&EPA
          United States
          Environmental Protection
          Agency
               Office of Water
               Washington, DC 20460
               www.epa.gov/npdes
EPA 833-R-10-005
August 2010
Final Report
Report to Congress:

Study of Discharges
Incidental to Normal JJT
Operation of Commercial
Fishing Vessels and Other
Non-Recreational Vessels
Less than 79 Feet
                                        II
                                       /
                                              fit/'
                                             L.
                                               '

-------
                            ACKNOWLEDGMENTS

       The EPA Office of Wastewater Management (OWM) presents this draft Vessels
Study Report to Congress conducted to meet the obligations of EPA under Public Law
(P.L.) 110-299 (July 31, 2008).  EPA would like to thank the numerous trade associations
and individual companies who contributed to this project. Those groups who provided
assistance to EPA are listed in Chapter 2 of this report.  The project could not have been
successful without the support by EPA Region 2, 3, and 5 laboratories, EPA Gulf
Ecology Division and other EPA program offices.  EPA would also like to thank the
United States Coast Guard for providing both logistical support and review of many of
the report's elements. Finally, EPA would also like to acknowledge the contractor
support for this project provided by individuals from Great Lakes Environmental Center,
Inc., Eastern Research Group, and Abt Associates.

-------
           The primary technical contacts for this document are:

                             Ryan Albert
                 U.S. Environmental Protection Agency
                  Office of Water (Mail Code: 4203M)
                    1200 Pennsylvania Avenue, NW
                        Washington, DC 20460
                      (202) 564-0763 (telephone)
                         (202) 564-6392 (fax)

                            Robin Danesi
                 U.S. Environmental Protection Agency
                  Office of Water (Mail Code: 4203M)
                    1200 Pennsylvania Avenue, NW
                        Washington, DC 20460
                      (202) 564-1846 (telephone)
                         (202) 564-6392 (fax)

   The primary EPA congressional relations contact for this document is:

                             Greg Spraul
                 U.S. Environmental Protection Agency
Office of Congressional and Intergovernmental Relations (Mail Code: 1301 A)
                    1200 Pennsylvania Avenue, NW
                        Washington, DC 20460
                      (202) 564-0255 (telephone)
                         (202) 564-1519 (fax)
                                 11

-------
                               TABLE OF CONTENTS

EXECUTIVE SUMMARY	xv
CHAPTER 1 Introduction to the Report	1
   1.1.       Congressional Study Charge	1
   1.2.       Organization of this report	2
   1.3.       Classes or Types of Vessels	2
       1.3.1. Commercial Fishing Vessels	3
       1.3.2. Tugs/Towing Vessels	8
       1.3.3. Water Taxis/Small Ferries	9
       1.3.4. Tour Boats	9
       1.3.5. Recreational Vessels Used for Non-Recreational Purposes	10
   1.4.       Vessel Population	11
       1.4.1. Vessel Characteristics Data	11
       1.4.2. Overview of Vessel Universe	12
   1.5.       Discharges from Vessels	24
       1.5.1. Bilgewater	25
       1.5.2. Deck Washdown and Deck Runoff.	25
       1.5.3. Engine Effluent	26
       1.5.4. Firemain Systems	28
       1.5.5. Fish Hold and Fish Hold Cleaning Effluent (Refrigerated Seawater Discharge or
             Fish Ice Slurry Discharge)	29
       1.5.6. Graywater	31
       1.5.7. Shaft Packing Gland Effluent	31
       1.5.8. Antifouling Hull Coatings	32
   1.6.       Pollutants Potentially Found in Vessel Discharges	32
       1.6.1. Classical Pollutants	33
       1.6.2. Nutrients	35
       1.6.3. Pathogen Indicators	36
       1.6.4. Metals	36
       1.6.5. Volatile and Semivolatile Organic Compounds	36
       1.6.6. Nonylphenols	37
   1.7.       Chapter Conclusions	37
CHAPTER 2 Study Design	39
   2.1        Data Sources	39
       2.1.1  Existing EPA Data Sources	39
       2.1.2  Industry Participation	40
       2.1.3  Vessel Sampling	42
       2.1.4  Literature Review	42
       2.1.5  Other Governmental Data Sources	42
   2.2        EPA Vessel Discharge Sampling Program	43
       2.2.1  Vessels Sampled and Locations	44
       2.2.2  Sampled Discharges	49
       2.2.3  Target Analytes	51
       2.2.4  Sampling Methods	53
       2.2.5  QA/QC	58
                                          in

-------
  2.3        Data Considerations and study limitations	61
      2.3.1  Voluntary Nature of the Sampling Program	61
      2.3.2  Vessels/Discharges Not Sampled	62
      2.3.3  Pollutants Not Sampled	63
      2.3.4  Application to Other Vessels, Including Larger Vessels Not Sampled for this
             Study	63
   CHAPTER 3 Analysis of Discharges and Potential Impact to Human Health and the
             Environment	67
  3.1        Approach to Analyses	67
      3.1.1  Data Reduction and Presentation	68
      3.1.2  Summary Statistics and Box Plots	69
      3.1.3  Calculation of Potential Hazard Quotients	70
  3.2        Characterization of Discharges	79
      3.2.1  Bilgewater	79
      3.2.2  Stern Tube Packing Gland Effluent	115
      3.2.3  DeckWashdown	143
      3.2.4  Fish Hold and Fish Hold Cleaning Effluent (Refrigerated Seawater and Ice
             Slurry)	188
      3.2.5  Graywater	234
      3.2.6  Engine Effluent	261
      3.2.7  Firemain Discharges	338
      3.2.8  Antifouling Hull Coatings	354
  CHAPTER 4 Potential Large-Scale Impacts of Study Vessels' Incidental Discharges to
             Human Health and the Environment	368
  4.1        Model Selection	370
  4.2        Fraction of Freshwater Model	371
      4.2.1  Step 1: Calculate Vessel Discharge Analyte Loading Rates	372
      4.2.2  Step 2: Calculate the Fraction of Freshwater in the Harbor	373
      4.2.3  Step 3: Calculate the Harbor Flushing Time	373
      4.2.4  Step 4: Calculate the Harbor Analyte Concentration	374
  4.3        Vessel Discharge Loading Rates	374
      4.3.1  Calculate the Average Analyte Concentrations	374
      4.3.2  Discharge Flow Rate Assumptions	375
      4.3.3  Number of Vessels Present in the Harbor	381
      4.3.4  Percentage of Vessels Discharging in the Harbor	385
      4.3.5  Vessel Discharge Loading Rates	387
      4.3.6  Dissolved Copper Loading Rates from Antifouling Paints	387
  4.4        Hypothetical Harbor	389
  4.5        Model Scenarios	391
  4.6        Model Results	393
      4.6.1  Dilution Factor Analysis	393
      4.6.2  Supplemental Model Run in Response to Comments	394
      4.6.3  Loading Rate Analysis	395
  4.7        Conclusions	399
CHAPTER 5 Summary of Findings	401
  5.1        Summary of Classes of Vessels Covered By this Study	401
                                          IV

-------
  5.2        Summary of Effluent Characterization of Select Discharges from the Study
             Vessels	401
      5.2.1   Estimated Volumes of Select Discharges from the Study Vessels	402
      5.2.2   Analytes of Potential Risk in Select Discharges from Study Vessels	404
  5.3        Summary of Predicted Impacts from Select Pollutants in Study Vessel
             Discharges	417
      5.3.1   Potential Watershed-Wide Impacts from Study Vessels	417
      5.3.2   Potential Localized or Near-Field Impacts of Vessel Discharges to Receiving
             Waters	418
  5.4        Possible Benefits to Human Health, Welfare, and the Environment from
             Reducing, Eliminating, Controlling, or Mitigating One or More of the Discharges
             from the Study Vessels	420
CHAPTER 6 Analysis of the Extent to Which Incidental Discharges are Currently Subject
              to Regulation Under Federal Law or a Binding International Obligation of
             the United States	422
  6.1        International Agreements	422
      6.1.1   The International Convention for the Prevention of Pollution from Ships
             (MARPOL 73/78)	422
      6.1.2   The International Convention on the Control of Harmful Anti-Fouling Systems on
             Ships	432
      6.1.3   International Convention for the Safety  of Life at Sea (SOLAS)	435
      6.1.4   Boundary Waters Treaty	436
      6.1.5   Great Lakes Water Quality Agreement	436
      6.1.6   St. Lawrence Seaway Regulations	439
  6.2        Federal Laws	440
      6.2.1   Act to Prevent Pollution from Ships (APPS)	440
      6.2.2   Clean Water Act (CWA) §§311, 312/Oil Pollution Control Act	444
      6.2.3   Organotin Antifouling Paint Control Act	446
      6.2.4   National Invasive Species Act	447
      6.2.5   Hazardous Materials Transportation Act	448
      6.2.6   National Marine Sanctuaries Act	448
      6.2.7   Resource Conservation and Recovery Act	449
      6.2.8   Federal Insecticide, Fungicide, and Rodenticide Act	450
  6.3        Additional International and Federal Laws	451
      6.3.1   International Convention on the Prevention of Marine Pollution by Dumping of
             Wastes and Other Matter	451
      6.3.2   International Convention on Oil Pollution, Preparedness, Response and
             Cooperation	452
      6.3.3   International Convention Relating to Intervention on the High Seas in Cases of
             Oil Pollution Casualties	452
      6.3.4   Comprehensive Environmental Response, Compensation,  and Liability Act.... 452
      6.3.5   CWA § 402, National Pollutant Discharge Elimination System (NPDES)	453
      6.3.6   Title XIV of the Consolidated Appropriations Act, 2001—Certain Alaskan Cruise
             Ship Operations	453
      6.3.7   Toxic Substances Control Act	453

-------
  6.4        Application of Legal Authorities to Discharges Incidental to the Normal
             Operation of Study Vessels	454
CHAPTER 7 References	458
Appendix A List of Acronyms	A-l
Appendix B Additional Characteristics of the P.L. 110 - 299 Vessel Population	B-l
  B.I        Vessel Subcategories	B-l
      B.I.I  Population of Vessels undergoing Discharge Analysis	B-2
  B.2        Vessel Geographical Area of Operation	B-4
  B.3        Other Vessel Characteristics: Construction and Propulsion	B-ll
      B.3.1  Vessel Age	B-12
      B.3.2  Hull Material Type	B-13
      B.3.3  Propulsion Method and Type	B-l5
      B.3.4  Horsepower Ahead	B-l6
  B.4        Distribution of the Study Vessel Universe versus the Recreational Vessel
             Universe	B-l 8
  B.5        Vessels Documented, Inspected, and/or State Registered	B-25
  B.6        Uncertainty	B-26
Appendix C Public Law 110-299 (S. 3298) and Public Law 110-288 (S. 2766)	C-l
Appendix D List of Target Analytes	D-l
Appendix E Analyte Concentrations and Summary Statistics from Ambient Water Samples
              	E-l
 Appendix F Analyte Concentrations and Summary Statistics from Source Water Samples
              	F-l
Appendix G Supporting Information for EPA's Screening-Level Water Quality Model. G-l
Appendix H Responsiveness Summary	H-l
Appendix I List of Preparers and Contributors	1-1
                                         VI

-------
                                  LIST OF TABLES

Table 1.1. Population of Operational, Domestic MISLE Vessels by Vessel Length	14
Table 2.1. Number of Vessels Sampled by Vessel Class and Discharge	46
Table 2.2. Analyte Groups by Discharge	52
Table 3.1. Water Quality and Other Benchmark Values Used to Screen the Vessel Discharge
       Data	75
Table 3.2. Major Cation Concentrations in Seawater	78
Table 3.3. Major Cation Concentrations in Freshwater	78
Table 3.1.1. Results of Bilgewater Sample Analyses for Dissolved Metals	85
Table 3.1.2. Results of Bilgewater Sample Analyses for Total Metals	86
Table 3.1.3. Results of Bilgewater Sample Analyses for Classical Pollutants	93
Table 3.1.4. Results of Bilgewater Sample Analyses for Pathogen Indicators	97
Table 3.1.5. Results of Bilgewater Sample Analyses for Nutrients	99
Table 3.1.6. Results of Bilgewater Sample Analyses for SVOCs	103
Table 3.1.7. Results of Bilgewater Sample Analyses for VOCs	108
Table 3.1.8. Results of Bilgewater Sample Analyses forNonylphenols	112
Table 3.1.9. Characterization of Bilgewater Discharge and Summary of Analytes that May Have
       the Potential to Pose Risk	114
Table 3.2.1. Results of Packing Gland Effluent Sample Analyses for Metals	118
Table 3.2.2. Results of Packing Gland Effluent Sample Analyses for Classical Pollutants	125
Table 3.2.3. Results of Packing Gland Effluent Sample Analyses for Nutrients	129
Table 3.2.4. Results of Packing Gland Water Sample Analyses for SVOCs and VOCs	133
Table 3.2.5. Results of Packing Gland Water Sample Analyses forNonylphenols	139
Table 3.2.6. Characterization of Packing Gland Effluent and Summary of Analytes that May
       Have the Potential to Pose Risk	142
Table 3.3.1. Results of Deck Washdown/Runoff Sample Analyses for Metals	152
Table 3.3.2. Dissolved-to-Total Metal Ratios (fas) in Paired Deck Washdown/Runoff
       Samples	154
Table 3.3.3. Minimum and Maximum Dissolved and Total Metal  Concentrations in Vessel
       Source and Ambient (Harbor) Water Relative to Median Sample Concentrations and
       Most Stringent Screening Benchmarks	154
Table 3.3.4. Comparison of Metal Concentrations in Deck Washdown Discharge Between
       Fishing Vessels and Non-Fishing Vessels	155
Table 3.3.5. Mean Concentrations of Dissolved and Total Heavy Metals from Deck Wash
       Discharges from Fishing Vessels and Nonfishing Vessels	156
Table 3.3.6. Results of Deck Washdown Water Sample Analyses  for Classical Pollutants	164
Table 3.3.7. Results of Deck Washdown Water Sample Analyses  for Pathogen Indicators	169
Table 3.3.8. Results of Deck Washdown Water Sample Analyses  for Nutrients	174
Table 3.3.9. Results of Deck Washdown Water Sample Analyses  for Long-Chain
       Nonylphenols	178
Table 3.3.10. Results of Deck Washdown Water Sample Analyses for VOCs  and SVOCs	182
Table 3.3.11. Characterization of Deck Washdown and Runoff Water and Summary of Analytes
       that May Have the Potential to Pose Risk	187
Table 3.4.1. Results of Fish Hold Effluent Sample Analyses for Total Metals	195
Table 3.4.2. Results of Fish Hold Effluent Sample Analyses for Dissolved Metals	196
                                         vn

-------
Table 3.4.3. Results of Fish Hold Cleaning Effluent Sample Analyses for Metals	202
Table 3.4.4. Results of Fish Hold Effluent Sample Analyses for Classical Pollutants	210
Table 3.4.5. Results of Fish Hold Cleaning Effluent Analyses for Classical Pollutants	211
Table 3.4.6. Results of Fish Hold and Fish Hold Cleaning Effluent Sample Analyses for
       Pathogen Indicators	217
Table 3.4.7. Results of Fish Hold (upper half) and Fish Hold Cleaning Effluent (lower half)
       Sample Analyses for Nutrients	221
Table 3.4.8. Raw Sewage Concentrations of Nutrients	221
Table 3.4.9. Results of Fish Hold Cleaning Effluent Sample Analyses for Long-chain
       Nonylphenols	226
Table 3.4.10. Means (and Standard Deviations) for Selected Analyte Concentrations, by
       Geographic Region. Units for All Analytes Expressed as |ig/L, Except for
       BOD(mg/L)	230
Table 3.4.11. Characterization of Fish Hold Effluent and Fish Hold Cleaning Effluent and
       Summary of Analytes that May Have the Potential to Pose Risk	233
Table 3.5.1. Results of Gray water Sample Analyses for Pathogen Indicators	236
Table 3.5.2. Results of Graywater Sample Analyses for Classical Pollutants	241
Table 3.5.3. Results of Graywater Sample Analyses for Nonylphenols (only long-chain NPEOs
       and OPEOs were detected)	245
Table 3.5.4. Results of Graywater Sample Analyses for Dissolved Metals	248
Table 3.5.5. Results of Graywater Sample Analyses for Total Metals	249
Table 3.5.6. Results of Graywater Sample Analyses for Nutrients	255
Table 3.5.7. Characterization of Graywater Effluent and Summary of Analytes that May Have
       the Potential to Pose Risk	260
Table 3.6.1. Sampled Engine Characteristics	263
Table 3.6.2. Results of Inboard Propulsion Engine Sample Analyses for Classical Pollutants . 267
Table 3.6.3. Results of Inboard Propulsion Engine Sample Analyses for Metals	271
Table 3.6.4. Comparison of Metals Results for EPA P.L. 110-299 and UNDS Engine Wet
       Exhaust Sampling	273
Table 3.6.5. Results of Inboard Propulsion Engine Sample Analyses for SVOCs	280
Table 3.6.6. Comparison of Phenol Results for EPA P.L. 110-299 and UNDS Engine Wet
       Exhaust Sampling	281
Table 3.6.7. Results of Inboard Propulsion Engine Sample Analyses for VOCs	286
Table 3.6.8. Results of Outboard Propulsion Engine Sample Analyses for Classical
       Pollutants	292
Table 3.6.9. Results of Outboard Propulsion Engine Sample Analyses for Metals	295
Table 3.6.10. Results of Outboard Propulsion Engine Sample Analyses for SVOCs	302
Table 3.6.11. Results of Outboard Propulsion Engine Sample Analyses for VOCs	305
Table 3.6.12. Results of Generator Engine Sample Analyses for Classical Pollutants	310
Table 3.6.13. Results of Generator Engine Sample Analyses for Metals	313
Table 3.6.14. Results of Generator Engine Sample Analyses for SVOCs	319
Table 3.6.15. Results of Generator Engine Sample Analyses for VOCs	324
Table 3.6.16. Mean Concentration Results, UNDS Engine Wet Exhaust Discharge and
       Background Samples for the LCPL	330
Table 3.6.17. Mean Concentration Results, UNDS Engine Wet Exhaust Discharge and
       Background Samples for the RIB	331
                                         Vlll

-------
Table 3.6.18. Comparison of Dewinterizing Effluent with Propulsion Effluent	333
Table 3.6.19. Comparison of Number of Detected Analytes in Engine Effluent	334
Table 3.6.20. Comparison of Results for Selected Analytes in Engine Effluent	334
Table 3.6.21. Characterization of Engine Effluent and Summary of Analytes that May Have the
       Potential to Pose Risk	336
Table 3.7.1. Results of Firemain System Sample Analyses for Metals	341
Table 3.7.2. Results of Firemain System Water Sample Analyses for Classical Pollutants	347
Table 3.7.3. Results of Firemain Water Sample Analyses for SVOCs	350
Table 3.7.4. Characterization of Firemain Discharge and Summary of Analytes that May Have
       the Potential to Pose Risk	353
Table 3.8.1. Rates of Passive Copper Leaching from Vessel AFSs	361
Table 3.8.2. Dissolved Copper Release from Vessel AFSs During an Underwater Hull Cleaning
       "Event"	362
Table 3.8.3. Estimated Dissolved Copper Mass Emissions from a 9.1m (30ft) Powerboat	362
Table 4.3.1. Offload Frequency by Fishing Vessel Subtype	376
Table 4.3.2. Examples of Field Data and Assumptions for Flow Rate Calculations by
       Discharge	377
Table 4.3.3. Vessel Flow Rates	379
Table 4.3.4. Vessel Population Scenario Representative Harbors Based on the Top 20 Hailing
       Ports Cited in the MISLE Database	382
Table 4.3.5. Percentage of Study Vessels Present in Representative Fishing Harbor	383
Table 4.3.6. Percentage of Study Vessels Present in Representative Large Metropolitan
       Harbor	383
Table 4.3.7. Percent of Study Vessels Present in Representative Recreational Harbor	384
Table 4.3.8. Vessel Population Scenarios	385
Table 4.3.9. Percentage of Vessels Discharging in the Harbor	386
Table 4.3.10. Estimated Average Vessel Length by Vessel Class	388
Table 4.4.1. Harbors Selected for Model Input Parameter Development	389
Table 4.4.2. Hypothetical Harbor Input Parameters	390
Table 4.4.3. Hypothetical Harbor Scenarios	391
Table 4.5.1. Fraction of Freshwater Model Scenarios	392
Table 4.6.1. "Tipping Point" Dilution Factors for Harbor Instantaneous Concentration to Equal
       the NRQWC Based on Vessel Population Scenario Loading Rates l	394
Table 4.6.2. Revised Model Assumptions	394
Table 4.6.3. Supplemental Model Run "Tipping Point" Dilution Factors for Harbor Instantaneous
       Concentration to Equal the NRQWC Based on Vessel Population Scenario Loading
       Rates	395
Table 4.6.4. Comparison of Model Loading Rates with Other Potential Point Source Loading
       Rates	398
Table 5.1. Analytes of Potential Risk by Discharge	411
Table 6. 1. International Treaties  and Federal Laws Applicable to Discharges Incidental to the
Normal Operation of Vessels	455
Table 6.2. International Treaties and Federal Laws Applicable to Vessels (by Length)	457
                                          IX

-------
                                 LIST OF FIGURES

Figure 1.1. MISLE Population of Operational, Domestic Non-Recreational Vessels by Vessel
             Service	13
Figure 1.2. Number of Study Vessels Recorded in MISLE, by Vessel Service (Type)	16
Figure 1.3. Relationship Between Vessel Gross Tons and Length	19
Figure 1.4. Distribution of MISLE Vessels by Length and Vessel Service (Type)	20
Figure 1.5. Cumulative Distribution of MISLE Vessels by Length and Vessel Service (Type).. 21
Figure 1.6. Distribution of Study Vessels by Length (in Feet) and Vessel Service (Type)	22
Figure 1.7. Distribution of Study Vessels by Gross Tons and Vessel  Service (for which gross ton
             data are given inMSLE)	23
Figure 3.1.1. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples ofBilgewater	87
Figure 3.1.2. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Bilgewater	88
Figure 3.1.3. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples ofBilgewater	89
Figure 3.1.4. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples ofBilgewater	90
Figure 3.1.5. Box and Dot Density Plot of Classical Pollutant Concentrations Measured in
             Samples ofBilgewater	94
Figure 3.1.6. Box and Dot Density Plot of Potential Hazard Quotients for Classical Parameters in
             Samples ofBilgewater	95
Figure 3.1.7. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of
             Bilgewater	100
Figure 3.1.8. Box and Dot Density Plot of Potential Hazard Quotients for Nutrients in Samples of
             Bilgewater	101
Figure 3.1.9. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
             Bilgewater	104
Figure 3.1.10. Box and Dot Density Plot of Potential Hazard Quotients for S VOCs in Samples of
             Bilgewater	105
Figure 3.1.11. Box and Dot Density Plot of VOC Concentrations Measured in Samples of
             Bilgewater	109
Figure 3.1.12. Box and Dot Density Plot of Potential Hazard Quotients for VOCs in Samples of
             Bilgewater	110
Figure 3.2.1. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Packing  Gland Effluent	120
Figure 3.2.2. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Packing Gland Effluent	121
Figure 3.2.3. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Packing  Gland Effluent	122
Figure 3.2.4. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples of Packing  Gland Effluent	123
Figure 3.2.5. Box and Dot Density Plot of Classical Pollutants Measured in Samples of Packing
             Gland Effluent	126

-------
Figure 3.2.6. Box and Dot Density Plot of Potential Hazard Quotients for Classical Pollutants in
             Samples of Packing Gland Effluent	127
Figure 3.2.7. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of
             Packing Gland Effluent	130
Figure 3.2.8. Box and Dot Density Plot of Potential Hazard Quotients for Nutrients in Packing
             Gland Effluent	131
Figure 3.2.9. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
             Packing Gland Effluent Samples	134
Figure 3.2.10. Box and Dot Density Plot of VOC Concentrations Measured in Samples of
             Packing Gland Effluent Samples	135
Figure 3.2.11. Box and Dot Density Plot of Potential Hazard Quotients for SVOCs in Samples of
             Packing Gland Effluent	136
Figure 3.2.12. Box and Dot Density Plot of Potential Hazard Quotients for VOCs in Samples of
             Shaft Packing Gland Effluent	137
Figure 3.2.13. Box and Dot Density Plot of Nonylphenol Concentrations Measured in Samples of
             Packing Gland Effluent	140
Figure 3.3.1. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Deck Washdown Water	157
Figure 3.3.2. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Deck Washdown Water	158
Figure 3.3.3. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Deck Washdown Water	159
Figure 3.3.4. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples of Deck Washdown Water	160
Figure 3.3.5. Box and Dot Density Plot of Classical Pollutants Measured in Samples of Deck
             Washdown Water	165
Figure 3.3.6. Box and Dot Density Plot of Potential Hazard Quotients for Classical Pollutants in
             Samples of Deck Washdown Water	166
Figure 3.3.7 Box and Dot Density Plot of Pathogen Indicator Concentrations Measured in
             Samples of Deck Washdown Water	170
Figure 3.3.8. Box and Dot Density Plot of Potential Hazard Quotients for Pathogens in Samples
             of Deck Washdown Water	171
Figure 3.3.9. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of Deck
             Washdown Water	175
Figure 3.3.10. Box and Dot Density Plot of Potential Hazard Quotients for Nutrients in Samples
             of Deck Washdown Water	176
Figure 3.3.11. Box and Dot Density Plot of Nonylphenol Concentrations Measured in Samples of
             Deck Washdown Water	179
Figure 3.3.12. Box and Dot Density Plot of Volatile Organic Chemical Concentrations Measured
             in Samples of Deck Washdown Water	183
Figure 3.3.13. Box and Dot Density Plot of Potential Hazard Quotients for VOCs in Samples of
             Deck Washdown Water	184
Figure 3.4.1. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Fish Hold Effluent	197
Figure 3.4.2. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Fish Hold Effluent	198
                                          XI

-------
Figure 3.4.3. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples of Fish Hold Effluent	199
Figure 3.4.4. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Fish Hold Effluent	200
Figure 3.4.5. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Fish Hold Cleaning Effluent	203
Figure 3.4.6. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Fish Hold Cleaning Effluent	204
Figure 3.4.7. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples of Fish Hold Cleaning Effluent	205
Figure 3.4.8. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Fish Hold Cleaning Effluent	206
Figure 3.4.9. Box and Dot Density Plot of Classical Pollutant Concentrations/Values Measured
             in Samples of Fish Hold Effluent	212
Figure 3.4.10. Box and Dot Density Plot of Classical Pollutant Concentrations/Values Measured
             in Samples of Fish Hold Cleaning Effluent	213
Figure 3.4.11. Comparison Between the BOD Secondary Treatment Limit from Sewage
             Treatment Facilities (30 mg/L), Average BOD Raw Sewage Concentrations, and
             BOD Concentrations from Fish Hold Effluent and Fish Hold Cleaning
             Effluent	214
Figure 3.4.12. Box and Dot Density Plot of Measured Pathogen Concentrations in Samples of
             Fish Hold Effluent	218
Figure 3.4.13. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of
             Fish Hold Effluent	222
Figure 3.4.14. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of
             Fish Hold Cleaning Effluent	223
Figure 3.4.15. Comparison of Concentrations of Ammonia, TKN, and Total Phosphorus in
             Different Fishing Vessel Platforms to those in the Lobster Tank (which has a live
             catch and continuously  circulating water)	224
Figure 3.4.16. Box and Dot Density Plot of Nonylphenol Concentrations Measured in Samples of
             Fish Hold Cleaning Effluent	227
Figure 3.5.1. Box and Dot Density Plot of Pathogen Indicator Values Measured in Samples of
             Graywater	237
Figure 3.5.2. Box and Dot Density Plot of Potential Hazard Quotients for Pathogen Indicators
             Measured in Samples of Graywater	238
Figure 3.5.3. Box and Dot Density Plot of Classical Pollutant Concentrations/Values Measured
             in Samples of Graywater	242
Figure 3.5.4. Box and Dot Density Plot of Potential Hazard Quotients for Classical Pollutants in
             Samples of Graywater	243
Figure 3.5.5. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Graywater	250
Figure 3.5.6. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Graywater	251
Figure 3.5.7. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Graywater	252
                                          xn

-------
Figure 3.5.8. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples of Graywater	253
Figure 3.5.9. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of
             Graywater	256
Figure 3.5.10. Box and Dot Density Plot of Potential Hazard Quotients for Nutrients in Samples
             of Graywater	257
Figure 3.6.1. Box and Dot Density Plot of Classical Pollutant Values Measured in Samples of
             Inboard Propulsion Engine Effluent	268
Figure 3.6.2. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Inboard Propulsion Engine Effluent	274
Figure 3.6.3. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Inboard Propulsion Engine Effluent	275
Figure 3.6.4. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Inboard Propulsion Engine Effluent	276
Figure 3.6.5. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples of Inboard Propulsion Engine Effluent	277
Figure 3.6.6. Box and Dot Density Plot of SVOC Concentrations Measured in P.L. 110-299
             Study Samples of Inboard Propulsion Engine Effluent	282
Figure 3.6.7. Box and Dot Density Plot of Potential Hazard Quotients for SVOCs in P.L. 110-
             299 Study Samples of Inboard Propulsion Engine Effluent	283
Figure 3.6.8. Box and Dot Density Plot of Volatile Organic Compounds Concentrations
             Measured in P.L. 110-299 Study Samples of Inboard Propulsion Engine
             Effluent	288
Figure 3.6.9. Box and Dot Density Plot of Potential Hazard Quotients for Volatile Organic
             Compounds in P.L. 110-299 Study Samples of Inboard Propulsion Engine
             Effluent	289
Figure 3.6.10. Box and Dot Density Plot of Classical Pollutant Values Measured in Samples of
             Outboard Propulsion Engine Effluent	293
Figure 3.6.11. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Outboard Propulsion Engine Effluent	297
Figure 3.6.12. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Outboard Propulsion Engine Effluent	298
Figure 3.6.13. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Outboard Propulsion Engine Effluent	299
Figure 3.6.14. Box and Dot Density Plot of Potential Hazard Quotients for Total  Metals in
             Samples of Outboard Propulsion Engine Effluent	300
Figure 3.6.15. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
             Outboard Propulsion Engine Effluent	303
Figure 3.6.16. Box and Dot Density Plot of Volatile Organic Compounds Concentrations
             Measured in Samples of Outboard Propulsion Engine Effluent	306
Figure 3.6.17. Box and Dot Density Plot of Potential Hazard Quotients for Volatile Organic
             Compounds in Samples of Outboard Propulsion Engine Effluent	307
Figure 3.6.18. Box and Dot Density Plot of Classical Pollutant Values Measured in Samples of
             Generator Engine Effluent	311
Figure 3.6.19. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Generator Engine Effluent	314
                                         Xlll

-------
Figure 3.6.20. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Generator Engine Effluent	315
Figure 3.6.21. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Generator Engine Effluent	316
Figure 3.6.22. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples of Generator Engine Effluent	317
Figure 3.6.23. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
             Generator Engine Effluent	321
Figure 3.6.24. Box and Dot Density Plot of Potential Hazard Quotients for SVOCs in Samples of
             Generator Engine Effluent	322
Figure 3.6.25. Box and Dot Density Plot of VOC Concentrations Measured in Samples of
             Generator Engine Effluent	326
Figure 3.6.26. Box and Dot Density Plot of Potential Hazard Quotients for VOCs in Samples of
             Generator Engine Effluent	327
Figure 3.7.1. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
             Samples of Firemain Water	342
Figure 3.7.2. Box and Dot Density Plot of Total Metals Concentrations Measured in Samples of
             Firemain Water	343
Figure 3.7.3. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals in
             Samples of Firemain Water	344
Figure 3.7.4. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
             Samples of Firemain Water	345
Figure 3.7.5. Box and Dot Density Plot of Classical Pollutants Measured in Samples of Firemain
             Water	348
Figure 3.7.6. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
             Firemain Water	351
                                          xiv

-------
                                                                                Executive Summary
EXECUTIVE SUMMARY

       This report to Congress provides information collected by the U.S. Environmental Protection
Agency (EPA) on the types of wastewater discharged from commercial fishing vessels and
nonrecreational vessels less than 79 feet in length. The report also provides information on the primary
pollutant concentrations in these discharges and the likelihood of any resulting environmental impacts
based on rate, frequency, volume, and location discharged. This study was conducted to meet the
obligations of EPA under Public Law (P.L.)  110-299 (July 31, 2008). The law provided for a
temporary two-year moratorium on National Pollutant Discharge  Elimination System (NPDES)
permitting of discharges from commercial fishing vessels, regardless of size, and other nonrecreational
vessels less than 79 feet long that were subject to the 40 CFR 122.3(a) exclusion. Except for ballast
water discharges (evaluated and assessed elsewhere in other Agency reports), discharges from these
vessels are not currently covered under the EPA's Vessel General Permit (VGP). During the two-year
moratorium, which began July 31,  2008, EPA was required to study the relevant discharges. EPA
believes that the results from this study will serve as an objective  source of information that Congress
can use for statutory decision-making and will provide other readers valuable technical analyses of
these vessels' incidental discharges.  EPA requested public comment on this draft report in March,
2010: this final report incorporates changes made in response to these comments.

       As directed by Congress, the goal of the study was to obtain sufficient information to address
the following six core objectives:
       •  A characterization of the nature, type, and composition of discharges for representative
          single vessels and for each class of vessel.
       •  A determination of the volumes of those discharges, including the average volumes for
          representative single vessels and for each class of vessel.
       •  A description of the locations, including the more common locations, of the discharges;
       •  An analysis  of the nature and extent of the potential effects of the discharges, including
          determinations of whether the discharges pose risks to human health, welfare, or the
          environment, and the nature of those risks.
       •  A determination of the benefits to human health, welfare, and the environment from
          reducing, eliminating, controlling, or mitigating the discharges.
       •  An analysis  of the extent to which the discharges are currently subject to regulation under
          federal law or a binding international obligation of the United States.

                                              xv

-------
                                                                                  Executive Summary

       EPA designed and conducted a sampling program of discharges from commercial fishing
vessels and other nonrecreational vessels less than 79 feet in length to provide information to achieve
the first two objectives of the study. As required in P.L. 110-299, the study specifically evaluated the
impacts of any 1) discharge of effluent from properly functioning marine engines; 2) discharge of
laundry, shower, and galley sink wastes; and 3) other discharges incidental to these vessels' normal
operation. In addition, EPA supplemented sample collection and analysis with the collection of
contemporaneous information regarding the shipboard processes, equipment, materials, and operations
that contribute to the discharges, as well as the discharge rates, duration, frequency, and location.

       EPA found that commercial fishing vessels and nonrecreational vessels discharge a wide
variety of effluents during their normal operation. The Agency decided to focus its evaluation on
discharges from engines, bilges, fish holds, decks, and graywater activities because such discharges
can release oils, heavy metals, toxic organics, oxygen-depleting substances, nutrients, and endocrine-
disrupting compounds to ambient waters in quantities that may exceed National Recommended Water
Quality Criteria (NRWQC). In some circumstances, some of these vessel discharges to water bodies
have the potential to impact the aquatic environment.

Vessel Types

       EPA estimates there are between 118, 000 and 140,000 vessels in the United States subject to
the permitting moratorium (i.e., study vessels).1 Figure ES. 1 presents the estimated number of study
vessels by vessel types (service). Approximately one-half of these vessels are commercial fishing
vessels involved in activities such as fish catching (e.g., longliner, shrimper, trawler), fish processing,
fishing tending, and charter fishing. The other half is distributed among a variety of vessel classes,
including passenger vessels (e.g., water taxis, tour boats, harbor cruise ships, dive boats), utility vessels
(e.g., tug/tow boats, research vessels, offshore supply boats), and freight barges.
1 Based on the U.S. Coast Guard Marine Information for Safety and Law Enforcement (MISLE) database. See discussion in
Chapter 1 and Appendix B of this report for detailed discussions about vessel estimates and limitations of these estimates.
                                              xvi

-------
                                                                                       Executive Summary
(/)
                                                                        Vessels with unspecified length

                                                                        Vessels less than 79 feet in length
                                                                    140,000 - Total number of vessels,
                                                                    including vessels of unspecified length
                                                                    118,000 - Number of vessels with
                                                                    specified length less than 79 feet
                                                                                                   27,375
                                             20,953
                         8,016
                                              18,660
                                                   11
                                    768
                                    579
                    622
                      923
                      287
                      779
                           CD
                           E
                           CO
                           CO
                           -I—'
                           .c
                           O)
                           'CD
O)
'CD
CD
O)
S
CO
CL
            CD
CD "O
CO CD
CO «=
                       CO
           >
           o
                                                         £=>
CD
CO
00
.*:
c
CO
O.
!c
W

CO
                                                  Other non-recreational

                                                     Vessel Service
      Figure ES.l. Estimated Number of Study Vessels by Vessel Service (Type)

             To select specific vessel classes for sampling, EPA first developed a list of commercial vessel
      classes based on published information and industry experience. Next, due to limited time and
      resources, EPA eliminated those vessel classes believed to consist primarily of vessels greater than 79
      feet in length, with the exception of commercial fishing vessels. Examples of vessel classes eliminated
      because of their size included cable laying ships, cruise ships, large ferries, and oil and petroleum
      tankers. Next, EPA eliminated vessel classes that have historically been subject to NPDES permitting,
      including stationary seafood processing vessels and vessels that can be secured to the ocean floor for
      mineral or oil exploration. After screening out these vessel classes, EPA selected a subset of priority
      vessel classes to study, including commercial fishing boats, tug/tow boats, water taxis, tour boats,
      recreational vessels used for nonrecreational purposes, and industrial support boats less than 79 feet in
      length. EPA selected these vessel classes because they represent a cross section of discharges and have
      the potential to release a broad range of pollutants.

             EPA sampled wastewater discharges and gathered shipboard process information from 61
      vessels in nine vessel classes. Vessels were sampled in 15 separate cities and towns in nine states
      across multiple geographic regions, including New England, the Mid-Atlantic, the Gulf Coast, the
                                                    xvn

-------
                                                                                Executive Summary
Mississippi River, and Alaska. Table ES.l presents the types of vessels from which EPA sampled and
gathered shipboard process information for this study. EPA sampled more commercial fishing vessels
than any other vessel class due to the large number of fishing vessels subject to the P.L. 110-299
permitting moratorium. EPA also sampled a few recreational vessels used for commercial purposes
(e.g, towboats) to: 1) provide a semi quantitative comparison of the discharges from these vessels and
the other study vessels, and 2) collect additional information for EPA's related Clean Boating Act (P.L.
110-288) work.

Table ES.l. Vessels Sampled by EPA
Vessel Class
Fishing:
Gillnetter
Lobster Tank
Longliner
Purse Seiner
Shrimp Trawler
Tender
Trawler
Trailer
Tugboat
Water Taxi
Tour Boat
Tow/Salvage :
Research :
Fire Boat
Supply Boat
Recreational
Total
Number of Vessels Sampled

5
1
o
J
5
6
o
J
4
6
9
4
o
J
6
2
1
1
2
61
(1) Consists primarily of recreational vessels used for commercial or governmental purposes.
Sampled Discharges

       EPA sampled a total of nine discharge types from the various vessel classes listed above. These
included:

       •  Bilgewater
       •  Stern tube packing gland effluent
       •  Deck runoff and/or washdown
       •  Fish hold effluent (both refrigerated seawater effluent and ice slurry)
       •  Effluent from the cleaning offish holds
       •  Graywater
       •  Propulsion and generator engine effluent
       •  Engine dewinterizing effluent
       •  Firemain

       EPA typically sampled one to four discharge types on each vessel, depending on applicability,
accessibility, and logistical considerations.  Vessel discharge samples were analyzed for a variety of
                                             xvin

-------
                                                                                Executive Summary

pollutants, including classical pollutants such as biochemical oxygen demand (BOD5), total suspended
solids (TSS), residual chlorine, and oil and grease; nutrients; total and dissolved metals; volatile and
semivolatile organic compounds (VOCs and SVOCs); nonylphenols (used as surfactants in detergents),
which are endocrine-disrupting compounds; and pathogen indicators (i.e., E. coli, enterococci, fecal
coliforms).

Summary of Findings

       EPA found that the sampled discharges with the greatest potential to impact surface water
quality include deck washdown, fish hold effluent, graywater, bilgewater, and marine engine effluent.
Though these discharges may have the potential to impact surface water quality, particularly on a
localized scale, a screening level model of a hypothetical large harbor indicates that most of these
discharges in and of themselves would not cause exceedences of national water quality criteria in large
water bodies (see additional discussion under environmental impacts below). Review of available
literature also indicates that leachate from antifouling hull coatings used on certain vessels to prevent
buildup of organisms, such as barnacles and algae, as well as underwater hull cleaning, likely impact
surface water quality in some situations.

       Deck washdown from utility vessels such as tug/tow boats, tour boats, water taxis, and supply
boats had elevated dissolved and total metal concentrations  (e.g., aluminum) likely associated with
particulate metal washing off metal decks or decks with significant metal components. Certain deck
washdown samples also contained pollutants such as BOD5, TSS, nonylphenols, total phosphorous,
and total residual chlorine, all of which are associated with detergents and disinfectants.

       Fish  hold effluent, which is either refrigerated seawater or ice slurry water found on fishing
boats, had BODs and chemical oxygen demand (COD) concentrations that were several times higher
than concentrations typically measured in raw domestic sewage. Nutrient levels in many fish hold
effluent samples were also similar to the concentrations normally found in raw domestic sewage, and
ammonia nitrogen was  occasionally detected at concentrations acutely toxic to aquatic life. While
small fishing boats periodically discharge only a few hundred gallons offish hold wastewater, large
fishing vessels, such as offshore trawlers,  can discharge thousands of gallons offish hold wastewater
in a matter of minutes.

       Most fishing vessel owners also clean the fish hold tanks with a detergent and/or disinfectant
after the fish have been off-loaded. Detergents are suspected of containing nonylphenols, which are
endocrine-disrupting compounds. Disinfectants such as chlorine bleach contain high concentrations of
total residual chlorine, which is toxic to aquatic organisms. The samples offish hold cleaning effluent
contained nonylphenols and total residual chlorine, along with the same pollutants measured in the fish
hold effluent.

       Galleys, sinks, showers, and laundry facilities onboard commercial vessels generate graywater,
which is typically discharged overboard. Graywater volumes vary considerably depending on the class
of vessel and its intended use, vessel size, the number of crew and passengers onboard, and the types
of gray water-generating activities. Pollutants associated with the various graywater sources depend on

                                             xix

-------
                                                                                Executive Summary

a variety of factors, such as the amount of food waste flushed into the graywater system, the level of
soiling on clothing being washed in the onboard laundry, and the use of showers. EPA did not sample
graywater mixed with sewage, so the results for this study are for graywater only. EPA's sampling data
found pathogens to be the primary pollutant of concern in graywater. The sampling data show that at
least one of the pathogenic organisms (fecal coliforms, enterococci, andE. coif) was found in all
graywater samples, and that levels of these indicators in most of these samples exceeded the water
quality benchmarks, some by as much as four orders of magnitude.

       Bilgewater effluent consists of the water that collects in the bottom of the vessel from sources
such as precipitation and spray, fuel  spills, leaking sewage and graywater piping, condensates, and
deck washing. Bilgewater contained the greatest variety (although not necessarily the highest
concentrations) of priority pollutants, including both total and dissolved metals, VOCs, and SVOCs. It
also contained pathogenic bacteria, nonylphenols, sulfide, total phosphorous, BOD5, TSS, and residual
chlorine. Both total arsenic and dissolved copper concentrations in bilgewater were consistently above
the most conservative screening benchmarks (e.g., EPA's 2006 NRWQC), and total arsenic
concentrations were nearly 1,000 times the safe human health standard.

       Propulsion and generation engine effluent varied dependent upon the type of engine. EPA
found that inboard propulsion engines discharge more pollutants in their cooling water than outboard
propulsion engines or generators. EPA also found that VOCs and SVOCs are the primary pollutants of
concern found in marine engine cooling water discharges. These pollutants (e.g., benzene and several
PAHs, including some that are carcinogenic, or cancer causing) are present in fuels and are products of
incomplete combustion. Dissolved copper was also measured in most inboard engine effluents at
concentrations that exceed the NRWQC. Some vessel owners in cold climates also add a solution of
propylene glycol (antifreeze) to the internal cooling system of inboard engines to protect them from
freezing during winter. In spring, the antifreeze solution may be discharged as the cooling system is
refilled with ambient water. EPA's sampling data showed that the spent antifreeze solution discharged
to surface water contained relatively high levels of metals, which are likely a result of corrosion within
the engine's cooling system.

       Stern tube packing gland effluent (from tug boats) and firemain discharges (limited to just two
tug boats, three tour boats,  and a fireboat) contained elevated levels of some metals (e.g., dissolved
copper, total aluminum, total arsenic). For both of these discharges (firemain in particular), the effluent
samples contained relatively small concentrations of pollutants, most of which could be attributed to
the ambient surrounding water predominating the discharge.  For example, stern tube systems have a
continual drip of ambient water while the shaft is turning to provide both cooling and lubrication for
the system. The source of the additional metals in stern tube packing gland effluent is likely
mechanical system wear or lubricants used in the vessels' power trains.

       Although not directly sampled, EPA gathered existing information from the literature to
characterize discharges from antifouling hull coatings. Antifouling hull coatings are specialized paints
and other coatings intended to retard the growth of algae; weeds; and encrusting organisms, such as
barnacles and zebra mussels, on the underwater portion of vessel hulls. The coatings retard growth by

                                              xx

-------
                                                                                Executive Summary

continuously leaching biocides into surrounding waters. The most commonly used biocide is cuprous
oxide. The biocide enters the water column through both passive leaching and underwater hull
cleaning and can accumulate in the water of poorly flushed boat basins to levels that may harm marine
life. For example, the leaching of copper from antifouling hull coatings used on recreational boats is a
major source of copper pollution in several large boat basins in Southern California. Copper from
antifouling coatings has created documented water quality concerns in areas such as the Chesapeake
Bay; Port Canaveral, Florida; and several harbors in the state of Washington.

Environmental Impacts

       Using the results obtained from this study, EPA modeled a large hypothetical harbor to
evaluate the environmental impacts from the nine above mentioned vessel discharge types that EPA
sampled. The screening-level model indicated that the study vessels' discharges would not, in
themselves,  exceed the  aquatic life or human health NRWQC; however, the model did not account for
background  loadings. Certain pollutants (e.g., total arsenic, dissolved copper) are more likely to
contribute to a water quality criterion being exceeded under real-world conditions in large-scale water
bodies. Additionally, many pollutants present in the vessel discharges were at concentrations that
exceed an NRWQC at end of pipe; therefore, they have the potential to contribute to an environmental
effect in the receiving water on a more localized scale. Based on the  study results and literature
reviews, EPA believes that total arsenic and dissolved copper represent the  greatest environmental
concern in vessel discharges, and that they are more likely than other pollutants to contribute to
exceedances of water quality standards. This is especially true if there are other sources of these
pollutants (e.g., stormwater runoff) or high concentrations of vessels in confined waters, or the
receiving waters already have high background  concentrations.

       Other notable pollutants of concern were found in fish hold effluent from fishing vessels. These
pollutants include total  phosphorus, BOD, COD, reactive nitrogen compounds, and pathogens. These
pollutants can exacerbate eutrophication in bays and estuaries, leading to poor surface water quality.

Analysis of Applicable Regulations

       This report to Congress includes EPA's  analysis of existing laws and treaties that apply to
vessels and their discharges. This analysis describes numerous domestic laws, including the Act to
Prevent Pollution from  Ships (APPS); the Clean Water Act (CWA); the Federal Insecticide, Fungicide,
and Rodenticide Act (FIFRA); and the  Organotin Antifouling Paint Control Act (OAPC). It also
summarizes key elements of several international treaties, including the International Convention for
the Prevention of Pollution from Ships (MARPOL 73/78), the International Convention on the Control
of Harmful Anti-Fouling Systems  on Ships, the International Convention on the Prevention of Marine
Pollution by Dumping of Wastes and Other Matter (London Convention), and the International
Convention  on Oil Pollution, Preparedness, Response and Cooperation (OPRC). The purpose of this
analysis is to summarize these existing regulations  and international  obligations and examine the
extent to which these discharges are subject to these obligations.
                                             xxi

-------
                                                                                Executive Summary

Conclusion

       Some vessel discharges from commercial fishing vessels and commercial vessels less than 79
feet in length may have the potential to impact the aquatic environment and/or human health. As noted
above, using the results obtained in this study, EPA modeled a large, hypothetical harbor to evaluate
how the nine vessel discharge types EPA sampled may impact water quality. Based on this evaluation,
EPA determined that the incidental discharges from study vessels to a relatively large water body are
not likely to solely cause an exceedance of any NRWQC.  This finding suggests that these discharges
are unlikely to pose acute or chronic exceedances of the NRWQC across an entire large water body.
However, since many of the pollutants present in the vessel discharges were at end-of-pipe
concentrations that exceeded an NRWQC, there is the potential for these discharges to contribute a
water quality impact on a more localized scale.  The study results indicate that total arsenic and
dissolved copper are the most significant water quality concerns for the study vessels as a whole, and
that they are more likely than other pollutants to contribute to exceedances of water quality criteria.
This is especially true if there are high concentrations of vessels in confined waters or other sources of
pollutants or the receiving water already has high background concentrations.

      Like an individual house in an urban watershed, most individual vessels have only a minimal
environmental impact. As in urban waters, however, the impacts caused by these vessels are
potentially  significant where there is high vessel concentration, low water circulation, or there are
environmentally stressed water bodies.  Targeted reduction of certain discharges or pollutants in
discharges  from these vessels in waters sensitive to the introduction of pollutants from vessels may
result in important significant environmental benefits to those waters.
                                             xxn

-------
                                                                           Chapter 1 - Introduction to Report
CHAPTER 1
INTRODUCTION TO THE REPORT
1.1.   CONGRESSIONAL STUDY CHARGE

       On July 31, 2008, Public Law (P.L.) 110-299l was signed into law. It provides a two-year
moratorium for nonrecreational vessels less than 79 feet in length and all commercial fishing vessels
regardless of length, from the requirements of the National Pollutant Discharge Elimination  System
(NPDES)2 program to obtain a permit for discharges incidental to the normal operation of those
vessels.3  Additionally, P.L.  110-299 directs the United States Environmental Protection Agency (EPA)
to study the environmental impacts of discharges incidental to the normal operation of those vessels.
Specifically, the law directs the agency to study and evaluate the impacts of:

    (1) Any discharge of effluent from properly functioning marine engines
    (2) Any discharge of laundry,  shower, and galley sink wastes
    (3) Any other discharge incidental to the normal operation  of a vessel

       Congress mandated that EPA include the following elements in the study:

    (1) Characterizations of the nature, type, and composition of the discharges for:
           a.  Representative single vessels
           b.  Each class of vessels
    (2) Determinations of the volume (including average volumes) of those discharges for:
           a.  Representative single vessels
           b.  Each class of vessels
    (3) A description of the locations (including the more common locations) of the discharges.
    (4) Analyses and  findings as to the nature and extent of the potential effects of the discharges,
       including determinations of whether the discharges pose a risk to human health, welfare, or the
       environment,  and the nature  of those risks.
    (5) Determinations of the benefits  to human health, welfare, and the environment from reducing,
       eliminating, controlling, or mitigating the discharges.
    (6) Analyses of the extent to which the discharges are currently subject to regulation under federal
       law or a binding international obligation of the United States.
1 P.L. 110-299, along with its companion law for recreational vessels, P.L. 110-288 ("The Clean Boating Act") are presented
in Appendix C of this report.
2 The NPDES program requires a permit when a point source discharges a pollutant to waters of the US. A NPDES permit
contains conditions and limitations on the rates, concentrations, and mass of a pollutant that can be discharged to a water
body. The limitations are based on available pollution control technologies and water quality standards that are established to
protect the designated uses of a water body, such as fishing or swimming.
3 Although this report focuses on the discharges from vessels subject to the moratorium, the Agency became aware during
interaction with congressional staff that some members may be interested in additional information on discharges incidental
to the normal operation of a larger universe of vessels—in particular, vessels currently subject to the NPDES General Permit
for Discharges Incidental to the Normal Operation of a Vessel ("Vessel General Permit"). Therefore, EPA has included some
additional information and analysis regarding those vessels where possible.

                                                 1

-------
                                                                         Chapter 1 - Introduction to Report
       The law expressly excludes certain discharges from the scope of the study: discharges from
vessels owned and operated by the Armed Forces;4 discharges of sewage5 from vessels, other than the
discharge of graywater from vessels operating on the Great Lakes; and discharges of ballast water.

       EPA conducted the  study required by P.L.  110-299 and is publishing this report to present its
findings. Due to the accelerated timeframe required to complete the study, EPA designed this analysis to
be accomplished quickly with existing resources. Limitations in the study design are discussed in
Chapter 2 of this report. Due to these factors, EPA focused its sampling efforts on the vessels that P.L.
110-299  specifically exempted. EPA henceforth refers to these vessels and vessel types as study vessels.
EPA sampled discharges from a few other vessel types, including commercial vessels that were
manufactured primarily for pleasure, where resources and logistics allowed.

1.2.   ORGANIZATION OF THIS REPORT

       The report is organized into seven chapters. In Chapter 1, EPA describes the universe of vessels
with discharges subject to the study, the types of discharges  generally thought to originate from those
vessels, and the types of pollutants or other constituents generally found in those vessel discharges. In
Chapter 2, EPA discusses the methods for sampling, the types of vessels sampled, the Quality Assurance
and Quality Control (QA/QC) measures taken in the course  of sampling, and the limitations of this
study. Chapter 3 is the most technical portion of the report, presenting the results from EPA's sampling
and other information gathered from literature reviews about the vessel discharges. Chapter 4 presents
the results of EPA's screening-level model, which was designed to look at the large-scale, cumulative
impacts of these vessel discharges on large harbor or estuarine systems in order to provide an initial
evaluation of the threat the  discharges pose to these ecosystems. Chapter 5 discusses the results and
identifies those key areas where EPA found discharges most likely to be a concern to human health,
welfare, or the environment. Chapter 6 provides a summary  of federal law and binding international
obligations to which discharges within the scope of the study are potentially  subject. To a certain extent,
Chapter 6 also discusses discharges described in the study that might be beyond the  scope of the
permitting moratorium in some circumstances.  Chapter 7 lists report references.

1.3.   CLASSES OR TYPES OF VESSELS

       The study required by P.L. 110-299 could potentially include numerous  classes or types of
vessels that vary greatly in  size. The smallest vessels include recreational boats used for commercial
purposes, which can be less than 20  feet in length. The largest vessels,  such as super oil tankers, can be
more than 1,200 feet in length. Characteristics of these vessels, including construction material,
designed purpose, onboard  activities, crewing requirements, engine type and power, and days in
4 The Clean Water Act defines "vessel of the Armed Forces" as any vessel owned or operated by the Department of Defense,
other than a time or voyage chartered vessel; and any vessel owned or operated by the Department of Transportation that is
equivalent to one owned by the Department of Defense. 33 U.S.C. § 1322(a)(14).
5 "Sewage" is defined as "human body wastes and the wastes from toilets and other receptacles intended to receive or retain
body wastes except that, with respect to commercial vessels on the Great Lakes, such term shall include graywater." 33
U.S.C. § 1322(a) (6).

-------
                                                                         Chapter 1 - Introduction to Report
operation vary widely. Consequently, the types and volumes of discharges generated by these different
classes or types of vessels also vary to a great extent.

       EPA identified many classes or types of nonrecreational vessels in the development of the 2008
Vessel General Permit (VGP). Examples include tank ships that transport large volumes of bulk liquids,
container ships that transport containerized cargo, barges that transport bulk goods, and large cruise
vessels that transport hundreds or thousands of passengers. In the VGP, EPA defines a "Cruise Ship" as
a passenger ship that is used commercially for pleasure cruises and provides overnight accommodations
to passengers. In a separate study, EPA prepared an extensive cruise ship discharge assessment report
characterizing five different discharge types from these vessels.6

       The moratorium of P.L. 110-299 applies to discharges from nonrecreational vessels less than 79
feet in length and all commercial fishing vessels. For some vessel classes or types, such as barges or
cruise ships, the  majority of that class or type are vessels longer than 79 feet. For other classes or types,
such as container ships or oil tankers, all the vessels would be expected to be longer than 79 feet. EPA
did not include such vessel classes in this study, as resources did not allow for representative sampling
of the larger vessels to provide an assessment of the discharges from those classes and still adequately
sample and assess the vessels specifically exempted by P.L. 110-299. In this study, EPA focused on
sampling discharges from the most prevalent classes or types of vessels defined by the moratorium
parameters, but sampled other vessels if the opportunity presented itself. The following subsections
briefly describe key characteristics of some of the vessels considered for sampling in the study, but this
list is not intended to be comprehensive.

   1.3.1.  Commercial Fishing Vessels

       As defined in P.L. 110-299, commercial fishing vessels are vessels that commercially engage in
the catching, taking, or harvesting offish or an activity that can reasonably be expected to result in the
catching, taking, or harvesting offish.  Commercial fishing vessels include any vessels harvesting fish,
crab,  lobster, shrimp, or other aquatic organisms for commercial sale. Commercial fishing vessels may
employ various methods of collection including nets, trawls, traps, or hook-and-line to capture the target
species.

       The commercial fishing industry is highly diverse, spanning a wide array of ocean and nearshore
conditions, differing by both region and fishery. For example, the State of Alaska alone manages 68
fisheries, and there are over three hundred combinations of species, gear, and regions. Approximately
half of the nearly 10,000 Alaska State permitted vessels are endorsed for only one fishery; vessels
working multiple distinct fisheries can be set up in a totally different manner depending on the target
species (United Fishermen of Alaska, 2010 and Alaska Trailers Association, 2010).  Types of fishing
vessels include, but are not limited to:
6 This report is available at: www.epa.gov/owow/oceans/cruise_ships/pdf/0812cruiseshipdischargeassess.pdf.

-------
                                                                        Chapter 1 - Introduction to Report
       Purse Seiner: Purse seiners catch fish that school close to the surface, such as salmon, herring,
and sardines, by encircling them with a long net and drawing (pursing) the bottom closed to capture the
fish. The net is retrieved using a winch. When most of the net has been retrieved, with the remainder
laying in a "bag" alongside the vessel, the fish are dipped from the bag and into the vessel's hold. Seine-
caught fish are delivered whole. Purse seiners are limited by Alaska law to 58 feet to more precisely
manage their fishing effort (Alaska Department of Fish and Game, 2007).
                                  Purse Seiner Fishing Vessel.
       Troller: Troll vessels catch fish such as salmon and tuna by "trolling" bait or lures on lines
through feeding concentrations offish. Trolling vessels come in a variety of sizes and configurations,
ranging from small, hand-trolling skiffs to large, ocean-going power trolling vessels of 50 feet or more
in length (Alaska Department of Fish and Game, 2007). Hand trailers fish fewer lines and bring fish
aboard with hand operated gurdies or rod and reel; power trailers use hydraulic gurdies to land their
catch. Fish caught in the troll fishery are landed one at a time. Most troll-caught fish are immediately
gilled, gutted, and iced; some are landed in the round and held in slush ice; and a small component is
frozen at sea. In Alaska, the troll fishery occurs nearly year-round; however, the vast majority offish are
delivered between July 1 and September 30 (Alaska Trailers Association, 2010).
                                               4

-------
                                                                        Chapter 1 - Introduction to Report
                                        Troller Fishing Vessel.

       Crabber/Lobster: Crabbers and lobster boats target crabs (Dungeness, King, Tanner, and Blue)
and lobsters using twine or wire-meshed steel pots (traps). A line extends from each pot to a surface
buoy that marks its location; a power winch is used to retrieve the pots. Baited pots are left to "soak" for
up to several days before retrieval. Once onboard, the pots are opened and sorted with legal
crabs/lobsters retained in aerated seawater tanks. Crabs/lobsters are delivered live to shore stations and
retail outlets. Crab and lobster boats come in  a variety of shapes and sizes, from aluminum skiffs with
outboard motors that fish the inside waters, to seagoing vessels 100 or more feet in length that fish the
Bering Sea and the Gulf of Alaska for King Crab (Alaska Department of Fish  and Game, 2007).

       Gillnetter: Gillnetters catch a variety offish, such as salmon, herring,  and  chum, by  setting
curtain-like nets perpendicular to the direction in which the fish are traveling as they migrate along the
coast toward their natal streams. Nets can be  set in place, such as at or near the mouths of rivers, or
allowed to drift freely in deep water. Mesh openings are just large enough to allow the male  fish, which
are usually larger, to get their heads suck in the mesh. Gillnet vessels are usually 30 to 40 feet long and
are easily recognized by the drum on either the  bow or the stern on which the net is rolled. Net retrieval
is by hydraulic power which turns the drum. Fish are removed from the net by hand as the net is reeled
aboard. Gillnet-caught fish are usually iced for  delivery (Alaska Department of Fish and Game, 2007).

-------
                                                                        Chapter 1 - Introduction to Report
                                    Gillnetter Fishing Vessel.

       Trawler: Trawlers, also occasionally called draggers, typically catch large quantities of mid-
water species, such as pollock or pink shrimp,  and bottom-fish, such as flounder, by towing a large,
cone-shaped net. The net is retrieved using winches and rolled onto a drum. The end of the net ("bag")
holds the fish and is pulled onto the back of the vessel via a slanted stern ramp. Fish such as flounder
may be processed onboard into fillets or minced. Shrimp are sorted by size and species and frozen either
whole or headed. Trawlers range in size from small shrimp trawlers to large 600-foot ocean pollock
trawlers that possess onboard processing facilities (Alaska Department of Fish and Game, 2007).
                                     Trawler Fishing Vessel.
                                               6

-------
                                                                        Chapter 1 - Introduction to Report
       Longliner: Longliners catch fish (primarily halibut, black cod, swordfish, and tuna) via a
longline that is either laid on the bottom or suspended in the water column. Each longline can be up to a
mile in length and have thousands of baited hooks. A longline vessel typically sets several lines for a 24-
hour "soak." The lines are retrieved over a side roller with a power winch, and the caught fish are
packed in ice in the vessel's hold and are delivered whole and bled, whole and gutted, or headed and
gutted dressed (Alaska Department of Fish and Game, 2007). Longliners range in size from 18-foot
open skiffs to 80-foot schooners (Alaska Longline Fishermen's Association, 2010).
                                   Longliner Fishing Vessel.
       Fishing Dredge: A fishing dredge, also known as a scallop dredge or oyster dredge, is a device
that is towed along the bottom of the sea by a fishing vessel to collect scallops, oysters, clams, crabs,
and even in some cases, sea cucumbers. The dredge is winched up into the vessel and emptied onto the
deck. Dredge boats used to collect clams, oysters, and crabs in near-shore estuarine waters range from
24 to 50 feet long. Large off-shore dredges used to collect sea scallops can be as long as 190 feet.

       Fish Tender: A fish tender vessel supports fishing vessels by providing supplies and storing,
refrigerating, or transporting fish, fish products, or other materials.

-------
                                                                        Chapter 1 - Introduction to Report
                                         Tender Vessel.
    1.3.2.  Tugs/Towing Vessels

       Tugboats and towboats serve many functions and include vessels that operate solely in river
systems to ocean-going vessels. Tugboats can be utilized to push or tow barges and rafts. Tugboats often
assist larger vessels in docking maneuvers in harbors and are generally powerful relative to their size.
Although tugboats and towboats can be over 200 feet in length, many are in the 40- to 100-foot range.

-------
                                                                         Chapter 1 - Introduction to Report
                                       Tugboat/Push Boat.
    1.3.3.  Water Taxis/Small Ferries

       Water taxis and small ferries (or water busses) are vessels employed to provide public transport
of people from one location to another. Small ferries are vessels for hire that are designed to carry
passengers and/or vehicles between two ports, usually in inland, coastal, or near-shore waters. Many of
these vessels can be found in the coastal harbors of New York, Baltimore, Boston, San Diego, Seattle,
and others. The sizes of the vessels in this class vary and can surpass 100 feet in length.

    1.3.4.  Tour Boats

       This  vessel class encompasses a variety of vessels used for activities such as dinner cruises,
ecotourism,  whale watching  excursions, and sightseeing trips. Vessels in this class can range from small
private vessels with just a few passengers to large vessels carrying 50 or more passengers. Large tour
boats designed for extended  excursions can include galley facilities, overnight accommodations, and
laundry.

-------
                                                                        Chapter 1 - Introduction to Report
                         A Tour Boat (left) and a Water Taxi (right).
    1.3.5.  Recreational Vessels Used for Non-Recreational Purposes

       This class includes vessels manufactured as recreational vessels that are used for nonrecreational
purposes, such as law enforcement vessels, fire/rescue vessels, towing and salvage vessels (not to be
confused with towboats above), and research vessels. This vessel class encompasses a broad range of
vessel types and sizes. Under the Clean Boating Act of 2008 (P.L. 110-288), vessels that are
manufactured or used primarily for pleasure are "recreational vessels" subject to regulation under that
Act.
                        Recreational Vessel Modified for Towing/ Salvage.
                                               10

-------
                                                                        Chapter 1 - Introduction to Report
1.4.   VESSEL POPULATION

       As discussed in Section 1.1, P.L. 110-299 requires EPA to characterize discharges for
representative single vessels and for each class of vessel in terms of its nature; type and composition;
average volume; location; nature and extent of the potential effects; and benefits of reducing,
eliminating, controlling, or mitigating the discharges. EPA focused its attention on the commercial
fishing vessels and other nonrecreational vessels less than 79 feet in length covered by the moratorium.
Understanding the characteristics of discharges from all commercial fishing vessels and nonrecreational
vessels less than 79 feet in length requires considering these vessels in term of their number, vessel type,
onboard equipment, type of service, and area of operation. A brief overview of the analysis on vessel
type and size is presented in this section. A more complete analysis, including a discussion regarding
vessel location (which impacts the location of vessel discharges) and other vessel characteristics, is
presented in Appendix B of this report.

    1.4.1.  Vessel Characteristics Data

       In evaluating and describing the vessel population, EPA primarily relied on data gathered by the
U.S. Coast Guard. The primary data source used in the vessel population analysis is the U.S.  Coast
Guard's Marine Information for Safety and Law Enforcement (MISLE) database (USCG, 2009). MISLE
provides a wide range of information regarding vessel and facility  characteristics,  accidents, marine
pollution incidents,  and other pertinent information tracked by the U.S. Coast Guard. Where possible,
EPA complemented the  data available in MISLE with information obtained from published sources or
from consultations with U.S. Coast Guard personnel or port authorities.

       MISLE includes data for nearly  1 million vessels that operate in U.S. waters. The database
covers a wide ensemble  of vessels (e.g., recreational vessels, commercial fishing vessels, freight barges,
tank barges, tank ships, passenger vessels, utility vessels), and provides data on various characteristics
for each individual vessel. These data include:

       •  Identification number(s)
       •  Vessel category (e.g., class, type, subtype, service)
       •  Size (e.g., tonnage,  length, breadth, depth)
       •  Area of operation (e.g., hailing port, route type)
       •  Passenger and crew capacity
       •  Propulsion (i.e., method, engine type, and horsepower)
       •  Construction material and design (e.g., hull material, design type, hull configuration/shape)
       •  Year built or age

       In compiling MISLE data, the U.S. Coast Guard largely relies on documents submitted by vessel
owners  or operators in accordance with vessel documentation requirements (e.g., certificate of
documentation) or on information gathered by U.S. Coast Guard staff directly (e.g., during inspections,
vessel boardings, or accident investigations). While the database scope is not limited to a certain size or
class of vessel, the scope of the data included in MISLE is driven in part by the regulatory requirements
to which different types  of vessels are subject or by activities conducted by Coast Guard offices.

                                               11

-------
                                                                           Chapter 1 - Introduction to Report
MISLE, therefore, is generally most comprehensive for those vessels that are documented, state
registered, and/or subject to inspection requirements.

       While MISLE represents the most comprehensive national dataset currently available, it does not
capture the entire universe of vessels operated on U.S. waters. As discussed at greater length in
Appendix B, only limited information is available for certain classes of vessels, such as smaller
recreational vessels, due to the way in which vessel data are gathered. Most recreational vessels are not
subject to documentation or regular inspection requirements and thus are not captured in MISLE.7 The
MISLE data set currently contains approximately 700,000 recreational vessels, approximately 36
percent of which are documented vessels; the other recreational vessels are present in MISLE because of
other U.S.  Coast Guard activities, such as boardings, nonmandatory inspections (e.g., voluntary
inspection program), or incident investigations.8 Shortcomings of the database mostly regard small
recreational vessels. Since recreational vessels are covered separately under the Clean Boating Act of
2008 (P.L. 110-288) and are therefore not the primary focus of this report, EPA believes that data
limitations do not preclude the use of the MISLE data for the current analysis to generally describe the
characteristics of study vessels.

       Additionally, while MISLE captures a wide range of characteristics for each vessel, the
information is at times incomplete (e.g.,  length may be missing or recorded as zero) or may be outdated
(e.g., a vessel may no longer be operating while its status in the database remains "active").
Consequently, the information provided by the database should be seen as approximate and as indicative
of the general characteristics of different populations or classes of vessels that operate in U.S. waters.

    1.4.2.  Overview of Vessel Universe

       Information is provided in MISLE for a total of 993,863 vessels. Based on information recorded
in the database,  976,649 of these vessels are presumed currently operational, of which 918,469 vessels
are identified as U.S.-flagged vessels (referred to as "domestic" vessels in the remainder of the
section).9'10 Nearly 80 percent of the 918,469 operational domestic vessels recorded in MISLE are
recreational vessels  (722,522 vessels), while 7.6 percent are identified as commercial fishing vessels.
The remainder of the MISLE universe is composed of other types of nonrecreational vessels
(10.5 percent) such as freight and tank barges and ships, passenger vessels, and utility vessels, and
vessels of unspecified service (3 percent). Figure 1.1. presents the MISLE population of operational,
domestic vessels for all vessel  service categories, excluding recreational vessels. While the P.L. 110-299
7 While the number of recreational vessels recorded in MISLE is high (over 700,000), the database accounts for only a small
fraction of the 16.9 million recreational vessels estimated to operate in U.S. waters, according to EPA's Economic Impact
Analysis of the Recreational Vessel Permit (USEPA, 2008a) and to the National Marine Manufacturers Association's
(NMMA's) 2007 U.S. Recreational Boat Registration Statistics (NMMA, 2009).
8 Personal communication with U.S. Coast Guard Representative, LCDR Scott Muller, on May 15, 2009.
9 Approximately 355,000 vessels do not provide a vessel status and 5,000 have an "unknown" status. Following guidance
from a Coast Guard representative (Source: Personal email communication with Harold Krevait of the U.S. Coast Guard.
March 13, 2009), EPA assumed that these vessels are currently operational.
10 This count is based on the flag of the vessel. However, the MISLE database records a U.S. hailing port for some vessels
that are foreign flagged. Additionally, approximately 57,000 vessel records do not identify the vessel flag. EPA assumed that
these are domestic vessels.

                                                 12

-------
                                                                               Chapter 1 - Introduction to Report
moratorium will generally apply to discharges from the vessel service categories shown in the figure,
many of the vessels presented in Figure 1.1. are not subject to the P.L. 110-299 permitting moratorium
since the law is limited to commercial fishing vessels (regardless of size) and other nonrecreational
vessels 79 feet or less. Approximately one-third of the operational, domestic, nonrecreational vessels are
commercial fishing vessels. The next largest vessel service category is freight barges, with
approximately 24 percent of vessels; however, many of these barges may exceed the 79-foot length
restriction.
      • Commercial Fishing Vessel
       69,944
       35%
                                                                                     D Freight Barge
                                                                                    r 46,129
                                                                                   / D Freight Ship
                                                                                  If i%55
                                                                                   /
                                                                                      Passenger Vessel
                                                                                  / r  22,851
                                                                                  ' /  12%

                                                                                  I/ DPublic Vessel, Unclassified
                                                                                    ,-  1,323
                                                                                   /  1%

                                                                                    • Tank Barge
                                                                                   ~- 7,813
                                                                                  \   4%

                                                                                  \ \ Blank Ship
                                                                                  \ \-  528
                                                                                   \   0%
                                                                                   \
                                                                                   \ • Utility Vessel
                                                                                    ^  16,338
Source: U. S. Coast Guard, MISLE database, 2009.
Note; The chart includes all commercial fishing vessels recorded in MISLE (69,944) and all 96,637 other non-recreational vessels, regardless of length.

Figure 1.1. MISLE Population of Operational, Domestic Non-Recreational Vessels by Vessel
Service11'12

        Table 1.1 further characterizes the vessel population in terms of length greater than or equal to or
less than 79 feet within each vessel service category. As shown in both Table 1.1 and Figure 1.1., the
vast majority of vessels contained in MISLE are less than 79 feet in length. For example, nearly
77 percent of commercial fishing vessels (54,176 vessels out of 69,944) recorded in MISLE have a
11 This figure does not include the 722,522 recreational vessels included in the MISLE population of operational, domestic
vessels.
12 Approximately 74,000 vessels have a vessel service indicated as "unclassified", "unknown", or "unspecified" in MISLE. In
approximately 44,000 of those instances, EPA was able to assign a vessel service for the purpose of this analysis based on
information provided in other data fields (i.e., using vessel class, vessel type, or vessel subtype information).
                                                    13

-------
                                                                         Chapter 1 - Introduction to Report
length less than 79 feet.13 Vessels less than 79 feet also are a vast majority (94 percent) of the
recreational vessels. Only the other nonrecreational vessel service category counts a majority of vessels
79 feet or longer.
Table 1.1. Population of Operational, Domestic MISLE Vessels by Vessel Length

Greater than or Equal to 79 ft
Less than 79 ft
ZeroorNull(1)
Total
Recreational
2,256
676,915
43,351
722,522
Commercial Fishing
2,231(2)(3)
54,176
13,537
69,944
Other Non-Recreational
54,142
32,799
9,696
96,637
Unspecified
1,991
15,011
12,364
29,366
Source: U. S. Coast Guard, MISLE database, 2009
(1) MISLE indicates a length of zero or the vessel length field is blank.
(2) A separate estimate provided by U.S. Coast Guard personnel suggests that commercial fishing vessels 79 feet
long or greater number approximately 1,800 to 1,900 vessels.14
(3) Columns with yellow background represent study vessels.
       Recreational vessels are generally excluded from many parts of our analysis because a separate
act (the Clean Boating Act of 2008 (P.L. 110-288)) exempts discharges incidental to the normal
operation of these vessels from NPDES permitting requirements. The Clean Boating Act defines
recreational vessels as those that are either 1) manufactured or used primarily for pleasure or 2) leased,
rented, or chartered to a person for the pleasure of that person. Furthermore, vessels that are subject to
U.S. Coast Guard inspection and that are either engaged in commercial use or that carry paying
passengers are not considered recreational vessels under the Clean Boating Act. This definition does not
necessarily correspond to the service categories used in MISLE to identify recreational versus
nonrecreational vessels because MISLE categories are based on the type of service the vessel is used for
rather than original manufacture purpose. There are additional instances in which MISLE may differ
from Clean Boating Act definitions, making the distinction between vessels listed in the database that
are within and outside the scope of this study not always clear. For example, certain vessels that would
appear to fall under the Clean Boating Act definition of recreational vessels because they are described
as uninspected vessels carrying fewer than 6 passengers are classified as "passenger vessels" in MISLE;
however, MISLE does not provide further specifications on whether the vessels are leased, rented or
chartered to a person or whether they carry paying passengers. A second example would include charter
fishing vessels. Often, these vessels are manufactured or used primarily for pleasure, or leased, rented,
or chartered to a person for the pleasure of that person.  Many are not inspected by the US Coast Guard.
Charter fishing vessels which are not inspected are exempted from NPDES permitting requirements by
the Clean Boating Act (P.L. 110-288).  Other charter fishing vessels are inspected by the US Coast
Guard.  These inspected, non-recreational vessels are not exempted from NPDES by the Clean Boating
Act, and are study vessels only if they are less than 79 feet.  Since EPA is unable to determine whether
certain charter fishing vessels are study vessels or Clean Boating Act vessels based on the information in
MISLE, all charter fishing vessels listed in MISLE are included in the following estimates of study
vessels.
13 According to a U.S. Coast Guard representative, the overall fraction of commercial fishing vessels that are less than 79 feet
in length is estimated to be approximately 95 percent (Personal communication with Jack Kemerer, Fishing Vessel Safety
Program, May 26, 2009).
14 Personal communication with Jack Kemerer, Fishing Vessel Safety Program, May 26, 2009.
                                                14

-------
                                                                            Chapter 1 - Introduction to Report
       1.4.2.1.    Study Vessel Type

       Once commercial, non-fishing vessels longer than 79 feet are removed from the analysis, the
relative makeup of the study vessels changes. EPA estimates there are between 118,000 and
140,000 vessels in the United States subject to the permitting moratorium established by P.L. 110-299.15
Figure 1.2. presents the estimated distribution of vessels within the study vessel population by vessel
service (type). Approximately one-half of these vessels are commercial fishing vessels involved in such
activities as fish catching (e.g., longliner, shrimper, and trawler), fish processing, fishing tenders, and
charter fishing. The other one-half are distributed among a variety of vessel classes, including passenger
vessels (e.g., water taxis, tour boats, harbor cruise ships, dive boats), utility vessels (e.g., tug/tow boats,
research vessels, offshore supply boats), and freight barges.
15 The range accounts for the exclusion and inclusion, respectively, of other non-recreational vessels for which MISLE does
not record the length or for which the recorded length is zero.

                                                 15

-------
                                                                                        Chapter 1 - Introduction to Report
      80,000
  I
  •s
  I
   E
   D
                                                                            Wesels with unspecified length
                                                                            • Vessels less than 79 feet in length
                                                                             140,000 - Total number of vessels,
                                                                             including vessels of unspecified length
                                                                             118,000 - Number of vessels with
                                                                             specified length less than 79 feet
                                                       Other non-recreational

                                                          Vessel Service

Note: The figure is based on operational, U.S.-flagged commercial fishing vessels (regardless of length) and other nonrecreational vessels
less than 79 feet in length.

Commercial fishing vessels also include fish processing vessels and fishing vessels. Passenger vessels include passenger (inspected),
passenger (uninspected), passenger barge (inspected), passenger barge (uninspected), and passenger ships. Public vessel, unclassified
includes military and other public service vessels. EPA notes that military vessels are specifically excluded in P.L. 110-299. Utility vessels
include towing vessels (i.e., tugs), school ships, research vessels/ships, mobile offshore drilling units, offshore vessels, offshore supply
vessels, oil recovery vessels, and industrial vessels. EPA notes that not all of all the vessels represented in this table are subject to the
moratorium at all times (e.g., small mobile offshore drilling units when engaged in industrial activity are required to have permits). Freight
barges include general barges, dry cargo barges, container barges, flat deck barges, and unspecified vessel types that operate in freight
barge service. Tank barges include bulk liquid (cargo) barges, bulk liquefied gas barges, and unspecified vessel types that operate in tank
barge service. Some vessel service categories did not fall into one of the listed categories. Therefore, based on the other classification fields
(class, type, subtype), EPA determined an appropriate service category.
Source: U. S. Coast Guard, MISLE database, 2009

Figure 1.2.  Number of  Study Vessels Recorded in MISLE, by Vessel Service (Type)

Commercial  Fishing Vessels

        As shown in Figure 1.2., approximately 70,000 commercial fishing vessels represent the largest
category of study vessels. Based on this information, EPA  sampled more commercial fishing vessels
than other nonrecreational vessels less than 79  feet in length (see discussion in Section 2.2.1). According
to the vessel service categories used by the U.S. Coast Guard in MISLE, "commercial fishing vessels"
                                                         16

-------
                                                                             Chapter 1 - Introduction to Report
are vessels involved in such activities as fish catching (e.g., longliner, shrimper, trawler), fish
processing, and charter fishing.16
       The U.S. Coast Guard generally describes commercial fishing vessels as including fishing
vessels, fish tender vessels, and fish processing vessels as follows:

       •   Fish processing vessel17 means a vessel that commercially prepares fish or fish products
           other than by gutting, decapitating, gilling, skinning, shucking, icing, freezing, or brine
           chilling.
       •   Fish tender vessel means a vessel that commercially supplies, stores, refrigerates, or
           transports fish, fish products, or materials directly related to fishing or the preparation offish
           to and from a fishing, fish processing, or  fish tender vessel or a fish processing facility.
       •   Fishing vessel means a vessel that commercially engages in the catching, taking, or
           harvesting offish or an  activity that can reasonably be expected to result in the catching,
           taking, or harvesting offish.

       While there is some overlap in service use for commercial fishing vessels and other vessel
categories, such as passenger vessels (e.g., charter fishing), EPA assumed that the categorization used in
MISLE generally follows the U.S.  Coast Guard definition of commercial fishing vessels.18

Other Nonrecreational Vessels

       Excluding the approximately 27,000 "unspecified" vessels shown in Figure 1.2., "passenger
vessels" have the second highest number of study vessels with approximately 21,000 vessels
(approximately 19,000 of these vessels have a length recorded as 79 feet or less). These vessels are
further divided into subtypes according to the types of activities in which they are involved (e.g., diving
vessels, charter fishing vessels, ferry, harbor cruise vessels, sailing vessels). Approximately
7,833 vessels are categorized as inspected under 46 CFR (7,753 under part T (small passenger vessels
under 100 gross tons), 69 vessels under part K (small passenger vessels carrying more than 150
passengers or with overnight accommodations for more than 49 passengers), and 11 vessels under part H
(passenger vessels 100 or more gross tons)19). The remaining 13,120 vessels are recorded as uninspected
passenger vessels.20
16 Several charter fishing vessels are categorized as "commercial fishing vessels" in MISLE even though they are generally
not considered commercial fishing vessels by the U.S. Coast Guard Fishing Vessel Safety Program. That program considers
these vessels to be passenger vessels (Source: Personal communication with Jack Kemerer, Fishing Vessel Safety Program,
May 26, 2009). According to the Coast Guard definition, the key difference between vessels formally classified as
commercial fishing vessels and recreational vessels or passenger vessels that may be used in fishing activities is whether the
catch is sold.
17 The moratorium provided by P.L. 110-299 applies only to discharges incidental to the normal operation of a vessel when
operating in a capacity as a means of transportation. EPA requires NPDES permits for seafood processing vessel discharges
when they are engaged in the processing of seafood (an industrial activity).
18 The MISLE classification also depends on the information provided directly by the vessel owner or operator on the
application for documentation or renewal (Source: Personal communication with Jack Kemerer, Fishing Vessel Safety
Program, May 26, 2009).
19 This last category appears to contain vessels potentially misclassified since their indicated gross tonnage is less than 100
gross tons.
20 The definition of passenger vessels used  in MISLE is broader than vessels subject to inspection under 46 CFR parts T, K,
and H. Therefore, depending on how these  vessels are operated (e.g., whether they carry paying passengers only or leased,

                                                  17

-------
                                                                               Chapter 1 - Introduction to Report
        The service category labeled "public vessel, unclassified" accounts for up to 600 study vessels
(e.g., lighthouse tender vessels, hospital ships, law enforcement vessels, ice breakers). The "utility
vessels" category covers remaining types of vessels, including tug/tow boats, school ships, research
vessels/ships, mobile offshore drilling units, offshore vessels,  offshore supply vessels, oil recovery
vessels, and industrial vessels. As many as 11,000 vessels are  classified as utility vessels in MISLE.21
Freight barges (4,288 to 8,016 vessels), freight ships (579 to 768 vessels), tank barges (67 to 622
vessels), and tank ships (49 to 179 vessels) account for the remaining nonrecreational study vessels.22
rented or are chartered to a person for the pleasure of that person), they could fall under the Clean Boating Act definition of
recreational vessels and would therefore be outside the scope of the permitting moratorium.
21 Some vessel service categories did not fall into one of the listed categories. EPA determined an appropriate service
category based on information provided in other vessel classification fields (class, type, subtype).
22 For each range, the minimum value represents the  number of vessels for which the length is non null or zero and is less
than 79 feet while the maximum value represents the total number of vessels when including vessels for which the length is
unspecified or zero.

                                                    18

-------
                                                                           Chapter 1 - Introduction to Report
       1.4.2.2.    Vessel Size

       Vessels can be characterized by size according to two metrics: length and gross tons. The two
metrics are related to each other (gross tonnage is a function of the ship's enclosed spaces as measured
to the outside of the hull framing), and Figure 1.3 presents a scatter plot of gross tons and lengths for
commercial fishing vessels and other nonrecreational vessels obtained from MISLE. In general, most
nonrecreational  vessels in MISLE have a length ranging between 26 and 50 feet, which translates into a
tonnage generally below 50 gross tons.  The 79-foot length threshold for other nonrecreational vessels
(the criterion for applicability of P.L. 110-299 moratorium) corresponds roughly to a tonnage of 150
gross tons. In Chapter 6 of this report, EPA uses this information in determining whether certain vessels
may be subject to regulation under federal law or a binding international obligation of the United States.
     1,000
                 Other Non-Recreational Vessels
                * Commercial Fishing Vessels
        10
                               10
     100
Gross Tons
1,000
10,000
Note: This chart is based on all operational, U.S.-flagged commercial fishing vessels and other nonrecreational vessels, regardless of
length.
Source: U. S. Coast Guard, MISLE database, 2009

Figure 1.3. Relationship Between Vessel Gross Tons and Length
                                                 19

-------
                                                                            Chapter 1 - Introduction to Report
       Approximately half of vessels documented in MISLE fall within the 26- to 50-foot-length
category, they have an average vessel length of 41 feet. Figure 1.4 and Figure 1.5 illustrate the
distribution of vessel length for commercial fishing vessels and other nonrecreational vessels in terms of
the vessel count (Figure 1.4) and cumulative distribution (Figure 1.5). In analyzing the cumulative
distribution of vessels by length (Figure 1.5), tank ships are the only vessel service category with a large
percentage of vessels longer than 300 feet.23 Several  other vessel service categories have a significant
fraction of vessels above the 79 feet threshold. Steps  in the cumulative distribution (Figure 1.5) indicate
common lengths for certain  categories of vessels: freight barges are generally around 200 feet in length,
while tank barges tend to be 200 feet or 300 feet in length. For almost all other vessel service categories
(commercial fishing vessels, passenger vessels, utility vessels and unspecified vessels), vessels less than
79 feet represent the majority of vessels within the overall population.
on nnn -,
18,000
16,000
14,000
I 12,000
to
to
2
"5 10,000
1
E
i 8,000
6,000
4,000
2,000

m- — • — Commercial Fishing Vessel JL J
» Freight Barge
Utility Vessel
r — X— Freight Ship
-*- UNSPECIFIED
— 1 — Tank Barge
	 Public Vessel, Unclassified
• r Passenger Vessel
/\
71 1
/ 1
I V It
x/A~\ \ A 79 feet
II 1 \/\^f\! \
"x^H^, .^ — . . s^\£\\ .... ./I
momomomomomo
T-TtT-TtT-Tt
T-T-T-T-CMCMCMCM
Vessel Length (feet)
Note: This figure is based on operational, U.S. -flagged commercial fishing vessels and all other nonrecreational vessels (no size exclusion).
Source: U.S. Coast Guard, MISLE database, 2009
Figure 1.4. Distribution of MISLE Vessels by Length and Vessel Service (Type)
23 Although a large percentage of tank ships are listed as greater than 300 feet long, this accounts for a very small number of
vessels when compared to the overall universe of vessels in the selected service categories; approximately 300 of the 391
tank ships that list a vessel length are longer than 300 feet.
                                                 20

-------
                                                                              Chapter 1 - Introduction to Report
     100%
      90%	
Commercial Fishing Vessel
Freight Barge
Utility Vessel
Freight Ship
UNSPECIFIED
Tank Ship
Tank Barge
Public Vessel, Unclassified
Passenger Vessel
          T-   Tt       T-
                            T-T-T-T-CMCMCMCMCOCOCO
                                               Vessel Length (feet)
Note: This figure is based on operational, U.S.-flagged commercial fishing vessels and all other nonrecreational vessels (no size exclusion).
Source: U. S. Coast Guard, MISLE database, 2009

Figure 1.5. Cumulative Distribution of MISLE Vessels by Length and Vessel Service (Type)
       As shown in the two previous figures, there is significant variability in vessel length across
categories of nonrecreational vessels. Most freight barges reported in MISLE are about 200 feet in
length and relatively few (10 percent) are under 79 feet in length.24 Hence, most freight barges are not
subject to the moratorium in P.L.  110-299 and are currently eligible for coverage under the VGP. In
contrast, the majority of utility vessels (e.g., towing vessels), passenger vessels, and commercial fishing
vessels overall are less than 79 feet in length. Figure 1.6 shows the distribution of all commercial fishing
vessels and only nonrecreational vessels less than 79 feet in length by length and vessel service
(focusing on the study vessels). The majority of commercial fishing vessels are relatively small
compared to other nonrecreational vessels such as barges or utility vessels, with 70 percent of
commercial fishing vessels for which the length is specified in MISLE in the 26- to 50-foot range. The
length of other nonrecreational vessels varies among the subcategories, with as many as 66 percent of
24 Freight barges less than 79 feet in length include a wide range of vessels used for freight barge service, as classified in
MISLE. These include, for example, barges operated by oil spill response companies or by dredging companies to transfer
recovered oil or dredged material.
                                                  21

-------
                                                                                 Chapter 1 - Introduction to Report
passenger vessels in the 26- to 50-foot range, compared to less than 1 percent of freight barges within
that same range (most freight barges less than 79 feet in length fall in the 50- to 79-feet range).
        100%
        90%  - -
        80%  - -
        70%  - -
        60%  - -
        50%  - -
        40%  - -
        30%  - -
        20%  - -
        10%  --
         0%
              Commercial
             Fishing Vessel
Freight Barge
Freight Ship
Passenger
 Vessel
Public Vessel,
 Unclassified
Tank Barge
Tank Ship
Utility Vessel
Unspecified
   DLess than 26
                5,631
                           52
                                      31
                                               1,489
                                                                     36
                                                                                         713
                                                                                                   5,598
   126-50
                39,262
                           214
                                     289
                                               13,496
                                                          41
                                                                    126
                                                                               23
                                                                                        3,966
                                                                                                   1,129
   D 50-79
                9,283
  4,021
                                     258
            3,674
                                                          16
                                                                    124
                                                                               23
                                           4,196
                                            1,284
   D79 or more
                2,231
                                                  NOT INCLUDED IN SELECTED POPULATION
Note: This figure is based on operational, U.S.-flagged commercial fishing vessels and other nonrecreational vessels less than 79 feet in
length.
The length field is not reported or provides a value of zero for approximately 36,000 vessels.
Source: U. S. Coast Guard, MISLE database, 2009
Figure 1.6. Distribution of Study Vessels by Length (in Feet) and Vessel Service (Type)

        Figure 1.7 presents the distribution of study vessels by gross tons and vessel service. Overall,
nearly 77 percent of study vessels  are less than 50 gross tons, while the remaining vessels generally fall
within the 50- to 300-gross-tons range. Very few vessels (less than  1 percent) within the selected vessel
population are greater than 300 gross tons. Note that some vessel service categories appear
underrepresented because the gross tons field is blank or is listed as zero in MISLE for approximately
56,000 vessels.
                                                     22

-------
                                                                            Chapter 1 - Introduction to Report
        100%
                                                            Figure insert: Distribution of Study Vessels
                                                       by Gross Tons and Vessel Service; for Size Categories
                                                            representing 10 Percent or Less of Vessels
Note: This chart is based on operational, U.S.-flagged commercial fishing vessels and other nonrecreational vessels less than 79 feet in
length.

The gross tons field is not reported or provides a value of zero for approximately 56,000 vessels.

Source: U. S. Coast Guard, MISLE database, 2009
Figure 1.7. Distribution of Study Vessels by Gross Tons and Vessel Service (for which gross ton
data are given in MSLE)

        To select specific vessel classes for sampling, EPA first developed a list of commercial vessel
classes based on published information and industry experience. Next, EPA eliminated those vessel
classes believed to consist of vessels greater than 79 feet in length, with the exception of commercial
fishing vessels. Examples of vessel classes eliminated because of their size include cable laying ships,
cruise ships, large ferries, and oil and petroleum tankers. Next, EPA eliminated vessel classes not
subject to VGP permitting, including stationary seafood processing vessels and vessels that can be
secured to the ocean floor for mineral or oil exploration (the CWA regulations separately require
NPDES permits for industrial  operations onboard vessels). After screening out these vessel classes, EPA
selected a subset of priority vessel classes to study, including commercial fishing boats, tug and tow
boats, water taxis, tour boats, recreational vessels used for nonrecreational purposes, and industrial
                                                  23

-------
                                                                         Chapter 1 - Introduction to Report
support boats less than 79 feet in length. EPA selected these vessel classes because they provide a cross
section of discharges and a broad range of potential pollutants.

       1.4.2.3.   Additional Vessel Characteristic Information

       Other vessel characteristics such as vessel age and engine power (horsepower ahead) likely
influence the characteristics and the volume of many vessel discharges. Intuitively, where a vessel is
located and operated can determine the impacts. Additionally, where there are more vessels, there is a
greater likelihood of cumulative impacts (e.g., where there are more vessels, there will be a greater
impact from vessel discharges).

       Appendix B presents additional vessel characteristic information, including summaries of vessel
subtypes, the hailing port of domestically flagged vessels, and information on construction and
propulsion of these vessels, including the vessel age and horsepower ahead. Appendix B also discusses
limitations in using the MISLE data. Appendix B lists the most common subtypes of vessel within each
vessel  type. For example, towing vessels are the most common type of utility vessel. Appendix B also
shows  where concentrations of vessel activities occur and what vessels are most predominant in those
assemblages. For example, the hailing port of New Orleans has the most registered vessels, including
significant numbers of commercial fishing vessels and other nonrecreational vessels. Finally, Appendix
B shows that most study vessels are relatively old, with the majority of them being more than 25 years
old. These analyses helped EPA qualitatively  and quantitatively analyze the cumulative impact of many
vessels' discharges (see Chapter 4 of this report), and to put the numbers and locations of study vessels
into perspective relative to other vessels, such as recreational vessels and other non-study vessels (e.g.,
nonrecreational, noncommercial vessels greater than 79 feet in length).
1.5.   DISCHARGES FROM VESSELS

       EPA developed a substantial list of discharges from vessels and pollutants of concern in each of
those discharges during the development and issuance process of the VGP in 2008. Starting with this
list, EPA developed a subset of discharges prevalent on fishing vessels and nonrecreational vessels less
than 79 feet in length that are expected to have pollutants of concern. The subset of discharges that EPA
selected included: bilgewater, deck washdown and runoff, propulsion engine effluent, generator engine
effluent, firemain systems, fish hold effluent, fish hold cleaning effluent, graywater, and shaft packing
gland effluent. While EPA did not sample antifouling hull-coating leachate, this discharge is discussed
as well because this is a significant discharge from many vessels and has been documented to cause
water quality impacts (see Section 3.2.8).

       EPA recognizes that there are additional discharges25 that also sometimes are present on study
vessels. Some of these were not conducive to sampling, such as cathodic protection, underwater ship
husbandry, and oil-to-sea interfaces. Some discharges are generally combined with other discharges and
25 EPA lists many discharges and descriptions of those discharges in the VGP and the accompanying fact sheet. Due to the
timeframe and resource limitations of this study, EPA chose to focus on the nine discharges that were a) conducive to
sampling and b) most likely to cause or contribute to impacts to human health, welfare, or the environment.

                                               24

-------
                                                                        Chapter 1 - Introduction to Report
are not typically available for independent sampling. An example of this is refrigeration system
condensate that is drained to the bilge. Other discharges are not expected to be commonly generated on
commercial fishing vessels or nonrecreational vessels less than 79 feet in length. These discharges are
typically associated with larger vessels,  such as those covered by the VGP, and were not sampled for in
this study due to resource limitations. Some examples include aqueous film-forming foam, distillation
and reverse osmosis brine, exhaust gas scrubber effluent, elevator pit effluent, and boiler/economizer
blowdown. A detailed discussion on the discharges EPA decided not to  sample is provided in Chapter 2.

   1.5.1.  Bilgewater

       Bilgewater is defined as the water that collects in the bottom of a vessel's hull. This includes
water from rough seas, rain, minor leaks (designed or accidental) in the  hull or stuffing box, condensate
from various types of equipment, spills onboard the  vessel, and leaks from pumps and seals. Bilgewater
can be found on almost every vessel; if too much water accumulates, it could threaten the safety and
stability of the vessel. For example, the U.S. Coast Guard requires that certain commercial fishing
vessels and fish-processing vessels have automated bilge pumping systems as part of their basic safety
features (46 CFR Part 28.255).

       A number of oily and non-oily wastewater sources sometimes drain intentionally or
unintentionally into the bilge. Oily wastewater sources include oil, fuel,  and antifreeze leaks from engine
and machinery operation and maintenance. To prevent floating oils typically found in bilgewater from
being discharged overboard, vessels can either use oil-adsorbent pads in the bilge compartment or pump
the bilgewater through a properly operating oil-water separation system  or oil absorbent filter prior to
overboard discharge.

       Non-oily wastewater sources include non-oily leaks from engine and machinery operation and
maintenance and various condensates. Vessels can have numerous sources of non-oily machinery
wastewater, including chilled water condensate drains, fresh- and saltwater pump drains, potable water
tank overflows, and leaks from propulsion shaft seals. Large vessels typically have separate systems to
collect non-oily machinery wastewater in dedicated  drip pans, funnels, and deck drains for subsequent
direct discharge. Small vessels  can also  generate non-oily machinery wastewater; however, these
wastewaters likely drain into the bilge.

   1.5.2.  Deck Washdown and Deck Runoff

       Deck washdowns are typically performed to  prevent slip and fall hazards; to prevent dirt, grit, or
other materials from harming the integrity of the deck surface; or to clean  the deck after pulling in a
catch or unloading cargo. Deck washdown is typically performed using  hoses and mops that move the
deck washdown water and cleaning agents (if any) to the scuppers through which the water is discharged
overboard.  Deck cleaning often occurs while the vessel is underway but is also performed pierside,
generally after loading or unloading catch or cargo.

       Deck runoff is typically related to either precipitation or surface  water spray that lands on the
deck and flows to the scuppers  where it  is discharged overboard. Operators of the vessel do not have

                                              25

-------
                                                                         Chapter 1 - Introduction to Report
control over the volume of discharge related to precipitation events or sea sprays, but they can minimize
the pollutants carried by the runoff by utilizing appropriate maintenance practices.

       Deck washdown and deck runoff have the potential to contain a variety of pollutants, including
oil and grease, nutrients, solids, metals, detergents, and solvents. Some or all of these pollutants could be
introduced to the deck from shipboard activities, storage of material on the deck, maintenance activities,
and the decking material itself.
                   Deck Washdown Activity of a Water Taxi (left) and a Towing and
                                         Salvage Vessel (right).

    1.5.3.  Engine Effluent

       Engines found on commercial vessels are typically used for two purposes: propulsion and
electricity generation. Engines used for vessel propulsion can be either outboard or inboard engines.
Outboard engines are self-contained units designed to be mounted outside the vessel hull at the stern
(rear) of the vessel. Inboard engines are enclosed within the hull of the vessel, usually connected to a
propulsion screw by a drive shaft. Outboard engines are typically fueled by gasoline, while inboard
motors can use either gasoline or diesel fuel. Gasoline or diesel engines can be either two stroke, which
require small amounts of oil to be mixed with the fuel to create a mixture that both lubricates and
provides combustion, or four stroke, which have separate lubrication systems.

       All combustion engines require cooling systems to remove  excess heat. Direct-cooled marine
engines draw raw water (either freshwater or seawater in which the vessel is floating) into the engine
and rely on the raw water to absorb the heat directly from the engine. Biocides sometimes are added to
the raw water to prevent biofouling of the heat exchange system (biofouling prevention). Indirect-cooled
marine engines use an enclosed cooling system that requires circulation of a freshwater-coolant solution
through the engine to absorb heat. The coolant solution passes through a closed heat exchanger where
the raw water absorbs the heat from the coolant solution and is then discharged.

       Vessels also use keel-cooling systems for indirectly cooling marine engines. A keel cooler is
essentially a heat exchanger mounted outside the vessel's hull beneath the waterline. Hot water from the
                                               26

-------
                                                                        Chapter 1 - Introduction to Report
marine engines is pumped through the keel cooler, which is in constant contact with the seawater. This
closed-circuit cooling system eliminates the need for an inboard heat exchanger, raw water pumps, and
strainers and does not result in a discharge.

       Some engines also use water to cool and quiet their exhaust, referred to as boat engine wet
exhaust. These engines inject spent cooling water from the engine into the exhaust stream, which results
in some of the gaseous and solid components of the exhaust being entrained into the cooling water
discharge.

       Vessels that require significant lighting or have electrical equipment, such as appliances and/or
electric motors, are likely equipped with engines used for electricity generation. Electrical generators on
these vessels are typically powered by diesel engines. The size of the electrical generators depends on
the electrical load requirements for the vessel, but could range from small generators used to power
navigation equipment and galley appliances to large generators used to power electric motors on deck
winches and cranes. Similar to vessel engines, electrical generators will require direct or indirect
cooling.
                      Collecting a Sample of Engine Effluent at Full Speed.
                                               27

-------
                                                                       Chapter 1 - Introduction to Report
                    Inboard Engine (left) and Outboard Engine (right)

    1.5.4.  Firemain Systems

       Some vessels are equipped with firemain systems to supply water for firefighting and to supply
water to other vessel systems. Vessels use either "wet type" or "dry type" firemain systems. The wet
type firemain piping is normally filled with water. Wet type systems are particularly used on vessels
where the firemain water is used frequently, typically for maintenance activities such as deck
washdown. In a dry type system, the piping is normally empty. Water is only introduced to the pipes
when actual firefighting takes place, or for testing or training.

       Aqueous film-forming foam (AFFF) can also be used on vessels as a fire suppression agent.
AFFFs are a combination of fluorochemical surfactants, hydrocarbon surfactants, and solvents (Koetter,
2008) that are injected into the water stream of a fire hose. These film-forming agents are capable of
forming water solution films  on the surface of flammable liquids, separating the fuel from the air
(oxygen). Systems that use AFFFs do not appear to be common on smaller vessels.
                               Firemain System on a Fire Boat.
                                              28

-------
                                                                        Chapter 1 - Introduction to Report
                                           Fire Boat.
    1.5.5.  Fish Hold and Fish Hold Cleaning Effluent (Refrigerated Seawater Discharge or Fish
          Ice Slurry Discharge)

       Commercial fishing vessels utilize different methods to keep seafood fresh after it is caught.
Most seafood is either dead when brought onboard or is killed shortly thereafter, before being stored in a
refrigerated seawater holding tank, with the exception of certain shellfish (e.g., crab, lobster), which
must be kept alive. The two most common methods of cooling seawater are by mechanical refrigeration
or by adding ice. Mechanical refrigeration is common on tenders, purse seiners, and trawlers, while
chipped and slurry ice tanks are more common on trailers, longliners, gillnetters, and some trawlers.

