UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON D.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
July 12, 2011
EPA-SAB-11-009
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
Subject: Efficacy of Ballast Water Treatment Systems: a Report by the EPA Science
Advisory Board
Dear Administrator Jackson:
This Advisory report responds to a request from EPA's Office of Water (OW) for EPA's Science
Advisory Board (SAB) to provide advice on technologies and systems to minimize the impacts of
invasive species in vessel ballast water discharge. Vessel ballast water discharges are a major source of
non-indigenous species introductions to marine, estuarine, and freshwater ecosystems of the United
States. Ballast water discharges are regulated by the EPA under authority of the Clean Water Act
(CWA) and by the United States Coast Guard (USCG) under authority of the National Invasive Species
Act (NISA). At present, federal requirements for managing ballast water discharges rely primarily on
ballast water exchange; however changes to federal ballast water regulations are under consideration. On
August 28, 2009, the USCG proposed revising their existing rules to establish numeric concentration-
based limits for live organisms in ballast water. The proposed rule would initially require compliance
with a "Phase 1 standard" that has the same concentration limits as the International Maritime
Organization (EVIO) D-2 standard and subsequently require compliance with a more stringent "Phase 2
standard."
The EPA's existing CWA general permit for vessels will expire on Dec. 19, 2013. In its revisions to the
vessel general permit, the EPA is considering numeric standards that limit the number of live organisms
in discharged ballast water. To assist in this, OW requested in their charge questions that the SAB
provide advice regarding the effectiveness of existing technologies for shipboard treatment of vessel
ballast water, how these technologies might be improved in the future, and how to overcome limitations
in existing data. This assessment was conducted by the SAB's Ecological Processes and Effects
Committee (EPEC) as augmented with additional Panel members having expertise in ballast water
issues and water treatment (collectively referred to as the Panel).
To prepare this report, the Panel reviewed a "Background and Issues Paper" prepared by OW and USCG
(June 2010) as well as information on 51 existing or developmental ballast water management systems
(BWMS) provided by OW and the public, although detailed data were available for only 15 BWMS.
-------
Hence this assessment is based on information available at a given point in time. The Panel used this
information as the source material for conducting its assessment of ballast water treatment performance
and, as requested by OW, used the numeric limits proposed by USCG and by some states as
performance benchmarks.
In response to the four specific charge questions, the Panel's findings were that:
(1) Based on the information provided, five of 34 categories of assessed BWMS achieved
reductions in organism concentrations sufficient to comply with the first standard proposed by
the USCG (i.e., the 'Phase 1' standard). Although current test methods and detection limits
preclude a complete statistical assessment of whether a BWMS meets any standard more
stringent than Phase 1, the Panel concluded that none of the assessed BWMS can meet a standard
that is 100 or 1000 times more stringent. Furthermore, it is not reasonable to assume that the
assessed BWMS are able to reliably meet or closely approach a "no living organism" standard.
(2) Current BWMS are based on reasonable engineering designs and standard water treatment
processes, but significant difficulties are encountered in adapting standard water treatment
technologies to shipboard operation (e.g., range of environmental conditions encountered, vessel
operational parameters, and vessel design characteristics).
(3) Reasonable changes in existing BWMS are likely to result in incremental improvements, but
are not likely to lead to 100 or 1000 times further reductions in organism concentrations.
Because of technological, logistical, and personnel constraints imposed by shipboard operations,
wholly new systems need to be developed to meet proposed standards that are 100 or 1000 times
more stringent than Phase 1. The Panel provided some ideas on designs for potential new
systems, recognizing that time will be required to develop and test new approaches to determine
their practicality and cost.
(4) The Panel reviewed the many limitations associated with existing data for ballast water
treatment performance and provided advice on how to correct these limitations in future
assessments; the Panel recommends using improved testing protocols for verifying discharge
concentrations, exploring the use of surrogate performance measures, and developing reliable
protocols for compliance monitoring.
However, the Panel's overarching recommendation is that the EPA adopt a risk-based approach to
minimize the impacts of invasive species in vessel ballast water discharge rather than relying solely on
numeric standards for discharges from shipboard BWMS. The Panel found that insufficient attention has
been given to integrated sets of practices and technologies that could be used to systematically advance
ballast water management. These practices include managing ballast uptake to reduce the presence of
invasive species, reducing invasion risk through operational adjustments and changes in ship design to
reduce or eliminate the need for ballast water, development of voyage-based risk and/or hazard
assessments, and treatment of ballast water in onshore reception facilities. The Panel recommended that
a comprehensive analysis be done to compare biological effectiveness, cost, logistics, operations and
safety associated with shipboard BWMS and onshore reception facilities.
-------
The SAB appreciates the opportunity to provide the EPA with advice on this important topic. We look
forward to receiving the Agency's response.
Sincerely,
/Signed/ /Signed/
Dr. Deborah L. Swackhamer Dr. Judith L. Meyer
Chair Chair
Science Advisory Board SAB Ballast Water Advisory Panel
Enclosure
-------
NOTICE
This report has been written as part of the activities of the EPA Science Advisory Board (SAB), a public
advisory group providing extramural scientific information and advice to the Administrator and other
officials of the Environmental Protection Agency. The SAB is structured to provide balanced, expert
assessment of scientific matters related to problems facing the Agency. This report has not been
reviewed for approval by the Agency and, hence, the contents of this report do not necessarily represent
the views and policies of the Environmental Protection Agency, nor of other agencies in the Executive
Branch of the Federal government, nor does mention of trade names of commercial products constitute a
recommendation for use. Reports of the SAB are posted on the EPA website at
http ://www. epa.gov/sab.
-------
U.S. Environmental Protection Agency
Science Advisory Board
Ecological Processes and Effects Committee
Augmented for the Ballast Water Advisory
(Ballast Water Advisory Panel)
CHAIR
Dr. Judith L. Meyer, Professor Emeritus, Odum School of Ecology, University of Georgia, Lopez
Island, WA
MEMBERS
Dr. E. Fred Benfield, Professor of Ecology, Department of Biological Sciences, Virginia Tech,
Blacksburg, VA
Dr. G. Allen Burton, Professor and Director, Cooperative Institute for Limnology and Ecosystems
Research, School of Natural Resources and Environment, University of Michigan, Ann Arbor, MI
Dr. Peter Chapman, Principal and Senior Environmental Scientist, Environmental Sciences Group,
Golder Associates Ltd, Burnaby, BC, Canada
Dr. William Clements, Professor, Department of Fish, Wildlife, and Conservation Biology, Colorado
State University, Fort Collins, CO
Dr. Loveday Conquest, Professor, School of Aquatic and Fishery Sciences, University of Washington,
Seattle, WA
Dr. Robert Diaz, Professor, Department of Biological Sciences, Virginia Institute of Marine Science,
College of William and Mary, Gloucester Pt, VA
Dr. Wayne Landis, Professor and Director, Department of Environmental Toxicology, Institute of
Environmental Toxicology, Huxley College of the Environment, Western Washington University,
Bellingham, WA
Dr. Thomas W. La Point, Professor, Department of Biological Sciences, University of North Texas,
Denton, TX
Dr. Amanda Rodewald, Associate Professor, School of Environment and Natural Resources, The Ohio
State University, Columbus, OH
Dr. James Sanders, Director and Professor, Skidaway Institute of Oceanography, Savannah, GA
AUGMENTED PANEL MEMBERS
Dr. JoAnn Burkholder, Professor, Department of Plant Biology, Center for Applied Aquatic Ecology,
North Carolina State University, Raleigh, NC
11
-------
Dr. Andrew N. Cohen1, Director, Center for Research on Aquatic Bioinvasions, Richmond, CA
Dr. Fred Dobbs, Professor and Graduate Program Director, Ocean, Earth and Atmospheric Sciences,
College of Sciences, Old Dominion University, Norfolk, VA
Dr. Lisa Drake, Physical Scientist, Center for Corrosion Science and Engineering, Naval Research
Laboratory, Key West, FL
Dr. Charles Haas, L.D. Betz Professor of Environmental Engineering, Civil, Architectural and
Environmental Engineering, College of Engineering, Drexel University, Philadelphia, PA
Mr. Edward Lemieux, Director, Center for Corrosion Science Engineering, Naval Research
Laboratory, Washington, DC
Dr. David Lodge, Professor, Biological Sciences, University of Notre Dame, Notre Dame, IN
Mr. Kevin Reynolds, Senior Marine Engineer, The Glosten Associates, Seattle, WA
Dr. Mario Tamburri, Associate Professor, Chesapeake Biological Laboratory, Maritime
Environmental Resource Center, University of Maryland Center for Environmental Science, Solomons,
MD
Dr. Nicholas Welschmeyer, Professor of Oceanography, Moss Landing Marine Laboratories, San Jose
State University, Moss Landing, CA
SCIENCE ADVISORY BOARD STAFF
Ms. Iris Goodman, Designated Federal Officer, U.S. Environmental Protection Agency, Washington,
DC
Dr. Cohen did not concur with the final draft report submitted to the chartered SAB for their quality review and approval.
iii
-------
U.S. Environmental Protection Agency
Science Advisory Board
CHAIR
Dr. Deborah L. Swackhamer, Professor and Charles M. Denny, Jr. Chair in Science, Technology and
Public Policy, Hubert H. Humphrey School of Public Affairs and Co-Director of the Water Resources
Center, University of Minnesota, St. Paul, MN
SAB MEMBERS
Dr. David T. Allen, Professor, Department of Chemical Engineering, University of Texas, Austin, TX
Dr. Claudia Benitez-Nelson, Full Professor and Director of the Marine Science Program, Department
of Earth and Ocean Sciences , University of South Carolina, Columbia, SC
Dr. Timothy J. Buckley, Associate Professor and Chair, Division of Environmental Health Sciences,
College of Public Health, The Ohio State University, Columbus, OH
Dr. Patricia Buffler, Professor of Epidemiology and Dean Emerita, Department of Epidemiology,
School of Public Health, University of California, Berkeley, CA
Dr. Ingrid Burke, Director, Haub School and Ruckelshaus Institute of Environment and Natural
Resources, University of Wyoming, Laramie, WY
Dr. Thomas Burke, Professor, Department of Health Policy and Management, Johns Hopkins
Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD
Dr. Terry Daniel, Professor of Psychology and Natural Resources, Department of Psychology, School
of Natural Resources, University of Arizona, Tucson, AZ
Dr. George Daston, Victor Mills Society Research Fellow, Product Safety and Regulatory Affairs,
Procter & Gamble, Cincinnati, OH
Dr. Costel Denson, Managing Member, Costech Technologies, LLC, Newark, DE
Dr. Otto C. Doering III, Professor, Department of Agricultural Economics, Purdue University, W.
Lafayette, IN
Dr. David A. Dzombak, Walter J. Blenko, Sr. Professor of Environmental Engineering, Department of
Civil and Environmental Engineering, College of Engineering, Carnegie Mellon University, Pittsburgh,
PA
Dr. T. Taylor Eighmy, Vice President for Research, Office of the Vice President for Research, Texas
Tech University, Lubbock, TX
Dr. Elaine Faustman, Professor and Director, Institute for Risk Analysis and Risk Communication,
School of Public Health, University of Washington, Seattle, WA
iv
-------
Dr. John P. Giesy, Professor and Canada Research Chair, Veterinary Biomedical Sciences and
Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
Dr. Jeffrey K. Griffiths, Professor, Department of Public Health and Community Medicine, School of
Medicine, Tufts University, Boston, MA
Dr. James K. Hammitt, Professor, Center for Risk Analysis, Harvard University, Boston, MA
Dr. Bernd Kahn, Professor Emeritus and Associate Director, Environmental Radiation Center, Georgia
Institute of Technology, Atlanta, GA
Dr. Agnes Kane, Professor and Chair, Department of Pathology and Laboratory Medicine, Brown
University, Providence, RI
Dr. Madhu Khanna, Professor, Department of Agricultural and Consumer Economics, University of
Illinois at Urbana-Champaign, Urbana, IL
Dr. Nancy K. Kim, Senior Executive, Health Research, Inc., Troy, NY
Dr. Kai Lee, Program Officer, Conservation and Science Program, David & Lucile Packard Foundation,
Los Altos, CA (Affiliation listed for identification purposes only)
Dr. Cecil Lue-Hing, President, Cecil Lue-Hing & Assoc. Inc., Burr Ridge, IL
Dr. Floyd Malveaux, Executive Director, Merck Childhood Asthma Network, Inc., Washington, DC
Dr. Lee D. McMullen, Water Resources Practice Leader, Snyder & Associates, Inc., Ankeny, IA
Dr. Judith L. Meyer, Professor Emeritus, Odum School of Ecology, University of Georgia, Lopez
Island, WA
Dr. James R. Mihelcic, Professor, Civil and Environmental Engineering, University of South Florida,
Tampa, FL
Dr. Jana Milford, Professor, Department of Mechanical Engineering, University of Colorado, Boulder,
CO
Dr. Christine Moe, Eugene J. Gangarosa Professor, Hubert Department of Global Health, Rollins
School of Public Health, Emory University, Atlanta, GA
Dr. Horace Moo-Young, Dean and Professor, College of Engineering, Computer Science, and
Technology, California State University, Los Angeles, CA
Dr. Eileen Murphy, Grants Facilitator, Ernest Mario School of Pharmacy, Rutgers University,
Piscataway, NJ
-------
Dr. Duncan Patten, Research Professor, Hydroecology Research Program, Department of Land
Resources and Environmental Sciences, Montana State University, Bozeman, MT
Dr. Stephen Polasky, Fesler-Lampert Professor of Ecological/Environmental Economics, Department
of Applied Economics, University of Minnesota, St. Paul, MN
Dr. Arden Pope, Professor, Department of Economics, Brigham Young University, Provo, UT
Dr. Stephen M. Roberts, Professor, Department of Physiological Sciences, Director, Center for
Environmental and Human Toxicology, University of Florida, Gainesville, FL
Dr. Amanda Rodewald, Professor of Wildlife Ecology, School of Environment and Natural Resources,
The Ohio State University, Columbus, OH
Dr. Jonathan M. Samet, Professor and Flora L. Thornton Chair, Department of Preventive Medicine,
University of Southern California, Los Angeles, CA
Dr. James Sanders, Director and Professor, Skidaway Institute of Oceanography, Savannah, GA
Dr. Jerald Schnoor, Allen S. Henry Chair Professor, Department of Civil and Environmental
Engineering, Co-Director, Center for Global and Regional Environmental Research, University of Iowa,
Iowa City, IA
Dr. Kathleen Segerson, Philip E. Austin Professor of Economics , Department of Economics,
University of Connecticut, Storrs, CT
Dr. Herman Taylor, Director, Principal Investigator, Jackson Heart Study, University of Mississippi
Medical Center, Jackson, MS
Dr. Barton H. (Buzz) Thompson, Jr., Robert E. Paradise Professor of Natural Resources Law at the
Stanford Law School and Perry L. McCarty Director, Woods Institute for the Environment, Stanford
University, Stanford, CA
Dr. Paige Tolbert, Professor and Chair, Department of Environmental Health, Rollins School of Public
Health, Emory University, Atlanta, GA
Dr. John Vena, Professor and Department Head, Department of Epidemiology and Biostatistics,
College of Public Health, University of Georgia, Athens, GA
Dr. Thomas S. Wallsten, Professor and Chair, Department of Psychology, University of Maryland,
College Park, MD
Dr. Robert Watts, Professor of Mechanical Engineering Emeritus, Tulane University, Annapolis, MD
Dr. R. Thomas Zoeller, Professor, Department of Biology, University of Massachusetts, Amherst, MA
VI
-------
SCIENCE ADVISORY BOARD STAFF
Dr. Angela Nugent, Designated Federal Officer, U.S. Environmental Protection Agency, Washington,
DC
Ms. Stephanie Sanzone, Designated Federal Officer, U.S. Environmental Protection Agency,
Washington, DC
vn
-------
Table of Contents
1. EXECUTIVE SUMMARY
2. INTRODUCTION 9
2.1. EPA's Charge to the SAB 9
2.2. SAB's Review Process 11
2.3. Regulatory Frameworks for Ballast Water Management 11
2.3.1. U.S. Federal rules 11
2.3.2. Other Regulatory Frameworks 12
2.3.3. Glossary of Terms Used 13
2.4. Applying Risk Assessment Principles to Ballast Water Management 16
2.4.1. Risk Assessment of Nonindigenous Species 16
2.4.2. Ballast Water Management Goals and the Decision Making, Risk Assessment Context 19
3. STATISTICS AND INTERPRETATION 22
3.1. Introduction 22
3.2. Assessing Whether Ballast Water Standards Can Be Met — The Statistics of Sampling 22
3.2.1. The Poisson Distribution 23
3.2.2. Spatially Aggregated Populations - Negative Binomial Distributions 27
3.3. Interactive Effects 28
3.4. Certainty of Results 28
3.5. Conclusions 29
4. PERFORMANCE OF SHIPBOARD SYSTEMS WITH AVAILABLE EFFLUENT TESTING DATA 30
4.1. EPA's Charge Questions 30
4.2. Assessment Process 30
4.3. Assessing the Reliability of Existing Data 30
4.4. Assessing the Ability of BWMS to Meet Discharge Standards 31
4.5. Assessment Results 36
4.6. Response to Charge Question 1 36
4.7. Response to Charge Question 2 38
4.8. Environmental Effects and Vessel Applications: Additional Constraints and Considerations that Influence
BWMS Performance 39
Vlll
-------
5. SYSTEM DEVELOPMENT 42
5.1. Introduction 42
5.2. Improving the Performance of Existing Systems 42
5.2.1. Combination Technologies 43
5.2.2. UV Radiation 44
5.2.3. Mechanical Separation and Cavitation 45
5.2.4. Deoxygenation 45
5.2.5. Oxidant-Based Systems 46
5.3. Principal Technological Constraints 49
5.3.1. Operational Challenges on Working Merchant Vessels 50
5.3.2. Idealized Designs for BWMS 53
5.4. Recommendations for Addressing Impediments and Constraints 56
5.5. Impediments Based on Organism Type 57
5.6. Sterilization of Ballast Water Discharge 57
6. LIMITATIONS OF EXISTING STUDIES AND REPORTS 58
6.1. EPA's Charge Question 58
6.2. Testing Shipboard Treatment Systems: Protocols, Analysis, and Reporting Practices that Could Be Improved.... 58
6.2.1 Confusion of Research and Development and Certification Testing 58
6.2.2 Lack of Standardized Testing Protocols 59
6.2.3. Compromises Necessary Because of Practical Constraints in Sampling and Available Methods 70
6.2.4. Testing Shipboard Treatment Systems: Inherent Mismatch between Viability Standard and Practical
Protocols 75
6.3. Approaches to Compliance/Enforcement of Ballast Water Regulations and Potential Application to Technology
Testing 79
6.4. Reception Facilities as an Alternative to Shipboard Treatment 80
6.4.1. Potential of Reception Facilities to Cost Effectively Meet Higher Standards 80
6.4.2. Challenges to Widespread Adoption of Reception Facilities in the U.S 85
6.5. Approaches Other than Ballast Water Treatment 88
6.5.1. Managing Ballast Uptake 88
6.5.2. Mid-ocean Exchange 89
6.5.3. Reducing or Eliminating Ballast Water Discharge Volumes 89
6.5.4. Temporal and Spatial Patterns 91
6.5.5. Combined Approaches 91
6.6. Risk Management Approaches to Reduce Invasion Risk: Hazard Analysis and Critical Control Points (HAACP).... 92
6.7. Summary and Recommendations 95
6.7.1. Principal Limitations of Available Data and Protocols 95
6.7.2. Alternatives to Shipboard Treatment of Ballast Water 95
6.7.3. Recommendations to Overcome Present Limitations 96
REFERENCES R-l
ix
-------
APPENDIX A: DOCUMENTS ON BALLAST WATER TECHNOLOGIES PROVIDED TO THE PANEL ....A-l
APPENDIX B: LITERATURE REVIEW OF RECEPTION FACILITY STUDIES B-l
APPENDIX C: FURTHER INFORMATION ON STATISTICS AND INTERPRETATION C-l
-------
1. EXECUTIVE SUMMARY
Vessel ballast water discharges are a primary source of introductions of nonindigenous species and
potentially harmful pathogens to marine, estuarine, and freshwater ecosystems of the United States. At
present, federal requirements for managing ballast water discharges rely primarily on ballast water
exchange. However, the U.S. Coast Guard (USCG) has proposed numeric concentration-based limits
for live organisms in ballast water that would initially require compliance with a "Phase 1 standard" that
has the same concentration limits as the International Maritime Organization (EVIO) D-2 standard and
subsequently require compliance with a more stringent "Phase 2 standard." In addition, the U.S.
Environmental Protection Agency is considering numeric standards that limit the number of live
organisms in discharged ballast water in its revision of the Vessel General Permit.
This Advisory report responds to a request from the EPA Office of Water (OW) for EPA's Science
Advisory Board (SAB) to provide advice on technologies and systems to minimize the impacts of
invasive species in vessel ballast water discharge. More specifically, the SAB was requested to provide
review and advice regarding whether existing shipboard treatment technologies can reach specified
concentrations of organisms in vessel ballast water, how these technologies might be improved in the
future, and how to overcome limitations in existing data. To conduct the assessment, the SAB's
Ecological Processes and Effects Committee (EPEC) was augmented with additional Panel members
(Ballast Water Advisory Panel, or "Panel") having expertise in ballast water issues, marine engineering,
and engineering treatment technologies.
In addition to the SAB assessment of ballast water management systems (BWMS), the EPA and the
USCG commissioned the National Academy of Sciences (NAS) to conduct a complementary study2 that
assesses the risk of successful invasions as a function of different concentrations of organisms in ballast
water discharges. Therefore, this SAB Advisory does not address the relationship between the various
concentrations of organisms as described in proposed standards and the likelihood of invasions or the
potential effects of pathogens. Rather, in this report, "effectiveness" refers to how well a technology
may meet a given numeric concentration limit for organisms, not to how well the technology may
protect the environment. Similarly, phrases such as "meets more stringent" or "meets less stringent"
proposed standards are used as descriptive conclusions and do not represent performance
recommendations.
To prepare this Advisory report, the Panel reviewed a "Background and Issue Paper" written by EPA's
OW and USCG (Albert et al. 2010). This paper provided an overview of information about major
categories of shipboard ballast water treatment technologies and presented proposed ballast water
discharge standards drawn from international sources, the USCG, and nine states. In addition, EPA's
OW and the public identified information on 51 existing or developmental ballast water treatment
technologies, although detailed data were available for only 15 specific BWMS. The Panel used this
information as the source material for its assessment of ballast water treatment performance and, as
requested by the EPA, used proposed ballast water discharge standards as the performance benchmarks.
2 Assessing the Relationship Between Propagule Pressure and Invasion Pdsk in Ballast Water, 6/2/2011,
http ://www. nap. edu/catalog.php?record_id= 13184)
-------
Regulatory context
Ballast water discharges are regulated by EPA under authority of the Clean Water Act (CWA) and by
the USCG under authority of the National Invasive Species Act (NTS A). In December 2008, EPA issued
a Vessel General Permit (VGP) for discharges incidental to the normal operation of commercial vessels,
including ballast water discharges. The VGP sets effluent limits for ballast water that rely on "best
management practices" (primarily use of ballast water exchange, or BWE) and do not include a numeric
discharge limit. The VGP will expire on Dec. 19, 2013. For subsequent iterations of the VGP, the EPA
has stated its intention to establish best available technology standards for the treatment of ballast water,
once such technologies are shown to be commercially available and economically achievable.
Existing USCG rules governing ballast water also primarily rely on BWE. In August 2009, the USCG
proposed revisions to their existing rules to establish numeric concentration-based limits for viable
organisms in ballast water. The proposed USCG rule would initially require compliance with a "Phase
1" standard, and, if a practicability review shows it is feasible, it would be followed by a "Phase 2"
standard that sets concentration limits at 1000 times more stringent than Phase 1 standards for viable
organisms >10 |im in minimum dimension. Phase 2 standards also set limits on the discharge
concentration for bacteria and viruses. Neither Phase 1 nor Phase 2 standards have been finalized. The
USCG Phase 1 standards have essentially the same concentration limits as those adopted in 2004 by the
International Maritime Organization's (EVIO) International Convention for the Control and Management
of Ships' Ballast Water and Sediments; thus both standards are often referred to as the "D-2/ Phase 1
standards." The U.S. is not a Party to the Convention, nor has the Convention yet entered into force.
However, manufacturers of BWMS have generally designed their equipment to meet these EVIO D-2
standards.
Ballast water management should be implemented using a risk-based systems approach
The Panel recommends that any ballast water management strategy to decrease the rate of successful
invasions by nonindigenous species or introduction of pathogens be part of an overall risk-based
management plan that includes methods to reduce invasion events, process and environmental
monitoring, containment, and eradication. Emphasis only on one aspect, the initial introduction of
organisms, is not likely to reduce the risk of invasions as effectively or as cost-efficiently as a risk
assessment approach that considers all the stages of the invasion process including survival after
introduction. Decisions on approaches to ballast water management should be viewed in the context of
risk management and should: (1) recognize the stochastic and non-linear nature of the invasion process,
(2) clearly define the management goals, and (3) evaluate the effectiveness of BWMS within the context
of other sources of nonindigenous species and other organisms found on the vessel and in the treatment
system, and with respect to specific receiving habitats.
Rigorous sampling and statistical verification of performance is essential
The Panel was asked to respond to charge questions that focused primarily on whether test data
demonstrated that BWMS met or "closely approached" proposed standards for discharge and whether
they did so "credibly" and "reliably." As benchmarks for performance, the Panel was asked to consider
proposed numerical standards as well as narrative descriptions such as "no living organisms,"
"sterilization," and "zero or near zero" discharge. In order to place its assessments of treatment
performance in appropriate scientific context, the Panel first had to consider statistical and sampling
-------
issues. While "zero detectable discharge" might initially seem a desirable standard, it is not statistically
verifiable. Further, verification of standards that set very low organism concentrations may require water
samples that are too large to be logistically feasible. However, when small sample volumes are used, the
probability of detecting an organism is low even when the actual organism concentration is relatively
high. These errors depend on the sample volume collected, and the relative errors are much larger for
small sample volumes.
A well-defined, rigorous sampling protocol is necessary to assess the effectiveness of BWMS at meeting
different standards. These sampling protocols should include consideration of the spatial distribution of
plankton in ballast water. The Poisson distribution is recommended as the model for statistical analysis
of treated water samples.
The Panel also concludes that the D-2/Phase 1 performance standards for discharge quality are currently
measurable, based on data from land-based and shipboard testing. However, current methods (and
associated detection limits) prevent testing of BWMS to any standard more stringent than D-2/Phase 1
and make it impracticable for verifying a standard 100 or 1000 times more stringent. New or improved
methods will be required to increase detection limits sufficiently to statistically evaluate a standard lOx
more stringent than EVIO D-2/Phase 1; such methods may be available in the near future. These
conclusions pertain to evaluating data from land-based and shipboard testing, although the same
statistical theory and practice applies to compliance testing by port state control officers.
Charge question 1: Performance of shipboard systems with available effluent testing data
a. For the shipboard systems with available test data, which have been evaluated with sufficient rigor to
permit a credible assessment of performance capabilities in terms of effluent concentrations achieved
(living organisms/unit of ballast water discharged or other metric)?
Evaluations of technologies are necessarily based on performance information for a given point in time
and the development and manufacture of ballast water treatment systems is a dynamic industry. For this
assessment, the Panel reviewed information provided by EPA's Office of Water and the public. This
information included peer-reviewed articles and publications; information provided directly from
individual manufacturers of BWMS (some included data reports, others provided only Type Approval
certificates); and public dossiers submitted to the IMO Group of Experts on the Scientific Aspects of
Marine Environmental Protection (GESAMP). This information was prepared or published prior to May,
2010. However, the majority of the documents were from 2008 to 2010, reflecting growth in the BWMS
industry. While other BWMS may exist, the Panel considered only those for which information was
provided.
From this information, the Panel identified 51 individual BWMS, which can be grouped into 34
categories of treatment technologies. Of the 51 BWMS identified, the Panel concluded that test data and
other information for 15 individual BWMS were credible and sufficient to permit an assessment of
performance capabilities. Of these 15 BWMS, nine systems (representing individual configurations of
five different categories of BWMS) achieved significant reductions in organism concentrations, and
were able to comply with the Phase 1 standard. These five categories of BWMS technologies are:
(1) Deoxygenation + cavitation; (2) Filtration + chlorine dioxide; (3) Filtration + UV; (4) Filtration +
UV + TiO2; and (5) Filtration + electro-chlorination.
-------
b. For those systems identified in (la), what are the discharge standards that the available data credibly
demonstrate can be reliably achieved? Furthermore, do data indicate that certain systems (as tested)
will not be able to reliably reach any or all of the discharge standards shown in that table?
The Panel concluded that the same five BWMS categories (listed above) have been demonstrated to
meet the EVIO D-2 discharge standard, when tested under the EVIO certification guidelines, and will
likely meet USCG Phase 1 standards, if tested under EPA's more detailed Environmental Technology
Verification (ETV) Protocol (U.S. EPA 2010). The Panel acknowledges the significant achievement of
several existing BWMS to effectively and reliably remove living organisms from ballast water under the
challenging conditions found on active vessels.
The detection limits for currently available test methods preclude a complete statistical assessment of
whether BWMS can meet standards more stringent than EVIO D-2/Phase 1. However, based on the
available testing data, it is clear that while five types of BWMS are able to reach EVIO D-2/Phase 1,
none of the systems evaluated by the Panel performed at 100 times or 1000 times the Phase 1 standard.
c. For those systems identified in (la), if any of the system tests detected "no living organisms " in any
or all of their replicates, is it reasonable to assume the systems are able to reliably meet or closely
approach a "no living organism " standard based on their engineering design and treatment processes?
The Panel concluded that it is not reasonable to assume that BWMS are able to reliably meet or closely
approach a "no living organism" standard. Available data demonstrate that current BWMS do not
achieve sterilization or the complete removal of all living organisms.
Charge question 2: Potential performance of shipboard systems without reliable testing data
Based on engineering design and treatment processes used, and shipboard conditions/constraints, what
types of ballast water treatment systems can reasonably be expected to reliably achieve any of the
proposed standards, and if so, by what dates? Based on engineering design and treatment processes
used, are there systems which conceptually would have difficulty meeting any or all of the proposed
discharge standards?
The Panel found that nearly all of the 51 BWMS evaluated are based on reasonable engineering designs
and treatment processes, and most are adapted from long-standing water treatment approaches.
However, the lack of detailed information on the great majority of BWMS precludes an assessment of
limitations in meeting any or all discharge standards. In particular, the Panel determined that the
following data are essential to future assessments: documentation that test protocols were followed; full
reporting of all test results; and documentation that rigorous Q A/QC methods were followed.
Although several BWMS appear to safely and effectively meet EVIO D-2/Phase 1 discharge standards,
the Panel notes that factors beyond mechanical and biological efficacy need to be considered as BWMS
technology matures. Several parameters will affect the performance or applicability of individual
BWMS to the wide variety of vessel types that carry ballast water. These include environmental
parameters (e.g., temperature and salinity), operational parameters (e.g., ballast volumes and holding
times), and vessel design characteristics (e.g., ballast volume and unmanned barges).
-------
Charge question 3: System development
a. For those systems identified in questions 1 a. and 2, are there reasonable changes or additions to
their treatment processes which can be made to the systems to improve performance?
The Panel defined "reasonable changes" as moderate adjustments that do not fundamentally alter the
treatment process. Based on information from available test results, such moderate adjustment could be
made to treatment processes, although it may add costs and engineering complexity. Examples of
moderate adjustments are:
• Deoxygenation + cavitation. It may be possible to reduce the time needed to reach severe
hypoxia, to increase holding time under severe hypoxia, and to increase the degree of
cavitation and physical/mechanical disruption of organisms.
• Mechanical separation + oxidizing agent. These systems could be optimized by improving
mechanical separation, increasing concentration and contact time for oxidizing agents, and
adjusting other water chemistry parameters (e.g., pH) to increase oxidizing agent efficacy.
• Mechanical separation + UV. These systems could be optimized by improved mechanical
separation and by increasing UV contact time and dosage.
The Panel concludes that moderate adjustments or changes to existing combination technologies are
expected to result in only incremental improvements. Reaching the Phase 2 standard, or even lOOx IMO
D-2/ Phase 1, would require wholly new treatment systems. Such new systems likely would use new
technological devices, including those drawn from the water treatment industry; employ multistage
treatment processes; emphasize technological process controls and multiple monitoring points; include
physical barriers to minimize the potential for cross-contamination of the system; and become part of an
integrated ballast water management effort. These new approaches likely will achieve higher
performance, but will require time to develop and test in order to determine their practicality and cost.
b. What are the principal technological constraints or other impediments to the development of ballast
water treatment technologies for use onboard vessels to reliably meet any or all of the discharge
standards?
Existing BWMS have been developed within the context of typical marine vessel constraints, including
restrictions on size, weight, and energy demands. The primary impediments to the ability of shipboard
systems to meet stringent discharge standards are that treatment processing plants likely will be large,
heavy, and energy intensive—many existing vessels may be unable to overcome these barriers through
retrofitting BWMS. Meeting more stringent performance standards may require a fundamental shift in
how ballast water is managed.
Existing and potential BWMS share several common impediments to development: (1) The focus has
been on engineering the technology with less attention to equally important issues such as training,
operation, maintenance, repair, and monitoring. (2) Without an established compliance monitoring and
enforcement regime to guide design requirements for technologies, incentives for further innovations are
dampened. (3) Facilities properly equipped to test BWMS technologies are few, so increased sharing of
data and testing protocols among such facilities is essential. (4) Discharge standards differ domestically
and internationally, giving manufacturers multiple targets. (5) Meeting more stringent standards requires
that BWMS consistently perform nearly perfectly; a fundamental shift in system design and operational
-------
practices would be needed to achieve this level of performance. (6) Once performance tests indicate that
a given BWMS meets IMO D-2/ Phase 1 standards, further efforts by manufacturers to improve design
and efficacy appear to decline.
c. What recommendations does the SAB have for addressing these impediments and constraints?
Clearly defined and transparent programs for compliance monitoring and enforcement are needed to
promote consistent, reliable operation of BWMS; such programs do not yet exist. Ideally, vessel crew
would have the technological capability to self-monitor BWMS efficacy and make real-time corrections
to maintain compliance. BWMS manufacturers should document performance metrics beyond discharge
treatment efficacy, such as energy consumption and reliability. This would enable vessel operators to
select systems that best integrate with their operations. Although meeting significantly higher standards
will likely require completely new treatment approaches, the Panel can neither predict which
combination of treatment processes will achieve the highest efficacy nor their ultimate performance. The
Panel recommends that one or more pilot projects be commissioned to explore new approaches to ballast
water treatment, including tests of ballast water transfer and treatment at an onshore reception facility.
d. Are these impediments more significant for certain size classes or types of organisms (e.g.,
zooplankton versus viruses)? Can currently available treatment processes reliably achieve sterilization
(no living organisms or viable viruses) of ballast water onboard vessels or, at a minimum, achieve zero
or near zero discharge for certain organism size classes or types?
Shipboard impediments apply to all size classes of organisms and specified microbes. Some treatment
systems or combinations are more effective for treating larger organisms and others for treating
unicellular organisms. The technology exists to remove or kill the great majority and in some cases, to
remove nearly all organisms > 50 jim from discharged water. Given the volumes of water involved,
onboard sterilization of ballast water is not possible using current technologies. It is not possible to
verify zero (sterilization) or near-zero discharge. Such values cannot be measured in a scientifically
defensible way.
Charge question 4: Development of reliable information
What are the principal limitations of the available studies and reports on the status of ballast water
treatment technologies and system performance and how can these limitations be overcome or corrected
in future assessments of the availability of technology for treating ballast water onboard vessels?
Existing information about ballast water treatment is limited in many respects, including significant
limitations in data quality, shortcomings in current methods for testing BWMS and reporting results,
issues related to setting standards and for compliance monitoring, and issues related to test protocols,
including the use of surrogate indicators.
More broadly, however, the Panel found that because of the lack of an overall risk management systems
approach, insufficient attention has been given to integrated sets of practices and technologies to reduce
invasion or pathogen risk by (1) managing ballast, (2) adjustments in operation and ship design to
reduce or eliminate the need for ballast water, (3) development of voyage-based risk assessments and
application of Hazard Analysis and Critical Control Points (HACCP) principles, and (4) options for
reception facilities for the treatment of ballast water. The Panel concludes that combinations of practices
-------
and technologies are potentially more effective and cost-efficient than sole reliance on shipboard
BWMS.
Principal limitations of available data and protocols
Data are not sufficiently compatible to compare rigorously across BWMS because standard test
protocols have been lacking. The most recent EPA ETV Protocol, published in 2010, will improve this.
Reporting of test failures during type approval testing is not required, although some independent test
facilities do report failures. This should be uniform across research and other test facilities so that it is
possible to draw conclusions about the consistency or reliability of BWMS.
Clear definitions and direct methods to enumerate viable organisms are missing for some organisms and
are logistically problematic for all size classes, especially nonculturable bacteria, viruses, and resting
stages of many other taxa. Methods to enumerate viruses are not included in the proposed USCG Phase
2 standard. The important size class of protists3 < 10 jim have not been considered adequately in
developing guidelines and standards, although some Panel members felt that other measurements may
indicate activity in that size class.
Alternatives to shipboard treatment of ballast water
Data on the effectiveness of practices and technologies other than shipboard BWMS are few. Use of
reception facilities for the treatment of ballast water appears to be technically feasible (given generations
of successful water treatment and sewage treatment technologies), and is likely to be more reliable and
more readily adaptable than shipboard treatment. Existing regional economic studies suggest that
treating ballast water in reception facilities would be at least as economically feasible as shipboard
treatment. However, these studies consider that vessels only call at those regional facilities; if vessels
also call at ports outside the region without reception facilities, they would need a shipboard BWMS4.
The effort and cost of monitoring and enforcement needed to achieve a given level of compliance is
likely to be less for a smaller number of reception facilities compared to a larger number of BWMS.
However, the Panel did not reach agreement on several issues related to treatment of ballast water in
onshore reception facilities. For further details on the Panel's views, see Section 6.4; for a review of the
literature review on treatment of ballast water in onshore reception facilities, see Appendix B.
Recommendations to overcome present limitations
As illustrated in the ETV protocol (U.S. EPA 2010), testing of BWMS in a research and development
mode should be distinct from testing for type approval certification and for verification. Certification
testing should be conducted by a party independent from the manufacturer with appropriate, established
credentials, approved by EPA/USCG. Test failures and successes during type approval testing should be
reported and considered in certification decisions. A transparent international standard format for
reporting, including specification of quality assurance and quality control (QA/QC) protocols and a
means to indicate QA/QC procedures were followed during testing, are needed. In addition, the EPA
3 Protists refers to various one-celled organisms classified in the kingdom Protista, and which includes protozoans, eukaryotic
algae, and slime molds.
4
Dr. Cohen, who read these studies, objected to this sentence as being untrue and misleading and felt there had not been
adequate opportunity for Panel discussion of the issue.
-------
should develop metrics and methods appropriate for compliance monitoring and enforcement as soon as
possible.
Limits for selected protists < 10 jim in minimum dimension should be included in ballast water
discharge standards and in BWMS test protocols. Suitable standard test organisms should be identified
for bench-scale testing, and surrogate parameters should be investigated to complement or replace
metrics that are logistically difficult or infeasible for estimating directly the concentration of living
organisms. Representative "indicator" taxa (toxic strains of Vibrio cholerae; Escherichia coli; intestinal
Enterococci) should continue to be used to assess BWMS. Estimates of the removal of harmful bacteria
will be improved when reliable techniques become available to account for active nonculturable cells as
well as culturable cells.
EPA should conduct a comprehensive analysis comparing biological effectiveness, cost, logistics,
operations, and safety associated with both shipboard BWMS and reception facilities. If the analysis
indicates that treatment at reception facilities is both economically and logistically feasible and is more
effective than shipboard treatment systems, it should be used as the basis for assessing the ability of
available technologies to remove, kill, or inactivate living organisms to meet a given discharge standard.
In other words, use of reception facilities may enable ballast water discharges to meet a stricter standard.
A risk management systems approach should be adopted, in which combinations of practices and
technologies should be considered as potentially more effective and potentially more cost-efficient
approaches than reliance on one ballast water treatment technology. Hazard Analysis and Critical
Control Points (HACCP) has been demonstrated to be an effective risk management tool in a variety of
situations and could be applied to ballast water management.
-------
2. INTRODUCTION
2.1. EPA's Charge to the SAB
EPA's Office of Water provided the following background and charge questions to the SAB:
Ballast water is typically drawn in from surrounding ambient water and used to assist with vessel draft,
buoyancy, and stability. Almost all large vessels have ballast tanks dedicated to this purpose; some
vessels may also ballast empty cargo holds. The ballast water discharge rate and constituent
concentrations of ballast water from vessels will vary by vessel type, ballast tank capacity, and type of
deballasting equipment. Under current U.S. regulation and permitting requirements (discussed in
greater detail in the White Paper), there are existing best management practices to reduce the potential
impacts of ballast water discharges. These include ballast water exchange and salt water flushing
(collectively referred to as BWE).
While useful in reducing the presence of potentially invasive organisms in ballast water, BWE can have
variable effectiveness and may not always be feasible due to vessel safety concerns. In order to make
progress beyond use of BWE, establishing a standard for the concentration of living organisms in
ballast water that can be discharged is necessary. The United States Environmental Protection Agency
(EPA) and the United States Coast Guard (USCG) both desire a stronger federal ballast water
management program.
To help develop the next Clean Water Act Vessel General Permit (VGP), EPA needs an objective
evaluation of the status and efficacy of ballast water treatment technologies and systems that are in
existence or in the development process. A second major scientific question for regulatory agencies is to
better under stand and relate the concentration of living organisms in ballast water discharges to the
probability of introduced organisms successfully establishing populations in U.S. waters. Given the
complexity of the issue, EPA 's Office of Water is seeking advice from the Science Advisory Board (SAB)
on the first issue and the National Academy of Sciences' National Research Council (NRC) on the
second issue. In particular, EPA is seeking advice from the SAB regarding the availability and efficacy
of ballast water treatment systems in neutralizing (killing) living organisms that might be discharged
from ballast tanks. For the other NRC study, EPA has requested that the NRC broadly assess and make
recommendations about various approaches for assessing the risk of establishment of new aquatic non-
indigenous species from ballast water discharges (see attachment 2 of the White Paper for the NRC
charge).
Charge Question 1: Performance of shipboard systems with available effluent testing data5
a. For the shipboard systems with available test data, which have been evaluated with sufficient
rigor to permit a credible assessment of performance capabilities in terms of effluent
concentrations achieved (living organisms/unit of ballast water discharged or other metric)?
b. For those systems identified in (la), what are the discharge standards that the available data
credibly demonstrate can be reliably achieved (e.g., any or all of the standards shown in Table 1
5 EPA and the US Coast Guard have provided data they currently have to the panel. Where feasible, the panel is encouraged
to find additional data if they have appropriate avenues to obtain those data.
-------
of the White Paper? Furthermore, do data indicate that certain systems (as tested) will not be
able to reliably reach any or all of the discharge standards shown in that table?
c. For those systems identified in (la), if any of the system tests detected "no living organisms "
in any or all of their replicates, is it reasonable to assume the systems are able to reliably meet
or closely approach a "no living organism " standard or other standards identified in Table 1 of
the White Paper, based on their engineering design and treatment processes?
Charge question 2: Potential performance of shipboard systems without reliable testing data
Based on engineering design and treatment processes used, and shipboard
conditions/constraints, what types of ballast water treatment systems (which may include any or
all the systems listed in Table 4 of the White Paper) can reasonably be expected to reliably
achieve any of the standards shown in Table 1 of the White Paper, and if so, by what dates?
Based on engineering design and treatment processes used, are there systems which
conceptually would have difficulty meeting any or all of the discharge standards in Table 1 of the
White Paper?
Charge question 3: System development
a. For those systems identified in questions 1 a and 2, are there reasonable changes or additions
to their treatment processes which can be made to the systems to improve performance?
b. What are the principal technological constraints or other impediments to the development of
ballast water treatment technologies for use onboard vessels to reliably meet any or all of the
discharge standards presented in Table 1 of the White Paper?
c. What recommendations does the SAB have for addressing these impediments/constraints?
d. Are these impediments more significant for certain size classes or types of organisms (e.g.,
zooplankton versus viruses)?
e. Can currently available treatment processes reliably achieve sterilization (no living organisms
or viable viruses) of ballast water onboard vessels or, at a minimum, achieve zero or near zero
discharge for certain organism size classes or types?
Charge question 4: Development of reliable information
What are the principal limitations of the available studies and reports on the status of ballast
water treatment technologies and system performance and how can these limitations be
overcome or corrected in future assessments of the availability of technology for treating ballast
water onboard vessels?
10
-------
2.2. SAB's Review Process
In response to the EPA request for an SAB assessment of shipboard ballast water management systems
(BWMS), the SAB's Ecological Processes and Effects Committee was augmented with additional
experts in ballast water issues, marine engineering, and engineering treatment technologies (the Panel).
The Panel met on July 29 - 30, 2010 to receive briefings from EPA's OW, to hear public comments, and
to begin discussing the charge questions. As requested, the SAB based its assessment and advice on
information provided by OW, including the background document, Availability and Efficacy of Ballast
Water Treatment Technology: Background and Issue Paper (June 2010), and a compilation of
information and data on BWMS (described in Appendix A and Section 4 of this report). Teleconferences
were held on October 26 and November 4, 2010, to discuss preliminary texts prepared by individual
subgroups of the Panel and to hear public comments. The full Panel met again on January 25-26, 2011,
to discuss the compiled draft report and to discuss preliminary conclusions and recommendations.
Additional teleconferences were held on March 15 and March 17, 2011, to discuss final revisions to the
draft report. The Panel considered public comments provided throughout the advisory process. The
Panel differed in their views regarding several issues related to onshore treatment of ballast water in
reception facilities. The potential use of reception facilities is discussed in Section 6.4, and a review of
the literature of onshore treatment in reception facilities is provided in Appendix B. One panel member
did not concur with the final draft report prepared for quality review by the chartered SAB. Public
comments were received and considered throughout the advisory process. The chartered SAB conducted
a quality review on June 16, 2011, and approved the report with clarifying edits.
2.3. Regulatory Frameworks for Ballast Water Management
2.3.1. U.S. Federal rules
In December 2008, the EPA issued a Vessel General Permit (VGP) for discharges incidental to the
normal operation of commercial vessels, including ballast water, as authorized under the Clean Water
Act (CWA). The VGP set technology-based effluent limits for ballast water that rely on "best
management practices" and do not include a numeric discharge limit. The required VGP practices
include flushing and exchange of ballast water by vessels in Pacific near-shore voyages and saltwater
flushing of ballast water tanks that are empty or contain only un-pumpable residual ballast water in
addition to mid-ocean exchange. The VGP expires on Dec. 19, 2013.
Existing U.S. Coast Guard (USCG) rules governing ballast water, as authorized under the National
Invasive Species Act (NISA), also rely primarily on ballast water exchange. Though the BWE
provisions are not identical, the general principle of BWE as used by the EPA and USCG is very similar.
NISA generally requires vessels equipped with ballast water tanks and bound for ports or places in the
U.S. after operating beyond the U.S. Exclusive Economic Zone either to conduct a mid-ocean ballast
water exchange (BWE), retain their ballast water onboard, or use an alternative environmentally sound
ballast water management method approved by the USCG. In August 2009, the USCG proposed
revising their existing rules to establish numeric concentration-based limits for organisms in ballast
water. The proposed rule initially would require compliance with a "Phase 1 standard" that has the same
concentration limits as the International Maritime Organization (EVIO) D-2 standard (see below) and
subsequently would require compliance with a "Phase 2 standard" that is 1000 times (lOOOx) more
stringent for organisms of more than 10 |im in minimum dimension. The Phase 2 standard also contains
11
-------
concentration limits for bacteria and viruses. As of July 2011, the USCG has not finalized this rule, and
in the meantime continues to require use of BWE.
