£,EPA
United States
Environmental Protection
Agency
Environmental Impact and Benefits
Assessment for Final Effluent
Guidelines and Standards for the
Construction and Development Category

November 2009

-------

-------
Environmental Impact and Benefits Assessment for the C&D Category
                U.S. Environmental Protection Agency
                       Off ice of Water (4303T)
                   1200 Pennsylvania Avenue, NW
                       Washington, DC 20460

                         EPA-821-R-09-012
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
                 ACKNOWLEDGMENTS AND DISCLAIMER


This document was prepared by U.S. Environmental Protection Agency Office of Water staff. The
following contractor provided assistance in performing the analyses supporting the conclusions
detailed in this document.

                                  Abt Associates, Inc.

Office of Water staff have reviewed and approved this document for publication. Neither the
United States Government nor any of its employees, contractors, subcontractors, or their
employees makes any warranty, expressed or implied, or assumes any legal liability or
responsibility for any third party's use of or the results of such use of any information, apparatus,
product, or process discussed in this document, or represents that its use by such a party would
not infringe on privately owned rights. References to proprietary technologies are not intended to
be an endorsement by the U.S Environmental Protection Agency.
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table of Contents
Table of Contents	v
List of Tables	xi
List of Figures	xvii
Acronyms and Abbreviations	xix
1     Introduction	1-1
2     Overview of Environmental Impacts from Construction Site Sediment Discharges	2-1
2.1   Sediment and Turbidity Discharge to Surface Waters	2-2
2.2   Sediment and Turbidity Behavior in Surface Waters	2-5
      2.2.1  Sediment Characteristics Affecting Surface Water Transport	2-6
      2.2.2  Sediment and Turbidity Behavior in Specific Waterbody Types	2-7
2.3   Aquatic Life Impacts of Sediment and Turbidity	2-11
      2.3.1  Primary Producers	2-13
      2.3.2  Invertebrates	2-15
      2.3.3  Fish	2-18
      2.3.4  Other Wildlife Dependent on Aquatic Ecosystems	2-22
      2.3.5  Threatened and Endangered Species	2-22
2.4   Sediment and Turbidity Impacts on Human Use of Aquatic Resources	2-25
      2.4.1  Navigation on Surface Waters	2-25
      2.4.2  Reservoir Water Storage Capacity	2-26
      2.4.3  Municipal Water Use	2-26
      2.4.4  Industrial Water Use	2-27
      2.4.5  Agricultural Water Use	2-27
      2.4.6  Stormwater Management and Flood Control	2-28
      2.4.7  Recreational  Uses and Aesthetic Value	2-28
      2.4.8  Recreational  and Commercial Fishing	2-29
2.5   Sediment and Turbidity Criteria	2-29
      2.5.1  Federal Sediment Criteria	2-30
      2.5.2  State Sediment Criteria	2-30
2.6   Surface Water Quality  Impairment from Sediment and Turbidity	2-36
      2.6.1  Current Total Suspended Solid Concentrations in U.S.  Surface Waters	2-38
3     Other Pollutants in Construction Site Discharges	3-1
3.1   Introduction	3-1
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
      3.1.1  Pollution Discharge and Transport Pathways	3-1
      3.1.2  National Scale and Cumulative Impact Concerns	3-2
3.2   Construction Site Sources of Pollutants Other Than Sediment and Turbidity	3-3
      3.2.1  Construction Materials and Equipment	3-3
      3.2.2  Historic Site Contamination	3-4
      3.2.3  Natural Site Constituents	3-5
      3.2.4  Altered Stormwater Discharge	3-6
3.3   Other Pollutants from Construction Activity	3-7
      3.3.1  Nitrogen and Phosphorus	3-7
      3.3.2  Organic Compounds and Materials	3-12
      3.3.3  Metals	3-16
      3.3.4  Dissolved Inorganic Ions	3-19
      3.3.5  pH Level	3-20
      3.3.6  Pathogens	3-21
4     Summary of Literature on Construction Site Discharges to Surface Waters	4-1
4.1   Overview of Impacts in Literature	4-1
4.2   Individual Study Summaries	4-17
4.3   State Reports of Construction Discharge Impacts	4-30
5     Overview of Benefits from Regulation	5-1
5.1   Conceptual Framework for Valuation of Environmental Services	5-1
      5.1.1  Applicable Benefit Categories	5-3
      5.1.2  Market Benefits	5-4
      5.1.3  Nonmarket Benefits	5-6
5.2   Summary of Effects	5-9
6     Water Quality Modeling	6-1
6.1   SPARROW Model Documentation	6-1
      6.1.1  General Overview	6-1
      6.1.2  Modeling Concept	6-2
      6.1.3  Model Infrastructure	6-3
      6.1.4  Model Specification	6-4
      6.1.5  Model Estimation	6-6
6.2   SPARROW Sediment Model	6-6
      6.2.1  Sediment Model Data Sources	6-6
      6.2.2  Model Estimation Results	6-8
      6.2.3  Model Simulation	6-9
6.3   Extension of SPARROW to Estuaries and Coastal Waters	6-10

November 2009                                                                                 vi

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.4   Modeling Changes in Nutrient Concentrations	6-12
      6.4.1  Modeling Approach Based on SPARROW Nutrient Models	6-13
6.5   Pollutant Load Modeling	6-15
6.6   Water Quality Modeling Results	6-17
      6.6.1  Current Water Quality Impacts to Surface Water from Construction Site Discharges	6-26
      6.6.2  Estimated Changes in TSS Concentrations in RF1 Watersheds with Most Developed
            Acres (1992-2001) Receiving Sediment Loading from Construction Sites	6-31
      6.6.3  Estimated Changes in TSS, TN, and TP Concentrations for all RF1 Watersheds
            Receiving Loading from Construction Sites	6-35
      6.6.4  Impacts on Reservoir Sedimentation	6-44
      6.6.5  Estimated Changes in Nutrient Concentrations for all RF1 Watersheds Receiving
            Loading from Construction Sites	6-44
7     Benefits to Navigation	7-1
7.1   Data Sources	7-1
7.2   Identifying Waterways That Are Dredged, the Frequency of Dredging, and the Quantity of
      Sediment Dredged	7-3
      7.2.1  Determining Dredging Job Locations	7-3
      7.2.2  Identifying the Baseline Frequency of Dredging	7-3
      7.2.3  Identifying the Amount of Sediment Dredged	7-5
7.3   Estimating the Navigational Maintenance Cost per Cubic Yard of Sediment Removed	7-5
7.4   Estimating the Total Cost of Navigable Waterway Maintenance Under Different Policy
      Scenarios	7-6
      7.4.1  Estimating the Reduction in Sediment Dredging in Navigable Waterways Due to the
            Reduction in Discharge from Construction Sites	7-6
      7.4.2  Estimating the Total  Cost of Navigable Waterway Maintenance Under the Baseline and
            Post-Compliance Scenarios	7-6
      7.4.3  Sensitivity Analysis	7-7
      7.4.4  Annualizing Future Dredging Costs	7-8
7.5   Estimating the Avoided Costs from Decreased Dredging of Navigable Waterways	7-10
7.6   Sources of Uncertainty and Limitations	7-13
8     Benefits to Water Storage	8-1
8.1   Review of Literature on Reservoir Sedimentation	8-1
8.2   Data Sources	8-3
8.3   Estimating the Unit Cost of Sediment Removal from Reservoirs	8-4
8.4   Estimating the Total Cost of Reservoir Dredging Under Different Policy Options	8-4
      8.4.1  Estimating Sediment Accumulation in Reservoirs and the Amount of Sediment
            Expected To Be Dredged	8-4
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
      8.4.2  Estimating the Total Cost of Reservoir Dredging Under the Baseline Scenario and
            Policy Options	8-5
8.5   Estimating the Avoided costs from Reduced Reservoir Dredging	8-6
8.6   Sources of Uncertainty and Limitations	8-8
9     Benefits to Drinking Water Treatment	9-1
9.1   Construction Discharge Effects on Drinking Water Treatment Costs	9-1
      9.1.1  Sediment	9-1
      9.1.2  Nutrients	9-1
      9.1.3  Increase in Primary Productivity and Associated Water Quality Impacts from Nutrient
            Discharges (Eutrophication)	9-2
9.2   Avoided Costs of Drinking Water Treatment from Reducing Sediment Discharge from
      Construction Sites	9-4
9.3   Data Sources	9-6
9.4   Modeling Sediment Concentrations and Reductions	9-6
9.5   Identifying  Reaches with Drinking Water Intakes and Their Intake Volumes	9-7
9.6   Estimating the Cost of Chemical Turbidity Treatment	9-8
9.7   Estimating the Cost of Sludge Disposal	9-10
      9.7.1  Estimating the Amount of Sludge Generated	9-10
      9.7.2  Calculating the Probability That the  Sludge Will Require Off-Site Disposal	9-11
      9.7.3  Estimating the Distance to the Disposal Location	9-11
      9.7.4  Calculating the Cost of Sludge Disposal	9-11
9.8   Sensitivity Analysis	9-12
9.9   Estimating the Total Costs of Drinking Water Treatment Under Different Policy Scenarios	9-13
9.10  Estimating Avoided costs from  Lower Sediment and Turbidity in Drinking Water Influent	9-13
9.11  Sources of Uncertainty/Limitations	9-16
10    Nonmarket Benefits from Water Quality Improvements	10-1
10.1  Water Quality Index	10-1
      10.1.1 Freshwater WQI for Rivers and Streams	10-3
      10.1.2 WQI Methodology for Estuaries	10-5
      10.1.3 Relation between WQI and Suitability for Human Uses	10-6
      10.1.4 Sources of Data on Ambient Water Quality	10-7
      10.1.5 Estimated Changes in Water Quality (AWQI) from the Regulation	10-9
10.2  Willingness to Pay for Water Quality Improvements	10-16
10.3  Estimating Total WTP for Water Quality Improvements	10-22
10.4  Uncertainty and Limitations	10-25
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
11    Total Estimated Benefits	11-1
11.1  Summary of the Estimated Benefits	11-1
11.2  Sources of Uncertainty and Limitations	11-5
      11.2.1 Water Quality Model Limitations	11-6
      11.2.2Focus on Selected Pollutants of Concern (Sediment and Nutrients)	11-7
      11.2.3 Omission of Several Benefit Categories from the Analysis of Monetized Benefits	11-7
      11.2.4Limitations Inherent in the Estimate of Nonmarket Benefits	11-8
12    References	12-1
Appendix A- Status of Available Data on Waterbody Assessments	A-l
Appendix B - Threatened and Endangered Aquatic Species	B-l
Appendix C - SPARROW Model Documentation	C-l
C.I General Overview	C-l
C.2 Modeling Concept	C-2
      C.2.1 Objectives of SPARROW Modeling	C-2
      C.2.2 SPARROW Mass Balance Approach	C-4
      C.2.3 Time and Space Scales of the Model	C-4
      C.2.4 Accuracy and Complexity of SPARROW Models	C-6
      C.2.5 Comparison of SPARROW with Other Watershed Models	C-7
      C.2.6 Model Infrastructure	C-8
      C.2.7 Use of Monitoring Data	C-10
      C.2.8 Model Specification for Monitoring Station Flux Estimation	C-ll
      C.2.9 Tools for Flux Estimation	C-ll
      C.2.10 Guidance for Specifying Monitoring Station Flux Models	C-12
      C.2.11 Stream Network Topology	C-13
      C.2.12 Watershed Sources and Explanatory Variables	C-15
C.3 Model Specification	C-17
      C.3.1 Model Equation and Specification of Terms	C-17
      C.3.2 Contaminant Sources	C-19
      C.3.3 Landscape Variables	C-20
      C.3.4 Stream Transport	C-21
      C.3.5 Reservoir and Lake Transport	C-21
C.4 Model Estimation	C-22
      C.4.1 Evaluation of Model Parameters	C-22
C.5 Model Prediction	C-23
Appendix D - A Preliminary SPARROW Model of Suspended Sediment for the Coterminous
      United States (Schwarz 2008a)	D-l
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
Appendix E - Estuary TSS Concentration Calculations Using Dissolved Concentration
      Potentials	E-l
Appendix F- TSS Subindex Curve Parameters	F-l
Appendix G - Meta-Analysis Results	G-l
G.I Literature Review of Water Resource Valuation Studies	G-l
      G.I.I Identifying Water Resource Valuation Studies	G-2
      G.I.2 Description of Studies Selected for Total WTP Meta-Analysis	G-4
G.2 Total WTP Meta-Analysis Regression Model and Results	G-7
      G.2.1 Metadata Total WTP Regression Model	G-8
      G.2.2 Total WTP Regression Model and Results	G-12
      G.2.3 Interpretation of Total WTP Regression Analysis Results	G-14
      G.2.4 Model Selection	G-18
G.3 Model Limitations	G-19
Appendix H- Water Quality Index Tables	H-l
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
List of Tables
Table 2-1: Sediment and Turbidity Terminology	2-2
Table 2-2: Sediment Grade Scale	2-6
Table 2-3: Suspended Sediment and Turbidity Criteria for Surface Water Quality by State	2-32
Table 2-4: Surface Waters Impaired by "Sediment," by EPA Region	2-37
Table 2-5: Surface Waters Impaired by "Turbidity" and "Suspended Solids," by EPA Region	2-37
Table 2-6: SPARROW Distribution of TSS Concentrations in RF1 Reaches1	2-39
Table 3-1: Nutrient-Related Impairment in 305(b)-Assessed Waters, by EPA Region	3-11
Table 4-1: Studies Documenting Construction Site Discharges to Surface Waters and Their Impacts	4-4
Table 4-2: Summary of Suspended Sediment Concentrations from Selected Studies	4-13
Table 4-3: Summary of Sediment Yield Data from Selected Studies	4-15
Table 4-4: Summary of Turbidity Findings from Selected Studies	4-16
Table 4-5: Construction Impairment in 305(b)-Assessed Waters	4-31
Table 5-1: Summary of Benefits from Reducing Sediment and Other Pollutant Discharges from
     Construction Sites	5-11
Table 6-1: Estimation Results for the SPARROW Suspended Sediment Model	6-9
Table 6-2: Nutrient-to-Sediment Ratios	6-14
Table 6-3: Literature Values of Nitrogen and Phosphorus in Soil and Sediment	6-14
Table 6-4: Baseline Construction Sediment Loading Summary and Post-Compliance Reductions	6-15
Table 6-5: Construction Sediment Loading Summary and Reductions by Option and EPA Region	6-16
Table 6-6: Distribution of TSS Concentrations based on SPARROW Output for 31,927 RF1 Reaches
     Receiving Construction Sediment Discharges	6-26
Table 6-7: Summary of Baseline TSS Concentration in RF1 Reaches Receiving Construction
     Sediment Discharges, by EPA Region	6-27
Table 6-8: Summary of TSS Concentrations Under the Hypothetical No Discharge Scenario, by EPA
     Region	6-27
Table 6-9: Water Quality Impacts:  Improvements in TSS Concentrations Under the Hypothetical No
     Discharge Scenario, by EPA Region	6-28
Table 6-10: Distribution of TN Concentrations based on SPARROW Output for 31,927 RF1 Reaches
     Receiving Construction Sediment Discharges	6-29
Table 6-11: Summary of Baseline TN Concentration in RF1 Reaches Receiving Construction
     Sediment Discharges, by EPA Region	6-29
Table 6-12: Summary of TN Concentrations Under the Hypothetical No Discharge Scenario, by EPA
     Region	6-29

November 2009                                                                               xi

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-13: Distribution of TP Concentrations based on SPARROW Output for 31,927 RF1 Reaches
      Receiving Construction Sediment Discharges	6-30
Table 6-14: Summary of Baseline TP Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-30
Table 6-15: Summary of TP Concentrations Under the Hypothetical No Discharge Scenario, by EPA
      Region	6-30
Table 6-16: Reaches with Largest Increase in Developed Land Area (1992-2001)	6-32
Table 6-17: Distribution of RF1 Watersheds that Receive Direct Loadings by Increase in Developed
      Land Area (1992-2001), by EPA Region	6-32
Table 6-18: Estimated Changes in RF1 Reach TSS Concentration by Policy Option: Top 1% of RF1
      Watersheds Receiving Direct Construction Loadings by Increase in Developed Land Area
      (1992-2001)	6-33
Table 6-19: Estimated Changes in RF1 Reach TSS Concentration by Policy Option: Top 10% of RF1
      Watersheds Receiving Direct Construction Loadings by Increase in Developed Land Area
      (1992-2001)	6-33
Table 6-20: Estimated Changes in RF1 Reach TSS Concentration by Policy Option: Top 25% of RF1
      Watersheds by Increase in Developed Land Area (1992-2001)	6-34
Table 6-21: Distribution of TSS Concentration based on SPARROW Output for 31,927 RF1 Reaches
      Receiving Construction Sediment Discharges	6-35
Table 6-22: Summary of Baseline TSS Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-36
Table 6-23: Summary of Option 1 TSS Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-36
Table 6-24: Summary of Option 2 TSS Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-37
Table 6-25: Summary of Option 3 TSS Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-37
Table 6-26: Summary of Option 4 TSS Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-38
Table 6-27: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS
      Concentrations	6-40
Table 6-28: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS
      Concentrations from Option 1, by EPA Region	6-40
Table 6-29: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS
      Concentrations from Option 2, by EPA Region	6-41
Table 6-30: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS
      Concentrations from Option 3, by EPA Region	6-42
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-31: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS
      Concentrations from Option 4, by EPA Region	6-43
Table 6-32: Sediment Accumulation in Reservoirs by Policy Option and EPA Region	6-44
Table 6-33: Distribution of TN Concentration Based on SPARROW Output for 31,927 RF1 Reaches
      Receiving Construction Sediment Discharges	6-45
Table 6-34: Distribution of TP Concentration Based on SPARROW Output for 31,927 RF1 Reaches
      Receiving Construction Sediment Discharges	6-45
Table 6-35: Summary of Baseline TN Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-46
Table 6-36: Summary Option 1 TN Concentration in RF1 Reaches Receiving Construction Sediment
      Discharges, by EPA Region	6-46
Table 6-37: Summary of Option 2 TN Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-47
Table 6-38: Summary of Option 3 TN Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-47
Table 6-39: Summary of Option 4 TN Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-48
Table 6-40: Summary of Baseline TP Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-48
Table 6-41: Summary of Option 1 TP Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-49
Table 6-42: Summary of Option 2 TP Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-49
Table 6-43: Summary of Option 3 TP Concentration in RF 1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-50
Table 6-44: Summary of Option 4 TP Concentration in RF1 Reaches Receiving Construction
      Sediment Discharges, by EPA Region	6-50
Table 7-1:  Dredging in U.S. Navigable Waterways, 1995-2008	7-2
Table 7-2:  Dredging of Navigable Waterways Performed or Contracted by USAGE, 1995-2008	7-3
Table 7-3:  Dredging Jobs and Recurrence Intervals, 1995- 2008	7-4
Table 7-4:  Costs of Recurring Dredging, 1995-2006	7-5
Table 7-5:  Annualized Dredging Costs Under the Baseline Scenario (millions of 2008$)	7-9
Table 7-6:  Reductions in Dredging and Annualized Avoided Costs Under Option 1	7-11
Table 7-7:  Reductions in Dredging and Annualized Avoided Costs Under Option 2	7-11
Table 7-8:  Reductions in Dredging and Annualized Avoided Costs Under Option 3	7-12
Table 7-9:  Reductions in Dredging and Annualized Avoided Costs Under Option 4	7-12
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 8-1: Average Annual Sedimentation Deposition Recorded by RESIS	8-3
Table 8-2: Estimated Cost of Reservoir Dredging Under the Baseline Scenario	8-6
Table 8-3: Reduction in Reservoir Dredging and Avoided Costs Under Option 1	8-7
Table 8-4: Reduction in Reservoir Dredging and Avoided Costs Under Option 2	8-7
Table 8-5: Reduction in Reservoir Dredging and Avoided Costs Under Option 3	8-8
Table 8-6: Reduction in Reservoir Dredging and Avoided Costs Under Option 4	8-8
Table 9-1: TN Concentrations Predicted by SPARROW for Reaches Serving as Drinking Water
      Sources1	9-2
Table 9-2: Public Water Intakes and Estimated Average Daily Flow by EPA Region	9-8
Table 9-3: Drinking Water Turbidity Treatment Costs Under the Baseline Scenario	9-13
Table 9-4: Reduction in Drinking Water Treatment Costs Under Option 1	9-15
Table 9-5: Reduction in Drinking Water Treatment Costs Under Option 2	9-15
Table 9-6: Reduction in Drinking Water Treatment Costs Under Option 3	9-16
Table 9-7: Reduction in Drinking Water Treatment Costs Under Option 4	9-16
Table 10-1: Freshwater Water Quality Subindices	10-4
Table 10-2: Original and Revised Weights for Freshwater WQI Parameters	10-4
Table 10-3: Estuarine Water Quality Subindices	10-5
Table 10-4: Weights for Estuarine WQI Parameters	10-6
Table 10-5: Water Quality Classifications	10-6
Table 10-6: Percentage of Reach Miles in Coterminous 48 States by WQI Classification for EPA
      Regions: Baseline Scenario	10-9
Table 10-7: Estimated Water Quality Improvements Under Option I1	10-12
Table 10-8: Estimated Water Quality Improvements Under Option 2:	10-13
Table 10-9: Estimated Water Quality Improvements Under Option 31	10-14
Table 10-10: Estimated Water Quality Improvements Under Option 4:	10-15
Table 10-11: Independent Variable Assignments	10-18
Table 10-12: Estimates of Annual Household Willingness to Pay for Water Quality Improvement by
      Region and Policy Option (2008$)	10-21
Table 10-13 :Regional Willingness to Pay for Water Quality Improvement (Millions 2008$)	10-24
Table 11-1: Annual Total National Benefits by Benefits Category (millions of 2008$)	11-3
Table 11-2: Annual Total National Benefits Under Option 1 (millions of 2008$)	11-4
Table 11-3: Annual Total National Benefits Under Option 2 (millions of 2008$)	11-4
Table 11-4: Annual Total National Benefits Under Option 3 (millions of 2008$)	11-5
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 11-5: Annual Total National Benefits Under Option 4 (millions of 2008$)	11-5
Table A-1: Year of Available Waterbody Assessment Data, as of September 17, 2009	A-1
Table B-l: List of Federal Threatened and Endangered Aquatic Species Potentially Impacted by
      Sediment	B-l
Table E-l: TSS Concentrations (mg/L) Estimated by DCP, Southeastern Estuaries	E-l
Table E-2: TSS Concentrations (mg/L) Estimated by DCP, Gulf of Mexico Estuaries	E-l
Table E-3: TSS Concentrations (mg/L) Estimated by DCP, Northeastern Estuaries	E-2
Table E-4: TSS Concentrations (mg/L) Estimated by DCP, West Coast Estuaries	E-3
Table E-5: Total Number Reaches in the Analysis located in Southeastern Estuaries that Use DCP to
      Calculate TSS by Estuary	E-3
Table E-6: Total Number Reaches in the Analysis located in Gulf of Mexico Estuaries that Use DCP
      to Calculate TSS by Estuary	E-4
Table E-7: Total Number Reaches in the Analysis located in Northeastern Estuaries that Use DCP to
      Calculate TSS by Estuary	E-4
Table E-8: Total Number Reaches in the Analysis located in West Coast Estuaries that Use DCP to
      Calculate TSS by Estuary	E-4
Table F-l: TSS Subindex Curve Parameters, by Ecoregion	F-l
Table G-l: Selected Summary Information for Studies	G-5
Table G-2: Variables and Descriptive Statistics for the Total WTP Regression Model	G-ll
Table G-3: Estimated Multilevel Model Results for the Trans-log and Semi-log Total WTP
      Regression Models: WTP for Aquatic Habitat Improvements	G-14
Table G-4: Comparison of WTP for Different Changes in WQI Based on Semi-log and
      Trans-log Models	G-19
Table H-l: Estimated Water Quality Improvements Under Option 1	H-2
Table H-2: Estimated Water Quality Improvements Under Option 2	H-5
Table H-3: Estimated Water Quality Improvements Under Option 3	H-8
Table H-4: Estimated Water Quality Improvements Under Option 4	H-ll
November 2009

-------

-------
Environmental Impact and Benefits Assessment for the C&D Category
List of Figures
Figure 2-1: EPA Regions	2-37
Figure 2-2: Extent of Stressors and Their Relative Risk to the Biological Condition of the Nation's
      Streams	2-38
Figure 2-3: TSS Concentrations by Total RF1 Reach Miles as Predicted by SPARROW for Current
      Conditions	2-39
Figure 3-1: Extent of Stressors and Their Relative Risk to the Biological Condition of the Nation's
      Streams	3-12
Figure 5-1: Calculation of Monetized Benefits from the Regulation	5-12
Figure 6-1: Schematic of the Major SPARROW Model Components	6-3
Figure 6-2: Schematic Illustrating a Vector Stream Reach Network with Node Topology and
     Water/Contaminant Reach-Node Routing Table	6-4
Figure 6-3: Location of 1,828 Water Quality Monitoring  Stations Used in the SPARROW Sediment
      Model, in Relation to the Reach File 1 (RF1) Reach Network	6-8
Figure 6-4: Plot of Construction Acres by RF1 Watershed Percentile Group (1992-2001)	6-18
Figure 6-5: Cumulative Plot of Construction Acres by RF1 Watershed (1992-2001)	6-19
Figure 6-6: EPA Region 1: Percent Urban Change 1992-2001 by RF1 Watershed	6-20
Figure 6-7: EPA Region 2: Percent Urban Change 1992-2001 by RF1 Watershed	6-20
Figure 6-8: EPA Region 3: Percent Urban Change 1992-2001 by RF1 Watershed	6-21
Figure 6-9: EPA Region 4: Percent Urban Change 1992-2001 by RF1 Watershed	6-21
Figure 6-10: EPA Region 5: Percent Urban Change 1992-2001 by RF1 Watershed	6-22
Figure 6-11: EPA Region 6: Percent Urban Change 1992-2001 by RF1 Watershed	6-22
Figure 6-12: EPA Region 7: Percent Urban Change 1992-2001 by RF1 Watershed	6-23
Figure 6-13: EPA Region 8: Percent Urban Change 1992-2001 by RF1 Watershed	6-23
Figure 6-14: EPA Region 9: Percent Urban Change 1992-2001 by RF1 Watershed	6-24
Figure 6-15: EPA Region 10: Percent Urban Change 1992-2001 by RF1 Watershed	6-24
Figure 9-1: Steps in Calculating the Cost of Treating Turbidity in Drinking Water	9-5
Figure 12-1: A  Simple Continuum of Model Types Based on the Level of Statistical and Mechanistic
      Descriptions of Contaminant Sources and Biogeochemical Processes	C-7
Figure 12-2: Schematic of the Major SPARROW Model  Components	C-9
Figure 12-3: Location of 1,828 Water Quality Monitoring Stations Used in the SPARROW Sediment
      Model, in Relation to the Reach File 1 (RF1) Reach Network	C-10
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
Figure 12-4: Schematic Illustrating a Vector Stream Reach Network with Node Topology and
      Water/Contaminant Reach-Node Routing Table	C-13
Figure 12-5: Schematic Illustrating Digital Overlay of Stream Reach Drainage Area and Polygonal
      Areas Associated with Diffuse Sources	C-15
Figure 12-6: Schematic Illustrating Digital Overlay of Stream Reach Drainage Area and Polygonal
      Areas Associated with Landscape Properties	C-16
Figure 12-7: Conceptual Illustration of a Reach Network for Five Incremental Watersheds. Model
      Equation C-4 Describes the Supply and Transport of Load within an Individual Reach and its
      Incremental Watershed	C-17
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Acronyms and Abbreviations
AGNPS - Agricultural Nonpoint Source Model
ASCE - American Society of Civil Engineers
ATS - Active Treatment Systems
ATTAINS - Assessment TMDL Tracking and Implementation System
AWWA - American Water Works Association
BMP - Best Management Practices
BOD - Biological Oxygen Demand
BUVD - Benefits Use Valuation Database
C&D - Construction and Development
cfs - cubic feet per second
ChA - Chlorophyll-a
cm - centimeter
CPI - Consumer Price Index
CS - Compensating Surplus
CWA-Clean Water Act
DAF - Dissolved Air Flotation
DCPs - Dissolved Concentration Potentials
OEMs - Digital Elevation Models
DO - Dissolved Oxygen
EDA - Estuarine Drainage Area
EMAP-NCA - Environmental Monitoring & Assessment Program, National Coastal Assessment
Monitoring Data
EPA - Environmental Protection Agency
EPT - Ephemeroptera, Plectoptera, and Trichoptera
ERF - Enhanced Reach File
EVRI - Environmental Valuation Resource Inventory
FC - Fecal Coliform
ft/s - feet per second
FTUs - Formazin Turbidity Units
g - Gram
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
GIS - Geographic Information System
HAB - Harmful Algal Bloom
HARC - The Houston Area Research Center
HSPF - Hydrologic Simulation Program-Fortran
HUC - Hydrologic Unit Code
JTU - Jackson Turbidity Units
kg/ha - Kilograms per hectare
LC - Lethal Concentration
LID - Low Impact Development
m - Meter
MCL - Maximum Contamination Limit
mgd - Millions of gallons per day
ug/L - Micrograms per liter
mg/L - Milligrams per liter
MIB - 2-methylisoborneol
mL/L - Milliliters per liter
mm - Millimeter
MRLC - Multi-Resolution Land Characteristics Consortium
NAWQA - National Water-Quality Assessment
NBER - National Bureau of Economic Research
NCEE - National Center for Environmental Economics
NEIWPCC - New England Interstate Water Pollution Control Commission
NERRS - The National Estuarine Research Reserve System
NHD - National Hydrogeography Dataset
NID - National Inventory of Dams
NLCD - National Land Cover Database
NMFS - National Marine Fisheries Service
NOAA - National Oceanic and Atmospheric Administration
NPDES - National Pollutant Discharge Elimination System
NRC - National Research Council
NRCS - Natural Resource Conservation Service
NRI - National Resources Inventory
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
NSQAN - National Stream Quality Accounting Network
NTUs - Nephelometric Turbidity Units
NWIS - National Water Information System
NWLS - Nonlinear Weighted Least Squares
PAC - Powdered Activated Carbon
PAHs - Polycyclic Aromatic Hydrocarbons
POTWs - Publicly Owned Treatment Works
PPI - Producer Price Index
PTS - Passive Treatment Systems
RESIS - Reservoir Sedimentation Survey Information System
RF1 - Reach File Version 1.0
RFF - Resources for the Future
RMSE - Root Mean Squared Error
RR-River Reach
RUSLE - Revised Universal Soil Loss Equation
RUSLE k-factor- Soil Erodibility
RUSLE r-factor - Precipitation
SABS - Suspended and Bedded Sediments
SAS -  Statistical Analysis System
SAS IML - Statistical Analysis System Interactive Matrix Language
SAV - Submerged Aquatic Vegetation
SCDHEC - South Carolina Department of Health and Environmental Control
SCDNR- South Carolina Department of Natural Resources
SDWIS - Safe Drinking Water Information System
SPARROW - Spatially Referenced Regressions on Watershed Attributes
SSC - Suspended Sediment Concentration
STATSGO -  State Soil Geographic database
STORET - Storage and Retrieval data warehouse
SWAT - Soil Water Assessment Tool
T&E - Threatened and Endangered
TDS - Total Dissolved Solids
TESS - Threatened and Endangered Species System database
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
TMDL - Total Maximum Daily Load
TM/ETM - Thematic Mapper/Embedded Trace Macrocells
TN - Total Nitrogen
TP - Total Phosphorus
TSS - Total Suspended Solids
U.S.  OMB - United States Office of Management and Budget
USAGE - United States Army Corps of Engineers
USDA - United Stated Department of Agriculture
EPA - United States Environmental Protection Agency
USFWS - United States Fish and Wildlife Service
USGS - United States Geological Survey
USLE - Universal Soil Loss Equation
WATERS - Watershed Assessment Tracking & Environmental Results
WHO - World Health Organization
WQI - Water Quality Index
WQL - Water Quality Ladder
WSUT - Weak Structural Utility Theoretic
WTP - Willingness To Pay
WWF - World Wildlife Fund
WY-Water Year
yd3 - cubic yard
November 2009

-------
Environmental Impact and Benefits Assessment for the C&D Category
        Introduction
This document presents information on environmental impacts associated with construction site
discharges to surface waters and benefits associated with their reduction. Additional information on the
construction industry and construction site discharges to surface waters is provided in the U.S.
Environmental Protection Agency's (EPA's) Development Document for Final Effluent Guidelines and
Standards for the Construction and Development Category (USEPA 2009b) and Economic Analysis for
Final Effluent Guidelines and Standards for the Construction and Development Category (USEPA
2009c).
Construction takes place on approximately 853,000 acres in the coterminous United States and discharge
more than 5 billion pounds of sediment each year (USEPA 2009b). All major surface water types receive
construction discharges, including streams, rivers, wetlands, lakes, estuaries, and other coastal waters. In
any given year, some level of construction activity occurs in the majority of U.S. watersheds. However,
most construction acreage is concentrated in a relatively limited number  of watersheds. Between 1992
and 2001, more than half of all construction activity took place in less than 5 percent of U.S. watersheds.1
A common development pattern is for new construction to concentrate in rural and suburban watersheds
adjacent to more densely developed urban areas. Because construction is a temporary activity, the
locations of the most highly impacted watersheds in the United States shift over time.
Construction activities significantly change the surface of the land.  Typical activities include clearing
vegetation and excavating, moving, and compacting earth and rock. Consequences from these activities
include reduced stormwater infiltration, increased runoff volume and intensity, and higher soil erosion
rates.
Construction sites have been documented to increase pollutant discharges to surface waters. The most
thoroughly documented pollutants are sediment and turbidity. Nitrogen and phosphorus are common soil
constituents that can also pollute receiving waters. Other pollutants can derive from a wide variety of
construction equipment and materials or from historic contamination of construction sites. These
pollutants include metals, trash and debris, nutrients, organic matter, pesticides, petroleum hydrocarbons,
polycyclic aromatic hydrocarbons (PAHs), and other toxic organics. Construction activity can also impact
receiving waters by increasing the volume and intensity of stormwater runoff from a site. These flows can
erode receiving water banks and beds, particularly when pipes, ditches, or other stormwater conveyances
concentrate discharges. Surface water erosion can alter waterbody morphology and elevate sediment and
turbidity levels downstream.
Most pollutants enter surface waters when precipitation erodes soil  and carries particulate matter to
receiving waters. A number of pollutants bind to soil particles and travel with them as they erode. Other
pollutants dissolve in precipitation and are carried to surface waters in solution. Construction site
pollutants can also enter surface waters during dry weather due to activities such as excavation
dewatering, construction equipment washout, wind erosion, and equipment operation in or near surface
waters.
1    Watersheds are those associated with the more than 62,000 surface water reaches in the coterminous United States in the
    Reach File Version 1 (RF1) surface water network (USEPA 2007f). See Chapter 4 for additional information.
November 2009                                                                                   1-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
Increases in pollutant and stormwater discharges from construction sites and subsequent increases in
waterbody pollutant levels can have adverse impacts on aquatic ecosystem function and human aquatic
resource use. Construction is a temporary activity in any given location, but impacts can range from
temporary to long-term. Impacts include both physical and chemical impacts on waterbodies and
biological impacts on aquatic organisms and communities. Impacts to human aquatic resource uses can
include impaired drinking water supplies, recreation, navigation, fishing, water storage, aesthetics,
property value, irrigation, industrial water supplies, and stormwater (including flood) management.
Though some surface waters can eventually recover from construction discharge impacts, they may
continue to be degraded by excess stormwater and pollutants from buildings, roads, and other structures
put in place by construction activity. This document, however, focuses solely on the impacts associated
with active construction sites. Chapters 2 and 3 provide additional information on the pollutants and
environmental impacts associated with construction site discharge to surface waters, with Chapter 2
focusing on sediment and turbidity and Chapter 3 presenting all other pollutants associated with
construction site stormwater discharges. Chapter 4 presents a review of literature documenting impacts of
construction site stormwater discharges.
EPA has established effluent limitations guidelines (ELGs) and new source performance standards
(NSPS) for stormwater discharges from the construction and development industry. These guidelines and
standards require discharges from certain construction sites to meet a numeric turbidity limit. The
guidelines and standards also require all construction sites currently required to obtain a National
Pollutant Discharge Elimination System (NPDES) permit to implement a variety of best management
practices (BMPs)  designed to limit erosion  and  control sediment discharges from construction sites. EPA
evaluated four options in developing the final rule. These options are described below:
        >  Option 1 establishes minimum  requirements for implementing a variety of erosion and
           sediment controls and pollution prevention measures on all construction sites that are
           required to obtain a permit.
        >  Option 2 contains the same requirements as Option 1. In addition, construction sites of 30 or
           more  disturbed acres would be  required to meet a numeric turbidity limit in stormwater
           discharges from the site. The technology basis for the numeric limit is Active Treatment
           Systems (ATS). The numeric turbidity standard would be applicable to stormwater discharges
           for all storm events up to the local 2-year, 24-hour event.
        >  Option 3 contains the same requirements as Option 1. Option 3 also requires all sites with 10
           or more acres of disturbed land to meet a numeric turbidity standard based on the application
           of ATS. The turbidity standard would apply to all stormwater discharges for all storm events
           up to the local 2-year, 24-hour  event.
        >  Option 4 contains the same requirements as Option 1. Option 4 also requires all sites with 10
           or more acres of disturbed land to meet a numeric turbidity standard of 280 NTU based on the
           application of Passive Treatment Systems (PTS). The turbidity standard would apply to all
           stormwater discharges for all storm events up to the local 2-year, 24-hour event, although
           only certain types of discharges would require monitoring.
EPA expects that reduced discharges of pollutants from construction sites will enhance environmental
services provided by the affected water bodies and, as a result, human welfare. Chapter 5 provides an
overview of EPA's approach to assessing and quantifying the benefits of reducing construction site
discharges to surface waters. Chapter 6 describes the methodology EPA used to quantify improvements
in surface water quality associated with each of the evaluated regulatory options. EPA used the Spatially
November 2009                                                                                  1-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
Referenced Regressions on Watershed Attributes (SPARROW) method, the Dissolved Concentration
Potential (DCP) method, and information on relationships between sediment and nutrient levels in surface
waters to quantify changes in surface water levels of total suspended solids (TSS), sedimentation, total
phosphorus (TP), and total nitrogen (TN) under each regulatory option. Chapter 6 also presents the results
of this analysis. EPA found available data on the location and magnitude of other types of construction
site pollutant discharges to be insufficient for inclusion in the water quality modeling analysis.
In analyzing benefits of the regulation, EPA quantified and monetized economic benefits from reduced
dredging of navigable waterways, reduced dredging of water storage facilities (reservoirs), and reduced
drinking water treatment costs. Chapters 7, 8, and 9 describe the methods EPA used to analyze these
benefit categories and present the results of EPA's analysis. Other benefit categories, including reduced
flood risk, increases in property values, industrial water use, agricultural water use, stormwater
management system management, and commercial fishing, are discussed qualitatively in Chapters 2 and
5. EPA estimates that Options 1, 2, 3, and 4 will reduce expenditures on navigable waterway dredging by
approximately $1.3, $2.6, $3.3, and $2.9 million per year respectively. Expenditures on reservoir
dredging are expected to decrease by $1.4 million per year under Option 1, $2.9 million per year under
Option 2, $3.6 million per year under Option 3, and $3.2 million per year under Option 4. Reductions in
drinking water treatment costs are estimated to amount to $1.2, $1.8 million, $2.1 million, and $1.8
million annually under Options 1, 2, 3 and 4, respectively. Overall, EPA expects this regulation to save
governments and private entities between $3.8 and $8.9 million in these three areas each year, depending
on the policy option.
EPA also expects that reductions in pollutant discharges to surface waters resulting due to regulation will
enhance or protect aquatic ecosystems. The drop in pollutant discharges is expected to improve the
protection of resident species; enhance the general health offish, invertebrate, plant and other aquatic
organism populations; increase their propagation in waters currently impaired; and expand fisheries for
both commercial and recreational purposes. Improvements in water quality such as decreased turbidity
will also favor increased recreational activities such as swimming, boating, nature observation, fishing,
camping and other outings, as well as overall aesthetic enjoyment. Improvements associated with reduced
nitrogen and phosphorus discharges include reduced eutrophication of surface waters, fewer beach
closings, greater fisheries productivity, and greater enjoyment of water resources. Finally, EPA expects
that the regulation will augment nonuse values (values that do not depend on use by humans, see Chapter
5 for details) of the affected water resources. EPA used a meta-analysis of surface water valuation studies
to estimate the total value of nonmarket benefits (values that arise outside of market transactions, such as
recreation at publicly accessible sites) stemming from the regulation. EPA estimates that mean total
willingness to pay (WTP) for water quality improvements resulting from regulation ranges from $210
million to $413 million, depending on the regulatory option under consideration. Chapter 10 of this report
provides details on EPA's analysis of nonmarket benefits. EPA expects that nonmarket benefits resulting
from Options 1, 2, 3, and 4 will be approximately $210, $353, $413, and $361 million per year,
respectively. Combining these nonmarket benefits estimates with the avoided cost estimate produces total
benefits estimates of $214 million per year for Option 1, $360 million per year for Option 2, $422 million
per year for Option 3, and $369 million per year for Option 4.
Sufficient data were available to monetize benefits for only a subset of:
        >  pollutants discharging from construction sites
        >  impacted surface waters
        >  ecological services from surface waters.
November 2009                                                                                  1-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
The monetized benefits estimates therefore represent only a portion of the total benefits of each of the
regulatory options. The scope of the monetized benefits analysis is discussed in more detail in Chapter
11.
November 2009                                                                                    1-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
2      Overview of Environmental Impacts from Construction  Site
Sediment and Turbidity Discharges
This chapter summarizes information available from the literature and other sources on the pollutants and
environmental impacts associated with construction site discharges of sediment.
Construction sites have been documented to increase pollutant discharges to surface waters in a number
of studies. Sediment and turbidity are the most thoroughly documented pollutants. This chapter discusses
the process of sediment and turbidity discharge from construction sites, their behavior in surface waters,
and their potential impacts on aquatic organisms and human use of aquatic resources. Documentation of
these discharges and impacts can be found in the literature reviewed and summarized in Chapter 4. A
number of other pollutants can also discharge from construction sites including nitrogen, phosphorus,
metals, toxic organic compounds, and others. These are discussed in Chapter 3.
Suspended and bedded sediments (collectively referred to as "SABS" or "sediment" in this document)
and turbidity are natural components of many aquatic ecosystems. Sediments contribute to the physical
structure of surface waters and help to transport nutrients and organic matter.  Natural turbidity can
modulate levels of aquatic photosynthetic activity and predator-prey relationships. Undisturbed
ecosystems contain species adapted to the sediment and turbidity levels naturally associated with those
ecosystems.
At excessive levels, however, sediment and turbidity become pollutants.  Modifications to the physical and
chemical composition of sediment and turbidity discharges can also transform them into pollutants.
Sediment and turbidity are the most commonly documented pollutants in construction site discharges and
impacted surface waters. In a number of documented cases, sites have discharged sediment and turbidity
at very high  levels (see Chapter 4).
Although suspended sediment, bedded sediment, and turbidity are distinct and separate water quality
properties, all describe impacts associated with eroded soil  discharge to  surface waters. For this reason,
many studies discuss these properties concurrently, as does this document. Soil and sediment are
composed of a variety of components including organic matter, phosphorus, nitrogen, metals, and other
compounds, both natural and anthropogenic. Many of these components travel with soil as it erodes and
discharges to surface waters. These components are discussed in more detail in Section 2.2.
There are multiple terms available to describe levels of sediment and turbidity in water. Table 2-1
presents terms and definitions used in this document.
November 2009
                                                                                           2-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 2-1: Sediment and Turbidity Terminology
   Sediment Metric	Description
 Bedded sediment      A general term for sediment that, at any given time, settles from the water column onto surfaces
                      within a surface water (e.g., channel bed or aquatic plant leaves) in a process commonly known as
                      sedimentation or siltation. Measures of bedded sediments include depth of deposition within a given
                      time period, percent fines, geometric mean diameter, Fredle number (a permeability index) (Berry et
	al. 2003), and others.	
 Settleable solids       A measure of the solids that will settle to the bottom of a cone-shaped container (called an Imhoff
                      cone) in a 60-minute period. Settleable solids are primarily a measure of particles that can be
	removed from water by sedimentation. Expressed as milliliters per liter (ml/L).	
 Suspended sediment   A general term for sediment that, at any given time, is either maintained in suspension by a surface
	water's turbulent currents or that exists in suspension as a colloid.	
 Suspended sediment   The velocity-weighted concentration of suspended sediment in the sampled zone in a surface water
 concentration (SSC)   defined as extending from the water's surface to a point approximately 0.3 feet above the bed. It is
                      determined by measuring the dry weight of all sediment from a known volume of sample. Expressed
	as milligrams of dry sediment per liter of water-sediment mixture (mg/L).	
 Total suspended       A dry weight measure of suspended inorganic and organic material in the water column. It is
 solids (TSS)           measured by filtering a subsample of water and measuring the weight of the dried solids. Expressed
	in milligrams of solids per liter of water-solids mixture (mg/L).	
 Turbidity            A measure of the scattering and absorption of light when it enters a water sample. The quantity of
                      suspended particles in water helps to determine turbidity levels as do particle shape, size,  and color
                      distributions. Suspended particles can include clay, silt, colloids, finely divided organic and
                      inorganic matter, soluble colored organic compounds, plankton, and other microscopic organisms. In
                      this document, turbidity levels are typically expressed in nephelometric turbidity units (NTUs).
	Higher NTU levels indicate more turbid water.	


Surface water turbidity levels are controlled by the quantity and nature of particulate matter suspended in
the water column. This particulate matter can consist of mineral particles  (sediment), algae and other
organisms, and organic detritus. Particle  shape, size, and color distributions influence total turbidity levels
as well. Suspended sediment contributes to surface water turbidity. However, because several factors
beyond the mass of suspended solids in the water column control surface  water turbidity, the quantitative
relationship between suspended particle concentrations and turbidity levels has been found to vary among
watersheds, surface waters, and precipitation events.

The sections below provide additional information on  the nature of construction site sediment and
turbidity discharges to surface waters  (Section 2.1), their behavior and transport in surface waters (Section
2.2), their impacts on aquatic ecosystems (Section 2.3), their impacts on human use of aquatic resources
(Section 2.4), appropriate levels in surface  waters as delineated by water quality criteria (Section 2.5), and
the extent to which they currently impair surface waters in the United States  (Section 2.6).


2.1    Sediment and Turbidity Discharge to Surface Waters

Construction activities increase soil vulnerability to erosion. Typical construction site activities  include
clearing vegetation and excavating, moving, and compacting earth and rock. Vegetation removal and
surface work loosens soil, removes protective root structures, and exposes soil directly to the erosive
powers of precipitation and stormwater runoff. Soil compaction reduces precipitation infiltration and
increases overland water flow, thereby increasing the quantity of stormwater runoff available to erode
soil. In addition, stockpiled construction  materials such as stripped topsoil, fill material, and soil from
foundation excavation are often placed in steep, uncovered  piles vulnerable to erosion. Construction
vehicles track soil onto roadways from which it can easily wash into storm sewer drainage systems and
subsequently to surface waters. Susceptibility to erosion remains high at construction sites until soil-
disturbing activities are complete and the land surface is revegetated or otherwise stabilized.
November 2009                                                                                         2-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
Precipitation events are the primary cause of construction site sediment and turbidity discharges to surface
waters. Raindrop impact energy and overland water flow detach soil particles from the land surface,
suspend them in surface flow, and transport them to other locations on the construction site or to
discharge points. Suspended soil particles in the stormwater flow create turbidity.
Sediment erosion rates are highly variable among sites and depend on a number of factors including site
topography (slope length, steepness, and shape), precipitation intensity and quantity (i.e., rainfall
erosivity), soil type  (particle size, credibility, land use (vegetation cover, erosion control practices), and
nature of the construction activity. Some phases of construction activity disturb soil more than others. For
a more detailed discussion of sediment and turbidity discharges from construction sites, see EPA's
Development Document for Final Effluent Guidelines and Standards for  the Construction and
Development Category (USEPA 2009b).
Mobilization of soil particles is dependent on many factors including soil particle size, soil cohesiveness,
and rainfall energy and duration. As flows grow larger and more powerful, they are able to transport
larger particles. Sand-sized and larger particles are more easily detached  from the soil surface because
they are generally not cohesive.  However, once mobilized, they more easily settle from stormwater runoff
because of their greater mass.  Smaller particles, such as clays, are generally harder to mobilize because
they are more likely to be cohesive. Once mobilized, however, individual smaller particles are more likely
to stay suspended in stormwater flow and be transported off-site because of their lesser mass (Reed 1980).
The soil fraction composed of smaller particles tends to contain a disproportionate quantity of the organic
matter and adsorbed materials and contaminants (e.g., metals, nutrients, pesticides, and other organic
compounds) found in soils. Smaller particles also contribute disproportionately to turbidity levels.
Suspended sediment and turbidity levels in construction site stormwater flows can be very high.
Suspended sediment concentrations up to 160,000 mg/L in construction site stormwater have been
documented (Daniel et al.  1979; Horner et al. 1990; Warner and Collins-Camargo 2001; Hedrick et al.
2006; USEPA 2009b). Turbidity levels up to 30,000 NTU have been documented (Lubliner and Golding
2005; Bhardwaj and McLaughlin 2008). These figures reflect suspended sediment and turbidity levels
prior to treatment on-site and/or prior to discharge to and dilution in surface waters.
In addition to elevated stormwater sediment and turbidity concentrations, stormwater runoff volume can
also increase on construction sites (Yorke and Davis 1972; Selbig et al. 2004; Clausen 2007; Line and
White 2007; Selbig  and Bannerman 2008; Montgomery  County DEP 2009) (see Chapter 4).
The path stormwater travels varies among construction sites. Stormwater can infiltrate soil, flow over
land, or travel through underground storm sewer systems. Vegetated areas can provide opportunities for
stormwater and pollutant capture on the land surface whereas paved surfaces tend to  efficiently transport
stormwater downslope. Storm sewer systems can also very efficiently move water away from
construction sites (and into surface waters). Storm sewer systems are often installed early in site
development when large areas of earth are still disturbed and highly prone to erosion (NRC 2008).
Distance to surface water also varies among construction sites. Some sites are far from surface waters,
whereas others are adjacent to or, in some cases, directly in water (e.g., bridge construction, stream
channelization). Longer travel paths prior to surface water discharge can  provide more opportunities for
stormwater and pollutant capture (Yorke and Herb 1978).
The combination of elevated sediment concentrations and elevated runoff volume results in higher total
export of sediment from a site. Erosion is a natural process; however, sediment yields from construction
sites can be many times higher than those from agricultural, forested, and mature developed sites.
November 2009                                                                                  2-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
Forested watersheds in the United States are estimated to yield 0.02 to 1 tons/acre/year of sediment.
Cropland in the United States is estimated to yield 0.65 to 15 tons/acre/year (Vice et al. 1969; Yorke and
Herb 1978; Osterkamp et al. 1998; Faucette et al. 2007). Paul and Meyer (2001) state that construction
activity generally increases sediment yields 100 to 10,000 times over those from forested areas, with
larger increases possible in basins with steep topography. Studies summarized in Table 4-2 report
sediment yields of 0.009 to 219 tons/acre/year from catchments containing construction. All studies of
construction site sediment yield report that construction activity increases sediment discharges from sites.
Because of the importance of precipitation and its subsequent flow over land to the transport of soil, most
sediment and turbidity discharges take place during or shortly after precipitation events. Once a
precipitation event ceases, discharges from construction sites generally cease within a relatively short
time period unless site terrain or stormwater management systems, such as stormwater ponds, attenuate
discharge flows. This  dynamic creates an intermittent discharge of pollutants from most construction
sites.
In addition to being an episodic discharger of pollutants, construction is a temporary activity at any one
site that typically lasts several months to several years. Individual construction sites are therefore transient
sources of pollutant discharges to surface waters. Sediment discharges from the land surface to surface
waters due to active construction at an individual site are greatly reduced once active construction ceases
and site soils have been stabilized (e.g., through revegetation or paving).
Surface water impacts such as elevated turbidity and suspended sediment levels generally cease
immediately downstream of construction sites soon after site discharges cease. Other impacts, however,
persist beyond the lifespan of individual precipitation events, individual construction sites, or even the
presence of construction activity in a watershed. This is due to the persistence of sediments and some
associated pollutants in surface waters as well as longer-term impacts on aquatic organism communities.
In addition, most construction activities in the United States are concentrated in a relatively small number
of watersheds during any single decade (see Chapter 6). Within these watersheds, the location of
individual construction sites changes from year to year, but the watershed's total annual construction
acreage remains elevated for a number of years until the watershed is "built out." Surface waters draining
these watersheds therefore experience a persistent, elevated  level of impact from construction activity for
several years, despite  the transient nature of individual sites  in the watershed.
Sediment and turbidity discharges can also take place during dry weather at some  construction sites.
Dewatering of site depressions and foundation excavations,  slope failure, attenuated drainage from
stormwater basins, construction equipment operation or other activity directly in or near surface waters
(e.g., channelization, pipeline crossing, culvert emplacement, bridge construction), vehicle and other
construction equipment washing, 6ndscape irrigation, and overland  flow from groundwater seeps can
create discharges during dry weather. Erosion of bare ground due to snowmelt in the spring can be high
(NRC 2008). Construction activity can also destabilize slopes under both dry and wet conditions. If
situated sufficiently close to a waterbody, slope failure can result in mass loading of sediment and
associated turbidity.
Requirements for construction site erosion and sediment control have gradually increased since the 1950s.
EPA, the U.S. Federal Highway Administration, the U.S. Federal Energy and Regulatory Commission,
the U.S. Department of Agriculture's Soil Conservation Service, states, counties, municipalities and other
entities have issued a variety of regulations and guidance. EPA has previously issued requirements for
construction sites under the National Pollutant Discharge Elimination System (NPDES)  stormwater
program. Phase I of the stormwater program was promulgated in 1990 and addresses large construction
November 2009                                                                                   2-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
activities disturbing 5 or more acres. Phase II of the stormwater program was promulgated in 1999 and
addresses construction sites disturbing 1 to 5 acres. EPA's Development Document for Final Effluent
Guidelines and Standards for the Construction and Development Category (USEPA 2009b) provides
additional information on the history of construction site erosion and sediment control requirements.
Under current practice, many construction site discharges are moderated to some degree by a variety of
sediment erosion and stormwater discharge control practices. Practices include the use of sedimentation
basins, silt fences, vegetative filter strips, grass swales, check dams, erosion control blankets, straw bale
barriers, gravel bag berms, sand bag barriers, straw mulch, hydraulic mulch, wood mulch, soil binders,
geotextiles and mats, rock filters, earth dikes and drainage swales, velocity dissipation devices,
sedimentation traps, and other methods. These practices and others are discussed in detail in EPA's
Development Document for Final Effluent Guidelines and Standards for the Construction and
Development Category (USEPA 2009b).

2.2    Sediment and Turbidity Behavior in Surface  Waters

Sediment and turbidity movement within surface waters is highly variable. Factors influencing sediment
and turbidity movement and persistence in surface waters include the intensity, quantity, and composition
of sediment and turbidity discharges and the nature of the receiving waters. Important waterbody
characteristics include type, size, and flow rate.
Sediment discharges from construction sites typically contain a large proportion of inorganic material that
can persist in surface waters for long periods of time. Under appropriate conditions (e.g., anaerobic),
organic material in construction site discharges can also persist. For these reasons, impacts from
construction site discharges can last beyond the life span of a single precipitation event, an individual
construction site, and of the presence of construction activity in a watershed (Yorke and Herb 1978).
Several researchers have stated that long-term monitoring beyond the cessation of a construction project
is necessary in some cases to fully document its impacts (Barton 1977; Taylor and Roff 1986; Chen et al.
2009; Lee et al. 2009).
While in the surface water network, sediments can cycle many times between suspension in the water
column, where they contribute to turbidity levels, and deposition on waterbody beds.  This process
depends on surface water currents and other disturbances. Larger sediment particles are more likely to
settle on surface beds, and smaller particles are more likely to remain suspended in the water column and
contribute to  surface water turbidity. Because of this dynamic, larger particles tend to take longer to move
through surface water systems than smaller particles.
Construction sites discharge sediment and turbidity to all major surface water categories: wetlands;
streams and rivers; lakes, reservoirs, ponds, and other impoundments; and estuaries, bays, and other
coastal waters. Some waterbodies have flow regimes and physical structures that allow them to purge or
absorb excess sediment and turbidity inputs within relatively short time frames. Other surface waters
allow excess  sediment and turbidity to persist for long periods of time. The ability of individual surface
waters to transport sediment varies as precipitation and other factors modify the nature of their flow.
Surface waters also vary widely in their size relative to construction site discharges. Larger surface waters
generally have more capacity for dilution of pollutant discharges. Sediment transport in several surface
water types are discussed in more detail in Section 2.2.
Sediments and turbidity migrate with surface water flow downstream and therefore affect waters beyond
the initial receiving water. Migration of construction sediment typically benefits the initial receiving water
November 2009                                                                                  2-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
but impacts downstream waters. Several studies have documented improvements in stream conditions
once construction site sediment and turbidity discharges cease and accumulated sediments migrate
downstream from the initial point of entry. The sediments, however, remain within the surface water
network and are redistributed to new waterbodies downstream. Because of sediment's ability to persist
and migrate downstream, surface waters downstream of a watershed containing a consistent level of
construction activity from year to year can experience a relatively consistent level of impact from
construction sediment and turbidity discharges, even though the timing of individual precipitation events
and the locations of individual construction sites within the watershed change from year to year.

2.2.1  Sediment Characteristics Affecting Surface Water  Transport

Important sediment characteristics that influence transport and fate include the size, shape, density, fall
velocity, and concentration of sediment particles (Shen  and Julien 1993).  Multiple size scales exist. Table
2-2 presents the broad size categories of one scale.
Table 2-2: Sediment Grade Scale
Category
Size range (mm)
Boulders,
Cobbles
64 - 4,000
Gravel
2-64
Sand
0.062-2
Silt
0.004-0.062
Clay
0.00024 - 0.004
Source: Julien (1995).
Sediment size is also loosely discussed in terms of "fine sediment" and "coarse sediment." Fine sediment
consists primarily of particles smaller than 0.85 mm. Coarse sediment consists primarily of particles from
0.85 to 9.5 mm (USEPA 2006b). Particles smaller than 1 micron (0.001 millimeter) in diameter are
sometimes referred to as colloids.
Sediment samples rarely contain particles of a single size. Particle size distributions describe the
percentage by weight of sediment materials in each size category  for a given sediment sample.
Particle size helps to determine how quickly a particle will settle out of suspension in the water column
onto the bed of a waterbody. The fall or settling velocity of sediment particles can be estimated using a
form of Stokes' law (Julien 1995; Chapra 1997):

                                                                                        (Eq. 2-1)
       Where:

       co  =   settling velocity (em's"1)

       a  =   dimensionless form factor accounting for geometry (a=l for sphere)

       g  =   gravitational acceleration (g=981 cm»s"2)

       ps  =   density of sediment particle (g»cm~3)

       pw  =   density of water (g» cm"3)
November 2009                                                                                  2-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
       /um =   dynamic viscosity (g'cm'^s"1)

       ds  =   particle diameter (cm)

This formula indicates that settling velocity increases as particle size increases, assuming all other factors
are equal. Stokes-type settling is most directly applicable to the behavior of sediment in detention basins
and some reservoirs and lakes, in which flow velocities are low and turbulent energy in the water column
is minimal. In streams and rivers, settling velocity operates in conjunction with turbulent energy affecting
particle suspension to determine sediment settling dynamics.  Since turbulent energy is required to offset
the force of gravity (settling) in order to keep a particle in suspension, a greater amount of turbulent
energy is required to suspend larger, heavier particles than for smaller particles.
The degree to which a particle remains suspended helps to determine the degree to which it will be able to
move with water flow within a surface water or whether it will be deposited on a surface water bed. It
also determines whether or not a sediment particle will contribute to surface water turbidity, since only
suspended particles increase turbidity levels. Particles smaller than 0.063 mm (silt and clay) tend to
remain suspended in flowing freshwater systems and are the primary contributors to water turbidity
(USEPA 2006b). However, small sediments can also be transported in the aquatic environment in the
form of aggregates bound together by living (e.g., bacteria and algae) and nonliving (e.g., organic detritus,
extra-cellular polymeric substances) material (Bilotta and Brazier 2008). These aggregates  have transport
behaviors closer to particles coarser than the particles composing the aggregate. Coarser particles are
more likely to settle from the water column, with the coarsest particles settling closest to their discharge
point to surface water.
The sections below provide additional information on transport dynamics associated with several surface
water types.

2.2.2  Sediment and Turbidity Behavior in Specific  Waterbody Types

2.2.2.1 Sediment and Turbidity in Streams and Rivers
A distinguishing characteristic of streams and rivers is their strong, primarily unidirectional flow. Their
transport of sediment is predominantly controlled by stream transport capacity and sediment
physiochemical characteristics and supply rate. Larger sediments generally experience more episodic
movement over longer time scales through watersheds. Smaller sediments generally move more
continuously and within a shorter time scale. This difference  is due to the fact that larger sediments rely
on larger, more powerful flows for transport, which occur episodically and less frequently than flows able
to move smaller particles.
Sediments transported by river and  stream channels are typically described as bed load and suspended
load. The boundary separating these categories changes with  flow conditions. Sediments not in transport
at a given time are often referred to as bedded sediments. If flow conditions change, these sediments can
become bed load or suspended load. Sediment particle size is the primary distinguishing factor among
these categories, with the largest particles deposited as bedded sediments, larger particles moving as bed
load, and finer particles moving as suspended load. Total sediment load consists of the sum of bed load
and suspended load.
Bed load consists of the movement  of particles along the streambed by rolling, sliding, and a hopping
process known as saltation which involves the brief suspension of sediment particles by turbulent flow.
Particles  moving as bed load often consist of sand and coarser size fractions. Bed load movement is
November 2009                                                                                  2-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
intermittent in most settings, and bed load particles may not move for long periods between high flow
events.
Suspended sediment load is the movement of sediment particles supported by turbulent motion in stream
or river flow and often consists primarily of clay and silt size particles. Suspended sediment (and the
turbidity it creates) move more quickly downstream with streamflow than bed load from a given
discharge point. If the suspended sediment input to the stream or river is a single pulse, rather than
continuous, the stream or river can often move it downstream within a relatively short time period.
Several hydraulic and geomorphologic factors determine stream transport capacity including channel
width, flow depth and cross-sectional geometry, bed slope and roughness, and discharge velocity and
volume. In general, the more turbulent energy available for suspension and mobilization of sediment, the
greater the sediment transport  capacity per unit of stream width and the larger the size of sediment
particles that can be moved. Several empirical formulas have been developed to estimate sediment
transport capacity on the basis of flow- and channel- related variables. One example, developed by
Simons et al. (1981), takes the form of a power law:

                                        qs=Cl-hc>-Vci                                (Eq.2-2)

       Where:

       qs       = sediment transport per unit stream area (ft2/s)

       h        = depth of flow in ft

        V       = flow velocity in ft/s

       ci, c2, c3  = empirical  coefficients calibrated to reflect channel slope (gradient) and median
                  particle size diameter.

The unit transport capacity is strongly related to flow velocity (V). Higher velocity flows are able to move
more sediment. Channels with lower stream gradients have, for a given volume of stream flow, lower
stream water velocities which  allow more time for settlement of sediment in that channel. Many, if not
most, of the hydraulic factors controlling sediment transport capacity, including channel geometry, depth,
slope, and velocity of flow, vary in time  and space within a given river system.
Many North American rivers and streams possess a strong seasonal discharge cycle with spring discharge
volumes typically many times  larger than those of late summer and autumn flows. Intense or prolonged
rainfall events can also generate flood pulses of hourly to daily duration, which often have significant
turbulent energy. In addition, as Equation 2-2 indicates, sediment transport capacity is a nonlinear
function of flow-related variables, so large flows have significantly greater transport energy. The
movement of sediments, both as bed load and as suspended load, is thus highly nonuniform in time for
most river systems. The majority of annual sediment flux, particularly the movement of coarse or highly
cohesive sediment particles, may occur over a relatively short period of time during a single flood event.
Between such events, sediments are typically stored within the stream or river channel. Episodic
movement and deposition in stream channels between periods of movement has been  documented for
sediments in watersheds  whose total sediment load has been substantially increased by construction
activity. Sediment inputs from construction sites can outpace the immediate transport capacity of the
stream and may not migrate downstream until a major flow event occurs  (Lee et al. 2009).


November 2009                                                                                2-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
Erosion and deposition of sediments within a river course also exhibit spatial patterns strongly related to
stream morphology. River reaches with smaller cross-sectional flow area, steeper slopes, and higher flow
velocities discourage the deposition of sediments. These traits tend to be characteristic of smaller streams
and rivers in upper elevation catchments, often at the headwaters of larger watersheds. These higher-
gradient streams may decrease their construction-associated sedimentation levels more quickly than other
aquatic ecosystems due to the occurrence of high flows, particularly in spring or after large storms, that
can resuspend and transport finer sediments downstream. Assuming cessation of additional sediment
input upstream, these stream dynamics may eventually restore a naturally coarse-grained channel bed
(Barton 1977; Berry et al. 2003), though some sediment may continue to persist in low-energy areas of
the stream (e.g., shallow side pools). By contrast, wider channels with lower bed slopes and flow
velocities act as regions of relative sediment deposition. Channel bottoms may be covered with finer
sediments, in contrast to the exposed rocks, boulders, and gravels seen in the channel beds of higher-
energy streams  and rivers. Natural sediment deposition is more characteristic of channels at lower
elevations in a watershed.
Stream and river hydraulic and geomorphologic variables provide one set of controls on sediment
transport capacity. Sediment transport is also be regulated by the rate and quality of sediment supply
(Julien 1995). Sediment supply can outpace, match, or fall below the ability of a channel to transport it.
Within a particular reach, sediment fluxes can originate from land surface erosion, streambank erosion,
upstream reach  sediment input, or remobilization of sediments previously deposited within the reach.
Channels whose sediment supplies outpace their transport capacity will accumulate sediments. The  size of
a channel can decrease as sediments accumulate, increasing the likelihood of flooding  and other  overbank
flow events. Channels with sediment supplies falling below transport capacity will work to mobilize
additional material from channel beds and banks.
In all streams, sediments are preferentially deposited in regions of low-energy flow, including pools and
the inside of bends (Chapra  1997). If sufficient quantities of sediment are deposited, the deposition
features can alter channel morphology and flow patterns, obstruct flow, and exacerbate flood events.
Increased sediment supply during construction activity has  converted some naturally meandering streams
to braided or straighter, more channelized forms (Paul and Meyer 2001). Fine sediments deposited on
stream and river beds may also impede water exchange with groundwater sources (both recharge and
discharge) (USFWS 1998). Sediment deposits can also provide substrate for the growth of plants in
channels in locations where  they would normally not occur. King and Ball (1964) and  Taylor and Roff
(1986) documented this effect  downstream of highway construction sites.
Individual sediment deposits are often not permanent features of streams and rivers, since they can be
scoured and moved downstream during major flow events.  Streams and rivers can also flow outside their
normal channels during major  flow events  and deposit sediments on low-lying areas adjacent to the
channel such as banks, floodplains, and terraces. These sediments may, at a later time, be remobilized
during an even larger flow event.
Lee et al. (2009) documented that watersheds undergoing construction activity may take many years or
even decades to move  sediments discharged from construction sites fully through watersheds. Sediment
yields from larger watersheds may therefore remain elevated for some time after the implementation of
enhanced sediment and erosion control measures and after the completion of most construction in the
watershed.
Within the time scale of a single precipitation event, a certain amount of time is also necessary to move
suspended sediment and turbidity from its original point of entry into the surface water network to
November 2009                                                                                 2-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
downstream waters. If construction is occurring primarily in the headwaters of a watershed, turbidity and
suspended sediment levels immediately downstream of construction sites will often decline within
relatively short periods of time following the cessation of precipitation events (assuming discharge from
the site is primarily precipitation-driven and not significantly attenuated). Turbidity and suspended
sediment levels further downstream, however, can remain elevated even after flow levels return to normal
because of the time required for their transport downstream.
An additional dynamic influencing short-term observation of impacts from upstream construction activity
is that levels of turbidity and sediment deriving from construction activity tend to decrease as they move
downstream from a construction site through the surface water network. This decrease is due to sediment
deposition during transport and dilution by waters containing lower levels of sediment and turbidity than
construction site  discharges (Wolman and Schick 1967; Lee et al. 2009).

2.2.2.2 Sediment and Turbidity in Lakes, Reservoirs, and Ponds
A number of lakes, ponds, and reservoirs have a unidirectional flow component and therefore have some
of the basic dynamics described for rivers and streams. However, most lakes, ponds, and reservoirs have
lower flow velocities, greater depth of flow, and longer water residence times than streams and rivers and
therefore act as deposition zones (sinks) for sediments. Longer flow residence times also mean that
influxes of turbid water can linger for longer periods of time than they would in a stream or river.
Residence times  vary widely among lakes, ponds, and reservoirs.
The lake and reservoir flow environment more closely approximates the still-water conditions under
which Stokes' law (Section 2.2.1} applies to sediment particles to predict settling velocity. The
assumption of Stokes-type settling is often used to predict the rate of sedimentation in lakes and
reservoirs by comparing the time required for a particle to settle with the particle's transit time through
the waterbody. Several surface water characteristics influence settling and transit times, including size,
shape, depth, and regulation of outflow (Chapra, 1997). These characteristics vary widely among lakes,
ponds, and reservoirs.
Particle transit time also varies with flow. Higher flows generally create shorter transit times as higher
volumes of water move more quickly through a surface water. Higher flows often transport both coarser
particles, which are more likely to settle in a lake, pond, or reservoir, and large quantities of fine
sediments, which may not have sufficient time to settle during a higher flow period. Sediments unable to
settle in lakes, ponds, and reservoirs due to insufficient settling time or turbulence from water flow, wind,
or human activity can continue their transport downstream.
Given sufficient  residence time for incoming flows, however, these waters can remove  significant
quantities of sediment and turbidity from incoming waters (Barton 1977). The deposition process
decreases sediment and turbidity loads to streams and rivers and  other surface waters downstream
(Renwick et al. 2005; Lee et al. 2009), though it increases sedimentation in the lake, reservoir, or pond.
Over time, sediment accumulation reduces a waterbody's volume and decreases  its capacity for sediment
removal.
Sediment deposition in lakes and reservoirs often begins with the formation of deltas at the point of water
inflow, where incoming stream flow decelerates and the heaviest particles settle  (Julien 1995).
Turbulence from wind, currents, and human activity can keep finer sediments in suspension, creating a
deposition pattern called focusing in which coarser particles are deposited in shallower waters and finer
particles in deeper waters, often in the centers of lakes (Chapra 1997). Given sufficient sediment
November 2009                                                                                2-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
deposition, ponds and shallow areas in lakes and reservoirs may be transformed into wetland
environments.

2.2.2.3 Sediment and Turbidity in Wetlands
Wetlands are natural sediment deposition zones. Water flow volumes and velocities are smaller than in
most other surface water types, allowing time for settling of sediment from the water column. Because of
lower flow-through volumes and speeds, influxes of turbid and sediment-laden water typically remain for
a longer period of time in a wetland than they do in a stream or river channel.
Wetland ecosystem structure is heavily influenced by the type of vegetation a wetland supports, which in
turn is influenced by water and sediment distribution in the wetland. Elevated sedimentation levels can
change the type of vegetation able to persist in a given wetland, with consequences for the organisms it no
longer supports. In severe sedimentation cases, excess sediment may partially or completely fill in a
wetland, creating dry land conditions.

2.2.2.4 Sediment and Turbidity in Estuaries
Estuaries are like lakes and reservoirs in that they vary widely in benthic geometries, residence times,
flushing rates, vertical mixing, stratification, wave exposure, and other factors that govern sediment
transport and deposition. Estuarine flows transition first from unidirectional turbulent channel flow
controlled primarily by topographic gradient and discharge rate to a tidal river reach zone in which
downstream flow is influenced by the tidal cycle. The flow regime transitions again in the estuary proper,
in which water discharge and tidal forces offset each other. Flow regime changes again in the bay or open
ocean where tidal, wave, or a combination of these forces dominate.
As turbidity and sediment-laden freshwater decelerates and encounters tidal cycles in the estuary, a null
zone is formed in which channel discharge and tidal action largely cancel each other, favoring the
deposition of suspended sediment (Chapra 1997). In addition, increasing salinity levels favor flocculation
of fine sediments. Large deltas may form where rivers deposit sediments in coastal waters. Sediment
deposited in estuaries can be disturbed and redistributed due to natural events (e.g., floods, high winds,
tidal action) or by human activity (e.g., dredging).

2.3   Aquatic Life Impacts of Sediment and Turbidity

Numerous aquatic ecosystem impacts from  construction site discharges have been documented (see
Chapter ¥).This section summarizes studies of aquatic organism and ecosystem impacts associated with
elevated sediment and turbidity levels in surface waters. EPA has reviewed available literature on the
biological effects of suspended and bedded  sediments (SABS) and turbidity. This review is not an
exhaustive summary of the available  literature, which is extensive, but instead provides an overview of
more commonly noted impacts.
A variety of organisms, including aquatic plants, invertebrates, amphibians, and fish, are affected by
elevated sediment and turbidity levels. High levels of sediment and turbidity affect aquatic ecosystems by
reducing photosynthetic activity, reducing food availability, burying habitat, and directly harming
organisms. Organisms may relocate,  sicken, or die. Organism loss can alter the composition of the aquatic
community.
Both the magnitude and duration of sediment and turbidity exposure are important. Organisms vary in
their ability to tolerate and recover from exposure. The proportion of organisms able to tolerate (or
November 2009                                                                                2-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
escape) periods of elevated sediment and turbidity levels can increase in impacted surface waters, while
the proportion of sensitive species declines.
The appropriate level of sediment or turbidity varies from waterbody to waterbody. Some aquatic
ecosystems contain organisms adapted to higher levels of sediment or turbidity (e.g., some large coastal
floodplain rivers), whereas other ecosystems contain organisms adapted to lower levels of sediment or
turbidity (e.g., alpine lakes, headwater streams). Smaller streams tend to have naturally clearer water,
particularly those in forested watersheds. Short-term increases in suspended sediment and turbidity levels
can naturally occur during spring thaws, storms, and other high flow events. However, even organisms
adapted to sediment or turbidity influx (whether episodic or constant) can be harmed if input levels rise
excessively and/or if their resiliency is taxed by other stressors.
A construction project can change natural sediment and turbidity dynamics by elevating sediment and
turbidity levels significantly beyond those associated with natural events, for longer periods of time, and
at times when  an aquatic ecosystem and its organisms  are unaccustomed to receiving such inflows (e.g.,
late summer low flow periods). Some waterbody types are better able to flush excess suspended sediment,
turbidity, and deposited sediment (e.g., high energy streams), whereas others may retain accumulated
sediments for years (e.g.,  lakes and wetlands) (see Section 2.2).
Sediment impacts have been researched through both laboratory dose-response and in-stream field
studies. Much  research has been done on the effects of sediment on fish (salmonids, in particular, because
of their economic importance and sensitivity). More studies have been conducted on suspended sediment
versus bedded sediment effects on organisms. Stream and coral reef habitats have been studied more
intensively than river, lake, and estuarine habitat, though available data indicate that biota sensitive to
elevated sediment levels exist in all of these environments. Many habitats with moderate and variable
levels of sediment also need additional research (Berry et al. 2003).
Though not addressed in detail in this discussion,  sediments and turbidity can also affect chemical, fungal,
and microbial  decomposition processes in aquatic ecosystems. Depending on the source and nature of the
sediments, sedimentation can change the organic content of surface water sediment, which can alter
fungal and microbial community activity. Rapid sediment deposition over an organic sediment layer can
also cause a shift to anaerobic conditions in the isolated strata and an accompanying shift in microbial
metabolism, methane generation, and nutrient dynamics (Tornblom and Bostrom 1995). In wetlands,
depressed levels of algal and microbial biomass and density can indicate chronically elevated
sedimentation  levels (USEPA 1995).
The studies cited in this section attribute observed impacts to sediment and turbidity discharges in
general, rather than to the specific source materials from which the sediments and turbidity derived. The
discussion in this section provides a qualitative summary of the types  of organism impacts associated with
elevated sediment and turbidity levels. Berry et al. (2003) discuss the possibilities, given available data,
for determining quantitative relationships (e.g., dose-response models) between sediment levels and
organism responses. For most organisms, there is  insufficient information currently available to create
complete dose-response models (Berry et al. 2003). A  model for clear water fish is discussed in Section
2.3.3.
The examples  below discuss effects of elevated sediment and turbidity levels acting on organisms in
isolation. However, many construction sites have been documented to discharge other pollutants in
addition to sediment and turbidity such as nitrogen, phosphorus, metals, and other organic and inorganic
compounds (see Chapter  3). A number of these other pollutants travel with sediment as it erodes and
moves through the surface water network. In addition, many construction sites discharge to surface waters
November 2009                                                                                 2-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
already impacted by stressors including pollutant discharges from other sources, invasive species, and
flow modification. Interactions among sediment, turbidity, and other stressors and their cumulative effects
on aquatic organisms have not been fully characterized (Berry et al. 2003).
Documented effects and relevant studies are organized in the discussion below by the type of aquatic
organism studied (primary producer, invertebrate, fish, and other wildlife) and by the type of organism
reaction. A section specific to threatened and endangered species follows this discussion (Section 2.3.5).

2.3.1   Primary  Producers

Primary producers, including algae, phytoplankton, submerged aquatic vegetation, and other macrophytes
(plants with large  leaves), are found in most aquatic ecosystems. Aquatic ecosystems vary in the degree to
which they derive energy and resources from primary producers. Some systems (e.g., small streams in
forested watersheds) derive a large proportion of their energy from external sources (e.g., leaf debris)
(Bilotta and Brazier 2008). Other systems are heavily dependent on the productivity of primary producers
(e.g., sea grass beds).
Primary producers can grow in a variety of forms: rooted to surface water substrate, free-floating, or
attached to rocks,  aquatic plants, or other structures. The term periphyton refers to the algae, small plants,
bacteria, microbes, and detritus attached to  submerged surfaces in aquatic ecosystems.
Primary producers transform sunlight into energy and provide food for other aquatic organisms. Aquatic
macrophytes and algae  influence water column chemistry, including dissolved oxygen, pH, and nutrient
levels; provide important habitat for many organisms; and reduce waterbody flow velocity. Loss of algae
and macrophytes due to elevated sediment and turbidity levels can compound these pollutants' effects
because fewer plant structures are available to slow water flow and facilitate settling of suspended
sediment from the water column. This dynamic can allow elevated suspended sediment and turbidity
levels to persist further downstream (Wood and Armitage 1997).
Turbidity is an important parameter for aquatic ecosystems because of its influence on the compensation
point (the depth at which carbon production from photosynthesis equals consumption through respiration
in aquatic plants). A decrease in a surface water's compensation point can translate into a decline in an
aquatic ecosystem's primary  producer community or a shift in its species composition. These changes can
affect the ecosystem's overall species composition, productivity, and health.
Impacts to primary producers have been noted in some studies of construction site impacts to surface
waters. Construction sites documented in these studies were discharging elevated levels of sediment and
turbidity. King and Ball (1964) documented a decline in primary productivity levels. Cline et al.  (1982)
noted an increase  in periphyton sediment content and a decline in periphyton abundance and algal
diversity.
Several ways in which turbidity and sediment affect primary producers are described below.

2.3.1.1  Light Reduction
The growth and distribution of macrophytes, algae, phytoplankton, and coral zooxanthellae can be limited
by excessive turbidity (Berry et al. 2003). Macrophytes with leaves that emerge above the water's surface
are much less affected by reduced light penetration than submerged species. Reduced primary production
due to  light limitation from elevated turbidity is a common impact that has been documented by
LaPerriere et al. (1983), Rivier and Seguier (1985), and Lloyd (1987) (all as cited in Waters 1995), Meyer
and Heritage (1941) and Edwards (1969) (both as cited in Kerr 1995).
November 2009                                                                                2-13

-------
Environmental Impact and Benefits Assessment for the C&D Category
LaPerriere et al. (1983) found that moderate turbidity levels of approximately 170 NTU reduced oxygen
production in a stream by 0.08-0.64 g/m2 per day relative to clear water conditions. Above 1,000 NTU,
primary production was undetectable. Lloyd (1987) developed a model relating turbidity and primary
production. The model indicated that in a clear, shallow stream, a turbidity increase of 5 NTU decreased
primary production 3 to 13 percent. A 25 NTU increase reduced primary production up to 50 percent. The
study documented stronger effects from turbidity in lakes due to greater water depths and the greater
importance of the phytoplankton community. Quinn et al. (1992,  as cited in Bilotta and Brazier 2008)
found a 40 percent reduction in phytoplankton biomass exposed to suspended sediment concentrations of
lOmg/L for 56 days.
Guenther and Bozelli (2004) investigated whether lower light levels or increased sinking of
phytoplankton due to binding to sediment particles was the cause of their reduced photosynthetic activity
in turbid waters. The authors identified lower light levels as the primary cause.
Stony corals typically live in clear, oligotrophic waters where their symbiotic association with
photosynthetic zooxanthellae provides more than 100 percent of the coral's daily metabolic carbon
requirements (Muscatine et al. 1981; Grottoli et al. 2006). Corals exposed to elevated turbidity levels have
reduced photosynthesis rates and must compensate for this loss of resources (Philipp and Fabricius 2003).
Sediments can also reduce photosynthetic activity by settling on and coating macrophytes and periphyton.
Even a thin layer of sediment can block enough light to inhibit the growth of algae attached to surface
water substrates (Welsh and Ollivier 1998). Lower primary production by macrophytes and periphyton
can result in their decline and replacement with suspended algal communities. A shift in primary producer
community composition can alter a surface water's invertebrate and fish community composition.

2.3.1.2  Direct Physical Damage
Algae and aquatic macrophytes can suffer physical damage from elevated suspended sediment levels
(Waters 1995). Lewis (1973, as cited in Kerr 1995) conducted an experiment in which suspended coal
particle concentrations greater than 100 mg/L caused severe abrasive damage to the leaves of the aquatic
moss Eurhynchium riparioides over a period of 3 weeks. The damage substantially lowered the moss's
ability to produce chlorophyll and photosynthesize. Periphyton is also susceptible to being scoured from
stream substrate by suspended sediment particles (Welsh and  Ollivier 1998). Francouer and Biggs (2006)
investigated the interaction of flow velocity and suspended solids in benthic algal communities and found
that high suspended sediment concentrations further increased algae removal above that due to flow
alone. Some taxa were more susceptible to removal than others. The results indicated that suspended
sediment scour may be an important mechanism for algae removal during flood events and that some
variability in biomass removal among flood events and benthic algal composition may be the result of
differences  in suspended sediment load. Birkett et al. (2007, as cited in Bilotta and Brazier 2008) found
that suspended sediment levels of 100 mg/L stimulated periphyton growth and filament length under low
flow conditions, but levels of 200 mg/L significantly reduced biomass and filament length.
Macrophytes, algae, and periphyton can also be buried and damaged or killed by excessive sedimentation,
with different species having varying abilities to cope  (Berry et al. 2003). Some organisms can outgrow
sediment damage, though this growth may consume additional organism resources. A reduction or
complete elimination of periphytic material and a reduction in component diversity due to elevated
sediment levels is described by Van Nieuwenhuyse and LaPerriere (1986), Pain (1987) (both as cited in
Waters (1995), and Samsel (1973, as cited in Kerr 1995).
November 2009                                                                               2-14

-------
Environmental Impact and Benefits Assessment for the C&D Category
2.3.1.3 Substrate Modification
Sediment deposition in surface waters can change the types of algae (Wood and Armitage 1997) and
macrophytes able to grow in a waterbody by modifying nutrient availability and substrate chemistry,
texture, stability, and, depth. In many waterbodies, the locations of macrophyte beds are highly correlated
with the spatial distribution of fine-grained sediments since these often provide better rooting substrate.
Sediment deltas that form rapidly as a result of sediment runoff to a waterbody are often susceptible to
colonization by nuisance invasive macrophyte species.
Some established plants respond to elevated sedimentation levels by extending rhizome growth upwards,
but such growth patterns are abnormal for some species and consume plant energy (USFWS 1998). In
some cases, plants may not be able to grow quickly enough to adjust to increased sedimentation. Over
time, substrate may change to the extent that it no longer supports the original primary producer
community. Wetlands are particularly vulnerable to changes in  sedimentation rate.

2.3.2  Invertebrates

Aquatic invertebrates, including zooplankton, insects, crustaceans, and bivalves (e.g., mussels and clams),
are widely distributed among aquatic ecosystems and can be found in even the smallest surface waters.
They provide an important link in the aquatic food chain between primary producers and larger organisms
such as amphibians, reptiles, fish, birds, and mammals (Waters  1995).
Because many benthic macroinvertebrates are relatively stationary and have limited powers to evade
polluted waters, their abundance, diversity, and species composition is widely used as an indicator for
overall aquatic ecosystem health (USEPA 2006d). The effects of elevated sediment and turbidity levels
on invertebrates, particularly on benthic macroinvertebrates, have been the focus of a large number of
studies over the past several decades.
Elevated sediment and turbidity levels impact invertebrates by reducing their health, disease and
parasitism resistance, and abundance, as documented in literature reviews conducted by Kerr (1995),
Waters (1995), and Henley et al. (2000). The density, diversity, and structure of aquatic invertebrate
communities can change as a consequence, to contain a higher proportion of sediment-tolerant species.
Gammon (1970, as cited in Kerr 1995) finds that TSS concentrations between 40 and 120 mg/L reduced
macroinvertebrate density by 25 percent, and greater TSS concentrations  reduced density by 60 percent.
Organisms  intolerant of turbid waters include mayflies and other Ephemeroptera, stoneflies (Plecoptera),
caddisflies  (Trichoptera), clam, and bryozoan species (Cooper 1987, as cited in Kerr 1995).
The severity of sediment and turbidity impacts on invertebrate populations is related to the intensity and
duration of exposure (Anderson et al. 1996, as cited in USEPA  2004a). Some populations are adapted to
the short-term increases in suspended sediment and turbidity levels that can occur during spring thaws,
storms, and other natural high flow events. However, high or sustained elevated levels of sediment and
turbidity can alter long-term community structure by degrading habitat, food sources, organism health,
and increasing "drift" (i.e., voluntary/involuntary release of substrate by aquatic macroinvertebrates with
subsequent transport downstream)
If disturbance to a macroinvertebrate community is intense enough, it may impact the ability of the
community to recover its former condition, particularly if sensitive organisms are eliminated from
surrounding areas, as well, or habitat has been modified to the point where it is no longer able to support
the former community (Montgomery County DEP 2009). Healthy invertebrate populations in unaffected
November 2009                                                                                 2-15

-------
Environmental Impact and Benefits Assessment for the C&D Category
parts of a surface water system can provide organisms to recolonize disturbed areas once construction
activity has ceased.
Multiple studies have documented impacts to invertebrates from construction site sediment and turbidity
discharges (King and Ball 1964; Peterson andNyquist 1972; Barton 1977; Reed 1977; Chisholm and
Downs 1978; Extence 1978; Lenat et al. 1981; Tsui and McCart 1981; Cline et al. 1982; Taylor and Roff
1986; Young and Mackie 1991; Ohio EPA 1996a, 1997f, 1998d, 1999d, 2003a, 2006b; Reid and
Anderson 1999; Fossati et al. 2001; Levine et al. 2003, 2005; Hedrick et al. 2007; Chen et al. 2009;
Montgomery County DEP 2009).
Several categories of impact identified by Kerr (1995) and Wood and Armitage (1997) are described
below.

2.3.2.1 Loss of Habitat
Increased sediment and turbidity levels in waterbodies can degrade aquatic invertebrate habitat, including
pools and interstitial spaces in waterbody substrate and wetland areas. Habitat alteration due to changes in
substrate composition can modify the distribution and density of different invertebrate species and
therefore the structure of an invertebrate community (Waters 1995; Berry et al. 2003). Sediment settles
from the water column onto the substrate to which sedentary organisms attach and also fills small crevices
in waterbody beds (interstitial spaces) where a variety of species shelter, reproduce, feed, and seek
protection from predators, high flow velocities, and low flow conditions. Large accumulations of
sediment are also anoxic at depth. When interstitial spaces fill with sediment, the microhabitat diversity of
the substrate significantly declines, as does the diversity  of organisms it can support (USEPA 2006d).
Embeddedness refers to the extent to which rocks or organic debris in a surface water are covered or
sunken into fine bottom sediments and is an important physical component of stream or river habitat
(USEPA 1999). Increased embeddedness due to sedimentation indicates habitat degradation. Studies have
found high correlations between benthic invertebrate  response to depth and degree of embeddedness
(Berry et al.  2003). Correlations have also been found between benthic invertebrate abundance and
substrate particle size (Waters 1995). Even small increases in sedimentation levels can impact caddisfly
pupa survival (Berry et al. 2003). Ryan (1991, as cited in Henley et al. 2000), concludes that as little as a
12 to  17 percent increase in interstitial fine sediment could result in  a 16 to 40 percent decrease in
macroinvertebrate abundance.
Siltation of previously coarse substrates is also implicated in the reduction of mussel populations by Ellis
(1936) and Marking and Bills (1980) (both as cited in Waters 1995). Very thin veneers of sediment can
decrease the settlement and recruitment of some bivalve  larvae (Berry et al. 2003). One study postulates
that many of the more than 30 extinct species of North American mussels became extinct due to a loss of
habitat, most likely from sedimentation of surface water  substrate (Henley et al. 2000). In coral reef
systems, most coral larvae seek firm substrates and will not settle and establish themselves on shifting
sediments (Berry et al. 2003).
Invertebrates can also be indirectly affected by impacts to primary producers in an aquatic community. A
shift in primary producers can impact the structure of invertebrate community by changing food and
shelter availability and quality. As discussed in the previous section, increased turbidity and sediment
levels can adversely affect the primary producer community and change its abundance and composition.
As turbidity  increases, the contribution of primary production from aquatic macrophytes may decline,
while that associated with phytoplankton comes to dominate the system. Sediments and turbidity can
adversely affect the growth, abundance, and species composition of the periphytic algal community,
November 2009                                                                                 2-16

-------
Environmental Impact and Benefits Assessment for the C&D Category
reducing food availability for invertebrate grazers that feed predominantly on periphyton. A shift in
primary producers can impact the structure of the invertebrate community by changing food and shelter
availability and quality. This can lead to shifts in the relative proportion of various feeding guilds and
habitat preferences in the resulting aquatic invertebrate community (Vannote et al. 1980; Merritt and
Cummins 1996).

2.3.2.2 Altered Movement and Increased Drifting
Movement upstream or downstream ("drift") of benthic macroinvertebrates has been found to increase
with elevated suspended sediment and turbidity concentrations (Kerr 1995; Berry et al. 2003). Increases
in invertebrate drift have been shown to be  specifically related to the turbidity caused by suspended
sediments (Waters 1995). One study found that suspended sediment levels of 120 mg/L were sufficient to
cause drift (Berry et al. 2003). Rosenberg and Wiens (1978, as cited in Bilotta and Brazier 2008) found
increased benthic macroinvertebrate drift after 2.5 hours of exposure to 8 mg/L suspended sediment.
Increased drift leads to increased predation on the displaced benthic invertebrates (Waters 1995) and
causes a decline in invertebrate abundance in the impacted surface water (White and Gammon 1977, as
cited in Waters 1995). Benthic macroinvertebrate drifting downstream of construction sites for pipeline
stream crossings has been documented in several studies (Tsui and McCart 1981; Reid and Anderson
1999).

2.3.2.3 Physiological Changes
The physical effects of elevated sediment levels on invertebrate species can be severe. Sediment particles
can clog an organism's filter feeding apparatus, filtration system, respiratory system, and digestive system
(Kerr 1995), resulting in compromised health or death of the organism. Affected organisms include taxa
with fragile or rudimentary gill structures, such as stoneflies (Merritt and Cummins  1996) and organisms
that primarily graze in the water column (e.g., zooplankton). Alabaster and Lloyd (1982, as cited in
Bilotta and Brazier 2008) found that Cladocera and Copepoda gills and guts were clogged after 72 hours
of exposure to suspended sediment levels of 300 to 500 mg/L.
Most species of coral are highly intolerant of sedimentation and will expend energy to flush excess
sediments. If a coral is unable to compensate for excess sedimentation, tissues can smother and die.
Sediments can also reduce zooxanthellae photosynthesis, reduce growth rates, cause temporary bleaching,
and eventually death. These impacts can affect other parts of the coral reef food web. Consequently,
increasing sedimentation levels are associated with reductions in coral reef diversity (Rogers 1990),
alterations in community composition by growth morphology (Rogers 1990, Stafford-Smith and Ormond
1992), decreased skeletal extension rates  (Dodge et al. 1974, Miller and Cruise 1995, Heiss 1996), and
reduced coral recruitment (Rogers 1990, McClanahan  et al. 2002, Fabricius et al. 2003).  Other coral reef
organisms, such as sponges, are thought to be sensitive to sedimentation as well (Berry et al. 2003).

2.3.2.4 Impaired Feeding Activity
Another effect of elevated suspended sediment and turbidity levels on aquatic invertebrates is a reduction
in their feeding activity or the efficiency of their feeding. Kerr (1995)  noted that this impact is
predominant in filter feeding species such as mussels and clams and cited Ellis (1936), who found that
under turbid conditions, mussels and clams reject silt-laden food and, under highly turbid conditions,
close their shells completely. Waters (1995) also discussed sediment's impact on feeding, citing Ellis
(1936) and Aldridge et al. (1987), who found that sediment additions impaired feeding in three species of
bivalves. Reactions to elevated sediment levels by freshwater mussels may be species specific. Mussels
compensate for increased suspended sediment levels by increasing filtration rates, increasing the
November 2009                                                                                 2-17

-------
Environmental Impact and Benefits Assessment for the C&D Category
proportion of filtered material that is rejected, and increasing the selection efficiency for organic matter.
This process increases the metabolic cost of feeding per unit of food and decreases resources available for
growth and reproduction.
Species that have evolved in turbid environments may be better able to select between organic and
inorganic particles during feeding. For example, one study found that a mussel from rocky coastal
environments (Mytilus californianus) was more sensitive to elevated suspended sediment levels than a
mussel of the same genus from a bay environment (Mytilus edulis), which is typically more turbid (Berry
et al., 2003). Many endangered freshwater mussel species evolved in fast flowing streams with
historically low levels of suspended sediment. Compared to other species that evolved in more turbid
environments, such rocky intertidal habitat species may not be able to differentiate between organic and
inorganic particles as well.
Light attenuation due to elevated turbidity can reduce feeding efficiency of invertebrates. Copepods and
daphnids have been found to reduce feeding activity in response to elevated suspended sediment levels
(Berry et al. 2003). Sediment coatings on macrophytes and periphyton can reduce their quality as food
sources because herbivores must consume elevated levels of sediment with the plant material (Ryan
1991).
Kerr (1995) cited three studies (Paffenhofer 1972; Appleby and Scarratt 1989; Kirk 1992) that found
reduced growth rates in invertebrates subjected to elevated sediment levels due, potentially, to reduced
feeding activity and/or reduced food value. Graham (1990, as cited in Waters 1995) finds that a decrease
in the percentage of organic matter in periphytic material may result from additional sediment particles
binding to the periphyton.

2.3.2.5 Direct Mortality
Sediment can directly or indirectly increase invertebrate mortality  levels. Forbes et al.  (1981, as cited in
Kerr 1995) found that elevated suspended sediment levels increased amphipod (shrimp-like crustacean)
mortality. Henley et al. (2000) concluded from a literature  review that increased sediment concentrations
have a negative effect on the survival rates of freshwater mussels with the magnitude of the impact being
species specific. Robertson (1957, as cited in Bilotta and Brazier 2008) found the survival and
reproduction of Cladocera to be harmed by 72 hours of exposure to suspended sediment concentrations of
82 to 392 mg/L.
Organisms can also be buried  and  smothered by heavy sedimentation. Large sediment accumulations
become anoxic at depth. Species vary in their ability to evade burial under moderate sedimentation
deposits.  Sediment grain size, depth, and bulk density and species  motility, living position, and tolerance
of anoxic conditions during burial  affect this ability (Berry et al. 2003).
Sediment organic matter is an important food source for many benthic macroinvertebrates. A number of
toxic pollutants bind to and travel with sediment and can be ingested by invertebrates with the organic
matter (Paul and Meyer 2001). The aquatic organism effects associated with these toxic pollutants are
discussed in Chapter 3.

2.3.3  Fish

A substantial amount of research has been conducted to identify and quantify the effects of elevated
sediment and turbidity levels on a  variety offish species, more so than other taxonomic groups (Berry et
al. 2003). Salmonids, in particular, have been well studied  because of their commercial and recreational
November 2009                                                                                 2-18

-------
Environmental Impact and Benefits Assessment for the C&D Category
importance and because of the concerns and well-documented impacts arising from logging in the Pacific
Northwest and elsewhere (Waters 1995).
The level of impact from  suspended sediments and turbidity is a function of the interaction of many
factors, including natural  sediment and turbidity levels for the area in question; pollutant concentration;
exposure duration; particle size; ambient temperature; physical and chemical properties of the sediments;
associated toxins; co-occurring stressors; and fish acclimatization, life history stage, migration, and
breeding season (Anderson et al. 1996, as cited in USEPA 2004a). In general, longer duration and higher
level exposures produce greater effects. Milder, primarily behavioral effects are observed at lower
magnitude and duration exposure levels. As exposure and duration levels increase, effects become
sublethal and lethal (Berry et al. 2003). Several researchers have sought to construct models linking
varying levels and durations of sediment exposure to fish responses, but responses vary widely among
different species, and data are generally insufficient  at this time to fully characterize many of them (Berry
et al. 2003).
Newcombe (2003) has created a model for clear water fishes that relates magnitude and duration of
exposure to turbid water to a variety of adverse effects. The model indicates that fish exposure to turbidity
levels of 7 to  1,100 NTU  for 1 to 48 hours can create effects ranging from small behavioral impacts to
death of a portion of the exposed population. This model does not describe fish that typically  inhabit more
turbid surface waters and that exhibit a different set of reactions to a given range of turbidity levels and
exposures.
Short-term increases in suspended sediment and turbidity levels can naturally occur during spring thaws,
storms, and other natural events. Aquatic ecosystems also vary in their natural sediment and turbidity
levels.  Studies have found that, in general, fish typically found in environments with naturally high
turbidity preferred more turbid environments in the laboratory (Berry et al. 2003). Other fish species are
very sensitive to elevated sediment levels (e.g., entire salmonid fisheries have been destroyed by elevated
sediment levels) (Berry et al. 2003).
A construction project may produce sediment and turbidity levels significantly higher those associated
with natural events, for longer periods of time, and at times when an  aquatic ecosystem is unaccustomed
to receiving such sediment inputs. Many fish species have seasonal migration and breeding behaviors
which can be disrupted by elevated  sediment levels.  Younger fish are more vulnerable to sediment
impacts and are more prevalent within a population at certain times of the year (Berry et al. 2003).
Observed impacts from elevated sediment and turbidity levels fall into several broad categories, discussed
below. The potential cumulative effect of these impacts includes reduced disease and parasite resistance,
reduced growth, and degraded health of individual organisms in the fish community. These impacts may
decrease fish  population levels in affected areas. Reductions can take place both through direct mortality
in the short term and reduced reproductive success in the long term. Newcombe (1994) and Newcombe
and Jensen (1996) (both as cited in Henley et al. 2000) found that elevated sediment concentrations are
associated with increased mortality  in at least 14 species  offish.  EPA (USEPA 2004a) states that
suspended sediment concentrations  between 500 and 6,000 mg/L can result in significant (greater than 50
percent) mortality. Berkman and Rabeni (1987, as cited in Wood and Armitage 1997) noticed a decline in
the overall abundance offish stocks as sediment levels increased in a river in northeastern Missouri. The
overall impact of these effects can be a change in fish community composition to one composed
predominantly of species  that are tolerant of sediment and turbidity, primarily bottom-feeders (Berry et al.
2003),  or an overall reduction in fish abundance.
November 2009                                                                                 2-19

-------
Environmental Impact and Benefits Assessment for the C&D Category
Impacts to fish communities downstream of construction sites discharging elevated levels of sediment and
turbidity have been documented in multiple studies (King and Ball 1964, Graves and Burns 1970, Barton
1977, Garton 1977, Reed 1977, Werner 1983, Taylor and Roff 1986, Ohio EPA 1997f, Ohio EPA 1999c,
Reid and Anderson 1999, Hunt and Grow 2001, California Department of Fish and Game 2004,
Montgomery County DEP 2009).

2.3.3.1  Avoidance and Other Behavioral Responses
The severity of behavioral responses is associated with the timing of disturbance, the level of stress, fish
energy reserves, phagocytes, metabolic depletion, seasonal variation, and habitat alteration (USEPA
2006b).
Because fish are generally more mobile than aquatic plants, plankton, and invertebrates, some species are
capable of actively avoiding waters with high levels of suspended sediment or turbidity if migration
routes to alternative habitats are available (mobility may be limited by culverts, small dams, and other
obstacles). While larger, nonlarval fish are generally capable of avoidance behavior (Berry et al. 2003),
smaller and younger fish are generally less mobile and therefore more vulnerable to elevated sediment
and turbidity levels. Avoidance behaviors may reduce or eliminate fish populations in stream reaches with
sustained elevated sediment levels. Barton (1977) noted a relocation offish populations away from areas
with elevated TSS levels due to construction near a small stream. Servizi and Martens (1987, as cited in
Reid et al. 2003) determined that turbidity levels in excess of 37 NTU elicit avoidance behavior among
fish populations. Berg (1982, as cited in Bash et al. 2001) documented that juvenile coho salmon exposed
to a short-term pulse of 60 NTU water left the water column and congregated at the bottom of the test
tank, returning to the water column once turbidity was reduced to 20 NTU. Boubee et al.  (1997, as cited
in Newcombe 2003) found inhibition of upstream migration of one species at turbidity  levels of 15 to 20
NTU. Because of avoidance behaviors, some fish may be excluded from otherwise desirable habitat due
to increased turbidity (Berry et al. 2003).
Other fish behavioral responses to elevated sediment and turbidity levels include increased frequency of
the cough reflex and temporary disruption of territoriality (USEPA 2006b).

2.3.3.2  Feeding and Hunting
Elevated suspended sediment and turbidity levels have been shown to reduce feeding rates in several
species offish (Kerr 1995). Elevated turbidity reduces the prey reaction distance for trout and therefore
reduces foraging success (Hedrick et al. 2006). This effect is generally believed to be a result of decreased
visibility caused by turbidity because many fish use vision to locate food. De Robertis et al. (2003) found
that turbidity level increases affected piscivorous fish feeding more than planktivorous fish feeding.
Piscivorous fish visually locate food  sources at much longer distances than planktivorous fish do. EPA
(USEPA 2004a) concluded that turbidities greater than 25 NTU or TSS levels of 2,000-3,000 mg/L or
greater are sufficient to reduce fish predation abilities. Sweka and Hartman (2001) observed that brook
trout prey reaction distance decreased curvilinearly as turbidities levels increased from 0 to 43 NTU.
Although most fish species' predation abilities are reduced by increased turbidity levels, some fish
species have been documented to hunt better as turbidity increases because of increased contrast between
the prey and surrounding water.  This advantage deteriorates as turbidity levels further increase, however
(Berry et al. 2003). Some larval fish species (e.g., striped bass -Morone saxatilis) appear to be able to
feed under extremely turbid conditions (Berry et al. 2003).
Reduction in predation can benefit prey species able to use turbid water to hide from predators (Rowe et
al. 2003). However, a paper by the Canadian DFO (2000) suggests that as turbidity levels continue to
November 2009                                                                                2-20

-------
Environmental Impact and Benefits Assessment for the C&D Category
increase, the advantage of reduced predation is outweighed by the prey species' own increased difficulty
in locating food.
An additional effect on feeding and hunting from elevated sediment and turbidity levels is decreased food
availability and quality due to invertebrate community impacts, decreased primary productivity, and
sediment coatings on macrophytes and periphyton, which increase the quantity of nonnutritive sediment
herbivores must consume (Waters 1995; Ryan 1991). Bottom-feeding fish can be adversely impacted if
sediment organic content and associated microbial and macroinvertebrate communities are degraded by
sedimentation.

2.3.3.3  Breeding and Egg Survival
Elevated sedimentation levels can reduce spawning habitat for multiple fish species, particularly benthic
spawners. Many fish species use well-aerated interstitial spaces in surface water beds to lay eggs.
Sedimentation impact levels vary with differences in species sensitivity, life stage impacted, sediment
particle size, and sedimentation rates and total magnitude. Eggs and larvae can be buried too deeply by
sediment to survive (Berry et al. 2003). Sediments that settle onto surface water beds, particularly finer-
sized particles, can reduce the level of dissolved oxygen available to eggs deposited in the substrate (Kerr
1995). Even thin layers of sediment only a few millimeters thick can disrupt the normal exchange of gases
and metabolic wastes between eggs and the water column (Berry et al. 2003). Bjornn et al. (1977), as
cited in Kerr (1995), found that successful emergence offish fry from eggs began to decline when fine
sediment levels in the substrate were in excess of 20 to 30 percent. Sediment deposition can result in
surface water substrate armoring, trapping larvae as they attempt to emerge after hatching (Berry et al.
2003). Emergence success of cutthroat trout was reduced from 76 percent to 4 percent when fine sediment
was added to spawning gravel redds (Weaver and Fraley, 1993).
Correlations have also been found between elevated suspended sediment and turbidity levels and reduced
breeding and hatching success. Saunders and Smith (1965, as cited in Kerr 1995) found that increased
suspended sediment levels reduced fish spawning activity. Other studies document lower survival rates
for eggs spawned in surface waters with elevated suspended sediment and turbidity levels (Chapman
1988, as cited in Canadian DFO 2000). Reynolds (1989, as cited in Henley et al. 2000) found that
increases in turbidity levels increased sac fry mortality in arctic grayling (Thymallus arcticus).

2.3.3.4  Habitat Loss
Like invertebrates, fish use crevices in waterbody beds for feeding, shelter from predators and high flow
events, and reproduction. Loss of these interstitial spaces degrades fish habitat (USEPA 2006b). Sediment
can also decrease the depth of or eliminate other important habitats such as stream pools and wetlands
adjacent to surface waters. Riparian wetlands are particularly  important for fish spawning. Fish are also
affected by loss or decline in aquatic macrophytes, which are  important sources of shelter and food for a
number offish species. Coral reef bleaching and decline severely impact fish habitat in coral reef
ecosystems.

2.3.3.5  Juvenile Growth and Survival
High sediment concentrations can impact juvenile fish more severely than adult fish. Smallmouth bass
reduced growth after a 24-hour exposure to suspended sediment levels as low as 11.4 mg/L (Berry et al.
2003). Reynolds et al. (1989, as cited in  Bilotta and Brazier 2008) found 6 to 15 percent mortality of
arctic grayling sac fry after 24 hours of exposure  to suspended sediment levels of 25 to 65 mg/L.
Mortality of 41 percent was documented after 72 hours of exposure to suspended sediment levels of 185
mg/L. Juvenile coho and Chinook salmon exhibited abnormal surfacing behavior under elevated
November 2009                                                                                2-21

-------
Environmental Impact and Benefits Assessment for the C&D Category
suspended sediment concentrations, rendering them more vulnerable to avian predators (Canadian DFO
2000). This study also notes that excessive sediment levels can stress juvenile fish and consequently
increase their predation risk. These types of effects can reduce the strength of a year class (Berry et al.
2003).

2.3.3.6 Physical Damage
At high levels, suspended sediment, particularly more angular particles, are capable of causing physical
damage to fish by clogging gills (which impairs breathing) and by skin abrasion (Bell 1973, as cited in
Waters 1995). Kerr (1995) notes that fish can withstand short episodes of high sediment concentrations by
producing mucus to protect skin and gills, but this reaction stresses fish. Excessive gill mucus can cause
gill tissue traumatization and asphyxiation. The severity of damage appears to be related to sediment
dose, exposure duration, and particle size and angularity. Fish can release stress hormones (i.e., cortisol
and epinephrine) in response to decreased gill function. Reid et al. (2003) finds that elevated sediment
levels reduces the ability of rainbow trout to withstand hypoxic conditions and thereby reduces their
average life span.

2.3.4  Other Wildlife Dependent on Aquatic Ecosystems

Birds, mammals, reptiles, and amphibians that consume aquatic plants, invertebrates, fish, and other
aquatic organisms or otherwise utilize aquatic habitats for shelter and reproduction can also be affected by
elevated sediment and turbidity levels in surface waters. Some species are sufficiently mobile that they
can avoid impacted aquatic communities and seek substitutes, if available and accessible (Berry et al.
2003).
Welsh and Ollivier (1998) documented lower densities of two salamander species and one frog species in
streams impacted by sedimentation from a construction site. The authors postulated that sedimentation of
interstitial spaces in the stream's substrate was the cause because of the organisms' dependence on these
spaces. Interstitial spaces provide amphibians with shelter from predators and high flows and habitat for
food (e.g., diatoms and periphyton) production. Concerns have been raised about the potential impact of
sedimentation on the endangered Barton Springs salamander (Eurycea sosorum), which depends on a
coarse substrate and healthy aquatic macrophyte population (USFWS 2005).
Sedimentation can also affect or eliminate riparian wetlands, important habitats for the laying of
amphibian egg masses (USEPA 2006b).
There are few studies available on the effects of elevated sediment and turbidity levels in aquatic
ecosystems on fundamentally terrestrial wildlife that utilize aquatic ecosystems. Most available studies
examine effects on birds. Loons appear to need clear water for hunting fish  and may avoid turbid waters
for nesting. Other studies documented birds modifying their fish hunting behaviors and distribution on a
surface water in order to avoid more turbid areas, probably because of increased foraging difficulty (Berry
et al. 2003).

2.3.5  Threatened and Endangered Species

Threatened and endangered (T&E) and other special status species can be adversely affected by elevated
turbidity  and sediment levels. The multiple ways in which elevated sediment and turbidity levels impact a
variety of organisms are discussed above. These impacts can reduce organism growth, health, survival,
and reproduction, thereby leading to further decline in an impacted T&E species population. The potential
November 2009                                                                                 2-22

-------
Environmental Impact and Benefits Assessment for the C&D Category
further decline of listed species or critical habitat for these species from construction activity is
particularly important because these species are already rare and at risk of irreversible decline.
The federal trustees for T&E species are the Department of the Interior's U.S. Fish and Wildlife Service
(USFWS) and the Department of Commerce's National Marine Fisheries Service (NMFS). USFWS is
responsible for terrestrial and freshwater species (including plants) and migratory birds, and NMFS deals
with marine species and anadromous fish (USFWS 2008). Various state agencies and departments have
state-level jurisdiction over T&E species and habitats of concern.
A species is federally listed as "endangered" when it is likely to become extinct within the foreseeable
future throughout all or part of its range if no immediate action is taken to protect it. A species is listed as
"threatened" if it is likely to become endangered within the foreseeable future throughout all or most of
its range if no action is taken to protect it. The 1973 Endangered Species Act outlines detailed procedures
used by the Services to list a species, including listing criteria, public comment periods, hearings,
notifications, time limits for final action, and other related issues (USFWS 1996).
A species is considered to be federally threatened or endangered if one or more of the  following listing
criteria apply (USFWS 2007):
        >  The species' habitat or range is currently undergoing or is jeopardized by destruction,
           modification,  or curtailment.
        >  The species is overused for commercial, recreational, scientific, or educational purposes.
        >  The species' existence is vulnerable because of predation or disease.
        >  Current regulatory mechanisms do not provide adequate protection.
        >  The continued existence of a species is affected by other natural or manmade factors.
States and the federal government have also included species of "special concern" on their lists. These
species have been selected because they are (1) rare or endemic, (2) in  the process of being listed,
(3) being considered for listing in the future, (4) found in isolated and fragmented habitats, or
(5) considered a unique or irreplaceable state resource.
The federal government and individual states develop and maintain lists of species that are considered
endangered, threatened, or of special concern. The federal and state lists are not identical: a state does not
list a species that is on the federal list if it is extirpated in the state. States may also list a species that is
not on the federal list if the species is considered threatened or endangered at the state, but not federal,
level.
Information on federally listed T&E species is available from the USFWS Threatened and Endangered
Species System (TESS) database (USFWS 2009), available at
http://www.fws.gov/endangered/wildlife.html. Information on both federal and state listed species is
available online  in the NatureServe database (NatureServe 2009) at http://www.natureserve.org/explorer/.
Additional information on state-listed species is available from state  T&E species coordinators.
Numerous physical and biological stressors have resulted in the listing  of multiple aquatic species as
threatened or endangered. Major factors include habitat destruction or modification, displacement of
populations by exotic species, dam building and impoundments, various point and nonpoint sources of
pollution, poaching, and accidental deaths due to human activity.
Because construction activity occurs in every state and in or near a variety of waterbody types, there are
many potentially affected T&E aquatic species. The current USFWS list of T&E species includes 255
November 2009                                                                                   2-23

-------
Environmental Impact and Benefits Assessment for the C&D Category
aquatic species in five species groups: fish, amphibians, mollusks, crustaceans, and corals (USFWS
2009). For a complete list of these species, see Appendix B. This list is only a partial representation of
potentially affected species because it does not include aquatic or aquatic ecosystem-dependent plants,
insects, arachnids, reptiles, birds, and mammals. Although the USFWS list of T&E species does not
provide information on causes of endangerment for each species, a review of the ecological literature
suggests that elevated sedimentation, turbidity, and suspended sediment levels may have adverse impacts
on species that are already in peril either directly or indirectly through impacts on supporting food chains
(Cloud 2004; NatureServe 2009; USFWS 2009).
Several T&E species are thought to be particularly susceptible to excessive sedimentation, turbidity, and
suspended sediment levels. One such example is Amblema neislerii, an endangered freshwater mussel
commonly known as the fat threeridge. As the NatureServe Explorer database notes, ".. .the species and
its habitats continue to be impacted by excessive sediment bed loads  of smaller sediment particles,
changes in turbidity,  [and] increased suspended solids..." (NatureServe 2009). Other similarly sediment-
susceptible species of mollusks include the oyster mussel (Epioblasma capsaeformis), purple bankclimber
(Elliptoideus sloatianus), Louisiana Pearlshell (Margaritifera hembeli), Alabama Heelsplitter (Potamilus
inflatus), and Ouachita rock pocketbook (Arkansia wheeleri) (USFWS 2009).
Many fish species are also vulnerable to sedimentation. Sedimentation is a risk factor for a number of
shiner species, including the Arkansas river shiner (Notropis girardi), Pecos bluntnose shiner (Notropis
simuspecosensis), and blue shiner (Cyprinella caeruled) (NatureServe 2009). The NatureServe Explorer
notes that the biggest protection need for the threatened blue shiner (Cyprinella caerulea), native to the
southeastern United States, is "...preventing] siltation of habitat, especially during the spawning
period..." (NatureServe 2009). Sedimentation is also listed as a threat factor to a number of minnow and
darter species, such as the loach minnow (Tiaroga cobitis), spikedace (Medafulgidd), fountain darter
(Etheostoma fonticola), leopard darter (Percina pantherind), watercress darter (Etheostoma nuchale), and
vermilion darter (Etheostoma chermocki) (USFWS 1998, 2009; Drennen 2003; Cloud 2004; NatureServe
2009).
Sedimentation has also contributed to the threatened status of some populations of rainbow trout and
several salmon species in the Pacific Northwest (NatureServe 2009).  For example, California coho
salmon and steelhead trout are listed as threatened and endangered by the state of California. The state has
expressed concern about the impact of sediment discharges from construction sites on habitat throughout
these species' range (McEwan and Jackson 1996; California Department of Fish and Game 2004). The
state noted, for example, that discharges from the construction of Interstate 5 in California depleted
spawning gravels for coho salmon (California Department of Fish and Game 2004).
Construction activities have been noted as a source of stress for the endangered Barton Springs
salamander (Eurycea sosorum) (USFWS 2005). The U.S. Fish and Wildlife Service (2004) also issued a
biological opinion addressing potential harm to the Arkansas river shiner (Notropis girardi) from a bridge
construction project in Texas. While USFWS concluded that the project would not jeopardize the
continued existence of the species, it did note that the river's fish community would be adversely affected
by temporary loss of habitat, seining and handling of individual fish necessary to remove them from the
immediate vicinity of the construction site, increased turbidity in the  river, and harassment due to
construction activity  (e.g., equipment usage, materials storage, foot and vehicle traffic, installation of
sediment and erosion controls, incidental deposition of debris in river). USFWS stated that these impacts
might affect habitat access and seasonal movement of the fish.
November 2009                                                                                 2-24

-------
Environmental Impact and Benefits Assessment for the C&D Category
Texas has identified the threatened and endangered aquatic plants Texas wild-rice (Zizania texand) and
puzzle sunflower (Helianthus paradoxus) as vulnerable to siltation. Texas wild-rice also grows best in
water with little or no turbidity. Little Aguja pondweed (Potamogeton clystocarpus) was identified as
vulnerable to substrate texture modification (USFWS 1998).

2.4    Sediment and Turbidity Impacts on Human Use of Aquatic Resources

As sediment eroded from construction sites settles in surface waters or elevates sediment concentrations
and turbidity in the water column, human uses of surface waters and human-built elements of the
environment can be damaged. Damage from discharge of sediment and turbidity to surface waters has
been recognized in the literature for several decades (Wolman and Schick 1967).  Impacts to several types
of human use of aquatic resources are described below.

2.4.1  Navigation on Surface Waters

Navigable waterways, including rivers, lakes, bays, shipping channels, and harbors, are an integral part of
the United States' industrial transportation network. Navigable channels are prone to reduced
functionality due to sediment build-up, which can reduce the navigable depth and width of the  waterway
(Clark et al.  1985). Increased sedimentation can lead to increased navigational difficulties  and
inefficiencies such as groundings and delays (Osterkamp et al.  1998). Shipping companies may switch to
lighter loads or smaller vessels, which are generally less efficient. Removing sediment to keep navigable
waterways passable requires dredging, which can be costly. The U.S. Army Corps of Engineers (USAGE)
spends an average of more than $572 million (2008$) every year to dredge waterways and keep them
passable  (USAGE 2009).
Dredging is itself an environmentally disruptive activity because it significantly disturbs the physical
structure of a surface water's bed and because dredged material may contain significant quantities of toxic
substances and heavy metals. Dredging can disturb contaminants that have settled to the bottom of the
waterway, increasing the potential for their uptake by fish and other aquatic organisms. Dredged sediment
may be disposed of in another section of the waterbody or watershed, relocating the problem rather than
removing it. Additionally, if the sediment removed from a site is contaminated, it can add  to pollution at
the disposal  site (Clark et al. 1985).
In addition, unless there is overdredging to compensate or sedimentation is monitored so that dredging
activity may be timed optimally, waterbodies will be on average be less navigable than if sedimentation
rates were reduced.
In areas with significant sediment contamination, dredging may not be a feasible  option because of high
disposal costs for the dredged sediment. This can lead to damage of shipping vessels or shipping
inefficiencies. An example is the Great Lakes Harbor and Indiana Harbor Ship Canal, where navigational
dredging has not been conducted since 1972 because of the difficulty of contaminated dredge spoil
disposal. Ships have trouble navigating the harbor and canal and must enter the harbor while loaded at
less than optimum vessel drafts. There is also restricted use of some docks, requiring unloading at
alternative docks and double handling  of bulk commodities to the preferred dock. In 1995, the  increased
transportation costs associated with the lack of dredging were estimated to be $12.4 million annually
(USEPA 2004c).
Chapter  7 of this report describes the methodology for and presents the results of EPA's analysis of
benefits to navigation from several alternative regulatory options.
November 2009                                                                                2-25

-------
Environmental Impact and Benefits Assessment for the C&D Category
2.4.2   Reservoir Water Storage Capacity

Reservoirs are water impoundments, often manmade, that serve many functions, including providing
drinking water, flood control, hydropower supply, and recreational opportunities. The National Inventory
of Dams includes more than 75,000 dams (Renwick et al. 2005). Renwick et al. (2005) estimated the
presence of 375,000-750,000 additional manmade impoundments in the United States, of which the
majority are smaller than 1 acre in area. These smaller impoundments are used for recreation, livestock
water supply, sediment trapping and erosion control, and other purposes. Sediment in streams can be
carried into reservoirs and smaller impoundments, where it can settle and build up layers of silt overtime.
Historically, the United States Geological Survey (USGS) has recorded an average of 2.6 billion pounds
of sediment deposition settles in reservoirs each year (USGS 2007c). An increase in sedimentation rates
will reduce reservoir capacity and utility. To replace this capacity, sediment must be dredged from
reservoirs, or new reservoirs must be constructed (Clark et al. 1985). Both dredging and new reservoir
construction can have a variety of environmental impacts. Crowder (1987) estimated that the United
States was losing about 0.22 percent of its reservoir capacity each year due to sedimentation. Clark et al.
(1985) noted that total U.S. reservoir capacity is filling up slowly and has enough excess capacity
dedicated to hold sediment build-up over hundreds of years. However, the study went on to conclude that
while total reservoir sedimentation is manageable, sedimentation is far from uniform and that, in
approximately 15 percent of U.S.  reservoirs, sedimentation rates exceeded 3 percent of capacity annually
and, in the more extreme cases, 10 percent per year (Clark et al.  1985).
Sediment can also affect reservoir evaporation rates. Turbid waters tend to be sharply stratified, with a
warm layer at the surface and a cooler layer below. Because warm water evaporates faster, turbidity can
cause higher rates of water loss from reservoirs. However, turbidity may also increase the reflectivity of
water, which makes water absorb solar heat more slowly than it otherwise would and reduces evaporation
rates (Clark et al. 1985). The overall effect from temperature stratification in an individual reservoir could
be positive or negative.
Wolman and Schick (1967) describe the loss of the use of a Massachusetts reservoir because of sediment
discharged from airport construction. Significant resources were expended over two successive years for
supply and pumping of water from alternative sources.
Chapter 8 of this report describes the methodology for and presents the results of EPA's analysis of
benefits from reduced sedimentation of larger reservoirs under several alternative regulatory options.