       For vessels with refrigerated seawater tanks, fish are typically extracted using a vacuum system
that removes both the fish and refrigerated seawater simultaneously. Any excess refrigerated  seawater
that is not required to assist in fish extraction is pumped overboard pierside. Vessels that use chipped or
slurry ice generally remove the seafood and then discharge the spent ice overboard pier side.
Occasionally, vessels that store their catch in ice slurry also use vacuum filtration systems (e.g., some
shrimping boats in the Gulf of Mexico). These discharges often contain pollutants generated by the
catch, such as biological wastes.

       Tanks used to keep lobster and crab catch alive pump surrounding water into the tank constantly
to maintain the highest water quality possible. The flow rate through these systems results in a nearly
continuous discharge offish hold effluent. Because the majority of the seafood product remains alive,
however, there is little biological decay or degradation in the tank. Furthermore, because these tanks
have reasonably rapid flushing times and a continuous discharge, there is a little accumulation of
pollutants.

       Fish holds are also often cleaned or disinfected by vessel crews between catches. To rinse the
tank, vessel crews use either municipal water from the pier or dock or they pump water from the
surrounding ambient water. Cleaning may simply involve rinsing the tanks with this water, or crews
                                               29

-------
                                                                       Chapter 1 - Introduction to Report
sometimes add detergents or disinfectants. Crews also often use scrub brushes to clean the walls and
floor of the fish hold to maximize the removal of organic material. Fish hold cleaning effluent is a
combination of residual fish hold water and ambient or municipal water and often contains soaps or
detergents.
                      Shoveling Fish Hold Ice Overboard From Ice Tank.
                           View of a Full Refrigerated Seawater Tank.
                                              30

-------
                                                                       Chapter 1 - Introduction to Report
    1.5.6.  Graywater

       Graywater is generated onboard vessels from domestic activities such as dish washing, food
preparation, laundry, and bathing. Graywater is discharged through either a single discharge port from a
collection system or through multiple, separate discharge ports for each graywater source (e.g., sink,
shower, washing machine). Graywater discharge is intermittent and occurs only when the specific
activity is performed. Most graywater processes use onboard potable water (service water).

       Smaller vessels can sometimes not generate any graywater. Many of these vessels are for day use
and do not provide any overnight quarters or heads (toilets). Smaller vessels that do generate graywater
(e.g., those that have accommodations, sinks, or showers) generally discharge graywater directly
overboard via ports typically located above the waterline. Most larger vessels used for overnight or
multiday travel have numerous graywater sources, including showers, bathroom and kitchen sinks, and
laundry. On these vessels, graywater discharges overboard by draining through gravity to either a
discharge port above the water line or to a small collection tank located in the vessel hull, where it is
immediately pumped to a discharge port above the waterline. Other vessels can collect their graywater
and treat it along with sewage in Marine Sanitation Devices (MSDs).

       Typical pollutants found in graywater often include metals, pathogens, total suspended solids,
biochemical oxygen demand, chemical oxygen demand, oil, grease, ammonia, nitrogen, and phosphates.
Graywater does not include sewage, or "blackwater", which is exclusively human waste from toilets and
urinals. Sewage is regulated under Section 312 of the Clean Water Act and 40 CFR Part 140 (see
Chapter 6 of this report for further discussion).
                            Collecting Graywater (Shower) Effluent.
    1.5.7.  Shaft Packing Gland Effluent

       For vessels with propeller shafts, a packing gland, or stuffing box, is used to provide a seal
around a propeller shaft at the point where it exits a boat's hull underwater.  This is a common method
for preventing water from entering the hull while still allowing the propeller shaft to turn. In a
conventional packing gland, the seal itself is provided by packing rings made of greased flax that is

                                              31

-------
                                                                         Chapter 1 - Introduction to Report
packed or wound tightly around the propeller shaft and compressed in place with a threaded nut and
spacer. The gland can also be fitted with an opening for periodic insertion of grease between the rings,
and sometimes includes a small grease reservoir.

       A packing gland packed with flax rings is designed to leak a small amount of water—a few
drops per minute—to provide lubrication when the shaft is turning.  Water that leaks through the seal
sometimes drips into a non segregated bilge or collects in a segregated area to avoid contact with oily
wastewaters. In the case of a segregated area, the water that collects (referred to as shaft packing gland
effluent) is automatically pumped overboard when levels reach a preset depth to prevent overflow.

   1.5.8.  Antifouling Hull Coatings-

       Vessel hulls are often coated with antifouling compounds to prohibit the attachment and growth
of aquatic life. Coatings are formulated for different conditions and purposes, and many contain
biocides. Those that contain biocides prevent the attachment of aquatic organisms to the hull by
continuously leaching substances into the surrounding water that are toxic to aquatic life. While a
variety of different biocides are used, the most commonly used is copper. Hull cleaning activities often
can cause additional releases of biocides, particularly if hulls are cleaned within the first 90 days
following application of the antifouling coating.

       A second metal-based biocide is organotin-based, typically tributyltin (TBT), which was
historically applied to vessel hulls. TBT and other organotins cause deformities in aquatic life, including
defects that disrupt or prevent reproduction. TBT and other organotins are also stable and persistent,
resisting natural  degradation in water bodies. As discussed in Chapter 6 of this report, the use of TBTs
and other organotins as biocides has been phased out on all vessels by domestic law and international
treaty.

1.6.   POLLUTANTS POTENTIALLY FOUND IN VESSEL DISCHARGES

       EPA developed groupings of pollutants of concern in the issuance process of the VGP in 2008. EPA
recognizes that while some discharges from all sizes of vessels are essentially the same, many will vary
due to the specific machinery and activities conducted on these vessels. EPA used slightly different
groupings of the pollutants from the discharges  sampled for this report to address differences from the
discharges covered by the VGP. The pollutants and constituents of concern are broken down into the
following groups: classical pollutants, nutrients, pathogen indicators, metals, volatile organic
compounds (VOCs), semivolatile organic compounds (SVOCs), and nonylphenols. Not all pollutants are
expected to be found in each discharge. For each discharge, EPA attempted to identify which pollutant
groups are of concern.
26 Though antifoulant hull coatings are present on some study vessels, particularly those operating in areas where there is a
significant potential for fouling, it was not feasible to sample discharges from these coatings for this study (see Chapter 2 for
further discussion).

                                               32

-------
                                                                        Chapter 1 - Introduction to Report
    1.6.1.  Classical Pollutants

       For purposes of this report, EPA uses the term "classical pollutants" for the following 14
pollutants: temperature; conductivity; salinity; turbidity; dissolved oxygen; total suspended solids (TSS);
biochemical oxygen demand (BOD); chemical oxygen demand (COD); total organic carbon (TOC); oil
and grease; pH; sulfide; and total residual chlorine (TRC). These include the CWA conventional
pollutants plus other common pollutants that are of general concern in a wide variety of contexts.

       Temperature changes can directly affect aquatic organisms by altering their metabolism, ability
to survive, and ability to reproduce effectively. Increases in temperature are frequently linked to
acceleration in the biodegradation of organic material in a water body, which increases the demand for
dissolved oxygen and can stress local aquatic communities. Thermal impacts from vessel discharges are
generally much smaller than those from traditional point sources, and the vessel discharge with the
greatest potential to alter receiving water temperature is engine cooling water.

       Conductivity and salinity measurements are related to ionic strength and can indicate what
specific ions are present in water or wastewater. Conductivity is a measure of the ability of water to pass
an electrical current. Conductivity in water is affected by the presence of inorganic dissolved solids (or
ions). Organic compounds like oil, phenol, alcohol, and sugar do not conduct electrical current very well
and therefore have a low conductivity when in water. Conductivity is also affected by temperature; the
warmer the water, the higher the conductivity. Salinity is a measure of the mass of dissolved salts (ions)
in solution. Ions commonly found in water include calcium, magnesium, potassium, and sodium cations
and bicarbonate, carbonate, chloride, nitrate, and sulfate anions. The average ocean salinity is
approximately 35 parts per thousand (ppt), while freshwater salinity is generally less than 0.5 ppt. The
salinity of brackish water, such as estuaries, is between 0.5 ppt and 17 ppt. Conductivity is a good
measure of salinity in water and vice versa.

       Both turbidity and TSS are assessments of the amount of suspended solids present in the water
column. Turbidity is an indicator of water clarity, measuring how much the material suspended in water
decreases the passage of light through the water. Higher turbidity increases water temperatures because
suspended particles absorb more heat. Suspended materials, also measured as the mass of TSS, can clog
fish gills, reducing resistance to disease in fish, lowering growth rates, and affecting egg and larval
development. As the particles settle, they can smother fish eggs and benthic macroinvertebrates on the
bottom substrate. Vessel discharges with relatively high turbidity and TSS concentrations include fish
hold effluent, bilgewater, graywater, and deck washdown.

       The oxygen content of water or wastewater is measured in its dissolved form as dissolved
oxygen (DO). Low DO levels (hypoxia) can impair animal growth or reproduction, and the complete
lack of oxygen (anoxia) will kill aquatic organisms. Organic material found in vessel discharges  (e.g.,
fish waste, bilgewater, graywater) that are easily biodegraded will result in depressed DO concentrations
in ambient receiving waters. The ability of the organic material in vessel discharges to biodegrade and
depress oxygen levels is measured as either BOD or COD. BOD measures the amount of oxygen used
by naturally occurring microorganisms to metabolize the organic material in the vessel discharge, while
COD measures the oxygen needed to chemically oxidize the organic material in the vessel discharge. If

                                               33

-------
                                                                         Chapter 1 - Introduction to Report
there is a large quantity of organic waste in water, there will also be a lot of bacteria present working to
decompose this waste. In this case, the demand for oxygen will be high (due to all the bacteria), so the
BOD level will be high. COD levels can often be correlated with BOD levels, though they are generally
higher because the measurement examines chemicals that are both biologically and chemically oxidized.
As the waste is consumed or dispersed through the water, BOD levels will begin to decline.

       Oil and grease are other known components of vessel discharges with potentially harmful
impacts to humans and to aquatic life. Oil and grease are measured using hexane extractable material
(HEM) and silica gel treated (SGT)-HEM. Vessels sometimes discharge oil, including lubricating oils,
hydraulic oils, and vegetable or organic oils, in everyday operation.  Oils produce a visible slick or
sheen27 on the water surface, which decreases natural oxygen transfer, resulting in depressed DO
concentrations. Also, oils might contain heavy metals and SVOCs, which can bioaccumulate in fish,
birds, marine mammals, and ultimately humans. Bilgewater, fish hold effluent (fish oils), and graywater
(galley wastewater) are the vessel discharges most likely to contain  oil and grease.

       The term pH is used to indicate the alkalinity or acidity of a  substance as ranked on a scale from
1.0 to 14.0. Substances with lower pH (i.e., less than 7) are acidic, while substances with higher pH (i.e.,
greater than 7) are basic.  pH affects many chemical and biological processes in the water. The largest
variety of aquatic animals prefers a range of 6.5 to 8.0. pH outside this range can reduce diversity
because it stresses the physiological systems of most organisms. Low pH can allow toxic elements and
compounds to become mobile and "available" for uptake by aquatic plants and animals. This can
produce conditions that are toxic to aquatic life, particularly sensitive species. Many vessel-cleaning
wastewaters can be either acidic (e.g., metal cleaners and tub, toilet, and sink cleaners) or basic (e.g.,
degreasers).

       Sulfide is a strong reducing agent typically generated during anaerobic decomposition of organic
materials. Sulfides are naturally present in groundwater as a result of leaching from sulfur-containing
mineral deposits. Surface water does not usually contain high sulfide concentrations.  Sulfide is a
pollutant that is commonly elevated in water distribution systems as well as sewers. Sulfur-reducing
bacteria, which use sulfur as an energy source,  are believed to be the primary producers of large
quantities of hydrogen sulfide. Ecologically, these bacteria are common in anaerobic environments (e.g.,
plumbing systems). For vessels, possible sources of sulfide include trace  constituents in the fuel,
products of incomplete combustion, or formations in anaerobic systems onboard the vessel. Sulfide
generated from anaerobic decomposition is suspected in graywater,  bilgewater, and fish holds.  Sulfide
may also be formed during fuel combustion in a vessel's engine. Sulfide, typically found as hydrogen
sulfide, poses a potential long-term hazard to aquatic life (USEPA, 1986b) at low concentrations.

       Chlorine is commonly used as a disinfectant in wastewater and drinking water. Chlorine,
measured as TRC, though toxic to humans at high concentrations, is of much greater concern to aquatic
species, which can experience respiratory problems, hemorrhaging,  and acute mortality. TRC is present
in potable water supplies, and consequently, any vessel  systems that use potable water could potentially
27 Visible slick or sheen means a "silvery" or "metallic" sheen, gloss, or increased reflectivity; visual color; iridescence, or oil slick on the
surface (58 FR 12507J.

                                               34

-------
                                                                         Chapter 1 - Introduction to Report
discharge TRC while conducting graywater activities and deck washing. Chlorine bleach can also used
as a disinfectant in cleaning activities, such as cleaning the fish hold, general vessel cleaning, and
laundry.
                       Measuring Total Residual Chlorine Immediately
                                   After Sample is Taken.
    1.6.2.  Nutrients
       Nutrients, including nitrogen, phosphorus, and numerous micronutrients, are constituents of
vessel discharges. Though traditionally associated with discharges from sewage treatment facilities and
runoff from agricultural and urban stormwater sources, small quantities of nutrients from vessels are
discharged from deck runoff, graywater, bilgewater, and fish hold tanks, among other sources. Although
outside the scope of this report, sewage discharge (blackwater) is likely one of the primary sources of
nutrients from vessels.

       When excessive amounts of phosphorus and nitrogen are added to the water, algae and aquatic
plants can be produced in large quantities and cause eutrophication of lakes or ponds. Eutrophication is a
natural process whereby primary producers (algae and aquatic plants) exhibit extreme growth due to
increased nutrient loading. Eutrophication can be greatly accelerated by human activities that increase
the rate at which nutrients enter the water. Increased nutrient discharges from human sources are a major
source of water quality degradation throughout the United States.

       Total nitrogen is a measure of all the various forms of nitrogen (nitrate, nitrite, and ammonia)
that are found  in a water sample. Nitrification is the biological oxidation of nitrogen compounds in both
water and soil: ammonia is oxidized to nitrite (via Nitrosomas bacteria) and further oxidized to nitrate
via Nitrobacter bacteria. Nitrite and ammonia are relatively toxic forms of nitrogen, while nitrate is
relatively nontoxic. Nitrogen in natural waters is usually found in the form of nitrate.

       Phosphorus can be measured in either the particulate phase or the dissolved phase. Particulate
matter includes living and dead plankton, precipitates of phosphorus, phosphorus adsorbed to
particulates, and amorphous phosphorus. The dissolved phase includes inorganic phosphorus and

                                               35

-------
                                                                       Chapter 1 - Introduction to Report
organic phosphorus. Phosphorus in natural waters is usually found in the form of phosphates.
Phosphates can be in inorganic form (including orthophosphates and polyphosphates) or organic form
(organically bound phosphates).

    1.6.3.  Pathogen Indicators

       Pathogens are microbes that cause disease. They include a few types of bacteria, viruses,
protozoa, and other organisms. Bacteria associated with human and  animal waste (e.g., total and fecal
coliforms, E. coli, enterococci) are often monitored in water and wastewater, and the detection of these
organisms can be a reliable indicator that other dangerous pathogens might be present. Pathogens are
often found in discharges from vessels, particularly in vessel sewage and graywater.

    1.6.4.  Metals

       Metals are a diverse group of pollutants, many of which are toxic to aquatic life and humans.
While some metals, including copper, nickel, and zinc, are known to be essential to organism function,
many others, including thallium and arsenic, are nonessential and/or are known to have only adverse
impacts. Even essential metals can do serious damage to organism function in sufficiently elevated
concentrations. Adverse impacts can include impaired organ function, impaired reproduction, birth
defects, and at extreme concentrations, acute mortality. For example, copper can inhibit photosynthesis
in plants and interfere with enzyme function in both plants and animals in concentrations as low as 4
ug/1. Additionally, through a process known as bioaccumulation, metals can accumulate in predator
organisms further up the food chain, including  commercially harvested fish species.

       The toxic potential of a metal depends on its bioavailability in a given aquatic environment. A
metal's bioavailability is determined by the characteristics of the surrounding environment (e.g.,
temperature, pH, salinity, TOC) and the species of the affected organism. The environmental conditions
determine a metal's tendency to either adsorb to suspended organic matter and clay minerals or to
precipitate out of solution and settle to the sediments. Benthic organisms can bioaccumulate metals by
consuming metal-enriched sediments and suspended particles or by uptaking ambient water containing
the dissolved form  of the metal.

       Vessel discharges can contain  a variety of metal constituents, which can come from a variety of
onboard sources. Graywater, bilgewater, and firemain systems have been shown to contain numerous
metals, the exact constituents of which vary depending on onboard activities and the materials used in
the construction of the vessel. Other metals, such as copper, are known to leach from the antifoulant
coatings on vessel hulls and can cause exceedances of water quality standards.

    1.6.5.  Volatile and Semivolatile Organic Compounds

       A variety of organic compounds have been found in vessel discharges, many  of which are known
to have a broad array of adverse impacts on aquatic species and human health. For this study, EPA
measured VOCs  and SVOCs, which can dissolve other substances and evaporate readily at room
temperature and atmospheric pressure. These carbon-containing compounds include a wide range of
                                              36

-------
                                                                      Chapter 1 - Introduction to Report
chemicals, such as aldehydes, ketones, and hydrocarbons, and are present in oily materials such as
gasoline, motor oil, engine coolants, and lubricants used on vessels. VOCs such as benzene, which is
found in fuel, have acute hematological toxicity (ATSDR, 2007) and many SVOCs such as
benzo(a)pyrene are persistent, bioaccumulative, and toxic compounds.

       EPA measured VOCs and SVOCs in vessel discharges from engines, bilges, and firemains for
this study. The most significant rates and levels of detection were phthalates (plasticizers added to
plastics to make them flexible) and components of or products of incomplete combustion of oil and fuel.
For example, VOCs  and SVOCs detected in engine effluent included multiple polycyclic aromatic
hydrocarbons (PAHs), straight-chain hydrocarbons, phenol and methyl phenols, trimethylbenzene,
phthalates, and the volatile constituents of fuel, commonly referred to as "BTEX" (benzene, toluene,
ethylbenzene, xylene). Many of these compounds are known to cause adverse impacts on aquatic  species
and human health.

    1.6.6.  Nonylphenols

       The general term nonylphenols in this report represents two distinct subsets of the broader family
of alkylphenols that  are commonly used in many products such as liquid detergents and soaps. They can
degrade to total nonylphenol, or NP, which is toxic to aquatic life. There are different types of
alkylphenols, such as nonylphenol polyethoxylates (NPEOs) and octylphenol polyethoxylates (OPEOs).
Because NPEOs and OPEOs are in the same family, they have similar chemical properties. These two
distinct subsets  of alkylphenols (nonylphenols and the closely related octylphenols) exist in these  and
other commercial products as mixtures of isomers (polyethoxylates) of different length chains. Different
isomers are distinguished by length of the branched alkyl side chain. The longer chain nonylphenol
polyethoxylates (of which there are 18 isomers distinguished by number,  e.g., NP18EO) and octylphenol
polyethoxylates (of which there are 12 isomers, e.g., OP12EO) eventually will degrade in the
environment to  isomers with shorter chained ethoxylate groups and ultimately, total nonylphenol (NP),
which, as discussed above, is toxic to aquatic life. In general, the hydrophobicity, persistence, and
toxicity of the substance all increase as the ethoxylate chain becomes  shorter. The short-chained isomers
may be quite persistent once they are buried in the sediment, and bottom-feeding fish can be
significantly exposed to these persistent and toxic compounds. Long-  and short-chain NPEOs and
OPEOs are expected to be found in several vessel discharges, including graywater, deck washing
wastewater,  and bilgewater.

1.7.   CHAPTER CONCLUSIONS

       The information summarized and referenced in this chapter provides an introduction to the study
vessel universe. It describes  the universe of study vessels, the types of discharges generally thought to
originate from those vessels, and the types of pollutants or other constituents generally found in those
vessel discharges. It  also references information contained in Appendix B of this report, which provides
more detailed information on the study vessel universe, such as vessel locations and characteristics. EPA
estimates that there are between 118,000 and 140,000 vessels in the United States subject to the NPDES
permitting moratorium established by P.L. 110-299. This chapter concludes that commercial fishing
                                              37

-------
                                                                       Chapter 1 - Introduction to Report
vessels are the most common type of study vessels, although there are significant numbers of other
commercial study vessels.

       The information contained in this chapter helped inform EPA's decisions of which discharges to
sample and the relative importance of each discharge (see Chapters 3, 4, and 5 for additional
discussion). Based on EPA's experience gained during the VGP process, the Agency believes
bilgewater, graywater, deck washdown, fish hold, engine effluent, and antifouling hull coating leachate
are the primary vessel discharges that could impact surface water quality. Pollutants in these discharges
might include metals, organics, nonylphenols, nutrients, oxygen depleting compounds, and pathogens.
The following chapters of this report present the methodology EPA used to characterize discharges from
vessels subject to the NPDES permitting moratorium, the results of that characterization, and the
potential environmental impacts to ambient waters that  could be caused by these discharges.
                                              38

-------
                                                                      Chapter 2 - Study Design
CHAPTER 2
STUDY DESIGN
       This chapter documents the methodology that EPA used to conduct this study of
discharges incidental to normal operation of study vessels. It describes the steps EPA took to
collect information on the nature and potential impacts of vessel discharges.

2.1    DATA SOURCES

       EPA collected data from a variety of sources, including existing data from other EPA
data collection efforts, meetings and telephone contacts with trade association representatives,
vessel visits and sampling, literature reviews, and other governmental data  sources. Each of these
data sources is discussed below.

    2.1.1  Existing EPA Data Sources

       A significant source of existing data regarding vessel discharges is EPA's administrative
record supporting EPA's 2008 Vessel General Permit (VGP). The administrative record is a
collection of all materials EPA considered in developing the VGP, including supporting
documents, references, and comments received on the proposed VGP. As a first step in
conducting this study, EPA reviewed these existing data sources to determine whether and to
what extent the data and information from these sources could be used to satisfy the study
objectives.  This review  also identified data and information gaps for EPA to target for additional
data collection efforts. In general, these existing data sources provided useful information
regarding the types of vessel discharges generated by vessel class, as well as the shipboard
processes that contribute to their generation; however, the existing data sources contained little
or no information regarding the nature, composition, and volume of discharges.

       Other existing data sources evaluated for this study included  supporting documents and
other materials from EPA's Uniform National Discharge Standards (HMDS) (USEPA, 1999) and
cruise ship  discharges (USEPA, 2008c) programs. These sources, which pertain to armed forces
vessels and large cruise ships, respectively,  have limited applicability to commercial fishing
vessels and nonrecreational vessels less than 79  feet in length; however, these data sources did
provide supplemental information regarding shipboard processes that result in wastewater
generation, as well  as information regarding the types and amounts of pollutants that may be
found in selected vessel discharges such as graywater and bilgewater. One  source directly
applicable to this study, however, is the UNDS document, Final Sampling Episode Report for
Small Boat Engine  Wet Exhaust Discharge from Compression Ignition Engines (USEPA,
2008b), which provides pollutant data and other relevant information (e.g.,  vessel power levels)
for wet exhaust discharges from two compression ignition engines. EPA used this report as a
primary source of information and data for this vessel discharge.
                                           39

-------
                                                                       Chapter 2 - Study Design
   2.1.2  Industry Participation

       EPA was contacted by or contacted, met with, or otherwise collaborated with trade
associations and individual companies. In the course of these meetings, EPA gathered the
following types of information regarding vessel discharges:

       •  Vessel classes within and outside the scope of this study.
       •  Typical vessel lengths by vessel class.
       •  Vessel operating seasons and locations.
       •  Shipboard systems and operations that contribute to vessel discharges.
       •  Vessel discharges and locations by vessel class.
       •  Volume, frequency, and nature of discharges.
       •  Vessel tours to inspect and observe vessel systems and operations that contribute to
          vessel discharges.

       Note that none of the trade associations or individual companies contacted were able to
provide pollutant data for vessel discharges.

       The trade associations that contacted EPA or that EPA contacted included:

       •  American Waterways Operators (represents over 250 members that operate carriers,
          tug boats, towboats, and barges).
       •  Passenger Vessel Association (represents approximately 600 members that operate
          vessels such as ferries, dinner cruises, whale watching expeditions, site seeing tours,
          and water taxis).
       •  National Association of Charterboat Operators (represents over 3,300 charterboat
          owner and operators who provide fishing, sailing, diving, eco-tours, and other
          excursion vessels that carry passengers for hire, as well as recreational for-hire
          vessels).
       •  Conference of Professional Operators for Response Towing (C-PORT) (represents
          over 170 members of the commercial marine assistance industry, providing services
          such as jump starts, fuel delivery,  and towing to boaters).
       •  Pacific Seafood Processors Association (represents 10 seafood processing companies
          in Alaska, Washington, and Oregon).
       •  At-Sea Processors Association (represents five companies that own and operate 19
          U.S.-flag catcher/processor vessels in the Alaskan pollock and West Coast Pacific
          whiting fisheries).
       •  Alaskan Longline Fishermen's Association (represents about 60 members of longline
          fishing vessel companies and salmon fishing vessels that operate in southeast Alaska).
       •  United Fishermen of Alaska (represents about 37 commercial fishing organizations
          and associations concentrated in Alaska representing thousands of fishing companies
          operating as harvesters throughout Alaska waters and the adjoining Exclusive
          Economic Zone).
       •  Southeast Alaska Fishermen's Alliance (represents commercial fishermen and the
          commercial fishing industry in southeast Alaska).
                                           40

-------
                                                                      Chapter 2 - Study Design
       •  Northeast Seafood Coalition (represents commercial groundfish fishermen and shore-
          side businesses from mid-coast Maine to Long Island, New York).
       •  Southern Shrimp Alliance (alliance of shrimp fishermen and processors in Alabama,
          Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, and Texas).
       •  Petersburg Vessel Owners Association (represents fishermen in Petersburg, Alaska).
       •  Alaska Trailers Association (represents southeastern Alaska trailers).
       •  Cordova District Fisherman United (represents Cordova, Alaska, area fishermen).

       Individual companies that provided additional information (generally after being
contacted by their respective trade groups) included:

       •  Potomac Marine, Woodbridge, Virginia.
       •  Vane Brothers Company, Mid-Atlantic.
       •  Potomac Riverboat Company, Alexandria, Virginia.
       •  Northeast Seafood Processors, Gloucester, Massachusetts.
       •  Vulcan Materials Company, Havre de Grace, Maryland.
       •  Sea Tow, Pensacola, Florida.
       •  EPA Gulf Ecology Division Laboratory, Gulf Breeze, Florida.
       •  Sea Tow, Slidell, Louisiana.
       •  AEP River Operations, Convent, Louisiana.
       •  Shrimp Charters, Pass Christian, Louisiana.
       •  Baltimore Water Taxi, Baltimore, Maryland.
       •  Sitka Sound Seafoods, Sitka, Alaska.
       •  Seafood Producers Co-op, Sitka, Alaska.
       •  Silver Bay Seafoods,  Sitka, Alaska.
       •  Argosy Cruises, Seattle, Washington.
       •  Tidewater Marine, LLC, Gulf Coast.
       •  E.N. Bisso & Son, Lower Mississippi River.
       •  Foss Maritime Company, California, Washington, Oregon, the Columbia River, and
          the Snake River.
       •  Taku Smokeries, Juneau, Alaska.
       •  Upper River Services, St. Paul, Minnesota.
       •  JB Marine Service, St. Louis, Missouri.
       •  Osage Marine Services, St. Louis, Missouri.
       •  AEP River Operations, New Orleans, Louisiana.
       •  Smith Shipyard, Baltimore, Maryland.
       •  Norfolk Tug Company, Norfolk, Virginia.
       •  Dann Marine, Baltimore, Maryland.
       •  Cape Fear Riverboats, Wilmington, North Carolina.
                                           41

-------
                                                                        Chapter 2 - Study Design
   2.1.3  Vessel Sampling

       EPA identified a critical need for pollutant data for vessel discharges following its review
of existing data sources. To satisfy this requirement, EPA designed and implemented a vessel
discharge sampling program, which is described in detail in Section 2.2 of this document.
Through this sampling program, EPA collected wastewater pollutant characterization data for
nine vessel discharges sampled from a total of 61 vessels (one to four discharges sampled per
vessel). These samples were collected in 15 different towns/cities in nine separate states,
representing several of the major regions of the United States. Another critical component of
EPA's sampling program was the collection of information regarding the shipboard processes,
equipment, materials, and operations that contribute to the discharges, as well as the discharge
rates, duration, frequency, and location.

   2.1.4  Literature Review

       EPA was not able to sample and characterize all study vessel classes and discharges
(discussed further in Section 2.3). To fill this data gap, EPA searched the literature (i.e.,
scientific and engineering journals or other academic publications) for relevant information. In
general, these  searches provided only general information regarding vessel classes and
discharges and little or no specific information, such as discharge composition and volumes.
EPA did, however, identify many relevant literature sources regarding vessel antifouling
leachate. EPA used these literature sources as the primary sources of information and data for
this vessel discharge.

   2.1.5  Other Governmental Data Sources

       EPA's primary data source for vessel information regarding population and other vessel
characteristics is the U.S. Coast Guard's Marine Information  for Safety and Law Enforcement
(MISLE) database. The MISLE provides data for nearly 1 million vessels that operate in U.S.
waters and is used to support the investigation and inspection activities of the U.S. Coast Guard
throughout the United States and its territories. Of the 1  million vessels identified in the
database, approximately 139,814 vessels comprise the study vessel population (see Chapter 1 for
additional discussion). Relevant vessel characteristics tracked in this database are vessel type,
length, geographical area of operation, age, hull material type, propulsion method and type, and
horsepower ahead.

       EPA used a screening-level analysis of a hypothetical estuarine harbor to evaluate the
potential environmental impacts from multiple vessels discharging to large U.S. water bodies,
specifically estuaries and brackish harbors (see Section 4.2). EPA used the characteristics of
harbor salinity, volume, and freshwater inflow from a variety of U.S. estuaries that receive vessel
discharges to develop the characteristics for the hypothetical estuary. EPA compiled these
characteristics from the following online sources:
                                            42

-------
                                                                      Chapter 2 - Study Design
       •  National Oceanic and Atmospheric Administration BookletChart™ List
       •  National Oceanographic Data Center World Ocean Database 2005 (WOD05)
       •  Southeast Environmental Research Center, Biscayne Bay Water Quality Monitoring
          Network, Miami, Florida.
       •  Cronick, T., and A. McGuire. Temperature and Salinity of the Yaquina Bay Estuary
          and the Potential Range ofCarcinus maenas, Corvallis, Oregon.
       •  Massachusetts Department of Environmental Protection, Total Maximum Daily
          Loads of Bacteria for Little Harbor, Worchester, Massachusetts.
       •  U.S. Geological Survey National Hydrography Dataset Plus.
       •  U.S. Geological Survey National Water Information System Surface Water Annual
          Statistics.

 2.2   EPA VESSEL DISCHARGE SAMPLING PROGRAM

       EPA conducted a sampling program of discharges from commercial fishing vessels and
other nonrecreational vessels less than 79 feet in length. EPA's sampling program was designed
to provide information to achieve the first two objectives of the study mandated by P.L. 110-299:

       •  A characterization of the nature, type, and composition of discharges for
          representative single vessels, and for each class of vessel.
       •  A determination of the volumes of those discharges, including the average volumes
          for representative single vessels, and for each class of vessel.

       Accordingly, EPA's sampling program included the sampling of large numbers and
varieties of vessel classes, vessels,  and discharges, and the analysis of target analytes as
discussed in the following subsections. In addition, EPA supplemented sample collection and
analysis with the collection of information regarding the shipboard processes, equipment,
materials, and operations that contribute to the discharges, as well  as the discharge rates,
duration, frequency, and location.

       Though the Agency was still in the final stages of drafting the 2008 VGP, EPA began
designing and planning the sampling program soon after enactment of P.L. 110-299. These
activities included developing the size and scope of the program considering overall  program
schedule and resources; identifying priority locations, vessel classes, discharges, and analyte
classes for sampling; developing a  detailed Generic Sampling Analysis Plan and Quality
Assurance Project Plan; procuring EPA regional laboratory and contract laboratory and sampling
support; and soliciting industry input and volunteers for participation in the program. Sample
collection was  conducted from March through July 2009. The remainder of this section provides
a further description of the sampling program, including the vessels sampled and their locations,
sampled discharges, target analytes, sampling methods, and quality assurance/quality control.
                                          43

-------
                                                                       Chapter 2 - Study Design
   2.2.1  Vessels Sampled and Locations

       EPA sampled discharges from a total of 61 vessels in nine vessel classes (see Table 2.1).
To select vessel classes for evaluation, EPA first developed a list of commercial vessel classes
based on published information and the EPA team's existing understanding of vessels. Next,
EPA narrowed the sampling scope to focus largely on those vessel classes believed to consist
primarily of vessels less than 79 feet in length.  Some examples of vessel classes on which EPA
did not focus, due to their size, include cable laying ships, cruise ships, large ferries, oil and
petroleum tankers, and freight ships/barges (most vessels in these classes are typically greater
than 80 feet in length). Next, EPA eliminated vessel classes outside the scope of study vessels,
including stationary seafood processing vessels and vessels that can be secured to the ocean floor
for mineral or oil exploration, because the industrial discharges from these vessels were outside
the scope of the previous 40 CFRPart 122.3 (a) exclusion (USEPA, 2008d). After eliminating
these vessels, the following common vessel classes were prioritized for evaluation:

       •  Commercial fishing vessels and tenders
       •  Tugs/towing vessels
       •  Water taxis/small ferries
       •  Tour boats
          Purse Seiner in Alaska (left) and a Shrimp Trawler in Louisiana (right).
                                           44

-------
                                                               Chapter 2 - Study Design
A Tugboat in Maryland (left) and a Tow/Salvage Vessel in Virginia (right).
    A Water Taxi in Virginia (left) and a Tour Boat in Virginia (right).
                                   45

-------
                                                                                                                 Chapter 2 - Study Design
Table 2.1. Number of Vessels Sampled by Vessel Class and Discharge
Vessel Class
Fishing:
Gillnetter
Lobster :
Longliner
Purse Seiner
Shrimp Trawler
Tender
Trawler
Trailer
Tugboat
Water Taxi
Tour Boat
Tow/Salvage
Research
Fire Boat
Supply Boat
Recreational
Total
Number of
Vessels
Sampled
5
1
3
5
6
3
4
6
9
4
3
6
2
1
1
2
61
Number of Vessels Sampled by Discharge
Bilge
Water
1
1

2
1
3




8
Stern Tube
Packing
Gland

9







9
Deck
Washdown
1
6
2
2
9
1
2
6

1
1
1
32
Fish Hold
o
J
1
3
5
2
3
3
6








26
Cleaning of
Fish Hold
1
1
2
4
1








9
Graywater
1
5
1





1
8
Propulsion
Engine Effluent
1
1

4
3
5
2
1

2
19
Generator
Engine Effluent
1

1
2


1


5
Firemain

2

3


1


6
(1) Sampled the lobster hold tank on a trawler.
                                                                    46

-------
                                                                       Chapter 2 - Study Design
       Additionally, EPA sampled recreational vessels used for nonrecreational purposes as part
of this study. This sampling was done for two purposes: 1) to provide a semi quantitative
comparison of the discharges from these vessels and the other study vessels, and 2) to collect
additional information for EPA's related work on the Clean Boating Act (P.L. 110-288). During
the execution of the sampling program, EPA also conducted opportunistic sampling of additional
non-priority vessel classes (e.g., fire boats, recreational boats,  a supply boat) when EPA had
access to these vessels and the resources to sample them. See Section 1.3 of this document for a
short description of different vessel classes or types.

       As discussed in Section 2.1.2, EPA was contacted by or otherwise developed contacts
with trade associations and individual companies. Many of these entities relayed the purpose of
the study to their constituents or peers, some of whom contacted EPA. Consequently, EPA
obtained a pool of individual companies who were willing  to volunteer their vessels for the
sampling program. EPA then selected specific companies and vessels within the volunteer pool
for sampling to obtain a variety of vessel classes, vessel platforms, companies, and geographic
distribution. In general, EPA selected the entire volunteer pool within the following geographic
areas to maximize the number and variety of sampled vessels based on available resources: New
England (Gloucester/New Bedford, Massachusetts); Mid-Atlantic (Woodbridge, Virginia;
Alexandria Virginia; Baltimore, Maryland; Havre de  Grace, Maryland; and Philadelphia,
Pennsylvania); Gulf Coast (Gulf Breeze, Florida; Pensacola, Florida; Bayou laBatre, Alabama;
Pass Christian, Mississippi;  Slidell, Louisiana; La Fitte, Louisiana; and Convent, Louisiana); and
Sitka, Alaska.

       EPA's vessel selection approach for commercial fishing vessels differed from that of
other vessel classes due to the nature of this industry. During the fishing season, fishing trips
typically last for multiple days with no preset schedule. The captain of each vessel determines
the end of each fishing trip,  returns to the seafood processor or tender to offload the catch, and
then typically immediately returns to the fishing grounds. Therefore, EPA identified seafood
processors, rather than specific fishing companies and vessels, as the means to obtain a pool of
active fishing vessels for sampling. Sampling was conducted at the docks of the seafood
processors during the offloading process, and EPA sampled all vessels that arrived while the
EPA sampling crew was at the docks (with the permission  of the captains). In this way,  sampling
of individual commercial fishing vessels was random. However, EPA did contact the seafood
processing facilities prior to sampling to provide sampling  details (e.g., nature of the study,
discharges of interest, sampling dates). It was the facilities' discretion whether or not to share
this information with the vessel fleets that use their offloading facilities.

       Due to the assistance of trade groups and others, vessel owner/operators were generally
very cooperative with EPA sampling teams. For example, the EPA vessel team found that most
of the fishermen with whom they spoke in Sitka, Alaska, were aware of the study and that EPA
would be sampling in the area during the summer. Other vessel owner/operators took EPA
                                           47

-------
                                                                       Chapter 2 - Study Design
underway to sample engine effluent, waited to wash their dishes or take showers until EPA was
able to collect the graywater discharge, and assisted EPA scientists in answering questions about
their vessel operations.

       During the public comment period for this report, EPA received comments noting that the
types of vessels sampled for this study were not necessarily representative of the industry as a
whole, specifically for commercial fishing vessels. Based on public comments, EPA evaluated
the representativeness of the sampled  commercial fishing vessels in terms of size, class, and
geographic distribution. The commercial fishing industry is highly diverse (spanning a wide
array of ocean and nearshore conditions, differing by both region and fishery). With respect to
vessel size, EPA notes that the sampled  commercial fishing vessel population does not represent
discharges from vessels less than 26 feet in length, which comprise an estimated 10% of the
overall commercial fishing vessel population. However, though EPA did  not physically sample
these vessels, we visually observed that  these small vessels typically  store their catch in coolers
(which do not have a discharge) rather than in refrigerated seawater or ice holding tanks, which
is a function of their relatively short fishing voyages. For larger fishing vessels, EPA believes the
sampled vessel population is reasonably representative of the overall vessel population, albeit
somewhat more heavily weighted toward the largest of commercial fishing vessels (i.e., 50 feet
or more).

       With respect to vessel  class, EPA's sampled commercial fishing vessel population
includes vessels in all of the fishing vessel classes. Furthermore, the percentage of sampled
vessels by class is similar to or greater than the percentage of the overall vessel population by
class for all vessel classes except for Pot/Trap vessels. Pot/Trap vessels include many of the
smallest commercial vessels that are not represented by EPA's sampled vessel population. EPA
also notes that it sampled a much higher percentage of Seiners than the overall vessel population
(15% of vessels sampled versus only 2% of the overall vessel population). Sampling in Sitka,
AK occurred at the start of the salmon fishing season, resulting in a preponderance of Seiners at
the docks of seafood processers.

       Finally, with respect to geographic distribution, EPA sampled commercial fishing vessels
in the following regions: Alaska (21 vessels); Gulf Coast (6 vessels); and New England (6
vessels). According to the National Marine Fisheries Service, in 2008 these three areas combined
represented approximately 77% of U.S.  domestic commercial landings in 2008. Among the
remaining geographic regions that EPA  was unable to sample, the Pacific Coast (excluding
Alaska) has the greatest landings in 2008 at 13% of U.S. domestic commercial landings;
commercials fishing vessels in this region are expected to be similar to those in Alaska. Finally,
many of other remaining geographic regions that EPA was unable to sample, such as the
Chesapeake Bay, Middle Atlantic, and the Great Lakes, are likely dominated by small fishing
vessels that most likely have low volumes of or no fish hold effluent  discharges.

       Hence, with the exception of the smallest commercial fishing vessels, EPA believes that
the sampled vessel population is a representative cross-section of vessels  and is adequate to
evaluate the vessel population for the  purpose of this study.
                                           48

-------
                                                                       Chapter 2 - Study Design
   2.2.2  Sampled Discharges

       EPA sampled a total of nine discharge types during the sampling program (see Table
2.1). To identify priority discharges for sampling, EPA first developed a list of vessel discharges
based on information collected from discussions with industry representatives (see Section
2.1.2), as well as EPA's understanding of vessel discharges. Next, EPA prioritized the list to
focus on the following discharges that are commonly generated by the vessel classes of interest
and that are amenable to sampling (see Chapter 1 for descriptions and locations of these
discharges):

       •  Bilgewater
       •  Stern tube packing gland effluent
       •  Deck runoff and/or washdown
       •  Fish hold effluent (including both refrigerated seawater effluent and ice slurry)
       •  Effluent from the cleaning offish  holds
       •  Graywater
       •  Propulsion engine effluent
       •  Generator engine effluent
       •  Firemain discharges

       Vessels routinely use ambient waters to conduct normal operational and cleaning
activities that lead to the generation of above  discharges. EPA collected samples of ambient
water where the vessels were operating. EPA also collected potable water used onboard the
vessels (service water) to characterize any background concentrations of pollutants that might be
detected in discharges from vessel operations that use service water.
                    Various Discharges Through Hull Discharge Ports.
                                           49

-------
                                                                      Chapter 2 - Study Design
       EPA did not select non-oily machinery wastewater as a priority discharge for sampling
because it was not expected to be discharged separately from bilgewater. The vessels that EPA
sampled during this program use the bilge system to manage non-oily machinery wastewater (if
there is any), such as fresh- and saltwater pump drains, chilled water condensate drains, and
potable water tank overflows, rather than installing dedicated drip pans, funnels, and deck drains
to provide for segregated discharge. Note, however, that EPA has not performed a
comprehensive investigation of whether or not certain non-oily machinery wastewaters may have
segregated discharges on other study vessels.

       EPA did not select the discharges listed below as priority discharges for sampling
because they were not reasonable or practical to sample within the overall program schedule and
available resources.

       •  Anti-fouling hull coatings.
       •  Cathodic protection.
       •  Controllable pitch propeller and thruster hydraulic fluid and other oil-to-sea
          interfaces.
       •  Underwater ship husbandry.

       EPA did not select the discharges listed below as priority discharges for sampling
because they were not expected to be commonly generated on commercial fishing vessels or
nonrecreational vessels less than 79 feet in length.

       •  Aqueous film-forming foam
       •  Boiler/economizer blowdown
       •  Distillation and reverse osmosis brine
       •  Elevator pit effluent
       •  Exhaust gas scrubber wash water
       •  Freshwater layup
       •  Gas turbine wash water
       •  Motor gasoline and compensating discharge
       •  Sonar dome discharge
       •  Welldeck discharges
       •  Graywater mixed with sewage

None of these discharges were sampled during the sampling program because none of the 61
vessels that EPA selected for sampling generated these discharges. Note, however, that EPA has
not performed a comprehensive investigation of whether or not these discharges are applicable to
other study vessels.
                                           50

-------
                                                                        Chapter 2 - Study Design
    2.2.3  Target Analvtes

       EPA's vessel discharge sampling and analysis program included 301 target analytes in
the following eight analyte groups:l

       •  Microbiologicals (pathogen indicators)
       •  Volatile and semivolatile organic compounds
       •  Total and dissolved metals
       •  Oil and grease
       •  Sulfide
       •  Long and short chain nonylphenol  and octylphenol ethoxylates (i.e., alkylphenol
          ethoxylates) and total nonylphenol (NP)
       •  Nutrients
       •  Other physical/chemical parameters

       Appendix D lists the target analytes included in each group, along with the analytical
methods. EPA selected this comprehensive list of analytes to perform a screening-level analysis
of the presence or absence of almost all priority pollutants (listed at 40 CFR Part 423, Appendix
A), conventional pollutants defined at Section 304(a)(4) of the Clean Water Act, and toxic
pollutants from EPA's 2006 National Recommended Water Quality Criteria for freshwater and
saltwater aquatic life and human health, as well as many other nonconventional pollutants.
Nearly half of these analytes (147) were never detected in any vessel discharge sample (see
Chapter 3).

       EPA did not analyze all vessel discharges for all selected analyte groups (see Table 2.2).
Analyte groups were selected for analysis based on their possible presence in discharges, as
determined from existing data sources and EPA's understanding of what constituents are
possibly present in the different vessel discharges. For example, long-chain nonylphenol and
octylphenol ethoxylates were only analyzed for in those discharges with the potential to contain
detergents (i.e., bilgewater, packing gland, deck washdown, fish hold cleaning effluent, and
graywater). Furthermore, short-chain nonylphenol and octylphenol ethoxylates and NP were only
analyzed for in those discharges for which the long-chain  structural isomers of these two subsets
of alkylphenol ethoxylates were analyzed,  and that also had a holding time onboard the vessel
that would allow for the possible degradation of the long-chain isomers to the short-chain
isomers and NP (e.g., bilgewater held in the bilge, graywater stored in a holding tank).
1 Due to overall program resource constraints and other factors, not all analyte groups of possible concern were
selected for this study (see Section 2.3.3).

                                            51

-------
                                                                                                                            Chapter 2 - Study Design
Table 2.2. Analyte Groups by Discharge








Vessel Class and Priority Discharge
Bilgewater
Stern tube packing gland effluent
Deck runoff and/or washdown
Fish hold effluent (including both refrigerated
seawater effluent and ice slurry)
Effluent from the cleaning of fish holds
Graywater
Propulsion engine effluent
Generator engine effluent
Firemain systems


<*>
C5
_0
0
.a

o
s
A/(C)

A/(d)
A/

A/
A/




.1
S U

§ o 1

•p ra a
3;§ s
> ® u
A/
A/





A/
A/
A/

•a
=
S5
c5
' — ' 9£
•• ™«
-*^ %
a> %
S3
A/
A/
A/
A/

A/
A/
A/
A/
A/



Is
O
•a
=

O
A/
A/
A/
A/

A/
A/
A/
A/






OJ
•a

!/5
A/
A/

A/

A/
A/
A/
A/

Q-
Z

=
111
U J3 "5,
-i S H
•• ^s O
i/5 •

— • 5«
-2 S
H^
A/
A/
A/
A/

A/
A/
A/
A/


CS
O
S £
Sp
"B a
*• .y s

O Q- a.
A/
A/
A/
A/

A/
A/
A/
A/
A/
(a)  Biochemical oxygen demand (BOD5), chemical oxygen demand (COD), and total organic carbon (TOC).
(b)  Other physical/chemical parameters include: conductivity; dissolved oxygen; pH; salinity; temperature; total residual chlorine; turbidity; and observations of
odor, color, and floating and settleable material.
(c)  Microbiologicals analyzed for only those vessels with potential for a source of these pollutants to enter the bilge (e.g., graywater piping, fish hold effluent).
(d)  Microbiologicals analyzed for only commercial fishing vessels.
(e)  Short-chain alkylphenol ethoxylates (i.e., nonylphenol and octylphenol ethoxylates) and NP analyzed only for graywater that has been stored in a holding
tank prior to discharge.
                                                                       52

-------
                                                                      Chapter 2 - Study Design
   2.2.4  Sampling Methods

       EPA used a variety of sample collection methods depending on the nature of the
discharge. This section describes the most commonly used sampling methods.

Discharge from a discharge port well above the water line.

       Samples of these types of discharges were typically collected directly into a 5-gallon
utility bucket lined with a new pail liner. For some samples, the bucket could be lowered by
hand, while for other samples, the bucket was lowered by rope. The sample in the pail liner was
then poured into the individual sample bottles. Whenever possible, samples for analysis of oil
and grease were collected directly into the sample bottles (either held by hand or attached to a
pole) to avoid the possible loss of oils to the sides of the sample transfer jar and pail liner.
However, when oil and grease sample bottles were filled directly by attaching to a pole, it was
typically necessary to "top off the sample bottles with sample from the pail liner to ensure
adequate sample volume for analysis.
                     Sample Collection Well Above the Water Line.

Discharge from a discharge port at or below the water line.

       Typically, samples of these types of discharges were impossible or too unsafe to collect.
In a few cases, however, EPA was able to collect samples upstream of the discharge port via a
sampling port. For example, on one vessel, engine effluent could be accessed from a petcock
valve on the muffler. Samples of these types of discharges were preferentially collected directly
into sample bottles. In some cases, the clearance between the sampling port and the vessel hull
                                           53

-------
                                                                      Chapter 2 - Study Design
was insufficient to accommodate the sample bottles and instead, required sample collection
directly into a new pail liner; the sample in the pail liner was then poured into the individual
sample bottles.
                               Close-up of Petcock Valve.

Deck washdown and runoff.

       Deck washdown and runoff wastewater is discharged through scuppers located along the
perimeter of the deck. To collect samples of this discharge, EPA generally directed the discharge
to one or more (up to four) select scuppers using a variety of methods. On some vessels, deck
washdown water naturally flowed by gravity to one or more scuppers at the lowest end of the
deck. On other vessels, EPA used either the spray from the hose used to wash the deck or the
broom used to wash the deck to direct the deck washwater to one or more selected scuppers.
Finally, on some vessels, EPA arranged the deck washing hose on the vessel deck such that it
directed and pooled deck washdown water to one or more selected scuppers. To collect the
discharge from a selected scupper, EPA held a new pail liner against the hull of the vessel to
capture the deck washdown water as it drained through the scupper. If deck washdown water was
discharged through multiple scuppers, EPA filled the pail liners proportionally from each
scupper (e.g., half from each of two scuppers, one-third from each of three scuppers). The
sample in the pail liner was then poured into the individual sample bottles.
                                          54

-------
                                                                      Chapter 2 - Study Design
       Deck Cleaning and Collecting Deck
               Washdown Sample.
   Collecting Deck Washdown
Sample with Close-up of Scupper
  (Indicated with Red Arrow).
       For select fishing vessels, EPA attempted to collect samples of runoff during actual
fishing operations. EPA arranged to travel with an overnight shrimping vessel on the Gulf Coast;
however, due to a temporary seasonal shrimp fishery closing, EPA obtained a research permit to
collect runoff from "demonstration" operations. Because these were demonstration operations
and the shrimp fishermen would be unable to keep the catch, the vessel operator used a smaller
net for shorter durations and did not handle the catch as he normally would. As a result, these
samples only partially resemble normal operations. While in Alaska, the U.S. Coast Guard
assisted EPA in attempting to sample deck washdown from fishing vessels immediately after
they pulled in their catch. EPA and the Coast Guard attempted to sample three to five vessels
during this operation. Due to weather conditions, however, they were only able to sample one
vessel successfully.

Fish hold tanks.

       Three types offish hold tanks were sampled during the program: tanks containing
refrigerated seawater, tanks containing ice slurry, and tanks containing chipped ice.  Refrigerated
seawater tanks were common to tenders, purse seiners, and trawlers, while slurry and chipped ice
tanks were common to trailers, long-liners, gillnetters, and some trawlers. For vessels with
refrigerated seawater tanks, fish are typically extracted using a vacuum system that removes both
the fish and refrigerated seawater simultaneously. Both fish and refrigerated seawater are
transferred to the seafood processing plant. The refrigerated seawater is generally recycled back
to the fish hold tank to provide the liquid needed to operate the vacuum system. Any excess
refrigerated seawater that is not required to assist in fish extraction is pumped overboard pier
side. EPA collected  samples of the refrigerated seawater directly into a 5-gallon utility bucket
lined with a new pail liner as the water was pumped overboard. The  sample in the pail liner was
then poured into the individual sample bottles. Because removal offish and refrigerated seawater
                                           55

-------
                                                                      Chapter 2 - Study Design
can take several hours depending on the vessel size, EPA collected the sample approximately
mid-way through the fish removal process. For vessels such as trailers and long-liners, which use
chipped or slurry ice, EPA collected a sample of the ice or slurry once the fish had been removed
from the fish hold tank. Ice/slurry was collected into a new pail liner and allowed to melt. Once
melted, the sample was poured from the pail liner into the individual sample bottles.
                     Collecting Fish Hold Samnle with a Lined Bucket.
Fish hold cleaning.
       After the fish hold has been evacuated, the vessel crew cleans the fish hold as described
in Section 1.3. For vessels with refrigerated seawater tanks or chipped ice tanks, the fish hold
cleaning wastewater is pumped overboard. EPA collected samples of the fish hold cleaning
wastewater directly into a 5-gallon utility bucket lined with a new pail liner as the cleaning water
was pumped overboard.

Firemain.

       EPA used valving on the firemain system to throttle the flow rate to allow firemain
samples to be collected from the fire hose directly into sample containers. None of the vessels
visited by EPA for this study tests its firemain system more frequently than once every two
weeks, and none operates its system for secondary purposes such as deck washing.  Of the six
firemain systems sampled, five were wet systems (the firemain piping is normally filled with
water) and one was a dry system (the firemain piping is normally filled with air). The resulting
                                          56

-------
                                                                      Chapter 2 - Study Design
sampling data are applicable to firemain systems that are operated infrequently with intake
provided by surrounding water and without additions to the discharge (e.g., no addition of foam-
forming agents).

Composite samples of multiple wastewaters.

       To better characterize some discharges, EPA decided to combine multiple samples of
wastewaters into a single sample for analysis. The most common example is a vessel that
operates its engines at multiple power levels—idle at the pier, half-speed when motoring through
the no wake zone, and three-quarter speed when performing harbor tours. Another example is a
vessel that generates two types of graywater—wastewater from a galley sink and wastewater
from a shower. In these cases, EPA filled a new pail liner proportionally based on the number of
wastewater sources (e.g., one-third from each of three power levels, half from each of sink and
shower water) using one or more of the sample collection methods described above. The sample
in the pail liner was then poured into the individual  sample bottles. Whenever possible, EPA
collected and analyzed separate samples for each discharge for oil and grease and for volatile
organics, rather than using the composite sample; this minimized the possible loss of these target
analytes from volatilization during sample transfer among multiple sampling equipment or due to
adherence of oils to the sides of multiple sampling equipment.
      Collecting Engine Effluent with a Transfer Jar.  Compositing the Sample in a
                                                     Lined Bucket.
                                          57

-------
                                                                       Chapter 2 - Study Design
   2.2.5  QA/OC

       Quality assurance/quality control (QA/QC) procedures applicable to EPA's vessel
sampling program are outlined in the Quality Assurance Project Plan for Discharges from
Commercial Fishing Vessels and Other Non-Recreational Vessels Less Than 79 Feet (Q APP),
which is included in the docket of the Federal Register notice announcing this study. This section
describes the QC practices used to assess the precision and accuracy of the analytical data.

       2.2.5.1    Analytical Quality Control

       Analytical chemistry support for this program was provided by EPA's own laboratories
in Regions 2, 3, and 5, as well as several subcontract laboratories. The EPA Regions were
responsible for the quality  of the work generated by their laboratories and for verifying that
laboratory performance was acceptable by conducting QC checks of the analytical data as
specified by the QAPP. Subcontract laboratories functioned within the quality system of EPA's
sampling contractor, who verified the acceptability of subcontract laboratory performance by
conducting QC checks of the analytical data as specified by the QAPP. Based on the data quality
review and evaluation of the analytical data under this sampling program, all analytical data were
deemed within or sufficiently close to the target analytical QC limits established for the study to
assure the data could be used for the specified intentions. QC failures were generally attributed to
matrix interference; these results are not uncommon for  complex wastewater samples.
Furthermore, the sample collection, handling, preparation, and analysis process utilized in this
sampling program was deemed acceptable for the matrices and conditions sampled.

       2.2.5.2    Field Quality Control

       Field QA/QC measures and results for the bottle  blanks, equipment blanks, trip blanks,
field blanks, and field duplicates are discussed in this subsection.

Bottle blank.

       A representative bottle and cap from the first lot  of bottles purchased for collection of
samples for analysis of pathogen indicators were analyzed for wide-spectrum contamination
prior to their use in the sampling program. Bottles were filled with sterile deionized water, and
100-milliliter (mL) aliquots were filtered by membrane  filtration. The filters were placed on
water agar, nutrient agar, modified mTEC agar (for E. coli cultures), and mEL agar (for
enterococci cultures). No pathogen indicators or other organisms (water or nutrient agar) were
detected in the bottle blank, indicating that the bottles were sterile.
                                           58

-------
                                                                     Chapter 2 - Study Design
Equipment blanks.

       Two equipment blanks were prepared and analyzed for volatile and semivolatile organic
compounds (SVOC), total and dissolved metals, nutrients, soaps and detergents, and other
physical/chemical parameters to assess the potential introduction of contaminants by sample
collection equipment. The sample collection equipment used to collect the equipment blanks was
the same as that used at the sampling points: 1) a new, factory-cleaned, Teflon® PFA pail liner
from the first lot of bags purchased from the vendor, and 2) a 3-foot segment of silicone tubing
connected to a 25-foot segment of Teflon tubing used in the peristaltic pump (only used on three
samples throughout the entire project). The pail liner equipment blank was prepared by rinsing
the bag with high performance liquid chromatography (HPLC) water and then pouring it into
sample bottles. The pump tubing equipment blank was prepared by pumping HPLC water
through this equipment and collecting directly into sample bottles. Of the 459 equipment blank
sample results, 29 (6.3 percent) were above the method reporting limit (RL). Of the cases where
the equipment blank exceeded the RL, 15 were for SVOC analytes and seven were for VOC
analytes. In all 22 of these cases, however, the analytes were tentatively identified compounds
(TICs), which are appropriately labeled in the analytical database as such. The remaining cases
where the equipment blank exceeded the RL were as follows: biochemical oxygen demand
(BOD) (two instances), chemical oxygen demand (COD)  (two instances), total Kjeldahl nitrogen
(TKN) (one instance), nitrate/nitrite nitrogen (one instance), and zinc (one instance). In each
instance, the vast majority (greater than 90 percent) of the associated discharge sample amounts
were significantly higher than the equipment blank levels.

Trip blanks.

       Trip blanks were prepared and analyzed for volatile organics to evaluate possible
contamination during shipment and handling of samples. These samples consisted of HPLC
water poured into the sample bottles and transported unopened to the field and finally to the
laboratory. One trip blank was prepared for each location-specific sampling event (e.g., Gulf
Coast, New England). Evaluation of the trip blanks indicated that of the 612 VOC results for
these samples, only two analytes were detected (tetramethylsilanol and tetrahydrofuran), and
these were at levels below the RL. Neither of these analytes was detected in any vessel discharge
samples, indicating that there was no sample contamination  during transport, field handling,  and
storage.

Field blanks.

       Field blanks were prepared and analyzed for all target analytes to monitor for the
contamination of samples during sample collection and handling.  These samples were prepared
aboard selected vessels at the location of greatest potential for contamination (e.g., the vessel
bilge space). The samples were prepared by pouring HPLC water into the sample bottles. One
field blank was prepared for each location-specific sampling event (e.g., Gulf Coast, New
                                          59

-------
                                                                      Chapter 2 - Study Design
England). Only six target analytes (conductivity, dissolved organic carbon, nitrate/nitrite, TKN,
turbidity, and total zinc) were detected in any of the 670 field blank results (0.3 percent) at levels
above the RL. In each instance, the associated discharge sample amounts were significantly
higher than the field blank levels.

Field duplicates.

       Field duplicate samples were collected and analyzed for all target analytes to assess the
precision of the entire sample collection, handling, preparation, and analysis process. Field
duplicate samples were collected simultaneously from the same location as the original samples
(i.e., poured from the pail liner as a split sample or sampled sequentially when collecting samples
directly into sample bottles from discharge ports). The relative percent difference (RPD) between
the two duplicate sample results was calculated and compared to the data quality objective. The
occurrence of field duplicate samples (number of samples exceeding out of total number of
duplicate samples) where one or more analytes within an analyte type (VOCs, SVOCs, dissolved
metals) exceeded the target RPD was 89 of 356 pairs  of field duplicate samples, or 25 percent.
The higher RPDs were calculated in samples where the concentrations of the analytes were
detected at levels at or near the detection level for the respective methods, mainly for VOCs,
Silica Gel Treated N-Hexane (SGT-HEM), and residual chlorine. For these methods, the
analytical variability increases as analyte concentrations approach their detection limits. These
results are not uncommon in complex wastewater samples.

       2.2.5.3    Database Development

       An Access database was created in which to collect and organize all analytical results.
This database contained data and associated qualifier  information. Although a number of EPA
and contractor staff were involved in reviewing the results, only one person had the authority to
make any changes to the database during its development. This one-person control system
eliminated the possibility of someone accidentally creating more than one current version of the
database  and minimized the risk of errors. Each time the database was updated, the current date
and time  stamp were used to name the new version, which was uploaded to a secure FTP server.

       After each sampling event, the chains of custody (COC) and field data sheets were used
to manually enter information into the "COC Information" table. This table contained identifiers
given to samples in the field (FieldlDs) associated with vessel name, location, and discharge
information, as well as the sample date and time. A second person performed a  100-percent
check of the data entered to ensure there were no transcription errors or mistakes made during
data entry.

       Four analytical chemistry and subcontract laboratories —EPA Region 2 (Edison, New
Jersey), EPA Region 3 (EPA Environmental Science Center, Fort Meade, Maryland),  TriMatrix
(Grand Rapids, Michigan), and Admiralty (Juneau, Alaska)—provided EPA's contractor with
                                           60

-------
                                                                      Chapter 2 - Study Design
electronic data deliverables (EDDs) in either Excel or delimited text format. EDDs were first
imported into the database as new tables that remained unaltered while the fields of interest,
contained therein, were appended to a table called "Vessel Results." The remaining fields were
populated using queries. Ten percent of the data in the Vessel Results table from three of the four
laboratories were compared to the original hard copy reports, if provided. This ensured
consistency between the EDD and hard copy report, as well as validated the importing
procedure. As a further quality assurance measure, a 100-percent check was done comparing
these PDF reports to database entries derived from the fourth lab's EDD reports.

       Data that were not received in EDD format (i.e., hard copies, PDFs, and field data sheets)
were manually entered directly into the Vessel Results table. These data were provided by six
additional analytical chemistry and subcontract laboratories: EPA Region 5 (Chicago, Illinois),
Biomarine (Gloucester, Massachusetts), EnviroChem (Mobile, Alabama), QC Laboratories
(Southampton, Pennsylvania), Northeast Environmental Laboratory (Danvers, Massachusetts),
and Sitka Water Treatment Plant (Sitka, Alaska). As with the COC information, a second person
did a 100-percent check of the accuracy of data entry.

       In addition to checking for reporting accuracy, a check of laboratory QC procedures was
performed.  EPA examined laboratory QC parameters, including method type, hold times,
laboratory blanks and duplicates, laboratory control samples, and surrogate recovery, where
applicable,  for all subcontract laboratories.

 2.3   DATA CONSIDERATIONS AND STUDY LIMITATIONS

   2.3.1  Voluntary Nature of the Sampling Program

       All vessel sampling performed for this study was conducted on a voluntary basis (i.e.,
vessel owners/operators voluntarily allowed EPA to sample their vessels). As such, the selection
process was not completely random from within the universe of study vessels, nor were the
vessels sampled unannounced, with the possible exception of fishing vessels (see Section 2.2.1).
These issues raise potential concerns regarding the representativeness of the sampling and the
statistical uncertainty of the resulting data analyses. To minimize these concerns, EPA provided
study volunteers with guidance for participation in the sampling program. This guidance stressed
EPA's desire to sample normal discharge cycles/events and requested that volunteers not alter
vessel operations from normal (typical) operation. The guidance specifically instructed that
volunteers should not perform any special cleaning in preparation for sampling, add or  eliminate
or alter any typical discharges, or increase  or decrease the volume or other characteristics of
discharges, etc. Also, as EPA preferred to collect samples pierside rather than underway, EPA
instructed volunteers to inform the Agency if conducting sampling pierside compromised, in any
way, the characteristics of discharges (sources, volumes, composition).
                                           61

-------
                                                                       Chapter 2 - Study Design
       As a further consideration, EPA assumed that most of the volunteers were generally
'good actors' who would have the best maintained vessels and be in compliance with all existing
applicable regulations, which could also affect the representativeness of the data collected for the
vessel class as a whole.

   2.3.2   Vessels/Discharges Not Sampled

       While this study included the sampling of a large number of discharges from a large
number of vessels, certain vessel classes and discharges were either not sampled at all or
received only limited sampling due to overall program schedule and resource constraints or other
factors (see Sections 2.2.1 and 2.2.2). EPA supplemented its sampling program with information
and data collected from other data sources to the extent possible; however, the Agency
acknowledges remaining gaps in achieving the study objectives for certain segments of the
industry. In particular, EPA has little or no information or data regarding freight barges, freight
ships, tank  barges,  and tank ships less than 79 feet in length (estimated to represent 7 percent of
study vessels). In addition, EPA has little information or data regarding the applicability of
several discharges listed in Section 2.2.2 to study vessels.

       EPA's ability to fully characterize certain discharges was limited by some practical
considerations. For example, on many vessels, discharges were too close to the waterline, or
even under the waterline, precluding the ability to collect an uncontaminated sample. Installation
of sample taps upstream of these discharge ports was either impossible (i.e., would compromise
system integrity) or impractical within time constraints for the sampling events. On other vessels,
collection of vessel discharges under normal operations was either impossible or unsafe. These
conditions included:

   •   Vessel configurations blocking access to  discharge ports
   •   Discharge volumes insufficient for sampling
   •   Discharges not generated during the sampling event
   •   Systems such as generators not operational during the sampling event
   •   Systems operated only during emergency
   •   Discharges requiring underway sampling
   •   Fishing vessel platforms inactive during the sampling schedule
   •   Fishing seasons closed or outside the sampling schedule
   •   Inability to sample all U.S. fisheries

       As an example, EPA was able to sample bilgewater on only eight of the 61 sampled
vessels (13  percent). Bilgewater sampling was infeasible for approximately three quarters of the
remaining vessels for three reasons. First, automatic bilge pumps operating while the vessel was
underway resulted  in an empty bilge when the vessel returned to pier. Manual activation of the
bilge pump on these vessels did not result in any discharge or only a small volume of discharge.
Second, as  a matter of policy, many vessels restrict bilgewater discharges to only while
                                           62

-------
                                                                       Chapter 2 - Study Design
underway or when outside U.S. waters due to possible concerns of exceeding existing Clean
Water Act § 311 requirements. Third, some bilgewater discharges were too close to the waterline
for sampling. For the remaining one quarter of vessels, sampling was not performed because the
vessels never discharge bilgewater. On these vessels, a contractor steam cleans the bilges once
per month, and the resulting cleaning waste is removed from the vessels for shoreside disposal.

   2.3.3  Pollutants Not Sampled

       A few candidate analyte groups (pesticides, polychlorinated biphenyls, dioxins/furans,
flame retardants, uranium, and asbestos) were not selected for analysis, as they are not
anticipated to be present in the vessel discharges due to the lack of a readily apparent source for
these pollutants.

       While EPA's list of target analytes includes many persistent, bioaccumulative, and toxic
chemicals (PBTs),  many other PBTs were not analyzed for due to the lack of test methods or
resources. In  general, these unanalyzed compounds either have no known use or source onboard
vessels or have no readily available means to enter the vessel discharges. Mercury was not
selected for analysis because it requires specialized sampling techniques inapplicable to vessel
sampling to minimize the potential for sample contamination (e.g., vessel sampling cannot be
conducted away from sources of metals or sources of airborne contamination such as engines or
generators).

       Test methods for pharmaceuticals and personal care products (PPCPs) have recently been
developed; however, EPA did not  select this analyte group for analysis due to a lack of
resources. These compounds are most likely to be found in sewage, which is outside the scope of
this study; however, they can also  be expected to be found in graywater sources, such as sink and
shower wastewater, albeit at very low concentrations.

       Although ballast water, and its assessment as a vector for aquatic  invasive species, was
specifically excluded from this study by the statutory language in P.L.  110-299 (see Appendix
C), EPA recognizes that other vessel discharges, such as bilgewater; stern tube packing gland
effluent; fish  hold effluent; and discharges from vessel hulls, propellers, and other exposed
surfaces are also potential vectors  for the spread of aquatic invasive species. EPA excluded any
aquatic invasive species characterization as part of this study in consideration of overall program
schedule and resources.

   2.3.4  Application to Other Vessels, Including Larger Vessels Not Sampled for this
          Study

       EPA's primary objective in conducting the vessel sampling program was to characterize
discharges specific to commercial  fishing vessels and nonrecreational vessels less than 79 feet
(i.e., study vessels). Some  data are applicable to other vessels, however, including larger vessels
not sampled for this study. This subsection discusses EPA's consideration of the applicability  of

                                           63

-------
                                                                        Chapter 2 - Study Design
these sample data to other vessels, as well as factors that data users should consider in
determining the broader applicability of the data.

Bilgewater.

       The composition and volume of bilgewater is highly dependent on the specific sources of
wastewater that accumulate in bilge, as well as vessel size, hull design and construction, vessel
operation, and a variety of additional factors (see Section 1.3). Any researcher, regulator, or
other stakeholder who subsequently uses the data collected in this study should evaluate and
compare the characteristics of the vessels sampled for this study to those of other vessels to
determine the applicability of EPA's sampling data. In general, EPA believes that the design,
construction, and operation of vessels not sampled for this study (e.g., cruise ships, ferries,
barges, tankers) differ considerably from those of the sampled vessels, which would result in
significantly different bilgewater characteristics. Hence, EPA cautions against applying the
limited bilgewater results from this study to all vessels.

Stern tube packing gland effluent.

       This discharge applies to vessels that collect the ambient water that leaks through the
stuffing box and packing gland that surround the propeller shaft in a segregated area from the
general bilge. During this study, EPA observed this segregated discharge onboard tugboats but
not on any other vessel classes. On tugboats, the stuffing box is packed with greased flax rings.
EPA's stern tube packing gland effluent data should be applicable to other vessel classes (if any)
that use this same type  of stern tube packing gland and that collect the resulting wastewater for
segregated discharge.

Deck runoff and/or washdown.

       Factors contributing to the volume and composition of deck runoff and/or washdown
include deck equipment and operations, deck surface material, and method of washing the deck
(see Section 1.3). Data users should evaluate and compare the characteristics of the vessels
sampled for this study to those of other vessels to determine the applicability of EPA's sampling
data. In general, EPA believes that deck operations performed on vessels outside the scope of
this study differ significantly from those of the sampled vessels. For example, deck washdown
generated by fishing vessels might be applicable only to this industry, particularly in cases where
these vessels are washing significant organic material from fishing operations overboard. As
another example, only one sampled vessel, a supply boat, is used to support the transfer and
handling of non-fish cargo. On the other hand, deck washdown from sampled passenger vessels
might apply to other vessels, such as larger tour boats,  water taxis, and possibly cruise ships.
                                            64

-------
                                                                       Chapter 2 - Study Design
Fish hold effluent (including both refrigerated seawater effluent and ice slurry) and effluent from
the cleaning offish holds.

       Since only commercial fishing vessels or tenders use fish holds for storing seafood
products or fish, EPA believes that fish hold effluent discharges are unique to commercial
fishing operations and are not applicable to other vessels.

Graywater.

       The graywater sources sampled by EPA for this study are "domestic" in nature, such as
sink water from washing hands and dishes, wastewater from shower stalls, and laundry water
from domestic washing machines. EPA cautions the data user against applying these sampling
data to non-domestic graywater operations, such as large-scale industrial dishwashing  and
laundry equipment. In addition, the graywater sources sampled by EPA were discharged
immediately upon generation; therefore, these data do not represent graywater that has been
retained in collection or storage tanks or graywater mixed with sewage. Finally, EPA's
graywater data do not apply to wastewater discharges from food waste processing operations,
such as food grinders or food pulping systems.

Propulsion and generator engine effluent.

       For this study, EPA sampled propulsion and generator effluent from a large number and
variety of engines. These include:

       •  Inboard and outboard.
       •  Two-stroke and four-stroke.
       •  Spark ignition and compression ignition.
       •  Diesel- and gasoline-fueled.
       •  New and existing.
       •  Direct cooling systems (raw water directly cools the engine) and indirect cooling
          systems (raw water cools antifreeze, which cools the engine).
       •  With and without wet engine exhausts (some raw water is injected into the  exhaust to
          cool and quiet the exhaust).
       •  Variety of manufacturers, sizes, and engine horsepower.
       •  Operation at varying engine power levels (i.e., idle, slow troll, half throttle, three-
          quarters throttle, and full throttle) depending on vessel use.

       EPA also observed a number of vessels, such as tug  boats and larger commercial fishing
vessels, that use keel-cooled propulsion engines and generators. The closed-loop cooling systems
used on these engines do not discharge any wastewater.

       Based on an evaluation of the engine effluent sampling results, EPA observed significant
differences in the nature and composition of discharges from inboard and outboard propulsion
engines and from generators. EPA may also have observed differences between diesel- and

                                           65

-------
                                                                      Chapter 2 - Study Design
gasoline-fueled inboard propulsion engines; however, the data set was too small to be
conclusive. Based on these findings, EPA believes the engine effluent data are applicable only to
engines of similar types, specifically inboard propulsion versus outboard propulsion versus
generators and diesel- versus gasoline-fueled engines.

Firemain systems.

       EPA sampled relatively few firemain systems for this study. Firefighting equipment
requirements are specified by the U.S. Coast Guard and differ by vessel type, size, construction
(e.g., open decks versus enclosed spaces with potential to entrap explosives, flammable gases, or
vapors), whether or not the vessel carries passengers for hire, and many other factors. Not all
vessels within the scope of this study are required to carry firefighting equipment. For those
vessels that require firefighting equipment, these requirements are often satisfied by carrying
hand-portable fire extinguishers rather than firemain systems. For vessels outfitted with firemain
systems, the systems are used during emergency and testing. None of the vessels visited by EPA
for this study tests its firemain system more frequently than once every two weeks,  and none
operates its system for secondary purposes such as deck washing. Operating personnel from
three tour boats and two tugboats that EPA visited agreed to engage their firemain systems for
EPA sampling. Most operated wet rather than dry systems. The resulting sampling data apply
primarily to wet-type firemain systems that are operated infrequently, with intake provided by
surrounding water and without additions to the discharge (e.g., no addition of foam-forming
agents).
                           Firemain System on a Passenger Vessel.

       EPA also sampled the firefighting system onboard a fire boat; however, these sampling
data may only apply to fire boats or other vessels equipped with high-pressure/high-volume fire
pumps.
                                           66

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


CHAPTER 3

ANALYSIS OF DISCHARGES AND POTENTIAL IMPACT TO
HUMAN HEALTH AND THE ENVIRONMENT	

       This chapter summarizes the results of the wastewater characterization data for the nine
types of vessel discharges sampled from the 61 vessels identified in Chapter 2. It includes the
characterization of the nature, type, and composition of discharges for each class of vessel, as
well as other relevant information collected regarding shipboard processes, equipment, materials,
and operations that might contribute to the level or explain the presence of pollutants in these
discharges.

       This chapter begins with a description of the approach used for the analyses of
contaminants in the various discharges of the vessel classes of interest in this sampling program,
and the specific procedures used to reduce, present, and interpret these data. Each section in the
chapter presents and discusses in detail the results found for each discharge type selected for
evaluation in the vessel classes of interest and summarizes the major findings for the discharges
associated with each major vessel type. The final section discusses anti-foulant hull coating,
which warrants discussion based on the results of other studies conducted on this discharge type
even though EPA did not sample this discharge in this study.

3.1           APPROACH TO ANALYSES

      EPA's approach was designed to ensure that the analyses conducted under this study
would be as comprehensive as possible and provide results that would represent the different
vessels and discharges to the greatest extent possible. EPA included the discharge data collected
from the vessels selected for this study (primary data) and any relevant data collected from other
studies (secondary data) (e.g., engine effluent from small Armed Forces vessels  covered under
EPA's sampling program for the Uniform National Discharge Standards (UNDS) rulemaking).
Where appropriate, EPA also assessed ambient (harbor) and potable waters at each geographic
location where vessels were sampled.

      EPA's analysis attempted to make full use of the primary and secondary  data collected
for this study, including data collected from ambient (harbor) and source (vessel service1 or city
water supply) waters. However, EPA recognizes that the analyses are based on a limited number
of samples; in some cases, on a sample size of fewer than five. These results should be regarded
as preliminary in nature due to statistical considerations related to small sample sizes.
1 Service water here means the vessel potable water supply. For study vessels, vessel service water generally
originates from municipal water supply rather than produced on board.

                                           67

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       EPA also attempted to identify where the analyses and results from this study could be
reasonably extrapolated to vessels other than those vessels sampled in this study. Many of the
discharges are not unique to vessels subject to the P.L. 110-299 moratorium in terms of the
expected pollutants or volumes and may also be found on larger nonrecreational vessels or
recreational vessels.

    3.1.1   Data Reduction and Presentation

       EPA compiled the data collected for the nine vessel discharges sampled from the 61
vessels (see Chapter 2, Section 2.2.5.3) into a Microsoft (MS) Access database developed
specifically for this study. For each discharge type, EPA reduced the data for summary according
to the following procedure.

       First, data were retrieved from MS Access by discharge group, using a query developed
specifically for this task. The queried data included the analytical result with the corresponding
screening benchmark (defined in Section 3.1.3) and ambient and source water  concentrations.
For each discharge group, the queried data were exported to MS Excel, and then resaved as tab-
delimited ASCII text (*.txt) files. Record counts were compared between the discharge group-
filtered MS Access query and the MS Excel and ASCII files to ensure that data were not lost.

       The ASCII data for each discharge group were read into an Interactive  Data Language
(IDL) (Research Systems Inc.,  2003) program that carried out a series of calculations for each
analyte, based on the following algorithm:

    1.  Identify and average concentrations measured for field replicate samples, including
       replacing below-detection concentrations with 1/2 of the reporting limit2 when at least
       one replicate was detected.
    2.  Identify and average concentrations measured for laboratory replicate samples, including
       replacing below-detection concentrations with 1/2 of the reporting limit when at least one
       replicate was detected.
    3.  Identify and average concentrations measured for vessel replicate samples (e.g., multiple
       deck wash, graywater, engine effluent samples from a single vessel), including replacing
       below-detection concentrations with 1/2 of the reporting limit when at least one replicate
       was detected.
    4.  Calculate potential hazard quotients (PHQs) by dividing the vessel average concentration
       by the corresponding screening benchmark, if one was available (see further details
       provided in Section 3.1.3).
 Laboratory analyses for low concentration pollutants report a detection limit (the presence or absence of a
pollutant) and a reporting limit (the level at which the concentration of a pollutant can be quantified with appropriate
certainty). Statistical methods often require replacement of values that are below the detection and reporting limits
of an analytical method (especially for zero values). EPA has established conventions on how to conduct this
replacement. In this study, certain labs were able to provide a reporting limit for only certain analytes, which is not
uncommon. For consistency, EPA chose to use a convention of replacing the nondetects with a value of 1A of the
reporting limit. These are referred to as replacement values below.

                                             68

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
    5.  Output vessel-average results to a comma-separated value (CSV) text file.
    6.  Calculate nonparametric percentiles of the distribution of vessel-average analyte
       concentrations using the algorithm of Hyndman and Fan (2003). Note that below-
       detection vessel average concentrations were not replaced at this step.
    7.  Replace below-detection vessel average concentrations for those analytes where at least
       one concentration was detected with 1/2 of the reporting limit. Calculate detected
       proportion of vessel concentrations for each analyte.
    8.  Output vessel-average results to a CSV text file.
    9.  Calculate average discharge group analyte concentrations from the vessel average
       concentrations, including replacement values.
    10. Output statistics for each analyte (number of samples, number and proportion detected,
       average,  and nonparametric percentiles) to a CSV text file.
    11. Read vessel-average results (including replacement) into SYSTAT Version 6.1 (SPSS,
       1996) to  generate box and density plots for each analyte class (see Section 3.2.1 below).
    12. Read these results into MS Excel and then reassemble into a workbook with the database
       query exported from MS Access. Generate summary data tables from these workbooks.
    13. For each discharge  category, reproduce by hand the data reduction and statistical
       calculations identified above for two or more randomly selected analytes as a QA
       procedure.

       All discharge-specific analytes summarized in subsequent sections of this chapter are
organized into the following major groups: classical  pollutants3, metals, nonylphenols, nutrients,
pathogen indicators, semivolatile organic compounds, and volatile organic compounds. For each
discharge type, the analyte groups  are generally presented according to the order of highest
expected significance or risk in that specific discharge (e.g., the graywater section begins with
pathogen indicators). The specific  list of target analytes by group is provided in Appendix D.
EPA did not analyze all vessel discharges for all selected analyte groups; see Table 2.2 for target
analyte groups by discharge type.

    3.1.2   Summary Statistics and Box Plots

       This chapter includes, for each analyte group within a specific discharge type (e.g.,
bilgewater, deck washdown water), tables that summarize the number of samples analyzed, the
number of times a specific  analyte within an analyte group was detected, the average
concentration (when only one sample was analyzed,  the average is equal to the measurement),
and additional standard summary statistics  related to the measured analyte concentrations
(median, min, max and selected (10th, 25th, 75th, and  90th) percentiles). These additional statistics
were only calculated when a sufficient number of samples had detected values for any given
3 The classical pollutants group of analytes combines several standard water quality parameters such as conductivity,
salinity, temperature, etc. with other parameters EPA defines as conventional pollutants (biochemical oxygen
demand (BOD), total suspended solids (TSS), pH, fecal conform, and oil and grease). For convenience, this group
also includes other common analytes such as total residual chlorine, or TRC. For simplicity, these conventional and
other common analytes and water quality parameters have all been grouped under the term "classical pollutants."


                                            69

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

analyte (usually five detected values or greater). In cases where some of the concentrations of an
analyte were reported as nondetect (censored), the concentration of that sample was estimated as
!/2 of the reporting limit for purposes of calculating average concentrations4.

       In addition to the summary tables, this section includes figures that graphically present
the analyte-specific concentrations that were detected (as well as any replacement values for
nondetects) for each analyte group within a discharge to better identify data trends related to
analytes of potential concern. These figures are shown as box and dot plots, with the names of
the analytes along the x (independent)-axis
and their associated vessel average
concentrations along the y(dependent)-
axis.
What is a Box Plot?
A box plot is a useful, simple statistical tool used
to show basic characteristics of a data set.  A box
plot can show the approximate center of a data set
and how those data are spread over a range in
values - in this case, a range of concentrations.
Below is an example box plot indicative of the type
of graphical data display used throughout this
chapter.
                                                        <8>
       For box plots, the bottom and top
of the box displays the 25th and 75th
percentile concentrations defined as the
interquartile range or IQR (i.e., the "box"
contains 50 percent of the data values),
respectively. The median is displayed as
the horizontal line within the box. The
"whiskers" show the relative distribution
of data points outside of the IQR and
represent 1.5 times the IQR.
Superimposed over each box plot are the
actual data points, shown as small open
circles. Circles surrounded by large circles
are outliers greater or less than  1.5 times
the IQR. Circles covered by asterisks are
outliers greater or less than three times the
IQR.
    3.1.3   Calculation of Potential Hazard Quotients

       To provide a context for the level of contaminant concentrations presented, EPA used
National Recommended Water Quality Criteria (NRWQC)5 and several other benchmarks as a
                                                       T
                                                             Data Point
 See footnote 2.
5 National Recommended Water Quality Criteria (NRWQC) include acute (short-term) and chronic (long-term)
criteria (toxicity threshold values) for the protection of aquatic life, as well as Human Health criteria for protection
of humans from consumption of contaminated water or contaminated water and aquatic organisms. EPA's most
recent compilation of NRWQC (2006) is presented as a summary table containing recommended water quality
criteria for the protection of aquatic life and human health in surface water for approximately 150 pollutants. These
criteria are published pursuant to Section 304(a) of the Clean Water Act (CWA) and provide guidance for states and
                                              70

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

preliminary screen for all discharge data with the potential to cause or contribute to the
nonattainment of a water quality standard in a given receiving water body. The "screening-level"
benchmarks chosen for this purpose are shown in Table 3.1 at the end of this subsection, and
generally represent EPA's most conservative (protective) concentration available for the specific
analyte of interest. Several "legacy" standards (for BOD, TSS and total phosphorus) are also
included with the screening benchmarks. For BOD and TSS, these benchmarks are EPA's
secondary treatment effluent limits for sewage treatment plants6. For total phosphorus, the
benchmark of 0.1 mg/L is from EPA's Gold Book (USEPA, 1986b) and represents a
concentration recommended to prevent nuisance algal blooms resulting from eutrophication in
flowing waters. EPA did not consider it appropriate to apply ecoregional nutrient criteria for this
project.

       EPA's NRWQC are recommended concentrations of analytes in a water body that are
intended to protect human health, aquatic organisms and the water body uses from unacceptable
effects from exposures to these pollutants. The NRWQC are not directly related to analyte
concentrations in a discharge for a number of reasons. First, NRWQC not only have a
concentration component, but also a duration and frequency component. Second, it is not always
necessary to meet all water quality criteria within the discharge pipe to protect the integrity of a
water body (USEPA,  1991). Under EPA's regulations at 40 CFR  122.44(d)(l)(ii), when
determining whether a discharge causes, has the reasonable potential to cause, or contributes to
an in-stream excursion above a narrative or numeric criteria within a state water quality standard,
the permitting authority will use procedures that account for, where appropriate, the dilution of
the effluent in the receiving water. A mixing zone allows for ambient concentrations above the
criteria in small areas near outfalls while dilution occurs. To ensure mixing zones do not impair
the integrity of the water body, the permitting authority will determine the mixing zone such that
it does not cause lethality to passing organisms and, considering likely pathways of exposure,
significant human health risks.
tribes to use in adopting water quality standards. EPA's 2006 NRWQC are available at:
http://www.epa.gov/waterscience/criteria/wqctable/. hereafter referred to as EPA's 2006 NRWQC.
6 Secondary treatment standards for sewage treatment plants were technology-based limits developed in the late
1970s and early 1980s, and are not the same as the water-quality-based criteria in the 2006 NRWQC. Thus, the
PHQs for BOD and TSS calculated as described below are not directly comparable to the PHQs based on criteria
designed to protect aquatic life or human health but, by design, such standards are imposed to limit ecological
impacts.
                                            71

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       Nevertheless, comparing analyte concentrations in vessel discharges to NRWQC (or
other equivalent screening benchmark) provides a conservative screen of whether these
discharges cause, have the reasonable potential to cause, or contribute to nonattainment of the
water quality standards in a water body. EPA calculated hazard quotients (HQs) by dividing the
concentration of a particular analyte7 by its corresponding water quality criterion or other
benchmark as an initial screen for the discharge-specific water sample data. If the concentration
of a given analyte in vessel discharge is less than the applicable screening criterion or benchmark
                                                        values (HQ<1), the discharge would
                                                        likely not cause, have the reasonable
Mitigating Conditions/Circumstances in a Water Body
Compared to the volume of a typical harbor, the effluent
volume of any particular vessel discharge is small (see
Chapter 4). Therefore, even when pollutant concentrations
of a particular effluent are high, the total loading of that
pollutant on the receiving water of the harbor can be
relatively small. Furthermore, most harbors are continually
flushed by freshwater and tidal activity. These dilution
factors, in addition to the mitigating capacity of saltwater
cations and organic matter, may reduce the toxicity  of
many of these pollutants.
                                                        potential to cause, or contribute to
                                                        nonattainment of a water quality
                                                        standard based on that value,
                                                        particularly after considering
                                                        assimilation and/or dilution by the
                                                        receiving water. If the HQ value is
                                                        greater than one, then there is the
                                                        possibility of ecological or human
                                                        health risk as the concentration of a
given analyte in vessel discharge is greater than the applicable screening criterion or benchmark
values (USEPA,  1997). However, because discharges in this study are measured at the "end of
pipe" before being released into a harbor where they are subsequently diluted, HQ values of
greater than one do not necessarily indicate that a discharge poses a significant risk or would be
likely to cause or contribute to a water quality standard exceedance.  Further, the presence of
additional environmental factors such as dissolved organic carbon can reduce the toxicity of
certain pollutants (e.g., metals and many organic pollutants) and reduce the likelihood of
ecological or health risk. Because of these additional considerations, EPA uses the term potential
hazard quotients  (PHQs) instead to indicate this difference, as the PHQs are only intended to
indicate that a screening benchmark was exceeded and the discharge thus warrants further
consideration regarding the potential to cause or contribute to nonattainment of water quality
standards8.

       Mobile sources such as vessels complicate the analysis because they discharge into many
different water bodies, but in general, greater mixing and dilution would be expected for
discharges from vessels than from stationary sources when they are in motion while discharging.
EPA acknowledges that vessel discharges to areas with high vessel traffic, areas with a low
7 PHQs were also calculated using replacement values for nondetected concentrations, so that such results would be
represented in the box and scatter plots. Note: all PHQ values in box plots that were calculated with replacement
values throughout this chapter are circled.
8 EPA does not consider a PHQ that exceeds 1 to signal that these discharges pose a potential risk to cause or
contribute to the non-attainment of a water quality standard when the PHQ is based on replacement values for
nondetected concentrations.
                                             72

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

degree of flushing, or impaired water bodies could reduce mixing and dilution. With these
factors in mind and assuming the data from this study are representative of the class of vessels as
a whole, a PHQ marginally above a value of 1 is most likely not of significant concern. On the
other hand, a PHQ value substantially above 1 (e.g., 10 or 100) may be more likely to be of
concern, particularly if the discharge is of significant volume, is in an area of low flushing, is in
an area where there is a high degree of vessel traffic, or is in a waterbody that is already impaired
or under other stress.

       EPA recognizes that one  of the key factors in evaluating metal toxicity is the
bioavailability of the metal to an organism. Exposure to metals at toxic levels can cause a variety
of changes in biochemical,  physiological, morphological, and behavioral patterns in aquatic
organisms. In the aquatic environment, elevated concentrations of dissolved metals can be toxic
to many  species of algae, crustaceans, and fish. Some metals have a strong tendency to adsorb to
suspended organic matter and clay minerals or to precipitate out of solution, thus removing the
metal from the water column. The tendency of a given metal to adsorb to suspended particles is
typically controlled by the pH and salinity of the water body, as well as the organic carbon
content of the suspended particles. If the metal is highly sorbed to particulate matter, then it is
not likely to be in a dissolved form that aquatic organisms can process (i.e., bioavailable)9.

       Accordingly, NRWQC for the protection of aquatic life for metals are typically expressed
in the dissolved form. Therefore, a high concentration of a metal measured in its total form
(dissolved and particulate) may not be an accurate representation of its toxic potential to aquatic
organisms. In contrast, human health criteria (for the consumption of organisms) for metals are
commonly expressed in the total metal form because human exposure to pollutants is assumed to
be through the consumption of organisms, where the digestive process is assumed to transform
all forms of metals to the dissolved phase, thus increasing the amount of biologically available
metals. EPA was mindful of this distinction between aquatic life and human health criteria for
metals when comparing the dissolved and total metals concentration data in the various
discharges to NRWQC and when calculating PHQs using the screening benchmarks.  In
particular, in considering the potential for vessel discharges to pose a risk to human health, EPA
also noted the likelihood of human exposure to such discharges (e.g., potential for receiving
water to be used as drinking water source).
9 Note that the bioavailability of metals is a relative term and depends on many factors. For example, particulate
metals complexed to suspended organic matter or clay minerals may be recycled into the water column and become
bioavailable due to physical resuspension (dredging activities) of bed sediments or bioturbation (the stirring or
mixing of sediment particles by benthic animals). Depending on conditions in the water column and microbiological
activity within the surficial sediment and overlying water surface layers, these physical and biological actions might
remobilize the metals in the dissolved bioavailable form for potential uptake by aquatic organisms. Likewise, certain
benthic organisms called deposit feeders might consume particulate-bound metals and re-release metals via
digestion and excretion or introduce metals into the food chain when consumed by predators.

                                             73

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       EPA chose to include the major cations calcium, magnesium, potassium and sodium in
the metals analysis to further characterize the vessel discharges. As common ions in surface
waters, the concentrations of these ions are indicative of the sample matrix (i.e., freshwater,
saltwater, brackish water) rather than pollutant loadings. Accordingly, major cation
concentrations are typically elevated (up to three orders of magnitude higher) relative to other
metals included in the metals analysis (e.g., copper, lead, and zinc). For example, the typical
concentrations of major cations in full and partial (brackish) strength seawater and in freshwater
of various total water hardness levels are listed in Tables 3.2 and 3.3 below. Major cations are
not toxic except at extreme, uncommon levels.

       For convenience, data tables for metals in this chapter segregate the presentation of major
cation concentration data from that of the other metals to clearly distinguish between the
naturally occurring cations and other metals of potential concern in vessel discharges. It is worth
noting that many of the samples collected for this study consist entirely or partially of sea water;
consequently, these samples can have high concentrations of components (e.g., salts) that can
interfere with the analytical measurement of the chemical of interest. EPA evaluated whether
measured concentrations of selenium and arsenic may have exhibited "positive interference"
(i.e., the measured concentration is higher than the actual concentration in the sample - see text
box for more technical information). EPA found that trace metal analysis using a conventional
ICP-MS-based analytical method may have resulted in positive interference for some samples of
selenium, and to a lesser extent, arsenic. However, for the majority of samples analyzed in this
study, either the samples contained few interferences (i.e., samples were from freshwater) or
alternate instrumentation, which had the capability of minimizing sample interferences, was used
for analyte measurement. Hence, the majority of arsenic and  selenium results did not have
positive interference.  EPA identified
the few samples analyzed using the
conventional ICP-MS method, which
may have yielded artificially  high
values for the measured concentration
of arsenic and selenium.  Therefore,
while such positive interferences were
not found to influence the overall
findings presented in this study, the
selenium and arsenic concentrations
potentially affected by positive
interference are identified (noted by
footnote in each instance) throughout
this chapter.
Explanation of Possible Positive Interference on Select
Arsenic and Selenium Measurements
Positive interference occurs when components of a
sample, other than the analyte, affect the measurement of
the analyte of interest by yielding an  artificially high value.
This occurs when components in the sample interfere with
the analytical methodology. Some of the components of
sea water (e.g., calcium, magnesium, potassium, sodium),
are known to cause positive interference with certain trace
elements, such as arsenic and selenium. The potential for
interference is based on the analytical method and
instrumentation used for the measurement. In these
situations, alternate sample preparation or analytical
instrumentation may be required to eliminate or reduce
sample interferences, in order to maintain analyte
sensitivity.
                                            74

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1. Water Quality and Other Benchmark Values Used to Screen the Vessel
Discharge Data
Analyte
1 ,1 ,2,2-Tetrachloroethane
1,1,2-Trichloroethane
1,1-Dichloroethene
1 ,2,4,5-Tetrachlorobenzene
1 ,2,4-Trichlorobenzene
1,2-Dichlorobenzene
1,2-Dichloroethane
1,2-Dichloropropane
1,2-Diphenyl hydrazine
1 ,3-Dichlorobenzene
1 ,3-Dichloropropane
1,4-Dichlorobenzene
2,3,7,8-TCDD (Dioxin)
2,4,5-Trichlorophenol
2,4,6-Trichlorophenol
2,4-Dichlorophenol
2,4-Dimethylphenol
2,4-Dinitrophenol
2,4-Dinitrotoluene
2-Chloronaphthalene
2-Chlorophenol
2-Methyl-4,6-Dinitrophenol
3,3'-Dichlorobenzidine
4,4'-DDD
4,4'-DDE
4,4'-DDT
4,6-Dinitro-2-methylphenol
Asbestos
Acenaphthene
Acrolein
Acrylonitrile
Aldrin
Alkalinity
alpha-BHC
alpha-Endosulfan
Aluminum, Total
Ammonia As Nitrogen (NH3-N)
Anthracene
Antimony, Total
Arsenic, Total
Arsenic, Dissolved
Barium, Total
Benzene
Benzidine
Screening
Benchmarks
0.17
0.59
330
0.97
35
420
0.38
0.5
0.036
320
0.34
63
5.0E-09
1800
1.4
77
380
69
0.11
1000
81
13
0.021
0.00031
0.00022
0.0010
13
7000000
670
6.0
0.051
1.3
20000
0.0026
0.0087
87
1.2
8300
5.6
0.018
36
1000
2.2
0.000086
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
fibers/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
mg/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
Source1
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH Org Only
2006 NRWQC HH Org Only
2006 NRWQC CCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC SW CMC
2006 NRWQC FW CCC
2006 NRWQC HH W+O
2006 NRWQC SW CCC
2006 NRWQC FWCCC
2006 NRWQCSWCCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQCSWCCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
                                            75

-------
Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Analyte
Benzo(a)Anthracene
Benzo(a)Fluoranthene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(k)Fluoranthene
beta-BHC
beta-Endosulfan
Biochemical Oxygen Demand (BOD)
Bis (2-Chloroethyl) ether
Bis (2-chloroisopropyl)ether
Bis(2-Chloroethyl)ether
Bis(2-Ethylhexyl) phthalate
Bromodichloromethane
Bromoform
Bromomethane
Butyl benzyl Phthalate
Cadmium, Dissolved
Carbon tetrachloride
Chlordane
Chloride
Chlorobenzene
Dibromochloromethane
Chloroform
Chlorophenoxy Herbicide (2,4,5,-TP)
Chlorophenoxy Herbicide (2,4-D)
Chloropyrifos
Chromium, Dissolved
Chrysene
Copper, Dissolved
Copper, Total
Cyanide
Demeton
Diazinon
Dibenz(a,h)Anthracene
Chlorodibromomethane
Dichlorobromomethane
Dieldrin
Diethyl Phthalate
Dimethyl phthalate
Di-n-butyl phthalate
Dinitrophenols
E. Coli by MPN
Endosulfan Sulfate
Endrin
Endrin Aldehyde
Enterococci by MPN
Ether, Bis(Chloromethyl)
Ethylbenzene
Screening
Benchmarks
0.0038
0.0038
0.0038
0.0038
0.0038
0.0091
0.0087
30
0.030
1400
0.030
1.2
0.55
4.3
47
1500
0.25
0.23
0.0040
230000
130
0.40
5.7
10
100
0.0056
11
0.0038
3.1
1300
1.0
0.10
0.17
0.0038
0.40
0.55
0.0019
17000
270000
2000
69
126
62
0.0023
0.29
33
0.00010
530
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
mg/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
MPN/100ml
ug/L
ug/L
ug/L
MPN/100ml
ug/L
ug/L
Source1
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006NRWQCSWCCC
1984 Secondary Treatment Effluent Limits
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC FWCCC
2006 NRWQC HH W+O
2006NRWQCSWCCC
2006 NRWQC FWCCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC SWCCC
2006 NRWQC FWCCC
2006 NRWQC HH W+O
2006 NRWQC SWCCC
2006 NRWQC HH W+O
2006 NRWQC SW CMC
2006 NRWQC FWand SWCCC
2006 NRWQC FW CMC and CCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC SWCCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
1986 NRWQC B FW
2006 NRWQC HH W+O
2006 NRWQC SWCCC
2006 NRWQC HH W+O
1986NRWQCBFW
2006 NRWQC HH W+O
2006 NRWQC HH W+O
                         76

-------
Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Analyte
Fecal Coliform by MF
Fluoranthene
Fluorene
Gamma-BHC (Lindane)
Guthion
Heptachlor
Heptachlor Epoxide
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclo-hexane-Technical
Hexachlorocyclopentadiene
Hexachloroethane
Hexane Extractable Material (HEM)
ldeno(1 ,2,3-cd)Pyrene
Iron, Total
Isophorone
Lead, Dissolved
Malathion
Manganese
Mercury
Methoxychlor
Methylene chloride
Mirex
Nickel, Dissolved
Nickel, Total
Nitrates
Nitrobenzene
Nitrosamines
Nitrosodibutylamine.N
Nitrosodiethylamine.N
Nitrosopyrrolidine,N
N-Nitroso Di-n-propylamine
N-Nitrosodimethylamine
N-Nitrosodiphenylamine
Pa rath ion
Pentachlorobenzene
Pentachlorophenol
Phenol
Phorphorus (as phosphate)
Polychlorinated Biphenyls (PCBs)
Pyrene
Selenium, Dissolved
Selenium, Total
Silica Gel Treated HEM (SGT-HEM)
Silver, Dissolved
Solids Dissolved and Salinity
Sulfide-Hydrogen Sulfide
Tetrachloroethene
Screening
Benchmarks
14
130
1100
0.16
0.010
0.0036
0.0036
0.00028
0.44
0.0123
40
1.4
15
0.0038
300
35
2.5
0.1
50
0.77
0.03
4.6
0.001
8.2
610
10000
17
0.0008
0.0063
0.0008
0.016
0.005
0.00069
3.3
0.013
1.4
7.9
21000
0.1
0.000064
830
5
170
15
1.9
250000
0.002
0.69
Units
MPN/100ml
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
mg/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
mg/L
ug/L
ug/L
ug/L
ug/L
mg/L
ug/L
ug/L
mg/L
ug/L
Source1
1976QCWSH
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC SW CMC
2006 NRWQC FWand SWCCC
2006 NRWQC SWCCC
2006 NRWQC SWCCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
MARPOL 73/78
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC FWCCC
2006 NRWQC FWand SWCCC
2006 NRWQC HH W+O
2006 NRWQC FWCCC
2006 NRWQC SWCCC
2006 NRWQC HH W+O
2006 NRWQC FWand SWCCC
2006 NRWQC SWCCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC FWCCC
2006 NRWQC HH W+O
2006 NRWQC SWCCC
2006 NRWQC HH W+O
EPA1986Goldbook
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC FWCCC
2006 NRWQC HH W+O
MARPOL 73/78
2006 NRWQC SW CMC
2006 NRWQC HH W+O
2006 NRWQC FWand SWCCC
2006 NRWQC HH W+O
                         77

-------
                           Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Analyte
Thallium, Total
Toluene
Total Nonylphenols
Total Phosphorus
Total Polychlorinated Biphenyls
Total Suspended Solids (TSS)
Total Residual Chlorine (TRC)
Toxaphene
trans-1 ,2-Dichloroethene
Tributyltin (TBT)
Trichloroethene
Vinyl chloride
Zinc, Dissolved
Zinc, Total
Screening
Benchmarks
0.24
1300
1.7
0.1
0.000064
30
0.0075
0.0002
140
0.0074
2.5
0.025
81
7400
Units
ug/L
ug/L
ug/L
mg/L
ug/L
mg/L
mg/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
Source1
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006NRWQCSWCCC
1986 NRWQC
2006 NRWQC HH Org Only
1984 Secondary Treatment Effluent Limits
2006 NRWQC SWCCC
2006 NRWQC FWand SWCCC
2006 NRWQC HH W+O
2006 NRWQC SWCCC
2006 NRWQC HH W+O
2006 NRWQC HH W+O
2006 NRWQC SWCCC
2006 NRWQC HH W+O
    (1)  Sources:
        MARPOL 73/78: International Convention for the Prevention of Pollution From Ships, 1973 as modified by the Protocol of
        1978 (MARPOL 73/88, 1978).
        1976 QCW SH (shellfish harvesting): Note MPN is most probable number and approximates the unit of measure for fecal
        coliform in this study of CPU (colony forming units) (USEPA, 1976).
        1984 Secondary Treatment Effluent Limits: 49 FR 37006, Sept. 20, 1984.
        1986 NRWQC B FW (bathing (full body contact) recreational waters-fresh water): (USEPA, 1986).
        Quality Criteria for Water 1986 (Goldbook) (USEPA, 1986b).
        2006 NRWQC FWCCC (freshwater chronic): (USEPA, 2006).
        2006 NRWQC SWCCC (saltwater chronic) (USEPA, 2006).
        2006 NRWQC SWCMC (saltwater acute) (USEPA, 2006).
        2006 NRWQC HH Org Only (human health for the consumption of organism only) (USEPA, 2006).
        2006 NRWQC HH W+O (human health for the consumption of water + organism) (USEPA, 2006).
Table 3.2. Major Cation Concentrations in Seawater
Seawater Salinity Level
Full Strength 1 (35 ppt
salinity)
Brackish2 (10 ppt
salinity)
Calcium, mg/L
400
114
Magnesium, mg/L
1,350
386
Potassium, mg/L
380
109
Sodium, mg/L
10,500
3,000
(1) Source: Mowka, 2009.
(2) Calculated from full strength seawater concentrations, assuming dilution by ion-free water.


