Ballast Water Sampling and Sample Analysis for Compliance Control - Global Maritime Transport and Ballast Water Management: Issues and Solutions - page 217

authorities. However, scientifi c studies have shown that sampling results may be
biased by different sampling approaches because of, e.g., the patchy distribution of
organisms in tanks, mortality of organisms during sampling etc. The lack of a com-
monly agreed ballast water sampling methodology or approach may impact repre-
sentative ballast water sampling so that certain vessels may be found in compliance
in one port, but not in another.
Different methods and sampling equipment may be used for different sampling
goals, e.g., D-1 or D-2 standards, indicative or detailed sampling. The selection of
appropriate sampling methods and equipment also depends on the ballast water
access points, i.e., in-tank via manholes, sounding pipes or air vents, or in-line
installed sampling points, and also on the target groups of organisms as stated in the
D-2 standard.
It is of prime importance to evaluate the appropriateness of a selected sampling
approach for compliance control according to the BWM Convention. The ballast
water sampling methods recommended here for compliance with the D-1 and D-2
standards are based on the author's sampling experience gained on more than 80
shipboard tests for type approval of 18 different BWMS. Very importantly, these
sampling methods were also scientifi cally tested, improved and validated during
three studies on representative BWS for compliance monitoring. These studies were
conducted for the Federal Maritime and Hydrographic Agency, Hamburg, Germany
in 2009 and 2012, and for the European Maritime Safety Agency, Lisbon, Portugal
in 2010 (Gollasch and David 2009 , 2010a , b , 2013 ).
For D-2 standard compliance tests it is suggested that samples should be taken
during discharge, i.e., from the ballast water discharge line after the pump prior to
the discharge overboard. This approach delivers the most representative and accu-
rate results regarding the organism concentration in the ballast water discharge from
a vessel as a side stream of the discharge is sampled. Samples should be taken in
two or more 10 min sampling sequences with a sample volume of 300-500 l in each
sequence. With this approach, organisms are less exposed to negative impacts dur-
ing sampling. Further, such a sampling approach is logistically more appropriate
than sampling over the entire discharge time, and our scientifi c studies have shown
this sampling method to be representative of the whole discharge. In reality, this
sampling method as per Guidelines G2 is not exactly the same as for the type
approval testing of BWMS according to Guidelines G8, but has all components of
it, e.g., the same sampling equipment may be used for the two or three recom-
mended sequential samples. This method is also in-line with the agreement at IMO
that sampling methods applied by PSC for compliance checks, i.e., G2 Guidelines
sampling, should be no more stringent than the methods applied for BWMS type
approval, i.e., G8 Guidelines sampling.
However, one key problem remains with the in-line sampling approaches and
this is that compliance or non-compliance can only be proven during discharge
while the ballast water is being pumped overboard. Consequently, the ballast water
may already have been discharged before it is clear whether or not it is in compli-
ance with the BWM Convention standards. Should high risk organisms be suspected
in the ballast water intended to be discharged, in-tank compliance control sampling