Determination of Radon in Drinking Water by
Liquid Scintillation Counting
Method 913.0
May 1991
DRAFT
Paul B. Hahn
Stephen H. Pia
Radioanalysis Branch
Nuclear Radiation Assessment Division
Environmental Monitoring Systems Laboratory
U. S. Environmental Protection Agency
Las Vegas, Nevada 89119
-------
DRAFT
Determination of Radon in Drinking Water by
Liquid Scintillation Counting
Method 913.0
Scope and Application
1.1 The following procedure is designed for the analysis
of radon in drinking water supplies from ground water
and surface water sources. Application of this
analytical procedure to matrices other than drinking
water have not been studied. Caution is advised if
this procedure is used for matrices other than
drinking water.
1.2 Diffusion of radon is affected by temperature and
pressure. Samples at other than laboratory
temperature should be allowed to equilibrate to room
temperature before processing.
1.3 Radon will be desorbed from water if the sample is
subject to turbulent flow or reduced pressure as is
encountered when using pipet tips or syringes with a
small orifice followed by a large fill space. Losses
can be minimized by using pipet tips with the tip
opening enlarged to permit laminar flow of sample
into and out of the tip fill space.
1.4 Precision and bias of the method is affected by the
background in the energy window used for analysis. A
procedure is provided for selection of the analytical
window which minimizes the background contribution to
measurement.
2. Summary of Method
2.1 In this procedure 10 mL of the water sample are
transferred into a 20 mL glass scintillation vial to
which has been added 10 mL of mineral oil
scintillation cocktail. Radon diffuses from the
sample into mineral oil for which it has a much
greater affinity than for water. The sample is then
equilibrated for three hours, and then counted for
50 minutes in a liquid scintillation counter using
energy discrimination for alpha particles. Results
are reported as pCiIL.
1
-------
_____ DRAFT
3. Interferences
3.1 There are no known chemical interferences from
species found in drinking water nor from the dilute
concentration of acid which may be present in the
standards. Uranium, radium or other radioactive
elements would cause positive bias if present in
significant quantities.
4. Safety
4.1 There are no unusual hazards associated with the
reagents used in this procedure. Routine safety
precautions ie: lab coat, plastic gloves, safety
glasses and use of a hood are required when
transferring samples and standards and preparing
standards when radium solutions are used.
5. Apparatus and Equipment
5.1 Pipet - 5mL mechanical pipet. Gas-tight syringe is
not recommended because of the large vacuum produced
when filling.
5.2 Scintillation cocktail dispenser - Adjustable to
deliver 10 mL.
5.3 Liquid scintillation counter - A system permitting
spectral analysis is recommended.
6. Reagents and Consumable Supplies
6.1 Scintillation vials — 20 mL glass vials with caps.
6.2 Mineral oil scintillator - High efficiency mineral oil
cocktail or equivalent.
6.3 Concentrated Nitric Acid - Reagent grade.
6.4 Water - Demineralized or equivalent.
6.5 Radium solution — Two sources are required for
calibration and check standards.
6.6 Sample collection bottles - 125 mL narrow mouth with
teflon-lined caps.
7. Sample Collection, Preservation and Shipment
7.1 Samples are collected from a non-aerated faucet or
spigot to minimize desorption of radon as follows.
2
-------
DRAFT
7.2 Fill a 1 liter glass beaker with water from the
source and allow to gently overflow. Open the sample
container and submerge inverted into the beaker.
7.3 Rotate the sample container so that it fills with
water. Cap with a teflon-lined cap while the sample
container is still under water to eliminate headspace.
7.4 Analysis should begin within three days after receipt
of sample in the laboratory.
8. Calibration and Standardization
8.1 Radium Solution Method
8.1.1 Prepare 100 mL of radium-226 in water standard
such that the final activity will be
approximately 35,000 pCi/L. Radium-226
traceable to the National Institute for
Standards and Technology (NIST) can be obtained
from the U.S. Environmental Protection Agency,
Environmental Monitoring Systems Laboratory,
NRA/STD, P.O. Box 93478, Las Vegas, Nevada 89193.
8.1.2 To a 100 mL beaker add 20 mL of water, 0.5 mL
of concentrated HNO 3 and place on an
analytical balance. Record the initial mass.