Table 3.3. Major Cation Concentrations in Freshwater
Freshwater Hardness
Level
Soft (40-48 mg CaCO3/L) 1
Moderately Hard (80-100
mg CaCO3/L) 1
Hard (160-180 mg
CaC03/L) l
Calcium, mg/L
6.99
14.0
27.9
Magnesium, mg/L
6.06
12.1
24.2
Potassium, mg/L
1.05
2.10
4.20
Sodium, mg/L
13.1
26.3
52.5
(1) Source: USEPA, 2007.
                                                   78

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
3.2          CHARACTERIZATION OF DISCHARGES

       Each subsection of Section 3.2 presents in detail the observed results for the discharge
types selected for evaluation in the study vessels: bilgewater; stern tube packing gland effluent;
deck runoff and/or washdown; fish hold effluent (both refrigerated seawater effluent and ice
slurry) and effluent from the cleaning offish holds; graywater; propulsion (inboard and
outboard) and generator engine effluent; and discharges from firemain systems. Tables and
figures are presented at the end of each subsection.

   3.2.1  Bilgewater

       Bilgewater can be found on board every vessel and describes the water that collects in the
bottom of a vessel. This water may be from rough seas, rain, minor leaks in the hull or stuffing
box, etc. Depending on the ship's  design and function, bilgewater sometimes contains
contaminants such as oil, fuel, graywater, detergents,  solvents, chemicals, pitch, and particulates.
For this study, EPA collected bilgewater samples from seven vessels: two tow/salvage vessels,
two water taxis, one longline fishing vessel, one shrimping vessel, and one tour boat.

       Based on data and field observations from EPA's vessel sampling program, as well as
information from secondary data sources, EPA estimates many commercial vessels generate, on
average, between 10 and 15 gallons per day (gpd) of bilgewater depending on the vessels'
configuration and intended use; however, EPA noted that vessels might generate as little as 2
gallons of bilgewater or as much as 750 gallons of bilgewater per day. For vessels such as small
tow/salvage vessels or water taxis with open bows, bilgewater pump-out can occur frequently
throughout the day, resulting in small volumes during each pump-out cycle (1-2 gallons). Larger
vessels such as commercial fishing boats are likely to pump less frequently due to larger storage
capacity in the bilge; however, the bilgewater discharge volume can be hundreds of gallons. For
example, EPA noted that a 26-foot,  center console Boston Whaler being used as a tow/salvage
vessel had accumulated only 2 gallons of bilgewater following a tow activity. However, a 62-
foot shrimp boat sampled by EPA in the Gulf of Mexico discharged approximately 750 gallons
of bilgewater during the daily pump-out.

       In general, the volume of bilgewater generated by commercial fishing boats and
commercial vessels depends on the following factors:

       •  Hull and deck construction
       •  Vessel size
       •  Precipitation
       •  Frequency of deck cleaning
       •  Amount of spray reaching the deck(s)
       •  Accidental  spills
       •  Integrity of hull and below-deck piping systems
       •  Potential for condensate formation in below-deck areas.
                                           79

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       Commercial vessels with open bow and stern areas (e.g., commercial fishing and
tow/salvage vessels) have relatively large deck areas that are exposed to precipitation, spray, and
cleaning water, which results in greater bilgewater volumes compared to vessels such as tour
boats or water taxis that have less exposed deck. Other sources that contribute to bilgewater
onboard commercial vessels include small leaks in potable water, graywater and sewage piping
systems, and condensates from the interior of the hull or refrigeration systems. The volume of
these additional bilgewater sources is also highly vessel-specific.

       In this vessel sampling program, EPA collected single grab samples of bilgewater
discharge from selected vessels for laboratory analysis. The results of the analysis were intended
to be representative of bilgewater pollutant concentrations over the range of normal vessel
operations. Collecting bilgewater samples proved difficult for EPA for a number of reasons
including: (1) automatic bilge pumps would discharge insufficient volumes of bilgewater in a
single operating cycle, (2) vessel operators were generally reluctant to discharge bilgewater for
fear of exceeding existing CWA §311  requirements (oily discharges), and (3) sampling was
often impractical because bilgewater was typically discharged via thru-hull openings located at
or near the vessel's waterline.

       Bilgewater samples were analyzed for a wide range of pollutants including metals,
classical pollutants, pathogen indicators,  nutrients, semivolatile and volatile organic compounds,
and nonylphenols. Results for each class  of pollutant are presented and discussed in the
following subsections.

       3.2.1.1     Metals

       Bilgewater samples were analyzed for dissolved10 and total (dissolved plus particulate)
concentrations of metals. The analytical results are summarized in Table 3.1.1 for dissolved
metals and in Table 3.1.2 for total metals that were detected in at least one bilgewater sample.
The following metals were measured in all bilgewater samples:

       •   Total aluminum
       •   Total arsenic11
       •   Dissolved and total barium
       •   Dissolved and total calcium
       •   Dissolved and total copper
       •   Dissolved and total magnesium
       •   Dissolved and total manganese
       •   Dissolved and total potassium
10 Dissolved metals were obtained by filtering the water sample.
11 Note that for three of the seven bilgewater samples analyzed, EPA suspects that the measured arsenic
concentrations are likely to be overestimated due to the positive interference of major cations in seawater (see
discussion on page 74).

                                            80

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment



       •   Dissolved and total sodium
       •   Dissolved and total zinc.


Concentrations of other metals were measured in 50 percent or more of the samples analyzed:


       •   Dissolved aluminum
       •   Dissolved arsenic12
       •   Dissolved and total chromium
       •   Total iron
       •   Total lead
       •   Dissolved and total nickel
       •   Dissolved and total
           selenium13.
Dissolved versus Total Metals
EPA recommends using dissolved metal to set and
measure compliance with water quality standards because
dissolved metal more closely approximates the
bioavailable fraction of metal in the water column than
does total recoverable metal (USEPA, 1993). EPA
considers that the primary mechanism fortoxicity to
organisms that live  in the water column to be adsorption to
or uptake across the respiratory surfaces of aquatic
organisms (i.e., the gills) as well as the carapace of certain
invertebrates, and this physiological  process requires
metal to be in a dissolved form. This is not to suggest that
particulate metals are nontoxic; rather, because toxicity of
particulate metals are primarily restricted to direct
ingestion via dietary exposure, they are less toxic overall
compared to dissolved metal (USEPA, 1996). There are
exceptions, however, particularly for bottom feeding
organisms, and for  metals that bioaccumulate (also see
footnote in Section  3.1.3 regarding physical  and biological
recycling of particulate metals). Dissolved metal is
operationally defined as that which passes through a 0.45-
um or a 0.40-um filter and particulate metal is operationally
defined as total recoverable  metal minus dissolved metal.
EPA typically uses the dissolved fraction, or fd, to express
the fraction of the total chemical concentration in water
that is dissolved. To calculate fd, divide the dissolved
concentration by the total concentration.  A chemical that is
entirely in the dissolved phase has a fd of 1, while  a
chemical that is entirely in the particulate phase has a  fdof
0.
       Figure 3.1.1 presents the range
of concentrations measured for
dissolved metals in the bilgewater
samples. The plots show that dissolved
metals concentrations range over six
orders of magnitude. Calcium,
magnesium, potassium and sodium
were the dissolved metals measured at
the highest concentrations. As discussed
in Chapter 1 and Section 3.1.3, these
cations naturally occur in seawater and
their levels in the discharges are similar
to levels seen in ambient seawater. As
many discharges use ambient water for
onboard activities,  and spray would
contribute to other discharges, it was
not unexpected to find these levels of
cations in the bilgewater samples as
most vessels were sampled in coastal
areas. At these concentrations, these cations are generally not toxic to aquatic organisms, which
is why there there are no NRWQC for these metals, and therefore, no PHQs were calculated (see
Section 3.1.3 for additional explanation). Dissolved aluminum, barium, copper, manganese,
selenium13 and zinc were also measured  at relatively high concentrations (tens to hundreds of
  Note that for three of the six bilgewater samples where dissolved arsenic was detected, EPA suspects that the
measured arsenic concentrations are likely to be overestimated due to positive interference (see discussion on page
74)
13 Note: EPA suspects positive interference for all concentrations of total and dissolved selenium detected in
bilgewater samples (see discussion on page 74). Reported values could be entirely due to this positive interference.
                                              81

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

|ig/L) in most bilgewater samples; dissolved arsenic14 and iron were also measured at
concentrations greater than 100 |ig/L in individual samples. Among the vessels from which
bilgewater was sampled, a tow/salvage boat had the highest concentrations of the most dissolved
metals (seven analytes), while the water taxi had only one dissolved metal.

       Figure 3.1.2 shows the total metals concentrations in the bilgewater samples. The box
plots show that the relative ranges of total metals concentrations are comparable to the
concentrations of dissolved metals. Among the vessels from which bilgewater was sampled, the
shrimper had the highest concentrations of the most total metals (11), while the longliner and the
water taxi had the fewest (one each). In general, total concentrations for each metal are similar to
or slightly higher than the dissolved concentrations. To explore this relationship further, EPA
calculated the average dissolved fraction fd of each metal  in the bilgewater samples to better
understand the potential for aquatic organism impacts. The metals with the highest average
dissolved fractions (fd> 90 percent) included barium, calcium, magnesium, potassium, selenium
(see footnote 13), and zinc. Metals having intermediate average dissolved fractions (90 percent >
fd > 50 percent) included antimony,  arsenic (see footnotes 11 and 12), cadmium, chromium,
cobalt, copper, iron, manganese, and nickel. Aluminum, lead, and vanadium had the lowest
average dissolved fractions (fd< 50 percent).

       Figure 3.1.3 shows the distributions of PHQs based on the most conservative screening
benchmark for each of the dissolved metals. Per Section 3.1.3 above, points on this plot above
the dashed line (demarcating a PHQ of one) indicate a dissolved metal concentration exceeding
the benchmark; two of the dissolved metals (cadmium and copper) have PHQs that include
values greater than 10, indicating that the measured concentrations were one (or more) order of
magnitude greater than the  screening benchmark. EPA suspects that the high PHQs for a third
dissolved metal in this discharge (selenium) was elevated due to positive interference  due to the
major seawater cations in these samples. The highest PHQ (113) was for dissolved copper,
measured in the bilgewater sample from the tour boat. EPA also found PHQs exceeding one for
dissolved arsenic, chromium, lead, nickel, and zinc, bringing to eight the number of dissolved
metals that exceeded the most stringent 2006 NRWQC in one or more  bilgewater sample.
Dissolved copper concentrations, ranging from 6.6 to 350 |ig/L, exceeded the saltwater acute
(4.8 |ig/L) and chronic (3.1 |ig/L) criteria in all seven bilgewater samples; concentrations in all
but one bilgewater sample also exceeded the freshwater acute (13 |ig/L) and chronic (9 |ig/L)
criteria. The single elevated dissolved cadmium concentration (10 |ig/L) exceeded the freshwater
acute (2.0 |ig/L) and chronic (0.25 |ig/L) criteria, and the  saltwater chronic (8.8 |ig/L) criterion.
In addition, the highest dissolved arsenic concentration (230 |ig/L) exceeded the 36 jig/L
saltwater chronic  criterion.  For the other dissolved metals (chromium, lead, nickel, and zinc),
14 As noted in footnote 11, some dissolved arsenic samples may experience positive interference. Dissolved arsenic
samples not suspected of having positive interference include the measured dissolved arsenic concentration of 230
ug/L for the tow/salvage vessel, the dissolved arsenic concentration of 10 ug/L for the tour boat as well as the
relatively low dissolved arsenic concentration of 1.1 ug/L for a sample of bilgewater from a tow/salvage vessel.
                                            82

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

concentrations in one or more bilgewater samples exceeded saltwater and/or freshwater criteria,
although in each of these cases the PHQs were less than five.

       Three of the total metals (aluminum, arsenic15, and iron) exceeded the most stringent
2006 NRWQC16 in one or more bilgewater samples as shown in Figure 3.1.4. PHQs for total
arsenic (those not suspected of significant positive interference) ranged from 306 to 16,170. The
total arsenic concentrations in samples associated with these PHQs all greatly exceeded the
human health criterion for consumption of water plus organism of 0.018 |ig/L, as well as the
human health criterion for organism consumption alone, 0.14 |ig/L. PHQs for aluminum and iron
did not exceed 11. Five of the seven total aluminum concentrations measured in bilgewater (at
concentrations ranging from 332 to 940 |ig/L) exceeded the freshwater chronic criterion (87
|ig/L, expressed as total recoverable metal). For total iron, concentrations in two of three
bilgewater samples exceeded the human health criterion for water plus organism consumption of
300 |ig/L; PHQs for total iron ranged from 0.17 to 6.3.

       To further evaluate the significance of the dissolved and total metals concentrations in the
bilgewater samples, EPA compared them to ambient dissolved and total metal concentrations in
surface water samples collected near the vessels. This was done because surface water might
occasionally leak into certain vessel bilges, be used onboard the vessel, or splash onto the vessel
and drain into the bilge. In these cases, the concentrations of metals (as well as other analytes)
measured in the bilgewater samples might be similar to or significantly influenced by the
ambient concentrations. Indeed, EPA found that the concentrations of many of the metals
(including aluminum, barium, calcium, chromium, magnesium, manganese, nickel, potassium,
selenium, and sodium) measured in multiple bilgewater samples were no more than double the
ambient concentrations. The similarity in the concentrations of many of these metals in
bilgewater and ambient samples suggests that some proportion of the water sampled in the  vessel
bilges may be from ambient water. It is less clear whether the significant background ambient
metals concentrations in the sampled harbors reflect the loading from the cumulative discharges
of the many vessels that operate there, or loadings from other point and/or nonpoint pollutant
sources to these water bodies.

       On the other hand,  the highest concentrations of some of the dissolved and total metals
measured in bilgewater were substantially elevated above the corresponding ambient
concentrations. For dissolved copper, the ambient concentration that accompanied the highest
bilgewater concentration (350 jig/L from a tour boat) was below the detection limit. The next
two highest dissolved copper concentrations in bilgewater (119 and 120 |ig/L) were from water
taxis with a somewhat higher corresponding ambient concentration of 24 |ig/L.
15 Measured total arsenic concentration where no positive interference is evident include 291 ug/L for a tow/salvage
vessel, 19and5.5 ug/Lforthe tour boat and longliner fishing vessel, respectively, and 1.3 ug/L for a sample of
bilgewater from a tow/salvage vessel.
16 PHQs for total metals are based on NRWQC for human health and not aquatic life, as stated in Section 3.1.3.

                                            83

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       For dissolved aluminum, the ambient concentration that accompanied the highest
bilgewater concentration (520 |ig/L from the longliner) was 870 |ig/L; in this case, and several
others, even the highest concentration for a metal in bilgewater was exceeded by the ambient
concentration.

       The data for total metals also demonstrate considerable variability in the relationships
between bilgewater and ambient concentrations. The highest total arsenic concentration in
bilgewater (291 |ig/L from a tow/salvage boat) exceeded the corresponding ambient
concentration (12 |ig/L) by a considerable margin. The ambient concentration that accompanied
the next highest total arsenic concentration in bilgewater (32 |ig/L from the shrimper) was a
comparable 29 |ig/L, although this moderately high concentration of total arsenic measured in
the bilgewater sample from the shrimper is likely an overestimate due to positive interference
(see discussion on page 74).

       The results shown here illustrate that relationships between metals concentrations in
bilgewater and ambient samples are quite variable, even for the highest concentrations of metals
measured in bilgewater. EPA acknowledges that such variability could be due to type of
bilgewater production and dilution onboard. For example, a shrimper might have used a
substantial amount of ambient water for washdown as  compared to a tow boat, and thus, dilute
what might be a similar actual bilge sample absent the washdown. Clearly the potential for
metals in bilgewater discharges to pollute receiving waters may be overestimated if the ambient
metals concentrations and other considerations (type and dilution of bilgewater) are not
appropriately considered.

       In summary, metals were frequently detected in bilgewater samples. EPA found
relatively high concentrations of a number of dissolved and total metals in these samples. Total
arsenic and dissolved copper concentrations were significantly elevated above the most
conservative screening benchmarks in individual samples, with PHQ values from greater than 10
to over 1,000. Dissolved cadmium concentrations in a  single bilgewater sample also generated
PHQs in this range. For these and other metals (including total aluminum and iron and dissolved
chromium, lead, nickel, and zinc), concentrations measured in one or several bilgewater samples
exceeded saltwater and/or freshwater criteria. EPA found that  the concentrations of many of the
metals measured in bilgewater samples (except for dissolved copper and total arsenic) were
comparable to the ambient receiving water concentrations.
                                           84

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.1. Results of Bilgewater Sample Analyses for Dissolved Metals1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum
Antimony
Arsenic3
Barium
Cadmium
Chromium
Cobalt
Copper
Iron
Lead
Manganese
Nickel
Selenium4
Vanadium
Zinc
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
7
5
7
5
7
7
5
7
5
7
7
7
7
5
7
6
1
6
5
1
5
2
7
1
3
7
6
4
1
7
86
20
86
100
14
71
40
100
20
43
100
86
57
20
100
150
0.66
41
49
1.9
12
1.0
100
75
2.3
34
9.2
24
0.62
130
37

10
43

1.6

56


28
8.8
30

100



38



6.6


3.9



53



38



6.6


3.9



53
9.7

1.1
39



25


13
4.7


72
420
0.65
21
62

17
1.8
120
87
4.2
50
14
36
0.55
190
520
1.3
230
64
10
56
2.5
350
170
7.2
79
15
57
1.1
250
520
1.3
230
64
10
56
2.5
350
170
7.2
79
15
57
1.1
250
NA
NA
36
NA
0.25
11
NA
3.1
NA
2.5
NA
8.2
5
NA
81
Cationic Metals
Calcium
Magnesium
Potassium
Sodium
ug/L
ug/L
ug/L
ug/L
7
7
5
5
7
7
5
5
100
100
100
100
76000
1 80000
67000
1 400000
76000
1 80000
65000
1 400000
33000
8300
9800
1 20000
33000
8300
9800
120000
47000
14000
37000
730000
100000
31 0000
98000
2000000
140000
420000
120000
2700000
1 40000
420000
1 20000
2700000
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) See footnotes 11 and 12.
(4) See footnote 13.
                                                                             85

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.2. Results of Bilgewater Sample Analyses for Total Metals1
Analyte
Units
No.
samples
No.
detected
Detected
proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum
Antimony
Arsenic3
Barium
Cadmium
Chromium
Cobalt
Copper
Iron
Lead
Manganese
Nickel
Selenium4
Vanadium
Zinc
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
7
5
7
5
7
7
5
7
5
7
7
7
7
5
7
7
1
7
5
1
6
1
7
3
6
7
6
4
2
7
100
20
100
100
14
86
20
100
60
86
100
86
57
40
100
370
1.3
53
50
2.6
25
1.3
150
520
9.6
53
12
25
2.6
160
330

12
44

3.5

130
250
7.5
52
9.4
25

87
26

1.3
38



8.5


7.4



56
26

1.3
38



8.5


7.4



56
28

5.5
38

2

50

2.3
37
6.2


72
640
0.65
32
66

37
0.7
210
1100
18
79
17
38
1.4
260
940
1.3
290
67
12
96
1.4
430
1900
26
97
24
66
1.7
360
940
1.3
290
67
12
96
1.4
430
1900
26
97
24
66
1.7
360
87
5.6
0.018
1000
NA
NA
NA
1300
300
NA
100
610
170
NA
7400
Cationic Metals
Calcium
Magnesium
Potassium
Sodium
ug/L
ug/L
ug/L
ug/L
7
7
5
5
7
7
5
5
100
100
100
100
76000
1 80000
68000
1 400000
77000
1 80000
65000
1 400000
36000
9200
9600
1 20000
36000
9200
9600
1 20000
47000
14000
37000
740000
1 1 0000
31 0000
100000
2000000
1 30000
390000
130000
2600000
1 30000
390000
130000
2600000
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3,  and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) See footnotes 11 and 12.
(4) See footnote 13.
                                                                             86

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
             100
        o
       "co
        CD
        O

        O
       O
10
                          i    i   r
                                                     ©
                                      ©
                                             T
        I   T   I   I
T   I
                                                        OOP
                                                                       ©
                                                                       T
                                                                          ./xO
                                    Dissolved Metals
Figure 3.1.1. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
Samples of Bilgewater
(Note: As discussed in footnotes 12 and 13, all but possibly one of the bilgewater samples analyzed for dissolved
selenium and three of the bilgewater samples analyzed for dissolved arsenic may be elevated due to positive
interference. The measured dissolved arsenic concentration of 230 ug/L for the tow/salvage vessel and measured
dissolved arsenic concentration of 10 ug/L for the tour boat are not expected to have had positive interference).
                                              87

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
        1000
    O)
    ^
    c
    o
   "co
    CD
    o
    c
    o
   O
100
  10
1
— Q
- (

-

-
-
1
-Q-


O
T
'

1 1 1 1
i




0












fife
*r
Q











c

)
o


^
-i


n ii
CtQ±>
- olo
-

^™
I
-
I
t

—
I
: | '







X
^


r












^
0
-e
o

c


0

9




o


£
1 1 ^




Q (
i^t





-&
^







i
©

3-
r i J







iC
-0-


r















61
*










£










i-.
3-


_£











i
1






9



<


o


HI

0
olo



1
1 1 1 1 1
_
f -
1 -
9
$ oeo E
^^Aolj>$^>vc
                                 Total Metals
Figure 3.1.2. Box and Dot Density Plot of Total Metals Concentrations Measured in
Samples of Bilgewater
(Note: As discussed in footnotes 11 and 13, all but one of the bilgewater samples analyzed for total selenium and
three of the bilgewater samples analyzed for total arsenic may be elevated due to probability of positive interference.
Exceptions are the total arsenic concentration of 291 ug/L for the tow/salvage vessel, concentrations of 19 and 5.5
ug/L for the tour boat and longliner fishing vessel, respectively, as well as the concentration of 1.3 ug/L for a
sample of bilgewater from a tow/salvage vessel).
                                        88

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
          100.0-
     CD
    "-I— •

    I
    -a
     CO
     N
     CD
    E
     CD
     o
    Q_
10.0-
                                  Dissolved Metals
Figure 3.1.3. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals
in Samples of Bilgewater
(Note: Values circled here and throughout the rest of this chapter indicate PHQs calculated based on replacement
values for non-detects. Non-detect (censored) concentrations were replaced with !/2 of the reporting limit for use in
these plots. Also, as discussed in footnotes 11 and 13, all but one of the bilgewater samples analyzed for total
selenium and three of the bilgewater samples analyzed for total arsenic may be elevated due to probability of
positive interference).
                                              89

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
10000.00
| 1000.00
"o
o 100.00
•a
N 10.00
05
05 1 nn
~ I .UU
CD
i -
§_ 0.10

0.01
I I 1 I I I I I I I I III
9
— o —
I cj> I
= s =
-
f& = ~\
••o m op
~ ¥ T ob ~
f c& ? o*1
: ® A<;¥^ ft "
r 	 * i i i ¥^ i f i
                                       Total  Metals
Figure 3.1.4. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
Samples of Bilgewater
(Note: Replacement values for non-detects are circled. Also, as discussed in footnotes 11 and 13, all but one of the
bilgewater samples analyzed for total selenium and three of the bilgewater samples analyzed for total arsenic may be
elevated due to probability of positive interference).
                                               90

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.1.2    Classical Pollutants

       Bilgewater samples were analyzed for 14 classical pollutants (see Table 3.1.3). These
pollutants include measurements that are qualitatively quite different: physical properties (pH,
temperature, conductivity, salinity, turbidity, TOC, TSS), oxygen consumption (BOD and COD),
oil and grease (hexane extractable material (HEM) and silica-gel treated hexane extractable
material (SGT-HEM)), as well as concentrations of several chemicals (sulfide, DO, TOC and
TRC).17 Figure 3.1.5 illustrates the variability of the concentrations/values measured for the
classical pollutant in bilgewater. The highest concentrations of BOD, COD and TOC (770, 2970,
and 732 mg/L, respectively), as well as HEM, SGT-HEM, and TRC, were measured in a single
bilgewater sample from a tow/salvage boat. BOD and TOC concentrations were highly variable
among the bilgewater samples, ranging from 2 to 770 mg/L for BOD and from 9 to 730 mg/L for
TOC.

       Oil and grease were measured as HEM and petroleum hydrocarbons were measured as
SGT-HEM. HEM and SGT-HEM were detected in all of the bilgewater samples, with
concentrations ranging from 1.1 to 43.6 mg/L (HEM) and 1.1  to 18.2 mg/L (SGT-HEM). These
concentrations were compared to the existing international and U.S. regulatory limit of 15 mg/L
of oil and grease that can be discharged from a moving ship when within 12 nautical miles from
land18. Some type of oil collector (sorbent pad, rags,  etc.) was used on four of the seven vessels
sampled for bilgewater. A single value taken from the tow/salvage boat exceeded the 15-mg/L
benchmark by threefold. Oil and grease discharges at this concentration are significant enough to
cause a visible sheen. The tow/salvage boat had no equipment or management practices in place
to remove oil or other pollutants prior to overboard discharge  of bilgewater.

       Sulfide was detected in two bilgewater samples, at concentrations of 0.015  and 0.2 mg/L.
These concentrations exceeded the NRWQC of 2 ug/L (0.002 mg/L) by factors of 7.5 to 100.
Sulfide (hydrogen sulfide) is a pollutant that is commonly elevated in water distribution systems
as well as sewers. Sulfur-reducing bacteria, which use sulfur as an energy source, are believed to
be the primary producers of large quantities of hydrogen sulfide in bilgewater. Ecologically,
these bacteria are common in anaerobic environments (e.g., plumbing systems). Sulfur-reducing
bacteria are apparently present in at least some of the vessels,  because sulfide was  not detected in
the ambient water sampled at the vessel locations.

       Figure 3.1.6 presents box and dot density plots of the PHQs for classical pollutants.
PHQs were calculated for the six classical pollutants for which benchmarks were available. As
this figure shows, all of the detected TRC concentrations exceeded the saltwater chronic
17 See Section 3.1.1 this chapter for the rationale to use this term for this large group of conventional,
nonconventional, and other physico-chemical factors.
18 International Convention for the Prevention of Pollution from Ships, 1973, as modified by the Protocol of 1978
relating thereto (MAPJ>OL).

                                           91

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

NRWQC benchmark of 0.0075 mg/L and yielded PHQs ranged from 6.7 to 21. The highest TRC
concentration (0.16 mg/L) was measured in a bilgewater sample collected from a tour boat.

       EPA compared classical pollutant concentrations in the bilgewater samples to ambient
concentrations in surface water samples collected near the vessels. Concentrations of a number
of the classical parameters (including conductivity, pH, salinity, temperature, and (to a varying
degree) turbidity in bilgewater were comparable with ambient water. This was expected,
considering the likelihood of ambient water leaking into vessel bilges. The concentration of DO
measured in one bilgewater sample (1.8 mg/L in the longliner) was hypoxic (<2 mg/L), although
the ambient DO value at this location (Sitka, Alaska) was also very low (1.0 mg/L). TRC
concentrations were elevated at 0.1 mg/L in two of the seven bilge samples; for the remaining
samples, TRC concentrations were comparable between bilgewater and ambient samples. For the
remaining classical pollutants (BOD, COD, HEM, SGT-HEM, sulfide, TOC, and TSS) the
concentrations measured in bilgewater greatly exceeded those measured in ambient samples.
BOD concentrations in three of the bilgewater samples (189, 325, and 770 mg/L) were high
enough to be comparable to values typical of raw domestic sewage (110 to 400 mg/L; Metcalf
and Eddy, 1979). These three bilgewater samples  also exceed EPA's secondary treatment
effluent limit of 30 mg/L for BOD. COD concentrations in four of the bilgewater samples (430,
546, 780, and 2,970 mg/L) were again high enough to compare with values for raw domestic
sewage (250 to 1,000 mg/L; Metcalf and Eddy, 1979). These high levels of BOD and COD in
bilgewater discharges could potentially cause stress on a water body (e.g., where there are many
sources of oxygen demand, where there may be limited circulation or flushing, or where the
water body is under existing hypoxic or anoxic stress). Although TSS concentrations in
bilgewater were not as high as values for raw sewage, four of the bilgewater samples exceeded
the 30 mg/L effluent limit for TSS by factors ranging from 1.2 to 3. EPA realizes that these
effluent limits are based upon the high removal efficiencies for BOD and TSS that are achievable
by land-based sewage treatment plants, and may be overly conservative as benchmarks for vessel
discharge. However, as discussed in Section 3.1.3, the benchmarks are  still useful in a screening
level analysis as a starting point for evaluating the potential of these pollutants to cause or
contribute to ecological stress on a water body.
                                          92

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.3. Results of Bilgewater Sample Analyses for Classical Pollutants1
Analyte
Biochemical Oxygen Demand (BOD)
Chemical Oxygen Demand (COD)
Conductivity
Dissolved Oxygen
Hexane Extractable Material (HEM)
PH
Salinity
Silica Gel Treated HEM (SGT-HEM)
Sulfide
Temperature
Total Organic Carbon (TOC)
Total Residual Chlorine
Total Suspended Solids (TSS)
Turbidity
Units
mg/L
mg/L
mS/cm
mg/L
mg/L
SU
ppt
mg/L
mg/L
C
mg/L
mg/L
mg/L
NTU
No.
samples
7
7
6
6
7
7
6
7
7
7
5
7
7
7
No.
detected
7
7
6
6
7
7
6
7
2
7
5
3
7
7
Detected
Proportion
(%)
100
100
100
100
100
100
100
100
29
100
100
43
100
100
Average
Cone.
190
740
5.0
5.3
9.3
7.2
5.5
4.4
0.034
20
200
0.077
39
41
Median
Cone.
14
430
6.9
5.5
5.2
7.0
4.5
2.4

21
110

38
20
Minimum
Cone.
2.0
91
0.017
1.8
1.1
6.9
0.40
1.1

9.0
8.9

3.7
3.5
10%
2.0
91
0.017
1.8
1.1
6.9
0.40
1.1

9.0
8.9

3.7
3.5
25%
4.1
98
0.56
3.4
1.2
6.9
3.1
1.2

14
16

5.5
5.2
75%
330
780
9.3
6.9
7.0
7.3
8.9
3.5
0.015
27
440
0.13
71
41
90%
770
3000
14
11
44
8.0
13
18
0.20
28
730
0.16
88
160
Maximum
Cone.
770
3000
14
11
44
8.0
13
18
0.20
28
730
0.16
88
160
Screening
BM2
30
NA
NA
NA
15
NA
NA
15
0.0020
NA
NA
0.0075
30
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                             93

-------
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
1000.00
•1— •
g 100.00
E
c<
^•v
o 10.00
g
1 1.00
CD
0
o 0.10
^•X
0.01
_ i i i i i i i i i i i i i _
- ° o -
" ° T fL ?
°» no -
1 fn ~
; ' P3 CpO J^l T T^ -
* ol lol
QD QD
: O -
O
r fl i
* 0
=111111 InfUtal 11111=
               0^ 0^\r^ CX^ C^ C^)<^ C^ CX^ C^ 0^ VP\^N^
             v^^o^^^^V^^^^-\$^
             -ov \v ^^^\^^\v^oH^x^\®};s^'  v
             I  ^^^     \^^»v*^   ^^  ~^^  \   \    »     ^^\^



                        Classical Pollutants
Figure 3.1.5. Box and Dot Density Plot of Classical Pollutant Concentrations

Measured in Samples of Bilgewater
                                       94

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
IUU.UU
fl\
^J^J
§ 10.00
0
•a
05
N
05
lo 1-00
-1— >
CD
-i— •
O
Q_
0.10
-
-
—
I
-
E
_
—

c
c
c



-£
C
£
(

)
)



3-
)_

I I I I I I T I I I I I I :
-
o
n -
, T
rl
^
C)

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


       3.2.1.3    Pathogen Indicators (Microbiological)

       Bilgewater samples19 from two commercial fishing vessels were analyzed for the
pathogen indicator bacteria E. coli, enterococci, and fecal coliform (commercial fishing vessels
only) (see Table 3.1.4). E.  coli and  enterococci were detected in a bilgewater sample collected
from a shrimping vessel, and fecal coliform were detected in bilgewater from two fishing vessels
(a longliner and the  shrimper).

       The NRWQC for pathogen  indicators references the bacteria standards in EPA's 1986
Quality Criteria for Water., commonly known as the Gold Book. NRWQC standards for bacteria
are described in terms of three different water body use criteria: freshwater bathing, marine water
bathing, and shellfish harvesting waters.

       For each of the pathogen indicators, the lowest NRWQC was  exceeded in one of the
bilgewater samples.  The E. coli value (393 MPN/100 mL) exceeds the freshwater bathing
NRWQC of 126 MPN/100 mL. The enterococci value (4,100 MPN/100 mL) exceeds the bathing
NRWQCs of 33  CFU/100  mL for fresh  water and 35 CFU/100 mL for salt water. One of the two
fecal coliform values (118 CFU/100 mL) exceeds the NRWQC of 14 MPN/100 mL for shellfish
harvesting20.

       Values of the pathogen indicators measured in these bilgewater samples  exceed the
values measured in nearby ambient surface water samples by factors ranging from 4 (for
enterococci) to 15 (E.  coli), suggesting that leakage or other entry of ambient water is not a
significant source of these pathogen indicators in bilgewater. EPA is unsure as to the source of
pathogen indicators  in bilgewater.
19 Logistics prevented EPA from delivering all bilgewater samples to laboratories within allowable holding times.
20 MPN is most probable number and approximates the unit of measure for fecal coliform in this study of CPU
(colony forming units).

                                           96

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.4. Results of Bilgewater Sample Analyses for Pathogen Indicators1
Analyte
£. Co//
Enterococci
Fecal Coliform
Units2
MPN/100ml
MPN/100ml
CFU/100ml
No.
Samples
1
1
2
No.
detected
1
1
2
Detected
Proportion (%)
100
100
100
Average
Cone.
390
4100
61
Median
Cone.


120
Minimum
Cone.


4.0
10%


4.0
25%


4.0
75%


120
90%


120
Maximum
Cone.


120
Screening
BM3
130
33
14
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) MPN =  Most Probable Number; CFU = Colony Forming Units.
(3) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            97

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.1.4    Nutrients

       Bilgewater samples were analyzed for four nutrient-related parameters: ammonia
nitrogen, nitrate/nitrite, total Kjeldahl nitrogen (TKN), and total phosphorus (see Table 3.1.5).
The box and dot density plots in Figure 3.1.7 illustrate the variability of the nutrient
concentrations measured in bilgewater. Ammonia, TKN and total phosphorus concentrations
were elevated in a single bilgewater sample collected from a longliner fishing vessel. The
elevated nutrient concentrations may be attributable to seepage from the/ice slurry in the fish
hold of the longliner. Water containing biological material (e.g., fish waste tissues, excreta)
might seep down into the bilge compartment, resulting in an increase in nutrient discharge.

       Ammonia is the only nutrient for which there are currently numeric NRWQC. EPA
established these numeric criteria based on chronic toxicity to aquatic life, not nutrient
enrichment. An ammonia-nitrogen concentration of 7.6 mg/L, measured in the bilgewater sample
from the longliner fishing vessel,  exceeded the NRWQC chronic criteria in both salt water (1.2
mg/L) and fresh water (1.24 mg/L). Three of the five bilgewater samples for total phosphorous
exceeded EPA's 0.1 mg/L 1986 Gold Book criterion. The highest total phosphorus
concentration, 13 mg/L, exceeded the benchmark by a factor of 130. Figure 3.1.8 presents box
and dot density plots of the PHQs calculated for the nutrient data.

       EPA compared nutrient concentrations in the bilgewater samples to ambient
concentrations in surface water samples collected near the vessels. Ammonia was detected in one
of the ambient samples at a concentration of 0.11 mg/L, comparable (within a factor of two) to
the concentration in the corresponding bilgewater sample, 0.13 mg/L. TKN was detected in three
ambient samples; in one, the ambient concentration of 0.60 mg/L marginally exceeded the
bilgewater concentration of 0.55 mg/L. However, ambient TKN concentrations were less than
the bilgewater concentrations in the other two cases. For total phosphorus, the comparison
showed the concentrations detected in two ambient samples were comparable to the
corresponding bilgewater concentrations; however, total phosphorus was not detected in the
ambient samples corresponding to the three bilgewater samples having the highest total
phosphorus concentrations. Thus, although ambient nutrient concentrations appear to be
comparable to the lower concentrations of nutrients in bilgewater and may be a partial source of
these nutrients in some samples, they  cannot explain the sources of the higher nutrient
concentrations measured in other samples.
                                           98

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.5. Results of Bilgewater Sample Analyses for Nutrients1
Analyte
Ammonia As Nitrogen (NH3-N)
Nitrate/Nitrite (NO3/NO2-N)
Total Kjeldahl Nitrogen (TKN)
Total Phosphorus
Units
mg/L
mg/L
mg/L
mg/L
No.
samples
5
7
5
5
No.
detected
4
5
5
5
Detected
Proportion
(%)
80
71
100
100
Average
Cone.
1.7
0.38
16
3.0
Median
Cone.
0.24
0.18
2.5
0.47
Minimum
Cone.


0.55
0.084
10%


0.55
0.084
25%
0.064

1.0
0.093
75%
4.0
0.36
39
7.1
90%
7.6
1.9
73
13
Maximum
Cone.
7.6
1.9
73
13
Screening
BM2
1.2
NA
NA
0.10
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3,  and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                             99

-------
       10.00
  D5

  E
        1.00
  03
 0)
 o
 c
 o
O
        0.01
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                                  oo
                                  OO
                                             ©
                                                       ©
                                                      Qt)
                                  Nutrients
Figure 3.1.7. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of

Bilgewater
                                       100

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
    0

    "o
    ID

    O
    03
    03


    ~C
        100.00
10.00
             .00
            0.10
            0.01
                               ^v


                                    *
                                               i         i
                                                        Q
                                             on	=
                                    Nutrients
Figure 3.1.8. Box and Dot Density Plot of Potential Hazard Quotients for Nutrients in

Samples of Bilgewater

(Note: Replacement values for non-detects are circled).
                                        101

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.1.5    Semivolatile Organic Compounds (SVOCs)

       Bilgewater samples were analyzed for 79 SVOCs. Out of the 79 analytes, 56 were not
detected in any of the bilgewater samples. Of the remaining 23 SVOCs, 18 were only detected in
a single bilgewater sample and five were found in multiple samples (see Table 3.1.6). Of these,
bis(2-ethylhexyl) phthalate was detected in more than 50 percent of the samples. This SVOC is a
manufactured chemical that is commonly added to plastics to make them flexible and can be
found in a variety of products used on vessels such as hoses, tubing, and gaskets. Di-n-butyl
phthalate, di-n-octyl phthalate, naphthalene, and phenanthrene were also detected in more than
one bilgewater sample. There was no obvious trend in the occurrence of SVOCs based on the
type of vessel sampled.

       Figure  3.1.9 presents the range of concentrations measured for SVOCs in the bilgewater
samples.  Concentrations of five SVOCs (2-butoxy ethanol, 2- methyl-naphthalene, dimethyl
phthalate, indole, and naphthalene) exceeded 100 |ig/L in single (but not the same) bilgewater
samples.  It was difficult for EPA to compare the concentration distributions between SVOCs
because the majority were detected in a  single sample. Bis(2-ethylhexyl) phthalate and
phenanthrene concentrations ranged  over nearly two orders of magnitude.

       The distributions of PHQs, based on the most conservative screening benchmarks, are
displayed for each SVOC in Figure 3.1.10. PHQs for two SVOCs, 2,4,6-trichlorophenol and
bis(2-ethylhexyl) phthalate,  exceeded the screening threshold of one. The 2,4,6-trichlorophenol
concentration (24 |ig/L) measured in a single bilgewater sample from a tour boat exceeded the
1.4 |ig/L  human health (water and organism consumption) criterion by a factor of 1721. Bis(2-
ethylhexyl) phthalate was detected in four of the seven bilgewater samples, at concentrations that
exceeded the 1.2 |ig/L human health (water and organism consumption) criterion by factors that
ranged from 1.1 to 59. As shown in Figure 3.1.10, the PHQs for four other SVOCs were orders
of magnitude less than 1, and therefore,  likely pose little risk as pollutants  from bilgewater
discharges.

       SVOCs were detected in two ambient samples, and for these chemicals (bis(2-ethylhexyl)
phthalate and Di-n-butyl phthalate) the ambient concentrations were only comparable to the
lowest concentrations  measured in bilgewater.
21 Because of elevated reporting limits for this SVOC in several samples, replacement values for the nondetected
concentrations exceed the benchmark (e.g., PHQ >1). However, these values were not based on measured
concentrations and are therefore uncertain.
                                           102

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.6. Results of Bilgewater Sample Analyses for SVOCs1
Analyte
2,4,6-Trichlorophenol
2-Butoxy ethanol
2-Methylnaphthalene
3-Methyl-butanoic acid
4-Methyl-pentanoic acid
Benzeneacetic acid
Benzenepropanoic acid
Benzothiazole
Bis(2-ethylhexyl) phthalate
Cholesterol
Dimethyl phthalate
Di-n-butyl phthalate
Di-n-octyl phthalate
Heptadecane
Indole
Naphthalene
n-Hexadecane
Nonadecane
p-Cresol
Phenanthrene
Phenol
Pyrene
Triethyl Phosphate
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
samples
7
1
5
1
1
1
1
1
7
1
7
7
7
1
1
7
1
1
5
7
7
7
1
No.
detected
1
1
1
1
1
1
1
1
4
1
1
2
2
1
1
3
1
1
1
2
1
1
1
Detected
Proportion
(%)
14
100
20
100
100
100
100
100
57
100
14
29
29
100
100
43
100
100
20
29
14
14
100
Average
Cone.
7.0
260
39
57
38
29
32
45
15
88
24
4.0
4.1
56
160
100
39
49
7.7
12
18
6.8
20
Median
Cone.








1.4














Minimum
Cone.























10%























25%























75%


88





21


1.4
3.1


2.3


8.7
1.3



90%
24

180





71

140
4.9
3.5


700


17
69
100
34

Maximum
Cone.
24

180





71

140
4.9
3.5


700


17
69
100
34

Screening
BM2
1.4
NA
NA
NA
NA
NA
NA
NA
1.2
NA
270000
2000
NA
NA
NA
NA
NA
NA
NA
NA
21000
830
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3,  and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                             103

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     O)
     ZJ
     c
     o
    "CD
     CD
     o
     c
     o
    O
          100
10
             1
                        i     r
                                         m
                           1    I     I    TT
             I	I
I	I
                                                          I
               0   2   4   6   8   10  12  14  16  18  20  22 24
                                      SVOCs
Figure 3.1.9. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
Bilgewater
SVOCs are identified as follows
(1) 2,4,6-Trichlorophenol
(2) 2-Butoxy Ethanol
(3) 2-Methylnaphthalene
(4) 3-Methyl-Butanoic Acid
(5) 4-Methyl-Pentanoic Acid
(6) Benzeneacetic Acid
(7) Benzenepropanoic Acid
(8) Benzothiazole
               (replacement values for non-detects are circled):
                      (9) Bis(2-Ethylhexyl) Phthalate
                      (10) Cholesterol
                      (11) Dimethyl Phthalate
                      (12) Di-N-Butyl Phthalate
                      (13) Di-N-Octyl Phthalate
                      (14) Heptadecane
                      (15) Indole
                      (16) Naphthalene
                             (17) N-Hexadecane
                             (18) Nonadecane
                             (19) P-Cresol
                             (20) Phenanthrene
                             (21) Phenol
                             (22) Pyrene
                             (23) Triethyl Phosphate
                                              104

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 
-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


       3.2.1.6    Volatile Organic Compounds (VOCs)

       Bilgewater samples were analyzed for 72 VOCs. Out of the 72 analytes, 46 VOCs were
not detected in any of the bilgewater samples. Of the remaining 26 VOCs, 11 were detected in
more than one bilgewater samples and 15 were detected only in one bilgewater sample (see
Table 3.1.7). Of the 11 VOCs that were detected in more than one bilgewater sample, the
following were detected in more than 50 percent of the samples:

       •   1,2,4-Trimethylbenzene
       •   1,3,5-Trimethylbenzene
       •   Acetone
       •   Benzene
       •   m-,p-Xylene (sum of isomers)
       •   Methylene chloride
       •   O-Xylene.

       2-butanone, ethylbenzene, styrene, and toluene were also detected in more than one
bilgewater sample.

       Figure 3.1.11 presents the range of concentrations measured for VOCs in the bilgewater
samples. The  VOC concentrations measured in bilgewater samples varied widely, with
concentrations of a half-dozen VOCs ranging over three orders of magnitude. The maximum
concentrations of four VOCs (1,2,4-trimethylbenzene, m-,p-xylene, o-xylene and toluene)
exceeded 1,000 |ig/L (1 mg/L), while the maximum concentrations of four other VOCs (1,3,5-
trimethylbenzene, benzene, ethylbenzene and n-propylbenzene) exceeded 100 |ig/L. Each of
these maximum VOC concentrations was measured in the bilgewater sampled from one
tow/salvage boat. These VOCs are commonly constituents of petroleum products, refining by-
products, and gasoline  additives, and are used as solvents.

       Figure 3.1.12 presents the distributions of PHQs for each VOC, based on the most
conservative screening benchmarks. The maximum PHQ for benzene, based on the 2.2 |ig/L
human health (water plus organism consumption) criterion benchmark, was 187. The maximum
PHQ for toluene was marginally higher than one; the highest concentration of toluene (1,700
|ig/L) exceeded the human health (water and organism consumption) criterion of 1,300 |ig/L. For
two other VOCs (chloroform and tetrachloroethene), only one of seven sample concentrations
were detected, and these detected concentrations were below the  screening benchmark. However,
because the method detection limits for these two compounds were more than double their
respective screening benchmarks, the resulting PHQs for these compounds, as reported in Figure
3.1.12, are greater than one when concentrations equal to !/2 of the detection limit are included.
Because these PHQ values were not based on detected concentrations, EPA considers them
highly uncertain.

                                         106

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       Finally, two VOCs (acetone and methylene chloride) were measured in ambient samples
at concentrations comparable to the corresponding bilgewater concentration. However, these
ambient concentrations were only comparable to the lowest concentrations of these VOCs
measured in some bilgewater samples. Therefore, it is unlikely that leakage or other entry of
ambient water is a significant source of the elevated acetone and methylene chloride
concentrations measured in bilgewater.
                                           107

-------
                                                                          Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.7. Results of Bilgewater Sample Analyses for VOCs1
Analyte
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
2-Butanone
4-lsopropyltoluene
Acetone
Benzene
Biphenyl
Carbon disulfide
Chloroform
cis-1 ,2-Dichloroethene
Cyclohexane
Ethylbenzene
Isopropylbenzene
m-,p-Xylene (sum of isomers)
Methyl tertiary butyl ether (MTBE)
Methylcyclohexane
Methylene chloride
Nonanal
n-Pentadecane
n-Propylbenzene
O-Xylene
Styrene
Tetrachloroethene
Toluene
Trichloroethene
Trichlorofluoromethane
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
samples
5
5
5
5
5
7
5
5
7
5
5
7
5
5
5
5
7
1
1
5
5
5
7
7
7
7
No.
detected
3
3
2
1
5
4
1
1
1
1
1
3
1
3
1
1
4
1
1
1
3
2
1
3
1
1
Detected
Proportion
(%)
60
60
40
20
100
57
20
20
14
20
20
43
20
60
20
20
57
100
100
20
60
40
14
43
14
14
Average
Cone. 3
220
65
2.6
3.3
13
61
4.5
2.0
3.3
2.3
5.8
68
9.9
370
2.0
5.4
1.6
3.1
58
26
240
11
2.6
240
2.6
3.5
Median
Cone.
0.50
0.10


10
0.10







0.50


0.10



0.20





Minimum
Cone.




2.3





















10%




2.3





















25%




4.3





















75%
540
160
2.8
3.3
23
1.3
0.87
0.050

0.75
9.5
1.3
20
930
0.050
8.5
0.20


60
590
20

0.30


90%
1100
320
3.7
6.5
31
410
1.7
0.10
4.1
1.5
19
460
40
1900
0.10
17
0.30


120
1200
39
0.40
1700
0.30
5.5
Maximum
Cone. 3
1100
320
3.7
6.5
31
410
1.7
0.10
4.1
1.5
19
460
40
1900
0.10
17
0.30


120
1200
39
0.40
1700
0.30
5.5
Screening
BM2
NA
NA
NA
NA
NA
2.2
NA
NA
5.7
NA
NA
530
NA
NA
NA
NA
4.6
NA
NA
NA
NA
NA
0.69
1300
2.5
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3,  and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) In some cases, the detected concentration(s) for an analyte could be lower than the replacement value (% of the reporting limit) for a concentration that was nondetected. In an
extreme (but possible) case, this could result in an average concentration for an analyte that is greater than the maximum detected concentration.
                                                                              108

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
         1000.0
                   _   0
    CO
    Z5
    C
    o
    "03
    CD
    O
    c
    o
    O
           100.0
            10.0
1.0
              0.1
                          i    r
                                            i    i
                   -*
       j	i
                                              I    I  I
I    I    I
                   02468  10121416182022242628
                                          VOCs
Figure 3.1.11. Box and Dot Density Plot of VOC Concentrations Measured in Samples of
Bilgewater
VOCs are identified as follows:
(1) 1,2,4-Trimethylbenzene
(2) 1,3,5-Trimethylbenzene
(3) 2-Butanone
(4) 4-Isopropyltoluene
(5) Acetone
(6) Benzene
(7) Biphenyl
(8) Carbon Disulfide
(9) Chloroform
                   (10) Cis-l,2-Dichloroethene
                   (11) Cyclohexane
                   (12) Ethylbenzene
                   (13) Isopropylbenzene
                   (14) M-,P-Xylene (sum of
                   isomers)
                   (15) Methyl Tertiary Butyl Ether
                   (Mtbe)
                   (16) Methylcyclohexane
         (17) Methylene Chloride
         (18) Nonanal
         (19) N-Pentadecane
         (20) N-Propylbenzene
         (21)O-Xylene
         (22) Styrene
         (23) Tetrachloroethene
         (24) Toluene
         (25) Trichloroethene
         (26) Trichlorofluoromethane
                                             109

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
      c
      Q)
     "o
      3
     O
      03
      N
      03
     X

     15
     "c
     JD
     "o
     Q_
100.0000

  10.0000

   1.0000

   0.1000

   0.0100

   0.0010
             0.0001
                         i    i   t   i    i    i    i   i    i    i    i   r
                      02468  101214161820222426
                                           VOCs
Figure 3.1.12. Box and Dot Density Plot of Potential Hazard Quotients for VOCs in
Samples of Bilgewater
VOCs are identified as follows (replacement values for non-detects are circled):
(1) 1,2,4-Trimethylbenzene            (10) Cis-l,2-Dichloroethene
(2) 1,3,5-Trimethylbenzene            (11) Cyclohexane
(3) 2-Butanone                     (12) Ethylbenzene
(4) 4-Isopropyltoluene               (13) Isopropylbenzene
(5) Acetone                        (14) M-,P-Xylene
(6) Benzene                        (sum of isomers)
(7) Biphenyl                       (15) Methyl Tertiary Butyl Ether
(8) Carbon Disulfide                 (Mtbe)
(9) Chloroform                     (16) Methylcyclohexane
                                                        (17) Methylene Chloride
                                                        (18) Nonanal
                                                        (19) N-Pentadecane
                                                        (20) N-Propylbenzene
                                                        (21)O-Xylene
                                                        (22) Styrene
                                                        (23) Tetrachloroethene
                                                        (24) Toluene
                                                        (25) Trichloroethene
                                                        (26) Trichlorofluoromethane
                                            110

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.1.7    Nonylphenols

       Bilgewater samples were analyzed for 34 long- and short-chain nonylphenol and
octylphenol ethoxylates (two discrete subsets of alkylphenol ethoxylates), as well as total
nonylphenol.  Of these analytes, 14 alkylphenol ethoxylates were not detected and 20 were
detected in a single bilgewater sample (see Table 3.1.8). Of the 20 distinct alkylphenol
ethoxylates detected, 16 were detected in the bilgewater from a tour boat, three were detected in
the bilgewater from a tow/salvage boat, and one was detected in the bilgewater from a shrimper.
Measured concentrations of ethoxylates in bilgewater ranged from less than 1 |ig/L for three of
the octylphenol ethoxylate isomers (OP10EO, OP12EO, and OP11EO) to 1,050 |ig/L for total
nonylphenol polyethoxylates (sum of NPEOs - NP5EO through NP18EO). This latter maximum
concentration was measured in the bilgewater sample from the tour boat. According to the
operator(s), the bilgewater discharged from this vessel is expected to possibly contain oil, grease,
fuel, cleaning solvents, detergent and water from deck washdown. Of these pollutants, detergents
are the most common source of NPEOs. Although there is no NRWQC for the sum of
alkylphenol ethoxylates, they can degrade to total nonylphenol, which does have a NRWQC, in
fresh and salt water.

       The one detected concentration for total nonylphenol (NP, representative of the same
nonylphenol isomers in the commercial mixture upon which EPA's NRWQC is based - CAS
#84852-15-3) of 4.9 |ig/L exceeded the saltwater chronic criterion of 1.7 |ig/L by a factor of 2.9.
Although the vessel operators added dish soap to the bilgewater prior to overboard discharge,
this detergent is not necessarily the primary or only source of the detected nonylphenol.
Lubricants also contain alkylphenol ethoxylates, and oil, grease, and fuel also might accumulate
in bilgewater. The operators of three of the other vessels where bilgewater was sampled also
reported using commercial bilge cleaners, yet NP was not detected in samples from those
vessels. Furthermore, the operator of the tour boat from which 16 of the long- and short-chain
alkylphenol ethoxylates were detected made no comment about using bilge cleaners. It is
unlikely that ambient water is the source of NP to bilgewater in the detected sample.
                                          Ill

-------
                                                                          Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.8. Results of Bilgewater Sample Analyses for Nonylphenols1
Analyte
Units
No.
samples
No.
Detected
Detected
Proportion
(%)
Average
Cone.3
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone. 3
Screening
BM2
Long-Chain Alkylphenol Ethoxylates
Total Nonylphenol Polyethoxylates
Nonylphenol octodecaethoxylate (NP18EO)
Nonylphenol heptadecaethoxylate (NP17EO)
Nonylphenol hexadecaethoxylate (NP16EO)
Nonylphenol pendecaethoxylate (NP15EO)
Nonylphenol tetradecaethoxylate (NP14EO)
Nonylphenol tridecaethoxylate (NP1 3EO)
Nonylphenol dodecaethoxylate (NP12EO)
Nonylphenol undecaethoxylate (NP1 1 EO)
Nonylphenol decaethoxylate (NP10EO)
Nonylphenol nonaethoxylate (NP9EO)
Nonylphenol octaethoxylate (NP8EO)
Nonylphenol heptaethoxylate (NP7EO)
Nonylphenol hexaethoxylate (NP6EO)
Nonylphenol pentaethoxylate (NP5EO)
Octylphenol dodecaethoxylate (OP12EO)
Octylphenol undecaethoxylate (OP1 1 EO)
Octylphenol decaethoxylate (OP10EO)
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
260
0.78
1.8
3.6
6.8
12
20
27
35
35
33
28
22
16
9.7
0.97
1.4
3.2








































































530
1.8
4.2
8.2
16
28
44
61
77
77
70
57
42
27
14
0.25
0.38
0.39
1100
3.6
8.4
16
31
56
88
120
150
150
140
110
83
53
27
0.49
0.77
0.78
1100
3.6
8.4
16
31
56
88
120
150
150
140
110
83
53
27
0.49
0.77
0.78
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Short-Chain Nonylphenols
Bisphenol A
ug/L
4
1
25
5.3




11
15
15
NA
Nonylphenols
NP
ug/L
4
1
25
9.2




3.7
4.9
4.9
1.7
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) In some cases, the detected concentration(s) for an analyte could be lower than the  replacement value (% of the reporting limit) for a concentration that was nondetected. In an
extreme (but possible) case, this could result in an average concentration for an analyte that is greater than the maximum detected concentration.
                                                                             112

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       3.2.1.8    Summary of the Characterization of Bilgewater Discharge

       Table 3.1.9 summarizes the specific analytes within bilgewater effluent that may have the
potential to pose risk to human health or the environment for these types of vessels based on
these samples. EPA's interpretation of a realized risk likely posed by these analytes, relative to
pollutant loadings, background ambient and source water contaminant levels and characteristics,
and other relevant information useful for this assessment, is presented in Chapter 5.

       In summary, among the metals, dissolved copperand zinc were consistently measured at
concentrations exceeding the most stringent NRWQC in fishing vessels, tow/salvage vessels,
water taxis, and tour vessels; total arsenic was also measured at concentrations exceeding the
most stringent NRWQC in a Longliner fishing vessel, a tow/salvage vessel, and tour vessel. The
classical pollutants BOD, COD, sulfide, TSS, and TRC exceeded the screening benchmarks in at
least one of the fishing vessels, tow/salvage vessels, water taxis, and tour vessels. Among the
pathogen indicators, enterococcus, E. coli and fecal coliform bacteria were all present at
concentrations exceeding NRWQC; these samples were collected only from fishing boats. Total
phosphorus was the only nutrient to exceed a screening benchmark in bilgewater from all vessel
types, while ammonia exceeded the screening benchmark in a fishing  vessel (longliner).
Concentrations of the SVOC bis(2-ethylhexyl) phthalate exceeded NRWQC in the bilgewater
discharges of fishing vessels, tow/salvage vessels, water taxis, and tour vessels, while 2,4,6-
trichlorophenol exceeded the screening benchmark in only the tour vessel. Benzene and toluene
sampled from tow/salvage vessels were the only VOCs found at concentrations exceeding the
most stringent NRWQC. The screening benchmark for total nonylphenol was exceeded in a
single sample collected from a fishing vessel.
                                           113

-------
                                                                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.1.9. Characterization of Bilgewater Discharge and Summary of Analytes that May Have the Potential to Pose Risk








Vessel Type (no. vessels)








Fishing (2)


Tow/Salvage (2)


Water Taxis (2)

Tour(1)



Analytes that May Have the Potential to Pose Risk in Bilgewater Discharge and Vessel Sources1'2








V)
re
o
O)
o
0
o
i.
.y
^
enterococcus,
E. co//,
Fecal Coliform













U)
•D
C
3
1
o
u
'E
re
O)
O
0)
13
o




Benzene,
Toluene







U)
•D
C
3
1
E
o
o
u
'E
re
s>
o
Q>
're
§

V
CO



Bis(2-ethylhexyl)-
phthalate

Bis(2-ethylhexyl)-
phthalate
Bis(2-ethylhexyl)-
phthalate,
2,4,6-Trichloro-
phenol







-1— ^
0)
~o
U)
n
* ^
Q. "S
5) re
5 re
Q.
TRC


TRC


TRC

TRC



(1) Analytes are generally bolded when a large proportion of the samples have concentrations exceeding the NRWQC (e.g., 25 to 50 percent), when several of the samples have
PHQs > 10 (e.g., two or three of five), when a few samples result in PHQs greatly exceeding the screening benchmark (i.e., 100s to 1,000s), or, in the case of oil and grease and for
nonylphenol, when one or more samples exceed an existing regulatory limit by more than a factor of 2. See text in Section 3.1.3 fora definition of PHQs and Table 3.1 for screening
benchmarks used to calculate these values.
(2) EPA notes that the conclusion of potential risk is drawn from a small sample size, in some cases a single vessel, for certain discharges sampled from some vessel classes.  EPA
included these results in the tables to provide a concise summary of the data collected in the study, but strongly cautions the reader that these conclusions, where there are only a few
samples from a given vessel class, should be considered preliminary and might not necessarily represent pollutant concentrations from these discharges from other vessels in this
class.
                                                                           114

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

   3.2.2  Stern Tube Packing Gland Effluent

       The packing gland or stuffing box surrounds the propeller shaft at the point it exits a
boat's hull underwater. Based on the vessels sampled for this analysis, using a packing gland is a
common method for preventing water from entering the hull while still allowing the propeller
shaft to turn. A stuffing box packed with greased flax rings is designed to leak a few drops per
minute of ambient water to cool the gland when a vessel is underway. Stuffing boxes are also
used to seal rudder stocks that penetrate the hull below the waterline. The packing gland effluent
water is often collected in a segregated section of the bilge that generally contains an automatic
bilge pump.

       During this study, EPA observed this segregated discharge onboard tugboats but not on
any other vessel classes. In most of the other vessels sampled, the packing gland effluent dripped
directly into the bilge. Possible constituents of concern in the packing gland effluent include
metals (from contact of the discharge with the drive shaft), hydraulic fluid, grease or lubricants
found in the gland, and fuel constituents since the packing gland is located in the engine
compartment.

       Based on field observations from EPA's vessel  sampling program, EPA estimated the
drip rate into the stuffing box at approximately  10 drips per minute, which is consistent with the
literature data (Casey, 2007; Chin, 2005). This equates to a stern tube effluent generation rate of
between 2 and 4 gpd. Since most tugboats had dual propeller systems, these boats are expected to
generate between 4 and 8 gpd of stern tube effluent.

       For this study, EPA collected samples from the packing gland effluent from nine
tugboats. Samples on these vessels were analyzed for metals (dissolved and total), classical
pollutants, nutrients, VOCs, SVOCs, and nonylphenols. Packing gland effluent samples were
collected by placing a glass transfer jar under the shaft to collect any water dripping and then
compositing the sample in a Teflon-lined pail. In some cases, EPA dipped the transfer jar into the
segregated bilge compartment. If the vessels had a dual propeller system, EPA collected samples
from each for the composite. However, samples for analysis of oil and grease and VOCs are not
appropriate to composite, so these samples were collected separately.

         3.2.2.1    Metals

       Packing gland effluent samples were analyzed for both total and dissolved concentrations
of 22 metals. Of the 22 metals, 18 total metals and 15 dissolved metals were detected in the EPA
sample set (see Table 3.2.1). Antimony, beryllium, silver,  and cadmium were not detected in any
samples in the total or dissolved form, while cobalt, iron,  thallium, and vanadium were not
detected in the dissolved form. Figures 3.2.1 and 3.2.2 present box and dot density plots of the
detected results for dissolved and total metals, respectively. The box and density plots in Figures
                                           115

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
3.2.3 and 3.2.4 present these same detected results for dissolved and total metals, respectively,
normalized by the lowest NRWQC where applicable. Points on these plots above the dashed line
(demarking a PHQ of 1) indicate metals concentrations exceeding the benchmark. With a few
exceptions, the metal concentrations normalized by the lowest NRWQC were below the PHQ of
1.

       Dissolved and total aluminum were found in all nine samples analyzed. Dissolved
aluminum was detected at concentrations ranging from 7.8 to 150 ug/L in the packing gland
effluent; however, no screening benchmark is available for dissolved aluminum. Total  aluminum
was detected at concentrations ranging from  50.7 to 6,400 ug/L and exceeded the screening
benchmark of 87 ug/L eight times. Arsenic, both total and dissolved, was detected in three of
nine samples in the packing gland effluent, although the sample with the highest measured total
and dissolved arsenic concentrations may be elevated resulting from positive interference (see
discussion in Section 3.1.3). All three total arsenic values exceeded the screening benchmark of
0.018 ug/L (based on the human health criterion for drinking water plus fish consumption) with
values of 2.8, 4.4, and 15.3 ug/L. None of the three detected dissolved arsenic values (1.2, 1.4
and 14.7 ug/L) exceeded the screening benchmark of 36 ug/L (based on the saltwater chronic
criterion for the protection of aquatic life). Dissolved  copper was detected in four of nine
samples with values ranging froml6.2 to 92  ug/L. All four sample values exceeded the screening
benchmark of 3.1 ug/L. Total copper was detected in  seven of the nine samples from the packing
gland effluent, with values ranging from 7 to 891 ug/L. None of the total  copper values exceeded
the screening benchmark of 1,300 ug/L.

       Dissolved and total nickel was detected in six  of nine and eight of nine packing gland
effluent samples  respectively. Two of the total nickel  results (1,670 and 3,230 ug/L) exceeded
the screening benchmark of 610 ug/L, while all of the dissolved nickel values exceeded the
screening benchmark of 8.2 ug/L. Zinc was found in seven of nine samples in the dissolved form
and eight of nine samples in the total form. One sample value of 120 ug/L for dissolved zinc
exceeded the screening benchmark of 81 ug/L. Selenium was found in three of nine samples in
the dissolved form and only one of nine samples in the total form (the latter an exceptionally
high concentration of 42.1 ug/L suspected of reflecting positive interference). Dissolved
chromium and lead were also detected in several samples. Chromium values exceeded  the
benchmark criteria of 11 ug/L in four detected samples. Dissolved lead was detected at a
concentration of 4.9 ug/L, which slightly exceeded  the benchmark of 2.5  ug/L.

       Total iron, manganese, and thallium were all detected at levels below the screening
benchmarks, except for one sample for total thallium that was detected at the reporting level of 1
ug/L.  This sample exceeded the benchmark of 0.24 ug/L for thallium and has a PHQ of 4.17 as
shown on Figure 3.2.4. Barium, sodium, and potassium, in both forms (total and dissolved) were
detected in three  of three samples, but did not exceed  benchmark criteria. The metals magnesium
and calcium, in both forms (total and dissolved); cobalt and vanadium (in total); and dissolved

                                          116

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
manganese were detected in one or more samples but no screening criteria exists for these
compounds.

       EPA analyzed ambient metal concentrations to determine if dissolved and total aluminum
concentrations found in packing gland effluent were contributed primarily by the vessel or
reflected contributions primarily by background ambient concentrations. For both dissolved and
total aluminum, sample concentrations were moderately influenced by ambient background
concentrations, with ambient concentrations as high as 130 ug/L (dissolved aluminum) and 3,950
ug/L (total aluminum). For both dissolved and total arsenic, sample concentrations from stern
tube packing gland effluent were strongly influenced by ambient background concentrations.
Ambient dissolved and total arsenic concentrations as high as 16.1 and 15.4 ug/L, respectively,
were measured in water surrounding one of the three vessels sampled (a vessel sampled in
Baltimore, Maryland), although these measured concentrations may be elevated due to positive
interference. Ambient background concentrations of both dissolved and total copper were
comparatively low relative to the packing gland effluent sample concentrations and therefore of
little influence (i.e., dissolved and total copper concentrations were largely from packing gland
effluent). As in the case of copper, nickel was not found at high levels in the surrounding
ambient water; thus, nickel is another metal that may have a significant source from the packing
gland effluent. All of the selenium values were consistent with concentrations in the surrounding
water. Neither chromium nor lead was strongly influenced by ambient concentrations in the
surrounding water. The concentrations barium, sodium, potassium, magnesium, calcium cobalt,
vanadium, and manganese generally reflect the concentrations in the surrounding water.
                                           117

-------
                                                          Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.2.1. Results of Packing Gland Effluent Sample Analyses for Metals1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum, Dissolved4
Aluminum, Total4
Arsenic, Dissolved3
Arsenic, Total3
Barium, Dissolved3
Barium, Total3
Chromium, Dissolved
Chromium, Total
Cobalt, Total4
Copper, Dissolved
Copper, Total
Iron, Total4
Lead, Dissolved4
Lead, Total
Manganese, Dissolved4
Manganese, Total4
Nickel, Dissolved
Nickel, Total
Selenium, Dissolved3
Selenium, Total3
Thallium, Total4
Vanadium, Total3
Zinc, Dissolved4
Zinc, Total
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
9
9
9
9
3
3
9
9
3
9
9
3
9
9
9
9
9
9
9
9
3
3
9
9
9
9
3
3
3
3
5
8
2
4
7
3
1
3
8
9
6
8
3
1
1
1
7
8
100
100
33
33
100
100
56
89
67
44
78
100
11
33
89
100
67
89
33
11
33
33
78
89
88
1200
3.3
3.8
53
88
19
230
3.0
22
140
3900
1.8
7.9
44
160
210
610
8.1
8.6
0.67
4.6
34
70
110
300


63
98
3.8
130
2.9

20
2700


9.6
110
13
45




18
73
7.8
51


30
32





710



79








7.8
51


30
32





710



79








31
170


30
32

9.7


3.5
710


2.9
93

12




3.3
11
140
1500
1.3
3.6
66
140
20
440
5.6
38
150
8300

8.9
53
230
370
970
1.2

1.0
13
53
120
150
6400
15
15
66
140
110
760
5.6
92
890
8300
4.9
43
250
350
1000
3200
41
42
1.0
13
120
180
150
6400
155
155
66
140
110
760
5.6
92
890
8300
4.9
43
250
350
1000
3200
41 5
425
1.0
13
120
180
NA
87
36
0.018
NA
1000
11
NA
NA
3.1
1300
300
2.5
NA
NA
100
8.2
610
5
170
0.24
NA
81
7400
Cationic Metals
Calcium, Dissolved3
Calcium, Total3
ug/L
ug/L
9
9
9
9
100
100
36000
37000
24000
24000
23000
22000
23000
22000
23000
23000
35000
39000
1 1 0000
1 1 0000
110000
110000
NA
NA
                                                             118

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 Table 3.2.1. Results of Packing Gland Effluent Sample Analyses for Metals1
Analyte
Magnesium, Dissolved3
Magnesium, Total3
Potassium, Dissolved3
Potassium, Total3
Sodium, Dissolved3
Sodium, Total3
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
samples
9
9
3
3
3
3
No.
detected
9
9
3
3
3
3
Detected
Proportion
(%)
100
100
100
100
100
100
Average
Cone.
40000
39000
39000
37000
810000
810000
Median
Cone.
7800
7900
4700
4600
20000
20000
Minimum
Cone.
6200
6000
4000
4600
18000
17000
10%
6200
6000
4000
4600
18000
17000
25%
6500
6300
4000
4600
18000
17000
75%
11000
12000
1 1 0000
100000
2400000
2400000
90%
290000
280000
1 1 0000
100000
2400000
2400000
Maximum
Cone.
290000
280000
110000
1 00000
2400000
2400000
Screening
BM2
NA
NA
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3,  and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) Sample concentrations are strongly influenced by background concentrations in ambient water, accounting for greater than 90% of sample concentrations in the majority of
samples.
(4) Sample concentrations are moderately influenced by background concentrations in ambient water, accounting for between 50 and 90% of sample concentrations in the majority of
samples.
(5) Maximum concentrations may be elevated as a result of positive interference.
                                                                             119

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
        1000
    O)
    ^

    c
    o

   "co
    CD
    o
    c
    o
   O
100
  10
—
-
—
-
—
-
-
-



&
00
o
	 £
c






1 1

t
9
©

oooooo
©
©




o
— e

T
1 1




{

_Q





CL







0000(









'










OOP








1
*
•

~a
-QJ


Q—
0—

-o
£
0—^
|







-
O


O
OOP


—
©
o
o 	
oooooo ojp _
©
-
                              Dissolved Metals
Figure 3.2.1. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in

Samples of Packing Gland Effluent

(Note: Maximum concentrations of arsenic and selenium may be elevated as a result of positive interference).
                                         120

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment





_ 1000
1
__l
~O)

^^j
s^^^x

o 100
"co


CD
S 10
O
O


1


_

~


—
_
_



—
_
_
_
-
—
~
-

_
—
~

i


c


)


0
-Q-

3£

(
C

)
)








I I




^





&
M

©

J

c








)
C

)


-G-



1
o

9
c
&
OOQOOO

|
c
I I



C


















































I




(




C






<
)




)
3n


0
3€

-6
rl
)

)


-€

(

5-

}







JJ


olo

|

1

-e-


I
































1












c


p



oc

|


;


q



0(

(

>

1 1 1 1
Q


C


CjCRJ f
'

_ _
W






)



~o~
o



*— '

0
0






O


C

^OQDOO

^
A
1 1 1 T '

















c






















)





i



































G
J
-e














)
)
c
3-



o
3

c






)





-

~


—
_
_



—
_
_
_
-
—
-
-

_
—
~























                                    Total  Metals
Figure 3.2.2. Box and Dot Density Plot of Total Metals Concentrations Measured in
Samples of Packing Gland Effluent
(Note: Maximum concentrations of arsenic and selenium may be elevated as a result of positive interference).
                                            121

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


100.0

^— •
c
"o
o 10-°
•a
CD
N
CD

- 1.0
CD
c
CD
-i— •
O
Q_
0.1




I I I

—
=
-

— G
-
_

0

_ 	
-
O

-
— (

c








)



p

)


-e-

K
J
_ ^~ 	 -^^
(QOQOOCy
" I <$> I

•)(
L.


<^c
^* —



D







G
C

Q









)
)

QL





)DC





























$r
^X.


c














G


)C










5


)f
)O(










I
c

C










-^ <(
-{
C




f
— t
c
I
)

)

)-]
)




•N
i—
>-<
)OC
VL.

>)



I

o





SoOdiCI
D
— *

eg)

c
k
I











C
1
d











)

0
o


-e-
o
>eu



n
w
--^



—
=
-

—
-
_

^

- —
-
-
-
-
—
N _
S —
-





















                                 Dissolved Metals
Figure 3.2.3. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals
in Samples of Packing Gland Effluent
(Note: Replacement values for non-detects are circled. Also, maximum concentrations of arsenic and selenium may
be elevated as a result of positive interference.)
                                            122

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

-£ 100.000
CD
§ 10.000
a
•a
i_
03 *i nnn
j^j I.UUU
03
I
| 0.100
c
£ 0.010


0.001
E I T I I I I I I
rdb>
~
- Q O
= 1 i
C)

- © ?
~
~_
r 6 r
G Q
(5€
' 1

r i i i i i i i i
1 1 1 1 1 1 E
-=
^
-
Q I
Q
^bd
atcrcn |
(3a
0 ~
«- E
(^OMO§> :
^@ pb^
2
o I
	 X-E
                                     Total  Metals
Figure 3.2.4. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
Samples of Packing Gland Effluent
(Note: Replacement values for non-detects are circled. Also, maximum concentrations of arsenic and selenium may
be elevated as a result of positive interference).
                                             123

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

        3.2.2.2     Classical Pollutants

       EPA sampled the packing gland effluent for numerous classical pollutants to further
characterize this discharge type for the tugboats sampled under this program. The classical
pollutants include measurements that are physical properties (temperature, conductivity, salinity,
turbidity, TSS), oxygen consumption (BOD, COD), oil and grease (HEM and SGT-HEM), as
well as chemical concentrations (pH, sulfide, DO, and TRC). Table 3.2.2 presents the data for
these parameters.

       Figure 3.2.5 illustrates the varied concentrations of measured for these parameters in the
packing gland effluent. Most of the concentrations and values reported reflect the concentrations
and values in the ambient water surrounding the vessel,  as this water is the source of the drive
shaft water. Two parameters (sulfide and TRC) were not detected in any samples.

       The PHQs were calculated for the classical pollutants for which they were available.
Only two pollutants exceeded these PHQ screening benchmarks (see Figure 3.2.6): oil and
grease and TSS. One of the vessel samples had values which exceeded the screening benchmark
for oil and grease measured as both FIEM and petroleum hydrocarbon (SGT-HEM). The
concentrations detected were 66.7 mg/L for FIEM and 55.8  mg/L for SGT-HEM, both of which
exceeded the benchmark of 15 mg/L. EPA noted a visible oily sheen on the surface of this
effluent and evidence of settled hydrocarbons on the bottom of the tank as this sample was
collected. Based upon conversations with the vessel engineer, the likely source is an oil leak that
was somehow making its way into this effluent. This seems a plausible explanation given that
background concentrations of FIEM and SGT-HEM in surrounding ambient water were very  low
(< 1.5 mg/L) relative to the measured sample concentrations.

       Total suspended solids were detected in all nine  samples collected from the packing
gland effluent. Two samples with concentrations of 269 and 134 mg/L exceeded the screening
benchmark of 98 mg/L.
                                          124

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.2.2. Results of Packing Gland Effluent Sample Analyses for Classical Pollutants1
Analyte
Conductivity
Dissolved Oxygen
Total Organic Carbon (TOC)
Biochemical Oxygen Demand (BOD)
Chemical Oxygen Demand (COD)
Hexane Extractable Material (HEM)
PH
Salinity
Silica Gel Treated HEM (SGT-HEM)
Temperature
Total Suspended Solids (TSS)
Turbidity
Units
mS/cm
mg/L
mg/L
mg/l
mg/l
mg/l
SU
ppt
mg/l
C
mg/l
NTU
No.
samples
9
9
7
9
9
9
9
8
9
9
9
9
No.
detected
9
9
7
9
4
5
9
7
5
9
9
9
Detected
Proportion
(%)
100
100
100
100
44
56
100
88
56
100
100
100
Average
Cone.
1.6
8.3
3.5
11
31
14
7.1
0.14
13
20
59
46
Median
Cone.
0.30
8.4
3.5
7.2

1.65
7.5
0.20
1.7
20
28
18
Minimum
Cone.
0.22
5.3
2.2
3.3


2.4


9.3
5.6
9.0
10%
0.22
5.3
2.2
3.3


2.4


9.3
5.6
9.0
25%
0.22
7.2
2.6
4.3


7.3
0.10

18
13
13
75%
0.55
9.3
4.8
13
53
23
8.0
0.20
19
23
81
70
90%
12
11
4.9
35
88
67
8.2
0.20
56
26
270
190
Maximum
Cone.
12
11
4.9
35
88
67
8.2
0.20
56
26
270
190
Screening
BM2
NA
NA
NA
30
NA
15
NA
NA
15
NA
30
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage  of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3,  and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            125

-------
     100.0
 o
 E
.o
"co
 CD
 O
 c
 o
O
      10.0
        1.0
        0.1
                    Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                        I    I
I    I
                               ©
1    I
               *
               _
               t2)
                               Parameter
Figure 3.2.5. Box and Dot Density Plot of Classical Pollutants Measured in Samples
of Packing Gland Effluent
                                      126

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     10
       8
       6

|    4
"o
O    2
       03
       N
       03
       03
       O
       Q_
                                        0
                                                    o
            ya
            -e-

             o
                   Q

                    I    I
                                          I    I
I    I
                   0°"
                                    Parameter
Figure 3.2.6. Box and Dot Density Plot of Potential Hazard Quotients for Classical
Pollutants in Samples of Packing Gland Effluent
(Note: Replacement values for non-detects are circled).
                                         127

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

         3.2.2.3    Nutrients

       Packing gland effluent samples were analyzed for four nutrient-related parameters:
ammonia nitrogen, nitrate/nitrite, TKN, and total phosphorus (see Table 3.2.3). Figures 3.2.7 and
3.2.8 illustrate the variability of the nutrients in the packing gland effluent. Ammonia,
nitrate/nitrite, and TKN were detected in most of the samples analyzed, but in relatively low
concentrations. Phosphorus was detected in seven of the nine tugboat samples collected.

       Only ammonia has a current numeric NRWQC value. The results for ammonia detected
in the packing gland effluent range from 0.07 to 0.23 mg/L, well below the benchmark of 1.2
mg/L. TKN and nitrate/nitrite were detected in all of the nine tugboat samples, with values
ranging from 0.40 to 1.8 mg/L for TKN to 0.62 to 1.5 mg/L for nitrate/nitrite. Total phosphorus
was detected in seven of the nine samples for packing gland effluent. The detected
concentrations ranged from 0.06 to 0.25 mg/L and only two values, 0.19 and 0.25 mg/L, exceed
the 0.1 mg/L benchmark.

       Most of these values for ammonia, TKN, and nitrate/nitrite are consistent with ambient
background results in each location. The background ambient for these total phosphorus samples
reported values from 0.06 to 0.19 mg/L, indicating a moderate influence of surrounding ambient
water on sample concentrations.

       In general, it appears that nutrient concentrations from packing gland effluent are
generally low and the wastestream does not appear to be adding significant nutrients to the
surrounding waters. Nutrient addition from packing gland effluent was not considered a likely
concern in this discharge relative to metals from contact of the discharge with the drive shaft,
hydraulic fluid, grease or lubricants from the gland, and fuel constituents.
                                           128

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.2.3. Results of Packing Gland Effluent Sample Analyses for Nutrients1
Analyte
Ammonia As Nitrogen (NH3-N)
Nitrate/Nitrite (NO3/NO2-N)
Total Kjeldahl Nitrogen (TKN)
Total Phosphorus
Units
mg/L
mg/L
mg/L
mg/L
No.
samples
9
9
9
9
No.
detected
7
9
9
7
Detected
Proportion
(%)
78
100
100
78
Average
Cone.
0.10
0.69
1.1
0.13
Median
Cone.
0.10
0.62
1.4
0.10
Minimum
Cone.

0.085
0.41

10%

0.085
0.41

25%
0.034
0.58
0.69
0.030
75%
0.14
0.80
1.4
0.22
90%
0.23
1.5
1.8
0.25
Maximum
Cone.
0.23
1.5
1.8
0.25
Screening
BM2
1.2
NA
NA
0.10
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall.  So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            129

-------
          2.0
          1.6
          1.2
          0.8
      O)
      03
 CD
 O
 c
 o
O
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                        oo
                                       ©
                                      ooloo
                                       o

                                       -Q-
                                                      T
                                                      oo
                                                      oo
                           <^C
                                  Nutrients
Figure 3.2.7. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of
Packing Gland Effluent
                                       130

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
      CD
      "-I—•
      O
      ^

      O
      03
      N
      03
      03
      CD
      -i—•
      O
      Q_
1.0
           0.5
                                                          OO
O
n


                                    Nutrients
Figure 3.2.8. Box and Dot Density Plot of Potential Hazard Quotients for Nutrients in
Packing Gland Effluent
                                         131

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
         3.2.2.4    Volatile and Semivolatile Organic Chemicals

       Packing gland effluent samples were analyzed for 70 VOCs and 73 SVOCs in nine
tugboats (see Table 3.2.4). Of the analytes tested, six VOC compounds and 10 SVOC
compounds were detected in the samples. Figures 3.2.9 and 3.2.10 illustrate the range of
concentrations measured for SVOCs and VOCs, respectively.

       Three VOCs, m-p-xylene, acetone, and methylene chloride, were detected in more than
one sample. Eight of the 10 SVOCs detected were found in one sample. Bis(2-ethylhexyl)
phthalate was found in the effluent of three vessels sampled and n-hexadecane was found in the
effluent of two of the vessels sampled. Bis(2-ethylhexyl) phthalate was detected at notably high
(compared to ambient surrounding water) values of 2.8, 5.4, and 23.5 |ig/L. The only other
compound with a screening benchmark is di-n-butyl phthalate,  which was detected in one  sample
with a concentration of 2.45 |ig/L, which is well below the screening benchmark of 2,000  |ig/L.
These two phthalate compounds are used as plasticizers, and bis(2-ethylhexyl) phthalate is used
as a hydraulic fluid and as a dielectric fluid in capacitors.

       Figure 3.2.12 presents the distributions of PHQs, based on the most conservative
screening benchmarks, for each VOC; none of the detected values exceed the screening
threshold22 . PQH was above one for all three samples of bis(2-ethylhexyl) phthalate, based on
the screening benchmark of 1.2 |ig/L (Figure 3.2.11).

       Of the six VOC and 10 SVOC compounds detected in packing gland effluent samples,
bis(2-ethylhexyl) phthalate was the only compound whose measured concentrations in the
discharge was substantially higher than in ambient water; all other VOCs and SVOCs detected in
packing gland effluent appear to reflect the similar  concentrations found in surrounding water.
  PHQs for benzene, methylene chloride and tetrachloroethene in multiple packing gland effluent samples were
based on replacement values of 1A of the reporting limit for nondetected concentrations. In Figure 3.2.12 the PHQs
based on replacement values for nondetected concentrations have been circled for identification. EPA does not
consider PHQs that exceed 1 to signal that these discharges pose a potential risk to cause or contribute to the non-
attainment of a water quality standard when the PHQs are based on replacement values for nondetected
concentrations.

                                            132

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.2.4. Results of Packing Gland Water Sample Analyses for SVOCs and VOCs1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone. 3
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone. 3
Screening
BM2
SVOCs
2,6,1 0,14-Tetramethyl Pentadecane
3,6-Dimethylundecane
5-Butyl-Hexadecane
Bis(2-ethylhexyl) phthalate
Di-n-butyl phthalate
Dodecane
Eicosane
n-Hexadecane
Nonanoic Acid
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
1
1
1
9
9
1
1
2
1
1
1
1
3
1
1
1
2
1
100
100
100
33
11
100
100
100
100
12
8.7
6.7
4.7
1.7
5.0
5.4
5.5
4.3







6.0








5.0








5.0








5.0




4.1



6.0




24
2.5


6.0




24
2.5


6.0

NA
NA
NA
1.2
2000
NA
NA
NA
NA
VOCs
Acetone
Benzene
m-,p-Xylene (sum of isomers)
Methylene chloride
n-Pentadecane
Sulfur dioxide
Tetrachloroethene
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
3
9
3
9
1
1
9
3
1
2
2
1
1
1
100
11
67
22
100
100
11
2.9
1.4
1.7
1.2
11
13
1.4
2.7

0.10




2.7






2.7






2.7






3.2

0.10
0.10



3.2
0.20
0.10
0.20


0.20
3.2
0.20
0.10
0.20


0.20
NA
2.2
NA
4.6
NA
NA
0.69
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) In some cases, the detected concentration(s) for an analyte could be lower than the replacement value (% of the reporting limit) for a concentration that was nondetected. In an
extreme (but possible) case, this could result in an average concentration for an analyte that is greater than the maximum detected concentration.
                                                                             133

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
            c
            g
           ~co
           "c
            0
            o
25
20
15
10
5
(
_iii
—
—
—
—
i i i
) 1 2 3
*



c
<




00
I





(.
4
I




T
;ujyt,
I
5
i i i i
—
—
—
--B_.
I I I I






6 7 8 9 10
SVOCs
Figure 3.2.9. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
Packing Gland Effluent Samples
SVOCs are identified as follows:
(1) 2,6,10,14-Tetramethyl Pentadecane
(2) 3,6-Dimethylundecane
(3) 5-Butyl-Hexadecane
(4) Bis(2-ethylhexyl) phthalate
(5) Di-n-butyl phthalate
(6) Dodecane
(7) Eicosane
(8) n-Hexadecane
(9) Nonanoic acid
                                               134

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
            10.0
        CO
        Z5
        C
        o
       "CD
        i_
       "c
        CD
        O
        c
        o
       O
1.0
              0.1
                  0      1
                        345
                           VOCs
678
Figure 3.2.10. Box and Dot Density Plot of VOC Concentrations Measured in Samples of
Packing Gland Effluent Samples
VOCs are identified as follows:
(1) Acetone
(2) Benzene
(3) m-,p-Xylene (sum of isomers)
(4) Methylene chloride
(5) n-Pentadecane
(6) Sulfur dioxide
(7) Tetrachloroethene
                                           135

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
    CD
    "o
    Z5
    O
    -a
    s__
    03
    N
    03

    "03
    o
   Q_
        10.000
1.000
0.100
          0.010
          0.001
                                            i     i     i     i     r
                                 I
                                           j	i
I
                 0    1
            )OQPO(
23456789   10
           SVOCs
Figure 3.2.11. Box and Dot Density Plot of Potential Hazard Quotients for SVOCs in
Samples of Packing Gland Effluent
SVOCs are identified as follows (replacement values for non-detects are circled):
(1) 2,6,10,14-Tetramethyl Pentadecane
(2) 3,6-Dimethylundecane
(3) 5-Butyl-Hexadecane
(4) Bis(2-ethylhexyl) phthalate
(5) Di-n-butyl phthalate
(6) Dodecane
(7) Eicosane
(8) n-Hexadecane
(9) Nonanoic Acid
                                          136

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                 4
                 3
             o
             ID
            O
             03
             N
             03
             o
            Q_
                                         oonnoo
                                            OP
                                              1
                   0     1
345
   VOCs
7      8
Figure 3.2.12. Box and Dot Density Plot of Potential Hazard Quotients for VOCs in
Samples of Shaft Packing Gland Effluent
VOCs are identified as follows (replacement values for non-detects are circled):
(1) Acetone
(2) Benzene
(3) m-,p-Xylene (sum of isomers)
(4) Methylene chloride
(5) n-Pentadecane
(6) Sulfur dioxide
(7) Tetrachloroethene
                                            137

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.2.5    Nonylphenols


       EPA analyzed samples of shaft packing gland effluent for long and short chain
nonylphenol and octylphenol ethoxylates and NP because of the possibility of alkylphenol-
containing water from the bilge or other areas of the vessel leaking into the shaft packing gland
effluent compartment. Table 3.2.5 presents the detected results.

       Of the nine samples for which long- and short-chain nonylphenol and octylphenol
ethoxylates were analyzed, only six long-chain isomers of the octylphenol polyethoxylate
(OPEO) type were detected: OP12EO, OP11EO, OP10EO, OP9EO, OP8EO, and OP7EO. All of
the detected OPEOs are long-chain octylphenols and were found in one tugboat sampled. The
OPEO with the longest ethoxylate chain (OP12EO) was detected at the lowest concentration
(Figure 3.2.13).  The OPEO isomers showed the trend of increasing concentrations as the size of
the ethoxylate chain is reduced (from OP12EO to OP7EO), indicating moderately advanced
degradation of the long-chain OPEOs in the packing gland.

       Average concentrations of OPEOs with the longest ethoxylate chains (OP12EO through
OP10EO) were similar to bilgewater effluent (see Table 3.1.8). In contrast to bilgewater effluent,
however, NP was not detected in packing gland effluent.

       None of the OPEOs detected in the packing gland effluent sample were detected in
ambient water, indicating a probable source from onboard the vessel (tugboat) - possibly from
seepage from  the bilge. Another possible source of OPEOs in packing gland effluent could be
from the use of lubricants for which octylphenol ethoxylates are common constituents.
                                          138

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.2.5. Results of Packing Gland Water Sample Analyses for Nonylphenols1
Analyte
Octylphenol dodecaethoxylate (OP12EO)
Octylphenol undecaethoxylate (OP1 1 EO)
Octylphenol decaethoxylate (OP10EO)
Octylphenol nonaethoxylate (OP9EO)
Octylphenol octaethoxylate (OP8EO)
Octylphenol heptaethoxylate (OP7EO)
Units
ug/L
Ug/L
ug/L
Ug/L
Ug/L
ug/L
No.
samples
9
9
9
9
9
9
No.
detected
1
1
1
1
1
1
Detected
Proportion (%)
11
11
11
11
11
11
Average
Cone.
1.7
2.6
4.8
5.3
10
13
Median
Cone.






Minimum
Cone.






10%






25%






75%






90%
12
15
22
26
30
28
Maximum
Cone.
12
15
22
26
30
28
Screening
BM2
NA
NA
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            139

-------
                    Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
           35


           30



       5 25




       o  20
       *—>
       05
       §  15
       o
       o
           10
                              Nonylphenols
Figure 3.2.13. Box and Dot Density Plot of Nonylphenol Concentrations Measured in

Samples of Packing Gland Effluent
                                      140

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

     3.2.2.6    Summary of the Characterization of Packing Gland Discharge

       Table 3.2.6 summarizes the specific analytes within packing gland effluent that may have
the potential to pose risk to human health or the environment for these types of vessels based
upon these samples. EPA's interpretation of a realized risk likely posed by these analytes,
relative to pollutant loadings, background ambient and source water contaminant levels and
characteristics, and other relevant information useful for this assessment, is presented in Chapter
5.

       To summarize the results of packing gland discharge measured in the nine tugboats,
metals were the constituents found most frequently and with the highest magnitudes of
exceedance of their respective screening benchmarks. Among the dissolved forms of metals,
concentrations of copper, chromiumand nickel exceeded the most stringent NRWQC
benchmarks. Among the total forms of metals, aluminum, arsenic, iron, manganese and nickel
exceeded the most stringent NRWQC benchmarks. However, concentrations of total iron and
total manganese in surrounding (ambient) waters were similar to concentrations measured in
packing gland discharge. Among the classical pollutants, most of the concentrations and values
reported reflect  the concentrations and values in the ambient water surrounding the vessel,  as this
water is the source of the drive shaft water. Exceptions were two samples for oil and grease
(HEM and SGT-HEM) values which exceeded screening benchmarks. Two (of nine) total
phosphorus samples also exceeded the screening benchmark; however, these concentrations were
similar to total phosphorus concentrations in the surrounding waters. Among the remaining
contaminants, the SVOC bis(2-ethylhexyl) phthalate had a PHQ of >10 for one of the vessels
sampled, and six of the relatively long-chained octylphenols were measured in one of the nine
vessels sampled.
                                          141

-------
                                                                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.2.6. Characterization of Packing Gland Effluent and Summary of Analytes that May Have the Potential to Pose Risk









Vessel Type (no. vessels)







Tugboats (9)

Analytes that May Have the Potential to Pose Risk in Packing Gland and Vessel Sources1









«
re
o

[o

o
u
i






•D
C
3
1

O
u
'E
re
O)
O
0)
*
o



U)
•D
C
3
1


O
u
c
re
O)
O
0)
IS
o
Qi

Bis(2-ethylhexyl)-
phthalate









f
°Q

H
U)
s

re ^
^ re
9 >
11
CO UJ









0
£
_Q.
^
c «
^ is
o >
i1^
i MI
X













in
'c
•3











o
e
•D
C
re
o
O
Q"
O
00
TP








•D
"o
(/)
•D
•D
V
Q.
U)
2
W
15
^
X








re
o
1
O
15
u
'to t/)

Q. "5
ll
5(0
Q.


(1) Analytes are generally bolded when a large proportion of the samples have concentrations exceeding the NRWQC (e.g., 25 to 50 percent), when several of the samples have PHQs > 10 (e.g.,
two or three of five), when a few samples result in PHQs greatly exceeding the screening benchmark (i.e.,  100s to 1,000s), in the case of oil and grease and for nonylphenol, when one or more
samples exceed an existing regulatory limit by more than a factor of 2, or when concentrations of analytes  are sufficiently high that they may have the potential to pose risks to local water bodies.
See text in Section 3.1.3 for a definition of PHQs and Table 3.1 for screening benchmarks used to calculate these values.
                                                                         142

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

   3.2.3  Deck Washdown

       Deck washdowns involve removing dirt, grit, or other materials that can impact the
integrity of the deck surface (for aesthetic and safety reasons) and are a common vessel
maintenance task. The process uses hoses and/or swabs (mops) to move the deck washdown
water, debris, and cleaning agents (if any) to the scuppers, which then discharge the water
overboard. EPA collected samples of deck water as it is drained through the scupper against the
hull of the vessel (see Section 2.2.4 Sampling Methods). More than half the vessels sampled
reported using detergents (dish soaps, ZEP™, Simple Green™) or other cleaners (chlorine
bleach) during the washdown process.  Depending on the vessel's design and function, deck
washdown water sometimes contains contaminants such as detergents, metals, oil, particulates,
and pathogens (the latter primarily from catch brought onboard fishing vessels).

       Deck washdowns can occur at any time onboard these classes of vessels. Fishing vessels
most often discharge while underway either into the nearshore (< 3 Nm from shore) or farshore
(> 3 Nm from shore). Washdowns are usually performed on fishing vessels after nets are pulled,
fish are brought onboard and cleaned, while returning to port, and after offloading the catch.
EPA notes that the majority of deck washdown samples from fishing vessels were taken while
the vessel was shoreside, and do not reflect constituents of deck washdown while the vessel is
engaged in fishing operations. Decks are washed less frequently for other types of vessels such
as water taxis, tour boats, and tow boats. Wash locations are generally pierside after excursions
or within the harbor for these types of vessels.

       The volume of deck washdown water generated by a vessel depends on the frequency of
deck washdown, the flow rate from the hose,  and the washdown time. Since most vessels use a
common garden-hose for deck washdowns, EPA estimated the flow rate to be between 10 and 12
gallons per minute (gpm). The time required for deck washdown varies depending on the type of
vessel and size. EPA observed during the vessel sampling program that most deck washdowns
were generally 15 minutes or less.

       Vessels such as tour boats, water taxis, and tow boats would generate an average
deckwash water volume between 20 and 30 gpd during the peak summer season, assuming their
decks are washed once every week. Deckwash water volume for fishing boats varies depending
on the type of boat. For example, trailers, trawlers, gill netters, and purse seiners sometimes
wash their decks three to four times per day while fishing, plus one additional time after
unloading seafood at the processing facility. For these vessels, deckwash volumes might range
between 750 and 900 gpd.
                                          143

-------
        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Collecting Deck Washdown Samples from a Tow and Salvage Vessel
       Deck Washdown Sample Collected in a Lined Bucket
                           144

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       For this study, EPA collected deck washdown samples from 32 vessels: 11 fishing
vessels (gillnetter, trawlers, and trailers), nine tugboats, six tow/salvage vessels, two tour boats, a
water taxi, a fire boat, a supply boat, and a recreational boat (see Table 2-1). EPA collected
single grab samples from one or more scuppers (composited sample if more than one accessible
scupper) on selected vessels for laboratory analysis in order to determine representative pollutant
concentrations for deck washdown across the range  of normal vessel operations.

       EPA also sampled a deck runoff discharge during a rain event. Deck runoff differs from
washdown in that the runoff discharge occurs because of precipitation or spray landing on the
deck in sufficient quantities to mobilize pollutants on the deck surface rather than an intentional
introduction of washdown water (often including detergents). However, deck runoff incorporates
pollutants that would have been included in an eventual washdown so the samples are
comparable. The deck runoff sample was collected from a fishing trawler that was being
unloaded at a fish processing facility in the Northeastern United States.

       EPA focused its sampling effort on the following analyte groups in deck
washdown/runoff that were expected to be present in the discharge: metals, classical pollutants,
pathogen indicators (commercial fishing vessels only), nutrients, nonylphenols, and semivolatile
and volatile organic compounds (tow/salvage vessels only). Results for each class of pollutant
are presented and discussed in the subsections below.

     3.2.3.1    Metals

       Deck washdown water samples were analyzed for dissolved and total  metals. The
analytical results are summarized in Table 3.3.1. The following metals were detected in 90
percent or more of the deck washdown water samples:

       •  Dissolved and total aluminum
       •  Dissolved and total barium
       •  Total chromium
       •  Total cobalt
       •  Dissolved and total copper
       •  Total iron
       •  Total lead
       •  Dissolved and total manganese
       •  Dissolved and total zinc.

       Concentrations of a number of other metals were detected in 50 percent or more of the
samples analyzed:

       •  Total antimony
                                           145

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


       •   Dissolved and total arsenic23 24
       •   Dissolved chromium
       •   Dissolved cobalt
       •   Dissolved iron
       •   Dissolved and total nickel
       •   Dissolved selenium25
       •   Dissolved and total vanadium.

       Figure 3.3.1 presents the concentration ranges for dissolved metals detected in the
samples. These plots show that dissolved metals concentrations span three orders of magnitude.
Aside from the alkali and alkali earth metals that are the major cations in seawater (Na, K, Ca,
Mg), average dissolved concentrations of iron, aluminum, and zinc were highest, followed by
dissolved barium, manganese, and copper. Concentrations of total metals are displayed in Figure
3.3.2, and follow the same general pattern, but are much higher than their corresponding
dissolved metal concentrations (f
-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       Deck washdown water samples collected from 29 of the 31 vessels sampled contained
dissolved copper concentrations that exceeded the saltwater chronic aquatic life criterion of 3.1
Hg/L. Dissolved copper concentrations ranged from 2.5 |ig/L for a tug and fishing (trawler) boat
to 204 |ig/L for the supply boat. The dissolved copper concentrations in deck washdown samples
from the tug and assorted fishing boats were evenly distributed across the entire range of
measured dissolved copper concentrations, while the tow/salvage, fire, taxi, tour, and supply
boats all had relatively high dissolved copper concentrations (above 30 |ig/L).

       Dissolved lead concentrations exceeding the freshwater chronic aquatic life criterion (2.5
|ig/L) were limited to just three (of nine) tugboats, five (of six) tow/salvage boats, one of the two
tour boats, and the fire and supply boats. Dissolved lead concentrations exceeding chronic
aquatic life criterion concentrations ranged from 2.9 |ig/L for one of the tugboats to 53.5 |ig/L
for the supply boat.

       Similar to dissolved copper, dissolved zinc in deck washdown samples collected from the
majority of vessels sampled (22 of 31) exceeded the most stringent 2006 NRWQC - the saltwater
chronic aquatic life criterion of 81 |ig/L. In contrast to dissolved copper, however, only the deck
washdown samples from the various types of fishing boats appeared to be evenly distributed
throughout the entire measured dissolved zinc concentration range, while dissolved zinc in deck
washdown water samples collected from all the tugboats exceeded the criterion. Dissolved zinc
concentrations in deck washdown water samples ranged from 16 |ig/L for a fishing vessel (the
gillnetter) to 1,200  jig/L for one of the tugboats. All but one of the tow/salvage boats produced
dissolved zinc in deck washdown water samples exceeding the criterion, as did the tour, fire, and
supply boats (the last with a measured dissolved zinc concentration of 465 |ig/L).

