I
EPA 520/1-75-002
ASSESSMENT OF POTENTIAL RADIOLOGICAL
POPULATION HEALTH EFFECTS FROM RADON IN
LIQUEFIED PETROLEUM GAS
1
^
II
WM
III
ill
i
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Radiation Programs
-------
ASSESSMENT OF POTENTIAL RADIOLOGICAL
POPULATION HEALTH EFFECTS FROM RADON IN
LIQUEFIED PETROLEUM GAS
Thomas F.Gesell
Raymond H. Johnson, Jr.
David E. Bernhardt
FEBRUARY 1977
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Radiation Programs
Washington, D.C. 20460
-------
FOREWORD
The Office of Radiation Programs carries out a national program
designed to evaluate the exposure of man to ionizing and nonionizing
radiation, and to promote development of controls necessary to pro-
tect the public health and safety and assure environmental quality.
Within the Office of Radiation Programs, the Environmental
Assessment Division conducts programs relating to sources and levels
of environmental radioactivity and the resulting population radiation .
dose. Results of these findings are published in appropriate scientific
journals and Division technical reports.
The technical reports of the Environmental Analysis Division allow
comprehensive and rapid publishing of the results of intramural and
contract projects. These reports are distributed to State and local
radiological health programs, universities, libraries, information ser-
vices, industry, hospitals, laboratories, technical and advisory committees
to the Office of Radiation Programs, the press, and other interested groups
and individuals. These reports are also included in the collections of
the Library of Congress and the National Technical Information Service.
Readers of these reports are encouraged to Inform the Office of
Radiation Programs of any omissions or errors. Comments or requests
for further information are also invited,
A
e^
W. D. Rowe, Ph.D.
Deputy Assistant Administrator
for Radiation Programs
iii
-------
ABSTRACT
Liquefied petroleum gas (LPG) contains varying amounts of radon-222
which becomes dispersed within homes when LPG is used in unvented appli-
ances. Radon-222 decays to alpha-emitting daughter products which are
associated with increased lung cancer when inhaled and deposited in the
respiratory system. The average dose equivalents to the bronchial
epithelium from the use of LPG in unvented kitchen ranges and space
heaters are estimated to be about 0.9 and 4.0 mrem/year, respectively.
When extrapolated to the United States population at risk, the estimated
tracheobronchial dose equivalents are about 20,000 and 10,000 person-rems/
year for these appliances, or a total of about 30,000 person-rems/year.
These doses are very small compared to other natural and man-made sources
of ionizing radiation. It is estimated that these low doses would result
in less than one lung cancer a year for the total U, S. population. Con-
sequently, the use of LPG containing radon-222 does not contribute signifi-
cantly to the incidence of lung cancer in the United States. Furthermore,
the cost for control of radon levels in LPG would be over $50 million for
reduction of one health effect, therefore it is concluded that a require-
ment for such controls would not be cost effective on a national basis.
This study did indicate that individual dose equivalents could possibly
exceed 500 mrem/year. However, existing data are not sufficient to
determine the significance of such potentially high individual doses.
ORP will be evaluating this matter further.
-------
ACKNOWLEDGMENTS
The assistance and cooperation of the following individuals
in preparing and reviewing this report is gratefully acknowledged,
Neal S. Nelson, D.V.M.
William A. Mills, Ph.D.
Floyd L. Galpin
Harry W. Galley
James M. Hardin
Bruce Mann
Off-ice of Radiation Programs
Environmental Protection Agency
Larry A. Franks, Ph.D.
E. G. & G»
Eddie W. Chew
El Paso Natural Gas Company
Charles Barton
Hollifield National Laboratory
Allan B. Tanner
U. S. Geological Survey
Robert Patzer, Ph.D.
National Environmental Research
Center - Las Vegas
Dr. Thomas F. Gesell was retained by the Environmental
Analysis Division.as a consultant in the preparation of this
report. He is an Associate Professor of Health Physics, at the
University of Texas, Health Science Center, P. 0. Box 20186
Houston, Texas 77025. Mr. Raymond fl. Johnson, Jr. is Chief of
the Surveillance Branch in the Office of Radiation Programs,
Environmental Protection Agency, 401 M Street, S.W
D.C 20460. Mr. David E. Bemhardt is Chief
vi
-------
CONTENTS
Page
FOREWARD '.. ill
ABSTRACT ^ . v
ACKNOWLEDGMENTS AND AUTHORS vi
INTRODUCTION 1
Radon In liquefied petroleum gas (LPG) 1
Approach 4. 1
Scope and objectives. .. 1
RADON CONCENTRATIONS IN NATURAL GAS 3
RADON CONCENTRATIONS IN LPG 6
At the processing plant . *.. 6
At the retail level 14
Radon concentration in home storage tanks 24
Radon in indoor air . 28
DOSIMETRY 29
222 222
Rn and Rn daughter dosimetry 29
Dose conversion factor for this study 32
Postulated exposure conditions « 35
Dose to an individual i * 37
Population dose 4. 38
Possible variations in population dose estimates ..* 40
POTENTIAL HEALTH EFFECTS 40
Conversion from dose to potential health effects 40
Health effects estimate 44
CONTROL COSTS . ....*.. 44
Comparison of radon control costs to reduction in potential
health effects *. 45
DISCUSSION i 47
Review of uncertainties * 47
Comparison with other sources of radiation 48
Interpretation of estimated health effects ». 48
CONCLUSIONS , * 50
REFERENCES i.. 51
vii
-------
FIGURES
Page
Figure 1. Model for estimating potential health effects from
radon in liquefied petroleum gas 2
Figure 2. General flow diagram for production and distribution
of natural gas and LPG 7
Figure 3. Distribution of gas processing plants in the
contiguous United States 11
Figure 4. Least squares fit of the radon-222 concentration in
propane vs. the radon-222 concentration in the
inlet gas 13
Figure 5. Flow diagram from wellhead to consumer of LPG in the
Houston, Texas area 15
Figure 6. Hypothetical flow diagram for LPG from wellhead to
market 15
Figure 7. Second order polynomial fit of the monthly average
values of 222Rn concentration found in weekly Houston,
Texas area retail LPG samples 19
Figure 8. Distribution of retail LPG sample sites 20
Figure 9. Distribution by region of 222Rn concentrations found
in retail LPG (pCi/liter) 22
Figure 10. Assigned values of 222Rn concentration in LPG delivered
to consumers for the purpose of estimating population
dose and health effects 27
TABLES
222
Table 1. Rn concentrations in natural gas at production
wells 5
Table 2. U.S. production of selected gas liquids from gas
processing plants and central fractionators 8
Table 3. Liquefied petroleum gas ^mt g
Table 4. Gas processing plants, propane production and estimated