In recent years, Congress has considered but not enacted legislation that would directly set
concentration-based ballast water discharge standards, including standards that would be 100 times
(lOOx) more stringent for organisms than the USCG proposed Phase 1 standard.
2.3.2. Other Regulatory Frameworks
U.S. States
Under the CWA, U.S. states have the authority to impose their own ballast water discharge standards
through the CWA section 401 certification process that applies to federally issued National Pollution
Discharge Elimination System (NPDES) permits such as the VGP. A number of states have exercised
that authority by setting numeric limits for ballast water discharges into their waters, and these numeric
limits are included as a condition in the VGP. In addition, several states (e.g., California and some Great
Lakes states) have enacted their own independent state laws to establish ballast water treatment
standards. Thus, in practice, EPA's VGP standards establish the minimum standard for ballast water
discharges, but states retain and have exercised their authority to set standards that are more stringent.
International Standards / Treaties
The International Convention for the Control and Management of Ships' Ballast Water and Sediments
(EVIO Ballast Water Management Convention), adopted by the International Maritime Organization
(EVIO) in February 2004, contains concentration-based limits on organisms in ballast water set out in its
Regulation D-2. The treaty will not come into force unless it is ratified by additional countries, and
implementation would then require the enactment and enforcement of appropriate laws or regulations by
the countries that are party to the treaty. However, equipment manufacturers are currently designing and
testing equipment to meet the D-2 standards. These and the other main concentration-based limits for
organisms in ballast water that have been proposed or adopted are shown in Table 2-1.
12
-------
Table 2-1. The range of concentration-based limits proposed or adopted for organisms in ballast water.
"US Negotiating Position" is what the U.S. argued for in the negotiations on discharge standards for the IMO Ballast
Water Management Convention. "California Interim" and "California Final" standards refer to the limits enacted by
the state of California in 2006.
A. Concentration limits for four organism classes
Organisms > 50 fim
in minimum
dimension
Organisms > 10-
<50 fim in
minimum
dimension
Bacteria
Viruses
IMO D-2
USCG Phase 1
US Negotiating Position
USCG Phase 2
California Interim
California Final
perm
10
10
0.01
0.01
no detectable"
zero detectable*
per ml
10
10
0.01
0.01
0.01
zero detectable
per ml
no limit
no limit
no limit
10
10
zero detectable
per ml
no limit
no limit
no limit
100
100
zero detectable
"For California's interim standard for organisms > 50 um, the "no detectable" standard is not associated with a
volumetric requirement, i.e., the standard is not "no detectable living organisms" per cubic meter.
b California's final standard is set as "zero detectable living organisms for all size classes." This final standard also
does not have a volume or organism concentration associated with it.
B. Public health protective concentration limits
IMO D-2
USCG Phase 1
US Negotiating Position
USCG Phase 2
California Interim
California Final
Toxicogenic
Vibrio cholerae
per ml
.01
.01
.01
.01
.01
no detectable
Escherichia coli
per ml
2.5
2.5
1.26
1.26
1.26
no detectable
Intestinal enterococci
per ml
1
1
0.33
0.33
0.33
no detectable
2.3.3. Glossary of Terms Used
To clarify the terms used in this report, the Panel provides the following definitions.
• BWMS refers to ballast water management systems as developed by vendors for installation on
vessels for treatment of ballast water prior to discharge. In this report, "systems" refers
specifically to commercial treatment units, not to a systems approach to ballast water
management.
13
-------
• Challenge conditions refer to the challenge water (influent) conditions as specified by the IMO
and the ETV Protocol. The IMO's G8 guidelines for challenge conditions are specified in
paragraph 2.2.2.5 (for shipboard testing) and paragraphs 2.3.3 and 2.3.17 - 2.3.22 (for land-
based testing) (at
http://www.regulations.gov/search/Regs/home.html#documentDetail?R=09000064807e8904).
The EPA's Environmental Technology Verification (ETV) draft Generic Protocol for
Verification of Ballast Water Treatment Technologies, protocol for challenge conditions are
specified in § 5.2. (at
http://standards.nsf.org/apps/group public/download.php/7597/Draft%20ETV%20Ballast%20W
ater%20Prot-v4%202.pdf).
• ETV Protocol refers to the U.S. EPA Generic Protocol for the Verification of Ballast Water
Treatment Technology, Version 5.1. 2010 (EPA/600/R-10/146), U.S. EPA Environmental
Technology Verification Program, Washington, DC.
• G9 approval, both "Basic Approval" and "Final Approval": Under Regulation D-3(2) of the
IMO Ballast Water Management Convention, ballast water treatment systems that make use of
"active substances" (biocides or other potentially harmful substances) are subject to approval by
the IMO's Marine Environment Protection Committee (MEPC) with respect to active substance-
related health, environmental, and safety issues. This review and approval is conducted under the
"G9 Guidelines" developed by MEPC, available at
http://www.regulations.gov/search/Regs/home.html#documentDetail?R=09000064807e890e.
Basic Approval" requires laboratory or bench-scale testing, while "Final Approval" requires
testing an actual piece of equipment. The Group of Experts on the Scientific Aspects of Marine
Pollution (GESAMP), an advisory body established by the United Nations in 1969, conducts the
technical reviews and makes approval or denial recommendations to MEPC. MEPC then makes
the G9 approval decisions.
• IMO refers to the International Maritime Organization, a subsidiary body of the UN whose
principal responsibility is to develop and maintain the international regulatory framework for
shipping with respect to safety, environmental concerns, legal matters, technical co-operation,
and maritime security. It accomplishes this through treaties negotiated under its auspices,
including the February 2004 International Convention for the Control and Management of Ships'
Ballast Water and Sediments. The Marine Environment Protection Committee (MEPC) is the
principal EVIO committee with responsibility for environmental issues associated with shipping.
For more information: http://www.imo.org/home.asp
• IMO D-2 refers to the ballast water discharge standards (expressed as concentrations of
organisms per unit of volume) that are contained in Regulation D-2 of the IMO Ballast Water
Management Convention. The U.S. has not ratified the treaty, and additional countries must
ratify it before it enters into force. Nonetheless, the D-2 standards have had an important
influence on the design of shipboard BWMS.
• IMO D-2 / Phase 1 are ballast water discharge standards that are sometimes used in combination
because their specifications are very similar. However, to be explicit, IMO D-2 is defined as
shown above.
14
-------
USCG Phase 1 (or P-l) refers to ballast water discharge standards contained in the U.S. Coast
Guard's August 28, 2009, notice of proposed rulemaking. Because this is a proposed rulemaking
that has not yet been finalized, these Phase 1 standards are not currently (as of July 1, 2011)
legally binding. For more information, refer to 74 Federal Register 44632. The table below
contains the text of the standards as stated in IMO D-2 and in the proposed USCG Phase 1,
arrayed so as to enable their direct comparison (blanks in table are not omissions, but rather are
arranged to highlight comparison of the texts).
Table 2-2. Comparing IMO D-2 with USCG Proposed Phase I Standard
IMO Regulation D-2 Standard
USCG Proposed Phase 1 Standards
Discharge less than 10 viable organisms
per cubic meter greater than or equal to 50
micrometers in minimum dimension
For organisms larger than 50 microns in
minimum dimension: Discharge less than
10 per cubic meter of ballast water;
Discharge less than 10 viable organisms
per milliliter less than 50 micrometers in
minimum dimension and greater than or
equal to 10 micrometers in minimum
dimension
For organisms equal to or smaller than 50
microns and larger than 10 microns:
Discharge less than 10 per milliliter (ml) of
ballast water; and
Discharge of the indicator microbes shall
not exceed the specified concentrations
described in the following paragraph:
Indicator microbes, as a human health
standard, shall include:
.1 Toxicogenic Vibrio choleras (Ol
and O139) with less than 1 colony
forming unit (cfu) per 100 milliliters
or less than 1 cfu per 1 gram (wet
weight) zooplankton samples ;
.2 Escherichia coli less than 250 cfu
per 100 milliliters;
.3 Intestinal Enterococci less than
100 cfu per 100 milliliters.
Indicator microorganisms must not exceed:
(i) For Toxicogenic Vibrio cholerae
(serotypes Ol and O139): A
concentration of <1 colony forming unit
(cfu) per 100ml;
(ii) For Escherichia coli: A
concentration of <250 cfu per 100 ml;
and
(iii) For intestinal enterococci: A
concentration of < 100 cfu per 100 ml.
• lOx D-2, lOOx D-2, lOOOx D-2 refer to concentration limits that are 10 times, 100 times, or 1000
times smaller (i.e., more stringent) than the concentration limits specified in IMO D-2, for one or
both of the organism size classes in IMO D-2 (i.e., organisms with minimum dimension > 50
|im; or > 10 jim and < 50 jim). The lOx, lOOx, and lOOOx notations do not apply to the D-2
indicator microorganisms. lOOx D-2 has been discussed in other fora such as past Congressional
bills and State requirements. lOOOx D-2 has been discussed in other fora such as the potential
Phase II standards in the USCG August 2009 proposed rule or as described in state requirements
• Type approval refers to the process under which a type of equipment is tested and certified by a
Flag state or its authorized representative (such as a Class society) as meeting an applicable
standard specified in treaty, law or regulation. This testing is conducted on a sample piece of
equipment which in all material respects is identical to the follow-on production units. For the
15
-------
IMO Ballast Water Management Convention, type approval testing (sometimes called "efficacy
testing") is conducted under the G8 Guidelines described in Regulation D-3(l) of the
Convention. The guidelines require both land-based and shipboard testing to verify the
equipment's ability to meet the IMO D-2 standards. In the U.S., a generally similar type approval
procedure was proposed as part of the USCG's August 28, 2009, notice of proposed rulemaking.
• Verification Organization (VO) refers to the party responsible for overseeing the test facility's
test Total Quality Assurance Plan (TQAP) development, overseeing testing activities, and
overseeing the development and approval of the Verification Report and Verification Statement
for the ballast water treatment system. Within the ETV Program, verification organizations are
the managers and operators of the various technology centers under cooperative agreements with
the EPA (U.S. EPA 2010).
2.4. Applying Risk Assessment Principles to Ballast Water Management
The charge to the Panel focused primarily on whether BWMS could meet specific discharge standards,
now or in the future. However, in its assessment the Panel found that combinations of practices and
technologies should be considered as potentially more effective than reliance on BWMS technology (see
section 6.5). Therefore this section sets the stage by exploring the use of risk assessment as a way to put
strategies for treatment of ballast water into a probabilistic decision-making process. This process should
be applied to the entire system of ballast water management and not to just one technique, device or
practice. Each step of the process, from taking on ballast water at the port of origin to its discharge into
the receiving port, depends upon the others. Risk assessment is a means of treating the entire ballast
water management process in a holistic fashion and changes to each step can be evaluated within a
defined risk-based process. A holistic approach to ballast water management includes the regulatory
environment, the training of personnel, quality control, and environmental sampling. Risk assessment
also provides the framework for risk management. One such framework is the Hazard Analysis and
Critical Control Points (HACCP). The HACCP process and its application to ballast water management
are described in section 6.6.
2.4.1. Risk Assessment of Nonindigenous Species
The establishment of a nonindigenous species is the joint probability of how often species are
introduced, the initial population size necessary to ensure reproduction, and the probability that
organisms would find a suitable environment for propagation. This joint probability is low for any one
species or specific shipping event. However, a large number of species can be transported via ship, and
thousands of ships arrive at U.S. ports, creating a substantial probability that a nonindigenous species or
a new pathogen will become established. Given that shipping is a major industrial activity that will
continue far into the future, even a small probability for each ship and for each species will result in
successful invasions. The goal of a ballast water management (BWM) program is to lower that
probability, especially for particularly damaging species and pathogens. For a BWM program to be
successful, the goals need to be specific and measurable, and the operational context needs to be
understood. First, a model of the relationship between the number of organisms in ballast water and the
likelihood of invasion or infection by a pathogen needs to be derived.
16
-------
Probabilistic Approach to Deriving Risk Due to Nonindigenous Species
A foundation for the risk assessment for invasive species has been established (Drake 2004, 2005;
Landis 2004). The process is density-dependent, with an increase in the density of organisms leading to
an increase in population growth rate (Drake 2004). Modeling also has demonstrated the importance of a
beachhead effect, where a population increases in a relatively isolated habitat patch before spreading to
the remainder of the environment (Deines et al. 2005). Both Drake (2005) and Deines et al. (2005)
recognize the importance of eliminating the organisms during the initial invasion event or destroying the
beachhead in order to implement control.
Deines et al. (2005) used spatially explicit stochastic difference models and Drake and Lodge (2006)
employed stochastic differential equations to model invasion events. In both studies the importance of
understanding the stochastic aspects of colonization, the initial population size (propagule pressure) and
density dependent effects on population growth were important in determining the probability of
invasion. The actual dynamics of invasions were sensitive to initial conditions. The combination of
stochastic and non-linear components results in a distribution of outcomes in both studies. This means
that any relationship between propagule pressure and probability of invasion will be a distribution of
outcomes. These foundations have been used to estimate risk in case studies (Drake et al. 2006; Colnar
and Landis 2007). A risk assessment for managing ballast water invasion should have as its foundation
the stochastic-non-linear nature of the invasion process.
Propagule Pressure and Invasion Relationships
It is possible to derive relationships between the number of organisms with an invasive potential
(propagules) and the probability of an invasion over a specified amount of time. Such a relationship is
described by the upper panel of Figure 2-1. It is assumed that the greater the number of propagules, the
greater the probability that an invasive potential will be established. In this instance, it is assumed that
the relationship is sigmoidal and has a threshold, but a number of curves are possible and the actual
curve may be specific to the type of organism or environment. The solid line represents the central
tendency of the relationship, with the dashed lines representing confidence intervals. Note that the
confidence intervals include the possibility of a successful invasion even without propagule pressure
from ballast water and also the likelihood of no invasion even with organisms escaping. After all,
organisms can come from a variety of sources other than ballast water.
The lower panel of Figure 2-1 illustrates a process for setting targets for the number of organisms in
ballast water. First a policy decision is made about an acceptable frequency of successful invasion over a
specified amount of time. Existing information may be inadequate, so derivation of this frequency may
require an assessment tailored to a specific habitat, species, endpoint, and location. Reading across the
graph to where this rate intersects with the concentration-response curve gives the numbers of organisms
corresponding to the low, expected and high values. Trade-offs can then be made on the likelihood of
success in meeting the specified target and the costs of achieving the goal.
Although these graphs were drawn to express the relationship with one species of concern, similar plots
may be derived for discharges with a large number of species. The greater the diversity of species and
life stages, the greater the probability of an invasion by at least a single species.
17
-------
100
B
g
'w c;n
CD ou
>
c
.Q
CD
8
Q.
0
Expected value
— — — • Confidence limit
Propagule Pressure (number of organisms)
100
-------
2.4.2. Ballast Water Management Goals and the Decision Making, Risk Assessment Context
In order to evaluate the various types of BWMS, it is important to understand how they fit into a
decision-making context. This means the management goal has to be clearly defined as in the graphical
model (Figure 2-1). In addition, the effectiveness of ballast water treatment has to be evaluated within
the context of a ship carrying cargo, human food and waste, and many organisms attached to the hull.
The sea chest (a portion of the ship where seawater can be loaded or discharged) also can be a source of
nonindigenous species. There is also the possibility of human error in the treatment process that may
lead to the escape of organisms or the release of toxic materials. Each of these items will be covered in
the paragraphs below.
The Goal: What Does "Zero" Mean?
What does "zero" discharge of nonindigenous species and other organisms mean as a goal, since such a
value is essentially not measureable directly (see section 3)? The required sample volumes are
enormous, there are refugia from treatment within the ballast water tanks, and the discharge is into an
environment with multiple sources of invasive species. Operational definitions are very important and
may prove more useful in making a decision about ballast water treatment options. For example, does
"zero" mean that a discharge from a specific ship will contain no organisms that will colonize or infect a
port environment for that one particular combination of disinfection treatment and vessel discharge?
This is a very specific criterion but it is not necessarily protective. Furthermore, given the logistics
needed to sample and enumerate organisms in a discharge, it will not be possible to meet this
requirement for every discharge of every ship.
On the other hand, "zero" could mean the treatment technology or system will prevent introduction of a
harmful invasive organism or disease to that port over a 10-year period. This is a very different criterion,
a performance-based requirement that states the goal (no invasion or infection) over a specified time
frame. Individual treatments on certain ships may fail, but an overall system would ensure that any
colonizing organisms were quickly eradicated or that other methods would be employed to prevent their
propagation. These two "zero" goals are very different and each puts onboard or land-based treatment
options into specific and differing contexts. In order to rank the various technologies and treatment
systems, therefore, the specific goals of the program need to be carefully defined.
There is also the question of specific goals for the protection of the port from nonindigenous species and
pathogens. Are there specific requirements for each category of organism or is a combination approach
to be attempted? Consider pathogenic organisms as an example. In ballast water, a large proportion of
the organisms likely are not pathogenic, but the human welfare implications may be higher for the
pathogenic organisms than non-pathogenic organisms. Is the goal protection against human pathogens or
those pathogens that may infect shellfish and fish populations, destroy important sea grass beds, or other
segments of the ecological structure of the receiving port? Depending upon the specific policy goals,
different propagule pressure-infection relationships may need to be considered.
Context of a Cargo or Tanker Ship and the Port Facility
As the specifications for the treatment process are made explicit, it is also important to understand the
context of a ship and its port facility. Ballast water would be only one of the potential sources of
nonindigenous species and pathogens brought by a vessel. Ships contain cargos of varied types, crew,
19
-------
food, human waste, and hull fouling organisms, each of which also could be sources of invasive species.
The port facility may also contain a variety of other vessels that may be sources of nonindigenous
species and pathogens. Understanding the efficacy of the treatment program needs to be placed into this
broader context.
Cargo may contain insects, fungi, seeds and spores that can be released to the environment as the cargo
is unloaded or transported. Food can be another source of nonindigenous species, especially if living
organisms are transported. Human waste can be a source of pathogens, but can be disposed of using
appropriate facilities. Fouling of the hull of a ship can be a source of nonindigenous species or
pathogens depending upon the origin of the ship, route and time of transit, and the effectiveness of the
anti-fouling paint and the overall condition of the hull. The sea chest is a repository of organisms from
across the travels of the vessel.
A confounding factor is that a number of other vessels will use the same port facilities, and all are
potential sources of invasion. Fishing fleets and pleasure craft, for example, often take very long
voyages and may transport nonindigenous species to a harbor. Also, these "other" vessels exist in
regulatory environments different than those of cargo ships, barges, and tankers, regulations that may be
less restrictive with respect to the transport of nonindigenous species. Although not directly affecting the
infection potential of any single ship, these "other" vessels can confound determination of treatment
effectiveness or identification of an invasive species' source. So although there may be zero propagules
in ballast water discharged at a facility, there will remain some probability of an invasion at the port.
Hence there is a non-zero confidence interval in the example considered in Figure 2-1.
The risks due to invasion are not the only risks to be considered in BWM. It will be important to assess
the potential impacts of decontamination and the effluent upon the environment. Does disinfection for
pathogens increase the risk to the environment from the treatment? The number of ships that use a port
may also contribute to the trade-off. Decontamination activities that release an effluent with some
residual toxicity may not pose an important risk to a facility that has a low volume, but may be
important in a busier port. Some ports are very specialized. Port Valdez, Alaska specializes in the
shipping of crude oil and some oil product. Cherry Point, Wash, is a port that currently receives crude
from a limited number of sites to the refineries and bauxite for the smelter. Other facilities, such as New
Orleans or Seattle-Tacoma, receive a variety of container ships and cargoes from across the world.
Shipboard emergencies, accidents, human error, and equipment failure should be considered in the risk
analysis and decision-making process. At times, weather conditions or shipboard emergencies may
preclude the operation of shipboard treatment facilities. Operator error or equipment failure may happen
on shipboard or on-shore facilities just as it does in waste-treatment facilities. However, in wastewater
treatment facilities, strong programs of operator training and certification are established, unlike for
shipboard BWMS. In parts of the U.S., hurricanes and northeasters can damage ships and on-shore
equipment. No matter the weather, accidents and equipment failure will occur and will introduce
nonindigenous species to a port facility. Maximizing reliability of the BWMS should be an important
part of the risk analysis process.
BWM in an Overall Management Program
Large-scale establishment of species have occurred from what appear to be multiple invasions. Kolar
and Lodge (2001, 2002) and Kolar (2004) describe examples for the Great Lakes in which populations
20
-------
of European fish have been established from multiple invasion events. European green crab was
established in San Francisco in the late 1980s and the species has spread north along the west coast
(Behrens and Hunt 2000). Invasions take time, often decades, are often due to multiple releases, and are
difficult to control once established. A BWM strategy to decrease the rate of successful invasions should
be part of an overall plan for the reduction of invasion events, monitoring, containment and eradication.
Emphasis only on one aspect, the initial invasion event, is not likely to reduce the risk of successful
invasions to an acceptable probability. The Hazard Analysis and Critical Control Points (HACCP)
approach incorporates the context of the risk, potential points of control, and is very flexible in
application. HACCP and its use for the management of invasive species are discussed in Section 6.6.
Summary
In summary, decisions about approaches to ballast water management can be viewed within a risk
assessment framework. This framework should incorporate the following features:
• Recognition of the stochastic and non-linear nature of the invasion process;
• Clear definition of management goal needs; and
• Evaluation of the effectiveness of BWMS within the context of other sources of invasive
species on the vessel, the treatment system, and the specific receiving habitat.
A ballast water management strategy to decrease the rate of successful invasions should be part of an
overall plan to reduce invasion events, and their subsequent monitoring, containment, and eradication
activities. Emphasis only on one aspect, the initial invasion event, is not likely to reduce the risk of
invasions to an acceptable probability. Such management systems are addressed in Sections 6.5 and 6.6
of this report.
21
-------
3. STATISTICS AND INTERPRETATION
3.1. Introduction
A consideration of statistical issues encountered in testing performance of BWMS is essential to the
credible evaluation of the performance data - whether that evaluation is done by expert panels, testing
facilities, or regulators. This section presents key statistical considerations relevant to conditions under
which BWMS performance is evaluated. A more detailed discussion of these issues is provided in
Appendix C. Testing conditions include the need to sample large volumes of water, particularly for the
size class of organisms > 50 jim in minimum dimension (nominally zooplankton, referred to hereafter as
'zooplankton-sized organisms') and to apply statistical methods that can quantitatively assess the
confidence of test results obtained from counts of low numbers of organisms. These discussions pertain
to both land-based and shipboard verification testing to determine conformity to a given performance
standard in a type approval process; the same statistical theory applies to compliance testing by port
state control officers for compliance or gross non-compliance (e.g., exceedance of a standard by orders
of magnitude).
Credible testing requires the following process: Water must be collected and filtered to concentrate
organisms into a manageable volume. The volume of ballast water carried by commercial ships ranges
from a few thousand m3 to more than a hundred thousand m3. The volume that must be sampled
following treatment is a small fraction relative to that total volume but, nonetheless, a large volume must
be filtered to determine the number of live zooplankton-sized organisms. This size class has the lowest
concentration threshold - that is, organisms per m3 vs. organisms per ml in the other two size classes -
and represents the most challenging size class in terms of sampling to achieve statistical rigor. Hence,
many of the examples in the following discussion will focus on zooplankton-sized organisms. The
required sample volumes for these organisms are in the range of five to tens of m3; the latter
approximates the volume of a city bus. In all size classes, subsamples of the concentrated volume are
analyzed for viable (living) organisms, because all standards are based on the number of organisms
surviving the treatment method. Once these counts are in hand, how reliably they portray conditions in
the ballast water discharge must be determined. To accomplish this task, the live organism counts are
analyzed using statistical methods to assess the uncertainty associated with the counts.
Assessing uncertainty in test results requires accounting for the spatial distribution of zooplankton-sized
organisms in the sampled volume of water. Different probability distributions apply depending upon
whether organisms are randomly distributed throughout a sample or are aggregated. Therefore, this
section illustrates how the use of appropriate probability distributions can characterize the level of
reliability in taking the important inferential step from observing actual organism counts to determining
whether a stated standard has been met.
3.2. Assessing Whether Ballast Water Standards Can Be Met — The Statistics of Sampling
Without a well-defined, rigorous protocol based upon probability sampling, any standard will be
difficult to assess and defend, and it will be impossible to compare the effectiveness of different BWMS.
To outline what a sampling scheme might entail, and what sorts of information it would yield, it is
necessary to investigate the probabilistic characteristics of plankton in ballast water.
22
-------
Organisms can have one of two spatial characteristics: they can be randomly dispersed or clumped
(aggregated) (see Lee et al. 2010). Because any sampling protocol is a function of the organisms' spatial
distribution, it is critical to understand the distribution in the tank and discharge pipe and then sample
accordingly. For randomly distributed organisms that are not abundant, the Poisson distribution can be
used to estimate probabilities and conduct statistical power analyses (the probability that the sampling
will find a vessel in or out of compliance when that is the case). Other hypervariable discrete alternatives
to the Poisson distribution are available, such as the Poisson Log Normal and Poisson Inverse Gaussian
distributions. The Panel chose to focus on the Poisson distribution because statisticians examining
samples of treated ballast water have used the Poisson distribution, and the theoretical determination and
empirical data collected thus far support its use. The Panel notes, however, that the negative binomial
distribution is appropriate as the underlying statistical model for concentrations of organisms that are
spatially aggregated.
3.2.1. The Poisson Distribution
Theoretical Considerations
The Poisson distribution has the property that its variance is equal to its mean, resulting in an increase in
variability at higher densities. One way to assess whether the Poisson distribution is appropriate is to
calculate the variance-to-mean ratio and compare it to 1.0. If a Poisson distribution is used, a single
representative sample must be collected. To meet this requirement, the EPA Environmental Technology
Verification (ETV) Generic Protocol for the Verification of Ballast Water Treatment Technology (U.S.
EPA 2010) specifies that the sample be collected continuously over the entire discharge of the ballast
tank and in an isokinetic manner. Assuming a given concentration, one can calculate the sample volume
needed to guarantee a stated probability of finding at least a single planktonic organism (plankter) in that
volume. An underlying assumption is that organisms are randomly distributed. Spatially aggregated
populations present additional difficulties (see below), but if all organisms are counted in a continuously
and isokinetically drawn representative sample, the issue of spatial distribution can be minimized or
eliminated.
A major challenge of sampling at low organism concentrations is that many samples will have
zero live organisms because the few live organisms present are missed. To improve the
probability of detecting them, impractically large volumes must be sampled and excellent
techniques must be used to enable detection (Figure 3-1). For example, when the water to be
evaluated has a known concentration of one organism in 1 m3, the probability of a 1 m3 sample
containing zero organisms is 36.8%; if the known concentration is 0.01 organism in 1 m3
(equivalent to one organism in 100 m3), the probability of obtaining a sample with zero
organisms is -99% (Lee et al. 2010; Appendix C). Furthermore, "If a small volume is used to
evaluate whether the discharge meets a standard, the sample may contain zero detectable
organisms, but the true concentration of organisms may be quite high.... The general point is
that more organisms may be released in ballast discharge using a stringent standard paired with
a poor sampling protocol than a more lenient standard paired with a stringent sampling
protocoT' (Lee et al. 2010, p.72, emphasis added).
23
-------
A
f£i:&&?\tjl§?!
t° * 4 °°°%t4i2'b, !^°% ""* °?*°. g"K°°°
$t "••"*'? V*****.?***^? "* ° r
V» *° "°'iS * °Irf***°'«'>°l> °«* c?'»*<'°°*°
f s*°&°» °°f &*«• ,° V 0°f°° '°*°"
iT °° *'»°*'i'o*°'° °* »" C$1\ * °"^ °-t° » ° "y*
t *• °f, 7" s\ . • * . CA»» if .„* */iC°°*»^
rjCjo;t:6'"EJ;; a0p°;J° .'^.
0.12
0.10
;& 0.08
ro
^ 0.06
^ 0.04
0.02
Onn
r
1
048
10 org m-2
1
L
12 16 20 24 28
B 1m
° "•"- '"""" 'in
Q ° / "• .0 E— 1
0 p *, ^
f D * a
0 0 * *
* g °
" *l^^^ *"
bli ° ° :
c a
"O. D ::: ;;':
0.35 -i
0.30
^,0.25
S 0.20
JD
o 0.15
D_
0.10
0.05
0.00
1 org m-2
i i .
0123456
C
'— '
1 — i
"— '
| |
0.8 -,
0.6
| 0.4 -
o
°- 0.2
0 0
0.01 org m-2
0 1 2
Number of organisms in sample
Figure 3-1. Illustration of the need to sample very large volumes to detect low concentrations of organisms
present, assuming random distribution: Probability distributions for random samples of 1 m2 for a randomly
distributed population with 10 (A), 1 (B), or 0.01 (C) organisms m"2. Red squares represent random samples. The
data are displayed in terms of area with units of m2, but the probabilities are the same for volumes. Plots on the
right indicate the probability that aim2 sample will contain a given number of organisms. At low
concentrations, the concentration of organisms likely will be estimated as 0 organisms m"2, unless very large
volumes are sampled. (Source: Lee et al. 2010).
The ETV Protocol stipulates that biological samples (for all three size classes) should be continuously
acquired on a time-averaged basis from a sampling port positioned in fully turbulent flow (U.S. EPA
2010), and are thus representative of the entire volume to be sampled. Organism abundance in BWMS
testing can be statistically represented by the Poisson distribution and, therefore, the cumulative or total
24
-------
count is the key test statistic (Lemieux et al. 2008; Miller et al. 2011). A Chi-square distribution can also
be used to approximate confidence intervals (CIs). However, experimental validation must be obtained
to ensure that testing organizations can accomplish detection of live organisms with quantified
uncertainty (see section 6.2.4 on viability).
The available methodologies for testing compliance with the IMO standards for zooplankton-sized
organisms are at or near the analytic detection limits. For example, based on the Poisson distribution for
a 95% CI from the Chi-square distribution, 30 m3 (30,000 L) must be sampled in order to find and count
< 10 organisms m"3 with the desired level of precision (U.S. EPA 2010; Appendix C).
The ETV Protocol provides examples of the sample size needed to provide the level of precision needed
to achieve a 95% upper confidence limit that is no more than twice the observed mean and does not
exceed the targeted concentration (Tables 3-1 and C-l, Appendix C; U.S. EPA 2010). If the volume of
subsample that is analyzed is increased, then validation experiments should be conducted to ensure that
counting accuracy is acceptably high. The Poisson distribution assumption still applies to organisms in
the next smaller class (here, referred to as "protist-sized" organisms, or organisms >10 jim and < 50 jim
in minimum dimension), and the ETV Protocol provides examples with a more stringent level of
precision than is used for the larger size class (Table C-l, Appendix C; U.S. EPA 2010). At present,
confirmation of the Phase 1 standard (< 10 protist-sized organisms mL"1) represents the practical limit
that can currently be achieved by testing facilities in the U.S. (e.g., MERC 2009a, 2010a, 2010b; Great
Ships Initiative 2010). In addition, determining viability of protist-sized organisms remains problematic
because many organisms do not move during time scales over which they are observed (as do many
zooplankton; see section 6.2.4 for a discussion of viability determination).
Table 3-1. Sample volume of treated ballast water required relative to treatment standards for organisms >50 um
(nominally zooplankton), assuming that the desired level of precision of the estimated density is set at the 95%
confidence interval of the Poisson distribution (= twice the observed mean and not greater than the standard limit).
These are the required whole-water sample volumes that must be concentrated to 1 L as a function of N, the number
of 20 1-mL subsamples analyzed. (Source: U.S. EPA 2010).
N= 1 3 5
Concentration (i.e., performance Sample Volume Required
standard] (individuals nr3) (m3)
0.01
0.1
1
10
60,000
6,000
600
60
20,000
2,000
200
20
12,000
1,200
120
12
It is expected that the statistics governing the smallest size classes—the < 10 jim protists proposed
below, and the indicator and pathogenic bacteria (Escherichia coli, Enterococci spp., and Vibrio
cholerae)—will be similar to the two size classes discussed here. That is, treated samples will be well-
mixed and will have sparse populations collected from representative samples of ballast water; thus, the
Poisson distribution would be applicable.
Laboratory experiments
Laboratory experiments with protist cultures support use of the Poisson distribution (Nelson et al. 2009;
Steinberg et al., accepted with revisions; Appendix C). Based on results of laboratory experiments with
25
-------
two sizes of standardized microbeads at different densities as imperfect proxies of protist- and
zooplankton-sized particles, Lemieux et al. (2008) recommend that samples for analysis of protist-sized
organisms be concentrated by at least a factor of five, and that at least four replicate counting chambers
should be analyzed for acceptable accuracy and precision (see Appendix C for details). Furthermore,
Lemieux et al. (2008) determined that the zooplankton size class requires a sample size of greater than 6
m3 concentrated to 0.5 L and analysis of at least 450 1-mL aliquots. Because these higher concentration
factors are likely unrealistic, Lemieux et al. (2008) suggest that larger sample sizes and improved
analytical methods be used.
When concentrations are close to the performance standard, a single sample may require too large a
volume of water to be logistically feasible. In that case, complete, continuous time-integrated sampling
(with the entire volume analyzed) and combining samples across multiple trials can improve resolution
while maintaining statistical validity. To that end, Miller et al. (2011) applied statistical modeling (based
on the Poisson distribution) to a range of sample volumes and plankton concentrations (see Appendix C
for details). They calculated the statistical power of various sample volume and zooplankton
concentration combinations to differentiate various zooplankton concentrations from the proposed
standard of < 10 live organisms m"3. They concluded that three trials of time-integrated sampling of 7 m3
from a ship's ballast water discharge theoretically can result in 80% or higher probability of detecting
noncompliant discharge concentrations of 12 vs. 10 live organisms m"3. Thus, pooling volumes from
separate trials will allow lower concentrations to be differentiated from the performance standard,
although the practicability and economic costs of doing so have not been evaluated.
It is important to note, however, the practical limits of increasing statistical sample sizes that may
already tax the capabilities of well-engineered land-based ballast water test facilities used in verification
testing. Shipboard testing in the U.S. has been done on a pilot scale to date (i.e., the USCG Shipboard
Testing Evaluation Program, STEP), but pooling volumes from multiple trials might also be problematic
on vessels used for shipboard verification testing and compliance testing. According to Table 3-1, to
meet a standard 10 times more stringent than D-2/ Phase 1 would require anywhere from 120-600 m3 of
whole-water sample volumes, which is impracticable; test facilities in the U.S. typically analyze ~5 m3
of water per test (e.g., MERC 2009a, 2010a, 201 Ob; Great Ships Initiative 2010).
Additional challenges of sampling large volumes
As outlined in Lee et al. (2010), the detection of viable organisms at very low concentrations is a major
practical and statistical challenge, partly because of the inherent stochasticity of sampling. Due to
random chance, the number of organisms in multiple samples taken from the same population will vary.
In addition, very large volumes of water must be sampled in order to accurately estimate the organism
densities. Three other considerations merit attention.
First, statistical approaches rest upon the premise that the samples realistically represent the actual
concentrations of organisms discharged. This premise is based upon two assumptions: all organisms are
detected in the analyzed volume (no human or equipment errors), and organisms are randomly
distributed in both ballast tanks and discharge water. Neither assumption will be true all of the time.
Human and equipment errors will occur, and organisms are typically "patchy" or non-random within the
water column of a tank or the stream of a large-volume discharge (Murphy et al. 2002; U.S. EPA 2010).
If appropriate quality control and assurance procedures were used in collecting the data, then ideally
human error and equipment malfunction would have been accounted for. The degree of randomness can
26
-------
be determined by calculating the variance-to-mean ratio from multiple samples and comparing the
resulting ratios to 1.0. Additionally, analysis of a time-averaged sample taken continuously in an
isokinetic manner renders this assumption moot.
Second, the logistics of managing large sampling containers, sample transport costs (since samples
usually are not processed aboard ship), analytical supplies, and personnel time would make it
impractical to process all of the volume of, for example, even one 100 m3 sample, much less multiple
samples, especially in type approval of BWMS when multiple successful tests are required. Lee et al.
(2010) calculated the probability of finding one or more organisms in a sample for a series of organism
concentrations and sample volumes (Table C-2, Appendix C). These calculations show that 100 L of
ballast must be sampled to have a > 99% probability of detecting at least 1 zooplankton-sized organism
when the true concentration is 100 organisms per m3. When small sample volumes are collected, the
probability of detecting an organism is low even at relatively high organism concentrations; for
example, organisms will be detected in fewer than 10 percent of subsamples if a 1-L sample is taken and
the "true" concentration is 100 organisms m"3. Lee et al. (2010) then estimated the upper possible
concentration (UPC, upper 95% CI) of organisms actually present in ballast water from the number of
organisms in a sample volume based on the Poisson distribution. Zero organisms detected in a 1-m3
sample could correspond to a true concentration of organisms in the ballast tank of up to -3.7 organisms
m"3; if the sample volume is only 1 L, zero organisms detected could correspond to a true concentration
of-3,700 organisms m"3 (Table C-3, Appendix C).
Third, in the above analyses, the true concentrations are known. The goal in sampling unknown
concentrations is to accurately assess whether a given BWMS treats water with true organism
concentrations that meet a given performance standard. Inherent stochasticity of sampling may result in
an indeterminate category, as well, and the probability of obtaining an indeterminate evaluation
increases with decreasing sample volume and increasing stringency of the ballast water standard (Figure
C-l, Appendix C). For example, it would be necessary to sample -0.4 m3 of ballast water to determine
whether the D-2/Phase 1 standard of < 10 zooplankton-sized organisms m"3 was met if fewer than
approximately 10 organisms were observed in the sample (Figure C-1B, Appendix C).
3.2.2. Spatially Aggregated Populations - Negative Binomial Distributions
This section illustrates how difficult statistical analyses can become when working with spatially
aggregated populations and emphasizes the gains made from doing a complete count of a representative
sample that has been continuously and isokinetically taken.
If organisms are aggregated rather than randomly distributed in a ballast tank, a different statistical
approach is required. For aggregated populations, the variance exceeds the mean (negative binomial
distribution, o > u); thus, as the variance increases, the number of organisms in a random sample is
increasingly unpredictable. Lee et al. (2010) recommend use of the negative binomial distribution to
model aggregated populations.
Because it is more difficult to accurately estimate the true concentration, more intensive sampling is
required. For a randomly distributed population with a true concentration of 1 organism m"3, -37% of
the subsamples from aim3 sample of treated ballast water would contain zero organisms; for an
aggregated population with a dispersion parameter of 0.1, -79% of the subsamples would contain zero
organisms (Figure C-2, Appendix C; Lee et al. 2010). As aggregation increases, the probability of
27
-------
samples containing either zero organisms or large numbers of organisms relative to the true
concentration also increases (Figure C-3, Appendix C). Thus, large numbers of subsamples from large
sample volumes must be taken to account for aggregated populations; otherwise, there will be a high
probability that the concentration estimates from sample analyses will be either much lower or much
higher than the true concentration.
Determination of whether a population is aggregated is complicated, since it is the scale of the
aggregation pattern relative to the size of the sampling unit that controls the estimate of aggregation
(Figure C-3, Appendix C). Lee et al. (2010) recommend the Taylor power law (Taylor 1961) as an
alternative to the negative binomial, because it can accommodate a wider range of aggregated
distributions than the negative binomial.
Overall, the degree of aggregation represents challenges in sampling sufficiently large volumes of
ballast water to determine whether a given BWMS passes or fails to meet standards more stringent than
the present EVIO guidelines, even if the true concentrations of organisms are 10 to 1000 times higher
than the performance standard. This remains a problem in quantifying many protist-sized organisms, but
becomes less of a problem with very small organisms such as bacteria, which have a tendency to clump
but are effectively counted as colonies and not individuals. However, laboratory experiments provide
data supporting use of the Poisson distribution to analyze ballast water samples (Lemieux et al. 2008;
Nelson et al. 2009; see Appendix C).
3.3. Interactive Effects
A final consideration regarding statistical analysis is the potential for covariance, or interactive effects
among environmental conditions - for example, a treatment system may perform well under high-
temperature or high-biomass conditions, but not both (Ruiz et al. 2006). To address this problem,
covariate measurements should be carefully addressed in experiments, and treatment evaluations should
consider the potential for interactions and target tests of especially challenging combinations.
3.4. Certainty of Results
As with all statements that are based upon statistical sampling, there is always a stated non-zero error
probability associated with a particular statistical conclusion. Thus, without a complete census of a
ship's entire volume of ballast water, one can never claim to be 100 percent certain that the
concentration of live zooplankton-sized organisms is below a discharge standard. Available
methodologies to test D-2/ Phase 1 compliance are presently at or near analytic detection limits for the
two largest organism size classes. The D-2/ Phase 1 performance standards are measurable at present
based on land-based and shipboard testing approaches, however, new or improved methodologies will
be required to increase detection limits.
From the examples above, statistical theory shows it is theoretically possible to detect adherence to a
very low discharge standard (e.g., lOOOx more stringent than D-2/ Phase 1). Measuring to a lOOOx more
stringent standard, however, is impracticable at the present time because of the logistics of collecting,
reducing, and counting organisms in all size classes within the volumes of water required. Detecting
achievement of a standard ten times more stringent may be possible (although consensus has not been
reached on this), but it seems unlikely for the reasons mentioned above that detecting achievement of a
lOOx more stringent standard is possible.
28
-------
3.5. Conclusions
• Rigorous statistical sampling protocols (that may include consideration of the spatial
distribution of plankton in ballast water) and subsequent statistical analysis are required to
assess whether a BWMS meets desired performance standards.
• Detecting organisms in low abundance is a difficult problem that requires sampling of very
large volumes of water, especially for the zooplankton-sized organisms (> 50 jim).
• The initial sample volumes needed are a function of the degree to which the sample volumes
are concentrated, the performance standard, and the desired level of confidence (e.g., 95%,
which is used most often in ecological investigations).
• The Poisson distribution is recommended as the model for statistical analysis of treated water
samples.
• Available methodologies to test D-2/ Phase 1 compliance are presently at or near analytic
detection limits for the two largest organism size classes. New or improved methodologies
will be needed to increase detection limits.
• The D-2/ Phase 1 performance standards are measureable at present. Because of the logistics
of collecting, reducing, and counting organisms in all size classes within the volumes of
water required to achieve a standard 1000 times more stringent than the D-2/ Phase 1
performance standard, measuring adherence to a lOOOx more stringent standard may be
impracticable. Measuring adherence to a standard that is lOx more stringent may be possible
if a continuously isokinetically taken representative sample is used. It seems unlikely, for
reasons mentioned above, that a lOOx more stringent standard could be measured at present.
• Statistical conclusions at a stated confidence level always have an associated error
probability; thus, "100 percent certainty" is not statistically possible without a complete
census of a ship's entire ballast water contents.
29
-------
4. PERFORMANCE OF SHIPBOARD SYSTEMS WITH AVAILABLE
EFFLUENT TESTING DATA
4.1. EPA's Charge Questions
This section responds to Charge Questions 1 and 2, which ask the Panel to assess the documented
performance of existing BWMS in terms of quality of discharged ballast water, and to assess the likely
future performance of BWMS based on their design and treatment processes.
4.2. Assessment Process
A subgroup of the Panel led the assessment of BWMS technologies. The subgroup considered only the
information compiled by EPA through solicitation of various Maritime Administrations that have
granted Type Approval certifications, direct communication with developers and manufacturers of
BWMS, and searches for publically available sources (such as journal or conference publications and
third-party reports available through the Internet). This information (listed in Appendix A) included data
packages, reports, publications, certification documents, and other available information on the
performance of BWMS.
Three subgroup members independently examined in detail all data packages, with two other members
providing review oversight and quality control. The type, amount, and quality of material in the data
packages varied—some contained only a type approval certificate, while others included land-based and
shipboard testing methods and data, documentation of G9 approval, a type approval certificate, and
press releases describing the sale of systems for use on commercial vessels. The Panel notes that BWMS
are still evolving with an ever-growing number of manufacturers developing systems. Thus, this analysis
represents a snapshot in time. No new data or information was considered beyond packages submitted to
the SAB by December 1, 2010.
4.3. Assessing the Reliability of Existing Data
The three primary reviewers independently scored each package as having "reliable" or "unreliable"
data. To earn a "reliable" rating, the data package had to include, at a minimum, methods and results
from land-based or shipboard testing. A BWMS holding a certificate of type approval without
supporting testing data was scored as having "unreliable" data because it was impossible to determine
the validity of the testing procedures and, therefore, the validity of the data. If a BWMS's data package
included one or more test reports, the data package was examined according to the following criteria:
• The operational type of system (e.g., deoxygenation + cavitation) was determined to be
generally appropriate for shipboard use (e.g., can it meet required flow capacities, size, and
power requirements).
• The technical literature supported the fundamental use of the technologies used (e.g., is it
well documented that using the approach will safely and effectively remove, kill, or
inactivate aquatic organisms).
• Laboratory testing was conducted with "reasonable and appropriate methods" (i.e., methods
commonly used in aquatic studies or alternative methods that appear rigorous and equivalent
to a standard, common approach).
30
-------
• Land-based testing was conducted with reasonable and appropriate methods; sample number
and size were appropriate; sample collection and handling was appropriate and documented;
analytical facilities were adequate; IMO or ETV (v. 5.1) challenge conditions were met; if
necessary, toxicological studies were conducted and demonstrated environmental safety; a
QA/QC policy was in place and followed; and ultimately, land-based testing produced
credible results.
• Shipboard testing was conducted with the same considerations as land-based testing
(described above) and produced credible results.
• If an active substance was included, the BWMS had credible toxicity and chemistry data and
G9 Basic approval or G9 Final Approval (which requires Basic approval).
• The BWMS had a type approval certificate.
• The BWMS was in operational use (i.e., not used only during shipboard type approval
testing) on one or more active vessels. A BWMS not yet having operational systems onboard
vessels was not automatically categorized as having "unreliable" data, but this information
was useful.
It is important to note that if the data packages were deemed "reliable, it was assumed that all protocols
and methods were followed exactly as described. For data packages that included clear QA/QC
procedures, there was a higher level of certainty that this was the case. In the absence of QA/QC
documentation, which was the case for most data packages, the level of rigor in following the protocols
and methods described was unknowable.
4.4. Assessing the Ability of BWMS to Meet Discharge Standards
For BWMS with reliable data, the system's ability to meet four discharge standards—IMO D-2/ USCG
Phase 1 and lOx, lOOx, and lOOOx more stringent than IMO D-2/USCG Phase 1—was determined, again
independently, by the three primary reviewers.
• lOx was evaluated based on BWMS ability to reduce concentrations of living organisms: (a)
>50 |im in minimum dimension to below 1 per m3, (b) >10 to <50 jim in minimum
dimension to below 1 per ml, and (c) a decrease in total bacteria.
• lOOx was evaluated based on BWMS ability to reduce concentrations of living organisms: (a)
>50 |im in minimum dimension to below 1 per 10 m3, (b) >10 to <50 jim in minimum
dimension to below 1 per 10 ml, and (c) and a significant reduction in total bacteria.
• lOOOx was considered the equivalent of the USCG Phase 2 standard, including: (a) >50 jim
in minimum dimension to below 1 per 100 m3, (b) >10 to <50 jim in minimum dimension to
below 1 per 100 ml, (c) total bacteria below 10 per ml, (d) total viruses below 100 per ml,
and below levels listed for indicator microbes (see Table 2-1).