       For the other dissolved metals (chromium, nickel, and selenium) where measured
concentrations exceeded the saltwater and/or freshwater criteria in one or more of the deck
washdown water samples, the PHQs were generally less than two (most likely less than one for
dissolved selenium after considering there may be elevated measured  concentrations as a result
of positive interference for the four samples with measured dissolved  selenium concentrations
exceeding 5 |ig/L). For both chromium and nickel, the tow/salvage vessel type had the greatest
number of dissolved metal concentration exceedances for their respective most stringent criteria.
No information was available concerning the frequency of deck washdowns for the supply
vessel, although this particular vessel is known to transport petroleum products, and its deck
appeared visibly "soiled" to the samplers. According to the surveys, the tow/salvage boats
generally undergo deck washdowns once to twice per week, about the same frequency as
tugboats, but less frequent than the fishing and tour boats.

       Four of the total metals (aluminum, arsenic, iron, and manganese) exceeded the most
stringent 2006 NRWQC in approximately half (manganese) or all the  deck washdown water
samples (aluminum, arsenic, and iron), although sample concentrations of these metals appear to

                                           147

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

be greatly influenced by surrounding ambient water concentrations (see Table 3.3.3). This
pattern was identical to the one observed for bilgewater discharge. In contrast to the bilgewater
samples, about half the deck washdown water samples for a fifth metal (antimony) exceeded the
most stringent 2006 NRWQC in the deck washdown water samples, as shown in Figure 3.3.4.
PHQs for total arsenic ranged from 56 to 4,600. All of the total arsenic concentrations exceeded
the most stringent human health  (water plus organism consumption) criterion of 0.018 |ig/L, as
well as the human health criterion for organism consumption alone, 0.14 |ig/L. The protective
human health criteria values for total arsenic are driven by the carcinogenic potential of this
metalloid. However, when compared to the less stringent saltwater chronic aquatic life criterion
for arsenic of 36 |ig/L, only five  of the 31 vessels produced total arsenic concentrations in deck
washdown water samples that exceeded this less stringent criterion, and the corresponding PHQs
ranged only from 1.0 to 2.3. These total arsenic exceedances were found on a shrimping vessel
(positive interference may have elevated the measured concentration, see footnote 24), three (of
the six) tow/salvage vessels, and the fire boat. In fact, the total arsenic concentrations in deck
washdown water samples from all six of the tow/salvage boats were close to  or within the upper
quartile of samples.

       Figure 3.3.3 displays the  distribution of PHQs based on the most conservative (most
protective) screening benchmark for each of the dissolved metals. PHQs for four of the dissolved
metals (cadmium, copper, lead, and zinc) include values from greater than 10 to  over 100,
indicating that the measured concentrations were one or more orders of magnitude greater than
the most conservative screening  benchmark. In addition, although the mean dissolved selenium
PHQ was less than one, there were two measured occurrences where PHQ exceeded 10,
however, the high measured concentration of dissolved selenium in these two samples are likely
due to positive interference, see footnote 25). PHQs exceeding one were also observed for
dissolved chromium and nickel, bringing to seven the number of dissolved metals that exceeded
the most stringent 2006 NRWQC in one or more deck washdown water samples.

       PHQs for total aluminum were also high, ranging from 7.5 to  150, followed closely by
total iron, with PHQs ranging from 3.1  to 48. For both metals, the majority of tug and
tow/salvage boats were consistently above the median (middle concentration  of the range) of
total metal concentrations (in addition to the fire and supply boats), while the fishing boats were
below the respective median total metal concentrations. Conversely, only three of the PHQs for
total manganese exceeded a value of 5 (a tugboat, the supply boat, and the water taxi).

       The frequency of PHQ exceedances for antimony, like total arsenic, are driven by the low
human health (water plus organism consumption) criterion of 5.6 |ig/L (the human health
criterion for organism consumption alone (640 |ig/L) is much higher). Only five of the 19 vessels
from which deck washdown water samples were obtained had PHQ below 1,  and were collected
from the supply, fire, recreational, and two of the salvage vessels. Among the PQHs for
antimony that were greater than  1, the low PHQ of 5.2 corresponded with the supply boat, and

                                           148

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

the high PHQ of 47 corresponded with a tow/salvage vessel - a PHQ value three times higher
than in the recreational vessel (PHQ = 15).

       From the perspective of potential risk, the discharges of metals where dissolved and total
concentrations exceed EPA's most stringent criteria correlate most strongly to utility, passenger,
or general service vessels such as the supply boat, tow/salvage boats, tugboats, water taxi, and
fire boat. Commercial fishing vessels may not be a source of concern except for metals such as
dissolved copper.

       EPA tested the hypothesis that the utility, service, and passenger vessels (referred to as
nonfishing vessels) discharged metals at higher concentrations than fishing vessels per discharge
event using two approaches. For both approaches, 20 nonfishing vessels (the tow/salvage boats,
tugboats, tour vessels, fire boat, water taxi, and supply vessels) were compared to the 10 fishing
vessels (six shrimping vessels, two trawlers, one trailer, and a gillnetter). For the analysis, when
multiple minimum detection limits were reported for a particular metal, the minimum
concentration was set to 1A of the highest reporting limit. This more conservative approach was
chosen to reduce the likelihood of detecting a difference that was not a "true" difference (Type I
error).

       For the first approach, a subset of the metals with the highest frequencies of screening
benchmark (NRWQC) exceedance from the nonfishing  vessels were compared to those from the
fishing vessels. Although there is no NRWQC for total lead, this metal was used in these
analyses because of the high proportion of nondetects in the dissolved form. This analysis was
performed using modified t-tests for unequal sample sizes and uneven variances (see Table
3.3.4). Concentrations of dissolved zinc and total lead were significantly higher in deck
washdown discharges of non-fishing vessels (e.g., tug boats) than fishing vessels. Although
concentrations of total arsenic were not significantly different between nonfishing and fishing
vessels, when the six tow/salvage vessels were compared to the remaining 24 vessels, total
arsenic concentrations in the tow/salvage vessels were significantly higher than in other vessels
(Table 3.3.4). When this analysis was performed for dissolved lead despite the occurrence of
nondetects, the results were the same (i.e., concentrations of dissolved lead in industrial vessels
were higher than in fishing vessels).

       For the second approach, mean concentrations for both dissolved and total forms of the
heavy metals (cadmium, chromium, copper, lead, nickel, zinc) were compared using an exact
binomial test. This approach assumes that, even if the difference in mean concentrations between
nonfishing and fishing vessels  for any given metal is not statistically significant, if the mean
metal concentrations from a particular vessel class are always or nearly always lower than those
of another class of vessels, then the overall trend may be statistically significant. Both dissolved
and total  metals concentrations of all six heavy metals were higher in nonfishing vessel
discharges than in fishing vessel discharges (see Table 3.3.5). A binomial test was then
                                           149

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

performed to determine whether the overall pattern of lower mean metal concentrations in
fishing vessel discharges could be attributed to chance, assuming an equal likelihood that
concentrations in fishing vessels or industrial vessels would be lower. The probability that mean
concentrations of all six metals (either dissolved or total) would be lower in fishing vessels
compared to nonfishing vessels, given an equal likelihood of either outcome occurring, was
statistically significant (P = 0.016). The probability of concentrations being lower in fishing
vessels for all 12 comparisons (six dissolved metals + six total metals) was also statistically
significant (P = 0.0002). The mean concentrations of these heavy metals by vessel class are
shown in Table3.3.5. Results of this analysis support the assertion that metals from deck
washdown discharges from nonfishing (utility, service, and passenger) vessels tend to be higher
than metals from deck washdown discharges from fishing vessels for each discharge event.

       One possible explanation for the higher metal concentrations in nonfishing vessels is that
the frequent washing of fishing vessels' decks may prevent metal build-up and keep metal
concentrations lower in each individual deck washdown discharge.

       With regard to assessing potential risk,  it is important to understand that, for most of the
metals identified above as of potential concern in deck washdown water, maximum metal
concentrations in the ambient or potable water used for deck washdown (see left two thirds of
Table 3.3.3) were higher than the median metal concentrations in deck washdown water samples
(Table 3.3.1). The ambient receiving waters  to which these deck washdown waters are being
discharged have metal concentrations that often exceed the most stringent NRWQC (see far right
column of Table 3.3.3). The relatively high metals concentrations for four dissolved metals
(copper, manganese, nickel, zinc) in potable water and four total metals (aluminum, arsenic, iron,
lead) in ambient water can at least partially account for the high concentrations of metals found
in some of the deck wash discharges.  Furthermore, based on corresponding concentrations of the
major seawater cations (calcium, magnesium, sodium, and potassium) in the deck washdown
water samples (see Table 3.3.1), few,  if any, of the potentially toxic dissolved metal
concentrations are likely to be bioavailable to biological organisms because of the high hardness
values, which reduce metal bioavailability.

       In summary, metals were frequently  detected  in deck washdown water samples, with
certain metals occurring much more frequently at levels that may have potential for risk than
others. EPA found high concentrations of a number of dissolved and total metals in these
samples. Dissolved cadmium, copper, lead, nickel and zinc were consistently elevated above the
most conservative screening benchmarks, with all the PHQ values in the 1 to 100 range.
However, dissolved cadmium concentrations measured in deck washdown water samples were
only detected in two of 31 vessels. For these and other metals (total aluminum, arsenic, iron, and
manganese), concentrations measured in most if not all of the water samples exceeded saltwater
and/or freshwater criteria, however they generally did not exceed concentrations in the ambient
                                           150

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


or potable water used for washdown, and would generally not be bioavailable to organisms in
seawater.
                                             151

-------
                                                          Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.1. Results of Deck Washdown/Runoff Sample Analyses for Metals1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum, Dissolved
Aluminum, Total
Antimony, Dissolved
Antimony, Total
Arsenic, Dissolved3
Arsenic, Total3
Barium, Dissolved
Barium, Total
Cadmium, Dissolved
Cadmium, Total
Chromium, Dissolved
Chromium, Total
Cobalt, Dissolved
Cobalt, Total
Copper, Dissolved
Copper, Total
Iron, Dissolved
Iron, Total
Lead, Dissolved
Lead, Total
Manganese, Dissolved
Manganese, Total
Nickel, Dissolved
Nickel, Total
Selenium, Dissolved4
Selenium, Total4
Thallium, Dissolved
Vanadium, Dissolved
Vanadium, Total
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
31
31
19
19
31
31
19
19
31
31
31
31
19
19
31
31
19
19
31
31
31
31
31
31
31
31
19
19
19
28
30
9
13
19
23
19
19
2
5
17
29
12
18
29
31
12
18
15
30
29
28
19
25
17
12
1
14
16
90
97
47
68
61
74
100
100
6
16
55
94
63
95
94
100
63
95
48
97
94
90
61
81
55
39
5
74
84
420
3400
7.3
26
6.4
18
63
270
1.3
2.0
5.1
34
2.7
6.0
42
130
520
4400
6.0
48
60
210
6.9
16
8.9
9.5
0.64
1.9
9.8
260
1900

1.9
2.3
8.3
42
100


2.3
24
1.3
4.1
23
110
190
2300

23
35
98
4.8
12
1.1


1.3
6.2






23
52







6.4













1.7
820




27
59



3.1

1.1
5.6
12

950

3.6
2.7
4.3







31
990




33
70



8.3

2.0
7.2
47

1700

8.0
11
55

6.2




2.9
570
4700
4.2
29
9.8
29
69
160


9.1
55
3.9
6.7
59
160
1100
5300
4.7
42
91
300
8.2
18
2.1
1.8

2.0
9.8
1100
8300
13
86
13
49
96
1300

1.7
16
84
8.2
20
120
340
1100
13000
19
160
200
540
13
27
25
23

5.2
20
1900
13000
91
260
28
83
280
1400
22
36
18
130
14
26
200
530
3000
15000
54
260
240
1300
17
100
82
96
3.2
7.6
58
NA
87
NA
5.6
36
0.018
NA
1000
0.25
NA
11
NA
NA
NA
3.1
1300
NA
300
2.5
NA
NA
100
8.2
610
5.0
170
NA
NA
NA
                                                            152

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 Table 3.3.1. Results of Deck Washdown/Runoff Sample Analyses for Metals1
Analyte
Zinc, Dissolved
Zinc, Total
Units
ug/L
ug/L
No.
samples
31
31
No.
detected
31
31
Detected
Proportion
(%)
100
100
Average
Cone.
260
580
Median
Cone.
120
330
Minimum
Cone.
16
20
10%
35
52
25%
51
150
75%
430
720
90%
620
1400
Maximum
Cone.
1200
4000
Screening
BM2
81
7400
Cationic Metals
Calcium, Dissolved
Calcium, Total
Magnesium, Dissolved
Magnesium, Total
Potassium, Dissolved
Potassium, Total
Sodium, Dissolved
Sodium, Total
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
31
31
31
31
19
19
19
19
31
31
31
31
19
19
19
19
100
100
100
100
100
100
100
100
73000
77000
1 30000
1 30000
30000
30000
510000
510000
34000
39000
14000
19000
8000
8100
79000
78000
5900
7300
6600
6800
3300
3600
26000
24000
25000
27000
7000
7800
4000
3900
38000
38100
32000
34000
7900
9200
5400
5600
45000
45000
83000
88000
59000
59000
24000
25000
410000
400000
1 90000
1 90000
510000
510000
1 40000
1 30000
2800000
2600000
320000
310000
1000000
1000000
1 80000
1 80000
3600000
3600000
NA
NA
NA
NA
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) See footnote 24.
(4) See footnote 25.
                                                                            153

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.2. Dissolved-to-Total Metal Ratios (fds) in Paired Deck Washdown/Runoff
Samples
Metal
Aluminum
Iron
Lead
Chromium
Vanadium
Manganese
Antimony
Copper
Arsenic
Cadmium
Nickel
Cobalt
Zinc
Selenium
Summary Statistics of DissolveckTotal
Metal Ratios Calculated for Select Metals
Geomean
0.10
0.12
0.14
0.16
0.25
0.25
0.30
0.33
0.38
0.48
0.50
0.53
0.54
0.89
Median
0.12
0.090
0.18
0.13
0.26
0.28
0.34
0.34
0.47
0.49
0.53
0.52
0.54
0.89
Min
0.010
0.050
0.030
0.060
0.12
0.010
0.14
0.050
0.060
0.36
0.17
0.26
0.18
0.61
Max
1.00
0.33
0.62
0.76
1.13
0.93
0.64
1.04
0.93
0.62
0.93
1.25
2.95
1.30
 Table 3.3.3. Minimum and Maximum Dissolved and Total Metal Concentrations in Vessel
 Source1 and Ambient2 (Harbor) Water Relative to Median Sample Concentrations and
 Most Stringent Screening Benchmarks
Metal
Aluminum, Dissolved
Aluminum, Total
Arsenic, Dissolved
Arsenic, Total
Copper, Dissolved
Copper, Total
Iron, Dissolved
Iron, Total
Lead, Total
Manganese,
Dissolved
Manganese, Total
Nickel, Dissolved
Nickel, Total
Selenium, Dissolved
Source
Water
Cone.
(min)
6.3
8.6
0
0
2.4
2.6
0
0
1.2
0
3.6
0
0
0
Source
Water
Cone.
(max)
310
250
1.9
1.8
36
51
0
801
6
33
37
3
2.7
1.6
N
6
6
3
3
5
4
1
4
2
6
6
4
4
3
Median
Cone.
From
Table
3.3.1
258
1900
2.3
8.3
23.1
109
189.5
2330
23
34.8
97.8
4.8
12
1.1
Ambient
Cone.
(min)
0
44.5
2
2.9
0
0
226
114
0
0
8.3
2.3
2.4
1.7
Ambient
Cone.
(max)
870
3950
26
28.9
24.2
23.3
259
4180
3.1
106
165
7.2
16.7
75.53
N
12
15
8
8
10
11
2
8
3
11
13
10
11
8
Most
Stringent
Screening
BM
NA
87
36
0.018
3.1
1300
NA
300
2.5**
NA
100
8.2
610
5
                                         154

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 Table 3.3.3. Minimum and Maximum Dissolved and Total Metal Concentrations in Vessel
 Source1 and Ambient2 (Harbor) Water Relative to Median Sample Concentrations and
 Most Stringent Screening Benchmarks
Metal
Selenium, Total
Zinc, Dissolved
Zinc, Total
Source
Water
Cone.
(min)
0
4.1
4.1
Source
Water
Cone.
(max)
1.9
1200
1100
N
2
6
6
Median
Cone.
From
Table
3.3.1
0
124
331
Ambient
Cone.
(min)
19.43
0
0
Ambient
Cone.
(max)
86. 53
116
23.9
N
6
13
15
Most
Stringent
Screening
BM
170
81
7400
N = sample size.
(1) Ambient water was collected from background water surrounding the vessel sampled.
(2) Source water was collected from the city tap water supply while pierside, except for one tugboat in
    Havre De Grace, Maryland, where source water was collected from a potable water storage tank on
    the vessel (service water) that was filled with city water.
(3) As discussed in footnote 25, EPA suspects positive interference may have resulted in these high
    measured concentrations of total and dissolved selenium detected in ambient deck wash samples.
Table 3.3.4. Comparison of Metal Concentrations in Deck Washdown Discharge Between
Fishing Vessels and Non-Fishing Vessels1
Metal
Copper
Nickel
Zinc
Arsenic^
Lead
Form
Dissolved
Dissolved
Dissolved
Total
Total
Average Metal Concentration (ug/L) by
Vessel Type
Fishing
27.7
6.19
161
14.0
5.48
Non-Fishing
50.7
7.23
314
20.5
70.7
Welch's Modified
2 -Sam pie t-Test
t
-1.68
-1.05
-2.15
-0.49
-3.76
df
18.2
20.7
14.9
19.8
19.1
P<|to/2|
0.110
0.306
0.049
0.629
0.001
Notes:
(1) Nonfishing vessels defined as tow/salvage vessels, tugboats, tour vessels, fire boat, water taxis, and
supply vessels. The recreational vessel is not a study vessel and was excluded from these analyses.
(2) Total arsenic concentrations discharged from the six tow/salvage boats were significantly higher than
for the other 24 vessels (Welch's Modified 2-Sample t-test, t=-5.26, P<0.001, on 16.7 df).
                                            155

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.5. Mean Concentrations of Dissolved and Total Heavy Metals from Deck Wash
Discharges from Fishing Vessels and Nonfishing Vessels1'2
Metal
Cadmium
Chromium
Copper
Lead
Nickel
Zinc

Cadmium
Chromium
Copper
Lead
Nickel
Zinc
Form
Dissolved
Dissolved
Dissolved
Dissolved
Dissolved
Dissolved

Total
Total
Total
Total
Total
Total
Cone.
Fishing
Vessels
0.750
3.79
27.7
2.00
6.19
161

2.00
15.7
93.2
5.48
8.65
207
Mg/L
n
10
10
10
10
10
10

10
10
10
10
10
10
in
Not Det.
(%)
100
70
0
100
40
0

100
20
0
10
40
0
Cone. (H9/L) in
Non-Fishing
Vessels
1.86
5.93
50.7
8.85
7.23
314

3.77
42.3
157
70.7
19.4
791
n
20
20
20
20
20
20

20
20
20
20
20
20
Not Det.
(%)
90
35
0
45
40
0

90
0
0
0
10
0
Notes:
(1) Nonfishing vessels defined as tow/salvage vessels, tugboats, tour vessels, fire boat, water taxis, and
supply vessels. The recreational vessel is not a study vessel and was excluded from these analyses.
(2) For these comparisons, minimum concentrations were set at 1/4 of the reporting limit of the highest
minimum detection level, when multiple detection limits were present. Average concentrations of
dissolved and total forms of all six heavy  metals were lower in fishing vessels than in nonfishing vessels.
                                             156

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

1000
„ 	 v
	 1
15)
^ 100
o
*T~:
CD
-i— •
C
CD
g 10
O
0


1

.
— ;
-
-•
—
i
-
- ;

—
-
-
_ *
-
—
:















L
I



*








0
•
r
















i






•

-i-
e
I

-8-
•
|
















I


*
]
a

T







i
i






©






]
T
I







1
',
I

g

r
I
















i










>
i

i-
r
















\














\

i




•














>
j >







s
•






<









1



I
I

A-
I






|
1



^
)


*

0
j

*

i
i



fe
^





/~.
0
j
-.
T 1














-
r
f
I _
—
_
h -

i —
J I
—
-

—
-
-
_
-
—
| =
                                 Dissolved Metals
Figure 3.3.1. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
Samples of Deck Washdown Water
(Note: As discussed in footnotes 24 and 25, EPA suspects positive interfence may have resulted in elevated
measured concentrations for a limited number of deck wash samples analyzed for dissolved arsenic and dissolved
selenium).
                                            157

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       10000
  05
  ^

  C
  o
 "-i—•
  03
  CD
  O
  C
  o
 O
        1000
100
  10

                                              ©
*
          JL
              #  ©
                                   Total Metals
Figure 3.3.2. Box and Dot Density Plot of Total Metals Concentrations Measured in

Samples of Deck Washdown Water

(Note: As discussed in footnotes 24 and 25, EPA suspects positive interfence may have resulted in elevated
measured concentrations for a limited number of deck wash samples analyzed for total arsenic and dissolved
selenium).
                                          158

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     CD
    -|   10.0

    o
     CD
     a     1.0
     CD
           0.1
                                 Ti   r
          \   \    \   \
T:
                    I
I   I    I   I   I    I   I
1
                                 Dissolved Metals
Figure 3.3.3. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals
in Samples of Deck Washdown Water
(Note: Replacement values for non-detects are circled. Also, as discussed in footnotes 24 and 25, EPA suspects
positive interfence may have resulted in elevated measured concentrations for a limited number of deck wash
samples analyzed for dissolved arsenic and dissolved selenium).
                                            159

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 0
"-I— •
 O
 ^

O
 03
 N
 03
 o
Q_
     1000.00
100.00
         10.00
           1.00
          0.10
          0.01
                       j	L
                          j	L
J	I
*   I
                 s^^^^^^^^^
                                   Total  Metals
 Figure 3.3.4. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in

 Samples of Deck Washdown Water
 (Note: Replacement values for non-detects are circled. Also, as discussed in footnotes 24 and 25, EPA suspects
 positive interfence may have resulted in elevated measured concentrations for a limited number of deck wash
 samples analyzed for total arsenic and dissolved selenium).
                                           160

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.3.2    Classical Pollutants

       Deck washdown water samples from 32 vessels were analyzed for 14 classical pollutants
(see Table 3.3.6). The classical pollutants include measurements that are physical properties
(temperature, conductivity, salinity, turbidity, TSS), oxygen consumption (BOD, COD), oil and
grease (HEM and SGT-HEM), as well as chemical concentrations (pH, sulfide, DO, and TRC).

       Measured values of the physical properties of the discharge (conductivity, dissolved
oxygen, pH, temperature, salinity) are unremarkable and appear to reflect conditions at the time
(seasonality) and location (geographical) of sampling. For instance, conductivity and salinity in
deck washdown water appear to reflect the type of source water used (ambient or potable
service/city26 water), as shown by the measured values of these two parameters. Half the fishing
vessels appear to have used ambient saltwater during normal operations (six of 11 vessels), while
the remaining fishing boats and nearly all other vessel types (tugs, tow/salvage, tour, supply
boats) used a freshwater source (aboard the vessel or pierside). Levels of pH were generally
about neutral (between 7 and 8), with the only exceptions being two tugboats where the pH was
9.1 and 9.8 (relatively high). Temperature of the deck washdown water ranged from 7.5 to 32 °C
and varied according to month (season) sampled and geographic location (warmer water samples
in southern United States and colder in mid-Atlantic and northern states). Dissolved oxygen
(DO) in deck washdown samples was sufficiently saturated (50 percent plus; DO ranged from
5.5 to 10.5 mg/L) in all samples, except for low DO concentrations from three fishing vessels
participating in the north Pacific fishery, which ranged from 1.6 to 1.9 mg/L.

       Figure 3.3.5 illustrates the variability of the values measured for the classical pollutants
in deck washdown water. Turbidity (measure of water clarity) and TSS are clearly related and
range from 4.1 to 460 NTU and 31  to 530 mg/L, respectively. Measured values above the
median concentrations were dominated by the tug, tow/salvage, supply, fire, and water taxi boats
for both parameters, while measured values below the median were largely from the fishing
boats (with only a few exceptions). EPA notes that the majority of deck washdown samples from
fishing vessels were taken while the vessel was shoreside, and do not reflect constituents of deck
washdown while the  vessel is engaged in fishing operations. Potable water measured during the
study was low in turbidity  (0 to 16 NTU) and TSS (0 mg/L), as  was ambient (harbor) water,
except for waters sampled in the Gulf Coast (Louisiana). Ambient turbidity and  TSS were as
high as 186 NTU and 98 mg/L, respectively, in a sample collected from one harbor in Louisiana.

       Of the remaining parameters, BOD, COD, and TOC have quite high concentrations (see
Figure 3.3.5). While the measured values for these parameters in deck washdown water samples
were generally evenly distributed for the different vessel types across the entire concentration
26 Service water here means the vessel potable water supply. For study vessels, vessel service water generally
originates from municipal water supply rather than produced on board. When deck washdown is performed pierside
most vessels used city water as their source water. Many fishing vessels and at least one tugboat use ambient water
as their water source when performing deck washdown offshore or underway.

                                           161

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


range, a select few vessels were clear standouts: three tugboats, a fishing (shrimping) vessel, and
the supply boat. The concentrations of all three parameters were highly variable and span two
orders of magnitude. In contrast, measured sulfide concentrations from deck washdown water
samples collected from two fishing boats and a tow/salvage boat were  all relatively low, but,
when compared to the most stringent NRWQC of 0.002 mg/L, had PHQs ranging from 2.5 to 8.5
(moderate exceedance  - data not shown).

       PHQs were calculated for three additional classical pollutants for which benchmarks
were available and are  shown in Figure 3.3.6. As the figure shows, the TRC concentrations
where TRC was detected above the reporting limit of 0.10 mg/L greatly exceeded the benchmark
(most stringent NRWQC of 0.0075 mg/L, the saltwater chronic aquatic life criterion) by factors
that ranged from 23 (tow/salvage vessel) to 106 (exceedance by 2  orders of magnitude - a
fishing vessel). These concentrations (ranging from 0.17 to 0.80 mg/L) were measured in deck
washdown water samples collected from three (of the 11) fishing vessels, the two tour boats, a
tugboat, and the tow/salvage boat. It is worth pointing out that in one instance (i.e., for a tugboat
with a high TRC concentration of 0.39 mg/L), the measured TRC concentration in the source
(potable) water was 0.70 mg/L. It is also worth noting that only one of 11 respondents (a fishing
vessel) indicated using chlorine bleach while washing decks, and this particular vessel had a
measured TRC concentration in the deck washdown sample of 0.38 mg/L and a PHQ of 51.

       TSS in most of the deck washdown water samples collected exceeded the secondary
treatment effluent limitation benchmark of 30 mg/L. However, 27 of 32 PHQs calculated for
these samples were below 10 (Figure 3.3.6), and all five TSS samples with PHQs>10 (max PHQ
= 17.7) were associated with tugboats. As discussed above, in the one potable water sample for
which TSS was measured, TSS was not detected.

       BOD was measured in 22 deck washdown water samples that exceeded EPA's secondary
treatment effluent limit of 30 mg/L (Figure 3.3.6). As indicated above, the vessels with the
highest level of exceedance (PHQs > 5) were associated with three tugboats, a fishing
(shrimping) vessel, and the supply boat.

       EPA compared HEM and SGT-HEM concentrations measured in deck washdown
samples to the existing international  and U.S. regulatory limit of 15 ppm (15 mg/L) for oil and
grease discharge. HEM and SGT-HEM were detected in all of the deck washdown water
samples, with concentrations ranging from 1.2 to a very high 133 mg/L for HEM and 0.91 to a
comparably high 84 mg/L for SGT-HEM. Based on the regulatory limit of 15 mg/L, PHQs
exceeded one in only six of 29 vessels sampled for HEM and two in the 29 vessels sampled for
SGT-HEM. The highest PHQs for both parameters corresponded with  the supply boat and a
tugboat, with PHQs of 4.7 and 8.9 for HEM  and 1.2 and 5.6 for SGT-HEM, respectively. Note,
oil and grease were not detected in the two potable water samples collected in this sampling
program.
                                          162

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       To summarize, just under a third of the vessels sampled had concentrations of TRC in
deck washdown samples above the reporting limit of 0.10 mg/L. Of these seven samples, the
measured TRC concentrations (as high as 0.80 mg/L) that exceeded the screening benchmark
were not associated with any one particular class of vessel. For TSS, however, one vessel type
(tugboats) had the highest number of exceedances. The elevated TSS in deck washdown water
samples from tugboats may be caused by a less frequent washdown on these vessels compared
with vessels such as fishing vessels. Just over two-thirds of vessels (22 out of 32) exceeded the
most stringent screening benchmark for BOD; however, as in the case with TRC, no one
particular class of vessels had a higher number of exceedances than other classes.

       Oil and grease are generally not of concern for this type of discharge, nor are any of the
other physical parameters that were measured (conductivity, dissolved oxygen, pH, temperature,
salinity). TOC was detected in all samples ranging from 3.6 to a very high 350 mg/L (one of the
tugboats with high HEM). Organic carbon strongly complexes metals in both freshwater and
saltwater matrices, and like the competing cations such as calcium and magnesium, renders
dissolved metals less bioavailable and less likely to be rapidly available for biological organisms.
                                          163

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.6. Results of Deck Washdown Water Sample Analyses for Classical Pollutants1
Analyte
Biochemical Oxygen Demand
(BOD)
Chemical Oxygen Demand
(COD)
Conductivity
Dissolved Oxygen
Hexane Extractable Material
(HEM)
PH
Salinity
Silica Gel Treated HEM
(SGT-HEM)
Sulfide
Temperature
Total Organic Carbon (TOC)
Total Residual Chlorine
Total Suspended Solids
(TSS)
Turbidity
Units
mg/L
mg/L
mS/cm
mg/L
mg/L
SU
ppt
mg/L
mg/L
C
mg/L
mg/L
mg/L
NTU
No.
samples
32
32
26
26
29
30
24
29
3
31
25
31
32
31
No.
detected
30
32
26
26
26
30
24
22
2
31
25
7
32
31
Detected
Proportion (%)
94
100
100
100
90
100
100
76
67
100
100
23
100
100
Average
Cone.
110
390
7.7
7.2
14
7.7
4.9
7.0
0.011
21
44
0.12
170
150
Median
Cone.
56
160
1.0
7.7
2.8
7.6
0.60
1.7
0.011
21
24

120
110
Minimum
Cone.

24
0.24
1.6

7.0
0.10


7.5
3.6

27
4.1
10%
4.7
52
0.37
1.8

7.0
0.20


9.0
5.0

31
36
25%
14
90
0.50
6.3
1.7
7.2
0.23
0.45

13
7.1

59
58
75%
92
570
13
8.9
12
7.9
8.0
3.8
0.017
29
52

250
190
90%
370
1200
30
9.7
39
8.5
21
13
0.017
31
96
0.37
470
380
Maximum
Cone.
830
1800
47
11
130
9.8
28
84
0.017
32
350
0.80
530
463
Screening
BM2
30
NA
NA
NA
15
NA
NA
15
NA
NA
NA
0.0075
30
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            164

-------
                   Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 CD
_^

 03
O

~03
 CD
 O

 O
O
    1000.00
      100.00
       10.00
         1.00
        0.10
        0.01
                 1   1   1   1
                                         1   1   1   1   1   1  =
                ^ C\^ «^ 
-------
   0

  "5
   ^
  a
   03
   N
   03
   o
  Q_
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       100.0
10.0
           1.0
           0.1
                   i    i   i    r
                                          \    \    \   \    \
                   I    I   I    I
                                       I   I    I    I   I    I   =
                  c\!^ c\^ ^ c\^ c\!^ c\^ O^ c\!^ c\^> c\!^ c\^> ^C ^ ^ t
                o°
                             Classical  Pollutants
Figure 3.3.6. Box and Dot Density Plot of Potential Hazard Quotients for Classical
Pollutants in Samples of Deck Washdown Water
(Note: Replacement values for non-detects are circled. Also, PHQs for sulfide are not shown in the figure, but are
mentioned in the text).
                                         166

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


     3.2.3.3    Pathogen Indicators (Microbiological)

       Selected deck washdown water samples were analyzed for the pathogen indicator
bacteria E. coli, enterococci, and fecal coliform. Sampling for pathogens was limited to fishing
vessels since EPA could not identify likely potential sources of pathogens in deck washdown
discharges on other vessel types. EPA targeted select fishing vessels to attain the best cross-
representation possible based on available funding and proximity to qualified subcontractor
laboratories to meet sample hold times (< 6 hours). The types of fishing vessels sampled
included three shrimping (trawler) boats in Louisiana, two ground fishery trawlers in
Massachusetts, and a gillnetter boat in Alaska. All vessels indicated that their decks are washed
frequently throughout the day (after or between catches, after unloading, etc.), and while pierside
and underway (nearshore and farshore). Table 3.3.7 summarizes the analytical results.
Concentrations were determined for each pathogen using the same (E. coli, enterococci) or
comparable methods (fecal coliform).

       Figure 3.3.7  shows the variability of the values measured for the pathogens in deck
washdown water samples from the various fishing vessels. Measured concentrations of E.  coli
range from 20 MPN/100 ml for one of the shrimping trawlers to 8,336 MPN/100 ml for one of
the ground fishery trawlers in Massachusetts. It should be noted, however, that the water the
ground fishery trawler used for desk washing was ambient (harbor) water receiving stormwater
and combined sewer overflow from a storm event. The measured concentration of E. coli in the
ambient water at that location was 24,200 MPN/100 ml. Excluding this outlier, the concentration
of E. coli from only  one vessel (shrimper;  concentration = 650 MPN/100 ml) exceeded EPA's
most stringent freshwater bathing NRWQC of 126 MPN/100 ml by more than a factor of five
(PHQ = 5.1),  as illustrated in Figure 3.3.8. EPA collected multiple samples from another
shrimping vessel in Louisiana to measure E. coli in prefishing deck washdown water, postfishing
water, without catch rinse water, and with catch rinse water. For this vessel, E.  coli
concentrations ranged from a low of 10 (prefishing sample) to a  high of only 50 MPN/100 ml
(without catch rinse). The concentrations ofE. coli were largely  unaffected by either the addition
of catch to the vessel (as E. coli concentrations in prefishing and postfishing deck washdown
samples were similar) or the process of rinsing the catch while on deck.

       The enterococci values measured in a deck washdown water samples ranged from  1.5 to
1,300 MPN/100 ml, and  follow the same general pattern as E. coli (Figure 3.3.7). Excluding the
previously described example of the trawler in Massachusetts, which was directly influenced by
high levels of enterococci in the ambient water resulting from storm-related combined sewage
overflow (5,100 MPN/100  ml), the deck washdown water samples from two vessels (both
shrimpers; concentrations = 637 and 914 MPN/100 ml, respectively)  exceeded EPA's most
stringent bathing NRWQC for enterococci of 33 MPN/100 ml (freshwater) and 35 MPN/100 ml
(saltwater) by factors of nearly 20 and 30 respectively (Figure 3.3.8). In contrast to E. coli,
however, analysis of the  multiple samples collected for enterococci in prefishing deck washdown

                                           167

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

water (540 MPN/100 ml), postfishing water (8 MPN/100 ml), without catch rinse (1,200
MPN/100 ml), and with catch rinse (801 MPN/100 ml) for the shrimping vessel in Louisiana
indicate that their deck washing process appeared to reduce the presence of the pathogen in deck
washdown discharge.

       The concentrations of fecal coliform bacteria measured in a deck washdown water
samples are all substantial (ranging from 91  to 600 CFU/100 ml27), except for the very low
concentration of 0.75 CFU/100 ml for the gillnetting vessel in Alaska (Figure 3.3.7). The
associated PHQs for fecal coliform range from 0.05 (gillnetter) to 43 (one of the shrimping
boats), as illustrated in Figure 3.3.8. The PHQs for this pathogen are based on the NRWQC of 14
MPN/100 ml for shellfish harvesting. As with enterococci, the multiple samples measured for
fecal coliform bacteria in prefishing deck washdown water (0 CFU/100 ml), postfishing water (6
CFU/100 ml), without catch rinse (1,630 CFU/100 ml), and with catch rinse (620 CFU/100 ml)
for the shrimping vessel indicate that their deck washing process did not increase (and seemed to
reduce) the presence of this pathogen in deck washdown discharge. The single potable water
sample taken while onboard a shrimping vessel pierside in Louisiana was free of all pathogens.

       The data collected for this study show that, while the three groups of pathogens are
present in deck washdown discharge samples from commercial fishing vessels, concentrations
are variable, and the source of the water used for deck washdown can greatly influence the
background bacteria levels. Of the three pathogen groups, fecal coliform are present at
concentrations exceeding EPA's most stringent criteria more often than enterococci andE. coli,
in that order.
27 Excluding the outlier value of 8,050 CFU/ml from the ground fishery trawler in Maine influenced by the storm
event.

                                           168

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.7. Results of Deck Washdown Water Sample Analyses for Pathogen Indicators1
Analyte2
£. Co// by MPN
Enterococci by MPN
Fecal Coliform by MF
Units3
MPN/100ml
MPN/100ml
CFU/100ml
No.
samples
5
5
6
No.
detected
5
5
6
Detected
Proportion
(%)
100
100
100
Average
Cone.
1900
580
1600
Median
Cone.
160
640
560
Minimum
Cone.
20
1.5
0.75
10%
20
1.5
0.75
25%
62
27
68
75%
4500
1100
2500
90%
8300
1300
8100
Maximum
Cone.
8300
1300
8100
Screening
BM4
130
33
14
Notes:
 (1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage  of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) MPN = Most Probable Number; MF = Membrane Filtration.
(3) CFU = Colony Forming Units.
(4) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                           169

-------
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       1000
   o
   o
   CD
   Q_
   O
  O
         100
10
                          t7
                                      -e-
                                       ©


                              <&*-        <<&
                           Pathogen Indicators
Figure 3.3.7 Box and Dot Density Plot of Pathogen Indicator Concentrations Measured in
Samples of Deck Washdown Water
(Note: Corresponding units are MPN/100 ml for E. coli and enterococci, and CFU/100 ml for fecal coliform).
                                       170

-------
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

£ 100.0
^_
_
^J
0

cr
E 10.0
/T*
TO
N
CD
^
CD
j=; ^ r\
±± \ .u

•i— •
o
Q_

0.1


-

:
-



—
_
-

_


-
-
-


—
-



6






c
V.

r
v.
r
v


^


CiC


5






~^
j

>
j
-\
j


5


NN

























x-
§
V.
r
V.




r
^







(
.^0
-

I
GO

V
^ O -
v ^
—
n -
w

A
j

—
-
-


—
& f -
^*\ .to
kj ^VV'^
                                  \0          A^
                             0^        ^e°

                          Pathogen  Indicators
Figure 3.3.8. Box and Dot Density Plot of Potential Hazard Quotients for Pathogens in
Samples of Deck Washdown Water
                                      171

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     3.2.3.4    Nutrients

       Deck washdown discharge was also characterized for nutrient levels. Nutrient pollution,
including nitrogen, phosphorus, and numerous micronutrients, is a component of certain vessel
discharges and a major source of water quality degradation throughout the United States (USGS,
1999). Deck washdown discharges from all vessel types were expected to contain potentially
high levels of phosphorus because of the wide-spread use of detergents for deck cleansing. Deck
washdown discharges from commercial fishing vessels were also expected to contain potentially
elevated ammonia concentrations for the same reason, as well as from biological wastes from
fish and shellfish catch. In addition to total phosphorus and total ammonia (as nitrogen), deck
washdown water samples were also analyzed for nitrate/nitrite nitrogen (inorganic nitrogen) and
TKN, the sum of organic nitrogen (including toxic ammonia nitrogen) (see Table 3.3.8).

       Concentrations of nitrate/nitrite nitrogen in deckwash discharge samples range from
0.025 to 6.5 mg/L (see Figure 3.3.9). An interesting note is that the deck washdown water
samples for commercial fishing vessels of all types did not exceed 0.50 mg/L while all other
vessels exceeded this value. The five highest nitrate/nitrite concentrations (ranging from 2.5 to
6.5 mg/L) were analyzed in samples from three tugs and two tow/salvage vessels.  It is important
to note, however, that most samples of deck washdown on fishing vessels were collected
onboard fishing vessels pierside and not when fishing activity was occurring. In the two cases
where deck washdown  samples were collected where fishing activities were taking place, the
samples were collected towards the end of the deck washdown activity and may not have
captured potentially higher levels of nitrate/nitrite from biological wastes.

       The concentrations determined for TKN (sum of organic nitrogen) show the
concentration range spans two orders of magnitude, from 0.05 to 40 mg/L (see Figure 3.3.9). In
contrast to the nitrate/nitrite samples, the TKN concentrations from all vessels were evenly
distributed across the entire concentration range. The two highest TKN concentrations (by more
than a factor of two) correspond to a trolling vessel and a tugboat, with TKN concentrations of
28 and 40 mg/L, respectively.

       Ammonia is the only  nutrient form for which there are currently numeric NRWQC
established to protect against its toxic effects. Only five of 31 vessels contained  ammonia in deck
washdown water samples slightly above (1.2 to 1.8 mg/L ammonia as nitrogen)  the most
stringent 2006 NRWQC of 1.2 mg/L, the freshwater chronic aquatic  life criterion for total
ammonia as nitrogen (see Figure 3.3.10). These values correspond with deck washdown water
samples collected from two tow/salvage boats, two fishing vessels, and the recreational  vessel.

       The benchmark for total phosphorus of 0.1 mg/L from the 1986 EPA Gold Book was
exceeded in samples collected from all but one of the 31 vessels. The highest total phosphorus
concentration of 22 mg/L from a tugboat exceeded the benchmark by a factor of 220 (see Figure

                                           172

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

3.3.10). This concentration was 6.5 times higher then the next highest measured concentration of
3.4 mg/L from a trolling vessel. The deck washdown water samples for phosphorus from all
vessels were generally evenly distributed across the entire concentration range.

       Total ammonia in ambient and service water ranged from below detection to 0.93 mg/L
and from below detection to 0.73 mg/L, respectively (all below the most stringent 2006 NRWQC
of 1.24 mg/L). Total phosphorus in ambient and service water ranged from below detection to
2.0 mg/L and from below detection to 0.52 mg/L, respectively (compared to 0.1 mg/L from the
1976 EPA Red Book).

       In summary, out of the four nutrient parameters, only total phosphorus is likely a
potential concern from deck washdown effluent. Twelve of the  19 respondents confirmed using
standard liquid detergents aboard their vessels for deck washing, the expected source of total
phosphorus in deck washdown discharges. However, ambient and domestic service water are
also likely sources  of phosphorus in a meaningful percentage of instances.
                                           173

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.8. Results of Deck Washdown Water Sample Analyses for Nutrients1
Analyte
Ammonia As Nitrogen (NH3-N)
Nitrate/Nitrite (NO3/NO2-N)
Total Kjeldahl Nitrogen (TKN)
Total Phosphorus
Units
mg/L
mg/L
mg/L
mg/L
No.
samples
31
32
31
31
No.
detected
31
27
30
31
Detected
Proportion (%)
100
84
97
100
Average
Cone.
0.53
1.4
6.0
1.7
Median
Cone.
0.32
1.5
3.6
0.79
Minimum
Cone.
0.058


0.060
10%
0.074

1.4
0.15
25%
0.10
0.16
1.8
0.44
75%
0.81
1.9
6.6
1.6
90%
1.5
2.7
11
2.9
Maximum
Cone.
1.8
6.5
40
22
Screening
BM2
1.2
NA
NA
0.10
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall.  So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculated PHQs.
                                                                            174

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
         10.0r
    ^

    o
    CD
    O

    O
    O
           1.0-
          0.1  ^
1
-
-
"t oc
0
—
00
~ o
00
_ oo
— 0 (
0
	 Ot 0
- ok
1 H
A ^
oX^ .&'


0
0
0


0


F
^











^
-t
•
(
t
t
0
C
0
C
0
0

-t




£
1
*
—
_
0
™
t
3 03
3
000 ^~
00
00 -
0030
-
* :
1
\ \\
^ ^?_x3
o

                                              \°N
                                    Nutrients
Figure 3.3.9. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of

Deck Washdown Water

(Note: NH3-N=Ammonia as Nitrogen, NO3/NO2-N= Nitrate/Nitrite Nitrogen, TKN=Total Kjeldahl Nitrogen, and

Total Phosph (truncated)=Total Phosphorus).
                                         175

-------
                  Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
    100.0

c
CD
"o

^   10.0
  03
  N
  CO
         1.0
  Q_
         0.1
                                                OC
                                              oooo
                     GOTO
                     oo
                     ooo
                     o
                     oo
                     ooo
                   oo
                   oo
                   oo
                    ooc
                     QUO
                      oo
                     V"»        V
                  rj?
                              Nutrients
Figure 3.3.10. Box and Dot Density Plot of Potential Hazard Quotients for Nutrients in

Samples of Deck Washdown Water
(Note: NH3-N=Ammonia as Nitrogen, NO3/NO2-N= Nitrate/Nitrite Nitrogen, TKN=Total Kjeldahl Nitrogen, and
Total Phosph (truncated)=Total Phosphorus).
                                  176

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

     3.2.3.5    Long-Chain Nonylphenols

       Deck washdown water samples from 29 vessels were analyzed for 27 long-chain
alkylphenol ethoxylates: 16 NPEOs and 5 OPEOs (see Table 3.3.9). The NPEOs with the longest
ethoxylate chains (i.e., less degraded products (NP18EO through NP10EO)) were detected in
slightly under a third of the vessels (nine of 29), with concentrations increasing as ethoxylate
chain is reduced (i.e., concentrations increasing from NP18EO to NP10EO because the longer-
chain products found in commercial formulations are quickly degraded). The OPEO with the
longest ethoxylate chain (OP12EO) was also detected in about a third of the vessels (see Table
3.3.9).  As with NPEOs, the OPEO concentrations generally increase as the ethoxylate chain is
reduced, except that no OPEOs with ethoxylate chains smaller than OP7EO were detected
(similar to the situation in packing gland effluent; see Section 3.2.2.5).

       Of the several vessels where NPEOs were detected in the longer (NP18EO through
NP10EO) ethoxylated compounds, only three of those vessels also had detectable concentrations
of NPEOs of the shortest chain (NP3EO), albeit at very low concentrations ranging from 0.80 to
29 |ig/L. These were tow/salvage vessels, one of which confirmed using liquid detergent
(Palmolive™) for deck washing (NP3EO concentration of 29 |ig/L in deck washdown sample).
A tugboat had the only measured concentration of OP8EO in its deck washdown water sample at
a concentration of 19 |ig/L.

       Total NPEO concentrations could be calculated from summed concentrations of
individual chain lengths for five of the 29 vessels: three tow/salvage vessels and two tour boats
(see Figure 3.3.11). The concentrations of total NPEOs ranged from 30 to 8,330 |ig/L.

       As discussed in previous subsections (see Sections 3.2.1.7 (bilgewater) and 3.2.2.5
(packing gland effluents)), while there are no NRWQC for the sum of individual ethoxylate
chains  of NPEOs or OPEOs, these compounds will ultimately degrade to NP in fresh and salt
water over time under all conditions. The NRWQC for NP in salt water based on chronic toxicity
to aquatic organisms is 1.7 |ig/L. EPA is uncertain as to exactly how much NP might be
generated from the degradation of NPEO and OPEO isomers under a given harbor scenario and
water quality condition (see Section 1.6.6 of this report). However, neither total NPEO or OPEO,
nor any of the different isomers, were detected in ambient water at the locations where the
vessels were sampled.  Service water (generally city tapwater pierside) was not analyzed for long-
or short- chain nonylphenol and octylphenol ethoxylates.
                                          177

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.9. Results of Deck Washdown Water Sample Analyses for Long-Chain Nonylphenols1
Analyte
Total Nonylphenol Polyethoxylates
Nonylphenol octodecaethoxylate (NP18EO)
Nonylphenol heptadecaethoxylate (NP17EO)
Nonylphenol hexadecaethoxylate (NP16EO)
Nonylphenol pendecaethoxylate (NP15EO)
Nonylphenol tetradecaethoxylate (NP14EO)
Nonylphenol tridecaethoxylate (NP13EO)
Nonylphenol dodecaethoxylate (NP12EO)
Nonylphenol undecaethoxylate (NP1 1 EO)
Nonylphenol decaethoxylate (NP10EO)
Nonylphenol nonaethoxylate (NP9EO)
Nonylphenol octaethoxylate (NP8EO)
Nonylphenol heptaethoxylate (NP7EO)
Nonylphenol hexaethoxylate (NP6EO)
Nonylphenol pentaethoxylate (NP5EO)
Nonylphenol tetraethoxylate (NP4EO)
Nonylphenol triethoxylate (NP3EO)
Octylphenol dodecaethoxylate (OP12EO)
Octylphenol undecaethoxylate (OP1 1 EO)
Octylphenol decaethoxylate (OP10EO)
Octylphenol nonaethoxylate (OP9EO)
Octylphenol octaethoxylate (OP8EO)
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
samples
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
No.
detected
5
12
9
10
9
9
9
8
9
9
8
8
7
6
6
4
3
8
2
4
5
1
Detected
Proportion
(%)
17
41
31
34
31
31
31
28
31
31
28
28
24
21
21
14
10
28
6.9
14
17
3.4
Average
Cone.
540
1.5
3.4
7.4
14
25
44
64
86
91
88
75
61
34
19
11
4.9
1.4
1.8
3.6
3.8
10
Median
Cone.






















Minimum
Cone.






















10%






















25%






















75%

0.15
0.21
0.89
0.91
1.8
2.9
4.5
6.1
6.9
3.1
3.2
0.99




0.98




90%
1400
5.0
13
27
55
75
180
260
350
350
330
280
220
140
40
2.6
0.80
2.4

1.8
1.3

Maximum
Cone.
8300
21
41
87
160
290
480
760
1100
1300
1300
1100
950
440
270
120
30
8.8
7.8
2.1
9.6
19
Screening
BM2
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                           178

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
O)
              1000.00
               100.00
          •-S     10.00
          co
          CD
          O
          c
          O
          O
        1.00

        0.10

        0.01
                           i    i   i    i    i    i    i   i    i    i    r
                           |    I   I    I    I    I    I   I    I    I    I
                        0  2   4   6   8  10  12  14 16 18 20  22 24
                                       Nonylphenols
Figure 3.3.11. Box and Dot Density Plot of Nonylphenol Concentrations Measured in
Samples of Deck Washdown Water
Nonylphenol parameters are identified as follows:
(1) Total Nonylphenol
Polyethoxylates
(2) Nonylphenol
octodecaethoxylate (NP18EO)
(3) Nonylphenol
heptadecaethoxylate (NP17EO)
(4) Nonylphenol
hexadecaethoxylate (NP16EO)
(5) Nonylphenol
pendecaethoxylate (NP15EO)
(6) Nonylphenol
tetradecaethoxylate (NP14EO)
(7) Nonylphenol
tridecaethoxylate (NP13EO)
(8) Nonylphenol
dodecaethoxylate (NP12EO)
                        (9) Nonylphenol
                        undecaethoxylate (NP11EO)
                        (10) Nonylphenol
                        decaethoxylate (NP10EO)
                        (11) Nonylphenol
                        nonaethoxylate (NP9EO)
                        (12) Nonylphenol
                        octaethoxylate (NP8EO)
                        (13) Nonylphenol
                        heptaethoxylate (NP7EO)
                        (14) Nonylphenol
                        hexaethoxylate (NP6EO)
                        (15) Nonylphenol
                        pentaethoxylate (NP5EO)
                        (16) Nonylphenol
                        tetraethoxylate (NP4EO)
(17) Nonylphenol triethoxylate
(NP3EO)
(18) Octylphenol
dodecaethoxylate (OP12EO)
(19) Octylphenol
undecaethoxylate (OP11EO)
(20) Octylphenol decaethoxylate
(OP10EO)
(21) Octylphenol
nonaethoxylate (OP9EO)
(22) Octylphenol octaethoxylate
(OP8EO)
                                             179

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

     3.2.3.6    Volatile and Semivolatile Organic Chemicals

       VOCs and SVOCs were not targeted for deck washdown water sample collection in this
study because these compounds were not expected to be found in common deck washdown on
most vessels28. In two cases during scheduled cleanings of the decks of two tow/salvage vessels,
however, there was a noticeable oily sheen and where fuel was spilled at the fueling location
while samplers were onboard the vessels. Samples of deck washdown water were taken in these
instances and analyzed for VOCs and SVOCs (see Table 3.3.10).

       Of the 70 VOCs that were analyzed for in the two deck washdown samples, only 12 were
detected in one or more of the two samples. Of these 12 VOCs, only acetone, chloroform,  and
toluene were detected in both samples. In one sample from the vessel with the oily sheen;
acetone was detected at 20 |ig/L. Figure 3.3.12 contains all the samples that were detected, the
other five samples were detected with very low values. Benzene, ethylbenzene, and xylene
(compounds associated with fuel oil spills) were detected in one of the two samples at
surprisingly low levels. The PHQ of 13 for the benzene sample that was below detection levels
was an artifact of the relatively high reporting limit of 25|ig/L compared to the screening
benchmark of 2.2 |ig/L. PHQs for only two VOCs, dibromochloromethane and
bromodichloromethane exceeded the benchmark (see Figure 3.3.13), which were artifacts  of the
reporting limits which were as high as 25|ig/L compared to the screening benchmarks of 0.4
Hg/L and 0.55 |ig/L, respectively. Both these were formerly used as flame retardants and as an
intermediate in chemical manufacturing.

       Similarly, of the 62 SVOCs that were analyzed for in the two deck washdown samples,
only three were detected in one or more of the two samples: bis(2-ethylhexyl) phthalate,
caprolactam, and di-n-butyl phthalate (data not shown due to so few analytes detected). Levels
detected in the latter two SVOCs are unremarkable. The concentration of bis(2-ethylhexyl)
phthalate in the one sample where it was detected (i.e., the tow/salvage vessel with the oily
sheen), however, was sufficiently high (6.7 |ig/L) such that the associated PHQ, based on the
most conservative screening benchmark of 1.2 |ig/L (human health criteria), was 5.6 (data not
shown). As previously noted, bis(2-ethylhexyl) phthalate is a manufactured chemical that is
commonly added to plastics to make them flexible. Phthalates in general are known to interfere
with reproductive health and liver and kidney function in both animals and humans (Sekizawa et
  It is worth noting that solvents in cleaning agents may be used for certain activities such as above-water-line hull
cleaning. Samples associated with above-water-line hull cleaning were not collected during this study because none
of the vessels engaged in such an activity while EPA's sampling crew was aboard the vessel. During a survey
collected while onboard the vessels, however, 11 of 16 respondents confirmed that they do perform above-water-line
hull cleaning occasionally on their vessels.
                                           180

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
al., 2003; DiGangi et al., 2002). Bis(2-ethylhexyl) phthalate was not detected in the associated
ambient water sample collected at the site corresponding with the two tow/salvage vessels, but
di-n-butyl phthalate was (ambient concentration of 1.1 |ig/L).

       Di-n-butyl phthalate was the only SVOC detected in ambient water samples collected in
association with the deck washdown samples collected in the study. No VOCs were detected in
ambient samples.
                                           181

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.10. Results of Deck Washdown Water Sample Analyses for VOCs and SVOCs1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
VOCs
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
Acetone
Benzene
Bromodichloromethane
Chloroform
Dibromochloromethane
Ethylbenzene
m-,p-Xylene (sum of isomers)
O-Xylene
Toluene
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
2
2
2
2
2
2
2
2
2
2
2
1
1
2
1
1
2
1
1
1
1
2
50
50
100
50
50
100
50
50
50
50
100
13
13
13
13
13
1.3
13
13
25
25
0.65
0.30
0.090
20
0.3
1.2
1.5
0.7
0.10
0.40
0.20
0.70


5.5


1.0




0.60


5.5


1.0




0.60


5.5


1.0




0.60
0.30
0.090
20
0.3
1.2
1.5
0.70
0.10
0.40
0.20
0.70
0.30
0.090
20
0.3
1.2
1.5
0.70
0.10
0.40
0.20
0.70
0.30
0.090
20
0.3
1.2
1.5
0.70
0.10
0.40
0.20
0.70
NA
NA
NA
2.2
0.55
5.7
0.4
530
NA
NA
1300
SVOCs
Bis(2-ethylhexyl) phthalate
Caprolactam
Di-n-butyl phthalate
Naphthalene
ug/L
ug/L
ug/L
ug/L
2
2
2
2
1
2
1
1
50
100
50
50
4.7
79
2.5
13
6.7
100
2.4
0.40

56



56



56


6.7
100
2.4
0.40
6.7
100
2.4
0.40
6.7
100
2.4
0.40
1.2
NA
2000
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage  of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3,  and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            182

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
         10.0
    05
    ^

    C
    o
   "-I—•
    03
    CD
    O
    C
    o
   O
1.0
          0.1
                  -Q-
                                      •e
                                    •e-
                                 1
                                                             t
f
               0        2       4       6        8       10      12

                                       VOCs
Figure 3.3.12. Box and Dot Density Plot of Volatile Organic Chemical Concentrations

Measured in Samples of Deck Washdown Water

VOCs are identified as follows:

(1) 1,2,4-Trimethylbenzene           (5) Bromodichloromethane           (9) m-,p-Xylene (sum of

(2) 1,3,5-Trimethylbenzene           (6) Chloroform                    isomers)

(3) Acetone                       (7) Dibromochloromethane           (10) O-Xylene

(4) Benzene                       (8) Ethylbenzene                  (11) Toluene
                                           183

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 CD
"-i— •
 O
 ^
O
 03
     10.0000
       1.0000
       0.1000
       0.0100
 CD
-i—•
 o
Q-     0.0010
       0.0001
                                             MS5
                                          I

                0       2       4       6       8      10     12
                                      VOCs
 Figure 3.3.13. Box and Dot Density Plot of Potential Hazard Quotients for VOCs in
 Samples of Deck Washdown Water
 VOCs are identified as follows (replacement values for non-detects are circled):
 (1) 1,2,4-Trimethylbenzene          (5) Bromodichloromethane           (9) m-,p-Xylene (sum of
 (2) 1,3,5-Trimethylbenzene          (6) Chloroform                   isomers)
 (3) Acetone                      (7) Dibromochloromethane           (10) O-Xylene
 (4) Benzene                      (8) Ethylbenzene                  (11) Toluene
                                         184

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

     3.2.3.7    Summary of the Characterization of Deck Washdown Water

       Table 3.3.11 summarizes the specific analytes within deck washdown and runoff water
that may have the potential to pose risk to human health or the environment. EPA's interpretation
of a realized risk likely posed by these analytes, relative to pollutant loadings, background
ambient and source water contaminant levels and characteristics, and other relevant information
useful for this assessment, is presented in Chapter 5.

       Metals were the class of pollutants found most frequently and at concentrations that
exceeded national water quality criteria in samples of deck washdown discharge. Several
dissolved metals were measured at PHQs>10, relative to the most stringent benchmarks. Among
the dissolved metals, copper was the most prevalent, and was measured at PHQ>10 in
tow/salvage, fire, taxi, tour, and supply vessels. Dissolved cadmium was rarely detected, but had
the highest exceedance, in a tow/salvage vessel. Dissolved lead exceeded NRWQC benchmarks
in five of six salvage vessels, three of nine tugboats, one of two tour vessels, the one fire vessel,
and the one supply vessel. Dissolved zinc exceeded NRWQC benchmarks in five of six
tow/salvage vessels, as well as in tug, tour, fire, and supply vessels. Among the total metals,
arsenic and aluminum were the most prevalent, particularly in deck washdown discharges of
tow/salvage boats (both metals), tugboats (aluminum), and fishing and fire vessels (arsenic).
Total iron exceedances were also common, with the highest PHQs for total iron occurring  in
tugboats and tow/salvage vessels. Finally, total antimony and manganese exceedances were
relatively rare, with PHQs in those instances associated mainly with the nonfishing vessels. In
general, metal discharges were higher in the industrial nonfishing vessels compared to fishing
vessels.

       Among the conventional pollutants, TRC was the most prevalent, with regard to high
concentrations and frequency of exceedance of the discharge. The highest PQHs for TRC were
observed in three of the 11 fishing vessels, the two tour boats, a tow/salvage vessel, and a
tugboat. TSS and turbidity were the next most important classical pollutants, with high
occurrences distributed across all vessel classes, but particularly tugboats. The highest
exceedances of BOD were found in three tugboats, one shrimp vessel, and the supply boat. COD
and TOC concentrations were similar to BOD  concentrations. Oil and grease and sulfide were
high in only a select few samples (in tugboat, tow/salvage boat, and the supply boat).

       Samples for pathogens were taken from only fishing vessels, with fecal coliform and
enterococci potentially having the highest concentrations. Levels were high in all vessels except
for the gillnetting vessel in Alaska. Differences in pathogen loads could be related to location or
method of fishing (gillnetting vs. trawling). Pathogen loads in deck wash declined after washing
in all cases.

       Total phosphorus was the only nutrient of potential concern, with high levels found in
almost all samples, presumably due to the use of detergents in the deck wash practices. Long-

                                           185

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


chain nonylphenol polyethoxylates of the smallest chain (i.e., NP3EO, most degraded form) were
found in only three of the tow/salvage vessels, and total nonylphenol polyethoxylates were found
at high concentrations in two tour vessels. Finally, a moderately high PHQ of 5.6 for bis(2-
ethylhexyl) phthalate was found in the discharge of a tow/salvage vessel with a noticeably oily
sheen.
                                            186

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.3.11. Characterization of Deck Washdown and Runoff Water and Summary of Analytes that May Have the Potential
to Pose Risk
Vessel Type
(no. vessels)
Fishing (11)
Tugboats (9)
Tow/Salvage
(6)
Tour (2)
Water Taxi (1)
Fire(1)
Supply (1)
Recreational
(1)
Analytes that May Have the Potential Risk to Pose Risk in Deck Washdown and Runoff Water and Vessel Sources 1'2'3A5
Microbiologicals
Fecal
coliform
Enterococci
£. co//







Volatile Organic
Compounds








Semivolatile
Organic
Compounds


Bis(2-
ethylhexyl)
phthalate





Metals (dissolved)
Cu.Zn
Cu.Pb.Zn
Cu,Cd,Cr,Pb,
Ni.Zn,
Cu.Pb.Zn
Cu
Cu.Cr.Pb
Cu.Cd.Pb,
Ni.Zn
Cu.Ni
Metals (total)
AI.As.Fe
Al,
As,Fe,Mn
Al, As,
Fe.Sb
AI.As
AI,As,Fe,M
n
As,AI,Fe,S
b
Al, As,
Fe.Mn.Sb
Al, As
%
re
V
3
•o
c
re
O

X




X

Sulfide
X

X





Short-Chain
Alkylphenol
Ethoxylates and NP








Long-Chain
Alkylphenol
Ethoxylates


X
X




Nutrients
TP,
NH3-N
TP
(including
one very
high PHQ
=220)
TP,
NH3-N
TP

TP
TP
TP
O
•o
c
re
O
O
O
m
BOD,
COD,
TOC
BOD,
COD,
TOC
BOD,
COD,
TOC
BOD,
COD,
TOC

BOD,
COD,
TOC
BOD,
COD,
TOC
BOD,
COD,
TOC
Total Suspended
Solids
X
X
X
X
X
X
X
X
Other
Physical/Chemical
Parameters
TRC,
DO
TRC,
turbidity
TRC,
turbidity
TRC

turbidity
TRC,
turbidity
turbidity
Notes:
(1) Analytes are generally bolded when a large proportion of the samples have concentrations exceeding the NRWQC (e.g., 25 to 50 percent), when several of the samples have PHQs > 10 (e.g.,
two or three of five), when a few samples result in PHQs greatly exceeding the screening benchmark (i.e., 100s to 1,000s), or, in the case of oil and grease and for nonylphenol, when one or more
samples exceed an existing regulatory limit by more than a factor of 2. See text in Section 3.1.3 for a definition of PHQs and Table 3.1 for screening benchmarks used to calculate these values.
(2) EPA notes that the conclusion of potential risk is drawn from a small sample size, in some cases a single vessel, for certain discharges sampled from some vessel classes.  EPA
included these results in the tables to provide a concise summary of the data collected in the study, but strongly cautions the reader that these conclusions, where there are only a few
samples from a given vessel class, should be considered preliminary  and might not necessarily represent pollutant concentrations from these discharges from other vessels in this
class. (3) All dissolved metals identified as possible risks are potentially influenced by the dissolved metal concentrations measured in source water (generally city tap water; used by
all vessel types), particularly dissolved Cu and Zn.
(4) All total metals identified as possible risks are influenced by total metal concentrations measured in surrounding ambient water (relevant only for vessels where ambient water is used for deck
washdown (i.e., many fishing vessels performing deck washdown while offshore,  certain tug boats (as indicated in vessel survey)).
(5) Elevated total  phosphorus concentrations in deck washdown samples likely influenced by ambient and source water concentrations.

                                                                             187

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

   3.2.4  Fish Hold and Fish Hold Cleaning Effluent (Refrigerated Seawater and
          Ice Slurry)

       Refrigerated seawater and ice/ice slurry are the two commonly used methods for
preserving fish in the fish hold of many fishing vessels. EPA noted that some vessels
(e.g., large shrimping vessels in the Gulf of Mexico) use dry freezers to preserve their
catches; however, these vessels do not produce significant amounts of effluent from the
hold that comes into contact with seafood product and that is later discharged. Lobster
and crab boats have seawater flow-through tanks used to keep lobsters and crabs alive.
Both the freezers and flow-through tanks might contain residual seafood material that
sometimes is discharged when the vessels clean their holds.

       The analytes and parameters detected in fish hold effluent come from the vessel,
ambient water and potable/service water. Additionally, many  of the constituents can
come from the seafood product itself. If the seafood (e.g., fish, shrimp) are not frozen, but
preserved in refrigerated seawater or ice slurry, small quantities of organic material from
the fish (e.g., lipids, protein) will be released as the fish degrade, thereby increasing the
concentration of those constituents in the discharge. Furthermore, different volumes of
blood, mucus,  and other matter can drain from the seafood into the hold, depending on
how the fish is butchered or cleaned on deck. For example, salmon, when caught via
gillnets on gillnetting vessels, are cut  at the gills and bled and then placed into the
refrigerated sea water tanks/on ice before being cleaned (resulting in their internal organs
and some blood leaking into the water). In contrast, salmon caught on trailers are cleaned
while the fishing vessel is still  at sea and the internal organs are discharged into the
surrounding waters. Hence, on the salmon trailers, the organs and most of the residual
blood are not in contact with refrigerated water/ice, and consequently,  lower quantities of
these materials are discharged when the vessel empties its hold at the dock.

       The volume offish hold water generated by a fishing vessel depends on the size
of the vessel and the method used for keeping the product fresh. Vessels such as small
salmon trailers or long-liners that frequent Alaska waters have around  1,500 gallons of
fish hold storage. Assuming a hold is  occupied by approximately 50 percent fish and 35
to 40 percent ice when  the vessel off-loads at the seafood processing facility, the ice,
which is thrown overboard daily after the fish are unloaded, would result in a fish hold
discharge of between 500 and 600 gallons for these types of fishing vessels every three to
seven days (70 to 200 gpd).
                                        188

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                        Collecting Fish Hold Ice from a Long Liner
                              Fish Hold Ice from a Trawler

       Mid-size fishing vessels, such as gill netters, and purse seiners found in Alaska,
and shrimp boats in the Gulf of Mexico, have fish hold volumes of between 3,000 and
5,000 gallons. Assuming a hold contains between 35 and 40 percent of ice/water slurry, a
                                        189

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


vessel discharges between 1,000 and 2,000 gallons offish hold water every two to three
days (333 to 1,000 gpd).

       Larger fishing vessels such as off-shore trawlers found off the coast of New
England and tenders found in Alaska can have refrigerated seawater holding tanks or ice
hold tanks with capacities as large as 15,000 gallons. Assuming these fish hold tanks
contain 30 to 40 percent refrigerated seawater or ice after the seafood is unloaded, the
fish hold discharge would be between 4,500 and 6,000 gallons. These vessels are
expected to unload seafood and discharge the fish hold water every three to five days
(900 to 2,000 gpd).
                Two Examples of Full Fish Hold Tanks on a Tender Vessel
                                       190

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       EPA collected effluent samples from 31 commercial fishing vessels for this study.
Samples were collected from the fish holds that were in use on 26 of these vessels. EPA
generally collected single grab samples from these vessels while the vessels were
dockside.  These samples were usually collected while the effluent was being discharged,
but they were occasionally collected directly from the fish hold. EPA analyzed samples
for both total and dissolved metals, classical pollutants, pathogens, and nutrients. EPA
also analyzed three samples from fish holds for nonylphenols.

       The fish hold tank is cleaned after the catch has been off-loaded at the seafood
processing facility, so the frequency offish hold cleaning depends on the type and
amount offish being caught. For example, off-shore trawlers in New England might only
clean the fish hold tank every three to five days when they return to the fish processing
facility. Small fishing vessels such as salmon trailers and long-liners in Alaska off-load
the catch every three to seven days. Fish tenders and purse seiners with refrigerated
seawater tanks might clean the tanks every couple of days when they return to the fish
processing facility.

       On small fishing boats such as trailers and long-liners, and mid-size fishing boats
such as gill netters, fish holds are typically cleaned using a garden hose at a flow rate of
between 10 and 12 gpm. Fish hold cleaning is completed in 15 minutes or less, resulting
in a discharge of between 150 and 200 gallons per day. Larger vessels  such as off-shore
trawlers found in New England and large tenders in Alaska also use a garden hose to
wash down the fish hold tanks; however, cleaning these tanks requires  approximately 30
minutes. EPA estimated the volume offish hold cleaning water discharge for these
vessels ranges between 300 and  400 gallons per cleaning (60 to 200 gpd depending on
frequency).

       EPA was able to collect samples of the fish hold cleaning water discharge from
nine vessels. These samples were analyzed for the same constituents as fish hold effluent
plus nonylphenols. Nonylphenols are suspected pollutants associated with cleaning
products.

       3.2.4.1    Metals

Fish Hold Effluent

       Samples of refrigerated cooling water and ice slurry from 26 fish holds were
analyzed for dissolved and total  concentrations of 22 metals. The analytical results are
summarized in Table 3.4.1 (total metals data) and Table 3.4.2 (dissolved metals data) for
the 19  metals that were detected in one or more fish hold effluent samples. Figures 3.4.1
and 3.4.2 present these  same results for total and dissolved metals, respectively,
                                       191

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

normalized by the lowest NRWQC where applicable. The following metals were detected
in all fish hold water samples:

       •  Total aluminum
       •  Dissolved and total barium
       •  Dissolved and total calcium
       •  Dissolved and total cobalt
       •  Dissolved and total iron
       •  Dissolved and total potassium
       •  Dissolved and total sodium
       •  Dissolved and total vanadium
       •  Dissolved and total zinc

       Concentrations of a number of other metals were measured for 50 percent or more
of the samples analyzed:

       •  Dissolved aluminum
       •  Total arsenic
       •  Dissolved and total copper
       •  Dissolved and total magnesium
       •  Dissolved and total manganese
       •  Dissolved and total potassium
       •  Total silver.

       Several metals for which EPA tested had concentrations that were notable. These
metals include dissolved and total arsenic, and dissolved copper, selenium, and zinc (see
Figures 3.4.1 and 3.4.2). A small percentage of the samples contained all the metals
which EPA regularly analyzes; however, metals such as lead, nickel, and selenium were,
with a few notable exceptions, in concentrations below PHQs at the point of discharge
(see Figures 3.4.3 and 3.4.4). EPA analyzed for and detected dissolved and total barium,
cobalt, iron, potassium, silver, sodium,  and  vanadium in only two samples. All of the
detected concentrations in the two samples were low, except for iron. EPA also analyzed
for antimony, beryllium, and thallium in these two samples and did not detect any of
these metals.

       The concentrations of many of the metals that were detected in fish hold
discharges are not unexpected as fish holds  generally have numerous exposed metal
surfaces. In addition, the pumps used to add water to the hold might also add low
concentrations of metals. Finally, metallic fishing equipment, deck surfaces, and  other
materials sometimes come in contact with the fish or water that runs into the hold.

       Some metal concentrations, particularly mineral salts, appear to be primarily a
result of background concentrations in the ambient water. For example, aluminum,

                                       192

-------
                 Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

barium, calcium, iron, magnesium, sodium, and potassium appear to be primarily
influenced by background concentrations. Other metals that had measurable
concentrations (e.g., arsenic, copper, manganese, and zinc) appear to result largely from
mechanically refrigerated water used to cool the sea water to preserve seafood catch,
adding ice to do the same, or possibly, from the seafood catch itself, or from any
combination of the three.

       Several metals were detected in at least one sample offish hold effluent with PHQ
values of greater than 1 (see Figures 3.4.3 and 3.4.4). For total metals, this included
aluminum, arsenic, copper, iron, and manganese. However, as discussed above,
aluminum concentrations appear to be primarily influenced by ambient water background
concentrations. Total copper concentrations exceeded the total copper benchmark based
on human health (for consumption of water and aquatic organisms) of 1,300 |ig/L by a
small fraction in two samples (Table 3.4.1). These total concentrations, however, could
pose potential risk to the  aquatic environment because the human health criteria of 1,300
Hg/L is significantly higher than the 3.1 |ig/L benchmark  used for dissolved copper based
on the saltwater chronic ambient water quality criterion for the protection of aquatic life.
When high levels of particulate copper are discharged, some of the paniculate copper will
likely convert to dissolved  copper and be made bioavailable to aquatic life. EPA collected
only two samples for analysis of total iron, one of which had a PHQ value of five and the
other  eight.

       Another metal with high PHQ values is total arsenic. The PHQ values for total
arsenic ranged from between more than 100 to more than 20,000 (Figure 3.4.3)29. One
reason for these extreme PHQ values is the exceptionally low screening benchmark of
0.018 |ig/L for total arsenic. Nonetheless, concentrations  of total arsenic in the upper end
ranges of these measurements are a possible environmental concern. These discharges
may have the potential to cause or contribute to exceedances of water quality standards,
particularly  in areas where multiple fishing vessels discharge their holds into the same
waters within the same time period.

       Several dissolved metals, including arsenic, cadmium, copper, iron, nickel, and
selenium, also had PHQs above 1 (see Figure 3.4.4). Dissolved arsenic samples resulted
in PHQs of approximately 9-10 for two discharges; one was from a shrimping vessel
from the Gulf Coast and the other from a ground fishery vessel in New England, while a
third boat ground fishery vessel in New England had a PHQ value of just over 2. There
was also only one sample which had a PHQ value for cadmium of approximately 5. Only
29 While EPA suspects the highest concentration of total arsenic (and total selenium) from a shrimping
vessel might be slightly elevated due to positive interference, measured concentrations of arsenic in fish
hold effluent from other similar vessels were absent positive interference and nearly as high. Therefore,
EPA believes the measured concentrations of total arsenic (and to a lesser extent selenium) from the
slumping vessel to reasonably represent true effluent concentrations for the discharge.

                                        193

-------
                 Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


four of the 26 values exceeded a PHQ value of 1 for dissolved nickel, and none exceeded
a value of 2. Dissolved selenium had 6 samples exceed a PHQ value of 1 (the highest
value of which was approximately 12). Dissolved zinc had numerous PHQ values of
greater than 1, but none greater than 10. Dissolved copper had numerous samples that
exceeded the PHQ value of 1, with more than 25 percent of these samples having a PHQ
value of greater than 10.

       The high dissolved arsenic concentrations were observed exclusively from three
vessels; a shrimping boat (345 |ig/L)30 and two ground fishery trawlers (74 and 310
|ig/L). Ambient water concentrations indicate that the arsenic likely did not come from
the surrounding water, although dissolved arsenic was measured at a substantial level of
26 |ig/L in the ambient water where the shrimping vessel was sampled. Another possible
explanation is entrainment of arsenic contaminated sediments on nets. Each of the vessels
with high arsenic values (trawlers and shrimp boats) use nets that drag the ocean floor.
When nets are retrieved and emptied on the deck of the vessel, entrained sediments from
the ocean floor could migrate into the fish holds along with the fish and shrimp. One
other possible source includes organic arsenic  compounds that are primarily found in
organisms living in the sea. Based on the limited data collected, EPA cannot identify the
specific source(s) of the high dissolved arsenic values at this time.

       In summary, some samples of dissolved copper in fish hold effluent discharges
were well above the PHQ screening benchmark of 3.1 |ig/L based on the 2006 NRWQC
saltwater chronic aquatic life criterion. Many of these concentrations resulted in PHQs of
greater than 10, with some upwards of 200. The three elevated concentrations of
dissolved arsenic could potentially pose an environmental concern, particularly if these
arsenic concentrations are common in these vessel discharges. Finally, concentrations of
total arsenic are also high relative to the benchmark, resulting in high PHQ values and
may have the potential to pose risks to human health if discharged into drinking water
sources, though almost all fishing vessels operate in marine or estuarine environments
that are not used for drinking water.
30 While EPA suspects the highest concentration of dissolved arsenic (and dissolved selenium) from a
shrimping vessel might be slightly elevated due to positive interference from major seawater cations,
measured concentrations of arsenic in fish hold effluent from other similar vessels were nearly as high, but
absent positive interference. Therefore, EPA believes the measured concentrations of dissolved arsenic
(and to a lesser extent selenium) from the slumping vessel to reasonably represent true effluent
concentrations for the discharge.

                                        194

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.1. Results of Fish Hold Effluent Sample Analyses for Total Metals1
Total Metal
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum
Arsenic
Barium
Cadmium
Chromium
Cobalt
Copper
Iron
Lead
Manganese
Nickel
Selenium
Silver
Vanadium
Zinc
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
26
26
2
26
26
2
26
2
26
26
26
26
2
2
26
26
16
2
3
7
2
24
2
9
15
5
7
1
2
26
100
62
100
12
27
100
92
100
35
58
19
27
50
100
100
827
40
98
0.99
4.3
3.7
190
2000
7.1
24
7.7
12
2.4
9.2
340
840
4.8
110


4.4
40
2500

6.6


2.7
10
230
89

83


2.9

1600





8.1
27
180

83


2.9
5.8
1600





8.1
46
420

83


2.9
12
1600





8.1
100
900
13
110

2.6
4.4
140
2500
5.6
17

13
2.7
10
450
1800
210
110
1.9
19
4.4
710
2500
31
130
17
29
2.7
10
940
2400
380
110
3.3
35
4.4
1700
2500
42
140
30
90
2.7
10
1700
87
0.018
1000
NA
NA
NA
1300
300
NA
100
610
170
NA
NA
7400
Cationic Metals
Calcium
Magnesium
Potassium
Sodium
ug/L
ug/L
ug/L
ug/L
26
26
2
2
26
25
2
2
100
96
100
100
1 50000
450000
330000
1 200000
1 90000
580000
480000
1 900000
1900

190000
370000
3000
1800
190000
370000
15000
14000
1 90000
370000
270000
840000
480000
1 900000
300000
980000
480000
1900000
31 0000
1 1 00000
480000
1900000
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            195

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.2. Results of Fish Hold Effluent Sample Analyses for Dissolved Metals1
Dissolved Metal
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum
Arsenic
Barium
Cadmium
Chromium
Cobalt
Copper
Iron
Lead
Manganese
Nickel
Selenium
Silver
Vanadium
Zinc
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
26
26
2
26
26
2
26
2
26
26
26
26
2
2
26
24
10
2
1
3
2
23
2
3
19
4
6
2
2
26
92
38
100
4
12
100
88
100
12
73
15
23
100
100
100
490
31
64
0.77
1.9
1.8
96
350
2.3
22
6.1
9.2
1.3
3.4
180
670

84


2.0
15
360

11


1.5
3.5
120


44


1.6

340




1.0
3.2
24
20

44


1.6

340




1.0
3.2
31
60

44


1.6
6.0
340




1.0
3.2
55
850
5.7
84


2.0
38
360

28

2.5
1.5
3.5
240
970
150
84

5.8
2.0
390
360
4.4
80
13
20
1.5
3.5
450
1000
350
84
1.4
7.9
2.0
920
360
8.0
110
17
61
1.5
3.5
790
NA
36
NA
0.25
11
NA
3.1
NA
2.5
NA
8.2
5.0
1.9
NA
81
Cationic Metals
Calcium
Magnesium
Potassium
Sodium
ug/L
ug/L
ug/L
ug/L
26
26
2
2
26
25
2
2
100
96
100
100
160000
480000
330000
1 200000
1 80000
560000
470000
2000000
1200

1 80000
360000
1900
770
1 80000
360000
9000
11000
1 80000
360000
290000
920000
470000
2000000
300000
990000
470000
2000000
31 0000
1 1 00000
470000
2000000
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            196

-------
              Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

1000
nr
~o>
^ 100
0
-t—l
CD
c
0 in
o ID

O

1

_
—
-
r $
-
—
-
-
—
-
-
_
—
:

t
>
&







oJcm


I I

k
a






©
©
j |
ky
i 
-------
              Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     1000
            :  i  i   i   i   i   i   f  i   r
 O)



 o
"-I—•
 CD
 CD
 O
 c
 o
O
      100
        10
                                      ©
                                               ©

                                               ©
                                            ©
                                            §  ©
                                            ©
                 I	I
                                     ^^^^
                                     &* Vc  >*
                         Dissolved Metals
Figure 3.4.2. Box and Dot Density Plot of Dissolved Metals Concentrations

Measured in Samples of Fish Hold Effluent
                                  198

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 0

"5
 ^

a

•a

 CD
 N
 CD


ID

"c
 CD
-I—•
 o
Q_
10000.000


  1000.000


   100.000


     10.000


       1.000


       0.100


       0.010
0.001
I	I
                                   1   1

                                        1   1
                  ^^^^^^^^
                  p>  ip   C/^rX^              ?*    *O    N\"2^
                                     Total Metals
        Figure 3.4.3. Box and Dot Density Plot of Potential Hazard Quotients for Total

        Metals in Samples of Fish Hold Effluent

        (Note: Replacement values for non-detects are circled. Also, as discussed in the text above, total arsenic is

        a potential concern; however, the exceptionally high PHQ values are due in part to the low human health

        value for total arsenic used as a benchmark).
                                           199

-------
 0

"5
 ^
a
 03
 N
 03
 o
Q_
               Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     100.0
      10.0
        1.0
        0.1
                 i   r
i   i   t   i   i
                 *     ©
           ©
©

©



©
                 ©


                           i   i
         I   i
                 ^1^^<$04$^
                  O^ov           ^jz^    *=>    ^

                         Dissolved Metals
 Figure 3.4.4. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved
 Metals in Samples of Fish Hold Effluent
 (Note: Replacement values for non-detects are circled).
                                  200

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

Fish Hold Cleaning Effluent

       EPA expected effluent from the cleaning offish holds to be fundamentally similar
to fish hold effluent with two exceptions: 1) many vessels used a soap or disinfectant,
which would not be expected to be present in the hold, and 2) cleaning fish holds brings
in either potable water from the local municipality via a pierside hose (service water) or
ambient water pumped from the surrounding waters. Table 3.4.3 presents summary
statistics for fish hold cleaning effluent. Figures 3.4.6 and 3.4.7 show the detected results
for total and dissolved metal concentrations, respectively, and Figures 3.4.8 and 3.4.9
shows the PHQ values for total and dissolved concentrations, respectively, where
applicable.

       Generally, average and maximum total and dissolved metals concentrations for
fish hold cleaning were slightly lower than for fish hold effluent. These lower values
could be due to any number of reasons: less contact time  with the vessel for fish hold
cleaning effluent, differences in source water (mechanically refrigerated and ice versus
city tap water), less contact time (or none at all) with the  seafood product or its residuals,
etc.

       The lower concentrations of metals for fish hold cleaning effluent resulted in
lower overall PHQ values for both total and dissolved forms, as well as a lower
percentage of samples that exceed a PHQ of 1. Not surprisingly, the metals (dissolved
copper, dissolved and total arsenic) identified as having high PHQs for fish hold effluent
also exhibited  higher PHQ values in fish hold cleaning effluent. Likewise, dissolved
copper occurs  in fish hold cleaning effluent at concentrations mostly above a PHQ value
of one, and dissolved arsenic was found in  two samples with PHQ values above one.
Dissolved zinc was also found in several samples with PHQ values above one, the
maximum being a PHQ value just below 10 (Figure 3.4.8).
                                       201

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.3. Results of Fish Hold Cleaning Effluent Sample Analyses for Metals1
Metal
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum, Dissolved
Aluminum, Total
Arsenic, Dissolved
Arsenic, Total
Cadmium, Total
Chromium, Dissolved
Chromium, Total
Copper, Dissolved
Copper, Total
Lead, Dissolved
Lead, Total
Manganese, Dissolved
Manganese, Total
Selenium, Dissolved
Selenium, Total
Zinc, Dissolved
Zinc, Total
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
5
5
1
1
3
8
9
1
4
4
5
1
2
8
8
100
100
56
56
11
11
33
89
100
11
44
44
56
11
22
89
89
780
1100
22
35
1.0
1.5
4.6
34
57
2.7
19
21
33
6.0
7.4
190
470
880
930
5.3
8.7



12
25



4.8


53
140
74
850






6.4








74
850






6.4








760
860





8.6
11






19
17
950
1500
38
64


5.4
32
61

37
39
61

7.0
420
890
1000
1700
97
150
3.0
3.4
23
180
290
8.7
79
64
110
14
18
640
1800
1000
1700
97
150
3.0
3.4
23
180
290
8.7
79
64
110
14
18
640
1800
NA
87
36
0.018
NA
11
NA
3.1
1300
2.5
NA
NA
100
5.0
170
81
7400
Cationic Metals
Calcium, Dissolved
Calcium, Total
Magnesium, Dissolved
Magnesium, Total
ug/L
ug/L
ug/L
ug/L
9
9
9
9
9
9
9
9
100
100
100
100
250000
260000
790000
780000
270000
280000
860000
880000
11000
13000
12000
13000
11000
13000
12000
13000
240000
260000
750000
710000
300000
310000
990000
1000000
320000
320000
1000000
1000000
320000
320000
1000000
1000000
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            202

-------
               Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     1000
 O)
 ^

 c
 o

"co
 CD
 o
 c
 o
o
100
  10

- oo
0^
^
)


_
-





oo



1























G
©

~
~

-

-


—
_
-
-


—
































c

f









0

)









Q






oo









o
0
00






t




0














0



00
|ooo_o|o_ooo









f





00



c





5



)

o
-A-




10f


c


















G

c









0(










G
3
0
3






0


X




3_T






1 <
c
t



3

0

O









r\
C
G
0





o
)
	

o

n




o

/-^^
o



j OOOQOOO c
_
—


~
^

-

-


—
_
5
-


1
—
                             Total Metals
Figure 3.4.5. Box and Dot Density Plot of Total Metals Concentrations Measured in

Samples of Fish Hold Cleaning Effluent
                                    203

-------
               Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
IUUU

nr
I5 100
•^_ X
._
0
"ro
c
CD
g 10
0
O




1
: E^3
-
1

— c-
I ©
-
-
~
-

—
_
-











0










oooo

1





©


oooooooo
1




1 1 1 1 :
c
*


1
,




o
0




c












(
c



©




o




o
,


oooooooo

1



-
o


0



©


oooooooo {


~
-
-
~
-

—
-
^




f 1 1
                           Dissolved  Metals
Figure 3.4.6. Box and Dot Density Plot of Dissolved Metals Concentrations
Measured in Samples of Fish Hold Cleaning Effluent
                                     204

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
1000.000
-1— •
c
| 100.000
a
-o 10.000
| 1.000
"-£ 0.100
CD
-1— •
£ 0.010
0.001
= I v' I I I
: o
=- o
r
I 0
- oo
0
- n
— OOO
1 T
r i i i i i
1111 =
?
ob -
o ^

v nn_2 ~
— -^QOOOOCIC^ ~

o :
on
I I I r^ ~=
                                Total Metals
Figure 3.4.7. Box and Dot Density Plot of Potential Hazard Quotients for Total
Metals in Samples of Fish Hold Cleaning Effluent
(Note: Replacement values for non-detects are circled).
                                       205

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
"c
•J= 100
o
a
_j
03
N
03
11 n
I .U
^^~
03
"c
CD
-i— •
O
Q_
01
.1
1
-

-

-
_
-
-

~- i (6



c
(


-f
c
f

0



3
)


v^
)
^
«
n
1






0

*~- --*•
2) 1
1
§
C
r
o
c

f



D

r-
)
)
1
O
)

&





\ \ \

O
o
-CSoooaacs
1QO_QQ_OQS>




i i i

o

r



c
0

(jS

o

1



)
0

3)
-
-

-

-
_
-
-

-
                             Dissolved Metals
Figure 3.4.8. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved
Metals in Samples of Fish Hold Cleaning Effluent
(Note: Replacement values for non-detects are circled).
                                      206

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.4.2    Classical Pollutants

       Table 3.4.4 presents analytical results for 14 classical pollutants detected in
samples from fish hold effluent (all classical pollutants analyzed for in the study were
detected). These detected results are also shown in Figure 3.4.9.

       Except for dissolved oxygen, other physical parameters measured (conductivity,
pH, salinity, and temperature) did not have results that were likely to result in any impact
on receiving water quality. Dissolved oxygen concentrations were low in several samples
offish hold effluent:  hypoxic (< 2 mg/L) in three cases and marginal (<5 mg/L) in 19
additional cases. These low oxygen conditions may be driven by the high BOD
concentrations found in many of the fish holds. Effluent with low dissolved oxygen
concentrations were also noted in the fish hold cleaning effluent, with six of nine samples
(67 percent) having concentrations of less than 5 mg/L (see Table 3.4.5 and Figure
3.4.10).

       EPA found BOD and COD to be highly elevated in fish hold effluent (Table
3.4.4). BOD was measured in several samples in concentrations in the thousands of
mg/L. High levels of BOD are almost certainly caused by the decay of the organic
material associated with the seafood product. As shown in Figure 3.4.9, the majority of
these concentrations are generally higher than those of raw sewage (which can range up
to a few hundred mg/L), and almost all are higher than a wastewater treatment plant's
secondary treatment limit of 30 mg/L for BOD. The median value for BOD discharge
was approximately 471 mg/L, indicating that BOD discharge from fish holds are
abnormally elevated (see Figure 3.4.11). The highest BOD value of 5,130 mg/L
approximates the concentrations found in sewage sludge (Metcalf and Eddy, 1979).

       These high levels of BOD in discharges could potentially pose environmental
problems in certain circumstances. For example,  high BOD concentrations in fish hold
effluents are potentially ubiquitous, and discharges could result in impacts to receiving
waters where there are numerous fishing vessels, poor flushing, or high levels of existing
hypoxic (low oxygen) stress in the water body. In stratified waters with hypoxic or
anoxic (no oxygen) conditions, the risk associated with elevated BOD is most likely to
occur in deeper waters under a thermocline or picnocline. When using refrigerated
seawater systems, fish hold effluent may be as saline (or more saline) than the
surrounding water. Where it is also cooler than the surrounding water, the fish  hold
effluent would be more likely to sink to the bottom of the stratified water under the
warmer water. This may deliver the BOD load to the deeper layers of the water body
where oxygen levels are likely to be lower in eutrophic waters. In contrast, where ice is
used to cool fish in the fish hold, the BOD load may be more likely to stay in the surface
layers since fresh water is less dense than salt water. Thus, a low salinity fish hold
                                       207

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

effluent discharge may prevent the BOD loading from having as significant an impact to
aquatic organisms in the receiving waters.

       The considerable variability in BOD concentrations from the 26 fish hold effluent
samples may be due to how fish are kept. The average concentration of BOD is lowest
for the lobster tank compared with the other fish hold types, which is logical since lobster
tanks have continuously circulating ambient water with live seafood inside. Hence, the
water is constantly being refreshed, while the seafood product generally has not begun
the process of degrading or bleeding into the tank. There could be other differences in
BOD concentrations based upon whether fish are kept on top of ice, in ice water slurry, or
in refrigerated seawater. New England trawlers and Gulf Coast shrimp boats had several
vessels with exceptionally high BOD concentrations.

       Whereas BOD measures oxygen demand from biodegradable material, COD
measures oxygen demand for both biodegradable material and nonbiodegradable
oxidizable material. Like BOD, COD discharge is elevated in fish hold effluent and fish
hold cleaning effluent (Tables 3.4.4 and 3.4.5). Occasionally, these values are
exceptionally high, which could potentially cause stress on a water body where there are
many discharges from fish holds and where there may be low circulation or flushing or
existing hypoxic or anoxic stress in the water body.

       Oil and grease as measured by HEM and SGT-HEM are generally discharged in
low concentrations from fish hold effluent, with the vast majority of samples from both
fish hold effluent and fish hold cleaning effluent having HEM and SGT-HEM being
discharged in quantities below 5 mg/L. However, there are a few discharges where the
concentrations exceed 15 mg/L. The highest of these values for either fish hold or fish
hold cleaning effluent (the HEM concentration was approximately 28 mg/L - slightly less
than twice the regulatory limit of 15 mg/L) are from the samples taken during a fish hold
cleaning event while onboard a New England ground fishing vessel. These values
demonstrate that while oil and grease discharges from fish holds sometimes occasionally
occur at elevated concentrations, they were generally not observed at concentrations that
are of particular concern.

       The concentrations of the classical pollutants EPA measured that are associated
with sediment or cloudiness (i.e., TSS and turbidity) were roughly equivalent to
concentrations observed in raw sewage effluent, but considerably lower than stormwater
runoff from construction sites or highly urbanized streams. TSS was elevated in both fish
hold effluent and fish hold cleaning effluent; however, concentrations were generally not
sufficiently elevated to alone exceed water quality standards. Just under 90 percent of
samples exceed the secondary treatment concentration of 30 mg/L for TSS  (the value
used to establish the PHQ benchmark), with a maximum concentration of 1,100 mg/L in
                                       208

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

a fish hold effluent sample. As with BOD, TSS appears to be more diluted in fish hold
cleaning effluent than in fish hold effluent. While it did not test for volatile suspended
solids (VSS) in this sampling program, EPA assumed that a significant percentage of the
TSS concentration is directly caused by organic material related to the seafood product.
Similar to TSS, turbidity concentrations were elevated in both fish hold effluent and fish
hold cleaning effluent, and slightly more concentrated in fish hold effluent than in fish
hold cleaning effluent.

       The concentrations of sulfide in fish hold and fish hold cleaning effluent were low
in most samples, with most values falling below a reporting limit value of 0.01 mg/L.
Sulfide was detected in only seven of 25 samples  where the parameter was tested, and in
only four of seven fish hold cleaning samples. However, a few samples had significantly
elevated sulfide concentrations, including a maximum fish hold concentration of 0.16
mg/L (PHQ value of 80) from fish hold discharges, and a maximum fish hold cleaning
value of 0.48 mg/L (PHQ value of 240). These high sulfide values cannot be attributed to
high background concentrations. A relatively higher percentage of detectable sulfide
concentrations were noted in New England ground fishery trawlers  compared with other
areas (seven out of the 11 detections). EPA is unable to determine why the New England
fishery vessels have higher concentrations of sulfide compared with vessels using other
fishing platforms or from other areas; however, one possible explanation is that the New
England fishery vessels are at sea for seven to 10  days, whereas Alaskan fishing vessels
are offloaded once every one to two days.

       TRC was detected with some prevalence (roughly a third to  two thirds of the
samples for fish hold and fish hold cleaning effluent, respectively), with maximum
concentrations of 0.3 mg/L (fish hold effluent) and  1.51 mg/L (fish  hold cleaning
effluent). PHQs for the fish hold and fish cleaning effluent ranged from one to 40 and one
to 200, respectively (data not shown). Such high concentrations might be expected
considering the source water (e.g., bag ice for keeping catch cold in fish holds) or use of
chlorine bleach for cleaning and disinfection (fish hold cleaning effluent). In both cases,
effluent volume is low relative to receiving waters for this volatile compound, and as
such, EPA does not expect significant risk to human health or the environment.

       TOC was detected in all of the 25 of the fish hold effluent samples for which it
was tested and all nine fish hold cleaning samples. Concentrations ranged from a low of
1.8 mg/L to an extreme high of 2,200 mg/L (see Table 3.4.4).  Background concentrations
of TOC (i.e., from mechanically refrigerated water or ice) are  much lower (in the range of
2 to 19 mg/L) and do not appear to be a significant cause of the high TOC loads in the
effluent. TOC levels are likely elevated by decay  and residuals from the seafood product.
                                       209

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.4. Results of Fish Hold Effluent Sample Analyses for Classical Pollutants1
Parameter
Biochemical Oxygen Demand
(BOD)
Chemical Oxygen Demand (COD)
Conductivity
Dissolved Oxygen
Hexane Extractable Material (HEM)
PH
Salinity
Silica Gel Treated HEM (SGT-HEM)
Sulfide
Temperature
Total Organic Carbon (TOC)
Total Residual Chlorine
Total Suspended Solids (TSS)
Turbidity
Units
mg/L
mg/L
mS/cm
mg/L
mg/L
SU
ppt
mg/L
mg/L
C
mg/L
mg/L
mg/L
NTU
No.
samples
26
26
26
26
26
26
26
26
25
26
25
26
26
26
No.
detected
24
26
26
26
18
26
26
15
7
26
25
10
26
26
Detected
Proportion
(%)
92
100
100
100
69
100
100
58
28
100
100
38
100
100
Average
Cone.
840
1500
25
4.3
3.2
7.0
13
3.4
0.017
7.0
290
0.096
210
96
Median
Cone.
440
940
30
3.9
1.5
6.8
17
0.98

6.9
140

130
63
Minimum
Cone.

52
0.20
1.7

6.0
0.10


-0.16
1.8

10
9.0
10%
25
340
0.35
2.0

6.3
0.47


0.098
8.3

29
16
25%
140
660
3.3
2.8

6.5
1.4


3.0
48

71
25
75%
830
1900
43
5.7
2.9
7.5
25
2.2
0.011
9.5
260
0.13
190
120
90%
3100
2600
46
8.2
6.4
7.8
28
3.7
0.045
16
970
0.22
690
310
Maximum
Cone.
5100
8700
61
9.2
16
8.3
28
4.4
0.16
26
2200
0.30
1100
450
Screening
BM2
30
NA
NA
NA
15
NA
NA
15
0.0020
NA
NA
0.0075
30
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            210

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.5. Results of Fish Hold Cleaning Effluent Analyses for Classical Pollutants1
Parameter
Biochemical Oxygen Demand (BOD)
Chemical Oxygen Demand (COD)
Conductivity
Dissolved Oxygen
Hexane Extractable Material (HEM)
PH
Salinity
Silica Gel Treated HEM (SGT-HEM)
Sulfide
Temperature
Total Organic Carbon (TOC)
Total Residual Chlorine
Total Suspended Solids (TSS)
Turbidity
Units
mg/L
mg/L
mS/cm
mg/L
mg/L
SU
ppt
mg/L
mg/L
C
mg/L
mg/L
mg/L
NTU
No.
samples
9
9
8
9
9
9
9
9
7
9
9
9
9
9
No.
detected
6
9
8
9
6
9
9
4
4
9
9
6
9
9
Detected
Proportion
(%)
67
100
100
100
67
100
100
44
057
100
100
67
100
100
Average
Cone.
470
1100
35
5.6
5.4
7.6
48
4.9
0.10
9.2
210
0.29
190
100
Median
Cone.
300
960
41
2.9
1.4
7.6
24

0.019
8.2
74
0.11
84
59
Minimum
Cone.

490
2.6
1.4

6.9
1.3


4.7
1.9

16
0.20
10%

490
2.6
1.4

6.9
1.3


4.7
1.9

16
0.20
25%

530
27
1.6

7.2
19


5.7
5.1

26
1.0
75%
770
1600
45
9.6
4.2
8.1
27
2.8
0.17
12
430
0.29
400
210
90%
1800
2400
46
15
28
8.6
260
12
0.48
15
730
1.5
460
330
Maximum
Cone.
1800
2400
46
15
28
8.6
260
12
0.48
15
730
1.5
460
330
Screening
BM2
30
NA
NA
NA
15
NA
NA
15
0.0020
NA
NA
0.0075
30
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            211

-------
              Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
_ 1000.00
E 100.00
c 10.00
o
"co
-£ 1.00
CD
o
S 0.10
0.01
El I I I I I I I I I I I I E
~ ° I ' ""
f ||:..._| ' ^Lj
: V * :
_ I ra © __
\. 	 J 	 :
              (^ (^ ^ c\^ c^ c\^ o^ c^ c\^ c^ c\^ ;OV\5V\S
            ^\^%°v^\^\^.V9^\^


           "O0^        ^°                     ^
                       Classical  Pollutants
Figure 3.4.9. Box and Dot Density Plot of Classical Pollutant Concentrations/Values

Measured in Samples of Fish Hold Effluent
                                   212

-------
            Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

1000.00



c
g 100.00
E
^

o 10.00
o
1 1.00
f—
CD
O
o 0.10
O

0.01

.
— ,
~ r°

-
I
—
:
-
~
—
-
—


-G






1
-
_
—
:
-
—
=


O_









JJ

















I
00
0
OQ.O















|
1






.S8.
9


©







1
1









i
T







1
1 1 1




©


u, 00
UaiJ
©
i °CT
M ©






1 1 1













c




















0

nJ

„











C
-e











0






































'
<

-Q-















o~
OOQ


























1



A
"
o
0
T
I









1
1




i




[~]


— G




-(
,
1




1
I _
—
~


-







—
:
-
~
ja.1
-
—
-
_
—
=
—
—
I =
      ^^^*^&&&*&
          o°      ^

                   Classical Pollutants
Figure 3.4.10. Box and Dot Density Plot of Classical Pollutant

Concentrations/Values Measured in Samples of Fish Hold Cleaning Effluent
                            213

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                BOD Concentrations in Waste water Effluent
   5000
=• 4000
D)

£

w
c
o
5 3000
o
o
c
o
O 2000

Q
O
CO
   1000
             Secondary Treatment Limit

             (Limit for Lreated Sewage)
Raw Sewage
Fish Hold

Effluent
Cleaning of

Fish Hold
                                      Discharge Type


     Raw Sewage Data Sources: Orford and Matusky (1959). Henry (1996). Medium value from Metcalf and Eddy (1979).
      Figure 3.4.11. Comparison Between the BOD Secondary Treatment Limit from

      Sewage Treatment Facilities (30 mg/L), Average BOD Raw Sewage Concentrations,

      and BOD Concentrations from Fish Hold Effluent and Fish Hold Cleaning Effluent
                                            214

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       3.2.4.3    Pathogen Indicators (Microbiological)

       Sampling pathogen indicators from fish holds presented logistical challenges for
the EPA sampling team. Many fishing vessels were sampled in locations remote from
laboratories and the holding times of tests for these three pathogens (< 6 hours) prevented
EPA from analyzing these samples from many of the sampling events. Nonetheless, EPA
was able to test for E. coli and enterococci in seven fish hold effluent samples and for
fecal coliform in 11 fish hold effluent samples. The results are summarized in Table 3.4.6
(upper half of table) and shown graphically in Figure 3.4.12.

