Vous êtes sur la page 1sur 18

INTERNATIONAL MARITIME ORGANIZATION

E
IMO

MARINE ENVIRONMENT PROTECTION MEPC 60/13


COMMITTEE 13 December 2009
60th session Original: ENGLISH
Agenda item 13

HARMFUL ANTI-FOULING SYSTEMS FOR SHIPS

The generation of biocide leaching rate estimates for anti-fouling coatings and their use in
the development of proposals to amend annex 1 of the AFS Convention

Submitted by International Paint and Printing Ink Council (IPPIC)

SUMMARY
Executive summary: In submission of a proposal to restrict a harmful anti-fouling coating,
Parties to the International Convention on the Control of Harmful
Anti-fouling Systems on Ships (AFS Convention) are required to
provide data and documentation on leaching rates of biocides from
the system. This document provides a review of the available
international standard methods for determination of biocide leaching
rates from anti-fouling coatings and suggests an appropriate route for
accurately and realistically estimating the leaching rate of biocides
from anti-fouling coatings to the environment as required in
annex 3 (3)(a) of the Convention.
Strategic direction: 7

High-level action: 7.1.2

Planned output: 7.1.2.8

Action to be taken: Paragraph 6

Related documents: None

Introduction

1 Article 6 of the International Convention on the Control of Harmful Anti-Fouling


Systems on Ships (AFS Convention) describes the process by which Parties can propose
restrictions on anti-fouling systems through amendment of annex 1 of the Convention, which lists
controls on harmful anti-fouling systems. In making a proposal to amend annex 1, Parties must
submit a proposal to the Organization containing information on an anti-fouling system as listed
in annex 2 of the Convention for an initial proposal and in annex 3 of the Convention for a
comprehensive proposal.

For reasons of economy, this document is printed in a limited number. Delegates are
kindly asked to bring their copies to meetings and not to request additional copies.

I:\MEPC\60\13.doc
MEPC 60/13 -2-

2 The most common way to control bio-fouling on immersed surfaces of ships is through
the use of biocidal anti-fouling coatings. Such anti-fouling products are applied to ship hulls and
rely on the presence of one or more biocidally Active Substances (biocides) in the coating to
provide protection from bio-fouling. In order to prevent fouling, biocides are gradually released
into the environment during the practical lifetime of the dry paint film creating a micro-layer of
biocidal activity at the paint surface.

3 In an assessment of the environmental risk associated with the use of a biocidal


anti-fouling coating, the rate of entry of biocide into the aquatic environment from the coating
applied to the ships hull, i.e. the biocide release rate (leaching rate), is an essential data.
Annex 3 (3)(a) of AFS Convention lists ‘leaching rates of biocides or active ingredients’ as
required data in a comprehensive proposal submitted by a Party in support of a proposal to
amend annex 1 of the Convention.

4 As biocide release rate is a key parameter in the assessment of the environmental impact
of a biocidal anti-fouling coating, the value used must be accurate and representative of the
biocide release rate from the coating when applied to the hulls of ships and immersed surfaces.

5 The annex of this document provides a review of the available international standard
methods for determination of biocide leaching rates from anti-fouling coatings, summarizes
recent work in this area and suggests an appropriate route for estimating the release rate of
biocides from anti-fouling coatings to the environment as required in annex 3 (3)(a) of
AFS Convention.

Action requested of the Committee

6 The Committee is invited to note the information on methods for determination of


biocides leaching rates from anti-fouling systems and, in particular, the suggested mass-balance
calculation method with suitable conservative default correction factors.

***

I:\MEPC\60\13.doc
MEPC 60/13

ANNEX

THE GENERATION OF BIOCIDE LEACHING RATE ESTIMATES FOR


ANTI-FOULING COATINGS AND THEIR USE IN THE DEVELOPMENT OF
PROPOSALS TO AMEND ANNEX 1 OF THE AFS CONVENTION

PRINCIPAL RELEASE RATE DETERMINATION METHODS

Laboratory methods – ASTM and ISO standard methods

1 The existing ISO and ASTM laboratory methods for determining the biocide release rate
can be considered as modifications of an earlier ASTM method for determining organotin release
rates (ASTM, 1990). All use the same generic rotating cylinder method, irrespective of the
biocide being studied (ISO, 2007a). In summary, triplicate test cylinders (6.4 cm diameter,
17.5 cm length) painted with a test-coating (typically 200 cm2 painted area) are immersed in a
holding tank containing standardised artificial seawater under controlled conditions of
temperature (24 to 26 oC), salinity (33 to 34 ppt), and pH (7.9 to 8.1). On specified measurement
days, each cylinder is transferred to a cylindrical vessel containing 1.5 litres of artificial seawater
and rotated at a surface velocity of 0.2 m sec-1 (about 0.4 knots) for up to 1 hour. The cylinders
are then returned to the holding tank until the next measurement point and the concentration of
biocide in the artificial seawater in the vessel at the end of the rotation period is determined by
chemical analysis. Testing is continued twice-weekly for a minimum of 45 days and may be
continued thereafter with once-weekly testing. From knowledge of the painted surface area, the
rotation time, and the resulting biocide concentration after rotation, the biocide release rate at
each measurement point can be determined. These individual results are arithmetically treated to
estimate the cumulative amount of biocide released over the test period (units: µg of biocide cm-2),
the average leaching rate from day 21 to day 45 (units: µg of biocide cm-2 d-1) and, if the trial is
continued beyond 45 days, from day 21 to the end of the trial. Additionally, if the coating
exhibits a ‘pseudo-steady state’ (defined as a 24-day minimum period to the end of the trial
during which the release rate varies by less than 15%), then the pseudo-steady state release rate is
also estimated.

2 At the present time, international standards using the rotating cylinder method have been
published by ASTM and/or ISO for organotins (ASTM, 1990), copper (ASTM, 2006;
ISO, 2007b), dichloro-N-octylisothiazolone (DCOIT), copper pyrithione, zinc pyrithione and
N-cyclopropyl-N’-(1,1-dimethylethyl)-6-(methylthio)-1,3,5-triazine-2,4-diamine (cybutryn)
(ASTM, 2007), zineb (ISO, 2007c), pyridine triphenylborane (ISO, 2008a), and dichlofluanid
and tolylfluanid (ISO, 2008b). Additionally a draft international standard method for tralopyril
has recently been published by ISO (ISO, 2008c). The analytical protocols for the published
methods, which avoid filtration prior to analysis, provide a measure of the total biocide release
(i.e. dissolved and particulate release).

