Part 2 – Test and Performance Specifications for Approval of Ballast Water Management Systems

The Administration decides the sequence of land-based and shipboard testing.

2.1 Quality Assurance and Quality Control Procedures

2.1.1 The testing body performing the tests should have implemented appropriate quality control measures in accordance with recognized international standards acceptable to the Administration.

2.1.2 The approval testing process should contain a rigorous quality control/quality assurance program, consisting of:

.1 Both a Quality Management Plan (QMP) and a Quality Assurance Project Plan (QAPP). Guidance on preparation of these plans, along with other guidance documents and other general quality control information are available from appropriate international organizations^footnote.
.2 The QMP addresses the quality control management structure and policies of the testing body (including subcontractors and outside laboratories).
.3 The QAPP is a project specific technical document reflecting the specifics of the BWMS to be tested, the test facility, and other conditions affecting the actual design and implementation of the required experiments.

2.2 Shipboard tests

2.2.1 A shipboard test cycle includes:

.1 the uptake of ballast water of the ship;
.2 the storage of ballast water on the ship;
.3 treatment of the ballast water in accordance with paragraph 2.2.2.3 by the BWMS, except in control tanks; and
.4 the discharge of ballast water from the ship.

Success criteria for shipboard testing

2.2.2 In evaluating the performance of BWMS installation(s) on a ship or ships, the following information and results should be supplied to the satisfaction of the Administration:

.1 Test plan to be provided prior to testing.
.2 Documentation that the BWMS is of a capacity within the range of the treatment rated capacity for which it is intended.
.3 The amount of ballast water tested in the test cycle on board should be consistent with the normal ballast operations of the ship and the BWMS should be operated at the treatment rated capacity for which it is intended to be approved.
.4 Documentation of the results of three consecutive, valid test cycles showing discharge of treated ballast water in compliance with regulation D-2.
.5 Valid tests are indicated by uptake water, for both the control tank and ballast water to be treated, with viable organism concentration exceeding 10 times the maximum permitted values in regulation D-2.1 and control tank viable organism concentration exceeding the values of regulation D-2.1 on discharge.
.6 Sampling regime:
- .1 For the control tank:
  - .1 three replicate samples of influent water, collected over the period of uptake (e.g., beginning, middle, end); and
  - .2 three replicate samples of discharge control water, collected over the period of discharge (e.g., beginning, middle, end).
- .2 For treated ballast water:
  - .1 Three replicate samples of discharge treated water collected at each of three times during the period of discharge (e.g., 3 x beginning, 3 x middle, 3 x end).
- .3 Sample sizes are:
  - .1 For the enumeration of organisms greater than or equal to 50 micrometres or more in minimum dimension, samples of at least one cubic metre should be collected. If samples are concentrated for enumeration the samples should be concentrated using a sieve no greater than 50 micrometres mesh in diagonal dimension.
  - .2 For the enumeration of organisms greater than or equal to 10 micrometres and less than 50 micrometres in minimum dimension, samples of at least one litre should be collected. If samples are concentrated for enumeration the samples should be concentrated using a sieve no greater than 10 micrometres mesh in diagonal dimension.
  - .3 For the evaluation of bacteria a sample of at least 500 millilitres should be taken from the influent and treated water. In the absence of laboratory facilities on board the toxicogenic test requirements should be conducted in an appropriately approved laboratory. However, this may limit the applicability of this test.
.7 The test cycles including invalid and unsuccessful test cycles are to span a trial period of not less than six months.
.8 The applicant is requested to perform three consecutive test cycles that comply with regulation D-2 and which are valid in accordance with paragraph 2.2.2.5. Any invalid test cycle does not affect the consecutive sequence.
.9 The source water for test cycles shall be characterized by measurement of salinity, temperature, particulate organic carbon and total suspended solids.
.10 For system operation throughout the trial period, the following information should also be provided:
- .1 documentation of all ballast water operations including volumes and locations of uptake and discharge, and if heavy weather was encountered and where;
- .2 the possible reasons for the occurrence of an unsuccessful test cycle, or a test cycle discharge failing the D-2 standard should be investigated and reported to the Administration;
- .3 documentation of scheduled maintenance performed on the system;
- .4 documentation of unscheduled maintenance and repair performed on the system;
- .5 documentation of engineering parameters monitored as appropriate to the specific system; and
- .6 documentation of functioning of the control and monitoring equipment.

2.3 Land-based testing

2.3.1 The test set-up including the ballast water treatment equipment should operate as described in the provided documentation during at least 5 valid replicate test cycles. Each test cycle should take place over a period of at least 5 days.

2.3.2 A land-based test cycle should include:

.1 the uptake of ballast water by pumping;
.2 the storage of ballast water for at least 5 days;
.3 treatment of ballast water within the BWMS, except in control tanks; and
.4 the discharge of ballast water by pumping.

2.3.3 Testing should occur using different water conditions sequentially as provided for in paragraphs 2.3.17 and 2.3.18.

