EPA/600/R-10/146
September, 2010
GENERIC PROTOCOL FOR THE
VERIFICATION OF BALLAST WATER
TREATMENT TECHNOLOGY
Produced by
NSF International
Ann Arbor, MI
For
U.S. Environmental Protection Agency
Environmental Technology Verification Program
In cooperation with
U.S. Coast Guard
Environmental Standards Division
Washington, DC
And
U.S. Naval Research Laboratory
Center for Corrosion Science and Engineering
Washington, DC
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
NOTICE
The U.S. Environmental Protection Agency, through its Office of Research and Development,
funded and managed, or partially funded and collaborated in, the research described herein. It has
been subjected to the Agency's peer and administrative review and has been approved for
publication. Any opinions expressed in this report are those of the author (s) and do not
necessarily reflect the views of the Agency, therefore, no official endorsement should be inferred.
Any mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
11
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
FOREWORD
The EPA is charged by Congress with protecting the nation's air, water, and land resources.
Under a mandate of national environmental laws, the Agency strives to formulate and implement
actions leading to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, the EPA's Office of Research and
Development provides data and science support that can be used to solve environmental problems
and to build the scientific knowledge base needed to manage our ecological resources wisely, to
understand how pollutants affect our health, and to prevent or reduce environmental risks.
The Environmental Technology Verification (ETV) Program has been established by the EPA to
verify the performance characteristics of innovative environmental technology across all media
and to report this objective information to permitters, buyers, and users of the technology, thus
substantially accelerating the entrance of new environmental technologies into the marketplace.
Verification organizations oversee and report verification activities based on testing and quality
assurance protocols developed with input from major stakeholders and customer groups
associated with the technology area. ETV consists of six environmental technology centers.
Information about each of these centers can be found on the Internet at http://www.epa.gov/etv
Under a cooperative agreement, NSF International has received EPA funding to plan, coordinate,
and conduct technology verification studies for the ETV "Water Quality Protection Center" and
report the results to the community at large. The WQP Center's primary technology areas address
surface water pollution concerns such as ship ballast water treatment, wastewater treatment,
stormwater runoff treatment, confined animal feeding operations (CAFOs), and urban
infrastructure rehabilitation.
in
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
ACKNOWLEDGEMENTS
EPA and NSF acknowledge and thank those persons who participated in the preparation and
review, subsequent validation, and the revision of this Generic Protocol for Verification of Ballast
Water Treatment Technologies. Without their hard work and dedication to the project, this
document would not have been possible. The original draft was prepared by Battelle, Duxbury,
MA, while significant input in later drafts was provided in a multi-year effort by the Naval
Research Laboratory, Key West, FL. Much input was also provided by the biological sub-group
of the Ballast Water Treatment Protocol Technical Panel as well as the Technical Panel in whole.
Revised Protocol Authors
Edward Lemieux
Jonathan Grant
Timothy Wier
Lisa Drake
NRL Technical Staff
Stephanie Robbins
Kevin Burns
Scott Riley
Luke Davis
Wayne Hyland
Bruce Nelson
Tiffanee Donowick
Robert Brown
Naval Research Laboratory
Battenkill Technologies, Inc.
EXCET, Inc.
Science Applications International Corporation
Science Applications International Corporation
Science Applications International Corporation
Science Applications International Corporation
Science Applications International Corporation
Azimuth Technical Consultants, Inc.
Battenkill Technologies, Inc.
Science and Engineering Technologies, Inc.
Azimuth Technical Consultants, Inc.
Original Protocol Writers
Deborah Tanis
Carlton D. Hunt
Battelle
Battelle
NSF Staff
Thomas Stevens
EPA Staff
Raymond Frederick
Project Manager, ETV Water Quality Protection Center
Project Officer, ETV Water Quality Protection Center
ETV Water Quality Protection Center — Technical Panel Participants for Current and Past
Versions
Donald Anderson Woods Hole Oceanographic Institute
Allegra Cangelosi Northeast-Midwest Institute
Dorn Carlson National Oceanic and Atmospheric Administration
Fred Dobbs Old Dominion University
Richard Everett U.S. Coast Guard
IV
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Maurya Falkner
Richard Fredricks
Stephan Gollasch
Frank Hamons
Richard Harkins
Penny Herring
Russell Herwig
Brian Howes
James Hurley
Thomas Mackey
Lucie Maranda
Kathy Metcalf
Richard Mueller
Gail Roderick
Andrew Rogerson
Terri Sutherland
Mario Tamburri
Fred Tsao
Thomas Waite
Nick Welschmeyer
California State Lands Commission
Maritime Solutions, Inc.
GoConsult and Chairman, ICES/IOC/IMO Working Group
on Ballast and Other Ship Vectors (WGBOSV)
American Association of Port Authorities
Formerly with the Lake Carriers Association, currently
with Keystone Shipping Company
U.S. Coast Guard R&D Center
University of Washington
University of Massachusetts
U.S. Coast Guard R&D Center
Hyde Marine, Inc.
University of Rhode Island
Chamber of Shipping of America
Northeast Technical Services Company, Inc.
U.S. Coast Guard R&D Center
California State University, Fresno
Fisheries and Oceans Canada
University of Maryland
U.S. Navy
Florida Institute of Technology
Moss Landing Marine Laboratories
v
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Table of Contents
Chapter 1 Introduction 1
1.1 The ETV Program 1
1.2 Objectives of Verification Testing 1
1.3 Purpose and Scope of the Protocol 2
1.4 Verification Testing Process 2
1.5 Policies and Program Specifications and Guidelines 2
Chapter 2 Responsibilites of Involved Organizations 3
2.1 Vendor 3
2.2 Testing Organization (TO) 3
2.3 TF Owner (Owner) 3
2.4 Verification Organization (VO) 4
2.5 Environmental Protection Agency (EPA) 4
2.6 Stakeholder Advisory Group (SAG) 4
2.7 Technology Panel 4
Chapter 3 Ballast Water Treatment System Capabilities and Description 5
3.1 Ballast Water Treatment System Definition 5
3.2 Technology or Treatment Performance Claims 5
3.3 Acceptability for Testing 6
3.4 Test BWTS Requirements 6
3.5 Operating and Maintenance (O&M) Evaluation 7
3.6 Biological Efficacy Evaluation with Standard Test Organisms 9
3.7 Calibration and Test Requirements 10
3.8 System Documentation Evaluation 10
3.9 Technical Data Package Submission 11
3.10 Format for the BWTS Technical Data Package 12
Chapter 4 Treatment Verification TQAP Development 14
4.1 Description of Ballast Water Treatment System 14
4.2 Required Elements of the TQAP 14
Chapters Experimental Design 16
5.1 Test Verification Factors 16
5.1.1 Biological Treatment Efficacy 17
5.1.2 Operation and Maintenance 17
5.1.3 Reliability 17
5.1.4 Cost Factors 17
5.1.5 Environmental Acceptability 17
5.1.6 Safety Factors 17
5.2 Challenge Conditions 18
5.2.1 Challenge Water - Water Quality Characteristics 18
5.2.2 Challenge Water - Biological Organism Conditions 21
5.2.3 Challenge Water - Flow Rates and Volumes 22
5.3 TF Physical Configuration 22
5.3.1 Overall experimental configuration 22
5.3.2 Sampling Methodology 23
vi
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
5.3.3 Test Organism & Water Quality Augmentation 29
5.3.4 Control & Instrumentation 29
5.4 Verification Testing 30
5.4.1 Treatment System Commissioning 31
5.4.2 Operation and Maintenance (O&M) Manual 32
5.4.3 Vendor and Test Organization Requirements 32
5.4.4 Toxicity Testing for Biocide Treatments 33
5.4.5 BE and O&M Verification Strategy: Test Duration and Coordination 33
5.4.6 Biological Efficacy (BE) Verification Testing 35
5.4.7 BE Validity Criteria 48
5.4.8 Alternative and Emerging Methods 50
5.4.9 Operation and Maintenance Verification Factor 51
Chapter 6 Reporting Verification Testing Results 57
Chapter 7 Quality Assurance/Quality Control (QA/QC) 58
7.1 Verification of Test Data 58
7.2 Project Management 58
7.3 Measurement and Data Acquisition 58
7.4 Assessment 59
Chapters Data Management, Analysis and Presentation 60
8.1 Data Management 60
8.2 Data Analysis and Presentation 60
Chapter 9 Environmental, Health, and Safety Plan 61
References 62
Appendix A Quality Assurance Project Plan 65
Appendix B Approach for Evaluating standard Test Organisms 70
Appendix C Report on Discharge Sampling 96
vn
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
List of Tables
Table 1. Acceptable Variations for Frequency and Voltage 7
Table 2. Recommended Standard Test Organisms for Bench-Scale Testing 10
Table 3. Water Quality Challenge Matrix for Verification Testing 19
Table 4. Minimum Criteria for Challenge Water Total Living Populations 22
Table 5. Accuracy and Precision Requirements for Potential Sensors 24
Table 6. Likely Treatment Sequences and Applications Inherent to Ballast Operations 31
Table 7. Criteria for Concentrations of Living Organisms in Control Tank Discharge 34
Table 8. Core and Potential Auxiliary Parameter and Measurement Techniques 36
Table 9. Sample Volumes, Containers and Processing 38
Table 10. Recommendation for Water Quality Sample Analysis Methods 39
Table 11. Data Quality Objectives for Water Quality Samples 40
Table 12. Density Confidence Intervals for Poisson Distributions Using the Chi-Square Statistic
43
Table 13. Sample Volume Required Relative to Treatment Standards-Organisms > 50 jim 44
Table 14. Sample Volume Required Relative to Treatment Standards Organisms > 10 jim and <
50 |im 44
Table 15. Challenge Test Validation Criteria by Location 50
List of Figures
Figure 1. Sampling design example for in-tank treatment 26
Figure 2. Sampling design example for in-line treatment 27
Vlll
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
GLOSSARY OF TERMS
Accuracy: The degree of agreement between an observed value and an accepted reference value,
including a combination of random error (precision) and systematic error (bias) components that
are due to sampling and analytical operations (EPA, 1992).
Ambient Populations: The biological organisms, including bacteria, protists, and zooplankton
that are naturally occurring in the water at the TF location.
Ballast Water Treatment System (or System): Prefabricated, commercial-ready, treatment
systems designed to remove, kill or inactivate (prior to discharge) organisms in ballast water. The
entirety of a vendor's ballast water treatment product will be used to achieve the vendor claims for
treatment efficacy or operational performance, and includes all components, in an integrated
fashion.
Bias: The systematic or persistent distortion of a measurement process that causes errors in one
direction.
Challenge Water: Water supplied to a treatment system under test. Challenge water must meet
specified ranges for living organism densities and water quality parameters and is used to assess
the efficacy of the treatment equipment under full-scale operational conditions.
Comparability: The measure of the confidence with which one data set can be compared to
another.
Completeness: The amount of data collected as compared to the amount needed to ensure that
the uncertainty or error is within acceptable limits.
Core Parameters: The measurements that are required as part of the ETV verification.
Cyst: The dormant cell or resting stage of microalgae, heterotrophic protists, and metazoans,
including but not limited to cysts of dinoflagellates, spores of diatoms, cysts of heterotrophic
protists, and cysts of rotifers.
Effluent: The treated discharge water produced by a ballast water treatment system.
Equipment: The ballast water treatment system, defined as either a package or a modular system,
which is tested in the Verification Testing Program.
ETV Testing: Testing of a technology under the EPA Environmental Technology Verification
Program following provisions of an established protocol and/or TQAP, with the final outcome
being a Verification Report, containing all findings of the test, and a Verification Statement,
signed by the US EPA and the Verification Organization (VO).
In-Line Treatment: A treatment system or technology used to treat ballast water during normal
flow of ballast during uplift or discharge.
IX
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
In-Tank Treatment: A treatment system or technology used to treat ballast water during the
time that it resides in the ballast tanks. This may involve treatment steps during uptake.
Mean Time Between Failure (MTBF): The predicted elapsed time between inherent failures of
a system during operation. MTBF can be calculated as the arithmetic mean (average) time
between failures of a system. The MTBF is typically part of a model that assumes the failed
system is immediately repaired (zero elapsed time), as a part of a renewal process. This is in
contrast to the mean time to failure (MTTF), which measures average time to system failure with
the modeling assumption that the failed system is not repaired.
Normally distributed data: Data that meet the following criteria: the data forms a bell shaped
curve when plotted as a graph, the mean is at the center of the distribution on the graph, the curve
is symmetrical about the mean, the mean equals the median, and the data are clustered around the
middle of the curve with very few values more than three standard deviations away from the mean
on either side.
Owner: The owner of a test site used for verification testing of a ballast water treatment system.
Performance Data: Removal efficacy and effluent concentration data for core and supplemental
parameters for a given set of Challenge conditions.
Precision: The degree to which a set of observations or measurements of the same property,
obtained under similar conditions, conform to themselves. Precision is usually expressed as
standard deviation, variance, or range, in either absolute or relative terms (NELAC, 1998).
Protocol: A written document that clearly states the objectives, goals, scope, and procedures for
the study of a particular group of similar technologies. A protocol shall be used for reference
during vendor participation in the verification testing program.
Proxy Measurement: A parameter used in lieu of another measurement (i.e., chlorophyll a as a
bulk measure of phytoplankton).
Quality Assurance Project Plan (QAPP): A written document that describes the
implementation of quality assurance and quality control activities during the life cycle of the
project (also see Test/quality assurance plan).
Representativeness: The degree to which data accurately and precisely represent a characteristic
of a population.
Sensitivity: The capability of a test method or instrument to discriminate between different levels
(e.g., concentrations) of a variable of interest.
Stakeholder Advisory Group (SAG): A group overseen by a Verification Organization (VO)
consisting of representatives from verification customer groups, technology developers and
vendors, the consulting engineer sector, the finance and export communities, and government
permitters and regulators.
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Standard Operating Procedure (SOP): A written document containing specific instructions and
protocols to ensure that quality assurance requirements are maintained.
Standard Test Organisms: Biological 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. The organisms are added to the challenge water during testing of ballast
water treatment technologies to determine treatment system effectiveness.
Start-Up: The period between the time the ballast water treatment system is activated and when
stable operating conditions are achieved.
Stable Operation: The time interval following a start-up period that the ballast water treatment
system performs consistently within the range of vendor-specified operating conditions.
Supplemental Parameters: A measurement taken that is specific to a particular treatment and
augments the results of the core parameter measurements.
Technical Panel: A group comprised of a subset of stakeholders and other individuals with a
technical expertise in various ballast water issues, such as fresh water and marine biologists,
environmental scientists, engineers, and ship architects.
Test Cycle: One fill/discharge cycle (including appropriate holding periods) designed to gather
data on treatment efficiency.
Test Facility: A site that provides the necessary infrastructure, systems and personnel to
complete the verification testing described in this protocol. The facility may be part of the Testing
Organization or may be independent from the Testing Organization, but in any case shall be
totally independent from technology vendors testing at their site.
Test/Quality Assurance Plan (TQAP): Also called a Quality Assurance Project Plan (QAPP),
this is a written document that describes the procedures for conducting a test or study according to
the verification protocol requirements for the application of a particular ballast water treatment
system at a particular site. At a minimum, the TQAP shall include detailed instructions for
sample and data collection, sample handling and preservation, precision, accuracy, goals, and
quality assurance and quality control requirements relevant to the particular site.
Testing Organization (TO): An organization qualified to conduct studies and testing of ballast
water treatment technologies in accordance with protocols and TQAPs.
Upset Conditions: Deviation or exception from normal or vendor-defined operating conditions,
for example, system faults or hardware failures.
Vendor: A business that manufactures, assembles, or sells ballast water treatment technologies.
Verification: The establishment of evidence on the performance of a ballast water treatment
system under specific conditions, following a predetermined study protocol(s) and TQAP(s).
XI
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Verification Organization (VO): The party responsible for overseeing TQAP development,
overseeing testing activities in conjunction with the Testing Organization, 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.
Verification Report: A detailed report on the testing results of a particular technology according
to an approved Test /Quality Assurance Plan and conducted under the ETV Program. The report
is typically prepared by the TO and contains a description of the test facility, photographs of
technology being tested methods and procedures, presentation of analyzed data including all
QA/QC data obtained during the test. Appendices include raw data sets and lab audit information,
TQAP, O&M Manual and other relevant information. Both the verification report and verification
statement are publically available on the ETV Program's web site and NSF's web site .
Verification Statement: An executive summary of the verification report, usually 4-6 pages in
length which is signed by EPA and the verification organization.. The verification statement is
intended to be used by the vendor for sales and marketing purposes.
Verification Test: A complete test of a treatment system, following a well defined TQAP which
includes enumeration of ambient and test populations in the challenge water to determine the
efficacy of the technology. Also see ETV Testing.
Viable: According to the EVIO G8 guidelines, "organisms and any life stages thereof that are
living".
xn
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Abbreviations and Acronyms
ATP Adenosine triphosphate
BE Biological efficacy
BWTS Ballast water treatment system(s)
CT Concentration-time relationship (curve) demonstrating the relationship between
concentration and time that achieves desired treatment effect.
m3 Cubic meter
CFR Code of Federal Regulations
DOC Dissolved organic carbon
DOM Dissolved organic matter
EPA U.S. Environmental Protection Agency
ETV Environmental Technology Verification
FRTJ Field replaceable unit
jig/L Micrograms per liter
mgd Million gallons per day
mg/L Milligrams per liter
MAWP Maximum allowable working pressure
MM Mineral matter
MOA Memorandum of agreement
MSDS Material safety data sheets
MTBF Mean time between failures
NRL U.S. Naval Research Laboratory
NSF NSF International (formerly National Sanitation Foundation)
NTTJ Nephelometric turbidity unit
xiii
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
O&M Operations and maintenance
OSHA Occupational Safety and Health Administration
Owner TF owner, if different from the Testing Organization (TO)
POM Particulate organic material
PSU Practical salinity units
QA Quality assurance
QAPP Quality assurance project plan
QC Quality control
QMP Quality management plan
SAG Stakeholder Advisory Group
SOP Standard operating procedure
STO Standard test organism
TF Test Facility
TO Testing Organization
TQAP Test/quality assurance plan
TSS Total suspended solids
USCG U.S. Coast Guard
VO Verification Organization
WQPC Water Quality Protection Center
xiv
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 1
Introduction
1.1 The ETV Program
The U.S. Environmental Protection Agency (EPA) established the Environmental Technology
Verification (ETV) Program in 1995. The goal of the ETV Program is to promote environmental
protection by accelerating the development and commercialization of improved and more cost-
efficient environmental technologies through third-party verification, performance reporting, and
information dissemination. The ETV Program neither certifies nor endorses environmental
technologies, but rather provides objective, high-quality, peer-reviewed performance data that
can be utilized by customer groups and regulators when selecting, permitting, or certifying the
use of environmental technology. The ETV Program's Water Quality Protection Center
(WQPC) addresses technologies to protect surface and ground water from chemical or biological
contamination and conducts performance verifications of technologies resulting in
comprehensive reports that are publically available on the ETV Program web site. Further
information on the ETV Program can be obtained at http://www.epa.gov/etv.
Through a formal Memorandum of Agreement (MOA) signed in 2001, the U.S. Coast Guard
(USCG) and EPA formed a partnership to develop procedures for evaluating the performance of
ballast water treatment systems (BWTS). The partnership also provided the Coast Guard a
pathway to begin the development of technical procedures for approving BWTSs for installation
on ships. EPA's interest includes the ecological, economic and public health impacts of ballast
water discharges. Ballast water treatment is viewed as an important step in mitigating the
proliferation of aquatic invasive species in U.S. coastal waters and the Great Lakes.
1.2 Objectives of Verification Testing
The objective of ETV ballast water treatment technology testing is to evaluate the performance
characteristics of commercial-ready treatment technologies with regard to specific verification
factors, including biological treatment performance, predictability/reliability, cost, environmental
acceptability, and safety. Given the variety of ship and ballast tank types, and potential treatment
system configurations, this protocol addresses the use of a land-based testing facility (TF) rather
than shipboard testing, to provide controlled conditions for verifying treatment performance.
Land-based BWTS verification testing will be conducted in a manner providing information that
is comparable to the maximum practical extent, to ensure that consumers and other stakeholders
can make informed choices in selecting appropriate ballast water treatment technology for
shipboard installations.
It is believed that ballast water treatment systems performing well under the controlled but
challenging conditions specified in this protocol at land-based testing facilities will have a
reasonable chance of performing as well in a shipboard installation. However, because of the
various designs used in ship ballasting systems and the water quality conditions encountered by
vessels in seaports around the world, any assumptions of shipboard technology performance
1
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
based solely on land-based testing results should be avoided. Thorough evaluation of ballast
water treatment technology must also include shipboard trials to monitor biological performance
and other ship-related verification factors over an extended period of time. The U.S. Coast
Guard's Environmental Standards Division http://www.uscg.mil/environmental_standards/
should be contacted for information concerning procedures for shipboard testing of ballast water
treatment technologies.
1.3 Purpose and Scope of the Protocol
The parties involved with ETV testing, including vendors, testing organizations, testing site
owners, and verification organizations, can use the information provided in this protocol as
guidance for BWTS verification testing. This protocol provides guidance on the necessary
elements of verification testing including: technology acceptability; vendor provided
specifications and information; and test/quality assurance plan (TQAP) development and
content. The protocol is intended for verification testing of entire BWTSs, not individual
component technologies that could be combined to form a system. The systems addressed by the
protocol could be in many configurations, such as treatment on uplift or discharge, treatment in-
transit (in-tank), or combinations of these options.
Periodic review and revision of protocols is a critical aspect of the ETV Program. As such, this
protocol will be reviewed periodically and revised as necessary. These efforts will keep the
protocol scientifically and functionally up to date.
1.4 Verification Testing Process
Verification testing is a three-step process, consisting of planning, verification, and data
assessment/reporting phases. The planning phase includes development of standardized
challenge conditions and the specific experimental design as it will be applied to the testing of
the vendor's BWTS. A site and treatment system-specific TQAP are prepared during the
planning phase in accordance with the guidance provided in Chapter 4 of this protocol. The
BWTS vendor, Testing Organization (TO), and Verification Organization (VO) collaborate on
the planning phase documents. The verification phase involves the testing of the BWTS by the
TO under the conditions and standard operating procedures specified in the TQAP. In the data
assessment and reporting phase, data are processed and analyzed by the TO, who prepares the
draft verification report and verification statement. The VO is responsible for QA review of the
data generated during the testing and coordination of the fmalization of the verification report
and statement.
1.5 Policies and Program Guidelines
Treatment system verification testing will be conducted in accordance with an approved TQAP
(for test specific activities) and with the policies and guidelines set forth by an established
Quality Management Plan (QMP) for the testing facility. Examples of ETV Center QMPs and
quality assurance plan documents for other testing activities can be viewed on the ETV
Program's web site. EPA also provides guidance documents for preparing QMPs and quality
assurance project plans (QAPPs) at http://www.epa.gov/quality
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 2
Responsibilities of Involved Organizations
Verification testing will involve several organizations with responsibilities divided among them.
These organizations may include the vendor of the treatment system, the TO (TO), the Test
Facility (TF) owner, the Verification Organization (VO), EPA, and sometimes the Technology
Panel and Stakeholder Advisory Group.
2.1 Vendor
The vendor of the ballast water treatment system will apply to the VO for verification testing.
The vendor must provide the VO and TO verification testing objectives and any existing relevant
performance data, along with the information required in Chapter 3. This information will be
considered during the development of the TQAP, which will be reviewed and approved by the
vendor. The vendor will provide a complete System along with any relevant operation and
maintenance manuals. Additionally, the vendor will be responsible for assuring proper
installation and set up of the equipment at the test site, training of TO personnel on BWTS
operation, and confirmation of the system's proper operation prior to commissioning and
commencement of maintenance or treatment efficacy testing. It is strongly recommended that
the vendor inspect the installation and operation of the system prior to the initiation of the
testing. The vendor will be available for logistical and technical support as required during the
planning and verification phases, but will not be directly involved in the testing. The vendor will
also be responsible for reviewing the verification report and statement generated from the TO.
2.2 Testing Organization (TO)
The TO is responsible for preparing the TQAP and working with the vendor and VO to assure
EPA approval of the TQAP, conducting the verification testing and all aspects of test data
management, and may be responsible for preparing drafts and final versions of the verification
report and verification statement. The TO is also responsible for coordinating all personnel and
testing activities, operating the vendor's equipment as specified in the equipment operations and
maintenance manual(s), and evaluating and reporting on the performance of the equipment.
Maintaining security for testing activities and site safety for all personnel is also the
responsibility of the TO.
2.3 TF Owner (Owner)
If different from the TO, the Owner of the verification TF may provide logistical and technical
support during planning and verification phases, as agreed upon by the TO, vendor, and Owner.
The Owner must notify the TO of any logistical or operational developments that may affect the
verification testing process and results.
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
2.4 Verification Organization (VO)
The VO is responsible for the technical and administrative operation of the ETV Program's
Water Quality Protection Center and all verification activities conducted on behalf of the ETV
Program. The VO is responsible for overseeing the development and approval of the TQAP, and
collaborating with the TO to administer testing activities at the TF. The VO is also responsible
for reviewing, revising and submitting the Verification Report and Statement to the EPA project
officer for final QA and technical review. The Report and Statement are typically drafted by the
TO, but they may be drafted by the VO or a contractor to the VO. The VO is also responsible
for initiating and coordinating periodic review and revision of this protocol.
2.5 Environmental Protection Agency (EPA)
The EPA Office of Research and Development (ORD), through the National Risk Management
Research Laboratory (NRMRL) in Cincinnati, Ohio oversees the Environmental Technology
Assessment, Verification and Outcomes Staff (ETAVOS) where the ETV Program is
headquartered. The ETV Program's Water Quality Protection Center is managed
administratively from ETAVOS. The Project Officer (PO) for the WQPC is assigned to the
Water Supply and Water Resources Division, Urban Watershed Management Branch. The
Project Officer is responsible for administrative and technical management of the cooperative
agreement with the VO. The PO is also responsible for obtaining EPA reviews of TQAPs for
BWTS verification testing, the verification report and statement generated from the testing, and
for assuring that the report and statement are posted on the EPA/ETV web site. EPA is also
responsible for coordinating review and approval of revisions that may be proposed to this
protocol.
2.6 Stakeholder Advisory Group (SAG)
Stakeholder Advisory Groups (SAGs) are established in each of the ETV Program's six Centers,
and consist of representatives from verification customer groups, such as buyers and users of
technology, developers and vendors, the consulting engineering sector, the finance and export
communities, and government regulators. The SAGs support generic verification protocol
development, prioritizing the types of technologies to be verified, and defining and conducting
outreach activities appropriate to the technology area and customer groups. In addition, the
SAGs may review WQPC-specific procedures and selected ETV verification reports emerging
from the ETV WQPC and serve as information conduits to the particular constituencies that each
member represents. The Ballast Water SAG, of the WQPC, is charged with addressing ballast
water treatment technologies.
2.7 Technology Panel
The Technology Panel is comprised of a subset of stakeholders and other individuals with
technical expertise in ballast water and environmental technology issues. Scientists, engineers,
technology vendors, naval architects, and regulators supported the development of this
Verification Protocol as participating panel members. In the future, the Technology Panel may
be responsible for reviewing TQAPs and verification reports and statements. The Panel will also
play a key role in working with the VO in reviewing and revising this protocol as needed.
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 3
Ballast Water Treatment System Capabilities and Description
3.1 Ballast Water Treatment System Definition
For the purposes of this verification testing program, ballast water treatment systems (BWTS)
are defined as:
Prefabricated, commercial-ready, treatment systems designed to remove, kill or
inactivate (prior to discharge) organisms in ballast water. This includes all components,
in an integrated fashion, required for shipboard operation.
Note that it is understood that many of the proposed regulatory discharge standards, and in fact
the desired effect of BWTSs, is that these technologies should render organisms unviable or
incapable of reproduction. In other words, to "kill, remove or inactivate" is technically
unnecessary when the objective is to eliminate the organism's capability for reproduction.
However, as the introduction of "viability" as a measure of efficacy significantly complicates the
Protocol and test methods, and since "kill, remove or inactivate" is a conservative approach, the
latter has been adopted as the measure of biological efficacy in this Protocol.
This definition includes both in-line (systems that treat the flow of ballast water either on uplift
or discharge) and in-tank systems (systems that treat ballast water during the time it resides in the
ballast tanks). Typically, BWTSs treat an average design flow between 1.4 - 17 m3 per minute
(370 - 4,490 gpm) or a total tank volume within a range of 20 - 14,500 m3 (5,280 - 3,830,000
US gal).
Systems that will be tested under this program will be capable of treating the entire discharge or
ballast water volume for biological organisms, either through a one-step treatment process or
through multi-step treatment processes, and will be capable of treating a wide range of source
water typical of ballast uplifted from fresh, coastal, estuarine and/or marine origins. These
technologies may be biological, physical, or chemical in nature or a combination of any or all of
the technologies. Treatment systems, or components of systems, that provide only partial
treatment of the discharge are excluded from verification testing.
3.2 Technology or Treatment Performance Claims
The vendor will supply a statement of treatment performance claims for the treatment or
technology. Discharge water quality specifications should reference current EPA regulations or
recommendations for shore discharge standards. The statement should include, as a minimum:
• Quantitative measures of biological treatment efficacy expressed as a concentration upon
discharge for a range of biological size groups as defined in Section 0; minimum
reporting parameters are specifically detailed in Section 5.4.6;
• Quantitative measures of operational performance requirements to achieve the biological
treatment performance stated above; these should include, as a minimum, the allowable
5
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
and treatable flow rate range and water quality (dissolved and participate matter
concentration and particulate size range, salinity, water temperature, turbidity and
dissolved oxygen content);
• Treatment capabilities over the anticipated range of maritime environmental conditions
must be identified by the vendor; the effects of extremes in temperature, turbidity,
biomass density, or other environmental conditions that may impact the treatment system
must be noted where these may cause variations in Vendor performance specifications;
• Quantify the concentration of disinfection residuals, by-products and toxicity for relevant
systems;
• The required operational and maintenance conditions (operator time, power requirements,
chemical consumption requirements, reliability, etc.) to achieve the biological
performance under a range of source water conditions typical to fresh, coastal, estuarine,
and marine ballast water (water conditions are detailed in Section 0); and
• The projected mean-time between failure (MTBF) for the technology given the operation
and maintenance schedules provided for the technology.
3.3 Acceptability for Testing
The treatment system must meet the definition of a BWTS and all existing environmental
regulatory requirements for operation and treatment byproduct discharge (including EPA
Registration under FIFRA for any antimicrobial chemical used in the system as active
substances). The system must be safe for the crew to operate and be compatible with other
shipboard systems as defined by marine equipment certification procedures by the American
Bureau of Shipping (ABS), or Det Norske Veritas (DNV). Only complete treatment systems will
be accepted for verification testing. Moreover, it is anticipated that a BWTS will have undergone
bench-scale testing with standard test organisms (STO) to validate treatment efficacy under
controlled laboratory conditions prior to the full scale standardized testing within the ETV
Program.
The VO has the right to reject a proposed system that does not satisfy the definition of a BWTS
in Section 3.1. A proposed treatment system may also be denied acceptance to the verification
testing program if, for technical or logistical reasons, it cannot be accommodated at the TF or its
use will result in non-compliance with the discharge requirements for the TF.
3.4 Test Requirements for BWTS
All piping, valves and fittings shall comply with regulations and marine industry standards as
contained in applicable sections of 46 CFR Subchapter F. Pressure piping shall be fitted with
relief valves set not to exceed maximum allowable working pressure (MAWP).
Electrical and electronic components in alternating current (AC) systems must be capable of
operating satisfactorily under normally occurring variations in voltage and frequency. Unless
otherwise stated, the variations from the rated value may be taken from Table 1. Direct current
(DC) system devices must be capable of operating satisfactorily at minus 15% voltage.
Conductors, power supply, and over-current protection shall be provided in accordance with 46
CFR Subchapter J and appropriate marine industry standards.
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Table 1. Acceptable Variations for Frequency and Voltage
Quantity in Operations Permanent Variation Transient Variation
Frequency ±5% ±10% (5 s)
Voltage ±6%, -10% ±20% (1.5s)
Operating conditions and tolerances for TO supplies of water pressure and flow, power
conditions, air pressure and flow, or any other requirements specific to the BWTS must be
clearly identified in system documentation.
System design should provide for appropriate lift and/or hoist points during installation. Center
of gravity, no step areas and other installation specific information should be clearly identified.
Any areas presenting a hazard to personnel during installation, checkout, and operation should be
visibly marked.
Recommendations to ensure post-installation operator access to maintenance ports, access
panels, and field replaceable units (FRUs) should be clearly identified in an installation guide
with appropriate layout diagrams.
3.5 Operating and Maintenance (O&M) Evaluation
The BWTS will be evaluated during the testing to determine if the system is:
• Designed and constructed to ensure that user access is restricted to essential controls for
normal operation of the system;
• If access beyond these controls is available for emergency maintenance and temporary
repair, and requires the breaking of security (lockout) seals or activation of another
device indicating an entry to the equipment;
• Provides capability for efficient maintenance and repair operations and provides a high
mean-time between failures (MTBF);
• If minor and major maintenance schedules, pre-requisite training, level of effort, and
recommended spares/supplies are detailed in the appropriate sections of the O&M
manual;
• If adequate documentation, including drawings, diagrams and instructions necessary for
routine maintenance, troubleshooting, and repairs, are provided;
• Designed to ensure any potential exposure to hazards or hazardous materials that are
involved in the maintenance or operation of the equipment are minimized;
• If explicit warning labels identifying the hazard are installed in accordance with OSHA
and/or other appropriate federal regulations;
• If procedures for working with stated hazards are clearly identified in the operating
instructions;
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
• If by-product, disposable component, or field replaceable unit (FRU) that presents a
safety or environmental hazard are explicitly identified, along with procedures for
material handling and disposal according to relevant regulations; and
• If the vendor provides technical support for this system via phone and internet, including
contact information for both methods.
The BWTS operation and control capability will be evaluated to determine the following:
• The control system ensures that services needed for the proper operation of the BWTS
are provided through automatic arrangements and operators are promptly alerted when
conditions warrant human intervention;
• The operator is able to control all BWTS functions through a single control unit;
• The control unit automatically monitors and adjusts optimal treatment dosages or
intensities, or other aspects of the BWTS, and/or provides control signals to the ballast
water system of the vessel to properly provide the necessary treatment;
• The control unit provides a continuous self-monitoring function when the BWTS is in
operation;
• The control unit includes a tamper-proof or tamper-evident recording device, located in a
position easily accessible to the person in charge of the BWTS, that provides the operator
the parameters listed below during ballast water treatment while continuously logging the
data:
o Proper functioning and status of all the services needed for the operation of the
BWTS;
o All parameters necessary to ensure the proper operation of the BWTS;
o Status of the valves present in the BWTS, including those leading to overboard
discharge;
o Total quantity of ballast water treated;
o Ballast water treatment rates;
o Alarm conditions;
o Date and time of start and end of the treatment operation;
o Ballast operation monitored (upload, discharge);
o Calibration and maintenance events;
o Other system events of interest;
o Relevant and necessary measurement information required for control and monitoring
operation of the BWTS;
o Meter and sensor accuracy to measure the suite of parameters appropriate and
necessary for control of the BWTS, representing the actual value of the parameters
being monitored within 10% despite the presence of contaminants normally expected
in ballast water and the operational environment of the BWTS;
o Diagnostics to enable the local operator to check the functioning of the electrical and
electronic circuitry, as well as the calibration of meters and sensors according to the
manufacturer's specifications;
o An emergency manual override function to be used in the event of failure of the
control unit;
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
o Audio and visual alarms and a recording in the event there is discharge of any
effluent or a component failure whenever the control unit is not fully operational; and
o The capability to print reports and logged data, as applicable, or stored electronically
with printout capability, upon the following events:
the BWTS is started
the BWTS is stopped
an alarm condition develops
normal conditions are restored
manual override is engaged
• In case of a single failure compromising the proper operation of the BWTS, audible and
visual alarm signals are given in all stations from which ballast water operations are
controlled, including, but not limited to, the following conditions:
o Power failure to the BWTS or any subsystem;
o Failure of any sensor, meter, or recording device;
o Hazardous condition detected by control system; and
o Operation outside set points of the BWTS for proper treatment.