8.1.3 Dispense with a pasteur pipet, or other
suitable dropper, the required mass of radium
solution into the beaker and record the mass.
The actual mass of radium solution added is
obtained by difference of the final and
initial weights.
8.1.4 Quantitatively transfer the solution into a
100 mL volumetric flask with a stream of water
and dilute to the mark.
8.1.5 Transfer 10.0 mL of the diluted standard into
the scintillation vial to which has been added
10 mL of mineral oil cocktail. Prepare at
least three standards and three backgrounds
using a distilled water sample.
8.1.6 Set the standards and background samples aside
for at least 30 days to allow radon to attain
secular equilibrium.
8.1.7 Determine the optimal analytical window as
outlined in: Selecting Optimal Window, below.
8.1.8 After 30 days let the sample dark adapt for 3
hours if necessary and count for 50 minutes.
3
-------
DRAFT
Repeat the counting two additional times.
8.1.9 From the pooled results calculate a system
calibration factor CF as cpm/pCi by the
following expression:
CF = S - B
Cxv
Where:
S = Counting rate of standard (mm 1 )
B Counting rate of background (mm ’)
C = Concentration of radium-226 standard
(pCi/L)
and V = Volume of standard used (eg. O.O100L)
8.2 Radon Generator Method
8.2.1 The radon generator consists of radium-226
strongly bound to a small amount of ion
exchange resin in a 60 milliliter bottle of
water. The radium remains fixed to the resin
while the radon formed by radioactive decay
diffuses into the water phase. The transfer
is practically quantitative and the generator
can be refilled and reused repeatedly after
ingrowth of radon to secular equilibrium
(30 days)
8.2.2 Agitate the radon in water solution by slowly
rotating the bottle and allowing the resin
beads to mix the solution. Avoid using
generators containing an air bubble.
8.2.3 Allow the resin beads to settle, remove cap,
and carefully pipet two 5-mL aliquots of
solution into a 20 mL glass scintillation vial
to which has been added 10 mL of mineral oil
cocktail. Prepare three replicates from each
of the generators. Shake the vial to extract
the radon into the organic phase.
8.2.4 Determine the optimal analytical window as
outlined in: Selecting Optimal Window, below.
8.2.5 Let the standards equilibrate for at least
three hours and count the samples for 50
minutes. Repeat the counting two additional
times. Prepare and count three background
samples.
4
-------
DRAFT
8.2.6 Correct the results to account for radon decay
occurring between sample preparation and the
midpoint of the counting period. From the
pooled results calculate a system calibration
factor as cpm/pCi by the following expression:
CF= SB
CxVx D
Where:
S and B are the standard and background
counting rates
C = The equilibrium radon-in-water
concentration of the source
V = The volume of the water transferred
and D = The decay correction factor (ext)
8.3 Selecting Optimal Window
8.3.1 The following procedure is for use with counters
which permit analysis of sample spectra.
8.3.2 Count a radon standard for five minutes and
generate a sample spectrum. For greater
clarity use a log scale for the channel number
if possible.
8.3.3 The region of greatest alpha activity will be
obvious as evidenced by one or two large peaks
at the high end of the energy spectrum (see
Figure 1) . The optimal window is formed by
extending the region by 10% on each side of
the alpha peaks. This window will be used for
subsequent calibration and analysis. The
calibration factor should be at least 6 cpm/pCi
with the background not exceeding 10 cpm.
9. Quality Control
9.1 Background Samples
One background sample having approximately the same
total dissolved solids as the samples is run with
every twenty samples. At least two backgrounds
should be counted with each batch.
A suitable background sample may be prepared by
boiling 2 L of laboratory radium— and uranium-free
tap water to remove residual radon if present. Store
the cooled tap water in a capped 2 L bottle.
5
-------
DRAFT
Channel Number
Figure 1. Liquid Scintillation Sprectrum of Radon
Cl)
-a
C
0
C-)
2500
2000
1500
1000
500
0
0 100 200 300 400 500 600 700
800 900 1000
6
-------
DRAFT
9.2 Duplicates
Duplicate analyses are required for one out of every
ten samples not including the standards arid quality
control check standards.
Duplicate analyses should have a relative percent
difference (RPD) equal to or less than the percent
2 sigma (%2o) counting error plus 10 percent.