2.3.4 The BWMS should be tested at its rated capacity or as given in paragraphs 2.3.13 to 2.3.15 for each test cycle. The equipment should function to specifications during this test.

2.3.5 The analysis of treated water discharge from each test cycle should be used to determine that the average of discharge samples does not exceed the concentrations of regulation D-2 of the Convention.

2.3.6 The analysis of treated water discharge from the relevant test cycle(s) should also be used to evaluate the toxicity of the discharged water for BWMS that make use of Active Substances and also for those BWMS that do not make use of Active Substances or Preparations but which could reasonably be expected to result in changes to the chemical composition of the treated water such that adverse impacts to receiving waters might occur upon discharge. Toxicity tests of the treated water discharge should be conducted in accordance with paragraphs 5.2.3 to 5.2.7 of the Procedure for approval of ballast water management systems that make use of Active Substances, as revised (resolution MEPC.169(57)).

Land-based testing objectives, limitations and criteria for evaluation

2.3.7 The land-based testing serves to determine the biological efficacy and environmental acceptability of the BWMS under consideration for Type Approval. The approval testing aims to ensure replicability and comparability to other treatment equipment.

2.3.8 Any limitations imposed by the ballast water management system on the testing procedure described here should be duly noted and evaluated by the Administration.

Land-based set-up

2.3.9 The test set-up for approval tests should be representative of the characteristics and arrangements of the types of ships in which the equipment is intended to be installed. The test set-up should therefore include at least the following:

.1 the complete BWMS to be tested;
.2 piping and pumping arrangements; and
.3 the storage tank that simulates a ballast tank, constructed such that the water in the tank should be completely shielded from light.

2.3.10 The control and treated simulated ballast tanks should each include:

.1 a minimum capacity of 200 m³;
.2 normal internal structures, including lightening and drainage holes;
.3 standard industry practices for design, construction and surface coatings for ships; and
.4 the minimum modifications required for structural integrity on land.

2.3.11 The test set-up should be pressure-washed with tap water, dried and swept to remove loose debris, organisms and other matter before starting testing procedures, and between test cycles.

2.3.12 The test set-up will include facilities to allow sampling as described in paragraphs 2.3.26 and 2.3.27 and provisions to supply influents to the system, as specified in paragraph 2.3.19 and/or 2.3.20. The installation arrangements should conform in each case with those specified and approved under the procedure outlined in section 7 of the main body to these Guidelines.

Ballast water treatment equipment scaling

2.3.13 In-line treatment equipment may be downsized for land-based testing, but only when the following criteria are taken into account:

.1 equipment with a TRC equal to or smaller than 200 m³/h should not be downscaled;
.2 equipment with a TRC larger than 200 m³/h but smaller than 1,000 m³/h may be downscaled to a maximum of 1:5 scale, but may not be smaller than 200 m³/h; and
.3 equipment with a TRC equal to, or larger than, 1,000 m³/h may be downscaled to a maximum of 1:100 scale, but may not be smaller than 200 m³/h.

2.3.14 The manufacturer of the equipment should demonstrate by using mathematical modelling and/or calculations, that any downscaling will not affect the ultimate functioning and effectiveness on board a ship of the type and size for which the equipment will be certified.

2.3.15 In-tank treatment equipment should be tested on a scale that allows verification of full-scale effectiveness. The suitability of the test set-up should be evaluated by the manufacturer and approved by the Administration.

2.3.16 Larger scaling may be applied and lower flow rates used than provided for in 2.3.13, if the manufacturer can provide evidence from full-scale shipboard testing and in accordance with 2.3.14 that scaling and flow rates will not adversely affect the ability of the results to predict full-scale compliance with the standard.

Land-based test design – inlet and outlet criteria

2.3.17 For any given set of test cycles (5 replicates is considered a set) a salinity range should be chosen. Given the salinity, the test water used in the test set up described above should have dissolved and particulate content in one of the following combinations:

	Salinity
	> 32 PSU	3 – 32 PSU	< 3 PSU
Dissolved Organic Carbon (DOC)	> 1 mg/l	> 5 mg/l	> 5 mg/l
Particulate Organic Carbon (POC)	> 1 mg/l	> 5 mg/l	> 5 mg/l
Total Suspended Solids (TSS)	> 1 mg/l	> 50 mg/l	> 50 mg/l

2.3.18 At least two sets of tests cycles should be conducted, each with a different salinity range and associated dissolved and particulate content as prescribed in paragraph 2.3.17. Tests under adjacent salinity ranges in the above table should be separated by at least 10 PSU^footnote.

2.3.19 Test organisms may be either naturally occurring in the test water, or cultured species that may be added to the test water. The organism concentration should comply with paragraph 2.3.20 below.