3.6 Biological Efficacy Evaluation with Standard Test Organism
Standard test organisms (STOs) shall be used in bench-scale tests to mimic and assess the
efficacy of the ballast water treatment system. Such tests occur in the laboratory prior to full-
scale testing. Recommended STOs are identified in Table 2 along with the recommended
densities to be added to the experimental water in the laboratory experiments.
The viability of STOs used in bench-scale tests should be determined with the following
parameters: using one organism from each size class listed in Table 2, treating the STOs in
conditions identical to the ballast water treatment system being tested (e.g., 18 ppm of sodium
hypochlorite) and following the experimental replication and use of controls as well as the
guidance for synthetic water preparation described in Anderson et al. (2008, which is included in
Appendix B of this document; e.g., tests are run in quadruplicate for bacteria and protists and run
at least in triplicate for zooplankton). If STOs are cultured rather than purchased from a vendor,
the methods described in Anderson et al. (2008) should be followed. Bench-scale tests may be
completed by the test facility or another organization; results should be included in the Technical
Data Package (Section 1.8, Test Results/Qualification Data) that is submitted to the TO following
full-scale testing.
If the STOs identified in Table 2 are unsuitable for use, alternatives may be considered and
utilized with completion of validation experimentation and the concurrence of the VO. Test
facilities wishing to replace any of the recommended STOs with other organisms should conduct
sufficient experimentation and provide evidence indicating a broad resistance to treatments as
outlined by Anderson, et al. (2008). The Anderson research identified the recommended
standard test organisms as a function of biological functional group and salinity. Similar
methods, as described in Appendix B, should be used by the TF to determine replacements for
those STOs.
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Table 2. Recommended Standard Test Organisms for Bench-Scale Testing
Size Class Marine/Brackish Water Fresh Water „ , ,. \
Concentration
Zooplankton
Protists
Bacteria
Artemia franciscana
Tetraselmis sp.
Geobacillus sp.
Ostracod
Green microalgae
Geobacillus sp.
105 organ! sms/m3
103 organisms/mL
103 organisms/mL
The volumes of water used in the laboratory do not have to match those in the table, but the concentration of
organisms should be equivalent (e.g., 102zooplankton/l is acceptable for 105zooplankton/m3).
3.7 Calibration and Test Requirements
The BWTS will be evaluated during the testing to determine if the system provides:
• Diagnostic routines and procedures to maintain accuracy of measured process
parameters, including:
o The degree to which diagnostics are automated;
o If self test routines are incorporated as part of the control unit;
o If the manufacturer specified appropriate diagnostic intervals; and
o If the diagnostics confirm that parameters are within specifications or that
calibration is required.
• Diagnostics for fault checking, system maintenance and repair;
• Automated diagnostics that also may be manually initiated by the operator;
• Diagnostics that isolate faults down to field replaceable units (FRUs);
• If the accuracy of the system components that take measurements are verifiable according
to the manufacturer's instructions; and
• If only the manufacturer or persons authorized by the manufacturer do the accuracy
checks.
3.8 System Documentation Evaluation
The documentation provided for the BWTS will be evaluated during verification to determine if
the specifications provide detailed requirements and tolerances for the following system
parameters:
• Ballast water turbidity, pressure, temperature and flow rate ranges (include any other
applicable criteria);
• Electrical power requirements;
• Air/pneumatic pressure and flow rate ranges;
• Weight;
• Dimensions;
• Environmental limitations (e.g., ambient temperature);
• Treatment limitations;
• Safety hazards; and
10
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
• The vendor provided list of procedures for unpacking and verifying contents of shipped
items.
The documentation of the installation procedures and requirements in the installation guide will
be evaluated to determine if:
• All areas of mechanical, electrical, hydraulic, pneumatic, and any other interface
requirements are addressed;
• Time estimates in man-hours provided for installation procedures are appropriate;
• If applicable standards are referenced and special precautions and hazards identified; and
• Appropriate diagrams, photographs and/or assembly drawings detail footprints,
attachment points, interfaces, and any referenced components or subassemblies.
The adequacy of the O&M manual(s) provided with the system will be evaluated during the
verification. If not included in the O&M manual, ancillary documentation provided with the
BWTS will be evaluated for the detail provided for the following items:
• Piping and instrumentation diagrams;
• Electrical schematics and wiring diagrams;
• Photographs;
• Guides for diagnostics and troubleshooting;
• Parts lists; and
• Operator training - minimal additional special training required to operate the system
(identified and supplied).
3.9 Technical Data Package Submission
A technical data package must be submitted to the TO by the Vendor of a BWTS to be
considered for verification. Vendor-specific performance claims should be identified along with
relevant existing performance data.
The information in the technical data package should demonstrate that the treatment processes
are well characterized and the equipment is designed to meet specific ballast water treatment
performance criteria at the intended operational scale. Photographs with appropriate reference
scales should be included. The data package shall also document operational and maintenance
requirements and conditions. At a minimum, the technical documentation provided by the
Vendor should address the items identified in the format outline in Section 3.9.
Much of the required information will likely be available in the Vendor O&M manual(s), which
are part of the required documentation. The information presented in an O&M manual will,
however, vary by vendor. To be considered for verification testing under this protocol, vendors
are required to submit a technical documentation package. This allows each vendor the
opportunity to incorporate those data most appropriate to the content topic. In addition to the
technical data package and the O&M manual(s), vendors may also provide ancillary reference
information through any combination of manuals, product literature, and electronic files. Any
11
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
ancillary information or proprietary information must be clearly identified as such, and the
intended purpose/relevance of providing the information must be clearly stated.
While not required for verification, but likely to be part of a submittal for regulatory compliance,
the manufacturer may provide certifications or quality assurance documentation for all vendor
QA/QC and factory testing that occurs during the manufacture of the equipment. If provided,
relevant standards traceability data should also be provided.
3.10 Format for the BWTS Technical Data Package
A Cover Page
B Table of Contents
C General Description & Capabilities (Marketing and technical specifications, and other items
below)
C. 1 System volume, weight, power & mechanical interface requirements
C.2 Vendor performance objectives (vendor should describe primary and non-primary
objectives of ETV testing, i.e., verification testing, or full scale evaluations)
D Target operating environments and conditions
D.I General Features
D.2 Permitting and Certifications
D.3 Scalability (no specified requirement - please address range of applicable ballast system
volumes and rates for the described treatment system)
E Installation Requirements and Instructions
E. 1 Hydraulic and mechanical connections
E.2 Electrical connections to mains
E.3 Hazard locations
E.4 Other special installation criteria / handling
E.5 Considerations for maintenance / consumables / repair
E.6 Shipping and delivery considerations (no specified requirement - vendor should
describe ability / methods to transport treatment system)
E.7 Interfacing for performance monitoring, alarms & controls (no specified requirement -
vendor should describe available options)
F Operating and Maintenance Instructions
F.I Operating and Maintenance Manual (may provide as standalone document(s), but - any
references in the text of the technical data package to the separate O&M manual
must/should be specific to page and paragraph.)
F.2 Training Materials
F.3 Repairs and Troubleshooting
F.4 Recommended Spares (and sources)
F.5 Safety Precautions and Issues
F.6 Environmental Hazards and Issues, Including By-Products
F.7 Expendables, Materials Handling, and Waste Disposal
F.8 Technical Support contact information
G System Performance Specifications
12
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
G. 1 Discharge water quality
G.2 Treatment capabilities vs. environmental conditions
G.3 Control features and capabilities
G.4 Factory testing criteria and procedures (for entire system and ancillary equipment)
G.5 Human operator requirements (special skills or training required to operate the system)
G.6 Data Storage
G.7 Automated capabilities
G.8 Alarms and safety capabilities
H Calibration and System Test Procedures
H. 1 Diagnostics
H.2 Quality assurance during operation
H.3 Calibration schedules and procedures
I Detailed Description of System Operation
1.1 Theory, processing and principles of operation (no specified performance requirement -
vendor should provide background on how and why treatment system works, including
explanation of any environmental limiting factors)
1.2 Selection of materials used in fabrication
1.3 Design considerations for marine applications
1.4 Ancillary Documentation Package (this section is for documentation not referenced
elsewhere)
1.5 Reference drawings and photographs
1.6 Materials / parts lists
1.7 Certifications (such as American Bureau of Shipping certifications)
1.8 Test Results / Qualification Data (no specified requirements - this should be results of
vendor and/or independent testing of system performance)
13
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 4
Treatment Verification TQAP Development
4.1 Description of Ballast Water Treatment System
Each ballast water treatment verification test will be completed following a written TQAP. From
the vendor-supplied treatment system documentation submitted as outlined in Chapter 3, the
TQAP should include those materials, data, and information that are necessary to describe the
treatment system's principle of operation, physical properties, installation and commissioning,
startup and operation, data collection, required actions during upset conditions and necessary
consumables. These may include, but are not limited to:
• Vendor treatment and operation claims as identified in Section 3.2
• Engineering description
• Process description including performance ranges and expectations
• Discharge characteristics
• Footprint
• Photographs
• TO physical and electrical interfaces
• Safety and Environmental Hazards and Precautions
4.2 Required Elements of the TQAP
The TQAP will detail test objectives, specific test procedures (including sample and data
collection, sample handling, analysis and preservation) and quality control and assurance
requirements (including measures of precision, accuracy, comparability, and representativeness).
The experimental approach for the ballast water treatment test, treatment system start-up, and
verification procedures will be presented. The TQAP will include a summary description of the
standardized water quality and biological challenge conditions established by the experimental
configuration as described in Section 5.3. The TQAP will summarize how the challenge
conditions will be implemented at the TF relative to the ballast water treatment system being
tested. Any modifications or supplements to the treatment verification protocols will be defined
and discussed in the Plan. The TQAP will also address quality assurance/quality control
(QA/QC) requirements, data handling and presentation, and environmental, health, and safety
issues.
The TO, with input from the vendor, is responsible for preparing the TQAP. If the vendor desires
data from ETV testing to be made available for type approval or other regulatory purposes, the
data required should be clearly identified in the TQAP. The VO shall review and coordinate the
approval of the TQAP prior to the start of verification testing.
14
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
The TQAP shall include:
• Title page/approval page with all project participants
• Table of contents
• Project description and treatment performance objectives
• Project organization and personnel responsibilities
• TF description
• Treatment system description
• Experimental design (including installation/start-up plan)
• Challenge water conditions and preparation (including TF standard operating procedures
(SOPs) for preparation)
• Sampling and analysis plan including sampling and analytical procedures
• Data management, analysis and reporting
• Environmental, health and safety plan
• References
• Appendices
o Quality Assurance Project Plan (QAPP)
o Vendor operation and maintenance manual
Content requirements for the QAPP are discussed in more detail in Appendix A.
15
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 5
Experimental Design
The primary purpose of ETV verification of BWTSs is to verify the biological treatment
performance according to an established protocol and specified challenge conditions identified in
an approved TQAP. Other factors pertinent to the treatment system's performance will also be
evaluated, including engineering and environmental metrics. To enable purchasers and other
stakeholders to make informed choices in selecting appropriate treatment systems, land-based
verification testing conducted in accordance with this protocol is intended to provide comparable
data sets for each technology or system to the maximum extent practical Standardized challenge
conditions included in this protocol address both water quality and the biological organism
concentrations used to evaluate treatment performance. Key water quality challenge conditions
are standardized under this protocol because the effectiveness of various treatment processes
may be influenced by certain water quality characteristics (e.g., salinity, turbidity, color, etc.).
Moreover, the natural environment (as would be encountered during shipboard BWTS
performance testing) has a large range of conditions, which may or may not provide adequate
information on a system's ability to perform in accordance with the Vendor's specifications
under non-ideal water quality conditions. Therefore, non-ideal water quality conditions form the
basis for challenging the treatment systems under this land-based verification testing protocol.
To this end, the protocol also includes the requirement for vendors to produce BWTS test data
using STOs at an appropriate scale in a controlled environment, and verification testing using
robust ambient species during full-scale tests to measure biological treatment efficacy.
The general objectives of the verification testing are to:
• Provide a comprehensive set of water quality and biological challenge conditions against
which treatment effectiveness can be quantitatively evaluated.
• Develop adequate data to document system performance against the verification factors.
The requirements for testing are described in the following sections, which provide guidance on
the four key elements of the protocol: 1) verification factors, 2) water quality and biological
challenge conditions, 3) the TF experimental configuration, and 4) verification testing, including
commissioning of the equipment and the measurement programs required under this protocol.
Variations in the protocol for specific treatment system types (e.g., in-line treatment versus in-
tank treatments) are also described.
5.1 Verification Factors
All treatment systems will be verified according to the following factors:
• Biological treatment efficacy
• Operation and maintenance
• Reliability
• Cost factors
16
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
• Environmental acceptability
• Safety
5.1.1 Biological Treatment Efficacy
Biological treatment efficacy is defined as the removal, inactivation, or death of organisms and
will be measured in terms of the concentration of selected organism size classes in the treated
discharge. Additional measures of efficacy may include measurements in terms of removal
efficiency (e.g., a percentage reduction of organisms present at uptake), against a threshold (e.g.,
a water quality standard), or in relation of treatment vs. control discharge concentrations. The
measurement program required by the protocol evaluates the primary treatment efficacy criteria
by measuring the quantity of living organisms in both the challenge water and the treated
discharge.
5.1.2 Operation and Maintenance
Operation and maintenance includes the labor expertise, equipment, and consumables required to
operate the system to achieve the stated performance goals and objectives. The quantitative
indicators to be considered during verification are described in detail in Section 5.4.9.1.
5.1.3 Reliability
Reliability is a statistical measure of the number of failures (either qualitative or quantitative) per
known quantity of test cycles. This is described in greater detail in Section 5.4.9.8.
5.1.4 Cost Factors
Cost factors include only those factors that can be verified, such as labor hours to operate and
maintain the system, expendable material, such as filter cartridges, and pounds or gallons of
chemicals consumed by the treatment system. Data is collected in units, to which unit prices,
which are likely to vary from location to location, can be applied to determine costs. These are
discussed further in Section 5.4.9.2.9.
5.1.5 Environmental Acceptability
Environmental acceptability assesses ballast water quality following treatment for factors other
than the abundance and viability of organisms. For example, this will determine if the treated
water meets acceptable water quality characteristics for such measures as dissolved oxygen,
temperature, treatment residuals, pH, etc. This is discussed in further detail in Section 5.4.4.
5.1.6 Safety Factors
Safety factors include any treatment-specific considerations that may pose a threat to the safety
of the operator or shipboard operations. These are not intended to be comprehensive in nature,
which is best evaluated by Classification Societies, such as the American Bureau of Shipping,
but are included as observations that can be made during the verification testing. Further
discussion of these observations are discussed in Section 5.4.9.2.11.
Performance test results will be reported using standard ETV formats to make certain the
reported information among treatment technologies tested is comparable. Flexibility is
permissible to ensure reporting for a specific treatment system type is appropriate and accurate.
17
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Some information supplied by vendors may not be verified under the protocol. This information
may be included in the verification test report and clearly identified as non-verified information.
Vendor-provided information may include shipboard compatibility (e.g., corrosion resistance,
system weight, system volume including clearances needed to perform maintenance and replace
vital components, and compatibility with other common shipboard systems such as operational
flow rates). Submission and reporting requirements for non-verified, Vendor-supplied
information is included under Chapter 3.
5.2 Challenge Conditions
This protocol recognizes that land-based testing cannot fully replicate actual treatment system
performance onboard ship. However, land-based verification testing can provide sufficient
information to verify the expected performance of treatment in the shipboard environment. It is
understood that all treatment technologies will face a range of physical/chemical water quality
conditions and biological organisms when operated onboard a ship. Therefore, each treatment
system's performance will be verified using a set of standard challenge conditions. This protocol
defines the following objectives for the challenge conditions:
• To verify a treatment system's performance using a set of challenging, but not rare, water
quality conditions representative of the natural environment.
• To verify removal or kill of bacteria, protists, and zooplankton using ambient organisms
as defined by size classes and analytical techniques that identify living quantities for
these organisms.
The standard challenge conditions are specified using two groups of factors that must be
addressed to properly challenge treatment technologies: water quality and living organisms. The
requirements for each group are presented in the following sections.
5.2.1 Challenge Water - Water Quality Characteristics
Since water quality conditions in ports and harbors around the world vary greatly, treatment
systems may encounter a wide range of conditions. Also, certain water quality conditions may
interfere with the ability of some treatment processes. It is therefore critical to evaluate the
effectiveness of a treatment system under water quality conditions that are challenging to the
technology being tested. Simulating all potential water quality conditions in a land-based testing
design would be prohibitively expensive1 and not essential for verifying the performance of a
treatment system. Because water quality parameters that can interfere with various treatments
are generally understood and few in number (e.g., salinity, turbidity, organic matter either as
dissolved or particulate forms), the number of water quality metrics that must be explicitly
included in the protocol can be limited. This protocol defines three possible challenge conditions
that represent some of the more challenging, natural conditions that may be encountered by
Similarly, shipboard testing of all potential water quality conditions will require extensive logistics to move a
treatment system to a matrix of natural conditions, as well as investment in methods and protocols by which the
treatment effectiveness is established using natural populations of organisms.
18
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
ballast water treatment systems. Challenge water quality characteristics to be used during testing
events are presented in Table 3.
Table 3. Water Quality Challenge Matrix for Verification Testing
Water Types Minimum Water Characteristics
Fresh (Salinity <1 PSU) DOM: 6 m§/L as DOC
POM: 4mg/LasPOC
Brackish (Salinity 10-20 PSU) MM: 20 m§/L
TSS=POM + MM: 24 mg/L
Marine (Salinity 28-36 PSU) Temperature: 4-35°C
Another basic premise in the design for this protocol is that ballast water treatment systems are
designed to function effectively in the full range of water quality characteristics that will be
encountered under shipboard operational conditions. By challenging the treatment systems with
these conditions, it is assumed treatment will be effective under less demanding conditions.
Challenge waters have been tailored to a minimum set of water quality conditions that may be
achieved either through naturally occurring conditions or through augmentation, if appropriate,
and validated by the TF. The challenge conditions are specified for three possible levels of
salinity, <1, 10 to 20, and 28 to 36 PSU (practical salinity units), and water quality characteristics
problematic for the range of technologies being developed to treat ballast water, namely,
suspended solids and dissolved organic matter (DOM).2
Suspended solid material that can interfere with treatment effectiveness is composed of several
types of particles, which can be of biological or mineral origin, specifically clay and silt. The
water quality challenge conditions defined by the solids content of the matrix include particulate
organic matter (POM) and mineral matter (MM). These two types of particles are both present
in natural waters at a range of concentrations and size distributions. Therefore, both forms are
included in the challenge conditions to address issues of parti culate removal and turbidity, which
can interfere with transmission of UV light or other treatment processes.
Various forms of dissolved chemicals and compounds, particularly organic material, can directly
affect the efficiency of some treatment processes. Dissolved organic matter (DOM) and
dissolved organic carbon (DOC) are two terms used to describe this component of natural water.
2
The protocol does not explicitly call for verification at a series of temperatures, even though some treatments may
have strong temperature dependence or may include temperature manipulation as part of the treatment procedure.
Rather than include temperature as a controlled water quality condition, which can have significant cost implications
for the TF, accurate and continuous monitoring of the source and treated water temperatures is required for all test
cycles. If temperature manipulations are to be included, the Test Plan will include protocols for these
manipulations. Temperature challenges should be addressed in shipboard testing, and in bench-scale tests of
treatment process.
19
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
DOM/DOC often contains many chromophores that contribute substantially to the color of the
water, another potential interference for treatments. Thus, the color of a water and DOM/DOC
concentration are often interrelated.
The measurement methods for evaluating the status of the challenge water quality conditions are
described in Section 5.4.6. They include standard analytical methods to document the
concentration of total suspended solids, paniculate organic matter or dissolved organic matter,
and methods that indirectly measure these parameters (e.g., turbidity measured by
electronic/optical measurement such as nephelometery (NTUs) or transmissometry (beam
attenuation) or fluorescence (color /DOM)).
Standardization of the water quality conditions for the verification testing requires a consistent
set of source water (e.g., fresh or marine water), as well as the use of well-characterized organic
matter and mineral matter. The TF will be responsible for providing these materials and
ensuring the water quality conditions are as described under this protocol. The water quality test
conditions will be standardized for salinity, particulate organic matter, mineral matter, and
dissolved organic matter as described in the following sections:
5.2.1.1 Salinity
Natural water of less than 1 PSU will be used for fresh water conditions, while natural seawater
will be used for marine conditions. Testing at multiple salinities at a given TF should only be
conducted if there are natural water sources with the differing salinities (e.g., fresh and brackish
waters). Artificial modification of the salinity of the waters should be used only if it can be
demonstrated that the concentrations, diversity and condition of organism populations required in
Section 5.2.2 will not be impacted by adjustment of the salinity.
5.2.1.2 Particulate and Dissolved Organic Matter
In the case of POM, if the natural waters have insufficient concentrations, the TO may augment
them through the addition of humic material (e.g., Micromate humates [Mesa Verde Resources,
Placitas, New Mexico]). Other sources include particulate carbon from sources such as ground
up seaweed or plankton detritus. DOM can be very difficult to adjust or augment if the natural
waters have insufficient content. There has been some success using Camellia sinesis
(decaffeinated iced tea mix) to augment natural DOM content, but a TF must assess the effect of
additives on the ambient and test organisms (if used) before using.
5.2.1.3 Mineral Matter (MM) - Clays and Silts
Mineral particles in the size range typically found in coastal and estuarine waters are readily
obtained from commercial sources and will be used as the source of the mineral matter. A study
of sediment size in ballast tanks suggests that particles are mostly fine grained (less than 63 jim)
and most vessels contain <10% sand (F. Dobbs, Pers. Com.). Thus, addition of the commercially
available clay minerals (with a majority of particles in the 10 to 50 jim size) addresses the
objective of having a prescribed level of non-biological particles as part of the water quality
challenge conditions. Specifically, ISO 12103-1, A3 MEDIUM TEST DUST and ISO 12103-1,
A4 COARSE TEST DUST can be used for this purpose. As these particles will tend to settle out
over time within the augmentation storage vessels, the test protocols must include a means of
20
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
maintaining any sediments in a homogeneous suspension prior to addition of the challenge water
(e.g., continuous mixing of the sediment augmentation tank).
The TO should verify that, whatever source of augmentation or delivery system is used, the
addition of that material should minimize to the extent possible biocidal or growth stimulant
response to the ambient organisms. The TF will be responsible for preparing the challenge water,
documenting the challenge conditions, and validating that the conditions are maintained at the
treatment system or control entry point. Challenge waters will be prepared under standard
operating procedures developed by the TF. The TQAP will include these SOPs and describe any
planned deviations from the SOPs.
5.2.1.4 Challenge Water- Water Quality Deviations
In some cases, a specific ballast water treatment system may be unable to operate with all of the
prescribed challenge water quality conditions as specified in Table 3. This may be either due to
mechanistic limitations of the technology (e.g., electrolytic chlorination (without brine addition)
is inoperable in fresh water) or by design (e.g., scale). In such cases, deviations may be
permitted provided that significantly challenging and realistic conditions are identified and
justified by the TO, and that the VO approves the deviation. In no case, however, shall the total
number of test cycles be reduced. All deviations will be specified in the verification report as
limiting conditions of the technology.
5.2.2 Challenge Water - Biological Organism Conditions
The death or removal of living aquatic organisms is central to the need to treat ballast water. To
ensure proper evaluation of a BWTS's performance, the effects on biological organisms living in
the challenge water will be measured for each treatment system tested. Biological efficacy will
be evaluated as function of a system's ability to kill or remove organisms that are naturally
occurring and represent the more robust ambient populations at the test site.
5.2.2.1 Organism Concentrations
A minimum total input concentration of living organisms, by size class, is defined in Table 3.
The two larger size classes must contain at least 5 different species from at least 3
phyla/divisions. Challenge water meeting these criteria shall be demonstrated for each test cycle
at 1) the influent point of the control tank and 2) immediately prior to the point of treatment for
systems that treat upon uptake or at the treatment tank influent point for systems that treat either
wholly in-tank or upon discharge.
21
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Table 4. Minimum Criteria for Challenge Water Total Living Populations
Organism Size Class 1 Total Concentration Diversity
> 50 jim 105 organ!sms/m3 5 species across 3 phyla
> 10 jim and < SOjim 103 organisms/mL 5 species across 3 phyla
<10 jim 103/mL as culturable aerobic n/a
3
r\
heterotrophic bacteria
Size is determined by the maximum dimension on the smallest axis.
2
Note it is assumed that the effects on culturable aerobic heterotrophic bacteria will be indicative of the effects on
all bacteria.
3 Diversity of bacteria by species or phyla is not applicable, and there is no diversity requirement for this size class.
5.2.3 Challenge Water - Flow Rates and Volumes
5.2.3.1 Flow Rate
Treatment tests will evaluate equipment at operational flow rates defined by the vendor's O&M
manual. The TF shall be capable of providing flow rates of at least 200 m3 per hour (880 gallons
per minute) and an available volume per test cycle of at least 400 m3. The TF shall provide
sufficient challenge water volume to meet these requirements, and the TQAP will identify the
rates that will be tested.
5.2.3.2 Volume
A minimum of 200 m3 shall be processed in each BE test cycle. The recommended minimum
volume for in-tank testing is 200 m3 (-52,800 gallons). The TF shall provide test and control
ballast tank configurations of at least 200 m3. Larger volumes may be used depending on vendor
specifications and availability of tanks at the TF.
5.3 Test Facility Physical Configuration
5.3.1 Overall experimental configuration
As a minimum, the TF should encompass four components: (1) fluid delivery capacity, (2) a
control tank and piping system, (3) a treatment tank and piping system, and (4) a discharge
collection tank and post-test treatment system. The fluid delivery systems include pumps,
piping, flow distribution controls, flow rate controls and relevant instrumentation to support the
challenge water requirements described in Section 5.2. The control tank shall be utilized to hold
untreated challenge water for each biological efficacy test cycle. The treatment tank will be
utilized to hold all test water subject to the BWTS during the test cycle. Both tanks shall be a
minimum of 200 m3 and suitably constructed to hold such volumes for at least one day. The tank
drains shall be located, to the extent possible, to minimize the retention of water following
discharge. Tank intake and discharge piping, fittings and relative configurations shall be
identical or the equivalent as validated by the TO. Finally, the discharge tank may be necessary
if the TF is required to post-treat on-site the control and treated challenge or test waters to
remove added inorganic and organic matter, disinfection by-products, or other regulated
22
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
substances prior to discharge back to the environment. The discharge tank should be of
sufficient volume to store at least 200 m3, but preferably large enough to store the cumulative
volume of the control and treatment tank.
There are multiple potential locations of ballast water treatment systems when used onboard
vessels. The TF must be arranged to support testing of systems, which operate at uptake,
discharge, in-tank or a combination of these. Examples of the arrangement for in-tank and in-
line treatment are shown in Figures 1 and 2. As shown, the test configuration includes a flow-
splitter such that challenge or test water is provided to both the control and treated legs
simultaneously. Note that in such an arrangement, the fluid pumping capacity of the TF would
be a minimum of 400 m3/hr. A sequential fill configuration may also allowable, in which the
treatment and control are filled or drained sequentially. The latter may result in reduced pump
capacity needs (but still requires a minimum pump capacity of 200 m3 per hour), less overall
logistic complexity, and reduced piping through the dual use of sampling apparatus, feed and
discharge plumbing, instrumentation and so on. In either case, the TF shall validate, to the
satisfaction of the VO, that significant differences between treatment and control lines in
biological and physical responses are minimal, and that there is no cross contamination by dual
use of the site infrastructure.
5.3.2 Sampling Methodology
Several types of samples are to be acquired during the verification testing of a ballast water
treatment system. During biological efficacy tests, discrete samples for water quality and
biological enumeration shall be acquired over the course of the test on a time averaged basis. A
minimum volume of 3 m3 shall be collected per location. In situ instrumentation to monitor water
quality and physiochemical parameters are also included. All sampling is assumed to be in-line,
whether discrete or in situ. Characterization of ambient waters may require discrete grab
samples, as described in Table 8
5.3.2.1 Sampling Locations
Required sample locations for various treatment scenarios are shown in Figures 1 and 2; samples
should be collected according to one of these test designs, unless otherwise accepted by the VO
in the TQAP. Samples (data) from the challenge water must be obtained, in accordance with the
guidance in Section 5.3.2.4, immediately prior to water entry to the control tank, and
immediately before entry to the BWTS (in-line treatment) or the ballast tank in the case of in-
tank treatments (if demonstrated as representative of the control and challenge water, a single
sample collected ahead of the splitter shown in Figures 1 and 2). For in-line BWTSs, samples of
treated water must be collected (1) immediately following the treatment system and (2)
following the holding tank at the end of the one-day hold time. For in-tank treatments, samples
of treated water must be collected from the ballast tank discharge following the vendor-defined
contact period. Further definition of hold times is described in Section 5.4.5. Systems that
incorporate treatment at multiple locations (e.g., upon uptake and discharge) will only require
sampling after the final stage of treatment. Sampling locations for the control tanks and BWTS
must exactly mimic the treatment tanks and system. Finally, in-tank sampling (e.g., via plankton
net tows) shall not be utilized for the purposes of verifying biological efficacy, as this method
may not result in representative samples. The exact locations, frequency, and methods to be used
to collect the samples must be defined in the TQAP.
23
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
5.3.2.2 Sample Collection Requirements - Frequency
Continuously recording in situ sensors (as available and feasible) may be used to measure water
quality and proxy parameters during verification testing. Description of the sensors, how they
operate, and how they are calibrated shall be included in the TQAP. Minimum instrument
performance requirements are provided in Table 5. Discrete samples for water quality
characterization will also be obtained during verification testing as discussed above, and they
will be collected at the time biological verification samples are collected. A higher frequency of
collection for discrete samples may be used if additional samples for calibrating the sensors are
necessary. The sample collection requirements and frequency of obtaining samples from the
control tanks and piping system will identically match those of the treatment tanks and system.
The appropriate frequency of discrete sample collections made in lieu of in situ sensing shall be
described in the TQAP.
Table 5. Accuracy and Precision Requirements for Potential Sensors
Sensor Reporting Units Range Accuracy Precision
Temperature
Conductivity (salinity)
Transmissometer (20-cm)
Dissolved oxygen
Fluorometer
°C
MS/cm
perm
mg/L
Hg/L
Oto30
0.5 to 65
Oto40
Oto20
0.03 to 75
0.1
0.1
0.20
0.10
50% of reading1
0.01
0.01
0.01
0.05
0.01
1 When compared to wet chemistry results.
5.3.2.3 Sample Replication
Verification testing will include replication only in analysis. Sample collection replicates are
based on the time integrated sample volumes collected during the test cycle (see examples shown
in Figures 1 and 2. These sample volumes form the minimum sample collection replication
required during each test cycle. Each of the integrated sample tanks will be sub sampled for the
Core parameters, which are discussed later. The TQAP will describe each type of analytical
replication planned, including acceptable ranges of variability.
5.3.2.4 In-line Sampling for Biological Efficacy
To obtain an accurate measurement of the organism concentration at the sample location, the
installation of an isokinetic sampling facility at each of these locations is recommended.
Isokinetic sampling is primarily intended for the sampling of water mixtures with secondary
immiscible phases (i.e., sand or oil) in which there are substantial density differentials. In such
conditions, convergence and divergence from sampling ports is of significant concern. Since
most organisms are relatively neutrally buoyant, true isokinetic sampling is likely unnecessary
for testing ballast water treatment systems. Nonetheless, the mathematics related to isokinetic
sampling are deemed useful for describing and specifying appropriate sampling geometries.
Isokinetic sampling is necessary to ensure that a sample contains the same proportions of the
various flowing constituents as the flow stream being sampled. During isokinetic sampling the
24
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
sampling device does not alter the profile or velocity of the flowing stream at the moment or
point at which the sample is separated from the main flow stream. To achieve isokinetic
sampling conditions, a sampler is designed to separate a subsection of the total flow-stream in a
manner that does not encourage or discourage water entry other than that which is otherwise in
the cross-section of the sampler opening. In other words, flow streams in the main flow of the
pipe should not diverge or converge as they approach the opening of the sampler.
Recommendations for the design and installation of appropriate sampling facilities are given
below. In any case, validation of the Test Facilities configuration should include verification that
the chosen sampling design, geometry and installation result in representative samples and
minimize organism mortality as a result of sample acquisition.
5.3.2.5 Design of In-line Sampling Apparatus
Through computational fluid dynamics modeling, it has been shown that the isokinetic diameter
calculation can provide guidance for sizing of sample ports for sampling of organisms (Richard
et al., 2008). Simulations showed that flow transitions from the main stream were best for sample
port diameters between 1.5 and 2.0 times the isokinetic diameter. Ports sized in this range had
smooth transitions and pressure profiles that allowed for direct sampling without the need of a
pump to induce sample collection. The isokinetic sample port diameter should therefore be
determined generally according to the equation:
= D.
25
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Challenge Water 450 tons
Challenge Water
Sample
Ballast Tank
> 200 m3
>30m3
Treatment Sample Tank
Flow
Splitter
Standard Test Organisms
Discharge
Control Holding
Tank
> 200 m3
Discharge
Control Sam pie Tank
Figure 1. Sampling design example for in-tank treatment.
26
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Challenge Water 450 tons
Challenge Water
Sample 3 ms
Post Treatment
Sample
(> 30 m3)
Treatment
Holding Tank
> 200 m3
>30m3
Post Treatment
Pump
t
Flow
Splitter
Standard Test Organisms
Holding Sample Tank Discharge
Figure 2. Sampling design example for in-line treatment.
Discharge ControlSample
27
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
where D;so and Dm are the diameters of the sample port opening and the main flow in the line to
be sampled, respectively; and Q;so and Qm represent the respective volumetric flow rates through
the two pipes. It is recommended that sample port size be based on the combination of maximum
sample flow rate and minimum ballast flow rate that yields the largest isokinetic diameter.
The opening of the sampling pipe should be chamfered to provide a smooth and gradual
transition between the inside and outside pipe diameters. The length of the straight sample pipe
facing into the flow can vary, but it should not usually be less than one diameter of the sampling
pipe. The sampling port should be oriented such that the opening is facing upstream and its
entrance leg flow is parallel to the direction of main pipe flow and concentric to the larger pipe,
which may require sampling pipes to be "L" shaped with an upstream facing leg, if installed
along a straight section of discharge pipe.
The need to be able to service the sample pipe is important and should be considered, taking
safety into consideration. Therefore, the sampling pipe should be retrievable either manually or
mechanically, or it must be in a system that can be isolated. Because of the potential for the
opening and interior of the sample pipe to become occluded by biological or inorganic fouling, it
is recommended that samplers be designed to be closable at the opening, removed between
sampling intervals, or be easily cleaned prior to sampling.
The sample pipe and all associated parts of the sampler that come into contact or near proximity
with the system piping should be constructed of galvanically compatible materials and generally
corrosion resistant. Any corrosion of the sampling system will affect sample flow rates and
potentially sample representativeness.
If flow control of the sample flow rate is required, ball, gate, and butterfly valve types should be
avoided as they may cause significant shear forces, which may result in organism mortality. For
flow control, it is recommended that diaphragm valves or similar valve types be used to
minimize sharp velocity transitions. For flow distribution, ball valves may be utilized only if
they are either fully open or fully closed
When sampling is conducted on the discharge of a tank through the use of a pump (i.e., a non-
gravity drain) and the sample port is located upstream of the pump, it may not be possible to
draw an adequately sized sample since the line will be under suction with a variable hydrostatic
pressure head. Therefore, maintenance of a time-averaged sample flow requires the sample to be
drawn from the discharge utilizing a pump. In such cases, a diaphragm pump is recommended to
minimize pump-induced organism mortality during sampling.
5.3.2.6 Installation of an In-line Sample Point
The sample taken should be removed from the main pipeline at a location where the flowing
stream at the sample point is representative of the contents of the stream. The sample port
entrance should be placed at a point where the flow in the main pipe is fully mixed and fully
developed.
The sampling point should be installed in a straight part of the system piping and the sampling
fixture should be positioned such that a representative sample of ballast water is taken. It is
28
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
recommended that the position of the sample point be determined using methods such as
computational fluid dynamics.
5.3.2.7 Operation of an In-line Sample Points
In-line biological samples will be collected on a time-integrated basis such that a composite
sample of the entire period of uptake or discharge is acquired. The sample flow rate should be
appropriately controlled to maintain an even distribution of samples acquisition over that time
period.