2.4.3   Municipal Water Use

Discharges from construction sites can affect the quality and cost of providing drinking water. Sediment,
turbidity, and other discharged pollutants negatively affect water quality and require increased spending
on treatment measures such as settlement ponds, filtration, and chemical treatment (Osterkamp et al.
1998). There is an additional cost associated with removing sludge that is created during the treatment
process  (USEPA 2007b). The presence of pollutants or dissolved minerals in drinking water may also
affect the flavor and odor of water.
Surface  water sedimentation can impede water flow into drinking water treatment facility intakes.
Facilities may need to expend resources to unblock intakes or rebuild them in a different part of the
drinking water supply surface water.
November 2009                                                                                2-26

-------
Environmental Impact and Benefits Assessment for the C&D Category
Construction site discharge impacts on municipal water supplies have been documented in the literature.
Wolman and Schick (1967) noted that erosion and deposition of sediment had compromised use of
surface waters as municipal water supplies in Maryland. They stated that where storage facilities were
lacking and water was pumped directly from a river, sediment concentration spikes following rain events
could force facilities to temporarily cease water withdrawals. This practice was practicable for only a very
short period of time. Highway officials in the state had compensated municipalities for the construction of
additional water storage facilities and for changing water intake locations to cope with sediment
discharges from highway construction sites. Tan and Thirumurthi (1978) documented increases in
turbidity, suspended solids, nitrogen, and conductivity levels in water supply lakes due to highway
construction in Nova Scotia, Canada. Buckner (2002) described contamination of a municipal water
supply due to runoff from an upstream highway construction site.
Chapter 9 of this report describes the methodology for and presents the results of EPA's  analysis of
benefits from reduced impacts to municipal water supplies under several alternative regulatory options.

2.4.4  Industrial Water Use

Elevated sediment and turbidity levels may have negative effects on industrial water users. Suspended
sediment increases the rate at which hydraulic equipment, pumps, and other equipment wear out, causing
accelerated depreciation of capital equipment. Sediment can also clog water intake  systems at power
plants and other industrial facilities and possibly require their relocation to another part and/or depth level
of the surface water. Some industrial facilities treat water before use, and elevated sediment and turbidity
levels may require additional treatment (Osterkamp et al. 1998) or make a surface water source unusable.
Wolman and Schick (1967) noted that erosion and deposition of sediment had led to turbid waters
unsuitable for industrial uses in the  state of Maryland. They stated that even small amounts of sediment
can cause problems for industrial operations such as vegetable processing or cloth manufacture. They also
noted that sediment had led to failure of pumping equipment. Excess sediment levels may require the use
of filters to improve water quality.
At least one positive effect from elevated turbidity levels is also possible, however.  Since turbidity may
reduce the rate at which waterbodies absorb solar heat, more turbid waterbodies may supply cooler water,
which in turn could improve the efficiency of cooling water systems at power plants and other industrial
facilities (Clark et al. 1985).

2.4.5  Agricultural Water Use

Irrigation water that contains sediment or other pollutants from construction sites can harm crops and
reduce agricultural productivity. Suspended sediment can form a crust over a field,  reducing water
absorption, inhibiting soil aeration,  and preventing emergence of seedlings. Sediment can also coat the
leaves of plants, inhibiting plant growth and reducing crop value and marketability. Furthermore,
irrigation water that contains dissolved salts or pollutants can harm crops and damage soil quality (Clark
etal. 1985).
Wolman and Schick (1967) noted that erosion and deposition of sediment in surface waters in the state of
Maryland had led to failure of pumping equipment. Pumps are often used for the movement of irrigation
water. Excess sediment levels may require the use of filters to improve water quality.
Surface water sedimentation can block irrigation water intake structures or require their relocation to
another part of the surface water. Sedimentation can also cause sediment build-up in irrigation water
November 2009                                                                                 2-27

-------
Environmental Impact and Benefits Assessment for the C&D Category
canals and other transport systems (Osterkamp et al. 1998), reducing their efficiency. Sediment can also
accumulate in livestock water supply impoundments and reduce their capacity (see Section 2.4.2).

2.4.6  Stormwater Management and Flood Control

Sediment in discharges from construction sites can clog or fill ditches, culverts, storm sewer pipes and
basins, Stormwater detention ponds, infiltration basins, and other stormwater management structures
(Osterkamp et al. 1998). Sedimentation also decreases the capacity of natural stream and river channels,
ponds, lakes,  and reservoirs, increasing the likelihood of overbank flow events and their magnitude (Paul
and Meyer 2001). Wolman and Schick (1967) noted that erosion and deposition of sediment in the state of
Maryland had led to sediment deposition in streams and overflow and clogging of storm drains.
If left in place, sedimentation can increase flooding. The U.S. Army Corps of Engineers (2007) stated that
channel sedimentation due to construction discharges contributed to flooding of a Virginia residential
neighborhood. Preventing flood damages from excessive sedimentation may require increased
maintenance efforts such as dredging (Clark et al. 1985), vacuuming, and other types of sediment removal
from stormwater management structures and surface waters. If sedimentation removal is performed in
surface waters or in structures directly adjacent to surface waters, it can create additional environmental
impacts such  as resuspension of sediments and associated turbidity and contaminants in the water column
and disturbance of the physical structure of the surface water bed and the associated benthic aquatic
community.
Surface water and stormwater management system sedimentation can increase the severity of property
damages  to bridges, roads, farmland, and other private and public property from flooding.  Additionally,
sediments carried by flood waters can damage property and can be expensive to remove once  deposited,
particularly in developed areas (Clark et al. 1985; Osterkamp et al. 1998). Clark et al. (1985) estimated
flooding  damages attributable to all sediment discharges to be $1.5 billion (2008$), annually.
The quantity of sediment captured in stormwater control structures is unknown or could be very variable.
Nelson and Booth (2002) did not observe significant long-term sediment accumulation in the stormwater
retention/detention facilities (including the ditches and pipes connecting  them to surface waters) of newer
residential developments in Issaquah, Washington. They did note, however, that a significant amount of
sediment was removed every year from catch-basins in the city of Issaquah, though a large fraction of this
sand may have derived from winter road sanding.

2.4.7  Recreational Uses and Aesthetic Value

Polluted water greatly reduces the aesthetic appeal of a variety of recreational activities that take place in
or near surface waters, including boating; swimming; hunting; and outings to hike, jog, picnic, camp, and
view wildlife. Turbidity  and suspended sediments are highly visible pollutants. Murky and visually
unpleasant water and odors can detract from the enjoyment gained through outdoor activities.
Sedimentation of streams, reservoirs, lakes, and bottomlands can reduce  their depth and thus their
capacity  for boating and swimming  (Osterkamp et al. 1998).
Turbidity caused by sediment discharges may affect the safety of boating. Turbidity may obscure
underwater obstacles, making collisions more likely. Increased sedimentation levels may create  sandbars,
increasing the chances of running aground. Clark et al. (1985) estimated that turbidity (from all sources)
may be responsible for as many as 200 boating fatalities and many more injuries each year. Turbidity may
also create safety hazards for swimmers by reducing the ability to see underwater hazards, increasing the
November 2009                                                                                2-28

-------
Environmental Impact and Benefits Assessment for the C&D Category
frequency of diving accidents by impairing the ability to gauge water depth, and making location of
swimmers in danger of drowning more difficult.
Wolman and Schick (1967) noted that erosion and deposition of sediment in the state of Maryland had
damaged recreation areas. Concern about potential highway construction impacts to aesthetic and
recreational values of a small lake was sufficient to motivate residents of an adjoining neighborhood to
initiate monitoring of lake water quality (Line 2009).
Aesthetic degradation of land and water resources resulting from sediment and turbidity discharges can
also reduce the market value  of property near surface waters and thus affect the financial status of
property owners. For example,  a hedonic price study by Bejranonda et al. (1999) found that "the rate of
sediment inflow entering the  lakes has a negative influence on lakeside property rent." Sediment
discharges also have a significant impact on stream morphology. For example, higher coarse-sediment
load leads to an increase in width of the river bed and, as a result, bank erosion (Wheeler et al. 2003). A
1993 study of Lake Erie's housing market found that "erosion-prone lakeshore property will be
discounted" (Kriesel et al. 1993). Stabilization of stream banks leads to an increase in the value of
surrounding property (Streiner  and Loomis 1996).

2.4.8   Recreational and  Commercial Fishing

Pollutants can negatively affect local flora and fauna, negatively impacting aesthetic appeal, wildlife
viewing, and hunting for game  (Osterkamp et al. 1998). By harming fish and shellfish communities,
sediment can reduce fish and shellfish numbers or cause more desirable fish and shellfish to be replaced
by less desirable fish and shellfish in a given location.
As discussed in Section 2.3, sediment, turbidity, and other discharges from construction sites can reduce
fish and shellfish populations by inhibiting their reproduction, growth, and survival. In some areas,
desirable sediment-sensitive  fish may be replaced by less-desirable, sediment-tolerant fish and shellfish.
These population changes and reductions would reduce the size of commercial and recreational harvest by
lowering both the total abundance of organisms and their individual size. These changes negatively affect
recreational anglers, subsistence anglers, commercial anglers, fish and shellfish sellers,  and consumers of
fish and shellfish products.
In addition, sediments, turbidity, and other pollutants reduce the aesthetics of a waterbody, which can
reduce anglers' enjoyment of their fishing experience and their choices of how often and where to fish.
Sediment and turbidity may also affect recreational anglers by reducing the distance over which fish can
see lures, resulting in lower catch rates (Clark et al. 1985).

2.5    Sediment and Turbidity  Criteria

Natural levels of sediment and  turbidity play important roles in aquatic ecosystems, providing habitat for
benthic species, protection from predators, nutrient transport, and other functions. Appropriate levels of
sediment are waterbody specific, as sediment levels vary naturally among different types of waterbodies
according to geology, topography, stream gradient, waterbody morphology, vegetative land cover,
climate, soil erodibility, and other landscape characteristics of the contributing watershed (USEPA
2006b). When sediment and turbidity enter surface  water at elevated levels, they can raise TSS and
turbidity levels in the water column, increase deposition of sediment on the waterbody's bed, and degrade
overall water quality (USEPA 2006b). Elevated sediment and turbidity levels also negatively impact
November 2009                                                                                 2-29

-------
Environmental Impact and Benefits Assessment for the C&D Category
human use of aquatic resources for recreation, drinking water supply, agricultural use, industrial use, and
navigation and can impair stormwater management system function and surface water aesthetics.
EPA and many states have issued criteria that describe appropriate sediment and turbidity levels for
surface waters. Many of these criteria are narrative and provide a qualitative description of appropriate
levels. Other criteria are numeric.

2.5.1   Federal Sediment Criteria

In 1987, EPA published the following guidance for developing numeric water quality criteria for
sediment and turbidity:
    Solids (Suspended, Settleable) and Turbidity - Freshwater fish and other aquatic life: Settleable and suspended solids
    should not reduce the depth of the compensation point for photo synthetic activity by more than 10 percent from the
    seasonally established norm for aquatic life (USEPA 1987).
This criterion has not been widely adopted by states. Guidance addressing primarily aesthetic properties,
however, has been adopted by several states.
    Aesthetic Qualities - All waters shall be free from substances attributable to wastewater or other discharges that: settle
    to form objectionable deposits; float as debris, scum, oil, or other matter to form nuisances; produce objectionable
    color, odor, taste, or turbidity; injure or are toxic or produce adverse physiological response in humans, animals, or
    plants; [or] produce undesirable or nuisance aquatic  life (USEPA 1987).
EPA has not issued new sediment or turbidity criteria since  1987. However, EPA has published a
document providing guidance on setting sediment criteria: Framework for Developing Suspended and
Bedded Sediments (SABS) Water Quality Criteria (USEPA 2006b).

2.5.2   State Sediment Criteria

Many states have developed their own criteria to designate appropriate sediment levels for waters within
the state. Some criteria vary by  waterbody type. Streams with hard substrates (e.g., gravel, cobble,
bedrock) or cold water fisheries typically have  more stringent criteria than streams with soft substrates
(e.g., sand, silt, clay) or warm water fisheries. Hawaii has separate criteria for reefs (Berry et al. 2003). In
addition, Total Maximum Daily Loads (TMDLs) addressing sediment and turbidity have been developed
for some surface waters and  contain information on acceptable levels specific to those waterbodies.
Many criteria are narrative statements describing the general nature of healthy sediment or turbidity levels
without attempting to provide numeric guidelines. Narrative criteria most frequently address turbidity or
surface water appearance (e.g.,  "free of substances that change color or turbidity"). Other criteria refer to
undesirable biological effects (e.g., "no adverse effects" or "no actions which will impair or alter the
communities") (Berry et al. 2003). EPA (USEPA 2006b) provides a summary of state narrative and
numeric criteria (though some criteria may have been modified since this information was compiled).
A number of states have also set numeric criteria for TSS or turbidity levels. Most of these numeric
criteria are for turbidity. Turbidity criteria exist as absolute values or ranges of values or as a permitted
exceedance beyond background turbidity levels. Some states have issued numeric criteria for suspended
sediments.
A number of states have criteria based on sedimentation levels over a time period or during a storm event.
Values are typically 5 mm during an individual event (e.g., during the 24 hours following a heavy
rainstorm)  for streams with hard substrate bottoms and 10 mm for streams with  soft bottoms. Hawaii's
reef criterion is 2 mm of deposited sediment after an event (Berry et al. 2003).
November 2009                                                                                   2-30

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 2-3 summarizes state suspended sediment and turbidity criteria for surface water quality. States
using narrative criteria are noted with a "Yes" in the "Narrative TSS" or the "Narrative Turbidity"
column. Values for numeric criteria and some details on their implementation are provided in the
"Numeric TSS" and "Numeric Turbidity" columns. This table is intended to provide an overview of state
numeric suspended sediment and turbidity water quality requirements and does not summarize all details
relevant to their applicability. In addition, states periodically update their water quality criteria. Readers
should consult individual state water quality criteria publications in order to obtain the most complete and
current requirements.
November 2009                                                                                 2-31

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 2-3:
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Suspended Sediment and Turbidity Criteria for Surface
Numeric TSS Narrative
(mg/L) TSS Numeric Turbidity Criteria
<50 NTU increase
<5 NTU increase
<10% increase (max. 25 NTU)3
10-50 NTU4
1 0-75 NTU 4 for baseflow values
<20% increase2
<10% increase3
Yes
<5 NTU increase
<10 NTU increase
Water Quality by State
Notes on Numeric Turbidity Criteria
Source: Alabama DEM (2009).
Source: Alaska DEC (2003)
Human Contact: 50 NTU in rivers, 25 NTU in lakes
Cold Water Fishery: 10 NTU
Warm Water Fishery: 50 NTU in rivers, 25 NTU in lakes
"Non-point source runoff shall not result in the
exceedance of the in stream all flows values in more
than 20% of the ADEQ ambient monitoring network
samples taken in not less than 24 monthly samples."
Source: Arkansas Pollution Control and Ecology
Commission (2007).
20% increase applies to Central Coast Region and is
measured in JTU
10% increase applies to San Francisco Region
Source: California EPA 2009a, b

Under ambient conditions. Class AA criteria
Source: Connecticut DEP (2002).
For all fresh waters and mixing zones
District of Columbia - -
Florida
Georgia
Hawaii
<29 NTU increase
Yes



Turbidity levels not to exceed NTU as specified below:
Water type , Geometric More than
mean 10° oof the time 2% of the time
Stream (wet 5 15 25
season)
Stream (dry 2 5.5 10
season)
1Q_557 yes 2"25 NTU f°r streams A11 Estuaries IZjTflll 3U> 	 Il-PZZZ
0.1-15 NTU for coastal/marine waters Pearl Harbor 4.0 8.0 15.0
Embayment 0.4 1.0 1.5

Open Coastal 0.5 1.25 2.0
(wet season)
Open Coastal 6.02 	 0^05 	 1.6
(dry season)
Oceanic 0.03 0.1 0.2
Marine 1 0.1 1 1

Narrative
Turbidity
Yes1
Yes
Yes
Yes
Yes
-
-
-
Yes5
Yes
Yes
Yes

November 2009
                                                                                                                                        2-32

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 2-3: Suspended Sediment
and
Turbidity Criteria for Surface Water Quality by State
Numeric TSS Narrative
State (mg/L) TSS Numeric Turbidity Criteria
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey 25
New Mexico
Yes
-
Yes
Yes
Yes
Yes
Yes
-
-
Yes
-
-
Yes
Yes
Yes
-
Yes
Yes
-
Yes
<50 NTU increase at any time, <25
NTU increase over 10 consecutive
days
-
-
25 NTU
-
-
25, 50, 150 NTU, or <10% increase4
-
150 NTU at any time
50 NTU monthly average
-
-
5, 10, or 25 NTU4
50 NTU
-
-
-
Waterbody specific
0-10 NTU increase4
10, 30, or 50 NTU4
<10 NTU increase2 or <20% increase3
Notes on Numeric Turbidity Criteria
For applicable mixing zones set by the department only
Source: Idaho DEQ (2008).





25 NTU: freshwater lakes, reservoirs, and oxbows;
designated scenic streams and outstanding natural
resource waters
50 NTU: Amite, Pearl, Ouachita, Sabine, Calcasieu,
Tangipahoa, Tickfaw, and Tchefuncte Rivers;
estuarine lakes, bays, bayous, and canals
1 50 NTU: Atchafalaya, Mississippi, and Vermilion
Rivers and Bayou Teche
10% Increase: All other waters




Domestic Consumption: Class A & B 5 NTU, Class C 24
NTU
Fisheries & Recreation: Class A 10 NTU, Class B & C
25 NTU
Industrial Consumption: Class A 5 NTU
Applicable to waters outside the limits of a 750-foot
mixing zone




Class A waters: No turbidity other than natural
Class B & C: 10 NTU increase over natural
Freshwater: 50 NTU any time, 15 NTU 30-day average
Saline water: 30 NTU any time, 10 NTU 30-day average
Coastal saline: 10 NTU
GA waterbody Turbidity shall not exceed 5 NTU
Source: New Mexico (2005)
Narrative
Turbidity
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
-
Yes
Yes
Yes
Yes7
Yes
Yes
-
Yes

November 2009
                                                                                                                                        2-33

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 2-3:
State
New York
North Carolina
North Dakota
Suspended Sediment and Turbidity Criteria for Surface
Numeric TSS Narrative
(mg/L) TSS Numeric Turbidity Criteria
Yes Waterbody specific
5009 Yes 10,25, or 50 NTU4
30
Water Quality by State
Notes on Numeric Turbidity Criteria

10 NTU in trout waters
25 NTU in other lakes and reservoirs
50 NTU in other streams
If background exceeds these levels, no increase


Narrative
Turbidity
Yes8
-
-
Ohio
Oklahoma
Oregon
Pennsylvania
Puerto Rico
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
10 or 50 NTU4
Yes < 10% Increase
Yes 40-100 NTU
Specific to waterbody class
5 NTU or < 1 0 NTU increase4
Yes 10 NTU or < 1 0% increase
53-263 (at any time);
30-1 50 (monthly Yes
average)4
Yes
Yes
35-904 Yes 10 or 15 NTU4
10 or 25 NTU4
Yes
Cold water aquatic communities/trout fisheries: 10 NTU;
Lakes: 25 NTU; Other Surface Waters: 50 NTU
If background exceeds these levels, point sources may not
cause increases above ambient levels
Exception for temporary increases to do "legitimate
activities" including dredging and construction,
providing all practicable turbidity control techniques
are applied.
Specific limits only apply to Neshaminy Basin
Class SB shall not exceed 10 NTU, except by natural
causes; Class SC shall not exceed 10 NTU; Class SD
shall not exceed 50 NTU, except when due to natural
phenomena
Source: Puerto Rico EQB (2003).
Class A: 5 NTU
Class B & C: 10 NTU increase
Providing existing uses are maintained
Specific to freshwaters suitable for supporting trout stocks



10 NTU for cold and warm water game fish
1 5 NTU for nongame fish, waterfowl, and other wildlife
10 NTU for Class A(l) ecological waters and Class B
cold water fish habitats
25 NTU for Class B warm water fish habitats

-
Yes
Yes
Yes
-
Yes
Yes
Yes
Yes
Yes10
Yes
Yes

November 2009
                                                                                                                                        2-34

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 2-3: Suspended Sediment and Turbidity Criteria for Surface Water Quality by State
State
Numeric TSS
(mg/L)
Narrative
TSS
Numeric Turbidity Criteria
Notes on Numeric Turbidity Criteria
Narrative
Turbidity
                                                                                                Class AA &A: 5 NTU2 or 10% increase3
                                                                                                Class B & C: 10 NTU2 or 20% increase3
                                                                                                Lakes:  5 NTU increase
 m  ,.   ,                                                 5 or 10NTU2or<10%or<20%       Also includes a flow-based criteria based on flow from
 Washington                   -                -                      -34                         +   +•    •+  ^         •   *    m*   i
                                                                      increase                       construction site. Flows ranging trom 10 to above
                                                                                                    100 CFS have turbidity measured at 100 to 300 feet
                                                                                                    from downstream activity
	Source: Washington (1997)	
 West Virginia                 -                -             10 NTU2 or < 10% increase3
 Wisconsin	-	Yes	-	Yes
                                                                                                <10 NTU increase for Class 1 and 2 waters that are cold
 „,     .                                      ,,                 ,„    IC-K-TTTT-       4              water fisheries                                            ,,
 Wyoming                     -               Yes             <10 or 15 NTU increase            IC-K-TTTT-       c   ™   i    j ^   +   a  +                 Yes
   J      °                                                                                      <15 NTU increase for Class 1 and 2 waters that are warm
	water fisheries and all Class 3 waters	
 1  "There shall be no turbidity of other than natural origin that will cause substantial visible contrast with the natural appearance of waters or interfere with any beneficial uses which they serve."
 2  If naturally less than 50 NTU.
 3  If naturally greater than 50 NTU.
 4  Varies based on waterbody classification.
 5  May not "produce objectionable odor, color, taste, or turbidity."
 6  Applies during dry season only. Geometric mean of 10 mg/L, not exceeding 30 mg/L 10% of the time and 55 mg/L 2% of the time.
    "To be aesthetically acceptable, waters shall be free from human-induced pollution which causes: 1) noxious odors; 2) floating, suspended, colloidal, or settleable materials that produce
    objectionable films, colors, turbidity, or deposits."
 8  Waterbody types AA, A, B, C, D, SA, SB, SC, SD, I: no increase may cause substantial visible contrast from natural conditions.
 9  Standard is for total dissolved solids (TDS).
 10 For mixing zones.
 Source:  USEPA (2006b), unless otherwise noted.	
November 2009                                                                                                                                                    2-35

-------
Environmental Impact and Benefits Assessment for the C&D Category
2.6    Surface Water Quality Impairment from Sediment and Turbidity

Section 305(b) of the Clean Water Act (CWA) requires states, territories, and other jurisdictions of the
United States to submit reports to EPA on the quality of their surface waters every two years. These
entities have determined the appropriate uses of each waterbody within their jurisdiction. Uses can
include recreation, drinking water source, navigation, cold water fishery, and wildlife habitat, among
others. States and other entities determine appropriate narrative and/or numeric water quality criteria for
each of the designated uses. The criteria describe the physical, chemical, and biological characteristics of
a surface water able to fulfill its designated uses. The sediment criteria described in the section above are
examples of such criteria. Typically, a surface water has criteria for multiple water quality parameters. If
a waterbody fails to meet any one of its designated uses, CWA Section 303(d) requires a state or other
entity to list the waterbody as "impaired." If a waterbody meets its designated uses but is in danger of
failing to do so in the future, the state or other entity must list the waterbody as "threatened" (USEPA
2005a).
The Assessment  TMDL Tracking and Implementation System (ATTAINS) provides information on water
quality conditions reported by the states to EPA under Sections 305(b) and 303(d) of the Clean Water
Act.  The information available in ATTAINS is updated as data are processed and are used to generate the
biennial National Water Quality Inventory Report to Congress. This information reflects only the status of
those waters that have been assessed. Appendix A provides information on the state water report year for
which data were  available for populating the tables below as of September 17, 2009.
According to ATTAINS, 49 percent of assessed reach miles (or 458,209 miles) have been identified as
impaired; sediment contributes to impairment in 107,231 miles, and turbidity contributes to impairment in
26,278 miles. For lakes and reservoirs, 66 percent of assessed lake acres (or 11,545,337 acres) have been
identified as impaired; sediment contributes to impairment in 715,558 lake acres, and turbidity contributes
to impairment in 1,008,276 acres. For bays and estuaries, 63  percent of assessed square miles (or 11,222
square miles) have been identified as impaired; sediment contributes to impairment in 209 square miles,
and turbidity contributes to impairment in 240 square miles. It should be noted that individual waters may
be impaired by more than one  pollutant. Although states tend to target their monitoring efforts to those
surface waters they believe to be impaired, the total area of impaired surface waters due to sediment and
turbidity is probably underestimated due to the low percentage of surface waters that were assessed. As of
September 17, 2009, states had assessed only 26 percent of the nation's  reach miles, 42 percent of its lake
acres, and 20 percent of its bay and estuary square miles.
Table 2-4 and Table 2-5 present information on "sediment" and "turbidity and suspended solids"
impairment by EPA Region as reported by the states in ATTAINS. This information is presented by EPA
Region. Figure 2-1 provides a map of the EPA Regions.
November 2009                                                                                2-36

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 2-4: Surface Waters Impaired by "Sediment," by EPA Region
EPA Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach (miles)
236
1,123
8,188
11,126
34,410
3,567
9,659
9,597
198
29,127
107,231
Source: EPA ATTAINS database (USEPA 2009a)
Lake/Pond/Reservoir (acres)
443
39,277
193
23,333
57,965
153,478
144,125
147,592
-
149,152
715,558
as of 9/1 7/09.
Bay/Estuary (sq. miles)
4
0
12
-
-
193
-
-
-
-
209

 Table 2-5: Surface Waters Impaired by "Turbidity" and "Suspended Solids," by EPA
 Region
EPA Region
1
2
3
4
5
6
7
8
9
10
Nation
Stream/ River (miles)
339
3,539
229
732
9,278
6,880
84
2,304
904
2,439
26,728
Source: EPA ATTAINS database (USEPA 2009a)
Lake/Pond/Reservoir (acres)
43,735
1,539
-
13,814
134,427
626,642
38,173
3,168
136,980
9,798
1,008,276
as of 9/1 7/09.
Bay/Estuary (sq. miles)
5
8
-
4
-
195
-
-
28
-
240

Figure 2-1: EPA Regions
   *.'-.'»K
•v
             [Quani
             IT rust rerrltoriaal
             American Samoa
             northern Mariana
             lalanda       I
November 2009
                                                                                     2-37

-------
Environmental Impact and Benefits Assessment for the C&D Category
Water quality in the United States has also been assessed through a series of national, probability-based
surveys known as the National Aquatic Resource Surveys. These surveys use randomized sampling
designs, core indicators, and consistent monitoring methods and laboratory protocols to provide
statistically defensible assessments of water quality at the national scale. The Wadeable Streams
Assessment (USEPA 2006d) is a statistical survey of the smaller perennial streams and rivers that,
according to the report, comprise 90 percent of all perennial stream miles in the United States. Excess
streambed sedimentation is ranked as one of the most widespread stressors examined in the survey. (The
survey did not analyze turbidity or suspended sediment levels.) According to the survey, 25 percent of
streams have "poor" streambed sediment condition, and 20 percent are in "fair" condition relative to
reference  streams. The survey also examined the association between stressors and biological condition,
and found that high levels of sediments more than double the risk for poor biological condition (see
Figure 2-2).

tFigure 2-2: Extent of Stressors and Their Relative Risk to the Biological Condition  of the
                                       Nation's Streams
                                Relative Extent
       Relative Risk to
Macroinvertebrate Integrity (IBI)
             Nitrogen
           Phosphorus
    Riparian Distrubance
    Streambed Sediments
    In-stream Fish Habitat
Riparian Vegetative Cover
               Salinity
           Acidification


h-H 30.9%
h-H 25.5%
J -i jt no/
1 | 1 24.9%
h-H 19.5%
1 	 1 19.3%
•42.9%
J 2.2%
3 10 20 30 40
Percentage Stream Length in Most

1 	 1 	 1 2.1
t 1 ^ ••)
\ \ 2..2.
\— I— 11-4
1 1 104

h-H 1.4
1 	 1 	 1 1.6
	 b-n.7
234
Relative Risk
                               Disturbed Condition
Source: USEPA (2006c).
Another National Aquatic Resource Survey, the National Coastal Condition Report III (USEPA 2008b),
assesses coastal aquatic habitat and found water clarity (related to turbidity) to be poor in 17 percent of
coastal waters. Data on excessive sedimentation was not collected for the report.

2.6.1   Current Total Suspended Solid Concentrations in U.S. Surface Waters

TSS concentrations vary from waterbody to waterbody due to differences in their contributing watersheds
created  by natural conditions and human activity. EPA used the SPARROW model to estimate current
TSS concentrations in the United States' Reach File Version 1 (RF1) surface water network (see Chapter
6 for more information). For this analysis, the RF1 network consists of approximately 650,000 miles of
the largest rivers and streams in the coterminous United States and associated lakes, reservoirs, and
estuarine waters. Due to data limitations, EPA did not model surface water turbidity levels. Table 2-6
summarizes the distribution of TSS concentrations found in individual RF1 reaches as estimated by
EPA's analysis. The median TSS concentration, weighted by reach length, is 302.9 mg/L. The range of
TSS concentrations falls between 26.7 mg/L and 6,154.5 mg/L at the 5th and 95th percentiles, respectively.
The average TSS concentration is 1,068.6 mg/L, indicating that concentrations in the 90th to 95th
November 2009
                                                                                             2-38

-------
Environmental Impact and Benefits Assessment for the C&D Category
percentile range are high enough to shift the average concentration significantly higher than the median
concentration.
 Table 2-6: SPARROW Distribution of TSS Concentrations in RF1  Reaches1
RFl
Reach
Count
62,370
RFl
Reach
Miles
650,043
Average
TSS
(mg/L)1
1,068.6
Distribution of TSS Concentrations in RFl Reaches1
5th
Percentile
26.7
25th
Percentile
102.9
50th
Percentile
302.9
75th
Percentile
1,079.6
95th
Percentile
6,154.5
  Concentrations weighted by reach length. Concentration estimates reflect replacement of potential outliers (defined as values above the 95
 percentile) with the 95th percentile value.
Figure 2-3 shows total RFl reach miles and their current TSS concentrations as predicted by SPARROW
(not weighted by reach length). As shown below, the majority of waters have TSS concentrations below
1,000 mg/L, though substantially higher average concentrations up to and exceeding 6,000 mg/L can be
found in some waters.
 Figure 2-3: TSS Concentrations by Total RF1 Reach Miles as Predicted by SPARROW for
	Current Conditions	
    225000
    200000
    175000
  y,
    150000
  t125000
  0)
  1*100000
  3
  o
  H  75000

     50000

     25000

          0
                 Total: 62,370 River Reaches (650,043 River Miles)
                                              95th Percentile: 6,154.51 mg/L
\
V
      N
                                                                           'V0   OV  
-------
Environmental Impact and Benefits Assessment for the C&D Category
year time periods. EPA uses these long-term estimates because the data and modeling resources currently
available to EPA do not permit the finer level of time resolution necessary to model individual storm
events. A study by Wong (2005) documents that long-term averages of TSS concentration sampling data
provide a different type of perspective on surface water conditions than data on observed 90th and 98th
percentile concentrations, which better reflect the types of concentrations observed after precipitation or
other erosion events.
EPA expects that, in general, surface water TSS concentrations during and immediately following storm
events would be higher than the concentrations presented here. During  periods when no precipitation
event is available to cause a sediment discharge from a construction site, surface water sediment
concentrations would generally be lower.
Current levels of sediment erosion and sediment and turbidity discharge to many surface waters in the
United States are much higher than would be observed under natural, undisturbed conditions due to
human activity. Sediment and turbidity levels in surface waters reflect these elevated levels of sediment
input.
Human activities that have increased sediment erosion and transport in surface waters include agriculture,
grazing, logging, construction, and mining. In addition, land use change has increased streambed erosion
in many urbanized areas, mobilizing significant quantities of sediment (Nelson and Booth 2002; Renwick
et al. 2005; Wilkinson 2005; Wilkinson and McElroy 2007). It is estimated that soil erosion due to human
activity is approximately 10 times greater than erosion from all natural forces combined (Wilkinson
2005). Approximately one-third of all eroded sediment in the United States derives from agricultural
activities (Osterkamp et al. 1998). Construction is considered to be another major source of erosion
(Wilkinson 2005). Nelson and Booth (2002) documented a 50 percent increase in annual sediment yield
due to human activity (primarily urbanization) in a western Washington watershed.
Only a portion of all eroded sediments are transported through the entire surface water system to the
ocean in any given year (Renwick et al. 2005; Wilkinson 2005; Wilkinson and McElroy 2007). A large
portion is stored on the land surface and in and adjacent to surface waters. Some surface waters are
currently mobilizing historical sediments stored on floodplain terraces and in small impoundments (e.g.,
from past farming, mining, and clear-cutting practices). These stored sediments provide a large and
continuing source of sediment input to the U.S. surface water network (Renwick et al. 2005).
November 2009                                                                                 2-40

-------
Environmental Impact and Benefits Assessment for the C&D Category
        Other Pollutants  in Construction  Site Discharges
In addition to mobilizing and discharging sediment and turbidity to surface waters, construction activity
can discharge many other kinds of pollutants. Whereas sediment and turbidity are nearly ubiquitous in
discharges from construction sites, the presence and quantity of other pollutants varies widely. These
pollutants can derive from construction materials and equipment, historic site contamination, and natural
soil and groundwater constituents. They may be carried in storm water in solution or adsorbed to
transported sediment particles. These "other pollutants," though not as well documented as sediment and
turbidity in the construction context, can also impact water quality and therefore aquatic life and human
use of water resources. This chapter discusses these pollutants and their potential sources, behavior in
surface waters, and potential to impact aquatic organisms and human use of aquatic resources. Section 3.1
provides an overview of the importance and concerns associated with these other pollutants. Section 3.2
provides additional information on potential sources of pollutants on construction sites and changes in
stormwater discharge.  Section 3.3 provides additional information on specific pollutants potentially found
in construction site discharges. Pollutant-specific impacts from laboratory testing and field studies  are
cited as appropriate, but further information is available in Chapter 4, which provides a review and
summary of studies from the literature that document specific cases of construction site discharges, the
pollutants they contain, and their impacts on the aquatic environment.

3.1     Introduction

In addition to mobilizing and discharging sediment and turbidity to surface waters, construction activity
can discharge many other kinds of pollutants. These pollutants can derive from construction materials and
equipment, historic site contamination, and natural soil and groundwater constituents. Construction
activity can also impact receiving waters by alteration of the hydrologic regime; for example, increasing
the volume and intensity of stormwater runoff but decreasing a surface water's dry weather baseflow.
These alterations may  lead to erosion and sediment mobilization further down in the watershed of the
construction activity. Section 3.2 provides additional information on potential sources of pollutants on
construction sites and changes in stormwater discharge.
The pollutants of interest include nutrients (nitrogen, phosphorus), organic compounds and materials
(organic material, petroleum hydrocarbons, polycyclic aromatic hydrocarbons (PAHs), pesticides),
metals, dissolved inorganic ions (major anions and cations), substances that can modify surface water pH,
and pathogens (bacteria, fungi). Studies of construction sites have  documented the presence of many of
these pollutants in construction site stormwater, discharges, and receiving waters. Section 3.3 describes
their potential sources, their transport behavior, and their potential to impact the aquatic environment and
resources.

3.1.1    Pollution Discharge and Transport Pathways

Most pollutants enter surface waters when precipitation erodes or dissolves materials and carries them to
receiving waters. Many of the pollutants described in this section adsorb to or are a component of soil
particles and travel with them as they erode (e.g., organic compounds, metals, nutrients). Stormwater flow
preferentially removes fine particles from soil because of their lesser mass and tendency to remain
suspended in flow once mobilized. Many of these pollutants form adsorption complexes that bind them to
clay minerals, other fine particles, and organic matter. For this reason, many pollutants that associate with
November 2009
                                                                                              3-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
sediments are present in higher levels in the sediments mobilized by stormwater than in the parent soil
(Novotny and Chesters 1989). The movement of these pollutants is therefore very similar to that
described for sediment in Section 2.2.
Other pollutants dissolve in precipitation or disassociate ("desorb") from sediment and move in a manner
that mimics that of stormwater flows (e.g., nitrate, chloride). Some pollutants may switch between
dissolved and particulate forms as conditions change. This may make control of these pollutants more
difficult since standard practices for limiting sediment discharge may not significantly reduce these
dissolved constituents.
Pollutants can also enter surface waters during dry weather due to excavation dewatering, groundwater
seepage,  construction equipment washout, landscape irrigation, and equipment operation in or near
surface waters. Dewatering, the removal and discharge of water from excavated areas or work areas
located within surface waters, is necessary at some construction sites to maintain dry work conditions.
Water can collect in excavated areas due to precipitation and groundwater seepage (e.g., Horner et al.
1990, Montgomery County DEP 2009). Levels of dewatering activity and groundwater contamination
vary widely among construction sites.
Excavation dewatering discharges can contain a variety of naturally occurring or human activity-derived
pollutants. Any pollutant that can enter groundwater flow can be present in a dewatering discharge.
Common groundwater contaminants include heavy metals, organic solvents and degreasers, pesticides
and herbicides, and nitrates. Naturally occurring constituents, such as iron and manganese, can form
extensive and unaesthetic deposits on surface water beds when groundwater under low reduction-
oxidation ("redox") conditions is exposed to ambient air.

3.1.2  National Scale and Cumulative Impact Concerns

The importance of the discharge of these other pollutants from construction sites is uncertain, but there
are strong indications that the cumulative effects of construction activities is of national significance.
Levels of pollutant discharge other than sediment and turbidity are highly variable among construction
sites. EPA quantified construction site contributions of nitrogen and phosphorus, both common soil
constituents, to surface waters. Chapter 6 describes the methodology EPA used for this analysis as well as
analysis results. Many other pollutants, however, may be less commonly found on construction sites. The
current literature contains limited information on the frequency and level at which these pollutants appear
at construction sites across the United States.
What literature is available indicates that individual construction sites typically do not discharge acutely
toxic levels of metals and toxic organic compounds unless a material spill occurs or  site soils and/or
groundwater are contaminated with sufficient levels of toxic pollutants. However, discharges of pollutants
below acutely toxic levels can nevertheless be  of concern.
Many potential construction site pollutants  are components of or adsorb to sediments (e.g., metals, toxic
organic compounds, nutrients). Pollutants with this behavior can accumulate over time in surface  water
sediments to levels many times higher than those found in the  adjacent water column. Pollutant levels can
eventually become high enough to affect aquatic organisms (USFWS 1998). Construction is a widespread
and frequent activity  in the United States. The  number of construction projects active each year (greater
than 80,000) is much larger than the number of active facilities in many other industrial sectors (USEPA
2009b). The multiplier effect of such a large number of potential discharge points means that the
construction sector's  cumulative pollutant contributions can be very high, even when contributions from
individual sites are relatively low.
November 2009                                                                                  3-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
In addition, construction sites can discharge mixtures of pollutants. While no one pollutant may be
present at a level sufficient to harm a surface water or its biota, the cumulative or synergistic effect of
several mixture components may be sufficient to do so, particularly if other environmental stressors are
present. Construction sites also discharge to many surface waters that are already impaired or in danger of
impairment. Construction sites contribute to the already elevated cumulative pollutant loads in these
waterbodies.

3.2     Construction Site Sources of Pollutants Other Than Sediment and Turbidity

3.2.1   Construction Materials and Equipment

An enormous variety of materials are used on construction sites. A number of these materials contain
chemicals or produce debris or residues that can cause harm if they enter surface waters in sufficient
quantities. These materials can contribute nutrients, metals, and toxic organic compounds to stormwater
discharges.
A partial list of common building and equipment materials includes paint; sealants; grouts; solvents;
adhesives; wood preservatives (e.g., creosote, copper azole, ammoniacal copper zinc arsenate); wood
shavings, scraps, and sawdust; drywall; packaging material; detergents (e.g., trisodium phosphate); metal
materials and debris; asphalt-based materials (e.g., roofing and paving materials); Portland cement;
concrete; concrete truck washout; aggregate; dust suppressants (e.g., calcium chloride); blasting
explosives; and construction equipment oil, fuel, lubricants, antifreeze, other fluids, and washwaters.
Pavement materials include asphalt, Portland cement, pavement sealers (e.g., methacrylate sealers, coal-
tar based sealants) and additives (e.g., plasticizers, air entraining agents, water reducers, strength
accelerating agents, corrosion inhibitors, colorants, pumping acids, rejuvenators, liquid and fibrous
polymers, carbon black, sulfur). Many industrial wastes/by-products are used as aggregate or additives in
asphalt and Portland cement pavement mixtures or as aggregate for unpaved roads. These materials
include recycled asphalt and concrete pavement, shredded scrap tires, blast furnace slag, plastic waste,
coal bottom ash, coal fly ash, steel slag, mine tailings and other wastes, municipal sludge, municipal solid
waste incineration bottom ash, phosphogypsum, wood waste, petroleum refinery residuals, foundry  sand,
bricks, and asphalt shingles (National Research Council 2001; Eldin 2002).
Additional materials, both newly manufactured and industrial wastes/by-products, are used as soil
amendments or fill on construction sites. Materials include imported soil and stone, lime (soil
stabilization), cement (soil stabilization), crushed limestone, asphalt emulsions, and a variety of industrial
by-products. Industrial by-products used as fill include coal combustion wastes (e.g., fly ash, bottom ash,
boiler slag), lime kiln dust, gypsum, gypsum wallboard, foundry sand,  sandblasting abrasives, cement kiln
dust, wood ash, glass, and wastewater filter sand (Eldin 2002;  USEPA  2005b; Iowa Waste Reduction
Center 2009). Landscaping materials include topsoil, soil sterilants in treated topsoil, fertilizer, mulch,
hydroseeding mixtures, lime, straw, and pesticides.
Pollutants from construction equipment and materials can be carried to surface waters in stormwater (or
other) flow in solution or by erosion of debris and particulate matter (City of Vancouver 2007).
Infiltration of construction site  stormwater into  soil can reduce transport of pollutants to surface water,
though many pollutants adsorb to soil particles and will travel  with them if they are subsequently eroded
and transported to surface water. Improper operation and maintenance of construction equipment and
insufficient storage and housekeeping practices (e.g., improper storage of oil,  gasoline, and chemical
November 2009                                                                                  3-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
products) can lead to additional leakage, spillage, or erosion of materials and their transport to surface
waters. Spills at construction sites have been documented (Garton 1977 (diesel fuel); Ohio  1997f
(construction debris); Kress 2007).
Some types of construction activity (e.g., stream channel realignment, culvert installation, bridge support
emplacement) take place directly in surface waters. In addition, at some sites construction equipment
must cross surface waters in order to access all parts of a site. These activities can provide easy access for
contaminants from construction materials and equipment to surface waters, as well as cause direct
disturbance of surface water sediments, physical structure,  and organisms if precautions are not taken.
Sedimentation ponds can become sources of pollutants if they are allowed to become host to large
amounts of algae, insect larvae, bacteria,  and other organisms or if they are allowed to heat water to high
temperatures before discharge.
The type and level of building, equipment, fill, soil amendment, and landscape material pollutants in
stormwater discharges from a specific construction site depend on the choice of materials and
construction practices used on the site. Many building, pavement, soil amendment, fill, and landscaping
materials vary in composition based on source materials and manufacturing method. Pollutants present in
a material from one source or manufacturer may not be present in the same category of material from
another source or manufacturer (NRC 200 Ic). Not every member of every material category will contain
constituents  of concern.

3.2.2  Historic Site Contamination

Construction soil and groundwater disturbance can mobilize pollutants already present on a site due to
prior human activity. A broad range of human activities and land uses can contaminate a site's soil and
groundwater.
Industrial activities can directly deposit a wide range of pollutants in soil and groundwater,  including
metals, nutrients, and toxic organic compounds  (Folkes et al. 2001; Seattle and King County Public
Health 2001; Area Wide Soil Contamination Task Force 2003, Blanco et al. 2009).
Agricultural lands can contain elevated levels of pesticide residues, nutrients from fertilizers and manure
(particularly phosphorus), and salts from  irrigation waters (New Jersey DEP 1999; Area Wide Soil
Contamination Task Force 2003; Renshaw et al. 2006; Robinson et al. 2007; Stone and Anderson 2009).
Approximately 40 percent of current developed land was previously used for agricultural activities (NRC
2008).
In urban areas, construction often begins  with demolition and removal of existing buildings and other
structures. Materials in preserved wood (e.g., creosote, chromated copper-arsenate); paint chips and dust
(including lead-based paints); brick and wood scraps; concrete debris; asbestos-based insulation and other
materials; polychlorinated biphenyls (PCBs) from old electricity transformers; metal scrap; waste
solvents; historic fuel spills; and fill material composed of coal ash and clinkers, brick and concrete
demolition debris, and other industrial by-products may be  present (Blanco et al.  2009). Pesticides have
been widely used in urban areas, as well,  and residues may persist in soil and other materials (Folkes et al.
2001). Erosion and leachate from these materials may elevate pH, metals, and toxic organic compounds
levels in stormwater runoff.
Several contaminants can be deposited on the land through atmospheric deposition, including metals (e.g.,
copper, chromium, lead, mercury, and zinc), nutrients (nitrogen, phosphorus), and organic compounds
(e.g., PAHs, PCBs, and pesticides). Sources of these contaminants include manufacturing facilities (e.g.,
November 2009                                                                                  3-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
smelters), power plants, motor vehicles, and wind-borne dusts (Folkes et al. 2001, NRC 2008).
Atmospheric deposition from these sources is more likely to occur near urbanized areas and downwind of
sources. However, atmospheric deposition can be very widespread and may affect areas that are
fundamentally undeveloped (e.g., undisturbed forest). Overtime, atmospherically deposited pollutants
may accumulate to significant levels (Seattle and King County Public Health 2001, Area Wide Soil
Contamination Task Force 2003). For example, lead emissions from several decades of leaded fuel use in
the United States elevated lead levels in many urban soils.
The frequency and magnitude at which soil contamination is present nationwide is unknown. A task force
sponsored by the state of Washington estimated that up to 676,550 acres in the state may contain soil
contaminated with low to moderate levels of lead and/or arsenic (Area Wide Soil Contamination Task
Force 2003). The state of New Jersey, which convened a similar task force, estimated that soil on up to 5
percent of the state's land was contaminated with lead and arsenic residues from the historical use of lead
arsenate pesticides (New Jersey DEP 1999). At the severe end of the contamination scale, approximately
300 sites on the Superfund National Priorities List have been identified as having some level of soil
contamination. At this time, EPA has determined that the soil contamination is severe enough at
approximately 150 of the sites to require cleanup under the Superfund program (USEPA 2004c).
The type and level of historical pollutants at a specific construction  site depend on the nature of the past
activities and pollutant release and varies widely  among sites. Not all sites contain constituents of concern
from historic contamination.

3.2.3  Natural Site Constituents

Naturally occurring soil, rock, and groundwater composition and constituents  vary widely among sites.
Soil is a complex mixture of inorganic and organic materials deriving from weathered rock, biological
activity, precipitation, and atmospheric deposition. Impacts from suspended solids,  turbidity, and
sedimentation  from eroded soil are discussed  in Chapter 2. Soil constituents also commonly include
nitrogen, phosphorus, partially and fully decomposed organic matter, bacteria, fungi, insects, metals (e.g.,
iron, aluminum, magnesium, manganese), and other inorganic materials (e.g.,  compounds containing
silicon, potassium, calcium, sodium, sulfur, inorganic carbon, and chlorides).  Some constituents, such as
zinc, copper, arsenic, chromium, nickel, barium, and lead, can be natural soil components but can be
highly uneven  in their distribution. Type and level of soil and rock constituents at specific construction
sites depend on the nature of the soil and rock present on the construction sites. For example, Hawaiian
soils have naturally higher levels of copper, zinc, and aluminum (Wong 2005).
Rock and soil constituents are natural components of sediments entering aquatic ecosystems. At elevated
levels, however, they can become pollutants. Construction activity modifies natural erosion processes by
pulverizing and exposing formerly sequestered rock and soil materials to erosion and leaching by
precipitation, making transport of their constituents to  surface waters much more likely.
Some sites also have natural groundwater seeps that may be further  exposed or captured in site drainage
by construction activity. These seeps, particularly those with naturally acidic waters or highly sodic
waters, may contain elevated levels of certain constituents, including metals. Kalainesan (2007)
documented elevated levels of dissolved solids and metals in a construction site groundwater seep
(aluminum, manganese, iron, magnesium, and sulfate).
Elevated levels of pollutants in stormwater and receiving surface waters due to natural rock, soil, and
groundwater constituents at construction sites has been documented in the literature (Huckabee et al.
November 2009                                                                                  3-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
1975; Burton et al. 1976, Chisolm and Downs 1978; Extence 1978; Tan and Thirumurthi 1978; Daniel et
al. 1979; Shields and Sanders 1986; Kucken et al.  1994; Kalainesan 2007; USEPA 2009b).
Naturally occurring organisms such as bacteria and fungi can also function as pathogen pollutants. For
example, Aspergillus sydowii, a common terrestrial fungus, can enter marine environments through soil
erosion and discharge to surface waters and is a widespread sea fan pathogen in the Caribbean (Lafferty et
al. 2004).

3.2.4  Altered Stormwater Discharge

An additional impact from construction activity is that vegetation removal, soil compaction, modified
surface topography, and installation of impervious surfaces (e.g., roads, parking lots, and roofs) can
increase the volume and intensity of water runoff discharging from a construction site during precipitation
events but decrease a surface water's dry weather baseflow.
The increased flow can cause changes in the hydrograph of a receiving stream, causing it to exhibit more
frequent and larger peak flows, followed by periods of reduced base flow and possibly desiccation. The
stream may incise or widen to accommodate periodic elevated flow volumes (Wheeler et al. 2003). The
changes in streambed morphology can increase sediment erosion, destabilize stream banks, elevate
sediment and turbidity levels in downstream waters, and cause long-term changes in aquatic and riparian
habitat. Elevated flows can also damage aquatic communities by scouring and washing downstream
aquatic plants, algae, invertebrates, and other organisms. These changes can degrade or destroy
previously existing aquatic communities.
Groundwater recharge is reduced by increased intensity and volume of stormwater runoff and reduced
soil infiltration. Reduced groundwater recharge results in less water available for surface water dry
weather baseflow. Reductions in dry weather baseflow can cause extended dry periods in streams and
stress aquatic organisms.
The type and level of runoff from a specific construction site depends on the nature of the construction
site, though many sites, particularly those that have been cleared of vegetation, have elevated surface
runoff. Surface runoff changes are predictable through the use of standard hydrologic models (e.g., Soil
Conservation Service TR-55, USDA 1986).
Line and White (2007) documented similarly elevated levels of stormwater runoff during all stages of a
residential construction project in North Carolina.  The authors postulated that soil compaction during the
clearing and grading stages of the project were sufficient to mimic the impervious surfaces in place during
later stages of the project. The authors also documented an increase in peak discharge rate and elimination
of stream baseflow in the affected area. Selbig and Bannerman (2008) found that a stormwater basin
serving a community under construction using Low Impact Development (LID) techniques discharged
more water than the original undeveloped landscape but significantly less than a stormwater basin serving
a neighborhood developed with conventional techniques. Clausen (2007) found that while runoff was
elevated from a traditional construction site, it actually declined at a site using LID practices. Other
studies documenting increases in surface runoff include Yorke and Davis (1972), Selbig et al. (2004), and
Montgomery County DEP (2009). Burton et al. (1976) and Chen et al. (2009) did not detect a change in
site runoff levels.
Construction sites can also concentrate previously diffuse surface runoff with ditches, pipes, and other
structures. When a high volume of runoff enters a surface water, particularly at too high a velocity, it can
November 2009                                                                                 3-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
erode the receiving water's banks and bed if not properly diffused. This erosion can modify local
waterbody morphology, transport sediment, and elevate turbidity levels in downstream waters.

3.3     Other Pollutants from Construction Activity

This section provides additional information on specific pollutants potentially found in construction site
discharges. The discussion describes potential sources of other pollutants, their transport behavior, and
their potential to impact the aquatic environment and resources.

3.3.1   Nitrogen and Phosphorus

Phosphorus and nitrogen compounds are natural components of most aquatic ecosystems. At excessive
levels, however, they become pollutants. Nitrogen and phosphorus are present at some level in nearly all
construction site soils. Information on levels of nitrogen and phosphorus compounds typically found in
soils and sediments is provided in Section 6.4.1. Several studies have documented discharge of nutrients
from construction sites believed to derive from natural soil constituents (Burton et al. 1976; Tan and
Thirumurthi 1978; Daniel et al. 1979; Shields and Sanders 1986).
Nutrients can also derive from construction materials and historic contamination. While some level of
nitrogen and phosphorus is naturally present in most soils, the presence of nutrients from other sources
varies widely. Roofing materials treated with phosphate washes and binders during preparation for use
may release phosphorus (NRC 2008). Phosphorus from blasting explosive residue on a construction site
has been documented (USEPA 2009b).
Fertilizers are also frequently used during the surface stabilization phase of construction to encourage
vegetation growth. They can be applied directly or as a component of a hydroseeding mixture or other
product. Organic materials such as mulch, compost, and straw are also  used to encourage vegetation and
contain  elevated nutrient levels relative to site soils. When insufficiently managed, they can contribute
nutrients to construction site discharges (Glanville et al. 2004). Faucette et al. (2007) documented
elevated nitrogen and soluble phosphorus levels in construction site stormwater runoff from areas treated
with straw blankets containing mineral fertilizer for erosion control. Hydroseeding operations, in which
seed, fertilizers, lime and, in some cases, tackifiers, are applied to soil in a one-step operation are more
likely to discharge nutrients than conventional seed-bed preparation operations during which fertilizers
and lime are tilled into the soil. Nutrient discharges from fertilizer use on construction sites have been
documented in several studies (Taylor and Roff 1986; Horner et al. 1990; Faucette et al. 2007; Clausen
2007; Kalainesan 2007; Faucette et al. 2008; Chen et al. 2009).
Decades of application of phosphorus-containing fertilizers and human, animal, and industrial waste to
the land surface has significantly elevated phosphorus levels in many surface soil horizons in developed
areas of the United States (Brady and Weil 1999). Land recently used as cropland may have elevated soil
phosphorus levels from many years of fertilizer application. This phosphorus can be mobilized through
erosion  (Paul and Meyer 2001).
Other studies documenting the presence of nutrients in construction site storm waters and/or their
elevation in receiving waters due to construction  activity include White (1976),  Barton (1977), Barrett et
al. (1995a), Harbor et al. (1995), Fossati et al. (2001), Kayhanian et al.  (2001), Wong (2005), and Selbig
and Bannerman (2008). Some studies monitored for and found no elevation in nutrient levels downstream
of construction sites (Peterson and Nyquist 1972; Extence 1978; Young and Mackie 1991; Cleveland and
Fashokun2006).
November 2009                                                                                  3-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
Phosphorus and nitrogen can be found in both particulate and soluble forms on construction sites. Soluble
nitrogen and phosphorus compounds are the typical form found in fertilizers so as to be immediately
available to vegetation. Relatively low levels of soluble nitrogen compounds and very low levels of
soluble phosphorus compounds also occur in natural terrestrial soils. Soluble nitrogen compounds leach
more readily through soil than phosphorus compounds. Soluble forms of nitrogen and phosphorus can
dissolve in and be transported with stormwater flows.
Most nitrogen in terrestrial soils is found in particulate organic form (Brady and Weil 1999). Phosphorus
is typically found in particulate organic matter or bound to sediment. Nutrients form adsorption
complexes with clay minerals and organic matter. Phosphorus readily binds to sediment, even when
initially introduced to soil in soluble form such as fertilizer. Once bound to soil, it does not leach readily.
Because phosphorus has such a strong association with sediment, discharges of total phosphorus tend to
increase and decrease with total suspended solid discharges from sites (Selbig and Bannerman 2008).
Discharges of organic particulate nitrogen also vary closely with suspended sediment discharges (Daniel
etal. 1979).
Some studies have documented strong associations between turbidity or sediment discharges and nutrient
discharges from construction sites, suggesting that discharged nutrients were associated with particulate
matter eroded from the construction sites (Daniel et al. 1979; Shields and Sanders 1986). Daniel et al.
(1979) found 99 percent of total phosphorus discharges and 90 percent of total nitrogen discharges from
three residential construction sites to be  associated with sediment. Because of storm water's tendency to
transport finer particles from eroded soils and nutrients' tendency to be associated with fine sediments
and organic matter, eroded sediments are frequently enriched in nutrient content relative to the parent
soils (Novotny and Chesters 1989). Once in  a waterbody, nutrients can, under certain conditions, detach
from sediments and enter the water column (Bilotta and Brazier 2008).

3.3.1.1  Nitrogen, Phosphorus, and Aquatic Resource Impacts
While nutrients are necessary for the primary production (organic matter derived from photosynthetic
activity) that forms the  base of many aquatic food chains, they can become pollutants at elevated levels.
Eutrophication is a process by which excessive nutrient levels increase the growth rates of primary
producers in aquatic systems, particularly algae, which can drastically alter a waterbody's ecological
balance  and impair human use of the waterbody (Wetzel 2000). Gradual eutrophication of some types of
surface waters is a natural process, but nutrient levels elevated by human activity can drastically change
the rate at which the process takes place (Goldman and Home 1994).
Both nitrogen and phosphorus must be available for eutrophication to proceed. Freshwater system
eutrophication is typically limited by phosphorus availability, and marine systems are typically limited by
nitrogen availability. Because many undisturbed stream phosphorus levels are naturally low, even small
increases can impact water quality. Nitrogen can also be the primary nutrient limiting freshwater primary
production in some regions of the United States, particularly portions of the Northeast and the Pacific
Northwest with granitic or basaltic geology (USEPA 2006d) and in other waters on a seasonal basis.
Elevated nutrient levels can change the quantity and quality of algal communities in both marine and
freshwater systems. High concentrations of algal biomass,  commonly referred to as algal "blooms,"
accumulate in the water column faster than natural processes (e.g., settling, zooplankton grazing, benthic
organism filtering, senescence) can remove them.
Algal blooms can significantly  reduce surface water transparency and therefore reduce the light available
for submerged aquatic vegetation (SAV) photosynthesis. Elevated nutrient levels in the water column also
November 2009                                                                                  3-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
encourage the growth of algae on the surface of submerged vegetation, which further decreases light
availability. Light attenuation reduces photosynthesis levels and the maximum depth at which SAVs can
survive. Deterioration in water quality due to eutrophication has been a major factor in the decline of
SAV in many ecosystems (e.g., Staver et al. 1996). A number of estuaries have experienced loss of high
quality SAV beds (e.g., eelgrass, turtlegrass) because of reduced light availability due to eutrophication
from elevated nitrogen loadings. Impacts to aquatic vegetation subsequently decrease food and habitat
availability for other aquatic organisms.
Excessive nutrients in shallow, wadeable streams can lead to dense mats of algal growth, particularly
filamentous algae (e.g., Cladophora spp.) on bottom substrates.  Such growths alter the habitat quality of
the substrate and lead to predictable changes in aquatic macroinvertebrate populations. The proportion of
pollutant-tolerant species in the community increases. Proportions and abundances of various feeding
guilds shift, changing food quality and availability for fish and other organisms. Elevated nutrient levels,
in combination with elevated levels of fine-grained sedimentation, can foster the growth of aquatic
macrophyte beds, particularly in stream pools (King and Ball  1964; Taylor and Roff 1986).
Decomposition of excessive algal biomass reduces dissolved oxygen levels in surface waters, lessening
the amount available to other aquatic organisms. This can impair organisms' functioning and health and
possibly kill them.  In surface waters stratified by temperature and/or salinity gradients, the settling of
large amounts of organic matter can create  low oxygen or anoxic conditions in bottom waters and allow
only a few pollution-tolerant species to persist in the benthic community. Large diurnal swings in surface
water oxygen content have also been observed in some waters. Supersaturation of dissolved oxygen takes
place during daylight hours when photosynthesis takes place. At night, oxygen levels fall dramatically
due to plant respiration. Such shifts have been shown to cause many summer fish kill events.
Coldwater fish species (e.g., salmonids) that depend on cool, well-oxygenated bottom waters as a refuge
during summer are particularly vulnerable to eutrophication impacts and may be extirpated from affected
systems. Low levels of dissolved oxygen can affect amphibians by reducing respiratory efficiency,
metabolic energy, reproductive rate, and ultimately survival (USFWS 2005).
Eutrophication can also lead to shifts in fish community composition due to modifications of trophic
feeding relationships. For example, the proportion of large, piscivorous fish species that rely on sight to
catch their prey may decline in a surface water as water clarity falls. Filter-feeding planktivores (e.g.,
gizzard shad) may greatly increase in abundance. If eutrophication conditions are extreme, a fish
community can be  reduced to pollution-tolerant bottom feeders (e.g., carp).
Eutrophic conditions can also foster the growth of certain types of algal blooms with toxic effects on
aquatic organisms and human beings. Brown or red tides are intense algal growths that can release
neurotoxins into the water column, harming some forms of aquatic life. Toxic effects on aquatic
organisms can cause population declines and localized extinctions (Burkholder 1998).
These types of blooms usually require beach closings and temporary bans on seafood harvesting because
consuming shellfish from affected waters can cause shellfish poisoning in humans (Burkholder 1998).
Human exposure to marine algal neurotoxins has been associated with a range of neurobehavioral
abnormalities and is an emerging area of study (Friedman and Levin 2005). Illnesses most frequently
linked to neurophysiological disturbance are Amnesic Shellfish Poisoning, Ciguatera Fish Poisoning, and
Possible Estuarine Associated Syndrome, which is associated with exposure to substances from the
dinoflagellate Pfiesteriapiscicida through the food chain.
November 2009                                                                                  3-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
In freshwater systems, toxins from blue-green algal (Cyanophyta) species (e.g.,Microcystis,
Cylindrospermum) can be of concern. Blue-green algae toxins are classified according to mode of action
and include hepatotoxins (e.g., microcystins), neurotoxins (e.g., anatoxins), skin irritants, and others
(WHO 2003). Increased levels of organic material from algal blooms can also increase concentrations of
chlorination disinfection by-products, including haloacetic acids and trihalomethanes, in treated drinking
water. Severe morbidity and mortality in domestic animals due to toxin-contaminated drinking water has
been documented (WHO 2003). Microcystins, in particular, have been associated with acute liver damage
and possibly liver cancer in laboratory animals.
Some algal blooms may not be directly harmful to humans but may still impair human use of aquatic
resources. Increased levels of organic material associated with algal blooms can clog drinking water
intakes and cause unpleasant tastes and odors in drinking water (Boyd 1990) (see Chapter 9). Excessive
algae growths can reduce the attractiveness and viability of surface waters for a variety of recreational
activities including swimming, boating, fishing, wildlife viewing, and other outings. Excessive growths
can also clog screens on water intakes for irrigation, industrial use, and drinking water.
Sedimentation levels also increase over the long term in eutrophic waterbodies, particularly lakes,
reservoirs, and other impoundments with quieter flow conditions. Sediment cores obtained from studies
of former lake environments have documented large increases in annual sediment deposits after the onset
of eutrophication (Goldman and Home 1994). These elevated sedimentation levels, acting in concert with
sediment deposition from construction activities and streambed erosion due to land use change, can
decrease depth and volume of surface waters overtime. Chapter 8 discusses EPA's analysis of economic
impacts associated with loss of reservoir capacity.
Algal growth also increases surface water turbidity, the various aquatic life and human resource impacts
of which are discussed in Sections 2.3 and 2.4.
The nitrogen compound ammonia can function as a nutrient, the impacts of which are discussed above.
Ammonia is also toxic to many aquatic organisms at relatively low levels under certain water conditions.
Ammonia toxicity is primarily attributable to the unionized form (NH3) as opposed to the ammonium ion
(NH4+) form. Ammonia toxicity typically increases with pH such that above pH levels of 9, the most
toxic form (unionized ammonia) is the predominant fraction.  Temperature also influences ammonia
toxicity. Low dissolved oxygen levels found in the bottom waters of eutrophic waterbodies can increase
the potential for ammonia toxicity, particularly when a surface water is stratified. The low redox
conditions found in these waters favor the anaerobic microbial conversion of nitrite or nitrate to ammonia.

3.3.1.2  Nitrogen and Phosphorus Criteria and Surface Water Impairment
As noted in Section 2.6, Section 305(b) of the Clean Water Act (CWA) requires states, territories, and
other jurisdictions of the United States to submit reports to EPA on the quality of their surface waters
every two years If a waterbody fails to meet any one of its designated uses, CWA Section 303(d) requires
a state or other entity to list the waterbody as "impaired" and  not meeting its designated uses. Excessive
nutrients (phosphorus, nitrogen) and its associated biological  response (i.e., chlorophyll a, DO levels,
transparency) are a common reason for a waterbody failing to meet its designated uses.
Appropriate levels of nitrogen and phosphorus vary among surface waters. Nutrient levels associated with
eutrophication vary among regions with the country due to differences in geology, climate, and soil types.
EPA has published a series of technical guidance documents that provide methods for setting nutrient
water quality criteria for (1) lakes and reservoirs (USEPA 2000c); (2) rivers and streams (USEPA 2000d);
(3) estuaries and coastal marine waters (USEPA 2001); and (4) wetlands (USEPA 2008c). EPA has used
November 2009                                                                                 3-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
this guidance to develop tables of recommended nutrient criteria for lakes/reservoirs and streams/rivers
for several ecoregions across the country (USEPA 2009d). States, tribes, and other entities have started to
develop their own numeric nutrient criteria, using EPA recommendations as a starting point.
The Assessment TMDL Tracking and Implementation System (ATTAINS) provides information on water
quality conditions reported by the states to EPA under Sections 305(b) and 303(d) of the Clean Water
Act. The information available in ATTAINS is updated as data are processed and are used to generate the
biennial National Water Quality Inventory Report to Congress. This information reflects only the status of
those waters that have been assessed.
Table 3-1 presents information on surface waters with nutrient-related impairments (algal growth,
ammonia, noxious aquatic plants, nutrients, and organic enrichment/oxygen depletion) by EPA Region as
reported by the states in ATTAINS. Appendix A provides information on the state water report year for
which data were available for populating the table below as of September  17, 2009.
It should be noted that individual waters may be impaired by more than  one pollutant. Although states
tend to target their monitoring efforts to those surface waters they believe to be impaired, the total area of
impaired surface waters due to nutrients is probably underestimated due to the low percentage of surface
waters that were assessed. As of September 17, 2009, states had assessed only 26 percent of the nation's
reach miles, 42 percent of its lake acres, and 20 percent of its bay and estuary square miles.
Table 3-1:
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Source: EPA
: Nutrient-Related Impairment in 305(b)-Assessed Waters, by EPA Region
Stream/River (miles)
2,267
10,485
8,009
11,522
72,350
13,162
22,369
11,298
20,816
31,722
204,000
ATTAINS database (USEPA 2009 a) as
Lake/Pond/Reservoir (acres)
229,728
219,203
61,179
262,978
1,244,029
742,606
487,346
995,262
804,315
506,696
5,553,342
of 9/17/09.
Bay/Estuary (sq. miles)
759
591
4,039
22

724


111

6,246

Water quality in the United States has also been assessed through a series of national, probability-based
surveys known as the National Aquatic Resource Surveys. These surveys use randomized sampling
designs, core indicators, and consistent monitoring methods and laboratory protocols to provide
statistically defensible assessments of water quality at the national scale. The Wadeable Streams
Assessment (USEPA 2006d) is a statistical survey of the smaller perennial streams and rivers that,
according to the report, comprise 90 percent of all perennial reach (i.e., stream and river) miles in the
United States.  Elevated nutrient and sedimentation levels are the top stressors of streams. According to
the survey, 32  percent of streams have "poor" nitrogen conditions, and 21 percent have "fair" conditions
relative to reference streams. Also, 31 percent have "poor" phosphorus conditions, and  16 percent have
"fair" conditions relative to reference streams.  The survey also examined the association between
stressors and biological condition and found that high levels of nitrogen or phosphorus more than double
the risk for poor biological condition (see Figure 3-1).
November 2009                                                                                 3-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Figure 3-1: Extent of Stressors and Their Relative Risk to the Biological Condition of the
                                       Nation's Streams
                                Relative Extent
       Relative Risk to
Macroinvertebrate Integrity (IBI)
             Nitrogen
           Phosphorus
    Riparian Distrubance
    Streambed Sediments
    In-stream Fish Habitat
Riparian Vegetative Cover
               Salinity
           Acidification

|— |— | 31.8%
1 — H 30.9%
HH 25.5%
1 — H 24.9%
1 	 1 19.5%
1 	 1 19.3%
H 2.9%
Jl 2.2%
3 10 20 30 40
Percentage Stream Length in Most

-1 T
1 1 J-
,
\ 	 1 	 12.2
1 	 1— 11.4
1 1
1 1
BH '-4
1 | 1 1.6
H ; 1.7

2
Relative
— 1 *7 A

3 4
Risk
                               Disturbed Condition
Source: USEPA (2006d).
Several threatened and endangered (T&E) species are vulnerable to eutrophication from nutrient
pollution. One study postulated that 25 percent of all current freshwater T&E species are adversely
impacted by eutrophication (Richter et al.  1997). Mollusks are frequently vulnerable to eutrophication
impacts. Examples include several species of freshwater snails and mussels (USFWS 2000; Biber 2002;
and Kozlowski and Vallelian 2009). Fish species and aquatic plants are also vulnerable. Threatened fish
species include the Waccamaw silverside (Menida extensd), paleback darter (Etheostoma pallididorsum),
and an arctic grayling subspecies (Thymallus arcticus montanus) found in Montana and Wyoming
(NatureServe Explorer 2009).
To quantitatively evaluate the water quality benefits from reducing nitrogen and phosphorus loadings
from construction sites, EPA developed an approach that relies on the empirical relationship between in-
stream sediment and nitrogen and phosphorus levels. A full discussion of this approach is provided in
Chapter 6.

3.3.2  Organic Compounds and Materials

A wide variety of organic compounds and materials can be discharged from construction sites. Organic
compounds are present on construction sites as natural soil constituents, components of a variety of
building and equipment materials, and historic soil contamination. The nature and frequency of these
materials' occurrences on and discharge from construction sites has not been well characterized in the
literature. Given the proprietary nature of the composition of many commercially available products and
the enormous variety of possible compounds in these materials, research in this area can be challenging.
Pavement materials such as asphalt and petroleum-based asphalt pavement rejuvenators contain a variety
of organic compounds. There are many concrete additives whose environmental effects have not been
thoroughly characterized (e.g., air entraining agents, water reducers, strength accelerating agents, and
pumping acids) (NRC 200Ic). Benzothiazole has been detected in leachate from shredded scrap tires
commonly used as pavement additives and is a likely toxin (Eldin 2002).  N, 4-dimethylbenzamine has
been detected in methacrylate bridge deck sealer and is also highly toxic (NRC 200Ic).
November 2009
                                                                                             3-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
PCB compounds have been banned for use in manufacturing in the United States due to their
carcinogenicity, persistence in the environment, and bioaccumulation potential. Because they were widely
used historically and are very stable in the environment, their residues are frequently found in soils of
areas where they have been manufactured or used, including many industrial and urban sites. PCBs bind
well to sediments and are often transported with them.
Organic compounds can have toxic effects on aquatic organisms and/or influence surface water oxygen
levels when present at sufficient levels. Discharges of organic compounds and materials of all types show
a strong correlation to suspended sediment discharges (Barrett et al. 1995b). Oil and grease, other
petroleum hydrocarbons, PAHs, and many organic pesticides adsorb to sediment and travel with it as it
erodes. As sediments accumulate in receiving waters, concentrations of associated organic pollutants can
increase to levels above those found in the adjacent water column. Over time these sediment levels can
become high enough to cause adverse impacts to organisms living in, interacting with, or connected
through the food chain to those sediments.
Information on some of the better characterized pollutants known to be of concern, including high organic
content material, petroleum hydrocarbons, PAHs, and organic pesticides, is provided below.

3.3.2.1  Organic Matter and Dissolved Oxygen
Organic matter consists of biodegradable carbon-based material. Organic material is a natural constituent
of many construction site soils. A common construction site practice is to stockpile topsoil, roots, and
other vegetative debris from site clearing on-site for use in final site landscaping. If not protected from
erosion, these stockpiles can contribute organic matter, sediment, turbidity, and adsorbed pollutants to
surface waters. Developers may also import topsoil for use during landscaping. Wood mulch, compost,
straw, organic geotextiles (i.e., biodegradable structural sheets used to reduce slope erosion), and other
organic materials used for soil erosion control and landscaping can increase organic material discharges if
improperly maintained and allowed to erode (Glanville et al. 2004). Taylor and Roff (1986) documented
extensive erosion of straw mulch from a construction site, sufficient to clog culverts downstream of the
site.
Elevated loadings of organic material  can increase levels of oxygen-demanding substances (e.g., as
measured by biological oxygen demand (BOD) and chemical oxygen demand (COD)) in receiving
waters. Microbes aerobically break down the organic compounds. Elevated BOD and COD levels can
lower dissolved oxygen levels in surface water, leading to several of the impacts associated with nutrient-
derived oxygen depletion discussed in Section 3.3.1.
Excessive algal growth has been documented in sedimentation basin waters at some construction sites
(Kalainesan 2007; USEPA 2009b). These blooms can deplete oxygen  levels diurnally and elevate organic
matter content in water discharged from these basins.
Some studies have documented a decline in the organic matter content of surface water sediment due to
construction activity (Barton 1977; Young and Mackie 1991). This effect is derived from the erosion of
largely inorganic soil particles from construction sites and the high levels of those sediments entering the
surface water. Some aquatic organisms depend on organic matter in surface water sediments as a food
source. Elevated levels of sediment containing  low levels of organic matter can reduce the food value of
this material for these organisms.
November 2009                                                                                3-13

-------
Environmental Impact and Benefits Assessment for the C&D Category
3.3.2.2  Petroleum Hydrocarbons
Petroleum hydrocarbons include a variety of crude and refined oil products, including gasoline, diesel
fuel, and lubricating oils. These products contain a diverse array of organic compounds, many poorly
characterized in terms of environmental behavior and impact. Petroleum hydrocarbons are often
monitored in water as a group, either as total petroleum hydrocarbons or as oil and grease.
Petroleum-derived aliphatic hydrocarbons on construction sites typically derive from human activity,
either current or historic. Many historic industrial sites have some level of petroleum hydrocarbon
contamination in site soils and, at some sites, groundwater as well. Improper operation, fueling, and
maintenance of construction equipment, and poor housekeeping practices (e.g., improper storage of oil,
diesel fuel, and gasoline products) can lead to leakage or spillage of materials containing petroleum
hydrocarbons (Garton 1977). Accidental rupturing of petroleum transport lines can also occur during
construction (USFWS 2005, citing two incidents during utility line trench construction in Texas).
Asphalt-based products (pavement material, asphalt shingles) are commonly used on construction sites
and contain a variety of organic compounds including n-alkanes, carboxylic acids such as n-alkanoic
acids, benzoic acids, thiaarenes, and PAHs (see below). Kayhanian et al. (2001) documented oil and
grease in construction site stormwaters.
Petroleum hydrocarbons tend to adsorb to soil and sediment particles and travel with them as they erode
and settle in surface waters. Petroleum hydrocarbons discharged directly to water spread quickly to cover
water surfaces. Microbial activity in surface waters eventually degrades the more structurally simple
components of petroleum materials (e.g., 40-80 percent of a crude oil) (Hoffman et al. 1995).
Petroleum hydrocarbons can affect aquatic organisms through direct toxicity, smothering, and changes in
water quality. Water quality impacts can include light reduction, pH alteration, dissolved oxygen
reduction, and dissolution of some product components in the water column. Oil, grease, fuel, and other
petroleum hydrocarbons typically contain toxic and carcinogenic constituents. Organisms may be
impacted directly or through impacts to food sources (USFWS 2005, Hoffman et al.  1995).
Fish can be exposed to petroleum hydrocarbons through contact with dissolved fractions in the water
column, particulate fractions and contaminated sediments and ingestion of contaminated food and water.
Floating egg masses can come into contact with surface hydrocarbon layers (Malins  and Hodgins 1981).
Garton (1977) documented a diesel spill at a construction site that resulted in a fish kill. These dynamics
affect other aquatic organisms, as well. Waterfowl can be affected by external oiling, ingestion, egg
oiling, and habitat changes.

3.3.2.3  Polycyclic Aromatic Hydrocarbons
PAHs are a large and diverse class of organic compounds deriving from both natural and anthropogenic
sources. Natural sources include forest fires, volcanic emissions, and oil seeps. Anthropogenic sources
include thermal combustion by industrial, municipal, power plant, vehicular, and household entities.
PAHs from these sources can be deposited from the atmosphere onto soil over wide  areas. They can
accumulate in soils, particularly in urbanized areas, and subsequently erode after ground disturbance.
PAHs are also found in construction materials. Coal tar-based sealcoat for driveway  and parking lot
surfaces contains very high levels of PAHs (Mahler et al.  2005; Van Metre et al. 2006) and has  been
found to contaminate stormwater and sediment in downstream surface waters (Science Daily 2009). It has
been inferred that these sealcoat products are contributing to PAH contamination of many U.S.  lakes.
November 2009                                                                                 3-14

-------
Environmental Impact and Benefits Assessment for the C&D Category
PAHs are also found in oil, asphalt, and tar products used for pavement and roofing projects, some
foundry sands, creosote wood preservatives, and shredded scrap tires used as a pavement additive. The
PAHs in asphalt include sulfur-containing PAHs, such as dibenzothiophene, which are more water
soluble and bioaccumulative than some other PAHs (NRC 2008). Several studies have noted road
construction sites as possible contributors of PAHs to surface waters (Ohio EPA 1998d,  1998e, 1999b).
A number of commonly used solvents (e.g., toluene, trichloroethane, and dichloroethane) can transport
PAHs and have frequently contaminated soil and groundwater at industrial and urban sites. Historic
industrial sites may contain PAHs from other sources as well, such as disposal of coal tar and combustion
by-products. Historic fill materials on industrial and urban sites may also contain PAHs.
In general, PAHs have low solubility in water, high melting and boiling points, and low vapor pressure
(Hoffman et al.  1995). PAHs are persistent in the environment and do not break down easily in water.
They also have bioaccumulation potential (USEPA 2008d). In the aquatic environment, most PAHs
associate with soil and sediment particles rather than partitioning to the water column. Their adsorption to
sediment increases with increasing organic content of the sediment. PAHs may accumulate in sediments
to concentrations much higher than those measured in the water column.
PAHs can adversely affect mammals (including human beings), birds,  fish, amphibians,  invertebrates, and
plants. PAHs at low concentrations can stimulate or inhibit growth of aquatic bacteria and algae. At
higher concentrations, however, they interfere with cell division and photosynthesis (Eisler  1987). Effects
on aquatic invertebrates include inhibited reproduction, delayed emergence, sediment avoidance, and
mortality. Egg and larval stages are more susceptible. Effects on amphibians and reptiles include impaired
reproduction, reduced growth and development, tumors, and cancer. Effects on fish include  fin erosion,
liver abnormalities, cataracts, and immune system impairments (Eisler 1987; Van Veld et al. 1990;
Mahler et al. 2005; USFWS 2005). PAHs from application of coal tar-based pavement sealcoat have been
documented as the cause of a fish kill (Maryland DEP 2000).
PAHs can transfer to humans through consumption offish from surface waters contaminated with
sufficient levels of PAHs (USEPA 2004b). PAHs have been found to cause birth defects and liver and
blood problems in some animals (USEPA 2008d). PAHs are a known carcinogen in many animals and are
suspected to be carcinogenic in human beings as well (USEPA 2008d).

3.3.2.4  Organic Pesticides
A large number of pesticides are organic compounds. Organic pesticides have been and continue to be
used extensively on both agricultural and urban lands (e.g., building foundation and lawn pest control).
Some pesticides, including many organochlorine pesticides used historically but since banned in the
United States, create residues that are able to persist in soils for long periods of time.  These  residues can
erode when soils are disturbed. Pesticide residues on construction sites can derive both from activities
associated with construction (e.g., landscaping  and vegetation removal activity) as well as from historical
pesticide applications on urban and agricultural lands that are now being  redeveloped. Chlorpyrifos and
diazinon have been detected in construction site storm waters in California (Kayhanian et al. 2001).
Organochlorine pesticides that strongly adsorb  to clay and organic matter and can be transported by
erosion and overland runoff include l,l,l-trichloro-2,2-bis(p-chlorophenyl)ethylene DDT, aldrin, mirex,
kepone, dieldrin, endosulfan, toxaphene, lindane, heptachlor, chlordane, and difocol (Alberta ARD 2009).
Except for endosulfan, use of these pesticides was generally banned in the United States  because of their
toxicity, persistence in the environment, and/or potential to bioaccumulate. For example, even though
DDT usage peaked and began to decline in the  1960s and was banned in  1972 in the United States, DDT
November 2009                                                                                3-15

-------
Environmental Impact and Benefits Assessment for the C&D Category
and its breakdown products - DDD (l,l-dichloro-2,2-bis(p-chlorophenyl)ethane) and DDE (1,1-dichloro-
2,2-bis(p-chlorophenyl)ethylene) persists widely in many agricultural soils and surface water sediments in
current and former agricultural watersheds (New Jersey DEP 1999).
Persistence, bioaccumulation, and toxicity properties vary among organic pesticides. Some organic
pesticides in current use are generally believed to degrade more quickly once applied in the environment
and have fewer toxic effects.
Due to their hydrophobic nature, pesticides tend to adsorb to sediments and organic materials and,
therefore, tend to erode, travel, and settle with sediment particles in surface waters. Similar to other
organic compounds, pesticides are often found at low or undetectable levels in the water column, but can
accumulate to problematic levels in surface water sediments, particularly fine-grained organic materials.
Biomagnification of pesticides occurs as higher trophic level organisms, such as fish-eating birds and
mammals, ingest pesticides in prey. Ingestion can also occur during feeding activities that involve
sediment contact (e.g., probing of sediment with bill, grooming).
Once in a surface water, pesticides can remain associated with sediment or can detach and enter the water
column given appropriate conditions (Bilotta and Brazier 2008). Aquatic organisms come into contact
with pesticide residues during contact with or ingestion of contaminated water, sediment, and prey.
Herbicides can remain active even at low concentration levels and damage  aquatic vegetation (USFWS
1998). Several studies have documented morphological, developmental, and biochemical alterations in a
variety of amphibians exposed to atrazine (USFWS 2005).
Most human sediment-related exposure to contaminants is through indirect routes in which pollutants
transfer from sediments to the water column or to aquatic organisms. A number offish consumption
advisories or fishing bans are due to chlordane, DDT, and its metabolites (DDD and DDE), all of which
are commonly found in sediments.

3.3.3  Metals

Metals derive from a variety of sources on construction sites. Iron, aluminum, and manganese are
common  constituents of natural soils. Other metals can also occur naturally, but their distribution is
highly variable. Elevated concentrations of these other metals are more commonly associated with human
activity. A variety of metals can  be a concern in the aquatic environment. These include aluminum,
antimony, arsenic, cadmium,  chromium, copper, iron, lead, manganese, mercury, nickel, selenium, silver,
and zinc.
Construction equipment and materials can be sources of metals. A wide variety of metal materials are
used for roofing, pipes, and structural frameworks and supports. Galvanized metal (e.g., for pipes, fence
supports, and roofing) can elevate zinc levels in construction site stormwater runoff (Horner et al. 1990;
NRC 2008). Cement and concrete products, wastes, and equipment washwaters can have  elevated metal
levels (e.g., of chromium) due to industrial wastes incorporated during the  manufacturing process
(USEPA  2006a). High levels of aluminum and iron have also been documented in cement wastes (NRC
2008). Preserved and waterproofed wood products can leach high levels of copper (NRC  2008). One
study attributed elevated levels of construction site metal discharge (chromium) to pressure-treated wood
products  (Ohio EPA 1997d).
A variety of materials used for road and highway construction contain metals, including pressure-treated
wood, coal  fly ash, phosphogypsum,  and asphalt mixes containing municipal incinerator ash, shredded
scrap tires, foundry sand, and recycled shingles. In laboratory studies, these materials have been found to
November 2009                                                                                3-16

-------
Environmental Impact and Benefits Assessment for the C&D Category
leach heavy metals at levels toxic to water fleas (Daphnia magnet) and green algae (Selenastrum
capricornutum). Leached metals included arsenic, copper, zinc, lead, aluminum, mercury, and vanadium
(USEPA 1999; NRC 2001c; Eldin 2002). Although adsorption to soil particles ameliorates the metal
leachates' toxic effects (NRC 2001), these soil particles may subsequently erode from construction sites
and enter surface waters. Loose particles of road and highway construction materials can also erode and
enter surface waters.
Landscaping materials can contain metals, as well. Some fertilizers contain elevated levels of metals due
to industrial wastes incorporated during their manufacture. Glanville et al. (2004) documented elevated
levels of several metals in composted biosolids, yard waste,  and bioindustrial waste relative to topsoil.
Compost can be used to reduce total erosion (and therefore total loadings of metals) from construction
sites. However, if compost products contain elevated metal levels and are allowed to erode, they can
contribute metals to construction site discharges. Metals content varies widely among composted products
and is heavily dependent on the source materials used.
Historic contamination can also be a source of metals. Lead  is a common soil contaminant near roads and
in urbanized areas due to its use in lead-based paint and as a gasoline additive for many years. Soil in fruit
orchards active prior to 1950 are frequently contaminated with high levels of lead and arsenic due to
widespread use of lead arsenate pesticides during the late 19th and early 20th centuries (New Jersey DEP
1999). These pesticides were also widely used for golf course and lawn care  (Folkes et al. 2001). Soils on
lands used to grow cotton prior to 1950 often contain high levels of arsenic due to use of arsenic acids to
defoliate cotton prior to harvest. Lead and arsenic  pesticide residues do not decay overtime and remain in
surface soil layers until a disturbance, such as construction activity, mobilizes them through erosion.
Metals have been documented in construction site stormwaters and at elevated levels in downstream
receiving waters due to construction activity. Documented metals include aluminum (Huckabee et al.
1975), cadmium (Kayhanian et al. 2001), chromium (Kayhanian et al. 2001), copper (Horner et al. 1990;
Barrett et al. 1995a; Kayhanian et al. 2001; Wong 2005; Clausen 2007), iron (Extence 1978; Shields and
Sanders 1986; Barrett et al. 1995; Chen et al. 2009), lanthanum (Huckabee et al. 1975), lead (Shields and
Sanders 1986; Horner et al. 1990; Kayhanian et al. 2001; Clausen 2007), manganese (Huckabee et al.
1975; Shields and Sanders 1986), nickel (Kayhanian et al. 2001), samarium (Huckabee et al. 1975), silver
(Kayhanian et al. 2001), and zinc (Huckabee et al. 1975; Shields and Sanders 1986; Horner et al. 1990;
Barrett et al. 1995a; Kayhanian et al. 2001; Clausen 2007). Some studies attributed metals to natural
constituents in site soils or groundwater (Huckabee et al. 1975; Extence 1978; Shields and Sanders 1986;
Yew and Makowski 1989; Barrett et al. 1995a). In many studies, the source of the metals was not
identified. The presence and level of metals is highly variable among site discharges. Some studies
examined but detected no elevation in metals levels in receiving waters downstream of construction sites
(Peterson andNyquist 1972).
Heavy metals can form complexes with clay minerals and organic matter in soil or can be a component of
particulate matter in soil. Organic matter, in particular, has a high capacity to bind with metals.  Sediments
with higher levels of organic matter typically have higher concentrations of zinc, lead, chromium, copper,
mercury, and cadmium (Paul and Meyer 2001).
Metals also tend to associate with finer sized sediment particles. Because stormwater tends to transport
finer particles from sites, the sediment in stormwater tends to be enriched in metals relative to the parent
soil. Novotny and Chesters (1989) discussed metals that can be present at elevated levels in sediment
relative to the original soil, including lead, copper, zinc, aluminum, iron, chromium, and  nickel.
November 2009                                                                                 3-17

-------
Environmental Impact and Benefits Assessment for the C&D Category
Metals also show a strong correlation with suspended solids discharges. One study of construction site
discharges noted strong correlations between total suspended solids (TSS) and particulate copper, zinc,
and chromium discharge levels, suggesting that the metals travel with sediment as it erodes (Kayhanian et
al. 2001). Barrett et al. (1995a) found strong correlations between iron and TSS discharge concentrations
downstream of a construction site. Barrett et al. (1995b) stated that lead and iron are the metals most
strongly associated with particulate matter, followed by copper and cadmium.
Metals tend to adsorb to sediment in the aquatic environment. Because metals do not biodegrade, they can
accumulate on waterbody beds as they settle from the water column with  sediments. Due to this
accumulation in bottom sediments over time, metals levels may often be below levels of concern or
detection in the water column but reach levels high enough to have adverse impacts on organisms living
in or near bottom sediments. For example, lead concentrations tend to be highest in benthic organisms,
perhaps due to feeding among bottom sediments (USFWS 1998). Metals can also desorb from sediments
and enter the water column under appropriate conditions (e.g., anaerobic conditions) (Bilotta and Brazier
2008).
During dredging or flooding, sediment containing metals may be reintroduced into the water column
because of the disturbance. Metals attached to suspended sediment can be transported downstream,
potentially contaminating areas far from their point of origin. Disturbance of contaminated sediments can
also release metal contaminants into water drawn from surface waters for drinking water supply. When
present at sufficiently high levels, metals can cause dredged sediment to be classified as toxic, making its
disposal more costly.
A variety of organisms in streams in urbanized areas have exhibited elevated metal concentrations,
including algae, mollusks, arthropods, and annelids. Organisms are directly exposed to both dissolved
metals in the water column and metals associated with ingested sediments and organic matter. The
potential for toxicity depends on the bioavailability of the metal, the means and length of organism
exposure, and the life stage exposed. The bioavailability of the metal depends on the nature of the metal
discharge and the aquatic environment.
Acute and chronic ambient water quality criteria have been established for most heavy metals and contain
adjustment factors to reflect dissolved and particulate metal forms and water column hardness (USEPA
2007d). A recent examination of the water quality criteria for copper by EPA uses the Biotic Ligand
Model (BLM) and incorporates data on receiving water levels of dissolved organic carbon, pH, major
cations and anions, alkalinity, and temperature. The presence of other types of pollutants or materials can
be important as some act additively, synergistically, or antagonistically with certain metals. For example,
the bioavailability of cationic metals in sediments decreases with increasing organic content and the
presence of binding sulfide compounds (USEPA 2005b).
While trace amounts of certain metals are necessary for normal biological functions in many aquatic
species, higher concentrations can be acutely or chronically toxic and adversely affect behavior, growth,
development, metabolism, reproductive success, and survival in aquatic species, waterfowl, and
mammals. Exposure to metals has been shown to reduce photosynthetic efficiency and algal colonization,
lowering primary production rates and diminishing food available to  other organisms. Elevated metal
levels can inhibit plant growth and can have adverse effects on the health  of benthic organisms
(Masterson and Bannerman 1994). Metals can be absorbed through skin, gills, intestines, and other
organs. Early organism life stages and organisms with longer exposure durations are more vulnerable
(USFWS 2005).  Some metals are able to progress and bioaccumulate through the food chain, imposing
November 2009                                                                                 3-18

-------
Environmental Impact and Benefits Assessment for the C&D Category
heavier metal exposures on organisms higher in the tropic structure. Impacts from elevated metals levels
can reduce organism abundance and alter community structure.
Iron and manganese are more common and less toxic to aquatic organisms than heavy metals. They are
highly sensitive to redox conditions and are more soluble under low dissolved oxygen conditions. High
levels of dissolved iron and manganese can be found seasonally in anaerobic bottom waters of stratified
lakes. High levels can also be found in low-oxygen groundwaters. In solution, these metals can create
highly colored water. Low-oxygen groundwaters containing high levels of iron and/or manganese may be
discharged from construction sites due to grading, excavation, and dewatering activities. If these
groundwaters are discharged to surface waters, aeration and rapid oxidation can deposit large amounts of
red iron oxides or black manganese oxides on surface water substrates. These deposits can coat stream
substrate and aquatic vegetation, fill interstitial spaces, and create turbid, highly colored water. Chisholm
and Downs (1978) discuss the discharge of such groundwater from a construction site in West Virginia.

3.3.4  Dissolved Inorganic Ions

A number of dissolved inorganic ions can discharge from construction sites, including calcium, chloride,
sodium, potassium, magnesium, sulfate, and bicarbonate. These ions are naturally occurring components
of many soils and, in some cases, of precipitation.  Levels vary geographically. Elevated levels relative to
freshwater surface waters can be found in some groundwater. Construction site activity may increase
discharge of ions to  surface waters through soil disturbance and erosion, groundwater disturbance, use of
dust suppressants (e.g., calcium chloride), fertilizers (which may contain chloride, potassium, or other
inorganic ions), lime (calcium carbonate), cement, concrete, and other building materials. High levels of
calcium, magnesium, and sodium have been documented in cement wastes (NRC 2008).
Several studies have documented the presence of dissolved inorganic ions in construction site
stormwaters and at elevated levels in downstream receiving waters due to construction activity. Ions that
have been documented include calcium (Huckabee et al. 1975; Tan and Thirumurthi 1978; Nodvin et al.
1986; Shields and Sanders 1986), bicarbonate (Tan and Thirumurthi 1978; Foassati et al. 2001), chloride
(Tan and Thirumurthi 1978; Shields and Sanders 1986; Chen et al. 2009), magnesium (Huckabee et al.
1975; Nodvin et al.  1986; Shields and Sanders  1986), potassium (Nodvin et al. 1986), sodium (Tan and
Thirumurthi 1978; Nodvin et al. 1986), and sulfate (Huckabee et al. 1975; Tan and Thirumurthi  1978;
Shields and Sanders 1986; Chen et al. 2009). Some studies monitored but did not detect changes in
dissolved inorganic  ion levels in receiving waters downstream of construction sites (Burton et al. 1976;
Cramer and Hopkins 1982; Hedrick et al. 2006).
Shields and Sanders (1986) and Chen et al. (2009) identified soil erosion as the source of ions. Chisholm
and Downs (1978) identified construction site use  of a calcium chloride dust suppressant as the source.
Huckabee et al. (1975) identified acid rock drainage as the source. Nodvin et al. (1986) documented
elevated levels of magnesium, sodium, and potassium in a stream due to leaching of material from
concrete freshly poured for bridge footers.
Dissolved inorganic ions generally exhibit low  toxicity to aquatic organisms. However, ions can influence
several surface water attributes including salinity and pH (discussed below). They can also influence the
toxicity of co-discharged metals. For example, large concentrations of calcium in surface waters can
decrease toxicity of several metals, including aluminum (Fredman 1995). In wetlands and anaerobic
sediments, however, microbial processes  can exchange calcium ions with mercury and facilitate the
release of mercury into these systems.
November 2009                                                                                3-19

-------
Environmental Impact and Benefits Assessment for the C&D Category
Elevated or widely fluctuating chloride levels can impair aquatic organism growth, reproduction, and
survival (New Hampshire DBS 2007). Some aquatic plants are sensitive to salinity changes (e.g., some
studies of the genus Potamogeton have shown it to be very sensitive to water column salinity and mineral
or ionic profile changes (USFWS 1998)). Irrigation water that contains dissolved salts or pollutants can
harm crops and damage soil quality (Clark et al. 1985).
Depending on whether fish are fresh or salt water species, they have been reported to tolerate chloride
levels of 400 to 30,000 mg/L. EPA has set an acute freshwater criterion for chloride at 680 mg/L and a
chronic criterion at 230 mg/L (USEPA 1988). Chloride toxicity increases when it co-occurs with other
ions such as potassium or magnesium. Dissolved inorganic ions can also affect organism osmoregulation.
Some sensitive species (e.g., mayflies) have been lost from surface waters due to elevated ion levels
(NRC 2008).

3.3.5   pH Level

Construction sites use materials that can alter surface water pH levels, including large quantities of
cement and concrete. Wastes from concrete truck and other equipment washouts have a high pH (12 pH
units) and can alter surface water pH if discharged at sufficient levels (Canadian DFO 2009). Storm water
can also become  elevated in pH if it comes into contact with freshly placed concrete. Sand-cement bags
may be used as temporary headwalls. Nodvin et al.  (1986) documented elevated pH levels in a stream in
Nevada downstream of a bridge construction site. pH levels increased from 6.83 to a maximum of 11.22.
The authors postulated that leaching of calcium oxide hydration products from the concrete poured for the
bridge footers was the source.
Some materials used on construction sites for soil stabilization and fill are also high in pH. These include
lime and cement materials, cement kiln dust, lime kiln dust, and crushed limestone. Lime can also be
added to soils as  a fertilizer.
Natural site soil and rock constituents, such as limestone or sulfide-containing rock and soil, can also alter
surface water pH if discharged at sufficient levels. Huckabee et al. (1975) documented acidification of a
small stream in Tennessee due to iron sulfides in rock used for roadbed fill. Upstream of the fill, stream
pH was 6.5 to 7.0. Downstream of the fill for several miles, stream pH was 4.5 to 5.9. Acidification took
place shortly after completion  of the project and continued to persist for more than 30 years. Yew and
Makowski (1989) also documented  stream acidification due to use of pyritic shales in road construction.
EPA (USEPA 2009b) described a highway construction site in Washington where disturbance of
sedimentary shale rock raised pH levels as high as 9.6 in stormwater and dewatered groundwater. Chen et
al. (2009) documented significant increases in sulfate levels downstream of a construction site. The
authors believed  the sulfate derived from construction site erosion and contributed to increased stream
acidity after project completion. Shields and  Sanders (1986) also attributed pH changes in downstream
receiving waters  to soil constituents.
Change in pH levels in receiving waters was highly variable among sites. In addition to the studies
described above, several studies monitored but detected no change in receiving water pH levels due to
construction activity (Peterson and Nyquist 1972; Barton 1977; Extence 1978; Lenat et al. 1981;  Cline et
al. 1982; Cramer and Hopkins 1982; Taylor and Roff 1986; Foasatti et al. 2001;  Hedrick et al. 2006).
Aquatic communities have low species richness at pH levels above or below 6 to 8.5 units (Kalff 2002).
Atmospheric acid deposition has been documented to eliminate species  as pH decreases and to extirpate
fish from lakes once long-term levels fall below 5 pH units. At high (alkaline) pH levels, LC50 (lethal
concentration) values for salmonids fall in the 9-10 pH unit range. Few fish survive even short-term
November 2009                                                                                3-20

-------
Environmental Impact and Benefits Assessment for the C&D Category
exposures to pH levels greater than 11 (Alabaster and Lloyd 1980). Some aquatic plants are also sensitive
to pH changes (e.g., some studies of the genus Potamogeton have shown it to be very sensitive to water
column pH changes) (USFWS 1998). Low species richness at high or low pH levels can be attributable,
in some cases, to other factors working in concert with altered pH levels. These factors can include
increased metal toxicity (e.g., aluminum), high salt levels, and high water temperature.

3.3.6  Pathogens

Pathogens may be present on construction sites due to the use of manure products for fertilization and
landscaping and improper disposal of on-site sanitary wastes. Pathogens are transported by attachment to
sediment particles. The transport and fate of bacterial indicator species Escherichia coli and Salmonella
spp. have been shown to be highly influenced by their relationship with flocculated suspended sediment
and bed sediments in a river (Droppo et al. 2009). Bacteria counts were consistently higher within
sediments than within the water column. Bed sediments were found to represent a possible reservoir of
pathogens for subsequent remobilization and transport
Fungi can also function as pathogens. Aspergillus sydowii, a common terrestrial  fungus, can enter marine
environments through upstream erosion. This fungus has emerged as a widespread sea fan pathogen in the
Caribbean. Another fungus dispersed in eroded soil is known to cause a disease called "valley fever" in
California sea otters (Lafferty et al. 2004).
November 2009                                                                                3-21

-------
Environmental Impact and Benefits Assessment for the C&D Category
4     Summary of Literature on  Construction Site Discharges to
Surface Waters
A large number of publicly available studies issued in peer-reviewed, government, and other publications
document pollutant discharges from construction sites to surface waters. EPA reviewed many of these
studies and refers to the information they contain throughout this document. This chapter provides an
overview of the studies EPA reviewed (Section 4.1) as well as a brief summary of each individual study
(Section 4.2). Section 4.3 presents state reports of construction site discharges.

4.1    Overview of Impacts in  Literature

Studies documenting the results of research on construction site discharges have been published for many
years. The earliest paper summarized in this chapter was published in 1959, and the most recent papers
were published in 2009. Research characterizing the nature and magnitude of construction site discharges
is ongoing.
Available studies cover a range of construction types, geographic locations, and surface water types.
Studies of the impact of highway construction are more numerous than studies of impacts from other
types of construction. Studies of impacts to streams and small rivers are more numerous than studies of
impacts to other types of surface waters.  Impacts to marine systems are poorly characterized. EPA was
unable to locate an explanation for these  aspects of the literature. One possible explanation for the
frequency of stream ecosystems in the  literature is the high frequency with which they are affected by
construction site discharges and the ease with which they can be divided into "altered" and "unaltered"
zones (i.e., controls) for study of construction site discharge effects.
Most of the studies EPA chose for review examined surface water impacts downstream of construction
sites. EPA also reviewed a smaller number of studies characterizing the nature of construction site
stormwaters prior to their discharge to  receiving waters.
The literature indicates that construction sites can affect a number of stormwater and receiving water
parameters, including physical, chemical, and biological parameters.  Most studies, however, characterized
only a subset of all potentially affected parameters, perhaps due to resource limitations. Suspended
sediment and turbidity were the most commonly monitored parameters. Basic water quality parameters
(pH, dissolved oxygen, some major dissolved ions, biological oxygen demand (BOD)), nutrients, metals,
and toxic organic compounds were also monitored. Nutrients were monitored in a wide variety of forms,
though only a subset of possible forms was studied in most cases. Likewise, when metals were monitored,
only a subset of possible metals was typically chosen for observation. Specific organic compounds were
rarely monitored. Summaries of study  findings for specific pollutants are provided in Chapters 2 and 3.
Studies varied in whether they measured in-stream pollutant concentrations, total pollutant loads, or both.
Measurement of pollutant concentrations was more common than measurement of pollutant loads. Each
measurement provides a different type of information. Measurement  of pollutant concentration is useful
for determining certain acute and chronic effects associated with individual discharge events (e.g., surface
water appearance, organism toxicity). Total load measurements provide information on the cumulative
effects of a discharge source and can be particularly important for pollutants such as sediment, metals,
and certain toxic and persistent organic compounds that can settle and persist in surface waters for several
or more years.
November 2009
                                                                                            4-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
A complication with the measurement of pollutant concentrations in construction site storm water is that
they can be diluted by the greater volume of stormwater runoff generated by many construction sites. An
observer relying solely on pollutant concentration data to determine the effects of construction activity on
stormwater discharges may incorrectly conclude that a construction site has no impact because
concentration measurements do not reflect the cumulative impact of the total volume of runoff from the
site. For example, Clausen (2007) observed that copper, lead, and zinc concentrations in stormwater
runoff were unchanged by construction activity at one site, even though the total load of these pollutants
discharged from the site increased.
Physical stream condition was discussed in a number of studies, particularly when sedimentation impacts
were noted. Observed impacts included increased substrate embeddedness, complete burial of substrate
beneath a sediment layer, elimination of pools and other stream microhabitats, changes in channel
morphology, and loss of stream baseflow.
Biological parameters were monitored less frequently than water quality parameters. Macroinvertebrates,
particularly benthic macroinvertebrates, were monitored most frequently, followed by monitoring offish.
Only a few studies examined impacts to primary producers or animals other than macroinvertebrates or
fish (e.g., amphibians, mollusks, reptiles). Impacts to organisms were noted in most studies in which they
were monitored. Impacts included reduction of abundance, diversity, and biomass. Increased drift and
avoidance behavior, change in community composition, and mortality were also documented.
Other study limitations included limits on the spatial and temporal extent of monitoring. Few studies were
designed to characterize the full spatial extent of impacts from construction sites' discharges. Most
studies examined conditions within surface waters less than a mile downstream of construction sites.
Several studies documented surface water impacts several miles downstream of construction sites (King
and Ball 1964; Wolman and Shick 1967; Vice et al. 1969; Huckabee et al.  1975; Hainly 1980; Nodvin et
al. 1986; Ohio EPA 1997f). One study documented impacts as far as 56 miles downstream (Fossati et al.
2001).
Many studies were also limited in the length of time they were able to monitor, and most were not
designed to characterize the full temporal extent of impacts. Most monitoring took place during active
construction. A number of studies included monitoring during pre-construction and post-construction
periods, as well. Elevated suspended sediment and turbidity concentrations in downstream surface waters
were generally noted to decline within a year or less of construction activity cessation and site surface
stabilization with sufficient vegetation.
A number of studies documented longer-term impacts from construction site discharges. These impacts
included elevated sedimentation of surface water substrates (Guy 1963; Chisholm and Downs 1978; Tsui
and McCart 1981; Cline et al.  1982; Taylor and Roff; Stout and Coburn 1989; Barrett et al. 1995a; Reid
and Anderson 1999).  Elevated sedimentation levels in areas immediately downstream of construction
sites were typically documented as requiring one  to several years to clear. Monitoring of sedimentation of
downstream surface water substrates as construction sediments migrated downstream was typically not
included in the studies. Lee et al. (2009) noted that many years may be required after completion of
construction in a watershed for construction-associated sediments to fully migrate downstream.
Longer-term effects to macroinvertebrate and fish populations were also documented (Peterson and
Nyquist 1972; Huckabee et al. 1975; Reed 1977;  Chisholm and Downs 1978; Tsui and McCart 1981;
Cline  et al. 1982; Taylor and Roff 1986; Reid and Anderson 1999). Macroinvertebrate population
improvements were usually noted to  occur as excess sedimentation was flushed from surface water
substrates, and fish population improvements were usually noted as substrate and macroinvertebrate
November 2009                                                                                 4-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
conditions improved. Recovery was often documented to require more than a year following construction
completion, with some studies noting recovery as still incomplete several years after construction
cessation (Taylor and Roff 1986; Montgomery County DEP 2009). In one case, damage was still noted 30
years after construction cessation because of stream acidification from iron sulfide rocks used as roadbed
fill (Kucken et al. 1994).
Implementation of sediment and erosion controls varied widely among sites described in the studies,
though a large number of sites, including many in older  studies, utilized them to some degree.
Requirements for use of sediment and erosion controls on construction sites have gradually become more
stringent overtime. Older studies tend to document discharges from sites with lower levels of sediment
erosion and control. However, although older studies often do not reflect current requirements, they do
provide valuable information on the nature of pollutant discharges from construction sites and their
effects on surface waters. In addition, current practices vary among states and localities and are not
always successfully implemented (particularly during large precipitation events) or maintained over time.
Because current requirements have reduced, rather than  eliminated, construction site discharges,
information from older studies continues to be relevant for full evaluation of potential impacts.
Construction  site discharges are dependent on site conditions (e.g., slope, vegetative cover, soil type),
rainfall, nature of construction activity, presence and effectiveness of sediment and erosion abatement
practices, and other factors. Given the large number of construction sites active every year (greater than
80,000) and limitations in resources available for construction site discharge and impact study and
documentation, the collection of studies summarized here only partially describes the nature of
construction site discharges taking place each year in the United States.
To quantify current national water quality impacts from  construction site sediment discharges and how
they would decrease under various regulatory options, EPA used several methodologies, including
Spatially Referenced Regressions on Watershed Attributes (SPARROW) models, to estimate surface
water sediment, nitrogen, and phosphorus levels deriving from construction activity. These
methodologies and the results of the analyses are described in more detail in Chapter 6 of this document.
Table 4-1 summarizes information from studies EPA reviewed that studied the effect of construction site
activity on physical, chemical, and/or biological parameters of site stormwaters or receiving waters.
Specific information on suspended sediment concentration, suspended sediment yield, and turbidity levels
are provided in Table 4-2, Table 4-3, and Table 4-4.
 "Altered Water Characteristics" as listed in Table 4-1 are water condition alterations identified by a study
as being due to construction activity. "Unaltered Water Characteristics" are water conditions identified by
a study as unchanged by construction activity. In some studies, conditions listed under "Unaltered Water
Characteristics" changed, but the cause was identified by the study as something other than construction
activity.
November 2009                                                                                  4-3

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 4-1 : Studies Documenting Construction Site Discharges to Surface Waters and Their Impacts
Study
Barrett et al.
(1995a)
Barton et al.
(1972)
Barton (1977)
Burton et al.
(1976)
Carline et al.
(2003)
Location Construction Monitoring Sediment &
Type Points Erosion Controls
„ ,,. . „, RockBerms
Texas Highway Stream o-n-n
fe J Silt Fence
Utah Highway Stream Unknown
„ , . „, Sedimentation Basin
Ontario, ,,. . Stream „, , , . .
„ , Highway „ , Straw Mulch
Canada & J Ponds ^ ,,.
Tuning
Earthen Berms
Hay Bales
Mulching
-„, ., TT. . Stream Plastic Sheeting
Flonda Highway , . „ ,. ,,•?,•
0 J Lake Sedimentation Basins
Seeding
Sod Placement
Visqueen Slope Drains
Rock Dam
Highway - Park „, Sedimentation Basin
A Stream C-H-T-
Avenue Silt Fence
Pennsylvania Vegetative Buffer
Highway - Rock „, Sedimentation Basin
T> j Stream C-H-T-
Road Silt Fence
Altered Water
Characteristics
BOD51
Chemical Oxygen Demand (COD)
Copper
Fecal Coliform1
Fecal Strep
Iron
Nitrate
Sedimentation
Total Coliform
Total Organic Carbon1
TSS
Turbidity
Volatile Suspended Solids
Zinc
Macroinvertebrates
Benthic Macroinvertebrates
Fish
Nitrate
Sedimentation
Sediment Organic Content
Suspended Sediment
Orthophosphate [load]
Phosphorus, Total Dissolved
[load]
Silicon, Dissolved [load]
Suspended Sediment [cone. +
load]
Turbidity
TSS
TSS
Unaltered Water
Characteristics
Cadmium
Chromium
Lead

Ammonia
Dissolved Oxygen
Hardness
Nitrite
PH
Phosphate
Dissolved Solids [load]
Stormwater Discharge
Benthic Macroinvertebrates
Substrate Composition
Trout Redds
Benthic Macroinvertebrates
Substrate Composition
Trout Redds
November 2009
                                                                                                                                         4-4

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 4-1 :
Study





Chen et al.
(2009)




Chisholm &
Downs (1978)








Clausen
(2007)







Cleveland &
Fashokun
(2006)
Studies Documenting Construction Site Discharges to Surface Waters and Their
Location Construction
Type




West „. ,
... . . Highway
Virginia




West
,r • • Highway
Virginia b J



Residential -
Traditional




^ 	 +._+




Residential -
Enhanced Best
Management
Practices
(BMPs)




Texas Highway
Monitoring Sediment &
Points Erosion Controls




Grass seeding
„, Mulch
Stream „ ,. ... „ .
Sedimentation Basin
Silt Fence




Stream Seeding



Storm Sewer Unspecified Practices









Earthen Berm
„., . Hay Bales
Ditch ci'li-n
Silt Fence
Topsoil stockpile cover




Ditch Rock Filter Dam
Altered Water
Characteristics
Acidity
Alkalinity (possible)
Calcium (possible)
Chloride
Conductivity (possible)
Iron
Macroinvertebrates
Nitrate
Sulfate
TSS
Turbidity
Dissolved Solids
Macroinvertebrates
Sedimentation
Stream Substrate
Ammonia [yield]
Copper [yield]
Lead [yield]
Nitrate + Nitrite [yield]
Runoff Volume
TKN [cone.]1
TKN [yield]
Total Phosphorus [cone.]
Total Phosphorus [yield]
TSS [yield]
Zinc [yield]
Ammonia [cone.]
Copper [cone.]
Lead [cone.]

Nitrate + Nitrite [cone.]
Runoff Volume1
TKN [cone.]
Total Phosphorus [cone.]
Total Phosphorus [yield]
TSS -[cone.]
TSS -[yield]
Zinc [yield]1

TSS
Phosphorus
Impacts
Unaltered Water
Characteristics




Ammonia
Phosphate
Temperature
Water Discharge








Ammonia [cone.]
Copper [cone.]
Lead [cone.]
Nitrate + Nitrite [cone.]
TSS [cone ]
Zinc [cone ]




Ammonia [yield]

BOD [cone ]
Copper [yield]
Fecal Coliform [cone.]
Lead [yield]
Nitrogen [yield]
TKN [yield]

Zinc [cone.]

Ammonia
Nitrite
Turbidity

November 2009
                                                                                                                                         4-5

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 4-1 : Studies Documenting Construction Site Discharges to Surface Waters and Their
Study
Cline et al.
(1982)
Cramer &
Hopkins (1982)
Downs &
Appell986
Duck (1985)
Eckhardt et al.
(1976)
Embler &
Fletcher (1983)
Extence(1978)
Fossati et al.
(2001)
Garton(1977)
Guy (1963)
Hainly(1980)
Location
Colorado
Louisiana
West
Virginia
Scotland,
Great Britain
Pennsylvania
South
Carolina
Great Britain
Bolivia
West
Virginia
Maryland
Pennsylvania
Construction
Type
Highway
Highway
Highway
Road
Highway
Unknown
Highway
Highway
Highway
Residential
Highway
Monitoring
Points
Stream
Wetland
Stream
Stream
Lake
Stream
Stream
Stream
River
River
Spring
Stream
Stream
Sediment &
Erosion Controls
Unknown
Unknown
Unknown
Unknown
Mulching
Seeding
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Altered Water
Characteristics
Epilithon
Macroinvertebrates
Substrate composition
TSS
Color
Turbidity
Suspended Sediment
Sedimentation
Sediment Load
Suspended Sediment
Turbidity
Suspended Sediment
Suspended Sediment
Turbidity
Iron
Macroinvertebrates
Sedimentation
Suspended Sediment
Bicarbonates
Macroinvertebrates
Phosphates
Sedimentation
Suspended Sediment
Turbidity
Fish
Petroleum Hydrocarbons
Suspended Sediment
Sedimentation
Suspended Sediment
Suspended Sediment
Impacts
Unaltered Water
Characteristics
Dissolved Oxygen
PH
Total Dissolved Solids
Dissolved Oxygen
pH
Salinity




BOD
Nitrate
Orthophosphate
pH
Calcium
Chlorides
Conductivity
Magnesium
Nitrates
PH
Potassium
Sodium
Sulfates
Temperature




November 2009
                                                                                                                                         4-6

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 4-1 : Studies Documenting Construction Site Discharges to Surface Waters and Their Impacts
Study


Hedrick et al.
(2006)


Hedrick et al.
(2007)
Helm (1978)




Helsel(1985)










Huckabee et al.
(1975)







Hunt & Grow
(2001)

Keller (1962)
Location Construction Monitoring Sediment &
Type Points Erosion Controls

Buffered Disposal of
^ TT. . „, Acidic Rock
Tennessee Highway Stream „ . „, . „
& J Rock Check Dams


West
,r • • Highway Streams Silt Fence
Virginia & 3
Pennsylvania Highway Stream Unknown
Benches
Dams
Dikes
Excelsior Matting
„. . TT. . Stream Hay Bales
Ohio Highway „. T / , , ...
fe J River Jute Matting
Mulching
Rock-Lined Channels
Sedimentation Basins
Seeding






North ,,. . „, ,, .
„ .. Highway Stream Unknown







Ohio Unknown Stream None

Maryland Multiple River Unknown
Altered Water
Characteristics

None


Benthic Macroinvertebrates (short
term)
Sediment Yield
Suspended Sediment




Suspended Sediment




Aluminum
Bicarbonate
Calcium
Fish
Lanthanum
Magnesium
Manganese
Periphyton
PH
Salamander
Samarium
Stream Substrate
Sulfate
Zinc
Fish
Habitat Quality Indices
Sedimentation
Substrate Embeddedness
Suspended Sediment
Unaltered Water
Characteristics

Conductivity
Magnesium
pH
Suspended Sediment
Benthic Macroinvertebrates
(long term)
Sediment Size Fractions











Arsenic
Cadmium
Chloride
Cobalt
Copper
Iron
Lead
Mercury
Phosphate
Potassium
Selenium
Sodium





November 2009
                                                                                                                                         4-7

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 4-1 : Studies Documenting Construction Site Discharges to Surface Waters and Their Impacts
Study Location
King & Ball . , .
(1964) Micnigan
Lee et al. „
(2009) Kansas
Lenat et al. North
(1981) Carolina
T • /innrA North
Line (2009) „ ..
Carolina
Line & White North
(2007) Carolina
Lubliner& „, . . ,
/-. u- ^nr\c\ Washington
Goldmg (2005) fe
Construction
Type
Highway
Commercial
Highway
Residential
Highway
Highway
Residential
Commercial
Residential
Transportation
Utility
Monitoring Sediment &
Points Erosion Controls
River Unknown
Stream Unknown
„ , Sedimentation Basin
Stream „ ... „
Silt Fence
Coir Baffles
Flocculent
Mulching
Rock Dams
„, Sediment Traps
Streams „ ,. . „ .
, . Sedimentation Basins
Lake „ , .
Seeding
Silt Fence
Skimmer Outlet
Slope Drains
Turbidity Curtain
Stream Sediment Trap
Blanket
Erosion Control
„ , Mulch Storm Drain
Stream _ , ,.
Protection
Sedimentation Basin
Vegetation
Altered Water Unaltered Water
Characteristics Characteristics
Fish
Macroinvertebrates
Primary Production
Sedimentation
Suspended Sediment
Turbidity
Sediment Yield
Macroinvertebrates Dissolved Oxygen
Substrate Composition pH
Suspended Sediment Temperature
Suspended Sediment Yield
Turbidity
Ammonia
Nitrate
Peak Runoff
Runoff- Rainfall Ratio
Stream Baseflow
TKN
Total Nitrogen
Total Phosphorus
TSS
Transparency
TSS
Turbidity
November 2009
                                                                                                                                         4-8

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 4-1: Studies Documenting Construction Site Discharges to
Study



Montgomery
County DEP
(2009)





Nodvin et al.
(1986)





Ohio EPA
(1997 f)


Owens et al.
(2000)





Peterson &
Nyquist(1972)






Location Construction
Type


Maryland Multiple





California Bridge





„. . Golf Course
Ohio „ ., ...
Residential


„,. . Residential
Wisconsin „ . .
Commercial





Alaska Highway
Bndge






Monitoring Sediment &
Points Erosion Controls
Baffles
Basin Forebays
Limit Disturbance
„ , Oversized + Dual
Streams _ „ , , „ •
Perforated Risers
Sedimentation Basin -
Skimmers
Stabilization




Stream TT .
T . Unknown
Lake





Stream Sedimentation Basin


Stormwater , , , . . , , . , ,
„ . , Not included in study
Discharge





Stream Unknown






Surface Waters and Their
Altered Water
Characteristics
Channel Aggradation
Channel Sinuosity1
Fish
Macroinvertebrates
Sedimentation
Stream "Flashiness"
Substrate Embeddedness
TKN
TSS
Alkalinity
Calcium
Conductivity
Magnesium
PH
Potassium
Sodium
Turbidity
Fish
Macroinvertebrates
Mollusks
Sedimentation
Substrate Embeddedness
Turbidity
Suspended Sediment





Conductivity
Macroinvertebrates
Turbidity






Impacts
Unaltered Water
Characteristics


















Alkalinity
Ammonia
Calcium
Carbon Dioxide
Dissolved Oxygen
Iron
Magnesium
Nitrate
Nitrite
Orthophosphate
pH
Silica
Sulfate
Temperature

November 2009
                                                                                                                                         4-9

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 4-1 : Studies Documenting Construction Site Discharges to Surface Waters and Their
Study



Reed (1977)



Reed (1980)



Reid &
Anderson
M999Ni
\1777)




Selbig et al.
(2004)


Selbig &
Bannerman



Location Construction Monitoring Sediment &
Type Points Erosion Controls
Hay Bales
Mulching
Virginia Highway Streams Rock Dam
Sedimentation Basin
Seeding
Mulching

Pennsylvania Highway Streams „ ,. ... „ .
J o j Sedimentation Basins
Seeding


Multiple
Sites in „. .. Streams TT .
,T ,. Pipeline „. Unknown
North F Rivers
America


Construction Phasing
Deep Tilling
Earthen Berms
Inlet Protection
Wisconsin Residential Stream Silt Fence
Stockpile Seeding
Stone Tracking Pads
Straw Bales
Vegetative Buffers
Earthen Berms
Erosion Fabric
„,. . „ ., ,. . Stormwater Rock Trenches
Wisconsin Residential „ . „ , . ,-,-.•
Basin Sediment Basins

Seeding
Silt Fence
Altered Water
Characteristics

Fish
Macroinvertebrates



Suspended Sediment

Benthic Macroinvertebrates
Channel Morphology
Fish
Sedimentation
Substrate Composition
Substrate Embeddedness
Suspended Sediment
Turbidity


Runoff Volume
Suspended Sediment
Total Sediment


Phosphorus - total
Runoff Volume
Total Solids

TSS

Impacts
Unaltered Water
Characteristics
















Fine Sediment
Fish
Habitat Quality
Macroinvertebrates
Stream Morphology
Temperature






November 2009
                                                                                                                                        4-10

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 4-1 : Studies Documenting Construction Site Discharges to Surface Waters and Their
Study
Shields &
Sanders (1986)
Stephens et al.
(1996)
Tan&
Thirumurthi
(1978)
Taylor & Roff
(1986)
Tsui & McCart
(1981)
Vice et al.
(1969)
Walling &
Gregory (1970)
Ward & Appel
(1988)
Wark & Keller
(1963)
Location
Mississippi
Utah
Nova Scotia,
Canada
Ontario
British
Columbia,
Canada
Virginia
Devon,
England
West
Virginia
Maryland,
Virginia, and
West
Virginia
Construction
Type
Waterway
Highway
Highway
Highway
Pipeline
Highway
Residential
Highway
Multiple
Monitoring
Points
Stream
Stream
Reservoir
Lake
Stream
Stream
Stream
Stream
Stream
Streams
Rivers
Sediment &
Erosion Controls
Polymer Flocculant
Sedimentation Basin
Disposal Area Design

Diversion Channel
Sedimentation Basin
Sedimentation Basin
Straw Mulch
Turfing
Unknown
Unknown
Unknown
Unknown
Unknown
Altered Water
Characteristics
Multiple2

Conductivity
Nitrate + Nitrite
Total Dissolve Solids
Turbidity
Fish
Macroinvertebrates
Macrophytes
Nitrate
Sedimentation
Sediment Organic Content
Suspended Sediment
Macroinvertebrates
Sedimentation
TSS
Turbidity
Suspended Sediment
Turbidity
Suspended Sediment
Suspended Sediment
Suspended Sediment
Impacts
Unaltered Water
Characteristics
Arsenic - total [cone.]
Cadmium - total [cone.]
Chromium - total [cone.]
Copper - total [cone.]
Mercury - total [cone.]
Nitrate - total [cone.]
Nitrate + Nitrite [cone.]
Organic Nitrogen [conc.[
Phosphorus - dissolved [cone.]
TKN [load]
Zinc - total [cone.]