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 2

Table 1 : Published and draft international standard methods for estimating biocide
release rate via the rotating cylinder laboratory method (all methods share a
common method of generating leachate and differ only in the analytical methods
required to quantify the amount of biocide in the leachate). Note that the ASTM
and ISO methods for copper are fully technically equivalent.

Method Description Status


ISO 15181 Generation of leachate Revised method published 2007
– Part 1 (generic rotating cylinder method) (fully harmonised with ASTM
– Part 2 Analysis for copper in leachate D6442-06)
– Part 3 Analysis for zineb in leachate Published 2007
– Part 4 Analysis of pyridine triphenylborane Published 2008
– Part 5 Analysis of dichlofluanid and tolylfluanidPublished 2008
– Part 6 Analysis of tralopyril in leachate In progress
ASTM D5108-90 Organotin release rate Published 1990
ASTM D6442-06 Copper release rate Revised 2006
(= ISO 15181Parts 1 & 2)
ASTM D6903-07 DCOIT, zinc pyrithione, copper pyrithione Published 2007
& cybutryn release rate

3 The rotating cylinder laboratory method was primarily developed in order to screen
experimental coatings during the manufacturer’s product development process and not to provide
reliable in-service estimates for use in environmental risk assessments or for regulatory purposes.
The test method is conducted over a relatively short period under closely controlled water
conditions and was never designed to reflect the actual biocide release rate from coatings on
ships for multi-year periods in natural waters with widely varying pH, temperature, and salinity.
In fact, Haslbeck & Ellor (2005) have shown that it is impossible for the copper release rates that
are measured by the ASTM/ISO method to reflect actual in-service release rates as they would
lead to complete exhaustion of the paint long before the end of the specified in-service paint
lifetime. In recognition of the general discrepancy between the rotating cylinder results and the
release rate to the aquatic environment, each of the ASTM and ISO copper and co-biocide
methods explicitly state that “The results of this test method do not reflect environmental
[biocide] release rates for anti-fouling products, and are not suitable for direct use in the process
of generating environmental risk assessments, environmental loading estimates, or for
establishing release rate limits for regulatory purposes.” This discrepancy has previously been
recognized by the US Environmental Protection Agency (EPA) which has acknowledged that
release rate results generated using the ASTM method “are not expected, or intended, to
correlate with actual, in-service leach rates” (Naval Sea Systems Command, U.S. Department of
the Navy, Office of Water – U.S. Environmental Protection Agency, 2003).

4 It is also notable that two ring trials have been conducted on the copper release rate
method to establish interlaboratory reproducibility by ISO in 1998 (Arias, 1999), and by ASTM
in 2001 (Haslbeck & Holm, 2005). Both trials showed that interlaboratory reproducibility was
relatively poor, and ranged from 24% to 81% depending on the coating type. No data is
available for the interlaboratory reproducibility of the method for other biocides.

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 3

FIELD METHODS

The Space and Naval Surface Warfare Center, San Diego (SSCSD) Dome

5 The SSCSD has developed a field method for measuring biocide release rates in situ
using a closed recirculating system (Seligman & Neumeister, 1983) called the ‘dome method’.
Using divers operating from a pilot boat, a polycarbonate dome (30.5 cm diameter) is filled with
ambient natural seawater and attached directly to an immersed anti-fouling paint film by light
suction in such a way that any biofilm growing on the surface of the paint film remains
essentially undisturbed. Water samples are periodically drawn off over a 1 hour period. The
dome is removed after sampling and analysis of the water samples allows the instantaneous
biocide release rate to be determined. No set sampling schedule is prescribed and the method has
been used to measure organotin and copper leaching rates either as single-point measurements or
through periodic testing over periods of up to several years directly from the hulls of ships
(Lieberman et al., 1985; Valkirs et al., 1994; Zirino & Seligman, 2002) and recreational vessels
(Schiff et al., 2004). When used to determine the copper release rate, the analytical protocol of
the published studies (filtration prior to analysis) provides a measure of the release rate of
dissolved copper, although this is not an inherent limitation of the method.

6 In contrast to the ASTM/ISO rotating cylinder method, the dome method was specifically
designed to allow in situ measurements of biocide release rates directly from the hull of a ship
under environmentally realistic conditions while in-service. The release rate would be expected
to vary from sampling location to sampling location on the vessel hull because of the differing
local hydrodynamic conditions, as well as differing biological conditions (e.g., the type and
degree of biofilm growth). Therefore provided that several samples are taken from different
locations on a vessel, the resulting average release rate can be considered to be a realistic and
reliable estimate of the actual release rate to the environment under the conditions existing at the
time of the measurement.

7 In a development of the original ship-hull dome method, domes have also been used to
measure the copper release rate from test-panels attached to a stationary exposure rack that is
immersed in natural waters (Valkirs et al., 2003). Measurements taken from smaller test panels
immersed in natural seawater from exposure racks are likely to be a less accurate reflection of
true environmental inputs than measurements taken directly from the hull of a vessel because of
the different hydrodynamic exposure conditions. In particular, ships and boats move whereas
immersed exposure racks are always stationary, and so panel testing of this kind necessarily
excludes the effect of vessel activity on the release rate.

8 Attempts by the US Navy to further develop the dome method are ongoing with the aim
of improving its practicality and utility (Earley, 2008). However, at present, the use of the dome
method has been largely restricted to the US Navy and its associated agencies and no studies
have been published for biocides other than organotins and copper. The practical, logistical and
economic challenges associated with its use (in particular the need to use divers) remain, and so
the dome method is not suitable for adoption by international standards bodies as a ‘standard test
method’ or for widespread use.