5.3.3 Test Organism & Water Quality Augmentation
Where the addition of organisms for the augmentation of ambient organisms) is required for
biological efficacy testing, a method for the injection or addition of organisms to the challenge
water must be provided. Similarly, water quality parameters that require adjustment from the
ambient conditions to the requisite challenge water properties will require some type of injection
process. Various means are available to inject or add organisms to the challenge water, for
instance, by a batch method to a large, discrete source volume or by injection into the flow
stream. In any case, the following requirements are applicable:
• Any organism addition or injection method must minimize, to the extent possible,
organism mortality as a result of its addition/injection mechanism.
• The method must result in a well-mixed and uniform distribution, spatially and
temporally, of organisms within the challenge water and at its introduction at the point of
treatment or tank intake.
• The concentration of living test (if used) organisms at the point of treatment or tank
intake must conform to the requirements given in Section 5.2.2.
• The point of addition or injection must be situated such that the flow is well mixed at the
subsequent point from which the first discrete sample is acquired to ensure a
representative sample is obtained; inclusion of substantial pipe lengths and/or a static
mixer may need to be considered.
• All methods for the injection or addition of organisms must be validated by the TF to
meet the above requirements.
For water quality additions, for example sediments or dissolved organics, the addition should
occur far enough upstream from the point of water quality sampling to ensure that the sample is
well mixed. Furthermore, the apparatus used for addition should minimize the system related
mortality on ambient and test organisms, to the extent possible.
5.3.4 Control & Instrumentation
The testing described throughout this protocol is complex and logistically challenging.
Moreover, since these tests are designed to provide a repeatable and accurate verification of
treatment system performance, it is important to ensure that each phase and measurement is
conducted with a high degree of reliability, repeatability, and accuracy. The verification testing
process is further complicated by the inclusion of biological organisms and related
measurements, which result in a variety of design and timing complexities. As a result, it is
recommended that the TF include a typical supervisory control and data acquisition (SCADA)
29
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
system to support the many operations and data acquisition tasks associated with this testing. A
typical SCADA system provides the TF with the ability to: 1) provide automatic control of
pumps, valves and sub-systems to maintain operational set points; 2) acquire and archive all
events, data and conditions; 3) provide controllable process control algorithms which improve
system efficiency, safety and repeatability; and 4) provide facility- and treatment-system
diagnostics during commissioning, testing, and upset conditions. Instituting such a system can
be expected to improve measurements, quality assurance, standardized reporting, and reduce
labor and analysis time.
Whether a SCADA system is utilized or not, the TF should include within its QAPP a discussion
of how the instrumentation associated with TF operation, process control (either manual or
automatic), and condition monitoring of the verification tests shall be operated, maintained and
calibrated. Also, as a minimum, the TF shall include sufficient instrumentation and condition
monitoring such that a substantive record is established which verifies that 1) challenge
conditions were obtained and maintained, 2) the treatment system was operated in accordance
with the Vendor's requirements and 3) no system or environmental effects occurred to perturb
the verification test or treatment system operation. The test instrumentation and test operating
procedures shall be documented in the TQAP.
5.4 Verification Testing
Verification testing will be separated into three distinct phases, 1) treatment system
commissioning, 2) biological efficacy (BE) tests, and 3) operating and maintenance (O&M)
tests. Commissioning tests are intended to validate, prior to the commencement of either BE or
O&M tests, that the treatment system is installed correctly and operating in accordance with the
vendor's requirements. A minimum of three BE tests shall be completed at each of two salinities
selected by the vendor (the vendor may complete testing at all three salinities if desired) and with
all of the challenge conditions specified in Section 5.2 to assess and verify the biological efficacy
of the treatment system under pre-established conditions. O&M testing shall be conducted with
ambient source water conditions, with the intention of operating the system with realistic
physical conditions, to assess the systems engineering performance.
Ballast water treatment system performance, operating conditions, and certain O&M criteria will
be recorded and monitored during verification testing by the TO. Results will be presented in the
verification report, described in Chapter 6. The factors to be verified during ballast water
treatment system verification testing include: biological treatment performance, operation and
maintenance, predictability/reliability, cost factors, environmental acceptability, and safety.
Any of several treatment sequences may be used by a particular treatment system (see Table 6),
including in-line treatment (during ballasting or deballasting), in-tank treatment, or a
combination of the two. The stage in the ballasting cycle at which treatment is applied may also
vary. This verification testing protocol accounts for these through flexibility in the TF and
Verification TQAP. The guidance in the following section provides the basic test requirements
and rationale for inclusion in the TQAP that will provide details specific to the treatment system
and its operation. The final verification report shall document the system sequence(s) completed
during testing.
30
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Table 6. Likely Treatment Sequences and Applications Inherent to Ballast Operations
Sequence Number Ballast Operation Application
1 Treatment applied during ballasting/ No treatment during
deballasting.
2 Treatment applied during ballasting/ Treatment applied during
deballasting.
3 No treatment applied during ballasting/ Treatment applied during
deballasting.
4 No treatment applied during ballasting/ Treatment applied during
transit/ No treatment during deballasting.
5 No treatment applied during ballasting/ Treatment applied during
transit/ treatment during deballasting.
6 Treatment applied during ballasting/ Treatment applied during
transit/ No treatment applied during deballasting.
7 Treatment applied during ballasting/ Treatment applied during
transit/ Treatment applied during deballasting.
The over-arching objectives of the verification testing (including all phases) are to:
• Evaluate the treatment performance of the ballast water treatment system relative to the
removal or kill of ambient and test organisms (if used), operating under vendor-specified
conditions;
• Evaluate the treatment system O&M criteria;
• Determine and record cost factor data; and
• Record and document test conditions, observations, and results.
Other testing objectives may be defined by the vendor and included in the TQAP. The
requirements for verification testing are described in the following sections and must be
addressed in the TQAP.
5.4.1 Treatment System Commissioning
The TQAP shall describe all the tests and start-up requirements required to validate that the
treatment system is installed correctly and operating in accordance with the vendor's
requirements. The objectives of the commissioning are to:
• Install and start the ballast water treatment system in accordance with the vendor O&M
manual;
• Reach stable operating conditions;
• Make modifications as needed to ensure stable operations under TF condition; and
• Record and document all installation and start-up conditions, observations and results.
31
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
The treatment system shall be installed at the TF according to the vendor instructions included in
the TQAP. Ideally, this phase of the verification will include close coordination between the
vendor and TO to quickly resolve discrepancies or malfunctions. Commissioning tests may
include small-scale tests of various vendor sub-systems or components, validation of treatment
system integration into the TF (e.g., a leak test or communication tests), or any other vendor-
required installation tests that may be expected during a shipboard installation. However, at least
one valid, full-scale verification test cycle, meeting all of the requisite challenge conditions,
should be conducted successfully by the TO without vendor assistance. While the challenge
conditions are to be employed during commissioning tests, it is not necessary to conduct a
complete suite of analytical measurements to assess biological treatment efficacy.
A successful commissioning is defined as one in which (1) all TF requirements and conditions
defined by the challenge conditions were met and (2) all components of the treatment system
operated in accordance with the vendor requirements. Subsequent BE and O&M verification
testing cannot commence until commissioning is successfully completed and agreed upon by the
TO and vendor. The verification report should document all of the small-scale, component-level
tests conducted and their results, any treatment system or TF deficiencies or failures, and their
successful resolution. Finally, the verification report should document in detail the challenge
conditions during the full-scale commissioning verification test cycle.
5.4.2 Operation and Maintenance (O&M) Manual
The O&M manual shall be incorporated into the TQAP and will be essential to the development
of the monitoring and maintenance plan incorporated in the TQAP. The vendor shall identify
factors that affect the operation of the BWTS, including any warm up or other requirements that
must be completed to achieve operational stability. The vendor's O&M manual shall specify
what constitutes stable operating conditions for the BWTS, factors that may affect operating
conditions, and any adjustments required to reach or to maintain a stable operating condition.
Adjustments made in the operating conditions will be presented in the final verification report.
5.4.3 Vendor and TO Requirements
An installation/start-up plan shall be prepared and included as part of the TQAP. The TO shall
conduct start-up procedures for the BWTS in accordance with the installation/start-up plan and
with the vendor O&M manual. At the end of the start-up period, the TO will assess whether the
BWTS is in a stable operating state, as specified in the O&M manual, and the vendor will certify
in writing that the system is installed and operating as intended. If the operation is stable, the
verification testing can begin. If not, start-up procedures will be repeated no more than two
additional times. If the BWTS does not achieve stable operating conditions after three start-up
cycles, the TO, in conjunction with the vendor, will review the start-up work plan for
applicability and determine where adjustments and modifications are required. In any case, the
TO will have the option of concluding or postponing further testing at the conclusion of three
failed start-up cycles.
The vendor will identify any additional equipment, system maintenance, changes to operating
conditions, or other modifications needed to ensure effective BWTS operation and to attain or
maintain stable operational conditions.
32
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
5.4.4 Toxicity Testing for Biocide Treatments
The Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), as amended, requires
registration by the U.S. Environmental Protection Agency (EPA) of pesticides sold or used in the
United States, which includes biocide products that might be used in BWTSs. The vendor is
required to provide information regarding FIFRA registration of any biocide to be used in their
BWTS.
The residual toxicity in the discharge from BWTSs employing a biocide is of concern to the TF
(as part of the TF's NPDES permit requirements), as well as for the environmental acceptability
of the treated ballast water from the BWTS in use. Toxicity testing of the water following
treatment and hold time, as appropriate, shall be conducted during the commissioning phase of
the verification testing according to the toxicity methods cited in Section 5.4.7.5. If the post-
treatment effluent passes the toxicity tests, then verification testing can proceed. If, however, the
effluent fails the toxicity test, verification testing shall not be initiated, and further toxicity tests
shall be required. The vendor shall be allowed no more than two additional attempts to pass the
toxicity tests within 30 days of the initial test. This may require modifications to the approach
for verifying the technology in the TQAP or other investigations to understand the toxicity
response. In the event a TF's NPDES permit requires a toxicity evaluation of the treated waters
at the end of each test with the addition of a biocide, or if the vendor requests additional toxicity
testing during the verification, the TQAP shall address the additional testing.
5.4.5 BE and O&M Verification Strategy: Test Duration and Coordination
A minimum of three valid BE tests, described in detail below, are required for each salinity
regime (defined in Section 5.2.1) under which the treatment system is verified. At a minimum,
testing at two salinity conditions shall be conducted. In addition, O&M testing of the treatment
system shall distribute testing of a minimum treated volume of 10,000 m3 amongst the BE test
cycles. These O&M test cycles are equivalent to -50 hours of operation at 200 m3 per hr (or
-65 hours of operation at 150 m3 per hr). Upon completion of the commissioning verification
tests, the next verification test shall be a BE test cycle. This sequence allows the testing to
validate operation of a new unit prior to substantial operational testing. For example, for the case
in which 6 BE test cycles will be conducted, each BE test cycle should be separated by 2,000 m3
in O&M testing. This approach also involves a substantial duration for the testing period and
associated range of ambient water conditions over this time.
During actual shipboard operation, ballasting procedures may occur over time periods ranging
from minutes to hours. For each in-line treatment BE or O&M verification cycle, a minimum
operational period of one (1) hour is required, although this may be extended if flow rates are
reduced from 200 m3/hr.
In addition to the uptake time, a minimum 1-day holding time within both the treatment and
control tank is required for each BE test cycle to simulate the time that water would reside in a
ballast tank. Thus, the duration of each test cycle will be defined by the operational approach
used by the treatment system. The holding time of the required BE test cycles may be extended
if the vendor requires such time as part of the BWTS or process.
33
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
The holding time included in this protocol is intended to provide a conservative assessment of
the BWTS's ability to treat ballast water according to the vendor's claims. For in-tank treatment
with additional in-line treatment during ballast water discharge, the duration will be equal to in-
tank treatment requirements and the deballasting time. Regardless, subsequent to the one-day
tank holding time, the control tank discharge must exhibit a minimum concentration of living
organisms, as defined in Table 7. These criteria are necessary, in addition to the input challenge
conditions, to constitute a valid BE test cycle. These control tank discharge concentrations are
intended to make certain that treatment efficacy measurements attributed to the BWTS in any
given BE test cycle are not the result of natural or non-treatment system related effects.
Table 7. Criteria for Concentrations of Living Organisms in Control Tank Discharge
Organism Size Class Minimum Concentration
> 50 |im 100 organisms/m3
> 10 |im and < 50|im 100 organisms/mL
5 x 102/mL as culturable aerobic heterotrophic
^ bacteria
Shorter or longer tank hold times may be utilized but must be justified in the TQAP.
Justifications for shorter tank hold times may include an inability to sustain organism
populations in the control tank to achieve the requirements in Table 7 because of natural
mortality. In such cases, tank hold times may be shortened, as appropriate and agreed upon, such
that an adequate assessment of the BWTS treatment efficacy may be made.
For in-tank treatments, test duration will include the minimum contact time the vendor prescribes
for effective treatment, but not less than a cumulative one-day holding time for each of the
required BE test cycles. As with the in-line approach, testing of the BWTS without active
ingredients may be run in parallel with the challenge test to reduce the overall duration of the
verification test. Modifications may be made according to vendor-specified requirements for
treatment, but they must be justified in the TQAP. For example, if holding water for a specified
time after the treatment's minimum contact time is required by the vendor, that time interval
would be added to each verification test incorporating challenge organisms. For combinations of
in-tank and in-line treatment, test duration will be equal to treatment time (in-line plus in-tank).
The O&M test cycles will provide data on the system's operation and support the assessment of
non-biological verification factors. In the case of in-tank treatment approaches, particularly
those using biocides or other chemical/physical means of achieving treatment, the TQAP may
elect to operate the BWTS during O&M cycle either eliminating or reducing dosage of the active
agent (i.e., to verify the electro/mechanical aspects of the BWTS). In some cases, it may be
possible to use inert substances in place of treatment chemicals to reduce the need for
conditioning prior to discharge back to the environment. Any such substitution must mimic the
operation of the BWTS when using treatment chemicals and must be agreed to by the VO.
34
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
5.4.6 Biological Efficacy (BE) Verification Testing
As discussed above, a minimum of three BE test cycles per salinity regime will be conducted;
each having a minimum tank holding time of one-day and having input challenge conditions as
described in Section 5.2. The BE verification test cycles are intended to measure the efficacy
with which the treatment system removes or kills organisms under challenging conditions. The
remainder of this section provides the detailed description of test parameters, measurements, and
analyses related to assessing biological efficacy and monitoring challenge water conditions. Due
to the nature of the verification tests, a set of core and auxiliary measurement parameters will
apply to each BE verification test. Core and auxiliary parameters, sampling location, and
sample/measurement approach are shown in Table 8. Core parameters are those that are required
during each BE test cycle and are the minimum measurements required to verify treatment
efficacy and the validity of the BE test cycle. Auxiliary parameters are: (1) useful indicators of
core parameters, (2) required by the vendor or VO, or (3) otherwise advisable to assess test
validity or treatment efficacy. Guidance on sampling methods, sample volumes, sample
container type, preservation method, and maximum holding time for each parameter is shown in
Table 9. Although the maximum holding times are listed, all analyses should be conducted as
soon as possible,
The TO, in conjunction with the TF and the vendor, will assess the use of continuous, in situ
(inline) biological or other process measurements during verification testing. Any selected
methods must be described and justified in the TQAP and approved for use by the VO.
The TO shall present a detailed schedule for verification test sample collection and analytical
methods in the TQAP. At a minimum, the TQAP shall contain the scheduled sample collection
times (expressed as time from start of test), parameters for testing, number of replicates, and
number of control samples.
5.4.6.1 Water Quality Parameters & Analysis
Water quality samples shall be collected as described in Section 5.3.2.1 and defined in the
TQAP, with the volumes described in Table 9. Note that some analyses, following methods
described in Table 10, must be performed within 6 hours of the sample collection. In cases where
water quality samples can be stored for appropriate time periods, TO logistics may warrant
outsourcing of water quality analyses to an independent, qualified laboratory, with agreement by
the VO. Reliable, continuously recording in situ sensors are available for temperature, salinity,
dissolved oxygen, chlorophyll a, and turbidity. Such sensors may, with VO approval, be used to
measure water quality parameters during verification testing. Discrete analytical samples shall
be collected to provide test-specific verification or calibration of the sensor data and to allow
comparison of sensor data to vendor-supplied information as appropriate. Sensor maintenance
and calibration shall be described in the test site operating procedures and the TQAP. Data
quality objectives for quality control and quality assurance purposes are provided in Table 11.
These data quality objectives and the related QA/QC measures should be discussed in the QAPP
of the TQAP as described in Appendix A.
35
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Table 8. Core and Potential Auxiliary Parameter and Measurement Techniques
Measurement Sample Location and Approach 1>2
Parameter
Class
Challenge Water
Post Treatment
Temperature
Salinity
Total suspended solids
Particulate organic
matter
Dissolved organic matter
Dissolved oxygen
pH
Ambient Organism
Concentration
Core
Core
Core
Core
Core
Core
Core
Core
In situ, continuous
In situ or Discrete grab
Discrete grab
Discrete grab
In situ, continuous,
discrete
In situ, discrete
In situ, continuous
Discrete
In situ, continuous
In situ or Discrete grab
Discrete grab
Discrete grab
In situ, continuous,
discrete
In situ, discrete
In situ, continuous
Discrete
Ballast System Flow Core
Rate
Ballast System Pressure Core
Sampling Flow Rates Core
Chlorophyll a (biomass) Core
Dissolved nutrients Auxiliary
(N, P, Si)
Turbidity Auxiliary
ATP (living material) Auxiliary
In situ, continuous
In situ, continuous
In situ, continuous
In situ, continuous
NA
In situ, continuous
In situ, continuous
In situ, continuous
In situ, continuous
Discrete
In situ, continuous In situ, continuous
Continuous as available Continuous as available
1 In Situ = in-line or in-tank measurements; Discrete Grab = an acquired sample for analysis at a specific place and
time; Continuous = measurement is continuous throughout the period of operation at some defined rate.
2 The frequency and means for calibrating and validating performance of in situ monitoring devices must be
addressed in the TQAP.
Discrete samples for determination of total suspended solids, particulate organic matter (as
carbon, POM), total dissolved organic matter (as carbon, DOC), and nutrient concentrations shall
36
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
be collected appropriate to the tests being conducted. The concentration of mineral matter may
be determined as the difference between the total suspended solids and the paniculate organic
matter concentration (mass per liter basis). In addition, when appropriate, samples should be
acquired or measurements made in situ to measure residual toxicity or the concentration of
chemical residuals or disinfection by-products. Guidance for such measurements and sample
collection are highly dependent on the chemicals of interest or in use; a qualified laboratory
should be consulted for the appropriate handling and measurement methods.
The analytical methods must be applied within defined holding times (Table ) after appropriate
preservation, per industry standard procedures. Where available, US EPA, Standard Methods or
other methods (i.e., ASTM) approved by the VO will be used to quantify each parameter. If
standardized methods are not available, the sampling and analytical methods to be used shall be
documented in the TQAP. These methods will follow accepted scientific practices and be
accepted by the VO.
5.4.6.2 Biological Parameters
Biological samples will be collected using methods and techniques appropriate to the size class
and anticipated concentration being measured. The samples for biological analyses will be
acquired from each of the time integrated sample volumes acquired during the test cycle. The TO
will ensure that the contents of the integrated sample collection tanks have been thoroughly
mixed to ensure homogeneity prior to sub-sampling.
The abundance of living ambient and test organisms (if used) will be quantified in (1) the uptake
challenge water just prior to treatment and entry into control tank, (2) the discharge of the control
tank after the appropriate hold time, (3) the discharge following an in-line BWTS and (4) the
discharge from the holding tank (for both types of treatment) of treated water after the
appropriate holding time. In the case of the control and treated discharge, biological samples
will be retrieved from a point upstream of any pumps or significant components which could be
expected to affect organism mortality or sample representativeness.
37
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Table 9. Sample Volumes, Containers and Processing
Parameter
Electronic in situ data
(Temperature, pH, Salinity,
etc.)2
Total suspended solids
Dissolved organic carbon
2 Particulate organic carbon
"5
1
£• Dissolved oxygen
a.
a
(j Phytoplankton Enumeration
(Live/Dead Analysis) 4
Zooplankton Enumeration
(Live/Dead Analysis) Low
Concentration/Discharge
Bacteria
„ Dissolved inorganic nutrients
%• £
| |
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Parameter
Table 10. Recommendation for Water Quality Sample Analysis Methods
Units Instrument Method/Reference
Dissolved ammonium
Dissolved inorganic nitrate and
inorganic nitrite
Dissolved inorganic phosphate
Dissolved inorganic silicate
Dissolved organic carbon
M Autoanalyzer
M Autoanalyzer
M Autoanalyzer
M Autoanalyzer
M Carbon Analyzer
Particulate organic matter
Chlorophyll a/phaeopigments
Total suspended solids
Dissolved oxygen
Carbon analyzer or
CHN Analyzer
g/L Fluorometer
Mg/L 5-place balance
Mg/L Radiometer TitraLab
APHA Standard Method No. 4500-NH3 / 20th edition
EPA Method No. 349.0 http://www.epa.gov/nerlcwww/m349_0.pdf
ESS Method No. 220.3 http://www.epa.gov/glnpo/lmmb/methods/methd220.pdf
APHA Standard Method Nos. 4500-NO2-B and 4500-NO3-F, 19th edition
EPA Method No. 353.4 http://www.epa.gov/nerlcwww/m353_4.pdf
ESS Method No. 310.1 http://www.epa.gov/glnpo/lmmb/methods/methd310.pdf
EPA Method No. 365.5 http://www.epa.gov/nerlcwww/m365_5.pdf
EPA Method 366.0 http://www.epa.gov/nerlcwww/m366_0.pdf
APHA Standard Method No. 5310-C, 20th edition
ASTM Method Nos. D6317, D2579, D4129, D4839, D513-02 and D5790
LMMB Method No. 096 http://www.epa.gov/glnpo/lmmb/methods/docanal2.pdf
LMMB Method No. 014 http://www.epa.gov/glnpo/lmmb/methods/pocdoc2.pdf
EPA Method No. 440.0 http://www.epa.gov/nerlcwww/m440_0.pdf
LMMB Method No. 097 http://www.epa.gov/glnpo/lmmb/methods/pocanal2.pdf
APHA Standard Method No. 5310-C, 20th edition
LMMB Method No. 014 http://www.epa.gov/glnpo/lmmb/methods/pocdoc2.pdf
EPA Method No. 440.0 http://www.epa.gov/nerlcwww/m440_0.pdf
EPA Method 445.0 http://www.epa.gov/nerlcwww/m445_0.pdf
EPA Method No. 446.0 http://www.epa.gov/nerlcwww/m446_0.pdf
EPA Method 447.0 http://www.epa.gov/nerlcwww/m447_0.pdf
ASTM Method No. 3731-87 (1998)
ESS Method No. 340.2 (LMMB Method No. 065)
http://www.epa.gov/glnpo/lmmb/methods/methd340.pdf
APHA Standard Method No. 2540D (1998)
EPA Method 160.2 http://www.epa.gov/region09/qa/pdfs/160_2.pdf
EPA Method No. 360.1 (Probe Method)
APHA Standard Method No. 4500-OG (Probe Method)
39
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Table 11. Data Quality Objectives for Water Quality Samples
Data Quality Indicator
Type/Acceptance
Criteria
Core Frequency of QC Sample
Parameter Collection
Method Detection
Limit
Dissolved Procedural blank
nutrients Two (2) per treatment cycle
Sample replicates
Three (3) sample replicates
per treatment cycle
Total Procedural blank
suspended Two (2) per treatment cycle
solids (DI Sample replicates
water and Three (3) sample replicates
seawater) per treatment cycle
DOC Procedural blank
Two (2) per treatment cycle
Sample replicates
Three (3) sample replicates
per treatment cycle
POM
Chlorophyll
a/phaeophytin
Procedural blank
Two (2) per treatment cycle
Sample replicates
Three (3) sample replicates
per treatment cycle
Procedural blank
Two (2) per treatment cycle
Sample replicates
Three (3) sample replicates
per treatment cycle
Dissolved Procedural blank
oxygen NA
Sample replicates
Three (3) sample replicates
per treatment cycle
Ammonia and silica
0.02 M
Nitrate, nitrite,
phosphate
0.01 M
0.1 mg/L
20 M
5.5 M
0.02 g/L
Procedural blank
<5 times MDL *
Sample replicates
2% PD 2
Procedural blank
<5 times MDL
Sample replicates <10%
RPD3
Procedural blank
15%PD
Sample replicates
10% RPD
Procedural blank
15%PD
Sample replicates
10% RPD
Procedural blank
<5% PD
Sample replicates
< 15% RPD
Procedural blank
NA
Sample replicates
<5% CV 4
1 MDL = method detection limit.
2 Percent Difference (PD) = [(true concentration - measured concentration)/true concentration] 100%.
3 Relative Percent Difference (RPD) = {[absolute value (replicate 1 - replicate 2)]/[(replicate 1 + replicate 2)12}}
100.
4 Filter blanks used for QC purposes only.
40
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
5.4.6.3 Sample Volumes & Data Quality
Samples from the discharges of successful treatments will likely have low concentrations of
organisms. Enumeration of the organisms from these samples (determined from 20 one-mL
subsamples from a concentrated whole water sample, as described below) is represented by the
Poisson distribution, and therefore the cumulative or total count is the key test statistic (Lemieux
et al., 2008b). Further, a chi-square transformation can be utilized to approximate the confidence
intervals.
Assuming, for organisms > 50 jim, that the desired minimum precision is that the upper bound of
the chi-square statistic should not exceed twice the observed mean (this corresponds to a
coefficient of variation of 40%), a count of 6 organisms is required.3
The volume required to successfully count 6 organisms is dependent on the whole water sample
volume, concentration factor, number of sub-samples counted, and the target concentration.
Table 12 provides the resultant upper bounds, based on the Poisson distribution for a 95%
confidence interval from the chi-square transformation for a variety of sample volumes at a
concentration factor of 3000 (3 m3 concentrated to 1 L) assuming 20 subsamples of 1 mL. Given
these assumptions, 30 m3 must be sampled to enumerate 10 organisms/m3, with the desired level
of precision given above. The total sample volume may be reduced accordingly if: the
concentration factor is increased, the confidence limit is lowered (e.g., from 95% to 90%),
or the volume of subsample analyzed is increased (e.g., from 20 mL to 40 mL). If the latter
is done, the TF should conduct validation experiments to ensure counting accuracy is high (e.g.,
using microbeads as described in Lemieux et al. (2010), which is found in Appendix C). It is the
responsibility of the Testing Organization to justify any changes to the volumes suggested in
Table 12. As discussed previously, sample replication is unnecessary as the Poisson distribution
pools the data to improve the measurement precision and assumes the organisms to be randomly
distributed. Note that this approach would not be appropriate if samples are not
continuously acquired on a time-averaged basis.
In any case, sample size should be selected relative to the targeted concentration and to provide
the level of precision required to supply a 95% upper confidence limit which is (1) no more than
twice the observed mean and (2) does not exceed the targeted concentration or as otherwise
defined by the TO. Examples are provided in Table 13 for standards for organisms larger than
50 |im compared to standards that are currently proposed or considered domestically and
internationally. The chart provides the volume of sample required assuming that the entire
sample is concentrated to 1 L and 6 organisms are counted. N is the number of samples
analyzed, with each sample dispensed into a well plate and having 20 one-mL subsamples
observed.
A similar approach for organisms > 10 jim and < 50 jim may be applied; however, the targeted
concentrations are considerably denser, and the anticipated total counts can be expected to be
Mathematically, that relationship can be represented as follows: coefficient of variation = standard
deviation/mean. For the Poisson distribution, the variance = (standard deviation)2 = mean, thus
substituting the critical value of the mean, 6, gives a coefficient of variation = 605/6 40%.
41
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
higher. The Poisson distribution assumption still applies, and a more stringent level of precision
may be applied. Specifically, if the desired level of precision is set at a coefficient of variation of
10% or the upper confidence limit is not more than approximately 20% of the estimated density,
then the volumes given in Table 14 result. These volumes are the required whole water sample
volume to be concentrated to 1 L as a function of the number of 1 mL sub-samples (N).
5.4.6.4 Zooplankton Enumeration
Time integrated in-line sample volumes should be concentrated at the time of sampling using 35
|im mesh plankton nets (50 jim in the diagonal). The concentrated contents of the cod-end
should be rinsed into a flask. The volume capacity of the flask will be dependent on the organism
density of the sample but typically requires a range of 1 to 4 L capacity. Fresh, artificial
seawater, filtered seawater, or freshwater, as appropriate, should be added to maintain oxygen
levels for the living organisms to be counted. If the initial sample has a low concentration of
zooplankton, the sample may need to be further concentrated before analysis. In this instance,
the sample should be concentrated using 35 jam mesh.
42
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
1
2
Table 12. Density Confidence Intervals for Poisson Distributions Using the Chi-Square Statistic
Count Data
Whole Water Sample
Scaled Densities and Confidence Intervals
(Assumes whole water sample volume is concentrated to 1L,
with analysis of 20 one-mL subsamples from the concentrate)
V=lmJ
V = 10 mj
V = 30 mj
V = 60 mj
Organism
Count
0
1
2
3
4
5
6
1
8
9
10
11
12
13
14
15
16
17
18
19
20
volume ^
•~ -O
^ *> =
fc a!
^PW
3.00
4.74
6.30
7.75
9.15
10.51
11.84
13.15
14.43
15.71
16.96
18.21
19.44
20.67
21.89
23.10
24.30
25.50
26.69
27.88
29.06
vj-
Upper
Bound /
Count
4.74
3.15
2.58
2.29
2.10
1.97
1.88
1.80
1.75
1.70
1.66
1.62
1.59
1.56
1.54
1.52
1.50
1.48
1.47
1.45
Mean
Density
(m3)
0
50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
850
900
950
1000
95%
Upper
Bound
(m3)
150
237
315
388
458
526
592
657
722
785
848
910
972
1033
1094
1155
1215
1275
1335
1394
1453
Mean
Density
(m3)
0
17
33
50
67
83
100
117
133
150
167
183
200
217
233
250
267
283
300
317
333
95%
Upper
Bound
(m3)
50
79
105
129
153
175
197
219
241
262
283
303
324
344
365
385
405
425
445
465
484
Mean
Density
(m3)
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
95%
Upper
Bound
(m3)
15
24
31
39
46
53
59
66
72
79
85
91
97
103
109
115
122
127
133
139
145
Mean
Density
(m3)
0
2
3
5
7
8
10
12
13
15
17
18
20
22
23
25
27
28
30
32
33
95%
Upper
Bound
(m3)
5
8
10
13
15
18
20
22
24
26
28
30
32
34
36
38
41
42
44
46
48
Mean
Density
(m3)
0
1
2
O
3
4
5
6
7
8
8
9
10
11
12
13
13
14
15
16
17
95%
Upper
Bound
(m3)
2.5
4.0
5.2
6.5
7.6
8.8
9.9
11.0
12.0
13.1
14.1
15.2
16.2
17.2
18.2
19.2
20.3
21.2
22.2
23.2
24.2
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Table 13. Sample Volume Required Relative to Treatment Standards-Organisms > 50 jim
N= 1 3 5
Concentration (individuals/m3) Sample Volume Required (m3) l
0.01
0.1
1
10
60,000
6000
600
60
20,000
2000
200
20
12,000
1200
120
12
1 Assumes the entire volume is concentrated to 1 L, 20 1-mL subsamples are analyzed, and the desired precision is
the 95% Confidence Interval of the Poisson distribution = 2 times the observed mean and not greater than the
Standard Limit.
Table 14. Sample Volume Required Relative to Treatment Standards Organisms > 10 jim
and < 50 jim
N1= 234
r\
Concentration (individuals/mL) Sample Volume Required (L)
0.01
0.1
1
10
6000
600
60
6
4000
400
40
4
3000
300
30
3
1 The number of 1 mL sub-samples analyzed.
2 Assumes the entire volume is concentrated to 1 L and the desired precision is that CV is not greater than 10%.
Subsamples should be analyzed immediately, and as analysis proceeds, the original sample
should be held at ambient water temperature. Previous work has shown zooplankton die-off
occurs in the sample after a hold time of 6 hours. The appropriate maximum hold time should be
validated at each test facility so that the detectable zooplankton mortality over the hold time does
not exceed 5%.
Subsamples should be extracted using 5-mL serological, graduated pipettes with an Eppendorf
pipette helper (or a similarly accurate instrument that can effectively capture swimming
zooplankton). Subsamples should then be examined in multi-well plates, Bogorov chambers,
Sedgewick Rafter Counting Chambers, or counting wheels. The subsample should be dispensed
into the counting chamber while still allowing for the addition of a narcotizing agent. In
addition, the counting chamber volume should be shallow enough to allow for adequate focusing
on the organisms during analysis. All direct counts should be done using counting chambers
placed under a stereo or compound microscope at magnifications ranging from 10x to 40x.
Lugol's iodine solution should be used as a euthanizing and preservation agent. It should be
noted that this agent works particularly well on the standard test organism Anemia spp. And for
ambient organisms that have chitinous exoskeletons. It has been documented, however, that
44
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Lugol's can have distorting effects on the preservation of some marine organisms, particularly if
their bodies lack chitin or other types of hard body structure. Given the choices of preservation
or euthanizing agents available, additional validation is advised when different zooplankton are
present in samples, or when dealing with organisms found at specific TFs, to determine which
fixative(s) work best in preserving the zooplankton concentrations for total direct counts.
In samples from challenge water or a control tank, the zooplankton should first be examined to
count the number of dead organisms, defined by a lack of visible movement during an
observation time of at least ten seconds. Unmoving but intact zooplankton may be living, so
they are gently touched with the point of a fine dissecting needle or probe to elicit movement.
Given that each dead organism is monitored for at least 10 seconds for visible movement,
viability measurements could be lengthy for samples with dense concentrations of dead
organisms, thereby increasing the potential for sample bias due to sample degradation.
Once the number of dead organisms has been tallied, the organisms within the wells should be
killed and/or preserved (to eliminate motion of the live organisms) and total counts obtained.
Live counts will then be calculated from the difference using the equation: Total #-# Dead = #
Live. Because samples collected following treatment are expected to have few living organisms,
the living organisms can be enumerated directly.
Note that samples collected to verify challenge conditions are met may require taxonomic
identification of dominant organisms.
5.4.6.5 Organisms >10 /urn and < 50 /urn (nominally protists)
Laboratory concentration of this size of organisms in the whole water sample can be
accomplished by gently passing the sample through a sieve with mesh < 10 jim in the diagonal.
Care should be taken to gently sieve organisms to ensure they are not killed in the process.
Techniques and standardized methods for the enumeration and viability analyses of protists
remain an active area of investigation. This protocol recommends use a combination of two vital
stains: Fluorescein Diacetate (FDA, Molecular Probes-Invitrogen Carlsbad, CA) and 5-
chloromethylfluorescein diacetate (CMFDA, CellTracker™ Green; Molecular Probes-Invitrogen
Carlsbad, CA). When non-specific esterases in living cells cleave the stains, the resultant
molecules fluoresce green when excited with a blue light (e.g., Selvin et al., 1988;
www.invitrogen.com).
This method utilizes manual epifluorescence microscopy to evaluate samples: FDA (final
concentration 5 jiM) and CMFDA (final concentration 2.5 jiM) are added to a 1 mL sample that
is incubated in the dark for 10 minutes, the sample is loaded into a Sedgewick Rafter Counting
Chamber, and it is examined under epifluorescence using a Fluorescein Isothiocyanate (FITC)
narrow pass filter cube (e.g., excitation 465-495 nm, dichroic mirror wavelength 505 nm, barrier
filter 515-555 nm; Drake et al., 2010). Samples should be examined for a maximum of 20
minutes because the signal fades as stain leaks from the cell. If a cell is labeled by either FDA or
CMFDA (as exhibited by a characteristic fluorescent green color) or moves, or both, it is scored
as viable. A photomicrograph should be taken of any such cells under fluorescent and brightfield
(white light) illumination to create a visual record of viable cells.