Multiple3
Ammonia
Dissolved Oxygen
Hardness
PH
Phosphorus






November 2009
                                                                                                                                        4-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
    Table 4-1: Studies Documenting Construction Site Discharges to Surface Waters and Their Impacts	
        Study        Location     Construction    Monitoring        Sediment &                Altered Water                 Unaltered Water
                                          Type           Points       Erosion Controls             Characteristics                   Characteristics
Welsh &
Ollivier(1998)
Werner (1983)
Whitney &
Bailey (1959)
Wolman &
Schick(1967)
Wong (2005)
Yew&
Makowski
(1989)
Yorke & Herb
(1978)
Young &
Mackie(1991)
California
West
Virginia
Montana
Maryland
Hawaii
Tennessee-
North
Carolina
Border
Maryland
Northwest
Territories,
Canada
Highway
Highway
Highway
Commercial
Highway
Highway
All types
Pipeline
Streams
Spring
Stream
Stream
Stream
Reservoir
Stream
Streams
Stream
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Stream Stabilization
Vegetative Buffer
Amphibians
Fine sediment depth
Substrate Embeddedness
Fish
Suspended Sediment
Fish
Sedimentation
Stream Substrate
Multiple4
Alkalinity
Fish
Metals
PH
Suspended Sediment [yield]
Macroinvertebrates
Sedimentation
TSS




Multiple5









Dissolved Oxygen
Hardness
Nitrogen - dissolved
PH
Phosphorus -dissolved
    1 Parameter declined between baseline and construction estimates.
     Alkalinity, Ammonia- total [cone.], Biological Oxygen Demand [cone.], Calcium - dissolved [cone.], Carbonate [cone.], Chemical Oxygen Demand [cone.], Chloride [cone.], Color,
    Hardness, Iron -total [cone. AND daily loading], Lead-total [cone. AND daily loading], Magnesium - dissolved [cone.], Manganese- total [cone. AND daily loading], Nitrate + Nitrite [daily
    loading], Nitrite -total [cone.], pH, Phosphorus-total [cone.], Specific Conductance, Sulfate [cone.], Total Dissolved Solids [cone. AND daily loading], Total Kjeldahl Nitrogen (TKN) [cone.],
    Total Nitrogen [cone. AND daily loading], TSS [cone.], Turbidity, and Zinc-total [daily loading].
    3 Alkalinity, ammonia, color, coliforms, dissolved oxygen, hardness, iron, manganese, pH, potassium, silica, temperature, total organic carbon
    4 Copper (possible), Nitrite + Nitrate, Sediment Yield, Specific Conductance, Total Phosphorus, Total Suspended Solids, Turbidity.
    5 2,4-D, 2,4,5-T, Aldrin, Aluminum, Arsenic, Barium, Beryllium, Cadmium, Chlordane, Chlorpyrifos, Chromium, Cobalt, DDD, DDE, DDT, DEF, Diazinon, Dichlorprop (2,4-DP), Dieldrin,
    Dissolved Oxygen, Disulfoton, Endosulfan-alpha, Endrin, Ethion, Fonofos, Heptachlor, Heptachlor epoxide, Iron, Lead, Lindane, Lithium, Malathion, Manganese, Mercury, Methyl-Parathion,
    Methoxychlor, Mirex, Molybdenum, Nickel, Parathion, Perthane, pH, Phorate, PCB, PCN, Selenium, Silver, Silvex (2,4,5-TP) Strontium, Temperature, Total Nitrogen, Toxaphene, Trithion,
    Vanadium, Zinc.
November 2009
                                                                                                                                                                  4-12

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
Table 4-2, Table 4-3, and Table 4-4 summarize findings on suspended sediment concentrations, sediment
yield, and turbidity levels, respectively, from the studies reviewed by EPA. Sediment in construction site
stormwaters or receiving waters was measured either as a concentration (Table 4-2) or a total yield from
the site or the surface water's watershed basin (Table 4-3). Suspended sediment concentrations are
measured in milligrams per liter (mg/L). Sediment yield is total sediment load expressed as a mass of
sediment normalized by the area of the contributing watershed. It represents the mass of sediment eroded
from the land, channels, and mass wasting, minus the sediment that is redeposited prior to the point of
measurement. Sediment yield is expressed in terms of total yield or in terms of an increase due to
construction site discharge and is sometimes specified per unit area or period of time. Turbidity is
expressed in terms of NTUs (Table 4-4).
For most studies, the data presented in Table 4-2, Table 4-3, and Table 4-4 reflect the averaging of
multiple samples rather than the comparison of single grab samples. The number of samples and the
length of time over which they are averaged vary widely  among studies.
Table 4-2: Summary of Suspended Sediment Concentrations from Selected Studies
Study
Barrett etal. (1995a)
Barton (1977)
Chisholm & Downs (1978)
Clausen (2007)
Cleveland & Fashokun (2006)
Cline etal. (1982)
Duck (1985)
Embler& Fletcher (1983)
Fossati etal. (2001)
Hedrick et al. (2006)
Helsel(1985)
Lubliner & Golding (2005)
Owens et al. (2000)
Location
Texas
Ontario, Canada
West Virginia
Connecticut - Enhanced site
Connecticut - Traditional site
Texas
Colorado
Colorado
Colorado
Colorado
Scotland
South Carolina
Bolivia
Tennessee
Ohio
Washington
Washington
Washington
Wisconsin - summer
construction
Wisconsin - winter
construction
Suspended Sediment Concentration (mg/L)
Unaffected Site Affected Site
13.9
2.8
—
42
No statistical difference.
212.5
16.4
2.2
3.8
2.6
0.3-3.4
<303
4
No statistical difference.
125.0
—
—
—
100
550
79
5-50.01
3,051.2
67

1410.3
10.8
18.9
91.0
121.6
7.4-712.8
60-1 303
5-2,5744

261.0
33.7
131.2
628.7
15,000
2,400
November 2009
                                                                                            4-13

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
Table 4-2: Summary of
Study
Reed (1980)
Shields & Sanders (1986)
Tay lor &Roff( 1986)
Tsui &McCart (1981)
Wolman & Schick (1967)
Wolman & Schick (1967)
Wong (2005)7
Young &Mackie( 1991)
Suspended Sediment Concentrations from Selected Studies
Location
Pennsylvania - Site 2
Pennsylvania - Site 2A
Pennsylvania - Site 2B
Pennsylvania - Site 3
Mississippi
Ontario, Canada
British Columbia, Canada
Maryland - Cockeysville
Multiple sites in Baltimore,
Maryland and District of
Columbia region
Hawaii -Site 16265700
Hawaii -Site 16266500
Hawaii -Site 16267500
Hawaii -Site 16269500
Hawaii -Site 16270900
Hawaii -Site 16275000
Northwest Territories, Canada
Suspended Sediment
Unaffected Site
7-85
42-1006
3-55
3S-636
11-175
6S-776
8-95
36-S76
119
5-10
0-7
1,500

5
3
3
2
5
3
<2
Concentration (mg/L)
Affected Site
12-185
146-2436
9-145
100-2096
58-705
309-4086
16-305
227-28S6
347
30-206
0-10,660
80,000
3,000 to 150,000 +
10
2
12
5
2
5
>300
 1 Average concentration varied with stage of construction activity.
 2 Modeled estimate.
 3 Peak values.
 4 Study sites extended up to 56 miles downstream of construction activity.
  Average daily-mean base-flow values.
 6 Average daily-mean storm event values.
 7 Wong (2005) measured concentration as a long-term geometric mean.
Studies vary in the degree to which construction activity increased suspended sediment yield for multiple
reasons, including variability in the percentage of the watershed under study containing construction
activity, nature of construction activity and associated land disturbance, precipitation intensity, site
characteristics (e.g., soil type and slope), and sediment and erosion control practices utilized. A number of
the yields in Table 4-3 were calculated for watersheds containing other land uses besides construction
(e.g., forest, mature development, pasture, cropland). This method of calculation lowers estimated
sediment yields because it incorporates lower sediment yields associated with these other land uses.
November 2009
                                                                                                      4-14

-------
Environmental Impact and Benefits Assessment for the C&D Regulation

Table 4-3: Summary of Sediment Yield Data from Selected Studies
Study
Burton etal. (1976)
Carlineetal. (2003)
Clausen (2007)
Daniel etal. (1979)
Downs and Appel (1986)
Duck (1985)
Eckhardt(1976)
Guy (1963)
Guy and Ferguson (1962)
Helsel(1985)
Line (2009)
Line and White (2007)
Nelson and Booth (2002)
Owens et al. (2000)
Reed (1980)
Selbig et al. (2004)
Location
Florida
Pennsylvania - Rock
Road
Connecticut -
Enhanced BMP site
Connecticut -
Traditional BMP site
Wisconsin - Site G2
Wisconsin - Site G3
Wisconsin - Site G5
West Virginia
Scotland, Great
Britain
Pennsylvania
Maryland
Virginia
Ohio
North Carolina -
Tilly Up
North Carolina -
Tilly Down
North Carolina -
Ellery Up
North Carolina -
Ellery Down
North Carolina -
King's Mill
North Carolina
Washington
Wisconsin - summer
construction
Wisconsin - winter
construction
Pennsylvania -Site 2
Pennsylvania - Site
2A
Pennsylvania - Site
2B
Pennsylvania - Site 3
Wisconsin - land
disturbance phase
Wisconsin - house
construction phase
Sediment
Unaffected Site


0.0021
0.00081
-


0.2
0.03

-
-
0.7-1.2
0.01
0.07
0.04
0.20
0.09
0.16
0.11
0.18
0.18
0.17
0.12
0.21
0.20
57.6
9.7
Yield
Affected Site
Up to 67
5
0.024
0.10
12.27
7.58
5.84
9.7
3.8
62-103
39
2
15-25
7.3
3.50
2.02
1.35
1.63
2.80
0.43
3.38
0.82
8
6
7
9.5
76.2
10.4
Units
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/storm event
tons/storm event

November 2009
                                                                                                  4-15

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
 Table 4-3: Summary of Sediment Yield Data from Selected Studies
           Study
    Location
        Sediment Yield
Unaffected Site    Affected Site
Units
 Selbig & Banner-man (2008)
Wisconsin-major     0.004
land disturbance
phase
                                                                  0.038-0.06
                                tons/acre/year
                             Wisconsin - house     0.004
                             construction phase
                                                    tons/acre/year
Vice etal. (1969)
Ward and Appel ( 1988)
Wolman and Schick (1967)
Yorke and Herb (1978)
Virginia 0.038
West Virginia 0.3
Maryland - Baltimore -
Maryland - Towson -
Maryland -
Cockeysville
Maryland - Gwynns -
Falls
Maryland - multiple
sites
63-76
4.4
219
125
112
18
7.2-100.8
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
tons/acre/year
1 Modeled estimate.
 Table 4-4: Summary of Turbidity Findings from Selected Studies
                                                                        Turbidity (NTU)
Study Author
Barrett, et al. (1995)
Chisholm & Downs (1978)
Embler& Fletcher (1983)
Line (2009)
Lubliner & Golding (2005)
Shields & Sanders (1986)
Tsui &McCart (1981)
Wong (2005)2
Location
Texas
West Virginia
South Carolina
North Carolina - Tilly Up
North Carolina - Tilly Down
North Carolina - Ellery Down
North Carolina - King's Mill
Washington
Washington
Washington
Washington
Washington
Washington
Mississippi
British Columbia
Hawaii -Site 16226200
Hawaii -Site 16265700
Hawaii -Site 16266500
Hawaii -Site 16267500
Hawaii -Site 16269500
Hawaii - Site 16270900
Hawaii -Site 16273950
Hawaii -Site 16275000
Unaffected site
6
2.3
<25'
25
54
140
41
2.2
5.2
5.5
6.4
9.8
17.6
18
0.5-2.3
5.6
1.4
0.8
3.5
3.8
3.1
13
0.7
Affected site
72
7.1
50- 801
1,530
1,197
504
593
2.2
6.4
5.2
25.7
45.0
12.0
40
0.7-5,000
9.0
11
1.8
14
2.2
2.2
3.8
1.0
  Peak values.
 2 Wong (2005) measures concentrations as a long-term geometric mean.
November 2009
                                                                                                4-16

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
4.2     Individual Study Summaries

This section provides brief summaries of studies of construction site discharges reviewed by EPA.
Additional information is available in the original studies. Individual study summaries are presented in
chronological order with the exception of state reports, which are summarized at the end of the section.
Whitney and Bailey (1959) documented large declines in fish abundance in a stream section that was
channelized as part of a highway construction project. Small fish declined 85 percent in number and large
fish declined 94 percent in number.
Guy and Ferguson (1962) described a lake in Virginia that accumulated approximately 235,000 tons of
sediment between 1938 and 1957 due to construction and development taking place in 68 percent of the
lake's watershed during this period.
Keller (1962) documented elevated suspended sediment discharge in a river in Maryland. Sediment
concentrations and total flux were significantly higher at a monitoring station downstream of a rapidly
developing area relative to those measured at a monitoring station in an upstream rural watershed.
Guy (1963) described a 58-acre residential construction site in Kensington, Maryland with an annual
sediment yield of 39 tons/acre from 1957 to 1962. Suspended sediment concentrations from 1,490 to
65,000 mg/L were measured in the receiving stream. Early in the project, sediment from the construction
site fully buried the receiving stream bed, which, under natural conditions, had a rock and gravel bottom.
Much of the sediment was flushed from the channel by 1962, approximately 2 years after the peak of
construction activity on the site.
Wark and Keller (1963), in a study of sediment sources in multiple Potomac River watershed subbasins,
found the highest sediment yields in subbasins with significant levels of construction activity. Rates were
10 to 50 times higher than those in rural  areas and varied with construction activity intensity. Annual
sediment yields in watersheds of 4.1 to 72.8 square miles with large amounts of construction activity
ranged from 1.7 to 3.6 tons per acre.
King and Ball (1964) documented impacts to a river in Michigan from highway construction. The
highway project crossed eight major tributaries to the river at distances of 2 to 5 miles from the river. The
authors observed elevated turbidity, suspended sediment, and sedimentation levels in the river. The
sediments were primarily inorganic. Primary production declined 70 percent once construction activity
began. Macroinvertebrate biomass declined. Smallmouth bass populations also declined, which the
authors attributed to the sedimentation of the river's pools, important habitat for the bass. The authors also
noted capture of sediments by patches of aquatic vegetation able to grow in the river because of nutrient
discharges from a sewage treatment facility. The authors expressed concern that the combination of
aquatic vegetation and sedimentation would create a mud flat ecosystem in place of the former coarse
substrate of the river.
Wolman and Schick (1967) examined multiple studies conducted in the Mid-Atlantic region and found
that sediment flux in streams receiving construction site discharges was two to several hundred times
greater than that in streams draining rural or wooded areas. The authors found annual sediment yields of
18 to 219 tons per acre from construction sites 1.2 to 151 acres in size. The authors stated that, under non-
construction conditions, annual sediment yields average 0.3-0.8 tons per acre, with lower levels (e.g., 0.02
tons per acre) in predominantly wooded watersheds. The authors noted that higher sediment yields were
found in those watersheds with the greatest proportion of land disturbed by construction and the least
"dilution" by lower sediment yields from other land uses. For example, the highest sediment yield noted
November 2009                                                                                4-17

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
in the study, 219 tons/acre/year, was associated with a drainage area of 1.2 acres that was completely
disturbed by construction activity. The study noted sediment concentrations in construction-affected
surface waters of 3,000 to greater than 150,000 mg/L, whereas the highest concentration found in waters
draining predominantly natural or agricultural areas was 2,000 mg/L.
The study also described in detail the impacts of sedimentation from an individual construction site in
Maryland. Virtually all stream sections contained sand or silt deposits, in contrast to natural conditions
consisting of a predominantly cobble streambed and riffle and pool sequences. Portions of the stream bed
had sand deposits up to 2 feet thick that traveled downstream in dunes. At all sites the authors examined,
construction sediment impacts were visible for the entire length of the stream between the construction
site and the stream's confluence with a reservoir (up to 2 miles). The authors noted that sediment from
construction activity was nearly absent from a different stream in the same area seven years after the
cessation of construction upstream. The authors did not describe the manner in which this sediment was
redistributed downstream of the site of original impact. The authors noted the difficulty of predicting the
duration and extent of sediment storage in different waters.
Vice et al. (1969) documented that highway construction constituted 1 to 10 percent of a 4.5 square mile
Virginia watershed's area at any given time but contributed 85 percent of total sediment yield.
Measurements were taken nearly 2 miles downstream from the construction area, indicating that the
construction impacts were not solely localized to the active construction site. The authors also noted
significant turbidity, which imparted a reddish color to the affected stream. Turbidity and suspended
sediment concentrations increased as construction activity increased and abated as construction activity
decreased. The stream carried a 90-day mean concentration of 20,000 to 25,000 mg/L of suspended
sediment during precipitation events for several months.
Walling and Gregory (1970) documented a 2- to 10-fold and occasionally up to 100-fold increase in
suspended sediment concentrations in waters draining areas undergoing construction.
Barton et al. (1972, as cited in Barton 1977 and Barrett et al. 1995) found that highway construction
extirpated invertebrates in a Utah stream  bottom. The stream bottom recolonized within 6 months of
construction completion.
Peterson and Nyquist (1972) studied highway and bridge construction impacts on an Alaskan stream. The
stream was still recovering from damage  from past placer mining activity. The authors documented
elevated turbidity and conductivity levels downstream of the site. No other water quality parameters
changed (see Table 4-1}. They also documented severe declines in the number and diversity of benthic
macroinvertebrates downstream of the site. Macroinvertebrate abundance and density approached normal
levels one year after construction.
Huckabee et al. (1975) documented impacts to a small stream  from a highway construction project in
Great Smoky Mountains National Park in Tennessee. Stream acidification took place shortly after
completion of the project and resulted in  a fish kill. Ten years  later, the authors documented elevated
sulfate, calcium,  and metal (aluminum, magnesium, zinc, manganese, lanthanum, and samarium) levels
and acidification in the stream due to iron sulfides in rock used for roadbed fill in 1963. Upstream of the
site, stream pH was 6.5 to 7.0. Downstream of the site, stream pH was 4.5 to 5.9. In addition, a dense
white-yellow precipitate coated stream substrate rocks and periphyton more than 1 mile downstream. The
impacted stream  was nearly devoid offish (e.g., brook trout), salamander larvae, and macroinvertebrates
for up to 4 miles  downstream of the fill. Conditions improved  gradually with increasing distance
downstream. Fish killed by the stream water showed clear evidence of gill hyperplasia. The elevated
acidity, sulfate, and heavy metal levels were documented as responsible for fish and salamander
November 2009                                                                                 4-18

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
mortality. Kucken et al. (1994) documented that these conditions continued to persist 30 years after
construction completion. Two species of stream-breeding salamander were nearly eliminated, and
abundances of two other species were halved (Kucken et al. 1994).
Burton et al. (1976)  documented impacts to a Florida stream and lake from highway construction. The
authors documented increases in turbidity and suspended sediment concentrations, as well as in mean
loads per storm for suspended solids, total dissolved phosphorus, dissolved silicon (silica), and
orthophosphate. Loads of nitrate, nitrite, and ammonia also increased, but the authors postulated that these
may have been associated with an upstream sewage discharges (even though these discharges affected
both the upstream and downstream monitoring stations). The authors attributed the elevated levels of
suspended solids, phosphorus, and dissolved silicon to soil  erosion from the construction site. Total
dissolved solids loads and total stormwater discharge did not change due to construction activity. The lake
downstream experienced elevated turbidity and sedimentation levels.
Eckhardt (1976, as cited in Barrett et al 1995b) documented impacts from highway construction on a
Pennsylvania stream. Suspended sediment discharge from the construction site comprised approximately
50 percent of the total suspended sediment load in the stream over a 3.5-year period. High sediment yields
from the  site continued after completion of construction because of delay in vegetation growth.
White (1976, as cited in Taylor and Roff 1986) documented increases in nitrogen and other inorganic ions
in a stream downstream of a road construction site.
Barton (1977) studied highway construction site impacts to an Ontario stream and associated ponds up to
approximately 1 mile downstream over the life of the construction project. The study documented
elevated suspended sediment and sedimentation levels (see Table 4-2). Sedimentation levels increased 10-
fold. The decline in  stream sediment organic matter content reflected the erosion of low organic matter
soil from the construction site. An increase in nitrate concentrations was observed during the final
construction stages.  Suspended sediment and sedimentation levels declined with increasing distance
downstream due to sediment accumulation in several ponds along the stream channel, though elevated
sedimentation levels were also seen downstream of the ponds after heavy spring flows. The number and
standing crop offish declined from 24 kg/ha to 10 kg/ha immediately downstream of the construction site
during construction but returned to normal within 8 months of construction completion. The authors
postulated that fish may have migrated from the area during construction to avoid high suspended
sediment levels. No  impacts on fish were documented further downstream, and no fish mortality was
observed at any point. Benthic macroinvertebrate abundance and species number remained steady, but
species composition changed. Areas completely denuded by construction activity were quickly
recolonized by species known to be strong recolonizers. The authors postulated that some invertebrates
may have sheltered in protected areas during periods of high sediment levels and others drifted in from
unaffected areas. No changes were recorded in ammonia, dissolved oxygen, hardness, nitrite, pH, or
phosphate levels.
Garton (1977, as cited in Barrett et al. 1995) documented an incident in which a highway construction site
in West Virginia discharged large quantities of sediment to a karst cavern system. Springs  discharging
from the caverns were being used as the water supply for a fish hatchery. Sediment accumulation on the
gills of trout in the hatchery killed 150,000 fish during one  event. Other fish kills were the result of a
diesel fuel spill on the site that subsequently washed into the caverns.
Reed (1977) documented impacts to several Virginia streams from highway construction. The authors
documented declines in both macroinvertebrate and fish diversity and abundance due to elevated
sediment discharges and consequent sedimentation from construction sites. Populations recovered
November 2009                                                                                4-19

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
somewhat after sediment discharges declined and excess sediments were flushed from the streams, though
many had still not fully recovered several to 22 months following construction completion.
Chisholm and Downs (1978) studied highway construction impacts to Turtle Creek in West Virginia, also
discussed in Downs and Appel (1986) (see below). Portions of the stream under study were in a relocated
channel. The authors documented elevated turbidity, dissolved solids, sedimentation, and TSS levels (up
to 11,100 mg/L). The authors recorded up to 8  to 10 inches of silt and clay on the stream bed.
Sedimentation levels varied during the study period as excess sediments were flushed downstream by
heavy winter and spring flows, followed by additional sedimentation during lower flow periods. The
authors postulated that the elevated dissolved solids levels derived from use of calcium chloride as a
wetting agent to reduce dust on the construction site during the early and middle phases of construction.
Calcium chloride is highly soluble and leaches readily. Dissolved solids concentrations decreased toward
the end of the construction project.
The authors also observed iron precipitates deriving from groundwater discharges. The precipitates
elevated stream turbidity and imparted a reddish-orange color to it. The authors stated that such
discharges are common in recently excavated areas in the watershed.
Heavy sedimentation levels and relocation of portions of the stream channel initially extirpated or heavily
degraded macroinvertebrate populations over several  miles of stream. Organism diversity and abundance
recovered within a year after construction completion as stream substrate was flushed of excess
sedimentation, then stabilized. The authors stated that organism drift from unimpacted tributaries
upstream and good quality construction and revegetation of the new stream channels aided community
recovery.
Extence (1978) documented elevated sedimentation levels and suspended solids and iron concentrations
in a stream and river in Great Britain downstream of a highway construction site. Sedimentation
proceeded to the point where fine sediment fully covered the former rocky substrate of the river. River
channel width decreased as a consequence of sediment accumulation. High flow events periodically
transported some of the sediments downstream. Sedimentation of the substrate reduced benthic
invertebrate density and diversity. The authors attributed the elevated iron levels to erosion of iron-
containing sediment from the construction site. The authors did not observe impacts to water column pH,
BOD, orthophosphate, or nitrate levels.
Helm (1978, as cited in Barrett et al. 1995b) documented impacts from highway construction on a stream.
Construction discharges contributed approximately 50 percent, or 8,000 tons, of the approximately 16,000
tons of suspended sediment discharged in the stream during the construction period. The stream already
contained elevated sediment levels due to steep slopes, fine coal wastes, coal-washing operations, and
other land uses in the upstream portion of the watershed.
Tan and Thirumurthi (1978) studied water quality changes in four lakes  during highway construction
activity taking place 30 to 400 feet from the lake shorelines. The lakes served as a drinking water supply
for the city of Halifax, Nova Scotia. In one lake, elevated turbidity, total dissolved solids, and nitrate +
nitrite-nitrogen levels were documented. Total dissolved solids were primarily composed of chlorides,
sulfate, and sodium and calcium carbonates. Elevated conductivity was documented in two lakes. No
water quality changes were documented in the  other two lakes. Table 4-1 lists the parameters for which
no water quality changes were detected. The authors postulated that construction activity was the source
of the water quality changes and that elevated nitrogen levels derived from the site's disturbed soils and
vegetation. The authors stated that the highway projects had caused serious erosion, producing suspended
sediment concentrations as high as 12,836 mg/L and very high turbidity concentrations in site runoff. The
November 2009                                                                                 4-20

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
authors noted that sediment control devices treating the runoff to the primary affected lake performed
poorly.
Yorke and Herb (1978) documented impacts to streams draining several Maryland watersheds over a
multi-year period of development. The authors documented elevated suspended sediment yields from
areas undergoing construction. The authors stated that an earlier study (Yorke and Davis 1972) had
documented a 30 percent increase in stormwater runoff to streams due to construction taking place in 15
percent of the watershed.
Daniel et al. (1979) documented elevated yields of suspended sediment, total phosphorus, organic
nitrogen, nitrate + nitrite, ammonia, and dissolved phosphorus from three residential construction sites in
Wisconsin.
Hainly (1980, as cited in USFWS 2003 and Barrett et al.  1995b) found that highway construction site
discharges elevated turbidity levels in a stream up to 6 miles downstream. The construction site
contributed 9,100 tons of suspended sediment to the stream or more than 25 percent of the total suspended
sediment load of 35,500 tons in the stream during the study period. The author observed little
sedimentation immediately downstream of the construction site.
Reed (1980) monitored a highway construction site in Pennsylvania before, during, and after construction
activity. The author documented elevated suspended sediment yields in four streams during construction.
These yields fell approximately to pre-construction levels during the 2- to 3-year period following project
completion. Suspended sediment concentrations also increased during construction and fell during the 2-
to 3-year period following project completion.
Lenat et al. (1981) documented elevated suspended solids levels in the water column and increased
fractions of sand and gravel in the substrates of two streams downstream of a construction project. These
discharges impacted invertebrates in the stream. A suite of benthic macroinvertebrate species adapted to
the new sandy substrate and associated periphyton downstream of the construction site was able to
establish itself in high numbers during low flow periods.  During high flow periods, however, this
community was unable to persist. On average, organism density and habitat availability was lower
downstream of the construction sites, but no species were entirely lost from the ecosystem. The authors
noted that all stream sites in the study were impacted by discharges from upstream agricultural activities
and gravel roads and so already contained a somewhat stress-tolerant benthic invertebrate community.
Tsui and McCart (1981) documented impacts to a stream in British Columbia due to construction of a gas
pipeline crossing. Suspended sediment levels were elevated up to 10,660 mg/L and turbidity levels up to
5,000 NTU within 300 feet downstream of the construction site. Sedimentation levels increased. The
authors documented a large reduction (up to 74 percent) in benthic macroinvertebrate abundance. Excess
sediment was flushed downstream within 2 years of pipeline installation. The invertebrate community had
not yet recovered after 2 years but was progressing in that direction.
Cline et al. (1982) examined stream sites within approximately  1,600 feet downstream of a highway
construction project in a multi-year study. Construction activity elevated TSS levels 10- to 100-fold at
downstream sites. The exception was one site where a reservoir release between the upstream and
downstream sampling points influenced observations. TSS levels returned to background conditions
within 2 weeks following cessation of construction. Construction also increased the percentage of fine
sediments in the stream's substrate. This effect dissipated within 1 year of construction cessation, in large
part due to high stream flow during the snowmelt period, which transported excess  sediment downstream.
Periphyton was generally less abundant and contained more detritus at downstream sites. The authors
November 2009                                                                                4-21

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
noted that fine sediment tended to collect in the periphyton. Downstream sites also had lower algal
diversity. These effects were still apparent 1 year after construction completion. Macroinvertebrate
density and biomass declined at some, but not all, downstream sites, and remained depressed for a year
following construction completion. Taxa composition changed to favor more pollution-tolerant species.
The authors postulated that construction activity timing had been such that high snowmelt flows  and
reservoir releases had transported sediments downstream and improved conditions at unaffected  stream
sites.
Cramer and Hopkins (1982, as cited in Barrett et al. 1995b) studied impacts to a Louisiana wetland from
highway construction. The authors documented elevated turbidity and a change in water color deriving
from construction discharges. Changes in salinity, dissolved oxygen, and pH were attributed to other
factors. Turbidity decreased and water color began to return to normal levels once construction ceased.
Embler and Fletcher (1983, as cited in Barrett et al. 1995b) studied construction site impacts to a South
Carolina stream. Turbidity levels increased from below 25 NTU to peak levels of 50 to 80 NTU.
Suspended sediment levels increased from below 30 mg/L to peak levels of 60 to 130 mg/L.
Werner (1983, as cited in Donaldson 2004) documented a case in which improper highway construction
practices  created multiple instances of turbid water in karst springs in West Virginia. The turbid water
clogged the gills of trout, killing a large number of them.
Helsel (1985) documented impacts to an Ohio stream and river from highway construction. Construction
significantly elevated sediment yield and suspended sediment loads downstream.
Duck (1985) documented impacts to a stream and lake from gravel road construction in Scotland. A
stream flowing into the lake carried a sediment load 20 times greater than normal. A turbid plume was
observed to extend over 1,600 feet into the lake. The author documented deposition of at least 2,010 tons
of sediment on more than 11 acres of lake bed during a 2-month period. The author estimated that the
lake, under natural conditions, would have otherwise taken 20 to 25 years to accumulate this quantity of
sediment.  Suspended sediment concentrations in the affected streams returned to normal levels after
completion of construction.
Downs and Appel (1986) documented construction site impacts to streams and rivers from highway
construction in West Virginia. Suspended sediment loads in Trace Fork, a small stream in the study,
increased from 830 to 2,385 tons downstream of the construction site for a 14-month period. Suspended
sediment concentrations of up to 7,520 mg/L were measured in the stream. The largest daily mean
concentration was 2,500 mg/L.  Turtle Creek, another small stream, had elevated suspended sediment
loads of 34,000 tons over 2 years. The authors were unable to gather sufficient data to determine impacts
to suspended sediment in the Coal River, a river downstream of the smaller streams affected by
construction activity. The authors postulated that the difficulty in measuring the impact to the larger river
was due to masking of the construction discharge signal by dilution from the river's large flow volume
and by large sediment contributions from other sediment sources in the river's watershed.
Nodvin et al. (1986) observed large sediment plumes extending across most of the length of a 2,200-foot
wide lake whose inlet lay approximately 5.1 miles  downstream of a bridge construction project. The
sediment derived from preparation of an upstream area on a tributary stream for bridge footer
emplacement. The authors documented elevated levels of alkalinity, conductivity, pH, and calcium in the
stream. Maximum changes observed  were 6.83 to 11.22 for pH,  9 to 341 micro-siemens per centimeter
for conductivity, 70 to 2001 micro-equivalents per liter for alkalinity, and 54 to 2227 micro-equivalents
per liter for calcium. Smaller increases in magnesium, sodium, and potassium levels were  also recorded.
November 2009                                                                               4-22

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
The authors postulated that leaching of calcium oxide hydration products from the cement in the concrete
poured for the bridge footers was the pollutants' source. Conductivity and alkalinity were found to be
elevated as far as the lake outlet approximately 5.6 miles downstream. Elevated pollutant levels ceased
and dispersed within the 24 hours following the concrete pouring. The authors postulated that the short-
term duration of the effects explains why the phenomena they documented have not been more widely
reported in the literature.
Shields  and Saunders (1986) documented stream impacts from waterway (canal) construction, a portion
of which required stream rechannelization. The authors recorded elevations in specific conductance,
turbidity, color, COD, total alkalinity, hardness, ammonia, phosphorus, sulfate, iron, lead and manganese
during construction. Mean values were 50 to 100 percent higher than levels measured before the start of
construction.  Concentrations of dissolved calcium, dissolved magnesium, chloride, total nitrogen, total
nitrite, and total Kjeldahl nitrogen also increased. Stream temperature and average daily loadings of total
metals (including iron, manganese, lead, and zinc), nutrients, and dissolved solids increased. Coliform
bacteria densities decreased.  The authors ascribed these changes to increased sediment input from the
construction site and reduced shading of the surface water due to removal of riparian vegetation during
construction.  Total organic nitrogen, phosphorus, zinc, iron, and manganese were all significantly
correlated with turbidity discharges. Average daily suspended solids loadings remained approximately the
same, though the authors ascribed this result to water quality sampling issues. The authors attributed the
water quality changes to erosion from the construction site, including excavations that exposed formerly
deeply buried materials to precipitation and erosion.
Taylor and Roff (1986) continued study of a small stream in southern Ontario impacted by both highway
and sewer trunk line construction, extending the work documented in Barton (1977) (summarized above).
Monitoring took place for six years following completion of highway building activities and included the
land stabilization and revegetation phase of construction activity. During this period, part of the
downstream study area was additionally impacted by sewage line construction.
The authors documented elevated suspended sediment and sedimentation levels downstream of the
construction site that declined over several years as vegetation stabilized the former construction site and
excess sediment migrated downstream. Site revegetation, in particular, reduced sediment discharges. The
proportion of organic material in the suspended and bedded sediments increased over time as
predominantly inorganic sediment from the construction site was flushed downstream. Stream sediment
conditions closest to the construction site returned to reference levels approximately 5 years after
construction site stabilization. Sediment conditions further downstream, however, did not fully recover
during the study period due to additional disturbance from sewage line construction and probable input of
sediments flushed downstream from sites closer to the original construction site. The stream bed remained
covered by up to 6 inches of inorganic sediment. The authors documented extensive growths of
previously absent aquatic plants downstream of the construction site, which they attributed to elevated
sedimentation levels.
Elevated nitrate levels were also detected downstream of the construction site. Levels were higher in areas
closer to the construction site and slightly lower at sites further downstream. The authors noted that
phosphorus and nitrogen fertilizers were applied with straw mulch as part of the highway site stabilization
process  and that heavy rains washed much  of the initial mulch application and, most likely, much of the
fertilizer from the site (site culverts were clogged with straw). Elevated phosphorus levels  were not
detected.
November 2009                                                                                 4-23

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
The authors documented that the changes in stream conditions favored some species of invertebrates and
fish and decreased or failed to favor others. For the first 2.5 years of the study, macroinvertebrate species
tolerant of sedimentation and low-level organic enrichment increased in abundance and biomass. The
authors postulated that the macroinvertebrates were responding to nutrient-enriched conditions in the
stream deriving from construction site nutrient discharges. Species less tolerant of sedimentation
increased in abundance more slowly or declined. Their loss led to an overall decline in community
diversity. The magnitude of modification to the macroinvertebrate community declined with distance
downstream from the construction site. In the final 3 years of the sturdy, diversity increased as excess
sediments were flushed from the stream and formerly declining species populations strengthened.
Fish abundance increased downstream of the construction site  during the study period. The authors
postulated that this effect was due to the elevated abundance of macroinvertebrates and the growth of
aquatic plants. The authors noted that fish populations declined downstream of the sewage line
construction site, probably due to migration of fish from the area to avoid high suspended sediment
levels. Population levels rebounded once construction was complete. Community composition changed to
favor mid-water feeder species over bottom feeder species.
The authors postulated that several of the changes to the stream community were "relatively permanent"
and that return to original conditions would take many years (assuming no further disturbance from new
construction sites).
Ward and Appel (1988) continued the study initially described in Downs and Appel (1986) (see above).
Over a period of nearly 5 years, the Trace Fork watershed saw completion and stabilization of
construction activity for one highway segment, a 2-year period without construction, followed by a
second construction phase. Sediment loads in Trace Fork were elevated during the first construction
phase, declined during the inactive phase, and increased again  during the second phase of construction.
Stout and Coburn (1989) studied macroinvertebrates that consume leaf detritus in a Tennessee stream
2 years after completion of highway construction upstream. Observation sites were located within
0.5 miles of the former construction site. The authors found stream substrate downstream of the
construction site to have returned to conditions similar to those of upstream sites. The total abundance  and
number of macroinvertebrate taxa was lower downstream of the former construction site than expected.
The authors attributed this observation not to construction sedimentation impacts but to the removal of
riparian vegetation during construction, which  reduced detrital inputs to the stream and elevated stream
temperature.
Yew and Makowski (1989, as cited in Barrett et al. 1995b) documented impacts to several streams near
the Tennessee-North Carolina border from highway construction. Exposure of pyritic shale during
construction resulted in discharge of sulfuric acid to the streams.  Conditions toxic to fish existed in the
streams due to low pH levels of 4.0 to 4.4, low alkalinity, and elevated toxic metal concentrations.
Young and Mackie (1991)  studied oil pipeline  construction impacts to a stream in the Northwest
Territories, Canada. Part of the construction process involved burying a portion of the pipeline beneath
the stream. Downstream sampling stations were located within 1,500 feet of the construction zone. The
authors documented elevated TSS concentrations downstream  of the construction site that peaked at 3,000
mg/L. These sediments contained significantly less organic matter than the usual stream sediments.
Macroinvertebrate drift increased as suspended sediment concentrations increased. The authors attributed
this effect to the increased levels of sediment on the stream bed and in the water column. Total biomass
and species richness recovered by the warm season following completion of construction. Sedimentation
also increased downstream of the construction  site, but excess  sediments were flushed downstream by  the
November 2009                                                                                4-24

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
end of the warm season following construction completion. Water chemistry parameters did not change
during the study.
Barrett et al. (1995) studied highway construction impacts to a stream in Texas. The authors calculated
percentage changes in concentrations of a number of parameters (see Table 4-1) but did not assess the
statistical significance of the observed differences. Large increases in concentrations of TSS, turbidity,
and iron were observed downstream of the site. Moderate increases were observed in concentrations of
zinc, volatile suspended solids, and nitrate. Smaller increases were observed in copper, COD, total
coliform, and fecal strep. Small declines were observed in concentrations of BODS, fecal coliform, and
total organic carbon. No change was observed in cadmium, chromium, or lead concentrations. The
authors found iron discharges to be highly correlated with TSS discharges from the construction site but
were not able to identify a source for the elevated zinc discharges. The authors observed elevated
sedimentation levels in the stream downstream of the construction site. This excess sediment was flushed
further downstream by storm flows within approximately 1.5 years following the  cessation of
construction.
Stephens et al. (1996) stated that road construction increased loadings of sediment, nutrients, and trace
elements to a receiving stream and downstream reservoir used as a drinking water supply.
Welsh and Ollivier (1998) documented deposition of sediment from a California highway construction
site into a stream. A single storm  event deposited sediment layers 0.1 to 2 inches thick on the  substrate of
several streams within 0.5 to 1 mile of the construction site.  Mean fine sediment depth and percentage of
substrate embeddedness was significantly greater in impacted stream pools. The sediment filled
interstitial spaces in the stream's substrate and significantly  reduced the density of two salamander
species and one frog species.
Reid and Anderson (1999) reviewed 27 studies of water quality impacts from pipeline stream and river
crossing construction at various sites in North America. Pipeline crossing construction frequently requires
construction activity directly in a  surface  water. Most earthmoving activity takes place over the course of
a few days to a few weeks, though site stabilization can require  additional time. Impacts varied among
sites. Multiple studies documented increases in downstream levels of turbidity, suspended sediment,
sedimentation, and stream embeddedness downstream during and after construction. Changes in channel
morphology and substrate composition were also observed. Turbidity levels up to 4,700 NTU and
suspended sediment levels up to 11,600 mg/L have been documented. Multiple studies also documented
reductions in benthic invertebrate abundance and diversity and reductions in fish abundance. Invertebrate
community compositions generally shifted to include a larger proportion of pollution-tolerant species.
Decreases in invertebrate abundance and  changes in community composition were attributed to increased
drift downstream and habitat degradation. The authors also found some studies documenting little or no
impact to fish communities.
Of those  studies that looked at post-construction recovery, multiple studies documented recovery of
downstream areas to pre-construction conditions  within 2 to 4 years following installation of the crossing.
Turbidity and suspended sediment levels typically decreased once major earthmoving activity was
complete. Excess sediments were usually flushed downstream within 6 weeks to 2 years. Recovery of in-
stream cover (pools, rocks, woody debris, and submersed vegetation) took 4 years in one study.
Invertebrate communities moved  toward recovery as excess  sediments were removed and recolonization
from undisturbed areas upstream took place. Fish populations also moved toward recovery with
improvements in sediment conditions and invertebrate communities.
November 2009                                                                                4-25

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
Owens et al. (2000) studied sediment yields in stormwater prior to its on-site treatment from one small
commercial and one small residential construction site in Wisconsin. The authors found that construction
activity significantly increased sediment yields from the sites relative to sediment yields from undisturbed
rural areas (see Table 4-3). Average event mean concentrations of suspended sediment also increased
significantly. Increases in sediment yield and event mean suspended sediment concentration were less at
the site that was active during winter, when rainfall events were less intense, than at the site that was
active during the summer, when rainfall events were more intense.
Fossati et al. (2001) studied a river in Bolivia impacted by highway construction. The small river
downstream of the construction site was transformed by excess sediments into a braided stream system.
The authors documented a 500-fold increase in suspended sediment concentrations, which ranged from
2,574 mg/L near the construction site to 5 to 267 mg/L further downstream. Turbidity impacts were
observable 56 miles downstream. Smaller increases in phosphate and bicarbonate concentrations were
observed. Sedimentation levels increased significantly. Aquatic invertebrate density declined 200-fold,
and the total number of taxa declined 6-fold. Authors attributed the invertebrate decline to various
impacts associated with increased suspended sediment concentrations. No changes in pH, temperature, or
concentrations of chlorides, sulfates, nitrates, calcium, magnesium, sodium, or potassium were observed.
Hunt and Grow (2001)  documented impacts to a high quality stream downstream of a construction site
that failed to comply with discharge requirements. Sedimentation created a silt and clay layer 12 to 16
inches thick on the stream bed. The stream's substrate, gravel and cobble under natural conditions, was
highly embedded 5,000 feet downstream. Number offish species decreased from 26 at the upstream site
to 19 at the downstream site. Fish abundance decreased from 525 to 230 individuals. Pollutant-tolerant
species increased 237 percent and pollutant-intolerant species decreased 33 percent. Fish biomass
declined  62 percent. Measures of overall habitat quality significantly declined. The authors attributed the
decline in aquatic community health to the severe embeddedness of the stream's substrate.
Kayhanian et al. (2001) studied 15 California highway construction sites with widely varying
characteristics over a 2-year period. In stormwater discharges from the sites, the authors documented
varying levels of particulate cadmium, nitrate, nitrite, ammonia, total Kjeldahl nitrogen,  suspended and
dissolved solids, turbidity, chemical oxygen demand, oil and grease, total and fecal coliform,
chlorpyrifos,  diazinon,  and dissolved and particulate chromium, copper, lead, nickel, silver, zinc, and
phosphorus. The authors attributed TSS and turbidity levels runoff to disturbed soils on the sites. They
stated that the sources of the other pollutants were unknown. Because levels of these pollutants varied
among sites, the authors speculated that site-specific soil and vegetation conditions influenced their
discharge. The authors  also noted correlations between TSS concentrations and particulate copper,
chromium, and zinc concentrations, indicating that these pollutants are traveling with sediment during
erosion.
Buckner  et al. (2002) stated that sediment is the top  source of impairment in the Tennessee, Cumberland,
and Mobile River basins,  a region containing high levels of aquatic organism diversity. The study noted
that a large number of aquatic species in this region are threatened, endangered, or being considered for
listing. Construction activity is one source of sediment in the region. The authors described an incident
where improper control of runoff from a road construction site in Tennessee resulted in stream vegetation
loss and stream bank collapse. Four feet of sediment accumulated on the stream bottom, extirpating the
existing biological community. The water supply  for a downstream town was contaminated. The authors
also noted a report of significant levels of sedimentation in the Cahaba River due to sewer trunk line and
residential home construction.
November 2009                                                                                 4-26

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
Nelson and Booth (2002) noted that construction sites had one of the highest sediment yields per unit of
land area of all land uses in western Washington, approximately four times greater than natural sediment
yield for the region. Construction comprised 0.3 percent of the area of the watershed under study and
contributed 0.7 percent of total watershed sediment yield.
Carline et al. (2003) documented statistically significant increases in TSS concentrations and suspended
sediment loads at sampling sites within 1,200 feet downstream of two different road construction sites.
The authors monitored the site during the entire construction process.  Total TSS export at one site
increased at least 13.5 percent relative to upstream levels after installation of concrete-lined drainage
channels on site. The authors noted that interpretation of data from this site was complicated by probable
elevated sediment export from an additional construction site upstream of the observed construction site.
Total TSS export at the second study site was  13.6 percent higher. The authors did not find impacts to
stream substrate  composition, benthic macroinvertebrate communities, or trout redd occurrence
downstream of either site.
Selbig et al. (2004) studied discharges from a residential construction site over the life of the project. The
authors found that discharge volumes and total and suspended sediment loads were significantly greater in
the stream downstream of a residential construction site compared to upstream of the site. The authors
noted that careful implementation of a wide range of sediment and erosion control practices significantly
reduced discharge volumes and sediment loads during the life of the project. The authors detected no
significant impacts on stream geomorphology, macroinvertebrate communities, and fish communities
downstream of the site. However, the stream drained a heavily agricultural area and was ranked very good
to good for macroinvertebrate community quality, poor for fish community quality, and fair for overall
habitat quality prior to the start of construction activity. Changes in temperature and fine sediment levels
were attributed to processes unrelated to construction activity.
Lubliner and Golding (2005), in a study designed to examine a wide range of construction sites, visited
183 sites in Washington. They found 44 to be discharging runoff at the time of their visit. Other sites were
not discharging due to the episodic nature of precipitation events, lower than normal rainfall, on-site
infiltration, and use of stormwater management practices. Of the 44 discharging sites, 6 discharged
directly to surface waters and 38 infiltrated discharges or directed them to a municipal stormwater sewer.
Two of the six sites' discharges elevated turbidity levels 100 feet downstream. The authors stated that the
low incidence of receiving water turbidity impacts might have reflected the lower level of precipitation
during the study  and efforts at construction sites to avoid discharging to surface waters.
Wong (2005), in a multi-year study, documented elevated total sediment yield, TSS concentrations, and
turbidity levels at multiple stream sites within approximately 0.5 miles of a highway construction project
that were most likely attributable to construction activity. Elevations were seen in both long-term
geometric means as well as 90th and 98th percentile values.  Some sites were also disturbed by golf course
construction. TSS and turbidity levels declined after completion of construction except for TSS levels at
two sites and turbidity levels at one  site.  Elevated nitrite + nitrate, total phosphorus, copper, and specific
conductance levels were also thought to derive from construction activity, but were detected at
substantially fewer stream sites. No effects from construction activity were seen on a reservoir
approximately 0.5 miles downstream of the project.
Cleveland and Fashokun (2006) documented a two- to  six-fold increase in TSS concentrations due to
highway construction activity.
Hedrick et al. (2006) found no statistically significant difference in water quality between sites 0.2 km
downstream of construction activity and unaffected sites. The authors speculated that the lack of elevated
November 2009                                                                                  4-27

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
suspended sediment levels was due to frozen soil conditions during the peak construction period,
decreased rainfall intensities, and/or effective sediment erosion and control practices. The authors also
noted that it was likely that the construction site elevated sediment discharges to the stream, but that they
were difficult to detect statistically because the construction site constituted a small portion of the
stream's total watershed, site discharges were diluted approximately 3,300-fold in the stream, and the
stream's suspended sediment concentration was naturally highly variable.
Hedrick et al. (2007) examined impacts from a West Virginia highway construction project during a
3-year period at two stream sites 300 to 800 feet downstream. The construction site significantly
increased total sediment export downstream, particularly prior to the installation of silt fences. During this
time, benthic macroinvertebrate community health declined in one stream. Community health recovered
once silt fences reduced sediment export from construction activity. Over the total duration of the study,
construction activity did not have a significant impact on the benthic macroinvertebrate community's
health. The authors postulated that stream areas downstream of the construction site were successfully
recolonized by organisms from unaffected stream areas. The authors also found little difference in the
percentage  of fine sediment present at upstream versus downstream sampling sites.
Clausen (2007) documented changes in stormwater runoff quality  in storm sewer and ditch discharges
from two watersheds under construction, one using traditional sediment and erosion control practices and
the other using more rigorous ("enhanced") control practices. Though runoff volume typically increases
during construction activity, it declined 97 percent from the enhanced BMP watershed. The authors
attributed this change to sediment and erosion control practices that retained water on-site and the
placement on-site of soil with higher infiltration values than the  site's original soil. Runoff volume
doubled in the traditional watershed. In the enhanced BMP watershed, both TSS and phosphorus total
export and concentrations increased. Total export of nitrogen stayed the same, though nitrogen discharge
concentrations increased. The author attributed increased nitrogen concentrations to fertilizer use. In the
traditional watershed, total export of both sediment and nutrients increased. TSS concentrations did not
change, which the author attributed to adequate sediment control practices. Nitrogen and phosphorus
concentrations stayed the same or declined. Zinc export  from the enhanced BMP watershed declined
during construction, and copper and lead export stayed the same. Metals export from the traditional
watershed increased. Fecal coliform and BOD levels did not change.
Line and White (2007) compared stormwater runoff immediately downstream of an area undergoing
residential development to a nearby undeveloped area. Construction activity elevated the percentage of
rainfall  leaving  the area as stream runoff to 55 percent versus 21 percent at the unaffected site. Runoff
volume was 68  percent greater. The site elevated TSS, total phosphorus, nitrate, TKN, ammonia, and total
nitrogen yields  (see Table 4-1}. Stream baseflow declined to 0 percent of flow versus 25 percent at the
unaffected site and peak discharge rates increased more than four times.
Selbig and Bannerman (2008) found that runoff volume  from an area undergoing residential construction
in Wisconsin increased relative to  pre-construction conditions. The authors documented that the Low
Impact Development (LID) techniques used to develop the area, however, had significantly reduced
runoff volume relative to conventional development techniques. TSS, total solids, and total phosphorus
yields also increased during construction.
Chen et al.  (2009) found statistically significant increases in TSS and turbidity levels at stream sites
approximately 100 to 3,400 feet downstream of construction activity. The authors noted that the
magnitude of the increase in TSS and turbidity levels was less than that recorded in many other studies
and thought this was possibly due to a combination of sediment and erosion controls at the construction
November 2009                                                                                 4-28

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
site and the predominance of woodland and pasture in the contributing watersheds. Total iron, sulfate,
chloride, and potentially other water quality parameters were elevated during construction activity, and
chloride, sulfate, nitrate, and acidity were elevated after project completion. The authors thought the
nitrate may derive from the use of mulches, fertilizer, and hydroseeding for erosion and sediment control
at the construction site. The authors documented a decline in overall stream ecosystem health as indicated
by the percentage of chironomidae, the Hilsenhoff Biotic Index, and Ephemeroptera, Plectoptera, and
Trichoptera (EPT) presence. Stream health remained good to excellent overall during and after
construction, however, which authors partially attributed to the overall wooded and rural nature of the
affected watershed and to the possibility that the stream may not have exceeded certain thresholds for fine
sediment accumulation critical to macroinvertebrate health.
Lee et al. (2009) documented substantially higher sediment yields per unit area from watersheds with
higher levels of road and building construction activity compared to watersheds with lower levels of
activity. Sediment yields were 1.6 tons/acre/year for the  watershed with the greatest level of construction
activity and fell as low as 0.2 tons/acre/year for already urbanized or less-developed watersheds. The
author also observed elevated levels of fine sediment deposition in areas with high levels of construction
activity. The watershed under study was listed as biologically impaired due, in part, to excess suspended
sediments.
Line (2009) described impacts to streams and a lake from highway construction in North Carolina. The
lakes were located in a residential neighborhood that valued them for their aesthetic and recreational
values. Construction discharges elevated suspended sediment and turbidity levels in tributary streams and
the lake. One watershed  ("Ellery-down") was also impacted by discharges from widening of a municipal
road and residential  construction. Larger increases in suspended sediment yield were found in watersheds
that contained a larger proportion of construction activity and/or contained construction activities
particularly prone to erosion.
Montgomery County Department of Environmental Protection (2009) is the most recent annual report for
a multi-year study of impacts to surface waters from construction and land use change on previously
undeveloped lands in the county. The report described changes seen in three large watersheds undergoing
development. Enhanced  sediment and erosion controls were in use and generally performed as expected.
Some monitoring was  conducted immediately downstream of construction sites, while other monitoring
was conducted further downstream in order to examine impacts deriving from larger areas containing
multiple construction projects whose locations changed over time. The report noted that impacts were in
part deriving from disturbed lands where construction had started but was subsequently halted due to an
economic downturn. Disturbed lands were left in an unstabilized condition. Stream chemistry was
monitored downstream of one construction project. TSS concentrations increased  during construction,
declined during site  stabilization, and returned to pre-construction levels during the post-construction
period. TKN levels may  have increased when sediment and erosion control structures from the
construction phase were  converted to permanent storm water control structures. Substrate embeddedness
impacts were documented downstream of two of six construction projects during 2007. Embeddedness
levels improved to pre-construction levels once construction was completed. Another stream underwent
channel aggradation and straightening during three years of active construction in the watershed. Total
water discharge to the  stream increased during this time, and flow was "flashier." As construction activity
declined in the watershed during two to three years, the channel started to erode accumulated sediments
and to widen. Flow became less flashy. Calculation of Index of Biological Integrity scores for benthic
macroinvertebrates and fish for multiple stations throughout the three  large study watersheds indicated
degradation of aquatic ecosystem health from construction activity. Scores for most areas dropped from
November 2009                                                                                 4-29

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
"good" or "excellent" to "fair" or "poor." Areas with more intense construction activity scored more
poorly. The authors noted a shift in macroinvertebrate community structure to include more generalist and
pollution-tolerant organisms. One area of one watershed showed indications of moving toward recovery
of pre-construction conditions.

4.3    State Reports of Construction Discharge Impacts

A number of states have identified construction as a source of damage or impairment to surface waters in
written reports and in required reporting of impaired waters under Sections 303(d) and 305(b) of the
Clean Water Act.
Ohio's state water quality reports from the early  1990s through 2006 include 29 reports identifying water
quality degradation due to construction activity (Ohio EPA 1994, 1995, 1996a, 1996b, 1996c, 1997a,
1997b, 1997c, 1997e, 1997f, 1998a, 1998b, 1998c, 1999a, 1999b, 1999c, 1999d, 2000, 2001b, 2003a,
2003b, 2004, 2005, 2006a, 2006b, 2006c, 2006d, 2007b, 2007c). Water quality reports from Ohio EPA
evaluate biological indicators of water quality and, where possible, identify the waterbody's sources of
impairment. Multiple reports identify fish and macroinvertebrate communities impaired by elevated
suspended sediment and/or sedimentation levels  from construction activity upstream. Several types of
construction are identified as sediment and/or turbidity sources in the studies, including road (Ohio EPA
1997e, 1998c, 1999b, 200la), bridge (Ohio EPA 2001a, 2001b), highway (Ohio EPA 1994, 1998b,
1998d, 1999d, 2005, 2005, 2007a, 2007b), residential (Ohio EPA 1997d, 2006d), commercial (Ohio EPA
2005), golf course (Ohio EPA 1997f, 1998e, 1999d, 2004, 2006b, 2007c),  and airport (Ohio EPA  1996a,
1998c, 2001b).
Ohio EPA (1997f) documented impacts to a stream from a combination golf course and residential
housing construction site. Elevated turbidity and sedimentation were observed up to 10 miles
downstream. Within a short distance downstream of the site, dissolved oxygen concentrations were
depressed, and TSS, fecal coliform, ammonia, and nitrate+nitrite concentrations were elevated. The
stream bed was covered with 4 to 8 inches of silt. The authors stated that the construction site was the
primary source of the pollutants but upstream agricultural sources also contributed. Discharges from the
site degraded stream habitat. Approximately 100 dead fingernail clam (Musculium transversum)
individuals were found in a downstream pool, and the macroinvertebrate community consisted of
pollution-tolerant species. Healthy populations of fingernail clams and pollution-sensitive
macroinvertebrates were found upstream and several miles downstream of the site. The fish community
was also somewhat impacted.
Ohio EPA documented road construction sites as possible contributors to elevated polycyclic aromatic
hydrocarbon (PAH) levels in at least three rivers: Little Cuyahoga River, Little Miami River, and Scioto
River (Ohio EPA 1998d, 1998e,  1999b). PAHs are found in asphalt and tar used in road construction and
repair, and it is possible for them to contaminate  sediment and construction site discharges during rain
events (Ohio EPA 1998d). Ohio EPA also attributed increased chromium concentrations in receiving
waters to pressure-treated wood products used at a nearby residential construction site (Ohio EPA 1997d).
The state of California has expressed concern about discharges from construction sites and has
specifically discussed the impacts of these discharges on threatened and endangered California coho
salmon and steelhead trout and their habitat (McEwan and Jackson 1996; California Department of Fish
and Game 2004). The state noted, for example, that discharges from the construction of Interstate 5 in
California depleted spawning gravels for coho salmon (California Department of Fish and Game 2004).
November 2009                                                                               4-30

-------
Environmental Impact and Benefits Assessment for the C&D Regulation
The states of Michigan and Missouri have also issued reports that identify construction as a source of
water quality degradation (Missouri DNR 2004, 2005; Michigan DEQ 2006).
Table 4-5 summarizes data from the ATTAINS database, a national database of water quality information
submitted to EPA by all states, territories, and the District of Columbia (see Section 2.6, for more
information). This table presents information on surface waters identified as impaired due to  construction
activity as of September 17, 2009. Though substantial, the numbers in Table 4-5 probably underestimate
the level of impairment due to construction activity because  states are not required to identify surface
water impairment sources. As of September 17, 2009, 29 states include "Construction" on their lists of
possible impairment sources. Because significant effort can be necessary to identify a surface water's
source of impairment,  it is likely that a number of construction sites contributing to surface water
impairment have not been identified as impairment sources.  In addition, as discussed in Section 2.6, a
large percentage of surface waters have not yet been assessed for impairment.

 Table 4-5: Construction Impairment in 305(b)-Assessed Waters

Stream/
River
(miles)
12,273
Lake/
Pond/
Reservoir
(acres)
300,855

Bay/
Estuary
(sq. miles)
16

Coastal
Shoreline
(miles)
3
Ocean/
Near
Coastal
(sq. miles)
3


Wetland
(acres)
13,971
Great
Lake
Shoreline
(miles)
-
Great
Lake Open
Water (sq.
miles)
-
 Source: EPA ATTAINS database (EPA 2009a) as of 9/17/09.
November 2009
                                                                                            4-31

-------
Environmental Impact and Benefits Assessment for the C&D Category
       Overview of Benefits from Regulation
This chapter provides an overview of the potential benefits to society related to the reduced sediment
turbidity and reduced nutrient discharges from construction sites that will result from the final regulation.
Sediments and other pollutants from construction sites (e.g., nutrients) may have a wide range of effects
on water resources located in the vicinity of construction sites. These environmental changes affect
economic productivity (e.g., drinking water supply and storage and navigation) as well as environmental
services valued by humans (e.g., recreation, public and private property ownership, existence services
such as aquatic life, wildlife, and habitat designated uses). Related market benefits (e.g., avoided costs of
producing various market goods and services) and nonmarket benefits are additive (Freeman 2003). In all
cases, benefits are conceptualized and estimated based on established welfare theoretic models (Freeman
2003; Just etal. 2004).
This chapter provides a conceptual framework for understanding the benefits likely to be achieved by the
regulation, a qualitative discussion of those benefits, and a review of the benefit estimation methods used
in EPA's analysis of this regulation. The following chapters quantify and estimate the economic value of
these benefit categories in greater detail.

5.1     Conceptual Framework for Valuation of Environmental Services

The economic benefits of the final regulation reflect the money-metric values associated with the
resulting improvements in water and other resources. The conceptual approach for estimating the
monetary value of benefits of a policy involves an evaluation of changes in social welfare realized by
consumers and producers, and quantification of these changes in money-metric terms. Such measures are
based on standardized and widely accepted concepts within applied welfare economics (Just et al. 2004).
They reflect the degree of well-being derived by economic agents (e.g., people and firms) given different
levels of goods and services, including those associated with environmental quality. For market goods,
analysts typically use money-denominated measures of consumer and producer surplus, which provide an
approximation of exact welfare effects (Freeman 2003). For nonmarket goods, such as aquatic habitat,
values must be assessed using nonmarket valuation methods (Freeman 2003). In such cases, valuation
estimates are typically restricted to effects on individual households  (or consumers), and either represent
consumer  surplus or analogous exact Hicksian welfare measures (e.g., compensating variation). The
choice of welfare measure (i.e., value) is often determined by the valuation context.
The economic value of the changes in environmental services provided by surface waters from the
regulation is a product of three sets of functional relationships (Freeman 2003):
>  The effect of the regulation on environmental quality (Q)
>  Human uses  of the affected resources and their dependence on environmental quality (X)
>  The economic value  of the uses of the environment (V).
The first functional relationship relates the final regulation to changes in environmental quality. In this
case, water quality is a function of sediment and other pollutant loadings (S) from construction sites, in-
stream pollutant concentrations (vector P),  and other relevant attributes (vector A) such as stream flow
and velocity. Because the final regulation affects private activities that influence sediment and other
November 2009
                                                                                             5-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
pollutant discharges from construction sites, pollutant loadings are a function of private responses to the
final regulation (R). This relationship may be represented as:

                                       Q = f(S(R),P,A)                                (Eq.5-1)

For this analysis, water quality changes are estimated in terms of sediment and nutrient concentrations in
receiving waters using SPARROW, an empirical relationship between sediment and nutrient
concentrations in surface waters, and DCP. (see Chapter 4 and Appendix C through Appendix F for
details).
The second set of functional relationships describes the human values of the affected water resources and
their dependence on water quality. In the following analyses, pollutant discharges are treated as an
"input" to the water quality, and, as a result, as an "input" to ecological services provided by the affected
water bodies. Like any other input, the value of changes in pollutant discharges from construction sites is
determined by its impact on water quality. Water quality, in turn, influences human uses of the affected
resources, leading to changes in use values. It may also lead to changes in ecosystem services that
generate nonuse values.
For example, water quality is often characterized based on its suitability for recreational activities (e.g.,
whether the water is beatable, fishable, or swimmable) or its ability to support specified uses (e.g.,
primary/secondary contact recreation, public drinking water supply, or commercial fishing). To link water
quality changes from reduced construction site discharges to effects on human uses and support for
aquatic and terrestrial species habitat, EPA used a Water Quality Index (WQI) for this analysis
(McClelland 1974; Dunnette 1979; Harrison et al. 2000; CCME 2001; Cude 2001; Carruthers and
Wazniak 2003; Gupta et al. 2003; USEPA 2004b). Section 10.1 provides detail on the WQI and its
application to the benefits analysis for the regulation. The WQI presents water quality in a way linked to
suitability for varying human uses, but does not in itself identify associated changes in human behavior.
Behavioral changes and associated welfare effects are implied in the proposed benefit transfer approach
for measuring economic values. Additional details of this approach are provided below and in the
following chapters.
In addition to water quality (Q), the level of human uses or other ecological services (vector X) of the
environmental resource depends on a variety of other factors (e.g., type and physical characteristics of the
waterbody, proximity to populated areas, presence of recreational amenities, scenery and etc). If vector/
represents these other inputs into the production of environmental services or uses based on the affected
water resource, the second functional relationship can be expressed as:

                                           X = f(QJ}                                   (Eq.5-2)

The vector of environmental services or human uses (X) affected by changes in water quality and
sediment discharges could include services received through purchases of market goods produced with
surface water resources (e.g., commercial fish), goods, services, and activities produced within the
household, such as water-based recreation, and existence values associated with water quality
improvements.
The third set of relationships defines the economic value (V) of the ecological services or human uses of
the environment. This relationship "embodies the value judgment society has adopted for economic
purposes" (Freeman 2003). For simplicity, and following standard approaches in benefit cost analysis
November 2009                                                                                   5-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
(Boardman et al. 2001), this analysis assumes that the value function is a simple aggregation of
individuals' values.2
The relationship between individual or household utility (U) and water quality (Q) can be more formally
described though the following utility function:

                                   U  =U(X(Q),Y,B)                            (Eq.5-3)

Individuals are assumed to derive utility fmmX, as well as from other goods and services (Y),  which are
not related to water resources. For example, a subset of household goods and services, such as recreation,
are "produced" with water quality (Q) and other inputs such as travel. To account for variation in this
utility structure across individuals, B denotes a vector of individual or household characteristics, such as
demographic characteristics, that influence the form or parameters of the utility function.
Maximization of Equation 4-3 with respect to an income (M) constraint produces the following indirect
utility function:

                                       V = V(Q,P,M;B)                               (Eq.5-4)

where P represents the vector of prices associated with goods and services (Xand Y).
The total dollar value associated with a change in water quality from go to Q\ (holding other variables
constant) can therefore be expressed using the compensating variation(CV) measure from the following
equation:

                             V(Q0 ,P0J0;B) = V(Q, ,P0J0-CV;B)                     (Eq. 5-5)

Other things being equal, an individual is better off with a higher level of water quality (Qi) than a lower
level (Qo). CV in this equation represents the individual's maximum willingness to pay for the specified
water quality improvement, leaving the individual just as well off after payment and higher water quality
as with lower water quality.
From the above, CV (or willingness to pay, WTP) may be derived as a function of pre- and post-change
environmental quality, prices, income, and individual attributes:

                                   CV = CV(Q0,Ql,P0,Y0;B)                           (Eq. 5-6)

This function also provides the basic conceptual foundation for constructing a benefit transfer function,
which can be used to predict WTP values for defined changes in water quality  (van Houtven et al. 2007).

5.1.1  Applicable Benefit Categories

There are likely to be several categories of benefits of the regulation; some benefits are linked to direct
use of market goods and services, and others pertain to nonmarket goods and services. The following
sections outline the most important categories of the expected benefits, and approaches used to estimate
each category. In cases where estimates were not possible due to data limitations or other constraints, this
is also noted below.
    This function could also incorporate weights or environmental justice concepts.
November 2009                                                                                  5-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
5.1.2   Market Benefits

Some of the social benefits expected from the regulation will manifest themselves in economic markets
through changes in price, cost, or quantity of market-valued activities. For benefit endpoints traded in
markets, such as increased yields from commercial fisheries, benefits can be measured using standard
market approaches based on data including market prices and quantities (Boardman et al. 2001).
Competitive prices can also be used to measure benefits related to avoided costs associated with market
goods and services. In analyzing benefits of the regulation, EPA used the avoided cost method to
monetize three benefit categories: (1) maintenance of navigational waterways, (2) reservoir dredging, and
(3) drinking water treatment. Other applicable market benefits (e.g., commercial fishing) are discussed
qualitatively. The avoided cost approach represents an appropriate measure of social benefits for cases in
which a policy change reduces costs to producers of a market-supplied good, but in which price effects
are minimal (Boardman et al. 2001, p. 70-74).

5.1.2.1  Maintenance of Navigational Waterways
The primary effect of pollutant discharges from construction sites on navigation and transportation is
accumulation of sediment in navigational channels and harbors (USEPA 2004b). Section 2.4.1 of this
report provides detail on sediment impacts on navigational waterways. Keeping these areas passable
requires extensive dredging, which can be costly. Implementation of the regulation is expected to result in
less frequent dredging of navigable waterways due to reduced sediment discharges from construction sites
and, as a result, cost savings to the government and private entities responsible for maintenance of
navigational shipping channels, harbors, and other waterways. EPA used the avoided cost of navigational
waterways dredging as a measure of benefits resulting from the final regulation. Chapter 7 of this report
details methods and presents results of EPA's analysis of benefits to navigation expected from the
regulation.

5.1.2.2  Water Supply and Use
>  Reservoir Dredging. Water storage facilities (e.g., reservoirs) serve many functions, including
    providing drinking water, flood control, hydropower supply, and recreational opportunities. Sediment
    in streams can be carried into reservoirs, where it can settle and build up layers of silt over time. An
    increase in sedimentation rates will reduce reservoir capacity. Section 2.4.2 of this report provides
    detail on sediment impacts on the reservoir water storage capacity. To replace this capacity, sediment
    must be dredged from reservoirs, or new reservoirs must be constructed (Clark et al. 1985). The
    regulation is expected to reduce the  amount of sediment entering reservoirs and, as a result, the need
    for sediment mitigation measures in these reservoirs and the cost of reservoir maintenance. Because
    sediment mitigation is often accomplished through dredging, which itself is environmentally
    destructive, an additional benefit to reduced reservoir sediment buildup is the reduced environmental
    impact of dredging sediment and disposing of it. EPA used the avoided cost of reservoir dredging as a
    measure of benefits resulting from the final regulation. Chapter 8 of this report details methods and
    presents results of EPA's analysis of benefits from reduced sedimentation of reservoirs due to the
    regulation.
>  Drinking Water Treatment and Household Water Use.  Discharges from construction sites can affect
    the quality and cost of providing drinking water. Sediment and other discharged pollutants negatively
    affect water quality, and require increased  spending on treatment measures such as settlement ponds,
    filtration, and chemical treatment. There is an additional cost of removing sludge that is created
    during the treatment process (USEPA 2007b). EPA used the avoided cost of treating drinking water
November 2009                                                                                  5-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
    for sediment and disposing of sludge resulting from reduced turbidity in source water as a measure of
    benefits resulting from the final regulation. Chapter 9 of this report details methods and presents
    results of EPA's analysis of benefits to drinking water treatment plants. Reducing nutrient loadings to
    surface waters is expected to reduce eutrophication which is one of the main causes of taste and odor
    impairment in drinking water. Taste and odor in drinking water has a major negative impact on the
    public perception of drinking water safety and the drinking water industry due to a significant
    increase in drinking water treatment costs from foul taste and odor in the source waters. The final
    regulation is expected to reduce the cost of drinking water treatment by improving taste and odor in
    the source waters. EPA, however, was unable to estimate a value of improved taste and odor of
    drinking water, due to data limitations (see Chapter 9 for more detail).
>  Industrial Water Use. Sediment may have negative effects on industrial water users. Suspended
    sediment increases the rate at which hydraulic equipment, pumps, and other equipment wear out,
    causing accelerated depreciation of capital equipment. Sediment can also  clog cooling water systems
    at power plants and other  large industrial facilities (Clark et al. 1985). Section 2.4.4 of this report
    provides detail on sediment impacts on industrial water uses. The final regulation is expected to
    benefit industrial water users by reducing sediment concentrations  in source waters and thus
    increasing the useful life of industrial equipment. The Agency, however, was not able to quantify this
    benefit category due to data limitations.
>  Agricultural Water Use. Irrigation water that contains sediment or other pollutants from construction
    sites can harm crops and reduce agricultural productivity (Clark et  al. 1985). Section 2.4.5 of this
    report provides more detail on sediment  impacts on agricultural water uses.  The final regulation is
    expected to benefit agricultural producers by reducing sediment discharges and, as a result, sediment
    deposition on farm land; this would lead to improvements in land productivity and enhanced
    marketability of agricultural products. The Agency, however, was not able to quantify this benefit
    category due to data limitations.

5.1.2.3  Commercial Fishing
Sediment and other discharges from construction sites can greatly reduce fish populations by inhibiting
reproduction and survival of an aquatic species. These population reductions would reduce the size of
commercial harvest. These changes negatively affect subsistence anglers, commercial anglers and fish
sellers, and consumers offish  and fish products. Chapters 2 and 3 of this report provide some detail on
aquatic life impacts from increased sediment and other pollutant concentrations in fresh and marine
waters. Reducing sediment and other pollutant discharges from construction sites is expected to enhance
aquatic life habitat and thus contribute to reproduction and survival of commercially harvested species,
which in turn will lead to an increase in producer and consumer surplus. EPA did not quantify market
benefits from improved commercial harvest  resulting from this regulation due to data limitations. The
nonmarket value of improved  habitat for commercial fish species may be implicitly accounted for in the
WTP for changes in environmental services provided by surface waters affected by construction site
discharges stemming from the regulation (see Chapter 10 of this report for details).

5.1.2.4  Public and  Private Property Ownership
>  Property Values. Aesthetic degradation of land and water resources resulting from sediment and other
    pollutant discharges (e.g.,  increased water turbidity) can reduce the market value of property and thus
    affect the financial status of property owners. For example, a hedonic price  study by Bejranonda et al.
    (1999) found that "the rate of sediment inflow entering the lakes has a negative influence on lakeside
    property rent" (p. 216). Sediment discharges also have a significant impact on stream morphology.
November 2009                                                                                  5-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
    For example, higher coarse sediment load leads to an increase in width of the river bed and, as a
    result, bank erosion (Wheeler et al. 2003). A 1993 study of Lake Erie's housing market found that
    "erosion-prone lakeshore property will be discounted" (Kriesel et al. 1993). Stabilization of stream
    banks leads to an increase in the value of surrounding property (Streiner and Loomis 1996).
    Therefore, the final regulation is expected to enhance nearby property value in two ways: (1) by
    improving the aesthetic quality of land and water resources (e.g., reduced turbidity) and (2) by
    improving stream morphology and stabilizing river banks.
    Due to data limitations, EPA did not estimate changes in market values of properties located near
    water bodies expected to benefit from reduced sediment discharges from construction sites as a result
    of this regulation. The nonmarket component (i.e., increased satisfaction with the property) may be
    implicitly accounted for in WTP for improvements in environmental services provided by surface
    waters affected by construction site discharges (see Chapter 10 of this report for details).
>  Flood Damages.  Sedimentation can also increase the severity of property damages from flooding.
    Sedimentation of river beds can reduce river capacity and result in higher flood levels and more
    frequent flooding. Additionally, sediments carried by flood waters can damage property and can be
    expensive to remove,  particularly in developed areas. Clark et al. (1985) estimated flooding damages
    attributable to sediment discharges to be $1.5 billion (2008$), annually. Section 2.4.6 of this report
    provides detail on sediment impacts on flood control and stormwater management. This regulation is
    expected to reduce flooding damages by reducing sediment discharges from construction sites. The
    Agency, however, was unable to estimate benefits from reduced flood damages due to data
    limitations.
>  Ditch Maintenance. Sediment in discharges from construction sites can clog ditches and culverts.
    Such sedimentation can increase flooding, which may result in damages to bridges, roads, farmland,
    and other structures in the nearby areas. Preventing these damages may require increased
    maintenance efforts such as dredging (Clark et al. 1985). Section 2.4.6 of this report provides more
    detail on sediment impacts on stormwater management. The final regulation is expected to reduce the
    costs of ditch maintenance by reducing the amount of sediment deposited in ditches. This benefit
    category is not quantified due to data limitations.

5.1.3  Nonmarket Benefits

EPA expects that this regulation will reduce the sediment and other pollutant discharges to surface waters
and will enhance or protect aquatic ecosystems currently under stress, effects for which humans may hold
value but cannot express this value in the form of a market transaction. The decrease in sediment and
other pollutant loadings is expected to improve protection of resident species, enhance the general health
offish and invertebrate populations, increase their propagation to waters currently impaired, and expand
fisheries for both commercial and recreational purposes or use values. Improvements in water quality
such as decrease in turbidity will also favor increased recreational uses such as swimming, boating, and
outings. Finally, the Agency expects that the regulation will augment nonuse values (e.g., option,
existence, and bequest values) of the affected water resources.
The expected nonmarket benefits of water quality improvements fall into two broad categories:
nonmarket use benefits and nonmarket nonuse benefits.  The following sections describe these benefit
categories and the nonmarket approaches to valuing these benefits in more detail.
November 2009                                                                                 5-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
5.1.3.1  Nonmarket Use Benefits
Direct use benefits include the value of improved environmental goods and services used and valued by
people (even if these services and goods are not traded in markets). Aesthetic degradation of water
resources resulting from sediment and turbidity discharges from construction sites can reduce owner
satisfaction with the property and the residential area in general. It can also adversely affect recreational
opportunities. Adverse impacts of sediment and other pollutant discharges and, as a result, increased
turbidity of surface waters on recreational activities are summarized below.
>  Outings. Activities that take place near water such as hiking, jogging, picnicking, and wildlife may be
    adversely affected by sediment and other pollutant discharges into the water. While these activities do
    not involve contact with the water, murky and visually unpleasant water and odors may greatly
    detract from the enjoyment gained through these activities. Decreases in fish populations may cause a
    reduction in wildlife near the resource, affecting wildlife viewing. Pollutants may negatively affect
    local flora and fauna, reducing the aesthetic appeal of the area near the resource and negatively
    impacting wildlife viewing.
>  Recreational Fishing. As noted in Chapter 2 of this report, sediment and other discharges from
    construction sites can reduce fish populations by inhibiting reproduction and survival of an aquatic
    species.  This may lead to fewer and smaller fish, and a reduction of the game fish population.  In
    addition, sediments and other pollutants reduce the aesthetics of the waterbody, which may reduce
    anglers' utility of their fishing experience. Additionally, turbidity caused by sediment and other
    pollutant discharges may affect recreational anglers by reducing the distance over which fish can see
    lures, resulting in lower catch rates (Clark et al. 1985).
>  Boating. Polluted water greatly reduces the aesthetic appeal of recreational boating activities.
    Turbidity caused by sediment and other pollutant discharges may affect the safety of boating.
    Turbidity may obscure underwater obstacles, making collisions more likely. Increased sediment
    concentrations may create sandbars, increasing the chances of running aground. Clark et al.  (1985)
    estimated that turbidity (from all sources) may be responsible for as many as 200 boating fatalities
    and many more injuries each year. Using the value for a statistical life's worth of risks ($7 million)
    and multiplying it by the number of boating fatalities yields the value of preventing these fatalities of
    $1.4 billion per year. Even if reducing discharges from construction sites prevents a few fatalities and
    injuries each year, the expected monetary value of preventing boating accidents can be significant.
>  Swimming.  Turbidity and other problems (e.g., eutrophication) associated with discharges from
    construction sites may greatly reduce a swimmer's aesthetic enjoyment of a resource. Additionally,
    turbidity may create safety hazards for the swimmer by reducing the  ability to see underwater hazards
    or increasing diving accidents by impairing the ability to gauge water depth.
>  Hunting. Similar to the effect on outings, discharged pollutants may greatly detract from the hunters'
    aesthetic enjoyment of a water resource. Damage to flora and fauna may also cause a reduction in the
    game population, reducing the number and quality of the game available.
Improved water quality from reducing sediment and other pollutant discharges from construction sites is
expected to enhance the quality of living in the areas affected by construction activities, thereby resulting
in welfare gain to the resident populations. The final regulation is also expected to enhance recreational
uses of water resources affected by sediment discharges from construction sites, thereby resulting  in
welfare gain to recreational users of these resources. Improved water quality from reducing sediment and
other pollutant discharges from construction sites may translate into two components of recreational
November 2009                                                                                    5-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
benefits: (1) an increase in the value of a recreational trip resulting from a more enjoyable experience, and
(2) an increase in recreational participation.

5.1.3.2  Nonmarket Nonuse Benefits
Even if no human activities (or uses) are affected by environmental changes caused by sediment and other
pollutants from construction sites, such environmental changes may still affect social welfare. For a
variety of reasons, including bequest, altruism, and existence motivations, individuals may value the
knowledge that water quality is being maintained, that ecosystems are being protected, and that
populations of individual species are healthy completely independent of their use value. It is often
difficult to quantify the relationship between changes in pollutant discharges and the improvements in
societal well-being that are not associated with current use of the affected ecosystem or habitat. That these
values exist, however, is indisputable, as evidenced, for example, by society's willingness to contribute to
organizations whose mission is to purchase and preserve lands or habitats to avert development (although
some portion of these donations may be motivated by use values). Notwithstanding challenges involved
in estimation of nonuse values, there is a substantial literature devoted to such issues (Bateman et al.
2002). This literature provides insight into analysts' ability to estimate nonuse values within various types
of policy contexts, and for various types of resources.

5.1.3.3  Nonmarket Valuation Methods
It is frequently difficult to quantify and attach economic values to ecological benefits. The difficulty
results from imperfect understanding of the relationship between changes in sediment, nutrient, and other
pollutant discharges from construction sites and the specific ecological changes, ranging from a lack of
water quality monitoring data for many locations, to time lags between water quality changes and changes
in biological community condition. In addition, it may be difficult to attach monetary values to these
ecological changes because they often do not occur in markets in which prices or costs are readily
observed.
A variety of nonmarket valuation methods exist for estimating nonmarket use value, including both
revealed and stated preference methods (Freeman 2003). Where appropriate data are available  or may be
collected, revealed preference methods can represent a preferred set of methods for estimating  use values.
These methods use observed behavior to infer users' value for environmental goods and services.
Examples of revealed preference methods include travel cost, hedonic pricing, and random utility (or site
choice) models. Compared to nonuse values, use values are often considered relatively easy to estimate,
due to their relationship to observable behavior, the variety of revealed preference methods available, and
public familiarity with the recreational services provided by surface water bodies.
In contrast to direct use values, nonuse values are often considered more difficult to estimate. Stated
preference methods, or benefit transfer based on stated preference studies, are the generally accepted
techniques for estimating these values (USEPA 2000b; OMB 2003). Stated preference methods rely on
carefully designed surveys, which either (1) ask people about their WTP for particular ecological
improvements, such as increased protection of aquatic species or habitats with particular attributes, or (2)
ask people to choose between competing hypothetical "packages" of ecological improvements and
household cost (Bateman et al. 2003). In either case, values are estimated by statistical analysis of survey
responses.
Nonuse values may be more difficult to assess than use values for several reasons. First, nonuse values
are not associated with easily observable behavior. Second, nonuse values may be held by both users  and
nonusers of a resource. Because nonusers may be less familiar with a resource, their values may be
November 2009                                                                                   5-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
different from the nonuse values for users of the same resource. Third, the development of a defensible
stated preference survey is often a time- and resource-intensive process. Fourth, even carefully designed
surveys may be subject to certain biases associated with the hypothetical nature of survey responses
(Mitchell and Carson 1989; Bateman et al. 2003). Finally, efforts to disaggregate total WTP into its use
and nonuse components have proved troublesome (Carson et al. 2000).
To evaluate nonmarket benefits of this regulation, EPA developed a benefit transfer approach based on a
meta-analysis of surface water valuation studies. Benefit transfer may be described as the "practice of
taking and adapting value estimates from past research ... and using them ... to assess the value of a
similar, but separate, change in a different resource" (Smith et al. 2002, p. 134). It involves adapting
research conducted for another purpose to estimate values within a particular policy context (Bergstrom
and De Civita 1999). Although the use of primary research to estimate values is generally preferred, the
realities of the policy process often dictate that benefit transfer is the only option for assessing certain
types of nonmarket values (Rosenberger and Johnston 2007).  In the benefit transfer used for analyzing
benefits of the regulation, the meta-analysis  is of WTP that incorporates both use and nonuse values.
Although the potential limitations and challenges of benefit transfer are well established (Desvousges et
al. 1998), the Agency also emphasizes that benefit transfers are a nearly universal component of benefit
cost analyses conducted by and for government agencies. As noted by  Smith et al. (2002, p. 134), "nearly
all benefit cost analyses rely on benefit transfers, whether they acknowledge it or not." Benefits transfer
methods may be placed into three general categories: (1) transfer of an unadjusted fixed-value estimate
generated from  a single study; (2) the use of expert judgment to aggregate or otherwise alter benefits to be
transferred from a site or set of sites, and (3) estimation of a value estimator model or benefits transfer
function, often based on data gathered from  multiple sites (Bergstrom and De Civita 1999).
Given the generally unreliable performance  of unadjusted single-site transfers, EPA used meta-analysis of
45 surface water valuation studies that use stated preference techniques to estimate a benefit function that
allows forecasting of total WTP estimates (including use and nonuse values) in a variety of policy
contexts, thereby providing more valid benefit estimates for transfer applications (USEPA 2000b;
Johnston et al. 2005). Because stated preference surveys describe a range of environmental services
corresponding to different levels of water quality (e.g., suitable for boating or swimmable), the total WTP
derived from meta-analysis implicitly accounts  for a wide range of nonmarket use and nonuse values,
including recreation, aesthetic value  of near-water properties,  enhanced quality of source water for
drinking and household use, as well as existence and bequest values.
The technical details involved in the estimation of original meta-analyses are presented in Appendix G as
well as in sources such as Bateman and Jones (2003),  USEPA (2004d), Johnston et al. (2005,  2006),
Rosenberger and Phipps (2007), and Shrestha et al. (2007).  Chapter 10 of this report describes application
of the meta-analysis results for estimating benefits of water quality improvements resulting from the
regulation.

5.2     Summary of Effects

Ultimately, changes in sediment and other pollutant loadings from construction sites related to the final
regulation affect humans through influences on the production of goods and services that enter into
individuals' utility functions. Table 5-1 summarizes some of the important ways in which sediment and
other pollutant discharges from construction sites affect the quantity and quality of services provided by
the natural environment and how these services in turn provide economic value to society through various
November 2009                                                                                  5-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
use and nonuse related societal welfare mechanisms.3 EPA was not able to bring the same depth of
analysis to all of the environmental services affected by the regulation categories, however, because of
imperfect understanding of the link between discharge reductions and benefit categories, and how society
values some of the benefit events. EPA was able to quantify and monetize some benefits (including
maintenance to navigable waterways, reservoir dredging, drinking water treatment, and WTP for
improvements in large river and streams and some coastal waters), quantify but not monetize other
benefits (changes in pollutant concentrations in lakes), and assess still other benefits only qualitatively
(e.g., commercial fishing, flood damages, and property values).
3   Underlying this table is a conceptual framework based on the valuation of ecosystem services (Daily 1997; Simpson and
    Christensen 1997; van Houtven et al. 2005).
November 2009                                                                                  5-10

-------
 Environmental Impact and Benefits Assessment for the C&D Category
Table 5-1: Summary of Benefits from Reducing Sediment and Other Pollutant Discharges from Construction Sites
Activity
Recreation
> Outings
> Boating
> Swimming
> Fishing
Commercial Fishing and
Shellfishing
Property Ownership
Water Conveyance and
Supply
> Water conveyance
> Water storage
> Water treatment
Water Transportation
Water Use
> Industrial
> Municipal
> Agricultural
Knowledge (No Direct
Uses)
Environmental Service Flows
Potentially Affected by
Discharges from Construction
Sites
Aesthetics, water clarity, water
safety, degree of sedimentation,
eutrophication, weed growth, fish
and shellfish populations
Fish and shellfish populations
Aesthetics, safety of property from
flooding, property value
Turbidity, degree of sedimentation
Degree of sedimentation of
navigational channels, harbors, and
other waterways
Turbidity
Environmental health
Entities Affected by Environmental Changes
Firms
^
S
S
S
S
^

Governments

S
S
S
S
^

Individuals
Use & Nonuse Values
Use & Nonuse Values
Use & Nonuse Values
Use Values
Use Values
Use Values
Nonuse
Methods Used in EPA's Analysis of Benefits
from the Regulation
Benefits Transfer
Qualitative Discussion
Benefits Transfer
Avoided cost to governments, including:
> Reservoir dredging
> Drinking water treatment
Avoided cost to firms and governments
Qualitative Discussion
Benefits Transfer
 November 2009
                                                                                                                      5-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
Figure 5-1 depicts data sources, analytic steps, and models in EPA's analysis of benefits from reducing
discharges of sediments and nutrients from construction sites. Chapters /through 11 detail methods used
in these analyses and present results.
Figure 5-1: Calculation of Monetized Benefits from the Regulation	
 Legend:
              Data input
           (Estimated Input
                               ^»sc-<21D
Abbreviation Key: USAGE- LJ S. Army Corps of Engineers TSS- Total Suspended
Solids SDWIS- Safe Drinking Water Information System DW- Drinking Water ASCE-
Americart Society of Civil Engineers TN- Total Nitrogen TP- Total Phosphorus WQI-
Waler Quality Index WTP- Willingness-to-pay
                                             Water Quality Modeling
                                    SPARROW predictions of sediment (concentration, yield, reservoir
                                                    s&dlrnentatlan)
Reservoirs
f ResBiyoirssdiment "V
\^ accumulation _j



Drinking Water
TSS'turbidity

,Xffidream\
I SDWIS Treat-
1 merit fee lil IRS

DW Treatment
                                                                                       Non-Market Benefits
                                                                                       STORET&NWISdataon
                                                                                       Bic4ogical Oxygen Demand,
                                                                                       Disserved Oxygen, & fecal
                                                                                       oolifbrm; SPARROW data on
                                                                                       TN and TP concentrations
                                    SPARROW predictions of sediment (concentration, yield, r
                                                    sedimentation}
                                      Change In future
                                     reservoir dredging
                                         costs
November 2009
                                                                                                             5-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
       Water Quality Modeling
Changes in water quality resulting from the various regulatory options were assessed using the U.S.
Geological Survey's SPARROW models (SPAtially Referenced Regressions On Watershed attributes).
SPARROW models for the coterminous United States for sediment, nitrogen, and phosphorus were used
to characterize baseline water quality conditions. For each regulatory option, reductions in sediment
discharge from constructions sites were predicted as described in Development Document for Final
Effluent Guidelines and Standards for the Construction and Development Category (USEPA 2009b).
These reductions were modeled in SPARROW and resulting changes in in-stream sediments and nutrient
concentrations were evaluated to assess the environment impact of each option.
EPA did not calculate reductions in discharge levels of other pollutants (e.g., turbidity, metals, pesticides)
under each of the alternative regulatory options due to insufficient data on the frequency and magnitude
of their presence in construction site discharges in the United States. It was therefore not possible to
quantify water quality improvements associated with reduction of their discharge. These pollutants are
instead discussed qualitatively in Chapters 2 and 3. Some of these pollutants erode and travel with
sediment discharges and are therefore addressed to some degree by the same practices that reduce
sediment discharges.
This chapter describes the methodology used to estimate the water quality impacts of the alternative
regulatory options and is organized as follows: Section 6.1 describes the general SPARROW modeling
approach. Section 6.2 provides details on the SPARROW sediment model. Section 6.3  describes the
approach used to estimate water quality impacts in estuaries where no model flow predictions were
available from SPARROW. Section 6.4 describes the approach for modeling nutrient reductions
associated with reduced discharges. Section 6.5 provides a brief summary of predicted  sediment loading
reductions that are used in the model. Section 6.6 discusses model results.

6.1     SPARROW Model  Documentation4

6.1.1   General Overview

SPARROW is a watershed modeling technique for estimating contaminant source contributions and
transport in surface waters. SPARROW employs a statistically estimated nonlinear regression model with
contaminant supply and process components, including surface-water flow paths, non-conservative
transport processes, and mass-balance constraints. Regression equation parameters are  estimated by
correlating stream water-quality records with GIS (Geographic Information System) data on pollutant
sources (point and nonpoint) and climatic and hydrogeologic properties (e.g., precipitation, topography,
vegetation, soils, and water routing) that affect contaminant transport. The SPARROW parameter-
estimating procedure also provides measures of uncertainty in water-quality predictions.
SPARROW model infrastructure consists of a detailed watershed - stream reach network to which all
monitoring stations and watershed properties are spatially referenced.  Digital elevation models (DEMs)
are used to delineate watershed  boundaries and to identify overland flowpaths.  The spatially distributed
model  structure allows separate statistical estimation of land- and water-related rates of pollutant delivery
    Discussion adapted from Schwarz et al. (2006).
November 2009                                                                                6-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
from sources to streams, and transport of pollutants to downstream locations within the stream network
(reaches, reservoirs and estuaries). Mechanistic separation of terrestrial and aquatic features of large
watersheds and improved parameter estimation techniques represent significant advances in water-quality
modeling of important contaminant sources and watershed properties that control transport over large
spatial scales. SPARROW has been applied in the analysis of suspended sediment, surface-water nitrogen
and phosphorus nutrients, pesticides, organic carbon, and fecal bacteria, and is potentially applicable to
other measures of water quality. The  SPARROW software is written in Statistical Analysis System
Interactive Matrix Language (SAS IML).
The remainder of this chapter provides a brief description of SPARROW modeling concepts, model
specification, data sources, and estimation results. Appendix C of this report provides more detail on the
nature of the general SPARROW methodology. Appendix D provides more detail on the  suspended
sediment SPARROW model developed by USGS (Schwarz 2008a). Appendix C also provides more detail
on the total nitrogen (TN) and total phosphorus (TP) SPARROW models developed by USGS (Smith et
al. 1997).

6.1.2   Modeling Concept

The broad objective of SPARROW modeling is to establish a mathematical relationship between water-
quality measurements made at a network of monitoring stations and attributes of the contributing
watersheds. The model may be used to satisfy a variety of water-quality information objectives. One
objective is to describe past or present water-quality conditions for a state or region on the basis of
monitoring data. A second objective is to identify and quantify the sources of pollution that give rise to
in-stream water-quality conditions. A third objective is water quality simulation: the ability to portray
counterfactual conditions for specified inputs. Water-quality simulations can depict the in-stream effects
of changes in contaminant sources associated with alternative potential pollution-control strategies,
providing an important step in the analysis of benefits associated with specific Best Management
Practices (BMP) or other regulatory approaches. For the analysis described in this document, SPARROW
is used to evaluate the  effectiveness of a range of regulatory options for controlling construction related
runoff and improving water quality as measured by in-stream Total Suspended Solids (TSS), TN and TP
concentrations. A fourth objective of SPARROW modeling is to test one or more hypotheses about the
nature and importance of factors and processes that may have influenced water quality at the locations
where samples were collected.
The dependent variable in SPARROW models is the mass flux, the mass of contaminant that passes a
specific stream location per unit time, and the laws of conservation of mass apply. Mass accounting in
SPARROW models is supported by the explicit spatial structure defined by the stream network. The
imposition of mass balance greatly improves the interpretability of model coefficients. For example,
assuming mass balance, the coefficient associated with the reach time-of-travel variable is interpreted as a
first-order decay rate. Mass balance provides a basis for contaminant mass flux accounting, whereby mass
flux can be allocated to its various sources, both spatially and topically.  Spatial variability is modeled as a
function of natural and human-related properties of watersheds that influence the supply  and transport of
contaminants. Estimates of long-term mean mass flux are developed from water-quality and streamflow
monitoring data that are regularly collected at fixed locations on streams and rivers.
November 2009                                                                                 6-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.1.3   Model Infrastructure

A flow diagram is provided in Figure 6-1 to illustrate the functional linkages between the major spatial
components of SPARROW models. Mass flux estimates at monitoring stations are derived from station-
specific models that relate contaminant concentrations from individual water-quality samples to
continuous records of streamflow and time. The stream reach, inclusive of its incremental contributing
drainage basin, is the most elemental spatial unit of the SPARROW model infrastructure. Explanatory
data (e.g., climate, topography, land use) are frequently compiled according to geographic units that are
not coincident with the drainage basin boundaries of river reaches. These data may be collected at
different spatial scales and according to spatial units that reflect political boundaries (e.g., counties) or
other non-hydrologic features of the landscape.

                               :k-§§!^
                                                        	.rilllllllllll
                                       Industrial / Municipal
                                         Point Sources
              SPARROW
         Model Components
                         Aquatic Transport
                           Streams
                           Reservoirs
                       In-Stream Flux Prediction
                        Calibration minimizes differences
                       between predicted and calculated
                          mean-annusi loacfs af the
                            monitoring stations
                                                  Water-Quality and
                                                     Flow Data
                                                  Periodic measurements
                                                   at monitoring stations
                  Monitoring
                  Station Flux
                  Estimation
Rating Curve Model
 of Pollutant Flux
 Station calibration to
   monitoring data
       Mean-Annual Pollutant
         Flux Estimation
                                                       Evaluation of Model
                                                    Parameters and Predictions
                Source: Modified from Alexander et a/. (2002a).
The estimation of a SPARROW model requires estimates of long-term mean mass flux from a spatially
distributed set of monitoring stations, having sufficiently long periods of record, inclusive of a wide range
of spatial scales and expressing considerable variation in predictor variable conditions. Long-term mean
mass flux is obtained by relating infrequently collected water-quality samples to continuous
measurements of streamflow and functions of time. In order to have consistency across stations having
different periods of record, it is necessary to detrend the mass flux estimates to a common base year prior
to computing the long-term mean.

A vector-  or raster-based digital representation of the stream and river network topology is the most
fundamental component of the spatial infrastructure that supports the SPARROW model (Figure 6-2).
The node topology is defined according to an upstream (from-) and downstream (to-) node attribute table.
November 2009
                                                                                               6-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
  Figure 6-2: Schematic Illustrating a Vector Stream Reach Network with Node Topology
 	and Water/Contaminant Reach-Node Routing Table	
  The reach-type indicator has possible values of "0" (Stream Reach), "1 "(Impoundment Reach), and "2" Outlet Reach for
                                          Impoundment.
                         12
                         11
                               Reservoir
                              "shorelines
                                 w _^[)ivergent
                             5   ~    Reach
                          Reach Node
                      Stream Reach
                                                     Reach-Node Attribute Table
Reach
Number
13
10
12
11
9
8
7
5
6
4
2
3
1
Up-
stream
Node
13
10
12
11
9
8
7
6
6
5
4
3
2
Down-
stream
Node
10
9
11
9
8
4
6
5
5
4
2
2
1
Diversion
Fraction
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.7
0.3
1.0
1.0
1.0
1.0
Reach -Type
Indicator
0
1
0
1
2
0
0
0
0
0
0
0
0
SPARROW separately estimates the sediment and contaminant attenuation in reservoirs and lakes. The
assessment of pollutant mass flux to coastal estuaries requires an expanded reach network that includes
shoreline features and the identification of reaches that terminate at estuaries. Shoreline reaches are used
to define coastal drainage areas - areas that discharge runoff directly to the estuary without transport
through a stream reach. The SPARROW node and routing architecture also can fully support the
modeling of contaminant transport along "off-reach" (landscape) flow paths according to flow directions
defined by landscape topography as reflected, for example, in DEMs.
The explanatory variables evaluated in SPARROW reflect current knowledge of natural and human-
related sources and the physical,  chemical, or biological properties of the terrestrial and aquatic
ecosystems that affect supply and transport of contaminants in watersheds. Point- and nonpoint-source
variables may include direct measures of the supply of contaminant mass to the landscape and streams
and reservoirs (e.g., municipal and industrial wastewater discharge, fertilizer application), or they may
serve as surrogate indicators of the contaminant mass supplied by point and diffuse sources in watersheds,
such as land-use/land-cover data or census data on human and livestock populations. Climatic and
landscape properties that affect contaminant transport may include measures of water-balance terms (e.g.,
solar radiation, precipitation, evaporation, evapotranspiration), soil characteristics (e.g., permeability,
moisture content), water-flow path properties (e.g., slope, hydraulic roughness, topographic index), or
management practices and activities (e.g., tile drains, conservation tillage, BMPs).

6.1.4   Model Specification

The specification of a SPARROW model consists of identifying the  explanatory variables and functional
forms.  Conceptually, the contaminant mass flux leaving a reach is the sum of two components. The first
component is the mass flux generated within upstream reaches that is delivered to the reach via the stream
network. Losses of contaminant mass from the stream network may  occur at points where flow is
November 2009
                                                                                            6-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
diverted, and contaminant mass will generally be attenuated by stream or reservoir processes. The second
component consists of source loading that is generated within the reach's incremental watershed and
delivered to the stream network within the reach. A number of source-dependent processes affect the
amount of source loading reaching the stream network and transported to the reach's downstream outlet
node. The processes affecting delivery to the stream network are called land-to-water delivery processes,
and may include both surface and sub-surface elements. Land-to-water delivery processes determine the
amount of contaminant generated within an incremental drainage area delivered to the area's
corresponding reach.
Large digital spatial data sets are now readily available, and advances in digital topographic and stream
network data [e.g., National Hydrography Dataset (NHD), National Land Cover Database (NLCD)] have
made spatially distributed modeling more feasible. Inclusion in the SPARROW model of detailed
information on the geographic locations where contaminants are released in watersheds has particular
relevance to the policy and management applications of the model. The source terms used in the model
can be generally classified as intensive and extensive measures of contaminant mass. Intensive measures
are direct measures of pollutant mass, such as fertilizer application, livestock waste, atmospheric
deposition, or sewage-effluent mass flux. The associated source parameter is a dimensionless coefficient
that, together with  standardized expressions  of the land-to-water delivery factor, describes the proportion
or fraction of the source  input that is delivered to streams. Extensive measures of contaminant mass are
surrogate indicators and  include measures of watershed properties such as specific land-use area and
sewered population, considered to be proportional to the actual contaminant mass generated by a
particular type of source. The associated model coefficients are expressed as the contaminant mass
generated per unit of the source type (e.g., kilograms kilometer"2 year"1; kilograms person"1 year"1).
Combined with the land-to-water delivery factor, the standardized source coefficient indicates the mean
quantity of contaminant mass per unit of the surrogate source measure  that is delivered to streams. For
land-use terms, the standardized coefficient gives what is frequently cited as an export coefficient.
Landscape variables in SPARROW describe properties of the landscape that relate to climatic, natural- or
human-related terrestrial processes affecting contaminant transport. The model structure allows tests of
hypotheses about the influence of specific features of the landscape on contaminant transport. Landscape
variables may include water-balance terms (e.g., precipitation, evapotranspiration) related to climate and
vegetation, soil properties (e.g., organic content, permeability, moisture content), topographic water flow-
paths variables (e.g., overland flow, topographic index, and slope), or management practices and
activities, including tile drainage, conservation tillage  practices, and BMPs related to stream  riparian
properties. Particular types of land-use classes, such as wetlands or impervious cover, may also be
potentially used to describe transport properties of the landscape.
Stream attenuation processes that act on contaminant mass as it travels along stream reaches  are
frequently modeled as first-order reaction rate processes. A first-order decay process implies that the rate
of removal of the contaminant from the water column per unit of time is proportional to the concentration
or mass that is present in a given volume of water (a zero-order process corresponds to a constant rate of
removal per unit of time). Accordingly, in basic forms of the SPARROW model, the fraction of the
contaminant mass originating from the upstream node and transported  along a reach to its downstream
node is estimated as a function of the mean water time of travel. Attenuation processes that act on
contaminant mass as it travels through a lake or reservoir are often modeled as a net removal process,
with the loss coefficient expressed as either a first-order reaction rate or a mass-transfer coefficient (also
referred to as an apparent settling velocity). These mass-balance models typically assume steady-state  and
uniformly mixed conditions in the waterbody.
November 2009                                                                                  6-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.1.5  Model Estimation

The SPARROW model equation is a nonlinear function of its parameters, and the model must be
estimated using nonlinear techniques. SPARROW utilizes a nonlinear weighted least squares (NWLS)
estimation method, implemented in SAS. Model parameters are evaluated for statistical significance and
the quantification of uncertainty. The key parameter statistics include the estimated mean coefficient
values, estimated variance of these coefficient estimators, and measures of statistical significance based
on the t statistics. Parameters are also evaluated for physical interpretability, to test hypotheses about the
importance of different contaminant sources and the hydrologic and biogeochemical processes that are
represented in the model. The interpretability of the parameters and their relation to specific processes are
enhanced in SPARROW by the use of a mass balance, mechanistic structure that explicitly separates the
terrestrial and aquatic properties of watersheds and accounts for nonlinear interactions among watershed
properties. The sign of SPARROW model coefficients can be evaluated to determine the direction of
influence of any explanatory variable on in-stream estimates of mean-annual mass flux. Coefficients
associated with source inputs expressed in areal units describe the mass per unit area delivered to streams
from these land areas. Other source coefficients that are expressed in dimensionless units provide a
measure of the fraction of the contaminant that is delivered from each source to streams, rivers, and
reservoirs.

6.2    SPARROW Sediment Model5

The following describes results from a SPARROW suspended sediment model developed for the
coterminous United States and described in Schwarz (2008a), which is reprinted in full in Appendix D of
this document. The analysis is based on mass flux estimates compiled from more than 1,800 long-term
monitoring stations operated by the U.S. Geological Survey (USGS) during the period 1975-2007. The
SPARROW model is structured on the Reach File 1 (RF1) stream network, consisting of approximately
62,000 reach segments. The reach network has been modified to include more than 4,000 reservoirs, an
important landscape feature affecting the delivery of suspended sediment. The model identifies six
sources of sediment, including the stream channel and five classes of land use (urban, forested, federal
non-forested, agricultural and other). The delivery of sediment from landform sources to RF1  streams is
mediated by soil permeability, credibility, slope, and rainfall; streamflow is found to affect the amount of
sediment mobilized from the stream channel. The  results show agricultural land and the stream channel to
be major sources of sediment. Per unit area, federal non-forested and urban lands are the largest landform
sediment sources. Reservoirs are identified as major sites for sediment attenuation.

6.2.1  Sediment Model Data Sources

The spatial framework of the SPARROW sediment model is the  vector-based 1:500,000 scale River RF1
hydrography,  originally developed by EPA (USEPA 1996b) and enhanced to include areal hydraulic load
information for selected reservoirs, shoreline reaches, and reach catchment areas derived from the USGS
Hydro  IK DEM (USGS 2006a). The enhanced network, consisting of 62,776 reach segments, including
shoreline reaches, 61,214 delineated reach catchments, and 2,171 individual reservoirs, has been used to
support numerous national  SPARROW modeling efforts for the coterminous United States. The RF1
reach network for the current SPARROW sediment model was further enhanced by inclusion of areal
hydraulic load information  for approximately 2,000 additional large reservoirs from the National
    Adapted from Schwarz (2008a).
November 2009                                                                                6-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
Inventory of Dams (NID, USAGE 1996). The dependent variable in the SPARROW sediment model is
given by long-term mean sediment mass flux, detrended to the base year 1992. In-stream sediment
concentrations and stream discharge measurements over the water year (WY) period 1975-2007, have
been obtained from the National Stream Quality Accounting Network (USGS 2006b), the USGS
NAWQA Program (USGS 2001), and the USGS National Water Information System (NWIS; USGS
2008), a database encompassing USGS water quality monitoring stations and water quality monitoring
activities done in cooperation with State governments. The sediment model is estimated using 1,828
monitoring stations on the RF1 stream network (Figure 6-3).
While the RF1 network provides reasonably comprehensive national coverage of major rivers, streams
and other surface water bodies, coverage is limited in certain important respects. First, RF1 network
coverage is limited to the coterminous United States, thus excluding Alaska and Hawaii. In addition,
while RF1 l:500,000-scale network reaches have associated data or estimates of stream discharge and
velocity that are required to specify the SPARROW model, the network excludes the majority of the
nation's total stream mileage, and smaller streams in particular. The linear coverage of the RF1 network is
approximately 700,000 miles (USEPA 2007e). By contrast, coverage of the USGS National Hydrographic
Dataset (NHD) at 1:24,000 - 1:100,000 scale is currently over 7 million miles (USGS 2007b). Non-
homogeneity of coverage may lead to an under-estimation of the impacts of construction-related sediment
on smaller stream reaches. As construction activities may be concentrated along lower-order streams not
included in the RF1 network, the relative share of total sediments contributed by construction activities in
these reaches may be high during active construction. By contrast, the specific impacts of construction
activities may diminish in magnitude relative to  contributions from spatially extensive and diffuse land
uses, including agriculture, at the spatial level of RF1 reaches.
Data on land cover and land use have been  developed from the 2001 USGS National Land Cover Data Set
Retrofit Change Product (Multi-Resolution Land Characteristics Consortium (MRLC) 2001), derived
from Landsat TM/ETM remotely sensed imagery at 30-meter resolution and classified according to the
eight Anderson Level I categories. For use within SPARROW, data were transformed to  1 km2 cells
within a Lambert map projection as consistent with Hydro IK. The 1 km2 cells were then resolved to
catchments associated with specific RF1 reaches. For model estimation, land use was given by the 1992
values of the 2001 NLCD  Retrofit Change Product. The 2001 values of the Retrofit Change Product were
used to simulate water quality conditions for 2001. Federal range and barren land was included in the
model separately from private land. Federal land extent, taken from the Federal Land coverage of the
National Atlas (USGS 2003) and transformed to 1 km2 cells, was apportioned using the 1 km2
transformation of the 1992 NLCD Change Product for Anderson Level 1 range and barren land classes.
For model simulation of 2001 conditions, federal range and barren land were similarly estimated using the
land use estimates from the 2001 NLCD Retrofit Change Product.
November 2009                                                                               6-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
  Figure 6-3: Location of 1,828 Water Quality Monitoring Stations Used in the SPARROW
 	Sediment Model, in Relation to the Reach File 1 (RF1)  Reach Network	
              *X   VAWi43£
Variables determining the delivery of contaminants from the land to RF1 streams include soil erodibility
(RUSLE k-factor), soil permeability, mean slope, and precipitation (RUSLE r-factor). Slope, soil
erodibility, and permeability were obtained from the State Soil Geographic (STATSGO) database (USDA
1994), converted to a 1-km2 grid and averaged over RF1 catchments. The RUSLE rainfall factor was
derived by interpolating a digitized national map of rainfall factor isoline contours, creating a continuous
1 km2 grid surface, which was subsequently averaged over individual RF1 catchments.

6.2.2  Model Estimation  Results

The nonlinear least squares estimation results of a preliminary version of the SPARROW suspended
sediment model are given in Table 6-1. The model includes six source terms: five measured by area of
specific land use (expressed in km2), and the sixth by the length of stream channel. The five land use
sources are urban land, forested land, federal nonforested land, agricultural land, and other land. The
federal land class excludes federal forested land, which is incorporated in the forested land class.
Agricultural land includes cropland, pasture land and orchards. Other land consists of nonfederal range
and barren land. Among all the land classes, only wetlands and land covered by water, ice or snow are
excluded as a potential source. The source described as "streambed" relates to stream channels as a direct
source of sediment, and is measured in terms of stream length (expressed in meters). The model specifies
two in-stream  sediment attenuation processes: attenuation in streams, distinguished by three streamflow
classes (below 500 cubic feet per second (cfs), 500-1,000 cfs, and greater than 1,000 cfs); and reservoir
attenuation, specified as a function of areal hydraulic load.
November 2009
                                                                                         6-8

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 6-1: Estimation Results for the SPARROW Suspended Sediment
Parameter
Source Coefficients
Urban land
Forested land
Federal nonforested land
Agricultural land
Other land
Streambed (reach length)
Land to Water Delivery
Slope
Soil permeability
R-factor
K-factor
Flow [< 500 cfs] (Reach)
Flow [> 500 cfs] (Reach)
Stream Attenuation Factors
Travel time (Q < 500 cfs)
Travel time (500 < Q < 1,000 cfs)
Travel time (Q > 1,000 cfs)
Reservoir settling velocity
Number of Observations
RMSE
R-square
Units
kg/km2/yr
kg/km2/yr
kg/km2/yr
kg/km2/yr
kg/km2/yr
kg/m/yr
Factors
day'1
day'1
day"1
m/yr
1,828
1.414
0.711
Estimate
47,130
634
64,344
18,047
11,343
28.80
0.804
-0.778
0.821
1.292
0.154
0.721
-0.007
-0.233
0.009
36.49

Standard Error
9,925
898
12,411
3,623
3,186
6.40
0.087
0.094
0.081
0.279
0.100
0.354
0.016
0.057
0.047
5.552

Model
p-value
0.000
0.480
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.125
0.042
0.673
0.000
0.854
0.000

Source: Schwarz (2008a).
With the exception of forested land, all of the source variables are highly statistically significant. The
largest intrinsic sediment yield is associated with federal range and barren land; urban land has the second
highest intrinsic yield. Stream channels are also a statistically significant source of sediment. Land-to-
water delivery for land sources is strongly mediated by the four delivery variables: soil permeability, soil
erodibility, slope, and rainfall. With the exception of soil permeability, an increase in these factors results
in an increase in the share of sediment delivered to streams. Permeable soils reduce the delivery of
sediment, presumably because more runoff infiltrates, leaving less overland flow to carry sediment.
Greater streamflow causes an increase in the amount of sediment generated from stream channels, with
the largest effect associated with streams with flow greater than 500 cfs. Reservoir retention is statistically
significant and indicates sediment settles at a mean velocity of 36 meters per year. The preliminary model
does not find evidence of sediment  attenuation in streams, as in-stream attenuation in small and large
streams is not statistically different  from  zero. Additional investigation would be necessary to determine
if this result is real, or if there are additional reach attributes, currently absent from the model, that
identify a subset of reaches where sediment attenuation takes place.

6.2.3  Model Simulation

The estimated SPARROW suspended sediment model for base water year 1992 is used to simulate water-
quality conditions for 2001, with and without proposed changes in the regulation of construction activity.
The simulation of suspended sediment mass flux for water year 2001 without changes in regulation is
obtained with all land use-related source  variables set to 2001 values based on the NLCD Retrofit Change
Product. Flow weighted concentration is  estimated by dividing simulated mass flux estimates by mean
streamflow over the period WY  1975-2006. Although the SPARROW model does not explicitly include a
term for construction loading, such  loading is implicitly accounted for in the urban land component of the
model. Therefore, the 2001 pre-compliance loading from construction (the "base-case" scenario loading)
November 2009
                                                                                              6-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
is incorporated in the 2001 mass flux attributed to urban land that is obtained by evaluating the urban land
variable in the SPARROW model using the 2001 NLCD Change Product value.
The absence in the SPARROW model of an explicit term for construction loading requires the
development of an indirect method for assessing changes in sediment mass flux arising from different
construction industry regulation scenarios. Suspended sediment loading under alternative regulation
scenarios has been estimated by EPA using a variation of the USLE. The USLE method determines the
amount of soil that is mobilized and delivered, under a regulatory scenario, to the edge of a construction
site. To evaluate the impact that changes in these loadings have on RF1 stream sediment mass flux and
concentration, it is first necessary to assess the rate at which "edge of site" loads are subsequently
delivered to RF1 streams. We use the estimated rate of delivery from agricultural land (explicitly included
in the model) that can be factored into a mobilization ("edge of site") delivery component and a stream-
delivery component. The method uses edge-of-site measurements of agricultural sediment erosion from
the 1992 National Resources Inventory (NRI) to isolate the stream-delivery component from the overall
rate of delivery from agricultural land and applies this component to the change in construction loading to
determine the change in loading to RF1  streams. Thus, the approach assumes that the delivery of sediment
from the edge of a site to an RF1 stream is the same for both construction and agriculture activities;
although the mobilization and delivery of sediment to the edge of the site between these activities is
allowed to differ.

6.3     Extension of SPARROW to Estuaries and Coastal Waters

EPA used SPARROW output to evaluate sediment deliveries to estuaries. While SPARROW predicts
mass flux in all reaches, it does not estimate concentrations in reaches where flow estimates are not
available. Of the 2,067 estuarine/coastal reaches in the RF1 dataset, SPARROW provided TSS
concentrations in more than 92 percent of the reaches. Of those reaches, 772 are retained for water quality
analysis based on the availability of data for other parameters needed in calculating Water Quality Index
values (see Chapter 10).  Of the 772 reaches used in the analysis, SPARROW provides TSS
concentrations for 729 reaches in this set directly. In the remaining 43 estuarine reaches, sediment
concentrations were calculated using the SPARROW predicted mass flux and dissolved concentration
potentials (DCPs) for estuaries (USEPA 1997). The National Oceanic and Atmospheric Administration
(NOAA) developed DCPs to provide an approach for predicting the ambient concentration of
conservative contaminants that are introduced into an estuary and are then subject to mixing and dilution.
Using this approach, EPA was able to estimate TSS concentrations associated with annual quantities of
sediment discharged into estuaries as predicted by SPARROW without detailed hydrodynamic modeling
of estuaries.
Concentration calculations using DCPs presume that water quality constituents are dissolved,
conservative pollutants dispersed under well-mixed, steady-state conditions. These criteria are not strictly
met for TSS, although the finest particle fractions approximately meet these criteria due to their extremely
low settling velocities and relatively stable composition. Because current and proposed sediment
abatement practices remove coarser particle fractions first, construction site sediment discharges are
expected to contain primarily finer particles. Therefore, the DCP methodology criteria are assumed to be
applicable for estimating changes in estuary water quality associated with changes in construction site
sediment discharge.
NOAA has calculated DCP values for most major estuaries in the Southeast (NOAA and EPA 1989a);
Northeast (NOAA and EPA 1989b); Gulf of Mexico  (NOAA and EPA 1989c); and West Coast (NOAA
November 2009                                                                               6-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
and EPA 1991) of the coterminous United States. For these estuaries, DCP values and baseline TSS
concentrations calculated using DCPs are shown in Appendix E in Table E-l through Table E-4 for
Southeastern, Gulf, Northeastern, and West Coast estuaries, respectively. Data on these and additional
estuaries are found via the NOAA Coastal Geospatial Data Project (http://coastalgeospatial.noaa.gov).
Total sediment loadings estimated in SPARROW for reaches draining to estuaries were aggregated on the
basis of 5-digit NOAA Estuarine Drainage  Area (EDA) codes. Estimated loadings were calculated in
SPARROW as Suspended Sediment Concentrations (SSC). Suspended Sediment Concentrations needed
to be adjusted before the application of DCP methods, for two reasons. First, reported SSC values include
all particle size fractions and coarse sediment fractions are not assumed to behave as conserved within the
water column due to higher settling velocities. Approximately 127,000 sediment records for which
particle size distributions were recorded, collected at 1805 water quality gauging stations obtained from
USGS  (Schwarz 2008b) were analyzed. It was found that on average over these  samples, 78.4 percent of
suspended sediment mass was classified as "fine" (diameter < 0.0625 mm). This percentage varied to
some extent by 2-digit USGS Water Resources Regions, from 71  percent (Region 14) to 89.5 percent
(Region 9); and was subject to even wider at-site inter-seasonal and inter-annual variations. It is assumed
that the implied sediment size distribution of SPARROW output is equivalent to observed distributions.
However, only 15 of the 1,805 water quality gauging stations for which sediment size fraction
information was available were terminal or coastal reaches (draining directly into bays or estuaries),  so
that no regional or estuary-specific sediment size distributions could be identified.
A second consideration is that USGS (2000) has determined that SSC and TSS, while carrying the same
physical interpretation, are subject to systematic differences in measurement due to  differing laboratory
protocols. SSC values in the SPARROW output are  divided by 1.3 to adjust for this difference, as
described in Section 6.6. The use of this adjustment  factor yields TSS concentrations that are roughly 77
percent of corresponding SSC values. Since this adjustment is reasonably close to the nation-wide
(average of samples) percentage of fine particles in SSC (78 percent), SPARROW SSC values were
converted to TSS equivalent, and the TSS values assumed to conform approximately to the assumptions
underlying  DCP calculations. In order to determine how well these assumptions are reflected in observed
estuarine water quality conditions, data on suspended sediment size distribution  for 623 samples taken
within  San  Francisco Bay were obtained from the USGS (http://sfbay.wr.usgs.gov/access/wqdata/). The
percentage  of fine particles (< 0.0625 mm) was 71 percent for these samples. It is evident that suspended
sediment accounting within estuaries is complex, and ambient TSS levels, in addition to the percentage of
fine particles, are observed to reflect many influences beyond the annual riverine input of sediment. As an
additional check on SPARROW outputs to estuaries, the aggregated annual loadings per unit of
contributing watershed area were calculated for each channel contributing sediment to estuaries, and
found to be reasonable in most cases.
Annual sediment loads as estimated by SPARROW were aggregated by estuary, unit-converted and
divided by  10,000 tons per year (the DCP unit loading), and multiplied by DCP coefficients to obtain
ambient estuarine TSS concentrations. Baseline estimates of estuarine TSS concentrations are presented
in Appendix E in Table E-l through Table E-4. Concentrations varied from below 1 mg/L (one
Southeastern and multiple Northeastern estuaries) to a maximum  of 833 mg/L (San Antonio  Bay, TX),
and averaged 71.5 mg/L over all  estuaries.
SPARROW-simulated TSS concentrations were compared to measured values from monitoring data.
Measured observations were obtained from the EPA National Coastal Assessment
(http://www.epa.gov/emap/nca/html/data/index.html). and augmented for San Francisco Bay with data
November 2009                                                                               6-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
from USGS (http://sfbav.wr.usgs.gov/access/wqdata/). Estimated TSS concentrations obtained from
SPARROW predictions and application of the DCP approach in general exceed measured values,
although agreement is remarkably good in many instances. In particular, DCP estimates of TSS for
Delaware Bay (40.0 mg/L) are very close to observed values (38.45 mg/L), and virtually identical for
Long Island Sound (9.7 mg/L predicted vs. 9.8 mg/L observed). Results for San Francisco Bay illustrate
the difficulty in comparing model predictions with observations. DCP TSS estimates of 89.3 mg/L are
slightly below the NOAA EMAP average of 71 observations (132 mg/L); but above the mean of the
larger USGS sample (34 mg/L). Results suggest that SPARROW sediment loading predictions,
transformed by the DCP approach, provide a reasonable basis for evaluating changes in TSS due to
specific regulatory options, although predicted TSS values may differ from physical observation in
absolute value in certain settings.
In the final analysis, the DCP approach was only used for the 43 reaches without flow predictions in
SPARROW. The number of reaches in each estuary where the DCP approach was used to calculate TSS
concentrations is shown in Appendix E in Table E-5 through Table E-8 .