2.3.20 The influent water should include:

.1 test organisms of greater than or equal to 50 micrometres or more in minimum dimension should be present in a total density of preferably 10⁶ but not less than 10⁵ individuals per cubic metre, and should consist of at least 5 species from at least 3 different phyla/divisions;
.2 test organisms greater than or equal to 10 micrometres and less than 50 micrometres in minimum dimension should be present in a total density of preferably 10⁴ but not less than 10³ individuals per millilitre, and should consist of at least 5 species from at least 3 different phyla/divisions;
.3 heterotrophic bacteria should be present in a density of at least 10⁴ living bacteria per millilitre; and
.4 the variety of organisms in the test water should be documented according to the size classes mentioned above regardless if natural organism assemblages or cultured organisms were used to meet the density and organism variety requirements.

2.3.21 The following bacteria do not need to be added to the influent water, but should be measured at the influent and at the time of discharge:

.1 Coliform;
.2 Enterococcus group;
.3 Vibrio cholerae; and
.4 Heterotrophic bacteria.

2.3.22 If cultured test organisms are used, then it should be ensured that local applicable quarantine regulations are taken into account during culturing and discharge.

Land-based monitoring and sampling

2.3.23 Change of numbers of test organisms by treatment and during storage in the simulated ballast tank should be measured using methods described in Part 4 of the annex, paragraphs 4.5 to 4.7.

2.3.24 It should be verified that the treatment equipment performs within its specified parameters, such as power consumption and flow rate, during the test cycle.

2.3.25 Environmental parameters such as pH, temperature, salinity, dissolved oxygen, TSS, DOC, POC and turbidity (NTU)^footnote should be measured at the same time that the samples described are taken.

2.3.26 Samples during the test should be taken at the following times and locations: immediately before the treatment equipment, immediately after the treatment equipment and upon discharge.

2.3.27 The control and treatment cycles may be run simultaneously or sequentially. Control samples are to be taken in the same manner as the equipment test as prescribed in paragraph 2.3.26 and upon influent and discharge. A series of examples are included in figure 1.

2.3.28 Facilities or arrangements for sampling should be provided to ensure representative samples of treated and control water can be taken that introduce as little adverse effects as possible on the organisms.

2.3.29 Samples described in paragraphs 2.3.26 and 2.3.27 should be collected in triplicate on each occasion.

2.3.30 Separate samples should be collected for:

.1 organisms of greater than or equal to 50 micrometres or more in minimum dimension;
.2 organisms greater than or equal to 10 micrometres and less than 50 micrometres in minimum dimension;
.3 for coliform, enterococcus group, Vibrio cholerae and heterotrophic bacteria; and
.4 toxicity testing of treated water, from the discharge, for BWMS that make use of Active Substances and also for those BWMS that do not make use of Active Substances or Preparations but which could reasonably be expected to result in changes to the chemical composition of the treated water such that adverse impacts to receiving waters might occur upon discharge.

2.3.31 For the comparison of organisms of greater than or equal to 50 micrometres or more in minimum dimension against the D-2 standard, at least 20 litres of influent water and l cubic metre of treated water, in triplicate respectively, should be collected. If samples are concentrated for enumeration, the samples should be concentrated using a sieve no greater than 50 micrometres mesh in the diagonal dimension.

2.3.32 For the evaluation of organisms greater than or equal to 10 micrometres and less than 50 micrometres in minimum dimension, at least 1 litre of influent water and at least 10 litres of treated water should be collected. If samples are concentrated for enumeration, the samples should be concentrated using a sieve no greater than 10 micrometres mesh in the diagonal dimension.

2.3.33 For the evaluation of bacteria, at least 500 millilitres of influent and treated water should be collected in sterile bottles.

2.3.34 The samples should be analysed as soon as possible after sampling, and analysed live within 6 hours or treated in such a way so as to ensure that proper analysis can be performed.

2.3.35 The efficacy of a proposed system should be tested by means of standard scientific methodology in the form of controlled experimentation, i.e. “experiments”. Specifically, the effect of the BWMS on organism concentration in ballast water should be tested by comparing treated ballast water, i.e. “treated groups”, to untreated “control groups”, such that:

.1 one experiment should consist of a comparison between control water and treated water. Multiple samples, but at a minimum of three, of control and treated water within a single test cycle should be taken to obtain a good statistical estimate of the conditions within the water during that experiment. Multiple samples taken during a single test cycle should not be treated as independent measures in the statistical evaluation of treatment effect, to avoid “pseudo-replication”.

2.3.36 If in any test cycle the average discharge results from the control water is a concentration less than or equal to 10 times the values in regulation D-2.1, the test cycle is invalid.

2.3.37 Statistical analysis of BWMS performance should consist of t-tests, or similar statistical tests, comparing control and treated water. The comparison between control and treated water will provide a test of unexpected mortality in the control water, indicating the effect of an uncontrolled source of mortality in the testing arrangement.

2.4 Reporting of test results

2.4.1 After approval tests have been completed, a report should be submitted to the Administration. This report should include information regarding the test design, methods of analysis and the results of these analyses.

2.4.2 The results of biological efficacy testing of the BWMS should be accepted if during the land-based and shipboard testing conducted as specified in sections 2.2 and 2.3 of this annex it is shown that the system has met the standard in regulation D-2 in all test cycles as provided in paragraph 4.7 below.