       Of these fish hold effluent samples, EPA detected bacteria concentrations above
the most stringent screening benchmarks for one (of the seven) E. coli samples, four (of
the seven) enterococci samples, and three (of the 11) fecal coliform samples. However,
EPA strongly suspects that all of these exceedances were due primarily or exclusively
due to background concentrations. For example, the fish hold effluent from a fishing
vessel sampled  in Gloucester, Massachusetts, exceeded all three stringent screening
benchmarks for E. coli, enterococci, and fecal coliform. However, ambient water
concentrations collected earlier in the day exceeded the concentrations in the later sample
taken from the fish hold. The likely source of the pathogenic bacteria in this case was a
combined sewer overflow (CSO) a few hundred feet above the location of the fishing
vessel. The fishing
vessel used ambient
water to wash off its
deck while
unloading cargo
(see section
3.2.3.3). Some of
this water likely
made its way into
the fish hold before
EPA sampled the
fish hold effluent
again at the later
time period; hence,
in this case, EPA
strongly doubts that
the vessel was the
source of the
extremely high
pathogen levels.
What are Combined Sewer Overflows (CSOs) and Sanitary Sewer
Overflows (SSOs)?

Combined sewer systems are sewers that are designed to  collect
rainwater runoff, domestic sewage,  and industrial wastewater in the
same pipe. Most of the time, combined sewer systems transport all of
their wastewater to a sewage treatment plant,  where it is treated and
then discharged to a water body. During periods of heavy rainfall or
snowmelt, however, the  wastewater  volume  in a combined  sewer
system can exceed the capacity of the sewer system or treatment
plant. For this reason,  combined sewer systems are designed to
overflow  occasionally  and discharge  excess  wastewater directly to
nearby streams, rivers, or other water bodies. These overflows, called
combined sewer overflows (CSOs), contain not only stormwater but
also untreated human and industrial waste, toxic materials, and debris.

Properly designed, operated, and maintained sanitary sewer systems
are meant to  collect and transport all of the sewage  that flows into
them  to  a  publicly  owned  treatment works  (POTW).  However,
occasional unintentional discharges of raw sewage from municipal
sanitary sewers  occur  in  almost  every  system.  These types of
discharges are called sanitary sewer overflows (SSOs). SSOs have a
variety of causes, including but not limited to severe weather, improper
system operation and maintenance, and vandalism. EPA estimates
that there are  at least 40,000 SSOs each year. The untreated sewage
from these  overflows  can contaminate our waters, causing serious
water quality problems.
                                       215

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       EPA encountered a similar situation while sampling a commercial fishing vessel
in New Bedford, Massachusetts. The samples from the fish hold exceeded water quality
criteria for enterococci (127 MPN/ 100 ml) and fecal coliform (125,000 CPU/ 100 ml).
However, this vessel was sampled immediately adjacent to an SSO that contained raw
fish waste and human sewage: the ambient water had enterococci concentrations of 4,342
MPN/ 100 ml and fecal coliform concentrations of 6,500 CPU/ 100  ml. This vessel also
used ambient water to hose off its deck, introducing the pathogenic bacteria to the fish
hold. Note that for fecal  coliform, this latter vessel's fish hold effluent did appear to add
to the high fecal coliform count in the sample.

       None of the concentrations of the three pathogens exceeded the most stringent
NRWQC set for the pathogens in cases where the ambient concentrations were also
below the stringent NRWQC. Although the results were based on this limited number of
samples, EPA believes it is unlikely that there is an onboard source of these pathogenic
bacteria in the fish hold.

       EPA was able to  test the effluent from three separate fish holds  from three vessels
while they were being cleaned (see Table 3.4.6, lower half of table). Two of the fish hold
cleaning effluent samples were from those vessels discussed above,  where ambient water
pathogen concentrations were impacted by the discharge from a CSO and an SSO. The
third sample was from a vessel sampled in Sitka, Alaska. Similar to  the fish hold effluent
results from Massachusetts, EPA found that the concentrations of the effluent from the
fish hold cleaning exceeded the NRWQC in one out of the three samples for E. coli, two
out of the three samples  for enterococci, and two out of three samples for fecal  coliform.
All the samples exceeding the most stringent screening benchmarks for the pathogens
were from the vessels located in Massachusetts. Pathogen concentrations  were below the
detection limit for all three pathogens for the fish hold  cleaning effluent from the vessel
in Sitka. In all cases, background concentrations in the ambient water exceeded the fish
hold cleaning effluent. Similar to what EPA observed with the fish hold effluent data,
pathogen contamination  in fish hold cleaning effluent from fishing vessels is not a likely
source of pathogen contamination to receiving waters.  Instead, EPA suspects that the
pathogen contamination  in these effluents might come  from the vessel pumping ambient
water with high levels of bacteria onboard.
                                       216

-------
                                                                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.6. Results of Fish Hold and Fish Hold Cleaning Effluent Sample Analyses for Pathogen Indicators1
Analyte2
Units3
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM4
Fish Hold
£. co// by MPN
Enterococci by MPN
Fecal Coliform by MF
MPN/100ml
MPN/100ml
CFU/100ml
7
7
11
6
5
6
86
71
55
83
380
11000
41
41
10






10


110
250
270
310
2200
1 00000
310
2200
130000
130
33
14
Fish Hold Cleaning
£. Co// by MPN
Enterococci by MPN
Fecal Coliform by MF
MPN/100ml
MPN/100ml
CFU/100ml
3
3
3
2
2
2
67
67
67
200
1000
1900
52
150
250









550
2800
5300
550
2800
5300
550
2800
5300
130
33
14
Notes:
 (1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on  detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) MPN = Most Probable Number; MF = Membrane Filtration.
(3) CFU = Colony Forming Units.
(4) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                          217

-------
                   Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     100000 -
       10000r
o
o
 CD
 o_
 o
O
1000r
                             Pathogen  Indicators
    Figure 3.4.12. Box and Dot Density Plot of Measured Pathogen Concentrations in
    Samples of Fish Hold Effluent
    (Note: All values were subtantially influenced by background concentrations in the ambient water, and of
    the 25 sample results presented (seven results for E. coli, seven for enterococci, and 11 for fecal conform),
    only two of the samples exceeded their background concentrations by more than 20 CFU/MPN 100 ml).
                                         218

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.4.4    Nutrients

       Samples offish hold effluent and fish hold cleaning were analyzed for four
nutrients or nutrient-related parameters:  ammonia nitrogen, nitrate/nitrite, TKN, and total
phosphorus (see Table 3.4.7). The corresponding nutrient concentrations detected in fish
hold and fish hold cleaning effluent samples are shown in Figures 3.4.13 and 3.4.14,
respectively.

       Concentrations of total ammonia nitrogen (NH3-N), nitrate/nitrite nitrogen
(NO3/NO2-N), TKN, and total phosphorus roughly compare to values of untreated raw
sewage (see values in Table 3.4.8). The fish hold effluent had average ammonia
concentrations of approximately 12 mg/L and the fish hold cleaning effluent had average
concentrations of 16 mg/L, which compares roughly to weak sewage as reported by
Metcalf and Eddy (1979) (see Table 3.4.8). However, there were several discharges in
which the ammonia concentration substantially exceeded these concentrations, and these
discharges could potentially result in acute toxic effects in the receiving water at and near
the point of discharge (see Figure 3.4.13). These high values increase the average
considerably  (the median values for fish hold and fish hold cleaning effluent are 2.1 and
4.8 mg/L, respectively). Most of the ammonia concentrations in  samples collected from
both fish hold and fish hold cleaning effluent exceed the PHQ screening benchmark of
1.2 mg/L based on the freshwater chronic aquatic life criterion of 1.2 mg N/L, with the
highest concentration resulting in a PHQ value of over 130.

       In contrast, average nitrate concentrations were near zero for both fish hold
effluent (maximum concentration of 0.39 mg/L) and fish hold cleaning effluent
(maximum concentration of max 0.53 mg/L). These concentrations are similar to those
expected in raw sewage effluent no matter the strength of the sewage  effluent (see Table
3.4.8). However, the average total phosphorus concentrations of 13 mg/L for the fish hold
effluent and 8.5 mg/L for fish hold cleaning effluent were similar to concentrations in
medium to strong raw sewage  (see Tables 3.4.7 and 3.4.8).

       TKN values averaged 110 mg/L  for fish hold  effluent and 59 mg/L for fish hold
cleaning effluent. These TKN results31 can be roughly compared with total nitrogen
results from Metcalf and Eddy (1979), showing that the nitrogen discharges are roughly
equivalent to strong sewage.

       Protein, free amino acids, and nucleotides from fish and fish by-products are all
potential sources of nitrogen. Inorganic phosphorus in the form of phosphate is a key
31 TKN includes ammonia (NH3) and ammonium (NH4+), and organic nitrogen values. Total nitrogen
includes ammonia (NH3) and ammonium (NH4+), organic nitrogen, and nitrate and nitrite values. Raw
sewage tends to have very low nitrate and nitrite values.

                                        219

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
element in DNA, RNA, and adenosine triphosphate (ATP) - key components present in
the tissue and blood of any animal.

       As shown in Figures 3.4.14 and 3.4.15, there is considerable variation exceeding
two orders of magnitude in the concentrations of three of the four nutrient and nutrient-
related parameters. EPA observed that nutrient concentrations showed some relationship
to the geographical location where the vessels operated. As shown in Figure 3.4.15,
concentrations of ammonia, TKN, and TP from the Gulf Coast shrimp boats and the New
England ground fishery trawlers appear to be higher than those from the fishing vessels
sampled in Alaska or the New England lobster tank. In addition, compared to the lobster
tank, whose water source is primarily flow-through water,  all fishing vessel platforms
appear to add nutrients to the effluent.
                                       220

-------
                                                                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.7. Results of Fish Hold (upper half) and Fish Hold Cleaning Effluent (lower half) Sample Analyses for Nutrients1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Fish Hold
Ammonia As Nitrogen (NH3-N)
Nitrate/Nitrite (NO3/NO2-N)
Total Kjeldahl Nitrogen (TKN)
Total Phosphorus
mg/L
mg/L
mg/L
mg/L
26
26
26
26
25
18
25
25
96
69
96
96
12
0.10
110
13
2.1
0.092
75
9.7




0.64

3.5
0.43
1.1

19
3.2
6.7
0.11
160
17
32
0.27
340
28
160
0.39
540
76
1.2
NA
NA
0.10
Fish Hold Cleaning
Ammonia As Nitrogen (NH3-N)
Nitrate/Nitrite (NO3/NO2-N)
Total Kjeldahl Nitrogen (TKN)
Total Phosphorus
mg/L
mg/L
mg/L
mg/L
9
9
9
9
7
8
6
7
78
89
67
78
16
0.24
59
8.5
4.8
0.27
40
11








0.034
0.070

0.025
18
0.35
140
17
97
0.53
170
20
97
0.53
170
20
1.2
NA
NA
0.10
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
Table 3.4.8. Raw Sewage Concentrations of Nutrients
Constituent
Ammonia as N
Nitrate as N
Total Nitrogen
Total Phosphorus
Concentration (expressed as me
Strong Sewage
50
0
85
15
Medium Sewage
25
0
40
8
/L)
Weak Sewage
12
0
20
4
Source: Metcalf and Eddy, 1979.
                                                                         221

-------
                 Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment



100.0
-- — •*•
— 1
***1^,
O)
E
§ 10.0
\£Z
m
%w
"c
CD
g 1.0
0
O


0.1


: ' '
oc
$ 0
_J
)0
o
o
= oooo
- Q- OOO
o
00
-
00 C
(
C)
-- oo c
: o d

)
)
)
o
- nn
w w
080
~

ooo
uuuu
— OQO
-
1 °
oo
[")
oo
- ^ OOuiJiJi, O

: ooo
1 0 1
1 :
™

—



0
oo
_
-
—
0
oo

oooo
ooo 	
oo ^
o -
OO'"' -
(

0
c
)

0 _
>
-


* -
=
1 =
                                   Nutrients
Figure 3.4.13. Box and Dot Density Plot of Nutrient Concentrations Measured in
Samples of Fish Hold Effluent
(Note: High maximum concentrations for certain samples for ammonia (160 mg N/L), total phosphorus (76
mg/L), and TKN (338 mg/L) are evident).
                                        222

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
    100.0
      10.0
 O)


 o
"co

 §      1.0
 o
O
        0.1
 \
 e


 o
oo
                       o
                                                      Q(1)Q
                                                       oo
                       C)
                                                        C)
                                  Nutrients
Figure 3.4.14. Box and Dot Density Plot of Nutrient Concentrations Measured in
Samples of Fish Hold Cleaning Effluent
(Note: For all parameters except ammonia, nutrient concentrations tend to be lower for fish hold cleaning
effluent).
                                      223

-------
                 Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
  1000
   100 -
    10-
c
o
I
"c
o
c
o
o
    1 -
   0.1 -
   0.01
              Nutrient Concentrations by Fishing Platform/Type
X
X
                        X

                                 *
                                            +
                                            X
                                           X
                                                              +
                                                              X
                                                              +
                                                 X Ammonia As Nitrogen
                                                 — Total Kjeldahl Nitrogen
                                                 + Total Phosphorus
                                                                        X

                                                                        $.
                                                                       X
            CSllnetter
                       Lobster
                       Tank
                                Longliner
                           Purse
                           Seiner
Shrimp
Trawler
Tender     N.E.
         Trawler
                                                                  +



                                                                  I
                                                                                X
                                                                                XN
                                                                               Trailer
Figure 3.4.15. Comparison of Concentrations of Ammonia, TKN, and Total
Phosphorus in Different Fishing Vessel Platforms to those in the Lobster Tank
(which has a live catch and continuously circulating water)
                                        224

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.4.5    Nonylphenols

       EPA analyzed three fish hold samples for nonylphenols. Short-chain nonylphenol
ethoxylates (e.g., NP2EO, NP1EO) and NP were not detected in any of these samples.
EPA expected this result because detergents should not be present when seafood catch is
stored in the vessel's fish hold compartment except for residual amounts from poor
rinsing after cleaning.

       As expected, several long chain nonylphenol and octylphenol ethoxylates (NPEOs
and OPEOs, respectively)) were detected in the fish hold cleaning samples collected from
eight vessels (see Table 3.4.9). As with deck washdown water, the NPEOs with the
longest ethoxylate chains were detected in approximately a third of the vessels, with
concentrations increasing as ethoxylate chain is reduced (i.e., concentrations increasing
from NP18EO to NP10EO). Of the vessels where long ethoxalate chain NPEOs were
detected, only one of the three vessels had detectable concentrations of NPEOs
representing the shortest chains (NP3EO through NP5EO); measured concentrations were
low in the  range of 12 to 32 |ig/L, respectively. The OPEO with the longest ethoxylate
chain (OP12EO) was detected in only one vessel, as were the lower ethoxylate chain
OPEOs. For OPEOs, the concentrations showed the same general trend as the NPEOs
with concentrations increasing as ethoxylate chain is reduced, although the
concentrations of the shorter chain OPEOs were much lower than the shorter chain
NPEOs.

       Total NPEO concentrations (from samples containing all 16 NPEO isomers)
could be calculated for only two of the eight vessels whose fish hold cleaning effluent
was sampled. The concentrations of total NPEOs ranged from 56 (a ground fishery
trawler in Massachusetts) to 4,540 |ig/L (another ground fishery trawler in
Massachusetts). These results are shown graphically in Figure 3.4.16.

       While there is no NRWQC for the sum of NPEOs or OPEOs, as indicated in
previous subsections, these compounds can degrade to NP in fresh and salt water (the
saltwater chronic aquatic life criterion for NP is only 1.7 |ig/L). EPA did not collect
samples of background levels for analysis of total NPEOs, OPEOs, and NP from ambient
or source water.
                                      225

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.9. Results of Fish Hold Cleaning Effluent Sample Analyses for Long-chain Nonylphenols1
Analyte
Total Nonylphenol Polyethoxylates
Nonylphenol octodecaethoxylate (NP18EO)
Nonylphenol heptadecaethoxylate (NP17EO)
Nonylphenol hexadecaethoxylate (NP16EO)
Nonylphenol pendecaethoxylate (NP15EO)
Nonylphenol tetradecaethoxylate (NP14EO)
Nonylphenol tridecaethoxylate (NP13EO)
Nonylphenol dodecaethoxylate (NP12EO)
Nonylphenol undecaethoxylate (NP11EO)
Nonylphenol decaethoxylate (NP10EO)
Nonylphenol nonaethoxylate (NP9EO)
Nonylphenol octaethoxylate (NP8EO)
Nonylphenol heptaethoxylate (NP7EO)
Nonylphenol hexaethoxylate (NP6EO)
Nonylphenol pentaethoxylate (NP5EO)
Nonylphenol tetraethoxylate (NP4EO)
Nonylphenol triethoxylate (NP3EO)
Octylphenol dodecaethoxylate (OP12EO)
Octylphenol undecaethoxylate (OP1 1 EO)
Octylphenol decaethoxylate (OP10EO)
Octylphenol nonaethoxylate (OP9EO)
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
samples
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
No.
detected
2
4
3
3
3
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
Detected
Proportion
(%)
25
50
38
38
38
25
25
25
25
25
25
25
25
25
13
13
13
13
13
13
13
Average
Cone.
620
1.6
3.1
6.9
14
25
39
56
75
75
73
74
66
51
32
21
12
2.8
2.7
4.5
4.9
Median
Cone.

0.15



















Minimum
Cone.





















10%





















25%





















75%
42
0.27
0.49
1.1
2.1
2.9
3.9
5.5
6.4
5.9
4.7
4.3
3.1
1.9







90%
4500
12
23
51
100
180
290
420
560
550
530
540
470
360
220
140
79
11
15
20
23
Maximum
Cone.
4500
12
23
51
100
180
290
420
560
550
530
540
470
360
220
140
79
11
15
20
23
Screening
BM2
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                           226

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     1000.00


73,   100.00


•S      10.00
 CD
"c
 §       1.00
 o
o
         0.10


         0.
                                I    I    I     I    I    I     I    I     I
                                          © ®  ©  ©
                                1    1    1     1    1    1     1    1     1
                      0   2   4   6   8   10  12 14  16  18 20  22
                                       Nonylphenols
Figure 3.4.16. Box and Dot Density Plot of Nonylphenol Concentrations Measured in
Samples of Fish Hold Cleaning Effluent
Nonylphenol parameters are identified as follows (nonylphenol and octylphenol ethoxylates in fish hold effluent
were not detected):
(1) Total Nonylphenol
Poly ethoxylates
(2) Nonylphenol
octodecaethoxylate (NP18EO)
(3) Nonylphenol
heptadecaethoxylate (NP17EO)
(4) Nonylphenol
hexadecaethoxylate (NP16EO)
(5) Nonylphenol
pendecaethoxylate (NP15EO)
(6) Nonylphenol
tetradecaethoxylate (NP14EO)
(7) Nonylphenol
tridecaethoxylate (NP13EO),
                          (8) Nonylphenol
                          dodecaethoxylate (NP12EO)
                          (9) Nonylphenol
                          undecaethoxylate (NP11EO)
                          (10) Nonylphenol
                          decaethoxylate (NP10EO)
                          (11) Nonylphenol
                          nonaethoxylate (NP9EO)
                          (12) Nonylphenol
                          octaethoxylate (NP8EO)
                          (13) Nonylphenol
                          heptaethoxylate (NP7EO)
                          (14) Nonylphenol
                          hexaethoxylate (NP6EO)
(15) Nonylphenol
pentaethoxylate (NP5EO)
(16) Nonylphenol
tetraethoxylate (NP4EO)
(17) Nonylphenol triethoxylate
(NP3EO)
(18) Octylphenol
dodecaethoxylate (OP12EO)
(19) Octylphenol
undecaethoxylate (OP11EO)
(20) Octylphenol decaethoxylate
(OP10EO)
(21) Octylphenol
nonaethoxylate (OP9EO)
                                             227

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       3.2.4.6    Regional variation in Fish Hold Effluent Discharges

       Based on public comments received on EPA's draft version of this report, EPA
conducted a regional analysis of vessel fish hold discharges. EPA was able to conduct this
analysis because there were 26 different hold discharges sampled. However, of the 26 fish hold
discharges sampled, twenty were from Alaska, four were from New England, and only two were
from the Gulf.  A sample size of two is the absolute minimum number that can be used for any
statistical comparisons,  and caution must be exercised before drawing any general conclusions
on the effects offish hold discharge for an entire geographic region based on only two samples.
Additionally, there is limited variation in the platforms sampled at New England and Gulf Coast
locations. EPA cautions these results must be considered preliminary in nature and cannot be
considered conclusive.  The limitations of this analysis are also discussed in the following
paragraphs.
       The potential regional
differences in concentrations offish
hold discharges were examined for
seven analytes (Total Copper, Total
Zinc, Total Arsenic, Ammonia, Total
Kjeldahl Nitrogen, Total Phosphorus,
and Biological Oxygen Demand).
Concentrations of each of these
analytes were compared for the three
regions in which vessels with fish hold
discharge were sampled (Alaska, Gulf
Coast,  and New England). Mean
analyte concentrations with
corresponding standard deviations are
shown for each of the regions in Table
3.4.10. Based on this preliminary
analysis, mean concentrations of all
seven analytes were lower in the fish
hold discharges from fishing vessels in
Alaska compared to the concentrations
discharged from fishing vessels in the
Gulf Coast and New England. A
preliminary statistical analysis
(Welch's t-tests accounting for
unequal variance and unequal sample
sizes - see accompanying text box for
details) comparing each of these
Regional Comparison of Fish Hold Discharge
Concentrations
A preliminary analysis was performed to assess the
effects of geographic region on seven selected analytes
listed in Table 3.4.10. Vessels were grouped into three
broad geographic regions, and concentration differences
between groups were evaluated using Welch's t-tests
accounting for unequal sample size and uneven variance.
Prior to analysis, all concentrations were log transformed
to stabilize sample variance. EPA performed t-test
analyses with and without subtracting background analyte
concentrations. The results were fundamentally similar.
For each analyte, three comparisons were made (Alaska-
Gulf, Alaska-New England, Gulf-New England), at a
Bonferroni adjusted significance level of .017 (.05/3), to
account for the effect of multiple comparisons.

Two additional analyses were performed to examine
whether the observed regional differences could be
explained by differences in fishing method (nets vs. no
nets), or fish hold cooling method (ice vs. refrigerated
seawatervs. both). These analyses also consisted of
Welch's t-tests for unequal sample size and unequal
sample variance, and were conducted using log
transformed concentrations after subtracting ambient
concentration with appropriate Bonferroni adjusted
significance levels to account for the effect of multiple
comparisons.
                                            228

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

groups suggests that regional differences in the fish hold discharge concentrations for these
seven analytes might be present, as concentrations from the two Gulf Coast vessels (shrimpers)
were statistically significantly higher than concentrations from the twenty Alaska vessels for six
of the seven analytes tested. The one exception was total arsenic, which did not differ
significantly between any of the three regions.

       Although these results suggest the possibility of regional differences, they may also
simply be a statistical artifact of: 1) a population of vessels that were both small in number and
highly unevenly distributed across regions, 2) vessel type, 3) fish hold cooling method, 4) fishery
type, or 5) some combination of the above. Both of the Gulf Coast vessels were shrimp trawlers,
and three of the four New England vessels were ground fishery trawlers. Trawling vessels may
also fish closer to the bottom of the water column, and could be expected to accumulate more
organic matter along with their catch, which could potentially  explain the higher concentrations
of the analytes examined in this analysis. A fourth New England sample was taken from a
lobster tank which consistently circulated ambient water.  The twenty Alaska vessels consisted of
five purse seiners, three gillnetters, three longliners, three tender vessels, and six trailers.  The
purse seiners and  gillnetters spread nets, which may tend to pull fish or other material from
closer to the bottom of the water column (though this is less likely in many deep waters off of
Alaska). A second analysis was performed to determine whether the regional differences
observed were an artifact of the distribution of vessel type; specifically, whether the vessel
fishing method employed nets (trawlers, purse seiners, gillnetters), or some other method (lobster
vessel, longliners, trailers). When the vessels were analyzed in two groups, those that do not use
nets versus those that use nets, concentrations of total ammonia were statistically significantly
higher, and concentrations of total arsenic and total copper were marginally (0.10
-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

be explained by fish hold cooling method. However, this result cannot be separated from the
effects of fishing method (nets vs. no nets), as seven of seven vessels that cooled fish holds with
refrigerated seawater also fished with nets, while nine often vessels that cooled with ice used
fishing methods that did not involve nets.

       Although results of this analysis suggest differences in fish hold discharge concentrations
are more pronounced between regions than between fishing method or fish hold cooling method,
these results should be considered preliminary, and additional information will be required to
draw any substantive conclusions regarding inter-region differences.  Both the Gulf Coast and
New England regions are represented by a small number of sample vessels. Not only is the Gulf
Coast region represented by the minimum number of vessels with which to perform any
statistical comparisons, the two vessels are similar with regard to vessel type and fish hold
cooling method; therefore it cannot be assumed that the two sampled discharges are
representative of the entire Gulf Coast fishery. While these analyses  suggest the possibility of
regional differences, the presence of true regional differences would require the sampling of a
larger number of vessels  from the Gulf and New England regions encompassing a broader, more
evenly distributed number of vessels to account for the effects of vessel and discharge (ice or
refrigerated water) type, as well as additional sampling of ambient receiving waters.
Table 3.4.10. Means (and Standard Deviations) for Selected Analyte Concentrations, by
Geographic Region. Units for All Analytes Expressed as ug/L, Except for BOD (mg/L).
Region
Alaska





Gulf Coast


New England

Mean Analyte Concentration Above Ambient (1 s.d.)
Total Cu
53.7
(69.0)




1640
(30.7)

112
(146)
Total Zn
253
(223)




446
(75.5)

706
(773)
Total As
4.79
(3.69)




186
(233)

132
(170)
Ammonia
2.72
(2.74)




19.0
(1.39)

55.2
(74.2)
TKN
75.4
(64.4)




397
(201)

165
(245)
Total P
8.40
(6.35)




24.7
(8.73)

26.0
(33.7)
BOD
416
(336)




3250
(499)

1720
(2370)
Vessel Type (no.)
Gillnetter (3)
Longliner (3)
Purse Seiner (5)
Tender Vessel (3)
Trailer (6)

Shrimp Trawler (2)


Lobster Tank (1)
Ground Fishery Trawler (3)
*As discussed in the text above, there are substantial limitations to this regional analysis which mean these results are preliminary in
nature. Additional information is needed before making firm conclusions.
                                           230

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.4.7     Summary of the Characterization of Fish Hold Effluent and Fish Hold
                  Cleaning Effluent

       Table 3.4.11 summarizes the specific analytes within fish hold and fish hold cleaning
effluent water that may have the potential to pose risk to human health or the environment.
EPA's interpretation of a realized risk likely posed by these analytes, relative to pollutant
loadings, background ambient and source water contaminant levels and characteristics, and other
relevant information useful for this assessment, is presented in Chapter 5.

       Total iron was sampled for in only two vessels, but PHQs were between 5 and 10.
Concentrations of dissolved copper exceeded NRWQC standards in all effluents sampled, with
PHQs>10 in four of the vessels sampled.

       The concentrations of certain total and dissolved metals, as well as many of the other
pollutants, measured in fish hold and fish hold cleaning effluent were elevated. Concentrations of
total arsenic were detected in 16 of 26 samples, and when detected were measured at levels
greatly exceeding its respective screening benchmark (i.e., NRWQC), resulting in PHQs of well
over 100.  Likewise, total copper concentrations, while only exceeding the NRWQC for human
health of 1,300 |ig/L in a few samples, were high in these few instances and might pose potential
acute toxicity risk to aquatic life32. To a large degree, total aluminum, iron, and manganese
concentrations could be explained by the respective metal concentrations in the surrounding
waters. Arsenic and copper, however, most likely originated from the fish hold effluent.
Concentrations of dissolved copper exceeded NRWQC standards in all effluents sampled, with
PHQs well above 10 in four of the vessels sampled. Samples with concentrations of dissolved
arsenic resulting in PHQs above 10 were limited to just two fishing vessels (a shrimper33 and a
ground fishing trawler) with a third vessel having a PHQ of approximately 2. Approximately 2/3
of the concentrations of dissolved zinc in fish hold effluent exceeded NRWQC benchmarks, but
no concentrations of dissolved zinc exceeded a PHQ of 10, and most concentrations were below
a PHQ of 3. Dissolved selenium was measured above reporting limits in only six discharges with
PHQs>l in all samples, and PHQs between 5 and 10 for two samples (including  a shrimping
vessel33). Total and dissolved metals concentrations were qualitatively similar in fish hold
cleaning effluents, but, in general, concentrations in cleaning effluent were lower than in
corresponding fish hold effluents.

       Several classical pollutants found in fish hold and fish hold cleaning effluent may have
the potential to pose risk. A classical pollutant found in fish hold and fish hold cleaning effluent
32 As discussed earlier in this chapter, total copper concentrations could pose potential risk to the aquatic
environment because the human health criteria of 1,300 ug/L is significantly higher than the 3.1 ug/L benchmark
used for dissolved copper based on the saltwater chronic ambient water quality criterion for the protection of aquatic
life. When high levels of paniculate copper are discharged, some of the paniculate copper will likely convert to
dissolved copper and be made bioavailable to aquatic life.
33 See discussion in footnotes 29 and 30 and Section 3.2.4.1.

                                            231

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

that poses one of the greatest potential risks to receiving waters is BOD, which was found at
elevated concentrations in all sampled vessels and, in many instances, was higher than
concentrations found in raw sewage (see Fig. 3.4.12). Concentrations of COD and  TOC
correlated with BOD concentrations and were similarly elevated in all fishing vessels. The high
BOD in these samples likely contributed to the pervasively low dissolved oxygen levels in these
samples. TSS and turbidity in fish hold and fish hold cleaning effluent are also equivalent to
levels found in raw sewage, and concentrations of sulfide, particularly in samples from the New
England ground fishery trawlers, exceeded the low PHQ screening benchmark (0.002 mg/L) for
this classical pollutant.

       The other pollutants of potential concern in fish hold and fish hold cleaning effluent were
the nutrient and nutrient-related parameters, particularly NH3-N, TKN,  and TP, all of which
were measured at concentrations similar to comparable concentrations typically measured in
strong (raw) sewage samples. Again, mean concentrations of BOD, COD, TOC, NH3-N, TKN,
and TP were highest in shrimping and trawling vessels.

       Aside from a select few samples, the high pathogen concentrations found in fish hold and
fish hold cleaning samples likely did not stem from the effluent itself, but rather, from the
excessively high concentrations measured in ambient background water contaminating the fish
holds from the deck washdown process.
                                           232

-------
                                                                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.4.11. Characterization of Fish Hold Effluent and Fish Hold Cleaning Effluent and Summary of Analytes that May
Have the Potential to Pose Risk









Vessel Type (no. vessels)









Fishing Vessels (31)


Analytes that May Have the Potential to Pose Risk in Fish Hold and Fish Hold Cleaning Effluent1










o
0)
IS
o




I/I
•o
3
o
Q.
E
o
o
u
'E
re
S1
o

0
IS
1
p
0
CO











__ _
V
.>
o
I/)
I/)

j2
re
ID
^
As.Cu, Zn














S
0
j2
re
'S
^
AI.AS,
(Cu)2,Fe











S
re
V
3
•O
C
re
—
O

















ID
;o

"5
CO
X








o
c
ID
^"D
< re

.E <"
re £z
^ re
9 >
0 °
^ •c
CO LLJ









Q
c
ID
Q.
>
•£

C «
re +j
•c J2

? o
o £
-J UJ
















U)
'c
0)
'Z
'S
•z.
NH3-N
TKN
TP








O
o
*~
c
re

o
O
Q
o
m
BOD
COD
TOO






1/1
;o
"o
w
•o
•D
C
Q.
U)
w
re
'o
H
X








re
u
1
O
15
u

^5
Q_ OJ
S E
^ re
*- re
O Q.
DO,
TRC,
turbidity
(1) Analytes are generally bolded when a large proportion of the samples have concentrations exceeding the NRWQC (e.g., 25 to 50 percent), when several of the samples have PHQs > 10 (e.g.,
two or three of five), when a few samples result in PHQs greatly exceeding the screening benchmark (i.e., 100s to 1,000s), in the case of oil and grease and for nonylphenol, when one or more
samples exceed an existing regulatory limit by more than a factor of 2, or when concentrations of analytes are sufficiently high that they may have the potential to pose risks to local water bodies.
See text in Section 3.1.3 for a definition of PHQs and Table 3.1 for screening benchmarks used to calculate these values.
(2) Only a few PHQs near or slightly exceeding 1, but concentrations (in excess of 1,000 ug/L) potentially acutely toxic to aquatic life, particularly to organisms living in the benthos.
                                                                         233

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

   3.2.5  Graywater

       EPA sampled graywater from eight vessels: five tugboats, a shrimper, a water taxi and a
recreational powerboat. The samples included graywater from sinks, dishwashers, and showers,
as well as graywater samples from several mixed or unspecified sources. Graywater samples
were analyzed for a range of pollutants including pathogen indicators, classical pollutants,
nonylphenols, metals, and nutrients. The analytical results were intended to provide
representative graywater pollutant concentrations over the range of normal vessel operations.

       Graywater volumes vary  considerably depending on the class of vessel and its intended
use, vessel size, the number of crew and passengers onboard, and the types of graywater-
generating activities onboard (e.g., galleys, sinks, showers, wash machines). Based on
observations made during the sampling program and from discussions with crew members, EPA
estimated that tugboats, some of which provide living quarters for three to five crew members,
generate approximately 130 gpd of graywater. Water taxis, which carry a significantly larger
number of crew and passengers,  but with fewer graywater-generating activities, generate
approximately 75 gpd of graywater. Graywater generation on commercial fishing boats might
range from a few gpd to hundreds of gpd, depending on the length of the trip and the size of the
crew. Due to the highly variable  graywater generation volumes possible within vessel classes,
EPA was unable to further define graywater generation rates.
                           The Sink and Shower Facilities of a Tugboat
                                          234

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.5.1    Pathogen Indicators (Microbiological)

       Graywater is generated from personal bathing, food preparation, and dish and clothes
washing, so EPA expected that this vessel discharge category could contain high levels of
pathogens. The analytical data for the pathogen indicator bacteria E. coli, enterococci and fecal
coliform confirm this expectation as the levels of pathogens measured in graywater were by far
the highest values measured in any  of the vessel discharges. However, it should also be noted
that for each of the pathogen indicators, a wide range of values were measured in the graywater
samples. EPA also noted that source water (generally municipal water transferred onto the vessel
(service water)) does not appear to account for any of the pathogen concentrations.

       The analytical results for pathogen indicators in the eight graywater samples are
summarized in Table 3.5.1 and displayed in Figure 3.5.1. For each of these parameters, the
highest levels (660,000 MPN/100 mL for E.  coli, 240,000 MPN/100 mL for enterococci, and
570,000 CFU/100 mL for fecal coliform) were measured in the mixed graywater sample from a
tugboat. For comparison, EPA measured average levels of 292,000 MPN/100 mL for E.  coli,
8,920 MPN/100 mL for enterococci, and 36,000,000 CFU/100 mL for fecal coliform in untreated
graywater, as reported in the 2008 Cruise Ship Discharge Assessment Report (USEPA, 2008).
Typical fecal coliform concentrations in untreated domestic wastewater are 10,000 to 100,000
MPN/100 mL34. The second highest concentration, of E. coli, was measured in a mixed
(dish/shower) graywater sample, while the second highest concentrations, for enterococci and
fecal coliform, were measured in a dishwashing sample. Samples of graywater from sinks  and
showers tended to have lower levels of the pathogen indicators. Pathogen indicators were not
detected in graywater samples from the  sink of one vessel, a water taxi.

       Figure 3.5.2 presents in box/scatter plots the PHQs for the three pathogen indicators in
graywater. As this figure shows, the majority of the values measured for each of the pathogen
indicators exceeded the water quality screening benchmarks, by up to four orders of magnitude
(or more, in the case of fecal coliform).
34 Note, as indicated above in Table 3.1 and elsewhere, units of MPN/100 ml for fecal coliform approximate similar
units of CFU/100 ml; therefore, the two units of expression are appropriate for comparison here.
                                           235

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.5.1. Results of Graywater Sample Analyses for Pathogen Indicators1
Analyte
£. Co//
Enterococci
Fecal Coliform
Units2
MPN/100ml
MPN/100ml
CFU/100ml
No.
samples
8
8
8
No.
detected
7
7
7
Detected
Proportion (%)
88
88
88
Average
Cone.
1 1 0000
40000
200000
Median
Cone.
16000
500
270000
Minimum
Cone.



10%



25%
180
70
74
75%
120000
57000
450000
90%
660000
240000
570000
Maximum
Cone.
660000
240000
570000
Screening
BM3
130
33
14
Notes:
 (1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage  of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) MPN = Most Probable Number; CFU = Colony Forming Units.
(3) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                           236

-------
                    Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


100000
1
E 10000
0
0
T —
5 1000
Q_

1 100

10
1

_ ^
-
—
—
-
~
—
-
-
(

r i op i
Q
)
O
O


o

o

r c
-
—
	 f
c






)

)




0
o
OO


C
\ f







OO


0


) (
L
^ v_/

-
-=
—
—
~
—
-
-
—
—
) —















0°^ cf& o^°
>". ^ _rS> . rv
                             <^       **"'
                          Pathogen Indicators
Figure 3.5.1. Box and Dot Density Plot of Pathogen Indicator Values Measured in Samples
of Graywater
                                     237

-------
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


10000.00
"c
~ 1000.00
o
^ 100.00
03
|! 10.00


03
"-£ 1.00
CD
-1— •
o
Q- 0.10
0.01

: I I ufr> :
-
r o
-
=-
-
=-
-
__
z
-
-
/->
c
/-N
Uv^

O

O

o


= c
I
_
-
=-
	 x'








)



o
0
oo




)

c

/d




uu










o


n




d
•N*.
)

^N
Cs^
^ I


o
-
-=
-
-=
-
-=
-
__
z
-
-
jr 	 =
—
_
-
-=
-


















                               <&p        <<*
                           Pathogen Indicators
Figure 3.5.2. Box and Dot Density Plot of Potential Hazard Quotients for Pathogen
Indicators Measured in Samples of Graywater
(Note: Replacement values for non-detects are circled).
                                       238

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       3.2.5.2    Classical Pollutants

       Graywater samples were analyzed for 14 classical pollutants (see Table 3.5.2). Figure
3.5.3 illustrates the variability of the concentrations/values measured for the classical pollutants
in graywater. There was no one vessel or graywater source that tended to have the highest level
of a majority of the classical pollutants, unlike the case for the pathogen indicators. The highest
concentrations of oil and grease (100 mg/L HEM and 35.3 mg/L SGT-HEM) were measured in
the sample of mixed dish/shower graywater on one tugboat; EPA speculates that the source of
the oil and grease are primarily oils from cooking and other food sources discharged with the
sink water. The highest levels of TSS (99 mg/L) and turbidity (128 NTU) were measured in the
dishwashing graywater from a second tugboat. The highest sulfide concentration (1.45 mg/L)
was measured in a shower graywater sample from a third tugboat.  The highest measured
concentrations of BOD (1200 mg/L), COD (4,040 mg/L), and TOC (440 mg/L) were measured
in the sample  of shower graywater from the recreational powerboat.

       Many  of the classical  pollutants that were elevated in the graywater samples likely reflect
the washing and bathing activities that generate graywater discharges. For example, sulfide35 is a
parameter that is commonly elevated in water distribution systems, especially on the hot water
side. Sulfur-reducing bacteria, which use sulfur as an energy source, are the primary producers of
large quantities of hydrogen sulfide. Sulfur-reducing bacteria can live in plumbing systems and
hot water heaters. A second example is the  high concentration of BOD measured in graywater
samples (mentioned above), which reflects the BOD generated  onboard the vessels sampled and
not from the service water used by that vessel.

       Figure 3.5.4 presents the PHQs for classical pollutants in graywater in box/scatter plots.
As this figure shows, the PHQ threshold of 1  was exceeded for sulfide, TRC (detected in only
one  sample (0.11 mg/L) above the reporting limit of 0.0075 mg/L for a PHQ of 15), BOD, oil
and grease (measured as HEM), and TSS. The highest PHQs were calculated for sulfide at 367
and BOD at 40. All of the graywater samples exceeded the 30 mg/L benchmark for BOD, and all
five of the detected concentrations  of sulfide exceeded the 0.002 mg/L benchmark.

       The source of water used on the sampled vessels was, in all cases, potable freshwater
bunkered in port (service water). Therefore, EPA did not consider it appropriate to compare the
  Although sulfide (S 2") is the analyte, hydrogen sulfide (H2S) is the nonpriority pollutant for which a NRWQC has been
established. Sulfides are commonly found as either hydrogen sulfide or hydrosulfide (HS").  EPA conservatively assumes that all of
the sulfide is in the form of hydrogen sulfide (H2S) is the form that is toxic to fish. However, the proportion of each depends on the
pH of the water. At pH 9 about 99 percent of the sulfide is in the form of HS"; at pH 7 the sulfide is equally divided between HS" and
H2S; and at pH 5 about 99 percent of the sulfide is present as H2S.  Unless heavily polluted, freshwater rivers typically tend to have
a pH which ranges from about 4.5 to about 7, marine environments have an average pH of around 8.1 (seawater is more basic
freshwater), while estuaries may have a pH between that of freshwater and seawater (approximately 5 to 8) dependent upon salinity
and other factors. Hence, the use of sulfide (S 2") as the analyte to detect for the presence of hydrogen sulfide (H2S) is more
conservative in marine and estuarine environments than in freshwater ones, but is a reasonable analyte to use due to variation
found in different aquatic ecosystems.


                                             239

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


concentrations of classical pollutants in graywater to ambient water body concentrations; rather,
EPA compared the concentrations of classical pollutants to those found in the service water.
None of the conventional parameters discussed here were consistently detected in service water.
                                             240

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.5.2. Results of Graywater Sample Analyses for Classical Pollutants1
Analyte
Biochemical Oxygen Demand
(BOD)
Chemical Oxygen Demand
(COD)
Conductivity
Dissolved Oxygen
Hexane Extractable Material
(HEM)
PH
Salinity
Silica Gel Treated HEM (SGT-
HEM)
Sulfide
Temperature
Total Organic Carbon (TOC)
Total Residual Chlorine
Total Suspended Solids (TSS)
Turbidity
Units
mg/L
mg/L
mS/cm
mg/L
mg/L
SU
ppt
mg/L
mg/L
C
mg/L
mg/L
mg/L
NTU
No.
samples
8
8
7
7
8
8
6
8
8
8
7
8
8
8
No.
detected
8
8
7
7
8
8
6
6
5
8
7
6
8
8
Detected
proportion
(%)
100
100
100
100
100
100
100
75
63
100
100
75
100
100
Average
Cone.
430
1000
0.43
7.4
39
7.4
0.25
8.1
0.11
27
140
0.12
52
74
Median
Cone.
260
440
0.41
7.1
29
7.2
0.20
1.5
0.017
27
83
0.020
58
89
Minimum
Cone.
99
180
0.22
6.0
9.4
6.1
0.10


21
27

14
40
10%
99
180
0.22
6.0
9.4
6.1
0.10


21
27

14
40
25%
110
270
0.30
6.3
14
6.7
0.18
0.33
0.0
24
66

37
45
75%
850
1700
0.50
8.3
68
8.5
0.40
9.4
0.035
29
160

69
110
90%
1200
4000
0.79
10
100
8.7
0.40
35
0.73
36
440
0.11
81
110
Maximum
Cone.
1200
4000
0.79
10
100
8.7
0.40
35
0.73
36
440
0.11
81
110
Screening
BM2
30
NA
NA
NA
15
NA
NA
15
0.0020
NA
NA
0.0075
30
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            241

-------
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
1000.000
§ 100.000
E
^ 10.000
o
1 1.000
CD
-1— •
£ 0.100
o
8 0.010
0.001
 %
e OOOSOOO
= oo^oo rn^ oosSo^
ooo
I 06 doL I
Q Q -
oc o
P 	 I , , , ,i

                            Classical  Pollutants
Figure 3.5.3. Box and Dot Density Plot of Classical Pollutant Concentrations/Values
Measured in Samples of Graywater
                                       242

-------
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       100.0
   o
   ^

  a

  •a
   i_
   03
   N
   03
  Q_
10.0
           1.0
          0.1
-
—
~
— ,
-
"

-

-
—


3
C
3_
C
a






0
)
c
)
o




1 1 1





(


1 1 1
1


Q
i
0
^€

I









t
1




Cc


D
3
T


i
o
0
o
Zi



1
1




2>



1 C

c

(
c
c

3

2

)

)
)
)

C

5
1 1 1 1 _
—
~
—
-
op
PM
?
i
^ 1 1 1 1 =
                 c\^ c\^ <^ r^ r^ r^ o^ c^ c\^ c\^ c^ vCJV\SW

               V^V^0^?^^ ^V^V%%3' \^U^
                O^  v^oQ ^-v^x A^ \^' ^v- XO oO c^xC^X)'   v


               "o0^        ^°                      ^
                            Classical Pollutants
Figure 3.5.4. Box and Dot Density Plot of Potential Hazard Quotients for Classical

Pollutants in Samples of Graywater

(Note: Replacement values for non-detects are circled).
                                        243

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.5.3    Nonylphenols

       Long- and short-chain nonylphenol and octylphenol ethoxylates and NP were expected in
graywater discharges given their use in soaps for hand and body washing and in liquid detergents
for dish washing. EPA anticipated that the long-chain alkylphenol ethoxylates would be present
in all graywater samples where detergents were used for cleaning, while the short-chain
ethoxylates (and possibly NP) would be present if detergents were used for cleaning and there
was a graywater holding tank that provided the additional residence time necessary for biological
activity to degrade the NPEOs and OPEOs.

       Graywater samples were analyzed for 34 long- and short-chain nonylphenol and
octylphenol ethoxylates, including 28 NPEOs and OPEOs, bisphenol A, and total nonylphenol
(NP).  Of these parameters, 25 were detected in one or more samples (see Table 3.5.3).  Average
concentrations for NP18EO-NP3EO and OP12EO-OP6EO ranged from approximately 0.1 to 10
|ig/L.  The average concentrations of total nonylphenol polyethoxylates (sum of NPEO isomers)
and total octylphenol polyethoxylates (sum of OPEO isomers) were 66 and 63 |ig/L,
respectively. All of the NPEOs were detected in the graywater sample  from the sink of one of the
tugboats and the graywater sampled from the shower on the recreational powerboat. All of the
OPEOs were detected in the graywater sampled from the shower on the recreational powerboat.
NPEOs and OPEOs were also occasionally detected in graywater samples from three of the other
vessels.

       EPA did not calculate any PHQs for the nonylphenol parameters measured in graywater.
The only screening benchmark available was the  saltwater chronic NRWQC for NP (1.7 |ig/L).
There were no analytical results for NP to compare to this screening benchmark, and no
NRWQC exist for the other nonylphenol parameters (individual or total long- and short-chain
NPEOs and OPEOs). None of the long- or short-chain nonylphenol or  octylphenol ethoxylates or
NP were detected in  the ambient water surrounding these vessels.
                                         244

-------
                                                                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.5.3. Results of Graywater Sample Analyses for Nonylphenols (only long-chain NPEOs and OPEOs were detected)1
Analyte
Total Nonylphenol Polyethoxylates
Nonylphenol octodecaethoxylate (NP18EO)
Nonylphenol heptadecaethoxylate (NP17EO)
Nonylphenol hexadecaethoxylate (NP16EO)
Nonylphenol pendecaethoxylate (NP15EO)
Nonylphenol tetradecaethoxylate (NP14EO)
Nonylphenol tridecaethoxylate (NP13EO)
Nonylphenol dodecaethoxylate (NP12EO)
Nonylphenol undecaethoxylate (NP1 1 EO)
Nonylphenol decaethoxylate (NP10EO)
Nonylphenol nonaethoxylate (NP9EO)
Nonylphenol octaethoxylate (NP8EO)
Nonylphenol heptaethoxylate (NP7EO)
Nonylphenol hexaethoxylate (NP6EO)
Nonylphenol pentaethoxylate (NP5EO)
Nonylphenol tetraethoxylate (NP4EO)
Nonylphenol triethoxylate (NP3EO)
Total Octylphenol Polyethoxylates
Octylphenol dodecaethoxylate (OP12EO)
Octylphenol undecaethoxylate (OP1 1 EO)
Octylphenol decaethoxylate (OP10EO)
Octylphenol nonaethoxylate (OP9EO)
Octylphenol octaethoxylate (OP8EO)
Octylphenol heptaethoxylate (OP7EO)
Octylphenol hexaethoxylate (OP6EO)
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
samples
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
No.
detected
2
2
3
3
3
3
3
3
3
2
2
2
1
1
2
2
1
1
4
2
2
1
1
1
1
Detected
Proportion (%)
25
25
38
38
38
38
38
38
38
25
25
25
13
13
25
25
13
13
50
25
25
13
13
13
13
Average
Cone.
66
0.084
0.31
0.59
1.1
2.2
3.5
5.4
7.0
6.7
7.3
7.9
7.8
7.3
5.8
4.7
2.8
63
1.5
2.0
3.5
3.3
7.6
10
10
Median
Cone.


















0.22






Minimum
Cone.

























10%

























25%

























75%
15
0.023
0.12
0.23
0.49
0.95
1.9
3.2
4.7
2.0
2.5
1.5


1.6
1.1


3.3
3.1
4.1




90%
53
0.041
1.0
1.6
2.4
5.8
9.3
14
16
6.9
7.3
7.6
6.5
5.5
3.7
2.7
0.99
37
3.5
5.2
7.2
7.8
7.3
6.3
4.1
Maximum
Cone.
53
0.041
1.0
1.6
2.4
5.8
9.3
14
16
6.9
7.3
7.6
6.5
5.5
3.7
2.7
0.99
37
3.5
5.2
7.2
7.8
7.3
6.3
4.1
Screening
BM2
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                          245

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.5.4    Metals

       Graywater samples were analyzed for dissolved (filtered) and total concentrations of
metals. The analytical results are summarized in Table 3.5.4 for the dissolved metals and Table
3.5.5 for the total metals that were detected in at least one graywater sample. The following
metals were detected in all of the graywater samples:

          •   Dissolved and total aluminum
          •   Total barium
          •   Dissolved and total calcium
          •   Dissolved and total copper
          •   Dissolved and total manganese
          •   Dissolved and total potassium
          •   Dissolved and total sodium
          •   Dissolved and total zinc.

Concentrations of other metals were measured in 50 percent or more of the graywater samples:

          •   Dissolved barium
          •   Total chromium
          •   Total iron
          •   Total lead
          •   Dissolved and total magnesium
          •   Dissolved and total nickel
          •   Dissolved and total selenium
          •   Total vanadium.

       Figures 3.5.5  and 3.5.6 present the ranges of concentrations measured for dissolved and
total metals in the graywater samples. The plots show that dissolved and total metals
concentrations range over five orders of magnitude. Calcium, magnesium, potassium and
sodium, which are the major cations present in seawater, were the dissolved metals measured at
the highest concentrations. Dissolved aluminum, copper, and zinc were also measured at
relatively high concentrations  (greater than 100 |ig/L) in most graywater samples. For these
dissolved metals, service water samples contained up to 80 percent of the graywater
concentration for aluminum, up to 100 percent for copper, and up to 170 percent for zinc.
Although the comparison of service water and graywater concentrations suggests that service
water might be the source of these metals in some of the graywater samples, this was not always
the case. In fact, service  water concentrations tended to be low in the samples that corresponded
to the highest metals  concentrations in graywater.
                                          246

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       Total concentrations for each metal were generally similar to or somewhat higher than
the dissolved concentrations. Aside from the major seawater cations, concentrations of total
metals in the graywater samples were highest for aluminum (912 |ig/L), copper (440 |ig/L), iron
(458 |ig/L), and zinc (3,470 |ig/L). For these total metals, EPA found that service water samples
contained up to 74 percent of the graywater concentration for aluminum, up to 115 percent for
copper, up to 175 percent for iron, and  up to 32 percent for zinc. As was the case for dissolved
metals, comparing the service water and graywater concentrations suggests that service water
might be the source  of these total metals in some, but not all, of the graywater samples.

       To quantify the relationship between dissolved and total metals concentrations, EPA
calculated the average dissolved fraction (fd) of each metal in the graywater samples. The metals
in graywater discharges with the highest average dissolved fractions (fd> 90 percent) included
arsenic, calcium, magnesium, nickel, potassium, and sodium. For all of the other metals where
dissolved fractions could be calculated (aluminum, barium, chromium, copper, iron, lead,
manganese, selenium, vanadium,  and zinc), the average values were in the intermediate (90
percent > fd > 50 percent) range.

       The plots in Figures 3.5.7 and 3.5.8 display the distribution of PHQs based on the
screening benchmark for each of the dissolved and total metals. For dissolved metals, copper and
zinc concentrations consistently exceed the screening benchmarks; the maximum PHQs for
copper and zinc were 90 and 18.5, respectively. For total metals, the measured concentrations of
aluminum consistently exceeded the screening benchmarks. PHQs for total aluminum varied
from 0.6 to 10.5. The PHQs based on measured concentrations of total arsenic were 160 and 110
(arsenic was detected in  only two of eight graywater samples, one of which may have an
elevated measured concentration due to positive interference(see discussion on page 74 for more
information); these high values reflect the very low NRWQC (0.018 |ig/L; human health for the
consumption of water +  organism) for this carcinogen.
                                          247

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.5.4. Results of Graywater Sample Analyses for Dissolved Metals1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion (%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum
Arsenic
Barium
Chromium
Copper
Iron
Lead
Manganese
Nickel
Selenium
Thallium
Vanadium
Zinc
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
7
8
3
8
8
3
8
8
8
8
3
3
8
7
2
2
2
8
1
4
8
4
1
1
1
8
100
25
67
25
100
33
50
100
50
13
33
33
100
190
1.9
26
1.4
55
83
2.5
17
5.5
3.5
0.80
0.73
400
160

27

17

1.1
8.8
2.1



240
24



5.3


4.7




70
24



5.3


4.7




70
86



7.6


6.4




80
300
1.1
45
1.4
60
150
4.2
35
70

1.4
1.2
610
460
4.5
45
2.2
280
150
6.0
42
9.8
1.4
1.4
1.2
1500
460
4.5
45
2.2
280
150
6.0
42
9.8
1.4
1.4
1.2
1500
NA
36
NA
11
3.1
NA
2.5
NA
8.2
5.0
NA
NA
81
Cationic Metals
Calcium
Magnesium
Potassium
Sodium
ug/L
ug/L
ug/L
ug/L
8
8
3
3
8
7
3
3
100
88
100
100
34000
9400
5500
79000
33000
11000
5700
48000
1800

4100
31000
1800

4100
31000
25000
6600
4100
31000
36000
13000
6700
160000
81000
18000
6700
160000
81000
18000
6700
1 60000
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            248

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.5.5. Results of Graywater Sample Analyses for Total Metals1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion (%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
Heavy and Other Metals
Aluminum
Arsenic
Barium
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Nickel
Selenium
Vanadium
Zinc
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
8
8
3
8
8
8
3
8
8
8
8
3
8
8
2
3
1
4
8
2
5
8
4
1
2
8
100
25
100
13
50
100
67
63
100
50
13
67
100
380
2.0
29
0.82
2.5
100
220
7.6
22
5.9
3.8
1.7
890
420

28

2.2
71
150
1.7
13
2.6

1.9
270
50

7.4


10


7.3



54
50

7.4


10


7.3



54
190

7.4


14


8.9



130
540
1.5
51

4.2
140
460
5.8
41
8.6

2.6
2000
910
2.93
51
2.0
4.9
440
460
43
51
10
1.7
2.6
3500
910
2.9
51
2.0
4.9
440
460
43
51
10
1.7
2.6
3500
87
0.018
1000
NA
NA
1300
300
NA
100
610
170
NA
7400
Cationic Metals
Calcium
Magnesium
Potassium
Sodium
ug/L
ug/L
ug/L
ug/L
8
8
3
3
8
7
3
3
100
88
100
100
35000
9700
5500
81000
36000
11000
6400
47000
1900

3400
36000
1900

3400
36000
26000
6500
340
36000
37000
13000
6600
1 60000
82000
18000
6600
1 60000
82000
18000
6600
160000
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3,  and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) EPA suspects that this measured concentration may be elevated due to positive interference, see Section 3.1.3.
                                                                            249

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

1000



'""P
1 100
o
"co
±±
S 10
0
O

1

I I I I I I I I I I I I I
— —
- O —
0 r-*-!
- <> ° -
I O
3d
o
- o n Si
9 r~i op
n °
- * n n -
so o
r T T y i i
() 0 o"p | obdoooapo _
1
op op 4- cioo ^f. _
\_\ ocko
-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
1000
5
O)
^
^ 100
.0
"ro
-i— •
§ 10
o
O
1

. I I I I I I I I I I I I • _
1 n ll =
- - 1 I -
_ LJ o _
C _o_ 0
~ rn U ~
: 1 T IE
: H ~ @ R :
oo
Tl °
e 1 _
mooo OQQOO
op T i ""
® ° 1
_ OQQQQO ^ ° 
-------
                    Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
mn nn
I UU.UU
® 10.00
o
a
S 1 nn
N | .UU
03
~03
^™
| 0.10
Q_
On*!
. U I

-
=

~
-
1
i i i i
^<\\°,^,,^

° i i i i i i i i
o
o
o
o
a®
o
°] 0
I °r
n ?«—* L
@> i
A o
Ci) A
1 1 1 1 1 1 1 1 1
,e\«o^ ^6 ^-v®x,^,^.,,\^^°

-
=

^
-
-
-










             J$$^«$*$i$$W
                           Dissolved Metals
Figure 3.5.7. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals
in Samples of Graywater
(Note: Replacement values for non-detects are circled).
                                     252

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


100.00
"c
CD
"•5 10.00
^2

o
•a
i_
03 1 r\r\
N I -UU
03


03
'•*-• n i n
n^
>Jx
"o
Q_


0.01



—
-
—
~

-
c

I
I
~
-
-

i
_
_
-
_

—
-


<

C
4
3

r


c














t

)
^
c

K


)













^^
OQDCX
• 	
CS5





3














|

|
D















?
-©-


I
I

I I





















I I













c

c

3



3
3














)



C



C
0















(










I







1


"®"
n^
I
•^*







|

1












>

C






1

1










DO
_\
O

o
~8(





c
1

1














3


QC

op (
5O§©£J
^~" f p-^/
T^

1

















_QJ

S)
I
T

1
















:T7^
29>



\












G
C




C
c
c
c
(
c













)
)




)
)
)
)
)
)



—3
—
—
~

-
_

z
I
™
-
-

-
_
-
-
_

—
-
























                                    Total  Metals
Figure 3.5.8. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
Samples of Graywater
(Note: Replacement values for non-detects are circled).
                                           253

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.5.5    Nutrients

       Gray water samples were analyzed for four nutrient and nutrient-related parameters:
ammonia nitrogen, nitrate/nitrite, TKN, and total phosphorus (see Table 3.5.6). The nutrient
concentrations measured in graywater samples are displayed in Figure 3.5.9. The highest nutrient
concentrations measured in graywater were: 4.5 mg/L (ammonia nitrogen), 2.4 mg/L
(nitrate/nitrite), 45 mg/L (TKN), and 3.4 mg/L (total phosphorus); all of these values were
measured in a single sample of shower graywater from a tugboat. A likely source of the
phosphorus in graywater could be phosphate detergents, although both phosphorus and nitrogen
parameters also reflect food and possibly other wastes. Of these maximum nutrient
concentrations, only TKN was high enough to fall within the range of concentrations typical of
untreated domestic wastewater (20 to 85 mg/L; Metcalf and Eddy, 1979). Although each of these
nutrients was occasionally detected in service water, only nitrate/nitrite was present in service
water at concentrations high enough to be comparable with those in graywater.

       Figure 3.5.10 presents the PHQs calculated for the nutrients. As shown in this figure,
total phosphorus PHQs ranged from 4.2 to 34 because concentrations in graywater consistently
exceeded the  screening benchmark. Graywater samples from three tugboats also had PHQs of
greater than 1 because the concentrations for ammonia nitrogen exceeded the screening
benchmark.
                                           254

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.5.6. Results of Graywater Sample Analyses for Nutrients1
Analyte
Ammonia As Nitrogen (NH3-N)
Nitrate/Nitrite (NO3/NO2-N)
Total Kjeldahl Nitrogen (TKN)
Total Phosphorus
Units
mg/L
mg/L
mg/L
mg/L
No.
samples
8
8
8
8
No.
detected
8
7
8
8
Detected
Proportion (%)
100
88
100
100
Average
Cone.
1.3
1.6
10
1.4
Median
Cone.
0.75
1.9
6.7
1.2
Minimum
Cone.
0.19

2.2
0.42
10%
0.19

2.2
0.42
25%
0.22
0.90
3.8
0.62
75%
1.8
2.3
7.7
2.2
90%
4.5
2.4
45
3.4
Maximum
Cone.
4.5
2.4
45
3.4
Screening
BM2
1.2
NA
NA
0.10
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            255

-------
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       10.00
  B     1.00
   03
 CD
 O



<§    0.10
      0.
                         I
                                  oo
                                  oo
                        A        A
                   XX
                                                      oto
                                  Nutrients
Figure 3.5.9. Box and Dot Density Plot of Nutrient Concentrations Measured in Samples of

Graywater
                                       256

-------
                    Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     100.00
 o

 §    10.00
O
 03
 o
Q_
        1.00
        0.10
        0.01
                                                     i
                                                     Q
^3£T~
                                 Nutrients
Figure 3.5.10. Box and Dot Density Plot of Potential Hazard Quotients for Nutrients in

Samples of Graywater
                                     257

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.5.6    Summary of the Characterization of Graywater Effluent Analyses

       Table 3.5.7 summarizes the specific analytes in graywater effluent that may have the
potential to pose risk to human health or the environment. EPA's interpretation of the realized
risk that may be posed by these analytes, relative to pollutant loadings, background ambient and
source water contaminant levels and characteristics, and other relevant information useful for
this assessment, is presented in Chapter 5.

       Pathogens were found at higher concentrations in graywater effluent than in any other
type of pollutant. The highest concentrations of all three pathogen groups (fecal coliforms,
enterococci, andE. coli) were found in the effluent of one of the five tugboats sampled, but were
found at high concentrations in all five sampled tugboats. For all eight vessels sampled, the
majority of PHQs for all three pathogen groups were greater than 1  (PHQs for all fecal coliform
samples were greater than 10), and, in many cases, were between 100 and 10,000. The fecal
coliform concentrations most often exceeded the water quality benchmarks, followed by E. coli
and enterococci concentrations, in that order. Pathogens were not detected in the one water taxi.

       BOD was the pollutant with the next highest concentrations that exceeded water quality
benchmarks, with PHQs>l in all eight vessels and PHQ values exceeding 9 for five of the
vessels. The highest BOD concentrations were found from the recreational powerboat (PHQ =
40). Concentrations of COD and TOC were positively correlated to BOD concentrations and
were found at high levels in all eight vessels. Sulfide was detected in five of the eight vessels and
exceeded benchmark concentrations in all five instances (PHQs of up to 367). Sulfides were
detected in graywater from all vessels  sampled, and elevated in the five tugboat discharges, with
PHQs ranging from 5-367. TSS and oil and grease (measured as HEM) concentrations were also
slightly elevated, particularly in tugboats. The highest HEM concentration (100 mg/L) was
observed in the graywater discharge from a tugboat. SGT-HEM was detected in six of eight
vessels, but only one sample had a PHQ greater than 2.

       Total nonylphenol polyethoxylates (sum of isomers from NP3EO to NP18EO) were
notable only in one tugboat and the recreational boat. Total NPEOs was highest in the graywater
sample collected from the recreational powerboat. No short-chain alkylphenol ethoxylates
(NP1EO,NP2EO or OP1EO or OP2EO) or bisphenol A were detected in any of the graywater
samples. Likewise, no NP was detected,  so no comparisons could be made to the screening
benchmark.

       Among the nutrients sampled, total phosphorus concentrations exceeded the benchmark
of 0.10 mg/L in all vessels  sampled, with PHQs ranging from 4.2 to 34.

       Concentrations of dissolved copper and zinc regularly exceeded NRWQC benchmarks,
with a maximum PHQ of 90 for dissolved copper and 18 for dissolved zinc.  The median
concentration for dissolved aluminum  was 160 |ig/L, though no benchmark exists. Service water

                                          258

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

concentrations of dissolved aluminum, copper, and zinc were moderately influential, but only in
the graywater samples with the lowest measured concentrations. Total arsenic was detected in
the shrimping36 and recreational vessel where concentrations exceeded NRWQC benchmarks
(PHQ values were 111 and 161, respectively). Total aluminum concentrations exceeded
NRWQC benchmarks in seven of the eight vessels, with one vessel exceeding a PHQ of 10.
36 See Section 3.2.5.4and footnote 3 in Table 3.5.5.
                                           259

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.5.7. Characterization of Graywater Effluent and Summary of Analytes that May Have the Potential to Pose Risk









Vessel Type (no. vessels)









Tugboat (5)


Shrimping Vessel (1)

Water Taxi (1 )


Recreational (1)


Analytes that May Have the Potential to Pose Risk in Graywater Effluent and Vessel Sources1'2









VI
re
o

o
o

o
,_
^
fecal coliform
enterococci
E. co//
fecal coliform
enterococci
£. co//



fecal coliform
enterococi
E. co//



i/i
•o
c
o
Q.
O
o
o
'E
re
O)

O
a)
13
o
>











i/i
•o
c
o
Q.
E
o
o
o
'E
re
O)
O


IS
o
>
E
0)
CO



















^^^
0)
"o
(/)
(/)
^5
""""
3
Q)
^
Al, Cu, Zn


Cu, Zn

Cu, Zn


Cu, Zn












^^
"re

2
' '
3
Q)
^
Al


As, Al

Al


As, Al












0)
U)
s
S.
o
•D
C
re
~
O
X







X
















V
;o
"5
CO
X
















0
c
a)
^z
< 0
= U)
'« 9)
•= ™

r |
^ £
(/) Lll

















o
c
_Q.
>
<
C «
••=  10 (e.g.,
two or three of five), when a few samples result in PHQs greatly exceeding the screening benchmark (i.e., 100s to 1,000s), in the case of oil and grease and for nonylphenol, when one or more
samples exceed an existing regulatory limit by more than a factor of 2, or when concentrations of analytes are sufficiently high that they may have the potential to pose risks to local water bodies.
See text in Section 3.1.3 for a definition of PHQs and Table 3.1 for screening benchmarks used to calculate these values.
(2) EPA notes that the conclusion of potential risk is drawn from a small sample size, in some cases a single vessel, for certain discharges sampled from some vessel classes.  EPA included these
results in the tables to provide a concise  summary of the data collected in the study, but strongly cautions the reader that these conclusions, where there are only a few samples from a given vessel
class, should be considered preliminary and might not necessarily represent pollutant concentrations from these discharges from other vessels in this class.
                                                                           260

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
   3.2.6  Engine Effluent

       Vessel engines are primarily used for two purposes: propulsion and electrical generation.
Engines used for vessel propulsion can be either outboard or inboard engines. Vessels that
require significant lighting or have electrical equipment such as appliances and/or electric motors
are likely equipped with engines used for electrical generation.

       Engine cooling systems include direct cooling, indirect cooling, and keel cooling. Direct
and indirect cooling systems discharge wastewater, while keel cooling systems are zero
discharge. Some engines with direct and indirect cooling systems also use water to cool and quiet
their exhaust, referred to as engine wet exhaust. These engines inject spent cooling water from
the engine into the exhaust stream, so that the cooling water directly contacts the engine exhaust.
Possible constituents of concern in engine effluent include the following: thermal loading; metals
from the discharge contacting the exhaust system, from erosion of moving engine components
(e.g., pistons), or from trace constituents of the fuel; and oil and grease and organic compounds
as constituents of fuel or possible products of incomplete fuel combustion.

       The volume of engine cooling water discharged depends on the type of engine and power
level of operation. Vessels with outboard propulsion engines discharge between 1 and 2 gpm of
raw cooling water per engine based on observations made during the sampling program. The
cooling water discharge rate  from  inboard marine  diesel engines varies based on power levels,
but typically averages around 20 gpm when engines operate between  1,500 and 2,000 rpm
(Sherwood Pumps, 2009). Marine  diesel generator sets require 5 to  6 gpm of cooling water for
smaller units  (9.5 kW) (Cummins, 2008), and up to 20 and 25 gpm  of cooling water for larger
marine generator sets (80 kW) (Cummins, 2004). Daily discharge rates for these engines are a
function of daily operating time.
                                           261

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                      Collecting the Engine Effluent of a Water Taxi at Idle
             Collecting the Engine Effluent of a Tow and Salvage Vessel at Full Speed

       For this study, EPA collected engine cooling water discharge samples from a variety of
vessel classes with different engine types, as summarized in Table 3.6.1. Note that two of the
sampled vessels are recreational vessels and are not study vessels. In addition, both of the
sampled research vessels and four of the six sampled tow/salvage vessels (those with outboard
propulsion engines) were manufactured for pleasure and therefore are also recreational vessels
and not study vessels. EPA sampled engine effluent from these vessels because all of the
sampled engines can be installed on either recreational or nonrecreational vessels and are
representative of engines on study vessels.

                                           262

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       Samples were analyzed for classical pollutants, metals (dissolved and total), SVOCs, and
VOCs. Engine discharge samples were typically collected from the discharge port using a sample
transfer jar attached to a pole. The contents of the sample transfer jar were poured into a lined
utility bucket. If the engines were operated at multiple engine levels (e.g., idle, half power, full
power), then equal portions of sample were collected from each power level and composited for
a single laboratory analysis. Ten of the 13  sampled vessels with inboard propulsion engines and
all six sampled vessels with outboard propulsion engines were operated at multiple power levels.
Similarly, if a vessel operated more than one engine, then  equal portions of sample were
collected from each engine and composited for a single laboratory analysis. However, samples
for analysis of oil  and grease and VOCs are not appropriate to composite. For these analytes,
samples were collected and analyzed separately for each engine power level or were collected
from only one of the multiple engines.

Table 3.6.1. Sampled Engine Characteristics
Fuel Type
Cooling Type
Engine Wet
Exhaust?
Number of Vessels
Sampled
Vessel Types
Inboard Propulsion Engines
Diesel
Diesel
Diesel
Diesel
Gasoline
Direct
Indirect
Unknown
Unknown
Indirect
Yes
Yes
Yes
Unknown
Yes
3
5
3
1
1
Water Taxi (2), Fishing
Tour Boat (2), Water Taxi, Tow/Salvage, Fire
Boat
Tour Boat, Water Taxi, Recreational
Fishing
Recreational
Outboard Propulsion Engines
Gasoline
Gasoline
Direct
Unknown
Yes
Yes
5
1
Tow/Salvage (4), Research
Research
Generator Engines
Diesel
Diesel
Diesel
Unknown
Direct
Indirect
Unknown
Indirect
Yes
Yes
Unknown
Yes
1
1
2
1
Tour Boat
Fire Boat
Fishing, Tour Boat
Water Taxi
       EPA also observed a number of vessels, particularly tug boats and larger commercial
fishing vessels, that use keel-cooled propulsion and generator engines. The vessels were not
sampled as these closed-loop cooling systems do not have a discharge. Approximately two-thirds
of the 61 vessels visited had keel cooled engine systems.

       An additional source of relevant engine effluent data is EPA's sampling program for the
Uniform National Discharge Standards (UNDS) rulemaking. In 2006, EPA sampled propulsion
engine wet exhaust discharges from two small Armed Forces vessels with inboard diesel engines
with engine wet exhaust: a 36-foot landing craft personnel large (LCPL) and a 7-meter rigid
inflatable boat (RIB) (USEPA, 2008b). This sampling program was specifically designed to
                                          263

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

characterize engine wet exhaust discharges by power level. While these Armed Forces vessels
are not study vessels, the engines used on these vessels are comparable to those used on study
vessels. Samples from both vessels were analyzed for eight classical pollutants and 92 volatile
and semivolatile compounds. Samples from the LCPL were also analyzed for seven total metals.
Grab samples of the engine discharge were collected from sample taps installed into the exhaust
lines of the vessels. Three replicate engine discharge samples were collected at each of five
different engine power levels: 0 percent (idle), 25 percent, 50 percent, 75 percent, and 100
percent (full power). Three replicate background seawater samples were also collected. Sampling
was conducted in the open ocean.

       3.2.6.1    Inboard Propulsion Engines

       For this study, EPA collected cooling water discharge samples from inboard propulsion
engines on 13 vessels: four water taxis, three tour boats, two fishing vessels, one tow/salvage
vessel, one fire boat, and two recreational vessels (Table 3.6.1). These engines included both
direct and indirect cooling discharges from both gasoline- and diesel-fueled engines. For the
UNDS program, EPA sampled engine wet exhaust from inboard propulsion engines on two
personnel  craft. Results for each class of pollutant are presented and discussed in the following
subsections.
                         The Inboard Propulsion Engine of a Fire Boat
                                           264

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.1.1   Classical Pollutants

       Table 3.6.2 presents analytical results for 11 classical pollutants detected in samples of
discharges from inboard propulsion engines. All of the classical pollutants analyzed for were
detected and the detected results are shown in Figure 3.6.1. Engine cooling water discharge
differs from all other discharges in that the water used in the engines is drawn from surrounding
waters and immediately discharged to the same waters. For this reason, EPA analyzed the sample
results to determine which pollutant concentrations were contributed primarily by engine
operations and which were contributed primarily by background ambient concentrations (see
footnotes on Table 3.6.2 and Figure 3.6.1). The remainder of this subsection discusses those
pollutants found to be contributed primarily by engine operations.

       Temperature increases in engine  effluent above background were generally less than 5°C.
However, on three vessels operated at higher power levels (recreational vessel, tow/salvage
vessel, and fire boat), temperature increases were greater than 20°C. EPA's findings were similar
for the UNDS sampling program, with temperature increases ranging from less than 3°C at idle
to a maximum of 27°C at full power.

       Oil and grease (measured  as FIEM) was detected in the majority of engine effluent
samples; however, detected concentrations were low (most were less than 5 mg/L). All sample
results were well below the 33 CFR § 151.10 and MARPOL prohibition of the discharge of oil
and oily mixtures with an oil content greater than 15 ppm into seawater from vessels. FIEM
values exceeded 5 mg/L in only three grab samples, and all three were collected  during engine
operation at relatively high power levels. For the UNDS sampling program, HEM was not
detected in any engine effluent samples,  regardless of power level (< 4 mg/L).

       Sulfide was detected in only two of 11 samples at concentrations of 0.013 and 0.016
mg/L. These measured concentrations are six to eight times greater than the most conservative
PHQ screening benchmark of 0.002 mg/L. Sulfide might be present as a trace constituent in the
fuel, as a product of incomplete combustion, or due to formation within the biofilm in the
cooling system piping. For the UNDS sampling program, sulfide was not detected in any engine
wet exhaust samples.

       For this study, TSS concentrations in effluent discharge samples were contributed
primarily by background ambient concentrations (i.e., sample concentrations ranged from <5 to
17 mg/L while ambient water concentrations ranged from 7.8 to 20 mg/L). For the UNDS
sampling program, TSS was not detected in any of the samples from the LCPL; however, TSS
was present in the RIB discharge  samples at concentrations ranging from 6 to 14 mg/L, which
were statistically greater than background for some power levels. UNDS TSS results correspond
with the field observations for samples from the RIB at the highest power levels  (i.e., the
samples were observed to be cloudy and contained settleable materials (resembling soot)).  In this
                                          265

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

study, EPA observed that some effluent engine samples were also cloudy and contained
settleable materials at higher power levels.

       TRC was detected in only one engine effluent sample collected from a fishing vessel at a
concentration of 0.17 mg/L. Fish hold effluent from this vessel, containing TRC at a
concentration of 0.27 mg/L, was discharged into the water surrounding the vessel just prior to
collection of engine effluent samples; the propulsion engine on this vessel utilizes the ambient
water for cooling. EPA believes that the TRC value for the engine effluent sample was likely
influenced by the fish hold  effluent discharge.
                                           266

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.2. Results of Inboard Propulsion Engine Sample Analyses for Classical Pollutants1
Analyte
Conductivity2
Dissolved Oxygen3
Hexane Extractable
Material (HEM)
pH2
Salinity2
Silica Gel Treated HEM
(SGT-HEM)
Sulfide
Temperature
Total Residual Chlorine2
Total Suspended Solids
(TSS)3
Turbidity3
Units
mS/cm
mg/L
mg/L
SU
ppt
mg/L
mg/L
C
mg/L
mg/L
NTU
No.
Samples
10
10
12
13
10
12
11
13
13
11
13
No.
Detected
6
6
8
13
10
7
2
13
1
8
13
Detected
Proportion (%)
100
100
66
100
100
58
18
100
7.7
73
100
Average
Cone.
11
6.8
3.0
6.9
6.9
4.0
0.0062
22
0.048
11
32
Median
Cone.
6.1
7.4
2.2
6.6
3.3
2.6

21

13
29
Minimum
Cone.
0.22
1.7

6.2
0.10


6.5


1.2
10%
0.22
2.0

6.2
0.10


9.9


2.7
25%
0.22
4.0

6.4
0.10


17


18
75%
17
9.3
3.8
7.4
9.9
3.6

26

16
45
90%
44
13
5.4
7.9
28
4.3
0.013
36
0.10
17
69
Maximum
Cone.
44
14
5.7
8.0
28
4.4
0.013
39
0.17
17
80
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in ambient water.
(3) Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background concentrations in ambient water.
                                                                           267

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment



1 10.00
0
E
° 1.00
o
"co
i_
c
8 0.10
o
O
0.01

:
o
-
—
—
-
o
-{
o
~c



O

)-,
O
r


QD

—
-
	
:
<


o











?
ob
oo
o
IOC
c








)








I I

o.
o o




o
-E
O
~c




1 1 1 1 T 1 i
rn -
o oloOQQO
\&e\ "1°^
>-i
o
r




00


I I I
o^o °W° 4
oo —
-^ ^c oooocsSsioot
olo
o
-
-
o
o _
oooonoooo I
o
o __
loooftooon) 1 1 1 1
                                Classical Pollutant
Figure 3.6.1. Box and Dot Density Plot of Classical Pollutant Values Measured in Samples
of Inboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
                                            268

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.1.2  Metals

       Inboard propulsion engine effluent samples were analyzed for 22 dissolved and total
metals. Table 3.6.3 presents analytical results for the 16 metals that were detected in one or more
engine effluent samples. The detected results are also shown in Figures 3.6.2 and 3.6.3 for
dissolved and total metals, respectively. Figures 3.6.4 and 3.6.5 display the distribution of PHQs
based on the screening benchmark for each of the dissolved and total metals.  EPA analyzed the
sample results to determine which metals were contributed primarily by engine operations and
which were contributed primarily by background ambient concentrations. The remainder of this
subsection discusses those metals found to be contributed primarily by engine operations.

       For most metals, concentrations for the dissolved  and total forms were similar, indicating
that engine operations contribute metals in dissolved rather that particulate form. Two exceptions
were iron and lead. A comparison of dissolved and total iron concentrations indicates that almost
all iron was present in particulate form. One possible source of particulate iron in engine effluent
is rust. Lead was detected in engine effluent samples from only four of the 13 vessels sampled
(three water taxis and a tow/salvage vessel). Total lead concentrations  (maximum measured
concentration = 9.6 ug/L) exceeded dissolved lead concentrations by three to four times.

       Dissolved and total copper were detected in almost all engine effluent samples at
concentrations ranging from 3 to 53 ug/L and 5 to 66 ug/L, respectively. Dissolved copper
concentrations exceeded the PHQ screening benchmark of 3.1 ug/L (saltwater chronic criterion)
by one to 17 times  (see Figure 3.6.4). In contrast, none of the total copper concentrations
exceeded the PHQ  screening benchmark of 1,300 ug/L (human heath for consumption of water
and aquatic organisms (see Figure 3.6.5)).

       Dissolved and total zinc were also detected in a majority of engine effluent samples.
Detected concentrations ranged from  12 to 120  ug/L and  11 to 95 ug/L for dissolved and total
zinc, respectively (see Figures 3.6.2 and 3.6.3).  However, only the two highest detected
dissolved zinc concentrations (83 and 120 ug/L) exceeded the PHQ screening benchmark of 81
ug/L (saltwater chronic criterion). None of the detected total zinc concentrations exceeded the
PHQ screening benchmark of 7,400 ug/L (human heath for consumption of water and aquatic
organisms).

       Dissolved and total nickel were detected in approximately half of the  engine effluent
samples, and dissolved and total chromium and lead were each detected in fewer than half of the
engine effluent samples. Detected concentrations were generally  within five times the reporting
limit and none exceeded the screening benchmarks for these analytes (see Figures 3.6.4 and
3.6.5). Note, however, that lead is a persistent bioaccumulative and toxic chemical (PBT) and the
long-term mass loading is more important than the discharge concentrations.
                                           269

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


       Dissolved manganese was detected in 11 of 13 engine effluent samples. Manganese was
predominantly in particulate form in background ambient water; therefore, EPA assumed
dissolved manganese concentrations in engine effluent samples to be contributed by engine
operations. NRWQCs or other PHQ screening benchmarks have not been determined for
dissolved manganese.

       Dissolved iron and dissolved and total vanadium were each detected in no more than
three engine effluent samples at measured concentrations close to the reporting limit. NRWQCs
or other PHQ screening benchmarks have not been determined for these analytes at this time.

       Finally, the concentrations in engine effluent discharges that exceeded the PHQ screening
benchmark concentrations for dissolved selenium, total aluminum, and total arsenic were caused
by high background concentrations  in ambient water (which exceeded benchmark
concentrations) and not by engine operations. Moreover, in the case of dissolved and total
arsenic and selenium, measured concentrations above their respective reporting limits (three
different water taxis and the recreational vessel) may be substantially elevated due to positive
interference  (see Section 3.1.3). After subtracting the contribution of ambient water (also
potentially elevated due to positive interference), none of the detected concentrations exceeded
their PHQ screening benchmarks.

       Comparing study sampling results with the metals data from the engine wet exhaust
sampling conducted for the UNDS program affirms EPA's sampling results. For the UNDS
program, EPA determined that five  of the seven total metals analyzed for were present at
concentrations statistically greater than background: cadmium, chromium, copper, lead, and
nickel. Total mercury was not detected in any samples, and total arsenic concentrations did not
exceed background concentrations.  Table 3.6.4 compares the metals results from this study and
the UNDS program.

       EPA notes that there were some important  differences between the UNDS sampling and
the sampling conducted in this study to consider when comparing the results. The UNDS
program used a different analytical method, as well as a different methodology to calculate mean
concentrations. Also, background metals concentrations in harbors for this study are greater than
those in the open ocean for the UNDS program.
                                          270

-------
                                                           Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.3. Results of Inboard Propulsion Engine Sample Analyses for Metals1
Analyte
Units
No.
Samples
No.
Detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Heavy and Other Metals
Aluminum, Dissolved2
Aluminum, Total2
Arsenic, Dissolved3'4
Arsenic, Total3'4
Barium, Dissolved2
Barium, Total2
Chromium, Dissolved
Chromium, Total
Copper, Dissolved
Copper, Total
Iron, Dissolved
Iron, Total3
Lead, Dissolved
Lead, Total
Manganese, Dissolved
Manganese, Total2
Nickel, Dissolved
Nickel, Total2
Selenium, Dissolved2'4
Selenium, Total3'4
Vanadium, Dissolved
Vanadium, Total
Zinc, Dissolved
Zinc, Total
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
13
13
13
13
7
7
13
13
13
13
7
7
13
13
13
13
13
13
13
13
7
7
13
13
12
13
4
6
7
7
3
3
12
11
1
6
3
4
11
11
7
7
4
4
3
2
9
11
92
100
31
46
100
100
23
23
92
85
14
86
23
31
85
85
54
54
31
31
43
29
69
85
200
340
4.2
4.5
35
36
1.2
1.3
16
18
64
250
1.5
3.0
43
55
4.4
4.6
11
11
0.90
1.4
38
38
100
300


32
34


6.6
9.3

250


44
53
2.5
3.1




23
29

59


23
24


















3.8
61


23
24


1.6















23
120


29
28


5.5
5.6

150


30
40







11
180
410
8.7
8.7
34
35
0.75
0.95
23
25

310
0.60
4.1
55
74
4.3
4.3
21
21
1.4
1.1
74
75
880
920
12
13
63
63
1.9
2.4
51
62
150
520
2.1
8.5
82
95
4.9
5.5
32
31
1.7
1.6
110
89
940
940
14
15
63
63
2.1
2.6
53
66
150
520
2.3
9.6
91
100
5.3
5.6
34
32
1.7
1.6
120
95
Cationic Metals
Calcium, Dissolved2
ug/L
13
13
100
80000
37000
24000
24000
26000
62000
310000
31 0000
                                                             271

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 Table 3.6.3. Results of Inboard Propulsion Engine Sample Analyses for Metals1
Analyte
Calcium, Total2
Magnesium, Dissolved2
Magnesium, Total2
Potassium, Dissolved2
Potassium, Total2
Sodium, Dissolved3
Sodium, Total3
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
Samples
13
13
13
7
7
7
7
No.
Detected
13
13
13
7
7
7
7
Detected
Proportion
(%)
100
100
100
100
100
100
100
Average
Cone.
81000
200000
200000
32000
32000
770000
860000
Median
Cone.
37000
12000
12000
39000
39000
860000
860000
Minimum
Cone.
26000
5200
5800
4000
3700
36000
35000
10%
26000
5200
5900
4000
3700
36000
35000
25%
29000
5900
6500
4100
3800
40000
39000
75%
62000
1 60000
1 60000
58000
58000
1600000
1600000
90%
310000
1 000000
1 000000
63000
65000
1 600000
2000000
Maximum
Cone.
31 0000
1 1 00000
1 1 00000
63400
65000
1600000
2000000
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Sample concentrations were almost completely accounted for (a 90 percent) by background concentrations in ambient water.
(3) Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background concentrations in ambient water.
(4) Values well above their respective reporting limits are suspected of being  elevated due to positive interference. (See discussion in Section 3.1.3).
                                                                            272

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.4. Comparison of Metals Results for EPA P.L. 110-299 and UNDS Engine Wet
Exhaust Sampling
Metal
Arsenic, Total
Cadmium, Total
Chromium, Total
Copper, Total
Lead, Total
Nickel, Total
Mean Inboard Propulsion Engine Effluent Concentration (ug/L)
EPA P.L. 110-299 Sampling
4.5
Not Detected (Reporting Limit = 1)
1.3
18
3.0
4.6
UNDS Engine Wet Exhaust
Sampling
2.2
0.024
0.33
24
0.2
6.8
                                           273

-------
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
        1000
    05
    ^

    C
   .o
   "-I—•
    03
    CD
    O
    C
    o
   O
         100
           10
          i    i    i    r
               i   i    i    r
                                      ©
                                         )oiooi
               I
I
                                         )OpO
                                                  I	I
             *$4^*$*&^
bs*  V*
                             Dissolved Metals
Figure 3.6.2. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
Samples of Inboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Values well above their respective reporting limits for dissolved arsenic and selenium are suspected of being
elevated due to positive interference).
                                      274

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
1000F
 O)

 c
.o
"-I—•
 CD
    CD
    O

    O
   O
           100
             10
                          I     I    I    I
                                                      I    I     I    I
                              *
                                                                       r
                                     Total  Metals
Figure 3.6.3. Box and Dot Density Plot of Total Metals Concentrations Measured in
Samples of Inboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Values well above their respective reporting limits for total arsenic and selenium are suspected of being
substantially elevated due to positive interference).
                                             275

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

10.00


CD
•~
0
| 1.00
03
N
03
|
^^^^
O3
c 0.10
-22
o
Q_

n ni
I
—
-
_
-
~
_

~
-

I









/•-,
"" v.
o

-

—
- .--
^-~
-
<
I

-€
j
o

3-


n*"1
<_)

K
WJ

K
|
I I



c

o






*

I I
^
c


id

a
c
c
c






)





\
•)


X

Q
)
)
)












I



)

O

<£5cx












1 1
—
-
_
-
~
_
^V' 	 — ^




5> (o*§)






I I
















 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Replacement values for non-detects are circled. Also, values well above their respective reporting limits for
dissolved arsenic and selenium are suspected of being elevated due to positive interference).
                                              276

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
~ 100.00
CD
§ 10.00
a
•a
03 1 DD
N llUU
03
« 0.10
-1— •
_CD
£ 0.01
0.00
= I II I I I
E oo
0
ri
"3
j-£ f
CD
E o
Op
nnop
po
; oopo
r i i i i i i
I I I I I I E
-
-=
_
OOT5O E
offio
ooToo_
<5oooeSiO£^^> z
o
I I I I I <&> 1
              ^
^
                                       Total  Metals
Figure 3.6.5. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
Samples of Inboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Replacement values for non-detects are circled. Also, values well above their respective reporting limits for
total arsenic and selenium are suspected of being elevated due to positive interference).
                                               Til

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.1.3   Semivolatile Organic Compounds

       Inboard propulsion engine effluent samples were analyzed for 76 SVOCs for the
sampling conducted as part of the P.L. 110-299 study. Table 3.6.5 presents analytical results for
the 31 SVOCs that were detected in one or more engine effluent samples. The detected results
are also shown in Figures 3.6.6 and 3.6.7 for analyte concentrations and for PHQs based on the
lowest NRWQC  or other PHQ screening benchmark where applicable, respectively. EPA
analyzed the sample results to determine which SVOCs were contributed primarily by engine
operations and which were contributed primarily by background ambient concentrations. All
were found to be contributed primarily by engine operations.

       Many of the detected SVOCs can be classified among the following pollutant classes:
polycyclic aromatic hydrocarbons or PAHs (14 analytes), straight-chain hydrocarbons (five
analytes), phenol and methyl phenols (five analytes), and phthalates (two analytes). These
include all of the SVOCs detected most frequently and at the highest concentrations.

       PAHs are present in fuel in small amounts and may be formed as products of incomplete
combustion. EPA has identified seven PAHs as probable human carcinogens, six of which were
detected in engine effluent collected from a recreational vessel with a gasoline engine
dewinterized immediately prior to sampling (see details below): benzo(a)anthracene,
benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, chrysene, and indeno(l,2,3-
cd)pyrene. Most  of these compounds exceed a PHQ of 1,000 as shown in Figure 3.6.7.

       Phthalates are plasticizers (chemicals added to plastics to make them flexible) and are
commonly detected in environmental samples (ATSDR, 2002). Bis(2-ethylhexyl) phthalate was
detected at concentration just above the screening benchmark of 1.2 ug/L (human heath for
consumption of water and aquatic organisms).

       Phenol and methyl phenols are present in petroleum products and may also be generated
as products of incomplete combustion. Discharges of phenol and methyl phenols are assumed to
not cause any environmental impacts as detected concentrations did not exceed the PHQ
screening benchmarks for these analytes. Straight-chain (alkane) hydrocarbons are also
components of fuel;  none of the straight-chain hydrocarbons detected in engine effluent have a
NRWQC or other PHQ screening benchmark, and they are not PBT chemicals

       It is important to note that 11 of the detected SVOCs were found only in one sample
collected from a recreational vessel (recreational vessels are not study vessels). These included
all six of the detected PAHs that  are probable human carcinogens, as well as four additional
PAHs. Engine  effluent from this  recreational vessel  also contributed the maximum detected
concentrations for six additional  analytes, including several additional PAHs as well as four of
the five detected  phenol/methyl phenols. (Maximum sample concentrations for 2,4-
dimethylphenol,  straight-chain hydrocarbons, and phthalates were contributed by other vessels.)

                                          278

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

This recreational vessel was the only sampled vessel that used gasoline as fuel rather than diesel;
however, the lack of replication precludes any determination as to whether fuel type is a critical
factor for engine effluent characteristics. In addition, the engines on this vessel were
dewinterized immediately prior to sampling. The lack of engine operation for several months
prior to sampling could have contributed to engine effluent characteristics.

       Comparing study sampling results with the results from the engine wet exhaust sampling
conducted for the UNDS program reveals some similarities. For the LCPL, phenol and bis(2-
ethylhexyl)phthalate were the only detected SVOCs; however, the presence of bis(2-ethyhexyl)
phthalate in LCPL effluent may be due to laboratory contamination and so data for the purpose
of comparison are not shown in this  report. For the RIB, phenol was the only detected SVOC.
EPA determined that phenol was present at concentrations statistically greater than background.
Table 3.6.6 compares the phenol results from this study to those from the UNDS program. Note
that the UNDS program used a different methodology to calculate mean concentrations.
                                           279

-------
                                                           Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.5. Results of Inboard Propulsion Engine Sample Analyses for SVOCsl
Analyte
1 ,2-Diethyl-Cyclobutane
1 ,6-Dimethyl
1 -Methylnaphthalene
2,4-Dimethylphenol
2-Methylnaphthalene
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Bis(2-ethylhexyl)
phthalate
Chrysene
Di-n-butyl phthalate
Eicosane
Fluorene
Heptadecane
lndeno(1,2,3-cd)
m-Cresol
Naphthalene
n-Hexadecane
Nonadecane
Nonanoic Acid
o-Cresol
Units
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
|jg/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
|jg/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
|jg/L
|jg/L
No.
Samples
1
1
2
12
8
12
12
12
12
12
12
12
12
12
12
12
2
12
4
12
4
12
3
2
1
3
No.
Detected
1
1
2
4
6
1
3
1
1
1
1
1
1
4
1
6
2
4
4
1
1
10
3
2
1
3
Detected
Proportion
(%)
100
100
100
33
75
8.3
25
8.3
8.3
8.3
8.3
8.3
8.3
33
8.3
50
100
33
100
8.3
25
8.3
100
100
100
100
Average
Cone.
10
35
13
3.7
17
2.0
7.0
3.3
3.3
3.2
2.8
2.6
3.1
1.7
3.3
1.7
19
3.5
29
2.5
13
30
26
27
11
6.6
Median
Cone.


24

13










1.1
28

27


6.6
17
38

5.8
Minimum
Cone.


3.2













10

3.5



3.1
15

5.7
10%


3.2













10

3.5



3.1
15

5.7
25%


3.2

0.90











10

3.8


1.9
3.1
15

5.7
75%


24
2.4
36

1.7






1.2

1.6
28
2.8
67

34
34
57
38

8.4
90%


24
16
46
1.5
44
12
13
11
7.8
6.9
11
1.8
12
3.5
28
14
80
5.6
45
160
57
38

8.4
Maximum
Cone.


24
22
46
2.2
61
18
18
16
11
9.8
15
20
18
3.8
28
18
80
8.0
45
210
57
38

8.4
                                                              280

-------
                                                                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.5. Results of Inboard Propulsion Engine Sample Analyses for SVOCsl
Analyte
Octadecane
p-Cresol
Phenanthrene
Phenol
Pyrene
Units
ug/L
ug/L
ug/L
ug/L
UQ/L
No.
Samples
2
7
12
12
12
No.
Detected
2
5
3
8
1
Detected
Proportion
(%)
100
71
25
67
8.3
Average
Cone.
10
26
6.1
27
6.6
Median
Cone.
17
17

3.7

Minimum
Cone.
3.1




10%
3.1




25%
3.1




75%
17
24
1.3
37

90%
17
110
35
140
40
Maximum
Cone.
17
110
48
170
57
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Sample concentrations were almost completely accounted for (a 90 percent) by background concentrations in ambient water.
(3) Sample concentrations were predominantly accounted for (> 50 percent and  <90 percent) by background concentrations in ambient water.
Table 3.6.6. Comparison of Phenol Results for EPA P.L. 110-299 and UNDS Engine Wet Exhaust Sampling
Analyte
Phenol
Mean Inboard Propulsion Engine Effluent Concentration (ug/L)
EPA P.L. 110-299 Sampling
27
UNDS Small Boat Engine Wet Exhaust Sampling
LCPL
13
RIB
14
                                                                         281

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
              100
          O)
          ^
          C
          o
         "co
          CD
          o
          c
          o
         O
                10
                  1
                      i    i   i   i   i    i   i   i   i    i   i   i    i   i   r
                                                   T
                                                 i
                           i   i   i   i    i  T   I  i    i   i   i    i   i
                   02468 101214161820222426283032
                                         SVOCs
Figure 3.6.6. Box and Dot
Study Samples of Inboard
SVOCs are identified as follows:
(1) 1,2-Diethyl-Cyclobutane
(2) 1,6-dimethylnaphthalene
(3) 1-methylnaphthalene
(4) 2,4-Dimethylphenol
(5) 2-Methylnaphthalene
(6) Acenaphthene
(7) Acenaphthylene
(8) Anthracene
(9) Benzo(a)anthracene
(10) Benzo(a)pyrene
(11) Benzo(b)fluoranthene
Density Plot of SVOC Concentrations Measured in P.L. 110-299
 Propulsion Engine Effluent
       (12) Benzo(g,h,i)perylene
       (13) Benzo(k)fluoranthene
       (14) Bis(2-ethylhexyl) phthalate
       (15) Chrysene
       (16) Di-n-butyl phthalate
       (17) Eicosane
       (18) Fluorene
       (19) Heptadecane
       (20) Indeno(l,2,3-cd)pyrene
       (21)m-Cresol
       (22) Naphthalene
(23) n-Hexadecane
(24) Nonadecane
(25) Nonanoic Acid
(26) o-Cresol
(27) Octadecane
(28) p-Cresol
(29) Phenanthrene
(30) Phenol
(31)Pyrene
                                             282

-------
                           Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 0
"-I— •
 O
 ^
O
 CD
 N
 CD
 CD
 o
Q_
1000.0000

  100.0000

    10.0000

     1.0000

     0.1000

     0.0100

     0.0010
          0.0001
                                     * i   i   i
                                                         i   r
                                                                        ©=
                                  I	I
                                          I	I
I	I
1
fib
                    02468  101214161820222426283032
                                          SVOCs
     Figure 3.6.7. Box and Dot Density Plot of Potential Hazard Quotients for SVOCs in P.L.
     110-299 Study Samples of Inboard Propulsion Engine Effluent
     SVOCs are identified as follows (replacement values for non-detects are circled):
     (1) 1,2-Diethyl-Cyclobutane
     (2) 1,6-dimethylnaphthalene
     (3) 1-methylnaphthalene
     (4) 2,4-Dimethylphenol
     (5) 2-Methymaphthalene
     (6) Acenaphthene
     (7) Acenaphthylene
     (8) Anthracene
     (9) Benzo(a)anthracene
     (10) Benzo(a)pyrene
     (11) Benzo(b)fluoranthene
                              (12) Benzo(g,h,i)perylene
                              (13) Benzo(k)fluoranthene
                              (14) Bis(2-ethylhexyl) phthalate
                              (15) Chrysene
                              (16) Di-n-butyl phthalate
                              (17) Eicosane
                              (18) Fluorene
                              (19) Heptadecane
                              (20) Indeno(l,2,3-cd)pyrene
                              (21)m-Cresol
                              (22) Naphthalene
         (23) n-Hexadecane
         (24) Nonadecane
         (25) Nonanoic Acid
         (26) o-Cresol
         (27) Octadecane
         (28) p-Cresol
         (29) Phenanthrene
         (30) Phenol
         (31)Pyrene
                                              283

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.1.4   Volatile Organic Compounds

       Inboard propulsion engine effluent samples were analyzed for 84 VOCs. Table 3.6.7
presents analytical results for the 38 VOCs that were detected in one or more engine effluent
samples. The detected results are also shown in Figures 3.6.8 and 3.6.9 for analyte
concentrations and for PHQs based on the lowest NRWQC or other PHQ screening benchmark
where applicable, respectively. EPA analyzed the sample results to  determine which VOCs were
contributed primarily by engine operations and which were contributed primarily by background
ambient concentrations. All were found to be contributed primarily by engine operations.

       Approximately one-third of the detected VOCs were frequently detected in engine
effluent (i.e., greater than half of the sampled vessels). Some of these compounds are volatile
constituents of fuel, specifically benzene, toluene, ethylbenzene, and xylene. Others are
trimethylbenzenes, which are naturally present in fuel, and ketones, which may be formed as
products of incomplete combustion. Among these compounds, only benzene and toluene have an
NRWQC. Approximately half of the detected benzene concentrations (from water taxis, tour
boats and a recreational vessels) exceeded the PHQ screening benchmark of 2.2 ug/L (human
heath for consumption of water and  aquatic organisms), including discharges from one vessel
that exceeded the benchmark by a factor of more than 50  (the next highest concentration that
exceeded the benchmark was by less than a factor of 4) (see Figure 3.6.9). None of the detected
toluene concentrations exceeded the PHQ screening benchmark of 1,300 ug/L (human health for
consumption of water and aquatic organisms).

       Approximately one-third of the detected VOCs were detected relatively infrequently (i.e.,
detected in fewer than half the sampled vessels). Among these compounds, only chloroform and
methylene chloride have an NRWQC. However, none of the detected concentrations for these
two analytes exceeded the PHQ screening benchmarks of 5.7 ug/L (human  heath for
consumption of water and aquatic organisms) and 1,300 ug/L (human heath for consumption of
water and aquatic organisms), respectively.

       The final  third of detected VOCs were detected in engine effluent from only one or two
vessels. None of these analytes have an NRWQC or are PBT chemicals, and are therefore not
expected to have the potential to pose risk to human health or the environment.

       It is important to note the maximum detected concentrations for 11 of the VOCs were
found in samples collected from a recreational vessel (recreational vessels are not study vessels).
These included benzene, toluene, ethylbenzene, xylene, and trimethylbenzenes (maximum
sample concentrations for ketones were contributed by other vessels). As noted above, this
recreational  vessel was the only sampled vessel that used  gasoline as fuel rather than diesel;
however, this data set is too small to demonstrate whether fuel type is a critical factor for engine
effluent characteristics. In addition, the engines on this vessel were dewinterized  immediately
                                          284

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

prior to sampling. The lack of engine operation for several months prior to sampling could have
contributed to engine effluent characteristics.