consumption in ranges and space heaters by states and
regions 9
viii
-------
Page
Table 5. Radon-222 concentration measured in gas plant processing
streams (pCl/llter at STP) 12
Table 6. 222Rn concentrations In gaseous retail LPG (pCi/liter at
STP) in the Houston, Texas area 17
Table 7. 222Rn concentration in non-Houston retail LPG samples
(at STP) 21
222
Table 8. Relationship between Rn concentations in LPG and
distance from a major gas processing area 23
Table 9. Relationship between 222Rn concentration in LPG type of
container from which the sample was obtained 24
Table 10. Simplified decay series for radon-222 31
Table 11. Summary of dose conversion factors for radon and radon
daughters 33
Table 12. Calculation of working levels. 34
Table 13. Exposure conditions employed in the estimation of dose
from radon in LPG 36
Table 14. Doses to individuals from radon-222 in LPG 37
Table 15. Estimated dose equivalent to the U.S. population due to
222Rn in LPG 39
Table 16. Corrections to adjust estimated population doses for
different exposure conditions 41
Table 17. Estimates of excess somatic and lung cancer deaths 43
Table 18. Costs for LPG storage (1974 basis) 46
Table 19. Annualized cost estimate for storage of LPG 46
Table 20. Comparison of bronchial epithelium doses from various
sources
ix
-------
ASSESSMENT OF POTENTIAL RADIOLOGICAL HEALTH
EFFECTS FROM RADON IN LIQUEFIED PETROLEUM GAS
INTRODUCTION
Radon in Liquefied Petroleum Gas (LPG)
Most natural gas is not used directly from production wells but
undergoes routine processing for removal of impurities and the heavier,
more valuable hydrocarbons. Part of these heavier hydrocarbons, con-
sisting principally of propane, are bottled under pressure for sale as
a fuel known as liquefied petroleum gas (LPG). The process for obtain-
ing LPG is of particular interest because the radioactive noble gas,
radon-222 a normal component of natural gas, is separated along with
LPG. When this LPG is burned in unvented appliances, such as kitchen
ranges and space heaters, the radon is released within the home. Here
radon-222 decays to alpha-emitting daughter products which can con-
tribute to lung dose when inhaled and deposited in the respiratory
system. At high exposure levels, which occurred in some early under-
ground uranium mines, radon-222 has been associated with the induction
of lung cancer. This report assesses the potential for sucH lung cancers
in the general population exposed to radon from LPG. The need for this
assessment was determined from earlier studies on the health signifi-
cance of radon in natural gas by Barton, et al. (l.,2), Gesell (3^),
and Johnson, et al. C5,^).
Approach
Potential health effects from radon in LPG are estimated according to
the model outlined in figure 1. The overall analysis involves determining
the radon-222 concentration in LPG at the points of home use, the defini-
tion of exposure conditions for calculating individual and population
dose equivalents, and finally the calculation of potential health effects.
Figure 1 also shows many of the factors to be considered in each step of
this analysis.
Scope and Objectives
This assessment is concerned with potential health effects which
may result from inhalation of radon daughter products in homes where LPG
containing radon-222 is used in unvented cooking ranges or space heaters.
The use of LPG in other home appliances is of much less importance (5)
and will not be included here.
-------
ro
incentration
>f tank
filling interval
Home
storage
tank
Gas use, heating, cooking, etc.
Use rate, venting, dilution
volume.
222 Rn concentration
Daughter product equilibrium and
free ion fraction, ventilation and
infiltration; aerosol properties,
dispersion and removal processes
1
1
1
1
1
Home use of
petroleum gas
Radon daughter
dosimetry
Critical mode of
exposure
Critical organ
Demographic data
Geographical gas use
Population dose
person rem
Morbidity
Mortality
Storage
time
I
I
SOURCE
TERM
1
Radon daughter
concentration to
dose equivalent
conversion
factor
EXPOSURE
CONDITIONS
Dose equivalent
to health effects
conversion
factor
I
POPULATION
EXPOSURE
HEALTH
EFFECTS
Figure 1. Model for estimating potential health effects from radon in liquefied natural gas
-------
Information on the following items is provided in this report:
(a) measured radon-222 concentrations in natural gas and in
LPG at processing plants and retail outlets,
(b) estimated radon-222 concentrations at points of consumer use,
(c) critical mode of exposure and radon dosimetry,
(d) individual and population dose equivalents, and
(e) estimation of potential health effects and interpretation
of their significance in relation to costs for control of
radon in LPG.
RADON CONCENTRATIONS IN NATURAL GAS
Since LPG and its associated radon are derived primarily from
natural gas, it is useful to briefly review the source of radon in
natural gas. Radon-222 is the gaseous daughter product of radium-226
which is part of the naturally occurring uranium-238 decay series.
According to Bell (J.) uranium minerals are distributed throughout the
earth's crust, often in association with carbonaceous materials. Sedi-
mentary formations that are good hosts for uranium deposits are generally
good for petroleums as well. The genesis of uranium deposits is not
certain, but it is believed that soluble uranium is deposited from
transporting ground water by chemical reducing conditions resulting from
decaying organic material (8). Adsorption and the formation of uranium
metallicorganic compounds may also be involved. Bell (7) noted that there
is no evidence that petroleum acts as an ore forming fluid for uranium
but the average uranium concentrations in rock formations do become elevated
by the above processes.
The deposits of uranium associated with many petroleum or natural
gas bearing formations result in greater amountts of radium-226 available
for decay to radon-222. In addition, it is common for radium to 1-each
from the adjacent uranium minerals so that more of the gaseous radon-222
is able to diffuse into the natural gas, i.e. it is not retained within
the solid minerals* (9), Since radon is an inert gas, it permeates porous
geological formations along with natural gas and is collected in production
wells with methane, the primary component of natural gas.
Radioactivity in natural gases was first reported for Canadian gas
in 1904 by Satterly and McLennan (10). At that time, it was suggested
that helium, a known product of radioactive decay, might be associated
with radioactivity. Somewhat later (1918), a systematic survey was made
of Canadian natural gases (10) in an effort to find a relation between
1
Tanner, A.B., U.S. Geological Survey, Reston, Virginia, personal
communications to D. E. Bernhardt, 1973 and 1974.