6.4     Modeling Changes in Nutrient Concentrations

Reducing soil erosion and sediment runoff from construction sites not only reduces sediment loadings
into streams, but can also reduce concentration of other pollutants that are present in the soil. To evaluate
the water quality benefits of these ancillary reductions, EPA developed an approach that relies on the
empirical relationship observed between baseline in-stream suspended sediment concentrations and TN
and TP concentrations to derive estimates of the post-compliance nutrient reductions.
In evaluating potential approaches to estimating in-stream nutrient concentration reductions EPA faced
several key challenges: (1) locating nationally available data on the nutrient content of soil or sediment at
sufficient resolution; (2) accounting for the impact of the sediment delivery processes from edge of site to
the stream on nutrient concentrations;  (3) and estimating total nutrients as compared with particulate,
dissolved, and other forms of nutrients. These challenges are in addition to the overall challenge of
modeling the complexity of the nitrogen or phosphorus cycles and delivery processes, on a national scale.
Based on literature, EPA determined that most of the total nitrogen and phosphorus in sediment
associated with  runoff are in the solid phase and therefore there is little difference between total and solid
phase nutrient concentrations (Schuman et al. 1973a, b, as cited in Sharpley 1985; Haith 2009). It is
estimated that up to approximately 90  percent of nutrients delivered to streams are bound to sediments
(Schuman et al. 1973a, b, as cited in Sharpley 1985; Daniel et al. 1979); this is particularly true in the case
of phosphorus and organic nitrogen. Several  studies of construction site discharges have shown high
correlations between sediment discharges and nutrients discharges (Daniel et al. 1979; Novotny and
Chesters 1989; Harbor et al. 1995).  Studies also indicate the ability of several construction site sediment
erosion and control technologies to reduce nutrient (and other pollutants, e.g., metal, hydrocarbons,
bacteria) discharges from construction sites (Homer et al.  1990; Stahre and Urbanos 1990; Bhaduri et al.
1995; Harbor et al. 1995; Lentz et al. 2002; USEPA 2006b; Bachand  et al. 2007; Faucette et al. 2008;
USEPA 2009b). Therefore, EPA assumed that the change in sediment-bound TN and TP would be
representative of the overall change in TN and TP delivered to streams. While this simplification does not
account for the complex processes that control nutrient fluxes through streams, EPA believes that it still
provides reasonable estimates of the scale of changes that may be  directly related to changes in sediment
discharge from construction sites.
November 2009                                                                                6-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.4.1   Modeling Approach Based on SPARROW Nutrient Models

USGS developed SPARROW models for TN and TP transport for the nontidal coterminous United States
as described in Smith et al. (1997). As with the sediment model, the SPARROW nutrient models
regression equations relate measured nutrient transport rates in streams to spatially referenced descriptors
of pollution sources and land-surface and stream-channel characteristics. Observed TN and TP transport
rates were derived from water-quality records for 414 stations in the National Stream Quality Accounting
Network. Nutrient sources identified in the regression equations include point sources, applied fertilizer,
livestock waste, nonagricultural land, and atmospheric deposition (TN model only). Surface
characteristics found to be significant predictors of land-water delivery include soil permeability, stream
density, and air temperature (TN model only).
The delivery processes, explanatory variables, and their coefficients are slightly different in each of the
three SPARROW models (sediment, TN, and TP). As a result,  it was  not possible to utilize the nutrient
models to predict changes in nutrient concentrations associated with construction sediment discharge by
specifying the same input parameters used in the SPARROW sediment model. However, since the three
models are based on the same hydrological network and baseline conditions, it is possible to relate
predicted concentrations at the level of individual stream reaches. In this application, EPA used the
baseline estimates for the three models to determine ratios between TN and sediment and between TP and
sediment that were then used in conjunction with the predicted change in sediment concentrations to
determine the change in nutrient concentrations, as represented by the following equation

                                                                                            . 6-1)
                                                               baseline

where A[nutrient] represents the change in nutrient concentration associated with the change in TSS
concentration (A[TSS]). The variable [nutrient] represents the in-stream concentration of either TN or TP.

The nutrient-to-TSS ratio ([nutrient]/[TSS]) is calculated using SPARROW model outputs of TSS, TN
and TP mass flux related to land-based runoff for the baseline conditions, as indicated in the equations
below. The ratios were calculated for each individual reach as follows (where IL = incremental sediment
mass flux per reach):

                   IX " \baseline   V  TN , total ~ ^^TN , population-related ~ ^^TN , atmospheric deposition )           (Ffl f\ ?^
                   [TSS]baseljne                      ILTSStotal


                             \_ll \baseline    \  TP,total ~   TP, population-related '
                            [TSS}haselme             IL
                                                                                         (Eq. 6-3)
                                                     'TSS,total
Note that, since the regulation specifically targets reducing stormwater runoff and the associated sediment
and nutrients and therefore does not address other potential sources of nutrients, EPA excluded from the
calculations of the nutrient-to-sediment ratios nutrient loadings associated with population and
atmospheric deposition sources.6 To mitigate the effects of outlier TSS concentration on the estimated
nutrient-to-sediment ratio, the top 5 percent of nutrient-to-TSS ratios were replaced by the 95th percentile
6   An example of population-related sources is effluent from publicly owned treatment works (POTWs). Atmospheric
    deposition is only a source in the nitrogen model.
November 2009                                                                                  6-13

-------
Environmental Impact and Benefits Assessment for the C&D Category
value of the distribution. This is similar to the approach used to summarize SPARROW-predicted TSS
concentrations at the national level which replaces values above the 95th percentile with the 95th percentile
value. National statistics of the reach-specific nutrient-to-sediment ratios are summarized in Table 6-2.
 Table 6-2: Nutrient-to-Sediment Ratios
                         Average
                                               Median
                                           Min
                                           Max
 TN/TSS
0.006
0.005
0.000
0.028
 TP/TSS
0.001
0.001
0.000
0.003
EPA averaged the reach-level ratios to the HUC8 watershed level to obtain a nutrient-to-sediment ratio
that represents the average relationship between in-stream nutrients and suspended sediments within each
basin.  EPA later uses these watershed-average ratios to estimate reach-specific changes in suspended
sediment concentrations for each scenario to yield estimated reach-level changes in nutrient
concentration.
EPA compared the ratios described above to values found in the literature, databases, and fate and
transport models (Table 6-3). As shown in the table, the nutrient-to-sediment ratios calculated as
described above  based on SPARROW model outputs fall within the range of values published in the
literature for both TN and TP. While differences are expected between soil and in-stream sediment
concentrations, the values for both range widely in the United States and are shown in Table 6-3 for
reference.
Table 6-3: Literature Values of Nitrogen and Phosphorus in Soil and Sediment
Data Source
Ratios used in this analysis
based on SPARROW1
Haithetal. (1996)
Sednet
(Wilkinson et al. 2004)
Tisdaleetal. (1985)
Haithetal. (1996)
Havlinetal. (1999)
NRCS (2008)2
SWAT (2005)
Data Type
Modeled values
based on Held data
Field data
Field data
Field data
Field data
Modeled values based
on field data
Field data
Modeled values based
on field data
Soil/Sediment
Sediment
Sediment
Subsoils
Natural, top 1 ft
Surface 30 cm
Surface soil
Soil (varied)
Surface soil
Nitrogen (%)
0 to 2.77
0.30
0.1
0.03 to 0.4
<0.05 to >0.2

0 to 3.57
min 0.0073
Phosphorus (%)
0 to 0.32
0.13
0.025

<0.09 to 0.68
0.005 to 0.1 5
0.0001 to 0.58
0.00254
 Notes: Shaded rows indicate ranges that reflect geographically-specific data. For example, the Natural Resources Conservation Service
 (NRCS 2008) dataset is based on over 4,000 and 3,500 data points for TN and TP concentrations in soil, respectively, with a distribution of
 samples across the continental United States. The data provide useful references when evaluating the range of ratios obtained from
 SPARROW outputs but are insufficient for extrapolation to a reach-specific analysis.
 1 Presented ranges are for reach-level values excluding top 5 percent.
  Range of soil values seen in the continental United States (TN: 4,176 data points; TP: 3564 data points).
 3 Nitrate only.
 4 Mineral pool concentrations only.
November 2009
                                                                                                     6-14

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.5    Pollutant Load Modeling

The general procedures that EPA used to estimate sediment load reductions from construction sites are
described in EPA's Development Document for Final Effluent Guidelines and Standards for the
Construction and Development Category (USEPA 2009b). The results of this analysis are also detailed in
that document. This analysis only considers stormwater sediment discharges to receiving waters and does
not analyze dry-weather increases in surface water sediment and turbidity levels from dewatering
discharges, wind deposition, construction activity taking place in surface waters, groundwater seepage,
and vehicle and equipment washwaters.
Table 6-4 presents a national summary of sediment loadings calculated by EPA for the baseline and four
regulatory options, as well as reductions under each regulatory option. Option 1 is expected to result in a
reduction of 34 percent of construction sediment discharges nationally, while Option 2, Option 3, and
Option 4 are expected to result in reductions of 70 percent, 87 percent, and 77 percent, respectively. Total
reductions under Option 2 are 1.80 million tons of sediment, 2.25 million tons under Option 3, and
1.98 million tons under Option 4 compared to 870,000 tons under Option 1.

 Table 6-4: Baseline Construction Sediment Loading Summary  and Post-Compliance
 Reductions
Scenario
Baseline
Option 1
Option 2
Option 3
Option 4
Sediment Loading
(million tons)
2.58
1.71
0.78
0.34
0.60
Sediment Discharge
Reduction (million tons)
—
0.87
1.80
2.25
1.98
Percent Reduction from
Baseline
—
34%
70%
87%
77%
Table 6-5 presents estimates of the amount of sediment entering reaches from construction related activity
under each of these four options, by EPA region. The table shows the baseline amount of sediment
entering reaches as well as post-compliance conditions and reductions for a regulatory option.
November 2009                                                                               6-15

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 6-5: Construction Sediment Loading Summary and Reductions by Option and EPA Region
EPA
Region
1
2
3
4
5
6
1
8
9
10
Total
Baseline
Loading
(mil. tons)
0.02
0.04
0.15
0.90
0.28
0.85
0.23
0.04
0.04
0.03
2.58
Option 1
Loading Reduction Percent
(mil. tons) (mil. tons) Reduction
0.01 0.01 34%
0.02 0.01 34%
0.10 0.05 34%
0.60 0.30 34%
0.19 0.10 34%
0.56 0.29 34%
0.15 0.08 34%
0.03 0.01 34%
0.02 0.01 34%
0.02 0.01 34%
1.71 0.87 34%
Option 2
Loading Reduction Percent
(mil. tons) (mil. tons) Reduction
0.01 0.01 69%
0.01 0.02 69%
0.05 0.11 70%
0.27 0.63 70%
0.09 0.20 70%
0.25 0.60 70%
0.07 0.16 70%
0.01 0.03 70%
0.01 0.03 70%
0.01 0.02 68%
0.78 1.80 70%
Option 3
Loading Reduction Percent
(mil. tons) (mil. tons) Reduction
0.00 0.02 85%
0.00 0.03 86%
0.02 0.13 87%
0.12 0.78 87%
0.04 0.25 87%
0.11 0.74 87%
0.03 0.20 87%
0.01 0.04 87%
0.00 0.03 87%
0.00 0.02 85%
0.33 2.25 87%
Option 4
Loading Reduction Percent
(mil. tons) (mil. tons) Reduction
0.01 0.01 58%
0.01 0.02 69%
0.04 0.11 73%
0.22 0.67 75%
0.07 0.21 74%
0.16 0.69 81%
0.05 0.18 79%
0.01 0.03 78%
0.01 0.03 77%
0.01 0.01 56%
0.60 1.98 77%
 Note: Due to rounding, percentage reduction may not match the shown (rounded) reduction divided by the shown (rounded) baseline.
November 2009
                                                                                                                              6-16

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.6    Water Quality Modeling Results

This section presents the results of water quality modeling using the SPARROW model for sediment and
the above described estimation approach for nutrients. The SPARROW sediment model simulations
generate estimates of water sediment content as Suspended Sediment Concentration (SSC), as consistent
with water quality gauging records used to estimate SPARROW sediment models. By contrast, ambient
sediment concentrations measured as Total Suspended Solids (TSS) are required in the calculation of the
water quality index (details of the water quality index can be found in Chapter 10). Although SSC and
TSS are interpreted in a similar manner as measures of the concentration of suspended solid-phase
material in surface water bodies, Gray et al. (2000) have determined that differences in laboratory
protocols result in systematic differences between SSC and TSS when measuring a given ambient
concentration of suspended sediment. SSC measurements are systematically higher than TSS, particularly
when the sand component of total sediments exceeds roughly 25 percent of total (dry) sediment mass.
Therefore, for the analysis described in this document, SPARROW estimates of SSC have been divided
by a factor of 1.3 to obtain the corresponding TSS values.
Construction activity discharges impact the majority of RF1 reaches in the coterminous United States.
During the nine-year period from 1992 through 2001, approximately 71 percent of RF1 reach watersheds
contained some level of construction. Construction discharges differ from other types of industrial
discharges in this respect. Other types of industrial dischargers tend to be fewer in  number and less widely
distributed across the United States.
The intensity of construction activity, however, varies widely among individual RF1  watersheds.
Approximately 43,900 RF1 watersheds had a net increase in urban acreage during  the 1992-2001 time
period. Approximately 17,200 RF1 watersheds had no change (-16,400 watersheds) or a minor decrease
(-800 watersheds) in urban acreage. Figure 6-4 illustrates the highly uneven distribution of construction
acres among RF1 watersheds during the 1992-2001 time period. Watersheds represented in the figure
account for 93 percent of all construction acres during the period, as shown in Figure 6-5, which shows
the cumulative distribution of acres of construction by RF1 watersheds in decreasing order of construction
acres by watershed and includes the  national total number of acres of construction  during the period. As
the two figures show, relatively few  watersheds account for a large fraction of acres of construction
nationally. In Figure 6-5, for example, the 5 percent of RF1 watersheds containing the largest number of
construction acres per watershed encompassed, as a group, over 2.4 million construction acres, or nearly
50 percent of all construction acreage for the 1992-2001 time period.
November 2009                                                                               6-17

-------
Environmental Impact and Benefits Assessment for the C&D Category
      	Figure 6-4: Plot of Construction Acres by RF1 Watershed Percentile Group (1992-2001)


       1,400,000  -,
    o
    O
    in
   1
       1,200,000
   tM
   a>
   a>

   ~  1,000,000


   e
   o
   £   800,000  -

   oi
   Q.
   ^

   SL



   |   600,000
        400,000
        200,000
                1%   6%   11%  16%  21%  26%   31%  36%  41%  46%   51%   56%  61%  66%   71%   76%  81%  86%   91%   96%


                Most construction acres per watershed                                                  Least construction acres per watershed


                                             RF1 Watershed Percentile Rank Based on Total Construction Acres
November 2009
                                                                                                                                   6-18

-------
Environmental Impact and Benefits Assessment for the C&D Category
                        Figure 6-5: Cumulative Plot of Construction Acres by RF1 Watershed (1992-2001)
      5,500,000 -,
      5,000,000
      4,500,000
                  Total Construction Acres (1992-2001) = 5,297,286
               1%   6%   11%  16%  21%   26%   31%   36%  41%  46%  51%  56%   61%   66%   71%  76%  81%  86%   91%   96%
               Most construction acres per watershed                                                   Least construction acres per watershed

                                            RF1 Watershed Percentile Rank Based on Total Construction Acres
November 2009
                                                                                                                                      6-19

-------
Environmental Impact and Benefits Assessment for the C&D Category
Figure 6-6 through Figure 6-15 illustrate the distribution of construction acreage by EPA region.
      Figure 6-6: EPA Region 1: Percent Urban Change 1992-2001 by RF1 Watershed
                                           % Developed from 1992-2001
                                                     ] Miles
     Figure 6-7: EPA Region 2: Percent Urban Change 1992-2001 by RF1 Watershed
                 '/„ Developed from 1992-2001
                     0.01%-0.1%
                     0.11%-0.5%
                    | 0.51%-2%
                    I >2%
                                           ] Miles
November 2009
                                                                                       6-20

-------
Environmental Impact and Benefits Assessment for the C&D Category
     Figure 6-8: EPA Region 3: Percent Urban Change 1992-2001 by RF1 Watershed
                                                   % Developed from 1992-2001
                                                       0,01% -0.1%
                                                       0.11% -0.5%
                                                       0.51% -2%
     Figure 6-9: EPA Region 4: Percent Urban Change 1992-2001 by RF1 Watershed
               % Developed from 1992-2001
                   < 0.01%
                   0.01%-0.1%
                   0.11%-0.5%
                   0.51% -2%
November 2009
                                                                                       6-21

-------
Environmental Impact and Benefits Assessment for the C&D Category
     Figure 6-10: EPA Region 5: Percent Urban Change 1992-2001 by RF1 Watershed
                     . Developed from 1992-2001
                       < 0.01%
                       0.01%-0.1%
                       0.11%-0.5%
                     • 0.51% -2%
                     • > 2%
                                                                       ] Miles
     Figure 6-11: EPA Region 6: Percent Urban Change 1992-2001 by RF1 Watershed
                                                      *    -
        % Developed from 1992-2001
            0.01% -0.1%
            0.11% -0.5%
          J! 0.51% -2%
                                                                       ] Miles
November 2009
                                                                                       6-22

-------
Environmental Impact and Benefits Assessment for the C&D Category
     Figure 6-12: EPA Region 7: Percent Urban Change 1992-2001 by RF1 Watershed
                      % Developed from 1992-2001
                         0.01% -0.1%
                         0.11% -0.5%
                         0.51% -2%
2:0
       D Miles
     Figure 6-13: EPA Region 8: Percent Urban Change 1992-2001 by RF1 Watershed
                                                           % Developed from 1992-2001
November 2009
                                                                                     6-23

-------
Environmental Impact and Benefits Assessment for the C&D Category
     Figure 6-14: EPA Region 9: Percent Urban Change 1992-2001 by RF1 Watershed
                                                % Developed from 1992-2001
                                                    •=0.01%
                                                    0.01%-0.1%
                                                    0.11%-0.5%
                                                    0.51%-2%
    Figure 6-15: EPA Region 10: Percent Urban Change 1992-2001 by RF1 Watershed
                                                     Developed from 1992-2001
                                                       = 0.01%
                                                       0.01%-0.1%
                                                       0.11%-0.5%
                                                      | 0.51%-2%
Because construction sites are distributed so unevenly among watersheds, this chapter summarizes water
quality information in two ways. One set of tables included in this chapter presents information for the
relatively small number of RF1 watersheds that contain most construction acreage for the 1992-2001 time
November 2009
                                                                                         6-24

-------
Environmental Impact and Benefits Assessment for the C&D Category
period. A second set of tables presents information for all RF1 reaches directly receiving discharges from
construction sites, regardless of the level of construction activity in their watersheds.
An important factor to consider when examining the data summarized in the tables below is that episodic
precipitation events are the primary cause of construction site discharges to surface waters. Most TSS and
associated nutrient discharges from construction sites, therefore, take place during or shortly after
precipitation events. Once a precipitation event ceases, discharges from construction sites generally cease
within a short time period. The estimates presented in the tables below, however, are not intended to
reflect surface water TSS and nutrient concentrations associated with individual, episodic storm events.
They instead represent average concentrations estimated to exist in RF1 reaches over multi-year time
periods. EPA uses these long-term estimates because the data and modeling resources currently available
to EPA do not permit a finer level of time resolution.
EPA expects that, in general, surface water TSS and nutrient concentrations in the short time periods
following storm events would be much higher than the concentrations presented in the tables below. TSS
concentrations as high as 7,000 mg/L have been documented due to construction site discharges in
downstream surface waters following precipitation events (Wolman and Schick 1967; Chisholm and
Downs 1978; Downs and Appel 1986). Sediment and nutrient concentration reductions for the evaluated
regulatory options would be higher, as well. During periods when there is no precipitation event to cause
a discharge from a construction site, surface water sediment and nutrient concentrations would generally
be lower.
A second important factor to consider when examining the data in the tables below is that construction
site sediment and nutrient discharges also have strong geospatial variability. Geospatial differences occur
as a result of variation in soil type, topography, distance from construction site to surface water, receiving
surface water characteristics, and weather patterns, among other factors. These differences  are reflected to
a certain extent by the differences in surface water TSS and nutrient concentrations among  different
groups of reaches, as reflected by reaches having progressively greater construction acreage within their
watershed (presented in  Table 6-16 through Table 6-20) and among the various EPA regions (presented in
Table 6-22 through Table 6-31). In general, however, the data presented below are averaged over large
geographic regions because of limitations in the data and modeling resources currently available to EPA.
When summarizing TSS and nutrient concentrations, EPA calculated statistics based on reach-length
weighted concentrations to account for the wide variation among reaches in length. EPA believes that this
adjustment provides a more representative characterization of the distribution of concentrations by giving
more weight to the longer reaches. Additionally, EPA has replaced concentrations that have
concentrations above the 95th percentile (e.g., TSS greater than 6,157.8 mg/L under baseline conditions)
with the 95th percentile value since these reaches could potentially be considered outliers. Finally, the data
presented in the tables are for those reaches receiving direct discharges from construction sites. The
decrease in loadings and concentrations in these reaches also translate into decreased concentrations in
downstream reaches. While improvements in these downstream reaches are included in the benefits
analysis, changes in concentration occurring downstream reaches that do not receive direct discharges
from construction sites are not included in the tables in this chapter.
While summaries cannot adequately describe the full extent of temporal  and geospatial variations in TSS
and nutrient concentrations, they are nevertheless useful for interpreting broad-scale changes in water
quality as a result of the  evaluated regulatory options.
November 2009                                                                                  6-25

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.6.1  Current Water Quality Impacts to Surface Water from Construction Site
       Discharges

To estimate water quality impacts associated with construction site discharges under current conditions,
EPA used SPARROW modeling of a hypothetical scenario in which all sediment discharges from
construction activity are prevented from entering surface waters. Under this hypothetical scenario, loading
from in-scope construction sites (one acre and greater in size) decrease by 2.58 million tons per year,
relative to baseline conditions. It is important to note that a certain level of sediment is present in surface
waters under natural, undisturbed conditions and that some sediment is necessary for the natural
biological function of surface waters. Current levels of sediment discharge to the U.S.  surface water
network are much higher, however, than would be observed under natural, undisturbed conditions  due to
sediment discharges from construction, agriculture, and eroding streambeds associated with land use
change and other human activity (see Section 2.1). Therefore, while construction sediment discharges
represent approximately 0.15 percent of total sediment in surface waters, they represent a greater
percentage of total sediment pollution in the nation's waters. Given the relatively small percentage of U.S.
land area dedicated to construction activity on an annual basis (-0.04 percent of coterminous United
States), construction sites have a disproportionately high rate of sediment contribution to surface waters
relative to other land uses.
Predicted changes, relative to baseline conditions, in TSS concentration for a hypothetical scenario with
zero discharge from construction sites are presented in Table 6-6 through Table 6-9. The tables show the
distribution of TSS concentrations and reach miles improved. Average TSS concentrations decrease from
956.4 mg/L under baseline conditions to 954.1 mg/L under the zero discharge scenario. This represents a
reduction of 2.4 mg/L or 0.3 percent. Median concentrations decrease from 289.3 to 287.5 mg/L,
representing a change of 1.8 mg/L or 0.6  percent. In examining the spatial distribution of concentrations
in the baseline and zero discharge scenarios, shown in Table 6-7 and Table 6-8, respectively, the highest
concentrations of TSS are seen in Regions 6, 7, 8, and 9, and this is also where the largest reductions are
predicted. Table 6-9 shows the distribution and magnitude of reductions in improved reach miles. Over 77
percent of improved reach miles are  expected to have a decrease in TSS concentration that is greater than
zero but less than 1 mg/L. Approximately 18 percent of improved reach miles will decrease in TSS
concentration by 1 to 5 mg/L. Greater than  5 mg/L reductions are estimated in 4.2 percent of improved
reach miles. These improvements are largely driven by Regions 4, 6, 7 and 10. Eliminating sediment
discharge from construction sites also reduces sediment accumulation in reservoirs by 3.7 million pounds
per year.

Table 6-6: Distribution of TSS Concentrations based on SPARROW Output for 31,927 RF1
Reaches Receiving Construction Sediment Discharges
Average Reduction in

Scenario
Baseline
Reach
Count
31,927
Reach
Miles
412,062
Distribution of TSS Concentrations by
TSS Average TSS
(mg/L)u
956.4
(mg/L)
-
5th
29.7
Percentile1'2 (mg/L)
25th
109.9
50th
289.3
75th
880.0
95th
6,157.8
Hypothetical
No Discharge
Scenario       31,927    412,062     954.1	2.4	28.6     108.7     287.5    877.9   6,156.8
'Reach-length weighted.
2 Any TSS concentration above the 95th percentile is assumed to be an outlier and was replaced with the 95th percentile value.
November 2009                                                                                 6-26

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-7: Summary of Baseline TSS Concentration in RF1
Construction Sediment Discharges, by EPA Region
EPA Reach
Region Count
1 1,213
2 1,111
3 2,012
4 7,564
5 3,918
6 4,634
7 3,063
8 4,300
9 1,439
10 2,673
Nation 31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TSS
(mg/L)u
77.5
110.2
180.6
236.8
296.0
1,567.7
1,800.8
2,024.6
1,486.7
239.6
956.4
Reaches Receiving
Distribution of TSS Concentrations by Percentile1'2 (mg/L)
5th 25th 50th 75th 95th
9.8
10.5
44.5
33.2
20.6
82.9
177.6
79.2
31.6
20.6
29.7
25.2
38.8
83.6
91.4
60.9
259.5
484.7
359.7
157.6
52.6
109.9
41.2
79.0
141.7
180.6
196.1
641.6
1,050.3
1,131.2
434.1
108.2
289.3
57.2
135.7
228.3
311.1
430.4
2,137.4
2,008.6
3,104.9
1,932.3
233.1
880.0
118.7
283.1
447.4
596.8
880.4
6,157.8
6,157.8
6,157.8
6,157.8
882.4
6,157.8
'Reach-length weighted.
2Any TSS concentration above the 95th percentile is assumed to be an outlier and was replaced with the 95th percentile value.
 Table 6-8: Summary of TSS Concentrations Under the Hypothetical No Discharge
 Scenario, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TSS
(mg/L)1'2
77.3
110.2
180.4
235.3
295.8
1,562.0
1,797.9
2,017.5
1,480.9
238.4
954.1
Distribution of TSS Concentrations by Percentile1'2 (mg/L)
5th 25th 50th 75th 95th
9.7
10.5
44.4
32.2
20.5
75.1
173.9
75.4
29.8
19.4
28.6
25.2
38.8
83.6
90.4
60.9
254.0
481.7
354.5
152.6
51.4
108.7
41.1
79.0
141.5
178.8
195.8
634.6
1,048.3
1,126.6
430.2
107.5
287.5
57.1
135.5
228.2
310.1
430.4
2,123.9
2,007.3
3,104.2
1,900.9
233.1
877.9
118.6
281.7
447.3
590.9
880.3
6,156.8
6,156.8
6,156.8
6,156.8
882.4
6,156.8
 Reach-length weighted.
 2Any TSS concentration above the 95th percentile is assumed to be an outlier and was replaced with the 95th percentile value.
November 2009
                                                                                                    6-27

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-9: Water Quality Impacts: Improvements in TSS Concentrations Under the
Hypothetical No Discharge Scenario, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Total
Improved
Reach
Miles1
10,740
12,560
25,471
83,028
56,906
70,630
46,859
57,322
22,204
25,920
411,640
Range of Reductions in TSS Concentration (mg/L)
0 < A TSS < 1
Reach % of
Miles Total
10,421 97.0%
12,488 99.4%
24,256 95.2%
64,885 78.1%
54,801 96.3%
33,122 46.9%
31,518 67.3%
47,070 82.1%
17,373 78.2%
23,283 89.8%
319,218 77.5%
1 < A TSS < 5
Reach % of
Miles Total
219 2.0%
72 0.6%
1,073 4.2%
14,320 17.2%
1,974 3.5%
27,840 39.4%
13,896 29.7%
9,967 17.4%
4,319 19.4%
1,527 5.9%
75,206 18.3%
5 < A TSS < 10
Reach % of
Miles Total
64 0.6%
0 0.0%
112 0.4%
2,236 2.7%
97 0.2%
4,506 6.4%
816 1.7%
227 0.4%
283 1.3%
559 2.2%
8,900 2.2%
10 < A TSS < 50
Reach % of
Miles Total
36 0.3%
0 0.0%
30 0.1%
1,401 1.7%
34 0.1%
4,501 6.4%
584 1.2%
49 0.1%
200 0.9%
478 1.8%
7,314 1.8%
50 < A TSS
Reach % of
Miles Total
0 0.0%
0 0.0%
0 0.0%
186 0.2%
0 0.0%
661 0.9%
44 0.1%
10 0.0%
29 0.1%
72 0.3%
1,002 0.2%
1 Does not include 40 freshwater reaches (422 reach miles) for which SPARROW does not predict a concentration.

Table 6-10 through Table 6-15 present the reductions in TN and TP concentrations associated with
reduced sediment discharge under the hypothetical zero discharge scenario. As described in Section 6.4,
reductions in TN and TP concentrations are calculated based on the predicted reduction of in-stream
sediment concentrations. The distribution of improved reach miles for TN and TP is therefore the same as
for TSS, which is presented above in Table 6-9.
Reductions in average TN are predicted to be 0.02 mg/L or 0.1 percent. Reductions in the median
concentrations are 0.01 mg/L or 0.1 percent and 95th percentile concentrations with 0.04 mg/L or 0.3
percent reductions.
Reductions in average TP concentrations are predicted to be 6.0 ug/L or 0.2 percent. Reductions in the
median concentrations are 2.2 ug/L or 0.1 percent and 95th percentile concentrations with 22.3 ug/L or
0.6 percent reductions.
The geographic distribution of TN and TP concentrations is similar to TSS concentrations, with the
highest concentrations seen in Regions 6, 7, 8,  and 9. In general, these regions are also predicted to see
the largest reductions in concentrations.
November 2009
                                                                                             6-28

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-10: Distribution of TN Concentrations based on SPARROW Output for 31,927 RF1
Reaches Receiving Construction Sediment Discharges
Reach Reach Average TN
Scenario Count Miles (mg/L)u
Baseline 31,927 412,062 18.22
Reduction Distribution of TN Concentrations by
in Average Percentile (mg/L) u
TN(mg/L) 5th 25th 50th 75th 95th
0.37 0.91 1.65 3.88 14.95
Hypothetical
No Discharge
Scenario       31,927   412,062     18.21	0.02      0.36     0.90      1.64      3.86     14.91
 Reach-length weighted.
 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was replaced
with the 95th percentile value.

Table 6-11: Summary of Baseline TN Concentration in RF1 Reaches Receiving
Construction Sediment Discharges, by EPA Region
EPA Region ^each
^ Count
1 1,213
2 1,111
3 2,012
4 7,564
5 3,918
6 4,634
7 3,063
8 4,300
9 1,439
10 2,673
Nation 31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average TN Distribution of TN Concentrations by Percentile
(mg/L)1 5th 25th 50th 75th
1.32
19.43
1.88
4.39
7.76
26.15
22.87
40.01
46.56
5.30
18.22
0.24
0.33
0.61
0.29
0.51
0.45
1.10
0.50
0.24
0.23
0.37
0.59
0.95
1.02
0.70
1.49
1.07
2.64
1.08
0.55
0.49
0.91
0.85
1.37
1.35
1.06
3.68
1.83
5.11
2.44
1.28
0.83
1.65
1.27
1.98
1.95
1.59
7.50
3.26
9.56
5.99
3.35
1.58
3.88
(mg/L)1
95th
2.57
3.85
4.35
3.06
15.86
12.38
21.50
28.66
54.43
5.29
14.95
'Reach-length weighted.

Table 6-12: Summary of TN
Scenario, by EPA Region
„_. „ . Reach
EPA Region
6 Count
1 1,213
2 1,111
3 2,012
4 7,564
5 3,918
6 4,634
7 3,063
8 4,300
9 1,439
10 2,673
Nation 31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Concentrations Under
Average
(mg/L)
1.32
19.43
1.88
4.38
7.75
26.11
22.85
39.99
46.51
5.29
18.21
the Hypothetical
No Discharge
TN Distribution of TN Concentrations by Percentile
u 5th 25th 50th 75th
0.23
0.33
0.61
0.28
0.51
0.42
1.08
0.48
0.23
0.23
0.36
0.59
0.95
1.02
0.70
1.48
1.06
2.61
1.07
0.54
0.49
0.90
'Reach-length weighted.
2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile
with the 95th percentile value.
0.85
1.37
1.35
1.06
3.68
1.82
5.10
2.42
1.28
0.83
1.64
1.27
1.98
1.95
1.58
7.49
3.23
9.49
5.96
3.26
1.58
3.86

(mg/L)u
95th
2.57
3.85
4.34
3.06
15.86
12.33
21.50
28.64
54.43
5.29
14.91
is assumed to be an outlier and was replaced
November 2009                                                                                  6-29

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-13: Distribution of TP Concentrations based on SPARROW Output for 31,927 RF1
Reaches Receiving Construction Sediment Discharges


Scenario
Baseline

Reach
Count
31,927

Reach
Miles
412,062
Average
TP
(Hg/L)1'2
3,921.7
Reduction
in Average
TP (jig/L)
-
Distribution of TP Concentrations by
Percentile (jig/L) u
5th
30.5
25th
104.4
50th
253.2
75th
685.3
95th
2,907.3
Hypothetical
No Discharge
Scenario       31,927   412,062    3,915.6	6.0	29.5     103.3    251.0     683.2    2,885.0
 Reach-length weighted.
2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was replaced
with the 95th percentile value.

Table 6-14: Summary of Baseline TP Concentration
Construction Sediment Discharges, by EPA Region
EPA Region
1
2
o
J
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TP (jig/L)
109.5
1,697.5
209.3
885.6
651.3
8,110.6
3,380.2
7,282.3
13,544.9
965.7
3,921.7
Distribution
1 5th
9.9
16.6
37.7
30.0
18.6
63.4
140.6
69.0
37.0
22.1
30.5
in RF1 Reaches Receiving
of TP
25th
29.6
47.2
80.2
81.8
75.6
208.4
377.8
249.5
114.9
56.3
104.4
Concentrations
50th
49.2
91.2
135.4
139.5
225.3
519.5
603.9
659.0
489.0
113.4
253.2
by Percentile (jig/L) 1
75th 95th
86.5
152.1
233.5
241.9
475.9
1,216.9
1,012.7
1,449.5
1,652.3
303.0
685.3
233.8
403.0
633.6
547.4
1,246.4
5,557.8
2,853.1
5,803.0
30,733.9
1,318.7
2,907.3
Reach-length weighted.

Table 6-15: Summary of TP
Scenario, by EPA Region
EPA
Region
1
2
o
5
4
5
6
7
8
9
10
National
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Concentrations Under the Hypothetical No
Average Distribution
TP (jig/L)1'2 5th
109.3
1,697.5
209.1
884.4
651.1
8,094.3
3,376.8
7,277.0
13,512.6
964.6
3,915.6
'Reach-length weighted.
2 Nutrient reductions are based on TSS reductions. Any TSS
with the 95th percentile value.
9.9
16.6
37.7
29.2
18.5
57.6
138.4
66.6
35.3
21.0
29.5
Discharge
of TP Concentrations by Percentile
25th 50th 75th
29.4
47.2
80.0
81.1
75.5
202.5
374.8
246.4
111.2
55.4
103.3
49.1
91.1
135.1
138.5
225.3
516.2
601.9
656.6
474.1
112.2
251.0
86.5
151.9
233.2
239.9
475.8
1,207.3
1,008.8
1,445.2
1,644.4
297.9
683.2

(Hg/L)1'2
95th
233.7
402.9
633.4
547.3
1,246.4
5,557.7
2,850.7
5,780.1
30,733.9
1,267.0
2,885.0
concentration above the 95th percentile is assumed to be an outlier and was replaced
November 2009                                                                                   6-30

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.6.2  Estimated Changes in TSS Concentrations in RF1 Watersheds with Most
       Developed Acres (1992-2001) Receiving Sediment Loading from Construction
       Sites

Table 6-16 summarizes the characteristics of different groups of modeled watersheds in terms of the
relative amount of development they contain. The table provides information for three particular
subgroups that represent the "top" watersheds in terms of the amount of developed land during the period
of 1992-2001: top 1 percent, top 10 percent, and top 25 percent. As discussed above, development is not
evenly distributed during the nine-year period under examination and each of the watershed subsets
therefore account for a relatively disproportionate share of construction acres. For example, the top 10
percent of watersheds account for over 58 percent of the increase in developed land in the coterminous
United States, while the top 25 percent account for 76 percent of all developed land. These watersheds are
also distributed unevenly across EPA regions both in terms of total reach miles and the number of
waterbody reaches, as shown in Table 6-17. For example, approximately 22 percent of reach miles within
the top 1  percent are found in Region 5 even though the region accounts for approximately 14 percent of
overall RF1 reach miles.
These differences are reflected in TSS concentration estimates shown in Table 6-16 through Table 6-20
for each of the three top groups, respectively, and across the four regulatory options (see Chapter 1 for
option descriptions). As shown in the tables, the top 1 percent of watersheds in terms of construction
acres tend to have the highest sediment concentrations, both under current conditions and under the four
options. While the three top groups result in a similar fraction of miles showing improvements in TSS
concentrations across the three groups, reaches within the top 1 percent of watersheds exhibit a higher
magnitude of TSS concentration reductions than reaches in watersheds that contain comparatively fewer
construction acres.
November 2009                                                                               6-31

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-16: Reaches with
Percent of RF1 Watersheds1
Top l%ofRFl Watersheds
Top 10% of RF1 Watersheds
Top 25% of RF1 Watersheds
All RF1 Watersheds in
Coterminous U.S.
(100% of Reaches)
Largest Increase
Number of RFl
Watersheds2
319
3,192
7,981
31,927
in Developed Land Area (1992-2001)
% of Total
Increase in
Coterminous
U.S. Developed
Land
21.7%
58.3%
76.0%
93.5%
Total Increase in
Developed Land
Area (1992-2001)
(million acres)
1.1
3.1
4.0
5.0
Average Increase
in Developed
Land Area per
RFl Watersheds
(1992-2001)
(acres)
3,596.4
968.0
504.4
155.1
Total Land Area
(million acres)
52.6
305.2
572.9
1,291.9
%of
Coterminous
U.S. Land Area
2.8%
16.1%
30.3%
68.2%
 Top 1% of RFl watersheds had 1,841 acres or greater of developed land area (1992-2001); Top 10% of watersheds had 321 acres or greater of developed land area (1992-2001); Top 25% of
watersheds had 124 acres or greater of developed land area (1992-2001).

 RF1 watersheds included in count are those that receive construction loadings and for which both NLCD and SPARROW data are available.
Table 6-17: Distribution of RF1 Watersheds that Receive Direct Loadings by Increase in Developed Land Area (1992-2001),
by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Top 1%
Reach Miles
21
50
410
1,472
1,590
1,540
736
450
947
172
7,387
of Watersheds
% of Miles
0%
1%
6%
20%
22%
21%
10%
6%
13%
2%
100%
Top 10%
Reach Miles
911
1,502
4,582
14,276
9,733
15,428
7,042
4,758
3,915
2,139
64,284
of Watersheds
% of Miles
1%
2%
7%
22%
15%
24%
11%
7%
6%
3%
100%
Top 25%
Reach Miles
2,088
3,646
9,774
33,331
19,670
31,300
16,350
13,362
6,827
5,239
141,587
of Watersheds
% of Miles
1%
3%
7%
24%
14%
22%
12%
9%
5%
4%
100%
Coterminous U.S.
(100% of Reaches)
Reach Miles % of Miles
10,740 3%
12,560 3%
25,471 6%
83,058 20%
56,909 14%
70,687 17%
46,903 11%
57,513 14%
22,280 5%
25,940 6%
412,062 100%
November 2009
                                                                                                                                             6-32

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-18: Estimated Changes in RF1 Reach TSS Concentration by Policy Option: Top 1% of RF1 Watersheds Receiving
Direct Construction Loadings by Increase in Developed Land Area (1992-2001)
Scenario
Baseline
Option 1
Option 2
Option 3
Option 4
Median
TSS
(mg/L)1
425.6
425.2
424.9
424.7
424.8
Average
TSS
(mg/L)1
1,073.3
1,069.2
1,067.0
1,065.9
1,066.5
Total Improved
%of
RF1 Total
Miles Miles2
-
6,932 93.8%
6,932 93.8%
7,372 99.8%
6,932 93.8%
Range of Reductions in TSS Concentrations (mg/L)
0 < A TSS < 1
% of Total
Improved
RF1 Miles Miles
-
4,891 70.6%
4,242 61.2%
4,457 60.5%
4,184 60.4%
1 < A TSS < 5
% of Total
RF1 Improved
Miles Miles
-
1,288 18.6%
1,509 21.8%
1,418 19.2%
1,359 19.6%
5 < A TSS < 10
% of Total
RF1 Improved
Miles Miles
-
354 5.1%
416 6.0%
618 8.4%
511 7.4%
10 < A TSS < 50
% of Total
RF1 Improved
Miles Miles
-
381 5.5%
683 9.9%
748 10.1%
776 11.2%
50 < A TSS
% of Total
RF1 Improved
Miles Miles
-
19 0.3%
82 1.2%
131 1.8%
103 1.5%
1 River-mile weighted.
2 Total miles (7,3 87) are based on top 1 % of reaches by area of urban development from 1992 to 2001.
Table 6-19: Estimated Changes in RF1 Reach TSS Concentration by Policy Option: Top 10% of RF1 Watersheds Receiving
Direct Construction Loadings by Increase in Developed Land Area (1992-2001)
Scenario
Baseline
Option 1
Option 2
Option 3
Option 4
Median
TSS
(mg/L)1
289.1
287.5
287.0
286.0
286.8
Average
TSS
(mg/L)1
947.7
945.0
943.6
943.9
943.3
Total Improved
%of
RF1 Total
Miles Miles2
-
60,992 94.9%
60,992 94.9%
64,202 99.9%
60,992 94.9%
Range of Reductions in TSS Concentrations (mg/L)
0 < A TSS < 1
% of Total
RF1 Improved
Miles Miles
-
48,452 79.4%
41,568 68.2%
42,342 66.0%
40,599 66. 6%
1 < A TSS < 5
% of Total
RF1 Improved
Miles Miles
-
9,547 15.7%
13,413 22.0%
14,308 22.3%
13,551 22.2%
5 < A TSS < 10
% of Total
RF1 Improved
Miles Miles
-
1,708 2.8%
2,970 4.9%
3,643 5.7%
3,355 5.5%
10 < A TSS < 50
% of Total
RF1 Improved
Miles Miles
-
1,067 1.8%
2,674 4.4%
3,419 5.3%
3,044 5.0%
50 < A TSS
% of Total
RF1 Improved
Miles Miles
-
218 0.4%
367 0.6%
490 0.8%
443 0.7%
1 River-mile weighted.
2 Total miles (64,284) are based on top 10% of reaches by area of urban development from 1992 to 2001.
November 2009
                                                                                                                       6-33

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-20: Estimated Changes in RF1  Reach TSS Concentration by Policy Option: Top 25% of RF1 Watersheds by Increase
in Developed Land Area (1992-2001)
Scenario
Baseline
Option 1
Option 2
Option 3
Option 4
Median
TSS
(mg/L)1
283.5
282.6
281.9
281.5
281.5
Average
TSS
(mg/L)1
929.9
928.2
927.3
926.8
927.1
Total Improved
%of
RF1 Total
Miles Miles2
-
134,556 95.0%
134,556 95.0%
141,498 99.9%
134,556 95.0%
Range of Reductions in TSS Concentrations (mg/L)
0 < A TSS < 1
% of Total
RF1 Improved
Miles Miles
-
114,258 84.9%
100,071 74.4%
101,252 71.6%
97,345 72.3%
1 < A TSS < 5
% of Total
RF1 Improved
Miles Miles
-
16,259 12.1%
25,718 19.1%
29,060 20.5%
27,184 20.2%
5 < A TSS < 10
% of Total
RF1 Improved
Miles Miles
-
2,336 1.7%
4,638 3.5%
5,898 4.2%
5,288 3.9%
10 < A TSS < 50
% of Total
RF1 Improved
Miles Miles
-
1,406 1.1%
3,635 2.7%
4,630 3.3%
4,132 3.1%
50 < A TSS
% of Total
RF1 Improved
Miles Miles
-
297 0.2%
493 0.4%
659 0.5%
608 0.5%
1 River-mile weighted.
2 Total miles (141,489) are based on top 25% of reaches by area of urban development from 1992 to 2001.
November 2009
                                                                                                                           6-34

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.6.3  Estimated Changes in TSS, TN, and TP Concentrations for all RF1 Watersheds
       Receiving Loading from Construction Sites

Table 6-21 summarizes national average reach-length weighted TSS concentrations in RF1 reaches as
estimated by SPARROW under both the baseline scenario representing current conditions and the four
regulatory options (see Chapter 1 for option descriptions). Information is presented for all reaches in the
RF1 network that receive sediment loading from construction sites.
Table 6-21 also presents information on the distribution of TSS concentrations under the baseline
scenario representing current conditions and the four regulatory options. TSS concentrations in the RF1
network vary over a wide range. Values for the 5th and 95th percentile bounds under baseline conditions
are 29.7 and 6,157.8 mg/L respectively. Median (50th percentile) TSS concentrations are approximately
289.3 mg/L under the baseline scenario. Under Option 1, the median concentration decreases 0.8 mg/L to
288.6 mg/L. Median concentration under Option 2 decreases 1.4 mg/L to 287.9 mg/L; under Option 3, it
decreases 1.6 mg/L to 287.7 mg/L; under Option 4 it decreases 1.5 mg/L to 287.8 mg/L. These numbers
represent the national average concentration reduction and do not provide information on the geospatial
variability in TSS concentration reduction, which is high.

Table 6-21: Distribution of TSS Concentration based on SPARROW Output for 31,927 RF1
Reaches Receiving Construction Sediment Discharges
Policy Reach Reach
Option Count Miles
Baseline 31,927 412,062
Option 1 31,927 412,062
Option 2 31,927 412,062
Options 31,927 412,062
Option 4 31,927 412,062
'Reach-length weighted.
2 Any TSS concentration above the
Percent of Reduction
total Average in
construction TSS Average
acres, '92-'01 (mg/L)1'2 TSS
93.05%
93.05%
93.05%
93.05%
93.05%
95th percentile
956.4
955.0
954.5
954.2
954.4
—
1.5
2.0
2.2
2.1
Distribution of TSS Concentrations, by
percentile (mg/L)u
5th 25th 50th 75th 95th
29.7
29.1
28.8
28.7
28.7
109.9
109.2
108.9
108.8
108.8
289.3
288.6
287.9
287.7
287.8
880.0
878.5
878.3
878.0
878.3
6,157.8
6,157.8
6,157.8
6,156.8
6,157.8
is assumed to be an outlier and was replaced with the 95th percentile value.
Table 6-22 through Table 6-26 detail TSS concentrations by EPA region under baseline conditions and
for each of the four regulatory options considered. The highest average TSS concentrations are found in
Regions 6, 7, 8, and 9 which have average reach-length weighted TSS concentrations greater than
1,400 mg/L under all scenarios. Median TSS concentrations in these regions—ranging from 434.1 to
1,131.2 mg/L under baseline conditions—are also higher than median concentrations in Regions 1, 2, 3,
4, 5, and 10 where the highest median TSS concentration is 196.1 mg/L. Under Option 1, TSS
concentrations decline a small amount (by less than 1 percent) from baseline concentrations. TSS
concentrations are further reduced under Option 2 and again under Option 3. Under Option 4,
concentrations are reduced relative to baseline and concentrations generally fall between those under
Option 2 and Option 3. The distribution of concentrations among regions is consistent with the baseline
under each of the options.
November 2009
                                                                                          6-35

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-22: Summary of Baseline TSS Concentration in RF1 Reaches Receiving
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TSS
(mg/L)1'2
77.5
110.2
180.6
236.8
296.0
1,567.7
1,800.8
2,024.6
1,486.7
239.6
956.4
Distribution of TSS
5th 25th
9.8
10.5
44.5
33.2
20.6
82.9
177.6
79.2
31.6
20.6
29.7
25.2
38.8
83.6
91.4
60.9
259.5
484.7
359.7
157.6
52.6
109.9
Concentrations,
50th
41.2
79.0
141.7
180.6
196.1
641.6
1,050.3
1,131.2
434.1
108.2
289.3
by percentile
75th
57.2
135.7
228.3
311.1
430.4
2,137.4
2,008.6
3,104.9
1,932.3
233.2
880.0
(mg/L)1'2
95th
118.7
283.1
447.4
596.8
880.4
6,157.8
6,157.8
6,157.8
6,157.8
882.4
6,157.8
'Reach-length weighted.
2Any TSS concentration above the 95th percentile is assumed to be an outlier and was replaced with the 95th percentile value.

Table 6-23: Summary of Option 1 TSS Concentration in RF1 Reaches
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
1
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TSS
(mg/L)1'2
77.5
110.2
180.5
236.2
295.9
1,564.7
1,798.8
2,017.8
1,481.5
239.1
955.0
Distribution of TSS
5th 25th
9.8
10.5
44.5
32.9
20.6
78.8
175.9
75.4
29.8
20.3
29.1
25.2
38.8
83.6
90.9
60.9
256.8
482.6
354.5
153.0
52.1
109.2
Concentrations,
50th
41.2
79.0
141.5
179.9
196.0
639.6
1,048.4
1,126.6
430.7
107.7
288.6
Receiving
by percentile
75th
57.1
135.6
228.3
311.0
430.4
2,130.3
2,007.4
3,104.2
1,902.6
233.2
878.5

(mg/L)1-2
95th
118.7
282.7
447.4
593.7
880.4
6,157.8
6,157.8
6,157.8
6,157.8
882.4
6,157.8
'Reach-length weighted.
2Any TSS concentration above the 95th percentile is assumed to be an outlier and was replaced with the 95th percentile value.
November 2009
                                                                                                 6-36

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table 6-24: Summary of Option 2 TSS Concentration in
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
1
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TSS
(mg/L)u
77.4
110.2
180.4
235.7
295.9
1,563.2
1,798.4
2,017.7
1,481.2
238.8
954.5
Distribution
5th
9
10
44
32
20
76
175
75
29
20
.7
.5
.5
o
.J
.6
.2
.8
.4
.8
.0
28.8
RF1 Reaches
of TSS
25th
25.
38.
83.
90.
60.
254.
482.
354.
152.
51.
108.
2
8
6
7
9
4
6
5
6
6
9
Concentrations,
50th
41.1
79.0
141.5
179.0
195.9
637.3
1,048.4
1,126.6
430.3
107.5
287.9
Receiving
by percentile
75th
57.1
135.5
228.3
310.7
430.4
2,128.6
2,007.3
3,104.2
1,901.1
233.2
878.3

(mg/L)
95th
118.
282.
447.
592.
880.
6,157.
6,157.
6,157.
6,157.
882.
6,157.

1,2
7
2
3
9
o
5
8
8
8
8
4
8
Reach-length weighted.
2 Any TSS concentration above the 95th percentile is assumed to be an outlier and was replaced with the 95th percentile value.

Table 6-25: Summary of Option 3 TSS Concentration in
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TSS
(mg/L)1'2
77.3
110.2
180.4
235.5
295.8
1,562.5
1,798.1
2,017.6
1,481.0
238.6
954.2
Distribution
5th
9.7
10.5
44.4
32.3
20.6
75
.4
175.8
75.4
29.8
20.0
28.7
'Reach-length weighted.
Any TSS concentration above the 95th percentile is assumed to be an outlier and
RF1 Reaches
of TSS
25th
25.
38.
83.
90.
60.
254.
481.
354.
152.
51.
108.
2
8
6
5
9
2
7
5
6
4
8
Concentrations,
50th
41.1
79.0
141.5
178.9
195.9
637.0
1,048.4
1,126.6
430.3
107.5
287.7
Receiving
by percentile
75th
57.1
135.5
228.2
310.2
430.4
2,124.3
2,007.3
3,104.2
1,901.0
233.2
878.0

(mg/L)
95th
118.
281.
447.
592.
880.
6,156.
6,156.
6,156.
6,156.
882.
6,156.

u
6
9
3
2
2
8
8
8
8
4
8
was replaced with the 95th percentile value.
November 2009
                                                                                                 6-37

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 6-26: Summary of Option 4 TSS Concentration in RF1 Reaches Receiving
 Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
„ , Average
Reach TSS
Miles 12
(mg/L)1'2
10,740
12,560
25,471
83,058
56,909
70,687 1,
46,903 1,
57,513 2,
22,280 1,
25,940
412,062
'Reach-length weighted.
2 Any TSS concentration above the 95th percentile
77.4
110.2
180.4
235.6
295.8
562.8
798.3
017.7
481.2
238.9
954.4
Distribution of TSS Concentrations,
5th 25th 50th
9.7
10.5
44.5
32.3
20.6
75.6
175.8
75.4
29.8
20.0
28.7
25.2
38.8
83.6
90.7
60.9
254.2
481.8
354.5
152.6
51.8
108.8
41.1
79.0
141.5
179.0
195.9
637.3
1,048.4
1,126.6
430.3
107.5
287.8
by percentile (mg/L)1'2
75th 95th
57.1
135.5
228.3
310.3
430.4
2,127.0
2,007.3
3,104.2
1,901.1
233.2
878.3
118.7
282.2
447.3
592.5
880.3
6,157.8
6,157.8
6,157.8
6,157.8
882.4
6,157.8
is assumed to be an outlier and was replaced with the 95th percentile value.
The following tables describe the number of RF1 reach miles improved and the distribution of TSS
concentration reductions for each of the four options analyzed for this regulation. As shown in Table
6-27, which summarizes the effects of sediment reduction at the national level, Options 1, 2 and 3
improve 94.7 percent of modeled RF1 reach miles receiving construction sediment loading. Options 4 is
predicted to lead to improvements in 99.9 percent of RF1 reach miles that receive direct construction
loadings. Under Option 1, approximately 93 percent of the improved reach miles show reductions in TSS
concentrations of greater than zero but less than 1 mg/L. Under Options 2 through 4, approximately 84 to
86 percent of the improved reach miles are predicted to see a decrease of less than 1 mg/L. While under
Option 1, only about 6 percent of improved reach miles are expected see a reduction of 1 to 5 mg/L, under
Options 2 through 4, 11 to 12 percent of improved reach miles are expected to improve by 1 to 5 mg/L.
Reductions greater than 50 mg/L are expected in less than or equal 0.2 percent of improved miles under
all options.
As discussed at the start of Section 6.6, the reduction in loadings and concentrations in reaches that
receive construction sediment discharge directly also translate into decreased concentrations in
downstream reaches. The total number of reach miles that are improved under the four options analyzed
is therefore greater than presented in the tables within this chapter. When considering all RF1 reaches,
including those that do not receive construction sediment discharge directly, the  results show a total of
431,074 reach miles improving under Options 1, 2, and 4. These miles account for 68.7 percent of the
total RF1 reach network. Option 3 has an even broader effect, reducing TSS concentrations in a total of
472,402 reach miles (75.3 percent of the RF1 reach network).
Table 6-28 summarizes Option 1's sediment reductions at the regional level. Nationwide, 93 percent of
improved RF1 reach miles under Option 1 experience TSS concentration reductions less than 1 mg/L.
While all but Region 2 will experience reduction of 1 to 5 mg/L, the number of regions experiencing
reductions decreases at higher reduction ranges. All but Regions 2, 3, and 5 will  experience reductions
between 5 and 10 mg/L, encompassing approximately 0.8 percent of all improved reaches nationally. All
but Regions 1, 2, 3, and 5 are predicted to see some reductions between 10 and 50 mg/L encompassing
0.5 percent of all improved reaches nationally. Only four regions (4, 6, 7, and 10) will experience
November 2009                                                                               6-38

-------
Environmental Impact and Benefits Assessment for the C&D Category
reductions in TSS concentrations greater than 50 mg/L, with a national total of 0.1 percent of improved
reaches.
As shown in Table 6-29, Option 2 is predicted to result in the same number of total improved reach miles
(390,374 miles), but the magnitude of reductions in TSS concentrations in these miles is expected to be
greater. All regions are predicted to see a reduction in TSS concentrations of up to 5 mg/L in some reach
miles. Approximately 86 and 11 percent of all reach miles improved will experience up to 1 and 5 mg/L
reductions, respectively. Some reach miles in all but Region 2 are predicted to improve by 5 to 10 mg/L
with approximately 1.6 percent of reach miles improved across the  nation in this range. Approximately
1.2 percent of all reach miles improved will see  a decrease of 10 to  50 mg/L. Regions 2,  3, and 5 are the
only regions without reductions in this range. Nationally, 0.2 percent of all improved reach miles will
experience a reduction of greater than 50 mg/L under Option 2. Among regions, Region  6 is predicted to
have the greatest number of reach miles with reductions greater than 50 mg/L, followed by Regions 4, 10,
7, and 8.
Option 3  (Table 6-30) is estimated to reduce TSS concentrations by more than Option 2 with an
additional 21,266 miles of improved rivers and generally a greater magnitude of reductions across the
reduction ranges and across regions. Reductions of up to  1 mg/L and 5 mg/L are expected for all regions
with a national total of 84 and 12 percent of improved reach miles,  respectively. Some reach miles are
expected to improve by 5 to 10 mg/L in all regions except Region 2, encompassing 2.0 percent of all
improved reach miles nationally. All regions but Regions 2, 3, and  5 are expected to experience
improvement in some reach miles by 10 to 50 mg/L including 1.5 percent of all improved reach miles
nationally. Greater than 50 mg/L reductions in TSS concentrations  are expected in 0.2 percent of all
improved reach miles nationally with improvements in Regions 4 and 7 through 10.
Option 4  (Table 6-31) is estimated to reduce TSS concentrations in the same number of reach miles as
Options 1 and 2, but the reductions are of slightly greater magnitude. Therefore, under Option 4, there is a
slightly smaller percentage of total miles improved by up to 1 mg/L and a larger percentage for all other
ranges. The spatial distribution of reductions, however, is the same  as Option 3, with respect to the
regions where reductions are  predicted in each of the reduction ranges. Approximately 11.9 and
1.8 percent of improved reach miles are expected to see 1 to 5 mg/L and 5 to 10 mg/L reductions,
respectively. Reductions of 10 mg/L to  50 mg/L are expected in 5,414 miles or 1.4 percent of all
improved reach miles, and 0.2 percent of improved reach miles are  expected to improve  by greater than
50 mg/L under Option 4.
November 2009                                                                                6-39

-------
 Environmental Impact and Benefits Assessment for the C&D Category
Table 6-27: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS Concentrations
Policy
Option
Option 1
Option 2
Option 3
Option 4
Percent of
Total
Reach
Total Miles Miles with
Total RF1 with TSS TSS
Miles Reductions1 Reductions1
412,062 390,374 94.7%
412,062 390,374 94.7%
412,062 411,640 99.9%
412,062 390,374 94.7%
Range of Reductions in TSS Concentrations (mg/L)
0 < A TSS < 1
%of
Reach Total
Miles Improved
Miles
361,569 92.6%
336,168 86.1%
346,379 84.2%
330,850 84.8%
1 < A TSS < 5
%of
Reach Total
Miles Improved
Miles
23,624 6.1%
42,481 10.9%
50,221 12.2%
46,339 11.9%
5 < A TSS < 10
%of
Reach Total
Miles Improved
Miles
2,994 0.8%
6,382 1.6%
8,084 2.0%
7,029 1.8%
10 < A TSS < 50
%of
Reach Total
Miles Improved
Miles
1,832 0.5%
4,738 1.2%
6,147 1.5%
5,414 1.4%
50 < A TSS
%of
Reach Total
Miles Improved
Miles
354 0.1%
605 0.2%
808 0.2%
742 0.2%
 Not included in the total are 40 reaches (422 reach miles) for which SPARROW does not predict concentrations. The 99.9% fraction represents all reach miles over which SPARROW calculates TSS
concentrations.
 Table 6-28: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS Concentrations from
 Option 1, by EPA Region	
Total
Improved
EPA Reach
Region Miles1
1 10,740
2 12,554
3 25,471
4 82,955
5 56,906
6 64,954
7 41,102
8 49,967
9 19,853
10 25,873
Nation 390,374
Range of Reductions in TSS Concentration (mg/L)
0 < A TSS < 1
Reach % of
Miles Total
10,610 98.8%
12,554 100.0%
25,164 98.8%
76,258 91.9%
56,474 99.2%
50,162 77.2%
37,682 91.7%
49,350 98.8%
18,943 95.4%
24,374 94.2%
361,569 92.6%
1 < A TSS < 5
Reach % of
Miles Total
110 1.0%
0 0.0%
307 1.2%
5,657 6.8%
432 0.8%
11,569 17.8%
3,040 7.4%
570 1.1%
778 3.9%
1,161 4.5%
23,624 6.1%
5 < A TSS < 10
Reach % of
Miles Total
20 0.2%
0 0.0%
0 0.0%
633 0.8%
0 0.0%
1,751 2.7%
257 0.6%
25 0.1%
78 0.4%
230 0.9%
2,994 0.8%
10 < A TSS < 50
Reach % of
Miles Total
0 0.0%
0 0.0%
0 0.0%
297 0.4%
0 0.0%
1,249 1.9%
109 0.3%
21 <0.1%
54 0.3%
102 0.4%
1,832 0.5%
50 < A TSS
Reach % of
Miles Total
0 0.0%
0 0.0%
0 0.0%
111 0.1%
0 0.0%
224 0.3%
14 <0.1%
0 0.0%
0 0.0%
6 0.1%
354 0.1%
  Total does not include 40 freshwater reaches (422 reach miles) for which SPARROW does not predict concentrations.
 November 2009
                                                                                                                                  6-40

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 6-29: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS Concentrations
 Option 2, by EPA Region	
from
                                                      Range of Reductions in TSS Concentration (mg/L)
Total
Improved
EPA Reach
Region Miles1
1 10,740
2 12,554
3 25,471
4 82,955
5 56,906
6 64,954
7 41,102
8 49,967
9 19,853
10 25,873
Nation 390,374
0 < A TSS < 1
%of
Total
Reach Improved
Miles Miles
10,564 98.4%
12,542 99.9%
24,751 97.2%
69,708 84.0%
55,616 97.7%
39,267 60.5%
33,876 82.4%
48,047 96.2%
18,041 90.9%
23,758 91.8%
336,168 86.1%
1 < A TSS < 5
%of
Total
Reach Improved
Miles Miles
128 1.2%
12 0.1%
691 2.7%
10,787 13.0%
1,236 2.2%
18,668 28.7%
6,323 15.4%
1,770 3.5%
1,515 7.6%
1,352 5.2%
42,481 10.9%
5 < A TSS < 10
%of
Total
Reach Improved
Miles Miles
29 0.3%
0 0.0%
30 0.1%
1,396 1.7%
54 0.1%
3,685 5.7%
495 1.2%
104 0.2%
165 0.8%
425 1.6%
6,382 1.6%
10 < A TSS < 50
%of
Total
Reach Improved
Miles Miles
20 0.2%
0 0.0%
0 0.0%
903 1.1%
0 0.0%
2,970 4.6%
394 1.0%
37 0.1%
132 0.7%
282 1.1%
4,738 1.2%
50 < A TSS
%of
Total
Reach Improved
Miles Miles
0 0.0%
0 0.0%
0 0.0%
161 0.2%
0 0.0%
364 0.6%
14 <0.1%
10 <0.1%
0 0.0%
56 0.2%
605 0.2%
  Total does not include 40 freshwater reaches (422 reach miles) for which SPARROW does not predict concentrations.
November 2009
                                                                                                                              6-41

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 6-30: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS Concentrations from
 Option 3, by EPA Region	
                                                      Range of Reductions in TSS Concentration (mg/L)
Total
Improved
EPA Reach
Region Miles1
1 10,740
2 12,560
3 25,471
4 83,028
5 56,906
6 70,630
7 46,859
8 57,322
9 22,204
10 25,920
Nation 411,640
0 < A TSS < 1
%of
Total
Reach Improved
Miles Miles
10,538 98.1%
12,501 99.5%
24,493 96.2%
66,798 80.5%
55,180 97.0%
40,665 57.6%
37,950 81.0%
54,799 95.6%
19,952 89.9%
23,503 90.7%
346,379 84.2%
1 < A TSS < 5
%of
Total
Reach Improved
Miles Miles
120 1.1%
59 0.5%
908 3.6%
12,850 15.5%
1,635 2.9%
21,400 30.3%
7,724 16.5%
2,273 4.0%
1,844 8.3%
1,408 5.4%
50,221 12.2%
5 < A TSS < 10
%of
Total
Reach Improved
Miles Miles
63 0.6%
0 0.0%
70 0.3%
2,021 2.4%
91 0.2%
4,178 5.9%
667 1.4%
203 0.4%
227 1.0%
565 2.2%
8,084 2.0%
10 < A TSS < 50
%of
Total
Reach Improved
Miles Miles
20 0.2%
0 0.0%
0 0.0%
1,186 1.4%
0 0.0%
3,845 5.4%
504 1.1%
37 0.1%
169 0.8%
388 1.5%
6,147 1.5%
50 < A TSS
%of
Total
Reach Improved
Miles Miles
0 0.0%
0 0.0%
0 0.0%
173 0.2%
0 0.0%
543 0.8%
14 <0.1%
10 <0.1%
12 0.1%
56 0.2%
808 0.2%
 1 Total does not include 40 freshwater reaches (422 reach miles) for which SPARROW does not predict concentrations.
November 2009
                                                                                                                              6-42

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 6-31: Total RF1 Miles that Receive Direct Construction Loadings with Improvements in TSS Concentrations from
 Option 4, by EPA Region	
                                                      Range of Reduction in TSS Concentration (mg/L)
Total
Improved
EPA Reach
Region Miles1
1 10,740
2 12,554
3 25,471
4 82,955
5 56,906
6 64,954
7 41,102
8 49,967
9 19,853
10 25,873
Nation 390,374
0 < A TSS < 1
%of
Total
Reach Improved
Miles Miles
10,581 98.5%
12,546 99.9%
24,662 96.8%
68,280 82.3%
55,617 97.7%
36,432 56.1%
33,170 80.7%
47,677 95.4%
17,940 90.4%
23,944 92.6%
330,850 84.8%
1 < A TSS < 5
%of
Total
Reach Improved
Miles Miles
136 1.3%
8 0.1%
779 3.1%
11,758 14.2%
1,235 2.2%
20,548 31.6%
6,887 16.8%
2,048 4.1%
1,610 8.1%
1,329 5.1%
46,339 11.9%
5 < A TSS < 10
%of
Total
Reach Improved
Miles Miles
4 0.1%
0 0.0%
30 0.1%
1,752 2.1%
54 0.1%
3,927 6.1%
568 1.4%
196 0.4%
149 0.8%
351 1.4%
7,029 1.8%
10 < A TSS < 50
%of
Total
Reach Improved
Miles Miles
20 0.2%
0 0.0%
0 0.0%
991 1.2%
0 0.0%
3,552 5.5%
462 1.1%
37 0.1%
142 0.7%
211 0.8%
5,414 1.4%
50 < A TSS
%of
Total
Reach Improved
Miles Miles
0 0.0%
0 0.0%
0 0.0%
173 0.2%
0 0.0%
495 0.8%
14 0.1%
10 0.1%
12 0.1%
37 0.1%
742 0.2%
 1 Total does not include 40 freshwater reaches (422 reach miles) for which SPARROW does not predict concentrations.
November 2009
                                                                                                                              6-43

-------
Environmental Impact and Benefits Assessment for the C&D Category
6.6.4  Impacts on Reservoir Sedimentation

Construction discharges affect not only sediment concentrations in receiving water bodies, but they also
affect the amount of sedimentation that takes place in reservoirs. As described in Section 6.2, reservoirs
provide major sites for sediment attenuation within the hydrographic network. This attenuation is
estimated in SPARROW for more than 4,000 reservoirs distributed throughout the RF1 network. The
model estimates that more than 444 million cubic yards accumulate under baseline conditions. The
accumulating sediment may hinder reservoir functions and need to be periodically dredged, as discussed
in greater detail in Chapter 2. The policy options reduce this amount to various degrees. As shown in
Table 6-32, Option 1 reduces the amount of sediment accumulating in reservoirs slightly as compared to
current conditions (by 558,800 cubic yards or approximately 0.1 percent). Most of this reduction is
concentrated in Regions 4, 6, and 7. Options 2, 3, and 4 show greater reductions in reservoir
sedimentation, with estimated reductions of 1.16, 1.45, and 1.32 million cubic yards, respectively (or
0.3 percent). They also show more geographically distributed effects, with significant reductions observed
in Regions 4, 5, 6, 7, 9, and 10.

 Table 6-32: Sediment Accumulation  in Reservoirs by Policy Option and EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Baseline
1,278
2,725
5,338
45,667
35,940
128,381
42,238
126,435
41,860
14,744
444,606
Sediment Accumulation
(thousand cubic yards)
Option 1 Option 2 Option 3
1,277
2,725
5,335
45,568
35,933
127,973
42,213
126,433
41,855
14,737
444,047
1,275
2,724
5,332
45,462
35,924
127,530
42,186
126,430
41,850
14,729
443,443
1,275
2,724
5,331
45,411
35,920
127,320
42,173
126,429
41,847
14,726
443,156
Option 4
1,276
2,724
5,332
45,440
35,923
127,397
42,179
126,430
41,849
14,732
443,283
Reduction in Sediment Accumulation
(thousand cubic yards)
Option 1 Option 2 Option 3 Option 4
1.4
0.5
2.6
98.8
7.9
408.5
24.8
2.2
4.8
7.3
558.8
2.9
1.1
5.3
205.1
16.3
851.2
51.6
4.5
10.0
14.8
1,162.7
3.6
1.4
6.6
255.7
20.3
1,061.7
64.3
5.6
12.4
18.4
1,449.9
2.1
1.0
5.5
226.5
17.2
984.4
59.1
4.7
10.5
11.8
1,322.9
6.6.5  Estimated Changes in Nutrient Concentrations for all RF1 Watersheds Receiving
       Loading from Construction Sites

Table 6-33 and Table 6-34 summarize national average reach-length weighted TN and TP concentrations
and their distribution under the baseline conditions and the four regulatory options (see Chapter 1 for
option descriptions) for reaches receiving construction sediment discharges. TN and TP concentrations in
the RF1 network vary over a wide range. Nutrient criteria established by EPA for rivers and streams vary
across ecoregions from 0.12 mg/L to 2.18 mg/L for TN and from 10 ug/L to 128 ug/L for TP (USEPA
2009d).
Estimated TN concentrations under baseline conditions range between 0.37 and 14.95 mg/L for the 5th
and 95th percentiles, respectively. Median (50th percentile) TN concentrations are approximately
1.65 mg/L under the baseline scenario. Under Options 1 and 2, the median concentration remains
relatively unchanged at 1.65 mg/L; under Options 3 and 4, it decreases 0.01 mg/L to 1.64 mg/L. Overall,
the largest reduction in TN concentrations is predicted under Option 3 with a 0.5 percent reduction in
November 2009
                                                                                           6-44

-------
Environmental Impact and Benefits Assessment for the C&D Category
median TN concentrations relative to baseline levels. These numbers represent the national average
concentration reduction, however, and do not reflect the wide geospatial variability in TN concentration
reductions.
 Table 6-33: Distribution of TN  Concentration Based on SPARROW Output for 31,927
 RF1 Reaches Receiving Construction Sediment Discharges
Policy
Option
Baseline
Option 1
Option 2
Option 3
Option 4
Reach
Count
31,927
31,927
31,927
31,927
31,927
Reach
Miles
412,062
412,062
412,062
412,062
412,062
Average
TN
(mg/L)u
18.22
18.21
18.21
18.21
18.21
Reduction
in Average
TN
-
0.01
0.02
0.02
0.02
Distribution of TN Concentrations, by Percentile
(mg/L)1'2
5th 25th 50th 75th 95th
0.37
0.37
0.36
0.36
0.36
0.91
0.91
0.91
0.90
0.91
.65
.65
.65
.64
.65
3.88
3.87
3.87
3.87
3.87
14.95
14.91
14.91
14.91
14.91
 'Reach-length weighted.
 2Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.

Values for TP concentrations for the 5th and 95th percentile bounds under baseline conditions are 30.5 and
2,907.3 ug/L, respectively. Median (50th percentile) TP concentrations are approximately 253.2 ug/L
under the baseline scenario. Under Option 1, the median concentration decreases 1.4 ug/L to 251.8 ug/L.
The median concentration under Option 2 decreases 1.9 ug/L to 251.3 ug/L and under Option 3 and 4, it
decreases 2.1 ug/L to 251.1 ug/L. These numbers represent the national average concentration reduction,
however, and do not reflect the wide geospatial variability in TP concentration reductions.

 Table 6-34: Distribution of TP Concentration Based on SPARROW Output for 31,927
 RF1 Reaches Receiving Construction Sediment Discharges
Policy
Option
Baseline
Option 1
Option 2
Option 3
Option 4
Reach
Count
31,927
31,927
31,927
31,927
31,927
Reach
Miles
412,062
412,062
412,062
412,062
412,062
Average Reduction Distribution of TP Concentrations, by Percentile
TP in Average (ng/L)1'2
(jig/L)1'2 TP 5th 25th 50th 75th 95th
3,921.6
3,916.3
3,915.9
3,915.7
3,915.9

5.3
5.7
5.9
5.8
30.5
29.9
29.6
29.6
29.6
104.4
103.8
103.4
103.3
103.4
253.2
251.8
251.3
251.1
251.1
685.3
683.7
683.6
683.5
683.5
2,907.3
2,886.4
2,886.4
2,885.1
2,886.4
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.	

Changes in nutrient concentrations are based on in-stream sediment reductions, as described in Section
6.4. Therefore, the distribution and magnitude of percent change in reach miles will be the same as for
TSS, as presented in Section 6.6.3. These tables are not recreated below for TN and TP.
Table 6-35 through Table 6-39 detail TN concentrations by EPA region under baseline conditions and for
each of the four regulatory options considered. The highest average TN concentrations are found in
Regions 8 and 9, which have average reach-length weighted TN concentrations between 40.01 and
46.56 mg/L under all scenarios. However, the highest median TN concentrations are found in Regions 5
and 7, with concentrations of 3.68 and 5.11 mg/L, respectively. Regions 8 and 9 appear to have a skewed
distribution with a smaller portion of reach miles at high concentrations skewing the average higher than
November 2009                                                                                 6-45

-------
Environmental Impact and Benefits Assessment for the C&D Category
other regions. Under all four options, TN concentrations decline a small amount, approximately
0.1 percent, from baseline concentrations.
Table 6-35: Summary of Baseline TN Concentration
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach Average
Miles TN (mg/L)1
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
1.32
19.43
1.88
4.39
7.76
26.15
22.87
40.01
46.56
5.30
18.22
in RF1 Reaches Receiving
Distribution of TN
5th 25th
0.24
0.33
0.61
0.29
0.51
0.45
1.10
0.50
0.24
0.23
0.37
0.59
0.95
1.02
0.70
1.49
1.07
2.64
1.08
0.55
0.49
0.91
Concentrations, by Percentile (mg/L)1
50th 75th 95th
0.85
1.37
1.35
1.06
3.68
1.83
5.11
2.44
1.28
0.83
1.65
1.27
1.98
1.95
1.59
7.50
3.26
9.56
5.99
3.35
1.58
3.88
2.57
3.85
4.35
3.06
15.86
12.38
21.50
28.66
54.43
5.29
14.95
'Reach-length weighted.

Table 6-36: Summary Option 1 TN Concentration in
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TN
(mg/L)u
1.32
19.43
1.88
4.39
7.76
26.12
22.86
39.99
46.51
5.29
18.21
RF1
Distribution of TN
5th 25th
0.24
0.33
0.61
0.28
0.51
0.43
1.08
0.48
0.23
0.23
0.36
0.59
0.95
1.02
0.70
1.49
1.06
2.62
1.07
0.54
0.49
0.91
Reaches
Receiving
Concentrations, by Percentile
50th 75th
0.85
1.37
1.35
1.06
3.68
1.83
5.10
2.42
1.28
0.83
1.65
1.27
1.98
1.95
1.58
7.49
3.24
9.53
5.96
3.26
1.58
3.87

(mg/L)1'2
95th
2.57
3.85
4.35
3.06
15.86
12.37
21.50
28.64
54.43
5.29
14.91
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.
November 2009
                                                                                                         6-46

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-37: Summary of Option 2 TN Concentration in RF1
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TN
(mg/L)1'2
1.32
19.43
1.88
4.38
7.76
26.12
22.85
39.99
46.51
5.29
18.21
Reaches Receiving
Distribution of TN Concentrations,
5th 25th 50th
0.24
0.33
0.61
0.28
0.51
0.43
1.08
0.48
0.23
0.23
0.36
0.59
0.95
1.02
0.70
1.48
1.06
2.62
1.07
0.54
0.49
0.91
0.85
1.37
1.35
1.06
3.68
1.83
5.10
2.42
1.28
0.83
1.65
by Percentile (mg/L)1'2
75th 95th
1.27
1.98
1.95
1.58
7.49
3.23
9.49
5.96
3.26
1.58
3.87
2.57
3.85
4.34
3.06
15.86
12.35
21.50
28.64
54.43
5.29
14.91
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.

Table 6-38: Summary of Option 3 TN Concentration in RF1 Reaches Receiving
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TN
(mg/L)1'2
1.32
19.43
1.88
4.38
7.75
26.11
22.85
39.99
46.51
5.29
18.21
Distribution
5th
0.24
0.33
0.61
0.28
0.51
0.42
1.08
0.48
0.23
0.23
0.36
ofTN
25th
0.59
0.95
1.02
0.70
1.48
1.06
2.61
1.07
0.54
0.49
0.90
Concentrations, by
50th
0.85
1.37
1.35
1.06
3.68
1.83
5.10
2.42
1.28
0.83
1.64
Percentile
75th
1.27
1.98
1.95
1.58
7.49
3.23
9.49
5.96
3.26
1.58
3.87
(mg/L)1'2
95th
2.57
3.85
4.34
3.06
15.86
12.34
21.50
28.64
54.43
5.29
14.91
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.
November 2009
                                                                                                                       6-47

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-39: Summary of Option 4 TN Concentration in RF1
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TN
(mg/L)1'2
1.32
19.43
1.88
4.38
7.76
26.11
22.85
39.99
46.51
5.29
18.21
Reaches Receiving
Distribution of TN Concentrations,
5th 25th 50th
0.24
0.33
0.61
0.28
0.51
0.42
1.08
0.48
0.23
0.23
0.36
0.59
0.95
1.02
0.70
1.48
1.06
2.62
1.07
0.54
0.49
0.90
0.85
1.37
1.35
1.06
3.68
1.83
5.10
2.42
1.28
0.83
1.64
by Percentile (mg/L)1'2
75th 95th
1.27
1.98
1.95
1.58
7.49
3.23
9.49
5.96
3.26
1.58
3.87
2.57
3.85
4.34
3.06
15.86
12.34
21.50
28.64
54.43
5.29
14.91
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.

Table 6-40 through Table 6-44 detail TP concentrations by EPA region under baseline conditions and for
each of the four regulatory options considered. The highest average TP concentrations are found in
Regions 2, 6, 7, 8, and 9, which have average reach-length weighted TP concentrations greater than
1,600 ug/L under all scenarios. These regions, with the exception of Region 2, also show the highest
median TP concentrations (ranging from 489.0 to 659.0 ug/L under baseline conditions). Option 1 shows
the smallest TP concentration reductions, with Options 2 and 3 showing slightly higher reductions.
Option 4 shows TP concentration reductions that are between those under Option 2 and Option  3, similar
to trends observed for TN.
Table 6-40: Summary of Baseline TP Concentration in RF1
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TP
(ng/L)1
109.5
1,697.5
209.3
885.6
651.2
8,110.6
3,380.2
7,282.3
13,544.9
965.7
3,921.6
Reaches Receiving
Distribution of TP Concentrations, by Percentile (jig/L)1
5th 25th 50th 75th 95th
9.9
16.6
37.7
30.0
18.6
63.4
140.6
69.0
37.0
22.1
30.5
29.6
47.2
80.2
81.8
75.6
208.4
377.8
249.5
114.9
56.3
104.4
49.2
91.2
135.4
139.5
225.3
519.5
603.9
659.0
489.0
113.4
253.2
86.5
152.1
233.5
241.8
475.9
1,216.9
1,012.7
1,449.5
1,652.3
303.0
685.3
233.8
403.0
633.6
547.4
1,246.4
5,557.8
2,853.1
5,803.0
30,733.9
1,318.7
2,907.3
'Reach-length weighted.
November 2009
                                                                                                 6-48

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-41: Summary of Option 1 TP Concentration in RF1 Reaches Receiving
Construction Sediment Discharges, by EPA Region
EPA Reach
Region Count
1 1,213
2 1,111
3 2,012
4 7,564
5 3,918
6 4,634
7 3,063
8 4,300
9 1,439
10 2,673
Nation 31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TP
(Hg/L)u
109.4
1,697.5
209.2
885.1
651.2
8,096.5
3,377.4
7,277.2
13,513.0
965.2
3,916.3
Distribution of TP Concentrations, by Percentile (ng/L)1'2
5th 25th 50th 75th 95th
9.9
16.6
37.7
29.8
18.6
61.3
138.4
66.6
36.2
21.5
29.9
29.5
47.2
80.1
81.4
75.5
205.2
375.7
246.4
111.7
56.1
103.8
49.2
91.2
135.3
139.1
225.3
517.0
603.3
656.7
476.2
112.3
251.8
86.5
152.0
233.4
241.1
475.8
1,213.2
1,010.4
1,445.2
1,644.4
303.0
683.7
233.8
403.0
633.6
547.4
1,246.4
5,557.7
2,853.1
5,780.2
30,733.9
1,314.0
2,886.4
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.

Table 6-42: Summary of Option 2 TP Concentration in RF1 Reaches Receiving
Construction Sediment Discharges, by EPA Region
EPA Reach
Region Count
1 1,213
2 1,111
3 2,012
4 7,564
5 3,918
6 4,634
7 3,063
8 4,300
9 1,439
10 2,673
Nation 31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TP
(Hg/L)u
109.4
1,697.5
209.1
884.7
651.1
8,095.2
3,377.2
7,277.2
13,512.8
964.9
3,915.9
Distribution of TP Concentrations, by Percentile (ng/L)1'2
5th 25th 50th 75th 95th
9.9
16.6
37.7
29.3
18.6
59.0
138.4
66.6
35.3
21.5
29.6
29.5
47.2
80.0
81.2
75.5
203.8
374.9
246.4
111.4
55.7
103.4
49.2
91.2
135.2
138.9
225.3
516.6
602.8
656.6
475.1
112.3
251.3
86.5
152.0
233.3
240.2
475.8
1,211.4
1,008.8
1,445.2
1,644.4
300.4
683.6
233.8
402.9
633.5
547.3
1,246.4
5,557.7
2,853.1
5,780.1
30,733.9
1,289.3
2,886.4
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.
November 2009
                                                                                                                       6-49

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 6-43: Summary of Option 3 TP Concentration in RF1 Reaches Receiving
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TP
(ng/L)1'2
109.4
1,697.5
209.1
884.5
651.1
8,094.6
3,376.9
7,277.0
13,512.7
964.7
3,915.7
Distribution of TP Concentrations, by Percentile (ng/L)1'2
5th 25th 50th 75th 95th
9.9
16.6
37.7
29.2
18.5
58.6
138.4
66.6
35.2
21.5
29.6
29.4
47.2
80.0
81.2
75.5
203.6
374.9
246.4
111.4
55.6
103.3
49.2
91.1
135.1
138.6
225.3
516.2
602.2
656.6
474.5
112.3
251.1
86.5
152.0
233.3
240.0
475.8
1,209.8
1,008.8
1,445.2
1,644.4
297.9
683.5
233.7
402.9
633.5
547.3
1,246.4
5,557.7
2,850.8
5,780.1
30,733.9
1,277.5
2,885.1
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.

Table 6-44: Summary of Option 4 TP Concentration in RF1 Reaches Receiving
Construction Sediment Discharges, by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Reach
Count
1,213
1,111
2,012
7,564
3,918
4,634
3,063
4,300
1,439
2,673
31,927
Reach
Miles
10,740
12,560
25,471
83,058
56,909
70,687
46,903
57,513
22,280
25,940
412,062
Average
TP
(Hg/L)u
109.4
1,697.5
209.1
884.6
651.1
8,094.9
3,377.1
7,277.1
13,512.8
965.0
3,915.9
Distribution of TP Concentrations, by Percentile (ng/L)1'2
5th 25th 50th 75th 95th
9.9
16.6
37.7
29.3
18.6
58.8
138.4
66.6
35.3
21.5
29.6
29.5
47.2
80.0
81.2
75.5
203.7
374.9
246.4
111.4
55.8
103.4
49.2
91.2
135.2
138.9
225.3
516.2
602.4
656.6
475.0
112.3
251.1
86.5
152.0
233.3
240.2
475.8
1,211.2
1,008.8
1,445.2
1,644.4
301.2
683.5
233.8
402.9
633.5
547.3
1,246.4
5,557.7
2,853.1
5,780.1
30,733.9
1,297.2
2,886.4
 'Reach-length weighted.
 2 Nutrient reductions are based on TSS reductions. Any TSS concentration above the 95th percentile is assumed to be an outlier and was
 replaced with the 95th percentile value.
November 2009
                                                                                                                       6-50

-------
Environmental Impact and Benefits Assessment for the C&D Category
        Benefits to Navigation
Navigable waterways, including rivers, lakes, bays, shipping channels and harbors, are an integral part of
the United States' industrial transportation network. Navigable channels are prone to reduced
functionality due to sediment build-up, which can reduce the navigable depth and width of the waterway
(Clark et al. 1985). Removing sediment to keep navigable waterways passable requires dredging. The
U.S. Army Corps of Engineers (USAGE) spends an average of more than $572 million (2008$) every
year to dredge waterways and keep them passable (USAGE 2009).
Implementation of the regulation is expected to result in less frequent dredging of navigable waterways
due to reduced sediment runoff from construction sites. This will result in avoided costs to the
government and private entities responsible for maintenance of navigational shipping channels, harbors,
rivers, and other waterways. For example, prior studies of benefits of the Conservation Reserve program
found that a reduction of 205 million tons of sediment erosion from farmland would result in
approximately $526 million (2008$) in benefits to navigation from decreased dredging (Ribaudo 1989).
Finally, unless there is overdredging to compensate or sedimentation is monitored so that dredging
activity may be timed optimally, water bodies will be on average less navigable even with a dredging
program.
This chapter presents EPA's analysis of the navigable waterway maintenance costs that would be avoided
by implementation of the regulation. This approach represents an appropriate measure of social benefits
for cases in which a policy change reduces costs to producers (in this case "producers" of navigable
waterways), but in which price effects are minimal (cf Boardman et al. 2001, p. 70-74). Additional
environmental benefits of reduced dredging are not estimated due to the highly place-specific nature of
these benefits and difficulty in estimating these costs over the national scale. Hence, in this regard the
Agency's benefit estimate in this chapter is understated. This analysis includes the following steps:
        >  Identifying navigable waterways that are regularly dredged and estimating the frequency of
           regular dredging  in each waterway
        >  Estimating the navigable waterway maintenance cost per cubic yard of sediment dredged
        >  Estimating the total cost of navigable waterway maintenance under the baseline and post-
           compliance scenarios
        >  Estimating cost saving from decreased dredging of navigable waterways due to the reduction
           in sediment runoff from construction sites.

7.1     Data Sources

EPA relied on the USAGE Dredging Information System in analyzing benefits to navigation from the
regulation. The dredging database catalogs all USAGE dredging contracts from  1990 to 2008 and all
USACE-conducted dredging jobs from 1995 to 2008, and provides information on the location, dates,
cost, and amount of sediment dredged for each dredging job or contract. For the purpose of this analysis,
a "dredging occurrence" refers to a single record in these data, containing a location and time of dredging,
while a "dredging job" refers to a single location of dredging that may have more than one dredging
occurrence, and thus be repeated in the data.
November 2009                                                                                 7-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
Where the USAGE Dredging Information System did not contain cost or quantity of sediment dredged for
a listed dredging occurrence, EPA estimated the missing information from other jobs. The Agency
calculated 10th, 50th, and 90th percentile costs and quantity boundaries and used these to fill in all
incomplete records in that region. EPA included 50th percentile (median) estimates in the summary data
presented in Table 7-1 through Table 7-4 and in the midpoint estimate of total dredging costs. The 10th
and 90th percentile cost and amount dredged estimates were used in the low and high total dredging cost
estimates, respectively. For more details, see Section 7.4.3: Sensitivity Analysis.
Table 7-1 summarizes dredging information for dredging jobs for the period 1995-2008, including
dredging performed by USAGE and dredging contracted to other firms. The cost data were adjusted to
2008$ using the construction cost index (CGI) (Construction Cost Index 2008). During this period,
USAGE funded nearly 3,400 dredging occurrences, worth more than $9 billion (2008$). This dredging
removed more than 2.6 billion cubic yards of sediment from more than 1,100 navigable reaches. The
region with the most dredging was Region 4, reporting more than 900 dredging occurrences between
1995 and 2008,  about 65 per year. Region 6 reported the largest volume of sediment removed, with nearly
1.5 billion cubic yards removed from navigable waterways over this 14-year period. Though benefits are
also dependent on the amount of sediment reduction in a region,  regions with more frequent dredging are
more likely to experience greater benefits from reduced sedimentation in navigable waterways.
                  Table 7-1: Dredging in U.S. Navigable Waterways,
                  1995-2008
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Number of
Dredging
Occurrences
79
278
291
921
739
647
22
1
129
290
3,397
Source: Dredging Information
Total Sediment
Removed
(millions of cubic
yards)
3.7
111.4
124.5
540.9
112.8
1,483.3
17.3
0.5
67.7
153.5
2,615.7
System (USAGE 2009).
Total Cost
(millions of 2008$)
$45.7
$801.1
$902.8
$3,121.3
$517.8
$2,706.4
$39.5
$2.3
$449.3
$499.9
$9,086.3

Table 7-2 summarizes dredging activity from 1995 through 2008 by type of entity performing dredging
work (private or government). As noted above, USAGE spent more than $9 billion (2008$) over the past
14 years to fund nearly 3,400 dredging occurrences. The majority of dredging work (58 percent) was
contracted out, with USAGE performing about 42 percent of the work over this period. The individual
costs of these dredging occurrences range considerably, from around $10,000 to more than $100 million,
but averaging about $3 million (2008$).
November 2009                                                                                7-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 7-2: Dredging
1995-2008

USAGE
Reported
Occurrences
of Dredging
1,414
Contract 1,983
Total
Source:
3,397
Dredging Information
of Navigable Waterways Performed or Contracted by USAGE,
Percent of
Total
Reported
Dredging
Occurrences
42%
58%
100%
System (USAGE 2009).
Reported
Sediment
Dredged
(millions of
cubic yards)
1,067.0
1,548.7
2,615.7

Percent of
Total
Reported
Sediment
Dredged
41%
59%
100%

Total
Reported
Cost (millions
of 2008$)
$3,963
$5,124
$9,086

Percent of
Total
Reported
Cost
44%
56%
100%

7.2    Identifying Waterways That Are Dredged, the Frequency of Dredging, and
       the Quantity of Sediment Dredged

7.2.1  Determining Dredging Job Locations

Location information is given by latitude-longitude coordinates for some jobs, but for others, location
information was limited to the name of the job (usually the waterway dredged) and the USAGE district
that performed the job. To identify the nearest reach segment, EPA used an unprojected version of the
ERF 1.2 (Enhanced Reach File) from USGS. For this analysis, EPA researched latitude-longitude
coordinates for jobs where they were not provided and linked them to Reach File Version 1.0 (RF1)
reaches. Each latitude/longitude of interest was matched to the nearest point in the ERF 1.2 universe of
points using a spherical model of the earth and a standard haversine distance formula. To associate the
point with an RF1 reach number, EPA used the "RR" (River Reach) attribute from the ERF 1.2 attribute
table. No reach types were excluded from consideration in the nearest reach calculation. Because these
coordinates may be inexact, dredging activity was aggregated to the state level for the calculations in this
analysis and is presented at a regional level.

7.2.2  Identifying the  Baseline Frequency of Dredging

Continuous sediment deposition in a navigable waterway is likely to require repeated dredging to
maintain navigability. Between  1995 and 2008, 1,272  sites required dredging on at least one occasion,
and 551 required dredging more than once. Table 7-3 shows these statistics for each EPA region. The
number of dredging jobs varies by region, and is likely to be influenced by the size of the region, its
number of navigable waterways, and their economic importance.  Because the data available from USAGE
reflect a relatively short period of time (14 years) and dredging data show a large degree of variability in
the frequency of dredging occurrences, EPA calculated an average frequency of recurrence for each
dredging job by dividing 14 years by the number of occurrences of each job over these 14 years.7 The
EPA-estimated regional average recurrence intervals range from 9.6 to 14 years, including jobs that
occurred only once in 14 years.  For all dredging jobs in the United States, the average frequency of
recurrence is about 10 years. Excluding jobs that occurred only once over this period, the regional average
recurrence interval ranges from 4.4 to 7 years, averaging slightly  less than 5 years at the national level.
7   As described later in this section, EPA used the number of days between dredging occurrences and imposed a minimum
    number of intervening days for a dredging occurrence to be considered unique. Occurrences separated by fewer than this
    minimum number of days were considered as a single occurrence.
November 2009                                                                                7-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
Prior studies (Clark et al. 1985; Ribaudo 1989) of the avoided costs from reduction in sedimentation of
navigable waterways, shipping channels, and harbors assumed a linear relationship between sediment
runoff and dredging activity, implying that every ton of sediment entering navigable waterways would
eventually be dredged up. Because dredging is an expensive and environmentally disruptive activity, it is
unlikely that every navigable waterway that receives sediment will eventually be dredged.
For this analysis, EPA assumed that navigable waterways that are regularly dredged would benefit from
reducing the amount of sediment that needs to be dredged up. Although other waterways may benefit
from reduced sedimentation, EPA based its estimation of reduced dredging activity based on historic
dredging data from USAGE. For dredging jobs that occurred only once between 1995 and 2008, EPA
assumed that this job or one similar to it will occur over the same time frame in the future. However,
these single occurrence jobs are excluded from the low range estimate of the analysis (see Section 7.4.3:
Sensitivity Analysis).

          Table 7-3: Dredging Jobs and Recurrence Intervals, 1995- 2008	
                                         Number of       Average     Average Interval
             EPA                        Recurring       Interval       for Recurring
            Region     Number of Jobs	Jobs	(years)	Jobs (years)
1
2
3
4
5
6
7
8
9
10
Total
46
105
133
294
293
226
18
1
45
111
1,272
Source: Dredging Information System
11
48
50
126
139
104
2
0
21
50
551
(USAGE 2009).
11.8
9.7
10.7
10.0
9.6
9.6
13.2
14.0
9.6
9.9
10.0

4.6
4.7
5.3
4.6
4.7
4.4
7.0
N/A
4.5
4.9
4.7

EPA used the USAGE dredging data to identify navigable waterways that are dredged (dredging jobs), as
well as the frequency at which dredging activity occurs. For each dredging job included in the analysis,
the Agency identified:
        >  The number of occurrences of this dredging job
        >  The time elapsed between occurrences of dredging.
EPA excluded from this analysis any occurrences where the time elapsed since the last dredging
occurrence was less than 30 days due to uncertainty as to  whether these occurrences were continuations of
an earlier occurrence of dredging. From this information,  EPA established a dredging schedule for each
recurring dredging job for the next 14 years. For the 721 jobs that only occurred once from 1995 to 2008,
the number of future dredging jobs is set to one, except in the low estimate, where it is set to zero (see
Section  7.4.3: Sensitivity Analysis). For other dredging jobs, EPA divided 14 years by the number of
occurrences of that job between 1995 and 2008 to obtain an average interval of dredging for that job in
years.
November 2009
                                                                                            7-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
7.2.3  Identifying the Amount of Sediment Dredged
For most jobs, the USAGE data reports sediment dredged (in cubic yards), but in cases where this data
was missing, EPA filled in the missing data with percentile values, as described in Section 7.7. In
addition, for occurrences of the same job that were within 30, 90, or 180s days of one another, the
occurrences were consolidated into one occurrence and total amount of sediment dredged was summed.
The total quantity of sediment dredged for each job (single  location) over the past 14 years was divided
by the number of occurrences of that job to calculate an average quantity of sediment dredged for an
occurrence of that job. EPA assumed that this quantity of sediment would be  dredged each time the job
occurs in the future under the baseline scenario, and that it would be reduced  due to the regulation (see
Section 7.4.1).

7.3    Estimating the Navigational Maintenance Cost per Cubic Yard of Sediment
       Removed

Table 7-4 shows the average cost of dredging for each EPA region. Average unit dredging costs vary
considerably over these regions, from around $1.80 per cubic yard in Region  6 to nearly $ 11 per cubic
yard in Region 1. The average unit cost of dredging for the  entire United States is $3.40 per cubic yard.
Table 7-4:
EPA Region
1
2
3
4
5
6
7
8
9
10
Total
Costs of Recurring Dredging, 1995-2006
Total Cost
(millions of 2008$)
$50.1
$1,057.3
$1,091.7
$3,862.2
$625.7
$3,443.4
$48.6
$2.3
$544.2
$717.7
$11,443.1
Source: USACE Dredging Information
Total Sediment Removed
(millions of cubic yards)
4.8
147.0
157.2
705.9
133.4
1,915.7
18.6
0.5
84.1
196.7
3,363.9
System (2009).
Average Cost per
cubic yard
$10.49
$7.19
$6.95
$5.47
$4.69
$1.80
$2.60
$4.49
$6.47
$3.65
$3.40

Though USACE does not disaggregate its reported dredging costs, the cost of sediment dredging typically
includes the following components (Sohngen and Rausch 1998):

       >  The cost of dredging sediment from the bed and loading onto the boat
       >  The cost of transporting the material to a disposal facility
       >  The cost of confining or disposing of the dredged sediment.
    The threshold for determining whether an occurrence was an extension of the previous occurrence varies between the low
    (180 days), midpoint (90 days), and high (30 days) estimates.
November 2009                                                                                7-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
7.4    Estimating the Total Cost of Navigable Waterway Maintenance Under
       Different Policy Scenarios

Avoided costs of navigational waterway dredging were estimated by comparing estimated future dredging
costs under the no policy scenario with estimated future dredging costs under the post-compliance
scenario.

7.4.1   Estimating the Reduction in Sediment Dredging in Navigable Waterways Due to
       the Reduction in Discharge from Construction Sites

EPA estimated changes in sediment deposition in navigable waterways under different regulatory
scenarios using output from SPARROW. Chapter 6 provides a description of SPARROW as well as input
data used in this analysis. For the post-compliance scenarios, EPA calculated the percentage change in
sediment yield from the baseline for each RF1 reach in the SPARROW analysis. The quantity of sediment
for each dredging job (calculated as described in Section 7.2.3) was multiplied by this percentage change
in sediment yield to obtain a post-compliance quantity of sediment to be dredged each time  a job occurs
in the next 14 years9. If an RF1 reach was not assigned to a dredging job due to data limitations or the
reach assigned to that dredging job was not included in the SPARROW model (such as coastal reaches),
the state-wide average reduction in sediment yield was applied to that job.

7.4.2   Estimating the Total Cost of Navigable Waterway Maintenance Under the Baseline
       and Post-Compliance Scenarios

EPA estimated future dredging costs over a period of 14 years from 2010 to 2023 under both the baseline
and post-compliance scenarios. For this cost calculation, average cost per cubic yard of sediment dredged,
total amount of sediment dredged, and the average interval between dredging jobs were calculated for
each relevant job in the USAGE Dredging Information System. For the post-compliance scenarios, total
amount of sediment dredged was determined as described in Section 7.4.1.
Because costs will occur whenever the waterbody is dredged rather than on an annual basis, each reach
will have a unique stream of costs derived from its individual dredging frequency, cost, dredging volume,
and reduction in sediment. Thus, EPA calculated the present value of dredging costs over the 14-year
period using  a dredging schedule for each dredging job and then discounted the estimated values using
both 3 and 7  percent annual interest rates. EPA estimated the present value of each dredging job as
follows:
       > Estimating the average recurrence interval for the job over the 14-year period for each job by
          dividing 14 years by its number of occurrences in the USAGE dredging data. This produces /
          in the equation below.
       > Entering this information along with the cost per cubic yard dredged, number of cubic yards
          dredged, and percentage of sediment requiring dredging in the post-compliance scenario into
          the formula below:
                                         14/1
                                             Qh*R   *C
                                                   pc
                                                                                    (Eq. 7-1)
    EPA estimated future dredging for 14 years to reflect the 14 years of historic data provided by USAGE.
November 2009                                                                              7-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
       Where:

       PV   = Present Value

       Z     is the sum of the expression in brackets for the number of dredging occurrences of this
                dredging job

       /      = recurrence interval in years (including fractions of a year) for this dredging job

        14/7   = number of periods between recurrence intervals for which costs will be discounted

       Qb    = cubic yards dredged under the baseline scenario

       Rpc   = percentage of cubic yards remaining in the post-compliance scenario

       C     = cost per cubic yards dredged

       d     = annual discount rate10

       d*I   = the discount rate for the period, being the annual rate d prorated for the length of the
                period / in years

       n      = the number of the specific dredging occurrence.

For example, if the Mississippi River is scheduled to be dredged every 2 years between 2010 and 2023,
there will be no dredging costs for year 1, but the dredging costs in year 2 will be equal to the cost per
cubic yard  multiplied by the amount of sediment dredged, then discounted by 2 years. There will again be
no dredging costs for year  3 . Dredging costs will then be incurred in year 4 as in year 2, but discounted by
4 years, and so on. By the end of 2023, the  Mississippi River will have been dredged seven times. The
sum of the  discounted costs of these six dredging occurrences is the total present value of the dredging
cost of the  Mississippi River for the next 14 years. Repeating this process for every dredging job in this
analysis yields an estimate of national dredging costs over the next 14 years.
Alternately, the frequency  can be interpreted as occurring at the beginning of each interval rather than at
the end, so that in the example above, the Mississippi River would be dredged in year 1, then in year 3,
etc., and discounted accordingly. This approach was taken for EPA's high  estimate  (see below).

7.4.3  Sensitivity Analysis

To account for uncertainty in projecting future dredging costs, EPA adjusted the assumptions described in
Section 7.2 to provide a range of benefits estimates. EPA estimated low, midpoint, and high levels of
dredging activity for this sensitivity analysis as follows:
Low estimate:
           Assuming a minimum recurrence interval (below which occurrences are assumed to be
           continuations of the previous occurrence of the same job) of 180 days
    EPA estimated costs using both 3 and 7 percent discount rates, in accordance with OMB (2003, 33-34).
November 2009                                                                                   7-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
        >   Assigning 10th percentile cost and cubic yards dredged estimates for jobs lacking these data
        >  Excluding jobs with only one occurrence over the 14 years from 1995 to 2008.
Midpoint estimate:
        >  Assuming a minimum recurrence interval of 90 days to include jobs with more closely spaced
           recurrences
        >  Using 50th percentile (median) cost and cubic  yards dredged estimates for jobs lacking these
           data
        >  Including jobs with only one occurrence over  the 14 years from 1995 to 2008 under the
           assumption that these jobs or similar jobs will occur once over the next 14 years.
High estimate:
        >  Assuming a minimum recurrence interval of 30 days
        >  Using 90th percentile cost and cubic yards dredged estimates for jobs lacking these data
        >  Including jobs with only one occurrence between 1995 and 2008 under the assumption that
           these jobs or similar jobs will occur once over the next 14 years.
This analysis also presents costs for all estimates using both 3 and 7 percent discount rates.

7.4.4   Annualizing Future Dredging Costs

Though dredging costs will not be incurred at a regular annual rate, EPA annualized its 14-year cost
projection in order to facilitate the comparison of these costs to other costs calculated for the analysis of
the regulation that are measured on an annual timeframe. This annualized cost was calculated  using the
same interest rates used to discount future dredging costs.  Table 7-5 presents estimated total annualized
dredging (cubic yards removed)  and the  estimated costs for navigable waterway dredging for the baseline
scenario, including low, midpoint, and high estimates.
Under the  baseline scenario, midpoint dredging costs projected over the period of 2010 to 2023 are
expected to range between $468 and 620 million per year (annualized using a 7 and a 3 percent discount
rate, respectively). The high and low estimates produce an overall range between $360 million and $655
million. Regions 4 and 6 are expected to incur the highest  costs, as they have the largest percentage of
dredging activity and amounts of sediment dredged. In the low estimate, Region 8 incurs no dredging
costs because the sole dredging job recorded in Region 8 in the USAGE Dredging Information System
occurs only once and was thus excluded. Regions  1 and 7  also have considerably lower dredging costs
than other regions, due to the small number of recorded dredging jobs in these areas.
November 2009                                                                                 7-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
  Table 7-5: Annualized Dredging Costs Under the Baseline Scenario (millions of 2008$)
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Sediment Dredged
(millions of cubic yards)
Low
1.4
84.5
95.5
405.5
85.4
1,240.5
7.4
0.0
60.5
112.2
2,093.0
Mid
3.7
111.5
124.6
540.8
112.7
1,483.3 1
17.3
0.5
67.7
153.5
2,615.7 2
High
6.6
130.5
134.8
575.9
149.2
,503.8
18.8
0.5
82.3
178.3
,780.8
Cost
Low
$1.4
$39.4
$47.5
$160.9
$23.6
$148.4
$1.0
$0.0
$27.6
$24.7
$474.5
using 3% discount
(millions of 2008$)
Mid
$3.0
$54.1
$61.6
$213.2
$34.9
$185.6
$2.5
$0.1
$31.1
$34.2
$620.4
rate
High
$3.6
$56.9
$63.2
$221.3
$44.6
$191.5
$2.8
$0.1
$33.1
$37.6
$654.7
Cost
Low
$1.0
$29.8
$36.2
$122.4
$17.9
$112.3
$0.7
$0.0
$21.1
$18.9
$360.3
using 7% discount
(millions of 2008$)
Mid
$2.2
$40.6
$46.5
$161.0
$26.2
$140.3
$1.8
$0.1
$23.6
$25.8
$468.1
rate
High
$2.6
$42.6
$47.6
$167.4
$33.5
$145.4
$2.0
$0.1
$25.2
$28.4
$494.9
November 2009
                                                                                                                       7-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
7.5     Estimating the Avoided Costs from Decreased Dredging of Navigable
        Waterways

The avoided costs for each post-compliance scenario are calculated as the difference in total annualized
dredging costs between the baseline and each post-compliance scenario, and are considered the benefits to
navigation resulting from the regulation. Table 7-6, Table 7-7, Table 7-8, and Table 7-9 present
annualized avoided costs from reduced dredging of navigable waterway for each of the Agency's policy
options, including low, midpoint, and high estimates for cost reductions under each policy option. Each of
these estimates was calculated using both 3 and 7 percent rates to discount and annualize costs. Because
the discount rate does not make a significant difference in the overall avoided costs, all values discussed
are those calculated assuming a 3 percent discount rate.
Annualized savings from reduced dredging activity range from $1.0 to $3.4 million. EPA estimates that
Regions 4 and 6 will benefit from the most substantial reductions in dredging costs under all policy
options. This is due to a large amount of dredging activity in these  regions, and a large percentage
reduction in sediment runoff expected as a result of the regulation.  Due to the lack of significant dredging
activity in Region 8, no noticeable benefits are expected in this region.
Option 1, which requires non-numeric effluent limitations for all sites, is EPA's least stringent policy
option. It is predicted to produce a range of avoided costs between  $1.0 and $1.3 million with a midpoint
estimate of slightly less than $1.3. EPA predicts that this option would prevent 8.5 million cubic yards of
sediment from entering navigable waters each year.
Option 2 requires ATS on sites with more than 30 acres disturbed at one time and imposes a 13 NTU
turbidity standard while requiring non-numeric effluent limitations on all sites and is similar to the option
EPA proposed previously in 2008. This option will prevent an estimated 17.6 million cubic yards of
sediment from entering navigable water bodies and requiring dredging. The midpoint estimate for avoided
costs under this option is $2.6 million per year, ranging from $2.1 to $2.8 million between the low and
high estimates.
Option 3, EPA's most stringent policy option, requires ATS on sites with more than  10 acres disturbed at
one time, imposes a 13 NTU turbidity standard on these sites, and requires non-numeric effluent
limitations on all sites. This option would prevent approximately 22.0 million cubic yards of sediment
from building up in navigable waterways each year. Avoided costs from this action range from $2.7 to
$3.4 million, with a midpoint of $3.3 million.
Option 4 requires passive treatment systems on all sites with 10 or  more acres disturbed at one time, and
establishes a numeric turbidity standard of 250 NTU (expressed as a daily maximum value) for sites
required to implement passive treatment. In addition, all sites will be required to meet non-numeric
effluent limitations. Avoided costs from Option 4 range between $2.4  and $3.0 million, with a midpoint
estimate of $2.9 million. The requirements of Option 4 produce larger reductions in dredged sediment
than those of Option 2, as turbidity treatment is required on more sites. Option 4 is not as effective as
Option 3 in reducing sediment in navigable waters, since the two options have the same criteria for
disturbed acres, but Option 4 has a less stringent turbidity standard. The total reduction in sediment
dredged from navigable waters is expected to be 20.0 million cubic yards under Option 4.
November 2009                                                                                 7-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 7-6: Reductions in Dredging and Annualized Avoided Costs Under Option 1
EPA
Region
I1
2
3
4
5
6
1
8
9
10
Total
Reduction in Sediment Dredged
(thousands of cubic yards)
Low Mid High
0.3 1.0 1.3
16.4 21.3 24.2
40.7 51.5 55.0
871.5 1,001.7 1,058.3
21.0 31.5 50.9
5,901.7 6,660.6 6,739.7
5.2 9.3 10.6
0.0 0.0 0.0
56.4 81.6 95.5
486.8 614.8 795.1
7,400.0 8,473.3 8,830.6
Avoided Costs Using 3% Discount Rate
(thousands of 2008$)
Low Mid
$0.3 $0.8
$7.6 $10.4
$20.0 $25.0
$262.5 $320.0
$8.3 $12.7
$628.8 $739.1
$0.8 $1.6
$0.0 $0.0
$28.7 $41.3
$69.7 $106.2
$1,026.7 $1,257.2
High
$0.9
$10.9
$25.6
$329.8
$18.3
$769.7
$1.7
$0.0
$43.3
$131.8
$1,331.9
Avoided Costs Using 7% Discount Rate
(thousands of 2008$)
Low Mid
$0.3 $0.8
$7.4 $10.1
$19.7 $24.4
$258.1 $312.5
$8.2 $12.4
$611.5 $720.8
$0.7 $1.5
$0.0 $0.0
$28.2 $40.0
$68.3 $102.9
$1,002.4 $1,225.3
High
$0.9
$10.5
$24.9
$322.3
$17.9
$754.8
$1.6
$0.0
$41.9
$127.5
$1,302.4
 Table 7-7: Reductions in Dredging and Annualized Avoided Costs Under Option 2
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Reduction in Sediment
(thousands of cubic
Low
0.7
33.8
84.5
1,806.7
43.2
12,309.2
10.8
0.0
116.9
990.2
15,395.9
Mid
2.1
43.8
106.7
2,076.3
65.0
13,891.9
19.4
0.1
168.5
1,251.4
17,625.0
Dredged
yards)
High
2.7
49.9
114.0
2,193.3
105.1
14,056.6
22.0
0.1
197.4
1,618.7
18,359.8
Avoided
Low
$0.7
$15.6
$41.4
$543.1
$17.1
$1,311.4
$1.6
$0.0
$59.5
$141.8
$2,132.1
Costs Using 3% Discount Rate
(thousands of 2008$)
Mid
$1.7
$21.4
$51.9
$661.9
$26.2
$1,541.4
$3.2
$0.0
$85.4
$216.1
$2,609.2
High
$1.9
$22.4
$53.0
$682.2
$37.6
$1,605.1
$3.6
$0.0
$89.4
$268.1
$2,763.4
Avoided
Low
$0.7
$15.2
$40.8
$534.0
$16.8
$1,275.4
$1.5
$0.0
$58.5
$138.8
$2,081.7
Costs Using 7% Discount Rate
(thousands of 2008$)
Mid
$1.6
$20.7
$50.7
$646.4
$25.5
$1,503.1
$3.1
$0.0
$82.6
$209.4
$2,543.0
High
$1.8
$21.7
$51.7
$666.7
$36.8
$1,574.1
$3.4
$0.0
$86.6
$259.4
$2,702.2
November 2009
                                                                                                                 7-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 7-8: Reductions in Dredging and Annualized Avoided Costs Under Option 3
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Reduction in Sediment Dredged
(thousands of cubic yards)
Low
0.9
42.1
105.3
2,251.4
53.8
15,356.2
13.5
0.0
145.7
1,229.5
19,198.2
Mid
2.6
54.5
132.9
2,587.4
81.0
17,330.6
24.1
0.1
209.8
1,554.0
21,977.0
High

3.4
62
142
2,733
130
17,536
27
0
245
2,010
22,891
1
0
0
8
2
5
1
9
3
2
Avoided
Low
$0.8
$19.4
$51.6
$676.6
$21.2
$1,636.0
$2.0
$0.0
$74.2
$176.0
$2,657.7
Costs Using 3% Discount Rate
(thousands of 2008$)
Mid
$2.1
$26.7
$64.7
$824.5
$32.6
$1,922.9
$4.0
$0.0
$106.3
$268.4
$3,252.1
High
$2.3
$27.9
$66.0
$849.8
$46.9
$2,002.4
$4.5
$0.0
$111.3
$333.0
$3,444.1
Avoided
Low
$0.8
$18.9
$50.8
$665.2
$20.9
$1,591.0
$1.9
$0.0
$72.9
$172.4
$2,594.9
Costs Using 7% Discount Rate
(thousands of 2008$)
Mid
$2.0
$25.8
$63.1
$805.1
$31.8
$1,875.1
$3.8
$0.0
$102.9
$260.0
$3,169.7
High
$2.2
$27.0
$64.5
$830.4
$45.7
$1,963.7
$4.3
$0.0
$107.8
$322.1
$3,367.8
 Table 7-9: Reductions in Dredging and Annualized Avoided Costs Under Option 4
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Reduction in Sediment Dredged
(thousands of cubic yards)
Low
0.6
33.7
92.1
1,964.1
43.4
14,395.8
12.4
0.0
127.8
815.4
17,485.3
Mid
1.7
43.5
115.9
2,252.6
67.2
16,242.8
22.3
0.1
174.1
1,043.3
19,963.5
High
2
49
123
2,374
107
16,434
25
0
205
1,351

3
6
8
9
2
8
3
1
7
7
20,675.5
Avoided
Low
$0.6
$15.5
$45.2
$574.7
$16.9
$1,531.7
$1.8
$0.0
$65.1
$116.0
$2,367.5
Costs Using 3% Discount Rate
(thousands of 2008$)
Mid
$1
$21
$56
$699
$26
$1,800
$3
$0
$88
$178

4
3
5
0
1
0
7
0
3
8
$2,875.1
High
$1.6
$22.3
$57.7
$720.3
$37.2
$1,874.4
$4.2
$0.0
$92.6
$222.5
$3,032.8
Avoided
Low
$0.6
$15.1
$44.5
$565.1
$16.7
$1,489.6
$1.8
$0.0
$64.0
$113.6
$2,310.9
Costs Using 7% Discount Rate
(thousands of 2008$)
Mid
$1.3
$20.6
$55.2
$682.8
$25.5
$1,755.3
$3.5
$0.0
$85.6
$173.1
$2,802.9
High
$1.5
$21.6
$56.3
$704.1
$36.3
$1,838.3
$3.9
$0.0
$89.9
$215.1
$2,967.1
November 2009
                                                                                                                 7-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
7.6    Sources of Uncertainty and Limitations

Key limitations of EPA's analysis of cost savings to drinking water treatment facilities are outlined
below. The SPARROW model for suspended sediments that was used for estimating TSS concentrations
in the source waters also has a number of limitations. These limitations are discussed in detail in
Chapter 6.
       >  The USAGE dredging database identifies dredging jobs by name, which is usually the name
           of the waterbody dredged. However, the data lack standardized naming conventions, so it is
           possible that the same waterbody is dredged under different job names. This may result in the
           exclusion of dredging job names that only appear once in the database, but in fact were
           carried out in the same water bodies as a differently named job, which would result in a
           downward bias in EPA's dredging frequency calculations and the project costs.
       >  The navigable waterway data provide latitude/longitude information for some dredging jobs,
           which are used to link dredging jobs to RF1 reaches, but these data are incomplete. In cases
           where latitude/longitude information was not available for a particular job, EPA matched it to
           an RF1 reach using the job name. This is a potential source of inaccuracy, as the job name is
           often the waterway name, and may not be very specific (in cases such as the Mississippi or
           Colorado rivers). Additionally, if no RF1 reach could be identified for the dredging job, the
           average reduction for the state in which the job occurred was used. It is unclear whether this
           would lead to an over- or underestimate of benefits.
November 2009                                                                               7-13

-------
Environmental Impact and Benefits Assessment for the C&D Category
        Benefits to Water Storage
Reservoirs are water impoundments, often manmade, that serve many functions, including providing
drinking water, flood control, hydropower supply, and recreational opportunities. Sediment in streams can
be carried into reservoirs, where it may settle and build up layers of silt overtime. An increase in
sedimentation rates will reduce the useful life of a reservoir, unless measures are taken to reclaim some of
its capacity. Historically, the United States Geological Survey (USGS) has recorded an average of 1.2
billion kilograms of sediment deposition per reservoir each  year (USGS 2007c).
The regulation is expected to reduce the amount of sediment entering reservoirs and, as a result, the cost
of reservoir maintenance. Though there are multiple options for sediment mitigation, based on a literature
review EPA used the avoided cost of reservoir dredging as a measure of the benefits of the regulation to
water storage. This approach represents an appropriate measure of social benefits for cases in which a
policy change reduces costs to producers (in this case producers of water supplies for human use), but in
which price effects are minimal (cf. Boardman et al. 2001, p. 70-74). A description of other amelioration
measures can be found in Section  8.1.
The analysis of these benefits includes the following steps:
    > Estimating the unit cost of sediment removal from reservoirs
    > Estimating the sediment accumulation in  reservoirs under the baseline scenario and the post-
       compliance EPA policy options
    > Estimating the cost that would be incurred to dredge this accumulated sediment under each
       scenario and the avoided cost of reservoir dredging  expected from this regulation.
The remainder of this chapter provides details of this analysis.

8.1    Review of Literature on Reservoir Sedimentation

Research into reservoir sedimentation was conducted in the 1980s as part of a larger research effort into
the economic effects of sediment runoff from agricultural land use. As sediment has similar effects
regardless of its source, many of the insights from these studies will be applicable to sediment discharge
from construction activities.
Crowder (1987) estimated that the United States was losing about 0.22 percent of its reservoir capacity
each year due to sedimentation. Clark et al. (1985) notes that total U.S. reservoir capacity is filling up
slowly and has enough excess capacity dedicated to hold sediment build-up over hundreds of years.
However, the study goes on to conclude that while total reservoir sedimentation is manageable,
sedimentation is far from uniform and that in about  15 percent of U.S. reservoirs, sedimentation rates
exceeded 3 percent of capacity annually, and in the more extreme  cases, 10 percent per year (Clark et al.
1985). The Reservoir Sedimentation Handbook (Morris and Fan 1997), an engineering guide, details
several methods of counteracting the effects of sedimentation to maintain reservoir functionality:
       >  Sediment routing encompasses a group of techniques that seek to allow sediment-laden water
           to pass around or through the reservoir, not allowing the sediment to settle. All of these
           techniques involve timing sediment-laden flows from events such as storms and floods and
           operating the pass-through or bypass  mechanism optimally. Pass-through techniques may be
November 2009                                                                                 8-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
           implemented at existing reservoirs; however, the bypass mechanisms are most cost-effective
           if incorporated into the initial design of the reservoir. SPARROW does not take into account
           any of these techniques in its assumptions about sediment removed from reaches by
           reservoirs, EPA was not able to find any information on either the prevalence or effectiveness
           of these techniques.
        >  Flushing involves scouring accumulated sediment from the reservoir by partially or fully
           draining the reservoir and allowing the erosive force of the water to carry the sediment
           through the dam and  downstream of the reservoir. This does not appear to be a very common
           practice in the United States, as this creates extraordinarily high sediment concentrations
           downstream of the flushed reservoir, sometimes in excess of 1,000,000 mg/L and thus may
           require special environmental permission according to Morris and Fan (1997). Flushing is not
           one-hundred percent  effective at removing settled sediment, and the reservoir may still
           require dredging. The process also requires interrupting the service of the reservoir.
        >  Dredging is a practical and common approach to sediment removal once a reservoir has been
           built. Morris and Fan (1997) note that with the exception of draining a reservoir and
           excavating settled sediment (which is often more expensive and less common than dredging),
           dredging may be the  only feasible option for sediment removal in many reservoirs.  Dredging
           may be limited by available funds, as well as by the amount of space available to dispose of
           the dredged material. Due to the cost of transporting dredged sediment, a lack of nearby
           disposal sites may limit the amount of sediment that can be removed using dredging.
In a study focused on the economic costs of reservoir sedimentation due to agricultural runoff, Crowder
(1987) describes several options for mitigating reservoir sedimentation. These options  are:
    >   Including sediment pools in the initial construction of the reservoir to collect sediment entering
        the reservoir and maintaining these pools to increase the useful life of the reservoir
    >   Dredging the reservoir
    >   Replacing the lost capacity by expanding or replacing the reservoir.
According to Crowder (1987), constructing sediment pools at the outset is more than five times less costly
than dredging the reservoir. This  study estimates that reservoir capacity costs between  $564 and $1,316
(2008$) per acre-foot to construct (regardless of whether it is built as excess capacity for sediment or to
replace capacity lost from sedimentation), and that dredging costs $4,700 (2008$) per acre-foot of
material dredged. He estimates that only 20,000 (32.3 million cubic yards) acre-feet of sediment are
removed from reservoirs each year by dredging, while sediment pools in reservoirs accumulate  1 million
acre-feet (1.6 billion cubic yards) of sediment per year, and that new capacity would have to account for
the remaining 620,000 acre-feet (1 billion cubic yards) lost to sediment each year (Crowder 1987). His
estimate of total economic damages from reservoir sedimentation is $1.3 billion (2008$) per year, based
on an assumed $94 million (2008$) in annual dredging expenditures and the discounted cost of replacing
all remaining lost capacity in 10 to 20 years.
The analysis used by Crowder (1987) assumes that all reservoir sediment that is not dredged or
accumulated by sediment pools is accounted for with the construction of new capacity. This assumption
implies that additional reservoir capacity can be built unlimitedly to maintain the useful capacity of
reservoirs, an assumption that the author acknowledges is unrealistic. Although Crowder (1987) finds that
building excess storage capacity may  be more cost effective than dredging reservoirs, it may not be a
feasible option in many cases. Moreover, in comparing costs of reservoir dredging and building new
November 2009                                                                                  8-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
reservoir capacity, Crowder (1987) does not take into account ecological effects of new reservoir capacity
construction. The value of ecological services lost due to new reservoir construction may outweigh
avoided costs compared to reservoir dredging. The study does also not take into account any opportunity
costs of increasing the amount of land used for water storage. The objective of Crowder (1987) is to
quantify the impact of sedimentation on water storage  reservoirs by assuming that all capacity lost to
sedimentation that is not currently recovered through dredging will need to be replaced by new capacity.
While this approach is appropriate for a theoretical estimation of the cost impact of reservoir
sedimentation, EPA's objective is to quantify the avoided costs of reduced reservoir sedimentation due to
reduced construction site stormwater discharges, and thus only takes into account actual measures being
taken to reduce reservoir sedimentation (i.e., dredging). Therefore, this analysis assumes that excess
sediment is removed from reservoirs by dredging rather than building new reservoir capacity.

