log KDOc
4
5.19
5.43
28
5.84
4.16
31
5.80
4.40
Note:
a. Averages by homologue reported by Burgess et al. (1996) for the 4-8 cm depth layer
Source: DEIR, Table 3-10a (USEPA, 1997)
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 2
Mean Length Weighted Average Concentration Estimate using 1984 Thiessen Polygons, 1994 LRC
and GE 1991 Composite Samples (from Table 363334-2 of White Paper - Sediment PCB Inventory Estimates)
Total PCB
Remediated
Not Remediated
Reach Wide
Contaminant (PCB) Average Concentration
Fine
Coarse All
Fine Coarse
All
Fine
Coarse
All
River Section 1 (> 3 g/m2) (2)
164.5
35.2 92.1
(3)
39.4 23.8
25.4
(3)
145.3
28.9
63.0
(3)
River Section 2 (> 10 g/m2) ^
146.5
146.5
(4)
14.8
14.8
(5)
59.3
12.1
40.4
(7)
River Section 3 (Select)(2)
-
31.7
(4)
-
9.6
(6)
-
-
9.8
(7)
Tri+
Remediated
Not Remediated
Reach Wide
Contaminant (PCB) Average Concentration
Fine
Coarse All
Fine Coarse
All
Fine Coarse All
River Section 1 (> 3 g/m2) (2)
46.2
12.4 27.2
(8)
12.7 8.9
9.3
(8)
41.1 10.4 19.4
(8)
River Section 2 (> 10 g/m2) ^
43.1
43.1
(9)
7
6.9
(5)
17.3
(7)
River Section 3 (Select)(2)
-
11.7
(10)
-
5.1
(6)
5.4
(7)
Notes
1. Average concentrations were constructed using Thiessen polygons and Length Weighted Average values for the individual
sampling locations. Note that the Total PCB values for section 1 represent the Sum of Aroclors 1242, 1254, and 1260.
2. Includes channel area to be dredged.
3. LWA concentration estimate based on 1984 Thiessen Polygons. (Concentrations based on the Sum of Aroclors 1242, 1254, and 1260).
4. Mean MVUE values estimated from 1994 coring data from Hot Spots 25, 28, 31, 34, 35 for Section 2 and from Hot Spots 37 and 39 for Section 3
(Table 4-7 Low Resolution Coring Report).
5. LWA concentration estimate based on GE 1991 Composite samples falling outside the remediation boundaries (exclusion for Rocky Areas). (Estimated
from a single composite sample)
6. LWA concentration estimate based on GE 1991 composite samples falling outside the remediation boundaries (no exclusion for Rocky Areas). (Estimated
from 45 composite samples)
7. LWA concentration estimate based on all GE 1991 Composite samples in the section.
8. LWA concentration estimate based on 1984 Thiessen Polygons. A factor of 0.944 is applied to the sum of Aroclors values to obtain estimates of Tri+
PCB values.
9. Tri+ values based on Total PCB estimates from 1994 coring data. A divider of 3.4 is applied to the Total PCB value.
10. Tri+ values based on Total PCB estimates from 1994 coring data. A divider of 2.7 is applied to the Total PCB value.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 3
Three-Phase Equilibrium Partitioning Model Results
PCB
Mass in truly
Total
Percent of
Congener
Mass in particulate
dissolved phase, Md
Mass in DOC-bound
Mass
Dissolved
dissolved
(BZ#)
phase, MP (mg)
Log Koc
(mg)
Log KDOc
phase, Mdc (mg)
(mg)
Mass (mg)
mass (%)
4
1.0E-01
5.19
3.5E-07
5.43
3.5E-06
1.0E-01
3.9E-06
0.0038%
28
5.0E-02
5.84
8.2E-07
4.40
4.4E-07
5.0E-02
1.3E-06
0.0025%
31
5.0E-02
5.80
9.0E-07
4.16
8.4E-07
5.0E-02
1.7E-06
0.0035%
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 4
Water-Column Instantaneous PCB Loading at TI Dam
TI Dam
Flow (m3/s)
Whole (total) water
PCBs (ng/L)
Dissolved phase PCB
(ng/L)
Suspended solids
PCBs (ng/L)
Ratio of dissolved to
total concentration
TI DAM
Transect 5
76
192
184
11.2
0.96
Transect 6
69
92
88
2.9
0.96
Schuylerville
Transect 5
85
160
150
15
0.94
Transect 6
74
89
84
4.8
0.94
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 5
Desorption Rate Constants from Literature
Compounds
Borglin et
al., 1996
day"1
ten Hulscher etal., 1999; 2002
Lobith susp. Matter
kfast (hr ) kyslow )
Ketelmeer
kyslow (fr" )
Rate Constants (k)
Cornelissen etal., 1997
^rapid
2 day
(hf1)
34 day
kslow )
2 day 34 day
Ghosh et al., 2000
kfast (day-1)
, (day")
Carrol et
al., 1994
k (hf1)
Monochlorobiphenyls
Trichlorobiphenyls
PCB-28 (trichloro)
PCB 65 (tetra)
T etrachlorobipheny Is
PCB 118 (penta)
Pentachlorobiphenyls
Hexachlorobiphenyl
Moderately PCB contaminated
Hudson River Sediment3
0.1174
0.2
2.25E-04
2.00E-04
0.058
0.045
0.117
0.112
2.54E-03
2.01E-03
1.74E-03
9.80E-04
0.0101
0.83
0.38
0.15
0.07
0.011
0.011
0.004
0.005
0.018
Note:
a As reported by Carrol etal1994. Moderately PCB contaminated sediment contained 64 mg/kg (dry weight) PCBs, with total organic carbon of 3.43%.
The PCB presents in the sediments consisted of primarily mono- and di-chlorinated biphenyls (60-70% or total).
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 6
*CBs Desorption Rate Constants and Partitioning Coefficients
Compound
Rate constant (k)
Half-
life
Estimated equ
llibrium time
Log Koc h
Log Kd1
hr"1
hr"1
hr
hr
PCB in equilibrium
5.05
Monochlorobiphenyls
0.0049 3
142 3
84 days
a
5.65
4.38
Trichlorobiphenyls
0.035 b
20 b
9 days
b
5.84
4.57
PCB-28 (trichloro)
0.2 c
3 c
26 hr
c
5.84
4.57
PCB 65 (tetra)
0.058 d'e
0.117 d'f
12 d'e
6 d'f
5.5 days
d,e
2.7 days d'f
6.27
5.00
Tetrachlorobiphenyls
0.016 b
44 b
14 days
b
6.27
5.00
PCB 118 (penta)
0.045 d'e
0.112 df
15 d'e
6
7 days
d,e
2.8 days d'f
6.41
5.14
Pentachlorobiphenyls
0.0063 b
111 b
50.7 days
b
6.41
5.14
Hexachlorobiphenyl
0.00042 3
0.0029 b
1664 3
238 b
980 days
a
108 days b
6.55
5.28
Moderately PCB contaminated8
0.0181 8
38 8
422 days
g
5.05
Notes:
3 Borglin et al. (1996)
b Ghosh etal. (2000)
c ten Hulscher et al. (1999; 2002)
d Cornelissen et al. (1997)
e k is for 2 day contact time
f k is for 34 day contact time
8 Carrolletal. (1994).Moderately PCB contaminated sediment contained 64 mg/kg (dry weight) PCBs,
with total organic carbon of 3.43%. The PCB presents in the sediments consisted of primarily mono- and
di-chlorinated biphenyls (60-70% or total).
h Partitioning coefficients were taken from DEIR Table 3-8 (USEPA, 1997)
1 foe of sediment is 5.38%
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 7
Background and Dredging Induced PCB Concentrations
Background Concentrations
Dredging Induced
Compound
Ratio to Total
Ratio to Total
PCB
(suspended
phase)b
Ratio to Total
PCB
(dissolved
phase)b
Csed b
TSSb
Ctotal b
Csusp b
Cdiss b
Csed d
TSSd
Csusp d
Ctotal b+d
PCB
(sediment)3
mg/kg
mg/L
ng/L
ng/L
ng/L
mg/kg
mg/L
ng/L
ng/L
PCB in equilibrium
1
1
1
5
1
50
5
45
50
5
250
300
Monochlorobiphenyls
0.14
0.0013
0.16
0.70
0.00131
8
9.11E-04
8.2
7
0.0065
0.0455
8
T richlorobiphenyls
0.30
0.0103
0.27
1.51
0.01034
13
0.02
13.2
15
0.0517
0.78
14
PCB-28 (trichloro)
0.30
0.0103
0.27
1.51
0.01034
13
0.02
13.2
15
0.0517
0.78
14
PCB 65 (tetra)
0.13
0.0072
0.13
0.63
0.00722
7
0.005
6.51
6.3
0.0361
0.23
6.7
T etrachlorobiphenyls
0.13
0.0072
0.13
0.63
0.00722
7
0.005
6.51
6.3
0.0361
0.23
6.7
PCB 118 (penta)
0.044
0.0032
0.026
0.22
0.00317
1
0.0007
1.28
2.2
0.0158
0.035
1.3
Pentachlorobiphenyls
0.044
0.0032
0.026
0.22
0.00317
1
0.0007
1.28
2.2
0.0158
0.035
1.3
Hexachlorobiphenyl
0.016
0.0021
0.0035
0.08
0.00208
0.17
0.00016
0.17
0.79
0.0104
0.0082
0.18
Moderately PCB contaminated8
1
1
1
5
1
50
5
45
50
5
250
300
Notes:
a Ratio of homologue to Total PCB in the sediment was taken from the low resolution coring data (USEPA, 1998)
b Ratio of homologue to Total PCB were taken from transect 6 water column data reported in DEIR (USEPA, 1997)
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 8
Dissolved Phase PCB Concentration Estimates
In 1 hour
Compound
% equilibrium
Cdiss due to dredge
Cdiss/Ctotal
Time
ng/L
%
(hour)
PCB in equilibrium
equil
100%
180
h
60.0%
h
Monochlorobiphenyls
1
0.49% 3
4.03E-02
0.5%
Trichlorobiphenyls
1
3.4% b
4.76E-01
3.4%
PCB-28 (trichloro)
1
18% c
2.54
18.1%
PCB 65 (tetra)
1
5.6% d'e
11% d'f
3.78E-01
7.42E-01
5.6%
11.0%
T etrachlorobiphenyls
1
1.6% b
1.06E-01
1.6%
PCB 118 (penta)
1
4.4% d'e
11% d'f
5.79E-02
1.39E-01
4.4%
10.6%
Pentachlorobiphenyls
1
0.6% b
8.20E-03
0.6%
Hexachlorobiphenyl
1
0.042% 3
0.29% b
7.60E-05
5.31E-04
0.0%
0.29%
Moderately PCB contaminated8
1
1.8% 8
3.23
1.1%
Note:
3 Borglin et al. (1996)
b Ghosh etal. (2000)
c ten Hulscher et al. (1999; 2002)
d Cornelissen et al. (1997)
e k is for 2 day contact time
k is for 34 day contact time
8 Carrolletal. (1994).Moderately PCB contaminated sediment contained 64 mg/kg (dry weight) PCBs,
with total organic carbon of 3.43%. The PCB presents in the sediments consisted of primarily mono- and
di-chlorinated biphenyls (60-70% or total).
h Assumed equilibrium was achieved
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 9
Summary of Field Samples and Analytical Data
from the Pre-Design Field Test - Dredge Technology Evaluation Report (8/6/2001)
Date
Type
Northing
Easting
Hour
Min
Turbidity (NTU)
Max Min Avg
TSS
(mg/L)
Total PCBs (ug/L) 18 Congeners
Particulate Dissolved Particulate+
Dissolved
Fraction
Particulate
Fraction
Dissolved
8/7/00
Grab
2704955
815354
16
26
Background Value - Acushnet Estuary 1000ft N
10
0.89
0.52
1.41
0.63
0.37
8/7/00
Grab
2703124
815820
16
36
Background Value - Acushnet Estuary 1000ft S
4
0.25
0.18
0.43
0.58
0.42
8/15/00
Grab
2704040
815356
17
52
Turbidity/TSS - Acushnet Estuary
26
26
26
53
8/15/00
Grab
18
5
Turbidity/TSS - Acushnet Estuary
12
12
12
22
8/15/00
Grab
18
8
Turbidity/TSS - Acushnet Estuary
3
5
4
5
8/16/00
Grab
2703129
815608
9
20
Up-Current reference sample
3
6
4.5
6
0.11
0.21
0.32
0.34
0.66
8/16/00
EBB
11
56
Sampling HR1 - Station 1 (50ft)
7
10
8.5
20
8/16/00
EBB
2703959
815530
12
2
Sampling HR1 - Station 2 (100ft)
16
21
18.5
24
8/16/00
EBB
2703621
815717
12
11
Sampling HR1 - Station 3 (500ft)
5
12
8.5
17
8/16/00
EBB
2704948
815379
12
22
Sampling HR1 - REF (1000ft up-current)
3
12
7.5
9
8/16/00
EBB
13
16
Sampling HR2 - Station 1 (50ft)
11
8/16/00
EBB
2703833
815506
14
6
Sampling HR2 - Station 2 (100ft)
43
8/16/00
EBB
2703647
815675
14
15
Sampling HR2 - Station 3 (500ft)
11
8/16/00
EBB
2704948
815379
14
22
Sampling HR2 - REF (1000ft up-current)
12
8/16/00
Composite
Composite Station 1
16
1.3
0.77
2.07
0.63
0.37
8/16/00
Composite
Composite Station 2
27
2.1
0.79
2.89
0.73
0.27
8/16/00
Composite
Composite Station 3
23
27
25
12
0.85
0.75
1.6
0.53
0.47
8/16/00
Composite
Composite -REF
10
17
13.5
9
0.89
0.9
1.79
0.50
0.50
8/16/00
FLOOD
2703995
815351
16
59
Sampling HR1 - Station 1 (50ft)
20
8/16/00
FLOOD
2704110
815393
17
17
Sampling HR1 - Station 2 (100ft)
20
20
20
17
8/16/00
FLOOD
2704375
815410
17
23
Sampling HR1 - Station 3 (500ft)
40
40
40
25
8/16/00
FLOOD
2702780
815578
17
44
Sampling HR1 - REF (1000ft up-current)
6
15
10.5
6
8/16/00
FLOOD
2704028
815329
17
56
Sampling HR2 - Station 1 (50ft)
21
27
24
12
8/16/00
Grab
17
56
Surface oil slick observed at HR1 - Station 1 (50ft)
8/16/00
FLOOD
2704140
815363
17
58
Sampling HR2 - Station 2 (100ft)
10
15
12.5
13
1.5
8/16/00
FLOOD
2704375
815410
18
19
Sampling HR2 - Station 3 (500ft)
39
42
40.5
9
8/16/00
FLOOD
2702780
815578
18
40
Sampling HR2 - REF (1000ft up-current)
38
42
40
7
8/16/00
Composite
Composite Station 1
27
2.6
0.66
3.26
0.80
0.20
8/16/00
Composite
Composite Station 2
10
0.99
0.58
1.57
0.63
0.37
8/16/00
Composite
Composite Station 3
16
1.1
0.52
1.62
0.68
0.32
8/16/00
Composite
Composite -REF
5
0.25
0.36
0.61
0.41
0.59
8/17/00
EBB
10
58
Sampling - Up-Current reference sample
23
27
25
5
0.29
0.46
0.75
0.39
0.61
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 10
Dissolved and Particulate Percent PCB Mass Loss
Dissolved Phase Maximum
Max >=100', no flood
0.95 ug/L
minus background
0.52 ug/L
0.43 ug/L
Maximum Flow Rate
10 cm/s
3.9
in/s
0.3
ft/s
wide
800 ft
deep
8.75 ft
Maximum Flow Rate
2297 cfs
2.8E-02
m3/cf
65.0
m3/s
65 m3/s
1000
L/m3
65032
Us
65032 Us
X
0.43 ug/L
27964 ug/s
Mass loss/second
2.8E-05 kg/s
time worked
17.5 hrs
3600
s/hr
63000
s
2.8E-05 kg/s
X
63000 s
PCB mass loss
1.8 kg
PCBs removed
1495 kg
Dissolved Phase Percentage
0.1%
Particulate Phase Maximum
Max >=100', no flood
2.6 ug/L
minus background
0.89 ug/L
1.71 ug/L
Maximum Flow Rate
10 cm/s
3.9
in/s
0.3
ft/s
wide
800 ft
deep
8.75 ft
Maximum Flow Rate
2297 cfs
2.83E-02
m3/cf
65.0
m3/s
65 m3/s
1000
L/m3
65032
Us
65032 Us
X
1.71 ug/L
111205 ug/s
Mass loss/second
1.1E-04 kg/s
time worked
17.5 hrs
3600
s/hr
63000
s
1.1E-04 kg/s
X
63000 s
PCB mass loss
7.0 kg
PCBs removed
1495 kg
Particulate Phase Percentage
0.5%
Percent Dissolved
20%
Percent Particulate
80%
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 9 Cont'd
Date
Type
Northing
Easting
Hour
Min
Turbidity (NTU)
Max Min Avg
TSS
(mg/L)
Total PCBs (ug/L) 18 Congeners
Particulate Dissolved Particulate+
Dissolved
Fraction
Particulate
Fraction
Dissolved
8/17/00
EBB
2703878
815379
11
7
Sampling HR1 - Station 1 (50ft)
11
18
14.5
6
8/17/00
EBB
2702964
815758
11
42
Sampling HR1 - Station 4 (1000ft)
10
17
13.5
12
8/17/00
EBB
2703218
815599
11
46
Sampling HR1 - Station 3 (700ft)
10
17
13.5
17
8/17/00
EBB
2703625
815534
11
50
Sampling HR1 - Station 2 (300ft)
11
18
14.5
12
8/17/00
EBB
2704948
815379
11
59
Sampling HR1 - REF (1000ft up-current)
9
18
13.5
9
8/17/00
EBB
2702964
815758
12
32
Sampling HR2 - Station 4 (1000ft)
6
10
8
8
8/17/00
EBB
2703218
815599
12
38
Sampling HR2 - Station 3 (700ft)
12
17
14.5
11
8/17/00
EBB
2703625
815534
12
45
Sampling HR2 - Station 2 (300ft)
11
17
14
15
8/17/00
EBB
2703878
815379
12
52
Sampling HR2 - Station 1 (50ft)
9
15
12
11
8/17/00
EBB
2704948
815379
13
1
Sampling HR2 - REF (1000ft up-current)
5
12
8.5
7
8/17/00
Grab
13
45
MIAMI II Plume (peak field turbidity)
60
70
65
300
26
2.7 28.7
0.91
0.09
8/17/00
EBB
2703878
815379
13
48
Sampling HR3 - Station 1 (50ft)
28
34
31
62
8/17/00
EBB
2703625
815534
13
58
Sampling HR3 - Station 2 (300ft)
19
23
21
29
8/17/00
EBB
2703218
815599
14
3
Sampling HR3 - Station 3 (700ft)
13
18
15.5
18
8/17/00
EBB
2702964
815758
14
8
Sampling HR3 - Station 4 (1000ft)
13
21
17
21
8/17/00
EBB
2704948
815379
14
38
Sampling HR3 - REF (1000ft up-current)
9
12
10.5
10
8/17/00
EBB
2703878
815379
14
47
Sampling HR4 - Station 1 (50ft)
26
29
27.5
39
8/17/00
EBB
2703625
815534
14
53
Sampling HR4 - Station 2 (300ft)
19
26
22.5
31
8/17/00
EBB
2703218
815599
14
57
Sampling HR4 - Station 3 (700ft)
27
29
28
37
8/17/00
EBB
2702964
815758
15
3
Sampling HR4 - Station 4 (1000ft)
13
18
15.5
22
8/17/00
Composite
Composite Station 1
10
16
13
19
2
2.7 4.7
0.43
0.57
8/17/00
Composite
Composite Station 2
21
29
25
21
2.2
0.83 3.03
0.73
0.27
8/17/00
Composite
Composite Station 3
18
24
21
18
1.3
0.79 2.09
0.62
0.38
8/17/00
Composite
Composite Station 4
20
24
22
15
1
0.67 1.67
0.60
0.40
8/17/00
Composite
Composite -REF
13
18
15.5
9
0.61
0.78 1.39
0.44
0.56
8/17/00
FLOOD
2704000
815324
16
49
Sampling HR1 - Station 1 (50ft)
13
16
14.5
17
8/17/00
FLOOD
2704266
815441
17
6
Sampling HR1 - Station 2 (300ft)
14
19
16.5
20
8/17/00
FLOOD
2704727
815455
17
12
Sampling HR1 - Station 3 (700ft)
60
70
65
210
8/17/00
FLOOD
2705097
815357
17
18
Sampling HR1 - Station 4 (1000ft)
10
13
11.5
10
8/17/00
FLOOD
2702805
815548
17
33
Sampling HR1 - Station 5 (1000ft up-current)
6
13
9.5
9
8/17/00
FLOOD
2704000
815321
18
0
Sampling HR2 - Station 1 (50ft)
6
13
9.5
8
8/17/00
FLOOD
2704266
815441
18
6
Sampling HR2 - Station 2 (300ft)
15
18
16.5
15
8/17/00
FLOOD
2704727
815455
18
12
Sampling HR2 - Station 3 (700ft)
11
19
15
16
8/17/00
FLOOD
2705097
815357
18
15
Sampling HR2 - Station 4 (1000ft)
12
17
14.5
14
8/17/00
FLOOD
2702805
815548
18
30
Sampling HR2 - REF (1000ft up-current)
11
13
12
6
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Table 9 Cont'd
Date
Type
Northing
Easting
Hour
Min
Turbidity (NTU)
Max Min Avg
TSS
(mg/L)
Total PCBs (ug/L) 18 Congeners
Particulate Dissolved Particulate+
Dissolved
Fraction
Particulate
Fraction
Dissolved
8/17/00
FLOOD
2704000
815321
19
4
Sampling HR3 - Station 1 (50ft)
12
15
13.5
13
8/17/00
FLOOD
2704266
815441
19
8
Sampling HR3 - Station 2 (300ft)
11
16
13.5
20
8/17/00
FLOOD
2704727
815455
19
12
Sampling HR3 - Station 3 (700ft)
8
13
10.5
11
8/17/00
FLOOD
2705097
815357
19
16
Sampling HR3 - Station 4 (1000ft)
12
19
15.5
19
8/17/00
FLOOD
2072805
815548
19
33
Sampling HR3 - REF (1000ft up-current)
4
9
6.5
3
8/17/00
Composite
Composite Station 1
11
0.91
0.55
1.46
0.62
0.38
8/17/00
Composite
Composite Station 2
16
1.6
0.77
2.37
0.68
0.32
8/17/00
Composite
Composite Station 3
18
2.6
0.95
3.55
0.73
0.27
8/17/00
Composite
Composite Station 4
12
1.1
0.92
2.02
0.54
0.46
8/17/00
Composite
Composite -REF
6
0.38
0.56
0.94
0.40
0.60
8/18/00
Grab
10
48
Sample Up-current-reference (Event scrubbed)
10
15
12.5
6
0.13
0.22
0.35
0.37
0.63
8/18/00
Grab
17
44
Sample inside moonpool during active dredging
44
50
47
120
23
4.6
27.6
0.83
0.17
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Figures
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment C - April 2004
-------
Figure 1
PCB, TSS and Turbidity vs. Distance from the Dredge
O)
D
in
ca
o
a.
-20
Particulate PCBs vs.Distance
499-
-49-
00
~ -i&c
-0r4-
tuu .
1000
2000
Distance from Dredge (ft)
Baseline shown at +/-1500 ft
Particulate PCBs vs.Distance
without Samples at the Dredge
Distance from Dredge (ft)
Baseline shown at +/-1500 ft
TSS vs.Distance
TSS vs.Distance
without Samples at the Dredge
O)
E
w
V)
100D~
<
inn i
~
~
~ in
~ 9 10
, q
U)
E
w
V)
r 49-
on
on
ซ~
* ~ *
~ in
\ ~ ~ ~
4 ~ ~ ~
~ 1 10
, 9-
-2000
-1000
1000
2000
-2000
-1000
1000
Distance from Dredge (ft)
Baseline shown at +/-1500 ft
Distance from Dredge (ft)
Baseline shown at +/-1500 ft
2000
Dissolved PCBs vs.Distance
Turbidity vs.Distance
O)
D
in
CQ q/~
O ^L
D.
-49-
iIdc
00 -1909
~ ~
t
~ ~ ~
1000 2000
94-
Distance from Dredge (ft)
Baseline shown at +/-1500 ft
O)
E
ฆg
!d
D
on
fin
~
ah
~
~
A ฐn
~~~ ~
w
~
w
, 9-
-2999 -1999 9 1999
Distance from Dredge (ft)
2999
Dissolved PCBs Fraction vs.Distance
Particulate PCBs Fraction vs.Distance
in
ca
o
a.
n qo
X n Rn
ป
~
n An
n ฐn ?
~
, 9-99
in
ca
o
a.
i on A
n fin
^ ~ n Rn
t ~
n 4n
~
n ฐn
T 0709
-2000
-1000
1000
2000
-2000
-1000
1000
2000
Distance from Dredge (ft)
Baseline shown at +/-1500 ft
Distance from Dredge (ft)
Baseline shown at +/-1500 ft
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment C - April 2004
-------
Attachment D
Modeling Analysis
Table of Contents
1.0 Introduction 1
2.0 Objectives 2
2.1 Near-Field Modeling 2
2.2 Far-Field Modeling 2
3.0 Selection of the Transport Models 3
3.1 Interaction Among the Transport Models 4
4.0 Near-Field Modeling 6
4.1 Parameters 6
4.1.1 Settling Velocities 7
4.1.1.1 Literature S earch 7
4.1.1.2 Selection of Settling Velocity 10
4.1.2 Lateral Dispersion Coefficient 11
4.2 RY1A2 12
4.2.1 Methodology 13
4.2.2 Results of RMA2 14
4.3 CSTR-Chem 14
4.3.1 Methodology 14
4.3.2 Results 20
4.3.3 Sensitivity Analysis 22
4.4 TSS-Chem 25
4.4.1 Methodol ogy 25
4.4.2 Relationship Between CSTR-Chem and TSS-Chem 34
4.4.3 Results 35
4.4.3.1 Average Source Strength Estimations 36
4.4.3.2 Particle Settling Results 39
4.4.3.3 Suspended Solids Near-field Standards and Monitoring
Locations 40
4.4.4 Sensitivity Analyses 44
4.4.4.1 Fine and Coarse-grained PCB Distributions 45
4.4.4.2 TSS-Chem Model Sensitivity Analysis 47
5.0 Far-Field Modeling 55
5.1 HUDTOX and FISHRAND: Fate, Transport, and Bioaccumulation
Modeling to Simulate the Effect of the Remedial Alternative 55
5.1.1 HUDTOX Input Values 56
5.1.2 Methodology 63
Hudson River PCB s Superfund Site
Engineering Performance Standards
l
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Attachment D
Modeling Analysis
Table of Contents
5.1.3 HUDTOX Input Study and Relationship Between Resuspension
Release and Export Rates 64
5.1.4 HUDTOX Results 71
5.1.5 FISHRAND Results for the Upper and Lower River 78
5.2 Relative Reduction In Human Health And Ecological Risks In The Upper
And Lower Hudson River 79
5.2.1 Introduction 79
5.2.2 Human Health Risk Reduction 80
5.2.2.1 Upper Hudson River 80
5.2.2.2 Mid-Hudson River 81
5.2.3 Ecol ogi cal Ri sk Reducti on 81
5.2.3.1 Upper Hudson River 81
5.2.3.2 Lower Hudson River 82
5.2.4 Conclusions 82
5.3 Suspended Solids Far-Field Standards 82
6.0 Modeling Studies Used 84
6.1 New Bedford Harbor Pre-Design Field Test Dredge Technology
Evaluation Report 84
6.2 Manistique River and Harbor, Michigan 86
7.0 Response to GE's Comments on Hudson River FS 88
7.1 Summary of GE's Conceptual Model and Results 88
7.2 Kinetics of PCB Desorption: Literature Review 88
7.3 CSTR-Chem Model 90
8.0 Case Studies - Dissolve Phase Releases and Export Rates 91
8.1 Introduction 91
8.2 New Bedford Harbor, Massachusetts 92
8.3 Fox River SMU 56/57 1999 And 2000 Dredging Projects, Wisconsin 94
8.4 Hudson Falls 96
8.5 Other Sites 97
8.6 Conclusions 98
9.0 References 99
Hudson River PCB s Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Attachment D
Modeling Analysis
Table of Contents
LIST OF TABLES
Table 1 Properties of Hudson River Sediments
Table 2 Summary of Settling Velocities
Table 3 Surface Water Elevation Slope in TI Pool based on USGS Gauge Data
Table 4 Estimated Shear Velocity and Lateral Dispersion Coefficient for Upper
Hudson River
Table 5 Silt Fractions in Hudson River Sections
Table 6 Summary of CSTR-Chem Model simulation results for dredging operations in
Section 1-3 of the Hudson River
Table 7 Summary of Sensitivity of Model Outputs to Model Parameter Inputs
Table 8 TSS-Chem Model Runs for the PCB 350 ng/L far-field Criterion with and
without Dissolved PCBs from Dredging as Modeled by CSTR-Chem
Table 9 TSS- Chem Model Runs for the PCB 350 ng/L far- field Criterion with and
without Coarse solids from Dredging as Modeled by CSTR- Chem
Table 10 Results for Average Source Strength Estimated Fluxes
Table 11 Increase in PCB Mass from Settled Material Estimated Using the TSS-Chem
Model Results
Table 12 TSS Average Concentration within the Plume at 300 m Downstream and
under 8000 cfs Flow
Table 13 Average Source Strength Estimated Fluxes and Concentrations for River
Section 1 with Various Flows and Total PCB Sediment Concentrations
Table 14 Range of Values and Relative Sensitivities of Each Parameter
Table 15 Effect on Model Output Values from Increase in Input Parameters
Table 16 Average Sensitivity Values and Individual magnitudes
Table 17 Average Baseline Conditions at Thompson Island Dam
Table 18 Average Baseline Conditions at Schuylerville
Table 19 Average Baseline Conditions at Waterford
Table 20 Daily Net Dredging Total PCB Flux for River Sections 1, 2, and 3 at the
Monitoring Stations
Table 21 HUDTOX Input for 350 ng/L with TSS Flux at 1 Mile Downstream of the
Dredge- Head
Table 22 TSS Flux Comparisons for Different Scenarios
Table 23 HUDTOX Input for 350 ng/L with TSS Flux at 1 Mile Downstream of the
Dredge- Head and Corrected Percent Reduction at the Monitoring Stations
Table 24 HUDTOX Schedule and Input Loading for 300 g/day Export Rate Scenario
Table 25 HUDTOX Schedule and Input Loading for 600 g/day Export Rate Scenario
Table 26 Percent Reduction at the Monitoring Locations Comparison for the 350 ng/L
Table 27 Expected versus Model Prediction of PCB Flux for Control Level 3-350
ng/L Scenario
Table 28 Annual Tri+ PCB Load Over TID
Hudson River PCBs Superfund Site
Engineering Performance Standards
ill
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Attachment D
Modeling Analysis
Table of Contents
LIST OF TABLES (continued)
Table 29 Tri+ PCB Load Over Schuylerville
Table 30 Tri+ PCB Load Over Waterford
Table 31 Resuspension Production, Release, and Export Rates from TSS- Chem and
HUDTOX Models
Table 32 Example of CSTR-Chem, TSS-Chem, and HUDTOX Application
Table 33 Expected versus Model Prediction of PCB Flux for Control Level - 600 g/day
Scenario
Table 34 Expected versus Model Prediction of PCB Flux for Evaluation Level - 300
g/day Scenario
Table 35 FISHRAND Forecast for Year to Reach Fish Tissue Concentration Difference
of 0.5 mg/kg Relative to the No Resuspension - Upper River
Table 36 FISHRAND Forecast for Year to Reach Fish Tissue Concentration Difference
of 0.05 mg/kg Relative to the No Resuspension - Lower River
Table 37 Upper Hudson Species-Weighted Fish Fillet Average PCB Concentration (in
mg/kg)
Table 38 Upper Hudson River Modeled Times (Years) of Compliance with Human
Health Risk-Based Concentrations Resuspension Scenarios
Table 39 Resuspension Scenarios - Long-Term Fish Ingestion Reasonable Maximum
Exposure and Central Tendency PCB Non-Cancer Hazard Indices Upper
Hudson River Fish - Adult Angler
Table 40 Resuspension Standard Scenarios - Long-Term Fish Ingestion Reasonable
Maximum Exposure and Central Tendency Cancer Risks Upper Hudson River
Fish - Adult Angler
Table 41 Mid-Hudson River Species-Weighted Fish Fillet Average PCB Concentration
(in mg/kg)
Table 42 Upper Hudson River Average Largemouth Bass (Whole Fish) PCB
Concentration (in mg/kg)
Table 43 Modeled Times of Compliance with River Otter Risk-Based Fish
Concentrations Upper Hudson River
Table 44 Lower Hudson River Average Largemouth Bass (Whole Fish) PCB
Concentration (in mg/kg)
Table 45 Modeled Times of Compliance with River Otter Risk-Based Fish
Concentrations Lower Hudson River
Table 46 Sediment Characteristics
Table 47 Impact of Dispersion Coefficient on Predicted Peak Concentration and Length
of Suspended Sediment Plume
Hudson River PCB s Superfund Site
Engineering Performance Standards
iv
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Attachment D
Modeling Analysis
Table of Contents
LIST OF FIGURES
Figure 1 Interaction Among the Transport Models
Figure 2 Sensitivity of Net Dissolved and Silt Fractions Exiting Near-Field with
Variations in Linear Velocity and Depth for CSTR-Chem
Figure 3 Sensitivity of Net Total PCB Flux and Net TSS Flux Exiting Near-Field with
Variations in Linear Velocity and Depth for CSTR-Chem
Figure 4 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Velocity for CSTR-Chem
Figure 5 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Depth for CSTR-Chem
Figure 6 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Near-Field Width for CSTR-
Chem
Figure 7 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Resuspension Rate for CSTR-
Chem
Figure 8 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Sediment Silt Fraction for
CSTR-Chem
Figure 9 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of PCB Sediment Concentration for
CSTR-Chem
Figure 10 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Dissolved PCB Fraction in the
Background and TSS Background Concentrations for CSTR-Chem
Figure 11 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Dissolved PCB Fraction in the
Background and Kd Value for CSTR-Chem
Figure 12 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Desorption Rate for CSTR-
Chem
Figure 13 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Background PCB Concentration
for CSTR-Chem
Figure 14 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Silt Settling Velocity for CSTR-
Chem
Hudson River PCB s Superfund Site
Engineering Performance Standards
v
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Attachment D
Modeling Analysis
Table of Contents
LIST OF FIGURES (continued)
Figure 15 Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net
TSS Flux Exiting Near-Field as Functions of Coarse Settling Velocity for
CSTR-Chem
Figure 16 Estimated TSS Concentration Downstream of the Dredge Head in Section 1
(Flow is 4000 cfs and PCB concentration is 500 ng/L at the far field station
Figure 17 Estimated TSS Concentration at 300 m Downstream of the Dredge Head in
Section 1 (PCB concentration at the far-field station is 500 ng/L)
Figure 18 Total PCBs Grouped by Total Organic Carbon (Figure 3-21 of LRC Report)
Figure 19 Grain Size, Organic Content and PCB Concentrations in Hudson River
Sediment collected near Moreau
Figure 20 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Riverwide Volumetric Flow
(Velocity- Depth Pairs) for the TSS- Chem
Figure 21 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Velocity for the TSS- Chem
Figure 22 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Depth for the TSS- Chem
Figure 23 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Source Strength for the
TSS- Chem
Figure 24 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Silt Fraction Entering for the
TSS- Chem
Figure 25 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Sediment PCB
Concentration for the TSS- Chem
Figure 26 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of TSS Background and PCB
Dissolved Fraction (Kd = 55,000) for the TSS- Chem
Figure 27 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Kd for the TSS- Chem
Figure 28 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Desorption Rate for the
TSS- Chem
Figure 29 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Lateral Dispersion for the
TSS- Chem
Hudson River PCBs Superfund Site
Engineering Performance Standards
vi
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Attachment D
Modeling Analysis
Table of Contents
LIST OF FIGURES (continued)
Figure 30 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of PCB Background
Concentration for the TSS- Chem
Figure 31 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Silt Settling Velocity for the
TSS- Chem
Figure 32 Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux
and Net TSS Flux at 1600 meters as Functions of Sand Settling Velocity for
the TSS- Chem
Figure 33 PCB Concentrations Downstream of Dredge for 350 ng/L scenario Section 1
at 1 mile and 3 miles
Figure 34 Whole Water Total PCB Concentration for Different 350 ng/L Input
Formulations
Figure 35 Tri+ PCB Cumulative Load for Different Dredging Scenarios
Figure 36 Total PCB Cumulative Load for Different Dredging Scenarios
Figure 37 Whole Water, Particulate, and Dissolved Total PCB Concentrations for 350
ng/L Dredging Scenario (sr04)
Figure 38 Whole Water, Particulate and Dissolved Total PCB Concentration for Control
Level - 600 g/day Total PCB Flux Dredging Scenario (srOl)
Figure 39 Tri+ PCB and Total PCB Cumulative Load for 600 g/day (srOl) Scenario
Figure 40 HUDTOX Forecast of Whole Water, Particulate, and Dissolved Total PCB
Concentrations for Evaluation Level - 300 g/day Scenario
Figure 41 Comparison Between Upper Hudson River Remediation Scenario (Various
Export Rates) and Monitored Natural Attenuation (MNA) Forecast for
Thompson Island Dam, Schuylerville, and Waterford
Figure 42 Total PCB Concentrations at Waterford for the Accidental Release Scenario
Figure 43 Composite Fish Tissue Concentrations for the Upper Hudson River
Figure 44 Composite Fish Tissue Concentrations for the Lower Hudson River
Figure 45 PCB Concentrations for New Bedford Harbor Pilot Dredging Study
Figure 46 Dissolved Fraction of PCBs for New Bedford Harbor Pilot Dredging Study
Figure 47 TSS and PCB Concentrations for New Bedford Harbor Pilot Dredging Study
Hudson River PCBs Superfund Site vii Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment D - April 2004
-------
Attachment D
Modeling Analysis
1.0 Introduction
Modeling of conditions expected during dredging operations was undertaken to evaluate
the short and long-term effects of remedial activities. Far-field models - consisting of
fate, transport and bioaccumulation models - were utilized to measure the long-term
effects of dredging and to determine the percent PCB mass loss that will result in
unacceptable river recovery and adverse impacts to downstream water supply intakes. In
addition to far-field modeling, near-field modeling was conducted to simulate dredging
and resulting river conditions near the dredge bucket/head and up to a mile downstream.
One near-field model (TSS-Chem) was used to estimate PCB water column conditions in
a lateral direction from the dredge (across the width of the river) up to one mile
downstream. The modeling results were used to aid in the determination of the best
location for monitoring points, the water column concentration near sensitive locations,
settling effects and rates of PCB flux for use in the long-term models. A second near-field
model (CSTR-Chem) was developed assuming that the conditions near the dredge are
similar to a continuous stirred tank reactor (CSTR). The model provided a basis for
assumptions regarding the dissolved phase PCB concentrations in the immediate vicinity
of the dredge.
Hudson River PCB s Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
2.0 Objectives
2.1 Near-Field Modeling
Near-field modeling was completed to simulate water column suspended solids and total
PCB concentrations in the vicinity of the dredge. The downstream models were applied
to determine the following:
Estimate monitoring locations for suspended solids and turbidity;
Estimate plume geometry of the resuspended sediment (sediment transport and
flux in close proximity to the dredge);
Estimate depositional patterns of the settled resuspended sediment, thickness of
the deposited material, and its impact on surficial sediments that are deposited
downstream;
Evaluate the potential PCB dissolved phase release downstream of the dredge.
2.2 Far-Field Modeling
Far-field modeling was completed to simulate water column, sediment and fish total PCB
concentrations in the Upper and Lower Hudson River as a result of the dredging
operation. The far-field model was applied to determine the following:
Estimate the impact of contaminant mass loss from resuspension during
remediation and its effect on water column concentrations at public water intakes;
Determine the acceptable mass loss for protection on downstream water resources
and public water intakes;
Evaluate the impact of accidental release scenario on resulting water column
concentrations at public water intakes and on the recovery of the river.
Hudson River PCB s Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
3.0 Selection of the Transport Models
Dredging operations are expected to release suspended sediment and PCBs into the water
column. As a result, modeling was needed to estimate the duration and intensity of these
impacts at sensitive downstream locations. Sensitive locations include the immediate
dredging area and downstream water supply intakes. Modeling at multiple scales was
conducted to estimate these impacts at all locations in the river system.
A far-field model was necessary to predict PCB concentrations over the extent of the
remediated area and downstream into the Lower Hudson River. The far-field model was
capable of estimating PCB concentrations during the years of dredging activities as well
as several years following the completion of dredging. In contrast, a near-field model
capable of estimating PCB water column concentrations over a short period of time
(weeks or months) was required to simulate river conditions in the vicinity of the dredge.
During preparation of the Hudson River Feasibility Study (FS) report (USEPA, 2000a)
and the Hudson River Responsiveness Summary (RS) report (USEPA, 2002), the USEPA
water quality model, HUDTOX, was developed to project current river conditions into
the future for comparison against model runs where active remediation such as capping
and dredging were simulated. This model forecasts future water column and sediment
PCB concentrations for various scenarios so the benefit of active remediation versus
monitored natural attenuation (MNA) could be compared and evaluated. The results of
the HUDTOX model were then utilized as input for the FISHRAND model to evaluate
fish bioaccumulation PCB levels as a result of the various scenarios. This model,
HUDTOX, was used to estimate far-field river and sediment concentrations for various
scenarios to allow for the development of a protective resuspension performance
standard.
An evaluation was conducted to determine if HUDTOX could be applied to simulate
dredging conditions near the dredge (near-field modeling) since HUDTOX already
reflects the conditions of the Hudson River. However, HUDTOX could not be readily
modified to obtain adequate resolution for estimating near-field river conditions,
therefore other models have been developed specifically for the near-field modeling.
A US ACE model, SED2D, was evaluated for use as the near-field model since it has
been proven to simulate near-field dredging conditions with similar accuracy as the
HUDTOX model only in a much shorter time frame. SED2D is part of the TABS-MD
(multi-dimensional) modeling system that was used in the development of HUDTOX. It
is a two-dimensional model that can be used for depth-averaged transport of cohesive or a
representative grain size of non-cohesive sediments and the deposition, erosion, and
formation of bed deposits. Until 1995, this model was distributed under the name of
STUDH. Sediment loading and bed elevation changes can be calculated when supplied
with a hydrodynamic solution computed by the model RMA2. RMA2 is a hydrodynamic
model that supports sub-critical flow analysis. The SED2D and STUDH models were not
selected for use, because of the limitations of the model, including modeling a single type
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
of solids. RMA2 was used to estimate the linear water velocities and depths at various
flowrates.
The near-field model used previously in the FS and ROD was DREDGE. DREDGE is a
module of the Automated Dredging and Disposal Alternatives Modeling System
(ADDAMS) distributed by the USACE through the Environmental Laboratory, USAE
Research and Development Center Waterways Experiment Station. DREDGE estimates
the rate at which bottom sediments become suspended into the water column as the result
of dredging operations and the resulting suspended sediment concentrations. TSS-Chem
was developed to model the downstream transport of solids and PCBs through the near-
field in the Hudson River. TSS-Chem is similar to the DREDGE model described in
Appendix E of the FS. It applies the same Gaussian plume for solids transport as
DREDGE but is able to model both coarse and fine solids and includes two phase
partitioning of PCBs from the solids into the dissolved phase. However, unlike the
DREDGE model, TSS-Chem is only applicable for dredging activities with 4-cy dredge
buckets. The TSS-Chem model provides estimates of PCB and solids concentrations and
fluxes across the river width from 10 meters downstream to approximately one mile
downstream.
Since TSS-Chem is unable to estimate conditions directly around the dredge bucket, a
second near-field model was necessary. CSTR-Chem models the area directly around the
dredge bucket as a continuous stirred tank reactor. The conditions in this area are
essential to the loading of TSS-Chem. By estimating the surroundings of the dredge
bucket, a basis for assumptions regarding the solids source of TSS-Chem was obtained.
3.1 Interaction Among the Transport Models
The main goal of the modeling effort is to study the long-term impacts of dredging
operations in the Upper and Lower Hudson River. As part of this, fish tissue recovery can
provide a threshold or limit to define an unacceptable impact due to dredging releases and
thereby a limit on the export rate is needed. The modeling efforts were focused on
examining the impact of running the dredging operation at the specified action levels in
the Resuspension Standard. The resuspension scenarios for the Resuspension Standards
are specified as the PCB export rate at the far-field monitoring stations. The
HUDTOX/FISHRAND model cannot be used for this purpose strictly since HUDTOX is
not designed to simulate the process of dredging releases. Due to the nature of the
HUDTOX model structure, PCB loads cannot be readily specified at far-field locations
(i.e., specifying the resuspension export rate). Rather, the input of PCBs is specified as an
input load at a location within the river, equivalent to a resuspension release rate. In order
to create a correctly loaded HUDTOX run, it is first necessary to estimate the local
resuspension release rate from the dredging operation; that is, the rate of Tri+ PCB, Total
PCB and solids transport at the downstream end of the dredge plume. At this location
most of the solids that are going to settle out, will have settled out and the suspended
solids will more closely resemble those simulated by HUDTOX. To estimate the input
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
loading term for HUDTOX, the two models designed to address the dredging release
process and near-field transport, CSTR-Chem and TSS-Chem, were used.
The three models were used to represent and link the three different scales of
resuspension. The immediate vicinity of the dredge (10 m radius) is simulated by the
CSTR-Chem. The region from the dredge to a distance of one mile (10 to 1610 m) is
represented by TSS-Chem with its solids transport and geochemical model. Finally, the
region beyond one mile is represented by HUDTOX. The choice of the TSS-Chem model
to represent a one-mile interval is related to the size of the individual HUDTOX cell,
which is approximately 2/3 of a mile long. Figure 1 shows the links among the transport
models and the different scales of resuspension they represent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
4.0 Near-Field Modeling
The near-field models are useful in determining the appropriate locations for monitoring
stations and provide a practical basis for defining criteria by estimating resuspension rates
that correspond to various action level scenarios. The resuspension rates were compared
to production rates and the ability to realistically resuspend solids at such rates from
dredge bucket operations were examined.
4.1 Parameters
The parameters required for HUDTOX and other long-term models are not directly
applicable to the near-field models. Many of the HUDTOX parameters were developed
empirically for long-term conditions. The near-field models only apply to periods of
dredge activities. Therefore, the parameters applied for use in the near-field models were
chosen based on extensive literature research, consideration of the unique conditions
found in the Upper Hudson River and a tendency towards conservative (greater release)
estimates.
For the near-field model simulations, the concentration of PCBs on the suspended
particles was estimated as the average sediment concentrations of the removed material
for each river section including the overcut. While in the water column the PCBs undergo
two-phase partitioning from the suspended to dissolved phase. The partitioning of the
PCBs between the two phases is based on the partition coefficient which dictates the
equilibrium fractions of the phases and the desorption rate which will determine how
quickly equilibrium is approached. The selection of the partition coefficient and the
desorption rate is discussed in Attachment C since they are not exclusively used for these
models.
With a given partition coefficient and desorption rate the time available for partitioning
will control the amount of desorption that occurs. The time that the particles remain
suspended is primarily a function of the sediment type. Generally the silt particles will
remain suspended longer than the coarse particles. In the model, the rate at which
particles fall through the water column is determined by the particle settling velocity. The
model includes different settling velocities for fine and coarse particles. In addition to the
time constraint, the concentration of suspended PCBs within the plume will also affect
the equilibrium conditions. In the TSS-Chem model dispersion of the solids within the
plume and thereby the concentration is dictated by the lateral dispersion coefficient. The
selection of both the settling velocities and lateral dispersion coefficient is discussed
below.
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
4.1.1 Settling Velocities
To accurately represent the solids concentrations and the time available for partitioning in
the CSTR-CHEM and TSS-CHEM models, settling velocities for both fine and coarse
resuspended sediments were researched. Eight references were examined and considered
in the selection of the settling velocities for the two models. The selection process took
into account the applicability of the studies to the Hudson River sediments and the
inclusion of significant dynamic aspects of settling solids {i.e., flocculation) in the
studies. Previous data analyses have been completed to define and characterize the
Hudson River sediments and the typical properties of the sediments are summarized in
Table 1.
4.1.1.1 Literature Search
As part of a literature search the following references that reported or used settling
velocities were examined:
(1) Estimating the Size-Dependent Settling Velocity of Suspended Particles Using
the LISST-ST. (Sequoia Scientific, Inc.)
The LISST-ST is a particle counter manufactured by Sequoia Scientific, which is
employed in the water column of rivers and used to count particle sizes and
measure the time it takes for the particle to settle out in the chamber of the
instrument. This data is then used to estimate the particle settling velocity. Data
generated from field studies is indicative of:
For particle of size 50 microns, Vs = 0.01 cm/s
For a particle of size 100 microns, Vs = 0.10 cm/s
For a particle of size 400 microns, Vs = 0.005 cm/s
(2) Transport and Transformation of Contaminants Near the Sediment-Water
Interface. (DePinto etal., 1994)
This reference examined both freshwater and saltwater sediment particles for
slightly flocculent New Bedford Harbor sediment and highly flocculent Passaic
Valley Sewage Sludge. Data generated from this study indicated:
New Bedford Harbor Freshwater sediment with a particle size of 21 |j,m: Vs
= 0.0124 cm/s
Passaic Valley Freshwater sewage sludge with a particle size of 22 |j,m: Vs =
0.0057 cm/s
(3) Filtration and Separation.com.
This web site has an interactive program that allows the user to enter in a
sediment particle size and density and then use the properties of water (density
and viscosity) to compute the particle settling rate. This program computes the
settling velocity using Stokes' Law, the Heywood Tables (valid for Reynolds
Numbers up to 100,000) and Archimedes correlation, which bases the estimated
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
settling velocity on the Reynolds number computed for the specific information in
the program. All results are provided as output with a recommendation of which
value is most applicable.
(4) Measurement Suspended Sediment Characteristics in an Embanked Flood
Plain Environment of the River Rhine. (Thonon and Van Der Perk, 2002)
This paper describes the study conducted on the River Rhine located in The
Netherlands. The study was conducted to help quantify the amount of sediment-
transported pollution that is occurring in the flood plains of the River Rhine. This
data is being used to calibrate flood plain sedimentation models and to assist in
the assessment of the fate and transport of sediment-associated pollutants in
riverine environments. Field studies were completed by deploying a LISST-ST
Type C portable particle counter manufactured by Sequoia Scientific at the main
distributary of the Rhine River.
Generally, this instrument measures particle sizes and settling velocities for
particles ranging from 2.5 to 500 um using laser diffraction principles. At the
beginning of each study, the settling tube is opened for four seconds and allowed
to fill with river water and suspended matter. It is then closed and the test is run
for a duration of 12 hours. The suspended matter size is then measured in the tube
71 times over the 12-hour period. Finally, the settling velocity is computed from
the decrease of the volume of concentration of the different particle fractions over
time. Results of this study were as follows:
For a particle of size 10 microns: Vs = 0.001 cm/s
For a particle of size 50 microns: Vs = 0.005 cm/s
For a particle of size 100 microns: Vs = 0.01 cm/s
For a particle of size 400 microns: Vs = 0.01 to 0.001 cm/s
(5) Model for Turbidity Plume Induced by Bucket Dredge (Kuo and Hayes, 1991)
This study employed a model to evaluate the plume created in a river by a
mechanically operated dredge. This study was completed for three river systems.
Sediment characteristics were provided for each of these river systems and the
settling velocity was computed using Stokes' Law.
St. John's River: Particle size of 39.6 microns (98% of sediment finer than 62
microns) and sediment density of 2.40 g/cc; Vs = 0.12 cm/s
Black River Harbor: Particle size of 36.3 microns and sediment density of
2.39 g/cc; Vs = 0.10 cm/s
Thames River: Particle size of 150 microns and sediment density of 2.50 g/cc;
Vs = 1.84 cm/s
Thames River: Particle size of 160 microns and sediment density of 2.50 g/cc;
Vs = 2.10 cm/s
(6) Dredge Induced Turbidity Plume Model. (Kuo et a/, 1985)
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
This paper examined a model to help describe the turbidity plume resulting from
dredging in a ship channel with a hydraulic dredge. The model was developed to
predict the sediment concentration within the plume and the resulting
sedimentation alongside the dredged channel. Results of the model are compared
with actual field measurements. It was concluded that the model calibrated
parameters agreed with field observations and measurements. The settling
velocity was computed for model input using the following equation:
Vs = w = l/18v * ((psp / pw) - 1)) * g * aA2
Where:
v = viscosity of water = 1.08 X 10"5 ft/s = 0.01 cc/s
psp = density of particle (g/cc)
pw= density of water = 1 g/cc
g = acceleration due to gravity = 32.2ft/s = 980 cm/s2
a = particle size (cm)
In the referenced paper, a = 20 microns = 20 X 10"4 cm and psp =
2.65 g/cc and Vs = 0.0359 cm/s
Applying this equation to the Hudson River Sediment Characteristics:
Silt assuming a particle size of 20 microns and range of particle densities from
2.2 -2.6 g/cc: Vs = 0.026 -0.035 cm/s
Fine Sand assuming a particle size of 100 microns and range of particle
densities from 2.2 -2.6 g/cc: Vs = 0.653 -0. 871 cm/s
Medium-Coarse sand assuming a particle size of 400 microns and a range of
particle densities from 2.2 - 2.6 g/cc: Vs = 4.0 - 8.5 cm/s
(7) New Bedford Harbor Water Quality Monitoring Pre-Design Field Test Dredge
Technology Evaluation Report, Appendix K. (US ACE, 2001)
An estimate of Vs using Stokes' Law and particle size for silts and clay was
provided as follows:
Silt with particle size of 0.02 mm; Vs = 3.21 X 10"6 cm/s
Clay with particle size of 0.002 mm; Vs = 3.21 X 10"8 cm/s
(8) 1999. PCBs in the Upper Hudson River Volume 2. A Model of PCB Fate,
Transport, and Bioaccumulation. (QEA, 1999)
For application of a model to predict PCB concentrations in the Hudson River, a
fate and transport model was applied. One of the parameters required for input
into this model was the specific Hudson River sediment characteristics including
the particle size, particle density, and the particle settling velocity. Settling
velocities for cohesive and non-cohesive sediments were estimated using different
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
methods. The settling velocity for cohesive sediment was computed utilizing the
following formula:
Vs = 3.3 * (CiG)A0.12 (EQ
Where:
Ci = particle concentration (mg/1)
G = water column bottom shear stress = Cf * qA2
(dynes/cm2)
This formula was developed for the fine particles when flocculation occurs among
particles during the settling procedure. Therefore, settling velocities may be
applied to silt particles since coarse/sand particles will not aggregate. Measured
settling velocities were plotted as a function of CiG and have a range from 4 to 9
m/day while the value of CiG ranges from 10 to 2000 (mg/L*dynes/cm2).
However, the study did not show a trend with particle density (within the silt
range used). In this study the non-cohesive settling velocity was estimated based
on particles size and particle density using Stokes' Law.
4.1.1.2 Selection of Settling Velocity
A summary of the settling velocities from the studies above is provided in Table 2. For
most of studies Stokes' Law is the theoretical basis for estimating the settling velocity of
sand particles. This approach is appropriate for discrete particles that do not aggregate.
For the fine sand sediments of the Hudson River, the settling velocity would be 0.6 - 0.8
cm/s assuming that the range of particle density is 2.2 to 2.6 g/cc and the particle size of
fine sand is 100 microns. Using the same range of particle density, the settling velocity of
medium-coarse sand in Hudson River sediments is 4.0 to 8.5 cm/s assuming that the
typical particle size is 400 microns. For the CSTR-Chem and TSS-Chem models 6 cm/s
was used as a conservative estimate of the typical settling velocity for the sand fraction of
Hudson River sediments.
Stokes' Law only applies to discrete particles settling and does not account for the
flocculation during settling. Flocculation increases the rate at which silts settle from the
water column, but the rate of flocculation depends on site specific conditions and
sediment properties. The silt settling velocities presented in QEA's report (1999) for
Hudson River sediments were used in the near-field models since these values were
directly applicable to Hudson River sediments and included the effects of flocculation.
Even though the settling velocity was presented as a function of Ci*G (particle
concentration * shear stress), the settling velocity varied in a very narrow range (4-9
m/day) while the value of Ci*G varied in 3 orders of magnitude (from single digit
"3
number to a couple thousands). Therefore, 7 m/day, equivalent to 8.1 x 10" cm/sec, was
chosen as the typical settling velocity for Hudson River silt/clay. The range of 4 m/day
and 9 m/day were applied to the sensitivity analyses of the models. It should be noted that
"3
8.1x 10" cm/sec is one order of magnitude less than the velocity estimated by Stokes'
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Law (0.026 - 0.035 cm/s) when assuming that the particle size is 20 microns and the
density is 2.2-2.6 g/cc.
Concern has been raised that a probability factor of settling should be applied to account
for the effects of near-bed turbulence on particle deposition. However, sediment particles
in the near-bed zone have effectively been removed from the water column. They are not
available for downstream transport within the water column and no longer contribute
significantly to water column exposure. Thus, the water quality models applied here do
not attempt to deal with complex near-bottom sediment erosion and deposition. It would
be reasonable to develop and apply models capable of considering a wider range of
processes, e.g. near-bed erosion and deposition, during the design phase when more
detailed analyses of the fate and transport of sediments and associated constituents are
appropriate.
4.1.2 Lateral Dispersion Coefficient
The lateral dispersion coefficient impacts the width of the solids plume and therefore the
concentration within the plume, as the solids are transported downstream. In order to use
TSS-Chem to model the movement of the solids plume downstream, a lateral dispersion
coefficient must be specified. Since the coefficient is dependent on the velocity of the
river water, more than one lateral dispersion coefficient value was required.
A time-of-travel study conducted by USGS in Upper Hudson River (USGS, 1969)
examined dye concentrations vs. time at both center and side channel stations located
near Schuylerville. The peak concentration at the center channel station occurred 0.5 to 1
hour earlier than the peak concentration at the side channel station, demonstrating the
lateral dispersion of the dye. Theoretically, the lateral dispersion coefficient can be
estimated based on the conservation of dye mass, but the locations of the center and side
channel stations and the raw data for the dye concentrations are not provided in the
report. Due to the limitation of available data and the difficulty of finding data from an
old report, the numerical solution was not pursued based on this report. Due to the
limitation of available data and the complexity of natural river systems, the results
presented below are considered to provide an order of magnitude estimate of the lateral
dispersion coefficient.
Fischer (1979) provides the practical rule that the lateral dispersion in a bounded channel
can be approximated as:
ฃt = 0.6du*
(EQ
Where:
d
u
2
lateral dispersion coefficient (m /s)
average depth of flow (m)
shear velocity (m/s), sfgdS
gravitational acceleration, 9.81 m/s
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
S = slope of the channel (unit less)
Since surface water elevation is the energy grade indicator of the river, surface water
elevation slope can also be used to calculate the shear velocity. USGS monitors the daily
water elevation at gauged stations throughout the year. Gauge 119 is located near Lock 7
and gauge 118 is located near TI Pool. The distance between these two gauges is about 6
miles. The surface water elevation slope between these two gauges represents the energy
slope within the TI Pool. The average water elevation difference was calculated on a
monthly basis for several years of data. Negative water elevation differences were
observed and treated as 0 in the averaging, which does not significantly change the
monthly average values. As summarized in Table 3, the maximum monthly average
elevation difference occurred in March due to high flows during spring run-off For the
dredging season (May through November), the monthly elevation difference is relatively
consistent. Using these months a dredging-period slope of 8 xlO"6 was obtained.
The hydrodynamic model RMA2 (described below in Section 4.2) was used to obtain
applicable depths and linear velocities for various river flowrates (2000-8000 cfs) and
locations (RM 190 and 193) along the Upper Hudson River. Equation 2 was used with
the applicable depths, velocities and average dredge-season slope to calculate the lateral
dispersion coefficients under different conditions. The results are presented in Table 4.
Dispersion coefficients calculated for the eastern segment at RM 190 were used as the
typical condition. The dispersion coefficients for the other conditions were investigated in
the sensitivity analysis.
4.2 RMA2
RMA2 is a hydrodynamic model created by the USACE that can be used to simulate
ambient water conditions such as velocity magnitude and direction at potential dredging
sites. Initially, LTI used the RMA2 model to simulate the flow patterns in the Thompson
Island Pool to develop the hydrodynamic portion of the HUDTOX model. These results
were presented in the Revised Baseline Model Report (USEPA, 2000b). The focus of the
LTI study was to derive the spatial distribution of the shear stresses, which in turn was
used to determine the depth of scouring and aggregate amount of re-suspension. The
amount of re-suspension was then partitioned to PCB loads and incorporated into a long-
term transport model {i.e., HUDTOX).
The LTI RMA2 model considered a wide range of flows, from an average flow of about
4,000 cfs to the 100-year flow of about 47,000 cfs. While the low to moderate flows were
confined within the Hudson River banks, the higher flows required the inclusion of the
Hudson River flood plains into the model. Therefore, the computational domain had to be
extended to include the flood plains even under low flow conditions.
Since the dredging activities are more likely to take place during normal summer flow
conditions, it is logical to reconfigure the computational model and allocate all available
computing resources, (i.e., memory, speed, and total number of elements) to normal flow
Hudson River PCB s Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
conditions only (excluding the flood plains). As a result, the narrowed flow range allows
the model to incorporate a refined resolution in the river and near the dredging sites. The
refined grid can also be used to incorporate more detailed bathymetric variations and to
reproduce higher accuracy flow patterns.
4.2.1 Methodology
The new computation grid for RMA2 reflected the following considerations:
(1) It essentially confined to the deep channel of the river and focused on the wet
boundary at low flow conditions;
(2) It uses highly refined spatial resolution (a typical resolution is about 15 feet in the
transverse direction of the flow);
(3) It represents the river bathymetry more realistically by incorporating the 1990
bathymetric survey data on the refined grids. Additionally, the new grid has
adopted quadratic elements to reduce numerical dispersion and enhance numerical
convergence at internal wet-dry boundaries.
The new configuration of the RMA2 model to depict dredging conditions was validated
by comparison to the LTI RMA2 model. To maintain continuity and consistency between
the two studies for comparison, the refined model and the previous model were both set
up to simulate the flow patterns and surface profiles with the same boundary conditions
and physical parameters. Comparable results from both models would indicate that the
refined model has inherited the characteristics of the previous model, and more
importantly the credentials that the previous model has built from a thorough calibration
process.
The is cross-model validation process was conducted for two flow conditions:
(1) The 100-year flow condition which was presented in the Revised Baseline
Modeling Report (USEPA, 2000b);
(2) A 4,000 cfs flow condition which approximates the average flow conditions.
For the previous LTI RMA2 model, the geometry file and boundary condition file were
obtained from LTI. The geometry file included both mesh and bathymetry information,
and the boundary condition files included physical and model control parameters. For the
refined model the boundary conditions and physical parameters were kept the same as the
previous model.
The refined model and the LTI RMA2 model were compared for flow patterns for 100-
year flow condition. The upstream flow is 47,330 cfs, and the downstream elevation is at
126 feet. Two Manning's n values were used in the previous model, 0.20 in the channel
and 0.60 in the flood plain. The refined model is mostly confined to the river channel,
therefore the Manning's n was kept at 0.20. Turbulent dispersion coefficient was 100 lb-
sec/ft and homogenous for both models. The previous and the refine models show
Hudson River PCBs Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
similar flow patterns and velocity magnitudes. The notable differences can be attributed
to the omission of flood plain in the refined model. Due to the relatively higher flow
depth, the more accurate representation of the bathymetry in the refined model does not
seem to contribute significantly to changes in flow pattern or the velocity magnitude.
In addition, the two models were compared for the flow patterns for 4,000 cfs. At this
flow rate, the downstream water surface elevation is at 119.2 feet. Because the flows are
confined mainly to the river channel, the omission of the flood plain is immaterial.
However, at this lower elevation, the effects of the more detailed representation of
bathymetry on the flow depth and velocities with the refined model became noticeable.
4.2.2 Results of RMA2
Once the model was validated with the previous model, it was used to simulate the flow
patterns at the normal summer flow range. Three representative flows were selected
based on the actual flow records - 2,000, 5,000 and 8,000 cfs. In all of these runs the
Manning's n value was kept at 0.2 and the turbulent dispersion coefficients was at 100 lb-
sec/ft2. The downstream elevations were at 118.6, 119.2 and the 120.6 feet respectively.
It can be seen that the magnitude of the velocity increases with flow and results an
increased water surface elevation upstream.
In addition to providing more detailed velocity magnitude and direction at potential
dredging sites, the RMA2 simulation results would provide a more accurate shear stress
representation and scouring analysis. Potentially the simulated flow field can be used
directly in contaminant and sediment transport models such as RMA4 and SED2D. As
dredging operations progress, the bathymetry in the model can be easily updated to
reflect the post-dredging bathymetry. The flow patterns can then be revised with the
updated geometry. The impact of the post-dredging bathymetry can become particularly
important when the dredged depth is comparable to the water depth and when the
dredging area is relatively large.
4.3 CSTR-Chem
4.3.1 Methodology
The objective of this analysis is to estimate the net contribution of solids, and dissolved
and suspended phase PCB to the water column in the immediate vicinity of the dredging
operations. This analysis describes the approximation of water quality impacts in the
immediate vicinity of a dredging operation using a mathematical model based upon the
CSTR concept. It assumes that the waters are completely mixed by ambient and induced
currents.
Hudson River PCB s Superfund Site
Engineering Performance Standards
14
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Ideal reactor configurations are used to simplify mathematical modeling of constituent
concentrations in surface waters. Two primary ideal reactor configurations are used -
continuous flow stirred tank reactors (CSTRs) and plug-flow reactors (PFRs). CSTRs
assume that a constant concentration and flow influent is instantaneously mixed as it
enters a confined, well-mixed tank. Physical and chemical reactions occur while the
water is within the ideal tank and the tank effluent is at the same flow as the influent and
at the uniform concentration within the tank. PFRs assume that constituent laden waters
travel downstream in a perfectly uniform pattern without lateral and vertical mixing;
physical and chemical reactions occur during downstream movement.
Real surface water systems do not have mixed flow conditions; i.e., the waters are never
completely mixed or travel downstream without lateral or vertical mixing. However,
representing sections of water bodies as one of these ideal reactors can provide useful
approximate results, often within errors associated with data available to support the
models. The CSTR concept is most appropriate to the analysis of dredging operations
because turbulence in the area of the dredge, coupled with ambient flows, may be
assumed to produce mixed conditions.
Water Column Mass Balance for Suspended Sediments1
Suspended sediment concentrations in the well-mixed water volume that can be
approximated as a CSTR can be approximated by:
Vnf^ = qmm-qm- vsAhm + MR (EQ
where:
3
Wnj = volume of the near-field area (m )
m = Suspended solids concentration in the near-field volume
approximated as a CSTR (mg/L)
t = elapsed time (sec)
3
q = flow through the near-field volume (m /sec)
min = Suspended solids concentration of flow entering the near-field
volume (mg/L)
?s = settling velocity of suspended particles in near-field volume
(m/sec)
Ah = cross sectional area perpendicular to the height (m ), and
Mr = rate of mass resuspension into the near-field area due to dredging
(g/sec).
Steady-state Conditions
1 This analysis consists of a mass balance for suspended sediments in the water column only.
Hudson River PCBs Superfund Site
Engineering Performance Standards
15
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
If q, Mr , and vs are constant for a relatively long period of time, steady-state conditions
will be reached, i.e., dm/dt = 0. Steady state suspended solids concentration can then be
estimated as:
and
m = !MlMjl (EQ4)
(EQ5)
where:
3
V(.;/ = volume of the near-field area (m )
T = hydraulic retention time within CSTR (sec)
H = water depth (m).
If the near-field area is assumed to be a square box over a water depth H, than the volume
can be expressed as:
Vnf=w2H
where:
w = width of the near-field area (m)
Hydraulic retention time is the volume divided by the flow rate
= 7T (EQ s>
Sdnf
It should be noted that the hydraulic retention time is only a function of the width and
linear velocity of the near-field. This is illustrated in the following equation.
w2H w
e"f=-n- = - (EQ?)
uHw u
where:
u = linear velocity of water (m/s)
The solids concentration inside the CSTR before settling can be expressed as:
Wadded =m,n+ (EQ 8)
and the solids concentration lost to settling is:
Hudson River PCBs Superfund Site
Engineering Performance Standards
16
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
msettled madded m0ut
(EQ 9)
Note that the concentration exiting the CSTR (mout) is equivalent to that in the CSTR (m).
In cases where the sediment type (i.e., silt, sand) is of importance, the suspended solids
mass balance can be applied to each sediment component, using the respective settling
velocities.
Toxic Constituents2
The transport, fate and impact of toxicants are intimately connected with how they
partition or associate with solid matter in or below the water body. This implies that the
two forms of the toxicant - the dissolved and suspended forms must be distinguished in
any analysis. This distinction has an impact on transport and fate because certain
mechanisms differently impact the two forms. In the analysis that follows, volatilization
and transformation of the contaminant are assumed to be negligible.
Recent studies have demonstrated that desorption of hydrophobic chemicals from
sediments can be quite slow and that chemical equilibrium may not be a good
approximation in many real situations. To be consistent with the literature on PCB
desorption, transient partitioning is assumed in the model, and the rate of PCB desorption
from solids is proportional to the difference between the PCB concentration of the
suspended sediments and the concentration that would be in equilibrium with the existing
soluble concentration. Therefore, a complete formulation of a mass balance under the
transient partitioning first requires the concentrations of PCB under equilibrium
conditions.
Contaminant Equilibrium Partitioning
It is assumed that equilibrium conditions exist in the near-field CSTR. A mass balance
for the concentration of total PCB under this condition can be expressed as:
die
^n/ = tfCTotal,in ~ tfCTotal ~ Vs^h^s,eqCTotal +^RCsed (EQ 10)
where:
3
Wnj = volume of the near-field area (m )
CTotal = total concentration of the contaminant (ng/L), which is the sum of the
dissolved
and suspended concentrations in the near-field volume
Cd,eq = equilibrium contaminant concentration in dissolved form in the near-
field
volume approximated as a CSTR (ng/L)
2
Porewater contributions are assumed to be negligible and are not considered in this analysis.
Hudson River PCB s Superfund Site
Engineering Performance Standards
17
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
cs,eq = equilibrium contaminant concentration in suspended form in the near-
field
volume approximated as a CSTR (ng/L)
t = elapsed time (sec)
3
q = flow through the near-field volume (m /sec)
^Total,in = tฐtal concentration of the contaminant in the flow entering the near-
field volume
(ng/L)
?s = settling velocity of suspended particles in near-field volume (m/sec)
Ah = cross sectional area perpendicular to the height (m )
Mr = rate of mass resuspension into the near-field area due to dredging
(g/sec)
csed = contaminant concentration on bottom sediments (mg/kg).
Fs,eq = fraction of contaminant mass in suspended form in equilibrium
(unitless)
This fraction of contaminant in suspended form under equilibrium partitioning can be
estimated:
F^ = ifjmXl0L-6 (EQ11)
\ + Kdxmx\0
where:
Kc/ = two-phase contaminant partition coefficient (L/kg)
m = suspended solids concentration in the near-field
Under steady state conditions:
*}cTotal, in +^RCsed
^ m A * sea (EQ 12)
q+vsAkFP,eq
The equilibrium concentrations in the dissolved phase and suspended phase along with
the concentration on the particles can then be computed as:
cd,eq = (EQ13)
\ + Kdxmx\0
Cp,eq = Cd,eq xKdx 1(T6 and Cseq = Cpeq xm (EQ 14)
where:
CP:Bq = contaminant equilibrium concentration on the particles (mg/kg)
If the background concentration is assumed to be in equilibrium and the suspended solids
and fraction of dissolved PCB are known then Kd may be computed as:
Hudson River PCBs Superfund Site
Engineering Performance Standards
18
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Kd=J7 1 (EQ 15)
Fd,in X X 10
where:
Fd in = fraction of contaminant mass in dissolved form in the background
(unitless).
For lipophilic contaminants such as PCBs, three-phase partitioning (adding partitioning
to dissolved organic carbon) may be important in determining the phase distribution of
contaminants. The equations presented above, however, remain valid if Cd,eq is interpreted
as the "apparent" dissolved concentration or the non-filterable portion that may include
both truly dissolved and DOC-sorbed PCBs.
Transient Contaminant Partitioning
Assuming that desorption from the suspended particles to the waster column occurs
during the residence time in the CSTR, mass balance expressions for both the dissolved
and suspended phases are:
Vnf = 1Cd,tn - 1Cd + kVnf{Cd,eq - Cd) (EQ 16)
Vnf ^ " hVnf(Cs,eq ~ C)~VACS +^RCsed (EQ 1 7)
where:
Cd
Cs
Cd,eq
Cs,eq
^d, in
contaminant concentration in dissolved form in the near-field volume
approximated as a CSTR (ng/L)
contaminant concentration in suspended form in the near-field volume
approximated as a CSTR (ng/L)
equilibrium contaminant concentration in dissolved form in the near-
field
volume approximated as a CSTR (ng/L). Obtained from equation 13.
equilibrium contaminant concentration in suspended form in the near-
field
volume approximated as a CSTR (ng/L). Obtained from equation 14
dissolved contaminant concentration of flow entering the near-field
volume
(ng/L)
Hudson River PCBs Superfund Site
Engineering Performance Standards
19
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
csin = suspended contaminant concentration of flow entering the near-field
volume
(ng/L)
k = rate of desorption of contaminant concentration from suspended form,
also
defined as the rate at which equilibrium is reached (1/sec).
If steady-state conditions exist in the near-field area, the dissolved and suspended
concentrations along with the concentration on the particles, under transient partitioning
can be estimated from equations 16 and 17 as follows:
-------
Mixing is less obvious with a hydraulic dredge, but should be a reasonable
assumption in relatively shallow water.
The diameter of the cylindrical area approximated as a CSTR should reflect the
extent to which well-mixed conditions exist. For the purposes of this analysis, a
CSTR width of 10 meters is used. Buckets expected to be used in the Hudson
River project are generally 2 to 3 m in diameter closed and somewhat more open.
It is reasonable to assume that velocities induced by bucket movement could
extend across most of a 10 m width used in this analysis.
The FS assumed that a 4-cy environmental bucket would be used to dredge the
Hudson River with a two-minute cycle time. Appendix E-6 estimated a sediment
resuspension rate of about 1 kg/sec.
This application also considered two sediment types - silt and coarse materials.
Appendix E of the FS contains information cohesive and non-cohesive fraction of
sediments, as well as the silt and coarse fraction. Tables 1 and 5 summarize this
information for the three sections of the river considered.
Newly suspended bed sediments are the primary source of new toxic constituents
to the water column during a dredging operation. Based upon the research of
Warren, Bopp, and Simpson (1997) equilibrium is reached at a rate of 0.20/hr or
less; a conservative estimate of 0.2/hr is used as the rate of PCB desorption in this
analysis. The selection of the desorption rate is discussed in more detail in
Attachment C.
The partitioning coefficients used for each river section were obtained by
assuming that background concentrations of dissolved and suspended PCB are in
equilibrium.
It is assumed that the inflow to the near-field consists only of silt particles. The
appropriate settling velocities for silt and sand particle were obtained from review
of literature on particle settling in similar systems. Sediments resuspended due to
dredging operation are assumed to have uniform particulate PCB content,
regardless of type.
Transient partitioning is assumed for desorption from resuspended sediments. All
other partitioning behavior is assumed to be adequately described by equilibrium
assumptions.
Table 6 presents the model inputs for the three sections along with model simulation
results. The results suggest that under transient partitioning conditions, which are
expected within the CSTR, over 98% of the resuspended PCBs are simulated to remain in
particle form.
Hudson River PCBs Superfund Site
Engineering Performance Standards
21
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
4.3.3 Sensitivity Analysis
The CSTR-Chem model was used to simulate the net suspended solids, net fraction
dissolved PCB and net total PCB flux in the near-field as a result of dredging operations.
Because models typically contain parameters, the simulation results can be highly
sensitive to small changes in the parameter values. Therefore, a sensitivity analysis was
performed to quantify the sensitivity of model outputs of greatest interest in the CSTR-
Chem model to uncertainty and variability in input parameters. This analysis is important
for checking the quality of the CSTR-Chem model, as well as the robustness and
reliability of CSTR-Chem modeling analysis.
The CSTR-Chem model parameters on which the sensitivity analysis was performed
include:
River Volumetric flow (thereby linear flow and depth),
Resuspension rate,
Silt fraction in the sediment,
PCB sediment concentration,
Near-field width,
Background conditions (suspended solids and PCB concentrations, and dissolved
PCB fraction),
Partition coefficient
Desorption rate
Silt and Coarse Settling Velocity
Four model output values were selected to assess the sensitivity of the above parameters.
These outputs of concern were:
The net fraction of dissolved PCBs from dredging, which is estimated as fraction
of the net total PCB that is dissolved. The net total PCB is the output total PCB
less the background total PCB.
Net fraction of silts, which is the fraction of net suspended solids (output
suspended solids less background suspended solids) that is silt.
Net total PCB flux exiting the near-field.
Net suspended solids flux exiting the near-field.
A deterministic approach, which assesses sensitivity of a model output to the range of
variation of a parameter, was used in this sensitivity analysis. This method involves
calculating the output for a few values of an input parameter. This analysis evaluates the
effect on model outputs exerted by individually varying only one of the model input
parameters across its entire range of plausible values, while holding all other inputs at
their nominal or base case values.
Hudson River PCBs Superfund Site
Engineering Performance Standards
22
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Results and Discussion
The results of the sensitivity analysis were presented using two techniques as follows:
A dimensionless sensitivity coefficient Sparameter,output for each parameter was
calculated as follows:
v AOutput / Output
Paramater, output ป ป-ป , n
Aparameter i Parameter
where,
Parameter = parameter value for the base case, which is the model default
value.
? Parameter = the absolute change in input parameter value.
Output = model simulated output for the base case input value.
? Output = the absolute change in model simulated output
The average of the Sparameter,output values was calculated for each output of concern and
the results are presented in Table 7. The higher the sensitivity coefficient for a
particular input parameter, the more sensitive the model output is to perturbation of
that parameter.
A graphical method, which gave a visual indication of how each output is affected
by variations in inputs, was also used to represent the results (Figures 2 through
15). These graphical representations depict the linearity or non-linearity of the
relationships between parameter values and model-simulated outputs.
The results of the parameter sensitivity analysis can be summarized as follows:
There were no significant differences between the River Sectons in the sensitivity
to most of the parameters (e.g. River wide flow and sediment PCB concentration).
Therefore, the sensitivity analysis is mainly focused on River Section 1.
The net fraction dissolved is most sensitive to changes in the width of the near-
field CSTR. The CSTR width directly affects the contaminant residence in the
near-field, and the residence time is important to the kinetics of particulate PCB
desorption. The net fraction dissolved is relatively less sensitive to changes in
width at lower CSTR widths (Figure 6). However the width becomes highly
sensitive at higher values, as indicated by the slope of the graph between the net
fraction dissolved and the CSTR width.