The following scores and interpretations were assigned:
A - Demonstrated to meet this standard in accordance with the approach suggested in the IMO
G8 guidelines (and G9 guidelines, if the BWMS employs an active substance).
B - Likely to meet this standard if the more detailed ETV Protocol (and corresponding sample
volumes) were to be used.
C - May have the potential to meet this standard with reasonable/feasible modifications to the
existing BWMS.
31
-------
D - Unlikely, or not possible, to meet this standard, even with reasonable/feasible modifications
to the exi sting BWMS.
To date, all BWMS adjudged to have reliable data have been tested in accordance with the G8
guidelines, which provide only general recommendations for how to evaluate performance with respect
to the D-2 standards. In late 2010, EPA's Environmental Technology Verification (ETV) Program
released the Protocol for the Verification of Ballast Water Treatment Technologies (Version 5.1, EPA
2010). Although no BWMS has yet been tested under the ETV Protocol, this protocol provides much
more detailed instructions for how to conduct BWMS tests that are scientifically rigorous and
statistically sound. In particular, the ETV Protocol has significantly improved sampling procedures. The
EVIO G8 guidelines suggest collecting replicate samples with volumes of at least 1 m3 for the size class
of organisms > 50 jim in minimum dimension (nominally zooplankton). ETV, and others have
demonstrated that a time-integrated sampling approach with larger sample volumes will increase
statistical confidence regarding whether zooplankton in sparse populations meet or exceed the EVIO D-
2/Phase 1 standard (Miller et al. 2011; Lee et al. 2010; section 3, above). As such, although D-2 and
Phase 1 standards are essentially the same, some BWMS were given a score of 'A' if the data showed
they met the D-2 standard by following the G8 guidance, and received a 'B' for Phase 1 if the number of
living organisms was consistently low and it seemed very likely the BWMS would still meet the
standard if ETV Protocols (including larger, integrated samples) were used.
Regarding the discharge standard lOx more stringent than the EVIO D-2/ Phase 1, the criterion used was
whether the number of living organisms in all size classes was consistently low following testing (below
the detection limit, often reported as zero, or not more than twice the standard). If so, the BWMS was
given a 'C', indicating it had the potential to meet the standard. However, as described in the response to
charge question 4 (Section 6), current testing methods do not provide the resolution required to conclude
that lOx standards can be met.
For the most stringent standards, lOOx and lOOOx more stringent than EVIO D-2/ Phase 1, if any living
organisms in any size class were found following treatment, the BWMS earned a 'D'. This score
indicates that it is extremely unlikely (or perhaps impossible) that the BWMS could meet a stricter
standard, again because the detection limit of the test methods used provide resolution to EVIO D-2/
Phase 1, at best. For example, if one viable zooplankter was found in testing using volumes of 1 m3, the
BWMS would be required to reduce the number of viable zooplankters to less than one in 10m3 or 100
m3 to meet the lOOx and lOOOx standards, respectively.
After each subgroup member completed his or her individual, independent assessments, they discussed
their scores collectively. All scores from the three primary reviewers were found to be identical and in
complete agreement with general assessments by the two subgroup oversight members, as well as other
members of the entire Panel. These consensus findings were used to create Table 4-1. Rather than
present the scores from individual, commercial BWMS units or models, the Panel categorized the
technologies by operation type (e.g., filtration + UV). The operation types were chosen from recently
published, third-party data reports (Albert et al. 2010; CSLC 2010; Lloyd's Register 2010) in order to
encompass all currently available operation types and to provide a standardized terminology. Thus,
while the data packages from individual BWMS were initially examined and scored, the results were
aggregated to represent a top-order status of the field. For a given operation type, if reliable data were
available for more than one commercial BWMS, the scores given to the operation type were the highest
32
-------
scores of any of the individual BWMS. In this manner, Table 4-1 represents the greatest potential for
each of the operational categories of technologies to meet various discharge standards.
33
-------
Table 4-1. Performance of Ballast Water Management Systems
Type or Category of BWMS
Deoxygenation
Deoxygenation+cavitation
Deoxygenation+bioactive agent
Electrochlorination
Electric pulse
Filtration
Filtration+chlorine
Filtration+chlorine dioxide
Filtration+coagulation
Filtration+UV
Filtration+UV+Ti O2
Filtration+ultrasound
Filtration+ozone+ultrasound
Filtration+UV+ozone
Filtration+electrochlorination
Filtration+UV+ozone+
electrochlorination
Filtration+electrochlorination+
advanced oxidation
Filtration+cavitation+
electrochlorination
Filtration+-electrochlorination+
ultrasound
Filtration+cavitation+ozone+
electrochlorination
#BWMS
2
1
1
2
1
1
2
1
1
10
1
1
1
1
5
1
1
1
1
1
# Type Approval Cert
0
1
0
1
0
0
0
0
1
3
1
0
0
0
1
0
0
0
0
1
# Available/Reliable Data
0
1
0
0
0
0
0
1
0
3
1
0
0
0
2
0
0
0
0
0
D-2
A
A
A
A
A
P-1
B
B
B
B
B
10x
C
c
C
c
c
100x
D
D
D
D
D
1000x
D
D
D
D
D
34
-------
Type or Category of BWMS
Filtration+plasma+UV
Filtration+cavitation+nitrogen+
electrochlorination
Filtration+hydrocyclone+
electrochlorination
Heat
Hydrocyclone+filtration+
peracetic acid **
Hydrocyclone+
electrochlorination
Hydrodynamic
shear+cavitation+ ozone
Hydrocyclone+filtration+UV
Menadione
Mexel
Ozone
Ozone+cavitation
Shear+cavitation+ozone
Shear+cavitation+peracetic acid
Totals
#BWMS
1
1
1
1
1
2
1
1
1
1
1
1
1
1
51
# Type Approval Cert
0
1
0
0
1
0
0
0
0
0
1
0
0
0
12
# Available/Reliable Data
0
0
0
0
1
0
0
0
0
0
0
0
0
0
9
D-2
P-1
10x
100x
1000x
Green rows designate the types of BWMS that had reliable data and whose performance was evaluated against various discharge standards.
Based on one or more reliable data sets, the type of BWMS:
(A) demonstrated to meet this standard in accordance with G8/G9
(B) likely to meet this standard if more detailed ETV Protocols were used
(C) potential to meet this standard with reasonable/feasible modifications
(D) unlikely, or not possible, to meet this standard
** Not scored because the one manufacturer has withdrawn this BWMS from market
35
-------
4.5. Assessment Results
The results of this assessment are presented in Table 4-1 and interpretations of the findings are provided
below. For this assessment, 51 individual BWMS were identified from prior reports (Albert et al. 2010;
CSLC 2010; Lloyds Register 2010) to show the breadth and diversity of treatment approaches.
However, it is important to note that of the 51 BWMS listed, a large proportion are at early
conceptual/development stages (only approximately 15 to 20 have been tested onboard an active vessel)
and a few have recently been discounted because of logistic or performance challenges. The Panel
received information packages on 15 individual BWMS, but just nine BWMS were considered to have
reliable data for an assessment of performance.
4.6. Response to Charge Question 1
The analysis described above formed the basis of the Panel's responses to charge Question la, 1 b, and 1
c; each of these sub-questions addresses different aspects of treatment capabilities for shipboard
systems. These questions and our responses are summarized below.
Question la: For the shipboard systems with available test data, which types or categories have been
evaluated with sufficient rigor to permit a credible assessment of performance capabilities in terms of
effluent concentrations achieved (living organisms/unit of ballast water discharged or other metric) ?
Conclusion la: Five types or categories of BWMS have been evaluated with sufficient rigor to permit a
credible assessment of performance capabilities. These technology combinations are:
• Deoxygenation + cavitation
• Filtration + chlorine dioxide
• Filtration + UV
• Filtration + UV + Ti O2
• Filtration + electrochlorination
Questionlb: For those types or categories of systems identified in la, what are the discharge standards
that the available data credibly demonstrate can be reliably achieved? Furthermore, do data indicate
that certain systems (as tested) will not be able to reliably reach any or all of the discharge standards?
The five types of BWMS listed above have been demonstrated to meet the IMO D-2 discharge standard,
when tested under the IMO G8 guidelines, and will likely meet USCG Phase 1 standards, if tested under
the more detailed ETV Protocol. This level of treatment efficacy results in a 10,000x reduction in
numbers of living organisms > 50 jim in minimum dimension (under land-based testing guidelines).
This represents an important achievement in the ability of these systems to effectively, reliably, and
dramatically remove live organisms from ballast water under the challenging conditions found on active
vessels.
The detection limits for currently available test methods and approaches prevent a complete statistical
assessment of whether BWMS can exceed the IMO D-2/ Phase 1 standard. However, one way to predict
the ability of a BWMS to meet lOx, lOOx, or lOOOx standards, is to consider the frequency with which
any live organisms are detected during testing. This approach provides insight into BWMS
36
-------
consistency/reliability and its lower performance limits. Three frequency categories were defined using
available data:
1. BWMS always produced "zero" or "non-detectable": The system is consistently exceeding
current detection limits and thus the IMO D-2/ Phase 1 standards (as described above).
However, if results for all test trials, for all categories of organisms, and for all samples from
a specific BWMS reported "zero" or "non-detectable," there is no way to determine if the
system is performing just below the IMO D-2/ Phase 1 standards or if it is approaching lOx,
lOOx or lOOOx.
2. BWMS produced "zero" or "non-detectable" most of the time, with only one or a very few
readings above the detection limit: The system appears to be operating near but below the
IMO D-2/ Phase 1 standards. It is also possible that the occasional or rare "detects" were a
result of BWMS malfunction or an error in sample collection, handling or analysis.
3. BWMS produced results in the detection limit most of the time, with only one or very few
"zero" or "non-detectable". The efficacy level of the system is clearly only at, or just below,
the IMO D-2/ Phase 1 standards.
Not one of the BWMS examined could be categorized in group 1 (i.e., consistently scored "zero" or
"non-detectable"). Instead, BWMS were roughly split between frequency categories 2 and 3. For all
BWMS, live organisms in the > 50 um and/or > 10 to < 50 um size classes were detected in at least two
independent test trials and in general, live organisms > 10 um were detected in 20% to 80% of test trials.
It is also important to note that when total bacteria was quantified during the testing of BWMS,
treatment did not reduce levels to that required in the lOOOx discharge standard (10/ml). In fact, it was
not uncommon to find an increase in total bacteria after treatment. Therefore, even if testing detection
limits are improved, the lower performance limits of current BWMS are not expected to change.
Conclusion Ib: The Panel concludes that five types of BWMS are currently able to reach IMO D-2/
Phase 1 standards. These same five types may be able to reach lOx EVIO D-2/ Phase 1 standards for the
>50 um and > 10 to < 50 um size classes in the near future, if both treatment performance and testing
approaches improve (see Section 5). Finally, no current BWMS types can meet a lOOx or lOOOx
discharge standard.
Question Ic: For those systems identified above, if any of the system tests detected "no living
organisms " in any or all of their replicates, is it reasonable to assume the systems are able to reliably
meet or closely approach a "no living organism " standard or other standards identified in Table 4.1 of
the EPA White Paper (June 2010), based on their engineering design and treatment processes?
To address this question, the phrase "no living organisms" was considered in two distinct ways: first, in
a literal sense to mean the sterilization of ballast water and second, from a scientific perspective to mean
results below method detection limits.
Based on the test data provided for several BWMS, it is clear that numbers of live organisms in
discharged ballast water are reduced dramatically relative to intake water and corresponding control
water. Five distinct BWMS types have been demonstrated to meet the EVIO D-2 standard and appear
very likely to meet the USCG Phase 1 standard (which demands, at minimum, a 4-log reduction from
37
-------
initial concentrations for the largest organism size class), not only in land-based testing but also under
the physically challenging conditions presented on active merchant vessels during shipboard testing.
However, even high levels of organism removal do not achieve sterilization or the complete removal of
all living organisms. The identification of just one live organism would indicate non-sterile conditions,
and all systems evaluated had at least one living organism in at least one treatment sample (and often
more, as described above). Unfortunately, in some cases, this low number of live organisms might result
from contaminated scientific sampling gear (nets, glassware, etc.) or human counting error.
Alternatively, it is possible to establish specific detection limits (e.g., 100, 10, 1.0, 0.1, live organisms
per m3 or ml) associated with the methods used to collect the current performance data available and
thus to conclude that, if numbers of live organisms are below those detection limits, they are statistically
indistinguishable from zero or no living organisms. Efforts have been made to calculate the probabilities
of meeting such specified detection limits under certain assumptions, such as whether the organisms are
randomly dispersed in space or spatially aggregated (see Lee et al. 2010 and Section 3 for details and
examples). Not surprisingly, increased statistical power comes not only from increased sample size, but
also from the difference between the mean established by regulation and the measured mean from a
sample—which indicates the degree of compliance (or noncompliance). (See Section 3 for a more
detailed discussion of sampling statistics and detection limits.)
Conclusion Ic: It is not reasonable to assume that BWMS are able to reliably meet or closely approach
a "no living organism" standard. Available data demonstrate that current ballast water management
systems do not achieve sterilization or the complete removal of all living organisms.
4.7. Response to Charge Question 2
Question 2: Based on engineering design and treatment processes used, and shipboard
conditions/constraints, what types of ballast water treatment systems can reasonably be expected to
reliably achieve any of the standards, and by what dates? Based on engineering design and treatment
processes used, are there types or categories of systems that conceptually would have difficulty meeting
any or all of the discharge standards?
A variety of BWMS types are being used to manage ballast water (Table 4-1). The data indicate that
several types or categories are proving reliable and effective, and Table 4-1 lists five types that have
been demonstrated to meet the EVIO D-2/ Phase 1 standard. These five BWMS also appear to be mature
technologies, with multiple active vessel installations, and are commercially available. Interestingly,
four of the five treatment approaches include a filtration step, although the inclusion of filtration does
not necessarily ensure that the BWMS will meet discharge standards. A large majority of BWMS also
appear to be adapted from technologies long applied to water treatment.
Given the data available, it is reasonable to assume that these same five systems have the potential to
meet a lOx EVIO D-2/ Phase 1 standard in the near future (see Section 5). As noted above, the Panel
makes this prediction based upon available data that show viable organisms sampled as low (usually,
below detection limits) but improvements to test methods/approaches will be required to demonstrate
conclusively that improved BWMS meet standards beyond EVIO D-2/ Phase 1. Given the data available,
it is highly unlikely that any of the systems listed in Table 4-1 could provide organism removal to the
level of lOOx or lOOOx the standard because all systems showed at least one observation of a living
organism within the sample volumes as specified in EVIO D-2 guidelines, thus clearly exceeding these
38
-------
more stringent standards. No BWMS reported zero living organisms in all samples analyzed following
treatment. In fact, most results showed an increase in total bacteria abundances after treatment, far
exceeding discharge levels proposed in the USCG Phase 2 standards. Ultimately, different technologies
or treatment approaches, and sampling strategies will be needed to achieve these higher levels of
removal. At this time, it is not possible to comment on the likelihood that the other treatment types listed
will, or will not, be able to meet either the EVIO D-2/Phase 1 or more stringent standards. All the BWMS
types listed in Table 4-1 have likely shown some potential for reducing the number of ballast water
organisms, but for most the data available for examination were deemed either to be absent or
unreliable. As such, it is not possible to predict the eventual performance of these BWMS.
Conclusion 2: Five types or categories of ballast water management systems can currently meet the
EVIO D-2 discharge standard and appear to meet USCG Phase I standard: Deoxygenation + cavitation;
Filtration + chlorine dioxide; Filtration + UV; Filtration + UV + TiO2; and Filtration +
electrochlorination. It is possible that the same five types could meet lOx EVIO D-2/Phase 1 sometime in
the near future if both treatment performance and testing methods and approaches (e.g., detection limits)
improve. Nearly all of the 51 treatment types or categories evaluated are based on reasonable
engineering designs and treatment processes, and most are adapted from longstanding industrial water
treatment approaches. However, the lack of detailed information on the great majority of BWMS
prevents an assessment of limitations in meeting any or all discharge standards.
4.8. Environmental Effects and Vessel Applications: Additional Constraints and Considerations
that Influence BWMS Performance
BWMS are still evolving with an ever-growing number of manufacturers developing systems. Although
several BWMS have received type approval certification, and appear to safely (e.g., received final G9
approval) and effectively meet EVIO D-2/Phase 1 discharge standards (Table 4-1), there are several
factors to consider beyond mechanical and biological efficacy. Perhaps the four most important
considerations for the broad applicability of BWMS are ambient water salinity (the ability to treat fresh,
brackish, and marine water) and temperature (the ability to work effectively and safely in a variety of
temperatures from warm equatorial to cold polar water), ship ballasting rate (the ability to treat water
moving at a variety of flow rates from < 200 m3/hr to > 4,000 m3/hr), and ballast volumes (the ability to
treat total volumes of ballast water from < 1,000 m3 to > 50,000 m3).
Another important vessel consideration is impacts of treatments on ballast tank and piping coatings and
substrate corrosion rates. Nearly all systems that alter the chemical composition or reactivity of ballast
water (e.g., heat, oxidants, and deoxygenation) can potentially affect corrosion of ship structures, piping,
fixtures and protective coatings. To a great extent, the potential effects of these BWMS have not been
consistently evaluated across the various modes of corrosion, including uniform or localized corrosion,
or for potential interactions with corrosion control systems including protective coatings and cathodic
protection systems. Some BWMS have provided data that indicate negligible impacts on corrosion rates
or even improvements. For example, deoxygenation, if operated properly, can dramatically reduce
uniform corrosion rates, but alternatively, deoxygenation may result in increased corrosion rates due to
either the cycling of hypoxic and aerated conditions or the formation of corrosion-causing sulfate
reducing bacteria if anoxic conditions are reached. Similarly, other BWMS utilizing strong oxidants
have been evaluated as having apparently negligible effects on coatings and steel corrosion rates.
However, it is also well documented in the water treatment and marine vessel industry that continuous
exposure to high doses of some oxidants, such as halogenated oxidants, can cause severe corrosion rates
39
-------
(depending on the specific oxidant, its concentration and contact period). On the other hand, while
heightened corrosion rates may be experienced shortly after treatment, corrosion rates on the whole may
not be significantly affected if the oxidant concentration declines rapidly.
Corrosion is already a significant concern for vessels operating in saltwater environments. As such,
coating failures and steel wastage are currently incorporated into periodic surveys and vessel service
periods. In the end, an increase in corrosion rates will impact the maintenance and repair costs borne by
the vessel owner; these potential increases in cost will need to be factored by the owner in selecting a
BWMS. In addition, corrosion control and mitigation strategies such as coatings and cathodic protection
should also be carefully considered since either or both of these may be employed to offset any
increased corrosion concerns. Although comprehensive assessments have not been conducted for all
BWMS, no major damage or casualties related to corrosion have been identified to date for BWMS
installed on ships.
In addition to specific environmental and vessel applications, vessel type and vessel operations can
dictate BWMS applicability. Although a multitude of vessel designs and operation scenarios exist, a few
important examples of specific constraints can greatly limit treatment options. Perhaps the most
dramatic limitations are found with the Great Lakes bulk carrier fleet that operates vessels solely within
the Great Lakes with large volumes of fresh, and often cold, ballast water ("Lakers"). The vessels in this
fleet have ballast volumes up to 50,000 m3, high pumping rates (up to 5,000 m3/hour), uncoated ballast
tanks (older vessels), and some vessels have separate sea chests and pumps for each ballast tank. A
further confounding issue is that voyages taken by Lakers average four to five days, with many less than
two days. Given these characteristics, a number of limitations are imposed: electrochlorination and
ozonation may only work in freshwater with the addition of brine (in particular Cl and Br, respectively);
oxidizing chemicals may increase the corrosion rate of uncoated tanks; deoxygenation and chemical
treatments that require holding times to effectively treat water (or for the breakdown of active
substances) may not be completely effective on short voyages; and the space and power needed for the
required numbers of filtration + UV treatments may simply not be available.
Another example of vessel-specific constraints is the sheer size of some vessels and the cargo they carry.
Very Large Crude Carriers (VLCC) and Ultra Large Crude Carriers (ULCC) can carry up to 100,000 m3
of ballast and can fill or discharge ballast water at over 5,000 m3/hour. While various BWMS may be
modular (perhaps providing the ability to add several units in a manifold design or in sequence), systems
that include a mechanical separation stage (e.g., filtration, hydrocyclone) or exposure to UV or
sonication may have difficulty addressing these large volumes and flow rates. Furthermore, given the
hazardous nature of the cargo carried on these ships (and other similar vessels, such as Liquefied Natural
Gas carriers), restrictions on the placement of a specific BWMS may apply, and system components will
likely have to satisfy classification society requirements for explosion-proof and intrinsically safe
construction, which might be more difficult for some BWMS types than others.
A final example is the treatment of ballast water on the tens of thousands of unmanned barges in the
U.S. that would fall under the ballast water discharge regulations. Inland waterways and coastal barges
are not self-propelled, but rather are moved by towing or pushing with tugboats. Because these vessels
have been designed to transport bulk cargo, or as working platforms, they commonly use ballast tanks or
fill cargo spaces with water for trim and stability, or to prevent excessive motions in heavy seas.
However, the application of BWMS on these vessels presents significant logistical challenges because
they typically do not have their own source of power or ballast pumps and are unmanned.
40
-------
Conclusion: While several BWMS appear to safely and effectively meet EVIO D-2/Phase 1 discharge
standards, there are several factors to consider beyond mechanical and biological efficacy. A variety of
environmental (e.g., temperature and salinity), operational (e.g., ballasting flow rates and holding times),
and vessel design (e.g., ballast volume and unmanned barges) parameters will impact the performance or
applicability of individual BWMS.
41
-------
5. SYSTEM DEVELOPMENT
5.1. Introduction
This section addresses issues related to potential future improvements in BWMS in response to charge
question 3, which focuses on three main issues: improving the performance of existing BWMS
technologies, identifying impediments to improved technological performance, and considering whether
technologies can achieve zero or near-zero discharge of organisms.
Before considering possible improvements, it is important first to consider what has been achieved to
date by existing BWMS technologies. The 2004 IMO D-2 standard has provided a stable target for
research, development, testing, and evaluation of practices and technologies to treat ballast water. Using
the proposed D-2 standard as a design goal, some developers of BWMS have:
• Integrated BWMS within marine vessel arrangements, weight and stability constraints,
electrical distribution and piping systems, and automated control systems.
• Integrated the operation of BWMS within the larger context of merchant vessel operations
such as ballasting rates and volumes, logistics requirements such as reliable chemical-supply
chains and service/support centers, safe operations such as hazardous-rated equipment and
chemical-handling procedures, and operational training.
• Tuned BWMS to achieve acceptable levels of disinfection by-products, residual toxicity,
within the limits of practicality and in compliance with the proposed D-2 standard.
• Packaged the technology for a competitive commercial market, which requires consideration
of life-cycle costs, reliability of equipment, maintenance issues, and acceptance of the
BWMS technologies by mariners.
These are important achievements. They also foreshadow the issues that must be considered when
evaluating technological options to improve BWMS technologies.
In general, technological changes to improve BWMS performance will proceed along one of two paths:
(1) incremental changes to existing designs with the goal of optimizing performance, or (2) designing
entirely new treatment methods. Incremental changes offer the faster path to improved performance, but
are likely to achieve only relatively modest improvements. Wholly new approaches for BWMS, possibly
drawn from the water treatment industry, also would improve performance—perhaps significantly—but
would take more time to develop and test in order to determine performance, practicality, and cost-
effectiveness. The Panel considers both pathways below.
5.2. Improving the Performance of Existing Systems
Charge question 3a. For those systems identified in questions La and 2, are there reasonable changes
or additions to treatment processes which can be made to the systems to improve performance?
The Panel defined "reasonable changes" as incremental adjustments or improvements that do not
fundamentally alter the treatment process. For example, design changes to increase UV radiation
intensity would be an incremental adjustment, whereas addition of a UV stage would not. In practice,
"reasonable changes" mean the same thing as "incremental improvements." Both are based on the
42
-------
concept of "turning up the dial" on existing technologies rather than creating wholly new systems or
adding processes to the treatment system. Although incremental changes can generally be implemented
more quickly, this approach is not necessarily simple or foolproof. First, it may be impossible or
impractical to further improve the baseline technology. Second, any changes could fundamentally alter
other aspects of the technology's performance or use; e.g., it could change its life cycle costs, affect
integration of the BWMS with vessel operation, or increase residual toxicity of ballast water discharges.
As described in Section 4.4, five BWMS have demonstrated compliance with EVIO D-2 standards, under
G8 testing conditions. Based on information from available test results, incremental improvements could
be made to these treatment processes, perhaps yielding performance greater than D-2. However, these
changes may add costs and engineering complexity.
Examples of incremental improvements to the BWMS judged to comply with the D-2 standard, as
shown in Table 4-1, are summarized below.
• Deoxygenation + cavitation. Current technology for these systems establishes severe
hypoxia, which kills larger organisms very effectively, although hypoxia has little or no
effect on some bacteria, pathogenic protozoa, or viruses. It is not possible to improve on
hypoxic conditions per se, however, it may be possible to reduce the time needed to reach
severe hypoxia and increase holding time under severe hypoxia. In addition, effectiveness
might be improved by increasing the degree of cavitation and physical/mechanical disruption
of organisms.
• Mechanical separation + oxidizing agent. These systems could be optimized in several
ways: improved mechanical separation (i.e., filtering) to remove higher percentages of
particles and particles of smaller size; increasing the concentration and contact time for
oxidizing agents; and adjusting other water chemistry parameters, such as pH, to increase the
efficacy of the oxidizing agent. (Note: this category corresponds to those BWMS using
filtration and chlorine dioxide, filtration plus UV + TiC>2, and filtration plus
el ectrochl orinati on).
• Mechanical separation + UV. These systems could be optimized by improved mechanical
separation (i.e., filtering) to remove higher percentages of particles and particles of smaller
size and by increasing UV contact time and dosage.
5.2.1. Combination Technologies
Most ballast water treatment systems, even those with a single primary component, are actually
combination technologies. For example, one company's BWMS relies primarily on deoxygenation, but
also has a venturi device that mechanically damages some of the organisms (as would cavitation) and
uses carbon dioxide, which forms carbonic acid, lowering the pH of the water. Another commercial
system is advertised as a combination technology that includes filtration, ultraviolet radiation, and
treatment by free radicals.
It is difficult to fully understand the interactions of treatment processes used in combination BWMS.
This makes it hard to predict the overall treatment effect of incremental improvements within individual
processes. For example, one company's system combines filtration, cavitation, ozone, and sodium
43
-------
hypochlorite. With four "primary" technologies at work, which should be the focus for "turning up the
dial" to improve performance? Further complicating matters is the great variability in the physical and
chemical properties of ballast water itself, which in turn creates complex interactions with individual
treatments as well as their combination.
To date, combination BWMS have been developed through research and testing. However, once a
technology has shown promise to meet the D-2 standard, its development is stopped to allow the device
to undergo certification testing. It is reasonable to assume that incremental improvements to
combination technologies could yield efficiencies in operation (less power, less cost, more reliability)
and moderate improvements in treatment effectiveness. Due to the complex interactions among
combined treatment processes, however, such possibilities can only be speculative until a more rational
understanding of the modes of inactivation/kill is developed and verified by experiments with
prototypes. Thus, the Panel's comments on likely improvements to BWMS technologies are restricted to
the primary treatment processes identified in Section 4 and described below: UV radiation, mechanical
separation + cavitation, deoxygenation, and oxidant based systems. Other BWMS processes, such as
ultrasound and electro-mechanical separation, are not as widely utilized and therefore were not
reviewed.
5.2.2. UV Radiation
There are several ways in which treatment by UV radiation may be improved. It may be possible to
deploy "over-sized" UV radiation treatment systems to improve performance. For example, a ballast
system that runs at 800 m3/hr could be paired with a treatment system rated for 1,000 m3/hr, thereby
presumably increasing UV exposure by 20 percent. Testing and analysis would be required to determine
if efficacy was actually increased and, if so, by how much, and to ensure there were no adverse impacts
to the residence time distribution or UV intensity field distribution of the UV chamber.
Similarly, UV system performance might be improved through increased intensity of UV lamps. The
length of time the ballast water is exposed to UV radiation could also be increased by increasing the size
of the chamber relative to the ballasting rate. Such improvements, however, would increase the size and
cost of BWMS equipment.
UV chambers could also be staged in series, though this would substantially increase cost, required
space, and maintenance. However, employing multiple UV chambers in series could provide the
following improvements to performance:
• Decreased chance that organisms could "slip" past untreated, assuming that each chamber
could independently provide treatment adequate to meet a given standard.
• Increased time during which organisms are exposed to UV.
The performance of UV systems also could be improved by using more effective mechanical separation
methods ("filtering") upstream of the UV chambers in order to enhance the transmissivity (clarity) of the
ballast water prior to UV treatment. Flocculants such as alum could further clarify the ballast water,
providing that these agents do not impart UV absorbance (as may be the case with iron-based
flocculants). Such improvements have drawbacks. For instance, advanced mechanical separation would
significantly increase costs and space requirements, and would likely significantly increase backpressure
44
-------
within the system, resulting in higher electrical power demands and a need for higher-head ballast
pumps.
5.2.3. Mechanical Separation and Cavitation
Many BWMS use mechanical separation as the primary precursor for other treatments such as UV
radiation or oxidants. The purpose of mechanical separation varies according to the treatment's
disinfection processes: e.g., screening to remove larger organisms resistant to disinfection, reduction of
organic matter to reduce oxidant demand, and reduction of turbidity to increase transmittance of UV
radiation.
Mechanical separation also has a secondary effect—physically damaging some of the organisms as they
pass through the device—which may inactivate or kill organisms or weaken their cellular structure such
that effective disinfection is more easily achieved. In this regard, mechanical separation is similar to
cavitation devices designed to impart physical damage to organisms. Some BWMS include cavitation
devices to damage cellular structure without having to handle separated filtrate.
Use of seawater filtration on vessels traditionally has been limited to protecting mechanical devices in
the piping system. For example, seawater might be "screened" to a one-eighth inch opening (3.175 mm)
to protect the narrow passages of a heat exchanger. Recently, however, several common and proprietary
devices have been developed for filtering and imparting cavitation effects on ballast water as part of the
treatment process: variations on back flushing of traditional screen filters; vibrating disc filters; multi
hydro-cyclone; and various cavitation devices. In general, the filter units target removal of particles
above 40 or 50 jim and have significant waste streams that are returned to the ambient water. Typically,
filtering takes place on ballast water uptake only.
Mechanical separation devices are advertised in terms of percentage removal. For example, two
companies claim filtration rates of approximately 90 percent removal of zooplankton. These removal
levels, although essential to support the disinfection process, by themselves are far from adequate to
meet the D-2 standard for the size class > 50 jim.
In summary, it is not reasonable to expect incremental improvements in mechanical separation devices
to achieve significant advances in performance over the D-2 standard. Such improvements would
require use of media filters, membrane filters, or other devices that have not yet been practically applied
to ballast water treatment. Similarly, cavitation devices alone cannot meet the D-2 standard. It is not
clear if improvements to cavitation devices will significantly increase the effectiveness of BWMS that
employ combined processes.
5.2.4. Deoxygenation
Two Type-Approved BWMS that have met the D-2 standard use deoxygenation as part of multiple
treatment processes. The first system lowers oxygen by pumping low-oxygen gas from a purpose-built
burner into the BWMS through a venturi device. The efficacy of this system relies on several
interrelated components: rapid application of the gas stream; creating carbonic acid from the carbon
dioxide in the gas stream, which lowers pH and makes the low-oxygen environment more lethal; and the
mechanical effect of the venturi on passing organisms.
45
-------
This system lowers the oxygen level to about 2 percent by volume (about 0.7 mg/L) by using a variation
on the traditional tank-ship combustion-based inert-gas generator. Traditional units produce a 5 percent
oxygen level (about 1.8 mg/L). For reference, 2 mg/L oxygen is considered the upper boundary for
environmental hypoxia and the point of mortality for sensitive species. Very few metazoans can survive
<1 mg/L oxygen for longer than 24 hours (Vaquer-Sonyer and Duarte 2008). Further improvements to
combustion-based units may not be practical given constraints of the combustion process.
The second system lowers oxygen levels through use of a nitrogen generator. The generator uses a
membrane to filter ambient air, resulting in high quality nitrogen gas. The system also uses mechanical
separation, cavitation and electrodialytic disinfection. Nitrogen generators are widely deployed in
industry and in some marine applications. On ships, they are generally regarded as expensive, high-
demand consumers of electrical power. It is possible to create very high quality nitrogen gas,
approaching 99.9 percent pure, but doing so requires significant space, capital costs, and high electrical
power demands.
As these examples show, BWMS that use deoxygenation often also use additional treatment processes.
Thus, it is difficult to predict the effect of incremental improvements on overall efficacy. In fact, some
changes might decrease efficacy, or worse, result in unanticipated adverse conditions; e.g., higher
populations of sulfate-reducing bacteria and subsequent increase in steel corrosion rates. Consequently,
it is not possible to assess whether incremental improvements will yield higher performance. For
example, with respect to treatment lethality to metazoans, there is little to no difference between oxygen
levels of 2 mg/L and 1 mg/L for the same contact period. Extending holding time would be more
efficient than additional efforts to reduce oxygen below 1 mg/L. However, there is some evidence
(Mario Tamburri, pers. comm.) that faster transitions to severe hypoxia are more lethal.
5.2.5. Oxidant-Based Systems
Oxidant-based systems introduce an oxidizing agent (such as chlorine) into the ballast water stream.
These systems are generally designed to target a given level of total residual oxidant (TRO) in the
treated ballast water. Oxidant-based systems pose issues for mechanical integrity and for worker safety,
since these systems require adding chemicals in bulk, on-site manufacture of sodium hypochlorite or
similar chemicals, or on-site production of ozone gas.
The consumption (or oxidant demand) of the introduced oxidant varies with the organic-matter content
of the ballast water uptake. After the initial instantaneous consumption of oxidant, any remaining
oxidant is pumped with the ballast water into the vessel's ballast tanks. There it is held for a prescribed
length of time at the TRO concentration. The TRO level will decay over time as a function of many
factors, including its initial concentration, salinity, temperature, motions of the vessel, and configuration
of the ballast tank and venting system. Depending upon the predicted or measured oxidant levels in the
ballast water, a neutralizing agent may be applied before or during its discharge to the environment.
The effectiveness of oxidant-based systems is a function of the concentration of the residual oxidants
and the holding time. Incremental changes that could improve effectiveness include: increasing initial
oxidant concentrations; maintaining a higher oxidant concentration during the holding period; and
increasing the holding period or contact time. The potential for improved performance for each of these
three options are considered below (the use of oxidants in combination BWMS treatments is considered
separately).
46
-------
Increasing Initial Oxidant Concentrations
Determining the initial oxidant concentration needed to reach the required efficacy is part of the "art" of
a BWMS. For example, TRO values for BWMS that the Panel reviewed varied in the type of oxidant
used (e.g., ozone, chlorite ion, free active chlorine) and TRO amounts varied by an order of magnitude.
Several oxidant-based systems also use some form of mechanical separation to remove larger organisms,
organisms entrapped within protective solids, and some paniculate organic matter and thereby reduce
oxidant demand. Regardless of the effectiveness of the mechanical separation, however, it is the residual
oxidants and other reactive disinfection byproducts that achieve the final treatment performance. Thus,
even though mechanical separation may reduce the amount of chemical required, it is unlikely to
improve the efficacy of the oxidant. Tertiary effects also occur, such as damage to organisms'
membranes during mechanical separation and subsequent membrane interaction with oxidant-based
systems. However, these tertiary effects are difficult to assess. Thus, they do not represent an obvious
method for incremental improvement.
However, it is possible to "turn up the dial" on existing BWMS by increasing the amount of oxidant
used, which should improve effectiveness. Doing so simply requires that a higher capacity ballast water
treatment system be installed. For example, concentrations could be increased 50 percent by installing a
system rated for 1200 m3/hr on a vessel that pumps ballast water at 800 m3/hr. Such an installation will
demand larger space and weight allowances, more power, and higher capital and operating costs. In
general, it should be possible to integrate higher capacity systems for new vessel designs, but it is more
of a challenge to retrofit on existing vessels.
Higher oxidant levels in ballast water can have a significant negative effect impact on piping-system
components and tank-coating systems. Valve packing, flange gaskets, and pump seals are made of a
variety of materials, some of which are not compatible with oxidants at low concentrations, and less so
at increasingly higher concentrations. Impacts on tank coatings are not yet well understood. TRO levels
up to 10 mg/L may be compatible with typical intact ballast-tank marine coatings. However, coatings
are frequently not intact, because they wear over time. In the case of freshwater shipping, a ballast water
tank may not be coated at all. Corrosion of exposed carbon-steel structures can lead to structural failures
and require expensive and complex repairs. Use of increased oxidant levels, therefore, would likely
increase rates for coating failures and corrosion of exposed carbon-steel structure.
Use of higher oxidant levels also increases concerns for safe handling on board vessels, due to the need
for handling and storage of additional bulk chemicals, lengthening the time required to make confined
tank spaces safe for entry for inspection and repair work, and generation of hydrogen gas. These
concerns can be handled through operating procedures but at the expense of increased time and effort.
As higher levels of oxidants are introduced into ballast water, complex chemical reactions take place,
resulting in potentially harmful disinfection byproducts through interactions between the oxidant level
and characteristics of the uptake water (such as its organic load, alkalinity, salinity, and chemical
contaminants). Further tests and analysis would be required to determine whether these byproducts
should and can be neutralized so that the ballast water discharge will meet acceptable toxicity limits.
47
-------
Maintaining Increased Oxidant Concentrations
Most oxidant-based systems rely on achieving residual oxidant levels that are adequate to meet D-2
standards and then maintaining that concentration for the duration of the holding period. The hold time
of ballast water can vary significantly, however, and shipping schedules, weather, equipment failure, and
cargo-handling changes frequently result in hold times that are longer or shorter than initially expected.
As hold times increase, TRO concentrations decay, which also reduces detoxification costs. Most
evaluation testing occurs during a prescribed holding period, typically for two to five days. In reality,
ballast water hold times routinely vary from one day to several weeks. In fact, some ballast tanks can
remain full, or partially full, for many months or even years.
There has been little development or testing of systems that monitor and maintain a specific oxidant
level in ballast water tanks. Indeed, automated monitoring of oxidant levels in ballast water tanks is not
currently practiced. Continuous or periodic monitoring would require either a network of sensors
installed in the tanks or a means of drawing a liquid sample on a periodic basis to a remote monitoring
device. Such sensors are common practice in the treatment of drinking water and wastewater. Either
approach requires significant cabling, possibly tubing and pumps, monitoring equipment, and data-
recording devices.
Current practice to maintain oxidant levels, if done at all, is to "top up" a ballast tank; i.e., to partially
discharge its contents, then refill the tank with freshly treated water. The objective is to achieve the
desired oxidant level by mixing the "new" water having a high concentration of oxidant with the water
remaining in the tank. Under such conditions, it is imprecise to determine whether desired inactivation
levels are achieved. Such efforts are similar in mechanical function to ballast water exchange, would
likely be performed while the vessel is at sea, and carry with them the same significant safety concerns
regarding vessel stability. A safer and more reliable approach for topping up oxidant levels would
require developing new systems. Such systems might include chemical dosing lines to deliver an
external supply to each ballast tank, combined with circulation devices internal to each ballast tank.
Increasing the Hold Period
Increasing the hold time of the ballast water while maintaining a certain oxidant level would likely
increase treatment efficacy. However, it is ship operations that will dictate the duration of this hold time
for most ballast water tanks. In particular, the largest mid-body ballast water tanks almost always have
to be discharged while tank ships or bulk carriers are being loaded. As such, the treatment process must
account for the expected hold period, but likely will not have the ability to alter it.
Summary of Oxidant-Based Treatments
Existing oxidant-based systems have been developed to meet the D-2 standard, and several have
received international approvals. Their efficacy could be improved by increasing initial residual oxidant
levels in ballast water during uptake. However, testing would be needed to determine the degree of
improvement and to determine toxicity effects from disinfection byproducts resulting from higher
oxidant doses.
Increasing residual oxidant levels also creates demands for space, weight, power, and capital and
operating expenses; in addition, these systems will increase piping-system compatibility issues, ballast-
48
-------
tank corrosion rates, and safe-handling concerns. Alternatively, it may be possible to increase
effectiveness by maintaining residual oxidant levels during holding time in the ballast water tanks.
Current systems, however, have only rudimentary methods for performing such operations. New
methods will need to be developed and tested to determine their practicality and effect.
5.3. Principal Technological Constraints
Charge question 3b. What are the principal technological constraints or other impediments to the
development of ballast water treatment technologies for use onboard vessels to reliably meet any or all
of the discharge standards?
Existing BWMS have been developed to the EVIO D-2 standard within the context of typical marine
vessel constraints, including restrictions on size, weight and energy demands. While practical for new
construction vessels, existing vessels may not be able to integrate such BWMS on a retrofit basis.
Meeting higher standards generally implies that the treatment processing plants will need to be large,
heavy, energy-intensive and expensive. At some point, constraints associated with the installation and
operation of such equipment may require a fundamental shift in how ballast water is managed.
Regardless of the applicable treatment standard, existing and potential BWMS share common
impediments to development.
(a) Technical constraints include:
• The shipboard marine environment is corrosive and subject to vibrations and ship motions.
Thus, one should not assume that shore-side systems can be transferred easily to shipboard
use; in fact, a strong shipboard service history will be an important guide to selecting system
components. Even so, the characteristics of water in some non-ballast water shipboard
applications may differ from ballast water (e.g., sediment concentrations may be greater in
ballast water). This makes it difficult to predict performance based on service history alone.
• Vessels are initially designed to have ballasting capabilities and procedures that meet their
intended service and voyage profile. BWMS intended for retrofit will need to fit within those
original parameters.
(b) Other impediments include:
• Lack of clear design goals. There is disagreement on discharge standards; they vary from
state to state within the U.S. and internationally. Thus, BWMS manufacturers have multiple
discharge standards as design targets. Further, there are no established compliance
monitoring, enforcement procedures. Such procedures would help focus future BWMS
development; e.g., they would encourage the creation and use of additional performance
metrics, such as system reliability, as contrasted with the current focus on discharge quality
as evaluated during certification testing.
• Limited experience and limited empirical data on life-cycle costs. The full cost of any
BWMS includes not only its initial purchase and installation costs, but also its long-term
operational costs. System reliability, durability, cost of spares, and ease of maintenance are
factors that contribute to determining the BWMS life-cycle costs.
49
-------
(c) Constraints to improved performance include:
• Shipboard BWMS are developing rapidly. The focus to date has been engineering the
treatment device for discharge performance. This focus has come at the expense of ensuring
integration of the BWMS with vessel mechanical systems and marine operational activities;
BWMS durability, maintenance and repair; training; and procedures for monitoring
technology performance.
• Ships crews are small in number and busy; therefore, any new system must be easy to
operate and maintain. Ideally, new systems would enable remote control from the ballast
control console and automatic operation in or near port, which is typically a busy time for
crew.
• Most importantly, BWMS should pose no unreasonable health risk for the crew nor create
higher risk for vessel safety, and require no exceptions to the safety procedures established
by the vessel owner. The BWMS installation and operation procedures must also meet the
requirements of control authorities, i.e., Classification Society, Flag State, and Port State.
• Facilities properly equipped to test BWMS technologies are few, which imposes a bottleneck
to swift verification and testing and thus hinders development. Increased sharing of data and
specific protocols among such facilities is essential.
5.3.1. Operational Challenges on Working Merchant Vessels
It is unlikely that current BWMS will be able to meet the most stringent proposed standards (e.g., lOOx
D-2 or lOOOx D-2). This is perhaps best understood in the context of required reductions in organisms.
Meeting the D-2 standards for zooplankton-sized organisms requires that the BWMS reduce the number
of zooplankton in challenge water (as defined by the EPA ETV program) by four orders of magnitude.
For a very large crude carrier (VLCC) tanker carrying roughly 90,000 m3 of ballast water, the D-2
standard would require reducing the number of zooplankton-sized organisms from 9 billion to 900,000
(Table 5-1). The USCG's proposed Phase 2 standard for zooplankton (in the column labeled "D-2/1000"
in Table 5-1) would require that BWMS reduce viable zooplankton by seven orders of magnitude
relative to values in ETV challenge water. (This is a 99.99999 percent reduction, referred to in reliability
engineering as "seven-nines"). For the VLCC example, the proposed Phase 2 standard would limit the
discharge of viable zooplankton to a maximum of 900 individuals (Table 5-1). To put this value in
perspective, it is fewer than half the number of zooplankton (2000 individuals) contained in a 20-liter
bucket of ETV challenge water (Table 5-1). Additional examples of allowable zooplankton discharges
associated with different discharge standards and for different types of vessels are summarized in Table
5-1, below.
50
-------
Table 5-1. Zooplankton Counts for Water and Increasing Log Reductions from D-2 Standard. The USCG's
proposed Phase 2 standard is represented by in the column labeled "D-2/1000".
Volume Basis
Test Standards
VLCC Tanker
Great Lakes Bulk Carrier
Handymax Bulk Carrier
Panamax Container
Feedermax Container
Passenger Ship
ETV Testing Tank
VLCC Pipe (2.2 meters)
Bucket (20 liters)
Glass (0.4 liters)
Volume
(m3)
l.OOE+00
9.00E+04
4.40E+04
1.80E+04
1.70E+04
3.50E+03
3.00E+03
2.00E+02
1.39E+00
2.00E-02
4.00E-04
Rate
(m3/hr)
NA
5.00E+03
l.OOE+04
1.30E+03
5.00E+02
4.00E+02
2.50E+02
2.00E+02
5.00E+03
NA
NA
Viable
Seawater
l.OOE+05
9.00E+09
4.40E+09
1.80E+09
1.70E+09
3.50E+08
3.00E+08
2.00E+07
1.39E+05
2.00E+03
4.00E+01
> Organisms
IMO D-2
l.OOE+01
9.00E+05
4.40E+05
1.80E+05
1.70E+05
3.50E+04
3.00E+04
2.00E+03
1.39E+01
2.00E-01
4.00E-03
>50 urn (Sea
D-2 x 10
l.OOE+00
9.00E+04
4.40E+04
1.80E+04
1.70E+04
3.50E+03
3.00E+03
2.00E+02
1.39E+00
2.00E-02
4.00E-04
water per Ul
D-2 x 100
l.OOE-01
9.00E+03
4.40E+03
1.80E+03
1.70E+03
3.50E+02
3.00E+02
2.00E+01
1.39E-01
2.00E-03
4.00E-05
>ETV)
D-2xlOOO
l.OOE-02
9.00E+02
4.40E+02
1.80E+02
1.70E+02
3.50E+01
3.00E+01
2.00E+00
1.39E-02
2.00E-04
4.00E-06
Table 5-1 expresses zooplankton treatment standards as maximum allowable numbers of viable organisms for various
volumes. The top row ("Test Standards") provides organism counts in 1 m3 of ETV challenge water (column labeled
"Seawater"), and maximum allowable counts in 1 m3 of water meeting the IMO D-2 standard and successive log reductions
beyond D-2. Several vessel types are listed showing their typical ballast-water volumes and discharge flow rates. For each
volume, the table shows the number of organisms it contains (column labeled "Seawater") and the maximum number of
organisms allowed by each of the discharge standards.
Table 5-1 also indicates the number of zooplankton in volumes of ETV challenge water equivalent to a beer glass, a bucket,
and that displaced by one second of untreated discharge from a VLCC. The colored highlights indicate when the glass,
bucket, or discharge contains more viable organisms than those in the total volume of water discharged from a vessel in
compliance with the various standards.