8.2     Data Sources

Information on reservoir locations, uses, and historic sedimentation rates is available from the  Reservoir
Sedimentation Survey Information System (RESIS) database (USGS 2007b). Formally known as RESIS-
II, this database was initially developed by the U.S. Department of Agriculture Natural Resources
Conservation Service, and was provided to EPA by USGS.
The database includes 4,227 sedimentation survey results from  1,819 reservoirs located across the
coterminous United States. Results from 1775 to 1992 are represented, with the  majority of measurements
covering the mid-20th century and on. RESIS does not include all U.S. reservoirs, nor does it contain
information about dredging  activity in reservoirs or about any other methods for recapturing lost water
storage capacity.
Table 8-1 details sediment deposition per year recorded by RESIS in each  EPA region. More than 1.2
billion kilograms of sediment are deposited in the average U.S.  reservoir each year. The most sediment
deposition is recorded in Region 6 with 6.6 billion kilograms per reservoir per year, almost four times
more than the national average. Region 9 also reports large masses of sediment deposition each year of
1.8 billion kilograms per reservoir, which is still greater than the national average. Other regions have
average annual deposition per reservoir ranging from 800,000 kilograms in Region 1 to 440 million in
Region 4.
Table 8-1: Average Annual Sedimentation Deposition
Recorded by RESIS
EPA Number of
Region Reservoirs
1 9
2 18
3 81
4 168
5 272
6 338
7 329
8 208
9 285
10 110
Total 1,818
Average Annual Sediment Deposition
per Reservoir (million kg/year)
0.8
25.1
71.5
440.4
25.9
6,568.9
57.2
436.0
1,817.6
67.2
1,234.0
Source: RESIS (USGS 2007b).
November 2009                                                                                  8-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
Changes in sediment attenuation in reservoirs will be predicted by the SPARROW model, the details of
which can be found in Chapter 6.

8.3     Estimating the Unit Cost of Sediment Removal from Reservoirs

Because RESIS does not provide information on reservoir dredging, EPA assumed for the purpose of this
analysis that the average unit cost of dredging sediment from a navigable waterway is approximately
equal to the average unit cost of dredging sediment from a reservoir. Crowder (1987) does provide an
estimate of reservoir dredging cost per unit of sediment dredged, but gives no empirical basis for this
estimate. Because the USAGE dredging database provides costs and quantity dredged for more than 3,300
dredging jobs in the past 12 years, this unit cost of dredging has more robust empirical foundations. EPA
does not expect there to be any significant difference between the costs of reservoir dredging and the
costs of navigable waterway dredging, as all of the variable costs for dredging are related to the quantity
of sediment dredged. Morris and Fan (1997) cite a unit cost of $4.45 (2008$) per cubic yard for sediment
dredging at Lake Springfield, Illinois, which is quite close to the $4.41 average cost per cubic yard for
EPA Region 5 in the navigational dredging data. Section 7.3  provides details on estimating the unit cost
of navigable waterway dredging that will be used as a surrogate for the unit cost of reservoir dredging.
This analysis will use average costs at the regional level.

8.4     Estimating the Total Cost of Reservoir Dredging Under Different Policy
        Options

8.4.1   Estimating Sediment Accumulation in Reservoirs and the Amount of Sediment
        Expected To Be Dredged

EPA estimated the annual sediment attenuation rate in each reservoir under the  baseline and the three
policy options using output data from the SPARROW model. Chapter 6 provides a description of the
model. This analysis uses data output from SPARROW that predicts the amount of sediment (in
kilograms) settling in reservoirs in a reach. To estimate the amount of sediment flowing into and settling
in reservoirs in each region, EPA summed the values in this  field under the baseline scenario and each
policy option.:: The amount of sediment settling in reservoirs under each scenario forms the basis for the
estimation of the quantity of sediment dredged from reservoirs and the costs of this dredging.
For the purposes of this analysis, EPA  assumed, consistent with Crowder (1987), that all sediment
entering reservoirs must be removed in order to maintain the current water storage  capacity in the United
States. The frequency of dredging is highly site-specific, depending on many factors including the
average sediment concentration of the river or stream, the size of the reservoir and  excess storage
capacity, and any sediment routing practices. For this analysis, EPA chose a general frequency of
reservoir dredging based on information presented by the USAGE in a Final Dredged Material
Management Plan and Environmental Impact Statement for reservoirs in Washington (USAGE 2002).
This report states that "dredging cycles may vary from 2 to 10 years" (USAGE  2002, p. 66). EPA used
these frequencies as high and low estimates, and employed their arithmetic mean of 6 years for a midpoint
11   EPA conducted a preliminary analysis of baseline costs including and excluding outliers, and found that the exclusion of
    outliers does not have a significant effect on the total cost of reservoir dredging. The values used for this analysis include
    reservoir sedimentation for all reservoirs on reaches modeled by SPARROW.
November 2009                                                                                8-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
estimate. This approach provides a range of benefits estimates to account for uncertainty in the frequency
of reservoir dredging.

8.4.2   Estimating the Total Cost of Reservoir Dredging Under the Baseline Scenario and
        Policy Options
The cost of dredging the sediment settling out in reservoirs is estimated by using the average dredging
cost per region as described in Section 8.3. Because this cost is given per cubic yard, the sediment
attenuation in reservoirs given by SPARROW will be converted from kilograms to cubic yards using a
sediment density of 1.5 g/cm3 (Hargrove 2007). This translates to a conversion of 1,147 kilograms per
cubic yard. The equation below summarizes the calculation of costs for one cycle of dredging before
discounting and annualization.
                                          10
                              TC       =Y
                                 reservoirs   / j
                                          R=l
                                                    1,147
(Eq. 8-1)
       Where:

       TCreservoirs= total cost of dredging all sediment settling in reservoirs in all regions

       R       = region number

       /        = the assumed interval in years of reservoir dredging; varied between 2, 6, and 10

       SR       = sum of all reservoir sediment settlement predicted by SPARROW in a region in
                  kilograms

       1,147    = kilograms in a cubic yard

       CR      = regional average of historic dredging job cost per cubic yard, 1995-2006

The resulting costs were discounted and annualized over the assumed interval using both 3 and 7 percent
discount rates, in accordance with OMB Circular A-4 (OMB 2003).
Table 8-2 presents the total amount of sediment that is estimated to be dredged per year and the estimated
cost of this dredging under the baseline scenario, including low, midpoint, and high estimates. The 445
million cubic yards predicted by SPARROW to settle annually in reservoirs is expected to cost between
$1.5 and $1.7 billion to dredge under the baseline scenario. Region 8 has the highest dredging costs, and
though less sediment is predicted to accumulate in Region 8 reservoirs than in those in Region 6, the
former has a much higher unit cost of dredging. Regions 1, 2, and 3 are all predicted to accumulate less
than 6 million cubic yards of sediment per year each, and accordingly are estimated to have  lower overall
dredging costs. For other regions, midpoint estimates of dredging costs range between $50 and $526
million.
November 2009                                                                                 8-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 8-2: Estimated Cost of Reservoir Dredging Under the Baseline Scenario
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Sediment Dredged (yd3)
1,278,280
2,725,285
5,337,766
45,666,958
35,940,396
128,381,390
42,237,651
126,434,846
41,859,743
14,743,876
444,606,190
Estimated
Low
$11.7
$17.1
$32.3
$218.0
$147.1
$201.3
$96.0
$495.0
$236.2
$46.9
$1,501.6
Cost (millions
Mid
$12.4
$18.2
$34.4
$231.8
$156.4
$214.0
$102.1
$526.4
$251.2
$49.9
$1,596.8
of 2008$)
High
$13.2
$19.3
$36.5
$246.2
$166.2
$227.3
$108.4
$559.1
$266.8
$53.0
$1,696.0
8.5     Estimating the Avoided costs from Reduced Reservoir Dredging

The difference between the anticipated dredging costs under the baseline and a particular policy option
represents the avoided costs of that policy option. Table 8-3, Table 8-4, Table 8-5, and Table 8-6 present
reductions in sedimentation and subsequent avoided costs from reduced reservoir dredging for each
policy option, including low, midpoint, and high estimates under these options. Because the range of
estimates is relatively small between the low and high estimates, the values presented below are midpoint
estimates unless otherwise stated. Benefits were estimated assuming both 3 and 7 percent discount rates,
consistent with OMB Circular A-4 (OMB 2003).
Avoided costs from a reduction in reservoir sedimentation range from $1,072 (Option 1, low estimate, 7
percent discount rate) to $3.6 million (Option 3, high estimate, 3 percent discount rate), depending on the
policy option, the assumed frequency of reservoir dredging, and the discount rate. The largest savings are
predicted in Region 6 under all options, as SPARROW predicts the largest overall reductions in sediment
accumulation in this region. Region 4 is also expected to benefit substantially relative to other regions due
to both large reductions in construction discharges in this region and a relatively high unit cost of
dredging (estimated from  USAGE data).
Option 1, which requires non-numeric effluent limitations for all sites, is EPA's least stringent policy
option. This  option is expected to reduce reservoir sedimentation by about 560 thousand cubic yards
nationally every year, which is estimated to save between $1.2 (7 percent discount rate) and $1.4 million
(3 percent discount rate) in dredging costs per year.
Option 2 requires ATS on sites with more than 30 acres disturbed at one time and imposes a 13 NTU
turbidity standard while requiring non-numeric effluent limitations on all sites and is similar to the option
EPA proposed previously. This action is estimated to prevent about 1.2 million cubic yards of sediment
from building up in reservoirs each year saving between $2.6 (at a 7 percent discount rate) and $2.9 (at a
3 percent discount rate) million in dredging costs per year.
Option 3, EPA's most stringent policy option, requires ATS on sites with more than 10 acres disturbed at
one time, imposes a 13 NTU turbidity standard on these sites, and requires non-numeric effluent
limitations on all sites. Its reduction in sediment deposition in reservoirs is estimated to be 1.5 million
cubic yards annually. The estimated avoided costs of reduced dredging are between $3.2 and $3.6 million.
November 2009
                                                                                             8-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
Option 4 requires passive treatment systems on all sites with 10 or more acres disturbed at one time, and
establishes a numeric turbidity standard of 250 NTU (expressed as a daily maximum value) for sites
required to implement passive treatment. In addition, all sites will be required to meet non-numeric
effluent limitations. Because Option 4 requires passive rather than active treatment of sediment and has a
higher turbidity standard, the estimated reductions in reservoir sediment buildup are lower than those
from Option 3. However, the estimated reductions from Option 4 are still greater than those expected
from Option 2, as that option requires treatment on fewer sites. Expected avoided costs from Option 4 are
estimated to be $2.9 million assuming a 7 percent discount rate, and $3.2 million assuming a 3 percent
rate. Regions 4 and 6 together account for more than half of the monetized benefits of reduced reservoir
sedimentation, due to the expected reductions in construction discharges in these regions. Option 4 is
estimated to reduce reservoir sedimentation by 1.3 million cubic yards annually.
 Table 8-3: Reduction in Reservoir Dredging and Avoided Costs Under Option 1
EPA
Region
1
2
o
3
4
5
6
7
8
9
10
Total
Annual
Reduction in
Sediment yd3)
1,423
535
2,554
98,832
7,878
408,507
24,768
2,165
4,833
7,293
558,788
Avoided Costs (thousands of 2008$)

Low
$13.0
$3.4
$15.5
$471.7
$32.2
$640.5
$56.3
$8.5
$27.3
$23.2
$1,291.5
3% Discount
Mid
$13.8
$3.6
$16.5
$501.6
$34.3
$681.1
$59.8
$9.0
$29.0
$24.7
$1,373.4
Rate
High
$14.7
$3.8
$17.5
$532.8
$36.4
$723.4
$63.6
$9.6
$30.8
$26.2
$1,458.7
7% Discount
Low Mid
$10.8 $12.5
$2.8 $3.2
$12.8 $14.9
$391.4 $453.6
$26.8 $31.0
$531.4 $615.9
$46.7 $54.1
$7.0 $8.2
$22.6 $26.2
$19.3 $22.3
$1,071.6 $1,241.9
Rate
High
$14.4
$3.7
$17.1
$522.5
$35.7
$709.4
$62.3
$9.4
$30.2
$25.7
$1,430.5
 Table 8-4: Reduction in Reservoir Dredging and Avoided Costs Under Option 2
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Annual
Reduction in
Sediment (yd3)
2,876
1,096
5,278
205,148
16,292
851,211
51,568
4,476
9,992
14,804
1,162,741
Avoided Costs (thousand 2008$)

Low
$26.3
$6.9
$32.0
$979.1
$66.7
$1,334.6
$117.2
$17.5
$56.4
$47.1
$2,683.8
3% Discount
Mid
$28.0
$7.3
$34.0
$1,041.2
$70.9
$1,419.2
$124.6
$18.6
$60.0
$50.1
$2,853.8
Rate
High
$29.7
$7.8
$36.1
$1,105.8
$75.3
$1,507.4
$132.3
$19.8
$63.7
$53.2
$3,031.2

Low
$21.8
$5.7
$26.5
$812.4
$55.3
$1,107.4
$97.2
$14.5
$46.8
$39.1
$2,226.8
7% Discount
Mid
$25.3
$6.6
$30.7
$941.5
$64.1
$1,283.3
$112.7
$16.8
$54.2
$45.3
$2,580.6
Rate
High
$29.1
$7.6
$35.4
$1,084.5
$73.9
$1,478.2
$129.8
$19.4
$62.5
$52.2
$2,972.6
November 2009
                                                                                           8-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 8-5: Reduction in Reservoir Dredging and Avoided Costs Under Option 3
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Annual
Reduction in
Sediment (yd3)
3,567
1,363
6,573
255,706
20,293
1,061,735
64,312
5,575
12,446
18,375
1,449,945
Avoided Costs (thousand 2008$)

Low
$32.6
$8.5
$39.8
$1,220.4
$83.1
$1,664.7
$146.1
$21.8
$70.2
$58.5
$3,345.8
3% Discount
Mid
$34.7
$9.1
$42.3
$1,297.7
$88.3
$1,770.2
$155.4
$23.2
$74.7
$62.2
$3,557.9
Rate
High
$36.9
$9.7
$45.0
$1,378.4
$93.8
$1,880.2
$165.0
$24.7
$79.3
$66.1
$3,778.9
7% Discount
Low Mid
$27.1 $31.4
$7.1 $8.2
$33.0 $38.3
$1,012.6 $1,173.5
$68.9 $79.9
$1,381.2 $1,600.7
$121.2 $140.5
$18.1 $21.0
$58.3 $67.5
$48.5 $56.2
$2,776.1 $3,217.2
Rate
High
$36.1
$9.5
$44.1
$1,351.7
$92.0
$1,843.8
$161.9
$24.2
$77.8
$64.8
$3,705.9
 Table 8-6: Reduction in Reservoir Dredging and Avoided Costs Under Option 4
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Annual
Reduction in
Sediment (yd3)
2,126
1,022
5,497
226,512
17,172
984,437
59,072
4,697
10,505
11,828
1,322,867
Avoided Costs (thousand 2008$)

Low
$19.4
$6.4
$33.3
$1,081.1
$70.3
$1,543.5
$134.2
$18.4
$59.3
$37.6
$3,003.6
3% Discount
Mid
$20.7
$6.8
$35.4
$1,149.6
$74.7
$1,641.3
$142.7
$19.6
$63.0
$40.0
$3,193.9
Rate
High
$22.0
$7.2
$37.6
$1,221.0
$79.4
$1,743.3
$151.6
$20.8
$67.0
$42.5
$3,392.4
7% Discount
Low Mid
$16.1 $18.7
$5.3 $6.2
$27.6 $32.0
$897.0 $1,039.5
$58.3 $67.6
$1,280.7 $1,484.2
$111.4 $129.1
$15.3 $17.7
$49.2 $57.0
$31.2 $36.2
$2,492.1 $2,888.1
Rate
High
$21.5
$7.1
$36.9
$1,197.4
$77.9
$1,709.6
$148.7
$20.4
$65.7
$41.7
$3,326.8
8.6    Sources of Uncertainty and Limitations

Key limitations of EPA's analysis of avoided costs to drinking water treatment facilities are outlined
below. The SPARROW model for suspended sediments that was used for estimating TSS concentrations
in the source waters also has a number of limitations. These limitations are discussed in detail in
Chapter 6.
   >  The reliance of the SPARROW model on the RF1 network is a significant limitation  because the
       reservoirs located off the RF1 network are omitted from this analysis. This omission  is likely to
       bias the estimated savings from reduced reservoir dredging downward.
   >  There is uncertainty as to the uniformity of sediment density because it is related to the type of
       soil in the area. Using a single density to convert volume to weight for all sediment may reduce
       the accuracy of the resulting cost estimates. However, the direction of this potential bias is
       uncertain.
November 2009
                                                                                         8-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
    >  The lack of data on reservoir dredging results in significant uncertainty as to the types of
       reservoirs that are dredged and the cost of this dredging. The appropriateness of the avoided cost
       approach is conditional upon the estimates representing actual dredging costs that are no longer
       imposed on reservoirs. If typical reservoirs do not reduce dredging, and hence costs, but rather
       undertake other actions in response to reduced sediment loads, then actual benefits may diverge
       from estimates provided  above.
    >  Neither RESIS nor the SPARROW model take into account sediment build-up in natural water
       storage facilities such as  glacial lakes and ponds, so any activity to mitigate sedimentation of
       these areas is not included in this analysis, resulting in a potential downward bias of this estimate.
November 2009                                                                                   8-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
       Benefits to  Drinking Water Treatment
Drinking water treatment plants provide the essential services of removing contaminants from water taken
from surface and ground sources, making the water potable. Sediment is one of the contaminants that
must be removed from drinking water before it can be consumed, so a reduction in sediment levels in
water taken up by drinking water treatment plants is expected to result in lower treatment costs for these
plants.  This section describes the methodology EPA used to estimate the reduction in drinking water
treatment costs that should result from a reduction of sediment discharge from construction sites. This
approach represents an appropriate measure of social benefits for cases in which a policy change reduces
costs to producers (in this case producers of drinking water), but in which price effects are minimal (cf
Boardman et al. 2001, p. 70-74).

9.1    Construction Discharge Effects on Drinking Water Treatment Costs

9.1.1   Sediment

The primary concern of sediment in terms of drinking water treatment is the turbidity that results from
suspended sediment in the water column. As turbidity increases in drinking water influent, the variable
costs of water treatment to mitigate this turbidity increase.
The variable costs associated with treating drinking water for turbidity are twofold:
       >  Adding chemicals coagulants that bond to the sediment particles and help them sink out of
           the water column, and
       >  Disposing  of the sludge which results from this treatment.
Several attempts have been made in the past  two decades to isolate these elements of drinking water
treatment costs:
       >  Forster et al. (1987) estimated $172 (2008$) per million gallons for drinking water sediment
           treatment, while Holmes (1988)  performed several estimates, with ranges between $7.45 and
           $212 (2008$) per million gallons. It is unclear in these studies what elements were included
           in these estimates.
       >  Moore  and McCarl (1987) estimated a total of $41 (2008$) per million gallons to treat water
           and dispose of the sludge for drinking water treatment plants in Oregon.
       >  Dearmont et al. (1998) put the chemical costs of turbidity treatment alone at $96 (2008$) per
           million gallons for Texas drinking water.
This broad range of cost estimates for sediment and turbidity treatment may be due to the fact that the
studies measure costs per million gallons of water, without respect to the specific sediment levels in
drinking water influent.

9.1.2  Nutrients

In addition to removing sediment from drinking water, treatment plants must also meet standards for
certain nutrients. Nitrogen and phosphorus are the two primary nutrients of concern in construction
November 2009
                                                                                           9-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
discharges. While there is no EPA standard for phosphorus levels in drinking water, nitrogen is regulated
by EPA with maximum contaminant levels (MCLs) of 10 mg/L for nitrate and 1 mg/L for nitrite; total
nitrogen (nitrate + nitrite) is limited to 10 mg/L (USEPA undated). DeZuane (1997) cites EPA surface
water surveys that found that only 1.2 percent of surveyed supplies had total nitrogen concentrations
exceeding 10 mg/L.
EPA recommends ion exchange, reverse osmosis, and electrodialysis to remove nitrogen from source
water (USEPA undated). Reverse osmosis and electrodialysis are systems with operating costs related to
the volume of water treated rather than the amount of contaminants, while ion exchange does have the
variable cost of the ion exchange resin to which the contaminant bonds. Therefore, drinking water
treatment facilities that use reverse osmosis and electrodialysis incur no additional cost associated with
nitrogen removal. Although facilities that use ion exchange may incur operating costs associated with
nitrogen removal, the process is also used to remove many other contaminants, and estimating the fraction
of variable cost attributable to nitrogen removal is not feasible within the national-level analysis.
Benefits from reduced nitrogen loadings to drinking water source reaches would only be realized if their
nitrogen loadings are  above the MCL for nitrate  and nitrite. Table 9-1 presents total nitrogen (TN)
concentrations as predicted by SPARROW for the reaches in the drinking water treatment benefits
analysis. As shown, the top 4 percent of these  reaches are predicted to have concentrations above the 10
mg/L threshold. However, because EPA's methodology treats values above the 95th percentile as outliers,
benefits for these reaches would not be able to be monetized.

            Table 9-1: TN Concentrations Predicted by SPARROW for Reaches
            Serving as Drinking Water Sources1
Distribution of TN Concentrations
Minimum
25th percentile
50th percentile
75th percentile
95th percentile
96th percentile
Maximum
Average
TN Concentration
(mg/L)
0.0
0.9
1.3
2.4
8.0
10.0
2,503.7
4.8
            Only includes reaches that are both modeled by SPARROW and included in the federal drinking water
            database SDWIS, see Section 9.5 for details.
9.1.3  Increase in Primary Productivity and Associated Water Quality Impacts from
       Nutrient Discharges (Eutrophication)
In addition to the meeting the nitrogen standards discussed above, indirect construction-related nutrient
impacts can result in the development of harmful algal blooms (HABs) in surface drinking water
reservoirs. These algal blooms, particularly those comprised of blue-green algae (Cyanobacteria), can
pose potential health threats and aesthetic problems for the reservoir water's potability. These include the
production of taste and odor compounds, production of cyanotoxins, and the potential for increase
disinfection by-product formation.
Cyanobacteria and actinomycetes produce the organic compound geosmin, which is typically released on
the death or lysis of the cell. This compound is responsible for the muddy smell and unpleasant taste in
drinking water that originates from surface water bodies experiencing algal blooms. The human sensory
November 2009                                                                                 9-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
system is highly sensitive to geosmin. Studies show that the nose can detect this compound at
concentrations as low as 5 parts per trillion and that taste can detect geosmin at 0.7 parts per billion. The
second taste and odor compound found in drinking water is 2-methylisoborneol (or MIB). It is also
produced by blue-green algae and detected by humans at very low levels of concentration. Together, these
two compounds are responsible for the taste and odor associated with drinking water treatment, which can
be quite strong in the presence of algal bloom. Taste and odor problems associated with drinking water
are a pervasive and significant problem for many municipalities. In 1996, a survey of 377 American and
Canadian water utilities conducted by the American Water Works Association reports that 43 percent of
the utilities had a taste and odor problem lasting more than one week and 16 percent consider their
drinking water taste and odor problem to be serious (Suffet et al.  1996).
Large concentrations of cyanobacteria can also be problematic because of the harmful toxins they  can
produce. Cyanobacterial toxins can be either found inside of the cell, "intracellular," or outside the cell,
"extracellular". The cyanotoxins can cause adverse impacts including death not only to aquatic organisms
that come in direct contact with them but also to livestock, domestic animals, waterfowl, and in some
cases to humans (WHO 1999). These toxins belong to a very diverse group of chemical substances that
can be classified according to their biological effects on the systems and organs that they affect most
strongly including: hepatotoxins, neurotoxins, cytotoxins, irritants and gastrointestinal toxins, and  other
cyanotoxins whose toxicological or ecotoxicological profile is still only partially known (Funari and
Testai 2008).
Additionally, eutrophication can lead to an increase in organic carbon (both particulate and dissolved) in
the waterbody. This additional carbon can result in the increased formation of trihalomethanes, a
dangerous by-product arising from disinfection of the waterbody with a halogen-containing compound.
The chlorine or bromine in the disinfectant reacts with organic compounds  in the water to form these tri-
substituted methanes that are considered environmental pollutants and are often carcinogenic to humans.
Currently, the EPA has set the limit of this total trihalomethanes at 80 ppb in treated water (USEPA
1998).
Common water treatment methods include filtration, softening, reverse osmosis, chlorination, and
distillation (Parsons and Jefferson 2006). Other, more specialized treatments include using powdered
activated carbon (PAC) to absorb pollutants, and ozonation. Using PAC treatment requires activated
carbon to be added to an aerobic or anaerobic treatment system where it serves as the "buffer" against
toxics in the wastewater. To treat geosmin and MIB with this method typically requires large doses of
activated carbon, and thus can be impractical, particularly for large facilities. Ozonation uses radical
formation to eliminate organic particles via coagulation or chemical  oxidation. However, like PAC
treatment, this method is expensive especially for large treatment facilities; it also can add harmful bi-
products to the water. Ozonation has stronger germicidal properties than chlorination however because it
rapidly reacts with bacteria, viruses, and protozoans.
The city of Waco, Texas is constructing a water clarification facility as part of the second phase of the
Water Quality and Quantity project as a result of a comprehensive five-year study of Lake Waco which
identified a significant cyanobacteria population that was dominating the lake as a result of nutrient
loading from the watershed. This abundance of blue-green algae further exacerbated the pre-existing taste
and odor issues that the reservoir already faced. This new facility, with a construction  cost of $40 million
dollars, proposes to improve the taste of the finished water by implementing Dissolved Air Flotation
(DAF) technology (Waco Water Utilities Services 2009). DAF works by attaching air bubbles to particles
suspended in water and floating them to the surface for tank collection. This technology has proven
November 2009                                                                                  9-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
successful in other systems and is particularly effective in waters with a significant amount of lightweight
particles (such as algae).
The effectiveness of a specific drinking water treatment for the removal of cyanobacterial toxins depends
on the total concentration of the algal toxins, the form of the toxin, and whether it is intracellular or
extracellular (Westrick 2008). Intracellular cyanotoxins can be removed by conventional treatment and
membrane filtration. Extracellular cyanobacterial toxins are more costly to remove because conventional
treatment (e.g., flocculation, coagulation, sedimentation and filtration) is usually not effective and
advanced treatment processes must be implemented unless the contaminant is oxidized through
disinfection (Westrick 2008).
Due to data limitations, EPA  was unable to estimate changes in drinking water treatment costs associated
with improvements in taste and odor of source waters resulting from reduced sediment and nutrient
discharges from construction sites.

9.2    Avoided Costs of Drinking Water Treatment from Reducing Sediment
       Discharge from Construction Sites

This section describes the estimation of the total expenditures to remove sediments from drinking water
and the avoided costs expected with the reduction of sediment discharge anticipated from the regulation.
It involves the following steps:
       >  Identifying RF1 reaches modeled by SPARROW that are sources for drinking water
           treatment plants
       >  Determining TSS reductions in these reaches
       >  Estimating the chemical cost of treating the turbidity caused by TSS in these reaches
       >  Estimating the cost of disposing  of the sludge generated from this turbidity treatment
       >  Estimating the total costs of drinking water treatment
       >  Estimating the total avoided costs between the baseline and post-compliance scenarios.
Error! Reference source not found, outlines these steps in more detail. The ovals at the top represent
primary data sources. The gray boxes contain calculations or conversions, and the white boxes contain the
results of these calculations or conversions that serve as input for the next step in the calculation. The left-
hand side of the diagram shows the calculation of the chemical cost of treating turbidity, while the right-
hand side shows the calculation of the cost of sludge disposal. Finally, the pale yellow boxes at the
bottom of the figure represent the costs under baseline and post-compliance scenarios and the difference
between them.
Sediment concentrations and  drinking water influent volumes were estimated for each surface drinking
water intake in the United States. EPA calculated the costs of treatment chemicals and sludge disposal by
surface drinking water intake; the specific steps are described in more detail in the following sections.
Costs under the baseline and post-compliance scenarios were compared, and the difference represents the
national avoided costs. To address uncertainty in its assumptions, EPA conducted a sensitivity analysis
that varies the assumptions described in the following sections, and the results of the following analyses
present low, midpoint, and high benefits estimates (see Section 9.8 for details).
November 2009                                                                                 9-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
     Figure 9-1: Steps in Calculating the Cost of Treating Turbidity in Drinking Water
         SPARROW results: TSS
          concentrations by RF1
         reach under baseline and
         post-compliance scenarios
         TSS—Turbidity conversion
                                                  SDWIS: Drinking water
                                                  influent volumes by RF1
                                                           reach
   Influent turbidity under baseline and post
            compliance scenarios
                    I
      Alum dose—Turbidity relationship
                    I
   Chemical usage under baseline and post-
           compliance scenarios
                    I
                                                      Sludge generation equation
               Alum unit cost
                                                            I
                                           Sludge generation under baseline and post-
                                                    compliance scenarios
Chemical costs under baseline and post-
        compliance scenarios
                                                                I
                                                    Probability of off-site disposal
                                                                I
                                                Sludge disposal under baseline and post-
                                                        compliance scenarios
                                                     Distance to disposal facility
                                                Sludge disposal costs under baseline and
                                                      post-compliance scenarios
                                                                _T
1
r
Total treatment costs by RF 1 reach under
baseline scenario

1 r
Total treatment costs by
post-compliance

RF 1 reach under
scenario
                   T
                                                             J
                  Reduction in total drinking water treatment costs by RF1
                     reach from baseline to post-compliance scenarios
November 2009
                                                                                           9-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
9.3    Data Sources

Information on drinking water intake points and volumes were provided by EPA's Safe Drinking Water
Information System (SDWIS), Federal Version. The federal version of SDWIS stores the information on
approximately 160,000 public water systems used to monitor the quality and safety of drinking water
(USEPA 2006b). For this analysis, EPA was only interested in surface water intakes, as these are the only
type that would be affected by discharge from construction sites. Location information and average daily
flow were estimated from the latitude-longitude and population data provided by SDWIS, respectively.
EPA obtained data on the chemicals used to treat turbidity based on data from the American Society  of
Civil Engineers (ASCE) and the American Water Works Association (AWWA). These data also contain a
formula for the calculation of sludge disposal costs (ASCE/AWWA 2005).
Estimations of coagulant dosing for various turbidity levels were taken from research reports by Cornell
University's AguaClara Project. This project focuses on improving drinking water quality through
engineering research and offers calculations and other engineering information for both researchers and
professionals in the field  (Menendez et al. 2007).
The costs of alum and alternative coagulant inputs were obtained by EPA for the Long Term 2 Enhanced
Surface Water Treatment Rule. The Technologies & Costs Document for this rule included research on
the expenses of public water systems and potential changes resulting from the rule (USEPA 2005c). For
this purpose, many details of the costs of drinking water treatment were obtained. This document reports
cost data in 1998 dollars, so its data are be adjusted to 2008 dollars.
The equation used to estimate the volume of sludge generated from turbidity and TSS treatment comes
from the EPA document Identification and Characterization of Typical Drinking Water Treatment
Residuals (USEPA  2007a), which examines the byproducts of drinking water purification from different
methods of treatment.
Calculations of the percentage of facilities that dispose of any of their residual sludge off-site and of the
percentage of total residual sludge generated that is disposed of off-site are obtained from the Drinking
Water Treatment Plant Master Survey Response Database (henceforth the Drinking Water Survey
Database) prepared by EPA for the Drinking Water Effluent Limitation Guidelines rule (USEPA 2007b).

9.4    Modeling Sediment Concentrations and  Reductions

EPA modeled sediment concentrations in RF1 stream reaches under both the current and final scenarios
using the SPARROW model. A detailed description of this model can be found in Chapter 6.
Drinking water treatment facilities located on waterbodies with typically higher levels of TSS and
turbidity may have systems in place that reduce total suspended solids and turbidity levels in influent
water before it is chemically treated. EPA assumed, as per Ngo and Nichols (1998), that facilities with
baseline TSS concentrations exceeding 1,000 mg/L would employ a process such as plain sedimentation,
during which influent water is allowed to settle for extended periods of days or weeks. The effectiveness
of this process depends on the size of the sedimentation basin and the amount of time the water is left to
settle. Studies by Yao (1973) and DeZuane (1997) suggest that the effectiveness of plain sedimentation
ranges from 30 and 90 percent with a midpoint estimate of 60 percent. Thus, EPA varied the effectiveness
November 2009                                                                                9-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
of plain sedimentation basins between 90, 60, and 30 percent TSS removal for the low, mid, and high
benefit estimates, respectively.12
This reduction is applied to influent water TSS concentrations at facilities that are estimated to have
influent water with TSS concentrations exceeding 1,000 mg/L in the baseline. EPA deems this approach
reasonable, as it brings treatment costs within the range of values published in existing literature
presented in Section 9.1.

9.5    Identifying Reaches with Drinking Water Intakes and Their Intake Volumes

To estimate the avoided costs of drinking water treatment resulting from the regulation, EPA used the
drinking water intake database compiled by USGS. The database provides information on public water
systems, including intakes, treatment plants, sampling sites, distribution systems, wells, and consecutive
connections from EPA's SDWIS database. Each drinking water intake is matched to the RF1 reach and
has the corresponding "USGS River Reach Number." The database, however, does not provide
information on the intake volume. For this analysis, EPA approximated each intake volume based on the
population served by each drinking water system and the number of intakes corresponding to that system.
The population served by a particular drinking water system was divided by the number of intakes in that
system to obtain the average population served per intake. From these data the daily flow of each intake
was estimated using the following equation determined by EPA for the Revised Arsenic Rule (USEPA
2000a):

                                      F = 0.067 *P1'062                               (Eq. 9-1)

       Where:

       F = average daily flow in thousands of gallons per day

       P = population served

Table 9-2 summarizes public surface drinking water intakes by EPA region and the total estimated daily
flow for all intakes in that region. Overall, public water systems in the United States withdraw nearly 18
billion gallons of water per day from surface water sources, an average of 2.8 million gallons per day per
intake.
12      The more effective the plain sedimentation process, the lower the treatment costs to the affected
facilities, and consequently, the lower their potential savings from reduced influent turbidity.
November 2009                                                                                9-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
                     Table 9-2: Public Water Intakes and Estimated
                     Average Daily Flow by EPA Region
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Source: SDWIS,
Number of Public
Water Intakes
639
478
881
868
525
821
353
549
748
487
6,349
Federal Version (USEPA
Total Estimated Daily
Flow (million
gallons/day)
1,105
2,286
2,301
2,913
2,373
2,309
715
701
2,543
560
17,809
2006c).
9.6    Estimating the Cost of Chemical Turbidity Treatment

Reducing sediment concentrations in drinking water influent is expected to reduce the turbidity level of
the water, requiring smaller doses of chemical coagulants to treat. EPA estimated the amount of
coagulants needed to treat a given level of turbidity according to the following steps:
       >  Converting sediment concentrations (given in mg/L) into nephelometric turbidity units
           (NTU) using an equation determined by ASCE/AWWA (2005):

                                                                                      (Eq.9-2)


        Where:

       T      = turbidity (in NTU)

       TSS   =  TSS  (in mg/L)

       b      = 0.8, 1.5, and 2.2 for the low , midpoint and high estimates,(as the value of b decreases,
              a given level of TSS generates more turbidity, leading to higher treatment costs).

       >  Determining the doses of alum (the primary coagulant used to treat turbidity) required to treat
           this turbidity level using relationships determined by researchers of the AguaClara Project
           (Menendez et al. 2007).
       >  Multiplying this coagulant dose (given per volume of water treated) by the total volume of a
           drinking water intake to obtain the total coagulant usage for water from that intake.
       >  Multiplying this amount of coagulant by the unit cost of the alum.
For this analysis, EPA assumed that alum is the coagulant used to treat turbidity, since ferrous sulfate is
likely to be used only on water with low turbidity (high doses of iron may have negative effects on water
November 2009                                                                               9-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
quality). No cost information was available for ferrous sulfate, but ferric chloride was quoted at $497 per
ton (2008$); additionally, dosing information for iron compounds could not be determined. For the Long
Term 2 Enhanced Surface Water Treatment Rule, EPA determined that alum costs between $286 and
$373 per ton (2008$) (USEPA 2005c). EPA's assumption that all turbidity will be treated with alum may
understate treatment costs and savings. EPA estimated avoided costs using the cost of both dry stock and
liquid stock alum to present a range of savings estimates in the sensitivity analysis. The lower cost of
liquid alum was used in the low estimate, but it should be noted that the sludge generation equation in
Section 9.7.1, calculates sludge volume based on dry alum as a coagulant. Therefore, the cost of dry alum
is used for the midpoint and high estimates. The approximate dose of alum used to treat water of a given
turbidity can be estimated using the following equation:

                                      Al = 331og(r)-28                              (Eq. 9-3)

       Where:

       Al  =   alum dose in milligrams per liter

       T  =   influent water turbidity in NTU (after any reduction from plain sedimentation).

After calculating the dose of alum necessary to treat the turbidity in the influent water at a given intake,
this dose was multiplied by the total daily influent volume of the intake (converted from gallons to liters
using a conversion factor of 3.8 liters to one gallon) to obtain the total amount of alum needed to treat
turbidity at the intake per day. This total daily amount of alum was multiplied by 365 to obtain an annual
estimate of the amount of alum needed to treat turbidity from the intake. EPA then used a conversion
factor of 907x106 milligrams per ton to estimate the amount of alum required for turbidity treatment in
tons per year. Finally, the estimated annual amount of alum was multiplied by the unit cost of alum  to
arrive at a final annual cost of alum for each drinking water intake. The following equation presents this
calculation of the annual cost of alum.
TCM = Q *
                                           907*106
                                                      * 365 * C
                                                               M
(Eq. 9-4)
       Where:

       TCA{ =     total annual alum cost in 2008$

       Q =       daily plant flow in gallons

       Al =       alum dose in milligrams per liter

       3.8 =      liters in one gallon

       907* 106 =  milligrams in one ton

       365 =      days in one year

       CAi=      cost of one ton of alum in 2008$
November 2009
                                                                                             9-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
The estimated annual cost of turbidity treatment was calculated under the baseline and post-compliance
scenarios for each surface drinking water intake in the SDWIS database where TSS concentration
information was available, a total of 6,132 intakes.

9.7    Estimating the Cost of Sludge Disposal

The alum added to the water to treat turbidity binds to the particles of sediment, making heavier lumps of
coagulated sediment that settle out of the water column. This residual material, known as sludge, is
proportional to the amount of sediment in the water and the amount of coagulant added. Drinking water
treatment facilities must regularly dispose of this sludge, either by discharging it to a receiving waterbody
or wastewater treatment plant, or by having it hauled away for off-site disposal. The cost of sludge
disposal can be significant.
The estimation of the cost of sludge disposal has several steps:
       >  Estimating the amount of sludge generated
       >  Calculating the probability that a generated quantity of sludge will be disposed of off-site (as
           opposed to discharged to a receiving waterbody)
       >  Estimating the distance to the disposal location
       >  Calculating the cost of sludge disposal

9.7.1  Estimating the Amount of Sludge Generated

A reduction in the turbidity of the water will result in a reduction in the amount of sludge generated from
the chemicals added to treat turbidity,  and thus a reduction in the cost of sludge disposal. Sludge
quantities are a function of the amount of chemical coagulants added to treat the water and the TSS levels
of the water. EPA (USEPA 2007c) derived the following relationship between sludge generation and
chemical coagulants:

S = 8.34*Q*[2.0CaCH + 2.6MgCH + CaNCH +\.6MgNCH + CO2 + 0.44AI + 2.9Fe + SS + A]           (Eq. 9-5)

       Where:

       S =        sludge production (Ib/day)

       Q =        plant flow (mgd)

       CaCH =   calcium hardness  removed as CaCO3 (mg/L)

       MgCH =   magnesium hardness removed as CaCO3 (mg/L)

        CaNCH=  noncarbonated calcium hardness removal as CaCO3  (mg/L)

       MgNCH =  noncarbonated magnesium hardness removed as CaCO2 (mg/L)

       CO2 =     carbon dioxide  removed by lime addition, as CaCO3

       Al =       dry alum dose (mg/L) (as 17.1 percent A12O3)
November 2009                                                                              9-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
       Fe =       iron dose as Fe (mg/L)

       SS =       raw water suspended solids (mg/L)

       A =         other additives (mg/L)

EPA used this equation in conjunction with the estimated sediment concentrations in the source water, the
quantity of water taken up by the drinking water intake, and the alum doses required to treat turbidity to
estimate the  amount of sludge generated from turbidity treatment.
EPA recognizes that the estimated volume of sludge takes into account the residuals generated from alum
treatment of TSS, and thus does not represent the actual volumes of sludge generated from the entire
water treatment process. Leaving out the other chemical additives does not affect the difference in sludge
generation from turbidity treatment under the baseline and post-compliance scenarios, as they would be
held constant under both scenarios.

9.7.2  Calculating the Probability That the Sludge Will Require Off-Site Disposal
Public water systems may use different sludge disposal methods. Some facilities dispose of their sludge
off-site; some discharge to water bodies under an NPDES permit or discharge to a wastewater treatment
plant. Facilities may also discharge a certain amount of their residual sludge and dispose of the remainder
off-site. Facilities discharging sludge into receiving waters or to a wastewater treatment plant do not incur
sludge disposal costs. To calculate the amount of sludge disposed of off-site and resulting in disposal
costs to a drinking water treatment facility, EPA calculated the probability of off-site disposal.
The Agency used facility data from the drinking water survey database to calculate the number of
facilities reporting any off-site disposal and the percent of sludge being disposed of off-site at these
facilities. The survey found 207 out of 511 facilities (41 percent) reporting some off-site disposal. At
these facilities, on average 85 percent of all residual solid sludge was disposed of off-site. To determine
the likelihood that a generated quantity of sludge will result in off-site disposal costs for a facility, EPA
multiplied the probability that the sludge was generated at a facility performing any off-site disposal (41
percent) by the probability that a unit of sludge from this facility will be disposed of off-site (85 percent).
The estimated probability (Pr) that a given volume of sludge generated will be disposed of off-site, thus
incurring sludge disposal costs, is 35 percent.

9.7.3  Estimating the Distance to the Disposal Location
The cost of sludge disposal is also a function of the distance to the disposal location. A sample of 10
facilities and their disposal locations was taken randomly from the Drinking Water Survey Database.
Because location information was available for both the facilities and their disposal locations, EPA
calculated the driving distance between the facility and its discharge locations for each of the 10 facilities.
If a facility had more than one listed discharge location, an average distance was used. Based on this
sample of 10 facilities, the distance to a disposal location can vary between 1 and 200 miles, and the
estimated average distance (Z) to disposal locations is 21.6 miles.

9.7.4  Calculating the Cost of Sludge Disposal
EPA estimated the annual cost of sludge disposal as a function of sludge volumes and distance to the
disposal facility using the following equation from EPA (USEPA 1996a):
November 2009                                                                                 9-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
              7 = [54.12(5 * 365 * .0005 * Pr)]+ [(3.83 + 0.25Z)(S * 365 * .0005 * Pr)]      (Eq. 9-6)
       Where:
       Y      =     annual cost of sludge disposal ($/year)
       54.12   =     disposal cost per ton of sludge ($)
       S      =     total sludge production from treating turbidity in drinking water (Ibs/day)
       365    =     days per year
       0.0005  =     tons per Ib
       Pr      =     the estimated probability of off-site sludge disposal (35 percent)
       3.83    =     fixed transportation cost per pound of sludge ($)
       0.25    =     variable transportation cost per mile per pound of sludge ($)
       Z      =     average transportation distance of 21.6 miles.
The cost of sludge disposal was estimated under both the baseline and post-compliance scenarios for all
surface drinking water intakes in SDWIS for which TSS concentrations were available.

9.8    Sensitivity Analysis
To account for uncertainty in drinking water treatment costs EPA varied several of the parameters of this
analysis, including the effectiveness of pre-sedimentation, the amount of turbidity generated by a unit of
TSS, and the cost of alum, to provide a range of benefits estimates. The values of these inputs in each
estimate range are as follows:
Low Estimate:
       >  TSS levels exceeding 1,000 mg/L in the baseline are reduced by 90 percent for all options
       >  TSS = 2.2 * Turbidity
       >  CM = $250
Midpoint Estimate:
       >  TSS levels exceeding 1,000 mg/L in the baseline are reduced by 60 percent for all options
       >  TSS = 1.5 * Turbidity
       >  CM = $327
High Estimate:
       >  TSS levels at exceeding  1,000 mg/L in the baseline are reduced by 30 percent for all options
       >  TSS = 0.8 * Turbidity
       >  CM = $327

November 2009                                                                                9-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
9.9    Estimating the Total Costs of Drinking Water Treatment Under Different
       Policy Scenarios
EPA estimated the total national costs of treating drinking water for turbidity by summing the estimated
annual costs of chemical coagulants added to reduce turbidity and the estimated annual cost of disposing
of the sludge resulting from this chemical treatment. Table 9-3 presents annual estimated total drinking
water treatment costs for the baseline scenario, including low, midpoint, and high estimates for avoided
costs. Estimated drinking water treatment costs for TSS and turbidity for facilities included in EPA's
analysis range from $343 to $651 million nationwide under the baseline scenario, with a midpoint value
of $521 million.
 Table 9-3: Drinking Water Turbidity Treatment Costs Under the Baseline Scenario
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Average
Low
26.7
50.3
71.7
110.8
136.8
151.5
166.0
121.8
123.8
43.9
104.0
Treated Turbidity
Mid
40.1
84.4
105.6
163.3
263.8
367.2
422.9
254.5
256.3
64.4
205.9
(NTU)
High
76.9
177.9
198.9
307.7
613.3
960.3
1,129.5
619.4
620.5
120.8
486.0
Cost per million gallons1
Treatment Cost
(millions of 2008$)
Low
$6.6
$21.3
$45.5
$60.0
$52.6
$60.3
$18.4
$18.3
$54.4
$5.1
$342.7
$58.4
Mid
$10.5
$33.9
$61.6
$82.5
$83.6
$90.1
$33.9
$25.5
$91.5
$7.8
$520.9
$88.7
High
$14.9
$44.7
$72.8
$98.2
$105.8
$110.8
$44.5
$30.3
$119.1
$10.2
$651.3
$110.9
 1 Calculated for the estimated 17,373 million gallons per day of intake for public water systems in SDWIS.

It is important to note that these avoided cost estimates do not take into account any treatment or disposal
avoided costs to public wastewater treatment facilities to which drinking water treatment facilities may
route their sludge, nor any environmental benefits from reduced volumes of sludge discharged directly
into surface waters by direct drinking water treatment facilities. Therefore, the total benefits of reducing
turbidity in drinking water treatment are understated.

9.10    Estimating Avoided costs from Lower Sediment and Turbidity in Drinking
        Water Influent

The total avoided costs from lowered turbidity resulting from lower TSS concentrations in drinking water
influent were estimated as the difference between drinking water turbidity treatment under the baseline
and post-compliance scenarios. Table 9-4, Table 9-5, Table 9-6, and Table 9-7 present estimated savings
in drinking water treatment costs from lowered turbidity for the three policy options EPA considered, as
well as EPA's final policy option. The tables include low, midpoint, and high estimates for savings under
each of these options.
The estimated savings from reduced TSS and turbidity treatment at drinking water facilities are between
$978,400 and $2.1 million, varying between the least and most stringent policy options and between the
low and high estimates. Region 5 benefits the most from the TSS reductions expected from this regulatory
November 2009
                                                                                          9-13

-------
Environmental Impact and Benefits Assessment for the C&D Category
action. The avoided costs in Region 5 under Option 1 account for more than half of the total national
savings. Other regions receive a larger portion of benefits under Options 2 through 4, though Region 5
still accounts for the greatest proportion of the total cost savings, though Regions 4 and 6 also show
significant savings under Options 2 through 4.
Option 1, which requires non-numeric effluent limitations for all sites, is EPA's least stringent policy
option. Average turbidity reductions are less than 1 NTU  for this policy option, and the estimated savings
drinking water treatment costs range between $978,400 and $1.3 million, with a midpoint estimate of $ 1.2
million.
Option 2 requires ATS on sites with more than 30 acres disturbed at one time and imposes a 13 NTU
turbidity standard while  requiring non-numeric effluent limitations on all sites and is similar to the option
EPA proposed previously. This option reduces turbidity between 0.4 and 1.3 NTU, translating to $1.4
million to $1.9 million in avoided costs, with a midpoint estimate of $1.8 million.
Option 3, EPA's most stringent policy option,  requires ATS on sites with more than 10 acres disturbed at
one time, imposes a 13 NTU turbidity standard on these sites, and requires non-numeric effluent
limitations on all sites. This option reduces treated turbidity by an average of 0.7 NTU in the midpoint,
ranging from 0.4 to 1.4 between the low and high estimates. Total estimated avoided costs for this option
are between $1.7 and $2.1 million, with a midpoint estimate just below $2.1 million.
Option 4, requires passive treatment systems on all sites with 10 or more acres disturbed at one time, and
establishes a numeric turbidity standard of 250 NTU (expressed as a daily maximum value) for sites
required to implement passive treatment. In addition, all sites will be required to meet non-numeric
effluent limitations. National average turbidity reductions from Option 4 range from 0.4 NTU to 1.3
NTU, with a midpoint estimate of 0.6 NTU. Total avoided costs for Option 4 range between $1.5 and
$1.9 million, with a midpoint estimate of $1.8  million. While Option 4 is less stringent than Option 3, it is
estimated to reduce turbidity in drinking water sources by nearly as much and to produce similar
monetized benefits.
As construction site discharge is more likely to contain smaller particles that contribute less to TSS and
more to turbidity, the high estimates for Options 2, 3, and 4 may be more relevant because EPA uses a
conversion factor between TSS and turbidity that takes this into  account. The high estimate for turbidity
reductions under Option 2 is 1.3 NTU nationwide. Under  Option 3, the high estimate is on average
around 1.4 NTU nationwide. For Option 4, the selected option, nationwide reductions are 1.4 NTU. In
Region 5,  high estimates of reductions under all options are 4.1 NTU.
November 2009                                                                                 9-14

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 9-4: Reduction in Drinking Water Treatment Costs Under Option 1
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Average
Low
0.0
0.1
0.0
0.4
1.2
0.3
0.1
0.0
0.1
0.1
0.2
Reduction in Treated
(NTU) 1
Mid
0.0
0.5
0.0
0.5
1.9
0.5
0.2
0.0
0.1
0.2
0.4
Turbidity
High
0.0
1.8
0.1
1.0
3.9
1.1
0.4
0.0
0.2
0.3
0.9
Avoided
Low
$4.0
$44.1
$8.0
$136.2
$515.0
$162.4
$11.3
$1.1
$7.6
$88.7
$978.4
costs (thousands
Mid
$5.0
$91.2
$9.2
$161.1
$603.4
$194.1
$14.1
$1.3
$9.9
$110.4
$1,199.8
of 2008$)
High
$5.8
$129.8
$9.3
$164.5
$616.6
$201.0
$15.8
$1.4
$11.3
$110.4
$1,265.9
 1 Average turbidity reductions shown as 0.0 are not actually zero, but not sufficiently large to show at this level of significant digits.





 Table 9-5: Reduction in Drinking Water Treatment Costs Under Option 2
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Average
Low
0.0
0.1
0.1
0.7
1.2
0.7
0.2
0.0
0.1
0.2
0.4
Reduction in Treated
(NTU) 1
Mid
0.0
0.6
0.1
1.0
2.0
1.1
0.4
0.0
0.2
0.3
0.6
Turbidity
High
0.1
1.8
0.2
1.9
4.1
2.3
0.8
0.1
0.4
0.6
1.3
Avoided
Low
$8.2
$50.8
$24.0
$258.9
$526.6
$341.5
$23.5
$2.2
$15.8
$195.9
$1,447.4
costs (thousands
Mid
$10.2
$99.4
$27.7
$307.6
$618.4
$408.3
$29.3
$2.7
$20.5
$244.9
$1,769.1
of 2008$)
High
$11.7
$138.4
$27.8
$318.3
$633.7
$422.5
$33.0
$2.9
$23.3
$244.9
$1,856.6
 1 Average turbidity reductions shown as 0.0 are not actually zero, but not sufficiently large to show at this level of significant digits.
November 2009
                                                                                                  9-15

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 9-6: Reduction in Drinking Water Treatment Costs Under Option 3
EPA
Region
1
2
3
4
5
6
1
8
9
10
Total
Average
Low
0.0
0.1
0.1
0.8
1.2
0.9
0.3
0.0
0.1
0.3
0.4
Reduction in Treated
(NTU)
Mid
0.1
0.6
0.1
1.2
2.0
1.4
0.5
0.0
0.2
0.4
0.7
Turbidity
High
0.1
1.8
0.2
2.2
4.1
2.8
1.0
0.1
0.5
0.7
1.4
Avoided
Low
$10.2
$54.1
$31.8
$312.7
$532.4
$428.6
$29.3
$2.7
$20.1
$237.0
$1,658.9
costs (thousands
Mid
$12.7
$103.7
$36.6
$370.7
$626.4
$513.1
$36.6
$3.4
$26.9
$320.3
$2,050.5
of 2008$)
High
$14.6
$143.2
$36.7
$383.2
$643.0
$531.2
$41.2
$3.7
$31.4
$320.3
$2,148.5
 Table 9-7: Reduction in Drinking Water Treatment Costs Under Option 4
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total
Average
Low
0.0
0.1
0.1
0.7
1.2
0.8
0.2
0.0
0.1
0.2
0.4
Reduction in Treated
(NTU)
Mid
0.0
0.6
0.1
1.1
2.0
1.3
0.4
0.0
0.2
0.2
0.6
Turbidity
High
0.1
1.8
0.2
2.0
4.1
2.6
0.9
0.1
0.4
0.4
1.3
Avoided
Low
$6.8
$50.3
$25.6
$281.1
$526.8
$396.9
$27.0
$2.2
$16.4
$145.5
$1,478.7
costs (thousands
Mid
$8.5
$98.8
$29.6
$333.5
$618.8
$474.5
$33.7
$2.8
$21.5
$181.9
$1,803.5
of 2008$)
High
$9.7
$137.8
$29.6
$344.8
$634.2
$490.6
$37.8
$3.1
$24.7
$181.9
$1,894.3
9.11   Sources of Uncertainty/Limitations

Key limitations of EPA's analysis of avoided costs to drinking water treatment facilities are outlined
below. The SPARROW model for suspended sediments that was used for estimating TSS concentrations
in the source waters also has a number of limitations. These limitations are discussed in detail in
Chapter 6.
       > EPA's analysis of reductions in TSS and turbidity in surface waters is limited to the RF1
          reaches modeled by SPARROW. Therefore, reductions at any facilities that are not located on
          RF1 reaches would not be included in this assessment of benefits to drinking water treatment,
          leading to an underestimation of benefits.
       > Sediment filtration  systems and pre-sedimentation (allowing water to sit and sediment to
          filter out before treatment) at drinking water treatment facilities reduce the sediment
          concentration of the water before it enters chemical treatment, so that the turbidity level of
November 2009
                                                                                         9-16

-------
Environmental Impact and Benefits Assessment for the C&D Category
           the water entering the facility is not the turbidity level that is eventually treated with
           coagulants. Assuming that the differential between pre- and post- compliance sediment
           concentration is proportional to the differential between pre- and post-compliance turbidity
           treatment introduces uncertainty, as the lower sediment levels may be more or less affected
           by the pre-sedimentation and filtration processes. EPA's analysis attempts to account for this
           uncertainty by varying the maximum amount of TSS and turbidity treated by a given drinking
           water treatment facility.
        >  If a drinking water treatment facility produces sludge that is toxic (due to other pollutants in
           the water besides sediment), its disposal costs may be significantly higher because toxic
           sludge disposal is more restricted and costly. If the facility cannot separate the sludge
           generated by sediment treatment from the sludge generated by treatment of toxics (which is
           likely the case), then all of its sludge will be characterized as toxic. This analysis may
           understate the cost of disposal (and thus the avoided costs of smaller quantities of sludge to
           be disposed of) for facilities that generate toxic sludge.
        >  Reducing nutrient loadings to surface waters is expected to reduce eutrophication which is
           one of the main causes of taste and odor impairment in drinking water. Taste and odor in
           drinking water has a major negative impact on the public perception of drinking water safety
           and the drinking water industry due to  a significant increase in drinking water treatment costs
           from foul taste and odor in the  source waters. Although the final regulation is expected to
           reduce the cost of drinking water treatment to improve taste and odor, the Agency was unable
           to monetize this benefit category due to data limitations.
        >  The analysis presumes that reduced costs of drinking water treatment will not result in
           appreciable changes in prices (costs) of drinking water paid by households. Given the
           marginal costs of reduced turbidity treatment relative to other costs of drinking water
           production, this assumption seems valid. However, if substantial price effects do occur,
           additional benefits  will accrue at the consumer level.
November 2009                                                                                 9-17

-------
Environmental Impact and Benefits Assessment for the C&D Category
10    Nonmarket Benefits from Water Quality Improvements
As discussed in the preceding chapters of this document, sediments, nutrients, and other pollutants from
construction sites may have a wide range of effects on water resources located in the vicinity of the
construction sites. These environmental changes affect environmental services valued by humans,
including recreation; commercial fishing; public and private property ownership; navigation; water supply
and use; and existence services such as aquatic life, wildlife, and habitat designated uses. These
nonmarket benefits (Freeman 2003) are in addition to market benefits (e.g., avoided costs of producing
various market goods and services) estimated in prior chapters.
This chapter describes the use of meta-analysis of surface water valuation studies for estimating benefits
of water quality improvements resulting from the regulation. The technical details involved in the
estimation of meta-analysis are presented in Appendix G as well as in sources such as  Bateman and Jones
(2003), Johnston et al. (2005, 2006), Shrestha et al. (2007), and Rosenberger and Phipps (2007).

10.1  Water Quality Index

To link water quality changes from reduced sediment discharge to effects on human uses and support for
aquatic and terrestrial species habitat, this analysis utilizes a water quality index (WQI). The index
translates water quality measurements, gathered for multiple parameters that are indicative of various
aspects of water quality, into a single numerical indicator. The parameters used in formulating the WQI
are determined based on waterbody type, scientific understanding of ecosystem response to varying
conditions, and available data.
Most importantly for the present analysis, the WQI provides the link between specific pollutant levels, as
reflected in individual index parameters (e.g., dissolved oxygen concentrations), and the presence of
aquatic species and suitability for particular recreational uses. The WQI value, which is measured on a
scale from 0 to 100, reflects varying water quality, with 0 for poor quality and 100 for excellent.
Numerous water quality indices have been developed and documented in the literature since the 1960s
(Horton 1965); however, no standardized approach has emerged. EPA Region 10 was one of the first
designers of the WQI framework in use today (Brown et al.  1970). The National Sanitation Foundation
WQI (McClelland 1974) and Oregon WQI (Dunnette 1979) proposed subsequent iterations that built on
this original framework. McClelland (1974) relied on a survey of water quality experts to identify water
quality parameters to be included in the index, the structure and the relationship between the parameters,
and the weights given to the parameters in the WQI equation. This is a procedure known as the Delphi
Method (Dalkey 1968).
The WQI used in this analysis builds on McClelland's work, and the methodology developed by Dunnette
(1979), which was subsequently updated by Cude (2001) to better account for spatial and morphologic
variability in the natural characteristics of streams. To support the analysis of changes in sediment
loadings from construction sites relative to baseline conditions, EPA adapted Cude's methodology to the
national scale by developing subindex curves for TSS at the level of ecoregions. EPA also  further
modified the WQI methodology to apply to estuarine reaches, building on prior applications of WQI
concepts to saltwater environments (Harrison et al. 2000; Canadian Council of Ministers of the
Environment 2001; Carruthers and Wazniak 2003; Gupta et al. 2003; and USEPA 2007a).
November 2009
                                                                                           10-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
Implementing the WQI methodology involves three key steps: (1) obtaining water quality measurements
for each parameter included in the WQI; (2) transforming measurements to subindex values expressed on
a common scale; and (3) weighting and aggregating the individual parameter subindices to obtain an
overall WQI value that reflects waterbody conditions across the parameters.
In compiling water quality data for each parameter, EPA used a combination of average water quality
field measurements and SPARROW modeling results as input values for each of the parameters included
in the WQI. Data sources for measurements are discussed in Section 10.1.4, and SPARROW modeling
results were discussed in Chapter 6.
The second step in the implementation of the WQI involves the transformation of parameter
measurements into subindex values that express water quality conditions on a common scale of 0 to 100.
With the exception of the TSS parameter, EPA used the subindex transformation curves developed by
Dunnette (1979)  and Cude (2001) for the Oregon Water Quality Index. In the case of TSS concentrations,
EPA adapted the approach developed by Cude (2001) to account for the wide range of natural or
background sediment concentrations that result from varying geologic and other region-specific
conditions, and to reflect the national context of the analysis. TSS subindex curves were developed for
each Level III ecoregion (USEPA 2009d) using baseline TSS concentrations calculated in SPARROW at
the RF1 reach level.13 For each of the 85 Level III ecoregions intersected by the RF1 reach network, EPA
derived the TSS transformation curve by assigning a score of 100 to the 25th percentile of the  reach-level
TSS concentrations in the ecoregion (i.e., using the 25th percentile as a proxy for "reference"
concentrations), and a score of 70 to the median concentration. An exponential equation was then fitted to
the two concentration points.
The final step in  implementing the WQI involves combining the individual parameter subindices into a
single WQI value that reflects the overall water quality across the parameters. Following McClelland's
approach, EPA calculated the overall WQI for a given reach using a weighted geometric mean function as
follows:
                                                 n
                                                      "                               (Eq. 10-1)
       Where:
       WQIr   =  the multiplicative water quality index (from 0 to 100) for reach r
       Qi      =  the water quality subindex measure for parameter /
       Wi      =  the weight of the i-th parameter
       n       =  the number of parameters
The WQI methodology used for rivers and streams and for estuaries are described in Section 10.1.1 and
Section 10.1.2, respectively.
13   The selected data exclude outlier TSS concentration values, defined as values that exceed the 95th percentile based on the
    national population of all RF1 reaches.
November 2009                                                                                10-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
10.1.1  Freshwater WQI for Rivers and Streams

The freshwater WQI used by EPA to evaluate water quality in rivers and streams includes six parameters
selected to represent major stream impairment categories: dissolved oxygen (DO), biochemical oxygen
demand (BOD), fecal coliform (FC), total nitrogen (TN), total phosphorus (TP), and total suspended
solids (TSS). These freshwater WQI parameters are based on the set of parameters used in the WQI
developed by Cude (2001) but exclude temperature and pH. Temperature and pH are not listed as
common causes of impairment nationally or with respect to the uses  assessed in this economic
valuation.14 Additionally, temperature may be considered to be indirectly reflected in the DO  parameter
because DO is highly temperature dependent. In addition, the WQI developed by Cude (2001) does not
explicitly account for turbidity associated with water quality impacts from TSS and eutrophication.
Table 10-1 presents parameter-specific functions used for transforming water quality data into water
quality subindices  for freshwater waterbodies. The equation parameters for each of the 85 ecoregion-
specific TSS subindex curves are provided in Appendix F.
    EPA (2009a) reports 38,228 miles and 26,569 miles of rivers or streams impaired or threatened due to temperature and pH,
    respectively, as compared to much higher incidence of impairment due to pathogens (112,084 miles), sediment (85,248
    miles), and nutrients (81,279 miles). Temperature and pH are similarly responsible for a relatively small share of
    impairments noted in estuaries and bays.
November 2009                                                                                   10-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 10-1: Freshwater Water Quality Subindices
       Parameter
Concentrations
Concentration Unit
          Subindex
                                       Dissolved Oxygen (DO)
 DO saturation <100%
DO
DO
DO
<3.3
3.3 < DO < 10.5
10.5  1,600
Lbs/lOOmL
Lbs/lOOmL
Lbs/100 mL
98
98 *exp(FC- 50)* -9.9178 E-4
10
                                         Total Nitrogen (TN)
TN
TN
<3
>3
Mg/L
Mg/L
100 * exp(TN * -0.4605)
10
Total Phosphorus (TP)
TP
TP
<0.25
>0.25
Mg/L
Mg/L
100- 299.5 *TP- 0.1384 * TP2
10
                                    Total Suspended Solids (TSS)1
TSS
TSS
TSS
 TSSioo
Mg/L
Mg/L
Mg/L
10
a * exp(TSS*b); where a and b are
ecoregion-specific values
100
                              Biochemical Oxygen Demand, 5-day (BOD)
BOD
BOD
<8
>8
Mg/L
Mg/L
100 *exp(BOD* -0.1993)
10
  TSSio and TSSioo are ecoregion-specific TSS concentration values that correspond to subindex scores of 10 and 100, respectively.
 Source: Cude (2001) (ecoregion-specific curves were developed for TSS based on Code's methodology).	

Table 10-2 lists the freshwater WQI parameters and the weights used in aggregating the subindex values
into the overall WQI for inland rivers and streams. EPA (USEPA 2002) revised the weights originally
developed by McClelland (1974) by redistributing the weights to the six parameters retained in the EPA
WQI so that the ratio among the parameters is maintained and the weights sum of 1.
Table 10-2: Original and Revised
Parameters
Dissolved oxygen (% saturation and mg/L)
Fecal coliform (colonies/100 mL)
Biochemical oxygen demand, 5-day (mg/L)
Total nitrogen (mg/L)
Total phosphorus (mg/L)
Total suspended solids (mg/L)
pH (standard)
Temperature (C)
Total Solids (mg/L)
Total
Weights for Freshwater WQI
Original Weights
(McClelland 1974)
0.17
0.16
0.11
0.10
0.10
0.08
0.11
0.10
0.07
1.00
Parameters
Revised Weights (USEPA
2002)
0.24
0.22
0.15
0.14
0.14
0.11
—
—
—
1.00
November 2009
                                                                                                10-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
10.1.2 WQI Methodology for Estuaries

The methodology used in calculating a WQI score for estuaries and coastal waterbodies adapts the
approach outlined above for rivers and streams to account for the different characteristics and response of
saltwater systems. EPA reviewed several applications of WQI concepts to saltwater systems (Harrison et
al. 2000; Carruthers and Wazniak 2003; Gupta et al. 2003; and USEPA 2007a) to identify water quality
parameters pertinent to estuaries and their relative importance in characterizing overall water quality.
The estuarine WQI includes parameters that are common to both freshwater and saltwater systems—
dissolved oxygen, nitrogen, phosphorus, fecal coliform, and suspended solids—and replaces biochemical
oxygen demand with chlorophyll-a (ChA). Chlorophyll-a assesses the amount of phytoplankton present in
the waterbody. While phytoplankton is a natural component of the waterbody's food web, changes in its
abundance (e.g., high ChA), species composition, and productivity are commonly the first biological
response to nutrient enrichment (Chesapeake Bay Program 2008). Subsequent decomposition of the
phytoplankton may reduce  available dissolved oxygen and induce hypoxia and anoxia.
Similarly to the WQI for rivers and streams, the data for each parameter included in the WQI for estuaries
are transformed to a standard (0-100) scale using subindex curves. Table 10-3 presents parameter-specific
functions used in transforming water quality data to subindices for estuaries. The curve used for ChA is
derived from the curves used for TN, since both parameters indicate trophic status.
Table 10-3: Estuarine Water Quality Subindices
Parameter
Concentrations
Concentration Unit
Subindex
Dissolved Oxygen (DO)
DO saturation <100%
DO
DO
DO
<3.3
3.3 1,600
Ibs/lOOmL
Ibs/lOOmL
Ibs/lOOmL
98
98 *exp(FC- 50)* -9.9178 E-4
10
Total Nitrogen (TN)
TN
TN
<3
>3
mg/L
mg/L
100 * exp(TN * -0.4605)
10
Total Phosphorus (TP)
TP
TP
<0.25
>0.25
mg/L
mg/L
100- 299.5 *TP- 0.1384 * TP2
10
                                  Total Suspended Solids (TSS)1
TSS
TSS
TSS
 TSSjoo
mg/L
mg/L
mg/L
10
a * exp(TSS*b); where a and b are
ecoregion-specific values
100
November 2009
                                                                                            10-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 10-3: Estuarine Water Quality Subindices
Parameter
Concentrations
Concentration Unit
Subindex
Chlorophyll-a (ChA)
ChA
ChA
<40
>40
pg/L
pg/L
100 *exp(ChA * -0.05605)
10
  TSSio and TSSioo are ecoregion-specific TSS concentration values that correspond to subindex scores of 10 and 100, respectively.
 Source: Cude (2001) (ecoregion-specific curves were developed for TSS based on Code's methodology).	


Table 10-4 shows the weighting scheme used to combine the parameters into an overall WQI. These
weights are a modification of the ones used for the freshwater WQI.

 Table 10-4: Weights for Estuarine WQI Parameters	
	Parameters	Estuary WQI Weights1
 Dissolved oxygen (% saturation and mg/L)	0.26	
 Fecal coliform (colonies/1 OOmL)	0.25	
 Total nitrogen (mg/L)	0.15	
 Total phosphorus (mg/L)	0.15	
 Total suspended solids (mg/L)	0.11	
 Chlorophyll-a (mg/L)	0.08	
 Total	1.00	
 1 Weights used in the freshwater WQI were redistributed among estuarine WQI parameters to account for the removal of BOD and the
 addition of ChA.
10.1.3 Relation between WQI and Suitability for Human Uses

Once an overall WQI value is calculated, it can be related to suitability for potential uses. Vaughan (1986)
developed a water quality ladder (WQL) that can be used to indicate whether water quality is suitable for
various human uses (i.e., boating, rough fishing, game fishing, swimming, and drinking without
treatment). Vaughan identified "minimally acceptable parameter concentration levels" for each of the five
potential uses. Water quality is deemed acceptable for each use if none of the six parameters exceeds the
threshold concentration levels. Vaughan used a scale of zero to 10 instead of the WQI scale of zero to  100
to classify water quality based on its suitability for potential uses. Table 10-5 presents water use
classifications and the corresponding WQL and WQI values.
Table 10-5: Water Quality Classifications
Water Quality Classification
. . . drinking without treatment
... swimming
. . . game fishing
... rough fishing
. . . boating
WQL Value
9.5
7.0
5.0
4.5
2.5
WQI Value1
95
70
50
45
25
1 The WQI value corresponding to a given classification of water quality equals the WQL value multiplied by 10.
Source: Vaughan (1986).
November 2009                                                                                     10-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
10.1.4  Sources of Data on Ambient Water Quality

For river and streams, EPA used the following data sources to obtain ambient concentrations for the six
parameters included in the WQI:
        >  The SPARROW model outputs provided data for baseline and post-compliance
           concentrations of total nitrogen, total phosphorus, and total suspended solids as discussed in
           Chapter 6.
        >  The USGS National Water Information System (NWIS) provided concentration data for three
           parameters: (1) fecal coliform, (2) dissolved oxygen, and (3) biochemical oxygen demand.
        >  EPA's Storage and Retrieval (STORET) data warehouse provided additional data on fecal
           coliform counts and biochemical oxygen demand (USEPA 2008e). The NWIS database did
           not include data for these two parameters for 88.6 percent and 91.3 percent of the 9,444 RF1
           reaches, respectively. To address these data gaps, EPA augmented fecal coliform and
           biochemical oxygen demand data by adding observations from the STORET data warehouse.
           The addition of these  observations increased the fecal coliform and biochemical oxygen
           demand data, so that 71.9 percent and 33.8 percent, respectively, of the 9,444 reaches were
           covered.
Baseline freshwater WQI values are based on baseline data for 60,017 reaches in the coterminous United
States.15 Baseline concentrations for all WQI parameters were available for a total of 9,444 reaches. EPA
used a successive average approach to address the data gaps in the remaining freshwater reaches. The
approach involves assigning the average of ambient concentrations for a WQI parameter within a
hydrologic unit to reaches within the same hydrologic unit with missing data and progressively expanding
the geographical scope of the hydrologic unit (HUC8, HUC6, HUC4, and HUC2) to fill in all missing
data.16 This approach assumes that waterbody reaches located in the same watershed generally share
similar characteristics. Using this  estimation approach, EPA was able to compile water quality data for a
total of 60,017 freshwater reaches (including 50,573 reaches estimated from HUC-based averages).
The following sources provided data for estuaries and coastal waterbodies:
        >  The SPARROW model outputs provided data for baseline and post-compliance
           concentrations of suspended sediment. Some sediment concentrations were calculated from
           sediment flux using Dissolved Concentration Potentials as discussed in Chapter 6.
        >  EPA's Environmental Monitoring & Assessment Program, National Coastal Assessment
           Monitoring Data (EMAP-NCA) database provided data for approximately 3,000 sampling
           stations in the United States, for all pertinent parameters except fecal  coliform (USEPA
           2008a).
    No ambient concentration data were available for Alaska and Hawaii.
    Hydrologic Unit Codes (HUCs) are cataloguing numbers that uniquely identify hydrologic features such as surface drainage
    basins. The HUCs consist of 8 to 14 digits, with each set of 2 digits giving more specific information about the hydrologic
    feature. The first pair of values designate the region (of which there are 21), the next pair the subregion (total of 222), the
    third pair the basin or cataloguing unit (total of 352), and the fourth pair the subbasin, or accounting unit (total of7,262)
    (USGS 2007a). Digits after the first eight offer more detailed information, but are not always available for all waters. In this
    discussion, a HUC level refers to a set of waters that have that number of HUC digits in common. For example, the HUC6
    level includes all reaches for which the first six digits of their HUC are the same.
November 2009                                                                                   10-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
       >  The National Estuarine Research Reserve System database (NERRS) provided data for 25
           estuarine reserves, each containing multiple water monitoring stations, for all relevant
           parameters except fecal coliform (NOAA 2008).
       >  EPA's STORET data warehouse provided additional data on all relevant parameters,
           including fecal coliform counts, but excluding chlorophyll-a (USEPA 2008e).
       >  EPA also obtained data from the following state and local agencies through special requests:
           •   The Houston Area Research Center (HARC) is a nonprofit organization that supports the
               water monitoring data collection for the Texas Commission on Environmental Quality.
               HARC provided data spanning the period of 1999 to 2005 (Gonzalez 2009).
           •   The New Hampshire Department of Environmental Quality collects coastal water quality
               data in support of its Coastal Program. Data include measurements made from 1999 to
               2008 (Hunt 2009).
           •   The Southern Carolina Estuarine and Coastal Assessment Program is a collaboration
               between the South Carolina Department of Natural Resources (SCDNR) and the South
               Carolina Department of Health and Environmental Control (SCDHEC) providing
               monitoring and periodic reports on the conditions of the state's estuarine habitats. Data
               are available for the period of 1999 to 2004 (VanDolah 2009).
           •   The Connecticut Department of Environmental Protection, on behalf of the collaboration
               with EPA on the Long Island Sound Study, conducts a Long Island Sound Water Quality
               Monitoring Program. In total, 28 stations are monitored each year. Data include
               measurements made from 1999 to 2008 (Lyman 2009).
           •   The Washington Department of Ecology conducts marine water quality monitoring at 85
               stations in Puget Sound, Grays Harbor, and Willapa Bay. About 40 stations are monitored
               each year on a monthly basis, with some stations monitored on a rotating schedule. Data
               cover the period of 1975 to 2008 (Thorn 2009).
           •   The Oregon Department of Environmental Quality collects coastal water quality data in
               support of its Water Quality program. Data include measurements made from 1991 to
               2008 (Marxer 2009).
In the case of estuaries and coastal reaches, EPA used a similar approach to estimate missing
concentration values based on the average concentrations available for reaches within consecutively
coarser HUC levels. Thus, for RF1 reaches where no empirical measurement could be obtained from any
of the sources listed above, the parameter value was estimated by averaging the corresponding values for
adjacent estuarine/coastal reaches within the same HUC8 watershed. Due to the smaller scale of estuarine
waterbodies, EPA used the average of concentration values for HUC8, HUC6, and HUC 4 to estimate
average parameter concentrations for reaches where data gaps existed and did not estimate values based
on HUC2 averages. Using this approach, EPA compiled ambient water quality data for all WQI
parameters for 772 reaches out of the total 2,067 estuarine/coastal reaches in the coterminous United
States. Due to data limitations  1,295 coastal reaches were excluded from the analysis.
EPA calculated baseline WQI for all reaches based on modeled, measured, or extrapolated parameter
values. The WQI was calculated using the methodology outlined above according to the type of reach
(either an inland stream or an estuarine reach). SPARROW identifies estuarine and coastal reaches using
November 2009                                                                               10-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
a specific flag.17 As done throughout the analysis (see Chapter 6), in calculating WQI values, EPA
replaced all outlier TSS concentrations equal or above the 95th percentile with the 95th percentile value
(6,157.8 mg/L in the baseline). Based on this threshold, a total of 3,106 reaches are considered outliers.
Table 10-6 shows the distribution of reach miles by WQI value and EPA region for the baseline scenario
(existing conditions).
Table 10-6: Percentage of Reach Miles
for EPA Regions: Baseline Scenario
EPA Region
1
2
3
4
5
6
7
8
9
10
National Average
WQK25
0.2%
0.5%
5.0%
0.8%
6.1%
1.0%
17.9%
5.0%
19.0%
0.1%
5.6%
in Coterminous
25
-------
Environmental Impact and Benefits Assessment for the C&D Category
calculated because they are so small EPA does not expect any benefits from these improvements. For
each combination of the baseline water quality category and the improvement range, the Agency
estimated average AWQI and the corresponding percentage of total reach miles in the state. In this
analysis, the numbers of reach miles improved under each option - i.e., miles that experience a change in
the WQI value - differ from the numbers presented in Chapter 6 which instead looked at absolute
changes in TSS, TN, or TP concentrations. Additionally, data presented in this chapter cover all reaches
within the RF1 network, including those that do not receive construction discharge directly but may
nevertheless see improvement due to changes in water quality in upstream reaches. As noted in Chapter
6, when considering all RF1 reaches, including those that do not receive construction sediment discharge
directly, the analysis shows a total of 431,074 reach miles improving under Options 1, 2, and 4. These
miles account for 68.7 percent of the total RF1 reach network. Option 3 has an even broader effect,
reducing TSS concentrations in a total of 472,402 reach miles (75.3 percent of the RF1 reach network).
Table 10-7, Table 10-8, Table 10-9, and Table 10-10 summarize changes in ambient water quality
resulting from Options 1, 2, 3, and 4 respectively. Appendix H provides more detail on water quality
improvements by the baseline water quality category.
EPA estimated that Option 1 will not result in significant water quality improvements; this option
improves water quality in only 73,054 reach miles18 or 11.6 percent of the 627,679 miles modeled in the
analysis. EPA Region 4 shows the most water quality improvements, with 29,911 reach miles (33.1
percent of reach miles modeled in this region) being affected. Region 6 is expected to have the next
largest water quality improvement with 23,355 reach miles being affected under the post-compliance
scenario.
EPA's Option 2 is estimated to improve ambient water quality in 112,429 reach miles (17.9 percent) that
receive construction discharges nationwide. EPA's  Regions 1, 4, and 6 are each estimated to experience
improved water quality on 20 percent or greater of RF1  reach miles included in the analysis. EPA's water
quality analysis predicts that Regions 4 and 6 will experience the most improvements in terms of both
reach miles (45,817  and 31,540) and percentage of total regional reach length (50.7 percent and
33.2 percent). EPA's analysis also indicates that Region 6 would benefit the most in terms  of the
estimated magnitude of water quality improvements with 3.1 percent of reach miles modeled estimated to
improve by greater than 0.5 WQI units. Conversely, EPA Region 2 is estimated to improve the least, with
1,532 reach miles benefiting from higher water quality.
Option 3 yields the most significant results overall with 129,747 reach miles expected to improve under
the post-compliance scenario. The estimated scale of improvements ranges from 3.2 percent to
55.2 percent of reach miles in EPA's Region 8 and 4, respectively. EPA estimated that Region 4 will
experience the largest water quality improvements,  with 49,945 reach miles improved (55.2 percent of the
reach miles modeled). EPA's analysis also shows that Region 6 will have the most reach miles (3,487)
improving by greater than 0.5 WQI units, while Region 2 will have the lowest water quality
improvements with 2,147 reach miles showing any  improvement under the post-compliance scenario.
Option 4 generates the second most improvements in water quality when compared to the other three
regulatory options. EPA calculates that atotal of 113,963 reach miles (18.2 percent of reaches in the
analysis) will be improved under this option. As with the other three options, Region 4 is expected to
have the most improved reaches with 47,130 reach miles (52.1 percent) experiencing some improvements
in water quality. EPA Region 2 is expected to improve the least in terms of reach miles, with 1,438 miles
    For the purpose of the discussion, "reach miles" refers to both freshwater and estuarine RF1 reaches.
November 2009                                                                               10-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
(9.5 percent) improving. EPA Region 8 is expected to have the least improvements in terms of percentage
of reach miles improved. Only 1.3 percent (1,739 miles) are expected to improve in the region under
Option 4. Region 6 has the most miles with water quality improvements greater than 0.5 WQI units,
3,313 miles.
November 2009                                                                               10-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 10-7: Estimated Water Quality Improvements Under Option 11
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Baseline Scenario
Total Reach
Miles in
Reach Miles RF1
Modeled Network
16,182 18,324
15,140 16,110
28,904 33,617
90,435 94,525
68,285 71,550
95,098 98,681
60,909 60,909
130,311 130,311
54,228 56,492
68,189 69,524
627,679 650,043
Water Quality Improvements by WQI Change
O.OKAWQKO.l
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
1,696 10.5% 9.3%
415 2.7% 2.6%
1,539 5.3% 4.6%
26,210 29.0% 27.7%
2,931 4.3% 4.1%
17,902 18.8% 18.1%
4,196 6.9% 6.9%
495 0.4% 0.4%
1,360 2.5% 2.4%
3,541 5.2% 5.1%
60,285 9.6% 9.3%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
31 0.2% 0.2%
0 0.0% 0.0%
114 0.4% 0.3%
3,172 3.5% 3.4%
132 0.2% 0.2%
4,227 4.4% 4.3%
562 0.9% 0.9%
64 0.1% 0.1%
134 0.3% 0.2%
1,322 1.9% 1.9%
9,757 1.6% 1.5%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
29 0.2% 0.2%
0 0.0% 0.0%
0 0.0% 0.0%
529 0.6% 0.6%
4 <0.1% <0.1%
1,227 1.3% 1.2%
168 0.3% 0.3%
362 0.3% 0.3%
151 0.3% 0.3%
542 0.8% 0.8%
3,012 0.5% 0.5%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
1,756 10.9% 9.6%
415 2.7% 2.6%
1,653 5.7% 4.9%
29,911 33.1% 31.6%
3,067 4.5% 4.3%
23,355 24.6% 23.7%
4,926 8.1% 8.1%
921 0.7% 0.7%
1,646 3.0% 2.9%
5,404 7.9% 7.8%
73,054 11.6% 11.2%
  Reach miles include both freshwater and estuarine RF1 reaches.
November 2009
                                                                                                                              10-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 10-8: Estimated Water Quality Improvements Under Option 21
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Baseline Scenario
Total Reach
Miles in
Reach Miles RF1
Modeled Network
16,182 18,324
15,140 16,110
28,904 33,617
90,435 94,525
68,285 71,550
95,098 98,681
60,909 60,909
130,311 130,311
54,228 56,492
68,189 69,524
627,679 650,043
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
3,035 18.8% 16.6%
1,527 10.1% 9.5%
4,483 15.5% 13.3%
37,936 42.0% 40.1%
5,889 8.6% 8.2%
21,442 22.6% 21.7%
6,719 11.0% 11.0%
1,365 1.1% 1.1%
2,088 3.9% 3.7%
4,295 6.3% 6.2%
88,779 14.1% 13.7%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
235 1.5% 1.3%
5 0.1% 0.1%
108 0.4% 0.3%
6,594 7.3% 7.0%
409 0.6% 0.6%
7,185 7.6% 7.3%
1,000 1.6% 1.6%
90 0.1% 0.1%
219 0.4% 0.4%
1,592 2.3% 2.3%
17,436 2.8% 2.7%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
41 0.3% 0.2%
0 0.0% 0.0%
30 0.1% 0.1%
1,288 1.4% 1.4%
85 0.1% 0.1%
2,912 3.1% 3.0%
261 0.4% 0.4%
367 0.3% 0.3%
186 0.3% 0.3%
1,044 1.5% 1.5%
6,214 1.0% 1.0%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
3,310 20.5% 18.1%
1,532 10.1% 9.5%
4,621 16.0% 13.7%
45,817 50.7% 48.5%
6,383 9.4% 8.9%
31,540 33.2% 32.0%
7,980 13.1% 13.1%
1,821 1.4% 1.4%
2,493 4.6% 4.4%
6,931 10.2% 10.0%
112,429 17.9% 17.3%
  Reach miles include both freshwater and estuarine RF1 reaches.
November 2009
                                                                                                                               10-13

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 10-9: Estimated Water Quality Improvements Under Option 31
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Baseline Scenario
Total Reach
Miles in
Reach Miles RF1
Modeled Network
16,182 18,324
15,140 16,110
28,904 33,617
90,435 94,525
68,285 71,550
95,098 98,681
60,909 60,909
130,311 130,311
54,228 56,492
68,189 69,524
627,679 650,043
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
3,717 23.0% 20.3%
2,142 14.2% 13.3%
6,084 21.1% 18.1%
40,092 44.3% 42.4%
7,228 10.6% 10.1%
23,679 24.9% 24.0%
7,473 12.3% 12.3%
3,677 2.8% 2.8%
2,834 5.2% 5.0%
4,405 6.5% 6.3%
101,332 16.1% 15.6%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
301 1.9% 1.6%
5 0.1% 0.1%
163 0.6% 0.5%
8,244 9.1% 8.7%
630 0.9% 0.9%
8,228 8.7% 8.3%
1,300 2.1% 2.1%
111 0.1% 0.1%
282 0.5% 0.5%
1,812 2.7% 2.6%
21,075 3.4% 3.2%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
41 0.3% 0.2%
0 0.0% 0.0%
30 0.1% 0.1%
1,610 1.8% 1.7%
85 0.1% 0.1%
3,487 3.7% 3.5%
319 0.5% 0.5%
376 0.3% 0.3%
222 0.4% 0.4%
1,170 1.7% 1.7%
7,340 1.2% 1.1%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
4,059 25.1% 22.2%
2,147 14.2% 13.3%
6,277 21.7% 18.7%
49,945 55.2% 52.8%
7,943 11.6% 11.1%
35,395 37.2% 35.9%
9,092 14.9% 14.9%
4,164 3.2% 3.2%
3,338 6.2% 5.9%
7,387 10.8% 10.6%
129,747 20.7% 20.0%
  Reach miles include both freshwater and estuarine RF1 reaches.
November 2009
                                                                                                                               10-14

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 10-10: Estimated Water Quality Improvements Under Option 41
EPA
Region
1
2
3
4
5
6
7
8
9
10
Nation
Baseline Scenario
Total Reach
Miles in
Reach Miles RF1
Modeled Network
16,182 18,324
15,140 16,110
28,904 33,617
90,435 94,525
68,285 71,550
95,098 98,681
60,909 60,909
130,311 130,311
54,228 56,492
68,189 69,524
627,679 650,043
Water Quality Improvements by WQI Change
O.OKAWQKO.l
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
2,338 14.5% 12.8%
1,433 9.5% 8.9%
4,620 16.0% 13.7%
38,507 42.6% 40.7%
5,979 8.8% 8.4%
21,693 22.8% 22.0%
6,860 11.3% 11.3%
1,282 1.0% 1.0%
2,166 4.0% 3.8%
3,894 5.7% 5.6%
88,772 14.1% 13.7%
O.KAWQK0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
156 1.0% 0.9%
5 <0.1% <0.1%
108 0.4% 0.3%
7,159 7.9% 7.6%
377 0.6% 0.5%
7,819 8.2% 7.9%
1,205 2.0% 2.0%
90 0.1% 0.1%
197 0.4% 0.4%
1,380 2.0% 2.0%
18,495 3.0% 2.9%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
29 0.2% 0.2%
0 0.0% 0.0%
30 0.1% 0.1%
1,465 1.6% 1.6%
85 0.1% 0.1%
3,313 3.5% 3.4%
280 0.5% 0.5%
367 0.3% 0.3%
186 0.3% 0.3%
943 1.4% 1.4%
6,696 1.1% 1.0%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
2,522 15.6% 13.8%
1,438 9.5% 8.9%
4,758 16.5% 14.2%
47,130 52.1% 49.9%
6,441 9.4% 9.0%
32,825 34.5% 33.3%
8,345 13.7% 13.7%
1,739 1.3% 1.3%
2,548 4.7% 4.5%
6,217 9.1% 8.9%
113,963 18.2% 17.5%
  Reach miles include both freshwater and estuarine RF1 reaches.
November 2009
                                                                                                                              10-15

-------
Environmental Impact and Benefits Assessment for the C&D Category
10.2   Willingness to Pay for Water Quality Improvements

To estimate nonmarket benefits of water quality improvements resulting from the regulation, EPA used a
benefits transfer function based on meta-analysis results presented in Appendix G of this report. The
general approach follows standard methods illustrated by Johnston et al. (2005) and Shrestha et al. (2007),
among many others (see Rosenberger and Phipps 2007). This function allows the Agency to forecast
willingness to pay (WTP) based on assigned values for model variables, chosen to represent a resource
change in the regulation's policy context. EPA's meta-analysis results imply a simple benefit function of
the following general form:

            In(WTP) = intercept + ^ (coefficient; )(Independent Variable Values;)    (Eq. 10-2)

Here, \n(WTP) is the dependent variable in the meta-analysis—the natural log of WTP for water quality
improvements. The metadata include independent variables characterizing specific details of the
resource(s) valued, such as the baseline resource conditions; the extent of resource improvements and
whether they occur in estuarine or freshwater; the geographic region and scale of resource improvements
(e.g., the number of waterbodies); resource characteristics (e.g., baseline conditions, the extent of water
quality change, and ecological services affected by resource improvements); characteristics of surveyed
populations (e.g., users, nonusers); and other specific details of each study. Table 10-11 provides the
estimated regression equation intercept (5.71), variable coefficients (coefficient?), and the  corresponding
independent variable names. Appendix G provides detail on the metadata, model specification, and
justification for the functional form.
EPA assigned a value to each model variable corresponding with theory, characteristics of the water
resources, and sites affected by the regulation and the policy context. Table  10-11 presents a complete list
of assigned variable values.
EPA followed Johnston et al. (2006) in assigning values for methodological attributes (i.e.,  variables
characterizing the study methodology used in the original source studies), which are set at mean values
from the metadata except in cases where theoretical considerations dictate alternative specifications. This
follows general guidance from Bergstrom and Taylor (2006) that meta-analysis benefit transfer should
incorporate theoretical expectations and structures, at least in a weak form. In this instance, three of the
methodology variables, discrete, WQI study, and outlier_bids are all included with an assigned value of
one. Year_index is given the value of 9.68, which corresponds to the mean year that the studies were
conducted, 2002. Nonparam is set to zero because most studies included in metadata used parametric
methods to estimate WTP values. Other study and methodology variables (volunt, mail, lump_sum,
non_reviewed, median_WTP} are assigned a zero value.
EPA used state-specific median household income, as provided by the U.S.  Census Bureau's 2006
American Communities Survey (U.S. Census Bureau 2006a), to assign a value for the income variable
(income). To be consistent with the WTP estimates, which are all estimated in 2008 dollars, EPA used the
Consumer Price Index to adjust the value from the 2006 survey to 2008 dollars. The variable nonusers
was  set to zero because water quality improvements resulting from reduced  sediment discharge from
construction sites would benefit both users and nonusers of the affected resources.
The  regulation is expected to affect water quality at a regional level because construction sites are located
throughout the entire United States. The Agency set a dummy variable denoting multiple regions (mr) to
zero because the expected magnitude of water quality improvement is relatively modest and, as a result,
November 2009                                                                                10-16

-------
Environmental Impact and Benefits Assessment for the C&D Category
EPA's analysis focuses primarily on in-state or local resource improvements. The Mountain Plain
regional dummy variable (mp) is set to zero because the magnitude of the regional effect suggests that
spurious or otherwise unexplained effects (e.g., the effect of specific researchers who appear more than
once in the data) may drive their overall magnitude. Appendix G provides more detail on regression
results.
To account for the regional scale of the water quality effect in freshwater bodies resulting from the
regulation, the variable regional Afresh is assigned a value of one. Other variables relating to waterbody
type (i.e.,  single_lake,  single_river, salt_pond, multiple_river, num_riv_pond) are set to zero. In
estimating WTP for water quality improvements in estuaries and coastal reaches, EPA set all waterbody
type variables (including regionalAfresh) to zero because the default value is a resource  change taking
place in estuaries.19
Water quality improvements resulting from the regulation are likely to enhance a variety of water
resource uses, including fishing, swimming, and boating. Therefore, variables denoting multiple uses
(allmulf) and recreational fishing (fish_use) are assigned a value of one, while the variable denoting
nonspecified uses (nonspec) is set to zero. However, the variable fishplus is given a value of zero because
it is unlikely that the regulation  will cause more than a 50 percent increase in the fish population. Baseline
water quality (baseline) and change in water quality (quality_ch) are assigned state-specific values as
described in Section 10.1. For a broader discussion of issues involved in the specification of variable
levels for  meta-analysis benefit  transfer, see Johnston  et al.  (2005, 2006), among others.
    The water quality model used in this analysis (SPARROW) does not go beyond the terminal reach to estimated water quality
    in large waterbodies, and these waterbodies are not fully characterized. However, EPA did calculate change in WQI for
    several more complex systems such as the Great Lakes and estuaries using ambient water quality data from other sources
    discussed in Section 11.1.4. EPA used mixing zone dilution factors to estimate the expected change in WQI in complex
    waterbodies under the post-compliance scenario (see Chapter 6 for details).
November 2009                                                                                   10-17

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 10-11: Independent Variable Assignments
Variable
Type
Study and
Methodology
Variables
Surveyed
Population
Variable
intercept
year index
discrete
volunt
mail
lump sum
WQI
nonparam
non reviewed
median_WTP
outlier bids
income
nonusers
Coefficient
5.7109
-0.08043
-0.1248
-1.3233
-0.2013
0.5569
-0.3275
-0.6698
-0.2718
-0.5358
-0.8837
0.0000027
-0.4036
Assigned
Value

9.68
1
0
0
0
1
0
0
0
1
Varies
0
Explanation

Set to 9.68 to reflect the mean year that the
studies in the dataset were conducted.
Set to one to reflect survey efforts that
employed discrete choice elicitation methods,
which are preferred over other approaches,
such as open-ended and payment card
methods.
Set to zero because hypothetical voluntary
payment mechanisms are not even potentially
incentive compatible (Mitchell and Carson
1989).
Set to zero because mail surveys are believed
to be of less quality than in-person interviews.
Set to zero because the policy option will be
paid for over a period of years.
Set to one because of the methodological use
of the WQI in the meta-analysis.
Set to zero because most studies used in the
meta analysis used regression analysis to
calculate willingness to pay values.
Set to zero because studies published in peer-
reviewed journals are preferred.
Set to zero because only average or mean
WTP values in combination with the number
of affected households will mathematically
yield total benefits if the distribution of WTP
is not perfectly symmetrical.
Set to one because survey data that exclude
such responses are preferable; outlier bids are
often excluded from the analysis of stated
preference data because these bids (often
identified as greater than a certain percentage
of a respondent's income) may indicate that
a respondent did not consider his or her
budget constraints and or supplementary
goods.
Median annual household income data from
the 2006 American Communities Survey;
median household income values assigned
separately for each state (U.S. Census Bureau
2006a).
Set to zero in order to estimate the total value
for aquatic habitat improvements, including
both use and nonuse values; for nonuser
population, the total value of water resource
improvements includes nonuse values only
(Freeman 2003).
November 2009
                                                                                      10-18

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table 10-11: Independent Variable Assignments
Variable
Type
Waterbody
Type
Variables
Geographic
Region and
Scale
Variables
Resource
Improvement
Variables
Variable
single river
single lake
multiple river
regional Afresh
salt_pond
num riv
pond
mr
mp
allmult
nonspec
Inquality ch
fish use
fishplus
Inbase
Coefficient
-0.4279
-0.06316
-1.4752
0.1588
0.9849
0.1173
-0.8846
1.6337
-0.3728
-0.4042
0.4065
-0.3317
0.4432
0.02610
Assigned
Value
0
0
0
Varies
0
0
0
0
1
0
Varies
1
0
Varies
Explanation
For valuing improvement in freshwater
bodies, regional Afresh is set to one because
the expectation is that multiple freshwater
bodies within a state would be affected by the
regulation. For saltwater reaches, all
waterbody type variables are set to zero
because the default value in the model is a
resource change taking place in estuaries.
Indicates the number of rivers or salt ponds
affected by a policy, and is set to zero because
this analysis assumes that the effluent
guidelines will affect the entire
watershed/region; this variable assignment is
constant across study regions.
Regional variables are omitted from the
predictive portion of the analysis (i.e., set to
zero) because the regression results suggest
that these variables may be picking spurious
or other unexplained effects (e.g., author's
effect).
Set to one because it is assumed that multiple
species would benefit from water quality
improvements
Set to zero because it is assumed that multiple
species would benefit from water quality
improvements.
Set to the natural log of average change in the
WQI for a given analytic scenario.
Set to one because the regulation is expected
to benefit a variety of aquatic species and
therefore enhance recreational fishing
opportunities.
Set to zero because the regulation is not
expected to result in a fish population change
of 50 percent or greater.
Set to the natural log of the average base
water quality for a given state.
Economic values of water quality improvements are calculated at the state level and organized by analytic
scenario. EPA calculated the state-level average of baseline and post-compliance WQI changes
corresponding to 12 analytic scenarios reflecting combinations of four levels of baseline water quality
conditions (WQI< 25, 25
-------
Environmental Impact and Benefits Assessment for the C&D Category
                                   WTP = exp(bi_WrP + oe2/2)                          (Eq. 10-3)

       Where:

       exp(-)   = the exponential operator

       In  WTP = the predicted natural log of WTP for a given analytic scenario

       oe2      = the model residual variance (0.1876) taken from Table G-3 in Appendix G.

The total WTP regression model presented above can be used to predict WTP for each of the studies in
the database; however, estimates derived from regression models are subject to some degree of error and
uncertainty. To better characterize the uncertainty or error bounds around predicted WTP, EPA used a
procedure described by Krinsky and Robb (1986). The procedure involves sampling the variance-
covariance matrix of the estimated coefficients, which is a standard output of the statistical package used
to estimate the meta-analysis model. WTP values are then calculated for each  drawing from the variance-
covariance matrix and an empirical distribution of WTP values is constructed. By varying the number of
drawings, it is possible to generate an empirical distribution with a desired degree of accuracy (Krinsky
and Robb 1986). The low and high estimate of WTP values is then identified based on the 10th and 90th
percentile of WTP values from the empirical distribution. These bounds may help decision-makers
understand the uncertainty associated with the benefit results.
EPA used the Krinsky and Robb (1986) procedure to estimate the 10th, 50th, and 90th percentiles of total
WTP for each EPA region, based on the results of the total WTP regression model.  Table 10-12 presents
the results of these calculations. The Agency notes that this analysis provides confidence limits for WTP
estimates related to the covariance matrix of meta-analysis parameter estimates. It does not, however,
assess the sensitivity of results to changes in meta-regression model assumptions or specifications (cf
Johnston et al. 2005, 2006) or assumptions implied in benefit aggregation (cf.  Loomis  1996; Loomis et al.
2000; Bateman et al. 2006). As noted above, however, the Agency has to the extent practicable made
assumptions and specifications that lead to conservative benefit estimates.
Table 10-12 presents 10th, 50th, and 90th percentiles of household WTP for water quality improvements
resulting form reduced sediment discharge from construction sites by EPA region and policy option.
November 2009                                                                                10-20

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 10-12: Estimates of Annual Household Willingness to Pay for Water Quality Improvement by Region and Policy
Option (2008$)
Option 1
EPA Low
Region 10%
1 $0.14
2 $0.04
3 $0.09
4 $1.14
5 $0.09
6 $1.19
7 $0.42
8 $0.45
9 $0.17
10 $0.64
National $0.49
Mid
50%
$0.87
$0.30
$0.56
$4.39
$0.49
$3.98
$1.62
$0.80
$0.64
$2.20
$1.85
High
90%
$1.96
$0.66
$1.24
$9.07
$1.07
$7.87
$3.32
$1.25
$1.30
$4.41
$3.78
Option 2
Low
10%
$0.36
$0.17
$0.28
$1.90
$0.21
$1.88
$0.70
$0.47
$0.26
$0.98
$0.83
Mid
50%
$2.13
$1.10
$1.71
$7.05
$1.13
$5.83
$2.63
$0.93
$0.98
$3.04
$3.10
High
90%
$4.74
$2.45
$3.79
$14.37
$2.45
$11.28
$5.34
$1.52
$2.01
$5.93
$6.33
Option 3
Low
10%
$0.50
$0.25
$0.40
$2.17
$0.27
$2.16
$0.82
$0.54
$0.30
$1.10
$0.97
Mid
50%
$2.94
$1.63
$2.40
$7.87
$1.43
$6.59
$2.97
$1.29
$1.16
$3.33
$3.63
High
90%
$6.54
$3.66
$5.31
$15.96
$3.09
$12.68
$5.98
$2.29
$2.37
$6.42
$7.41
Option 4
Low
10%
$0.30
$0.16
$0.29
$1.96
$0.22
$2.05
$0.76
$0.47
$0.26
$0.84
$0.86
Mid
50%
$1.82
$1.08
$1.76
$7.23
$1.16
$6.23
$2.80
$0.91
$0.99
$2.68
$3.17
High,
90%
$4.08
$2.42
$3.90
$14.73
$2.51
$11.98
$5.64
$1.48
$2.02
$5.26
$6.45
November 2009
                                                                                                                 10-21

-------
Environmental Impact and Benefits Assessment for the C&D Category
As shown in Table 10-12 the mid-point estimates of household WTP for water quality improvements
resulting from the regulation range from $1.85 to $3.63 per household per year under different policy
options. The estimated WTP values vary greatly across EPA regions depending on the expected water
quality improvements and water resource and population characteristics in a given region.
Option 1 is estimated to improve ambient water quality in 11.6 percent of RF1 reach miles modeled in the
analysis. This option results in the least water quality improvements across all EPA regions. EPA Region
4 is expected to the have the greatest improvements in water quality under this option. The estimated
annual WTP per household in Region 4 is $4.39 (mid-point estimate). Region 2 has the smallest
household WTP, with a mid-point estimate of $0.30 per household. Nationwide, the estimated annual
household WTP is $1.85 (mid-point estimate).

Option 2 is estimated to improve ambient water quality in 17.9 percent of RF1 reach miles modeled. The
estimated national average WTP for water quality improvements resulting from the regulation is $3.10
per household per year. Regions 4 and 6 are estimated to see improvements in water quality in 50.7 and
33.2 percent of their total RF1 reach miles (Table 10-8). EPA's analysis suggests that Region 4
households would be willing to  pay the most, with a mid-point estimate of $7.05 per household per year,
for water quality improvements  resulting from the regulation. Region 6 has the second largest household
WTP of $5.83. Conversely, households located in Region 2 are estimated to have the lowest WTP for
water quality improvements from the regulation, $1.10.
Policy Option 3 yields the most significant results overall in terms of reach miles expected to improve
under the post-compliance scenario (20.7 percent).  The estimated scale of improvements ranges from
3.2 percent to 55.6 percent of total reach miles in Region 8 and 4, respectively (Table 10-9). Nationwide,
the estimated annual per-household WTP for water quality improvements under Option 3 is $3.63 (mid-
point estimate).
Option 4 is estimated to generate improvements in  18.2 percent of reach miles nationally. These
improvements result in a national household WTP of $3.17 per year (mid-point estimate).. As with the
other options, Region 4 has the largest annual household WTP value with a mid-point estimate of $7.23.
Conversely, EPA Region 8 has the lowest WTP value with a mid-point estimate of $0.91.

10.3    Estimating  Total WTP for Water Quality Improvements

For each policy option, EPA calculated reach-level WTP as follows. First EPA estimated mean state-level
per-household WTP for each combination of the baseline water quality category (WQI baseUne) and the
expected change in WQI (AWQI). Then, the Agency assigned each reach in the  analysis a mean
household WTP value  based on reach location, baseline water quality, and change in water quality. The
WTP was then multiplied  by the number of households in a given state in 2006  and the percentage of
reach miles in that state that comprise a given reach. The number of households per state was calculated
by taking U.S. Census Bureau population estimates for 2006 for each state and dividing by the average
number of people per household for a given state as reported in U.S. Census Bureau (2006a, 2006b). The
total WTP equation for each reach is provided below  (Equation 10-4):

               TWTPreach = WTP(WQIbaselme, AWQI) x StateHHx PercentRivrMiles)      (Eq. 10-4)

        Where:
        TWTP'reach          = the reach-level welfare change from improved water quality
November 2009                                                                              10-22

-------
Environmental Impact and Benefits Assessment for the C&D Category
       WTP
       StateHH
= the estimated state-level per-household WTP for water quality
  improvement for a given combination of the baseline water quality
  category (WQI baseiine) and the expected change in water quality under the
  post-compliance scenario (AWQI)
= the number of households in a given state
       PercentRiverMiles  = the percentage of total reach miles in a given state that are comprised of a
                            given reach.
Finally, EPA aggregated reach-level benefits to the regional level. The regional benefits for the ten EPA
regions were then combined to calculate the national benefit of the regulation. Table 10-13 presents
estimated benefits of the regulation by EPA region and policy option.
November 2009
                                                                                            10-23

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table 10-13 :Regional Willingness to Pay for Water Quality Improvement (Millions 2008$)
Option 1
Lower
EPA 10%
Region Bound
1 $0.81
2 $0.45
3 $1.07
4 $26.41
5 $1.78
6 $15.62
7 $2.28
8 $1.76
9 $2.67
10 $2.90
National $55.75
Mean
$4.93
$3.10
$6.45
$101.44
$9.92
$52.44
$8.86
$3.17
$9.95
$10.01
$210.27
Upper
90%
Bound
$11.10
$6.94
$14.23
$209.34
$21.64
$103.61
$18.12
$4.91
$20.26
$20.09
$430.24
Option 2
Lower
10%
Bound
$2.06
$1.76
$3.27
$43.94
$4.27
$24.72
$3.84
$1.85
$4.04
$4.45
$94.19
Mean
$12.07
$11.50
$19.70
$162.70
$22.97
$76.74
$14.37
$3.66
$15.32
$13.86
$352.90
Upper
90%
Bound
$26.84
$25.67
$43.59
$331.84
$49.85
$148.49
$29.15
$5.97
$31.28
$27.03
$719.72
Option 3
Lower
10%
Bound
$2.84
$2.63
$4.58
$50.10
$5.43
$28.45
$4.49
$2.12
$4.72
$5.03
$110.39
Mean
$16.67
$17.14
$27.61
$181.69
$29.01
$86.74
$16.24
$5.08
$18.09
$15.16
$413.41
Upper
90%
Bound
$37.03
$38.36
$61.08
$368.63
$62.88
$166.91
$32.67
$9.03
$36.97
$29.25
$842.81
Option 4
Lower
10%
Bound
$1.71
$1.71
$3.36
$45.36
$4.37
$27.01
$4.17
$1.84
$4.08
$3.82
$97.44
Mean
$10.32
$11.30
$20.30
$167.04
$23.52
$82.05
$15.27
$3.60
$15.43
$12.21
$361.04
Upper
90%
Bound
$23.09
$25.35
$44.94
$340.13
$51.02
$157.72
$30.81
$5.83
$31.47
$23.97
$734.34
November 2009
                                                                                                                    10-24

-------
Environmental Impact and Benefits Assessment for the C&D Category
EPA estimates that total annual benefits of water quality improvements resulting from reduced sediment
discharge from construction sites range from $210.3 million under Option 1 to $413.4 million under
Option 3. The estimated mean regional benefits vary from $3.1 to $ 181.7 million per year, depending on
the level of construction activity, average rainfall, water resource characteristics, and resident population
in a given region and stringency of the policy option.
As shown in Table 10-13, Option 1 generates the least water quality improvements of the four regulatory
options. Thus, this option yields the smallest benefits at the regional and national levels. The national
benefits of water quality improvements under this option range from $55.8 million to $430.2 million per
year, with a mid-point estimate of $210.3 million. Region 4 gains the most benefit from water quality
improvement, with a total value of $101.4 million per year (48.2 percent of the total national benefits).
Region 6 has the second largest benefits ($52.4 million per year), which account for 24.9 percent of the
total national benefits.
Under Option 2, the mid-point estimate of national benefits is $352.9 million per year, with low and high
estimates of $94.2 million and $719.7 million per year. Region 4 gains the most from water quality
improvements resulting from the regulation  ($162.7 million per year). EPA Region 8 receives the smallest
annual benefits, $3.7 million
Under Option 3, the estimated national benefits of water quality improvement from the regulation are
$413.4 million per year, with alow estimate of $110.4 and a high estimate of $842.8 million. As with the
other policy options Region 4 receives the largest benefits from water quality improvements, accounting
for 43.9 percent ($181.7 million per year) of the total national benefits. Region 8 is estimated to gain least
under Option 3, with the total regional benefits estimated at $5.1 million per year.
Option 4 is expected to generate national benefits of $361.0 million per year with a low estimate of $97.4
million and a high estimate of $734.3 million. EPA Region 4 gains the largest benefits, with a total of
$159.9 million per year. Region 8 has the smallest improvements in water quality under this option, with
improvements in only 1.3 percent of reach miles, and thus has the smallest annual benefits ($3.6 million).

10.4    Uncertainty and Limitations

A number of issues are common to all benefit transfers. Benefit transfer involves adapting research
conducted for another purpose in the available literature to address the policy questions at hand. Because
benefits analysis of environmental regulations rarely affords enough time to develop original stated
preference surveys that are specific to the policy effects, benefit transfer is often the only option to inform
a policy decision. As a result, they are nearly universal in benefit-cost analyses (Smith et al. 2002).
Benefit transfers are by definition characterized by a difference between the context in which resource
values are estimated and that in which benefit estimates are desired (Rosenberger and Phipps 2007). The
ability of meta-analysis to adjust for the influence of study, economic, and resource  characteristics on
WTP can minimize, but not eliminate, potential biases (Smith et al. 2002; Rosenberger and Stanley 2006;
Rosenberger and Phipps 2007). As is typical in applied benefit transfers, the meta-analysis model used in
this analysis provides a close, but not perfect, match to the context in which values are desired. Therefore,
some beneficial effects associated with reducing sediment discharges from construction sites are not
accounted for in the estimated WTP values.  For example, surface water valuation studies included in
meta-data focused primarily on changes in ambient water quality. Other effects associated with sediment
discharges (e.g., sedimentation of river beds, stability and erosion of river banks, and changes in the
November 2009                                                                                 10-25

-------
Environmental Impact and Benefits Assessment for the C&D Category
stream carrying capacity) were not included in the original study scenarios and, as a result, from the
estimated WTP for reducing sediment discharges from construction sites.
Some related and additional limitations inherent to the meta-analysis model and the subsequent benefit
transfer include:
        >  The Agency notes, as detailed by Loomis (1996), Loomis et al. (2000) and Bateman et al.
           (2006), among others, that there are numerous uncertainties and associated assumptions
           required to aggregate WTP across spatial jurisdictions. While these uncertainties are well
           known, the literature does not agree on appropriate, standardized guidance for benefit
           aggregations, and applied benefit-cost analysis almost universally requires simplifying
           assumptions in order to generate defensible welfare aggregations. In an ideal context, analysts
           would have information necessary to estimate spatially referenced distance decay
           relationships for all changes resulting from policies under consideration (cf Bateman et al.
           2006). However, the Agency notes that even the most advanced literature provides only
           simple illustrations of such issues, and none methodologically sufficient to support regulatory
           analysis. In analyzing benefits of the  final regulation, EPA assumed that households would
           gain no benefits from water quality improvements in aquatic resources located outside  of
           their state of residence. As a result, the population considered in the benefits analysis of the
           regulation does not represent all the households that are likely to hold values for water
           resources in a given state. Residents of other states may hold values for water resources
           outside of their home state, in particular if such resources have personal, regional, or national
           significance. Even if per household WTP for out-of-state residents are small they can be very
           large in the aggregate if these values are held by a substantial fraction of the population
        ^  Some resource valuation studies have found that respondents in the typical contingent market
           situation may overstate their WTP compared to their likely behavior in a real-world situation.
           However, the magnitude of hypothetical bias on the estimated WTP is uncertain. Following
           standard benefit transfer approaches,  including meta-analytic transfers, this analysis proceeds
           under the assumption that each source study provides a valid, unbiased estimate of the
           welfare measure under consideration (cf. Moeltner et al. 2007; Rosenberger and Phipps
           2007). To minimize potential  hypothetical bias, EPA set independent variable values to
           reflect best benefit transfer practices.
        >  The estimation of WTP may be sensitive to differences in the environmental water quality
           measures. Studies that did not use the WQI were mapped to the WQI so a comparison could
           be made across studies. The dummy variable (WQI) captures the effect of a study using
           (WQI=Y) or not using the WQI (WQI=0). It was found that studies that did not use the WQI
           had lower WTP values. This may indicate that there may have been some systematic biases in
           the mapping of studies that did not use the WQI. In analyzing the benefits of this regulation,
           EPA set the WQI to one to reduce uncertainty in WTP estimates  associated with studies that
           did not include WQI as a native survey instrument.  See Appendix G for a detailed discussion
           of water quality measures used in the original studies included in the meta-analysis.
        >  Transfer error may occur when benefit estimates from a study site are adopted to forecast the
           benefits of a policy site. Rosenberger and Stanley (2006) define transfer error as the
           difference between the transferred and actual, generally unknown, value. While meta-analysis
           is fairly accurate when estimating benefit function, transfer error may be a problem in cases
           where the sample size is small. While meta-analyses have been shown to outperform other
November 2009                                                                                 10-26

-------
Environmental Impact and Benefits Assessment for the C&D Category
           function-based transfer methods in many cases, this result is not universal (Shrestha et al.
           2007). This notwithstanding, results reviewed by Rosenberger and Phipps (2007) are "very
           promising" for the performance of meta-analytic benefit transfers relative to alternative
           transfer methods.
Additional limitations and uncertainties are associated with the use of WQI to link water quality changes
from reduced sediment discharges to effects on human uses and support for aquatic and terrestrial species
habitat:
        >  The estimated changes in WQI reflect only water quality improvements resulting directly
           from reductions in total  suspended solid and nutrient loadings. They do not include
           improvements in water quality indicators indirectly associated with pollutant loadings, (e.g.,
           dissolved oxygen). This is likely to result in underestimation of the expected water quality
           changes resulting from the proposed regulation because the combined impact of several
           pollutants on ambient water quality is likely to be greater than the sum of the individual
           impacts of reducing concentrations of sediments and nutrients.
        >  The WQI index used in  EPA's analysis uses TSS concentrations as a proxy for water
           turbidity. It does not include turbidity as a separate parameter in calculation of the WQI. This
           omission may understate the expected change in WQI resulting from the final regulation.
        >  The methodology used to translate in-stream sediment and nutrient concentrations into sub-
           index scores employs nonlinear transformation curves. Water quality changes that fall outside
           of the sensitive part of the transformation curve (i.e., above/below the upper/lower bounds)
           yield no benefit in the analysis.
Limitations and uncertainty associated with the use of SPARROW are discussed in Chapter 6 of this
report.
November 2009                                                                                 10-27

-------
Environmental Impact and Benefits Assessment for the C&D Category
       Total Estimated Benefits
This chapter summarizes findings from EPA's analysis of the expected benefits from the regulation. EPA
considered a wide range of policy options in developing this regulation and presents here the three policy
options for which a full cost-benefit analysis was performed. These options are:
       >  Option 1: nonnumeric effluent limitations on all sites.
       >  Option 2: in addition to the requirements of Option 1, active sediment treatment is required
           on sites with 30 or more acres disturbed at one time and treated discharge is subject to a
           turbidity standard of 13 NTU.
       >  Option 3: in addition to the requirements of Option 1, active sediment treatment is required
           on sites with 10 or more acres disturbed at one time and treated discharge is subject to a
           turbidity standard of 13 NTU.
       >  Option 4: in addition to the requirements of Option 1, passive sediment treatment is required
           on sites with 10 or more acres disturbed at one time, and treated discharge is subject to a
           turbidity standard of 280 NTU.
Chapter 1 describes these policy options in more detail.

11.1  Summary of the Estimated Benefits

The Agency analyzed benefits of the regulation in four benefits categories: navigation, water storage,
drinking water treatment, and nonmarket benefits of water quality improvement. The previous chapters of
this report provide the details of the methods and data used in the analysis of the four monetized benefit
categories (i.e., list benefit categories). See Chapters 7, 8, and 9 for a discussion of the avoided cost
methods used to estimate benefits to navigation, water storage, and treatment. Chapter 10 and Appendix  G
provide a discussion of the methods used to estimate the willingness to pay (WTP) for water quality
improvements resulting from the policy options considered for the regulation.
EPA was unable to quantify benefits stemming from:
       >  Reduced damages to industrial and agricultural water users
       >  Improved market value of properties near surface waters
       >  Reduced cost of stormwater system maintenance and flood damages
       >  Benefits to commercial fishing and shellfishing
       >  Reduced cost of drinking water treatment to improve taste and odor resulting from
           eutrophication of drinking water sources.
Chapter 5 of this report describes these benefits qualitatively. In addition, limitations inherent in EPA's
analysis of water quality improvements and monetization of benefits are likely to result in understatement
of benefits resulting from the final regulation. Section 77.2 below summarizes key limitations of EPA's
analysis of benefits stemming from reducing construction site discharges..
Table 11-1 presents low, midpoint, and high estimates of benefits under each policy option, consisting of
benefits to navigation, water storage, drinking water treatment, and WTP. Table 11-2, Table 11-3, Table
November 2009                                                                                11-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
11-4,and Table 11-5 detail total benefits by EPA region, policy option, and estimate range. It should be
noted that these tables incorporate the confidence intervals of 10, 50, and 90 percent from the WTP
analysis into the low, mid, and high sensitivity analyses performed for the avoided cost estimates. Though
these  are conceptually different, they are both intended to present a range of values to account for some of
the uncertainty inherent in these estimates. The sensitivity analyses create a range by varying EPA's
assumptions underlying the analysis, while the confidence interval presents high and low bounds from the
meta-analysis regression.
All tables present benefits for navigable waterway and reservoir dredging calculated using both 3 and 7
percent discount rates. Because the discount rate only applies to two of the four monetized benefits
categories, which represent at most 5 percent of total benefits, varying it has little effect on the total
benefits estimate. EPA calculated benefits for drinking water treatment and WTP using a single-year
timeframe, which did not require  discounting or additional calculations to present annual values. All
benefits presented reflect annual values. The remaining discussion presents the benefits estimates
assuming a 3 percent discount rate; the associated tables present results for both discount rates.
Total national benefits vary significantly among the three regulatory options. Under Option 1, the
estimated benefits range from approximately $59.0 million to approximately $434.2 million, with a
midpoint estimate of $214.0 million. Estimated avoided costs range from $3.3 million to $4.1 million,
with a midpoint of $3.8 million, and WTP ranges from $55.8 to $430.2 million , with a midpoint estimate
of $210.3 million.
For Option 2, the estimated benefits  range from $100.5 million to  $727.4 million, with a midpoint
estimate of $360.1 million. The estimated WTP for water quality improvements from reduced sediment
discharges from construction sites under Option 2 ranges from $94.2 to $719.7 million, with a midpoint
value of $352.9 million. Estimated cost savings range from $6.3 million to $7.7 million per year, with a
midpoint estimate of $7.2 million.
Under Option 3, total benefits are estimated to be between $118.0  and $852.2 million, with a midpoint
estimate of $422.3 million. The avoided costs are estimated to be between $7.7 and $9.4 million per year,
with a midpoint estimate of $8.9 million. WTP under Option 3 ranges from $110.4 million to $842.8
million , with a midpoint estimate of $413.4 million.
Under Option 4, the final regulation, the estimated benefits range from $104.3 million to $742.7 million,
with a midpoint estimate of $368.9 million. Nonmarket benefits estimated based on household WTP for
surface water quality improvements account for 93,  98, and 99 percent of total benefits from the
regulation in the low, mid, and high estimates, respectively. The estimated WTP for water quality
improvements from reduced sediment discharges from construction sites under Option 4 ranges from
$97.4 to $734.3 million, with a midpoint estimate of $361.0 million. The estimated cost savings to
industry and government through reduced costs of navigable waterway maintenance, reservoir dredging,
and drinking water treatment range from $6.8 million to $8.3 million per year, with a midpoint estimate of
$7.9 million. Under Option 4, avoided cost benefits  account for 7, 2, and 1 percent of total benefits in the
low, mid, and high estimates, respectively. Because  this option requires passive treatment at sites with
more  than 10 acres of land disturbed and establishes a numeric effluent limit, its benefits are more than
double those of Option 1, which does not establish numeric criteria for sediment discharge. It also
produces more benefits than Option 2, which requires active  treatment of sediment but on fewer sites.
Benefits under Option 4 are lower than those under  Option 3, which would require active sediment
treatment on the same sites where Option 4 requires passive treatment, which is less burdensome.
November 2009                                                                                 11-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
     Table 11-1: Annual Total National Benefits by Benefits Category (millions of
     2008$)
Benefit Category
3% Discount Rate
Low
Mid
High
7% Discount Rate
Low
Mid
High
                                              Option 1
Navigation
Water Storage1
Drinking Water1
Avoided Costs
WTP1
Total2
$1.0
$1.3
$1.0
$3.3
$55.8
$59.0
$1.3
$1.4
$1.2
$3.8
$210.3
$214.1
$1.3
$1.5
$1.3
$4.1
$430.2
$434.3
$1.0
$1.1
$1.0
$3.1
$55.8
$58.8
$1.2
$1.2
$1.2
$3.7
$210.3
$213.9
$1.3
$1.4
$1.3
$4.0
$430.2
$434.2
                                              Option 2
Navigation
Water Storage1
Drinking Water1
Avoided Costs
WTP1
Total2
$2.1
$2.7
$1.4
$6.3
$94.2
$100.5
$2.6
$2.9
$1.8
$7.2
$352.9
$360.1
$2.8
$3.0
$1.9
$7.7
$719.7
$727.4
$2.1
$2.2
$1.4
$5.8
$94.2
$99.9
$2.5
$2.6
$1.8
$6.9
$352.9
$359.8
$2.7
$3.0
$1.9
$7.5
$719.7
$727.2
                                              Option 3
Navigation
Water Storage1
Drinking Water1
Avoided Costs
WTP1
Total2
$2.7
$3.3
$1.7
$7.7
$110.4
$118.0
$3.3
$3.6
$2.1
$8.9
$413.4
$422.3
$3.4
$3.8
$2.1
$9.4
$842.8
$852.2
$2.6
$2.8
$1.7
$7.0
$110.4
$117.4
$3.2
$3.2
$2.1
$8.4
$413.4
$421.8
$3.4
$3.7
$2.1
$9.2
$842.8
$852.0
                                              Option 4
Navigation
Water Storage1
Drinking Water1
Avoided Costs
WTP1
Total2
$2.4
$3.0
$1.5
$6.8
$97.4
$104.3
$2.9
$3.2
$1.8
$7.9
$361.0
$368.9
$3.0
$3.4
$1.9
$8.3
$734.3
$742.7
$2.3
$2.5
$1.5
$6.3
$97.4
$103.7
$2.8
$2.9
$1.8
$7.5
$361.0
$368.5
$3.0
$3.3
$1.9
$8.2
$734.3
$742.5
     These savings were calculated for a one-year timeframe, did not require discounting, and are equal under both discount rates.
    2 Totals may not equal the sum of categories due to rounding.