-------
helium concentrations and radioactivity. The radioactivity was reported
as due to "radium emanation," now known to be radon-222, which decays by
alpha particle emissions, as do its daughter products, polonium-214 and
218. Each alpha particle consists of two protons and two neutrons which
is Identical with the structure of a helium nucleus without its electrons,
He . When this alpha particle or helium nucleus loses its decay energy
and comes to rest, it picks up electrons and becomes a neutral helium
atom.
Although no strong correlation between radon-222 and helium
concentrations was found in the 1904-1918 studies, interest in a possible
correlation was revived in the 1940*s and early 1950Ts resulting in a
series of papers dealing with radon and helium concentrations in U.S.
natural gases (11,12). A comprehensive survey of radon-222 and helium
concentrations in the Texas Panhandle Field gases was issued by the U.S.
Geological Survey in 1964 (13). The hypothesis that most of the helium
found in natural gases is of radiogenic origin is still widely accepted.
Nevertheless, no correlation has been found between radon-222 and helium
concentrations in individual wells. This is partly the result of rapid
equilibration of radon with radium deposits which may be moved in time by
ground water, whereas the helium has been collecting over geologic time.
Subsequent to the reports dealing with possible radon-222 and helium
concentration correlations, reports were made on radon-222 concentrations
in wellhead natural gas in connection with nuclear gas well stimulation
experiments (14,15). These reports dealt with radon concentrations in
natural gas in San Juan Basin, located in southwestern Colorado and north-
western New Mexico. These papers are significant because the potential
for exposure of population groups to radon-222 via the natural gas pathway
was explicitly recognized. These investigations determined that the
Project Gasbuggy nuclear stimulation experiment did not raise the radon-222
concentrations in the neighboring wells above the naturally occurring levels.
In addition to the primary sources of data on radon in natural gas
mentioned above, there are other data available in published and un-
published report form 2,3 (i8-22) which have been reviewed by Gesell (3,4,
16,17) and by Johnson et al. (5.,6). A summary of the presently available
information on radon-222 at the wellhead is given in table 1. The radon
concentrations are seen to range from nearly zero to 1450 pCi/liter with
averages for groups of wells up to about 100 pCi/liter in the U.S. and
about 169 pCi/liter in Canada. The differences in radon concentrations
are not due just to the amount of radium-226 present in the rock formations,
but also to the pressure of the gases in the well. The production of radon
in a given gas reservoir volume will be fairly stable with time, but as
the reservoir pressure is reduced by removal of natural gas, less gas
remains to dilute the radon. Thus radon concentrations will increase as
gas reservoirs near depletion.^ Radon concentrations are also influenced
by gas flow rates which are functions of geological formation permeabilities
as well as pressure.
2Tanner, A.B., op. cit.
Skipp, B., U.S. Geological Survey, unpublished data reports from
work in the 1950's in Kansas, Oklahoma, and Texas.
4
-------
Table 1. 2^2Rn Concentrations in Natural Gas at Production Wells
Area
Ontario, Canada
Alberta, Canada
British Columbia
Colorado, New Mexico
Texas, Kansas, Oklahoma
Texas Panhandle
Colorado
Project Gas buggy Area
Project Gasbuggy Area
California
Gulf Coast (Louisiana, Texas)
Kansas
Wyoming
222Rn level (pCi/1)
Average
169
62
473
25
<100
25.4
15.8-19.4
29.4
5
100
10
Range
4-800
10-205
390-540
0.2-160
5-1450
10-520
11-45
12-59
1-100
Reference
10,18
10,18
10,18
14
13,18
11
15,19,20
15
21
22
4*
5*
5*
*4Kaye, S. V., Oak Ridge National Laboratory, Memorandum to distribution.
Progress Report for October 1 - December 31, 1972. January 17, 1973.
*5
Bernhardt, D. E., Memorandum to D. W. Hendricks, Radon-222 in natural aas
{May 24, 1973).
-------
RADON CONCENTRATIONS IN LPG
At the Processing Plant
A generalized flow sheet for the production and distribution of
natural gas and LPG is shown in figure 2. Some of the factors affecting
the concentration of radon-222 are also given in this figure. In par-
ticular, the behavior of radon in the processing, storage, and sale of
LPG will be considered in detail.
Natural gas from the wellhead is usually processed by thermal frac-
tionation and other means to remove the heavier, more valuable hydrocarbons
and impurities (23). Gas processing plants are usually located near the
gas fields which supply them. After processing, the "residue gas" as the
processed natural gas is called, may be sold for use as a fuel or chemical
feedstock or it may be pumped back underground to sweep the oil fields.
The latter process is called a cycle operation and produces little or no
methane for sale. After the field has been swept and the cycling operation
is no longer profitable, residue gas is sold and the cycling operation cur-
tailed. The typical products of a gas processing plant together with the
average production for 1971 are given in table 2.
While the details of the processing method may vary, the physical
principle behind the separation of the heavier hydrocarbons from the inlet
stream mixture is thermal fractionation. Table 3 gives the boiling points
for several of the products of gas processing together with that of radon.
It is clear that the boiling point of radon is bracketed by those of ethane
and propane, and radon will therefore tend to separate with these fractions.
The makeup of the LPG sold for domestic fuel may be principally propane
with small amounts of the other hydrocarbons, or it may be principally
butane with small amounts of the other hydrocarbons, or mixtures of propane
and butane. The use of butane is limited to regions or seasons where the
temperature is not likely to go below 0°C (32°F) because below this tempera-
ture the butane will not produce sufficient vapor pressure to be useful
This disadvantage coupled with the increasing value of butane as a chemical
feedstock has curtailed its use as a fuel. Thus most of the LPG used as a
domestic fuel is propane-based. Gesell 04) analyzed the composition of six
samples of LPG collected from different retail dealers in the Houston, Texas
area; the results are shown in table 3. These measurements indicate the
predominant use of propane with some ethane in these LPG samples.
Liquefied petroleum gas is produced principally in the same regions
of the country where natural gas is produced. Table 4 gives a listing of
total gallons of propane and LPG mixture produced by region and by state
together with estimates of domestic consumption in ranges and space heaters.
These data were developed from Cannon (23_) and from U.S. Census Eureau
statistics (24^. Just as with natural gas (5), the overwhelming majority of
LPG is produced in the west south central region (Texas, Oklahoma, Louisiana,
Arkansas).