In contrast, incremental adjustments to existing technologies are expected to result in only slightly
greater reductions of viable organisms in BWMS discharge. In part, the inability to achieve huge
reductions stems from the design characteristics of present-day BWMS technology, which is placed "on
top of existing ballast-piping systems. The treatment devices (e.g., filters, UV lamps, cavitation
devices) are added to the standard ballast-piping, which was originally designed solely for the efficient
uptake and discharge of ballast water. Further, ballast water is still taken up, held, and discharged in
essentially the same manner as in the past.
It is also instructive to consider the challenges of meeting the proposed, more stringent standards within
the context of a working merchant vessel. Table 5-1 shows that VLCC tankers discharge ballast water at
a rate of 5,000 m3/hr. At this rate, one second of discharge yields 1.39 cubic meters of water. Assuming
ETV challenge water conditions, this one second of discharge would contain 139,000 zooplankton - a
number that would exceed the allowable discharge of organisms for the entire VLCC ballast water
capacity by the following amounts for the proposed more stringent standards: 1.5 times greater than the
lOx D-2, 15 times greater than the lOOx D-2, or 154 times greater than the lOOOx D-2.
Meeting these more stringent standards will require the following technical challenges be overcome:
51
-------
Controls to Avoid Discharge of Untreated Water
At a minimum, ballast water piping systems must be carefully designed to avoid discharge of any
untreated ballast water, however minimal in volume. Doing so likely would require separate uptake and
discharge ballast water piping. Current standard practice is to use a common piping system for both
uptake and discharge. In addition, guarding against discharges during brief interruptions in treatment
during start-up or shut-down may require that BWMS be designed to re-circulate treated ballast water to
confirm its treatment status before discharge.
Controls to Avoid Cross-Contamination
• Isolating the ballast-piping system. Many ships have a cross-over to fire mains, black and
grey water drains, bilge water lines, and cooling water circuits.
• Maintaining a high level of tank structure integrity. Especially in aging vessels, tank
structures can permit transfer of fluids from adjacent tanks, piping systems running through
tanks, fluids pooling on tank tops, and directly from ambient water through seams or pipe
fittings in the vessel's side shell.
• Protecting tank vents. Ballast tanks vents are typically fitted with only a rough screen or a
ball check device to minimize entry of seawater. Protecting these vents from ingress of
untreated seawater will become more critical if standards become more stringent.
In-tank Monitoring, Treatment and Mixing
Careful monitoring of in-tank conditions will be very important under the following conditions: when
ballast water hold times are very long, thus enabling surviving organisms to reproduce, or when they are
very short, thereby reducing time for treatment to take effect; when in-tank sediment loads provide a
protective layer for organisms, shielding them from the disinfection process; and when highly
heterogeneous ("patchy") uptake ballast water overwhelms the treatment process.
Overcoming these challenges requires developing means to:
• Monitor tank conditions, although doing so is difficult because ballast water tanks are
typically complex and are known to have hydrodynamic "dead zones" that are not flushed
out during a typical ballast cycle.
• Treat (or re-treat) a full ballast water tank, such as would be needed when the ballast water
uptake is ineffective, when it has been contaminated from external sources, or when the
expected hold time has been exceeded.
• Mix a full ballast water tank. An ideal mixing system would suspend sediment loads, permit
even treatment of the tank's entire contents, and permit representative monitoring of the tank.
52
-------
Improving the Efficacy of Mechanical Separation and Disinfection Technology
The performance of current "filter and disinfect" treatments is especially limited in circumstances when
the ballast water uptake is patchy or has a high sediment load.
5.3.2. Idealized Designs for BWMS
The water treatment industry is an obvious place to turn for developing new BWMS. This industry has
developed methods to disinfect large volumes of water to very high standards. New approaches adapted
from that arena may be very efficacious and able to achieve the proposed, more stringent standards, but
it would take time to develop, test, and determine their practicality and cost impacts. Nonetheless, in
thinking about an idealized design for ships, it is a useful thought exercise first to consider elements
from a shore-based treatment system.
To that end, the Panel developed a hypothetical design for an onshore ballast water treatment plant with
a design capacity of 20,000 m3 of ballast water per day. This is equivalent to -800 m3 per hour, roughly
similar to a "low ballast dependent" vessel such as a containership. ("High ballast dependent" vessels,
such as Great Lakes bulkers and large tank ships, would require a treatment plant 5 to 12 times larger.)
The design requirements for this idealized, hypothetical treatment plant were estimated as:
• Equalization tanks of volume 20,000 m3.
r\
• Plain sedimentation area of-1,000 m .
• Granular media filtration of-120 m2.
• Three UV units each at -800 m3 per hour.
• Sludge and backwash handling.
• Possibly to include a membrane-filtration unit.
Based on the long history of water treatment plants, the Panel thinks it likely that such an idealized
system could meet EVIO D-2, and indeed, lOOOx D-2 standards for all size fractions, including the IMO-
specified bacteria. Nonetheless, pilot scale testing would be needed to confirm optimum design
parameters. Such pilot testing programs are common practice in water treatment plant design.
Using the Idealized Design as a Basis for Conceptualizing New Shipboard BWMS
The previous section describes opportunities for incremental improvements in BWMS. Here the Panel
illustrates the second pathway for improved BWMS - that is, the design of wholly new systems - for the
purpose of meeting the proposed, more stringent standards. The example concept discussed below draws
upon the operational particulars of the idealized system just described within the context of the technical
and operational constraints for shipboard BWMS.
A wholly new treatment design would significantly increase the operational burden on ship operators,
but it is technically feasible to integrate wholly new treatment systems into new vessel designs.
Integrating such an idealized system into existing vessels would be technically challenging on most, and
not possible on many, existing vessels. Finally, such a conceptual system and processes would need
better definition and specification in order to develop cost-benefit analyses; neither capital nor operating
costs have been estimated.
53
-------
In order to better convey the distinction between incremental improvements to BWMS and new designs
for BWMS, an illustrative, hypothetical concept sketch of a new design for shipboard BWMS is
presented in Figure 5-1. This concept sketch is based on a Panamax container ship having a ballast
volume of 17,000 m3 and a discharge rate of 500 m3/hr. This sketch is illustrative only. It is presented
solely to assist in the evaluation of how more stringent treatment standards might impact vessel
arrangements, operations and costs. For example, this system would likely require at least three to four
times the number of components, space, expense, and effort compared to existing BWMS.
By way of overview, this conceptual system would be capable of achieving higher filtration levels,
provide greater control of oxidant levels in tanks, and enable a final disinfection using UV radiation. The
treatment process would be integrated through use of large media filters integral to the vessel hull for
ballast water uptake and discharge and through recirculation of the ballast water in the ballast water
tanks in order to dose, monitor, and maintain oxidant levels in the ballast water tanks. For ballast water
discharge, a residence tank would be considered adequate to ensure neutralization of the oxidant. A final
UV disinfection step would be handled using a dedicated ballast water discharge connection. Details for
each of these steps for this idealized system are described below.
LOW
'SEA CHEST
HIGH
SEA CHEST
Figure 5-1. Concept Sketch of a New Approach to Shipboard Ballast Water Treatment (TYP means
"typical").
Ballast Water Uptake
Two traditional, but oversized, sea chests (intake structures for ballast water in ships' hulls) would serve
to take up ballast water. Piping would generally be 300 mm nominal. Each sea chest would include
standard skin-valve isolation and piping materials. The sea chests would be located port and starboard,
one high and one low, with a cross-over suction main connecting each. This would provide flexibility
for avoiding sediment when the ship is close to the bottom, and algal blooms when the ship is light and
the high sea chest is close to the surface. The Panel recognizes that sea chests can provide refuge for
nonindigenous species, but methods for keeping them and adjacent hull areas free of fouling organisms
were not considered, as such considerations were beyond the charge.
54
-------
The cross-over suction main would discharge by gravity into two large media chambers plumbed in
parallel and each sized for full flow. This arrangement would allow one to be by-passed during back-
flush cycles. Each would be built into a one-meter-height double bottom in the ship's hull and 8 m2 for a
volume of 64 m3 each. Industrial wastewater industry media with tolerance for velocities approaching 60
m/hr, and a useful life of six years between dry dock periods would be considered. Six-year servicing of
media would be through manhole covers.
Ballast water leaving the media filter would be disinfected prior to entering the ballast water tanks,
either by a UV or an oxidant chemical. This transfer would be possible by using ballast water pumps, or
through gravity when there is adequate head pressure from the sea. The piping would be direct, through
a pipe tunnel for ease of monitoring condition and servicing, and have no cross-connects.
In Tank
Once a ballast water tank is full or partially full, it would be periodically mixed through the use of low-
pressure high-volume air bubbles, or in-tank eductors. This mixing would allow the application of an
oxidant to a prescribed level, and the monitoring and the maintenance of that oxidant level. Mixing
frequency would be based on detected oxidant decay levels, as well as calculations to prevent sediment
from settling.
The tanks would be fitted with pressure-vacuum relief valves that only open when the ballast water is
being transferred or occasionally to relieve built-up pressure or vacuum from a diurnal cycle. The
gauging system would be a closed system to limit contaminants from entering the tanks. At least two
tank vents would be installed. Each vent would be fitted for ready connection to ventilation blowers to
facilitate gas-freeing tanks to make safe for personnel entry.
Depending on the required oxidant level, the ballast tanks might also require a special coating system. In
addition, piping system gaskets and valve seals might also require special materials not typically used in
seawater applications.
Discharge
Each tank would be fitted with piping for deballasting with a high suction at approximately 300 mm
above the tank bottom, and a low suction at approximately 75 mm above the tank bottom. The high
suction would be used for ballast tank discharge, such that the discharge does not contain sediment. The
low suction would be used for stripping sediment from tanks when suitable disposal facilities are
available.
The discharge piping would be independent from the uptake piping. Each tank would be outfitted with
an isolation valve connecting it to the discharge main header. The header would lead to a reactor tank of
one-meter height built into the ship's double bottom with at least 25 m3 capacity, allowing a contact time
of at least three minutes. During the contact time, the oxidant level would be neutralized and water
quality confirmed prior to discharge. The system would be failsafe, returning the ballast water to the
ballast water storage tank if needed.
A dedicated seawater overboard, designed to avoid contamination from ballast water uptake or other
sources, would be fitted for discharging the ballast water. The disinfection step would be as close as
55
-------
practical to the overboard. This final disinfection step would provide assurance against contaminants in
the reactor tank where the oxidant was neutralized, as well as providing a measure of caution in treating
the ballast water a second time by a different process.
The ballast water could be moved through the discharge by gravity if there is adequate head in the
ballast tank. At any time, a pump would take suction on the reactor tank, avoiding pump contact with the
oxidants. The pump would then discharge to the UV unit and overboard.
Summary
In summary, reaching the USCG Phase-2 standard, or even 100 times the EVIO D-2/ Phase 1 standard,
will likely require wholly new treatment systems. Such new systems will have many attributes different
from existing BWMS. They will use new technological devices, including those drawn from the water
treatment industry; employ multistage treatment processes; emphasize technological process controls
and multiple monitoring points to ensure desired performance, rather than rely on end-of-pipe testing;
include physical barriers to minimize the potential for cross-contamination of the system; and become
part of an integrated ballast water management effort. These new approaches will achieve higher
performance, but will require time to develop, test, and determine their practicality and cost.
In addition, new BWMS technologies will need to become more energy-efficient. Driving factors
include rising fuel costs, potential future valuations or other constraints on air emissions and other
pollutants, and potential future taxes of carbon sources from maritime shipping. To date, attempts to
meet proposed discharge standards generally have increased the energy required for ballast
management. New BWMS methods should attempt to reverse this trend. Recent innovations have
significantly reduced the volume of discharged ballast water, and in some cases eliminated discharges in
all routine operations. Such direct approaches should continue and their reduced environmental impact
should be recognized and encouraged in regulatory, monitoring and enforcement efforts. These
approaches are discussed further in response to charge question 4.
5.4. Recommendations for Addressing Impediments and Constraints
Charge question 3c. What recommendations does the SAB have for addressing these impediments and
constraints?
Several existing technologies have demonstrated compliance with the D-2 standard during testing
periods. However, it is not clear that these BWMS will operate consistently at this level of performance
on board the many thousands of vessels that will require their use. Clearly defined and transparent
programs for compliance monitoring and enforcement are needed to promote consistent, reliable
operation of BWMS; such programs do not yet exist. Ideally, vessel crew members would have the
technological capability to monitor BWMS efficacy, and make real-time corrections to maintain
compliance. Further, it is important that BWMS manufacturers document and report performance
metrics beyond discharge treatment efficacy. This information would enable vessel operators to select
systems that best integrate with their operations. For example, the ETV protocol provides guidance for
third-party evaluation of factors such as energy consumption and reliability. These and similar metrics
should be encouraged.
56
-------
Although meeting significantly higher standards will likely require completely new treatment
approaches, the Panel can predict neither which combination of treatment processes will achieve the
highest efficacy nor their ultimate performance. The Panel recommends that one or more pilot projects
be commissioned to explore new approaches to ballast water treatment, including tests of ballast water
transfer and treatment at a reception facility
5.5. Impediments Based on Organism Type
Charge Question 3d. Are these impediments more significant for certain size classes or types of
organisms (e.g., zooplankton versus viruses)?
Shipboard impediments apply to all size classes of organisms and specified microbes. This broad
conclusion is based on analysis of BWMS test results, as well as general considerations of the treatment
processes and the vessel application constraints.
With regard to specific technologies, however, BWMS performance varies across target organisms. For
example, existing BWMS are capable of removing (e.g., mechanical separation) or killing (e.g.,
deoxygenation, UV, chlorine dioxide) the great majority and in some cases, nearly all organisms > 50
|im; UV irradiation kills or inactivates unicellular organisms and viruses more efficiently than it does
metazoans; and deoxygenation does not eliminate bacteria but rather alters microbial communities.
Such variation among organisms is exemplified by testing data. Section 4 of this report reviews results
of seven BWMS that "reliably met" the EVIO D-2 standard. All treatment systems were limited in their
ability to reach extremely stringent, proposed standards for total bacteria. In addition, although they met
EVIO D-2 standards, some live organisms were found in either one or both of the > 10 to < 50 jim and >
50 |im size classes. In summary, these data indicate that current technology is broadly challenged by
bacterial counts and sometimes selectively challenged by both > 10 to < 50 jim and > 50 jim size
classes.
5.6. Sterilization of Ballast Water Discharge
Charge Question 3e. Can currently available treatment processes reliably achieve sterilization (no
living organisms or viable viruses) of ballast water onboard vessels or, at a minimum, achieve zero or
near zero discharge for certain organism size classes or types?
It is an unrealistic and unattainable goal for current BWMS to yield ballast discharge that is "sterile",
i.e., "free from living organisms and viruses" (Madigan and Martinko 2006). Given the volumes of
water requiring treatment, sterilization is not possible using current technologies; there simply is not
enough energy on a vessel to implement steam autoclaving of its ballast water tanks and piping systems.
With respect to "zero or near zero discharge", however, technology exists to remove most organisms in
the size classes > 10 to < 50 jim and > 50 jim. As a practical matter, the Panel notes that it is not
possible to measure zero (sterilization) or near zero discharge—especially for microorganisms such as
phytoplankton, bacteria, and viruses, which are especially difficult to differentiate as "live" or "dead" on
the basis of physiological certainties. If such values cannot be measured, a BWMS cannot be controlled
to ensure zero or near-zero discharge at the "end of pipe" for a working vessel.
57
-------
6. LIMITATIONS OF EXISTING STUDIES AND REPORTS
6.1. EPA's Charge Question
This section responds to Charge Question 4: "What are the principal limitations of the available studies
and reports on the status of ballast water treatment technology and system performance and how can
these limitations be overcome or corrected in future assessments of the availability of technology for
treating ballast water onboard vessels?" Bearing in mind the broader charge to "provide advice on
technologies and systems to minimize the impacts of invasive species in vessel ballast water discharge"
(Feb. 2010 Federal Register notice), this section addresses aspects of ballast water discharge not covered
in the responses to earlier charge questions. Several themes emerged which the Panel discusses in the
following sections. First, improved methods for testing and reporting are needed to ensure that high
quality data are available with which to assess BWMS performance. Second, improved data also are
important to the development of effective approaches to enforcement and compliance. Third, existing
data and reports on the effectiveness of practices and technologies other than shipboard BWMS are
inadequate because insufficient attention has been given to integrated practices and technologies that
could reduce the risk of invasions. These include managing ballast water uptake to reduce presence of
invasive species, reducing invasion risk through operational adjustments and changes in ship design to
reduce or eliminate the need for ballast water, and consideration of land-based reception facilities for
ballast water treatment. Voyage-based risk assessments could be used to integrate such practices,
through use of applied risk management principles such as Hazard Assessment and Critical Control
Points methods.
6.2. Testing Shipboard Treatment Systems: Protocols, Analysis, and Reporting Practices that
Could Be Improved
This section applies to test facilities both in the U.S. and abroad and was informed by the ETV Protocol
for land-based verification of BWMS performance (U.S. EPA 2010, hereafter the Protocol). The Panel
acknowledges the many efforts put forth by various technical panels and stakeholder groups over many
years to draft, validate and finalize the Protocol. Most of this section focuses on land-based verification
testing (used to gain Type Approval from Maritime Administrations) rather than shipboard verification
testing (also used to gain Type Approval) or compliance testing (used to determine adherence to any
discharge standard when a vessel enters a port of call). This is because, to date, programs that address
these types of testing have not been finalized in the U.S.
6.2.1 Confusion of Research and Development and Certification Testing
In some cases, little if any distinction is made between research and development (R&D) testing and
verification testing. Adjustments to BWMS often are made during testing of prototypes and, in some
cases, only the most favorable results are reported. Thus, certification may be gained on the basis of
unrealistically favorable results that may not be representative of replicated testing with multiple
commercially available units of a BWMS. To address this problem, the Protocol requires that BWMS
undergoing verification testing are "prefabricated, commercial-ready treatment systems" and that all test
results be reported (U.S. EPA 2010). Given the early state of the BWMS industry, mass-produced
58
-------
assembly line systems are not currently tested. As indicated in the Protocol, R&D testing should be
barred from use in certification testing.
To ensure that the performance of ballast water treatment systems is objectively and thoroughly
evaluated during verification testing, experienced specialists in an independent testing organization
should conduct the tests (as required in the Protocol), rather than the system manufacturers. This is
important because research has shown that it is extremely difficult for system creators—who have
constructively designed their systems—to change their perspective and instead view their system from
the "deconstructive" state of mind that is focused on finding flaws and exposing weaknesses and
limitations (Myers 1979). Thus, it is critically important that verification testing be conducted by
independent specialists in order to assess system performance in a scientifically rigorous way. Further,
as noted in the Protocol, the credentials of these personnel should be approved by the Verification
Organization (the entity that oversees testing preparation, testing and the Verification Report issued by
the test facility at the conclusion of testing). In sum, testing should be conducted by a party that is
independent from the manufacturer and has appropriate, established credentials.
6.2.2 Lack of Standardized Testing Protocols
Comparative evaluations of the performance of different BWMS are hampered by inconsistencies in
discharge standards and in testing protocols. As shown in Table 2-1, there are diverse state, national and
international discharge standards for ballast water—including differences in limits that vary by orders of
magnitude for similar categories of organisms. This range of standards not only results in confusion for
the regulated industry but also provides significant challenges for testing of BWMS. Performance
standards set requirements for technology to achieve and should help to advance progress in treatment
system designs, but only if a set of standardized, practical, scientifically rigorous assessment techniques
is available to evaluate system performance. The EVIO standards are based upon different size groups of
organisms, and all size groupings pose challenges for assessing performance.
Comparison of the performance of different ballast water treatment technologies requires consistent
testing protocols regardless of the target discharge standard (Phillips 2006; Ruiz et al. 2006). To date, all
BWMS have been evaluated using the basic approaches provided by the EVIO Guidelines for Approval
of Ballast Water Management Systems (G8) and the Procedure for Approval of Ballast Water Systems
that Make Use of Active Substances (G9) (EVIO 2008a,b). While G8 and G9 suggest a basic framework,
the level of detail required for rigorous and comparable BWMS testing is lacking. The state of
California also has developed ballast water treatment technology testing guidelines that are intended to
provide a standardized approach for evaluating treatment system performance (Dobroski et al. 2009).
Procedures also are are being developed for verifying vessel compliance with California performance
standards.
The 2010 Protocol is a federal program that is much more detailed and proscriptive regarding test
facility design, sampling design and volume, sample handling, analytical methods, data reporting and
QA/QC requirements. However, this Protocol has yet to be implemented in practice or broadly adopted.
Thus, at present there is no broad international program that includes performance standards, guidelines,
and protocols to verify treatment technology performance, and no standardized sets of methods for
sampling and analysis of ballast water to assess compliance. The existing federal and various state
standards lack consistency as well. Treatment evaluations generally are designed to test whether a given
59
-------
technology can meet IMO D-2 standards in accordance with the IMO G8 or G9 Guidelines (IMO 2008
a,b).
With exception of BWMS installed aboard vessels enrolled in the U.S. Coast Guard's (USCG)
Shipboard Technology Evaluation Program (STEP), BWMS presently are not approved for use in
compliance with proposed federal ballast water management requirements. Thus, while there are various
state ballast water management requirements, there is no formal U.S. type approval program for BWMS
(although one is described in the USCG proposed final rule). The EPA has, however, included
provisions in the draft NPDES Vessel General Permit for vessels with treatment systems that discharge
ballast water containing biocides or chemical residues.
Performance standards set requirements for technology to achieve and should help to advance progress
in treatment system designs, but only if a set of standardized, practical, scientifically rigorous
assessment techniques is available for use in assessments. All existing and proposed performance
standards are based upon different size groups of organisms, and all size groupings pose challenges for
accurate assessments of treatment performance (see below). In the IMO D-2 performance standard,
organisms in the < 10 jim size class are represented by a subset of taxa consisting of three indicator
bacteria or bacteria groups (Vibrio cholerae, Escherichia coli, and Enterococci). Assessment has relied
upon a subset group of organisms as representative of treatment of all bacteria (see Section 6.2.3,
below). There is as yet no strong evidence for suitable proxy organisms to represent the virus size class,
and no acceptable methods to verify compliance with a total virus standard (which is in the proposed
USCG Phase 2 performance standard).
The following section provides Panel recommendations for future versions of the ETV Protocol, thus it
focus on differences from the Protocol rather than reiterating recommendations made in the Protocol.
Because the Protocol pertains only to land-based verification testing (not shipboard testing), the Panel's
recommendations focus on land-based testing for verifying treatment performance by independent
testing operations. This section also comments on shipboard testing.
Test Verification Factors
The Protocol recommends that all treatment systems be verified using the following factors: biological
treatment efficacy (or BE, defined as the removal, inactivation or death of organisms), operation and
maintenance (O&M), reliability as measured by the mean time between failures (MTBF), and
environmental acceptability including residual toxicity and safety. The Panel agrees with the Protocol
that biological treatment efficacy should be measured as the concentration, in the treated ballast water
discharge, of the organism size classes indicated in the EVIO D-2 and USCG Phase 1 performance
standards, with a minimum concentration of organisms in the control tank discharge. Other
measurements can include water quality parameters in comparison to appropriate water quality
standards. Verification protocols should include detailed descriptions of on-site sampling, sample
handling (chain of custody), QA/QC, in-place mechanisms for selecting independent laboratories with
appropriate expertise and certification to conduct the sample analyses, and requirements for compliance
reporting.
The Panel also agrees with the Protocol that tests and species selected for toxicity testing during
commissioning need to have carefully justified protocols detailed in the Test Plan. BWMS that involve a
chemical mode of action are regulated under the National Pollutant Discharge Elimination System
60
-------
(NPDES) permit process (Albert et al. 2010), which requires demonstration of "no adverse effects" as
evaluated through chemical-specific parameters and standardized Whole Effluent Toxicity (WET)
testing (U.S. EPA 2002a-c; 40 CFR 136.3, Table 1 A). WET experiments are designed to assess the
effects of any residual toxicity on beneficial organisms in receiving waters. Standardized acute and
chronic toxicity assays have been developed by the EPA for a limited number of freshwater and marine
species (Table 6-1). The Protocol does not include specific freshwater assays, but recommends that
toxicity tests for biocide treatments in brackish and marine waters should be selected from the EPA
acute toxicity assay for mysids (EPA OPPTS Method 850.1035;
http://www.epa.gov/opptsfrs/OPPTS_Harmonized/850_Ecological_Effects_Test_Guidelines/Drafts/850-
1035.pdf), and the chronic toxicity assays for the inland silverside Menidia beryllina (larval survival
and growth, EPA Method 1006.0; http://www.epa.gov /OST/WET/diskl/ ctml3.pdf) and the sea urchin,
Arbaciapunctulata (fertilization, EPA Method 1008.0; http://www.epa.gov/OST/WET/diskl/ctml5.pdf
). The Panel recommends that freshwater assays also be included in toxicity testing.
The Protocol also recommends that complete results of verification testing, including equipment
failures, be reported as standard practice. These data are needed to enable realistic evaluation of a given
BWMS. At present, there is no requirement under the EVIO G8 guidelines to report tests in which a
BWMS does not perform to the D-2 performance standard. The Panel strongly recommends that reports
should include all test results, and that criteria for approval should consider the failure rate (proportion
of tests that were successful).
Table 6-1. Freshwater and marine species for which the U.S. EPA has developed standardized acute
and chronic toxicity assays (http://www.epa.gov/waterscience/WET').
Habitat
Acute Toxicity
Chronic Toxicity
Freshwater
Algae
Zooplankton
Fish
Marine
Mysid shrimp
Sea urchin
Fish
Ceriodaphnia dubia
Daphnia magna
Daphnia puplex
Bannerfin shiner (Cyprinellale edsi)
Brook trout (Salvelinus fontinalis)
Fathead minnow (Pimephale spromelas)
Rainbow trout (Oncorhynchus mykiss)
Americamysis bahia
Sheepshead minnow (Cyprinodon
variegatus)
Silversides (Menidia beryllina, M.
menidia, M. peninsulae)
Selenastrum capricornutum (growth)
Survival, reproduction
Larval survival, growth; embryo-larval
survival, teratogenicity
Survival, growth, fecundity
Arbacia punctulata - fertilization
Larval survival, growth; embryo-larval
survival, teratogenicity
M. beryllina - larval survival, growth
61
-------
Challenge Conditions
The Panel recommends that testing should be applied across the gradient of environmental conditions
(temperatures, salinities) represented by the Earth's ports; to address this concern, the ETV Protocol
requires testing at a minimum of two salinities (U.S. EPA 2010), although some Panel members argued
the minimum should be three. All treatment technologies should function well across the range of
physical/chemical conditions and densities/types of biological organisms that a ship encounters. Thus,
BWMS ideally should be verified using a set of standard challenge conditions that encompass the suite
of water quality conditions, and that capture environmental conditions represented by ports and a range
of densities of the organisms and organism size classes (unless a BWMS is designed, and certified, for
only a specific subset of conditions).
The ETV Protocol states that the objectives for challenge conditions are to verify treatment system
performance using a set of "challenging, but not rare, water quality conditions representative of the
natural environment," and to verify removal or kill of organisms ranging in size from bacteria to
zooplankton, using natural assemblages and appropriate analytical techniques that enable quantification
of densities of live organisms (U.S. EPA 2010, p.18). It is important to evaluate the effectiveness of
treatment systems under conditions that challenge the technology because certain water quality
conditions can interfere with some treatment processes. These physical/chemical environmental
conditions are generally understood and relatively few in number, which helps to limit the number of
water quality metrics that must be included in the protocol (Table 6-2).
62
-------
Table 6-2. Comparison of the ETV Protocol's recommendations (U.S. EPA 2010) and the alternatives the Panel recommends that EPA consider,
with respect to minimum criteria for challenge water total living populations, criteria for a valid biological efficiency (BE) test cycle at land-based
facilities (living organisms in control tank discharge after a holding time of at least 1 day), and water types (salinity groupings) for completion of BE
tests.1 Three salinity ranges are recommended for BWMS that are planned for use in freshwater, brackish, and marine waters.
Minimum Criteria for Challenge Water Total Living Populations; and
Criteria for a Valid BE Test Cycle - Living Organisms in Control Tank Discharge After 1 Day Holding Time
Size Category
> 50 urn
> 10 um and < 50 um
Other3
ETV Protocol
105 organisms m"3, 5 species in 3 phyla
103 organisms mL"1, 5 species in 3 phyla
< 10 um: 103 ml/1 as culturable
aerobic heterotrophic bacteria
Water Types (Salinity Groupings) for Completion of BE Tests
Fresh (salinity < 1)
Brackish
Marine
Two salinity ranges;
brackish = salinity 10-20;
marine = salinity 28-36
Physical/Chemical4
Environmental:
Others of Specific Interest:
Panel Alternatives
same
same
< 10 um: 103 selected protists mL"1
< 2 um: same as ETV for < 10 um
Two or three salinity ranges;
brackish = salinity 1 to < 28
marine = salinity > 28
Temperature (4-3 5°C), DOC, POC, TSS, MM, pH, DO
Example - nutrient concentrations (TN, TP, TKN, NHX, NOX
SRP)
Abbreviations: DOC, dissolved organic carbon; POC, particulate organic carbon; TSS, total suspended solids = particulate organic
matter (POM) + MM (mineral matter); DO, dissolved oxygen; TN, total nitrogen; TP, total phosphorus; TKN, total Kjeldahl nitrogen;
NHX, ammonia + ammonium; NOX, nitrate + nitrite; SRP, soluble reactive phosphorus.
Size = maximum dimension on the smallest axis.
3 Effects on culturable aerobic heterotrophic bacteria are assumed to be indicative of effects on all bacteria.
The ETV Protocol's water quality challenge matrix for verification testing includes the following minimum water characteristics for the three
salinity water types as: Dissolved organic matter, 6 mg L"1 as DOC, and particulate organic matter, 4 mg L"1
as POC; MM 20 mg I/1 and TSS 24 mg I/1; and temperature range 4-35°C.
63
-------
In recognition of the difficulties that can be encountered, especially in ship-based testing, tests of the
three salinity ranges could include two land-based tests and one ship-based test. A rationale for
recommending tests of all three salinity ranges is that if a given BWMS is planned for use across all
three salinity ranges, but testing indicates that its efficiency at organism removal is poor under one or
more of the salinity groupings, then that system should not be used by ships visiting ports that are
characterized by such conditions. Similarly, if a BWMS is planned for use across other environmental
gradients (e.g., temperatures from cold to warm waters, or salinities from fresh to marine), but tests
indicate that it has poor efficiency in removing biota under part of the natural range, then that system
should not be used by ships visiting ports that have such conditions. Indeed, the USCG proposed rule
indicates that "at least 2 sets of test cycles should be conducted with different salinity ranges and
associated dissolved and particulate content as described. BWMS not tested for each of the 3 salinity
ranges and water conditions listed in this section may be subject to operational restrictions within a
certificate of approval" (USCG 2009, p. 44666). A fully crossed design should be used where possible,
for example, if natural water can be obtained at the desired salinity range. As another example, cold
water testing may be critical to understand the breakdown of chemical treatments (i.e., active
substances), but testing with natural water in the winter would encounter relatively few organisms in the
challenge water, making it difficult to achieve recommended challenge conditions.
There are other major practical constraints on such tests. First, alterations to establish the natural range
of physical and chemical conditions should be imposed without affecting the concentrations, diversity
and viability of the biota present. For that reason, natural water sources ideally should be used to impose
the levels of salinity, rather than artificially modified salinity. Artifactual interactions may occur
between biota and artificial media (e.g., artificial seawater prepared with commercially available "sea
salts"). The Panel thus diverges from Anderson et al. (2008) in recommending that a source of filtered,
high-quality natural freshwater or seawater should be used to prepare treatments insofar as possible.
There are pros and cons with either approach: Artificial sea salts are expensive but enable routine
preparation of media. However, caution is warranted in using artificial sea salts because some
ingredients that are not found in natural seawater, such as phthalate esters (e.g., di(2-
ethylhexyl)phthalate, a commonly used plasticizer in Instant Ocean aquarium salts), are abundant and
can be toxic to aquatic life, resulting in spurious data (e.g., Peal 1975; Moeller et al. 2001).
In addition, various dissolved organic compounds that are important to the nutrition and the life histories
of aquatic organisms likely will be missing from artificially constructed media. While use of natural
waters avoids such problems, the natural water source should be as free as possible from toxic
pollutants, which are increasingly ubiquitous in fresh, brackish, and coastal marine waters (Kay 1985;
Pate et al. 1992; Loganathan and Kannan 1994; Hoff et al. 1996; U.S. EPA 2000; Shaw and
Kurunthachalam 2009), or contain at most only trace levels of such pollutants. Final selection of natural
versus an artificial water sources requires careful consideration of these issues. In addition, when using
artificial water sources or otherwise modifying environmental conditions, timing is important; care
should be taken to avoid imposing rapid environmental changes that, alone, could stress or kill the biota
tested.
Similarly, the Protocol recommends adjusting parti culate organic matter (POM), if natural waters do not
meet challenge conditions, by adding commercially available humic materials, plankton, detritus, or
ground seaweed; commercially available clays can be added to adjust the mineral matter concentration
(U.S. EPA 2010). However, the Panel is concerned that the cation exchange capacity of the dried, then
rehydrated, clays can significantly alter plankton communities (Avnimelech et al. 1982; Burkholder
64
-------
1992; Cuker and Hudson 1992). Artificial modification of dissolved organic carbon (DOC) is difficult to
achieve without a strong potential of affecting the biota present, especially the smaller size-fraction
components. The Panel believes that the testing organization should be required to verify, insofar as
possible, that in preparing the test water, any materials added had minimal effects on the biota, and
"minimal effects" should be clearly defined.
The EVIO (2008a, b), the Protocol, and other suggested standards (e.g., California VGP 401
certification/State regulations; see Albert et al. 2010) make no mention of protists in the < 10 jim size
range. Many harmful organisms occur in this size range (e.g., harmful "brown tide" pelagophytes
Aureococcus anophagefferem and Aureoumbra lagunensis, many harmful cyanobacteria, and certain
potentially toxic dinoflagellates; see Burkholder 1998, 2009). The selected bacteria presently targeted
for standards are not useful as indicators for these taxa which, as a general grouping, can adversely
affect both environmental and human health (Burkholder 1998, 2009). Thus, failure to consider this size
class represents a serious omission in efforts to protect U.S. estuaries, marine waters and the Great
Lakes from harmful species introductions. For some of these taxa, such as toxigenic Microcystis spp.
affecting some of the Great Lakes (e.g., Boyer 2007), the tendency of the cells to aggregate into colonies
can sometimes "boost" them into the > 10 jim size range during filtration processes, but the size
measurements are based on individual cells.
There is a critical need to consider harmful representative protists (which should be expected to vary
depending on the geographic region) from this size class in developing protective ballast water
standards. Depending on the salinity and the region, and based on the smallest cell dimension, examples
of candidates could include selected toxigenic cyanobacteria such asAnabaenaflos-aquae, the
haptophyte Prymnesium parvum, brown tide organisms Aureococcus anaphogefferens or Aureoumbra
lagunensis, small toxigenic dinoflagellates such as Karlodinium veneficum., and the pathogenic
protozoans Giardia spp. and Cryptosporidiumparvum that are found across the salinity gradient.
Accordingly, protists in this size class should be included in standards for assessing the performance of
BWMS in land-based testing if they were naturally occurring at the test facility. For shipboard
verification testing or compliance testing, where the source water is unknown until sampling occurs,
appropriate organisms for evaluation can be selected accordingly.
Verification Testing
The Panel offers considerations that differ from the Protocol on some points, including specifics for
collecting water quality and biological samples for verification testing of BWMS (Table 6-3). Some
panel members argue that these points should be a part of verification testing; others argue that such
approaches could be incorporated into future revisions of the Protocol if their utility, effectiveness, and
practicability were deemed appropriate. Some Panel members have concerns that such analyses would
(1) provide qualitative indications of viability, not the quantitative data on the density of organisms
necessary to assess BWMS performance in accordance with a discharge standard, or (2) overestimate
the number of viable organisms. Also, conducting new analyses, in addition to those required in the
Protocol, might not be practicable by already busy testing teams.
For zooplankton, phytoplankton and other protists, the Panel supports the need for collecting at least 3-6
m3 of sample volume at each required location on a time-averaged basis over the testing period. Field
quality control samples and field blanks should be taken under actual field conditions to provide
information on the potential for bias from problems with sample collection, processing, shipping and
65
-------
analysis (Ruiz et al. 2006). Accepted scientific methods should be used for all analyses (e.g., for water
quality parameters, U.S. EPA 1993, 1997; American Public Health Association (APHA) et al. 2008).
Biological samples should be collected in a time-integrated manner during the tests, and sample
collection tanks should be thoroughly mixed prior to sampling to ensure homogeneity (U.S. EPA 2010).
Samples collected from control and treated tank discharges should be taken upstream from pumps or
other apparatus that could cause mortality or other alterations, and if pumps and valves must be used
upstream of sample collection, they should previously have been tested and shown not to damage
organisms (U.S. EPA 2010). Note that analysis of some parameters is extremely time-sensitive (Table 6-
3). For example, zooplankton die-off occurs in some samples held for 6 hours or more (U.S. EPA 2010);
the timing of die-off likely varies depending on the zooplankton community, and the upper limit should
be determined at each test facility. The approximate maximum hold times should maintain detectable
zooplankton mortality over time at < 5%.
66
-------
Table 6-3. Sample volumes, containers, and processing for core parameters and auxiliary nutrients (nitrogen, N;
phosphorus, P; silicate, Si; carbon, C). Note that HDPE = high-density polyethylene, and POC information is from
Baldino (1995). Recommendations that differ from those in the ETV (U.S. EPA 2010) are indicated in bold.1
Parameter
TSS
DOC
Minimum Sample
Volume
100 mL
25 mL
Containers
HDPE or glass
glass
Processing/Preservation
Holding Time
Process immediately or
store at 4°C
Pre-combusted GF/F filters;
Maximum
1 week
28 days
POC
MM
DO
Chlorophyll a,1
pheopigments
Phytoplankton No.2
(viable, < 10 u,m -
selected harmful taxa)
500 mL
= TSS - POC
300 mL
or
in situ sensor
400 mL
500 mL
preserve filtrate with H3PO4
(pH < 2), hold at 4° in darkness
(APHA et al. 2008)
HDPE Filter (GF/F in foil); freeze filter 28 days
until analysis
glass BOD Fix (Oudot et al. 1988); titrate in 24 hours
bottles 2-24 hours; or
Continuously recording
dark HDPE Filter (GF/F); fix with saturated 3 weeks
MgCO3 solution; freeze filter until
analysis
dark HDPE Filter (Nuclepore or Anotech); process
assess autofluorescence (e.g. immediately
Maclsaac and Stockner 1993), or
Filter, fix (e.g. 0.2% (v/v) 3-4 weeks
formalin), freeze filter; or
filter, fix, followed by selected months
molecular techniques (e.g.
Karlson et al. 2010)
67
-------
Table 6-3 (cont.)
Parameter Minimum Sample
Volume
Phytoplankton No . 3 m3 ( 1 ,000 L)
(viable, nano-/ — > 1 L
micro-plankton,
> 10 to 50 (inland
> 50 urn) 3
Containers Processing/Preservation
Holding Time
60 mL Viable: No preservative; stain
dark HDPE with FDA, CMFDA; or,
fix with acidic Lugol' s solution
(Vollenweider 1974), store at
4°C in darkness, and quantify
as viable when collected
and
combine with various molecular
techniques to confirm harmful
taxa of interest (e.g. Karlson et
al. 2010)
Maximum
process
immediately; or
28 days,
preferably
1 week
Other protists (#) 1
(viable heterotrophs,
< 10 u,m - selected
harmful taxa)
Zooplankton #
(viable, > 50 |im)
Zooplankton #
(viable) (cont'd.)
Bacteria
(active culturable,
aerobic heterotrophic •
selected taxa)
Nutrients1 -
TN, TP total
Kjeldahl N (TKN)
500 mL
3 m3 (3,000 L)'
>1000mL
60 mL
100 mL, Techniques appropriate for the variable
dark HDPE selected taxa (e.g. U.S. EPA
2005)
1-L flask No preservative; subsample Process
450 1-mL wells3 and probe; immediately
fix with buffered formalin and (< 6 hr)
Rose Bengal's solution to
quantify; or
fix as above and quantify as Process within
formerly viable (Johnson and 1 month
Allen 2005)
sterile HDPE Plate on appropriate media Process
(U.S. EPA 2010)6 Immediately
Varies - see standard methods varies (mostly
(U.S. EPA 1993, 1997; APHAet 28 days)
al. 2008; and U.S. EPA 2010, p.39)
68
-------
Table 6-3 (cont.)
Parameter
NOXN, NHXN, SRP,
Si02
Minimum Sample
Volume
60 mL
Containers Processing/Preservation
Holding Time
varies Varies - see standard methods
as above
Maximum
varies (mostly
28 days)
1 Methods in the ETV Protocol that differ from those and recommended above by the Panel for consideration are as follows: DOC - pass sample through a GF/F filter and freeze
filtrate until analysis; chlorophyll a and pheopigments - listed as an auxiliary parameter rather than a core parameter; protists (phytoplankton, protozoans) in the < 10 |Jm size class
are not considered; TN, TP, total Kjeldahl N, and silica are not addressed; and dissolved inorganic phosphate is referred to here as soluble reactive phosphorus (SRP).
In situ sensors are available for measuring chlorophyll a as relative fluorescence units, but not as chlorophyll a concentrations. Chlorophyll a may be considered as a core parameter
or as an auxiliary parameter, used as a collective indicator for algal biomass. The Panel also recommends assessment of nutrients (TN, TP, total Kjeldahl N, and silica) if possible,
although nutrients are not considered as core parameters by the ETV and the Panel recognizes that core parameters should have top priority.
The Panel also recognizes the fact that present performance standards are for living (viable) organisms. Because of widely acknowledged practical limitations in techniques to assess
living (viable) organisms, in this table and explained in the writing below, the Panel suggests alternate techniques for quantifying the organisms that were viable in samples when
collected.
This size category has not been considered for ballast water treatment standards by IMO (2008a,b), the ETV (U.S. EPA 2010), etc. Because many harmful organisms occur in the <
10 |im grouping, this size class should be considered for inclusion in assessment of BWMS.
3 FDA, fluorescein diacetate; CMFDA, 5-chloromethylfluorescein. Delicate protists (e.g. wall-less flagellates) mostly would not be expected to survive the process of rapid
concentration of large- volume samples. The Panel recognizes that samples collected for protists are not concentrated in the same way as samples collected for the larger size class: For
the P-l standard, 3-6 L (taken as whole water, isokinetically from the discharge of control or treatment tanks) are concentrated to 1 L, which can be done with a sieve. Nevertheless,
this process can lyse delicate protists. As a more practical alternative than attempting to quantify viable algae and other protists from unpreserved samples, an option is to preserve
samples immediately upon collection and then assessing intact organisms as "viable when collected," based on the fact that protists such as most algae in this general size class are
known to lyse and/or decompose rapidly (minutes to several hours) after death, so that the cell contents become distorted or are lost even if the cell coverings remain (Wetzel 2001). It
should be noted that vital (or mortal) stains address the question, "Is this alga living?" in a way that is substantially different than the presence/absence of intact chloroplasts and other
cell contents. Thus, the two methods sometimes do not yield the same quantitative answer and require careful calibration.
It should be noted that phototrophic organisms in this size class should be quantified using the protocols for phytoplankton, above.
5
Zooplankton die-off occurs in samples held for 6 hours or more (Naval Research Laboratory, unpubl. data; U.S. EPA 2010).
6 Media suggested by the U.S. ETV (2010, p.47) for brackish/marine taxa include 2216 Marine Agar and salt-modified R2A agar; media for freshwater species may include Plate
Count Agar and Nutrient broth (plus agar (15 g L"1).
69
-------
An alternative to quantifying viable organisms from unpreserved samples is preserving samples
immediately upon collection and then assessing intact organisms as "viable when collected." Some
Panel members recommend this as a more practical alternative, whereas others note vigorous validation
is needed before it can be recommended. This approach is commonly used in characterizations of
microflora and microfauna assemblages in the peer-reviewed literature, based on the fact that protists
and zooplankton deteriorate quickly once dead (within minutes to hours: Wetzel 2001; Johnson and
Allen 2005; and see Section 6.2.3). Effective "fast-kill" preservatives can be used that cause death
before distortion or cell lysis can occur. Standardized, accepted techniques are available for quantifying
"viable when collected" protists and zooplankton from preserved material (e.g., Lund et al. 1958; Wetzel
and Likens 2000; Johnson and Allen 2005; and see Section 6.2.3). A shortcoming of this approach is
that dying organisms which still contain apparently intact cellular contents would be included in the
"viable" estimate. Thus, the number of viable organisms would be overestimated if a large fraction of
the sample was dead. In addition, as for counts based on unpreserved material, it is difficult to assess
whether some resistant structures such as thick, opaque cysts contain organisms with intact cell contents.
Because of practical and environmental health/safety constraints, neither approach avoids the problem of
likely major losses of viable organisms that occur during rapid concentration of large sample volumes. If
a "formerly viable" approach were used, the Panel recognizes that it would need to be validated and
approved by the Verification Organization.
6.2.3. Compromises Necessary Because of Practical Constraints in Sampling and Available
Methods
Ideally, the goal of standard challenge conditions would include the full range of (a) challenging
conditions present in the world's ports, (b) organism density, (c) taxonomic diversity, and (d) organism
size classes. Meeting this ideal goal is impeded by several serious practical constraints in sampling large
ballast tanks effectively, and in the methods that presently are available for quantifying viable
organisms. As Lee et al. (2010, p. 19) pointed out, "perfect compliance and no failure is practically, if
not theoretically, impossible, particularly for microbiological organisms unless ballast water is
discharged into a land-based treatment facility or ships are redesigned to eliminate the need to discharge
ballast water." This section considers how the ideal can be modified to accommodate practical
considerations while accomplishing a meaningful evaluation of the efficacy of BWMS.
Standardization of Choices of Standard Test Organisms
The Protocol defines standard test organisms (STOs) as "organisms of known types and abundance that
have been previously evaluated for their level of resistance to physical and/or chemical stressors
representing ballast water technology" that are used in bench-scale testing (U.S. EPA 2010, p. 74). The
Protocol (U.S. EPA 2010) requires that prior to full-scale verification testing, laboratory experiments be
conducted to evaluate post-treatment viability of STO taxa used to assess the biological effectiveness of
BWMS in removing zooplankton, protists (heterotrophic and phototrophic), and bacteria.
The selection and development of STOs that are broadly resistant to treatments for use in testing BWMS
performance is a fertile field of research because of the practical need (Hunt et al. 2005; Anderson et al.
2008; U.S. EPA 2010). The Panel urges caution in the use of STOs, however, since results from a very
small number of taxa are broadly applied to all of the organisms in the same general grouping (e.g.,
protozoans in a certain size class). An assumption that first must be validated is that the selected taxa are
among the most resistant to treatment, so that most organisms are eliminated when the surrogate taxa are
70
-------
eliminated (Ruiz et al. 2006). The fundamental challenge is to identify the best species that are
"representative" of a broad range of organisms within a given size class. Good candidates are considered
to be easily and economically cultured in large numbers for future full-scale testing in experimental
ballast water tanks; tolerant of a wide range of environmental conditions; reliable and consistent in their
response to treatment across culture batches; and resilient in withstanding ballast water tests and
sampling (Ruiz et al. 2006; Anderson et al. 2008). A list of suggested STOs is provided in the ETV
Protocol. An obvious risk is spurious results from surrogate taxa that poorly represent the larger group
of organisms in a given size class. Because testing with STOs is part of a larger testing process (land-
based and shipboard verification testing) that employs a range of organisms, this risk is somewhat
ameliorated.
