5.2.5 Aeration
If a 250mL analysis reads low, a common reason is
because the glass frit was not at the bottom of the
bottle, but set for a 40mL vial, thus incompletely
aerating the 250mL sample. Care should be taken to
check that the frit is close to the bottom of the vial. If
a mistake is discovered a
ft
er the system is properly set
up, it is permissible to allow some aeration to take
place before the WAT250 measurement is started.
Th
at way, the water can be partly aerated, with some
of the radon already in the closed loop at the start of
the test, allowing a more complete overall aeration
a
ft
er the
fi
ve-minutes of aeration in the
fi
rst 5
minutes of the WAT250 measurement. Section 6.5
provides instructions on how to begin aerating a
water sample in advance of the actual test.
5.2.6 Counting Time
Longer counting times improve sensitivity and
precision by accumulating a greater total number
counts above background, which gives more
favorable counting statistics. Increasing the usual 20
minute count time to 80 minutes (4 times 20) will
improve counting statistics by a factor of 2 (square
root of 4). For this to work, however, it is necessary
that the RAD7 be thoroughly dried out, so that the
relative humidity does not climb too high during the
80 minutes of count time. It is possible, during a
measurement, to set the pump from GRAB to ON,
which will turn it on, thus moving air through the
desiccant and into the RAD7. When the relative
humidity is down once more, the pump must be set
back to GRAB.
5.2.7 Temperature
Th
e temperature e
ff
ect on accuracy is very small with
the 40mL sample vial, but may begin to become
noticeable with the 250mL vial at very low or high
temperatures.
Th
e RAD H
2
O system has been
calibrated for a sample analysis temperature of 20° C
(68° F). At colder temperatures the water "holds back"
a little more of the radon during the aeration process,
and at warmer temperatures the water "gives up" the
radon more readily.
Th
e maximum temperature e
ff
ect at equilibrium for
the 40mL sample is about +/-1% over the range of 0
to 40°C (32° to 104° F).
Th
e maximum temperature
e
ff
ect at equilibrium for the 250mL sample is about
+/-6% over the same range.
5.2.8 Relative Humidity
If the RAD7 is thoroughly dried out before use, the
relative humidity inside the instrument will stay
below 10% for the entire 30 minutes of the
measurement. If not, then the humidity will rise
during the 25 minutes that the RAD7 is counting and
the pump is stopped, and may rise above 10% before
the end of the measurement period. High humidity
reduces the e
ffi
ciency of collection of the
polonium-218 atoms, formed when radon decays
inside the chamber. At around 60% humidity, the
collection e
ffi
ciency may be only half that at 10%
relative humidity or below. However, the 3.05 minute
half life of polonium-218 means that almost all the
decays that are actually counted come from atoms
deposited in the
fi
rst 20 minutes of the measurement.
So a rise in humidity above 10% over the last ten
minutes of the counting period will not have a
signi
fi
cant e
ff
ect on the accuracy of the result.
If the
fi
rst two counting periods are below 10%
relative humidity, you may ignore humidity e
ff
ects.
On the other hand, if the humidity rises above 10%
before the end of the
fi
rst counting cycle, there will be
an error whose size is indeterminate. However, you
can be sure that any error due to high humidity will
be in a direction to reduce the reading, so that the
true value must be higher than the observed value.
For accurate readings, the RAD7 should be dried out
thoroughly before making the measurement (See
section 1.4.1). In addition, when viewing RAD7 data
using the CAPTURE so
ft
ware, the Correct For
Humidity checkbox may be checked to compensate
for the reduced e
ffi
ciency of the RAD7 at higher
humidity levels.
5.2.9 Background E
ff
ects
By paying careful attention to details, you can reduce
the background in the RAD H
2
O system to
insigni
fi
cant levels. We previously discussed how to
control the background due to purge air radon
content and residual radon and its progeny.
Th
e
uncontrollable, or "intrinsic", background of the
RAD7 is low enough to ignore in all but the most
demanding cases.
Th
e intrinsic background of the
RAD7 is less than 1 count per hour, corresponding to
a 40mL water sample concentration of less than 2
pCi/L (even lower for the 250mL sample). In
principle, you can achieve a background this low if
you completely eliminate all radon and progeny from
the system before a measurement, but that will
Section 5
Accuracy and Quality Control
22