The net fraction of dissolved PCB is also sensitive to changes in the PCB
partitioning coefficient and the rate of PCB desorption. The partitioning
coefficient controls the equilibrium concentrations of dissolved and suspended
phases, while the rate of desorption control the PCB desorption kinetics. Both
parameters had no effect on the other outputs simulated.
Hudson River PCB s Superfund Site
Engineering Performance Standards
23
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
The net total PCB concentration is only sensitive to changes in the concentration
of PCB in sediment, and rate of resuspension. Note that the net fraction dissolved
is sensitive to changes in resuspension rates and sediment PCB concentrations
under conditions of very low resuspension rates (Figure 7) and very low sediment
PCB concentrations (Figure 9), respectively.
The settling velocities of suspended particles were not sensitive parameters
especially for silt particles. However, all the outputs of concern are moderately
sensitive to the specification of the sediment silt fraction.
The sensitivity analysis suggests that the CSTR width, the PCB partitioning coefficient
and the PCB desorption rate are the three most important parameters controlling the
release of suspended PCB to the dissolve phase. The width of the CSTR depends on the
dimensions of the dredge bucket, and a conservative input of 10 m is used as the base
value in the model. The Hudson river FS presented detailed values of the partitioning
coefficient of PCB for several congeners suggesting that values of this parameter are well
constrained. Therefore, the rate of the PCB desorption is the only parameter that can
significantly affect the reliability of the CSTR-Chem model simulations.
Recent studies have demonstrated that desorption of hydrophobic chemicals from
sediments can be quite slow and that chemical equilibrium may not be a good
approximation in many real situations. In the CSTR-Chem model the rate of PCB
desorption from solids is proportional to the difference between the PCB concentration of
the suspended sediments and the concentration that would be in equilibrium with the
existing soluble concentration. Several studies (Carroll el al, 1994, Borglin el al, 1996;
Cornelissen et al., 1997; ten Hulscher et al., 1999, 2002; and Ghosh et al., 2000) have
characterized the kinetics of PCB desorption as a two stage process: 1) the desorption of
a fast desorbing labile fraction and 2) a slow desorption of a non-labile fraction. A
representative value for desorption rate of the fast fraction of PCB from these studies is
0.2 hr"1. The rate of desorption of the slow fraction is over an order of magnitude lower
that that given for the fast fraction. In order to be conservative, the CSTR-Chem model
simulation for the base case were performed using a constant desorption rate of 0.2 hr"1.
Conclusions
The sensitivity analysis indicates that model simulations using conservative values of
PCB desorption and CSTR width should not affect the reliability of model conclusions.
Given the small residence time within the CSTR, most of the silt particles are expected to
exit the CSTR. However, no significant release dissolved phase release of PCBs is
expected.
Hudson River PCBs Superfund Site
Engineering Performance Standards
24
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
4.4 TSS-Chem
4.4.1 Methodology
TSS-Chem is intended to provide a model of the downstream transport of solids and
PCBs through the near-field (approximately 1 mile). TSS-Chem contains both a solids
component and a PCB component. The solids considered are from the silt and coarse
resuspended sediments and PCB concentrations modeled are both suspended and
dissolved.
TSS-Chem uses the solids source strength of dredging activities to model downstream
transport of suspended solids. The source strength differs from the resuspension rate
since resuspended sediments settle around the dredgehead, and only a fraction of the
suspended solids will be available for downstream transport. As was shown in the CSTR
model, the solids that settle within this area are primarily coarse material. Due to the high
settling velocity of coarse solids, they do not supply a significant amount of solids or
PCB transport. In order to show that the coarse material will not supply a significant
amount of solids or PCBs, the solids downstream transport model in Appendix E and
Resuspension White Paper of the RS, was modified in TSS-Chem to include the
contribution of coarse solids as well as fine solids.
During the downstream transport PCBs adsorbed to the solids will partition into the
water-column. In this model two-phase partitioning from the suspended phase into the
dissolved phase is estimated. As shown in the CSTR the initial dissolved phase available
for downstream transport is not significant and the initial PCB concentration on the solids
available for transport downstream (known as the source strength) is not significantly
different from the sediment concentration.
Suspended Solids - Kuo and Hayes Model (General Equation)
The current suspended solids plume model utilizes the Kuo and Hayes (1991) Gaussian
equation (Equation 24) for modeling the downstream transport of resuspended sediments
with clamshell bucket dredges. This equation assumes no lateral or downstream barriers,
uniform and unidirectional flow, and constant water depth.
x = distance downstream of source (m)
y = distance across stream from the source (m)
g = sediment loss rate (kg/s)
u = ambient linear velocity in the x- direction (m/s)
h = depth (m)
(EQ
Where:
Hudson River PCBs Superfund Site
Engineering Performance Standards
25
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
ky = lateral (y-direction) dispersion coefficient
w = settling velocity
The model presented in Equation 24 is a continuous mathematical function/model that
models transport in the x-direction by advection only. Dispersion in the x direction is not
considered a significant factor. It computes a concentration for a given x, y location. That
value is valid at that x,y point only. However, it is not unreasonable to assume that
concentration represents an approximate average of the concentration between some x-
distance before the point and a similar x-distance beyond the point. Simple averaging in
the lateral direction yields a less correct answer. In fact, over the centerline, it can yield
an extremely incorrect answer. Equation 26 computes concentrations out to infinity, as
discussed below, a cut-off concentration is necessary to limit the width of the plume to
within the river. However, with a cut-off concentration the mass outside the designated
plume width will not be accounted for and the model will not conserve mass. Therefore
to conserve mass the integration of this function should be used obtain an average
concentration of a transect (x=constant).
Suspended solids - Kuo and Hayes Model (Integrated Equation)
In order to conserve mass the average concentration along a transect is calculated using
the integrated version of Equation 24. The following known integral (CRC Handbook of
Chemistry and Physics) can be applied to Equation 24 to obtain the product of the
average concentration and width of the plume with total reflection of solids along the
shorelines (no mass lost past the shorelines).
<*Q
Applying Equation (25) to Equation (24) and multiplying by two for both sides of the
plume yields:
wx I wx
q y = 2 ^ q hu * _ & g hu (EO
avg pIume uh-yjAnkjX/u 2^ju/(4nkyx) uh
Where: ypiume = width of the plume (lateral extent of the plume)
(m)
Suspended solids - Kuo and Hayes Model - Determining ypiume (General Equation)
To determine the width of the plume Equation 24 can be modified. The width can then be
bound by a cut-off concentration or a percentage of the concentration at x=0. Equation 24
may be used to calculate the suspended concentrations for various locations along a river
transect (x=constant). If the width of the river is given than a y-increment can be chosen
to estimate the average concentration along the transect. The width is separated into
discreet boxes each with a width equal to the y-increment, except for the outer two boxes.
Hudson River PCBs Superfund Site
Engineering Performance Standards
26
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
For instance, if the source is located 2 meters from the shoreline and a y-increment of 1 is
chosen the boxes are:
y = 2 to 1.5 (represented by y=2, width=0.5),
y = 1.5 to 0.5 (represented by y=l, width=l),
y = 0.5 to -0.5 (represented by y=0, width=l),
y = -0.5 to-1.5 (represented by y=-l, width=l), etc.
Since the model will be used to calculate the solid concentrations for a source close to
one shoreline Equation 24 must be modified to include shoreline reflection. In this model
it was assumed that there is total reflection. Therefore the solids that would be 1 meter
outside the shoreline were added to the solids 1 meter within the shoreline. For instance
in the example above:
Outside Shoreline Inside
River y=2 River
=3.5 y=2.5
II
=1.5 y=0.5
>>>>>>>>>
<<<<<<<<<
>>>>>>>>>
<<<<<<<<<
>>>>>>>>>
<<<<<<<<<
>>>>>>>>>
<<<<<<<<<
>>>>>>>>>
<<<<<<<<<
>>>>>>>>>
>>>>
< < < <
:ฆ :ฆ :ฆ :ฆ
ฆ: ฆ: ฆ: ฆ:
:ฆ :ฆ :ฆ :ฆ
ฆ: ฆ: ฆ: ฆ:
:ฆ :ฆ :ฆ :ฆ
ฆ: ฆ: ฆ: ฆ:
:ฆ :ฆ :ฆ :ฆ
ฆ: ฆ: ฆ: ฆ:
:ฆ :ฆ :ฆ :ฆ
y = 2 to 1.5 would also include the solid concentration from y=2.5 to 2,
y = 1.5 to 0.5 would also include the solid concentration from y=3.5 to 2.5,
etc.
Equation 24 then becomes:
c(x,y) =
2 4 kvx hu
f s J +-
h^AnkyX I
_ "yout , m
ฃ 4 kyX hu
e L
a,
u
c(x,y) = -
g
uh^AnkyX I u
or
f
2
2
uy
Wont
4 kYx
+ e
4 kYx
L
L J
Where:
uh^AnkyX I u
(EQ 27)
yout = the lateral distance the reflected solids would have traveled
without reflection (m)
The yout can be expressed in terms of y as:
Hudson River PCBs Superfund Site
Engineering Performance Standards
27
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
yout = (y Skore ~y)x2+y
(EQ
Where:
y shore
the distance to the shoreline from the source (m)
When the cut-off to determine the width of the plume (ypiUme) is expressed as a percentage
of the solids concentration at x=0, ypiUme is calculated as the sum of the box widths that
contain solid concentrations above the cut-off or:
y piume =YjWidthy
([box,y=i)
(EQ
Where:
n and -n
Widthbox,y=I
= furthest y distance that has a concentration greater than the cutoff
= width of the box represented by solids concentration at y=i (m)
For this model the plume was confined to solid concentrations greater or equal to 1% of
the concentration at x = 0.
Suspended solids - Kuo and Hayes Model (Two Settling Velocities)
If the source is assumed to contain both silts and coarse grain materials Equations 24 and
26 need to be modified to include a second settling term. It the two sediment types are
assumed to have the same lateral dispersion coefficient than Equation 24 may be
modified to:
c(x,y) = -
g.
silt
uy + wsutx
4kYx hu
go
W + wcoarsex
4kYx hu
ih^AnkyX 11
ih^jiTtkyX /1
or
c(x,y) = -
g,
total
ซy2 f
4 kYX
ih-^AnkyX I
u
fsnfilhu J +(!-/ )e M
and
fsilt
ง silt ^
gl
total
8,
total
(EQ
Where:
fsiit = fraction of silt in released sediment (unitless)
gtotai = total sediment loss rate (kg/s)
To account for both reflection from one shoreline and two settling velocities Equation 24
becomes:
Hudson River PCBs Superfund Site
Engineering Performance Standards
28
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
c(x,y) = -
g
silt
a,
u
'h^4nkyx I
ง coars
uy + wsmx
4kYx hu
uy wcoarse x
4kYx hu
g
silt
a,
u
'h^4nkyx I
ง coc
Wont + wsiltx
4kYx hu
Uypul wcoarsex
4kvx hu
ihyj 47rk yx /1
ih-^AnkyX 11
or
c(x,y) = -
g,
total
ih^jAnkyX 11
f r wsjit
^ ^ hu
silt
V
fsute 1 hu J +(1 ~fsut)e
"coarse"
hu
2
2
uy
Wont
4 kYx
+ e
4 kYx
L
L J
(EQ
The integral already accounts for total reflection therefore Equation 26 only needs to be
modified to account for two settling velocities. Equation 26 is modified as:
_ gsilt hu
uh
g
c v , =^^-e "" +" "Jฐ e
avgJ plume
coarse hu ^total
uh
uh
wsUtx
hu
fsme 1 hu J +(1 -fslit)e
hu
(EQ
Two-Phase Partition Model for PCBs
The two-phase partition model is used to estimate PCB concentrations in the water
column based on the sediment releases from dredging, the PCB concentrations of the
suspended sediments and the background conditions. Both the dissolved and suspended
(particulate) PCB concentrations are modeled using equilibrium partitioning. As shown
from the CSTR model runs, the initial fraction of the dissolved PCBs is not significant
and may be assumed to be zero. For the initial conditions of the two-phase partitioning
model, partitioning between dissolved and suspended has not reached equilibrium and
PCBs will continue to be transferred from the particles to the dissolved phase as they are
carried downstream. To estimate the progression towards equilibrium the two-phase
partitioning model factors in the residence time of the sediment in the water column (time
available to reach equilibrium). A conceptual depiction of the model is shown below.
Hudson River PCBs Superfund Site
Engineering Performance Standards
29
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
CSp
Where:
TSS, = Concentrations of TSS (mg/1)
CSi = PCB concentration on the suspended particles (mg/kg)
CDj = Dissolved PCB concentration in the water column (ng/1)
CTi = Total PCB concentration in the water column (ng/1)
Qi = Volumetric flowrate of box i (m3/s)
x = Distance traveled by the water and solids within each box (m)
y, = width of the plume (m)
in, out and BKG apply to the entering, exiting and background conditions respectively
The path of PCBs being transported downstream of the dredge head is divided into
segments. Each segment is addressed as a box. The width of the box equals to the width
of the suspended solids plume at the location of the box (its distance downstream of the
dredge head). It is assumed that the width of the plume does not change within a box and
therefore the volume and flowrate of the box remains constant. The incremental distances
downstream (x-increments) used in the model determine the residence time of suspended
solids within the boxes, since the residence time is equal to the length of the box divided
by the linear velocity. The suspended solids concentration entering each box is assumed
to be the average concentration inside the plume. The following assumptions are made in
the calculations:
(1) The solids entering the box remain suspended. Settling only occurs after the
particles exit. Therefore the PCB concentration on the settled solids equals the
PCB concentration on the particles exiting the box.
(2) The change in plume width occurs between boxes. Therefore both the dissolved
phase and the suspended PCBs are diluted before entering a subsequent larger
box. Additional background PCBs would be included at this point since the larger
plume width spreads into areas with a baseline concentration.
(3) Besides the partitioning between dissolved phase and suspended solids and loss
through settling, no other mechanism or reaction exists to affect the fate of PCB
Hudson River PCBs Superfund Site
Engineering Performance Standards
30
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
in the water column {i.e. volatilization, transformation, and reactions are not being
considered in this model).
The equations for the two-phase partitioning model based on the conceptual model and
assumptions above are listed below.
Equations for Entering Conditions
The volumetric flowrate (Q) must be calculated for each box (since it is dependent on the
width of the plume). The volumetric flowrate is calculated using:
Qi=i**h*yi (EQ 33)
Where:
u = ambient water velocity (m/s)
h = water depth (m)
The concentration of suspended solids within the plume must also be calculated for each
box. The suspended solids concentration given by the Kuo and Hayes Model above is
without background; therefore, the background concentration must be added for each
segment.
TSShm = TSSKuoIlayesi + TSSbkg (EQ 34)
The flux into the first segment
The total PCB concentration and the dissolved fraction for the background are known. In
addition, the dissolved fraction of PCBs from dredging activities is given either by the
CSTR model or by assuming it is zero. The concentration of PCBs from dredging
activities may also be given from the CSTR model or calculated by using:
p- * c'si * i o3
cTMsปs =- (EQ 3 5)
Where:
g = sediment loss rate (kg/s)
CSsed = concentration of the suspended sediment (mg/kg)
The total, dissolved and suspended PCB fluxes into the first segment are:
FcT.BKG, 1 Q.CTbkg
^CT, Dredge,\ Dredging
7
CT,Dredge, 1
(EQ 36)
FcTXin ^CT,Dredge,! + ^CT,BKG,\
FcD.BKG, 1 fBKG-^CT ,BKG, 1
-^CD,Dredge,! fDredge,\-^CT,Dredge,\
^CD,\,in ~ FCD,Dredge,\ + ^CD,BKG,\
(EQ 37)
FcS.BKG, 1 (1 IbKG ~)FCt,BKG, 1
^CS, Dredge, 1 0 fDredge,CT,Dredge,\
Fes,\,in ~ Fes,.Dredge,1 + FCS BKG,\
(EQ 38)
Hudson River PCBs Superfund Site
Engineering Performance Standards
31
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Where:
F = Flux (g/s)
/ = PCB fraction dissolved (unitless)
Subsequent segments:
For subsequent segments an additional flux from background will be added if the
plume width has increased. The additional background contribution and total flux
into box i+1 may be calculated as follows:
FCT,BKGMl = (ฃ?.ฆ+1 ~Qr)CTBKG*W6
FCT,i+\,in ~^CT,i,out ^CT,BKG,i+\
FcD,i+\,in ~ ^CD J,out + JBKO^CT ,BKG,i+\
Fcs,i+\,in ~ Fes,i,out + 0 ~~ fBKG ~)^CT,BKG,i+\
The average total and dissolved concentrations in the plume are calculated by dividing
the flux by the volumetric flowrate as:
(EQ 39)
(EQ 40)
(EQ 41)
(EQ 42)
F F
CT ฆ = * 106 CD ฆ = * io6 (EQ 43, 44)
' a M a
The average concentration on the particles is calculated by dividing the flux by the
volumetric flowrate and suspended solids concentration.
F
CS, ฆ = cs'hm * 106 (EQ 45)
' Qt*TSSUn
Equations for Inside Conditions (Approaching Equilibrium)
Inside the box Q, suspended solids, and the fluxes remain the same as the entering
conditions. The concentrations change as the PCBs begin to partition off of the particles
and into the dissolved phase. The retention time within the box is determined by:
Ax * v *h
a = ih l * 3 600 (EQ 46)
a
Where:
?i = retention time/suspended solids contact time (hr)
Hudson River PCBs Superfund Site 32 Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment D - April 2004
-------
If the retention time were long enough equilibrium would be achieved and the dissolved
and suspended concentrations would be:
CD = 7 (-^ n CS = CD x Kd x 10"6 (EQ 47, 48)
qi (l + ^x/XS^xlO ') q' q-
Where:
Kd= partitioning coefficient (L/Kg)
Before equilibrium is reached the dissolved and suspended concentrations must be
calculated using the following equations for net desorption:
CD, = CD,,, + (CO,, - CD,,,,) x (1 - e~u<) (EQ 49)
CS, = CS, -(cs-cs,ti )x (1 -) (EQ 50)
Where:
?= desorption rate constant (hr"1)
Equations for Exiting Conditions
The exiting dissolved and suspended (concentration on the particles mg/kg) are equal to
the concentrations inside the box or:
CDhOUt = CD, and CSUout = CS, (EQ 51, 52)
To calculate the total concentration, the suspended solids lost to settling must be
calculated. The suspended solids loss must be calculated using the suspended solids flux
since the plume volume increases in the next segment and the suspended solids
concentration is being diluted, therefore the suspended solids concentration in the i+1
box will not equal the suspended solids out of i. Suspended solids loss to settling can be
calculated as:
(/xv^a-yxsv^a,)
TSSSettledj = ' Q '+1 ^'+i; (EQ 53)
and
ISS: ::l ISS. TSS:.^ . : (EQ 54)
The total PCB concentration may be calculated as:
CTUottt = CD, + - TSS^out (EQ 55)
Hudson River PCB s Superfund Site
Engineering Performance Standards
33
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
The total, dissolved, and suspended fluxes are:
FCT^t=CThOUt*Qt* 10- (EQ 56)
FCD^t=CDiout*Qi*lO-6 (EQ 57)
Fes,out = CSKOUt * a * TSSUout * 1(T6 (EQ 58)
Equations for Net Conditions
To get the effects from dredging alone, the contributions from background must be
subtracted. The equations for the concentrations are as follows:
CT(net)!ml =CTIohi -CTm:a (EQ 59)
CD{net)i ojU = CD, -CDa:a (EQ 60)
-r<\
i,out i,out BKG 100 BKG
CS(net)i out = = (EQ 61)
l^Uout
Equation for the K,j value
From previous studies the background conditions are well defined. It is assumed that the
conditions of the background represent equilibrium. When the fraction of dissolved and
suspended concentrations is given and a background suspended solids value the Kd value
can be calculated by:
CD 1
B,:a = fBKG = t n (EQ 62)
CTBKa (l+^j x75S-0 *10 )
4.4.2 Relationship between CSTR-Chem and TSS-Chem
The objective of the models was to determine the relationship between suspended solids
and PCB (dissolved and particulate) fluxes downstream and resuspension rates. TSS-
Chem is useful for the near-field downstream transport of solids and PCBs but is
inadequate for modeling the resuspension from dredging activities. Therefore the CSTR-
Chem model must be used to translate the resuspension rate, and sediment characteristics
to the source strength and suspended solid characteristics used in the TSS-Chem model.
The source strength and suspended solid characteristics will in turn determine the
suspended solids and PCB fluxes downstream. The resuspension rate of sediments (input
to CSTR-Chem) and source strength of suspended solids (output of CSTR-Chem, input to
TSS-Chem) are not directly related since the CSTR-Chem model will provide a source
strength which has a width dependent on the dredge used and the TSS-Chem models a
point source. However, the CSTR-Chem can provide estimations of the initial conditions
Hudson River PCBs Superfund Site
Engineering Performance Standards
34
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
of the TSS-Chem, specifically the silt and coarse fractions within the sediment and source
strength and the initial dissolved fraction of PCBs in the source strength.
Dissolved PCBs from Dredging Activities
The results of the CSTR-Chem model showed that the suspension time of the solids
around the dredge head was not long enough to achieve equilibrium conditions. Though
some partitioning occurred between the PCBs on the resuspended sediments and the
water column, the results indicated that the amount of partitioning was negligible and the
dissolved PCB fraction exiting was insignificant. However, it was necessary to determine
the impact of an initial dissolved PCB source (other than background) on the PCB and
suspended solids fluxes downstream. Therefore, the TSS-Chem model was run for the
350 ng/1 far-field criteria scenario in River Sections 1 (2007) and 2 (2009) with and
without the dredging dissolved PCB concentrations obtained from the corresponding
CSTR-Chem runs. The results are shown in Table 8. The source strengths for the scenario
runs did not require adjustments since the PCB flux at one mile experienced a negligible
change. The suspended solids flux did not change given that it is not dependent on the
dissolved PCB concentration and the source strength was not adjusted. Therefore the
dissolved concentration directly around the dredgehead from the partitioning of
resuspended material has a negligible effect on the downstream PCB concentration and
could be assumed to be zero for the TSS-Chem model runs.
Silt and Coarse Fractions
When the fractions of silt and coarse material in the sediments were applied to the CSTR-
Chem model the residence time of the solids within the model was long enough to allow
a significant amount of coarse material to settle. For instance, the silt fraction in River
Section 1 sediments is approximately 0.37. When the resuspension of this material is
modeled using CSTR-Chem, the solids exiting the area around the dredge have a silt
fraction of 0.66. To determine the impact of the silt and coarse fractions on the source
strengths and fluxes, the TSS-Chem model was run for the 350 ng/1 far-field criteria
scenarios in sections 1 (2007) and 2 (2009) with and without coarse solids. The results for
these runs are shown in Table 9. As the table shows the effect of adding coarse solids
does not significantly affect the suspended solids or PCB flux. The total source strength
without coarse materials, however, must change to equal the silt source strength when
coarse solids are present. This illustrates that while the coarse materials will not have a
significant contribution on the relationship between PCB and suspended solids fluxes
downstream, they will affect the resuspension rates required to obtain those fluxes.
Therefore in calculating the different resuspension rate requirements it is necessary to
consider the coarse material.
4.4.3 Results
The results of the TSS-Chem analysis indicated that a significant amount of PCBs
released would partition off of the solids and become dissolved by a distance of one mile.
Hudson River PCBs Superfund Site
Engineering Performance Standards
35
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
The dissolved fraction at one mile is greater when the source strength is decreased. The
majority of the PCB load at one mile was contributed by the silt fraction, since the coarse
material generally fell to less than 0.1 percent of the total solids within the plume within
30 meters downstream. The results for the average source strength analyses and near-
field suspended solids criteria are discussed below.
4.4.3.1 Average Source Strength Estimations
The resuspension rate is the rate at which sediments directly around the dredge will be
suspended into the water column. Before the sediments are available for transport
downstream resettling in the dredge area occurs. The resettled material is predominately
coarse sediment. The particles that do not resettle around the dredge move downstream.
The rate at which the particles are transported downstream out of the immediate dredge
area is the source strength.
As outlined in Appendix E.6 of the FS and White Paper: Resuspension of PCBs During
Dredging (336740) of the RS, the average resuspension rate is based on a combination of
field data from other sites and a resuspension model. The downstream transport rates
(source strengths) only apply to silts and finer particles (65 percent of cohesive and 20
percent of non-cohesive sediments for the Hudson River) within the sediment. The use of
only silts does not significantly affect the PCB flux estimates since the silt resuspension
rate (which is essentially equal to the silt source strength) is the driving source term for
the PCB flux downstream. This aspect of the models is discussed in Section 4.4.2 of this
attachment.
The average source strength in the FS was originally based on the cohesive sediments.
An estimate of 0.3 percent of cohesive sediments was expected to be available for
transport downstream. Since this only applies to silt, the percentage can be normalized to
the silt fraction in cohesive sediments as 0.003^-0.65 to yield 0.5 percent of silts and finer
particles. The contribution to the average source strength from non-cohesive sediments
must also be added to the average source strength since they are 20 percent silts. The
overall fraction of non-cohesive sediments is 0.005x0.2 or 0.1 percent of cohesive
sediments. Since silt fractions can be estimated for each section based on the percentages
of silts in cohesive and non-cohesive sediments (given above) the source strengths can be
calculated as 0.5 percent of the production rates of silty sediments.
The production rates were based on a total of five dredging seasons (two half and four
full seasons). Given the amount of sediment removal necessary and the time limitations
involved, the average production rates for each river section were calculated. The silt
fractions in each river section were applied to yield an average source strength. Each
source strength was run through TSS-Chem to estimate the resulting flux and
concentration increases at one mile. The production rates, source strengths, and results
are shown in Table 10.
Hudson River PCBs Superfund Site
Engineering Performance Standards
36
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Model Revisions from FS Appendix E.6 and RS White Paper Semi-Quantitative
Estimates
As part of the FS and RS semi-quantitative assessments of water quality impacts
associated with dredging activities were performed. The assessments utilized the
DREDGE model (discussed in section 3.0 of this attachment) which is similar to TSS-
Chem, however the assessments were not as extensive as those performed for the
resuspension performance standard modeling. The semi-quantitative assessments had
several assumptions that were modified by the new models. In the analysis of the FS and
RS, a model similar to the TSS-Chem model was used to estimate the solids plume within
10 meters of the source term. The estimates of the plume in this model and the TSS-
Chem model use the same modeling equations for solids but differ in the modeling of
PCB concentrations. The modeling of solids for the TSS-Chem calculations does not use
the same parameters as the solids modeling in Appendix E.6. The parameters were
revised as part of an extensive literature search since the publication of the FS. The
various parameters (i.e. dispersion coefficient and settling velocity) and the rationale for
their current values are discussed in Section 4.4.1 of this attachment. The differences
between the analyses and the individual effects of the differences (overall effects will
vary) are discussed below.
The three differences that had the greatest effects on the estimates were:
Mass was conserved - The suspended solids plume equations will predict
concentrations to infinity. In the previous analyses the solids concentration was cut-
off at 1 mg/L (or 0.5 mg/L if no values were greater than 1). Therefore the mass
outside the cut-off concentration was not accounted for in the suspended solids or
PCB flux. In order to preserve mass the TSS-Chem model uses the integrated form
of the suspended solids plume equation. The new method increases the suspended
solids and PCB concentration and flux estimates for any given resuspension rate.
Even if all the other parameters had remained the same the suspended solids Flux
estimates at 10 meters with mass conserved in River Section 1 increases from 11.5
to 40 g/sec and in River Sections 2 and 3 from 30.1 to 52 g/sec.
PCB phase partitioning was included - The TSS-Chem model estimates the phase
partitioning of PCBs from suspended to dissolved phases. When partitioning is
taken into account the PCB flux and water column concentrations increase relative
to the approach used in the FS and RS since the particles settling have a lower
concentration and more PCBs remain in the water column. For the average source
strengths, the TSS-Chem model estimates net PCB fluxes that contain more than
one third dissolved PCBs.
Settling velocity of silts was decreased - A decrease in the settling velocity of the
silts, causes an increase in PCB concentration and flux estimates. After an
extensive literature search the settling velocity was estimated to be an order of
Hudson River PCBs Superfund Site
Engineering Performance Standards
37
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
magnitude lower than was previously predicted. The revised settling velocity
greatly increased the amount of solids and PCBs lost to downstream transport.
Other differences that affected the solids and PCB estimates are:
Plume width concentration was decreased - The former models defined the plume
width as described above (greater than 1 mg/L or greater than 0.5 mg/L if no values
were above 1 mg/L). TSS-Chem defines the width of the plume by concentrations
greater than 1% of the center concentration. The plume width is greater using the
current method, however, the volumetric flow rate of the plume varies accordingly
and width will not directly affect flux. The concentration in the plume is dependent
on the width (concentration will decrease with increasing width), however due to
the difference in plume concentration estimated (see "mass was conserved" above)
the new method did not decrease the plume concentrations. This increase in the
plume width is a model constraint and is not directly related to the change in the
lateral dispersion coefficient discussed below.
Dispersion coefficient was decreased - A decrease in dispersion coefficient
increases the PCB concentration within the plume by decreasing the width, but
does not change the average river-wide concentration or the flux.
Linear velocity was increased - An increase in velocity results in an increase in the
PCB concentration and flux estimates.
Depth was decreased - A decrease in depth results in a decrease in the PCB
concentration and flux estimates.
River-wide volumetric flow was increased - The flow examined was changed from
3,000 cfs to 4,000 cfs, since 4,000 cfs is approximately the average flow of the
summer months across the three river sections. An increase in flow decreases the
PCB concentration but increases the PCB flux.
Distance downstream was increased - The suspended solids plume concentrations
in Appendix E.6 were taken for a distance downstream of 10 meters from the
source term. No further removal by settling was permitted. For the revised PCB
flux, the TSS-Chem model was extended to one mile downstream allowing for
further settling between 10 meters and one mile. An increase in distance, and
thereby in settling, will decrease estimates of PCB concentration and flux.
PCB basis changed from Tri+ to Total - The Tri+ PCB concentrations were used in
the former analysis while the new estimates are based on Total PCB concentrations.
This would not change the Total PCB flux unless the PCB sediment concentrations
and Tri+ to Total PCB ratio were revised. Both the sediment concentrations and the
Tri+ to Total PCB ratios were revised from the FS values as part of the RS. The
values from the RS were used in this analysis.
Hudson River PCBs Superfund Site
Engineering Performance Standards
38
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
4.4.3.2 Particle Settling Results
Some fraction of the sediment resuspended from the dredge will settle downstream. If the
material is contaminated, this will add to the PCB mass and concentration in the
surrounding downstream areas. Using the modeled suspended solids concentrations in the
water column downstream of the dredge with the associated PCB concentration on the
suspended solids, it is possible to estimate the increase in PCB mass in these areas. The
increase in mass per unit area and the length-weighted average concentration of the top
six inch bioavailable layer will be used to measure the effect of the settled material.
The amount of settled material is estimated by calculating the mass of suspended solids in
the water column at each modeled location. The mass at each cross section is summed.
The difference in mass between each cross section is the amount of solids that has settled
downstream. The loss for each section is distributed in the cross section in the same
proportion as the amount of mass in the water column along the cross section. The rate of
deposition is calculated considering the flow rate. Using the PCB concentration estimated
for the suspended sediment, the rate of PCB deposition is estimated at each modeled
location.
The spatial distribution of the settled contamination will vary according to the shape of
the target area and the rate of dredging. For this estimate, the target area is assumed to be
5 acres, 200 ft across and approximately 1,100 ft long, because the areas of
contamination are typically located in the shoals of the river and are narrow. From the
FS, a time needed to dredge a 5-acre area with 1 m depth of contamination would take 15
days operating 14 hours per day. It is assumed that the dredge will move in 50 ft
increments across and down the target area. With this assumption, the dredge will
relocate approximately every two hours. To simulate the deposition of settled material,
the amount of PCB mass per unit area, the mass of the settled material and the thickness
of the settled material that is deposited in two hours downstream at each modeled
location is added on a grid as the dredge moves across and down the area.
The TSS-Chem results for each river section and action levels were used to estimate the
additional mass per unit area and length weighted average concentration in the target
area, 100 feet to the side of the target area and approximately 2 acres downstream. The
remediation could operate continuously at Evaluation Level of 300 g/day or the Control
Level of 600 g/day but not Control Level of 350 ng/L. The results are shown in Table 11.
The increase in mass per unit area can be compared to the mass per unit areas values used
to select the target areas in River Sections 1 and 2. Areas in River Section 3 are not
selected on the basis of a single mass per unit area value. The Tri+ PCB mass per unit
area values for River Sections 1 and 2 are 3 g/m2 and 10 g/m2 Using the conversion
factors for Tri+ PCBs to total PCBs (USEPA, 2002), the total PCB mass per unit area for
River Sections 1 and 2 are 6.6 g/m2 and 34 g/m2. It is estimated that only a small amount
of PCBs will be deposited in the area to the side of the target area with the greatest
increase in mass per unit area being only 0.004 g/m2 in River Section 3.
Hudson River PCBs Superfund Site
Engineering Performance Standards
39
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
In the target area, the increase in mass per unit area is more substantial. The mass per unit
area increases by 1.8 g/m2 in River Section 1 for the Control Level of 600g/day, which is
nearly a third of the value used to select the areas. In River Section 2, the increase in
mass per unit area is nearly the same as in River Section 1, but this increase is only 4
percent of the value used to select the areas. For Control Level of 350 ng/L, the increase
in mass per unit area is 3.9 g/m2 in River Section 1 (65 percent of the value used to select
2 2
the areas), 4.7 g/m in River Section 2 and 5.6 g/m in River Section 3.
In the area immediately downstream of the target area, in River Sections 1, 2 and 3 for
Evaluation Level, the increase in mass per unit area is 0.2, 0.1 and 0.2 g/m2, respectively.
The mass per unit area increases another 2 to 3 times for the 600 g/day Total PCB
scenario over the Evaluation Level and increases another two to four times between the
600 g/day and 350 ng/L Total PCB scenarios. These increases in mass per unit area are
only significant for Control Level Total PCB criterion of 350 ng/L in River Section 1,
which is 17 percent of the value used to select the areas.
The length weighted area concentrations were calculated assuming that the PCB
concentration in the sediment underlying the settled material is 1 mg/kg. The ROD
defines 1 mg/kg as the acceptable residual concentration. In the area to the side of the
target area, no increase in concentration was found. In the target area, the concentrations
range from 5 to 29 mg/kg. In the 2 acres below the target area, the concentrations range
from 2 to 9 mg/kg. These increases suggest that dredging should proceed from upstream
to downstream if no silt barriers are in place so that settled material can be captured by
the dredge. Also, silt barriers may be needed to prevent the spread of contamination to
areas downstream of the target areas have already been dredged or are not selected for
remediation. This settled material is likely to be unconsolidated and easily resuspended
under higher flow conditions.
4.4.3.3 Suspended Solids Near-field Criteria and Monitoring Locations
Introduction
PCB criteria for resuspension are set in terms of concentration or load at the far-field
monitoring stations. Achieving these criteria requires controlling the PCB concentration
and flux from the dredging operation. Paired with the far-field PCB monitoring,
suspended solids will be measured at the near-field locations in order to provide the real-
time or near real-time monitoring for the potential contaminant flux from the dredging
operation. High levels of suspended solids in the near-field may result in exceedances of
the PCB criteria at the far-field stations, and therefore should trigger some level of
concern. The near-field suspended solids criteria have been developed corresponding to
the far-field PCB action levels. HUDTOX and TSS-Chem models were utilized to
simulate the connection between the far-field PCB concentrations and loads and the near-
field suspended solids concentrations.
Hudson River PCB s Superfund Site
Engineering Performance Standards
40
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Approach
The HUDTOX model was used to predict the PCB levels at the far-field stations.
Therefore, for the Control Level, the regulated PCB load of 600 g/day is the output flux
simulated by HUDTOX. Similarly for the total PCB concentration criterion of this action
level, (350 ng/L), the PCB loads were calculated (at different flows) and were the output
fluxes of the HUDTOX model (Hout).
HUDTOX simulates an effective rate of PCB loss during transport, due to volatilization
and settling. The percentage reduction (1 - output flux/input flux) during transit through a
river section varies by section and by year of operation. The percentage reduction
obtained from previous HUDTOX runs was used to estimate the input of HUDTOX runs
(H;n) which will result in the PCB level at the far-field stations corresponding to the
action levels. When performing the near-field and far-field model simulation, it is
assumed that PCB flux 1 mile downstream of the dredge head estimated by the TSS-
CHEM model (Timiie) is the input flux for the HUDTOX model (Hni), The input flux for
TSS-Chem (T;n) was determined by trial and error, until the simulated plume at one mile
(Timiie) matched the targeted input to the HUDTOX model. The resulting suspended
solids concentrations in these simulations was used as the basis to develop the near-field
criteria.
Since some of the TSS-Chem input parameters, such as lateral dispersion coefficient and
flow velocity, are flow-dependent, the resulting suspended solids and PCB concentrations
and loads are also flow-dependent. As mentioned above, when the output concentration is
set as the target value at the far-field stations, the associated load will be calculated and
used as the controlling value in the whole process of estimation. Load varies with flow
when the concentration is constant. Therefore, it is expected that different flows will
generate different plumes at the near-field locations, which means that at the same
location, the estimated suspended solids concentration can be significantly different when
the flow varies. Suspended solids concentrations at different flows were fully investigated
and the most reasonable value, which provides the best representation of the near-field
conditions, was chosen as the basis to develop the near-field suspended solids standard.
Since the model simulation determines the values and no actual data is available, other
uncertainty factors were taken into account while finalizing the criteria. Criteria were
only formulated for the Evaluation Level and Control Level to avoid unnecessary
shutdowns.
Results
Multiple TSS-Chem runs were used to simulate the suspended solids plume in the near-
field using the one mile downstream PCB flux as the controlling factor. The estimated
suspended solids concentrations downstream of the dredge head for River Section 1 at
4,000 cfs and a far-field PCB concentration of 500 ng/L is shown in Figure 16. The
profile shown in Figure 16 is a good representation of the estimated suspended solids
plumes under all scenarios. The suspended solids concentration decreases and the width
Hudson River PCB s Superfund Site
Engineering Performance Standards
41
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
of plume increases as the solids area transported downstream. The suspended solids
concentration at 300 m downstream is about Vi to 1/3 of the concentration at 50 m
downstream. Assuming that the boundary of the plume is the location where the
suspended solids concentration is 5 mg/L higher than the background level (2.3 mg/L in
River Sectionsl and 2, 1.7 mg/L in River Section3), the width of the plume at 50 m, 100
m, 300 m and 600 m downstream is 21 m, 29 m, 47 m and 61 m, respectively, for the
scenario shown in Figure 16. The plume widths at these locations for other scenarios are
within the same scale. Since the plume is wider further downstream there is more
assurance that a sample collected at 300 m is within the plume than a sample collected a
50 m. At 50 m downstream, due to the narrow width, it is possible to miss the plume
when collecting a sample. This could potentially cause a large exceedance at the far-field
stations without any indication in the near-field. In addition, the curved shape of the river
channel at some points will make it more difficult to predict the direction and the location
of the center of the plume when going further downstream. However, further downstream
the plume is more diluted and less visible. Therefore it is possible to miss the plume
when collecting a sample. In order to counter balance the requirements, ease of sampling
within the plume and ease of identifying the plume, two near-field locations are
necessary. From the results of this analysis 100 m and 300 m were chosen as the near-
field monitoring locations downstream of the dredge.
As mentioned in the approach section, flow will change the current velocity and the
lateral dispersion coefficient, which result in different suspended solids concentrations
corresponding to the same PCB level at the far-field station. Figure 17 presents the
suspended solids concentration at 300 m downstream when only flow varies. Consistent
with intuition considering the dilution caused by the flow, a 2,000 cfs flow results in the
highest concentration and the lowest concentration occurs with the 8,000 cfs flow. But
the difference in concentration is not directly proportional to the flow mainly due to the
changes in the lateral dispersion coefficient. Since the flow will vary during dredging a
conservative criteria was selected. Therefore the criteria were based on the lowest
suspended solids level at 8,000 cfs flow.
Estimated suspended solids concentrations within the plume are used to set the criteria.
As mentioned above, the boundary of the plume is determined by the location where the
suspended solids concentration is 5 mg/L above the background level. The average flow
during the dredging period is assumed to be 4,000 cfs. To provide a common basis for
comparing the concentration at different flows, the width of the plume determined by the
4,000 cfs flow is applied to other flow conditions. That is, if the width of plume at 300 m
downstream is 47 m when the flow is 4,000 cfs, the widths of plume at the same location
under other flows are 47 m as well. As noted above, suspended solids concentration
under the high flow is lower than the suspended solids under the low flow. Since the
width of the plume is determined by the concentration at the 4,000 cfs flow and the
plume at 8,000 cfs is actually not as wide, the average concentration calculated at 8,000
cfs is underestimated. This results in lower values and thereby conservative criteria.
Mean suspended solids concentrations within the plume at 300 m downstream at 8,000
cfs are summarized in Table 12 for each section, corresponding to each far-field action
Hudson River PCB s Superfund Site
Engineering Performance Standards
42
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
level. The suspended solids levels are similar in River Sections 1 and 3, while the
concentrations in River Section 2 are approximately half of the values for River Sections
1 and 3. This is due to the higher average PCB sediment concentration in River Section 2.
The average PCB concentration on the dredged sediment is 27, 62 and 29 ppm for
Section 1, 2 and 3 respectively. Since the PCB far-field criteria are the same for all three
river sections, and dredging in River Section 2 is expected to suspended solids with
higher PCB concentrations, section specific SS criteria are necessary. The same criteria
may be applied to River Sections 1 and 3 since the average PCB sediment concentrations
in these sections are similar.
Suspended solids concentrations reported for the water column monitoring samples
collected during the dredging operations in the Lower Fox River SMU 56/57 and New
Bedford Harbor pre-design field test were reviewed and compared to the numbers
simulated by the models. During the SMU 56/57 work, the downstream suspended solids
samples were collected at fixed locations within 800 ft downstream of the dredge head.
Most suspended solids numbers fall between 20 and 40 mg/L, with one greater than 100
mg/L and two around 80 mg/L. During New Bedford Harbor pre-design field test,
suspended solids samples were collected at different locations within 1000 ft down
current of the dredge head. These data were in the range of 10 -30 mg/L. Assuming that
the suspended solids concentrations in the Hudson River during dredging are similar to
these two projects, the action level corresponding to the 600 g/day of total PCBs at the
far-field stations exceed too frequently and possibly cause unnecessary contingencies.
Therefore, the SS action level criteria are not based on the numbers determined by 600
g/day of total PCBs, but are based on the numbers corresponding to 350 ng/L at the far-
field stations
The near-field suspended solids standard assuming hourly samples is finalized and
summarized below.
River Sections 1 and 3 (100 mg/L) and River Section 2 (60 mg/L)
Evaluation Level 6 hrs continuously or 9 hrs in a 24 hour period
Control Level daily dredging period or 24 hour period
Monitoring of suspended solids at near-field stations is intended to provide timely
feedback and allow prompt adjustments to be implemented in order to avoid any
significant impact on the far-field stations. Decisions to shutdown operations will be
made based on the PCB levels at the far-field station.
The concentration limits (100 mg/L and 60 mg/L) are based on model predictions of a
total PCB concentration of 350 ng/L at the far-field station as listed in Table 12.
Evaluation Level and Control Level use the same concentration limit but different
durations. The duration is chosen based on engineering judgment with an emphasis on the
cumulative impact of resuspension on the water quality due to dredging. The impact of a
long period with a relatively low concentration is more significant than one sample with a
high concentration. It should be noted that the suspended solids concentration regulated
Hudson River PCBs Superfund Site
Engineering Performance Standards
43
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
herein is the net suspended solids concentration increase, which is the suspended solids
concentration 300 m downstream of the dredge head minus the suspended solids
concentration upstream of the dredge head, in order to control the suspended solids
increase from resuspension and thereby maintain consistent correlation between the PCB
concentrations and loads and sediment concentrations.
According to the monitoring plan, the near-field suspended solids sample will be
collected at 5 stations, one upstream, one close to the side channel, and three
downstream. The upstream sample will provide the background suspended solids level
necessary to calculate the net suspended solids increase caused by dredging. The sample
for the side channel is intended to provide information on the suspended solids caused by
river traffic. For the three samples collected downstream, one will be located at 100 m
downstream of the dredge operation and two will be located at 300 m downstream. Even
though the criteria are based on the suspended solids level at 300 m downstream, a
sample collected 100 m downstream will provide information on how the suspended
solids are being transported downstream, and may be useful for Phase 2 work if
modifications based on Phase 1 results are necessary. The higher concentration between
the two samples collected 300 m downstream will be used for determining compliance
with performance standards.
In addition to the performance standards above, a second Evaluation Level criteria is set
at 700 mg/L for over three hours at 100 m downstream. This concentration limit is
estimated based on the maximum concentration within the plume at 100 m downstream
corresponding to a total PCB concentration of 500 ng/L at the far-field station and a flow
of 8000 cfs. Collection of PCB samples at the nearest far-field station should be designed
to sample the suspended solids release of concern based on the travel of time and any
necessary engineering contingencies will be based on the PCB results.
In the formulation of the criteria above no assumptions were made for solid control
measures. At any location where a solid control measure such as a silt curtain is used, as
described in the monitoring section, the near-field downstream location should be 150 m
downstream of the most exterior silt control barrier. Under these conditions the single-
level concentration standard (700 mg/L) is not applicable.
4.4.4 Sensitivity Analyses
Two sensitivity analyses were performed. The first analysis examines the distribution of
PCBs on the fine and coarse-grained sediments, to determine if they should be modeled
with different concentrations. The second sensitivity analysis varies all the inputs one at a
time to determine which parameters have the greatest impact on the model outputs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
44
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
4.4.4.1 Fine and Coarse-grained PCB Distributions
The analysis presented below uses published data from River Section 1 sediment to
examine the
relationships between grain size, organic content and Total PCB concentration. The
limited data set was used to provide a ratio of Total PCBs for the fine and coarse-grained
sediments. Using these ratios dredging-related PCB resuspension (assuming the average
source strength) was modeled for different fine-grained Total PCB sediment
concentrations.
The original analysis of the source strength, modeled at 4000 cfs with an average Total
PCB sediment concentration of 27 mg/kg, yielded a Total PCB flux of 78 g/day.
Published grain-size, organic content and PCB data indicated that the Total PCB
concentration on the fine-grained sediments may range from 30 to 36 Total PCB mg/kg.
The TSS-Chem transport model indicated that these concentrations on the fine-grained
sediments for flows ranging from 2000 to 5000 cfs have PCB fluxes at one mile of 44 to
115 g (Total PCB) /day. Therefore, the model indicated that the Total PCB
concentrations investigated do not represent a significant change in the flux or the water
column concentration increase, particularly when the uncertainties in sediment
homogeneity and river-wide flowrates are considered.
Although the results suggest that the original estimate may not be as conservative as
possible, there are many other conservative assumptions in the model. Due to limitations
of modeling, the resuspension criteria and action levels were based on the MCL and fish
body burdens in the Lower Hudson. The modeling was used as an aid in estimating the
resuspension rates each of the criteria may represent. During Phase 1 the model will be
reevaluated and possibly modified.
Discussion
While USEPA recognizes that PCB concentrations are generally higher in fine-grained
sediments relative to coarse-grained sediments when classified as a whole sample, it is
not clear that the organic carbon content within a sample can approximate this
relationship. That is, it is not clear that within a given sample, the PCB content of each
grain-size fraction is well approximated by the organic carbon content for the sample.
The lack of a direct correlation between organic carbon content and PCB concentration
can be seen in Figure 3-21 of the Low Resolution Sediment Coring Report (USEPA,
1998), included in this attachment as Figure 18. This figure shows that PCB
concentration does not increase linearly with TOC and that significant variation can be
found at any organic carbon concentration. The USEPA agrees that there may be some
enhancement of PCB concentration with smaller particles but it is not clear that the
response is linear.
According to a study of contaminated Hudson River sediments conducted by General
Electric Corporate Research and Development and MIT published in Environmental
Hudson River PCBs Superfund Site
Engineering Performance Standards
45
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Science and Technology (Carroll et al, 1994) the Hudson River sediments greater than
0.069 |im (sand) had % TOC values from 3.2 to 7.3 while the sediments less than 0.069
|im (silt/clay) had a %TOC value of 3.9, indicating little if any difference. These data
suggest that the organic carbon content is relatively homogeneous in fine-grained
sediments. The data set presented in the paper represents a limited number of samples so
it is unclear how far this data can be extrapolated. Nonetheless, it indicates that organic
carbon content may not vary with grain size fraction in fine-grained sediments.
Furthermore the PCB concentrations for these sediment fractions did not substantively
differ. The sand fraction PCB concentrations ranged from 203-284 ppm and the silt/clay
concentration was 338 ppm. The data are shown in Figure 19. If the ratio of these samples
(which were all taken from Moreau NY, and therefore only represent River Section 1) were
assumed to be applicable to the average sediment concentration in River Section 1 (27 ppm), the
silt Total PCB concentration would range from 30 to 36 ppm. The equations used to estimate this
range are shown below (River Section 1 has an estimated silt fraction of 37%).
Csilt f silt ^ Ccoarse fcoarse CTotal
Csilt fsilt R-Qtio coarse -to-silt Csilt fcoarse ~ CTotal
or,
Csilt = :<^EฃmL (EQ
fsilt + Rdtio coarse-to - silt 0 fsilt )
Where:
C= PCB concentration (mg/kg)
f = fraction (kg sediment type/kg total)
RatioCoarse-to-siit = Ratio of PCB concentrations on coarse-grained and silty
sediments
Further TSS-Chem model runs were performed using River Section 1 Total PCB silt
concentrations of 27, 30 and 36 mg/kg and river-wide flows of 2000, 4000, and 5000 cfs.
The results are shown in Table 13.
Results
The PCB flux using the values from the previous source strength modeling (27 Total
PCB mg/kg and 4000 cfs) was 78 g (Total PCB) /day at one mile. With the different
concentrations and flows the PCB fluxes ranged from 44 to 115 g (Total PCB) /day. The
Total PCB water-column concentration modeled in the original analysis was 14 ng/L at
one mile. With the different flows and sediment concentrations the water-column
concentration was modeled to range from 13-19 ng/L. Given the dependency of Total
PCB flux on flow, the uncertainty introduced by using the average sediment
concentrations instead of the silt concentrations (exhibited by the data from Carroll et al,
1994) is not significant.
Hudson River PCB s Superfund Site
Engineering Performance Standards
46
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Conclusions
Although these results suggest that the estimates originally presented may not be as
conservative as possible, they are still quite conservative based on other assumptions
made in the development of the standard. In particular, the model transport mechanisms
themselves are quite conservative. For example, the source strength term is derived from
an upper-bound estimate of the releases due to dredging. Secondly, the transport
mechanisms have been idealized and further settling of particles is expected relative to
the model predictions.
4.4.4.2 TSS-Chem Model Sensitivity Analysis
The sensitivity of four modeled outputs were examined for the TSS-Chem model. The
four output values selected to assess the sensitivity of the above parameters are defined
as:
The net fraction of dissolved PCBs from dredging is equal to the dissolved PCB
concentration minus the dissolved background concentration, divided by the total
PCB concentration minus the background PCB concentration.
The distance downstream from the dredge at which the coarse material is less than
0.1 percent of the net suspended solids from dredging.
The net total PCB flux at one mile, which is the flux at one mile minus the
contribution from background.
The net suspended solids flux at one mile, which is estimated as the flux at one
mile minus the contribution from background.
Two of the outputs, the net suspended solids and PCB fluxes, are inputs in HUDTOX.
The other two outputs examined are the net dissolved PCB fraction and the distance
downstream at which the coarse material is less than 0.1 percent of the net suspended
solids. To test the sensitivity of these outputs, each input parameter was varied within
reasonable ranges while the others were held constant and the effect on each output was
examined. The ranges used for each input parameter are shown in Table 14.
The model parameters on which the sensitivity analysis was performed include:
Volumetric flow (thereby linear flow, depth, and lateral dispersion),
Source strength,
Silt fraction of the entering solids (from dredging),
PCB sediment concentration,
Background conditions (suspended solids and PCB concentrations, and dissolved
PCB fraction),
Partition coefficient,
Desorption rate,
Hudson River PCBs Superfund Site
Engineering Performance Standards
47
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Lateral dispersion coefficient, and
Settling velocities of silt and coarse solids being transported downstream
Along with the general effects on modeled outputs, the relative change caused by varying
each input was examined. The relative change of an input parameter on the output (X)
was calculated by the sensitivity of the parameter Sparameter,x as defined by Gbondo-
Tugbawa et al., 2001:
s (Output, - Outputdefault) / Outputdefault
Paramater,Output / r* > > \ / ^
(Paramteri - Parameterdefault) / Parameterdefault
The higher the value of the average Sparameter,Output, the more sensitive the model output is
to that parameter. The relative sensitivities of the parameters were ranked by the
magnitude of their average Sparameter, Output- If the parameter was among the top 30 percent
in the ranking the relative sensitivity was labeled as "high", within 60 percent was
"moderate" and below that was "low". If the output was not sensitive to the parameter it
was labeled as "none".
Results
The input ranges are presented in Table 14. Direct and indirect relationships between the
various inputs and outputs are indicated in Table 15. The relative sensitivities are
qualitatively given in Table 13. The average of the absolute Sparameter,Output values are
presented in Table 16.
Flow
The first parameter examined was the river-wide volumetric flow since this is an
environmental parameter and is likely to vary continuously. The river-wide volumetric
flow was varied from 2000 to 8000 cfs which is consistent with the natural variation
between low and high flow in the Hudson River. However, it should be noted that
dredging activities are not expected to occur at such high flow rates (8000 cfs). The
default value is 4000 cfs since this is the average flow for the summer months. By
changing the river-wide volumetric flows, three model parameters (linear velocity, depth
and lateral dispersion) were varied. Using the RMA2 model (at RM 190 and RM 193) the
linear velocities and depths for these river flows were acquired as input for the TSS-
Chem model. River-wide flows have specific linear velocity-depth pairs, however since
the width of the river is not constant there is more than one depth-velocity pair for each
river-wide flow. In addition, the lateral dispersion is a function of linear velocity since it
is dependent on the shear forces. The results for various river-wide flows are shown in
Figure 20. Due to the variations in the other input parameters there is no consistent effect
of varying the river-wide flow. In order to provide a clear representation of the effects
each input parameter (velocity, depth and dispersion coefficient) was examined
separately.
Hudson River PCBs Superfund Site
Engineering Performance Standards
48
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Velocity
The velocity was varied separately in the range of linear velocities that apply to the river-
wide flow rates discussed above. The results of varying the velocity are shown in Figure
21. By varying the velocity, the solids will reach one mile downstream in less time.
Therefore, the PCBs on the solids will have less time to partition into the water column
and the net dissolved PCB fraction will decrease. Likewise, the solids will have less time
to settle and the distance at which the coarse solids are less than 0.1 percent of the net
solids and the net suspended solids flux will increase. The net PCB flux increases as well
since a large fraction of the PCBs are associated with the solids flux. As shown in Figure
21 the net suspended solids flux and net PCB flux are closely correlated to each other.
Depth
The depth was varied separately using the depths that apply to the river-wide flow rates
discussed above. The results are shown in Figure 22. For this model the depth affects the
amount of settling that will take place and the volumetric flow inside the plume. With
increasing depth the amount of solids lost to settling decreases therefore the solids remain
suspended in the water column for a longer period of time and have more time to
partition, increasing the dissolved fraction. The decrease in settling also increases the
fluxes and the distance at which coarse materials are less than 0.1 percent of the net
solids. As shown in Figure 22 there is still a strong correlation between PCBs and
suspended solids with varying depths.
Source Strength
The source strength was varied from 0.01 kg/s to 40 kg/s. This upper limit was chosen
since the production rates in the various river sections are expected be around 40 kg
solids/s. It should be noted that this upper bound is unrealistic as a source strength since
at this rate the dredge would be resuspending all of the material it is collecting,
furthermore the reduction of suspended solids in the near-field due to settling (as
exhibited by the CSTR-Chem model) is not being taken into account. For the TSS-Chem
runs used to obtain HUDTOX inputs this parameter is set by the standard being
examined. For instance if the HUDTOX output of 600 g/day was being examined the
source term in the TSS-Chem model was increased until the PCB flux out of HUDTOX
equaled 600 g/day. Therefore there is no clear default value and 1 kg/s was chosen.
The results of varying the source strength are shown in Figure 23. As the source strength
is increased the net dissolved concentration increases. The net dissolved fraction however
decreases since the system is being overwhelmed by solids and the PCBs associated with
them. The distance that the coarse material becomes less than 0.1 percent of the net solids
remains constant since it is only a function of the flow, settling rates and initial silt
fraction. Both the net total PCB flux and the net suspended solids flux have a direct linear
relationship to the source strength.
Hudson River PCBs Superfund Site
Engineering Performance Standards
49
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Silt Fraction Entering
The silt fraction entering was varied from 0 to 1. It is anticipated that the fraction will be
closer to unity since the coarse materials are less prone to resuspension and have a greater
settling velocity. However due to the heterogeneous nature of sediments within a river the
full range including all coarse material was applied. The default value of 0.66 was
obtained by entering the fractions of silt and coarse in the sediments of Section 1 into the
CSTR-Chem model with the same parameter values used in the TSS-Chem model runs.
The net silt fraction exiting the CSTR-Chem model (0.66) was then used as the input of
the TSS-Chem model.
The results with varying silt fractions are shown in Figure 24. Since silt has a lower
settling rate than coarse solids, an increase in the silt fraction entering the system will
cause more solids to remain in the water column longer. With increasing silt fractions, the
solids are available for partitioning longer and the dissolved PCB concentration increases.
However by increasing the initial silt fraction, the suspended PCB fraction at one mile
also increases. The overall effect tends to drive the dissolved PCB fraction down, as is
shown in Figure 24.
The distance to 0.1 percent coarse material decreases as less coarse material is added into
the system. The relationship is not linear and the distance is noticeably less sensitive
between initial silt fractions of 0.1 to 0.9 in which the distance only changes by 18
meters.
As shown in Figure 24, both the net PCB flux and the net suspended solids flux linearly
increase with increasing silt fraction entering. As was discussed above the increases are
due to the lower settling velocity (less settling) and the greater time period available for
partitioning.
PCB Sediment Concentration
Due to the heterogeneous nature of the sediments the PCB concentration may have large
variations and therefore the range used for the sensitivity analysis is also large (1 to 1000
mg/kg). The default value of 27 mg/kg is the average concentration of the sediments that
will be removed in River Section 1. The results for the varying sediment concentrations
are shown in Figure 25.
Neither the distance at which the coarse material is less than 0.1 percent of the net solids
nor the net TSS flux are dependent on PCB sediment concentrations. The net dissolved
fraction increases with increasing sediment concentration, however the sensitivity of the
parameter is greatest between 1 and 20 mg/kg. As shown in Figure 25, above 20 mg/kg
the fraction begins to plateau. The reason this occurs can be shown by examining the
calculations for the net dissolved fraction. Equation 65 below is the equation for the net
dissolved fraction (for a small ? x):
Hudson River PCB s Superfund Site
Engineering Performance Standards
50
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
f
CPout-CPBKG
CTout-CTBKG
CD, +
CT
(i+^(xrs,xio-s)
-CD,
x(l -elt)-CD
BKG
CT -CT -CT
in settled ^1 BKG
(EQ 65)
Where:
TSS = Concentrations of suspended solids (mg/1)
CD = Dissolved PCB concentration in the water column (ng/1)
CT = Total PCB concentration in the water column (ng/1)
x = Distance downstream (m)
Kd = partitioning coefficient (L/Kg)
? = desorption rate constant (hr"1)
BKG = Background, and
In, out and settled apply to the concentrations in, out and settling for ?x.
The equation can be simplified by grouping some of the parameters that are not
dependent on the sediment concentration such as Kd, TSSin, e" .
C/A, +
f CT ^
" -CD,
x E - CDbkc
in nKLr
f net,dissolved * 1 rI' f 1 rI' f 1 rI' (66)
in settled BKG
As the sediment concentration increases CTinปCDin>CDBKG, and CTinปCTsettied and the
fraction begins to approach CTin/CTin*constants.
The net PCB flux is highly sensitive to the PCB sediment concentration as is exhibited in
Figure 25. Since the relationship is a linear one and deviations from the average value are
equally likely in either direction (though lower values will probably be more common
due to over cutting), the fluctuations within a day would most likely balance out the daily
loads to those anticipated with the average sediment concentration.
Dissolved PCB Fraction in the Background
The dissolved PCB fraction in the background, the background suspended solids
concentration and the partition coefficient are interrelated by the following equation:
CD 1
BKG = fBKG = 1 n (EQ 67)
CTbkg (l + KdxTSSBKGx\0-6)
Hudson River PCB s Superfund Site
Engineering Performance Standards
51
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Therefore in order to vary the dissolved fraction in the background the partition
coefficient was held constant at the literature value of 5,500 L/kg and the suspended
solids concentration in the background was varied from 0.5 to 40 mg/L. These values
determined background PCB dissolved fraction between 0.31 and 0.97.
The results for the various PCB dissolved fractions are shown in Figure 26. Neither of the
net solid outputs (distance to 0.1 percent net coarse and net suspended solids flux) are
dependent on the background PCB dissolved fraction or the suspended solids
concentration. The net dissolved fraction increases with an increasing background
fraction since a higher background fraction will limit the partitioning and therefore the
particles that settle will have a higher concentration. By the time the solids have reached
one mile so many solids with higher concentrations have settled out of the water column
that the conditions have moved further away from equilibrium. Therefore the dissolved
concentration and net dissolved fraction at one mile increases with an increasing
dissolved background fraction. However, by removing more concentrated solids through
settling, the overall PCB concentration (and thereby the flux) decreases.
Partition Coefficient
As noted above, the partition coefficient, dissolved PCB fraction in the background and
the background suspended solids concentration are interrelated. In order to test the model
3 5
sensitivity to the partition coefficient, the coefficient was varied from 5x10 to 5x10 and
the suspended solids background concentration was held constant (therefore the dissolved
PCB fraction in the background varied from 0.99 to 0.47). This range was used since it is
not uncommon to find partition coefficients given as log values, and therefore likely to
vary by an order of magnitude. The default value is given by the measured dissolved PCB
fractions and suspended solids concentrations in the background.
As is shown in Figure 27 neither the distance at which the coarse material becomes less
than 0.1% of the net, nor the net suspended solids flux is effected by the varying partition
coefficient (and background PCB dissolved fraction). It should be noted that a log scale is
used in Figure 27 for the partition coefficient. The net dissolved fraction is highly
sensitive to the partition coefficient since it indicates the equilibrium fractions. However,
the net PCB flux is not highly sensitive to the magnitude changes in the partition
coefficient, since most of the total PCB concentration is dominated by the suspended
concentration and the suspended solids concentration is not being affected. Given that
most of the criteria are determined by the total PCB value and the confidence in the
default partition coefficient is fairly high, variations in the partition coefficient are not
expected to limit the usefulness of the TSS-Chem model.
Desorption Rate
The range of desorption rates was obtained through a literature search which is described
in attachment C in this attachment. The default value was set at the maximum of the
range since this is a conservative assumption and will allow the partitioning to approach
Hudson River PCB s Superfund Site
Engineering Performance Standards
52
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
equilibrium conditions more quickly. The results for the various desorption rates are
shown in Figure 28. As with many of the other parameters there is no effect on the two
solids outputs.
The net dissolved fraction increases with increasing desorption rate since the system
approaches equilibrium conditions more quickly. The net PCB flux increases with
increasing desorption rate since both the dissolved concentration is increasing and the
concentration on the settled solids is decreasing.
Lateral Dispersion
The range and default value of the lateral dispersion coefficient was obtained through a
literature search, which is described Section 4.4.1 in this attachment. The results for the
various coefficients are shown in Figure 29. It should be noted that a log scale is used in
Figure 29.
With an increase in lateral dispersion the net dissolved fraction increases since the ratio
of the volume of water to the solids becomes larger. The slope of the increase in the net
dissolved fraction decreases as the solids begin to disperse so quickly that the width of
the plume becomes the width of the river well before it is a mile downstream. The net
PCB flux increases due to the increase in dissolved PCBs and decrease in the PCB
content of settled solids. As is shown in Figure 29, the net PCB flux is less sensitive than
the net dissolved fraction to changes in the lateral dispersion coefficient, due to the
significance of the suspended PCB concentrations.
PCB Background Concentration
The range of background PCB water column concentrations is based on the variations
experienced throughout the years. The default value is based on the average background
value for June to November. The results for the various PCB Background concentrations
are shown in Figure 30.
The PCB background concentration has a linearly indirect effect on both the net
dissolved fraction and the net PCB flux. The high PCB background values introduce
more dissolved PCBs into the system and limit the partitioning of the solids in the water
column. Therefore there is a decrease in the net dissolved PCBs and the net fraction
decreases. Similarly, the net total PCB flux decreases due to low dissolved
concentrations, and high PCB concentrations on settled particles.
Settling Velocity of Silts
The range and default value of the settling velocity of silts was obtained through a
literature search, which is described in Section 4.4.1 in this attachment. The results for
the various coefficients are shown in Figure 31.
Hudson River PCBs Superfund Site
Engineering Performance Standards
53
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
The settling velocity of the silt determines the residence time of silty solids in the water
column, thereby affecting the time available for partitioning. As the silt settling velocity
increases, the net dissolved concentration will decrease. However, the suspended PCB
concentration is also decreasing as particles settle more quickly with higher
concentrations. As shown in Figure 31, the decrease in the net dissolved concentration is
smaller than the decrease in the net total PCB concentration and the net fraction thereby
increases. The decrease in the total PCB concentration and flux is a result of less
partitioning and therefore lower dissolved PCB concentrations and greater PCB
concentrations on settled particles.
The settling velocity of the silt also affects the two solid outputs, by determining how
long the silty solids will remain in the water column. Since the silt settling velocity is
much greater than the coarse settling velocity and the distance at which the coarse
fraction becomes 0.1 percent is limited by the incremental nature of the model (the value
is only given to the nearest meter), the effect of increasing the silt settling velocity is
negligible and not exhibited in Figure 31. The net suspended solids flux decreases with
increasing settling velocities since the silt particles are settling from the water column at a
faster rate.
Settling Velocity of Coarse Particles (Sand)
The range and default value of the settling velocity of sand was obtained through a
literature search, which is described in Section 4.4.1 in this attachment. The results for
the various coefficients at one mile are shown in Figure 32.
The distance at which the coarse material is less than 0.1 percent decreases as the coarse
particles settle more quickly. The settling velocity of the coarse particles does not have a
significant effect on the net dissolved PCB fraction, net PCB flux, or net suspended solids
flux at one mile, since the coarse material settles out of the water column within 60
meters. Therefore the contributions of the coarse materials at one mile, to both PCB
partitioning and solids presence are minimal.
Hudson River PCB s Superfund Site
Engineering Performance Standards
54
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
5.0 Far-Field Modeling
5.1 HUDTOX and FISRAND: Fate, Transport, and Bioaccumulation
Modeling to Simulate the Effect of the Remedial Alternative
HUDTOX models suspended sediment and PCB transport from Fort Edward through the
Thompson Island Pool and downstream to the Federal Dam at Troy, New York.
HUDTOX consists of a 2-dimensional vertically-averaged hydrodynamic mathematical
model (the US ACE RMA-2V model) and a 2-dimensional water quality model with
sediment resuspension and scour submodels.
The RMA-2V half of the model simulates water movement by applying conservation of
mass and momentum to a finite element mesh overlaying the water surface. It computes
water depth and the depth-averaged velocity, both magnitude and direction, in each cell
under a specific set of conditions. The finite element mesh used for the Thompson Island
Pool consisted of about 6,000 cells connected at approximately 3,000 nodes. Nodes were
spaced about 92 m apart in the downstream direction and 15 m apart laterally (see Figure
3-2 from Revised Baseline Modeling Report (BMR) (USEPA, 2000b). RMA-V2 was
calibrated by adjusting Manning's n (flow resistance) values to match available water
level and velocity data for steady flow conditions at 30,000 cfs. This flow represents the
highest values associated with both the upstream and downstream rating curves. The
model was validated using data from a 29,800 cfs event that occurred in April 1993.
HUDTOX's submodel is used to estimate sediment deposition and erosion based upon
the results of the hydrodynamic model. Variations in bottom velocities within Thompson
Island Pool and bottom sediment characteristics - both laterally and vertically - dictated
careful consideration of sediment dynamics to accurately estimate changes in water
column concentrations due to bottom sediments scour or suspended sediment deposition.
PCB concentrations in some areas of the river are higher at depth than at the surface.
Thus the exposure of these buried deposits is of particular concern. The Depth of Scour
Model (DOSM) with a 2 cm vertical discretization was used to assess bottom sediment
dynamics and changes in bottom sediment PCB concentrations due to river flows.
Fate and transport modeling within HUDTOX is based upon EPA's WASP4/TOXI4
models. One-dimensional, transient water quality models considering advection,
diffusion, external loadings (e.g., sediment releases) and transformation (e.g., settling)
were applied to both suspended solids and PCBs assuming vertical (z-domain) and lateral
(y-domain) homogeneity. A finite difference solution was used to predict average water
column concentrations in adjoining segments over time. The finite-difference derivation
of the general WASP mass balance equations and the specific solution technique
implemented to solve these equations are described in Ambrose et al. (1993).
Details on all components of the HUDTOX model along with calibration and validation
procedures can be found in the Revised Baseline Modeling Report (USEPA, 2000b).
Hudson River PCBs Superfund Site
Engineering Performance Standards
55
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
To examine the PCB transport and fish body burdens of PCB, fate, transport, and
bioaccumulation models were used. The FISHRAND model requires surface sediment
and dissolved water Tri+ PCB concentrations corresponding to the three river sections as
described in the FS. FISHRAND is a time-varying mechanistic model based on the
modeling approach presented in Gobas (1993 and 1995). The model relies on solutions of
differential equations to describe the uptake of PCBs over time, and incorporates both
sediment and water sources to predict the uptake of PCBs based on prey consumption and
food web dynamics.
5.1.1 HUDTOX Input Values
The resuspension performance standard consists of a Resuspension Standard threshold
and action levels. This action level covers operations in the immediate vicinity of
dredging operations (near-field) and at the main fixed monitoring locations (far-field) so
that water quality responses to the dredge operation, site conditions, engineering controls
and other management efforts can be quickly identified. The action levels include both
mass and concentration criteria, and apply to suspended solids and Total PCBs. The
action levels for Total PCBs are:
Load Criterion of Evaluation Level
Load Criterion of Control Level
Concentration Criterion of Control Level
The net increase in Total PCB mass
transport due to dredging-related activities at
any downstream far-field monitoring station
exceeds 300 g/day.
The net increase in Total PCB mass
transport due to dredging-related activities at
any downstream far-field monitoring station
exceeds 600 g/day.
The total PCB concentration at any
downstream far-field monitoring station
exceeds 350 ng/L.
Because of the different scale of resuspension (near-field vs. far-field), the following
terms have been defined in the preliminary draft of the resuspension performance
standard:
Resuspension production rate. Dredging-related disturbances suspend PCB-bearing
sediments in the water column. The rate at which this occurs is the resuspension
production rate.
Resuspension release rate. Since most of the sediments to be remediated in the Upper
Hudson are fine sands, a significant fraction and often the majority of this material
will settle in the immediate vicinity of the dredge. Materials that remain in the water
Hudson River PCBs Superfund Site
Engineering Performance Standards
56
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
column are then transported away by river currents. The rate of sediment transport
from the immediate vicinity of the dredge is defined as the resuspension release rate.
Resuspension export rate. Beyond roughly 1,000 yards, further PCB removal from
the water column by particle settling becomes small and most of the PCB in the water
column is likely to travel long distances before it is removed or captured by natural
geochemical processes. The rate at which PCBs are transported beyond 1,000 yards is
defined as the resuspension export rate.
The Evaluation Level and the load criterion of the Control Level specify the Total PCB
load at the far-field monitoring stations and the concentration criterion of the Control
Level specifies the Total PCB concentration at the far-field monitoring stations. These
resuspension criteria are the targeted export rates. During dredging operations, it is
necessary to specify the load to the water column in the near-field that yields the targeted
export rate at the far-field stations. However, there is no prior knowledge of the
relationship among the resuspension production, release and export rates. For this reason,
computer models will be utilized to estimate the relationship between the far-field and the
near-field dredging-induced PCB transport and loss. These computer models are CSTR-
Chem, a Gaussian plume model with its associated geochemical component (TSS-Chem),
and HUDTOX. The three models will be used to represent and link the three different
scales of resuspension. The resuspension production rate in the immediate vicinity of the
dredge (30 m) is simulated by the CSTR-Chem. The resuspension release rate in the
region from the dredge to a distance of one mile (30 to 1600 m) is represented by TSS-
Chem model. Finally, the resuspension export rate in the region beyond one mile is
represented by HUDTOX. The choice of the TSS-Chem model to represent a one-mile
interval is related to the size of the individual HUDTOX cell, which is approximately 2/3
of a mile long. In addition to the fate and transport models, a series of model simulations
is also needed to assess the impacts of dredging to the fish tissue concentrations in the
Upper and Lower River. For this purpose, FISHRAND will be used to predict the fish
trajectory in the Upper and Lower River and the Farley model will be used to predict the
water column and sediment concentrations in the Lower River.
This series of computer models was used to simulate all action levels at the far-field
monitoring stations. For the purpose of the modeling effort, all the far-field monitoring
for River Section 1 will be done at Thompson Island Dam (TID) and all monitoring for
River Sections 2 and 3 will be done at Schuylerville and Waterford, respectively. The
one-mile exclusion for the monitoring purposes as stated in the performance standard is
not considered in the model runs.
Since the Total PCB action levels are specified as the export rate, HUDTOX is expected
to simulate the upper river dredging conditions that caused the conditions at the far-field
monitoring stations as specified in the action levels {i.e., 300 g/day, 600 g/day and 350
ng/L). Due to the inherent nature of the HUDTOX model structures, PCB loads cannot be
readily specified at far-field locations. Rather, the input of PCBs is specified as an input
load at a location within the river, equivalent to a resuspension release rate. For the initial
supporting model runs completed for the performance standard, the resuspension release
Hudson River PCBs Superfund Site
Engineering Performance Standards
57
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
rate was set equal to the desired export rate, recognizing that this yields export rates less
than the desired test value. In order to create a correctly loaded HUDTOX run, it is first
necessary to estimate the resuspension release rate from the dredging operation, that is,
the rate of PCB and solids transport at the downstream end of the dredge plume. At this
location most of the solids that are going to settle out, will have settled out and the
suspended solids will more closely resemble those simulated by HUDTOX. Therefore, to
estimate the input loading term for HUDTOX, the CSTR-Chem and TSS-Chem models
were used.
From the initial model runs, it was observed that the HUDTOX model yields an
approximately 25 percent reduction (75 percent throughput) of the resuspension release
rate to the export rate at the far-field monitoring stations. Therefore, based on these initial
runs, the input loading of the HUDTOX model was corrected.
The model formulations for each action level will be discussed in the next sections. The
Control Level Total PCB criterion of 350 ng/L will be discussed first since in the
preliminary draft of the performance standard at this level, engineering solutions were
mandatory and they were only suggested for the other two levels.
Control Level - 350 ng/L at the Far-Field Monitoring Stations
The Control Level of the performance standard specifies that the Total PCB
concentration at any downstream far-field monitoring station (compliance point) should
not exceed 350 ng/L. The 350 ng/L action level will include both mass flux and
concentration criteria, and apply to total suspended solids (suspended solids) and Total
PCBs.
To calculate the total flux based on the maximum concentration of 350 ng/L, the
following formula is used:
7, ^rr^nK 1000 Z 10 ~9 g
Ft = 350 xqx x
L m ng
where:
Ft = total Total PCB flux (g/sec)
350 ng/L = Maximum Total PCB concentration (ng/L)
"3
q = flow rate (m /sec)
1000 L/m = conversion factor from m3 to L
10"9 g/ng = conversion factor from ng to g
The 350 ng/L resuspension criterion includes ambient PCB loads as well as loads from all
sources upstream of the monitoring location. To obtain the load as a result of dredging
only, the ambient Total PCB loads (mean baseline loads) should be subtracted from the
total flux of Total PCB. Mean baseline load is calculated as follows:
Hudson River PCBs Superfund Site
Engineering Performance Standards
58
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
r ^ 1000 L 10~9 g
Fmb = CS^ xqx x
m ng
where:
Fmb = Mean baseline Total PCB flux (g/sec)
CSmb = Mean baseline Total PCB concentration (ng/L)
and other parameters as described above.
The mean baseline Total PCB concentrations were analyzed for TID and Schuylerville
based on the water column samplings collected by GE in their on-going weekly sampling
program. The methodology and results of the baseline concentrations analysis can be
found in Attachment A of the Resuspension Performance Standard. The mean baseline
Total PCB concentration for TID and Schuylerville stations can be found in Tables 17
and 18, respectively. Due to limited data available for Waterford, the mean baseline
concentrations at this station were estimated by applying a dilution factor of 0.75 to the
Schuylerville data. The dilution factor was based on the drainage area ratio of
Schuylerville (3440 ft2) to that of the Waterford (4611 ft2). The drainage areas for
Schuylerville and Waterford were obtained from USGS. The mean baseline Total PCB
concentration for Waterford can be found in Table 19.
The net dredging export flux at the monitoring station is then:
Fnd ~ FT ~ Fmb
where:
Fnd = Net dredging Total PCB flux (g/sec)
and other parameters as described above
The net dredging flux in a day depends on the length of the production or the working
hours and is described as follows:
S6C
FNDdaily ~ Fnd X tw X 3600
where:
FNDdaUy = Daily net dredging Total PCB flux (g/day)
tw = production/working hours in one day (hr/day)
3600 sec/hr = conversion factor from seconds to hour
The daily net dredging Total PCB flux was calculated for all river sections using the
above equations for both 14-hour and 24-hour workdays. Table 20 summarizes the daily
net dredging flux for River Sections 1, 2, and 3. For the modeling purposes, a 14-hour
workday was used to be consistent with the productivity standard.
Dredging operations are scheduled from 2006 to 2011 with a dredging season from May
1 to November 30 each year, except for the last year of dredging which ends on August
Hudson River PCB s Superfund Site
Engineering Performance Standards
59
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
15, 2011. For the purpose of the modeling effort, May conditions are excluded in the
daily average of the net dredging Total PCB flux since flow conditions in May are not
representative of the remainder of the dredging season (i.e., May has high flow rates
compared to other months). The average is only from June to November. In the model
simulation, using this average Total PCB flux will also be protective for May conditions.
As mentioned above, the resuspension criterion of 350 ng/L is specified at the far-field
monitoring stations. This means the export rate at the monitoring stations should not
exceed 350 ng/L. In order to simulate the 350 ng/L Total PCB concentration at the far-
field monitoring stations, the Total PCB flux at the near-field location or station that
causes the 350 ng/L at the far-field monitoring station is needed. Once the Total PCB flux
that represents the 350 ng/L at the far-field monitoring station was obtained using the
above equations, the value was increased based on the fraction remaining of the
HUDTOX input to the Total PCB flux at the monitoring stations. For the first attempt, a
75 percent fraction remaining at the monitoring station was used based on the previous
HUDTOX model runs (Table 21). The input to HUDTOX is calculated by applying the
average daily flux for the specific river section for the whole dredging period (May to
November) divided by the fraction remaining at the monitoring stations and is described
as follows:
F
77 NDave
NDinput ^
where
FNDinput = Daily net dredging Total PCB flux input to HUDTOX (g/day)
FNDave = June to November average of daily net dredging Total PCB
flux (g/day)
y = fraction remaining at the far-field monitoring station (%)
Table 21 summarizes the Total PCB flux input to the HUDTOX segments. For the first
year of dredging, the resuspension release is applied to June 1 to September 15, 2006
only to account for the half-speed production during that period.