       Comparing these sampling results with the results from the engine wet exhaust sampling
conducted for the UNDS program reveals some similarities. For the LCPL, no VOCs were
detected. For the RIB, 1,2,3-trimethylbenzene, and 1,3,5-trimethylbenzene were the detected
VOCs. However, EPA determined that the trimethylbenzenes were not present at concentrations
statistically greater than background.
                                           285

-------
                                                            Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.7. Results of Inboard Propulsion Engine Sample Analyses for VOCsl
Analyte
(2-Methyl-1-Propenyl)-
Benzene
1 ,2,3,4-Tetrahydro-5-
Methylnaphthalene
1 ,2,3,4-Tetrahydro-6-
Methylnaphthalene
1 ,2,3,4-Tetrahydro
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
1 ,3-Methylnaphthalene
1 ,7-Methylnaphthalene
2,3-Dihydro-4-Methyl-1 H-
Indene
2,6-Dimethyl
2-Butanone
2-Ethyl-1 ,3,5-Trimethyl-
Benzene
2-Ethyl-1 ,4-Dimethyl-
Benzene
2-Hexanone
4-lsopropyltoluene
4-Methyl-2-Pentanone
Acetone
Benzene
Benzocycloheptatriene
Biphenyl
Chloroform
Dimethocxymethane
Ethylbenzene
Isopropylbenzene
Units
|jg/L
|jg/L
|jg/L
M9/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
M9/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
|jg/L
No.
Samples
1
1
2
2
7
7
1
1
1
1
7
1
1
7
7
7
8
12
1
8
12
1
12
7
No.
Detected
1
1
2
2
7
5
1
1
1
1
7
1
1
5
3
3
8
9
1
6
4
1
6
3
Detected
Proportion
(%)
1.00
1.00
1.00
1.00
1.00
0.71
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.71
0.43
0.43
1.00
0.75
1.00
0.75
0.33
1.00
0.50
0.43
Average
Cone. 2
3.2
24
19
12
6.1
2.1
4.2
19
53
41
17
4.4
20
2.1
1.8
1.9
58
12
39
4.1
1.7
89
2.3
1.9
Median
Cone.


33
22
1.8
0.70




7.8


1.1


34
2.3

3.0


0.10

Minimum
Cone.


4.6
3.2
0.12





2.6





6.0







10%


4.6
3.2
0.12





2.6





6.0







25%


4.6
3.2
0.30





3.0





15
0.17

0.27




75%


33
22
3.8
0.92




32


2.9
1.3
0.80
110
5.4

4.5
1.0

0.83
1.4
90%


33
22
32
7.2




40


3.2
1.4
1.6
150
84

12
2.1

12
1.5
Maximum
Cone. 2


33
22
32
7.2




40


3.2
1.4
1.6
150
120

12
2.1

16
1.5
                                                              286

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.7. Results of Inboard Propulsion Engine Sample Analyses for VOCsl
Analyte
m-,p-Xylene (sum of
isomers)
Methyl acetate
Methyl tertiary butyl ether
(MTBE)
Methylene chloride
n-Butylbenzene
n-Pentadecane
n-Propylbenzene
n-Tetradecane
O-Xylene
sec-Butylbenzene
Styrene
Toluene
Trichlorofluoromethane
Vinyl acetate
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
Samples
7
7
7
12
7
2
7
2
7
7
7
12
12
7
No.
Detected
7
1
1
4
3
2
4
2
7
1
7
8
1
1
Detected
Proportion
(%)
1.00
0.14
0.14
0.33
0.43
100
57
100
100
14
100
67
8.3
14
Average
Cone. 2
11
2.4
2.4
1.2
1.8
24
1.5
20
5.5
2.3
6.1
11
2.1
2.4
Median
Cone.
1.8




31
0.15
33
1.5

1.3
0.90


Minimum
Cone.
0.30




16

6.5
0.20

0.13



10%
0.30




16

6.5
0.20

0.13



25%
0.90




16

6.5
0.65

0.50



75%
2.0


0.14
1.0
31
0.40
33
1.8

3.4
2.8


90%
70
1.5
1.9
0.19
1.1
31
2.2
33
32
1.4
35
80
1.9
1.9
Maximum
Cone. 2
70
1.5
1.9
0.20
1.1
31
2.2
33
32
1.4
35
110
2.7
1.9
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) In some cases, the detected concentration(s) for an analyte could be lower than the replacement value (% of the reporting limit) for a concentration that was nondetected. In an
extreme (but possible) case, this could result in an average concentration for an analyte that is greater than the maximum detected concentration.
                                                                             287

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
             100.0
 O)

 C.
 O
"co
         CD
         o
         c
         o
        o
               10.0
        1.0
                 0.1
                i  i   i   i  i   i   i  i.  i  i   i   i  i   i   i  i   i  i   i
                                                            ©
                                              ©
                                                                     U-
                                                                       ©
                     0246  810121416182022242628303234363840
                                            VOCs
Figure 3.6.8. Box and Dot
Measured in P.L. 110-299
VOCs are identified as follows:
(1) (2-Methyl-l-Propenyl)-
Benzene
(2) l,2,3,4-Tetrahydro-5-
Methylnaphthalene
(3) l,2,3,4-Tetrahydro-6-
Methylnaphthalene
(4) 1,2,3,4-
Tetrahydronaphthalene
(5) 1,2,4-Trimethylbenzene
(6) 1,3,5-Trimethylbenzene
(7) 1,3-Methylnaphthalene
(8) 1,7-Methylnaphthalene
(9) 2,3-Dihydro-4-Methyl-lH-
Indene
(10) 2,6-dimethylnaphthalene
(ll)2-Butanone
                   Density Plot of Volatile Organic Compounds Concentrations
                   Study Samples of Inboard Propulsion Engine Effluent
                          (12) 2-Ethyl-l,3,5-Trimethyl-
                          Benzene
                          (13) 2-Ethyl-l,4-Dimethyl-
                          Benzene
                          (14) 2-Hexanone,
                          (15) 4-Isopropyltoluene
                          (16) 4-Methyl-2-Pentanone
                          (17) Acetone
                          (18) Benzene
                          (19) Benzocycloheptatriene
                          (20) Biphenyl
                          (21) Chloroform
                          (22) Dimethoxymethane
                          (23) Ethylbenzene
                          (24) Isopropylbenzene
(25) m-,p-Xylene (sum of
isomers)
(26) Methyl acetate
(27) Methyl tertiary butyl ether
(MTBE)
(28) Methylene chloride
(29) n-Butylbenzene,
(30) n-Pentadecane
(31) n-Propylbenzene
(32) n-Tetradecane
(33) O-Xylene
(34) sec-Butylbenzene
(35) Styrene
(36) Toluene
(37) Trichlorofluoromethane
(38) Vinyl acetate
                                              288

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
10.0000
"c
CD
g 1.0000
0
•D
J3 0.1000
cc
X
H 0.0100
-2
£ 0.0010
n nnni
: I I I I I I I I T I I I
T
do
! C3feii
I C°^o>
r
: *°
r
I I I I I I I :
-
000 ;
Cg£302> -
0 ~=
000
—
gTnnnnfg> oo
Oy0 j '
i i i i i i i i i i i i
I I I I I I I
                       0246  810121416182022242628303234363840
                                             VOCs
Figure 3.6.9. Box and Dot Density Plot of Potential Hazard Quotients for Volatile Organic
Compounds in P.L. 110-299 Study Samples of Inboard Propulsion Engine Effluent
VOCs are identified as follows (replacement values for non-detects are circled):
(1) (2-Methyl-l-Propenyl)-
Benzene
(2) l,2,3,4-Tetrahydro-5-
Methylnaphthalene
(3) l,2,3,4-Tetrahydro-6-
Methylnaphthalene
(4) 1,2,3,4-
Tetrahydronaphthalene
(5) 1,2,4-Trimethylbenzene
(6) 1,3,5-Trimethylbenzene
(7) 1,3-Methymaphthalene
 (8) 1,7-Methylnaphthalene
(9) 2,3-Dihydro-4-Methyl-lH-
Indene
(10) 2,6-dimethylnaphthalene
(ll)2-Butanone
(12) 2-Ethyl-l,3,5-Trimethyl-
Benzene
(13) 2-Ethyl-l,4-Dimethyl-
Benzene
(14) 2-Hexanone
(15) 4-Isopropyltoluene
(16) 4-Methyl-2-Pentanone
(17) Acetone
(18) Benzene
(19) Benzocycloheptatriene
(20) Biphenyl
(21) Chloroform
(22) Dimethoxymethane
(23) Ethylbenzene
(24) Isopropylbenzene
(25) m-,p-Xylene (sum of
isomers)
(26) Methyl acetate
(27) Methyl tertiary butyl ether
(MTBE)
(28) Methylene chloride
(29) n-Butylbenzene
(30) n-Pentadecane
(31) n-Propylbenzene
(32) n-Tetradecane
(33) O-Xylene
(34) sec-Butylbenzene
(35) Styrene
(36) Toluene
(37) Trichlorofluoromethane
(38) Vinyl acetate
                                               289

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       3.2.6.2    Outboard Propulsion Engines

       For this study, EPA collected samples of discharges from outboard propulsion engines on
six vessels: four tow/salvage vessels and two research vessels (see Table 3.6.1 above). It is
important to note that all six of these vessels were confirmed by the vessel owners/operators to
be manufactured for pleasure. Vessels manufactured for pleasure are defined as recreational
vessels under P.L. 110-288 and are not study vessels. Nonetheless, EPA has included the results
here assuming they are representative of vessels with outboard propulsion engines, some  of
which may be study vessels. EPA also collected these results so that the Agency could later
compare results between study vessels and recreational vessels if appropriate.
                       The Outboard Engine of a Tow and Salvage Vessel

       3.2.6.2.1  Classical Pollutants

       Outboard propulsion engine effluent samples were analyzed for 11 classical pollutants.
Table 3.6.8 presents analytical results for the eight classical pollutants that were detected in one

                                           290

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


or more engine effluent samples. The detected results are also shown in Figure 3.6.10. EPA
analyzed the sample results to determine which pollutants concentrations were contributed
primarily by engine operations and which were contributed primarily by background ambient
concentrations (see footnotes on table and figure). The remainder of this subsection discusses
those pollutants found to be contributed primarily by engine operations.

       Temperature increases in engine effluent above background were less than 5°C for all
vessels. Engine effluent temperatures were only slightly higher (approximately 1°C) when
vessels were operated at higher power levels as compared to idling.

       Oil and grease (measured as HEM) was not detected in any of the engine effluent
samples. SGT-HEM was detected in only two of 16 grab samples at concentrations significantly
less than the reporting limit (sample concentrations of 0.86 mg/L and  0.94 mg/L, compared to
reporting limit of 10 mg/L).
                                           291

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.8. Results of Outboard Propulsion Engine Sample Analyses for Classical Pollutants1
Analyte
Conductivity3
Dissolved Oxygen2
PH2
Salinity3
Silica Gel Treated HEM
(SGT-HEM)
Temperature
Total Suspended Solids
(TSS)3
Turbidity2
Units
mS/cm
mg/L
SU
ppt
mg/L
C
mg/L
NTU
No.
Samples
5
5
6
5
6
6
6
6
No.
Detected
5
5
6
5
2
6
2
6
Detected
Proportion
(%)
100
100
100
100
33
100
33
100
Average
Cone.
167
6.2
7.4
11
4.5
28
8.1
13
Median
Cone.
17
6.3
7.3
12

31

10
Minimum
Cone.
7.3
5.7
7.0
3.9

14

6.5
10%
7.3
5.7
7.0
3.9

14

6.5
25%
9.2
5.9
7.1
7.3

25

8.0
75%
22
6.4
7.7
14
3.6
31
13
21
90%
25
6.4
7.9
16
3.6
32
17
25
Maximum
Cone.
25
6.4
7.9
16
3.6
32
17
25
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage  of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in ambient water.
(3) Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background concentrations in ambient water.
                                                                           292

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
         0°
            ^


-1— •
0
E
•<
»
o
o
"co
I":
c
CD
O
0
O


35
30
25
20

15


10





5


i i i i i i i i
- -
— Q Q —
~ o i -
o 9
- * -
o
aa 1
o
^ ^
Q
00
0 I obi©b
6
O^D
_ OOOO GOOD _
O
1 1 1 1 1 1 1 1
                                Classical  Pollutant
Figure 3.6.10. Box and Dot Density Plot of Classical Pollutant Values Measured in Samples
of Outboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
                                           293

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.2.2  Metals

       Outboard propulsion engine effluent samples were analyzed for dissolved and total
concentrations of 22 metals. Table 3.6.9 presents analytical results for the 14 metals that were
detected in one or more engine effluent samples. The detected results are also shown in Figures
3.6.11 and 3.6.12 for dissolved and total metals, respectively. Figures 3.6.13 and 3.6.14 display
the distribution of PHQs based on the screening benchmark for each of the dissolved and total
metals. EPA analyzed the sample results to determine which metals were contributed primarily
by engine operations and which were contributed primarily by background ambient
concentrations (see footnotes on table and figures). The remainder of this subsection discusses
those metals found to be contributed primarily by engine operations.

       Dissolved and total concentrations for both vanadium and zinc are similar, which
indicates that engine operations contribute these metals in dissolved rather that particulate form.
Dissolved zinc was detected in all engine effluent samples at concentrations two to five times the
reporting limit; none of the concentrations exceed the PHQ screening benchmark (a value of 81
ug/L based on the chronic saltwater criterion for aquatic life). Dissolved vanadium was detected
in engine effluent from four of the six sampled vessels at concentrations close to the reporting
limit (<2 times reporting limit of 1 ug/L). Dissolved vanadium does not have an NRWQC or
other PHQ screening benchmark.

       Total arsenic was detected in engine effluent from five of the six sampled vessels at
concentrations two to five times the reporting limit (reporting limit = 8  ug/L), however, EPA
suspects the measured concentrations of total (and dissolved) arsenic are elevated due to positive
interference. Likewise, dissolved selenium was detected in all engine effluent samples at
concentrations ranging from 2.4 to 100  ug/L; however, EPA suspects the measured
concentrations of dissolved (and total) selenium are elevated due to positive interference.

       Finally, concentrations in engine effluent discharges for dissolved copper, total
aluminum, total iron, and total manganese that exceed benchmark concentrations appear to be
caused by background concentrations in ambient water and not by engine operations.
                                           294

-------
                                                           Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.9. Results of Outboard Propulsion Engine Sample Analyses for Metals1
Analyte
Units
No.
Samples
No.
Detected
Detected
Proportion
(%)
Average
Cone.
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Heavy and Other Metals
Aluminum, Dissolved2
Aluminum, Total2
Arsenic, Dissolved3'4
Arsenic, Total3'4
Barium, Dissolved2
Barium, Total2
Copper, Dissolved3
Copper, Total3
Iron, Total3
Manganese, Dissolved2
Manganese, Total3
Nickel, Dissolved3
Nickel, Total3
Selenium, Dissolved3'4
Selenium, Total3'4
Vanadium, Dissolved
Vanadium, Total
Zinc, Dissolved3
Zinc, Total
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
5
6
5
5
6
6
6
5
2
6
6
6
6
6
6
2
3
6
6
83
100
83
83
100
100
100
83
33
100
100
100
100
100
100
33
50
100
100
7.4
160
25
24
25
27
3.3
3.6
200
6.0
57
5.6
11
76
72
0.87
1.7
11
11
8.2
58
32
30
15
16
3.4
3.4

5.4
35
6.6
7.7
97
94

1.2
11
8.3

34


13
14
2.8


1.0
29
3.2
3.3
2.4
1.5


3.5
3.5

34


13
14
2.8


1.0
29
3.2
3.3
2.4
1.5


3.5
3.5
5.1
38
8.6
9.9
14
14
3.1
2.4

1.2
29
3.6
5.6
24
22


7.1
6.4
9.7
320
37
34
41
43
3.5
3.8
460
10
91
7.1
14
110
100
1.5
1.4
14
14
10
570
41
41
57
65
3.5
3.9
560
18
140
7.4
33
130
120
1.8
1.5
19
28
10
570
41
41
57
65
3.5
3.9
560
18
140
7.4
33
130
120
1.8
1.5
19
28
Cationic Metals
Calcium, Dissolved3
Calcium, Total3
Magnesium, Dissolved3
ug/L
ug/L
ug/L
6
6
6
6
6
6
100
100
100
130000
130000
380000
1 60000
1 60000
480000
43000
43000
31000
43000
43000
31000
50000
51000
1 20000
1 70000
1 70000
520000
200000
1 90000
630000
200000
190000
630000
                                                             295

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.9. Results of Outboard Propulsion Engine Sample Analyses for Metals1
Analyte
Magnesium, Total3
Potassium, Dissolved3
Potassium, Total3
Sodium, Dissolved3
Sodium.Total3
Units
ug/L
ug/L
ug/L
ug/L
ug/L
No.
Samples
6
6
6
6
6
No.
Detected
6
6
6
6
6
Detected
Proportion
(%)
100
100
100
100
100
Average
Cone.
370000
130000
130000
2900000
2900000
Median
Cone.
480000
1 60000
1 60000
3800000
3700000
Minimum
Cone.
31000
11000
11000
220000
220000
10%
31000
11000
11000
220000
220000
25%
1 20000
48000
48000
1000000
1 1 00000
75%
520000
1 90000
1 80000
4100000
4000000
90%
600000
220000
21 0000
4700000
4700000
Maximum
Cone.
600000
220000
210000
4700000
4700000
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Sample concentrations were almost completely accounted for (a 90 percent) by background concentrations in ambient water.
(3) Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background concentrations in ambient water.
(4) Measured concentrations well above their respective reporting limits,  are suspected of being  elevated due to positive interference (See section 3.1.3).
                                                                            296

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
      O)
      O
     "co
=     10
 CD
 O

 O
O
                       I      I
                             fo
 Q
oo
on
                       ^
                       I	I
                                         I      I
                                              Q
                                              o
                                              0
                                         I	I
                                          I      I
                                                        Cprnri
                                        obdo
                                          I      I
                                  Dissolved  Metals
Figure 3.6.11. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in
Samples of Outboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Measured concentrations well above their respective reporting limits for dissolved arsenic and selenium are
suspected of being elevated due to positive interference).
                                             297

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
           100
      o
      "co
 CD
 O
 c
 o
O
             10
                     QD
                                 Q
                       i     i
                                                            I      I     I
                                          ooo
                                                      ?
                                                     nrbn
                                                 --
                                                       n
                                    OBflO
                                       i     i
                                            *     *
                                                                       0
                                                                 Q
i    T    i
                                      Total  Metals
Figure 3.6.12. Box and Dot Density Plot of Total Metals Concentrations Measured in
Samples of Outboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Measured concentrations well above their respective reporting limits for total arsenic and selenium are
suspected of being elevated due to positive interference).
                                             298

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


^ 10.00
CD
"o
a
j= 1 nn
03 I .UU

03

03
"c
CD
•+± n i n
o U. IU
Q_
1 1 1 1 1 1 o 1 1
nnn -
n ~

"OQTu C^u" -
_ uu nn ~

010
n
1 -
UU
oo
1 1 1 1 1 1 II
                               ^x
'1T
                                  Dissolved Metals
Figure 3.6.13. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals
in Samples of Outboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Replacement values for non-detects are circled. Also, measured concentrations well above their respective
reporting limits for dissolved arsenic and selenium are suspected of being elevated due to positive interference).
                                              299

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
1000.000
1 100.000
-1— •
o
0 10.000
•a
M 1 nnn
CD 1-uuu
|j 0.100
CD
f 0.010
0.001

- I Q|U I I I I I I I I -
- $ -
_ i^r, —
=- -=
- rL -
P°i Q
= so r®n oooo =
Oodo
= 1111111111 =
***¥*¥* ** *x\r
                                       Total  Metals
Figure 3.6.14. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
Samples of Outboard Propulsion Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Replacement values for non-detects are circled. Also, measured concentrations well above their respective
reporting limits for total arsenic and selenium are suspected of being elevated due to positive interference).
                                               300

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.2.3   Semivolatile Organic Compounds

       Outboard propulsion engine effluent samples were analyzed for 62 SVOCs. Table 3.6.10
presents analytical results for the seven SVOCs that were detected in one or more engine effluent
samples. The detected results are also shown in Figure 3.6.15. EPA analyzed the sample results
to determine which SVOCs were contributed primarily by engine operations and which were
contributed primarily by background ambient concentrations. All were found to be contributed
primarily by engine operations.

       The detected SVOCs can be classified among the following pollutant classes: polycyclic
aromatic hydrocarbons (PAHs) (one analyte), phenol and methyl phenols (four analytes),
phthalates (one analyte), and methylnaphthalenes (one analyte). All of these SVOCs were
frequently detected in engine effluent (i.e., more than half of the sampled vessels). However, all
of the detected SVOC concentrations are well below any applicable PHQ screening benchmarks.
For example, the maximum PHQ for any of the detected SVOCs was 2,4-dimethylphenol with a
PHQ of approximately 0.005. Therefore, SVOCs in engine effluent are highly unlikely to have
the potential to pose risk to human health or the environment.
                                          301

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.10. Results of Outboard Propulsion Engine Sample Analyses for SVOCs1
Analyte
2,4-Dimethylphenol
2-Methylnaphthalene
Di-n-butyl phthalate
m-Cresol
Naphthalene
p-Cresol
Phenol
Units
UQ/L
UQ/L
M9/L
M9/L
M9/L
M9/L
M9/L
No.
Samples
6
6
6
6
6
6
6
No.
Detected
1
2
3
2
5
2
2
Detected
Proportion
(%)
17
33
50
33
83
33
33
Average
Cone.
2.5
2.4
2.4
2.6
7.8
3.7
4.6
Median
Cone.


1.2

2.0


Minimum
Cone.







10%







25%




1.4


75%
0.49
1.5
2.3
1.9
12
3.9
5.9
90%
2.0
2.8
3.5
4.2
35
9.8
14
Maximum
Cone.
2.0
2.8
3.5
4.2
35
9.8
14
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage  of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
                                                                           302

-------
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
              40

              30


              20
              10
          o

         "CD
          CD
          o
          c
          o
         O
                                    *


                                    I       I
                                      SVOCs
Figure 3.6.15. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of

Outboard Propulsion Engine Effluent

(Note: two analyte names were truncated: 2-Methylnaphalene and Di-n-butyl phthalate).
                                         303

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.2.4   Volatile Organic Compounds

       Outboard propulsion engine effluent samples were analyzed for 70 VOCs. Table 3.6.11
presents analytical results for the 18 VOCs that were detected in one or more engine effluent
samples. The detected results are also shown in Figures 3.6.16 and 3.6.17 for analyte
concentrations and for PHQs based on the lowest NRWQC or other PHQ screening benchmark
where applicable, respectively. EPA analyzed the sample results to determine which VOCs were
contributed primarily by engine operations and which were contributed primarily by background
ambient concentrations. All were found to be contributed primarily by engine operations.  Some
of these compounds are volatile constituents of fuel, specifically benzene, toluene, ethylbenzene,
and xylene. Others are trimethylbenzenes, which are naturally present in fuel, and one is a
ketone, which may be formed as a product of incomplete combustion. Among these compounds,
benzene, ethylbenzene,  and toluene have an NRWQC. Most of the detected benzene
concentrations exceeded the PHQ screening benchmark of 2.2 ug/L (human heath for
consumption of water and aquatic organisms), including discharges from the two research
vessels that exceed the benchmark by factors of five and 28. None of the detected ethylbenzene
and toluene concentrations exceeded the PHQ screening benchmarks.

       The final  one-third of the detected VOCs were detected relatively infrequently (i.e.,
detected in fewer than half the sampled vessels). Among these compounds, only methylene
chloride has an NRWQC. However, none of the detected methylene chlorine concentrations
exceeded the screening benchmark.
                                          304

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.11. Results of Outboard Propulsion Engine Sample Analyses for VOCs1
Analyte
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
2-Butanone
2-Hexanone
4-Methyl-2-Pentanone
Acetone
Benzene
Cyclohexane
Ethylbenzene
Isopropylbenzene
m-,p-Xylene (sum of
isomers)
Methyl tertiary butyl ether
(MTBE)
Methylcyclohexane
Methylene chloride
n-Propylbenzene
O-Xylene
Styrene
Toluene
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
Samples
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
No.
Detected
6
5
2
1
1
5
6
1
6
2
6
1
1
5
4
6
5
6
Detected
Proportion
(%)
100
83
33
17
17
83
100
17
100
33
100
17
17
83
67
100
83
100
Average
Cone. 2
13
4.6
3.8
2.5
2.3
7.8
13
2.4
8.2
2.4
28
2.3
2.3
0.58
3.2
15
4.9
52
Median
Cone.
2.3
1.9



2.5
2.4

2.1

3.4


0.20
1.7
4.0
3.4
3.8
Minimum
Cone.
0.30





0.13

0.90

0.33




0.17

0.40
10%
0.30





0.13

0.90

0.33




0.17

0.40
25%
0.53
0.75



1.4
0.76

0.92

0.43


0.15

0.43
0.22
0.75
75%
24
6.5
3.8
0.56
0.35
11
24
0.41
14
1.3
52
0.34
0.36
0.20
3.6
26
6.6
98
90%
63
18
12
2.3
1.4
34
62
1.7
38
3.8
140
1.4
1.5
0.20
9.4
70
16
260
Maximum
Cone. 2
63
18
12
2.3
1.4
34
62
1.7
38
3.8
140
1.4
1.5
0.20
9.4
70
16
260
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) In some cases, the detected concentration(s) for an analyte could be lower than the replacement value (% of the reporting limit) for a concentration that was nondetected. In an
extreme (but possible) case, this could result in an average concentration for an analyte that is greater than the maximum detected concentration.
                                                                            305

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
        100.0
   05
   ^
   C
  .o
  "-I—•
   03
   CD
   O
   C
   o
  O
         10.0
           1.0
           0.1
                     i     r
                                            \      \
                        ©
                           ©
                        ©
                     j	L
              ©
                                                 ©  ©
j	L
J	I
                0    2    4    6    8    10  12   14  16  18  20
                                        VOCs
Figure 3.6.16. Box and Dot Density Plot of Volatile Organic Compounds Concentrations
Measured in Samples of Outboard Propulsion Engine Effluent
VOCs are identified as follows:
                                (8) Cyclohexane
                                (9) Ethylbenzene
                                (10) Isopropylbenzene
                                (11) m-,p-Xylene (sum of
                                isomers)
                                (12) Methyl tertiary butyl ether
                                (MTBE)
(1) 1,2,4-Trimethylbenzene
(2) 1,3,5-Trimethylbenzene
(3) 2-Butanone
(4) 2-Hexanone
(5) 4-Methyl-2-Pentanone
(6) Acetone
(7) Benzene
                    (13) Methylcyclohexane
                    (14) Methylene chloride
                    (15) n-Propylbenzene
                    (16) O-Xylene
                    (17) Styrene
                    (18) Toluene
                                           306

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 0
"§
a
•a
 N
 03

~03
Q_
      10.000
        1.000
       0.100
       0.010
       0.001
                     i     r
                                           \     \
\\
                                  (Hi
                                                   oonoo
                                       SL
                                       dxd
                                                                CH:
                     I	I
                                           I	I
I
               0    2    4    6    8   10  12   14  16   18  20
                                       VOCs
Figure 3.6.17. Box and Dot Density Plot of Potential Hazard Quotients for Volatile Organic
Compounds in Samples of Outboard Propulsion Engine Effluent
VOCs are identified as follows (replacement values for non-detects are circled):
(1) 1,2,4-Trimethylbenzene           (8) Cyclohexane                   (13) Methylcyclohexane
(2) 1,3,5-Trimethylbenzene           (9) Ethylbenzene                  (14) Methylene chloride
(3) 2-Butanone                    (10) Isopropylbenzene              (15) n-Propylbenzene
(4) 2-Hexanone                    (11) m-,p-Xylene (sum of            (16) O-Xylene
(5) 4-Methyl-2-Pentanone            isomers)                         (17) Styrene
(6) Acetone                      (12) Methyl tertiary butyl ether        (18) Toluene
(7) Benzene                      (MTBE)
                                          307

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       3.2.6.3    Generator Engines

       For this study, EPA collected cooling water discharge samples from engines on generator
sets onboard five vessels: a fishing vessel, a fire boat, two tour boats, and a water taxi (Table
3.6.1). These engines included both direct and indirect cooling discharges from both gasoline-
and diesel-fueled engines.
                                   The Generator on a Fire Boat

       3.2.6.3.1  Classical Pollutants

       Table 3.6.12 presents analytical results for 11 classical pollutants detected in samples of
discharges from generator engines (all classical pollutants analyzed for were detected). The
detected results are also shown in Figure 3.6.18. EPA analyzed the sample results to determine
which pollutant concentrations were contributed primarily by generator engine operations and
which were contributed primarily by background ambient concentrations (see footnotes on table
and figure). The remainder of this subsection discusses those classical pollutants found to be
contributed primarily by generator engine operations.

       Temperature increases in generator engine effluent above background were
approximately 5°C for the fishing vessel, fire boat, and water taxi. For the two tour boats,
temperature increases were 9 and 13°C.
                                           308

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


       Oil and grease (measured as HEM) was detected in engine effluent from three of the five
sampled generators; however, detected concentrations were low, ranging from less than the
reporting limit to just above the reporting limit (reporting limit = 5 mg/L).  All sample results
were well below the 33 CFRPart 151.10 prohibition of the discharge of oil and oily mixtures
with an oil content greater than 15 ppm into seawater from vessels.

       Sulfide was detected in only one of five samples at a concentration of 0.012 mg/L, which
is slightly above the reporting limit of 0.01 mg/L. This concentration is six times greater than the
most conservative PHQ screening benchmark - a 2006 NRWQC value of 0.002  mg/L for the
protection of aquatic life. Sulfide could be present due to entrainment in fuel, as a product of
incomplete combustion, or due to formation within the biofilm in the cooling system piping.

       TRC was detected in only one generator engine effluent sample collected from a water
taxi at a concentration of 0.15 mg/L. This detected concentration is 20 times greater than the
PHQ screening benchmark of 0.0075 mg/L. There is no known source of TRC for this vessel  as
background concentration of the ambient water at this location was below  detection and the
generator did not use service water that might contain TRC.
                                          309

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.12. Results of Generator Engine Sample Analyses for Classical Pollutants1
Analyte
Conductivity2
Dissolved Oxygen2
Hexane Extractable
Material (HEM)
PH2
Salinity3
Silica Gel Treated HEM
(SGT-HEM)
Sulfide
Temperature
Total Residual Chlorine
Total Suspended Solids
(TSS)3
Turbidity2
Units
mS/cm
mg/L
mg/L
SU
ppt
mg/L
mg/L
C
mg/L
mg/L
NTU
No.
Samples
4
4
5
5
4
5
4
5
5
4
5
No.
Detected
4
4
3
5
4
1
1
5
1
3
5
Detected
Proportion
(%)
100
100
60
100
100
20
25
100
20
75
100
Average
Cone. 4
11
5.3
2.9
6.5
6.5
4.2
0.0068
21
0.060
9.0
27
Median
Cone.
0.31
6.2
1.1
6.6
0.20


20

12
33
Minimum
Cone.
0.23
1.9

5.7
0.10


18


1.3
10%
0.23
1.9

5. 7
0.10


18


1.3
25%
0.23
2.6

5.9
0.10


19

2.1
14
75%
32
7.7
4.3
7.0
19
0.55
0.0090
24
0.075
13
38
90%
43
8.2
5.8
7.0
25
1.1
0.012
26
0.15
13
39
Maximum
Cone. 4
43
8.2
5.8
7.0
25
1.1
0.012
26
0.15
13
39
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in ambient water.
(3) Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background concentrations in ambient water.
(4) In some cases, the detected concentration(s) for an analyte could be lower than the replacement value (% of the reporting limit) for a concentration that was nondetected. In an
extreme (but possible) case, this could result in an average concentration for an analyte that is greater than the maximum detected concentration.
                                                                             310

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

- 10.00
o
E
o 1.00
c
0
"ro
-i— •
§ 0.10
0
O
0.01
:
^~
—
-

)


o
HE



—
~
—
I
"O"
I




|
I I
c
° OQQO
OOP
0 0


I I I I pb I -
"o" ~=
i oc"°;

o
^
o
Q

-
I I I fffl, I I I I I :
         0°
            ^
                               Classical Pollutants
Figure 3.6.18. Box and Dot Density Plot of Classical Pollutant Values Measured in Samples
of Generator Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
                                            Ill

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.3.2  Metals

       Generator engine effluent samples were analyzed for dissolved and total concentrations
of 22 metals. Table 3.6.13 presents analytical results for the 11 metals that were detected. The
detected results are also shown in Figures 3.6.19 and 3.6.20 for dissolved and total metals,
respectively. Figures 3.6.21 and 3.6.22 display the distribution of PHQs based on the screening
benchmark for each of the dissolved and total metals. EPA analyzed the sample results to
determine which metals were contributed primarily by generator engine operations and which
were contributed primarily by background ambient concentrations (see footnotes on table and
figures). The remainder of this subsection discusses those metals found to be contributed
primarily by generator engine operations.

       Dissolved and total metals concentrations are similar, which indicates that engine
operations contribute metals in dissolved rather that particulate form. Dissolved copper was
detected in all five generator effluent samples at concentrations ranging from 2.4 to 13 ug/L.
Total copper was detected in two of the five samples at concentrations of 2.4 and 11 ug/L
(reporting limit = 5 ug/L). Dissolved copper concentrations exceeded the PHQ screening
benchmark of 3.1 ug/L (2006 NRWQC saltwater chronic aquatic life criterion) by as much as
five times. In contrast, none of the total copper concentrations exceeded the PHQ screening
benchmark of 1,300 ug/L (human health criterion based on consumption of water and aquatic
organisms).

       Dissolved manganese was detected in four of the five generator engine effluent samples.
Manganese was predominantly in particulate form in background ambient water; therefore,
dissolved manganese concentrations in engine effluent samples are assumed to be contributed by
engine operations. NRWQCs or other PHQ screening benchmarks have not been determined for
dissolved manganese.

       Dissolved zinc was detected in two of the five generator engine effluent samples.
Detected concentrations were 21 to 29 ug/L, which are substantially lower than the screening
benchmark of 81 ug/L (2006 NRWQC saltwater chronic aquatic life criterion).

       Finally, concentrations in generator engine effluent discharges that exceed benchmark
concentrations for total aluminum are likely  caused or heavily influenced by higher
concentrations in ambient water (which exceeded benchmark concentrations).
                                           112

-------
                                                                          Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.13. Results of Generator Engine Sample Analyses for Metals1
Analyte
Units
No.
Samples
No.
Detected
Detected
Proportion
(%)
Average
Cone. 4
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone. 4
Heavy and Other Metals
Aluminum, Dissolved2
Aluminum, Total2
Barium, Dissolved2
Barium, Total2
Copper, Dissolved
Copper, Total
Iron, Total2
Manganese, Dissolved
Manganese, Total3
Nickel, Dissolved3
Nickel, Total3
Zinc, Dissolved
Zinc, Total3
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
5
5
1
1
5
5
1
5
5
5
5
5
5
5
5
1
1
5
2
1
4
4
1
1
2
3
100
100
100
100
100
40
100
80
80
20
20
40
60
280
420
37
37
6.5
4.2
200
33
40
4.5
3.5
13
11
160
390


5.6


36
43



12
11
120


2.4








11
120


2.4








86
220


3.9


16
17




540
640


9.5
6.7

49
59
1.4
1.4
25
15
870
890


13
11

53
63
2.7
2.7
29
19
870
890


13
11

53
63
2.7
2.7
29
19
Cationic Metals
Calcium, Dissolved2
Calcium, Total2
Magnesium, Dissolved2
Magnesium, Total2
Potassium, Dissolved2
Potassium, Total2
Sodium, Dissolved2
Sodium, Total2
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
5
5
5
5
1
1
1
1
5
5
5
5
1
1
1
1
100
100
100
100
100
100
100
100
80000
82000
180000
180000
4000
3600
37000
36000
26000
28000
5900
6600




23000
27000
5200
5900




23000
27000
5200
5900




24000
27000
5200
5950




1 60000
1 60000
440000
450000




290000
290000
870000
890000




290000
290000
870000
890000




Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10%  of the observations were found.
(2) Sample concentrations were almost completely accounted for (a 90 percent)  by background concentrations in ambient water.
(3) Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background concentrations in ambient water.
(4) In some cases, the detected concentration(s) for an analyte could be lower than the  replacement value (% of the reporting limit) for a concentration that was nondetected. In an
extreme (but possible) case, this could result in an average concentration for an analyte that is greater than the maximum detected concentration.
                                                                             313

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
g
 c
 o

"co

"c
 CD
 o
 c
 o
O
o
0
-^
0
                                                oh
                                                       0000   000
                                         n
                                         9
               **
                                 ^

                            Dissolved Metals
Figure 3.6.19. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in

Samples of Generator Engine Effluent

* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in

ambient water.

** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background

concentrations in ambient water.
                                             114

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Concentration (ug/L)
-^ o
0 0
: T I I I I I I :
o
OTO

Q ^
Q
00
- O -
00 00
„ ~ _ _ f\ S~\ f\

\ \ T~ 1 1 1 1
* * <***«*«*
_x\ _x\ c»N o. * \* ~*
                                        ^N
                                     Total  Metals
Figure 3.6.20. Box and Dot Density Plot of Total Metals Concentrations Measured in
Samples of Generator Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
                                             115

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
u
4
3
-i— •
.1 2
0
o 1
3! Hazard
vw
"c
CD
O
Q_

1 1 1 1 1 1
— —
cea


o 9

dlOCL)
1 1 1 1 1 ^~T"

                                                  ^e     _\*     ...oP
                                  Dissolved Metals
Figure 3.6.21. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals
in Samples of Generator Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Replacement values for non-detects are circled).
                                              116

-------
 o
 ^
a
•a
 i_
 03
 N
 03
 o
Q_
      10.000
1.000
0.100
       0.010
       0.001
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                       I      I
                    QJDTD
                                                 i      r
                                                               0
                                                              on
                       I      I
                                                      Cjjr!
                       *      *       {
                  .^ AS^  0X
                ,<^ ^   o°v
                                  Total  Metals
Figure 3.6.22. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in
Samples of Generator Engine Effluent
* Sample concentrations were almost completely accounted for (> 90 percent) by background concentrations in
ambient water.
** Sample concentrations were predominantly accounted for (> 50 percent and <90 percent) by background
concentrations in ambient water.
(Note: Replacement values for non-detects are circled).
                                          117

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


       3.2.6.3.3   Semivolatile Organic Compounds

       Generator engine effluent samples were analyzed for 79 SVOCs. Table 3.6.14 presents
analytical results for the 26 SVOCs that were detected in one or more engine effluent samples
(14 of the detected SVOCs were analyzed for and detected in only one generator effluent
sample). The detected results are shown in Figures 3.6.23 and 3.6.24 for analyte concentrations
and PHQs based  on the lowest applicable NRWQC or other PHQ screening benchmark. EPA
analyzed the sample results to determine which SVOCs were contributed primarily by generator
engine operations and which were contributed primarily by background ambient concentrations.
All were found to be contributed primarily by generator engine operations.

       Many of the detected SVOCs can be classified among the following pollutant  classes:
PAHs (five analytes), straight-chain hydrocarbons (six analytes), phenol and methyl phenols
(five analytes), and phthalates (two analytes). These include all of the SVOCs analyzed for and
detected most frequently and at the highest concentrations.

       PAHs are present in fuels in small amounts and may be formed as products of incomplete
combustion. However, none of the detected PAH concentrations exceeded the screening
benchmarks for these analytes, indicating that they are unlikely to have the potential to pose risk
to human health or the environment.

       Straight-chain (alkane) hydrocarbons are also components of fuel. None of these analytes
has an NRWQC or other PHQ screening benchmark, and they are not PBT chemicals. Therefore,
the straight-chain hydrocarbons detected in engine effluent are unlikely to have the potential to
pose risk to human health or the environment.

       Phenol and methyl phenols are also present in petroleum products and may also be
generated as products of incomplete combustion. Discharges of phenol and methyl phenols are
assumed not to result in any environmental impacts as detected concentrations did not exceed the
screening benchmarks for these analytes.

       Phthalates are plasticizers (chemicals added to plastics to make them flexible) and are
commonly detected in environmental samples  (ATSDR, 2002). Bis(2-ethylhexyl) phthalate was
detected at concentration just above the screening benchmark of 1.2 ug/L (human heath for
consumption of water and aquatic organisms).

       The generator engine effluent sample from the fire boat contained the maximum
concentration of  12 of the detected SVOCs. These include all five of the detected PAHs, four of
the five detected  phenols and methyl phenols, and both of the detected phthalates.  The generator
effluent sample from a tour boat contained the maximum concentration of all six of the detected
straight-chain hydrocarbons.
                                          118

-------
                                                           Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.14. Results of Generator Engine Sample Analyses for SVOCsl
Analyte
1-methylnaphthalene
2,4-Dimethylphenol
2-Cyclopenten1-one
2-Hydroxy-Benzaldehyde
2-Methylnaphthalene
2-Naphthalene
3-Methyl-Benzaldehyde
3-Methylphenol
3-Phenyl-2-Propenal
Acenaphthylene
Acetophenone
Bis(2-ethylhexyl) phthalate
Di-n-butyl phthalate
Eicosane
Fluorene
Heneicosane
Heptadecane
m-Cresol
Naphthalene
n-Hexadecane
Nonadecane
Octadecane
p-Cresol
Units
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
M9/L
No.
Samples
3
5
2
2
4
2
1
1
1
5
1
5
5
1
5
1
3
1
5
1
1
1
1
No.
Detected
3
1
2
2
4
2
1
1
1
1
1
1
1
1
1
1
3
1
4
1
1
1
1
Detected
Proportion
(%)
100
20
100
100
100
100
100
100
100
20
100
20
20
100
20
100
100
100
80
100
100
100
100
Average
Cone. 2
6.7
2.6
8.5
11
16
18
18
12
8.1
1.8
11
1.3
1.3
32
2.0
22
30
18
17
46
40
44
43
Median
Cone.
5.4

13
17
10
20










8.9

7.3




Minimum
Cone.
3.8

3.9
4.3
4.6
16










4.1






10%
3.8

3.9
4.3
4.6
16










4.1






25%
3.8

3.9
4.3
5.5
16










4.1

2.3




75%
11
4.0
13
17
32
20



1.9

0.63
0.59

2.4

76

36




90%
11
7.9
13
17
40
20



3.8

1.3
1.2

4.9

76

61




Maximum
Cone. 2
11
7.9
13
17
40
20



3.8

1.3
1.2

4.9

76

61




                                                              319

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.14. Results of Generator Engine Sample Analyses for SVOCsl
Analyte
Phenanthrene
Phenol
Pyrene
Units
ug/L
UQ/L
ug/L
No.
Samples
5
5
5
No.
Detected
3
4
1
Detected
Proportion
(%)
60
80
20
Average
Cone. 2
3.9
23
1.4
Median
Cone.
3.2
13

Minimum
Cone.



10%



25%

2.1

75%
6.8
48
0.90
90%
9.7
75
1.8
Maximum
Cone. 2
9.7
75
1.8
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) In some cases, the detected concentration(s) for an analyte could be lower than the replacement value (% of the reporting limit) for a concentration that was nondetected. In an
extreme (but possible) case, this could result in an average concentration for an analyte that is greater than the maximum detected concentration.
                                                                             320

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
         O)
         ^
         c
         o
         "co
         CD
         o
         c
         o
         O
              80
              60
              40
20
                 i    r
i    r
i    r
   *
                      I
                 I	I
              I	I
              I
                 0   2   4   6   8   10  12 14  16  18 20 22  24 26
                                         SVOCs
Figure 3.6.23. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
Generator Engine Effluent
SVOCs are identified as follows:
(1) 1-methylnaphthalene
(2) 2,4-Dimethylphenol
(3) 2-Cyclopentenl-one
(4) 2-Hydroxy-Benzaldehyde
(5) 2-Methylnaphthalene
(6)2-
Naphthalenecarboxaldehyde
(7) 3-Methyl-Benzaldehyde
(8) 3-Methylphenol
                    (9) 3-Phenyl-2-Propenal
                    (10) Acenaphthylene
                    (11) Acetophenone
                    (12) Bis(2-ethylhexyl) phthalate
                    (13) Di-n-butyl phthalate
                    (14) Eicosane
                    (15) Fluorene
                    (16) Heneicosane
                    (17) Heptadecane
                       (18) m-Cresol
                       (19) Naphthalene
                       (20) n-Hexadecane
                       (21)Nonadecane
                       (22) Octadecane
                       (23) p-Cresol
                       (24) Phenanthrene
                       (25) Phenol
                       (26) Pyrene
                                             321

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
      1.0000
 0
"-I— •
 o
 ^
O
 CD
 N
 CD
 CD
 o
Q_
0.1000
                    o
0.0100
      0.0010
      0.0001
                    i    i    i    i     r
                                                     \     \    \
                                                O
                     I    I    I    I     I
                                                 I    I     I    I
                0   2   4   6   8  10  12  14 16  18  2022 2426
                                       SVOCs
Figure 3.6.24. Box and Dot Density Plot of Potential Hazard Quotients for SVOCs in
Samples of Generator Engine Effluent
SVOCs are identified as follows (replacement values for non-detects are circled):
(1) 1-methylnaphthalene
(2) 2,4-Dimethylphenol
(3) 2-Cyclopentenl-one
(4) 2-Hydroxy-Benzaldehyde
(5) 2-Methymaphthalene
(6) 2-Naphthalene-
carboxaldehyde
(7) 3-Methyl-Benzaldehyde
(8) 3-Methylphenol
(9) 3-Phenyl-2-Propenal
                          (10) Acenaphthylene
                          (11) Acetophenone
                          (12) Bis(2-ethylhexyl) phthalate
                          (13) Di-n-butyl phthalate
                          (14) Eicosane
                          (15) Fluorene
                          (16) Heneicosane
                          (17) Heptadecane
                          (18) m-Cresol
                          (19) Naphthalene
(20) n-Hexadecane
(21)Nonadecane
(22) Octadecane
(23) p-Cresol
(24) Phenanthrene
(25) Phenol
(26) Pyrene
                                           322

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.3.4   Volatile Organic Compounds

       Generator engine effluent samples were analyzed for 80 VOCs. Table 3.6.15 presents
analytical results for the 28 VOCs that were detected. The detected results are also shown in
Figures 3.6.25 and 3.6.26 for analyte concentrations and for PHQs based on the lowest NRWQC
or other PHQ screening benchmark, where applicable, respectively. EPA analyzed the sample
results to determine which VOCs were contributed primarily by generator engine operations and
which were contributed primarily by background ambient concentrations. All were found to be
contributed primarily by generator engine operations.

       Twenty-two of the detected VOCs were analyzed for in only one sample. None of these
compounds has an NRWQC or are PBT chemicals. Of the seven detected VOCs that were
analyzed for in more than one sample, three have an NRWQC: benzene, ethylbenzene, and
toluene. All of the detected benzene concentrations (from three of the five samples) exceeded the
PHQ screening benchmark of 2.2 ug/L by factors ranging from one to nine. The single detected
concentration for each of ethylenebenzene and toluene did not exceed their respective PHQ
screening benchmarks.

       Note that the generator effluent sample from the fire boat contained the maximum
concentration of 19 of the detected VOCs. These include benzene, toluene, ethylbenzene, xylene,
trimethylbenzenes, and ketones.
                                          323

-------
                                                           Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.15. Results of Generator Engine Sample Analyses for VOCsl
Analyte
(E)-2-Butenal
1 ,2,3,4-Tetrahydro-5-
Methylnaphthalene
1 ,2,3,4-Tetrahydro-6-
Methylnaphthalene
1 ,2,4-Trimethylbenzene
1 ,3,5-Trimethylbenzene
2,6-Dimethyl
2-Butanone
2-Butenal
2-Ethyl-1 ,4-Dimethyl-
Benzene
4-lsopropyltoluene
4-Methyl-2-Pentanone
Acetone
Benzaldehyde
Benzene
Benzofuran
Biphenyl
Ethylbenzene
Isopropylbenzene
m-,p-Xylene (sum of
isomers)
Methyl acetate
n-Pentadecane
n-Propylbenzene
n-Tetradecane
O-Xylene
sec-Butylbenzene
Units
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
No.
Samples
1
1
1
1
1
1
1
1
1
1
1
2
1
5
1
1
5
1
1
1
1
1
1
1
1
No.
Detected
1
1
1
1
1
1
1
1
1
1
1
2
1
3
1
1
1
1
1
1
1
1
1
1
1
Detected
Proportion
(%)
100
100
100
100
100
100
100
100
100
100
100
100
100
60
100
100
20
100
100
100
100
100
100
100
100
Average
Cone.
12
5.9
7.2
8.0
1.6
5.5
83
19
5.7
0.40
1.7
120
4.2
5.9
6.9
12
1.4
0.50
5.3
0.80
40
0.90
20
3.4
0.50
Median
Cone.











220

3.1











Minimum
Cone.











22













10%











22













25%











22













75%











220

12


1.0








90%











220

21


2.0








Maximum
Cone.











220

21


2.0








                                                              324

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.15. Results of Generator Engine Sample Analyses for VOCsl
Analyte
Styrene
Toluene
Vinyl acetate
Units
ug/L
ug/L
ug/L
No.
Samples
1
5
1
No.
Detected
1
1
1
Detected
Proportion
(%)
100
20
100
Average
Cone.
8.9
3.5
1.5
Median
Cone.



Minimum
Cone.



10%



25%



75%

6.2

90%

12

Maximum
Cone.

12

Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall.  So the 10th percentile is the concentration below which at least 10% of the observations were found.
                                                                            325

-------
                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
           100
      05
      ^
      C
      o
      "co
      CD
      O
      c
      o
     O
10
                         i   i    i    r
                                       i    i    i    r
                                             IQ    OQQO
                         I    I    I    I
                                        I    I    I    I
                 02468  1012141618202224262830
                                           VOCs
Figure 3.6.25. Box and Dot Density Plot of VOC Concentrations Measured in Samples of
Generator Engine Effluent
VOCs are identified as follows:
(1) (E)-2-Butenal
(2) l,2,3,4-Tetrahydro-5-
Methylnaphthalene
(3) l,2,3,4-Tetrahydro-6-
Methylnaphthalene
(4) 1,2,4-Trimethylbenzene
(5) 1,3,5-Trimethylbenzene
(6) 2,6-dimethylnaphthalene
(7) 2-Butanone
(8) 2-Butenal
                     (9) 2-Ethyl-l,4-Dimethyl-
                     Benzene
                     (10) 4-Isopropyltoluene
                     (11) 4-Methyl-2-Pentanone
                     (12) Acetone
                     (13) Benzaldehyde
                     (14) Benzene
                     (15) Benzofuran
                     (16) Biphenyl
                     (17) Ethylbenzene
                     (18) Isopropylbenzene
(19) m-,p-Xylene (sum of
isomers)
(20) Methyl acetate
(21) n-Pentadecane
(22) n-Propylbenzene
(23) n-Tetradecane
(24) o-Xylene
(25) sec-Butylbenzene
(26) Styrene
(27) Toluene
(28) Vinyl acetate
                                              326

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
        10.00E
   0
   "-I—•
   o
g     1.00
   o
   •a
   03
   N
   03

   ~03
   O
  Q_
       0.10
   3     0.01
          0.00
                       \    \
                     \    \
                       I	I
I    I    I
                 02468  10121416182022242628
                                          VOCs
Figure 3.6.26. Box and Dot Density Plot of Potential Hazard Quotients for VOCs in
Samples of Generator Engine Effluent
VOCs are identified as follows (replacement values for non-detects are circled):
(1) (E)-2-Butenal
(2) l,2,3,4-Tetrahydro-5-
Methylnaphthalene
(3) l,2,3,4-Tetrahydro-6-
Methylnaphthalene
(4) 1,2,4-Trimethylbenzene
(5) 1,3,5-Trimethylbenzene
(6) 2,6-dimethylnaphthalene
(7) 2-Butanone
(8) 2-Butenal
                               (9) 2-Ethyl-l,4-Dimethyl-
                               Benzene
                               (10) 4-Isopropyltoluene
                               (11) 4-Methyl-2-Pentanone
                               (12) Acetone
                               (13) Benzaldehyde
                               (14) Benzene
                               (15) Benzofuran
                               (16) Biphenyl
                               (17) Ethylbenzene
                               (18) Isopropylbenzene
                         (19) m-,p-Xylene (sum of
                         isomers)
                         (20) Methyl acetate
                         (21) n-Pentadecane
                         (22) n-Propylbenzene
                         (23) n-Tetradecane
                         (25) sec-Butylbenzene
                         (26) Styrene
                         (27) Toluene
                         (28) Vinyl acetate
                                             327

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.4    Comparison of Effluent Generated at Different Propulsion Engine Power
                 Levels

       Although inboard and outboard propulsion engines were often sampled during operation
at different power levels (e.g., idle, half power, full power), these samples were generally
composited for a single analysis. Exceptions include samples for analysis of HEM/SGT-HEM
and VOCs, which were collected and analyzed separately for each power level of engine
operation (composite samples for these analytes are not appropriate). EPA reviewed the
HEM/SGT-HEM and VOC data to determine whether there were any trends in the resulting data
based on engine power level  of operation.

       HEM was detected in the majority of inboard engine effluent samples; however, detected
concentrations were low (the majority were less than the reporting limit of 5 mg/L). Of the eight
vessels with inboard engines with detected HEM concentrations that were sampled at different
power levels, engine effluent samples from six had higher HEM concentrations at higher engine
levels than at idle. Data for the remaining two vessels were inconclusive. Note, however, that
differences in HEM concentrations among power levels were small, ranging from 0.1 to 5 mg/L.
For outboard engines, HEM was not detected  in any of the engine effluent samples.

       Regarding VOC results for inboard engines, EPA reviewed benzene, toluene,
ethylbenzene, and xylene concentrations as these compounds were the most frequently detected.
Of the eight vessels with inboard engines with detected benzene concentrations that were
sampled at different power levels, engine effluent samples from five contained higher benzene
concentrations at higher engine levels than at  idle. Data for the remaining three vessels showed
the opposite pattern, with higher benzene concentrations at idle than at higher engine levels. For
seven of these sampled vessels, differences in benzene concentrations among the power levels
were small, ranging from 0.1 to 4.7 ug/L. In contrast, for the remaining vessel (a recreational
vessel), the difference in benzene concentrations from idle to three-quarter speed was 89 ug/L,
with the higher concentration detected at  idle. As discussed previously, this recreational vessel
was the only sampled vessel that used gasoline as fuel rather than diesel. In addition, the engines
on this vessel were dewinterized immediately prior to sampling.

       The differential among detected concentrations of ethylbenzene, xylene, and toluene at
different power levels is too small to draw any conclusions, except for the engine effluent data
for the recreational vessel.  Differences in detected concentrations between idle and three-quarter
power were 18 ug/L for ethylbenzene, 73 ug/L for m-,p-xylene, 31 ug/L for o-xylene, and 84
ug/L for toluene. The higher concentrations were found at idle for all four analytes.

       The UNDS sampling program provides a useful comparison for this study as it was
specifically designed to evaluate engine wet exhaust characteristics among power levels,
including the separate collection and analysis  of three replicate samples at each of five different
                                          328

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

power levels. Tables 3.6.16 and 3.6.17 present sample results from the UNDS study by power
level for the LCPL and RIB, respectively.

       EPA made several conclusions for the LCPL based on a review of the engine effluent
results. Chromium, copper, lead, and nickel were all detected at concentrations significantly
greater than background concentrations for all five power levels. For copper and nickel,
concentrations were highest at idle, second highest at 100 percent power, and then generally
decreased with decreasing  power levels (decreasing engine RPM). Chromium concentrations
were highest at 100 percent power and then also decreased with decreasing power levels, with
the lowest chromium concentrations found at idle. Lead concentrations were not significantly
different at the various power levels. For TOC and phenol, only idle concentrations were
significantly greater than background  concentrations.

       For the RIB, only TOC concentrations were significantly greater than background
concentrations for all five power levels. TOC concentrations were highest at 100 power and then
generally decreased with decreasing power levels; TOC concentrations were lowest at idle.
                                          329

-------
                                                                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.16. Mean Concentration Results, UNDS Engine Wet Exhaust Discharge and Background Samples for the LCPL1
Analyte
Model
RPM 2050
(100% Power)
Mean
Mode 2
RPM 1850
(75% Power)
Mean
Mode3
RPM 1650
(50% Power)
Mean
Mode 4
RPM 1300
(25% Power)
Mean
ModeS
RPM 750
(0% Power)
Mean
Background Water
Mean
Units
Classical Parameters
Nitrate/Nitrite (NO2+ NO3-N)
Total Organic Carbon (TOO)
ND (0.010)
1.15
0.011
1.03
0.011
0.933
ND (0.010)
0.858
0.012
1.73
ND (0.010)
0.992
mg/L
mg/L
Metals
Arsenic, Total
Cadmium, Total
Chromium, Total
Copper, Total
Lead, Total
Nickel, Total
2.22
0.032
0.574
21.7
0.369
4.12
1.98
0.028
0.431
26.0
0.188
4.79
1.92
0.024
0.313
17.2
0.145
3.04
2.38
0.022
0.310
13.5
0.118
2.81
2.21
0.022
0.260
40.1
0.127
14.8
2.29
0.020
N 0(0.100)
0.780
0.030
0.477
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
SVOCs
Bis(2-ethylhexyl)phthalate
Phenol
ND(10.0)
ND(10.0)
ND(10.0)
ND(10.0)
ND (10.18)
ND (10.18)
ND(10.0)
ND(10.0)
20.4
19.7
ND(10.0)
ND(10.0)
ug/L
ug/L
Source: USEPA, 2008b.
(1) Mean values were estimated based on the replicate concentrations for each mode or background sample using a lognormal or modified-delta lognormal distribution.
ND - Not detected (number in parentheses is reporting limit).
                                                                   330

-------
                                                                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.17. Mean Concentration Results, UNDS Engine Wet Exhaust Discharge and Background Samples for the RIB1
Analyte
Model
RPM 2450
(100% Power)
Mean
Mode 2
RPM 2270
(75% Power)
Mean
Mode3
RPM 1720
(50% Power)
Mean
Mode 4
RPM 1290
(25% Power)
Mean
ModeS
RPM 400
(0% Power)
Mean
Background
Water
Mean
Units
Classical Parameters
Biochemical Oxygen Demand (BOD)
Nitrate/Nitrite (NO2 + NO3-N)
Total Organic Carbon (TOC)
Total Suspended Solids (TSS)
ND (2.00)
0.017
1.67
11.9
ND (2.00)
ND (0.010)
1.55
12.4
ND (2.00)
0.015
1.27
ND (5.00)
4.8
0.012
1.15
5.3
3.3
0.013
1.29
ND (5.00)
3.3
ND (0.010)
0.832
ND (5.00)
mg/L
mg/L
mg/L
mg/L
SVOCs
Phenol
32.4
24.6
ND(10.0)
ND(10.0)
ND(10.0)
ND(10.0)
ug/L
VOCs
1 ,2,3-Trimethylbenzene
1 ,3,5-Trimethylbenzene
12.3
12.3
ND(10.0)
ND(10.0)
ND(10.0)
ND(10.0)
ND(10.0)
ND(10.0)
12.6
12.6
ND (10.0)
ND(10.0)
ug/L
ug/L
Source: USEPA, 2008b.
(1) Mean values were estimated based on the replicate concentrations for each mode or background sample using a lognormal or modified-delta lognormal distribution.
ND - Not detected (number in parentheses is reporting limit).
                                                                   331

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.5    Engine Dewinterizing Effluent

       Marine engines used in cold climates typically require maintenance prior to winter
storage to prevent engine damage caused by freezing or corrosion. The indirect cooling systems
in inboard engines are typically winterized by draining the water from the ambient water cooling
system and refilling the system with approximately 5 gallons of antifreeze. Marine engine
antifreeze contains propylene glycol37, corrosion inhibitors, and other additives. In spring, the 5
gallons of antifreeze is emptied by starting the engine, which discharges the glycol solution and
replaces it with ambient water. EPA sampled dewinterizing effluent from an inboard engine on a
recreational vessel as it was converted from winter storage. This sample was collected in the
same manner as that used for sampling other engine effluents. The sample was analyzed for
select classical pollutants and metals.

       Table 3.6.18 presents the collected dewinterizing effluent data, together with the mean
inboard propulsion engine effluent concentrations from Tables 3.6.2 and 3.6.3. The source of the
biochemical oxygen demand concentrations is the propylene glycol in the antifreeze. Elevated
metals concentrations in dewinterizing effluent compared to those in inboard engine effluent
could have been due to prolonged contact of the antifreeze with  the engine cooling system and
associated piping.

       Outboard engines are winterized by  spraying an oily aerosol, commonly referred to as
"fog,"  into the combustion air intake while the motor is running. Therefore, the engine
dewinterizing effluent sample results in this subsection are not applicable to outboard engines.
37 Ethylene glycol is not used for marine applications due to its higher toxicity as compared to propylene glycol.
                                           332

-------
                            Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 Table 3.6.18. Comparison of Dewinterizing Effluent with Propulsion Effluent
Analyte
Units
Dewinterizing Effluent
Inboard Propulsion Engine Mean Concentration
from Tables 3.6.2 and 3.6.3
Classical Parameters
Biochemical Oxygen Demand
(BOD)
Total Residual Chlorine
Turbidity
mg/L
mg/L
NTU
11
2.8
350
Not analyzed
0.0481
322
Metals
Aluminum, Dissolved
Aluminum, Total
Antimony, Dissolved
Antimony, Total
Arsenic, Dissolved
Arsenic, Total
Barium, Dissolved
Barium, Total
Calcium, Dissolved
Calcium, Total
Chromium, Dissolved
Chromium, Total
Cobalt, Dissolved
Cobalt, Total
Copper, Dissolved
Copper, Total
Iron, Dissolved
Iron, Total
Lead, Dissolved
Lead, Total
Magnesium, Dissolved
Magnesium, Total
Manganese, Dissolved
Manganese, Total
Nickel, Dissolved
Nickel, Total
Potassium, Dissolved
Potassium, Total
Selenium, Dissolved
Selenium, Total
Sodium, Dissolved
Sodium, Total
Vanadium, Dissolved
Vanadium, Total
Zinc, Dissolved
Zinc, Total
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
560
3,700
2.1
2.4
24
32
43
59
21 ,000
25,000
820
720
8.7
12
370
820
3,300
20,000
19
64
5,200
6,400
160
400
7.2
18
23,000
23,000
45
54
690,000
630,000
230
190
570
900
2001
3401
Not detected
Not detected
4.22'4
4.52'4
351
361
80.0001
81.0001
1.2
1.3
Not detected
Not detected
16
18
64
2502
1.5
3.0
200.0001
200.0001
43
551
4.4
4.61
32.0001
32.0001
111'4
112'4
770.0002
860.0002
Not detected
Not detected
38
38
(1)  Sample concentrations were almost completely accounted for (>90 percent) by background concentrations in ambient water.
(2)  Sample concentrations were predominantly accounted for (>50 percent and <90 percent) by background concentrations in
    ambient water.
(3)  Measured concentrations well above their respective reporting limits for dissolved arsenic and selenium are suspected of being
    elevated due to positive interference
                                                    333

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

       3.2.6.6    Summary of the Characterization of Engine Effluent Analyses

       Tables 3.6.19 and 3.6.20, and Table 3.6.21 at the end of this subsection, compare effluent
characteristics for inboard and outboard propulsion engines and generator engines. Specifically,
Table 3.6.19 compares the number of analytes detected in effluent from these engines, while
Table 3.6.20 compares engine effluent analyte concentrations for those pollutants that may have
the potential to lead to environmental impacts. Finally, Table 3.6.21 summarizes the specific
analytes within each engine effluent type with the potential to pose risk to human health or the
environment. The Table 3.6.21 is presented here to help interpret a realized risk likely posed by
these analytes in engine  effluent as  summarized in Chapter 5.

Table 3.6.19. Comparison of Number of Detected Analytes in Engine Effluent
Analyte Class
Classical Parameters
Metals
SVOCs
VOCs
Total
Number of Analytes Detected in Engine Effluent
Inboard Propulsion
11
16
31
38
96
Outboard Propulsion
11
14
7
18
50
Generator
11
11
26
28
76
Table 3.6.20. Comparison of Results for Selected Analytes in Engine Effluent
Analyte
Temperature Differential
Oil and Grease (HEM)
Arsenic, Total
Copper, Dissolved
Lead, Dissolved
Lead, Total
Selenium, Dissolved
Zinc, Dissolved
PAHs
Benzene
Units
°C
mg/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
Mean Concentration
Inboard Propulsion
5 (low power levels)
20 (high power levels)
3.0
4.51
16
1.5
3.0
111
38
1 4 total detected
6 carcinogens
12
Outboard Propulsion
<5
Not detected
241
3.3
Not detected
Not detected
761
11
1 detected
0 carcinogens
13
Generator
<5to13
2.9
Not detected
6.5
Not detected
Not detected
Not detected
13
5 detected
0 carcinogens
5.9
(1) Measured concentrations well above their respective reporting limits for dissolved arsenic and
selenium are suspected of being elevated due to positive interference.

       Among all engine types, the SVOCs and VOCs were the most frequently detected
pollutants (Table 3.6.19). Concentrations of PAHs were potentially high in inboard engine
effluent. Fourteen PAHs were detected, including six of the seven PAHs classified as known
carcinogens (Table 3.6.20), but these were only detected in a single inboard engine effluent from
a gasoline engine of a recreational vessel (not a study vessel) dewinterized immediately prior to
sampling. PAH concentrations in this sample were  several hundred to over 1,000 times greater
than their associated benchmarks. PAHs were also detected in outboard engine and generator
                                           334

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

effluents, but at concentrations lower than their associated benchmarks. Furthermore, none of the
probable human carcinogens were detected in generator or outboard propulsion engine effluent
samples.

       The plasticizer bis(2-ethylhexyl) phthalate was found in the effluents of all engine types,
PHQs were just above 1; however, the measured concentrations appear to be largely reflective of
ambient concentrations. The VOC benzene was also found at concentrations above the PHQ
screening benchmarks in all engine effluents. Trimethylbenzenes and ketones (VOCs) were
frequently detected in the effluents of inboard engines, but no screening benchmarks exist for
these compounds. Despite the high frequency of concentrations of benzene that exceeded
screening benchmarks in engine  effluent of all types, rarely were PHQs in excess of 5.

       Among the classical pollutants, inboard propulsion engines increase cooling water
temperatures by moderate amounts (<5°C) at low power levels, but by as much as 20°C at higher
power levels. In contrast, outboard propulsion engines increase cooling water temperatures by
<5°C, regardless of engine level. Most of the generator engine effluent samples increased
cooling water temperature by <5°C; however, two of the generator engine effluent samples had
greater temperature differentials.

       Oil and grease was  not detected in effluent from outboard propulsion engines, but was
detected at concentrations just above reporting limits in  effluent from inboard propulsion and
generator engines. Such concentrations were well below PHQ screening benchmarks for
saltwater discharge. However, EPA did occasionally observe a sheen in receiving waters where
marine engines  were operating.

       Table 3.6.21 lists those metals that were found to be contributed primarily by engine
operations (elevated above ambient water concentrations)  and were detected at concentrations
that exceed a NRWQC, indicating that they may have the potential to cause environmental
impacts. After accounting for background concentrations,  dissolved concentrations of copper
exceeded NRWQC in most inboard engine and generator effluents. The highest PHQ for
dissolved copper was 17. Several effluents from inboard and outboard  engines had dissolved
selenium at concentrations approximately two to seven times higher than NRWQC benchmarks;
however, these measured concententrations are suspected of being elevated due to positive
interference. Among the total metals, PHQs for arsenic were much greater than 1 in both inboard
and outboard engines. However,  as in case of dissolved  selenium, all of the arsenic values
measured above reporting limits  are suspected of being elevated due to positive interference.
Total arsenic was not detected in generator effluents.
                                          335

-------
                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.6.21. Characterization of Engine Effluent and Summary of Analytes that May Have the Potential to Pose Risk
Vessel Type (no. vessels)
Inboard Engines
Water Taxis (4)
Tour Boats(3)
Fishing Vessels (2)
Tow/Salvage Vessel (1)
Fire Boat(1)
Recreational Vessel (2)4
Outboard Engines
Tow/Salvage Vessel (4)
Research (2)
Analytes that May Have the Potential to Pose Risk in Engine Effluent Discharge and Vessel Sources1 2
Microbiologicals










Volatile Organic Compounds

Benzene
Benzene



Benzene

Benzene
Benzene
Semivolatile Organic Compounds

Bis(2-ethylhexyl)
phthalate



Bis(2-ethylhexyl)
phthalate
PAHs5



Metals (dissolved)

Cu
Cu
Cu
Cu
Cu
Cu



I
V)
1










0)
V)
re
£
O
•o
c
re
6










Sulfide






X



Short -Chain Alkylphenol
Ethoxylates or NP










Long-Chain Alkylphenol
Ethoxylates










Nutrients










o
e
•o
c
re
O
O
O
00










Total Suspended Solids










Other Physical/Chemical
Parameters

Temp3
Temp3
Temp3
Temp3
Temp3
Temp3



                                                            336

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
 Table 3.6.21. Characterization of Engine Effluent and Summary of Analytes that May Have the Potential to Pose Risk

Generator Engines (5)
Analytes that May Have the Potential to Pose Risk in Engine Effluent Discharge and Vessel Sources1 2

Benzene
Bis(2-ethylhexyl)
phthalate
Cu


X





Temp,
TRC
Notes:
(1) Analytes are generally bolded when a large proportion of the samples have concentrations exceeding the NRWQC (e.g., 25 to 50 percent), when several of the samples have
PHQs > 10 (e.g., two or three of five), when a few samples result in PHQs greatly exceeding the screening benchmark (i.e., 100s to 1,000s), or, in the case of oil and grease and for
nonylphenol, when one or more samples exceed an existing regulatory limit by more than a factor of 2. See text in Section 3.1.3 for a definition of PHQs and Table 3.1 for screening
benchmarks used to calculate these values.
(2) EPA notes that the conclusion of potential risk is drawn from a small sample size, in some cases a single vessel, for certain  discharges sampled from some vessel classes.   EPA
included these results in the tables to  provide a concise summary of the data collected in the study, but strongly cautions the reader that these conclusions, where there are only a few
samples from a given vessel class,  should be considered preliminary and might not necessarily represent pollutant concentrations from these discharges from other vessels in this
class.
(3) At full (100%) power.
(4) For inboard engine effluent, higher measured concentrations and concentrations that exceeded the screening benchmarks were consistently from the recreational vessel, which
was de-winterized immediately prior to sampling (see text). The recreational vessel was the only vessel sampled that used gasoline instead of diesel fuel. PHQs for the majority of
samples were less than 5.
(5) All PAHs detected (6 of which are  probable human carcinogens) were from one sample collected from a recreational vessel with a gasoline engine dewinterized immediately prior
to sampling and after a winter of non-use.
                                                                            337

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
   3.2.7  Firemain Discharges

       The primary purpose of the firemain system is to supply water for fire fighting, although
this system can also be used for other secondary purposes (deck washing, various maintenance
and training activities, anchor chain washdown, or to create bypass flow from the firemain
pumps to cool auxiliary machinery equipment) onboard the vessels of interest in this study. The
firemain systems (see Section 1.5) sampled by EPA on three tour boats, two tug boats, and the
single fire boat for this study are generally only used during emergencies and during biweekly
testing. The firemain system intake water sampled on the vessels selected in this study was taken
from the surrounding (ambient) water without addition of foam-forming agents such as aqueous
film-forming  foam (AFFF) or other chemical  additions.
                               The Firemain Hose on a Tour Boat

       It should be noted that AFFF agents could potentially be used on the vessels of interest
in this study, although none of the vessels were outfitted with systems that used AFFF. AFFF
agents are used for fire suppression and are a combination of fluoro-chemical surfactants,
hydrocarbon surfactants, and solvents that are injected into the water stream of a fire hose. These
film-forming agents can form water solution films on the surface of flammable liquids,
separating the fuel from the air (oxygen).

       EPA focused on analyzing the samples of firemain discharge water for metals, VOCs,
and SVOCs. Metals were selected for analysis because water in the "wet type"  firemain system
passes through a significant amount of metal pipe onboard most vessels. EPA initially selected
                                          338

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

VOCs and SVOCs to characterize the AFFF, which, as noted, none of the vessels sampled in the
study used while testing their firemain systems. Despite the lack of AFFF use while testing
firemain systems, EPA decided to analyze for VOCs and SVOCs in firemain system discharge
water anyway.

     3.2.7.1   Metals

       Only half the total number of metals analyzed for in water samples from firemain systems
were detected in the six vessels sampled.

       Figure 3.7.1  presents the concentration ranges for dissolved metals detected in firemain
water samples. The figure shows that dissolved metals concentrations span two orders of
magnitude. Average dissolved concentrations of aluminum and zinc were highest, followed, in
order of decreasing concentration, by barium, copper, manganese, nickel, and lead.

       Figure 3.7.2  presents the concentration ranges of total metals detected in firemain water
samples. Except for barium (dissolved:total metal ratio, or fd, of 0.96), total metal concentrations
were much higher than their corresponding dissolved metal concentrations, particularly for lead
and copper. For the other total metal concentrations detected at higher levels, a disproportionate
amount of the metals in ambient water is in the particulate form (i.e., aluminum, manganese and
probably iron).

       Arsenic, cadmium, selenium, antimony, beryllium, cobalt, silver, thallium, and vanadium
were not detected in the firemain discharges.

       Dissolved and total aluminum and total manganese were detected in the firemain effluent
of all  six of the vessels sampled. These metal concentrations are moderately to strongly
influenced by ambient water concentrations. Dissolved zinc, also moderately influenced by
ambient water, was detected in five of the samples. Dissolved and total copper, as well as
dissolved manganese, were detected in four of the samples and  were generally not affected by
ambient water concentrations. Total lead was detected in three of the samples, and only one of
the firemain systems had dissolved lead and chromium at detectable levels. Dissolved and total
barium and total iron were also detected in one sample from a firemain system.

       Disparities between dissolved:total metal  concentrations sampled in firemain water
versus ambient water suggest chromium, lead,  and iron detected in firemain samples at least
partially originated from the network of pipes within the firemain system. The dissolved:total
metal ratio for copper was lower in the firemain water samples than in the  ambient water samples
(fdS of 0.79), suggesting the possibility that some of the total copper detected in firemain samples
originated from the network of pipes within the vessels that support the firemain system - most
likely due to corrosion. Dissolved:total concentrations in firemain samples for the remaining
metals (aluminum, barium, zinc, manganese, nickel) were similar to corresponding ambient

                                           339

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

dissolved:total concentration ratios, suggesting most of these metals detected in firemain samples
originated from the ambient water. Ambient harbor water data are not shown.

       Figures 3.7.3 and 3.7.4 display the distribution of PHQs based on the most conservative
(most protective) screening benchmark for each of the dissolved and total metals. PHQs for only
one of the dissolved metals (copper) include a value of greater than 10 (one dissolved copper
concentration from the firemain system analyzed from a tour boat resulted in a PHQ of 24).
PHQs with values of slightly higher than 1 were found for two other dissolved metals (lead and
zinc) when using the most conservative (most stringent 2006 NRWQC) screening benchmark. In
contrast, all of the concentrations for total aluminum and the concentrations for the single
detected total iron value exceeded the most stringent 2006 NRWQC; however, none of these
PHQs exceeded 11.

       In summary, the concentration of metals in firemain water was generally lower than some
other discharges (e.g. bilgewater, deck washdown water). The water used in the vessel firemain
systems analyzed in this study was ambient water, and the concentrations of most of the
dissolved and total metals in firemain water reflect these surrounding ambient concentrations.
Aluminum, manganese, and iron had high concentrations in the ambient water from which the
firemain withdrew water and were generally higher or the same as other discharges. Dissolved
and total copper, dissolved and total lead, and to a lesser degree, nickel and zinc, were found in
concentrations higher than the ambient water. Of these metals, dissolved copper is the only metal
also found at concentrations consistently above the most conservative screening benchmarks,
albeit only with PHQ values in the 1 to 11 range, which is considerably lower than values found
in most other discharge types discussed in this report.
                                          340

-------
                                                                                  Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.7.1. Results of Firemain  System Sample Analyses for Metals1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone. 5
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone. 5
Screening
BM2
Heavy and Other Metals
Aluminum, Dissolved3
Aluminum, Total4
Barium, Dissolved3
Barium, Total3
Chromium, Total4
Copper, Dissolved
Copper, Total
Iron, Total
Lead, Dissolved
Lead, Total
Manganese, Dissolved4
Manganese, Total4
Nickel, Dissolved4
Nickel, Total4
Zinc, Dissolved4
Zinc, Total
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
6
6
1
1
6
6
6
1
6
6
6
6
6
6
6
6
6
6
1
1
1
4
4
1
1
3
4
6
1
2
5
6
100
100
100
100
17
67
67
100
17
50
67
100
17
033
83
100
110
330
36
37
1.7
23
150
3800
2.1
50
17
86
4.9
7.0
120
490
140
360



15
70


7.6
16
98


58
280
15
180









49



20
15
180









49



20
72
200









59


5.3
26
150
440


1.2
40
290

1.1
81
31
120
1.1
11
270
1200
160
650


4.9
74
580

4.3
270
47
120
4.4
11
370
1600
160
650


4.9
74
580

4.3
270
47
120
4.4
11
370
1600
NA
87
NA
1000
NA
3.1
1300
300
2.5
NA
NA
100
8.2
610
81
7400
Cationic Metals
Calcium, Dissolved3
Calcium, Total3
Magnesium, Dissolved3
Magnesium, Total3
Sodium, Dissolved3
Sodium, Total3
Potassium, Dissolved3
Potassium, Total3
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
ug/L
6
6
6
6
1
1
1
1
6
6
6
6
1
1
1
1
100
100
100
100
100
100
100
100
27000
30000
6500
7300
38000
37000
3800
3600
25000
29000
6500
6600




23000
23000
5200
5500




23000
23000
5200
5500




24000
23000
5700
6200




29000
38000
7200
9200




37000
40000
9000
9800




37000
40000
9000
9800




NA
NA
NA
NA
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with Vi of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated when analytes were
detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank cell reflects a situation when a median or
percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at least that percentage of the values fall. So the 10th percentile is the
concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
(3) Sample concentrations were strongly influenced by background concentrations in ambient water, accounting for greater than 90% of sample concentrations in the majority of samples.
(4) Sample concentrations were moderately influenced by background concentrations in ambient water, accounting for between 50 and 90% of sample concentrations in the majority of samples.
(5) In some cases, the detected concentration(s) for an analyte could be lower than the replacement value (1/2 of the reporting limit) for a concentration that was nondetected. In  an extreme (but possible) case,
this could result in an  average concentration for an analyte that is greater than the maximum detected concentration.
                                                                                      341

-------
         100
     O)
     ^

     c
     .o
     "-I—•
     03
     CD
     O
     C
     o

     O
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
                    i      r
                    j	L
                                 Q
-©•
oo
                                             oo
                                                          o
                          O
                oorgroo
                   J	I
                             Dissolved Metals
Figure 3.7.1. Box and Dot Density Plot of Dissolved Metals Concentrations Measured in

Samples of Firemain Water
                                       342

-------
                    Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
        1000
   O)
   ^

   c
   o

   "co
   CD
   o
   c
   o
   o
        ooo
100
  10
                   1\
                   Q


                  nip
                                            1\
                                            Q
              -©-
                        O
                                  O
                                 oo
                                       p
                                       oTo
                                                          O
                                                          o.o
                                  o
                                          ODOO
                           oono
                   I     I     9
                                            oc
                              I	I
J	L
                               Total Metals
Figure 3.7.2. Box and Dot Density Plot of Total Metals Concentrations Measured in

Samples of Firemain Water
                                      343

-------
       CD
      "-I—•
       O
       ^

      O

      •a

       CD
       N
       CD


      ID
      "-I—•
       c
       CD
      -i—•
       O
      Q_
                      Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
           10.0
1.0
             0.1
                                                   o
                                                                o
                                      I       I      I       I
                ^
                   ^
                                 Dissolved Metals
Figure 3.7.3. Box and Dot Density Plot of Potential Hazard Quotients for Dissolved Metals

in Samples of Firemain Water
(Note: Replacement values for non-detects are circled).
                                         344

-------
  O
  ^

  a

  •a

  CD
  N
  CD
  o
  Q_
                     Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
      10.000
1.000
0.100
        0.010
        0.001
            i     r
                    Q

                    Q|Q.
             I	I
                                         \     \
                                        -Q-
              i     r
        .CHJC1	
                                          O.
                           •0-
                                   O
                                   O
                                                       CIO.
                                                             O
                                                            bol
I	I
I	I
                                 Total  Metals
Figure 3.7.4. Box and Dot Density Plot of Potential Hazard Quotients for Total Metals in

Samples of Firemain Water

(Note: Replacement values for non-detects are circled).
                                      345

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

     3.2.7.2   Classical pollutants

       The firemain system water samples were analyzed for 10 classical pollutants (BOD,
COD, TOC, and sulfide were not analyzed for as they were not expected in firemain system
discharge (see Table 2.2). Of the  10 classical pollutants analyzed for, oil and grease (measured as
HEM and SGT-HEM) were not detected in any samples (Table 3.7.2). The concentrations of all
other pollutants, with the possible exception of turbidity, were not elevated.

       The conductivity, pH, and low salinity (ranging from 0.01 to 0.2 parts per thousand) in
the firemain water samples are consistent with freshwater ambient water (all firemain samples
were taken from vessels  operating in fresh water). The pH of these waters was between 7 and 8,
and turbidity and TSS was low, under 90 NTU and 20 mg/L, respectively.  The firemain system
effluent was sampled in the spring, and the temperature was in a seasonal range of 14 to 22°C
and varied according to geographic location  (warmer water samples in southern United States
and colder in mid-Atlantic and northern states).  Dissolved oxygen in firemain system water
ranged from a low of 4.1 mg/L (slightly less then 50 percent  saturation) to a high of 13 mg/L
(super-saturated).  All of these values were, to a  large degree, consistent with concentrations of
these parameters found in respective ambient water.

       Figure 3.7.5 illustrates the variability of the values measured for the classical pollutants
in firemain system water, which is relatively low given the relative similarities in ambient water
quality (freshwater harbors sampled during springtime) for the three locations where vessels
were sampled. The only  other parameters detected in this category were TRC and turbidity. TRC
was only detected in one of the six samples collected (measured at the reporting limit = 0.10
mg/L; PHQ =  13). All of the other TRC concentrations were below the reporting limit of 0.10
mg/L, which, when reported at half the reporting limit or 0.05 mg/L, still exceeds the most
stringent 2006 NRWQC for TRC of 0.0075 mg/L. In contrast, turbidity ranged from a low of 4.6
to a high of 89 NTU, concentrations similar  to the range of turbities (3 to 180 NTU) observed in
estuaries. In contrast, turbidity in raw sewage can be several hundred NTUs or more. There is no
screening benchmark for turbidity from which to assess potential to cause or contribute to
adverse effects on water quality.

       To summarize, the concentrations of classical pollutants in firemain system water
samples are within the normally expected ranges for the given season and geographical location
where vessels were sampled. It appears that the classical pollutant concentrations primarily
reflect concentrations found in the ambient water.
                                          346

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.7.2. Results of Firemain System Water Sample Analyses for Classical Pollutants1
Analyte
Conductivity
Dissolved Oxygen
PH
Salinity
Temperature
Total Residual Chlorine
Total Suspended Solids (TSS)
Turbidity
Units
mS/cm
mg/L
SU
ppt
C
mg/L
mg/L
NTU
No.
samples
5
5
6
5
5
6
1
6
No.
detected
5
5
6
5
5
1
1
6
Detected
Proportion
(%)
100
100
100
100
100
17
100
100
Average
Cone.
0.32
7.7
7.4
0.12
18
0.05
16
33
Median
Cone.
0.24
6.8
7.4
0.10
19


27
Minimum
Cone.
0.23
4.1
6.9
0.010
14


4.6
10%
0.23
4.1
6.9
0.010
14


4.6
25%
0.24
4.9
7.0
0.055
15


16
75%
0.43
11
7.8
0.20
21
0.025

48
90%
0.47
13
7.9
0.20
22
0.10

89
Maximum
Cone.
0.47
13
7.9
0.20
22
0.10

89
Screening
BM2
NA
NA
NA
NA
NA
0.0075
30
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            347

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

-1— •
c
g 10.00
E
<
i_
o
.1 10°
"co
-i— •
c
CD
O
§ 0.10
o

n ni
: 1 1 1 1 1 1 T 1 i
00
r^ m ("i (_j (_}
,^ f
Q

I OOifffiiOO I
oo
I $ ~
- © ^


— —
I on I

ono
oo


oo o
~ ~
on ~

OO
1 1 l 1 1 1 1 1
         0°"
                               Classical  Pollutants
Figure 3.7.5. Box and Dot Density Plot of Classical Pollutants Measured in Samples of
Firemain Water
(Note: Concentrations reflect ambient water concentrations and values because ambient water was used as the
source of water for all fireman systems in the vessels sampled in the study program).
                                           348

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
     3.2.7.3   Volatile and Semivolatile Organic Chemicals

       VOC and SVOCs were targeted in firemain systems for this program because of the
expectation that AFFF agents might be injected into the water stream of a fire hose to practice
potential fire suppression scenarios. AFFF was not used, however, by any of the vessels sampled
for this study.

       Of the 57 SVOCs that were analyzed for in the six firemain system water samples, only
six were detected, none of which were detected in more than one sample (Table 3.7.3 and Figure
3.7.6). Similarly, of 37 VOCs analyzed for, only five were detected, and as with the SVOCs,
none were detected in more than one sample (Table 3.7.3). When SVOC and VOC
concentrations were above detection levels, concentrations were relatively low. Of these, only
bis(2-ethylhexyl) phthalate was measured at a sufficiently high concentration of 4.6 |ig/L that
exceeded the associated PHQ of 3.8, based on the most conservative screening benchmark of 1.2
|ig/L (human health criterion). Bis(2-ethylhexyl) phthalate was also the only SVOC or VOC
detected in ambient water, but interestingly, at a slightly higher concentration of 13 |ig/L.
                                          349

-------
                                                                         Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.7.3. Results of Firemain Water Sample Analyses for SVOCs1
Analyte
Units
No.
samples
No.
detected
Detected
Proportion
(%)
Average
Cone.1
Median
Cone.
Minimum
Cone.
10%
25%
75%
90%
Maximum
Cone.
Screening
BM2
SVOCs
2,6,1 0,1 4-Tetramethyl
Pentadecane
2-Mercaptobenzothiazole
Benzothiazole
Bicyclo[2.2.1 ]heptane,1 ,7,7-
Trimethyl-
Bis(2-ethylhexyl) phthalate
lsopropylbenzene-4,methyl-1
ug/L
ug/L
ug/L
ug/L
M9/L
Ug/L
1
1
1
1
4
1
1
1
1
1
1
1
100
100
100
100
25
100
9.9
4.1
7.2
14
2.1
9.9




























3.4





4.6





4.6

NA
NA
NA
NA
1.2
NA
VOCs
1 -Methyl-2-(1 -Methylethyl)-
Benzene
1 -Methyl-4-(1 -Methylidene)-
Cyclohexane
Limonene
n-Pentadecane
n-Tetradecane
ug/L
ug/L
ug/L
ug/L
ug/L
1
1
1
1
1
1
1
1
1
1
100
100
100
100
100
97
6.8
9.5
3.8
3.5



































NA
NA
NA
NA
NA
Notes:
(1) Nondetect (censored) concentrations were replaced with % of the reporting limit for calculating average concentrations. The remaining statistics in this table were only calculated
when analytes were detected at a sufficient frequency. For example, if an analyte was detected in fewer than 50% of samples, then a median concentration was not calculated. A blank
cell reflects a situation when a median or percentile could not be computed based on detected concentrations. The percentiles are the concentrations of each analyte below which at
least that percentage of the values fall. So the 10th percentile is the concentration below which at least 10% of the observations were found.
(2) Screening BM represents the screening benchmark referred to in Section 3.1.3, and is the most stringent 2006 NRWQC or other conservative benchmark used to calculate PHQs.
                                                                            350

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
        CD
       -i—•
        C
        CD
        O
        c
        o
       O
16
14
12
10
  8

  6

  4
                                                         0
                                                      Q0O
                0       1
                            345
                           SVOCs
7
Figure 3.7.6. Box and Dot Density Plot of SVOC Concentrations Measured in Samples of
Firemain Water
SVOCs are identified as follows:
(1) 2,6,10,14-Tetramethyl             (4) Bicyclo[2.2.1]Heptane,l,7,7-        (5) Bis(2-Ethylhexyl) Phthalate,
Pentadecane,                       Trimethyl-,                        (6) Isopropylbenzene-4, Methyl-
(2) 2-Mercaptobenzothiazole,                                           1
(3) Benzothiazole,
                                            351

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment

     3.2.7.4   Summary of the Characterization of Firemain System Water Analyses

       Table 3.7.4 summarizes the specific analytes in firemain system effluent that may have
the potential to pose risk to human health or the environment. EPA's interpretation of a realized
risk likely posed by these analytes, relative to pollutant loadings, background ambient and source
water contaminant levels and characteristics, and other relevant information useful for this
assessment, is presented in Chapter 5.

       The proportion of dissolved to total metals for firemain system discharge was low
overall, relative to other discharge types. Among the dissolved metals, copper was detected in
the highest concentrations and exceeded the NRWQC in the largest number of samples (four of
six samples). The corresponding PHQs for dissolved copper ranged from approximately  4 to
over 20. Dissolved lead and zinc had concentrations that exceeded the most conservative
NRWQC in one and three samples, respectively, but none of the PHQs were above 10. Total
aluminum and iron concentrations exceeded NRWQC benchmarks in all samples, with PHQs
ranging from 1-5 (aluminum) and of approximatley 13 (iron; single sample from a fire boat).
However, most of the aluminum in firemain discharge can be attributed to aluminum in the
ambient waters. Overall, the concentrations of metals in firemain discharge were low compared
to other discharge types.

       Among the classical pollutants,  TRC was the only pollutant of potential concern.
However, TRC was detected right at the reporting limit of 0.10 mg/L in only one of six samples
and the concentration likely reflects an  elevated TRC concentration in the ambient water.

       Finally, the concentration of bis(2-ethylhexyl) phthalate (an SVOC) exceeded the
NRWQC (PHQ = 3.8) in one discharge sample; however, most SVOCs and VOCs sampled for
were below detection limits, and when they were detected, occurred at very low concentrations.
It is noteworthy to reiterate that bis(2-ethylhexyl) phthalate was also the only SVOC or VOC
detected in ambient water, and at a slightly higher concentration (13 |ig/L) than in the one
firemain water sample.
                                          352

-------
                                                                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
Table 3.7.4. Characterization of Firemain Discharge and Summary of Analytes that May Have the Potential to Pose Risk









Vessel Type (no. vessels)








Tour (3)
Tug (2)
Fireboat(1)
Analytes that May Have the Potential to Pose Risk in Firemain Discharge and Probable Source1 7









(/>
re
u
O)
"o

o
u
ii







•O
C
3
i
o

'E
O)
O
0)
•—
15
o





i)

Bis(2-ethylhexyl)
phthalate






. — .


"o
U)
w
•D
C
re
15
I


1
^
Cu(dissolved);
Fe (total)

Cu (dissolved)










0)
U)
re
3
•D
C
re
O
















a)
•D
C
3
CO










0
i)
"^" ^
< s
= U)
(0 u
^ re
O >,
•e x
II
^

C «
'p-s
O >•>
1 ^
0 1
1 III















I/I
c
0
•3
^











o
0

•D
C

Q
O
O

O
00










U)
•D
"o
CO
•D
•D
0)
a.
in
V)

re
'o











re
u
i
o
"(5
'(7) O


>- E
* re
*- re
O Q.
TRC


Notes:
(1) EPA notes that the conclusion of potential risk is drawn from a small sample size, in some cases a single vessel, for certain discharges sampled from some vessel classes.  EPA
included these results in the tables to provide a concise summary of the data collected in the study, but strongly cautions the reader that these conclusions, where there are only a few
samples from a given vessel class, should be considered preliminary and might not necessarily represent pollutant concentrations from these discharges from other vessels in this
class.
(2) Analytes are generally bolded when a large proportion of the samples have concentrations exceeding the NRWQC (e.g., 25 to 50 percent), when several of the samples have
PHQs > 10 (e.g., two or three of five), when a few samples result in PHQs greatly exceeding the screening benchmark (i.e., 100s to 1,000s), or, in the case of oil and grease and for
nonylphenol, when one or more samples exceed an existing regulatory limit by more than a factor of 2. See text in Section 3.1.3 for a definition of PHQs and Table 3.1 for screening
benchmarks used to calculate these values.
                                                                           353

-------
                        Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment


    3.2.8   Antifouling Hull Coatings

       Antifouling hull systems  (AFSs) are specialized paints and other coatings intended to
retard the growth of algae, weeds, and encrusting organisms such as barnacles and zebra mussels
on the underwater portion of vessel hulls. These organisms may foul hulls and other underwater
parts, increasing corrosion and drag, reducing safety and maneuverability, decreasing fuel
efficiency and economy, and lengthening transit times (WHOI, 1952). Vessel hull fouling is
often significant as vessels can move between a diverse range of aquatic environments and
remain in the photic zone that is the most productive region of the water body (Chambers et al.,
2006). Exposed to a variety of organisms, vessel hulls can transfer the organisms into other water
bodies, where they can become invasive species38.
       Figure 3.8.1. Encrusting organisms (left) and weeds (right) growing on vessel hulls
       (figures from the Naval Surface Warfare Center's Carderock Division, West
       Bethesda, Maryland, and the Boating Industry
       Association of Victoria, South Melbourne,
       Australia39).
       The development of AFSs  has a  long history,  as
mariners have tried for centuries to keep vessel bottoms free
of barnacles and  other fouling  growth (Yebra et  al., 2004;
Readman, 2006). Ancient civilizations of the Greeks and the
Romans coated their vessels with lead  sheathing secured  by
What is a Biocide?
A biocide is a chemical
substance capable of
killing living organisms,
usually in a  selective
way.
  For this report, EPA did not evaluate the relationship between Anti Foulant Systems, fouled vessel hulls and the
transport/spread of invasive species  Other studies have shown that fouled vessel hulls contribute to the spread of
invasive species and increase fuel consumption, thereby increasing greenhouse gas emissions and vessel operator
cost. Though it is beyond the scope of this study, preventing vessel hull fouling provides important environmental
and economic benefits, however, as discussed in this section, biocidal anti-foulant paints can also contribute to
environmental degradation.
39 See http://www.dt.naw.mil/sur-str-mat/fun-mat/pai-pro-bra/fou-con-tec/images/fouling.jpg and
http://www.biavic.com.au/files/weedunderhull.jpg. respectively, for access to figures.
                                             354

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
copper nails. These heavy metals were early  examples of using biocides to control  fouling.
Columbus' ships are thought to have been coated with pitch and tallow. In the United Kingdom,
lead sheathing was abandoned by the Navy in the late 1600s,  and antifouling paints containing
tar, grease, sulphur pitch and brimstone were developed (Carberry, 2006). One hundred years
later, copper sheathing was used that prevented fouling through  dissolution of the toxic metal
ions (Readman, 2006). With the introduction of iron ships in the mid-1800s, different antifouling
paints were needed because the copper sheathing reacted with the  hull material to hasten
corrosion of the iron. New paints were developed by adding toxic biocides such as copper oxide,
arsenic,  and mercury oxide to resin binders. Following the  Second World War, the introduction
of petroleum-based resins and health and  safety concerns relating to arsenic- and mercury-
containing paints meant that copper-based paints became most popular (Readman, 2006).

      In the late 1950s and early 1960s, new antifouling paint formulations using tributyltin
(TBT) proved to be excellent in preventing hull fouling. TBT, especially in "self-polishing"
formulations, proved very efficient, and the application of TBT-based paints rapidly expanded.
TBT was frequently formulated together with cuprous  oxide to  control a broader range of
organisms. Not only was antifouling performance improved, but tin-based formulas (without
copper components) are noncorrosive to aluminum, which was  being used more in the
construction of vessel hulls and propulsion systems. Unfortunately, the use of TBT also had
severe and unexpected environmental consequences (Carberry,  2006). As the popularity of TBT
grew, oyster producers in France reported shell malformations caused by paint leachate
containing TBT that rendered their harvest worthless. Wild populations of other mollusc species
were also affected at very low concentrations of TBT in the water and sediment (Evans et al.,
1994). For example, female dog whelks (Nucella sp.) developed male characteristics (termed
imposex) at these levels (Bryan et al., 1986). This masculinization of female gastropods was also
reported in the open North  Sea (Ten Hallers-Tjabbes et al., 1994). TBT use on small vessels was
phased out in the late 1980s, when EPA and  other regulatory agencies (including those in
Canada, Australia, and many in Europe) restricted use  of TBT-based  AFSs to ships longer than
25 meters (see Section 6.2.3 of this report for further discussion about regulatory elimination of
TBT).

      Restrictions on the use of TBT-based AFSs  opened the market for paint manufacturers
and chemical companies developing new biocides for new antifouling paints to be used on
vessels. Other metallic species, such as copper (copper hydroxide, copper thiocyanate) and zinc
(zinc pyrithione),  are currently used as substitutes for TBT. Copper oxide (in formulations
without TBT) is by far the most common of the metallic biocides, used in more than 90 percent
of the approximately 180 AFS products registered in California (Singhasemanon, 2008). A single
AFS product might actually contain multiple biocides,  with "booster biocides" incorporated  to
increase the duration and functionality of copper-based AFSs (Chambers et al., 2006). Irgarol is
currently the organic biocide booster most frequently formulated into AFS products. As was the
case for TBT, the biocides used in AFSs today can be toxic to a range of aquatic organisms,  not

                                          355

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
just fouling organisms. In the subsections below, EPA discusses the literature on studies of
adverse effects of these AFS biocides to aquatic resources as well as alternatives to using
biocidal AFSs.

       EPA did not sample antifouling systems as part of this study because of lack of time and
resources available for this study. Assessing AFS discharge involves isolating a commercial
vessel within a confined body of water (a "boat bag" or slip liner), and measuring the release,
discharge, or leaching of the AFS biocide(s) over a period of time (weeks or months); the
amount of time needed for the study would impose economic hardship on the vessel's owners
and operators. Rather, EPA elected to rely on the significant secondary data on anitifouling
systems available in the literature.

       3.2.8.1    Copper Biocides

       Copper is typically the biocide added to antifouling paints to prevent biofouling
organisms from attaching to the hull. The most common form of copper used in AFSs is cuprous
oxide, which acts as a preventative biocide by leaching into the water body. Cuprous oxide
concentrations in marine antifouling paints range from 26 to 76 percent by weight, with most
paints in the 40- to 70-percent range. Since cuprous oxide is 89 percent copper by weight, typical
cuprous oxide marine antifouling paints are 36 to 62 percent copper by weight (TDC
Environmental, 2004). Two additional copper biocides are occasionally used in AFSs: copper
thiocyanate and copper hydroxide. These formulations are not as common,  although copper
thiocyanate has the advantage of being compatible with aluminum.  The contribution of copper
from these paints to receiving water is small relative to AFSs containing cuprous oxide (TDC
Environmental, 2004).

       Conventional copper-based AFSs fall into several general categories: copolymer or
ablative paints and hard contact leaching paints (Conway and Locke, 1994). Copolymer paints
release biocide at a constant rate, ablating (wearing away) much like a bar of soap, which is
intended to reduce the need for cleaning. Hard contact leaching paints are usually modified
epoxy paints that leach biocide upon contact with water, and, over time, the biocide is released at
a decreasing rate.  Each of these  coating formulations can benefit from periodic hull cleaning to
remove fouling growth, maintain a smooth surface, and improve the copper release on vessel
hulls, but underwater hull cleaning can be a source of pollution or introduce non-native species if
not done carefully. Cleaning frequencies and methods vary by paint type, area of vessel
operation, frequency and conditions of operation, and vessel operator's needs. Techniques that
capture removed fouling growth and paint residue reduce negative impacts  on the environment.

       Passive leaching rates from antifouling paint, including those that are copper-based,
depend on a number of factors, including the paint matrix (e.g., vinyl, epoxy), copper content,
age of the paint, time since last hull cleaning, and frequency of painting. Leaching rates also vary
                                           356

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
with environmental conditions such as pH, temperature, salinity, and the existing slime "biofilm"
layer (CRWQCB, 2005).

       Rates of passive leaching of dissolved copper from AFSs on seven recreational vessels
painted with epoxy copper antifouling paints were investigated in studies conducted in Southern
California by the U.S. Navy, under test conditions intended to represent realistic vessel
conditions. Copper release rates were found to range from 2 to 14 |ig/cm2/day, with an average
leaching rate of 8.2 |ig/cm2/day40. In another study of copper-based AFSs on recreational
vessels, researchers with the Southern California Coastal Water Research Project (SCCWRP)
measured the mass emissions of dissolved copper from both passive leaching and underwater
hull cleaning (Schiff et al., 2003). Fiberglass panels were painted with copper-based antifouling
paints and immersed in seawater in a harbor environment. SCCWRP researchers determined the
average flux rates for epoxy and hard vinyl copper antifouling paints to be approximately 4.3 and
3.7 |ig/cm2/day over the course of a month, respectively. In the SCCWRP study, the authors also
discussed the comparability of the results between the U.S. Navy and SCCWRP studies.
According to the authors, the range of passive leaching measurements from the U.S. Navy study
was within the range of measurements obtained in the SCCWRP study. By combining the results
from the two studies, an average passive leaching rate for vessels at the Shelter Island Yacht
Club (SIYB) was determined to be 6.5 |ig/cm2/day (CRWQCB, 2005). In the United Kingdom,
Thomas et al. (1999) found higher copper leaching rates for ablative copper antifouling paint
ranging from 18.6 to 21.6  |ig/cm2/day in 17 day experiments (Schiff et al., 2003). Table 3.8.1
summarizes the passive leaching rates for vessel AFSs found in the literature. The copper
leaching rates summarized in this table were measured in experiments designed to simulate
environmentally relevant conditions. However, more recently developed types of AFSs may
leach at different rates, and the actual rates of copper leaching from many vessels and real-world
environmental conditions may differ from those in Table 3.8.141.

       Estimates  of copper released from AFS leaching and underwater hull cleaning were
calculated based upon the  6.5 |ig/cm2/day average flux rate cited above, which was  extrapolated
to vessels using the underwater surface area of the hull42, and then to marinas (or harbors) based
on the number of vessels in the marinas. Despite the caveats and limitations discussed above,
EPA uses these estimates in Chapter 4 to calculate loadings from vessel hull AFSs to attempt to
40 EPA notes that a calculated average for release rates will not reflect real-world conditions for many vessels and
environmental conditions.
  Additional test data for copper AFC leaching rates were provided to EPA by the Antifouling Coatings Work
Group (AFWG) of the American Coatings Assoc iation (ACA) during the public comment period. These data
substantially agree with EPA's best estimate of copper AFC leaching rate (6.5 ug/cm2/day) used for water quality
modeling in Chapter 4.
42
  Hull surface area can be estimated using the following equation: Hull Surface Area = VesselLength*Beam*0.85
(Interlux, 1999).
                                           357

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
understand the impacts of this source of copper discharge from certain vessels on large water
bodies.