Table 11-2, Table 11-3, Table  11-4, and Table 11-5 detail total monetized benefits (including benefits to
navigation, water storage, drinking water, and water quality) by region. Region 4 benefits the most from
this regulation under all policy options, as it experiences the most widespread changes in terms of
improved reach miles and the most significant reductions in sediment concentrations in these reach miles.
This leads to higher WTP estimates in Region 4, which account for the largest proportion of benefits.
Region 6 benefits second most, though monetized benefits in this region are about half of those in Region
4. Regions 4 and 6 together account for more than half of the benefits under all of the options. For Region
4, midpoint benefits estimates are $102.4, $164.7, $184.2, and $169.2 million, respectively under the four
policy options.  For Region 6, midpoint benefits estimates for the four options are $54.1, $80.1 $90.9, and
$86.0 million, respectively.
November 2009
                                                                                               11-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
       Table 11-2: Annual Total National Benefits Under Option 1 (millions of 2008$)
EPA
Region
1
2
3
4
5
6
7
8
9
102
Total3
Avoided
Low
$0.8
$0.5
$1.1
$27.3
$2.3
$17.1
$2.3
$1.8
$2.7
$3.1
$59.0
Dredging Costs
30/01
Mid
$5.0
$3.2
$6.5
$102.4
$10.6
$54.1
$8.9
$3.2
$10.0
$10.3
$214.1
Discounted at
High
$11.1
$7.1
$14.3
$210.4
$22.3
$105.3
$18.2
$4.9
$20.3
$20.4
$434.3
Avoided
Low
$0.8
$0.5
$1.1
$27.2
$2.3
$16.9
$2.3
$1.8
$2.7
$3.1
$58.8
Dredging Costs
1%1
Mid
$5.0
$3.2
$6.5
$102.4
$10.6
$54.0
$8.9
$3.2
$10.0
$10.2
$213.9
Discounted at
High
$11.1
$7.1
$14.3
$210.4
$22.3
$105.3
$18.2
$4.9
$20.3
$20.4
$434.2
        Only avoided costs of dredging navigable waterways and reservoirs required discounting and annualization. Avoided costs of
       drinking water treatment and willingness-to-pay were estimated on a single-year basis.
       2 Benefits estimates in this region are not zero, but less than $500 annually.
       3 Totals may not be equal to the sum of regional data because the WTP model estimates the national total rather than summing
       regional totals.
        Table 11-3: Annual Total National Benefits Under Option 2 (millions of 2008$)
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total2
Avoided
Low
$2.1
$1.8
$3.4
$45.7
$4.9
$27.7
$4.0
$1.9
$4.2
$4.8
$100.5
Dredging Costs
30/01
Mid
$12.1
$11.6
$19.8
$164.7
$23.7
$80.1
$14.5
$3.7
$15.5
$14.4
$360.1
Discounted at
High
$26.9
$25.8
$43.7
$333.9
$50.6
$152.0
$29.3
$6.0
$31.5
$27.6
$727.4
Avoided
Low
$2.1
$1.8
$3.4
$45.5
$4.9
$27.4
$4.0
$1.9
$4.2
$4.8
$99.9
Dredging Costs
1%1
Mid
$12.1
$11.6
$19.8
$164.6
$23.7
$79.9
$14.5
$3.7
$15.5
$14.4
$359.8
Discounted at
High
$26.9
$25.8
$43.7
$333.9
$50.6
$152.0
$29.3
$6.0
$31.5
$27.6
$727.2
        1 Only avoided costs of dredging navigable waterways and reservoirs required discounting and annualization. Avoided costs
        of drinking water treatment and willingness-to-pay were estimated on a single-year basis.
        2 Totals may not be equal to the sum of regional data because the WTP model estimates the national total rather than summing
        regional totals.
November 2009
                                                                                                             11-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
       Table 11-4: Annual Total National Benefits Under Option 3 (millions of 2008$)
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total2
Avoided
Low
$2.9
$2.7
$4.7
$52.3
$6.1
$32.2
$4.7
$2.1
$4.9
$5.5
$118.0
Dredging Costs
30/01
Mid
$16.7
$17.3
$27.8
$184.2
$29.8
$90.9
$16.4
$5.1
$18.3
$15.8
$422.3
Discounted at
High
$37.1
$38.5
$61.2
$371.2
$63.7
$171.3
$32.9
$9.1
$37.2
$30.0
$852.2
Avoided
Low
$2.9
$2.7
$4.7
$52.1
$6.0
$31.9
$4.6
$2.1
$4.9
$5.5
$117.4
Dredging Costs
70/01
Mid
$16.7
$17.3
$27.8
$184.0
$29.7
$90.7
$16.4
$5.1
$18.3
$15.8
$421.8
Discounted at
High
$37.1
$38.5
$61.2
$371.2
$63.7
$171.2
$32.9
$9.1
$37.2
$30.0
$852.0
       1 Only avoided costs of dredging navigable waterways and reservoirs required discounting and annualization. Avoided costs
       of drinking water treatment and willingness-to-pay were estimated on a single-year basis.
       2 Totals may not be equal to the sum of regional data because the WTP model estimates the national total rather than summing
       regional totals.
       Table 11-5: Annual Total National Benefits Under Option 4 (millions of 2008$)
EPA
Region
1
2
3
4
5
6
7
8
9
10
Total2
Avoided
Low
$1.7
$1.8
$3.5
$47.3
$5.0
$30.5
$4.3
$1.9
$4.2
$4.1
$104.3
Dredging Costs
30/01
Mid
$10.4
$11.4
$20.4
$169.2
$24.2
$86.0
$15.4
$3.6
$15.6
$12.6
$368.9
Discounted at
High
$23.1
$25.5
$45.1
$342.4
$51.8
$161.8
$31.0
$5.8
$31.7
$24.4
$742.7
Avoided
Low
$1.7
$1.8
$3.5
$47.1
$5.0
$30.2
$4.3
$1.9
$4.2
$4.1
$103.7
Dredging Costs
To/o1
Mid
$10.4
$11.4
$20.4
$169.1
$24.2
$85.8
$15.4
$3.6
$15.6
$12.6
$368.5
Discounted at
High
$23.1
$25.5
$45.1
$342.4
$51.8
$161.8
$31.0
$5.8
$31.7
$24.4
$742.5
       1 Only avoided costs of dredging navigable waterways and reservoirs required discounting and annualization. Avoided costs
       of drinking water treatment and willingness-to-pay were estimated on a single-year basis.
       2 Totals may not be equal to the sum of regional data because the WTP model estimates the national total rather than summing
       regional totals.
11.2   Sources of Uncertainty and  Limitations
EPA notes that quantifying and monetizing impacts of reducing sediment discharges from construction
sites is challenging. As a result, total national benefits estimates of the regulation are subject to the
limitations and uncertainties inherent in the valuation approaches used for assessing benefits to
navigation, water storage, drinking water treatment, and nonmarket benefits of water quality
improvement. Because the combined effect of these limitations and uncertainties is likely to
November 2009
                                                                                                    11-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
underestimate the national level of benefits of this regulation, the estimated benefits should be interpreted
in the context of these limitations.
The preceding sections of this report discuss specific limitations and uncertainties associated with
estimating benefits to navigation, water storage, drinking water treatment, and nonmarket benefits of
water quality improvement. This section summarizes the limitations inherent in EPA's analysis of
reducing discharges from the construction sites.

11.2.1 Water Quality Model Limitations

To estimate benefits of reduced sediment loadings to surface water, EPA relied on SPARROW. The
SPARROW model for suspended sediments has a number of limitations, some of which are inherent to
the methodology and some the result of the particular model application. The key model limitations are:
       >  Reliance on the Reach File 1 network. While the RF1 network provides reasonably
           comprehensive national coverage of major rivers, streams, and other surface waterbodies,
           coverage is limited in certain important respects. First, RF1 network coverage is limited to
           the coterminous United States, thus excluding Alaska and Hawaii. In addition, while the RF1
           l:500,000-scale network reaches have associated data or estimates of stream discharge and
           velocity that are required to specify the SPARROW model, the network excludes the majority
           of the nation's total stream mileage, and smaller streams in particular. The linear coverage of
           the RF1 network is approximately 700,000 miles (USEPA 2007e). By contrast, coverage of
           the USGS National Hydrographic Dataset, at 1:24,000 - 1:100,000 scale, is currently over 7
           million miles (USGS 2007b). Given that RF1 accounts only for 10 percent of the total reach
           miles, the impacts of construction-related sediment on smaller stream reaches  are likely to be
           significantly understated. As construction activities may be concentrated along lower-order
           streams not included in the RF1 network, the relative share of total sediments contributed by
           construction activities may be high on these reaches during active construction phases. By
           contrast, the specific impacts of construction activities may diminish in importance relative to
           contributions from spatially extensive and diffuse land uses, including agriculture, at the level
           of RF1 reaches. Moreover, approximately 20-30 percent of sediment discharged from
           construction sites does not reach the RF1 network and is instead retained on the land surface
           and in smaller streams and reservoirs. These loads are omitted from the current analysis.
       >  Estimation of changes in nutrient concentrations associated with changes in sediment loading.
           The approach used to estimate changes in nutrient concentration resulting from reduction in
           sediment loadings is based on modeled long-term relationships between sediment and
           nutrient loadings within each of the modeled watersheds. This assumption follows
           observations from case studies discussed in Chapter 4 of this report, which suggest a
           correlation between sediment loadings from construction sites and elevated nitrogen and
           phosphorus loadings. By using a fixed relationship, the approach assumes that nutrients are
           bound to the sediments and that methods used to reduce sediment runoff from construction
           sites are equally effective in reducing nitrogen and phosphorus runoff from these sites. There
           is currently insufficient information to assess the extent to which actual reductions may differ
           from this assumption. While phosphorus is often attached to the sediment and therefore may
           be more readily addressed by control measures that retain sediments on site, nitrogen is
           typically found in soluble forms, and sediment control measures may be less effective in
           reducing nitrogen loading.
November 2009                                                                                11-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
        >  Omission of all ponds and lakes and reservoirs located off the RF1 network from the water
           quality analysis. All lakes, ponds, and reservoirs located off the RF1 network are not included
           in the SPARROW model and thus are excluded from estimation of monetized benefits. The
           National Water Quality Inventory Report to Congress (USEPA 2009e) reports 40.6 million
           acres of lakes and reservoirs in the coterminous United States. The RF1 network includes
           approximately 3.9 million acres or 9.5 percent of the total lakes and reservoir acres in the
           United States (USEPA 2007e).20 Omission of these waterbody types from the analysis of
           monetized benefits is likely to lead to understatement of benefits in two benefit categories:
           (1) nonmarket benefits of water quality improvements resulting from the regulation and
           (2) reservoir dredging.
        >  Restriction of the water quality analysis to the description of long-term mean water quality
           conditions. Construction activities are, by contrast, transient in nature, extending over weeks
           or months.  Construction activities (unlike agricultural activities) are spatially compact, so
           they are a sub-grid phenomenon with respect to the specification of the national scale of the
           SPARROW model. The restriction to mean water quality conditions precludes an analysis of
           the frequency with which conditions of extreme sediment transport conditions (e.g., during an
           active construction period) occur. Although the predicted changes in average water quality
           conditions may be small, the expected changes in  sediment concentrations under extreme
           sediment transport conditions may be significant.

11.2.2  Focus  on Selected Pollutants of Concern (Sediment and Nutrients)

Existing case studies of environmental impacts associated with construction activities demonstrated that a
number of pollutants are found in construction site discharges, including turbidity, BOD, metals, toxic
organics, trash and debris, and other miscellaneous pollutants.  However, EPA's analysis of benefits from
reduced construction site discharges focuses only on water quality improvements resulting directly from
reductions in total suspended solid and nutrient loadings. It does not include improvements resulting from
reductions in other pollutant loadings, nor does it include improvements in water quality indicators
indirectly associated with pollutant loadings (e.g., increase in primary productivity such as algae growth,
changes in dissolved oxygen, and turbidity). This is likely to result in underestimation of the expected
water quality changes resulting from the regulation because the combined impact of several pollutants on
ambient water quality conditions is likely to be greater than the sum of the individual  impacts of reducing
concentrations of sediments and nutrients.

11.2.3  Omission of Several Benefit Categories from the Analysis of Monetized Benefits

Due to data limitations, EPA did not estimate benefits in several benefit categories. Although the
magnitude of benefits in the omitted categories is unknown, they may not be trivial. Chapter 5 of this
report provides a qualitative discussion of the omitted benefit categories. A brief summary of the omitted
benefit categories is provided below:
        >  Market values of properties located near waterbodies. Reducing sediment discharges from
           construction sites is likely to increase market values of properties located in the vicinity of
    The estimated total lake/reservoir acres do not include the Great Lakes.
November 2009                                                                                 11-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
           construction sites by enhancing the aesthetic quality of the affected land and water resources
           (e.g., reducing erosion of river banks and improving water clarity).21
        >  Flood damages. Reducing sediment discharges from construction sites is expected to reduce
           flooding damages by decreasing sedimentation of river beds and improving river capacity.
           Clark et al. (1985) estimated flooding damages attributable to sediment discharges to be $1.5
           billion (2008$), annually. Therefore, even a small reduction in the frequency and severity of
           flooding is likely to generate significant benefits.
        >  Ditch maintenance. The regulation is expected to reduce the costs of ditch maintenance by
           reducing the amount of sediment deposited in ditches.
        >  Industrial water use. The regulation is expected to benefit industrial water users by reducing
           sediment concentrations in source waters and thus increasing the useful life of industrial
           equipment.
        >  Agricultural water use. The regulation is expected to benefit agricultural producers by
           reducing sediment discharges  and, as a result, sediment deposition on farm land; this would
           lead to improvements in land productivity and enhanced marketability of agricultural
           products.
        >  Drinking water treatment.22 The regulation is expected to reduce the cost  of drinking water
           treatment to improve taste and odor. Reducing nutrient loadings to surface waters is expected
           to reduce eutrophication which is one of the main causes of taste and odor impairment in
           drinking water. Taste and odor in drinking water has a major negative impact on the public
           perception of drinking water safety and the drinking water industry due to a significant
           increase in drinking water treatment costs from foul taste and odor in the source waters.
           Chapter 9 of this report provides detail on eutrophication impacts on portable water.

11.2.4  Limitations Inherent in the  Estimate of Nonmarket Benefits

To estimate nonmarket benefits of water quality improvements resulting from the regulation,  EPA used a
benefits transfer function based on meta-analysis of surface water valuation studies. Key limitations of
this approach that are likely to lead to underestimation on nonmarket benefits are outlined below:
        >  Using benefit transfer from  existing surface water valuation studies to estimate nonmarket
           benefits of the regulation. A number of issues are common to all benefit transfers. Benefit
           transfer involves adapting research conducted for another purpose in the available literature
           to address the policy questions at hand. As is typical in applied benefit transfers, the meta-
           analysis model used in the analysis of C&D benefits provides a close, but not perfect, match
           to the context in which values are desired. Therefore, some beneficial effects associated with
           reducing sediment discharges  from construction sites are not accounted for in the estimated
           WTP values. For example, surface water valuation studies included in meta-data focused
           primarily on changes in ambient water quality. Other effects associated with sediment
21   The nonmarket component (i.e., increased satisfaction with the property) may be implicitly accounted for in WTP for
    improvements in environmental services provided by surface waters affected by construction site discharges (see Chapter 10
    for detail), however this does not encompass increased market values of real estate.
22 Although EPA estimated a change in drinking water cost associated with reducing turbidity in the source water this estimate
    does not account for any changes in the cost of treating T/O.
November 2009                                                                                   11-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
           discharges (e.g., sedimentation of river beds, stability and erosion of river banks, and changes
           in the stream carrying capacity) were not included in the original study scenarios and, as a
           result, from the estimated WTP for reducing sediment discharges from construction sites.
        >  Restriction of WTP to the state-level improvements in water quality. In analyzing benefits of
           the C&D rule, EPA assumed that households would gain no benefits from water quality
           improvements in aquatic resources located outside of their state of residence. Although
           empirical literature shows that WTP for environmental quality improvements is inversely
           related to the distance from the water resource in question, WTP values are likely to be
           positive for out-of-state residents if improvements occur in aquatic resources of regional or
           national significance (e.g., Chesapeake Bay, Great Lakes). It is also possible that some of the
           out-of-state households have non zero WTP for water quality improvements in water
           resources that have only local significance. Even if per household WTP for out-of-state
           residents are small they can be very large in the aggregate if these values are held by a
           substantial fraction of the population.
        >  Percent of the total reach miles expected to improve maybe understated. Excluding smaller
           streams from water quality analysis may understate the percentage of both (1) the total reach
           miles estimated to experience water quality improvements and/or (2) reach miles associated
           with higher reductions in pollutant concentrations. Because WTP for regional level water
           quality improvements is a function of the magnitude of the expected water quality change and
           the geographic scale of improvement (i.e., percent of reach miles expected to improve), the
           estimated WTP value may be understated. In addition, approximately 67  percent of
           estuarine/coastal reaches are  excluded from the analysis  due to data limitations.
        >  Proximity of water quality improvements to residential areas is not accounted for. Most
           construction impacts are to surface waters near major population centers. Because people
           tend to place a higher value on improvements in waterbodies closer to where they live
           omitting the distance effect from the analysis may lead to understatement of benefits.
November 2009                                                                                 11-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
Aiken, R.A. 1985. Public Benefits of Environmental Protection in Colorado. Masters thesis, Colorado
       State University.
Alabama Department of Environmental Management (DEM). 2009. Water Division - Water Quality
       Program. Chapter 335-6-10: Water Quality Criteria. Available at
       http://www.epa.gov/waterscience/standards/wqslibrary/al/al 4  wqs.pdf
Alabaster, J.S. and Lloyd, D.S. 1982. "Finely divided solids." In: Alabaster, J.S., Lloyd, D.S. (Eds.),
       Water Quality Criteria for Freshwater Fish. Butterworth, London, pp. 1-20.. As cited in Bilotta
       and Brazier 2008.
Alaska Department of Environmental Conservation (DEC). 2003. 18 AAC 70 Water Quality Standards
       As amended through June 26, 2003. Available at
       http://www.epa.gov/waterscience/standards/wqslibrary/ak/ak 10 wqs.pdf
Government of Alberta, Agriculture and Rural Development (ARD). 2009. A Primer on Water Quality:
       Pollutant Pathways. Available at
       http://wwwl.agric.gov.ab.ca/Sdepartment/deptdocs.nsf/all/wat3350
Aldridge, D. W., B. S. Payne, and A. C. Miller. 1987. "The effects of intermittent exposure to suspended
       solids and turbulence on three species of freshwater mussels." As cited in Waters  1995.
Alexander R.B., A.H. Elliott, U. Shankar U., and McBride G.B. 2002a.  "Estimating the sources and
       transport of nutrients in the Waikato River basin." New Zealand: Water Resources Research
       38:1268-1290. As cited in Schwarz etal. 2006.
Alexander R.B., R.A. Smith, and G.E. Schwarz. 2000. "Effect of stream channel size on the delivery of
       nitrogen to the Gulf of Mexico. "Nature 403:758-761.^5 cited in Schwarz etal. 2006.
Alexander, R.B., J.W. Brakebill, R.E. Brew, and RA.  Smith, R.A. 1999, ERF1 - Enhanced river reach
       file 1.2: U.S. Geological Survey Open-File Report 99-457. As cited in Schwarz et al. 2006.
Alexander, R.B., P.J.  Johnes, E.A. Boyer, and R.A. Smith. 2002b. "A comparison of methods for
       estimating the riverine export of nitrogen from large watersheds." Biogeochemistry 57/58:295-
       339. As cited in Schwarz et al. 2006.
Alexander, R.B., R.A. Smith, and G.E. Schwarz. 2004. "Estimates of diffuse phosphorus sources in
       surface waters of the United States using a spatially referenced watershed model." Water Science
       and Technology 49:1-10. As cited in Schwarz et al. 2006.
Alexander, R.B., R.A. Smith, G.E. Schwarz, E.W. Boyer, J.V. Nolan, and J.W. Brakebill, J.W. 2008.
       "Differences in  phosphorus and nitrogen delivery to the Gulf of Mexico from the Mississippi
       River Basin." Environmental Science and Technology 42(3):822-830.
Alexander, R.B., RA. Smith, G.E. Schwarz, S.D. Preston, J.W. Brakebill, R Srinivasan, and P.A.
       Pacheco. 2001.  Atmospheric nitrogen flux from the watersheds of major estuaries of the United
       States - An application of the SPARROW watershed model, in Valigura, R.A. As cited in
       Schwarz et al. 2006.
November 2009                                                                               12-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
American Society of Civil Engineers (ASCE)/American Water Works Association (AWWA). 2005.
       Water Treatment Plant Design, 4th Edition, McGraw-Hill, NYC, NY Page 17-6.
Anderson, G.D. and S.F. Edwards. 1986. Protecting Rhode Island's Coastal Salt Ponds: "An Economic
       Assessment of Downzoning." Coastal Zone Management 14( i/z): 67-91.
Anderson, P. G., B. R. Taylor, and G. C. Balch. 1996. Quantifying the Effects of Sediment Release on
       Fish and their Habitats. As cited in EPA 2004.
Appleby, J. P., and D. J. Scarratt. 1989. "Physical effects of suspended solids on marine and estuarine fish
       and shellfish with special reference to ocean dumping: A literature review." Canadian Technical
       Report of Fisheries and Aquatic Sciences No. 1681. A? cited in Kerr 1995.
Area Wide Soil Contamination Task Force. 2003. Area Wide Soil Contamination Task Force Report
       Submitted to: Washington State Department of Agriculture, Washington State Department of
       Ecology, Washington State Department of Health, Washington State Department of Community,
       Trade and Economic Development. June 30, 2003.
Arkansas Pollution  Control and Ecology Commission. 2007.Regulation Number 2: Regulation
       Establishing Water Quality Standards for Surface Waters of the State of Arkansas. #014.00-002.
       Available at http://www.epa.gov/waterscience/standards/wqslibrary/ar/ar_6_wqs.pdf
Arnold, J.G., J.R. Williams, and D.R. Maidment. 1995. "Continuous time water and sediment routing
       model for large basins." Journal of Hydraulic Engineering 121(2): 171-183.
Arnold, J.G., R. Srinivasan, R.S. Muttiah, and J.R. Williams.  1998. "Large Area Hydrologic Modeling
       and Assessment: I. Model Development." J. American Water Resources Association 34:73-89.
Arrow, K., R. Solow,  E. Learner, P. Portney, R. Rander, and H. Schuman. 1993. Report of the NOAA
       Panel on Contingent Valuation, Federal Register 58 (Jan.): 4602-4614.
Azevedo, C., J.A. Herriges, and C.L. Kling. 2001. Valuing Preservation and Improvements of Water
       Quality in Clear Lake. Center for Agricultural and Rural Development (CARD), Iowa State
       University, Staff Report 01-SR 94.
Bachand, P., Heyvaert, J. Reuter, I. Werner, S. Bachand, S.J.  The. 2007.  Chemical Treatment Methods
       Pilot (CTMP) System for Treatment of Urban Runoff- Phase I. Feasibility and Design. Final
       Report for the City of South Lake Tahoe. February 22, 2007.
Barrett, M.E., J.F. Malina, Jr., R.J. Charbeneau, and G. Ward. 1995a. "Effects of Highway Construction
       and Operation on Water Quality and Quantity in an Ephemeral Stream in the Austin, Texas
       Area." Center for Research in Water Resources, University of Texas at Austin.
Barrett, M.E., R.D.  Zuber, and E.R. Collins III.  1995b. "A Review And Evaluation of Literature
       Pertaining to the Quantity and  Control of Pollution from Highway Runoff and Construction."
       Center for Research in Water Resources, Bureau of Engineering  Research, University of Texas at
       Austin. Available at http://www.ce.utexas.edu/centers/crwr/reports/online.html
Barton, J.R.., D.A. White, P.V. Winger, and E.J. Peters. 1972. "The effects of highway construction on
       the fish habitat in the Weber River near Henefer, Utah. Engineering Resources Center, Colorado
       Bureau of Reclamation, Report No. REC-ERC 72-17. June,  1972: pp. 17-28. As cited in Barton
       1977.
November 2009                                                                               12-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
Barton, B.A. 1977. "Short-term effects of highway construction on the limnology of a small stream in
       southern Ontario." Freshwater Biology 7: 99-108.
Bash, J. C. Berman, and S. Bolton. 2001. Effects of Turbidity and Suspended Solids on Salmonids.
       Center for Streamside Studies, University of Washington.
Bateman, I.J. and A.P. Jones. 2003. "Contrasting Conventional with Multi-Level Modeling Approaches to
       Meta-Analysis: Expectation Consistency in U.K. Woodland Recreation Values." Land Economics
       79(2): 235-258.
Bateman, I.J., B.H. Day, S. Georgiou and I. Lake. 2006. "The Aggregation of Environmental Benefit
       Values: Welfare Measures, Distance Decay and Total WTP'." Ecological Economics 60(2): 450-
       460.
Beaulac, M.N. and K.H. Reckhow. 1982. "An Examination of Land Use - Nutrient Export
       Relationships." Water Resources Bulletin 18:1013-1024. As cited in Schwarz et al. 2006.
Bejranonda, S. F.J. Hitzhusen; and D. Hite. 1999. "Agricultural Sedimentation Impacts on Lakeside
       Property Values." Agricultural and Resource Economics Review 28(2): 208-18.
Bell, M.C. 1973. Fisheries handbook of engineering requirements and biological criteria. As cited in
       Waters  1995.
Berg, L. 1982. The effect of exposure to short-term pulses of suspended sediment on the behavior of
       juvenile salmonids. P. 177-196 in G.F. Hartman et al. (Eds.) Proceedings of the Carnation  Creek
       Workshop: a ten-year review. Department of Fisheries and Oceans, Pacific Biological Station,
       Nanaimo, California. As cited in Bash et al. 2001.
Bergstrom, J.C. and L.O. Taylor. 2006. "Using Meta-Analysis for Benefits Transfer: Theory and
       Practice." Ecological Economics 60(2):351-360.
Bergstrom, J.C. and P. De Civita.  1999. "Status of Benefits Transfer in the United States and Canada: A
       Review." Canadian Journal of Agricultural Economics 47(1): 79-87.
Berkman, H. E., and C. F. Rabeni. 1987.  "Effects of siltation on stream  fish communities. "As cited in
       Wood and Armitage 1997.
Berry, W., N. Rubinstein,  B. Melzian, and B. Hill.  2003.  The Biological Effects  of Suspended and
       Bedded Sediment (SABS) in Aquatic Systems: A Review. Internal Report of the U.S. EPA
       Office of Research and Development. Narragansett, RI. Available at
       http://www.epa.gov/waterscience/criteria/sediment/appendixl.pdf
Beven, K. 2002. "Towards an alternative blueprint for a physically based digitally simulated hydrologic
       response modeling system."HydrologicalProcesses 16:189-206. As1 citedin Schwarz etal. 2006.
Beven, K. and M. Kirkby. 1979. A physically-based variable contributing area model of basin hydrology.
       Hydrologic Science Bulletin 24:43-69.
Bhaduri, B.L., J.M. Harbor, P. Maurice. 1995. "Chemical Trap Efficiency of a Construction Site
       StormWater Retention Basin." Physical Geography 16 (5): 389-401.
Bhardwaj, A.K. and R.A. McLaughlin. 2008. "Simple Polyacrylamide Dosing Systems For Turbidity
       Reduction In Stilling Basins." Transactions of the ASABE 51(5): 1653-1662.
November 2009                                                                               12-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
Biber, E. 2002. "The application of the Endangered Species Act to the protection of freshwater mussels: a
       case study." Environmental Law 32(1): 118-120.
Bicknell, B.R., J.C. Imhoff, J.L. Kittle, T.H. Jobes Jr., and A.S. Donigian Jr. 2001. Hydrological
       simulation program-Fortran, HSPF version 12 user's manual: Mountain View, CA, AQUA
       TERRA Consultants, 873 p. As cited in Schwarz et al. 2006.
Bilotta, G.S. and R.E. Brazier. 2008. "Understanding the influence of suspended solids on water quality
       and aquatic biota." Water Research 42: 2849-2861.
Binger, R.L., F.D. Theurer, and Y. Yuan. 2009. "AnnAGNPS Technical Processes Documentation
       Version 5.0, August 2009. Available at
       http://www.wsi.nrcs.usda.gov/products/w2q/h&h/tools models/agnps/downloads .html
Birkett, C., E.W. Tollner, and D.K. Gattie. 2007. "Total suspended solids and flow regime effects on
       periphyton development in a laboratory channel." Transactions of the American Society of
       Agricultural Engineers 50(3): 1095-1104. As cited in Bilotta and Brazier 2008.
Bjornn, T. C., M. A. Brusven, M. P. Molnau, J. H. Milligan, R. A. Klamt, E. Chacho, and C.  Schaye.
       1977. "Transport of granitic sediment in streams and its effects on insects and fish." Completion
       Report Project B-036-IDA, Bulletin 17, Idaho Cooperative Fishery Research Unit, University of
       Idaho, Moscow, Idaho. 43 p. As cited in Kerr 1995.
Blanco, H., D. Carlson, M. Chalana, F. Wagner, and C. Wolfe. 2009. Final Report Linking Toxics
       Cleanup and Redevelopment Across the States: Lessons for Washington State. Prepared for The
       Toxics Cleanup Program, Washington State Department of Ecology.
Boardman, A.E., D.H. Greenberg, A.R. Vining, and D.L. Weimer. 2001. Cost-Benefit Analysis: Concepts
       and Practice, 2nd edition. Upper Saddle River, NJ: Prentice Hall.
Bockstael, N.E., McConnell, K.E., and Strand, I.E. 1988. Benefits from improvements in Chesapeake Bay
       Water Quality. Department of Agricultural & Resource Economics, University of Maryland,
       1988.
Bockstael, N.E., K.E. McConnell and I.E. Strand. 1989. "Measuring the Benefits of Improvements in
       Water Quality: The Chesapeake Bay" Marine Resource Economics 6:  1-18.
Boubee, J.A.T., T.L. Dean, D.W. West, and R.F.G. Barrier. 1997. "Avoidance of Suspended  Sediment by
       the Juvenile Migratory Stage of Six New Zealand Native Fish Species." New Zealand Journal of
       Marine and Freshwater Research (31):61-69. As cited in Newcombe 2003.
Boyd, W. 1990. "Effects of poultry waste on ground and surface water." In: Proceedings of the 1990
       National Poultry Waste Management Symposium, Blake, J., and R.M. Hulet (eds.).
Boyle, K.J. and J.C. Bergstrom. 1992. "Benefit Transfer Studies: Myths, Pragmatism and Idealism."
       Water Resources Research 28(3):657-663.
Brady, N.C.  and RR Weil .  1999. The Natures and Properties of Soils, 12th ed. New Jersey:  Prentice
       Hall. Pp. 494, 540, 545.
Brakebill, J.W. and S.E. Preston. 1999.  Digital data used to relate  nutrient inputs to water quality in the
       Chesapeake Bay watershed, version 1.0: U.S. Geological  Survey Water Open-File Report 99-60.
       Available at http://md.usgs.gov/publications/ofr-99-60/. As cited in Schwarz et al. 2006.
November 2009                                                                               12-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
Breffle, William S., Morey, E., Rowe, R., Waldman, D. and Wytinck, S. 1999. Recreational Fishing
       Damages from Fish Consumption Advisories in the Waters of Green Bay. Prepared by Stratus
       Consulting Inc. for U.S. Fish and Wildlife Service, U.S. Department of Interior and U.S.
       Department of Justice, 1999.
Brock, T.D. 1985.  "Life at Higher Temperatures." Science 230: 132-138.
Brown, R.M., N.I. McClelland, R.A. Deininger, and RG. Tozer. 1970. "A Water Quality Index - Do We
       Dare?" Presented at the National Symposium on Data and Instrumentation for Water Quality
       Management, Madison, Wisconsin and published in Water & Sewage Works. As cited in
       McClelland 1974.
Buckner, M. M., Wendy  Smith, and Judy A. Takats (World Wildlife Fund). 2002. Tennessee,
       Cumberland, and Moblie River Basins at Risk: A Biological Assessment and Vision for the
       World Wildlife Fund's Southeast Rivers and Streams Project.
Burkholder J.  M. 1998. "Implications of harmful marine microalgae and heterotrophic dinoflagellates in
       management of sustainable fisheries" Ecological Applications (Supplement) 8: pp. S37-S62.
Burton, T.M., R.R. Turner, and R.C. Harris. 1976. "The impact of highway construction on a north
       Florida watershed." Water Resources Bulletin 12: 529-538.^45 cited in Nodvin etal. 1986.
California Department of Fish and Game. 2004. "Recovery Strategy for California Coho Salmon."
       Species Recover Strategy 2004-1. February, 2004.
California Environmental Protection Agency (EPA). 2009a. Central Coast Regional Water Quality
       Control Board Water Quality Objectives. Available at
       http://www.swrcb.ca.gov/rwqcb3/publications_forms/publications/basin_plan/chapter_3/chapter3
       .shtml
California Environmental Protection Agency (EPA). 2009b. San Francisco Regional Water Quality
       Control Board Water Quality Objectives. Available at
       http://www.swrcb.ca.gov/sanfranciscobay/water_issues/programs/basin_plan/docs/bp_ch3+tables
       .pdf
Cameron, T.A. and D.D.  Huppert. 1989.  "OLS versus ML Estimation of Non-market Resource Values
       with Payment Card Interval Data." Journal of Environmental Economics and Management 17,
       230-246.
Cameron, Trudy Ann. 1988. Using the Basic "Auto-validation" Model to Assess the Effect of
       Environmental Quality on Texas Recreational Fishing Demand: Welfare Estimates. EPA National
       Center for Environmental Economics, 1988.
Canadian Council  of Ministers of the Environment (CCME). 2001. Canadian water quality guidelines for
       the protection of aquatic life: CCME Water Quality Index 1.0, Technical Report. In: Canadian
       environmental quality guidelines, 1999, Canadian Council of Ministers of the Environment,
       Winnipeg.
Canadian Department of Fisheries and Oceans (DFO). 2000. Effects of sediment  on fish and their habitat.
       DFO Pacific Region Habitat Status Report 2000/01.
Canada Department of Fisheries and Ocean (DFO). 2009. Concrete Wash-Water Characteristics.
       Available at http://www-heb.pac.dfo.gc.ca/water quality/ffish and pollution/cone char e.htm.
November 2009                                                                               12-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
Caraco, N.F., J.J. Cole, G.E. Likens, G.M. Lovett, and K.C. Weathers, K.C. 2003. "Variation in NO3
       export from flowing waters of vastly different sizes - Does one model fit all?" Ecosystems 6:344-
       352. As cited in Schwarz et al. 2006.
Carline, Robert F., Adam Smith, Curtis Gill, and Kirk Patten. 2003. "Effects of Bridge and Road
       Construction on Spring Creek: Route 26 Transportation Improvements Project." Pennsylvania
       Department of Transportation, Pennsylvania Fish and Boat Commission.
Carruthers, T. and Wazniak, C. 2003. Development of a Water Quality Index for the Maryland Coastal
       Bays. Maryland's Coastal Bays: Ecosystem Health Assessment, Chapter 4.4. 2003.
Carson, R.T. and R.C. Mitchell. 1993. "The value of clean water: the public's willingness to pay for
       beatable, fishable, and swimmable quality water." Water Resources Research 29(7): 2445-2454.
Carson, R.T., T. Groves, and M.J. Machina. 2000. Incentive  and Informational Properties of Preference
       Questions. Draft Paper. University of California, La Jolla.
Carson, R.T., W.M. Hanemann, R.J. Kopp, J.A. Krosnick, RC. Mitchell, S. Presser, P.A.  Ruud, and V.K.
       Smith. 1994. Prospective Interim Lost Use Value due to DDT and PCB Contamination in the
       Southern California Bight. Volume 2. Report to the National Oceanic and Atmospheric
       Administration, Produced by Natural Resources Damage Assessment Inc., LA Jolla, CA.
Chapman, D. W. 1988. "Critical review of variables used to  define effects of fines in redds of large
       salmonids."^45 cited in Canadian Department of Fisheries and Oceans 2000.
Chapra, S.C. 1997. Surface water-quality modeling. McGraw-Hill, New York. 844 p. As cited in Schwarz
       et al. 2006.
Chen, Y., R.C. Viadero Jr., W. Xinchao, R. Fortney, L.B. Hedrick, S.A. Welsh, J.T. Anderson, and L.
       Lin.  2009. "Effects of Highway Construction on Stream Water Quality and Macroinvertebrate
       Condition in a Mid-Atlantic Highlands Watershed, USA." Journal of Environmental Quality 38:
       1672-1682.
Chisholm, J.L., and S.C. Downs (USGS).  1978. Stress and Recovery of Aquatic Organisms as Related to
       Highway construction along Turtle Creek, Boone County, West Virginia. Geological Survey
       Water Supply Paper 2055.
Clark, E., J.A. Haverkamp, and W.Chapman. 1985. "Eroding Soils: The Off-Farm Impacts." Washington,
       DC:  The Conservation Foundation.
Clausen, J.C. 2007.Jordan Cove Watershed Project: Final Report. Department of Natural Resources
       Management and Engineering, College of Agriculture and Natural Resources, University of
       Connecticut.
Cleveland, T.G., and A. Fashokun. 2006. "Construction-Associated Solids Loads With a Temporary
       Sediment Control BMP." Journal of Construction Engineering and Management  132(10): pp.
       1122-1125.
Cline, L.D., R.A. Short, and J.V. Ward. 1982. "The influence of highway construction on the
       macroinvertebrates and epilithic algae of a high mountain stream." Hydrobiologia 96: pp. 149-
       159.
November 2009                                                                               12-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
Clonts, H.A., and J.W. Malone. 1990. Preservation Attitudes and Consumer Surplus in Free Flowing
       Rivers. In: Social Science and Natural Resource Recreation Management, Joanne Vining, editor.
       Westview Press, Boulder, CO. pp. 301-317.
Cloud, T.J. Jr, U.S. FWS, to Allen Ryan, Department of the Army. February 2, 2004. Letter. Available at
       http://www.fws.gov/southwest/es/Documents/R2ES/Ark shiner BNSF railroad BO  COE.pdf.
       Accessed April 2, 2008.
Cohn, T.A. 2005. "Estimating contaminant loads in rivers: An application of adjusted maximum
       likelihood to type 1 censored data." Water Resources Research 41:W07003,
       doi:10.1029/2004WR003833.
Cohn, T.A., D.L. Caulder, E.J. Gilroy, L.D. Zynzuk, and R.M. Summers. 1992a. The validity of a simple
       statistical model for estimating fluvial constituent loads: an empirical study involving nutrient
       loads entering Chesapeake Bay. Water Resources Research 28(9):2353-2363. As cited in Schwarz
       et al. 2006.
Cohn, T.A., E.J. Gilroy, and W.G. Baier. 1992b. Estimating fluvial transport of trace constituents using a
       regression model with data subject to censoring. Proceedings of the Section on Statistics and the
       Environment of the American Statistical Association, p. 1-9. As cited in Schwarz et al. 2006.
Connecticut  Department of Environmental Protection (DEP). 2002. Water Quality Standards. Available at
       http://www.epa.gov/waterscience/standards/wqslibrary/ct/ct_l_wqs.pdf
Construction Cost Index. 2008. "Construction Economics." Engineering News-Record March 2008.
Cooke, S.J.,  C.M. Bunt and J. F. Schreer. 2004. "Understanding fish behavior, distribution, and survival
       in thermal effluents using fixed telemetry arrays: A case study of smallmouth bass in a discharge
       canal during winter." Environmental Management, vol. 22:140-150.
Cooper, C. M. 1987. "Benthos in Bear Creek, Mississippi: Effects of habitat variation and agricultural
       sediments." Journal of Freshwater Ecology, Vol. 4: pp  101-114. As cited in Kerr 1995.
Cordone, A.  J., and S. Pennoyer. 1960. "Notes on silt pollution in the Truckee River Drainage, Nevada
       and Placer counties." As cited in Waters 1995.
Cramer, G.H., II, and W.C.  Hopkins, Jr. 1982. "Effects of Dredged Highway Construction on Water
       Quality in a Louisiana Wetland." Transportation Research Record 896: 47-50. As cited in Barrett
       etal. 1995b.
Croke, Kevin, R.G. Fabian, and G. Brenniman.  1986-87. Estimating the Value of Improved Water
       Quality in an Urban River System.  Journal of Environmental Systems 16, No. 1.  pp. 13-24.
Crowder, B.  M. 1987. "Economic Costs of Reservoir Sedimentation: A Regional Approach to Estimating
       Cropland Erosion Damages." Journal of Soil and Water Conservation 42(3): 194-197.
Cude, C. G. 2001. Oregon Water Quality Index: "A Tool for Evaluating Water Quality Management
       Effectiveness." Journal of the American Water Resources Association 37(1): 125-137.
Cude, C.G. 2009. Personal Communication.
Dalhuisen, J.M., R.J.G.M. Florax, H.L.F. de Groot, and P. Nijkamp. 2003. "Price and income elasticities
       of residential water demand: A meta analysis." Land Economics 79(2):292-308.
Dalkey, N.C. 1968. DELPHI. The Rand Corporation.
November 2009                                                                               12-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
Daniel, T. C., P. E. McGuire, D. Stoffel and B. Miller. 1979. "Sediment and Nutrient Yield from
       Residential Construction Sites." Journal of Environmental Quality 8(3): pp. 304-308.
De Robertis, A., C.H. Ryer, A. Veloza and Richard D. Brodeur. 2003. "Differential effects of turbidity on
       prey consumption of piscivorous and planktivorous fish." Canadian Journal of Fisheries and
       Aquatic Science Vol. 60: pp. 1517-1526.
De Zoysa, A.D.N. 1995. A Benefit Evaluation of Programs to Enhance Groundwater Quality, Surface
       Water Quality and Wetland Habitat in Northwest Ohio. Dissertation, Ohio State University.
Dearmont, D., B.A. McCarl, and D.A. Tolman. 1998. "Cost of water treatment due to diminished water
       quality: A case study in Texas." Water Resources Research 54(4): pp 849-853.
Desvousges, W.H., F.R. Johnson, and H.S. Banzhaf. 1998. Environmental Policy Analysis with Limited
       Information: Principles and Applications of the Transfer Method. Northampton, MA: Edward
       Elgar Publishers.
Desvousges,W.H., V.K. Smith and A. Fisher. 1987. "Option Price Estimates for Water Quality
       Improvements: A Contingent Valuation Study for the Monongahela River." Journal of
       Environmental Economics and Management. 14: 248-267.
DeZuane, John. 1997. Handbook of Drinking Water Quality, 2nd Ed. New York: John Wiley & Sons.
Dodds, W.K. et al. 2009. "Eutrophication of U.S. Freshwaters: Analysis of Potential Economic
       Damages." Environ. Sci. Tech. 43(1): 12-19. Available at
       http://pubs.acs.org/doi/full/10.1021/es801217q.
Dodge, R.E., R.C. Aller and J.  Thomson. 1974. Coral growth related to resuspension of bottom
       sediments. Nature 247: 574-576.

Donaldson, B.M. 2004. Final Report Highway Runoff in Areas of Karst Topography. Virginia
       Transportation Research Council. VTRC 04-R13
Downs, S.C., and D.H. Appel (USGS). 1986. Progress Report on the Effects of Highway Construction on
       Suspended-Sediment Discharge in the Coal River and Trace Fork, West Virginia, 1975-81. Water
       Resources Investigations Report  84-4275.
Drennen, D.J. 2003. "The Urban Life of Darters." Endangered Species Bulletin 28(2): 12-13.
Droppo, I.A., S.W. Liss, D. Williams, T.  Nelson, C. Jaskot, and B. Trapp. 2009. "Dynamic Existence of
       Waterborne Pathogens within River Sediment Compartments. Implications for Water Quality
       Regulatory Affairs" Environmental Science and Technology 43: 1737-1743.
Duck, R.W.  1985. "Short Communications: The Effect of Road Construction on Sediment Deposition in
       loch Earn, Scotland." Earth Surface Processes andLandforms 10: 401-406.
Dunnette, D. A. 1979. A Geographically  Variable Water Quality Index Used in Oregon. As cited in Cude
       2001.
Edwards, D. 1969. "Some effects of siltation upon aquatic macrophyte vegetation in rivers."
       Hydrobiologia, 34: 29-36. As cited in Kerr 1995.
Eisler, R. 1987. Polycyclic Aromatic Hydrocarbon Hazards to Fish, Wildlife, and Invertebrates: A
       Synoptic Review. Biol. Report 85(1.11), U.S Fish and Wildlife Service, Washington D.C.
November 2009                                                                              12-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
Eckhardt, D.A.V., 1976, Sediment Discharge from an Area of Highway Construction, Applemans Run
       Basin, Columbia County, Pennsylvania, USGS Water-Resources Investigations 76-111. As cited
       in Barrett et al.  1995b.
Eldin, N.N. 2002. "Road Construction: Materials and Methods." Journal of Environmental Engineering
       128(5): 423-430.
Elliott, A.H., R.B. Alexander, G.E. Schwarz, U. Shankar, J.P.S. Sukias, and G.B. McBride. 2005.
       Estimation of nutrient sources and transport for New Zealand using the hybrid physical-statistical
       model SPARROW. Journal of Hydrology (New Zealand) 44:1-27. As cited in Schwarz et al.
       2006.
Ellis, M.M. 1936. "Erosion silt as a factor in aquatic environments. "As citedin Waters 1995.
Embler, P.P., and Fletcher, M.O., 1983, Effects of Highway Construction on Stream Turbidity and
       Suspended Solids: A Case Study, FHWA/SC-81/3, Federal Highway Administration,
       Washington, D.C. As cited in Barrett et al.  1995b.
Extence, C.A. 1978. "The Effects of Motorway Construction on an Urban Stream." Environmental
       Pollution 17: 245-252.
Fabricius K., C. Wild, E. Wolanski and D. Abele. 2003. "Effects of transparent exopolymer particles and
       muddy terrigenous sediments on the survival of hard coral recruits."  Estuarine, Coastal and Shelf
       Science 56: 1-9.
Faucette L.B., J. Governo, C.F. Jordan, E.G. Lockaby, H.F.  Carino, and R. Governo. 2007. "Erosion
       control and storm water quality from straw with PAM, mulch, and compost blankets of varying
       particle sizes" Journal of Soil and Water Conservation 62(6): 404-413.
Faucette, L.B., K.A. Sefton, A.M. Sadeghi, and R.A. Rowland. 2008. "Sediment and phosphorus removal
       from simulated storm runoff with compost filter socks and silt fence." Journal of Soil and Water
       Conservation 63(4): 257-264.
Folkes, D.J., S.O. Helgen, and R.A.  Litle. 2001. "Impacts of Historic Arsenical Pesticide Use on
       Residential Soils in Denver, Colorado." In: Proc. 4th Int. Conf. On Arsenic Exposure and Health
       Effects, SanDiego.
Forbes, A. M., J.  J. Magnuson, and D. M. Harrell. 1981. Effects of habitat modifications from coal ash
       effluent on stream macrobenthos: A synthesis. As cited in Kerr 1995.
Forster, D. L., C. P. Bardos, and D. D. Southgate. 1987. "Soil Erosion and Water Treatment Costs."
       Journal of Soil and Water Conservation 42: 349-352. As cited in Dearmont et al. 1998.
Fossati, O., J.  G. Wasson, C. Hery, R. Marin, and G. Salinas. 2001. "Impact of sediment releases on water
       chemistry and macroinvertebrate communities in clear water Andean streams (Bolivia)." Archiv
       fur Hydrobiologie 151(1): 33-50.
Francoeur, S.N. and J.F.B. Biggs. 2006. "Short-term effects of elevated velocity and  sediment abrasion on
       benthic algal communities." Hydrobiologia 561: 59-69.
Fredman, W. 1995. Environmental Ecology. The ecological effects of pollution, disturbance, and other
       stressors. Academic Press. San Diego, CA.
Freeman III, A.M. 2003. The Measurement of Environmental and Resource Values: Theory and Methods.
       Washington, DC: Resources for the Future.
November 2009                                                                               12-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
Friedman, M. A. and B. E. Levin. 2005. "Neurobehavioral effects of harmful algal bloom (HAB) toxins:
       A critical review." Journal of the International Neuropsychological Society 11(3): 331-338.
Funari E. and E. Testai. 2008. "Human health risk assessment related to cyanotoxin exposure." Critical
       Reviews in Toxicology 38: 97-125.
Gammon, J. R. 1970. "The effect of inorganic sediment on stream biota." Water Pollution Control
       Research Series 18050 DWC 12/70, U.S. Environmental Protection Agency, Washington, DC.
       141 pp. As cited in Kerr 1995.
Garton, E.R., Jr. 1977.  "The Effects of Highway Construction on the Hydrologic Environment at
       Bowden, West Virginia," In: Hydrologic Problems in Karst Regions, edited by Ronald R.
       Dilamarter and Sandor C. Csallany, Western Kentucky University, Bowling Green, KY. As cited
       in Barrett et al. 1995b.
Glanville, T.D., RA. Persyn, T.L. Richard, J.M. Laflen, and P.M. Dixon. 2004. "Environmental Effects
       of Applying Composted Organics to New Highway Embankments: Part 2. Water Quality"
       Transactions of the ASAE 47(2): 471-478.
Glass, G.V. 1976.  "Primary, Secondary, and Meta-Analysis of Research." Educational Researcher 5(10):
       3-8.
Goldman C.R. and A. Home.  1994. Limnology. 2nd edition.  McGraw-Hill International.
Goldstein, H.  \995.Multilevel Statistical Models. 2nd ed. Edward Arnold, London, UK.
Gonzalez, L. 2009. Personal Communication. Houston Advanced Center for Research. Data received via
       Excel file. April 6, 2009.
Graham, A. A. 1990. "Siltation of stone-surface periphyton in rivers by clay-size particles from low
       concentrations in suspension. "As  cited in Waters 1995.
Graves, D.S., and J.W. Burns  (The Resources Agency of California, Department of Fish  and Game).
       1970. Comparison of the Yields of Downstream Migrant Salmonids Before and After Logging
       Road Construction on the South Fork Caspar Creek, Mendocino County. Inland Fisheries
       Administrative Report No. 70-3.
Gray, J.R., Glysson, G.D., Turcios, L.M., and Schwarz, G.E. 2000.  Comparability of Suspended-
       Sediment Concentration and Total Suspended Solids Data.  U.S. Geological Survey Water-
       Resources Investigations Report 00-4191.
Griffiths, C. undated. The Use of Benefit-Cost Analysis in Environmental Policy Making. Working
       Paper. National Center for Environmental Economics, U.S. Environmental Protection Agency,
       Washington, DC.
Grottoli A.G., L.J. Rodrigues and J.E. Palardy. 2006. "Heterotrophic plasticity and resilience in bleached
       corals."Nature 440: 1186-1189.
Guenther, M. and R. Bozelli. 2004. "Effects of inorganic turbidity on the phytoplankton of an Amazonian
       Lake impacted by bauxite tailings." Hydrobiologia 511: 151-159.
Gupta, A.K., Gupta, S.K., and Patil, RS. 2003. "A Comparison of Water Quality Indices for Coastal
       Water." Journal of Environmental Science and Health 38(11): 2711-2725. 2003.
November 2009                                                                              12-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
Gupta, V.K. and S. Sorooshian. 1985. "The relationship between data and the precision of parameter
       estimates of hydrologic models." Journal of Hydrology 81:57-77. As cited in Schwarz et al. 2006.
Guy, H. P. 1963. Residential construction and sedimentation at Kensington, Maryland. Paper presented at
       Federal Interagency Sedimentation Conference, Jackson, Mississippi. January 28-February 1.
Hainly, R.A. 1980. The effects of highway construction on sediment discharge into Blackhouse Creek and
       Stream Valley Run, Pennsylvania. U.S. Geological Society Water Resource Investigation 80-68.
       As cited in Barrett et al. 1995b.
Haith, D. A., Mandel, R. and Wu, R S. 1996. GWLF User's Manual.
Haith, D. A. 2009. Personal email communication.
Harbor, J.M., J. Snyder, and J. Storer. 1995. "Reducing nonpoint source pollution from construction sites
       using rapid seeding and mulching."Physical Geography 16(5): 371-378.
Hargrove, W.W. 2007. Brashear Island Volumetric Interpolation. Oak Ridge National Library. Accessed
       April 4, 2008. Available at http://research.esd.ornl.gov/CRERP/SEDIMENT/BRASHEAR.HTM.
Harrison, T.D., Cooper, J.A.G., and Ramm, A.E.L., 2000. Geomorphology, Ichthyofauna, Water Quality
       and Aesthetics of South African Estuaries. South Africa Department of Environmental Affairs &
       Tourism, Pretoria, South Africa, ENV-DC 2000-01.
Havlin, J. L., J. D. Beaton, S. L. Tisdale and W. L. Nelson. 1999. Soil Fertility and Fertilizers. 6th
       Edition. Prentice Hall. Upper Saddle River, NJ. As cited in Binger et al. 2009.
Hayes, D.B., H.R. Dodd, and J.L. Lessard. 2006. "Effects of small dams on cold water stream fish
       communities. "American Fisheries Society Symposium: 587-601.
Hayes, K.M., T.J. Tyrell and G. Anderson.  1992. "Estimating the Benefits of Water Quality
       Improvements in the Upper Narragansett Bay."Mar/we Resource Economics 7: 75-85.
Hedrick, K.P., R.B. Robinson, B. Tschantz, and S.E. Moore. 2006. "Impact of Tunnel Reconstruction on
       Stream Water Quality in Great Smoky Mountains National Park." Journal of Hydrologic
       Engineering 11(6): 570-577.
Hedrick, L.B., S.A. Welsh, and J.T. Anderson. 2007. "Effects of Highway Construction on Sediment and
       Benthic Macroinvertebrates in Two Tributaries to the Lost River, West Virginia." Journal of
       Freshwater Ecology 22(4): 561-569.
Heiss, G.A. 1996.  "Annual band width variation inPorites sp.  from Aquaba, Gulf of Aquaba, Red Sea."
       Bulletin of Marine Science 59: 393-403.
Helm, R.E., 1978, Sediment Discharge from Highway  Construction Near Port Carbon, Pennsylvania,
       USGS Water-Resources Investigations 78-35. As cited in Barrett et al. 1995b.
Helsel, D.R. (USGS). 1985. Contributions of Suspended Sediment From Highway Construction and
       Other Land Uses to the Olentangy River, Columbus, Ohio. Water Resources Investigations
       Report 84-4336.
Henley, W. E., M. A. Patterson, R. J. Neves, and A. Dennis Lemly. 2000. "Effects of Sedimentation and
       Turbidity on Lotic Food Webs: A Concise Review for Natural Resource Managers." Review in
       Fisheries Science 8(2): 125-139.
November 2009                                                                              12-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
Herriges, J.A. and J.F. Shogren. 1996. "Starting point bias in dichotomous choice valuation with follow-
       up questioning." Journal of Environmental Economics and Management 30(1):112-131.
Kite, D. 2002. Willingness to Pay for Water Quality Improvements: The Case of Precision Application
       Technology. Department of Agricultural Economics and Rural Sociology, Auburn University,
       Auburn, AL; 8 August 2002.
Hoffman, D.J., B.A. Rattner, G.A. Burton and J. Cairns. 1995. Handbook of Ecotoxicity. Lewis
       Publishers.
Holmes, T.P. 1988.  "The Offsite Impact of Soil Erosion on the Water Treatment Industry." Land
       Economics  64(4): 356-366.
Horton, R.K. 1965.  "An Index-Number System for Rating Water Quality." Journal of Water Pollution
       Control Federation 37(3): 300-306. As cited in Cude 2001.
Hoos, A.B. 2005. Evaluation of the SPARROW model for estimating transport of nitrogen and
       phosphorus in streams of the Interior Low Plateau Ecoregion, Tennessee, Kentucky, and
       Alabama, during 1992-2002. In Tennessee Water Resources Symposium, 15th, Burns, Tenn.,
       2005, Proceedings. Tennessee Section of the American Water Resources Association, p. 2A-47.
       As cited in Schwarz et al. 2006.
Horner, R.P., Guedry, J., and Kortenhof, M.H.  1990. Improving the cost effectiveness of highway
       construction site erosion and pollution control. Final Report. Washington State Transportation
       Commission.
Howarth, R.W., G. Billen, D. Swaney, A. Townsend, N. Jaworski, K. Lajtha, J.A. Downing, R. Elmgren,
       N. Caraco, T. Jordan, F. Berendse, J. Freney, V. Kudeyarov, P. Murdoch, and Z. Zhao-Liang.
       1996. "Regional nitrogen budgets and riverine N and P fluxes for the drainages to the North
       Atlantic Ocean - natural and human influences." Biogeochemistry 35:75-139. As cited in Schwarz
       et al. 2006.
Huang, J.C., T.C. Haab and J.C. Whitehead. 1997. "Willingness to Pay for Quality Improvements: Should
       Revealed and Stated Preference Data Be Combined?" Journal of Environmental Economics and
       Management 34(3): 240-255.
Huckabee, J.W., C.P. Goodyear, and R.D. Jones. 1975. "Acid Rock in the  Great Smokies: Unanticipated
       Impact on Aquatic Biota of Road Construction in Regions of Sulfide Mineralization."
       Transactions of the American Fisheries Society, 1975(4): 677-684
Hunt, C. 2009. Personal Communication. University of New Hampshire. Data received via Excel file.
       March 13, 2009.
Hunt, C.L. and J.K.  Grow. 2001. Impacts of Sedimentation to a Stream by a Non-Compliant Construction
       Site. American Society of Civil Engineers Conference Proceedings, World Water Congress 2001.
Hushak, L. and M. Bielen. 1999. Valuing the Ottawa River: The Economic Values and Impacts of
       Recreational Boating. Ohio Sea Grant College Program Technical Bulletin Series Publication
       OSHU-TB-042 2000, 2002.
Idaho Department of Environmental Quality (DEQ). 2008. Idaho Administrative Code, IDAPA 58.01.02:
       Water Quality Standards. Available at
       http: //www. epa. gov/waterscience/standards/wq slibrary/id/id 10 wq s .pdf
November 2009                                                                             12-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
Iowa Department of Natural Resource (IA DNR). 2009. WQS revisions: chloride, Sulfate and TDS.
       Powerpoint presentation. Available at http://www.iowadnr.gov/water/standards/files/ws_pres.pdf
Iowa Waste Reduction Center. 2009. Beneficial Reuse of Solid Wastes. University of Northern Iowa. IAC
       567-108. May 2009.
Jakeman, A. and G. Hornberger.  1993. "How much complexity is warranted in a rainfall-runoff model?"
       Water Resources Research 29:2637-2649. As cited in Schwarz et al. 2006.
Jobson, H.E. 1996. Prediction of travel time and longitudinal dispersion in rivers and streams.
       U.S. Geological Survey Water-Resources Investigations Report 96-4013, 69 p. As cited in
       Schwarz et al. 2006.
Johnes, P.J., B. Moss, and G.L. Phillips, G.L. 1996. "The determination of water quality by land use,
       livestock numbers and population data - testing of a model for use in conservation and water
       quality management." Freshwater Biology 36:951-473. As cited in Schwarz et al. 2006.
Johnston, R.J., E.Y. Besedin and R.F. Wardwell. 2003. "Modeling Relationships Between Use and
       Nonuse Values for Surface Water Quality: A Meta-Analysis." Water Resources Research 39(12):
       1363-1372.
Johnston, R.J., E.Y. Besedin, R. lovanna, C.J. Miller, RF. Wardwell and M.H. Ranson. 2005.
       "Systematic Variation  in Willingness to Pay for Aquatic Resource Improvements and
       Implications for Benefit Transfer: A Meta-Analysis." Canadian Journal of Agricultural
       Economics 53(2-3): 221-248.
Johnston, R.J., J.J. Opaluch, T.A. Origanums, and M.J. Mazzotta. 2001. "Estimating Amenity Benefits of
       Coastal Farmland." Growth and Change 32(summer): 305-325.
Johnston, R.J., M.H. Ranson, E.Y. Besedin and E.G. Helm. 2006. "What Determines Willingness To Pay
       Per Fish? A Meta-Analysis of Recreational Fishing Values. "Marine Resource Economics 21:1-
       32.
Jowett, I.G. 1998. "Hydraulic geometry of New Zealand rivers and its use as a preliminary method of
       habitat assessment." Regulated Rivers - Research and Management 14: 451-466. As cited in
       Schwarz et al. 2006.
Julien, P. 1995. Erosion and Sedimentation. Cambridge University Press.
Just, R.E., D.L. Hueth and A. Schmitz. 2004. The Welfare Economics of Public Policy: A Practical
       Approach to Public Policy. Cheltenham, UK: Edward Elgar.
Kalff, J. 2002. Limnology. Inland Water Ecosystems.  Prentice-Hall, NJ.
Kalainesan, S. 2007. Best Management Practices for Highway Construction Site Sedimentation Basins.
       Doctoral dissertation, University of Pittsburgh School of Engineering.
Kansas Department of Health and Environment. 2004 Surface water nutrient reduction plan. Available at
       http://www.kdhe.state.ks.us/water/download/ks nutrient reduction_plan  12 29 final.pdf.
       Accessed 3/1/2005). As cited in Schwarz et al. 2006.
Kaoru, Y. 1993. "Differentiating Use and Nonuse Values for Coastal Pond Water Quality
       Improvements." Environmental and Resource Economics 3: 487-494.
November 2009                                                                              12-13

-------
Environmental Impact and Benefits Assessment for the C&D Category
Kayhanian, M, K. Murphy, L. Regenmorter, and R. Haller. 2001. "Characteristics of Storm-Water Runoff
       from Highway Construction Sites in California." Transportation Research Record 1743: 33-40.
       Paper No. 01-3447.
Keller, F.J. 1962. Effect of urban growth on sediment discharge, Northwest Branch Anacostia River
       basin, Maryland. U.S. Geological Survey Professional Paper 450-C, pp. C129-131.
Kelly, C.A., J.W.M. Rudd., R.H. Hesslein, D.W. Schindler, P.J. Dillon, C.T. Driscoll, S.A. Gherini, and
       R.E. Hecky. 1987. "Prediction of biological acid neutralization in acid-sensitive lakes. Part 1: A
       Theoretical and Practical Introduction to SPARROW 119." Biogeochemistry 3:  129-140. As cited
       in Schwarz et al. 2006.
Kerr, S. J. 1995. Silt, Turbidity and Suspended Sediments in the Aquatic Environment: An Annotated
       Bibliography and Literature Review.  Ontario Ministry of Natural Resources, Southern Region
       Science & Technology Transfer Unit Technical Report TR-008.
King, D.L. and R.C. Ball. 1964. "The Influence of Highway Construction on a Stream. "Michigan State
       University Agricultural Experiment Station East Lansing 19: 1 -4.
Kirk, K. L. 1992. "Effects of suspended clay on Daphnia body growth and fitness." Freshwater Biology,
       Vol. 28: pp 103-109.^5 cited in Kerr 1995.
Kozlowski, G. and S. Vallellian. 2009. "Eutrophication and endangered aquatic plants: an experimental
       study onBaldellia ranunculoides (L.) Parl. (Alismataceae) " Hydrobiologia 635(1):  181-187.
Kress, B. 2007. "Spill into stream puts spotlight on construction work." The Columbus Dispatch  18
       August 2007. Available at
       http://www.dispatch.com/live/content/local_news/stories/2007/08/18/waterpollute.ART_ART_08
       -18-07 B7 K77LAUJ.html
Kriesel, W., A. Randall, and F. Lichtkoppler. 1993. "Estimating the Benefits of Shore Erosion Protection
       in Ohio's Lake Erie Housing Market." Water Resources Research. 29(4). 795-801.
Krinsky, I. and A.L. Robb.  1986. "Approximating the statistical properties of elasticities." Review of
       Economics and Statistics 68(4):715-719.
Kucken, D.J., J.S. Davis, J.W. Petranka, and C.K. Smith. 1994. "Anakeesta Stream acidification and
       metal contamination: Effects on a salamander community." Journal of Environmental Quality 23:
       1311-1317.
Lafferty, K.D., J.W. Porter, and S.E. Ford.  2004. "Are Diseases Increasing in the Ocean?" Annual Review
       of Ecology, Evolution, and Systematics 35: 31-54.
Lant, C.L., and G.A. Tobin. 1989. "The Economic Value of Riparian Corridors in Cornbelt Floodplains:
       A Research Framework."Professional Geographer (41): 337-349.
Lant, C.L., and R. S. Roberts. 1990. "Greenbelts in the Cornbelt: Riparian Wetlands, Intrinsic Values, and
       Market Failure." Environment and Planning 22: 1375-1388.
LaPerriere, J. D., D. M. Bjerklie, R. C. Simmons, E. Van Nieuwenhuyse,  S. M. Wagener and J. B.
       Reynolds. 1983.  "Effects of gold placer mining on the interior Alaska stream ecosystems." As
       cited in Waters 1995.
November 2009                                                                              12-14

-------
Environmental Impact and Benefits Assessment for the C&D Category
Lee, C.J., P.P. Rasmussen, A.C. Ziegler, and C.C. Fuller. 2009. Transport and Sources of Suspended
       Sediment in the Mill Creek Watershed, Johnson County, Northeast Kansas, 2006-07. USGS
       Scientific Investigations Report 2009-5001.
Lenat, D.R., D.L. Penrose, and K.W. Eagleson. 1981. "Variable effects of sediment addition on stream
       benthos." Hydrobiologia 79(2), pp. 187-194.
Lentz, R.D., Sojka, R.E., and B.E. Mackey. 2002. "Fate and Efficacy of Polyacrylamide Applied in
    Furrow Irrigation: Full-Advance and Continuous Treatments." Journal of Environmental Quality 31:
    661-670.

Levine, J.F., A.E. Bogan, K.H. Pollock, H.A. Devine, L.L. Gustafson, C.B. Eads, P.P. Russell and E.F.
       Anderson. 2003. Distribution of Freshwater Mussel Populations in  Relationship to Crossing
       Structures. North Carolina Department of Transportation/U.S. Department of Transportation.
       Report no. HWY-2003-02.
Levine, J.F., W. G.  Cope, A. E. Bogan, M. Stoskopf, L. L. Gustafson, B. Showers, D. Shea, C. B. Eads, P.
       Lazaro, W. Thorsen, D. Forestier and E. F. Anderson. 2005. Assessment of the Impact of
       Highway Runoff on Freshwater Mussels in North Carolina Streams. North Carolina Department
       of Transportation/U.S. Department of Transportation. Report no. FHWA/NO/2004-03.
Lewis, K. 1973. "The effect of suspended coal particles on the life forms of the aquatic moss,
       Eurhynchium ripariodus." Freshwater Biology 3(73): 251-257. As cited in Kerr 1995.
Lichtkoppler, F.R. and T.W. Elaine. 1999. "Environmental Awareness and Attitudes of Ashtabula County
       Voters Concerning the Ashtabula River Area of Concern:  1996-1997." Journal of Great Lakes
       Resources 25: 500-514.
Lindsey, G. 1994. "Market Models, Protest Bids, and Outliers in Contingent Valuation." Journal of Water
       Resources Planning and Management 12: 121-129.
Line. D. 2009. "Saving Lake Springfield from Sediment." Stormwater Solutions 4(5).
Line, D.E., and N.M. White. 2007. "Effects of Development on Runoff and Pollutant Export." Water
       Environment Research 79(2): 185-190.
Lipton, D. 2004. "The Value of Improved Water Quality to Chesapeake Bay Boaters" Marine Resource
       Economics  19: 265-270
Lloyd, D. S. 1987. "Turbidity as a water quality standard for salmonid habitats in Alaska." North
       American Journal of Fisheries Management 1: 18-33.^4s citedinWaters 1995.
Loomis, J., P. Kent, L. Strange, K. Fausch and A. Covich. 2000. "Measuring the total economic value of
       restoring ecosystem services in an impaired river basin: results from a contingent valuation
       survey" Ecological Economics 33: 103-117.
Loomis, J.B. 1996.  "How Large is the Extent of the Market for Public Goods: Evidence from a Nation
       Wide Contingent Valuation Survey." Applied Economics 28(7): 779-782.
Louisiana Department of Environmental Quality. Baton Rouge, LA. "Surface Water Quality Standards."
       Louisiana Administrative Code (LAC). Title 33, Part IX, Chapter 11. August 6, 2007
Lubliner, Brandi, and Steven Golding (Washington State Department of Ecology). 2005. Stormwater
       Quality Survey of Western Washington Construction Sites, 2003-2005. Publication No. 05-03-
       028.
November 2009                                                                              12-15

-------
Environmental Impact and Benefits Assessment for the C&D Category
Lyke, A.J. 1993. Discrete Choice Models to Value Changes in Environmental Quality: A Great Lakes
       Case Study. Dissertation submitted to the Graduate School of the University of Wisconsin,
       Madison.
Lyman, M. 2009. Personal Communication. Connecticut Department of Environmental Protection. Data
       received via Excel file. March 25, 2009.
Mahler, J.J., P.C. Van Metre, T.J. Bashara, J.T. Wilson and D.A. Johns. 2005. "Parking Lot Sealcoat: An
       Unrecognized Source of Urban Polycyclic Aromatic Hydrocarbons." Environmental Science &
       Technology 39: 5560-5566.
Malins, D.C. and H.O. Hodgins. 1981. "Petroleum and marine fishes: a review of uptake, disposition and
       effects" Environmental Science and Technology 15: 1272.
Mallin, M. A., K.L. Stone, and M.A. Pamperi. 1994. "Phytoplankton community assessments of seven
       Southeast U.S. cooling reservoirs." Water Research 28: 665-673.
Marking, L. L., and T. D. Bills. 1980. "Acute effects of silt and sand sedimentation on freshwater
       mussels." As cited in Waters 1995.
Marxer, L. Personal Communication. Oregon State Department of Environmental Quality. Data received
       via Excel file. March 31, 2009.
Maryland Department of Environmental Protection (Maryland DEP), Montgomery County. 2000.  County
       Investigators Cite Contractor for Toxic Spill in Turkey Branch." Press Release, July 28, 2000.
Masterson, J.P. and R.T. Bannerman. 1994. Impacts of Stormwater Runoff on Urban Streams in
       Milwaukee County, Wisconsin. In Proceedings of American Water Resources Association,
       National Symposium on Water Quality, November 6-10, 1994, Chicago, IL. Pp. 123-133.
Matthews, L.G., F.R. Romans, and K.W. Easter. 1999. Reducing Phosphorous Pollution in the Minnesota
       River: How Much is it Worth? Staff Paper. Department of Applied Economics, University of
       Minnesota.
McClanahan, T., N. Polunin and T. Done. 2002. "Ecological states and the resilience of coral reefs."
       Conservation Ecology 6: 77-103.
McClelland, N. I. 1974. Water Quality Index Application in the Kansas River Basin. U.S. Environmental
       Protection Agency Region 7, Kansas City, Missouri.
McEwan, D. and T. Jackson. 1996. Steelhead Restoration and Management Plan for California. California
       Department of Fish and Game. February, 1996.
McMahon, R.F. 1975. "Effects of artificially elevated water temperatures on the growth, reproduction and
       life cycle of a natural population ofPhysa virgata (Gould)." Ecology 56: 1167-1175.
McMahon G., R.B. Alexander, and S. Qian. 2003. "Support of TMDL programs using spatially
       referenced regression models." ASCE Journal of Water Resources Planning and Management
       129: 315-329. As cited in Schwarz et al. 2006.
McMahon, G. and C. Roessler. 2002. A regression-based approach to understand baseline total nitrogen
       loading for TMDL planning. In Proceedings of Water Environment Federation National TMDL
       Science and Policy Conference, Phoenix, Arizona, November 13-16, 2002.  Phoenix, AZ, Water
       Environment Federation, 28  p. As cited in Schwarz et al. 2006.
November 2009                                                                             12-16

-------
Environmental Impact and Benefits Assessment for the C&D Category
Menendez, D. et al. 2007. AguaClara: Ojojona Sub-team, Final Report. Department of Environmental and
       Civil Engineering, Cornell University. Available at
       http://eswserver.cee.cornell.edu/aguaclara/documents/reports.htm.
Merritt R.W. and K.W. Cummins. 1996. Aquatic Insects of North America. Kendall/Hunt Publishing
       Company, Dubuque, Iowa.
Meyer, B. C., and A. Heritage. 1941. "Effect of turbidity and depth of immersion on apparent
       photosynthesis in Ceratophyllum demersum" Ecology 22: 17-22. As cited in Kerr 1995.
Michigan Department of Environmental Quality (Michigan DEQ). 2006. Water Quality and Pollution
       Control in Michigan-2006 Sections 303(d), 305(b), and 314 Integrated Report. MI/DEQ/EB-
       06/019.
Miller, R.L. and Cruise, J.F. 1995. "Effects of suspended sediments on coral growth: evidence from
       remote sensing and hydrologic modeling." Remote Sensing Environment 53: 177-187.
Missouri Department of Natural Resources (Missouri DNR). 2004. Stream Survey Sampling Report:
       Hinkson Creek Stream Study, Columbia Missouri, Boone County.
Missouri Department of Natural Resources (Missouri DNR). 2005. Stream Survey Sampling Report:
       Phase II, Hinkson Creek Stream Study, Columbia Missouri, Boone County.
Mitchell, R.C. and R.T. Carson. 1989. Using Surveys to Value Public Goods: The Contingent Valuation
       Method. Resources for the Future, Washington, DC.
Moeltner, K., K.J. Boyle, and R.W. Paterson. 2007. "Meta-Analysis and Benefit Transfer for Resource
       Valuation—Addressing Classical Challenges with Bayesian Modeling." Journal of
       Environmental Economics and Management 53(2): 250-269.
Molot, L.A. and P.J. Dillon. 1993. "Nitrogen mass balances and denitrification rates in central Ontario
       Lakes." Biogeochemistry 20: 195-212. As cited in Schwarz et al. 2006.
Montgomery County Department of Environmental Protection (DEP). 2009. Special Protection Area
       Program Annual Report 2007.
Moore, R.B., C.M. Johnston, K.W. Robinson, and J.R.  Deacon. 2004. Estimation of total nitrogen and
       phosphorus in New England streams using spatially referenced regression models. U.S.
       Geological Survey Scientific Investigations Report 2004-5012, 50 p. Available at
       http://water.usgs.gov/pubs/sir/2004/5012/. As cited in Schwarz et al. 2006.
Moore, W.B., and B.A. McCarl. 1987. "Off-site Costs of Soil Erosion: A  Case Study in the Willamette
       Valley." Western Journal of Agricultural Economics 12(1): 42-49.
Morris, G.L., and J. Fan. 1997. Reservoir Sedimentation Handbook: Design and Management of Dams,
       Reservoirs, and Watersheds for Sustainable Use. New York: McGraw-Hill.
Multi-Resolution Land Characteristics Consortium (MRLC). 2001. National Land Cover Database 2001
       (NLCD 2001). Available at http://www.mrlc.gov/mrlc2k  nlcd.asp.
Muscatine L., McCloskey L.R., Marian R.E. 1981. "Estimating the daily contribution of carbon from
       zooxanthellae to coral animal respiration." Limnology and Oceanography 26:  601-611.
National Resources Conservation Service (NRCS). 2008. USDA-NRCS Soil Geochemistry Spatial
       Database March 2008. Database on CD available from NRCS.
November 2009                                                                             12-17

-------
Environmental Impact and Benefits Assessment for the C&D Category
National Oceanic and Atmospheric Administration (NOAA). 2008. National Estuarine Research Reserve
       System (NERRS). Accessed 07/11/08. Available at http://cdmo.baruch.sc.edu/.
National Oceanic and Atmospheric Administration (NOAA) and U.S. Environmental Protection Agency
       (USEPA). 1989a. National Oceanic and Atmospheric Administration and U.S. Environmental
       Protection Agency. Strategic assessment of near coastal waters. Susceptibility of east coast
       estuaries to nutrient discharges: Albemarle/Pamlico Sound to Biscayne Bay. Rockville, MD:
       Strategic Assessment Branch, NOAA.
National Oceanic and Atmospheric Administration (NOAA) and U.S. Environmental Protection Agency
       (USEPA). 1989b. National Oceanic and Atmospheric Administration and U.S. Environmental
       Protection Agency. Strategic assessment of near coastal waters. Susceptibility of east coast
       estuaries to nutrient discharges: Passamaquoddy Bay to Chesapeake Bay. Rockville, MD:
       Strategic Assessment Branch, NOAA.
National Oceanic and Atmospheric Administration (NOAA) and U.S. Environmental Protection Agency
       (USEPA). 1989c. National Oceanic and Atmospheric Administration and U.S. Environmental
       Protection Agency. Strategic assessment of near coastal waters. Susceptibility and status of Gulf
       of Mexico estuaries to nutrient discharge. Rockville, MD:  Strategic Assessment Branch, NOAA.
National Oceanic and Atmospheric Administration (NOAA) and U.S. Environmental Protection Agency
       (USEPA). 1991. National Oceanic and Atmospheric Administration and U.S. Environmental
       Protection Agency. Strategic assessment of near coastal waters. Susceptibility and status of west
       coast estuaries to nutrient discharges: San Diego Bay to Puget Sound. Rockville, MD: Strategic
       Assessment Branch, NOAA.
National Research Council (NRC). 2000. Clean coastal waters - understanding and reducing the effects of
       nutrient pollution.  As cited in Schwarz et al. 2006.
National Research Council (NRC). 200la. Assessing the TMDL approach to water quality management.
       As cited in Schwarz et al. 2006.
National Research Council (NRC). 200 Ib. Ecological indicators for the nation. As cited in Schwarz et al.
       2006.
National Research Council (NRC). 200 Ic. Environmental Impact of Construction and Repair Materials on
       Surface and Ground Waters: Summary of Methodology, Laboratory Results, and Model
       Development. National Cooperative Highway Research Program Report 448.
National Research Council (NRC). 2002a. Review of the USGS National Water-Quality Assessment
       Program. As cited  in Schwarz et al. 2006.
National Research Council (NRC). 2002b. Estimating water use in the United States - a new paradigm for
       the National Water-Use Information Program. As cited in Schwarz et al. 2006.
National Research Council (NRC). 2008. Urban Stormwater Management in the United States.
       Washington, D.C.: National Academies Press. 553 pp.
Natural Resources Conservation Service (NRCS), United States Department of Agriculture. 2005.
       National Resources Inventory—a statistical survey of land use and natural resource  conditions
       and trends on U.S. non-Federal lands. As cited in Schwarz et al. 2006.
Natural Resources Conservation Service (NRCS), U.S. Department of Agriculture. 2008. USDA-NRCS
       Soil Geochemistry Spatial Database  2008. Received on CD from NRCS May 2009.
November 2009                                                                             12-18

-------
Environmental Impact and Benefits Assessment for the C&D Category
NatureServe. 2009. NatureServe Explorer database. Available at http://www.natureserve.org/explorer/.
       Accessed November, 2009.
Navrud, S. and R. Ready. 2007. "Review of Methods for Value Transfer." In: Navrud, S. and R. Ready
       (eds.), Environmental Value Transfer: Issues and Methods. Dordrecht, The Netherlands:
       Springer, pp. 1-10.
Neff, J.M. 1985. Fob/cyclic aromatic hydrocarbons. In: Fundamentals of Aquatic Toxicology. Rand G.M.
       and S.R. Petrocilli (Eds.) Hemisphere. New York.
Nelson, E.J., and D.B. Booth. 2002. "Sediment Sources in an urbanizing, mixed land-use watershed."
       Journal of Hydrology 264: 51-68.
New England Interstate Water Pollution Control Commission (NEIWPCC). 2004. New England
       SPARROW water quality model. Interstate Water Report 1(3): 6-7. As cited in Schwarz et al.
       2006.
New Hampshire Department of Ecological Services (DES). 2007. Hazard Identification for Human and
       Ecological Effects of Sodium Chloride Road Salt. Watershed Management Bureau. Concord, NH.
New Jersey Department of Environmental Protection (DEP). 1999. Findings and Recommendations for
       the Remediation of Historic Pesticide Contamination. Historic Pesticide Contamination Task
       Force.
State of New Mexico. 2005. Standards for Interstate  and Intrastate Waters  20.6.4 NMAC. 2005. Available
       at http: //www. epa. gov/waterscience/standards/wqslibrary/nm/nm_6_wqs .pdf
Newcombe, C.P. 1994. Suspended Sediment in Aquatic Ecosystems: 111 Effects as a Function of
       Concentration and Exposure. Victoria, BC: Habitat Protection Branch, Ministry of Environment,
       Lands and Parks. As cited in Henley et al. 2000.
Newcombe, C.P. 2003. "Impact Assessment Model for Clear Water Fishes Exposed to Excessively
       Cloudy Water" Journal of the American Water Resources Association June 2003: 529-544.
Newcombe, C.P., and J. O. Jensen. 1996. "Channel suspended sediment and fisheries: a synthesis for
       quantitative assessment of risk and impact." North. American Journal of Fisheries Management
       16: 693-727. As cited in Henley et al. 2000.
Ngo, D. and J. Nichols.  1998. Design of Sedimentation Basins. Water Treatment Primer: CE 4124:
       Environmental Information Management. Civil Engineering Department, Virginia Tech.
       Accessed September 29, 2008. Available at
       http://www.cee.vt.edu/ewr/environmental/teach/wtprimer/sedimen/sedimen.html.
Nodvin, S.C., L.B. Weeks, E.P.E. Thomas, and L.J. Lund. 1986. "Alkalization of a High-Elevation Sierra
       Nevada Stream." Water Resources Research 22(7): 1077-1082.
Novotny, V., and G. Chesters. 1989. "Delivery of sediment and pollutants  from nonpoint sources: A
       water quality perspective." Journal of Soil and Water Conservation 44(6): 568-576.
Office of Management and Budget (OMB).  2003. OMB Circular A4. p. 33-34.
Ohio Environmental Protection Agency (Ohio EPA). 1994. Biological and Water Quality Study of the
       Cuyahoga River. Ohio EPA Technical Report EAS/1992-12-11.
November 2009                                                                             12-19

-------
Environmental Impact and Benefits Assessment for the C&D Category
Ohio Environmental Protection Agency (Ohio EPA). 1995. Biological and Water Quality Study of the
       Middle and Lower Great Miami River and Selected Tributaries, 1995. Ohio EPA Technical
       Report MAS/1996-12-8.
Ohio Environmental Protection Agency (Ohio EPA). 1996a. Biological and Water Quality Study of Mill
       Creek - 1995. Ohio EPA Technical Report MAS/1996-7-4.
Ohio Environmental Protection Agency (Ohio EPA). 1996b. Biological and Water Quality Study of the
       Mahoning River Basin. Ohio EPA Technical Report MAS/1995-12-14.
Ohio Environmental Protection Agency (Ohio EPA). 1996c. Biological and Water Quality Study of the
       Upper Great Miami River and Selected Tributaries. Ohio EPA Technical Report MAS/1995-12-
       13.
Ohio Environmental Protection Agency (Ohio EPA). 1997a. 1995-96 Biological and Water Quality Study
       of the Chagrin River and Selected Tributaries. Ohio EPA Technical Report MAS/1996-12-6.
Ohio Environmental Protection Agency (Ohio EPA). 1997b. Biological and Water Quality Study of Big
       Walnut Creek. Ohio EPA Technical Report MAS/1997-10-2.
Ohio Environmental Protection Agency (Ohio EPA). 1997c. Biological and Water Quality Study of Mill
       Creek (Scioto River Basin) and Selected Tributaries. Ohio EPA Technical Report MAS/1996-12-
       11.
Ohio Environmental Protection Agency (Ohio EPA). 1997d. Biological and Water Quality Study of
       Raccoon Creek.  Ohio EPA Technical Report MAS/1996-12-14.
Ohio Environmental Protection Agency (Ohio EPA). 1997e. Biological and Water Quality Study of the
       Upper Hocking River and Selected Tributaries. Ohio EPA Technical Report MAS/1996-12-10.
Ohio Environmental Protection Agency (Ohio EPA). 1997f. Biological and Water Quality Study of Twin
       Creek and Selected Tributaries 1995. Ohio EPA Technical Report MAS/1996-12-9.
Ohio Environmental Protection Agency (Ohio EPA). 1998a. Biological and Water Quality Study of
       Blacklick Creek and Selected Tributaries. Ohio EPA Technical Report MAS/1997-12-7.
Ohio Environmental Protection Agency (Ohio EPA). 1998b. Biological and Water Quality Study of Little
       Beaver Creek and Big Beaver Creek - 1997. Ohio EPA Technical Report MAS/1998-5-1.
Ohio Environmental Protection Agency (Ohio EPA). 1998c. Biological and Water Quality Study of
       Lower Big Walnut Creek and Walnut Creek Tributaries 1996. Ohio EPA Technical Report
       MAS/1997-12-10.
Ohio Environmental Protection Agency (Ohio EPA). 1998d. Biological and Water Quality Study of the
       Little Cuyahoga River and Tributaries. Ohio EPA Technical Report MAS/1997-12-9.
Ohio Environmental Protection Agency (Ohio EPA). 1998e. Biological and Water Quality Study of the
       Little Miami River Basin, 1998. Ohio EPA Technical Report MAS/1999-12-3.
Ohio Environmental Protection Agency (Ohio EPA). 1999a. Biological and Water Quality Study of the
       Cuyahoga River and Selected Tributaries. Ohio EPA Technical Report MAS/1997-12-4.
Ohio Environmental Protection Agency (Ohio EPA). 1999b. Biological and Water Quality Study of the
       Middle Scioto River and Alum Creek. Ohio EPA Technical Report MAS/1997-12-12.
November 2009                                                                             12-20

-------
Environmental Impact and Benefits Assessment for the C&D Category
Ohio Environmental Protection Agency (Ohio EPA). 1999c. Biological and Water Quality Study of the
       Rocky River and Selected Tributaries. Ohio EPA Technical Report MAS/1998-12-3.
Ohio Environmental Protection Agency (Ohio EPA). 1999d. Biological and Water Quality Study of
       Walnut Creek and Selected Tributaries 1996. Ohio EPA Technical Report MAS/1997-12-5.
Ohio Environmental Protection Agency (Ohio EPA). 2000. Biological and Water Quality Study of
       Marion Area Streams 1998. Ohio EPA Technical Report MAS/1999-12-6.
Ohio Environmental Protection Agency (Ohio EPA). 200 la. Biological and Water Quality Study of the
       Olentangy River and Selected Tributaries 1999. Ohio EPA Technical Report MAS/2000-12-6.
Ohio Environmental Protection Agency (Ohio EPA). 200 Ib. Biological and Water Quality Study of the
       Stillwater River Watershed. Ohio EPA Technical Report MAS/2001-12-8.
Ohio Environmental Protection Agency (Ohio EPA). 2003a. Biological and Water Quality Study of the
       Big Walnut Creek Basin, 2000. Ohio EPA Technical Report EAS/2003-11-10.
Ohio Environmental Protection Agency (Ohio EPA). 2003b. Biological and Water Quality Study of the
       Sandusky River and Selected Tributaries 2001. Ohio EPA Technical Report EAS/2003-4-6.
Ohio Environmental Protection Agency (Ohio EPA). 2004. Biological and Water Quality Study of the Big
       Darby Creek Watershed, 2001/2002. Ohio EPA Technical Report EAS/2004-6-3.
Ohio Environmental Protection Agency (Ohio EPA). 2005. Biological and Water Quality Study of the
       Olentangy River, Whetstone Creek and Select Tributaries, 2003-2004. Ohio EPA Technical
       Report EAS/2005-12-6.
Ohio Environmental Protection Agency (Ohio EPA). 2006a. Biological and Water Quality Study of the
       Chagrin River and Selected Tributaries 2003-04. Ohio EPA Technical Report EAS/2006-12-7.
Ohio Environmental Protection Agency (Ohio EPA). 2006b. Biological and Water Quality Study of the
       Grand River Basin 2003-2004. Ohio EPA Technical Report EAS/2006-11-6.
Ohio Environmental Protection Agency (Ohio EPA). 2006c. Recreational Use Water Quality Survey of
       the Sugar Creek Watershed, 2005. Ohio EPA Technical Report NEDO/2006-02-01.
Ohio Environmental Protection Agency (Ohio EPA). 2006d. Biological and Water Quality Study of
       Walnut Creek and Tributaries. Ohio EPA Technical Report EAS/2006-12-8.
Ohio Environmental Protection Agency (Ohio EPA). 2007a. Biological and Water Quality Study of
       Furnace Run Watershed. Ohio EPA Technical Report NEDO/2007-02-01.
Ohio Environmental Protection Agency (Ohio EPA). 2007b. Biological and Water Quality Study of the
       BlanchardRiver.  Ohio  EPA Technical Report EAS/2007-6-2.
Ohio Environmental Protection Agency (Ohio EPA). 2007c. Biological and Water Quality Study of Twin
       Creek and Select Tributaries 2005. Ohio EPA Technical Report EAS/2007-10-03.
Olsen, D., Richards, Jack, and Scott, R. D. 1991. Existence and Sport Values for Doubling the Size of
       Columbia River Basin Salmon and Steelhead Runs. Rivers. Vol. 2, No.  1. pp. 44-56.
Opaluch, J.J., T. Origanums, M.J. Mazzotta, J. Diamantides, and R. Johnston. 1998. Resource and
       Recreational Economic Values for the Peconic Estuary. Report prepared for Peconic Estuary
       Program, Suffolk County Department of Health Services, Riverhead, NY, by Economic Analysis,
       Inc., Peace Dale,  Rhode Island.
November 2009                                                                             12-21

-------
Environmental Impact and Benefits Assessment for the C&D Category
Osterkamp, W.R., P. Heilman, and L.J. Lane. 1998. "Economic Considerations of a Continental
       Sediment-monitoring Program." InternationalJournal of Sediment Research 13(4): 12-24.
Owens, D.W., P. Jopke, D.W. Hall, J. Balousek and A. Roa. (USGS). 2000. Soil Erosion from Two Small
       Construction Sites, Dane County, Wisconsin. USGS Fact Sheet FS-109-00.
Paffenhofer, G.A. 1972. "The effects of suspended 'red mud' on mortality, body weight and growth of the
       marine planktonic copepod Calanus helgolandius." Water, Air and Soil Pollution 1: 314-321.^45
       cited in Kerr 1995.
Pain, S. 1987. "After the gold rush." As cited in Waters 1995.
Parsons S. and B. Jefferson. 2006. Introduction to Potable  Water Treatment Processes. Blackwell
       Scientific. Oxford.
Paul, M.J. and J.L. Meyer. 2001. "Streams in the Urban Landscape" Annual Review of Ecology and
       Systematics 32: 333-65.
Peierls, B.L., N.F. Caraco, M.L. Pace,  and J.J.  Cole. 1991.  "Human influence on river nitrogen." Nature
       350: 386-387. As cited in Schwarz et al. 2006.
Peterson, L.A. and D. Nyquist. 1972. "Effects of Highway  Bridge Construction on a Subarctic Stream."
       The Northern Engineer 4: 18-20.
Philipp E, and K. Fabricius. 2003. "Photophysiological stress in scleractinian corals in response to short-
       term sedimentation." Journal of Experimental Marine Biology and Ecology 287: 57-78
Poe, G.L., K.J. Boyle, and J.C. Bergstrom. 2001. "A Preliminary Meta Analysis of Contingent Values for
       Ground Water Quality Revisited." In: J.C. Bergstrom, K.J. Boyle and G.L. Poe, eds. The
       Economic Value of Water Quality. Northampton, MA: Edward Elgar Publishers.
Preston, S.D. and J.W. Brakebill. 1999. Application of spatially referenced regression modeling for the
       evaluation of total nitrogen loading in the Chesapeake Bay watershed: Baltimore, MD, U.S.
       Geological Survey Water Resources Investigations Report 99-4054, lip. Available at
       http://md.water.usgs.gov/publications/wrir-99-4054/. As cited in Schwarz et al. 2006.
Puerto Rico Environmental Quality Board (EQB). 2003. Puerto Rico Water Quality Standards
       Regulation. Available at http://www.epa.gov/waterscience/standards/wqslibrary/pr/pr_2_wqs.pdf
Quinn, J.M., RJ. Davies-Colley, C.W. Hickey, M.L. Vickers, and P.A. Ryan. 1992. "Effects of clay
       discharges on streams." Hydrobiologia 248(3): 235-247. As cited in Bilotta and Brazier 2008.
Reckhow, K.H. and S.C. Chapra. 1983. Engineering approaches for lakes management, vol. 1 - data
       analysis and empirical modeling. Butterworth publishers, Boston, MA. 340 p. As cited in Schwarz
       et al. 2006.
Reed, J.R. 1977. Stream Community Response to Road Construction Sediments. Virginia Water
       Resources Research Center, Virginia Polytechnic Institute and State University.
Reed, Lloyd A.  (USGS). 1980. Suspended-Sediment Discharge, in Five Streams near Harrisburg,
       Pennsylvania, Before, During, and After Highway  Construction. Geological Survey Water-
       Supply Paper 2072.
November 2009                                                                              12-22

-------
Environmental Impact and Benefits Assessment for the C&D Category
Reid, S.M. and P.O. Anderson. 1999. "Review of the effects of sediment released during open-cut
       pipeline water crossings on stream and river ecosystems." Canadian Water Resources Journal
       24: 23-29.
Reid, S.M., G. Isaac, S. Metikosh and J. Evans. 2003. "Physiological Response of Rainbow Trout to
       Sediment Released during Open-Cut Pipeline Water Crossing Construction." Water Quality
       Resources Journal of Canada 38(3): 473-481.
Renshaw, C.E., B.C. Bostick, X.  Feng, C.K. Wong, E.S. Winston, R. Karimi, C.L. Folt, and C.Y. Chen.
       2006. "Impact of Land Disturbance on the Fate of Arsenical Pesticides." Journal of
       Environmental Quality 35: 61-67.
Renwick, W.H., S.V. Smith, J.D. Bartley, and R.W. Buddemeier. 2005. "The role of impoundments in the
       sediment budget of the conterminous United  States." Geomorphology 71(1-2):  99-111.
Reynolds, J.B., R.C. Simmons and A.R. Burkholder.  1989. "Effects of placer mining discharge on health
       and food of Arctic grayling." Water Resources Bulletin 25: 625-635. As cited in Henley et al.
       2000.
Ribaudo, M.O. 1989. Water Quality Benefits from the Conservation Reserve Program.  Agricultural
       Economic Report No. 606. Resources and Technology Division, Economic Research Service,
       U.S.D.A., Washington, D.C.
Richter, B.D., D.P. Braun, M.A. Mendelson and L.L. Master. 1997. Threats to imperiled freshwater
       fauna. Conservation Biology 11: 1081-1093.  as cited in Dodds et al. 2009.
Rivier, B., and J. Seguier.  1985. "Physical and biological effects of gravel extraction in river beds." As
       cited in Waters 1995.
Roberts, L.A., and J.A. Leitch. 1997. Economic Valuation of Some Wetland Outputs of Mud Lake.
       Agricultural Economics Report No. 381, Department of Agricultural Economics, North Dakota
       Agricultural Experiment Station, North Dakota State University.
Robertson, M.. 1957. "The effects of suspended material on the productive rate of daphnia magna."
       Publications of the Institute  of Marine Science (University of Texas) 4, 265-277. As cited in
       Bilotta and Brazier 2008.
Robinson, G.R., P.  Larkins, C.J. Boughton, B.W. Reed, and P.L. Sibrell. 2007. "Assessment of
       Contamination from Arsenical Pesticide Use  on Orchards in the Great Valley region, Virginia and
       West Virginia, USA." Journal of Environmental Quality 36:  654-663.
Rogers C.S.  1990. "Responses of coral reefs and reef organisms to sedimentation. "Marine Ecology
       Progress Series 62:  185-202
Rosenberg, D.M. and A.P. Wiens. 1978. "Effects of sedimentation on macrobenthic invertebrates on a
       northern Canadian river." Water Research. 12(10), 753-763. As cited in Bilotta and Brazier 2008
Rosenberger, R.S. and J.B. Loomis.  2000a. "Using Meta-Analysis for Benefit Transfer: In-Sample
       Convergent Validity Tests of an Outdoor Recreation Database." Water Resources Research
       36(4): 1097-1107.
Rosenberger, R.S. and J.B. Loomis.  2000b. "Panel  Stratification in Meta-Analysis of Economic Studies:
       An Investigation of Its Effects in the Recreation Valuation Literature." Journal of Agricultural
       and Applied Economics 32(3): 459-470.
November 2009                                                                              12-23

-------
Environmental Impact and Benefits Assessment for the C&D Category
Rosenberger, R.S. and R.J. Johnston. 2007. Selection Effects in Meta-Valuation Function Transfers.
       Meta-Analysis in Economics Conference, September 27-30, 2007, Sandbjerg Manor, S0nderborg,
       Denmark.
Rosenberger, R.S. and T.D. Stanley. 2006. "Measurement, Generalization, and Publication: Sources of
       Error in Benefit Transfers and Their Management." Ecological Economics 60(2): 372-378.
Rosenberger, R.S. and T.T. Phipps. 2007. "Correspondence and Convergence in Benefit Transfer
       Accuracy: Meta-Analytic Review of the Literature." In: Navrud, S. and R. Ready (eds.),
       Environmental Value Transfer: Issues and Methods. Dordrecht, The Netherlands: Springer.
       pp.23-43.
Rowe, D., E. Graynoth, G. James, M. Taylor and L. Hawke. 2003. "Influence of turbidity and fluctuating
       water levels on the abundance and depth distribution of small, benthic fish in New Zealand alpine
       lakes" Ecology of Freshwater Fish 12: 216-227.
Rowe, R.D., W.D. Schulze, B. Hurd and D. Orr. 1985. Economic Assessment of Damage Related to the
       Eagle Mine Facility. Energy and Resource Consultants, Inc., Boulder, CO.
Runkle, R.L., C.G. Crawford, and T.A. Cohn. 2004. Load estimator (LOADEST) - A FORTRAN
       program for estimating constituent loads in streams and rivers: U.S. Geological Survey
       Techniques and Methods 4-A5, 75 p.. As cited in Schwarz et al. 2006.
Ryan, Paddy A. 1991. "Environmental effects of sediment on New Zealand streams: a review." New
       Zealand Journal of Marine and Freshwater Research 25: 207-221.
Samsel Jr., G.L. 1973. "Effects of sedimentation on the algal flora of a small recreational impoundment."
       Water Resources Bulletin 9: 1145-1152. As cited in Kerr  1995.
Sanders, L.  B., R.G. Walsh, and J.B. Loomis. 1990. "Toward Empirical Estimation of the Total Value of
       Protecting Rivers." Water Resources Research. 26(7): 1345-1357.
Santos, J.M.L.  1998. The Economic Valuation of Landscape Change. Cheltenham, UK: Edward Elgar.
SAS. 1993.  SAS/ETS user's guide, version 6 (2nded.). Gary, NC, SAS Institute Inc., 1022 p. As citedin
       Schwarz et al. 2006.
SAS. 1999.  SAS OnlineDoc, version 8. Available at http://v8doc.sas.com/sashtml/. Accessed 4/29/2005.
       SAS Inc. As cited in Schwarz et al. 2006.
Saunders, J.W., and M.W. Smith. 1965. "Changes in a stream population of trout associated with increase
       silt." Journal of Fisheries Research Board Canada 22: 395-404. As cited in Kerr 1995.
Science Daily.  2009. Pavement sealcoat a source of toxins in storm runoff. April 14, 2009. Available at
       http://www.sciencedailv.com/releases/2009/04/090408145548.htm
Schuman, G.E., R.G. Spomer, and R.F. Piest. 1973a. "Phosphorus losses from four agricultural
       watersheds on Missouri Valley loess." Soil Science Society of America Proceedings 37: 424-427.
       As citedin Sharpley 1985.
Schuman, G.E., R.E. Burwell, R.F. Piest, and R.G. Spomer. 1973b. "Nitrogen losses in surface runoff
       from agricultural watersheds on Missouri Valley loess." Journal of Environmental Quality 2:
       299-302. As citedin Sharpley 1985.
November 2009                                                                             12-24

-------
Environmental Impact and Benefits Assessment for the C&D Category
Schulze, W.D., R.D. Rowe, W.S. Breffle, R.R. Boyce, and G.H. McClelland. 1995. Contingent Valuation
       of Natural Resource Damages Due to Injuries to the Upper Clark Fork River Basin. State of
       Montana, Natural Resource Damage Litigation Program. Prepared by: RCG/Hagler Bailly,
       Boulder, CO.
Schwarz, G.E (USGS). 2008a. A Preliminary SPARROW model of suspended sediment for the
       conterminous United States: U.S. Geological Survey Open-File Report 2008-1205, 8 p.,
       Available at http://pubs.usgs.gov/of/2008/1205
Schwarz, G.E. (USGS). 2008b. Personal Communication. June 4, 2008.
Schwarz, G.E., R.A. Smith, R.B. Alexander, and J.R. Gray. 2001. A Spatially Referenced Regression
       Model (SPARROW) for Suspended Sediment in Streams of the Coterminous U.S. U.S.
       Subcommittee on Sedimentation, Proceedings of the Seventh Federal Interagency Sedimentation
       Conference, March 25-29 2001, Reno, Nevada USA. p. VII-80-7. As cited in Schwarz et al. 2006.
Schwarz, G.E., A.B. Hoos, R.B. Alexander, and R.A Smith. 2006. The Sparrow Surface Water Quality
       Model: Theory, Application and User Documentation. U.S. Geological Survey Techniques and
       Methods Book 6, Section B, Chapter 3.
Seattle and King County Public Health. 2001. Vashon/Maury Island Soil Study 1999-2000, Final Report.
Seitzinger, S., RV. Styles, E.W. Boyer, R.B. Alexander, G. Billen, RW. Howarth, B. Mayer, andN.
       vanBreemen. 2002. "Nitrogen retention in rivers - model development and application to
       watersheds in the Eastern U.S." Biogeochemistry 57: 199-237. As cited in Schwarz et al. 2006.
Selbig, W.R., P.L. Jopke, D.W. Marshall and M.J. Sorge (USGS).  2004. Hydrologic, Ecologic, and
       Geomorphic Responses of Brewery Creek to Construction of a Residential Subdivision, Dane
       County, Wisconsin, 1999-2002. Scientific Investigations Report 2004-5156.
Selbig, W.R., and R.T. Bannerman. 2008. A comparison of runoff quantity and quality from two small
       basins undergoing implementation of conventional- and low-impact-development (LID)
       strategies: Cross Plains, Wisconsin, water years 1999-2005. U.S. Geological Survey Scientific
       Investigations Report 2008-5008.
Servizi, J.A., and D.W. Martens. 1987. "Some effects of suspended Fraser River sediment on sockeye
       salmon, Oncorhynchus nerka." p. 254-264. As cited in Reid et al. 2003.
Sharpley, A.N. 1985. "The selective erosion of plant nutrients in runoff." Soil Science Society of America
       Journal 49: 1527-1534.
Shen, H.W., and P.Y. Julien. 1993. "Erosion and Sediment Transport." In: Maidment, David (Ed.).
       Handbook of Hydrology. McGraw-Hill, Inc. Ch. 12.
Shields, F.D. and T.G.  Sanders. 1986. "Water Quality Effects of Excavation and Diversion." Journal of
       Environmental Engineering 112(2): 211-228.
Shrestha, R. K., J.R.R. Alavalapati. 2004. "Valuing Environmental Benefits of Silvopasture Practice: A
       Case Study of the Lake Okeechobee Watershed in Florida." Ecological Economics 49: 349-359.
Shrestha, R., R.  Rosenberger and J. Loomis. 2007. "Benefit Transfer Using Meta-Analysis in Recreation
       Economic Valuation, in Benefit Transfer Accuracy: Meta-Analytic Review of the Literature." In:
       Navrud, S. and R. Ready (eds.). Environmental Value Transfer: Issues andMethods. Dordrecht,
       The Netherlands: Springer, pp. 161-177.
November 2009                                                                             12-25

-------
Environmental Impact and Benefits Assessment for the C&D Category
Simons, D.B., R.M. Li and W. Fullerton. 1981. Theoretically derived sediment transport equations for
       Pima County, Arizona. Prepared for Pima County DOT and Flood Control District, Tucson,
       Arizona. Ft. Collins, Colorado: Simons, Li and Assoc..
Singer, J.D. 1998. "Using SAS PROC MIXED to fit multilevel models, hierarchical models, and
       individual growth models." Journal of Educational and Behavioral Statistics 23(4): 323-355.
Singh, V.P. 1995. Computer models of watershed hydrology: Highlands Ranch, CO. Water Resources
       Publications, 1,144 p. As cited in Schwarz et al. 2006.
Smith, R.A., G.E. Schwarz, and R.B. Alexander. 1997. "Regional interpretation of water-quality
       monitoring data." Water Resources Research 33: 2781-2798. As citedin Schwarz et al. 2006.
Smith, R.A., RB. Alexander, G.D. Tasker, C.V. Price, C.V., K.W. Robinson, and D.A. White. 1994.
       Statistical modeling of water quality in regional watersheds. In: U.S. Environmental Protection
       Agency. Proceedings of watershed '93 - a national conference on watershed management,
       Alexandria, Virginia, March 21-24, 1993: Washington, DC, U.S. Environmental Protection
       Agency, EPA 840-R-94-002, p.  751-754. As cited in Schwarz et al. 2006.
Smith, V.K., and L. Osborne. 1996. "Do Contingent Valuation Estimates Pass the Scope Test? A Meta
       Analysis." Journal of Environmental Economics and Management 31(3):  287-301.
Smith, V.K., and S.K. Pattanayak. 2002. "Is Meta-Analysis a Noah's  Ark for Non-Market Valuation?"
       Environmental and Resource Economics 22: 271-296.
Smith, V.K., and Y. Kaoru. 1990. "Signals or Noise? Explaining the Variation in Recreation Benefit
       Estimates." American Journal of Agricultural Economics 42(2): 419-433.
Smith, V.K., G. Van Houtven, and S.K.  Pattanayak. 2002. "Benefit Transfer via Preference Calibration:
       'Prudential Algebra' for Policy." Land Economics 78(1): 132-152.
Sohngen, B. and J. Rausch. 1998. "Soil Erosion in the Maumee River Basin: A Case Study Using Market
       Methods to Value Environmental Externalities." Draft. Department of Agricultural Economics,
       Ohio State University. June 5.
Spash, C.L. and A. Vatn. 2006. "Transferring Environmental Value Estimates: Issues and Alternatives."
       Ecological Economics 60(2): 379-388.
Srinivasan, R., J.G. Arnold, R.S. Muttiah, D. Walker, and P.T. Dyke. 1993. "Hydrologic unit modeling of
       the United States (HUMUS)" In: Yang, S. (ed.). Advances in hydro-science and engineering, vol
       I, part A: University, MS, Center for Computational Hydroscience and Engineering, School of
       Engineering, University of Mississippi, p.  451-456. As cited in Schwarz et al. 2006.
Stafford-Smith, M.G. and R.F.G. Ormond.  1992. "Sediment-rejection mechanisms of 42 species of
       Australian scleractinian corals. "Australian Journal of Marine Freshwater Research 43: 683-705.
Stahre, P. and B. Urbonas. 1990. Stormwater Detention for Drainage, Water Quality, and CSO
       Management. Prentice Hall, Englewood Cliffs, NJ.
Staver, L., K. Staver and J. Stevenson. 1996. "Watershed discharge effects on water quality in the
       Choptank River estuary: Implications for watershed management. "Estuaries 19: 342-358.
Stephens, D.W., K.R. Thompson and J.B. Wangsgard (USGS). 1996. Potential Effects of Coal Mining
       and Road Construction on the Water Quality of Scofield Reservoir and its Drainage Area, Central
       Utah, October  1982 to October 1984. Water Resources Investigations Report 96-4020.
November 2009                                                                              12-26

-------
Environmental Impact and Benefits Assessment for the C&D Category
Stream Solute Workshop. 1990. "Concepts and methods for assessing solute dynamics in stream
       ecosystems" Journal of the North American Benthological Society 9: 95-119. As cited in Schwarz
       et al. 2006.
Streiner, C.F. and Loomis, J.B. 1996. "Estimating the Benefits of Urban Stream Restoration Using the
       Hedonic Price Method." Rivers. 5(4): 267-278.
Stone, D. and K. Anderson. 2009. Yesterday's Orchard...  Today's Home: Legacy Pesticides on Former
       Orchard Properties. Oregon State University Extension Service. EC1513-E.
Stout, B.M. and C.B. Coburn, Jr. 1989. "Impact of highway construction on leaf processing in aquatic
       habitats of eastern Tennessee." Hydrobiologia 178: 233-242.
Stumborg, B.E., K.A. Baerenklau and R.C. Bishop. 2001. "Nonpoint Source Pollution and Present
       Values: A Contingent Valuation of Lake Mendota." Review of 'Agricultural Economics. 23(1):
       120-132.
Stumm and Morgan. 1981. Aquatic Chemistry 3rd edition. Wiley.
Suffet, I.H., C. Ana, D. Chou, M.J. Maguire and S. Butterworth. 1996. "American Water Works
       Association (AWWA) Taste and Odor Survey." Journal of the American  Water Works
       Association 88:  168-180.
Sullivan, K., D.J. Martin, R.D. Cardwell, J.E. Toll and S. Duke. 2000. "An analysis of the effects of
       temperature on salmonids of the Pacific Northwest with implications for selecting temperature
       criteria." Sustainable Ecosystems Institute, Portland, Oregon.
Sutherland, R. J., and R.G. Walsh. 1985. "Effect of Distance on the Preservation Value of Water
       Quality." Land Economics 61(3): 282-291.
SWAT (Soil and Water Assessment Tool).  2005. Theoretical Documentation. Available at
       http://www.brc.tamus.edu/swat/downloads/doc/swat2005/SWAT%202005%20theorv%20final.pd
       f
Sweka, J.A. and K.J. Hartman. 2001. "Influence of Turbidity on Brook  Trout Reactive Distance and
       Foraging Success." Transactions of the American Fisheries Society 130: 138-146.
Tan, K. and D. Thirumurthi. 1978. "Highway construction impact on water-supply lakes." Journal of the
       Environmental Engineering Division (American Society of Civil Engineers) 104(EE-5): 997-
       1011., as citedin Kalainesan 2007.
Taylor, B.R. and J.C. Roff. 1986. "Long-Term Effects of Highway Construction on the Ecology of a
       Southern Ontario  Stream." Environmental Pollution (Series A)  40: 317-344.
Thorn, R. 2009. Personal Communication. Pacific Northwest National Laboratory. Directed to University
       of Washington Department of Ecology. Available at
       http://www.ecy.wa.gov/programs/eap/mar_wat/data.html. March 16. 2009.
Tisdale, S.L., W.L. Nelson and J.D. Beaton. 1985. Soil Fertility and Fertilizers. London: Macmillan
       Publishing. As cited in Binger, Theurer, and Yuan 2009.
Tornblom, E. and B. Bostrom. 1995. "Benthic microbial response to a sedimentation event at low
       temperature in sediments of a eutrophic lake" Marine and Freshwater Research 46:  33-43.
November 2009                                                                              12-27

-------
Environmental Impact and Benefits Assessment for the C&D Category
Trimble, S.W. and P. Crosson. 2000. "U.S. soil erosion rates - myth and reality." Science 289: 248-250.
       As cited in Schwarz et al. 2006.
Tsui, P.T.P. and P.J. McCart.  1981. "Effects of stream-crossing by a pipeline on the benthic
       macroinvertebrate communities of a small mountain stream." Hydrobiologia 79: 271-276.
U.S. Army Corps of Engineers (USAGE). 1996. USAGE National Inventory of Dams. Available at
       http://crunch.tec.army.mil/nidpublic/webpages/nid.cfm.
U.S. Army Corps of Engineers (USAGE). 2002. Final Dredged Material Management Plan and
       Environmental Impact Statement (DMMP/EIS) McNary Reservoir and Lower Snake River
       Reservoirs. Final, July 2002.
U.S. Army Corps of Engineers (USAGE). 2007. June 2006 Flood Investigation for Cameron Run, Fairfax
       County, Virginia. Final Report January 2007.
U.S. Army Corps of Engineers (USAGE). 2009. "Dredging Information System." Navigational Data
       Center. Accessed 16 September 2009. Available at
       http://www.iwr.usace.army.mil/ndc/data/datadrgsel.htm
U.S. Census Bureau. 2006a. The 2006 American Community Survey. Available at
       http://www.census.gov/acs/www/index.html (Accessed April 1, 2008).
U.S. Census Bureau. 2006b. US Population Estimates by State. Population Estimates. Available at
       htto://www.census.gov/popest/states/tables/NST-EST2007-01 .xls (Accessed April  1, 2008)
U.S. Department of Agriculture (USDA) 1986. Soil Conservation Service TR-55: Urban Hydrology for
       Small Watersheds.
U.S. Department of Agriculture (USDA) .1994.  State Soil Survey Geographic (STATSGO) Data Base:
       U.S. Department of Agriculture, Soil Conservation Service, National Cooperative Soil Survey,
       Miscellaneous Publication Number 1492. Available at
       http://www.ncgc.nrcs.usda.gov/products/datasets/statsgo/.
U.S. Department of Labor. 2008.  Consumer Price Index (CPI): All Urban Consumers. Bureau of Labor
       Statistics. Available at ftp://ftp.bls.gov/pub/special.requests/cpi/cpiai.txt. Accessed April 1, 2008.
U.S. Environmental Protection Agency (USEPA) Science Advisory Board. 2007. Advisory on EPA's
       Issues in Valuing Mortality Risk Reduction (EPA-SAB-08-001). U.S. EPA, Office of the
       Administrator, Washington, DC, October 2007.
U.S. Environmental Protection Agency (USEPA). Undated. Technical Factsheet on: Nitrate/Nitrite.
       Available at http://www.epa.gov/safewater/pdfs/factsheets/ioc/tech/nitrates.pdf
U.S. Environmental Protection Agency (USEPA).  1987. Quality Criteria for Water 1986 [The Gold
       Book].Office of Water. EPA 440/5-96-001. Available at
       http://www.epa.gov/waterscience/criteria/library/goldbook.pdf
U.S. Environmental Protection Agency (USEPA).  1988. Ambient Water Quality Criteria for Chloride —
       1988. Office of Water, Regulations and Standards, Criteria and Standards Division, EPA 440/5-
       88-001.
U.S. Environmental Protection Agency (USEPA).  1995. "Algal and microbial communities as indicators
       of prairie wetland integrity." In: Bioindicators for Assessing Ecological Integrity of Prairie
       Wetlands. EPA/600/R-96/082. September, 1995.
November 2009                                                                              12-28

-------
Environmental Impact and Benefits Assessment for the C&D Category
U.S. Environmental Protection Agency (USEPA). 1996a. Technology Transfer Handbook: Management
       of Water Treatment Plant Residuals (EPA 625-R-95-008), Office of Research and Development,
       April 1996.
U.S. Environmental Protection Agency (USEPA). 1996b. The USEPA Reach File Version 1.0 (RF1) for
       the Conterminous United States (CONUS): Availble at
       http://www.epa.sov/waterscience/BASINS/metadata/rfl.htm. As cited in Schwarz et al. 2006.
U.S. Environmental Protection Agency (USEPA). 1997. Development of Mixing Zone Dilution Factors.
       Final report, EPA Contract No. 68-D3-0013 (Work Assignment No. 2-12) and EPA Contract No.
       68-W6-0023 (Work Assignment No. 1-2). Prepared by Versar, Inc. for the U.S. Environmental
       Protection Agency, Economics, Exposure and Technology Division, Office  of Pollution
       Prevention and Toxics, Washington, DC.
United States Environmental Protection Agency (USEPA). 1998. Stage  1 Disinfectants and Disinfection
       Byproducts Rule. 63 FR 69390 - 69476, December 16, 1998.
U.S. Environmental Protection Agency (USEPA). 1999. Rapid Bioassessment Protocols for Use in
       Wadeable Streams and Rivers. Periphyton, Benthic Macroinvertebrates and Fish. 2nd. Edition.
       Office of Water, Washington, D.C. EPA 841-B-99-002. July 1999.
U.S. Environmental Protection Agency (USEPA). 2000a. Arsenic in Drinking Water Rule- Economic
       Analysis (EPA 815-R-00-026). U.S. EPA, Office of Water, Washington, DC, December, 2000.
U.S. Environmental Protection Agency (USEPA). 2000b. Guidelines for Preparing Economic Analyses.
       (EPA 240-R-00-003). U.S. EPA, Office of the Administrator, Washington, DC, September 2000.
U.S. Environmental Protection Agency (USEPA). 2000c. Nutrient Criteria Technical Guidance Manual.
       Lakes and Reservoirs. 1st ed. Office of Water. EPA-822-800-001. April 2000.
U.S. Environmental Protection Agency (USEPA). 2000d. Nutrient Criteria Technical Guidance Manual.
       Rivers and Streams. 1st ed. Office of Water. EPA-822-B-00-002. July 2000.
U.S. Environmental Protection Agency (USEPA). 2001. Nutrient Criteria Technical Guidance Manual.
       Estuaries and Coastal Marine Waters. Office of Water. EPA-822-B-01-003. October 2001.
U.S. Environmental Protection Agency (USEPA). 2002. Estimation of National Economic Benefits Using
       the National Water Pollution Control Assessment Model to Evaluate Regulatory Options for
       Concentrated Animal Feeding Operations (EPA-821-R-03-009). U.S. EPA,  Office of Water,
       Washington, DC, December 2002.
U.S. Environmental Protection Agency (USEPA). 2004a. Biological Opinion for Threatened and
       Endangered Species USEPA Superfund Program Revised Proposed Clean-up Plan for the
       Milltown Reservoir Sediments Operable Unit of the Milltown Reservoir/Clark Fork River
       Superfund Site. Endangered Species Act- Section 7 Consultation.
U.S. Environmental Protection Agency (USEPA). 2004b. Development Document for Final Action for
       Effluent Guidelines and Standards for the Construction and Development Category. EPA-821-B-
       04-001.
U.S. Environmental Protection Agency (USEPA). 2004c. The Incidence and Severity of Sediment
       Contamination in Surface Waters of the United States: National Sediment Quality Survey Second
       Edition. Office of Science and Technology. EPA-823-R-04-007.
November 2009                                                                             12-29

-------
Environmental Impact and Benefits Assessment for the C&D Category
U.S. Environmental Protection Agency (USEPA). 2004d. Regional Analysis for the Final Section 316(b)
       Phase II Existing Facilities Rule (EPA-821-R-04-006). U.S. EPA, Office of Water, Washington,
       DC, February 2004.
U.S. Environmental Protection Agency (USEPA). 2005a. Guidance for 2006 Assessment, Listing and
       Reporting Requirements Pursuant to Sections 303(d), 305(b) and 314 of the Clean Water Act.
       U.S. EPA, Office of Water, Washington, DC. July 29, 2005.
U.S. Environmental Protection Agency (USEPA). 2005b. Procedures for the Derivation of Equilibrium
       Partitioning Sediment Benchmarks (ESBs) for the Protection of Benthic Organisms: Metal
       Mixtures (Cadmium, Copper, Lead, Nickel, Silver and Zinc). Office of Research and
       Development. Washington, D.C. USEPA-600-R-02-011.
U.S. Environmental Protection Agency (USEPA). 2005c. Technologies and Costs Document for the Final
       Long Term 2 Enhanced Surface Water Treatment Rule and Final Stage 2 Disinfectants and
       Disinfection Byproducts Rule. EPA 815-R-05-013. U.S. EPA, Office of Water, Washington, DC.
       December 2005.
U.S. Environmental Protection Agency (USEPA). 2006a. Concrete Washout: National Pollutant
       Discharge Elimination System Factsheet. Available at
       http://cfpub.epa.gov/npdes/stormwater/menuofbmps/index.cfm?action=browse&Rbutton=detail&
       bmp=l 17&minmeasure=4
U.S. Environmental Protection Agency (USEPA). 2006b. Framework for Developing Suspended and
       Bedded Sediments (SABS) Water Quality Criteria. EPA-822-R-06-001.Office of Water, Office of
       Research & Development.
U.S. Environmental Protection Agency (USEPA). 2006c. "Safe Drinking Water Information System /
       Federal Version: About SDWIS/FED." Office of Ground Water and Drinking Water. Updated 22
       Feb. 2006. Available athttp://www.epa.gov/ogwdw/sdwisfed/about_fed.html.
U.S. Environmental Protection Agency (USEPA). 2006d. Wadeable Streams Assessment. EPA/841-B-
       06-002. U.S. Environmental Protection Agency, Office of Water, Washington, DC.
U.S. Environmental Protection Agency (USEPA). 2007a. Aquatic Life Ambient Freshwater Quality
       Criteria-Copper: 2007 Revision.  Office of Water. EPA-822-R-07-001. Feb. 2007.
U.S. Environmental Protection Agency (USEPA). 2007b. "Identification and Characterization of Typical
       Drinking Water Treatment Residuals." Peer Involvement for the Drinking Water Treatment
       Effluent Guidelines. September 2007.
U.S. Environmental Protection Agency (USEPA). 2007c. Master Survey Response Database for the
       Water Treatment Plant Questionnaire (EPA Form No. 2040-0269).
U.S. Environmental Protection Agency (USEPA). 2007d. National Recommended Water Quality Criteria:
       2002. Office of Water. Office of Science and Technology. Washington D.C. EPA-822-R-02-047.
U.S. Environmental Protection Agency (USEPA). 2007e. U.S. EPA Reach File 1 (RF1) for the
       Conterminous United States in BASINSRF1. Available at
       http://www.epa.gov/waterscience/BASINS/metadata/rfl .htm - Metadata Reference Information.
U.S. Environmental Protection Agency (USEPA). 2008a. Environmental Monitoring and Assessment
       Program: National Coastal Assessment (EMAP-NCA). Accessed October 1, 2008. Available at
       http://www.epa.gov/emap/nca/html/data/index.html.
November 2009                                                                             12-30

