SWRHL-77
RADIUM-226 IN SELECTED PUBLIC WATER
SUPPLIES - A COMPARATIVE STUDY
by
Robert N. Snelling and Richard E. Jaquish
Technical Services Program
Western Environmental Research Laboratory
ENVIRONMENTAL PROTECTION AGENCY
Published June 1971
-------
SWRHL-77
RADIUM-226 IN SELECTED PUBLIC WATER
SUPPLIES - A COMPARATIVE STUDY
by
Robert N. Spelling and Richard E. Jaquish
Technical Services Program
Western Environmental Research Laboratory*
ENVIRONMENTAL PROTECTION AGENCY
Published June 1971
*Formerly Southwestern Radiological Health Laboratory, part of U. S.
Department of Health, Education, and Welfare, Public Health Service,
Environmental Health Service, Environmental Control Acliiv n1 strati on,
Bureau of Radiological Health
-------
ABSTRACT
The objective of this study was to resample selected public water
supplies which from a previous survey had exhibited elevated radium-
226 concentrations, and to provide a more definitive picture of the
radium concentrations in these supplies. Water samples were collected
from the selected supplies on a monthly basis over the period of one
year. Comparison of data from the two surveys indicated that the
results from the initial survey were not in most cases representative
of the water system as a whole.
-------
ACKNOWLEDGEMENTS
Sincere appreciation is expressed to the State Health Departments
of Colorado, Minnesota, Oklahoma, and South Dakota for the coopera-
tion and assistance provided during this study. Without this
cooperation, the study could not have been adequately performed.
-------
TABLE OF CONTENTS
Page
ABSTRACT i
ACKNOWLEDGEMENTS 1i
BACKGROUND 1
PROJECT OBJECTIVE - 1969-1970 SURVEY 2
SAMPLING AND ANALYTICAL PROCEDURES 4
RESULTS 4
DISCUSSION 6
SUMMARY 11
REFERENCES 14
LIST OF TABLES
Table Page
I. Elevated Radium-226 Concentrations Observed in Public 3
Water Supplies, 19660)
II. Radium-226 in Selected Public Water Supplies March 1969 - 5
April 1970
III. Comparison of Representative Radium-226 Concentrations for 13
Selected Water Supplies
-------
Background
During 1966-1967 a project was conducted to sample and analyze for
radium-226 in approximately 900 public water supnlies serving
3.8 million people in 20 states."*'' The study was initiated by the
Protective Action Section (Radiation Bio-Effects Program, National
Center for Radiological Health) to locate population groups that
consumed water containing relatively high concentrations of radium.
These could be used for epidemiological studies on the effects of
chronic radium ingestion. Relatively high, in the context of this
project, was defined as levels of radium greater than 3 picocuries per
liter. This concentration of radium-226 is greater than those predom-
inant in the United States but lower than the microcurie quantities
associated with the radium dial painter studies.
To permit broad geographic coverage, utilizing the laboratory facilities
available, only a single one-liter sample was collected from each
community. However, certain selection criteria and sampling procedures
were employed which, hopefully, made the single sample more representa-
tive than it might have been otherwise. Results of this study indicated
that 41 supplies, serving 144,000 people in 10 states, contained more
than 3 pCi/liter radium-226. Of these, four supplies serving 3,000 people
were found to contain more than 10 pCi/liter. The highest concentration
observed was 24.1 pCi/liter.
-------
The original plans called for resampling those supplies identified
by the initial survey as having radium concentrations in excess of
10 pCi/1. Although the preliminary survey achieved its objective
of locating water supplies containing elevated radium levels, there
was still a need to determine if the preliminary results reflected
an accurate estimate of the radium content of the water consumed by
the total population in the communities over a meaningful period of
time. The report of this survey was published in the July 1968 issue
of Public Health Reports. ^
Project Objective - 1969-1970 Survey
In 1969 a survey was conducted by the Western Environmental Research
Laboratory (WERL) to resample selected water supplies which from the
1966 survey had exhibited elevated radium-226 concentrations. The
objective was to provide a more definitive picture of the radium concen-
tration in these supplies. This was particularly important where munici-
palities used multiple wells for their water supplies and only one well
was sampled during the 1966 survey.