-------
Radon Concentration,
Well Location, Depth
and Pressure, Seasonal
Variations
Radon Separation
with LPG
See Ref. (1-6)
for Radon in
Natural Gas
Natural Gas
Production Wells
Impurities
C02,N2,H20,
He,H2S, etc.
Gas
Processing Plant
I
Natural Gas
Distribution System
Other Usable Fractions,
Butane, Ethane, Gasoline, etc.
1
Liquefied Petroleum
Gas
Distribution System
Wholesale Pipeline or
Truck Delivery, Retail
Delivery or Retail Pickup
Transport and
storage time
1
Home Storage
Tanks
Figure 2. GENERAL FLOW DIAGRAM FOR PRODUCTION AND DISTRIBUTION OF
NATURAL GAS AND LPG.
-------
Table 2
U.S. Production of Selected Gas Liquids from Gas Processing
Plants and Central Fractionators (23)
Product
Production - 1971
(109 gal/yr)*
Ethane
Propane
Isobutane
Normal and unsplit butane
Natural gasoline
3.67
8.48
1.43
3.81
4.28
*1 gallon of liquid propane yields 1.25 m3 of gas at STP.
Table 3
Liquefied Petroleum Gas (4)
Component
Methane
Ethane
Radon
Propane
Butane
Percent
of LPG*
0.2-2.0
2.4-9.5
88-96
0.5-1.5
Boiling
Point (°C)
-161.5
- 88.3
- 61.8
- 42.2
- 0.5
'Measured in a survey of six retail sources of LPG in the Houston, Texas area.
-------
Table 4
Gas Processing Plants, Propane Production and Estimated Consumption In Ranges and
Space Heaters by States and Regions
Division and State
United States(b)
New England
Connecticut
Maine, New Hampshire, Vermont
Massachusetts
Rhode Island
Middle Atlantic
New Jersey
New York
Pennsylvania
East North Central
Illinois
Indiana
Michigan
Ohio
Wisconsin
West North Central
Iowa
Kansas
Minnesota
Missouri
Nebraska
North Dakota
South Dakota
South Atlantic
Delaware
Florida
Georgia
Maryland, District of Columbia
North Carolina
South Carolina
Virginia
West Virginia
East South Central
Alabama
Kentucky
Mississippi
Tennessee
West South Central
Arkansas
Louisiana
Oklahoma
Texas
Mountain
Arizona
Colorado
Idaho
Montana
Nevada
New Mexico
Utah
Wyoming
Pacific
Alaska
California
Hawaii
Oregon
Washington
3as Processing
Plants
805
Percentage
0
0
0
0
0
«*
0
0.2
0.7
0.1
0
0.6
0
0
4.2
0
3.5
0
0
0.2
0.4
0.1
0.6
0
0.1
0
0
0
0
0
0.5
1.6
OTT
0.4
1.1
0
75.4
~074
16.9
11.6
46.5
10.5
~OTT
1.4
0
0.7
0
4.5
0.4
3.4
6.8
O
6.6
0
0
Q
Propane
Production
8.48 x 109
of United States
0
0
0
0
0
.02
0
0
.02
3.3
O
0
0.5
0
0
7.0
0~
6.2
0
0
0.2
0.3
0.3
1.2
0
0.30
0
0
0
0
0
0.90
1.4
0~
1.2
0.2
0
74.8
"O
16.9
10.8
46.8
8.3
o~
0.9
0
0.?
0.0
3.5
0.4
3.3
3.9
0.1
3.8
0
0
0
Use<»
2.24 x 109
3.5
0.8
1 .5
1.0
0.2
7.0
079
4.0
2.1
20.2
~5~4
3.8
3.5
3.1
4.4
20.7
~O
2.9
4.2
8.1
2.6
1.1
0.5
13.8
~072
3.5
4.0
1.1
2.2
1.3
1.1
0.4
9.8
370
2.7
2.7
1.4
12.9
~O
1.4
3.0
5.7
6.6
676
1.9
0.5
O.r'
0.6
1.1
0.4
0.6
5.5
oTT
3.9
0.1
0.7
0.7
(a)Note that this usage Is for domestic ranges and space heaters only.
not include other domestic appliances, commercial or industrial uses.
(fa)propane production and consumption 1h gallons per year.
It does
-------
The factors which affect the 222Rn content of propane as produced
in the gas processing plant include the 222Rn content in the wellhead
gas, the transit time from wellhead to processing plant, the makeup of
the inlet gas, and the type of processing.
Gesell (40 has surveyed nine gas processing plants in the U.S.
The locations of these surveys are shown by the circles in figure 3.
Also shown in figure 3 are the locations of gas processing plants in
the United States. The survey included measurement of the 222Rn content
of the inlet stream(s), some of the products and the residue (outlet) gas.
The results of this survey are given in table 5. It can be seen that
222Rn concentrations in the inlet gases are in the expected range for
wellhead gases (see table 1). Furthermore, the 222Rn concentrations in
the propane are much higher than those in the inlet natural gas, ranging
up to 1119 pCi/liter. Figure 4 shows a plot of the 222Rn concentration
in propane vs. the 222Rn concentration in the inlet gas for eight of the
nine plants (the propane sample was lost from one plant). A least squares
fit to the data gave a slope of 8.2 and a linear correlation coefficient
of °'94',,,Jhus £or a variety of processing plants, the average concentra-
tion of z/2Rn in propane is about eight times the concentration in the
inlet gas.
Bernhardt reported that a radon concentration of 56 pCi/liter in
the inlet natural gas was increased to 1100 pCi/liter in the propane
fraction at a San Juan gas processing plant which indicates an increase
in radon concentration by a factor of nearly 20. In addition, Fries and
Kilgren (22) report data for a plant in California, which indicate a con-
centration factor for radon of about 14. The differences in these factors
may be attributed to differences in hydrocarbon make-up of the inlet
natural gas or to differences in plant operations and the composition of
the LPG product, i.e., the relative quantities of ethane, propane, and
butane. The factor of 20 reported by Bernhardt6 was for a basically pro-
pane stream. The butane stream for this facility had a radon concentra-
tion factor of 7. The factor of 14 determined by Fries and Kilgren (22)
appears to be for a mixed propane-butane stream. When these factors are
combined with the data from Gesell (4), the average concentration of
radon in propane is about 10+5 times that in natural gas at one standard
deviation.