45
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Research on the dual staining method at four locations (including marine and estuarine sites) in
the U.S. has shown the method to yield variable degrees of false positives (Type I error) from as
little as 3% to nearly 40% (Steinberg et al., 2010). Thus, before Tos use the dual staining
method or any other alternative method, it is necessary that it undergo on-site validation by
preparing, examining, and analyzing ambient samples that are killed (i.e., negative controls).4
From the perspective of environmental protection, this type of error is conservative, as it
overestimates the number of viable organisms. In contrast, Type II errors (false negatives)
underestimate the number of viable organisms. Encouragingly, the Type II error rate was
uniformly low across all study sites: 0% in three locations and 1% at the remaining two
locations (Steinberg et al., 2010). Nonetheless, the Type II error rate should also be determined
during initial site validation of this method or alternative method validation on a seasonal basis,
and as part of the on-going QA program.
The advantages of using manual microscopy with vital stains are: (1) the instrumentation
required is available in most research laboratories, (2) the cost of materials is low, (3) sample
incubation times are relatively short, (4) the protocol is straightforward, and (5) results can be
generated fairly rapidly. The disadvantages of this method are that it takes several hours (4-5
hours) to completely characterize the subsamples within a sample, and unless the microscope is
equipped with a camera (which is recommended), there is no archive of the data collected.
Additionally, manual counts are subject to errors from operator-specific biases as well as from
fatigue effects during extended observation periods.
Within this size class fall dormant cells or resting stages exhibited across a broad phylogenetic
range of microalgae, heterotrophic protists, and metazoans (e.g., Marret and Zonneveld, 2003;
Matsuoka and Fukuyo, 2000). To encompass this group, the term 'cysts' is used, which includes
but is not limited to cysts of dinoflagellates, spores of diatoms, cysts of heterotrophic protists,
and cysts of rotifers. Notably, spores of bacteria and fungi are not included; they are smaller in
minimum dimension than the lower limit of the size class considered here (10 jim).
With focus on dinoflagellates alone, many authors (e.g., Dobbs and Rogerson, 2005; Doblin and
Dobbs, 2006; and references in both) have made the point that cysts in ships' ballast water
represent robust ecological hazards. Given their resistance to physiological stress, killing cysts
may be the best, i.e., most stringent, test of ballast-water technologies. If cysts can be killed,
then there is excellent reason to assume vegetative cells or non-resting stages will also be killed.
But because of their low metabolic state and relative impermeability to stains, it may be difficult
to assess the viability of cysts on an individual basis without painstaking, cultural analyses,
which, if possible at all, may require weeks or months to complete. At present, no rapid, reliable
4 For example, heat-killed, negative control samples are prepared by placing ambient water samples in a 50 °C water
bath. Once the sample temperature reached 50 °C, it is held in the bath for an additional 10 minutes (Drake et al.,
2010; Steinberg et al., 2010). The sample is cooled to room temperature before being stained. Organisms should
not show a green, fluorescent signal after heat killing; those that do fluoresce represent false positives and indicate
the Type I error associated with the dual-stain method.
5 One approach is to collect ambient protists and place them in one of four categories based on an organism's
fluorescence signal and movement: (1) fluorescent and moving, (2) fluorescent and non-moving, (3) non-fluorescent
and moving, and (4) non-fluorescent and non-moving (Drake et al., 2010; Steinberg et al., 2010). Organisms binned
as non-fluorescent and moving are obviously viable, but the combination of stains fails to indicate viability,
representing the Type II (false negative) error.
46
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
method to determine cysts' viability is in widespread use, and the FDA-CMFDA method has
yielded variable results with dinoflagellates and cyst-like objects. This protocol allows for
additional techniques for plankton assemblages to be developed (Section 5.4.8); should a method
reliably indicating cyst viability become available, it is assumed that it would allow all viable
organisms within the Protest size class to be enumerated.
5.4.6.7 Organisms <10 /urn
Bacteria samples should not need to be concentrated from the whole water sample prior to
analysis. Sample analysis will be conducted according to standard microbial techniques.
Multiple bacterial growth media will be used to assess the effectiveness of a treatment for
bacteria6. Use of multiple types of media enables measurement of the response of different
portions of the ambient bacterial community7. The minimum number of media used will include
two general-purpose (1 marine, 1 nutrient agar) media for culturable aerobic heterotrophic
bacteria. Other media may be added during the development of the TQAP. The rationale and
methods will be described in the TQAP.
For culturable, aerobic, heterotrophic bacteria, 1 mL samples should be diluted in a 10-fold
dilution series in sterile Phosphate Buffered Saline (PBS) or sterile ambient water. Next, 100 jil
of each appropriate dilution should be spread onto the media recommended in the protocol, with
triplicate plates for each dilution. Plates should be incubated at 25 °C and monitored during the
incubation time to ensure overgrowth does not occur. Colonies should be monitored and counted
after 5 days (or after 3-5 days, if colony overgrowth appears imminent on all plates) and
recorded as colony forming units (CFUs) per 100 mL of sample water.
For E. coli samples, USEPA Method 1603 should be used: ImL, 10 mL and 100 mL water
samples should be passed through 0.45 jim membranes, which should be placed on modified
thermo-tolerant E. coli agar (mTEC) plates (Becton Dickson, Sparks, MD). Plates should be
incubated at 35 ± 0.5°C for 2 hours to allow for cell wall repair. Next, plates should be incubated
at 44.5°C in a waterbath for 22-24 hours. Total red and magenta colonies should be scored and
data reported as E. coli colonies per 100 mL of sample water. Alternatively, an IDEXX Colilert
kit (Westbrook, ME) can be used according to the manufacturer's protocol.
For Enterococci samples, a modified version of USEPA Method 1106.1 should be used: 10 mL
and 100 mL water samples should be passed through 0.45 jim membranes, the membranes
transferred onto mEnterococcus agar (mEA) plates, and the plates incubated at 35 ± 2°C for 24
hours. Membranes with light and dark red colonies should be transferred to bile esculin agar
(BEA) plates, which should be incubated for 4 hours at 35 ± 2°C. After incubation, colonies
with black halos should be scored and data reported as Enterococci per 100 mL. Alternatively,
an IDEXX Enterolert kit (Westbrook, ME) can be used according to the manufacturer's protocol.
Toxigenic Vibrio choleras densities should determined by a DNA colony blot hybridization
method that detects ctxA gene (Huq et al., 2006). Briefly, colonies are grown on TCBS agar,
6 The suggested media for marine water include 2216 Marine Agar and salt-modified R2A agar; media for fresh
water species may include Plate Count Agar and Nutrient broth (plus agar (15 g/L)').
.ed, a
47
7 Note it is assumed that if all culturable bacteria are killed, all non-culturable bacteria are also killed.
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
purified, and inoculated with 2.5% yeast extract and nalidixic acid and fixed after incubation
overnight. Viable V. choleras Ol and O139 cells are enumerated using a direct-fluorescent
antibody kit (New Horizons Diagnostics; Columbia, MD) for serogroups Ol and O139 using
monoclonal antibodies tagged with fluorescein isothiocyanate (FITC) under an epifluorescence
microscope.
Appropriate controls (e.g., heat to remove vegetative cells for tests using resting stages or spores)
for microbial plates will be used throughout the verification testing. Steps will also be taken to
ensure the action of any treatment (e.g., a biocide) is stopped at the time of sample collection
(i.e., treatment does not continue after sample collection). Any steps and controls used to verify
the effectiveness of a neutralizer will be described and justified in the TQAP.
5.4.6.8 Auxiliary Parameters
Sampling and analysis of supplemental parameters may be required depending on vendor-
specified information. For example, a vendor may define an additional treatment effectiveness
based on removal of fecal coliform bacteria or other microorganisms of public health concern.
In such cases, the TO, with VO acceptance, will determine the appropriate supplemental
parameters, based on vendor-specific information, and shall determine sampling and analysis
requirements for inclusion in the sample collection schedule in the TQAP.
5.4.7 BE Validity Criteria
At the conclusion of each BE verification test cycle, the TO should verify that all criteria
necessary for a valid BE test were established and maintained, as appropriate. As a minimum,
the test validity criteria should consist of: (1) operational parameters that demonstrate the
requisite volumes were transferred and sampled and Vendor-specified flows, pressures, or other
validation criteria were maintained, (2) water quality challenge conditions for uptake and
discharge waters, including any toxicity sampling as required by the TQAP, were met, and (3)
biological challenge conditions for ambient organism concentration and diversity in treatment
and control samples were met. Note that the vendor-specified validation criteria should be
limited to operational parameters; that is, the criteria should be employed to ensure the system
was operated correctly and in accordance with the provided training and O&M manual.
Parameters that document a system failure under proper usage do not invalidate a test. The types
and locations for measurements in each category are summarized in Table 15, and requisite
criteria are discussed by category.
Each of the measurement criteria and its valid ranges are to be documented by the TO in the
TQAP. The declared ranges should accommodate variability of ambient source water conditions
as well as possible ranges for ambient organisms in challenge water. The declared range of valid
conditions also indicates the degree to which the TO can control the test parameters. Following
each individual test, the TO will produce a test validation matrix that summarizes valid ranges
(from the TQAP) and corresponding mean values obtained during the test. Any significant
deviations from the mean noted during the testing shall be discussed in the verification report.
48
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
5.4.7.1 Uptake Operations
The ETV protocol provides minimum requirements for volume and flow in Section 5.2.3, with
final ranges for volume, pressure and flow to be identified in the TQAP by the TO and the
vendor. For all water transfers, the minimum volume is 200 m3, and the minimum sample
collection volume associated with each transfer is 3 m3. Acceptable ranges for sample collection
volumes, pressures, and flows are also to be identified in the TQAP. The test validation matrix
should provide the valid ranges and the resulting mean values to verify if valid, in-range
conditions were met over the duration of the test at the locations specified in Table 16.
5.4.7.2 Water Quality Conditions
Minimum water quality conditions for BE tests are provided in Table 3 for dissolved and
particulate organics, mineral matter, total suspended solids, and temperature for the two salinity
ranges. The TO, in conjunction with the vendor, will declare valid test ranges for these and other
relevant water quality parameters of interest (e.g., pH, DO, etc.) in the TQAP and provide a list
of the valid ranges and mean results for each test in the test validation matrix. If there is reason to
measure these parameters in the discharge waters as well, these measurements should also be
presented.
5.4.7.3 Biological Diversity and Concentrations
Table 3 presented requirements for biological challenge conditions to include a minimum of 5
species from 3 separate phyla across the three requisite size classes. This specification refers to
populations at the point of treatment and entry to the control tank. Valid population densities and
diversity ranges for ambient organisms should be defined by the TO in the TQAP. These ranges
are envisioned to be fairly broad to accommodate variations in ambient populations and
dominant species over the duration of the ETV testing. The anticipated ranges and the measured
mean values for the sampled populations are to be presented in the test validation matrix.
Organism population densities and diversity are also to be measured in the discharge samples for
both treated and control waters, where minimum living concentrations are required in the control
discharge as noted in Table 7. These criteria are given by totals for each size class and should be
shown in the test validation matrix with results broken out for the dominant 5 species present in
each size class.
5.4.7.4 Biological Treatment Efficacy Determination
Treatment efficacy will be determined by the measurement of living ambient organism
concentrations in the treatment discharge for the three size classes identified in Table 3.
5.4.7.5 Toxicity Test for Biocide Treatments
Toxicity tests conducted during the start-up for treatments involving biocides in marine and
brackish waters will be selected from the following:
Inland Silverside, Menidia beryllina, Larval Survival and Growth (EPA Method 1006.0):
http://www.epa.gov/OSTAVET/diskl/ctml3.pdf)
49
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Table 15. Challenge Test Validation Criteria by Location
Parameter
Volume
Pressure
Flow
Vendor-specified parameters
Temperature, Salinity, pH, DO
(can monitor source waters)
Total Suspended Solids (TSS)
Dissolved Organic Carbon
(DOC)
Paniculate Organic Material
(POM)
Mineral Matter (MM)
Environmental Contaminants
Ambient organisms/m3;> 50(im
(live/dead)
Ambient organisms/mL;
> 10 and < SO^m (live/dead)
Ambient organisms/mL;
< 10 (im (live/dead)
Control
Sample
Tank
Treatment
Sample
Tank
Control
Tank
Ballasting Operations
XXX
XXX
XXX
Water Quality Conditions
X X
X X
X X
X X
X X
Biological Diversity and Concentrations
Treatment
Tank
x
x
X
X
X
X
X
Control Treatment
Discharge Discharge
Sample Sample
Tank
x
x
X
X
X
X
X
X
X
X
Tank
x
X
X
X
X
X
X
X
X
X
X
Sea Urchin, Arbaciapunctulata, Fertilization Test (EPA METHOD 1008.0:
http://www.epa.gov/OSTAVET/diskl/ctml5.pdf)
Mysid Acute Toxicity Test (EPA OPPTS Method 850.1035:
http://www.epa.gov/opptsfrs/OPPTS_Harmonized/850_Ecological_Effects_Test_Guidelines/D
rafts/850-103 5.pdf)
Additional guidance can be found in ASTM (1996a, 1996b) and Klemm, et al. (1994). Tests and
species selected for toxicity testing during commissioning will be specified in the TQAP in
accordance with the salinity ranges identified for testing.
5.4.8 Alternative and Emerging Methods
New methods for analysis and enumeration of living plankton communities are being developed
to meet the relatively complex and demanding needs of ballast water treatment system testing.
These methods include, but are not limited to, rapid analytical measurements, vital stains and
dyes, and molecular probes. The inclusion or substitution of these techniques to those described
above is acceptable. However, at a minimum the method(s) selected for any given size class
50
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
should provide a quantitative measurement of the concentration of living organisms. If non-
standard methods are selected, they should be validated by the TO to the satisfaction of the VO.
5.4.9 Operation and Maintenance Verification Factor
The operation and maintenance (O&M) of the ballast water treatment system will be verified.
The verification has been designed as a minimum volume requirement, allowing sufficient time
to verify operation and maintenance of the ballast water treatment system. It is anticipated that
O&M testing commences in between BE test cycles to ensure some equipment run-time occurs
prior to each BE verification test. In this manner, any changes in treatment efficacy due to
equipment operation over time may be observed.
The TO is responsible for monitoring and maintaining the system, in accordance with the
Vendor's O&M manual, throughout the testing to ensure stable operating conditions (as mutually
agreed to by the vendor) and proper operating effectiveness. All system components will be
monitored for proper operation throughout the test period. All maintenance activity completed
during the verification testing shall be documented for inclusion in the verification report.
All required monitoring and maintenance activities should be coordinated with the TO in
advance of verification testing, and detailed in a monitoring and maintenance plan included in
the TQAP. The monitoring and maintenance plan shall address the following requirements, as
applicable:
• A monitoring and maintenance schedule for the testing period (as shipboard systems are
generally designed to require minimal regular maintenance, visual inspections by the
operator may be all that is required);
• Equipment and component calibration methods and frequencies;
• Monitoring and maintenance activities and procedures shall be described and
documentation forms provided - maintenance documentation forms must identify the TF,
date and time, describe the work performed, observations of the treatment system, and
results of the work; and
• Operating characteristics and vendor-specified ranges required for proper operating
conditions shall be described (e.g., system temperature, flows entering and exiting the
system, power levels).
Other information that must be addressed in the TQAP includes:
• Monitoring requirements to ensure a proper operating environment;
• Continuous on-line O&M monitoring requirements, as specified by the vendor; and
• Credentials of all personnel involved in operating, monitoring and maintaining the
treatment system.
All monitoring and maintenance documentation must be maintained in a written record at the TF
and will be included in the verification report.
To help address predictability and reliability verification factors, qualitative and quantitative
O&M performance indictors will be evaluated. The means and methods to evaluate or quantify
51
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
O&M performance indicators shall be included in the TQAP and described in a schedule for
collecting this information.
5.4.9.1 Qualitative O&M Performance Indicators
Qualitative O&M performance indicators will include, but are not limited to:
5.4.9.2 Visual Observations
Visual inspections of the treated ballast water quality (e.g., turbidity, color) and treatment system
conditions (e.g., foaming, floating material, settled solids) will be performed at each maintenance
or monitoring event. Visual observations will also include the inspection of the treatment system
prior to, during and following each test cycle for equipment and process failures, corrosion,
leaks, impediments of flow (entering or exiting the system) and any other system issues that
could impact performance. Specific visual indicators shall be defined in the TQAP.
5.4.9.3 Operability
Observations regarding the ease of start-up and operation during testing and the ease of
monitoring system performance shall be noted and recorded.
5.4.9.4 O&M Manual
The TO shall evaluate the usefulness and quality of the O&M manual, and a written report on the
evaluation shall be prepared.
5.4.9.5 Operator Skills
The level of operator expertise required to operate and maintain the treatment system shall be
noted and compared with that indicated by the vendor.
5.4.9.6 System Accessibility
The ease of access and required clearances for system operation and required maintenance shall
be noted.
5.4.9.7 Quantitative O&M Performance Indicators
Quantitative O&M performance indicators shall include, but are not limited to:
5.4.9.8 Time demand
Personnel time required to start-up, shutdown, operate, and maintain the treatment system shall
be recorded in the monitoring and maintenance log.
5.4.9.9 Residual
Volumes of residual materials, (e.g., solids removed via filtration systems, etc.), mass generation
rates, and concentrations shall be determined during verification testing. Results will be
recorded in m3, gallons or pounds per m3, or gallons of water treated, as appropriate. Factors
related to the disposal of residuals (such as storage requirements and handling hazards) shall also
be addressed.
52
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
5.4.9.10 Chemical Use
Usage rates and concentrations of any chemicals (e.g., biocides) used as part of the treatment
system and its operation during verification testing (per test cycle) will be measured and
recorded. Results shall be reported for residuals and possible by-products.
5.4.9.11 Power consumption
The power consumed per test cycle by the treatment system will be monitored and recorded (e.g.,
kWh per m3 of water treated shall be calculated for use in cost factors below). The peak
electrical load at system start-up will also be monitored and recorded as will fluctuations in
consumption during test cycles.
5.4.9.12 Other Consumables
The use of any other consumables, such as filter cartridges, shall be monitored, documented, and
reported.
5.4.9.13 Supplemental Parameters
Depending on vendor claims, supplemental monitoring, maintenance, and O&M performance
indictors may be required. These will be described, along with requirements for performance
monitoring, in the TQAP.
5.4.9.14 Upset Conditions
Upset conditions are those events or occurrences outside the operating parameters defined in the
TQAP that result in either malfunctioning of the equipment, exception from normal operating
conditions, or conditions causing alarms that indicate the system is producing or discharging
treated water that exceeds the stated set points or limits for effective treatment. The cause of
upset conditions may be due to conditions at the TF or the technology. These events may
include both events in which the system is operating within the manufacturer's specifications and
those that are within specification but do not result in adequate treatment. The TO shall notify
the vendor and the VO immediately when an upset condition is identified. The TO shall correct
the upset condition as soon as possible to bring the treatment system back on line. For unusual
upset conditions, the TO will work with the vendor to identify and correct the problem. The
occurrence of all upset conditions, the causes, the results, and the means to correct the upset shall
be documented at the time of the occurrence and shall be described in the verification report.
As sampling is continuous over the course of the test, any upset conditions during the test need to
be noted and a post-test review conducted to determine their cause and assess the impact on test
results. (This task will be done by the TO and approved by the VO.) This review will determine
where inclusion of these data is appropriate for performance assessment and the statistical
analysis presented in the verification report. If the cause of an upset condition cannot be
determined or the condition cannot be qualified as a true upset, then the sampling results shall be
used in the statistical analysis for the verification report.
5.4.9.15 Reliability
53
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
The mechanical reliability of the technology will be determined by comparing the Vendor
projected mean-time between failure (MTBF) with the maintenance events observed during
testing. The comparison will be reported in the verification report.
The reliability of the treatment system to achieve treatment will be determined by (1) the number
of instances where the treatment system or technology does not achieve the stated performance
goal per the total number of test cycles, and (2) the standard deviation of the mean for biological
performance data (e.g., percent removal).
Reliability performance measures will take into consideration any vendor provided information
that assists in the projection of the performance such as CT (concentration-time) disinfection
information or power/energy curves. Any adjustments made to the system, outside of the
vendor-specified operation and maintenance claims, to achieve the performance goals will be
noted in the maintenance log and specified in the verification report.
Specific performance reliability indicators along with the planned methods for evaluating and
reporting them will be identified in the TQAP.
5.4.9.16 Cost Factors
Verified cost factors will include the following as applicable:
5.4.9.17 Power consumption
Power consumption will be reported as total kWh necessary to operate all equipment to achieve
desired biological treatment performance.
5.4.9.18 Consumable or expendable materials
Amounts of all consumables or expendables, including chemicals or other items required for
treatment, shall be itemized and reported.
5.4.9.19 Replacement parts used during normal maintenance
The number of replacement parts will be itemized and reported. Any unanticipated replacement
of parts will be specified separately.
5.4.9.20 Labor time to start-up, operate, and maintain the treatment system
The total number of hours for each activity will be recorded and reported.
5.4.9.21 By-product or waste materials produced
By-products that require treatment or disposal will be reported as an expression of total volume
treated or disposed.
5.4.9.22 Environmental Acceptability
Two performance indicators will determine the environmental acceptability of a treatment
system: water quality and treatment residuals.
54
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
The data used to evaluate the environmental acceptability of a system will be taken from the
water quality data collected at the point of discharge as detailed in Section 5.3.2. These data will
include but may not be limited to the following parameters:
• Temperature
• pH
• Salinity
• Total suspended solids
• Particulate organic matter
• Dissolved organic matter
• Dissolved oxygen
• Dissolved nutrients
• Biochemical oxygen demand
• Biological efficacy
The results of these tests at the point of discharge will be compared to the range of expected
natural conditions and reported in the verification report.
Additional analytical parameters will be included as necessary for reporting on any residual
material that may result from treatment; for example residual biocides and disinfection
byproducts. The additional parameters, the potential impact to the environment, and the
analytical methods will be detailed in the TQAP
It will be the responsibility of the TF to obtain NPDES discharge permits and to ensure that
discharge is within permitted limits. Additionally, toxicity testing of any biocide treatment will
be conducted, as discussed under Section 5.4.4. Verification testing will not begin unless the
results of the toxicity tests are acceptable.
5.4.9.23 Safety
Safety is of concern during the operation of any equipment or machinery and during the use of
potential hazardous materials, but it is of particular concern while on board ship, where staff is
limited and access to land based emergency infrastructure is unavailable. Therefore, the safety
of the treatment system will be evaluated during verification testing.
The performance indicators for this verification factor will be technology specific, but, to the
extent possible, required indicators shall include:
• Listing of all dangerous or hazardous materials, including submittal of Material Safety
Data Sheets (MSDS);
• Potential to compromise the normal ship ballasting or deballasting cycle (i.e., impediment
of flow);
• Visual indicators of potential threats to shipboard operations, such as exposed or
improper housing of power cables, structural stability of the system, external
temperatures of the treatment system, and any other treatment-specific factors that may
pose a threat to the operator or compromise the safety of ship operations; and
55
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
• Review of the vendor provided O&M manual for adequacy of cautions and guidance on
ways to minimize the potential for, and directives to mitigate, a hazardous situation.
The method for evaluating these and other items identified by the TO in reviewing the
technology documentation shall be described in the TQAP.
56
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 6
Reporting Verification Testing Results
Deviations from this protocol or any TQAP prepared for BWTS testing shall be described in the
verification report, which shall include supporting documentation that provided the basis for
acceptance of the deviations. All testing results will be presented in the report, including all data
regarding challenge conditions, results of verification testing for all verification factors, and any
vendor supplied data or information. A summary verification statement will also be prepared.
The outline for the report shall include:
• Verification Statement
• Executive Summary
• Introduction and Background
• Description of the Treatment System or System
• Experimental Design (including a description of all deviations from the protocol and the
basis for accepting the deviations)
• Description of Challenge Conditions
• Methods and Procedures
• Results and Discussion
• Verification Testing Operation and Monitoring QA/QC
Appendices:
• TQAP
• BE Test Validation Matrix
• Vendor-supplied Operation and Maintenance Manual
• Data Generated During Testing
• QA/QC Records
• Maintenance Logs
• Any other records maintained during testing, such as chain of custody forms
• Any other information provided by the Vendor, which may be of use to the stakeholder
community
Upon completion of the draft report the VO, the vendor, and the TF QA manager will review the
document and supporting data, and provide comment. The comments will be addressed or
stricken with approval of all parties and the final report will be submitted to NSF International
(ETV Water Quality Protection Center partner) and EPA for QA and technical review. The final
verification report and statement will be processed for clearance and posted on the ETV
Program's web site.
57
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 7
Quality Assurance/Quality Control (QA/QC)
To ensure the quality and integrity of data gathered during testing activities, a Quality Assurance
Project Plan (QAPP) will be prepared by the TO and included as part of the TQAP. The QAPP
will describe the project scope, management, procedures for measurements and data acquisition,
project assessment and oversight, and data validation and usability assessments necessary to
meet the project goals. The written document will communicate all decisions related to project
design and completion to the project team so work is performed according to written
specifications. The generic format for a QAPP is included in Appendix A. EPA also provides
guidance in preparing quality management plans, QAPPs and other quality management
documents on their web site: http://www.epa.gov/quality.
7.1 Project Management
The QAPP will list all project participants and clearly define their roles and responsibilities. In
addition, this section will describe project scheduling, data quality objectives, training and
certification requirements (as applicable), and required documentation. The information
included in this section will ensure that all participants understand the scope of the study and
their explicit roles. Due to the complexity of testing in accordance with these protocols, it is
advisable that each test cycle be preceded by a briefing or meeting in which the TF personnel
critically review the plan of action, test operation, and conduct so they are familiar with the
TQAP and their responsibilities. It is further recommended that this briefing or review be
accompanied by a standardized test form that identifies the specific, quantitative set points and
objectives that may be actively used throughout the test cycle to identify or record specific
events, measurements or alarms. The consolidated and completed test form from each cycle
should be included in an Appendix of the test report.
7.2 Measurement and Data Acquision
A detailed description of the experimental design and its components will be included in the
QAPP. Specific requirements with regard to use, maintenance, and calibration of equipment,
analytical procedures, chain-of-custody procedures, sample collection, data management and
documentation, records management, project scheduling, experimental design assumptions, and
disclosure of non-standard techniques or equipment will be discussed.
7.3 Verification of Test Data
The data quality objective process will be used to develop the QAPP and establish the locations,
types and numbers of samples to be collected, the quality control samples (duplicates, blanks,
spikes, etc.) required for both field and laboratory samples, and will establish the data quality
criteria and measures of acceptability that are appropriate for the project. The TQAP will also
detail a corrective action plan to describe actions to be taken if acceptance criteria for accuracy,
precision and completeness are not met.
58
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
7.4 Assessment
The effectiveness of QA/QC will be monitored through assessments of general and project-
specific activities. The QAPP will include detailed information on the types of assessments to be
utilized (e.g., management, technical, and/or quality assurance assessments), appropriate
response actions, reporting requirements, and assessment and reporting authority. To increase
facility-to-facility and test-to-test comparisons, and TF internal QA/QC, standardized spiked and
blank samples shall be incorporated into the sample analysis procedures. Spikes may be
accomplished using inert objects such as stained, killed organisms or microbeads of appropriate
size for the specific analyses. The methods to be used for spiked and blank samples shall be
described in the QAPP.
59
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 8
Data Management, Analysis and Presentation
8.1 Data Management
Any data collected during testing activities must be capable of withstanding challenges to its
validity, accuracy, and legibility. Data will be recorded in standardized formats and in
accordance with the following minimum requirements:
• Data are entered directly, promptly, and legibly;
• Hand-entered data are recorded legibly in ink; all original data records include, as
appropriate, a description of the data collected, the unit, the unique sample identification,
the name of the person collecting the data, and the date and time of data collection;
• Any changes to the original entry do not obscure the original entry, document the reason
for the change, and are initialed and dated by the person making the change;
• All deviations from the QAPP must be documented in writing, and approved by the TO;
documentation and communication include an assessment of the impact the deviation has
on data quality; and
• Data in electronic format shall be included in a commercially available program for word
processing, spreadsheet calculations, database processing, or commercial software
developed especially for the data collection and processing on a specific hardware
instrument or piece of equipment; backup of computer databases should be performed on
a daily basis, if possible.
Project-specific data management requirements, including the types of data to be collected and
managed and how they will subsequently be reported, shall be defined in the data handling
section of the TQAP. QA/QC activities for data management will be described in the QAPP and
included in the TQAP.
8.2 Data Analysis and Presentation
Hand-recorded data gathered during verification testing will be entered into electronic format (a
spreadsheet or other database product capable of performing graphical and simple statistical
analyses). Following reduction, data will be presented in a graphical, tabular, or other logical
format and accompanied by a detailed discussion to be included in the verification report.
Treatment effectiveness will be calculated for each size class of ambient organisms as
concentration per unit volume in the discharge and may be related to relevant standards as
identified by the vendor or the vendor's claims. In addition, viability data will be reported for
STOs used in bench-scale experiments. Additional measures or comparisons may also be used
to assess treatment efficacy, including percent organism removal by size class, or as a
comparison of treated discharge to the control tank discharge. All methods will be described in
the TQAP. The treatment effectiveness will be discussed in the verification test report with raw
data included as an appendix.
60
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chapter 9
Environmental, Health, and Safety Plan
The TO shall develop an Environmental, Health, and Safety (EHS) Plan to be included in the
TQAP. The EHS Plan shall identify all environmental concerns and potential hazards associated
with the verification testing process and the TF, as well as the required measures to prevent
exposure to the identified hazards. The TO shall be responsible for informing all personnel at
the test site, including employees, contractors, and visitors, of the potential hazards and safety
measures to be employed at the test site. The EHS plan shall address the following issues, as
applicable:
• Permitting requirements for equipment operation, effluent discharge, and waste disposal;
• Biological, chemical, mechanical, electrical, and other hazards;
• Environmental hazards will be defined in accordance with local, state and federal
regulations;
• Handling, storage, and disposal of all biological material and chemicals associated with
the testing;
• Safeguards and protocols to prevent the accidental release to the environment of any non-
ambient organisms if used in the test process; protocols of the form supplied in Part II
ANS Task Force ANS Research Evaluation Protocol are recommended
(http://www.anstaskforce.gov/Documents/Research_Evaluation_Protocol_ANSTF.pdf);
• Material Safety Data Sheets (MSDS);
• Conformance with the local electrical code;
• Conformance with the local plumbing code;
• Ventilation of equipment, trailers, or buildings housing equipment, if gases generated by
the equipment could present a safety hazard;
• Confined space entry hazards;
• Fire safety; and
• Emergency contacts for 911, the nearest hospital (provide directions), local fire
department, the site manager, and all other important contacts.
Any other environmental, health, or safety issues specific to the test location or ballast water
treatment system to be tested must be addressed. A copy of the EHS plan, including all MSDS,
shall be maintained and readily accessible at the test site. A one-page summary of emergency
contacts shall be placed inside a clear plastic cover and kept at the verification-testing unit.
61
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
References
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.
ASTM. 1996a. Standard Guide for Conducting Acute Toxicity Tests Starting with Embryos of
Four Species of Saltwater Bivalve Mollusks. In Annual Book of ASTM Standards,
Section 11, Water and Environmental Technology, Volume 11.05 (724-94), West
Conshohocken, PA.
ASTM. 1996b. Standard Guide for Conducting Acute Toxicity Tests with Echinoid Larvae, In
Annual Book of ASTM Standards, Section 11, Water and Environmental Technology,
Volume 11.05 (E1563-95). West Conshohocken, PA.
Dobbs, F.C., and A. Rogerson. 2005. Ridding ships'ballast water of microorganisms. Environ.
Sci. Technol. 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. Mar. Pollut. Bull. 52: 259-263.
Drake, L.A., M.K. Steinberg, S.H. Robbins, S.C. Riley, B.N. Nelson, and EJ. Lemieux. 2010.
Development of a method to determine the viability of organisms > 10 jim and < 50 jim
(nominally protists) in ships' ballast water: a combination of two vital, fluorescent stains.
NRL Letter Report 6130/1011, Washington, D.C.
Garthright, W.E. 1993. Bias in the logarithm of microbial density estimates from serial
dilutions. Biometrical Journal 35: 299-314.
Huq, A., C.A. Grim, R.R. Colwell, G.B. Nair. 2006. Detection, Isolation, and Identification of
Vibrio cholerae from the Environment. Current Protocols in Microbiology: 6A.5
Klemm, D.J., G.E. Morrison, TJ. Norberg-King, W.H. Peltier, and M.A. Heber. 1994. Short-
Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Waters to
Marine and Estuarine Organisms (Second Edition) EPA/600/4-91/003, Washington, DC.
Lemieux, E.J., T.P. Wier, M.K. Steinberg, S.H. Robbins, S.C. Riley, B.C. Schrack, W.B. Hyland, J.F.
Grant, C.S. Moser, and L.A. Drake. 2010. Design and preliminary use of a commercial filter skid to
capture organisms > 50 (im in minimum dimension (nominally zooplankton) for evaluating ships'
ballast water management systems at land-based test facilities. NRL Letter Report 6130/1029,
Washington, DC.
Lemieux, E.J., J. Grant, T. Wier, S. Robbins, S. Riley, W. Hyland, L. Davis, T. Donowick, B.
Stringham, WJ. Kinee, B. Brown, andR. Everett. 2008a. Pilot Environmental
Technology Verification (ETV) Test Report of the Severn Trent DeNora BalPure™
Ballast Water Treatment System, NRL Letter Report 6130/6098, Washington, DC.
62
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Lemieux, E.J., S. Robbins, K. Burns, S. Ratcliff, and P. Herring. 2008b. 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.
Marret, F., and K.A.F. Zonneveld. 2003. Atlas of modern organic-walled dinoflagellate cyst
distribution. Rev. Palaeobot. Palynol. 125: 1-200.
Matsuoka, K., and Y. Fukuyo. 2000. Technical Guide for Modern Dinoflagellate Cyst Study.
http://dinos.anesc.u-tokyo.ac.jp/technical_guide/main.pdf.
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.
Limnol. Oceanogr. 33: 146-150.
Richard, R.V., J.F. Grant, and EJ. Lemieux. 2008. Analysis of Ballast Water Sampling Port
Designs Using Computational Fluid Dynamics. Report No. CG-D-01-08. US Coast
Guard Research and Development Center, Groton, CT.
Selvin, R., B. Reguera, I. Bravo, and C.M. Yentsch. 1988. Use of Fluorescein Diacetate (FDA)
as a single-cell probe of metabolic activity in dinoflagellates cultures. Biol. Oceanogr. 6:
505-511.
Steinberg, M.K., S.C. Riley, EJ. Lemieux, and L.A. Drake. 2010. Mult-site Validation of a
Method to Determine the Viability of Organisms > 10 |im and < 50 jim (Nominally
Protists) in Ships' Ballast Water Using Two Vital, Fluorescent Stains. Letter Report No.
6130/1016, Published by the US Naval Research laboratory, Washington, D.C.
63
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Appendix A: Quality Assurance Project Plan Outline
64
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Quality Assurance Project Plan
A Quality Assurance Project Plan (QAPP) shall be prepared as part of the TQAP for evaluating
the performance of ballast water treatment technologies. Information on preparing QAPPs is
provided on the EPA web site: http://www.epa.gov/quality. The generic format for QAPPs
include:
A.I Project Descriptions, Objectives and Organization
• The purpose of the study shall be clearly stated.
• The processes to be evaluated will be described.
• The TF, apparatus and technology set-up will be fully described.
• Project objectives shall be clearly stated and identified as being primary or non-primary.
• Responsibilities of all project participants shall be identified. Key personnel and their
organizations shall be identified, along with the designation of responsibilities for planning,
coordination, sample collection, measurements (i.e., analytical, physical, and process), data
reduction, data validation (independent of data generation), data analysis, report preparation,
and quality assurance.
A.2 Experimental Approach
• Technology installation and shakedown procedures will be identified.
• Technology startup procedures will be identified. Startup will comprise a number of tasks to
implement and check operating and sampling protocols. Tasks will include establishing feed
makeup (including and procedures for challenge water preparation) and performing
calibration checks on monitoring systems, identifying sampling and monitoring points and
identifying the types of samples to be collected.
• Physical, analytical or chemical measurements to be taken during the study will be provided.
Examples include flow rates, pH, salinity, total suspended solids, particulate organic matter,
dissolved organic matter, dissolved oxygen, dissolved nutrients, biochemical oxygen
demand, biological organisms, O&M performance indicators, etc.
• Sampling and monitoring points for each test unit and the type of sample to be collected
(grab or composite) will be identified.
• The frequency of sampling and monitoring as well as the number of samples required will be
provided. This includes the number of samples needed to meet QA/QC objectives.