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 4

Chugoku Marine Paint’s On-site Sampler

9 A new developmental method for measuring the release rate directly from ship hulls has
recently been described by workers at Chugoku Marine Paints in Japan (Honma et al., 2008).
In this method, a plastic cylinder equipped with a motorised paddle-stirrer is attached by
electromagnets to a ship’s hull after it enters dry-dock. The cylinder (approximately 10 cm
diameter, 510 ml volume) is filled with artificial seawater, stirred for 1 hour and quantification of
the amount of biocide in the water is then expressed as a release rate. The release rate
measurement cannot be made while the ship is in the water but it is done as soon as practical
after the dry-dock is drained. In principle, the determined release rate should therefore closely
correlate with the release rate to the environment before the dry-dock was drained.

10 The method has so far been used to measure the release rate of copper from a single paint
applied to several different vessels. The simultaneous use of several ‘on-site samplers’ allows
replicate measurements to be made. The method and apparatus is still at the early stages of
development and further work will be required to demonstrate its utility and understand its
limitations. However, it is expected that the method could potentially be further extended in
future to determine the release rate of other biocides although, unlike the dome method, it is not
anticipated that this method will allow direct in-water measurements from ship hulls.

Laboratory/field combination methods – Harbour exposed panels (HEP) (static and dynamic)

11 SSCSD has also developed a combined laboratory/field method known as the Harbour
Exposed Panel (HEP) method, whereby 45 x 45 cm test-panels are immersed in natural waters
from a floating platform or raft under tidal flow conditions. The panels are periodically
transported to a laboratory where they are immersed in a measurement chamber containing a
known volume of fresh natural seawater (typically 8 or 11.5 litres depending on the size of the
panel) for up to 6 hours under gentle agitation using a rotating cam device. The panels are then
returned to the platform/raft and re-immersed.

12 In a modification to this general procedure, smaller curved test-panels are used which
alternate between periods of immersion in natural waters in tidal flow, and dynamic rotation at a
peripheral speed of 9.25 m sec-1 (17 knots) using a rotating drum apparatus. The panels are
periodically transferred to the laboratory for release rate measurement using a similar protocol to
that used for tidal flow-only immersion. No set sampling schedule is prescribed and the method
has been used to measure copper leaching rates over several years under both tidal flow-only and
alternating tidal/dynamic immersion conditions (Lindner, 1993; Valkirs et al., 2003).

Calculation Methods

ISO Mass-balance calculation method

13 The mass-balance calculation method for determining biocide leaching rates was
originally developed by the European Council of the Paint, Printing Ink and Artists’ Colours
Industry (CEPE)(CEPE, 2003). It is currently being further developed by ISO with a view to its
publication as an international standard method. The ISO mass-balance calculation method
provides a generic empirical model of biocide release and is based on the underlying fact that the
total amount of biocide released by a coating cannot exceed the amount of biocide that was
present when the coating was manufactured and applied. The inputs to the model are the
specified service lifetime of the paint (months), the amount of biocide in the coating formulation
(% by weight), the weight fraction of active ingredient in the biocide, the volume solids of the
I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 5

wet paint (%), the density of the wet paint, the dry film thickness of paint applied for the
specified lifetime (micrometres), and the fraction of active ingredient in the dry film released
during the specified lifetime of the paint.

14 The original CEPE model was based on a substantial industry dataset of organotin and
copper release from different paint types that had been generated using the ASTM/ISO rotating
cylinder method. This showed that, for most anti-fouling paints, the release rate would typically
fall sharply over the first 14 days or so of testing before settling to a relatively constant rate for
the remainder of the test period (usually 45 days). To reflect this, the CEPE model assumed that
the release rate falls linearly over the first 14 days in service, and thereafter remains constant for
the remainder of the specified lifetime, such that the ratio of X, the total amount of biocide
released during the first 14 days, to Y, the release rate during the remainder of the specified
lifetime, is 30 (CEPE, 2003, Finnie, 2006). However, it is important to note that these
assumptions are not inherent in the mathematics of the mass-balance model. In essence, the core
calculation divides the estimated total amount of biocide that can be released from the coating by
the specified lifetime of the coating to give an estimate of the average release rate over the full
coating lifetime. Additionally, in a refinement of the original CEPE model by ISO, calculation of
the amount of biocide released in the first 14 days has been separated from the calculation of the
average release rate over the remaining paint lifetime. In reality, the ISO mass-balance
calculation is therefore wholly independent of whether the release rate remains constant, falls,
rises or varies in any regular or irregular way over full specified lifetime of the coating, including
the first 14 days.

15 Based on the experience of the industry that anti-fouling paints are generally not
exhausted at the end of their specified lifetime, CEPE originally recommended that a value of 0.7
is used for the fraction of active ingredient in the dry film released during the specified lifetime.
In contrast, the position of European regulatory authorities has been that a more conservative
value of 0.9 should be used (i.e. this assumes that 90% of the biocide in the paint is released over
its specified lifetime) and this value has been incorporated as a further refinement of the method
by ISO. As a result, the calculated estimate of the total amount of biocide released by the coating
over its lifetime is a realistic worst-case for the total amount of biocide released and so the
calculated average release rate value should also be considered as the maximum possible,
realistic worst-case average release rate over the lifetime of the paint (ISO, 2009).

16 The reliability of the mass-balance calculation will, of course, be strongly influenced by


the reliability of all the input parameters used in the calculation. In particular, it is important that
the input values for the specified service lifetime of the paint and the dry film thickness of paint
applied for that specified lifetime properly reflect the pattern of use for any given anti-fouling
product. Nevertheless, because the mass-balance calculation model is independent of how the
release rate changes over the paint lifetime, its validity is not limited to paints and biocides that
show the same ‘typical behaviour’ as copper and organotins. The ISO mass-balance calculation
of the maximum possible average release rate over the lifetime of the paint is therefore inherently
valid for all biocides and all anti-fouling paint types.

17 The ISO mass-balance calculation method is in the final stages of the ISO standards
development process and it is anticipated that the method will be formally published in final form
in 2010.