Four public water supplies, identified in the 1966 survey as exhibiting
radium-226 concentrations greater than 10 pCi/1, were selected for re-
sampling. In addition, at the request of the South Dakota Health Depart-
ment, five locations were resampled in that state which had displayed
radium concentrations between 3 and 10 pCi/1 in the 1966 survey. At
the suggestion of the Colorado Health Department the water supply of
Rye, Colorado, was added to the study. A summary of the water supplies
-------
chosen for resampling along with the radium concentrations determined
in the 1966 sampling is shown in Table I.
TABLE I
ELEVATED RADIUM-226 CONCENTRATIONS
OBSERVED IN PUBLIC WATER SUPPLIES, 1966
STATE AND COMMUNITY
Colorado - Wiley
Colorado - Rye
Minnesota - Howard Lake
Minnesota - Montrose
Oklahoma - Afton
South Dakota - Canton
South Dakota - Edgemont
South Dakota - Hot Springs
South Dakota - Howard
South Dakota - Ipswich
ESTIMATED
POPULATION
400
600
1,000
400
1,250
2,500
1,770
4,950
1,200
1,130
226 Ra pCi/iiter|
15.7
9.72*
15.2
24.1
10.3
6.93
5.33
4.82
3.52
3.02
^Determined by Colorado Health Department, 1968
-------
Sampling and Analytical Procedures
A one-gallon water sample was obtained from each well. In addition,
if a community was served by more than two wells, a sample representa-
tive of the distribution system was obtained. The sampling was conducted
by the respective state health departments on a monthly basis for the
period of one year. Each sample was forwarded to the WERL for radium-
226 analysis. The radium analyses were performed by the radon emanation
technique.(2)
Particular attention was given to quality control procedures on the
analyses of samples from this project. The WERL is a regular partici-
pant in the Analytical Quality Control Service (AQCS) program of the
Environmental Protection Agency (EPA). During the period.of this survey
the AQCS cross-check results were all satisfactory. Twenty percent of
the samples were analyzed in duplicate. The standard deviations were
estimated from the duplicates with the following results.
Activity range - pCi/1 Standard deviation pCi/1
less than 1 0.035
1-5 0.39
5-10 1.49
10-15 1.90
The allowable standard deviation for radium-226 in water used by AQCS
is ± 15%.
Results
A summary of results by station location is found in "able II.
-------
TABLE II
RADIUM-226 IN SELECTED PUBLIC WATER SUPPLIES
MARCH 1969 - APRIL 1970
LOCATION
Colo., Rye
Rye
Colo. , Wiley
Minn. , Howard Lk.
Minn. , Montrose
Okla., Afton
S. D., Canton
Canton
Canton
Canton
S. D., Edgemont
Edgemont
Edgemont
Edgemont
S. D., Hot Springs
Hot Springs
S. D. , Howard
Howard
Howard
Howard
S. D., Ipswich
Ipswich
SAMPLING
POINT
Dist. Syst.
Well
Dist. Syst.
Well #2
Well #1
Dist. Syst.
Well #1
Well #4
Well #5
Dist. Syst.
Well #1
Well #2
Well #4
Dist. Syst.
New
Spring
Well #1
Well #3
Well #5
Dist. Syst.
North
South
NUMBER OF
SAMPLES
8
1
9
11
11
7
6
7
10
11
9
12
11
11
12
11
8
9
10
2
10
10
RADIUM-221
MEAN
7.5
120
4.44
16.9
12.2
8.56
0.56
0.94
0.64
0.78
4.22
4.15
6.46
3.76
0.37
0.24
3.45
2.28
0.25
1.30
1.32
3.30
5, PCi/l
RANGE
2.3- 25
—
3.4 - 5.3
13 - 21
8.7 - 15
6.9 - 9.6
0.4 - 0.7
0.7 - 1.2
0.2 - 1.0
0.4 - 0.9
3.6 - 4.5
3.5 - 4.7
2.9 - 21
2.5 - 4.4
0.3 - 0.5
0.1 - 0.3
2.1 - 4.7
0.2 - 3.4
0.1 - 0.5
0.6 - 2.4
1.1 - 1.5
2.1 - 4.3
-------
Discussion
Rye, Colorado
The Rye, Colorado, water supply is composed of two sources: Greenhorn
Creek and an 800-ft. well. The water from the two sources is mixed
at a treatment station (chlorination only) and held in a common storage
tank. The percentage of water contributed by the well varies from
approximately 50% in the summer to 10% or less in the winter. These
seasonal fluctuations are dependent on the condition of the Greenhorn
Creek watershed and are therefore highly variable. For this reason,
sampling was done from the distribution system.
