Groundwater Sampling Devices
21
and finally TNT. These results were not unexpected as USGS guidance
(Vroblesky 2001) recommends that this device be used only for a specified
list of VOCs.
Pesticide study
As expected with lower concentrations, the variability in this data set was
greater than in the previous study. The RSDs ranged from 3% to 22% for the
controls and all the devices except for the PDB samplers, where the RSDs ranged
from 29% to 138% (Table 5).
There were no statistically significant differences between concentrations in
the control samples vs. those taken with the HydraSleeve, Kabis Sampler,
Discrete Interval Sampler, or PneumoBailer (Table 4). Although there also was
no significant difference between the control values and those found in the PDB
samplers, mean concentrations in these samplers were substantially lower for two
analytes (27% for aldrin and 66% for dieldrin) after 26 days' contact. Again,
these findings support current guidance (Vroblesky 2001) that does not recom-
mend using the PDB samplers for these types of analytes.
Volatile organics studies
The variability in these data sets was generally low with RSDs less than 5%,
except for the late eluting PCE where RSDs ranged from 2 to 16% (Tables 6 and
7).
In both of these experiments, the Discrete Interval Sampler provided samples
that were not significantly different from the controls (Table 4), although in the
low-level study, loss of PCE was nearly 16% with this device.
In the low-level study, significantly lower concentrations of several analytes
(TCE, ODCB, MXYL) were found with the HydraSleeve. However, in all cases
these losses were minimal, less than 5%. There were no significant differences
between the controls and sampler concentrations in the high-level study.
The results for the Kabis Sampler were mixed. In some cases, concentrations
were significantly lower in the Kabis Sampler while in others the opposite was
true. However, all these differences were small (<5%) except for loss of TCE
(18%), which was one of the most volatile analytes in this study.
The results were also mixed for the PneumoBailer. There were statistically
significant differences for three analytes but these differences were very small,
less than 2%.
Previous Page