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 6

Polishing rate mass-balance method

18 An alternative method of calculation is to relate the biocide release rate to the in-service
polishing rate for the anti-fouling paint (Haslbeck & Ellor, 2005). Where the volume fraction of
biocide in the paint film is known, then the release rate of active ingredient can be simply
calculated if the polishing rate (i.e. the rate of loss of film thickness, sometimes known as the
erosion or ablation rate) is also known. The volume fraction of biocide in the paint film can be
simply calculated if the weight fraction of biocide in the coating formulation is known, along
with the weight fraction of active ingredient in the biocide, the volume solids of the wet
paint (%), and the specific gravity of the wet paint. Unlike the other laboratory, field and
calculation methods discussed here, which are applicable to all anti-fouling paint types, the
polishing rate-mass balance method can only be used with anti-fouling coatings that undergo
systematic in-service reduction in film thickness. The use of this method is therefore restricted to
self-polishing, erodible/ablative and hybrid anti-fouling coatings. The resulting calculated value
can be considered to be the theoretical average release rate for the biocide over the entire lifetime
of the coating, although this does not take account of short-term changes in release rate due to the
vessel’s sailing pattern, etc.

19 In a closely related method, panels with a known film thickness of anti-fouling paint have
been applied to a ship’s bilge-keel while in dry-dock (Kojima et al., 2007). The panels were
recovered when the vessel re-entered dry-dock at the end of its in-service period and the copper
release rate can be inferred from the measured thickness and copper content of the intact residual
paint.

METHOD-DEPENDENCE OF BIOCIDE RELEASE RATE ESTIMATES

20 A statistical comparison of the estimated biocide release rate for a variety of laboratory,
field and calculation methods has shown that the choice of method has a highly significant effect
on the magnitude of the resulting release rate estimate (Finnie, 2006). To summarise, this paper
reviewed the available data for the determined copper release rate for InterspeedTM 640
(BRA640), a widely studied ablative soluble matrix anti-fouling paint manufactured by
International Paints Ltd. which contains cuprous oxide. The results of nine separate ASTM/ISO
rotating cylinder trials were compared with the results from 5 dome studies on ships while
pier-side, which covered in-service periods of up to 758 days and where each study was the result
of between 5 and 9 individual measurements per ship. These results were also compared against
the results of the CEPE calculation method, the release rate range estimated from the known
typical in-service polishing rate for this paint, single measurements for harbour exposed panel
testing under static/dynamic cycling and static-only conditions, and dome measurements taken
from panels immersed in harbour waters on a static test rack.

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 7

Figure 1 : Comparison of mean release rate data for InterspeedTM 640 determined using
different methods (error bars for ASTM/ISO data and for Dome data represent
95% confidence intervals and n indicates the number of measurements; error
bars for polishing rate data represent the calculated range; no bars indicates a
single measurement); data from Finnie (2006).

70

n=8
60
copper release rate (µg cm-2 d-1)

50

40

30

20

10
n=5

ASTM/ISO Calculation HEP Calculation HEP Dome Dome


(CEPE) (static/dynamic) (polishing rate) (static) (ships) (panels)

21 As shown in figure 1, this study found that the value of the determined copper release rate
for InterspeedTM 640 is ranked in the order ASTM/ISO rotating cylinder >> Mass-balance
calculation ≈ HEP (static/dynamic) ≈ Polishing rate calculation > HEP (static) > Dome (ships)
> Dome (racked test panels). Similar qualitative trends had previously been reported for
organotin release rate data (Schatzberg, 1996; Lewis & Baran, 1993) and also for copper
(Haslbeck & Ellor, 2005).

22 Importantly, the differences between the release rate values estimated by different
methods in figure 1 are statistically significant and can be precisely quantified. Of these
methods, it should be remembered that the dome measurements from ships are the only direct
measurements of the rate of release of biocide to the aquatic environment. These results
demonstrated that the ASTM/ISO rotating cylinder method typically overestimates the release
rate of copper to the environment from these ships by about an order of magnitude and, from a
statistical analysis of the variance in the results, it was concluded that the minimum
overestimation is by a factor of at least 5.4. These results also demonstrated that the mass-
balance calculation method generally provides an almost 4-fold overestimate of the
environmental copper release rate from these ships, and from a similar statistical analysis of the
variance, it was concluded that the minimum overestimation is by a factor of at least 2.9.

23 Finnie (2006) also compared the available results from these methods for a series of coatings
which included representatives of most of the main classes of anti-fouling paints, namely insoluble
matrix paints (1 example), ablative soluble matrix paints (2 additional examples), and tin-free

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 8

SPC products (2 examples). There was insufficient data to allow a statistically conclusive
comparison of the data for these paints but, without exception, the same qualitative
method-dependent trends were seen for these paints as were seen for the soluble matrix paint,
InterspeedTM 640. Wherever direct comparison with dome measurements of the environmental
copper release rates were possible, the limited available data consistently indicated that the
discrepancy between the ASTM/ISO method, the mass-balance calculation, and the environmental
copper release rate were at least as large as those seen for InterspeedTM 640 (see table 2).

Table 2 : Mean method-dependent copper release rates for a series of anti-fouling paints
showing the relationship between the ASTM/ISO rotating cylinder method, the
mass-balance calculation, and direct measurement of the release rate to the
aquatic environment via the SSCSD Dome (n/d = no data available); data from
Finnie (2006).

InterspeedTM Sol. matrix Insol matrix


Method 640 BRA540 prod. A prod. SPC prod. A SPC prod. B
ASTM/ISO 48.6 ± 13.9 58.8 87.0 131.2 100.0 66.1
Mass-balance
calculation 18.4 18.3 15.5 30.4 30.7 30.2
Dome (ship) 4.7 ± 1.7 4.6 3.9 5.7 n/d n/d
Dome (raft) 2.2 n/d n/d n/d 1.6 1.3

Ratio of n/d n/d


ASTM/ISO:
Dome 10.4 12.8 22.3 23.0
Ratio of n/d n/d
CEPE:Dome 3.9 4.0 4.0 5.3

24 It should be noted that the biocide release rate from an anti-fouling paint is generally
higher when a vessel is sailing than when it is moored (Kiil et al., 2002) but the calculated
release rate, which represents the maximum possible average release rate over the in-service
lifetime, provides no direct information on how the release rate changes with time. This is true
both in the long-term over the full paint lifetime and in the shorter term as the vessel alternates
between the periods in motion and period at rest. It is not surprising therefore, that the
mass-balance calculation significantly overestimates the environmental release rate for stationary
vessels.