In order to conduct forecast simulations with the HUDTOX model, it was necessary to
specify suspended solids and Tri+ PCB flux instead of Total PCB flux. To obtain the
Tri+ PCB flux, the Total PCB values were divided by the sediment Total to Tri+ PCB
ratio estimated in the Responsiveness Summary to the Record of Decision (USEPA,
2002). The ratio of Total to Tri+ PCB in the sediment for River Section 1 is 3.2, River
Section 2 is 3.4 and River Section 3 is 2.7 (USEPA, 2002).
There is no existing data on how to load the suspended solids flux associated with the
Total PCB flux for the HUDTOX input. One way to obtain the suspended solids flux is to
assume instantaneous equilibrium for PCBs in the water column and use the sediment
PCB concentrations in each section of the river to come up with the suspended solids flux
(Table 22). However, in dredging scenario, the residence time (contact time) of the
Hudson River PCBs Superfund Site
Engineering Performance Standards
60
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
sediment in the water column is relatively short, on the order of hours. For this period of
time, it is unlikely that PCB reaches equilibrium. Therefore, the suspended solids flux
was estimated using TSS-Chem model that accounts for the non-equilibrium partitioning
for the desorption of the Total PCBs. The suspended solids flux one mile downstream of
the dredge-head was first chosen based on the size of the HUDTOX cells. The suspended
solids flux at one mile downstream of the dredge-head was about 3 to 6 percent lower
than that of the full equilibrium scenario, depending on the river section (Table 22).
From the Total PCB concentrations downstream of dredge-head plot, it was shown that at
three miles downstream, both particulate and dissolved Total PCBs are closer to the
equilibrium conditions (Figure 33). Since the HUDTOX far-field model assumes
equilibrium partitioning of PCBs, the second attempt of simulating the 350 ng/L
resuspension criterion is to take the suspended solids flux from TSS-Chem at three miles
downstream of the dredge-head. The suspended solids flux values are slightly smaller
than those at the one-mile downstream location (Table 22). To bound the model estimate,
a scenario of 350 ng/L without suspended solids flux added to the model was also
simulated.
Based on initial HUDTOX runs, the fraction of PCBs remaining at the monitoring station
differs by reach of the river, and the fraction remaining is higher closer to the monitoring
stations (Table 23). Discussions on the HUDTOX results for the first attempt of 350 ng/L
can be found in the Section 5.1.4 of this attachment. Based on the first attempt results, the
fraction remaining at the monitoring station was adjusted accordingly (Table 23). The
final 350 ng/L scenario was simulated based on the corrected fraction remaining of total
PCBs at the monitoring stations and the suspended solids flux at one mile downstream of
the dredge-head. The input to the HUDTOX model for the 350 ng/L can be found in
Table 23.
Evaluation Level - 300 g/day Total PCB Flux Export Rate
In Evaluation Level, the Total PCB flux at the downstream monitoring stations should
not exceed 300 g/day. To examine the effect of running the dredging operation at this
action level for the entire dredging period, the Total PCB flux at the downstream
monitoring stations was set to be 300 g/day. The input loading for the HUDTOX model
was then calculated using the corrected fraction remaining at the monitoring stations. The
suspended solids flux associated with the Total PCB flux was calculated using the TSS-
Chem model at one mile downstream of the dredge-head. The schedule and the input
functions of the 300 g/day resuspension criterion can be found in Table 24.
Control Level - 600g/day Total PCB Flux Export Rate.
Similar to Evaluation Level , the load criterion of the Control Level specified that the
Total PCB flux at the downstream monitoring stations should not exceed 600 g/day.
Therefore, to study the effect of running the dredging operation at 600 g/day for the entire
dredging period, the Total PCB flux at the downstream monitoring stations was set at 600
g/day. Just like the Evaluation Level scenario, the 600 g/day scenario was based on the
Hudson River PCBs Superfund Site
Engineering Performance Standards
61
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
corrected fraction remaining at the monitoring stations with suspended solids flux at 1
mile downstream of the dredge-head obtained from TSS-Chem. Table 25 summarizes the
schedule and input functions of the 600 g/day resuspension criterion.
Accidental Release Scenario
HUDTOX was used to model an accidental release scenario. The purpose of modeling
this scenario is to demonstrate the short-term and long-term impact to the public water
intakes. The following accidental release scenarios were proposed:
1. A hopper barge containing 870 tons of silty sand (barge capacity is 1000 tons,
with 87 percent sediment and 13 percent water) from River Section 2 is
damaged and releases the entire load in the area just above Lock 1. The
contents fall in a mound and no effort is made to remove or contain the
material. Over a period of one week, the entire load is swept downstream. The
sediment had been removed by mechanical dredging. The background
concentrations are at the 600 g/day Total PCB flux at the River Section 3
monitoring location. For this scenario, there will be additional release of
113,000 kg/day suspended solids, with a baseline condition of 20,000 kg/day
for a one week period from July 1 through 7, 2011.
2. A hydraulic pipe bursts. The dimension is 3-mile long and 16 inch diameter.
The pipe consists of 20 percent solids (USEPA, 2002; Herbich and Brahme,
1991). For this scenario, the additional suspended solids flux will be
approximately 125,000 kg/day for a one-day period.
Case 1 is more severe than case 2. In addition, the case 1 scenario is quite conservative in
that the average concentration from River Section 2 is higher than in the TI Pool because
areas with mass per unit area greater than 10 g/m are targeted whereas, in the TI Pool,
areas greater than 3 g/m2 are targeted. The hopper barge was used because it has a larger
capacity than the deck barge (200 tons), which was also proposed in the FS. The location
of the accident is just above the public water intakes at Halfmoon and Waterford,
minimizing any reductions that may occur in the water column concentration resulting
from settling and dilution. Because the sediment was removed by a mechanical dredge,
the entire weight is attributed to sediment with no dilution with water. The already
elevated water column concentrations result in water column concentrations at the public
water intakes greater than the MCL.
Hudson River PCB s Superfund Site
Engineering Performance Standards
62
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
5.1.2 Methodology
The resuspension criteria are defined as Resuspension Standard threshold and action
levels. The standard threshold is the maximum total PCB concentration of 500 ng/L at the
far-field monitoring stations and represents the acute component of the criteria. The
secondary action levels represent a chronic component (i.e., control of long-term impacts
to fish and related receptors). For the chronic component, a modeling effort was
performed to define a basis for a Total PCB flux standard in terms of Total PCB mass
export per year as well as a total mass exported due to dredging for the entire remedial
period.
Long term impacts of dredging focus largely on annual rates of PCB transport and
changes in fish body burdens of PCBs. For an unacceptable rate of release of resuspended
sediments the model would forecast impacts that deviate from the selected alternative.
That is, fish at downstream locations exhibit a slower recovery as a result of PCB
resuspension losses relative to the original no-resuspension scenario.
To study the long-term impacts of dredging, far-field modeling was completed to
simulate water column, sediment and fish Tri+ PCB concentrations in the Upper and
Lower Hudson River. The modeling efforts were focused on examining the impact of
running the dredging operation at the specified action levels in the resuspension
performance standard. The water column, sediment and fish total PCB concentrations
were forecast using USEPA's coupled, quantitative models for PCB fate, transport and
bioaccumulation in the Upper Hudson River, called HUDTOX and FISHRAND, which
were developed for the Reassessment RI/FS. HUDTOX was developed to simulate PCB
transport and fate for 40 miles of the Upper Hudson River from Fort Edward to Troy,
New York. HUDTOX is a fate and transport model, which is based on the principle of
conservation of mass. The fate and transport model simulates PCBs in the water column
and sediment bed, but not in fish. For the prediction of the future fish PCB body burdens,
the FISHRAND model will be used. FISHRAND is a mechanistic time-varying model
incorporating probability distributions and based on a Gobas approach and it predicts
probability distributions of expected concentrations in fish based on mechanistic mass-
balance principles, an understanding of PCB uptake and elimination, and information on
the feeding preferences of the fish species of interest. Detailed descriptions of HUDTOX
and FISHRAND models can be found in the Revised Baseline Modeling Report
(USEPA, 2000b).
For the Lower Hudson River, the Farley et al. (1999) fate and transport model was used.
The water and sediment concentrations from the Farley fate and transport model are used
as input for FISHRAND to generate the PCB body burdens for fish species examined in
the Lower Hudson.
Hudson River PCBs Superfund Site
Engineering Performance Standards
63
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
5.1.3 HUDTOX Input Study and Relationship Between Resuspension Release and
Export Rates
HUDTOX Total PCB and Suspended Solids Flux Input Study
As part of the long term impacts study, a measure of fish tissue recovery that can provide
a threshold or limit to define an unacceptable impact due to dredging releases and thereby
a limit on the export rate needs to be determined. The lower bound will be the ideal
conditions of dredging, where there is no sediments being spilled (no resuspension) and
the upper bound will be the MNA scenario. The HUDTOX/FISHRAND model runs that
exist cannot be used for this purpose strictly since HUDTOX is not designed to simulate
the process of dredging releases. From the previous HUDTOX model runs for the RI/FS
and the Responsiveness Summary of the FS, the model runs appear to be correctly
executed but it is clear from the HUDTOX's handling of the solids that the application of
the model is not entirely correct. Essentially HUDTOX is exporting too many suspended
solids from dredging operation. This happens because the boundary conditions
formulations were not done properly. Therefore, the specification of dredging releases to
HUDTOX needs to be refined.
During dredging operations, it is necessary to specify the load to the water column in the
near-field that yields the targeted export rate at the far-field stations. However, there is no
prior knowledge of the relationship between the near-field load and export rates at the
far-field stations. Due to the inherent nature of the HUDTOX model structure, PCB loads
cannot be readily specified at far-field locations {i.e., specifying the resuspension export
rate). Rather, the input of PCBs is specified as an input load at a location within the river,
equivalent to a resuspension release rate. In order to create a correctly loaded HUDTOX
run, it is first necessary to estimate the local export rate from the dredging operation, that
is, the rate of Total PCB and solids transport at the downstream end of the dredge plume.
At this location most of the solids that are going to settle out, will have settled out and the
suspended solids will more closely resemble those simulated by HUDTOX.
Unfortunately, there is no prior knowledge on the relationship between the resuspension
release and export rates. In addition to the lack of knowledge on the relationship between
the resuspension release and export rates, there is no existing data on how to load the
suspended solids flux associated with the Total PCB flux for the HUDTOX input. To
estimate the suspended solids flux input loading term for HUDTOX, the TSS-Chem
model was used. The total PCB input loading term for HUDTOX (the resuspension
release rate) was derived iteratively. The resuspension release rate was obtained by
checking the resuspension export rate (output from HUDTOX) until the model output
gives the desired total PCB export rate. Once the resuspension release rate that creates the
desired resuspension export rate was obtained, the corresponding suspended solids flux
associated with the total PCB release rate is estimated using TSS-Chem model. These
iterations also took into account the different river sections, with their corresponding
target sediment properties {i.e., silt fraction), PCB concentrations and hydrodynamics.
The simulations also accounted for the changes in dredging location as the remediation
progresses.
Hudson River PCBs Superfund Site
Engineering Performance Standards
64
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
To study the effect of different formulations of suspended solids flux input to the
HUDTOX model, the Control Level (350 ng/L at the far-field monitoring stations) was
modeled and examined in detail. The following scenarios were considered for the 350
ng/L export rate HUDTOX input:
Suspended solids and Total PCB flux at one mile downstream of the dredge-head
from TSS-Chem model (HUDTOX run number d006). The choice of the TSS-Chem
model to represent a one-mile interval is related to the size of the individual
HUDTOX cell, which is approximately 2/3 of a mile long.
Suspended solids and Total PCB flux at three miles downstream of the dredge-head
from TSS-Chem model (HUDTOX run number d007). This scenario was chosen
based on TSS-Chem model results where the Total PCB concentrations (both
particulate and dissolved phase) at 3 miles downstream of dredge-head are closer to
the equilibrium conditions (Figure 33). Since the HUDTOX model assumes
equilibrium partitioning of PCBs, the second attempt of simulating the 350 ng/L
resuspension criterion is to take the suspended solids flux from TSS-Chem at 3 miles
downstream of the dredge-head. The suspended solids flux values for the 3-mile
scenario are slightly lower than those of the 1-mile location (Table 10).
No suspended solids associated with Total PCB flux (HUDTOX run number sr03).
This scenario is essentially the pure dissolved phase Total PCB release during
dredging and was chosen to serve as an upper bound for the 350 ng/L simulation. The
model simulation for this scenario is carried out to the year 2020 only.
Suspended solids and Total PCB flux at one mile downstream of the dredge-head
from TSS-Chem model with a corrected of the fraction remaining at the far-field
monitoring stations (HUDTOX run number sr04). This scenario was simulated based
on the first three runs of the 350 ng/L (d006, d007, and sr03).
From the previous HUDTOX runs, it was estimated that there is an approximately 25
percent reduction (75 percent throughput) of the resuspension release rate to the export
rate. For the first attempt of simulating the export rate represented by the 350 ng/L, the
input to HUDTOX model was obtained by taking the suspended solids and Total PCB
flux at 1 mile downstream of the dredge-head from TSS-Chem model (d006). The
suspended solids and PCB flux input to the HUDTOX model segments can be found in
Table 20. The Tri+ PCB input flux was calculated based on the maximum Total PCB
concentration of 350 ng/L at the monitoring locations. Detailed calculations can be found
in the Sections 5.1.1 and 5.1.2 of this attachment.
The HUDTOX results are in the form of Tri+ PCB at the monitoring stations and they
are:
Tri+ PCB daily flux.
Integrated daily flow.
Suspended solids daily flux.
Hudson River PCBs Superfund Site
Engineering Performance Standards
65
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved phase Tri+ PCB daily flux.
The Tri+ PCB HUDTOX output includes both the ambient Tri+ PCB loads, as well as
loads from all sources upstream of the monitoring location, and the load resulted from
dredging operations. The baseline (background) Tri+ PCB flux can be obtained from the
no-resuspension scenario (d004) model run. Since the output of HUDTOX model is in
Tri+ PCB, conversions are needed to get the Total PCB concentrations. Baseline Tri+
PCB concentrations are on a 24-hour basis. The Total PCB baseline concentrations can
be calculated as follows:
Baseline Tri+
PCc-jFTn+nฐ-resusp;, 1 hour 1 day 1ft3 1 m3 1012 ng
q 3600 sec 24 hour 0.02832 m3 1000 L 1kg
where
Baseline Tri+ PCB = Tri+ PCB concentration in the water column (ng/L)
Fin+ no - resusp = HUDTOX Tri+ PCB flux output for no-resuspension
scenario (kg/day)
"3
q = Flow rate (ft /sec)
1 hour/3600 sec = Conversion factor from seconds to hours
1 day/24 hour = Conversion factor from hours to days
1 ft3/0.02832 m3 = Conversion factor from ft3 to m3
3 3
1 m /1000 L = Conversion factor from m to Liters
12
10 ng/l kg = Conversion factor from kg to ng
To estimate the Total PCB baseline concentrations, the ratios of Total PCB to Tri+ PCB
in the water column are used. The Total PCB to Tri+ PCB ratios in the water column are
presented in the Responsiveness Summary (RS) to the FS, Table 424694-1 (USEPA,
2002). Using the water column Total PCB to Tri+ PCB ratios, the Total PCB baseline
concentrations can be calculated as follows:
Baseline Total PCB = Baseline Tri+ PCB x water column ratio
Where:
Baseline total PCB = Total PCB concentration in the water column (ng/L)
water column ratio = Water column ratio of Total PCB to Tri+ PCB.
The value is
2 for River Sections 1 and 2;
1.4 for River Section 3
and other parameters as defined above.
The net addition of Tri+ PCB concentration due to dredging is based on the 14-hour work
period since the dredging operations are assumed to be 14 hours in one day, and it is
estimated as follows:
Hudson River PCB s Superfund Site
Engineering Performance Standards
66
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
NctTn l PCe-^"^;; 1 hฐU1" :: 1 day ::
q 3600 sec 14 hour 0.02832 m3 1000 L 1kg
where:
Net Tri+ PCB = Net additional Tri+ PCB concentration from the model
run output (ng/L)
A/"|ri = dredge scenario - no resuspension = Net Tri+
PCB flux output from dredging scenario (kg/day)
q = Flow rate (ft3/sec)
1 hour/3600 sec = Conversion factor from hours to seconds
1 day/14 hour = Conversion factor from hours to days, taking into
account 14-hour work period.
1 ft3/0.02832 m3 = Conversion factor from ft3 to m3
3 3
1 m /1000 L = Conversion factor from m to Liter
12
10 ng/l kg = Conversion factor from ng to kg
To calculate the net additional Total PCB in the water column due to dredging, the
sediment ratios of Total PCB to Tri+ PCB are used. The net addition of Total PCB due to
dredging is calculated using the following formulas:
Net Total PCB = Net Tri+ PCB x sediment ratio
Where:
Net total PC = Net additional Total PCB concentration in the water column (ng/L)
sediment ratio = Sediment ratio of Total PCB to Tri+ PCB.
The value is
3.2 for River Section 1;
3.4 for River Section 2;
2.7 for River Section 3;
and other parameters as defined above
The whole water Total PCB concentration is then:
Total PCB concentration = Baseline Total PCB + Net total PCB
Where:
Total PCB concentration = Whole water Total PCB concentration (ng/L)
and all other parameters as defined above.
From the first attempt of the 350 ng/L scenario (d006), it was found that the fraction
remaining at the monitoring station was different for different section of the river. The
fraction remaining is higher closer to the monitoring stations (Table 25). This happens
because in the model simulations, the monitoring station for all River Section 1 dredging
Hudson River PCB s Superfund Site
Engineering Performance Standards
67
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
was assumed to be at Thompson Island (TID). And all the monitoring for River Sections
2 and 3 dredging were assumed to be at Schuylerville and Waterford, respectively. The
one-mile monitoring exclusion from the dredging operations location was not considered
in the modeling effort. Therefore, as the dredging operations moved downstream (closer
to the monitoring location), the amount of Total PCB flux transported downstream were
getting higher. In other words, there is less settling taking place due to the distance from
the dredge-head to the monitoring station.
The model results showed that the HUDTOX model is not sensitive to the suspended
solids flux input. Three different suspended solids flux inputs were modeled (Table 26).
The suspended solids flux input for the 350 ng/L for the 3-mile downstream of the
dredge-head scenario is about 6 to 23 percent lower than that of the 1-mile scenario.
However, HUDTOX predicted that the Total PCB flux and concentrations at the far-field
monitoring stations are almost the same. Figure 34 shows the Total PCB concentration in
the water column for TID, Schuylerville, and Waterford, respectively for different 350
ng/L Total PCB concentration scenarios. The scenario with the suspended solids flux at
three miles downstream of the dredge-head resulted in a slightly lower Total PCB flux at
the monitoring stations than that of the 1-mile scenario. However, the difference is less
than 2 percent (Table 26). The upper bound estimate is the model scenario with pure
dissolved phase total PCB release (sr03). The model estimated a higher Total PCB flux
for this scenario. However, the difference is less than 15 percent.
The effect of different suspended solids flux input to the model can also be seen from the
predicted annual Tri+ PCB loads. The predicted annual Tri+ PCB loads over the TID,
Schuylerville, and Waterford for each of the HUDTOX forecast scenarios are shown in
Tables 28through 30. The annual loads for the 1- and 3-mile scenarios (d006 and d007)
are practically the same. The predicted Tri+ PCB cumulative loads for the no suspended
solids flux scenario (sr03) are higher compared to the 1- and 3-mile scenarios. However,
the predicted increase in loads is less than 3 percent. Figure 35 shows the predicted Tri+
PCB cumulative loads over the TID, Northumberland Dam, and Waterford, respectively.
Due to the model's insensitivity to the amount of suspended solids flux input and to be
consistent with the scale of the HUDTOX and TSS-Chem models, the 350 ng/L (sr04)
scenario was simulated based on the suspended solids flux at 1 mile of the dredge-head
and the fraction remaining at the far-field monitoring stations was adjusted based on the
1-mile (d006) model run results.
Similarly, the Total PCB load criterion for the Evaluation Level and Contorl Level were
simulated based on the 1-mile suspended solids flux and the fraction remaining at the far-
field monitoring stations was based on d006 run.
Relationship Among the Resuspension Production, Release, and Export Rates
As mentioned before, there is no prior knowledge of the relationship on the amount of
sediment being suspended to the water column to the suspended solids and PCB fluxes
downstream of the dredge-head. For this reason, computer models were utilized to
Hudson River PCB s Superfund Site
Engineering Performance Standards
68
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
estimate the relationship between the far-field and the near-field dredging-induced PCB
transport and loss. The TSS-Chem and HUDTOX models were used to represent and link
the resuspension production (at the dredge-head), release, and export rates. The
resuspension production rate is represented by the source strength of the TSS-Chem
model. The resuspension release rate in the region from the dredge to a distance of one
mile is represented by TSS-Chem model and the resuspension export rate in the region
beyond one mile is represented by HUDTOX.
The TSS-Chem and HUDTOX models were used to examine the amount of sediment
being suspended to the water colum at the dredge-head, the suspended solids and Total
PCB flux at one mile downstream of the dredge-head and the Total PCB flux at the far-
field monitoring stations for all three action levels. Table 31 shows the resuspension
production, release, and export rates for the simulated action levels. Because HUDTOX
predicted that the fraction remaining at the monitoring station was different for different
reach of the river, the TSS-Chem model was run to simulate the Total PCB flux at 1 mile
for each year of dredging. From the results it was predicted that to create an export rate of
300 g/day of Total PCB at the TID, the amount of sediments need to be suspended is
approximately 1 to 1.3 kg/s depending on the location of the dredge-head to the
monitoring stations. The farther away the dredge-head from the monitoring location, the
larger the amount of solids may be suspended to the water column (Table 31). In order to
get the same result, the resupension production rates that create an export rate of 300
g/day are on the order of 2 to 3 percent of the solids production rate, which is 42 kg/s. In
River Section 2, the solids production rate is lower than that of the River Section 1, with
a value of approximately 37 kg/s. For this river section, the amount of solids suspended
to the water column to create the 300 g/day Total PCB flux is approximately 0.3 kg/s,
which is on the order of one percent of the solids production rate. River Section 3 has the
lowest solids production rate, with a value approximately 31 kg/s. The resuspension
production rate that creates the 300 g/day of Total PCB flux is approximately 0.9 kg/s
when the dredge-head is farther away from the monitoring location and it is around 0.7
kg/s when the dredge-head moves downstream closer to the monitoring station.
For the Control Level load criterion (600 g/day Total PCB flux), the required amount of
solids suspended into the water column in River Section 1 ranges from 2 to 2.7 kg/s (on
the order 5 to 6 percent of the solids production rate). In River Section 2, to obtain an
export rate of 600 g/day, approximately 0.6 to 0.7 kg/s of solids need to be suspended to
the water column (approximately 2 percent of the solids production rate). For River
Section 3, approximately 1.4 to 1.9 kg/s of solids need to be suspended to the water
column to create an export rate of 600 g/day Total PCB flux (on the order of 2 percent).
Finally, the Control Level criterion of 350 ng/L Total PCBs was also simulated. The
Total PCB flux at the TID, Schuylerville, and Waterford that represents the 350 ng/L is
1200, 2000, and 2300 g/day, respectively. The resuspension production rates correspond
to the 350 ng/L Total PCB concentration at TID are approximately 4 to 5.6 kg/s, which is
approximately 10 to 13 percent of the solids production rate. For River Section 2, the
resuspension production rates are approximately 0.6 to 0.75 kg/s (approximately 6 to 7
percent of the solids production rate). In River Section 3, approximately 6 to 7.5 kg/s of
Hudson River PCBs Superfund Site
Engineering Performance Standards
69
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
solids need to be suspended to the water column to create an export rate of 350 ng/L
Total PCB concentrations. These resuspension production rates are approximately 19 to
24 percent of the solids production rate.
As for the resuspension release rates, under the 300 g/day (sr02) and 600 g/day (srOl)
scenarios, HUDTOX predicted that the values are approximately 1 to 1.3 times the
resuspension export rate (Table 31). For example, during the second year of dredging in
River Section 1 (2007), a 400 g/day Total PCB flux resuspension release creates an
export rate of 300 g/day. For the 350 ng/L scenario, HUDTOX predicted that the
resuspension release rates are approximately 1 to 1.4 times the resuspension export rates.
Example of CSTR-Chem, TSS-Chem and HUDTOX Application
As an example of the use of CSTR-Chem, TSS-Chem and HUDTOX to simulate the fate
and transport of PCBs during dredging operations, the development of the 350 ng/L (i.e.,
the Control Level) dredging scenario is discussed in this section. To simulate the Control
Level, the water column at the far-field monitoring stations was specified to be 350 ng/L.
The models were used in a backward sense, first determining the desired conditions to be
simulated (in this case 350 ng/L at the far-field stations) and then iterating through the
use of the models to determine the fluxes and dredging resuspension terms that would
yield the desired condition. The far-field monitoring stations for River Sections 1, 2, and
3 were assumed to be the Thompson Island Dam (TID), Schuylerville, and Waterford,
respectively. The PCB fate and transport model analysis was done in the following
sequence:
1. The expected Total PCB fluxes based on the 350 ng/L scenario at these three
monitoring stations are 1,200 g/day, 2,000 g/day, and 2,300 g/day, respectively
based on mean flow at these stations and the desired water column concentration
"3
(Table 31) . These are the resuspension export rates to be produced by HUDTOX
model when driven by input conditions derived from the near-field models.
HUDTOX input is the suspended solids and Total PCB flux at the upstream of the
far-field monitoring stations plus the resuspension loading terms derived from
TSS-Chem.
2. For HUDTOX to give the most reliable results, the Total PCB flux and the
corresponding suspended solids to the water column in the near-field need to be
determined. The Total PCB flux input was estimated based on previous HUDOX
runs. The near-field suspended solids load derived from the TSS-Chem model run
at the desired Total PCB output flux. Based on the previous HUDTOX runs, the
Total PCB flux at the near-field (i.e., the resuspension release rate) is
approximately 5 to 30 percent higher than the flux at the far-field monitoring
stations (i.e., the resuspension export rate), depending on the river section and the
dredging season (Table 31). For example, in River Section 1 during May 1 to
November 30, 2007 dredging season, the input Total PCB flux was predicted to
3 Note that the target loads and concentrations for HUDTOX were estimated for mean flow conditions and
the desired concentrations. The model was not run attempting to attain exactly 350 ng/L on each day of the
period of simulation. This approach is consistent with the long-term framework of HUDTOX, i.e., the
model was designed to address annual scales and longer.
Hudson River PCBs Superfund Site
Engineering Performance Standards
70
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
be approximately 27 percent higher than the output (Table 31). Therefore, for an
expected Total PCB flux of 1,200 g/day, the input Total PCB flux (i.e., the
resuspension release rate) has to be approximately 1,600 g/day. The 1,600 g/day
Total PCB flux is the value to be attained as the output of the TSS-Chem model.
The TSS-Chem output of 1,600 g/day was taken at approximately 1 mile
downstream of the dredge-head to be consistent with the size of the HUDTOX
model grid size. As mentioned above, the corresponding suspended solids load for
the 1,600 g/day Total PCB flux was obtained from TSS-Chem model.
3. Since the target for the TSS-Chem model is to produce as output the Total PCB
flux needed as input to HUDTOX, the TSS-Chem model was run iteratively to
determine the corresponding suspended solids and Total PCB input to TSS-Chem.
Once the suspended solids input rate to TSS-Chem yielded the desired Total PCB
flux (i.e., approximating the resuspension release rate), the flux of suspended
solids at 1 mile downstream of the dredge-head was taken as the suspended solids
load input to HUDTOX model. For example, in River Section 1 during the May 1
to November 30, 2007 dredging season, the corresponding suspended solids input
flux to TSS-Chem that creates the 1,600 g/day Total PCB output flux was
approximately 60,000 kg/day.
4. To determine the resuspension production rate at the dredge-head, the CSTR-
Chem model was used. The suspended solids input flux to the CSTR-Chem model
the resuspension production rate. The TSS-Chem suspended solids input flux is
the output of the CSTR-Chem model. Knowing the desired suspended solids
output flux for CSTR-Chem, the input to the CSTR-Chem was obtained
iteratively. For example, in River Section 1 during the May 1 to November 30,
2007 dredging season, the suspended solids input flux to the CSTR-Chem model
that creates a 60,000 kg/day suspended solids flux was approximately 280,000
kg/day.
5.1.4 HUDTOX Results
HUDTOX was used to simulate the following scenarios:
Control Level - 350 ng/L Total PCB concentrations at the monitoring stations
(HUDTOX run number sr04).
Contorl Level - 600 g/day Total PCB flux at the monitoring stations (HUDTOX run
number srOl).
Evaluation Level 1 - 300 g/day Total PCB flux at the monitoring stations (HUDTOX
run number sr02).
Accidental release (HUDTOX run number srAl).
The following sections summarize the results from the HUDTOX model simulations.
Control Level - 350 ng/L HUDTOX Simulation Results
The Total PCB concentration criterion of the Control Level specificies that the Total PCB
concentration at any downstream far-field monitoring station (compliance point) should
Hudson River PCB s Superfund Site
Engineering Performance Standards
71
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
not exceed 350 ng/L. The suspended solids and PCB flux input to the model can be found
in Sections 5.1.1 and 5.1.2 of this attachment. The 350 ng/L (sr04) scenario simulation
showed that the predicted Total PCB flux at the far-field monitoring stations is within 5
percent of the expected values (Table 20). The Tri+ PCB loads for this scenario are lower
than the previous two 350 ng/L model runs (d006 and d007). The HUDTOX model
predicted that the Tri+ PCB loads over the TID for the 350 ng/L scenario is lower than
the monitored natural attenuation (MNA) scenario by 2034 (Table 2827). The loads are
higher during dredging period (2006 to 2011) and 20 years beyond the completion of
dredging (Figure 35). However, by approximately 2033, the Tri+ PCB loads are the
same. Similarly, the amount of Tri+ PCB loads over the Schuylerville station is higher
than that of the MNA until approximately 2034 (Figure 35), where they become lower
than the MNA beyond that year. The Tri+ PCB loads over the Waterford (transported to
the Lower River) are predicted to be slightly higher than that of the MNA (Figure 35).
However, the predicted increase is minimal, less than 4 percent.
In terms of total PCB, the loads in the water column for the 350 ng/L scenario (sr04) are
predicted to be much higher than that of the MNA for all the monitoring stations (TID,
Schuylerville, and Waterford). The Total PCB loads over TID, Schuylerville, and
Waterford can be found in Figure 36. The Total PCB loads are higher because in order to
obtain the Total PCB loads for the MNA scenario, the multiplier is the water column ratio
of Total to Tri+ PCB while the multiplier for the 350 ng/L scenario is the ratio of the
Total to Tri+ PCB ratio for the sediment. The ratio or Total to Tri+ PCB for the sediment
is much higher than that of the water column ratio. Even though the Total PCB loads are
much higher, the impact to the fish tissue is expected to be minimal. Only Tri+ PCBs
include the PCB congeners that bioaccumulate in fish and hence are key to the risk
assessment (USEPA, 2000b).
Figure 37 shows the whole water, particulate, and dissolved Total PCB concentrations at
TID for the 350 ng/L (sr04) scenario during the dredging period (2006 through 2011).
The HUDTOX model predicted that the average whole water Total PCB concentrations
during dredging period in the first three years of River Section 1 is less than 350 ng/L. By
the end of the River Section 1 dredging, the whole water column Total PCB
concentrations are very low (Figure 37). The amount of dissolved phase Total PCB in the
water column is about 40 to 50 percent of the whole water total PCB. The amount of
particulate phase Total PCB increase in the reach closer to the monitoring stations
(Figure 37).
During River Section 2 dredging, the predicted Total PCB concentrations in the water
column are high. This is because the flow during that dredging period (August 16 to
November 30, 2009), on average is about 15 percent lower than the historical flow based
on the USGS data. Therefore, the high concentrations are expected. However, the average
concentrations during the whole dredging period for River Section 2 (August 16 to
November 30, 2009 and May 1 to August 15, 2010) is around 380 ng/L (Figure 37).
HUDTOX predicted that the amount of dissolved phase Total PCB during the first period
of River Section 2 dredging is about the same as the particulate phase (approximately 50
percent). During the next period of dredging (May 1 to August 15, 2010) the model
Hudson River PCBs Superfund Site
Engineering Performance Standards
72
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
predicted a slightly higher dissolved phase than the particulate phase Total PCB (Figure
37). This is probably due to the model prediction of flows that is low for that particular
year and section of the river.
In River Section 3, there are some high whole water Total PCB concentrations during the
last year of the dredging period. However, the average Total PCB concentration in the
water column during the whole dredging period is less than 350 ng/L (Figure 37). Again,
the amount of dissolved phase Total PCB is about the same as the particulate phase in the
dredging period of August 16 to November 30, 2010. The next period of the dredging
operations, the dissolved phase is less than the particulate because the location of the
dredging operations is closer to the monitoring station (Waterford) and hence there is less
settling.
Control Level - 600 g/day HUDTOX Simulation Results
The PCB load criterion of the Control Level, specifies that the Total PCB flux at any
downstream monitoring station should not exceed 600 g/day. To examine the effect of
running the dredging operation at this action level for the entire dredging period, the
Total PCB flux at the downstream monitoring stations was set to be 600 g/day. Based on
the first attempt of the 350 ng/L scenario and to be consistent with the scale of HUDTOX
and TSS-Chem models, the suspended solids flux for this model simulation was based on
the 1-mile TSS-Chem model results. The input suspended solids and PCB flux can be
found in Sections 5.1.1 and 5.1.2 of this attachment.
The HUDTOX model predicted that the Total PCB flux at the far-field monitoring
stations are within 10 percent of the expected Total PCB flux values (Table 33). The
whole water Total PCB concentrations at TID during the dredging period (2006 to 2011)
are predicted to be less than 250 ng/L except for few days in June 2008 (Figure 38). The
whole water Total PCB concentrations at the Schuylerville and Waterford monitoring
stations are predicted to be lower than 200 and 150 ng/L, respectively (Figure 38). For
this scenario, HUDTOX predicted a higher fraction of dissolved phase Total PCB in the
water column compared to the particulate phase total PCB. At TID, the amount of
dissolved phase is slightly higher than the particulate phase Total PCB during the first
and second year dredging period (May 1 to November 30, 2006 and May 1 to November
30, 2007). As the dredging operations moved downstream in the subsequent years (May 1
to November 30, 2008 and May 1 to August 15, 2009), the particulate phase Total PCB
increases and the amount of dissolved and particulate phase Total PCB are almost the
same (Figure 38). The fraction of dissolved phase in the water column is even higher in
River Section 2 (Schuylerville monitoring station). The amount of dissolved phase in the
water column is about 70 percent of the whole water Total PCB concentrations (Figure
38). The dissolved phase Total PCB in the water column at Waterford is approximately
50 percent of the whole water Total PCB concentrations (Figure 38).
The predicted annual Tri+ PCB loads over the TID, Schuylerville, and Waterford
monitoring stations for the 600 g/day (srOl) scenario are shown in Figure 39. The
predicted Tri+ PCB cumulative loads over TID and Schuylerville for 600 g/day scenario
are below the MNA by the year 2014 (Figure 39). The predicted Tri+ PCB cumulative
Hudson River PCB s Superfund Site
Engineering Performance Standards
73
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
load over Waterford is slightly above the MNA for another year, to approximately 2015.
Tables 28 through 30 summarize the predicted annual Tri+ PCB loads over TID,
Schuylerville, and Waterford. In terms of total PCB, the annual loads for the 600 g/day
(srOl) scenario stays higher than that of the MNA for a longer period of time (Figure 39).
Similar to the 350 ng/L scenario, this is due to the sediment ratios used in converting the
Tri+ PCB to total PCB.
The Total PCB Load criterion of the Control Level requires that the net increase in Total
PCB mass transport due to dredging-related activities at any downstream far-field
monitoring station cannot exceed 600 g/day. Look-up tables of PCB concentrations that
correspond to the 600 g/day Total PCB flux as a function of river flow and month are
provided in the resuspension performance standard. The concentrations that correspond
to the 600 g/day Total PCB flux in these look-up tables were calculated based on the GE
water column samples data at TID and Schuylerville. Since the concentrations were
calculated based on the historical data, the reduction of the baseline concentrations at the
subsequent section of the river due to the completion of the previous section of the river
was not accounted. The HUDTOX simulation for the 600 g/day takes into account the
reduction of the baseline concentrations in River Section 2 after dredging River Section
1. After completion of River Section 1 dredging, the baseline water column Total PCB
concentrations in River Section 2 are lower since the source upstream at the Thompson
Island Pool (TI Pool) has been removed. Control Level 1 as it is currently written
assumed the baseline of whole water Total PCB concentrations at Schuylerville as if the
TI Pool has not been dredged. In other words, the action level as specified in the
resuspension performance standard is too high. The mean baseline Total PCB
concentrations were analyzed for TID and Schuylerville based on the water column
samples collected by GE in their on-going weekly sampling program. The methodology
and results of the baseline concentrations analysis can be found in Attachment A of the
Resuspension Performance Standard.
To examine the additional loading that might be added due to this discrepancy, the
HUDTOX results for the 600 g/day are adjusted as follows. Assuming the baseline water
column monitoring will be performed from 2003 through 2005, the average monthly
Total PCB concentrations were estimated based on the MNA scenario results.
The difference of the average monthly Total PCB concentrations between the MNA and
the 600g/day (srOl) scenarios are calculated using the following formula:
ATPCB, = MNAbasC( - SrOlbase,
where:
ATPCB, = Average difference in Total PCB concentrations in
month i (ng/L).
MNAbase,. = Average baseline Total PCB concentration from MNA
scenario for month i (ng/L).
srO 1 base; = average baseline Total PCB concentration from 600
g/day (srOl) scenario for month i (ng/L)
Hudson River PCB s Superfund Site
Engineering Performance Standards
74
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
For River Section 2, the difference in Total PCB concentrations was calculated for
September through November 2009 and May through August 2010. Once the average
monthly difference in Total PCB was obtained, the Total PCB flux was calculated using
the following formula:
ATPCBflux, = ATPCB, x qms. x 0.02832 ft3/m3 x 3600 sec/hour x 14 hour/day x 1000
m3/L x 10"9 g/ng
where:
ATPCBflux, = Average difference in Total PCB flux for month i
(g/day).
t/avc, = Average flow rate for month i (ft3/sec).
0.02832 ft3/m3 = Conversion factor from ft3 to m3.
3600 sec/hour = Conversion factor from second to hour
14 hour/day = Conversion factor from hour to day
1000 m3/L = Conversion factor from m3 to liter
10"9 g/ng = Conversion factor from gram to nanogram
From the average Total PCB flux difference, the average Total PCB flux difference for
the whole dredging period (August 16 - November 30, 2009 and May 1 - August 15,
2010) in River Section 2 was calculated. May conditions are excluded in the average of
the difference in Total PCB flux since flow conditions in May are not representative of
the remainder of the dredging period. From the calculations above, the average difference
in Total PCB flux for River Section 2 is approximately 200 g/day. The 200 g/day Total
PCB flux was then added to the Total PCB flux of River Section 2 from HUDTOX
results (srOl).
Similarly, to account for the reduction in the baseline whole water column Total PCB
concentrations at Schuylerville during dredging River Section 3, the difference in Total
PCB flux was calculated using the above formulas. For River Section 3, the Total PCB
concentrations difference was calculated for September through November 2010 and
May through August 2011. The estimated Total PCB flux that needs to be added to the
Waterford Total PCB loads is approximately 300 g/day. During River Section 2 dredging,
the sediments from Schuylerville are being transported downstream to River Section 3.
HUDTOX predicted that 45 percent of the sediment from Schuylerville is transported to
River Section 3. Therefore, during River Section 2 dredging period, 45 percent of the
additional flux to the Schuylerville (95 g/day) will be transported to River Section 3.
Overall, the adjustment for Total PCB loads at Waterford is an additional 95 g/day Total
PCB flux from September through November 2009 and May through August 2010 and an
additional of 300 g/day Total PCB flux from September through November 2010 and
May through August 2011.
By adding this difference, the Total PCB loads over Schuylerville and Waterford stations
are predicted to increase by approximately 2 and 3 percent, respectively. However, the
70-year forecast Total PCB loads for this scenario are still lower than that of the MNA
Hudson River PCB s Superfund Site
Engineering Performance Standards
75
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
(Figure 39). The adjusted Tri+ PCB loads over Schuylerville and Waterford are also
plotted.
Evaluation Level - 300 g/day HUDTOX Simulation Results
Similar to the Control Level, the Evaluation Level specified that the Total PCB flux at the
downstream monitoring stations should not exceed 300 g/day. Therefore, to study the
effect of running the dredging operation at 300 g/day for the entire dredging period, the
Total PCB flux at the downstream monitoring stations was set at 300 g/day. The
suspended solids flux for this model simulation was based on the 1-mile TSS-Chem
model results. The input suspended solids and PCB flux can be found in Sections 5.1.1
and 5.1.2 of this attachment.
The HUDTOX model predicted that the Total PCB flux at the far-field monitoring
stations is within 13 percent of the expected Total PCB flux values of 300 g/day (Table
34). Figure 40 shows the whole water Total PCB concentrations in the water column at
TID, Schuylerville, and Waterford. The HUDTOX model predicted that by running the
dredging operations at the load criterion of the Control Level (total PCB flux of 300
g/day), the whole water column Total PCB concentrations at TID are less than 160 ng/L.
At Schuylerville and Waterford, the HUDTOX model predicted that the whole water
column concentrations are less than 120 and 80 ng/L, respectively (Figure 40). The
model predicted that the fraction of dissolved phase in the water column is approximately
60 to 70 percent depending on the location of the dredging operations relative to the
monitoring stations for River Sections 1 and 2 (Figure 40). At Waterford, the fraction of
dissolved phase Total PCB in the water column is estimated to be approximately 50
percent of the whole water column Total PCB (Figure 40).
Tables 28 through 30 summarize the predicted annual Tri+ PCB loads over the TID,
Schuylerville, and Waterford stations. HUDTOX predicted that the 300 g/day (sr02)
scenario has the lowest annual Tri+ PCB loads for all stations (Figure 41). Similar to the
600 g/day (srOl) scenario, the annual Total PCB loads for the 300 g/day (sr02) scenario
remain higher than that of the MNA for a longer period (Figure 41). Again, this is due to
the ratios of Tri+ PCB to Total PCB used in converting the Total PCB loads.
Similar to the Control Level, the 300 g/day Total PCB flux is the net increase in Total
PCB mass transport due to dredging-related activities. To be consistent with the
performance standard, in which it does not take into account the reduction of the mean
baseline Total PCB concentrations after completion of River Sections 1 and 2 dredging
operations, the Tri+ PCB and Total PCB loads for the 300 g/day Total PCB flux results
from HUDTOX need to be adjusted. Based on the 600 g/day Total PCB flux (srOl)
scenario results, the adjustment is expected to be small (on the order of 2 to 3 percent).
Hudson River PCB s Superfund Site
Engineering Performance Standards
76
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Comparison of the Water Column PCB Concentrations for Different Resuspension
Criteria
Figure 41 presents comparisons over 70-year forecast period of predicted HUDTOX Tri+
PCB concentrations in the water column at various locations throughout the Upper
Hudson River for the MNA, no resuspension and three action levels scenarios.
The effect of running the dredging operations at the Total PCB load criteria of the
Evaluation Level and Control Level on predicted water column Tri+ PCB concentrations
is largely confined to the six-year active dredging period (2006 through 2011). Outside of
the period of scheduled dredging, impacts on water column Tri+ PCB concentrations are
minimal. However, running the dredging operations at the PCB concentration criterion of
the Control Level results in significantly higher water-column concentrations during the
dredging period and slightly elevated water-column concentrations for approximately 10
years in River Section 3 (Figure 43).
The fraction of dissolved phase Total PCB in the water column is higher for dredging
scenario with lower suspended solids flux introduced to the water column (compare
Figures 37, 38, and 40). For example, the dissolved phase Total PCB for the 600 g/day
(srOl) scenario is higher than that of the 350 ng/L (sr04) dredging scenario. This is
because the amount of suspended solids flux to the water column for the 600 g/day
scenario is relatively lower than that of the 350 ng/L scenario. Compared to the 600 g/day
and 350 ng/L dredging scenarios Total PCB flux, the predicted Total PCB flux for the
300 g/day scenario is higher because the amount of solids introduced to the water column
is less than both 600 g/day and 350 ng/L scenarios. The smaller the amount of solids
introduced to the water column due to dredging, the higher the fraction of dissolved
phase Total PCB in the water column.
HUDTOX Results for Accidental Release Scenario
An accidental release scenario was simulated based on a hopper barge running aground
just above Lock 1 during dredging Section 3 of the river. The barge carried dredged
sediment from River Section 2. The accidental release scenario was assumed to happen
when dredging operations were operated under the Control Level criterion of 600 g/day
Total PCB flux. The Tri+ PCB loads over TID and Schuylerville remain the same as the
600 g/day (srOl) scenario (Figure 39). The Tri+ PCB load over Waterford was predicted
to increase due to the accidental release. The Tri+ PCB load increase is minimal, less
than 1 percent. Due to this small increase, the impact to the fish body burdens is expected
to be minimal and FISHRAND was not used to model the long-term impact of this
release to the fish concentrations.
HUDTOX provided the whole water, particulate bound, and dissolved phase PCB
concentrations in the water column. The model predicted that the accidental release
scenario results in a short-term increase of the whole water Total PCB above the MCL in
the water column at Waterford (Figure 42). However, the highest dissolved phase Total
PCB concentration was less than 350 ng/L (Figure 42). These concentrations can be
Hudson River PCB s Superfund Site
Engineering Performance Standards
77
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
examined against minimal treatment such as filtration and activated carbon to give an
indication if the public water supply will be adversely affected, even in the short term.
The impact of the elevated solids in the water column during the one-week period can be
examined versus the capacity of the Waterford treatment plant to cope with solids.
5.1.5 FISHRAND Results for the Upper and Lower River
FISHRAND model was used to simulate the dredging operations at the Control Level
only. FISHRAND modeling results for the Upper River show, similar to the HUDTOX
modeling, that the impact of running the dredging operations at load based criterion of
the Control Level to the fish tissue concentrations are largely confined to the dredging
period in River Sections 1 and 2 (Figure 43). In River Section 3, the impact to the fish
tissue concentrations lasts about three years beyond the dredging period to approximately
2014. Table 35 shows the years where FISHRAND model forecasted that the fish tissue
concentrations difference to the no resuspension dredging scenario is approximately 0.5
mg/kg. By 2009, the predicted fish tissue concentrations in River Section 1 are within 0.5
mg/kg of the no-resuspension scenario fish tissue concentrations. For River Section 2, the
fish tissue concentrations are within less than 0.5 mg/kg of the no-resuspension scenario
in 2008. The fish tissue concentrations difference in River Section 3 are predicted to be
always less than 0.5 mg/kg. The 0.5 mg/kg difference in fish tissue concentrations was
used because this number is within the measurement variability.
The impact of dredging operations at the Control Level criterion of 350 ng/L Total PCBs
is larger than running the dredging operations at the 600 g/day scenario (Figure 43).
Predicted fish tissue concentrations for the 350 ng/L scenario are within less than 0. 5
mg/kg to the no-resuspension scenario by approximately 2010 in River Section 1 (Table
37). The impact of the 350 ng/L scenario is slightly longer lasting in River Section 2
compared to that for River Section 1. The predicted fish tissue concentrations in
River Section 2 are greater than 0. 5 mg/kg of the no-resuspension scenario until
approximately 2010. However, in River Section 3, the predicted fish tissue concentration
under the 350 ng/L scenario is within 0.05 mg/kg of the no-resuspension scenario in
approximately 2011.
The Evaluation Level was not simulated since the Tri+ PCB loads to the Lower River are
lower than the load and concentration based criteria of the Control Level (Figure 32 and
Table 30). The results for the load based criterion of the Control Level show that the fish
tissue concentrations are only slightly impacted and there is only about four years delay
for the fish tissue concentrations to be the same as the no-resuspension scenario. In
addition, the annual average Tri+ PCB concentrations in the water column for the
Evaluation Level scenario are almost the same as that of the no-resuspension scenario by
the end of dredging period. Therefore, the Evaluation Level was not simulated and the
impact of running the dredging operations at this level is expected to have no adverse
impact.
Hudson River PCBs Superfund Site
Engineering Performance Standards
78
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
For the Lower Hudson River, the FISHRAND model predicted that the fish recovery is
slightly longer further downstream (Figure 44). Note that the fish tissue concentrations in
the Lower River are lower than those of the Upper River. The predicted fish tissue
concentrations for the 600 g/day (Control Level) scenario are within less than 0.05 mg/kg
relative to the no-resuspension scenario between 2013 and 2014 for all river miles
(Figure 44 and Table 36). As for the 350 ng/L (Control Level) scenario, the fish tissue
concentrations are within less than 0.05 mg/kg relative to the no-resuspension scenario
between 2016 and 2017 at RMs 152 and 113. Further downstream, at RMs 90 and 50, the
predicted fish tissue concentrations are within 0.05 mg/kg of the no-resuspension
scenario in 2018 (Table 36).
5.2 Relative Reduction In Human Health And Ecological Risks In
The Upper And Lower Hudson River
Human health hazards and risks and ecological risks in the Upper and Lower Hudson
River were calculated for the no resuspension, 350 ng/L Total PCB, 600 g/day Total
PCB, and monitored natural attenuation (MNA) scenarios. All active remediation
scenarios showed reductions in human and ecological risks, as compared to the MNA
scenario, with minimal differences generally seen between most active remediation
scenarios.
5.2.1 Introduction
PCB body burdens in fish under various resuspension scenarios were used to calculate
long term long-term risks (i.e., after completion of dredging) to anglers and ecological
receptors (as represented by the river otter [Lutra canadensis]). The following four
scenarios and their run designations (e.g., d004) were modeled:
? No resuspension (d004).
? 350 ng/L Total PCB (sr04).
? 600 g/day Total PCB (srOl).
? Monitored natural attenuation.
Risks were calculated with the same exposure durations used as those used the for the
Hudson River PCBs Reassessment RI/FS reports (e.g., 40 years for evaluating cancer
risks to the reasonably maximally exposed [RME] adult angler, 7 years for evaluating
non-cancer health hazards to the RME adult angler). Start years for calculating risks were
set to begin one year after the year in which dredging will be completed in the each
section of the river and the average of the upper river. All other risk assumptions,
locations, toxicity values, receptors, and fate, transport, and bioaccumulation models (i.e.,
HUDTOX, FISHRAND, and Farley) used to evaluate risks under various resuspension
scenarios are the same as those used for baseline conditions in the Revised Human Health
Hudson River PCBs Superfund Site
Engineering Performance Standards
79
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Risk Assessment, the Revised Baseline Ecological Risk Assessment, the Feasibility
Study, and the Responsiveness Summary for the Record of Decision, except where noted.
5.2.2 Human Health Risk Reduction
5.2.2.1 Upper Hudson River
Table 37 presents annual species-weighted fish fillet PCB concentrations in the Upper
Hudson River, as compared to the risk-based remediation goal (RG) for the protection of
human health of 0.05 mg/kg PCBs in fish fillet. The RG is based on non-cancer hazard
indices for the RME adult fish consumption rate of one half-pound meal per week, but
this level is protective of cancer risks as well. Other target concentrations are 0.2 mg/kg
PCBs in fish fillet, which is protective of human health at a fish consumption rate of one
half-pound meal per month and 0.4 mg/kg PCBs in fish fillet, which is protective of the
CT or average angler, who consumes one half-pound meal every two months.
FISHRAND, the model used to calculate fish body burdens, models fish tissue PCBs on a
Tri+ basis. PCB contamination in fish tissue has been shown to contain almost
exclusively Tri + PCB homologues (USEPA, 2002). Therefore EPA's fish forecasts and
modeling analyses, based on Tri+ PCB, require no revision for comparison to total PCB
toxicity values.
The time to reach human health fish target concentrations of 0.2 mg/kg Tri+ PCB and 0.4
mg/kg Tri+ PCB in the Upper Hudson River was shorter for all resuspension scenarios as
compared to monitored natural attenuation in the upper river as a whole, and in each
individual river section (Table 38). The remediation goal of 0.05 mg/kg Tri+ PCB was
only reached in Section 3. The greatest differences seen in the time to achieve fish target
concentrations between the active remediation scenarios and MNA were seen in River
Sections 1 and 2, where the MNA scenarios took up to 17 years longer to achieve some
target concentrations. Smaller differences were seen between scenarios in River Section
3.
Using fish fillet concentrations based upon the three resuspension scenarios (i.e., no
resuspension, 350 ng/L, and 600 g/day) human health fish consumption cancer risks and
noncancer hazards show at least a 50 percent reduction in the upper river as a whole,
Section 1 (River Mile 189), and Section 2 (River Mile 184) compared to monitored
natural attenuation for both RME and average exposures (Tables 39 and 40). Risk
reductions in Section 3 were seen for the no resuspension and 600 g/day scenarios as
compared to monitored natural attenuation, but not for the 350 ng/L Total PCB scenario.
Hudson River PCBs Superfund Site
Engineering Performance Standards
80
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
5.2.2.2 Mid-Hudson River
Based on site-specific angler surveys, the Human Health Risk Assessment determined
that Mid-Hudson River anglers have a different diet than anglers in the upper river,
consisting of 53 percent brown bullhead, 15 percent largemouth bass, 1.4 percent yellow
perch, 7.6 percent white perch, and 23 percent striped bass (USEPA, 2000). Striped bass
concentrations were modeled using the Farley model for the Hudson River RI/FS reports.
However, the Farley model was not run for fish tissue concentrations for resuspension
scenarios and therefore precise estimated of human health cancer risks and noncancer
hazards for Mid-Hudson River anglers could not be calculated.
To provide an estimate of relative risks amongst the resuspension scenarios, angler intake
was calculated using fish concentrations from the FISHRAND model. Striped bass intake
was proportionally divided between the remaining fish species (i.e., 69 percent brown
bullhead, 19 percent largemouth bass, 2.0 percent yellow perch, and 10 percent white
perch) and white perch concentrations from the FISHRAND model were used in the
absence of Farley model data. Calculated fish exposure concentrations were used only for
comparison between alternatives and do not represent predicted intake concentrations
based on mid-river angler consumption patterns. As expected, fewer differences were
seen between the resuspension scenarios in the lower river than in the upper river, with
long-term cancer risks and non-cancer hazards differing by a maximum of 32 percent.
The no resuspension and 600 g/day Total PCB scenarios showed the greatest risk
reductions as compared to monitored natural attenuation scenario. The 350 ng/L Total
PCB showed lower and sometimes no reductions in risk, owing to elevated
concentrations of PCBs predicted in fish tissues for several years following dredging
operations under the 350 ng/L scenario (Table 41).
5.2.3 Ecological Risk Reduction
5.2.3.1 Upper Hudson River
Risks to ecological receptors, as represented by the river otter, were evaluated by
examining largemouth bass whole fish PCB concentrations and comparing them to
toxicity reference value (TRV) based target levels using lowest-observed-adverse-effect-
level (LOAEL) and no- observed-adverse-effect-level (NOAEL) concentrations. In the
Upper Hudson River the LOAEL target levels were reached within the modeling
timeframe for the upper river as a whole and in Section 3 for all scenarios (Table 42). All
resuspension scenarios, reached the LOAEL target level of 0.3 PCBs mg/kg 17 years
prior to the MNA scenario for the upper river as a whole (Table 43). Ecological target
levels were not reached within the modeling timeframe for Sections 1 and 2 of the river.
In Section 3, all scenarios reached the LOAEL target level within five years of one
another.
Hudson River PCBs Superfund Site
Engineering Performance Standards
81
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
5.2.3.2 Lower Hudson River
Largemouth bass PCB concentrations in the Lower Hudson River were lower under all
resuspension scenarios than under the MNA scenario (Table 44). The LOAEL PCB target
concentration in largemouth bass was reached 4 to 11 years sooner under the various
resuspension scenarios than under MNA (Table 45).
5.2.4 Conclusions
Resuspension may temporarily increase PCB concentrations locally, resulting in slight
increases in fish PCB concentrations. However, human health noncancer hazards and
cancer risks and ecological risks under active remediation scenario were calculated to be
well below those under the monitored natural attenuation scenario. Minor differences
were seen between the various resuspension scenarios indicating the human health and
environmental impacts from dredging are predicted to be minimal, particularly since
levels of resuspension approaching the performance criteria are expected to occur on an
intermittent, rather than continuing basis. In general, human health and ecological target
concentrations are achieved within similar time frames under active remediation. Non-
cancer hazards, cancer risks, and ecological toxicity quotients showed minimal
differences between scenarios. Increased resuspension results in a maximum delay of five
years to achieve human health target concentrations under active remediation, as
compared to up to 17 year delays under monitored natural attenuation.
5.3 Suspended solids Far-Field Criteria
The far-field suspended solids criteria are based on the PCB far-field criteria. The
suspended solids concentration was calculated based on the PCB increase of the criteria,
assuming the solids concentrations were equal to the dredged material. For a total
concentration of 500 ng/L, and a background concentration of 100 ng/L, the net increase
would be 400 ng/L. As stated in the FS, the average PCB concentration on the dredged
sediment across all three River Sections is approximately 34 ppm. Therefore, the
suspended solids concentration for 500 ng/L was calculated to be about 12 mg/L.
Considering the uncertainty associated with some of the calculation assumptions, the TSS
criterion for Control Level was set at twice the estimated concentration or 24 mg/L, and
the TSS criterion for the Evaluation Level was set at 12 mg/L. Two-tiered far-field
suspended solids criteria, applicable to all the far-field stations, are established and
summarized below. It should be noted that the concentration of PCBs at the far-field
station with a suspended solids concentration of 12 mg/L is modeled by TSS-Chem to be
greater than 500 ng/L Total PCBs since the PCB dissolved phase would also contribute to
the concentration. The far-field suspended solids criteria are specified in Chapter 2 of
Volume 1.
Hudson River PCBs Superfund Site
Engineering Performance Standards
82
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
No standard was formulated for Resuspension Standard to avoid unnecessary shutdown
of operations. Exceedance of the far-field suspended solids criteria will not cause any
engineering contingency except for additional monitoring of PCBs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
83
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
6.0 Modeling Studies Used
6.1 New Bedford Harbor Pre-Design Field Test Dredge Technology
Evaluation Report
A numerical model of Upper New Bedford Harbor was used to predict concentrations of
suspended sediments in the water column resulting from dredging activities. The model
was based on previous hydrodynamic modeling of New Bedford Harbor performed by
the US Army Corps of Engineers (USACE 1998; USACE 2001). The computer models
RMA2 and SED2D were used to simulate hydrodynamics and sediment transport,
respectively.
Methods
Hydrodynamic Model (RMA2)
RMA2 is a two-dimensional depth averaged finite element model that simulates free
surface flow. The mesh size for this model ranged from 30 meters (98 feet) over most of
the domain (from Cogeshall Bridge at the south to Wood Bridge at the north) to 5 meters
(16 feet) in the vicinity of the dredging area (refer to Appendix K of the Pre-Design Field
Test Report, Figure K-3). This model, used at the New Bedford Harbor in 1988, was
calibrated to two sets of conditions: a spring high tide (March 1986), and a tide between
mean high tide and mean spring tide (April 1986). The model was rerun in 2000 to study
the potential impact of confined disposal facility construction on the hydrodynamics of
New Bedford Harbor. The predicted water surface elevation at the Cogeshall Bridge was
used to drive the new Upper New Bedford Harbor hydrodynamic model at the southern
boundary, while the same freshwater inflow used in the initial model was used at the
northern boundary.
Sediment Transport Model (SED2D)
The SED2D model was used to simulate sediment transport resulting from dredging
activities. The model calculates suspended sediment concentration and change in bed
elevation. For the application of the model to dredging it was assumed that the only
sediment source was due to dredging operations, and the bed surface was assumed to be
non-erodible due to waves, tidal currents, precipitation run-off etc.
Sediment source was defined as a constant input mass rate of sediment released in the
water column at four mesh elements. The resolution of the model mesh in the dredging
area is roughly 5 m (16 feet) square. The source was assumed to cover an area of four
mesh elements at any time, an area approximately equal to that of the dredge moon pool
(10 meters x 10 meters or 33 feet x 33 feet). The source strength was estimated from the
expected production rate of 69 m3/hr (90 yd3/hr), and the fraction of sediment lost to the
water column by the environmental bucket used (estimated 1 percent). Combining the
production rate and the percent lost, the total sediment release rate to the water column
was calculated to be about 482 kg/hr (1063 lb/hr).
Hudson River PCBs Superfund Site
Engineering Performance Standards
84
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
The sediments were assumed to be composed of three main sediment fractions which
were assumed to be non-cohesive with fall velocities calculated using Stokes' Equation,
as shown in Table 466. Since the SED2D model can only simulate one sediment type at a
time, each fraction was run independently, and the results were combined to obtain the
total suspended solids concentration.
Model Parameters and Variables
In the absence of field measurements to calibrate the present model, a series of
simulations were performed with dispersion coefficient values of 0.1, 1.0, 10 and 100
2 2
m /s (1, 11, 108, 1076 ft /s). It was confirmed that the dispersion coefficient had a major
impact on the extent of the suspended sediment plume and predicted concentrations.
Results
The model was run with a constant sediment source at the point of dredging for two tide
cycles, and the results for each sediment fraction were combined to predict the total
suspended sediment concentration throughout Upper New Bedford Harbor at half-hour
intervals. Modeled suspended sediment concentrations for flood tide and ebb tide are
shown in the Pre-Design Field Test Report, Figures K-4 and K-5, respectively. Figure K-
6 of the Pre-Design Field Test Report presents a time series of predicted suspended
sediment concentration at specified distances north and south of the dredge, along with
water surface elevations at the Cogeshall Street Bridge.
Numerous scenarios were considered with different combinations of dredge location
within the test area, mass release rate, and dispersion coefficients. Predicted local
suspended solids concentrations were greatest when the dredge was in the shallower
waters (at the eastern end of the dredge area). However, far-field suspended solids levels
were similar to those levels predicted to be present when dredging in deep waters. The
peak concentration predicted (immediately adjacent to the sediment release/dredge
location) decreased with increasing dispersion coefficients and varied from a maximum
2 2
of about 390 mg/L for dispersion coefficient of 0.1 m /s (1 ft /s), to less than 5 mg/L for a
coefficient of 100 m2/s (1076 ft2/s). The later value was within the variability of
background measurements; therefore it was difficult to detect above ambient conditions.
Table 47 presents the peak suspended sediment concentration predicted for different
dispersion coefficient values. In all cases, the results predicted no re-suspended sediment
transport under the Cogeshall Street Bridge to the Lower Harbor while the dredged
operation within the designated Pre-Design Field Test area.
Comparison of Predictive Modeling and Field Measurements
The predictive transport of suspended solids using a dispersion coefficient of 10 m /s
(108 ft2/s) provided a reasonable match with the results of field monitoring. The model
predicted a maximum elevation of suspended solids over background of 13 mg/L, and an
elevation of 5 mg/L extending approximately 400 feet (122 m) down current. The
Hudson River PCBs Superfund Site
Engineering Performance Standards
85
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
suspended solids levels measured in the samples collected during the field test displayed
some elevations above background that were slightly higher and extended further
downstream than the predictions. In addition, the turbidity measurements and suspended
solids data revealed much greater variability in the distribution of elevations than the
model predictions of suspended solids. These differences between predictions and
measured values are understandable given the following:
Dredging source term differences - The model assumed a constant, steady source
of sediment introduced to the water column while actual dredging proceeds at a
highly variable pace. The model also assumes release of the sediment over the
entire water column of the designated source cells. The actual release of material
during the dredging process can be much more focused at a particular location
(both x-y space in the depth).
Additional source terms - The model did not include additional source terms from
support activities in the area. In particular, the operation and grounding of the
support vessel (shallow draft tender tug) Miami II during the monitoring period
are thought to have contributed to some of the elevations noted in the suspended
solids data.
Comparison of the model predictions with field measurements provided two additional
insights that are important in planning additional modeling and monitoring efforts in the
Upper Harbor:
Three-dimensional flow field - Despite the shallowness of the Upper Harbor (i.e.,
generally 1 to 4 feet), the field measurements revealed distinct variations in the
flow field over depth. Although a two-dimensional simulation provides a
reasonable approximation for overall circulation, consideration must be given to
the vertical variation in flow when addressing transport issues.
Environmental factors - Even the moderate winds that occurred during the field
test had a measurable impact on the current regime. This highlights the
importance of the use of field measurements to assess model predictions and
sample collection locations on a daily basis.
6.2 Manistique River and Harbor, Michigan
The USACE RECOVERY model is employed to predict the temporal responses of
surface water to contaminated sediment. This model is generally employed to simulate
natural recovery of the river system. Input data to the RECOVERY model consists of
sediment contaminant concentration data from the sediment mixed-layer and
corresponding surface water concentrations. Output data consist of contaminant and
water concentration concentrations over a projected period of time. For the Manistique
River system,
Hudson River PCBs Superfund Site
Engineering Performance Standards
86
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
A second USACE model employed is the TGU (turbidity generating unit) model. This
model projects the amount of suspended mass per unit volume that will result from
dredging operations {i.e. resuspension). Typically, values of TGU range from 2 to 50
kg/m3 based various dredges and a variety of sediment bed types. This model assumes
that the dredge operates within a volume of water (m3) and using a solid mass balance
once can estimate the solids concentration in the water column surrounding the dredge
assuming the use of permeable vertical barriers both upstream and downstream of the
dredge. This set-up bases its analysis on the theory that the turbidity barriers will retain
all solids while allowing water to pass through the area. This assumes that the solids must
eventually settle out onto the stream body when the system reaches a steady state.
Once output is generated from the TGU model, the Equilibrium Model (EQUIL) is
utilized. EQUIL is a chemical release model that determines chemical equilibrium
between the particle bound solid and within the water column or aqueous phase. An end
result of this model is an estimate of the soluble fraction partitioning from the
resuspended solid and the constituent concentration in the dredged suspended sediment
on the river bottom.
The combination of these three models was used to simulate the dredging operation at
Manistique harbor. The RECOVERY model was used to simulate natural recovery
following dredging (the pre-dredge condition) and the TGU/EQUIL models were used to
predict the water concentration increase and the dredge suspended sediment deposit
increase {i.e. residual from dredging). Lastly, the results from the TGU/EQUIL models
were set as the starting or boundary condition into the RECOVERY model to simulate
the post-dredge sediment and water quality conditions projected into the future or for a
set period following the completion of dredging.
Results of the TGU/EQUIL model predicted a PCB water concentration during dredging
of 460ng/L. In comparison, actual water quality samples collected during dredging in
1997 resulted in an average PCB concentration in the water column of 230ng/L and
81ng/L in 1998 or an overall average for these two dredge seasons of 170ng/L. With
regard to sediment concentrations within the sediment mixed-layer following dredging,
the model predicted sediment PCB concentrations would increase to 30 ppm immediately
following dredging but assuming a natural depositional rate of 1 inch per year, the PCB
concentration in the sediment reduced to 10 ppm in the year 2000 (two years after
dredging), and to 0.012 ppm by the year 2020 (22 years after dredging). As indicated
previously, the average PCB concentrations measured in the sediment following dredging
in 1997 was 18.1 ppm while the average sediment PCB concentrations measured in the
year 2000 by the FIELDS team following the completion of all dredging activities was
7.06 ppm. Thus, it can be concluded that the TGU/EQUIL model overestimated dredging
resuspension and sediment residual concentrations following dredging activities.
Hudson River PCB s Superfund Site
Engineering Performance Standards
87
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
7.0 Response to GE's Comments on Hudson River FS
7.1 Summary of GE's Conceptual Model and Results
In Appendix A (Assessment of Sediment Resuspension and PCB Release During
Dredging Activities) of GE's comments on the FS (GE, 2001) Section 3.1, GE's
consultants presented a conceptual model of the near-field dredging area. Their analysis
assumed the following:
The near-field area can be approximated as a CSTR
Steady state condition exist in the near-field area
Equilibrium partitioning between the suspended phase and dissolved phase PCB.
Using these assumptions GE concluded that significant losses of resuspended PCBs are
expected. While the first two assumptions are reasonable, the third assumption does not
accurately represent the PCB desorption kinetics of this system.
7.2 Kinetics of PCB Desorption: Literature Review
Recent studies have demonstrated that desorption of hydrophobic chemicals from
sediments can be quite slow and that chemical equilibrium may not be a good
approximation in many real situations. In a dredging scenario, the residence time (contact
time) of the resuspended sediment in the water column is relatively short, on the order of
hours. For this period of time, it is unlikely that PCB reaches equilibrium.
Many researchers showed evidence that desorption of contaminants takes place in at least
two steps, a fast step and a slow step as discussed in Attachment C of this document. The
desorption of PCBs from Hudson River sediments was studied by Brown (1981) and
Carroll and co-workers (Carroll et al., 1994). Brown developed and tested a method for
the analysis of rates of PCB desorption from sediment suspended by dredging activities.
The data used were taken from dredging operations in the Hudson River at the town of
Fort Edward during 1977. The monitoring stations were placed in the east channel of
Rogers Island. Brown used the Freundlich isotherms model to obtain the sinking and
sorption-desorption rate constants of Aroclor 1016. In the report, the author used a term
sinking rate constant for the first order decay settling coefficient. In this study, the
sinking and sorption-desorption rates were chosen by trial and error method to fit the
measured concentration of Aroclor 1016 during the low and high flow conditions. For
low flow conditions, it was found that a sinking rate of -0.08 hr"1 and desorption rate
constants ranging from 0.025 hr"1 to 0.05 hr"1 fitted the measured data well. Under the
high flow conditions, a reasonable fit was obtained using a sinking rate of -0.4 hr"1 and
desorption rate constants on the order of 1.0 hr"1. Brown concluded that in the model, the
rate of PCB desorption from solids is proportional to the difference between the PCB
burden of the suspended sediments and the burden that would be in equilibrium with the
existing soluble concentration.
Hudson River PCBs Superfund Site
Engineering Performance Standards
88
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Carroll and co-workers studied desorption of PCBs from Hudson River sediment using
XAD-4 resin as a PCB adsorbent. They used sediments contaminated with high, medium,
and low levels of PCBs from the Hudson River near Moreau, NY. The three Hudson
River sediment used in their study contained 25, 64, and 205 mg/kg (dry weight) PCBs
with total organic carbon contents of 0.96, 3.43, and 4.59 percent, respectively. They
reported that the PCBs present in the sediments consisted primarily mono- and di-
chlorinated biphenyls (60-70 percent of total). Both a rapidly desorbing labile component
and a more slowly desorbing resistant component were observed. Rate constants for the
labile (fast) and resistant (slow) fractions were obtained using a model developed by
Berens and Huvard (1981). For the purpose of our study, the desorption rate constant of
the untreated moderately (64 mg/kg dry weight PCB) PCB-contaminated Hudson River
sediment is considered. The desorption rate constant obtained from Carrol and co-
workers study was approximately 0.018 hr"1 (refer to Table 5 in Attachment C).
Borglin and co-workers studied parameters affecting the desorption of hydrophobic
organic chemicals from suspended sediments (Borglin et al., 1996). In their paper,
Borglin and co-workers presented the results from the long-term experiments performed
for three hydrophobic organic chemicals (hexachlorobenzene and two polychlorinated
biphenyls). They concluded that the desorption times are on the order of a month to
several years and they observed that the desorption rates are dependent on the
particle/floc size and density distributions, the type of water, the amount of organic
carbon in the sediments, the time of adsorption before desorption, and the chemical
partition coefficient. Borglin and co-workers presented the results of the amount of PCBs
(monochlorobiphenyl and hexachlorobiphenyl) desorbed over time. From these results,
the rate constants obtain are on the order of 0.0049 hr"1 and 0.00042 hr"1 for
monochlorobiphenyl and hexachlorobiphenyl, respectively.
Cornelissen and co-workers studied the desorption kinetics of chlorobenzenes, PAH, and
PCBs for different contact times and solute hydrophobicity (Cornelissen et al., 1997).
They used a technique employing Tenax TAฎ beads as "sink" for desorbed solute to
measure the kinetics of desorption of the compounds mentioned above. For PCBs, they
studied PCB-65 (2,3,5,6-tetrachlorobiphenyl) and PCB-118 (2,3',4,4',5-
pentachlorobiphenyl). The sediment used was taken from Lake Oostvaardersplassen, The
Netherlands. They observed two stages of desorption rates, the rapid release of the
"labile" sorbed fraction and slow release of the "nonlabile" fraction. Two different
contact times were considered in this study, 2 and 34 days. The desorption rate constants
were varied for the different contact times for both the rapid and slow release. The values
are summarized in Attachment C.
In 1999, ten Hulscher and co-workers studied desorption kinetics and partitioning of
chlorobenzenes, PCBs, and PAHs in long term field contaminated sediment cores and top
layer sediment (ten Hulscher et al., 1999). They concluded that the desorption from
sediment was triphasic: fast, slow, and very slow. In this study, they used the sediment
from Lake Ketelmeer, The Netherlands. Only core results were presented for PCB-28.
Hudson River PCBs Superfund Site
Engineering Performance Standards
89
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
"3
They reported the desorption rate constant for very slow fraction with values of 0.21x10"
hr"1 and 0.19xl0"3 hr"1.
Ghosh and co-workers studied the relationship between PCB desorption equilibrium,
kinetics, and availability during land biotreatment (Ghosh et al., 2000). For this purpose,
they conducted a study of the equilibrium partitioning and desorption kinetics using
industrial lagoon sediments containing 0.91 percent oil and grease as a function of
biotreatment duration. A two compartment model was used to model the desorption of
PCBs from sediment. Tri-, tetra-, penta-, and hexa-chlorobiphenyls desorption rate
constants were reported. The values for the untreated sediment are summarized in
Attachment C.
Recently, ten Hulschler and co-workers studied desorption kinetics of in-situ
chlorobenzenes and 2,4,4'-trichlorobiphenyl (PCB-28) from River Rhine suspended
matter in Lobith, The Netherlands (ten Huschler et al., 2002). They observed fast, slow
and very slow desorption rates for PCB-28. Rate constants observed were on an average
of 0.2 hr"1 for fast, 0.0004 hr"1 for slow, and 0.00022 hr"1 for very slow desorption rates.
7.3 CSTR-Chem Model
A near-field CSTR model (CSTR-Chem) was developed to understand the net effect of
dredging on solids, fraction of dissolved PCB and total PCB flux. The model description,
its application and sensitivity are presented in section 4.3 of this attachment. CSTR-Chem
used a conservative rate of desorption of 0.2 hr"1. This desorption rate was applied to the
difference between the PCB concentration of the suspended sediments and the
concentration that would be in equilibrium with the existing soluble PCB concentration.
This formulation is consistent with the theory presented above.
Model simulations using CSTR-Chem suggest that the net fraction of dissolved PCB
from dredging operations under river flows of 4,000 cfs, is approximately 0.03 percent.
This net fraction of dissolved PCB of 0.03 percent was consistent for all near-field
velocity and river depth values simulated in the sensitivity analysis. Therefore, negligible
losses of PCBs are expected in the near-field dredging area.
Hudson River PCBs Superfund Site
Engineering Performance Standards
90
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
8.0 Case Studies - Dissolve Phase Releases and Export Rates
Every Superfund site represents a unique setting, with different hydrologic and geological
conditions, different discharge histories, and different contaminants. However, a study of
other dredging sites can provide information on the conditions that may be encountered
during this dredging project. In particular these other sites provide a basis to determine
what distances are reasonable for monitoring, what export rates are achievable and what
type of releases will occur. As the Hudson River PCBs Site is one of the largest
Superfund sites, identical or near-identical conditions would not be expected at other
sites. However, taken together, data from these other sites demonstrate the feasibility of
achieving the individual components of the Hudson River remedy.
The previous examination of the export rates for the case studies in the Responsiveness
Summary to the ROD (RS, USEPA, 2002) indicated:
The range of resuspension rates modeled as the average source strengths (best
engineering estimates) was reasonable. Furthermore, the data from the case
studies indicated that the export rates estimated are likely to overestimate the
anticipated export rate under routine conditions in the Hudson River.
The releases observed at other sites have been predominately associated with the
solids. As the solids are transported downstream dissolution will occur. The
magnitude of the dissolution is dependent on the sediments concentrations,
distance downstream and flow. The case studies with reliable split phase
concentrations support the conclusion that dredging-related PCB releases are
predominately solids.
Given the limitations of these case studies they are not used directly to infer the
conditions that will occur during dredging in the Hudson River. Therefore, the Remedial
Design should provide contingencies and dredging techniques to deal with site-specific
factors.
8.1 Introduction
None of the case studies examined provide specific estimates for the conditions in the
Hudson River. Rather, the studies presented evidence for:
The range of export rates achieved and how the export rates can be accurately
determined;
The type of releases (i.e. solid or dissolved phase) that generally occur.
In the case studies reviewed, the monitoring plans, sediment
concentrations/classifications, the nominal flows and weather conditions were different
than those anticipated in the Hudson River. It is acknowledged that the case studies do
not provide perfect templates, and therefore they were not used as such.
Hudson River PCBs Superfund Site
Engineering Performance Standards
91
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
The three case studies examined in depth were New Bedford Harbor, Fox River, and
Hudson Falls. Since these sites were examined previously for the Feasibility Study
(USEPA, 2000a) and RS (USEPA, 2002), only new analyses or further clarification for
each of these three sites is provided below. Other case studies were also examined, but
either there was not enough information concerning resuspension or the conditions were
too dissimilar to be relevant to the Hudson River PCBs Site; these are discussed briefly.
8.2 New Bedford Harbor, Massachusetts
The New Bedford Harbor Pre-Design Field Test dissolved phase releases were also
discussed in Attachment C. The discussion provided here is specific to the modeling
results presented in this attachment. The New Bedford Harbor Superfund Site is located
in Bedford, Massachusetts, about 55 miles south of Boston. The site is contaminated with
PCBs, heavy metals, and other organic chemicals from industrial discharges. Removal of
PCB-contaminated sediments in hot spots located on the west side of the Acushnet River
estuary was completed between April 1994 and September 1995. Dredging of the hot
spots was performed using a hydraulic dredge, and the slurry was subsequently pumped
into a confined disposal facility (CDF). Following the hot spot dredging, a pre-design
field test using mechanical dredging equipment was performed in August 2000 and
documented in the Pre-Design Field Test Final Report (USACE, 2001). During the Pre-
Design Field Test the area directly around the dredge was referred to as the moonpool. At
times oily sheens and oily slick releases were noticed. The report contains detailed
information regarding the dredging operation, water quality monitoring for turbidity,
particulate PCBs, dissolved PCBs, threshold water column levels, and contingency plans
to be put in effect in the event that the action level was detected at one of the monitoring
stations. Since the hot spot removal has been previously discussed in depth in the RS
(USEPA, 2002), only the pre-design study is considered in this analysis.
Export Rate
A rough estimate of the PCB loading was provided in Attachment C. However due the
lack of flow data, the results are not discussed any further in this attachment.