-------
Environmental Impact and Benefits Assessment for the C&D Category
U.S. Environmental Protection Agency (USEPA). 2008b. National Coastal Condition Report III.
       EPA/842-R-08-002. US Environmental Protection Agency, Office of Research and Development
       and Office of Water, Washington, DC.
U.S. Environmental Protection Agency (USEPA). 2008c. Nutrient Criteria Technical Guidance Manual.
       Wetlands. Office of Water. EPA-822-B-08-001. June 2008.
U.S. Environmental Protection Agency (USEPA). 2008d. Polycyclic Aromatic Hydrocarbons (PAHs).
       Factsheet.
U.S. Environmental Protection Agency (USEPA). 2008e. Welcome to STORET, EPA's largest
       computerized environmental data system. Available at http://www.epa.gov/storet/.
U.S. Environmental Protection Agency (USEPA). 2009a. ATTAINS. Available at
       http://www.epa.gov/waters/ir/index.html
U.S. Environmental Protection Agency (USEPA). 2009b. Development Document for Proposed Effluent
       Guidelines and Standards for the Construction and Development Category. EPA-821-R-08-007.
U.S. Environmental Protection Agency (USEPA). 2009c. Economic Analysis of Proposed Effluent
       Guidelines and Standards for the Construction & Development Industry. EPA-821-R-08-008.
U.S. Environmental Protection Agency (USEPA). 2009d. "Ecoregional Nutrient Criteria Documents for
       Rivers & Streams." Office of Water. Available at
       http://www.epa.gov/waterscience/criteria/nutrient/ecoregions/rivers/index.html
U.S. Environmental Protection Agency (USEPA). 2009e. National Water Quality Inventory:  Report to
       Congress: 2004 Reporting Cycle. EPA 841-R-08-001. Available at
       http://www.epa.gov/305b/2002report/report2002305b.pdf
U.S. Fish and Wildlife Service (USFWS). 1996. The Endangered Species Act of 1973. A Summary of the
       ESA and Implementation Activities. Updated October 1996. U.S. Fish and Wildlife Service.
       http://endangered.fws.gov/esasum.html. Accessed April 2, 2008.
U.S. Fish and Wildlife Service (USFWS). 1998. Letter to Gregg Cooke, U.S. Environmental  Protection
       Agency.
U.S. Fish and Wildlife Service (USFWS). 2000. Recovery Plan for Mobile River Basin Aquatic
       Ecosystem. Southeast Region. Available at http://ecos.fws.gov/docs/recovery_plan/001117.pdf
U.S. Fish and Wildlife Service (USFWS). 2004. Letter to Allen Ryan, Army Corps of Engineers
       (Biological opinion on BNSF railroad bridge construction impacts on threatened and endangered
       species).
U.S. Fish and Wildlife Service (USFWS). 2005. Barton Springs Salamander Recovery Plan.
U.S. Fish and Wildlife Service (USFWS). 2007. Listing a Species as Threatened or Endangered: Section
       4 of the Endangered Species Act. Available at
       http://www.fws.gov/endangered/factsheets/listing.pdf. Accessed April 2, 2008.
U.S. Fish and Wildlife Service (USFWS). 2008. The Endangered Species Act and What We Do.
       Available at http://www.fws.gov/endangered/whatwedo.html. Accessed April 2, 2008.
U.S. Fish and Wildlife Service (USFWS). 2009. Endangered Species Program: Species Information.
       Available at http://www.fws.gov/endangered/wildlife.html. Accessed November 2009.
November 2009                                                                             12-31

-------
Environmental Impact and Benefits Assessment for the C&D Category
U.S. Geological Survey (USGS). 2000. Comparability of suspended-sediment concentration and total
       suspended solids data. Water-Resources Investigations Report 2000-4191.
U.S. Geological Survey (USGS). 2001. USGS National Water Quality Assessment Program (NAWQA).
       Available at http://water.usgs.gov/nawqa/data/.
U.S. Geological Survey (USGS). 2003. Federal lands of the United States, available at
       http://nationalatlas.gov/mld/fedlanp.html. (Accessed June 5, 2008.)
U.S. Geological Survey (USGS). 2006a. HYDROlk Documentation, available at
       http://edc.usgs.gov/products/elevation/gtopo30/hydro/readme.html. Accessed June 5, 2008.
U.S. Geological Survey (USGS). 2006b. National Stream Water Quality Network (NASQAN) Published
       Data. Available at http://water.usgs.gov/nasqan/data/finaldata.html. (Accessed January 30, 2008.)
U.S. Geological Survey (USGS). 2007a. "Hydrologic Unit Codes (HUCs) Explained." Accessed
       November 5, 2008. Available at http://nas.er.usgs.gov/hucs.asp.
U.S. Geological Survey (USGS). 2007b. National Hydrography Dataset (NHD). Available at
       http://nhd.usgs.gov/data.html
U.S. Geological Survey (USGS). 2007c. Reservoir Sedimentation Information System (RESIS).
U.S. Geological Survey (USGS). 2008. National Water Information System - Web Interface. USGS
       Surface Water Data for the Nation Available at http://waterdata.usgs.gov/nwis/sw. Accessed June
       5,2008.
VanDolah, B. 2009. Personal Communication. Marine Resources Research Institute. Available at
       http://www.dm. sc. gov/marine/scecap/. March 11, 2009.
Van Metre, P.C., B.J .Mahler, M. Scoggins and P.A. Hamilton. 2006. Parking Lot Sealcoat: A Major
       Source of Poly cyclic Aromatic Hydrocarbons (PAHs) in Urban and Suburban Environments.  U.S.
       Geological Survey (USGS) Fact Sheet 2005-3147. January 2006.
Van Nieuwenhuyse, E.E., and J.D. LaPerriere. 1986. "Effects of placer gold mining on primary
       production in subarctic streams of Alaska. "As cited in Waters 1995.
Van Veld, P.A., D.J. Westbrook, B.R. Woodin, R.C. Hales, C.L. Smith, R.J. Huggett, and J.J. Stegeman.
       (1990). "Induced cytochrome P-450 in intestine and liver of spot (Leiostomus xanthurus} from a
       poly cyclic aromatic hydrocarbon contaminated environment." Aquatic Toxicology 17: 119-132 .
Vancouver, City of. 2007. Bulletin 2002-001-EV: Construction Site Wastes. Community Services Group,
       Office of the Chief Building Official.
VandenBerg, T.P., G.L. Poe and J.R. Powell. 2001. "Assessing the accuracy of benefits transfers:
       Evidence from a multi-site contingent valuation study of ground water quality." In: The Economic
       Value of Water Quality. Bergstrom, J.C., K.J. Boyle and G.L. Poe (eds.). Edward Elgar
       Publishers, Northampton, MA.
Vannote, R.L., G.W. Minshall, K.W. Cummings, J.R. Sedell and C.E. Gushing. 1980. "The River
       Continuum Concept.: Canadian Journal of Fisheries and Aquatic Sciences 37: 130-137.
Vaughan, W.J. 1986. "The RFF Water Quality Ladder." In: Mitchell, R.C. and R.T. Carson. The Use  of
       Contingent Valuation Data for Benefit/Cost  Analysis in Water Pollution Control, Final Report.
       Washington: Resources for the Future. Appendix B.
November 2009                                                                             12-32

-------
Environmental Impact and Benefits Assessment for the C&D Category
Vecchia, A.V. 2000. Water-quality trend analysis and sampling design for the Souris River,
        Saskatchewan, North Dakota, and Manitoba: U.S. Geological Survey Water- Resources
        Investigations Report 00-4019, 77 p. As cited in Schwarz et al. 2006.
Vermont Water Resources Board. Montpelier, VT. "Vermont Water Quality Standards." January 25,
        2006.
Vice, R.B., H.P. Guy, and G.E. Ferguson (USGS). 1969. Sediment Movement in an Area of Suburban
        Highway Construction, Scott Run Basin, Fairfax County, Virginia, 1961-64. Geological Survey
        Water-Supply Paper 1591-E. U.S. Government Printing Office.
Vincoli, J. 1997. Risk Management of Hazardous Materials. Vol. II. CRC Press. Boca Raton, FL.
Viscusi, K.W, J. Huber and J. Bell. 2008. "The Economic Value of Water Quality." Environmental and
        Resource Economics 41: 169 - 187
Vollenweider, R.A. 1976. Advances in defining critical loading levels for phosphorus lake eutrophication.
        Memorie dell 'Institute Italiano di Idrobiologia 33:53-83. As cited in Schwarz et al. 2006.
Waco Water Utility Services. 2009. Dissolved Air Floatation.  Available at
        http://www.wacowater.com/dafplant.html
Walling D.E. and K.J. Gregory. 1970. "The Measurement of the Effects of Building Construction on
        Drainage Basin Dynamics." Journal of Hydrology 11: 129-144.
Ward, S.M., and D.H. Appel. 1988. Suspended-Sediment Yields in the Coal River and Trace Fork Basins,
        West Virginia, 1975-84. Water-Resources Investigations Report 87-4171.
Wark, J.W. and F.J. Keller. 1963. Preliminary study of sediment sources and transport in the Potomac
        River basin. Interstate Commission on the Potomac River Basin, Technical Bulletin 1963-11.
Warner, R. and  Collins-Comargo, F. 2001. Erosion Prevention and Sediment Control Computer Modeling
        Project. Submitted to the Chattahoochee-Flint River Regional Development Center. Available at
        http://www.gaepd.org/Files_PDF/techguide/wpb/dirt2/full_model.pdf
State of Washington. 1997. Washington Administrative Code Chapter 173-201A Water Quality Standards
        for Surface Waters of the State of Washington. Available at
        http://www.epa.gov/waterscience/standards/wqslibrary/wa/wa 10 chapterl73-201a.pdf
Washington Department of Ecology.  Olympia, WA. "Water Quality Standards for Surface Waters of the
        State of Washington." Washington Administrative Code (WAC). Chapter 173-201A. November
        18, 1997.
Waters, T. F. 1995. Sediment in Streams: Sources, Biological Effects, and Control. Bethesda, MD;
        American Fisheries Society Monograph 7.
Weaver, T.M. and J.F. Fraley. 1993. "A method to measure emergence success of Westslope cutthroat
        trout fry from varying substrate compositions in a natural stream channel." North American
        Journal of Fisheries Management 13: 817-822.
Welle, P.G. 1986. Potential Economic Impacts of Acid Deposition: A Contingent Valuation Study of
        Minnesota. Dissertation, University of Wisconsin-Madison.
Welsh, H.H. and L.M. Ollivier. 1998. "Stream Amphibians as Indicators of Ecosystem Stress: A Case
        Study from California's Redwoods." Ecological Applications 8(4): 1118-1132.
November 2009                                                                              12-33

-------
Environmental Impact and Benefits Assessment for the C&D Category
Werner, E.  1983. "Effects of highways on karst springs: An example from Pocahontas County, West
       Virginia." In: Dougherty, P.H. (Ed.). Environmental Karst. Geospeleo Publications, Cincinnati,
       Ohio, pp. 3-13. As cited in Donaldson 2004.
Westrick, J.A. 2008. "Cyanobacterial toxin removal in drinking water treatment processes and
       recreational waters." In: Hudnell, H.K. (Ed.). Cyanobacterial Harmful Algal Blooms.  State of the
       Science and Research Needs. Advances in Experimental Medicine and  Biology, 619.
Wetzel, R. 2000. Limnology. Lake and River Ecosystems. 3rd edition. Springer-Verlag.
Wey, K.A.  1990. Social Welfare Analysis of Congestion and Water Quality of Great Salt Pond, Block
       Island, Rhode Island. Dissertation, University of Rhode Island.
Wheeler, A.P., A.E. Keiter Rosenberger and P.L. Angermeier. 2003. Potential Impacts of 1-73 on Stream
       Habitat and Biota,with Emphasis on the Federally Endangered Roanoke Logperch. Department of
       Fisheries and Wildlife services, Virginia Polytechnic Institute and State University.
White, C.S. 1976.  "Factors influencing natural water quality and changes resulting from land-use
       practices." Water, Air, and Soil Pollution 6: 53-69. As cited in Taylor and Roff 1986.
White, D.S. and J.R. Gammon. 1977. "The effect of suspended solids on macroinvertebrate drift in an
       Indiana creek. "As cited in Waters 1995.
Whitehead, J.C. and G.C. Blomquist. 1991a. "A Link Between Behavior, Information, and Existence
       Value."Leisure Sciences 13: 97 -109.
Whitehead, J.C. and G.C. Blomquist. 1991b. "Measuring contingent values for  wetlands: Effects of
       information about related environmental goods." Water Resources Research 27(10): 2523-2531.
Whitehead, J.C., and P.A. Groothuis. 1992. "Economic Benefits of Improved Water Quality: a case study
       of North Carolina's Tar-Pamlico River." Rivers 3:  170-178.
Whitehead, J.C., G.C. Blomquist, T.J. Hoban and W.B. Clifford. 1995. "Assessing the Validity and
       Reliability of Contingent Values: A Comparison of On-Site Users, Off-Site Users, and Non-
       users." Journal of Environmental Economics and Management 29: 238-251.
Whitehead, J.C., T. Hoban and W. Clifford. 2002. Landowners' Willingness to  Pay for Water Quality
       Improvements: Jointly Estimating Contingent Valuation and Behavior with Limited Information.
       White paper developed in part by U.S. EPA, NCDENR and the College of Agriculture and Life
       Sciences at NSCU.
Whitney, A.N. and J.E. Bailey. 1959. "Detrimental Effects of Highway Construction on a Montana
       Stream." Transactions of the American Fisheries Society 88: 72-73.
Whittington, D., G. Cassidy, D. Amaral, E. McClelland, H. Wang and C.  Poulos. 1994. The Economic
       Value of Improving the Environmental Quality of Galveston Bay. Department of Environmental
       Sciences and Engineering, University of North Carolina at Chapel Hill. GBNEP-38, 6/94.
Wilkinson,  S., Henderson, A., Chen, Y. and Sherman, B. (2004). SedNet User Guide. Client Report,
       CSIRO Land and Water; Canberra.
Wilkinson,  B.H. 2005. "Humans as geologic agents: a deep-time perspective." Geology 33(3): 161-164.
Wilkinson,  .B.H and B.J. McElroy. 2007. "The impact of humans on continental erosion and
       sedimentation." Geological  Society of America Bulletin 119(1-2): 140-156
November 2009                                                                              12-34

-------
Environmental Impact and Benefits Assessment for the C&D Category
Wolman, M.C., and A.P. Schick. 1967. "Effects of Construction on Fluvial Sediment, Urban and
       Suburban Areas of Maryland." Water Resources Research 3(2): 451-464.
Wong, Michael F. (USGS). 2005. Water Quality in the Halawa, Haiku, and Kaneohe Drainage Basins
       Before, During, and After H-3 Highway Construction, Oahu, Hawaii, 1983-99. Scientific
       Investigations Report 2004-5002.
Wood, P.J. and P.O. Armitage. 1997. "Biological Effects of Fine Sediment in the Lotic Environment."
       Environmental Management 21(2): 203-217.
Woodward, R.T. and Y.S. Wui. 2001. "The economic value of wetland services: ameta-analysis."
       Ecological Economics  37(2): 257-270.
World Health Organization (WHO). 1999. Toxic Cyanobacteria in Water: A guide to public health
       consequences, monitoring and management. Chorus, I and P. Bartram (eds.) E & FN Spon,
       London.
World Health Organization (WHO). 2003. Cyanobacterial toxins: microcystin-LR in drinking-water.
       Background document for development of WHO Guidelines for Drinking-water Quality. Geneva,
       Switzerland. (WHO/SDE/WSH/03.04/57)
Yao, K.M. (ASCE). 1975.  "Extended Plain Sedimentation." Journal of the Environmental Engineering
       Division 101(3): 413-423.
Yapo, P.O., H.V. Gupta and S. Sorooshian. 1996. 'Calibration of conceptual rainfall-runoff models -
       sensitivity to calibration data." Journal of Hydrology 181:  23-48. As cited in  Schwarz et al. 2006.
Yew, C.P. and P.B. Makowski. 1989. "Water Quality Monitoring of Highway Construction," Proceedings
       of the 1989 National Conference on Hydraulic Engineering, ASCE, New York, NY, pp. 157-162.
       As cited in Barrett et al. 1995b.
Yoder, C.O. 1997. Important concepts and elements of an adequate state watershed monitoring  and
       assessment program, ASIWPCA Standards and Monitoring Task Force: prepared for USEPA,
       Office of Water (Cooperative agreement CX 825484-01-0), 38 p. As cited in  Schwarz et al. 2006.
Yorke, T.H. and W.J. Davis.  1972. Sediment yields of urban construction sources, Montgomery County,
       Maryland. USGS Open-File Report, 39 p. As cited in Yorke and Herb 1978
Yorke, Thomas H., and William J. Herb (USGS). 1978. Effects of Urbanization on Streamflow and
       Sediment Transport in the Rock Creek and Anacostia River Basins, Montgomery County,
       Maryland, 1962-74. Geological Survey Professional Paper 1003.
Young, R.A., C.A. Onstad, D.D. Bosch and W.P. Anderson. 1995. AGNPS user's guide, version 5.00,
       1995. Morris, MN, Agricultural Research Service, U.S. Department of Agriculture. As cited in
       Schwarz et al. 2006.
Young, R.J. and G.L. Mackie. 1991. "Effect of oil pipeline construction on the benthic invertebrate
       community structure of Hodgson Creek, Northwest Territories." Canadian Journal of Zoology
       69:  2154-2160.
November 2009                                                                             12-35

-------
Environmental Impact and Benefits Assessment for the C&D Category
Appendix A - Status of Available  Data on Waterbody Assessments
This appendix presents the year of the data available from each state on waterbody assessments. These are
the data used in the National Water Quality Inventory Report to Congress (USEPA 2009e) and presented
inSections 2.6and 3.3.1.2.
 Table A-1: Year of Available Waterbody Assessment Data, as of September 17, 2009
           State Name
Assessed Waters Report Year
Impaired Waters Report Year
 Alabama
           2008
           2008
 Alaska
           2008
           2008
 American Samoa
           2008
           2008
 Arizona
           2008
           2004
 Arkansas
           2004
           2004
 California
           2004
           2006
 Colorado
           2008
           2008
 Connecticut
           2008
           2008
 Delaware
           2006
           2006
 District Of Columbia
           2008
           2008
 Florida
           2002
           2002
 Georgia
           2008
           2008
 Guam
           2008
           2008
 Hawaii
           2006
           2006
 Idaho
           2008
           2008
 Illinois
           2006
           2006
 Indiana
           2008
           2008
 Iowa
           2008
           2008
 Kansas
           2008
           2008
 Kentucky
           2008
           2008
 Louisiana
           2008
           2006
 Maine
           2008
           2008
 Maryland
           2002
           2006
 Massachusetts
           2006
           2006
 Michigan
           2008
           2008
 Minnesota
           2008
           2008
 Mississippi
           2008
           2008
 Missouri
           2008
           2008
 Montana
           2006
           2006
 N. Mariana Islands
           2008
           2008
 Nebraska
           2008
           2008
 Nevada
           2006
           2004
 New Hampshire
           2006
           2006
 New Jersey
           2006
           2006
 New Mexico
           2008
           2008
 New York
           2008
           2008
 North Carolina
           2006
           2006
 North Dakota
           2008
           2008
 Ohio
           2008
           2008
 Oklahoma
           2008
           2008
 Oregon
           2006
           2006
 Pennsylvania
           2006
           2004
 Puerto Rico
           2008
           2008
 Rhode Island
           2008
           2008
 South Carolina
           2008
           2008
 South Dakota
           2008
           2008
 Tennessee
           2008
           2008
November 2009
                                                                                             A-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Texas	2008	2008
 Utah	2006	2006
 Vermont	2008	2008
 Virgin Islands	2008	2008
 Virginia	2008	2008
 Washington	2008	2008
 West Virginia	2006	2006
 Wisconsin	2006	2006
 Wyoming                                         2008                               2008
November 2009                                                                                         A-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
Appendix B - List of Federally Threatened and Endangered Aquatic
Species Potentially Impacted by Sediment and Turbidity
 Table B-1: List of Federal Threatened and Endangered Aquatic Species Potentially
 Impacted by Sediment	
         Common name
Scientific Name
Threatened/Endangered
      Status
                                     Fish
Ala balik (trout)
Alabama cavefish
Alabama sturgeon
Amber darter
Apache trout
Arkansas River shiner
Ash Meadows Amargosa pupfish
Ash Meadows speckled dace
Asian bonytongue
Atlantic salmon
Ayumodoki (loach)
Bayou darter
Beautiful shiner
Beluga sturgeon
Big Bend gambusia
Big Spring spinedace
Blackside dace
Blue shiner
Bluemask (= jewel) Darter
Bonytail chub
Borax Lake chub
Boulder darter
Bull Trout
Cahaba shiner
Cape Fear shiner
Catfish
Cherokee darter
Chihuahua chub
Chinook salmon
Chum salmon
Cicek (minnow)
Clear Creek gambusia
Clover Valley speckled dace
Coho salmon
Colorado pikeminnow (=squawfish)
Comanche Springs pupfish
Conasauga logperch
Cui-ui
Delta smelt
Desert dace
Desert pupfish
Devils Hole pupfish
Devils River minnow
Duskytail darter
Etowah darter
Foskett speckled dace
Fountain darter
Salmo platycephalus
Speoplatyrhinus poulsoni
Scaphirhynchus suttkusi
Percina antesella
Oncorhynchus apache
Notropis girardi
Cyprinodon nevadensis mionectes
Rhinichthys osculus nevadensis
Scleropagesformosus
Salmo salar
Hymenophysa curia
Etheostoma rubrum
Cyprinella formosa
Huso huso
Gambusia gaigei
Lepidomeda mollispinis pratensis
Phoxinus cumberlandensis
Cyprinella caerulea
Etheostoma sp.
Gila elegans
Gila boraxobius
Etheostoma -wapiti
Salvelinus confluentus
Notropis cahabae
Notropis mekistocholas
Pangasius sanitwongsei
Etheostoma scotti
Gila nigrescens
Oncorhynchus (=Salmo) tshawytscha
Oncorhynchus (=Salmo) keta
Acanthorutilus handlirschi
Gambusia heterochir
Rhinichthys osculus oligoporus
Oncorhynchus (=Salmo) kisutch
Ptychocheilus lucius
Cyprinodon elegans
Percina jenkinsi
Chasmistes cujus
Hypomesus transpacificus
Eremichthys acros
Cyprinodon macularius
Cyprinodon diabolis
Dionda diaboli
Etheostoma percnurum
Etheostoma etowahae
Rhinichthys osculus ssp.
Etheostoma fonticola
Endangered
Endangered
Endangered
Endangered
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Threatened
Threatened
Endangered
Threatened
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Endangered
Endangered
Threatened
Threatened
Endangered
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Threatened
Endangered
Endangered
Threatened
Endangered
Endangered
Threatened
Endangered
November 2009
                                                                          B-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table B-1: List of Federal Threatened and Endangered Aquatic Species Potentially
 Impacted by Sediment
Common name
Gila chub
Gila topminnow (incl. Yaqui)
Gila trout
Goldline darter
Greenback cutthroat trout
Gulf sturgeon
Hiko White River springfish
Humpback chub
Hutton tui chub
Ikan temoleh (minnow)
Independence Valley speckled dace
June sucker
Kendall Warm Springs dace
Lahontan cutthroat trout
Leon Springs pupfish
Leopard darter
Little Colorado spinedace
Little Kern golden trout
Loach minnow
Lost River sucker
Maryland darter
Mexican blindcat (catfish)
Miyako tango (=Toyko bitterling)
Moapa dace
Modoc Sucker
Mohave tui chub
Nekogigi (catfish)
Neosho madtom
Niangua darter
North American green sturgeon
Okaloosa darter
Oregon chub
Owens pupfish
Owens tui chub
Ozark cavefish
Pahranagat roundtail chub
Pahrump poolfish
Paiute cutthroat trout
Palezone shiner
Pallid sturgeon
Pecos bluntnose shiner
Pecos gambusia
Pygmy madtom
Pygmy Sculpin
Railroad Valley springfish
Razorback sucker
Relict darter
Rio Grande silvery minnow
Roanoke logperch
San Marcos gambusia
Santa Ana sucker
Scioto madtom
Shortnose sturgeon
Shortnose Sucker
Slackwater darter
Slender chub
Scientific Name
Gila intermedia
Poeciliopsis occidentalis
Oncorhynchus gilae
Percina aurolineata
Oncorhynchus clarki stomias
Acipenser oxyrinchus desotoi
Crenichthys baileyi grandis
Gila cypha
Gila bicolor ssp.
Probarbus jullieni
Rhinichthys osculus lethoporus
Chasmistes liorus
Rhinichthys osculus thermalis
Oncorhynchus clarki henshawi
Cyprinodon bovinus
Percina pantherina
Lepidomeda vittata
Oncorhynchus aguabonita whitei
Tiaroga cobitis
Deltistes luxatus
Etheostoma sellare
Prietella phreatophila
Tanakia tanago
Moapa coriacea
Catostomus microps
Gila bicolor mohavensis
Coreobagrus ichikawai
Noturus placidus
Etheostoma nianguae
Acipenser medirostris
Etheostoma okaloosae
Oregonichthys crameri
Cyprinodon radiosus
Gila bicolor snyderi
Amblyopsis rosae
Gila robusta jordani
Empetrichthys lotos
Oncorhynchus clarki seleniris
Notropis albizonatus
Scaphirhynchus albus
Notropis simus pecosensis
Gambusia nobilis
Noturus stanauli
Coitus paulus (=pygmaeus)
Crenichthys nevadae
Xyrauchen texanus
Etheostoma chienense
Hybognathus amarus
Percina rex
Gambusia georgei
Catostomus santaanae
Noturus trautmani
Acipenser brevirostrum
Chasmistes brevirostris
Etheostoma boschungi
Erimystax cahni
Threatened/Endangered
Status
Endangered
Endangered
Threatened
Threatened
Threatened
Threatened
Endangered
Endangered
Threatened
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Threatened
Threatened
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Endangered
Threatened
Endangered
Endangered
Threatened
Endangered
Endangered
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Endangered
Endangered
Threatened
Threatened
November 2009
                                                                                    B-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table B-1:  List of Federal Threatened and Endangered Aquatic Species Potentially
 Impacted by Sediment
Common name
Smalltooth sawfish
Smoky madtom
Snail darter
Sockeye salmon
Sonora chub
Spikedace
Spotfin Chub
Steelhead
Thailand giant catfish
Tidewater goby
Topeka shiner
Totoaba (seatrout or weakfish)
Unarmored threespine stickleback
Vermilion darter
Virgin River Chub
Waccamaw silverside
Warm Springs pupfish
Warner sucker
Watercress darter
White River spinedace
White River springfish
White sturgeon
Woundfin
Yaqui catfish
Yaqui chub
Yellowfin madtom
Scientific Name
Pristis pectinata
Noturus baileyi
Percina tanasi
Oncorhynchus (=Salmo) nerka
Gila ditaenia
Meda fulgida
Erimonax monachus
Oncorhynchus (=Salmo) mykiss
Pangasianodon gigas
Eucyclogobius newbetryi
Notropis topeka (=tristis)
Cynoscion macdonaldi
Gasterosteus aculeatus williamsoni
Etheostoma chermocki
Gila seminuda (=robusta)
Menidia extensa
Cyprinodon nevadensis pectoralis
Catostomus warnerensis
Etheostoma nuchale
Lepidomeda albivallis
Crenichthys baileyi baileyi
Acipenser transmontanus
Plagopterus argentissimus
Ictalurus pricei
Gila purpurea
Noturus flavipinnis
Threatened/Endangered
Status
Endangered
Endangered
Threatened
Endangered
Threatened
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Threatened
                                       Amphibians
African viviparous toads
Arroyo (=arroyo southwestern) toad
Barton Springs salamander
California red-legged frog
California tiger Salamander
California tiger Salamander (Sonoma)
Cameroon toad
Cheat Mountain salamander
Chinese giant salamander
Chiricahua leopard frog
Desert slender salamander
Frosted Flatwoods salamander
Golden coqui
Goliath Frog
Guajon
Houston toad
Israel painted frog
Japanese giant salamander
Mississippi gopher Frog
Monte Verde golden toad
Mountain yellow-legged frog
Panamanian golden frog
Puerto Rican crested toad
Red Hills salamander
Reticulated flatwoods salamander
San Marcos salamander
Santa Cruz long-toed salamander
Shenandoah salamander
Sonora tiger Salamander
Nectophrynoides spp.
Bufo califomicus (=microscaphus)
Eurycea sosorum
Rana aurora draytonii
Ambystoma californiense
Ambystoma californiense
Bufo superciliaris
Plethodon nettingi
Andrias davidianus (=davidianus d.)
Rana chiricahuensis
Batrachoseps aridus
Ambystoma cingulatum
Eleutherodactylus jasperi
Conraua goliath
Eleutherodactylus cooki
Bufo houstonensis
Discoglossus nigriventer
Andrias japonicus (=davidianus j .)
Rana capita sevosa
Bufo periglenes
Rana muscosa
Atelopus varius zeteki
Peltophryne lemur
Phaeognathus hubrichti
Ambystoma bishopi
Eurycea nana
Ambystoma macrodactylum croceum
Plethodon shenandoah
Ambystoma tigrinum stebbinsi
Endangered
Endangered
Endangered
Threatened
Endangered
Endangered
Endangered
Threatened
Endangered
Threatened
Endangered
Threatened
Threatened
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Threatened
Endangered
Threatened
Endangered
Endangered
Endangered
November 2009
                                                                                     B-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table B-1:  List of Federal Threatened and Endangered Aquatic Species Potentially
 Impacted by Sediment
Common name
Stephen Island frog
Texas blind salamander
Wyoming Toad
Panamanian golden frog
Scientific Name
Leiopelma hamiltoni
Typhlomolge rathbuni
Bufo baxteri (=hemiophrys)
Atelopus varius zeteki
Threatened/Endangered
Status
Endangered
Endangered
Endangered
Endangered
                                        Mollusks
Cumberland elktoe
Dwarf wedgemussel
Appalachian elktoe
Fat three-ridge (mussel)
Ouachita rock pocketbook
Birdwing pearlymussel
Fanshell
Tampico pearlymussel
Dromedary pearlymussel
Chipola slabshell
Tar River spinymussel
Purple bankclimber (mussel)
Cumberlandian combshell
Oyster mussel
Curtis pearlymussel
Yellow blossom (pearlymussel)
Tan riffleshell
Upland combshell
Catspaw (=purple cat's paw pearlymussel)
White catspaw (pearlymussel)
Southern acomshell
Southern combshell
Green blossom (pearlymussel)
Northern riffleshell
Tubercled blossom (pearlymussel)
Turgid blossom (pearlymussel)
Shiny pigtoe
Finerayed pigtoe
Cracking pearlymussel
Pink mucket (pearlymussel)
Finelined pocketbook
Fh'ggins eye (pearlymussel)
Orangenacre mucket
Arkansas fatmucket
Speckled pocketbook
Shinyrayed pocketbook
Alabama lampmussel
Carolina heelsplitter
Scaleshell mussel
Louisiana pearlshell
Alabama moccasinshell
Coosa moccasinshell
Gulf moccasinshell
Ochlockonee moccasinshell
Nicklin's pearlymussel
Ring pink (mussel)
Littlewing pearlymussel
White wartyback (pearlymussel)
Orangefoot pimpleback (pearlymussel)
Clubshell
James spinymussel
Alasmidonta atropurpurea
Alasmidonta heterodon
Alasmidonta raveneliana
Amblema neislerii
Arkansia wheeleri
Conradilla caelata
Cyprogenia stegaria
Cyrtonaias tampicoensis tecomatensis
Dromus dramas
Elliptic chipolaensis
Elliptic steinstansana
Elliptoideus sloatianus
Epioblasma brevidens
Epioblasma capsaeformis
Epioblasma florentina curtisii
Epioblasma florentinaflorentina
Epioblasma florentina walkeri (=E. walkeri)
Epioblasma metastriata
Epioblasma obliquata obliquata
Epioblasma obliquata perobliqua
Epioblasma othcaloogensis
Epioblasma penita
Epioblasma torulosa gubernaculum
Epioblasma torulosa rangiana
Epioblasma torulosa torulosa
Epioblasma turgidula
Fusconaia cor
Fusconaia cuneolus
Hemistena lata
Lampsilis abrupta
Lampsilis altilis
Lampsilis higginsii
Lampsilis perovalis
Lampsilis powellii
Lampsilis streckeri
Lampsilis subangulata
Lampsilis virescens
Lasmigona decorata
Leptodea leptodon
Margaritifera hembeli
Medionidus acutissimus
Medionidus parvulus
Medionidus penicillatus
Medionidus simpsonianus
Megalonaias nicklineana
Obovaria retusa
Pegiasfabula
Plethobasus cicatricosus
Plethobasus cooperianus
Pleurobema clava
Pleurobema collina
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
November 2009
                                                                                     B-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table B-1: List of Federal Threatened and Endangered Aquatic Species Potentially
 Impacted by Sediment
Common name
Black clubshell
Southern clubshell
Dark pigtoe
Southern pigtoe
Cumberland pigtoe
Flat pigtoe
Ovate clubshell
Rough pigtoe
Oval pigtoe
Heavy pigtoe
Fat pocketbook
Alabama (=in£lated) heelsplitter
Triangular Kidney shell
Rough rabbitsfoot
Winged Mapleleaf
Cumberland monkeyface (pearlymussel)
Appalachian monkeyface (pearlymussel)
Stirrupshell
Pale lilliput (pearlymussel)
Purple bean
Cumberland bean (pearlymussel)
Scientific Name
Pleurobema curtum
Pleurobema decisum
Pleurobema furvum
Pleurobema georgianum
Pleurobema gibberum
Pleurobema marshalli
Pleurobema perovatum
Pleurobema plenum
Pleurobema pyriforme
Pleurobema taitianum
Potamilus capax
Potamilus inflatus
Ptychobranchus greenii
Quadrula cylindrica strigillata
Quadrula fragosa
Quadrula intermedia
Quadrula sparsa
Quadrula stapes
Toxolasma cylindrellus
Villosa perpurpurea
Villosa trabalis
Threatened/Endangered
Status
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
                                        Crustaceans
Madison Cave isopod
Conservancy fairy shrimp
Longhorn fairy shrimp
Vernal pool fairy shrimp
San Diego fairy shrimp
Cave crayfish
Cave crayfish
Illinois cave amphipod
Noel's Amphipod
Vernal pool tadpole shrimp
Lee County cave isopod
Nashville crayfish
Shasta crayfish
Squirrel Chimney Cave shrimp
Alabama cave shrimp
Kentucky cave shrimp
Kauai cave amphipod
Riverside fairy shrimp
Peck's cave amphipod
Hay's Spring amphipod
California freshwater shrimp
Socorro isopod
Antrolana lira
Branchinecta conservatio
Branchinecta longiantenna
Branchinecta lynchi
Branchinecta sandiegonensis
Cambarus aculabrum
Cambarus zophonastes
Gammarus acherondytes
Gammarus desperatus
Lepidurus packardi
Lirceus usdagalun
Orconectes shoupi
Pacifastacusfortis
Palaemonetes cummingi
Palaemonias alabamae
Palaemonias ganteri
Spelaeorchestia koloana
Streptocephalus woottoni
Stygobromus (=Stygonectes) pecki
Stygobromus hayi
Syncaris pacifica
Thermosphaeroma thermophilus
Threatened
Endangered
Endangered
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Threatened
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
Endangered
                                          Corals
Coral, elkhorn
Coral, staghom
Acropora palmata
Acropora cervicornis
Threatened
Threatened
 Source: USFWS 2009
November 2009
                                                                                      B-5

-------
Environmental Assessment and Benefits Document for the C&D Category
Appendix C - SPARROW Model Documentation
C.1  General Overview
SPARROW (SPAtially Referenced Regressions On Watershed attributes) is a watershed modeling
technique for estimating contaminant source contributions and transport in watersheds and surface waters.
SPARROW employs a statistically estimated nonlinear regression model with contaminant supply and
process components, including surface-water flow paths, non-conservative transport processes, and mass-
balance constraints. Regression equation parameters are estimated by correlating stream water-quality
records with GIS (Geographic Information System) data on pollutant sources (atmospheric, point,
nonpoint) and climatic and hydrogeologic properties (e.g., precipitation, topography, vegetation, soils,
water routing) that affect contaminant transport. The procedure for estimating SPARROW parameters
also provides measures of uncertainty in model coefficients and in water-quality predictions. The
SPARROW software is written in SAS (Statistical Analysis System) IML (Interactive Matrix Language).
SPARROW model infrastructure consists  of a detailed watershed - stream reach network to which all
monitoring stations and GIS data on watershed properties are spatially referenced. Digital elevation
models (DEMs) are used to delineate watershed boundaries and to identify overland flowpaths. The
spatially distributed model structure allows separate statistical estimation of land- and water-related
parameters that quantify rates of pollutant delivery from sources to streams,  and transport of pollutants to
downstream locations within the stream network (i.e., reaches, reservoirs and estuaries). Mechanistic
separation of terrestrial and aquatic features of large watersheds, and improved parameter estimation
techniques represent significant advances in water-quality modeling to objectively evaluate alternative
hypotheses about important contaminant sources and watershed properties that control transport over
large spatial scales.
SPARROW has been applied in the analysis of suspended sediment, surface-water nutrients, pesticides,
organic carbon, and fecal bacteria, and is potentially applicable to other measures of water quality. Recent
applications of SPARROW have provided estimates of nutrient sources and  long-term rates of nutrient
removal in surface waters (e.g., Smith et al. 1997; Alexander et al. 2000, 2001, 2002a, 2002b). The model
has demonstrated particular utility for quantifying long-distance transport and delivery of nutrients to
sensitive downstream locations (e.g., estuaries, reservoirs, drinking water intakes). The earliest version of
the SPARROW model was developed to describe contaminant transport in surface waters of the State of
New Jersey (Smith et al.  1994). Federal and State environmental managers are currently using
SPARROW to assess the sources of nutrient loadings in streams, including targeting of nutrient reduction
strategies in the Chesapeake Bay watershed (Preston and Brakebill 1999) and in waters of the State of
Kansas (Kansas Dept. Health and Environment 2004) as well as for developing TMDLs (Total Maximum
Daily Loads) in the Connecticut River Basin (NEIWPCC 2004). Other applications encompass New
England watersheds (Moore et al. 2004), New Zealand river basins (Alexander et al. 2002a; Elliott et al.
2005), North Carolina coastal watersheds (McMahon et al. 2003), and watersheds in Tennessee and
Kentucky (Hoos 2005). SPARROW models are currently under development for the Delaware River
Basin and are being planned for selected regions of the U.S. under the U.S. Geological Survey (USGS)
National Water-Quality Assessment (NAWQA) Program.
November 2009
                                                                                           C-1

-------
Environmental Assessment and Benefits Document for the C&D Category
C.2 Modeling Concept

C.2.1 Objectives of SPARROW Modeling
The broad objective of SPARROW modeling is to establish a mathematical relation between water-
quality measurements made at a network of monitoring stations and attributes of the watersheds
containing the stations. Once constructed, the model may be used to satisfy a variety of water-quality
information objectives. One common modeling objective is to describe past or present water-quality
conditions for a state or region on the basis of monitoring data. The underlying challenge is to extrapolate
from a sample of water-quality measurements made at a finite number of stream and river locations to an
area containing un-sampled locations. The usual limitations in attempting this are: (1) sparse sampling,
reflecting the high cost of monitoring; and/or (2) unrepresentative (nonrandom or targeted) sampling
undertaken to characterize water quality at specific locations, especially those suspected of having water-
quality problems. In the absence of an interpretive model such as SPARROW, a single monitoring design
cannot be optimal for both of these distinct objectives. Because the Federal Clean Water Act requires
state governments to collect and report both types of information, the distinction between the two types of
monitoring is of great practical importance. "Probabilistic" monitoring  has been promoted by the
U.S. Environmental Protection Agency (EPA) (Yoder 1997) to obtain a spatially unbiased, broad
overview of water-quality conditions in State waters.  An important limitation of this approach is that the
monitoring data alone do not provide detailed information on the geography of water-quality conditions
and give little understanding of the factors (sources and processes) that explain those conditions. Targeted
monitoring has been used extensively by the States to identify specific streams with water-quality
problems and to gage compliance with State water-quality regulatory standards and criteria. These data,
however, provide a spatially biased description of water-quality conditions in watersheds.  SPARROW is
effective in integrating samples from these different monitoring approaches to provide both a
geographically representative description of water-quality conditions as well as insight into the sources
and watershed processes that control water quality.
A second objective of SPARROW modeling is to identify and quantify the sources of pollution that give
rise to in-stream water-quality conditions. SPARROW distinguishes between  source categories, such as
point sources, atmospheric sources, and agriculture; and individual sources, defined as the rate of supply
of contaminant of a particular category originating in the watershed and draining to a specific stream
reach. The ability to develop quantitative information on pollution sources in SPARROW models stems
from the ability to trace, for each contaminant category, the predicted in-stream flux through a given
stream reach to the individual sources in each of the upstream reach watersheds contributing
contamination to that reach. Sources may be quantified either in  mass units or in terms of their percent
contribution to the total contaminant flux to the reach. An example application of SPARROW in
quantifying pollution sources is in TMDL analyses (e.g., McMahon and Roessler 2002). In general, the
Clean Water Act requires TMDL analyses for any stream reach in which the concentration of a
contaminant exceeds the applicable water-quality standard when all pollution discharge limits are met.
The ultimate objective of TMDL-related modeling is  to establish a hypothetical waste-load allocation for
all individual sources affecting the reach in question that would meet the standard. In theory, an infinite
number of hypothetical load allocations will satisfy the standard, and choosing the official allocation
requires comparing many possible solutions, utilizing simulation (see below) in search of a least-cost or
other optimum solution. Prior to conducting the hypothetical analyses, however, a great deal of
preliminary quantitative information on the actual or baseline relations  between watershed sources and in-
stream conditions is useful. The percentage contributions (shares) of individual point sources are of
November 2009                                                                                 C-2

-------
Environmental Assessment and Benefits Document for the C&D Category
particular interest because point sources are usually the only sources subject to direct regulation. The
share contributions of individual nonpoint sources are also of interest as a means of identifying the
important stakeholders to include in discussions of voluntary pollution reductions. This describes the
current application of SPARROW: to evaluate the effectiveness of a range of regulatory options for
controlling and mitigating sediments associated with construction activities, with explicit focus on the
impacts of such options on in-stream water quality as measured by Total Suspended Solids (TSS).
A third objective is water quality simulation. Simulation refers to the use of a calibrated model to predict
conditions on the basis of a set of altered, typically hypothetical model inputs. The ability to portray
counterfactual conditions for specified inputs is one of the most powerful justifications for models; there
are often no alternative methods for conducting controlled experiments on complex systems. Water-
quality model simulations can depict the in-stream effects of changes in contaminant sources associated
with alternative future pollution-control strategies, providing a critical  step in the analysis of the costs and
benefits associated with specific Best Management Practices (BMPs) or regulatory approaches.
Simulating alternative waste-load allocations in a TMDL analysis (see  above) is a prime example of a
water-quality simulation application. In the present application, SPARROW simulation is used to conduct
experiments focused on specific regulatory options (BMPs) for managing construction site sediment that
would not be feasible to conduct physically outside a limited number of test sites.
A fourth objective of SPARROW modeling is  hypothesis testing. One  common feature of the preceding
modeling objectives is that they each make use of predictions of the dependent variable by a calibrated
model. Another class of modeling objectives focuses on the calibration process  and its results directly.
The SPARROW estimation process explores the predictive value of a set of potential explanatory
variables and may also compare alternative mathematical forms. The selection of a final set of predictors
and mathematical form usually has the primary objective of maximizing the accuracy of model
predictions of the dependent variable, but an alternative modeling objective may be to test one or more
hypotheses about the nature and importance of factors and processes that may have influenced water
quality at the locations where samples were collected. Hypothesis tests are performed for each of the
model parameters estimated in the calibration,  and these serve as  indicators of an empirical relation
between the independent variables associated with each parameter and the dependent variable of the
model. Because the coefficients in a SPARROW model are specified to conform to physical processes,
and the potential explanatory variables are selected on the basis of theoretical or logical connections to the
dependent variable, a statistically strong parameter provides evidence of the physical relationship. For
example, if multiple categories of potential sources of the contaminant are being evaluated, the model
estimation process provides useful hypothesis tests on the importance of each category. "Importance" is
measured by the correlation between contaminant inputs from the source category and downstream
monitored loads of the contaminant. That correlation will tend to  be stronger when the mass contribution
of the category to the total mass of contaminant flowing past many of the monitoring stations is large, but
model estimation may also indicate a source category is important even when the mass contribution is
small, provided the amount contributed to stream loads by each unit of source is consistent from place to
place.
The importance of factors and processes potentially related to the transport of contaminants from sources
to stream channels  and within stream  channels may be tested through calibration of the "land to water"
and "in-stream decay" terms in the model. These terms  dictate the fraction of the contaminant mass that
completes the terrestrial and aquatic phases of transport within the watershed draining to each stream
reach. The land-to-water terms describe the land-surface characteristics that influence both overland and
subsurface transport from sources to stream channels. Similarly, the in-stream decay terms describe the
November 2009                                                                                  C-3

-------
Environmental Assessment and Benefits Document for the C&D Category
effects of channel characteristics on downstream transport. In sum, an important objective in building and
calibrating SPARROW models is to gain insight and to test hypotheses concerning the role of specific
contaminant sources and hydrologic processes in supplying and transporting contaminants in watersheds.

C.2.2 SPARROW Mass Balance Approach
SPARROW models constructed to date, like most water-quality models, are expressed in the form of a
mass balance. Such models describe the movement of mass in space and/or the change of mass in time.
The law of conservation of mass implies that certain basic accounting rules must apply to a mass balance,
water-quality model, such as: (1) the sum effluxes entering the confluence of two streams equals the flux
leaving the confluence; (2) the sum of the fluxes attributable to each contaminant source must equal total
flux; and (3) a doubling of all sources in the model results in an exact doubling of the predicted flux at
each location. Because the dependent variable in SPARROW models (the mass of contaminant that
passes a specific stream location per unit time) is, in mathematical terms, linearly related to all sources of
contaminant mass in the model, all accounting rules relating to the conservation of mass will apply. Mass
accounting in SPARROW models is also supported by the explicit spatial structure defined by the stream
network.
There are a number of advantages to the mass balance approach. Because of the linear relation between
flux and sources, there is an expectation that the estimation of flux over spatial scales smaller and larger
than that covered by the model's sample data will yield reasonably accurate results. The imposition of
mass balance greatly improves the  interpretability of model coefficients. For example, assuming mass
balance, the coefficient associated with the reach time-of-travel variable is interpreted as a first-order
decay rate and, because point-source loadings are delivered directly to the  stream network, a reasonable
null hypothesis for  coefficients associated with point sources is that they equal 1.0 (Smith et al. 1997).
The enhanced interpretability of the model coefficients  in turn facilitates the comparison of coefficient
estimates from the model with other estimates described in the literature. These comparisons have been
generally favorable, especially for the model components that quantify in-stream nitrogen decay rates
(Alexander et al. 2000, 2002a; McMahon et al. 2003), nutrient and suspended sediment removal rates in
reservoirs (Schwarz et al. 2001; Alexander et al. 2002a), and the nutrient export associated with various
land uses and pollutant sources (e.g., Alexander et al. 2001, 2002a, 2004; McMahon et al. 2003). Mass
balance provides a basis for flux accounting, whereby flux can be allocated to its various sources, both
spatially  and topically. For example, mass balance makes it possible to attribute nutrients discharged to
the Gulf of Mexico to specific sources within the Mississippi basin (Alexander et al. 2000), thereby
providing guidance in managing the reduction of this discharge. This approach is also sensitive to the
effects of natural and human-related processes that supply and remove contaminants from watersheds
over long periods. Thus, this approach de-emphasizes the quantification of short-term cycling and
transformation processes, which are often central to the functioning of many dynamic mechanistic
models, in favor of processes that have  a long-term impact on elemental budgets in aquatic ecosystems.

C.2.3 Time and Space Scales  of the Model
SPARROW models are structurally designed to explain spatial variability in the long-term mean-annual
or mean-seasonal flux of contaminants in streams. Spatial variability is modeled as a function of natural
and human-related properties of watersheds that influence the supply and transport of contaminants.
Estimates of the long-term mean flux are developed from water-quality and streamflow monitoring data
that are regularly collected at fixed locations on streams and rivers. The basic form of SPARROW models
is structured to describe the long-term, steady-state water-quality and flow conditions in streams.
November 2009                                                                                C-4

-------
Environmental Assessment and Benefits Document for the C&D Category
Contaminant source inputs are assumed to be in balance with the estimated sinks and measured in-stream
water-quality load such that there is conservation of mass among the model components. The principal
objective supported by this model structure is the quantification of the location and rates (and statistical
uncertainties) of the supply, transport, and fate of contaminants within the terrestrial and aquatic
ecosystems of watersheds. In the current specifications of SPARROW models, temporal variability in
contaminant loads, including intra- and inter-annual variations in water quality and streamflow reflected
in the monitoring data, are explicitly modeled and accounted for in a step prior to modeling spatial
variability in loads with SPARROW.
The computation of mean-annual or mean-seasonal flux (the SPARROW response variable) requires the
prior application of a water-quality flux-estimation model constructed on the basis of streamflow and
water-quality records from regularly monitored stream locations. The flux-estimation model explicitly
accounts for temporal variability in contaminant loads related to streamflow, season of the year, and
trends (either continuous or abrupt) with time. A base-year load estimate ensures that the stream water-
quality loads and the contaminant source data (which are commonly reported only periodically, e.g., the
U.S. Agricultural Census reports every five years) are contemporaneous. Therefore, the mean-annual
loads  used to calibrate  SPARROW models describe the mean load that would be expected to occur during
a particular base year under long-term mean streamflow conditions.
The steady-state mass-balance structure of the basic SPARROW model quantifies the long-term net
effects of biogeochemical and hydrologic processes on contaminant transport in terrestrial and aquatic
ecosystems. Modeling the effects of these processes is typically of greatest interest for non-conservative
chemical, physical, and biological properties of water, such as nutrients, pesticides, fecal coliform
bacteria, organic carbon, and suspended sediment. Many of these constituents are subject to chemical
transformations or degradation during transport and may be stored over short or long periods. In
SPARROW models that are estimated using long-term water-quality records, biogeochemical cycling and
storage processes that temporarily immobilize or remove contaminants from flow paths are generally in
steady state with those processes that mobilize or release contaminants from storage. Hence, the effects of
transformation and removal processes that operate on relatively short intra-annual time scales (e.g., daily,
weekly, seasonal), or over multi-year time scales that are less frequent than the period of model
estimation, are not likely to be detected as contaminant losses in the steady-state form of SPARROW
models. Limitations in our knowledge of temporal lags in chemical transport and their causes also create
uncertainties in the periods over which steady-state  conditions apply.  For example, the time scales for the
transport of sediment in streams reflect erosion, storage, and transport processes that operate from
seasonal to decadal or even longer (e.g., century) periods (Trimble and Crosson 2000).
Temporal changes in mean conditions also can be explicitly modeled  in SPARROW using the current
model structure and software. One approach is to estimate multiple steady-state models that explain
mean-annual or mean-seasonal stream contaminant loads for separate multi-year periods. Alternatively,
the current software permits the estimation of a single SPARROW model having time-dimensioned
dependent and independent variables - each assumed to pertain to different steady-state conditions. In this
approach, temporal changes in mean-annual stream contaminant flux  may be modeled as a function of
temporal changes in contaminant sources, land use, and climatic factors (e.g., precipitation, streamflow,
and temperature) over similar multi-year periods  as those used to estimate stream loads at each
monitoring station. The coefficients in this model can be time dependent or restricted to take common
values over the full period of the analysis. This type of model structure has considerable data demands
that require the development of historical data on contaminant sources and climatic/hydrologic variability
in the watersheds and reaches (including stream velocity, which is sensitive to average streamflow and
November 2009                                                                                 C-5

-------
Environmental Assessment and Benefits Document for the C&D Category
thus varies in response to changes in mean streamflow across different periods). Such a model allows
users to explicitly test for temporal changes in model parameters to determine whether changes are
evident in process rates or contaminant export coefficients with time. As an alternative to the explicit
estimation of a SPARROW model with time-dimensioned coefficients, existing SPARROW models can
be used to simulate changes in mean-annual stream water-quality loads as a function of changes in source
inputs. This approach assumes that the process rates and contaminant export coefficients of the model do
not change with time.

C.2.4 Accuracy and  Complexity of SPARROW Models
It is generally the case for SPARROW, as for any model with statistically estimated parameters, that
model accuracy (bias and precision) and complexity (number of statistically significant or sensitive
parameters) are dependent on the information content of the water-resources data used in model
calibration. Investigations of hydrologic models have demonstrated that both the quantity and quality of
calibration data define the information content and have important effects on parameter estimation and
precision (e.g., Gupta and Sorooshian 1985; Yapo et al. 1996). Increasing the quantity of data can
improve the precision provided the data give new, independent information about the values of the model
parameters. Data quality, as defined by Gupta and Sorooshian (1985), generally increases as the data
become more "representative" of the range of watershed properties that affect transport and the range of
conditions present in the sampled watersheds. An independent set of measurements are preferred for
estimating parameter values, such that the data reflect the most extreme combinations of watershed
conditions for the various properties. These general statistical guidelines have implications for the time
and space scales required to develop SPARROW data sets and accurate models. First, a sufficiently large
number of water-quality monitoring stations are required. In SPARROW models, the monitoring-station
loads serve as the response variable observations in the nonlinear spatial regressions. The number of
stations has a demonstrable effect on the statistical power of the regression - i.e., the capacity of the
model to detect the effect of an explanatory factor on stream loads. Models with more stations generally
have greater power, which typically supports more complex models  - i.e., models with a larger number of
statistically significant parameters and functional components. Second, the amount of spatial variability in
the stream monitoring data and explanatory factors should reflect as broad and representative a range of
watershed conditions as possible. The most complex SPARROW models typically have been developed
for regions that have relatively large spatial variability (greater than  one order of magnitude) in the
watershed properties that affect contaminant transport. Watershed properties that vary over a wide range
within a modeled region generally provide more information about the response of stream loads to
different levels of a given watershed property and are more likely to be statistically significant in
SPARROW models.
The spatial variability of a given variable is most readily increased in SPARROW models by expanding
the spatial domain of the model to include larger drainage areas. This, of course, has the effect of
increasing the number of monitoring stations, which also contributes to the potential for greater model
complexity. In selecting water-quality monitoring sites for modeling, it is especially important to obtain
sites that are located on a wide range of stream sizes (especially small streams) and are inclusive of
impoundments of varying size.  Sediment and nutrient removal rates  are highly responsive to the hydraulic
characteristics of streams and reservoirs, such as flow volume and velocity (Alexander et al. 2000,
2002a). Accounting for the wide variation in these properties can provide more accurate estimates of
nutrient removal and improve the separation of land and water effects on transport in the model.
Estimation errors are much larger for constituents that are most affected by high flows, such as suspended
November 2009                                                                                 C-6

-------
Environmental Assessment and Benefits Document for the C&D Category
sediment, total phosphorus, and fecal coliform bacteria. These constituents are generally more difficult to
measure and exhibit larger variability in concentrations in streams, which can produce less precise
estimates of the mean-annual flux. By contrast, dissolved substances, such as nitrate, sulfate, and total
dissolved solids exhibit less variability with flow, and their fluxes can generally be more accurately
estimated.

C.2.5 Comparison of SPARROW with Other Watershed Models
A wide variety of hydrologic and water-quality models have been used to describe contaminant sources
and transport in watersheds and surface waters. These models can be characterized on the basis of their
process complexity and temporal and spatial scales (e.g., Singh  1995). The level of complexity or process
detail represented by model descriptions of hydrologic and biogeochemical processes commonly varies
with the extent to which deterministic (mechanistic) and statistical/empirical methods are used to
describe and estimate these processes (Figure 12-1; Alexander et al. 2002b). All models reflect some
blend of these methods, but most place greater emphasis on one or the other type of model structure and
process specification.
  Figure 12-1: A Simple Continuum of Model Types Based on the Level of Statistical and
    Mechanistic Descriptions of Contaminant Sources and Biogeochemical Processes
                 Statistical                                  Deterministic


In general, purely statistical models tend to reflect more simplistic model constructs. These models have a
simple correlative mathematical structure and typically assume limited a priori knowledge of various
processes. Conventional versions of these models are expressed as simple linear (or log linear)
correlations of stream measurements with watershed sources and landscape properties (e.g., Peierls et al.
1991; Howarth et al. 1996; Caraco et al. 2003). The methods have the advantage of being readily applied
in large watersheds (often relying on generally available stream monitoring records) and can readily
quantify the errors in model parameters and predictions. Simple correlative approaches, however, offer
little mechanistic explanation of contaminant sources and transport. They generally lack spatial detail on
the distribution of sources and sinks within watersheds, rarely account for nonlinear interactions between
sources and loss processes, and do not impose mass-balance constraints on contaminant transport. The
most purely statistical approaches are found in artificial neural network and kriging techniques. These
models commonly provide an excellent fit to the observations, but provide little understanding of the
processes that affect contaminant transport.
November 2009                                                                               C-7

-------
Environmental Assessment and Benefits Document for the C&D Category
By contrast, mechanistic water-quality models have complex mass-balance structures that simulate
hydrologic and contaminant transport processes, often according to relatively fine temporal scales
[e.g., Soil Water Assessment Tool (SWAT), Srinivasan et al. 1993; Agricultural Nonpoint Source Model
(AGNPS), Young et al. 1995; Hydrologic Simulation Program-Fortran (HSPF), Bicknell et al. 2001]. The
components of these models frequently provide a highly detailed temporal description (e.g., daily, hourly)
of the response of stream contaminants to climatic variability, and the effects of coarser temporal
variations in land use and management activities are often superimposed on the more detailed climatic
variations. The mathematical descriptions of these responses are frequently based on a priori assumptions
about the dominant processes and their reaction rates. The complexity of mechanistic simulation models
creates intensive data and calibration requirements, which generally limits their application to relatively
small watersheds.  Because mechanistic water-quality models are frequently calibrated manually,  robust
measures of uncertainty in  model parameters and predictions cannot be quantified. Despite the common
use of mechanistic models, there are growing concerns about whether sufficient water-resources data and
knowledge of biogeochemical processes exist to reliably support the general use of such highly complex
descriptions of processes (Jakeman and Hornberger 1993; Beven 2002). Without sufficient data, there is
limited ability to apply formal parameter-estimation techniques required to quantify model uncertainties
and to identify unique models having sensitive and uncorrelated parameters.
By comparison to  other types of water-quality models,  SPARROW may be best characterized as a hybrid
process-based and statistical modeling approach for estimating pollutant sources and contaminant
transport in surface waters. The mechanistic mass transport components of SPARROW include surface-
water flow paths (channel time of travel, reservoirs), non-conservative transport processes  (i.e., first-order
in-stream and reservoir decay), and mass-balance constraints on model inputs (sources), losses (terrestrial
and aquatic losses/storage), and outputs (riverine nutrient export).  The statistical features of the model
involve the use of nonlinear parameter-estimation for spatially correlating stream water-quality records
with geographic data on pollutant sources (e.g., atmospheric, point-, nonpoint) and climatic and
hydrogeologic properties (e.g., precipitation, topography, vegetation, soils, water routing). Parameter
estimation ensures that the  calibrated model will not be more complex than can be supported by the data.
This provides an objective  statistical approach for evaluating alternative hypotheses about important
contaminant sources and controlling transport processes over large spatial scales in watersheds.
SPARROW has been shown to improve the accuracy and interpretability of model parameters and the
predictions of nutrient loadings and sources in streams  as compared with conventional statistical
modeling approaches (Smith et al. 1997; Alexander et al. 2000, 2002a, 2002b).
A number of inter-model comparisons have been previously conducted that compare the performance and
properties of SPARROW to those of other water-quality models. These include national comparisons
with SWAT and comparisons in the Chesapeake Bay watershed with HSPF (see Alexander et al.  2001).
Inter-model comparisons also have been conducted with statistical and quasi-deterministic models in
watersheds of the northeastern U.S.  (see Alexander et al. 2002b; Seitzinger et al. 2002). Finally,
evaluations of the  SPARROW technique also  are included as part  of a number of recent National
Research Council  (NRC) reports (2000, 200 la, 200 Ib,  2002a, 2002b). At least two of the NRC reports
(2000, 200la) have noted the advantages of statistical modeling approaches, such as SPARROW, for
general water-quality assessment and use in TMDL assessments.

C.2.6 Model Infrastructure
The parameters of SPARROW models are statistically  estimated with nonlinear regression techniques by
spatially correlating water-quality flux estimates at monitoring stations with watershed data on sources,
November 2009                                                                                 C-8

-------
Environmental Assessment and Benefits Document for the C&D Category
and landscape and surface-water properties that affect transport. The calibrated models are then used to
predict flux, total and disaggregated by contributing source, for stream reaches throughout a river
network. A flow diagram is provided in Figure 12-2 to illustrate the functional linkages between the
major spatial components of SPARROW models. Pre-processing steps are required to develop reach-level
information for the major components  of the SPARROW model infrastructure as shown in Figure 12-3.
Monitoring station flux estimation refers to the estimates of long-term flux used as the response variable
in the model. Flux estimates at monitoring stations are derived from station-specific models that relate
contaminant concentrations from individual water-quality samples to continuous records of streamflow
and time. The stream reach, inclusive of its  incremental contributing drainage basin, is the most elemental
spatial unit of the SPARROW model infrastructure. Stream reaches typically define the length of stream
channel that extends from one stream tributary junction to another. Explanatory data (e.g., climate,
topography, land use) are frequently compiled according to geographic units that are not coincident with
the drainage basin boundaries of river reaches. These data may be collected at different spatial scales and
according to spatial units that reflect political (e.g., counties) or other non-hydrologic features of the
landscape.
                                        Industrial / Municipal
                                          Point Sources
              SPARROW
         Model Components
                         Aquatic Transport
                            Streams
                           Reservoirs
                        In-Stream Flux Prediction
                        Calibration minimizes differences
                        between predicted and calculated
                          mean-annual loads at the
                             monitoring stations
                                                   Water-Quality and
                                                       Flow Data
                                                   Periodic measurements
                                                    a! monitoring stations
                   Monitoring
                  Station Flux
                   Estimation
Rating Curve Model
 of Pollutant Flux
 Station calibration to
   monitoring data
       Mean-Annual Pollutant
          Flux Estimation
                                                        Evaluation of Model
                                                     Parameters and Predictions
Source: Modified from Alexander et al. (2002).
November 2009
                                                                                                 C-9

-------
Environmental Assessment and Benefits Document for the C&D Category
 Figure 12-3: Location of 1,828 Water Quality Monitoring Stations Used in the SPARROW
	Sediment Model, in Relation to the Reach File 1 (RF1) Reach Network	
C.2.7 Use of Monitoring Data
The estimation of a SPARROW model requires estimates of long-term mean flux from a spatially
distributed set of monitoring stations within the study area, having sufficiently long periods of record. In
addition, it is necessary to include data from a diverse set of monitoring stations, inclusive of a wide range
of spatial scales and expressing considerable variation in predictor variable conditions. Often, monitoring
stations have different periods of record. Long-term variations in hydrologic conditions, combined with
long-term trends in water quality, imply mean fluxes computed over different periods may not be directly
comparable. If there are trends in water quality, the estimate of long-term mean water-quality  flux will
depend on the period or window through which the water-quality data are acquired. Additionally, water
quality is rarely measured with the frequency necessary to directly estimate  long-term mean flux, so that
indirect methods are needed to estimate flux from available measurements to account for hydrologic
conditions during periods when water-quality measurements are not made. The estimation procedure must
account for the fact that water-quality measurements are not always collected in a random manner,
implying the arithmetic average of flux is not a reliable estimate of mean flux. Rather, the measurements
of flux must be related in some way to other information, collected on a continuous basis, which exhibits
a close relation to flux, typically streamflow. The limitation, therefore, in constructing a mass-balance
model using long-term mean flux is that only water-quality measurements that can be associated with a
continuous record of streamflow will be suitable for model estimation.
Trends in water quality also create problems for the mass balance approach  in relating water quality to its
predictor variables, principally the source variables. Ideally, the source variables will consist of a time
series of estimates,  and these source variables would be included in SPARROW as long-term  averages in
the same way as flux estimates. However, source information is rarely available for multiple periods.
Even if source information is available over time, the period over which it is compiled will rarely match
the period of the flux information; for example, point-source and land-use data are infrequently available
due to the high cost of compiling this information. To address the problem of incompatible periods of
record, a SPARROW model is typically specified for a single base year; with all water-quality and source
November 2009
                                                                                           C-10

-------
Environmental Assessment and Benefits Document for the C&D Category
information assumed to pertain to a given point in time. The severest constraint on the period of the
analysis is imposed by source information that is available only for a single year. Consequently, the base
year is generally selected to be in the middle of the range of available years for this 'one-time'
information.

C.2.8 Model Specification for Monitoring Station Flux Estimation
The extrapolation of infrequently sampled water-quality data and removal of trend require the
specification of a model of flux. The model must relate infrequently measured data to variables that are
measured continuously over time; and to accommodate de-trending, the model must include a function of
time. These requirements suggest a model that relates infrequently measured concentration to the
variables streamflow and time. The inclusion of streamflow as an explanatory factor serves another
purpose in the analysis. Because flux is typically positively related to streamflow, water-quality sampling
is commonly biased towards high flow events. The inclusion of flow in the water-quality model
effectively conditions the estimation of flux so as to remove the effects of high-flow sampling bias. The
estimation of mean flux by a station-specific model need not account for all processes affecting flux
within a basin; as this task is assigned to the SPARROW model. All that is required is that the estimated
mean flux at a station be reflective of long-term average processes within the basin. Station-specific flux
models need not be structurally accurate; they need only be predictively accurate. A causal explanation of
flux is ultimately obtained through application of a SPARROW model, whereby variations in mean flux
conditions across stations, estimated either explicitly or implicitly, are correlated with variations in basin
attributes across space.
Cohn et al. (1992a) have suggested a simple seven-parameter model in which the logarithm of
contaminant concentration  (c) is related via a linear model to an intercept, the logarithm of flow, the
square of the logarithm of flow, decimal time T, decimal time squared, and a seasonal harmonic
consisting of two trigonometric terms - the sine and cosine of 2 times decimal time. Vecchia (2000) has
argued that there are important long-term  lags affecting the relation between water quality and flow. He
suggests a specification that relates the log of contaminant concentration to a set of compound flow terms
consisting of moving averages of flows, of various lengths, in addition to time trend terms. A
specification that generalizes both the seven-parameter and the lag flow models takes the form

                              ct = M(Qt)j3Q + h(Tt)j3T + Xtj3x +et                     (Eq. C-l)

Where Qt is a/>-elenient row vector consisting of current and lagged logarithms of flow, M(Q) is a vector
function that transforms the />-element logged flows into a Kq element row vector; fiQ  is a Kq element
vector consisting of coefficients associated with the transformed flow terms; h(Tt) is a KT element row
vector function of decimal time; fir is a KT element vector of coefficients associated with the transformed
decimal time terms; Xt is a Kx element row vector of other exogenous variables affecting water quality; fix
is a Kx element vector of coefficients associated with the other exogenous variables; and et is the normally
distributed error term, independent over time and uncorrelated with each of the predictor variables.

C.2.9 Tools for Flux Estimation
The water-quality models described above can be estimated using ordinary least squares if the water-
quality data do not include  any censored observations. For cases in which observations are censored,
Cohn et al. (1992b) suggest estimation via adjusted maximum likelihood. The standard maximum
likelihood method for type  I censored data (data for which the censoring threshold is known) is the Tobit
November 2009                                                                               C-11

-------
Environmental Assessment and Benefits Document for the C&D Category
model. The method of adjusted maximum likelihood combines the Tobit model with an adjustment to
correct for first-order bias in the coefficient estimates caused by estimation using a small sample. The
method of adjusted maximum likelihood is implemented within the USGS program Load Estimator
(LOADEST) 2000 (Runkle et al. 2004). In addition to estimates of the parameters and their covariance
matrix, the program uses retransformation methods to produce unbiased estimates of daily and annual
flux. A simple averaging of the daily or annual estimates over all days or years yields an estimate of long-
term mean flux. The estimation of the model used to detrend flow requires a maximum likelihood method
capable of correcting for serial correlation in the errors. This capability is included in the PARMA model
developed by Vecchia (2000), and in standard statistical packages such as SAS (SAS  1993). The more
recently developed program Fluxmaster (Schwarz et al. 2006) includes methods to estimate the time-
series flow model using maximum likelihood, detrend flow, and estimate the water-quality model via
adjusted maximum likelihood. The program also computes unbiased, detrended estimates of long-term
mean flux, and provides an estimate of the associated standard error. The exact methods used in
Fluxmaster differ from those used in LOADEST 2000, but they are a close approximation, and identical if
there are no censored observations. Fluxmaster (Schwarz et al. 2006) is used in the present study.

C.2.10 Guidance for Specifying Monitoring Station Flux Models
The ideal water-quality record has sufficient observations to reliably estimate the coefficients in the
model. Models that include both flow and time require a fairly long record, one that includes the base date
for detrending, and encompassing a wide range of hydrologic conditions. One consideration that places
limits on the length of the record is the need for flux estimates to be representative of base-year
conditions. This does not imply that the record must be representative of the hydrologic conditions in the
base year: the SPARROW model is estimated using mean estimates of water quality with the intention of
removing variations due to hydrologic conditions. Rather, long-term patterns in water quality should be
adequately captured by the model specification. This implies the period of record should not be so long as
to violate the assumption that water-quality trend is reasonably approximated by a simple linear function
of time. It is generally true that longer water-quality records display more complex patterns of trend. A
record that is too long runs the risk of misrepresenting the trend in water quality for any given year,
adding noise to the detrended flux estimate. For national analyses, Smith et al. (1997) and Alexander et al.
(2000) have based mean flux on water-quality data that span a 15 to 20 year period. Regional analyses by
Preston and Brakebill (1999)  and Moore et al. (2004) have used water-quality data that span 20 to
25 years. Shorter periods of about six years were used by Alexander et al. (2002a). Shorter records may
certainly be used to obtain an estimate of mean flux, although a shorter record will generally result in a
larger standard error in the mean flux estimate. For very short records, however, the specification of trend
becomes unreliable. It is  recommended that trend terms be excluded from the model if record spans less
than three years.
One of the principal determinants of accuracy in flux estimation is the number of observations input to the
water-quality model. Generally, it is recommended that an estimate of flux depend on no fewer than
15 uncensored water-quality observations. However, it may be necessary to deviate from this threshold in
situations in which a site is subject to high variability in streamflow. The final consideration for inclusion
of data for a station in the SPARROW analysis is that the standard error of the mean flux estimate not be
too large. SPARROW has the capability of weighting observations according to their standard errors;
therefore wholesale exclusion of high-error flux estimates is unnecessary. If, however, a high-error flux
estimate enters the model as an upstream source for some downstream flux observation  (the nested basin
arrangement) the potential for biasing estimates increases. High-error flux estimates also present
November 2009                                                                               C-12

-------
Environmental Assessment and Benefits Document for the C&D Category
problems in assessing the error of SPARROW predictions if the predictions are conditioned on observed,
upstream flux estimates. In national analyses, we have frequently excluded mean flux estimates that have
a standard error greater than about 20 to 30 percent.
C.2.11 Stream Network Topology
A vector- or raster-based digital representation of the stream and river network topology is the most
fundamental component of the spatial infrastructure that supports the SPARROW model (Figure 12-4).
Vector representations are based on point and line (arc) GIS features, whereas raster representations are
based on a cellular (areal) structure. In either instance, a stream reach network explicitly defines surface-
water flow paths that spatially connect contaminant sources and landscape features with water quality
observations at downstream monitoring stations. In a vector-reach topology, a stream reach represents the
length of stream channel that extends from one tributary junction to another. Reach nodes are point
features that are associated with the location of tributary junctions, hence reach boundaries. Reach nodes
will also occur at the locations where reaches overlay with the shorelines of impoundments (reservoirs,
lakes), and with stream water-quality monitoring sites. In a raster representation of streams, nodes may
also define various intermediate locations along the stream reach between tributary junctions. Whether
digitally represented in vector or raster form, the reach topology must define either a set of reach nodes or
raster cells that are linked hydrologically to indicate the direction of water flow. The node topology must
be defined according to an upstream (from-) and downstream (to-) node attribute table (Figure 12-4). This
tabulation of surface-water flow paths is required for the routing of water and contaminants through the
river network by SPARROW navigation software during estimation and application of the model. Thus,
the reach-node table defines the fundamental data infrastructure of the model.

  Figure 12-4: Schematic Illustrating a Vector Stream Reach Network with Node Topology
	and Water/Contaminant Reach-Node Routing Table	
    The reach-type indicator has possible values of "0" (Stream Reach), "1 "(Impoundment Reach), and "2"
                                  Outlet Reach for Impoundment
            13
                        12
                         12
                                                     Reach-Node Attribute Table
                               Reservoir
                              "shorelines
                                      Divergent
                                      Reach
                           Reach Node
                      'Stream Reach
Reach
Number
13
10
12
11
9
8
7
5
6
4
2
3
1
Up-
stream
Node
13
10
12
11
9
8
7
6
6
5
4
3
2
Down-
stream
Node
10
9
11
9
8
4
6
5
5
4
2
2
1
Diversion
Fraction
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.7
0.3
1.0
1.0
1.0
1.0
Reach -Type
Indicator
0
1
0
1
2
0
0
0
0
0
0
0
0
The SPARROW model structure supports distributary and diversion reaches in the stream network;
including braided channels or reaches where water is diverted to canals or other water bodies. The
November 2009
                                                                                           C-13

-------
Environmental Assessment and Benefits Document for the C&D Category
SPARROW model assumes that contaminants are diverted in proportion to volume of flow. Therefore, an
estimate is required for each reach of the diversion fraction - a measure of the fraction of streamflow that
is diverted in distributary reaches. A diversion fraction of 1.0 is assigned to reaches without water
diversions. In the example distributary reaches shown in Figure 12-4, the fraction of the flow in reach 7
that is diverted to the two downstream reaches is 0.7 for reach 5 and 0.3 for reach 6. The SPARROW
reach topology also supports the separate designation of impoundments (e.g., lakes, reservoirs) associated
with stream reaches. This designation is used in SPARROW to separately estimate the sediment and
contaminant attenuation in reservoirs and lakes. In Figure 12-4, reaches 9-11 are associated with a
reservoir. A "reach-type" indicator is used to identify reaches associated with impoundments separately
from conventional stream reaches. The outlet reach of an impoundment is coded separately from other
interior impoundment reaches to facilitate the pollutant attenuation calculations.
In addition to the reach properties listed in Figure 12-4, the reach length and an estimate of the mean-
annual streamflow of the reach is also required. Estimates of mean water velocity are required to estimate
in-stream contaminant attenuation as a function of the water  time of travel (alternatively, in-stream
attenuation can be estimated as a function the reach  length). Measures of the areal water load are needed
for impoundments for use in estimating the contaminant attenuation in these water bodies; the areal water
load is computed as the quotient of the outflow to surface area of the waterbody (see Alexander et al.
2002a). Digital representations of the drainage basin boundaries associated with river reaches are also
needed to  estimate the incremental and total drainage area of the reaches and to support the digital overlay
of drainage boundaries with polygonal boundaries (vector or raster) that define the locations of
contaminant sources (point and diffuse) and various landscape properties.
The assessment of pollutant loadings to coastal estuaries requires an expanded reach network that
includes shoreline features and the identification of reaches that terminate at estuaries. The node points of
shoreline reaches include the downstream nodes of the terminating reaches. Shoreline reaches are used to
define  coastal drainage areas - areas that discharge runoff directly to the estuary without transport
through a  stream reach. Because the discharge for a  shoreline reach does not accumulate from any
upstream location, the diversion fraction for these features is set to zero. Shoreline reaches also have no
stream attenuation so travel time is set to zero.
Spatial referencing is a critical step in SPARROW modeling that is necessary to construct the watershed
attribute data used as explanatory variables and to verify the  accuracy of the hydrologic connectivity of
the stream reaches, which governs routing and accumulation of mass in the model. Spatial referencing
entails the use of GIS techniques to digitally establish the geographic relation between stream reaches in
the river network and the various watershed attributes that are used to specify the model. Watershed
attribute data sets (e.g., topography, land use, climate) are frequently compiled and reported according to
geographic units that are not coincident with the drainage basin boundaries of river reaches; for example
at finer spatial scales or according to spatial units that reflect cultural (e.g., counties, states) or other non-
hydrologic features of the landscape. Watershed attributes may also be geographically located according
to precise  locations coded as latitude and longitude,  which must be digitally linked to a watershed or
nearby river reach. These may include the locations  of stream monitoring gages or municipal wastewater
treatment outfalls. Verification of automated GIS operations  is  often required to ensure accurate spatial
referencing; including, for example, comparisons of the river name of a gage with that of a river reach or
the comparison of the reported drainage area of a gage with that estimated for a river reach. Finally, the
SPARROW node and routing architecture also can fully support the modeling of contaminant transport
along "off-reach" (landscape) flow paths according to flow directions defined by landscape topography as
reflected, for example, in DEMs, although this capability is not currently implemented in the sediment
November 2009                                                                                 C-14

-------
Environmental Assessment and Benefits Document for the C&D Category
model. Such architecture facilitates the incorporation of high-resolution spatial data sets that delineate
sources and other explanatory variables at scales finer than an incremental drainage area for the reach.

C.2.12 Watershed Sources and Explanatory Variables
The explanatory variables evaluated in SPARROW models reflect current knowledge of natural and
human-related sources and the important physical, chemical, or biological properties of the terrestrial and
aquatic ecosystems that affect supply and transport of contaminants in watersheds, along with practical
considerations of availability. Point- and diffuse-source variables may include direct measures of the
introduction or supply of contaminant mass to the landscape and streams and reservoirs (e.g., municipal
and industrial wastewater discharge, fertilizer application). Alternatively, source variables may serve  as
surrogate indicators of the contaminant mass supplied by point and diffuse sources in watersheds, such as
land-use/land-cover data or census data on human and livestock populations. Watershed data on sources
and other properties (e.g., climate, topography, land-use) must be spatially referenced to the drainage
basins and stream reaches of the SPARROW river network. Figure 12-5 shows the relation between areas
of source polygons and the incremental drainage basin of a hypothetical stream reach. Estimates of the
diffuse sources associated with the drainage basins of stream reaches can be obtained using GIS
operations to digitally overlay the drainage basin boundaries of stream reaches with the polygonal areas
associated with the diffuse source data. Once quantitative measures of the overlap in watershed areas  are
obtained (Figure 12-5), estimates of the area-weighted sum of source characteristics (S ) for reach/ and
source type n can be calculated as:

                                                                                       (Eq.  C-2)
where P(j) is the set of all source-related polygons that intersect the incremental drainage polygon for
stream reach j, Snjk is the quantity of source-type n associated with polygon k, Ajik is the sub-area of reach
/s incremental drainage that intersects the source-related polygon k, ar\dA*k is the total area of the source-
related polygon k.

   Figure 12-5: Schematic Illustrating Digital Overlay of Stream Reach Drainage Area and
_
                      Polygonal Areas Associated with Diffuse Sources
                                 Area,
                                AreaML
                 Stream Reach)

Reach Drainage
    in Boundary
                                            Area//(=t
                                                                 Diffuse Source
If it is known that a particular contaminant source is associated with a specific land use (or some grouping
of land uses) - for example, fertilizer is associated with cultivated land - and if the spatial scale of land-
use information is similar to the scale at which watersheds are delineated, then the method described
above can be modified to obtain a more refined estimate of sources within a reach incremental drainage
November 2009
                                                                                            C-15

-------
Environmental Assessment and Benefits Document for the C&D Category
area. The enhanced method requires only that the area term Ajik in Equation C-2 be redefined to represent
the area of the associated land use that intersects the reach j incremental drainage and source polygon k,
and that the area term A "k be redefined to equal the total area of the associated land use within the source
polygon k.
Climatic and landscape properties that affect contaminant transport may include measures of water-
balance terms (e.g., solar radiation, precipitation, evaporation, evapotranspiration), soil characteristics
(e.g., permeability, moisture content), water-flow path properties (e.g., slope, hydraulic roughness,
topographic index), or management practices and activities (e.g., tile drains, conservation tillage, BMPs).
Estimates of various climatic and landscape characteristics or properties are often calculated as an area-
weighted mean estimate  (Z!>;) for stream reach j and landscape property /' according to:
                                    ZU=  IX
                                          (Eq. C-3)
where P(j) is the set of all land-characteristics polygons that intersect the incremental drainage polygon
for stream reach j, Zik is the landscape property / associated with polygon k, Ajik is the sub-area of reach
/s incremental drainage that intersects the landscape property's polygon k, and A*j is the total drainage
area of the reach watershed/. Figure 12-6 shows the relation between areas of landscape polygons and the
incremental drainage basin of a hypothetical stream reach that is described in Equation C-3. Note that the
area-weighted mean estimate defined in Equation C-3 differs from the area-weighted sum given by
Equation C-2 in that the area ratio terms sum to 1.0 in Equation C-3 but not in Equation C-2.

  Figure 12-6: Schematic Illustrating Digital Overlay of Stream Reach Drainage Area and
                  Polygonal Areas Associated with Landscape Properties

              Stream Reachy
               Area,
Area,
                                                          ^Reach Drainage
                                                           Basin Boundary
                                             Area/i/c=7
                          Landscape
                          Polygon k
The basic SPARROW models described are designed to model long-term mean contaminant loadings in
streams, implying explanatory variables in the model should be computed to reflect long-term conditions.
Variables that describe contaminant sources and landscape characteristics may be averaged over multiple
years, corresponding to the available period of record for water-quality monitoring data, or may reflect
the conditions during a specified base year used to estimate stream water-quality loadings, as in the
sediment study.
November 2009
                                                                                           C-16

-------
Environmental Assessment and Benefits Document for the C&D Category
C.3 Model Specification
The specification of a SPARROW model consists of identifying the explanatory variables and functional
forms for the associated processes the model is to include. The process is described here in generic terms,
supplemented by specific descriptions of the specification of the SPARROW sediment model.
C.3.1 Model Equation and Specification of Terms
Conceptually, the contaminant load or flux leaving a reach is the sum of two components. The first
component is the load generated within upstream reaches that is delivered to the reach via the stream
network. Losses of flux from the stream network may occur at points where flow is diverted.
Additionally, in moving through the reach, flux will generally be attenuated by stream or reservoir
processes. The second component consists of source flux that is generated within the reach's incremental
watershed and delivered to the stream network somewhere along the reach segment. A number of source-
dependent processes, in addition to stream attenuation processes, affect the  amount of source flux
reaching the stream network and transported to the reach's downstream outlet node. For flux originating
on the landscape, the processes affecting delivery to the stream network are called land-to-water delivery
processes, and may include both surface and sub-surface elements. A conceptual illustration of the
pertinent spatial relations is given  in Figure 12-7. A connecting reach is generally defined as a stream
segment that connects the confluence of two stream segments; a headwater reach is a reach that is defined
without an upstream confluence. Figure 12-7 shows five complete reaches, two of which are headwater
reaches, with one of the headwater reaches classified as a reservoir. Each reach is embedded within a

Figure 12-7: Conceptual Illustration of a Reach Network for Five Incremental Watersheds.
   Model Equation C-4 Describes the Supply and  Transport of Load within an Individual
                           Reach and its Incremental Watershed.
                     Stream
                     reach
                     segment
                    Downstream
                    monitoring
                    station, X
                                                                 Upstream
                                                                 monitoring
                                                                 station, Y
                                                              Reservoir
                                                                Reach
                                                                contributing
              area
Point source
Source: McMahon et al. (2003).
color-coded area representing the reach's incremental drainage - the area that drains directly to the reach
without passing through another reach. Because there are five reaches, there are five incremental drainage
areas. Land-to-water delivery processes determine the amount of contaminant generated within an
incremental drainage area, excluding contaminant generated directly on a reach, which is then delivered
to the area's corresponding reach. In the figure, two monitoring stations, X and Y, form the boundaries of
a nested basin, defined as all reaches above a monitored reach /', containing monitoring station X, but
exclusive of reaches above and including all upstream monitoring  stations - in this case the single
November 2009
                                                                                          C-17

-------
Environmental Assessment and Benefits Document for the C&D Category
monitoring station Y. In-stream attenuation processes associated with headwater reaches affect only the
transport of contaminants from their incremental drainage; in-stream attenuation processes for all non-
headwater reaches affect the transport of contaminants from their own incremental drainage and also from
the incremental drainage of all upstream reaches (including reaches in upstream nested basins).
This conceptual model can be formalized through a mathematical equation. Let F*] be the model-
estimated flux for contaminant leaving reach /'. This flux is related to the flux leaving adjacent reaches
upstream of reach /', denoted by Fj where/ indexes the set J(i) of adjacent reaches upstream of reach /',
plus additional flux that is generated within the incremental reach segment /'. In most cases, the set of
adjacent upstream reaches J(i) will consist of either two reaches, if reach /' is the result of a confluence, or
no reaches if reach /' is a headwater reach (see Figure 12-7). The functional relations determining reach /'
flux are given by:

                                                                                         (Eq.C-4)
The first summation term represents the amount of flux that leaves upstream reaches and is delivered
downstream to reach /', where F]- equals measured flux, FjM, if upstream reach/ is monitored or, if it is
not, is given by the model-estimated flux F*j. St is the fraction of upstream flux delivered to reach /'. If
there are no diversions, then St is set to 1. Otherwise, this fraction is defined by the fraction of streamflow
leaving upstream reaches that is delivered to reach /'. A(.) is the stream delivery function representing
attenuation processes acting on flux as it travels along the reach pathway. This function defines the
fraction of flux entering reach /' at the upstream node that is delivered to the reach's downstream node.
The factor is a function of measured stream and reservoir characteristics, denoted by the vectors Z5 and
2s, with corresponding coefficient vectors  9S and OR. If reach /' is a stream, then only the Z5 and 9$ terms
determine the value of A(.); conversely, if reach /' is a reservoir then the terms that determine A(.)  consist
of Z* and 9R
The second summation term represents the amount of flux introduced to the stream network at reach /'.
This term is composed of the flux originating in specific sources, indexed by n = 1,... JVS. Associated with
each source is a source variable, denoted Sn. Depending on the nature of the source, this variable could
represent the mass of the source available for transport to streams, or the area of a particular land  use. The
variable an is a source-specific coefficient. This coefficient retains the units that convert source variable
units to flux units. The function /)„(.) represents the land-to-water delivery factor. For sources associated
with the landscape, this function along with the source-specific coefficient determines the amount of
contaminant delivered to streams. The land-to-water delivery factor is a source-specific function of a
vector of delivery variables, denoted by Z,D, and an associated vector of coefficients 6D. For point sources
that are described by a measured (same units as flux) discharge of mass directly to the stream channel
(e.g., municipal wastewater effluent measured in kilograms year ), the delivery factor should be 1.0, with
no underlying factors acting as determinants, and the source-specific coefficient should be close to 1.0.
The last term in the  equation, the function A '(.), represents the fraction of flux originating in and delivered
to reach / that is transported to the reach's downstream node. This function is similar in form to the stream
delivery factor defined previously; however, the default assumption in SPARROW models is that if reach
/' is classified as a stream (as opposed to a reservoir reach), the contaminants introduced to the reach from
its incremental drainage area receive the square root of the reach's full in-stream delivery. This
assumption is consistent with the notion that contaminants are introduced to the reach network at the
midpoint of reach /' and thus experience only half of the reach's time of travel. For reaches classified as
November 2009                                                                                  C-18

-------
Environmental Assessment and Benefits Document for the C&D Category
reservoirs, the default assumption is that the contaminant receives the full attenuation defined for the
reach.
The nonlinear model structure in Equation C-4 contains several key features. The additive contaminant
source components and multiplicative land and water transport terms are conceptually consistent with the
physical mechanisms that explain the supply and movement of contaminants in watersheds. Total
modeled flux for a reach is shown to be decomposed into its individual sources. Because this same
decomposition is done for all upstream reaches, it is possible in this framework to perform flux
accounting, whereby total flux is attributed to its source components. All processes are spatially
referenced with respect to the stream network according to the reach in which they operate. This means,
for example, that a reservoir at reach /' affects the transport of all contaminants entering the reach network
upstream (but not downstream) of reach /'. The additive source components also provide a mathematical
structure in the model that preserves mass. This can be seen by noting that a doubling of each of the
source variables SH:i, along with a doubling of all upstream sources, as represented by a doubling of F]-
results in an exact doubling of modeled flux F*t. Finally, the modeled flux at any reach / is conditioned on
monitored fluxes entering the stream network anywhere upstream of reach /'. This approach to nested
basins serves to isolate errors introduced in any upstream basin from incremental errors that arise in a
downstream basin, making it defensible to treat nested basins as independent observations.

C.3.2 Contaminant Sources
The selection of potential contaminant source variables in SPARROW models initially depends upon a
user's particular knowledge of a watershed as well as inferences that can be derived from the research
literature about the major sources that contribute pollutants to watersheds Knowledge of the geography of
these sources, based on direct measures or surrogate indicators of the contaminant mass supplied to the
land surface and surface waters, is also critical to provide an effective use of the spatially distributed
structure in SPARROW. The compilation and availability of large digital spatial data sets has become
much more common (e.g., Brakebill and Preston 1999) and the advances in many types of digital
topographic and stream network data [e.g., National Hydrography Dataset (NHD), National Land Cover
Database (NLCD)] have made spatially distributed modeling more feasible. The inclusion of detailed
information in the SPARROW model on the geographic locations where contaminants are released in
watersheds has particular relevance to the potential policy and management applications of the model.
The source terms used in the  model can be generally classified as intensive and extensive measures of
contaminant mass. The former measures are typically descriptive of direct measures of pollutant mass,
such as fertilizer application, livestock waste, atmospheric deposition, or sewage-effluent loadings. In
these cases, the source-specific parameter (a) is expressed as a dimensionless coefficient that, together
with standardized expressions of the  land-to-water delivery factor, describes the proportion or fraction of
the source input that is delivered to streams. This fraction would be expected to be less than 1.0 but
greater than zero, reflecting the removal of contaminants in soils and ground water.
Extensive measures of contaminant mass also may be used in SPARROW models. These are surrogate
indicators of contaminant mass and include measures of watershed properties such as specific land-use
area and sewered population that are considered to be proportional to the actual mass  loadings generated
by a general type of a contaminant source. The empirical estimates of the source coefficients in the model
provide a quantitative measure of the proportion of the mass loading that is associated with a specified
source. If extensive measures are used, the associated model coefficients are expressed as the contaminant
mass generated per unit of the source type (e.g., kilograms kilometer"2 year"1; kilograms person"1 year"1). If
combined with the land-to-water delivery factor and expressed as a standardized source coefficient, the
November 2009                                                                                C-19

-------
Environmental Assessment and Benefits Document for the C&D Category
coefficient indicates the mean quantity of contaminant mass per unit of the surrogate source measure that
is delivered to streams. For land-use terms, the standardized coefficient gives what is frequently cited as
an export coefficient (Beaulac and Reckhow 1982; Johnes et al. 1996). Observed values of export
coefficients have been reported in the literature for various land-use/land-cover types, such as crop, urban,
and forested lands (Beaulac and Reckhow 1982); and often compare favorably with corresponding
estimates by SPARROW (e.g., Alexander et al. 2004). Other sources of guidance include sediment
erosion estimates generated from the Universal Soil Loss Equation (USLE) as part of the
U.S. Department of Agriculture  (USDA) National Resources Inventory (NRI) (Natural  Resources
Conservation Service 2005).

C.3.3 Landscape Variables
Landscape variables in SPARROW describe properties of the landscape that relate to climatic, natural- or
human-related terrestrial processes affecting contaminant transport. These include properties that are
assumed on some conceptual or empirical basis to control the rates of contaminant processing and
transport, and which are widely  available. The model structure allows tests of hypotheses about the
influence of specific features of the landscape on contaminant transport. Landscape variables may include
water-balance terms (e.g., precipitation, evapotranspiration) related to climate and vegetation, soil
properties (e.g., organic content, permeability, moisture content), topographic water flow-paths variables
(e.g., TOPMODEL overland flow, topographic index, and slope) (Beven and Kirkby 1979), or
management practices and activities, including tile drainage, conservation tillage practices, and BMPs
related to stream riparian properties. Particular types of land-use classes,  such as wetlands or impervious
cover, may also be potentially used to describe transport properties of the landscape.

The land-to-water delivery factor in Equation C-4, Z)n(Z!D;^D), is a source-specific function of a vector of
delivery variables, denoted by Z?, and an associated vector of coefficients 0D. In basic  SPARROW
models, the land-to-water delivery factor has been expressed in exponential functional form. For source n,
the fraction of contaminant mass generated in the incremental reach drainage area and delivered to the
reach (excluding source coefficient term) is estimated as
                              Dn(ztD- 9D} = expIf]®nmZm°0Dm \                      (Eq. C-5)
where Zm!D represents delivery variable m for the incremental drainage of reach /', 6Dm is its corresponding
coefficient, conm is an indicator variable that is 1.0 if delivery variable m affects source n and zero
otherwise, andMD is the number of delivery variables. Log-transformed delivery variables,  such as the
logarithm of soil permeability, may provide an improved fit to the data. Under the log-transformation, the
land-to-water coefficient is interpreted as the percent change in flux delivered to streams, derived from all
sources to which the land-to-water variable is applied, from a one-percent increase in the land-to-water
delivery variable. The full effect of landscape variables on the delivery of source n to streams is
determined by the product of the delivery factor in Equation C-5 and the source coefficient otn.  Therefore,
modification of delivery factor specification can be expected to change the mean of the delivery factor,
resulting in a corresponding counter-adjustment of the source coefficient otn. In order to improve the
interpretability of the source coefficient, and to provide some stability in its value across alternative
specifications of the land-to-water delivery factor, it is recommended that the delivery variables Zm® in
Equation C-5 be expressed as differences from their mean value over all reaches.
November 2009                                                                                 C-20

-------
Environmental Assessment and Benefits Document for the C&D Category
C.3.4 Stream Transport
Stream attenuation processes that act on contaminant flux as it travels along stream reaches are frequently
modeled as first-order reaction rate processes (Chapra 1997). A first-order decay process implies that the
rate of removal of the contaminant from the water column per unit of time is proportional to the
concentration or mass that is present in a given volume of water (a zero-order process corresponds to a
constant rate of removal per unit of time). In a first-order decay process, the fraction of contaminant
removed over a given stream distance is estimated as an exponential function of a first-order reaction rate
coefficient (in reciprocal time units) and the cumulative water time of travel over this distance. A reaction
rate is estimated on a volumetric basis, and therefore is expected to depend on properties of the water
column  that are proportional to water volume, such as streamflow and water-column depth (Stream Solute
Workshop 1990). Accordingly, in basic forms of the SPARROW model, the fraction of the contaminant
mass originating from the upstream node and transported along reach /' to its downstream node is
estimated as a function of the mean water time of travel (Tcs{, units of time) in reach /' and stream class c
defined  according to discrete intervals of mean streamflow or depth (in this case, Zf = {T/j}, c= I,... ,CS,
where T/t is nonzero only for the streamflow class corresponding to reach /'), and a stream-size dependent
loss rate coefficient (0Sc; units of time) such that:
                                     ,R ;0S,0R)= expj- f; 0SCTCS, }                     (Eq. C-6)
Mean water time of travel is estimated as the quotient of the reach channel length and mean water
velocity. The most accurate estimates of water velocity are obtained from time-of-travel dye studies
(Jobson 1996), but in some cases may be obtained from instantaneous measurements of water velocity
taken during flow gage site visits. Empirical geomorphic relations, which use regression methods to relate
time-of-travel measurements to channel and basin properties (e.g., streamflow, slope), also are available
for regions of the U.S. and other countries (e.g., Jobson 1996; USEPA 1996b; Jowett 1998; Alexander et
al. 1999) and can be used to estimate the time-of-travel of stream reaches for a given river network.
Alternatively, channel length may be used in Equation C-6 where water velocity estimates are not
available. This assumes that the water time of travel is proportional to the channel length, and estimates of
the removal rate are expressed as reciprocal length (e.g., see Alexander et al. 2002a; McMahon et al.
2003).

C.3.5 Reservoir and Lake Transport
Attenuation processes that act on contaminant mass as it travels through a lake or reservoir are often
modeled as a net removal process, with the loss coefficient expressed as either a first-order reaction rate
or a mass-transfer coefficient (also referred to as an apparent settling velocity) (Chapra 1997). Both of
these loss expressions have been used in empirical mass-balance lake models for phosphorus (e.g.,
Vollenweider 1976; Reckhow and Chapra 1983) and nitrogen (e.g., Kelly et al. 1987; Molot and Dillon
1993), although the use of the mass-transfer rate is more common. These mass-balance models typically
assume steady-state and uniformly mixed conditions in the waterbody. The reaction rate is expressed in
reciprocal time units and its estimation and use are dependent on knowledge of the water depth and
surface area of the waterbody (volume). The apparent settling velocity is expressed in units of length per
time and is estimated as a function of the ratio of the reservoir outflow and the surface area of the
reservoir sediments (assumed equal to the  surface area of the waterbody); this ratio, denoted q,R for reach
/', is termed the areal hydraulic load and is  a measure of the water displacement or velocity in the
November 2009                                                                                C-21

-------
Environmental Assessment and Benefits Document for the C&D Category
reservoir. The term "apparent" indicates that the settling velocity measures the net effect of various
processes that remove the contaminant from the water column and deliver it to the sediments, and
processes that may add contaminant to the water column. The function that relates the areal hydraulic
load to the fraction of flux attenuated in  a reservoir is

                               A(Z*,Z?;QS,QR) =	lr-^-                       (Eq.C-7)
                                                   1 + ^ofa)

where ORO is a parameter, to be estimated in the model, representing the apparent settling velocity.

C.4 Model  Estimation
The SPARROW model equation, given in Equation C-4, is a nonlinear function of its parameters, and the
model must be estimated using nonlinear techniques. The model errors are  assumed to be independent
across observations with zero mean, while the variance of each observation may be observation specific.
SPARROW utilizes a nonlinear weighted least squares (NWLS) estimation method, which does not
assume the precise distribution of the residuals. Several algorithms exist for obtaining NWLS estimates,
and the Levenberg-Marquardt Least-Squares method, implemented in SAS, is used to estimate the
SPARROW model (SAS 1999). Unlike linear models, the statistical properties of the estimated
parameters for nonlinear models are not precisely known in finite samples. For this reason, much of the
theory  of nonlinear estimation has focused on characterizing asymptotic properties - the statistical
properties of the coefficient estimates as the sample size goes to infinity. Emphasis in the following
sections is on interpretation of estimated model parameters.

C.4.1 Evaluation of Model Parameters
The objective of parameter evaluation in SPARROW modeling is to determine whether a converged
model provides statistically sound and physically interpretable coefficient values. The first objective
entails  the appraisal of model parameters for statistical significance and the quantification of uncertainty.
This provides important information for identifying unique model specifications (parameters and values
for which the model predictions are sensitive) and determining the level of model complexity (number
and types of explanatory variables and model functions) that can be empirically supported by the stream
monitoring data. The  emphasis on parameter estimation in SPARROW models reflects the objective of
identifying the important contaminant sources and factors affecting mean-annual contaminant transport
over large spatial scales in soils and in ground and surface waters. The key parameter statistics include the
estimated mean coefficient values, estimated variance of these coefficient estimators, and measures of
statistical significance based on the t statistics (ratio of the coefficient value to its standard error). These
statistics are biased in finite samples but consistent as sample size goes to infinity; the t statistics are
asymptotically distributed as standard normal. The/"-values are based on a two-tailed probability from a
Student's t distribution. The/"-values can be used to identify statistically significant model coefficients -
those that are statistically distinguishable from zero - and can be used to refine the parameter set to
identify parsimonious SPARROW models.
The second complementary objective is the evaluation of the parameters for physical interpretability. This
objective entails the evaluation of the sign and magnitude of model coefficients to test hypotheses about
the importance of different contaminant sources and the hydrologic and biogeochemical processes that are
represented by the explanatory variables of the model. SPARROW model parameters reflect the net
November 2009                                                                                C-22

-------
Environmental Assessment and Benefits Document for the C&D Category
effects over large spatial scales of an aggregate set of hydrologic and biogeochemical processes and
human-related activities. The interpretability of the parameters and their relation to specific processes is
enhanced in SPARROW by the use of a mass balance, mechanistic structure that explicitly separates the
terrestrial and aquatic properties of watersheds and accounts for nonlinear interactions among watershed
properties, together with an emphasis on the statistical estimation of parameter values. The sign of
SPARROW model coefficients can be evaluated to determine the direction of influence of any
explanatory variable on in-stream estimates of mean-annual flux.  The direction of influence is assessed
for consistency with the anticipated response based on available theoretical or empirical information
about processes related to individual explanatory factors. For example, a negative  sign on the soil
permeability coefficient indicates that total loads in streams are inversely related to permeability - i.e., in-
stream loads of contaminants are generally lower in watersheds with highly permeable soils. The sign of
the coefficient is also important in estimating physically meaningful contaminant source terms in
SPARROW, which are generally expected to  contribute positive contaminant mass to the watershed
system.
Coefficients associated with source inputs expressed in areal units describe the mass per unit area
delivered to streams from these land areas. These areal expressions of contaminant transport (export) can
be directly compared with ranges of export coefficients reported in the  literature (e.g., Beaulac and
Reckhow 1982). Estimated SPARROW coefficients associated with specific land uses generally compare
favorably with published export coefficients. Other source coefficients  that are expressed in
dimensionless units provide a measure of the fraction of the contaminant that is delivered from each
source to streams, rivers, and reservoirs. These coefficients can be evaluated to determine how reasonably
they reflect the net mean rates of contaminant removal by a source as part of the delivery to aquatic
systems.

C.5 Model Prediction
SPARROW output contains prediction results paired with measures of accuracy. A number of technical
issues can arise in the derivation of these statistics; most of these caused by the nonlinear nature of the
SPARROW model. The prediction equation is similar to the calibration Equation C-4. Because
predictions are generated for specific sources, however, it is necessary to decompose flux  into source-
specific components. Let F*in denote the reach / model-estimated flux associated with source n, and let
Fj:H be the source-^ flux from upstream reach/. If reach7 is monitored (that is,/ El, where / is the set of
monitored reaches), and predictions are conditioned on measured flux, then F7-_n is apportioned from the
measured flux F}M according to
                                                                                        (Eq.C-8)
November 2009                                                                                C-23

-------
Environmental Assessment and Benefits Document for the C&D Category
Otherwise, if reach7 is not a monitored reach (that is, j &.I) or predictions are not conditioned on
measurements, then F7-_n is set equal to F*j>n. The equation defining the model-estimated flux for source n,
F*in, is given by
          F\n =
(Eq.C-9)
November 2009
                                                                                              C-24

-------
Environmental Impact and Benefits Assessment for the C&D Category
Appendix D - A Preliminary SPARROW Model of Suspended Sediment
for the Coterminous United States (Schwarz 2008a)
November 2009
                                                                  D-1

-------
    uses
science fora changing world
A                               of
the
Open-File Report 2008-1205
U.S. Department of the Interior
U.S.       Sureef

-------
A                                               of
By Gregory E. Schwarz
National Water-Quality Assessment Program
Open-File Report 2008-1205
U.S.           of the Interior
U.S.          Sureef

-------
U.S.
DIRK KEMPTHORNE, Secretary

U.S.
Mark D. Myers, Director
U.S. Geological Survey, Reston, Virginia:  2008
For product and ordering information:
World Wide Web: http://www.usgs.gov/pubprod
Telephone:  1-888-ASK-USGS

For more information on the USGS—the Federal source for science about the Earth, its natural and living resources,
natural hazards, and the environment:
World Wide Web: http://www.usgs.gov
Telephone:  1-888-ASK--USGS
Suggested citation:
Schwarz, G.E., 2008, A Preliminary SPARROW model of suspended sediment for the conterminous United States: U.S.
Geological Survey Open-File Report 2008-1205, 8 p., available only online at http://pubs.usgs.gov/of/2008/1205.
Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the
U.S. Government.

Although this report is in the public domain, permission must be secured from the individual copyright owners to
reproduce any copyrighted materials contained within this report.

-------
                                                                                                            Hi

Abstract	1
Introduction	1
Sediment Model Data Sources	1
Preliminary Model Estimation Results	3
Model Simulation	5
Model Limitations	6
References Cited	6

1. Location of 1,828 water-quality monitoring stations used in the SPARROW sediment model, in
        relation to the Reach File 1 (RF1) stream  network	2
Tables
1. Preliminary estimation results for the SPARROW suspended sediment model	4

-------
A                                                            of
By Gregory E. Schwarz
Abstract

    This report describes the results of a preliminary Spatially
Referenced Regression on Watershed attributes (SPARROW)
model of suspended sediment for the conterminous United
States. The analysis is based on flux estimates compiled from
more than  1,800 long-term monitoring stations operated by the
U.S. Geological Survey (LISGS) during the period 1975-2007.
The SPARROW model is structured on the Reach File 1 (RF1)
stream network, consisting of approximately 62,000 reach seg-
ments. The reach network has been modified to include more
than 4,000 reservoirs, an important landscape feature affect-
ing the delivery of suspended sediment. The model identifies
six sources of sediment, including the stream channel and
five classes of land use: urban, forested, Federal nonforesled,
agricultural and other, noninundatcd land. The delivery of
sediment from landform sources to RF1 streams is mediated
by soil permeability, erodibility, slope, and rainfall; stream-
flow is found to affect the amount of sediment mobilized from
the stream channel. The results show agricultural land and
the stream channel to be major sources of sediment flux. Per
unit area, Federal nonforested and urban lands are the largest
landform sediment sources. Reservoirs are identified as major
sites for sediment attenuation. This report includes a descrip-
tion for how the model results can be used to assess changes
in instream sediment flux and concentration resulting from
proposed changes in (lie regulation of sediment discharge from
construction sites.
Introduction

    This report describes the results of a preliminary Spatially
Referenced Regression on Watershed attributes (SPARROW)
model of suspended sediment for the conterminous United
States.
    The spatial framework of the SPARROW sediment model
is the vector-based l:500,000-scale River Reach File (RF)
1 hydrography, originally developed by the U.S. EPA (U.S.
Environmental Protection Agency, 1996) and subsequently
enhanced to include areal hydraulic load information for
selected reservoirs (Ruddy and Hitt, 1990), shoreline reaches,
and reach catchment areas derived from the U.S. Geological
Survey (USGS) HYDRO Ik Digital Elevation Model (DEM)
(U.S. Geological Survey, 2006a). The enhanced network
(Nolan and others, 2002), consisting of 62,776 reach seg-
ments, including shoreline reaches, 61,214 delineated reach
catchments, and 2,171 individual reservoirs,  has been used to
support numerous national SPARROW modeling efforts for
the conterminous United States (for example, Alexander and
others, 2000). The RF1. reach network for the current SPAR-
ROW sediment model was further enhanced  by the inclusion
of area! hydraulic load information for approximately 2,000
additional  large reservoirs (reservoir storage  greater than 500
acre-feel), identified from the National Inventory of Dams
(NID) (U.S. Army Corps of Engineers, 2006) and linked to the
RF1 network according to dam geographic coordinates, river
name, and drainage area.
     Included in the original RF1 network are reach estimates
of mean strcamflow and mean velocity, the latter being con-
verted to reach time of travel using the RF1 measure of reach
length. The attributes of mean streamflow and mean velocity
are used to assess various sediment mobilization and attenu-
ation processes associated with the stream channel. Because
catchment areas used to derive the original RF1 estimates of
mean streamflow are not compatible with catchments included
in the enhanced RF1 network, an alternative  measure of mean
streamflow was used to compute flow-weighted sediment con-
centration. The alternative measure is  based on an interpola-
tion of USGS streamgage estimates from the 1975-2006 water
years (WYs), with extrapolation of slreamflow upstream of
gages based on runoff estimated at downstream or neighbor-
ing stations and apportioned to the land surface according  to
the enhanced RF1 catchments (David  Wolock,  U.S. Geologi-
cal. Survey, 2008, written commun.). The available period for
these data is 1 year less than the period for this study, and a
water year consists of the period October 1st of the previous
calendar year through September 30th of the enumerated water
year.
     The dependent variable in the SPARROW sediment
model is given by long-term mean sediment flux. Long-term

-------
     A Preliminary SPARROW Model of Suspended Sediment for the Conterminous United States
mean sediment flux is estimated using the maximum likeli-
hood approach developed by Cohn (2005), as implemented in
the Fluxmaster program (Schwarz and others, 2006). Instream
sediment concentrations and stream discharge measurements
over the WY period 1975-2007 have been obtained from the
National Stream Quality Accounting Network (NASQAN)
(Alexander and others, 1996; U.S. Geological Survey, 2006b),
the USGS National Water-Quality Assessment (NAWQA)
Program (Mueller and Spahr, 2005), and the USGS National
Water Information System (NWIS) (U.S. Geological Survey,
2008), a database encompassing USGS water-quality moni-
toring stations as well as water-quality monitoring activities
done in cooperation with State governments. Sampling for
suspended sediment (USGS water-quality parameter 80154)
is typically done periodically, not daily. Sample data are
weighted by channel cross-sectional flow geometry (depth,
width) and correlated with stream discharge at the time of
sampling. A linear regression model is estimated that relates
log-transformed instantaneous suspended-sediment concentra-
tion to log-transformed mean daily streamflow, which is mea-
sured continuously via water-surface height (stage) coupled
with a previously estimated relation between surface height
and instantaneous  flow. Included in the regression are the sine
and cosine of decimal time to capture a seasonal signal, and
a linear time trend to be used for detrending flux.  To support
the detrending of flux, a companion model of daily streamflow
is estimated for each water-quality station. The streamflow
model relates the logarithm of daily streamflow to a second-
order harmonic of the sine and cosine of decimal time, and
a linear time trend term. To account for serial correlation in
the daily values, the model is estimated using time-series
methods that assume a 30-day autoregression in the residuals.
The water-quality and streamflow models are used to simu-
late daily flux, with both water-quality and streamflow trends
removed, for all days within the 33-year period WY 1975-
2007 for which a daily streamflow value is available for every
day in the same WY. Thus, if streamflow is not available for
any day within a given WY, no simulated water-quality flux is
computed for any day in that WY. The simulated estimates of
flux, detrended to the base year 1992, for all complete WYs
within the 3 3-year period, are averaged to obtain a detrended
flux reflecting long-term mean hydrologic conditions.
    The sediment model is estimated using 1,828 monitoring
stations located on the RF1 stream network (fig. 1). Sta-
tions were selected for inclusion in the model if they had at
least 15 concentration measurements during the period WY
1975-2007, and the standard error of the flux estimate did not
exceed 80 percent of the flux estimate. Of the stations included
in the analysis, 90 percent had streamflow records in excess
of 5 years, 70 percent had records exceeding 19 years, and 50
percent had records that exceeded 32 years. Approximately
700 monitoring stations have been indexed to the RF1 stream
      Figure 1. Location of 1,828 water-quality monitoring stations used in the SPARROW sediment model, in relation to the
      Reach File 1 (RF1) stream network.

-------
                                                                            Preliminary Model Estimation Results
network as pail of previous studies (Alexander and others,
2000). The remaining monitoring stations were linked to RF1
reaches via their association with USGS streamgages, which
are linked to the National Hydrography Dataset (NHD) reach
network (Stewart and others, 2006). The location of these
streamgages was transferred from NHD to RF1 using informa-
tion on streamgage latitude and longitude (based on the NHD
location), stream name, and reported drainage area. If multiple
stations were present for the same RF1 reach, the alternative
stations were first ranked in terms of drainage area, number
of days predictions were made, the number of water-quality
observations used in estimation, the coefficient of variation of
the flux estimate, and whether the water-quality and stream-
flow records were sufficient to support detrending of flux to
the base year. (To be detrended, the water-quality and stream-
flow records must span at least 3 years and, if extended no
more than 15 percent in duration, must include the detrending
date June 30, 1992.) The station with the lowest sum of these
ranks was selected for inclusion in  the model.
     Most of the source variables for the SPARROW sediment
model are expressed as extensive measures of land use. Data
on land cover and land use have been developed from the 2001
USGS National Land Cover Data (NLCD) Set Retrofit Change
Product [Multi-resolution Land Characteristics (MRLC),
2001], derived from Landsat Thematic Mapper/Embed-
ded Trace Macrocells (TM/ETM) remotely sensed imagery
at 30-meter resolution and classified according to the eight
Anderson Level 1 categories. These data were transformed to
1-square kilometer (km2) cells within a Lambert map projec-
tion as consistent with HYDROlk for use within SPARROW.
The 1-km2 cells are then resolved to catchments associated
with specific RF1 reaches. For model estimation, land use was
assigned the 1992 values of the 2001 NLCD Retrofit Change
Product; the  2001 values of the Retrofit Change Product were
used to simulate water-quality conditions for 2001.
     Federal nonforested (range and barren) land was included
in the model separately from private land. Federal land extent,
taken from the Federal Land coverage of the National Atlas
(U.S. Geological Survey, 2003) and transformed to 1-km2 cells
in Lambert projection, was apportioned into Federal range
and Federal ban-en land using the 1-km2 transformation of
the 1992 NLCD Change Product for Anderson Level I range
and barren land classes  (see above). For model simulation of
2001 conditions, Federal range and barren land were similarly
estimated using the land-use estimates from the 2001 NLCD
Retrofit Change Product.
     Variables governing the estimated delivery of contami-
nants from the land to RF1 streams include soil erodibility
[the revised universal soil loss equation (RUSLE) K factor],
soil permeability (inches/hour; depth integrated), mean slope
(percent), and precipitation (RUSLE R factor). Slope, soil
erodibility, and permeability were obtained from the State Soil
Survey Geographic (STATSGO) database (U.S. Department of
Agriculture,  1994), converted to a 1-km2 grid in the Lambert
projection, and averaged over RF1  catchments. The  RUSLE
rainfall factor was derived by interpolating a digitized national
map of rainfall factor isoline contours (Wischmeier and Smith,
1978), creating a continuous 1-km2 grid surface in the Lambert
projection. The grid coverage was subsequently averaged over
individual RF1 catchments.
    The nonlinear, leasl-squares estimation results of a pre-
liminary version of the SPARROW suspended sediment model
are given in table 1. The preliminary model includes six source
terms, five of which are measured by area of specific land
use (urban, forested, Federal nonforesled, agricultural, and
other land, expressed in km2), and an additional source given
by the length of the stream channel. The Federal land class
consists only of Federal range and barren land; it excludes
Federal, forested land, which is incorporated in the forested
land class. Agricultural land includes cropland, pasture land,
and orchards. Other land consists of non-Federal range and
barren land. Among all the land classes, only wetlands and
land covered by water, ice, or snow are excluded as a potential
source. The source described as "strcambed" relates to stream
channels as a direct source of sediment, and is measured in
terms of stream length (expressed in meters).
    The transport of sediment from the land surface to RF1
rivers is mediated by a land-to-water delivery factor that is
expressed as a function (see Equation 1.28 in Schwarz and
others, 2006) of logarithm-transformed values of soil permea-
bility, soil credibility (USLE K factor), land-surface slope, and
the USLE rainfall factor.  The strcambed source is mediated
by the logarithm of streamflow, distinguished by streamflows
above and below 500 cubic feet per second (ft-Vs). Although
the mediation of the streambed source by slreamflow is not
a land process, the manner in which the process is speci-
fied in SPARROW is mathematically equivalent to treating
streamflow as a land-to-water variable affecting  the streambed
source, and for this reason streamflow is listed as a land-lo-
water variable in table 1.  To facilitate interpretation of the
source coefficients, the delivery variables are all expressed as
deviations from their mean value.
    Streamflow was used as the mediating factor affecting
the mobilization of sediment from the stream channel. There
are two principal physical factors affecting the mobilization
of sediment from the streambed:  the energy of the stream,
represented by stream velocity; and the availability of channel
material, which is proportional to  the area of the streambed  per
unit of channel length. Direct measurements of channel veloc-
ity, width, and depth (the determinants of streambed area per
unit channel  length) are not available for most reaches in  the
RF1 network. Streamflow was deemed to be a viable surrogate
for these variables because it is highly correlated with them
(Schwarz and others, 2006) (in fact, the estimate of velocity
included with the RF1. network is derived from slreamflow),
and because  estimates of streamflow exist for all reaches.