IMPLICATIONS FOR ENVIRONMENTAL RISK AND EXPOSURE ASSESSMENT

25 When conducting an environmental risk or exposure assessment on an anti-fouling


product, it is a well established principle that the biocide release rate value that is selected for the
assessment should be representative for the scenario and risk assessment case in question
(OECD, 2005). However, an explicit conclusion of the statistical comparison of the biocide
release rate estimates from different methods was that neither the ASTM/ISO rotating cylinder
method nor the mass-balance calculation method provides copper release rate values that are
representative of the release rate from ships. It was therefore further concluded that neither
method gives release rate values that are appropriate for direct use in environmental risk
assessments, environmental loading estimates or establishing release rate limits for regulatory
purposes (Finnie, 2006).

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 9

26 This leads to a problem for environmental risk and exposure assessment: how should
representative release rate values be routinely obtained if the available practical methods do not
provide representative values, and the best available method for directly measuring representative
environmental release rates is impractical for routine use? In an attempt to resolve this paradox,
it has been proposed that suitable correction factors should be applied to ASTM/ISO rotating
cylinder and ISO mass-balance release rate values in order to provide more reliable and
representative estimates of the release rate to the aquatic environment.

27 The aforementioned analysis showed that the rotating cylinder method and mass-balance
calculation overestimate the rate of copper release to the aquatic environment from pier-side
vessels for a widely studied ablative soluble matrix paint by a factor of at least 5.4 and at least 2.9
respectively. Indicative data for other paints (SPCs, insoluble matrix and other soluble matrix
paints) shows that these methods are likely to overestimate the environmental copper and
organotin release rate from such vessels by at least a similar amount. Although only statistically
reliable comparative data for copper was available, it was argued that the same physical and
chemical processes underpin the release of co-biocides as for copper. It was therefore proposed
that 5.4 and 2.9 are suitable for use as conservative default correction factors that should be
applied to rotating cylinder and mass-balance release rate data for any anti-fouling product where
direct measurements of the environmental release rate are absent (Finnie, 2006). Clearly, as
these correction factors are based on a comparison of release rates from pier-side vessels, their
use is most valid in related emission scenarios, such as commercial harbours and marinas.

28 This principle of using correction factors has since been accepted by regulatory
authorities in the European Union for Active Substance authorisation under the Biocidal Products
Directive. The EU regulatory authorities have agreed that where environmental exposure
assessments for any Active Substance are carried out on the basis of mass-balance calculation
release rate data, then a correction factor of 2.9 can be applied for the marina scenario if
uncorrected data indicates potential adverse environmental effects. A weight of evidence
approach will then be adopted to enable a decision on Active Substance authorisation to be made
(European Chemicals Board, 2006).

29 While the EU have accepted the principle of a universal default value for all biocides and
all anti-fouling paints, it is recognized that there has hitherto been little substantive data available
for biocides other than copper and, to a more limited extent, organotins. In part to address this, a
recent study concurrently determined the release rate of copper and co-biocides from a series of
anti-fouling paints (Finnie, 2008). In this study, the release rate of four self-polishing copolymer
paints, each containing a mixed biocide system of cuprous oxide and either copper pyrithione or
zinc pyrithione, was characterised over a 1 year period using the relevant published international
standard rotating cylinder methods. At each measurement point the leachate sample was
concurrently analyzed for copper and for the co-biocide. The four SPC paints were considered to
be representative examples of this type of paint which is in widespread commercial use.

30 In summary, while the absolute copper release rate was invariably higher than the
absolute co-biocide release rate for these paints, a statistical evaluation of the results showed that
relative copper and co-biocide release rates closely reflected the relative contents of copper and
co-biocide in the paints, as shown in figure 2. Expressing this as the ratio of the relative copper
and co-biocide release rates to the relative copper and co-biocide contents of the paint, this ratio
remains consistently close to 1 over the 1 year test period. Although the correlation is less close
in the first 14 days, this ratio ranged from 0.91 to 1.09 for the four paints from day 21 onwards
with a relative standard deviation (RSD) of less than 20%, with an overall mean value of 1.02
and a mean RSD of less than 17% (see table 3).
I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 10

Figure 2 : Illustrative copper and co-biocide release rate data concurrently determined
by the published standard rotating cylinder methods; release rates have
been normalised relative to the respective biocide contents of the paint
(arbritrary release rate units); data from Finnie (2008).

PAINT A

5
Normalised release rate

Cu
3
CuPY

0
0 50 100 150 200 250 300 350
time (days)

31 Even though the water-solubilities of copper pyrithione and zinc pyrithione are
respectively about one and three orders of magnitude higher than that of cuprous oxide
(Health and Safety Executive, 1999; 2003; 2005), there was no preferential progressive release of
the more soluble biocide from the mixed biocide paint films and there was no discernible
difference between the release rate behaviours of the more soluble and the less soluble biocides
over the 1 year test period.

Table 3 : Summary statistics for the variation of the ratio of the relative copper and
co-biocide release rates to the relative copper and co-biocide content of the paint
(= (Cu:CoB)norm). The 162 and 190 day datapoints for Paint C have been
excluded as outliers (linked to an observation of small paint flakes); data taken
from Finnie (2008).

0-14 days 21-360 days


Paint Co-biocide Mean Mean St deviation RSD (%)
(Cu:CoB)norm (Cu:CoB)norm
A CuPT 2.32 1.05 0.13 12.7
B CuPT 2.15 0.91 0.17 18.3
C CuPT 0.87 1.09 0.21 19.4
D ZnPT 1.55 1.04 0.18 17.4

Mean for 1.72 1.02 0.17 16.9


Paints A-D

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 11

32 These results demonstrate that the paint film functions as a controlled release medium for
the biocides that are present, and it is the overall properties if the paint film rather than the properties
of the biocides themselves that control the biocide release characteristics. These release rate
characteristics will be independent of whether the paint film is immersed under laboratory or
in-service conditions, although the magnitude of the determined release rate will of course be
dependent on the selected test-method. In other words, the same synchronous release rate
behaviour will also be shown for the release of biocides to the aquatic environment from a ship hull.