Dissolved Phase Release
In the Pre-design Field Report it was noted that New Bedford Harbor contains free oil
phase PCBs as well as sediment-bound PCBs. For this analysis (and the analysis in the
Performance Standard Report), the data from the oil releases and moonpool were not
included since these samples represent a multiphase system, and multi-phase systems are
not applicable to the lower PCB concentrations typical of the Hudson. Essentially,
samples labeled as "oily sheen" or "oil slick" do not apply to the sediment resuspension
processes anticipated for the Hudson. Exclusion of these oil-bearing samples provides a
more consistent picture of the PCB release process at New Bedford Harbor.
In Figure 45, the total, suspended, and dissolved phase PCB concentrations are presented
as a function of distance upstream and downstream of the dredging operations. For each
Hudson River PCBs Superfund Site
Engineering Performance Standards
92
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
PCB form (total, suspended and dissolved), two plots are presented - one showing all
data, and a second showing an expanded scale. In each case, samples within the
"moonpool" around the dredging operation (0 distance from the dredge) show very high
levels relative to baseline (i.e., upstream) conditions. These samples represent conditions
in the immediate vicinity of the dredge. Examining the expanded scale graphs allows a
comparison of the upstream and downstream conditions. In this comparison, it is clear
that all three forms of PCB (total, dissolved, and suspended) increased downstream of the
dredge, indicative of resuspension release. These conditions represent the near-field
conditions referred to in the standard. However, it is also clear that the suspended matter
concentration has increased substantially more than the dissolved phase, indicating that
the primary form of the net PCB increase took place in suspended matter form, consistent
with the analysis provided in the standard. The suspended matter concentration increased
by more than 100 percent from approximately 500 ng/L to 1000-1500 ng/L. The
dissolved phase increased form about 500 ng/L to about 750 ng/L or about 50 percent.
The impact of the dredging related release can also be seen in Figure 46, which presents
the fraction of the dissolved phase as a function of total PCB concentration and distance
from the dredge. In the diagram comparing dissolved fraction to total PCB, there is a
clear trend toward lower dissolved fractions as the total PCB concentration increases [i.e.,
the fraction of the Total PCB load in the dissolved fraction decreases as the Total PCB
load (sum of dissolved and suspended) increases]. This trend correlates with the decrease
in dissolved fraction PCB that occurs from upstream to downstream, as also shown in the
figure. These data all support the assertion that PCB releases due to dredging occur
primarily as a suspended matter release and thus can be tracked in the near field by
suspended matter or possibly turbidity measurements. This also shows that PCBs enter
the water column as suspended matter, a process that is independent of the baseline
dissolved phase PCB concentration.
Subsequent to the resuspension, greater dissolution of PCBs takes place but the elevated
PCB suspended matter fraction remains, indicating that it is possible to track PCB
releases by suspended matter or turbidity. Additionally, as shown in Figure 45, the total
PCB concentrations increased by roughly 1,000 ng/L or about 100 percent. Of the 1000
ng/L increase, roughly 750 ng/L is particle-borne and 250 ng/L is dissolved phase-borne.
This corresponds to an increase in TSS of roughly 100 percent, consistent with the PCB
gain. This TSS signal would be readily detected by the monitoring scheme required for
the standard. Notably, the dissolved baseline PCB concentrations, while elevated at 500
ng/L, are not so far above those typically found in the Hudson during peak summer time
conditions (150 to 200 ng/L). Thus, similar behavior of PCBs is expected in the Hudson
with respect to the downstream distribution on dissolved and suspended matter fractions.
Results
As noted, the Pre-Design Field Test was not used to estimate the magnitude of dredging
related PCB releases. Only the nature of the releases was examined. Nonetheless the data
clearly show elevated mean concentrations of PCBs downstream of the dredge, regardless
of the downstream distance. Additionally, the data show increased mean PCB
concentrations on the suspended matter, as well as an increase in suspended solids at all
points downstream (see Figure 47). The examination of these data shows that the
Hudson River PCBs Superfund Site
Engineering Performance Standards
93
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
suspended solids would be clear indicators of the PCB releases and that the dredging-
related PCB releases are predominately from solids.
8.3 Fox River SMU 56/57 1999 And 2000 Dredging Projects,
Wisconsin
The Fox River sediment management unit (SMU) 56/57 is located along the Fox River
adjacent to the Fort James Plant. This river system is part of the Great Lakes Area of
Concern. Approximately 80,000 cy of PCB-contaminated sediment were targeted for
removal using a hydraulic cutter head dredge. After one week of dredging activities, the
dredge was switched to an IMS 5012 Versi dredge in attempt to increase the solids
content of the dredge slurry. The dredge was upgraded two more times during the first
month of dredging in an attempt to meet an optimum production rate of 200 cy/hr. The
Fox River SMU 56/57 was divided into 100 x 100 foot subunits. Dredging was conducted
from August 1999 to December 1999. It was determined at the end of Phase I (December
1999) that unacceptably high residuals were left in the area dredged due to mounds of
sediment left behind between dredge passes. As a result, the dredging equipment was
switched to a horizontal auger dredge for Phase II, which was carried out from late
August 2000 to the end of November 2000. Phase I subunits were re-dredged to meet a 1
ppm PCB residual concentration. The activities were documented in the Final Summary
Report for Sediment Management Unit 56/57 (September 2000) and the Environmental
Monitoring Report (July 2000). The reports contain information regarding water quality
monitoring, PCB water column levels and loading, turbidity measurements, and post-
dredge sampling. Since, the export rate was estimated in the Responsiveness Summary
(RS, USEPA 2002) the discussion below only discusses why the export estimation is
likely an overestimate of the conditions anticipated during dredging in the Hudson.
Export Rate
The export rate determined for the Fox River site is not directly applicable to the export
rates anticipated in the Hudson due to difference in the monitoring locations, dredge type
used, and sampling technique. However, the Fox River export estimate obtained is within
the range considered in the performance standard criteria.
As noted in the Resuspension White Paper in the RS (USEPA, 2002), the Fox River
studies were complicated by the location of the monitoring stations. The fact that
significant loss of PCBs only occurred when the dredging area was close to the sampling
cross-section suggests that settling of any resuspended matter occurs within a short
distance of the dredging operation. Only when the monitoring location was close to the
dredging could this signal be detected. This suggests that the loads obtained by this study
do not represent PCB released for long-distance transport. Rather, the PCBs appear to be
quickly removed from (settle out of) the water column a short distance downstream. As
such, it is inappropriate to use these results to estimate downstream transport from a
dredging site.
Hudson River PCBs Superfund Site
Engineering Performance Standards
94
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Furthermore, as discussed in the white paper, the higher resuspension rates may also be a
result of the dredge used in these operations. In fact, the New Bedford pilot study
compared the sediment resuspension characteristics of a horizontal auger dredge (used in
Fox River) with a conventional hydraulic cutterhead suction dredge and found a disparity
similar to that observed between the Fox River and average source strength estimates.
The sample compositing may not have been performed in such a manner as to account for
flow. As noted it the Resuspension White Paper in the RS:
The sample compositing strategy [of the Fox River Studies], designed to reduce the
number and cost of PCB analyses, was contrary to the mass flux analysis attempted. The
equal volume composites do not allow consideration of flow variation across the cross-
section. USGS (2000) states that stagnant areas and even reversed flows were observed
during sampling operations, confirming the errors associated with the composite PCB
samples. The TSS sample composites induce less error and provide a more accurate
estimate of downstream TSS flux, yet they showed an unexplained decrease in suspended
sediment across the dredging operation. The decrease is almost certainly an artifact
associated with compositing equal volume samples from 20 percent and 80 percent depth.
Even though it has long been established that velocity measurements from these depths
represent the average velocity in an open channel, there is no justification for suggesting
that a composite sample from these depths represents the average concentration along the
profile. This is particularly true in deeper water where the two samples represent 25 feet
or more of water depth. (USEPA, 2002)
As discussed previously in the Responsiveness Summary for the ROD (USEPA, 2002),
Attachment C, there were several reasons why the field estimates for Fox River were
considered overestimations. The most significant of these is that the proximity of the
monitoring locations to the dredging operations did not allow for export to be reliably
calculated. The sampling locations were located too close to the operations, and therefore
export estimates from these samples did not account for settling. In addition, the samples
taken in the cross-sections were not composited in a manner representative of the entire
load. Despite these reservations, a rate of loss equivalent to 2.2 percent was obtained
from the previous analysis. It should be noted that a short-term (days to weeks) export
rate of 2.2 percent would not cause exceedances of the Resuspension Standard (i.e., 500
ng/L) in any of the river sections. Furthermore, the models indicate that a release of 2.2
percent would only represent a concern for the 350 ng/L Total PCB criterion in River
Section 2 due to the higher sediment concentrations. However, according to the modeling
this resuspension rate would represent loads greater than 600 g/day Total PCB, thus
prompting additional sampling and possibly additional engineering controls if these
levels are sustained. Ultimately, the Resuspension Standard has been designed to allow
for occasional large loads without prompting immediate cessation of the operation.
Dissolved Phase Release
It is unclear how much time elapsed between sample collection and separation of the
sample into dissolved and particulate fractions, confounding conclusions with regard to
dissolved and suspended loads. The data provide evidence of this lag in separations. As
Hudson River PCB s Superfund Site
Engineering Performance Standards
95
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
noted in the RS, the data are not consistent with a large dissolved phase release based on
the lack of change in PCB congener pattern across the dredging area. A large dissolved-
phase PCB contribution from the sediments, either by porewater displacement or
sediment-water exchange, should yield a gain whose PCB congener pattern is similar to
that of the filter supernatant. The fact that the congener pattern is unchanged across the
study area suggests a direct sediment addition. Yet the TSS data do not document an
increase in suspended sediment concentrations. Please refer to the Resuspension White
Paper in the RS (USEPA, 2002) for further details.
Results
The measurements provided in the Fox River report are not applicable or appropriate for
use directly in the Resuspension Performance Standard for a variety of reasons. As noted
in the Resuspension White Paper in the RS, the Fox River study was complicated by the
location of the monitoring stations. In this case study there was a paper mill close by that
significantly affected the monitoring results. Furthermore,
The fact that significant loss of PCBs only occurred when the dredging area was close to
the sampling cross-section suggests that settling of any resuspended matter occurs within
a short distance of the dredging operation. Only when the monitoring location was close
to the dredging could this signal be found. This suggests that the loads obtained by this
study do not represent PCB released for long-distance transport. Rather, the PCBs appear
to be quickly removed from the water column a short distance downstream. As such, it is
inappropriate to use these results to estimate downstream transport from a dredging site.
(USEPA, 2002)
The data are not particularly useful for analysis of the PCB release mechanisms during
dredging either, since the times lag prior to phase separation of the split samples may
have allowed for further dissolution between the phases. Despite the analysis performed
in the Resuspension Standard Report as well as previous reports suggesting no significant
dissolved release will exist at the dredge, the resuspension criteria do not rely on this (i.e.,
assuming that the dissolved phase releases are small relative to the suspended phase). The
criteria downstream are for total PCBs, both dissolved and particulate, and therefore
releases in either phase (dissolved or suspended) will be detected.
8.4 Hudson Falls
Hudson River sediments were removed from the vicinity of the GE pump house near
Hudson Falls. Sediments in this area contained high levels of PCBs, as well as pure PCB
oil. Dredging was accomplished by diver-directed suction hoses over a total period of
about seven months (October through December, 1977, and August through November,
1998). During this period, GE conducted its regular monitoring at Bakers Falls and
Rogers Island, which can be used to estimate the effects of dredging to the downstream.
Since the original analysis of the export rate was provided in the previous analysis
(USEPA, 2002), the following discussion is only provided to further clarify the
conservative assumptions incorporated in that analysis.
Hudson River PCBs Superfund Site
Engineering Performance Standards
96
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Export Rate
In the Hudson Falls dredging project, PCBs were present in the non-aqueous phase liquid
(NAPL) form as well as on sediments. The presence of this NAPL PCB has the potential
to escape on its own or to supersaturate the water column. As a result, the anticipated
release and export rates should be higher than that expected from sediment resuspension
alone. The mass of sediment removed from Hudson Falls was provided by the NYSDEC
and the average PCB concentrations were taken from cores in the dredged area. Even if
the calculations of the mass were off by a factor of two, the export rate would still be less
than 1 percent. PCB export at this rate would not exceed the Resuspension Standard in
any river section, based on the modeling analysis Furthermore the export rates estimated
for the Hudson Falls site represent upper bounds on the losses due to dredging because of
the historical sources between Bakers Falls and Rogers Island, (i.e., the Hudson Falls and
Fort Edward facilities). While the baseline is considered relatively well constrained as a
result of controls implemented by GE at Hudson Falls, the addition of PCBs by the GE
facilities was still occurring at the time, thus potentially adding to the total load and
yielding an overestimate of the export from the Hudson Falls site. Overall, the conditions
noted for the Hudson Falls dredging project suggest that the conditions experienced were
likely to have been much worse than those to be anticipated for this dredging project. The
means of estimating loads represents a conservative approach and thus provides a useful
upper bound on the actual PCB export. For these reasons it was a useful site for inclusion
in the analysis for the resuspension standard.
Dissolved Phase Release
Split phase data were not available for this site.
Results
Since the export rate estimations for the Hudson Falls dredging operations were based on
conservative assumptions, it is likely that the export rate has been overestimated.
8.5 Other Sites
Data from Fox River Area N and Manistique Harbor were not used for comparison to the
modeled dissolved phase release and export rates based on the project size as well as the
application of a dredging technology that was deemed inappropriate for the Hudson and
unlikely to be used (based on its apparent loss rate). For the Fox River Area N study only
slightly more than 100 lbs of PCBs were removed, suggesting that operations were too
small to become routine. Much of the loss may have been associated with start-up. It is
likely that the larger project on the Fox River (Areas 56/57 with nearly 1,500 lbs of PCBs
removed) is more reflective of the dredging related losses even though these are probably
overestimated as well. The data for Manistique are not available, however it is known
that dredging at Manistique was primarily accomplished with a cable arm bucket dredge
(although other dredges were used as well). In the dredged locations, extensive areas of
dense, coarse sediments and debris inhibited the effectiveness of the dredge bucket. The
cable arm bucket is designed to dredge soft sediments and does not perform as well
Hudson River PCBs Superfund Site
Engineering Performance Standards
97
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
where either consolidated materials or debris are present. Thereby, the Remedial Design
will have to consider the type of dredge as well as the other engineering contingencies,
particularly in areas identified as likely to resuspend.
8.6 Conclusions
The export rates obtained from the case studies are not directly applicable for comparison
to the resuspension criteria since these represent daily averages and the criteria pertain to
running averages. The long-term effects on the river will be dependent on the export rates
downstream. The case studies exhibit that the monitoring stations should be sufficiently
downstream to correctly measure the release rate (i.e., the load to the Lower Hudson
River). As the near-field transport model of the Performance Standard Report and the Fox
River case study indicated much of the TSS settle close to the dredging operations. It is
likely that these solids will be removed as the dredge moves downstream.
Ultimately, these studies are not expected to be comprehensive templates for dredging on
the Hudson since the conditions of dredging (operations, engineering contingencies, etc.)
may have been different from those to be used on the Hudson River PCBs Site. The case
studies are used to show that dredging operations at other sites (even in the Hudson) have
had success with minimizing export through various techniques and engineering
contingencies.
When taken together, these sites demonstrate a consistent level of site clean-up and
resuspension release when viewed on a relative basis. The Resuspension Standard as
developed for the Hudson River PCBs Site does not require greater degree of control for
resuspension than that achieved by other remedial efforts at other sites.
Hudson River PCBs Superfund Site
Engineering Performance Standards
98
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
9.0 References
Ambrose, R. B., Jr., T. A. Wool, and F. Martin. 1993. The water quality analysis
simulation program,WASP5,part A: model documentation. U.S. Environmental
Protection Agency, Athens, GA.
Berens, A.R. and G.S. Huvard. 1981. Particle Size Distribution of Polymers by Analysis
of Sorption Kinetics, Journal of Dispersion Science and Technology, Vol 2, pp. 359-387,
1981. As cited in Carroll, et al. (1994).
Borglin, S., A. Wilke, R. Jepsen, and W. Lick. 1996. Parameters Affecting the
Desorption of Hydrophobic Organic Chemicals from Suspended Sediments. Env. Tox.
Chem. Vol. 15, No. 10, pp. 2254-2262.
Brown, M. 1981. PCB Desorption from River Sediments Suspended During Dredging:
An Analytical Framework. New York State Department of Environmental Conservation,
Technical Paper No. 65. April 1981.
Carroll, K.M., M.R. Harkness, A. A. Bracco, and R.R. Balcarcel. 1994. Application of a
Permeant/Polymer Diffusional Model to the Desorption of Poly chlorinated Biphenyls
from Hudson River Sediment. Environ. Sci. Technol. Vol 28, pp. 253-258. 1994.
Cornelissen, G., P.C.M. Van Noort, and A. J. Govers. 1997. Desorption kinetics of
chlorobenzenes, polycyclic aromatic hydrocarbons, and polychlorinated biphenyls:
sediment extraction with Tenaxฎ and effects of contact time and solute hydrophobicity.
Environ. Toxicol. Chem. Vol 16, No. 7, pp. 1351-1357, 1997
DePinto, Joseph V., Wilbert Lick and John F. Paul. 1994. Transport and Transformation
of Contaminants Near the Sediment-Water Interface. CRC Press, Inc. Boca Raton. 1994.
Farley, K.J., R.V. Thomann, T.F. Cooney, D.R. Damiani, and J.R. Wand. 1999. An
Integrated Model of Organic Chemical Fate and Bioaccumulation in the Hudson River
Estuary. Prepared for the Hudson River Foundation. Manhattan College, Riverdale, NY.
Filtration & Separation.com, 2003. website address, http://www.filtration-and-
separation.com/ settling/settling.htm accessed on February 23, 2003.
Fischer, H.B, E.J. List, R. C.Y. Koh, J. Imberger, and N.H. Brooks. 1979. Mixing in
Inland and Coastal Waters. Academic Press, New York, 1979.
Fox River Remediation Advisory Team (FRRAT). 2000. Evaluation of the Effectiveness
of Remediation Dredging: The Fox River Deposit N Demonstration Project November
1998 - January 1999 Madison, Wisconsin, Water Resources Institute Special Report,
WRI SR00-01. June 2000.
Hudson River PCB s Superfund Site
Engineering Performance Standards
99
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Gbondo-Tugbawa, S.S., CT Driscoll, JD Aber, GE Likens. 2001. Evaluation of an
integrated biogeochemical model (PnET-BGC) at a northern hardwood forest ecosystem.
Water Resources Research 37:1057-1070.
General Electric (GE). 2001. Comments of General Electric Company on the Feasibility
Study and Proposed Plan for the Hudson River PCBs Superfund Site. April 2001. With
Appendix A: Assessment of Sediment Resuspension and PCB Release During Dredging
Activities, prepared for GE by QEA.
Ghosh, U., A.S. Weber, J.N. Jensen, and J.R. Smith. 2000. Relationship between PCB
desorption equilibrium, kinetics, and availability during land biotreatment. Environ. Sci.
Technol. Vol. 34, No. 12, pp. 2542-2548, 2000.
Gobas, F.A.P.C., M.N. Z'Graggen and X. Zhang. 1995. Time response of the Lake
Ontario ecosystem to virtual elimination of PCBs. Env. Science Technol. 29(8):2038-
2046.
Gobas, F.A.P.C. 1993. A model for predicting the bioaccumulation of hydrophobic
organic chemicals in aquatic food-webs: Application to Lake Ontario. Ecological
Modelling 69:1-17.
Herbich, J.B., and S.B. Brahme. 1991. Literature review and technical evaluation of
sediment
resuspension during dredging. Contract Report HL-91-1. Prepared for the US Army
Engineer
Waterways Experiment Station, Vicksburg, MS. Texas A&M University, College
Station, TX.
Kuo, A. Y., C. S. Welch and R. J. Lukens. 1985. Dredge Induced Turbidity Plume
Model, ASCE Journal of Waterways, Ports, Coastal and Ocean Engineering, Vol. Ill,
pp 476-494.
Kuo, A. Y. and D. F. Hayes. 1991. Model for Turbidity Plume Induced by Bucket
Dredge, Journal of Waterway, Port, Coastal, and Ocean Engineering, Vol. 117, No. 6,
Nov/Dec.
Martin, J. L. and S. C. McCutcheon. 1999. Hydrodynamics and Transport for Water
Quality Modeling. Lewis Publishers, New York, 1999.
Orton P.M. and G.C. Kineke. 2001. Comparing Calculated and Observed Vertical
Suspended-Sediment Distributions from a Hudson River Estuary Turbidity Maximum.
Estuarine, Coastal and Shelf Science, Vol 52, pp. 401-410.
QEA. 1999. PCBs in the Upper Hudson River. Volume 2: A Model of PCB Fate,
Transport, and Bioaccumulation. Prepared for General Electric, Albany, New York by
Quantitative Environmental Analysis, LLC. May 1999, and amended June 1999.
Hudson River PCBs Superfund Site
Engineering Performance Standards
100
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
ten Hulscher, Th.E.M., B.A. Vrind, H. Van den Heuvel, L.E. Van der Velde, P.C.M. Van
Noort, J.E.M. Beurskens, and H.A.J. Govers. 1999. Triphasic desorption of highly
resistant chlorobenzenes, polychlorinated biphenyls, and polycyclic aromatic
hydrocarbons in Field Contaminated Sediment. Environ. Sci. Technol. Vol. 33, No. 1, pp.
126-132, 1999.
ten Hulscher, Th.E.M., B.A. Vrind, P.C.M. van Noort, and H.A.J. Grovers. 2002.
Resistant sorption of in situ chlorobenzenes and a polychlorinated biphenyl in river Rhine
suspended matter. Chemosphere, Vol 49, pp. 1231-1238, 2002.
Thonon, Ivo, and Marcel Van Der Perk. 2002. Measurement of Suspended Sediment
Characteristics in an Embanked Flood Plain Environment of the River Rhine, Erosion
and Sediment Transport Measurement: Technological and Methodological Advances.
Workshop in Oslo 19 - 21 June 2002.
U. S. Army Corps of Engineers (USACE). 1990. New Bedford Harbor Superfund Pilot
Study: Evaluation of Dredging and Dredged Material Disposal. US Army Corps of
Engineers, New England Division. May 1990.
USACE. 1998. Entrainment by Hydraulic Dredges - A Review of Potential Impacts.
Dredging Operations and Environmental Research Technical Note, DOER-E1. U.S.
Army Engineers Waterways Experiment Station (USAEWES).
USACE. 2001. Final Pre-Design Field Test Dredge Technology Evaluation Report, New
Bedford Harbor Superfund Site, New Bedford, Massachusetts. Prepared by Foster
Wheeler Environmental Corporation, Boston, Massachusetts. August 2001.
US Environmental Protection Agency (USEPA). 1997. Phase 2 Report, Further Site
Characterization and Analysis, Volume 2C - Data Evaluation and Interpretation Report
(DEIR), Hudson River PCBs RI/FS. Prepared for USEPA Region 2 and USACE by
TAMS Consultants, Inc., the Cadmus Group, Inc., and Gradient Corporation. February
1997.
USEPA. 1998. Further Site Characterization and Analysis. Volume 2C-A Low
Resolution Sediment Coring Report (LRC), Addendum to the Data Evaluation and
Interpretation Report, Hudson River PCBs Reassessment RI/FS. Prepared for USEPA
Region 2, New York by TAMS Consultants, Inc., Gradient Corporation, and TetraTech,
Inc. July 1998.
USEPA, 2000a. Phase 3 Report: Feasibility Study, Hudson River PCBs Reassessment
RI/FS. Prepared for EPA Region 2 and the US Army Corps of Engineers (USACE),
Kansas City District by TAMS Consultants, Inc. December 2000.
USEPA. 2000b. Further Site Characterization and Analysis, Revised Baseline Modeling
Report (RBMR), Hudson River PCBs Reassessment RI/FS Volume 2D. Prepared for
USEPA Region 2 and USACE, Kansas City District by TAMS Consultants, Inc., Limno-
Tech, Inc., Menzie-Cura & Associates, Inc., and Tetra-Tech, Inc. January 2000.
Hudson River PCBs Superfund Site
Engineering Performance Standards
101
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
USEPA. 2000c. Phase 2 Report Further Site Characterization and Analysis, Volume 2E -
Revised Human Health Risk Assessment Hudson River PCBs Reassessment RI/FS.
Prepared by TAMS Consultants, Inc. and Gradient Corporation. November.
USEPA, 2002. Responsiveness Summary Hudson River PCBs Site Record Of Decision.
Prepared for USEPA Region 2 and USACE by TAMS Consultants, Inc. January 2002.
U.S. Geological Survey (USGS). 1969. Time-of Travel Study, Upper Hudson River, Fort
Edward, New York to Troy Lock and Dam, Troy, New York. Report of Investigation,
Rl-10. 1969.
USGS. 2000. A mass-balance approach for assessing PCB movement during remediation
of a PCB-contaminated deposit on the Fox River, Wisconsin. USGS Water Resources
Investigations Report 00-4245. December 2000.
Warren, S.D., Boff, R.F., and Simpson, H.J. 1997. Volatilization of PCBs from
Contaminated Sediments and Water. Final Report to NY State Department of
Environmental Conservation, Lamont-Doherty Geological Observatory of Columbia
University, Palisades, NY. October 1997.
Hudson River PCBs Superfund Site
Engineering Performance Standards
102
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Tables
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment D - April 2004
-------
Table 1
Properties of Hudson River Sediments
Non-cohesive sediments
Cohesive sediments
Typical location
Deeper areas and channel
Shallower areas
Fine sand or coarser (%)
80
35
Silt or finer (%)
20
65
Solids (%)
76
58
In-situ Density (gm/cc)
1.74
1.45
Organic content (%)
1 to 2
3 to 4
Average Particle Size
62 |im - 250 |im
< 1 |im to 62 |im
Particle Density
2.2-2.6
2.2 -2.6
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 2
Summary of Settling Velocities
Reference
Particle Density
Particle Size
Vs or w (cm/s)
Sequoia Scientific, Inc
Not Indicated
50 microns
0.01
100 microns
0.1
400 microns
0.005
DePinto et al, 1994
DePinto et al, 1994
20.7 microns
0.0124
Passaic Valley
Freshwater Sewage
Sludge
22 microns
0.0022
Filtration & Separation.com,
2003
2.2 g/cc
100 microns
0.603
2.6 g/cc
100 microns
0.789
2.2 g/cc
400 microns
4.7
2.6 g/cc
400 microns
5.8
Thonon and Van Der Perk,
2002
Not Indicated
10 microns
0.001
50 microns
0.005
100 microns
0.01
400 microns
0.001-0.1
Kuo and Hayes, 1991
St. John's River
2.40 g/cc
39.6 microns
0.12
Black Rock Harbor
2.39 g/cc
36.3 microns
0.1
Thames River
2.50 g/cc
150 microns
1.84
160 microns
2.1
Kuo et al, 1985
From paper: 2.65 g/cc
20 microns
3.59 X 10"2
HR: 2.2 g/cc
20 microns
0.026
HR: 2.6 g/cc
20 microns
0.035
HR: 2.2
100 microns
0.653
HR: 2.6 g/cc
100 microns
0.871
HR: 2.2 g/cc
400 microns
10.453
HR: 2.6 g/cc
400 microns
13.938
US ACE, 2001
Silt
20 microns
3.21 X 10"6
Clay
2 microns
3.21 X 10"8
QEA, 1999
Silt
Based on cohesive
Hudson River
0.005 to 0.01
(4-9m/day)
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 3
Surface Water Elevation Slope in TI Pool based on USGS Gauge Data
Monthly Average Elevation Difference (ft)
Slope
including
negative values
(6 mile
Month
negative values
treated as 0
distance)
3
1.05
1.05
3.00E-05
4
0.676
0.694
2.00E-05
5
0.416
0.436
1.00E-05
6
0.223
0.244
8.00E-06
7
0.151
0.169
5.00E-06
8
0.147
0.168
5.00E-06
9
0.166
0.185
6.00E-06
10
0.234
0.254
8.00E-06
11
0.336
0.349
1.00E-05
12
0.577
0.582
2.00E-05
Dredging period
Average
0.239
0.258
8.00E-06
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 4
Estimated Shear Velocity and Lateral Dispersion Coefficient for
Upper Hudson River
Based on Water Elevation Slope
Shear
Lateral Dispersion
Flow
Velocity
Coefficient
RM
(cfs)
Location
Depth (m)
Slope
(m/s)
(cmA2/s)
overall
2.4
8.00E-06
0.01
190
west
0.9
8.00E-06
0.01
40
o
o
center
3.5
8.00E-06
0.02
350
o
(N
east
2.4
8.00E-06
0.01
190
overall
2.6
8.00E-06
0.01
200
west
1.1
8.00E-06
0.01
100
o
o
center
3.7
8.00E-06
0.02
400
o
east
2.6
8.00E-06
0.01
200
overall
2.7
8.00E-06
0.01
240
west
1.2
8.00E-06
0.01
70
o
o
center
3.9
8.00E-06
0.02
410
o
east
2.7
8.00E-06
0.01
240
overall
3
8.00E-06
0.02
280
cn
Q\
west
1.6
8.00E-06
0.01
110
o
o
center
4.2
8.00E-06
0.02
460
ง
o
00
east
3.1
8.00E-06
0.02
280
overall
2.9
8.00E-06
0.02
260
west
3
8.00E-06
0.02
280
o
o
center
4
8.00E-06
0.02
420
o
(N
east
1.7
8.00E-06
0.01
120
overall
3.1
8.00E-06
0.02
290
west
3.2
8.00E-06
0.02
310
o
o
center
4.2
8.00E-06
0.02
450
o
east
1.9
8.00E-06
0.01
140
overall
3.2
8.00E-06
0.02
300
west
3.3
8.00E-06
0.02
320
o
o
center
4.3
8.00E-06
0.02
470
o
east
2
8.00E-06
0.01
150
overall
3.5
8.00E-06
0.02
350
o
On
west
3.6
8.00E-06
0.02
370
o
o
center
4.6
8.00E-06
0.02
520
ง
o
00
east
2.3
8.00E-06
0.01
190
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 5
Silt Fractions in Hudson River Sections
Section
Cohesive Sediment
Fraction 1
Non-Cohesive Sediment
Fraction1
Silt Fraction2
1
0.37
0.63
0.37
2
0.62
0.38
0.48
3
0.62
0.38
0.48
Note:
1. Sediment in each river section is consisted of cohesive sediment and non-cohesive sediment.
The sum of cohesive sediment fraction and non-cohesive sediment fraction is equal to 1.
2. It is assumed that the percentage of silt is 65% in the cohesive sediment and 20% in the non-cohesive sediment.
Therefore, the silt fraction in Section 1 is 0.37*0.65+0.63*0.2 = 0.37 and in Section 2 and 3 is 0.65*0.62+0.2*0.38 = 0.48.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 6
Summary of CSTR-Chem Model simulation results for dredging operations in
Section 1-3 of the Hudson River
Ambient River Characteristics
min Ambient TSS - Silt (mg/L)
CTotaUn Ambient PCB (ng/L)
Fd in Fraction Dissolved in BKG
Q River flow (cfs)
H Water Depth (m)
u Upstream velocity (m/s)
Dredging and Sediment Characteristics
V! Settling Velocity Silt (m/s)
v2 Settling Velocity Sand (m/s)
Fsilt Fraction Sediment Silt
csed Sediment PCB (mg/Kg)
M dot R Resuspension rate (kg/sec)
CSTR Conditions
wnf width of the near field (m)
qnf CSTR flow (m3/s)
Anf Horizontal Area (m )
Vnf CSTR Volume (m3)
0 nf Retention time (s)
PCB Geochemistry
Kd Partition Coefficient (L/kg)
k Desorption Rate (1/hr)
Model Simulation Results
Total TSS (Combined silt and coarse materials)
m(dredge) TSS from dredge
m(loss) TSS lost to settling (mg/L)
m(out) TSSout (mg/L)
Sediment Type 1 - Silt
m(dredge) TSS from dredge
m(loss) TSS lost to settling (mg/L)
m(out) TSSout (mg/L)
River Sections
Section 1 Section 2 Section 3
2.3
2.3
1.7
122
76
57
0.9
0.9
0.92
4000
4000
4000
1.88
1.88
1.88
0.131
0.131
0.131
1.00008
0.00008
0.00008
0.06
0.06
0.06
0.3665
0.479
0.479
27
62
29
1
1
1
10
10
10
2.4623
2.4623
2.4623
100
100
100
188.4
188.4
188.4
77
77
77
48309
48309
51151
0.2
0.2
0.2
406
406
406
183
151
151
226
258
257
149
195
195
0
1
1
151
196
196
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 6
Summary of CSTR-Chem Model simulation results for dredging operations in
Section 1-3 of the Hudson River
Model Simulation Results (cont.)
Sediment Type 2 - Coarse materials
m(dredge) TSS from dredge
257
212
212
m(loss) TSS lost to settling (mg/L)
182
150
150
m(out) TSSout (mg/L)
75
62
62
Equilibrium Conditions
Cd,eq Equilibrium Dissolved Cone (ng/L)
535
1218
541
Cs eq Equilibrium Suspended Cone (ng/L)
10552
24037
11293
Cp eq Equilibrium Particle cone (mg/kg)
25.8
58.9
27.7
Fd,eq Equilibrium Dissolved Fraction
0.048
0.048
0.046
Fs eq Equilibrium Particulate Fraction
0.952
0.952
0.954
Transient Partitioning Conditions
CTotal Exiting Total Cone (ng/L)
6172
15966
7483
Cd Exiting Dissolved Cone (ng/L)
111.6
73.3
54.5
Cs Exiting Suspended Cone (ng/L)
6060
15893
7428
Cp Exiting Particle Cone (mg/kg)
26.9
61.7
28.9
Fd Exiting Fraction Dissolved
0.01808
0.00459
0.00729
Fp Exiting Fraction Particulate
0.982
0.995
0.993
NET DREDGING Contribution
Ckmnet Net Total Cone (ng/L)
6050
15890
7426
Cdnet Net Dissolved Cone (ng/L)
1.8
4.88
2.07
Cs net Net Suspended Cone (ng/L)
6048
15885
7424
Cp net Net Particle Cone (mg/kg)
27.1
62.2
29.1
TSSnet Net TSS Cone (mg/L)
223
255
255
Fd,net Net Fraction Dissolved
2.98E-04
3.07E-04
2.79E-04
Fp,net Net Fraction Particulate
0.9997
0.9997
0.9997
Fsilt,net Net Fraction Silt Exiting
0.66
0.76
0.76
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 7
Summary of Sensitivity of Model Outputs to Model Parameter Inputs
Sensitivity Coefficient (S)
Input Parameter
Range of Values
Model Default Value
Net Fraction
Dissolved PCBs
Net Fraction Silt
Net PCB Flux
Net TSS Flux
River-wide Volumetric Flow (Velocity &
2000 - 8000 cfs
4000 cfs
0.14
0.16
0.16
0.16
Depth)
0.11
0.11
0.11
0.11
0.1
0.11
0.11
0.11
Velocity (alone)
0.08 - 0.25 m/s
0.131 m/s
0.27
0.22
0.23
0.23
Depth (alone)
0.9-2.3 m
1.88 m
0.73
0.26
0.25
0.25
Near-Field Width
1-100 meters
10 meters
5.34
0.15
0.17
0.17
Resuspension Rate
0.5 - 40 kg/s
1 kg/s
0.25
<0.01
1
1
Sediment Silt Fraction
0-1
0.37 (Section 1)
0.46
0.52
0.47
0.47
0.48 (Sections 2 & 3)
Sediment PCB Concentration
1 - 1000 mg/kg
27 mg/kg (Section 1)
0.62
<0.01
1
<0.01
62 mg/kg (Section 2)
0.33
<0.01
1
<0.01
29 mg/kg (Section 3)
0.28
<0.01
1
<0.01
Dissolved Fraction in Background (& TSS
0.15 -1
0.9 (Sections 1 & 2)
0.16
<0.01
0.11
<0.01
Concentration in Background)1
0.92 (Section 3)
Partition Coefficient (& PCB Dissolved
5E3 - 5E5 L/kg
4.8E4 (Sections 1 & 2)
2.95
<0.01
<0.01
<0.01
Fraction in Background)
5.1E4 (Section 3)
Desorption Rate
1.6E-4 - 0.2 hr"1
0.2 hr"1
1
<0.01
<0.01
<0.01
Total PCB Concentration in Background
0 - 500 ng/L
122 ng/L (Section 1)
0.24
<0.01
<0.01
<0.01
76 ng/L (Section 2)
57 ng/L (Section 3)
Silt Settling Velocity
4.1-9 m/day
6.9 m/day (8E-5 m/s)
0
<0.01
<0.01
<0.01
Coarse Settling Velocity
0.03 - 0.08 m/s
0.06 m/s
0.25
0.26
0.27
0.27
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 8
TSS-Chem Model Runs for the PCB 350 ng/L far-field Criterion
with and without Dissolved PCBs from Dredging as Modeled by CSTR-Chem
River
Section
Year
Dissolved
PCBs from
dredging
(ng/L)
g (source
strength)
(kg/s)
SS Flux
(1 mile)
(kg/day)
TPCB Flux
(1 mile)
(g/day)
Fraction
Dissolved
(unitless)
Section 1
Section 1
2007
2007
0
1.89
3.052
3.052
60,593
60,593
1,684
1,684
0.09
0.09
Section 2
Section 2
2009
2009
0
5.06
1.669
1.669
37,841
37,841
2,466
2,466
0.14
0.14
Table 9
TSS-Chem Model Runs for the PCB 350 ng/L far-field Criterion
with and without Coarse solids from Dredging as Modeled by CSTR-Chem
Sediment
CSTR-Chem
Silt Fraction
TSS-Chem
Silt
Resuspension
from
source
Silt source
SS Flux
TPCB Flux
Fraction
Fraction
Rate
dredging
strength
strength
(1 mile)
(1 mile)
Dissolved
liver Sectio:
Year
(unitless)
(kg/s)
(unitless)
(kg/s)
(kg/s)
(kg/day)
(g/day)
(unitless)
Section 1
2007
0.37
5.6
0.66
3.1
2.0
60,593
1,684
0.09
Section 1
2007
1
2.0
1
2.0
2.0
60,609
1,684
0.09
Section 2
2009
0.48
2.7
0.76
1.7
1.3
37,841
2,466
0.14
Section 2
2009
1
1.3
1
1.3
1.3
37,847
2,466
0.14
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 10
Results for Average Source Strength Estimated Fluxes
INPUT
TSS-Chem RESULTS
PERCENT LOSS
Net Total PCB
Net Fraction
Concentration
PCB Production
Sediment
SS Silt Source
Net TSS Flux at
Flux at 1 mile
Dissolved PCBs
increase at 1
SS Loss at
PCB Loss
rate
production rate
Silt Fraction
Strength (1,2)
1 mile (2)
(2)
at 1 mile
mile
1 mile
at 1 mile
ks> PCB/dav
ksi solids/day
(kg/s)
(kg/day)
(g/day)
unilless
(ng/1)
River Section
Section 1
57
2,099,921
0.37
0.077
2,303
78
0.35
14
0.11
0.14
Section 2
116
1,857,493
0.48
0.088
2,642
209
0.39
37
0.14
0.18
Section 3
45
1,563,927
0.48
0.074
2,225
81
0.40
14
0.14
0.18
Notes:
1. Source strengths apply to silt and finer particles only
2. Production rates are based on 7 days/week, 14 hours per day, 630 days in Section 1 and 210 days each in River Sections 2 & 3.
3. Values are based on river-wide volumetric flow of 4000 cfs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 11
Increase in PCB Mass from Settled Material Estimated Using the TSS-Chem Model Results
Increase in PCB Mass from Settled
Length Weighted Average
Material (g/sq. m)
Concentration (ppm)
Management
Condition at Far Field Station
River
Target
Sides of
2-Acres
Target
Sides of
2-Acres
Level
Section
Area
T arget Area
Below the
Area
Target Area
Below the
Target Area
Target Area
Evaluation
300 g/day PCB Mass Loss
1
0.9
6E-04
0.2
7.0
1.0
2.6
Control
600 g/day PCB Mass Loss
1
1.8
1E-03
0.5
12
1.0
4.2
Control
350 ng/L
1
3.9
3E-03
1.0
14
1.0
6.6
Evaluation
300 g/day PCB Mass Loss
2
0.6
4E-04
0.1
5.0
1.0
2.0
Control
600 g/day PCB Mass Loss
2
1.2
8E-04
0.3
10
1.0
3.3
Control
350 ng/L
2
4.7
3E-03
1.2
29
1.0
9.1
Evaluation
300 g/day PCB Mass Loss
3
0.6
4E-04
0.2
5.5
1.0
2.2
Control
600 g/day PCB Mass Loss
3
1.4
9E-04
0.4
10
1.0
3.5
Control
350 ng/L
3
5.6
4E-03
1.5
15
1.0
8.6
1. Mass/Area used to define the lateral extent of dredging in River Sections 1 and 2 is approximately 6.6 g/sq. m and 34 g/sq. m,
respectively. In River Section 3, a mass/area was not used to select the areas in this way.
2. The length weighted average concentration was calculated assuming the concentration below the deposited PCBs is 1 ppm.
Table 12
TSS Average Concentration within the Plume at
300 m Downstream and under 8000 cfs Flow
Management Levels
River
Section 1
River
Section 2
River
Section 3
350 ng/L
94
54
110
600 g/day
23
11
22
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 13
Average Source Strength Estimated Fluxes and Concentrations for River Section 1
with Various Flows and Total PCB Sediment Concentrations
INPUT
TSS-Chem RESULTS
PERCENT LOSS
Silt
Net
Sediment
TSS Silt
Net Total
Fraction
Concentrat
PCB
Source
Net TSS
PCB Flux
Dissolved
ion
Concentrat
Silt
Strength
Flux at 1
at 1 mile
PCBs at 1
increase at
TSS Loss
PCB Loss
ion
Fraction
(1.2)
mile (2)
(2)
mile
1 mile
at 1 mile
at 1 mile
(mg/kg)
unitless
(kg/s)
(kg/day)
(g/day)
unitless
(ng/1)
%
%
4000 cfs
27
0.37
0.077
2,303
78
0.35
14
0.11
0.14
30
0.37
0.077
2,303
87
0.36
15
0.11
0.15
36
0.37
0.077
2,303
105
0.37
18
0.11
0.18
2000 cfs
27
0.37
0.077
671
39
0.55
14
0.03
0.07
30
0.37
0.077
671
44
0.56
15
0.03
0.08
36
0.37
0.077
671
53
0.57
19
0.03
0.09
5000 cfs
27
0.37
0.077
2,721
86
0.27
12
0.13
0.15
30
0.37
0.077
2,721
95
0.28
13
0.13
0.17
36
0.37
0.077
2,721
115
0.28
16
0.13
0.20
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 14
Range of Values and Relative Sensitivities of Each Parameter
Relative Model Sensitivity
Input parameter
Range of Values
Default Value
Net Fraction Dissolved
PCBs at 1 mile
Distance where
coarse <0.1%
Net PCB Flux
at 1 mile
Net TSS Flux
at 1 mile
River-wide Volumetric Flow (Velocity &
Depth)
Q
2000-8000 cfs
4000 cfs
moderate
low
moderate
low
Velocity (alone)
u
0.08-0.25 m/s
0.131 m/s
moderate
moderate
moderate
low
Depth (alone)
h
0.9-2.3 m
1.88 m
low
moderate
moderate
moderate
Source Strength
g
0.01-40 kg/s
1 kg/s
moderate (high at low
values of source strength)
none
high
high
Silt Fraction Entering
fsilt,sed
0-1
0.66 (Section 1)
moderate
low
high
high
Sediment PCB Concentration
Csed
1-1000 mg/kg
27 mg/kg (Section 1)
high (low at high
concentrations)
none
high
none
Dissolved Fraction in Background (& TSS
Concentration in Background)1
fd,bkg
0.31-0.97
0.9 (Sections 1)
low
none
low
none
Partition Coefficient (& PCB Dissolved
Fraction in Background)2
Kd
5E3-5E5 L/kg
4.8E4 (Sections 1)
high
none
low
none
De sorption Rate
X
1.6E-4 to 0.2 hr"1
0.2 hr"1
high
none
low
none
Lateral Dispersion (alone)
k(y)
1E-4 to 1E2
0.014 m2/s
low (high at low
coefficients)
none
low
low
Total PCB Concentration in Background
PCB(bkg)
0-500 ng/L
122 ng/L (Section 1)
low
none
low
none
Silt Settling Velocity
w(silt)
4.1-9 m/day
6.9 m/day (8E-5 m/s)
low
none
moderate
moderate
Coarse Settling Velocity
w(coarse)
0.03-0.08 m/s
0.06 m/s
low
high
low
none
Notes:
1. The dissolved PCB fraction in the background and the TSS concentration were varied, with Kd held constant at 5,500 L/kg.
2. The partition coefficient (Kd) and PCB dissolved fraction in the background was varied with TSS background concentration
held constant at 2.3 mg/L.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 15
Effect on Model Output Values from Increase in Input Paramters
Input parameter
Effect on Net
Fraction Dissolved
PCBs at 1 mile
Effect on distance
where coarse <
0.1%
Effect on Net PCB
Flux at 1 mile
Effect on Net TSS
Flux at 1 mile
River-wide Volumetric Flow (Velocity,
Depth and Lateral Dispersion)
Q
Varies
Varies
Varies
Varies
Velocity (alone)
u
Decrease
Increase (linear)
Increase
Increase
Depth (alone)
h
Increase
Increase (linear)
Increase
Increase
Source Strength
8
Decrease
No Effect
Increase (linear)
Increase (linear)
Silt Fraction Entering
fsilt,sed
Decrease
Decrease
Increase (linear)
Increase (linear)
Sediment PCB Concentration
Csed
Increase
No Effect
Increase (linear)
No Effect
Dissolved Fraction in Background (& TSS
Concentration in Background)1
fd,bkg
Increase
No Effect
Decrease
No Effect
Partition Coefficient (& PCB Dissolved
Fraction in Background)2
Kd
Decrease
No Effect
Decrease
No Effect
Desorption Rate
X
Increase
No Effect
Increase
No Effect
Lateral Dispersion (alone)
k(y)
Increase
No Effect
Increase
No Effect
Total PCB Concentration in Background
PCB(bkg)
Decrease (linear)
No Effect
Decrease (linear)
No Effect
Silt Settling Velocity
w(silt)
Increase (linear)
Increase
Decrease
Decrease
Coarse Settling Velocity
w(coarse)
No Effect
Decrease
No Effect
No Effect
Notes:
1. The dissolved PCB fraction in the background and the TSS concentration were varied, with Kd held constant at 5,500 L/kg.
2. The partiton coefficient (Kd) and PCB dissolved fraction in the background was varied with TSS background concentration held constant a
3. Due to the stepwise characteristic of the model (particularly with the distance to 0.1% coarse material), linearity was defined as an r-squarei
greater than 99%.
Table 16
Average Sensitivity Values and Individual magnitudes
Average Spa^^eroutput
Input parameter
Net Fraction
Dissolved PCBs at 1
mile
Distance where
coarse <0.1%
Net PCB Flux at 1
mile
Net TSS Flux at 1
mile
River-wide Volumetric Flow (Velocity &
Depth)
Q
-0.51 (-)
0.69 (+/-)
0.32 (+/-)
0.37 (+/-)
Velocity (alone)
u
-0.71 (-)
0.97 (+)
0.43 (+)
0.52 (+)
Depth (alone)
h
0.17 (+)
1.07 (+)
0.57 (+)
0.61 (+)
Source Strength
8
-0.49 (-)
0
0.96 (+)
1(+)
Silt Fraction Entering
fsilt,sed
-0.71 (-)
-0.72 (-)
0.96 (+)
1(+)
Sediment PCB Concentration
Csed
0.9 (+)
0
1.02 (+)
0
Dissolved Fraction in Background (& TSS
Concentration in Background)1
fd,bkg
0.27 (+)
0
-0.09 (+/-)
0
Partition Coefficient (& PCB Dissolved
Fraction in Background)2
Kd
-0.93 (-)
0
-0.05 (-)
0
Desorption Rate
X
0.87 (+)
0
0.03 (+)
0
Lateral Dispersion (alone)
k(y)
0.2 (+)
0
0.02 (+)
-5.44E-17 (+/-)
Total PCB Concentration in Background
PCB(bkg)
-0.23 (-)
0
-0.02 (-)
0
Silt Settling Velocity
w(silt)
0.33 (+)
0
-0.45 (-)
-0.53 (-)
Coarse Settling Velocity
w(coarse)
-0.0002 (-)
-1.25 (-)
-0.0009 (-)
0
Notes:
1. The dissolved PCB fraction in the background and the TSS concentration were varied, with Kd held constant at 5,500 L/kg.
2. The partition coefficient (Kd) and PCB dissolved fraction in the background was varied with TSS background concentration
held constant at 2.3 mg/L.
3. The sign (+/-) indicates that the individual Sensitivity values were both positive and negative.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 17
Average Baseline Conditions at Thompson Island Dam
Month
Mean flow, q1
Mean baseline concentrations2
Mean baseline Load
(cfs)
TSS (mg/L)
TPCB (ng/L)
(g/sec)
May
7,800
4
128
0.028
June
4,200
5
169
0.020
July
3,000
2
138
0.012
August
3,000
2
96
0.008
September
3,100
2
75
0.007
October
4,300
2
127
0.015
November
5,500
2
127
0.020
June - Nov Average3
3,900
2.3
122
0.014
Notes:
1 Mean flow was estimated based on USGS flow data from 1977 to 2002 at Thompson Island Dam.
2 TSS and TPCB values are arithmetic means obtained from the baseline analysis study. See Attachment A for detail analysis.
3 Only June to November mean baseline concentrations were used. May was excluded since flow is not typical.
Table 18
Average Baseline Conditions at Schuylerville
Month
Mean flow, q1
Mean baseline concentrations2
Mean baseline load
(cfs)
TSS (mg/L)
TPCB (ng/L)
(g/sec)
May
8,800
3
106
0.026
June
4,900
5
106
0.015
July
3,400
2
82
0.008
August
3,400
2
74
0.007
September
3,600
2
52
0.005
October
4,800
2
75
0.010
November
6,200
2
67
0.012
June - Nov Average
4,400
2.3
76
0.009
Notes:
1 Mean flow was estimated based on USGS flow data from 1977 to 2002 at Schuylerville
2 TSS and TPCB values are arithmetic means obtained from the baseline analysis study. See Attachment A for detail analysis.
3 Only June to November mean baseline concentrations were used. May was excluded since flow is not typical.
Table 19
Average Baseline Conditions at Waterford
Month
Mean flow, q1
Mean baseline concentrations2
Mean baseline load
(cfs)
TSS (mg/L)
TPCB (ng/L)
(g/sec)
May
11,300
2
79
0.025
June
6,400
3
79
0.014
July
4,200
1
61
0.007
August
4,000
1
55
0.006
September
4,200
1
39
0.005
October
6,500
1
56
0.010
November
8,300
1
50
0.012
June - Nov Average3
5,600
1.7
57
0.009
Notes:
1 Mean flow was estimated based on USGS flow data from 1977 to 2002 at Waterford
2 TSS and TPCB values were obtained by multiplying a dilution factor based on drainage area ratio.
Drainage areas were obtained from USGS data. Drainage area for Schuylerville and Waterford is 4611 and 3440 ft2, respectively.
3 Only June to November mean baseline concentrations were used. May was excluded since flow is not typical.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 20
Daily Net Dredging Total PCB Flux for River Sections 1, 2, and 3 at the Monitoring Stations
River Sect
on 1 (TID)
River Section 2
(Schuylerville)
River Section
3 (Waterford)
Net Dredge
Net Dredge
Net Dredge
Net Dredge
Net Dredge
Net Dredge
Month
TPCB Flux (14-
TPCB Flux (24-
TPCB Flux
TPCB Flux
TPCB Flux (14-
TPCB Flux (24-
hr basis)
hr basis)
(14-hr basis)
(24-hr basis)
hr basis)
hr basis)
g/day
g/day
g/day
g/day
g/day
g/day
May
2,500
4,200
3,000
5,200
4,400
7,500
June
1,100
1,900
1,700
2,900
2,500
4,200
July
900
1,600
1,300
2,300
1,700
3,000
August
1,100
1,800
1,300
2,300
1,700
2,900
September
1,200
2,100
1,500
2,600
1,900
3,200
October
1,400
2,300
1,900
3,200
2,700
4,700
November
1,700
3,000
2,500
4,300
3,600
6,100
June - Nov Average
1,200
2,100
1,700
2,900
2,300
4,000
Note:
Numbers are rounded to 2 significant digits
Bold italic numbers - values were used as the TPCB flux representing the 350 ng/L at the monitoring stations.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 21
HUDTOX Input for 350 ng/L with TSS Flux at 1 Mile Downstream of the Dredge-Head
Sediment removal
season
Dredging
Location
speed
Cubic
yards of
sediment
removed1
Monitoring
Location
(Compliance
Point)2
Sediment Total
PCB
concentration3
(mg/kg)
Expected Total
PCB export
rate at
compliance
point (g/day)
Percent
remaining at
the monitoring
location5
TPCB input
flux to
HUDTOX
(g/day)
Ratio of
Total to
Tri+ PCB6
Tri+ PCB Flux
input to
HUDTOX7
(g/day)
TSS Flux input
to HUDTOX8
(kg/day)
HUDTOX
Segment(s)
Jun. 1 - Sep. 15, 20069
Sec. 1
half
260,000
TID
27
1,237
75%
1,649
3.2
520
58,800
5 & 7
May 1 - Nov. 30, 2007
Sec. 1
full
520,000
TID
27
1,237
75%
1,649
3.2
520
58,800
11 & 13
May 1 - Nov. 30, 2008
Sec. 1
full
520,000
TID
27
1,237
75%
1,649
3.2
520
58,800
20 & 22
May 1 - Aug. 15, 2009
Sec. 1 &
full
260,000
TID
27
1,237
75%
1,649
3.2
520
58,800
26 & 28
Aug. 16-Nov. 30, 20C
Sec. 2
290,000
Schuylerville
62
2,034
75%
2,712
3.4
670
34,300
30
May 1 - Aug. 15, 2010
Sec. 2 &
full
290,000
Schuylerville
62
2,034
75%
2,712
3.4
670
34,300
31
Aug. 16-Nov. 30, 201
Sec. 3
255,000
Waterford
29
2,334
75%
3,112
2.7
1,150
104,500
38
May 1 - Aug. 15, 2011
Sec. 3
full
255,000
Waterford
29
2,334
75%
3,112
2.7
1,150
104,500
45
Notes:
1 Volume of sediment removed is taken from Table 8-9 of the Feasibility Study.
2 All TIP monitoring is done at TID, all River Section 2 monitoring is done at Schuylerville, and all River Section 3 monitoring is done at Waterford. 1 mile exclusion is not considered.
3 Total PCB concentration in the sediment is for the dredge material and was taken from Table 363334-6 of the Sediment Inventory White Paper of the Resp. Summ.
Total PCB Flux is the average net flux for June to Nov at the compliance point (TID, Schuylerville, & Waterford). PCB flux in May was excluded since flow is not typical.
5 Percent reduction at the monitoring location was obtained from the initial HUDTOX runs performed for the preliminary draft of the resuspension performance standard
6 Ratio of Total to Tri+ PCB is based on the amount of Total PCB and Tri+ PCB removed for each river section (USEPA 2002).
7 Tri+ PCB flux is calculated based on the Total PCB flux 1 mile downstream of the dredgehead divided by the ratio of Total to Tri+ PCB for each section.
8 TSS flux from TSS-Chem model, 1 mile downstream of the dredge-head
9 Actual dredging period is from May 1 - Nov. 30, 2006. The PCB and TSS flux is loaded only from June 1 to Sep. 15, 2006 to account for half speed operation.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 22
TSS Flux Comparisons for Different Scenarios
Sediment removal season
Dredging Location
speed
Cubic yards of
sediment
removed
Monitoring Location
(Compliance Point)1
Full TSS Flux2
(kg/day)
TSS Flux3 @ 1
mile (kg/day)
TSS Flux3 @ 3
mile (kg/day)
TSS Flux4 @ 1
mile with
corrected percent
reduction (kg/day)
Jun. 1 - Sep. 15, 20065
Sec. 1
half
260,000
TID
60,602
58,800
51,200
61,030
May 1 - Nov. 30, 2007
Sec. 1
full
520,000
TID
60,602
58,800
51,200
60,575
May 1 - Nov. 30, 2008
Sec. 1
full
520,000
TID
60,602
58,800
51,200
53,423
May 1 - Aug. 15, 2009
Sec. 1 &
full
260,000
TID
60,602
58,800
51,200
45,599
Aug. 16-Nov. 30, 2009
Sec. 2
290,000
Schuylerville
36,595
34,300
26,500
37,814
May 1 - Aug. 15,2010
Sec. 2 &
full
290,000
Schuylerville
36,595
34,300
26,500
32,242
Aug. 16-Nov. 30, 2010
Sec. 3
255,000
Waterford
107,575
104,500
98,400
106,675
May 1 - Aug. 15,2011
Sec. 3
full
255,000
Waterford
107,575
104,500
98,400
82,308
Notes:
1 All TIP monitoring is done at TID, all River Section 2 monitoring is done at Schuylerville, and all River Section 3
monitoring is done at Waterford. 1 mile exclusion is not considered.
2 TSS flux using the concentrations of the dredged sediment in each section of the river
3 TSS flux is obtained from TSS-Chem model output.
4 TSS flux is obtained from TSS-Chem model output at 1 mile with corrected percent reduction at the monitoring stations.
5 Actual dredging period is from May 1 - Nov. 30, 2006. The PCB and TSS flux is loaded only from June 1 to Sep. 15, 2006
to account for half speed operation.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 23
HUDTOX Input for 350 ng/L with TSS Flux at 1 Mile Downstream of the Dredge-Head and Corrected Percent Reduction at the Monitoring Stations
Sediment removal
season
Dredging
Location
speed
Cubic
yards of
sediment
removed1
Monitoring
Location
(Compliance
Point)2
Sediment Total
PCB
concentration3
(mg/kg)
Expected Total
PCB export
rate at
compliance
point (g/day)
Percent
remaining at
the monitoring
location5
TPCB input
flux to
HUDTOX
(g/day)
Ratio of
Total to
Tri+ PCB6
Tri+ PCB Flux
input to
HUDTOX7
(g/day)
TSS Flux input
to HUDTOX8
(kg/day)
HUDTOX
Segment(s)
Jun. 1 - Sep. 15, 20069
Sec. 1
half
260,000
TID
27
1,237
73%
1,697
3.2
530
61,030
5 & 7
May 1 - Nov. 30, 2007
Sec. 1
full
520,000
TID
27
1,237
73%
1,684
3.2
526
60,575
11 & 13
May 1 - Nov. 30, 2008
Sec. 1
full
520,000
TID
27
1,237
83%
1,490
3.2
466
53,423
20 & 22
May 1 - Aug. 15, 2009
Sec. 1 &
full
260,000
TID
27
1,237
97%
1,278
3.2
399
45,599
26 & 28
Aug. 16-Nov. 30, 20C
Sec. 2
290,000
Schuylerville
62
2,034
82%
2,466
3.4
725
37,814
30
May 1 - Aug. 15, 2010
Sec. 2 &
full
290,000
Schuylerville
62
2,034
96%
2,117
3.4
623
32,242
31
Aug. 16-Nov. 30, 201
Sec. 3
255,000
Waterford
29
2,334
74%
3,150
2.7
1,167
106,675
38
May 1 - Aug. 15, 2011
Sec. 3
full
255,000
Waterford
29
2,334
96%
2,441
2.7
904
82,308
45
Notes:
1 Volume of sediment removed is taken from Table 8-9 of the Feasibility Study.
2 All TIP monitoring is done at TID, all River Section 2 monitoring is done at Schuylerville, and all River Section 3 monitoring is done at Waterford. 1 mile exclusion is not considered.
3 Total PCB concentration in the sediment is for the dredge material and was taken from Table 363334-6 of the Sediment Inventory White Paper of the Resp. Summ.
Total PCB Flux is the average net flux for June to Nov at the compliance point (TID, Schuylerville, & Waterford). PCB flux in May was excluded since flow is not typical.
5 Percent remaining at the monitoring location was obtained from the initial HUDTOX runs performed for the preliminary draft of the resuspension performance standard
6 Ratio of Total to Tri+ PCB is based on the amount of Total PCB and Tri+ PCB removed for each river section (USEPA 2002).
7 Tri+ PCB flux is calculated based on the Total PCB flux 1 mile downstream of the dredgehead divided by the ratio of Total to Tri+ PCB for each section.
8 TSS flux from TSS-Chem model, 1 mile downstream of the dredge-head
9 Actual dredging period is from May 1 - Nov. 30, 2006. The PCB and TSS flux is loaded only from June 1 to Sep. 15, 2006 to account for half speed operation.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 24
HUDTOX Schedule and Input Loading for 300 g/day Export Rate Scenario
River Section
Dredging Season
Monitoring
Station
Tri+ PCB Flux
input to HUDTOX1
(g/day)
TSS Flux input
to HUDTOX2
(kg/day)
TPCB input flux
to HUDTOX3
(g/day)
Percent
remaining at the
monitoring
location4
Expected Total
PCB at monitoring
station5 (g/day)
May 1-Nov 30, 2006
TID
129
13,948
411
73%
300
Section 1
May 1-Nov 30, 2007
TID
128
13,828
408
73%
300
dredging
May 1-Nov 30, 2008
TID
113
12,130
361
83%
300
May 1-Aug 15,2009
TID
97
10,311
310
97%
300
Section 2
Aug 16-Nov 30, 2009
Schuylerville
107
4,873
364
82%
300
dredging
May 1 - Aug 15,2010
Schuylerville
92
4,118
312
96%
300
Section 3
A.ug 16-Nov 30, 2010:
Waterford
150
12,725
405
74%
300
dredging
May 1 - Aug 15,2011
Waterford
116
9,702
314
96%
300
Notes:
1 Tri+ PCB flux is calculated by dividing the TPCB flux with the Total to Tri+ PCB ratio estimated in the RS. The ratio is 3.2 for Section 1, 3.4 for Section 2, and 2.7 for Section 3
2 TSS flux from TSS-Chem model, 1 mile downstream of the dredge-head
3 Total PCB input is based on the expected flux at monitoring locations divide by the percent reduction. Same as Gaussian plume output at 1 mile.
4 Percent remaining at the monitoring location was obtained from the initial HUDTOX runs performed for the preliminary draft of the resuspension performance standard
5 Expected net export rate of TPCB flux at monitoring station (300 g/day).
Table 25
HUDTOX Schedule and Input Loading for 600 g/day Export Rate Scenario
River Section
Dredging Season
Monitoring
Station
Tri+ PCB Flux
input to HUDTOX1
(g/day)
TSS Flux input
to HUDTOX2
(kg/day)
TPCB input flux
to HUDTOX3
(g/day)
Percent
reduction at the
monitoring
location4
Expected Total
PCB at monitoring
station5 (g/day)
May 1-Nov 30, 2006
TID
257
28,975
823
73%
600
Section 1
May 1-Nov 30, 2007
TID
255
28,676
817
73%
600
dredging
May 1-Nov 30, 2008
TID
226
25,179
723
83%
600
May 1-Aug 15,2009
TID
194
21,582
620
97%
600
Section 2
Aug 16-Nov 30, 2009
Schuylerville
214
10,379
728
82%
600
dredging
May 1-Aug 15,2010
Schuylerville
184
8,799
625
96%
600
Section 3
Aug 16-Nov 30, 2010
Waterford
300
26,398
810
74%
600
dredging
May 1-Aug 15,2011
Waterford
232
20,193
627
96%
600
Notes:
1 Tri+ PCB flux is calculated by dividing the TPCB flux with the Total to Tri+ PCB ratio estimated in the RS. The ratio is 3.2 for Section 1, 3.4 for Section 2, and 2.7 for Section 3
2TSS flux from TSS-Chem model, 1 mile downstream of the dredge-head
3Total PCB input is based on the expected flux at monitoring locations divide by the percent reduction. Same as Gaussian plume output at 1 mile.
4 Percent reduction at the monitoring location was obtained from the initial HUDTOX runs performed for the preliminary draft of the resuspension performance standard
5Expected net export rate of TPCB flux at monitoring station (600 g/day).
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment D - April 2004
-------
Table 26
Percent Reduction at the Monitoring Locations Comparison for the 350 ng/L Scenario
River Section
Dredging Season
Monitoring
Station1
d006 percent
remaining2
d007 percent
remaining3
sr03 percent
remaining4
sr04 percent
remaining5
May 1-Nov 30, 2006
TID
73%
74%
82%
73%
Section 1
May 1-Nov 30, 2007
TID
73%
74%
85%
73%
dredging
May 1-Nov 30, 2008
TID
83%
83%
91%
83%
May 1-Aug 15,2009
TID
97%
97%
99%
97%
Section 2
Aug 16 - Nov 30, 2009
Schuylerville
82%
84%
92%
83%
dredging
May 1 - Aug 15,2010
Schuylerville
96%
97%
99%
96%
Section 3
Aug 16-Nov 30, 2010
Waterford
74%
75%
85%
71%
dredging
May 1 - Aug 15,2011
Waterford
96%
96%
99%
95%
Notes:
1 All TIP monitoring is done at TID, all River Section 2 monitoring is done at Schuylerville, and all River Section 3 monitoring is done
at Waterford. 1 mile exclusion is not considered.
2 d006 is the 350 ng/L model run with TSS and TPCB flux at 1 mile downstream of the dredge-head obtained from TSS-Chem.
3 d007 is the 350 ng/L model run with TSS and TPCB flux at 3 mile downstream of the dredge-head obtained from TSS-Chem.
4 sr03 is the 350 ng/L model run without any TSS flux associated with the TPCB flux.
5 sr04 is the 350 ng/L model with TSS and TPCB flux at 1 mile downstream of the dredge-head and corrected percent reduction.
Table 27
Expected versus Model Prediction of PCB Flux for Control Level 3 - 350 ng/L Scenario
Total PCB @ monitoring station
River Section
Dredging Season
Monitoring
Station
Expected
(g/day)2
d006 - model
estimate3
(g/day)
d007 - model
estimate4
(g/day)
sr03 - model
estimate5
(g/day)
sr04 -
model
estimate6
(g/day)
May 1-Nov 30, 2006
TID
1237
1213
1224
1360
1234
Section 1
May 1-Nov 30, 2007
TID
1237
1222
1233
1410
1244
dredging
May 1-Nov 30, 2008
TID
1237
1381
1389
1519
1252
May 1-Aug 15,2009
TID
1237
1611
1615
1653
1245
Section 2
Aug 16 - Nov 30, 2009
Schuylerville
2034
1879
1909
2097
2049
dredging
May 1 - Aug 15,2010
Schuylerville
2034
2189
2200
2261
2029
Section 3
Aug 16-Nov 30, 2010
Waterford
2334
2276
2290
2619
2223
dredging
May 1 - Aug 15,2011
Waterford
2334
2969
2974
3083
2302
Notes:
'Output loading from HUDTOX d006 run at the assigned monitoring station.
2Total PCB flux at the monitoring station based on max concentration of 350 ng/L minus baseline concentrations.
3 d006 is the 350 ng/L model run with TSS and TPCB flux at 1 mile downstream of the dredge-head obtained from TSS-Chem.
4 d007 is the 350 ng/L model run with TSS and TPCB flux at 3 mile downstream of the dredge-head obtained from TSS-Chem.
5 sr03 is the 350 ng/L model run without any TSS flux associated with the TPCB flux.
6 sr04 is the 350 ng/L model with TSS and TPCB flux at 1 mile downstream of the dredge-head and corrected percent reduction.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 28
Annual Tri+ PCB Load Over TID
Year
No Resuspension
Total PCB 350 ng/L
Total PCB 350 ng/L
Total PCB 350 ng/L
fraction remaining
Total PCB 350 ng/L
with no Solids
Total PCB
Total PCB 300
Accidental Release
MNA (p3nas2)
(d004)
@ 1 mile (d006)
@ 3 mile (d007)
adjusted (sr04)
(sr03)
600g/day (sr01)
g/day (sr02)
(srA1)
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
2005
0
0
0
0
0
0
0
0
0
2006
57
56
97
97
98
102
77
67
77
2007
114
106
228
230
231
246
169
138
169
2008
152
133
348
349
341
374
237
185
237
2009
190
154
423
425
405
452
279
217
279
2010
241
180
450
452
431
478
305
244
305
2011
284
203
474
475
455
501
328
266
328
2012
325
224
495
497
477
523
350
288
350
2013
365
246
517
519
498
545
371
309
371
2014
398
264
536
538
517
564
390
328
390
2015
429
282
554
556
535
582
408
346
408
2016
454
297
569
571
550
597
423
361
423
2017
476
311
583
586
564
612
437
375
437
2018
503
327
599
601
580
627
453
391
453
2019
524
340
612
614
593
641
466
404
466
2020
546
354
626
629
607
655
480
418
480
2021
567
368
640
642
621
494
494
2022
584
380
652
655
633
506
506
2023
601
392
664
666
644
518
518
2024
622
405
677
680
658
531
531
2025
639
417
689
692
670
543
543
2026
656
429
701
704
682
555
555
2027
671
440
712
715
693
566
566
2028
686
452
724
727
705
578
578
2029
702
463
735
738
716
589
589
2030
716
475
747
750
728
601
601
2031
732
486
758
761
739
612
2032
747
497
769
772
750
623
2033
760
508
780
783
761
634
2034
774
519
791
794
771
645
2035
787
529
801
804
782
656
2036
801
540
812
815
793
666
2037
814
551
823
826
803
677
2038
826
561
832
836
813
687
2039
841
571
843
846
824
698
2040
852
581
853
856
834
707
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 28
Annual Tri+ PCB Load Over TID
Year
MNA (p3nas2)
No Resuspension
(d004)
Total PCB 350 ng/L
@ 1 mile (d006)
Total PCB 350 ng/L
@ 3 mile (d007)
Total PCB 350 ng/L
fraction remaining
adjusted (sr04)
Total PCB 350 ng/L
with no Solids
(sr03)
Total PCB
600g/day (sr01)
Total PCB 300
g/day (sr02)
Accidental Release
(srA1)
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
2041
864
591
863
866
844
717
2042
874
600
872
875
853
726
2043
887
611
882
886
863
737
2044
899
621
893
896
873
747
2045
911
631
902
906
883
757
2046
921
640
912
915
893
766
2047
932
649
921
924
902
776
2048
944
659
930
934
911
785
2049
955
668
939
943
920
794
2050
967
677
949
952
930
804
2051
979
687
959
962
940
813
2052
989
696
968
971
949
822
2053
999
705
976
980
957
831
2054
1009
714
985
988
966
840
2055
1019
723
995
998
975
849
2056
1028
731
1003
1006
984
858
2057
1038
740
1012
1015
993
867
2058
1047
749
1021
1024
1002
876
2059
1057
758
1030
1033
1010
884
2060
1067
767
1039
1042
1020
894
2061
1078
777
1049
1052
1030
904
2062
1087
786
1057
1061
1038
912
2063
1096
794
1066
1069
1047
921
2064
1105
803
1075
1078
1056
930
2065
1114
812
1084
1087
1065
939
2066
1123
821
1092
1096
1073
947
2067
1132
829
1101
1104
1081
956
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 29
Tri+ PCB Load Over Schuylerville
Year
Total PCB 350 ng/L
Total PCB 350 ng/L
Total PCB 600
No Resuspension
Total PCB 350 ng/L
Total PCB 350 ng/L
fraction remaining
with no Solids
Total PCB
Total PCB 300
Accidental Release
g/day corrected to
MNA (p3nas2)
(d004)
@ 1 mile (d006)
@ 3 mile (d007)
adjusted (sr04)
(sr03)
600g/day (sr01)
g/day (sr02)
(srA1)
MNA (sr01)
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
2005
0
0
0
0
0
0
0
0
0
0
2006
78
77
110
111
111
117
94
86
94
94
2007
162
155
256
258
258
280
208
183
208
208
2008
207
190
362
365
357
404
276
234
276
276
2009
253
221
496
501
488
551
344
285
344
359
2010
327
263
610
615
596
667
405
337
405
442
2011
390
291
640
645
626
697
434
365
434
471
2012
444
316
668
673
654
723
460
390
460
496
2013
499
341
695
701
681
750
485
416
485
522
2014
540
361
717
723
703
772
507
437
507
543
2015
578
381
738
744
723
793
527
457
527
564
2016
607
397
755
761
740
809
543
473
543
580
2017
632
412
770
776
755
825
558
488
558
595
2018
666
429
788
794
773
843
575
505
575
612
2019
690
443
802
808
787
857
589
519
589
626
2020
717
458
818
824
803
873
604
534
604
641
2021
742
472
832
839
817
619
619
655
2022
761
485
845
851
830
631
631
668
2023
779
496
857
863
842
643
643
679
2024
804
511
872
878
857
658
658
694
2025
824
523
884
891
869
670
670
707
2026
843
536
897
904
882
682
682
719
2027
859
547
908
915
893
693
693
730
2028
877
559
920
927
905
705
705
742
2029
894
570
932
938
917
717
717
754
2030
910
582
943
950
928
728
728
765
2031
929
594
955
962
940
741
777
2032
945
605
967
974
952
752
789
2033
959
616
977
984
962
762
799
2034
974
627
988
995
973
773
810
2035
988
638
999
1006
984
784
821
2036
1003
649
1010
1017
995
795
832
2037
1018
659
1021
1028
1006
806
843
2038
1030
669
1031
1038
1016
816
853
2039
1046
680
1042
1049
1027
827
864
2040
1058
690
1052
1059
1037
837
873
2041
1070
700
1062
1069
1047
846
883
2042
1079
708
1070
1077
1055
855
891
2043
1093
719
1081
1088
1066
866
902
2044
1106
730
1091
1099
1076
876
913
2045
1119
739
1101
1108
1086
886
923
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 29
Tri+ PCB Load Over Schuylerville
Year
Total PCB 350 ng/L
Total PCB 350 ng/L
Total PCB 600
No Resuspension
Total PCB 350 ng/L
Total PCB 350 ng/L
fraction remaining
with no Solids
Total PCB
Total PCB 300
Accidental Release
g/day corrected to
MNA (p3nas2)
(d004)
@ 1 mile (d006)
@ 3 mile (d007)
adjusted (sr04)
(sr03)
600g/day (sr01)
g/day (sr02)
(srA1)
MNA (sr01)
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
2046
1130
749
1111
1118
1096
896
932
2047
1140
758
1120
1127
1105
905
942
2048
1152
767
1129
1136
1114
914
951
2049
1163
776
1138
1145
1123
923
960
2050
1175
786
1147
1155
1132
932
969
2051
1188
795
1157
1164
1142
942
979
2052
1198
804
1166
1173
1151
951
988
2053
1208
812
1174
1181
1159
959
996
2054
1217
821
1183
1190
1168
968
1005
2055
1228
830
1192
1199
1177
978
1014
2056
1237
838
1200
1207
1185
985
1022
2057
1247
847
1209
1216
1194
994
1031
2058
1256
855
1217
1224
1202
1003
1039
2059
1265
864
1226
1233
1211
1011
1048
2060
1275
873
1235
1242
1220
1021
1057
2061
1286
883
1245
1252
1230
1031
1067
2062
1295
892
1253
1261
1238
1039
1076
2063
1304
900
1262
1269
1247
1047
1084
2064
1313
908
1270
1277
1255
1056
1092
2065
1322
917
1279
1286
1264
1064
1101
2066
1331
925
1287
1294
1272
1073
1109
2067
1339
933
1295
1302
1280
1081
1117
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 30
Tri+ PCB Load Over Waterford
Total PCB 350 ng/L
Total PCB 350 ng/L
Total PCB 600
Year
No Resuspension
Total PCB 350 ng/L
Total PCB 350 ng/L
fraction remaining
with no Solids
Total PCB
Total PCB 300
Accidental Release
g/day corrected to
MNA (p3nas2)
(d004)
@ 1 mile (d006)
@ 3 mile (d007)
adjusted (sr04)
(sr03)
600g/day (sr01)
g/day (sr02)
(srA1)
MNA (sr01)
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
2005
0
0
0
0
0
0
0
0
0
0
2006
102
102
116
117
117
122
110
106
110
110
2007
205
201
250
251
251
266
227
214
227
227
2008
254
245
325
327
324
352
287
267
287
287
2009
301
285
404
408
401
445
342
315
342
349
2010
393
353
607
612
601
664
451
404
451
484
2011
464
397
782
788
754
843
524
463
535
580
2012
528
437
848
854
818
908
572
508
584
628
2013
595
478
906
912
875
967
618
551
631
674
2014
643
508
949
955
917
1010
652
583
665
708
2015
687
536
987
993
954
1047
683
612
696
738
2016
714
553
1010
1017
977
1069
702
631
715
757
2017
738
569
1032
1038
998
1090
719
648
733
775
2018
771
588
1055
1061
1021
1113
739
667
753
795
2019
793
602
1072
1079
1039
1130
754
681
768
810
2020
821
620
1094
1100
1059
1151
772
699
786
828
2021
847
636
1112
1119
1078
789
803
845
2022
865
648
1127
1133
1092
802
816
858
2023
882
659
1140
1146
1105
813
827
869
2024
911
677
1160
1166
1125
832
846
888
2025
930
689
1174
1180
1139
845
859
901
2026
949
702
1188
1194
1153
858
872
913
2027
964
712
1199
1205
1164
868
882
924
2028
982
724
1211
1218
1177
880
894
936
2029
999
736
1224
1230
1189
892
906
948
2030
1015
747
1236
1242
1201
903
917
959
2031
1033
759
1248
1255
1213
916
972
2032
1048
769
1259
1266
1224
926
982
2033
1061
779
1269
1276
1234
936
992
2034
1077
790
1281
1287
1246
947
1003
2035
1100
809
1292
1298
1257
958
1014
2036
1134
839
1303
1310
1268
970
1026
2037
1164
864
1316
1324
1281
1001
1057
2038
1185
882
1341
1349
1307
1023
1079
2039
1212
905
1372
1380
1338
1050
1106
2040
1228
919
1391
1399
1357
1067
1123
2041
1243
932
1408
1416
1374
1082
1138
2042
1253
941
1420
1428
1385
1093
1149
2043
1272
958
1440
1447
1405
1111
1166
2044
1292
974
1457
1465
1423
1128
1184
2045
1308
987
1471
1479
1437
1141
1197
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 30
Tri+ PCB Load Over Waterford
Year
Total PCB 350 ng/L
Total PCB 350 ng/L
Total PCB 600
No Resuspension
Total PCB 350 ng/L
Total PCB 350 ng/L
fraction remaining
with no Solids
Total PCB
Total PCB 300
Accidental Release
g/day corrected to
MNA (p3nas2)
(d004)
@ 1 mile (d006)
@ 3 mile (d007)
adjusted (sr04)
(sr03)
600g/day (sr01)
g/day (sr02)
(srA1)
MNA (sr01)
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
Cumulative Load
2046
1322
1000
1484
1492
1450
1154
1210
2047
1334
1010
1496
1503
1461
1165
1221
2048
1346
1020
1507
1514
1472
1176
1232
2049
1356
1028
1516
1523
1481
1185
1241
2050
1369
1039
1527
1535
1492
1195
1251
2051
1382
1049
1539
1546
1504
1207
1262
2052
1392
1057
1547
1555
1513
1215
1271
2053
1400
1065
1555
1562
1520
1222
1278
2054
1409
1072
1563
1570
1528
1230
1286
2055
1419
1081
1572
1579
1537
1239
1295
2056
1426
1087
1579
1586
1544
1245
1301
2057
1435
1095
1587
1594
1552
1254
1310
2058
1443
1103
1595
1602
1560
1261
1317
2059
1451
1110
1602
1609
1567
1269
1325
2060
1462
1120
1612
1619
1577
1278
1334
2061
1473
1130
1622
1629
1587
1289
1345
2062
1481
1137
1629
1636
1594
1296
1352
2063
1488
1144
1636
1643
1601
1303
1359
2064
1495
1151
1643
1650
1608
1310
1366
2065
1503
1158
1650
1658
1616
1317
1373
2066
1510
1165
1658
1665
1623
1324
1380
2067
1517
1172
1664
1671
1629
1331
1387
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 31
Resuspension Production, Release, and Export Rates from TSS-Chem and HUDTOX Models
TPCB Flux at
Monitoring
TPCB flux at 1
Stations from
Resuspension
Resuspension
Resuspension
Net TSS Flux at 1
mile
HUDTOX4
Solids
Source Strength as
Export Rate as
Dredging Location and
Production Rate
Production Rate
mile from TSS-
(Resuspension
(Resuspension
TPCB
Production
Percentage of
Percentage of
Scenario
Sediment Removal Period
Monitoring Station
of Sediment1
of Total PCB2
Chem
Release Rate)
Export Rate)
Production Rate
Rate7
TPCB Removed8
TPCB Removed9
(kg/s)
(g/day)
(kg/s)
(g/day)
(g/day)
(g/day)
(kg/s)
(%)
(%)
May 1 - November 30, 2006
Section 1, TID
1.3
1,700
0.28
410
320
5.7.E+04
42
3%
0.56%
Evaluation
May 1 - November 30, 2007
Section 1, TID
1.3
1,700
0.27
410
320
5.7.E+04
42
3%
0.56%
Level - 300
May 1 - November 30, 2008
Section 1, TID
1.1
1,500
0.24
360
300
5.7.E+04
42
3%
0.53%
g/day TPCB
May 1 - August 15, 2009
Section 1, TID
0.9
1,300
0.20
310
310
5.7.E+04
42
2%
0.54%
Flux at
August 16 - November 30, 200'
Section 2, Schuylerville
0.3
1,100
0.10
360
330
1.2.E+05
37
1%
0.29%
Monitoring
May 1 - August 15, 2010
Section 2, Schuylerville
0.3
900
0.08
310
300
1.2.E+05
37
1%
0.26%
Stations
August 16 - November 30, 2011
Section 3, Waterford
0.9
1,300
0.25
400
340
4.5.E+04
31
3%
0.75%
May 1 - August 15, 2011
Section 3, Waterford
0.7
1,000
0.19
310
340
4.5.E+04
31
2%
0.75%
May 1 - November 30, 2006
Section 1, TID
2.6
3,600
0.57
820
620
5.7.E+04
42
6%
1.1%
Control Level -
600 g/day
TPCB Flux at
Monitoring
Stations
May 1 - November 30, 2007
May 1 - November 30, 2008
May 1 - August 15, 2009
Section 1, TID
Section 1, TID
Section 1, TID
2.6
2.3
2.0
3,600
3,100
2,700
0.57
0.50
0.43
820
720
620
630
620
590
5.7.E+04
5.7.E+04
5.7.E+04
42
42
42
6%
6%
5%
1.1%
1.1%
1.0%
August 16 - November 30, 200'
Section 2, Schuylerville
0.7
2,300
0.21
730
620
1.2.E+05
37
2%
0.5%
May 1 - August 15, 2010
Section 2, Schuylerville
0.6
1,900
0.17
630
590
1.2.E+05
37
2%
0.5%
August 16 - November 30, 2011
Section 3, Waterford
1.9
2,700
0.52
810
660
4.5.E+04
31
6%
1.5%
May 1 - August 15, 2011
Section 3, Waterford
1.4
2,100
0.40
630
650
4.5.E+04
31
5%
1.4%
May 1 - November 30, 2006
Section 1, TID
5.6
7,600
1.2
1,700
1,200
5.7.E+04
42
13%
2.1%
Control Level -
May 1 - November 30, 2007
Section 1, TID
5.6
7,600
1.2
1,700
1,200
5.7.E+04
42
13%
2.1%
350 ng/L
May 1 - November 30, 2008
Section 1, TID
4.9
6,700
1.1
1,500
1,300
5.7.E+04
42
12%
2.3%
TPCB
May 1 - August 15, 2009
Section 1, TID
4.2
5,700
0.91
1,300
1,200
5.7.E+04
42
10%
2.1%
Concentrations
August 16 - November 30, 200'
Section 2, Schuylerville
2.7
8,300
0.75
2,500
2,000
1.2.E+05
37
7%
1.7%
at Monitoring
May 1 - August 15, 2010
Section 2, Schuylerville
2.3
7,100
0.64
2,100
2,000
1.2.E+05
37
6%
1.7%
Stations
August 16 - November 30, 2011
Section 3, Waterford
7.5
10,900
2.1
3,100
2,200
4.5.E+04
31
24%
4.9%
May 1 - August 15, 2011
Section 3, Waterford
5.8
8,400
1.6
2,400
2,300
4.5.E+04
31
19%
5.1%
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 31
Resuspension Production, Release, and Export Rates from TSS-Chem and HUDTOX Models
TPCB Flux at
TPCB flux at 1
Monitoring
Resuspension
Resuspension
Resuspension
Net TSS Flux at 1
mile
Stations10
Solids
Source Strength as
Export Rate as
Dredging Location and
Production Rate
Production Rate
mile from TSS-
(Resuspension
(Resuspension
TPCB
Production
Percentage of
Percentage of
Scenario
Sediment Removal Period
Monitoring Station
of Sediment1
of Total PCB2
Chem
Release Rate)
Export Rate)
Production Rate
Rate7
TPCB Removed8
TPCB Removed9
(kg/s)
(g/day)
(kg/s)
(g/day)
(g/day)
(g/day)
(kg/s)
(%)
(%)
May 1 - November 30, 2000
Section 1, TIL)
9.4
12,800
2.0
2,800
2,100
5.7.E 04
42
23".j
3.7".j
Resuspension
May 1 - November 30, 2007
Section 1, TID
9.3
12,700
2.0
2,800
2,100
5.7.E+04
42
22%
3.7%
Standard - 500
May 1 - November 30, 2008
Section 1, TID
8.2
11,200
1.8
2,500
2,100
5.7.E+04
42
20%
3.7%
ng/L TPCB
May 1 - August 15, 2009
Section 1, TID
7.1
9,600
1.53
2,100
2,100
5.7.E+04
42
17%
3.7%
Concentrations
August 16 - November 30, 200'
Section 2, Schuylerville
3.5
10,900
0.99
3,200
2,700
1.2.E+05
37
9%
2.3%
at Monitoring
May 1 - August 15, 2010
Section 2, Schuylerville
3.0
9,300
0.84
2,800
2,700
1.2.E+05
37
8%
2.3%
Stations
August 16 - November 30, 2011
Section 3, Waterford
11
16,600
3.2
4,800
3,500
4.5.E+04
31
37%
7.7%
May 1 - August 15, 2011
Section 3, Waterford
8.8
12,800
2.5
3,700
3,500
4.5.E+04
31
28%
7.7%
Notes:
Numbers are rounded to 2 significant digits.