-------
4     A Preliminary SPARROW Model of Suspended Sediment for the Conterminous United States
     The model specifies two instream sediment-attenuation
processes:  attenuation in streams (see Equation 1.30 in
Schwarz and others, 2006), distinguished by three streamflow
classes (less than 500 ftYs, 500-1,000 ftVs, and greater than
1,000 ft3/s); and reservoir attenuation, specified as a function
of area! hydraulic load (see Equation 1.34 in Schwarz and
others, 2006). The three streamflow classes used to distin-
guish instream decay are characteristic of the streamflow at
the  1,828 monitoring stations:  1,040 stations have streamflow
less than 500 ftVs, and 539 stations have streamflow exceeding
1,000 ftVs. Attenuation in reservoirs is specified to be a func-
tion of the ratio of the reservoir settling velocity, the estimated
mean rate at which sediment moves vertically in water, to area!
hydraulic load (the ratio of streamflow to reservoir surface
area), which represents the velocity a particle at the surface of
the  stream would need to travel in order to reach the bottom of
the  reservoir within the average period that the  streamflow is
impounded in the reservoir.
     The estimation results given in table 1 characteristically
reflect the large uncertainty associated with sediment model-
ing. The model root mean squared error (RMSE) is 1.4, imply-
ing that predicted sediment flux or concentration in any given
reach has an error of approximately 140 percent (Schwarz
and others, 2006). This compares with the much smaller 0.3
RMSE obtained with total nitrogen models (Alexander and
others, 2008). Despite this uncertainty, many of the model
coefficients are statistically significant. With the exception of
forested land, all of the source variables are highly statistically
significant. The largest intrinsic sediment yield is associated
with Federal range and barren land; urban land has the second
highest intrinsic yield. Stream channels are also a statistically
significant source of sediment.
     Land-to-water delivery for land sources is strongly
mediated by the four delivery variables—soil permeability,
soil credibility, slope, and rainfall. As would be expected and
with the exception of soil permeability, the presence of higher
levels of these factors results in  greater sediment delivery to
streams. Conversely, permeable soils reduce the delivery of
sediment, presumably because more of the water runoff infil-
trates into the ground leaving less overland flow to transport
     Table 1.  Preliminary estimation results for the SPARROW suspended sediment model.
[Kg/km2/yr = kilograms per square kilometer per year; kg/m/yr = kilograms per meter per year; fVYs = cubic feet per second; and m/yr = meters per
year!
Parameter
Source Coefficients
Urban land
Forested land
Federal non-forested land
Agricultural land
Other land
Strcambcd (reach length)
Land-to- Water Delivery Factors
Slope
Soil permeability
R-factor
K-factor
Flow |< 500 ft3/s] (Reach)
Flow [> 500 ft3/sj (Reach)
Stream Attenuation Factors
Travel time (Q < 500 ft3/s)
Travel time (500 < Q < 1.000 ft'/s)
Travel tune (Q > 1,000 ftVs)
Reservoir settling velocity
Number of Observations
Root mean squared error (RMSE)
R-square
Units

kg/knvVyr
kg/km2/yr
kg/krrrVyr
kg/km2/yr
kg/km2/yr
kg/m/yr

-
-
-
-
-
-

day-1
day'
day-1
m/yr
1.828
1.414
0.711
Estimate

47.130
634
64,344
18,047
11,343
28.80

0.804
-0.778
0.821
1.292
0.154
0.721

-0.007
-0.233
0.009
36.49



Standard Error

9,925
898
12,411
3,623
3,186
6.40

0.087
0.094
0.081
0.279
0.100
0.354

0.016
0.057
0.047
5.552



p-¥alue

0.000
0.480
0.000
0.000
0.000
0.000

0.000
0.000
0.000
0.000
0.125
0.042

0.673
0.000
0.854
0.000




-------
                                                                                              Model Simulation
sediment. Greater streamflow causes an increase in the amount
of sediment generated from stream channels, with the largest
effect associated with streams having flows greater than 500
ft3/s.
     Reservoir retention is statistically significant and indi-
cates sediment settles at a mean velocity of 36 meters per
year (m/yr), comparable to estimates of reservoir attenuation
obtained for phosphorus (Alexander and others, 2008). The
preliminary model indicates medium-sized streams (flow
500-1,000 ft3/s) have a statistically significant negative rate
of instream attenuation, indicating that medium streams are a
source of sediment, in addition to the stream-channel source
identified above. Unlike the streambed source of sediment (see
above), which is dictated by channel length and, thus, explic-
itly tied to a physical entity, the implied "source" of sediment
arising from a negative rate of instream attenuation represents
a proportional "enhancement" of sediment already suspended
in the stream (see  Equation 1.30 in Schwarz and others, 2006).
Because this proportional enhancement of sediment is not
associated with a physical source, it is inconsistent with mass
balance and represents an anomalous finding of the model that
is not yet explainable. Instream attenuation in small and large
streams is not significantly statistically  different from zero.
Thus, the preliminary model does not find evidence for sedi-
ment loss in streams.
Model Simulation

     The estimated SPARROW suspended sediment model for
base WY 1992 can be used to simulate water-quality condi-
tions for 2001, with and without U.S. EPA proposed changes
in the regulation of construction activity. The simulation of
suspended sediment flux for WY 2001, without changes in
regulation, is obtained using the model prediction equation,
described as Equation 1.120 in Schwarz and others (2006),
with all land use-related source variables set to 2001 values
according to the 2001 NLCD Retrofit Change Product. Flow-
weighted average sediment concentration is estimated by
dividing simulated flux estimates by mean streamflow over the
period WY 1975-2006, obtained from USGS streamgages and
interpolated to RF1 reaches (Wolock, U.S. Geological Survey,
2008, written commun.). Although the SPARROW model
does not explicitly include a source term for construction, such
loading is implicitly accounted for in the urban land compo-
nent of the model. Therefore, the 2001 precompliance loading
from construction (that is, the "base-case" scenario loading) is
incorporated in the 2001  loading attributed to urban land that
is obtained by evaluating the urban land variable in the SPAR-
ROW model using the 2001  NLCD Change Product value.
     The absence of an explicit term for construction loading
in the SPARROW model necessitates the development of an
indirect method for assessing changes in sediment loading
arising from different construction industry regulation scenar-
ios. Suspended  sediment loading under alternative regulation
scenarios has been estimated by U.S. EPA using a variation
of the Universal Soil Loss Equation (USLE). The USLE
method determines the amount of soil that is mobilized and
delivered, under a proposed regulation scenario, to the edge
of a construction site. To evaluate the impact that changes in
these loadings have on RF1 stream-sediment flux and flow-
weighted concentration, it is first necessary to assess the rate
at which "edge of site" loads are subsequently delivered to
RF1  streams. To do this, we use the estimated rate of delivery
from agricultural land, a source that is explicitly included in
the model, and that can be factored into a mobilization, "edge
of site" delivery component and a stream-delivery component.
The method described below isolates the stream-delivery com-
ponent from the overall rate of delivery from agricultural land
and applies this component to  the change in construction load-
ing to determine the change in loading to RF1 streams. Thus,
the approach assumes that the  delivery of sediment from the
edge of a site to an RF1 stream is the same for both construc-
tion and agriculture activities;  the mobilization and delivery
of sediment to the edge of the  site between these activities
is allowed to differ. Given that urban areas,  as compared to
agricultural areas, generally exhibit higher rates of runoff, with
compressed runoff duration periods for a given precipitation
event, the assumption probably leads to an underestimate in
the change in stream-sediment flux from proposed regulation
of the construction industry. It would not be necessary to make
this assumption if the analysis was based instead on a factor-
ization of urban land delivery; however, as is indicated below,
the information necessary to do this is not available.
     The amount of sediment mobilized from agricultural
land, delivered to the edge of field, and subsequently trans-
ported to an RF1 stream, is estimated in SPARROW as the
product of the agricultural land-source coefficient and the
associated land-to-water delivery factor. This quantity, denoted
KAG and expressed in units of yield as kilograms per square
kilometer per year (kg/km2/yr), is conceptually divided into
two components:  a component representing the amount of
sediment mobilized from agricultural land and delivered to the
edge of site, denoted ^"EOS.AG, and a component representing
the fraction of this material that is subsequently delivered to an
RF1  stream, denoted A"RF1. If ALEOS(S) represents the change in
construction sediment loading to the edge of site, as estimated
by U.S. EPA using the USLE method, and ALRF1(S) represents
the change in sediment loading to RF1 streams associated with
construction regulation scenario S, then the two loadings are
related according to
                                                      (D
                                      K
                                        EOS-AG
     Given estimates of the regulation-induced change in
sediment delivered to each RF1 reach, instream processes
associated with attenuation in channels and reservoirs, as
described by the SPARROW model estimates, can be applied
to estimate changes in sediment flux and concentration for
all RF1 reaches. Additionally, because SPARROW imposes

-------
6     A Preliminary SPARROW Model of Suspended Sediment for the Conterminous United States
mass balance, the reservoir attenuation process can be used to
assess changes in the amount of sediment retained in each of
the approximately 4,000 reservoirs linked directly to the RF1
network.
    To implement the method described by Equation 1, it
is necessary to have estimates of KAG, the delivery factor
for agricultural land, and KEOS_Aa, the amount of soil erosion
mobilized from agricultural land and delivered to the edge of
site. K^c is reach specific and is  estimated in SPARROW as
a function of the land-to-water delivery factors for agricul-
ture (see Equation 1.28 in Schwarz and others, 2006) and the
empirically estimated values of the agricultural land-source
and land-to-water delivery coefficients. The value for KEOS_AO
is obtained using information on soil erosion included in the
1992 National Resources Inventory (NRI) (U.S. Department
of Agriculture, 1994), which reports county estimates of soil
erosion rates for cropland, pasture, and orchards—the land
classes encompassed by the class labeled "agricultural land" in
the NLCD Retrofit Change Product. ttTRF1 is based on agricul-
tural land because erosion rates for urban land are not reported
by the NRI.) The USLE-based county erosion rates for each
of the three land classes were weighted according to the share
of county land in the respective class (as reported in the 1992
NRI) and then averaged. If there was no county estimate for
cropland erosion, an average erosion rate for that county was
not computed. The county average erosion rate was appor-
tioned to RF1 catchments according to a 1-km2 grid of agricul-
tural land area derived from the  1992 NLCD Retrofit Change
Product. For 6,329 catchments where an NRI erosion rate was
not available, the erosion rate was estimated by determining
an agricultural land weighted average of all available erosion
rates for catchments in the same 8-digit hydrologic cataloging
unit. Lack of an 8-digit cataloging unit average necessitated
using a 6-digit cataloging unit average  for 713 catchments, and
the remaining 54 catchments were estimated using a 4-digit
cataloging unit average.
     The preliminary SPARROW model for suspended sedi-
ment described above has certain limitations, some of which
are inherent to the methodology and some the result of the
particular model application. An example of the former is the
restriction of the analysis to the description of long-term mean
water-quality conditions. As explained in Schwarz and others
(2006), this restriction is a consequence of imposing mass
balance on the predictions. One of the benefits of the mass
balance methodology is that it facilitates the interpretation of
model coefficients, and enables the comparison of coefficient
estimates to estimates obtained by other studies in the litera-
ture; however,  the restriction to mean water-quality conditions
precludes an analysis of  the frequency with which conditions
of extreme sediment transport occur.
     A second example of a methodology-imposed limita-
tion concerns the use of the statistical method to estimate
model coefficients. The statistical method provides consider-
able  insight into the evaluation of model fit and the empirical
relevance of individual model processes. It also enables the
estimation of prediction uncertainty. However, reasonable
precision in the statistical estimation of model coefficients
is generally possible only if the number of specified model
parameters is limited to those associated with sources and
delivery processes that have the greatest influence on water
quality. The resulting model is parsimonious, but may be
overly simplistic in terms of the range of processes affecting
sediment transport.
     The RF1 reach network is fairly coarse, and its use in the
present application limits the ability to predict water-quality
conditions in smaller streams. The median headwater catch-
ment area in RF1 is 88 km2, implying water-quality conditions
in streams with smaller catchments are unresolved. Addition-
ally,  the smallest monitored catchment in the SPARROW sedi-
ment model is 13 km2, with only five percent of the monitored
catchments less than 100 km2. A SPARROW analysis struc-
tured on a denser reach network, such as the National Hydrog-
raphy Dataset, would relax the reach network limitation, but
the large number of reaches associated with this network
would make it difficult to conduct a national analysis.
     The large error obtained in the present analysis implies
the prediction of sediment flux or concentration in any given
reach segment is imprecise. Although this error compromises
the ability to describe water-quality conditions in any given
reach, it does not preclude using the model to characterize
water quality in a large grouping of reaches. As long as the
error across reaches is sufficiently independent, the assessment
of mean water quality in a group of reaches becomes more
precise as the size of the group increases.
     With the exception of reservoirs, the preliminary model
does not find evidence of sediment attenuation in streams. The
result implies that sediment transport in streams is not in a
steady state. Additional investigation is necessary to determine
if this result is real, or if there are additional reach attributes,
currently absent from the model, that identify a subset of
reaches where sediment attenuation takes place.
Alexander, R.B., Slack, J.R., Ludtke, A.S., Fitzgerald, K.K.,
  Schertz, T.L., Briel, L.I., and Buttleman, K.P., 1996, Data
  from selected U.S. Geological Survey National Stream
  Water-Quality Networks (WQN): U.S. Geological Sur-
  vey Digital Data Series DDS-37, available online at
  http://pubs.usgs.gov/dds/wqn96cd/. (Accessed January 30,
  2008.)

-------
                                                                                              References Cited
Alexander R.B., Smith, R.A., and Schwarz, G.E., 2000, Effect
  of stream channel size on the delivery of nitrogen to the
  Gulf of Mexico:  Nature, v. 403, no. 6771, p. 758-761.

Alexander, R.B., Smith, R.A., Schwarz, G.E., Boyer, E.W.,
  Nolan, J.V., and Brakebill, J.W., 2008, Differences in phos-
  phorus and nitrogen delivery to the Gulf of Mexico from the
  Mississippi River Basin: Environmental Science & Tech-
  nology, v. 42, issue 3, p. 822-830.

Conn, T.A., 2005, Estimating contaminant loads in rivers—An
  application of adjusted maximum likelihood to type 1
  censored data: Water Resources Research 41:W07003,
  doi: 10.1029/2004 WR003833.

Mueller, O.K., and Spahr, N.E., 2005, Water-quality, stream-
  flow, and ancillary sata for nutrients in streams and
  rivers across the Nation, 1992-2001: U.S. Geological
  Survey Digital Data Series DDS-152, available online at
  http://pubs.tisgs.gov/ds/2005/152/. (Accessed 1/30/2008.)

Nolan, J.V., Brakebill, J.W., Alexander, R.B., and
  Schwarz, G.E., 2002, ERF1_2—Enhanced River
  Reach File 2.0: U.S. Geological Survey Open-
  File Report 02-40. (Also available online at
  http://water.usgs.gov/GIS/metadata/usgswrd/XMIJerfl _2.
  xml.)

Ruddy, B.C., and Hitt,  K.J., 1990, Summary of selected char-
  acteristics of large reservoirs in the United States and Puerto
  Rico, 1988: U.S. Geological Survey Open-File Report
  90-163.

Schwarz, G.E., Hoos, A.B., Alexander, R.B., and Smith,
  R.A, 2006, The SPARROW surface water-quality
  model: Theory, application and user documentation:
  U.S. Geological Survey Techniques and Methods, sec-
  tion B, chapter 6-B3, 248 p., available only online at
  http://pubs. usgs.gov/tm/2006/tm6b3/. (Accessed June 10,
  2008.)

Steffen, L.J., 1996, A reservoir sedimentation survey informa-
  tion system—RESIS, in Proceedings of the Sixth Federal
  Interagency Sedimentation Conference, Las Vegas. NV,
  March 10-14, 1996: Sponsored by the Subcommittee on
  Sedimentation, Interagency Advisory Committee on Water
  Data, p. 29-37.

Stewart, D.W., Rea, A.M., and Wolock, D.M., 2006,  USGS
  streamgages linked to the medium resolution NHD:  U.S.
  Geological Survey Data Series DS-195.

U.S. Army Corps of Engineers (USACE), 2006,
  USAGE National Inventory of Dams Web site at
  http://cnincli.tec.anny.mil/nidpublic/webpages/nid.cjm.
  (Accessed June 17, 2005.)
U.S. Department of Agriculture, 1994, State Soil Survey Geo-
  graphic Data Base: U.S. Department of Agriculture, Soil
  Conservation Service, National Cooperative Soil Survey,
  Miscellaneous Publication Number 1492.

U.S. Environmental Protection Agency. 1996. The U.S.
  Environmental Protection Agency Reach File Version 1.0
  (RF1) for the Conterminous United Slates in BASINS:
  U.S. Environmental Protection Agency Web page at
  http://www.epa.gov/waterscience/BASINS/metadata/rfl.httn.
  (Accessed June 5, 2008.)

U.S. Geological Survey, 2003, Federal lands
  of the United States, available online at
  http://nationalatlas.gov/inld/fedlanp. html. (Accessed June
  5, 2008.)

U.S. Geological Survey, 2006a, HYDROlk
  Documentation, available online at
  http://edc.usgs.gov/products/elevation/gtopo30/hydro/
  readtne.html. (Accessed June 5, 2008.)

U.S. Geological Survey, 2006b, National Stream Water Qual-
  ity Network (NASQAN) Published Data, available online at
  http://waler.usgs.gov/nasqan/dala/jinalclata.html. (Accessed
  January 30, 2008.)

U.S. Geological Survey, 2008, USGS Water Data for the
  Nation, available online at
  http://waterdata.usgs.gov/nwis. (Accessed February 4,
  2008.)

Wischmeier, W.H., and Smith, D.D., 1978, Predicting rainfall
  erosion losses—A guide to conservation planning: USDA
  Agriculture Handbook 537, 58 p.

-------
                                                                                                                                                                                        13

                                                                                                                                                                                        m
                                                                                                                                                                                        3
                                                                                                                                                                                        a>
                                                                                                                                                                                       eo
                                                                                                                                                                                        3D
                                                                                                                                                                                        3D
                                                                                                                                                                                        s
                                                                                                                                                                                        O
                                                                                                                                                                                        =:..
                                                                                                                                                                                        to
                                                                                                                                                                                       C/J
                                                                                                                                                                                       B
                                                                                                                                                                                       trt
                                                                                                                                                                                       T3
                                                                                                                                                                                       09
                                                                                                                                                                                       I]
                                                                                                                                                                                       a,
                                                                                                                                                                                       CO
                                                                                                                                                                                       =:..

                                                                                                                                                                                       C/J
                                                                                                                                                                                       CD
                                                                                                                                                                                       O.

                                                                                                                                                                                       i"
                                                                                                                                                                                       (B
                                                                                                                                                                                        =T
                                                                                                                                                                                        (B

                                                                                                                                                                                        O
                                                                                                                                                                                        O
                                                                                                                                                                                        3
                                                                                                                                                                                        <-*
                                                                                                                                                                                        CD
                                                                                                                                                                                        •«{
                                                                                                                                                                                        3_

                                                                                                                                                                                        5"
                                                                                                                                                                                        o
                                                                                                                                                                                        c
                                                                                                                                                                                        «/)

                                                                                                                                                                                        c
                                                                                                                                                                                        =1
                                                                                                                                                                                       C/J
                                                                                                                                                                                       **>
                                                                                                                                                                                       a>
                                                                                                                                                                                       s
                                                                                                                                                                                       o
                                                                                                                                                                                       T3
                                                                                                                                                                                       O3
                                                                                                                                                                                       ~a
                                                                                                                                                                                       o

                                                                                                                                                                                       3
                                                                                                                                                                                       I-O
                                                                                                                                                                                       CD
                                                                                                                                                                                       CD
                                                                                                                                                                                       oo


                                                                                                                                                                                       ISJ
                                                                                                                                                                                       C3
                                                                                                                                                                                       cn
Printed on recycled paper

-------
Environmental Impact and Benefits Assessment for the C&D Category
Appendix E - Estuary TSS Concentration Calculations Using
Dissolved Concentration Potentials
As described in Section 6.3, EPA used SPARROW predicted sediment loadings in combination with
dissolved concentration potentials (DCPs) for calculating sediment concentrations where flow estimates
were not available from SPARROW. This appendix includes all the tables referenced in Section 6.3.
DCP values and baseline TSS concentrations calculated using DCPs are shown in Table E-l through
Table E-4 for Southeastern, Gulf, Northeastern, and West Coast estuaries, respectively. Data on these and
additional estuaries are found via the NOAA Coastal Geospatial Data Project
(http ://coastalgeospatial .noaa.gov).

Table E-1 : TSS Concentrations
DCP, Southeastern Estuaries
Estuary
Albemarle/Pamlico Sound, NC, VA
Altamaha River, GA
Biscayne Bay, FLA
Bogue Sound, NC
Broad River, SC
Cape Fear River, NC
Charleston Harbor, SC
Indian River, FL
N and S Santee Rivers, SC
New River, NC
Ossabaw Sound, GA
Savannah River, SC,GA
St Catherines/Sapelo Sound, GA
St Helena Sound, SC
St Johns River, FL
St. Andrews/St. Simons Sound, GA
Winyah Bay, SC, NC
(mg/L) Estimated by
DCP
0.14
0.37
0.4
1.47
4.92
0.61
0.44
1.02
2.91
7.68
1.99
0.43
7.56
0.95
0.83
2.43
0.39
Base TSS
(mg/1)
33.7
82.8
0.2
3.0
16.3
69.7
2.5
3.4
456.8
15.3
81.0
61.2
17.2
10.8
13.3
47.4
109.0

                 Table E-2: TSS Concentrations (mg/L) Estimated by
                 DCP, Gulf of Mexico Estuaries
Estuary
Apalachee Bay, FL, GA
Apalachicola Bay, FL
Arkansas Bay, TX
Atchafalaya and Vermillion Bays, LA
Brazos River, TX
Calcasieu Lake, LA
Charlotte Harbor, FL
Choctawhatchee Bay, FL, AL
Corpus Cristi Inner Harbor, TX
Galveston Bay, TX
DCP
0.36
0.17
6.02
0.04
1.11
1.18
0.76
0.7
4.67
0.4
Base TSS
(mg/1)
8.2
21.1
152.7
254.8
598.3
48.3
15.1
51.7
555.0
93.4
November 2009
                                                                                    E-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
                Table E-2: TSS Concentrations (mg/L) Estimated by
                DCP, Gulf of Mexico Estuaries
Estuary
Laguna Madre, TX
Matagorda Bay, TX
Mississippi River, LA, MS
Mississippi Sound, MS, LA, AL
Mobile Bay, AL
Pensacola Bay, FL
Perdido Bay, FL, AL
Sabine Lake, LA, TX
San Antonio Bay, TX
St Andrew Bay, FL
Suwanee River, FL
Tampa Bay, FL
Ten Thousand Islands, FL
DCP
0.34
0.81
0.01
0.17
0.08
0.46
2.89
0.38
1.30
0.76
0.38
1.03
1.94
Base TSS
(mg/1)
21.7
212.0
198.4
127.5
87.7
44.1
34.5
76.3
833.3
2.2
24.5
10.4
1.6

                Table E-3: TSS Concentrations (mg/L) Estimated by
                DCP, Northeastern Estuaries
Estuary
Barnegat Bay, NJ
Blue Hill Bay, ME
Buzzards Bay, MA
Cape Cod Bay, MA
Casco Bay, ME
Chesapeake Bay, VA, MD, DE, PA, DC
Chincoteague Bay, MD, VA
Delaware Bay, DE, NJ, PA, MD
Englishman Bay, ME
Gardiners Bay, NY
Great Bay, NE, NH
Great South Bay, NY
Hudson River/Raritan Bay, NY, NJ, MA, CT
Long Island Sound, NY, CT, MA
Massachusetts Bay, MA
Merrimack River, NH., MA
Muscongus Bay, ME
Narragansett Bay, MA, RI
Narragauges Bay, ME
Passaquoddy Bay, ME
Penobscot Bay, ME
Saco Bay, ME, NH
Sheepscop Bay, ME, NH
DCP
1.36
1.03
1.04
0.69
0.61
0.072
3.08
0.14
0.92
1.77
1.54
5.07
0.2
0.054
0.27
1.01
2.25
0.52
1.54
0.27
0.13
0.45
0.088
Base TSS
(mg/1)
3.7
1.7
3.8
0.5
3.1
91.6
2.0
40.0
4.7
2.0
7.6
12.0
52.4
9.7
2.3
32.0
6.5
8.2
4.2
1.5
7.1
4.2
0.2
November 2009                                                                        E-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table E-4: TSS Concentrations (mg/L) Estimated by
DCP, West Coast Estuaries
Estuary
Alsea River, OR
Columbia River, WA, OR
Coos Bay, OR
Eel River, CA
Grays Harbor, WA
Humboldt Bay, CA
Klamath River, OR, CA
Monterey Bay, CA
Nehalam River, OR
Netarts Bay, OR
Puget Sound, WA
Rogue River, OR
San Diego Bay, CA
San Francisco Bay, CA
San Pedro Bay, CA
Santa Monica Bay, CA
Siletz Bay, OR
Siuslaw River, OR
Tillamook Bay, OR
Umpqua River, OR
Willapa Bay, WA
Yaquina Bay, OR
DCP
2.086
0.032
2.27
0.39
0.278
8
0.31
1.05
1.556
12.564
0.039
0.556
12.31
0.104
4.41
1.371
2.424
1.853
1.049
0.793
0.466
3.356
Base TSS (mg/1)
37.4
77.0
50.9
65.3
21.2
38.7
79.8
148.2
38.5
13.0
9.7
125.4
88.9
89.3
104.0
30.0
42.8
46.5
21.3
110.6
12.1
25.7
In the final analysis, the DCP approach was only used in 43 reaches without flow predictions. The
number of reaches in each estuary where the DCP approach was used to calculate TSS concentrations is
shown in Table E-5 through Table E-8 .
Table E-5: Total Number Reaches in the Analysis located
Use DCP to Calculate TSS by Estuary
Estuary
Albemarle/Pamlico Sound, NC, VA
Altamaha River, GA
Bogue Sound, NC
Broad River, SC
Cape Fear River, NC
Charleston Harbor, SC
Indian River, FL
N and S Santee Rivers, SC
New River, NC
Ossabaw Sound, GA
Savannah River, SC, GA
St Catherines/Sapelo Sound, GA
St Helena Sound, SC
St Johns River, FL
WinyahBay,SC,NC
in Southeastern
Estuaries that
DCP Number of Reaches
0.14
0.37
1.47
4.92
0.61
0.44
1.02
2.91
7.68
1.99
0.43
7.56
0.95
0.83
0.39
2
0
0
0
0
1
0
1
1
1
0
0
0
2
1
Note: Only estuaries with reaches included in the analysis are listed in the table.
November 2009
                                                                                             E-3

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table E-6: Total Number Reaches in the Analysis
that Use DCP to Calculate TSS by Estuary
Estuary
Apalachicola Bay, FL
Charlotte Harbor, FL
Choctawhatchee Bay, FL, AL
Galveston Bay, TX
Pensacola Bay, FL
Perdido Bay, FL, AL
Sabine Lake, LA, TX
St Andrew Bay, FL
Suwanee River, FL
Tampa Bay, FL
located in Gulf of Mexico Estuaries
DCP
0.17
0.76
0.7
0.4
0.46
2.89
0.38
0.76
0.38
1.03
Number of Reaches
0
0
2
0
0
0
0
1
0
2
Note: only estuaries with reaches included in the analysis are listed in the table.

Table E-7: Total Number Reaches in the Analysis
that Use DCP to Calculate TSS by Estuary
Estuary
Barnegat Bay, NJ
Buzzards Bay, MA
Cape Cod Bay, MA
Casco Bay, ME
Delaware Bay, DE, NJ, PA, MD
Gardiners Bay, NY
Great Bay, NE, NH
Great South Bay, NY
Hudson River/Raritan Bay, NY, NJ, MA, CT
Long Island Sound, NY, CT, MA
Massachusetts Bay, MA
Narragansett Bay, MA, RI
Saco Bay, ME, NH
Sheepscop Bay, ME, NH
located in Northeastern Estuaries
DCP
1.36
1.04
0.69
0.61
0.14
1.77
1.54
5.07
0.2
0.054
0.27
0.52
0.45
0.088
Number of Reaches
1
0
0
1
0
0
0
0
2
1
0
2
1
0
Note: Only estuaries with reaches included in the analysis are listed in the table.

Table E-8: Total Number Reaches in the Analysis
Use DCP to Calculate TSS by Estuary
Estuary
Alsea River, OR
Coos Bay, OR
Grays Harbor, WA
Nehalam River, OR
Netarts Bay, OR
Puget Sound, WA
Rogue River, OR
San Diego Bay, CA
San Francisco Bay, CA
San Pedro Bay, CA
Santa Monica Bay, CA
Siletz Bay, OR
Siuslaw River, OR
Tillamook Bay, OR
Umpqua River, OR
Willapa Bay, WA
Yaquina Bay, OR
located in West
DCP
2.086
2.27
0.278
1.556
12.564
0.039
0.556
12.31
0.104
4.41
1.371
2.424
1.853
1.049
0.793
0.466
3.356
Coast Estuaries that
Number of Reaches
0
5
3
0
0
4
1
1
2
1
0
0
0
2
0
1
0
Note: Only estuaries with reaches included in the analysis are listed in the table.

November 2009                                                                                      E-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
Appendix F - TSS Subindex Curve Parameters
This appendix provides the ecoregion-specific parameters used in estimating the TSS water quality
subindex, as follows:
       If [TS S] < TSS! oo                    Subindex =100
       If TSS100 < [TSS] < TSS10              Subindex = a exp(/> [TSS])
       If [TSS] > TSS,0                     Subindex = 1 0
       where [TSS] is the measured concentration and TSS10, TSS10o, a, and b are specified in Table F-l.
Table
ID
10.1.2
10.1.3
10.1.4
10.1.5
10.1.6
10.1.7
10.1.8
10.2.1
10.2.2
10.2.4
11.1.1
11.1.2
11.1.3
12.1.1
13.1.1
15.4.1
5.2.1
5.2.2
5.3.1
5.3.3
6.2.10
6.2.11
6.2.12
6.2.13
6.2.14
6.2.15
6.2.3
6.2.4
6.2.5
6.2.7
6.2.8
6.2.9
7.1.7
7.1.8
F-1: TSS Subindex Curve Parameters, by Ecoregion
Ecoregion Name
Columbia Plateau
Northern Basin and Range
Wyoming Basin
Central Basin and Range
Colorado Plateaus
Arizona/New Mexico Plateau
Snake River Plain
Mojave Basin and Range
Sonoran Desert
Chihuahuan Desert
California Coastal Sage, Chaparral, and Oak Woodlands
Central California Valley
Southern and Baja California Pine-Oak Mountains
Madrean Archipelago
Arizona/New Mexico Mountains
Southern Florida Coastal Plain
Northern Lakes and Forests
Northern Minnesota Wetlands
Northern Appalachian and Atlantic Maritime Highlands
North Central Appalachians
Middle Rockies
Klamath Mountains
Sierra Nevada
Wasatch and Uinta Mountains
Southern Rockies
Idaho Batholith
Columbia Mountains/Northern Rockies
Canadian Rockies
North Cascades
Cascades
Eastern Cascades Slopes and Foothills
Blue Mountains
Strait of Georgia/Puget Lowland
Coast Range
a
126.56
112.42
123.36
121.22
144.44
126.76
146.39
119.34
112.39
214.39
127.97
171.86
115.12
261.35
120.98
116.95
157.76
154.99
174.99
245.15
144.64
238.9
185.36
124.28
153.42
184.23
180.7
396.62
240.95
192.94
178.82
148.35
181.06
174.78
b
-0.0038
-0.0007
-0.001
-0.0018
-0.001
-0.0004
-0.0027
-0.0015
-0.0002
-0.0005
-0.0012
-0.0044
-0.0007
-0.0005
-0.0004
-0.0405
-0.0233
-0.0186
-0.0261
-0.0176
-0.0038
-0.0068
-0.0116
-0.0014
-0.0031
-0.0142
-0.0168
-0.0308
-0.0193
-0.0181
-0.0145
-0.0037
-0.0224
-0.0114
TSSioo
63
160
220
109
363
668
142
121
567
1,419
205
122
197
2,053
477
4
20
24
21
51
98
129
53
160
140
43
35
45
46
36
40
107
27
49
TSS10
668
3,457
2,513
1,386
2,670
6,349
994
1,653
12,097
6,130
2,124
646
3,491
6,527
6,233
61
118
147
110
182
703
467
252
1,800
881
205
172
119
165
164
199
729
129
251

November 2009
                                                                                      F-1

-------
Environmental Impact and Benefits Assessment for the C&D Category

Table
ID
7.1.9
8.1.1
8.1.2
8.1.3
8.1.4
8.1.5
8.1.6
8.1.7
8.1.8
8.1.10
8.2.1
8.2.2
8.2.3
8.2.4
8.3.1
8.3.2
8.3.3
8.3.4
8.3.5
8.3.6
8.3.7
8.3.8
8.4.1
8.4.2
8.4.3
8.4.4
8.4.5
8.4.6
8.4.7
8.4.8
8.4.9
8.5.1
8.5.2
8.5.3
8.5.4
9.2.1
9.2.2
9.2.3
9.2.4
9.3.1
9.3.3
9.3.4
9.4.1
9.4.2
9.4.3
9.4.4
9.4.5
9.4.6
F-1: TSS Subindex Curve Parameters, by Ecoregion
Ecoregion Name
Willamette Valley
Eastern Great Lakes and Hudson Lowlands
Lake Erie Lowland
Northern Appalachian Plateau and Uplands
North Central Hardwood Forests
Driftless Area
S. Michigan/N. Indiana Drift Plains
Northeastern Coastal Zone
Maine/New Brunswick Plains and Hills
Erie Drift Plain
Southeastern Wisconsin Till Plains
Huron/Erie Lake Plains
Central Com Belt Plains
Eastern Com Belt Plains
Northern Piedmont
Interior River Valleys and Hills
Interior Plateau
Piedmont
Southeastern Plains
Mississippi Valley Loess Plains
South Central Plains
East Central Texas Plains
Ridge and Valley
Central Appalachians
Western Allegheny Plateau
Blue Ridge
Ozark Highlands
Boston Mountains
Arkansas Valley
Ouachita Mountains
Southwestern Appalachians
Middle Atlantic Coastal Plain
Mississippi Alluvial Plain
Southern Coastal Plain
Atlantic Coastal Pine Barrens
Aspen Parkland/Northern Glaciated Plains
Lake Manitoba and Lake Agassiz Plain
Western Corn Belt Plains
Central Irregular Plains
Northwestern Glaciated Plains
Northwestern Great Plains
Nebraska Sand Hills
High Plains
Central Great Plains
Southwestern Tablelands
Flint Hills
Cross Timbers
Edwards Plateau
a
210.3
144.62
112.79
322.68
148.68
117.97
191.44
158.48
156.02
133.08
121.34
145.17
187.95
235.18
175.82
149.68
220.47
224.11
205.3
492.49
184.36
162.32
186.83
166.76
183.67
216.16
175.16
329.77
283.25
212.77
207.09
182.17
131.35
138.62
283.76
136.43
174.13
135.01
201.19
133.98
130.6
289.85
125.61
156.84
137.77
270.93
134.97
173.77
b
-0.0114
-0.0104
-0.0049
-0.0113
-0.0108
-0.0012
-0.0143
-0.0164
-0.025
-0.0037
-0.0042
-0.0058
-0.0033
-0.003
-0.0042
-0.0013
-0.0037
-0.0048
-0.0085
-0.0048
-0.0045
-0.0013
-0.0063
-0.0062
-0.0032
-0.0087
-0.0018
-0.0062
-0.004
-0.0048
-0.0071
-0.0178
-0.0029
-0.0144
-0.0463
-0.0005
-0.0042
-0.0009
-0.001
-0.0006
-0.0004
-0.0066
-0.0005
-0.0005
-0.0003
-0.0009
-0.0006
-0.001
TSSioo
65
36
25
103
37
141
46
28
18
78
46
65
191
282
135
303
217
169
85
333
136
362
99
82
190
89
317
193
261
157
103
34
93
23
23
640
131
347
673
483
636
162
507
925
1,280
1,084
523
544
TSS10
267
257
494
307
250
2,057
206
168
110
700
594
461
889
1,053
683
2,081
836
648
356
812
648
2,144
465
454
910
353
1,591
564
836
637
427
163
888
183
72
5,226
680
2,892
3,002
4,325
6,424
510
5,061
5,505
8,743
3,666
4,337
2,855

November 2009
                                                                                                  F-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
   Table F-1: TSS Subindex Curve Parameters, by Ecoregion
     ID
Ecoregion Name
                                                                 a
  b
TSS,
TSS,
    9.4.7   Texas Blackland Prairies
                                    134.23
-0.0005
    624
  5,194
    9.5.1    Western Gulf Coastal Plain
                                    124.47
-0.0025
            1,009
    9.6.1    Southern Texas Plains/Interior Plains and Hills with
   	Xerophytic Shrub and Oak Forest	
                                    166.67   -0.0003
            1,602
            9,378
November 2009
                                                                                                   F-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
Appendix G - Meta-Analysis Results
EPA used function-based benefit transfer to estimate benefits of surface water quality improvements due
to reductions in sediment runoff from construction sites under the regulatory analysis options considered
for the analysis of the regulation. The benefit function was derived using meta-analysis, following the
general approach of Johnston et al. (2005), Shrestha et al. (2007), and others, following conceptual
methods outlined by Bergstrom and Taylor (2006). The recent literature has given increasing emphasis to
the potential use of meta-analysis to conduct and inform function-based benefit transfer (Johnston et al.
2005; Bergstrom and Taylor 2006; Rosenberger and Stanley 2006; Shrestha et al. 2007). For the present
analysis, the meta-regression model was based on a model specification and data developed originally for
the 316(b) Phase II Cooling Water Intake rule, and extended to generate a benefit function more suitable
for assessing changes in ambient water quality from reducing sediment discharges. Chapter A12,
"Methods for Estimating Non-use Benefits," in the Regional Analysis document for the final Phase II rule
provides details on the original meta-analysis (USEPA 2004d); revisions are detailed below. These
revisions included adding new studies to the metadata, re-estimating the willingness-to-pay (WTP)
function to better account for ecological services potentially affected by sediments and nutrients, and
testing additional functional forms and statistical approaches.
As stated by Rosenberger and Johnston (2007, p. 1-2):
    One of the primary advantages of meta-analysis as a benefit transfer tool relates to its capacity to allow more
    appropriate adjustments of welfare measures based on patterns observed in the literature. Within a benefit transfer
    context, transfer error is often inversely related to the correspondence between a study site and a policy site among
    various dimensions (Rosenberger and Phipps 2007). The probability of finding a good fit between a single (or multiple)
    study site and a policy site, however, is usually low (Boyle and Bergstrom 1992; Spash and Vatn 2006). If, on the other
    hand, empirical studies contribute to a body of WTP estimates (i.e., metadata), and if empirical  value estimates are
    systematically related to variations in resource, study, and site characteristics, then meta-regression analysis may
    provide a viable tool for estimating a more universal transfer function with distinct advantages over unit value or other
    function-based transfer methods (Johnston et al. 2003; Rosenberger and Loomis 2000a; Rosenberger and Stanley
    2006). More specifically, Rosenberger and Phipps (2007) posit a meta-valuation function as the envelope of a set of
    empirically-defined valuation functions reported in the literature.
In the present case, EPA identified 45 valuation studies that use stated preference techniques to elicit
benefit values for water quality improvements. To examine the relative influence of study, economic, and
resource characteristics on WTP for improving surface water quality, the Agency conducted a regression-
based meta-analysis  of 115 estimates of WTP for water quality improvements, provided by the 45 original
studies. Analytic methods and model specifications follow established methods in  the published literature
(e.g., Johnston et al.  2005; Bergstrom and Taylor 2006). The estimated econometric model is used as the
basis of a function-based benefit transfer, to calculate WTP for improving water quality in waterbodies
affected by sediment runoff from construction sites.
The following discussion summarizes the results of EPA's meta-analysis of surface water valuation
studies and the use of the resulting benefit function for transfer.

G.1 Literature  Review of Water Resource Valuation Studies
As outlined in the introduction, EPA conducted a meta-analysis of water resource  valuation studies to
examine the relative influence of study, economic, and resource characteristics on  total WTP for water
quality improvements. The Agency analyzed 45 studies, published between 1981 and 2008, that applied
November 2009
                                                                                                 G-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
generally accepted, standard valuation methods to determine total (including use and nonuse) values
associated with aquatic habitat improvements. These 45 studies all used stated preference techniques to
assess WTP, but varied in other respects, including the survey administration methods used, the specific
environmental change valued, and the geographic region affected by the environmental changes. Studies
using stated preference approaches are preferred to studies using revealed preference approaches, because
they elicit total household WTP (including use and nonuse values). Revealed preference studies allow to
estimate use values only. Data from the 45 studies result in a total of 115 observations for the meta-
analysis because 30 studies provide more than one usable estimate of total WTP for aquatic habitat
improvements.
When constructing metadata for subsequent meta-analysis, analysts must determine the optimal scope of
the metadata (Rosenberger and Johnston 2007), interpreted as the exact definition of the dependent
variable in the meta-regression model, which, in turn, defines the set of source studies that can be
considered for inclusion in the metadata. The primary tradeoff is often between maintaining close
similarity among dependent variables versus including additional information (i.e., observations) in the
metadata. Similarity in dependent variable definition and study attributes within the metadata can be
important for two reasons. First, theory may dictate that certain types of estimated values are not strictly
comparable (e.g., Hicksian compensating variation from a stated preference model versus Marshallian
consumer surplus from a travel cost model). Second, model fit may be improved by narrowing the
metadata, for example to include only valuation studies that use a particular valuation approach (e.g.,
stated preference methods, as in Johnston  et al. (2005); or travel cost model estimates, as in Smith and
Kaoru (1990)). Such study selection issues may be framed in terms of a requirement that at a minimum,
studies included in metadata satisfy both commodity consistency and welfare consistency (Bergstrom and
Taylor 2006). The former implies that "the commodity (Q) being valued should be approximately the
same within and across studies" (Bergstrom and Taylor 2006, p. 353). The  latter implies that "measures
of WTP within and across studies ... should represent the same ... welfare  change measure, or ex-post
calibrations [are] made to account for theoretical differences between welfare change measures"
(Bergstrom and Taylor 2006, p. 355).
The requirement of welfare consistency implies that—outside of preference calibration approaches that
explicitly account for theoretical differences between welfare constructs (e.g., Smith et al. 2002)—meta-
analyses should not combine data representing theoretically distinct welfare measures. For example, the
ad hoc combination of stated and revealed data for meta-analysis would generally violate the condition of
welfare consistency. Although one may introduce independent variables in regression models (typically
dummy variable intercept shifters) to account for such theoretical differences, Smith and Pattanayak
(2002) argue that such methods are unlikely to represent appropriate adjustments for fundamental
differences in theoretical welfare measures.

G.1.1 Identifying Water Resource Valuation Studies
EPA identified surface water valuation studies used in the total WTP meta-analysis by conducting an in-
depth search of the economic literature. EPA used a variety of sources and  search methods to identify
relevant studies:
       >  Review of EPA's research and bibliographies dealing with non-market benefits associated
           with water quality changes
November 2009                                                                                  G-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
        >  Systematic review of recent issues of key resource economics journals (e.g., Land
           Economics, Marine Resource Economics, Journal of Environmental Economics and
           Management)
        >  Searches of online reference and abstract databases (e.g., Environmental Valuation Resource
           Inventory (EVRI), Benefits  Use Valuation Database (BUVD), AgEcon Search)
        >  Visits to home pages of authors known to have published stated preference studies and/or
           water quality valuation research
        >  Searches of Web sites of agricultural and resource economics departments at several colleges
           and universities
        >  Searches of Web sites of organizations and agencies known to publish environmental and
           resource economics valuation research (e.g., Resources for the Future (RFF), National Center
           for Environmental Economics (NCEE)).
From this review, EPA identified approximately 300 surface water valuation studies that were potentially
relevant for this analysis and compiled a bibliographic database to organize the literature review process.
Sixty-seven of these studies met the criteria identified for inclusion in the meta-analysis.23 These criteria
were designed to ensure both commodity and welfare consistency as noted above, and include:
        >  Specific amenity valued: Selected studies were limited to those in which the environmental
           quality change being valued affects ecological services provided by surface waterbodies,
           including aquatic life support, recreational activities (such as fishing, boating, and
           swimming), and nonuse value
        >  Values estimated: Selected studies were limited to those that used stated preference
           techniques to elicit household WTP.
        >  Study location: Selected studies were limited to those that surveyed U.S. and Canadian
           populations to value resources
        >  Research methods: Selected studies were limited to those that applied research methods
           supported by journal literature.
The Agency compiled extensive information from the 67 selected studies. Of these studies, 45 were
utilized in the model estimation. Reasons for the difference between the total number of studies in the
final metadata (67) and studies represented  in the final model (45) include unavailability of information
for certain key regressors for all studies.
The tradeoff between the number of regressors or independent variables that may be included in a meta-
regression analysis (K) and the number of studies that are included in the metadata (TV) is a fundamental
tradeoff in most meta-analyses in the valuation literature (Moeltner et al. 2007). That is, if a study
considered for inclusion in the metadata does not provide information for a certain regressor that analysts
might wish to include in the meta-regression, analysts must generally choose between omitting the
regressor from the meta-regression or omitting the study from the metadata.  As a result, researchers
wishing to increase the number of explanatory variables in meta-regression models (increasing K) often
do so at the cost of reductions in the number of studies or observations in the metadata (reducing N).
23   The remaining studies were either earlier unpublished versions or slightly modified versions of the included studies, focused
    on water resources outside of the United States and Canada, used secondary research methods, or valued environmental
    quality changes that were not directly linked to changes in water quality (e.g., change in recreational catch rates).
November 2009                                                                                   G-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
Conversely, increases in the number of observations in meta-regression models are sometimes only
possible if one reduces the number of independent variables in the model. Hence, there is a tradeoff
between the quantity of information in the metadata (i.e., the number of observations or studies) and the
possible risk of omitted variables bias due to the omission of influential regressors. See Moeltner et al.
(2007) for additional discussion of this "TV versus K" tradeoff in meta-analysis.
The complete data set used in the meta-analysis is provided in the docket for the regulation (DCN 6-
7900), and includes the following information:
       >  Full study citation
       >  Study location
       >  Sample data and description (e.g., size, response rate, income)
       >  Resource characteristics (e.g., affected waterbody type, recreational uses, baseline quality)
       >  Description of environmental quality change, including geographic scale, affected species,
            and affected recreational uses (e.g., water quality change from fishable to beatable)
       >  Quantitative measure of environmental quality change measured in terms of improvements in
            Water Quality Index (WQI)24 and/or percentage reduction in pollutant concentration
       >  Study WTP values updated to 2007 dollars
       >  WTP estimation characteristics (i.e., parametric versus non-parametric, inclusion of protest
            bids and outlier bids, WTP description).

G.1.2 Description of Studies Selected for Total WTP Meta-Analysis
The 45 studies that EPA used in the total WTP meta-analysis were conducted between 1981 and 2008,
and applied  standard, generally accepted stated preference valuation methods to assess WTP. Studies
were excluded if they did not conform to general tenets of economic theory, or if they applied methods
not generally accepted in the literature.
All selected studies focus on environmental quality changes that affect surface water resources in the
United States. Beyond this general similarity, the  studies vary in several respects. Differences include the
specific environmental change valued, scale of environmental improvement, geographic  region affected
by environmental changes, types of values estimated, survey administration methods, demographics of
the survey sample, and statistical methods employed. The 45 studies include 25 journal articles, 6 reports,
5 Ph.D. dissertations, 7 academic or staff papers, 1 book, and 1 master's thesis.
The 45 studies selected for the meta-analysis provided 115 observations in the final data set because
multiple estimates of WTP were available from 30 studies. The availability of multiple observations from
single studies is common in meta-analyses of this type (e.g., Bateman and Jones 2003; Johnston et al.
2005). Some of the characteristics that allowed multiple  observations to be derived from a single study
include the extent of the amenity change, the respondent population type, elicitation  method(s),
    Additional details on the WQI and the use of the WQI in survey instruments are provided by McClelland (1974), Vaughan
    (1986), and Mitchell and Carson (1989, p. 342). This index is linked to specific pollutant levels, which in turn are linked to
    presence of aquatic species and suitability for particular recreational uses. The WQI allows the use of objective water quality
    parameters (e.g., dissolved oxygen concentrations) to characterize ecosystem services or uses provided by a given
    waterbody.
November 2009                                                                                    G-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
waterbody type, number of waterbodies affected, recreational activities affected by the quality change,
and species affected by the quality change. These variations are often due to experimental design driven
by the key research questions or hypotheses. Table G-l lists key study and resource characteristics and
indicates the number of observations derived from each study.
Surveys in 26 studies were administered by mail; 9 studies collected information through personal
interviews in the home, onsite, or in a centralized location, 9 surveys were conducted by telephone, and 1
conducted by computer administration. Study sample sizes range from 96 to 4,033 responses.

Table G-1: Selected Summary Information for Studies
Author(s) and Year
Aiken(1985)
Anderson and
Edwards (1986)
Azevedoetal. (2001)
Bockstael et al.
(1988)
Bockstael et al.
(1989)
Breffleetal. (1999)
Cameron (1988)
Cameron and
Huppert(1989)
Carson and Mitchell
(1993)
Carson etal. (1994)
Clonts and Malone
(1990)
Croke etal. (1986-
87)
DeZoysa(1995)
Desvousges et al.
(1987)
Hayes etal. (1992)
Herriges and Shogren
(1996)
Kite (2002)
Huang etal. (1997)
Hushak and Bielen
(1999)
Kaoru(1993)
Lant and Roberts
(1990)
Lant and Tobin
(1989)
Lichtkoppler and
Elaine (1999)
Lindsey(1994)
Lipton (2004)
Loomis(1996)
Loomis et al. (2000)
Lyke(1993)
Observations
1
1
5
1
2
2
1
1
4
2
3
9
2
12
2
2
2
2
2
1
3
9
1
8
1
1
2
2
State
CO
RI
IA
DC,MD,
VA
MD
WI
TX
CA
National
CA
AL
IL
OH
PA
RI
IA
MS
NC
OH, MI
MA
IA,IL
IA,IL
OH
MD
MD
WA
CO
WI
Waterbody
Type
all freshwater
salt
pond/marshes
lake
estuary
estuary
estuary
Estuary
river/stream
multiple
estuary
river/stream
river/stream
river and lake
river/stream
estuary
lake
river/stream
estuary
river/stream
salt
pond/marshes
river/stream
river/stream
multiple
estuary
estuary
river/stream
river/stream
lake
Type of Water Quality
Improvement
general water quality
general water quality
nutrients
general water quality;
phosphorus and nitrogen
general water quality
general water quality
general water quality
wildlife habitat
general water quality
DDT and PCBs
general water quality
general water quality
sediment and nutrients;
wildlife habitat
general water quality
general water quality
general water quality
general water quality
general water quality
general water quality
fecal coliform
sediment
general water quality
PCBs and general water
quality
nutrients
general water quality
general water quality
general water quality
general water quality
Affected
Recreational Uses
fishing
fishing and swimming
fishing and swimming
swimming, beach,
boating, fishing,
outings
swimming
fishing
fishing
game fishing
boating; fishing;
swimming
fishing
multiple uses
multiple uses
multiple uses
multiple uses
swimming; fishing
boating and fishing
multiple uses
fishing
multiple uses
fishing
boating, fishing, and
swimming
boating, fishing
all recreational uses
multiple uses
boating
fishing
multiple uses
fishing
November 2009
                                                                                            G-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table G-1: Selected Summary Information for Studies
Author(s) and Year
Matthews et al.
(1999)
Olsenetal. (1991)
Opaluchetal. (1998)
Roberts and Leitch
(1997)
Roweetal. (1985)
Sanders etal. (1990)
Schulzeetal. (1995)
Shrestha and
Alavalapati (2004)
Stumborg et al.
(2001)
Sutherland and
Walsh (1985)
Viscusi et al. (2008)
Welle (1986)
Wey(1990)
Whitehead et al.
(2002)
Whitehead and
Groothuis(1992)
Whitehead et al.
(1995)
Whittington et al.
(1994)
Observations
2
3
1
1
1
4
2
2
2
1
2
6
2
1
3
2
1
State
MN
ID, MT,
OR,WA
NY
MN, SD
CO
CO
MT
FL
WI
MT
National
MN
RI
NC
NC
NC
TX
Waterbody
Type
river/stream
river/stream
estuary
lake
river/stream
river/stream
river and lake
multiple
lake
river and lake
river and lake
all freshwater
salt pond/
marshes
river/stream
river/stream
estuary
estuary
Type of Water Quality
Improvement
phosphorus
wildlife habitat
general water quality
general water quality
general water quality
general water quality
hazardous pollutants
phosphorus and wildlife
habitat
phosphorus
general water quality
general water quality
acid rain
general water quality
general water quality
sediment and nutrients
general water quality
heavy metals and pesticides
Affected
Recreational Uses
boating and fishing
fishing
shellfishing
multiple uses
boating, fishing, and
swimming
swimming
boating, fishing, and
swimming
multiple uses
multiple uses
swimming
multiple uses
game fishing and
wildlife viewing
Other
fishing, boating,
swimming
multiple uses
boating, fishing, and
swimming
multiple uses
The Agency's review of the relevant economic literature showed that available surface water valuation
studies focus primarily on general water quality rather than specific pollutants or changes. Even in cases
in which specific pollutants are the primary policy issue, the stated preference surveys from which
welfare estimates are derived often characterize water quality changes only in general (i.e., non-pollutant
specific) terms. Hence, the associated welfare measures are conditioned on this general description. Of
the 45 studies, 26 presented only WTP values for changes in general water quality (approximately 60
percent). Of the studies that did address specific changes, eight specified nutrients and/or sediment, four
addressed hazardous pollutants including heavy metals and pesticides, three addressed wildlife habitat,
one addressed acid deposition, one addressed fecal coliform bacteria, one presented values for both
changes in general water quality and nutrient reductions, and one presented values for both changes in
general water quality and wildlife  habitat. Preliminary model estimates showed no evidence that the type
of pollutant considered had a statistically significant influence on WTP across and within studies. For this
reason, EPA used a standardized scale to define both the baseline water quality and the water quality
change valued in the original study. Additional details are provided below.
From these 45 studies, the Agency compiled a data set for the meta-analysis of total WTP values. EPA
specified a regression model based on these data to estimate a range of total household benefits for
surface water and aquatic habitat improvements. General empirical methods follow those outlined by
Johnston et al. (2005), following standard approaches in the meta-analysis literature. The model and
results are described in the next section.
November 2009
                                                                                              G-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
G.2 Total WTP Meta-Analysis Regression Model and Results
EPA estimated both trans-log and semi-log meta-regression models based on 115 WTP estimates for
improvements in water resources, derived from 45 original studies.25 These metadata, the model
specification, model results, and interpretation of the results are described in the following sections. EPA,
however, notes that only the trans-log model is used in the analysis of benefits  from reduced sediment
discharges. The alternative specification (semi-log) is presented for comparative purposes only.
In a frequently cited work, Glass (1976) characterizes meta-analysis as "the statistical analysis of a large
collection of results for individual studies for the purposes of integrating the findings. It provides a
rigorous alternative to the casual, narrative discussion of research studies which is commonly used to
make some sense of the rapidly expanding research literature" (p. 3; cited in Poe et al. 2001, p. 138).
Meta-analysis is being increasingly explored as a potential means to estimate resource values in cases
where original targeted research is impractical, or as a means to reveal systematic components of WTP
(e.g.,  Smith and Osborne 1996; Santos 1998; Rosenberger and Loomis 2000a;  Poe et al. 2001; Woodward
and Wui 2001; Bateman and Jones 2003). While the literature often urges caution in the use and
interpretation of benefit transfers for direct policy application (e.g., Desvousges et al. 1998; Poe et al.
2001; Navrud and Ready 2007), such methods are  "widely used in the United States by government
agencies to facilitate benefit-cost analysis of public policies and projects affecting natural resources"
(Bergstrom and De Civita 1999). Transfers based on meta-analysis are likewise common in both the
United States and Canada (Bergstrom and De Civita 1999; Bergstrom and Taylor 2006).
Depending on the suitability of available data, a meta-analysis can provide a superior alternative to the
calculation and use of a simple arithmetic mean WTP over the available observations, as it allows
estimation of the systematic influence of study, economic, and natural resource attributes  on WTP
(USEPA 2000b; Rosenberger and Phipps 2007; Shrestha et al. 2007). The primary advantage of a
regression-based (statistical) approach is that it accounts for differences among study characteristics that
may contribute to changes in WTP, to the extent permitted by available data (Johnston et al. 2005;
Rosenberger and Phipps 2007). An additional advantage is that meta-analysis can reveal systematic
factors influencing WTP, allowing analysts to assess whether, for example, WTP estimates are (on
average) sensitive to scope (Smith and Osborne  1996).
There is, however, some controversy regarding whether regression-based meta-analyses should be used
for direct benefit transfer. Many contemporary sources in the literature note the potential ability of
regression-based meta-analyses to generate benefit functions better able to adjust and forecast benefits at
policy sites in question, and either explicitly  or implicitly favor the use of meta-analysis over alternative
benefit transfer approaches  (e.g., Johnston et al.  2005; Bergstrom and Taylor 2006; Moeltner et al. 2007;
Rosenberger and Johnston 2007; Rosenberger and Phipps 2007; Shrestha et al. 2007). EPA (USEPA
2000b) characterizes meta-analysis as "the most rigorous" benefit transfer method.  In contrast, the EPA
Science Advisory (2007) Board's "Advisory on EPA's Issues in Valuing  Mortality Risk Reduction"
recommends against the use of regression-based meta-analysis for VSL (value  of statistical life) transfers,
and other authors advise caution in such uses (Navrud and Ready 2007). The primary disagreement is
whether it is appropriate to use meta-analysis results as a reduced form model to estimate  benefits, and
    In its analysis of nonuse benefits for the final 316(b) Phase II rule, EPA also specified trans- and semi-log regression models
    similar to the models estimated for Phase II discussed in this section. See Chapter A12, "Methods for Estimating Non-use
    Benefits," in the Regional Analysis document for further details regarding both the log-log and semi-log regression models
    estimated in EPA's analysis of nonuse benefits for the final Phase II rule (USEPA 2004d);
    http://www.epa. gov/ost/316b/casestudv/final.htm).
November 2009                                                                                   G-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
whether the empirical ability of meta-analysis in many cases to generate benefit transfers with reduced
transfer errors offsets the lack of an underlying theoretical model to "calibrate" benefit estimates across
studies (cf Smith and Pattanayak 2002; Smith et al. 2002). While the Agency recognizes this ongoing
controversy, it notes that there are a large number of practitioners and publications supporting the use of
regression-based meta-analysis for benefit transfer. It also removes the element of subjective judgment
associated with selecting a single study or value for benefit transfer.  Hence, the following model is
presented as a means to provide a benefit function that capitalizes on the substantial information available
for existing water quality valuation studies, notwithstanding potential concerns voiced by some regarding
the use of meta-analyses for such purposes.

G.2.1 Metadata Total WTP Regression Model
Meta-analysis is largely an empirical, data-driven process, but one in which variable and model selection
is guided by theory (Bergstrom and Taylor 2006).  Given a reliance on information available from the
underlying studies that comprise the metadata, meta-analysis models most often represent a middle
ground between model specifications that would be most theoretically appropriate and those
specifications that are possible given available data. Smith and Osborne (1996), Rosenberger and Loomis
(2000a), Poe et al. (2001), Bateman and Jones (2003), Dalhuisen et al. (2003), Johnston et al. (2005,
2006), Bergstrom and Taylor (2006), Moeltner et al. (2007), and others provide insight into the mechanics
of specifying and estimating meta-equations in resource economics applications.
Past meta-analyses have incorporated a range of different statistical methods, with none universally
accepted as superior (Johnston et al. 2005). EPA followed recent work of Bateman and Jones (2003) and
Johnston et al. (2005) in applying a multilevel model specification to the metadata, to  address potential
correlation among observations gathered from single  studies. Also following prior work (e.g., Smith and
Osborne  1996; Poe et al. 2001), EPA applied the Huber-White robust variance estimation. As described
by Smith and Osborne (1996, p. 293), "this approach treats each study as the equivalent of a sample
cluster with the potential for heteroskedasticity.. .across clusters." Weighted models are avoided
following the arguments of Bateman  and  Jones (2003).26
To guide development of the model and variable specifications, EPA relied upon a set of general
principles.  These principles are designed to help prevent excessive data manipulations and other factors
that may lead to misleading model  results. The general principles include, all else being equal:
       >  Fewer and simpler data transformations are preferred to  more extensive ones.
       >  In the  absence of overriding theoretical considerations, continuous variables are generally
           preferred to discrete variables derived from underlying continuous distributions.
       >  Models should attempt to capture elements of the scope  and scale of resource changes.
       >  Models should distinguish WTP associated with different types of resources and resource
           uses, particularly where relevant to the policy question at hand.
       >  Following the "weak structural utility theoretic" (WSUT) approach of Bergstrom and Taylor
           (2006, p. 352), exogenous structural constraints are avoided to afford the flexibility necessary
26   For comparison, models were also estimated using ordinary least squares (OLS) with robust variance estimation, weighted
    least squares with robust variance estimation, and multilevel models with standard (non-robust) variance estimation. None
    of these models outperformed the illustrated model in terms of overall model significance and fit, and statistical significance
    of individual coefficients (see Section G.2.2 for further details concerning the specification of the model).
November 2009                                                                                   G-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
           to appropriately model empirical patterns that may not necessarily flow from an underlying
           theoretical modeling structure. The dependent variable in the meta-analysis is the natural
           logarithm of estimated household WTP for water quality improvements in aquatic habitat, as
           reported in each original study. For this analysis, original study values were adjusted to
           2007$ based on the relative change in Consumer Price Index (CPI) from the study year to
           2007. Total WTP over the sample ranged from $5.33 to $502.70, with a mean value of
           $83.09.
As noted above, two model specifications are estimated (cf Johnston et al. 2005). For the first
specification, all right-hand-side variables are linear, resulting in a standard semi-log functional form. The
second specification is identical to the semi-log model, except for the specification  of the explanatory
variable measuring water quality change and baseline as natural logs. This results in a trans-log functional
form, also common in empirical applications (Johnston et al. 2001, 2005). Both the semi-log and trans-log
models have advantages related to (1) their fit to the data, (2) the intuitive results that are provided, and
(3) their common  use in  the empirical valuation and meta-analysis literature (e.g., Smith and Osborne
1996; Santos 1998; Johnston et al. 2001, 2005, 2006). The trans-log model, however, has the additional
structural advantage that estimated WTP is necessarily zero when water quality change is also zero (cf.
Johnston et al. 2001)—a property suggested by theory that analysts may wish to weight against model fit
considerations when choosing a model for benefit estimation. While linear forms are also common in this
literature (Rosenberger and Loomis 2000a, 2000b; Poe et al. 2001; Bateman and Jones 2003),
specifications requiring more intensive data transformations (e.g., Box-Cox, log-log) are less common.
As noted in the preceding section,  the metadata include independent variables characterizing specific
details of the resource(s) valued  such as the baseline resource conditions; the extent of resource
improvements and whether they occur in estuarine or freshwater; the geographic region and scale of
resource  improvements (e.g., the number of waterbodies); resource characteristics (e.g., baseline
conditions, the extent of water quality change, and ecological services affected by resource
improvements); characteristics of surveyed populations (e.g., users, nonusers); and  other specific details
of each study. For ease of exposition, these variables are categorized into those characterizing (1) study
and methodology, (2) surveyed populations, (3) geographic  region and scale, and (4) resource
improvements. Attributes included within each category are summarized  below.
Study and methodology variables characterize such features as:
       >  The year in which a study was conducted
       >  The payment vehicle and elicitation format (e.g., discrete choice versus open-ended,
           voluntary versus non-voluntary, interview versus mail versus phone)
       >  WTP  estimation methods and conventions (e.g., approaches to protest and outlier bids, use of
           parametric versus non-parametric statistical methods, estimation of mean or median WTP, the
           use of annual or lump-sum payments)
       >  Whether the original survey represented water quality changes using the WQI.
Surveyed populations variables characterize  such features as:
       >  The average income of respondents
       >  Whether the survey specifically targeted nonusers.
Geographic region and scale variables characterize such features as:
November 2009                                                                                 G-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
        >  The number of waterbodies affected by the policy
        >  Whether the study considered water quality improvements in all waterbodies in a region
        >  The geographic area of the country in which the study was conducted.
Resource improvement variables characterize such features as:
        >  The extent of water quality change estimated as a difference between the baseline and post-
           change water quality index
        >  Baseline water quality index
        >  Those studies for which recreational uses such as fishing are specifically noted in the survey
        >  Aquatic species affected by resource improvements (e.g., game fish and shellfish)
        >  Those studies identifying large increases in fish populations (i.e., greater than 50 percent).
Although the interpretation and calculation of most independent variables requires little explanation, a
few variables require additional detail. These include the variables characterizing surface water quality
and its measurement. Many (23) observations in the metadata characterize quality changes using variants
of the WQI (e.g., Mitchell and Carson 1989). This scale is linked to specific pollutant levels, which, in
turn, are linked to the presence of aquatic species and recreational uses. However, some observations
provide water quality measures using other, primarily descriptive, means that differ from the WQI.
To allow consistent comparisons of water quality change using a single scale, EPA mapped all water
quality measures to the original WQI developed by McClelland (1974). WQI values were therefore
developed for those studies that did not originally use this index. This scale was chosen for two reasons:
        >  WQI values are linked to specific pollutant levels including sediment concentrations, which,
           in turn, are linked to the presence of aquatic species and suitability for particular recreational
           uses. Therefore, the WQI can be used to link water quality changes from reduced sediment
           runoff to effects on human uses and support for aquatic species habitat. Chapter 10 of this
           report provides detail on application of the WQI to estimating resource improvements from
           reduced sediment runoff from construction sites.
        >  A large number of the original  studies in the metadata included WQI measures as "native"
           components of the original surveys. Hence, for these studies, no additional transformations
           were required.
While not all studies in the metadata included the WQI as a native survey component, in most cases the
descriptions of water quality (present in the studies that did not apply the WQI) rendered mapping of
water quality measures to the WQI straightforward. In cases where baseline and improved (or declined)
water quality was not defined by suitability for recreational activities (e.g., boating, fishing, and
swimming) or corresponding qualitative measures (e.g., poor, fair, good), EPA used descriptive
information available from studies  (e.g., amount/indication of the presence of specific pollutants,
historical decline of the quality of the resource) to approximate the baseline level of water quality and the
magnitude of the change.27 For studies that valued discrete changes in the size of species populations,
EPA characterized the baseline quality based on the current presence and prevalence of the species at
    For example, a study by Huang et al. (1997) described current water quality as degraded from 1981 levels in terms of
    reduced fish catches (60 percent) and reduced number of open shellfish beds (25 percent). However, because the water
    resource was still supporting recreational fishery, the baseline water quality was set to "fishable" on the WQI.
November 2009                                                                                   G-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
hand, and assumed population increases to correspond to modest increases in water quality in order to be
conservative.28 To account for the uncertainty involved in mapping those studies that are not based on the
WQI, EPA introduced the binary variable WQI, which indicates those studies in which WQI
measurements were an original component of the survey instrument. This approach is based largely on the
published methods of Johnston et al. (2005), drawn from prior Agency work for the 316(b) Phase II
Cooling Water Intake rule.
Variables incorporated in the final model are listed and described in Table G-2.

Table G-2: Variables and Descriptive Statistics for the Total WTP Regression Model
Variable
ln_WTP
year_indx
discrete
volunt
mail
lump_sum
nonparam
quality_ch
lnquality_ch'
WQI
non-reviewed
outlier bids
median_WTP
income
Description
Natural log of WTP for specified resource improvements.
Year in which the study was conducted, converted to an index by
subtracting 1980.
Binary (dummy) variable indicating that WTP was estimated
using a discrete choice survey instrument.
Binary (dummy) variable indicating that WTP was estimated
using a payment vehicle described as voluntary as opposed to, for
example, property taxes.
Binary (dummy) variable indicating that the survey was
conducted through mail (default value for this dummy is a phone
survey).
Binary (dummy) variable indicating that payments were to occur
on something other than an annual basis over a long period of
time, such as property taxes. For example, some studies specified
that payments would occur over a five-year period.
Binary (dummy) variable indicating that WTP was estimated
using non-parametric methods.
The change in mean water quality, specified on the WQI
(McClelland 1974; Mitchell and Carson 1989). Defined as the
difference between baseline and post-compliance quality. Where
the original study (survey) did not use the WQI, EPA mapped
water quality descriptions to analogous levels on the WQI to
derive water quality change (see text).
Natural log of the change in mean water quality (quality ch),
specified on the WQI (McClelland 1 974; Mitchell and Carson
1989).
Binary (dummy) variable indicating that the original survey
reported resource changes using a standard WQI.
Binary (dummy) variable indicating that the study was not
published in a peer-reviewed journal.
Binary (dummy) variable indicating that outlier bids were
excluded when estimating WTP.
Binary (dummy) variable indicating that the study reported
median, not mean, WTP.
Mean income of survey respondents, either as reported by the
original survey or calculated by EPA based on U.S. Census
Bureau averages for the original surveyed region.
Units and
Measurement
Natural log of dollars
(Range:2.12to6.22)
Year index (Range: 1
to 28)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
WQI units
(Range: 2.5 to 65)
WQI units
(Range: 0.92 to 4. 17)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Dollars
(Range: 34,955 to
158,347)
Mean
(Std. Dev.)
4.42
(0.75)
9.68
(6.42)
0.28
(0.45)
0.05
(0.22)
0.48
(0.52)
0.13
(0.34)
0.50
(0.50)
21.3
(10.4)
2.91
(0.59)
0.34
(0.48)
0.32
(0.47)
0.94
(0.24)
0.04
(0.20)
5,7049.59
(13,946.64)
    For example, a study by Lyke (1993) describes the baseline conditions as follows: (1) "there are no naturally reproducing
    lake trout in Lake Michigan; all lake trout found there are from hatcheries." (2) "Lake Superior stocks of self-reproducing
    lake trout were much reduced, but not wiped out, and both natural and hatchery-raised lake trout are found there." These
    baseline conditions correspond to the "game-fishable" level on the WQI. The study estimates WTP for restoring natural
    populations of lake trout to the Wisconsin Great Lakes. Therefore, the expected change that will occur within the "game-
    fishable" category is likely to be small.
November 2009
                                                                                                  G-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
Table G-2: Variables and Descriptive Statistics for the Total WTP Regression Model
Variable
nonusers
single^ver2
single_lake3
regional_fresh
multiple_river
salt_pond
num riv_pond
mr
mp
allmult
nonspec
fish use
fishplus
baseline
Inbase
Description
Binary (dummy) variable indicating that the survey was
implemented over a population of nonusers (default category for
this dummy is a survey of any population that includes users).
Binary (dummy) variable indicating that resource change
explicitly took place over a single river (default is a change in an
estuary or that takes place on a national scale).
Binary (dummy) variable indicating that resource change
explicitly took place over a single lake (default is a change in an
estuary or that takes place on a national scale).
Binary (dummy) variable indicating that resource change
explicitly took place over an entire region such as a state (default
is a change in an estuary or a change that takes place on a national
scale).
Binary (dummy) variable indicating that resource change
explicitly took place over multiple rivers (default is a change in an
estuary or that takes place on a national scale).
Binary (dummy) variable indicating that resource change
explicitly took place over multiple salt ponds (default is a change
in an estuary or that takes place on a national scale).
Number of rivers or salt ponds affected by policy; if unspecified
num riv_pond = 0. (In the present data, only studies addressing
rivers and lakes specified >1 number of waterbodies. All others
specified either 1 waterbody, or the number was unspecified.)
Binary (dummy) variable indicating that the survey included
respondents from more than one of the EPA regions.
Binary (dummy) variable indicating that the survey included
respondents from the Mountain Plain region.4
Binary (dummy) variable indicating that either all or multiple
aquatic species are affected by the resource change.
Binary (dummy) variable indicating that the study did not specify
what species would be affected by water quality improvements.
Binary (dummy) variable identifying studies in which changes in
fishing use are specifically noted in the survey.
Binary (dummy) variable identifying studies in which a fish
population or harvest change of 50 percent or greater is reported
in the survey.
Baseline water quality, specified on the WQI.
Natural log of baseline water quality, specified on the WQI.
Units and
Measurement
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Number of specified
rivers or ponds
(Range: 0 to 15)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
WQI units (Range:
1.61 to 5.2)
WQI units (Range: 5
to 70)
Mean
(Std. Dev.)
0.09
(0.29)
0.23
(0.42)
0.08
(0.28)
0.36
(0.48)
0.06
(0.24)
0.03
(0.18)
1.01
(2.98)
0.08
(0.26)
0.02
(0.13)
0.18
(0.39)
0.42
(0.50)
0.56
(0.50)
0.08
(0.28)
39.79
(20.37)
3.46
(0.80)
1 The variable quality_ch is defined earlier in this table as the difference between baseline and post-compliance quality, specified on the WQI
(Mitchell and Carson 1989).
2 Examples of rivers and streams considered in the studies include the Columbia, Potomac, Elwha, Eagle, and Tar-Pamlico rivers.
3 Includes one study that focused on a segment of the Lake Erie shoreline.
4 The Mountain Plain region includes the following states: Colorado, Iowa, Kansas, Missouri, Montana, Nebraska, North Dakota, South
Dakota, Utah, Wyoming.
G.2.2 Total WTP Regression Model and Results
As noted above, EPA estimated the meta-analysis regression using a multilevel, random-effects
specification. This model follows the general approaches of Bateman and Jones (2003) and Johnston et al.
(2005), among others. Multilevel (or hierarchical) models may be estimated as either random-effects or
random-coefficients models, and are described in detail elsewhere (Singer 1998). The fundamental
distinction between these and classical linear models is the two-part modeling of the equation error to
account for hierarchical data. Here, the metadata are comprised of multiple observations per study, and
November 2009
                                                                                                  G-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
there is a corresponding possibility of correlated errors among observations that share a common study or
author.
The common approach to modeling such potential correlation is to divide the residual variance of
estimates into two parts, a random error that is independently and identically distributed across all studies
and for each observation,  and a random effect that represents systematic variation related to each study.
The model is estimated as a two-level hierarchy, with level one corresponding to WTP estimates
(individual observations), and level two corresponding to individual studies. The random effect may be
interpreted as a deviation  from the mean equation intercept associated with individual studies (Bateman
and Jones 2003). The model is estimated using a maximum likelihood estimator, assuming that random
effects show a multivariate normal distribution. Following Bateman and Jones (2003), observations are
unweighted. Covariances  are obtained using the Huber-White covariance estimator (Smith and Osborne
1996). Random-effects models  such as the multilevel model applied here are becoming increasingly
standard in resource economics applications, and are estimable using a variety of readily available
software packages.

A Note on Model Specification
As noted above, EPA considered two functional forms in this analysis: semi-log and trans-log. In both
cases, the  dependent variable is the natural log of estimated household WTP for surface water quality
improvement, as shown in Table G-2. For Model One, all right-hand-side variables are linear, resulting in
a semi-log functional form common in meta-analysis (e.g., Smith and Osborne 1996; Johnston et al.
2003). While linear forms are also common (Rosenberger and Loomis 2000a, 2000b; Poe  et al. 2001;
Bateman and Jones 2003), the semi-log and trans-log forms were chosen based on statistical performance
and ability to capture curvature in the valuation function, and because they allow independent variables to
influence WTP in a multiplicative rather than additive manner.
Model Two is a trans-log  model, identical to the semi-log specification save for the inclusion of water
quality measures (baseline and  quality_ch) as natural logarithms. This form—common in the hedonic
modeling literature (Johnston et al. 2001) and illustrated by Johnston et al. (2005) within the meta-
analysis literature—shares many advantages of the semi-log functional form, but also incorporates the
desirable quality that WTP is constrained to zero when quality change is also equal to zero.
Following standard econometric practice and the "weak structural utility theory" approach to  meta-
analysis summarized above (Bergstrom and Taylor 2006), the final models are specified based on
guidance from theory and prior literature. For example, Arrow et al. (1993) made a fundamental
distinction between discrete choice and open-ended payment mechanisms (such as iterative bidding and
payment cards). Hence, this distinction is made in the final model (i.e., including the variable
discrete_ch). Similarly, other "survey methodology" variables in the model were chosen based on
theoretical considerations and prior findings in the literature (e.g., voluntary versus mandatory payment
vehicles, parametric versus non-parametric, treatment of protest and outlier bids, use of mean versus
median WTP). Also included are variables characterizing the scope and scale of the resource change,
based on theoretical expectations that such factors should be relevant to welfare estimates.
Few variables were excluded solely because of lack of statistical significance. Individual variables were
only excluded if they could not be  shown to be statistically significant in any version of the model
(restricted or unrestricted), and  there was no overriding rationale for retaining the variable in the model.
For example, variables distinguishing different types of discrete choice instruments (e.g., conjoint versus
dichotomous choice) added no significant explanatory power to the model (p = 0.58).
November 2009                                                                                G-13

-------
Environmental Impact and Benefits Assessment for the C&D Category
It is important to note that although empirical considerations play a role in model development, certain
variables were retained in the model for theoretical reasons, even if significance levels were low. Such
specification of meta-analysis models using a combination of theoretical guidance and empirical
considerations is standard in modeling efforts (Bergstrom and Taylor 2006).
Table G-3 presents results of the total WTP regression model.

Table G-3: Estimated Multilevel Model Results for the Trans-log and Semi-log Total
WTP Regression Models: WTP for Aquatic Habitat Improvements
Variable
intercept
year indx
discrete
volunt
nonparam
income
WQI
outlier bids
single river
single lake
multiple river
regional fresh
salt_pond
num riv_pond
mr
mp
nonusers
allmult
nonspec
quality ch (lnquality_ch
in trans-log model)
fish use
fishplus
baseline (Inbase in
trans-log model)
mail
lump sum
non reviewed
median wtp
-2 Log Likelihood
Trans-log Model
Parameter Estimate
5.7109 3
-0.08043 3
-0.1248
-1.32333
-0.66983
2.698 x 10'6
-0.3275
-0.88373
-0.4279 3
-0.06316
-1.47523
0.1588
0.98493
0.11733
-0.88463
1.63373
-0.40363
-0.37282
-0.40422
0.40651
-0.33172
0.44322
0.02610
-0.2013
0.55692
-0.27181
-0.53583
133.93
Standard Error
0.9352
0.01482
0.2230
0.1653
0.1434
3.9 xlO'6
0.2692
0.2855
0.1412
0.2386
0.3540
0.1505
0.3580
0. 02806
0.1832
0.2980
0.1314
0.1644
0.1731
0.1488
0.1291
0.1820
0.1183
0.1466
0.2387
0.1625
0.1875

Semi-log Model
Parameter Estimate
6.0946 3
-0.06707 3
-0.1696
-1.30493
-0.68923
-1.16X10'7
-0.363 11
-0.78293
-0.27242
-0.1268
-1.50543
0.22191
1.13573
0.11453
-0.79323
1.51683
-0.445 13
-0.40443
-0.2988
0.032083
-0.44803
0.40173
0.005205
-0.30733
0.71883
-0.37443
-0.46752
115.52
Standard Error
0.5127
0.01430
0.1735
0.1661
0.1164
3.238 xlO'6
0.2096
0.2164
0.1222
0.2223
0.3184
0.1168
0.3142
0.02510
0.1343
0.2574
0.1301
0.1266
0.1351
0.006337
0.1034
0.1217
0.003739
0.1155
0.1873
0.1234
0.1898

Covariance Factors:
Study Level (ou)
Residual (o e)
0
0.18763


0
0.15993


1 Significant at the 0.10 level.
2 Significant at the 0.05 level.
3 Significant at the 0.01 level.
G.2.3 Interpretation of Total WTP Regression Analysis Results
Regression results reveal strong systematic elements influencing WTP. The analysis finds both
statistically significant and intuitive patterns that influence WTP for surface water quality improvements.
In general, the statistical fit of the three estimated equations is good; model results suggest a considerable
November 2009
                                                                                           G-14

-------
Environmental Impact and Benefits Assessment for the C&D Category
systematic component of WTP variation that allows forecasting of WTP based on site and study
characteristics. Likelihood ratio tests (Table G-3) show that model variables are jointly significant at the
p<0.01 level for the trans-log model and the p < 0.05 level for the semi-log model. In both models, the
majority of independent variables are statistically significant at p<0.05, with most statistically significant
at p<0.01  (Table G-3). Signs of significant parameter estimates generally correspond with intuition,
where prior expectations exist. As shown in Table G-3, the random effect is statistically insignificant (i.e.,
study level covariance factors are zero). Considering these factors, the statistical performance of both
models compare favorably to prior meta-analyses in the valuation literature.
Despite differences in the functional form of the two models, statistical results are robust across models.
In most cases, coefficient magnitudes and standard errors vary to only a small degree. Measures of
equation fit are  similar, and both  models are significant at the p<0.05 level or better. Such results mirror
those of Johnston et al. (2005), whose earlier meta-analysis of WTP for water quality improvements finds
a high degree of robustness to changes in model specification.
The initial discussion emphasizes results of the trans-log model. Although policy implications of both
model specifications are nearly identical for moderate to relatively large water quality improvements
(e.g., more than 5 percent increase in WQI), the trans-log model provides more accurate WTP estimates
when the expected water quality change is very small (e.g., less than 1 percent increase in WQI).
One  of the primary means to assess the validity of benefit transfers—and the only one that may be applied
in cases wherein the true value for the study site is unknown—is value surface tests (Bergstrom and  De
Civita 1999). These tests involve assessments of ways in which "different factors may cause values to
vary across sites, providing guidance for adjustments needed to make a valid transfer of value estimates
from the study site(s) to the policy site" (Bergstrom and De Civita 1999). Following general approaches
for such value surface assessments—which generally involve comparisons of empirical patterns found via
meta-analysis to theoretical expectations or norms—the Agency concluded that most results of the
estimated value surface suggest an appropriate benefit function. Results of these value surface
assessments are detailed in the following sections.

Resource Improvement Effects
Seven variables characterize resource improvements and uses; most are of the expected sign. The
coefficient on the quality_ch variable is positive and statistically significant (p<0.01), indicating that
larger water quality improvements generate larger WTP. This is an important result, and indicates that
WTP is sensitive to the scope of water quality improvements. The estimated model  showed that WTP
values are not sensitive to the baseline water quality from which water quality change would occur. The
estimated parameter on the variable baseline representing the baseline water quality from which water
quality change would occur is not statistically significant (p>0.1). This finding differs from that of
Johnston et al. (2005), which shows that WTP for marginal water quality improvements declines as
baseline water quality improves. Here, the value is positive but is only significant at p<0.17 in the semi-
log model. In the trans-log model, this parameter is not statistically significant. The reason for this result
is unknown, but may be related to highly valued uses that are associated with larger values on the WQI.
For example, increases beginning at higher levels on the WQI may cross thresholds allowing such highly
valued uses as swimming and drinking, such that increases at these high-quality  levels may be valued
more highly than otherwise similar changes at lower baselines, which may not allow such uses. Such
thresholds, or other non-convex preference patterns, may lead to unexpected  results for the baseline  water
quality variable (baseline).
November 2009                                                                                 G-15

-------
Environmental Impact and Benefits Assessment for the C&D Category
Both models reveal that water quality changes associated with recreational fishing uses lead to a
significant decrease in total WTP values, compared to improvements that do not affect fishing. The
variable fish_use identifies those studies that specified effects of water quality improvements on
recreational fishing (e.g., increase in catch rates). The associated parameter estimate is significant
(p<0.05) and has the negative sign. In contrast, the variable fishplus identifies those studies for which the
associated survey identified particularly large gains in fish populations or harvest rates (>50 percent). The
positive and statistically significant result (p<0.05) indicates that large gains in fish populations or
harvests are associated with statistically significant increases in total WTP. Results suggest that while
water quality improvements targeting fishing uses may not be valued particularly highly on average, very
large resultant improvements in fish populations or harvests are associated with increases in WTP, ceteris
paribus.
The variables alljnult and nonspec indicate that water quality improvements affect multiple species
(alljnulf) or unspecified species (nonspec), respectively. The default category from which these variables
allow systematic variations in WTP is a focus on particular aquatic species affected by water quality (e.g.,
shellfish or game fish). The associated coefficients are negative, indicating that WTP is lower when a
survey instrument does not specify what aquatic species would be affected (nonspec) or when all or
multiple species are affected (alljnult). The latter finding  seems to be counterintuitive at first. However,
when a survey instrument focuses on the effect of water quality on particular species, it is a likely
indication that these effects are of significant concern to the affected communities, which typically leads
to a higher WTP. That is, this result suggests that WTP is higher when water improvements can be shown
to offer targeted benefits to specific, and often high-profile, species groups—as opposed to cases in which
improvements benefit an often poorly characterized group  of species.

Geographic Region and Scale Effects
Ten binary variables characterize geographic region and scale; seven are statistically significant at
p<0.10. The default category from which these variables allow systematic variations in WTP is an
estuarine waterbody. Also included in this default are a small number of observations addressing national
level improvements. Compared to this baseline, WTP associated with rivers is lower (single_river and
multiple_river both have negative and significant values). Single_lake has a negative value, but it is not
significant. WTP for water quality gains in salt ponds (salt_pond)  is higher than for estuaries (p<0.05).
This is not surprising since water quality gains in salt ponds correspond to an increase in the number of
acres of shellfish beds.
Of particular importance for the general validity of empirical findings, the model results further suggest
that WTP is sensitive to the number ofwaterbodies under consideration and geographic scale of
improvement (regional Jresh). Of the waterbody categories distinguished above, both rivers and salt
ponds allowed variation in numbers of affected waterbodies explicitly described by the survey. This
variation is captured by the variable num_riv_pond (see Table G-3).29 The associated parameter estimate
is statistically significant (p<0.01) and indicates that WTP  increases with the number ofwaterbodies
considered. The parameter estimate on the regionaljresh variable is positive and significant (p<0.10) in
the semi-log model, indicating that large-scale regional water quality improvements lead to an increase in
WTP. These results, combined with the statistical significance of the water quality change variables noted
above, suggest that WTP values (in this case for water quality improvements) are strongly sensitive to
29   Technically, this variable is the sum of two interaction variables: (1) an interaction between multiple _river and the number
    ofwaterbodies noted in the survey (0 if unspecified) and (2) an interaction between salt_pond and the number of
    waterbodies noted in the survey (0 if unspecified).
November 2009                                                                                   G-16

-------
Environmental Impact and Benefits Assessment for the C&D Category
scope, both in terms of the number of waterbodies considered, geographic scale of improvement, and the
magnitude of water quality change. In the trans-log model, the regionaljresh variable is positive but not
statistically significant (p>0.30).
Finally, the regional indicator variables mp and mult_reg are statistically significant at p<0.01, suggesting
that there are significant differences among WTP estimates from surveys in different geographical regions
of the United States. The parameter estimate on the mult_reg variable is negative, indicating that WTP for
non-local water quality improvements (e.g., out of state) are lower compared to in-state or local resource
improvements. This is consistent with prior findings that WTP for water quality improvements declines
with the distance from the  resource (Bateman et al. 2006). The magnitude  of the Pacific Mountain (mp)
regional effect suggests that spurious or otherwise unexplained effects (e.g., the effect of specific
researchers who appear more than once in the data) may drive their overall magnitude. For example, the
size of the positive parameter estimate associated with WTP in the Pacific Mountain dummy (mp)  leads in
many cases to relatively large increases in WTP for Pacific Northwest policies. Hence, EPA believes that
particular, spurious, or unexplained aspects of studies from this region may have caused the associated
parameter estimate to have a larger-than-expected influence on WTP. Although effects of regional
dummy variables often escape simple, intuitive characterization, EPA notes that they  are often
statistically significant in meta-analysis found in the valuation literature. Similar issues are found by
Johnston et al. (2005), for example.

Surveyed Populations Effects
Only two variables, nonusers and income, are used to characterize surveyed populations. In particular, the
nonusers variable is of substantial policy relevance. The negative and strongly significant (p<.01)
parameter estimate indicates that surveys of nonusers only, who by definition have only nonuse values for
the resource improvements in question (cf Freeman 2003, p. 142), generate lower WTP values than
surveys that include users, who may have both use and nonuse values. Based on this statistically
significant result, it is possible to use this model to estimate nonuse values, interpreted as the mean WTP
values estimated by surveys of nonusers only. Such methods, however, may underestimate nonuse values
of the general population, if the nonuse values of users exceed those of nonusers (Whitehead and
Blomquist 1991a,b).
The income parameter estimate is positive in the  trans-log model, as expected, but is not statistically
significant. Such lack of statistical significance for income parameters is not uncommon in meta-analyses
found in the literature (e.g., Johnston et al. 2005).

Study and Methodology Effects
As often found in meta-analyses within the valuation and benefit transfer literature (Navrud and  Ready
2007), a variety of study and methodology effects can be shown to influence WTP for water quality
improvements. While expected, this does indicate that the methodological approach influences WTP, as
argued by Arrow et al. (1993). Of nine variables  characterizing study and methodological effects, eight
are statistically significant at p<0.10. Among these is the year in which a study was conducted (yearjndx,
a continuous variable), with later studies associated with lower WTP. This is the expected result, as the
focus of survey design overtime has often been on the reduction of survey biases that would otherwise
result in an overstatement of WTP (Arrow et al. 1993).
Model results reveal that voluntary (voluntary=\) payment vehicles (i.e., surveys that describe
hypothetical payments as voluntary) are associated with reduced WTP estimates. This result counters
common intuition and empirical findings that voluntary payment vehicles  are  associated with
November 2009                                                                                 G-17

-------
Environmental Impact and Benefits Assessment for the C&D Category
overstatements of true WTP (Carson et al. 2000). The reason for this counter-intuitive finding is
unknown, but may reflect an unwillingness among respondents to offer large voluntary payments, given
the fear that others will free-ride (Johnston et al. 2005). Reduced WTP estimates are also associated with
studies applying non-parametric methods to WTP estimation (nonparam). Survey elicitation method does
not have a strong effect in this model; studies using discrete choice formats have lower WTP values, but
this difference is not statistically significant.
Smaller WTP estimates are associated with studies that eliminate or trim outlier bids when estimating
WTP (outlier_bids=l; p<0.01). Studies that report median WTP (median_WTP; p<0.01) have lower WTP
values.
Lower WTP is associated with the use of the WQI in the original survey (WQI=\). This parameter is,
however, significant in the semi-log model only (p<0.1). As is the case with a variety of study design
variables,  there  is no necessary expectation with respect to the direction of this effect. Nonetheless, this
finding might suggest the capacity of such scales to clarify the specific magnitude and implications of
water quality change, and hence (perhaps) reduce methodological misspecification or symbolic biases that
might act to systematically inflate estimated WTP.
Survey format variables also have an effect on WTP, as might be expected. Mail has a negative and
statistically significant coefficient (p<0.01) in the semi-log model, compared to the default of telephone
surveys or interviews. This parameter is negative, but not statistically significant (p>0.17) in the trans-log
model. It may be possible that the interview and telephone survey format results in larger WTP values
either because the respondents are better able to understand the valuation scenario, or because
respondents may feel pressure from interviewers to bias their WTP estimates upward. Finally, studies that
ask respondents to  report an annual payment (as opposed to  a lump_sum payment) have higher WTP
estimates (p<0.05). This likely to reflect the fact that annual payments are regarded as an infinite
contribution and may reflect the respondent's uncertainly regarding his future income and budget
constraints.

G.2.4 Model Selection
To select the model for estimating benefits of water quality improvements from the regulation, EPA
calculated WTP values for a range of WQI changes using both semi-log and trans-log models. In all cases
the baseline WQI is set to 50, which approximates the average WQI value across all RF1 reaches in the
United States. EPA assigned values to other independent variables corresponding with theory,
characteristics of the water resource, and the policy context. Table 10-11 in Chapter 10 provides a
complete list of values assigned to the remaining independent regressors.
As shown in Table G-4, both the semi-log and trans-log models yield similar WTP for water quality
changes greater than five points as measured by WQI. However, the semi-log model is not sensitive to
very small water quality changes (i.e., changes less than one point on the WQI index). Because the
expected water quality changes from the regulation are  relatively small, the Agency selected the trans-log
specification for estimating benefits from reducing sediment runoff from construction sites, as a more
conservative option.
November 2009                                                                                 G-18

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table G-4: Comparison of WTP for Different Changes in WQI Based on Semi-log and
 Trans-log Models
Model
Trans-log
Semi-log
20 Point
Change in
WQI
105.0186
89.0567
10 Point
Change in
WQI
79.2313
64.6167
5 Point
Change in
WQI
59.7760
55.0407
1 Point
Change in
WQI
31.0738
48.4122
0.50 Point
Change
in WQI
23.4436
47.6419
0.10 Point
Change
in WQI
12.1869
47.0345
0.01 Point
Change
in WQI
4.7796
46.8989
 Semi-log Error Term = 0.1599.
 Trans-log Error Term =0.1876.
G.3 Model Limitations
The validity and reliability of benefit transfer—including that based on meta-analysis—depends on a
variety of factors. While benefit transfer can provide valid measures of use and nonuse benefits, tests of
its performance have provided mixed results (e.g., Desvousges et al. 1998; Vandenberg et al. 2001; Smith
et al. 2002; Shrestha et al. 2007). Nonetheless, benefit transfers are increasingly applied as a core
component of benefit cost analyses conducted by EPA and other government agencies (Bergstrom and De
Civita 1999; Rosenberger and Phipps 2007). Moreover, Smith et al. (2002, p. 134) argue that "nearly all
benefit cost analyses rely on benefit transfers, whether they acknowledge it or not." Given the increasing
[or as Smith et al. (2002) might argue, universal] use of benefit transfers, an increasing focus is on the
empirical properties of applied transfer methods and models.
Although the statistical performance of the model is good, EPA notes several limitations of the model.
These limitations stem largely from information available from the original studies, as well as degrees of
freedom and statistical significance. An important factor in any benefit transfer is the ability of the study
site or estimated valuation equation to approximate the resource and context under which benefit
estimates are desired. As is common, the meta-analysis model presented here provides a close but not
perfect match to the context in which values are desired. Although all of the studies used in the meta-
analysis valued changes  in water quality improvements, many studies did not specify the cause of water
quality impairment in the baseline or focused on causes that are different from the pollutant of concern in
the regulation (i.e., sediment). Preliminary models, however, suggest no systematic patterns in WTP
associated with such factors, at least in the present metadata.
Additional  limitations relate to the paucity of demographic variables available for inclusion in the model.
The only demographic variable incorporated in the analysis (income) was not statistically significant.
Moreover, other demographic variables are unavailable.
The estimated model is statistically significant and allows estimation of WTP based on study and site
characteristics. However, strictly speaking, model findings are relative to the specific case studies
considered, and must be  viewed within the context of the 115-observation data set, with all the
appropriate caveats. Although this represents a fairly standard-to-large sample size for a meta-analysis in
this context (the 45 studies in the analysis gather data from a total of 23,589 respondents), it is relatively
small relative to other statistical applications in resource and environmental economics. Model results are
also subject to choices regarding functional form and statistical approach, although many of the primary
model effects are robust  to reasonable changes in functional form and/or statistical methods. The rationale
for the specific functional form chosen here (the  semi-log form) is detailed above.
As in all cases, results of the meta-analysis are dependent on the sample of studies available for the given
resource change (Navrud and Ready 2007), and may be subject to various selection biases if the available
November 2009
                                                                                             G-19

-------
Environmental Impact and Benefits Assessment for the C&D Category
literature does not provide a representative, unbiased perspective on welfare estimates associated with
resource changes (Rosenberger and Johnston 2007). In this case, however, the Agency took various steps
to ameliorate such potential biases, including the incorporation of both peer-reviewed and gray literature
to avoid possible publication biases (Rosenberger and Johnston 2007), and the use of a comprehensive
literature review in the attempt to avoid—as much as possible—other types of selection biases.
The relatively large (positive) magnitude of the parameter estimate for the Pacific Mountain U.S. regional
dummy variable (mp) leads EPA to question the appropriate interpretation of this effect. While it is
theoretically possible that WTP for water quality changes is substantially higher in the Pacific Northwest
(e.g., people who live in this region are outdoor enthusiasts), the magnitude of the effect suggested by the
model seems unlikely from an intuitive perspective. As suggested above, it is possible that spurious,
unexplained factors influence the magnitude of this parameter in the present model. However,
assessments of preliminary model runs suggest that this effect is relatively robust given the present data
and selection of variables available. Nonetheless, EPA recommends that the magnitude of the predicted
shift in WTP associated with the Pacific Mountain region should be viewed with caution.
Finally, as noted above, there is some controversy over the appropriateness of meta-analysis for benefit
transfer. While recognizing this controversy, the Agency emphasizes that the broader literature provides
support for the use of various types of meta-analysis for benefit transfer (cf. USEPA 2000b; Bergstrom
and Taylor 2006; Shrestha et al. 2007).
November 2009                                                                                  G-20

-------
Environmental Impact and Benefits Assessment for the C&D Category
Appendix H - Water Quality Index Tables
This appendix presents detailed versions of Table 10-7 through Table 10-10, showing WQI reach mile
and percentage improvements by baseline water quality index range, improvement range, and EPA
Region for each policy option.
November 2009
                                                                                      H-1

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-1:  Estimated Water Quality Improvements Under Option 1
EPA
Region
1
2
3
4
5
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
33 37
1,343 1,520
13,387 15,159
1,420 1,608
16,182 18,324
82 87
1,958 2,084
5,507 5,860
7,592 8,078
15,140 16,110
15448 1,684
12,552 14,599
13,006 15,127
1,898 2,207
28,904 33,617
714 746
32,338 33,801
38,191 39,918
19,192 20,060
90,435 94,525
4,175 4,375
43,315 45,386
18,508 19,393
2,287 2,396
68,285 71,550
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
77 5.76% 5.09%
1,481 11.06% 9.77%
138 9.69% 8.56%
1,696 10.48% 9.25%
0 0.00% 0.00%
36 1.82% 1.71%
177 3.21% 3.01%
202 2.67% 2.51%
415 2.74% 2.57%
51 3.49% 3.00%
645 5.14% 4.42%
780 5.99% 5.15%
64 3.39% 2.91%
1,539 5.33% 4.58%
135 18.93% 18.11%
7,903 24.44% 23.38%
12,232 32.03% 30.64%
5,939 30.95% 29.61%
26,210 28.98% 27.73%
0 0.00% 0.00%
899 2.08% 1.98%
1,859 10.04% 9.58%
173 7.56% 7.22%
2,931 4.29% 4.10%
O.KAWQK0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
0 0.00% 0.00%
31 0.23% 0.21%
0 0.00% 0.00%
31 0.19% 0.17%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
21 0.17% 0.14%
93 0.71% 0.61%
0 0.00% 0.00%
114 0.39% 0.34%
0 0.00% 0.00%
1,003 3.10% 2.97%
1,696 4.44% 4.25%
473 2.46% 2.36%
3,172 3.51% 3.36%
0 0.00% 0.00%
33 0.08% 0.07%
99 0.53% 0.51%
0 0.00% 0.00%
132 0.19% 0.18%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
9 0.69% 0.61%
20 0.15% 0.13%
0 0.00% 0.00%
29 0.18% 0.16%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
154 0.48% 0.45%
282 0.74% 0.71%
94 0.49% 0.47%
529 0.59% 0.56%
0 0.00% 0.00%
4 0.01% 0.01%
0 0.00% 0.00%
0 0.00% 0.00%
4 0.01% 0.01%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
87 6.45% 5.70%
1,532 11.44% 10.10%
138 9.69% 8.56%
1,756 10.85% 9.58%
0 0.00% 0.00%
36 1.82% 1.71%
177 3.21% 3.01%
202 2.67% 2.51%
415 2.74% 2.57%
51 3.49% 3.00%
666 5.30% 4.56%
872 6.71% 5.77%
64 3.39% 2.91%
1,653 5.72% 4.92%
135 18.93% 18.11%
9,060 28.02% 26.80%
14,210 37.21% 35.60%
6,506 33.90% 32.43%
29,911 33.07% 31.64%
0 0.00% 0.00%
937 2.16% 2.06%
1,958 10.58% 10.09%
173 7.56% 7.22%
3,067 4.49% 4.29%
November 2009
                                                                                                                      H-2

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-1:  Estimated Water Quality Improvements Under Option 1
EPA
Region
6
7
8
9
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
931 966
61,473 63,789
26,731 27,738
5,963 6,188
95,098 98,681
10,877 10,877
43,158 43,158
6,199 6,199
675 675
60,909 60,909
6,557 6,557
66,701 66,701
32,368 32,368
24,685 24,685
130,311 130,311
10,282 10,712
39,696 41,353
3,946 4,111
304 316
54,228 56,492
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
101 10.82% 10.43%
7,303 11.88% 11.45%
8,464 31.66% 30.52%
2,034 34.10% 32.86%
17,902 18.82% 18.14%
13 0.12% 0.12%
2,096 4.86% 4.86%
1,933 31.18% 31.18%
155 22.95% 22.95%
4,196 6.89% 6.89%
0 0.00% 0.00%
29 0.04% 0.04%
363 1.12% 1.12%
103 0.42% 0.42%
495 0.38% 0.38%
128 1.24% 1.19%
1,170 2.95% 2.83%
62 1.57% 1.51%
0 0.00% 0.00%
1,360 2.51% 2.41%
O.KAWQK0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
7 0.71% 0.68%
995 1.62% 1.56%
2,419 9.05% 8.72%
805 13.51% 13.02%
4,227 4.44% 4.28%
0 0.00% 0.00%
212 0.49% 0.49%
328 5.29% 5.29%
22 3.28% 3.28%
562 0.92% 0.92%
0 0.00% 0.00%
28 0.04% 0.04%
36 0.11% 0.11%
0 0.00% 0.00%
64 0.05% 0.05%
0 0.00% 0.00%
128 0.32% 0.31%
0 0.00% 0.00%
6 2.06% 1.97%
134 0.25% 0.24%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
13 1.43% 1.38%
440 0.72% 0.69%
587 2.19% 2.12%
187 3.14% 3.02%
1,227 1.29% 1.24%
11 0.10% 0.10%
132 0.31% 0.31%
26 0.41% 0.41%
0 0.00% 0.00%
168 0.28% 0.28%
68 1.04% 1.04%
190 0.28% 0.28%
100 0.31% 0.31%
4 0.02% 0.02%
362 0.28% 0.28%
7 0.07% 0.07%
144 0.36% 0.35%
0 0.00% 0.00%
0 0.00% 0.00%
151 0.28% 0.27%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
121 12.96% 12.49%
8,738 14.21% 13.70%
11,470 42.91% 41.35%
3,026 50.74% 48.90%
23,355 24.56% 23.67%
24 0.22% 0.22%
2,439 5.65% 5.65%
2,286 36.88% 36.88%
177 26.22% 26.22%
4,926 8.09% 8.09%
68 1.04% 1.04%
246 0.37% 0.37%
500 1.54% 1.54%
107 0.43% 0.43%
921 0.71% 0.71%
135 1.32% 1.26%
1,442 3.63% 3.49%
62 1.57% 1.51%
6 2.06% 1.97%
1,646 3.03% 2.91%
November 2009
                                                                                                                      H-3

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-1:  Estimated Water Quality Improvements Under Option 1
EPA
Region
10
Nation
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
39 40
13,116 13,373
24,190 24,664
30,844 31,448
68,189 69,524
35,137 36,080
315,650 325,764
182,033 190,537
94,859 97,662
627,679 650,043
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
263 2.01% 1.97%
1,043 4.31% 4.23%
2,235 7.24% 7.11%
3,541 5.19% 5.09%
427 1.22% 1.18%
20,421 6.47% 6.27%
28,393 15.60% 14.90%
11,042 11.64% 11.31%
60,285 9.60% 9.27%
O.KAWQK0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
189 1.44% 1.41%
410 1.69% 1.66%
723 2.34% 2.30%
1,322 1.94% 1.90%
7 0.02% 0.02%
2,609 0.83% 0.80%
5,112 2.81% 2.68%
2,030 2.14% 2.08%
9,757 1.55% 1.50%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
28 0.21% 0.21%
213 0.88% 0.86%
301 0.98% 0.96%
542 0.79% 0.78%
99 0.28% 0.27%
1,100 0.35% 0.34%
1,227 0.67% 0.64%
586 0.62% 0.60%
3,012 0.48% 0.46%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
480 3.66% 3.59%
1,666 6.89% 6.75%
3,259 10.57% 10.36%
5,404 7.93% 7.77%
533 1.52% 1.48%
24,131 7.64% 7.41%
34,732 19.08% 18.23%
13,658 14.40% 13.99%
73,054 11.64% 11.24%
November 2009
                                                                                                                      H-4

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-2: Estimated Water Quality Improvements Under Option 2
EPA
Region
1
2
3
4
5
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
33 37
1,343 1,520
13,387 15,159
1,420 1,608
16,182 18,324
82 87
1,958 2,084
5,507 5,860
7,592 8,078
15,140 16,110
1,448 1,684
12,552 14,599
13,006 15,127
1,898 2,207
28,904 33,617
714 746
32,338 33,801
38,191 39,918
19,192 20,060
90,435 94,525
4,175 4,375
43,315 45,386
18,508 19,393
2,287 2,396
68,285 71,550
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
3 8.79% 7.76%
176 13.10% 11.57%
2,606 19.46% 17.19%
250 17.63% 15.57%
3,035 18.75% 16.56%
0 0.00% 0.00%
90 4.59% 4.31%
766 13.91% 13.07%
671 8.84% 8.31%
1,527 10.09% 9.48%
144 9.93% 8.54%
1,907 15.19% 13.06%
2,174 16.72% 14.37%
258 13.59% 11.69%
4,483 15.51% 13.34%
221 30.97% 29.63%
11,822 36.56% 34.97%
17,431 45.64% 43.67%
8,462 44.09% 42.18%
37,936 41.95% 40.13%
31 0.75% 0.71%
2,136 4.93% 4.71%
3,268 17.66% 16.85%
454 19.84% 18.94%
5,889 8.62% 8.23%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
5 0.40% 0.35%
229 1.71% 1.51%
0 0.00% 0.00%
235 1.45% 1.28%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
5 0.07% 0.06%
5 0.03% 0.03%
0 0.00% 0.00%
28 0.22% 0.19%
80 0.62% 0.53%
0 0.00% 0.00%
108 0.37% 0.32%
13 1.89% 1.80%
1,959 6.06% 5.80%
3,450 9.03% 8.64%
1,171 6.10% 5.84%
6,594 7.29% 6.98%
0 0.00% 0.00%
89 0.21% 0.20%
320 1.73% 1.65%
0 0.00% 0.00%
409 0.60% 0.57%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
9 0.69% 0.61%
32 0.24% 0.21%
0 0.00% 0.00%
41 0.25% 0.22%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
30 0.23% 0.20%
0 0.00% 0.00%
30 0.10% 0.09%
0 0.00% 0.00%
458 1.42% 1.36%
639 1.67% 1.60%
191 0.99% 0.95%
1,288 1.42% 1.36%
0 0.00% 0.00%
25 0.06% 0.05%
60 0.33% 0.31%
0 0.00% 0.00%
85 0.12% 0.12%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
3 8.79% 7.76%
191 14.19% 12.53%
2,866 21.41% 18.91%
250 17.63% 15.57%
3,310 20.46% 18.06%
0 0.00% 0.00%
90 4.59% 4.31%
766 13.91% 13.07%
676 8.91% 8.37%
1,532 10.12% 9.51%
144 9.93% 8.54%
1,934 15.41% 13.25%
2,285 17.56% 15.10%
258 13.59% 11.69%
4,621 15.99% 13.74%
234 32.85% 31.43%
14,239 44.03% 42.13%
21,520 56.35% 53.91%
9,824 51.19% 48.97%
45,817 50.66% 48.47%
31 0.75% 0.71%
2,250 5.19% 4.96%
3,648 19.71% 18.81%
454 19.84% 18.94%
6,383 9.35% 8.92%
November 2009
                                                                                                                      H-5

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-2: Estimated Water Quality Improvements Under Option 2
EPA
Region
6
7
8
9
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
931 966
61,473 63,789
26,731 27,738
5,963 6,188
95,098 98,681
10,877 10,877
43,158 43,158
6,199 6,199
675 675
60,909 60,909
6,557 6,557
66,701 66,701
32,368 32,368
24,685 24,685
130,311 130,311
10,282 10,712
39,696 41,353
3,946 4,111
304 316
54,228 56,492
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
138 14.86% 14.32%
11,393 18.53% 17.86%
8,353 31.25% 30.11%
1,558 26.13% 25.18%
21,442 22.55% 21.73%
47 0.43% 0.43%
3,854 8.93% 8.93%
2,546 41.07% 41.07%
273 40.40% 40.40%
6,719 11.03% 11.03%
13 0.20% 0.20%
253 0.38% 0.38%
773 2.39% 2.39%
326 1.32% 1.32%
1,365 1.05% 1.05%
239 2.33% 2.23%
1,741 4.39% 4.21%
102 2.58% 2.48%
6 2.00% 1.92%
2,088 3.85% 3.70%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
1,896 3.08% 2.97%
4,035 15.09% 14.55%
1,255 21.04% 20.27%
7,185 7.56% 7.28%
0 0.00% 0.00%
420 0.97% 0.97%
557 8.99% 8.99%
22 3.28% 3.28%
1,000 1.64% 1.64%
0 0.00% 0.00%
39 0.06% 0.06%
51 0.16% 0.16%
0 0.00% 0.00%
90 0.07% 0.07%
0 0.00% 0.00%
213 0.54% 0.51%
0 0.00% 0.00%
6 2.06% 1.97%
219 0.40% 0.39%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
46 4.93% 4.76%
825 1.34% 1.29%
1,488 5.57% 5.36%
554 9.29% 8.95%
2,912 3.06% 2.95%
11 0.10% 0.10%
179 0.42% 0.42%
71 1.15% 1.15%
0 0.00% 0.00%
261 0.43% 0.43%
68 1.04% 1.04%
190 0.28% 0.28%
105 0.32% 0.32%
4 0.02% 0.02%
367 0.28% 0.28%
7 0.07% 0.07%
179 0.45% 0.43%
0 0.00% 0.00%
0 0.00% 0.00%
186 0.34% 0.33%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
184 19.79% 19.07%
14,114 22.96% 22.13%
13,876 51.91% 50.02%
3,366 56.45% 54.40%
31,540 33.17% 31.96%
57 0.53% 0.53%
4,453 10.32% 10.32%
3,174 51.21% 51.21%
295 43.68% 43.68%
7,980 13.10% 13.10%
81 1.24% 1.24%
482 0.72% 0.72%
928 2.87% 2.87%
330 1.34% 1.34%
1,821 1.40% 1.40%
246 2.40% 2.30%
2,133 5.37% 5.16%
102 2.58% 2.48%
12 4.06% 3.89%
2,493 4.60% 4.41%
November 2009
                                                                                                                      H-6

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-2: Estimated Water Quality Improvements Under Option 2
EPA
Region
10
Nation
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
39 40
13,116 13,373
24,190 24,664
30,844 31,448
68,189 69,524
35,137 36,080
315,650 325,764
182,033 190,537
94,859 97,662
627,679 650,043
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
220 1.68% 1.64%
1,538 6.36% 6.24%
2,537 8.23% 8.07%
4,295 6.30% 6.18%
836 2.38% 2.32%
33,591 10.64% 10.31%
39,557 21.73% 20.76%
14,795 15.60% 15.15%
88,779 14.14% 13.66%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
257 1.96% 1.92%
250 1.03% 1.01%
1,084 3.52% 3.45%
1,592 2.33% 2.29%
13 0.04% 0.04%
4,907 1.55% 1.51%
8,972 4.93% 4.71%
3,543 3.74% 3.63%
17,436 2.78% 2.68%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
117 0.89% 0.87%
420 1.74% 1.70%
508 1.65% 1.62%
1,044 1.53% 1.50%
132 0.38% 0.37%
1,981 0.63% 0.61%
2,844 1.56% 1.49%
1,257 1.32% 1.29%
6,214 0.99% 0.96%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
594 4.53% 4.44%
2,208 9.13% 8.95%
4,129 13.39% 13.13%
6,931 10.16% 9.97%
982 2.79% 2.72%
40,479 12.82% 12.43%
51,373 28.22% 26.96%
19,595 20.66% 20.06%
112,429 17.91% 17.30%
November 2009
                                                                                                                      H-7

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-3: Estimated Water Quality Improvements Under Option 3
EPA
Region
1
2
3
4
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
33 37
1,343 1,520
13,387 15,159
1,420 1,608
16,182 18,324
82 87
1,958 2,084
5,507 5,860
7,592 8,078
15,140 16,110
1,448 1,684
12,552 14,599
13,006 15,127
1,898 2,207
28,904 33,617
714 746
32,338 33,801
38,191 39,918
19,192 20,060
90,435 94,525
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
3 8.79% 7.76%
223 16.60% 14.66%
3,142 23.47% 20.73%
349 24.60% 21.73%
3,717 22.97% 20.29%
0 0.00% 0.00%
155 7.90% 7.42%
1,026 18.63% 17.51%
961 12.66% 11.90%
2,142 14.15% 13.29%
155 10.74% 9.23%
2,449 19.51% 16.77%
3,135 24.10% 20.72%
346 18.21% 15.66%
6,084 21.05% 18.10%
193 27.00% 25.83%
12,663 39.16% 37.46%
18,276 47.85% 45.78%
8,960 46.68% 44.66%
40,092 44.33% 42.41%
0.1 < AWQI < 0.5
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
5 0.40% 0.35%
295 2.21% 1.95%
0 0.00% 0.00%
301 1.86% 1.64%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
5 0.07% 0.06%
5 0.03% 0.03%
0 0.00% 0.00%
35 0.28% 0.24%
121 0.93% 0.80%
7 0.38% 0.33%
163 0.56% 0.48%
44 6.18% 5.92%
2,401 7.43% 7.10%
4,249 11.12% 10.64%
1,549 8.07% 7.72%
8,244 9.12% 8.72%
0.5 < AWQI
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
9 0.69% 0.61%
32 0.24% 0.21%
0 0.00% 0.00%
41 0.25% 0.22%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
30 0.23% 0.20%
0 0.00% 0.00%
30 0.10% 0.09%
0 0.00% 0.00%
618 1.91% 1.83%
771 2.02% 1.93%
220 1.15% 1.10%
1,610 1.78% 1.70%
Total Improved Reaches
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
3 8.79% 7.76%
238 17.70% 15.63%
3,469 25.91% 22.88%
349 24.60% 21.73%
4,059 25.08% 22.15%
0 0.00% 0.00%
155 7.90% 7.42%
1,026 18.63% 17.51%
966 12.72% 11.96%
2,147 14.18% 13.33%
155 10.74% 9.23%
2,483 19.78% 17.01%
3,285 25.26% 21.72%
353 18.59% 15.99%
6,277 21.72% 18.67%
237 33.18% 31.75%
15,683 48.50% 46.40%
23,296 61.00% 58.36%
10,729 55.90% 53.48%
49,945 55.23% 52.84%
November 2009
                                                                                                                      H-8

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-3: Estimated Water Quality Improvements Under Option 3
EPA
Region
5
6
7
8
9
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
4,175 4,375
43,315 45,386
18,508 19,393
2,287 2,396
68,285 71,550
931 966
61,473 63,789
26,731 27,738
5,963 6,188
95,098 98,681
10,877 10,877
43,158 43,158
6,199 6,199
675 675
60,909 60,909
6,557 6,557
66,701 66,701
32,368 32,368
24,685 24,685
130,311 130,311
10,282 10,712
39,696 41,353
3,946 4,111
304 316
54,228 56,492
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
58 1.39% 1.32%
2,595 5.99% 5.72%
3,983 21.52% 20.54%
592 25.88% 24.70%
7,228 10.59% 10.10%
185 19.90% 19.18%
13,385 21.77% 20.98%
8,577 32.09% 30.92%
1,532 25.69% 24.76%
23,679 24.90% 24.00%
34 0.31% 0.31%
4,404 10.21% 10.21%
2,718 43.85% 43.85%
317 46.97% 46.97%
7,473 12.27% 12.27%
26 0.39% 0.39%
1,348 2.02% 2.02%
1,826 5.64% 5.64%
477 1.93% 1.93%
3,677 2.82% 2.82%
295 2.87% 2.75%
2,334 5.88% 5.64%
199 5.03% 4.83%
6 2.00% 1.92%
2,834 5.23% 5.02%
0.1 < AWQI < 0.5
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
137 0.32% 0.30%
493 2.67% 2.54%
0 0.00% 0.00%
630 0.92% 0.88%
0 0.00% 0.00%
2,174 3.54% 3.41%
4,786 17.90% 17.25%
1,268 21.27% 20.50%
8,228 8.65% 8.34%
13 0.12% 0.12%
527 1.22% 1.22%
732 11.81% 11.81%
28 4.13% 4.13%
1,300 2.13% 2.13%
0 0.00% 0.00%
39 0.06% 0.06%
72 0.22% 0.22%
0 0.00% 0.00%
111 0.09% 0.09%
0 0.00% 0.00%
275 0.69% 0.67%
0 0.00% 0.00%
6 2.06% 1.97%
282 0.52% 0.50%
0.5 < AWQI
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
25 0.06% 0.05%
60 0.33% 0.31%
0 0.00% 0.00%
85 0.12% 0.12%
66 7.11% 6.86%
1,010 1.64% 1.58%
1,741 6.51% 6.28%
671 11.25% 10.84%
3,487 3.67% 3.53%
11 0.10% 0.10%
233 0.54% 0.54%
76 1.22% 1.22%
0 0.00% 0.00%
319 0.52% 0.52%
68 1.04% 1.04%
190 0.28% 0.28%
114 0.35% 0.35%
4 0.02% 0.02%
376 0.29% 0.29%
7 0.07% 0.07%
215 0.54% 0.52%
0 0.00% 0.00%
0 0.00% 0.00%
222 0.41% 0.39%
Total Improved Reaches
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
58 1.39% 1.32%
2,757 6.36% 6.07%
4,537 24.51% 23.40%
592 25.88% 24.70%
7,943 11.63% 11.10%
252 27.02% 26.04%
16,569 26.95% 25.97%
15,104 56.50% 54.45%
3,471 58.21% 56.09%
35,395 37.22% 35.87%
57 0.53% 0.53%
5,164 11.96% 11.96%
3,526 56.88% 56.88%
345 51.10% 51.10%
9,092 14.93% 14.93%
94 1.43% 1.43%
1,577 2.36% 2.36%
2,011 6.21% 6.21%
482 1.95% 1.95%
4,164 3.20% 3.20%
302 2.94% 2.82%
2,825 7.12% 6.83%
199 5.03% 4.83%
12 4.06% 3.89%
3,338 6.16% 5.91%
November 2009
                                                                                                                      H-9

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-3: Estimated Water Quality Improvements Under Option 3
EPA
Region
10
Nation
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
39 40
13,116 13,373
24,190 24,664
30,844 31,448
68,189 69,524
35,137 36,080
315,650 325,764
182,033 190,537
94,859 97,662
627,679 650,043
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
224 1.70% 1.67%
1,616 6.68% 6.55%
2,566 8.32% 8.16%
4,405 6.46% 6.34%
949 2.70% 2.63%
39,780 12.60% 12.21%
44,498 24.44% 23.35%
16,106 16.98% 16.49%
101,332 16.14% 15.59%
0.1 < AWQI < 0.5
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
314 2.39% 2.35%
358 1.48% 1.45%
1,140 3.69% 3.62%
1,812 2.66% 2.61%
57 0.16% 0.16%
5,908 1.87% 1.81%
11,106 6.10% 5.83%
4,004 4.22% 4.10%
21,075 3.36% 3.24%
0.5 < AWQI
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
117 0.89% 0.87%
446 1.85% 1.81%
607 1.97% 1.93%
1,170 1.72% 1.68%
152 0.43% 0.42%
2,416 0.77% 0.74%
3,270 1.80% 1.72%
1,502 1.58% 1.54%
7,340 1.17% 1.13%
Total Improved Reaches
%of
% of Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
654 4.99% 4.89%
2,420 10.01% 9.81%
4,312 13.98% 13.71%
7,387 10.83% 10.63%
1,158 3.30% 3.21%
48,104 15.24% 14.77%
58,874 32.34% 30.90%
21,612 22.78% 22.13%
129,747 20.67% 19.96%
November 2009
                                                                                                                      H-10

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-4: Estimated Water Quality Improvements Under Option 4
EPA
Region
1
2
3
4
5
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
33 37
1,343 1,520
13,387 15,159
1,420 1,608
16,182 18,324
82 87
1,958 2,084
5,507 5,860
7,592 8,078
15,140 16,110
15448 1,684
12,552 14,599
13,006 15,127
1,898 2,207
28,904 33,617
714 746
32,338 33,801
38,191 39,918
19,192 20,060
90,435 94,525
4,175 4,375
43,315 45,386
18,508 19,393
2,287 2,396
68,285 71,550
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
131 9.76% 8.62%
1,982 14.81% 13.08%
224 15.80% 13.95%
2,338 14.45% 12.76%
0 0.00% 0.00%
90 4.59% 4.31%
713 12.95% 12.17%
630 8.30% 7.80%
1,433 9.47% 8.90%
144 9.93% 8.54%
2,031 16.18% 13.91%
2,229 17.14% 14.74%
217 11.41% 9.81%
4,620 15.98% 13.74%
234 32.85% 31.43%
12,208 37.75% 36.12%
17,484 45.78% 43.80%
8,581 44.71% 42.77%
38,507 42.58% 40.74%
58 1.39% 1.32%
2,184 5.04% 4.81%
3,242 17.52% 16.72%
495 21.65% 20.66%
5,979 8.76% 8.36%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
0 0.00% 0.00%
156 1.16% 1.03%
0 0.00% 0.00%
156 0.96% 0.85%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
5 0.07% 0.06%
5 0.03% 0.03%
0 0.00% 0.00%
28 0.22% 0.19%
80 0.62% 0.53%
0 0.00% 0.00%
108 0.37% 0.32%
0 0.00% 0.00%
2,084 6.44% 6.16%
3,785 9.91% 9.48%
1,290 6.72% 6.43%
7,159 7.92% 7.57%
0 0.00% 0.00%
68 0.16% 0.15%
308 1.67% 1.59%
0 0.00% 0.00%
377 0.55% 0.53%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
9 0.69% 0.61%
20 0.15% 0.13%
0 0.00% 0.00%
29 0.18% 0.16%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
0 0.00% 0.00%
30 0.23% 0.20%
0 0.00% 0.00%
30 0.10% 0.09%
0 0.00% 0.00%
531 1.64% 1.57%
714 1.87% 1.79%
220 1.15% 1.10%
1,465 1.62% 1.55%
0 0.00% 0.00%
25 0.06% 0.05%
60 0.33% 0.31%
0 0.00% 0.00%
85 0.12% 0.12%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
140 10.45% 9.23%
2,158 16.12% 14.24%
224 15.80% 13.95%
2,522 15.59% 13.77%
0 0.00% 0.00%
90 4.59% 4.31%
713 12.95% 12.17%
635 8.37% 7.87%
1,438 9.50% 8.93%
144 9.93% 8.54%
2,058 16.40% 14.10%
2,339 17.98% 15.46%
217 11.41% 9.81%
4,758 16.46% 14.15%
234 32.85% 31.43%
14,823 45.84% 43.86%
21,982 57.56% 55.07%
10,091 52.58% 50.30%
47,130 52.12% 49.86%
58 1.39% 1.32%
2,277 5.26% 5.02%
3,611 19.51% 18.62%
495 21.65% 20.66%
6,441 9.43% 9.00%
November 2009
                                                                                                                      H-11

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-4: Estimated Water Quality Improvements Under Option 4
EPA
Region
6
7
8
9
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
931 966
61,473 63,789
26,731 27,738
5,963 6,188
95,098 98,681
10,877 10,877
43,158 43,158
6,199 6,199
675 675
60,909 60,909
6,557 6,557
66,701 66,701
32,368 32,368
24,685 24,685
130,311 130,311
10,282 10,712
39,696 41,353
3,946 4,111
304 316
54,228 56,492
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
185 19.90% 19.18%
11,913 19.38% 18.68%
8,090 30.26% 29.17%
1,504 25.22% 24.31%
21,693 22.81% 21.98%
34 0.31% 0.31%
3,922 9.09% 9.09%
2,604 42.01% 42.01%
300 44.40% 44.40%
6,860 11.26% 11.26%
26 0.39% 0.39%
250 0.37% 0.37%
711 2.20% 2.20%
296 1.20% 1.20%
1,282 0.98% 0.98%
251 2.44% 2.34%
1,807 4.55% 4.37%
101 2.56% 2.46%
6 2.00% 1.92%
2,166 3.99% 3.83%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
2,085 3.39% 3.27%
4,456 16.67% 16.07%
1,278 21.44% 20.66%
7,819 8.22% 7.92%
13 0.12% 0.12%
524 1.22% 1.22%
639 10.31% 10.31%
28 4.13% 4.13%
1,205 1.98% 1.98%
0 0.00% 0.00%
39 0.06% 0.06%
51 0.16% 0.16%
0 0.00% 0.00%
90 0.07% 0.07%
0 0.00% 0.00%
191 0.48% 0.46%
0 0.00% 0.00%
6 2.06% 1.97%
197 0.36% 0.35%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
46 4.93% 4.76%
945 1.54% 1.48%
1,698 6.35% 6.12%
624 10.46% 10.08%
3,313 3.48% 3.36%
11 0.10% 0.10%
193 0.45% 0.45%
76 1.22% 1.22%
0 0.00% 0.00%
280 0.46% 0.46%
68 1.04% 1.04%
190 0.28% 0.28%
105 0.32% 0.32%
4 0.02% 0.02%
367 0.28% 0.28%
7 0.07% 0.07%
179 0.45% 0.43%
0 0.00% 0.00%
0 0.00% 0.00%
186 0.34% 0.33%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
231 24.84% 23.94%
14,943 24.31% 23.43%
14,245 53.29% 51.35%
3,406 57.12% 55.05%
32,825 34.52% 33.26%
57 0.53% 0.53%
4,640 10.75% 10.75%
3,319 53.55% 53.55%
328 48.53% 48.53%
8,345 13.70% 13.70%
94 1.430/0 L43o/0
479 0.72% 0.72%
866 2.68% 2.68%
300 1.22% 1.22%
1,739 1.33% 1.33%
258 2.51% 2.41%
2,177 5.48% 5.26%
101 2.56% 2.46%
12 4.06% 3.89%
2,548 4.70% 4.51%
November 2009
                                                                                                                      H-12

-------
Environmental Impact and Benefits Assessment for the C&D Category
 Table H-4: Estimated Water Quality Improvements Under Option 4
EPA
Region
10
Nation
Baseline
Water
Quality
<26
26-50
50-70
>70
Total
<26
26-50
50-70
>70
Total
Baseline Scenario
Total
Reach
Reach Miles in
Miles RF1
Modeled Network
39 40
13,116 13,373
24,190 24,664
30,844 31,448
68,189 69,524
35,137 36,080
315,650 325,764
182,033 190,537
94,859 97,662
627,679 650,043
Water Quality Improvements by WQI Change
0.01 < AWQI < 0.1
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
236 1.80% 1.77%
1,342 5.55% 5.44%
2,317 7.51% 7.37%
3,894 5.71% 5.60%
932 2.65% 2.58%
34,773 11.02% 10.67%
38,497 21.15% 20.20%
14,569 15.36% 14.92%
88,772 14.14% 13.66%
0.1 < AWQI < 0.5
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
180 1.37% 1.35%
259 1.07% 1.05%
941 3.05% 2.99%
1,380 2.02% 1.99%
13 0.04% 0.04%
5,199 1.65% 1.60%
9,735 5.35% 5.11%
3,548 3.74% 3.63%
18,495 2.95% 2.85%
0.5 < AWQI
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
96 0.74% 0.72%
385 1.59% 1.56%
461 1.50% 1.47%
943 1.38% 1.36%
132 0.38% 0.37%
2,168 0.69% 0.67%
3,087 1.70% 1.62%
1,309 1.38% 1.34%
6,696 1.07% 1.03%
Total Improved Reaches
% of % of
Reach Total
Reach Miles Reach
Miles Modeled Miles
0 0.00% 0.00%
513 3.91% 3.83%
1,986 8.21% 8.05%
3,719 12.06% 11.82%
6,217 9.12% 8.94%
1,077 3.06% 2.98%
42,140 13.35% 12.94%
51,319 28.19% 26.93%
19,427 20.48% 19.89%
113,963 18.16% 17.53%
November 2009
                                                                                                                      H-13

-------