• Planned approach for evaluation objectives (data analysis). This will include formulas, units,
and definition of terms and statistical analyses to be performed in the analysis of the data.
Example graphical relationships will be provided.
• Demobilization of the technology, including scheduling and site restoration requirements,
will be described.
A.3 Sampling Procedures
• Whenever applicable or necessary to achieve project objectives, the method used to establish
steady-state conditions shall be described.
65
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
• Each sampling/monitoring procedure to be used shall be described in detail or referenced. If
compositing or splitting samples is required, those procedures shall be described.
• Sampling or monitoring procedures shall be appropriate for the matrix or analyte being
tested.
• If sampling/monitoring equipment is used to collect critical measurement data (e.g., used to
calculate the final concentration of a critical parameter), the QAPP shall describe how and at
what frequency the sampling equipment is calibrated.
• If sampling/monitoring equipment is used to collect critical measurement data, the QAPP
shall describe how cross-contamination between samples is avoided.
• When representativeness is essential for meeting a primary project objective, the QAPP shall
include a discussion of the procedures to be used to assure that representative samples are
collected.
• A list of sample quantities to be collected, and the sample amount required for each analysis,
including QC sample analysis, shall be specified in the QAPP.
• Containers used for sample collection for each sample type shall be described in the QAPP.
• Sample preservation methods (e.g., refrigeration, acidification, etc.) and holding times shall
be described in the QAPP.
• A sample of the chain of custody form to be used during testing shall be provided, including
records of times and other critical parameters such as storage temperatures, light condition,
etc.
A.4 Testing and Measurement Protocols
• Each measurement method to be used shall be described in detail or referenced in the QAPP.
Modifications to EPA-approved or similarly validated methods shall be specified.
• For unproven methods, the QAPP shall provide evidence that the proposed method is capable
of achieving the desired performance.
• For measurements that require a calibrated system, the QAPP shall include specific
calibration procedures, and the procedures for verifying both initial and continuing
calibrations (including frequency and acceptance criteria, and corrective actions to be
performed if acceptance criteria are not met).
A.5 QA/QC Checks
A.5.1 Data Quality Indicators
• Statistical analyses shall be carried out on data obtained for all performance measurements.
As part of the assessment of data quality, six data quality indicators (DQIs) can be used to
interpret the degree of acceptability or utility of the data. At a minimum, the QAPP shall
include a protocol for assessing the following DQIs, and acceptable limits and criteria for
each of these indicators: representativeness, accuracy, precision, bias, comparability, and
completeness.
• The TO shall determine acceptable values or qualitative descriptors for all DQIs in advance
of verification testing as part of the experimental design. The assessment of data quality will
require specific field and laboratory procedures to determine the data quality indicators. All
details of DQI selection and values shall be documented in the QAPP.
66
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
A.5.1.1 Representativeness
Representativeness refers to the degree to which the data accurately and precisely represent the
conditions or characteristics of the parameter represented by the data. In this testing,
representativeness will be ensured by executing consistent verification procedures.
Representativeness will also be ensured by using each method at its optimum capability to
provide results that represent the most accurate and precise measurement it is capable of
achieving. For equipment operating data, representativeness entails collecting a sufficient
quantity of data during operation to be able to detect a change in operations.
A.5.1.2 Accuracy
For water quality analyses, accuracy refers to the difference between a sample result and the
reference or true value for the sample. Loss of accuracy can be caused by such processes as
errors in standards preparation, equipment calibrations, loss of target analyte in the extraction
process, interferences, and systematic or carryover contamination from one sample to the next.
Loss of accuracy for microbial species can be caused by such factors as error in dilution or
concentration of microbiological organisms, systematic or carryover contamination from one
sample to the next, improper enumeration techniques, etc. The TO shall discuss the applicable
ways of determining the accuracy of the chemical and microbiological sampling and analytical
techniques in the TQAP.
For equipment operating parameters, accuracy refers to the difference between the reported
operating condition and the actual operating condition. For water flow, accuracy may be the
difference between the reported flow indicated by a flow meter and the flow as actually
measured on the basis of known volumes of water and carefully defined times. Meters and
gauges must be checked periodically for accuracy, and when proven dependable over time, the
time interval between accuracy checks can be increased. In the TQAP, the TO shall discuss the
applicable ways of determining the accuracy of the operational conditions and procedures.
From an analytical perspective, accuracy represents the deviation of the analytical value from the
known value. Since true values are never known in the field, accuracy measurements are made
on the analysis of QC samples analyzed with field samples. QC samples for analysis shall be
prepared with laboratory control samples, matrix spikes, and spike duplicates. It is
recommended for verification testing that the TQAP includes laboratory performance of one
matrix spike for determination of sample recoveries. Recoveries for spiked samples are
calculated in the following manner:
1000SSK SR) .. ^
% Recovery = (A-l)
SA
where: SSR = spiked sample result
SR = sample result
SA = spike amount added
Recoveries for laboratory control samples are calculated as follows:
67
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5. 1
n/ _, lOO(foundconcentmtion) .. _.
% Recovery = - ^ - '- (A-2)
trueconcen tration
For acceptable analytical accuracy under the verification testing program, the recoveries reported
during analysis of the verification testing samples must be within control limits, where control
limits are defined as the mean recovery plus or minus three times the standard deviation.
A.5.1.3 Precision
Precision refers to the degree of mutual agreement among individual measurements and provides
an estimate of random error. Analytical precision is a measure of how far an individual
measurement may be from the mean of replicate measurements. The standard deviation and the
relative standard deviation recorded from sample analyses may be reported as a means to
quantify sample precision. The coefficient of variation (CV) may be calculated in the following
manner:
(A.3)
X
average
where: S = standard deviation
= the arithmetic mean of the recovery values
Standard Deviation is calculated as follows:
X)
— — (A-4)
n 1
where: X; = the individual recovery values
X = the arithmetic mean of the recovery values
n = the number of determinations
The QAPP shall list and define all other QC checks and/or procedures (e.g., detection limits
determination, blanks, spikes, surrogates, controls, etc.) used for the project.
For each specified QC check or procedure, required frequencies, associated acceptance criteria,
and corrective actions to be performed if acceptance criteria are not met shall be included in the
QAPP.
A.6 Data Reporting, Data Reduction, and Data Validation
• The reporting requirements (e.g., units) for each measurement and matrix shall be identified
in the QAPP.
• Data reduction procedures specific to the project shall be described, including calculations
and equations.
68
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
• The data validation procedures used to ensure the reporting of accurate project data to
internal and external clients should be described.
• The expected product document that will be prepared shall be specified.
A.7 Assessments
Whenever applicable, the QAPP shall identify all audits (i.e., both technical system audits
[TSAs] and performance evaluations [PEs]) to be performed, who will perform these audits, and
who will receive the audit reports.
A.8 References
References shall be provided in the QAPP in the body of the text as appropriate.
69
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Appendix B: Anderson, et al. Approach for Evaluation of Standard
Test Organisms
Taken From:
Final Report
To NSF International
For
Contract No 03/06/394
Woods Hole Oceanographic Institute
Woods Hole, MA 02543
"Screening of Surrogate Species for
Ballast Water Treatment"
3 July, 2008
70
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
MATERIALS AND METHODS
Common Methods
Synthetic Water Preparation
The freshwater and seawater used in these experiments were prepared in each Pi's lab using
synthetic solutions for uniform consistency and replication purposes. The seawater for these
experiments was prepared by dissolving Instant Ocean® salts in Milli-Q or equivalent deionized
water to achieve the desired salinity. Freshwater was prepared either as WC medium (Guillard,
1975) (for heterotrophic protists and phytoplankton) or according to the EPA's recommended
aquatic toxicology testing protocol for freshwater organisms
(http://www.epa.gov/waterscience/WET/disk2/) (for bacteria and zooplankton). Freshwater and
seawater were prepared using Milli-Q water or equivalent water (bacteria and zooplankton only)
and was filtered through glass fiber filters (marine zooplankton) or filter-sterilized (all other taxa
and water combinations) before use. Additional details are found in the taxon-specific methods
sections.
Treatments
The treatments, their associated vendors, dose, and literature citation are summarized in Table 1.
Phase I tests separately involved thermal, glutaraldehyde, and hypochlorite treatments. All other
treatments were administered in Phase II tests. The methods common to all taxa tested are
provided in the text following the table. Please note there occasionally were deviations from the
specific concentrations or times listed here, and those deviations, as well as other specific
methods, are given in the taxon-specific section of the Materials and Methods. In particular, the
data sets for the zooplankton species are more extensive than for the other three taxa.
Thermal treatment: Regardless of the organisms or the volume of the experimental container in
which they were held, the thermal treatment was begun by immersing the container into a water
bath (accurate to ± 0.5° C) heated to the test temperatures (35°, 40°, 45°, and 50° C). Timing of
the treatment began once the water in the container reached the criterion temperature. After 4
hours (bacteria and zooplankton) or 8 hours (heterotrophic protists and phytoplankton) of
treatment, viability was assessed. Controls consisted of vessels containing organisms held for
the same time at room temperature.
Chlorine (sodium hypochlorite): Household bleach (UltraClorox® Regular Bleach) has a
concentration of 6.0% sodium hypochlorite. Appropriate volumes were added to generate four
experimental concentrations of sodium hypochlorite (0.25, 0.5, 1.0, and 2.0 mg/L). Additional
concentrations were tested in the zooplankton study; see details in the taxon-specific section.
After 24 hours of exposure, viability was assessed. Control vessels containing organisms were
not treated with hypochlorite, but were held for the same time at room temperature.
71
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Table 1. Summary of treatment and experimental conditions.
Treatment
Thermal
treatment
Chlorine
(sodium
hypochlorite)
Chlorine
dioxide
(Ecochlor™)
Glutaraldehyde
UV light
Ozone
Hydrogen
peroxide
Deoxygenation
SeaKleen®
PeraClean®
Ocean
Vendor or
source
Water bath
Chlorox bleach
Ecochlor, Inc.
Fisher Scientific
UV collimator
designed and built
by Dr. E. "Chip"
Blatchley, Purdue
University
Enaly OZX-300U
or Clear Water
Tech UV-275
Local drugstore
BBL GasPak
System™
Vitamar, Inc.
Degussa AG
Concentration or
Intensity
35°, 40°, 45°,
50° C
Aqueous solution of
sodium hypochlorite.
Final cones, of 0.25,
0.5, 1.0, 2.0mg/L
Final cones, of 1 , 2,
4, 6 ppm
Final cones, of 50,
100, 500 and 1000
mg/L
UV light (256 nm) at
10, 25, 50, 100
mJ/cm
Total initial residual
oxidant (TRO) level
of 0.25, 0.5, 1.0, and
2.0 mg Br2/L in
seawater
Final cones, of 0.5,
1, 10 and 20 ppm
Anoxia (0 mg/L
oxygen)
0.25,0.5, 1.0,2.0
mg/L active
ingredient
Final cones, of 50,
100, 200 and 400
ppm
Exposure
time
4 hours or 8
hours
24 hours
24 hours
24 hours
Dose inde-
pendent of
exposure time
between c. 30
sec. to 2 min.
24 hours after
achieving initial
level of TRO
24 hours after
achieving intial
cone.
24, 48, 72 hours
24 hours
24 hours
References
Hallegraeffetal. 1997;Rigby
etal. 1999
Sano et al. 2004; Bolch and
Hallegraeff 1993
T. Perlich, Echochlor, Inc.
(pers. comm, 19 & 20 Oct.
2004)
Sano et al. 2003
Azanza et al. 2001; Montani
et al. 1995; Sutherland et al.
2001; Sutherland et al. 2003
Hoigne 1998; Langlais et al.
1991; Cooper etal. 2002
Kuzirian et al. 2001
Tamburri et al. 2002; P. D.
McNulty, NEI Treatment
Systems, Inc. (pers. comm..
27 Oct. 2004)
Cutler et al. 2003; Sano et al.
2004
http://dmses.dot.gov/docimag
es/pdf81/175321_web.pdf
Note: NB SeaKleen® was tested only with zooplankton species and Geobacillus stearothermophilus, as it was not
provided by its manufacturer for general distribution to the Pis.
Chlorine dioxide (Ecochlor1™): Species were tested against four concentrations (1.0, 2.0, 4.0,
and 6.0 ppm) of chlorine dioxide (Ecochlor™). Reagents were provided by Ecochlor, Inc.
(Acton, MA). A concentrated stock solution (3,000 ppm) was prepared as per manufacturer
directions by dissolving the two-part reagent into one liter of distilled water. This stock was
maintained in the refrigerator in a glass bottle until required. To make working solutions,
appropriate volumes were added to either WC or sterile artificial seawater to generate the four
experimental concentrations listed above. Additional concentrations were tested in the
72
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
zooplankton study; see details in the taxon-specific section. Treatment times were 24 hours in all
cases, during which time experimental and control organisms were incubated at room
temperature. After 24 hours, viability was assessed.
Glutaraldehyde: Glutaraldehyde (Fisher Scientific, Cat. No. G151-1 or for the zooplankton and
bacteria studies, 25% solution from Sigma Chemical Company) was used at concentrations of
50, 100, 500, and 1000 mg/L in either artificial seawater or WC medium. Additional
concentrations were tested in the zooplankton study; see details in the taxon-specific section.
Controls consisted of organisms incubated in all ways the same, save the addition of
glutaraldehyde. After 24 hours of treatment at room temperature, the viability of the organisms
was determined.
Ultraviolet light: This treatment utilized a collimator built by Dr. E. "Chip" Blatchley, Purdue
University, which has a 256 nm UV bulb to provide controlled dosages of UV light onto the test
surface. Different dosages were delivered by varying the length of time organisms were
exposed. The protocols employed were developed by Drs. Russell Herwig and Adelaide Rhodes
(University of Washington). The test organisms and medium were transferred into disposable 50
mm diameter petri dishes, which were positioned ca. 1 cm beneath the end of the collimator.
Dosages used for zooplankton, protists, and phytoplankton were: 10 mJ/cm2 (72 seconds); 25
mJ/cm2 (180 seconds); 50 mJ/cm2 (360 seconds); and 100 mJ/cm2 (720 seconds). Similar
dosages were used for bacteria but since these experiments were performed later in the contract
period and following extensive usage of the same UV lamp, the exposure time needed to be
slightly increased to achieve the similar doses. The doses (time of exposure) for the bacteria
were: 10 mJ/cm2 (80 seconds); 25 mJ/cm2 (200 seconds); 50 mJ/cm2 (400 seconds); and 100
mJ/cm (800 seconds). Additional higher dosages were tested in the zooplankton study since
some of the species were tolerant of 100 mJ/cm2; see details in the taxon-specific section.
During the UV exposure, organisms suspended in freshwater or seawater in the petri dishes were
stirred with a small magnetic stir bar to ensure equal exposure of all organisms placed in the
dish. The control exposures were conducted in the same way, but the UV light was not turned
on.
Ozone: Ozone (Os) is a gas that is also an oxidizing biocide. It is commonly used for the
disinfection of freshwater and seawater, such as in drinking water, freshwater and seawater
aquaria. Ozone can be generated in a corona-discharge tube by passing an electrical current
through an atmosphere that is enriched in oxygen. A variety of commercial ozone generators are
available, but we decided to use a small inexpensive unit that could generate ozone for bench-
scale experiments. We chose the Enaly (http://www.ozone.enaly.com/, Shanghai, China) OZX-300 U
unit with air dryer and built-in aerator.
We calibrated the unit using synthetic seawater and freshwater (R. Herwig and A Rhodes,
University of Washington). To calibrate the instrument, ozone was bubbled into test water for
various times, and Total Residual Oxidant (TRO) levels were measured for seawater and the
ozone concentrations were measured for freshwater. Ozone in freshwater is toxic by direct
contact with the organism. In seawater, ozone quickly reacts with bromide ions (Br") to form
bromines, compounds that are long lasting disinfectants. In freshwater, this transformation does
not occur. The concentration of bromines was measured using the DPD (N,N-diethyl-p-
73
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
phenylenediamine) colorimetric test (Hach DREL/2010 spectrophotometer with the DPD Total
Chlorine Powder Pillows (US EPA Method 8016), Hach Company, Loveland, CO). Using a
calibration curve (Figure 2), the times required to achieve the desired concentration of ozone
were calculated. A similar procedure was used to calibrate freshwater ozone concentrations. We
directly measured ozone concentrations in freshwater at various times using a Hach DREL/ 2010
spectrophotometer with Ozone Accuvac ampoules that measure the formation of indigo (US
EPA Method 8311). The freshwater ozone test was difficult to perform because of the rapid
disappearance of ozone, so we repeated the calibration three times (Figure 3).
To expose the zooplankton, phytoplankton, and protists to ozone, we placed the test organisms in
500 ml of seawater or freshwater, inserted the airstone provided by the manufacturer (Enaly) into
the liquid, and let the generator run for the required time. Ozone testing with bacteria followed a
slightly different protocol because small volumes of liquid were treated. For freshwater testing,
ozone was produced using the Enaly ozone generator. The UV-275 (ClearWater Tech, LLC, San
Luis Obispo, California), which produces ozone using UV light, was used for testing seawater
conditions, because the Enaly generator produced ozone too rapidly.
TRO Concentration for Marine Treatments vs. Time
9 I
S 7 -
|6H
r s H
o
o
1 -
0
y= 1.1803X+ 0.203
R2 = 0.9981
O
O Trial 1
D Trial 2
O Trial 3
X Average of All Trials
— Linear (Average of All Trials)
0
3 4
Time
Figure 2. TRO concentrations (measured as mg Cli/L) for synthetic seawater treated with
ozone versus time using the Enaly ozone generator.
74
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
1.2
o>
E 0.8
0.6
0.4
0.2
0
Ozone Concentrations for Freshwater
Treatments vs. Time
y = 0.4923Ln(x) - 0.4348
R2 = 0.9763
10 15 20
Time
25
Figure 3.
generator.
Ozone concentrations for freshwater versus time using the Enaly ozone
Exposure times for each experimental run were based on a standard curve of ozone
concentrations versus time (Figure 4) (Russell Herwig and Jake Perrins, University of
Washington). Containers holding test bacteria were injected with ozone to four experimental
TRO concentrations (0.25, 0.50, 1.0, and 2.0 mg Br2/L). Following exposure, organisms were
held for 24 hours at room temperature, after which their viability was determined. Controls were
similarly treated, but ozone was not injected into the test containers. Additional ozone
concentrations were tested in the zooplankton study; see details in the taxon-specific section.
Hydrogen Peroxide: Hydrogen peroxide (3%) was purchased locally at drug stores by the Pis.
Only unopened bottles were used to make treatment solutions. Appropriate volumes were added
to either WC or sterile artificial seawater to generate four experimental concentrations of
hydrogen peroxide (0.5, 1.0, 10.0, and 20.0 ppm). Additional concentrations were tested in the
bacteria and zooplankton studies; see details in the taxon-specific section. All treatments lasted
24 hours at room temperature. Viability was assessed thereafter.
Deoxygenation: The deoxygenation tests used a BBL™ GasPak™ anaerobic system (Becton
Dickinson, Becton Drive, Franklin Lakes, NJ 07417). The GasPak™ system is self-contained
and produces anaerobic conditions when used with hydrogen and CO2 generating GasPak™
envelopes (BD# 271040). Test organisms and media were placed in Petri dishes and the dishes
placed in the chamber, together with GasPak™ dry anaerobic test strips (BD # 271051). The
strips were monitored for color change when conditions became anaerobic, approximately 2
hours after activation. The containers were sealed to prevent infiltration of any oxygen over the
time course of the experiment, and treatments lasted 24, 48, and 72 hours at room temperature.
75
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
m
V)
F1
i
Ozone diffusion in saltwater
2 5
9
1 5 -
1
0 5
n
y = 0.5033X - 0.2673
R2 = 0.9914 *
£^
^^^
.S +
^^
^^r^
+^
01 23456
Time (min)
Ozone diffusion in freshwater
t
0.75
0.25
Time (min)
Figure 4. Ozone standard curves used to test bacteria in seawater and freshwater. (For
seawater, bromine was measured as TRO. For freshwater, ozone was directly measured.)
Experimental controls consisted of containers not exposed to anaerobic conditions for same time
periods. After treatment, the viability of the organisms was determined.
SeaKleen® (zooplankton and Geobacillus stearothermophilus only): As mentioned earlier,
Vitamar, Inc. chose not to distribute SeaKleen® to the Pis. The zooplankton and bacteria
investigators, however, had some material remaining from previous work and were able to
perform a series of experiments. See the taxon-specific methods for details.
PeraClean® Ocean: The manufacturer of PeraClean® (Degussa Corporation, 379 Interpace
Parkway Parsippany, NJ 07054) supplied the product, certified to contain 15.4% peracetic acid
and 14.2% hydrogen peroxide. PeraClean® volumes were added either to WC or sterile artificial
seawater to generate four experimental concentrations (50, 100, 200, and 400 ppm). Treatment
76
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
times were 24 hours in all cases and tubes were incubated at room temperature. Controls
consisted of containers identical in all ways save the addition of PeraClean® Ocean.
Experimental Replication and Controls
Each combination of treatment level and species tested were run in quadruplicate, as were the
controls, for the bacteria, heterotrophic protists, and phytoplankton. Zooplankton experimental
combinations were replicated at least three times. In any case where high mortality (defined as
>20%) was determined for the controls, the experiment was repeated until controls achieved an
acceptable level of survival.
Viability Determinations
For the potential surrogate species studied here, viability was determined either though an
assessment of their growth potential or in the case of zooplankton species, their ability to move
or respond to mechanical stimulation. Thus, two different approaches were followed, one
culture-based and the other behavioral.
Bacterial viability was determined based on the ability of the species to grow on agar plates
comprised of a standard general or selective bacteriological medium. This classic technique
yields so-called "colony-forming units" (CPUs) that can be pointed to and counted.
Effectiveness of a treatment was expressed as the log-scale mortality in the treatment relative to
the control.
Viability of both heterotrophic protists and phytoplankton was assessed using the "most probable
number" (MPN) method (Throndsen 1978), also known as the "extinction dilution method" of
Imai et al. (1984). The MPN data yield an estimate of the abundance of cells (or cysts) capable
of dividing (or germinating and dividing). It is important to understand the dormancy and
excystment characteristics of the species being investigated if this method is to be used, since
cysts sometimes will not germinate, but are nevertheless viable in the long term.
Briefly, the MPN counting method begins with pipetting medium into a set of tubes, one set for
each of the four replicate tubes of organisms (experimental or control). The inoculum volume
added to each MPN tube differs by a factor of 10. After mixing, samples of the MPN series are
then transferred to multiwell plates and following an appropriate incubation period, wells are
examined by inverted microscopy to determine cell viability. The number of positive scores
(i.e., the presence of viable cells in a well) is entered into a Most Probable Number program
(Blodgett, 2003) to determine the number of organisms per ml. Percent survival is calculated by
comparing the MPN from the treated samples with the MPN of the control after incubation. We
emphasize this MPN technique is suitable for pure cultures of heterotrophic protists or
phytoplankton, but is not a useful tool for mixed cultures.
To assess viability of zooplankton following treatment, animals were placed in a counting wheel
and examined under at least 10* magnification to determine their survival (indicated by motion)
or mortality (lack of movement after a few seconds of observation or prodding with a wire
pointer).
77
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Bacteria Methods
General
The Phase I assessment utilized two Gram negative pathogen-indicator organisms, Enterococcus
avium (ATCC 14025) and Vibrio cholerae (ATCC 14033), and two Gram positive spore-forming
organisms, Geobacillus stearothermophilus (ATCC 7953), and Bacillus subtilis (ATCC 6633).
E. avium and V. cholerae were purchased from the American Type Culture Collection (ATCC)
(www.atcc.org) as freeze dried pellets of cells. G. stearothermophilus and B. subtilis were
purchased as spores from SGM Biotech, Inc. (www.sgmbiotech.com). Spores of these species are
also available from other commercial sources. Spore suspensions from these and other bacterial
species are commonly used for disinfection experiments and the testing of disinfectants, so the
suspensions are routinely available. Phase II experiments were conducted solely with G.
stearothermophilus, based on its superior performance (as a potential surrogate species) in Phase
I of the investigation. G. stearothermophilus is a thermophilic bacteria that has a recommended
incubation temperature of 55° C.
At the beginning of an experiment, sterile containers (test tubes or petri dishes) were inoculated
with a live bacterial culture or spore suspension to achieve a final concentration of 104 to 105
cells/mL. Two methods were used to provide this concentration. In the first, freeze-dried pellets
of E. avium and V. cholerae were rehydrated and streaked onto agar media, as suggested by
ATCC, and incubated for 24 hours at 37° C. Colonies were harvested with a sterile inoculating
loop, and plated onto general purpose agar media, Trypticase Soy Broth Agar for E. avium, and
Nutrient Agar for V. cholerae. To collect cells for the surrogate challenge tests following a 24 hr
incubation, colonies on the agar media were harvested with a sterile inoculating loop and
suspended in Phosphate Buffered Saline (PBS). The cell suspensions were placed in a
spectrophotometer and the cell concentration was determined by measuring the absorbance at
450 nm. An absorbance of 0.05 was equivalent to a concentration of 106 cells/mL. The bacteria
suspension was then distributed into sample containers to yield a final concentration of 104 to 105
cells/mL.
In the second method, G. stearothermophilus and B. subtilis spore suspensions were purchased at
1 &
the vendor-reported spore concentrations of 10 and 10 spores/mL. Spores were diluted to an
estimated 106 spores/mL in PBS. An aliquot of the PBS spore suspension was then added to the
sample containers to have a final concentration of 104 spores/mL.
In either case, to ensure uniform bacteria distribution among replicate treatments and controls, all
bacterial cell and spore aliquots were removed from the same intermediate PBS suspension for
use in a specific surrogate test. The stressor was added or applied after bacteria were diluted into
the synthetic freshwater or seawater in the sample containers. The total experimental volume
(after bacterial cell or spore, and stressor addition) was 10.0 mL for experiments conducted in 16
x 125 mm screw-capped test tubes, and 5.0 mL for the two experiments (UV and anaerobic
treatment) conducted in 50 mm diameter sterile plastic petri dishes.
After treatment, culturable bacteria were enumerated as colony forming units (CPU) using the
agar spread-plate method. Petri dishes (100 mm diameter) containing agar medium were
78
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
inoculated directly with a 0.1 mL (10"1 dilution) aliquot from the treatment or control sample,
which was spread evenly over the agar surface using a sterile, bent-glass rod. Nutrient Agar was
used to enumerate G. stearothermophilus, B. subtilis, and V. choleras cells, and Trypticase Soy
Broth Agar (TSBA) was used to enumerate E. avium. Two serial dilutions (10"2 and 10"3
dilutions) were prepared as necessary to ensure an accurate count when the number of culturable
bacteria was greater than 200 to 300 CPU per inoculated petri dish on the 10"1 dilution. The
spread-plate samples were inoculated in triplicate (three plates per dilution) and incubated in the
dark for 24 hours. Petri plates for E. avium, V. cholerae, and B. subtilis were incubated at 37° C,
and G. stearothermophilus plates were incubated at 55° C in a dry incubator. After incubation,
bacterial colonies were counted and the data were reported as CFU/mL.
Thermal Treatment (Phase I)
As described in "Common Methods". Following treatment, bacteria were enumerated (as above)
after the incubation period.
Biocide Treatments (Phase I and II)
Biocides (sodium hypochlorite, chlorine, chlorine dioxide (Ecochlor™), glutaraldehyde,
hydrogen peroxide, Seakleen®, and PeraClean® Ocean—all except ozone, see below) were
added to test tubes as aqueous solutions, and were prepared from liquid or solid stocks using
deionized water. Biocide stocks were prepared at ten times their final concentrations. A volume
of 1.0-mL of the stock solution was added to treatment test tubes containing 9.0 mL of the
bacteria or spore suspension in synthetic freshwater or seawater, resulting in the desired final
biocide concentration. Control test tubes received 1.0 mL of deionized water in place of the
treatment solution. After addition of the biocide, treatment and control test samples were mixed
with a vortexer and held in the dark at room temperature for 24 hours. Bacteria were enumerated
(as above) following this incubation period.
Ultraviolet Light Treatment (Phase II)
The UV treatment tests were performed as described in "Common Methods". Specifically for
bacteria, UV treatment was conducted by putting 5.0-mL samples containing 104 bacterial cells
or spores/mL in 50 mm sterile plastic petri dishes, and placing the dish, with the lid removed,
under the collimated UV beam. Bacteria in the dish were gently stirred with a small clean
magnetic stir bar during treatment. After the treatment, the petri dishes were covered with the
plastic lid, and held in the dark at room temperature for 24 hours before enumerating culturable
bacteria by counting colonies on the inoculated media.
Ozone Treatment (Phase II)
Prior to ozone treatment, 1.00-L volumes of synthetic seawater and freshwater were filter
sterilized and amended with bacteria or spores to achieve a final concentration of 104 cells/mL.
Fifty (50.0) mL of these bacterial suspensions were placed into sterile, 500-mL graduated glass
cylinders using a sterile pipet, and ozonated for the required time to achieve the target ozone
dose or Total Residual Oxidant (TRO) level. The graduated cylinder reduced the water volume
79
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
necessary to submerge the air stone diffuser and allowed for a tall water column, which increased
ozone absorption efficiency. After treatment, a 10-mL aliquot of ozonated water was transferred
from the graduated cylinder into a screw-cap test tube and held in the dark at room temperature
for 24 hours. Control samples (also 10 mL) were taken from the original 1.00-L volume of the
bacteria-amended water, and held in the dark at room temperature for 24 hours. Viable bacteria
remaining in both control and treatment test tubes after 24 hours were plated on agar media.
Viable organisms were reported as CPUs.
Deoxygenation Treatment (Phase II)
The deoxygenation treatment was generally performed as described in "Common Methods".
Specifically for bacteria, 5.0-mL samples containing 104 bacterial cells or spores/mL of a given
species were placed in 50 mm sterile plastic petri dishes. The combination of the small sample
volume and its placement in a petri dish yielded a large surface area for the sample, reducing the
time required to achieve anoxic conditions. Petri dishes were placed inside a BD BBL GasPak®
Jar and held for 48 hours in the dark at room temperature. The 5.0 mL control samples
containing 104 bacteria cells/mL also were placed in 50 mm plastic petri dishes and held in the
dark, not under anaerobic conditions, at room temperature for 48 hours. After 48 hours, petri
dishes were removed from the dark and the GasPak® jars. Plates were agitated before removing
an aliquot for plating, incubation, and enumeration of culturable bacteria.
Heterotrophic Protist Methods
Sources of Organisms
Acanthamoeba sp. was obtained from stocks maintained by the Oceanographic Center (OC) of
Nova Southeastern University (NSU). This strain was originally isolated from soil adjacent to
Hollywood beach, Florida. Chilomonas sp. was purchased from Carolina Biological Supply.
Rhynchomonas sp. was isolated from mangrove water in the vicinity of the OC of NSU. Axenic
Tetrahymena pyriformis were purchased from the American Type Culture Collection (ATCC,
Maryland, USA) and Vannella anglica and Vanella platypodia from the Culture Collection of
Algae and Protozoa (CCAP, Dunstaffnage, UK). Uronema sp. was isolated from a beach in
Oregon by A. Hartz.
Culture Methods
The seven heterotrophic protists tested (Figure 5) were cultured and maintained in various media
formulations as summarized in Table 2. The freshwater amoebae, Acanthamoeba sp. and
Vannella platypodia, were cultured on non-nutrient agar plates (NNAS) streaked with E. coli as a
food source. In the case of the acanthamoebae, when food was depleted cyst formation was
induced; a typical Petri dish generated ca. 10 million cysts within one week. The freshwater
flagellate Chilomonas sp. was cultured in amoeba saline (AS) with an added sterile rice grain.
Nutrients leaching from the rice stimulated attendant bacteria in the culture to grow and provided
prey for this bactiverous protist. The freshwater ciliate Tetrahymena pyriformis was maintained
in axenic culture in tubes of proteose peptone yeast (PPY) medium. The marine flagellate
80
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Rhynchomonas sp. and the ciliate Uronema sp. were cultured in filtered sterile seawater seeded
with rice grains (again to stimulate the growth of bacterial prey). The saltwater amoeba Vannella
anglica was cultured on malt yeast agar plates made with 75% seawater (MY75S). All cultures
were incubated at 21°C and subcultured weekly to maintain healthy, exponentially growing
stocks. The one exception was Acanthamoeba sp., which survived well in encysted form. This
amoeba was cultured every three weeks.
NNAS was made by dissolving 15 g non-nutrient agar in 1 L amoeba saline. AS was prepared
from five stock solutions: NaCl, 1.20 g/100 mL; CaCl2, 0.04 g/100 mL; MgSO4, 0.04 g/100 mL;
Na2HPO4, 1.42 g/100 mL; KH2PO4, 1.36 g/100 mL. Ten (10) mL of each stock solution was
added to 950 mL dH2O. PPY was made with proteose peptone (20.0 g), yeast extract (2.5 g),
and dH2O (1 L). MY75S comprised sterile filtered natural seawater (750 mL), dH2O (250 mL),
malt extract (0.1 g), yeast extract (0.1 g), and bacteriological agar (15.Og).
Table 3. Growth media used for the routine cultivation of surrogate protists.
Surrogate Growth Conditions
1 Acanthamoeba sp. cysts
2 Chilomonas sp.
3 Rhynchomonas sp.
4 Tetrahymena pyriformis
5 Uronema sp.
6 Vannella anglica
7 Vannella platypodia
NNAS agar seeded with E. coll
Amoeba saline enriched with rice grains
Filtered sterile seawater enriched with rice grains
PPY
Filtered sterile seawater enriched with rice grains
MY75S
NNAS agar seeded with E. coll
Figure 5. Photomicrographs of heterotrophic protists used for Phase I testing. (Top row,
left to right: Acanthamoeba sp. cysts, freshwater, 15-18 (im; Chilomonas sp., freshwater, 13-15 ^im;
Rhynchomonas sp., marine, 4-7 (im; Tetrahymena pyriformis, freshwater, 60 (im; Bottom row, left to
right: Uronema sp., marine, 24 (im; Vannella anglica, marine, 21-24 (im; Vannella platypodia,
freshwater, 16-21 (im. For Phase II testing, only Acanthamoeba sp. cysts, T. pyriformis, and Uronema sp.
were used.)
81
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Thermal Treatment (Phase I)
Six protozoan species plus Acanthamoeba sp. cysts were tested against four temperature
stressors, 35°, 40°, 45°, and 50° C. Three of the trophic protists were freshwater and the others
were marine. The acanthamoebae cysts were from a freshwater source, however these cysts are
unaffected by marine conditions and previous work has shown that trophs even replicate in
seawater (Booton et al., 2004). In all cases, the treatment time was 8 hours at the experimental
temperature. Each experimental run was tested in quadruplicate. The organisms were
transferred to a sterile tube (16 mL) with either 10 mL artificial seawater (Instant Ocean®,
Aquarium Systems, Mentor, OH), in the case of the marine protists, or 10 ml WC medium (Table
3). To concentrate organisms, cells were either washed off agar plates into the respective
medium or rinsed from the base of Petri dishes after pouring off excess medium. Tetrahymena
pyriformis were maintained in axenic PPY medium. Prior to experimentation, cells were
pipetted into amoeba saline and centrifuged three times (3 min at 3000 rpm) to concentrate and
wash cells free of this organically rich medium. Acanthamoeba sp. cysts were tested in both
artificial seawater (SW Acanthamoeba) and WC medium (FW Acanthamoeba). In all cases,
densities of cells in the experimental tubes was sufficient to allow reliable enumeration but not
so excessive as to promote interference between cells. The tubes were immersed in a water bath
to maintain the correct incubation temperature (+/- 0.1 °C). After the 8 hours incubation, the
number of viable organisms was determined using an MPN method (see "Common Methods"
above and next paragraph as well). For each experimental run, a control set of 4 tubes was set up
to ensure that cells remained viable over the experimental period. The controls were kept with
the minimum of bacteria (to limit cell replication) in either artificial seawater or WC medium.
Incubation was at 21° C, a temperature chosen to optimize the survival of the protozoa.
Table 3. Preparation of inorganic medium WC.