Source strength represents the amount of solids being suspended to the water column at the dredge-head in kg/s. The value is obtained from the TSS-Chem model.
TPCB flux for source strength is obtained by multiplying the solids source strength with the TPCB concentration in the sediment. The TPCB concentration for River Sections 1, 2, and 3 is 27, 62, and 29 mg/kg, respectively.
Net TSS flux is the TSS-Chem model result at a distance 1 mile downstream of the dredge-head.This number is also the TSS flux input to the HUDTOX model.
Values represent the amount of TPCB flux at the monitoring stations as predicted by HUDTOX.
TPCB flux is obtained from TSS-Chem model. It is the TPCB flux at 1 mile downstream of the dredge-head. This is also the input TPCB flux to the HUDTOX model.
TPCB production rate based on the total TPCB being removed in each river section (36,000 kg, 24,300 kg, and 9,500 kg of TPCB for River Sections 1, 2, and 3, respectively);
assuming 7days/week, 14 hours/day, 630 days in River Section 1 and 210 days each in River Sections 2 and 3.
Solids production rate based on the total sediment being removed including overcut (1.5x10A6 cy, 5.8x10A5 cy, and 5.1x10A5 cy of solids in River Sections 1, 2, and 3, respectively);
assuming 7days/week and 14 hours/day, 630 days in River Section 1 and 210 days each in River Sections 2 and 3.
Percentage is calculated as TPCB source strength divide by the TPCB production rate.
Percentage is calculated as TPCB flux at the monitoring station divide by the TPCB production rate.
TPCB flux is calculated based on the 500 ng/L at the far-field monitoring stations minus the mean baseline TPCB concentrations based on the GE water column samples data.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 32
Example of CSTR-Chem, TSS-Chem, and HUDTOX Application
Sediment removal
season
Dredging
Location
speed
Cubic
yards of
sediment
removed1
Monitoring
Location
(Compliance
Point)2
Expected Total
PCB export rate
at compliance
point3 (g/day)
Percent
remaining at
the monitoring
location4
Total PCB
input flux to
HUDTOX
(g/day)
TSS-Chem
Output at 1
Mile of Dredge
head5 (kg/day)
CSTR-Chem
Input6
(kg/day)
Jun. 1 - Sep. 15, 20067
Sec. 1
half
260,000
TID
1,237
73%
1,697
61,030
281,965
May 1 - Nov. 30, 2007
Sec. 1
full
520,000
TID
1,237
73%
1,684
60,575
279,856
May 1 - Nov. 30, 2008
Sec. 1
full
520,000
TID
1,237
83%
1,490
53,423
246,754
May 1 - Aug. 15,2009
Sec. 1 &
full
260,000
TID
1,237
97%
1,278
45,599
210,718
Aug. 16-Nov. 30,200
Sec. 2
290,000
Schuylerville
2,034
82%
2,466
37,814
133,724
May 1 - Aug. 15,2010
Sec. 2 &
full
290,000
Schuylerville
2,034
96%
2,117
32,242
114,014
Aug. 16-Nov. 30,201
Sec. 3
255,000
Waterford
2,334
74%
3,150
106,675
377,052
May 1 - Aug. 15,2011
Sec. 3
full
255,000
Waterford
2,334
96%
2,441
82,308
290,921
Notes:
1 Volume of sediment removed is taken from Table 8-9 of the Feasibility Study.
2 All TIP monitoring is done at TID, all River Section 2 monitoring is done at Schuylerville, and all River Section 3 monitoring is done at Waterford. 1 mile exclusion is not considered.
3 Total PCB Flux is the average net flux for June to Nov at the compliance point (TID, Schuylerville, & Waterford). PCB flux in May was excluded since flow is not typical.
4 Percent remaining at the monitoring location was obtained from the initial HUDTOX runs performed for the preliminary draft of the resuspension performance standard
5 Input to HUDTOX
6 CSTR-Chem suspended solids flux is the resuspension production rate.
7 Actual dredging period is from May 1 - Nov. 30, 2006. The PCB and TSS flux is loaded only from June 1 to Sep. 15, 2006 to account for half speed operation.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 33
Expected versus Model Prediction of PCB Flux for Control Level - 600 g/day Scenario
Loading period
Tri+ PCB Input
Tri+ PCB Output
Total PCB @ monitoring station
River Section
Dredging Season
Monitoring
(g/day)
(g/period)
(g/day)
Expected
Model estimate
From
To
Station
(g/period)1
(g/day)2
(g/day)
May 1-Nov 30, 2006
1 - Jun-06
15-Sep-06
TID
260
27,820
195
20,853
600
624
Section 1
May 1-Nov 30, 2007
l-May-07
30-Nov-07
TID
260
55,640
197
42,114
600
630
dredging
May 1-Nov 30, 2008
l-May-08
30-Nov-08
TID
230
49,220
195
41,740
600
624
May 1-Aug 15, 2009
l-May-09
15-Aug-09
TID
190
20,330
186
19,865
600
594
Section 2
Aug 16 - Nov 30, 2009
16-Aug-09
30-Nov-09
Schuylerville
210
22,470
183
19,573
600
622
dredging
May 1 - Aug 15, 2010
1-May-10
15-Aug-10
Schuylerville
180
19,260
174
18,609
600
591
Section 3
Aug 16-Nov 30, 2010;
16-Aug-10
30-Nov-lO
Waterford
300
27,300
243
22,373
600
657
dredging
May 1 - Aug 15, 2011
1-May-11
15-Aug-l 1
Waterford
230
24,610
240
25,680
600
648
Notes:
'Output loading from HUDTOX
2Total PCB flux at the monitoring station based on 1% export rate at the monitoring stations
September output from HUDTOX appears to have incorrect loading, 15 days instead of 30 days. Input loading was adjusted to reflect this.
Table 34
Expected versus Model Prediction of PCB Flux for Evaluation Level - 300 g/day Scenario
Loading period
Tri+ PCB Input
Tri+ PCB Output
Total PCB @ monitoring station
River Section
Dredging Season
Monitoring
(g/day)
(g/period)
(g/day)
Expected
Model estimate
From
To
Station
(g/period)1
(g/day)2
(g/day)
May 1-Nov 30, 2006
1 - Jun-06
15-Sep-06
TID
130
13,910
100
10,664
300
319
Section 1
May 1-Nov 30, 2007
l-May-07
30-Nov-07
TID
130
27,820
101
21,667
300
324
dredging
May 1-Nov 30, 2008
l-May-08
30-Nov-08
TID
110
23,540
95
20,287
300
303
May 1-Aug 15, 2009
l-May-09
15-Aug-09
TID
100
10,700
98
10,492
300
314
Section 2
Aug 16 - Nov 30, 2009
16-Aug-09
30-Nov-09
Schuylerville
110
11,770
98
10,456
300
332
dredging
May 1 - Aug 15, 2010
1-May-10
15-Aug-10
Schuylerville
90
9,630
89
9,565
300
304
Section 3
Aug 16-Nov 30, 2010;
16-Aug-10
30-Nov-lO
Waterford
150
13,650
125
11,464
300
336
dredging
May 1 - Aug 15, 2011
1-May-11
15-Aug-l 1
Waterford
120
12,840
125
13,421
300
339
Notes:
'Output loading from HUDTOX
2Total PCB flux at the monitoring station based on 0.5% export rate at the monitoring stations
September output from HUDTOX appears to have incorrect loading, 15 days instead of 30 days. Input loading was adjusted to reflect this
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 35
FISHRAND Forecast for Year to Reach Fish Tissue Concentration Difference of 0.5
mg/kg Relative to the No Resuspension - Upper River
River Section
Total PCB 600 g/day (srOl)
Total PCB 350 ng/L (sr04)
Section 1 (RM 189)
2008-2009
2009-2010
Section 2 (RM 184)
2008
2010
Section 3 (RM 154)
Always <0.5 mg/kg
2011
Table 36
FISHRAND Forecast for Year to Reach Fish Tissue Concentration Difference of 0.05
mg/kg Relative to the No Resuspension - Lower River
River Section
Total PCB 600 g/day (srOl)
Total PCB 350 ng/L (sr04)
RM 152
RM 113
RM 90
RM 50
2013-2014
2014
2014
Always <0.05 mg/kg
2016-2017
2016-2017
2018
2018
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 37
Upper Hudson Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
Year
No Resuspension (d004)
350 ng/L (sr04)
Upper River
Average
River Section 1
(RM 189)
River Section 2
(RM 184)
River Section 3
(RM 154)
Upper River
Average
River Section 1
(RM 189)
River Section 2
(RM 184)
River Section 3
(RM 154)
1998
3.317
6.813
9.271
1.537
3.316
6.807
9.276
1.537
1999
3.328
6.908
9.406
1.510
3.328
6.909
9.410
1.509
2000
2.866
5.747
8.346
1.300
2.865
5.751
8.338
1.300
2001
2.582
5.098
7.588
1.177
2.583
5.104
7.585
1.177
2002
2.370
4.841
6.925
1.053
2.372
4.848
6.924
1.054
2003
2.182
4.340
6.471
0.978
2.182
4.338
6.474
0.978
2004
2.290
5.285
6.356
0.946
2.290
5.286
6.354
0.946
2005
1.905
3.912
5.712
0.816
1.911
3.910
5.740
0.821
2006
1.617
2.996
5.119
0.716
1.703
3.111
5.350
0.770
2007
1.487
2.838
4.669
0.647
1.709
3.461
5.141
0.739
2008
1.297
2.318
4.226
0.571
1.673
3.762
4.743
0.694
2009
0.964
1.573
2.949
0.489
1.323
2.317
3.769
0.687
2010
0.595
0.899
1.355
0.398
0.928
1.012
1.835
0.753
2011
0.447
0.661
0.847
0.332
0.817
0.736
1.122
0.781
2012
0.404
0.723
0.786
0.269
0.631
0.774
0.999
0.537
2013
0.342
0.568
0.717
0.229
0.515
0.600
0.883
0.433
2014
0.318
0.593
0.669
0.199
0.453
0.602
0.803
0.361
2015
0.289
0.520
0.638
0.178
0.400
0.524
0.751
0.312
2016
0.294
0.586
0.651
0.170
0.391
0.589
0.750
0.287
2017
0.296
0.671
0.612
0.161
0.379
0.672
0.704
0.260
2018
0.272
0.606
0.574
0.149
0.344
0.605
0.665
0.233
2019
0.281
0.710
0.567
0.140
0.341
0.702
0.656
0.210
2020
0.243
0.584
0.502
0.125
0.292
0.579
0.584
0.180
2021
0.217
0.471
0.482
0.117
0.260
0.468
0.557
0.164
2022
0.215
0.476
0.477
0.114
0.253
0.473
0.548
0.155
2023
0.216
0.529
0.454
0.108
0.247
0.524
0.514
0.142
2024
0.195
0.484
0.417
0.094
0.219
0.480
0.463
0.122
2025
0.176
0.415
0.391
0.088
0.196
0.413
0.426
0.110
2026
0.163
0.357
0.377
0.084
0.180
0.355
0.405
0.103
2027
0.183
0.490
0.380
0.083
0.197
0.488
0.403
0.100
2028
0.177
0.509
0.353
0.076
0.189
0.508
0.371
0.090
2029
0.158
0.414
0.337
0.072
0.168
0.412
0.351
0.084
2030
0.143
0.326
0.326
0.072
0.152
0.325
0.342
0.082
2031
0.151
0.422
0.303
0.067
0.159
0.421
0.320
0.075
2032
0.138
0.362
0.288
0.064
0.145
0.362
0.305
0.071
2033
0.133
0.349
0.277
0.061
0.138
0.349
0.295
0.066
2034
0.132
0.368
0.259
0.060
0.134
0.368
0.276
0.060
2035
0.123
0.279
0.249
0.068
0.116
0.279
0.266
0.056
2036
0.148
0.356
0.242
0.087
0.124
0.356
0.258
0.051
2037
0.137
0.297
0.234
0.086
0.115
0.298
0.250
0.053
2038
0.140
0.337
0.221
0.083
0.130
0.337
0.235
0.068
2039
0.128
0.270
0.214
0.083
0.132
0.271
0.227
0.087
2040
0.124
0.262
0.214
0.079
0.132
0.262
0.225
0.087
2041
0.140
0.359
0.219
0.079
0.150
0.360
0.228
0.091
2042
0.143
0.400
0.223
0.074
0.153
0.401
0.229
0.087
2043
0.123
0.318
0.202
0.068
0.132
0.318
0.206
0.080
2044
0.108
0.245
0.191
0.064
0.114
0.246
0.193
0.073
2045
0.112
0.282
0.190
0.063
0.118
0.283
0.191
0.070
BOLD-ITALICIZED - First occurrence of species-weighted fish fillet average PCB concentration below risk-based remediation goal of
0.05 mg/kg. Target concentrations of 0.2 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/month)
and 0.4 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/ 2 months) are also italicized.
Upper Hudson River average is weighted by river section length:
River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 37
Upper Hudson Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
No Resuspension (d004)
350 ng/L (sr04)
Upper River
River Section 1
River Section 2
River Section 3
Upper River
River Section 1
River Section 2
River Section 3
Year
Average
(RM 189)
(RM 184)
(RM 154)
Average
(RM 189)
(RM 184)
(RM 154)
2046
0.105
0.258
0.184
0.058
0.109
0.256
0.184
0.064
2047
0.109
0.284
0.187
0.058
0.112
0.271
0.187
0.065
2048
0.115
0.329
0.188
0.057
0.118
0.318
0.187
0.064
2049
0.116
0.339
0.190
0.055
0.120
0.340
0.189
0.062
2050
0.105
0.289
0.183
0.052
0.109
0.290
0.182
0.057
2051
0.101
0.286
0.180
0.047
0.104
0.287
0.178
0.052
2052
0.094
0.244
0.181
0.047
0.097
0.246
0.180
0.051
2053
0.113
0.359
0.187
0.048
0.116
0.359
0.185
0.052
2054
0.105
0.311
0.185
0.047
0.107
0.311
0.184
0.050
2055
0.098
0.274
0.182
0.045
0.100
0.274
0.180
0.048
2056
0.105
0.307
0.195
0.046
0.106
0.307
0.193
0.048
2057
0.105
0.323
0.185
0.045
0.107
0.324
0.183
0.047
2058
0.095
0.253
0.188
0.045
0.096
0.253
0.186
0.047
2059
0.109
0.356
0.181
0.043
0.110
0.356
0.181
0.045
2060
0.091
0.256
0.175
0.040
0.092
0.256
0.175
0.042
2061
0.086
0.234
0.169
0.040
0.087
0.233
0.169
0.042
2062
0.091
0.261
0.171
0.040
0.091
0.261
0.170
0.042
2063
0.091
0.261
0.172
0.041
0.091
0.260
0.171
0.041
2064
0.093
0.268
0.175
0.041
0.093
0.268
0.174
0.042
2065
0.092
0.255
0.178
0.043
0.093
0.255
0.177
0.043
2066
0.105
0.353
0.172
0.041
0.105
0.353
0.171
0.041
2067
0.095
0.275
0.180
0.042
0.095
0.275
0.179
0.042
BOLD-ITALICIZED - First occurrence of species-weighted fish fillet average PCB concentration below risk-based remediation goal of
0.05 mg/kg. Target concentrations of 0.2 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/month)
and 0.4 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/ 2 months) are also italicized.
Upper Hudson River average is weighted by river section length:
River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 37
Upper Hudson Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
Year
600 g/day (srOl)
Monitored Natural Attenuation
Upper River
Average
River Section 1
(RM 189)
River Section 2
(RM 184)
River Section 3
(RM 154)
Upper River
Average
River Section 1
(RM 189)
River Section 2
(RM 184)
River Section 3
(RM 154)
1998
3.316
6.807
9.276
1.537
3.353
6.774
9.659
1.529
1999
3.328
6.909
9.410
1.509
3.212
6.621
8.877
1.501
2000
2.865
5.751
8.338
1.300
2.791
5.563
8.028
1.292
2001
2.583
5.104
7.585
1.177
2.504
4.924
7.210
1.171
2002
2.372
4.848
6.924
1.054
2.301
4.705
6.571
1.047
2003
2.182
4.338
6.474
0.978
2.129
4.290
6.090
0.980
2004
2.290
5.286
6.354
0.946
2.204
5.084
5.934
0.942
2005
1.908
3.909
5.726
0.819
1.852
3.739
5.523
0.812
2006
1.666
3.076
5.237
0.746
1.574
2.890
4.904
0.716
2007
1.614
3.225
4.920
0.697
1.474
2.862
4.489
0.654
2008
1.525
3.216
4.582
0.634
1.371
2.774
4.168
0.586
2009
1.106
1.907
3.140
0.583
1.262
2.616
3.877
0.519
2010
0.707
0.943
1.411
0.535
1.116
2.321
3.533
0.440
2011
0.568
0.697
0.901
0.483
0.971
1.921
3.164
0.388
2012
0.469
0.747
0.818
0.350
0.878
1.851
2.879
0.324
2013
0.389
0.572
0.734
0.291
0.791
1.682
2.601
0.287
2014
0.353
0.582
0.675
0.248
0.742
1.666
2.396
0.258
2015
0.316
0.506
0.638
0.219
0.686
1.535
2.229
0.237
2016
0.317
0.573
0.648
0.205
0.680
1.610
2.126
0.231
2017
0.315
0.660
0.610
0.190
0.649
1.573
1.978
0.221
2018
0.289
0.595
0.577
0.173
0.593
1.437
1.765
0.210
2019
0.295
0.694
0.572
0.161
0.577
1.497
1.619
0.200
2020
0.253
0.571
0.507
0.142
0.512
1.270
1.480
0.182
2021
0.226
0.459
0.486
0.131
0.460
1.080
1.365
0.171
2022
0.222
0.464
0.482
0.126
0.450
1.093
1.296
0.166
2023
0.222
0.517
0.461
0.118
0.435
1.088
1.225
0.158
2024
0.200
0.474
0.427
0.102
0.385
0.939
1.123
0.139
2025
0.181
0.406
0.402
0.094
0.350
0.842
1.019
0.129
2026
0.166
0.347
0.388
0.089
0.325
0.757
0.952
0.124
2027
0.186
0.483
0.387
0.088
0.339
0.888
0.920
0.121
2028
0.179
0.504
0.353
0.080
0.322
0.863
0.875
0.111
2029
0.159
0.407
0.332
0.076
0.287
0.720
0.801
0.105
2030
0.143
0.320
0.322
0.075
0.261
0.620
0.735
0.103
2031
0.152
0.418
0.302
0.069
0.257
0.679
0.675
0.095
2032
0.139
0.357
0.289
0.066
0.234
0.602
0.610
0.091
2033
0.133
0.343
0.279
0.063
0.219
0.560
0.564
0.086
2034
0.132
0.366
0.261
0.059
0.208
0.545
0.521
0.082
2035
0.114
0.275
0.251
0.055
0.191
0.443
0.475
0.089
2036
0.125
0.352
0.244
0.055
0.209
0.504
0.446
0.104
2037
0.125
0.295
0.237
0.070
0.190
0.427
0.410
0.101
2038
0.140
0.335
0.224
0.083
0.189
0.456
0.386
0.098
2039
0.131
0.268
0.218
0.087
0.173
0.382
0.363
0.096
2040
0.128
0.260
0.217
0.085
0.164
0.352
0.346
0.092
2041
0.146
0.358
0.222
0.087
0.180
0.461
0.347
0.092
2042
0.148
0.399
0.225
0.081
0.178
0.486
0.337
0.084
2043
0.129
0.320
0.205
0.075
0.155
0.386
0.316
0.078
2044
0.114
0.256
0.195
0.069
0.136
0.301
0.289
0.074
2045
0.118
0.301
0.194
0.066
0.137
0.329
0.278
0.071
BOLD-ITALICIZED - First occurrence of species-weighted fish fillet average PCB concentration below risk-based remediation goal of
0.05 mg/kg. Target concentrations of 0.2 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/month)
and 0.4 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/ 2 months) are also italicized.
Upper Hudson River average is weighted by river section length:
River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 3 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 37
Upper Hudson Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
600 g/day (srOl)
Monitored Natural Attenuation
Upper River
River Section 1
River Section 2
River Section 3
Upper River
River Section 1
River Section 2
River Section 3
Year
Average
(RM 189)
(RM 184)
(RM 154)
Average
(RM 189)
(RM 184)
(RM 154)
2046
0.110
0.273
0.187
0.062
0.131
0.319
0.269
0.067
2047
0.112
0.285
0.190
0.062
0.153
0.474
0.261
0.066
2048
0.116
0.316
0.190
0.061
0.175
0.612
0.263
0.066
2049
0.117
0.328
0.192
0.059
0.166
0.574
0.259
0.063
2050
0.106
0.283
0.185
0.055
0.151
0.498
0.251
0.060
2051
0.104
0.294
0.182
0.050
0.140
0.457
0.242
0.055
2052
0.099
0.263
0.184
0.049
0.130
0.402
0.236
0.054
2053
0.118
0.379
0.189
0.050
0.146
0.494
0.244
0.055
2054
0.109
0.327
0.187
0.049
0.134
0.430
0.235
0.053
2055
0.101
0.287
0.183
0.047
0.125
0.383
0.231
0.052
2056
0.108
0.322
0.195
0.047
0.129
0.407
0.233
0.051
2057
0.108
0.337
0.186
0.046
0.126
0.397
0.231
0.050
2058
0.097
0.264
0.188
0.046
0.116
0.337
0.226
0.050
2059
0.111
0.366
0.182
0.044
0.127
0.422
0.228
0.047
2060
0.093
0.266
0.175
0.041
0.106
0.316
0.209
0.044
2061
0.087
0.241
0.169
0.041
0.100
0.286
0.200
0.043
2062
0.092
0.268
0.170
0.041
0.102
0.297
0.197
0.043
2063
0.092
0.266
0.171
0.041
0.101
0.296
0.196
0.043
2064
0.094
0.273
0.175
0.042
0.103
0.306
0.196
0.044
2065
0.093
0.260
0.177
0.043
0.100
0.283
0.195
0.045
2066
0.106
0.358
0.171
0.041
0.113
0.377
0.195
0.043
2067
0.096
0.279
0.179
0.043
0.101
0.301
0.183
0.044
BOLD-ITALICIZED - First occurrence of species-weighted fish fillet average PCB concentration below risk-based remediation goal of
0.05 mg/kg. Target concentrations of 0.2 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/month)
and 0.4 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/ 2 months) are also italicized.
Upper Hudson River average is weighted by river section length:
River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 4 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 38
Upper Hudson River Modeled Times (Years) of Compliance with
Human Health Risk-Based Concentrations Resuspension Scenarios
No Resuspension
(d004)
350 ng/L (sr04)
600 g/day (srOl)
MNA
Upper River Average
Human Health risk-based RG 0.05 mg/kg
>2067
>2067
>2067
>2067
Fish Target Concentration 0.2 mg/kg
2024
2025
2024
2035
Fish Target Concentration 0.4 mg/kg
2013
2015
2013
2024
River Section 1- RM 189
Human Health risk-based RG 0.05 mg/kg
>2067
>2067
>2067
>2067
Fish Target Concentration 0.2 mg/kg
>2067
>2067
>2067
>2067
Fish Target Concentration 0.4 mg/kg
2026
2030
2026
2043
River Section 2- RM 184
Human Health risk-based RG 0.05 mg/kg
>2067
>2067
>2067
>2067
Fish Target Concentration 0.2 mg/kg
2044
2044
2044
2061
Fish Target Concentration 0.4 mg/kg
2025
2028
2026
2038
River Section 3- RM 154
Human Health RG 0.05 mg/kg
2051
2055
2051
2059
Fish Target Concentration 0.2 mg/kg
2014
2020
2017
2019
Fish Target Concentration 0.4 mg/kg
2010
2014
2012
2011
Note: RG = risk-based remediation goal
Upper Hudson River average is weighted by river section length. River Section 1:6.3 miles = 15.4%;
River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 12.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 39
Resuspension Scenarios - Long-Term Fish Ingestion
Reasonable Maximum Exposure and Central Tendency PCB Non-Cancer Hazard Indices
Upper Hudson River Fish - Adult Angler
Remedial
PCB Cone.
Intake
Reference
Hazard
Alternative
in Fish
(Non-Cancer)
Dose
Index
(mg/kg ww)
(mg/kg-day)
(mg/kg-day)
Reasonable Maximum Exposure
Upper Hudson Average
No Resuspension d004
0.30
1.4E-04
2.0E-05
6.9
350 ng/L sr04
0.58
2.6E-04
2.0E-05
13
600 g/day srOl
0.50
2.3E-04
2.0E-05
11
MNA
1.4
6.4E-04
2.0E-05
32
River Section 1 (RM 189)
No Resuspension d004
0.62
2.8E-04
2.0E-05
14
350 ng/L sr04
0.64
2.9E-04
2.0E-05
15
600 g/day srOl
0.62
2.8E-04
2.0E-05
14
MNA
1.7
7.7E-04
2.0E-05
39
River Section 2 (RM 184)
No Resuspension d004
0.66
3.0E-04
2.0E-05
15
350 ng/L sr04
0.79
3.6E-04
2.0E-05
18
600 g/day srOl
0.67
3.1E-04
2.0E-05
15
MNA
2.3
1.0E-03
2.0E-05
52
River Section 3 (RM 154)
No Resuspension d004
0.18
8.0E-05
2.0E-05
4.0
350 ng/L sr04
0.30
1.4E-04
2.0E-05
6.8
600 g/day srOl
0.21
9.7E-05
2.0E-05
4.8
MNA
0.23
1.1E-04
2.0E-05
5.4
Central Tendency
Upper Hudson Average
No Resuspension d004
0.27
1.2E-05
2.0E-05
0.6
350 ng/L sr04
0.52
2.4E-05
2.0E-05
1.2
600 g/day srOl
0.46
2.1E-05
2.0E-05
1.0
MNA
1.2
5.5E-05
2.0E-05
2.8
River Section 1 (RM 189)
No Resuspension d004
0.60
2.7E-05
2.0E-05
1.4
350 ng/L sr04
0.61
2.8E-05
2.0E-05
1.4
600 g/day srOl
0.59
2.7E-05
2.0E-05
1.4
MNA
1.50
6.9E-05
2.0E-05
3.5
River Section 2 (RM 184)
No Resuspension d004
0.59
2.7E-05
2.0E-05
1.4
350 ng/L sr04
0.70
3.2E-05
2.0E-05
1.6
600 g/day srOl
0.60
2.7E-05
2.0E-05
1.4
MNA
1.9
8.7E-05
2.0E-05
4.4
River Section 3 (RM 154)
No Resuspension d004
0.15
6.8E-06
2.0E-05
0.3
350 ng/L sr04
0.24
1.1E-05
2.0E-05
0.5
600 g/day srOl
0.18
8.0E-06
2.0E-05
0.4
MNA
0.21
9.4E-06
2.0E-05
0.5
Notes: The RME non-cancer exposure time frame is seven years, while the CT time frame is 12 years.
Upper Hudson River average is weighted by river section length. River Section 1: 6.3 miles = 15.4%;
River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 40
Resuspension Standard Scenarios - Long-Term Fish Ingestion
Reasonable Maximum Exposure and Central Tendency Cancer Risks
Upper Hudson River Fish - Adult Angler
Remedial
PCB Cone.
Intake
Cancer Slope
Cancer
Alternative
in Fish
(Cancer)
Factor
Risk
(mg/kg ww)
(mg/kg-day)
(mg/kg-day)
Reasonable Maximum Exposure
Upper Hudson Average
No Resuspension d004
0.18
4.6E-05
2
9.3E-05
350 ng/L sr04
0.32
8.3E-05
2
1.7E-04
600 g/day srOl
0.30
7.7E-05
2
1.5E-04
MNA
0.60
1.7E-04
2
3.3E-04
River Section 1 (RM 189)
No Resuspension d004
0.43
1.1E-04
2
2.2E-04
350 ng/L sr04
0.43
1.1E-04
2
2.2E-04
600 g/day srOl
0.42
1.1E-04
2
2.2E-04
MNA
0.86
2.2E-04
2
4.5E-04
River Section 2 (RM 184)
No Resuspension d004
0.36
9.3E-05
2
1.9E-04
350 ng/L sr04
0.40
1.0E-04
2
2.1E-04
600 g/day srOl
0.36
9.4E-05
2
1.9E-04
MNA
0.90
2.4E-04
2
4.9E-04
River Section 3 (RM 154)
No Resuspension d004
0.09
2.4E-05
2
4.8E-05
350 ng/L sr04
0.12
3.2E-05
2
6.4E-05
600 g/day srOl
0.10
2.7E-05
2
5.3E-05
MNA
0.12
3.2E-05
2
6.4E-05
Central Tendency
Upper Hudson Average
No Resuspension d004
0.27
2.1E-06
1
2.1E-06
350 ng/L sr04
0.52
4.0E-06
1
4.0E-06
600 g/day srOl
0.46
3.6E-06
1
3.6E-06
MNA
1.2
9.5E-06
1
9.5E-06
River Section 1 (RM 189)
No Resuspension d004
0.60
4.7E-06
1
4.7E-06
350 ng/L sr04
0.61
4.8E-06
1
4.8E-06
600 g/day srOl
0.59
4.7E-06
1
4.7E-06
MNA
1.5
1.2E-05
1
1.2E-05
River Section 2 (RM 184)
No Resuspension d004
0.59
4.7E-06
1
4.7E-06
350 ng/L sr04
0.70
5.5E-06
1
5.5E-06
600 g/day srOl
0.60
4.7E-06
1
4.7E-06
MNA
1.9
1.5E-05
1
1.5E-05
River Section 3 (RM 154)
No Resuspension d004
0.15
1.2E-06
1
1.2E-06
350 ng/L sr04
0.24
1.9E-06
1
1.9E-06
600 g/day srOl
0.18
1.4E-06
1
1.4E-06
MNA
0.21
1.6E-06
1
1.6E-06
Notes: The RME cancer exposure time frame is 40 years, while the CT time frame is 12 years.
Upper Hudson River average is weighted by river section length. River Section 1: 6.3 miles = 15.4%;
River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 41
Mid-Hudson River
Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
No Resuspension (d004)
350 ng/L (sr04)
River Section 1 (RM
River Section 2 (RM
River Section 3 (RM
River Section 1 (RM
River Section 2 (RM
River Section 3 (RM
Year
152)
113)
90)
152)
113)
90)
1999
1.150
0.963
0.792
1.150
0.963
0.792
2000
1.080
0.851
0.712
1.080
0.851
0.712
2001
1.154
0.821
0.656
1.154
0.821
0.656
2002
0.972
0.745
0.611
0.972
0.745
0.611
2003
0.837
0.658
0.553
0.837
0.658
0.553
2004
0.622
0.537
0.485
0.622
0.537
0.485
2005
0.592
0.462
0.420
0.598
0.463
0.420
2006
0.605
0.435
0.373
0.661
0.453
0.375
2007
0.522
0.398
0.337
0.641
0.441
0.349
2008
0.386
0.335
0.301
0.560
0.407
0.326
2009
0.316
0.278
0.263
0.537
0.370
0.296
2010
0.308
0.250
0.231
0.734
0.420
0.294
2011
0.307
0.234
0.208
1.119
0.558
0.325
2012
0.247
0.205
0.187
0.570
0.464
0.329
2013
0.253
0.192
0.170
0.443
0.381
0.308
2014
0.217
0.172
0.155
0.330
0.305
0.274
2015
0.181
0.152
0.140
0.259
0.245
0.238
2016
0.136
0.127
0.125
0.186
0.190
0.201
2017
0.118
0.110
0.111
0.138
0.149
0.168
2018
0.110
0.098
0.099
0.118
0.123
0.139
2019
0.093
0.086
0.088
0.095
0.099
0.115
2020
0.108
0.084
0.080
0.109
0.090
0.098
2021
0.101
0.081
0.075
0.101
0.084
0.086
2022
0.087
0.075
0.071
0.087
0.077
0.078
2023
0.080
0.070
0.066
0.080
0.071
0.071
2024
0.085
0.069
0.064
0.085
0.070
0.066
2025
0.088
0.070
0.063
0.088
0.071
0.064
2026
0.083
0.068
0.061
0.083
0.068
0.062
2027
0.069
0.063
0.059
0.069
0.063
0.060
2028
0.068
0.060
0.056
0.068
0.060
0.057
2029
0.076
0.060
0.055
0.076
0.060
0.055
2030
0.074
0.060
0.054
0.074
0.060
0.054
2031
0.068
0.058
0.054
0.068
0.058
0.054
2032
0.067
0.058
0.053
0.067
0.058
0.053
2033
0.063
0.056
0.052
0.063
0.056
0.052
2034
0.064
0.055
0.051
0.064
0.055
0.051
2035
0.095
0.063
0.052
0.095
0.063
0.052
2036
0.126
0.078
0.056
0.126
0.078
0.056
2037
0.141
0.091
0.063
0.141
0.091
0.063
2038
0.138
0.093
0.068
0.138
0.094
0.068
2039
0.122
0.091
0.070
0.122
0.091
0.070
2040
0.106
0.086
0.070
0.106
0.086
0.070
2041
0.081
0.075
0.067
0.081
0.075
0.067
2042
0.069
0.066
0.063
0.069
0.066
0.063
2043
0.079
0.064
0.059
0.079
0.064
0.059
2044
0.091
0.067
0.058
0.091
0.067
0.058
2045
0.085
0.067
0.057
0.085
0.067
0.057
2046
0.076
0.063
0.056
0.076
0.063
0.056
BOLD-ITALICIZED - First occurrence of species-weighted fish fillet average PCB concentration below risk-based remediation goal of
0.05 mg/kg. Target concentrations of 0.2 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/month)
and 0.4 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/ 2 months) are also italicized.
Note: Fish concentrations were not available for all species used to model Mid-Hudson River angler consumption.
Therefore, the concentrations here provide only an estimate of fish concentrations.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 41
Mid-Hudson River
Species-Weighted Fish Fillet Average PCB Concentration (in mg/kg)
600 g/day (srOl)
Monitored Natural Attenuation
River Section 1 (RM
River Section 2 (RM
River Section 3 (RM
River Section 1 (RM
River Section 2 (RM
River Section 3 (RM
Year
152)
113)
90)
152)
113)
90)
1999
1.150
0.963
0.792
1.126
0.952
0.788
2000
1.080
0.851
0.712
1.093
0.848
0.708
2001
1.154
0.821
0.656
1.138
0.822
0.654
2002
0.972
0.745
0.611
0.959
0.742
0.607
2003
0.837
0.658
0.553
0.826
0.655
0.549
2004
0.622
0.537
0.485
0.611
0.532
0.482
2005
0.595
0.463
0.420
0.580
0.459
0.417
2006
0.633
0.444
0.374
0.598
0.432
0.370
2007
0.589
0.420
0.343
0.500
0.394
0.334
2008
0.478
0.374
0.314
0.388
0.333
0.299
2009
0.420
0.321
0.280
0.333
0.279
0.261
2010
0.467
0.316
0.259
0.367
0.272
0.236
2011
0.549
0.335
0.251
0.391
0.276
0.222
2012
0.343
0.284
0.235
0.338
0.256
0.210
2013
0.313
0.252
0.215
0.352
0.252
0.199
2014
0.253
0.216
0.193
0.303
0.230
0.188
2015
0.205
0.182
0.171
0.246
0.202
0.173
2016
0.152
0.147
0.149
0.185
0.167
0.155
2017
0.130
0.124
0.130
0.171
0.145
0.138
2018
0.120
0.109
0.113
0.165
0.136
0.125
2019
0.100
0.094
0.098
0.143
0.120
0.112
2020
0.115
0.090
0.088
0.168
0.120
0.104
2021
0.106
0.085
0.081
0.153
0.115
0.098
2022
0.091
0.079
0.075
0.127
0.106
0.093
2023
0.084
0.073
0.070
0.119
0.097
0.087
2024
0.088
0.072
0.066
0.127
0.097
0.084
2025
0.091
0.073
0.065
0.131
0.100
0.083
2026
0.085
0.070
0.063
0.119
0.095
0.081
2027
0.070
0.065
0.061
0.098
0.087
0.078
2028
0.070
0.061
0.058
0.098
0.081
0.073
2029
0.078
0.061
0.056
0.106
0.081
0.070
2030
0.075
0.061
0.055
0.102
0.079
0.068
2031
0.069
0.059
0.054
0.094
0.077
0.067
2032
0.068
0.058
0.054
0.092
0.077
0.068
2033
0.064
0.056
0.052
0.084
0.073
0.066
2034
0.062
0.055
0.051
0.086
0.071
0.063
2035
0.069
0.056
0.050
0.118
0.079
0.064
2036
0.068
0.055
0.050
0.145
0.093
0.067
2037
0.111
0.070
0.053
0.166
0.107
0.074
2038
0.152
0.088
0.060
0.156
0.107
0.079
2039
0.141
0.094
0.066
0.139
0.105
0.080
2040
0.123
0.093
0.070
0.120
0.098
0.080
2041
0.094
0.083
0.069
0.092
0.085
0.076
2042
0.078
0.072
0.066
0.078
0.073
0.070
2043
0.088
0.071
0.063
0.093
0.073
0.066
2044
0.097
0.073
0.062
0.107
0.077
0.064
2045
0.087
0.070
0.060
0.098
0.076
0.064
2046
0.078
0.066
0.058
0.086
0.072
0.062
BOLD-ITALICIZED - First occurrence of species-weighted fish fillet average PCB concentration below risk-based remediation goal of
0.05 mg/kg. Target concentrations of 0.2 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/month)
and 0.4 mg/kg PCBs (protective at a fish consumption rate of 0.5 lbs/ 2 months) are also italicized.
Note: Fish concentrations were not available for all species used to model Mid-Hudson River angler consumption.
Therefore, the concentrations here provide only an estimate of fish concentrations.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 42
Upper Hudson River Average Largemouth Bass (Whole Fish)
PCB Concentration (in mg/kg)
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
Upper River
Section 1
Section 2
Section 3
Upper River
Section 1
Section 2
Section 3
Year
Average
(RM 189)
(RM 184)
(RM 154)
Average
(RM 189)
(RM 184)
(RM 154)
1998
7.13
16.73
17.22
3.33
7.13
16.70
17.24
3.33
1999
7.04
17.11
16.80
3.20
7.04
17.12
16.83
3.20
2000
5.84
13.71
14.51
2.66
5.84
13.74
14.47
2.66
2001
5.29
12.01
13.33
2.47
5.30
12.04
13.32
2.47
2002
4.91
11.63
12.30
2.20
4.92
11.66
12.29
2.20
2003
4.43
10.12
11.39
2.01
4.43
10.11
11.40
2.01
2004
5.12
14.37
11.49
2.04
5.12
14.38
11.48
2.04
2005
3.94
9.68
9.91
1.67
3.95
9.67
9.97
1.68
2006
3.14
6.44
8.80
1.45
3.38
6.61
9.48
1.63
2007
2.96
6.45
8.04
1.33
3.63
8.59
9.25
1.59
2008
2.59
5.37
7.38
1.17
3.88
11.02
8.77
1.51
2009
2.00
4.08
5.15
1.02
3.06
6.90
7.31
1.50
2010
1.35
2.88
2.56
0.81
2.14
3.17
3.68
1.66
2011
1.00
2.02
1.57
0.68
1.94
2.18
2.05
1.86
2012
0.94
2.35
1.48
0.55
1.38
2.45
1.85
1.07
2013
0.76
1.69
1.30
0.47
1.08
1.75
1.59
0.85
2014
0.72
1.80
1.22
0.41
0.97
1.81
1.44
0.71
2015
0.64
1.52
1.16
0.37
0.85
1.53
1.35
0.62
2016
0.68
1.72
1.26
0.36
0.87
1.72
1.43
0.59
2017
0.73
2.17
1.18
0.35
0.89
2.16
1.34
0.54
2018
0.66
1.93
1.09
0.32
0.79
1.91
1.24
0.48
2019
0.72
2.34
1.13
0.30
0.83
2.32
1.28
0.43
2020
0.59
1.89
0.92
0.26
0.68
1.86
1.06
0.36
2021
0.51
1.44
0.90
0.25
0.59
1.43
1.03
0.33
2022
0.51
1.43
0.92
0.24
0.58
1.43
1.04
0.33
2023
0.54
1.69
0.88
0.24
0.60
1.67
0.98
0.30
2024
0.49
1.58
0.79
0.20
0.53
1.57
0.87
0.25
2025
0.43
1.29
0.74
0.19
0.46
1.29
0.80
0.23
2026
0.38
1.08
0.71
0.18
0.41
1.07
0.75
0.21
2027
0.47
1.60
0.74
0.18
0.50
1.59
0.78
0.21
2028
0.46
1.69
0.65
0.16
0.48
1.69
0.68
0.18
2029
0.39
1.34
0.63
0.15
0.41
1.33
0.65
0.17
2030
0.35
0.99
0.63
0.16
0.36
0.98
0.65
0.18
2031
0.40
1.42
0.58
0.15
0.41
1.41
0.61
0.16
2032
0.35
1.18
0.55
0.14
0.36
1.18
0.58
0.15
2033
0.34
1.14
0.53
0.13
0.35
1.13
0.56
0.14
2034
0.34
1.23
0.49
0.13
0.35
1.23
0.52
0.13
2035
0.29
0.88
0.47
0.14
0.28
0.87
0.50
0.12
2036
0.40
1.21
0.48
0.22
0.33
1.21
0.50
0.11
2037
0.36
0.98
0.46
0.21
0.29
0.98
0.49
0.11
2038
0.36
1.13
0.43
0.19
0.33
1.13
0.45
0.14
2039
0.33
0.89
0.42
0.19
0.34
0.89
0.44
0.21
2040
0.31
0.86
0.42
0.17
0.33
0.86
0.44
0.20
Notes:
1. Fish fillets multiplied by 2.5 to obtain whole fish concentrations.
2. All whole fish PCB concentrations are above target fish concentration of 0.3 mg/kg and/or 0.03 mg/kg based on the
river otter lowest-observed-adverse-effects-level (LOAEL) and no-observed-adverse-effects-level (NOAEL), respectively.
Upper Hudson River average is weighted by river section length:
River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 42
Upper Hudson River Average Largemouth Bass (Whole Fish)
PCB Concentration (in mg/kg)
Year
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
Upper River
Average
Section 1
(RM 189)
Section 2
(RM 184)
Section 3
(RM 154)
Upper River
Average
Section 1
(RM 189)
Section 2
(RM 184)
Section 3
(RM 154)
2041
0.37
1.23
0.44
0.18
0.40
1.23
0.45
0.22
2042
0.39
1.40
0.46
0.16
0.42
1.40
0.47
0.20
2043
0.33
1.10
0.39
0.15
0.35
1.10
0.40
0.18
2044
0.28
0.82
0.37
0.14
0.29
0.82
0.37
0.16
2045
0.30
0.97
0.38
0.14
0.31
0.97
0.38
0.16
2046
0.27
0.86
0.36
0.13
0.28
0.86
0.36
0.14
2047
0.28
0.93
0.37
0.13
0.29
0.91
0.37
0.14
2048
0.30
1.08
0.37
0.13
0.31
1.07
0.37
0.14
2049
0.31
1.14
0.39
0.12
0.33
1.15
0.39
0.14
2050
0.28
0.96
0.36
0.12
0.29
0.96
0.36
0.13
2051
0.27
0.96
0.36
0.10
0.28
0.96
0.36
0.11
2052
0.24
0.80
0.36
0.10
0.25
0.80
0.36
0.11
2053
0.32
1.26
0.38
0.11
0.32
1.26
0.38
0.12
2054
0.29
1.08
0.38
0.11
0.29
1.08
0.38
0.11
2055
0.26
0.93
0.36
0.10
0.26
0.93
0.36
0.11
2056
0.28
1.03
0.41
0.10
0.29
1.02
0.40
0.11
2057
0.29
1.14
0.37
0.10
0.30
1.14
0.37
0.10
2058
0.25
0.85
0.37
0.10
0.25
0.85
0.37
0.10
2059
0.31
1.27
0.36
0.10
0.31
1.26
0.36
0.10
2060
0.24
0.88
0.35
0.09
0.25
0.87
0.35
0.09
2061
0.23
0.79
0.33
0.09
0.23
0.79
0.33
0.09
2062
0.25
0.89
0.34
0.09
0.25
0.89
0.34
0.09
2063
0.24
0.89
0.35
0.09
0.25
0.89
0.34
0.09
2064
0.25
0.92
0.36
0.09
0.25
0.92
0.36
0.09
2065
0.25
0.88
0.36
0.10
0.25
0.87
0.36
0.10
2066
0.30
1.25
0.34
0.09
0.30
1.25
0.34
0.09
2067
0.26
0.95
0.37
0.09
0.26
0.95
0.37
0.09
Notes:
1. Fish fillets multiplied by 2.5 to obtain whole fish concentrations.
2. All whole fish PCB concentrations are above target fish concentration of 0.3 mg/kg and/or 0.03 mg/kg based on the
river otter lowest-observed-adverse-effects-level (LOAEL) and no-observed-adverse-effects-level (NOAEL), respectively.
Upper Hudson River average is weighted by river section length:
River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 42
Upper Hudson River Average Largemouth Bass (Whole Fish)
PCB Concentration (in mg/kg)
Total PCB 600 g/day (srOl)
Monitored Natural Attenuation
Upper River
Section 1
Section 2
Section 3
Upper River
Section 1
Section 2
Section 3
Year
Average
(RM 189)
(RM 184)
(RM 154)
Average
(RM 189)
(RM 184)
(RM 154)
1998
7.13
16.70
17.24
3.33
7.19
16.61
18.04
3.29
1999
7.04
17.12
16.83
3.20
6.76
16.16
15.91
3.17
2000
5.84
13.74
14.47
2.66
5.74
13.09
14.57
2.64
2001
5.30
12.04
13.32
2.47
5.13
11.34
12.94
2.45
2002
4.92
11.66
12.29
2.20
4.76
11.11
11.84
2.18
2003
4.43
10.11
11.40
2.01
4.33
9.92
10.73
2.03
2004
5.12
14.38
11.48
2.04
4.88
13.63
10.57
2.02
2005
3.94
9.67
9.95
1.68
3.85
9.04
10.09
1.66
2006
3.28
6.57
9.17
1.55
3.06
5.97
8.70
1.46
2007
3.35
7.78
8.73
1.47
2.96
6.39
7.95
1.36
2008
3.40
9.02
8.30
1.36
2.78
6.45
7.30
1.21
2009
2.49
5.39
5.93
1.27
2.60
6.16
6.88
1.10
2010
1.65
3.00
2.76
1.17
2.31
5.51
6.40
0.92
2011
1.34
2.12
1.67
1.11
1.95
4.24
5.61
0.83
2012
1.07
2.41
1.54
0.70
1.78
4.21
5.16
0.68
2013
0.85
1.71
1.34
0.59
1.55
3.47
4.60
0.61
2014
0.79
1.80
1.23
0.50
1.46
3.49
4.23
0.55
2015
0.70
1.51
1.16
0.44
1.33
3.13
3.87
0.50
2016
0.73
1.71
1.26
0.43
1.36
3.53
3.65
0.50
2017
0.77
2.16
1.18
0.40
1.38
3.73
3.60
0.49
2018
0.70
1.92
1.10
0.37
1.24
3.29
3.21
0.46
2019
0.75
2.33
1.14
0.34
1.25
3.68
2.94
0.43
2020
0.61
1.87
0.93
0.29
1.08
3.02
2.71
0.38
2021
0.53
1.42
0.91
0.27
0.93
2.43
2.40
0.36
2022
0.53
1.42
0.93
0.27
0.93
2.51
2.26
0.36
2023
0.55
1.68
0.89
0.25
0.94
2.67
2.21
0.35
2024
0.50
1.57
0.81
0.21
0.82
2.26
2.05
0.29
2025
0.44
1.28
0.76
0.20
0.73
1.98
1.82
0.28
2026
0.39
1.06
0.72
0.19
0.66
1.69
1.68
0.26
2027
0.48
1.59
0.75
0.19
0.75
2.29
1.66
0.27
2028
0.46
1.69
0.66
0.17
0.73
2.33
1.61
0.23
2029
0.40
1.33
0.62
0.16
0.62
1.83
1.44
0.22
2030
0.35
0.98
0.62
0.17
0.55
1.45
1.33
0.23
2031
0.40
1.41
0.58
0.15
0.59
1.86
1.27
0.21
2032
0.35
1.18
0.55
0.14
0.53
1.59
1.13
0.20
2033
0.34
1.13
0.53
0.13
0.49
1.47
1.04
0.18
2034
0.34
1.23
0.49
0.13
0.48
1.50
0.98
0.17
2035
0.28
0.87
0.48
0.12
0.41
1.12
0.87
0.18
2036
0.33
1.20
0.48
0.12
0.51
1.43
0.85
0.26
2037
0.32
0.98
0.47
0.15
0.45
1.19
0.75
0.24
2038
0.37
1.13
0.43
0.20
0.45
1.32
0.72
0.22
2039
0.34
0.89
0.42
0.21
0.41
1.09
0.68
0.22
2040
0.32
0.86
0.42
0.19
0.38
0.98
0.63
0.20
Notes:
1. Fish fillets multiplied by 2.5 to obtain whole fish concentrations.
2. All whole fish PCB concentrations are above target fish concentration of 0.3 mg/kg and/or 0.03 mg/kg based on the river
otter lowest-observed-adverse-effects-level (LOAEL) and no-observed-adverse-effects-level (NOAEL), respectively.
Upper Hudson River average is weighted by river section length:
River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 3 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 42
Upper Hudson River Average Largemouth Bass (Whole Fish)
PCB Concentration (in mg/kg)
Year
Total PCB 600 g/day (srOl)
Monitored Natural Attenuation
Upper River
Average
Section 1
(RM 189)
Section 2
(RM 184)
Section 3
(RM 154)
Upper River
Average
Section 1
(RM 189)
Section 2
(RM 184)
Section 3
(RM 154)
2041
0.39
1.23
0.44
0.20
0.45
1.42
0.66
0.21
2042
0.41
1.40
0.46
0.18
0.46
1.56
0.65
0.19
2043
0.34
1.10
0.40
0.16
0.39
1.22
0.62
0.17
2044
0.28
0.83
0.37
0.15
0.32
0.88
0.55
0.16
2045
0.31
1.00
0.38
0.15
0.34
1.04
0.52
0.16
2046
0.28
0.88
0.36
0.14
0.32
0.95
0.51
0.15
2047
0.29
0.93
0.37
0.14
0.35
1.17
0.49
0.15
2048
0.31
1.07
0.37
0.13
0.39
1.42
0.50
0.15
2049
0.32
1.13
0.39
0.13
0.38
1.39
0.50
0.14
2050
0.28
0.95
0.37
0.12
0.34
1.21
0.49
0.13
2051
0.27
0.96
0.37
0.11
0.32
1.12
0.47
0.12
2052
0.25
0.82
0.36
0.11
0.29
0.98
0.44
0.12
2053
0.33
1.28
0.38
0.11
0.37
1.41
0.49
0.12
2054
0.30
1.10
0.38
0.11
0.32
1.18
0.46
0.12
2055
0.27
0.95
0.36
0.10
0.30
1.06
0.44
0.11
2056
0.29
1.04
0.41
0.10
0.32
1.16
0.45
0.11
2057
0.30
1.15
0.37
0.10
0.32
1.17
0.46
0.11
2058
0.25
0.87
0.38
0.10
0.27
0.91
0.43
0.11
2059
0.31
1.28
0.36
0.10
0.33
1.31
0.46
0.10
2060
0.25
0.89
0.35
0.09
0.26
0.93
0.40
0.10
2061
0.23
0.80
0.33
0.09
0.25
0.84
0.38
0.09
2062
0.25
0.90
0.34
0.09
0.26
0.91
0.38
0.10
2063
0.25
0.89
0.35
0.09
0.26
0.91
0.37
0.10
2064
0.25
0.92
0.36
0.09
0.27
0.97
0.38
0.10
2065
0.25
0.88
0.36
0.10
0.25
0.87
0.38
0.10
2066
0.30
1.25
0.34
0.09
0.31
1.26
0.40
0.09
2067
0.26
0.95
0.37
0.09
0.27
0.97
0.37
0.10
Notes:
1. Fish fillets multiplied by 2.5 to obtain whole fish concentrations.
2. All whole fish PCB concentrations are above target fish concentration of 0.3 mg/kg and/or 0.03 mg/kg based on the river
otter lowest-observed-adverse-effects-level (LOAEL) and no-observed-adverse-effects-level (NOAEL), respectively.
Upper Hudson River average is weighted by river section length:
River Section 1: 6.3 miles = 15.4%; River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 4 of 4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 43
Modeled Times of Compliance with River Otter
Risk-Based Fish Concentrations Upper Hudson River
River Otter - RI/FS TRVs (whole fish tissue)
LOAEL 0.3 PCBs
mg/kg
NOAEL 0.03 PCBs mg/kg
Upper Hudson River Average
2035
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2035
>2067
Total PCB 600 g/day (srOl)
2035
>2067
Monitored Natural Attenuation
2052
>2067
Upper Hudson River Section 1
>2067
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
>2067
>2067
Total PCB 600 g/day (srOl)
>2067
>2067
Monitored Natural Attenuation
>2067
>2067
Upper Hudson River Section 2
>2067
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
>2067
>2067
Total PCB 600 g/day (srOl)
>2067
>2067
Monitored Natural Attenuation
>2067
>2067
Upper Hudson River Section 3
2019
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2024
>2067
Total PCB 600 g/day (srOl)
2020
>2067
Monitored Natural Attenuation
2024
>2067
Notes:
First year in which fish target concentrations are achieved are provided.
Upper Hudson River average is weighted by river section length. River Section 1:6.3 miles = 15.4%;
River Section 2: 5.1 miles = 12.5%; and River Section 3: 29.5 miles = 72.1%.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 44
Lower Hudson River Average Largemouth Bass (Whole Fish)
PCB Concentration (in mg/kg)
Year
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
River Mile
152
River Mile
113
River Mile 90
River Mile 50
River Mile
152
River Mile
113
River Mile 90
River Mile 50
1998
7.15
5.21
3.55
3.26
7.15
5.21
3.55
3.26
1999
4.53
4.12
3.30
3.01
4.53
4.12
3.30
3.01
2000
3.81
3.56
2.93
2.73
3.81
3.56
2.93
2.73
2001
4.50
3.54
2.66
2.49
4.50
3.54
2.66
2.49
2002
3.97
3.19
2.49
2.31
3.97
3.19
2.49
2.31
2003
3.42
2.82
2.26
2.10
3.42
2.82
2.26
2.10
2004
2.42
2.26
1.97
1.89
2.42
2.26
1.97
1.89
2005
2.27
1.95
1.69
1.67
2.27
1.95
1.69
1.67
2006
2.37
1.85
1.49
1.48
2.53
1.89
1.49
1.49
2007
1.93
1.71
1.35
1.34
2.37
1.86
1.40
1.36
2008
1.54
1.41
1.22
1.20
2.33
1.77
1.33
1.25
2009
1.21
1.15
1.06
1.05
2.03
1.53
1.18
1.12
2010
1.10
1.02
0.92
0.94
2.55
1.71
1.16
1.06
2011
1.25
1.01
0.84
0.86
5.16
2.57
1.35
1.10
2012
0.92
0.86
0.75
0.77
2.17
2.06
1.38
1.13
2013
1.02
0.82
0.68
0.71
1.78
1.63
1.28
1.11
2014
0.86
0.74
0.62
0.64
1.33
1.29
1.12
1.04
2015
0.72
0.65
0.56
0.59
1.04
1.04
0.96
0.94
2016
0.55
0.53
0.50
0.53
0.76
0.78
0.79
0.83
2017
0.46
0.45
0.44
0.48
0.54
0.60
0.65
0.73
2018
0.43
0.41
0.39
0.44
0.45
0.50
0.54
0.63
2019
0.34
0.35
0.35
0.40
0.35
0.39
0.44
0.54
2020
0.42
0.35
0.32
0.36
0.42
0.37
0.38
0.46
2021
0.41
0.34
0.30
0.34
0.41
0.36
0.34
0.41
2022
0.35
0.32
0.29
0.32
0.35
0.33
0.31
0.37
2023
0.30
0.29
0.27
0.30
0.30
0.29
0.28
0.33
2024
0.32
0.28
0.25
0.28
0.32
0.28
0.26
0.31
2025
0.35
0.30
0.25
0.27
0.35
0.30
0.26
0.29
2026
0.33
0.29
0.25
0.27
0.33
0.29
0.25
0.28
2027
0.26
0.26
0.24
0.26
0.26
0.26
0.24
0.27
2028
0.24
0.24
0.23
0.25
0.24
0.25
0.23
0.26
2029
0.29
0.25
0.22
0.24
0.29
0.25
0.22
0.25
2030
0.29
0.25
0.22
0.24
0.29
0.25
0.22
0.24
2031
0.25
0.24
0.21
0.23
0.25
0.24
0.21
0.23
2032
0.25
0.24
0.21
0.23
0.25
0.24
0.21
0.23
2033
0.23
0.23
0.21
0.23
0.23
0.23
0.21
0.23
2034
0.22
0.22
0.20
0.22
0.22
0.22
0.20
0.22
2035
0.35
0.25
0.21
0.22
0.35
0.25
0.21
0.22
2036
0.48
0.32
0.23
0.23
0.48
0.32
0.23
0.23
2037
0.57
0.39
0.26
0.24
0.57
0.39
0.26
0.24
2038
0.58
0.40
0.28
0.26
0.58
0.40
0.28
0.26
2039
0.48
0.39
0.29
0.27
0.48
0.39
0.29
0.27
2040
0.43
0.37
0.29
0.27
0.43
0.37
0.29
0.27
2041
0.30
0.32
0.28
0.27
0.30
0.32
0.28
0.27
2042
0.25
0.27
0.26
0.26
0.25
0.27
0.26
0.26
2043
0.29
0.26
0.24
0.25
0.29
0.26
0.24
0.25
2044
0.35
0.28
0.23
0.24
0.35
0.28
0.23
0.25
2045
0.33
0.28
0.23
0.24
0.33
0.28
0.23
0.24
2046
0.29
0.26
0.22
0.24
0.29
0.26
0.22
0.24
Notes:
Fish fillets multiplied by 2.5 to obtain whole fish concentrations.
All whole fish PCB concentrations are above target fish concentration of 0.3 mg/kg and/or 0.03 mg/kg based on the
river otter lowest-observed-adverse-effects-level (LOAEL) and no-observed-adverse-effects-level (NOAEL), respectively.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 1 of2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 44
Lower Hudson River Average Largemouth Bass (Whole Fish)
PCB Concentration (in mg/kg)
Year
Total PCB 600 g/day (srOl)
Monitored Natural Attenuation
River Mile
152
River Mile
113
River Mile 90
River Mile 50
River Mile
152
River Mile
113
River Mile 90
River Mile 50
1998
7.15
5.21
3.55
3.26
7.54
5.30
3.55
3.24
1999
4.53
4.12
3.30
3.01
4.37
4.06
3.28
2.99
2000
3.81
3.56
2.93
2.73
4.01
3.56
2.91
2.71
2001
4.50
3.54
2.66
2.49
4.51
3.54
2.65
2.47
2002
3.97
3.19
2.49
2.31
3.91
3.17
2.47
2.28
2003
3.42
2.82
2.26
2.10
3.39
2.82
2.25
2.08
2004
2.42
2.26
1.97
1.89
2.39
2.23
1.96
1.88
2005
2.27
1.95
1.69
1.67
2.25
1.94
1.68
1.66
2006
2.49
1.86
1.49
1.49
2.34
1.86
1.49
1.47
2007
2.20
1.79
1.38
1.34
1.89
1.70
1.35
1.32
2008
1.97
1.60
1.27
1.23
1.57
1.42
1.21
1.20
2009
1.62
1.34
1.12
1.08
1.27
1.16
1.06
1.05
2010
1.73
1.30
1.02
1.00
1.36
1.13
0.94
0.95
2011
2.43
1.49
1.01
0.96
1.63
1.22
0.91
0.89
2012
1.32
1.20
0.96
0.90
1.30
1.11
0.86
0.83
2013
1.27
1.08
0.88
0.84
1.48
1.13
0.83
0.79
2014
1.01
0.92
0.78
0.77
1.27
1.03
0.79
0.74
2015
0.82
0.78
0.69
0.70
1.00
0.90
0.73
0.70
2016
0.61
0.61
0.60
0.63
0.76
0.72
0.65
0.64
2017
0.51
0.51
0.51
0.56
0.68
0.62
0.57
0.59
2018
0.47
0.45
0.45
0.50
0.65
0.58
0.51
0.53
2019
0.37
0.38
0.39
0.45
0.52
0.50
0.46
0.49
2020
0.45
0.37
0.35
0.40
0.68
0.51
0.42
0.44
2021
0.44
0.36
0.32
0.36
0.63
0.49
0.40
0.41
2022
0.37
0.34
0.30
0.34
0.51
0.45
0.38
0.39
2023
0.32
0.30
0.28
0.32
0.46
0.41
0.35
0.37
2024
0.33
0.29
0.26
0.30
0.48
0.40
0.34
0.35
2025
0.37
0.31
0.26
0.29
0.53
0.43
0.34
0.34
2026
0.34
0.29
0.25
0.28
0.48
0.40
0.33
0.33
2027
0.26
0.27
0.24
0.27
0.37
0.36
0.32
0.32
2028
0.25
0.25
0.23
0.26
0.35
0.34
0.30
0.31
2029
0.30
0.25
0.22
0.25
0.42
0.34
0.28
0.30
2030
0.30
0.25
0.22
0.24
0.40
0.34
0.28
0.29
2031
0.25
0.24
0.22
0.24
0.34
0.32
0.27
0.28
2032
0.26
0.24
0.22
0.24
0.35
0.32
0.27
0.28
2033
0.23
0.23
0.21
0.23
0.31
0.30
0.26
0.28
2034
0.22
0.22
0.20
0.23
0.29
0.29
0.25
0.27
2035
0.27
0.23
0.20
0.22
0.42
0.33
0.25
0.26
2036
0.23
0.22
0.20
0.22
0.54
0.38
0.27
0.26
2037
0.40
0.28
0.21
0.22
0.69
0.46
0.30
0.28
2038
0.65
0.38
0.24
0.23
0.65
0.47
0.32
0.29
2039
0.56
0.41
0.27
0.25
0.55
0.44
0.33
0.30
2040
0.51
0.40
0.29
0.26
0.48
0.42
0.33
0.31
2041
0.35
0.35
0.28
0.27
0.35
0.36
0.31
0.30
2042
0.29
0.30
0.27
0.27
0.28
0.30
0.28
0.29
2043
0.33
0.29
0.25
0.26
0.35
0.30
0.26
0.28
2044
0.38
0.31
0.25
0.25
0.42
0.32
0.26
0.27
2045
0.34
0.30
0.24
0.25
0.38
0.32
0.26
0.26
2046
0.30
0.27
0.23
0.24
0.33
0.30
0.25
0.26
Notes:
Fish fillets multiplied by 2.5 to obtain whole fish concentrations.
All whole fish PCB concentrations are above target fish concentration of 0.3 mg/kg and/or 0.03 mg/kg based on the
river otter lowest-observed-adverse-effects-level (LOAEL) and no-observed-adverse-effects-level (NOAEL), respectively.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 45
Modeled Times of Compliance with River Otter
Risk-Based Fish Concentrations Lower Hudson River
River Otter - RI/FS TRVs (whole fish
tissue)
LOAEL 0.3 PCBs
mg/kg
NOAEL 0.03 PCBs
mg/kg
Lower Hudson River RM 152
2027
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2027
>2067
Total PCB 600 g/day (srOl)
2027
>2067
Monitored Natural Attenuation
2034
>2067
Lower Hudson River RM 113
2023
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2023
>2067
Total PCB 600 g/day (srOl)
2024
>2067
Monitored Natural Attenuation
2034
>2067
Lower Hudson River RM 90
2021
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2023
>2067
Total PCB 600 g/day (srOl)
2023
>2067
Monitored Natural Attenuation
2028
>2067
Lower Hudson River RM 50
2023
>2067
No Resuspension (d004)
Total PCB 350 ng/L (sr04)
2025
>2067
Total PCB 600 g/day (srOl)
2024
>2067
Monitored Natural Attenuation
2029
>2067
Notes:
First year in which fish target concentrations are achieved are provided.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Table 46
Sediment Characteristics
Fraction Name
Fraction by
Weight (%)
Mass Release
Rate (kg/hr)
Representative
Grain Diameter
(mm)
Fall velocity from
Stokes
equation1 (m/s)
Comments
Middle of ASTM 1990
Sand
0.19
91.5
2
3.21
"fine sand"
Middle of ASTM 1990
Silt
0.53
255.2
0.02
3.21 x 10-4
"silt"
Middle of ASTM 1990
Clay
0.28
134.8
0.002
3.21 x 10-6
"clay"
1. Stokes equation: Fall velocity (w) = gd2/rs-r)/18m, where g is 9.81 m/s2, d is the diameter of a spherical grain (m),
rs is the density of sediment particles (kg/m3), r is the water density (999 kg/m3), and m is the dynamic viscosity
of water (1.12 x 10-3 N-s/m2 at 15.6oC). A dry density of 700 kg/m3 was assumed for all sediments.
Table 47
Impact of Dispersion Coefficient on Predicted Peak
Concentration and Length of Suspended Sediment Plume
Dispersion
Coefficient (m2/s)
Peak suspended sediment
concentration in immediate vicinity
of dredge (mg/L) - above ambient
conditions
Approximate
length of plume at
5 mg/L contour
(m)
0.1
390
900
1
72
800
10
13
120
100
2
0
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figures
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment D - April 2004
-------
CSTR-Chem
(0-10 m)
TSS-Chem
(10-1600 m)
HUDTOX
(Greater than 1 mile)
Figure 1. Interaction Among the Transport Models
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
M
S
H ffl
w u
ฆa ซ
o JS
ฐ =1
-M 75
0.001
0.0009 -
0.0008 -
0.0007 -
0.0006 -
0.0005 -
0.0004 -
0.0003 -
0.0002 -
0.0001 -
0
0.000
~
~
o
o
~
~
o
o
~~
(P
0.050 0.100 0.150 0.200 0.250
Linear Velocity (m/s)
r 1
- 0.9
- 0.8
- 0.7
- 0.6
- 0.5
- 0.4
- 0.3
- 0.2
- 0.1
0
0.300
ONet Fraction Dissolved
~Net Fraction Silt Exiting
0.001
17 0.0009
a1 a
ฆB 8 0.0008
h w
W U 0.0007
a PM
S
ฃ ง
-a <->
ฃ
0.0006 -
0.0005 -
0.0004 -
ฆง ji 0.0003 -
g | 0.0002 -
u "2 0.0001 -
z g
=- o -
5? a
S o
ฃ3 u
Xfl
W M
S H
$
S3 -S
2 3
t5 o
S3 ซ->
fe S3
z g
~
~
~
~
o
o
o
o
Sฐ a
s ฐ
a ซ 0 0008
* CQ
1*1 cj
g a! 0.0007
ฉ +*
t3 a 0.0006
ฃ | 0.0005
ฆa u
ฆa 0.0004
0.000 0.050
r 1
- 0.9
- 0.8
oฐ
0.100 0.150 0.200
Linear Velocity (m/s)
0.250 0.300
WW
0 7 .9 O
0.6 td !ซ
ซ H
0.4
0.3
0.2
a <:
.2 g
t5 o
03 U
Si
o.i z g
0.000 0.050 0.100 0.150 0.200 0.250
Linear Velocity (m/s)
1- 0
0.300
Note: Net concentrations exclude background.
Figure 2
Sensitivity of Net Dissolved and Silt Fractions Exiting Near-Field
with Variations in Linear Velocity and Depth for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
X
3
e
ca
u
a.
o
H
ts
z
0.05 n
0.045 -
0.04 -
0.035 -
_ 0.03 -
tj
& 0.025 -
M
w 0.02 -
0.015 -
0.01 -
0.005 -
0
0.000
~
~
9
~
~
6>
~~
cP
0.050 0.100 0.150 0.200 0.250
Linear Velocity (m/s)
ONet Total Cone
~Net TSS Load
1
- 0.9
- 0.8
- 0.7
0.6
m
_s
E
0.5 M
0.4
0.3
0.2
0.1
0
A
0.300
0.05 i
0.045 -
0.04 -
0.035 -
0.03 -
ฃ 0.025 -
5/5
M 0.02 -
0.015 -
0.01 -
0.005 -
0
0.000
t
@
0.4
in
H
$
0.3
z
0.2
0.1
0
0.050 0.100 0.150 0.200
Linear Velocity (m/s)
0.250 0.300
Note: Net concentrations exclude background.
Figure 3
Sensitivity of Net Total PCB Flux and Net TSS Flux Exiting Near-Field
with Variations in Linear Velocity and Depth for CSTR-Chem
1
0.9
0.8
0.7
- 0.6
n <; M
0.5
h 0.4
0.3
0.2
0.1
0
0.250 0.300
A
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.001
0.0009
2 0.0008
a a
C3 O
U ^
z ^
a u
S Ph
ii
* 0>
W C
a g
ฉ .2
'H
cj C3
C3 ^
rv i G
0.0007
0.0006
0.0005
73 r<
^ 5
> O
ซr> ~
(A CJ
5 ฃ
a ฐ
. cซ
^ !S
3 0.0004
0.0003
0.0002
0.0001
Section 1 Fraction
o
0.000
o - - Net Fraction Dissolved
~ Net Fraction Silt Exiting
*-o.
ฆ O.
O-.
ฆฉ..
ฆ0-.
0.050
0.100 0.150 0.200 0.250
Linear Water Velocity (m/s)
0.9
0.8
ฆ3 ฃ
0.7 "3 b
ฃ ง
iL u
b g
ซ 2
0-6 ฃ ฃ
bD {/j
.3 H
.IS +*
x o>
0.5 ฃ fl
G
S3 o
J/5
_ cs
0.4 .2 -ง
0.3 ซ
0.2
0.1
a> +3
ฃ S
01
8
0
0.300
m
_s
ฃ
CO
u
a.
o
H
0.03
0.025
0.02
0.015
0.01
0.005
Section 1 Flux
O- -Net Total Cone
~ Net TSS Load
0
0.000
0.050 0.100 0.150 0.200 0.250
Linear Water Velocity (m/s)
0.9
0.8
0.7
0.6
U'5 ?ซ M
0.4
0.3
0.2
0.1
0
0.300
Note: Net concentrations exclude background.
Figure 4
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Velocity for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.001
0.0009
Section 1 Fraction
Section 1 Flux
2 0.0008
a a
C3 O
U ^
z ^
a u
S Ph
ii
* 0>
W C
a g
ฉ .2
'H
cj C3
C3 ^
rv. C
0.0007
0.0006
0.0005
73 r<
ฉ
ซr> ~
(A CJ
5 ฃ
a ฐ
y cซ
^ !S
3 0.0004
0.0003
0.0002
0.0001
0.00
o - - Net Fraction Dissolved
~ Net Fraction Silt Exiting
.ฉ-
.O-'
JO-
JO
ฉ'
0.50
1.00 1.50
Depth (m)
2.00
0.9
0.8
ฆ3 ฃ
0.7 "3 b
ฃ ง
iL u
b S3
ซ 2
0-6 fc ฃ
bD {/j
.3 H
.IS
x o>
0.5 ฃ ซ
c
S3 O
J/5
_ C3
0.4 .ฉ -ง
0.3 ซ
0.2
0.1
2.50
& ii
ฃ S
o>
c
0.03
0.025
0.02
x
a
E
PQ
S ฃ 0.015
-2 ฃ5
o
H
0.01
0.005
O- -Net Total Cone
~ Net TSS Load
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.00
0.50
1.00 1.50
Depth (m)
2.00
2.50
Note: Net concentrations exclude background.
Figure 5
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Depth for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.03
fl 0.025
o
2 S
ai
CO
fl y
s ซ
H o
ฐซ
ซr> ~
sซ CJ
s ฃ
^ *8
0>
c
0.02
0.015
0.01
0.005
0 -)
Section 1 Fraction
Q
Q
O
Q'
.O'
DO-?'