Relative percent difference is calculated by taking
the absolute value of the following expression:
RPD = ( Analysis 1 - Analysis 2) x 200
Analysis 1 + Analysis 2
If the RPD exceeds this value, recount the samples.
If the new RPD still exceeds %2 , + 10 percent, prepare
another duplicate sample or duplicate low-level (100 -
300 pCi/L) standard and analyze. If the RPD for the
reanalysis exceeds the limit, check for instrument
problems or possible losses in transferring the water
to the scintillation vial.
9.3 Quality Control Check Sources (QCCS)
QCCS are prepared from a dilution of radium different
than that used to prepare the standards and should
have a nominal activity of 5000 pCi/L. Three check
sources are analyzed with each batch of samples. One
check source is run with the first ten samples, the
second is placed approximately in the middle of the
batch and the last is run with the last ten samples
of the batch.
9.4 Records
Collect and maintain the results from the duplicate
pairs and check standards in a bound notebook. The
record should include the date, the results, the
analyst and any comments relevant to the evaluation
of these data.
Plot the results of the check standards on a control
chart for that counter.
10. Procedure
10.1 Carefully pipet a 10-mL or two 5-mL aliquots of
sample into a glass scintillation vial to which has
been added 10 niL of mineral oil cocktail.
10.2 Cap and shake the samples and set aside in the dark
for at least three hours to equilibrate the radon
progeny and dark adapt before counting. The time of
7
-------
DRAFT
sample collection is the initial time for decay
correction.
10.3 Count the samples for 50 minutes using the optimized
window settings for alpha counting.
11. Calculations
11.1 Calculate the concentration of radon-222 in pCi/L
from the following equation:
pCi/L = CF x D x V
Where:
G = Gross counting rate of the sample
B = Background counting rate
CF = Calibration factor (see 8.1.9)
and D = Decay factor for Rn-222 between time of
collection and midpoint of counting period
11.2 Calculate the 2 sigma (95% confidence level) counting
uncertainty from:
(
2 - --+ --
2 = B
CF x D x V
Where:
TG = The duration of the sample count
and P 5 = The duration of the background count
12. Precision and Accuracy
12.1 The intralaboratory and interlaboratory precision and
accuracy of the method has not been fully evaluated.
12.2 At concentrations near zero the counting uncertainty
will be the primary factor in the lack of precision
of the method. Figure 2 plots the percent 2-sigma
(95% confidence level) counting uncertainty as
functions of concentrations and counting times.
12.3 Interlaboratory differences are estimated to be on
the order of 25% based on a 12-laboratory study
using the Ra-226/resin radon generators. Figure 3
plots percent recoveries (measured/expected values)
8
-------
DRAFT
Counting Time Minutes
Figure 2.
Counting Uncertainty as a Function of
Concentration and Counting Time.
w
V.)
a)
0
>
C
a)
C-)
C
C
C
0
0
E
C /)
1
1 10 100
1000
9
-------
DRAFT
Percent Recovery at 500 pCi/L Rn222
Figure 3.
Analytical Recoveries for 12-Laboratory
Collaborative Study
c J
c. 1
-J
0
0
0
>1
0)
>
0
C.)
a)
C
a)
C.)
1
a)
0.
150
10
-------
DRAFT
for a 100 pCi/L sample versus a 500 pCi/L sample for
each of the 12 laboratories. The four laboratories
biased low at the 100 pCi/L level are suspected of
using an open window.
Bibliography
1. Whittaker, E.L, J.D. Akridge and J. Giovino, Two Test
Procedures for Radon in Drinking Water, U.S. Environmental
Protection Agency, EPA/600/2—87/082 (1987)
2. Vitz, E., Toward a Standard Method for Determining
Waterborne Radon, Health Physics, 60:817 (1991)
3. Lowry, J.D., Measuring Low Radon Levels in Drinking Water
Supplies, J. Amer. Water Works Assoc., 4/1991:149 (1991).
4. Prichard, H.M. and T.F. Gesell, Rapid Measurements of Rn-222
Concentrations in Water with a Common Liquid Scintillation
Counter. Health Physics J. 33:577 (1977).
5. Youden, W.J., and E.H. Steiner, Statistical Manual of the
Association of Official Analytical Chemists, 1975.
11
------- |