33 As this behaviour was common to all the paints and all the biocides in this test-series,
then it is expected that the same general synchronous release rate behaviour would also be
exhibited by other paints of this type and by other mixed biocide systems. The extended duration
of this trial also indicates that the synchronous biocide release behaviour for these mixed biocide
systems will continue to be maintained over the full in-service lifetime of the anti-fouling paints.
The implications of this for environmental risk and exposure assessment are that it validates the
use of the same correction factors for copper and for co-biocides for such paints where this
approach is applied to ASTM/ISO rotating cylinder or mass-balance calculation data.

34 Although the results for these paints cannot definitively demonstrate that similar
synchronous biocide release behaviour will always be show by all other paints types and all
potential mixed biocide combinations, similar synchronous biocide release behaviour can be
expected unless the paint incorporates an active transport mechanism to facilitate the preferential
loss of one biocide in a mixed-biocide system. The results for these paints therefore supports the
general use of the same correction factors for all paint types and all biocides to allow the most
reliable and representative estimates of the environmental release rate to be made where direct
measurements of the environmental release rate are absent.

35 Another interesting aspect of this extended ASTM/ISO rotating cylinder study was that
the measured long term release rate is substantially lower than the release rate over the
first 45 days (the normal length of a trial of this type). In this case, the 21-45 day mean release
rate in the final month was 3-4 times higher than the mean release rate in the final month of the
one-year trial (see figure 2). These results are consistent with other studies on different paints
which have shown similar trends (Haslbeck & Ellis, 2005; Honma et al, 2008; Forsberg &
Haglund, 2008). The routine extension of rotating cylinder trials to some arbitrary point in time
beyond the required minimum period could, in theory, appear to be a way to experimentally
generate data that more closely reflects the release rate to the environment. However, the costs
involved in such trials are substantial, particularly when the requirement to generate data
according to Good Laboratory Practice standards if it is to be used to support product
authorizations is taken into account: the longer the trial, then the higher the cost.

36 Perhaps more importantly, extended testing in this study over 1 year led to a 3-4 fold
reduction in the measured release rate compared with the short term 21-45 day average, the
discrepancy between the short term average and the environmental release rate is typically about
an order of magnitude for pier-side vessels (Finnie, 2006; Valkirs et al., 2003). Hence, even a 1
year test period does not directly produce environmentally relevant release rate estimates from
the ASTM /ISO rotating cylinder method. As yet, any recommendation of a preferred extended
duration for the ASTM/ISO rotating cylinder method would be purely arbitrary and would still
require a suitable revised correction factor to be identified and applied to the results to enable
representative estimates of the release rate to the environment to be obtained. The application of
the proposed conservative default correction factor to short-term data is therefore currently
considered to be preferred approach whenever rotating cylinder data is used.

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 12

CONCLUSION
37 It is generally recognized that the reliability of any environmental exposure or risk
assessment is limited by the reliability of the input data. Biocide release rate data is a key input
to the environmental exposure and risk assessment process and so it is vital that the estimated
biocide release that is used is both accurate and representative of the release rate to the
environment under the scenario and risk assessment case in question (OECD, 2005). The
ongoing work to further develop the dome method and the recent work in Japan to develop new
methods to generate environmentally pertinent biocide release rate data are to be warmly
welcomed. At the present time though, there is unfortunately no practical method currently
available for the routine direct measurement of the biocide release rate to the environment from
ship hulls and so other indirect methods must be used.

38 Of these other methods, the most widely used laboratory experimental method is the
ASTM/ISO rotating cylinder method and a number of published international standard methods
are available for many of the commonly used copper-based and organic co-biocides. However,
all available data suggests that the ASTM/ISO rotating cylinder method does not provide
accurate or representative release rate values. Results indicate that the method typically
overestimates the environmental release rate by an order of magnitude, and consequently the
published international standard methods explicitly state that they are unsuitable for direct use for
the purposes of environmental risk assessment.

39 In order to generate representative estimates of the environmental release rate by the


ASTM/ISO rotating cylinder method, it is necessary to apply a suitable correction factor. Based
on a statistical analysis of the available data, a default correction factor of 5.4 has been proposed
as a conservative realistic worst-case for commercial harbour and marina scenarios
(Finnie, 2006). Recent work has shown that the release rate of the different biocides in certain
mixed biocide systems is closely coupled and that, even though the different component biocides
in the paint may have wildly differing water solubilities, neither biocide is preferentially released
(Finnie, 2008). By inference, this is expected to be the general case rather than the exception and
this supports the use of this default correction factor for all paints and all biocides. This default
correction factor should therefore be routinely applied to ASTM/ISO rotating cylinder data
whenever such data is used in relation to environmental exposure or risk assessment.

40 An alternative to the rotating cylinder method is the mass-balance calculation model


originally developed by CEPE and now being refined and further developed by ISO. Unlike the
rotating cylinder method, which normally only characterises the release rate over a short period
(typically 45 days), the ISO mass-balance calculation model provides an estimate of the maximum
possible average release rate over the full lifetime of the paint (which may be up to 5 or more years).
The underlying mathematics of this calculation means that it is universally applicable to all biocides
and all paints with the proviso that the input values (particularly for film thickness and paint lifetime)
are representative of actual use. Suitable correction factors are therefore also required whenever
the ISO mass-balance calculation method is used for environmental risk assessment purposes.
Based on a statistical analysis of the available data, a default correction factor of 2.9 has been
proposed as a conservative realistic worst-case for commercial harbour and marina scenarios
(Finnie, 2006). Again, this default correction factor should be routinely applied whenever ISO
mass-balance calculation data is used in relation to environmental exposure or risk assessment.