1. Prepare 7 Stock Solutions:
a. NaNO3: 85.01 g/L
b. CaCl2-2H2O: 36.76 g/L
c. K2HPO4: 8.71 g/L
d. MgSO4-7H2O 36.97 g/L
e. Na2O3Si-9H2O 14.21 g/L
f. NaHCO3 12.6 g/L
g. Trace Metal Mix
MnCl2-4H2O 0.18 g/L
ZnSO4- 7H2O 0.022 g/L
(NH4)6Mo7O24- 4 H2O 0.0046 g/L
CoCl2- 6H2O 0.012 g/L
CvSO4- 5H2O 0.01 g/L
H3BO3 0.006 g/L
2. Add 1 mL of each of the 7 stock solutions to 993 mL diH2O
82
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
The most probable number (MPN) counting method was used to count the number of protozoa
after treatment (in the experimental tubes and the control tubes). For each tube, medium was
pipetted into a set of MPN tubes containing either amoeba saline or filtered, sterilized seawater
(depending on whether the cells were freshwater or marine). The inoculum volume added to
each MPN tube differed by a factor of 10. One set of MPN tubes were inoculated with 10 mL of
water, 1 mL of water, 0.1 mL of water and 0.01 mL of water. A separate MPN series was set up
for each of the 4 replicate experimental or control tubes. The MPN tubes were sub-sampled after
briefly vortexing (5 sec) to ensure thorough mixing of cells. Nine replicate aliquots of each tube
(20 |il) were added to 1 ml of amoeba saline (or sterile filtered seawater, as appropriate) held in
nine wells of a 24 well tissue culture plate and incubated at 21°C. To promote excystment and
growth of surviving cysts, nutrients were added to stimulate bacterial prey. Two 2 jiL Bacto
Casitone medium (Difco) was added to each well. After 1,3, and 5 days, wells were examined
by inverted microscopy to detect the presence or absence of a population of growing protozoa.
The number of positive scores were entered into a Most Probable Number (MPN) program
(Blodgett, 2003) to determine the number of organisms per ml. Percent survival was calculated
by comparing the MPN from the treated samples with the MPN of the control after incubation.
Statistical analyses were performed on the MPN data using single-factor ANOVA and the
Tukey-Kramer procedure for determining difference in mean. Data was analyzed using the
PHStat2 add-in for Microsoft Excel (Version 10, Prentice Hall, 2001).
Chlorine (sodium hypochlorite) Treatment (Phase I)
Household bleach was added to either amoeba saline or sterile artificial seawater to generate the
four experimental concentrations of sodium hypochlorite. The organisms were transferred to a
sterile tube (16 mL) containing either 10 mL artificial seawater, in the case of the marine protists,
or 10 mL WC medium for freshwater protists. After 24 hours at room temperature, the number
of surviving organisms was determined using the MPN method.
Glutaraldehyde Treatment (Phase I)
As described in "Common Methods". After 24 hours of treatment, the viability of the organisms
was determined using the MPN method.
Chlorine Dioxide (Ecochlor™) Treatment (Phase II)
Three heterotrophic protists (Acanthamoeba sp. cysts, Tetrahymena pyriformis, and Uronema
sp.) were selected for Phase II testing based on the results of the Phase I tests. These three
species were easy to culture in large numbers, were robust during harvesting, and grew out
rapidly in the MPN counting protocols. Moreover, they were easy to observe under the
microscope when scoring positive or negative growth.
The three species were tested against four concentrations of chlorine dioxide (Ecochlor™). The
organisms were transferred to a sterile tube (16 mL) containing either 10 mL artificial seawater
32 ppt (Instant Ocean®, Aquarium Systems, Mentor, OH), in the case of the marine protists, or
10 mL WC medium for freshwater protists.
83
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
To concentrate Acanthamoeba sp. cysts, cells were harvested off agar plates by scraping the
surface with Falcon™ Cell Scrapers. Cysts were decanting into 100 mL of WC medium.
Tetrahymena pyriformis was maintained in axenic protease peptone and yeast (PPY) medium.
Prior to experimentation, cells were pipetted into WC medium and centrifuged three times for
three minutes at 3000 rpm to concentrate and wash. Uronema sp. was harvested from Petri
dishes by washing cells into 100 mL of artificial seawater. In all cases, densities of cells in the
experimental tubes was sufficient to allow reliable enumeration but not so excessive to promote
interference between cells.
After 24 hours incubation, the number of viable organisms was determined using the MPN
method described above. For each experimental run, a control set of 4 tubes was set up to ensure
that cells remained viable over the experimental period. Like the experimental runs, the controls
were kept with the minimum of bacteria (to limit cell replication) in either artificial seawater or
WC medium. Incubation was 24 hours at 21°C, a temperature chosen to optimize the survival of
the protozoa.
The enrichment of surviving cells was similar to protocols used in Phase I with a few minor
modifications. For the ciliates Tetrahymena pyriformis and Uronema sp., 20 uL aliquots were
inoculated into nine wells of a 24 well tissue culture plate containing 1 ml of appropriate media
enriched with the prey bacterium E. coli. Plates were incubated at 21°C for 7 days.
Acanthamoeba sp. cysts were inoculated (20 jiL aliquots) onto non-nutrient agar plates (NNAS)
streaked with E. coli prey. After 7 days, agar plates were examined by dissecting microscopy
with transmitted light to detect the presence or absence of a population of growing amoebae.
The number of positive scores were entered into a Most Probable Number (MPN) program
(Blodgett, 2003) and percent survival was calculated by comparing the MPN from the treated
samples with the MPN of the controls after incubation.
PeraClean® Ocean Treatment (Phase II)
As described in "Common Methods". PeraClean® Ocean volumes were added to either WC or
sterile artificial seawater to generate four experimental concentrations. The organisms were
transferred to sterile tubes (16 mL), each containing 10 mL of the experimental (or control)
media. After 24 hours the number of surviving organisms was determined using the MPN
method.
Hydrogen Peroxide Treatment (Phase II)
As described in "Common Methods". After 24 hours of treatment, the viability of the organisms
was determined using the MPN method.
Ultraviolet Light Treatment (Phase II)
The test organisms were transferred into sterile 60x15 mm polystyrene Petri dishes along with 10
mL of appropriate media. The dish was positioned 1 cm beneath the end of the collimator box
and dosages were delivered as indicated in "Common Methods". After exposure, medium and
84
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
organisms were transferred to sterile 16 mL glass tubes for 24 hours incubation at room
temperature, after which viability of the organisms was determined using the MPN method.
Ozone Treatment (Phase II)
Generally as described in "Common Methods." The test organisms contained in 500 mL of
appropriate medium were placed in 500 mL glass media bottles. Holes were drilled in bottle lids
to accommodate tubing leading to an airstone that was used to deliver the ozone. Ozone was
injected into marine media for 9, 18, 35, and 90 sees to generate four experimental
concentrations of ozone (0.25, 0.50, 1.0, and 2.0 mg Br2/L). Since ozone reaction chemistry
differs in freshwater and seawater (see "Common Methods"), only three experimental
concentrations could be achieved in the case of the freshwater treatments. Here, ozone was
injected into WC media for 4, 8, and 16 minutes to generate the different experimental
concentrations of ozone (0.25, 0.50, and 1.0 mg Br2/L).
After treatment exposure, suspensions were transferred to sterile 16 mL glass tubes for 24 hours
incubation at room temperature. Each treatment, including controls, was replicated 4 times.
After 24 hours of treatment, the viability of the organisms was determined using the MPN
method.
Deoxygenation Treatment (Phase II)
Generally as described in "Common Methods." Test organisms were transferred to sterile 100 x
15 mm polystyrene Petri dishes along with 10 mL of appropriate media and dishes were placed
in the GasPak® chamber. Each treatment, including controls, was replicated 4 times. After 24
hours of treatment, the viability of the organisms was determined using the MPN method.
Phytoplankton Methods
Phase I
Phase I experiments testing the impact of hypochlorite, glutaraldehyde, and thermal stressors
were conducted on eight phytoplankton species of which five were marine forms [Chlorella sp.
(CCMP256), Chaetoceros a/finis (CCMP158), Skeletonema costatum (CCMP780), Scrippsiella
trochoidea (CCMP1599), Scrippsiella lachrymosa (both vegetative and cyst forms, clone: B10-
1)] and four were freshwater forms \Prymnesium parvum (CMS204), Microcystis aeruginosa
(UTEX B 2672) and Fragilaria crotensis (UTCC269)]. The marine species were cultured in
modified f/2 medium with or without silicate (Guillard and Ryther 1962), where Na2SeO3 has
been added to a final concentration of 10"8 M and the concentration of CuSO/iS^O has been
reduced to the same level. For culture upkeep, the medium was prepared in autoclaved, 0.2 jim
filtered natural seawater. The freshwater species were grown in WC medium (Guillard, 1975)
made with Milli-Q water. All cultures were maintained at 20 C on a 14: lOh light: dark cycle (ca.
200 jimol photons-m^-sec"1 irradiance provided by cool white fluorescent bulbs). All marine
culture experiments were conducted in f/2 medium made with sterile filtered Instant Ocean®
water adjusted to a salinity of 32 as the base.
85
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Sterile, disposable polystyrene tubes (17 x 100 mm) filled with 7 mL of mid-exponential growth
phase culture were used for all experiments except for the UV light and deoxygenation trials, for
which sterile, polystyrene Petri dishes (50 x 11 mm) containing 5 mL of culture were used.
The thermal trial tubes were supported on floating racks and incubated in separate water baths
adjusted to the criterion temperatures. After 8 hours, the cultures were transferred to room
temperature and individually pipetted into 96 well tissue culture plates for most probable number
(MPN) determinations. Phytoplankton species were exposed to the other Phase I treatments,
glutaraldehyde and hypochlorite, for 24 hours at room temperature, then pipetted into culture
plates preparatory to MPN determinations.
For MPN estimates, each well of the tissue culture plate was filled with 270 jiL off/2 medium,
with the exception of wells in row H, which were loaded with 300 jiL of treatment culture (6
replicate wells per replicate treatment tube). 30 jiL of the culture from row H wells were then
serially diluted into the wells above using a 12 position multi-channel pipette so that a final
dilution series ranging from 10"° to 10"7 was achieved. The plates were sealed with tape around
the perimeter to minimize evaporative loss, incubated at 20°C and monitored after 21 days with
an inverted microscope at 100x to determine cell viability. A gridded tally sheet was scored with
the results and the data were entered into a most probable number calculator Excel spreadsheet
(Dr. R. Blodgett, Division of Mathematics, FDA/CFSAN) to derive the MPN for the experiment.
Phase II
Based on the results of the Phase I trials, 3 species, S. lachrymosa cysts (B10-1), Chaetoceros
affmis (CCMP158), and Chlorella sp. (CCMP256) were advanced for Phase II testing.
Experiments with chlorine dioxide (Ecochlor™), hydrogen peroxide, and Peraclean® Ocean
were conducted in the same manner described for the glutaraldehyde and hypochlorite Phase I
trials.
For the UV light exposures, 5 mL of culture was loaded into a 50 mm diameter polystyrene Petri
dish and exposed, uncovered, to the collimator UV light. Samples were then transferred into
disposable 15 mL centrifuge tubes so that the samples could be mixed well prior to MPN
dilutions.
For the deoxygenation treatment, 5 mL samples were pipetted into Petri dishes, which were
placed on a perforated plastic support within a glass desiccator. Four desiccators in total were
used - 1 each for the 12, 24, 48, and 72 hour incubations. Two BD BBL™ GasPak sachets,
clipped onto the support in an upright position, were then activated within each desiccator per the
manufacturer's protocol. The desiccators were covered with the airtight lid and were closely
monitored until duplicate indicator strips revealed that anaerobic conditions were achieved. This
event marked the "time 0" for the experiments. At the termination of the incubation time, the
indicator strips were again checked to ensure that anaerobic conditions had persisted through the
course of the incubation. The Petri dish culture contents were then transferred into disposable 15
mL centrifuge tubes as above for MPN dilutions.
86
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Zooplankton Methods
Species Selection
We obtained and isolated representative species from the list of organisms provided by the ETV
Technical Panel and suggested in the "Request for Proposals" (Table 4). This list was developed
with the understanding that in some cases, species might not be obtainable and the final list
would substitute equivalent taxa more readily available in the Pacific Northwest. For example,
the European harpacticoid copepod species, Tisbe battagliai, was replaced with a common and
abundant Tisbe species of the Pacific Northwest, Tisbe cf.fiircata.
After collecting or purchasing the organisms, cultures were established using a standardized
protocol based on EPA recommendations and refined to accommodate the large number of
organisms being screened. Many of the suggested organisms were not easy to obtain or to
culture to the numbers required for bench-top testing, so the list was augmented with organisms
either easily obtained from the field in the Pacific Northwest region or ordered from a reputable
provider of live organisms (Table 5a,b). We verified species identifications of all purchased
species. In the course of the project, we screened 35 species distributed among 12 major taxa as
potential surrogate species: copepods (17 species); cladocerans (7); rotifers (4); amphipod (1);
barnacle (1); isopod (1); branchiopod (1); annelid (1); abalone (1); and insect (1).
Three criteria were used in the Phase I testing to reduce the list of surrogates for Phase II tests:
performance in the preliminary trials; ability to culture the organisms to high enough densities
for future full-scale testing in experimental ballast water tanks; and ability to determine the
efficacy of the treatments, i.e., the organisms' viability, using a microscope. Application of these
three criteria allowed us to narrow the list of potential surrogates to four marine and three
freshwater zooplankton species representing three marine classes (one branchiopod, two
copepods, and one annelid) and three freshwater classes (one cladoceran, one copepod, and one
ostracod) (Table 6). These seven species were used in the Phase II trials.
Table 4. Potential surrogate zooplankton species identified by the ETV program and
related studies.
Fresh Water
Marine Water
Daphniapulex (resting)
Daphnia magna
Brachionus calyciflorus (resting)
Culex (insect larvae)
Diaptomus pallidus (adult)
Ceriodaphnia lacustris (adult)
Acartia hudsonica (warm - resting)
Acartia tonsa (cold - resting)
Brachionus calyciflorus (adult)
Tisbe battagliai (adult)
Nitokra lacustris
Artemia
Mussel larvae
87
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Table 5a. Potential surrogate zooplankton species ordered from commercial sources or
provided by other researchers.
Organism and Type
Acquisition Procedure
Experimental Status
Harpacticoid copepods
Tisbe sp. cf.furcata (CA)
Nitokra lacustris
Essential Live Feeds, WA
Essential Live Feeds, WA
Phase I
Phase I & II
Calanoid copepods
Acartia tonsa
Algagen, LLC, in Vero Beach, FL
Phase I
Cyclopoid copepods
Acanthocyclops robustus
Macrocyclops albidus
Carolina Biological Supply, NC
Essential Live Feeds, WA
Phase I
Phase I & II
Cladocerans
Daphnia pulex
Daphnia magna
Ceriodaphnia dubia
Chydorus sphaericus *
Moina sp.
Carolina Biological Supply, NC
Frieda Taub's Lab, UW
Carolina Biological Supply, NC
Carolina Biological Supply, NC
Florida Aquafarms, FL
Phase I
Phase I
Phase I
Phase I & II
Rejected for experimentation; very
low densities in culture.
Mollusks
RedAbalone
Carolyn Friedman's Lab, UW
Phase I
Rotifers
Brachionus plicatilis
Brachionus calyciflorus
Philodina citrine
Lecane monostyla
Carolina Biological Supply, NC
Florida Aquafarms, FL
Carolina Biological Supply, NC
Carolina Biological Supply, NC
Phase I
Phase I
Rejected for experimentation
because it is not easy to filter intact
for observation.
Rejected for experimentation
because it is not easy to filter intact
for observation.
Branchiopods
Artemia salina
Brine Shrimp Direct
Phase I & II
Ostracods
Ostracod sp.*
Carolina Biological Supply, NC
Phase I & II
Annelids
Nereis virens
Sea Bait Ltd., Maine
Phase I & II
Mosquito larvae
Culex sp.
Florida Aquafarms, FL
Phase I
*While these organisms were started from cultures from Carolina Biological Supply, these were not the
organisms ordered. They were contaminants in with the D. pulex.
88
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Table 5b. Potential surrogate zooplankton species collected by the University of
Washington ballast water team.
Organism and Type
Acquisition Procedure
Experimental Status
Harpacticoid copepods
Tigriopus californicus
Tisbe sp. Washington
Harpacticus uniremis
Mesochra sp.
Deception Pass State Park, WA.
April, 2005
Field sampling at NOAA
Fisheries Manchester Research
Station, WA. April, 2005
NOAA Fisheries Manchester
Research Station, WA. April,
2005
Jakle's Lagoon, San Juan Island,
June, 2005
Phase I & II
Rejected for experimentation because
needs to be kept at 10 degrees C.
Rejected for experimentation due to
length of life cycle and univoltine
reproduction limit the culture of this
species throughout the year
Rejected for experimentation due to
low densities in culture, and because
three harpacticoid copepods were
already available for testing.
Calanoid copepods
Eurytemora affmis
Eurytemora americana
Calanus pacificus
Acartia hudsonica
Diaptomus nevadensis
Diaptomus sp.
Columbia River sampling. April,
2005
Jakle's Lagoon on San Juan
Island, June, 2005
NOAA Fisheries Manchester
Research Station, WA. April,
2005
Jakle's Lagoon on San Juan
Island, June, 2005
Soap Lake, WA. August, 2005.
Collected from Lake Washington,
October 2005
Rejected for experimentation due to
extremely low densities in culture.
Phase I
Rejected for experimentation because
it Could not be kept alive on the
variety of diets available
Rejected for experimentation due to
extremely low densities in culture.
Rejected for experimentation because
it Could not be kept alive on the
variety of diets available
Phase I
Cyclopoid copepods
Corycaeus anglicus
NOAA Fisheries Manchester
Research Station, WA. April,
2005
Rejected for experimentation because
it did not adapt easily to the laboratory
conditions
Cladoceran
Bosmina longirostris
Daphnia villosa
Lake Washington sampling,
October, 2005
Soap Lake, WA, August 2005
Phase I
Rejected for experimentation because
it could not be kept acclimated to full
strength salinity or freshwater.
Cirripedia
Cirripedia larvae
Columbia River sampling. April,
2005
Rejected because it is not easy to
culture nor is it available commercially
89
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Organism and Type
Acquisition Procedure
Experimental Status
Amphipods
Eogammarus
confervicolus
Duwamish River Estuary, WA.
April, 2005
Rejected for experimentation due to
extremely low densities in culture.
Isopods
Gnorimosphaeroma sp.
Duwamish River Estuary, WA.
April, 2005
Rejected for experimentation due to
extremely low densities in culture.
Table 6. Zooplankton species used in Phase II treatments.
Fresh Water
Macrocyclops albidus (copepod)
Chydorus sphaericalis (cladocerans)
ostracod sp.
Marine Water
Artemia salina (branchiopod)
Tigriopus californicus (copepod)
Nitokra lacustris (copepod)
Nereis virens (annelid)
Culture Conditions
Organisms were acclimated to standard laboratory conditions by placing them in an incubator
held at 20°C with a 12 hours light-12 hours dark photoperiod. Depending on the densities
achievable, organisms were kept either in 1 L of media in a half-gallon glass jar, or in a 150 mL
vented tissue culture flask. For regular maintenance, animals were screened and placed in fresh
culture media biweekly. (Twenty-four hours before experiments, animals were placed in fresh
media and allowed to evacuate their guts.) All cultures were kept in batch conditions, meaning
that no aeration was provided and all animals were manually filtered when the culture media was
changed.
Marine organisms were fed a combination of the live algae Tetraselmis suecica, freeze-dried
phytoplankton (Phytoplan™, Two Little Fishies, FL), and powdered fish food (Boyd's Vita-diet,
Boyd Enterprises, FL) on a weekly basis. Freshwater animals were provided with the live alga
Scenedesmus sp., freeze-dried phytoplankton (Phytoplan™, Two Little Fishies, FL), and
powdered fish food (Boyd's Vita-diet, Boyd Enterprises, FL) on the same schedule. The same
media was used for the experiments and the maintenance conditions, with the exception two
commercial products added to the freshwater maintenance media: Kent Marine R/O Right
powdered formula (a mixture of major salts of sodium, magnesium, calcium, and potassium;
Kent Marine, WI) and Kent Freshwater Essential liquid (a mixture of essential trace minerals to
reproduce natural freshwater; Kent Marine, WI). Organisms were held up to one year using
these techniques.
Experimental Conditions and Methods
The experiments for zooplankton differed from those for phytoplankton, where the dilution
media required the presence of nutrients to determine whether the organisms were affected by
the treatments and not starvation. An elevated level of nutrients was not required to keep the
90
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
animals alive in the zooplankton trials. The seawater comprised filter-sterilized artificial
seawater (Instant Ocean®) without nutrient supplementation. Salts were dissolved in glass fiber-
filtered reverse osmosis water to a salinity of 30.
For freshwater organisms, the test medium consisted of filter-sterilized (glass fiber-filtered)
reverse osmosis water with a testing medium based on the EPA's recommended aquatic
toxicology testing dilution protocols for freshwater organisms. Depending on the sensitivity of
the organisms, one of two EPA protocols was utilized to prepare the dilution water for testing
(see http://www.epa.gov/waterscience/WET/disk2/). In the toxicity-test methods, synthetic water is
referred to as diluted mineral water (DMW).
Whether seawater or fresh, the pH of test water was adjusted to match the culture conditions of
the organism to be tested. Water was used within 7 days of preparation.
Freshwater and marine experiments were conducted at room temperature in containers of
volumes appropriate to the size of the organisms being tested. Except for the thermal treatments,
which were terminated directly after 4 hours of exposure to the target temperature, and the
deoxygenation treatments, which were terminated after 48 hours in the degassing units, percent
survivals were determined after 24 hours in the dark. Biocides were treated as instantaneous
stressors, but the time of exposure to the treatment stressors for the ozone and UV treatments
varied based on desired exposure levels. Table 7 provides detailed information on the source,
concentration, and exposure time.
Because so many zooplankton species were resistant to the levels originally proposed for this
project, many of them were tested at higher levels or concentrations to provide a basis for
comparison with the performance of other zooplankton.
Phase I testing identified the seven most promising organisms among the different classes to be
tested (Table 7). Only the organisms listed in Table 6 were carried forward into Phase II testing.
91
-------
Protocol for the Verification of Ballast Water Treatment Technologies
Version 5.1
September 2010
Table 7. Zooplankton treatments (proposed and realized) and experimental conditions.
Treatment
stressor
Thermal
treatment
Chlorine (sodium
hypochlorite)
Chlorine dioxide
(Ecochlor™)
Glutaraldehyde
UV light
Ozone
Hydrogen
peroxide
Deoxygenation
SeaKleen®
PeraClean®
Ocean
Vendor or
source
Water bath
Chlorox bleach
Ecochlor, Inc.
Fisher Scientific
UV collimator
designed and
built by Dr. E.
"Chip"
Blatchley,
Purdue Univ.
Aquatic Eco-
Systems, UV-
type, 2.4 g with
air. Connected
to tubing and
an air stone.
Fisher Scientific
Sparge with N2
(95%) and CO2
(5%) mixture,
at levels to
reduce pH to
5.5, then seal
container.
Vitamar, Inc.
Degussa AG
Concentration
or Intensity
35°, 40°, 45°,
SOT
Aqueous
solution of
sodium hypo-
chlorite. Final
cones, of 0.25,
0.5, 1.0,2.0
mg/L
Final cones, of 1 ,
2, 4, 6 ppm
Final cones, of
50, 100, 500 and
1000 mg/L
UV light (256
nm) at 10, 25,
50, 100mJ/cm2
Total initial
residual oxidant
(TRO) level of
0.25,0.5,1.0,
and 2.0 mg Br2/L
Final cones. Of
0.5, 1, 10, and 20
ppm
Anoxia (0 mg/L
oxygen)
0.25,0.5, 1.0,2.0
mg/L active
ingredient
Final cones, of
50, 100, 200, and
400 ppm
Test volume
1 50 mL or 1 L
1 50 mL or 1 L
150mL
150mL
50 mL
0.5 L
150mL
50 mL
150mL
150mL
Actual Conditions and
Exposure Times
35°, 40°, 45° for 4 hours
Final cone, of 1, 2, 4, 8 mg/L for
24 hours
Final cone, of 1, 2, 4, 8 and 16
ppm for 24 hours
Final cone, of 50, 100, 200, 400,
800, 1200 and 1600 mg/L for 24
hours
Final exposures of 50, 100, 200,
and 400 mJ/cm2 - dose depends on
exposure time
Final exposure of 0.5 and 1 mg Br
L"1 for FW and 1,2, 4, and 8 mg Br
L"1 residual oxidant level for
Marine for 24 hours after
achieving initial level of TRO
Final cone, of 10, 20, 40, 80, 160,
and 320 ppm for 24 hours after
achieving initial concentration
48 hours after achieving anoxia
Final cone, of 1, 2, 4, 8, and 16
mg/L for 24 hours
Final cones, of 50, 100, 200, and
400 ppm
Note: Entries in bold indicate alterations from the originally proposed protocol (see Table 2 of this
report).
92
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Thermal Treatment (Phase I)
For zooplankton, only three temperatures were tested, 35°, 40°, and 45°C. Regardless of
organism and volume of culture container (150 mL of medium in 200 mL beakers or 1 L of
medium in half-gallon glass jars), the test container was filled either with artificial seawater or
freshwater medium and the test organisms were screened and added. After 4 hours of treatment
in the dark, the viability of organisms was immediately assessed. Controls were incubated
simultaneously at room temperature in the dark.
In the case of 1 L treatments, animals were recollected by passing the test medium twice through
a 20um Nitex filter screen. As this method proved to be inefficient in recollecting soft-bodied
organisms such as rotifers or animals which disintegrated due to the treatment, we switched to
150 mL treatments in 200 mL beakers, which enabled us to count organisms without having to
recollect them on a filter screen. Animals were placed in a counting wheel and examined under
at least lOx magnification to determine survival (indicated by motion) or mortality (lack of
movement after a few seconds of observation or prodding with a wire pointer).
Biocide Treatments (Phase I and II)
Biocide (sodium hypochlorite, glutaraldehyde, chlorine dioxide (Ecochlor™), SeaKleen®,
PeraClean® Ocean, hydrogen peroxide) experiments with zooplankton were conducted in
microcosms of non-reactive materials, incubated at natural ambient temperatures in the dark for
24 hours. At least 20 animals were placed in the microcosms and tested at various doses of the
treatments with a no-treatment control. Since variability was low in biocide treatments, three
replicates at each dosage level were adequate for comparison. Experimental controls consisted
of organisms held either in artificial seawater or freshwater medium for 24 hours in the dark. All
treatments were analyzed for survival and mortality after 24 hours.
When the originally proposed doses of biocides were not sufficient to kill more than 50% of
treatment organisms, dosages were increased until 50% mortality (or more) was obtained or until
the dose of biocide reached more than 100 times the maximum dose originally proposed. Details
about individual biocides follow.
Chlorine (sodium hypochlorite)
As described in "Common Methods". At the start of the experiment and after 24 h, the level of
total residual oxidant (TRO) was measured using a Hach Kit 2010.
Glutaraldehyde
As described in "Common Methods", except that organisms were exposed to a wider range of
glutaraldehyde concentrations: 50, 100, 200, 400, 800, 1200 and 1600 mg/L (Table 7).
93
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Chlorine dioxide (Ecochlor™)
As described in "Common Methods", except that organisms were exposed to a wider range of
chlorine dioxide concentrations: 1, 2, 4, 8, and 16 ppm (Table 7).
SeaKleen®
An organic biocide, SeaKleen® is available in several formulations and concentrations from the
manufacturer. We examined the sensitivity of prospective surrogate organisms to a wettable
powder that contains 85% active ingredient. Fresh stock solutions were prepared from the
powder before each series of trials. SeaKleen® was tested at 1,2, 4, 8, and 16 mg/L of the active
ingredient (Table 7).
PeraClean® Ocean
As described in "Common Methods".
Hydrogen peroxide
As described in "Common Methods", except that organisms were exposed to a wider range of
hydrogen peroxide concentrations: 10, 20, 40, 80, 160, and 320 ppm (Table 7).
Ultraviolet Light Treatment (Phase II)
Generally as described in "Common Methods", except that organisms were exposed to a greater
range of exposures: 50, 100, 200, and 400 mJ/cm2 (Table 7). A small polystyrene petri dish (60
x 15 mm, Falcon) containing the test organisms and less than 0.5 cm of medium was placed
beneath the collimator light beam. We determined there was no significant production of heat
generated by the lamp. Organisms were generally checked for mortality only once, 24 hours
after exposure. When no mortality was observed, we occasionally held onto samples for further
analysis at 48 and 72 hours. Four replicates were conducted for each combination of treatment
level and species.
Ozone Treatment (Phase II)
Generally as described in "Common Methods". To expose zooplankton to ozone, we placed
them in one half liter of media, placed the air stone in the media, and let the generator run for the
predetermined periods. In the case of the marine treatments, times were adjusted if exposure
level was less than anticipated. We were not able to do this for the freshwater experiments. A
lid through which the ozone delivery hose was fitted was placed on freshwater containers to
ensure maximum contact between the organisms and ozone for each 24 hour experiment. After
treatment, we replaced the hose-fitted lid with a sealed lid. All treatment jars were placed in the
dark for 24 hours before analyzing organism survival.
94
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Deoxygenation Treatment (Phase II)
Generally as described in "Common Methods", except that only a 48 hours treatment was run.
At least 20 organisms from each taxon being tested were placed in a small polystyrene petri dish
(60 x 15 mm, Falcon), and the dish itself was placed inside the BD BBL GasPak® Jar System.
95
-------
Protocol for the Verification of Ballast Water Treatment Technologies September 2010
Version 5.1
Appendix C: Design and Preliminary Use of a Commercial Filter
Skid to Capture Organisms > 50 jim in Minimum Dimension
(Nominally Zooplankton) for Evaluating Ships' Ballast Water
Management Systems at Land-based Test Facilities
96
-------
REPORT DOCUMENTATION PAGE
Form Approved
OMB No. 0704-0188
The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources,
gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of
information, including suggestions for reducing the burden, to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188),
1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any
penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number.
PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY)
22-06-2010
2. REPORT TYPE
NRL Letter Report
3. DATES COVERED (From - To)
March2009-April2010
4. TITLE AND SUBTITLE
Design and Preliminary Use of a Commercial Filter Skid to Capture
Organisms > 50 pjn in Minimum Dimension (Nominally Zooplankton) for
Evaluating Ships' Ballast Water Management Systems at Land-Based Test
Facilities
5a. CONTRACT NUMBER
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S)
Edward J. Lemieux, Code 6130, NRL, DC, Lisa A. Drake, SAIC, Key West,
FL, Stephanie H. Robbins, SAIC, Key West, FL, Mia K. Steinberg, Code
6130, NRL, DC, Scott C. Riley, SAIC, Key West, FL, Elizabeth C. Schrack,
University of Virginia, Charlottesville, VA, Wayne B. Hyland, Azimuth
Technical Consultants, Key West, FL, Jonathan F. Grant, Battenkill
Technologies, Inc., Manchester Center, VT, Cameron S. Moser, EXCET, Inc.,
Key West, FL, Tim P. Wier, EXCET, Inc., Seattle, WA
5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
US Naval Research Laboratory
4555 Overlook Ave., SW
Washington, DC 20375
8. PERFORMING ORGANIZATION
REPORT NUMBER
6130/1029
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)
United States Coast Guard
Environmental Standards Division (CG-5224)
2100 2nd Street, S.W.
Washington, DC 20593
10. SPONSOR/MONITOR'S ACRONYM(S)
USCG
11. SPONSOR/MONITOR'S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Distribution Statement A: Approved for public release, distribution is unlimited.
13. SUPPLEMENTARY NOTES
14. ABSTRACT
International and national standards dictating the number of living organisms discharged in ships' ballast water are in the process of
being enacted. Developing a method to accurately determine whether how many organisms are present in treated water (e.g., with
sparse concentrations of organisms) with statistical confidence requires large volume (e.g., > 10 mA3) are concentrated and
examined. Here, we describe efforts to (1) develop and validate a procedure for retaining organisms > 50 pjn in minimum
dimension (nominally zooplankton) in large volumes of water used for testing ballast water management systems, (2) design and
build the equipment to do so, and (3) validate the equipment. Importantly, zooplankton needed to be kept alive and minimally
affected by filtering and handling, since ballast water discharge standards prescribe the number of viable (living) organisms.
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF:
a. REPORT
u
b. ABSTRACT
U
c. THIS PAGE
U
17. LIMITATION OF
ABSTRACT
Unclassified
18. NUMBER
OF
PAGES
46
19a. NAME OF RESPONSIBLE PERSON
Edward Lemieux
19b. TELEPHONE NUMBER (Include area code)
202-404-2123
Standard Form 298 (Rev. 8/98)
Prescribed by ANSI Std. Z39.18
-------
DEPARTMENT OF THE NAVY
NAVAL RESEARCH UOCfWOKf
45S5 CV6HLOQIC AV6 SW
WASHINGTON DC 2ft375-S32fl
IN HiPiV REFER TO
j90;)
3e r 61 30 / *. 029
27, Jun» 2010
-'it,
! - i> >r
he NRl, points of cor.tacc are LISJ DraKc, Cocso 5136, 305-293-4215,
: 11 SA . clrak& . ct i'i?ni 1. na vy , mi i ana Edward Lo-ntit'-.x, Code 6130, 240-4
e -'mail : e c»wa r d . i em i e u x 9 n r 1 . r. a v y . m i 1..
WARREN W. Sv.ji.JI-i »'
Py 'Jj rect ion
-------
6130/1029
21 JUN2010
Naval Research Laboratory
4555 Overlook Ave, S.W.
Washington, DC 20375-5320
Center for Corrosion
Design and Preliminary Use of a Commercial Filter Skid to
Capture Organisms > 50 um in Minimum Dimension
(Nominally Zooplankton) for Evaluating Ships' Ballast
Water Management Systems at Land-Based Test Facilities
Edward J. Lemieux, Code 6130, NRL, DC
Lisa A. Drake, SAIC, Key West, FL
Stephanie H. Robbins, SAIC, Key West, FL
Mia K. Steinberg, Code 6130, NRL, DC
Scott C. Riley, SAIC, Key West, FL
Elizabeth C. Schrack, University of Virginia, Charlottesville, VA
Wayne B. Hyland, Azimuth Technical Consultants, Key West, FL
Jonathan F. Grant, Battenkill Technologies, Inc., Manchester Center, VT
Cameron S. Moser, EXCET, Inc., Key West, FL
Tim P. Wier, EXCET, Inc., Seattle, WA
DISTRIBUTION STATEMENT A. Approved for public release, distribution is unlimited.
Encl (1) to NRL Ltr
3900
6130/1029
-------
Table of Contents
1 Introduction 1
2 Objectives 2
3 Experimental Approach 2
3.1 Ballast Water Treatment Test Facility (BWTTF) Description 2
3.2 Microbeads 3
3.2.1 Microbead Disintegration 4
3.2.2 Method of Microbead Counting 4
3.2.3 Intercalibration 5
3.3 Filter Bags 6
3.3.1 Filter Bags Field Trials—Water Flow Rate 7
3.3.2 Filter Bags—Laboratory and Field Validation with Microbeads 7
3.3.3 Filter Bags—Seams 8
3.3.4 Filter Bags—Toxicity of Sealant on the Seams 8
3.4 Filter Skid 9
3.4.1 Filter Skid Requirements 9
3.4.2 Toxic Effects of Stainless Steel Filter Housings 9
3.4.3 Comparison of Zooplankton in the Ballast Tank vs. the Filter Skid 10
3.4.4 Effect of Crowding on Ambient Zooplankton 11
4 Results 11
4.1 Microbeads 11
4.1.1 Microbead Disintegration 11
4.1.2 Method of Microbead Counting 11
4.1.3 Intercalibration 12
4.2 Filter Bags 12
4.2.1 Filter Bag Field Trials—Water Flow Rate 12
4.2.2 Filter Bags—Field and Laboratory Validation Using Microbeads 13
4.2.3 Filter Bags—Seams 15
4.2.4 Filter Bags—Toxicity of Sealant on the Seams 15
4.3 Filter Skid 16
4.3.1 Filter Skid Design 16
4.3.2 Filtration Area of Filter Skid vs. PlanktonNet 22
4.3.3 Flow Velocity through the Filter Skid vs. PlanktonNet 22
4.3.4 Toxic Effects of Stainless Steel Filter Housings 24
4.3.5 Comparison of Zooplankton in the Ballast Tank vs. the Filter Skid 25
4.3.6 Effect of Crowding on Ambient Zooplankton 25
5 Discussion 25
6 Acknowledgements 26
7 References 26
Appendix 1 28
Appendix 2 29
Appendix 3 30
Appendix 4 31
Appendix 5 32
Appendix 6 38
Appendix 7 40
-------
Appendix 8 41
Appendix 9 41
Table of Figures
Figure 1. Two 41 cm long filter bags in Hayward housing arranged in series; water entered the
system via the green hose at the top of the left-hand housing 6
Figure 2. Recovery after microbeads were counted using the =dry' and =wet' methods. Orange
bars represent dry trials, and blue bars represent wet trials; numbers above bars indicate the
percent microbead recovery; letters on the x-axis labels indicate the researcher's initials 12
Figure 3. Recovery efficiency of microbeads in laboratory trials using 25 um filter bags (FB).