- - O - -Net
Fraction
Dissolved
~ Net
Fraction
Silt Exiting
0.8
0.7 .2
0.5
10 20 30 40 50 60 70
Near-Field Width (m)
90 100
ฆa 2
0.6 Z
in
M !ซ
.5 H
s +*
H 2
w -5
ซ s
a o
0.4 o
0.3
0.2
0.1
0
03
-
Z -a
Section 1 Flux
0.005
20 40 60
Near-Field Width (m)
100
Note: Net concentrations exclude background.
Figure 6
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Near-Field Width for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Section 1 Fraction
Section 1 Flux
0.001
0.0009
s
S 0.0008
C3
S *->
ai
0.0007
C3
-
X 73
s ^
y 1*1
^ -3
bD (J
.5 Ph 0.0006
'B "5
o ง 0.0005
O
0.0004
0.0003
g 0.0002
s
0.0001
0
ฆO - -Net Fraction Dissolved
~ Net Fraction Silt Exiting
'Q-0-0-Q.Q
o
10 20 30 40 50 60 70 8(
Resuspension Rate (kg/sec)
0.9
0.8
ฆa
0.7 ซ
tu
0.6 Z
iz
gp
0.2 S
0.7
0.6
0.5
*
_3
^ 0.4
n
- -I
0
H 0.3
01
Z
0.2
0.1
O- -Net Total Cone
~ Net TSS Load
25
20
15
10
m
_s
E
VI
VI
H
ts
Z
M
A
10 20 30 40
Resuspension Rate (kg/sec)
50
Note: Net concentrations exclude background.
Figure 7
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Resuspension Rate for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.001
0.0009
Section 1 Fraction
Section 1 Flux
s-
r ฎ
s- a
C3 o
ซ u
fl U
s *
II
* 0>
W C
a g
ฉ .2
*-C
cj cs
C3 *-<
fV I S
o
ซr> ~
(A CJ
5 ฃ
a ฐ
y cซ
55 ^3
4)
a
2 0.0008
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
0.0001
- O - - Net Fraction Dissolved
~ Net Fraction Silt Exiting
0.00 0.20 0.40 0.60 0.80
Sediment Silt Fraction
1.00
a
_o
2 2
a
tu S
iL u
b s
a
u S
go iซ
S H
s +*
H 2
w -5
ซ a
a o
03
-
6t
Z -a
_s
E
CO
u
a.
o
H
0.05
0.045
0.04
0.035
0.03
0.025
0.02
0.015
0.01
0.005
o- -Net Total Cone
~ Net TSS Load
0.00 0.20 0.40 0.60 0.80
Sediment Silt Fraction
1.00
0.9
0.8
0.7
0.6 ,
C3
O .
-J (J
4>
n c Z/l jg
ฃ ^
4>
0.4 Z
0.3
0.2
0.1
1.20
Note: Net concentrations exclude background.
Figure 8
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Sediment Silt Fraction for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Fractions
Section 1 Fluxes
ฃ a
C3 O
ซ U
z ^
a u
S Ph
ii
* 4>
W C
a g
ฉ .2
*-C
cj cs
C3
rv i B
o
5ซ ~
(A CJ
5 ฃ
a ฐ
y cซ
^ !S
0.0004
0.00035 -
ฆ3 ง 0.0003
0.00025
0.0002 -
0.00015
0.0001
0.00005 -
0 -4
0
"Section 1 Variation
Section 1 Value
Section 2 Variation
Section 2 Value
ฆ Section 3 Variation
Section 3 Value
0.6
0.5
0.4
m
_s
E
CO
ฃ ^0.3
-S 3
o
H
0.2
0.1
ฆNet Total Cone
"Net TSS Load
200 400 600 800 1.000
PCB Sediment Concentration (mg/kg)
200 400 600 800
PCB Sediment Concentration (mg/kg)
0.9
0.8
0.7
0.6 _
C3
ฐ
1-1 u
OS ^
' H J?
1.000
Note: Net concentrations exclude background.
Figure 9
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of PCB Sediment Concentration for CSTR-Chem
Ol
0.4
0.3
0.2
0.1
0
Hudson River PCBs Superfiind Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.001
0.0009
Section 1 Fraction
Section 1 Flux
2 0.0008
a a
C3 O
U ^
z ^
a u
S Ph
ii
* 0>
W C
a g
ฉ .2
'H
cj C3
C3 ^
rv. C
0.0007
0.0006
0.0005
73 r<
ฉ
ซr> ~
(A CJ
5 ฃ
a ฐ
y cซ
^ !S
3 0.0004
0.0003
0.0002
0.0001
0.00
o - - Net Fraction Dissolved
~ Net Fraction Silt Exiting
~ ~
~ ~ ~ ~
ฉ- o o O-O- O- o. -O-I
oo
0.9
0.8
0.7 u -B
0.5
V A
0.6 ฃ
i/5
bD x/i
.3 H
.IS +*
w 5
c
S3 O
J/5 'U
_ cs
0-4 .2 "B
ฆO a
0.3 ซ
0.2
0.1
Z -a
0>
c
0.03
0.025
0.02
x
a
E
PQ
* ฃ 0.015
-2 ฃ5
o
H
0.01
0.005
ฆ6~o--o
~ ~ ~ ~O _
O-O -O -O - -Q-OO 0.5 $
O- -Net Total Cone
~ Net TSS Load
0
0.9
0.8
0.7
0.6
0.4
01
Z
0.3
0.2
0.1
1.00
0.00
0
0.20 0.40 0.60 0.80 1.00
Dissolved PCB Fraction in Background
0.20 0.40 0.60 0.80
Dissolved PCB Fraction in Background
Note: Net concentrations exclude background.
Kd was held constant at 5.500 L/kg and Background TSS was varied from 0.5 to 40 mg/L.
Figure 10
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Dissolved PCB Fraction in the Background and TSS Background Concentrations for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Section 1 Fraction
Section 1 Flux
0.0025
s
.2 0.002
C3
-
73
r ฎ
^ 8
s- a
cs o
ซ u
fl u
S *
ii
* 0>
w a
a g
.2 -2
0.0015
0.001
5 ฃ
a ฐ
55 ^3
u 0.0005
s
ฆO - - Net Fraction Dissolved
~ Net Fraction Silt Exiting
*o.
- 0.9
0.8
ฆa 2
0.7
0.6 Z
M
S
0.5
1000
10000
100000
1000000
0.4 .2
03
-
- 0.3 ซ
- 0.2
- 0.1
0.03
0.025
0.02
_s
E
CO
u
a.
o
H
a>
z
0.01
0.005
^ ฃ 0.015 -O O G
O- -Net Total Cone
~ Net TSS Load
0.9
- 0.8
0.7
- 0.6 _
cs
O .
hJ
4>
- 0 S ^ 45
U !> ^ ^
H &
Ol
- 0.4
0.3
- 0.2
0.1
10000
100000
Kd Value (L/kg) - Log Scale
1000000
K,, Value (L/kg) - Log Scale
Note: Net concentrations exclude background.
Background TSS was held constant at 2.3 mg/L
Figure 11
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Dissolved PCB Fraction in the Background and Kd Value for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.001
0.0009
e
2 0.0008
2
"3
'&
u
03
ซ
^ 03
fl V,
s ซ
H
0.0007
0.0006
0.0005
ฆa _
ฃ ง
ฐ ซ
5ซ ~
sซ CJ
s ฃ
^ "3
3 0.0004
0.0003
0.0002
0.0001
Section 1 Fraction
ฆO - -Net Fraction Dissolved
~ Net Fraction Silt Exiting
.ฉ
.ฉ
.ฉ
.ฉ
-O-'
>.ฉ
.ฉ
.ฉ
-ฉ-,
.ฉ
ฉ
.ฉ
-ฉ-'
CfSP
-ฉ-'
0.9
0.8
0.7
2 2
- a
fe 8
0.6 Z
0.5
in
gp iซ
.5 H
a ซ
h g
w -5
in
0.4 ง
0 0.3 ซ
z -s
0.2
0.1
0.05
0.1
0.15
0.2
Desporption Rate (hr')
Section 1 Flux
0.05
0.045
0.04
0.035
m
2 0.03
En
CO
i ฃ 0.025
ซ M
O
H
"S 0.02
0.005
O- -Net Total Cone
~ Net TSS Load
0 015 -QQQ-Q -O -O -O -O -O -O -O -O -O -O -O -O -O -O -O -O -O
0.01
0.9
0.8
0.7
0.6 _
03
> ฐ
-J (J
4>
OS M ^
ฐ-5 ฃ -s>
0.4
H A
4>
z
-GD.3
0.2
0.1
0.05 0.1 0.15
Desporption Rate (hr-1)
0.2
Note: Net concentrations exclude background.
Figure 12
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Desorption Rate for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.001
0.0009
Section 1 Fraction
Section 1 Flux
2 0.0008
2 E
E S
I U
CS o
ซ S
^ 03
18
0.0007
0.0006
H ซ
o ง 0.0005
C3
-
X 73
s ฃ
y 1*1
^ -3
0.0004
0.0003
u 0.0002
s
0.0001
- - O - - Net Fraction Dissolved
~ Net Fraction Silt Exiting
~~~~~~
o
100 200 300 400 500
Background PCB (ng/L)
0.9
0.8
ฆ ฐ-7 3
5 2
ง
ss ฎ
0.6 z ^
.|f H
.-*2 ~
3 o>
-0.5 w ซ
c
ฃ3 o
J/5 *-C
_ cs
0.4 .2 -g
C3
-
0.3 3 ~
ฃ ^
Ol
S
0.2 w
0.1
600
0.05
0.045
0.04
0.035
0.03
m
_s
E
CO
- j* 0.025
0
H
01
Z
0.005
~~~~~~
0.02
0.015 40000088000 OOOOOO O O
0.01
O- -Net Total Cone
~ Net TSS Load
0 100 200 300 400 500
Background PCB (ng/L)
- 0.9
0.8
- 0.7
"0.6 ,
cs
O .
hJ
4>
- 0 S ^
H A
01
- 0.4
0.3
0.2
- 0.1
600
Note: Net concentrations exclude background.
Figure 13
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Background PCB Concentration for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.001
0.0009
2 0.0008
2 E
01
S- fl
03 o
ซ s
^ 03
fl ^
s ซ
H o
ฐ ซ
ซr> ~
sซ <3J
s ฃ
^ -3
0.0003
0.0002
0.0001
0
Section 1 Fraction
ฆO - -Net Fraction Dissolved
~ Net Fraction Silt Exiting
O- . . O-
o- - - o
0.9
0.8
5 2
0.7 "3
E 41
sL u
ซ a
0.6 Z ฃ
.|f H
.-*2 ~
3 o>
0.5 w 5
e
ฃ3 o
JZ5 *-C
_ 03
0-4 .2 -g
u
03
t-
fe
0.3 3 ~
ฃ ^
01
8
0.2 w
0.1
0
0.05
0.045
0.04
0.035
m
2 0.03
En
CO
i ^0.025
-S ฃ5
o
H
u 0.02
z
0.015
0.01
0.005
0
2.0E-05
Section 1 Flux
-G - - -G G G ฆ
O- -Net Total Cone
~ Net TSS Load
-O- - -O
4.0E-05 6.0E-05 8.0E-05 1.0E-04
Silt Settling Velocity (m/s)
0.9
0.8
0.7
0.6
ฆa
03
0-5 S g
l?
0>
0.4 Z
0.3
0.2
0.1
2.0E-05 4.0E-05 6.0E-05 8.0E-05 1.0E-04 1.2E-04
Silt Settling Velocity (m/s)
Note: Net concentrations exclude background.
Figure 14
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Silt Settling Velocity for CSTR-Chem
o
1.2E-04
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
0.001
0.0009
s
.2 0.0008
Section 1 Fraction
Section 1 Flux
s-
r ฎ
s- a
CS o
ซ u
fl U
S *
ii
* 0>
W C
a g
ฉ .2
*-C
cj C3
C3 ^
rv i B
T3 ซ
^ s
> ฉ
arI ~
(A CJ
5 ฃ
a ฐ
. cซ
55 ^3
0>
B
0.0007
0.0006
0.0005
0.0004
0.0003
0.0002
0.0001
ฆO - -Net Fraction Dissolved
~ Net Fraction Silt Exiting
ฉ- '
.Q-'
O- - -o- - -ฐ
Q- -
0.02 0.04 0.06 0.08
Coarse Settling Velocity (m/s)
0.9
0.8
0.7
0.6 Z
0.1
ฆa 2
in
M !ซ
S H
S +*
3 oi
0.5 w ^
ซ s
a o
0.4 o
0.3
0.2
0.1
03
-
Z -a
_s
E
CO
o
H
0.05
0.045
0.04
0.035
0.03
0.025
<5 0.02
z
0.015
0.01
0.005
*-o
- -o.
-o.
o. . o
O- -Net Total Cone
Net TSS Load
0.02 0.04 0.06 0.08
Coarse Settling Velocity (m/s)
0.9
0.8
0.7
0.6 ,
C3
O .
-J (J
4>
n c m
H &
4>
0.4 Z
0.3
0.2
0.1
0.1
Note: Net concentrations exclude background.
Figure 15
Net Dissolved PCB Fraction, Net Silt Fraction, Net Total PCB Flux and Net TSS Flux Exiting Near-Field
as Functions of Coarse Settling Velocity for CSTR-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 16
Estimated TSS Concentration Downstream of the Dredge Head in Section 1
(Flow is 4000 cfs and PCB concentration is 500 ng/L at the far field station)
Figure 17
Estimated TSS Concentration at 300 m Downstream of the Dredge Head
in Section 1 (PCB concentration at the far-field station is 500 ng/L)
hJ
"Sid
S
a
#o
""83
-
c
o
o
c
o
u
in
m
H
Cross River Distance, m
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Piniie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Total Organic Carbon (%)
Note: 1) See text for discussion
Source: TAMS/Gradient Database, Release 3.5 TAMS
Figure 18 (Figure 3-21 of LRC Report)
Total PCBs Grouped by Total Organic Carbon
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
350
Figure 19
Grain Size, Organic Content and PCB Concentrations
in Hudson River Sediment collected near Moreau
<69
300-
293-900
250"
200-
69-123
123-293 nm
900-3327 nm
150 -
100"
50-
T
2
T
4
T
7
% TOC
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5 n r 70
0.45
0.4
s
a>
CJ
c
o
ซJ
fi pa
.2 u
"2 ^
> -2
5 8
01 =
> o
CJ
0.35
ฆa
Ol
>
0.3
0.25
0.2
0.15
3 01
s
0.05
B Fraction Dissolved at 1600m
~Distance downstream (coarse < 0.1%)
60
50
40
30
20
10
2000
4000
6000
8000
0
10000
Notes: Flow (cfs)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
C3
O
U
.c
u
2
*
C3
s
C3
O
<*3
4>
ฃ
o
4>
-
C3
900
800
700
600
x
a
E
PM ^
S "5b
o w
H
"5
z
500
400
300
200
100
Fluxes
~
ฆ
~
ฆ
~
~ ฆ
ฆ
*
1
ฆ
~
BTCB Flux at 1600m
*SS Flux at 1600m
30000
25000
20000
ฆa
03
O ^
15000 $ 5
zl bx
10000
5000
2000
4000 6000
Flow (cfs)
8000
10000
Figure 20
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Riverwide Volumetric Flow (Velocity-Depth Pairs) for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
Fluxes
0.5
0.45
ฆ5 0.4
s
o
(j
a pa
.2 5
I ^
- -g
ฆa
0.35
0.3
s
.2 0.25
0.2
0.15
ฆg 0.1
0.05
- -O - - Fraction Dissolved at 1600m
~ Distance downstream
(coarse < 0.1%)
70
- 60
50
0.05
0.1
0.15
0.2
0.25
0.3
Notes: Velocity (m/s)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
40
30
20
10
o
Vi
4>
Z
900
800
700
600
m
_s
E
V >>
0H ซ
S "Slo
o w
500
400
01
Z
300
200
100
O- - TPCB Flux at 1600m
~ TSS Flux at 1600m
30000
25000
20000
ฆa
03
O
hJ
15000 (Jj
10000
5000
03
"Si
0.05
0.1
0.15
0.2
0.25
0.3
Velocity (m/s)
Figure 21
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Velocity for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5 n r 70
Fluxes
0.45
s
2 0.4
e
o
cj
fi pa
.2 5
I ^
ฃ "5
0.35
0.3
*e
o>
>
e
o 0.25
2 2 0.2
e
o
CJ
73
4>
0.15
0.1
0.05
ฆO - -Fraction Dissolved at 1600m
~ Distance downstream (coarse <0.1%)
O o
- - - -O
60
50
40
30
0.5
1.5
2.5
Notes: Depth (m)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
03
O
U
.S
(j
IS
ฃ
20 "X iซ
10
900
800
700
600
_s
E
U ^
S "5b
o w
500
400
01
Z
300
200
100
O- - TPCB Flux at 1600m
-TSS Flux at 1600m
30000
25000
20000
ฆa
03
O
hJ
15000$
10000
5000
03
"5b
a
0.5
1 1.5
Depth (m)
2.5
Figure 22
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Depth for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5 t r 70
Fluxes
a pa
.2 5
u Ph
a
> -2
_
0 cs
-ฆ* J-
(*)
S I
01 a
> o
CJ
ฆa
01
0.45
8
.2 0.4
. . jO -
- Fraction Dissolved at
1600m
~
Distance downstream
(coarse < 0.1%)
0.35
0.3
0.25
0.2
ฃ 0.15
0.1
0.05
fm~-
b
o
o
o
'o
O
60
50
40
30
03
o
U
.S
tj
IS
is
C3
S
C3
0 5 10 15 20 25 30
Notes: g (source strength) (kg/s)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
35
40
20
10
O
Ifl
4>
ฃ
O
4>
U
C3
25000
O- - TPCB Flux at 1600m
~ TSS Flux at 1600m
20000
X
_s
e
V >>
0h ซ
S "5b
o w
H
15000
10000
5000
900000
600000
700000
600000
ฆa
50000ง
hJ
in
VI
H
400008
z
300000
200000
100000
03
"5b
a
5 10 15 20 25 30 35
g (source strength) (kg/s)
Figure 23
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Source Strength for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5 r 70
Fluxes
0.45 0
s
o
CJ
fi pa
.2 u
fe a
0.35
*e
o>
ฃ
ฉ
.2 0.25
S I
oi =
>ฆ ฐ
CJ
ฆa
Ol
j;
o
ฃ 0.15
0.05
. . JO -
- Fraction Dissolved at
1600m
~
Distance downstream
(coarse < 0.1%)
0 0.2 0.4 0.6 0.8
Notes: Fraction of Silt Entering (unitless)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
- 60
900
800
700 -
600 -
O- - TPCB Flux at 1600m
~ TSS Flux at 1600m
X
E
U >5
^ pS
o w
H
o>
z
500
400 -
300 -
200
100 -
0000
25000
20000
ฆa
03
O ^
h-1
15000 % 5
H g
+* vw
0>
z
10000
5000
0 0.2 0.4 0.6 0.8 1
Fraction of Silt Entering (unitless)
Figure 24
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Silt Fraction Entering for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5
0.45
s
.2 0.4
s
o
(j
a pa
.2 5
I ^
- -g
ฆa
0.35
0.3
s
.2 0.25
0.2
> 0.15
0.1
- -O - - Fraction Dissolved at 1600m
~ Distance downstream (coarse < 0.1%)
MM*-*
JO
-O-
-O
-~
o
0.05
0 200 400 600 800 1000
Notes: PCB Sediment Concentration (mg/kg)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
70
60
50
40
30
1200
20
10
O
in
4>
ฃ
03 .
ซ O
0>
5ฃ U,
S3 cs
& G
M 0>
g
CJ .S
C 73
C3 ซ
25000
20000 -
E
PQ
U
15000
5?
5
$ 3
o
H
10000 -
5000
Fluxes
O- - TPCB Flux at 1600m
~ TSS Flux at 1600m
~o
30000
25000
20000
-a
C3
0 /-v
15000 5
rj bx
01
z
- 10000
- 5000
, , , , 1- 0
0 200 400 600 800 1000 1200
PCB Sediment Concentration (mg/kg)
Figure 25
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Sediment PCB Concentration for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5 t r 70
Fluxes
s
0>
CJ
C
0
ซJ
fi pa
.2 5
I ^
ฃ "5
"3 fi
ฃ -2
0 cs
cซ S-
s I
01 =
Zo
u
ฆa
01
>
"o
0.45
s
.2 0.4
0.35
0.3
0.25
0.2
ฃ 0.15
0.1
. . o -
- Fraction Dissolved at
1600m
Distance downstream
(coarse < 0.1%)
- 60
50
~~ ~-
~ ~ ~ ~
jO-ฐ
.o-
O'
0.05
0 0.2 0.4 0.6 0.8
Notes: Background Dissolved Fraction (unitless)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
40
30
20
10
I Z
^ SI-
'S ฐ
s?
9 ฃ
O
a -a
03 O)
900
800
700 -
600 -
\o
~~-
20000
_s
E
w _
u >>
0H ซ
S "Slo
o w
H
ฆ -O-OC)
500
400 -
300 -
200
100 -
O- - TPCB Flux at 1600m
-TSS Flux at 1600m
30000
25000
-a
C3
O /-s
hJ %
15000 % 5
H g
+* vw
0>
z
10000
5000
0.2 0.4 0.6 0.8 1
Background Dissolved Fraction (unitless)
Figure 26
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of TSS Background and PCB Dissolved Fraction (Kd = 55,000) for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5
0.45 -
s
.2 0.4
s
a>
CJ
e
o
ซj
fi pa
.2 5
I ^
ฃ "5
"S fi
ฃ -2
0 cs
cซ S-
s I
01 =
Zo
u
0.35
0.3 -
0.25 -
0.2
0.15 -
0.1 -
0.05
- -O - - Fraction Dissolved at 1600m
~ Distance downstream
(coarse < 0.1%)
~ ~-
0 -I 1 1
1000 10000 100000
Notes: Partition Coefficient (L/kg)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
70
60
Fluxes
Ol
-50 2
ง 3
u -3
.s <*>
U -M
"S ^
-9 z
-40 ^ s*.
a ฃ
30
fi cs
(A
O +S
o ง
ซ g
O) .3
a -a
03 O)
20 3
10
1000000
900
800 -
700 -
600
X
a
E
PM ^
o
-Q.
-o.
-~
-o
500 -
400
300 -
200
100
30000
- 25000
20000
ฆa
03
0
15000 $ 5
Zl sx
01
z
- 10000
- 5000
O- - TPCB Flux at 1600m
-TSS Flux at 1600m
0 -I , , 1- 0
1000 10000 100000 1000000
Partition Coefficient (L/kg)
Figure 27
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Kd for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5
0.45
a
2 0.4
e
o
cj
fi pa
.2 5
I ^
ฃ "5
0.35
0.3
*e
o>
>
e
o 0.25
2 2 0.2
e
o
CJ
73
4>
0.15
0.1
0.05
- - -Q - - Fraction Dissolved at 1600m
~ Distance downstream
(coarse <0.1%)
Vฉ'
vO'
-O
.ฉ
.O-
.o-
G'
O'
70
60
50
40
0 0.05 0.1 0.15
Notes: Desoprtion Rate 1/hr
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
0.2
03
o
U
X
(j
IS
ฃ
30
a e3
o "S
O ง
a -a
53
20 -S <*>
10
Fluxes
900
800
700 -
30000
- 25000
600 -<
_s
E
CO
u
a.
o
H
~ ~~~~~~~~~~~~~~~~~
^QO-O-O-O-O-O-O-O-O-O-O-O-O-O-O-O-O-O-O
500
03
-s
400 -
300 -
200
100 -
O- - TPCB Flux at 1600m
TSS Flux at 1600m
ฆtfOOOO
O
- 15000
C3
O
hJ
J/5
J/5
H
4>
ฃ
C3
"bh
- 10000
- 5000
0.05 0.1 0.15
Desoprtion Rate 1/hr
0.2
Figure 28
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Desorption Rate for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5
a
o
(j
a pa
.2 5
I ^
"5
a
0.45
8
.2 0.4
0.35 -
0.3
En
ฆ3 a
> .2 0.25
oi a
Zo
u
ฆa
Ol
>
> 0.15
0.1
0.05
-O - - Fraction Dissolved at
1600m
~ Distance downstream
(coarse < 0.1%)
70
- 60
ocd- q-oo
- 50
f
$
&
O'
0
0.0001
- 40
30
20
10
0.001 0.01 0.1 1
Notes: Lateral dispersion mA2/s
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
10
100
O 3
U -3
ซ-ป **
f ฃ
fi
ฐ H
p 5
V g
CJ .S
C 73
C3 <3J
Fluxes
900
800 -
700
600
E
CP
u
O
H
Q -O-OEฎ
500
03
ฆa
400
300
200
100 -
O- - TPCB Flux at 1600m
~ TSS Flux at 1600m
0 -I 1 1 1 1
0.0001 0.001 0.01 0.1 1
Lateral dispersion mA2/s
30000
25000
ฆ 20000
-a
C3
O /-s
hJ %
15000 % s
H $
0>
z
10000
5000
10
100
Figure 29
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Lateral Dispersion for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5
0.45
s
2 0.4
c pa
.2 u
"2 ^
a
> -2
S |
01 =
> o
CJ
ฆa
Ol
>
0.35
0.3
0.25
0.2
ฃ 015
0.1
0.05
- - -O - - Fraction Dissolved at
1600m
~ Distance downstream
(coarse < 0.1%)
~~~~~~
*G
70
60
- 50
40
30
- 20
10
0 100 200 300 400
Notes: PCB Background (ng/L)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
500
600
O
in
4>
ฃ
O
4>
U
C3
Fluxes
900
800
700 -
30000
600
~~~~~~
m
_s
E
V >>
0H ซ
S "Slo
o w
H
4>
z
-oooooo.
o
500 -
400 -
- 25000
20000
*e
C3
o /-s
hJ %
15000 % s
H $
300
200 -
100
10000
5000
O- - TPCB Flux at 1600m
TSS Flux at 1600m
100 200 300 400 500
PCB Background (ng/L)
600
Figure 30
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of PCB Background Concentration for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5
0.45
0.35
a pa
.2 5
u Ph
a
> -2
_
0 cs
-ฆ* J-
(*)
S I
01 a
> o
CJ
ฆa
Ol
>
0.3
0.25
0.2
0.15
ฆa
-g 0.1
a
0.05
0
ฆ Fraction Dissolved at
1600m
"Distance downstream
(coarse < 0.1%)
O-
. JO- -" "
.O
~
70
60
0>
!*!
50^
o
U
40
30 a
>
o
O
20"
10
0
0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 0.0001 0.00011
Notes: Settling Velocity of silt (m/s)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
O
5/5
4>
ฃ
4>
-
(Z5
Fluxes
900
800
700
600
E
U >5
^ pS
S "5b
o w
H
500
400
300
200
100
- O- - TPCB Flux at 1600m
TSS Flux at 1600m
0
0.00004 0.00005 0.00006 0.00007 0.00008 0.00009 0.0001 0.00011
Settling Velocity of silt (m/s)
30000
25000
20000
a
ฉ /-N
hJ %
"OCR ^
H $
4)
Z
10000
5000
0
Figure 31
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Silt Settling Velocity for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Dissolved Fraction and Distance to 0.1% Coarse
0.5
Fluxes
s
o
(j
a pa
.2 5
I ^
- -g
ฆa
s
01
(J
S
0
(J
ฆa
01
>
"o
0.45
ฆS 0.4
0.35
0.3
s
.2 0.25
0.2
0.15
ฃ 0.1
0.05
-ฆฆO-
- Fraction Dissolved at
1600m
~
Distance downstream
(coarse < 0.1%)
O- - Q- - ฉ- -
70
- 60
50
40
30
0.02
0.04
0.06
0.08
0.1
Notes: Settling Velocity of sand (m/s)
1. Net concentrations exclude background.
2. Fluxes are based on 14 hours per day.
- 20
10
o
in
4>
ฃ
900
800
700
600
*
a
E
& ^
PM ^
S "5b
o w
500
400
01
Z
300
200
100
-Q . -o - --O - -o - --o - - o
O- - TPCB Flux at 1600m
TSS Flux at 1600m
30000
25000
20000
15000
*e
C3
0 ^
J s-
5/5 "Sri
c_, OJj
S e
01
10000
5000
0.02
0.04
0.06
0.1
Settling Velocity of sand (m/s)
Figure 32
Net Dissolved PCB Fraction, Distance to Coarse < 0.1%, Net Total PCB Flux and Net TSS Flux
at 1600 meters as Functions of Sand Settling Velocity for the TSS-Chem
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Distance Downstream (m)
Figure 33
PCB Concentrations Downstream of Dredge for 350 ng/L scenario
Section 1 at 1 mile and 3 miles
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 34
Whole Water Total PCB Concentration for Different 350 ng/L Input Formulations
Thompson Island Dam
'15 per. Mov. Avg. (d006-350 ng/1 @ lmile)
'15 per. Mov. Avg. (d007-350 ng/1 @ 3 miles)
"15 per. Mov. Avg. (sr03-350 ng/1, no solids)
'15 per. Mov. Avg. (sr04-350 ng/1, fraction
remaining adjusted)
Note:
Lines represent 15 day
moving averages.
Date
700
ox
ฃ
s
o
e
600
500
400
300
i
u
-------
Figure 35
Tri+ PCB Cumulative Load for Different Dredging Scenarios
Thompson Island Dam
bl
O
iJ
P3
U
Oh
+
"MNA (P3NAS2)
"No Resuspension (d004)
-Total PCB 350 ng/L @ 1 mile (dOO6)
-Total PCB 350 ng/L @ 3 mile (d007)
"Total PCB 350 ng/L fraction remaining adjusted (sr04)
"Total PCB 350 ng/L with no Solids (sr03)
Total PCB 600g/day (srOl)
"Total PCB 300 g/day (sr02)
"Accidental Release (srAl)
2005
2015
2025
2035
2045
2055
2065
2075
Year
Srhnvlerville
2075
Year
WatprfnrH
2075
Year
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 36
Total PCB Cumulative Load for Different Dredging Scenarios
Thompson Island Dam
Sf 2500
6-
"MNA (p3nas2)
"No Resuspension (d004)
"Total PCB 350 ng/L @ 1 mile (d006)
"Total PCB 350 ng/L @ 3 mile (d007)
"Total PCB 350 ng/L fraction remaining adjusted (sr04)
"Total PCB 350 ng/L with no Solids (sr03)
- Total PCB 600g/day (srOl)
"Total PCB 300 g/day (sr02)
"Accidental Release (srAl)
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 37
Whole Water, Particulate, and Dissolved Total PCB Concentrations for the 350 ng/L Dredging
Scenario (sr04)
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 38
Whole Water, Particulate and Dissolved Total PCB Concentration for Control Level
Total PCB Flux Dredging Scenario (srOl)
- 600 g/day
Thompson Island Dam
o
o
o
o
r-
o
r-
o
r-
o
r-
o
00
o
00
o
00
o
00
o
C\
o
On
o
On
o
On
o
o
o
O
o
o
ฉ
Date
i>
o
O
ฉ
Date
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 39. Tri+ PCB and Total PCB Cumulative Load for 600 g/day (srOl) Scenario
Total PCB Cumulative load
Schuylerville
Year
Total PCB Cumulative load
Waterford
Year
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 39 (Cont'd). Tri+ PCB and Total PCB Cumulative Load for 600 g/day (srOl)
Scenario
Tri+ Cumulative Load
2005 2015 2025 2035 2045 2055 2065
Year
Tri+ Cumulative Load
Year
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 40
HUDTOX Forecast of Whole Water, Particulate, and Dissolved Total PCB Concentrations for
Evaluation Level - 300 g/day Scenario
180
Thompson Island Dam
15 per. Mov. Avg. (Whole Water)
15 per. Mov. Avg. (Particulate)
15 per. Mov. Avg. (Dissolved)
00
o
OS
o
OS
o
OS
o
OS
o
o
o
o
o
-------
Figure 41
Comparison Between Upper Hudson River Remediation Scenario (Various Export Rates) and
Monitored Natural Attenuation (MNA) Forecast for Thompson Island Dam, Schuylerville, and
Waterford
2008
2048
2058
2068
Thompson Island Dam
2018 2028 2038
Year
U
* t?
+ m 30
M S
J 25
S ? ..
0
1998
MNA (p3nas2)
'No-Resuspension (d004)
ฆTotal PCB 300 g/day (sr02)
ฆTotal PCB 600 g/day (srOl)
'Total PCB 350 ng/L(sr04)
2008
2018
2048
2058
2068
40
S 35
* tJ
+ gJ03O
m S
Si I25
S
o
1998 2008 2018 2028 2038 2048 2058 2068
Year
y 33
^ 3
.1 WS0
- s
a!25
2 1
% ฃ20
ซ< 5
ซ al5
s o
5 ^
= 10
5
0
1998
2028 2038
Ypar
Waterford
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 42
Total PCB Concentrations at Waterford for the Accidental Release Scenario
"Whole Water
ฆDissolved Phase
Particulate phase
5/31/2011
6/30/2011
7/30/2011
8/29/2011 9/28/2011
Date
10/28/2011
11/27/2011
12/27/2011
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 43
Composite Fish Tissue Concentrations for the Upper Hudson River
Composite Fish - River Section 1 (RM 189)
Year
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Page 1 of 2
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 43 (Cont.)
Composite Fish Tissue Concentrations for the Upper Hudson River
a
u
a.
+
a
H
=
6.0
5.0
4.0
3.0
2.0
1.0
0.0
2005
Composite Fish - River Section 3 (RM 154)
MNA (p3nas2)
No Resuspension (dOQ4)
Total PCB 350 ng/L (sr04)
Total PCB 600 g/day (srOl)
2010
2015
2020
2025
Year
Composite Fish - River Section 3 (RM 154)
1.0
$
0.8
0.6
09
U
Ph
H 0.4
0.0
2015
1
2020
Year
2025
Notes:
Fish composite is 47% largemouthbass + 44% brown bullhead + 9% yellow perch
The bottom figure is portion of the top figure.
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Figure 44
Composite Fish Tissue Concentrations for the Lower Hudson River
Composite Fish - RM 113
2005 2010 2015 2020 2025 2030 2035 2040 2045
Year
Note:
Fish composite is 47% largemouthbass + 44% brown bullhead + 9% yellow perch
Hudson River PCBs Superfimd Site Malcolm Pimie/TAMS-Earth Tech
Engineering Performance Standards Page 1 Of 2 Volume 2: Attachment D - April 2004
-------
Figure 44 (Cont.)
Composite Fish Tissue Concentrations for the Lower Hudson River
Year
Composite Fish - RM 50
Year
Note:
Fish composite is 47% largemouthbass + 44% brown bullhead + 9% yellow perch
Hudson River PCBs Superfimd Site
Engineering Performance Standards
Page 2 of 2
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
Moonpool Samples Included
30,000-
25,000-
20,000-
g 15,000-
10,000-
5,000-
> Total PCBs
>
* t *
1
li
1
0-
-2000 -1500 -1000 -500 0 500 1000 1500 2000
Distance from Dredge (ft)
30,000-
25,000-
20,000-
CQ
15,000-
10,000-
5,000-
Particulate Phase
*
U t *
S 1
a> *
30,000
25,000-
j 20,000
0-
-2000 -1500 -1000 -500 0 500 1000 1500 2000
Distance from Dredge (ft)
5 10,000
5,000-
0
-
Dissolved Phase
ป
f f
m
-2000 -1500 -1000 -500 0 500 1000 1500 2000
Distance from Dredge (ft)
Expanded
5,000
4,000
3,000
CQ
U
2,000-
1,000
n i r
-2000 -1500 -1000 -500 0 500 1000 1500 2000
Distance from Dredge (ft)
CQ
U
5,000
4,000
3,000
2,000
1,000
Particulate Phase
#
>
/
' .
' I ^
.
5,000 -
0
-2000 -1500 -1000 -500 0 500 1000 1500 2000
Distance from Dredge (ft)
3,000 -
o 2,000 -
Q
Dissolved Phase
^ *
o-
-2000 -1500 -1000 -500 0 500 1000 1500 2000
Distance from Dredge (ft)
Notes
1,The points at a distance of 0 ft are in the moonpool. The plots with the expanded scales do not include the moonpool samples,
2, Baseline is shown at -1,500 ft,
3, The expanded plots have weighted curves as visual aids to show the approximate mean conditions.
Figure 45
PCB Concentrations for New Bedford Harbor
Hudson River PCBs Superfund Site Pilot Dredging Study Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment D - April 2004
-------
CQ
U
CQ
U
Moonpool Samples Included
Dissolved Fraction vs. Total Concentration
0.8-
0.6- >
H
0.4
0.2-
0 5,000 10,000 15,000 20,000 25,000 30,000
Total PCBs (ng/L)
0.6
0.4
Dissolved Fraction
i
k
-
ป
1 1 1
>
i i i
0-
-2,000 -1,500 -1,000 -500 0 500 1,000 1,500 2,000
Distance from Dredge (ft)
CQ
U
CQ
U
Expanded
0.8-
0.6-
0.4-
0.2-
Dissolved Fraction vs. Total Concentration
#.* I
I I I I I
1,000 2,000 3,000 4,000 5,000 6,000
Total PCBs (ng/L)
0.8
0.6
0.4
0.2
Dissolved Fraction
/
/
%/
y
^
* /*
0
-2,000 -1,500 -1,000 -500 0 500 1,000 1,500 2,000
Distance from Dredge (ft)
Notes
1.The points at a distance of 0 ft are in the moonpool. The plots with the expanded scales do not include the moonpool samples.
2. Baseline is shown at -1,500 ft.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Figure 46
Dissolved Fraction of PCB for
New Bedford Harbor Pilot Dredging Study
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment D - April 2004
-------
30-
25-
20-
15-
10-
Suspended Solids vs. Distance from the Dredge
oH 1 1 1
-2000 -1500 -1000 -500 0 500 1000 1500 2000
Distance from Dredge (ft)
0
O
U
200-
Particulate Concentration vs. Distance from the Dredge
150 -
100-
50-
OH 1 r-
-2000 -1500 -1000
-500
500 1000 1500 2000
Distance from Dredge (ft)
Notes
1 .The points at a distance of 0 ft are in the moonpool.
2. Baseline is shown at -1,500 ft.
3. The plots have weighted curves as visual aids to show the approximate mean conditions.
Figure 47
TSS and PCB Concentrations for New Bedford Harbor
Hudson River PCBs Superfund Site Pilot Dredging Study Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment D - April 2004
-------
Attachment E
Engineering Contingencies Considerations
Table of Contents
1.0 Introduction 1
1.1 Performance Standard Monitoring Locations 1
1.2 Resuspension Criteria 2
2.0 Monitoring and Contaminant Control Technologies Used At Other Sites 3
2.1 St. Lawrence River Remediation Project at the Alcoa, Inc. Massena East Smelter
Plant, New York (Reynolds Metals) 3
2.2 New Bedford Harbor (Pre-Design Field Test), New Bedford, Massachusetts 5
2.3 Grand Calumet River, Gary, Indiana 6
3.0 Engineering Contingencies for the Remediation 8
3.1 Monitoring Contingencies 8
3.2 Engineering Evaluations 8
3.3 Barriers 9
3.4 Operation and Equipment Modifications 11
3.5 Scheduling Changes 11
4.0 Implementation Strategies 12
5.0 References 13
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
Attachment E
Engineering Contingencies Considerations
1.0 Introduction
This attachment describes engineering contingencies that may be applied in the event that
the action levels or the Resuspension Standard threshold are exceeded. The levels of the
performance standard were developed using the statistical analysis of historical data,
surface water quality modeling and applicable federal standards. The resuspension
criteria will be used to govern the implementation of various engineering contingencies to
minimize the release of PCBs during the remediation, to achieve the remediation goals as
set forth in the Record of Decision (ROD) (USEPA, 2002), and to minimize the potential
impact of dredging on ambient water quality. In the event that the resuspension criteria
are exceeded, engineering contingencies will be implemented as necessary to minimize
the potential impact of dredging on ambient water quality. A series of contingencies,
ranging from increased monitoring frequency to cessation of dredging operation, have
been proposed. These engineering contingencies will be implemented based on near-field
and far-field water quality monitoring results.
The performance standard requires additional monitoring under certain conditions, the
frequency and parameters for this additional monitoring of which are defined as a part of
the performance standard. For other contingencies {i.e., contingencies not specifically
addressed in the performance standard), the specific technology cannot be selected, but
must be a judgment that is specific to the problem encountered. Contingencies must be
developed during the design stage for use in the event that water column concentrations
exceed the performance standard. The performance standard does specify that if certain
levels are exceeded, the cause of the exceedance will be examined and necessary changes
must be made to the existing operations.
This attachment provides a brief overview of the performance standard (including a
discussion of the monitoring locations needed to assess compliance with the standard), a
summary of engineering contingencies used during similar projects, and a discussion of
the engineering contingencies that may be applicable to the remediation.
Engineering contingencies for public and agricultural water intakes will be addressed in
the Community Health and Safety Plan.
1.1 Performance Standard Monitoring Locations
Two types of monitoring locations are discussed throughout this attachment. Definitions
are provided below:
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
Far-Field (Upper River and Lower River)
Far-field stations are fixed locations, typically located at dams and bridges. The primary
contaminants to be monitored at these stations are PCBs and suspended solids. The
results from monitoring at the far-field stations are the primary measure of PCB loss due
to dredging, based on the assumption that only PCBs escaping each river section have the
potential to cause significant downstream impacts.
Near-Field
Near-field monitoring locations are located within a short distance of the remedial
operations, typically within a mile or so downstream. Depending upon the proximity of
the various ongoing remedial operations to one another and the use of barriers, each
remedial operation may have near-field monitoring locations associated with it. These
near-field stations will be monitored continuously to determine the local impacts of
dredging activities. The primary measurements in the near-field will be suspended solids
concentrations and turbidity.
1.2 Resuspension Criteria
The resuspension criteria consist of three action levels and one standard providing limits
on the PCB and suspended solids concentrations. Each of the resuspension criteria has
associated monitoring requirements and engineering contingencies. The monitoring plan
is summarized in Tables 1-2, 1-3 and 1-4 of the main document, showing the parameters
required at each station and the frequency of sampling. Table 1-1 of the main document
lists the concentration or load limits for each action level. Monitoring and resuspension
criteria are fully described in the main body of the text and in Attachment F. An
engineering evaluation of conditions in the river leading to elevated concentrations is
recommended for Evaluation Level, but is mandatory for the Control Level and
Resuspension Standard threshold. Similarly, implementation of engineering
contingencies to reduce contaminant levels in the river is recommended at the Evaluation
Level, but is mandatory for other the two other levels.
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
2.0 Monitoring and Contaminant Control Technologies Used At
Other Sites
The monitoring and contaminant control technologies employed at three other PCB
remediation sites are described below. The three sites are:
St. Lawrence River Remediation Project at the Alcoa, Inc. Massena East Smelter
Plant, New York, (Bechtal Environmental, 2000; 2002);
New Bedford Harbor (Pre-Design Field Test), New Bedford, Massachusetts,
(USACE, August, 2001); and
Grand Calumet River, Gary, Indiana, (Earth Tech, Inc., 2002).
The technologies implemented at these three sites and reviewed in this attachment are
containment (St. Lawrence River), dredging system design [hydraulic bucket design]
(New Bedford Harbor), and monitoring (Grand Calumet River).
2.1 St. Lawrence River Remediation Project at the Alcoa, Inc.
Massena East Smelter Plant, New York (Reynolds Metals)
In order to control the export of PCB-contaminated sediment at the St. Lawrence River
Alcoa site, a containment system was installed as part of the remedial design. The
containment system at this site included:
A sheet pile wall that enclosed the entire remediation area;
Silt curtains that provided secondary containment for the more highly contaminated
Area C and also isolated uncontaminated portions of Area B from dredging areas;
and
Air gates (air curtain technology) that created an air-bubble curtain that acted as a
circulation barrier while allowing for barge and tugboat access to areas enclosed by
the silt curtain and pile wall.
Each of these components is discussed below.
Sheet Pile Wall
The wall consisted of interlocking steel sheeting embedded several feet or more into
sediments and supported by H-beams ("king piles") driven to greater depths. The
sheeting and king piles were held together by a welded and bolted framework of steel
Hudson River PCB s Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
braces and walers. The 3,800-foot finished wall consisted of about 200 king piles and
2,200 sheets. The maximum depth of water along the wall was about 32 feet.
The original design of the sheet pile wall specified that every fifth sheet would be driven
to the water's surface to balance any differences in hydrostatic pressure between the
inside of the wall and the outside environment. However, this was later changed and all
sheets were raised to a height of about 2 ft above the river surface, minimizing the
connection of turbid water inside the sheet pile wall with the river water outside the
enclosure.
After the installation, a video survey was conducted to verify that there were no openings
along the bottom of the wall or open seams in the sheeting. This survey identified a few
small holes that were patched using sandbags. In addition, some of the sheeting was
trimmed to reduce the height of all the sheets down to the 2 ft above water level (after
installation) to reduce the surface area exposed to wind forces. Environmental monitoring
data showed that the sheet pile wall functioned as designed and effectively contained the
turbidity and suspended sediments generated during the dredging activities within the
remediation area.
Silt Curtains
Silt curtains, consisting of 22-oz. PVC sheeting weighted on the bottom and suspended
by polystyrene floatation buoys, were installed around Area C and a portion of Area B.
The silt curtains were tied to H-beam anchor posts driven at a spacing of 100 feet, and
anchored on the shoreline using a driven post or tree. The ballast for the curtains
consisted of 3/8-in. galvanized anchor chain within a sealed pocket in the sheeting that
could adapt to the bottom contours, thereby providing a complete vertical barrier. The
curtain was suspended by cables attached to tensioners and anchor plates with reefing
lines connected to the lower ballast chain to adjust the vertical height. A total of 1,222
feet and 996 feet silt curtains were used in Area B and Area C, respectively. The silt
curtains effectively isolated the more contaminated Area C and prevented contamination
of the clean portion of Area B.
The original design called for the installation of the silt curtain H-beam piles after the
sheet pile wall was completed. However, due to the additional time required to install the
sheet pile wall, this plan was changed for the clean part of Area B, and the silt curtain H-
beam piles were driven while the sheet pile wall was being installed. A similar change for
the contaminated part of Area B was not approved by the United States Environmental
Protection Agency (USEPA).
Another change to the design of the silt curtain involved the addition of dual H-beams
rather than a single H-beam to anchor the curtain. The original design specified that one
H-beam would be placed at intervals along the inside of the curtain and timbers would be
attached to the top of the beam to prevent barge traffic from hitting the curtain from the
outside. The silt curtain manufacturer recommended placing dual H-beams at a spacing
of 90 feet and then anchoring the curtain between the beams.
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
Air Gates
Air gates (air curtain technology) were used to create vertical circulation barriers that
allowed boats to pass but restricted the movement of water between various parts of the
remediation area. The air curtains consisted of 2-in. outside diameter (OD) steel pipe
fitted with diffuser orifices on a helical, 9-inch spacing. The pipes had leg supports that
raised them about a foot off the bottom. Geomembrane was laid beneath the pipes to
minimize the disturbance of nearby sediment. Divers were used to place the liner, pipe
and anchors, connect the supply lines and verify proper operation once the equipment
was in place. A compressor station supplied air to the gates at a flow rate of about 1,000
cubic feet per minute (cfm) with flow pressures of 90 to 100 psig. The gates allowed for
barge transit and limited the migration of turbid water across the barrier. A major
objective of the gates was to contain the turbidity generated during the removal of Area C
sediment. The gates accomplished this objective and otherwise functioned as designated
for the duration of the project.
2.2 New Bedford Harbor (Pre-Design Field Test), New Bedford,
Massachusetts
A pre-design field test was conducted at the New Bedford Harbor site to assess the
effectiveness of hydraulic dredging as an engineering contingency to minimize the
release of PCB contaminated material to the water column and to limit the transport of
sediment away from the dredging area. The water quality monitoring data obtained
during dredging activities indicated that the actual dredging process using a hydraulic
excavator appeared to have a limited impact on the water column. The factors that
minimized the release of material to the water column included the design of the bucket
(tight closing with limited leakage), the configuration of the dredge (with a "moon-pool"
work area enclosed behind a 36-inch silt curtain), and the controlled manner in which the
operation was executed.
Factors that limited the transport of contaminated material away from the dredging area
included the shallowness of the area (maximum depth of the dredged area was less than
10 feet (3 m)) at high tide and the limited currents (maximum currents generally
measured less than 0.5 feet/sec.).
Activities performed in support of the dredging (e.g., the operation of support vessels
such as tug boats) appeared to have a much greater impact on water quality than the
dredging operations due to shallowness of the water, which measured about 4 to 5 feet in
depth.
Normal fluctuations that occur in Upper Harbor due to changing environmental
conditions appeared to be similar to, or greater in scale than, the overall impacts related
to the actual dredging process.
Hudson River PCB s Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
2.3 Grand Calumet River, Gary, Indiana
Dredging activities are scheduled for completion in December 2003. The following was
extracted from the Water Quality Certification Work Plan, dated July 2002.
Three water quality monitoring locations (Sites A, B, and C) are defined as the primary
monitoring sites. A fourth monitoring location (Site D) is defined as the verification site.
Site A is intended to monitor water quality upstream of dredging (located mid-
channel of the Grand Calumet River at Transect 4 and will be re-located to Transect 2
as dredging progresses),
Site B is located mid-channel, approximately 200 yards downstream of the open
water dredge in Transect 12 to 36, and will be re-located as dredging progresses
through cell D (or from Transect 12 to 36),
The third station, Site C, is the downstream sample site and is located mid-channel
downstream of Transect 36 (downstream of the limit of dredging), and
A fourth sample location, Site D, also known as the verification sample site, will be
situated 200 yards upstream of the open water dredge in transects 12 to 36, will be
used to verify water quality exceedances, and used to determine if the exceedance is a
result of the dredging operation or a different point source. This station was proposed
in lieu of performing background sampling prior to initiating dredging. All water
samples will be equal volume composites created from a total of three samples per
location. The three samples per location will be taken from the water surface, at 50
percent of the water depth and at 80 percent of the water depth.
Three levels of monitoring will be utilized, including Level 3 Monitoring (i.e., collection
of composite water samples once per month from automatic samplers at Sites A and C
and manually at Sites B and D for analysis of PCBs and other specified parameters). If
results indicate no exceedances at Sites A, B and C, or if monitoring indicates
exceedances at all three sites (A, B, and C), then it will be concluded that dredging is not
the source and normal sampling will be conducted (once per month). If, however, results
indicate exceedances at Sites B and C but not site A, then the water sample collected at
Site D will be analyzed. If the sample from Site D indicates the parameters exceeded at
Sites B and C are also exceeded at Site D, it will be assumed that the downstream
exceedances at these sites are not a result of dredging and the normal frequency sample
will be conducted. However, if no exceedances are found at Site D, it will be concluded
that dredging is the source and enhanced monitoring consisting of additional sample
collection at Sites A, B and C will be implemented at a rate of three times per week.
When results indicate that the parameters of concern are less than the criterion for two
months of consecutive samples, enhanced monitoring will be discontinued and the
normal monitoring frequency will be resumed.
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
In addition to the increased sampling frequencies as a result of exceedances determined
to be due to dredging, the Indiana Department of Environmental Management and the US
Army Corps of Engineers will also implement a response action. If it is thought that an
immediate threat to human health or aquatic life exists, the required response action will
be issued within 72 hours, and this action will be implemented as quickly as possible,
with a maximum time limit of one week to complete the implementation. If this schedule
is not met, enhanced monitoring will be automatically implemented as described above,
based on the parameters exceeded and the level of monitoring utilized when the
exceedances occurred.
Possible response actions may consist of the following engineering contingencies:
Decrease dredging operation,
Install additional turbidity barriers or control mechanisms,
Temporary cessation of dredging activities, and
Conduct additional monitoring.
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
3.0 Engineering Contingencies for the Remediation
The required engineering contingencies for the Resuspension Performance Standard are
described below. These include an increased monitoring frequency, engineering studies,
containment technologies, operational modifications, equipment modifications and
scheduling changes. With the exception of the monitoring frequency, specific
implementations of the engineering contingencies must be planned during design.
The applicability of many of the containment technologies was evaluated in Appendix
E.5 of the FS (USEPA, 2000). The advantage and limitation of each type of turbidity
barrier were discussed. This information will be useful when choosing the appropriate
containment system for a specific area to address the engineering contingency during the
remediation.
3.1 Monitoring Contingencies
Monitoring frequency of the far-field stations will be increased at higher levels of
exceedance to gain more information from which to evaluate conditions. The degree of
increased frequency is detailed in Table 1-2, 1-3 and 1-4 of the main document for non-
routine monitoring. The sampling method also changes for some stationsfrom grab
samples to composites of hourly samplesto better capture the average water column
concentration at the nearest representative far-field stations and to limit the number of
analytical samples required.
3.2 Engineering Evaluations
In instances where water quality measurements exceed a resuspension criteria based on
PCB concentration or load, an evaluation of the remedial operations should be conducted
to determine the possible source and mechanism causing the exceedance, including:
Examination of the barrier, if it is in use, for leaks and stability,
Examination of the sediment transport pipeline if a hydraulic dredge is used,
Examination of the turbidity associated with sediment transport barges and
other support vehicles, and
Sampling of PCB concentrations in the near-field.
The above-listed engineering studies will be mandatory in the event the Control Level
and Resuspension Standard threshold are exceeded.
Hudson River PCB s Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
3.3 Barriers
Several types of barrier systems are described below:
Fixed Structural Barriers,
Non-Structural (Portable) Barriers,
Other Portable Barrier Systems, and
Control Zone Technology.
Fixed Structural Barriers
Fixed structural barriers, such as sheet piling, are particularly suitable for areas where
there is a potential for high levels of resuspension. Sheet piling consists of a series of
interlocking steel sections. The piles and panels are all driven into the riverbed to
approximately the same depth. It is anticipated that turbidity barriers comprised of sheet
piling will not be used areas where relatively shallow rock is present.
While fixed structural barriers provide considerable structural capacity, these systems are
relatively expensive and usually require significant planning, equipment and manpower
resources to install.
Non-Structural Barriers (Portable Barriers)
Non-structural barriers, such as silt curtains and silt screens (sediment curtains), can be
considered for use to prevent the transport of sediments resuspended as a result of
dredging activities Silt curtains are constructed of impervious materials that block or
deflect the passage of water and sediments. Silt screens are similar to silt curtains;
however, these barriers allow water to flow through while impeding the passage of a
fraction of the suspended load. Typically, a silt curtain and silt screen are suspended by a
flotation unit at the water's surface and held in a vertical position by a ballast chain
within the lower hem of the skirt. Anchors attached to the barrier also serve to hold it in
place.
The advantage of using non-structural barriers is that they can easily be deployed and re-
located to new work areas after dredging at a specific location has been completed. Silt
curtains are not considered appropriate in situations where the river current is greater than
approximately 1.5 feet per second and/or where the depth of the river exceeds 21 feet.
However, it should be noted that if the silt curtain is set up in a configuration that is
closely parallel to river flow, the curtain could function effectively in currents
approaching 3 feet per second.
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
Other Portable Barrier Systems
Other commercial products such, as the Portadam and Aqua-Barrier systems, are
also available for construction site containment, diversion of water flow, erosion control,
and flood control. These systems are low-cost alternatives to building earthen dams or
using sheet piles, and are relatively easy to set up. These systems are generally applicable
to water depths of less than 10 feet.
The Portadam system utilizes a free-standing steel support structure in conjunction
with an impervious fabric membrane. The support members transfer fluid loading to an
approximately vertical downward load, allowing for installation on a solid impenetrable
foundation. This structure stands independently on the existing bed, eliminating the need
for pile-driving equipment, cross bracing, or anchorage. The membrane is placed on the
outer section of the support structure, and is rolled out all the way down to the level of
the bed. Hydraulic loading on the membrane assists in the sealing and stability of the
entire structure. Once installed, the work area enclosed by the structure can be de-
watered.
The Aqua-Barrier and GeoCHEM Water Structures systems utilize water-filled,
vinyl, polyester-reinforced tubes to provide mass for stability and they can be coupled
together to form a barrier of any length. Punctures in the material can be easily patched
with repair kits. They are lightweight, easy to transport, and re-usable. While these
systems are not as sturdy as the Portadam system, they can be used in cold weather
conditions and are reasonably resistant to sunlight exposure.
Air gates are used to facilitate the passage of dredging-related traffic to and from an
enclosed (i.e., sheet piled or silt curtained) area. The technology employs a continuous
release of bubbles to reduce the flow of water to and from an enclosure. The air is
supplied from a blower or compressed air source. The effort and cost associated with the
deployment and operation of air gates are low and the performance of air gates appear to
be superior when compared to silt curtain gates.
Control Zone Technology
A Control zone is a secure dredging area that is maintained and sealed off to prevent the
release of contaminants generated inside the zone. Application of control zone
technology (CZT) allows the excavation of contaminated sediments without the release
of particulate and soluble contaminants into the surrounding water environment. It also
establishes an area that can be easily monitored to confirm that remediation goals are
met. This type of technology is more stringent than other barrier technology, since it
requires additional water treatment. CZT has only been tested on a pilot scale and the cost
is likely to be prohibitive. This type of technology could be considered for limited use in
the most highly contaminated areas.
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
3.4 Operation and Equipment Modifications
Depending on the level of resuspension observed, operational control and equipment
modification which include the following should be considered:
Limiting boat speeds to reduce prop wash,
Restricting the size of boats that can be used in certain areas,
Loading barges to less than their capacity where it is necessary to reduce draft,
Selecting an alternate dredge with a lower resuspension rate,
Selecting alternate equipment or method for placing backfill or capping
material, and
Use of smaller, shallow draft boats for the transport of crewmembers and the
inspection of dredges.
3.5 Scheduling Changes
The baseline PCB water column concentrations are high during the months of May and
June relative to the remainder of the dredging season. As documented in the baseline
water column level study (Attachment A), the 95 percent upper confidence limits (95%
UCL) on the mean of PCB concentrations at the TI Dam and Schuylerville ranged from
110 ng/L to 200 ng/L in May and June. Remedial activities in high-concentration areas
during high flow conditions may result in increased water column PCB concentrations
that fall above resuspension criteria, resulting in the implementation of engineering
contingencies, for example a containment system capable of containing enough of the
resuspended material to maintain acceptable water column concentrations. Areas with
higher sediment concentration may need to be scheduled for remediation in later months
of each year (i.e., under low flow conditions, when the baseline level of PCB
concentration is relatively low) if the engineering contingencies chosen are not effective.
Baseline water column concentrations should also be considered when scheduling
remediation in areas nearest water treatment plants in order to maintain a margin of safety
for the public water supply.
Hudson River PCB s Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
4.0 Implementation Strategies
Flowcharts depicting the implementation of the Resuspension Performance Standard are
provided in Figures 3-1, 3-2 and 3-3 of the main document for the near-field suspended
solids, far-field total PCBs and far-field suspended solids, respectively. These flowcharts
present the interaction between the three aspects of the Resuspension Performance
Standard: resuspension criteria, monitoring requirements and engineering contingency
requirements.
Hudson River PCBs Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
5.0 References
Bechtel Environmental, Inc./Metcalf & Eddy, Inc. 2000. Final Dredging Program Design
Report for the River Remediation Project at the Reynolds Metals Company, St. Lawrence
Reduction Plant, Massena, New York, Revision 3. Prepared for Reynolds Metals
Company. May 2000.
Bechtel Environmental, Inc./Bechtel Associates Professional Corporation, 2002. Draft
Interim Completion Report for the St. Lawrence River Remediation Project at the Alcoa,
Inc., Massena East Smelter Plant, New York, Volume 1, Revision 0. Prepared for Alcoa.
March 2002
Chesner, W.H., 2002. Control zone technology for dredging. Presentation and technology
demonstration on July 9. 2002. Southold, NY.
Earth Tech, Inc. 2002. Grand Calumet River Section 401 Water Quality Certification
Work Plan. Prepared for US. Steel. July 2002.
U. S. Army Corps of Engineers (USACE). 2001. Final Pre-Design Field Test Dredge
Technology Evaluation Report, New Bedford Harbor Superfund Site, New Bedford,
Massachusetts. Prepared by Foster Wheeler Environmental Corporation, Boston,
Massachusetts. August 2001.
U.S. Environmental Protection Agency (USEPA). 2000. Phase 3 Report: Feasibility Study,
Hudson River PCBs Reassessment RI/FS. Prepared for EPA Region 2 and the US Army
Corps of Engineers (USACE), Kansas City District by TAMS Consultants, Inc.
December 2000.
USEPA. 2002. Responsiveness Summary: Hudson River PCBs and Record of Decision,
Prepared for USEPA Region 2 and United States Army Corps of Engineers by TAMS
Consultants. January 2002.
Hudson River PCBs Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment E - April 2004
-------
Attachment F
Measurement Technologies
Table of Contents
1.0 Introduction 1
2.0 Measurement Techniques 2
2.1 Correlations Between Turbidity and Suspended Solids 4
2.2 In-situ Turbidity Measurement 6
2.3 In-situ Total Suspended Sediment Measurement 8
2.4 Semipermeable Membrane Devices (SPMDs) 10
2.5 Trace Organic Platform Sampler (TOPS) 14
2.6 ISCO Portable Water And Wastewater Sampler 18
3.0 References 20
LIST OF FIGURES
Figure 1 Laser diffraction principles - a cut away view of the basic LISST-100
instrument.
Figure 2 SPMD Apparatus
APPENDIX
Attachment F-l Literature Review
Attachment F-2 PCB Analytical Methods Detection (Reporting) Limits in Water
Attachment F-3 Memo Regarding PCB Analyses; Whole Water Extracts vs. Separated Particle
and Filtrate Extracts
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS -Earth Tech
Volume 2: Attachment F - April 2004
-------
Attachment F
Measurement Technologies
1.0 Introduction
This section provides detailed descriptions of specific measurement techniques for the
general continuous monitoring prescribed in the performance standard. These include:
In-situ Turbidity Measurement,
In-situ Total Suspended Sediment Measurement,
Semipermeable Membrane Devices (SPMDs),
Trace Organic Platform Sampler (TOPS), and
ISCOฎ Portable Water and Wastewater Sampler.
The above-listed instruments are presented as examples of technology that may be used
during construction to satisfy the requirements of the standard, but the selection of
appropriate technology will be a part of the design process.
Several other issues related to the monitoring are presented in this attachment.
Correlations between turbidity and suspended solids measurements are discussed, and
development of a correlation between these parameters will be required in order to obtain
a real-time indication of dredge-related impacts on the water column. Attachment F-l
presents the results of a literature search on this topic. Attachment F-2 provides a
synopsis of PCB analytical methods and associated detection limits. The detection limits
for PCB congener analysis will be low in order to obtain detections at each station and to
allow for identification in congener patterns.
Hudson River PCB s Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
2.0 Measurement Techniques
All types of dredging (navigational and environmental remedial action) create sediment
plumes in the water column. Of particular interest for the Hudson River remedial action
are plumes associated with the following activities:
The mechanical and/or hydraulic dredging (sediment removal) operation,
Material handling of dredged materials,
Boat and barge movements, and
Open-water placement of backfill materials.
The regulatory agencies and the public are concerned about potential adverse effects
caused by these plumes on humans and biological resources, either through impact to
water quality or increased siltation. To gain a better understanding of the temporal and
spatial dynamics of sediment plumes, and in order to implement the performance
standard for resuspension, it is necessary to monitor the plumes created to determine their
composition, extent, and duration. Numerous techniques have been used to monitor
sediment plumes, ranging from collection of water samples using simple water samplers
to highly complex systems involving state-of-the-art instrumentation. Given the variety
of techniques available to monitor dredging-related plumes, it is necessary to understand
the advantages and limitations of the various techniques in order to determine which
techniques provide the most cost-effective approach for each specific monitoring
requirement.
The resuspension performance standard (as defined in Section 1) includes specifications
for both PCBs and suspended sediment monitoring. The PCB standard requires
measurement of the total PCB concentration, specifically, the measurement of the
dissolved phase and suspended phase concentrations for all PCB congeners
(monochlorbiphenyls through decachlorobiphenyl). Suspended solids (sediment)
standards have been defined in order to serve as a surrogate for the amount of PCBs in
the water column in order to provide a real-time indication of PCB concentration.
Another parameter that could potentially serve as a surrogate for PCBs is turbidity.
Although turbidity has been historically used during the monitoring of dredging activities
to estimate suspended sediment using empirical correlations, it is well recognized that
achieved correlations are site-specific and subject to significant error.
The objectives of measuring the water quality parameters discussed above (PCBs,
suspended solids, and turbidity) are twofold:
First, determine the water quality associated with the plume; and
Second, track the plume both in space and time.
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
Knowledge of the spatial extent of a given plume is necessary to determine areas of
potential plume impact. Similarly, knowledge of the time history of a plume provides
information on how long a plume is present in a particular area and the time required for
the plume to dissipate. It is clear that both near-field and far-field monitoring are
necessary.
It may be important to measure various physical parameters not directly associated with
water quality such as currents, waves, and water elevations. Currents carry plumes from
the area in which they were generated into adjacent waters. Therefore, data on the current
structure can be used to estimate the movement and spatial extent of the plume. Waves
increase turbulence in the water column, which can potentially introduce additional
sediment into suspension and prevent material in suspension from settling out.
Measurement techniques for monitoring of plumes involve one of the following:
The collection of water samples from the water column for analysis either in the
field or the laboratory (ex-situ methods), or
The placement of instruments in the water column to directly measure water
quality parameters or other physical parameters (in-situ methods).
Off-site laboratory analysis is time-consuming, expensive, and cannot provide data in the
short term (i.e., within a few hours or less of sample collection). At present, there are no
in-situ methods available for directly measuring PCB congeners in the water column,
therefore, sample collection and laboratory analysis are required.
Concentrations of PCBs in the water column are often present at parts-per-billion (|~ig/L)
or parts-per-trillion (ng/L) levels. Conventional sampling, extraction, and analysis
methods like liquid-liquid extraction or solid-phase extraction can require the sampling
and processing very large volumes of water (e.g., 50 liters) for an analysis of adequate
sensitivity to detect low concentrations. (See Attachment F-2 for a synopsis of PCB
analytical methods and associated detection limits.)
The limitations inherent in methods for the direct measurement of contaminant water
concentrations have often prompted the use of biomonitoring to assess the exposure of
organisms in the water column to trace or ultra-trace levels of hydrophobic chemicals like
PCBs. Because certain organisms often bioconcentrate these trace or ultra-trace levels of
PCBs to relatively higher concentrations (parts per million) in their lipids, determination
of the bioavailable portion of environmental pollutants like PCBs is critical to assessing
the potential for detrimental biological impacts.
This organism-based approach also has inherent problems, including biotransformation
and depuration of contaminants, and inapplicability in many exposure situations due to
the effects of stress on the biomonitoring organisms that often lead to a lack of
proportionality between the biomonitoring organism tissue concentrations and ambient
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
exposure concentrations (Petty et al., 2000). Therefore, innovative approaches for
sampling and analyzing trace and ultra-trace levels of water-borne PCBs are needed.
The major mechanisms that result in relatively high concentrations of PCBs in organisms
are passive processes and include the following:
biomembrane diffusion, and
partitioning of the chemical between an organism's lipid tissue and its
environment.
Employing a mimetic chemistry approach (i.e., use of processes in simple or uniform
media to mimic complex biological systems), scientists at the United States Geological
Survey's (USGS) Columbia Environmental Research Center (CERC) have developed a
passive, integrative sampler that simulates hydrophobic chemical bioconcentration. The
uncertainty of estimating ambient exposure concentrations from tissue concentrations in
biomonitoring organisms is thereby avoided. This sampler, the semipermeable membrane
device (SPMD), measures the concentration of dissolved phase PCBs in the water
column. A second type of integrating sampler, the Trace Organic Platform Sampler
(TOPS), has been developed by New York State Department of Environmental
Conservation (NYSDEC). The TOPS concentrates hydrophobic organic compounds from
surface waters and is designed to collect suspended and dissolved-phase organics.
2.1 Correlations Between Turbidity and Suspended Solids
This section describes techniques traditionally used to measure turbidity and suspended
solids in waters, how the two parameters relate to each other and to various
environmental impacts, and why one cannot be routinely substituted for the other. An
additional literature review is presented in Attachment F-l.
The term total suspended solids (TSS), sometimes referred to simply as suspended solids
(SS), encompasses both inorganic solids such as clay, silt, and sand, and organic solids
such as algae and detritus. It is a measure of the dry weight of suspended solids per unit
volume of water, and is reported in milligrams of solids per liter (mg/L). The suspended
solids concentration is determined by filtering a known volume of water through a filter
of specific pore size (45 |j,m), and then drying and weighing the material retained on the
filter. USEPA Method 160.2 is often used for this 'TSS' measurement.
Although popularly called suspended solids (the terminology used in this report), this
method is more accurately termed non-filterable solids (or residue), because the size of
separation (about 0.45 |j,m) is not the same as the boundary between suspended and
dissolved solids. The suspended solids/dissolved solids boundary varies among
molecules, but is generally around 0.1 |j,m. Another drawback of this method is that
laboratories often perform this analysis using a 100 mL aliquot of the total sample
provided, typically a 250-mL sample bottle. There is the potential that some of the solids
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
adhered or adsorbed to the surfaces of the container, yielding a reported result with a low
bias relative to the 'true' value. The method used by USGS to measure suspended
sediment, ASTM Method D3977-97, may be preferable.
Turbidity is an optical property of water that causes light to be scattered and absorbed
rather than transmitted in straight lines through the sample. It is caused by water
molecules, dissolved substances, and organic and inorganic suspended matter.
Turbidity measurements can be used as an operational aid in monitoring dredging and
backfill placement operations as an adjunct to the more costly and time-consuming
suspended solids measurements in a laboratory. The primary reason for wanting to use
turbidity measurements instead of suspended solids is that turbidity measurements are
immediate; nephelometric turbidity readings can be performed in a matter of minutes.
The collection and analysis of a suspended solids sample requires the following actions:
Transport to the laboratory
Analysis of the sample
- filtering
- drying
- weighing
Calculation of the suspended solids value
The transport and analysis process can take from 3 to 24 hours to complete. In the time it
takes to get results of the laboratory analysis, the suspended solids of the discharge or
water body of interest will have changed. Therefore, laboratory measurements for
suspended solids cannot be easily used to detect and correct short-term problems or
performance standard violations.
It is for this reason that turbidity measurements have historically been substituted for
suspended solids. Turbidity is easy to measure quickly, but the following problems are
associated with using turbidity as a surrogate for suspended solids:
There is no universal relationship between it and suspended solids
There is no universal relationship among turbidity measurements made on
different water-sediment suspensions
There is no universal relationship among turbidity measurements made on the
same suspension with different instruments.
Turbidity does not correlate well with many categories of environmental impact.
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
However, turbidity can be used to indicate suspended solids concentration on a site-
specific basis, if certain specific techniques are used.
Two factors that prevent the development of a simple, universal relationship between
suspended solids and turbidity are that the two parameters measure different things and
their values are functions of different variables, . The suspended solids parameter
depends on the total weight of particles in suspension, and is a direct function of the
number, size, and specific gravity of the particles. In contrast, while turbidity is a direct
function of the number, surface area, and refractive index of the particles, it is also an
inverse function of their size (for constant suspended solids) (Thackston and Palermo,
2000).
The problems associated with the correlation of turbidity and suspended solids are based
on two factors associated with calibration:
The calibration changes with changes in grain size of the sediment.
The calibration changes with sediment color.
A landmark paper co-authored by the inventor of one of the most widely used turbidity
meters noted the following:
The calibration changed by a factor of 10 based on color alone
The change in calibration that is linear with sediment grain size (Sutherland et al.,
2000).
For example, the calibration would change by a factor of 20 between white 5 micron
sediment particles and gray 10 micron sediment particles. Such changes in sediment
properties are not uncommon in nature. Since sediment color and grain size are not
characteristics that are generally known during the course of a monitoring period, spot
calibrations from samples are likely to contain unknown errors as sediment properties
change in space and time (Agrawal and Pottsmith, 2000).These errors can reach several
hundred percent and greater. Laser sensors described below in Section 2.3: In-situ Total
Suspended Sediment Measurement overcome both these errors, making it easier to
monitor suspended sediments.
2.2 In-situ Turbidity Measurement
Turbidity is the apparent "cloudiness" of water produced as light is scattered by
particulate matter or dissolved material in the water. Presently established methods for
measuring suspended sediments via optical turbidity are rooted in the research performed
by Whipple and Jackson around the year 1900, which lead to a candle-based turbidity
standard called the Jackson Turbidity Unit (JTU). Devices commonly used to measure
turbidity include the Jackson candle turbidimeter, absorptimeters, transmissometers, and
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
nephelometers (McCarthy, Pyle, and Griffin 1974). All but the nephelometer measure the
effects of both the absorption and the scattering of light. The nephelometer measures
scattered light only, and is the most commonly used device in colloidal chemistry,
drinking water treatment, and water quality management. Turbidity measured by such an
instrument is reported in nephelometric turbidity units (NTU).
A transmissometer projects a narrow beam of light through a volume of water and
measures the intensity of the beam as it exits the volume of water. If particles are present
in the water, they will attenuate the beam of light such that the light exiting the volume of
water is less than the light entering the volume of water. The amount of attenuation can
be measured, and with the appropriate calibration, these measurements can be used to
estimate the suspended-particle concentration using an empirically-derived calibration
curve. At low particle concentrations, transmissometers are very sensitive to small
changes in particle concentration and/or size; however, at high-particle concentrations,
transmissometers become saturated and lose their sensitivity to variations in
concentration. Therefore, while transmissometers are very useful at measuring low-
particle concentrations, they are inadequate for measurements at suspended solids levels
above approximately 150 mg/L (Zaneveld, Spinrad, andBartz 1979).
Nephelometers project a beam of light into a volume of water and measure the amount of
light scattered out of the beam. The amount of light scattered is almost entirely dependent
on the amount and size of particulate matter present in the volume of water. Ideally, a
nephelometer would measure the amount of light scattered at all angles. Such a
nephelometer is impractical, however, and standard nephelometers measure the scattered
light at only one angle. Nephelometers use a device such as a photomultiplier tube or
silicon photodiode to measure light that has been scattered at a specific angle, usually 90
degrees, from the main light path. The light source is usually a tungsten filament lamp or
a light-emitting diode, and the light path is designed to minimize the amount of stray light
that reaches the detector. Thus, a zero signal means that no light scattered at 90 degrees
from the main light path and implies no turbidity.
Nephelometers used for in-situ measurements are, in general, referred to as optical
backscatter sensors (OBSs). OBSs measure the amount of infrared light backscattered
from a volume of water. While suspended sediment will reflect infrared light, organic
matter will not (Tubman 1995). This characteristic of OBSs makes them well suited for
the monitoring of sediment plumes because it does not bias the data by including organic
matter. Because an OBS measures backscatter, its design is simple and compact relative
to that of a transmissometer. More importantly, an OBS is capable of measuring
significantly higher particle concentrations than a transmissometer, though it lacks the
accuracy of the transmissometer at low-particle concentrations. Like the transmissometer,
particle concentrations in the water can be estimated from OBS measurements using
empirically determined calibration curves.
The ability of a particle to scatter light depends on the size, shape, and relative refractive
index of the particle, as well as on the wavelength of the light (Lillycrop, Howell, and
White 1996). The reading taken by the instrument depends on many design parameters,
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
including the light source, detector, electrical circuit, sample container, and optical
arrangement. As a result, two samples with equal suspended solids concentrations but
different size distributions of particles will produce very different turbidity readings on
the same nephelometer, and two different nephelometers may produce different turbidity
readings on the same sample, even if they were calibrated on the same standard (Vanous
1978; Hach 1972). Although the original Jackson candle turbidimeter was standardized
with a specific fine silica suspension in which one JTU equaled 1.0 mg/L of suspended
solids, modern turbidimeters are no longer standardized against the Jackson candle, and
the term JTU is no longer used. The Jackson candle turbidimeter is no longer an accepted
standard method (StandardMethods 20th edition, APHA et al.).
Modern turbidimeters are standardized against a formazin suspension with a value of 40
NTU. The standards should be prepared according to Standard Methods 20th edition
(APHA et al.) The 400-NTU stock suspension should be prepared monthly, and the 40-
NTU standard turbidity suspension should be prepared daily. Experience shows that this
turbidity can be repeatedly prepared within an accuracy of +1 percent (Hach 1972). The
formazin turbidity standard is assigned a value of 40 NTU and can be diluted to any
desired value.
One of the main benefits of measuring turbidity is that turbidity sensors are relatively
simple, inexpensive, and robust. The objective of most turbidity measurements is to
identify the presence of suspended solids and quantify the suspended solids based on a
correlation between turbidity and suspended solids. Historically, the standard practice has
been to use turbidity measurements to estimate suspended solids. Such estimates are
accurate only under the following conditions:
All measurements being compared are taken using the same turbidity sensor.
The turbidity sensor is calibrated with a reference standard and suspended
material from the area where the measurements are being taken.
Particle size and composition of the suspended material do not change
significantly during the measurement period.
Turbidity can also be measured in the field by collecting water samples and using
portable instruments to analyze the samples. While these instruments are typically less
expensive than in-situ sensors, the measurements take longer and may not represent true
in-situ conditions, since particles may settle out of suspension prior to analysis. The cost
of these instruments is approximately $1,500 to $2,000.
2.3 In-situ Total Suspended Sediment Measurement
Historically, suspended solids have been measured by collecting water samples and
analyzing the samples in an off-site laboratory. Water samples can be collected using a
bottle sampler or a submerged pump. Independent of the collection method, care must be
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
taken to ensure that suspended particulate matter does not settle out of suspension or
flocculate during collection or prior to analysis. Off-site laboratory analysis is time-
consuming and cannot provide data in the short term. However, this approach is
considered to be the most accurate and reliable method for measuring suspended solids.
The alternative has been to estimate suspended solids based on other measurements such
as turbidity or acoustic backscatter, both of which have limitations as discussed above.
New technology, Laser In-Situ Scattering and Transmissometry (LISST) has been
recently developed that can measure suspended solids in-situ more accurately than can be
achieved using correlations with turbidity. The instrument, the LISST-25, measures the
scattering of a laser beam by particles in a volume of water. It should be noted, that laser
diffraction measurements have been used to measure and characterize suspended
sediments and floc-sizes in situ since 1985 (see, e.g. Bale and Morris, 1987; McCabe et
al.,1993; Agrawal and Pottsmith, 1994; Gentien et al.,1995; Bale, 1996; van der Lee,
1998).
The LISST-25 is a small, self-contained unit that is suitable for field deployment and has
real-time data return capabilities (Sequoia Scientific, Inc.). The instrument is capable of
measuring particle total volume, particle total area, and the Sauter mean diameter within
a particle range of 1.2 to 250 mm. These parameters are defined as follows:
Particle total volume is the volume of material per volume of water.
Particle total area is the projected cross-sectional area of the particles per volume
of water.
Sauter mean diameter is the ratio of the particle total volume to the particle total
area.
If the density of the suspended particulate matter is assumed, it is possible to calculate the
suspended solids concentration by multiplying the particle total volume by the assumed
density.
Other models in the product line of the LISST instrument are capable of measuring the
particle size distribution, in addition to the above-listed capabilities. The LISST-100 is
the first in-situ laser that simultaneously measures the beam attenuation coefficient, the
volumetric concentration (ml/L), and in-situ particle size spectra. It is designed to be
submerged to a maximum depth of 300 m and is equipped with a built-in data logger.
