require a fair amount of e
ff
ort and patience. A more
realistic background to shoot for in routine analysis
might be between 10 and 20 pCi/L. Remember, if you
know the background well enough, you can subtract
it o
ff
and have reasonable con
fi
dence in the result.
5.3 Comparison of RAD H
2
O with Other
Methods
Figure 7 provides a basis for comping di
ff
erent
methods of measuring radon in water samples.
Th
e
numbers are typical, and some laboratories may be
able to get better results than this table indicates,
while others may not.
Th
e precision
fi
gures include
counting statistics only, with no adjustment for
sampling variation or decay of the sample.
Note that standard laboratory analysis o
ft
en entails a
long delay between sampling and analysis, which can
signi
fi
cantly increase the error and raise the detection
limit (DL) and the lower limit of detection (LLD).
Also note that the background
fi
gure used to
calculate the RAD H
2
O precision, DL, and LLD is
conservatively estimated to re
fl
ect typical
fi
eld usage.
Th
e most demanding and patient RAD H
2
O operator
should be able to reduce background to less than 0.02
cpm (rather than the 0.10 cpm given in the table),
which will allow for DL's and LLD's lower than those
listed.
5.4 Quality Assurance
A proper quality assurance plan should follow the
guidelines set by the USEPA as described in [Goldin].
Compliance with future certi
fi
cation programs will
certainly require an approved quality assurance plan.
Th
e elements of a quality assurance plan include
blank samples, duplicate samples, and spiked
samples. O
ft
en, the plan provides for blind samples to
be measured in an inter-comparison program. If a
quality control measurement deviates beyond the
acceptable range, the operator must cease to make
measurements until the cause of the deviation has
been discovered and corrected.
Th
erefore, the quality
assurance plan should specify the range of acceptable
measurement deviations, o
ft
en in the form of a
"control chart".
Th
e operator should maintain
complete records of the quality control measurements
and their deviations.
Section 5
Accuracy and Quality Control
23