Blue, red, and green bars represent 50 um, 100 um, and 150 um microbeads, respectively 13
Figure 4. Recovery efficiency of microbeads in field trials. Unless noted, filter bags' mesh was
25 um. Blue, red, and green bars represent 50 um, 100 um, and 150 um microbeads,
respectively. * = a notable amount of green algae was collected in filter bags 14
Figure 5. Recovery efficiency of 50 um microbeads in laboratory and field trials conducted after
improvements were made to the microbead protocol. Numbers in parentheses represent
replicates; FB = filter bag 15
Figure 6. Filter skid design for flow-through sampling of discharged ballast water 16
Figure 7. Eaton Topline™ filter housing 17
Figure 8. Nylon mono filament mesh filter bag secured in an Eaton Topline™ filter housing. ..18
Figure 9. Plan view of the prototype filter skid at NRLKW. Blue arrow shows water entering
the filter skid; yellow arrow shows water exiting the skid 18
Figure 10. YamadaNDP-80 Air diaphragm pump as installed 20
Figure 11. Pressure drop of filter skid vs. flow rate (figure from Eaton Filtration, LLC) 21
Figure 12. Flow velocity model for water flow through an Eaton Topline™ Filter Housing at 50
gpm, entering though the top of the housing and exiting at the bottom 23
Figure 13. Flow velocity model illustration for water flow through an Eaton Topline™ Filter
Housing at 25 gpm, entering though the top of the housing and exiting at the bottom 24
in
-------
1 Introduction
In 2004, the International Maritime Organization adopted the International Convention for the
Control and Management of Ships' Ballast Water and Sediments, which establishes standards for
ballast water discharge and the performance of ballast water management systems (IMO, 2004).
Among the criteria for ballast water discharge is the density of organisms > 50 um (nominally
zooplankton), which is set at < 10 viable organisms m"3. This numerical standard, intended to
j>ievent, minimize and ultimately eliminate the risks to the environment, human health, property
and resources arising from the transfer of harmful aquatic organisms and pathogens', is also
proposed by the US Coast Guard (2009).
Because the probability of finding a living zooplankter in a water sample meeting the proposed
discharge standard follows a Poisson distribution (Lemieux et al., 2008b), the key test statistic is
the number of organisms counted. With respect to land-based and shipboard testing of ballast
water management systems, accurate enumeration of organisms in ballast tanks depends on
several factors: the volume of water sampled, the volume of sample analyzed (e.g., the entire
sample or a sub-sample), the level of desired precision, and the relevant discharge standard (e.g.,
< 10 viable organisms m"3 or < 0.1 m"3). Previous work at the Naval Research Laboratory in Key
West (NRLKW) has shown for a 3 m3 zooplankton sample treated by a ballast water
management system and concentrated to a volume of 1 1, only a 20 ml subsample of the
concentrate could be evaluated before zooplankton in the subsample died because they were held
in artificial conditions in the laboratory (i.e., concentrated and gently aerated in a flask; Lemieux
et al., 2008a). If only 20 ml of a concentrated sample can be evaluated, the required sample
volume is large. For example, to know with 95% confidence a ballast tank contains < 10 living
zooplankters m"3, a sample volume of 60 m3 would need to be concentrated to 1 1 and a 20 ml
subsample examined (Lemieux et al., in review). Several factors affect the statistics; if a larger
volume of subsample (or the entire concentrated sample) could be analyzed, the sample volume
concentrated to 1 1 would decrease accordingly. Nonetheless, given the work at NRLKW and
other test facilities, it appears large sample volumes—at least 10m -will need to be concentrated
and evaluated.
The large volumes result in two challenges: adverse effects on organisms filtered through a fine
mesh must be minimized, and the size of the individual nets or filters must be manageable (the
latter is especially relevant for shipboard testing). Traditionally, marine scientists capture
organisms by filtering water through a plankton net. Alternatively, a set of filter bags enclosed
in individual housings—a configuration commonly used in water-treatment facilities—can be
employed. An argument can be made that filter bags are superior to a plankton net for ballast
water testing because filter assemblies do not require a mechanism to lift them from a tank to
collect samples. Additionally, the filter bags are enclosed in a housing, so they are less
vulnerable to inclement weather than plankton nets.
-------
2 Objectives
The objectives of this work were to (1) develop and validate a procedure for retaining organisms
> 50 um in minimum dimension (nominally zooplankton) in large volumes of water used for
testing ballast water management systems, (2) design and build the equipment to do so, and (3)
validate the equipment. Importantly, zooplankton needed to be kept alive and minimally
affected by filtering and handling, since the ballast water discharge standard prescribes the
number of'viable (living) organisms (IMO, 2005).
3 Experimental Approach
All laboratory and field experiments were conducted at the Ballast Water Treatment Test Facility
(BWTTF) at NRLKW. An initial field experiment compared the recovery of zooplankton
proxies (red, 50 um diameter microbeads) between a plankton net and two filter bags enclosed in
individual filter housings. After microbeads were added to the net or filter bag, ambient
seawater was pumped through them, and the filtrand of each was examined for the presence of
microbeads (see Appendix 1 for experimental details). Although NRLKW is surrounded by
oligotrophic seawater with relatively low levels of plankton, debris, or colored dissolved organic
matter, the samples were clouded with material so it was difficult to see the microbeads.
Recovery efficiency of the microbeads was very low (Appendix 1; 25 and 28% from the
plankton net and filter bags, respectively).
To address this outcome, trials using microbeads were conducted in the laboratory using filter
bags or sieves. The initial laboratory trials also yielded low microbead recovery efficiencies, so
a number of improvements was made to the experimental approach (e.g., as described in sections
3.2 and 3.3.3 below). Next, field trials were conducted using fresh water and filter bags.
Considering the initial field trial with a plankton net—and taking into account previous
experiences at NRLKW—it was concluded using plankton nets to process the large volumes of
water required to evaluate treated water at land-based testing facilities would not be feasible.
Nets are cumbersome, prone to tearing, open to the atmosphere and thus vulnerable to wind and
weather, and difficult to configure in a flow-through sampling apparatus suitable for large
volumes. In response, engineers at NRLKW designed and built a filter skid containing multiple
filter bags, each enclosed in a stainless steel housing. Here, we describe preliminary trials to
evaluate the retention efficiency of filter bags using microbeads, the design and construction of
the filter skid, and initial validation trials.
3.1 Ballast Water Treatment Test Facility (BWTTF) Description
As part of the United States Environmental Protection Agency's Environmental Technology
Verification (ETV) program and in partnership with the U.S. Coast Guard, a BWTTF was
established at NRLKW, where research is conducted to provide technical guidance for the ETV's
Generic Protocol for the Verification of Ballast Water Treatment Technologies (Lemieux et al.,
-------
in review). The BWTTF is located on Fleming Key at Trumbo Point Annex, Naval Air Station
Key West, FL. As part of the Naval Research Laboratory's Center for Corrosion Science and
Engineering, it functions as a scientific test platform for the standardized assessment of
technologies designed to reduce or eliminate aquatic nuisance species in shipboard ballast water.
If land-based test facilities incorporate a standard set of challenge conditions, then comparable
and defensible results are generated. To that end, the BWTTF is designed to conduct
experiments and develop methods to inform the ETV verification protocol and thereby has a
fully instrumented seawater storage and transfer system that replicates the volumes, flows, and
pressures typical of ballast water systems on marine vessels. The BWTTF pumps can provide
flow rates up to 5110 1 min"1 (1350 gpm), and on-site tanks include a 382 m3 (101,000 gal)
ballast tank, a 151 m3 (40,000 gal) ballast test tank, and a 394 m3 (104,000 gal) discharge tank.
The facility provides for the injection of specified test organisms as well as the means to monitor
test conditions, conduct sampling, and analyze samples in the laboratory. All water from
experiments is treated prior to final discharge to remove all added challenge components and any
by-products of any treatment systems tested. To accommodate yet undefined and unidentified
technologies, the system can be reconfigured to accommodate treatment systems at uplift, in-
tank, or discharge locations in the flow path.
Control of all systems and instrumentation is provided by the Honeywell Plantscape Industrial
Control and Automation system (Honeywell, Morris Township, NJ), which has been customized
by engineers at NRLKW to provide BWTTF-specific data acquisition, operational control, and
system monitoring with alarm and interlock functionality. The industrial control hardware
consists of standalone control panels with programmable logic controllers, relays, and analog-to-
digital converters connected to a redundant computer server system to log all actions and data. A
computer in a control room adjacent to the ballast tanks with five displays provides the primary
operator interface to control of the BWTTF. This control system provides a series of graphical
control screens from which the operator can select and view the overall system with key
parameters displayed.
3.2 Microbeads
The proxies used for zooplankton were red, ChromoSphere-T NIST Certified microspheres
(Microgenics Corporation, Fremont, CA). Depending on the trial, one or two sizes of
microbeads were used: with a nominal diameter of 49 um (CV = 7.8%), 96 um (CV = 7.8%) or
150-um ±3.6 um (CV = 6%). All microbeads were composed of a cross-linked polystyrene
divinylbenzene copolymer. Microbeads were counted in 15 ml Bogorov counting chambers or
1-ml Sedgewick Rafter counting chambers using an Olympus SZH10 dissecting microscope, a
Nikon AZ100 Multizoom microscope, or a Nikon Eclipse TS100 inverted microscope.
Usually, microbeads were counted the day before an experiment and placed in a beaker with
Type II water (5-um filtered and treated by reverse osmosis and deionization). The contents of
the beaker were poured into the plankton net, filter bag, or sieve with care to minimize splashing
and loss of microbeads. Initial trials used 600 microbeads to approximate a discharge volume of
60 m meeting the IMO zooplankton discharge standard (10 zooplankton m" x 60 m = 600
-------
microbeads); subsequent trials typically used 100 or 200 microbeads (which allowed more trials
to be run, as less time was spent counting microbeads). In all cases, the beaker containing the
microbeads was rinsed 3-5 times with Type II water, and the beaker and wash water were
examined for microbeads that may have adhered to the beaker wall.
At the end of laboratory trials or field runs, the sieve or filter bags were rinsed with Type II water
or artificial seawater (Instant Ocean®; Aquarium Systems, Inc., Mentor OH) into beakers.
Subsamples of approximately 15 ml were pipetted into Bogorov counting chambers. Two
researchers counted subsamples until the entire volume of each beaker was counted.
Additionally, surfaces that touched the sample were examined for microbeads: beakers, pipet
tips, and funnels.
For each Bogorov chamber analyzed, counts were made by a researcher until two subsequent
counts agreed. In most experiments, two researchers examined at least one of the Bogorov
Chambers to ensure counts were the same. If the final count differed between the researchers,
which was rare (approximately 20% of the time), it differed by one or two microbeads. Because
it was assumed microbeads were much more likely to be undercounted than over counted, the
highest number was used in the final tally.
3.2.1 Microbead Disintegration
When the initial recovery rates were low in field and laboratory trials, it seemed unlikely, but
possible, the microbeads dissolved in seawater or Type II water. An assay was conducted: in
separate wells of a 12-well plate, 10 50-um microbeads and 10 150 um microbeads
were dispensed into 5 ml of filtered seawater (0.22 um), and ten microbeads of each size class
were dispensed into 5 ml of Type II water. The four wells were covered and placed on a
laboratory bench for 7 days, and then microbeads in each well were counted.
3.2.2 Method of Microbead Counting
The microbead recovery efficiency was low in initial trials using 50 um diameter microbeads in
field and laboratory experiments, which was surprising, as the microbeads had been accounted
for in all possible places: the filtrate, filtrant (material retained in the filter bag), residue in the
beaker that initially held the microbeads, and residue in the pipette used to dispense samples.
Given this unbalance, it seemed possible the method for initially counting microbeads to add to
filter bags, sieves, or plankton nets was somehow undercounting them. Until this point, the
microbeads had been counted 'dry', that is, a sterile pipette tip was dipped into the bottle
containing the microbeads to deliver a pile of microbeads to a gridded Sedgewick Rafter (SR)
counting chamber, which did not have water on it. After the microbeads were counted, the SR
was gently rinsed with Type II water from a squirt bottle into a beaker, and the beaker's contents
were poured into a filter bag or sieve for an experiment. To ensure the SR did not have any
remaining microbeads, it was placed on a Kim wipe and examined under the
microscope. Perhaps rinsing the dry microbeads into a beaker caused them to be dispersed into
the air rather than the beaker.
-------
Addressing this hypothesis, recovery efficiency of 100 50-um microbeads was determined in 5
trials with =dry' counting and 4 trials with =wet' counting. In the latter method, a pipette tip was
dipped into the bottle containing the microbeads and tapped into a drop of Type II water on a SR
slide to remove the microbeads from the pipette tip. The water did not touch the edge of the
slide, nor was a cover slip used. After the microbeads were dispensed (dry or wet) into the SR,
they were counted and rinsed from the SR with a squirt bottle containing Type II water into a
beaker. To determine that 100 microbeads were, in fact, delivered to the beaker,
its contents were pipetted into Bogorov chambers and counted. When the entire sample was
counted, the pipette and beaker were examined for any residual microbeads. For each trial, the
name of the scientist counting the microbeads was recorded to determine if the poor recovery
was attributable to a given researcher.
3.2.3 Intercalibration
To ensure all microbeads were counted, an intercalibration exercise was conducted in which two
observers counted the same 15 ml sample dispensed in a Bogorov counting tray. Four trays
(each with a different sample from the filter bag) were counted by each observer, and then each
tray was recounted by each observer. An additional 10 trays (each with a different sample from
the filter bag) were counted by both observers, for a total of 18. Although the seawater used for
the experiment had been filtered, debris (sediment and plankton) was present in the samples.
Thus, extra care was required to find the microbeads, and each tray took approximately 45 min to
examine.
To determine the difference between observers the mean of the observers' counts for a single
tray was determined, and the percent difference for each observer's count was calculated as
follows:
Equation 1. Calculation to determine percent difference between microbead counts.
Mean of Counts - Observer 1 Count]
Mean of Counts x 100
Because there were two observers, the percent difference was the same for each observer for
each tray counted.
In a separate trial, one observer dispensed 15 ml of sample from the first filter bag and counted it
five times. Each time, she shook the tray before counting to determine if the movement of debris
and microbeads resulted in a different microbead count.
-------
3.3 Filter Bags
In laboratory trials using microbeads, nylon filter bags (18 cm [7"] in diameter at the top and 41
cm [16"] long) with various mesh sizes (100 um, 50 um, 25 um, or 10 um) were used singly
(Universal Filters, Inc., Asbury Park, NJ). In some cases, sieves (20 cm [8"] in diameter) with
25 um and 10 um nylon mesh were used in lieu of filter bags to test recovery in a very
straightforward way (i.e., from a fiat surface free of seams and dimples found in filter bags).
In field trials, filter bags 18 cm (7") in diameter and 41 cm (16") long were placed in filter
housings arranged in series. A variety of mesh sizes was used: 25 um, 35 um, 50 um, 100 um.
Each filter bag was enclosed in a fiberglass filter housing (Hayward Flow Control Systems,
Clemmons, NC) containing a polypropylene filter basket (Figure 1). Flow and pressure were
monitored using a paddle wheel flow sensor downstream of the second housing and with three
pressure sensors (one before and after the first housing and one after the second housing; all
sensors were manufactured by GF Signet, El Monte, CA). Flow was controlled manually using a
5 cm (2") diaphragm valve at the bottom of the right-hand filter housing (Figure 1, black handle).
Ambient seawater was taken up by a 30-hp, horizontal centrifugal pump, passed through a 15 cm
(6"), polyvinyl chloride (PVC) line, passed through a manifold, and delivered to the Hayward
units by a 5 cm (2"), reinforced PVC plastic hose. Freshwater was added to a 3-m3 storage tank
using a hose and pumped to the Hayward units through a 5 cm (2"), reinforced PVC plastic hose.
Figure 1. Two 41 cm long filter bags in Hayward housing arranged in series; water entered the
system via the green hose at the top of the left-hand housing.
Early field trials employed filter bags with 25-um mesh. Later trials used filter bags with 35 um
mesh, because it better approximated the mesh size used by researchers at test facilities in the US
and abroad (37 um) and was available in commercially manufactured filter bags. To capture
organisms in the > 50 um size class, the IMO Convention (2004) states the hypotenuse of the
-------
mesh must be no longer than 50 um; that requirement is met using 35 um mesh (hypotenuse =
49.5 um).
When microbeads were used as proxies for zooplankton, immediately prior to the trial, they were
rinsed into a filter bag placed in the second filter housing in series (downstream of a housing
with a filter bag used as a prefilter). The cover of the second filter housing was removed, and the
housing had been filled with freshwater. After the cover was secured on the filter housing and
the air bled from the housing using a valve on its cover, water was pumped through the system.
The filter bags were retrieved and the microbeads counted from the filter bag to which the
microbeads had been added and from the filter bag downstream of it. Often, a filter bag was
used as a prefilter upstream of the filter bag containing the microbeads to remove plankton,
sediment, and detritus.
3.3.1 Filter Bags Field Trials—Water Flow Rate
To determine the appropriate mesh size for filter bags arranged in series and the appropriate
water flow rate through them, five trials were run. The flow rates varied (25 gpm, 125 gpm, 150
gpm), as did the filter bags' mesh sizes (25 um, 50 um, 100 um). The differential pressure
between filter housings was monitored, and whether the filter bags clogged was noted.
3.3.2 Filter Bags—Laboratory and Field Validation with Microbeads
In laboratory trials, microbeads were counted using a Sedgewick Rafter counting chamber in the
wet or dry manner, rinsed into a beaker with Type II water, and the beaker's contents gently
rinsed into a filter bag or onto a sieve. The material retained on the filter bag or sieve was rinsed
with Type II water into a beaker, and its contents were transferred with a serological pipet into
Bogorov counting chambers. The recovery efficiency of 50 um microbeads was low in the first
laboratory (and field) trials, so 150 um and 100 um microbeads were added to experiments
because their large size rendered them easier to find, thus they served as quasi-positive controls.
The following parameters were recorded to account for any biases: type of filter bag (plastic-
topped vs. felt-topped), the glue used to seal the seams, how the seams were glued (e.g., once on
the inside and once on the outside), the method of counting, and the filter bag number (to record
the number of uses).
After initial field trials with ambient seawater yielded low microbead recovery efficiencies,
freshwater was used exclusively in the field. One 25-um filter bag was placed in each of three
Hayward filter housings arranged in series (with the pressure and flow sensors described above).
A filter bag served as a prefilter to remove debris from the water; a known number of
microbeads in a beaker was rinsed with Type II water into the second filter bag in series. After
approximately 3785 liters (1000 gal) was pumped through the system at a flow rate of 25 gpm,
the second and third filter bags were removed, rinsed with Type II water, and the material
retained in the bags was rinsed into a beaker, pipette into Bogorov counting chambers, and
examined for microbeads. The filter bags' seams used in field trials were sealed with 3M®
5200: 25 um filter bags were sealed twice on the inside. To reduce handling time, the 35 um
7
-------
filter bags were sealed once on the inside with twice as much sealant typically used for the 25
um filter bags; a laboratory trial showed 100% recovery of 50 um microbeads.
In laboratory and field trials, the filter bags' seams, the beakers, and pipettes were examined for
residual microbeads.
3.3.3 Filter Bags—Seams
After initial experiments showed a low recovery efficiency of microbeads, it was discovered that
the holes at the seams were much greater than the nominal mesh size, and it was hypothesized
the microbeads passed through the filter bags via the holes. A survey of filter bag manufacturers
showed bags with heat-welded seams were available, which would ameliorate the problem, but
bags were available only in microfilament or polypropylene felt material. Neither was
appropriate, as organisms could not be retrieved from these surfaces.
Closing off the bags' seams with a marine sealant was a viable solution, as long as the cured
sealant was malleable, not tacky (so it would not trap zooplankton), and non-toxic to marine
organisms over the short time they would be sequestered in the filter bags during sampling
(approximately 2 h). Seven sealants were evaluated: Marine Goop® (Eclectic Products, Inc.,
Eugene, OR), INSTANT Adhesive (GC™ Electronics, Rockford, IL), Weld-On PVC 717™
Plastic Pipe Cement (IPS Corporation, Gardena, CA), 2-part epoxy (John C. Dolph Company,
Monmouth Junction, NJ), white 3M™ Marine Fast Cure 5200 Adhesive Sealant (3M, St. Paul,
MN), and Elmer's® No-Wrinke Rubber Cement (Elmer's Products, Inc., Columbus, OH).
Using a-41 cm (16") long filter bag with 10 um mesh, each adhesive was applied in
approximately 5-cm long sections the outside and inside of the bags, and the adhesives cured for
36 h (longer than any of the recommended curing times).
3.3.4 Filter Bags—Toxicity of Sealant on the Seams
To test if zooplankton were killed by brief exposure to the filter bags' sealant (3M™ 5200,
chosen after the trial described in section 3.3.3), bioassay experiments were conducted with brine
shrimp Artemia franciscana (previously used at NRLKW as a standard test organism) and
subsequently with ambient zooplankton (copepods). Cysts of A. franciscana were purchased
from a vendor (Brine Shrimp Direct, Ogden, UT) and incubated in 5-um-filtered, aerated
artificial seawater (salinity = 36) for 24 h in 25°C with a 12:12 lightdark cycle under fluorescent
bulbs (72 uM Einsteins m"2 s"1). Hatched nauplii approximately 12 h old and 300 - 400 um long
were removed from the culture in 1 ml aliquots and dispensed into each of eight Petri dishes (47
mm diameter) with 10 ml of 0.22 urn-filtered seawater. Next, the dishes were examined using a
dissecting microscope to ensure all A. franciscana were living (determined by movement), which
was the case. Four Petri dishes served as controls, and to the remaining four, a strip
(approximately 4 cm x 1 cm) of 25 um, nylon filter bag with a bead of sealant (approximately 3
cm x 0.5 cm) that had been cured for 24 h was added.
-------
Petri dishes were arranged haphazardly on a shelf in the incubator (described above) for 2 h, the
estimated amount of time zooplankton would be sequestered in a filter bag during collection.
After the incubation, the number of dead A. franciscana was counted. If an organism was not
moving, it was gently prodded with a probe, and if it did not move within 10 s, it was scored as
dead. Following the tally, all samples were fixed with Lugol's iodine solution, the total number
of A. franciscana counted, and the number of living organisms determined by subtraction.
The trial was repeated with ambient copepods, which were concentrated from a seawater hose
onto a 31-um sieve. Copepods of all-life history stages (> 50 um in minimum dimension) were
removed from the sample individually using a pipet. To each of eight Petri dishes (47 mm
diameter) containing 10 ml of 0.22 um filtered seawater, 20 copepods were added. Four dishes
served as controls, and strips of mesh with sealant were added to the other four Petri dishes
(treatment). Dishes were examined to ensure all copepods were living (they were), and dishes
were arranged haphazardly on a shelf in the incubator and incubated for 2 h at the settings from
the previous trial. Following incubation, the number of dead copepods was counted and
preserved as above. Statistics were calculated with SigmaPlot® v. 11 (Systat Software Inc., San
Jose, CA).
3.4 Filter Skid
3.4.1 Filter Skid Requirements
The filter skid was designed with several considerations in mind: first, it was desirable to
construct an apparatus using common, commercially available products so the skid could be
replicated easily at other test facilities and did not require special fabrication techniques.
Second, the prototype was developed to eliminate large, cumbersome plankton nets currently
used by test facilities, so it needed to have nets in self-contained filter housings. Third, the
apparatus was designed so it might be accommodated in future shipboard sampling programs,
e.g., a small footprint was required. Because the skid would be used with seawater, the materials
needed to withstand a corrosive environment (e.g., stainless steel or PVC). Both the cost and
availability of corrosion-resistant metals were considered. Grade 316 stainless steel was chosen
for the filter housings, and PVC was chosen for the piping.
3.4.2 Toxic Effects of Stainless Steel Filter Housings
A bioassay test was conducted to determine if the brine shrimp Artemia franciscana or ambient
zooplankton would die from exposure to stainless steel filter housing. Cysts of A. franciscana
were hatched as in section 3.3.4, and 3 1-ml subsamples of 12-h old Nauplii (300 - 400 um long)
were removed from the culture and evaluated qualitatively using a dissecting microscope. If a
nauplius was not moving, it was gently prodded with a probe, and if it did not move within 10s,
it was scored as dead. All nauplii (70 - 77 per subsample) were living. Two ml aliquots of the
culture were dispensed into each of five 1000 ml beakers containing 800 ml of 0.45 um-filtered
seawater.
-------
Ambient zooplankton (primarily copepods) were concentrated from a seawater hose onto a 37-
um sieve. The filtrand was suspended in 0.45 urn-filtered seawater. A 1 ml sample was
removed and zooplankton (> 50 urn in minimum dimension) were qualitatively assessed: 1 was
dead, 6-10 were living. Four-mi aliquots of the sample were placed into each of the five beakers.
Two beakers served as controls, with only seawater, A. franciscana, and ambient zooplankton.
To the remaining three beakers, 10 washers (grade 316 stainless steel, 1.9 mm thick, and 38.1
mm in diameter with a hole in the center 15.8 mm in diameter) were added. Their surface area
(SA) in the beaker was calculated to equal the S A: volume ratio of water in a filter skid housing.
Beakers were arranged haphazardly on a shelf and covered with a black tray in the incubator for
2.5 h, the estimated amount of time zooplankton would be sequestered in a stainless steel filter
housing during collection. After the incubation, the washers were removed from the treatment
beakers using forceps, and the contents of each beaker were gently poured through a 37 urn
sieve, the filtrand rinsed with filtered seawater into a beaker, and the entire volume transferred to
a Petri dish. The number of dead A. franciscana and zooplankton were counted. Following the
tally, all samples were fixed with Lugol's iodine solution, the total number of organisms
counted, and the number of living organisms determined by subtraction.
3.4.3 Comparison of Zooplankton in the Ballast Tank vs. the Filter Skid
In a preliminary experiment to determine if collecting zooplankton (copepods) using the filter
skid killed them, the percentage of living, ambient copepods was quantified after they were
pumped into a ballast tank at NRLKW, and that number was compared to the percentage of
living zooplankton collected in the filter bags in the skid as the tank was drained. The ballast
tank was filled with 219m3 (57,775 gal) of ambient seawater at a flow rate of 3834 1 min"1 (1013
gpm), and immediately afterwards, a vertical plankton tow was taken in the tank using a 25 urn
mesh, 0.75 m mouth diameter, nylon plankton net. Originally, a quantitative sample was to be
collected, but when the Niskin bottle used to collect the sample broke, a qualitative plankton tow
was taken instead. The net was rinsed with ambient seawater and its contents suspended in
filtered seawater (0.22 urn) and diluted lOx with filtered seawater. After the container was
sealed and gently inverted 3x, 2 subsamples were removed and evaluated in Bogorov counting
chambers. The number of dead copepods (nauplii; copepodites and adults) was counted, the
samples fixed in Lugol's iodine solution, and the number of living copepods determined by
subtraction.
The water was held in the ballast tank for two hours (as the net tow was analyzed) and then
drained at flow rate of 3834 1 min"1 (1013 gpm), with 65 m3 (17226 gal) diverted to flow through
the filter skid at a rate of 454 1 min"1 (120 gpm), with 114 1 min"1 (30 gpm) through each of the 8
housings containing filter bags (all flows are approximate averages). The first two filter
housings in the filter skid were empty (except for the metal liners with 3 mm holes); the 8
housings downstream each held a 25-um filter bag with the seams glued twice on the inside with
3M® 5200 sealant. After the drain operation, filter bags were removed from the housings, rinsed
with ambient seawater, and their filtrands consolidated into a 2 1 graduated cylinder. The
cylinder was gently inverted 5x, and, using 10 ml serological pipettes, three researchers each
10
-------
immediately and simultaneously took a 10 ml sample from the center of the cylinder and
dispensed their samples into a 50 ml centrifuge tube. The inversion was repeated, and the
researchers took a 5-ml sample and placed it into their centrifuge tube, bringing each tube's
volume to 15 ml. This process was undertaken to ensure subsamples were representative of the
sample in the 2 1 graduated cylinder. Due to the high concentrations of plankton, sediment, and
debris, it was necessary to dilute the samples lOx using filtered seawater before they were
analyzed in the same manner as the samples from the plankton tow.
3.4.4 Effect of Crowding on Ambient Zooplankton
The effect of crowding during sampling was evaluated by comparing (1) a sample simulating a
parcel of water 60 m meeting the zooplankton discharge standard (10 living organisms m" ) and
concentrated during sampling to a volume of 1 1 (= 600 zooplankters I"1) to (2) a control sample
with a much lower concentration of zooplankton (10 I"1). Six hundred or ten copepods (Acartia
tonsa, purchased from LiveCopepods.com, Seattle, WA; salinity = 33.5), were manually counted
and added to a beaker containing 1 1 of artificial seawater (salinity = 34). Beakers were placed in
an incubator set at 25°C and illuminated by fluorescent bulbs (72 uM Einsteins m"2 s"1), gently
aerated, and incubated for 4 h. Previous work at NRLKW has shown die-off occurs in samples
held longer than six h (Riley et al., 2006), so the zooplankton handling time and incubation time
was a total of 6 h. After incubation, the copepods were recovered from the beakers using a 50-
um sieve. The dead A tonsa were counted. Next, the samples were fixed with Lugol's iodine
solution, the total number of copepods counted, and the number of living copepods was
determined by subtraction. Because the sample handling time was too long to prepare and
analyze replicate samples in a single day, the experiment was repeated the following day. In that
trial, only 572 Acartia tonsa were available, so 28 brine shrimp (Artemiafranciscand) were
added to bring the treatment beaker's total zooplankton count to 600. Because no A. franciscana
were added to the control sample, the survival was determined only for A. tonsa in both
treatments.
4 Results
4.1 Microbeads
4.1.1 Microbead Disintegration
The counts of microbeads (10 50 um microbeads and 10 150 um microbeads) at the start of the
trial and 7 days later were the same in both the filtered seawater and the Type II water (data not
shown). Therefore, microbeads did not disintegrate on the time scales used for trials.
4.1.2 Method of Microbead Counting
The nine trials comparing =dry' to jvet' counting of microbeads showed the wet method to be
superior (Figure 2). In only one of five trials using the dry method, microbead recovery was >
90%; however, in all four trials using the wet method, microbead recovery was > 90%. No bias
11
-------
was attributed to any researchers, as all three researchers achieved both good and poor recovery.
Henceforth, microbeads were counted only in the wet manner.
01
o
u
Ol
te.
-o
ra
01
.a
o
95
96
«**
Trials and Observers
Figure 2. Recovery after microbeads were counted using the =dry' and =wet' methods. Orange
bars represent dry trials, and blue bars represent wet trials; numbers above bars indicate the
percent microbead recovery; letters on the x-axis labels indicate the researcher's initials.
4.1.3 Intercalibration
When 18 Bogorov trays with a seawater sample were counted by two observers to determine the
difference from the mean of the counts, the grand mean was 12% (SD = 14%; data not shown).
When one observer counted the same tray with a seawater sample five times, the number of
microbeads counted varied over three-fold: 5, 5, 4, 2, 7 microbeads.
The intercalibation showed good agreement between observers, but because different counts
resulted when a sample with seawater (and corresponding debris) was counted repeatedly, it was
necessary to conduct validation trials—in both the laboratory and the field—using freshwater, so
microbeads would be visible.
4.2 Filter Bags
4.2.1 Filter Bag Field Trials—Water Flow Rate
The filter bags clogged immediately in the first trials, which were conducted at relatively high
flow rates (e.g., 150 gpm; Appendix 2). Subsequent trials showed a flow rate of 25 gpm allowed
sustained flow with low pressure differentials across filter bags having relatively small mesh
sizes (50 um or 25 um). In subsequent trials, the flow rate was set at 25 gpm. With the
12
-------
Hayward units, a 5 cm (2") hose supplied water to the system; at 25 gpm, water flowing through
-i
the pipe would travel at 0.78 m s" (1.7 mph).
4.2.2 Filter Bags—Field and Laboratory Validation Using Microbeads
Recovery efficiencies of 50 um microbeads in the laboratory ranged from 41% - 84% when the
microbeads were counted in the wet fashion and the filter bags' seams were unglued or glued
insufficiently with Marine Goop® (Figure 3). Usually, no microbeads were found in the filtrate;
on one occasion, 7 microbeads were found (100 were added to the filter bag; recovery
efficiencies exclude microbeads found in the filtrate). Recovery of the 50-um microbeads
generally improved after the microbeads were counted in the wet fashion and the seams were
glued more thoroughly. When 3M 5200® was used, recovery efficiency of the smallest
microbeads ranged from 82% - 110% (mean = 94%). The larger two size classes showed
excellent recovery efficiencies, at least 95%, with one exception (87%, Trial 21). A laboratory
trial using 50 um microbeads in a 50 um filter bag with seams glued twice on the inside with 3M
5200® sealant yielded a 20% recovery efficiency (data excluded from Figure 3).
110
100
90
-------
Although the tank that acted as a freshwater reservoir for field trials was cleaned beforehand, a
small amount of green algae from the tank and the piping in the system was collected and
concentrated in the filter bags (Figure 4, Trial 1; all trials except the first one had a 25-um filter
bag placed upstream to filter material from the water). During sample analysis, removing algae
from one Bogorov counting chamber with many microbeads and re-counting it led to a 20%
increase in the number of microbeads recovered (Trial 7). The little bit of algae in the samples, it
seems, obscured the 50 um microbeads, making recoveries appear to be low even though
microbeads were likely captured in the filter bags. The tank was rigorously scrubbed to remove
algae, and the experiment was repeated with 200 50 um and 200 150 um microbeads. The
results were better: 93% and 100% recovery efficiency (50 um and 150 um microbeads,
respectively); there was noticeably less algae in the sample than in previous experiments.
Subsequent trials with very clean tanks showed recovery efficiencies > 87% of 50 um
microbeads. Recovery efficiency of larger (100 um and 150 um) microbeads prior to trial 7
ranged from 74% to 101% (the latter represents a counting error or a microbead leftover from a
previous trial, although filter bags were examined for residual microbeads on a dissecting
microscope prior to each trial). Following Trial 7, efficiencies were > 99%. In all trials, nearly
all of the microbeads were recovered from the filter bag to which they had been added; in one
instance, 13 microbeads were found in the filter bag downstream of it.
Algae Removed (> 87% recovery)
>
O
Figure 4. Recovery efficiency of microbeads in field trials. Unless noted, filter bags' mesh was
25 um. Blue, red, and green bars represent 50 um, 100 um, and 150 um microbeads, respectively.
* = a notable amount of green algae was collected in filter bags.
14
-------
Because microbead recovery efficiencies increased after improvements were made to the
protocol for trials in the laboratory (i.e., counting using the wet method and sealing the filter
bags' seams) and the field (i.e., removing algae from the reservoirs used to hold freshwater), the
data collected after improvements were compiled to give a true picture of microbead recovery
(Figure 5). With the exception of the laboratory trial using 50 um microbeads in a 50-um filter
bag, the average recovery efficiency was > 89%.
10 um
sieve
lab
(2)
2 5 (.1111
FB
lab
(15)
25 (.1111
FB
field
(3)
35 [.nil
FB
field
(2)
37 um
sieve
lab
(1)
50 nr
FB
lab
(1)
Trial
Figure 5. Recovery efficiency of 50 um microbeads in laboratory and field trials conducted
after improvements were made to the microbead protocol. Numbers in parentheses
represent replicates; FB = filter bag.