It should be noted that the study by Fries and Kilgren (22) indicated
significant external gamma radiation levels may be found within the gas
processing plant, arising from the short-lived radon daughters. However,
these radiation levels were quite localized in particular process equip-
ment. In some pieces of equipment, measureable quantities of long-lived
radon daughters were also found. Fries and Kilgren (22) concluded that
these quantities did not present a radiation hazard t~plant maintenance
personnel. Nevertheless, several precautions were recommended. Namely.
process equipment with significant external radiation should be shut down
6Bernhardt, D.E., "Radon in natural gas productsSan Juan Plant,"
memorandum to C.L. Weaver, August 31, 1973.
10
-------
Figure 3. DISTRIBUTION OF GAS PROCESSING PLANTS IN THE CONTIGUOUS UNTIED STATES.
EACH SMALL DOT REPRESENTS ONE GAS PROCESSING PLANT AND EACH CIRCLE
REPRESENTS A SURVEYED GAS PLANT (4)
-------
Table 5
Radon-222 Concentration Measured in Gas Plant Processing
Streams (pCi/liter at STP) (4)
Source
Inlet gas
Ethane
Propane
Butane
Sales gas (methane)
Plant Code
7
10.5
368.2
177.6
1.5
0.6
2
26.3
(b)
13.6
3
78.4
386.1
3.2
4
118.5
1,119.0
93.6
5
35.4
97.7^
237.7
3.1
6
18.4
306.7
56.8
0.9
7
2.9
5.0
5.0
61.5
1.2
1.0
8
1.0
1.5
9.0
0.9
0.5
9
23.6
41.3
17.3
28.6
213.3
---
Is}
(a)
(b)
Mixture of ethane and propane.
Sample lost.
-------
- 1200
i
o
LJJ
Q.
O
CC
Q.
O
z
O
O
c
cc
CNJ
evi
CM
800
400
40
80
120
222Rn CONC. IN INLET GAS (pCi/liter)
Figure 4. LEAST SQUARES FIT OF THE RADON-222
CONCENTRATION IN PROPANE VS. THE RADON-222
CONCENTRATION IN THE INLET GAS (4)
-------
at least three hours "before disassembly to allow decay of short-lived
nuclides. Then for equipment requiring brushing or scraping, gloves
and a face mask should be worn to prevent inhalation or ingestion of
long-lived radon daughters. They furthermore recommended that decay
time for radon in LPG be maximized by selling the oldest product first.
In addition to propane produced in domestic gas processing plants,
propane is also produced by oil refineries and is imported. For the
year ending in November 1973, gas processing plants produced approxi-
mately 9 x 109 gallons of propane, refineries produced 14 x 10 gallons
and M. x 1(K gallons were imported (25) .
Refinery propane is not expected to contain high levels of radon-
222, but concentrations should be checked. No information on radon-222
content of imported propane is available. However, radon levels should
be low since shipping time would reduce the radon-222 content from the
level at manufacture.
At the Retail Level
Although a knowledge of radon in wellhead gas and radon in LPG at
the point of manufacture is basic to an understanding of the 222Rn in
LPG problem, it does not provide direct information on radon levels to
be expected in LPG as purchased ty or delivered to the consumer. Trans-
portation and storage networks are complex and the relatively short
half-Life of 222Rn can result in substantial decay of the activity. An
indication of the complexity of the supply, transit, storage and delivery
system is given by Gesell (4^) who has studied the system for Houston,
Texas. A flow diagram for a typical retailer is given in figure 5.
Distribution systems in other areas may be more or less complex. For
example, rail shipment is utilized for markets far from production, and
for markets very near to production, the retail delivery truck may pick
up directly from the gas plant on the day of delivery. Figure 6 shows a
hypothetical delivery system which includes the various known sources,
storage, transport, and delivery modes. It is clear from this figure
that, because of the unknown storage and transport times, the best infor-
mation on 222Rn levels in LPG at the consumer level would be obtained at
the retail dealer level.
Gesell (3.»A»!i»!2.) ^as conducted a year-long weekly survey of LPG
at the retail level in Houston, Texas, as well as a grab sample survey
of retailers in 14 southern and southwestern states. He used the Lucas
(26) method for radon analysis.
The year-long study of seven retailers was performed to examine any
seasonal trends in radon concentration. The results of this survey (4)
are given in table 6. Also included in table 6 are column and row aver-
ages. Retailer No. 7 was consistently low, so sampling was discontinued
after six months. For purposes of the averages, retailer No. 7 was taken
as 0.5 pCi/liter for February through June. The data are characterized
14
-------
Gas Wells - (South Texas)
Gas Processing - (South Texas)
Retail Surface
Storage Facility
(Houston Area)
T
pipeline
LPG Wholesale
Distribution
(Houston Area)
Other
Customers
Retail Truck A
Retail Truck B
NG and Other
Products
Surface
Storage
Underground
Storage
Residentrial, Commercial, and Recretional Customers
Figure 5. Flow Diagram from Wellhead to Consumer of LPG in the Houston,
Texas Area (4)
15
-------
Industrial
and
Commercial
Customers
Prod
W
\
uction
ells
V
U.S. Gas
Processing
Plants and
Refineries
\
Retail ^
Truck S^
Nearby
Customers
)
Pipeline
Rail
i i i i
i i
I 1 1 r1 i i
Transport Truck
Local Storage
/
X
Bottle Station
/
Industrial
and
Commercial
Customers
1
/
Intermediate
Distance
Reg
Distr
Cei
onal
ibution
iters
\
Retail
Truck
\
/
\
Intermediate
Distance
Customers
/
i
Imports
'\
Pipeline
Rail
i i i
1 i i i i
Transport Truck
^ Local Storage
s
^^ 1
Bottle Station
/
Industrial
and
Commercial
Customers
X I
Distant
Regional
Distribution
Centers
Retail
Truck
,
\
/
\
Distant
Customers
\
Local Storage
S
\
Bottle Station
'
Figure 6. HYPOTHETICAL FLOW DIAGRAM FOR LPG FROM WELLHEAD TO MARKET.
-------
T.M. < z22Rn Con«ntr.t1on in Gaseous Retail LPG (pCI/Httr at STP) In the Houston. Texas Area (4)
Meek of
July 23 1972
July 30
Aug 6
Aug 13
Aug 20
Aug 27
Sept 10
Sept 17
Sept 24
Oct 1
Oct 8
Oct 15
Oct 22
Oct 29
Nov 5
Xov 12
Nov 19
Nov ?6
Dec 3
Dec 10
Pec 17
Dec 24
Dec 31
Jan 7 1973
Jan 14
Jan 21
Jan 26
Fch 4
Feb 11
Feb 18
Feb 25
Mar 4
Mar 12
Mar 26
Apr 1
Apr 8
Apr 15
Apr 22
Apr 29
May 6
May 13
May 20
May 27
June 3
June 17
June 24
Average
Retailer Code Number
1
9.7
9.0<»)
23.9
77. 00)
106.30)
45.3
28.6
30.9
23.7
67.4
5.4
12.9
3.6
11.9
0.7
72,8
8.9
1.4
0.6
<0.5
<0.5
0.9
28.fi
6.3
4.3
2.1
0.8
5.5
7.7
197.?