The LISST-100 measures the particle size distribution in 32 logarithmically-spaced size
classes in the range of 1.25 to 250 //m (a LISST-100 type B). Other versions of the
instrument can measure size ranges of 2.5 to 500 //m and 7.5 to 1,500 //m, spanning a
200:1 dynamic range in all cases. A detailed description of the design and the operational
principles of the LISST-100 can be found in Agrawal and Pottsmith (1994), Agrawal et
al.(1996) or Traykovski et al. (1999). However, the basic principles are explained very
briefly below.
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
The LISST-100 measures the angular distribution of forward-scattered light energy over
a path length of 5 cm, using a collimated laser beam with a wavelength of 670 nm. The
energy of the scattered light is detected on 32 logarithmically-spaced ring detectors and
stored in a built-in data logger. When data collection is complete, these raw data are off-
loaded and mathematically inverted. The inversion yields the area distribution of the
suspended particles in 32 size classes. By multiplying the area distribution by the
diameter of each size class, the particle volume distribution is obtained. The absolute
volume concentration (ml/L), is found by summing the volume distributions in all 32 size
classes and dividing by an instrument-dependent calibration constant. The part of the
light not scattered is detected by a photo-diode in the center of the ring detector, thus
yielding the optical transmission, T, of the water. From the optical transmission, the beam
attenuation coefficient at 670 nm, c(670), can be calculated using Eq. (1)
c(670)(m"1) = -1/0.05 m x In (T) (1)
The processed data output from the LISST-100 thus consists of a particle volume
distribution in 32 size classes, an absolute volume concentration (ml/L) and a beam
attenuation coefficient at 670 nm. The LISST-100 also records the temperature and
pressure. From the particle volume distribution, statistical parameters such as the mean
and standard deviation can then be calculated. All software necessary for obtaining and
analyzing raw data is supplied by the manufacturer of the LISST-100, Sequoia Scientific
Inc., USA. (See Figure 1.)
Although the LIS ST instruments have not been used extensively in field studies of
plumes when compared with turbidimeters, some documented information on their
performance does exist (Melis, T., 2002, Mikkelsen, O., 2000). A recent study comparing
the LIS ST to traditional methods of measuring suspended-sediment concentrations
indicated that the LIS ST provided accurate measurements of the total volume
concentration of suspended sediments (Traykovski et al, 1999). Once the accuracy and
limitations of these systems are thoroughly documented by site-specific testing at the
Hudson River PCBs Superfund Site, which could occur during the two-to three-year
baseline/pre-dredge monitoring period if this device is selected for use, this instrument
could prove very useful for the in-situ monitoring of sediment plumes in the Hudson
River during Phase 1 and Phase 2 dredging activities. The cost of the monitoring
equipment is approximately $15,000 to $30,000 for the LISST-25 and LISST-100.
Because of the cost, some limited use of these instruments is warranted such at the far-
field stations and for daily readings at the near-field stations.
2.4 Semipermeable Membrane Devices (SPMDs)
An SPMD is a passive sampling device that consists of a thin film of the neutral lipid,
triolein, sealed inside a layflat, thin-walled tube of nonporous (i.e, no fixed pores; only
transient thermally mediated cavities) low-density polyethylene (LDPE). The diameters
of the transient cavities range up to approximately 10 A, effectively precluding the
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
sampling of any contaminant molecules associated with dissolved organic matter or
particulates. This cavity size limitation has an important consequence: in general, only
dissolved chemicals with molecular masses less than about 600 are sampled by SPMDs;
this molecular mass limitation is very similar to that imposed by the pores of
biomembranes.
At saturation, the capacity of an SPMD for a hydrophobic compound like PCBs is
generally related to the compound's octanol-water partition coefficient (Kow)- The higher
the Kow is for a compound, the greater the capacity for that compound the SPMD has.
Due to the very high concentration factors that are possible using an SPMD, even ultra-
trace levels of the hydrophobic contaminants in the water column are readily analyzed.
Standard SPMDs are designed to sequester and concentrate hydrophobic compounds like
PCBs and PAHs, but not ionic species such as ionic metals, salts of organic acids, or very
polar organic chemicals. Neutral organic chemicals that are hydrophobic (i.e., with log
Kow values > 3) will be concentrated significantly above ambient levels. In reality, any
compound with a log Kow > 1 will be concentrated by the SPMD, but for compounds with
log Kow values < 3, there is no significant advantage in using SPMDs in lieu of other
sampling techniques.
When placed in an aquatic environment, SPMDs passively accumulate hydrophobic
organic compounds, such as PCBs. The LDPE tubing mimics a biological membrane by
allowing the selective diffusion of organic compounds. Triolein is a major nonpolar lipid
found in aquatic organisms. The passive sampling of the hydrophobic organic chemicals
is driven by the mechanism of membrane- and lipid-water partitioning (See Figure 2).
SPMDs can be deployed for long periods of time (on the order of days to months) and
can be used to estimate the time-weighted mean concentrations of the hydrophobic
organic compounds in the water body. The SPMD is placed on a rack, which is then
placed within a protective "shroud." Once the rack is added to the shroud, the device is
ready for use in the water. An SPMD can be oriented vertically or horizontally as
illustrated below:
An SPMD will effectively sample 0.5 to 10 L of water per day, depending on the
chemical's hydrophobicity (as quantified by its water solubility or octanol-water
partitioning coefficient, Kow) and other factors. A compound with log Kow of 6 would
need 200 days at an effective sampling rate of 10 L per day to reach 90 percent of
equilibrium. However, during the first 50 days, the uptake rate into the SPMD is linear.
The concentrations of these chemicals in rivers can change daily or even hourly. To get a
true picture of the concentration of contaminants present in the water column, it would be
necessary to collect and analyze and significant number of samples. The SPMD allows
the calculation of a cumulative time-average of the concentration of each contaminant
while the SPMD was in the water.
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
The ambient "truly dissolved" water concentration (Cw) can be estimated based on the
concentration in the SPMD (Cspmd), the volume of the SPMD (vspmd), the effective
sampling rate (Rs), and the time of deployment (t):
Cw = Cspmd Vspmd/ (Rs*t)
After a typical deployment period of approximately 15 to 30 days, the SPMDs are
removed from the aquatic environment and recovered via dialysis using a nonpolar
solvent such as hexane. This extract is then reduced, cleaned up, and enriched. The
cleanup procedure typically includes gel permeation chromatography. This process
removes any lipid and polyethylene waxes that might have carried over during the
dialysis extraction. Further clean-up can be performed during enrichment on an activated
alumina and silica gel column. The enriched extract is then analyzed for target
compounds using chromatographic techniques.
A major portion of the sequestered residues can be recovered by opening the ends of the
SPMD polyethylene tube and rinsing out the lipid with an organic solvent. However,
analytes are generally recovered by dialyzing the intact SPMD (which requires removing
periphytic growths, minerals, and debris from the exterior membrane surface) in an
organic solvent such as hexane. Using this approach, contaminant residues present in the
membrane (sometimes representing as much as 50 percent of the total) are also recovered
for analysis and the dialysis process separates nearly all of the bulk lipid from the
chemicals of interest.
A problem inherent with the deployment of SPMDs lies in the biofouling layer, the
coating found on the membrane exterior. This biofouling layer can impede flux across the
membrane, thus slowing the effective sampling rate (Rs). This impedance factor is
specific to each SPMD at any given point in time. Impedance for a specific deployment
can be quantified by measuring the loss of a surrogate compound (contained within the
SPMD) during deployment.
The SPMD sampling rates are directly proportional to the SPMD membrane surface area.
For example, a standard 1-g triolein SPMD (surface area about 450 cm2) may extract 5 L
of water per day for a PCB congener, whereas a standard triolein SPMD with half the
surface area (225 cm2) (0.5-g of lipid) can be expected to extract 2.5 L of water per day
of the same congener, assuming similar conditions of exposure.
Due to the highly sensitive nature of the SPMDs, assembly and placement of the devices
requires considerable care. According to Huckins et al. (1996), the following quality
control (QC) procedures must be followed during the SPMD preparation phase:
Use of synthetic triolein or lipid, with all new lots or batches analyzed for
contaminants, amputated, and stored in a freezer until use;
Accurate delivery of small volumes of triolein requires the use of a micropipettor
equipped with a total displacement plunger;
Hudson River PCB s Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
Batch-extraction of SPMD tubing with nanograde hexane or cyclohexane just
prior to use in SPMD construction;
Enclosure of triolein in SPMD layflat tubing using a heat sealer, which results in a
molecular weld;
After assembly, SPMDs are sealed in clean, gas-tight paint cans (solvent rinsed to
remove cutting oils) or gas phase sampling bags (Tedlarฎ) for transport to
deployment sites.
Placement of the devices is important due to a variety of factors. According to Huckins et
al. (1996), the following quality control (QC) procedures must be followed during the
deployment phase:
Use of plastic components should be minimized, with the exception of Teflon,
due to the possible presence of leachable organic residues;
The design of the structure to hold the SPMD should minimize abrasion of the
membrane; and
Since the SPMD membrane generally controls uptake, current velocity is usually
only a concern in terms of abrasion and tethering.
Another important phase to consider is the recovery and storage of SPMDs. According to
Huckins et al. (1996), the following QC procedures must be followed during this phase:
As soon as SPMDs are recovered from the environment, they should be sealed in
the original can or Tedlar bag and placed on ice. The devices should be shipped to
the processing laboratory overnight; and
SPMDs should be stored in the original container at -20ฐ C until they are
analyzed.
During dredging activities in the Hudson River, SPMDs could be deployed at the far-field
stations for periods of 15 days. The dissolved phase PCB concentration in the water
column over the two- week period can then be determined. It should be noted that these
measurements should be regarded as qualitative and used to measure relative changes in
the water column concentration over successive two-week periods.
Hudson River PCB s Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
2.5 Trace Organic Platform Sampler (TOPS)
The detection of trace organic compounds in the water column is generally very
problematic because many target compounds are typically present at concentrations that
are below the detection limits of conventional analytical methods. In these instances, a
non-detect result generally represents a failure in field sampling and/or laboratory
analysis to measure these target compounds at environmentally relevant concentrations.
Available analytical methods require large sample volumes to resolve concentrations in
the picogram to femtogram per liter range. In environmental settings where
concentrations are known to be exceedingly low, collection of large grab samples can be
logistically difficult and cumbersome. Field processing of samples in these settings
greatly simplifies the collection process while significantly lowering detection limits.
In order to overcome these difficulties, the NYSDEC developed the TOPS as a tool to
obtain water column samples. The TOPS is composed of a set of plumbing, pumps, and
sensors that concentrate hydrophobic organic compounds from surface waters. The
device is designed to collect suspended solids using glass fiber cartridge filters (1 micron
pore size), and to capture dissolved-phase organics (e.g., dissolved phase hydrophobic
organic compounds like PCBs) using the synthetic resin Amberlite XAD-2 (XAD). The 1
micron pore size filters were chosen for two reasons:
They are readily available in desirable configurations, and
They were assumed to be efficient at capturing most of the suspended solids in
river settings.
XAD is a polymeric adsorbent of hydrophobic cross-linked polystyrene copolymer
supplied as 20-60 mesh beads. The beads are an agglomeration of many microspheres,
providing a continuous gel phase and a continuous pore phase. The XAD surface area is
300 m /g, and the open cell porous structure allows water to easily penetrate the pores of
the resin. During the adsorption process, the hydrophobic portion of the adsorbate
molecule is preferentially adsorbed on the hydrophobic polystyrene surface of the resin,
while the hydrophilic section of the adsorbate remains oriented in the aqueous phase.
Compounds adsorbed do not penetrate into the microsphere phase; they remain at the
surface where they can be easily eluted. Unlike liquid/liquid extraction procedures, it is
easy to scale up XAD sampling systems to treat exceptionally large volumes of water.
These large water volumes have a greater likelihood of containing a detectable mass of
the target organic analyte than smaller volumes.
The best use of the TOPS is to obtain whole-water concentrations of extremely dilute
hydrophobic organic compounds. With adequate support, the TOPS is a very powerful
field tool that can be deployed from ships or fixed locations where sample size is
unlimited. In such cases, there is virtually no detection limit as more analyte can be
obtained simply by pumping more water. The TOPS typically processes more than 5,000
liters in order to achieve adequate detection of target compounds. Where field setup is
inconvenient and concentrations are expected to be relatively high, TOPS can be used in
Hudson River PCBs Superfund Site
Engineering Performance Standards
14
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
bench-top mode. Samples on the order of tens of liters can be brought in from the field
and batch processed.
In its original configuration, the TOPS was run by an on-site operator for a fixed length
of time (as short as one day) or at fixed intervals to sample wastewater effluent, coastal
waters, and other environments with a low level of suspended sediments. The USGS, in
cooperation with the NYSDEC, modified the TOPS for operation in river environments
where suspended sediment concentrations are relatively high. Additional TOPS
modifications allow for remote, automated, and flow-weighted operation (USGS, 2003
and NYSDEC, 2003).
The TOPS uses 110 VAC and processes water through cartridge filters (available in 4 or
10 inch lengths), and through XAD columns at a maximum rate of 620 mL/min. The
TOPS can process water at a much greater rate through the filter (3,200 mL/min) than
through the XAD, so significant amounts of suspended solids may be captured even in
waters with low levels of suspended solids. Since the pump rates through the glass fiber
filters and through the XAD are independent, sampling rates can be adjusted depending
on the turbidity of the water.
Remote and automated operation was made possible by adding a Campbell CR10X data
logger that performs the following tasks:
Monitors stream stage,
Triggers sample collection based on stream discharge, and
Monitors flow through the XAD resin and filter and backpressure associated with
the filter.
A modem connected to the data logger allows a user to dial into the site to initiate,
monitor, or stop sampling. Hydrologic events rarely occur at convenient times, so data
logger programming includes a set of conditions under which the TOPS will begin
sampling automatically. These conditions usually take the form of a threshold change in
river stage over time, but could include a variety of other programmable triggers,
including river discharge. Ending the sampling activities performed by the TOPS can also
be accomplished either manually or automatically. Automatic termination based on river
stage is set for when the stage falls 80 percent of the difference between the event start
stage and peak stage.
The collection of composite samples during periods of changing river discharge is best
accomplished by flow-weighting the volume of water collected. Flow-weighting is a
method by which the volume of sample water collected is proportional to the volume of
water passing the sample station. Flow-weighting, as compared to fixed interval
sampling, avoids the over-representation of conditions present during the beginning and
end of the event and under-sampling of the the mass flux of contaminants passing during
the hydrologic peak.
Hudson River PCBs Superfund Site
Engineering Performance Standards
15
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
The contaminant flux can be determined by multiplying the contaminant concentration
derived from the flow-weighted samples by the mean river discharge during the sampling
period, and then converting to the appropriate units. In practice, flow-weighting is
accomplished by collecting a fixed volume or sub-sample of river water every time a pre-
set volume of river water passes the sampling station. This pre-set river water volume is
an educated guess based on the anticipated river discharge maximum, expected duration
of the event, and minimum sample volume required. Real time discharge data is required
to collect a flow-weighted sample. The interval for collection of discharge data is
dependent upon a variety of factors, but is principally dictated by the pre-set volume of
river water used to trigger a sub-sample; in NYSDEC's application under the
Contaminant Assessment and Reduction Project (CARP), discharge data were typically
collected once per minute.
To allow sampling when suspended sediment concentrations are high, another pump was
added to the sampling system that delivers a flow-weighted sample to a
settling/compositing tank. In this configuration, the TOPS draws water from the tank
instead of directly from the river. The tank sits on a scale which is monitored by the data
logger. The mass of the water in the tank is used to control when the TOPS turns on and
off and when the river water pump should turn off. The tank allows material that would
otherwise prematurely clog the TOPS filter to settle. Settled material in the tank is
collected and filtered at the end of the event, and composited for analysis with the TOPS
filter. The advantages of using an additional pump in the sampling process include:
The use of pumping rates that keep material in suspension without compromising
the integrity of the TOPS filter;
The ability to purge the sampling line before and after a sampling interval; and,
The removal of the TOPS from the role of collecting a flow-weighted sample.
The addition of the settling tank to the TOPS system is primarily designed to extend the
life of the TOPS cartridge filter; material settling to the bottom of the tank avoids TOPS
filtration, thereby reducing the amount of material on the filter and prolonging filter life.
Besides this obvious advantage, the tank has several additional benefits that improve the
quality of sample collection. Without the tank, the main TOPS pump must collect and
process the sample directly from the river, which requires the main pump to pull water
from the river at a rate of at least 2 ft/sec to keep material in suspension. The filter may
be able to process the volume of water required, but when the filtration is time
constrained, the result is an increase in backpressure from the filter to the point where the
TOPS shuts down. Additionally, as the filter accumulates sediment and backpressure
builds, the effective pumping rate decreases with time, introducing bias into the sample
collection in that the efficiency of the point intake to collect suspended material changes
over time.
Hudson River PCBs Superfund Site
Engineering Performance Standards
16
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
By removing the main TOPS pump from service as the direct collector of river water, the
pump rate of the main TOPS pump can be significantly slowed. Slower filtration reduces
backpressure from the cartridge filter and extends processing time. Slower pump rates
also reduce the formation of air bubbles in the sampling line produced from the degassing
of sample water under rapidly changing pressure conditions. Air bubbles can adversely
impact the accuracy of the flow meters, which are critical in determining contaminant
concentrations. The tank also gives the operator time to get to the site in the event that
maintenance is needed. Sub-samples can be composited in the tank at the following
times:
At the beginning of the event before,
During installation of the TOPS cartridge filter and XAD, and
During the event to change a clogged filter.
By remotely monitoring river conditions and TOPS backpressure, sub-samples can be
collected without interruption over the course of the hydrologic event.
A further advantage the tank- and sub-sample pump combination has over direct TOPS
pumping is that the sub-sample pump can flush excess water remaining in the line
following collection of a sub-sample without adversely affecting TOPS processing or
pumping sample water back to the river. Without intake line flushing, the sub-sample
water collected directly by the TOPS may be partially or entirely made up of water that
remains in the sample line from the previous sub-sample. In addition, part or all of the
sample water collected may not adequately represent the suspended sample fraction in
that settling of suspended material occurs in the sample intake line between sub-samples
- this is particularly a problem in locations requiring long sections of vertical or near
vertical sample line.
Wound glass fiber cartridge filters are capable of filtering large volumes of water without
clogging, but have the disadvantage of allowing more suspended material to pass through
than conventional plate filters with the same nominal pore size. Experiments conducted
to test the efficiency of the 10-inch cartridge filters (both 0.5 and 1 micron nominal pore
size) indicate the efficiency changes with the volume of water processed, often times in
unexpected ways, but generally in response to material loading of the filter. Over the
course of these tests, both filter pore sizes trapped between 85 and 89 percent of the total
mass of sediment sampled with pre-filter concentrations ranging from 3 to 82 mg/L. The
TOPS can be equipped with a series of solenoid valves to periodically divert a sub-
sample of water to a sample container. These valves and containers can be placed after
the filter to assess the overall trapping efficiency of the filter.
A conventional automatic sampler is used with the TOPS to help interpret and support the
organics data collected by the TOPS. This sampler collects discrete sample pairs for
analysis of suspended sediment concentration and particulate and dissolved organic
carbon concentrations. Sediment and organic carbon samples are collected at the
Hudson River PCBs Superfund Site
Engineering Performance Standards
17
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
beginning, end, and peak of the hydrologic event, in addition to measured changes in
stage (e.g. every 0.5 feet of stage change).
2.6 ISCOฎ Portable Water and Wastewater Samplers
All of the portable samplers manufactured by ISCO can be divided into two groups: the
full-size sampler and the compacted sampler. The compacted samplers are specially
designed for locations with limited access, for example a manhole. The full-size sampler
requires a larger space for installation. The open-channel flow conditions at the far-field
monitoring stations in the Hudson River would be appropriate for a full-size sampler.
The Model 3710 sampler is a composite-only portable sampler that combines simple
operation and high volume capacity for single-bottle sampling. The unit collects
composites samples, based on time or flow interval, in a 2.5 gallon glass or polyethylene
bottle or a 4 gallon polyethylene bottle. Up to 24 sampling stop and resume times can be
preset for unattended, automatic sampling. The controller can be set up for uniform time
interval, non-uniform time interval, and flow-paced sampling with or without time delay.
The Model 3700 sampler collects sequential or composite samples based on time, flow
rate, or storm conditions. It is an ideal choice if the parameter monitoring and logging
capabilities are not needed. The exclusive LD90 provides automatic compensation for
changes in head height, plus automatic suction line rinsing to prevent cross
contamination. Basic and extended programming modes are provided for uniform time
intervals, non-uniform time intervals, stormwater runoff sampling, multiple bottle
compositing, and split sampling. The bottle configurations for composite sampling are
the same as for Model 3710. Sequential sampling bottle configurations include 24 x 1
liter polypropylene or 350 ml glass, 12 x 1 liter polyethylene or glass, and 4x1 gallon
polyethylene or glass.
Both the Model 3710 and Model 3700 pumps maintain the USEPA-recommended 2 feet
per second (fps) line velocity at head heights up to 16 ft, with '/4-inch suction line. For
higher lifts, the 6700 series is recommended. The 6712 Portable Sampler is the most
sophisticated full-size sampler that ISCO produces. Samples can be delivered at the
USEPA-recommended velocity of 2 fps, even at a head height of 26 feet.
The plug-in 700 Series Modules and the new SDI-12 interface make it easy to add flow
and parameter monitoring to the basic system. The 6712 Controller allows the user to
select different programming modes to assure the most suitable routine for specific
application. The included 4MB of memory gives the user great flexibility for logging
environmental data. Choice of 11 different glass and plastic bottle configurations ranges
from 24 x 1 liter to 1 x 5.5 gallon.
All the samplers require the power of 12 VDC. Ni-cad lead-acid batteries can be
purchased from ISCO. But depending on the sampling frequency and the volume of one
sample, the battery can last only 1 to 3 days. To meet the 2-week continuous sampling
Hudson River PCBs Superfund Site
Engineering Performance Standards
18
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
requirement set for the routine monitoring, the most convenient and economic way to
provide the power for the sampler will be to provide the electricity to the sampling
location.
The purchasing costs are as follows:
$1,975 for Model 3710,
$2,425 for Model 3700; and
$2,700 for Model 6712.
To analyze PCB appropriately, the laboratory requires a 16-L sample. The 5-gallon
container is needed to collect sufficient amount of water sample. Given the features of
these samplers and the needs of this project, Model 3700 and Model 3712 would be the
better choice. The details regarding how to deploy the samplers during remediation
monitoring should be fully addressed in the design phase.
Hudson River PCB s Superfund Site
Engineering Performance Standards
19
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
3.0 References
Agrawal, Y.C., and Pottsmith, H.C., "Laser diffraction particle sizing in STRES," Cont.
Shelf Res. 14, pi 101-1121 (1994).
Agrawal, Y. C, Pottsmith, H.C., Lynch, J., Irish, J., "Laser instruments for particle size
and settling velocity measurements in the coastal zone," Proc. IEEE Ocean. Engineer.
Soc., 1135-1142(1996).
Agrawal, Y.C., and Pottsmith, H.C., "Instruments for Particle Size and Settling Velocity
Observations in Sediment Transport," Marine Geology, 168 (l-4):89-l 14 (2000).
American Public Health Association (APHA), American Water Works Association, and
Water Environment Federation. Standard Methods for the Examination of Water and
Wastewater. (20th Edition)
Bale, A.J., and Morris, A.W., "In situ measurement of particle size in estuarine waters,"
Estuar. Coast. Shelf Sci. 24, p253-263 (1987).
Bale, A.J., "In situ laser optical particle sizing," J. Sea Res. 36, p31-36 (1996).
Gentien, P., Lunven, M., LeHaitre, M., and Duvent, J.L., "In-situ depth profiling of
particle sizes," Deep-Sea Res. 142, 1297-1312 (1995).
Hach, C.C., "Understanding turbidity measurement," Industrial Water Engineering, 18-22
(1972).
Huckins, J.N., Petty, J.D., Lebo, J.A., Orazio, C.E., Prest, H.F., Tillitt, D.E., Ellis, G.S.,
Johnson, B.T., and Manuweera, G.K. Semipermeable Membrane Devices (SPMDs) for
the Concentration and Assessment of Bioavailable Organic Contaminants in Aquatic
Environments. Chapter in Techniques in Aquatic Toxicology. G.K. Ostrander, Ed. CRC
Lewis Publishers, Boca Raton, FL. (1996)
Lillycrop, L.S., Howell, G.L., and White, T.E., "Development and evaluation of of an in-
situ, long-term turbidity sensor," Technical Report CERC-96-9, U.S. Army Research and
Development Center, Vicksburg, MS. (1996).
McCabe, J.C., Dyer, K.R., Huntley, D.A., and Bale, A.J., "The variation of floe sizes
within a turbidity maximum at spring and neap tides," Coast. Engineer. 3, 3178-3188
(1993).
McCarthy, J.C., Pyle, T.E., and Griffin, G.M., "Light transmissivity, suspended
sediments and and the legal definition of turbidity," Estuarine and Coastal Marine Marine
Science, 2, 291-299 (1974).
Hudson River PCBs Superfund Site
Engineering Performance Standards
20
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
Melis, T.S., Topping, D.J., and Rubin, D.M., "Testing laser-based sensors for continuous,
in-situ monitoring of suspended sediment in the Colorado River, Arizona." Proceedings
of the Eighth Federal Interagency Sediment Conference. Reno, Nevada. (2002).
Mikkelsen, O.A., and Pejrup, M., "In situ particle size spectra and density of particle
aggregates in a dredging plume," Marine Geology, 170, p443-459 (2000).
NYSDEC, 2003. Personal communication from Simon Litten of the New York State
Department of Environmental Conservation to Maheyar Bilimoria of TAMS, an
EarthTech Company.
Petty, J. D., Orazio, C. E., Huckins, J.N., Gale, R. W., Lebo, J.A., Meadows, J.C., Echols,
K. R., and Cranor, W.L., Journal of Chromatography A, 879 p 83-95, (2000).
Sutherland, T.F., Lane, P.M., Amos, C.L., and Downing, J., "The calibration of optical
backscatter sensors for suspended sediment of varying darkness levels," Marine Geology,
162,p587-597 (2000).
Thackston, E.L., and Palermo, M. R., "Improved Methods for correlating turbidity and
suspended solids monitoring," DOER Technical Note E8, U.S. Army Research and
Development Center, Vicksburg, MS. (2000).
Traykovski, P., Latter, R.J., and Irish, J.D., "A laboratory evaluation of the laser in situ
scattering and transmissometery instrument using natural sediments," Marine Geology
159, p355-367 (1999).
Tubman, M.W., "Plume measurement system (PLUMES) technical manual and data
analysis project," Technical Report DRP-95-1, U.S. Army Research and Development
Center, Vicksburg, MS. (1995).
USGS, 2003. Personal communication from Gary R. Wall of United States Geological
Survey, New York District to Maheyar Bilimoria of TAMS, an EarthTech Company.
Van der Lee, W.T.B., "The impact of fluid shear and the suspended sediment
concentration on the mud floe size variation in the Dollard estuary, The Netherlands.
Sedimentary processes
in the intertidal zone, Black, K.S., Paterson, D.M., Cramp, A. (Eds.). Geol. Soc. Lond.
Spec. Publ. 139, pl87-198 (1998).
Vanous, R.D., "Understanding nephelopmetric instrumentation," American Laboratory,
67-79 (1978).
Zaneveld, J.R., Spinrad, R.W., and Bartz, R., "Optical properties of turbidity standards,"
Society of Photo-Optical Instrumentation Engineers, Vol. 208, Ocean Optics VI, 159-168
(1979).
Hudson River PCBs Superfund Site
Engineering Performance Standards
21
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
Figures
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment F - April 2004
-------
Figure 1 Laser diffraction principles - a cut away view of the basic LISST-100 instrument.
A collimated laser beam illuminates particles (left to right). Multi-angle scattering is sensed by a specially constructed photo-
diode array placed in the focal plane of the receiving lens. The array detector has 32 concentric rings, placed in alternate
quadrants. An aperture in the center passes the attenuated beam for measurement of optical transmission.
Hudson River PCBs Superfiind Sits
Engineering Performance Standards
Malcolm Pirnie/lTAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
Semipermeable Membrane Device (SPMD)
t*-wrei
ปMa .* hi
(IlKfi5 '
4t#' -
SPMD
E
Water
will* *p
II
Tft* lipid tdn'ainlnfl t >tn tpi FrncaLMo rttombfiHc' dซฅicซ IS PUD)
and ฆ typical daplcym-snl apoaralu*..
A VERTICAL DEPLOYMENT APPARATUS FOR SPMDs
Designed by Barry Pout Ion and Brad Mustier at CERC
I
V /
Floats
if
Removable
Doors
Rear View
v
ป
Top View
Anchor
20 meter*
of Palyrape
Cable
Harness
X
I'empmentor
Case
Steel Rebar
Side f te*
A HORIZONTAL DEPLOYMENT APPARATUS FOR SPMDs
Designed by Jon Lobo. of CSFfC
Til
jE
k-
Figure 2. SPMD Apparatus
Hudson River PCBs Superfiind Sits
Engineering Performance Standards
Malcolm Pirnie/lTAMS-Earth Tech
Volume 2: Attachment F - April 2004
-------
Attachment F-l
Literature Review
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
Attachment F-l
Literature Review
1.0 Introduction
PCB concentrations cannot be measured quickly or easily in the field, requiring time-
consuming laboratory analyses. Turbidity and total suspended solids (TSS) can be
measured relatively quickly and easily using real-time monitoring devices. To develop an
estimate of the real-time PCB concentration in the vicinity of the dredging operations, the
development of relationships between turbidity and TSS and TSS and PCB
concentrations will be investigated.
Analysis of TSS and PCB data from a set of GE water column monitoring samples did
not yield a correlation between the two parameters. Based on this observation, the PCB
concentrations in the near-field will be projected using modeled solids concentrations
(obtained using the DREDGE and/or SED20 models), consideration of the travel time,
average concentrations in each river section, and an estimate of the time to reach
equilibrium between the dissolved and suspended phases. It is not anticipated that PCB
concentrations will be measured in the near-field during remediation.
PCB concentrations will be measured at the far-field stations, via sampling and analysis,
and the levels will be compared with the TSS levels from the near-field stations to
determine if a correlation exists. Phase 1 of the remediation will provide information that
can be used to further refine any observed relationship between near-field solids and far-
field PCB concentrations; refinements could be incorporated in the Final Phase 2
Engineering Performance Standards. The papers below were reviewed to investigate the
feasibility and applicability of such a correlation.
Hudson River PCB s Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
2.0 Paper List
1. Chattooga River Watershed Ecological/Sedimentation Project (Pruitt at al., 2001)
2. Improved Methods for Correlating Turbidity and Suspended Solids for
Monitoring (Thackston et al., 2000)
3. St. Lawrence River Sediment Removal Project Environmental Monitoring Plan:
Section 2: Pre-Sediment Removal Data Collection (BBL Environmental Services,
Inc., 1995)
4. Use of Acoustic Instruments for Estimating Total Suspended Solids
Concentrations in StreamsThe South Florida Experience (Patino et al., USGS,
2003)
5. Appendix K: Water Quality Monitoring Pre-Design Field Test Dredge
Technology Evaluation Report, New Bedford Harbor Superfund Site, Section
K.6.2 (USACE, 2001)
6. Suspended Solids Flux Between Salt Marsh and Adjacent Bay: A Long-term
Continuous Measurement (Suk et al., 1999)
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
3.0 Chattooga River Watershed Ecological/Sedimentation Project
(Pruitt at al., 2001)
The purpose of this study was to conduct a sediment-yield evaluation and analyses to
determine if sediment was a primary cause of physical and biological impairment to
streams within the Chattooga River watershed, located in northeast Georgia, northwest
South Carolina, and southwest North Carolina. This goal was achieved by sampling
sediments and aquatic ecology from different areas of the watershed and correlating the
data by site.
For the aquatic ecological analysis, a total of three reference sites and 56 other sites from
six subwatersheds (Headwaters, Lower Chattooga, Middle Chattooga, Stekoa Creek,
West Fork, and Warwoman Creek) were sampled. Biological sampling methods were
focused on benthic macroinvertebrates and used modified rapid bioassessment protocols.
Reference sites were chosen prior to sampling based on habitat condition, in-situ water
chemistry, and surrounding land use. Two of the reference sites were located on the
Chattooga River, and one on the upper Chattahoochee River located outside of the
Chattooga watershed. Data from all stations were analyzed using a multimetric approach:
17 metrics were calculated from the raw data, and ultimately the five of those that had the
greatest ability to detect impairment were selected.
For sediment sampling, 17 stream reaches were selected for storm flow investigations
based on the following criteria: relative degree of biological impairment as measured
using modified rapid bioassessment protocols, position within the watershed, relative
geomorphic condition, and access logistics. Storm flow investigations were performed
during three storm events in March 1998, June 1999, and March 2000. A total of 58
observations were made across the 17 stations.
Total suspended sediment (TSS) was analyzed through the filtration of whole water
samples and in accordance with USEPA Method 160.2. Bedload samples were collected
using a 6-inch cable-suspended bedload sampler or a 6-inch wading type bedload
sampler. The samples were transported to the laboratory in 1-liter containers, and
processed for particle size determination in the laboratory using the EPA-SESD wet sieve
method. Laboratory results of dry-weight bedload samples (Mb, grams) were converted to
bedload transport rate (Qb, tons/day) by the following equation:
Qb = K(Wt/T)Mt
where: Qb = bedload discharge (tons/day)
K = converts grams/second/foot to tons/day/foot
Wt = wetted surface (ft)
T = total time sampler on bottom (seconds)
Mt = total mass of samples (grams)
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
The amount of bedload sediment measured over the course of the three storm events
averaged 13.32 tons/day, with mean particle sizes ranging from fine sand to very coarse
sand. On average, the bedload sediments only accounted for 14% of the total sediment
load. The TSS averaged 85.3 tons/day over the course of the three storm events, making
up 86% of the total sediment load on average. Total sediment load (bedload sediment +
TSS) was compared to discharge and road density (road length/corresponding drainage
area). Road density is a factor that represents the net impacts of road construction and
maintenance, interception of subsurface interflow, routing of other non-point sources to
the stream, and entrainment, mobilization, and transport of sediment to the stream.
Study results indicated that the biological conditions in most of the streams sampled
showed little or no impairment due to sedimentation effects. 78% rated "very good" or
"good," 19% rated "fair," and 3% rated "poor." None rated as very poor. Although some
sedimentation or habitat effects of sedimentation were evident at many sites, a negative
biological response was not always presented. The most degraded biological community
was observed in the Stekoa Creek subwatershed. Data indicated that impaired streams
contained a higher concentration of bedload and suspended load sediments when
compared to the reference streams. Study results also indicated that the road density and
sediment sources associated with the road density were the source of 51% of the total
sediment loading.
Good correlation was observed between the biological index and the normalized TSS
data. Data suggest that a TSS concentration normalized to discharge/mean discharge
greater than 284 mg/1 adversely affected the biological community structure. However,
based on regional concentrations, a normalized TSS concentration of 58 mg/1 or less
during storm flow provides an adequate margin of safety and is protective of aquatic
macroinvertebrates in the area. Corresponding turbidity limits of 22 and 69 NTU
represent the margin of safety and threshold of biological impairment.
Reference
Pruitt, B. A.; Melgaard, D. L; Howard, H.; Flexner, M. C.; Able, A. "Chattooga River
Watershed Ecological/Sedimentation Project," FISC Proceedings, Federal Interagency
Sedimentation Conference, Reno, Nevada, March 26-30, 2001.
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
4.0 Improved Methods for Correlating Turbidity and Suspended
Solids for Monitoring (Thackston et al., 2000)
This article describes techniques that are traditionally used to measure turbidity and
suspended solids in water, how the two parameters relate to one another and to various
environmental impacts, and why one cannot be routinely substituted for the other. This
paper also outlines techniques describing the use of quick turbidity measurements as aid
to monitoring dredging and dredged material disposal operations.
Turbidity and suspended solids are common parameters of concern for regulatory
agencies, and thus are often included in the environmental monitoring plans for dredging
operations. Because suspended solids measurements cannot be made quickly and easily
in the field, turbidity measurements are often taken instead. While turbidity can be
measured quickly, there is no universal correlation between the two parameters, or
between turbidity measurements taken from different suspensions or the same suspension
with a different instrument. However, turbidity can be used as an indicator on a site-
specific basis.
Total suspended solids (TSS) include both inorganic solids and organic solids. TSS is a
measure of the dry weight of suspended solids per unit volume of water, and is reported
in milligrams of solids per liter of water (mg/1).
Turbidity is an optical property of water that causes light to be scattered and absorbed
rather than transmitted in straight lines through the sample, and is reported in
Nephelomatic Turbidity Units (NTUs). The source of turbidity in a sample includes
suspended inorganic and organic matter, water molecules, and dissolved substances. The
ability of a particle to scatter light depends on the size, shape, relative refractive index of
the particle, and the wavelength of the light.
There is no universal correlation of TSS and turbidity, but sediment-specific correlations
are useful as a real-time indicator of suspended solids. Such correlations have been
developed in the laboratory using whole sediment samples. Generally, any samples used
to produce a correlation between TSS and turbidity must be suspension-specific, not just
site-specific. The sample must approximate the suspension to be representative of the
size, number, shape, and type of particles present.
Most discharge or monitoring permits that are associated with dredging operations are
based on TSS rather than turbidity because TSS correlates well with environmental
impact and is at least roughly comparable from site to site and sediment to sediment.
It has been suggested that there are three general situations where a TSS-turbidity
correlation curve may serve as an aid in the routine monitoring of a dredging operation:
Solids resuspension in the immediate vicinity of the dredge (20-50m) where
most solids will be continuously replenished by dredging actions.
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
Containment area effluent, where only the finer particles will be present due to
the settling of larger, heavier particles near the point of inflow for the
contaminant disposal facility. For this case, a laboratory settling column and test
procedure would be required to obtain a representative sample.
Open-water dredged material placement where the larger, heavier solids will
begin to settle to the bottom immediately upon leaving the dredge discharge
pipe, hopper, or barge usually in a well-defined plume. This case requires the
use of a laboratory column-settling test to obtain a representative sample.
Reference
Thackston, E. L.; Palermo, M. R. "Improved Methods for Correlating Turbidity and
Suspended Solids for Monitoring," DOER Technical Notes Collection (ERDC TN-
DOER-E8), U.S. Army Engineer Research and Development Center, Vicksburg, MS,
2000.
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
5.0 St. Lawrence River Sediment Removal Project Environmental
Monitoring Plan: Section 2: Pre-Sediment Removal Data
Collection (BBL Environmental Services, Inc., 1995)
The goal of the pre-sediment removal data collection program was to verify bottom
conditions, obtain background water quality information, and obtain a location survey of
the sediment control system in the St. Lawrence River at the GM Massena site. One of
the tasks planned to accomplish these objectives was pre-dredging turbidity monitoring.
To perform real-time monitoring that allowed for a rapid response to changing river
conditions, a water quality parameter that is easily measured and correlates with sediment
resuspension during removal activities must be chosen. Turbidity was the parameter
selected in this case.
A downstream total suspended solids (TSS) maximum limit of 25 mg/1 above background
was defined as the conservative action limit based on two variables: previous
environmental dredging projects and a 1994 site-specific bench-scale laboratory
correlation between TSS and turbidity.
The 1994 bench scale experiment established a site-specific correlation between TSS and
turbidity for the GM Massena site, resulting in the use of real-time turbidity
measurements as a surrogate for TSS measurements. The laboratory-produced
correlation, which is based on a combination of all data points from the treatability test
(including some elevated TSS results (> 300 mg/1) from the beginning of the settling
test), is described by the equation 1 below:
Turbidity (NTU) = 7.3745 + (0.61058 X TSS) + (0.00094375 X TSS2) (1)
with a correlation coefficient of r =0.941
Turbidity monitoring data collected in 1994 indicated that the St. Lawrence River can be
characterized as having a relatively low suspended solids content (based on the
evaluation of background river water samples, which contain < 10 mg/1 TSS) and low
turbidity readings. A regression analysis was rerun by BB&L only including data that fell
within the expected working range, defined as: TSS < 60 mg/1 and turbidity > 60 NTU.
The regression equation 2 calculated is defined below:
TSS (mg/1) = [0.63x x (turbidity in NTU)] + 6.8
(2)
with a correlation coefficient of r = 0.43
Based on the revised regression (2), a turbidity of 28 NTU would correlate to a value less
than 25 mg/1 TSS concentration. Dredging activities would not take place when the
measured TSS background was above 60 mg/1. So, due to the nearly linear relationship
that exists between turbidity and TSS for the St. Lawrence River in the subject area, a
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
turbidity increase of 28 NTUs from upstream to downstream was defined as the action
level for the St. Lawrence Sediment Removal Project during waterborne activities.
Real-time turbidity measurements were obtained from three monitoring locations, one 50
feet upstream of the western extent of the control system and two between 200 and 400
feet downstream of the eastern-most active installations, during the mobilization and
installation of the Phase I sediment control system to evaluate any potential short-term
effects of the operations. Measurements were collected near 50% water depth. Turbidity
was also monitored if visible sediment releases were observed during sheet pile
installations.
Reference
"St. Lawrence River Sediment Removal Project Environmental Monitoring Plan."
Prepared for General Motors Powertrain by BBL Environmental Services, Inc. May
1995.
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
6.0 Use of Acoustic Instruments for Estimating Total Suspended
Solids Concentrations in StreamsThe South Florida Experience
(Patino et al., USGS, 2003)
An acoustic velocity meter (AVM) and an acoustic Doppler velocity meter (ADVM)
were used in a study to estimate the total suspended solids (TSS) concentration in two
southern Florida streams. The AVM system provides information on automatic gain
control (AGC), which is an index of the strength of the acoustic signal recorded by the
instrument as the acoustic pulse travels across a stream. The ADVM system provides
information on acoustic backscatter strength (ABS), which is an index of the strength of
return acoustic signals recorded by the instrument. Both the AGC and the ABS values
increase as the concentration of suspended material increases.
The AVM system was installed in 1993 in the L-4 Canal, a man-made channel in
northwestern Broward. The canal is approximately 40 feet wide and averages between 7
and 8 feet in depth. The water velocities in this canal range from -0.5 to 2.5 feet per
second. The ADVM system was installed in 1997 in the North Fork Stream (a tidal
channel), located in Veterans Park in southeastern Florida. The stream is about 280 feet
wide and averages 8 feet in depth, with water velocities that range from about -1.5 to 1.5
feet per second and a salinity that varies from fresh to brackish (0.2 to 15 mg/1).
Depth integrated samples for TSS were collected at the L-4 Canal site using a DH-59
sampler and equal discharge increment (EDI) methodology, and samples at the North
Fork site were collected using a point sampler at the same depth as the ADVM system
and located 9 feet away from the transducer faces (near the start of the sampling volume).
TSS concentrations ranged from 22 to 1,058 mg/1 at the L-4 Canal site, and from 3 to 25
mg/1 at the North Fork site.
Regression analysis techniques were used to develop empirical and site-specific
relationships between the AGC and ABS results and the TSS and the two sites. The
equation below describes those relationships:
rpgg _ JQ {A*[a + b* log (salinity) + C * log (temperature)] + d * log (velocity) + e}
The relationships obtained using the site-specific equations produced good correlations,
with coefficients of 0.91 and 0.87 at the L-4 Canal and North Fork sites, respectively.
The results suggest that this technique is feasible for estimating TSS concentrations in
streams using information from acoustic instruments.
Reference
Patino, E.; Byrne, M. J. "Use of Acoustic Instruments for Estimating Total Suspended
Solids Concentrations in StreamsThe South Florida Experience," U.S. Geological
Survey, Ft. Myers, FL. Available at
http://water.usgs.gov/osw/techniques/TSS/Patino.pdf. downloaded in Feburary 2003.
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
7.0 Section K.6.2 - Correlation Analysis found in Appendix K:
Water Quality Monitoring Pre-Design Field Test Dredge
Technology Evaluation Report, New Bedford Harbor Superfund
Site (USACE, 2001)
A Pre-Design Field Test was undertaken in order to evaluate the performance of a dredge
system under consideration for use at the New Bedford Harbor Superfund Site. The
objectives of the test focused on the performance of the dredge system. This report
section evaluates the impacts on water quality associated with the test; the following
tasks were performed for the evaluation:
Predictive modeling was used to aid in the design of the water quality
monitoring field program and to assess the utility of modeling for the full-scale
remediation effort.
Field monitoring was performed to assess sediment resuspension during the
dredging operation, to collect water samples for laboratory analysis, and to
ground-truth the predictive modeling.
Laboratory analysis of water samples for TSS and PCBs was performed to
assess water quality impacts.
A correlation assessment between the field and laboratory data was performed.
Three correlation studies were performed on the data obtained from the monitoring
samples:
TSS vs. total particulate PCBs - Analysis of the data revealed an excellent
correlation between the two parameters. The study yielded a coefficient of fit for
the linear relationship of 0.84, suggesting that TSS could serve as a good
indicator of the particulate PCB concentrations associated with operations
similar in scope to the pre-design work.
Total particulate PCBs vs. total dissolved PCBs - Analysis of the data yielded a
poor correlation between these parameters. An exponential function provided a
better fit to the data.
TSS vs. total dissolved PCBS - Analysis of the data provided a poor correlation
between these parameters. An exponential function provided a better fit to the
data.
A review of the individual dissolved/particulate data pairs indicated the following:
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
For the reference samples, the dissolved phase and particulate PCB
concentrations were generally similar on a per liter basis, with the dissolved-
phase concentration sometimes exceeding the particulate concentration.
For the samples impacted by the dredging operations, the total particulate PCB
concentration was generally increased to a much greater degree than the
dissolved-phase PCB concentration.
Analysis of the monitoring data also suggested the following:
A moderate correlation between the total suspended solids measured in the lab
and the turbidity measured in the field. The linear coefficient of fit for these data
was 0.56. Measurement of both parameters from the same water parcel would be
expected to increase the strength of the correlation.
Given the different correlations indicated by the data, turbidity to TSS and TSS
to PCB, the results suggest that field measurement of turbidity could be used as
an indicator of the mobilization and transport of particulate-bound PCBs during
the full-scale remediation activity.
Reference
USACE. 2001. "Appendix K: Water Quality Monitoring Pre-Design Field Test Dredge
Technology Evaluation Report, New Bedford Harbor Superfund Site," Pre-Design Field
Test - Dredge Technology Evaluation Report, New Bedford Harbor Superfund Site, New
Bedford, Massachusetts. Prepared by Foster Wheeler Environmental Corporation,
Boston, Massachusetts. August 2001.
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
8.0 Suspended Solids Flux Between Salt Marsh and Adjacent Bay: A
Long-Term Continuous Measurement (Suk et al., 1999)
The goal of this study was to establish an improved methodology to determine the
suspended solids flux between Schooner Creek, NJ, a tidal salt marsh, and Great Bay,
adjacent to it. The most significant difference in methods used in this study was related to
data collection. Field data were collected continuously from March to October 1996.
A suite of instruments, including a current velocity sensor, a turbidity sensor, an
automatic water sampler, a pressure transducer, and a data logger were placed in (and
around) a location 300 m from the mouth of Schooner Creek, to measure the velocity,
water surface elevations, and suspended solids concentrations of the creek. Water
velocity was measured at a depth corresponding to the mid-depth of the creek at high
tide. The instruments were placed in the water on the deeper side of the creek so that they
would remain submerged.
Total suspended solids (TSS) in the stream were quantified using turbidity as an
indicator. A feasibility study performed prior to the experiment's initiation that examined
593 water samples over 25 different time periods found that that the measured suspended
solids concentrations were statistically related to the measured turbidity. The average
correlation coefficient for flood and ebb time periods averaged 0.827, indicating that
turbidity measurements would provide surrogate measurements of the suspended solids
concentration.
The water flux rate was derived from measurements taken by the submerged instruments
and calculated as a product of the current velocity and the area of the wetted cross
section, and cumulative flow volumes were calculated using the average flow rate for
successive time intervals.
The TSS flux was calculated as the product of the water flux and the TSS concentration.
Two TSS fluxes were calculated:
TSS fluxes for the entire recording period (periods of balance and imbalance)
using TSS concentrations derived from the overall regression relationship.
TSS fluxes for periods of time where the calculated water fluxes were more
balanced, yielding net flux values that were not strongly impacted by a water
imbalance.
Analysis indicated that the flow data are not continuous, and there are several different
natural and artificial factors that may attribute to a water imbalance, though the
researchers decided that net water import or export during a particular time was most
likely due to the measurement of an incomplete cycle of water exchange across marsh
boundaries other than the creek mouth.
Hudson River PCBs Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
The study also calculated a minimum number of water sample sets needed to produce a
reasonably good TSS-turbidity regression relationship. To do so, varying combinations of
water sample sets were used to develop a number of different regression relationships.
The regression relationships were then used in the flux calculations, and the relative error
was calculated.
The following observations were produced from the study:
Data analysis indicated that the cumulative and cycle fluxes calculated for the
entire recording period are considerably uncertain due to an imbalance in the
calculated water fluxes.
Data analysis indicated that the coefficient of correlation between the
cumulative TSS fluxes per tidal cycle and the average TSS concentration
differences was 0.71. The flow-weighted average TSS concentration resulting
from all of the water balance periods during the flood tide was higher than that
during the ebb tide, contributing to a net import of TSS.
Data suggested that, for this study, a reasonably good overall TSS-turbidity
regression was established when five data sets with correlation coefficients
greater than or equal to 0.80 were used.
Reference
Suk, N. S.; Guo, Q.; Psuty, N. P. "Suspended Solids Flux Between Salt Marsh and
Adjacent Bay: A Long-term Continuous Measurement," Estuarine, Coastal, and Shelf
Science, Vol. 49, pp. 61-81, 1999.
Hudson River PCBs Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-l - April 2004
-------
Attachment F-2
PCB Analytical Methods
Detection (Reporting) Limits in Water
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-2 - April 2004
-------
Attachment F-2
PCB Analytical Methods
Detection (Reporting) Limits in Water
1. CLP Method OLM04.1 (September 1998)
Contract-required quantitation limit is 1 //g/L for all Aroclors
(CRQL for Aroclor 1221 is 2 //g/L)
Laboratories can report lower detections (e.g., 0.5 J [//g/L])
2. SW-846 Method 8082 (Rev 0, December 1996)
MDLs (method detection limits) for Aroclors range from 0.054 to 0.90 //g/L
(Method provides no data as to Aroclor-specific MDLs)
3. PCB Congeners - Dual Column GC/ECD (Laboratory-specific)
STL/Colchester Vt (formerly Aquatec)
Detects inidividual PCB congeners at a detection limit of 0.001 //g/L
(Monochlorobiphenyls at 0.005 //g/L)
(Other labs have other methods with varying detection limits)
4. NYSDEC Analytical Services Protocol Low-Concentration Method (91-6)
CRQL is 0.2 //g/L for Aroclors except for 1221 (0.4 //g/L)
5. USEPA Method 505, Revision 2.1 - 1995 (Organohalide Pesticides and PCBs by
mi croextracti on/GC)
MDL for Aroclors 1016, 1248, 1254 - about 0.1 //g/L
MDL for Aroclor 1260 - about 0.2 //g/L
MDL for Aroclor 1242 - about 0.3 //g/L
MDL for Aroclor 1232 - about 0.5 //g/L
MDL for Aroclor 1221 - about 15.0 //g/L
(from Method 505 Revision 2.0, USEPA EMSL, 1989)
6. USEPA Method 508, Revision 3.1 (1995). Determination of Chlorinated
Pesticides in Water by GC/ECD.
Note to method summary states that the extraction is similar to Method
608 (q.v.), and the extract can be analyzed by 508, 525, or 608;
however, no performance data for Aroclors were collected as part of
method development for 508.
EDLs (reporting limits) for most single-component pesticides are in
the 0.01 //g/L to 0.05 //g/L range (a few are higher and a few are
lower).
Hudson River PCBs Superfund Site 1 Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment F-2 - April 2004
-------
This method is supposedly being used by Waterford for monitoring its
drinking water supply. The detection and reporting limits would have
to be developed on a laboratory-specific basis. Multi-component
analytes (such as Aroclors, and also toxaphene and chlordane)
typically have higher reporting limits than single-component
pesticides.
7. USEPA Method 680 (PCBs by GC/MS)
Arocolor detection limits are about 100 //g/L
8. USEPA Method 608 (Pesticides/PCBs by dual column GC)
Aroclor Detection limits 0.5 //g/L (1.0 //g/L for Aroclor 1221)
9. USEPA Method 525.2 (1995 revision)
Method uses solid/liquid extraction by either disk or cartridge and analysis using
quadropole MS or ion trap. MDLs are presented for method analytes for each of
the four possible combinations; except Aroclor MDLs only by disk and ion trap.
Sensitivity is better for more chlorinated arolcors. MDLs range from 0.018 //g/L
for 1260 to 0.054 //g/L for Aroclor 1221.
10. USEPA Method 1668A (December 1999) - Chlorinated Biphenyl Congeners in
Water, Soil, Sediment, and Tissue by HRGC/HRMS.
Detection limits (EMDLs) and reporting limits (EMLs) are provided
for more than 150 congeners in both water and non-aqueous matrices.
Method is more sensitive for less-chlorinated congeners.
Reporting limits for individual congeners range from 50 to 1000 pg/L
(10 pg/L for BZ#2) in water (detection limits [EMDLs] are typically
1/3 to '/2 the reporting limit [EML]).
Reporting limits range from 5 to 100 ng/kg (except 1 ng/kg for BZ#2)
in non-aqueous samples (detection limits [EMDLs] are typically 1/5 to
'/2 the reporting limit [EML]).
11. Green Bay Method. Original method not reviewed (or obtained). Not included
in the GE August Design Support Sediment Sampling and Analysis Plan (Revision
1, August 2002). Reportedly a single-column PCB congener GC/ECD method.
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-2 - April 2004
-------
General notes on units of measure:
"3
g/L = parts per thousand (10" );
mg/L = parts per million (10"6);
//g/L = parts per billion (1CT9);
12
ng/L = parts per trillion (10" );
pg/L = parts per quadrillion.(10"15).
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-2 - April 2004
-------
Attachment F-3
Memo Regarding PCB Analyses; Whole Water Extracts vs. Separated
Particle and Filtrate Extracts
Hudson River PCB s Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
February 25, 2003
To: Kelly Robinson, Earthtech (TAMS)
From: Richard Bopp, RPI
Re: PCB Analyses; Whole Water Extracts vs Separated Particle and Filtrate Extracts
Background
Since I first analyzed Hudson River water samples for PCBs in the late 1970s, I have
been interested in particle/water partitioning. Consequently, I have always filtered the
samples and extracted and analyzed the particles and filtrate separately. In addition,
based on considerations of analytical sensitivity, I have always analyzed large volume
(typically 18 liter) water samples. These procedures were adopted by the USEPA for the
water column PCB samples that we collected and processed as part of the Hudson River
PCBs Reassessment.
Several other important datasets rely on an EPA-approved whole water extraction and
analysis of much smaller volume (typically 1 liter) samples. These include
The USGS monitoring in the upper Hudson. This program provides the longest
historical record of water column PCB levels.
The GE monitoring between Rogers Island and Schuylerville conducted under
consent order with the NYSDEC as part of the remnant deposits monitoring
program. This set of samples, collected approximately weekly since 1997
provides, by far, the most detailed picture of PCB transport ever developed (J.
Tatten, Master's Project, RPI, 2000; Task 3 Final Report to NYSDEC, Contract
C003844, 2000).
In 1993 I was at RPI and supervising the collection and processing of the water column
samples for the Hudson River PCBs Reassessment. As I recall, I suggested that on one of
the transects we collect duplicate samples for PCB analysis through NYDSEC at the
NYSDOH labs. In addition, since their standard procedure was whole water extraction, it
was arranged that at least some of the samples also be analyzed as separate particle and
filtrate fractions. This would allow a more direct comparison with the EPA sample
analysis and provide a test of my general impression that whole water extraction would
not be particularly efficient at recovering particle-associated PCBs. The suggestion was
welcomed at NYSDEC and collaboration was facilitated by the fact that I had been
employed there in 1990-91.
Analysis and interpretation of the data from this exercise was to form the basis of the
Master's project of Christine Juliano. After an initial data gathering and analysis effort,
Christine decided to work on a different project and completed her Master's. My
preliminary look at the data indicated that whole water extraction missed a significant
fraction of the particle-associated PCBs. Although based on very limited data, I have
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
used this observation often to support my geochemical bias toward separate particle and
filtrate extraction and analysis.
Over the past month, I have had two requests for a more quantitative assessment of this
data. Both were related to water column monitoring associated with the proposed
dredging. The first was from Kelly Robinson at Earth Tech (TAMS), the primary EPA
contractor on the Upper Hudson River PCB project. A few days later, Roger Sokol of the
NYSDOH requested similar information specifically for monitoring the Waterford, NY
drinking water supply and raw water intake on the Hudson. I was able to locate files
prepared by Christine Juliano that contained water column PCB data from the upper
Hudson consistent with events described above.
More Detailed Information
The sample ID format and numbering used in the files indicates that the samples were
collected during EPA transect 4 (April 12 to April 14, 1993) at Stillwater (0007),
Waterford (0008), the Hoosic River (0012), Mohawk River (0013), and Green Island
Bridge (site 0014). Two of the samples, Waterford and Green Island, have data for whole
water and separate particle and filtered water analyses. Further confirmation of the
identification of these samples comes from the fact that the TSS levels in the files
prepared by Christine Juliano are identical to those reported for samples TW-0004-0008
(34.0 mg/1) and TW-0004-0014 (39.8 mg/1) in the EPA Database. More specific
collection information can most likely be retrieved from the detailed field notes kept by
Rensselaer personnel and submitted to TAMS a part of the official record of our work
with EPA on the reassessment. The rest of this report will refer to the Waterford (004-
0008, 04/13/93) and Green Island (004-0014, 04/13/93) samples.
As I recall, I was informed that the separation of particulate and dissolved phases for the
NYSDOH analysis was accomplished by pouring the water sample through a soxhlet
extraction thimble. This simple procedure should be comparable to separation by more
standard filtration techniques that typically employ pre-fired glass fiber filters. The
corresponding EPA samples that we collected were filtered by Kevin Reed of RPI
through pre-fired Whatman GF/F filters. Soxhlet extraction thimbles used in PCB
analyses are also treated to minimize blanks. Paper thimbles are typically pre-extracted
and glass fiber thimbles are pre-fired.
Results
In terms of total PCBs, the DOH values reported for the whole water extracts
were about half of the (particulate + dissolved) PCBs in the replicate samples
(Table 1).
At the congener level, whole water extraction yielded results lower than (P + D)
in every case with only one exception (BZ 24, 27). Figures 1 (Waterford) and 2
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
(Green Island) present data for a range of more abundant congeners that together
comprise over half the total PCBs.
The figures also show that the differences between whole water and (P + D)
results tend to be less for the lower chlorinated congeners. This is consistent with
a simple model of the whole water extraction process - complete recovery of
dissolved PCBs and less efficient recovery of particulate phase PCBs.
Based on this first order model applied at the congener level, the whole water
extraction missed 61 ฑ 20% of the particle-associated PCBs in the Waterford
sample (Table 2) and 72 ฑ 13% in the Green Island sample (Table 3).
Implications
The above analysis provides support for the logical assumption that whole water
extraction will result in an underestimate of total PCBs. It is also logical to
assume that the degree of under-recovery would depend significantly on the
details of the procedure (the number of extraction cycles, the solvent used, the
percentage of solvent removed between extraction cycles, the degree of sample
agitation etc.).
If the simple model presented above is applied, the degree of under-recovery will
also depend on the TSS in the sample. Using an average particle extraction
efficiency of 33% (based on the DOH analyses) and an average upper Hudson
PCB particle/water distribution coefficient of 105 (Bopp et al., Final Report to
NYSDEC, Contract C00708, 1985), first-order error estimates can be made.
TSS (mg/1)
% of PCB on Particles
% under-recovery of
total PCBs
2
17
11
10
50
33
40
80
53
100
91
61
This analysis raises the possibility that historical (USGS) estimates of PCB
transport in the upper Hudson that focused on high flow, high TSS, high transport
events may be low by on the order of 50% and suggests a low bias to any
transport estimates that utilize the weekly GE water column monitoring data.
The potential for significant under-recovery of PCBs when using whole water
extractions should be considered in the design of any future monitoring program.
Cc: Roger Sokol, NYSDOH
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
Tables
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment F-3 - April 2004
-------
Table 1. Total PCBs in samples collected April 12-14, 1993 (all PCB concentrations in ng/l)
Particulate
Dissolved
P + D
Whole Water
TSS (mg/l)
Waterford (0008)
DOH EPA
225.4 159.8
74.4 75.0
299.8 234.8
159.9
34.0
Green I (0014)
DOH EPA
227.7 144.5
50.9 53.5
278.6 198.0
110.6
39.8
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
Table 2. 'Waterford
004-0008 004-0008 004-0008
Whole Water Particulate Filtered Water
CONGENER
BZ-10,BZ-4
17
9.5
13
BZ-19
4.8
2.3
3.6
BZ-18
12
5.1
6.9
BZ-15,BZ-17
11
8.5
5.8
BZ-16,BZ-32
4.3
3.5
2.8
BZ-31
7.5
12
4.2
BZ-28
8.3
14
4.6
BZ-20, BZ-33, BZ-53
3.6
5.6
2.1
BZ-52
6.4
9.5
3
BZ-49
5.6
8.8
2.2
BZ-47
4.4
7.6
1
BZ-44
3.7
6.2
1.5
BZ-37, BZ-42, BZ-59
1.2
3.2
1
BZ-41
3.1
5.3
1.2
BZ-70
4.8
11
1.6
BZ-66,BZ-95
7.8
18
2.1
BZ-110,BZ-77,BZ-136
2.3
6.4
0.5
Totals
107.8
136.5
57.1
BZ-24,BZ-27
7.7
2.7
4
Hudson River PCBs Superfund Site
Engineering Performance Standards
sum P+F
%P missed
%T missed
22.5
5.9
12
14.3
6.3
16.2
18.6
7.7
12.5
11
8.6
7.7
4.2
6.5
12.6
20.1
6.9
193.6
Std. Dev.
6.7
24
19
0
23
32
54
55
53
49
49
49
52
71
52
62
61
67
45
Std. Dev. 19
-15
58
48
0
39
57
73
74
73
64
61
55
65
94
64
71
68
72
61
20
-37
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
Table 3. Green Island
004-0014
004-0014
004-0014
Whole Water
Particulate
Filtered Water
sum P +
CONGENER
ng/L
ng/L
ng/L
BZ-10,BZ-4
13
8.5
9.9
18.4
BZ-19
3.8
1.7
3
4.7
BZ-18
7.9
5.1
5.2
10.3
BZ-15,BZ-17
7.4
8.9
4.1
13
BZ-16,BZ-32
3.3
3.5
2.2
5.7
BZ-31
5.3
13
2.5
15.5
BZ-28
5.9
15
2.7
17.7
BZ-20, BZ-33, BZ-53
3
5.7
1.7
7.4
BZ-52
4.6
8.6
2.1
10.7
BZ-49
4.1
8.6
1.5
10.1
BZ-47
3.2
6.7
0.5
7.2
BZ-44
3.7
6.2
1.5
7.7
BZ-37, BZ-42, BZ-59
1
3.1
1
4.1
BZ-41
2.2
9.4
1
10.4
BZ-70
3.3
10
1
11
BZ-66,BZ-95
5.3
15
1.1
16.1
BZ-110,BZ-77,BZ-136
1.7
6.4
1
7.4
TOTALS
78.7
135.4
42
177.4
BZ-24,BZ-27
5.2
2.3
2.8
5.1
Hudson River PCBs Superfund Site
Engineering Performance Standards
%P missed
%T missed
64 29
53 19
47 23
63 43
69 42
78 66
79 67
77 59
71 57
70 59
60 56
65 52
100 76
87 79
77 70
72 67
89 77
72 55
Std. Dev. 13 Std. Dev. 18
-4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
Figures
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment F-3 - April 2004
-------
25
20
15
O)
ฃ
10
5
0 4
CD
00
N
ฆ*7
"*7
m
N
N
cT
CD
CD
N
CO
Hudson River PCBs Superfund Site
Engineering Performance Standards
Figure 1
WATERFORD
~Whole Water
ฆ Part + Dissolved
rl
CO
LO
CM
LO
CD
h-
CD
LO
O
hp
LO
CD
N
CO
N
CD
N
CD
N
CD
N
CD
N
CD
N
CD
N
CD
N
CD
CO
c?
CM
CD
CD
N
CO
N
CD
N
CD
cf
CN
K
c?
N
CO
N
CD
CD
CO
N
CD
N
CD
N
CD
Malcolm Pimie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
Figure 2
GREEN ISLAND
20
18
16
14
12
o) 10
ฃ
8
6
4
~Whole Water
ฆ Part + Dissolved
U
N
CD
o"
N
CD
G)
00
h-
CN
T
00
CO
CM
0)
h-
0)
O
LO
C?
c?
(\l
LO
LO
LO
hp
O)
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
CO
CD
CD
CD
CD
CD
CD
CD
CD
CD
CD
CD
CD
CD
CD
N
CD
N
CD
CO
CO
N
CD
cf
(\l
N
CD
CM
N
CD
K
co
N
CD
CD
CD
N
CD
CD
CO
N
CD
N
CD
N
CD
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Piniie/TAMS-Earth Tech
Volume 2: Attachment F-3 - April 2004
-------
Attachment G
Statistical Justification of the Sampling Frequency
for Phase 1 Monitoring Program
Table of Contents
1.0 Introduction 1
2.0 Estimates of the Tolerable Error for the Monitoring Sampling Frequency Using
Decision Error Feasibility Trials (DEFT) Software 1
3.0 Development of Data Quality Objectives 2
3.1 Statement of the Problem 2
3.2 Identify the Decision 3
3.3 Identify the Inputs to the Decision 3
3.4 Define the Boundaries of the Study 4
3.5 Develop a Decision Rule 6
3.6 Specify Tolerable Limits on Decision Errors 6
3.7 Optimize the Design for Obtaining Data: Results of the Analysis 7
4.0 References 11
LIST OF TABLES
Table 1 Summary of Sampling Frequency Requirements and Expected Error
Rates
Table 2 Summary of Sampling Frequency Requirements and Expected Error
Rates for Automatic Sampler
LIST OF FIGURES
Figure 1 Routine to Evaluation Level - Action level of 300 g/day
Figure 2 Routine to Control Level - Action Level of 600 g/day
Figure 3 Confirmation of the 600 g/day - Action Level of 600 g/day
Figure 4 Routine to Control Level - Action Level of 350 ng/L
Figure 5 Confirmation of the 350 ng/L - Action Level of 350 ng/L
Figure 6 Evaluation Level to Control Level - 300 g/day to 600 g/day
Figure 7 Resuspension Threshold - Confirmation of 500 ng/L
Figure 8 Resuspension Threshold - Confirmation of 500 ng/L (24 hours; 4
samples of 6 aliquots)
Figure 9 Routine to Control Level (350 ng/L, 2-week deployment) or
Evaluation Level to Control Level (350 ng/L, 1-week deployment)
Continuous total PCB sampling requirements
Figure 10 Routine to Evaluation Level - Far-field Baseline to >12 mg/L with
discrete samples every 3 hrs for 24 hrs
Figure 11 Routine to Evaluation Level - Far-field baseline to >12 mg/L with
continuous sampling every 15 min for 24 hrs
Figure 12 Routine to Control Level - Far-field Baseline to >24 mg/L with
discrete samples every 3 hrs for 24 hrs
Figure 13 Routine to Control Level - Far-field baseline to >24 mg/L with
continuous sampling every 15 min for 24 hrs
Hudson River PCBs Superfund Site
Engineering Performance Standards
l
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Attachment G
Statistical Justification of the Sampling Frequency
for Phase 1 Monitoring Program
Table of Contents
LIST OF FIGURES (continued)
Figure 14 Evaluation to Control Level - Far-field Evaluation to Control Level
with discrete samples every 3 hours for 24 hours
Figure 15 Evaluation to Control Level - Far-field Evaluation to Control Level
with continuous sampling every 15 min for 24 hours
Figure 16 Routine to Control Level Near-field River Sections 1 and 3 - baseline
to >100 mg/L with discrete samples every 3 hours for 6 hours
Figure 17 Routine to Control Level Near-field River Sections 1 and 3 - baseline
to >100 mg/L with continuous sampling every 15 min for 6 hrs
Figure 18 Routine to Control Level Near-field River Section 2 - baseline to >60
mg/L with discrete samples every 3 hours for 6 hours
Figure 19 Routine to Control Level Near-field River Section 2 - baseline to >60
mg/L with continuous sampling every 15 min for 6 hrs
Figure 20 Evaluation to Control Level Near-field River Sections 1 and 3 -
baseline to >100 mg/L with discrete samples every 3 hours for 15
hours
Figure 21 Evaluation to Control Level Near-field River Sections 1 and 3 -
baseline to >100 mg/L with continuous sampling every 15 min for 15
hrs
Figure 22 Evaluation to Control Level Near-field River Section 2 - baseline to
>60 mg/L with discrete samples every 3 hours for 15 hours
Figure 23 Evaluation to Control Level Near-field River Section 2 - baseline to
>60 mg/L with continuous sampling every 15 min for 15 hrs
Figure 24 Routine to Evaluation Level - Near-field baseline to >700 mg/L with
discrete samples every 3 hrs for 3 hrs
Figure 25 Routine to Evaluation Level - Near-field baseline to >700 mg/L with
continuous sampling every 15 min for 3 hrs
Figure 26 Automatic Sampler at the Evaluation Level (300 g/day) - 1 sample per
hour for 24 hours
Figure 27 Automatic Sampler at the Control Level (600 g/day) - 1 sample per
hour for 24 hours
Figure 28 Automatic Sampler at the Control Level (350 ng/L) - 1 sample per
hour for 24 hours
Figure 29 Routine to Evaluation Level with Automatic Sampler - Action level of
300 g/day
Figure 30 Routine to Control Level with Automatic Sampler - Action level of
600 g/day
Figure 31 Confirmation of the 600 g/day with Automatic Sampler - Action level
of 600 g/day
Figure 32 Routine to Control Level with Automatic Sampler - Action level of
350 ng/L
Hudson River PCBs Superfund Site ii Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment G - April 2004
-------
Attachment G
Statistical Justification of the Sampling Frequency
for Phase 1 Monitoring Program
Table of Contents
LIST OF FIGURES (continued)
Figure 33 Confirmation of the 350 ng/L with Automatic Sampler - Action level
of 350 ng/L
Figure 34 Evaluation to Control Level with Automatic Sampler - Action Level
600 g/day
Hudson River PCBs Superfund Site
Engineering Performance Standards
m
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Attachment G
Statistical Justification of the Sampling Frequency
for Phase 1 Monitoring Program
1.0 Introduction
The monitoring plan for the Resuspension Performance Standard is summarized in
Tables 1-2, 1-3 and 1-4 of the main document. This attachment describes the adequacy of
the sampling frequencies required as part of the routine monitoring programs, which are
derived using United States Environmental Protection Agency (USEP A)-defined methods
for assessing statistical uncertainty (USEPA, 2000). The analyses cover only routine
monitoring and the minimum levels of contingency monitoring as defined in the
Resuspension Standard. Additional monitoring related to the required engineering studies
at the Control Levels (as well as exceedance of the standard threshold) may be required,
depending on the anticipated cause of the exceedance. The design of these additional
monitoring programs may be developed during the remedial design period. Alternatively,
ad hoc monitoring plans may be developed by the design team during the actual dredging
operation in response to observations made at the time.
A particular limitation to the analysis presented in this attachment is that little
information on the variance of river conditions in response to dredging-related releases.
Little data exist on which to develop the estimate of variance. As a result, the variation of
baseline conditions was used as a means to estimate the variance for dredging operations.
These estimates for sampling requirements and the associated error rates will require
review once additional data become available during Phase 1.
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
2.0 Estimates of the Tolerable Error for the Monitoring Sampling
Frequency Using Decision Error Feasibility Trials (DEFT)
Software
The USEPA's guidance on data quality objectives (USEPA, 2000) was used in the
development of the monitoring program for the Phase 1 dredging operation. This
guidance describes a seven-step process for identification of the decision points and data
needs associated with the environmental problem to be addressed. With regard to PCB
releases via resuspension during the Phase 1 operation, there is a major concern to be
resolved: How can the USEPA verify that PCB concentrations in the Upper Hudson
River are in compliance with the resuspension criteria?
The focus of this analysis will be to design the appropriate sampling program,
particularly the optimal sampling frequency that must be implemented to address the
above-mentioned concern.
In the following discussion, the data quality objectives (DQO) process (USEPA QA-G4;
USEPA, 2000) is applied as outlined below:
1. State the Problem
2. Identify the Decision
3. Identify the Inputs to the Decision
4. Define the Boundaries of the Study
5. Develop a Decision Rule
6. Specify Tolerable Limits on Decision Errors
7. Optimize the Design for Obtaining Data
A separate discussion is provided for each question. A summary of the sampling
requirements is provided in Section 1 of the Resuspension Performance Standard.
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
3.0 Development of Data Quality Objectives
3.1 Statement of the Problem
The LSEPA needs to verify that water column concentrations of PCBs in the Upper
Hudson are below the Resuspension Standard criteria, thereby permitting unfettered
dredging operations. If PCB concentrations are not within acceptable levels, then
additional monitoring and possible modifications to the engineering operations may be
required.
The USEPA staff represents the decision makers who will consult with General Electric
Company (GE), the New York State Department of Environmental Conservation
(NYSDEC), water supply operators, local government representatives, and non-
government organizations.
The conceptual model is defined as follows:
PCB loads and concentrations within the Upper Hudson are currently derived
from sediment-based sources that contribute about 50 to 200 ng/L to the water
column under typical flow conditions. These concentrations constitute baseline
conditions. Dredging of contaminated sediments will add to this water column
burden to some degree. Anticipated load additions due to dredging are expected to
be less than 300 g/day (Evaluation Level threshold) under normal routine
dredging for a 6-year remediation program. This is especially true for Phase 1,
since the operation is planned at only half of the annual production rate
anticipated for Phase 2.
Although the mean daily Total PCB load increase due to dredging is expected to
be well below 300 g/day, instantaneous conditions may result in momentary
fluxes that are much higher. Consistent Total PCB loads higher than 300 g/day are
considered indicative of problems in the dredging operation and warrant further
study. Exceedance of the 300 g/day threshold does not constitute an immediate
risk to human or ecological health but rather will delay the recovery of the river if
allowed to continue for long periods of time. Similarly, exceedance of the 600
g/day action level does not represent an immediate risk to human or ecological
health, but, as is the case a 300 g/day load, an extended amount of time above this
action level will delay the river's recovery.
Total PCB concentrations in excess of 350 ng/L alone do not represent a risk to
downstream users so long as levels remain below the drinking water maximum
contaminant level (MCL) of 500 ng/L (total) PCBs. However, the proximity of
this level (350 ng/L) to the MCL warrants more careful scrutiny and closer
observation if 350 ng/L is exceeded due to the short transit time from the dredging
area to the nearest public water supply intakes (two to seven days).
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
Suspended solids data will provide an indication of increased PCB contamination
in the water column. Net far-field suspended solids concentrations must be below
12 mg/L to be at routine levels and below 24 mg/L to be at or below the
Evaluation Level. Net near-field suspended solids concentrations (as defined in
the Resuspension Standard) must be below 60 mg/L, 100 mg/L, or 700 mg/L,
depending on the location of the station relative to the dredge and the river section
in which dredging b occurring. The duration of the exceedances provides an
indication of the severity of the exceedance and the required response.
3.2 Identify the Decision
Depending on the magnitude of the dredging-related PCB load increase, the USEPA may
decide to do one or more of the following as described in Section 1 of this document:
Increase monitoring frequency;
Modify monitoring techniques;
Modify dredging operations;
Add additional engineering controls to the dredging operation; or
Suspend the dredging operation until the PCB release problem has been resolved
The primary question governing this decision is: Are water column concentrations in
compliance with the resuspension criteria? If water column concentrations are not in
compliance, required actions involve collection of additional samples to further define the
PCB loads if the requirements of the first decision statement are met, with further
increases in monitoring and the possibility or requirement of engineered modifications to
the operation, as described in the standard.
3.3 Identify the Inputs to the Decision
To determine net PCB loads due to dredging (i.e., the total load less the baseline), the
following data are needed:
Instantaneous and mean daily river flow at all monitoring locations
PCB concentrations at multiple monitoring locations, including the first far-field
station downstream of the dredging operation and extending to Waterford.
PCB concentration at a location upstream of the dredging operation (specifically
Rogers Island)
Suspended solids concentrations
Total organic carbon (TOC) on suspended solids
Dissolved organic carbon content (DOC; i.e., TOC on filtered water samples)
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
Historical concentrations of PCBs, suspended solids, TOC on suspended solids at
each of the main monitoring locations
The first six items listed above are used to characterize the actual conditions during
dredging. The seventh item is used to provide a basis for comparison to establish the net
load relative to the historical baseline conditions. The difference between baseline
conditions and conditions measured during dredging is the net increase in PCB
concentration due to dredging activities at each monitoring location. The product of the
mean daily flow and this concentration difference yields an estimate of the net load
increase for comparison against the load-based criteria. Suspended solids and PCB
concentration data will be used together to examine the usefulness of a suspended solids-
PCB correlation to estimate PCB levels based on suspended solids monitoring alone.
The methods for sample analysis include:
PCB congeners with a detection limit of 0.5 ng/L total PCBs. The effective
congener detection limit is roughly 0.05 ng/L Currently this can only be achieved
by one of the following: EPA's dual column GC/ECD method, Standard Method
1668A or GE's modified Green Bay Method.
Total Suspended Sediment with a detection limit of 0.1 mg/L, by Analytical
Method ASTM D3977-97, Standard Test Method for Determining Sediment
Concentration in Water Samples, or equivalent. No subsampling of a sampling
container is permitted.
Organic carbon on the suspended solids can be done via a Total Organic Carbon
method or by a combustion technique but must be sensitive down to 0.1% (1000
mg/kg) on the suspended solids.
Dissolved organic carbon method should have a detection limit of 0.5 mg/L, such
as ASTM Method D4839-03 [0.1 mg/L] or EPA 415.2 [.05 mg/L],
3.4 Define the Boundaries of the Study
The boundaries of the site are defined as the shorelines of the Hudson River, excluding its
tributaries, between the Fennimore Bridge at Hudson Falls and the Federal Dam at Troy.
The Fennimore Bridge is included as the upper boundary, rather than the northern end of
Rogers Island, because of the potential for PCB releases associated with the remediation
of the GE Hudson Falls facility that will be taking place at the same time or just prior to
the sediment remediation.
In recognition of the need to simplify monitoring, both project data needs and ease of
access will be considered when choosing monitoring locations. The following stations, all
of which are accessible by bridge, were selected based on access considerations:
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
Fennimore Bridge
Rogers Island
Schuylerville
Stillwater
Waterford
These locations also roughly divide the river into 10 to 15 mile segments, providing
sufficient resolution to identify potential PCB sources by location. The separation of
these locations also allows natural hydrodynamic processes to homogenize PCB
concentrations in the river, simplifying the sampling process.