Protocols for STOs should include clear justification for use of these taxa under a defined set of
conditions; careful consideration of potential confounding interactions between the STOs and natural
species; and the percentage ratio of challenge organisms that are STOs versus naturally occurring taxa in
the challenge water. Selection of a specific combination of STOs should be based upon extensive testing
at bench and mesocosm scales, preferably by several laboratories located in different geographic
regions, of a wide range of STO species, life histories, habitats, and source regions across environmental
gradients (Ruiz et al. 2006). Consistent use of the same protocols is needed in order to minimize
confounding factors and strengthen comparability. Ideally, several STOs or taxonomic subgroups,
including several life stages, should be included in the tests since confidence in interpretations can be
strengthened by this redundancy. It would also be best to include multiple strains (populations) of
candidate STOs if possible, to account for significant intraspecific variability in response to
environmental conditions that is commonly documented, particularly among protists (Ruiz et al. 2006;
Burkholder and Gilbert 2006). Nevertheless, the Panel recognizes that practical and economic
considerations may prevent testing beyond use of one or two STOs to represent each size class.
Standardization of Choices of Indirect Metrics (Surrogate Parameters)
There are practical and logistical limitations involved in obtaining statistically meaningful estimates of
concentrations of specific organisms per unit volume in compliance testing, as required by the EVIO or
proposed by the USCG. Given these limitations, the Panel recommends adding to future compliance
protocols parameters that are much more rapidly and easily assessed. Examples of candidate "surrogate
parameters" are shown in Table 6-4. They can be calibrated with organism numbers in laboratory tests
on microcosm "ecosystems," but would be much more difficult, if not impractical, to calibrate for use
with unknown types and numbers of organisms in ballast tanks. Therefore, surrogate parameters could
be useful as bulk measurements in compliance testing. These parameters also could be used to augment
existing measurements in land-based and shipboard verification testing. It will be critical to carefully
calibrate all potential surrogate parameters with natural populations of ballast water flora and fauna
before they can be used to evaluate the performance of BWMS - especially at the resolution of very low
organism densities.
Increased Use of Tests at Multiple Spatial Scales
Instead of relying solely on full ship-scale testing, for practical reasons the Panel recommends that
testing be conducted at a combination of scales as needed to address particular issues. Such tests would
be done by the vendor of the BWMS prior to validation testing. For example, full-scale tests can pose
71
-------
Table 6-4. Examples of candidate "surrogate" parameters for quantifying viable organisms in ballast water, and an analysis
of their utility considering methods that are presently available.*
Parameters)
Description
Suitability, Considering Methods Presently Available
Chlorophyll a (chla)
"Signature" or marker
pigments
Adenylates, especially
ATP (adenosine
triphosphate), and
Total adenylates
(ATP + ADP, adenosine
di-phosphate, + AMP,
adenosine mono-
phosphate)
"Universal" plant pigment,
found in all phototrophic algae1;
widely used as an indicator of
total algal biomass2'3
Diagnostic for cyanobacteria
(zeaxanthin) and major eukaryotic
algal groups (e.g. diatoms, other
heterokontophytes - fucoxanthin;
dinoflagellates - peridinin; chloro-
phytes and euglenophytes - chl 6)3'13
Indicator of total microbial biomass
in plankton, sediments17
Pros: Allows rapid processing of large numbers of samples;
standardized methods widely available2"5.
Cons: Cannot discern cell numbers per unit sample volume;
not sensitive enough to detect < 10 cells ml"1 of small algae6;
cellular chla content highly variable (0.1-9.7% fresh weight)
depending on the species7 and the light conditions8'9; methods,
and results depending on the method, vary widely10"12.
Present status: Available methods do not allow reliable
calibration with algal cell numbers in natural samples; improved
methods are needed.
Pros: Potentially superior to chki as algal biomass indicators;
more specificity to algal groups (divisions or classes);
standardized methods available13"16.
Cons: Techniques must be applied carefully to avoid
artifacts and sample bias17; low taxonomic resolution (can
sometimes be improved by screening samples using
microscopy18 to identify abundant taxa5).
Present status: Available methods do not allow reliable
calibration with algal cell numbers in natural samples;
improved methods are needed
Pros: —All microbial taxa have a —constant ratio of ATP to total
cell carbon18; easily extracted from microbial assemblages; not
associated with dead cells or detritus19'20.
Cons: Cell ATP content varies for cells under environ-
mental stress21'22; encysted cells with low metabolic
activity have low ATP content (difficult to detect);
total adenylates considered a better indicator of microbial
biomass than ATP17 within a given size class, but
extrapolation from small sample volumes would lead to
large error factors in estimating organism numbers.
Present status: Available methods do not enable accurate
assessment of small numbers of viable organisms per
unit volume in stressed conditions within ballast tanks;
improved methods are needed.
72
-------
Table 6-4 (cont.)
Parameters)
Description
Suitability, Considering Methods Presently Available
INT (2-^-iodophenyl)-
3-(p-nitrophenyl)-5-
phenyltetrazolium
chloride
RNA, DNA
Commonly used tetrazolium
salt used to measure microbial
activity (electron transport chain
activity indicating viable organisms)
in surface waters, biofilms, and
sediments (freshwater to marine)23"25
Quantitative PCR and related
techniques; molecular and
genomic probes29"31
Pros: INT accepts electrons from dehydrogenase enzymes
and is reduced to a reddish colored formazan (INTF) - can
be quantified by simple colorimetric analysis after a very
short incubation time26; total cell numbers are quantified
under epifluorescence microscopy using a counter-stain (e.g.
acridine orange27) - proportion of population that is
metabolically active is then estimated; very sensitive method
even at low microbial biomass and low temperatures27, with
resolution at the level of individual cells28.
Cons: Can miss cells with very low respiration (e.g. cysts), or
cells that do not use INT as an electron acceptor28; still requires
microscopy (tedious, time-consuming).
Present status: Shows promise for use with various size classes
of microorganisms in ballast water.
Pros: Reliable quantification of targeted taxa from environ-
mental water samples if PCR inhibitors can be removed and
molecular material can be efficiently recovered29'30.
Cons: More research is needed to test the degree to which these
can reliably discern between viable and non-viable cells,
or infective and non-infective cells, or toxic and nontoxic
cells, unless supplemented by other techniques31"34.
Present Status: Available methods do not allow reliable
calibration with living algal cell numbers in natural samples;
quantitative methods are emerging35'36.
* References used: ' Graham et al. (2009);2'3'4'5 Wetzel (2001), Jeffrey et al. (1997), US EPA (1997), Sarmento and Descy (2008);
6 MERC (2009c); 7 Boyer et al. (2009);8'9 U.S. EPA (2003), Buchanan et al. (2005); 10"12 Bowles et al. (1985), Hendrey et al.
(1987), Porra (1991); 13 Schluter et al. (2006);
' Mackey et al. (1996), Jeffrey and Vest (1997), Schmid et al. (1998), Schluter
et al. (2006); 17 Sandrin et al. (2009); 18 Karl (1980); 19'20 Holm-Hansen (1973), Takano (1983);21'22 Inubushi et al. (1989), Rosaker
and Kleft (1990);23"25 Songster-Alpin and Klotz (1995), Posch et al. (1997), Blenkinsopp and Lock (1998);26 Mosher et al. (2003);
27 Sandrin et al. (2009);28 Posch et al. (1997); 29Caron et al. (2004); 30Kudela et al. (2010); 31Karlson et al. (2010); 32Guy et al.
(2003), 33Audemard et al. (2004), 34Burkholder et al. (2005), 35Jones et al (2008), 36Bott et al. (2010).
73
-------
extreme practical and logistical limitations and/or high risk in efforts to assess the effectiveness of
treatment systems in removing maximal densities of harmful organisms, or mixes of representative
organisms within certain density ranges. These risks support the use of sized-down mesocosm
treatments (hundreds to thousands of liters; Ruiz et al. 2006; MERC 2009a-c) that are larger and
therefore more realistic than bench-scale microcosms, but more manageable in volume than ballast
tanks.
Sized-down treatments help to reduce risks to human health safety and receiving aquatic ecosystems for
testing treatment system effectiveness at removing toxic substances and residues that are part of the
treatment process. As Ruiz et al. (2006, p. 10) stated:
Economy of small scale and ease of manipulating environmental variables and community assemblage at
the laboratory and intermediate scales make it possible and practical to estimate if a ballast water
treatment process and system is likely to be effective over the full range of physical [chemical,] and
biological conditions expected in the field;.. .the same regime on a ship would prove logistically and
financially very unwieldy. Thus, smaller scale tests demonstrate the treatment's performance and capacity
across a wide range of relevant state variables....
This approach also allows more precise, controlled sampling during test trials (MERC 2009b). At larger
scales, practical limitations restrict the number of conditions that can reasonably be tested, and testing is
directed more toward ensuring functionality of the engineered system rather than understanding the
treatment process under various conditions.
Small-scale (benchtop or laboratory) experiments minimize logistics and expense, and they can provide
proof of concept in assessing whether a given treatment meets expectations (Ruiz et al. 2006). For
example, if a BWMS is planned for use across the salinity gradient, then its efficacy should be tested
across all three salinity ranges (Table 6-2). Logistically, however, it may be feasible to test two salinity
ranges at full scale, but not the third. In such cases, small-scale and intermediate-scale (see below) tests
could be completed using the third salinity range. Likewise, the Panel recommends that bench-top and
mesocosm experiments complement full-scale testing.
Practical Limitations of Challenge Water Conditions
While it is important to evaluate BWMS under diverse and challenging biological, physical and
chemical conditions to understand system performance and the broad applicability and reliability of
BWMS, all biological and chemical challenge conditions may not be achievable during a series of tests
using natural waters. As described above, artificial manipulations of biological, physical and chemical
conditions may introduce significant artifacts. Therefore, without rigorous validation that a modification
to challenge water is representative of natural conditions and does not cause artifacts (e.g., stressing or
killing organisms), only natural ambient conditions should be used. The following options are available
to address the difficulty in meeting challenge conditions using un-augmented challenge water:
• Make the challenge conditions somewhat less stringent by allowing challenge water conditions
in some replicate trials during testing of a given BWMS to fall outside of target values. For
example, accept test results if all challenge water conditions for some replicate tests are within
70 percent of target values, as long as more than half the replicate trials are above all threshold
values;
74
-------
• Loosen the requirement that all the biological, physical and chemical challenge water conditions
must be met for each replicate test of a given BWMS, as long as the majority of threshold
conditions are met for an individual replicate test; or
• Allow certain manipulations or alterations to challenge water during testing of BWMS if
approved by the EPA or the Verification Organization. Acceptance of the conditions would be
based on either the test facility's test data or experimental data from other test facilities, showing
the manipulations do not affect the validity of the test.
6.2.4. Testing Shipboard Treatment Systems: Inherent Mismatch between Viability Standard
and Practical Protocols
The previous section reviewed features of current procedures for testing BWMS that could be improved
with existing knowledge and technology. In this section, the Panel reviews additional aspects of current
procedures that may not accomplish the stated goals because of inherent limitations in current
knowledge and technology. All of the six issues considered below stem from the difficulty—perhaps the
impossibility, given current technology—of accurately enumerating only those organisms that are viable
(living). Current practices result from trying to directly assess the legal standards (which focus on viable
organisms). This section is aimed especially at organisms ^50 jim, because the challenge of
determining viability of larger organisms may be secondary to the problem of sampling an adequate
volume to assess the concentration aspect of the legal standard (see section 3, Statistics and
Interpretation). The Panel recommends that new approaches be developed, including procedures that
address the standards indirectly, but have the benefit of practicality. In general, the Panel recommends
that the limitations of testing protocols for determining "viability" and/or "living" be assessed. Where
they are found to be lacking, the Panel recommends development of improved standardized protocols. If
indirect metrics can be reliably correlated with the concentration of viable organisms at land-based test
facilities, they should also be considered for adoption (as the Panel noted in Table 6-4 above).
As Lee et al. (2010, p. 72) aptly state, "A discharge standard of 'zero detectable organisms' may appear
very protective; however, the true degree of protection depends on the sampling protocol." Here, a
viable or living organism is defined as in U.S. EPA (1999), namely, as an organism that has the ability
to pass genetic material on to the next generation. The percentage of non-viable cells can vary markedly,
for example, from 5-60 percent among phytoplankton taxa, and in general, non-viable organisms are
believed to represent a substantial component of the total plankton (Agusti and Sanchez 2002). There are
several fundamental problems with present attempts to quantify viable organisms to evaluate ballast
water treatment efficiency, outlined as follows.
Death of Organisms by Rapid Concentration from Large Volumes
A major issue confounding the realistic representation of viable organism concentrations is that the rapid
concentration of organisms from large volumes into small volumes (which is a necessary prerequisite of
enumeration) causes the death of many organisms across size classes. This concentration step must be
accomplished quickly before organisms die, e.g., within six hours (or less) for zooplankton. Because of
the need to evaluate relatively large volumes of water in order to be confident about the concentrations
of sparse organisms in treated water, there is a fundamental disconnect in these requirements. It is
difficult if not impossible to rapidly concentrate microflora and microfauna from relatively large
volumes (hundreds of liters for zooplankton; liters for protists) by available filtration or centrifugation
75
-------
techniques without killing some of the organisms (e.g., Turner 1978; Cangelosi et al. 2007). Such rapid
concentration techniques can cause the loss of a major fraction of the viable organisms, even when
dealing with small sample volumes such as 1 liter (Darzynkiewicz et al. 1994). This problem affects
zooplankton and protist size classes, especially delicate species such as wall-less algal flagellates. Thus,
even if viable organisms can be distinguished from dead organisms when counted, what cannot be
known is the proportion of the dead organisms that were actually living at the time of sampling. These
problems illustrate the critical importance of a test facility validating all steps of any method it uses. It
should be noted that concentration-related losses do not affect the smallest size classes, bacteria and
viruses, because they are so abundant in most fresh, estuarine, and marine waters that it usually is not
necessary to concentrate them from whole water samples prior to analysis by standard microbial
techniques (U.S. EPA 2010; see below).
Organism Viability is Difficult to Determine
Organism viability is not easily detected by a single morphological, physiological, or genetic parameter,
making it advantageous to use more than one approach (Brussaard et al. 2001). In the context of meeting
a numeric standard, this approach is problematic because more than one 'answer' is generated.
Furthermore, the procedures used to determine viability are specific to some taxonomic groups (e.g.,
vital stains) and have varying degrees of uncertainty in categorizing live versus dead. Even procedures
recommended in the Protocol for land-based verification testing have practical limitations because of
time constraints. For example, the Protocol defines dead zooplankton operationally as individuals that
do not visibly move during an observation time of at least 10 seconds. Since live zooplankton may not
move over that short period, death is verified by gently touching the organism with the point of a fine
dissecting needle to elicit movement. However, the Protocol acknowledges that if every apparently dead
zooplankter in a concentrated subsample was probed and monitored for at least 10 seconds, the length of
time to complete analysis of the sample could be extended enough to increase the potential for sample
bias due to death of some proportion of individuals that had survived the sampling and concentration
procedures. In other words, the currently applied methods have serious limitations in some situations.
The results are therefore most appropriately viewed as an index of the number of viable organisms.
The Protocol is a living document, and as better methods are developed, they will be incorporated. The
Panel recommends consideration of a wider variety of indices that have the potential to be more rapidly
completed, if not more accurate. These include parameters that may be correlated to the abundance of
viable organisms, as discussed in the previous section (Table 6-4), and techniques to distinguish living
from dead individuals prior to enumeration by other methods (e.g., microscopy). The latter is elaborated
upon here.
Fluorescent stains have shown promise in detecting some live organisms or groups. For example, the
fluorogenic substrate Calcein-AM (Molecular Probes Inc.) is used to stain live cells that have metabolic
esterase activity (Kaneshiro et al. 1993; Porter et al. 1995). Once the colorless, nonfluorescent substrate
is inside the living cell, its lipophilic blocking groups are cleaved by nonspecific esterases to a charged
green fluorescen product that cannot pass across the plasma membrane. Dead cells cannot hydrolyze the
Calcein-AM or retain the fluorescent product. Use of FDA, sometimes in combination with CMFDA
(Table 6-3), is based on measuring intracellular esterases in live cells (Laabir and Gentien 1999; Hampel
et al. 2001). FDA was described as a reliable, efficient method to quantify concentrated viable
freshwater organisms in the >_10 to < 50 jim size class from ballast discharge (Reavie et al. 2010).
However, various algal species differ in their uptake of FDA and CMFDA, and other particles in a given
76
-------
sample can also fluoresce (Garvey et al. 2007; MERC 2009c). The vital stain propidium iodide (PI), in
combination with molecular probes, has been used to discern live from dead bacteria (Williams et al.
1998), but the number of false positives can vary widely (Steinberg et al. 2010), and this stain cannot be
used to assess algal viability because its emission spectrum overlaps that of chlorophyll (Veldhuis et al.
2001). Consistent with the ETV Protocol (U.S. EPA 2010), the Panel recommends completion of on-site
validation before selecting a viability method, including evaluation of false positives and false negatives.
As mentioned above, detection of infective viruses has received relatively little attention in ballast water
treatment. Waterborne illnesses can involve a wide array of viruses; for example, enteric viruses that can
be transmitted by water include poliovirus, coxsackievirus, echovirus, human caliciviruses such as
noroviruses and sapoviruses, rotaviruses, hepatitis A virus, and adenoviruses (Howard et al. 2006).
Considering pathogens of aquatic organisms, aquatic ecosystems are poorly understood with respect to
the diversity of viral pathogens of beneficial aquatic life (Suttle et al. 1991; Griffin et al. 2003; Munn
2006; Suttle 2007). Viruses also cross size classes; in environmental samples, many are in range of
nanometers, but some can be nearly 3 jim in maximum dimension (Bratbak et al. 1992); their tendency
to adsorb to sediment particles (> 6 jim) means they can be captured with larger particles during sample
preparation involving filtration (Bosch et al. 2005).
The U.S. EPA (2001) requires a 99.9 percent reduction in the total number of human enteric viruses in
water for human consumption. In practice, this requirement is met based on treatment alone, although
the EPA acknowledges that removal actually can only be accurately assessed by monitoring finished
waters over time. Ultra-filtration protocols have been developed for concentrating and enumerating
human enteric viruses (Fout et al. 1996; U.S. EPA 2001), but these techniques do not discern potentially
infectious from non-infectious viruses. Environmental water samples also have been evaluated for
human viral pathogens using standard techniques for in vitro cultivation, an approach that is affected by
the same problems confronted for detection of viable bacteria—the techniques are expensive, time-
consuming, labor-intensive, and can easily miss various groups of infectious viruses (Fout et al. 1996).
Rapid, sensitive molecular methods for viral nucleic acid detection have been recently developed but,
again, most cannot discern potential infectivity. The intercalating dye propidium monoazide (PMA) has
shown promise in detecting potentially infective coxsackievirus, poliovirus, echovirus, and Norwalk
virus (Parshionikar et al. 2010). In other promising research, Cromeans et al. (2005) included additional
processing steps such as specific capture by cell receptors for Coxsackie B viruses in vitro, followed by
molecular detection of viral nucleic acids in the captured viruses; or selection/detection of specific RNA
present in host cells only during virus replication. Real-time assays (30-90 minutes) were also developed
for enterovirus, hepatitis A virus, Adenovirus, and Norovirus detection. There remains a need for new,
commercially available technology that can discern infectious from non-infectious viruses (Cromeans et
al. 2005; Parshionikar et al. 2010) although current and proposed standards do not distinguish the two.
Special Challenges of Resistant or Nonculturable Stages in Attempts to Assess Viability
Resting stages (e.g., cysts) of some bacteria, phytoplankton, protists, zooplankton and metazoans are
particularly resistant to motility, staining, and any other tests. For example, the protist size class (> 10
|im to < 50 |im) includes many species (microalgae, heterotrophic protists, metazoans) that form
dormant cells or resting stages, or cysts (Matsuoka and Fukuyo 2000; Marrett and Zonneveld 2003).
Cysts from potentially toxic dinoflagellates are commonly found in ballast waters and sediments
(Hallegraeff and Bolch 1992; Dobbs and Rogerson 2005; Doblin and Dobbs 2006). These cysts have
been used as model indicator organisms to assess ballast water treatment efficiency (Anderson et al.
77
-------
2004; Stevens et al. 2004), based on the premise that treatments which can eliminate the cysts likely also
eliminate other, less-resistant organisms (Bolch and Hallegraeff 1993; Hallegraeff et al. 1997).
Because resistant cells often have a low metabolic state and thick, multi-layered walls that are
impermeable to many stains (Romano et al. 1996; Kokinos et al. 1998; Connelly et al. 2007), their
viability can be difficult to assess without culture analyses that may require weeks to months (Montresor
et al. 2003; U.S. EPA 2010). Improved methods have been developed for some algal groups (Binet and
Stauber 2006; Gregg and Hallegraeff 2007) but, overall, as the ETV Protocol (U.S. EPA 2010, pp.46-
47) states, "At present, no rapid, reliable method to determine cysts' viability is in widespread use, and
the FDA-CMFDA method has yielded variable results with dinoflagellates and cyst-like objects." The
ETV Protocol recommends use of this method as a "place holder" until more effective methods become
available.
The effectiveness of ballast water treatment in removing viable bacteria is evaluated by using multiple
bacterial media in combination with taxon-specific molecular techniques (MERC 2009c; U.S. EPA 2010
and references therein). Colonies are monitored and quantified after ~1 to 5 days, depending upon the
organism and its growth. These methods enable detection and quantification of viable, culturable cells.
However, it has been repeatedly demonstrated that bacterial consortia across aquatic ecosystems
commonly have a substantial proportion of cells that are active (viable) but nonculturable (Oliver 1993;
Barcina et al. 1997 and references therein). These cells obviously would be overlooked in culturing
techniques, a problem that would result in failure to detect viable cells of bacterial pathogens in treated
ballast water. Under some conditions, the nonculturable organisms can regain activity and virulence
(Barcina et al. 1997 and references therein).
Biased Counts Due to Live, Motile Species Changing their Location in Counting Chambers
At the other extreme from resting stages are living organisms that are difficult to enumerate because
they are highly mobile. Organisms are typically enumerated in counting chambers, based upon an
underlying premise that the cells do not change their location in the chamber. However, many protists
move rapidly by means of flagella or other structures. Because they do not maintain their position in a
counting chamber, as live cells they could be counted multiple times. Moreover, their sudden movement
can disrupt the locations of other cells in the chamber, mixing cells that may have been counted with
others that have not yet been counted. For these reasons, reliance on live counts can easily yield
unreliable data. This consideration underscores the need for vigorous validation of protocols used to
quantify viable organisms.
Indirect Metrics for Enumeration of Viable Cells Should be Investigated for Use in Standard
Protocols
Consideration of the above points—death during concentration of organisms, lack of reliable procedures
to assess viability (especially for resting stages of many taxa), movement of live organisms in counting
chambers that can result in serious quantification errors—leads the Panel to recommend that alternative
approaches, including enumeration of preserved organisms and indirect metrics of the concentration of
viable organisms, be tested. Should they be validated as superior to present protocols, then the argument
is strong to consider elevating these alternative approaches to standard protocols. These inherent
limitations add weight to the more practical considerations in section 6.2.3 above: the practical and
inherent limitations converge as an argument for the greater development, testing and implementation of
78
-------
indirect metrics of the concentration of viable organisms, including both STOs and surrogate
parameters, particularly in compliance testing. Adding parallel testing of indirect metrics to land-based
testing currently underway in test facilities from different geographic regions could rapidly yield
comparisons on which decisions for future testing could be made. Possibly, a combination of approaches
will prove to be the most advantageous in estimating the concentration of viable organisms of different
taxonomic groups.
6.3. Approaches to Compliance/Enforcement of Ballast Water Regulations and Potential
Application to Technology Testing
The EPA has extensive experience in effective compliance and enforcement of discharge regulations,
and has committed to work with the USCG to develop and implement compliance and enforcement
measures for ballast water regulations (2011 MOU between EPA and USCG6). However, given the
nature of ship ballast water discharge, new approaches likely will be needed. Both initial testing of
treatment systems (6.2.2 - 6.2.4) and methods currently available for potential compliance and
enforcement monitoring are complex, slow and expensive. Statistical (see section 3, this report) and
logistical limitations related to collection of appropriate sample volumes and detection/quantification of
live organisms in practice, mean that it may often be impossible to directly assess whether a vessel can
meet all the numerical standards for viable organisms (King and Tamburri 2010). No information was
provided to the Panel on whether protocols and systems for compliance monitoring (whether voluntary
by ship operator or legally required) and enforcement were being considered alongside the development
and testing of treatment systems. The Panel considers it essential that these be developed so enforcement
can commence as soon as a U.S. ballast water performance standard is finalized.
The practical and inherent limitations suffered by the full protocols for verification testing of BWMS
(6.2.2 - 6.2.4) have even greater force in the context of routine inspections (either self-inspections or
regulatory inspections) (King and Tamburri 2010). They are simply not practicable to use in the
compliance and enforcement context. Unless alternative protocols that are practical for inspections are
developed, neither self-compliance efforts nor regulatory enforcement will be possible once a system is
installed on a ship. For example, treatment system malfunctions are inevitable. If some types of
mechanical failure are not obvious to the operator or inspector, release of organisms may reach and
maintain non-compliant levels for long periods of time with no detection of the malfunction, no penalty,
and therefore no incentive to detect and fix the system. Unenforceable rules are bound to fail to meet the
goal of reducing invasions. Therefore, the Panel recommends that the EPA develop an approach for
BWMS that includes metrics appropriate for compliance monitoring and enforcement.
A potential solution is the use of a step-wise compliance reporting, inspections, and monitoring
approach, described below, which involves a series of steps that increase the likelihood of detecting non-
compliance but also increase in cost and logistic challenges (King and Tamburri 2010).
• Reporting. Vessel owner or ship master submits reports on the type of certified treatment
system onboard and documentation demonstrating appropriate use and maintenance.
6 February 2011 Memorandum of Understanding Between the U.S. Environmental Protection Agency, Office of
Enforcement and Compliance Assurance and the U.S. Coast Guard, Office of Marine Safety, Security and Stewardship for
Collaboration on Compliance Assistance, Compliance Monitoring, and Enforcement of Vessel General Permit Requirements
on Vessels (available at http://epa.gov/compliance/resources/agreements/cwa/mou-coastguard-vesselpermitrequirements.pdf)
79
-------
• Inspections. Enforcement official boards vessel and inspects the certified treatment systems
to verify use and appropriate operations and maintenance.
• Measures of system performance. Indirect or indicative water quality measures are collected
autonomously (using commercially available instruments), or by inspectors, that demonstrate
appropriate treatment conditions have been met.
• Indirect measures of non-compliance. Indirect metrics (e.g., Table 6-5) of abundances of live
organisms are collected autonomously, or by inspectors, for indications of clear non-
compliance.
• Measures of performance standard. Direct measures of concentration of live organisms in
the various regulated categories are made by specially trained technicians, with statistically
appropriate sampling and validated analyses and methodologies.
Protocols assessing indirect surrogate measures to quantify viable organisms should be further
developed for quick, easy, and defensible shipboard compliance monitoring (see 6.2.2 - 6.2.4).
6.4. Reception Facilities as an Alternative to Shipboard Treatment
Proposed federal regulations and the EVIO Ballast Water Management Convention allow for the transfer
of vessel ballast water to reception facilities, where the organisms in ballast water would be removed or
inactivated. Various studies have envisioned reception facilities as either built on land or installed on
port-based barges or ships. The discussion here refers to the use of on-land facilities, unless otherwise
specified.
Ballast water would be pumped off a vessel to a reception facility through main deck fittings and piping
or hoses similar to those currently used to transfer oil or other liquid cargo, oil-contaminated ballast
water, or fuel oil between vessels and shore. Vessels would need to be outfitted with appropriate pipes
and pumps to move ballast water to the deck and off the ship at a fast enough rate so the vessel is not
unduly delayed in port. The reception facility would store and treat the ballast water before discharging
it to local waters.
Vessel architecture and operations are principal impediments to the development of shipboard BWMS.
Challenging factors include vibration, small and busy crews, limited space and weight allowances,
limited power, potentially increased corrosion rates and sometimes short voyages. Reception facilities,
relieved of many or all of those constraints, show promise to achieve more stringent ballast water
treatment standards than shipboard BWMS.
The Panel did not reach consensus on certain issues and analyses related to treatment of ballast water at
reception facilities. Issues on which the Panel reached consensus are described in sections 6.4.1 and
6.4.2 (although two points of view are included for one of the issues). Panel conclusions on reception
facilities are presented in section 6.7.
6.4.1. Potential of Reception Facilities to Cost Effectively Meet Higher Standards
Though various studies, regulations and guidelines recognize the potential of reception facilities to treat
ballast discharges, the EPA and USCG reports on ballast water treatment have not addressed reception
facilities (EPA 2001; Albert et al. 2010; USCG 2008a,b). The literature on onshore treatment is
reviewed in Appendix B. Some studies conclude that reception facilities are a technically feasible option
80
-------
for either the industry as a whole or for some part of the industry (Pollutech 1992; NRC 1996; Oemke
1999; CAP A 2000; California SWRCB 2002; Brown & Cal dwell 2007, 2008). Other studies conclude
that cost or other factors could limit the use of reception facilities to part of the industry (Victoria ENRC
1997; Dames & Moore 1998, 1999; Rigby and Taylor 200la, b; California SLC 2009, 2010).
Four studies compared the effectiveness or costs of reception facilities and shipboard treatment.
Pollutech (1992) ranked reception facilities second in terms of effectiveness, feasibility, maintenance
and operations, environmental acceptability, cost, safety and monitoring out of 24 ballast water
management approaches for Great Lakes vessels; this study ranked shipboard filtration through a 50 jim
wedgewire strainer higher, and 17 other shipboard treatments lower, than treatment by a reception
facility7. AQIS (1993 a) found reception facilities (considering both on land and barge-mounted
facilities) to be less expensive than shipboard treatment in both single-port and nation-wide scenarios in
Australia, concluding that reception facilities "are more economic and effective than numerous ship-
board plants." Aquatic Sciences (1996) estimated the costs of using barge-mounted reception facilities in
the Great Lakes, and concluded that it is technically feasible, "more practical and enforceable" than
shipboard treatment, and offers "the best assurance of prevention of unwanted introductions." California
SWRCB (2002) found reception facilities to be the only approach to have acceptable performance in all
three categories of effectiveness, safety, and environmental acceptability in a qualitative comparison
with 10 shipboard treatments. Cost estimates compiled by the U.S. Coast Guard (e.g., USCG 2002)
showed reception facilities to be generally less expensive on a per metric ton basis than shipboard
treatment, although these estimates predate the establishment of discharge regulations and the most
recent generation of BWMS. In fact, most of the existing studies and estimates use outdated assumptions
or data, or are based on specific regions; therefore their conclusions may not apply to the current U.S.
situation, nor do they address international shipping issues.
The potential advantages of reception facilities over shipboard treatment systems include: fewer
reception facilities than shipboard systems would be needed; smaller total treatment capacity would be
needed; and reception facilities would be subject to fewer physical restrictions, and would therefore be
able to use more effective technologies and processes such as those commonly used in water treatment.
A shift from shipboard treatment to reception facilities is in some ways analogous to a shift from
household septic tanks to centralized wastewater treatment plants. These advantages are discussed in
greater detail in the following paragraphs.
Treatment Capacity
The EPA estimates that approximately 40,000 cargo vessels and 29,000 other vessels will be subject to
ballast water discharge requirements in the U.S. over the five-year VGP period (Albert and Everett
2010); approximately 7,000 ocean-going vessels called at U.S. ports in 2009 (MARAD 2011). Using
reception facilities would reduce the number of treatment plants and the total treatment capacity needed
for ballast water management. In shipboard treatment, a plant is installed on each vessel, and for nearly
all types of BWMS this must be large enough to treat the vessel's maximum ballast uptake or discharge
rate (Lloyd's Register 2010). The total treatment capacity needed is thus equal to the sum of the
maximum uptake or discharge rate of all ships. In contrast, reception facilities serve a number of vessels,
and since all vessels do not arrive and discharge ballast water simultaneously, the treatment capacity
needed would be less. Ballast water storage tanks at reception facilities would further lower the needed
treatment capacity, potentially to the average ballast water discharge rate. However, existing studies do
The remaining 5 management approaches involved neither shipboard treatment nor reception facilities.
81
-------
not appear sufficient to reliably estimate total treatment capacities required for individual BWMS versus
a national U.S. network of reception facilities. If undertaken, such studies would benefit from explicit
statements of assumptions about what drives treatment capacity needs and from comparisons of capacity
estimates derived from a range of assumptions.
Constraints on Treatment
Constraints on onboard treatment include limited space, power and treatment time, and ship stability
challenges (Pollutech 1992; AQIS 1993a; Aquatic Sciences 1996; NRC 1996; Cohen 1998; Oemke
1999; Reeves 1999; California SLC 2010; Albert and Everett 2010). These constraints are largely absent
in reception facilities,
Efficacy of Treatment Methods
Any treatment used on vessels could be used in reception facilities; alternatively, there are methods
available for reception facilities that cannot be used on vessels because of the space and other constraints
listed above. Such technologies include common water or wastewater treatment processes, such as
settling tanks and granular filtration, and less common processes including membrane filtration (AQIS
1993a; Gauthier and Steel 1996; NRC 1996; Victoria ENRC 1997; Reeves 1999; Cohen and Foster
2000; California SWRCB 2002; California SLC 2010).
The following information illustrates what could be achieved by using available water/wastewater
technologies in reception facilities; it describes what can be achieved in drinking water treatment
systems, recognizing that reception facilities would have to deal with a much greater taxonomic
diversity of organisms (from large zooplankton to microbes) and be able to effectively treat all possible
salinities (not just freshwater). For example, testing protocols for shipboard BWMS require at least a
four-log reduction in the > 50 um size class, and at least a two-log reduction in the > 10 to < 50 jim size
class when compared to ballast water uptake conditions. These metrics, however, do not account for
organism mortality in ballast water that occurs even with untreated ballast water. As this mortality varies
significantly (in some cases resulting in one-log reductions), it is difficult to quantify efficacy in terms
of log reductions. As such, Table 6-6 below compares the level of treatment that would be required for
two discharge standards relative to mean organism counts taken from vessels after a voyage. The Panel
recognizes that there may be valid ways of assessing efficacy other than basing it on mean
concentrations.
Table 6-6. Log reductions required by different discharge standards. Reductions are from mean values reported
by IMO (2003) for unexchanged and untreated ballast water sampled from vessels at the ends of voyages, for
zooplankton (n=429, collected with 55-80 um mesh nets, corresponding approximately to organisms in the >50 um
size class), phytoplankton (n=273, collected with <10 um mesh sieves or counted in unconcentrated samples,
corresponding approximately to organisms in the 10-50 um class), bacteria (n=l 1) and virus-like particles (n=7).
Organism/size class:
Discharge standard
IMO D-2 and USCG Phase 1
USCGPhase2
>50 um
perm3
2.7
5.7
> 10 - < 50 um
per ml
1.5
4.5
Bacteria
per ml
no reduction
4.9
Viruses
per ml
no reduction
4.9
82
-------
EPA requires that drinking water treatment systems be capable of at least 3-5 log reductions in Giardia8,
3-5.5 log reductions in Cryptosporidium9, and 4-6 log reductions in viruses, depending on the source
water (US EPA 1991, 2006). Several common drinking water filtration technologies are capable of 3-4
log reductions in protozoans and bacteria and 2-4 log reductions in viruses, and membrane filtration can
achieve >4-7 log reductions (U.S. EPA 1991, 1997b; NESC 2000a; LeChevallier and Au 2004; Wang et
al. 2006; WHO 2008). UV disinfection can achieve 2-3 log reductions in protozoans and 3-4 log
reductions in bacteria and viruses; biocides can achieve at least three-log reductions in Giardia, 3-6 log
reductions in bacteria, and 3-4 log reductions in viruses depending on dose and contact time (U.S. EPA
1997b; Sugita et al. 1992; NESC 2000b; LeChevallier and Au 2004). Filtration and disinfection are
generally considered additive processes: that is, a filtration process can produce a 3 log reduction, and a
disinfection process can produce a two-log reduction, and in sequential combination could potentially
produce a five-log reduction (U.S. EPA 1991).
Thus, even without a disinfection step, it appears that several common drinking water filtration
technologies available for reception facility use could achieve the 1.5-2.7 log reductions from mean
ballast water concentrations needed to meet the EVIO D-2 and USCG Phase 1 standards, although this
has not been tested with ballast water. It has not been demonstrated that these technologies could
address the extremely high numbers of organism found in the ballast water of vessels after some
voyages. Several combinations of filtration plus a single disinfection process appear to be able to
achieve the 4.5-4.9 log reductions needed to meet the USCG Phase 2 requirements for viruses, bacteria
and organisms in the > 10 to < 50 jim size class, and perhaps also the 5.7 log reduction needed to meet
the USCG Phase 2 standard for organisms > 50 jim Treating with one or more additional disinfection
processes could produce greater log reductions.10
Some membrane filtration technologies that could be used in reception facilities have produced results
of no detectable organisms in different organism classes. For example, the microfiltration unit used in
the conceptual design for a reception facility at the Port of Milwaukee (Brown & Caldwell 2008) would
likely result in no detectable organisms in both the > 50 jim and > 10 to < 50 jim size classes (based on
microfiltration results cited in U.S. EPA 1997b and LeChevallier and Au 2004). On the other hand,
ultrafiltration or nanofiltration might be needed to leave no detectable bacteria or viruses in the effluent,
although the time required to filter water to this level and its effect on vessel operations has not been
evaluated.
Plant Operation by Trained Water/Wastewater Treatment Personnel
Shipboard BWMS likely would be operated and maintained by regular crew members as added duties
(NRC 1996; California SLC 2010). Studies have noted that many of these crews are already
overburdened. Operation by trained, dedicated personnel in reception facilities would likely result in
more reliable performance (Cohen 1998; California SWRCB 2002; Brown & Cal dwell 2007; California
SLC 2010). Maintenance and repair work are more likely to be done reliably, and replacement parts
obtained more quickly, in reception facilities (AQIS 1993a; Aquatic Sciences 1996; Cohen 1998).
A protozoan pathogen with an active form measuring approximately 3x9x15 um and an ellipsoid cyst averaging 10-14
um long.
9 A protozoan pathogen with round cysts 4-6 um in diameter.
10 Sequential combinations of some disinfectants produce reductions even greater than the sum of the disinfectants'
reductions when examined separately (LeChevallier and Au 2004).
83
-------
Safety
Restricted working spaces and difficult or hazardous working conditions at sea (AQIS 1993 a; Cohen
1998; Cohen and Foster 2000) increase the risk of accidents with shipboard treatment. The storage and
use of biocides or other hazardous chemicals pose greater risks to personnel on vessels than in reception
facilities (AQIS 1993a; Carlton et al. 1995; Reeves 1998; Cohen 1998) and greater risk of accidental
discharge to the environment (Pollutech 1992; AQIS 1993a; Carlton et al. 1995). Because some physical
treatment processes cannot be used onboard, shipboard systems might rely on biocides more than would
reception facilities.
On the other hand, increased safety risk and a risk of spills or leaks of untreated ballast water may
accompany transfers of ballast water to reception facilities. Although liquid transfer is common practice
for tank ships, many other ships do not have crews experienced in these operations, and safety training
would be needed.
Reliability
Operation and maintenance by dedicated wastewater treatment staff should make the reliability of
reception facilities greater than that of shipboard BWMS. Extensive, long-term experience with water
and wastewater treatment technologies provides a basis for estimating the expected long-term
performance of these technologies if employed in reception facilities, while the brief and limited
experience with shipboard BWMS provides little basis for assessing whether they are likely to perform
adequately over a 20-to-30-year vessel lifetime. Since many BWMS treat ballast water on uptake
(Lloyd's Register 2010), which many vessels hold in dedicated ballast tanks where cysts or other resting
stages may be retained in sediments for long periods (Cohen 1998), failure to operate a BWMS or to
operate it effectively at any time could contaminate treated ballast water on later voyages (AQIS 1993a;
Reeves 1998). In addition, reception facilities would have more flexibility to build redundancy into the
system design than would shipboard systems.
Adaptability
Because of space restrictions on vessels and structural cost factors that make treatment components a
smaller part of the total cost of reception facilities, it is likely to be both physically and financially easier
to retrofit, replace or upgrade reception facilities than shipboard systems. Reception facilities "provide
treatment flexibility, allowing additional treatment processes to be added or modified as regulations and
treatment targets change" (Brown & Caldwell 2008).
Compliance Monitoring and Regulation
Although the requirements for demonstrating compliance with ballast water discharge regulations have
yet to be established, the effort and cost of monitoring and enforcement needed to meet a given standard
could be much less for a small number of reception facilities compared to a larger number of mobile,
transient, shipboard plants, most of which are foreign-owned or foreign-flagged, which are accessible
only when in U.S. ports, usually for brief periods (AQIS 1993a; Ogilvie 1995; Aquatic Sciences 1996;
Cohen 1998; Dames & Moore 1999; Oemke 1999; California SWRCB 2002; Brown & Caldwell 2007;
California SLC 2010). Some studies noted that only reception facilities put the responsibility for
monitoring, control and effectiveness entirely in the hands of the authorities responsible for protecting
84
-------
the receiving waters, without reliance on marine vessel logs or on authorities in originating ports (AQIS
1993b; Dames & Moore 1999; California SWRCB 2002).
6.4.2. Challenges to Widespread Adoption of Reception Facilities in the U.S.
Although reception facilities offer advantages as just discussed, the Panel recognizes that there are
challenges to their adoption. The Panel reached consensus on all but one of the challenges presented in
this section, and opposing views are presented for that one.
Ballast Discharge Before Arrival to Reduce Time Spent at Berth
Some vessels may discharge part of their ballast water before arriving at berth so they can complete
discharge by the time the cargo is loaded (AQIS 1993a; Oemke 1999; Cohen and Foster 2000; CAPA
2000; Rigby and Taylor 200la). Alternatively, a ship's ballast water system can be outfitted with pipes
and pumps that are large enough to allow the ship to unload ballast water as quickly as it loads cargo
(AQIS 1993b). Glosten (2002) and Brown & Caldwell (2007, 2008) identified technical solutions for
retrofitting a variety of vessels (but not all types) to allow them to deballast at berth during the time they
load cargo. However, this issue has not been studied with respect to costs or feasibility for handling as
yet uncertain estimates of the numbers of vessels expected to require treatment at different ports.
Ballast Discharge to Reduce Draft Before Arriving at Berth
Several studies noted that some vessels discharge ballast water before arriving at berth to reduce draft to
cross shallows (Cohen 1998; Dames & Moore 1998, 1999; Oemke 1999; CAPA 2000, Rigby and Taylor
200la; California SWRCB 2002; California SLC 2010). The frequency of these occurrences has not
been quantified (one authority the Panel consulted stated that they are rare whereas another indicated
that some Great Lakes operators may perform such discharges routinely). Possible solutions include
offloading ballast water to barges as is done for some liquid cargos (AQIS 1993a; Carlton et al. 1995;
Dames & Moore 1999; CAPA 2000; Rigby and Taylor 200la; Glosten 2002; California SWRCB 2002),
or importing cargo in shallower-draft ships. Dames & Moore (1998) suggested that a barge- or ship-
mounted reception facility could service deep-drafted arrivals that need to deballast during approach.
Some panel members point out that this issue has not been studied with respect to costs or feasibility for
handling as yet uncertain estimates of the numbers of vessels expected to require treatment at different
ports.
Ballast Discharge by Lightering Vessels
Large tankers that arrive on the U.S. coast carrying crude oil or other liquid cargo may transfer part of it
to lightering vessels (smaller tankers or barges) in designated anchorages or lightering zones. These
lightering vessels often discharge ballast as they load cargo. In many cases, the discharged ballast water
is from nearby sites (CDR Gary Croot, U.S. Coast Guard, pers. comm.; National Ballast Information
Clearinghouse data), and depending on how the regulations are written may not require treatment.n In
cases where the ballast water is from more distant sites, solutions might include offloading ballast water
11 EPA's current Vessel General Permit requires vessels on nearshore Pacific Coast voyages to conduct ballast water
exchanges only if they cross international boundaries or cross from one Captain of the Port Zone to another (VGP §2.2.3.6).
Similarly the U.S. Coast Guard's proposed discharge standards would not apply to vessels operating within a Captain of the
Port Zone (USCG2008c).
85
-------
to lightering vessels that have been ballasted with local water, or importing cargo in smaller tankers. The
frequency, volumes, and uptake and discharge locations associated with lightering have not been
quantified, so the significance of this issue with respect to invasions or feasibility of technical solutions
is not known.
Implementation Schedule
It typically takes up to 30 months to design, permit and construct a sewage treatment plant larger than 10
mgd, and potentially much longer if sites are scarce, or if there are issues related to permit approvals
(Robert Bastian, U.S. EPA Office of Water, pers. comm.). Most ballast water reception facilities needed
in the U.S. would be smaller. Vessel modifications are needed for either shipboard or reception facility
approaches, either to install a BWMS or to allow rapid discharge to a reception facility. This process is
almost exclusively undertaken while the vessel is out of service, which occurs infrequently; dry
dockings, by marine vessel classification society requirement, must be no less than once every five years
(ABS SVR 7/2/1-11). To accommodate vessel modifications, proposed standards include phase-in
periods (8 years for EVIO D-2, 9 years for USCG Phase 1). The critical path for both reception facility
and shipboard treatment is the vessel modification work, where the governing factor is the frequency
with which the vessel is taken out of service. This is the same for either approach.
A more comprehensive comparison of potential implementation schedules for both shipboard BWMS
and reception facilities is needed.
The Current Regulatory Framework
Challenges associated with the regulatory framework are included in this section even though the Panel
did not reach consensus on this issue, because many Panel members thought that this is the major
challenge to reception facilities; therefore leaving it out would result in an unbalanced portrayal of
advantages and challenges. Two views of the issue are presented here.
View 1: Although reception facilities are allowed in policy and rules and have identified
advantages relative to BWMS, there are no reception facilities currently available in the U.S. to
remove organisms from ballast water. At the same time, there are 10 internationally Type
Approved BWMS of which many have been sold. This appears to be a result of the framework of
the 2004 EVIO Convention that phases in performance standards by marine vessel ballast water
capacity and construction date of marine vessels rather than on a port-by-port basis. To avoid the
risk of arriving in a port without an operational reception facility, operators are opting to install
shipboard BWMS. The U.S. proposed Phase 1 timetable would require all new vessels
constructed starting in 2012 to meet performance standards upon delivery. To be in compliance
using only reception facilities, the marine vessel operator must be assured that there will be an
operational reception facility at all anticipated ports-of-call where ballast water discharge might
be expected for the lifetime of the vessel. On the other hand, vessels engaged solely in regional
trade may benefit from the reception facility approach if reception facilities are operational in the
region and will not need to invest in a shipboard BWMS.
• View 2: The alternative view holds that current federal regulations governing ballast discharges
under NISA and CWA are based on mid-ocean exchange (Albert et al. 2010), and thus favor
neither BWMS nor reception facility treatment. Various states have adopted discharge standards,
86
-------
of which some might be met by BWMS, while others would require reception facility treatment
because their requirements are more stringent. Regulatory agency dismissal of or opposition to
reception facilities, and encouragement of BWMS (including the sponsorship and funding of
research), has contributed to the focus on BWMS. Equipment manufacturers have invested in the
development of BWMS because they expect that discharge standards that can be met by BWMS
will be implemented and enforced, thereby creating a large enough market to allow them to
recoup their investments and turn a profit. Ports have not promoted the development of reception
facilities because they are not convinced that discharge standards requiring treatment in reception
facilities will be implemented and enforced effectively. If equipment manufacturers and ports
come to believe that standards will be implemented that will need to be met by treatment in
reception facilities—then the current focus on BWMS will shift. It is the decisions, actions and
communications of regulatory agencies that will mold these expectations about the future
direction and implementation of discharge standards.