11.1
109.8
9.6
122.7
109.6
120.1
89.7
81.4
128.0
62.9
33.5
58.5
160.2
64.8
148.8
131.7
49.2
2
5.8
5.3«>>
77.90)
46.8(1")
55.70)
68.2
22.8
80.6
24.1
7.4
31.3
7.3
95.1
101.1
2.4
100,7
1.0
8.1
215.1
2.8
2.2
<0.5
2.6
,,1.1
0.8
1.2
130.0
7.2
41.0
12.3
83.6
28.6
6.0
84.9
29.8
88.0
58.1
63.0
42.S
14.4
8.3
4.6
16.0
98.7
5.5
8.3
38.5
3
8.8
3.2<>>
84.10)
20.30)
42.20)
19.9
6.2
2.1
21.8
13.6
35.5
24.8
46.1
31.1
71.0
33.2
35.9
114.0
10.6
55.4
33.7
5.6
22.5
...
'0.5
4.3
1.9'
0.7
1.7
6.1.4
19.0
5.3
23.2
3.0
8.6
3.1
1.9
7.9
8.5
...
58.7
10.6
69.8
3.0
66.6
25.9
4
141.4
120.5
137.4
140.4
142.5
221.4
5
79.2
35. 9<")
37.50)
35.00>
40.20)
17.5
70.2 14.3
110.1
51.0
86.8
11.8
149.4
81.3
100.3
71.1
55. E
65.0
2.2
11.9
<0.5
0.8
<0.5
0.7
<0.5
0.7
<0.5
0.7
16.4
15.2
6.2
8.6
20.3
52.1
48.0
88.6
91.9
102.2
46.7
59.2
107.2
69.5
51.4
4.8
30.1
20.1
27.1
43.4
18.0
32.9
67.8
14.1
14.0
22.8
54.1
15.9
23.8
17.3
3.4
1.8
1.0
2.1
1.0
«2.5
15.5
3.1
2.2
145.7
6.8
22.5
4.R
8.5
'1,9
5.0
3.7
11.6
52.9.
27.9
26.0
32.6
46.9
12.3
25.5
70.4
12.1
17.1
41.0
26.7 .
6
316.3
150.4<«>
181 .9
214.lO>
236.50)
155.3
155.6
139.0
153.2
133.8
17.0
306.5
151.6
211.1
173.1
277.1
177.4
, ?1B.4
?09.9
170.4
202.1
178.2
121.3
147.7
133.2
157.9
115.fi
123.8
194.5
1R4.0
169.4
147.8
274.4
195.0
124.0
120.2
215.6
148.1
138.8
146.0
231.6
143.6
191.5
105.9
262.9
152.4
170.4
7
<0.5
<0.5<»)
<0.5
3.9
0.7
<0.5
0.8
5.5
4.1
2.9
4.9
6.5
0.7
1.2
7.4
7.6
1.5
6.7
3.6
3.7
1.5
...
...
...
...
...
...
...
...
...
...
...
...
1,8
Average
80.2
46.4
77.7
76.2
89.1
76.0
42.6
54.5
43.9
54.6
17.0
74.0
57.9
73.7
47.8
80.6
44.7
50.3
64.5
33.9
35.0
27.2
31.4
26.2
20.1
24.1
56.4
23.0
4P.4
66.9
43.0
44.9
53.0
6S.4
53.2
68.1
70.8
53.4
58.6
58.8
59.2
42.1
73.3
45.0
74.5
61.1
50.8
,
0)
Average of two samples.
Average of three samples.
17
-------
by a large amount of scatter within and among dealers. This scatter
is not too surprising in view of the previously discussed factors that
control the radon in LPG.
Of the seven retailers sampled, No. 6 was on the average much
higher in 222Rn concentration than the others and No. 7 was on the
average much lower than the others. Inquiry revealed that retail
dealership No. 6 obtains its LPG directly from a gas processing plant
and that retail dealership No. 7 receives its LPG from an underground
storage well. The other five retailers obtain their LPG from various
wholesale distributors. Thus, the radon concentrations in the LPG
from the several retailers are consistent with their individual supply
practices which govern storage decay times for radon-222.
Figure 7 shows a second order polynomial fit of the monthly average
data from the seven sampled Houston area dealers. Gesell (4) interprets
the pronounced dip in the activity during the winter months~as being due
to the higher winter consumption rate and consequent appearance on the
market of previously stored LPG. This interpretation is consistent with
the known practices for manufacture, storage, and distribution of LPG.
The production of LPG is maintained at a relatively constant level, with
production going to storage or consumption according to demand.
It is worthwhile noting that the average concentration during the
heating season is approximately one-half of the concentration during the
summer months . Gesell (4) notes that the existing practice (at least in
the Houston, Texas area) of storing during the summer and using from
storage during the winter serves to reduce the exposure of those who utilize
LPG in unvented heating from what it would be if production were simply
adjusted to demand.
Gesell' s (^) grab sample survey of retailers in southern and south-
western states included LPG samples at the locations indicated in figure
8. The results for these LPG samples are presented individually in table 7,
and are grouped by region in figure 9. In each region of figure 9 the
top number is the maximum concentration observed, the middle number is
the average of all samples in the region and the bottom number is the
lowest concentration observed. The most remarkable feature of this dis-
tribution is difference in radon concentrations in the Eastern and Western
parts of the United States. Two major factors affect the concentration
of radon in retail LPG, concentration at the point of manufacture and the
time required to deliver the gas to the consumer. The low concentrations
found in the two eastern-most regions could potentially be explained in
terms of longer transport and storage times since these areas are fairly
far from any major production. The low concentrations found in Louisiana
and Arkansas, in view of the large amount of gas processing in Louisiana,
invite interpretation in terms of low concentrations in the sources. This
interpretation is consistent with the low 222Rn concentrations found in
the gas processing plant surveyed in Louisiana and the reported low radon
concentration in Gulf Coast gas (5,18).