       Even when an effective AFS is used, the biofouling could accumulate over time to
unacceptable levels. If the AFS is still viable, this accumulated growth can be removed from
vessel hulls by a number of methods, most frequently by underwater hull cleaning. Several
studies have investigated the release of copper from copper-based AFSs into water bodies during
underwater hull cleaning. The amount of copper released depends on cleaning frequency, method
of cleaning, type of paint, and frequency of painting (SWRCB, 1996). Valkirs et al. (1994) found
that underwater hull cleaning resulted in elevated total copper concentrations near the vicinity of
the operation as dissolved copper was released during and shortly after hull cleaning. Smaller
amounts of dissolved copper also leached from debris and sediments after cleaning. The
particulate form of copper was rapidly incorporated into the bottom sediment, likely rendering it
unavailable to aquatic organisms. The biologically active species of copper complexed rapidly,
and dissolved copper levels returned to precleaning conditions within minutes to hours after the
hull cleaning. Valkirs et al. (1994) concluded that potential adverse effects of hull cleaning on
aquatic organisms from the increased dissolved copper concentrations were relatively short-term
and pulsed in nature, while the potential adverse effects of increased particulate copper were
probably long-term in nature, and dependent on resuspension or sediment uptake from benthic
organisms.

       McPherson and Peters (1995) also studied the effects of underwater hull cleaning on
water body copper concentrations and toxicity to aquatic life. In the study, an underwater hull
cleaning operation was performed in Shelter Island Yacht Basin using Best Management
Practices (BMPs) that used less abrasive techniques to remove fouling growth (e.g., hand-wiping
with a soft cloth). Most of the copper released during the cleaning was in the dissolved form.
Researchers found that the plume of copper released by cleaning moved with the current, and
that the degree of plume contamination depended on fouling extent and exertion by the diver.
McPherson and Peters (1995) concluded that underwater hull cleaning elevates concentrations in
the vicinity of the operation, which return to background levels within minutes. The researchers
did not identify the type of antifouling paint (ablative or contact leaching paint), the age of the
antifouling paint on the vessel, or the time since last hull cleaning. While the study provided
important information regarding impacts of underwater hull cleaning on water quality, it did not
provide copper emission rates associated with hull cleaning.

       Schiff et al. also estimated dissolved copper emissions rates associated with underwater
hull cleaning. Fiberglass panels were painted with copper antifoulants to simulate the hulls of
recreational vessels. The study objective was to estimate the flux rates of dissolved copper from
underwater hull cleaning of vessels painted with two commonly used types of copper-based
antifouling paints in San Diego Bay. Schiff found that hull cleaning released between 3.8 to 17.4
ug/cm2 per event (see Table 3.8.2), with an average release of 8.6 |ig/cm2/event. The researchers
concluded that underwater hull cleaning results in a greater daily load of copper to the
                                           358

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
environment than passive leaching. In terms of mass loading, the authors concluded that
approximately 95 percent of dissolved copper from antifouling paint enters the environment via
passive leaching (CRWQCB, 2005). EPA notes, however, that this does not include loading rates
from paniculate copper, which may also impair the environment in the benthos due to
biogeochemical cycling.

       AFSs that are applied to vessel hulls are one of the most  commonly identified major
sources for copper in marinas. A number of studies have been carried out to estimate the loading
of copper from vessel AFSs. EPA estimated that copper loading from AFS use in California's
Lower Newport Bay (LNB) area, which harbors approximately 10,000 boats, contributed more
than 62,000 pounds of copper (via passive leaching and underwater hull cleaning) into LNB
waters annually (USEPA, 2002). EPA believed that this load could account for as much as 80
percent of all copper input into LNB.

       The U.S. Navy and private researchers conducted two copper source loading studies for
the  San Diego Bay in the late 1990s (Johnson et al., 1998; PRC,  1997). Both studies concluded
that AFSs accounted for the majority of dissolved copper loading to the bay. The San Diego
Regional Water Quality Control Board (SDRWQCB) estimated  that passive leaching and
underwater hull cleaning of the 2,400 boats berthed in the SIYB marina combine to contribute 98
percent of the copper load to the basin (Singhasemanon et al., 2009). Of the approximately  1.8
pounds of copper estimated released per boat per year (TDC Environmental, 2004), about 95
percent is believed to leach from AFS while boats are moored at the dock;  the remaining 5
percent is believed to be released during monthly underwater hull cleaning activities.

       The constant input of copper by leaching from the AFSs  applied to pleasure, commercial,
and military vessels has been cited as a likely primary source of copper in  San Diego Bay.
Sediment concentrations measured at the SIYB were relatively high (from  133 to 212 mg/kg)
compared to other areas in San Diego Bay (Valkirs et al., 1994). Elevated copper concentrations
(108 to 270 mg/kg) were found throughout San Diego Bay,  with small boat harbors, commercial
shipping  berths, and military berths most affected. This distribution pattern is expected,
considering the historical use of copper-based antifouling paints in the area.

       Marinas in general tend to have elevated levels of pollutants in the  water and sediments,
including copper, as explained later in this subsection. For example, monitoring in the Southern
California Bight demonstrated that sediment from marinas throughout southern California had
consistently elevated copper levels compared to surrounding waters (Bay et al., 2000). The
National  Oceanic and Atmospheric Administration (NOAA , 1991) found the highest sediment
concentrations, reaching over 104 mg copper/dry kg, in marinas, compared to other areas
throughout the Southern California Bight. Sediment quality surveys around the United States
routinely find high copper concentrations in marinas and harbors (USEPA, 1996; NOAA, 1994).
                                          359

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       A recent study of AFS biocides in California marinas found dissolved copper
concentrations ranging from 0.1-18.4 ug/L (Singhasemanon, 2008) in the water. Concentrations
were significantly higher in salt- and brackish water marinas than in freshwater marinas.
Dissolved copper concentrations in many of the salt- and brackish water marinas exceeded
established water quality standards. Thus, there are ecological risks due to copper in many salt
and brackish water marinas (Singhasemanon, 2008).

       Copper contamination from vessel hulls is a water quality problem that is not unique to
California. Within the United States, other areas of current concern to regulators include
Chesapeake Bay, Maryland; Port Canaveral  and Indian River Lagoon, Florida; and various
harbors in the state of Washington (Carson et al, 2009).

       Elevated copper levels in marinas may be attributable to a number of factors. Marinas are
home to high concentrations of recreational and commercial vessels. Since recreational vessels
spend much of their time moored in marinas, most of the biocide from the antifouling paints on
the vessel hulls is released in the marinas. Moreover, marinas  are purposefully constructed to
shelter boats from currents and waves, so they are not flushed well. Elevated trace metal
concentrations in marinas are partly the result of the lack of mixing and dispersion. Thus, AFS
pollution at these locations would represent some of the worst-case scenarios with regard to
water quality (Singhasemanon et al., 2009; CRWQCB, 2005).

        The biocides leached from AFSs can accumulate in the water of poorly flushed boat
basins to levels that might harm marine life, especially mollusks, crustaceans, and echinoderms
(Johnson and Gonzalez, 2006). At relatively low concentrations, copper is toxic to a wide range
of aquatic organisms, not just fouling organisms (CRWQCB, 2005). Concentrations as low as 5
to 25 ug/L can be lethal for marine invertebrates (Chambers et al., 2006). Elevated copper levels
affect growth,  development, feeding, reproduction, and survival at various life stages offish,
mussels, oysters, scallops, crustaceans, and sea urchins.  High copper levels also change the types
of phytoplankton that thrive in boat basins (Calabrese et al., 1984). Low levels of dissolved
copper affect the olfactory capabilities in juvenile Coho salmon, which is  critical for homing,
foraging, and predator avoidance (Baldwin et al., 2004). The effect of copper on olfaction of
juvenile salmonids suggests that copper might affect other fish species, too. Most effects on fish
are sublethal (e.g., they may hinder metabolic processes, reproduction, development, activity
levels and behavior). Thus, the damage is chronic and less noticeable than, for example, fish kills
caused by sudden oxygen depletion (Evans et al., 1994).

        In the California marina study, significant toxicity was measured in eight of 47 water
samples; seven of the toxic samples came from Marina del Rey (MdR) in Los Angeles
(Singhasemanon et al., 2009). The authors concluded that copper was the most likely cause of
the toxicity in these samples. Two models of copper bioavailability and toxicity to aquatic
organisms, the Biotic Ligand Model (BLM) and dissolved organic carbon (DOC) model,  were
                                           360

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
used to confirm these findings. The BLM and DOC model predictions agreed favorably with the
actual toxicity data, although both models tended to slightly overpredict toxicity, especially when
close to the toxic effect concentration (i.e., EC50) (Singhasemanon, 2008).

       Rivera-Duarte et al. (2003) also investigated the bioavailability and toxicity of copper in
San Diego Bay and found that toxicity was based on chemical speciation and followed the free
ion activity model. The ECso for mussel larval development was observed near 10"11 molar (i.e.,
0.64 ng/L) free copper ion.  The toxic threshold concentration of free copper ion was independent
of spatial and temporal effects, indicating the need to study chemical speciation of copper
released from antifouling paints in order to determine its environmental  effects (Rivera-Duarte et
al., 2003).

       Table 3.8.1. Rates of Passive Copper Leaching from Vessel AFSs
Study
UK (Thomas etal., 1999)
U.S. Navy (Zirino and
Seligman, 2002)
U.S. Navy (Valkirs etal.,
2003)
SCCWRP(Schiffetal.,
2003)
Test Method
Not reported
Not reported
7 recreational
vessels in
recirculating dome
system
Fiberglass panels
in recirculating
dome system
AFS
Ablative copper
antifouling paint
Ablative copper
antifouling paint
Epoxy copper
antifouling paint
Epoxy copper
antifouling paint
Hard vinyl/Teflon copper
antifouling paint
Biocide-free coating
Leaching Rate
(ug/cnWday)
18.6-21.6
Average = 3.9
2-14
(average = 8.2)
4.3
3.7
0.24
                                           361

-------
                Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
  Table 3.8.2. Dissolved Copper Release from Vessel AFSs During an Underwater
  Hull Cleaning "Event"
AFS
Epoxy copper antifouling paint
Hard vinyl/Teflon copper
antifouling paint
Biocide-free coating
Cleaning Method
Less abrasive management
practices
No management practices
Less abrasive management
practices
No management practices
Less abrasive management
practices
No management practices
Copper Release
([jg/cm2/event)
8.6
17.4
3.8
4.2
0.03
0.05
  Source: Schiff et al., 2003

  Table 3.8.3. Estimated Dissolved Copper Mass Emissions from a 9.1m (30ft)
  Powerboat
Source
Passive leaching
(min-max)
Underwater hull
cleaning with BMPs
(min-max)
Total emissions
(min-max)
Dissolved Copper Emission (grams/month)
Epoxy Copper Antifouling Paint
24.9
(23.3-27.8)
1.8
(1 .7-2.0)
26.7
(20.5-33.6)
Hard Vinyl/Teflon Copper
Antifouling Paint
21.4
(15.7-24.5)
0.8
(0.5-1.2)
22.2
(15.0-31.5)
Biocide-Free
Coating
1.4
(0.9-1 .8)
<0.01
(0-0.01)
1.4
(0.9-1.8)
Source: Schiff et al., 2003
                                    362

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       3.2.8.2    Irgarol and Other Organic Biocide Boosters

       Irgarol (Irgarol 1051, 7V-tert-butyl-7V-cyclopropyl-6-methylthio- l,3,5-triazine-2,4-
diamine) is a highly effective biocide used in AFSs to prevent the growth of autotrophic (e.g.,
plants and algae) organisms on vessel hulls. After the ban of tributyltin (TBT) on vessels shorter
than 25 meters, the use of TBT-free paints containing copper compounds and organic booster
biocides such as Irgarol increased considerably and became more widespread (Mohr et al.,
2009).  Other organic biocides, including Diuron (3-(3,4-dichlorophenyl)-l,l-dimethylurea),
dichlorofluanid (1,1 -dichloro-N-(dimethylamino)sulfonyl)-1 -fluoro-N-
phenylmethanesulfenamide), and Sea-Nine (4,5-dichloro-2-n-octyl-4-isothiazolino-3-one) are
also added to AFS preparations to boost performance (Thomas et al., 2001). The use of biocide
boosters is in part a response to concerns about performance, environmental impacts, and,
according to Chambers et al., (2006), a reported increasing tolerance of some macrophytes and
algae to copper. Freshwater locations such as the Great Lakes are plagued primarily by algae
(West Marine, 2008), and booster biocides such as Irgarol are used to restrict the growth of algae
by blocking photosynthesis near the water surface. To date, however, most studies on Irgarol
have focused on marine areas and toxicity tests with marine organisms (Mohr et al., 2009).

       Irgarol has been detected with increasing frequency at ecologically sensitive levels in
coastal waters worldwide, as reviewed by Konstantinou and Albanis (2004). In ports and marinas
in coastal waters, it has been detected in relevant effect concentrations of up to 4.2 ug/L (Basheer
et al., 2002). Levels of up to 1.4 and 2.4 ug/L have been reported from UK marinas and
freshwater sites (Thomas et al., 2002). In the United States, Irgarol and its major metabolite Ml
have been detected in the Chesapeake Bay and Florida (Hall and Gardinali, 2004). In the
California marina study, Irgarol and Ml were detected in all 45 marina samples (Singhasemanon
et al., 2009); Irgarol concentrations ranged from 12 to 712 ng/L,  and Ml concentrations ranged
from 1.6 to 217.1  ng/L. Higher concentrations of irgarol and Ml  were found in salt water
marinas.

       Although Irgarol was predicted to  easily dissipate under natural conditions (Hall et al.,
2005),  it is the most frequently detected antifouling biocide worldwide (Konstantinou and
Albanis, 2004). Published values of the half-life of Irgarol in water are between 24 and 200 days
(Mohr  et al., 2009).
                                           363

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
       EPA has expressed concern over the potential toxic effects of Irgarol on aquatic plants
and algae (USEPA, 2003a). Compared to other triazines like atrazine and simazine, Irgarol is a
more potent inhibitor of algal photosynthesis, and is
therefore highly toxic to macrophytes, phytoplankton,
and periphyton (Mohr et al., 2008). Irgarol is likely to
be much less toxic to animals than flora (Mohr et al.,
2009). The main metabolite Ml is also toxic to aquatic
plants and algae, but in many cases much more than 10
times less toxic than Irgarol.
       Although Irgarol is formulated in AFSs to
control periphyton on vessel hulls, the range of
environmental concentrations measured in freshwater
can be toxic to nontarget macrophytes. The results of
the Mohr et al. (2009) study indicate that Irgarol is
likely to have serious impacts on natural macrophyte
communities at environmentally relevant
concentrations. The fact that Irgarol accumulates in
macrophytes, especially at lower concentrations,
suggests the expected toxicity of Irgarol may be
underestimated (Mohr et al., 2009).
What are Macrophytes,
Phytoplankton, and
Periphyton?
A macrophyte is an aquatic
plant that grows in or near
water and is either emergent,
submergent, or floating.
Phytoplankton are planktonic
algae that live in water bodies.
Periphyton is a complex
mixture of algae,
cyanobacteria, heterotrophic
microbes, and detritus that is
attached to submerged
surfaces in most aquatic
ecosystems.
       Irgarol concentrations at many of the marinas in the California study were high enough to
be toxic to some phytoplankton and aquatic plants (Singhasemanon et al., 2009). For example,
the range of observed Irgarol concentrations (12 to 712 ng/L) exceed aquatic benchmark values
that are protective of 90 percent of aquatic plant species. The Irgarol metabolite Ml never
exceeded the  aquatic benchmark value (Singhasemanon, 2008).


       3.2.8.3    Zinc Biocides


       In  recent years, there has been an increase in the registration of AFS products with zinc
pyrithione (bis(N-oxopyridine-2-thionato)zinc(II)), also commonly known as zinc omadine, as
the primary biocide (Singhasemanon  et al., 2009).

       In  a California marina study, dissolved zinc concentrations from paints containing zinc
omadine ranged from 1.0-66.6 ug/L with a concentration distribution that was similar to
dissolved  copper (Singhasemanon, 2008). Dissolved zinc concentrations were much higher in
saltwater marinas than brackish and freshwater marinas. Zinc concentrations did not exceed
California Toxics Rule (CTR) fresh- and saltwater standards. If zinc pyrithione becomes more
popular as an AFS biocide in the future (e.g., as a replacement for copper AFSs), the
                                           364

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
contributions of zinc AFSs to the marina zinc load will increase and potentially lead to zinc-
related toxicity (Singhasemanon et al., 2009).
       3.2.8.4    Emerging Biocides

       As mentioned in the introduction to this subsection, AFSs using copper-containing
biocides are the most common substitutes for TBT. However, paint manufacturers continue to
search for new antifouling biocides. One promising development is ECONEA, a metal-free
biocide developed by a pharmaceutical company. According to the paint manufacturers,
ECONEA is rapidly biodegradable and does not accumulate in the marine environment, and is
reported by the manufacturer to very effectively control a wide range of invertebrate fouling
organisms in significantly less amounts compared to conventional biocides. However, AFSs
formulated with ECONEA have not entered the market, and independent testing data are not
currently available.
       3.2.8.5    Biocide-Free (Nonbiocidal) AFSs

       In recent years, biocide-free coatings designed to prevent fouling growth from adhering
to boat hulls have entered the market. Biocide-free coatings are designed to produce a slick
surface that prevents fouling organisms from firmly adhering to the hull. Currently available
nonbiocidal bottom coatings may be silicone-based, epoxy-based, water (urethane)-based, or
polymer-based. They do not include biocidal components. Epoxy coatings are durable, and are
expected to last for many years, but require frequent and aggressive cleaning (Johnson and
Miller, 2002). The most commonly used nonbiocidal  coatings are silicone elastomeric coatings,
which are rubbery and are more easily nicked or abraded than epoxy, although recent advances
have improved their durability. They are sometimes called "fouling release" coatings, because
fouling growth is  sheared off the hull once the vessel  exceeds a certain speed (e.g., 20 knots).
Movement of a foul-release-coated vessel through the water dislodges organisms that do adhere.
The utility of these coatings depends on vessel speeds and the proportion of time the vessel is
underway (rather than at dock). Foul-release coatings are typically more expensive than biocidal
AFSs. Because of their expense and operational requirements, foul-release systems generally are
not used on recreational vessels at this time.

       To date, nontoxic AFS alternatives have not been widely accepted in the boating
industry, due to concerns about practicality and cost. If adopted, these alternatives would
eliminate the leaching of biocides from marine antifouling paint, as well as biocide release
during underwater hull  cleaning.

       A number of projects are underway to develop new biocide-free AFSs. The European
Commission is collaborating with industry with the goal of developing a nonbiocidal antifouling
                                          365

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
coating that relies on nanostructuring to impede the adhesion of fouling organisms (Ambio,
2008). The U.S. Navy is sponsoring research by University of Florida engineers to develop a
biocide-free hull coating based on the geometry of shark skin  scales. Chambers et al. (2006)
provide a review of these and other biomimetic approaches to environmentally effective AFSs.

       Because nonbiocidal coatings do not affect fouling growth, they may need more frequent
cleaning than biocide-based AFSs, and can be more effective when used with other practices
designed to increase the amount of shearing and decrease exposure to fouling organisms during
times of inactivity: using the vessel more often and/or operating it at higher speeds; storing it on
land or on a hoist at the slip when not in use; and, surrounding the vessel with a slip liner and
adding 10 to 15 percent fresh water to reduce salinity (Johnson and Gonzalez, 2006).

       3.2.8.6    BMPs

       The most effective way to reduce biocide emissions from AFSs on recreational vessels is
by carefully selecting the AFS. The owner/operator should match antifouling performance with
how the vessel typically operates. Choosing a nonbiocidal AFS can eliminate emissions from
vessels that, for example, operate at high speeds when they are underway. Slow-release
formulations or formulations with lower biocide content may also reduce the release of biocides
into the aquatic environment. As noted previously, passive leaching accounts for most of the
biocide release from recreational vessels, but biocide also could leach into the water body during
underwater hull cleaning and AFS application and removal.

       In addition to AFS selection, other BMPs may be used to limit emissions of toxic
components from AFSs. These BMPs include specifications for capturing and treating materials
removed during underwater hull cleaning, properly managing wastes from AFS application
processes, and capturing and appropriately disposing of old hull coating residue prior to
repainting. When nonbiocidal coatings are used, companion strategies can be used to reduce
fouling including slip liners, boat lifts, and frequent hull cleaning (Johnson and Gonzalez, 2006).

       BMPs for underwater hull cleaning must also address the potential introduction of aquatic
nuisance species (ANS). EPA notes that small vessels are strongly suspected of contributing to
the spread of numerous invasive species including zebra and quagga mussels. Prohibitions on
biocide-containing AFSs could potentially exacerbate the spread of ANS as the toxicity of vessel
hull coatings declines and as water quality improves as a result.

       Pollutants from passive leaching and hull cleaning can be reduced by implementing other
BMPs, such as using nontoxic (or less toxic) antifouling paints to replace copper-based paints.
Switching to nontoxic and less toxic antifouling paints will reduce the loading from both passive
leaching and underwater hull cleaning. For example, if all new boats entering the Shelter Island
Yacht Basin use nontoxic or less toxic coatings and existing boats replace copper coatings with
nontoxic or less toxic coatings at the next routine hull-stripping (as assumed in their total

                                           366

-------
                       Chapter 3 - Analysis of Discharges and Potential Impact to Human Health and the Environment
maximum daily load), the basin's water quality is expected to dramatically improve (CRWQCB,
2005). Additionally, nontoxic or less toxic coatings will require companion strategies such as
slip liners, boat lifts, and frequent hull cleaning to control fouling (Johnson and Gonzalez, 2006).

       3.2.8.7    Conclusion

       Antifouling systems currently used on the majority of recreational and commercial
vessels are paints that prevent and retard fouling growth by leaching biocides, most frequently
cuprous oxide, onto the hull. Biocides can enter a water body through passive leaching,
underwater hull cleaning, hull painting, and AFS removal processes. Biocides leached from
vessel AFSs can accumulate in the water of poorly flushed boat basins to levels that could harm
marine life. Copper from vessel hulls in particular is a water quality concern in many near-
coastal waters of the United States,  including the waters of Southern California, the Chesapeake
Bay, Port Canaveral and Indian River Lagoon in Florida, and in various harbors in the state of
Washington. Copper leaching from vessel hulls has also been reported as a problem in several
European countries, including Sweden, the Netherlands, and Denmark.

       Concerns about impacts to aquatic ecosystems from both TBT and  copper have led to the
development of AFSs that use alternative biocides or are biocide-free. At this time, these
alternatives are relatively costly and have  not been widely accepted by boaters. Releases of
biocidal components of AFS can be reduced by implementing BMPs, including the use of
nontoxic (or less toxic) antifouling paints to replace copper-based paints.
                                           367

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment

CHAPTER 4

POTENTIAL LARGE-SCALE IMPACTS OF STUDY VESSELS'

INCIDENTAL DISCHARGES TO HUMAN HEALTH AND THE

ENVIRONMENT	

       In Chapter 3, EPA described the variety of vessel discharges and the scope and
magnitude of pollutants discharged by 'study vessels.' EPA discussed whether these discharges
of pollutants exceeded a National Recommended Water Quality Criteria (NRWQC) at end-of-
pipe or contained persistent bioaccumulative toxic (PBT) chemicals which could indicate a
potential for environmental effects. Public Law (P.L.) 110-299 tasks EPA with assessing the
potential for discharges incidental to the normal operation of vessels to pose a risk to human
health, welfare, or the environment from all sizes of commercial fishing vessels and other
nonrecreational vessels less than 79 feet in length. As part of this assessment, EPA used a
screening-level model as a tool to evaluate the  cumulative effects of discharges from a
population of such vessels operating in a large  receiving water body.

       EPA developed the screening-level water quality model to assess the impacts of vessel
discharges on a hypothetical harbor environment1. For purposes of the model, EPA developed
several vessel population scenarios that included multiple vessels from numerous vessel classes,
such as fishing vessels, tour boats, water taxis,  and tugboats discharging various waste streams
(e.g., antifouling leachate, bilgewater, engine effluent, graywater). EPA then modeled numerous
scenarios combining the different vessel populations in different hypothetical harbors to
represent a range of environmental conditions potentially observed in harbors  across the United
States.

          Due to the limitations  of this screening-level model, EPA assumed that the
background concentration for all analytes in the harbor water was zero. Although this assumption
is likely unrealistic, removing other loading considerations from model calculations allowed
EPA to evaluate whether incidental discharges from study vessels alone have the potential to
exceed National Recommended Water Quality Criteria (NRWQC) in receiving waters without
any additional sources of pollution. Vessel discharges may have a potential to contribute to
water body impairment when vessel discharge  pollutant concentrations exceed the NRWQC at
end-of-pipe, depending on the quantity of pollutant in the discharge, what other potential sources
of pollution are present, and the characteristics of the  waters in which the vessel is operating.
For example, if a group of vessels contributes a significant quantity of a given pollutant via a
1 For this analysis, the "harbor environment" refers to a large body of water that could potentially have 175 to 300
commercial vessels simultaneously discharging. EPA assumed that the harbor area extended beyond the defined
vessel docking area to include the surrounding water body with an estimated surface area ranging from one to three
square miles.

                                          368

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
discharge into a water body, the impact of the vessel discharge is more likely to contribute to a
water quality exceedance. If a group of vessels contributes only a very small quantity of a given
pollutant via a discharge, the impact of the vessel discharge is less likely to contribute
meaningfully to a water quality exceedance. EPA believes that assessing the potential for vessel
discharges to contribute to water-body impairment is best conducted on a site-specific basis and
is beyond the scope of this screening-level analysis.

       Based on this assessment, EPA determined that incidental discharges from study vessels
do not solely cause any NRWQC to be exceeded in the modeled hypothetical large estuaries and
harbors. This determination suggests that these discharges alone are unlikely to cause
impairments to relatively large water bodies. However, if a large water body already contains
select pollutants, then vessels that contribute significant quantities of these pollutants might
contribute to  such an NRWQC exceedance. Furthermore, as discussed in Chapter 3, many
pollutants detected in the vessel discharges were present at concentrations that exceed an
NRWQC at the end of pipe,  and therefore have the potential to negatively impact the receiving
water on a more localized scale. Because the screening model assumes instantaneous and
universal dilution in a large hypothetical harbor, the model is not designed to examine impacts
on a local scale, in small water bodies with many vessels, or in water bodies with little to no
flushing (i.e.  dilution). These discharges may cause environmental concerns in areas such as
small side embayments or marinas where flushing rates are low (see discussion in Section 4.6).
As discussed above, EPA further notes that this model does not take into account any loadings
from vessels that are not study vessels or other point/nonpoint sources that discharge pollutants
that contribute to the loadings in the water body.

       For the purpose of this study, EPA selected a simple screening-level model to provide a
coarse "big picture" assessment of the overall potential for discharges from study vessels to
cause or contribute to an impact on human health, welfare, or the environment. Although a
screening-level model has several limitations, it identifies any major water quality issues,
provides valuable information on pollutants of concern, identifies data gaps, and serves as a
starting point for any future site-specific studies that are beyond the scope and  objectives of this
study.

       The remainder of this chapter details EPA's cumulative effects assessment and is
organized as follows:

       •  Section 4.1: Model  Selection - Presents EPA's rationale for selecting the Fraction of
          Freshwater Screening-Level Model  for the analysis.

       •  Section 4.2: Fraction of Freshwater Model - Describes the "fraction of freshwater
          model" and presents the equations and input parameters required for the screening-
          level analysis.
                                           369

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment

       •  Section 4.3: Vessel Discharge Loading Rates - Describes the methodology for
          developing the input parameters required to calculate the total analyte-specific
          loading rates for each vessel population scenario.

       •  Section 4.4: Hypothetical Harbor - Describes the methodology for developing
          hypothetical harbor input parameters.

       •  Section 4.5: Model Scenarios - Presents the 24 model scenarios represented in the
          model.

       •  Section 4.6: Model Results  - Presents the results from the "fraction of freshwater
          model."

       •  Section 4.7: Conclusions - Presents EPA's conclusions on the potential for vessel
          discharges from study vessels to solely impact large-scale harbors or estuaries (e.g.,
          to solely pose  a risk to human health, welfare, and the environment).

4.1    MODEL SELECTION

       Study vessels discharge into coastal harbors throughout the United States. Estuarine
models, which are commonly used to assess harbor water quality, consist of two primary
components: hydrodynamics (i.e., water transport processes) and water quality. Estuarine models
are generally classified into the following four levels according to  the temporal and spatial
complexity of the hydrodynamic component of the model:

       •  Level I - Desktop screening models that calculate seasonal or annual mean
          concentrations based on steady-state conditions and simplified flushing time
          estimates.

       •  Level II - Computerized steady-state or tidally averaged quasi-dynamic simulation
          models, which generally use a box or compartment-type network.

       •  Level III - Computerized one-dimensional (i.e., estuary is well-mixed vertically and
          laterally) and quasi-two-dimensional (i.e., a link-node system describes estuary
          longitudinal and lateral mixing) dynamic simulation models.

       •  Level IV - Computerized two-dimensional (i.e., represents estuary longitudinal and
          lateral mixing) and three-dimensional (i.e., represents estuary longitudinal, lateral,
          and vertical mixing) dynamic simulation models (EPA  2001).

       The sheer number of different coastal harbor environments potentially impacted by these
vessels precludes using the more complex and data-intensive Level II, III, and IV models for the
cumulative impacts analysis. For these reasons, EPA selected a Level I screening-level model,
the "fraction of freshwater model,"  for the environmental assessment of vessel discharges from
study vessels.

                                           370

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
       In addition to coastal harbors, study vessels also discharge to freshwater environments
such as the Great Lakes and major river systems (e.g., Mississippi River). The "fraction of
freshwater model" is applicable to only estuarine or saltwater-influenced environments;
therefore, the modeling approach presented in this chapter does not address the potential
environmental impact of vessel discharges in completely freshwater environments. Additional
screening-level modeling approaches would be required to assess possible impacts of vessel
discharges in these environments. EPA assumes that discharges to freshwater systems represent a
smaller percentage of the total load from study vessels based on hailing port information
provided in the Marine Information for Safety and Law Enforcement (MISLE) database
maintained by the U.S. Coast Guard. Based on these data, commercial fishing vessels are almost
exclusively located along U.S. coastal waters, and only about a third of other nonrecreational
vessels less than 79 feet in length cite an inland waterway as their hailing port.
4.2    FRACTION OF FRESHWATER MODEL

       The "fraction of freshwater model" is a series of equations that represent the harbor
environment in zero dimensions and at a steady state (USEPA, 2001). These calculations are
zero-dimensional in that they estimate concentrations at a given point in a water body within a
specified, spatially homogenous volume. For example, the calculations assume instantaneous and
homogeneous mixing of vessel discharges within the defined volume of a given harbor. It does
not account for gradients of concentrations that would occur with distance from discharge
source(s) such as plumes from vessels and other sources2.  Specifying plumes and accounting for
locations of numerous discharge sources would require a two- or three-dimensional model,
which is beyond this Level I screening-level analysis.

       Steady state means that the calculations provide an instantaneous estimate of the
concentration under the assumption of chemical and physical equilibrium. Chemical equilibrium
means that the water body salinity and the vessel discharge analyte concentrations do  not change
over time, while physical equilibrium means that the volume of water in the water body, tides,
currents, and vessel discharge flow rates do not change over time.  The assumption is that every
process occurs instantaneously; therefore, temporal variability is not a factor. Accounting for
changes in tides, currents, river flow, vessel discharge flow rates, and discharge concentrations
over time would require a dynamic model, which is beyond this Level I screening-level model.
This aspect of the model may cause it to underestimate  localized environmental impacts,
 Discharge plumes can be highly structured, especially in low-flushing environments; therefore, the development of
a worst-case scenario using a screening-level model is not entirely conservative due to the assumptions of
instantaneous and homogenous mixing within the entire volume of the harbor. A true worst-case scenario would
likely include the concentration of pollutants within a small area of the harbor due to minimal dispersion of
discharge plumes across the harbor. It would also include background concentrations and take other pollutant
loadings into account (e.g., sewage treatment facilities, recreational vessels and other large vessels, stormwater,
agricultural runoff).

                                           371

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
especially in areas with inadequate flushing. However, in estimating quantities of pollutants
discharged from the various discharge types, EPA has tended to use conservative parameter
estimates (i.e., estimates that may overstate the average value) for variables such as flow and
pollutant concentration.

       The "fraction of freshwater model" calculates the analyte concentration in a harbor
resulting from vessel discharges using the following four steps:

       •   Step  1: Calculate vessel discharge analyte loading rates (Equations 4-1 and 4-2)
       •   Step 2: Calculate the fraction of freshwater in the harbor (Equation 4-3)
       •   Step 3: Calculate the harbor flushing time (Equation 4-4)
       •   Step 4: Calculate the harbor analyte concentration (Equation 4-5)

       The following subsections describe the input requirements, assumptions, and calculations
for each step in the "fraction of freshwater model."

   4.2.1   Step 1: Calculate Vessel Discharge Analyte Loading Rates

       Analyte-specific total discharge loading rates (We) are required as input values in the
"fraction of freshwater model" to calculate the instantaneous analyte concentrations in the harbor
(Cx). In this analysis, analyte loading rates were based on the following four input parameters:

       •   Average analyte concentrations for each vessel class discharge type;
       •   Estimated flow rate for each discharge type within a vessel class;
       •   Number of vessels per vessel class present in the harbor; and
       •   Percentage of vessels per vessel class discharging each discharge type in the harbor
           (Equation 4-1).

We,z = Z( Ce,y,z* Qy,z * Nz * Py,z)                                               Equation 4-1

       Where:
               WBiZ   =    Discharge loading rate for analyte e from vessel class z (mass/time)
               Ce,y,z  =    Average concentration of analyte e in discharge^ from vessel class z
                          (mass/volume)
               QyiZ   =    Flow rate for discharge y from vessel class z (volume/time)
               N,z   =    Number of vessels in vessel class z present in the harbor
               Py,z   =    Percentage of vessels in vessel class z discharging discharge^

       EPA calculated the  analyte-specific total discharge loading rate by summing the
discharge loading rates for that analyte from each vessel class (Equation  4-2). Section 4.3
describes EPA's methodology for calculating this loading rate in more detail.

We = Z( WSrZ)                                                               Equation 4-2

                                            372

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment

       Where:
               We    =   Total discharge loading rate for analyte e from study vessel
                         discharges (mass/time)
               We,z   =   Discharge loading rate for analyte e from vessel class z (mass/time)


    4.2.2   Step 2: Calculate the Fraction of Freshwater in the Harbor

       The "fraction of freshwater model" estimates analyte concentrations in one dimension
using information on freshwater inflow and by comparing salinity in the harbor with salinity in
the seawater at the mouth of the harbor (USEPA, 2001). The fraction of freshwater (fx) at any
location in the estuary is calculated as:

fx = (Ss - Sx)/Ss                                                                Equation 4-3

       Where:
              fx    =  Fraction of freshwater at location x in the model harbor (unit-less)
               Ss    =  Seaward boundary salinity at the mouth of model harbor  (PSU)
               Sx    =  Salinity at location x in model harbor (PSU)

       EPA states that this ratio (fx) ".. .can be viewed as the degree of dilution of the freshwater
inflow (as well as pollutants) by seawater" from tidal influx in the harbor (USEPA, 2001).

    4.2.3   Step 3: Calculate the Harbor Flushing Time

       Harbor flushing time is defined as the amount of time required to replace the freshwater
volume of the harbor by the river freshwater input. The flushing time (i) of the model  harbor is
calculated using Equation 4-4:

       t = (V * /x)/Qfw                                                        Equation 4-4

       Where:
               t     =  Model harbor flushing time
               V    =  Volume of model harbor
              fx    =  Fraction of freshwater at location x in model harbor (unit-less)
                    =  Inflow of freshwater to model harbor from the model river
                       (volume/time)
                                           373

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
   4.2.4  Step 4: Calculate the Harbor Analyte Concentration

       The concentration of an analyte at location x (Cx) is the analyte-specific total loading rate
(We in mass/time) divided by the flow rate away from location x, described by the volume of the
harbor (V) divided by the flushing time (t) (USEPA 2001):

Cx = Wel(Vlf)                                                             Equation 4-5

       Where:
               Cx   =  Instantaneous analyte concentration at location x in model harbor
                       (mass/volume)
               We   =  Analyte-specific loading rate (mass/time) as calculated under Step 1
               V    =  Volume of the model harbor as defined in Step 3
               t     =  Model harbor flushing time as calculated in Step 3

4.3    VESSEL DISCHARGE LOADING RATES

       Step 1 in the "fraction of freshwater model" calculates a range of analyte-specific total
loading rates (Wein mass/time) from fishing and nonrecreational vessels less than 79 feet based
on the analyte concentration in a given discharge, the estimated flow rate for a given discharge,
and assumptions on the number of vessels present in a harbor and percentage of vessels
discharging each discharge type in the harbor. The following subsections present EPA's
methodology for developing the modeling input parameters to calculate the analyte-specific total
discharge loading rate.

   4.3.1  Calculate the Average Analyte  Concentrations

       As described in Chapter 2, EPA  collected wastewater characterization data for nine
vessel discharges sampled from a total of 61 vessels (See Table 2.1). The objective of EPA's
sampling program was to provide information on the nature, type, and composition of discharges
from representative single  study vessels and study vessel classes. EPA calculated vessel-class-
specific analyte concentrations by averaging all of the discharge effluent sampling data by
discharge type and by analyte. Replicate samples from a single vessel were averaged together
prior to calculating a vessel-class-specific average. Certain analytes were not detected above the
sample reporting limit in some wastewater samples. To fully represent the variability of pollutant
concentrations in vessel discharges, EPA included both nondetected and detected results in
calculating average vessel-class-specific analyte concentrations. For nondetected results, EPA
assumed the analyte concentration was equal to one-half the sample reporting limit for that
analyte. EPA based this assumption on the expectation that the analyte was present in
wastewater, albeit at a concentration less than the sample reporting limit.
                                           374

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
    4.3.2  Discharge Flow Rate Assumptions

       EPA calculated discharge-specific flow rates for each of the 593 study vessels sampled
based on the following information for each discharge type:

    •   Known or estimated flow rates for the pump or mechanism controlling the discharge
    •   Assumptions on the frequency of discharge
    •   Assumptions on the duration of the discharge

       EPA estimated vessel-specific discharge flow rates based on data and field observations
from EPA's vessel sampling program, as well  as information from secondary data sources. EPA
developed frequency and duration assumptions based on interview responses from the vessel
crew or observations from EPA's vessel sampling team. For example, EPA reviewed interview
responses from a tow/salvage vessel operator to estimate bilge discharges based on the
observation that the bilge pump discharges 60  gallons per minute for an approximate duration of
five seconds per pump-out with an average frequency of one pump out every 10-minutes. As
another example, the frequency at which fishing vessels discharge fish hold water into a harbor is
generally dictated by how often the vessel offloads its catch. EPA used vessel sampling team
field observations to develop the discharge frequency for each fishing vessel subclass (Table
4.3.1).

       In addition, many of the study vessel classes discharge different amounts in different
seasons. For example, fishing vessels operate during certain times of the year to coincide with
different peak fishing seasons. As a conservative estimate, to account for the seasonal nature of
these discharge loadings, EPA developed vessel flows to represent the loading rate that would
typically occur during peak vessel activity for  each vessel class. Specifically, EPA calculated the
loading rates to represent the summer (or peak) season for all vessels, which is the time of
greatest fishing activity in the major harbors across the United States and is generally the peak of
recreational and tourist activity. 4
3 As previously discussed, EPA excluded the sampling data from the two recreational vessels in the model because
these vessels are not study vessels.
4 Vessel flow rates presented in the screening-level analysis are not intended to be used to estimate annual loads.
Additional seasonal considerations, such as the length of different fishing seasons, are required to calculate annual
loads, which is beyond the scope of the screening-level analysis.

                                            375

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
                Table 4.3.1. Offload Frequency by Fishing Vessel Subtype
Fishing Vessel Subclass
Purse Seiners
Trailers
Gillnetters
Tenders
Longliners
Shrimpers
Trawlers
Frequency of Offloads1
Daily
Daily
Daily
Once every 2 days
Once every 2 days
Once every 3 days
Once every 3 days
                (1) Based on sampling team observation in the field.


       Table 4.3.2 provides examples of the known or estimated field data parameters and
assumptions used to calculate the vessel-specific discharge flow rates for each discharge type.
Where data parameter information were unknown, EPA used information from a similar vessel
discharge type or used best professional judgment to estimate the required information.
Appendix G provides a detailed description of the data and assumptions used to calculate the
discharge-specific flows for each of these 59 sampled vessels. EPA averaged the vessel-specific
discharge flows presented in Appendix G by vessel class and discharge type to calculate the
vessel class-specific flow rates (Qy,z) used in the model (Table 4.3.3).
                                           376

-------
                                             Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment

Table 4.3.2. Examples of Field Data and Assumptions for Flow Rate Calculations by Discharge
Discharge
   Type
 Example Data Parameters
       Example Assumptions
              Example Discharge Flow Calculation
Bilgewater
- Flow rate of bilge pump
- Frequency of bilge pump
out
- Duration of a single pump
out
-12 volt bilge pump at 20 gpm1
- Discharged all year
- 5 min to pump bilge
- 2 pumpouts per day
- 5 min to pump bilge
-1 pump per week
- Discharged 365 days a year
-12 volt bilge pump at 20 gpm

20 gal per min X 5 min X 1 pump/7 days = 14.3 gal/day (0.05 nrVday)
Deck Wash
- Volume of water used
during deck wash down
- Frequency of deck washes
- Duration of deck washes
- Flow rate of garden hose or
high-pressure sprayers used
to wash decks
- Garden hose flow rate is 11.67 gpm2
- 1 wash every 2 weeks
- 15 minutes per deck wash
- Cleaned with hose
-15 minute per deck wash
- Garden hose flow rate is 11.67 gpm
-1 wash every 2 weeks

11.67 gal per min X 15 min X 1 wash/14 days = 7.21 gal/day
(0.03 nrVday)	
Fish Hold
- Volume of holding tanks
- Volume of fish
- Whether the tanks hold fish
in water or ice
- Amount of ice
- Frequency of offloads
- Length of fishing season
- Density offish is 0.9 kg/liter
- Holding tank is 70% shrimp, 30%
water3
- Ice tank holds 50% fish, 35% ice,
15% air4
- 5,000-gallon tank
-75% full at offload
- Holding tank is 70%shrimp, 30% water
-1 offload every 3 days

5000 gal X 30% X 3/4 full X loffload/3 days = 375 gal/day (1.42
nrVday)	
Fish Hold
Clean
- Frequency of tank
cleanings
- Length of fishing season
- Washed with garden hose
- 30-minute wash for tenders and purse
seiners
- 15-minute wash for all other fishing
vessels
- Wash done after each offload
- Garden hose flow rate is 11.67 gpm
- 15-minute hose down after each offload
-1 offload every 3 days
- Garden hose flow rate is 11.67 gpm

11.67 gal per min X 15 min X 1 wash/ 3 day = 33.66 gal/day
(0.13 nrVday)	
Graywater
- Number of crew onboard
- Types of gray water
generated
- Frequency of laundry
washed
- Frequency of showers
- Laundry - front-load washer uses 25
gal/load
- Laundry - standard washer uses 40
gal/load
- Shower -17.2 gal per shower5
- Shower - 0.8 showers per person per
day5
- Sink - 30 min of sink use per crew per
week
- Sink - 2.2 gal per min in standard sink
- 3 crew
-17.2 gal per shower
- 0.8 showers per person per day

3 crew X 17.2 gal per shower X 0.8 showers per person per day = 41.28
gal/day (0.16 m3/day)
                                                                   377

-------
                                              Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment

Table 4.3.2. Examples of Field Data and Assumptions for Flow Rate Calculations by Discharge
Discharge
   Type
Example Data Parameters
       Example Assumptions
              Example Discharge Flow Calculation
Generator
Engine
- Engine type
- Cooling system type
- Hours of use per year
- 2 gpm cooling flow for a standard
generator6
-17,000 hours over 15 years
- 2 gpm cooling flow

2 gal/minX 60 min/hrX 17000hrs/15 years/365 days = 372.6 gal/day
(1.41 m3/day)	
Propulsion
Engine
- Engine type
- Cooling system type
- Hours of use per year
- Number of engines
onboard
-1 gpm cooling water flow rate for
outboard engine
- 20 gpm cooling water flow rate for
inboard engine6
- Cummins inboard 380hp diesel engine
- 463 hours in last 2 years
- 20 gpm cooling water flow rate

20 gal per minX 231.5 hours/year = 761.1 gal/day (2.88 nrVday)
Shaft
Water
- Duration of boat operation
-10 mL/min constant drip (3.8 gal/day
drip)4
- operates 5 days/week
-10 mL/min constant drip (3.8 gal/day drip)

3.8 gal per day X 5 days/week = 2.71 gal/day (0.01 nrVday)
(1) Estimate based on commonly used 12-volt bilge pumps. Flow rates ranged from 5 gpm to 30 gpm via Google.
(2) EPA used http://www.uiweb.uidaho.edu/extension/lawn/Files/Garden Hose.htm to calculate the average flow rate of a garden hose (i.e., 11.67 gpm). EPA
calculated the flow rate as the average flow for all three sizes of standard garden hose (1/2, 5/8, and 3/4 inches in diameter), assuming a water pressure of 40 PSI
and a hose length of 100 feet.
(3) Based on data from one of the sampled vessels: 2,700 cubic feet per tank, 3 tanks (229,461.75 liters of tanks space), holds 325,000 Ibs of salmon (163,798
liters offish assuming density offish is 0.9 kg/L). 163,798 liters offish/229,461.75 liters of tanks space = 70% offish. Assume remaining is hold water.
(4) Based on sampling team observation in the field.
(5) WaterSense Showerhead Factoids, Draft Date 7/27/09.
                                                                    378

-------
Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment




    Table 4.3.3. Vessel Flow Rates
Vessel Class
Fire Boat
Fire Boat
Fire Boat
Fire Boat
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Research
Supply Boat
Tour Boat
Tour Boat
Tour Boat
Tour Boat
Tour Boat
Tow/Salvage
Tow/Salvage
Tow/Salvage
Tugboat
Tugboat
Tugboat
Tugboat
Vessel Subclass
NA
NA
NA
NA
Gillnetter
Gillnetter
Lobster Boat
Longliner
Longliner
Longliner
Purse Seiner
Purse Seiner
Purse Seiner
Purse Seiner
Shrimper
Shrimper
Shrimper
Shrimper
Tender Vessel
Tender Vessel
Trawler
Trawler
Trawler
Trailer
Trailer
Trailer
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Discharge
Deck Wash
Engine Effluent
Fire Main Effluent
Generator Effluent
Engine Effluent
Fish Hold Effluent
Fish Hold Effluent
Bilgewater
Fish Hold Effluent
Fish Hold
Cleaning Effluent
Engine Effluent
Fish Hold Effluent
Fish Hold
Cleaning Effluent
Generator Effluent
Bilgewater
Deck Wash
Fish Hold Effluent
Graywater
Fish Hold Effluent
Fish Hold
Cleaning Effluent
Deck Wash
Fish Hold Effluent
Fish Hold Clean
Deck Wash
Fish Hold Effluent
Fish Hold
Cleaning Effluent
Engine Effluent
Deck Wash
Bilgewater
Deck Wash
Engine Effluent
Fire Main Effluent
Generator Effluent
Bilgewater
Deck Wash
Engine Effluent
Deck Wash
Fire Main Effluent
Graywater
Shaft Water
Flow Discharged to
Harbor per Vessel
(m3/day) 1
0.0100
36.3
0.00 2
1.80
14.9
0.800
2.83
0.450
2.83
0.00 2
16.6
16.3
1.07
1.41
2.84
0.344
1.25
0.00 2
19.3
0.660
0.344
1.25
0.220
0.470
3.04
0.660
0.0900
0.0300
0.0400
0.140
42.2
0.00 2
3.82
1.39
0.0240
0.952
0.0978
0.00 2
0.478
0.0100
                                           379

-------
Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment

    Table 4.3.3. Vessel Flow Rates
Vessel Class
Water Taxi
Water Taxi
Water Taxi
Water Taxi
Water Taxi
Vessel Subclass
NA
NA
NA
NA
NA
Discharge
Bilgewater
Deck Wash
Engine Effluent
Generator Effluent
Graywater
Flow Discharged to
Harbor per Vessel
(m3/day) 1
0.130
0.0650
39.8
9.08
0.280
    NA - Not applicable.
    (1) EPA estimated discharge flow rates for each vessel class based on data and field
    observations from EPA's vessel sampling program, as well as information from secondary
    data sources. EPA assumes that discharges not listed for a given vessel class are either not
    generated by a given vessel class or are discharged outside of the hypothetical harbor area.
    (2) These waste streams are all discharged in the harbor; however, the relatively small
    volume and infrequency of the discharge results in an insignificant daily discharge volume.
                                            380

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment

    4.3.3   Number of Vessels Present in the Harbor

       The total number of vessels present in any given harbor and the distribution of vessels
among the different vessel classes operating in that harbor vary significantly across the United
States. The number and distribution of vessels among the different classes depend on factors
such as the regional economic base (e.g., fishing versus recreation), size of the city supporting
the harbor, and geographic location (e.g., Alaska versus Gulf of Mexico). To represent the
variety of vessel combinations potentially present in a harbor, EPA developed the following
three vessel population scenarios for the model:

       •   Scenario 1: Fishing Harbor - A harbor where fishing is the primary economic driver
           in the region, and fishing vessels represent the majority of vessels present in the
           harbor5.

       •   Scenario 2: Large Metropolitan Harbor - A harbor where there are nonrecreational
           study vessels associated with a large metropolitan city that would require a greater
           number of support vessels such as supply boats, tow/salvage vessels, and tugboats. In
           addition, EPA assumed that there would be a higher level  of vessel activity within the
           hypothetical harbor compared to the activity assumed for Scenarios 1 and 3. Note that
           this screening analysis does not include large non study vessels such as container
           ships, tankers, bulk carriers, or other larger vessels, which would be present in almost
           any large port6.

       •   Scenario 3: Recreational Harbor - A harbor where the primary economic driver is the
           tourist or recreation industry. Although recreational vessels are not study vessels,
           EPA assumed that a recreational  harbor would have a high concentration of
           nonrecreational support vessels such as tow/salvage, tour boats, and water taxis
           associated with the regional recreational and tourist industry. However, as noted
           previously, this analysis does not consider discharges from non study vessels and
           other sources.

       EPA used  data from the MISLE database maintained by the U.S. Coast Guard to develop
the number of vessels present in the hypothetical harbors for the three scenarios and the
distribution among the different vessel classes. The MISLE database includes a wide range of
information regarding vessel and facility characteristics, accidents, marine pollution incidents,
and other pertinent information tracked by the U.S. Coast Guard from investigation and
5 Charter fishing vessels are not modeled as part of this analysis. Charter fishing vessels are generally either
manufactured or used primarily for pleasure, or leased, rented, or chartered to a person for the pleasure of that
person. Many are not inspected by the US Coast Guard.  These vessels are exempted from NPDES permitting
requirements by the Clean Boating Act (P.L. 110-288). Other charter fishing vessels are inspected by the US Coast
Guard.  These inspected, non-recreational vessels are not exempted from NPDES by the Clean Boating Act, and are
study vessels only if they are less than 79 feet. As a general matter, therefore, EPA anticipates that a significant
portion of charter fishing vessels are not study vessels.
6 Due to time and resource constraints, EPA did not sample these large vessels for this study. Therefore, EPA did not
calculate loadings from these larger vessels for this screening analysis.

                                             381

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
inspection activity. While MISLE represents the most comprehensive national dataset currently
available, it may not capture the entire universe of study vessels that operate in U.S. waters (see
Chapter 1 of this report for further discussion about the vessel universe in this study and the
MISLE database).

       EPA identified and  compiled hailing port and vessel class distribution data on the top 20
hailing ports cited in the MISLE database. Based on the identified harbors, EPA selected
representative harbors for each vessel population scenario to develop the vessel distributions in
the model (see Table 4.3.4).

             Table 4.3.4. Vessel Population Scenario Representative Harbors
             Based on the Top 20 Hailing Ports Cited in the MISLE Database
Top 20 Hailing Ports
Cited in MISLE
Boston, MA
Cordova, AK
Gloucester, MA
Homer, AK
Houma, LA
Houston, TX
Juneau, AK
Ketchikan, AK
Key West, FL
Kodiak, AK
Miami, FL
New Orleans, LA
New York, NY
Norfolk, VA
Petersburg, AK
Portland, OR
San Diego, CA
San Francisco, CA
Seattle, WA
Sitka, AK
Vessel
Population
Scenario 1
Fishing
Harbor

X
X
X

X
X
X

X

X


X
X


X
X
Vessel
Population
Scenario 2
Large
Metropolitan
Harbor
X









X
X
X



X



Vessel
Population
Scenario 3
Recreational
Harbor




X
X


X

X
X
X
X

X
X
X
X

       For each representative harbor, EPA calculated the percentages of fishing vessels and
non-fishing study vessels reported in the MISLE database (see Table 4.3.5, Table 4.3.6, and
Table 4.3.7). EPA averaged the percentages of fishing and non-fishing vessels to develop the
overall proportion of these vessel types for each vessel population scenario.
                                           382

-------
Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment

          Table 4.3.5. Percentage of Study Vessels Present  in
          Representative Fishing Harbor
Hailing Port
New Orleans, LA
Seattle, WA
Houston, TX
Juneau, AK
Houma, LA
Cordova, AK
Homer, AK
Sitka, AK
Kodiak, AK
Portland, OR
Ketchikan, AK
Gloucester, MA
Petersburg, AK
Average
Percentage of Fishing
Vessels
26%
69%
56%
82%
39%
94%
82%
76%
91%
51%
62%
84%
93%
70%
Percentage of
Non-fishing Study
vessels
74%
31%
44%
18%
61%
6%
18%
24%
9%
49%
38%
16%
7%
30%
          Source: MISLE database.
          Table 4.3.6. Percentage of Study Vessels Present in
          Representative Large Metropolitan Harbor
Hailing Port
New Orleans, LA
New York, NY
Miami, FL
Boston, MA
San Diego, CA
Average
Percentage of Fishing
Vessels
26%
21%
43%
55%
37%
36%
Percentage of
Non-fishing Study
vessels
74%
79%
57%
45%
63%
64%
          Source: MISLE database.
                                      383

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
                Table 4.3.7. Percent of Study Vessels Present in
                Representative Recreational Harbor
Hailing Port
New Orleans, LA
Seattle, WA
New York, NY
Houston, TX
San Francisco, CA
Miami, FL
Norfolk, VA
Houma, LA
San Diego, CA
Portland, OR
Key West, FL
Average
Percent of Fishing
Vessels
26%
69%
21%
56%
64%
43%
28%
39%
37%
51%
47%
44%
Percent of
Non-fishing Study
vessels
74%
31%
79%
44%
36%
57%
72%
61%
63%
49%
53%
56%
                Source: MISLE database.

       EPA established the total number of vessels present in each vessel population scenario
based on:

       •  Field observations from EPA's vessel sampling program.

       •  Total vessel population data for the top 20 hailing ports as reported in the MISLE
          database.

       •  An assumption that the hypothetical harbor is representative of up to 10 miles of
          shoreline.

       •  An assumption that the vessel distributions reflect vessel populations during peak
          activity for each scenario (i.e., summer season during peak fishing, recreational, and
          tourist activity).

       Based on these assumptions, EPA selected  a total vessel population of 175 vessels for
Scenarios 1 and 3 and 300 vessels for Scenario 2 (see Table 4.3.8). Table 4.3.8 presents the
distribution of vessels among the different vessel classes for each vessel population scenario
developed using the vessel ratios discussed above,  assumptions on the total vessel population,
field observations, and best professional judgment.
                                            384

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
Table 4.3.8. Vessel Population Scenarios
Vessel Class
Fire Boat
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Research
Supply Boat
Tour Boat
Tow/Salvage
Tugboat
Water Taxi
Vessel Subclass
NA
Gillnetter
Lobster Boat
Longliner
Purse Seiner
Shrimper
Tender Vessel
Trawler
Trailer
NA
NA
NA
NA
NA
NA
Total Number of Vessels
Vessel Population
Scenario 1
Fishing Harbor
1
12
12
24
12
10
20
20
12
2
12
10
6
12
10
175 1
Vessel Population
Scenario 2
Metropolitan Harbor
5
10
10
16
10
8
10
16
10
10
55
20
40
60
20
300 2
Vessel Population
Scenario 3
Recreational Harbor
1
9
9
15
9
5
9
13
9
8
10
24
20
10
24
175 3
NA - Not applicable.
(1) Fishing harbor-percentage of fishing vessels is 70%, percentage of non-fishing vessels is 30%.
(2) Large metropolitan harbor - percentage of fishing vessels is 30%, percentage of non-fishing vessels is 70%.
(3) Recreational harbor - percentage of fishing vessels is 45%, percentage of non-fishing vessels is 55%.
    4.3.4   Percentage of Vessels Discharging in the Harbor

       In addition to the number of vessels present in the harbor, EPA also established the
percentage of vessels within each vessel class and discharge type that discharge into the harbor.
The purpose of this is to account for the fact that not all vessels within a vessel class discharge
all waste streams. EPA developed and selected the percentage of vessels discharging to the
harbor (see Table 4.3.9) based on interview responses and data collected during EPA's vessel
sampling program. EPA assumed all  sampled vessels generate all discharges unless otherwise
noted by the vessel operators as follows:

       •   Vessel does not have the system or process responsible for the discharge  (e.g., the
           vessel does not generate graywater as it does not have sinks, showers, or washing
           machines).
       •   System has no discharge (e.g., vessel propulsion and generator engines are keel-
           cooled).
       •   Vessel typically discharges outside U.S. waters (e.g., fishing vessel washes decks
           after each catch at fishing grounds greater than 12 nautical miles from shore).

       Based on these criteria, EPA calculated the percentage of vessels (Py,z) in each vessel
class that discharge each discharge type into the harbor using the following equation:
                                            385

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
Py,z= Sample Ny,z/ Sample Nz

       Where:
Equation 4-6
               PV:z          = Percentage of vessels in vessel class z discharging discharge^
               Sample Ny,z   = Number of vessels in vessel class z discharging discharge^
                              from EPA's vessel sampling program
               Sample Nz    = Number of vessels from vessel class z from EPA's vessel
                              sampling program

       Appendix G includes the field data and assumptions used to develop the percentage of
vessels input parameter (Py:Z) for each vessel class and discharge stream.
           Table 4.3.9. Percentage of Vessels Discharging in the Harbor
Vessel Class
Fire Boat
Fire Boat
Fire Boat
Fire Boat
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Vessel Subclass
NA
NA
NA
NA
Gillnetter
Gillnetter
Lobster Boat
Longliner
Longliner
Longliner
Purse Seiner
Purse Seiner
Purse Seiner
Purse Seiner
Shrimper
Shrimper
Shrimper
Shrimper
Tender Vessel
Tender Vessel
Trawler
Trawler
Trawler
Trailer
Trailer
Trailer
Discharge
Deck Wash
Engine Effluent
Fire Main Effluent
Generator Effluent
Engine Effluent
Fish Hold Effluent
Fish Hold Effluent
Bilgewater
Fish Hold Effluent
Fish Hold
Cleaning Effluent
Engine Effluent
Fish Hold Effluent
Fish Hold
Cleaning Effluent
Generator Effluent
Bilgewater
Deck Wash
Fish Hold Effluent
Graywater
Fish Hold Effluent
Fish Hold
Cleaning Effluent
Deck Wash
Fish Hold Effluent
Fish Hold Clean
Effluent
Deck Wash
Fish Hold Effluent
Fish Hold
Cleaning Effluent
Percentage of
Vessels
Discharging Flow
in Harbor 1
100%
100%
100%
100%
80%
80%
100%
33%
100%
100%
40%
100%
100%
40%
50%
80%
80%
100%
100%
67%
80%
80%
40%
17%
100%
33%
                                           386

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
            Table 4.3.9. Percentage of Vessels Discharging in the Harbor
Vessel Class
Research
Supply Boat
Tour Boat
Tour Boat
Tour Boat
Tour Boat
Tour Boat
Tow/Salvage
Tow/Salvage
Tow/Salvage
Tugboat
Tugboat
Tugboat
Tugboat
Water Taxi
Water Taxi
Water Taxi
Water Taxi
Water Taxi
Vessel Subclass
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Discharge
Engine Effluent
Deck Wash
Bilgewater
Deck Wash
Engine Effluent
Fire Main Effluent
Generator Effluent
Bilgewater
Deck Wash
Engine Effluent
Deck Wash
Fire Main Effluent
Graywater
Shaft Water
Bilgewater
Deck Wash
Engine Effluent
Generator Effluent
Graywater
Percentage of
Vessels
Discharging Flow
in Harbor 1
100%
100%
67%
67%
100%
100%
67%
33%
100%
83%
100%
100%
67%
89%
75%
100%
100%
25%
25%
           NA - Not applicable.
           (1) The percentages of vessels discharging to the harbor were determined based on
           field observations of sampled vessels. As a conservative estimate, it was assumed that
           100% of vessels in sampled vessel classes with no information available discharge in
           the harbor.
    4.3.5   Vessel Discharge Loading Rates

       EPA calculated the vessel class-specific loading rates for each analyte (WB:Z) using
Equation 4-1 for each of the three vessel population scenarios described in Section 4.3.3. EPA
then calculated the total analyte-specific load rates (We) for each vessel population scenario using
Equation 4-2. Appendix G presents the total analyte-specific loading rates for each of the three
vessel population scenarios represented in the model (i.e., fishing harbor, large metropolitan
harbor, and recreational harbor).

    4.3.6   Dissolved Copper Loading Rates from Antifouling Paints

       In addition to the loading rates calculated based on EPA's vessel sampling program data,
EPA also considered the additional dissolved copper load to receiving waters associated with
antifouling paints used on vessel hulls. As described in Chapter 3, antifouling systems (AFSs)
are designed to release biocide over time to retard growth and maintain a smooth underwater
surface (Schiff et al., 2003). Copper oxide is the most common biocide added to AFSs to prevent
                                            387

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
biofouling organisms from attaching to the hull. Numerous studies have investigated the leaching
rate of copper from both passive leaching and underwater hull cleaning (Thomas et al., 1999;
Zirino and Seligman, 2002; Valkirs et al., 2003; Schiff et al., 2003). Based on estimates
produced in these studies, EPA selected a dissolved copper leaching rate of 8.2 |ig/cm2/day to
estimate the additional dissolved copper load to the harbor from vessel AFSs. EPA estimated the
average vessel length for each vessel class based on information available in the MISLE database
and field observations from EPA's vessel sampling program (Table 4.3.10). EPA assumed that
the beam of the vessel beam (i.e., width) was equal to approximately one-third its length and
used Equation 4-7 (Interlux, 1999) to estimate the hull surface area for each vessel class:
   = LZ*(LZ/3)*0.85
Equation 4-7
       Where:
               A2   =   Hull surface area for individual vessels in vessel class z (area)
               Lz   =   Average length of vessels in vessel class z (distance)
                     Table 4.3.10. Estimated Average Vessel Length
                     by Vessel Class
Vessel Class
Fire Boat
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Research
Supply Boat
Tour Boat
Tow/Salvage
Tugboat
Water Taxi
Vessel Subclass
NA
Gillnetter
Lobster Boat
Longliner
Purse Seiner
Shrimper
Tender Vessel
Trawler
Trailer
NA
NA
NA
NA
NA
NA
Vessel Length
(feet) 1
50
35
35
35
50
50
100
50
35
40
50
50
40
79
79
                     NA - Not applicable.
                     (1) - EPA estimated the average vessel length for each vessel
                     class based on information available in the MISLE database
                     and field observations during EPA's vessel sampling
                     program.

       EPA calculated the dissolved copper loading rate from AFSs for each vessel population
scenario using Equation 4-8, and then added these loadings to the dissolved copper loading rates
calculated in Section 4.3.5 for the other vessel discharges to determine the total dissolved copper
                                            388

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
load introduced into the harbor for each loading scenario7. EPA calculated that AFSs contribute
approximately 2.79 Ibs/day of dissolved copper under Vessel Population Scenario 1 (fishing
harbor), 4.86 Ibs/day under Vessel Population Scenario 2 (large metropolitan harbor), and 2.63
Ibs/day under Vessel Population Scenario 3 (recreational harbor8). Appendix G presents the total
dissolved copper loading rates represented in the model.
AFC Wcopper = I Nz * Az * 8.2 ng/cm2/day

       Where:
Equation 4-8
               AFS Wcopper  =    AFS discharge loading rate for dissolved copper (mass/time)
               Nz          =    Number of vessels in vessel class z present in the harbor
               Az          =    Hull surface area for individual vessels in vessel class z (area)

4.4    HYPOTHETICAL HARBOR

       Given the wide variety of coastal harbor environments potentially impacted by study
vessel discharges, EPA developed several hypothetical harbors for the vessel discharge
environmental assessment to represent a range of environmental conditions that could potentially
be impacted. To develop input values that represented realistic environmental conditions, EPA
identified and collected environmental data on eight harbors (Table 4.4.1) that represented a
geographically and environmentally diverse group of water bodies,  had the potential for a high
density of study vessels, and received freshwater inflow from a major river system.

     Table 4.4.1. Harbors Selected for Model Input Parameter Development
Harbor Name
Cohasset Harbor
Dorchester Bay
Auke Bay
Biscayne Bay
Mobile Bay
Yaquina Bay
Craford Bay
Eastern Channel
City Name
Boston
Boston
Juneau
Miami
Mobile
Newport
Norfolk
Sitka
State
Massachusetts
Massachusetts
Alaska
Florida
Alabama
Oregon
Virginia
Alaska
River Name
Gulf River
Neponset River
Mendenhall River
Miami River
Tensaw, Blakeley, and Mobile River
Yaquina River
Eastern and Southern Branch Elizabeth River
Indian River
       The "fraction of freshwater model" requires the following four input parameters to define
the water body characteristics:
7 Note that some hull cleaning methods can release a plume of antifouling paint, which contains copper in paniculate
form, in the water. The paniculate copper can settle into the sediments and over time reenter the water body in the
dissolved form. EPA did not include the potential dissolved copper load from paniculate copper resulting from hull
cleaning.
 1 As noted above, these loading rates do not include the loading from nonstudy vessels.
                                            389

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
       •  Seaward boundary salinity at the mouth of the harbor (Ss)
       •  Salinity at location x in the harbor (Sx)
       •  Volume of the harbor (V)
       •  Inflow of freshwater to the harbor
       EPA collected data on the four input parameters for the harbors listed in Table 4.4. 1 and
calculated a flushing time using Equation 4-4 in Section 4.2.3. Appendix G presents the
environmental data identified by EPA for each harbor listed in Table 4.4.1. EPA selected the
input parameters for the hypothetical harbors' salinity, volume, and river flow based on the
environmental data collected for the harbors with the minimum and maximum flushing times
(Table 4.4.2). EPA assumed an average ocean salinity of 35 PSU for the salinity at the seaward
boundary of the hypothetical harbor.

                   Table 4.4.2. Hypothetical Harbor Input Parameters
Model Parameter
Harbor Salinity (Sx~) Minimum
Harbor Salinity (Sx) Maximum
Ocean Salinity (Ss)
Harbor Volume (V) Minimum
Harbor Volume (V) Maximum
River Flow (Q^) Minimum
River Flow (Q^) Maximum
Model Input
Value
26.1
31
35
3,090,000
38,500,000
352,000
2,900,000
Units
PSU
PSU
PSU
m3
m3
nrVday
m3/day
       Using the input parameters in Table 4.4.2, EPA developed eight hypothetical harbors for
the vessel discharge environmental assessment (see Table 4.4.3). For each harbor scenario, EPA
calculated the fraction of freshwater (fx) and flushing time (i) using Equations 4-3 and 4-4 in
Sections 4.2.2 and 4.2.3, respectively. Flushing times for the hypothetical harbors ranged from
less than a day (0.122 days or 2.9 hours) to 27.8 days.
                                           390

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
Table 4.4.3. Hypothetical Harbor Scenarios
Hypothetical
Harbor Scenarios
Harbor Scenario 1
Harbor Scenario 2
Harbor Scenario 3
Harbor Scenario 4
Harbor Scenario 5
Harbor Scenario 6
Harbor Scenario 7
Harbor Scenario 8
Harbor
Salinity (Sx)
26.1PSU
SxMin
26.1PSU
SxMin
26.1PSU
SxMin
26.1PSU
SxMin
31PSU
SxMax
31PSU
SxMax
31PSU
SxMax
31PSU
SxMax
Ocean
Salinity (Ss)
35PSU
35PSU
35PSU
35PSU
35PSU
35PSU
35PSU
35PSU
Harbor
Volume (V)
3,090,000 m3
VMin
3,090,000 m3
VMin
38,500,000 m3
VMax
38,500,000 m3
VMax
3,090,000 m3
VMin
3,090,000 m3
VMin
38,500,000 m3
VMax
38,500,000 m3
VMax
River Flow ((?>)
352,000 m3/day
QfwMin
2,900,000 nrVday
QfwMax
352,000 nrVday
QfwMin
2,900,000 nrVday
QfwMax
352,000 nrVday
QfwMin
2,900,000 nrVday
QfwMax
352,000 nrVday
QfwMin
2,900,000 nrVday
QfwMax
/*
0.254
0.254
0.254
0.254
0.114
0.114
0.114
0.114
Flushing
Time
(Days)
2.23
0.271
27.8
3.38
1
0.122
12.5
1.52
4.5    MODEL SCENARIOS
       EPA developed a total of 24 model scenarios (see Table 4.5.1) for the screening-level
analysis based on the three vessel population scenarios and the eight hypothetical harbors
discussed in Sections 4.3.3 and 4.4, respectively. EPA calculated the estimated harbor dilution
for each model scenario using the following equation:
Dx=(Vli)ITj(Qy.z*Ny.Z*Py.Z)
                                                 Equation 4-9
          Where:
               Dx
               V
               t
               pv
Harbor dilution at location x
Volume of model harbor
Model harbor flushing time
Flow rate for discharge j/ from vessel class z (volume/time)
Number of vessels in vessel class z discharging discharge y
Percent of vessels in vessel class z discharging discharge^
                                           391

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
Table 4.5.1. Fraction of Freshwater Model Scenarios
Model
Scenario
Model Scenario 1
Model Scenario 2
Model Scenario 3
Model Scenario 4
Model Scenario 5
Model Scenario 6
Model Scenario 7
Model Scenario 8
Model Scenario 9
Model Scenario 10
Model Scenario 1 1
Model Scenario 12
Model Scenario 13
Model Scenario 14
Model Scenario 15
Model Scenario 16
Model Scenario 17
Model Scenario 18
Model Scenario 19
Model Scenario 20
Model Scenario 21
Model Scenario 22
Model Scenario 23
Model Scenario 24
Total Loading Rate (W,)
Scenario
Vessels Population Scenario 1 Fishing Harbor
Vessels Population Scenario 1 Fishing Harbor
Vessels Population Scenario 1 Fishing Harbor
Vessels Population Scenario 1 Fishing Harbor
Vessels Population Scenario 1 Fishing Harbor
Vessels Population Scenario 1 Fishing Harbor
Vessels Population Scenario 1 Fishing Harbor
Vessels Population Scenario 1 Fishing Harbor
Vessels Population Scenario 2 Metropolitan Harbor
Vessels Population Scenario 2 Metropolitan Harbor
Vessels Population Scenario 2 Metropolitan Harbor
Vessels Population Scenario 2 Metropolitan Harbor
Vessels Population Scenario 2 Metropolitan Harbor
Vessels Population Scenario 2 Metropolitan Harbor
Vessels Population Scenario 2 Metropolitan Harbor
Vessels Population Scenario 2 Metropolitan Harbor
Vessels Population Scenario 3 Recreational Harbor
Vessels Population Scenario 3 Recreational Harbor
Vessels Population Scenario 3 Recreational Harbor
Vessels Population Scenario 3 Recreational Harbor
Vessels Population Scenario 3 Recreational Harbor
Vessels Population Scenario 3 Recreational Harbor
Vessels Population Scenario 3 Recreational Harbor
Vessels Population Scenario 3 Recreational Harbor
Hypothetical
Harbor
Scenario
Harbor Scenario 1
Harbor Scenario 2
Harbor Scenario 3
Harbor Scenario 4
Harbor Scenario 5
Harbor Scenario 6
Harbor Scenario 7
Harbor Scenario 8
Harbor Scenario 1
Harbor Scenario 2
Harbor Scenario 3
Harbor Scenario 4
Harbor Scenario 5
Harbor Scenario 6
Harbor Scenario 7
Harbor Scenario 8
Harbor Scenario 1
Harbor Scenario 2
Harbor Scenario 3
Harbor Scenario 4
Harbor Scenario 5
Harbor Scenario 6
Harbor Scenario 7
Harbor Scenario 8
Dilution (Dx)
705
5,810
705
5,810
1,570
12,900
1,570
12,900
506
4,170
506
4,170
1,130
9,280
1,130
9,280
494
4,070
494
4,070
1,100
9,050
1,100
9,050
       As shown in Table 4.5.1, there are duplicate dilution factor values for different model
scenarios (e.g., Model Scenarios 1 and 3 both have a dilution factor of 705). Hence, there are
effectively 12 unique model scenarios and not 24 presented in this screening-level analysis. The
duplicate dilution factors are an artifact of EPA's decision to calculate dilution factors and
instantaneous harbor concentrations using all combinations of the input parameters in Table
4.4.2. In calculating the dilution factor, the volume of the harbor (V) cancels out of the dilution
equation (Equation 4-9) and is not a consideration (see below).
          Where:
               (Vlt)
                          * P
                            r
                             y,z
(V/(V*fx/Qfw)
Total discharge flow from all vessels
                                            392

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
       EPA used three total discharge flows (I*(Qy,z * Ny,z* Py>z) (i.e., vessel flows in a fishing
harbor, large metropolitan harbor, and recreational harbor) and four different volume-to-
flushing-time (V/t) ratios (i.e., assumed two fx values in the model and two Qfw values) in the
model. Section 4.6 discusses the results from the 12 unique model scenarios and presents the
results of the duplicate scenarios as one result (i.e., harbor concentrations from Model Scenarios
1 and 3).

4.6    MODEL RESULTS

       EPA calculated the instantaneous concentration (Cx) in the hypothetical harbor using
Equation 4-5 presented in Section 4.2.4 for each of the 12 model scenarios defined in Table
4.5.1. Appendix G presents the concentrations for all model scenarios for each vessel population
scenario. EPA compared the instantaneous concentrations in the hypothetical harbor with the
NRWQC to evaluate the potential for the cumulative effect of study vessel incidental discharges
to impact aquatic life or human health. EPA determined that none of the modeled concentrations
in the hypothetical harbor for the 12 scenarios exceeded an aquatic life or human health
NRWQC.

   4.6.1  Dilution Factor Analysis

       The model scenario dilutions factors calculated for the 12 unique scenarios ranged from
494 to 12,900. EPA performed a sensitivity analysis to determine the dilution factor at which
point NRWQC would be exceeded. EPA calculated the "tipping point" dilution in the
hypothetical harbor where the instantaneous concentration in the harbor would equal the most
stringent NRWQC for aquatic life or human health using the three vessel population scenario
loading rates discussed in Section 4.3.5. Table 4.6.1 presents the tipping point dilution factors for
the top 10 analytes with the highest dilution factor requirements to avoid exceeding an NRWQC.
Based on the results of the dilution analysis, a harbor dilution  factor of greater than 358  is
required to avoid exceeding any NRWQC for aquatic life or human health, which is below the
range of calculated model scenario dilution factors (i.e., 494 to 12,900). This sensitivity analysis
also demonstrates that dissolved copper and total arsenic represent the most significant
environmental risk from study vessels incidental discharges. These two analytes have relatively
stringent range of dilution requirements depending on the vessel population scenario selected to
avoid exceeding a NRWQC (i.e., dilution factors of greater than 144 to 266 for dissolved copper
and 284 to 358 for total arsenic)  and represent the highest dilution requirements for all the
analytes detected in vessel discharges. Following dissolved copper, the required  dilution factors
drop off significantly with a dilution  of greater than 33.7 required to avoid exceeding all other
NRWQC with most of the remaining dilution factors below one.
                                           393

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
       Table 4.6.1. "Tipping Point" Dilution Factors for Harbor Instantaneous
       Concentration to Equal the NRQWC Based on Vessel Population Scenario
       Loading Rates l
Class
Metals
Metals
Metals
Classicals
Metals
Classicals
Metals
voc
Metals
Analyte
Arsenic, Total 2
Copper, Dissolved
Arsenic, Dissolved 2
Total Residual Chlorine
Aluminum, Total
Sulfide
Selenium, Total 2
Benzene
Manganese, Total
Vessel
Scenario 1
Fishing Harbor
Dilution (Dx)
358
214
31.4
12.4
6.77
1.75
1.13
0.756
0.684
Vessel
Scenario 2
Metropolitan
Harbor
Dilution (Dx)
331
266
33.7
16.2
5.15
2.36
1.46
1.57
0.983
Vessel
Scenario 3
Recreational
Harbor
Dilution (Dx)
284
144
29.6
12.2
4.83
1.65
1.52
1.34
1.04
       (1) Table includes only those analytes that required a dilution factor of greater than one to avoid
       exceeding a NRWQC.
       (2) EPA suspects a limited number of the samples analyzed for selenium (and even fewer for
       arsenic) for bilgewater, packing gland effluent, propulsion engine effluent, graywater and deck
       washdown water may have elevated measured concentrations due to positive interference. Despite
       these limited instances of interference, EPA believes the fish hold concentrations reasonably
       represent true effluent concentrations for the discharge (see discussion in Sections 3.1.3 and 3.2.4.1
       for further information). EPA considered these interferences when interpreting the potential for
       vessel discharges to pose a risk to human health, aquatic life, or the environment and determined
       that such cationic interference does not influence the major findings presented in the modeling
       analysis.

    4.6.2  Supplemental Model Run in Response to Comments

       In response to public comments submitted for the draft version of this report, EPA
performed a supplemental model run using revised values based on information submitted by
commenters to assess the impacts of these alternative values on the model results. EPA adjusted
the model  assumptions presented in Table 4.6.2 and recalculated the associated discharge flows
and loads.  EPA observed no significant change in model results based on the revised  values.
Table 4.6.3 presents the revised "tipping point" dilution factors for the supplemental model run.

Table 4.6.2. Revised Model Assumptions
Vessel
Class
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Fishing
Tour Boat
Vessel
Subclass
Gillnetter
Longliner
Toller
Toller
Toller
Toller
Toller
Shrimping
NA
Discharge
Fish Hold
Fish Hold
Fish Hold Clean
Fish Hold
Fish Hold
Fish Hold
Deck Wash
Bilge Water
Bilge Water
Old Assumption
Offloads daily
Offloads once per two days
Offloads daily
Offloads daily
840 ft3 fish hold
5.5 tons of ice per offload
125 gallons per deck wash
150 gallons per minute bilge
pump rate
14.3 gallons per day
New Assumption
Offloads once per five days
Offloads once per five days
Offloads once per seven days
Offloads once per seven days
595 ft3 fish hold
2 tons of ice per offload
50 gallons per deck wash
20 gallons per minute bilge
pump rate
5 gallons per day
                                              394

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment


      Table 4.6.3. Supplemental Model Run "Tipping Point" Dilution Factors for
      Harbor Instantaneous Concentration to Equal the NRQWC Based on Vessel
      Population Scenario Loading Rates l
Class
Metals
Metals
Metals
Classicals
Metals
Classicals
Metals
voc
Metals
Analyte
Arsenic, Total 2
Copper, Dissolved 3
Arsenic, Dissolved 2
Total Residual Chlorine
Aluminum, Total
Sulfide
Selenium, Total 2
Benzene3
Manganese, Total 3
Vessel
Scenario 1
Fishing Harbor
Dilution (DJ
349
225
31.1
12.4
6.77
1.68
1.03
0.790
0.696
Vessel
Scenario 2
Metropolitan
Harbor
Dilution (Dx)
325
273
33.6
16.2
5.10
2.33
1.42
1.61
0.997
Vessel
Scenario 3
Recreational
Harbor
Dilution (Dx)
279
147
29.5
12.1
4.79
1.60
1.50
1.37
1.05
       (1) Table includes only those analytes that required a dilution factor of greater than one to avoid
       exceeding a NRWQC.
       (2) EPA suspects a limited number of the samples analyzed for selenium (and even fewer for
       arsenic) for bilgewater, packing gland effluent, propulsion engine effluent, graywater and deck
       washdown water may have elevated measured concentrations due to positive interference.  Despite
       these limited instances of interference, EPA believes the fish hold concentrations reasonably
       represent true effluent concentrations for the discharge (see discussion in Sections 3.1.3 and 3.2.4.1
       for further information). EPA considered these interferences when interpreting the potential for
       vessel discharges to pose a risk to human health, aquatic life, or the environment and determined
       that such cationic interference does not influence the major findings presented in the modeling
       analysis.
       (3) The revised model assumptions (see Table 4.6.2) did not significantly impact the total loads for
       this analyte; however, these assumptions lowered the total discharge volume from these vessels.
       Therefore, the dilution factors for the supplemental model run for this analyte are higher than the
       original model run due to the same mass loading rate being divided by a smaller total discharge
       flow.
    4.6.3  Loading Rate Analysis

       EPA compared the three analyte-specific loading rates used in the model with other
known loading rates to provide perspective on their magnitude and on their relative contribution
to the possible impairment of receiving waters (see Table 4.6.2 and Table 4.6.3). EPA selected
the following loading sources for comparison:

       •  Loads From Publicly Owned Treatment Works (POTW)
       •  Dissolved copper loads discharged to the Shelter Island Yacht Basin
       •  Estimated metal loading rates from urban stormwater

       EPA generated estimates for hypothetical medium-sized sewage treatment facilities with
a discharge rate of 10 million gallons per day (MOD).  These estimates were derived from the
                                              395

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
National Research Council's 1993 report "Managing Wastewater in Urban Areas". EPA
calculated loadings by multiplying an effluent volume of 10 MOD times the low and high
effluent concentrations for selected parameters using four types of wastewater treatment
(chemically-enhanced primary plus biological treatment, primary or chemically enhanced
primary plus nutrient removal, primary or chemically enhanced primary plus nutrient removal
plus gravity filtration, or primary or chemically enhanced primary plus nutrient removal  plus
high lime plus filtration)9. Values presented in Table 4.6.2 present the lowest and highest
derived loadings for these medium systems. EPA determined that the nutrient loads from the
175 to 300 study vessels were comparable to the low end estimates for Ammonia as Nitrogen
and total phosphorus, but notably  lower than those from the high end treated effluent estimates
from sewage treatment facilities.  As noted above, the model nutrient loadings from study vessels
do not include sewage discharges  (which is likely a source of nutrients from these vessels)10,
whereas these data are from POTW effluent, which has a significant sewage component. Table
4.6.2 shows that a medium sewage treatment facility discharges a higher volume of metals than
these 175 to 300 study vessels.  Finally, these study vessels discharge comparable levels of
BOD; though sewage treatment facilities are discharging a larger volume of effluent, they
remove significant quantities of BOD from the effluent. On the other hand, study vessels'
incidental discharges are untreated waste, some of which has notably high BOD concentrations
(e.g., fish hold effluent).

       EPA also obtained nutrient loading estimates from a sewage treatment facility with
advanced nutrient removal capabilities to provide real world example nutrient loadings that may
be associated with POTW discharges (Albert, 2007). This facility discharges approximately 40 to
50 MGD. EPA determined that the nutrient loads (i.e., ammonia as nitrogen, nitrate/nitrite as
nitrogen, total Kjeldahl nitrogen, and total phosphorus) from the 175 to 300 study vessels used to
establish the vessel loads in the screening-level analysis were notably lower than the nutrient
loads from this sewage treatment facility. It is important to note that these model nutrient loads
do not include nutrient contributions from vessel sewage discharges (possibly a significant
source of nutrients), as sewage discharges are excluded from the scope of P.L.  110-299.
9 A number of systems exist which are both smaller and larger than 10 MGD; for example, the Blue Plains POTW in
Washington DC is the largest advanced wastewater treatment system in the word and discharges an average of
approximately 330 MGD.  The wastewater treatment facilities in nearby Arlington County discharge less than 40
MGD. In comparison, the sewage treatment facility in Sitka, Alaska is designed to discharge only 1.8 MGD.
10 Sewage from vessels within the meaning of CWA section 312, which includes graywater in the case of
commercial vessels operating on the Great Lakes, is exempt from the CWA definition of "pollutant". 33 U.S.C.
1362(6); 33 U.S.C. 1322(a)(6). As a result, vessel sewage discharges are not subject to NPDES permitting. Instead,
Congress enacted a separate non-permitting scheme - CWA section 312 - to regulate the discharge of sewage from
vessels.
Under section 312 of the CWA, all vessels equipped with installed toilet facilities must also be equipped with an
operable U.S. Coast Guard-certified marine sanitation device (MSD). 33 U.S.C. 1322(h). The provisions of section
312 are implemented jointly by EPA and the Coast Guard: EPA sets performance standards for MSDs, and the Coast
Guard is responsible for developing regulations governing the design, construction, certification, installation and
operation of MSDs, consistent with EPA's standards. 33 U. S.C. 1322(b). Current performance standards which
apply to MSDs have standards for solids and fecal coliform. Generally speaking, most MSDs currently installed on
study vessels are not designed to remove nutrients from sewage.
                                             396

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
Therefore, these estimates are not a complete representation of vessel nutrient loadings; rather,
they are merely an estimate of nutrient loadings from incidental discharges.

       As described in Chapter, 3 dissolved copper concentrations resulting from study vessels'
incidental discharges potentially pose a risk to aquatic life. A significant contribution of the
dissolved copper load is from copper leaching from antifouling coatings on vessel hulls. In 2005,
the California Regional Water Quality Control Board examined the dissolved copper loads to
Shelter Island Yacht Basin from recreational vessel antifouling hull coatings and other source
loads in support of a Total Maximum Daily Load (TMDL) analysis for the impaired water. EPA
compared the dissolved copper loads from Shelter Island Yacht Basin TMDL to the vessel
population scenario loading rates (Table 4.6.2). EPA determined that the estimated dissolved
copper loads from 175 to 300 study vessels used in the model (i.e., 2.75 to 4.97 Ib/day) were
consistent with the combined dissolved copper loads from passive leaching and hull cleaning
from 2,363 recreational vessels present in Shelter Island Yacht Basin (i.e., 12.7 Ib/day). EPA also
compared the model dissolved copper loads to the combined estimated  contributions from urban
runoff, background, and atmospheric deposition in Shelter Island Yacht Basin (i.e., 0.381
Ib/day). The model dissolved copper loads from hull leaching and other discharge streams were
significantly larger than the other  source contributions present in Shelter Island Yacht Basin,
suggesting that dissolved copper from study vessels incidental discharges can represent a
significant portion of the dissolved copper load in a water body.

       EPA also estimated metal loading rates for urban stormwater runoff based on reported
loading rates from a 2001 literature study by Davis et al.  and an assumed watershed area of
approximately 17 square  miles (watershed area determined from readily available information on
watersheds' drainage areas for the water bodies discussed in Table 4.4.1). As shown in Table
4.6.2, EPA determined that urban  stormwater likely represents a greater load of total copper,
total lead, zinc, and cadmium to receiving waters than discharges from  175 to 300 study vessels.
However, the model results indicate that dissolved copper loads from study vessels are
significant.
                                           397

-------
                                                  Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
Table 4.6.4. Comparison of Model Loading Rates with Other Potential Point Source Loading Rates
Analyte
Ammonia as
Nitrogen
(NH3-N)
Biochemical
Oxygen
Demand (BOD)
Nitrate/Nitrite
(N03+N02-
N)
Total
Phosphorus
Total Kjeldahl
Nitrogen
(TKN)
Arsenic, Total
Cadmium,
Total
Copper,
Dissolved
Copper, Total
Lead, Total
Zinc, Total
Model Loading Rates from
Vessel Population Scenarios 1
Fishing
Harbor
(Ib/day)
8.52
635
0.127
13.8
97.8
0.0279
0.000749
2.88
0.158
0.0108
0.758
Large
Metropolitan
Harbor
(Ib/day)
6.07
481
0.203
8.91
68.5
0.0359
0.000657
4.97
0.179
0.0154
0.613
Recreational
Harbor
(Ib/day)
5.07
392
0.102
7.74
59.0
0.0315
0.000551
2.75
0.165
0.0142
0.516
POTW
Loading
Rates
10
mg/day2
(Ib/day)
8.35-41.7
250.4-
751.1
NA
8.35-125.2
NA
0.117-1.17
0.117-
0.609
NA
1.25-4.17
1.50-4.01
3.34-9.35
POTW
Loading
Rates
-40
mg/day3
(Ib/day)
36.2
NA
1,320
22.0
285
NA
NA
NA
NA
NA
NA
Shelter Island Yacht Basin Loading Rates 4'5'6
Passive
Leaching
(Ib/day)
NA
NA
NA
NA
NA
NA
NA
12.1
NA
NA
NA
Hull
Cleaning
(Ib/day)
NA
NA
NA
NA
NA
NA
NA
0.604
NA
NA
NA
Urban
Runoff
(Ib/day)
NA
NA
NA
NA
NA
NA
NA
0.181
NA
NA
NA
Background
(Ib/day)
NA
NA
NA
NA
NA
NA
NA
0.181
NA
NA
NA
Atmospheric
Deposition
(Ib/day)
NA
NA
NA
NA
NA
NA
NA
0.0181
NA
NA
NA
Estimated
Urban
Runoff
Loading
Rates 7
(Ib/day)
NA
NA
NA
NA
NA
NA
0.032
NA
1.0
1.8
17
NA- Not available.
(1) Model loading rates do not include contributions from study vessel sewage waste streams as these discharges are not covered under P.L. 110-299.
(2) Estimated loadings from concentrations for medium sewage treatment facilities (-10 mg/d) derived from concentrations presented in National Research Council (1993).
(3) Estimated nutrient loads from an actual sewage treatment facility with advanced nutrient removal capabilities with an average of approximately 40 mgd discharge (Albert, 2007).
(4) Estimated point source loads to Shelter Island Yacht Basin (California Regional Water Quality Control Board, 2005).
(5) Passive leaching and hull cleaning loading rates were based on an assumption of 2,363 recreational vessels present in Shelter Island Yacht Basin.
(6) Urban runoff contributions were based on a watershed area of 0.84 mi2 draining to Shelter Island Yacht Basin, and the atmospheric deposition loads were based on a surface area of
Shelter Island Yacht Basin of 0.27 mi2.
(7) Estimated urban stormwater loads were based on loading rates presented in Davis et al, 2001 and an assumed watershed area of 17 mi2 (MA DEP, 2006).
The loading rates presented are average annual daily loads.
                                                                           398

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
4.7    CONCLUSIONS

       This screening-level analysis evaluated the potential for discharges incidental to the
normal operation of vessels to pose a risk to human health, welfare, or the environment in large
water bodies. The analysis includes all sizes of commercial fishing vessels and other
nonrecreational vessels less than 79 feet in length. EPA selected a Level I screening-level model
(see Section 4.1) to help assess the potential impacts from study vessels' incidental discharges
and modeled several scenarios combining different vessel assemblages and different hypothetical
harbors to represent a range of environmental conditions potentially observed in harbors across
the United States. The modeled constituent concentrations from the discharges into the
hypothetical harbor for the 12 scenarios did not exceed an aquatic life or human health NRWQC
solely from study vessel discharges; however, the model  did not account for background
loadings. Certain pollutants (e.g., arsenic and dissolved copper) are more likely to contribute to a
water quality criterion being exceeded under real-world conditions. Furthermore, the model's
capabilities do not allow for the evaluation of whether these discharges cause localized impacts
(see Section 4.2), nor do they allow an analysis of issues  such as bioaccumulation or persistent
toxicity in water bodies or accumulation of pollutants in sediments.

       As discussed in the introduction, EPA's fraction of freshwater analysis is only intended to
evaluate environmental effects from vessel discharges at the water body or harbor scale and does
not address the environmental effects that could potentially occur in localized areas such as  small
side embayments or marinas. As discussed in Section 4.1, the "fraction of freshwater model"
does not describe concentration gradients within plumes from vessels. Accounting for  spatial and
temporal variability in a harbor would require a more data intensive dynamic model and is
beyond a Level I screening-level model. EPA acknowledges that incidental discharges from
study vessels may pose an environmental  threat in confined areas with low receiving water
flushing rates and a large population of vessels. In the dilution analysis discussed in Section 4.6,
EPA determined that a "tipping point" dilution factor of greater than 358  would be required to
avoid exceeding any NRWQC based  on the estimated loading rates used in the model (see Table
4.6.1). These results suggest that the loading rates represented in the model may have the
potential to cause a water quality criterion to be exceeded on  a localized scale either before
complete mixing is achieved in the receiving water (i.e., as the plume dissipates) or if the
discharges are released in a receiving water with a dilution potential of lower than 358. The
model further suggests that these vessels may be more likely to contribute to an NRWQC being
exceeded (particularly where the diluting factor is high for a pollutant) where the ambient
concentrations or other sources of pollutants are significant. On the other hand, EPA has tended
to use conservative estimates of some parameters (e.g., flow and pollutant concentrations) in its
modeling.

       In the "fraction of freshwater model," EPA calculated the instantaneous concentration in
the hypothetical harbor based solely on pollutant contributions from discharges from study

                                           399

-------
      Chapter 4 - Potential Large Scale Impacts of Study Vessels Incidental Discharges to Human Health and the Environment
vessels. Although the assumption that harbor background pollutant concentrations are zero for all
analytes is likely unrealistic, removing other loading considerations from model calculations
allows for the assessment of the potential for study vessel incidental discharges alone to cause an
NRWQC  to be exceeded. Although the "fraction of freshwater model" results suggest that study
vessels' incidental discharges will not cause an environmental impact on their own, the fact that
pollutants are present in the vessel discharges at concentrations that exceed the NRWQC at end-
of-pipe may support a determination that  some of these discharges have the potential to
contribute to a water quality standard exceedence.

       Based on the dilution results, the two pollutants that represent the greatest risk for
contributing to an environmental effect or water body impairment are total arsenic and dissolved
copper. EPA determined that the loading  rates from the metropolitan harbor (i.e., Model
Scenarios 9 and 11) were at the greatest risk of exceeding the NRWQC for these pollutants.
However, the minimum dilution factors required to avoid exceeding the NRWQC for these
pollutants (i.e., 284 for total arsenic and 144 for dissolved copper in the recreational harbor) are
similar to the lowest dilution factor represented in the hypothetical harbor scenarios (i.e., 494).
This suggests that study vessel's incidental  discharges may be contributing a significant load of
these two pollutants to the water body. Given the right environmental conditions (i.e., low
flushing) or pollutant loadings from other point/nonpoint sources (e.g., recreational vessels, large
commercial vessels, stormwater runoff, and industrial and municipal point sources), the
concentrations of these pollutants may have a potential to cause or contribute to an exceedence
of the NRWQC, regardless of vessel class distributions. These results are consistent with real-
world observations that metals are frequently  associated with vessel discharges in concentrations
of potential environmental concern (see Chapter  3). In particular, environmental impacts from
dissolved copper leaching from hull coatings has been well documented in low flushing
environments such as Shelter Island Yacht Basin near San Diego, California, and Marina Del
Rey Harbor in Los Angeles, California.

       Nutrients from study vessels' incidental discharges represent another pollutant class with
the  potential to contribute to deleterious environmental effects. Nutrients differ from other
pollutants present in vessel discharges in  that the environmental effects are driven by site-
specific environmental conditions (e.g., water temperature, types of algae present, limiting
nutrient). For example, the estimated nutrient loads used in the modeling analysis may contribute
to an environmental effect in one  water body, but not another depending on a variety of factors
that control eutrophication. EPA has not developed an NRWQC for nutrients; however, some
states have established water-body-specific or state-wide standards for nutrients based  on site-
specific evaluations.
                                           400

-------
                                                                Chapter 5 - Summary of Findings
CHAPTER 5
SUMMARY OF FINDINGS
       This chapter summarizes the major findings of EPA's detailed analyses described in
Chapters 1, 3, and 4. It describes findings on vessel classes that are covered by this study. It
summarizes major findings from the characterization of select discharges from the study vessels,
including EPA's interpretation of these findings in the context of the level of potential risk from
these pollutant loadings. Additionally, it discusses major findings of EPA's assessment of the
predicted impacts of these discharges to a hypothetical harbor. This chapter also briefly
summarizes possible benefits to human health, welfare, and the environment from reducing,
eliminating, controlling, or mitigating discharges from study vessels.

5.1    SUMMARY OF CLASSES OF VESSELS COVERED BY THIS STUDY

       EPA estimates there is a population of approximately  140,000 study vessels. According
to the U.S. Coast Guard's Marine Information for Safety and Law Enforcement (MISLE)
database, there are approximately 70,000 commercial fishing vessels operating in the United
States.  These vessels represent the largest category of study vessels. Passenger vessels comprise
the second highest number of vessels  within the study population, with approximately 21,000
vessels. These vessels are further classified by  subtypes according to the types of activities in
which they are involved,  such as diving vessels, charter fishing vessels, ferries, harbor cruise
vessels, and sailing vessels. The study population also includes over 11,000 utility vessels,
including tugs/towing vessels, school  ships, research vessels/ships,  mobile offshore drilling units,
offshore vessels, offshore supply vessels, oil recovery vessels, and industrial vessels.  Other
vessel categories such as  freight barges (approximately 8,000 vessels), tank barges
(approximately 900 vessels), freight ships (approximately 800 vessels), unclassified public
vessels (approximately 600 vessels), and tank ships (approximately 200 vessels) account for the
remainder of other non-recreational study vessels. An additional 27,375 vessels in the MISLE
database are also believed to be study vessels; however, the database does not indicate their type
of service. See Chapter 1  for additional discussions of the study vessel and recreational vessel
populations.
5.2    SUMMARY OF EFFLUENT CHARACTERIZATION OF SELECT DISCHARGES
       FROM THE STUDY VESSELS
       The major findings of EPA's analysis of the vessel discharge characterization data for
study vessels are summarized below. For this study, EPA sampled 61 vessels in nine states
generating over 22,000 data points. EPA tested for 301 analytes and detected 154 of these
analytes in at least one sample; therefore, 158 of the tested analytes were never found in the
                                         401

-------
                                                                 Chapter 5 - Summary of Findings
discharges. Section 5.2.1 discusses the estimated volumes of the discharges and Section 5.2.2
discusses the detected pollutants that may have the potential to pose a risk to human health or the
environment. See chapters 3 and 4 for more technical, in-depth discussions of these results.

   5.2.1  Estimated Volumes of Select Discharges from the Study Vessels

       EPA estimated volumes for each discharge from the study vessels based on data and field
observations from EPA's vessel sampling efforts, as well as information from secondary data
sources. Discharge volumes are important to both characterize the discharge and to analyze the
potential risk of the pollutant concentrations discharged from vessels. EPA also used these
discharge volumes to calculate flow rates for the modeling of pollutant loadings to a hypothetical
harbor in Chapter 4.

       Bilgewater generation rates are highly variable. EPA observed as little as 2 gallons of
bilgewater discharged from a tow/salvage vessel following a tow activity to as much as 750
gallons of bilgewater discharged during the daily bilge pump-out from a 62-foot shrimp boat
from the Gulf of Mexico. In general, based on observations from dozens of vessel operations,
EPA estimates that small (less than 79 feet), nonrecreational vessels typically generate between
10 and 15  gallons per day (gpd) of bilgewater.

       Stern tube packing gland effluent is by nature limited to the small amount of water
needed to provide cooling and lubrication to the gland around the drive shaft. The range in
estimated discharge for stern tube packing gland effluent is approximately 4 to 8 gpd.

       For deckwash water from tour boats, water taxis, and tow boats, EPA estimates a
discharge volume of between 20 and 30 gpd. Fishing boats are estimated to generate  more
deckwash water and the volumes generated vary with the type of boat. Trailers, trawlers,
gillnetters, and purse seiners may wash their decks three to four times per day while fishing,
producing as much as an estimated 750 to 900 gpd of deckwash water.

       The volume offish hold effluent generated by a fishing vessel depends on the size of the
vessel and the method used to keep the product fresh. Smaller fishing vessels such as small
salmon trailers or long-liners may discharge an estimated fish hold volume ranging from 70 to
200 gpd. Mid-size fishing vessels, such as gill netters and purse seiners found in Alaska and
shrimp boats in the Gulf of Mexico may discharge approximately 333 to 1,000 gpd. Larger
fishing vessels such as off-shore trawlers found in New England and tenders found in Alaska,
however, can have refrigerated seawater holding tanks  or ice hold tanks as large  as 15,000
gallons. These vessels are expected to offload seafood and discharge the fish hold effluent every
three to five days, resulting in an estimated flow rate ranging from 900 to 2,000 gpd.  EPA
estimates the volume offish hold cleaning effluent discharged by certain fishing vessels to be
anywhere  from 300 to 400 gallons per cleaning, which occur typically every three to five days
                                          402

-------
                                                                 Chapter 5 - Summary of Findings
when the fish holds are emptied (discharge volumes range from an estimated 60 to 200 gpd
depending on frequency of offloading).
       Fisherman unloading their catch to the dock from a trawler (dragger) in Massachusetts.

       Graywater volumes also vary considerably depending on the class of vessel and its use,
size, number of crew and passengers onboard, and types of gray water-generating activities
onboard (e.g., galleys, sinks, showers, and washing machines). For example, EPA estimated that
tugboats, some of which provide living quarters for three to five crew members, generate
approximately 130 gpd of graywater. Water taxis typically have considerably more people
onboard, but less graywater is generated per person because the discharge is typically limited to
bathroom sinks with an estimated 75-gpd discharge. Graywater generation on commercial
fishing boats might range from a few to hundreds of gpd, depending on the length of the trip and
the size of the crew.

       Finally, the volume of engine effluent discharged depends on the type of engine and
power level of operation. Vessels with outboard propulsion engines are estimated to discharge
between  1 to 2 gallons per minute (gpm) of raw cooling water per engine. The cooling water
discharge rate from inboard marine diesel engines varies based on power levels, but typically
averages around 20 gpm for the study vessels. Marine diesel generator sets require
approximately 5 to 6 gpm of cooling water for smaller units, and up to 20 to 25 gpm of cooling
water for larger marine generator sets. Daily  discharge rates for these engines are a function of
the daily operating time.