Issues regarding treatment in reception facilities for which the Panel did not reach consensus
There were several additional issues regarding treatment in reception facilities for which the Panel did
not reach consensus. Included below is a brief summary of discussions that may be helpful to the ballast-
water community in the future.
• Need for further study. The Panel discussed the need for further study of BWMS treatment
options. Some Panelists, noting the scarcity of reliable test data, discussed the need for further
study of long-term performance of shipboard BWMS under the challenging conditions of actual
shipboard use. Some Panelists discussed the need for pilot studies of reception facilities to assess
their cost, operations, and safety issues in order to assess systemic challenges and to support
operational solutions for creation of networks of onshore reception facilities.
• Cost comparison. The Panel did not reach agreement on issues relating to estimating and
comparing the cost of treating ballast water in shipboard BWMS and the cost for treating in
onshore reception facilities. There were differing opinions on the assumptions needed to develop
screening estimates for either option. These included assumptions about capacity requirements,
applicability of existing cost data, extrapolation methods, inclusion of operational costs that
could be incurred by vessel owners if they were delayed due to unavailability of reception
facilities or from inoperable BWMS, and costs for some vessel owners that might be required to
install a shipboard BWMS as well as pay for use of a reception facility, depending on the port of
call.
• Implementation issues. The Panel discussed issues that could affect the time needed to
implement treatment of ballast water by individual shipboard BWMS or for developing a
network of reception facilities, but did not come to agreement on their implications for
implementation timelines. Some panel members said that reliance on shipboard BWMS would
require a potential lag of several years for large-scale production of BWMS and time needed to
develop effective monitoring and enforcement. Some members said that timelines for developing
reception facilities would need to consider implications of land availability adjacent to port
terminals, and time to acquire and permit newly designed treatment facilities and required
support services.
87
-------
International issues. The Panel briefly discussed international issues related to potential new
U.S. standards for ballast water discharge. These issues included the complexities of
implementing standards for vessels engaged in international maritime trade. The Panel also
briefly discussed whether setting U.S. standards based on the likely higher performance of
reception facilities would introduce potentially varying levels of protection against introduction
of invasive species among the U. S. and other countries.
6.5. Approaches Other than Ballast Water Treatment
Several approaches other than the treatment of ballast water could help to reduce the risk of biological
invasions from ballast water discharges, and contribute to the achievability of performance standards
and permit requirements. While these approaches are often recommended, including by EVIO, they are
not often required or incentivized in practice. These approaches include ballasting practices to reduce
the uptake of organisms, ballast water exchange to reduce the concentration of exotic organisms,
reductions in the volume of ballast water discharged in U.S. waters, and management of the rate, pattern
or location of ballast water discharge to reduce the risk of establishment. Although the charge questions
to the Panel focused on shipboard treatment, the Panel considered these other approaches because, when
used in combination with shipboard treatment, they appear to be capable of achieving a greater level of
risk reduction than shipboard treatment alone.
6.5.1. Managing Ballast Uptake
Several studies have recommended various ballasting practices—sometimes referred to as ballast micro-
management (Carlton et al. 1995; Oemke 1999; Dames & Moore 1998, 1999; Cohen and Foster 2000),
shipboard management measures (Gauthier and Steel 1996), or precautionary management measures
(Rigby and Taylor 2001a,b)—to reduce the number of organisms, or the number of harmful or
potentially harmful organisms (such as bloom-forming algae and human pathogens found in sewage),
that are taken up with ballast water It is suggested that this can be accomplished by managing the time,
place and depth of ballasting. Some of these measures have been included in laws, regulations or
guidelines, including EVIO guidelines and the USCG rules implementing the National Invasive Species
Act. Although some of these regulations or guidelines have been in effect for nearly 20 years, there
appear to be no data on levels of compliance and no studies of the effectiveness of any of these measures
in reducing the uptake of organisms.
While there may be reasons for skepticism regarding the effectiveness or feasibility of several of these
measures (AQIS 1993b; Cohen 1998; Dames & Moore 1998, 1999; Cohen and Foster 2000; Rigby and
Taylor 200 Ib), some could be helpful in meeting stringent standards if vessels had sufficient incentive to
implement them. The effectiveness of alternative ballasting (e.g., at locations low in harmful organisms)
and deballasting practices (e.g., locations and practices to reduce concentrating propagules) should be
quantified. As an example of the former, researchhas shown that taking up ballast water in areas affected
by toxic dinoflagellate blooms, followed by deballasting in another location, can result in distribution of
those blooms to previously unaffected areas (Hallegraeff and Bolch 1991). Clearly, such action should
be avoided as routine practice, and can also help to meet BWMS standards.
The value of such practices could be evaluated with models using currently available data on organism
distributions or by experimental approaches. To the extent these practices would reduce the uptake of
organisms, they could be used by vessels to help them meet any performance standards that might be
88
-------
adopted. From the perspective of overcoming technical limitations on the feasibility of meeting different
performance standards, such practices might allow the adoption of — and vessel compliance with —
more stringent standards than would otherwise be achievable. Thus, there are valid reasons for the EPA
to consider the potential for employing these practices in combination with ballast water treatment to
further reduce the risk of releasing exotic organisms in U.S. waters.
6.5.2. Mid-ocean Exchange
Mid-ocean ballast water exchange has the potential, in combination with the other approaches discussed
here, to further reduce the concentration of exotic organisms (though not necessarily reduce the
concentration of all organisms) in ballast discharges. There is general agreement that when properly
done, ballast water exchange can reduce the concentration of initially loaded organisms by about an
order of magnitude on average (Minton et al. 2005). It is not however, always possible, especially for
short coastal voyages. Additionally, conducting exchange represents an additional cost to the vessel.
6.5.3. Reducing or Eliminating Ballast Water Discharge Volumes
Invasion risk is positively related to the total number of propagules released in a given time and place.
Thus, risk is positively related to the concentration of propagules times the volume of the discharge.
Even if the concentration of propagules is unmanaged, reducing discharge volumes will reduce invasion
risk in ways that are predictable across taxa (Drake et al. 2005). Given this, various alternatives to the
use of "conventional" ballast water management systems have been proposed and studied since the 2004
EVIO Ballast Water Management Convention. These emerging alternatives to shipboard BWMS include
concepts and designs for "ballastless" or "ballast-free" ships, "ballast-through" or "flow-through" ships,
the use of "solid-ballast", and the use of "freshwater ballast". In fact, Regulation B-3, of the EVIO
Convention predicts and allows for the development and future use of such approaches to prevent the
transport of invasive species by ships. These approaches are summarized below.
Ballastless ship designs constitute a fundamental paradigm shift in surface vessel design. Rather than
increasing the weight of vessels by adding water to ballast tanks, these new designs use reduced
buoyancy to get the ship down to safe operating drafts in the no-cargo condition. For example, the
Variable Buoyancy Ship design (Parsons 1998; Kotinis et al. 2004; Parsons 2010) achieves this end by
having structural trunks of sufficient volume that extend most of the length of the ship below the "ballast
waterline" and then opening these trunks to the sea in the no-cargo condition. When the ship is at speed,
the natural pressure difference between the bow and the stern induces flow through the open trunks,
resulting in only local water (and associated organisms) within trunks at any point during a voyage.
While showing promise, and worthy of further considerations, ballastless ship designs appear feasible
only for new vessels being built in the future and may result in an overall increase in vessel biofouling
(another significant source of invasive species), if surfaces in open flow-through spaces are more
accessible and hospitable than traditional ballast tank surfaces (which are rarely fouled by higher
organisms). Similarly, a return to a historic approach of using solid ballast (commonly iron, cement,
gravel or sand) has been discussed recently but may not be feasible or cost effective for most vessels in
the modern merchant fleet.
Marine vessels that carry cargo in bulk, such as oil tankers or dry bulk carriers, cannot generally avoid
discharging ballast water in a cargo loading port. Part of the weight of the discharged bulk cargo,
typically 50 percent, must be replaced with ballast water to maintain stability. However, there are other
89
-------
vessel types, such as passenger ships and container ships, which do not experience the same bulk shift in
cargo that demands immediate ballast water replacement. These vessels provide opportunities for
innovative designs and operational practices that can significantly reduce or even eliminate ballast water
discharges in port.
Innovations to reduce ballast have also occurred in other types of vessels. Some vessels only require
ballast water to replace fuel oil consumption. A recent research vessel design was able to use the
processed effluent from the marine sanitation devices as ballast water. The mass balance between the
crew's gray and black water waste was similar to the amount of ballast water required to account for
consumed fuel oil. This approach eliminated traditional sea water ballast from the vessel design. The use
of freshwater as ballast has also been proposed, either the onboard production of potable water as ballast
for smaller vessels to replace fuel consumption or the transportation of freshwater from one port to
another that might have limited supplies of drinking or agricultural water (e.g., Suban et al. 2010). Using
a similar principle of only local water being onboard a vessel at any one time, other sorts of flow-
through ballast systems have also been proposed. These approaches would likely require modifications
to the existing ballast systems to actively and continuously pump water in and out of the ballast tanks
throughout voyages, resulting in complete tank turnover in an hour or two.
Container ships can sometimes balance operations between loaded cargo and discharged cargo. Even
when not balanced, the weight differential may often be within the margins of the vessel trim and
stability requirements. One company has built and is operating two trailer-ships, similar to container
ships that used design trades to eliminate the use of seawater ballast in all cases except emergencies. The
ships are a bit wider, and potentially burn some additional fuel to account for their increased size.
However, they have eliminated ballast water movements, as well as maintenance efforts associated with
salt water piping systems and ballast tanks. Trim corrections are accounted for by shifting ballast water
between tanks.
Given increased scrutiny and demands for ballast water exchange, it appears that many operators have
been able to reduce or eliminate their discharges through careful operational practices, e.g., members of
the Pacific Merchant Shipping Agency (PMSA) "all practice ballast water management protocols to
reduce or eliminate the risk of introduction of aquatic invasive species in state waters . . . Over 80
percent of vessels hold all ballast water in port to eliminate this risk. Those vessels that must discharge
ballast ensure that it is exchanged with mid-ocean water prior to entering coastal waters, dramatically
reducing the risk of carrying invasive species" (Pacific Merchant Shipping Agency, National
Environmental Coalition on Invasive Species, NECIS) . Similarly, an industry led initiative, Marine
Vessel Environmental Performance (MVeP), provides a numerical score to rate the environmental
soundness of ballast water management. This score accounts for both the volume and the concentration
of the ballast water discharged.
While many of these alternatives are conceptual at this point and may be limited to only specific vessels
and/or routes, future ballast water management approaches to minimize the risk of invasive species may
involve a variety of options and combination approaches. Regulatory frameworks for ballast water
management that address both the volume and concentration of organisms in ballast discharges could
further facilitate these alternative management approaches.
90
-------
6.5.4. Temporal and Spatial Patterns
Independent of practices of ballast water uptake and total volume of a given discharge (previous
sections), operational adjustments that modify the temporal and spatial patterns of ballast water
discharge also may reduce the probability that discharged propagules will found a self-sustaining
population (Drake et al. 2005). At least for sexually reproducing populations of planktonic species, for a
given concentration of a given species in ballast discharge, the greater the volume discharged in a given
time at a given location, the greater the probability of population establishment. If a total discharge
volume for a given port of call can be broken up in space or time, invasion risk will be lowered. Thus, if
a given discharge volume can be spread over space (e.g., as a vessel approaches harbor), be
discontinuous in time (with scheduled breaks in discharge), or be discharged in a mixing environment
(to dilute the concentration of propagules), the risk of invasion will be lowered (Drake et al. 2005).
For the same reasons, infrastructure modifications within ports that increase the rate and/or magnitude of
dilution of discharged propagules also would decrease the risk of population establishment by
discharged propagules. If discharges could be made in or piped to locations of greatest mixing within the
harbor (e.g., closer to the tidal channels instead of in partially enclosed ship slips), then the rate of
diffusion would be more likely to overcome the rate of reproduction. For example, low-velocity low-
energy propellers, oloid mixers, or other mixing methods are routinely used in sewage treatment plants,
industrial applications, and lakes. Such devices could be used in ports to increase the severity of Allee
effects and other population hurdles faced by newly discharged propagules to minimize the probability
of population establishment.
6.5.5. Combined Approaches
It may be possible to meet more stringent performance standards, or otherwise reduce the risk of
invasions from ballast water discharges, by combining the approaches discussed in previous sections
with either shipboard or onshore treatment. For example, a study by Fisheries and Oceans Canada
suggests that conducting a mid-ocean exchange combined with BWMS for Great Lakes bound carriers
may result in at least a lOx reduction in density of high risk taxa (Examining a combination treatment
strategy: ballast water exchange PLUS treatment, Sarah Bailey, Fisheries and Oceans Canada). After
considering the best science and technology now available, the state of Wisconsin is proposing to
continue requiring ships to flush their ballast tanks at sea and require oceangoing ships to use BWMS to
reduce remaining organisms to a level that meets the international numerical standard. This approach of
combining ballast water exchange with shipboard ballast water treatment is targeting an enhanced level
of protection for freshwater environments, similar to what has been proposed by Canada.
Each step from ballasting to deballasting, including the choice of procedures and the choice of
technologies, contributes to the probability of an invasion occurring (see below). Recognizing and better
quantifying the probability associated with each step could better target management efforts and achieve
reductions in the overall probability of invasion at lower cost than relying only on BWMS.
91
-------
6.6. Risk Management Approaches to Reduce Invasion Risk: Hazard Analysis and Critical
Control Points (HAACP)
The Panel provides the following analysis as an example of one potential risk management approach
that could be applied to ballast water management.
What is HACCP?
Risk assessment for decision-making can be implemented using the Hazard Analysis and Critical
Control Points (HACCP) approach. HACCP was developed in the late 1950s to assure adequate food
quality for the nascent NASA program, further developed by the Pillsbury Corporation, and ultimately
codified by the National Advisory Committee on Microbiological Criteria for Foods in 1997. The
framework consists of a seven-step sequence:
1) Conduct a hazard analysis.
2) Determine the critical control points (CCPs).
3) Establish critical limit(s).
4) Establish a system to monitor control of the CCPs.
5) Establish the corrective action to be taken when monitoring indicates that a particular CCP is
not under control.
6) Establish procedures for verification to confirm that the HACCP system is working
effectively.
7) Establish documentation concerning all procedures and records appropriate to these
principles and their application.
In international trade, these principles are important parts of the international food safety protection
system. The development of HACCP ended reliance on the use of testing of the final product as the key
determinant of quality, and instead emphasized the importance of understanding and control of each step
in a processing system (Sperber and Stier 2009). HAACP principles also appear applicable to
operationalize risk management for ballast water.
Basic Definitions
Hazard: The hazard under HACCP is the constituent whose risk one is attempting to control.
Critical control point: A critical control point (defined in the food sector) is "any point in the
chain of food production from raw materials to finished product where the loss of control could
result in unacceptable food safety risk"(Unnevehr and Jensen 1996).
Performance criteria: An important task in the HACCP process is to set performance criteria
(critical limits) at each of the critical control points (CCP). The minimum performance criteria
for each of the CCPs is set based on the final desired quality These criteria are determined using
experimentation, computational models or a combination of such methods (Notermans et al.
1994). Then, readily measurable characteristics for each process needed to assure the desired
quality are established and coupled to the control points.
92
-------
Application in Food and Water
HACCP has been applied in the food safety area for 50 years, and in the past decade guidelines and
regulations in the U.S. have been written that require an approved HACCP process in a number of
applications. For example, the U.S. Food and Drug Administration has developed a HACCP process
applicable to the fish and shellfish industries (21 CFR 123). HACCP has also been widely adopted in the
EU, Canada and a number of other developed and developing nations for food safety (Ropkins and Beck
2000).
Havelaar (1994) was one of the first to note that the drinking water supply/treatment and distribution
chain has a formal analogy to the food supply/processing/transport/sale chain, and therefore that
HACCP would be applicable. However, in effect, the development of the U.S. surface water treatment
rule under the Safe Drinking Water Act (40 CFR 141-142) and subsequent amendments incorporate a
HACCP-like process. Under this framework, an implicitly acceptable level of viruses and protozoa in
treated water was defined. Based on this, specific processes operated under certain conditions (e.g., filter
effluent turbidity for granular filters) were "credited" with certain removal efficiencies, and a sufficient
number of removal credits needed to be in place depending on an initial program of monitoring of the
microbial quality of the supply itself. This approach (of a regulation by treatment technique) is chosen
when it is not "economically or technically feasible to set an MCL [maximum contaminant level]" (Safe
Drinking Water Act section 1412(b)(7)(A)).
How HACCP Might be Applied to Ballast Water Management
Shipboard BWMS and onshore treatment of ballast water differ in a number of characteristics that
would affect their respective HACCP processes. Implementation of a HACCP program would need to
account for different regulatory agencies and their scope of enforcement, the training of personnel, and
the operational factors of each type of treatment. Figure 6-2 illustrates the control points for managing
ballast water to reduce invasion risk; these are elaborated in the following examples of steps in applying
the HAACP process:
• Identify the critical control points (which might include each particular treatment process as
well as the method and type of intake water used)
• Determine the needed total reduction of organisms needed for the totality of the treatment
system given the nature of the intake water (to achieve D-2, lOx D-2, etc.), and allocate these
reductions amongst individual treatment processes.
• Given criteria in the discharged treated ballast water (D-2, lOx D-2, etc.), determine the
minimum performance criteria for each treatment process, as well as criteria that determine
whether or not particular intake water might be suitable. Note that these performance criteria
should be based on easily measurable parameters that can be used for operational control.
Research may be needed to determine relationships for each process between such surrogate
parameters and removal of each of the size classes of organisms.
• A given ship having a set of processes with designated removal credits would only be
allowed to take in ballast water that does not exceed the capacity of the controlled process
train to meet the discharge criteria under the controlled operation.
• A QA process would be established for periodic validation and auditing (possibly by a third-
party organization). Operational procedures would need to be developed to indicate the
corrective actions needed for a particular process in the event that surrogate parameters fall
93
-------
outside acceptable limits, e.g., for additional holding time, recirculating for additional
treatment, or some other measure.
A blind testing procedure for the treatment products could be added to ensure that testing
laboratory is not biased.
Control points could also be identified for the various steps associated with transfer of an
invasive species to a new habitat.
Control Points for the Management of Invasives
Treatment of
Ballast Water
Shipboard Receiving
Treatment ~ Discharges - habitat Invasion
Port of Open Ocean Transfer to
Departure exchange treatment
System
\
On-shore _ Eff|uent Receiving |nvgsion
Treatment habitat
Figure 6-2. Some control points for the control of invasive species. Each of the processes may have imbedded control
points.
The overall context for applying HACCP methods varies depending upon the treatment envisioned. One
criterion for deciding which option is more straightforward to implement may be the ease with which a
risk management scheme can be applied and control ensured. In the consideration of shipboard and
reception facilities, the Panel notes there are uncertainties in both approaches. The Panel has
recommended that such uncertainties for ballast water treatment be assessed using a risk management
framework such as HACCP. For this report, none of the BWMS or alternative ballast water approaches
have been evaluated with respect to the risk of species invasions nor have critical control points been
identified within the full sequence of ballast water management activities. Given the similarities among
onshore treatment in reception facilities and water treatment facilities, some preliminary and partial
extrapolations may be possible. However, the lack of information precludes further analysis of this issue
by the Panel.
Figure 6-2 can be used as a heurist for identifying potential control points. For example, the
characteristics of the port of origin could be included in the consideration of the types of propagules
likely to be included in the ballast water. Known hazards from particular ports could be identified and
the protocol for the control process modified for those ports. Open-ocean (or water) exchange is a way
to reduce the number of propagules from the original port. Sea conditions or other factors may preclude
an exchange. The control process may require modification to allow for this contingency. Next, there is
transfer from the ballast tanks to the treatment system, which could be shipboard or on-shore. For either
treatment, multiple control points could be identified. The role of sea chests, filter systems, oxidizing
systems and plumbing could be identified. The HACCP approach would also take into account that
onboard and on-shore treatment facilities will differ in the number and location of discharge points.
94
-------
Likely outside of an engineering-based HACCP, but part of an overall strategy, is the consideration of
the receiving waters for the ballast water and the types of habitat. Receiving habitats that are similar to
those of the original port are likely to provide more opportunity for the establishment of an invasive
species or pathogen. This information may be useful in establishing a site-specific treatment
recommendation. These habitats could also be monitored as part of an overall plan for reducing the
likelihood of successful invasion.
HACCP can also be used to set priorities for the implementation of alternative means for managing
ballast water. Most current BWMS are built with a one-size-fits-all approach and designed to be adopted
by thousands of ships at some future time. There are defensible reasons for this one-size-fits-all
approach, but as considered above, additional reasons exist to consider more flexible and combination
approaches. This is especially true in the face of tight budgets and the constant need to prioritize
spending on the most cost-effective strategies to reduce invasion risk. Setting priorities using HACCP
principles could provide a basis for guiding the deployment of combinations of technologies and
practices now and in the future (Keller et al. 2010). For example, to minimize invasion risk most cost-
effectively while BWMS are being phased in, the highest risk ships that conduct the highest risk
voyages could be retrofitted first. Likewise, ship-voyage specific risk assessments could guide the
schedules for compliance monitoring of the operation and condition of installed water treatment
systems.
6.7. Summary and Recommendations
6.7.1. Principal Limitations of Available Data and Protocols
• Data are not sufficiently compatible to compare rigorously across ballast water treatment
systems because accepted standard protocols for testing ballast water treatment systems have
been lacking, although they have been under development at multiple testing sites. The EPA
ETV Protocol (U.S. EPA 2010) will improve this situation.
• No international requirement exists to report failures in type approval testing. On the basis of
typically reported results, therefore, it is impossible to draw reliable conclusions about the
consistency or reliability of some BWMS.
• The important size class of protists < 10 |im previously has been ignored in developing
guidelines and standards.
• Clear definitions and direct methods to enumerate viable organisms in the specified size
classes at low concentrations are missing for some size classes and indicator organisms, and
logistically problematic for all size classes, especially nonculturable bacteria, viruses, and
resting stages of many other taxa.
6.7.2. Alternatives to Shipboard Treatment of Ballast Water
• Data on the effectiveness of practices and technologies other than shipboard ballast water
treatment systems are inadequate because insufficient attention has been given to integrated
sets of practices and technologies, including: (1) managing ballast uptake to reduce presence
of invasives, (2) reducing invasion risk from ballast discharge through operational
adjustments and changes in ship design to reduce or eliminate the need for ballast water, (3)
95
-------
development of voyage-based risk assessments and/or HACCP principles, and (4) options for
treatment in reception facilities.
• Use of reception facilities for the treatment of ballast water appears to be technically feasible
(given generations of successful water treatment and sewage treatment technologies), and is
likely to be more reliable and more readily adaptable than shipboard treatment. Existing
regional economic studies suggest that treating ballast water in reception facilities would be
at least as economically feasible as shipboard treatment. However, these studies consider
only vessels calling at those regional facilities; if vessels also call at ports outside the region
without reception facilities, they would need a shipboard BWMS12. The effort and cost of
monitoring and enforcement needed to achieve a given level of compliance is likely to be
less for a smaller number of reception facilities compared to a larger number of BWMS.
6.7.3. Recommendations to Overcome Present Limitations
• Testing of BWMS in a research and development mode should be distinct from testing for
certification, and certification testing should be conducted by a party independent from the
manufacturer with appropriate established credentials, approved by EPA/USCG.
• Reported results from type approval testing of BWMS should include failures as well as
successes during testing (as per the Protocol) so that the reliability of systems can be judged.
This would be aided by the adoption of a transparent international standard format for reporting,
including specification of QA/QC protocols.
• Consideration should be given to including protist-sized organisms < 10 jim in dimension in
ballast water standards, and therefore in protocols to assess the performance of ballast water
treatment systems.
• Consideration should be given to expanding test protocols recommended by the ETV to
include components highlighted in Table 6-4.
• Suitable standard test organisms should be identified for bench-scale testing, and surrogate
parameters should be investigated to complement or replace metrics that are logistically
difficult or infeasible for estimating directly the concentration of living organisms.
• Use of representative "indicator" taxa (toxic strains of Vibrio cholerae; Escherichia coli;
intestinal Enterococci) should continue as a sound approach to assess BWMS for effective
removal of harmful bacteria. These estimates will be improved when reliable techniques
become available to account for active nonculturable cells as well as culturable cells.
• U.S. EPA is urged to develop metrics and methods appropriate for compliance monitoring
and enforcement as soon as possible.
12 Dr. Cohen, who read these studies, objected to this sentence as being untrue and misleading and felt that there had not been
adequate opportunity for Panel discussion of the issue.
96
-------
• Combinations of practices and technologies should be considered as potentially more
effective approaches than reliance on one ballast water treatment technology. For example,
ship-specific risk assessments (based on the environment and organisms present in previous
ports of call) could be used to help prioritize the use of risk management practices and
technologies, as well to target compliance and enforcement efforts.
• EPA should conduct a comprehensive analysis comparing biological effectiveness, cost,
logistics, operations, and safety associated with both shipboard BWMS and reception
facilities. If the analysis indicates that treatment at reception facilities is both economically
and logistically feasible and is more effective than shipboard treatment systems, then it
should be used as the basis for assessing the ability of available technologies to remove, kill,
or inactivate living organisms to meet a given discharge standard. In other words, use of
reception facilities may enable ballast water discharges to meet a stricter standard.
• Risk management is critical to ensure the efficacy of the entire spectrum of ballast water
management; that is, not just the specific treatment process but also management practices,
logistics and testing. Hazard Analysis and Critical Control Points (HACCP) has been
demonstrated to be an effective risk management tool in a variety of situations and could be
applied to ballast water management. HACCP methods are well understood and flexible.
HACCP can be used to set priorities for ballast water management and can be applied to
shipboard or shore-based systems or alternative management measures.
97
-------
REFERENCES
Agusti, S. and M.C. Sanchez. 2002. Cell viability in natural phytoplankton communities quantified by a
membrane permeability probe. Limnology and Oceanography 47: 818-828.
Albert, R. and R. Everett. 2010. Capabilities of shipboard ballast water treatment technology.
Presentation to EPA's Science Advisory Board, July 29, 2010, Washington, DC.
Albert, R., R. Everett, J. Lishman, and D. Smith. 2010. Availability and Efficacy of Ballast Water
Treatment Technology: Background and Issue Paper. U.S. EPA and US Coast Guard, Washington, DC.
Available on the SAB website, at
http://yosemite.epa.gov/sab/sabproduct.nsf/02ad90bl36fc21ef85256eba00436459/9E6C799DF254393A
8525762C004E60FF/$File/OW_Paper_Ballast_water_technology_issues_and_background_June_2010.
pdf
American Bureau of Shipping (ABS). 2010. Ballast Water Treatment Advisory. American Bureau of
Shipping, Houston, TX.
American Public Health Association (APHA), American Water Works Association and Water
Environment Federation. 2008. Standard Methods for the Examination of Water and Wastewater, 21st
edition. APHA, Washington, DC.
Anderson, A.B., B.O. Johannessen, and E. Dragsund. 2004. Ballast water treatment verification protocol
- DNV. In: Proceedings of the 2nd International Ballast Water Treatment R&D Symposium, by J.T.
Matheickal and S. Raaymakers (eds.). EVIO, London, UK.
Anderson, A., J. Cordell, F. Dobbs, R. Herwig, and A. Rogerson. 2008. Screening of Surrogate Species
for Ballast Water Treatment. Final Report to NSF International, Ann Arbor, MI.
Aquatic Sciences. 1996. Examination of Aquatic Nuisance Species Introductions to the Great Lakes
through Commercial Shipping Ballast Water and Assessment of Control Options Phase II. Final Report.
A Report for the Canadian Coast Guard. Aquatic Sciences Inc., St. Catharines, Ontario, Canada, 60 pp.
Audemard, C., K.C. Rease and E.M. Burreson. 2004. Real-time PCR for detection and quantification of
the protistan parasite Perkinsus marinus in environmental waters. Applied and Environmental
Microbiology 70:6611-6618.
Australian Quarantine and Inspection Service (AQIS). 1993 a. Ballast Water Treatment for the Removal
of Marine Organisms. Prepared by Gutheridge, Haskins and Davey Pty Ltd. Ballast Water Research
Series Report No. 1, AQIS, Canberra, Australia, 106 pp.
Australian Quarantine and Inspection Service (AQIS). 1993b. Ballast Water Management
Study. Prepared by Thompson Clarke Shipping Pty Ltd. Ballast Water Research Series Report No. 4,
AQIS, Canberra, Australia, 254 pp.
Avnimelech, Y., B.W. Troeger and L.W. Reed. 1982. Mutual flocculation of algae and clay evidence
and implications. Science 216:63-65.
R-l
-------
Baldino, R. 1995. Outline of Standard Protocols for Particulate Organic Carbon (POC) Analyses.
University of Wisconsin - Madison, Madison, WI.
Barcina, I, P. Lebaron, and J. Vives-Rego. 1997. Survival of allochthonous bacteria in aquatic systems:
a biological approach. FEMSMicrobiology Ecology 23:1-9.
Behrens, Y.S. and C. Hunt. 2000. The arrival and spread of the European green crab, Carcinus maenas,
in the Pacific Northwest. Dreissena: The Digest of National Aquatic Nuisance Species Clearinghouse
11:1-7.
Binet, M.T. and J.L. Stauber. 2006. Rapid flow cytometric method for the assessment of toxic
dinoflagellate cyst viability. Marine Environmental Research 62:247-260.
Blenkinsopp, S.A. and M.A. Lock. 1990. The measurement of electron transport system activity in river
biofilms. Water Research 24:441-445.
Bolch, CJ. and G.M. Hallegraeff. 1993. Chemical and physical treatment options to kill toxic
dinoflagellate cysts in ships' ballast water. Journal of Marine and Environmental Engineering 1:23-29.
Bosch, A., F.X. Abad and R.M. Pinto. 2005. Human pathogenic viruses in the marine environment,
pp. 109-131. In: Oceans and Health: Pathogens in the Marine Environment, by S. Belkin and R.R.
Colwell (eds.). Springer, New York, NY.
Bott, N.J., K.M. Ophel-Keller, M.T. Sierp, Herdina, K.P. Rowling, A.C. McKay, M.G.K. Loo, I.E.
Tanner, and M.R. Deveney. 2010. Toward routine, DNA-based detection methods for marine pests.
Biotechnology Advances 28:706-714.
Bowles, N.D., H.W. Pearl, and J. Tucker. 1985. Effective solvents and extraction periods employed in
phytoplankton carotenoid and chlorophyll determination. Canadian Journal of Fisheries and Aquatic
Science 42:1127-1131.
Boyer, G.L. 2007. The occurrence of cyanobactedal toxins in New York lakes: Lessons from the
MERHAB-Lower Great Lakes Program. Lake and Reservoir Management 23:153-60
Boyer, J.N., C.R. Kelbe, P.B. Ortner, and D.T. Rudnick. 2009. Phytoplankton bloom status: Chlorophyll
a biomass as an indicator of water quality condition in the southern estuaries of Florida, USA.
Ecological Indicators 9s: s56-s67.
Bratbak, G., O.K. Haslund, M. Heldal, A. Naess and T. Roeggen. 1992. Giant marine viruses? Marine
Ecology Progress Series 85:201-202.
Brown & Caldwell. 2007. Port of Milwaukee Onshore Ballast Water Treatment - Feasibility Study
Report. Prepared for the Wisconsin Department of Natural Resources. Brown & Caldwell, Milwaukee,
WI, 114pp.
Brown & Caldwell, and Bay Engineering. Inc. 2008. Port of Milwaukee Onshore Ballast Water
Treatment - Feasibility Study Report. Phase 2. Prepared for the Wisconsin Department of Natural
Resources. Brown & Caldwell, Milwaukee, WI, 98 pp.
R-2
-------
Brussaard, C.P.D., D. Marie, R. Thyrhaug, and G. Bratbak. 2001. Flow cytometric analysis of
phytoplankton viability following viral infection. AquaticMicrobialEcology 26:157-166.
Buchanan, C., R. Lacouture, H.G. Marshall, M. Olson, and J. Johnson. 2005. Phytoplankton reference
communities for Chesapeake Bay and its tidal tributaries. Estuaries 28:138-159.
Burkholder, J.M. 1992. Phytoplankton and episodic suspended sediment loading: Phosphate partitioning
and mechanisms for survival. Limnology and Oceanography 37:974-988.
Burkholder, J.M. 1998. Implications of harmful marine microalgae and heterotrophic dinoflagellates in
management of sustainable marine fisheries. Ecological Applications 8: S37-S62.
Burkholder, J.M. 2009. Harmful algal blooms, pp. 264-285. In: Encyclopedia of Inland Waters, Volume
7, by G.E. Likens (ed.). Elsevier, Oxford, UK.
Burkholder, J.M. and P.M. Gilbert. 2006. Intraspecific variability: An important consideration in
forming generalizations about toxi genie algal species. South African Journal of Marine Science 28:177-
180.
Burkholder, J.M., A.S. Gordon, P.O. Moeller, J.M. Law, K.J. Coyne, A.J. Lewitus, J.S. Ramsdell H.G.
Marshall, N.J. Deamer, S.C. Gary, J.W. Kempton, S.L. Morton, and P.A. Rublee. 2005. Demonstration
of toxi city to fish and to mammalian cells by Pfiesteria species: Comparison of assay methods and
multiple strains. Proceedings of the National Academy of Sciences (U.S.A) 102: 3471-3476.
Burkholder, J.M., G.M. Hallegraeff, G. Melia, A. Cohen, H.A. Bowers, D.W. Oldach, M.W. Parrow,
M.J. Sullivan, P.V. Zimba, E.H. Allen, and M. A. Mallin. 2007. Phytoplankton and bacterial assemblages
in ballast water of U.S. military ships as a function of port of origin, voyage time and ocean exchange
practices. Harmful Algae 6: 486-518.
California Association of Port Authorities (CAP A). 2000. feasibility of Onshore Ballast Water
Treatment at California Ports. Prepared by URS Corporation/Dames & Moore. CAP A, Sacramento,
CA.
California State Lands Commission (SLC). 2009. Assessment of the Efficacy, Availability and
Environmental Impacts of Ballast Water Treatment Systems for Use in California Waters. California
SLC, Sacramento, CA, 180 pp.
California State Lands Commission (SLC). 2010. 2010 Assessment of the Efficacy, Availability and
Environmental Impacts of Ballast Water Treatment Systems for Use in California Waters. CSLC,
Sacramento, CA, 141 pp.
California State Water Resources Control Board (SWRCB). 2002. Evaluation of Ballast Water
Treatment Technology for Control of Nonindigenous Aquatic Organisms. SWRCB, California
Environmental Protection Agency, Sacramento, CA, 70 pp.
Cangelosi, A.A., N.L. Mays, M.D. Balcer, E.D. Reavie, D.M. Reid, R. Sturtevant, and X. Gao. 2007.
The response of zooplankton and phytoplankton from the North American Great Lakes to filtration.
Harmful Algae 6: 547-566.
R-3
-------
Carlton, J.T., D.M. Reid and H. van Leeuwen. 1995. Shipping Study. The Role of Shipping in the
Introduction of Nonindigenous Aquatic Organisms to the Coastal Waters of the United States (other
than the Great Lakes) and an Analysis of Control Options. Report No. CG-D-11-95, U. S. Coast Guard,
Groton CT and U. S. Dept. of Transportation, Washington DC.
Caron, D.A., P.D. Countway and M.V. Brown. 2004. The growing contributions of molecular biology
and immunology to protistan ecology: Molecular signatures as ecological tools. Journal of Eukaryotic
Microbiology 51:38-48.
Cohen, A.N. 1998. Ships' Ballast Water and the Introduction of Exotic Organisms into the San
Francisco Estuary: Current Status of the Problem and Options for Management. A report for the
Collaboration Among State and Federal Agencies to Improve California's Water Supply (CALFED) and
the California Urban Water Agencies. San Francisco Estuary Institute, Richmond, CA.
Cohen, A.N. and B. Foster. 2000. The regulation of biological pollution: Preventing exotic species
invasions from ballast water discharged into California coastal waters. Golden Gate University Law
Review^: 787-883.
Colnar, A.M. and W.G. Landis. 2007. Conceptual model development for invasive species and a
regional risk assessment case study: the European Green Crab, Carcinus maenas, at Cherry Point,
Washington USA. Human and Ecological Risk Assessment 13:120-155.
Connelly, S.J., E.A. Wolyniak, K.L. Dieter, C.E. Williamson, and K.L. Jellison. 2007. Impact of
zooplankton grazing on the excystation, viability, and infectivity of the protozoan pathogens
Cryptosporidium parvum and Giardia lamblia. Applied and Environmental Microbiology 73:7277-7282.
Cromeans, T., J. Narayanan, K. Jung, G. Ko, D. Wait, and M. Sobsey. 2005. Development of Molecular
Methods to Detect Infections Viruses in Water. American Water Works Association Research
Foundation Report 90995F. ISBN 9781843398912, 128 pp.
Cuker, B.E. and L. Hudson Jr. 1992. Type of suspended clay influences zooplankton response to
phosphorus loading. Limnology and Oceanography 37:566-576.
Dames & Moore. 1998. Ballast water management technical memorandum for the Port of Oakland. Job
No. 02801-028-086, November 6, 1998. Dames & Moore, San Francisco, CA.
Dames & Moore. 1999. Phase I Final Report Ballast Water Exchange and Treatment. A Report for the
California Association of Port Authorities, Pacific Merchant Shipping Association, Steamship
Association of Southern California, and Western States Petroleum Association. Dames & Moore, San
Francisco, CA.
Darzynkiewicz, Z., X. Li and J. Gong. 1994. Assays of cell viability: Discrimination of cells dying by
apoptosis. In: Methods in Cell Biology, by Z. Darzynkiewicz, J.P. Robinson, and H.A. Crissman (eds.).
Academic Press, New York.
Deines, A.M., V. Chen and W.G. Landis. 2005. Modeling the risks of non-indigenous species
introductions using a patch-dynamics approach incorporating contaminant effects as a disturbance. Risk
Analysis 6:1637-1651.
R-4
-------
Drake, J.M. 2004. Allee Effects and the Risk of Biological Invasion. Risk Analysis 24:795-802.
Drake, J.M. 2005. Risk Analysis For Invasive Species And Emerging Infectious Diseases: Concepts
And Applications. Am. Midi. Nat. 153:4-19.
Dobbs, F.C. and A. Rogerson. 2005. Ridding ships' ballast water of microorganisms. Environmental
Science and Technology 39: 259A-264A.
Doblin, M.A. and F.C. Dobbs (2006) Setting a size-exclusion limit to remove toxic dinoflagellate cysts
in ships' ballast water. Marine Pollution Bulletin 52:259-263.
Dobroski, N., C. Scianni, D. Gehringer, and M. Falkner. 2009. 2009 Assessment of the Efficacy,
Availability and Environmental Impacts of Ballast Water Treatment Systems for Use in California
Waters. Produced for the California State Legislature. Available at:
http://www.slc.ca.gov/spec_pub/mfd/ballastwater/documents/2009cslctechreportfmal.pdf.
Dobroski, N., Scianni, C., Takata, L., & Falkner, M. 2009. October 2009 update: Ballast water treatment
technologies for use in California waters. California State Lands Commission, Marine Invasive Species
Program. Produced for the California State Legislature. Final Report available at:
http://www.slc.ca.gov/Spec Pub/MFD/Ballast Water/Reports Presentations.html.
Drake, J.M. and D.M. Lodge. 2006. Allee effects, propagule pressure and the probability of
establishment: risk analysis for biological invasions. Biological Invasions 8:365-375.
Drake, J.M., D.M. Lodge and M. Lewis. 2005. Theory and preliminary analysis of species invasions
from ballast water: controlling discharge volume and location. American Midland Naturalist 154: 459-
470.
Elliot, J.M. 1971. Some Methods for the Statistical Analysis of Samples of Benthic Invertebrates.
Scientific Publication No. 25, Freshwater Biological Association, Windermere, UK.
Fout, G.S., F.W. Schaefer III, J. W. Messer, D.R. Dahling and R.E. Stetler. 1996. ICRMicrobial
Laboratory Manual. U.S. Environmental Protection Agency, Washington, DC.
Garvey, M., B. Moriceau and U. Passow. 2007. Applicability of the FDA assay to determine the
viability of marine phytoplankton under different environmental conditions. Marine Ecology Progress
Series 352: 17-26.
Gauthier, D. and D.A. Steel. 1996. A Synopsis of the Situation Regarding the Introduction of
Nonindigenous Species by Ship-Transporting Ballast Water in Canada and Selected Countries.
Canadian Manuscript Report of Fisheries and Aquatic Sciences No. 2380. Marine Environmental
Sciences Division, Department of Fisheries and Oceans, Maurice Lamontagne Institute, Mont-Joli
(Quebec), Canada.
Glosten Associates.2002. Pacific Ballast Water Treatment Pilot Project: Ballast Water Transfer Study:
Technical Feasibility with Associated Capital Costs. Report prepared for the Port of Seattle, Seattle,
WA. File No. 01080. Glosten Associates, Seattle, WA, 36 pp.
R-5
-------
Great Ships Initiative. 2010. Report of the Land-Based Freshwater Testing by the Great Ships Initiative
of the Siemens SiCURETM Ballast Water Management System for Type Approval According to
Regulation D-2 and the Relevant EVIO Guidelines. Great Ships Initiative.
http://www.nemw.org/GSI/GSI-LB-F-A-l.pdf.
Gregg, M.D. and G.M. Hallegraeff 2007. Efficacy of three commercially available ballast water
biocides against vegetative microalgae, dinoflagellate cysts and bacteria. Harmful Algae 6:567-584.
Griffin, D.W., K.A. Donaldson, J.H. Paul and J.B. Rose. 2003. Pathogenic human viruses in coastal
waters. Clinical Microbiology Reviews 16:129-143.
Guy, R.A., P. Payment, UJ. Krull and P.A. Horgen. 2003. Real-time PCR for quantification of Giardia
and Cryptosporidium in environmental water samples and sewage. Applied and Environmental
Microbiology 69:5178-5185.
Hallegraeff, G.M. and CJ. Bolch. 1991. Transport of toxic dinoflagellate cysts via ships' ballast water.
Marine Pollution Bulletin 22: 27-30.
Hallegraeff, G.M. and CJ. Bolch. 1992. Transport of dinoflagellate cysts in ship's ballast water:
Implications for plankton biogeography and aquaculture. Journal of Plankton Research 14:1067-1084.
Hallegraeff, G.M., J.P. Valentine, J.-A. Marshall and CJ. Bolch. 1997. Temperature tolerances of toxic
dinoflagellate cysts: application to the treatment of ships' ballast water. Aquatic Ecology 31:47-52.
Hampel, M., I. Moreno-Garrido, C. Sobrino, H.L.M. Lubia and J. Blasco. 2001. Acute toxicity of LAS
homologues in marine microalgae: Esterase activity and inhibition growth as endpoints of toxicity.
Ecotoxicology and Environmental Safety 48:287-292.
Havelaar, A.H. 1994. Application of HACCP to Drinking Water Supply. Food Control 5(3): 145-152.
Hendry, G.A.F., J.D. Houghton and S.B. Brown. 1987. The degradation of chlorophyll - a biological
enigma. New Phytologist 107: 255-302.
Hoff, R.M., W.MJ. Strachan, C.W. Sweet, C.H. Chan, M. Shackleton, T.F. Bidleman, K.A. Brice, D.A.
Burniston, S. Cussion, D.F. Gatz, K. Harlin, and W.H. Schroeder (1996) Atmospheric deposition of
toxic chemicals to the Great Lakes: A review of data through 1994. Atmospheric Environment 30:3505-
3527.
Holm-Hansen, O. 1973. The use of ATP determinations in ecological studies. Bulletin of the Ecological
Research Commission (Stockholm) 17:215-222.
Howard, G., J. Bartram, S. Pedley, O. Schmoll, I. Chorus and P. Berger. 2006. Groundwater and public
health, pp. 3-19. In: Protecting Groundwater for Health: Managing the Quality of Drinking-Water
Sources, byO. Schmoll, G. Howard, J. Chilton, and I. Chorus (eds.). IWA Publishing, London, U.K.
Hunt, C.D., D.C. Tanis, T.G. Stevens, R.M. Frederick and R.A. Everett. 2005 (August 1). Verifying
ballast water treatment performance. Environmental Science and Technology 321A-328A.
R-6
-------
International Maritime Organization (IMO). 2004. International Convention for the Control and
Management of Ships' Ballast Water and Sediments, 2004. IMO, International Conference on Ballast
Water Management for Ships.
International Maritime Organization (IMO). 2005a. Guidelines tor Ballast Water Management
Equivalent Compliance (G3). IMO, Marine Environment Protection Committee, Resolution MEPC
123(53), adopted 22 July 2005.
International Maritime Organization (IMO). 2005b. Guidelines for Ballast Water Management and
Development of Ballast Water Management Plans (G4). IMO, Marine Environment Protection
Committee (MEPC), Resolution MEPC 127(53), adopted 22 July 2005.
International Maritime Organization (IMO). 2008a. Guidelines for Approval of Ballast Water
Management Systems (G8). Marine Environment Protection Committee (MEPC), Annex 4 Resolution
MEPC. 174(58). Available at:
http://www.regulations.gov/search/Regs/home.html#documentDetail?R=09000064807e8904.
International Maritime Organization (IMO). 2008b. Procedure for Approval of Ballast Water
Management Systems that Make Use of Active Substances (G9). Marine Environment Protection
Committee (MEPC), Annex 1 Resolution MEPC. 169(57). Available at:
http://www.regulations.gov/search/Regs/home.html#documentDetail?R=09000064807e890e.
Inubushi, K., P.C. Brookes and D.S. Jenkinson. 1989. Adenosine 5'-triphosphate and adenylates energy
charge in waterlogged soil. Soil Biology and Biochemistry 21:733-739.
Jeffrey, S.W. and M. Vest. 1997. Introduction to marine phytoplankton and their pigment signatures, pp.
37-87. In: Phytoplankton Pigments in Oceanography, by S.W. Jeffrey, R.F.C. Mantoura, and S.W.
Wright (eds.). The United Nations Educational, Scientific, and Cultural Organization (UNESCO), Paris,
France.
Jeffrey, S., R. Mantoura and S. Wright (eds.). 1997. Phytoplankton Pigments in Oceanography:
Guidelines to Modern Methods. UN Educational, Scientific & Cultural Organization, Paris, France.
Johnson, W.S. and D.M. Allen. 2005. Zooplankton of the Atlantic and Gulf Coasts - A Guide to Their
Identification and Ecology. John Hopkins University Press, Baltimore, MD.
Jones, W.J., C.M. Preston, R. Marin III, C.A. Scholin and R.C. Vrijenhoek. 2008. A robotic molecular
method for in situ detection of marine invertebrate larvae. Molecular Ecology Resources 8:540-550.
Kaneshiro, E.S., M.A. Wyder, Y.-P. Wu, and M.T. Cushion. 1993. Reliability of calceinacetoxy methyl
ester and ethidiumhomodimer or propidium iodide for viability assessment of microbes. Journal of
Microbiological Methods 17:1-16.
Karl, D.M. 1980. Cellular nucleotide measurements and applications in microbial ecology.
Microbiological Reviews 44:739-796.
R-7
-------
Karlson, B., C. Cusack and E. Bresnan. 2010. Microscopic and Molecular Methods for Quantitative
Phytoplankton Analysis. Intergovernmental Oceanographic Commission (IOC) Manuals and Guides
#55. IOC, United Nations Educational, Scientific and Cultural Organization, Paris, France, 114 pp.
