Jökull - 01.12.2007, Page 78

Thorsteinsson et al.
sediment or cells derived from 15 ml aliquots from
each sample were cultured on PCA agar at 30!C for 3
days.
Significant microbial growth was observed in all
samples (Table 2), but those that had been exposed to
the UV radiation (samples 3–5) contained far fewer
cultivable cells than the other samples. The lowest
cell counts were found in a sample collected from the
drill stem. The growth at 30!C is probably due to
mesophilic contaminates, whereas the low tempera-
ture growth indicates the presence of psychrotolerant
bacteria in the snow (and hence in all samples taken
from the drilling system during operations).
Additional testing of the sterilization efficiency
of the drilling system was carried out in two sepa-
rate tests at the NEA workshop during spring 2006
(Table 3). The system was filled with 70% ethanol
2 hours before the start of each of these two tests.
The heater within the high pressure pump (unit 2D)
was now run at maximum temperature (150!C), and
the water/steam mixture emerging at the drill tip was
found to be at 99!C.
The snow used in Test 1 (March 8) was collected
near Reykjavík. For Test 2 (May 22) no snow could
be obtained and the snowmelter was thus filled with
tap water. In Test 2, the water was deliberately spiked
with a known number of viable Escherichia coli bac-
teria and samples were collected when the concentra-
tion of bacteria in the system had reached 3.2!104
cells/ml. Standard methods accredited by the Ice-
landic standard ÍST EN ISO/IEC 17025 and New
York State Department of Health, Environmental Lab-
oratory Approval Program (ELAP) were used in the
biological analysis of the samples collected during
these tests.
In NEA Test 1, culturing at 3!C showed no ev-
idence for growth of psychrotolerant bacteria in the
snow sample or the samples collected from the sys-
tem. Relatively few mesophilic contaminants, indi-
cated by growth at 22!C, were found in the system.
The number of mesophilic bacteria decreased slightly
after in-line filtering and dropped significantly after
UV radiation. In this test and the Langjökull test,
the mesophiles likely originated from the snowmelter
which was not washed with 70% ethanol prior to the
tests like other parts of the system. In the test with
spiked tap water (NEA Test 2), the number of viable
cells was dramatically attenuated by UV and reduced
to undetectable levels by heating.
It is not straightforward to explain the difference
between the results shown in Table 3 and those ob-
tained in the Langjökull test. Two possible factors are
the higher temperature in the test performed at NEA
and the use of 70% ethanol (a more effective disin-
fectant than the 95% solution used in the Langjökull
test) in the system before the start of operations. The
absence of psychrotolerant bacteria in the snow col-
lected near Reykjavík (Test 1) is also unexplained.
Chemical analysis of drilling water
During the drill test on May 22 (Test 2 in Table 3),
samples were collected from the system for chem-
ical analysis to determine the background concen-
trations of ions released into the meltwater by drill
system components. The samples were taken from
the water container (filled with tap water), from the
sampling location S1 (Figure 1) and from the drill
stem outlet. The water samples were filtered on
site through 0.2µm Teflon filters into polypropylene
bottles. Samples for determination of cation and
trace metal concentrations were acidified (Teflon dis-
tilled Suprapur!, Merck), but samples for anion con-
stituents were not treated. Amber glass bottles were
used to collect samples for determination of pH and
total dissolved CO2. To prevent any organic growth
or decay, samples for the determination of nutrients
were frozen (-18!C) within hours of collection and
kept frozen until analyzed. The concentrations of ma-
jor cations and trace metals were analysed using ICP-
AES and ICP-SFMS techniques, respectively, and the
concentrations of major anions and nutrients were
analysed using RF-IC. The results are given in Ta-
ble 4.
The composition of the water did not change sig-
nificantly as it was run from the snowmelter through
the drill system. The concentrations of major ele-
ments, most trace metals as well as nutrients are, es-
sentially, the same in the container, at the inlet and
outlet of the drill system. A few metals are, however,
slightly enriched at the inlet and outlet of the drill sys-
tem compared to the snowmelter, including Zn, Pb,
76 JÖKULL No. 57