4.2.3 Filter Bags—Seams
Of the seven sealants examined, Marine Goop® appeared best for sealing the filter bags' seams
because it was the most pliable but not sticky. After being used in field tests, however, the clear
sealant began to flake apart, allowing microbeads to become trapped in the sealant and
potentially uncovering the relatively large holes along the bags' seams. 3M™ 5200 was re-
evaluated, and after using it in a field trial and noticing no breakdown, it was used in all
subsequent experiments. Like Marine Goop, it was also pliable but not sticky, and it was white,
so red microbeads were visible along the seams.
4.2.4 Filter Bags—Toxicity of Sealant on the Seams
There was no apparent, negative effect of the 3M™ 5200 sealant onArtemiafranciscana or
ambient copepods. In both trials, in all replicates, the percentage of living organisms was at least
15
-------
90% (Appendix 3). In the Artemia franciscana experiment, because the data were not distributed
normally, a Mann-Whitney Rank Sum Tests was used, and no significant difference was found
(p = 0.49). In the copepod experiment, a t-test showed no significant difference between
treatment and control groups (p = 0.54). When all of the A. franciscana results (from both
control and treatment groups) were averaged into a grand mean and compared to the grand mean
of the copepod results, the numbers were nearly identical (98.3% and 98.1%, respectively).
4.3 Filter Skid
4.3.1 Filter Skid Design
After investigating commercially available filtration devices, the Eaton Topline™ Type II
housing (Eaton Filtration, LLC, Iselin, NJ) was chosen. To capture zooplankton effectively in
large volumes of water, the filter skid was designed with 10 cylindrical filter housings, each
containing a metal filter basket with holes 3 mm (0.12") in diameter (Figure 6; schematic in
Appendix 4 and operations manual in Appendix 5). As water entered the skid, it was split to
flow into one of two pre filters, which contained no filter bags, but the metal filter baskets served
as prefilters to collect large organisms, sediment, and rust flakes from the ballast tank.
Following each prefilter, water flow was split again, into two filter housings, thus evenly
distributing the water flow between the housings and reducing the filtration load for each filter
by a factor of four. Each of the four housings was followed by another filter housing, which was
intended to capture anything organisms that may have gone through the first set of housings. All
filter housings, excluding the prefilters, contained a filter bag designed to capture all organisms
greater than 50-um minimum dimension in each experiment. A key feature of the Eaton design
is the tight seal among the filter bag, housing, and housing cover.
Figure 6. Filter skid design for flow-through sampling of discharged ballast water.
16
-------
4.3.1.1 Filter Housings
The cylindrical Eaton Topline™ bag filter housings, constructed of corrosion-resistant 316
stainless steel, had dimensions of 18 cm (7") diameter x 91 cm (36") long (Figure 7; drawing
created using SolidWorks 2010 3D Design Software, Concord, MA). The height and orientation
of the housings could be varied using their support legs.
Figure 7. Eaton Topline™ filter housing.
Water entered the side of the housing through a 5 cm (2") flanged connection and flowed over
the top of the filter bag, sealing it in place. This design resulted in a minimal volume of
unfiltered liquid (i.e., between the housing's lid and the top of the filter bag) and provided
optimum sealing of the filter bag so a minimal amount of water bypassed the filter bag.
Although each of the Eaton Topline™ filter housings are manufactured for a maximum flow rate
of 681 1 m"1 (180 gpm) and 10,500 g cm"2 (150 pounds per square inch, psi) maximum pressure,
the filter skid was designed so each of the housings would realize approximately 95 1 min"1 (25
gpm) during normal operation (see section 3.3.1 for rationale). Additionally, the manufacturer's
operation instructions indicated low-flow conditions lead to good filtration results. By splitting
the flow four ways, it was possible to slow velocity through the skid by a factor of four.
A main component of the filter skid is the filter bags that were snapped into place in the eight
filter housings after the two prefilters. The material of the filter bags was chosen to (1) mimic
the filtration efficiency of plankton nets, which are commonly used by test facilities, and (2)
capture organisms and allow them to be gently removed from the filter bag surface without
increasing mortality. Felt, micro fiber, and mono filament meshes were considered, and nylon
monofilament filter bags best met the criteria (Eaton part number BNMO35P2P, Figure 8; the
filter bag's polypropylene ring is snapped into place at the top of the filter housing. The 3 mm
holes in the metal liner are visible through the filter bag's mesh).
17
-------
Figure 8. Nylon mono filament mesh filter bag secured in an Eaton Topline™ filter housing.
4.3.1.2 Piping, Valves, and Diaphragm Pump
All piping between the filter housings was 5 cm (2") diameter, Schedule 40, PVC (Spears®
Manufacturing Company, Sylmar, CA), with glued fittings (Figure 9). Piping was connected to
the filter housings using 5 cm (2") diameter 68 kg (150 Ib) Van Stone flanges with ethylene
propylene diene Monomer (M-class) rubber (EPDM) gaskets.
Figure 9. Plan view of the prototype filter skid at NRLKW. Blue arrow shows water entering
the filter skid; yellow arrow shows water exiting the skid.
18
-------
To allow filter housings to be individually isolated, 5 cm (2") diameter valves were fitted to the
inlet and outlet of each filter housing (Spears® Manufacturing Company, model Compact 2000
ball valves). Because all isolation valves were left in the 100% open position during operation,
they did not add shear forces to the organisms during sampling.
At the base of every filter housing, a threaded 1.3 cm (0.5") diameter PVC ball valve (Spears®
Manufacturing Company) was installed to serve as a drain valve. The valves remained closed
during filtration operations but were opened at the end of a run to drain the filter housings from
the bottom. In this manner, filter bags were removed from the housings without the bottoms of
the filter bags floating upwards and potentially losing part of the sample.
A threaded, 0.64 cm (0.25") diameter PVC ball valve was installed atop of each filter housing,
and it was used to manually bleed air trapped within the housings prior to use. This important
safety feature prevented air from being trapped within filter housings and potentially building to
several times greater than atmospheric pressure. This feature allowed the system operator to
know when the filter housings were full of water (i.e., water exits the valves), thereby ensuring
the filter bags were wet prior to sampling.
A Yamada NDP-80 8 cm (3") diaphragm pump (Yamada® America, Inc., Arlington Heights, IL)
was installed to draw a sample from the discharge line exiting the ballast tank into the filter skid
(Figure 10). The pump was needed because the discharge line was under suction (as water was
pumped to a discharge tank for filtration prior to discharge into ambient waters); therefore, to
collect a representative sample of the line, a diaphragm pump was been installed. It was
constructed of polypropylene with an EPDM rubber diaphragm in contact with the water passing
through the pump. Previous testing by engineers at NRLKW on various pump types on
organism mortality has shown diaphragm pumps to be the superior choice (Riley et al., 2009).
Although it was capable of flowing at a rate of 220 gpm, the pump was set at 100 gpm and
operated by compressed air supplied from a portable Sullair® 185SCFM air compressor
(Michigan City, IN). A time-averaged sample was removed from the discharge line from the
ballast tank to the line into the filter skid using a tee; an isokinetic sample port was not
necessary, as the sample volume was greater than 10% of the tank volume (i.e., 60 m3 of 200 m3;
Richard et al., 2008).
19
-------
Figure 10. Yamada NDP-80 Air diaphragm pump as installed.
4.3.1.3 Pressure Drop Calculations vs. Flow Rate
Maintaining a low pressure drop across a filter bag is critical because it reduces the chances
organisms are killed during sampling, and it enhances filtration efficiency and increases the
service life of the filter bag, thus reducing the operating cost of the system. During the design
phase, the filter skid was analyzed to determine where the bulk of the friction losses occurred.
Data for the filter housings and filter bags were available from the manufacturer (Figure 11).
The analysis was completed using data from 25 um filter bags, since data for 35 um nylon
monofilament filter bags were unavailable. As expected, the pressure drop increased as the flow
rate increased. At a flow rate of 95 1 min"1 (25 gpm), the pressure drop across an individual
housing with a 25 um nylon monofilament bag was slight, approximately 35 g cm" (0.5 psi). It
is anticipated the pressure drop would be a bit smaller using a 35-um filter bag.
20
-------
EatonTopline Filter Housing Pressure Drop (psi)vs. Flow(gpm)
-Housing pressure drop (psi)
-clean filter pressure drop for 25um NMO {psi)
-Total pressure drop across housing {psi)
-Linear {Total pressure drop across housing {psi))
100 150
Flow Rate (GPM)
Figure 11. Pressure drop of filter skid vs. flow rate (figure from Eaton Filtration, LLC).
Because the skid was designed with a small footprint, there were several elbows and fittings, as
well as flow direction changes, which contributed to the pressure drop across the skid. When all
pressure changes were calculated, the pressure drop across the entire skid was 631 g cm"2 (9 psi)
at 379 1 min"1 (100 gpm; calculations not shown).
4.3.1.4 Flow Rate, Flow Control, and Volume Measurement through the Filter
Skid
The flow rate was monitored using a GF Signet 2551 Magmeter, which was situated on the inlet
line to the skid using a saddle fitting. The total volume sampled was calculated from a totalizer
installed downstream of the skid that used the electrical signal generated by the Magmeter to
calculate the total volume by multiplying the flow rate by time and summing it over the entire
test run Because the sampling was a continuous operation, rather than a batch process, the
totalizer was used to sum the flow rate. Flow was controlled through the filter skid using a
linear, pneumatic-actuated, 7 cm (3") diameter diaphragm valve. In turn, valve position was
controlled using Proportional, Integral, Derivative (PID) feedback control via Honeywell
controls logic based on the Magmeter reading to maintain a flow of 100 gpm through the skid.
21
-------
4.3.1.5 Cost and Weight of the Filter Skid
When the cost of the materials and labor to build the filter skid was tabulated, the total cost of the
prototype was $19,851, and it took approximately 60 hours to construct it (Appendix 6).
The skid was analyzed to determine its weight with and without water. Using the respective
weight of each of the skid's components from the manufacturer's literature, the weight of the
skid without water was estimated to be 228 kg (503 Ib); when seawater is contained in all of the
housings and the piping system, the total weight of the skid was an estimated 556 kg (1225 Ib).
4.3.2 Filtration Area of Filter Skid vs. Plankton Net
The effective surface area of the filter skid was compared to that of typical plankton net. The
effective surface area of a single filter bag within the filter skid was 0.43 m (calculations not
shown). Because the sample flow was split into four filter housings in parallel, the effective
surface area is four times the surface area of one housing, or 1.70 m2. The surface area could
easily be increased by increasing the number of filter housings arranged in parallel.
In the past, a Sea-Gear Model 9000 Plankton Net with a mouth opening of 60 cm and a length to
mouth ratio of 3:1 was used atNRLKW (Sea-Gear Corporation, Melbourne, FL). Its surface
area was 1.72 m2, nearly identical to the filter skid (calculations not shown).
4.3.3 Flow Velocity through the Filter Skid vs. Plankton Net
The filter skid was designed to obtain discharge water samples at a volumetric water flow rate of
100 gpm, and the flow through the skid was plumbed to split into a set of pre-filters at a
volumetric flow rate of 50 gpm. After the two pre-filters, the flow split again to flow into a set
of four filter housings that contained 35 um filter bags to capture organisms > 50 um. The
volumetric flow rate to these four housings was 25 gpm. The flow velocity of the water as it
flows through the pre-filters and filters was calculated and compared to velocities encountered in
theoretical, horizontal plankton tows using plankton nets.
4.3.3.1 Flow Velocity through Two Prefilters (No Filter Bags)
Water was supplied to the filter housings through a 5 cm (2") inner diameter PVC pipe. The
flow velocity of the water entering the pre-filter housings is 1.55 m s"1 (3.5 mph) (Appendix 7,
Equation 1).
As the water enters the filter housing, the flow velocity is reduced because the cross sectional
area increases from the 5 cm (2") pipe to the 18 cm (7") housing (Figure 12; model created using
SolidWorks). Correspondingly, the flow velocity decreases from 1.55 m s"1 (3.47 mph) to 0.51
m s"1 (1.13 mph; Appendix 7 Equation 2). As the water leaves the housing, the flow velocity
returns to 1.55 m s"1 (3.47 mph).
22
-------
»• J
.
B"'»*
M'»)
«>»!»
»••»»
7J»1««
t* MM
I
VMOC*m«
Figure 12. Flow velocity model for water flow through an Eaton Topline™ Filter Housing at 50
gpm, entering though the top of the housing and exiting at the bottom.
4.3.3.2 Flow Velocity through Housings with Filter Bags
Downstream of the pre-filters, the water flow split to enter the next four housings, which
contained filter bags. The volumetric flow rate to each of these filter housings is 25 gpm. The
flow velocity entering these filters is 0.77 m s"1 (1.7 mph; Appendix 7, Equation ).
As seen in the first set of pre-filters, as the water enters into the housing, the flow velocity
decreases because the cross sectional area increases from 5 cm to 18 cm (2" to 7", Figure 13).
Here, the flow velocity slows from 0.77 m s"1 (1.72 mph) to 0.25 m s"1 (0.56 mph; the velocity of
the water when contacting the filter bags; Appendix 7, Equation 4). As the water leaves the
housing, it returns to 0.77 m s"1.
23
-------
'
.
Figure 13. Flow velocity model illustration for water flow through an Eaton Topline™ Filter
Housing at 25 gpm, entering though the top of the housing and exiting at the bottom.
4.3.3.3 Flow Velocity through a Plankton Net
The flow velocity of water entering a plankton net as it is towed behind a boat was calculated.
For this example, it was assumed the velocity of the boat during sampling was 0.51 - 3.1 m s"1 (1
- 6 kts; e.g., Aron, 1965), the mouth diameter of the net was 1 m, and the exit diameter near the
cod end was 10 cm. Because of its conical shape, the net tends to concentrate organisms strained
from the water sample in the cod end and uses surface filtration to separate organisms and
particles > 35 um. The entrance velocities of the water entering the plankton net for speeds 1 kt,
2 kts, and 6 kts are 0.51 m s"1, 1.03 m s"1, and 3.1m s"1, respectively. The velocity of water
entering the filtration housings in the filter skid (0.77 m s"1, Appendix 7, Equation ) is similar to
a horizontal plankton tow between 0.51 m s"1 - 1.03 m s"1 (1-2 kts).
4.3.4 Toxic Effects of Stainless Steel Filter Housings
Stainless steel did not induce mortality on 12-hour old Artemiafranciscana nauplii or ambient
zooplankton (> 50um) over a 2.5 h exposure time, as there was no significant difference between
the percentage of living organism in the control and treatment groups (Appendix 8; t-test p =
0.59 for A. franciscana, p = 0.33 for zooplankton). Although the grand mean from all A.
franciscana measurements (both control and treatment groups) was greater than the grand mean
24
-------
from all ambient zooplankton (99.1% vs. 87.8%), there was no statistical difference between
them (Mann-Whitney Rank Sum Test, p = 0.15).
4.3.5 Comparison of Zooplankton in the Ballast Tank vs. the Filter Skid
3
A sample collected using a plankton net tow from a ballast tank holding 200 m of ambient
3
seawater was compared to a time-averaged, 60-m sample collected as the tank was drained. The
percentage of living zooplankton was nearly identical (Appendix 9). The mean percentage of
living copepod nauplii was 98.3% and 97.8% (plankton tow and filter skid, respectively), and for
copepod adults and copepodites, it was 84.2% and 83.7% (plankton tow and filter skid,
respectively).
4.3.6 Effect of Crowding on Ambient Zooplankton
When approximately 600 copepods (Acartia tonsa) or 10 copepods were added to 1 1 of artificial
seawater and incubated for 4h, the percentage of living copepods was > 90% in each treatment
for both trials: 98.0% and 100% (600, 10 A. tonsa) and 97.5% and 90% (600, 10 A.tonsa; data
not shown).
5 Discussion
A filter skid was successfully designed to sample large volumes of water to determine if its
zooplankton concentration meets a ballast water discharge standard of < 10 organisms m"3.
Using off-the-shelf components, it was built to meet the requirements of having a low water-flow
rate, self-contained nets, and a small footprint. Calculations of the filtration surface area and
flow velocity show the skid is comparable to a standard plankton net, and the number of filter
housings in the skid could be increased to allow more surface area. The trade off, of course, is a
greater handling time of the sample, as more filter bags would require rinsing.
Initial trials using zooplankton proxies (microbeads 50 um in diameter) to validate filter bags'
efficiencies revealed issues that were addressed: the method for counting microbeads was
improved, the filter bags' seams were sealed to prevent microbeads from slipping through the
holes, and the small amount of algae found in holding tanks used in freshwater field trials was
removed to ensure all microbeads were visible within samples. The latter point illustrates the
importance of employing fluorescent stains or movement or both to quantify zooplankton.
Otherwise, zooplankton may be undercounted when small organisms are obscured by dead
organisms, sediment, and debris in samples. Although not addressed in this study, the same is
true of the protist (> 10 um and < 50 um) size class. In the end, microbeads approximating the
lower end of the zooplankton size class could be recovered from filter bags in field and
laboratory trials with good efficiency, > 87%.
To address concerns that the sealant used to fuse the filter bags' seams and the stainless steel of
the filter housings may be toxic to plankton as they are sequestered (albeit for a short time) in the
25
-------
filter bags within the housings, short-term toxicity tests were conducted. Both toxicity tests
exposed the test organism Artemia franciscana and ambient zooplankters to the potential
toxicant. No immediate or apparent negative effect on the crustaceans was evident. In each
instance, no significant difference between treatment or control groups was detected.
Validating nets' or filters' retention efficiencies is, as far as we know, an uncommon practice,
both in oceanographic research in general, and in ballast water treatment testing specifically.
Regarding the latter case, it seems especially relevant given the potential for fines to be levied
when ballast water management systems exceed a discharge standard. Microbeads are not a
perfect proxy for living organisms, which may be squeezed through a net more readily than
polystyrene spheres. The microbeads do, however, represent a good metric, as they can be
purchased at the lower size class (e.g., with a diameter of 50 um) and have no spines or setae to
cling to a net or filter. A logical—and necessary—next step is to compare retention efficiencies
of natural aquatic communities between the filter skid and plankton nets. Those trials are
underway at NRLKW.
6 Acknowledgements
William =BF Kinee and Barren Stringham helped construct the filter skid and run field trials, and
we are grateful for their assistance. We appreciate and acknowledge the support of the US Coast
Guard Environmental Standards Division (CG-5224) for funding this research.
7 References
Aron W, Ahlstrom EH, Bary BMcK, Be AWH, Clarke WD (1965) Towing characteristics of
plankton sampling gear. Limnol Oceanogr 3:33-340
Federal Register (2009) Standards for Living Organisms in Ships' Ballast Water Discharged in
U.S. Waters; Draft Programmatic Environmental Impact Statement, Proposed Rule and
Notice, 74 FR 44631-44672 (28 August 2009). National Archives and Records
Administration, Washington, DC
International Maritime Organization (2004) Convention BWM/CONF/36 Jnternational
Convention for the Control and Management of Ships' Ballast Water and Sediments, 2004'
International Maritime Organization (2005) Resolution MEPC. 125(53) =Guidelines for Approval
of Ballast Water Management Systems (G8)'
Lemieux EJ, Grant J, Wier T, Drake L, Robbins S, Burns K. In review. Generic Protocol for the
Verification of Ballast Water Treatment Technologies, Version 4.2. Submitted to the EPA
Environmental Technology Verification Program
Lemieux, EJ, Grant J, Wier T, Robbins S, Riley S, Hyland W, Davis L, Donowick T, Stringham
B, Kinee WJ, Brown B, Everett R (2008a) Pilot Environmental Technology Verification
26
-------
(ETV) Test Report of the Severn Trent DeNora BalPureTM Ballast Water Treatment System,
NRL Letter Report 6130/6098, Washington, DC
Lemieux EJ, Robbins S, Burns K, Ratcliff S, Herring P (2008b) 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
Richard RV, Grant JF, Lemieux EJ (2008) Analysis of Ballast Water Sampling Port Designs
using Computational Fluid Dynamics. Report No. CG-D-01-08 issued by the US Coast
Guard Research and Development Center, Groton, CT
Riley SC, Grant JF, Lemieux EJ, Robbins SH, Hyland WB (2006) Biological in-tank sampling
and sample degradation for standardized ballast water treatment technology sampling.
Abstract in the Proc 14th International Conference on Aquatic Invasive Species, Key
Biscayne, Fl
Riley SC, Lemieux EJ, Grant JF, Robbins SH, Hyland WB, Davis LE (2009) Validation of
biological methods for full-scale treatment testing. Abstract in Proc 16th International
Conference on Aquatic Invasive Species, Montreal, Quebec, Canada
27
-------
Appendix 1
Data from initial field trials conducted at NRLKW comparing microbead recovery efficiency
between a plankton net and filter bags arranged in series.
Type of
Filtration
Plankton
net
Filter bags
(2 in series)
Parameter
Mesh size
35 um net with two filter
bags in series as prefilters:
50 um (first in series) and
25 um (second)b
100 um (first) and 25 um
mesh (second)
Dimensions
99 cm (39")
diameter at the top
and 162 cm (64")
long
1 8 cm (7") diameter
at the top and 41
(16") cm long
Flow rate
(gpm)
100
25
Number of
50 um-
diameter
microbeads
recovereda
150/600
(25%)
168/600
(28%)
aThe 600 microbeads, in Type II water in a beaker, were gently poured into the cod cup of the
plankton net or the first filter bag in series. Afterwards, the beaker was examined for residual
microbeads, and in both cases, 0 were found.
bA previous trial conducted with ambient seawater pumped through the plankton net with no
prefilters in place yielded a sample so loaded with debris and sediment that it would have taken
days to analyze all of it. In the 8% of sample analyzed, 1 microbead (of 600 added) was found.
All subsequent trials used prefilters with ambient seawater (as was the case for data in this table)
or freshwater.
28
-------
Appendix 2
Results of preliminary trials to determine the appropriate flow rate through the filter bags to
allow maximum flow and minimum clogging.
Mesh in
first bag
in series
(urn)*
25
50
100
50
25
Mesh in
second
bag in
series
(urn)
25
50
25
25
25
Flow
rate
(1 min '*)
(gpm)
568
(150)
474
(125)
95
(25)
95
(25)
95
(25)
Duration
of the
run
(min)
9
46
64
80
72
Outcome
The bags clogged and water flow nearly stopped
almost immediately; a total of 87 gal (330 1)
flowed
The bags clogged, pressure differential between
filter housings increased (PI = 5.3 psi, P2 =
19.1, P3 = 59.4 at the end of the run), and the
flow rate slowed to 62.1 gpm; 3312 gal (12,550
1) flowed
The pressure differential remained slight (PI =
64.8 psi, P2 = 63.9, P3 = 64.3 at 53 min); 1485
gal (5627 1) flowed with little clogging of bags
The pressure differential remained slight (PI =
65.2 psi, P2 = 63.9, P3 = 64.3 at 70 min); 1460
gal (5533 1) flowed with little clogging of bags
The pressure differential remained slight (PI =
64.5, P2 = 63.2, P3 = 63.3); 1833 gal (6946 1)
flowed with little clogging of bags
''All experiments were conducted using Hayward Filtration units with 41-cm long filter bags.
29
-------
Appendix 3
Data from toxicity tests evaluating the effect of 3M™ 5200 sealant on brine shrimp Artemia
franciscana (top table) and ambient zooplankton (bottom table) after two hours of exposure time.
Replicate
Cl*
C2
C3
C4
Tl
T2
T3
T4
Live
Af
78
60
43
48
49
56
50
66
Total
Af
79
60
43
48
54
56
50
68
% living
Af
98.7
100.0
100.0
100.0
90.7
100.0
100.0
97.1
C = control group, T = treatment group, Af = Artemia franciscana, cultured brine shrimp.
Replicate
Cl*
C2
C3
C4
Tl
T2
T3
T4
Live
copepods
(> 50 jim)
20
19
19
20
20
19
19
20
Total
copepods
(> 50 um)
20
20
20
20
20
20
19
20
% living
copepods
100.0
95.0
95.0
100.0
100.0
95.0
100.0
100.0
C = control group, T = treatment group
30
-------
Appendix 4
Eaton Filter Skid drawing with dimensions.
Eaton Filter Skid "EPS"
Tim Wier NRL-ANS Project
6/14/2010 Rage 1
31
-------
Appendix 5
Eaton Filter Skid operations manual.
These instructions are specific to operations using the Eaton Filter Skid (EPS, shown in Figure 1)
and performed at the Naval Research Laboratory ballast water treatment test facility (BWTTF) in
Key West, Florida. The procedures typically used in trials with the EPS are broadly categorized
into preparation, operation, and shutdown.
Figure 1. Eaton Filter Skid showing air bleed and drain ports and housing manual inlet and
outlet ball valves.
Only a trained system operator shall perform operations specified in this document; improper
setup or operation can jeopardize personnel safety or damage equipment.
Preparation
1. Verify there is sufficient capacity in the discharge tank to receive the entire volume of
water contained in the ballast tank.
2. Isolate the EPS by closing the manual 3" butterfly inlet and outlet valves (Figure 2,
Valves 510 and 404, respectively), which are upstream and downstream of the EPS.
32
-------
CO
CD
S>
CD
_c
o
>
: control v
;m valve
CL) TO
C -C
— Q.
1!
Q. Si
E B
£ <
8 a
rn rn
CD
>
0)
—
=
1
0)
o
m
g
1"
s
control v;
0)
1
c
So
TO
-5
i5
s
f
o
oj
Q.
1
^
1 0
butterfli
ain line ir
m E
3 to
c E
i t
' h-
0) -0
"t= ^
~ g>
m iZ
O
-------
3. Close 1/2" drain port valves on each of the individual Eaton Topline™ filter housings at
their posterior ends (Figure 1 and Figure 3, balloon 5).
£2.000
|RD11
1
?
3
4
5
A
I
1.
1/4
h
we* Me?
d clamp
failed per
for cover
W drain pen
n
tor Oullel
Figure 3. Eaton Topline™ Filter Housing used in the Eaton Filter Skid.
Open each housing by unscrewing the four lid clamps at the anterior end and lifting the
hinged lid by its handle (Figure 1 or Figure 3, balloon 2).
Install a filter bag in each of the filter housings (Figure 4). The filter bag will slide into
the housing, and the top plastic ring of the filter bag should seat into the top of the
housing. Take care when installing the filter bags to prevent their ripping on sharp edges
during installation.
Fill each of the housings from the top using a hose with running seawater. Filter the
seawater by placing the end of the hose in a 25-|am mesh filter bag and directing the
water that passes through the filter bag into all filter housings.
Close each of the individual filter housings by ensuring O-rings are in the proper
position, which is determined by ensuring each O-ring lies flat in its designated circular
slot (Figure 4). Shut the lid and re-tighten the four lid clamps.
34
-------
Figure 4. Filter housing with cover open and filter bag installed; O-rings are properly seated.
8. Open %" bleed port valves at the anterior end of each housing (Figure 3, balloon 3).
9. Open all manual and control valves to create a flow path from the ballast tank to the
discharge tank with an exception of the EPS and discharge tank inlets, which are to
remain closed.
10. The EPS draws flow from sample port 1 l(Figure 2, SP-11) which connects to the main 6"
ballast tank discharge line by a T fitting. Slowly open the EPS inlet valve to
approximately 50% until water under head pressure from the ballast tank generates a
steady stream flowing from the open %" filter housing bleed ports, indicating a full prime
is achieved.
a. At this point, if a housing lid is not properly sealed, a leak (due to a loose lid
clamp or unaligned O-ring) will be apparent. This situation must be remedied by
closing the EPS inlet valve, isolating the leaking housings at their individual inlets
and outlets by closing 2" valves (Figures 1 and 2), and then correcting the issue.
b. Re-open housing inlet and outlet valves then repeat step 8.
11. Throttle back the manual EPS inlet valve until the flow in step 10 is reduced to a
minimum, then close all bleed port valves.
12. Open the EPS inlet valve completely. Pressure from the ballast tank will equalize
throughout the system upstream of the discharge tank inlet valve such that opening it
would invoke flow, i.e., this valve should be the last obstacle preventing head pressure
from the ballast tank from causing a gravity-induced flow into the discharge tank.
13. Check the fuel level for portable air compressor to ensure there is an ample amount for
the desired run time.
35
-------
14. Hook up the %" air hose from the diaphragm pump to the portable air compressor. The
%" air hose has Chicago-type twist-lock fittings. These fittings should align and mate
correctly.
15. The diaphragm pump, powered from high-pressure air generated by the compressor,
feeds into and is controlled binarily using an electrically actuated solenoid valve located
toward the top end of the pump. Ensure the power plug from the solenoid valve is
plugged into a power receptacle near the diaphragm pump.
16. Start the compressor and then open the corresponding manual valves that feed air to the
diaphragm pump (Figure 2, Diaphragm Pump).
17. At the BWTTF in Key West, a Honeywell Plantscape Control and Automation system is
integrated such that all switches, sensors, and machinery send inputs and receive outputs
from a programmable logic controller that can be monitored, operated, and controlled
from a human-machine interface (HMI) in a centralized control room. Using an identical
or similar system is required to enter proper parameters for desired pressure and ballast
tank discharge and EPS sampling flow set points into control operation input fields in the
HMI. Depending on the location of the controlling flow meter, the set point should be
adjusted to account for sampling flow to achieve total desired flow, e.g., if the primary
pumps are controlled using a flow meter upstream from where sampling flow rejoins
main flow; the set point should be the total designed flow minus the sampling flow.
Operation
1. Partially open the inlet control valve to the discharge tank (Figure 2, Valve 405) using the
HMI; verify that flow is initialized from the ballast tank to the discharge tank.
Note: If little or no flow observed, an issue needs further investigation by the
system operator. Causes may include valves set in an improper position, an
inoperable flow meter, or a higher level of water in the discharge tank than the
ballast tank.
2. Begin the diaphragm pump and sampling automation from the HMI by clicking =Start'
under the appropriate control operations. Pumps and valves will actuate to achieve
desired set points.
3. Continually monitor EPS pressure and flow to ensure filter bags do not clog; pressure
should not exceed 30 pounds per square inch.
Shutdown
1. As head pressure from the ballast tank decreases, the primary pumps (Figure 2, Pumps 8
and 9) will start to overpower the suction from the diaphragm pump drawing water from
the sample port (Figure 2, SP-11). This situation must be prevented by closely
monitoring EPS flow as the ballast tank level becomes low (relative to the total flow and
tank dimensions, i.e., approximately 10 minutes remaining at the current flow rate and
remaining volume) until it suddenly begins to decrease dramatically. At this point:
a. Shut down the primary pumps (Figure 2).
b. Close the primary pumps' inlet control valve (Figure 2, Valve 303) to isolate all
flow through EPS.
c. Open the control valve to the discharge tank inlet to 100% (Figure 2, Valve 405).
d. Normal flow through EPS should be restored and stable.
36
-------
2. Depending on tank dimensions, flow will continue for a period until it again begins to
decrease drastically. At this point:
a. Close the automated diaphragm valve (Figure 2, Valve 313) at the outlet of the
diaphragm pump.
b. Immediately shut down the diaphragm pump.
c. Immediately isolate EPS from the system by closing the six 2" ball valves at the
farthest upstream and downstream filter inlets and outlets (Figure 5).
Figure 5. Ball valves (in red circles) closed immediately after the diaphragm pump is shut down.
3. Isolate the remaining housings by closing all 2" ball valves.
4. Release pressure at the posterior end 1/2" drain valves such that flow channels through
the filters and out through the drain valves.
5. Once pressure releases, open the 1A" bleed valves at the top of the housings. The
remaining water will drain onto the ground until no more water exits the housings.
6. Open the lid of each housing by unscrewing lid clamps and lifting at the handle.
The filter bags are ready for extraction.
37
-------
Appendix 6
Eaton Filter Skid bill of materials and calculation of weight.
Filter Housing
M
ro
CD
1_
-------
•JO
to
T3
O
O
^
0)
ro
ro
to
o>
c
TO
5?
u to
O 00
-J C
£ S
TO LL.
0
Equipment
4x8 treated plywood 3/4"
2x6 treated 8' board
4x4 treated 8' board
Wood Screws
5/8-11x3" 316 stainless steel hex cap screw
partially threaded
5/8-11 316 Stainless steel hex cap nut
Type 316 SS Type A SAE Flat Washer 5/8" Screw
Size, 1-5/16" OD, .07"-. 13" Thick, Packs of 10
Type 316 Stainless Steel Split Lock Washer 5/8"
Screw Size, 1.08" OD, .15" min Thick, Packs of
10
Type 316 Stainless Steel Hex Head Cap Screw
5/8"-ll Thread, 3-1/4" Length, Packs of 1
Aluminum Cam-and-Groove Hose Coupling
Plug, PFA Adapter, 3 Coupling Size, 3" Pipe Size
Quantity
2
3
2
40
100
22
20
12
10
2
Cost
$32.00
$5.50
$9.50
$0.10
$4.32
$8.00
$8.54
$5.37
$4.69
$62.30
Total Cost
$64.00
$16.50
$19.00
$4.00
$432.00
$176.00
$170.80
$64.44
$46.90
$124.60
Cost of Equipment Components $19,851.57
o
TO
Hours to build skid per design drawing
80
$150.00
$12,000.00
Total Discharge Skid Cost $19,851.57
39
-------
Appendix 7
Calculations of flow velocity in the filter skid and during a hypothetical plankton tow.
Equation 1 . Flow velocity calculation for 50 gpm in a 5 cm (2") diameter pipe.
/ ft \ Volumetric Flow Rate (50 <& * °'13 3 (jd*) * 144&)
Flow Velocity — = - - - = - 7 - - - — - -
Vsec/ Cross Sectional Area, I fsec\ /2in\^\
60 1 — — * TT* 1-^-
\ \minJ \ 2 J I
i ft \ fm\ / m \
= 5.081- — * 0.3048 f — = i-33 ( -
Vsec/ \ft/ Vsec/
Equation 2. Flow velocity calculation for 50 gpm in an 18 cm (7") diameter filter housing.
//^ Volumetric Flow Rate <50 * °'133 $>* 144<&>
Flow Velocity [ — } =
Vsec/ Cross Sectional Area
/ sec \ f3 5in\~
601 - — *TT*|-L= -
\ Vmin/ V 2 )
f ft \ /m\ / tn \
= 1,66 — * 0.3048 (— = 0.51 ( -
Vsec/ \ftJ \secs
Equation 3. Flow velocity calculation for 25-gpm flow in a 5 cm (2") diameter pipe.
/ ft \ Volumetric Flow Rate
Flow Velocity { — } = - =
* °'133 *
Vsec/ Cross Sectional Area f^f\/-sec-\^ .„ ,-2-m,,,
(6U^ ;—J * 71 * \ •) )"
f ft \ /m\ m
= 2.541— * 0.3048 — = 0.77 ( )
Vsec/ \ftJ secj
Equation 4. Flow velocity calculation for 25-gpm flow in an 18 cm (7") diameter filter housing
1 JT 3 *"*
/ ft \ Volumetric Flow Rate (25 fe^) * °'133 C^j) * 144(
Flow Velocity I = ; = r-p-;
Vsec/ Cross Sectional Area fff.fsec^ ,3.5zn-,,x
(60(——) * TT * f—=—)-)
\ "'JYt^Yt *-
/ft \ /m\ m
= .83 (— * 0.3048 — = 0.2S ( )
Vsec/ \ft) secj
40
-------
Appendix 8
Results of toxicity tests evaluating the effect of stainless steel on brine shrimp Artemia
franciscana and ambient zooplankton for 2.5 h of exposure.
Replicate
Cl*
C2
Tl
T2
T3
Live
Af
264
262
353
229
348
Total
Af
265
267
354
232
350
% living
Af
99.6
98.1
99.7
98.7
99.4
Live
zoopl
(> 50 jim)
34
15
34
28
30
Total
zoopl
(> 50 um)
38
20
38
33
30
% living
zoopl
89.5
75.0
89.5
84.9
100
* C = control group, T = treatment group, Af' = Artemia franciscana; zoopl = ambient
zooplankton > 50 um.
Appendix 9
Results of the comparison between living copepods in the ballast tank and the filter skid.
Replicate
PT1*
PT2
FS1
FS2
Live
nauplii
83
104
293
279
Total
nauplii
86
104
299
286
% living
nauplii
96.5
100
98
97.6
Live
A + C
39
28
79
98
Total
A + C
43
36
97
114
% living
A + C
90.7
77.8
81.4
86
* PT = plankton tow; FS = filter skid; A + C = adult and copepodite stages. All copepods were > 50 um
in minimum dimension.
41
------- |