41 Environmental risk assessment guidelines generally indicate that experimental data are
usually preferable to calculation data (OECD, 2005). In relation to biocide release rate data for
anti-fouling products, in the absence of a direct measurement of the release rate to the aquatic
environment from a ship or boat under the emission scenario and location of concern, this would
I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 13

normally mean that experimental ASTM/ISO rotating cylinder data generated in the laboratory
would be preferred to ISO mass-balance calculation data. As has been discussed above, the
discrepancy between the release rate measured via the rotating cylinder method and direct in situ
measurement of the release rate from ship hulls is significantly greater than the discrepancy
between mass-balance calculations and direct in situ measurements. Consequently, in this case,
there is no obvious compelling reason to prefer the experimental laboratory data over the
calculation data, particularly as both require the application of suitable correction factors in order
to generate representative environmental release rate estimates.

42 Moreover, the ASTM/ISO rotating cylinder methods are expensive and time consuming to
perform, require specialist equipment and analytical chemistry skills, and are poorly reproducible.
In addition, methods are not available for all of the biocides in current use, and new methods would
be required to be developed for any new biocides that are introduced in future. In contrast, the
ISO mass-balance calculation is inexpensive, quick and simple to perform in a fully transparent
manner and is intrinsically reproducible. It is also universally and immediately applicable to all
current and future biocides and anti-fouling paint types. As both methods require the application
of suitable correction factors to provide representative estimates of the release rate to the
environment, employment of the ISO mass-balance calculation method clearly makes best use of
resources and avoids unnecessary costs. Given this, it is concluded that the mass-balance
calculation method, when used in conjunction with suitable conservative default correction
factors, is the most appropriate route to generate representative biocide release rate estimates for
anti-fouling products. This approach enables more accurate environmental risk and exposure
evaluations to be made, allowing any decisions on the restriction of anti-fouling products to be
made with greater confidence, without subjecting the aquatic environment to greater risk.

REFERENCES

Almeida E, Diamantino TC, de Sousa O. 2007. Marine paints: the particular case of anti-fouling
paints. Progress in Organic Coatings 59:2-20.

Anderson C, Atlar M, Callow M, Candries M, Townsin RL. 2003. The development of


foul-release coatings for seagoing vessels. J Marine Design Operations 84: 11-23.

Anon. 1952. Marine Fouling and its Prevention, Woods Hole Oceanographic Institute, United
States Naval Institute, Annapolis. 388 p.

Arias S. 1999. Round robin test for anti-fouling paints. European Coatings Journal.
March 1999: 122-128.

ASTM. 1990. Standard test method for organotin release rates of anti-fouling coating systems in
sea water, ASTM Method D 5108-90, 6 p.

ASTM. 2006. Standard test method for determination of copper release rate from anti-fouling
coatings in substitute ocean water, ASTM Method D 6442-06, 10 p.

ASTM. 2007. Standard test method for determination of organic biocide release rate from anti-
fouling coatings in substitute ocean water, ASTM Method D 6903-07, 21 p.

CEPE. 2003. Provision of biocide leaching rate data for anti-fouling products: a discussion
document from the Anti-Fouling Working Group of CEPE. In: Emission scenario document on
anti-fouling products: Annex, Environment Directorate of the organisation for Economic
Co-operation and Development, 2005. Available from http://www.oecd.org.
I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 14

Chambers LD, Stokes KR, Walsh FC, Wood RJK. 2006. Modern approaches to marine
anti-fouling coatings. Surface & Coatings Technology 201: 3642–3652.

Earley P. 2009. Personal communication.

European Chemicals Board. 2006. Harmonisation of leaching rate determination for anti-fouling
products under the biocidal products directive, Workshop Report, Ispra, Italy, 12 December 2006.

European Commission. 2004. European Commission Director-General Environment,


Harmonisation of environmental emission scenarios: an emission scenario document
for anti-fouling products in OECD countries, final report. Available from
http://ecb.jrc.it/Documents/Biocides/ENVIRONMENTAL_ EMISSION_SCENARIOS.

Finnie AA. 2006. Improved estimates of environmental copper release rates from anti-fouling
products. Bio-fouling 22(5): 279 – 291.

Finnie AA. 2008. Copper and co-biocide release from anti-fouling paints and implications for
environmental risk assessment. 14th International Congress on Marine Corrosion and Fouling,
Kobe, Japan, 27-31 July 2008, Paper 29A-4-1.

Finnie AA, Williams DN. 2009. Paint and coatings technology for the control of marine fouling.
In: Bio-fouling, (Eds Dürr S & Thomason JC). Wiley-Blackwell. Oxford, Chapter 13 (in press).

Forsberg A, Haglund, K. 2007. Prolonged copper release rate study with three pleasure boat
anti-fouling paints, Nordic Council of Ministers TemaNord report 2008:518, 40 p.

Honma M, Kiseki Y, Nishi T, Teragaki H. 2008. On-site sampler for measuring release rate of
antifoulant from ship’s bottom. 14th International Congress on Marine Corrosion and Fouling,
Kobe, Japan, 27-31 July 2008, Paper 29A-4-2.

Haslbeck E, Ellor JA. 2005. Investigating tests for antifoulants: variation between laboratory
and in-situ methods for determining copper release rates from navy-approved coatings.
J. Protect. Coat Linings. August 2005: 34-44.

Haslbeck EG, Holm ER. 2005. Standard methods: tests on leaching rates from anti-fouling
coatings reveal high level of variation in results among laboratories. European Coatings Journal.
October 2005: 26-31.

Health and Safety Executive. 2003. Evaluation on zinc pyrithione: use as a booster biocide in
anti-fouling products. Advisory Committee on Pesticides evaluation of fully approved or
provisionally approved products. Issue No 208, 172 p.

Health and Safety Executive. 2005. Evaluation on copper pyrithione as a new active ingredient
in professional anti-fouling products. Advisory Committee on Pesticides evaluation of fully
approved or provisionally approved products. Issue No 223, 181 p.

Health and Safety Executive. 2003. Evaluation on copper compounds, 1st evaluation of their use
in anti-fouling products. Advisory Committee on Pesticides evaluation of fully approved or
provisionally approved products. Issue No 183, 151 p.

I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 15

ISO. 2007a. Determination of the release rate of biocides from anti-fouling paints – Part 1:
General method for extraction of biocides. International Standard ISO 15181-1: 2007.