Kay, S.H. 1985. Cadmium in aquatic food webs. Residue Reviews 96:13-43.
Keller, R.P., J.M. Drake, M.B. Drew and D.M. Lodge. 2010. Linking environmental conditions and ship
movements to estimate invasive species transport across the global shipping network. Diversity and
Distributions DOI: 10.1111/j. 1472-4642/2010/00696x
Kieft, T.L. and L.L. Rosaker. 1991. Application of respiration and adenylates-based soil microbiological
assays to deep subsurface terrestrial sediments. Soil Biology and Biochemistry 23:563-568.
King, D.M. and M.N. Tamburri. 2010. Verifying compliance with ballast water discharge regulations.
Ocean Development & International Law 41:152-165.
Kokinos, J.K., T.I. Eglington, M.A. Goni, JJ. Boon, P.A. Martoglio and D.M. Anderson. 1998.
Characterization of a highly resistant biomacromolecular material in the cell wall of a marine
dinoflagellate resting cyst. Organic Geochemistry 28:265-288.
Kolar, C.S. 2004. Risk assessment and screening for potentially invasive species. New ZealJ Mar Fresh
38:391-7.
Kolar, C.S. and D.M. Lodge. 2001. Progress in invasion biology: predicting invaders. Trends EcolEvol
16:199-204.
Kolar, C.S. and D.M. Lodge. 2002. Ecological predictions and risk assessment for alien species. Science
298:1233-6.
Kotinis, M., M. G. Parsons, T. Lamb and A. Sirviente. 2004. Development and investigation of the
ballast-free ship concept. Transactions Society of Naval Architects andMarine Engineers 112.
Kudela, R.M., M.D.A. Howard, B.D. Jenkins, P.E. Miller and GJ. Smith. 2010. Using the molecular
toolbox to compare harmful algal blooms in upwelling systems. Progress in Oceanography 85:108-121.
Laabir, M. and P. Gentien. 1999. Survival of toxic dinoflagellates after gut passage in the Pacific oyster
CrassostreagigasThunberg. Journal of Shellfish Research 18: 217-222.
Landis, W.G. 2004. Ecological risk assessment conceptual model formulation for nonindigenous
species. Risk Analysis 24:847-858.
LeChevallier, M.W. and K.K. Au. 2004. Water Treatment and Pathogen Control; Process Efficiency in
Achieving Safe Drinking Water. World Health Organization, Geneva. IWA Pubishing, London.
Lee II, H., D.A. Reusser, M. Frazier and G. Ruiz. 2010. Density matters: Review of approaches to
setting organism-based discharge standards. U.S. EPA, Office of Research and Development, National
Health and Environmental Effects Research Laboratory, Western Ecology Division, Washington, DC.
EPA/600/R-10/031.
R-8
-------
Lemieux, E.J., S. Robbins, K. Burns, S. Ratcliff, and P. Herring. 2008. Evaluation of Representative
Sampling for Rare Populations using Microbeads. Report No. CG-D-03-09, United States Coast Guard
Research and Development Center, Groton, CT.
Lemieux, E.J., L.A. Drake, S.C. Riley, B.C. Schrack, W.B. Hyland, J.F. Grant and C.S. Moser. 2010.
Design and Preliminary Use of a Commercial Filter Skid to Capture Organisms > 50 ^m in Minimum
Dimension (Nominally Zooplankton) for Evaluating Ships' Ballast Water Management Systems at Land-
Based Test Facilities. Report 6130/1029. Center for Corrosion Science and Engineering, Naval Research
Laboratory, Washington, DC, 41 pp.
Lloyd's Register. 2010. Ballast Water Treatment Technology: Current Status. Lloyd's Register, London.
Loganathan, E.G. and K. Kannan. 1994. Global organochlorine contamination trends: an overview.
Ambio 23:187-191.
Lund, J.W.G., C. Kipling and E.D. LeCren. 1958. The inverted microscope method for estimating algal
numbers and the statistical basis of estimations by counting. Hydrobiologia 11:143-170.
Maclsaac, E.A. and J.G. Stockner. 1993. Enumeration of phototrophic picoplankton by autofluorescence
microscopy, pp. 187-198. In: Handbook of Methods in Aquatic Microbial Ecology., by P.P. Kemp, B.F.
Sherr, E.B. Sherr, and JJ. Cole (eds.). Lewis Publishers, CRC Press, Inc., Boca Raton, FL.
Mackey, M.D., DJ. Mackey, H.W. Higgins and S.W. Wright. 1996. CHEMTAX - A program for
estimating class abundances from chemical markers: Application to HPLC measurements of
phytoplankton. Marine Ecology Progress Series 144:265-283.
Madigan, M.T. and J.M. Martinko. 2006. Biology of Microorganisms, 11th ed. Pearson Prentice Hall,
Upper Saddle River, NJ, p. G-14.
Maier, R.M., I.L. Pepper, and C.P. Gerba (eds.). Environmental Microbiology. Elsevier, New York, NY.
MARAD. Feb. 2011. U.S. Water Transport Statistical Snapshot. US DOT Maritime Administration.
http://www.marad.dot.gov/library landing_page/data and statistics/Data and Statistics.htm
Marine Environment Protection Committee (MEPC). 2006a (21 April). Harmful Aquatic Organisms in
Ballast Water - Basic Approval of Active Substances Used by Pure Ballast Management System.
International Maritime Organization, 55th session, Agenda Item 2, MEPC 55/2/5, International Maritime
Organization.
Marine Environment Protection Committee (MEPC). 2006b (28 Feb.). 54th Session, Agenda item 2 -
Harmful Aquatic Organisms in Ballast Water - Report of the first meeting of the Group of Experts on
the Scientific Aspects of Marine Environmental Protection Border-Wide Workgroup (GESAMP-
BWWG). MEPC 54/2/12, International Maritime Organization.
Maritime Environmental Resource Center (MERC). 2009c. Test Plan for the Performance Evaluation of
the Severn Trent DeNoraBalPure™ BP-2000 Ballast Water Management System. Chesapeake
Biological Laboratory, University of Maryland Center for Environmental Science, Solomons, MD.
R-9
-------
Maritime Environmental Resource Center (MERC). 2009a. Land-Based Evaluations of the Maritime
Solutions, Inc. Ballast Water Treatment System. MERC ER02-09, [UMCES] CBL 09-138.
Maritime Environmental Resource Center (MERC). 2009b. M/FCape Washington Ballast Water
Treatment Test Facility Validation. Chesapeake Biological Laboratory, University of Maryland Center
for Environmental Science, Solomons, MD.
Maritime Environmental Resource Center (MERC). 2010a. Land-Based Evaluations of the Siemens
Water Technologies SiCURE Ballast Water Management System. MERC ER02-10, [UMCES] CBL 10-
038.
Maritime Environmental Resource Center (MERC). 201 Ob. Land-Based Evaluations of the Severn Trent
De Nora BalPure BP-1000 Ballast Water Management System. MERC ER01-10, [UMCES] CBL 10-
015.
Marret, F. and K.A.F. Zonneveld. 2003. Atlas of modern organic-walled dinoflagellate cyst distribution.
Review ofPalaeobotanyandPalynology 125:1-200.
Matsuoka, K. and Y. Fukuyo. 2000. Technical Guide for Modern Dinoflagellate Cyst Study. Available
at: http://dinos.anesc.u-tokyo.ac.jp/technical guide/main.pdf.
Miller, A.W., M. Frazier, G.E. Smith, E.S. Perry, G.M. Ruiz and M.N. Tamburri. 2011. Enumerating
Sparse Organisms in Ships' Ballast Water: Why Counting to 10 is Difficult? Environmental Science and
Technology 45:3539-3546.
Minton, M.S., E. Verling, A.W. Miller and G.M. Ruiz. 2005. Reducing propagule supply and coastal
invasions via ships: effects of emerging strategies. Frontiers in Ecology and the Environment 3: 304-
308.
Moeller, P.D.R., S.L. Morton, B.A. Mitchell, S.K. Sivertsen, E.R. Fairey, T.M. Mikulski, H.B. Glasgow,
N.J. Deamer-Melia, J.M. Burkholder and J.S. Ramsdell. 2001. Current progress in isolation and
characterization of toxins isolated from Pfiesteria spp. Environmental Health Perspectives 109:739-743.
Montresor, M., L. Nuzzo and M.G. Mazzocchi. 2003. Viability of dinoflagellate cysts after the passage
through the copepod gut. Journal of Experimental Marine Biology and Ecology 287:209-221.
Mosher, J.J., B.S. Levison, and C.G. Johnston. 2003. A simplified dehydrogenase enzyme assay in
contaminated sediment using 2-(p-iodophenyl)-3(p-nitrophenyl)-5-phenyl tetrazolium chloride. Journal
of Microbiological Methods 53:411-415.
Munn, C.B. 2006. Viruses as pathogens of marine organisms - from bacteria to whales. Journal of the
Marine Biological Association of the UK 86:453-467.
Murphy, K.R., D. Ritz, and C.L. Hewitt. 2002. Heterogeneous zooplankton distribution in a ship's
ballast tanks. Journal of Plankton Research 24:729-734.
Myers, GJ. 1979. The Art of Software Testing. John Wiley & Sons, New York, NY, pp. 12-13,128.
R-10
-------
National Environmental Coalition on Invasive Species (NECIS). July 2008. New Ballast Tank
Legislation. http://www.pollutionengineering.com/Articles/Feature Article/BNP GUTD 9-5-
2006 A 100000000000003 80477
National Environmental Services Center (NESC). 2000a. Slow Sand Filtration. Tech Brief: A National
Drinking Water Clearinghouse Fact Sheet. National Environmental Services Center, West Virginia
University. Morgantown, WV.
National Environmental Services Center (NESC). 2000b. Ultraviolet Disinfection. Tech Brief: A
National Drinking Water Clearinghouse Fact Sheet. National Environmental Services Center, West
Virginia University. Morgantown, WV.
National Research Council (NRC) Marine Board. 1996. Stemming the Tide: Controlling Introductions
of Nonindigenous Species by Ships' Ballast Water. Committee on Ships' Ballast Operations, Marine
Board, Commission on Engineering and Technical Systems, NRC. National Academy Press,
Washington, D.C.
Nelson, B.N., EJ. Lemieux, L. Drake, D. Anderson, D. Kulis, K. Burns, D. Anderson, N. Welschmeyer,
S. Smith, C. Scianni C, T. Weir, S. Riley and P. Herring. 2009. Phytoplankton Enumeration and
Evaluation Experiments. Report No. CG-D-06-09, United States Coast Guard Research & Development
Center, Groton, CT.
Notermans, S., G. Gallhoff, M.H. Zweitering and G.C. Mead. 1994. The HACCP concept:
Identification of potentially hazardous micro-organisms. Food Microbiology 11:203-214.
Oemke, D. 1999. The Treatment of Ships' Ballast Water. EcoPorts Monograph Series No. 18, Ports
Corporation of Queensland, Brisbane, Australia, 102 pp.
Ogilvie, DJ. 1995. Land-based treatment options. Pages 113-123 in: Ballast Water. A Marine Cocktail
on the Move. Proceedings of the National Symposium., 27-29 June 1995, Wellington, New Zealand. The
Royal Society of New Zealand. Miscellaneous Series 30.
Oliver, J.D. 1993. Formation of viable but nonculturable cells, pp. 239-272. In: Starvation in Bacteria.,
by S. Kjelleberg (ed.). Plenum Press, New York, NY.
Oudot C., R. Gerard, P. Morin and I. Gningue. 1988. Precise shipboard determination of dissolved
oxygen (Winkler procedure) for productivity studies with a commercial system. Limnology and
Oceanography 33:146-150.
Parshionikar, S., I. Laseke and G.S. Fout. 2010. Use of propidiummonoazide in reverse transcriptase
PCR to distinguish between infectious and noninfectious enteric viruses in water samples. Applied and
Environmental Microbiology 76:4318-4326.
Parsons, M. G. 2010. The Variable Buoyancy Ship: A Road to the Elimination of Ballast. Proceedings
from EVIO and World Maritime University Research and Development Forum: Emerging Ballast Water
Management Systems, January 26-29, Malmo, Sweden.
R-ll
-------
Parsons, M.G. 1998. Flow-through ballast water exchange. Society of Naval Architects and Marine
Engineers (NAME) Transactions 106:485-493.
Pate, A.S., A.E. De Souza and D.R.G. Farrow. 1992. Agricultural Pesticide Use in Coastal Areas: A
National Summary. NOAA, Rockville, MD.
Peal, D.B. 1975. Phthalate Esters: Occurrence and Biological Effects. Springer-Verlag, New York, NY,
1-41.
Phillips, S. 2006. Ballast Water Issue Paper. Report for the Pacific States Marine Fisheries Commission.
Portland, OR. 34 pp. (http://www.aquaticnuisance.org/wordpress/wp-
content/uploads/2009/01/ballast_water_issue_paper.pdf).
Pollutech. 1992. A Review and Evaluation of Ballast Water Management and Treatment Options to
Reduce the Potential for the Introduction of Non-native Species to the Great Lakes.
Pollutech Environmental, Ltd., Sarnia, Ontario, prepared for the Canadian Coast Guard, Ship Safety
Branch, Ottawa, Canada.
Porra, RJ. 1991. Recent advances and re-assessments in chlorophyll extraction and assay procedures for
terrestrial, aquatic, and marine organisms, including recalcitrant algae, pp. 31-57. In: Chlorophylls, by
H. Scheer (ed.). CRC Press, Boca Raton, FL.
Porter, J., J. Diaper, C. Edwards, and R. Pickup. 1995. Direct measurements of natural planktonic
bacterial community viability by flow cytometry. Applied and Environmental Microbiology 61: 2783-
2786.
Posch, T., J. Pernthaler, A. Alfreider and R. Psenner. 1997. Cell-specific respiratory activity of aquatic
bacteria studied with the tetrazolium reduction method, cyto-clear slides, and image analysis. Applied
and Environmental Microbiology 63:867-873.
Reavie, E.D., A.A. Cangelosi andL.E. Allinger. 2010. Assessing ballast water treatments: Evaluation of
viability methods for ambient freshwater microplankton assemblages. Journal of Great Lakes Research
36: 540-547.
Reeves, E. 1998. Protection of the Great Lakes from Infection by Exotic Organisms in Ballast Water.
Report for Ninth US Coast Guard District, Cleveland OH.
Reeves, E. 1999. Exotic Policy: An IJC White Paper On Policies for the Prevention of the Invasion of
the Great Lakes by Exotic Organisms. Documents on Water Law, College of Law, University of
Nebraska - Lincoln. 132 pages.
Rigby, G. and A.H. Taylor. 200 la. Ballast Water Treatment to Minimize the Risks of Introducing
Nonindigenous Marine Organisms into Australian Ports. Ballast Water Research Series Report No. 13,
Australian Government Department of Agriculture, Fisheries, and Forestry, Canberra, Australia, 93 pp.
Rigby, G. and A. Taylor. 2001b. Ballast water management and treatment options. Transactions of the
Institute of Marine Engineers 113: 79-99.
R-12
-------
Romano, G., A. lanora, L. Santella, and A. Miralto. 1996. Respiratory metabolism during embryonic
subitaneous and diapause development in Pontellamediterranea (Crustacea, Copepoda). Journal of
Comparative Physiology 166: 157-163.
Ropkins, K. and A. J. Beck. 2000. Evaluation of worldwide approaches to the use of HACCP to control
food safety. Trends in Food Science & Technology 11(1):10-21.
Rosacker, L.L. and T. L. Kieft. 1990. Biomass and adenylate energy charge of a grassland soil during
drying. Soil Biology and Biochemistry 22:1121-1127.
Ruiz, G.M., G.E. Smith, andM. Sytsma. 2006. Workshop Report on Testing of Ballast Water Treatment
Systems: General Guidelines and Step-wise Strategy Toward Shipboard Testing. Aquatic Bioinvasion
Research and Policy Institute, a joint program of the Smithsonian Environmental Research Center
(SERC) and Portland State University (PSU). Submitted to the Pacific States Marine Fisheries
Commission, Alaska Department of Fish and Game, and Prince William Sound Regional Citizens
Advisory Council, Anchorage, AK.
Sandrin, T.R., D.C. Herman, andR.M. Maier. 2009. Physiological methods, pp. 191-223. In:
Environmental Microbiology, by R.M. Maier, I.L. Pepper, and C.P. Gerba (eds.). Elsevier, New York,
NY.
Sarmento, H. and J.-P. Descy. 2008. Use of marker pigments and functional groups for assessing the
status of phytoplankton assemblages in lakes. Journal of Applied Phycology 20:1001-1011.
Schliiter, L., T.L. Lauridsen, G. Krogh and T. J0rgense. 2006. Identification and quantification of
phytoplankton groups in lakes using new pigment ratios - a comparison between pigment analysis by
HPLC and microscopy. Freshwater Biology 51:1474-1485.
Schmid, H., F. Bauer and H.B. Stich. 1998. Determination of algal biomass with HPLC pigment
analysis from lakes of different trophic state in comparison to microscopically measured biomass.
Journal of Plankton Research 20:1651-1661.
Shaw, S. and K. Kurunthachalam. 2009. Polybrominateddiphenyl ethers in marine ecosystems of the
American continents: Foresight from current knowledge. Reviews on Environmental Health 24:157-229.
Songster-Alpin, M.S. and R.L. Klotz. 1995. A comparison of electron transport system activity in stream
and beaver pond sediments. Canadian Journal of Fisheries and Aquatic Sciences 52:1318-1326.
Sperber, W. H. and R. F. Stier. 2009. Happy 50th Birthday to HACCP: Retrospective and Prospective.
foodsafetymagazine.com (December 2009).
Steinberg, M.K., M.R. First, EJ. Lemieux, L.A. Drake, B.N. Nelson, D. Kulis, D. Anderson, N.
Welschmeyer, and P. Herring (accepted with revisions) Comparison of techniques used to count single-
celled viable phytoplankton. Journal of Applied Phycology.
R-13
-------
Steinberg, M.K., S.C. Riley, S.H. Robbins, B.N. Nelson, EJ. Lemieux and L.A. Drake. 2010. Multi-site
validation of a method to determine the viability of organisms > 10 um and < 50 um (nominally protists)
in ships' ballast water using two vital, fluorescent stains. NRL Letter Report 6130/1016, Washington,
DC.
Stevens, T.G., R.M. Frederick, R.A. Everett, J.T. Hurley, C.D. Hunt and D.C. Tanis. 2004. Performance
verification of ballast water treatment technologies by USEPA/NSF Environmental Technology
Verification Program. In: Proceedings of the 2n International Ballast Water Treatment R&D
Symposium, by J.T. Matheickal and S. Raaymakers (eds.). EVIO, London, UK.
Suban, V., P. Vidmar and M. Perkovic, 2010. Ballast Water Replacement with Fresh Water - Why Not?
Proceedings from EVIO and World Maritime University Research and Development Forum: Emerging
Ballast Water Management Systems, January 26-29, Malmo, Sweden.
Sugita, H., T. Mitsaya, K. Amanuma, K. Hayashi, T. Asai, C. Maruyama and Y. Degechi. 1992.
Ultraviolet susceptibility of three marine fish pathogens. Bulletin of the College of Agriculture and
Veterinary Medicine 49:117-121.
Suttle, C.A. 2007. Marine viruses - major players in the global ecosystem. Nature Reviews -
Microbiology 5:801-812.
Suttle, C.A., A.M. Chan and M.T. Cottrell. 1991. Use of ultrafiltration to isolate viruses from seawater
which are pathogens of marine phytoplankton. Applied and Environmental Microbiology 57:721-726.
Takano, C.T., C.E. Folsome, and D.M. Karl. 1983. ATP as a biomass indicator for closed ecosystems.
BioSystems 44:739-796.
Taylor, L.R. 1961. Aggregation, variance and the mean. Nature 189:732-735.
Turner, R.E. 1978. Variability of the reverse-flow concentration technique of measuring plankton
respiration. Estuaries 1:65-68.
United States Coast Guard. 1999. Implementation of the National Invasive Species Act of 1996 (NISA).
Interim rule. U.S. Federal Register 64(94): 26672-26690 (May 17, 1999).
United States Coast Guard. 200 la. Implementation of the National Invasive Species Act of 1996
(NISA). Final rule. U.S. Federal Register 66(225): 58381-58393 (November 21, 2001).
United States Coast Guard. 2001b. Report to Congress on the Voluntary National Guidelines for Ballast
Water Management. Washington D.C.
United States Coast Guard. 2002. Standards for Living Organisms in Ship's Ballast Water Discharged in
U.S. Waters: A Proposed Rule by the Transportation Department and the Coast Guard. 67 FR 9632,
March 4, 2002.
United States Coast Guard. 2004. Mandatory Ballast Water Management Program for U.S. Waters. U.S.
Federal Register 69(144): 44952-44961 (July 28, 2004).
R-14
-------
United States Coast Guard. 2008. Ballast Water Discharge Standard - Draft Programmatic
Environmental Impact Statement. U.S. Coast Guard, Washington, DC, 218 pp.
United States Congress. 1990. Nonindigenous Aquatic Nuisance Prevention and Control Act of 1990.
U.S. Congress, Public Law 101-6461, enacted on November 29, 1990.
United States Congress. 1996. National Invasive Species Act of 1996. U.S. Congress, Public Law 104-
332, enacted on October 26, 1996.
United States Environmental Protection Agency (U.S. EPA). 1993. Methods for the Determination of
Inorganic Substances in Environmental Samples. Report 600/R-93-100. Office of Research and
Development, U.S. EPA, Washington, DC.
United States Environmental Protection Agency (U.S. EPA). 1997. Method445.0: In vitro
Determination of Chlorophyll a and Pheophytina in Marine and Freshwater Algae by Fluorescence.
U.S. EPA, Cincinnati, OH.
United States Environmental Protection Agency (U.S. EPA). 1999. Criteria for Evaluation of Proposed
Protozoan Detection Methods. Report #EPA 815-K-99-02. Office of Research and Development, U.S.
EPA, Cincinnati, OH. Available at: http://www.epa.gov/microbes/critprot.pdf.
United States Environmental Protection Agency (U.S. EPA). 2000. Deposition of Air Pollutants to the
Great Waters. Third Report to Congress. U.S. EPA, Washington, DC.
United States Environmental Protection Agency (U.S. EPA). 2001. Aquatic Nuisance Species In Ballast
Water Discharges: Issues and Options. Draft Report for Public Comment - September 10, 2001. Office
of Water, Office of Wetlands, Oceans and Watersheds, and Office of Wastewater Management, U.S.
EPA, Washington, D.C.
United States Environmental Protection Agency (U.S. EPA). 2002a. Methods for Measuring the Acute
Toxicity of Effluents to Freshwater and Marine Organisms, 5 Edition. Report # EPA 821-R-02-012.
Office of Water, U.S. EPA, Washington, DC.
United States Environmental Protection Agency (U.S. EPA). 2002b. Short-Term Methods for Estimating
the Chronic Toxicity of Effluents and Receiving Waters to Freshwater Organisms, 4th Edition. Report #
EPA 821-R-02-013. Office of Water, U.S. EPA, Washington, DC.
United States Environmental Protection Agency (U.S. EPA). 2002c. Short-Term Methods for
Estimating the Chronic Toxicity of Effluents and Receiving Waters to Marine andEstuarine Organisms,
3rd edition. Report # EPA 821-R-02-014. Office of Water, U.S. EPA, Washington, DC.
United States Environmental Protection Agency (U.S. EPA). 2003. Ambient Water Quality Criteria for
Dissolved Oxygen, Water Clarity and Chlorophylls/or the Chesapeake Bay and Its Tidal Tributaries.
Report EPA 903-R-03-002. Office of Water, U.S. EPA, Washington, DC.
United States Environmental Protection Agency (U.S. EPA). 2010 (July). Environmental Technology
Verification Program (ETV) Draft Generic Protocol for the Verification of Ballast Water Treatment
Technologies, Version 5.1 .U. S. EPA ETV in cooperation with the U. S. Coast Guard Environmental
R-15
-------
Standards Division (CG-5224) and the U.S. Naval Research Laboratory. National Sanitation Foundation
International, Ann Arbor, MI, 62 pp.+ appendices.
Unnevehr, L. J. and H.H. Jensen. 1996. HACCP as a Regulatory Innovation to Improve Food Safety in
the Meat Industry. American Journal of Agricultural Economics 78(3):764-769.
Vaquer-Sunyer, R. and C.M. Duarte. 2008. Thresholds of hypoxia for marine biodiversity. Proceedings
of the National Academy of Sciences of the United States of America 105:15452-15457.
Veldhuis, M.J.W., G.W. Kraaij, and K.R. Timmermans. 2001. Cell death in phytoplankton: correlation
between changes in membrane permeability, photosynthetic activity, pigmentation and growth.
European Journal ofPhycology 36:1-13.
Vezie, C., J. Rapala, J. Vaitomaa, J. Seitsonen, and K. Sivonen. 2007. Effect of nitrogen and phosphorus
on growth of toxic and nontoxic Microcystis strains and on intracellular microcystin concentrations.
MicrobialEcology 43:443-454.
Victoria Environmental and Natural Resources Committee (ENRC). 1997.Report on Ballast Water and
Hull Fouling in Victoria. Parliament of Victoria, ENRC, Victorian Government Printer, Melbourne,
Australia.
Vollenweider, R. A. (ed.). 1974. A Manual on Methods for Measuring Primary Production in Aquatic
Environments, 2nd edition. International Biological Program Handbook No. 12, Blackwell Scientific,
Oxford, U.K.
Wang, X.C., F.G. Qiu and P.K. Jin. 2006. Safety of treated water for re-use purposes—comparison of
filtration and disinfection processes. Water Science & Technology 53(9):213-220.
Wetzel, R.G. 2001. Limnology, 3rd edition. Academic Press, New York, NY.
Wetzel, R.G. and G.E. Likens. 2000. Limnological Analyses, 3rd edition. Springer-Verlag, New York,
NY.
Williams, S.C., Y. Hong, D.C.A. Danavall, M.H. Howard-Jones, D. Gibson, M.E. Frischer and P.G.
Verity. 1998. Distinguishing between living and nonliving bacteria: Evaluation of the vital stain
propidium iodide and its combined use with molecular probes in aquatic samples. Journal of
Microbiological Methods 32:225-236.
World Health Organization (WHO). 2008. Guidelines for Drinking-water Quality. Third Edition.
Incorporating the First and Second Addenda. Volume 1: Recommendations. World Health Organization,
Geneva.
R-16
-------
APPENDIX A: DOCUMENTS ON BALLAST WATER TECHNOLOGIES PROVIDED TO THE PANEL
This Appendix lists the documents available to the Panel for its assessment of ballast water technologies. These documents are
available in the EPA Docket: Science Advisory Board Review of the Availability and Efficacy of Ballast Water Treatment Technology for
EPA's Office of Water and the United States Coast Guard (at www.regulations.gov under docket number EPA-HQ-OW-2010-05 82). Shaded
rows indicate those documents that the Panel used as reliable sources of credible data for their assessment.
Table A-l. Documents Available to the Panel for its Assessment of Ballast Water Technologies
System
Group 1: Third-Party Reviews
General
General
General
General
General
General
Document Title
Ballast Water Treatment Technology: Current Status
2009 Assessment of the Efficacy, Availability and Environmental Impacts of
Ballast Water Treatment Systems for Use in California Waters
October 2010 Update: Ballast Water Treatment Technologies for Use in
California Waters
Density Matters: Review of Approaches to Setting Organism -Based Ballast
Water Discharge Standards
International Convention for the Control and Management of Ships' Ballast
Water and Sediments, 2004 - List of ballast water management systems that
make use of Active Substances which received Basic and Final Approvals
Ballast Water Treatment Advisory
Group 2: Direct Data Reports and Supporting Information
Ecochlor® Ballast Water Treatment System
Ecochlor® Ballast Water Treatment System
Ecochlor® Ballast Water Treatment System
(Filtration+chlorine dioxide)
STEP 2006 Application Form - Section 4.0: Proof of Ballast Water Treatment
Performance
Final Environmental Assessment Review of the Application by Atlantic
Container Lines for Acceptance of the Vessel M/V Atlantic Compass and the
Ecochlor™ Inc. Technology into the USCG Shipboard Technology Evaluation
(STEP) Program
Final report of the land-based testing of the Ecochlor®-system, for Type
Approval according to regulation-D2 and the relevant IMO guideline (April -
July 2008)
Date
2/1/2010
1/1/2009
10/15/2009
7/2/2005
9/24/2009
6/8/2010
6/28/2005
8/1/2008
2/1/2009
A-l
-------
System
Electro-Clean Ballast Water Management
System
Electro-Cleen™ System
GloEn-Patrol™ Ballast Water Management
System
Greenship's Ballast Water Management System
Hyde GUARDIAN Ballast Water Treatment
System
Hyde GUARDIAN Ballast Water Treatment
System (Filtration+UV)
Hyde GUARDIAN Ballast Water Treatment
System
Hyde GUARDIAN Ballast Water Treatment
System (Filtration+UV)
Hyde GUARDIAN Ballast Water Treatment
System
MSI (Filtration+UV)
MH Systems (Deoxygenation)
NEI Venturi Oxygen Stripping (VOS)
NEI Venturi Oxygen Stripping (VOS)
NEI Venturi Oxygen Stripping (VOS)
NEI Venturi Oxygen Stripping (VOS)
Document Title
Development of technologies on test facility and procedures for the land-based
test as a type approval test at ballast water treatment system
Information on the Type Approval Certificate of the Electro-Cleen™ System
(ECS)
Type Approval Certificate of Ballast Water Management System
Landbased Test Report - Test Cycle Summary
Environmental Acceptability Evaluation of the Hyde GUARDIAN Ballast Water
Treatment System as Part of the Type Approval Process
Final report of the land-based testing of the Hyde-Guardian™ -System, for Type
Approval according to the Regulation D-2 and the relevant IMO Guideline
(April - July 2008)
Type Approval Certificate of Ballast Management System
Shipboard Trials of Hyde "Guardian" System in Caribbean Sea and Western
Pacific Ocean, April 5th - October 7th, 2008
Type Approval of the Hyde GUARDIAN™ Ballast Water Management System
MERC Land-Based Evaluations of the Maritime Solutions, Inc. Ballast Water
Treatment System
Ballast water treatment by De-oxygenation with elevated CO2 for a shipboard
installation
Short-term Toxicity Testing of a De-oxygenation Ballast Water Treatment to
Receiving Water Organisms. Final Report.
Short-term Chronic Toxicity Testing of a De-oxygenation Ballast Water
Treatment to Receiving Water Organisms. Final Report.
STEP 2006 Application Form.
Type Approval Certificate of Ballast Water Management System; Ballast Water
Management System Type Approval Compliance Certificate
Date
6/30/2005
2/20/2009
12/4/2009
6/29/2005
4/20/2009
1/1/2009
4/29/2009
4/1/2009
5/7/2009
11/1/2009
7/23/2003
8/29/2008
3/27/2009
3/1/2006
7/6/2009;
7/8/2007;
1/19/2010
A-2
-------
System
NEI Venturi Oxygen Stripping (VOS)
(deoxygenation & decavitation)
NEI Venturi Oxygen Stripping (VOS)
(deoxygenation & decavitation)
OceanSaver® Ballast Water Management
System
OceanSaver® Ballast Water Management
System
OceanSaver® Ballast Water Management
System
OptiMarin Ballast System
Optimarin (Filtration+UV)
Peraclean
PureBallast 250-2500
PureBallast (Filtration+UV+TiO2)
PureBallast (Filtration+UV+TiO2)
SEDNA ® 250
SEDNA® ballast water treatment system using
PERACLEAN® Ocean
SEDNA®-System
SEDNA®-System
Severn Trent De Nora (BalPure)
Severn Trent De Nora (BalPure)
Severn Trent De Nora (BalPure)
Document Title
Application for Type Approval Certification: NEI Treatment Systems' Venturi
Oxygen Stripping Ballast Water Management System.
Evaluations of a Ballast Water Treatment to Stop Invasive Species and Tank
Corrosion.
Det Norske Veritas Type Approval Certificate
Type Approval Certificate of the OceanSaver ® BWMS
Information on the Type Approval Certificate of the OceanSaver® Ballast Water
Management System
Det Norske Veritas Type Approval Certificate
Land based testing of the OptiMarin ballast water management system of
OptiMarin AS - Treatment Effect Studies
Toxic Shock as New Ballast Water Treatment Fails Test
Det Norske Veritas Type Approval Certificate
Land-based testing of the PureBallast Treatment System of AlfaWall AB
Shipboard testing of the PureBallast Treatment System of AlfaWall AB
Type Approval Certificate of Ballast Water Management System
Effective Protection Against "Stowaways": Ballast Water Management System
of Hamann and Evonik Receives Final Approval
Final report of the land-based and shipboard testing of the SEDNA®-system
Summary of Additional Provisions of the Type Approval Certificate of Ballast
Water Management System SEDNA 250 of Hamann AG
Washington State Dept. of Fish and Wildlife Application Package Ballast Water
Treatment System
Marrowstone Sodium Hypochlorite Mesocosm September 2004
Environmental Assessment Review of the Application for Acceptance of the
SeaRiver Maritime Inc. S/R American Progress and Severn Trent de Nora
BalPure™ System into the Shipboard Technology Evaluation Program (STEP)
Date
3/1/2007
6/27/2005
4/8/2009
4/17/2009
5/6/2009
11/12/2009
8/1/2008
2/9/2010
6/27/2008
9/1/2008
5/1/2008
8/16/2008
6/11/2008
3/1/2008
8/1/2008
8/8/2005
9/1/2004
2/1/2009
A-2
-------
System
Severn Trent (Filtration+electrochlorination)
Severn Trent (Filtration+electrochlorination)
Siemens SICURE Ballast Water Management
System
Siemens (Filtration+electrochlorination
Siemens SICURE Ballast Water Management
System (Filtration+electrochlorination)
Group 3: G9 Files
"ARA Ballast" Ballast Water Management
System (formerly Blue Ocean Guardian
BWMS)
Alfa Laval Ballast Water Management System
(PureBallast)
Alfa Laval Ballast Water Management System
(PureBallast)
AquaStar Ballast Water Management System
AquaTriComb Ballast™ Water Treatment
System
AquaTriComb Ballast™ Water Treatment
System
ATLAS-DANMARK, TG Ballastcleaner and
TG Environmentalguard,
Sunrui Ballast Water Management System,
DESMI Ocean Guard,
Blue Ocean Guard (BOG)
BalClor ™ ballast water management system
(formerly Sunrui BWMS)
Document Title
Final report of the land-based testing of the BalPure®-BWT-System
MERC Land-Based Evaluations of the Severn Trent De Nora BalPure™ BP-
1000 Ballast Water Management System
A Great Lake Relevancy Preamble to the GSI Report on Land-Based Testing
Outcomes for the Siemens SICURE Ballast Water Management System
MERC Land-Based Evaluations of the Siemens Water Technologies SiCURE
Ballast Water Management System
Report of the Land-Based Freshwater Testing of the Siemens SiCURE Ballast
Water Management System
Application for Final Approval of "ARA Ballast" Ballast Water Management
System
Basic Approval of Active Substances used by PureBallast management system
Application for Final Approval of a ballast water management
system using Active Substances
Application for Basic Approval of AquaStar Ballast Water Management System
Application for Basic Approval of the AquaTriComb Ballast Water Treatment
System
Application for Basic Approval of the AquaTriComb™ Ballast Water Treatment
System Corrigendum
Report of the eleventh meeting of the GESAMP -Ballast Water Working Group
(GESMP-BWWG)
Application for Final Approval of BalClor ™ ballast water management system
Date
1/1/2010
7/1/2009
4/28/2010
11/1/2009
5/15/2010
3/23/2010
4/21/2006
12/15/2006
3/18/2010
12/16/2008
6/29/2009
12/1/2009
3/22/2010
A-4
-------
System
BalPure®
BalPure®
Blue Ocean Guardian (BOG) Ballast Water
Management System
Blue Ocean Shield Ballast Water Management
System
BlueSeas Ballast Water Management System
CleanBallast!
CleanBallast!
CleanBallast!
ClearBallast,
Greenship Sedinox,
AquaTriComb
DESMI Ocean Guard Ballast Water
Management System
EcoBallast
EcoBallast
Ecochlor® Ballast Water Treatment System
Ecochlor® Ballast Water Treatment System
Ecochlor® Ballast Water Treatment System
EctoSys™
Document Title
Application for Basic Approval of the Severn Trent DeNora BalPure® Ballast
Water Management System
Application for Final Approval of the Severn Trent DeNora BalPure® Ballast
Water Management System
Application for Basic Approval of Blue Ocean Guardian (BOG) Ballast Water
Management System
Application for Basic Approval of the Blue Ocean Shield Ballast Water
Management System
Application for Basic Approval of the BlueSeas Ballast Water Management
System
Comments on the report of the fourth meeting of the GESAMP-BWWG
Application for Final Approval of a ballast water management system using
Active Substances
Application for Final Approval of the RWO Ballast Water Management System
(CleanBallast)
Report of the ninth meeting of the GESAMP -Ballast Water Working Group
(GESMP-BWWG)
Application for Basic Approval of the DESMI Ocean Guard Ballast Water
Management System
Application for Basic Approval of the FIHI Ballast Water Management System
(EcoBallast)
Application for Final Approval of FIHI Ballast Water Management System
"EcoBallast"
Application for Basic Approval of the Ecochlor® Ballast Water Treatment
System
Application for Final Approval of the Ecochlor® Ballast Water Management
System
Application for Final Approval of the Ecochlor® Ballast Water Management
System
A Swedish Disinfection System
Date
8/28/2009
3/28/2010
8/24/2009
12/5/2008
3/31/2010
2/4/2008
9/7/2007
1 1/28/2008
5/5/2009
8/19/2009
12/9/2008
8/20/2009
3/20/2008
12/16/2008
3/28/2010
1/13/2006
A-5
-------
System
EctoSys™
Electo Clean System, Clear ballast System,
CleanBallast! System
Electro-Clean Ballast Water Management
System
Electro-Clean Ballast Water Management
System
Electro-Clean Ballast Water Management
System
Electro-Clean Ballast Water Management
System
En-Ballast
ERMA FIRST
General
GloEn-Patrol, Ecochlor,
SiCURE, Resource Ballast Technologies
System
GloEn-Patrol™ Ballast Water Management
System
GloEn-Patrol™ Ballast Water Management
System
Greenship Sedinox Ballast Water Management
System
Greenship's Ballast Water Management System
HiBallast
Document Title
Basic Approval of Active Substances used by EctoSys™ electrochemical system
Report of the fourth meeting of the GESAMP -Ballast Water Working Group
(GESAMP-BWWG)
Application for Basic Approval of Active Substances used by Electro-Clean
(Electrolytic Disinfection) Ballast Water Management System
Application for Final Approval of a ballast water management system using
Active Substances (Electro-Clean Electrolytic Disinfection)
Application for Final Approval of a ballast water management system using
Active Substances (Electro-Clean Electrolytic Disinfection). Corrigendum
Application for Final Approval of the Electro-Clean System (ECS)
Application for Basic Approval of Kwang San Co., Ltd. (KS) Ballast Water
Management System "En-Ballast"
Application for Basic Approval of the ERMA FIRST Ballast Water
Management System
Guidelines on the Installation of Ballast Water Treatment Systems
Report of the tenth meeting of the GESAMP-Ballast Water Working Group
(GESMP-BWWG)
Basic Approval of Active Substance used by GloEn-Patrol™
Application for Final Approval of the GloEn-Patrol™ Ballast Water Treatment
System
Application for Final Approval of the Greenship Sedinox Ballast Water
Management System
Application for Basic Approval of a combined ballast water management system
consisting of sediment removal and an electrolytic process using seawater to
produce Active Substances (Greenship Ltd)
Application for Basic Approval of Hyundai Heavy Industries Co., Ltd. (HHI)
Ballast Water Management System (HiBallast)
Date
4/21/2006
12/19/2007
12/16/2005
9/7/2007
3/12/2008
3/20/2008
8/25/2009
3/29/2010
3/1/2010
10/30/2009
9/7/2007
12/16/2008
12/12/2008
12/20/2007
8/24/2009
A-6
-------
System
HiBallast, En-Ballast,
OceanGuard, Severn Trent DeNora
Hitachi Ballast Water Purification System
(ClearBallast)
Hitachi Ballast Water Purification System
(ClearBallast)
Hybrid Ballast Water Treatment System using
Seawater Electrolytic Process
Hybrid Ballast Water Treatment System using
Seawater Electrolytic Process, NKO3 BWTS,
PureBallast, PureBallast
Kuraray Ballast Water Management System
MES Ballast Water Management System
(FineBallast MF)
NK Ballast Water Treatment System
NK Ballast Water Treatment System
NK-O3 BlueBallast System
NK-O3 BlueBallast System
OceanGuard™ Ballast Water Management
System
OceanGuard™ Ballast Water Management
System
OceanSaver, Ecochlor,
NK-O3 BlueBallast System
OceanSaver® Ballast Water Management
System
OceanSaver® Ballast Water Management
System
Document Title
Report of the twelfth meeting of the GESAMP-Ballast Water Working Group
(GESMP-BWWG)
Application for Basic Approval of Active Substances used by Hitachi Ballast
Water Purification System (ClearBallast)
Application for Final Approval of the Hitachi Ballast Water Purification System
(ClearBallast)
Basic Approval of Active Substances used by the Hybrid Ballast Water
Treatment System using Seawater Electrolytic Process
Report of the third meeting of the GESAMP-Ballast Water Working Group
(GESMP-BWWG)
Application for Basic Approval of Kuraray Ballast Water Management System
Application for Basic Approval of the MES Ballast Water Management System
(FineBallast MF)
Request for re-evaluation of the proposal for the approval of Active Substances
Basic Approval of Active Substances used by NK Ballast Water Treatment
System
Application for Final Approval of the NK-O3 BlueBallast System (Ozone)
Application for Final Approval of the NK-O3 BlueBallast System (Ozone)
Application for Basic Approval of the OceanGuard™ Ballast Water
ManagementSystem
Application for Final Approval of the OceanGuard™ Ballast Water
ManagementSystem
Report of the seventh meeting of the GESAMP-Ballast Water Working Group
Application for Basic Approval of a ballast water management system using
Active Substances
Application for Final Approval of the OceanSaver® Ballast Water Management
System (OS BWMS)
Date
2/8/2010
9/7/2007
12/11/2008
12/14/2006
4/13/2007
3/25/2010
3/17/2010
8/18/2006
4/20/2006
3/21/2008
12/8/2008
8/26/2009
3/25/2010
7/28/2008
9/7/2007
3/19/2008
A-7
-------
System
Peraclean Ocean, ElectroClean
Peraclean® Ocean
Peraclean® Ocean
Peraclean® Ocean & Sedna system
Purimar™ Ballast Water Management System
Resource Ballast Technologies System
(cavitation combined with Ozone and Sodium
Hypochlorite treatment)
Resource Ballast Technologies System
(cavitation combined with Ozone and Sodium
Hypochlorite treatment)
Resource Ballast Technologies System, GloEn
Patrol, SEDNA using Percaclean Ocean,
OceanSaver
Siemens SiCURE
Special Pipe Ballast Water Management
System (combined with Ozone treatment), NK
Ballast Water Treatment System, EctoSys
Document Title
Report of the first meeting of the GESAMP-Ballast Water Working Group
(GESMP-BWWG)
Application for approval of an Active Substance for Ballast Water Management
Application for approval of an Active Substance for Ballast Water Management.
Corrigendum
Application for Final Approval of a ballast water management system using
Active Substances
Application for Basic Approval of Techwin Eco Co., Ltd. (TWECO) Ballast
Water Management System (Purimar™)
Basic Approval of Active Substances used by Resource Ballast Technologies
System (Cavitation combined with Ozone and Sodium Hypochlorite treatment)
Application for Final Approval of the Resource Ballast Technologies System
(Cavitation combined with Ozone and Sodium Hypochlorite treatment)
Report of the fifth meeting of the GESAMP-Ballast Water Working Group
(GESMP-BWWG)
Application for Basic Approval of the Siemens SiCURE Ballast Water
Management System
Report of the second meeting of the GESAMP-Ballast Water Working Group
(GESMP-BWWG)
Date
2/28/2006
4/15/2005
5/27/2005
9/7/2007
3/9/2010
4/6/2007
12/19/2008
1/25/2008
12/19/2008
7/7/2006
A-8
-------
System
Special Pipe Hybrid Ballast Water
Management System (with Ozone),
CleanBallast,
NK-O3 BlueBallast System, Blue Ocean
Shield, EcoBallast
Special Pipe Hybrid Ballast Water
Management System combined with Ozone
treatment version
Special Pipe Hybrid Ballast Water
Management System combined with Ozone
treatment version
Special Pipe Hybrid Ballast Water
Management System combined with Ozone
treatment version
Special Pipe Hybrid Ballast Water
Management System combined with
PERACLEAN ® Ocean (SPO-SYSTEM)
Sunrui ballast water management system
TG Ballastcleaner and TG Environmentalguard
TG Ballastcleaner and TG Environmentalguard
TG Ballastcleaner and TG
Environmentalguard, Greenship's Ballast
Water Management System, Electro-Clean
System (ECS)
Document Title
Report of the eighth meeting of the GESAMP-Ballast Water Working Group
(GESMP-BWWG)
Basic Approval of Active Substances used by Special Pipe Ballast Water
Management System (combined with Ozone treatment)
Application for Final Approval of the Special Pipe Hybrid Ballast Water
Management System (combined with Ozone treatment)
Application for Final Approval of the Special Pipe Hybrid Ballast Water
Management System combined with Ozone treatment version (SP-Hybrid
BWMS Ozone version)
Application for Final Approval of the Special Pipe Hybrid Ballast Water
Management System combined with PERACLEAN ® Ocean (SPO-SYSTEM)
Application for Basic Approval of Sunrui ballast water management system
Application for Basic Approval of the ballast water management system using
"TG Ballastcleaner and TG Environmentalguard" as Active Substances
(Toagosei Group)
Application for Final Approval of the JFE Ballast Water Management System
(JFE-BWMS) that makes use of "TG Ballastcleaner® and TG
Environmentalguard®"
Report of the sixth meeting of the GESAMP-Ballast Water Working Group
Date
4/8/2009
4/12/2006
12/4/2008
3/17/2010
3/29/2010
8/24/2009
12/26/2007
8/20/2009
7/14/2008
A-9
-------
APPENDIX B: LITERATURE REVIEW OF RECEPTION FACILITY STUDIES
The Panel identified and evaluated a number of studies in the published and gray literature that discuss
reception facilities (RF) (Table B-l).
Five studies have compared the effectiveness or costs of RF and shipboard treatment. In a study for the
Canadian Coast Guard, Pollutech (1992) scored and ranked a variety of ballast water management
approaches for vessels entering the Great Lakes, including ballast water exchange and several shipboard
and land-based treatments, in terms of effectiveness, feasibility, maintenance and operations,
environmental acceptability, cost, safety and monitoring. RF with discharge to a sanitary sewer (the only
RF treatment scenario analyzed) ranked second out of 24 treatment and management approaches
analyzed in the report.
In a second study for the Canadian Coast Guard, Aquatic Sciences (1996) considered RF alternatives
(referred to as "pump off options") for Great Lakes shipping and found them to be "technically feasible"
and to "undoubtedly offer the best assurance of prevention of unwanted introductions." The report
further found that when installed onshore, "treatment options could have a more practical and
enforceable application" than in shipboard installations, and concluded that "ship board treatment of
ballast water appears to be logistically, economically, and particularly from the aspect of control, the
least attractive method of ballast water treatment." The report estimated that treatment ships could be
provided at key ports throughout the Great Lakes to receive discharged ballast water and heat it to
>65°C at an annualized cost of around $17 to $51 million, or alternatively a single treatment ship could
operate at a site en route to the Great Lakes to treat all incoming ballast water at an annualized cost of
$2.7-2.8 million. Retrofitting costs to enable ships to discharge their ballast water to treatment ships
were estimated at approximately $40,000 to over $200,000 per ship.