The number of users is estimated to be approximately 600 persons.
A single sampling from the well in April, 1969, showed a radium-226
concentration of 120 pCi/1. Radium concentrations in the distribution
system averaged 7.50 pCi/1 and ranged from a minimum of 2.3 pCi/1 (100%
creek water) to a maximum of 25 pCi/1- The high variability observed is
the result of the fluctuating contribution of the well water.
Wiley, Colorado
The Wiley water supply consists of two wells, 500 feet and 700 feet deep,
respectively. The two wells together have a total capacity of 144,000
gallons per day and serve a population of approximately, 400 people. The
water is composited in an equalizing tank prior to distribution.
-------
All sampling was done from the distribution system and represents the
composited water from the two wells. The average radium-226 concentra-
tion for nine samples was 4.44 pCi/1. Individual samples ranged from
a minimum of 3.4 pCi/1 to a maximum of 5.3 pCi/1.
Afton, Oklahoma
The Afton water supply is produced from two 1,000-ft. wells operating
together at a combined capacity of 150,000 gallons per day. The
supply serves approximately 1250 people. Seven samples were obtained
from the distribution system over the sampling period and showed an
average radium-226 concentration of 8.56 pCi/1 with a range of 6.9 -
9.6 pCi/1.
Howard Lake, Minnesota
The Howard Lake water supply consisted of two wells until May 1969
when well #1 was abandoned. Prior to that time, 95% of the water was
supplied by well #2. This well is 1038 feet deep and produces 100,000
gallons per day serving approximately 1100 people. Because only the
single well is in use, water from that well is assumed to be representa-
tive of the distribution system.
Eleven samples were collected over the year. The radium-226 concentra-
tion ranged from 13 pCi/1 to 21 pCi/1 with an average of 16.9 pCi/1.
Montrose. Minnesota
The Montrose water supply consists of a single 690-ft. well which
produces 35,000 gallons per day. Approximately 400 people are served by
the system. Again, since only one well is used, water from that well
7
-------
is assumed to be representative of the entire distribution system.
Eleven samples were collected over the year averaging 12.2 pCi/1
of radium-226. The individual samples ranged from 8.7 pCi/1 to
15 pCi/1.
Canton, South Dakota
The Canton water supply consists of seven wells and serves a
population of 2500 people. Only three wells (No's 1, 4, 5) were
directly available for sampling, the others not having sampling taps.
The wells were all approximately 400 feet deep and ranged in capacity
from 70,000 - 200,000 gallons per day. The pumps are run in parallel
in varying combinations throughout the year. A routine sample was also
obtained at the water plant pump house which is considered to be most
representative of the water to the distribution system.
The average radium concentration of the three individual wells was
0.72 pCi/1. The total range of results for all three wells (23 samples)
was from a minimum of 0.2 pCi/1 to a maximum of 1.2 pCi/1. The average
radium-226 concentration of eleven samples taken from the distribution
system was 0.78 pCi/1 with a range of from 0.4 pCi/1 to 0.9 pCi/1. The
average of the three individual wells shows good agreement with the
average of the distribution system.
Edgemont, South Dakota
The Edgemont water system is supplied by four wells, each with a capacity
of approximately 500,000' gallons per day. Three of the wells (1, 2, 4)
8
-------
were sampled during the study along with a sample from the distribution
system. The population served by the system is 1700 people. The
average radium-226 concentration obtained from the three wells was
4.94 pCi/1 with a total range of from 2.9 pCi/1 to 21 pCi/1. The
eleven samples from the distribution system averaged 3.76 pCi/1 with
a range of from 2.5 pCi/1 to 4.4 pCi/1.
Hot Springs, South Dakota
The Hot Springs water supply comes from two wells; "New" well (860,000
gal/day) and "Spring" well (720,000 gal/day). The population served by
the system is approximately 4950 people. Samples were obtained from
the discharge of each well. Both wells contribute more or less equally
to the distribution system. Twelve samples obtained from the "New" well
averaged 0.37 pCi/1 of radium-226 with a range of from 0.3 to 0.5 pCi/1.
"Spring" well averaged 0.24 pCi/1 with a range of 0.1 to 0.3 pCi/1.