18
-------
100
80
o
60
z
o
z
UJ
O
I
40
20
1
I I I I I I I I I 1
JUL SEP NOV JAN MAR MAY
AUG OCT DEC FEB APR JUN
1972-73
Figure 7. SECOND ORDER POLYNOMIAL FIT OF THE MONTHLY
AVERAGE VALUES OF 222RN CONCENTRATION FOUND
IN WEEKLY HOUSTON, TEXAS, AREA RETAIL LPG SAMPLES (4)
19
-------
ro
o
Figure 8. DISTRIBUTION OF RETAIL LPG SAMPLE SITES (4)
-------
Table 3. Radon-222 Concentration In Retail LPG Samples (at STP) (1)
Location
ALABAMA
Dotnan
ARIZONA
Vina
Vuma
ARKANSAS
N Little Rock
Forrest City
CALIFORNIA
Angels Camp
Bakersfield
BakersfieTd
Bakersfield
Carlsbad
Delano
El Cajon
El Centro
El Toro
El Toro
EscondldQ
Fresno
Santa Barbara
Stockton
Vlsalla
King City
Lancaster
Livingston
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Gatos
Mercedes
Modesto
Paso Robles
FLORIDA
Bonlfay
Chlpley
Oe Funlak
Springs
PensacoTa
Tallahassee
TENNESSEE
Jackson
Knoxvllle
La Follette
Memphis
Memphis
Nashville
Sevlerville
TEXAS
Alice
Angle ton
Angloton
Arlington
Atlanta
Austin
Austin
Austin
Eal linger
Bay City
Bay City
Beaumont
Beaumont
Brownsville
Brownsville
Brownwood
Borger
Carthage
Columbus
Corpus Christi
Corpus Christi
Corslcina
Cortul la
Dal las
Dallas
Denton
Dunas
Enn1 s
El Paso
FredeHcksburg
Galveston
George West
Gollad
Harlingen
Harlingen
Date
08/02/73
08/30/73
08/30/73
08/08/73
08/08/73
08/24/73
08/28/73
08/28/73
08/28/73
08/22/73
08/27/73
08/22/73
08/30/73
08/22/73
08/23/73
08/28/73
07/27/73
08/23/73
08/25/73
08/27/73
08/23/73
08/28/73
08/27/73
08/22/73
08/22/73
08/22/73
08/22/73
08/24/73
08/27/73
03/27/73
08/23/73
08/02/73
08/02/73
03/02/73
08/01/73
08/02/73
03/08/73
08/06/73
08/06/73
OR/08/73
W 08/73
08/07/73
08/06/73
01/19/73
0)/!8/73
01/18/73
07/17/73
04/13/73
05/11/73
05/11/73
05/11/73
07/26/73
01/18/73
01/18/73
12/18/72
12/18/72
02/23/73
02/23/73
07/26/73
07/19/73
04/13/73
05/10/73
01/19/73
01/19/73
07/T6/73
07/13/73
07/16/73
07/16/73
07/17/73
07/19/73
07/16/73
08/30/73
05/11/73
12/18/72
07/12/73
07/12/73
02/23/73
02/23/73
Radon-222
Concentration
(PCI/liter)
<0.5
6.8
34.1
1.8
1.6
7.9
55.4
15.1
133.8
84.0
102.0
24.5
3.4
45.6
41.9
27.9
22.3
47.0
3.3
710.8
1049.3
8.5
2.0
144.2
663.1
92.8
88.0
5.9
283.5
269.7
5.4
<0.5
1.3
<0,5
3.8
1.2
<0.5
<0.5
1.1
10.3
04/13/73
Livingston
longvie*
Lufkln
Lufkin
Lufkln
Luling
Marshall
Me A lien
Midland
Nacogdoches
Port Arthur
Port Arthur
Robs town
San Angel o
San Antonio
San Antonio
04/12/73
04/13/73
04/12/73
04/12/73
04/12/73
05/10/73
04/13/73
02/23/73
07/25/73
04/12/73
01/19/73
12/18/72
02/22/73
07/25/73
05/11/73
05/11/73
San Antonfo 05/11/73
San Antonfo 05/11/73
San Benlto 02/23/73
San Benito 02/23/73
San Bent to | 02/23/73
Seguln 06/10/73
Slnton 02/22/73
Temple 07/26/73
Te*arkana '04/13/73
Texarkan* : 04/13/73
Texllne '07/20/73
Victoria 03/19/73
ffdor 12/18/72
Weimar 05/10/73
Weslaco 02/23/73
ZMJtL .07/13/73
Radon-222
Concentration
(pCi/Hter)
1.1
0.6
<0.5
2.7
<0.5
<0.5
<0.5
0.8
4.B
<0.5
3.B
-------
Figure 9. DISTRIBUTION BY REGION OF 222Rn CONCENTRATIONS FOUND IN RETAIL LPG (pCi/liter)
IN EACH REGION. THE TOP NUMBER IS THE MAXIMUM VALUE FOUND, THE LOWER NUMBER IS THE
LOWEST VALUE FOUND AND THE MIDDLE NUMBER IS THE MEAN OF ALL SAMPLES IN THE REGION. <4>
-------
Gesell (6) examined the variation of 222Rn in LPG with distance
from a major gas processing area by dividing the data from table 7
into three categories according to whether the samples were taken
less than 25 miles, between 25 and 200 miles, or greater than two-
hundred miles from a major gas processing area. The results are
shown in table 8. The differences were tested by the non-parametric
Wilcoxon two sample test (27). The mean for the 25-200 mile category
was found to be significantly less (P < 0.1) than the mean for the
<25 mile category and the .mean for the >200 mile category was found
to be significantly less (P < 0.01) than the mean for the 25-200 mile
category. The results suggest that the^unsampled northern states
would tend to have lower radon concentrations than the sampled states
since they are, with few exceptions farther from the major gas pro-
cessing areas than the sampled states. This suggestion assumes that
an insignificant amount of Canadian LPG enters the northern states
market.
TABLE 8
RELATIONSHIP BETWEEN 222RN CONCENTRATION IN LPG
AND DISTANCE FROM A MAJOR GAS PROCESSING AREA (4).
Average
Distance Number of Concentration
(miles) Samples (pCi/liter
<25 103 87.8
25-200 23 63.7
>200 31 45.6
Actually, about 7 percent of the annual LPG consumed in the United
States is imported, mainly from Canada. Most of this Canadian LPG is
used in the states nearest to Canada. Since the Canadian LPG has a
generally higher radon-222 content than that from the United States (as
shown in table 1), then the average radon content of LPG consumed in the
northern states might be higher than would be predicted for LPG from
the United States alone. However, no specific information was available
for this study on radon concentrations in LPG in the northern states.