AQIS (1993a) developed designs and cost estimates to compare shipboard, land-based and treatment
ship-based treatment at a port serving 140,000-ton bulk carriers. The shipboard design consisted of a 50-
|im strainer, with high-level ballast tank off-take pipes to reduce the discharge of ballast sediments and
settled cysts or spore stages. The land-based designs included either 4,000 or 52,000 MT of storage, with
coagulation, flocculation, granular filtration, UV disinfection, and thickening, dewatering and disposal
of solids. The treatment ship design included 4,000 MT of storage, pressurized granular filters, UV, and
solids management and disposal. Annualized costs were reported as $0.69/MT for shipboard treatment,
$0.55/MT for treatment in a treatment ship, and $0.35-$0.62 for treatment in a land-based facility
(depending on the type and size of storage used).13 Some costs (pipelines to transport ballast water from
berths to treatment plants, and land costs) were not included in the RF alternatives, which reduced their
estimated cost relative to the shipboard alternative. On the other hand, the Panel notes that the RF
treatment analyzed here (granular filtration with coagulation and flocculation, followed by UV
disinfection) would treat ballast water to a substantially higher standard than the shipboard alternative (a
50 |im strainer with no disinfection); and that basing the analysis on large bulk carriers, which typically
discharge the largest volumes of ballast water of the vessels using Australia's ports (Table 4.1 in AQIS
1993a), favored shipboard treatment.
13 Unless stated otherwise, the cost estimates cited in this appendix were converted from foreign currencies in the original
publications into US dollars at the daily average interbank transfer rates reported at
http://www.oanda.com/currency/historical-rates on the date of publication or presentation, or on the first day of the month
where only the month of publication was given, and adjusted for inflation from the date of original publication to June 1,
2010 using the calculator at http://inflationdata.com/inflation/Inflation_Calculators/InflationCalculator.asp, which is based on
the U.S. Bureau of Labor Statistics' Consumer Price Index for all Urban Consumers (CPI-U).
B-l
-------
Table B-l. Reports that discuss reception facilities (RFs) for onshore treatment of ballast water.
Report
Polluted! 1992
AQIS1993a
AQIS1993b
Ogilvie 1995
Aquatic Sciences 1996
NRC 1996
Gauthier & Steel 1996
Victoria ENRC 1997
Greenman et al. 1 997
Cohen 1998
Reeves 1998, 1999
Oemkel999
Dames & Moore 1998,
1999
Cohen & Foster 2000
CAPA 2000
Rigby & Taylor
2001a,b
US EPA 2001
California SWRCB
2002
Glosten 2002
NSF 2003
Milliard 2006; Milliard
&Matheickal2010
Brown & Caldwell
2007, 2008
California SLC 2009,
2010
Pereira et al. 2010
Dormer 20 1 Oa,b,c
Discussion
Compares and ranks various shipboard and RF
treatment approaches.
Compares shipboard, land-based and treatment ship
approaches in Australia.
Briefly discusses treatment ship and land-based
treatment in Australia.
Reviews possible treatment methods and estimates
some costs for RFs.
Compares shipboard, treatment ship, land-based and
external source treatment.
Briefly discusses advantages and disadvantages of RF.
Mentions shipboard, treatment ship and land-based
approaches.
Briefly discusses RFs.
Student report commissioned by the U.S. Coast Guard.
Briefly discusses advantages and disadvantages of
RFs.
Briefly discusses RFs.
Briefly discusses advantages and disadvantages of RF.
Briefly discusses RF.
Briefly discusses advantages and disadvantages of RF.
EPA-funded study estimates the cost of RF for
California.
Briefly discusses RF.
Briefly mentions RF.
Briefly discusses RF.
Estimates upper-bound retrofit costs to discharge
ballast to RFs.
Mentions shipboard, RF and operational options for
the longer term.
Compares and ranks various shipboard and RF
treatment approaches for the Black Sea-Caspian Sea
Waterway.
Develops designs and estimates costs for RF at the
Port of Milwaukee.
Briefly discusses advantages and disadvantages of RF.
Uses simulation model to assess RF operation at a
Brazilian port.
Compares RF and shipboard treatment.
Conclusions
RF ranks 2nd out of 24 options.
Land-based and treatment ship are cheaper and more
effective than shipboard.
RF is unlikely except in special circumstances.
Several methods show promise for RF.
RF is technically feasible and the most effective and
cheapest approach.
RF remains an option.
RF is considered a poor option.
RF is probably too costly at a large scale; may be
viable at a smaller scale.
RF is feasible at all sites considered.
Lists RF as an alternative.
RF is feasible for some parts of the industry, such as
VLCCs.
RF may be good option at oil export terminals with oil
stripping plants.
RF is technically feasible.
Cost, availability, quality control may prevent RF
development, but it might work for tankers that
discharge oily ballast to RFs.
RF is an attractive option, at least for some parts of the
industry.
Shipboard seems the most challenging approach.
RF ranks 1st out of 16 options.
RF is feasible; treatment ship is cheaper than land-
based.
RF might be suitable for terminals with regular vessel
calls such as cruise ships, or for the Port of Milwaukee.
RF treatment would not affect port operations
negatively.
RF is more efficient, cheaper and safer.
B-2
-------
AQIS (1993a) also developed a scenario for RF treatment of all the ballast water discharged in
Australia, using both treatment ships and land-based treatment plants. Total capital costs for RF
treatment were estimated at $330 million and annual operating costs at $6.7 million. The capital cost for
outfitting one-year's worth of visiting ships for shipboard treatment was estimated at $1 billion
"ignoring the fit out of new ships in future years," with estimated annual operating costs working out to
$5.4 million. The study concluded that "land-based or port-based [=treatment ship] facilities are more
economic and effective than numerous ship-board plants."
California's State Water Resources Control Board (California SWRCB 2002) qualitatively evaluated
RFs and ten shipboard treatment alternatives for effectiveness, safety, and environmental acceptability.
RF was the only approach rated acceptable in all three categories. There were reservations or unresolved
questions about the effectiveness of all shipboard alternatives, about the safety of 80 percent of the
shipboard alternatives, and about the environmental acceptability of 90 percent of the shipboard
alternatives.
Hilliard (2006) (also reported in Hilliard and Matheickal 2010) compared an RF using conventional
water treatment methods (such as granular filtration with disinfection) to 15 shipboard treatment
approaches for vessels transiting the Black Sea-Caspian Sea Waterway. Based on scores for 13 technical
factors, RF treatment was ranked first. The study concluded that an RF, using standard water industry
methods, would provide a cost effective solution if based at an appropriate port, but cautioned that this
might be a less useful approach for some vessels.
In each of these comparative studies, RF was judged to be as effective or more effective, and generally
cheaper, than shipboard treatment. As noted, there are limitations to these studies and grounds for
criticism; in particular, some were done over a decade ago and do not reflect current BWMS costs.
However, these studies comprise the most detailed published comparisons of RF and shipboard
treatment approaches available. In addition, the U.S. Coast Guard compiled a table of cost estimates
from different studies (U.S. Coast Guard 2002). Figure B-l shows all the estimates that were expressed
as costs per metric ton or cubic meter of ballast water, and thus in a form that can be compared. In these
estimates, RF treatment is generally more expensive than ballast water exchange and less expensive than
shipboard treatment, though there is considerable overlap. These cost estimates also do not reflect
current BWMS costs.
B-3
-------
1000
100
Figure B-l. Cost estimates listed in U.S. Coast Guard (2002). The Coast Guard converted Australian estimates to U.S.
dollars at the Oct. 16, 2001 exchange rate, but did not adjust estimates for inflation. One m3 of ballast water is assumed to
weigh 1 MT. Cost estimates (note log scale) for ballast water exchange are in blue, for RF treatment in green, and for
shipboard treatment in red.
In three recent papers, Donner (2010a,b,c) argued that RF treatment is more efficient, less expensive,
safer for the crew and the environment, and easier to monitor and verify than shipboard treatment, but
that shipboard treatment has been pursued because it is "the solution of least resistance" politically.
Other comparisons of RF and shipboard treatment in the literature consist of lists or brief discussions of
their relative merits. These reports variously concluded that RF treatment is probably a superior or
probably an inferior option compared to shipboard treatment, or that RF treatment is suitable for a
particular part of the cargo fleet (Table B-l), but none provided analysis or data to support these
conclusions.
Two studies (in addition to AQIS (1993a) and Aquatic Sciences (1996), discussed above) provided
conceptual designs and cost estimates for RF treatment for specific regions. CAPA (2000), an EPA-
funded study conducted for the California Association of Port Authorities, developed conceptual designs
and cost estimates for constructing and operating ballast water treatment plants at cargo ports in
California (Table B-2). These plans and estimates included pipelines from berths to plants; storage
tanks; coagulation, flocculation, filtration and UV disinfection; thickening, dewatering and landfill
disposal of residual solids; and discharge of effluent through an outfall pipeline. They did not estimate
costs for land, permits, seismic evaluation, or retrofitting vessels to enable them to discharge ballast
B-4
-------
water to an RF. The study concluded that onshore treatment would be technically and operationally
feasible, though there could be delays to vessels in some circumstances.
Table B-2. Cost estimates for onshore treatment in California (Source: CAPA 2000).
System Component
Pipelines
Storage Tanks
Treatment Plants
Outfalls
Total
Capital Costs
146,950,000
76,235,000
22,510,000
1,380,000
247,075,000
Annual O&M
-
-
2,018,000
-
2,018,000
Brown & Caldwell (2007, 2008) developed designs and cost estimates for land-based and treatment ship
approaches to treat ballast discharges from oceangoing ships arriving at the Port of Milwaukee. The first
report assessed four land-based treatment systems:
• 100-|im screening followed by UV treatment;
Coarse screening followed by ozonation;
• 500-|im screening followed by membrane filtration to remove particles > 0.1 |im;
• 500-|im screening followed by hydrodynamic cavitation.
These were analyzed along with two systems for transferring and storing the discharged ballast water:
discharge at berths into pipelines to land-based RF with storage tanks; and discharge to a barge that
would store and carry the water to a land-based RF. Design criteria required a system capable of
receiving ballast water at 680 MT/h, storage capacity of 1,900 MT, and treatment at 80 MT/h. The
report concluded that all four treatment systems and both transport/storage systems are feasible, with
UV treatment and hydrodynamic cavitation having the most promise for treating viruses (Brown &
Caldwell 2007). The second report (Brown & Caldwell 2008) developed a design and cost estimate for
retrofitting a barge to serve as a treatment ship, which would collect, store and treat ballast water.
Treatment included a cloth media disk filter with a nominal pore size of 10 jim, and UV disinfection at
an estimated minimum dose of 30 mJ/cm2. The design criteria for this analysis included the capacity to
receive ballast discharges at 2,300 MT/h, storage of 10,000 MT, and treatment at 230 MT/h, which is
around three times the flow rates and five times the storage required in the first report. Estimated costs
from both studies are shown in Table B-3.
B-5
-------
Table B-3. Cost estimates for onshore treatment for oceangoing ships at the Port of Milwaukee
(Source: Brown & Caldwell 2007,2008).
Treatment [1]
Transport
— — — Capital Costs — — — . ,
r Annual
Pipelines [4] Storage Treatment Total O&M
100-um screening & UV [1]
Ozone [1]
0.1-um membrane filter [1]
Hydrodynamic cavitation [1]
100-um screening & UV [1,2]
Ozone [1,2]
0.1-um membrane filter [1,2]
Hydrodynamic cavitation [1,2]
10-um filter &UV [3]
Pipelines
Pipelines
Pipelines
Pipelines
Barge
Barge
Barge
Barge
Treatment ship
2,973,000 1,252,000
2,973,000 1,252,000
2,973,000 1,252,000
2,973,000 1,252,000
261,000 522,000
261,000 522,000
615,000 4,840,000 11,500
835,000 5,060,000 9,800
1,096,000 5,321,000 15,600
2,608,000 6,833,000 20,900
615,000 1,398,000 367,000
835,000 1,617,000 365,000
261,000
261,000
522,000
522,000
1,096,000
2,608,000
1,879,000
3,390,000
376,000
0 2,695,000
866,000 3,561,000 514,000
[1] Design criteria are maximum ballast discharge of 680 MT/h, 1,900 MT storage, and treatment rate of 80 MT/h.
[2] "Storage" refers to barge purchase and modification to use for ballast water transfer and storage, exclusive of the
treatment system.
[3] Design criteria are maximum ballast discharge of 2,300 MT/h, 10,000 MT storage, and treatment rate of 230
MT/h.
[4] Includes collection pumps, pipelines, a lift station and coarse screening.
Besides the need for facilities to receive, store and treat ballast water from ships, ships must be modified
so they can safely and rapidly discharge ballast water to RFs. This requires modification of a ship's pipe
system and possibly larger ballast pumps, to raise the water to deck level and/or to discharge it quickly
enough. Cost estimates have ranged from around $15,000 to $540,000 for container ships (Pollutech
1992; Glosten 2002), $15,000 to $500,000 for bulkers (Pollutech 1992; CAPA 2000), and less than
$140,000 to around $2.3 million for tankers (Victoria ENRC 1997; Glosten 2002) (Table B-4). Most of
these estimates explicitly included the replacement of existing pumps with more powerful pumps where
needed (AQIS 1993a; Aquatic Sciences 1996; Dames & Moore 1998; CAPA 2000; Glosten 2002;
Brown & Caldwell 2008 4). The cost to outfit a new ship was estimated to be less than the cost to
retrofit an existing ship (AQIS 1993b), perhaps by an order of magnitude (CAPA 2000). Some reports
provided little or no explanation of their retrofit/modification estimates (Pollutech 1992; AQIS 1993 a;
Aquatic Sciences 1996; Dames & Moore 1998). Victoria ENRC (1997) provided a materials list for a
bulk carrier, and noted that a tanker "with its ballast lines running on deck would have a considerable
lower installation cost." CAPA (2000) provided a cost breakdown for modifying a bulker, and stated
that modifying a tanker would generally cost more.
14 Glosten (2002) designed the pumps and pipes to be large enough to enable ships to deballast completely at berth during a
typical cargo loading period. Brown & Caldwell (2008) found, based on dynamic head vs. flow curves, that Great Lakes
bulkers would not need larger ballast pumps—that is, with their existing pumps the ships could fully deballast while at berth
during the time it takes to load cargo.
B-6
-------
Table B-4. Cost estimates for retrofitting ships to discharge ballast water to a treatment facility. Where the
data are available, length is given in feet, size in deadweight tons (DWT), ballast water capacity in metric tons (MT),
and maximum ballast discharge rate in metric tons per hour (MT/h), in parentheses following the ship type.
Ship Type
Great Lakes bulker, break-bulk or container
Small container
Large bulker (140,000 DWT; 45,000 MT; 4,000 MT/h)
Great Lakes bulker
Handysize bulker (520'; 22,000 DWT)
Container
Container or bulker (1,000 MT/h)
Tanker (869'; 123,000 DWT; 75,850 MT; 6,400 MT/h)
Bulker (735'; 67,550 DWT; 35,000 MT; 2,600 MT/h)
Break-bulk (644'; 40,300 DWT; 26,850 MT; 3,000 MT/h)
Container (906'; 65,480 DWT; 19,670 MT; 2,000 MT/h)
Car carrier (570'; 13,847 DWT; 6,600 MT; 550 MT/h)
Bulker (469'; 5,700 MT; 570 MT/h)
Bulker (722'; 18,000 MT; 2,300 MT/h)
[1] Estimate developed by the Pacific Merchant Shipping
Capital Cost
$13,200-26,500
$20,400
$204,000
$40,400-202,00
$142,000
$53,200-173,000
$502,000
$2,3230,000
$131,000
$373,000
$540,000
$198,000
$60,000
$203,000
Association.
Report
Polluted! 1992
AQIS 1993a
AQIS 1993a
Aquatic Sciences 1996
Victoria ENRC 1997
Dames & Moore 1998 [1]
CAPA 2000
Glosten 2002
Glosten 2002
Glosten 2002
Glosten 2002
Glosten 2002
Brown & Caldwell 2008
Brown & Caldwell 2008
Glosten (2002) and Brown & Caldwell (2008) provided the most detailed estimates. Glosten (2002)
estimated ship retrofit/modification costs for five ships representing common vessel types in Puget
Sound (Table B-4). The modifications were designed to "allow ballast transfer with minimal disruption
to current operations" including sizing them for deballasting completely at berth during the time needed
to load cargo, eliminating any need to start deballasting before arriving at berth. For each vessel type,
the authors selected ships that "had ballast systems with capacities on the upper end of vessels that call
on Puget Sound to attempt to establish an upper-bound on retrofitting costs." In selecting pipe sizes and
other elements "every attempt was made to capture an upper bound on the modification costs associated
with each vessel type surveyed," including the installation of "a completely new piping system to
provide the ability to fill and empty each ballast tank separately." Notably, this new piping system was
included in the tanker estimate even though it is not needed on crude oil tankers, the type of tanker
analyzed, where "a simpler, lower-cost solution" exists. It was included because it could be needed on
some other ships (i.e. product tankers) in the same general category, and this produced by far the highest
cost estimate in the study.15 The modifications were also designed to allow ballast water transfer in
either direction between a ship and a RF (either onto or off a ship),16 which in some cases may raise the
cost over what is needed to only discharge ballast water to RFs.
Brown & Caldwell (2008) provided analyses, conceptual designs, drawings and cost estimates for
modifying two sizes of ocean-going bulkers serving the Great Lakes, based on a smaller actual ship and
a larger hypothetical ship (Table B-4). These designs were also sized to allow the ship to initiate and
complete deballasting at berth during cargo loading.
15 Consistent with the study's aim of quantifying "the capital cost required to provide the maximum capability in a ballast
transfer system, to represent a maximum capital investment" for each vessel category (Glosten 2002).
16 This ability was included to accommodate the possibility of loading "clean" ballast, an approach that is not considered to
be onshore treatment in this report.
B-7
-------
The potential for treating ballast discharges with RFs has also been recognized in laws, regulations,
guidelines and treaty conventions. The U.S. Nonindigenous Aquatic Nuisance Prevention and Control
Act (NANPCA) of 1990 and the National Invasive Species Act (NTSA) of 1996 directed the U.S. Coast
Guard to fund research on ballast water management, specifically noting that technologies in "land-
based ballast water treatment facilities" could be included, and to investigate the feasibility of using or
modifying onshore ballast water treatment facilities used by Alaskan oil tankers to reduce the
introduction of exotic organisms (§§1101(k)(3), 1104(a)(l)(B), 1104(a)(2) and 1104(b)(3)(A)(ii) in U.S.
Congress 1990, 1996). In the interim and final rules implementing NTS A, the U.S. Coast Guard
specifically included discharge to a RF as a means of meeting NISA's ballast discharge requirements,
and required ships to keep records of ballast water discharged to RFs (US Coast Guard 1999, 2001),
although the Coast Guard eliminated these provisions when it concluded that it did not have the
authority to regulate or approve RFs (US Coast Guard 2004). The U.N. International Maritime
Organization's 1991 Guidelines stated that "Where adequate shore reception facilities exist, discharge of
ship's ballast water in port into such facilities may provide an acceptable means of control" (EVIO 1991
and EVIO 1993, §7.5 Shore Reception Facilities). The EVIO's 1997 Guidelines stated that "Discharge of
ship's ballast water into port reception and/or treatment facilities may provide an acceptable means of
control. Port State authorities wishing to utilize this strategy should ensure that the facilities are
adequate... If reception facilities for ballast water and/or sediments are provided by a port State, they
should, where appropriate, be utilized" (EVIO 1997, §7.2.2, §9.2.3). The EVIO's 2004 Convention stated
that "The requirements of this regulation do not apply to ships that discharge ballast water to a reception
facility designed taking into account the Guidelines developed by the Organization for such facilities"
(EVIO 2004, Regulation B-3.6). The EVIO adopted specific guidelines for RFs (EVIO 2006), and
recognized RFs as an alternative in EVIO 2005b (§1.2.3), as have Australia, New Zealand and Canada in
their ballast water regulations (AQIS 1992; New Zealand 1998, 2005; Canada 2000, 2007).
Additional Literature Cited
Australian Quarantine and Inspection Service (AQIS). 1992. Controls on the Discharge of Ballast Water
and Sediment from Ships Entering Australia from Overseas. AQIS Notice (Barrier Co-ordination) 92/2.
AQIS, Canberra, Australia.
Canadian Marine Advisory Council (CMAC). 2000. Guidelines for the Control of Ballast Water
Discharge from Ships in Waters under Canadian Jurisdiction. TP 13617, CMAC, Ottawa, Canada.
Canadian Marine Advisory Council (CMAC). 2007. A Guide to Canada's Ballast Water Control and
Management Regulations. TP 13617E, Transports Canada, Ottawa, Canada.
Donner, P. 2010a. Ballast Water Treatment Ashore Brings More Benefits. Pages 97-105 in: Emerging
Ballast Water Management Systems. Proceedings of the EVIO-WMU Research and Development Forum,
26-29 January 2010, Malmo, Sweden. A GloBallast-Global Industry Alliance and World Maritime
University Initiative. Bellefontaine, N., F. Haag, O. Linden and J. Matheickal (eds.). WMU Publication,
Malmo, Sweden.
Donner, P. 2010b. Ballast water treatment ashore—better for the environment and for seafarers. WMU
Journal of Maritime Affairs 9(2): 191-199.
-------
Donner, P. 2010c. Is there a case for shore-based ballast water treatment facilities? 5th International
Conference on Ballast Water Management 2010. Singapore, Nov. 1-4, 2010. Session 1: Ballast Water
Management Plans by Ports and Flag States.
Greenman, D., K. Mullen and S. Parmar. 1997. Ballast Water Treatment Systems: A Feasibility Study.
Produced in cooperation with ENS Chris Freise, U.S. Coast Guard Office of Response. Commissioned
by the U.S. Coast Guard and Worcester Polytechnic Institute. Submitted to Professor Matthew Ward,
Project Center, Worcester Polytechnic Institute, Washington, DC, 59 pp.
Hilliard, R.W. 2006. Assessment of Shipping Traffic and Ballast Water Movements to and from Caspian
Sea, and Preliminary Appraisal of Possible Ballast Water Management Options. Draft Report. Prepared
for the International Maritime Organization. URS Corporation. Accessed at:
http://www.caspianenvironment.org/Newsite/Activities-Ballast%20Waters.htm
Hilliard R.W. and J.T. Matheickal. 2010. Alternative Ballast Water Management Options for Caspian
Region Shipping: Outcomes of a Recent CEP/EVIO/UNOPS Project. Pages 119-138 in: Emerging
Ballast Water Management Systems. Proceedings of the EVIO-WMU Research and Development Forum,
26-29 January 2010, Malmo, Sweden. A GloBallast-Global Industry Alliance and World Maritime
University Initiative. Bellefontaine, N., F. Haag, O. Linden and J. Matheickal (eds.). WMU Publication,
Malmo, Sweden.
International Maritime Organization (EVIO). 1991. International Guidelines for Preventing the
Introduction of Unwanted Aquatic Organisms and Pathogens from Ships' Ballast Water and Sediment
Discharges. EVIO, Marine Environment Protection Committee, Resolution MEPC 50(31), adopted 4 July
1991.
International Maritime Organization (EVIO). 1993. Guidelines for Preventing the Introduction of
Unwanted Aquatic Organisms and Pathogens from Ships' Ballast Water and Sediment Discharges. EVIO,
Resolution A.774(18), adopted 4 November 1993.
International Maritime Organization (EVIO). 1997. Guidelines for the Control and Management of Ships'
Ballast Water to Minimize the Transfer of Harmful Aquatic Organisms and Pathogens. EVIO, 20th
Assembly, Resolution A.868(20), adopted 27 November 1997.
International Maritime Organization (EVIO). 2006. Guidelines for Ballast Water Reception Facilities
(G5). EVIO, Marine Environment Protection Committee, Marine Environment Protection Committee
(MEPC), Resolution MEPC 153(55), adopted 13 October 2006.
National Science Foundation (NSF) (2QQ?>) Engineering Controls for Ballast Water Discharge:
Developing Research Needs. Report of a workshop held on April 28-30, Seattle, WA. Bioengineering
and Environmental Systems Division of the Engineering Directorate, NSF, Arlington, VA.
New Zealand Ministry of Agriculture and Forestry (NZMAF). 1998. Import Health Standard For Ships'
Ballast Water From All Countries (Biosecurity Act 1993). NZ MAP, Wellington, New Zealand.
New Zealand Ministry of Agriculture and Forestry (NZMAF). 2005. Import Health Standard For Ships'
Ballast Water From All Countries. Issued pursuant to Section 22 of the Biosecurity Act 1993. NZ MAP,
Wellington, New Zealand.
B-9
-------
Ogilvie, DJ. 1995. Land-based treatment options. Pages 113-123 in: Ballast Water. A Marine Cocktail
on the Move. Proceedings of the National Symposium, 27-29 June 1995, Wellington, New Zealand. The
Royal Society of New Zealand. Miscellaneous Series 30.
Pereira, N.N., R.C Better, H.L. Brinati and E.F. Trevis. 2010. A Study of Ballast Water Treatment
Applied on Iron Ore Ports in Brazil Using Discrete Simulation. Pages 77-91 in: Emerging Ballast Water
Management Systems. Proceedings of the IMO-WMU Research and Development Forum, 26-29
January 2010, Malmo, Sweden. A GloBallast-Global Industry Alliance and World Maritime University
Initiative. Bellefontaine, N., F. Haag, O. Linden and J. Matheickal (eds.). WMU Publication, Malmo,
Sweden.
B-10
-------
APPENDIX C: FURTHER INFORMATION ON STATISTICS AND
INTERPRETATION
A major challenge of sampling at low organism concentrations is that many samples will have zero live
organisms because the few live organisms present are missed. Therefore to improve the probability of
detecting them, large volumes must be sampled and excellent techniques must be used to enable
detection (Figure C-l).
Consider the following examples from Lee et al. (2010): from the Poisson distribution, if 1 m3 of ballast
water was sampled from a ballast water discharge that had a known average concentration of 10
zooplankton-sized organisms m"3 over the entire ballast water volume, about 95 percent of the samples
would contain 4-17 organisms m"3. As the concentration of organisms decreases, the frequency
distribution becomes increasingly skewed, and there is a high probability of obtaining a sample with
zero organisms. Thus, if the sample concentration is 1 organism m"3, the probability of a 1 m3 sample
containing zero organisms is 36.8 percent. If the sample concentration is only 0.01 organism m"3, or 1
organism in 100 cubic meters of ballast water, the probability of obtaining a sample with zero organisms
is -99 percent. Moreover,
If a small volume is used to evaluate whether the discharge meets a standard, the sample
may contain zero detectable organisms, but the true concentration of organisms may be
quite high... .For example, even with a relatively high concentration of 100 organisms m"
3, only about 10% of 1-L samples will contain one or more organisms. Furthermore, even
if zero organisms are detected in a 1-L sample, the upper possible concentration, based on
a 95% confidence interval, is about 3,000 organisms m"3... .The general point is that more
organisms may be released in ballast discharge using a stringent standard paired with a
poor sampling protocol than a more lenient standard paired with a stringent sampling
protocol. (Lee et al. 2010, p.72).
The available methodologies for testing compliance with the EVIO standards for zooplankton-sized
organisms are at or near the analytic detection limits. The following example from the ETV Protocol
(U.S. EPA 2010) illustrates the problem: For the desired minimum precision in quantifying
zooplankton-sized organisms, consider an example where the upper bound of the Chi-square statistic
should not exceed twice the observed mean (corresponding to a coefficient of variation of 40 percent,
which is relatively high). Then, if 6 or fewer live organisms are counted, the upper bound of the 95% CI
for the volume sampled does not exceed the EVIO/Phase-1 performance standard for zooplankton-sized
organisms (< 10 viable individuals per m3):
• Coefficient of variation (CV) = standard deviation (SD) divided by the mean (M).
r\
• For the Poisson distribution, the variance (V) = SD = M.
• Substituting the critical value of the mean, 6: CV = 61/2/6 ~ 40%.
The volume needed to find and quantify 6 live organisms per m3 depends on the whole-water
sample volume, the concentration factor, and the number of subsamples examined. Very large
sample volumes (tens of m3) are required to quantify viable zooplankton-sized organisms
(assuming 20 mL of the concentrated sample is analyzed), and each sample must be concentrated
down to a manageable volume (concentrating 3 m3 to 1 L would yield a concentration factor of
3,000). Based on the Poisson distribution for a 95% CI from the Chi-square distribution, 30 m3
C-l
-------
(30,000 L) must be sampled in order to find and count < 10 organisms m"3 with the desired level
of precision. The total sample volume can be reduced if the concentration factor is increased (and
the same subsample volume analyzed), if the CI is also lowered (e.g., from 95% to 90%) or the
subsample volume analyzed is increased (e.g., from 20 mL to 40 mL). Notably, as the
concentration factor increases, the likelihood of losing organisms or inadvertently killing them in
the sample processing steps increases, thus creating an artifact that overstates the effectiveness of
the treatment (see Section 6.2.4 (A)).
The ETV Protocol provides examples of the sample size needed to provide the level of precision needed
to achieve a 95% upper confidence limit that is no more than twice the observed mean and does not
exceed the targeted concentration. If the volume of subsample that is analyzed is increased, then
validation experiments should be conducted to ensure that counting accuracy is acceptably high. The
problem is exacerbated for zooplankton-sized organisms because they are sparse compared to organisms
in the next smaller class (here, referred to as "protist-sized" organisms, or organisms >10 |im and < 50
|im in minimum dimension). The Poisson distribution assumption still applies to this smaller size class,
and the ETV Protocol provides examples with a more stringent level of precision than is used for the
larger size class (Table C-2; U.S. EPA 2010). At present, confirmation of the Phase 1 standard (< 10
protist-sized organisms mL"1) represents the practical limit that can currently be achieved by testing
facilities in the U.S. (e.g., MERC 2009, 2010a, 2010b; Great Ships Initiative 2010).
Table C-l. Sample volume of treated ballast water required relative to performance standards for
organisms > 10 um and <50 um (nominally protists), assuming that the desired level of precision is set at a
CV of < 10%. These are the required whole-water sample volumes that must be concentrated to 1 L as a
function of N, the number of 1-mL subsamples analyzed. (Source: U.S. EPA 2010).
N= 2 3 4
Concentration (i.e. performance Sample Volume Required (L)
standard] (individuals ml/1)
0.01
0.1
1
10
6,000
600
60
6
4,000
400
40
4
3,000
300
30
3
Laboratory experiments withprotist cultures support use of the Poisson distribution
A workshop was held to evaluate four methods for enumerating living protists in treated ballast water
(Nelson et al. 2009, Steinberg et al., accepted with revisions). Live and dead cells were counted using
flow cytometry, an enhanced flow-through system with imaging capacity (FlowCAM®, Fluid Imaging
Technologies, Yarmouth, ME), direct counts of samples collected on membrane filters, and direct counts
using a Sedgewick Rafter counting chamber. All techniques used fluorescent stains to differentiate
between live and dead cells. Counting methods were tested with several ratios and densities of live and
dead Tetraselmis sp., a small phytoflagellate. In these trials, comparisons were conducted under ideal
conditions with no debris (except for one sample) or particulate matter and with a single target species.
Data were evaluated to determine whether they conformed to a Poisson distribution by determining if
the variance was equal to the mean. At low concentrations of living cells (approximately 10 mL"1 to 100
mL"1), there was no evidence to reject the Poisson hypothesis (Nelson et al. 2009).
C-2
-------
Accuracy and precision in sparse samples following a Poisson distribution
A series of laboratory experiments was conducted to assess the accuracy and precision of enumerating
zooplankton- and protist-sized organisms at a variety of densities (Lemieux et al. 2008). Inert 10-|im
standardized microbeads at densities of 1, 5, 10, 50, 100, 500, and 1,000 microbeads per mL of artificial
seawater represented protist-sized organisms, and 150-|im microbeads at 10, 30, and 60 microbeads per
500 mL represented zooplankton-sized organisms. Such inert, standardized polymer microbeads were
used rather than organisms to eliminate any potential bias, and artificial (rather than natural) seawater
was used to avoid inclusion of various organic particles (e.g., detritus) that could interact with the
microbeads and confound interpretations. Here, microbeads served as proxies, and it is acknowledged
they are an imperfect representation of living organisms (e.g., microbeads do not exhibit the swimming
behavior of many planktonic organisms; living organisms can cling to nets or filters and can also be
squeezed through a net more readily than microbeads). Nonetheless, if the items of interest in a sparse
concentration - be they microbeads or living organisms in treated ballast water - are well mixed, the
Poisson distribution should be applicable. In addition, as stated previously, a continuously isokinetically
taken sample whose contents are completely counted avoids problems associated with the spatial
distribution of the organisms.
At each microbead density, the percent difference of the observed mean from the expected mean
indicated counting accuracy, and the CV indicated the level of precision. In this study, benchmarks for
acceptable accuracy and precision were established at a percent difference of 10 percent and a CV of 0.2
(20 percent), respectively. For the "protist" microbeads, the 50 - 1,000 mL"1 concentrations were not
significantly different, with acceptable accuracy and precision below the 10 percent and 20 percent
benchmarks, respectively. Unfortunately, however, analysis of the "zooplankton" microbead populations
at all densities showed poor precision, with CVs well above 20 percent, and only counts at the highest
density showed a CV < 100 percent. All densities of "zooplankton" microbeads showed acceptable
accuracy (i.e., a percent difference < 10%) after sufficient aliquots were examined to result in a stable
mean.
From this work, Lemieux et al. (2008) recommended that samples for analysis of protist-sized organisms
should be concentrated by at least a factor of five, and that at least four replicate counting chambers
(e.g., four Sedgewick Rafter slides) should be analyzed for acceptable accuracy and precision, including
evaluation of at least 10 random rows (from a total of 20) of each counting chamber. Importantly, for
zooplankton-sized organisms, Lemieux et al. (2008) determined the earlier (draft) ETV protocol
recommendations for sample sizes as inadequate to achieve acceptable precision. The data from these
microbead experiments indicated, instead, that this size class requires a sample size of greater than 6 m3,
concentrated to 0.5 L (i.e., concentrated by a factor of 12,000), and analysis of at least 450 1-mL
aliquots, considering that CVs at the highest volumes were > 20%. As higher concentration factors were
likely unrealistic, it was suggested that larger sample sizes and improved analytical methods should be
used. Lemieux et al. (2008) also noted that these laboratory trials represented a "best case" situation
because the study was conducted under simplified, "ideal" conditions rather than with natural organism
assemblages in natural seawater.
When concentrations are close to the performance standard, a single sample may require too large a
volume of water to be logistically feasible. In that case, complete, continuous time-integrated sampling
(with the entire volume analyzed) and combining samples across multiple trials can improve resolution
while maintaining statistical validity. To that end, Miller et al. (2011) applied statistical modeling (based
on the Poisson distribution) to a range of sample volumes and plankton concentrations. They calculated
the statistical power of various sample volume and zooplankton concentration combinations to
C-3
-------
differentiate various zooplankton concentrations from the proposed standard of < 10 m"3. Their study
involved a two-stage sampling approach. Stage 1 checked compliance based on a single sample, which
was expected to be effective when the degree of noncompliance was large. Stage 2 combined several
samples to improve discrimination (1) when concentrations are close to the performance standard, or (2)
when a large-volume, single-trial sample would be logistically problematic, or both. The Stage 2
approach took advantage of the fact that the sum of several Poisson random variables is still a Poisson
distribution, and is called the "summed Poisson method." Stage 2 also compared the summed Poisson
approach to power calculations using standard t-tests, the nonparametric Wilcoxon Signed Rank test
(WSRT), and a binomial test, all well-known statistical techniques. The summed Poisson approach had
more statistical power relative to the other three statistical methods. Not surprisingly, as noncompliant
concentrations approached the performance standard, the sampling effort required to detect differences
in concentration increased. The major conclusions from this study are presented in Section 3.2.1.
The major finding from Miller et al. (2011) is that three trials of time-integrated sampling of 7 m3 (and
analyzing the entire concentrated sample from the 21 m3) from a ship's BW discharge can theoretically
result in 80% or higher probability of detecting noncompliant discharge concentrations of 12 vs. 10 live
organisms m"3. Thus, pooling volumes from separate trials will allow lower concentrations to be
differentiated from the performance standard, although the practicability and economic costs of doing so
have not been evaluated. Moreover, the practical limits of increased statistical sample sizes may already
tax the capabilities of well-engineered land-based ballast water test facilities used in verification testing.
Shipboard testing in the U.S. has been done on a pilot scale to date (i.e., the USCG Shipboard Testing
Evaluation Program, STEP), but we imagine that pooling volumes from multiple trials might also be
problematic on vessels used for shipboard verification testing and compliance testing. According to
Table C-l, to meet a standard ten-fold more stringent than the EVIO D-2/ Phase 1 standard would require
anywhere from 120-600 m3 of whole-water sample volumes, which is impracticable at this point - test
facilities in the U.S. typically analyze ~5 m3 of water per test (e.g., MERC 2009a, 2010a, 2010b; Great
Ships Initiative 2010).
Additional challenges of sampling large volumes
Lee et al. (2010) calculated the probability of finding one or more organisms in a sample as l-e° v(l
minus the probability of finding no organisms) for a series of organism concentrations and sample
volumes, where e is the natural log, c is the true concentration of organisms, and v is the sample volume
(Table C-3). The authors used the following assumptions:
• Performance standards are for the concentration of organisms in the ballast discharge (rather
than the maximum number of organisms), so that the purpose of sampling is to estimate the
"true" concentration of organisms in the discharge, referred to as average-based sampling;
• The organisms are randomly distributed and therefore amenable to modeling with the
Poisson distribution, as above;
• All organisms are counted, with no human or instrumentation errors, so that any variation
among samples for a given population (species) is from the natural stochasticity of sampling;
• The sample volume is calculated from the total volume of ballast water filtered
(concentrated) and the filtrate volume that is subsampled. For example, following Lemieux et
al. (2008): 100 m3 of ballast water is filtered through a net to retain the zooplankton-sized
organisms; the organisms are rinsed from the net, collected, and diluted to 1 L of water to
give a concentration factor of 100,000:1. The organisms from 20 1-mL subsamples are
C-4
-------
counted: Total sample volume = 20 mL subsamples/1000 mL concentrated sample x 100 m3
ballast water filtered = 2 m3.
Table C-2. Probability of detecting > 1 zooplankton-sized organism for sample volumes (100 mL to
300 m3) and ballast water concentrations (0 to 100 organisms m"3). Gray boxes indicate probabilities
of detection^ 0.95. (Source: Lee et al. 2010).
Sample volume, in3
0.0001 (100 mL)
0.001 (1 L)
0.01 (10 L)
0.1 (100 L)
1
5
10
25
50
100
300
True concentratiou (organisms per m3)
0
0
0
0
0
0
0
0
0
0
0
0
0.001
<0.001
<0.001
<0.001
<0.001
0.001
0.005
0.010
0.025
0.049
0.095
0.259
0.01
<0.001
<0.001
<0.001
0.001
0.01
0.049
0.095
0.221
0.393
0.632
0.950
0.1
O.001
<0.001
0.001
0.01
0.095
0.393
0.632
0.918
>0.99
>0.99
>0.99
1
O.001
0.001
0.01
0.095
0.632
>0.99
>0,99
>0.99
>0.99
>0.99
>0.99
10
0.001
0.01
0.095
0.632
>0.99
>0.99
>0.99
>0.99
>0.99
>0.99
>0.99
100
0.01
0.095
0.632
>0.99
>0.99
>0.99
>0.99
>0.99
>0.99
>0.99
>0.99
As Table C-2 illustrates, 100 L of ballast must be sampled to have a > 99% probability of detecting at
least 1 zooplankton-sized organism when the true concentration is 100 organisms per m3. When small
sample volumes are collected, the probability of detecting an organism is low even at relatively high
organism concentrations; for example, organisms will be detected in fewer than 10% of subsamples if a
1-L sample is taken and the "true" concentration is 100 organisms m"3. This analysis also illustrates that
when no organisms are detected from a relatively small sample, the true concentration in the ballast tank
may still actually be large - it depends on the sample volume collected.
Lee et al. (2010) then estimated the upper possible concentration (UPC, upper 95% CI) of organisms
actually present in ballast water from the number of zooplankton-sized organisms in a sample volume
(ranging from 100 mL to 100 m3) based on the Poisson distribution. As Table C-3 shows, 0 organisms
detected in 1 m3 of sample could correspond to a true concentration of organisms in the ballast tank of
up to -3.7 organisms m"3. The error is much larger for a small sample volume of 1 L; 0 organisms
detected could correspond to a true concentration of-3,700 organisms m"3.
C-5
-------
Table C-3. Upper possible concentration (UPC) of zooplankton-
sized organisms based on one and two tailed 95% exact confidence
intervals when zero organisms are detected in a range of sample
volumes. (Source: Lee et al. 2010).
Sample volume, in3
0.0001 m3 (lOOiiiL)
0.001 m3 (1 L)
0.01 m3 (10L)
0.1 m3 (lOOL)
0.5 m3 (500 L)
1 m5
10 ill3
100 m3
Upper possible
concentration, org m""
one-tailed
29,960
2.996
299.6
29.96
5.992
2.996
0.300
0.030
two-tailed
36,890
3.689
368.9
36.89
7.378
3.689
0.369
0.037
Third, in the above analyses, the true concentrations of zooplankton-sized organisms are known.
The goal in sampling unknown concentrations of organisms in ballast water is to accurately
assess whether a given BWMS treats water with true organism concentrations that meet a given
performance standard. Inherent stochasticity of sampling may result in an indeterminate
category, as well, and the probability of obtaining an indeterminate evaluation increases with
decreasing sample volume and increasing stringency of the ballast water standard (Figure C-l).
Based on this analysis, it would be necessary to sample -0.4 m3 of ballast water to determine
whether the EVIO standard of < 10 zooplankton-sized organisms m"3 was met if fewer than
approximately 10 organisms were observed in the sample (Figure C-1B).
C-6
-------
A Discharge standard s 0,01 org rrr3
B Discharge standard s 10 org nr3
15
>
•1
I
10
0 -
PASS
0.01 0.1
10
100 1000
0.01
Log,0 Sample volume, m3
Figure C-l. Determining whether ballast water discharge exceeds or meets a performance standard of <
0.01 (A) and <10 (B) organisms m-3 (note: axes have different scales). Red regions indicate total
organism counts that exceed the standard. Green regions indicate total organism counts that meet the
standard. White regions indicate indeterminate results; counts in this region do not pass or fail inspection
based on two-tailed 95% confidence intervals. (Source: Lee et al. 2010).
Spatially Aggregated Populations - Negative Binomial Distributions
This section illustrates how difficult statistical analyses can become when working with spatially
aggregated populations. It further emphasizes the gains made from doing a complete count of a
representative sample that has been continuously and isokinetically taken.
If organisms are aggregated (i.e., in clumped or contagious populations) rather than randomly distributed
in a ballast tank, a different statistical approach is required. For aggregated populations, the variance
exceeds the mean (negative binomial distribution, o > u); thus, as the variance increases, the number of
organisms in a random sample is increasingly unpredictable. Because it is more difficult to accurately
estimate the true concentration, more intensive sampling is required. Lee et al. (2010) recommend use of
the negative binomial distribution to model aggregated populations. This distribution can be used to
predict the probability of finding a certain number of organisms in a sample. It is defined by the mean
(u) and the dispersion or size parameter (0 = (//(o2 - u), where o2 = the variance; the smaller the
dispersion parameter, the more aggregated the population.
The problem of having to sample multiple subsamples from large volumes to accurately assess low
densities of organisms is compounded by aggregated distributions (Figure C-2). In the comparison given
in Lee et al. (2010), for a randomly distributed population with a true concentration of 1 zooplankton-
sized organism m"3, -37% of the subsamples from aim3 sample of treated ballast water would contain
zero zooplankton-sized organisms. For an aggregated population with a dispersion parameter of 0.1,
however, -79% of the subsamples would contain zero organisms (Figure C-2). The relationship between
the probability of finding zero organisms in a sample and the amount of aggregation is also illustrated
(Fig. C-3) for the concentration of 1 organism m"3. As variance (a2) increases, the dispersion parameter 9
C-l
-------
decreases, indicating more aggregation, with increasing probability of finding no organisms in a sample.
With more aggregation, the probability of samples containing large numbers of organisms relative to the
true concentration also increases. Thus, large numbers of subsamples from large sample volumes must
be taken to account for aggregated populations; otherwise, there will be a high probability that the
concentration estimates from sample analyses will be either much lower or much higher than the true
concentration.
0.6-,
0.4-
ra
jfj
D
0,2-
0.0 J
Poisson
Aggregated
I
lit
01234567
Number of organisms in sample
Figure C-2. Comparison of sample probabilities from a randomly distributed population (Poisson distribution) vs. an
aggregated population with a dispersion parameter of 0.1 (negative binomial distribution) for a sample volume of 1 m3 and
concentration of 1 organism m"3. For low organism numbers (3 or fewer m"3), the probability that a sample will contain zero
organisms tends to be much greater for the aggregated population. (Source: Lee et al. 2010).
ta 60
variance Sigma Sua-cd
i -i
Figure C-3. The probability of finding zero organisms in a sample volume of 1 m3 and concentration of (i = 1 organism m"3.
The probability of 0 organisms = (1 + 9)"e, where dispersion parameter 9 = l/(o2 - 1). When o2 = 1.0, organisms are randomly
distributed, at which the probability of 0 organisms in the sample = 0.37 (Poisson distribution) (Elliott 1971).
-------
Determination of whether a population is aggregated is complicated, since it is the scale of the
aggregation pattern relative to the size of the sampling unit that controls the estimate of aggregation
(Fig. C-3). If organisms form clumps that are randomly distributed, the population may be highly
aggregated, but in a small sample volume containing 0 or 1 organisms, the population will appear
randomly distributed or only slightly aggregated. With increasing sample volume, the variance in the
number of organisms increases in comparison to the mean, and maximum variance is encountered when
the sample volume is equal to the volume of a single cluster of organisms (Elliott 1971). For larger
sample volumes, a sample unit will include several clusters, so the variance decreases in comparison to
the mean and the observations will approach a Poisson distribution. Lee et al. (2010) recommend the
Taylor power law (Taylor 1961) as an alternative to the negative binomial, because it can accommodate
a wider range of aggregated distributions than the negative binomial.
Overall, the possibility for and degree of aggregation represent challenges in sampling sufficiently large
volumes of ballast water to determine whether a given BWMS passes or fails to meet standards more
stringent than the present EVIO guidelines, even if the true concentrations of organisms are 10 to 1,000
times higher than the performance standard. This remains a problem in quantifying many protist-sized
organisms, but becomes less of a problem with very small organisms such as bacteria, which have a
tendency to clump but are effectively counted as colonies and not individuals. However, in Lemieux et
al. (2008), data from protist-sized microbeads at various concentrations were analyzed and
concentrations of 100 mL"1 and lower were found to adhere to a Poisson distribution. The flasks of
microbeads were well mixed, as would be samples of ballast water collected from the sample ports and
collected to be representative of the entire volume sampled (e.g., over the entire discharge operation of
the tank). Likewise, monocultures of protists in low densities (-10 to 30 mL"1) adhered to a Poisson
distribution (Nelson et al. 2009). These data lend support to using the Poisson distribution to analyzed
ballast water samples.
C-9
------- |