                                          403

-------
                                                                   Chapter 5 - Summary of Findings
    5.2.2  Analytes of Potential Risk in Select Discharges from Study Vessels

       EPA compared the measured concentration of any given analyte to its most stringent
benchmark1 (Table 3-1) as one means to identify pollutants in vessel discharges that may have a
potential to pose a risk to human health or aquatic life. EPA  divided the concentration of an
analyte by its corresponding benchmark to calculate a potential hazard quotient (PHQ). If a PHQ
is less than 1, there is less of a concern that the pollutant in the discharge will have impacts to
human health or aquatic life. An exception to this determination is when the pollutant is
persistent and/or bioaccumulative and may increase in concentration within the ecosystem food
chain to harmful levels. If a PHQ is equal to or greater than 1, then there is more of a concern.
However, PHQs of greater than 1 do not provide conclusive  evidence of risk to human health or
the environment for the following reasons:

    1.  Samples were collected at the "end of pipe" as the vessels discharged into larger waters
       (e.g., harbors, rivers). However, the discharge is typically diluted in the water body.
       Therefore, accounting for possible dilution in the receiving water could result in ambient
       PHQ of less than  1 (except possibly small harbors or marinas with limited or no flushing
       or where the receiving water PHQ is already above 1 due to other factors).

    2.  The benchmarks used to evaluate the potential for risk were always the most protective,
       even if it was not the most commonly applicable screening benchmark for that particular
       analyte. Given this, the potential for risk might be over-stated.

    3.  The surrounding ambient water or source water (vessel service2 or city water supply)
       used in the vessel systems that generated these discharges (e.g., engine cooling water
       drawn from ambient water or potable water used for  deck cleaning) may already contain
       high concentrations of some of these analytes. In these instances, a high analyte
       concentration measured at the "end of pipe" may not originate from vessel activities, but
       rather from the water used in these operations.

       EPA made the following general  observations based  on its review of the vessel discharge
data (see Chapter 3 for EPA's detailed analysis of the data):

    •  Dissolved copper was the analyte detected in vessel discharges at concentrations that
       consistently posed the greatest potential risk for local impacts and for contributing to
       exceedances of water quality standards in larger water bodies.  Copper is a heavy metal
       that can restrict the growth and reproduction of plants and algae and can produce both
       acute (short-term) and chronic (long-term) toxic effects on reproduction, growth, and
1 To provide a context for the level of contaminant concentrations presented, EPA used National Recommended
Water Quality Criteria (NRWQC) and several other benchmarks as a preliminary screen for all discharge data with
the potential to cause or contribute to the nonattainment of a water quality standard in a given receiving water body.
2 Service water here means the vessel potable water supply. For study vessels, vessel service water generally
originates from municipal water supply rather than produced on board.

                                            404

-------
                                                                  Chapter 5 - Summary of Findings
       survival in fish and shellfish. Prolonged exposure to elevated copper concentrations can
       lead to long-term liver and kidney damage in humans. Concentrations of dissolved
       copper exceeding the most protective screening benchmark were found in at least some
       samples for every sampled discharge type, except for outboard engine and generator
       engine effluents.

       Dissolved copper was detected at the highest concentrations in the deck washdown,
       graywater, fish hold, and bilgewater discharges from most vessel classes, particularly
       utility vessels (e.g., towboats, supply boats). PHQs for mean dissolved copper
       concentrations ranged from a low of 1.1 in graywater discharges to a high of
       approximately 200 in fish hold effluent. Based on concentration and average discharge
       volume, deck washdown and fish hold discharges contribute the most dissolved copper.

       Copper is released (leached) from antifouling hull coatings used on certain vessels to
       prevent buildup of organisms such as barnacles and algae. Copper can also be released
       via underwater hull cleaning, hull coating removal operations, and paint application.
       Although copper antifouling discharges were not measured, previous studies have shown
       it can be a major contributor to copper concentrations  in harbors, especially marinas with
       large vessel populations (see Section 3.2.8.1).

       Average ambient dissolved copper concentrations in the harbors sampled in this study
       were also slightly higher than the most protective benchmark (mean PHQ of 1.6).
       However, discharge concentrations still exceeded the benchmark even after subtracting
       the potential contribution of copper from ambient waters.

       Total arsenic3 concentrations in vessel discharges were also notably higher than the most
       protective screening benchmark. In samples where arsenic was detected, PHQs for mean
       total arsenic concentrations ranged from a low of 110  in graywater discharge to a high of
       2,900 in bilgewater discharge. Arsenic is a metalloid (a nonmetallic element with some
       metal properties) that is easily absorbed by  aquatic plants, algae, fish, and shellfish.
       Arsenic can cause a variety of acute and chronic toxic effects in aquatic organisms,  as
       well as in humans who ingest arsenic via drinking water and contaminated seafood.
       Arsenic is a known carcinogen, and prolonged high exposures via ingestion  can cause
       cancer, skin irritation, kidney and liver damage, and neurological damage.

       Despite the high potential toxicity of total arsenic, the risk posed to aquatic life is lower
       than what is  suggested by this analysis for two reasons. First, the screening benchmark
       for total arsenic is a human health criterion to prevent cancer-causing agents in drinking
       water and is  over 100 times lower than that of any other metal in this study. The high
3 See discussion in Section 3.1.3 regarding potential positive interference which may have resulted in elevated
measured concentrations of arsenic for a subset of samples.
                                           405

-------
                                                              Chapter 5 - Summary of Findings
   total arsenic PHQs in vessel discharges are the result of this low benchmark for human
   health, which is 2,000 times lower than the dissolved arsenic benchmark that is based on
   chronic, long-term toxicity to saltwater aquatic life. Many of the waters where many
   study vessels operate,  particularly for certain vessel types such as commercial fishing
   vessels, are not typically used as drinking water sources (i.e., ocean and coastal waters).
   However, some waters where study vessels operate (e.g., the Mississippi River) do serve
   as drinking water sources and high arsenic loadings in these waters could contribute to
   human health concerns.

   Second, between 20 to 100 percent of the total arsenic measured in the various vessel
   discharges can be attributed to ambient water that is used as source water for vessel
   systems. Vessel discharges most influenced by ambient total arsenic concentrations
   include those from stern tube packing glands, outboard engines,  and firemain systems.
   However, less than half of the total arsenic measured in bilgewater, deckwash, and fish
   hold discharges appears to be contributed by concentrations in ambient water, indicating
   that these discharges potentially contribute to arsenic toxicity in  receiving waters. Based
   on concentration and average discharge volume, deck washdown and fish hold discharges
   appear to contribute the most total arsenic.

•  Total aluminum concentrations exceeded benchmark concentrations in at least some
   samples for all discharge types; however, some of the aluminum concentrations in the
   discharge may be due  to background concentrations (e.g., not  added to the discharge by
   the vessel). Average PHQs for total aluminum ranged from a high of 39 in deck
   washdown discharge to a low of 1.8 in outboard engine effluent. The metalloid aluminum
   is most toxic to aquatic organisms in acidic conditions (i.e., waters with a pH < 7). When
   pH is neutral (7) or higher, aluminum can still inhibit growth of aquatic organisms but to
   a lesser extent. The pH measured in the vessel discharges and  ambient water sampled in
   this study was generally 7 or higher. Chronic exposure to high concentrations can cause
   aluminum to accumulate in bones offish (and humans) and lead to loss of kidney
   function.

   Indications are that the potential risk from total aluminum is greatest in deck washdown
   discharges, followed by fish hold discharges, and then stern tube packing gland
   discharges. For deck washdown, there is  an elevated risk because of the high aluminum
   concentrations (possibly from the leaching of the abundant amount of aluminum found on
   the surfaces of many vessels), as well as the potentially large discharge volume (up to
   900 gpd). Fish hold discharge also contains high total aluminum concentrations with
   discharges up to 1,000 gpd.  Although concentrations in the stern tube packing gland
   discharge are nearly as high as those in fish hold effluent, potential for risk from stern
   tube packing gland effluent is lower due to the lower volume of the discharge.
                                       406

-------
                                                                  Chapter 5 - Summary of Findings
       Ambient concentrations of total aluminum were high (ranging from 29 to 3,950 |ig/L -
       see Appendix E) in all of the sampled harbors for this study. The average concentration
       of total aluminum in ambient water is higher than the average concentration of total
       aluminum for all discharge types except for deck washdown, fish hold effluent, and stern
       tube/packing gland effluent. For fish hold4 and  stern tube packing gland discharges, it
       appears half of the measured total aluminum likely originates from the ambient water.
       Deck washdown discharge from vessels that use ambient water to clean decks have an
       estimated 20 percent of the measured total aluminum concentrations contributed by
       ambient water. In contrast, only 2 percent of the measured total aluminum concentrations
       were attributable to background concentrations  for vessels that used service water to
       clean their decks (primarily tugboats/utility vessels).

   •   Concentrations of other metals such as total iron and manganese  and dissolved cadmium,
       lead and zinc above their respective screening benchmarks were measured in some
       samples of deck washdown effluents (PHQs ranging from 1 to 11). These heavy metals
       are all known to produce acute and chronic toxic effects in aquatic organisms and
       humans, in the following order: cadmium is more toxic than lead, which is more toxic
       than zinc, which is more toxic than iron. These  elevated concentrations were particularly
       prevalent in the deck washdown discharges from utility vessels. However, decks of utility
       vessels (tugboats) are washed less frequently than fishing vessel decks, so overall metal
       loads from the two types of vessels are more comparable than concentrations alone might
       suggest. Although background concentrations of these metals in the ambient and service
       waters used to wash decks were generally low (except for  dissolved zinc in some
       background samples), average PHQs of all these metals in vessel discharges were not
       significantly greater than 1, indicating that these metals likely pose minimal potential risk
       to the environment.

   •   Total phosphorus concentrations were elevated  in some samples of bilgewater, deck
       washdown, fish hold, and graywater discharges. Average PHQs for total phosphorus in
       these discharge categories ranged from a high of 130 in fish hold effluent to a low of 14
       in graywater. Total phosphorus in some vessel discharges comes from detergents and
       soaps. Other total phosphorus loadings come from  decaying seafood (in fish hold) or
       leftover food (graywater). Based on concentration and average discharge volume, fish
       hold  effluent contributes the most total phosphorus.

       Phosphorus is an important macronutrient limiting reproduction and growth of plant
       material and algae (so called "primary production"). Elevated levels of phosphorus can
4 The assertion that background concentrations contribute approximately half of aluminum concentrations for fish
hold effluent assumes that vessels either took in the original fish hold water from the surrounding harbor waters, or
that the fishing grounds where the vessel took in the fish hold water share similar characteristics with surrounding
harbor waters.

                                           407

-------
                                                              Chapter 5 - Summary of Findings
   contribute to nuisance algal blooms, eutrophication (nutrient enrichment), and low
   dissolved oxygen levels in the water column (hypoxia). Ambient concentrations of total
   phosphorus, averaged across all sampled harbors, were twice the concentration of the
   PHQ screening benchmark.

•  The concentrations of reactive nitrogen compounds (e.g., nitrate, nitrite, ammonia) and
   the parameter TKN were generally not significantly elevated; except for in fish hold and
   fish hold cleaning effluents. Concentrations of ammonia exceed the most stringent
   recommended acute aquatic life criterion. Concentrations of TKN also exceeded the most
   stringent screening value. TKN in fish hold and fish hold cleaning effluent were also
   typical of concentrated raw sewage.

•  Biochemical oxygen demand (BOD) and chemical oxygen demand (COD) were elevated
   in bilgewater, deck washdown, fish hold, and graywater discharges. BOD and COD  are
   measures of oxygen-demanding substances present in the discharges (e.g., organic
   matter) that can contribute to hypoxia (low dissolved oxygen) in receiving waters.
   Average BOD concentrations were highest in fish hold effluents (as high as 25 times the
   concentrations in raw sewage), followed by graywater and then bilgewater and deck
   washdown water. The BOD levels in fish hold effluent and graywater are comparable to
   BOD concentrations in raw sewage.  Fish hold effluent also has a relatively high
   discharge volume, so this discharge can contribute a significant BOD/COD loading to
   receiving waters, particularly when multiple vessels discharge at the sample location
   (e.g., pierside at a fish processing facility).  Hence, depending upon receiving water
   characteristics, BOD and COD from fish hold effluent may significantly impact the local
   environment and contribute to water quality exceedances in receiving waters.

•  Pathogen indicators, E. coli, enterococci, and fecal coliforms, were also found in elevated
   concentrations in some samples of bilgewater and deck washdown (fishing vessels only),
   fish hold, and graywater discharges.  These  three types of bacteria are all found in animal
   digestive tracts. Epidemiological studies suggest a link between high concentrations of E.
   coli and enterococci in ambient waters and  incidents of gastrointestinal illnesses
   associated with swimming. Accordingly, they are used as indicators of the possible
   presence of intestinal pathogens. The highest concentrations by far of all three pathogen
   indicators were found in graywater, with PHQs of around 1,000 for all three bacteria. The
   estimated discharge volume of graywater from study vessels, however, is relatively small
   (130 gpd maximum). Larger vessels with additional crew or passengers are expected to
   generate considerably more graywater (see EPA's Cruise Ship Discharge Assessment
   Report, USEPA, 2008c). Fish hold effluent contained the second highest concentrations
   of these pathogen indicators and may pose  a potential level of risk considering the
   relatively high volume of this discharge and possible discharge by multiple vessels in the
   same location. However, EPA notes that most of the pathogen concentrations in fish hold

                                       408

-------
                                                                   Chapter 5 - Summary of Findings
       effluent were well below or similar to ambient water concentrations, and this study is
       inconclusive as to whether fish hold effluent results in additional discharge of pathogen
       indicators5.

    •   The semivolatile organic compound bis(2-ethylhexyl) phthalate was found in elevated
       concentrations in some samples of bilgewater, stern tube packing gland, deck washdown,
       firemain, and inboard engine and engine generator discharges. The highest PHQ of 59 for
       bis(2-ethylhexyl) phthalate was found in a bilgewater discharge sample. Even though
       bis(2-ethylhexyl) phthalate was found at elevated concentrations in multiple discharges,
       the overall frequency of detection was low and generally detected at concentrations just
       slightly above the benchmark. This compound is a plasticizer that is added to an ever-
       increasing variety of plastics to provide flexibility and is the most common phthalate in
       the environment. Although no conclusive evidence exists demonstrating bis(2-
       ethylhexyl) phthalate affects humans, high concentrations have been shown to feminize
       males of other species. Bis(2-ethylhexyl) phthalate was not analyzed for in fish hold or
       graywater discharge  samples.

    •   Benzene was the only volatile organic compound found with any frequency at
       concentrations above, but generally close to, the PHQ benchmark. Benzene is a known
       carcinogen that is a common constituent of fuel. Benzene can also be formed as a product
       of incomplete combustion of fuel. Elevated concentrations of benzene were detected in a
       bilgewater sample and in samples from both outboard engine and generator engine
       discharges.

    •   Long- or short-chain nonylphenol and octylphenol ethoxylates (two distinct subsets of
       alkylphenol  ethoxylates) were detected in some samples of bilgewater, stern tube/packing
       gland, deck washdown, and graywater discharges, and total nonylphenol  was detected in
       one sample from bilgewater. Nonylphenols were not analyzed for in samples of the
       remaining discharge  types.

       Nonylphenols (a term used generally here to identify a specific group of alkylphenols of
       potential human and  environmental concern which also includes the octylphenols) are
       manmade organic compounds that are used in a wide  variety of applications, such as the
       manufacturing of detergents, because of their surfactant properties. Nonylphenols are
       synthetic estrogens, which means they can mimic the natural vertebrate hormone estrogen
5 Fish hold water may also serve as a potential pathway for the spread of aquatic nuisance species (ANS). This might
occur where fish and water are taken onboard in one place and then transported significant distances for sale or
unloading and the water is discharged. Organisms discharged with the water may include parasites and commensals
taken in with fish, as well as organisms taken in with water used for refrigerated seawater. EPA did not study the
potential for these discharges to transport ANS; however, the Agency is identifying this as a potential area of
concern that may warrant further research.

                                           409

-------
                                                         Chapter 5 - Summary of Findings
and evoke an estrogen-like response. An example of such a response is the disruption of
male sexual development, causing female characteristics to emerge.

Commercial nonylphenol is most accurately described by CAS number 84852-15-3
(phenol, 4-nonyl-branched), but CAS numbers 104-40-5 (phenol, 4-nonyl-) and 25154-
52-3 (phenol, nonyl) have also been used to describe these compounds. The commercial
nonylphenol mixtures tested that correspond with EPA's criteria are those with CAS
numbers 84852-15-3 and 25154-52-3. The analyte category named "total nonylphenol" in
the database generated for this study is directly equivalent to the commercial mixture of
nonylphenol isomers specified under CAS Number 84852-15-3,  and thus is directly
comparable to the NRWQC.

Total nonlyphenol (or NP) was not detected, except in one bilgewater sample with a PHQ
of 4. Long-chain nonylphenol and octylphenol ethoxylates were  detected with far greater
frequency, but these longer chain compounds are more water soluble and less toxic than
NP. The long-chain nonylphenol and  octylphenol ethoxylates will all degrade into NP
over time; however, research is ongoing with regard to proportion and duration of the
conversion.

Table 5.1 summarizes the major findings discussed above.
                                   410

-------
Table 5.1. Analytes of Potential Risk by Discharge1



Discharge Type
(Volume/vessel)



Bilgewater
(2 to hundreds
of gpd; average
between 10 and
15 gpd)












Stern tube
packing gland
effluent
(from 4 to 8 gpd)








Analyte Group
 
-------
Table 5.1. Analytes of Potential Risk by Discharge1



Discharge Type
(Volume/vessel)




Deck
washdown
and/or runoff
(from 20 to 30
gpd utility; 750
to 900 gpd
fishing)




























Analyte Group
 W O O



































^
a>

8
«
i/i
re
ID
^
Copper
(detected in 29
of 31 samples -
PHQs from 2 to
65; highest for
utility vessels)

Zinc
(detected in all of
31 samples -
67% of PHQs
between 1 and
10. Max. cone, of
1 ,200 ug/L in
tugboat; PHQ =
14)

Lead
(detected in 1 5
of 31 samples -
PHQs in 2
samples from 10
to 21)

Cadmium
(detected in 2 of
31 samples -
max. cone. = 22
ug/L in tow boat;
PHQ = 90)







—
I
i/i
re
ID
^
Arsenic
(detected in 23 of
31 samples;
cone, from 4 to
83 ug/L - PHQs
from 200 to 4,000
because very low
NRWQC)

Aluminum
(detected in 30 of
31 samples -
PHQs between
7.5 and 150.
Max. cone, of
13,000 ug/L in
tugboat)

Iron
(detected in 1 8 of
19 samples -
PHQs between
3.1 and 48. Max.
cone, of 14,500
ug/L in tugboat)

Manganese
(detected only in
tugboats and
water taxi -
PHQs between
1.2 and 13. Max.
cone, of 12,800
ug/L in water
taxi)2


re
V
3
•o
c
re

O
HEM
(detected in 26
of 29 samples -
only 3 samples
with PHQ > 2.
Range of
concentrations
1 .1 to a max of
1 33 mg/L in a
tugboat)

SGT HEM
(detected in 22
of 29 samples -
only 1 sample
with PHQ > 2.
Range of
concentration
0.91 to a max
of 84 mg/L in a
tugboat)













412



U)
8 =
'« t?
« _

o S.
BOD and COD
(detected in 29 of 31
BOD samples- all
COD samples;
concentrations
roughly equivalent to
raw sewage- all
vessel types)

TSS
(detected in all of 32
samples- PHQs
between 1 and 17;
max. concentrations
in tugboats)

TRC
(detected in 7 of 31
samples- PHQs
between 23 and 100.
Max. cone, of 0.8
mg/L in fish trolling
boat















i/i
'c
V
'^
"tj
Z
Total
phosphorus
(detected in all of
31 samples;
PHQ as high as
220 in a tugboat)































Q
c
a.
^
o







































Comments




Concentrations of many
dissolved metals in so-
called "utility" or non-fishing
vessels statistically higher
compared with fishing
vessels.

Elevated concentrations of
total arsenic, aluminum
and iron strongly
influenced by surrounding
ambient water
concentrations.

TOC detected in all of 25
samples at concentrations
from a low of 3.5 to a very
high 350 mg/L (tugboat).

Bis(2-ethylhexyl)-phthalate
detected in only 1 sample
(PHQ = 5.6).

Only 3 of 29 vessels
sampled had detectable
concentrations of NPEOs
of the shortest chain
(NP3EO) indicative of
detergents; concentrations
ranging from 0.80 to 29
ug/L.





-------
Table 5.1. Analytes of Potential Risk by Discharge1



Discharge Type
(Volume/vessel)



Fish hold/
Fish hold
cleaning
effluent -
Fishing vessels
only
(tens to several
hundred or a few
thousand gpd
(or gallons per
several days
dependent upon
offloading
frequency)
based on fishing
vessel type and
platform)














Analyte Group
 W O O
NA






























^
>
8

i/i
re
ID
^
Copper
(detected in 23
of 26 samples -
PHQs from 1 to
300. Max. cone.
of 921 ug/L in
shrimper)


























—
I
i/i
re
ID
^
Arsenic
(detected in 1 6 of
26 samples;
cone, from 3.1 to
380 ug/L -PHQs
from 170 to
potentially 21,000
because very low
NRWQC)























Q)
re
V
3
•o
c
re

O


































U)
re c
" re
« -3
.2 'o
0 Q.
BOD and COD
(detected in 24 of 26
BOD samples- all
COD samples;
median
concentrations of
BOD and COD were
440 and 940 mg/L
with max of 5,100
and 8,700 mg/L
equivalent to
sewage sludge)

Sulfide
(detected in 7 of 25
samples- PHQs
between 5 and 80;
max. concentrations
= 0.16 mg/L in fish
trawler)

TSS
(detected in all of 26
samples- PHQsof
4 samples between
1 7 and 37)
DO
(hypoxic, i.e., <2.0
mg/L in 3 of 26
samples- all purse
seiners)



i/i
'c
0)
•3
Z
Ammonia
(detected in 25
of 26 samples;
cone, from 0.087
to 1 60 ug/L -
PHQ at max
cone. = 133)

TKN
(detected in 25
of 26 samples;
values indicative
of strong
sewage)

Total
phosphorus
(detected in 25
of 26 samples -
all but 3 samples
resulting in
PHQs above 10;
highest
PHQ=760)








m
°Q
C
a.
0
Z
NA

































Comments



Level of detection of all
analytes similar in fish hold
cleaning effluent, although
concentrations somewhat
reduced.

HEM detected at a max
cone. = 1 6 mg/L; PHQ = 1 .























                                                               413

-------
Table 5.1. Analytes of Potential Risk by Discharge1



Discharge Type
(Volume/vessel)



Graywater
(tugs- 130 gpd;
taxis -75 gpd;
fishing - a few to
few hundred
gpd)















Analyte Group
  O O
NA




















^
a>

~o
«
1/1
re
ID
^
Copper
(detected in all of
8 samples -
PHQs from 1 .7
to 90. Max. cone.
of 280 ug/L in
tugboat)

Zinc
(detected in all of
8 samples - max.
cone. = 1 ,500
ug/L in sink
water from a
water taxi; PHQ
= 19)







, — ,
I
i/i
re
ID
^























re
3
•o
c
re

O
HEM
(detected in all
of 8 samples -
4 samples with
PHQ 2 or more.
Concentrations
from 9. 4 to a
max of 100
mg/L in
tugboats)

SGT HEM
(detected in 6
of 8 samples -
only 1 sample
with PHQ >2.
Range of
concentration
1 .3 to a max of
35 mg/L from a
tugboat)



U)
8 =
'« •§
.2 'o
0 Q.
Sulfide
(detected in 5 of 8
samples- PHQs
between 4.8 and
370; max.
concentration = 0.73
mg/L in tugboat)

BOD and
COD
(detected in all of 8
samples; median
concentration of
BOD and COD were
260 and 440 mg/L,
respectively with
max of 1,200 and
4,000 mg/L
indicative of strong
sewage)




i/i
'c
V
•3
Z
Total
phosphorus
(detected in 8 of
8 samples -all
but 3 PHQs
above 10;
highest PHQ=34
in tugboat)















J2
o
c
a.
0
Z
























Comments



Only 1 of 8 vessels
sampled had detectable
concentrations of NPEOs
of the shortest chain
(NP3EO) indicative of
detergents; concentration
of 0.99 ug/L.














                                                               414

-------
Table 5.1. Analytes of Potential Risk by Discharge1



Discharge Type
(Volume/vessel)




Propulsion
Engine Effluent
- inboard
(20 gpm - high
power)











Propulsion
Engine Effluent
- outboard
(1 to 2 gpm)












Analyte Group
 v> O O
PAHs
(6 probable
carcinogenic
PAHs
detected in
sample from
a recreational
vessel with a
gasoline
engine.
Measured
concentration
s result in
PHQs from
2,100to
4,800)
Benzene
(detected in 6
of 6 samples
-only one
sample with a
PHQ above
10; value of
28 based on
a max. cone.
of 62 ug/L in
sample from
research
vessel
averaged
from variable
speeds)
^
a>

"o
~
i/i
re
ID
^
Copper
(detected in 1 2
of 13 samples -
PHQs from 3
samples from 1 1
to 17. Max. cone.
of 53 ug/L in
sample from
water taxi at idle)

























, — ,
I
i/i
re
ID
^


































re
3
•o
c
re

O



































U)
S =
'« t?
« _

o £
Temperature
(high idle only;
temperature
increases of up to
20°C)






























i/i
'c
V
'^
"3
Z
NA















NA















r*i
M
J2
o
c
a.
^
o

NA















NA


















Comments




Note: Though the
recreational vessel with the
gasoline engine is not a
"study vessel", it
represents EPA's only
samples from a gasoline
engine. EPA assumes
gasoline engines from
similarly designed study
vessels would have similar
characteristics.





















                                                               415

-------
Table 5.1. Analytes of Potential Risk by Discharge1


Discharge Type
(Volume/vessel)




Engine Effluent
- Generator
(5 to 25 gpm)










Firemain
Systems
(no volume
estimated- used
infrequently)

Analyte Group
 = Q.
IS E re p
"5 ? o
> W O O
Benzene
(detected in 3
of 5 samples
-only one
sample with a
PHQ
approaching
10; value of 9
based on a
max. cone, of
21 ug/L in
sample from
a fire boat)






^
a>
1
«
1/1
re
ID
^













Copper
(detected in 4 of
6 samples -
PHQs from 3.8
to 23; highest for
a tour boat)

. — .
I
i/i
re
ID
^




















Si
re
V
3
•o
c
re

O





















U)
S =
'« 'S

2 "5
o o.





















i/i
'c
V
'^
•3
Z
NA












NA






S
j«
Q
c
S-
^
0
Z
NA












NA







Comments























       Notes:
       (1) Generally includes analytes when a large proportion of the samples have concentrations exceeding the NRWQC, when several of the samples have PHQs > 10, when a few
       samples result in PHQs greatly exceeding the screening benchmark (i.e., 100s to 1,000s), or, in the case of oil and grease and for nonylphenol, when one or more samples exceed an
       existing regulatory limit by more than a factor of 2. See text above and in Section 3.1.3 for a definition of PHQs and Table 3.1 for screening benchmarks used to calculate these values.
       (2) Longer chain nonylphenol and octylphenol ethoxylates degrade to shorter chained ethoxylates under aerobic conditions. In general, the shorter the chain, the more hydrophobic,
       persistent, and toxic the substance becomes. 4-Nonylphenol (NP) is a shorter-chain nonylphenol that has been found in surface water and is toxic to aquatic life.  NP is formed from
       the longer chain nonylphenol and octylphenol ethoxylates as they break down. The time span from time of use on the vessel to time of sampling of the discharge was probably not long
       enough for this to occur, except for discharges of bilgewater.
       gpm = gallons per minute
       gpd = gallons per day
       NA - Not applicable; discharge  not analyzed for this analyte group.
                                                                                    416

-------
5.3    SUMMARY OF PREDICTED IMPACTS FROM SELECT POLLUTANTS IN STUDY
       VESSEL DISCHARGES

   5.3.1  Potential Watershed-Wide Impacts from Study Vessels

       Using estimated discharge volumes and average pollutant concentrations, EPA evaluated
the potential for cumulative effects of the discharges from an assemblage of study vessels on a
large hypothetical harbor. The evaluation used a screening-level water quality model to estimate
the pollutant concentration into several hypothetical harbors based on different scenarios of
vessel groups. Model assumptions included instantaneous and universal dilution of vessel
discharges in the harbor and a background concentration of zero for all analytes in the harbor
environment (i.e.,  the model is not able to evaluate whether vessel discharges are likely to cause
environmental or human health impacts in the immediate vicinity of the vessel discharges or in
small water bodies). Instead, the model can only analyze potential vessel loadings to and impacts
on hypothetical large water bodies. Furthermore, the model is not able to analyze parameters that
do not have numeric aquatic life or human health based criteria such as BOD or nutrients.

       The model did not predict that discharges from the study vessels solely exceeded aquatic
life or human health NRWQC for any of the hypothetical harbor scenarios evaluated. This is
primarily due to the large dilution predicted in these large harbors (even with low flushing).
However, some of these pollutants from these vessels could reasonably have more significant
local impacts (although determining this is outside the scope of the model used in this study). In
smaller water bodies with many vessels or in more confined areas of a harbor with little to no
flushing, EPA believes study vessel discharges have the potential to cause or contribute to
exceedances of NRWQC in receiving waters.

       Under the low-dilution scenarios,  dissolved copper and total arsenic discharges represent
the greatest environmental concern and are more likely than other pollutants to contribute to
exceedances of water quality standards, particularly if there are other sources of these pollutants
(e.g., stormwater runoff) present. These results are summarized below.

Dissolved Copper

       EPA determined that the loading rates of dissolved copper from a metropolitan harbor
likely posed the greatest potential risk to human health and aquatic  life from study vessels on a
large scale. Compared to other types of harbors, a metropolitan harbor has a higher level of
activity from its vessel population and has more support utility vessels such as supply boats,
tow/salvage vessels, and tugboats. The model predicted that discharges from study vessels have
the reasonable potential to contribute a significant load of dissolved copper to a water body.
Furthermore, when considering the loadings of dissolved copper from other  sources (e.g.,
recreational vessels, large commercial vessels, stormwater runoff, and industrial and municipal
                                          417

-------
point sources), the model results suggest a reasonable potential for the concentrations of
dissolved copper to exceed the NRWQC in this type of harbor.

       The results of this study are consistent with real-world observations that metals are
frequently associated with vessel discharges in concentrations of potential environmental
concern. Environmental impacts from dissolved copper leaching from antifouling hull coatings
have been well documented in low-flushing environments in harbors with large numbers of
recreational vessels, such as the Shelter Island Yacht Basin near San Diego, California, and
Marina Del Rey Harbor in Los Angeles, California (see Section 3.3.8.1 of this report). The
impacts from the high levels of dissolved copper include reduced primary production and
productivity;  accumulation of copper in sediments, reducing sediment quality; and chronic low-
level toxicity to aquatic organisms, especially sensitive mollusks, crustaceans, and echinoderms.

Total Arsenic

       EPA determined that the loading rates of total  arsenic (and to a certain extent, dissolved
arsenic) may  pose a potential risk to human health and the environment in low-dilution or low-
flushing environments. Arsenic was found to be ubiquitous in this study, both in vessel
discharges and in ambient water6. Although arsenic concentrations in ambient water can be quite
high in select harbors, certain discharges from study vessels contribute to the overall arsenic
load. While the source of total arsenic in vessel discharges is unknown, EPA suspects that
atmospheric deposition contributes to total arsenic concentrations in deck washdown and
possibly in bilgewater. Total arsenic in fish hold discharges may be biological in origin (from
seafood catch) or from sediment entrained in the catch. The biological contribution of arsenic
may be significant in that total arsenic concentrations  are substantially greater in seawater
organisms than in freshwater organisms (Francesconi  and Kuehnelt, 2002;  USEPA 2003b).

       The greatest impact of high total arsenic in harbors is primarily via the food chain and
subsequent bioaccumulation to high levels in seafood consumed by humans. Arsenic exposure
through drinking water is also of concern where receiving water is used as  a source for drinking
water.  Arsenic is strongly linked to cancer in humans  and a potent inhibitor of certain enzymes in
vertebrates.

    5.3.2  Potential Localized or  Near-Field Impacts of Vessel Discharges to Receiving
          Waters

       EPA found that some study vessel incidental discharges may pose an environmental
threat in confined water bodies with low flushing rates and a large population of vessels, in water
bodies that are hypoxic or hypereutrophic, and/or where the background concentrations or other
6 See discussion in Section 3.1.3 regarding potential positive interference which may have resulted in elevated
measured concentrations of arsenic for a subset of samples.
                                           418

-------
sources of these pollutants are significant. In addition to the parameters (copper and arsenic)
discussed in Section 5.3.1, the following classical pollutants may likely exhibit near-field effects.

BOD and COD

       In general, oxygen-demanding compounds in vessels discharges (measured as BOD and
COD) are expected to pose little risk to the environment due to the relatively low volume of
vessel discharges that contain these pollutants. However, the frequency and magnitude of BOD
and COD in certain discharges (as much as 25 times the concentrations found in raw sewage)
warrant additional discussion.

       Specifically, the relatively high  BOD and COD concentrations in fish hold and fish hold
cleaning  effluent could pose a localized  water quality impact in areas such as small side
embayments where flushing rates are low or where portions of the water body are already low in
dissolved oxygen. The high levels of BOD result from the degradation of organic material and its
by-products in the fish hold. Higher volume discharges with high BOD concentrations (e.g.,
certain fish hold effluent) may contribute to localized hypoxic conditions in receiving waters,
depending on the volume of effluent discharged, the number of vessels discharging in confined
areas, and other factors such as season and water temperature.

Pathogen Indicators

       Bacteria such as E. coli, enterococci, and fecal coliforms are generally of limited concern
for most  discharges where the pathogens were present (i.e., bilgewater, deck washdown, and fish
hold discharges). However, high levels of pathogens in graywater (and potentially other
discharge types) may pose some risk to human health and larger vessels with additional crew or
passengers are expected to generate considerably more graywater than smaller vessels. However,
looked at on a relative basis, the risk from pathogens in graywater is substantially lower than
risks from other sources that cause very  high concentrations of pathogen indicators in
surrounding ambient water. For example, during sampling for this study in Massachusetts that
took place in wet weather, a sanitary sewer overflow and a combined sewer overflow caused
extremely high pathogen indicator counts in two different harbors, relative to what would be
expected from graywater discharges from study vessels.

Total Phosphorus

       Nutrients in vessel discharges are generally expected to pose little risk to the
environment. However, the frequency and magnitude of total phosphorus in certain discharges
warrants  some additional discussion.

       The environmental effects of nutrients are driven by site-specific environmental
conditions (e.g., receiving water temperature, types of algae present, and limiting nutrient
conditions). For example, nutrients in vessel discharges may contribute to an environmental
                                          419

-------
effect in one water body, but not another depending on a variety of environmental conditions that
control eutrophication (excess productivity in a water body). While EPA has not developed
NRWQC for total phosphorus and other nutrients in coastal waters, some states have established
water-body-specific or state-wide standards for nutrients based on site-specific evaluations.

       As mentioned above, the water quality impact of concern for total phosphorus is
eutrophi cation. The first indications of potential problems are the increased ambient levels of
total phosphorus, often followed by an immediate increase in the density (biomass) of the
planktonic algal community. This increased algal biomass usually blocks light and reduces water
clarity and may contribute to nuisance algal blooms and declining dissolved oxygen. Of note in
this study was that the mean total phosphorus concentration in the 15 ambient water samples
collected was two times  the screening benchmark, suggesting that the incremental effect of
discharges from study vessels may be small.

5.4    POSSIBLE BENEFITS TO HUMAN HEALTH, WELFARE, AND THE
       ENVIRONMENT FROM REDUCING, ELIMINATING, CONTROLLING, OR
       MITIGATING ONE OR MORE OF THE DISCHARGES FROM THE STUDY
       VESSELS

       Some vessel discharges from commercial fishing vessels and commercial vessels less
than 79 feet in length may have the potential to impact the aquatic environment and/or human
health. As noted above, using the results obtained in this study, EPA modeled a hypothetical
large harbor to evaluate the potential water quality impacts caused by the nine vessel discharge
types EPA sampled. Based on this evaluation,  EPA determined that the incidental discharges
from study vessels to a relatively large water body are not likely to solely cause an exceedance of
any NRWQC (i.e., these discharges are unlikely to pose acute or chronic excursions of the
NRWQC across an entire large water body). However, many of the pollutants in the vessel
discharges were at end-of-pipe concentrations that exceeded an NRWQC, and therefore have the
potential to contribute to an exceedance of water quality  standards at a more localized scale. The
study results indicate that total arsenic and dissolved copper are the most significant water
quality concern for the study vessels as a whole. These pollutants are more likely than other
pollutants to contribute to exceedances  of water quality standards, particularly if there are other
sources of pollutants or the receiving water already has high background concentrations.
                                         420

-------
       Gloucester Harbor faces many environmental stressors including Combined Sewer
       Overflows and Urban Stormwater Runoff. For most pollutants, the impact of these sources
       may be more significant than from study vessels. However, some pollutants, such as copper
       or BOD are discharged in notable quantities from certain study vessel discharges,

       Like an individual house in an urban watershed, most individual vessels have only a
minimal environmental impact. However, the impacts caused by these vessels is potentially
significant where there are high vessel concentrations, low circulation in waters, additional
environmental stressors, or pollutant loadings from  other sources (e.g., recreational vessels, large
commercial vessels, Stormwater runoff, and industrial and municipal point sources).  Reducing
certain discharges or certain pollutants in discharges from these vessels in sensitive waters may
result in significant environmental benefits to those waters; however, EPA did not analyze the
feasibility or cost of managing these discharges as part of this study.
                                           421

-------
                           Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

CHAPTER 6

 ANALYSIS OF THE EXTENT TO WHICH INCIDENTAL

DISCHARGES ARE CURRENTLY SUBJECT TO REGULATION
UNDER FEDERAL LAW OR A BINDING INTERNATIONAL

OBLIGATION OF THE UNITED STATES	

       As discussed in Chapter 1, Congress directed EPA, in consultation with the U.S. Coast
Guard and other interested federal agencies, to conduct a  study of discharges incidental to the
normal operation of all fishing vessels and nonrecreational vessels less than 79 feet in length
(study vessels). Among other things, the study's charge directed EPA to include an "analysis of
the extent to which the discharges are currently subject to regulation under federal law or a
binding international obligation of the United States" (Public Law (P.L.) 110-299 § 3(b)(6)).
This chapter and accompanying tables present that analysis. Note, however, that as discussed in
Chapter 1, this chapter includes some discussion of treaties and statutes that pertain to nonstudy
vessels for information purposes. In accordance with P.L. 110-299, this study does not include
significant discussion about discharges of sewage or ballast water.1

       This chapter is organized into four sections. Section 6.1 offers brief overviews of the
international obligations addressing vessel discharges, while Section 6.2 summarizes applicable
federal statutes and regulations. Section 6.3 includes a brief overview of other international  and
federal laws that do not directly regulate discharges incidental to the normal operation of a
vessel, but which the Agency felt merited some discussion. Finally, Section 6.4 provides tables
identifying which applicable laws apply to specific incidental discharges.

6.1   INTERNATIONAL AGREEMENTS

6.1.1  The International Convention for the Prevention of Pollution from Ships (MARPOL
       73/78)

       The International Convention for the Prevention of Pollution from Ships, 1973, as
modified by the Protocol of 1978 (MARPOL 73/78), is the primary international instrument for
regulating and preventing pollution from vessels. A total of 150 countries are Parties to
1 As of the writing of this report, ballast water discharges are regulated by the U. S. Coast Guard under the National
Invasive Species Act of 1996 (NISA), by EPA under Section 402 of the Clean Water Act, and by several states
under state law. NISA is discussed briefly in this analysis to the extent that it addresses invasive species from
sources other than ballast water. Furthermore, the International Convention for the Control and Management of
Ships' Ballast Water and Sediments (BWM Convention), adopted by the International Maritime Organization (IMO)
in 2004, establishes ballast water discharge standards. The Convention has not yet attracted the requisite number of
Parties necessary for its entry into force. For further discussion, see Standards for Living Organisms in Ships'
Ballast Water Discharged in U.S. Waters (74 FR 44,631 (Aug. 28, 2009)).


                                         422

-------
                           Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

MARPOL. MARPOL includes six annexes, covering six categories of vessel discharges: oil
(Annex I), noxious liquid substances (Annex II), harmful packaged substances (Annex III),
sewage (Annex IV), garbage (Annex V), and air emissions (Annex VI).

       Before entering into force, the Convention required ratification by 15 member states, with
a combined merchant fleet of not less than 50 percent of the total world shipping fleet, measured
by gross tonnage. To ratify the convention, member states are required to ratify only Annexes I
and II; the remaining annexes are optional.  The United States has ratified Annexes I, II, III, V,
and VI (the United States has not ratified Annex IV, which regulates sewage discharges from
ships; the United States regulates sewage under Section 312 of the Clean Water Act, which is
discussed in Section 6.2, Federal Laws).

       In the United States, MARPOL is primarily implemented through the Act to Prevent
Pollution from Ships (APPS), 33 U.S.C. §§ 1901-1915. APPS implements Annexes I, II, V, and
VI. Annex III of MARPOL is implemented through the Hazardous Materials Transportation Act,
49 U.S.C.  § 5101 etseq. These implementing statutes are discussed in depth in Section 6.2,
Federal Laws.

       6.1.1.1    MARPOL Annex I: Prevention of Pollution by Oil

       MARPOL Annex I establishes requirements for the control of oil pollution from vessels.
As previously discussed in this report, small to large amounts of oil can be found in numerous
vessel discharges, including bilgewater, deck runoff, and engine effluent. The requirements of
this Annex apply to all ships operating in the marine environment, unless expressly provided
otherwise.

       Every oil tanker of 150 gt and above and every other ship of 400 gt and above is required
to undergo a series of surveys to ensure that the ship's structure, equipment, systems, fittings,
arrangements, and material are in full compliance with all applicable Annex I requirements and
do not pose "an unreasonable threat of harm to the marine environment" (Annex I, Regulations
6.1  and 6.4.1). The surveys are required before the ship is put in service (or before an
International Oil Pollution Prevention Certificate [IOPP Certificate], explained below, is issued
for the first time); for IOPP Certificate renewal purposes; at certain intervals surrounding the
anniversary date of the ship's IOPP Certificate; and after certain repairs or renewals are
completed (Annex I, Regulation 6).

       Oil tankers of 150 gt and above and ships of 400 gt and above that travel to ports or
offshore terminals under the jurisdiction of other Parties to Annex I are required to have an IOPP
Certificate, which indicates completion of and compliance with Annex I's inspection
requirements. These certificates are issued or endorsed by the government of the state, or any
persons or organizations authorized by it, under whose authority the ship is operating (Annex I,
                                         423

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

Regulation 7). The IOPP Certificate shall not be issued for a time period exceeding five years,
subject to various survey provisions contained in the Annex (Annex I, Regulation 10).

       Annex I prohibits the discharge of oil or oily mixtures into the sea, except under the
following circumstances:

    •   Ships of 400 gt and above, whether inside or outside a special area where:
           o  The ship is proceeding en route.
           o  The oily mixture is processed through area-appropriate oil  filtering equipment
              (under Regulation 14).
           o  The oil content of the effluent without dilution does not exceed 15 parts per
              million (ppm).
           o  The oily mixture does not originate from cargo pump-room bilges on oil tankers.
           o  The oily mixture, in case  of oil tankers, is not mixed with oil cargo residues.

    •   Ships of less than 400 gt, whether inside or outside a special area where:
           o  The ship is proceeding en route.
           o  The ship has in operation equipment of a design approved by the government
              under whose authority the ship is operating, that ensures that the oil content of the
              effluent without dilution does not  exceed 15 ppm.
           o  The oily mixture does not originate from cargo pump-room bilges on oil tankers.
           o  The oily mixture, in the case of oil tankers, is not mixed with oil cargo residues
              (Annex 1, Regulation 15).

    •   Discharges of oil or oily mixtures from cargo areas of oil tankers outside special areas
       where:
           o  The tanker is more than 50 nautical miles from the nearest  land.
           o  The tanker is proceeding en route.
           o  The instantaneous rate of discharge of oil content does not  exceed 30 liters per
              nautical mile.
           o  For tankers delivered on or before December 31, 1979, the total quantity of oil
              discharged into the sea does not exceed 1/15,000 of the total quantity of the
              particular cargo of which the residue formed a part, or for tankers delivered after
              December 31, 1979, 1/30,000 of the total quantity of the particular cargo of which
              the residue formed a part.
           o  The tanker has in operation an oil  discharge monitoring and control system and a
              slop tank arrangement (under Regulations 29 and 31). (Annex 1, Regulation 34).

       Discharges of oil or oily mixtures from the cargo area of an oil  tanker while in a special
area are prohibited (Annex 1, Regulation 34).

       Discharging oil or oily mixtures from any ship in the Antarctic area is expressly
prohibited. No discharge into the sea may contain substances in quantities or concentrations that
are hazardous to the marine environment or substances introduced for the  purpose of
circumventing the conditions of discharge specified in Annex 1 (Annex 1, Regulation 15).
                                          424

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

       The prohibition against the discharge of oil and oily mixtures does not apply where the
discharge is necessary for the purpose of securing the safety of a ship or saving life at sea. The
prohibition also does not apply where the discharge resulted from damage to the ship or its
equipment, provided that all reasonable precautions were taken after the occurrence of the
damage or discovery of the discharge and the damage was not caused intentionally or recklessly
with knowledge that damage would probably result. Ships may discharge substances containing
oil when those substances are being used to combat specific pollution incidents in an effort to
minimize damage from the pollution, subject to relevant governments'  approvals (Annex I,
Regulation 4).

       Every oil tanker of 150 gt and above  and every other ship of 400 gt and above must
maintain an Oil Record Book Part I.2 The Oil Record Book Part I must be completed whenever
any of the following machinery-space events occur: ballasting or cleaning of oil fuel tanks;
discharge of dirty ballast water or cleaning water from oil fuel tanks; collection and disposal of
oil residues; discharge overboard or disposal otherwise of bilgewater that has accumulated in
machinery spaces; bunkering  of fuel or bulk  lubricating oil; accidental or other exceptional
discharge of oil; and failure of oil filtering equipment. The Oil Record Book Part I must be
readily available for inspection. A Party  to Annex I may request inspection of the  Oil Record
Book Part I while any ship to  which this Annex applies is in its port or  offshore terminal and
require the master of the ship  to certify that any copies made of the  Oil  Record Book Part I are
true. (Annex I, Regulation 17).

       Oil tankers of 150 gt and above and all other ships of 400 gt and above must carry
onboard a shipboard oil pollution emergency plan approved by the government under whose
authority the tanker is operating. The plan must include the procedures  for ship operators to
follow to report an oil pollution incident, the list of authorities or people to be contacted in the
event of an oil pollution incident, a detailed description of the actions to be taken immediately to
reduce the discharges of oil following  an incident, and a contact onboard responsible for
coordinating with authorities to combat the pollution. This plan may be combined with the
emergency response plan required by MARPOL Annex II (discussed below). Oil tankers of
5,000 tons deadweight or more must have prompt access to computerized damage stability and
residual structural strength calculation programs (Annex I, Regulation 37).

       Governments of Parties to Annex I must ensure that there are adequate reception facilities
for discharging oil and oily residues and comply with various requirements related thereto,
including capacity and location requirements (Annex I, Regulation 38).
: Oil tankers must also maintain an Oil Record Book Part II (Annex I, Regulation 36).
                                         425

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

       Although ballast water falls outside the scope of P.L. 110-299, the Agency notes that
Annex I includes regulations governing ballast water.  These regulations establish when ships
must have segregated ballast tanks and under what circumstances ballast water may be carried in
oil fuel tanks or cargo tanks (Annex I, Regulations 16 and 18).

       In addition to the requirements discussed above, Annex I includes a number of
requirements applicable to oil tankers alone. Since oil tankers would not generally be expected to
be study vessels, EPA has omitted an in-depth discussion of these requirements, which include:

    1.  New-build protective cargo tank arrangements (including double-hull/double-bottom
       requirements) for certain tankers (Annex I, Regulations 19-20).
    2.  Double-bottom pump-room requirements for oil tankers of 5,000 tons deadweight and
       above constructed on or after January 1, 2007 (Annex I, Regulation 22).
    3.  Requirement that oil tankers delivered on or after January 1, 2010, be built in such a way
       that if they are damaged, oil will not spill from them at a rate greater than  MARPOL
       allows (Annex I, Regulations 23-25).
    4.  Limitations on the size and arrangement of cargo tanks for oil tankers of 150 gt and
       above, depending on delivery  date (Annex I, Regulation 26).
    5.  Subdivision, damage stability, and intact stability criteria (Annex I, Regulations 27-28).
    6.  Cargo tank cleaning requirements, including requirements relating to slop tanks (Annex I,
       Regulation 29).
    7.  Pumping, piping, and discharge arrangement regulations governing the discharge of dirty
       ballast water or oil-contaminated water (Annex I, Regulation 30).
    8.  Oil discharge monitoring and control system requirements, including requirements for
       effective government-approved oil/water interface detectors (Annex I, Regulations 31-
       32).

       Also outside the  scope of this  study, but worth noting, is that Annex I includes
requirements applicable to fixed or floating platforms. Specifically, fixed or floating platforms
must comply with the requirements of the Annex applicable to ships of 400 gt and above, other
than oil tankers, except that they shall be equipped only to the extent practicable relating to tanks
for oily residue and oil filtering equipment. Records involving oil or oily mixture discharges
must be kept in a form approved by the government under whose authority the vessel is
operating, and the discharge of oil or oily mixtures to the sea is prohibited except when the oil
content of the discharge without dilution does not exceed 15 ppm (Annex I, Regulation 39).
                                         426

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

       6.1.1.2    MARPOL Annex II: Control of Pollution by Noxious Liquid Substances
                 in Bulk

       MARPOL Annex II addresses pollution caused by "noxious liquid substances" (NLS)
carried in bulk. Substances regulated as NLS under MARPOL are categorized into four
categories3 based on their potential to cause harm:

        •   Category X:       Substances that, if discharged into the sea from tank cleaning or
                              deballasting operations, present a major hazard to either marine
                              resources or human health and therefore justify the prohibition
                              of the discharge into the marine environment.
        •   Category Y:       Substances that, if discharged into the sea from tank cleaning or
                              deballasting operations, present a hazard to either marine
                              resources or human health or cause harm to amenities or other
                              legitimate uses of the sea and therefore justify a limitation on the
                              quality and quantity of the discharge into the marine
                              environment.
        •   Category Z:       Substances that, if discharged into the sea from tank cleaning or
                              deballasting operations, will present a minor hazard to either
                              marine resources or human health and therefore justify less
                              stringent restrictions on the quality and quantity of the discharge
                              into the marine environment.
        •   Other Substances:  Substances that fall outside of categories X, Y, or Z because
                              they are considered to present no harm to marine resources,
                              human health, amenities, or other legitimate uses of the sea
                              when discharged into the sea from tank cleaning or deballasting
                              operations. The discharge of bilge or ballast water or other
                              residues or mixtures containing these substances are not subject
                              to any requirements under MARPOL Annex II. (Annex II,
                              Regulation 6).

       All ships certified to carry one or more of these  substances in bulk must follow the
requirements established in Annex II unless the discharge is necessary for the purpose of
securing the safety of a ship or saving life at sea (Annex II, Regulations 2-3). The Annex's
requirements also do not apply where the discharge resulted from damage to the ship or its
equipment, provided that reasonable precautions were taken after the occurrence of the damage
or discovery of the discharge, and the damage was not caused intentionally or recklessly with
knowledge that damage would probably result. Discharges of other substances may also be
exempted from Annex IFs requirements if they are government-approved (by both the
government under whose authority the ship is operating and any government in whose
3 This categorization scheme was developed when Annex II was revised; it entered into force in January 2007. The
United States Coast Guard's implementing regulations, discussed below, have not yet been revised to reflect this
new scheme.
                                         427

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

jurisdiction the discharge will occur) and being used to combat specific pollution incidents in an
effort to minimize damage from the pollution (Annex II, Regulation 3). Regulation 4 of Annex II
provides for a number of other specific exemptions to the Annex's requirements.

       Ships intending to carry NLS in bulk to other Parties to MARPOL must obtain an
International Pollution Prevention Certificate for the Carriage of Noxious Liquid  Substances in
Bulk ("Certificate"). The Certificate records the results of the various inspections to which NLS-
carrying ships are subject. The government under which the ship is registered is typically
responsible for issuing the Certificate, using a form provided in Appendix 3 to Annex II (Annex
II, Regulation 9). Certificates are  issued for a period of time not to exceed five years (Annex II,
Regulation 10).

       Prior to and at periodic intervals after a ship is issued a Certificate, it is subject to a
complete inspection of its structure, equipment, systems, fittings, arrangements, and materials to
ensure compliance with Annex II. The government of the country under whose authority a ship is
operating is responsible for having these inspections conducted. If a ship or its equipment is
found to not correspond substantially with the particulars of the Certificate,  corrective action
must be taken. If corrective action is not taken, the ship's Certificate should be withdrawn.
Conformity with these MARPOL requirements is necessary to ensure that the ship does not pose
an unreasonable threat of harm to the marine environment (Annex II, Regulation  8).

       Ships that are certified to carry NLS in bulk that are identified in chapter 17 of the
International Bulk Chemical Code must generally ensure that their design, construction,
equipment, and operation  are in conformance with the requirements of that Code  (Annex II,
Regulation 11).

       Ships constructed prior to July 1, 1986, must have a pumping and piping arrangement
ensuring that each tank certified to carry substances in Category X or Y does not  retain more
than 300 liters of residue in the tank and its associated piping. Each tank certified to carry
substances in Category Z must not retain more than 900 liters in the tank and its associated
piping (Annex II, Regulation 12(1)).  Ships constructed on or after July 1,  1986, but before
January 1, 2007, must not retain residue greater than 100 liters for Category X or Y substances or
300 liters for Category Z substances in the tank and its associated piping (Annex II, Regulation
12(2)). Ships constructed after January 1, 2007, must not retain residue in a quantity greater than
75 liters in the tank or its associated piping for Category X, Y, or Z (Annex II, Regulation 12(3)).

       Ships certified to carry Category X, Y, or Z substances, except ships constructed before
January 1, 2007, and certified to carry Category Z substances, must have at  least one underwater
discharge outlet, which must be located within the cargo area in the vicinity of the turn of the
bilge and arranged to avoid the re-intake of residue/water mixtures by  the ship's seawater
intakes. The residue/water mixture discharged into the sea must not pass through  the ship's
boundary layer (Annex II, Regulation 12 (6)-(9)).
                                          428

-------
                           Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

       Ships are prohibited from discharging into the sea residues of Category X, Y, or Z
substances or ballast water, tank washings, or other mixtures containing these substances unless
the discharges fully comply with the applicable operational requirements of Annex II.
Specifically, 1) the ship must be proceeding en route at a speed of at least 7 knots in the case of
self-propelled ships or at least 4 knots for other ships, 2) the discharge must be made below the
waterline through the underwater discharge outlets at a rate not to exceed what the outlet was
designed for, and 3) the discharge must be made no less than 12 nautical miles from the nearest
land and in water not less that 25 meters deep (Annex II, Regulation 13(l)-(2)). For Category Z
substances on ships not required to have an underwater discharge outlet, the requirement that
discharges occur below the waterline does not apply. Annex II also sets out requirements for the
discharge  of NLS residues (Annex II, Regulation 13(6)-(7)). Any discharge of NLS or mixtures
into the Antarctic area is prohibited (Annex II, Regulation 13(8)).

       Every ship certified to carry Category X, Y, or Z substances must have a government
approved Manual onboard. The Manual is meant to inform the ship's officers of the physical
arrangements and operational procedures necessary to comply with Annex II (Annex II,
Regulation 14).  Ships must also carry with them a Cargo Record Book to record where NLS
substances were loaded and unloaded and the circumstances of the loading and unloading. If any
accidental or emergency discharges occur, those must also be recorded in the Cargo Record
Book (Annex II, Regulation 15).

       Ships certified to carry NLS in bulk that weigh 150 gt or above must carry onboard a
marine pollution emergency plan for NLS. The plan must be government approved and must
include the procedures for ship operators to follow to report an NLS pollution incident, the list of
authorities and people to be contacted in the event of an NLS pollution incident, a detailed
description of the actions to be taken immediately to reduce the discharges of NLS following an
incident, and a contact onboard responsible for coordinating with authorities to combat the
pollution (Annex II, Regulation 17).

       The Government of each Party to MARPOL must ensure that its ports and terminals have
adequate NLS reception facilities for the ships utilizing those ports and terminals to meet the
requirements of Annex II (Annex II, Regulation 18).

       6.1.1.3     MARPOL Annex III: Prevention of Pollution by Harmful Substances
                 Carried by Sea in Packaged Form

       MARPOL Annex III establishes requirements for preventing pollution caused by harmful
substances that are carried in packaged form. "Harmful substances" are defined as those
substances that are identified as marine pollutants in the International Maritime Dangerous
Goods Code (EVIDG Code). "Packaged form" is defined as the forms of containment specified
for harmful substances in the IMDG Code (Annex III, Regulation 1(1)). Although the
requirements of this Annex do not directly regulate discharges incidental to the normal operation
                                         429

-------
                           Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

of a vessel, they play a critical role in preventing harmful substances from entering into such
discharge streams.

       Annex III prohibits the carriage of harmful substances from all ships unless the
requirements of the Annex are followed (Annex III, Regulation 1(2)). Empty packages that were
previously used to carry harmful substances and contain harmful residue are themselves
considered harmful substances and must be treated as such (Annex III, Regulation 1(4)).
Additionally, the Government of each Party to MARPOL is required to issue detailed
requirements on packing, marking, labeling, documentation, stowage, quantity limitations, and
exceptions for preventing or minimizing pollution of the marine environment by harmful
substances (Annex III, Regulation 1(3)).

       The Annex requires that packages containing harmful substances be adequate to
minimize the hazard to the marine environment, having regard to their specific contents (Annex
III, Regulation 2). They must be durably marked with the correct technical name (trade names
alone are prohibited), must indicate that the substance is a marine pollutant, and should be
supplemented where possible by other means (e.g., use of the relevant United Nations  number).
The durability of both the package and the markings must be considered because the Annex
requires that the markings be able to withstand at least three months immersed in the sea (Annex
III, Regulation 3).

       In all documents relating to the carriage of harmful substances at sea, the correct
technical name of each substance must be used, and the substance  must be identified with the
words "MARINE POLLUTANT." The  shipping documents provided by the shipper must be
accompanied by a signed certificate declaring that the shipment is  properly packaged and marked
and in proper condition for carriage to minimize the hazard to the marine environment. Every
ship must keep, both onboard and onshore, a list or manifest detailing the harmful substances
onboard and where they are stowed (Annex III, Regulation 4).

       Packages containing harmful substances must be stowed and  secured so as to minimize
the hazards to the marine environment, without impairing the safety of the ship and the people
onboard (Annex III, Regulation 5). Some harmful substances may  face restrictions, for sound
scientific and technical reasons, as to the quantity that can be carried onboard, and in some cases,
carrying them might be prohibited altogether. These determinations will take into account the
size, construction, and equipment of the ship, as well as the packaging and nature of the
substance (Annex III, Regulation 6).

       Except where necessary to protect the ship or saving life at sea, the jettisoning  of harmful
substances carried in packaged form is prohibited (Annex III, Regulation 7).
                                         430

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

       6.1.1.4     MARPOL Annex IV: Prevention of Pollution by Sewage from Ships

       Annex IV of MARPOL establishes requirements for the prevention of pollution caused
by sewage from ships. The discussion of discharges of sewage from vessels was specifically
excluded from the scope of this study; therefore, the summary of this section is omitted. See P.L.
110-299,  § 3(c)(2). It should also be noted that, as mentioned above, the United States is not a
party to Annex IV and is therefore not obligated to follow its requirements.

       6.1.1.5     MARPOL Annex V: Prevention of Pollution by Garbage from Ships

       Annex V of MARPOL regulates garbage pollution from ships. Under the Annex,
"Garbage" is defined as all kinds of victual, domestic, and operational waste (excluding fresh
fish and fish parts) generated during the normal operation of the ship and liable to be disposed of
continuously or periodically (Annex V, Regulation 1(1)). Although the requirements of this
Annex do not directly regulate discharges subject to this report ("garbage" is not subject to the
former NPDES permit exclusion at 40 CFR 122.3(a)), they play a critical role in preventing
garbage from entering into and contaminating discharge streams subject to this report.

       The Annex establishes different disposal requirements depending on the type of garbage
being disposed of. Disposal into the sea of dunnage—lining and packing materials that will
float—is prohibited if the ship is closer than 25 nautical miles from the nearest land. The
disposal of food wastes and all other garbage, including paper products, rags, glass, metal,
bottles, crockery, and similar refuse is prohibited less than 12 nautical miles from the nearest
land; however, it may be permitted if it has passed through a comminuter or grinder, is small
enough that it can pass through a screen with openings no greater than 25 mm, and is disposed of
as far as practicable from the nearest land (but no closer than 3 nautical miles). The disposal of
plastics, including but not limited to synthetic ropes,  synthetic fishing nets, and plastic garbage
bags, is prohibited. Where garbage is mixed, the more stringent requirements will apply (Annex
V, Regulation 3). Additional  special requirements are in place for discharges into certain defined
areas.4

       None of the disposal regulations described above apply where: 1) the disposal was
necessary for the purpose of securing the safety of the ship or those onboard or saving life at sea;
2) the garbage escaped as the result of damage to the ship (provided all reasonable precautions
were taken before and after the incident to prevent or minimize the escape); or 3) disposal was
the result of an accidental loss of synthetic fishing nets (provided that all reasonable precautions
were taken to prevent the loss) (Annex V, Regulation 6).
4 For the purposes of Annex V, the special areas are the Mediterranean Sea area; the Baltic Sea area; the Black Sea
area; the Red Sea area; the "Gulfs" area; the North Sea area: the Antarctic area; and the wider Caribbean region,
including the Gulf of Mexico and the Caribbean Sea (although the rules have not entered into force with respect to
all of these areas yet). For the specific requirements, see Annex V, Regulation 5.

                                          431

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

       The Parties to the Annex must ensure that ports and terminals have adequate facilities for
the reception of garbage (Annex V, Regulation 7).

       Each ship 12 meters or more in length must display placards that notify those onboard of
the various disposal requirements. The placards must be written in the working language of the
ship's personnel and, for ships engaged in voyages to ports or offshore terminals under the
jurisdiction of other Parties to the Convention, shall also be in English, French, or Spanish
(Annex V, Regulation 9(1)).

       Every ship 400 gt and above and every ship certified to carry 15 or more people must
carry a garbage management plan for the crew to follow.  The plan must describe procedures  for
collecting, storing, processing, and disposing of garbage,  including the use of equipment
onboard. It must be written in the working language of the crew and identify the person in charge
of carrying out the plan (Annex V, Regulation 9(2)). Ships of this size or certification that travel
to ports or offshore terminals under the jurisdiction of other countries party to MARPOL, and
every fixed and floating platform engaged in  exploration  and exploitation of the seabed, must
also carry a Garbage Record Book onboard. The Garbage Record Book must include a record of
every discharge operation or incineration, including the date and time of the discharge, the
position of the ship, a  description of the garbage, and the  estimated amount discharged or
incinerated. Any escapes or accidental losses must also be noted in the Garbage Record Book,
along with a description of the circumstances of the loss (Annex V, Regulation 9(3)).

       6.1.1.6    MARPOL Annex VI: Prevention of Air Pollution from Ships

       Annex VI of MARPOL regulates air emissions from ships. Air emissions are outside  the
scope of this study, therefore, the summary of this Annex has been omitted.

       6.1.1.7    MARPOL Summary

       The earlier chapters of this study describe a number of pollutants detected by EPA in
incidental vessel discharges that have the potential to pose a risk to human health or the
environment. Of these pollutants of concern,  it appears that oil and grease are the only pollutants
found in incidental discharges that would be directly regulated by MARPOL, through Annex I.
However, MARPOL may indirectly regulate  other pollutants found in incidental  discharges,  such
as metals, to the extent that they are found in any of the noxious liquid substances categorized
under Annex II or garbage under Annex V and prevented from entering incidental discharge
streams . In all cases, the requirements of MARPOL only apply to those vessels that are large
enough to meet the size thresholds established in the treaty.

6.1.2  The International Convention on the Control of Harmful Anti-Fouling Systems on
       Ships

       The International Convention  on the Control of Harmful Anti-Fouling Systems on Ships
was adopted by the IMO on October 5, 2001, and entered into force on September 17, 2008.  The
                                         432

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

U.S. Senate gave its consent to ratify the Convention on September 26, 2008; however, the
United States will not deposit its instrument of ratification with the IMO until Congress adopts
the necessary implementing legislation. Implementing legislation was introduced on September,
24, 2009. See Clean Hull Act of 2009, H.R. 3618, lllth Congress (1st Session 2009). If passed,
this new legislation would replace the Organotin Anti-Fouling Paint Control Act of 1988
(OAPC), discussed below.

       Parties to the Convention are required to take steps to reduce or eliminate adverse effects
on the marine environment and human health caused by antifouling systems. Under the
Convention, an "antifouling system" is any coating, paint, surface treatment, surface, or device
used on a ship to control or prevent the attachment of unwanted organisms (Article 2(2)).

       The Convention applies to any ship entitled to fly the flag of a Party; ships not entitled to
fly the flag of a Party but that operate under the authority of that Party; and ships that enter a
port, shipyard, or offshore terminal of a Party but do not fall under one of the earlier categories.
Warships, naval auxiliary, or other ships owned or operated by a Party are exempted when used
only for noncommercial government service. However, each Party must ensure that these
exempted ships operate in a manner consistent with the Convention, where reasonable and
practicable. Parties must also ensure that favorable treatment is not given to ships registered to
countries that are not Parties to the Convention (Article 3).

       Under the Convention, Parties must prohibit and/or  restrict the application, re-
application, installation, or use of environmentally harmful  antifouling systems on ships
registered under them, as well as on ships that enter its ports, shipyards, or offshore terminals
(Article 4). As of January  1, 2008, ships are prohibited from having any organotin compounds on
their hulls that act as biocides, unless the compounds have been sealed so that no leaching occurs
(Annex I).

       Parties to the Convention must take measures to require that wastes generated by the
application or removal  of an antifouling system are collected, handled, treated,  and disposed of
in a safe and environmentally sound manner (Article 5). In  the United States., this provision
would be implemented through the Solid Waste Disposal Act,  33 U.S.C. §§ 6901-6992, and the
Clean Water Act, 33 U.S.C. §§ 1251-1387.

       Any Party can propose an amendment to the Convention, including proposals to prohibit
antifouling systems other than organotins. The process for proposing an amendment, and
subsequently considering and adopting it, is described in Articles 6, 7, and 16.

       Parties must take appropriate measures to promote and facilitate scientific and technical
research on the effects of antifouling systems, as well as monitoring these effects. The research
should include observation, measurement,  sampling, evaluation, and analysis of the effects of
                                         433

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

antifouling systems. Parties should share the information learned in these studies with other
Parties to the Convention when requested (Article 8).

       The Convention requires Parties to report to the IMO a list of all surveyors and
organizations that are authorized to act on behalf of that Party in administration of matters
relating to the control of anti-fouling systems. Parties must also annually report information
regarding any antifouling systems that were approved, restricted, or prohibited under domestic
law. For antifouling systems that were approved, registered, or licensed by a Party, that Party
must provide to other Parties upon request relevant information on which that decision was made
(alternatively, a Party could require the manufacturers of approved, registered, or licensed
antifouling systems to provide this information) (Article 9).

       A Party must ensure that ships entitled to fly under its flag or operate under its authority
are surveyed and certified in accordance with the requirements of Annex 4 (Article 10). Annex 4
requires that ships of 400 gt and above that are engaged in international voyages be surveyed
before the ship is put into service and whenever the antifouling systems are changed or replaced.
The survey is intended to ensure the ship's  antifouling system fully complies with the
Convention (Annex 4, Regulation 1). At the conclusion of the survey, the ship will be issued an
International Anti-Fouling  System Certificate (Annex 4, Regulation 2). Ships less than 400 gt
and 24 meters or more in length and that are engaged in international voyages must carry a
Declaration, signed by the owner or his  agent, declaring that the antifouling system used on the
ship complies with the requirements of the  Convention (Annex 4, Regulation 5).

       Ships to which the Convention applies may be inspected in any port, shipyard, or
offshore terminal of a Party. Unless there are clear grounds for believing that a ship is in
violation of the Convention, the inspection  is limited to: 1) verifying that, where required, there
is a valid International Anti-Fouling System Certificate or Declaration onboard; and/or 2) a brief
sampling of the ship's antifouling system, taking into account IMO guidelines. If there are clear
grounds to believe that a ship is in violation of the Convention, a more thorough inspection is
permitted, taking into account IMO guidelines.  Additionally, a Party may take steps to warn,
detain, dismiss, or exclude from its ports any  ship that is found to be in violation but must
immediately notify  the country under whose flag the ship is registered (Article 11).

       Parties must, through domestic laws, prohibit violations of the Convention and establish
sanctions severe enough to discourage violations.5 If a violation occurs within the jurisdiction of
a Party, that Party must either cause proceedings to be taken in accordance with its domestic
laws or furnish any  information or evidence it has showing a violation has occurred to the
government under whose authority the ship concerned is operating. If that government finds the
5 For vessels larger than 79 feet, EPA has prohibited the discharge of tributyltin and other organotins under the
Agency's Vessel General Permit (see Section 6.2.3).
                                          434

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

evidence sufficient to enable proceedings to be brought, it must do so as soon as possible, in
accordance with its laws, and notify both the IMO and the reporting Party that it has done so. If
the government does not take action within one year after receiving the information, it must so
inform the Party that reported the alleged violation (Article 12).

       Parties must make every effort to avoid unduly detaining or delaying ships while
conducting inspections or investigating potential violations. If a ship is unduly detained or
delayed, it is entitled to compensation for any loss or damage suffered (Article 13).

       The Convention does not prejudice the rights or obligations of any country under
customary international law as reflected in the United Nations Convention on the Law of the Sea
(Article 15).

       Presently, the International Convention on the Control of Harmful Anti-Fouling Systems
on Ships only regulates the use of organotin tributyling (TBT) in antifouling coatings. Effective
January 2003, new applications of antifouling coatings containing TBT were prohibited by the
treaty, and as of January 2008, all vessels with an existing TBT antifouling coating  on their hulls
are required to apply a protective coating over the TBT to prevent leaching.

       Since the use of TBT has been prohibited, vessel operators have turned to anti-fouling
systems that contain other potentially harmful pollutants, such as copper. Copper is not currently
regulated under the International Convention on the Control of Harmful  Anti-Fouling Systems
on Ships; however, the treaty provides a  system whereby Parties may propose that a specific
anti-fouling system be regulated under the treaty. Through this mechanism,  copper may one day
be regulated under the treaty if parties to the treaty determine it is necessary.

6.1.3   International Convention for the Safety of Life at Sea (SOLAS)

       The International Convention for the Safety of Life at Sea (SOLAS) is considered the
most important international treaty concerning the safety of merchant ships. The first version was
adopted in 1914 in response to the Titantic disaster and has been amended many times since
then, most recently in 1974. The primary objective of SOLAS is to establish minimum standards
for the construction, equipment, and operation of ships, in consideration of their safety. The
responsibility for ensuring compliance rests with the individual flag  states, although contracting
governments do have limited authority to inspect ships of other contracting governments if there
are clear grounds for believing the SOLAS requirements are not being met. For additional
information on SOLAS, please see the IMO's discussion of the Convention at www.imo.org

       While SOLAS does not directly regulate vessel discharges, it does provide environmental
benefits through its regulations and through adoption of the International Safety Management
(ISM) Code, all of which assist in preventing spills and other accidental  discharges. The ISM
Code provides an international standard for safely managing and operating ships and for
                                         435

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

preventing pollution. In addition to other requirements, under the Code, companies or individuals
responsible for operating vessels must establish a safety and environmental-protection policy and
ensure that the policy is implemented and maintained at all levels of the organization, both ship-
based and shore-based. These operators must also create a safety management system, which is a
structured and documented system that enables company  personnel to effectively implement the
company's safety and environmental protection policy (ISM Code, Part A).

       SOLAS could be used to address any of the pollutants of potential concern identified by
EPA though this study, to the extent that the individual policies adopted by vessel operators
address specific pollutants found in incidental discharges.

6.1.4   Boundary Waters Treaty

       The Boundary Waters Treaty is an agreement the United States and Canada entered into
in 1919 to govern the management of boundary waters. Among other things, the treaty provides
that "boundary waters" - defined as "waters from main shore  to main shore of the lakes and
rivers and connecting waterways, or the portion thereof, along with the international boundary"
between the U.S. and Canada - "and waters flowing across the boundary shall not be polluted on
either side to the injury of health or property on the other" (Preliminary Article and Article IV.2).

       The Treaty established the International Joint Commission (IJC), composed of three
commissioners from each country, to assist in the resolution of boundary water issues (Article
III). Since 1919, the IJC has addressed a variety of water-use and water-quality issues. The
Treaty is a foundational backdrop  for other bilateral agreements between the United States and
Canada, such as the Great Lakes Water Quality Agreement.

       As a foundational agreement, the Boundary Waters Treaty does not directly regulate
specific pollutants, which means it does not directly regulate specific pollutants in incidental
discharges.

6.1.5   Great Lakes Water Quality Agreement

       The Great Lakes Water Quality Agreement, first signed in 1972, and revised in 1978 and
1987, expresses the commitment of both the United States and Canada to restore and maintain
the chemical, physical, and biological integrity of the waters of Great Lakes Basin Ecosystem. It
also reaffirms the rights and obligations of both countries under the Boundary Waters Treaty.
The Great Lakes Water Quality Agreement is primarily implemented through Section 118 of the
Clean Water Act.

       One of the stated policies in the Agreement is the  prohibition of discharges of toxic
substances in toxic amounts and the virtual elimination of discharges containing any or all
persistent toxic substances (Article II). The general objectives of the agreement are to ensure that
the waters in the Great Lakes  System are free from pollutants resulting from human activity,
                                         436

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

such as substances that will settle to form sludge deposits or harm aquatic life or waterfowl;
floating materials (e.g., debris, oil, scum, other immiscible substances); materials or heat that
produce color, odor, taste, or other conditions that will interfere with beneficial uses or are toxic
or harmful to human health or the environment; and nutrients that create growths of aquatic life
that interfere with beneficial uses (Article III).

       Vessel discharges are directly addressed through Annexes 4 (discharges of oil and
hazardous polluting substances from vessels), 5 (discharges of vessel wastes), and 6 (review of
pollution from shipping sources) of the Agreement. In all of these annexes, "vessel" is defined as
"any ship, barge or other floating craft, whether or not self-propelled" (Annex 4(l)(e), Annex
       Annex 4 addresses discharges of oil and hazardous polluting substances from vessels.
Within this annex, the term "discharge" includes, but is not limited to, any spilling, leaking,
pumping, pouring, emitting, or dumping; it does not include unavoidable direct discharges of oil
from a properly functioning vessel engine (Annex 4(l)(a)). The annex requires that each country
adopt regulations to prevent discharges of harmful quantities of oil and hazardous substances
from vessels into the Great Lakes System. Specifically:

        •   Discharges of harmful quantities of oil or hazardous substances, including those
            contained in ballast water, must be prohibited and made subject to appropriate
            penalties.
        •   As soon as any person in charge, including a vessel owner/operator, becomes aware
            of a discharge, or probable discharge, of harmful quantities of oil or hazardous
            substances, he/she must immediately notify the appropriate agency in the
            jurisdiction where the discharge occurred. Failure to give this notice must be subject
            to appropriate penalties (Annex 4(2)).

       A "harmful quantity of oil" is defined as "any quantity of oil that, if discharged from a
ship that is stationary into clear calm water on a clear day, would produce a film or a sheen upon,
or discoloration of, the surface of the water or adjoining shoreline, or would cause a sludge or
emission to be deposited beneath the surface of the water or upon the adjoining shoreline"
(Annex 4(1)).

       Annex 4 also requires both countries to adopt regulations for the design, construction,
and operation of vessels, as well as programs to  ensure that merchant vessel personnel are trained
in the use, handling, and stowage of oil and abatement of oil pollution, thereby preventing the
discharge of harmful quantities of oil or hazardous polluting substances. For oil, the regulations
must ensure that each vessel has a suitable means for containing spills of oil and oily wastes and
retaining those wastes onboard for off-load at a reception facility. Oil loading, unloading, and
bunkering systems must be suitably designed to  minimize the possibility of failure (Annex 4(3)).
                                          437

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

       For hazardous polluting substances, each country must adopt programs and measures to
prevent discharges of harmful quantities of hazardous polluting substances carried as cargo. Such
regulations include ensuring that all vessels have a suitable means of containing onboard spills
caused by loading or unloading operations and have the capability of retaining onboard wastes
accumulated during vessel operation for off-loading to a reception facility. The regulations must
also provide for the identification of vessels carrying cargos of hazardous substances and for the
identification in vessel manifests of all the hazardous substances those vessels are carrying
(Annex 4(4)). A list of hazardous polluting substances and potential hazardous polluting
substances can be found in Appendices 1 and 2 to  Annex 10.

       Additionally, under Annex 4, both countries must ensure that there are adequate facilities
for the reception, treatment, and subsequent disposal of oil and hazardous polluting substances
from all vessels (Annex 4(5)).

       Annex 5 addresses discharges of vessel wastes, including garbage, sewage, and waste
water. "Garbage" is defined as "all kinds of victual, domestic, and operational wastes, excluding
fresh fish and parts thereof generated during the normal operation of the ship and liable to be
disposed of continually or periodically." "Wastewater" encompasses any water combined with
other substances, "including ballast water and water used for washing cargo hold, but excluding
water in combination with oil, hazardous polluting substances, or sewage" (Annex 5(1)).

       The agreement requires both countries to adopt regulations that will:

        •   Prohibit the discharge of garbage from vessels and make such discharges subject to
            appropriate penalties.
        •   Prohibit the discharge of wastewater in harmful amounts or concentrations and make
            such discharges subject to appropriate penalties.
        •   Ensure that each vessel operating in boundary waters, and that has toilet facilities, is
            equipped with a device to contain, incinerate, or treat sewage to an adequate degree.
            Appropriate penalties must be provided for failure to comply (Annex 5(2)).

       Within the Great Lakes System, certain critical use areas may be designated where the
discharge of wastewater or sewage will be limited or prohibited (Annex 5(3)). Both countries
must take measures to ensure there are adequate facilities for the reception,  treatment, and
subsequent disposal of garbage, wastewater, and sewage from vessels (Annex 5(5)).

       Annex 6 calls on both the Canadian and U.S. Coast Guards to review "services, systems,
programs, recommendations, standards, and regulations relating to shipping activities for the
purpose of maintaining or improving Great Lakes  water quality" (Annex 6(1)). The two Coast
Guards must meet at least annually to consult on implementing the Agreement (Annex 6(2)).

       Of the pollutants of potential concern identified by EPA earlier in this study, oil and
grease  are the pollutants most directly addressed under the Great Lakes Water Quality
                                         438

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

Agreement. The Agreement also addresses wastewater, which may include some incidental
discharges. Under the Clean Water Act, the Great Lakes National Program Office is tasked with
developing and implementing specific action plans to carry out the responsibilities of the U.S.
under the Great Lakes Water Quality Agreement. (33 U.S.C. § 1268(c)(l)(A)). EPA might be
able to address incidental discharges through these action plans to the extent those discharges are
"wastewater" as that term is defined in the Agreement.

Annex I lists a number of chemicals and pollutants that are  specific objectives of the agreement,
including metals such as arsenic, cadmium, chromium, copper, and nickel, among others. Annex
I also provides standards for the concentration of total dissolved solids, hydrogen sulfide,
phosphorus, and other pollutants in  the Great Lakes.

6.1.6   St. Lawrence Seaway Regulations

       In 1954, the United States statutorily created the Saint Lawrence Seaway Development
Corporation to construct, operate, and maintain the section of the St. Lawrence Seaway between
the Port of Montreal and Lake Erie that falls within the territorial limits of the United States (33
U.S.C. § 981). The mission of the wholly government-owned corporation, which is under the
direction and supervision of the Department of Transportation, is to improve the operation and
maintenance of a deep-draft waterway in cooperation with a Canadian counterpart.6

       The Department of Transportation's regulations governing the Seaway can be found at 33
CFR Part 401. The regulations define the St. Lawrence Seaway as the "the deep waterway
between the Port of Montreal and Lake Erie and including] all locks, canals and connecting and
contiguous waters that are part of the deep waterway, and all other canals and works, wherever
located, the management, administration and control of which have been entrusted to the
Corporation or the Manager" (33 CFR § 401.2(j)).

       While the regulations are primarily geared toward maintaining and using the Seaway,
they do include provisions designed to lessen the impacts of vessel pollution to the Great Lakes,
including a provision that prohibits  the discharge of garbage, ashes, ordure, litter, or other
materials into the Seaway (33 CFR  § 401.59(d)). The regulations also prohibit any vessel from
emitting sparks or excessive smoke, or from blowing boiler tubes (33 CFR § 401.59(a)). In
addition, the regulations contain a blanket requirement that no discharge is allowed that is not in
6 In addition to the authorities under its enabling statute, 33 U.S.C. § 981 et seq., the St. Lawrence Seaway
Development Corporation has authority to "operate, maintain, improve or expand vessel traffic services consisting
of measures for controlling or supervising vessel traffic or for protecting navigation and the marine environment"
pursuant to the Ports and Waterways Safety Act of 1978, at 33 U.S.C. 1223-1225, 1229. The U.S. Coast Guard has
this authority in all other navigable waters of the United States, except for the area under the jurisdiction of the
Corporation.
                                          439

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

conformity with all applicable U.S. and Canadian regulations, except within certain areas of the
Welland Canal, where no discharges are allowed at all (33 CFR § 401.59(b)).

       Although ballast water is not a focus of this study, it should be noted that the St.
Lawrence Seawater Regulations also include provisions relating to ballast water, including a
recently passed regulation requiring all oceangoing vessels entering the Seaway to conduct
saltwater flushing  (Seaway Regulations and Rules: Periodic Update, Various Categories, 73 FR
9950 (February 25, 2008)).

       The St. Lawrence Seaway Regulations only regulate specific pollutants to the extent they
are found in ballast water.

6.2   FEDERAL LAWS

6.2.1   Act to Prevent Pollution from Ships (APPS)

       The Act to Prevent Pollution from Ships (APPS) is the United States law implementing
Annexes I, II, V, and VI of MARPOL (Annex III is implemented through the Hazardous
Materials Transportation Act). The U.S. Coast Guard has the primary authority to implement and
enforce the majority of provisions within APPS. EPA was also given specific authorities in
certain sections of APPS, the most extensive  of which relate to MARPOL Annex VI. The Coast
Guard's implementing regulations, found at 33 CFR Part 151, are addressed below.

       APPS applies to U.S.-registered ships regardless of where in the world they are operating.
With respect to the implementation of Annexes I and II, APPS additionally applies to all foreign-
flagged ships operating in navigable U.S. waters.  The implementation of Annex V applies to all
U.S.-registered ships, as well as all foreign-flagged ships in navigable U.S. waters or the
exclusive economic zone of the United States (33 U.S.C.  § 1902(a)). Warship, naval auxiliary,
and ships owned by the United States that are engaged in noncommercial service are exempted
from the requirements of APPS, except for certain provisions implementing Annex V.7 Ships
that are specifically exempted from MARPOL, or the Antarctic Protocol, are also exempted from
the requirements of APPS.

       In addition to implementing the requirements of MARPOL, described above, APPS
establishes a number of administrative requirements regarding inspections, penalties for
violations, procedures for legal actions, and public education requirements (33 U.S.C. §§ 1907,
1908, 1910, and 1915).
7 However, all surface ships and submersibles owned or operated by the Department of the Navy are required to
comply with the special area requirements of Annex V. Unique vessels that cannot fully comply with the
requirements of Annex V are permitted to discharge some types of garbage without regard to the requirements of
Annex V. See 33 U.S.C. § 1902(d)(2).


                                         440

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

       6.2.1.1     U.S. Coast Guard Implementing Regulations

       The U.S. Coast Guard implements APPS through its regulations at 33 CFR Part 151.
These regulations apply to every ship required to comply with Annex I, II, or V of MARPOL (33
CFR§ 151.03).8

       6.2.1.1.1   Annex I Implementation—Prevention of Oil Pollution

       The requirements of Annex I of MARPOL, pertaining to the prevention of oil pollution
from ships, are implemented by the U.S. Coast Guard though its regulations at 33 CFR §§
151.09-151.29. This section of the regulations, with the exception of the oil pollution emergency
plan requirements,9 is applicable to ships that are operated under the authority of the United
States and that engage in international voyages, are certificated for ocean service, are certificated
for coastwise service beyond three nautical miles from land, or are operated at any time seaward
of the outmost boundary of the territorial seas of the United States. The regulations also apply to
ships  operated under the authority of another country while in the navigable waters of the United
States or while at a port or terminal under U.S. jurisdiction (33 CFR §  151.09(a)). The
regulations do not apply to warships, naval auxiliary, or other ships owned or operated  by a
country when engaged in noncommercial service; Canadian or U.S.  ships operating  exclusively
on the Great Lakes or their connecting tributary waters or on any internal waters of the  United
States or Canada; or any ships specifically excluded by MARPOL.

       The Coast Guard's requirements  for oil discharges from ships other than oil tankers10  are
very similar to Annex  Fs requirements. The Coast Guard's regulations apply to the same size
ships  regulated under MARPOL; however, the Coast Guard also  distinguishes vessels depending
on how far off shore they are operating:

   •   When more than 12 nautical miles from the nearest land, any discharge of oil or oily
       mixtures must meet the following conditions:
          o  The discharge must not originate from cargo pump  room bilges.
          o  The discharge must not be  mixed with oil cargo residues.
          o  The ship must not be within a special area.
 On December 18, 2009, EPA finalized regulations to implement the air emission requirements of APPS (which
themselves implement MARPOL Annex VI). The final rule is not scheduled to appear in the Federal Register until
the end of February 2010. To see a pre-publication copy of the rule, please visit EPA's website at
http://www.epa.gov/OMS/oceanvessels.htnrfregs.
9 The shipboard oil pollution emergency plan requirements at 33 CFR §§ 151.26-151.29 apply to all U.S.- and
foreign-operated oil tankers of 150 gt and above and all other ships of 400 gt and above. The same exceptions
described in the text apply, with the additional exception that barges or other ships constructed or operated in such a
manner that no oil in any form can be carried aboard are also exempted from the requirements (33 CFR § 151.09(c)-
(d)).
10 The requirements for oil tankers are found in a separate section of the regulations (33 CFR Part 157).

                                           441

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

         o   The ship must be proceeding en route11
         o   The oil content of the effluent without dilution must be less than 15 ppm.
         o   The ship must be operating oily-water separating equipment, a bilge monitor, a
             bilge alarm, or a combination of the three (33 CFR § 151.10(a)).

   •   When within 12 nautical miles from the nearest land, any discharge of oil or oily
       mixtures must meet all of the above requirements, with the additional requirement that
       the oily-water separating equipment be equipped with a U.S. government- or IMO-
       approved 15 ppm bilge alarm (33 CFR § 151.10(b)).

       Ships of 400 gt or above and oil tankers are prohibited from discharging oil or oily
mixtures while operating in a  special area,  as defined in 33 CFR §  151.13(a). However, if the
discharge is of processed bilgewater from machinery space bilges, ships of this size may
discharge in special areas if all of the above requirements are met and the ship is equipped with
an automatic shut-off device that will engage when the oil content of the effluent exceeds 15
ppm (33 CFR § 151.13). Ships of 400 gt or less, other than oil tankers, may discharge in special
areas only if the undiluted oil  content of their effluent is 15 ppm or less. If a ship  cannot meet the
discharge requirements, the oily mixtures must be retained onboard or discharged to a reception
facility (33 CFR § 151.10(f)).

       As with MARPOL, these discharge requirements do not apply where the discharge is
necessary to secure the safety of the ship or save life at sea, or if the discharge results from
damage to the ship (provided  reasonable precautions were taken after the occurrence of the
damage or discovery of the discharge to prevent or minimize the discharge, and the owner or
master of the ship did not act with intent to cause damage, or recklessly and with knowledge that
damage would probably result) (33 CFR §  151.11).

       The regulations also implement the reporting, survey, certification, inspection and
enforcement, recordkeeping, and planning  requirements of Annex I, described above (33 U.S.C.
§§ 151.15, 151.17, 151.19, 151.23, 151.25-151.28).

       6.2.1.1.2  Annex II Implementation—Prevention of Pollution from Noxious Liquid
                 Substances

       The requirements of Annex II of MARPOL, pertaining to the discharges of noxious
liquid substances from ships, are implemented by the U.S. Coast Guard primarily through its
regulations at 33 CFR §§ 151.30-151.49, although some requirements are  also at 46 CFR Parts
11 A ship not traveling en route may discharge oil and oily mixtures, provided it is equipped with a U.S.
government-or IMO-approved 15 ppm bilge alarm and complies with other requirements of 33 CFR § 151.10. 33
CFR§ 15
                                         442

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

151 and 153.12 Which regulations are applicable to a particular vessel depends on the specific
substance(s) the ship is carrying (33 CFR § 151.31).

       The primary regulations at 33 CFR §§ 151.30-151.49 are applicable to the same ships
subject to the implementing regulations for Annex I (i.e., all ships operated under the authority of
the United States that are engaged in international voyages, certificated for ocean service,
certificated for coastwise service beyond 3 nautical miles from land, or operated seaward of the
outermost boundary of the territorial sea of the United States). These requirements also apply to
ships operated under the authority of another country while in U.S. waters or while at a port or
terminal under U.S. jurisdiction (33 CFR §  151.30(a)). The same exemptions that apply to Annex
Fs implementing regulations also apply here, with an added exemption for tank barges whose
certificates are endorsed by the Coast Guard for a limited short protected coastwise route if the
barge is constructed and certificated primarily for service on inland routes (33 CFR § 151.30(b)).

       U.S. oceangoing ships are prohibited from carrying certain Category C and D NLS,
identified at 33 CFR §§ 151.47-151.49, in cargo tanks unless those tanks have been endorsed
through a Certificate of Inspection to carry those  substances. Foreign ships and ships traveling to
foreign destinations must meet additional certification requirements (33 CFR §§ 151.33-151.35).
Ships carrying Category C or D oil-like substances must also meet additional operating
requirements, such as having monitoring and control equipment installed and meeting damage
stability requirements (33 CFR § 151.37).

       To discharge NLS residue to the  sea, the ship must be at least 12 nautical miles from the
nearest land. Additional depth restrictions and maximum rates of discharge also apply for
particular types of residue (46 CFR § 153.1128). Discharges of NLS residue from slop tanks are
also  subject to additional  restrictions (46 CFR § 153.1126). If a ship cannot meet these discharge
requirements, the NLS residue must be retained onboard or discharged to a reception facility. If
the NLS cargo or residue is being transferred at a port or terminal of the United States, the
operator of the ship must notify the port  or terminal at least 24 hours in advance of the name of
the ship and the name, category, and volume of the NLS cargo that will be unloaded (33 CFR §
151.43).

       6.2.1.1.3  Annex V Implementation - Prevention of Garbage Pollution from Ships

       The requirements of Annex V of MARPOL, pertaining to garbage pollution from ships,
are implemented by the Coast Guard through regulations found at 33 CFR §§ 151.51-151.77.
These regulations apply to all ships of U.S.  registry or nationality, all ships operated under the
authority of the United States (including recreational and uninspected vessels), and all ships
12 Coast Guard regulations currently implement a prior version of Annex II. Parts 151 and 153 are currently under
revision to implement revised Annex II, dated November 1, 2004. Navigation and Vessel Inspection Circular No.
03-06 contains guidance on the Coast Guard's implementation of revised Annex II.


                                          443

-------
                           Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

operating in the navigable waters or the Exclusive Economic Zone of the United States. They do
not apply to warships, naval auxiliary, other ships owned or operated by the United States when
engaged in noncommercial service, or any ship specifically excluded by MARPOL (33 CFR §
151.51).

       The regulations prohibit the discharge of garbage into the navigable waters of the United
States by any person onboard any ship unless the requirements of MARPOL are followed.
Commercial ships are permitted to discharge bulk dry cargo residues into the Great Lakes
provided certain requirements are met (33 CFR § 151.66). As with Annex V, the discharge of
plastic or garbage mixed with plastic into the sea or navigable waters of the United  States is
flatly prohibited (33 CFR § 151.67).

       The Coast Guard's regulations also implement the recordkeeping, waste management
plan, placard, inspection for compliance and enforcement, and reporting requirements of
MARPOL (33 CFR §§ 151.55, 151.57, 151.59, 151.61, and 151.65).

       As with MARPOL, oil and grease are pollutants of potential concern found in incidental
discharges that would be directly regulated by APPS and the relevant implementing regulations.
Like MARPOL, APPS and its relevant implementing regulations may indirectly regulate other
pollutants found in incidental vessel discharges, such as metals, to the extent that they are found
in any of the noxious liquid substances or garbage categorized under Annex II and Annex V or
the Coast Guard's implementing regulations and are prevented from entering incidental
discharge streams. As with MARPOL, APPS and the relevant implementing regulations only
apply to those vessels that meet the size thresholds established in MARPOL.

6.2.2   Clean Water Act (CWA) §§ 311, 312/Oil Pollution Control Act

       6.2.2.1    CWA § 311, Oil and Hazardous Substances

       Clean Water Act (CWA) § 311 (Oil and Hazardous Substances Liability) states that it is
U.S. policy that there should be no discharges of oil or hazardous substances into waters of the
U.S., adjoining shorelines, into or upon the waters of the contiguous zone, and in certain other
specified instances, except where permitted under MARPOL/APPS or where in quantities the
president has, by regulation, determined not to be harmful (33 U.S.C. §§ 1321(b)(l)-(b)(3)). The
term "discharge" excludes discharges in compliance with a National Pollutant Discharge
Elimination System (NPDES) permit under CWA § 402; discharges anticipated in the NPDES
permitting process; and discharges incidental to mechanical removal authorized by the president
to remove or mitigate a discharge (33 U.S.C. § 1321(a)(2)). A list of hazardous substances EPA
has designated under the CWA can be found at 40 CFR §  116.1.13
13 EPA's regulations implementing § 311 are located at 40 CFR § 110-117.

                                        444

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States
       Any person in charge of a vessel or onshore facility must immediately notify the
appropriate federal agency upon discovering any harmful discharge of oil or hazardous substance
from the vessel or facility under their control. The federal agency will then notify appropriate
state agencies. Any person in charge of a vessel or onshore facility who discharges in violation
of the CWA and fails to provide  immediate notification to the appropriate federal agency shall,
upon conviction, be fined or imprisoned, or both  (33 U.S.C. § 1321 (b)(5)). Owners or operators
must respond immediately to any discharge or threat of discharge of oil (33 U.S.C. § 1321
       This section of the CWA also requires the president to prepare and publish a National
Contingency Plan (NCP) for the removal of oil and hazardous substances (33 U.S.C. §
1321(d)(l)). The NCP must include:
       •  an assignment of duties and responsibilities among federal departments and agencies;
       •  identification, procurement, maintenance, and storage of equipment and supplies;
       •  establishment of Coast Guard strike teams; a system of surveillance and notice;
       •  establishment of a national center to provide coordination and direction for operations
          in carrying out the plan;
       •  procedures and techniques to be employed in identifying, containing, dispersing, and
          removing oil and hazardous substances;
       •  a schedule of which chemicals and dispersants may be used in which waters to
          mitigate any spills;
       •  a system for states affected by a discharge to act to remove the discharge;
          establishment of criteria and procedures to ensure immediate and effective federal
          identification of and response to discharges or threats  of discharges that will endanger
          public health;
       •  establishment of procedures and standards for removing a worst case discharge of oil;
       •  designation of federal officials to act as on-scene coordinators; establishment of
          procedures for the coordination of activities; and a fish and wildlife response plan (33
          U.S.C. § 1321(d)(2)). The full text of the NCP can be  found at 40 CFR Part 300.
       6.2.2.2    Oil Pollution Control Act
       The Oil Pollution Control Act of 1990 (OPA), 33 U.S.C. §§ 2701-2762, was passed as an
almost immediate response to the Exxon Valdez tanker accident, which caused more than 11
million gallons of crude oil to spill into Alaska's Price William Sound. The OPA expanded the
                                         445

-------
                           Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

federal government's authority to respond to oil spills, provided the money and resources
necessary for the government to exercise its authority, and required revisions to the National Oil
and Hazardous Substances Pollution Contingency Plan to broaden coordination and preparedness
planning requirements. The OPA also increased penalties for regulatory noncompliance,
broadened the response and enforcement authorities of the federal government, and preserved
state authority to establish laws governing oil spill prevention and response. Additionally, the
OPA created the Oil Spill Liability Trust Fund to help fund some of the cleanup costs and repair
damage resulting from oil discharges (discussion  on the exact requirements of the Fund has been
omitted). The requirements of the OPA apply to all vessels, onshore facilities, offshore facilities,
deepwater ports, and pipelines.

       The OPA is implemented by both EPA and the U.S. Coast Guard. EPA regulations on oil
spill prevention and response are found in 40 CFR Parts 112 and 300. U.S. Coast Guard
regulations regarding oil spill prevention and response plans are located at 33 CFR §§ 155.1010-
155.2230 and 49 CFR §§ 130.1-130.33.

       6.2.2.3    CWA § 312, Marine Sanitation Devices

       The CWA also requires EPA, in consultation with the Coast Guard, to promulgate federal
performance standards for marine sanitation devices. These standards must be designed to
prevent the discharge of untreated or inadequately treated sewage into or upon the navigable
waters from vessels (33 U.S.C. § 1322(b)). Both the EPA and Coast Guard have promulgated
regulations implementing this provision. The Coast Guard's regulations can be found at 33 CFR
Part 159, while EPA's can be found at 40 CFR Part 140.

       Because discharges of sewage were exempted by  Congress from this study, as such
discharges are not incidental to the normal operation of a vessel, an in-depth discussion of this
provision and its implementing regulations has been omitted.

6.2.3  Organotin Antifouling Paint Control Act

       The Organotin Antifouling Paint Control Act of 1988, 33 U.S.C.  §§ 2401-2410, prohibits
the use of antifouling paints containing organotin such as tributyltin (TBT) on vessels that are 25
meters or less in length, unless the vessel hull is aluminum or the paint is applied to an outboard
motor (33 U.S.C. § 2403(b)). The term vessel is defined to include "every description of
watercraft or other artificial contrivance used, or capable  of being used, as a means of
transportation on water" (33 U.S.C. § 2402(11)).

       The Act also prohibits the sale, purchase, and application of antifouling paint containing
organotin unless the paint has been approved by EPA as being a qualified antifouling paint.
Under the Act, "antifouling paint" includes any "coating, paint, or treatment that is applied to a
vessel to control fresh water or marine fouling organisms" (33 U.S.C. § 2402(2)). A qualified
                                         446

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                             Regulation under Federal Law or a Binding International Obligation of the United States

antifouling paint is one that has a release rate of not more than 4.0 micrograms per square
centimeter per day (33 U.S.C. § 2402(6)).

       As noted in Section 6.1.2, in September 2008 the United States Senate gave its advice
and consented to ratification of the International Convention on the Control of Harmful Anti-
Fouling Systems on Ships. However, the United States will not deposit its instrument of
ratification with the IMO until Congress adopts the necessary implementing legislation.
Implementing legislation is pending. See Clean Hull Act of 2009, H.R. 3618, 111th Congress
(1st Session 2009). EPA has already canceled all U.S. FIFRA registrations for TBT antifouling
paints. The last cancellation became effective December 31, 2005. Any current use  of these
products is dwindling because there are very limited or no stocks of the products remaining on
the market. Also, the International Convention has made it difficult and undesirable for vessel
owners/operators to use TBT antifouling paints.

       Additionally, as  discussed above, EPA has prohibited the use of TBT or other organotins
as biocides on any vessel covered by the Vessel General Permit.

       The Organotin Antifouling Paint Control Act only regulates the use of organotin, it does
not extend to other pollutants of potential risk that may be present in antifouling hull coatings.
Although the Act banned new applications of antifouling hull coatings containing organotin,
organotin may still be present in residual quantities on older vessels.

6.2.4   National Invasive Species Act

       The primary purpose of the National Invasive Species Act of 1996 (NISA), which
reauthorized and amended the Non-Indigenous Aquatic Nuisance Prevention and Control Act of
1990, is to prevent, monitor, and control the unintentional introduction and dispersal of
nonindigenous species into waters of the United States through ballast water and other pathways
(16 U.S.C. § 4701(b)). The voluntary guidelines and mandatory regulations required by NISA
apply, with only few exceptions, to all vessels equipped with ballast water tanks that operate in
waters of the United States.14

       Because ballast water was specifically exempted from this study by P.L.  110-299, an in-
depth discussion of the ballast water requirements of NISA has been omitted.15 In addition to
ballast water guidelines, however, NISA requires the development of guidelines to prevent the
spread of nonindigenous species from other vessel operations, such as hull fouling.
14 The Act requires that the Coast Guard and the Department of Defense implement ballast water management
programs for seagoing vessels under their control (16 U.S.C. § 4713).
15 For the ballast water requirements, see the text of the Act at 16U.S.C. §§ 4701-4751 as well as the Coast Guard's
regulations at 33 CFR Part 151, subparts C and D.


                                          447

-------
                            Chapter 6 - Analysis of the Extent to which Incidental Discharges are currently subject to
                            Regulation under Federal Law or a Binding International Obligation of the United States

       For example, the Coast Guard's regulations require that all vessels equipped with ballast
water tanks that operate in the waters of the U.S. have fouling organisms removed from their
hulls, piping, and tanks on a regular basis, and that any removed substances be disposed of in
accordance with local, state, and other federal regulations (33 CFR § 151.2035(a)(6)).

       The National Invasive Species Act does not directly regulate any of the pollutants of
potential concern discussed in this study; the Act is focused solely on preventing the spread of
nonindigenous species.

6.2.5  Hazardous Materials Transportation Act

       The Hazardous Materials Transportation Act, 49 U.S.C. §§ 5101 etseq., regulates the
transportation of hazardous material in interstate, intrastate, and foreign commerce. The Act,
which implements MARPOL Annex III, includes registration, reporting, and recordkeeping
requirements and applies to any vessel involved in transporting hazardous material in commerce.

       The Act, and its implementing Hazardous Materials Regulations (HMR), 49  C.F.R. parts
171-180, apply to anyone who transports hazardous material in commerce, causes hazardous
material to be transported in commerce, is involved in any way in the design and manufacture of
containers used to transport hazardous material, prepares or accepts hazardous material for
transport in commerce, is responsible for the safety of transporting hazardous material in
commerce, or certifies compliance with an