Given that most of the dredging is scheduled for the Thompson Island (TI) Pool, an
additional monitoring location is identified at the TI Dam so as to better identify loads
originating in this reach.
Because dredging-related releases will depend on many factors related to dredge
operation, sediment type and location with in the river, the PCB load is expected to vary
significantly over time. Daily monitoring is considered a minimum basis for determining
compliance with the lowest (most stringent) secondary criterion of 300 g/day. When this
threshold is exceeded, a higher frequency of monitoring will used to document and
understand the sources of PCBs to the water column.
The loads released by dredging are expected to vary rapidly over time and thus will need
to be reviewed daily. Sampling when routine conditions are expected will measure the
daily variability. The weekly variability, as defined by a 7-day running mean calculated
daily, will be used to test compliance with the load-based criteria. This technique will
allow confirmation of compliance with the long-term load criterion while also collecting
data to demonstrate that more significant exceedances of PCB concentration criteria (e.g.,
exceeding 350 or 500 ng/L) have not occurred.
The transit time of water from the TI Pool to Waterford is expected to vary from two to
seven days, depending inversely on flow. As a result of the normal dispersion and settling
processes, the intensity of any short-term PCB release is expected to be diminished as the
river travels from TI Pool to Waterford. Thus, for a dredging operation in the TI Pool, the
discrete sample collected at TI Dam has not undergone the same level of integration as a
sample obtained at Waterford. Thus collecting samples along the Upper Hudson serves to
examine both short-term (one hour duration) and longer-term (one- to two-day duration)
PCB loads and PCB concentrations. Both measures are needed to assess the success of
the resuspension controls.
The sampling program must reflect the need to assess gradual increases in long-term
impacts, such as PCB mass transported downstream and the consideration of acute PCB
concentrations at downstream public water supplies. The long-term averages (7-day
period) and daily results are required to assess such long term impacts. To address the
protection of downstream water supplies, 24-hour turn-around times are needed for the
two monitoring stations downstream of, but closest to, the dredge operation. For Phase 1,
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
these are expected to be the TI Dam and Schuylerville stations. Based on the above
considerations, and those of the standard, the decision units are the loads as measured
weekly and the concentrations measured daily.
The results from the two far-field stations closest to the dredging operations provide
some indication of what the downstream PCB levels. However, due to the highly variable
nature of the PCB release process, samples must still be collected from locations farther
downstream and the concentrations confirmed to be in compliance with the standard.
These samples can have a longer turn around time, on the order of 7 days from collection
to result, since their role is primarily confirmational. These samples are necessary during
Phase 1 but may be dropped in Phase 2, depending on the success of the suspended solids
monitoring and the actual PCB loss rates.
3.5 Develop a Decision Rule
The decision rules are derived from the performance standard criteria described in
Volume 1 of the document and justified in Sections 2 and 3 of Volume 2 of the
document. The decision rule is designed to test compliance with the standard criteria.
The arithmetic mean is selected as the primary measure since it reflects an
integration of several measures and representative of the integrated PCB load over
the averaging period. Compliance with each of the resuspension criteria is the
primary focus of this DQO discussion.
3.6 Specify Tolerable Limits on Decision Errors
Current estimates of PCB release due to dredging, as developed in other attachments to
this document, indicate that PCB loads and concentrations are likely to fall below the
action level criteria during most of the operation. More specifically, the estimates of PCB
release indicate that when the PCB loads and concentrations are viewed on a daily or
weekly basis, momentary flux variations will average out so as to fall below the action
level criteria. Additionally, the threshold criteria developed for the decision rules do not
represent conditions immediately dangerous to human health or the environment. Based
on this, the null hypothesis for the decision rule is taken as the condition that the river is
in compliance (z'.e., the river flux or concentration of total PCBs is below the criteria
value). This approach also takes into consideration that daily monitoring will continue,
and that confirmation of any day's decision about dredging releases and water column
concentration will be obtained in the next sample taken.
USEPA's Decision Error Feasibility Trials Software (DEFT (USEPA, 2001)) was used to
develop the sampling requirements for this program. The results of this analysis are
presented in Table l.As defined in USEPA (2001):
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
A false acceptance decision error occurs when the sample data lead you to
decide that the baseline condition is probably true when it is really false.
A false rejection decision error occurs when the limited amount of sample
data lead you to decide that the baseline condition is probably false when
it is really true.
The gray region is a range of true parameter values within the alternative
condition near the Action Level where it is "too close to call."
False acceptances were minimized because it is the more serious error. In general,
decisions that were more critical, such as confirmation of exceedance of the
Resuspension Standard which requires the shut down of operations, or exceedance of the
Control Level which requires intense monitoring and implementation of engineering
evaluations and solutions, required a large number of samples and had greater certainty
than the less critical decisions. For the suspended solids measurements, it was clear that
the implementation of a continuous monitor capable of estimating suspended solids
concentrations would be needed to provide a reasonable amount of certainty in these
decisions. The low level of certainty is tolerable only because any decisions made as a
result of an exceedance of the suspended solids will be confirmed by measurements of
PCB concentrations in the impacted water column.
For PCB measurement-based decisions, a false acceptance rate of 5 percent or less was
sought, with lower rates sought when an incorrect decision would yield an unnecessary
halting of the operation or an engineering improvement. The rate of 5 percent was
selected as an acceptable error for the lower action level criteria, since exceedance of the
action level criteria only initially induces additional monitoring which will quickly
confirm the exceedance. This error rate reflects a balance between setting the monitoring
requirements as low as possible while still providing protection.
3.7 Optimize the Design for Obtaining Data: Results of the Analysis
The final sampling requirements for the standard were developed using DEFT (USEPA,
2001), a program to estimate sampling requirements based on a project-specific error
rate. Table 1 summarizes the analysis of the various criteria, acceptable gray region
around each criterion, the sampling frequency required by the resuspension standard, and
the false acceptance and false rejection levels. The table is organized by measurement
type (i.e., PCB and suspended solids). For all criteria except the confirmation of the 500
ng/L exceedance, the null hypothesis assumed that river conditions were in compliance.
Two important assumptions were made to develop the error rate values in the table. There
is no site-specific data on the expected variance of water column conditions related to
dredging. As a result, the extensive analysis of variance compiled in Attachment A was
used. A nominal coefficient of variance was assumed for PCBs and suspended solids
based on the variance observed under baseline conditions. For PCB measurements (both
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
Total PCBs and Tri+), the coefficient of variance is assumed to be 25 percent. For
suspended solids, the coefficient of variance was assumed to be 75 percent.
This section also includes a set of figures illustrating the statistical calculations used to
estimate the error rates. Figures 1 to 25 represent the calculations for each line in Table 1.
Table 1 shows that the higher level of sampling associated with the higher action levels
and the and Resuspension Standard yield low false error rates, reflecting the need to be
accurate before taking costly actions or improvements. In some instances, the false
rejection rate is fiirly high, indicating that additional sampling may be unnecessarily
triggered. However, this represents a protective approach from the perspective of
ensuring the safety of public water supplies. Additionally, the higher monitoring rates
will quickly confirm the need to remain at the action level thought to be exceeded.
Higher error rates were estimated in the transition from routine conditions to the
Evaluation and Control Levels, reflecting the relative low sampling rate required for
routine sampling. Also shown in the table is the one week confirmation result (i.e., the
error rate for the combination of one week of routine monitoring and one week at the
action level). In each instance, the false acceptance error was brought below 5 percent,
thereby confirming the need to sample at the higher rate or indicating that sampling at the
routine rate may be resumed.
The results for the monitoring requirements implemented after exceedance of the
standard demonstrate the need for the intensive sampling specified. In this instance the
river is assumed be in exceedance of the standard. Four additional discrete samples
(Figure 7) do not provide sufficient certainty given that the next day's decision will
involve the temporary halting of the dredging operations. However, by collecting hourly
composites, the power of the same four analyses is greatly improved and the 5 percent
false acceptance rate is attained.
Table 1 also presents the results for the long-term integrative samples. These samples will
serve to confirm the results of the daily routine monitoring, or indicate that more frequent
sampling is warranted. The results assume the automated collection of eight samples per
day over a one- to two-week period.
The results for suspended solids illustrate the need to use a continuous sampling system
such as a turbidity probe. In the lower portion of the table, results for the discrete
sampling program are compared with those that can be achieved with a continuous probe
taking a reading once every 15 minutes. In almost all cases, the continuous reading probe
provided more than an order of magnitude improvement in the expected error rate. Better
rates can be achieved using the continuous probes by simply taking data more frequently.
Note that this analysis does not consider any uncertainty introduced by use of a probe
over discrete samples. Nonetheless, given a semi-quantitative relationship between the
probe and the actual suspended solids levels, it is highly likely that the probes will
provide a substantial reduction in the expected error rates for suspended solids
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
monitoring, reducing unnecessary additional PCB sampling prompted by a false
indication.
Figures 26 through 28 show the Total PCB sampling requirements for the evaluation and
control levels to achieve 5 percent false acceptance and false rejection rates if automatic
samplers were used. Using the automated sampler, one composite sample with 24
aliquots (i.e., 1 aliquot per hour) is collected each day. At the evaluation level, to achieve
the false acceptance and false rejection rate of 5 percent, 2 composite samples with 24
aliquots of each sample are needed (Figure 26). This means that data from at least two
days are needed to be certain that the evaluation level is exceeded. Three composite
samples with 24 aliquots each sample are needed to be certain that that 600 g/day Total
PCB load action level is exceeded at the control level (Figure 27). For a concentration
exceedance at the control level, four composite samples with 24 aliquots each sample are
needed to achieve false acceptance and false rejection rates of 5 percent (Figure 28).
Table 2 summarizes the various criteria, the associated gray region, the sampling
frequency required by the resuspension standard, and the false acceptance and false
rejection levels when the automatic sampler is used. Figures 29 through 34 illustrate the
statistical calculations used to estimate the error rates for each line of Table 2. Using the
automatic sampler, the error rates for most of the sampling requirements are less than 1
percent. The highest error rate was about 2 percent for the false rejection of the sampling
requirement from evaluation to control level. However, this value is still below 5 percent
error rate. This analysis shows that the power of the sampling program for Total PCB
using automatic sampler is greatly improved.
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
4.0 References
USEPA, 2000. Guidance for the Data Quality Objectives Process EPA/600/R-96/055.
August 2000.
USEPA, 2001. Data Quality Objectives Decision Error Feasibility Trials Software
(DEFT) - USER'S GUIDE. EPA/240/B-01/007. September 2001.
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G-April 2004
-------
Tables
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment G - April 2004
-------
Table 1
Summary of Sampling Frequency Requirements and Expected Error Rates
Analysis Transition
Detail
Sampling Time Period
Action Level
Number of Samples1
Grey Region
Limit
False
Rejection
Error Limit -
a (%)
False
Acceptance
Error Limit -
P(%)
Figure
Number
Total PCB Sampling Requirements (25% CV)
Far Field
Routine to Evaluation Level
Routine to Control Level
Confirmation of the Control Level
Routine to Control Level
Confirmation of the Control Level
Routine to > 300 g/day
Routine to >600 g/day
Confirmation of > 600 g/day
Routine to > 350 ng/L
Confirmation of > 350 ng/L
1 week
1 week
1 week routine + 1 week
1 week
1 week routine + 1 week
300 g/day
600 g/day
600 g/day
350 ng/L
350 ng/L
7 (1 sample/day for 1 week)
7 (1 sample/day for 1 week)
28 (7 samples routine + 21 samples
control level)
7 (1 sample/day for 1 week)
28 (7 samples routine + 21 samples
control level)
400 g/day
700 g/day
700 g/day
400 ng/L
400 ng/L
7.5
25
5
27.5
10
5
15
4
20
5
1
2
3
4
5
Evaluation to Control Level
300 g/day to > 600 g/day
1 week evaluation + 1 week
600 g/day
35 (14 samples evaluation level + 21
samples control level)
700 g/day
4
2
6
Resuspension Standard Threshold
Confirmation of > 500 ng/L2
1 day routine + 1 day
500 ng/L
5 (1 sample routine + 4 samples
confirmation)
400 ng/L
15
30
7
Confirmation of > 500 ng/L (24 hours)2
1 day
500 ng/L
4 composites of 6 aliquots each
400 ng/L
5
7
8
Routine to Control Level
Continuous Total PCB 1-week or 2-week
deployment
1 week or 2 weeks
350 ng/L
2 composites of 56 aliquots each
400 ng/L
6.5
5
9
Suspended Solids Sampling Requirements (75% CV)
Far Field
Routine to Evaluation Level
Far-field - Baseline to > 12 mg/L
1 day (3 hrs for 24 hrs)
1 day (15 min for 24 hrs)
14 mg/L
14 mg/L
8 (discrete)
96 (continuous)
21 mg/L
21 mg/L
27.5
0.1
12.5
0.1
10
11
Routine to Control Level
Far-field - Baseline to > 24 mg/L
1 day (3 hrs for 24 hrs)
1 day (15 min for 24 hrs)
26 mg/L
26 mg/L
8 (discrete)
96 (continuous)
39 mg/L
39 mg/L
27.5
0.1
12.5
0.1
12
13
Evaluation to Control Level
Far-field -12 mg/L to > 24 mg/L
1 day evaluation + 1 day
1 day evaluation + 1 day
26 mg/L
26 mg/L
16 (discrete)
192 (continuous)
39 mg/L
39 mg/L
15
0.5
5
<0.5
14
15
Near Field
Routine to Control Level
Near Field - River Sections 1 and 3
Baseline to > 100 mg/L
6 hours (1 sample per 3 hours)
6 hours (1 sample per 15 min)
100 mg/L
100 mg/L
3 (discrete)
24 (continuous)
150 mg/L
150 mg/L
35
6.6
25
5
16
17
Routine to Control Level
Near Field - River Section 2
Baseline to > 60 mg/L
6 hours (1 sample per 3 hours)
6 hours (1 sample per 15 min)
60 mg/L
60 mg/L
3 (discrete)
24 (continuous)
90 mg/L
90 mg/L
35
6.6
25
5
18
19
Evaluation to Control Level
Near Field - River Sections 1 and 3
Baseline to > 100 mg/L
1 day (3 hrs for 15 hrs)
1 day (15 min for 15 hrs)
100 mg/L
100 mg/L
5 (discrete)
60 (continuous)
150 mg/L
150 mg/L
27.5
0.7
20
0.5
20
21
Evaluation to Control Level
Near Field - River Section 2
Baseline to > 60 mg/L
1 day (3 hrs for 15 hrs)
1 day (15 min for 15 hrs)
60 mg/L
60 mg/L
5 (discrete)
60 (continuous)
90 mg/L
90 mg/L
27.5
0.7
20
0.5
22
23
Routine to Evaluation Level
Near Field
Baseline to > 700 mg/L
3 hours (1 sample per 3 hours)
3 hours (1 sample per 5 min)
700 mg/L
700 mg/L
2 (discrete)
36 (continuous)
1000 mg/L
1000 mg/L
40
16.5
30
5
24
25
Note
1 Sampling frequency at the different action level can be found in Table 1-2 of Volume 1 of the document
2 Null hypothesis for the 500 ng/L assumed that river conditions were not in compliance, for all other action levels, the null hypohesis assumed that river conditions were in compliance. See text for discussions.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Table 2
Summary of Sampling Frequency Requirements and Expected Error Rates for Automatic Sampler
An alysis T ransition
Detail
Sampling Time Period
Action Level
Number of Samples
Grey Region
Limit
False
Rejection
Error Limit -
a (%)
False
Acceptance
Error Limit -
P(%)
Figure
Number
Total PCB Sampling Requirements (25% CV)
Far Field
Routine to Evaluation Level
Routine to > 300 g/day
1 week
300 g/day
7 composites of 24 aliquots each (1
sample/day for 1 week)
400 g/day
0.1
<0.1
29
Routine to Control Level
Routine to > 600 g/day
1 week
600 g/day
7 composites of 24 aliquots each (1
sample/day for 1 week)
700 g/day
0.5
0.1
30
Confirmation of the Control Level
Confirmation of > 600 g/day
1 week routine + 3 day
600 g/day
10 (7 samples routine + 3 samples
control level)
700 g/day
0.5
<0.5
31
Routine to Control Level
Routine to > 350 ng/L
1 week
350 ng/L
7 composites of 24 aliquots each (1
sample/day for 1 week)
400 ng/L
1
1
32
Confirmation of the Control Level
Confirmation of > 350 ng/L
1 week routine + 3 day
350 ng/L
10 (7 samples routine + 3 samples
control level)
400 ng/L
0.5
<0.5
33
Evaluation to Control Level
300 g/day to > 600 g/day
2 day evaluation + 3 day
600 g/day
5 (composite sampling every 1 hour, 1
sample/day)
700 g/day
2
1
34
Hudson River PCBs Superfiind Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figures
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment G - April 2004
-------
Figure 1
Routine to Evaluation Level
Action level of 300 g/day
FR
Ho: mean < 300
i 1 1 "Ti 1 1 1 1
70.0 140.0 210.0 280.0 350.0 420.0 490.0 560.0 630.0 700.0
ฆ0.050
FA
0.075
0.0
Simple
Act ion
Cost =
Sample
True Mean Concentration
Random Sampling Decision Error Limits
Level = 300.000 concentration prob(E) type
$2100.00 300.000 0.075 FR
Size = 7 400.000 o.oso fa
Figure 2
Routine to Control Level
Action Level of 600 g/day
CD
FR
Ho: mean < 600
ฆ0.150
FA
0.250
ฐ-ฐi 1 1 1 1 1 1 r
0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.01000.0
True Mean Concentration
Simple Random Sampling Decision Error Limits
Action Level = 600.000 concentration prob(E) type
Cost = $2100.00 600.000 0.250 FR
Sample Size = 7 700.000 0.150 fa
Note: Figures generated from DQO - DEFT using a coefficient of variation for all total
PCB cases of 25 percent.
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment G - April 2004
-------
Figure 3
Confirmation of the 600 g/day
Action Level of 600 g/day
CD
sz
4ป
O)
c
33
o
CD
"O
ฆ4
o
&
CO
_Q
O
%
CL
1.0-1
ฃZ
o
0.9 -
4l
o
0.8 -
CD
FR
Ho: mean < 60 0
T
T
ฆ0.040
FA
0.050
Simple
Act ion
Cost =
Sample
-1 1 1 r
100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.01000.0
True Mean Concentration
Decision Error Limits
concent ration prob(E)
Random Sampling
Level = 600.000
$8400.00
Size = 28
600.000
700.000
0.050
0.040
type
FR
FA
Figure 4
Routine to Control Level
Action Level of 350 ng/L
Estimated Performance Curve
CD
E J5
CD c
sz o
CD
O) sz
s s
"O 5r-
o ro
2r ฃ
=
jd to
m
JD C
O
-------
Figure 5
Confirmation of the 350 ng/L
Action Level of 350 ng/L
FR
Ho: mean < 350
J.
ฆ0.050
FA
0.100
0.0
1 1 1 1 I 1 "1 1 1
50.0 100.0 150.0 200.0 250.0 300.0 350.0 400.0 450.0 500.0
True Mean Concentration
Random Sampling Decision Error Limits
Level = 350.000 concentration prob(E) type
$8400.00 350.000 o.ioo fr
gj_ze 28 400.000 0.050 FA
Figure 6
Evaluation Level to Control Level
300 g/day to 600 g/day
Estimated Performance Curve
OJ CD
Z5 >
m
CD
o CG
CL
FR
Ho: mean < 600
FA
Simple Random Sampling
Action Level = 600.000
Cost = 510500.00
Sample Size = 35
300.0 400.0 500.0 600.0 100.0 800.0 300.0 1000.0
True Mean Concentration
Decision Error Limits
concentration proto (E) type
600.000 0.040 FR
700.000 0.020 FA
Note: Figures generated from DQO - DEFT using a coefficient of variation for all total
PCB cases of 25 percent.
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment G - April 2004
-------
Figure 7
Resuspension Threshold
Confirmation of 500 ng/L
Estimated Performance Curve
0) _ฃ=
2 ฉ
O ro
& ฃ
= cs
1q m
CD
_G ฃ=
O CO
CL
CD
1.0
0.9
0.8 ฆ
0.7 ฆ
0.6'
0.5
0.4 ฆ
0.3 ฆ
0.2'
0.1'
0.0 '
FA
Action Level
FR
Ho: mean > 500
o.ISO
0.300
Simple
Act ion
Cost =
Sample
o.o
Random
Level
$15 00.
Size =
100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 1000.0
True Mean Concentration
Sampling Decision Error Limits
= 500.000 concentration prob(E) type
00 400.000 0.300 FA
5 500.000 0.150 FR
Figure 8
Resuspension Threshold
Confirmation of 500 ng/L (24 hours; 4 samples of 6 aliquots)
a? ฎ
CD r-
Estimated Performance Curve
500
0.070
I i i i i i r
0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.01000.0
True Mean Concentration
site Sampling Decision Error Limits
n Level = 500.000 concentration prob(E) type
= $1220.00 400.000 0.070 fa
0 Sizs = 4 500.000 0.050 FR
composites of 6 aliquots)
Note: Figures generated from DQO - DEFT using a coefficient of variation for all total
PCB cases of 25 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 9
Routine to Control Level (350 ng/L, 2-week deployment) or
Evaluation Level to Control Level (350 ng/L, 1-week deployment)
Continuous total PCB sampling requirements
S3
U
CO
E R
Ho: mean < 350
ฆ0.050
FA
0.065
Compo
Actio
Cost
Sam pi
(
I I I I I I I I I
0.0 70.0 140.0 210.0 2B0.0 350.0 420.0 490.0 560.0 630.0 700.0
True Mean Concentration
Decision Error Limits
concentration prob(E) type
site Sampling
n Level = 350.000
$610.00
e Size = 2
composites of 56 aliqnots)
350.000
400.000
0.065
0.050
FR
FA
Note: Figures generated from DQO - DEFT using a coefficient of variation for all total
PCB cases of 25 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 10
Routine to Evaluation Level
(Far-field Baseline to >12 mg/L with discrete samples every 3 hrs for 24 hrs)
CD ฉ
g 5
S3
U)
5
S ฃ
Tj i
h! ฎ
O "K
^ ฃ
= 01
JD 07
CO
JD C
O CO
ฃL
CD
Estimated Performance Curve
ion Level
0. 125
FA
0.275
0.0
10.0
90.0 100.0
Simple
Act ion
Cost =
Sample
1 1 1 1 1 r
20.0 30.0 40.0 50.0 60.0 70.0 80.0
True Mean Concentration
Random Sampling Decision Error Limits
Level = 14.000 concentration prob(E)
$160. 00 14.000 0.275
= 0 21.000 0.125
type
FR
Fit
Figure 11
Routine to Evaluation Level
(Far-field baseline to >12 mg/L with continuous sampling every 15 min for 24 hrs)
ฎ ฉ
2 ฉ
O
CO
ฆ*-' ฉ
O) XI
a
"o
TJ
ฉ
o to
ฉ
O)
LTl
2?
-Q
_Q ฃ=
O CO
CL
ฉ
Estimated Performance Curve
1.0
0.9 -
0.8
0.7
0.6
0.5
0.4
0.2
0.1
0.0
ฆ0.001
ilction Level
FR
mean <
FA
o.ooi
Simple
Act ion
Cost =
Sample
i 1 1 1 1 r
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0
True Mean Concentration
Random Sampling Decision Error Limits
Level = 14.000 concentration prob(E) type
$1920.00 14.000 0.001 FR
Size = 96 21.000 o.ooi fa
Note: The analysis is based on a baseline of Schuylerville conditions (Average TSS
concentration from May-Nov of 2.4 mg/L with an average standard deviation from May-
Nov of 1.87 mg/L) and coefficient of variation equal to 75 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 12
Routine to Control Level
(Far-field Baseline to >24 mg/L with discrete samples every 3 hrs for 24 hrs)
FR
o; mean < 26
0. 125
FA
0.275
90.0 100.0
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
True Mean Concentration
Simple Random Sampling Decision Error Limits
Action Level = 26.000 concentration prob(E)
Cost = $160. 00 26.000 0.275
Sample Size = 8 39.000 0.125
type
FR
FA
Figure 13
Routine to Control Level
(Far-field baseline to >24 mg/L with continuous sampling every 15 min for 24 hrs)
FR
o; mean < 26
T
ฆ0.001
FA
0.0
Simple
Act ion
Cost =
Sample
1 1 1 1 1
10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0
True Mean Concentration
Random Sampling Decision Error Limits
Level = 26.000 concentration prob(E)
$1920. 00 26.000 o.ooi
Size = 96 39.000 o.ooi
0.001
type
FR
FA
Note: The analysis is based on a baseline of Schuylerville conditions (Average TSS
concentration from May-Nov of 2.4 mg/L with an average standard deviation from May-
Nov of 1.87 mg/L) and coefficient of variation equal to 75 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 14
Evaluation to Control Level
(Far-field Evaluation to Control Level with discrete samples every 3 hours for 24
hours)
Estimated Performance Curve
True Mean Concentration
Decision Error Limits
concentration prob(E) type
26,000 o.xso m.
29,000 0.060 F&
Simple Random Sampling
Action Level = 2 6.000
Cost = $320.00
Sample Size = 16
Figure 15
Evaluation to Control Level
(Far-field Evaluation to Control Level with continuous sampling every 15 min for 24
hours)
1.0 -
0.9-
0.S-
0.7-
0.6-
Estimated Performance Curve
Act? ion Level >
FR
o: mean <26
FA
30.0
40.0
30.0
ฃ0.0
70.0
80.0
30.0
100.0
Simple Random Sampling
Action Level - 2 6.000
Cost ~ $3840,00
Sample Size = 192
True Mean Concentration
Decision Error Limits
concentration prob (E) type
ฃ6.000 0.005 FR
39.000 0.000 Fh
Note: The analysis is based on a baseline of Schuylerville conditions (Average TSS
concentration from May-Nov of 2.4 mg/L with an average standard deviation from May-
Nov of 1.87 mg/L) and coefficient of variation equal to 75 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 16
Routine to Control Level Near-field River Sections 1 and 3
(baseline to >100 mg/L with discrete samples every 3 hours for 6 hours)
Estimated Performance Curve
Action Level
0.250
FA
0.350
1 1 1 1 1 1 1 1
100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.01000.0
True Mean Concentration
Simple Random Sampling Decision Error Limits
Action Level = 100.000 concentration prob(E) type
Cost = $60 . 00 100.000 0.350 FR
Sample Size = 3 iso.ooo 0.250 fa
Figure 17
Routine to Control Level Near-field River Sections 1 and 3
(baseline to >100 mg/L with continuous sampling every 15 min for 6 hrs)
Estimated Performance Curve
0.050
Action Level
FA
0.066
100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.01000.0
True Mean Concentration
Simple Random Sampling Decision Error Limits
Action Level = 100.000 concentration prob(E) type
Cost = S480.00 100.000 0.066 FR
Sample Size = 24 iso.ooo o.oso fa
Note: The analysis is based on a coefficient of variation equal to 75 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 18
Routine to Control Level Near-field River Section 2
(baseline to >60 mg/L with discrete samples every 3 hours for 6 hours)
=4an
0.2
0. 1
Action Level
ฆ0.250
R
FA
60
0.350
0.0
A
CL
F
0.0
T rue
Simple Random Sampling
Action Level = 60.000
Cost = $60.00
Sample Size = 3
1 I I I I I I I T
100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.01000.0
Mean Concentration
Decision Error Limits
concentration prob(E) type
60.000 0.350 FR.
90.000 0.250 FA
Figure 19
Routine to Control Level Near-field River Section 2
(baseline to >60 mg/L with continuous sampling every 15 min for 6 hrs)
CD
CD
TJ
O
4<
U
CD
>
CO
-------
Figure 20
Evaluation to Control Level Near-field River Sections 1 and 3
(baseline to >100 mg/L with discrete samples every 3 hours for 15 hours)
Estimated Performance Curve
CD
1.0 -
cz
0.9 -
o
41
i >
0.8 -
100 mg/L with continuous sampling every 15 min for 15 hrs)
Estimated Pertormance Curve
o ro
2
O)
tn
1.0
0.9
0.8
0.7
0.6
0.5
0.4
h^:321
0.
0.1-
0.0 ฆ
ฆ0.005
01 Action Level
FR
FA
I I I
600.0 700.0 800.0 900.0 1000.0
Simpl
Actio
Cost
Sampl
1 I I
0.0 100.0 200.0 300.0 400.0 500.0
True Mean Concentration
e Random Sampling Decision Error Limits
n Level = 100.000 concentration prob(E)
$12 00. 0 0 ioo.ooo 0.007
Size = 60 iso.ooo o.oos
0.007
type
FR
FA
Note: The analysis is based on a coefficient of variation equal to 75 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 22
Evaluation to Control Level Near-field River Section 2
(baseline to >60 mg/L with discrete samples every 3 hours for 15 hours)
>
o>
1.0 -
c:
0.9
o
o
0.8 -
Q>
0.7
4ซ
0.6-
ฃZ
sz
0.5 i'
i_
0.4 -
CD
Hob.^
CO
0.2 -
tn
0.1-
i
kU
0.200
FA
60
0.275
-
0.0
Random
Level
$100.0
Size =
r i r
100.0 200.0 300.0 400.0
True Mean
Sampling
= 60.000
t i i r
500.0 600.0 700.0 800.0 900.0 1000.0
Concentration
Decision Error Limits
concentration prob(E)
60.000
90.000
0.275
0.200
type
FR
FA
Figure 23
Evaluation to Control Level Near-field River Section 2
(baseline to >60 mg/L with continuous sampling every 15 min for 15 hrs)
Estimated Performance Curve
ฉ ฉ
ฉ ฃZ
ฃ
4-' ฉ
O) sz
sz
^ ง
2 ฉ
o
-Q
0.005
Action Level
FA
60
1 I r
100.0 200.0 300.0 400.0
True Mean
Sampling
= 60.000
00
60
1 I I l
500.0 600.0 700.0 800.0 900.0 1000.0
Concentration
Decision Error Limits
concentration prob(E)
0.007
60.000
90.000
0.007
0.005
type
FR
FA
Note: The analysis is based on a coefficient of variation equal to 75 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 24
Routine to Evaluation Level
(Near-field baseline to >700 mg/L with discrete samples every 3 hrs for 3 hrs)
CD <1>
g a
Estimated Performance Curve
0.300
0.400
Simple
Act ion
Cost =
Sample
i 1 r
0.0 150.0 300.0 450
True
Random Sampling
Level = 7 0 0.000
S 4 0.00
Size = 2
1 I 1 I 1 1 1
0 600.0 750.0 900.01050.01200.01350.01500.0
Mean Concentration
Decision Error Limits
concentration prob(E) type
700.000 0.400 FR
1000.000 0.300 FJ
Figure 25
Routine to Evaluation Level
(Near-field baseline to >700 mg/L with continuous sampling every 15 min for 3 hrs)
CD ฎ
CD s-
& 9.
41 ฆ*1
11
CD +~
"O *-
FR
Ho: mean < 75 0
i 1 r
0.0 150.0 300.0 450,
True
Simple Random Sampling
Action Level = 750,000
Cost = $72 0.0 0
Sample Size = 3 6
ฆ0.050
FA
0. 165
r 1i 1 1 1
0 600.0 750.0 900.01050.01200.01350.01500.0
Mean Concentration
Decision Error Limits
concentration prob(E) type
750.000 0.165 FR
1000.000 0.050 FJ
Note: The analysis is based on a coefficient of variation equal to 75 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 26
Automatic Sampler at the Evaluation Level (300 g/day)
(1 sample per hour for 24 hours)
Estimated Performance Curve
Q
ฃ ฃ
Actjioxn Level -->
FR
Ho: mean < 3 00
FA
Composite Sampling
Action Level = 300.000
Cost = 3610.00
Sample Size = 2
(2 compos it=es of ฃ4 al i^uots)
200.0 400.0 500.0 600.0 700.0 800.0 900.0 1000.0
True Mean Concentration
Decision Error Limits
concentration probiE) type
300.000 0.050 FT!
400.000 0.050 Th
Figure 27
Automatic Sampler at the Control Level (600 g/day)
(1 sample per hour for 24 hours)
Estimated Performance Curve
Act; ion Level
FR
Ho: mean < 600
FA
-+-
Composite Sampling
Action Level = 600.000
Cost = $915.00
Sample Size = 3
(3 composites of 24 aliq^iots)
True Mean Concent ration
Decision Error Limits
concentration proto (E) type
600.000
700.000
0.050
0.050
Note: Figures generated from DQO - DEFT using a coefficient of variation for all total
PCB cases of 25 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 28
Automatic Sampler at the Control Level (350 ng/L)
(1 sample per hour for 24 hours)
Estimated Performance Curve
1.0
0.9 ฆ
o.s
0.7
0.6
0.5
0.4'
0.3
0.2
0.1-
0.0
Action Ltvซl
FR
Ho: mean < 350
FA
150.0 200.0 250.0 300.0 350.0 400.0
1 1
450.0 500.0
Composite Sampling
Action Level = 350.000
Cost = $ 1220.00
Sample Size = 4
(3 compos itr-4 5 of ฃ4 aliquots)
True Mean Concentration
Decision Error Limits
concentration prob(E) type
3-50.000 0.0-50 FR
400.000 0.0-50 FA
Figure 29
Routine to Evaluation Level with Automatic Sampler
Action level of 300 g/day
Estimated Performance Curve
CD QJ
js s
*- CD
ZSjJ=
0.8
0.7 ฆ
0.5
O.S
0.4-
0.3
0.2
0.1-
0.0
Action. Level -->
FR
Ho: mean < 300
FA
-h
Composite Sampling
Action Level = 300,000
Cost = $2135.00
Sample Size = 7
(7 composites oฃ ฃ4 a.1 i
-------
Figure 30
Routine to Control Level with Automatic Sampler
Action level of 600 g/day
Estimated Performance Curve
OJ CD
3 >
ฆ!=
o> c
1.0 -
0.9
0.8
0.7 ฆ
o.e
0.5 ฆ
0.4-
0.3
0.Z -
0.1-
0.0
Composite Sampling
Action Level = 600.000
Cost = $2135.00
Sample Size = 7
(7 compos of 24 al iqu.ot?5)
TS ซ-
*
O a;
O 05
ฃ S
Act?ioi% Level
FR
Ho: mean < 600
FA
300.0 400.0 500.0 SOO.O 700.0 800.0 SOO.O 1000.0
True Mean Concentration
Decision Error Limits
concentration prob(E) type
soo.ooo o.ooi m
100.000 0.001 FA
Figure 31
Confirmation of the 600 g/day with Automatic Sampler
Action level of 600 g/day
Estimated Performance Curve
i.o
o.s
0.8 -
0.7 -
0.6
0.5 ฆ
FR
Ho: mean < 600
FA
Composite Sampling
Action Level = 600,000
Cost = ^3050.00
Sample Size = 10
(10 compos itซ5 of 24 al iquotjs)
300.0 400.0 500.0 600.0 700.0 800.0 900.0 1000.0
True Mean Concentration
Decision Error Limits
concentration proto (E) type
500.000 0.005 FP.
700.000 0.000 FA
Note: Figures generated from DQO - DEFT using a coefficient of variation for all total
PCB cases of 25 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 32
Routine to Control Level with Automatic Sampler
Action level of 350 ng/L
Estimated Performance Curve
Act; ion Level
FR
Ho: mean < 350
FA
250.0 300.0
50.0
400.0 450.0
500.0
Composite Sampling
Action Level = 350.000
Cost = $2135.00
Sample Size = 7
True Mean Concentration
Decision Error Limits
concentration prob (E) type
350.000 o.oio m
400.000 0.010 FA
( 7 compos it:*
of ฃ4 alicfuofcs)
Figure 33
Confirmation of the 350 ng/L with Automatic Sampler
Action level of 350 ng/L
Estimated Performance Curve
il> CD
3 >
O 05
ฃ S
1.0 -
0.9 ฆ
0.8 ฆ
0.7 -
0.6
0.5
0.4-
0.3 ฆ
0.2 ฆ
0.1-
0.0
Act; ion Level -->
FR
Ho: mean < 350
FA
Composite Sampling
Action Level = 350.000
Cost = $3050.00
Sample Size = 10
(10 compos it? e s of ฃ4 aliquot: 5)
150.0 200.0 250.0 300.0 350.0 400.0 450.0 50
True Mean Concentration
Decision Error Limits
concentration prob (E)
350.000 0.005
400.000 0.000
type
FR
FA
Note: Figures generated from DQO - DEFT using a coefficient of variation for all total
PCB cases of 25 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Figure 34
Evaluation to Control Level with Automatic Sampler
Action Level 600 g/day
Estimated Performance Curve
1.0 -
0.9
o.s
0.7
0.6 -
0.5 -
0.4
o.s
0.2
0.1
0.0
Act-ion. Level -->
FR
Ho: mean < 600
FA
Composite Sampling
Action Level = 600.000
Cost = $1525.00
Sample Size = 5
(J coivipos ites of 24 al iqucCs)
SOO.O 400.0 500.0 500.0 700.0 800.0 300.0 1000.0
True Mean Concentration
Decision Error Limits
concentration probjE) type
COO.OOO 0,020 FR
700.000 0.010 FA
Note: Figures generated from DQO - DEFT using a coefficient of variation for all total
PCB cases of 25 percent.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment G - April 2004
-------
Attachment H
Estimated Cost and Feasibility of the Phase 1 Monitoring Program
Table of Contents
1.0 Abstract 1
2.0 Introduction 2
3.0 Phase 1 Compliance Monitoring Cost Estimate 4
3.1 Labor Costs - Level of Effort (LOE) 4
3.2 Routine Monitoring with Automated Suspended Solids Collection 4
3.2.1 Far-Field (Including Baker's Falls) 4
3.2.2 Near-Field 5
3.2.3 Routine Monitoring LOE Summary 5
3.3 Non-Routine Monitoring with Automated Suspended Solids Collection 5
3.3.1 Far-Field (Including Baker's Falls) 5
3.3.2 Near-Field 6
3.3.3 Non-Routine Monitoring LOE Summary 6
4.0 Cost Parameters 7
5.0 Special Study 8
5.1 Near-field PCB Release Mechanism (Dissolved vs. Particulate) 8
5.2 Development of a Semi-Quantitative Relationship between TSS and a Surrogate
Real-Time Measurement for the Near-field and Far-field Stations
(Bench Scale) 9
5.3 Non-Target, Downstream Area Contamination 10
6.0 Reasonable Estimate of Monitoring Program Cost 11
7.0 Feasibility and Other Considerations 13
8.0 Conclusions 15
9.0 References 16
LIST OF TABLES
Table 1 Sampling Cost on a Weekly Basis - Lower River Far-Field Stations
Table 2 Sampling Cost on a Weekly Basis - Upper River Far-Field Stations
Table 3 Sampling Cost on a Weekly Basis - Upper River Near-Field Stations
Table 4 Labor Cost on a Daily Basis - Far-Field Stations
Table 5 Labor Cost on a Daily Basis - Near-Field Stations
Table 6 Near-Field Total PCB Concentration Special Studies
Table 7 Summary of Labor and Lab Analytical Costs by Action Level
Table 8 Reasonable Estimate of Phase 1 Season Monitoring Plan Costs
Hudson River PCBs Superfund Site
Engineering Performance Standards
i
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
Attachment H
Estimated Cost and Feasibility of the Phase 1 Monitoring Program
1.0 Abstract
Cost estimates for the Phase 1 monitoring program were calculated assuming that the
major costs for the monitoring program are the labor costs to collect the samples and the
analytical costs. On this basis, the estimated cost of the Phase 1 monitoring program is
approximately $3,000,000. The estimated cost of the Phase 1 monitoring program cannot
be used as a basis for estimating the monitoring costs for the remainder of the
remediation. The Phase 1 monitoring program is designed to measure compliance with
the standard and to evaluate and refine the implementation of the standard. The sampling
efforts for the second objective are designated as "special studies." The results of the
monitoring in Phase 1 will determine the extent to which the Phase 1 monitoring program
requirements can be reduced (after the completion of Phase 1) and still measure
compliance with the resuspension criteria with an acceptable degree of certainty.
Hudson River PCBs Superfund Site
Engineering Performance Standards
1
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
2.0 Introduction
A number of different sampling and data collection events of which the Phase 1
monitoring program is included, will occur as part of the remediation of the Hudson
River PCBs Site. Components of the Phase 1 monitoring program include various water
column sampling and analyses to assess different techniques and measurement types for
monitoring and verifying compliance with the Resuspension Performance Standard; and
also to generate additional data to improve understanding of the sediment and
contaminant transport processes which may occur during the dredging program. Ongoing
monitoring during dredging operations subsequent to Phase 1 (Phase 2 monitoring) will
include monitoring conducted from the second year of the dredging program through its
completion. It is anticipated that the Phase 2 monitoring program will not be as intensive
as the Phase 1 program, as it is expected that data obtained during Phase 1 will enable
either the number of samples, or the analytical parameters, to be reduced while still
ensuring compliance with the resuspension criteria.
In addition to compliance monitoring, the Phase 1 monitoring program includes five
special studies for the resuspension standard. These are as follows:
Near-field PCB Release Mechanism (Dissolved vs. Particulate)
Development of a Semi-Quantitative Relationship between TSS and a
Surrogate Real-Time Measurement for the Near-field and Far-field
Stations (Bench Scale)
Development of a Semi-Quantitative Relationship between TSS and a
Surrogate Real-Time Measurement for the Near-field and Far-field
Stations (Full Scale)
Non-Target, Downstream Area Contamination
Phase 2 Monitoring Plan
These studies are intended to:
Determine the Total PCB water column concentrations and the nature
of contaminant release from the remedial operations (dissolved or
suspended phase release).
Determine and maintain a semi-quantitative relationship between TSS
and a real-time surrogate measurement.
Determine the extent of downstream contamination in the non-target
areas.
Establish alternate strategies to more efficiently handle the
requirements of the monitoring program.
Costs for these special studies are also provided where sufficient scope for the study is
available.
Hudson River PCBs Superfund Site
Engineering Performance Standards
2
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
The estimated costs of the Phase 1 program cannot be used to project the monitoring
costs for the rest of the remedial program, since it is likely that the Phase 1 program is be
more sample- and analytical-intensive than the Phase 2 monitoring program will be.
The cost estimate provided in this analysis focuses on the two main elements of the
program: labor and laboratory analytical cost. The cost estimate for the Phase 1
monitoring program is based on specific scenarios for implementing the monitoring
program, which are described in detail below. Standard laboratory rates are used to
estimate the analytical costs; however, it is likely that lower rates can be negotiated for
this program (due to the large quantity of analyses being performed). The final cost of the
Phase 1 monitoring program will also be dependent on the degree to which the operations
are in compliance with the resuspension criteria.
Alternate strategies may be developed to more efficiently handle the requirements of the
monitoring program. Other modifications to the monitoring program, which reduce the
costs of the program, will be acceptable, as long as all data quality objectives are met and
the modification is not so substantial as to cause the resuspension criteria to be
reevaluated. The standard requires that a special study establishing any proposed alternate
strategies for sampling be demonstrated concurrently with the Phase 1 program.
Hudson River PCBs Superfund Site
Engineering Performance Standards
3
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
3.0 Phase 1 Compliance Monitoring Cost Estimate
It is assumed that the primary costs for the Phase 1 monitoring program will be labor
costs associated with the sample collection and laboratory analytical costs. It is also
assumed that the quality assurance/quality control requirements will be limited due to the
quick turnaround requirements. Estimated costs for these elements for the monitoring
program described in the Resuspension Performance Standard were developed and are
described below. The labor costs are a function of two variables: the level of effort (i.e.,
the personnel-hours required to collect the samples), and the labor rates (dollars per
hour). Similarly, the analytical costs are a function of the number of analyses of each type
performed (e.g., PCB analysis, TSS, total organic carbon), and the unit cost for each of
these analyses.
The calculated cost estimates for the Phase 1 monitoring presented assumes that two field
laboratories will be established to perform the total suspended solids (TSS) analyses. As
the facilities (a mobile office trailer) and equipment (scale, oven, filters, and glassware)
are relatively simple and inexpensive, costs for the field laboratories (which will likely be
less than $10,000 for each) are not included in this estimate. Costs for the technicians to
perform the analyses are not included in this estimate; however, the costs for the TSS
analysis are addressed as a laboratory analytical cost (based on the cost of an off-site
laboratory performing the TSS analyses). The estimated samples required and the
laboratory analytical costs for routine and non-routine monitoring are provided in Tables
1 through 3.
In the discussion below, a number of the sampling activities are discussed relative to the
'operations' which are occurring at the time. In this context, 'operations' means any
remedial activities that involve sediment disturbance. These activities will be primarily
the dredging activites, but may also include other activities such as debris removal and
installation or removal of containment other than silt curtains.
3.1 Labor Costs - Level of Effort (LOE)
The level of effort for both the routine monitoring and non-routine monitoring efforts are
presented below. Each (routine and non-routine) is further subdivided into the LOE
estimate for near-field and far-field sample collection.
3.2 Routine Monitoring with Automated Suspended Solids Collection
3.2.1 Far-Field (Including Baker's Falls)
If the 1 suspended sample per every 3 hours is collected by mechanical means (i.e., by an
ISCO sampler), there is a significant reduction in LOE requirements of the Ft Edward, TI
Dam, Schuylerville, Stillwater and Waterford stations, for all action levels. Using
Hudson River PCBs Superfund Site
Engineering Performance Standards
4
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
automated samplers will not change the LOE requirements of the remaining stations.
Under these conditions, the field crews would manually collect the whole-water PCB,
DOC and grab suspended solids samples. The field crew would also be responsible for
picking up the automated suspended solids samples, replenishing sample vials in the
ISCO devices and delivering the samples to the field laboratories. One field crew could
sample multiple stations during each shift, with a reasonable breakdown of stations being
Ft Edward/TI Dam and Schuylerville/Stillwater/ Waterford.
The LOE breakdown for routine monitoring is shown in Table 4.
It may be possible for 1 crew to collect the samples at all 5 stations, but when considering
possible problems related to collection at the TI Dam, a more conservative estimate is
more appropriate.
3.2.2 Near-Field
One crew should be able to handle up to five operations of near field sampling. Above
that, a second crew will be required. Each crew will consist of two samplers and one
boat operator. The crew will collect the samples, fill out required paperwork and
transport the samples to the field labs described above.
The LOE breakdown (for five operations) for routine monitoring is shown in Table 5.
The major assumption of this estimate is that the dredging operations are within close
proximity to one another (i.e., all are within the same pool). Additional personnel will be
required if operations are being conducted in two or more pools.
3.2.3 Routine Monitoring LOE Summary
Based on the near-field and far-field estimates and the assumptions listed above, the LOE
for routine monitoring is between 10 and 16 people per day (the variability is contingent
on specifics of operations) to collect samples, fill out paperwork and transport the
samples to one of two field labs for the duration of the program.
3.3 Non-Routine Monitoring with Automated Suspended Solids
Collection
3.3.1 Far-Field (Including Baker's Falls)
For non-routine sampling with automated suspended solids collection, the station
assignments of the field crews must change as a result of the additional sampling
requirements and consideration of river mile. The realignment would be Ft Edward/TI
Dam/Schuylerville and Still water/Waterford. The Sampling Level dictates the number of
additional crews required. Under most instances, 1 additional crew is added for each
Hudson River PCBs Superfund Site
Engineering Performance Standards
5
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
sampling event. For example, a second crew is added during Evaluation Level monitoring
and a third crew is added for Concern Level evaluation. For Threshold Level monitoring,
a fourth crew is added to collect the 4 required samples.
The LOE breakdown is for non-routine monitoring is shown in Table 4.
Additional reductions in LOE requirements for both routine and non-routine monitoring
may be possible if technicians at the field laboratories are made responsible for picking
up automated suspended solids samples from the ISCO samplers.
3.3.2 Near-Field
The hourly suspended solids sample collection requirement of the non-routine monitoring
would require one crew per two operations, with an additional person added to each crew
to shuttle samples to the field laboratories.
The LOE breakdown (assuming six operations) is for non-routine monitoring is shown in
Table 5.
With two or fewer operations, only one additional person (relative to routine monitoring)
per shift would be required; five additional people per shift would be required for three or
four operations; nine people per shift for five or six operations, and so on. The maximum
number of additional people would be 17 people per shift at a maximum of 10 operations.
The major assumption of this estimate is that dock space can be accessed nearby the
operations so that the time required to get the samples to shore for transport to the labs is
not a significant factor. As with Routine Monitoring, the estimate assumes that operations
are being conducted in the same pool, and the LOE is estimated only for sample
collection, documentation and transport to the field labs.
A concern of the non-routine sampling is the immediate need for the additional personnel
if the surrogate relationship is not in compliance. The range of people required for non-
routine sampling (personnel in addition to the full-time staff doing routine monitoring) is
significant, starting at 9 people (Evaluation Level, one or two operations) up to a
maximum of 33 additional personnel (Control Level, 10 operations). At the maximum
level, the size of the field crew essentially doubles. From a resource management
standpoint, maintaining a pool of 30 qualified and trained individuals to be ready to
sample with less than 12 hours notice would be difficult, at best.
3.3.3 Non-Routine Monitoring LOE Summary
Based on the near-field and far-field estimates and the assumptions listed above, the LOE
for non-routine monitoring is between 14 and 56 people per day (the variability is
contingent on specifics of operations) to collect samples, fill out paperwork and transport
the samples to one of two field labs for the duration of the program.
Hudson River PCBs Superfund Site
Engineering Performance Standards
6
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
4.0 Cost Parameters
Labor Rates
It is assumed that the average cost for sampling technicians during an 8-hour shift will be
$416 ($52/hour loaded rate, based on a $20/hour direct rate and an overhead factor of
1.6).
Laboratory Analysis - Estimated Quantities
The estimated laboratory analysis quantities for far-field (Upper Hudson River and
Lower Hudson River) and near-field laboratory analyses are provided in Tables 1 through
3.
Laboratory Analysis - Unit Costs
The estimated unit costs for laboratory analyses are listed below.
PCB Congeners (standard turnaround time)
$
300
24-hour Turnaround Time
$
600
72-hour Turnaround Time
$
525
Suspended Solids 3-hour Turnaround Time
$
20
Dissolved Organic Carbon
$
35
Suspended Organic Carbon
$
60
The PCB congener rates above assume a 100 percent surcharge for 24-hour turnaround
time, and a 75 percent surcharge for 72-hour turnaround.
Hudson River PCBs Superfund Site
Engineering Performance Standards
7
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
5.0 Special Studies
The monitoring programs for the resuspension and residual standards are organized to
separate sampling necessary to measure compliance with the standard from sampling
efforts needed to evaluate and refine the implementation of the standard. This has been
accomplished by designating the second category of sampling efforts as "special studies."
The special studies will be conducted for limited periods of time to gather information for
specific conditions that may be encountered during the remediation or to develop an
alternate strategy for monitoring. Specific conditions may include different dredge types,
contaminant concentration ranges, and varying sediment textures. Each of these studies is
integral to the Phase 1 evaluation, the development of Phase 2, and is also tied to
compliance issues.
There are a total of five special studies for the resuspension standard. These are as
follows:
Near-field PCB Release Mechanism (Dissolved vs. Particulate)
Development of a Semi-Quantitative Relationship between TSS and a
Surrogate Real-Time Measurement for the Near-field and Far-field
Stations (Bench Scale)
Development of a Semi-Quantitative Relationship between TSS and a
Surrogate Real-Time Measurement for the Near-field and Far-field
Stations (Full Scale)
Non-Target, Downstream Area Contamination
Phase 2 Monitoring Plan
Costs for the near-field PCB, semi-quantitative relationship (bench scale) and non-target
downstream area contaminant special studies are provided. The full scale semi-
quantitative relationship is included in the cost of compliance monitoring. No cost
estimates can be calculated for the Phase 2 monitoring plan special studies because the s
scope of the study has not been defined.
5.1 Near-field PCB Release Mechanism (Dissolved vs. Particulate)
The special study to characterize near field PCBs will consist of collecting samples in the
vicinity of a remedial operation once a day for approximately seven days. This would be
one event. It is estimated that approximately five events would be needed to characterize
the different conditions during remediation. The samples would be collected from
upstream, in two transects across the plume downstream from the dredge and within the
containment, if present. The samples will be depth integrated. There will be five sample
locations along each transect. Most of the samples will be collected for whole water
analysis, but a subset will be filtered and both the suspended and dissolved phase sent for
analysis. An acoustic sensor will be used to define the extent of the plume
Hudson River PCBs Superfund Site
Engineering Performance Standards
8
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
It is assumed that two boats with a crew of three technicians will be required each day of
sampling at a single location for a full day. One boat will be responsible for defining the
extent of the plume and identifying the sampling locations while the other boat collects
the samples.
Thus the LOE breakdown (for the 7-day operations) is:
2 crew x 3 people x 1 shifts per day = 6 people per day for the duration of the program.
The estimated costs are provided in Table 6.
5.2 Development of a Semi-Quantitative Relationship between TSS
and a Surrogate Real-Time Measurement for the Near-field and
Far-field Stations (Bench Scale)
To determine an initial relationship between TSS and a surrogate (turbidity or laser
particle analysis) it is proposed that three types of sediment (silty, fine sand and medium
sand) be collected for detailed analysis. For each sediment type one bucket full of
sediment will be required.
For labor costs, assume that one field crew of 2 people can collect and handle the 3
buckets of sediment in an 8-hour day. Thus:
1 crew x 2 people x 1 day = 2 man days, to collect the material.
Therefore the labor costs are approximately $832.
To conduct the bench study from this material, the following cost estimate is based on the
USACE Long Tube Settling Test (LTST) and the batch test as described in a paper by
Earhart (Earhart, 1984).
As per the USACE methodology, the LTST takes a full 15 days to determine
compression settling results, however, the test could be run for just a few days to
determine the turbidity-TSS correlations only.
Assuming three sediment samples are tested, the costs are as follows:
$4000 Column construction (one LTST column as per EM 111-2-5027)
$3000 Labor and supplies to do a column settling test for one sediment sample
assuming a column run of 3 days (note: the cost for multiple columns in
use simultaneously would be less) x 3 samples
$3000 Earhart method correlations (for each sediment sample) x 3 samples
Hudson River PCBs Superfund Site
Engineering Performance Standards
9
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
$10000 report preparation and QC reviews for all testing.
Summarizing, for 3 samples:
$4000 + $9000 + $9000 + $10000 = $32,000.
The total estimated cost for this study is approximately $33,000.
5.3 Non-Target, Downstream Area Contamination
For a study involving 40 sediment trap locations that are distributed over a 5 acre area,
and with each of the 40 locations has co-located sediment traps, the following is an
estimate of costs including the cost of the sampling equipment, labor, and sample
handling.
There are very few easily identifiable vendors for sediment traps. However, based on a
quote from Aquatic Research Instruments from Hope ID, the price for one sediment trap
would be about $175. It was indicated that purchasing by volume could affect this price,
but using that quote for estimation purposes:
$175 per sediment trap x 80 traps = $14,000. Add another $500 for necessary sundry
equipment (stakes & rope which are necessary for trap deployment), and the total cost for
equipment is $19,000.
For labor costs, assume that it will take one 2-man crew 3 days to deploy the 80 traps, 1
day per week to collect and manage the samples for the 3-week program, and one day for
demobilization, the labor estimate would be:
1 crew x 2 people x 7 days = 14 man days.
This estimate assumes 10 or 12-hour days. Therefore, the labor costs can be estimated as
approximately $12,000 (14 man days, 12 hours with over time).
Assuming the study duration is three weeks approximately 160 samples, the lab
analytical costs will be approximately $48,000.
The total estimated cost per study is approximately $79,000. For five studies (assuming
the traps can be reused), the total cost for this special study is approximately $319,000.
Hudson River PCBs Superfund Site
Engineering Performance Standards
10
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
6.0 Reasonable Estimate of Monitoring Program Cost
The weekly costs for far-field (Upper Hudson River and Lower Hudson River) and near-
field laboratory analyses are provided in Tables 1 through 3. The daily cost for far-field
and near-field labor are provided in Tables 4 and 5. The costs per day are summarized in
Table 7.
The cost of the monitoring program will depend on the amount of time that is spent at
each monitoring level. It is assumed that Phase 1 will last for 30 weeks and have 210
days of operation. Far-field monitoring will be conducted every day during Phase 1.
Near-field monitoring will be conducted only on the days of operation.. During Phase 1,
on average four operations will be ongoing throughout to meet the production goal of half
the annual production rate. If the monitoring level is routine through Phase 1, the cost of
the monitoring program will be approximately $4,000,000.
Cost if Routine Throughout Phase 1
Upper River Far-Field $987,425
Lower River Far-Field $ 14,400
Near-Field $527,072
Total $1,528,897
It is likely that some amount of non-routine monitoring will be required during Phase 1,
although extended periods of higher level monitoring (Control Level or Threshold) are
not foreseen because the amount of resuspension export can be controlled by changes to
the remediation like maintaining strict adherence to operating procedures. It is unlikely
that the concentrations at Waterford will exceed 350 ng/L Total PCB if Phase 1 is
conducted in River Section 1 and the baseline concentrations stay relatively low.
Therefore, it is likely that the Lower River Far-Field monitoring will be at the Routine
Level throughout Phase 1. For a reasonable estimate of Upper River Far-Field
monitoring, it is assumed that Routine Level monitoring will be needed for 26 of the 30
weeks and Control Level monitoring will be needed for the remaining four weeks.
Similarly for Near-Field monitoring, it is assumed that all stations will be in compliance
for 26 weeks and non-routine monitoring will be required for four weeks. This near-field
non-compliant monitoring is somewhat high assuming four stations will be out of
compliance at each of the 4 operations, but this additional cost may address the limited
far-field monitoring that will accompany exceedances of the near-field suspended solids
resuspension criteria and engineering evaluations. The estimated costs the special studies
(for near-field PCBs, bench scale for a semi-quantitative relationship and non-target
contaminant) are presented in Table 8. A reasonable estimate of the monitoring program
cost for Phase 1 is also provided in Table 8.
The present worth cost estimated for the selected remedy in the feasibility study (FS,
[USEPA, 2000]) is $470,000,000. During Phase 1, approximately 10 percent of the total
volume to be removed will be dredged. Assuming that the cost of Phase 1 will be in
proportion to the amount of sediment dredged, the cost for the Phase 1 operations will be
Hudson River PCBs Superfund Site
Engineering Performance Standards
11
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
approximately $47,000,000. For both the minimum monitoring requirements and the
reasonable estimate, the monitoring program represents less than 10 percent of the total
cost of the Phase 1 program.
The Phase 1 monitoring encompasses more than merely demonstrating compliance with
the resuspension criteria and has been developed to provide answers to questions such as
the nature of the PCB releases. This data generated during the Phase 1 monitoring
program can be used throughout the remediation and justifies the cost of the program.
The water column monitoring cost estimated in the FS for the selected remedy was
substantially lower than the estimated cost of the Phase 1 program presented herein;
however, the performance standard requirements were added during development of the
ROD in response to public comments and the additional costs associated with meeting
fixed standards and answering the questions raised by the public are accounted for in this
estimate. One important goal of the monitoring program during Phase 1 is to gather data
to demonstrate that the water column concentrations and loads can be assessed with
confidence using fewer or less costly measurements (suspended solids or turbidity, as
opposed to PCB analysis). If a semi-quantitative relationship is demonstrated during
Phase 1, the monitoring program can be reduced accordingly for Phase 2.
The costs used in this estimate are conservative. The analytical costs used in these
estimates are higher than what may be negotiated given the large amount of samples. The
amount of labor needed for the monitoring program could differ from what is estimated
here. For instance, if the laboratory were to filter the whole water samples for the levels
other than routine, there would not be a need to add additional people for far-field
sampling (with perhaps an addition of two people to shuttle samples to the lab). In
addition, the monitoring program has been developed to conform to a series of data
quality objectives. This allows for alteration of the monitoring plan as long as all of the
data quality objectives are met. As a result, less costly means of achieving these
objectives may be developed. Similarly, the costs for operating two field laboratories for
seven months (assuming staffing by one technician each for 24 hours per day, seven days
per week) may be on the order of about $550,000 (total for two field labs - based on the
same labor rates as above; and trailer rental and equipment costs of about $10,000 for
each field lab); this may be less costly than the estimate herein, which is based on off-site
laboratory costs for the TSS analyses.
Hudson River PCBs Superfund Site
Engineering Performance Standards
12
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
7.0 Feasibility and Other Considerations
The benefit of using ISCO samplers lies in labor cost savings, as the collection of the 8
daily TDS samples at the far-field stations will be automated. Under this proposed
sampling plan, whole-water PCB and the associated TDS sample will still be collected
manually, with depth-integrated samplers. Field personnel will also be required to gather
the ISCO-generated samples and to replenish sample bottles in the ISCO. The schedule
for this must be determined so as to accommodate overall QC requirements. This task
could effectively be shared between the crews collecting the PCB samples and the field
laboratory personnel.
The ISCO samplers must be positioned in locations that are within product specifications
(e.g., distance from and height above the river) and, to prevent tampering, the ISCOs
must be properly secured. Electric power will have to be provided to the locations, unless
models employing low-voltage DC-current are employed.
Another benefit of the automated samplers is the elimination of variation between
samples, caused by differences in sampling technique of the individual sampler or by
differences in sampling location. They also eliminate the need for people to be out on
bridges or near dams in the dark or inclement weather.
However, the primary advantage of this program is the elimination of managing large
pools of samplers, many being "on-call" for extended periods. Coordinating personnel
required to collect samples at increasingly higher action levels becomes much easier than
the previous program. Under the current plan, a small pool of individual collect the
whole-water PCB samples, reducing the potential for variability in sample technique and
thereby providing the best opportunity to meet data quality objectives.
The use of ISCOs will require the inlet lines be permanently mounted within the river and
safe from recreational traffic. At the 4 bridge stations, inlet tubing could be attached to
bridge abutments or to buoys near the bridge. At the TID station, recreational traffic is
not necessarily an issue; so weighted tubing could be strung from the sampler to buoys
positioned at a safe distance upriver of the dam. This will allow for precise positioning of
intakes to address concerns about flow at that station.
Routine maintenance will likely be required on the ISCO intake ports, as well as after
storm events, to clear accumulated debris carried by the current. This may also involve
repositioning the intake ports due to drift or to high flows. This task can be best
accomplished through the use of a johnboat transported by vehicle between stations and
launched nearby the station.
This proposed program makes efficient use of sampling crews by pairing up locations for
each PCB sampling event. For example, under Routine monitoring conditions, 1 crew of
2 individuals will sample the Ft. Edward and TID stations (upriver crew), while another
crew of 2 people (downriver crew) will sample the Schuylerville, Stillwater and
Waterford stations. The travel time between the Ft Edward and TID stations is
Hudson River PCBs Superfund Site
Engineering Performance Standards
13
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
approximately 10 minutes, meaning that this crew should have a relatively easy time
collecting the samples from both stations. The upriver crew only samples 2 stations due
to unique problems presented at the TID station. The downriver crew will sample 3
stations, however the short travel time (approximately 10 minutes between each station)
and the relative ease of collection at these stations warrant the additional station. Each of
these stations require sampling from bridges that have wide sidewalks and guard rails for
safety.
Another factor to consider is the placement of the mobile labs. If the labs are situated
near the TID station and near the Waterford station, the crew could deliver the upstream
sample to the lab for processing then move downriver to collect samples at the next
station.
With respect to meeting the required turn around times for sample analysis, the proposed
extraction method for PCB analysis is solid-phase extraction (SPE). Although the
extraction time varies somewhat based on the physical characteristics of the sample (e.g.,
suspended matter which may tend to slow down the process), it appears that the actual
process can be completed in an hour or so. Add to that the analysis itself, which may take
a minimum of an hour (based on the time from injection of the sample through the
completion of the analysis). However, it needs to be considered that extraction for 1-L
samples is fairly automated; the 8-L extraction requires manual intervention during the
extraction.
It thus appears that 24-hour TAT for PCBs for the proposed 1-L method is at least
theoretically achievable. Whether or not this would require the lab to run additional shifts
(add a second and/or third shift) or weekend shifts is a separate issue that would be
contingent upon the scheduling of sample delivery to the laboratory, as well as how many
samples had to be processed at once.
A hidden advantage in the decrease in sampling staff is a net decrease in the potential for
safety incidents. The smaller pool of samplers in this program will become acquainted
with specific safety issues at each site, which will help in minimizing accidents. For
additional safety measures, a communication system should be employed, such as hand-
held radios or a higher gain system that could be tied into the field laboratories, as well.
All field personnel should be required to carry cellular telephones (with service that
covers the area in question) to contact local authorities. The Hudson valley presents
unique problems to cellular customers, so care should be paid as to which cellular carrier
is chosen. The Minimal safety equipment for the crewmembers will be steel-toed boots,
hard hats, safety glasses, nitrile gloves and PFDs.
Hudson River PCBs Superfund Site
Engineering Performance Standards
14
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
8.0 Conclusions
The Phase 1 monitoring plan developed for the performance standard measures
compliance with the resuspension criteria and provides important information on the
nature and impact of the remediation on the river. The estimate cost of the water column
monitoring is approximately $3,000,000. The costs developed for Phase 1 cannot be
applied to the entire remediation, because modifications to the monitoring program may
be made for Phase 2; it is likely that these modifications will result in cost reductions
after the Phase 1 program data are reviewed and the Phase 2 monitoring program is
optimized.
Hudson River PCBs Superfund Site
Engineering Performance Standards
15
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
9.0 References
Earhart, H.G. 1984. "Monitoring total suspended solids by using nephelometry,"
Environmental Management 8(1), pp. 81-86.
USEPA, 2000. Phase 3 Report: Feasibility Study, Hudson River PCBs Reassessment
RI/FS. Prepared for USEPA Region 2 and the US Army Corps of Engineers (USACE),
Kansas City District by TAMS Consultants, Inc. December 2000.
Hudson River PCBs Superfund Site
Engineering Performance Standards
16
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
Tables
Hudson River PCBs Superfund Site Malcolm Pirnie/TAMS-Earth Tech
Engineering Performance Standards Volume 2: Attachment H - April 2004
-------
Table 1
Sampling Cost on a Weekly Basis - Upper River Far-Field Stations
Routine Monitoring
Lab
Laboratory Analyses
Laboratory Analyses
Number of Samples per
Day Only
Turn-
Around
Time (hr.)
Congener-Spec. PCBs
Whole Water
DOC &
Susp. OC
SS
SS (1/3-
hours)3
Integrating
Sampler for
PCBs
Congener-Spec. PCBs
Whole Water
DOC &
Susp. OC
SS
SS (1/3-
hours)3
Integrating
Sampler for
PCBs
RM 197.0 - Bakers Falls Br.
72
1
1
1
525
95
20
RM 194.2 - Ft Edward
72
7
7.5
7.5
56
0.5
3,675
713
150
20
150
RM 188.5 - TIDam
24
7
7.5
7.5
56
0.5
4,200
713
150
150
150
RM 181.4 - Schuylerville
24
7
7.5
7.5
56
0.5
4,200
713
150
150
150
RM 163.5 - Stillwater
72
7
7.5
7.5
56
0.5
3,675
713
150
150
150
RM 156.5 -Waterford
72
7
7.5
7.5
56
0.5
3,675
713
150
150
150
Analytical Cost/Week
36
38.5
38.5
280
2.5
19,950
3,658
770
620
750
Total Analytical Cost/Week
38.5 or
5.5 /day |
25,748 or
3,678 /day |
Evaluation Level
Lab
Laboratory Analyses
Laboratory Analyses
Number of Samples per
Day Only
Turn-
Around
Time (hr.)
Congener-Spec. PCBs
Whole Water
DOC &
Susp. OC
SS
SS (1/3-
hours)3
Integrating
Sampler for
PCBs
Congener-Spec. PCBs
Whole Water
DOC &
Susp. OC
SS
SS (1/3-
hours)3
Integrating
Sampler for
PCBs
RM 197.0 - Bakers Falls Br.
72
1
1
1
525
95
20
RM 194.2 - Ft Edward
72
7
7.5
7.5
56
0.5
3,675
713
150
20
150
RM 188.5 - TIDam
24
14
0.5
0.5
56
0.5
8,400
48
10
150
150
RM 181.4 - Schuylerville
24
14
0.5
0.5
56
0.5
8,400
48
10
10
150
RM 163.5 - Stillwater
72
7
7.5
7.5
56
0.5
3,675
713
150
10
150
RM 156.5 - Waterford
72
7
7.5
7.5
56
0.5
3,675
713
150
150
150
Analytical Cost/Week
50
24.5
24.5
280
2.5
28,350
2,328
490
340
750
Total Analytical Cost/Week
52.5 or
7.5 /day |
32,258 or
4,608 /day |
Control Level
Lab
Laboratory Analyses
Laboratory Analyses
Number of Samples per
Day Only
Turn-
Around
Time (hr.)
Congener-Spec. PCBs
Whole Water
DOC &
Susp. OC
SS
SS (1/3-
hours)3
Integrating
Sampler for
PCBs
Congener-Spec. PCBs
Whole Water
DOC &
Susp. OC
SS
SS (1/3-
hours)3
Integrating
Sampler for
PCBs
RM 197.0 - Bakers Falls Br.
72
1
1
1
525
95
20
RM 194.2 - Ft Edward
72
7
7.5
7.5
56
0.5
3,675
713
150
20
150
RM 188.5 - TIDam
24
21
1
1
56
1
12,600
95
20
150
300
RM 181.4 - Schuylerville
24
21
1
1
56
1
12,600
95
20
20
300
RM 163.5 - Stillwater
72
7
7
56
7
665
140
20
3,675
RM 156.5 - Waterford
72
7
7
56
7
665
140
140
3,675
Analytical Cost/Week
50
24.5
24.5
280
16.5
29,400
2,328
490
350
8,100
Total Analytical Cost/Week
66.5 or 9.5 /day |
40,668 or
5,810 /day |
Threshold
Lab
Laboratory Analyses
Laboratory Analyses
Number of Samples per
Turn-
SS (1/3-
Integrating
SS (1/3-
Integrating
Around
Congener-Spec. PCBs
DOC &
Sampler for
Congener-Spec. PCBs
DOC &
Sampler for
Time (hr.)
Whole Water
Susp. OC
SS
hours)3
PCBs
Whole Water
Susp. OC
SS
hours)3
PCBs
RM 197.0 - Bakers Falls Br.
72
1
1
1
525
95
20
RM 194.2 - Ft Edward
72
1
1
1
8
1/2-weeks
525
95
20
20
21
RM 188.5 - TIDam
24
4
1
1
8
1
2,400
95
20
20
600
RM 181.4 - Schuylerville
24
4
1
1
8
1
2,400
95
20
20
600
RM 163.5 - Stillwater
24
4
5
5
8
1
2,400
475
100
20
600
RM 156.5 - Waterford
24
4
5
5
8
1
2,400
475
100
100
600
Analytical Cost/Day
18
14
14
40
4
10,650
1,330
280
180
2,421
Total Analytical Cost/Day
22 /day |
14,861 /day |
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tecl
Volume 2: Attachment H - April 2004
-------
Table 2
Sampling Cost on a Weekly Basis - Lower River Far-Field Stations
Lower River Sampling Requirements on a Weekly Basis
Routine Monitoring
No. of Analyses/Week
Cost of Analyses/Week
Lab
Turn-
Around
Time (hr.)
Congener-
specific
PCBs
Whole
DOC &
Susp. OC
SS
Congener-
specific
PCBs Whole
Water
DOC &
Susp. OC
SS
Mohawk R. at Cohoes
72
0.25
0.25
0.25
131
24
5
RM 140-Albany
72
0.25
0.25
0.25
131
24
5
RM 77 - Highland
72
0.25
0.25
0.25
131
24
5
Analytical Cost/Week
0.75
0.75
0.75
394
71
15
Total Analytical Cost/Week
480
Non-Routine Monitoring
No. of Analyses/Week
Cost of Analyses/Week
Lab
Congener-
Congener-
Turn-
specific
specific
Around
PCBs
DOC &
PCBs Whole
DOC &
Time (hr.)
Whole
Susp. OC SS
Water
Susp. OC SS
Mohawk R. at Cohoes
24
1
1 1
600
95 20
RM 140-Albany
24
1
1 1
600
95 20
RM 77 - Highland
24
1
1 1
600
95 20
Analytical Cost/Week
3
3 3
1800
285 60
Total Analytical Cost/Week
2145
Note:
(1) Non-routine monitoring will be triggered only when Waterford or Troy have total PCB concentration greater than 350 ng/L.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
Table 3
Sampling Cost on a Weekly Basis - Upper River Near-Field Stations
Near-Field Sampling Requirements on a Weekly Basis
Routine Monitoring (with use of continuous reading probe to indicate suspended solids concentrations)
No. of SS
Cost of SS
No. of
Laboratory
Laboratory
Operations
Analyses
Analyses
1
35
700
2
70
1400
3
105
2100
4
140
2800
5
175
3500
6
210
4200
7
245
4900
8
280
5600
9
315
6300
10
350
7000
Non-Routine Monitoring
Number of SS Laboratory Samples with 4-Hour Turn-Around per Week
No. of
Number of Stations with Exceedences of the Standard
All Stations
Operations
1
2
3
4
5
1
49
98
147
196
245
2
98
196
294
392
490
3
147
294
441
588
735
4
196
392
588
784
980
5
245
490
735
980
1,225
6
294
588
882
1,176
1,470
7
343
686
1,029
1,372
1,715
8
392
784
1,176
1,568
1,960
9
441
882
1,323
1,764
2,205
10
490
980
1,470
1,960
2,450
Cost of SS Laboratory Samples with 4-Hour Turn-Around per Week
No. of
Number of Stations with Exceedences of the Standard
All Stations
Operations
1
2
3
4
5
1
980
1,960
2,940
3,920
4,900
2
1,960
3,920
5,880
7,840
9,800
3
2,940
5,880
8,820
11,760
14,700
4
3,920
7,840
11,760
15,680
19,600
5
4,900
9,800
14,700
19,600
24,500
6
5,880
11,760
17,640
23,520
29,400
7
6,860
13,720
20,580
27,440
34,300
8
7,840
15,680
23,520
31,360
39,200
9
8,820
17,640
26,460
35,280
44,100
10
9,800
19,600
29,400
39,200
49,000
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
Table 4
Labor Cost on a Daily Basis - Far-Field Stations
Routine Monitoring
Station
No.of No. of No. of Labor
people shift/day people/day /day
Baker's Falls (1)
Ft Edward/TID
Schuyl/Still/Wat
Total
2 0.1 0.2
2 1 2
2 1 2
4.2 $ 1,747
Evaluation Level
Station
No.of No. of No. of Labor
people shift/day people/day /day
Baker's Falls (1)
Ft Edward/TID/Schuyl
Still/Wat
2 0.1 0.2
2 2 4
2 1 2
Total
6.2 $ 2,579
Contol Level
Station
No.of No. of No. of Labor
people shift/day people/day /day
Baker's Falls (1)
Ft Edward/TID/Schuyl
Still/Wat
2 0.1 0.2
2 3 6
2 1 2
Total
8.2 $ 3,411
Threshold
Station
No.of No. of No. of Labor
people shift/day people/day /day
Baker's Falls (1)
Ft Edward/TID/Schuyl
Ft Edward/TID/Schuyl
Still/Wat
Still/Wat
2 0.1 0.2
2 3 6
2 1 2
2 3 6
2 1 2
Total
16.2 $ 6,739
Notes:
(1) Other stations includes Bakers Falls Bridge and Lower Hudson.
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
Table 5
Labor Cost on a Daily Basis - Near-Field Stations
Near-Field Sampling Requirements on a Weekly Basis
Routine Monitoring
No. of
No. of
No. of
No. of
Labor
Operations
people
shift/day
people/day
/day
1-5
3
2
6
$ 2,496
5-10
6
2
12
$ 4,992
Non-Routine Monitoring
No. of
No. of
No. of
No. of
Labor
Operations
people
shift/day
people/day
/day
1-2
4
2
8
$ 3,328
3-4
8
2
16
$ 6,656
5-6
12
2
24
$ 9,984
7-8
16
2
32
$ 13,312
9-10
20
2
40
$ 16,640
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
Table 6
Near-Field Total PCB Concentration Special Studies
Assumptions:
Assume 4 different types of dredges.
One sampling event will be conducted for each dredge type and one debris removal
Sampling will be conducted once per day for one full work week.
There are 7 days per work week.
5 locations are occupied in the transect.
There are 2 transects:one outside the containment and one 100 m downstream of containment
One subsample is located in water depth greater than 10 ft., others less than 10 ft.
At the one deeper location one sample is collected 0-10 ft, one deeper than 10 ft.
At least three samples will be taken within containment and composited
Samples will be vertically integrated.
All work is done in containment.
Number of upstream samples
Number of samples per transect
Number of transects
Number of samples with containment
Number of Analyses per Day
Number of Days per Event
Total Number of Analyses per Event
Analytical Cost Per Event
Total Analytical Cost per Event
Number of Technicians per Day
Total Labor Costs per Event
Total Cost per Event (Labor+Analytical):
Number of Events
Total for Study
Congener-Specific PCBs
Whole Dissolved Suspended DOC & Probe
Water Phase Phase Susp. OC SS Turbidity
1 111
4 2 2 6 6 6
2
11111
10 5 5 15 15 15
7
70 35 35 105 105
$ 21,000 $ 10,500 $ 10,500 $ 9,975 $ 2,100
$ 54,075
6 (2 boat crews of 3)
$17,472
$ 71,547
5
$ 357,735
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
-------
Table 7
Summary of Labor and Lab Analytical Costs by Action Level
Phase 1 Costs/Day
Upper River Far-Field
Lower River Far-Field
Level
Analytical
Labor
Total
Level
Analytical
Routine
3,678
1,747
5,425
Routine
69
Evaluation
4,608
2,579
7,187
Non-Routine
306
Control
5,810
3,411
9,221
Threshold
14,861
6,739
21,601
Near Field
Routine
Non-Routine
Non-Compliant
Analytical
Analytical
Stations
Labor
1
2
3
4
Labor
1
100
2,496
140
280
420
560
3,328
2
200
2,496
280
560
840
1,120
3,328
3
300
2,496
420
840
1,260
1,680
6,656
4
400
2,496
560
1,120
1,680
2,240
6,656
5
500
2,496
700
1,400
2,100
2,800
9,984
6
600
4,992
840
1,680
2,520
3,360
9,984
7
700
4,992
980
1,960
2,940
3,920
13,312
8
800
4,992
1,120
2,240
3,360
4,480
13,312
9
900
4,992
1,260
2,520
3,780
5,040
16,640
10
1,000
4,992
1,400
2,800
4,200
5,600
16,640
Table 8
Reasonable Estimate of Phase 1 Season Monitoring Plan Costs
Assume:
Half Production (4-operations on average)
Far-Field Sampling on all days; Near-Field on days of operation
210 days of operation
30 weeks/Phase 1
7 days/week far-field sampling
26 weeks of Routine Monitoring Upper River Far-Field
987,425
4 weeks of Control Monitoring Upper River Far-Field
258,184
30 weeks of Routine Monitoring Lower River Far-Field
14,400
26 weeks of Routine Near-Field Monitoring
527,072
4 weeks of Non-Routine Near-Field Monitoring at 4 Stations
249,088
Monitoring Cost:
2,036,169
Special Study for Total PCBs
357,735
Special Study for Total PCBs
33,000
Special Study for Total PCBs
319,000
Monitoring Cost & Special Study:
2,745,904
Hudson River PCBs Superfund Site
Engineering Performance Standards
Malcolm Pirnie/TAMS-Earth Tech
Volume 2: Attachment H - April 2004
------- |