ISO. 2007b. Determination of the release rate of biocides from anti-fouling paints – Part 2:
Determination of copper-ion concentration in the extract and calculation of the release rate.
International Standard ISO 15181-2: 2007.

ISO. 2007c. Determination of the release rate of biocides from anti-fouling paints – Part 3:
Calculation of the zinc ethylenebis(dithiocarbamate) (zineb) release rate by determination of the
concentration of ethylenethiourea in the extract. International Standard ISO 15181-3: 2007.

ISO. 2008a. Determination of the release rate of biocides from anti-fouling paints – Part 4:
Determination of pyridine-triphenylborane (PTPB) concentration in the extract and calculation of
the release rate. International Standard ISO 15181-4: 2008.

ISO. 2008b. Determination of the release rate of biocides from anti-fouling paints – Part 5:
Calculation of the tolylfluanid and dichlofluanid release rate by determination of the
concentration of dimethyltolylsulfamide (DMST) and dimethylphenylsulfamide (DMSA) in the
extract. International Standard ISO 15181-5: 2008.

ISO. 2008c. Determination of the release rate of biocides from anti-fouling paints – Part 6:
Determination of tralopyril release rate by quantitation of its degradation product in the extract.
International Standard ISO/CD 15181-6.

ISO. 2009. International Standard ISO/DIS 10890.2.

Kiil S, Dam-Johansen K, Weinell CE, Pedersen MS, Codolar SA. 2002. Dynamic simulations
of a self-polishing anti-fouling paint exposed to seawater. Journal of Coatings Technology
74 (929): 45-54.

Kojima R, Miyata O, Shibata T, Senda T, Shibata K. 2007. Leaching phenomena of anti-fouling


agents from ships’ hull paints. Proceedings of the International Symposium Shipbuilding
Technology 2007, September 6-7 2007, Osaka, Japan, pp85-92.

Lewis JA, Baran IJ. 1993. Biocide release rates from anti-fouling coatings. Oebalia 19: 449-456.

Lieberman SH, Homer V, Seligman PF. 1985. In-situ determination of butyltin release rates
from anti-fouling coatings on navy test ships. Naval Oceans Center Technical Report 1027. 43 p.
Available from U.S. Department of Commerce National Technical Information Service,
Springfield, VA 22161.

Lindner E. 1993. Dynamic and static exposure tests and evaluations of alternative copper-based
antifoulant coatings. Surface Warfare Systems San Diego Technical Report 1628. 96 p. Available
from U.S. Department of Commerce National Technical Information Service, Springfield, VA 22161.

Naval Sea Systems Command, U.S. Department of the Navy, Office of Water – U.S. Environmental
Protection Agency. 2003. Draft characterization analysis report hull coating leachate, August 2003.
Available from http://www.epa.gov/owow/oceans/regulatory/unds/hullcoating.

OECD. 2005. Emission scenario document on anti-fouling products. OECD Environmental


Health and Safety Publications Series on Emission Scenario Documents, No. 13, 166p.
I:\MEPC\60\13.doc
MEPC 60/13
ANNEX
Page 16

Omae I. 2003. General aspects of tin-free anti-fouling paints. Chem. Rev. 103: 3431-3448.

Schatzberg P. 1996. Measurement and significance of the release rate for tributyltin.
In: Organotin, MA Champ and PF Seligman editors, Chapman Hall, London, p 383-403.

Schiff K, Diehl D, Valkirs A. 2004. Copper emissions from anti-fouling paint on recreational
vessels. Marine Pollution Bulletin 48: 371-377.

Schultz MP. 2007. Effects of coating roughness and bio-fouling on ship resistance and
powering. Bio-fouling 23:5 331-341.

Seligman PF, Neumeister JW inventors; The United States of America as represented by the
Secretary of the Navy, assignee. In-situ leach measuring system, US Patent 4375451, 1983 March 1.

Stein J, Truby K, Darkangelo-Wood C, Takemori M, Vallance M, Swain G, Kavanagh C,


Kovach B, Schultz M, Wiebe D, Holm E, Montemarano J, Wendt D, Smith C, Meyer A. 2003.
Structure-property relationships of silicone bio-fouling-release coatings: effect of silicone
network architecture on pseudo-barnacle attachment strengths. Bio-fouling 19: 87-94.

Thomas KV. 2001. The environmental fate and behaviour of anti-fouling paint booster biocides:
a review. Bio-fouling 17: 73-86.

Turley PA, Fenn RJ, Ritter JC, Callow ME. 2005. Pyrithiones as antifoulants: environmental
fate and loss of toxicity. Bio-fouling 21: 31-40.

Valkirs AO, Davison BM, Kear LL, Fransham RL, Zirino AR, Grovhoug JG. 1994.
Environmental effects from in-water cleaning of ablative copper anti-fouling coatings. Naval
Command Control and Ocean Surveillance Center RDT&E Division San Diego, Technical
Document 2662. 82 p. Available from U.S. Department of Commerce National Technical
Information Service, Springfield, VA 22161.

Valkirs AO, Seligman PF, Haslbeck E, Caso JS. 2003. Measurement of copper release rates
from anti-fouling paint under laboratory and in situ conditions: implications for loading
estimation to marine water bodies. Marine Pollution Bulletin 46: 763-779.

Van Hattum B, Baart A, Boon J. 2005. Emission estimation and chemical fate modelling of
antifoulants. In: The Handbook of Environmental Chemistry, Volume 5, Springer-Verlag.

Yebra DM, Kiil S, Dam-Johansen K. 2004. Anti-fouling technology – past, present and future
steps towards efficient and environmentally friendly anti-fouling coatings. Progress in Organic
Coatings 50: 75-104.

Zirino A, Seligman PF. 2002. Copper chemistry, toxicity, and bioavailability and its
relationship to regulation in the marine environment: Office of Naval Research Workshop
Report, Surface Warfare Systems San Diego Technical Document 3140. 97 p. Available from
U.S. Department of Commerce National Technical Information Service, Springfield, VA 22161.

___________

I:\MEPC\60\13.doc

Vous aimerez peut-être aussi