The radium-226 concentration most representative of the distribution
system is the average of the two individual wells. This average
calculates to 0.31 pCi/1.
Howard, South Dakota
Four wells are utilized for the Howard water supply. Three of these
(1, 3, 5) were available for sampling and were sampled during this study.
The wells have an average capacity of 120,000 gallons per day each. In
addition, two samples were obtained from the distribution system. The
population served by the supply is estimated at 1200 oeople. Wells
number 1 and 3 averaged 3.45 pCi/1 and 2.28 pCi/1 respectively while
well number 5 averaged 0.25 pCi/1. Since only two samples were obtained
9
-------
from the distribution system itself, the average radium-226 concentra-
tion of the three wells sampled is considered to be most representative
of the distribution system. This average is 1.99 pCi/1.
Ipswich, South Dakota
The Ipswich water supply consists of two wells; north well at 29,000
gallons per day and south well at 72,000 gallons per day. The popula-
tion served is 1100 people. Each well discharges into a unique
equalizing reservoir and is then pumped into the distribution system.
Each well was sampled at the discharge pipe.
The north well showed an average radium-226 concentration of 1.32
pCi/1 with a range of 1.1 pCi/1 to 1.5 pCi/1- The south well showed
an average concentration of 3.30 pCi/1 with a range of 2.1 pCi/1 to
4.3 pCi/1.
The radium concentration most representative of the distribution system
is the average of the two wells weighted according to their capacity.
The weighted average radium-226 concentration is determined to be 2.73
pCi/1.
10
-------
Summary
Table III shows a comparison of the radium concentrations derived
from the single sampling in 1966 with the average concentration
observed in the 1969 study. The values shown for 1969 are the
values considered most representative of the concentrations of
radium-226 in each community's water system during 1969. The
values reported for the 1966 survey were reportedly obtained from
the distribution systems and were assumed to be representative thereof.
No comparison is shown in Table III for Rye, Colorado, since this
community was not sampled in the 1966 survey.
It is noted that in only three cases (Howard and Ipswich, S. D., and
Howard Lake, Minn.) were the single sampling values of 1966 within
two standard deviations of the mean determined by the 1969 survey.
In two instances (Hot Springs, S. D., and Canton, S. D.), the single
sampling was greater than the average by an order of magnitude.
Although significant differences might be expected between a single
sample and the average of a sampling, differences of an order of
magnitude are difficult to explain.
It is not possible at this time to determine the reason for the diff-
erences in the two surveys. Samples from the 1966 survey are not
available and the laboratory that analyzed the samples is no longer in
existence. The discrepancy could have been due to differences in sampling
11
-------
technique or in the analyses. It can only be concluded that the
radium-226 concentration obtained from individual samples collected
and analyzed in 1966 cannot be considered representative of the water
supplies in those six communities where there was a wide discrepancy
between the two studies.
12
-------
TABLE III
COMPARISON OF REPRESENTATIVE RADIUM-226
CONCENTRATIONS FOR SELECTED WATER SUPPLIES
1966 vs. 1969
LOCATION
Colo., Wiley
*Minn., Howard
Lake
Minn., Montrose
Okla., Afton
S.D., Canton
S.D., Edgemont
S.D., Hot Springs
*S.D., Howard
*S.D., Ipswich
SAMPLING 226Ra, pCi/l*<
POINT, 1969 1966
Dist. Syst.
Well #2
Well #1
Dist. Syst.
Dist. Syst.
Dist. Syst.
Ave.t
Ave.t
Ave.t
15.7
15.2
24.1
10.3
6.93
5.33
4.82
3.52
3.02
" 226Ra, pCi/1, 1969
MEAN
4.44
16.9
12.2
8.56
0.78
3.76
0.31
1.99
2.73
±2S
1.16
5.68
3.78
1.66
0.30
1.26
0.13tt
1.58tt
0.96tt
t Ave. represents average of individual wells sampled
tt Determined from pooled variance
** Single sample from distribution system
* Single sample within ± 2S of x of 1969
13
-------
REFERENCES
1. Mickey, 0. L., S. D. Campbell. High Radium-226 Concentration
in Public Water Supplies, Public Health Reports, Vol. 83,
No. 7, pp. 551-558, July 1968.
2. Johns, F- B. Southwestern Radiological Health Laboratory,
Handbook of Radiochemical Analytical Methods, SWRHL-11,
March, 1970.
14
------- |