Therefore, the contribution of radon from Canadian LPG was not included
specifically in this study, although the general significance of Canadian
LPG will be considered in the overall interpretation of data (see discus-
sion section).
23
-------
Gesell (4) also examined the variation of 222Rn in LPG with the
type of container (retail storage tank, truck or small tank) from
which the sample was drawn. Only Texas was included in this examina-
tion because in the other regions very few samples were obtained from
trucks. Retailers fill bottles brought to their locations from the
same, generally large (>10,000 gal.) tanks from which they fill their
own delivery trucks, from the delivery trucks themselves, or from
special "bottle stations" which are supplied by small (generally
<2,000 gal.) tanks. The "bottle stations" are typically filled from
the retail trucks. Thus, the LPG in the bottle stations would tend to
be older, on the average, than the LPG in the trucks and that in the
trucks slightly older than the LPG in the large storage tanks. The
Texas samples (non-Houston) were categorized and averaged and the
results are presented in table 9. The mean of the truck samples, al-
though smaller than that of the large tank samples was not significantly
smaller. The mean of the small tank samples is significantly smaller
(P <0.01) than both the truck samples and the large tank samples. Signi-
ficance was tested with the non-parametric Wilcoxon two sample test (27).
This finding is consistent with the operational practices of the LPG
retailers described above in that the samples which are anticipated to
be older have, on the average, lower radon concentrations.
TABLE 9
RELATIONSHIP BETWEEN 222RN CONCENTRATION IN LPG
TYPE OF CONTAINER FROM WHICH THE SAMPLE WAS OBTAINED
Container
Large Tank
Truck
Small Tank
Number of
Samples
21
20
30
Average Concentration
(pCi/liter)
162.9
125.8
31.4
Radon Concentration in Home Storage Tanks
In order to estimate population dose and utimately potential health
effects, it is necessary to know or estimate the average radon levels in
LPG contained in the consumers1 home storage facilities. This could be
best accomplished by a suitably large, well-distributed (geographically
and in time) sampling of LPG in consumers' home storage tanks. Such data
do not exist, however, so an alternative method will be employed to esti-
mate the radon levels in delivered LPG as a function of geographical
24
-------
location. This method is based upon Gesell's (4) data and knowledge
of the locations of gas processing plants. Referring to figure 9, it
is seen that California LPG has an average of M.50 pCi/liter. Texas
exhibits an average of MOO pCi/liter and Arizona-New Mexico exhibit
an average of <50 pCi/liter. These regions have the highest radon in
LPG levels, with the remainder of the sampled states exhibiting average
radon in LPG levels of <5 pCi/liter.
It should be noted however that with the exception of Texas and
California, very limited sampling was done in Gesell's study (4). For
example, only three samples were gathered in Oklahoma, Even without
resorting to statistical techniques, one can examine the data for Texas
(table 7) and readily see that if only three samples were selected at
random from Texas there would be a very good chance of missing the
average value by a wide margin. Thus, excessive confidence should not
be placed in the average concentration values for regions with only a
few samples.
In addition to the measured concentration values in the fourteen
states, the seasonal behavior must be considered. Furthermore, a
rationale for estimating the average concentrations for the states not
included in Gesell's survey must be developed. Gesell (_4) found that
for Houston, Texas the average winter 222Rn concentration was approxi-
mately one-half the average summer content. Because the fourteen state
survey was performed in the summer months, one should consider applying
a seasonal adjustment which would affect especially the dose due to
unvented heaters which are used mainly in winter months. However, since
the principle of radiation protection is to estimate on the conservative
side, when reliable information is not available to indicate otherwise,
and especially since the experience in Houston may not be directly appli-
cable to other regions of the country, no seasonal adjustment is made for
calculations in this study. This matter will be considered further in
the discussion of uncertainties later in the report.
A scheme for estimating the radon concentration in LPG retailed
in the unsampled states should reflect the known information in the
sampled states as well as the distance from gas producing areas.
Gesell has shown (table 8) that an expected relationship exists be-
tween average radon content and distance from a major gas processing
area. Since this relationship indicates that concentrations decline
as distance from a gas processing area increases, LPG from unsampled
states far from gas processing areas should be estimated as lower in
222Rn content than states containing significant gas processing opera-
tions .
Based on the foregoing considerations, the following state average
radon concentrations in LPG delivered to consumers have been assigned for
the purpose of estimating potential health effects.
1. Based solely on Gesell's data (4), the following assignments have
been made:
California: 150 pCi/liter
Arizona-New Mexico: 50 pCi/liter
Texas: 100 pCi/liter
25
-------
2, Based partly on Gesell rs data (4) and partly on the desire to be
conservative, Oklahoma, Arkansas, Louisiana and the remaining
surveyed southeastern states were assigned a value of 10 pCi/liter.
3. Based on proximity to states sampled by Gesell (4_), Nevada, Utah
and Colorado have been assigned a value of 50 pCi/liter.
4. Based primarily on the large distances from u^ajor gas processing
areas and upon Gesell 's data for states with gas processing, all
remaining states have been assigned a value of 10 pCi/liter. (The
significance of possible higher values in northern states due to
import of Canadian LPG will be considered in the conclusions of this
study.)
These radon concentrations are summarized in figure 10 which gives
the state by state assigned values. All the calculations and estimates
carried out later in the report utilize these values.
In previous studies on the potential radiological health effects of
radon in natural gas (1-6), it has been generally assumed that the radon
concentrations measured in the distribution systems are equivalent to the
concentrations at the point of consumption. This assumption is a valid
one because of the magnitude of the uncertainties in other aspects of
the calculations and because transport pipelines do not represent more
than a few hours to a few days delay between points of sampling and use.
In the case of LPG, however, direct application of the concentations
found at the retail level would result in a serious over-estimation of
the population dose. This over-estimation would occur because substantial
storage time is involved in the residential storage tanks. If we consider
that residential LPG storage tanks are filled at time intervals (t) with
LPG of 222Rn concentration CQ, then the concentration C of the 222Rn in
LPG at time t' after delivery would be
C - C e-Xt'
C - CQ a ,
where X is the decay constant for 222Rn. The average 222Rn content, ~C>
can be obtained by integrating C over the delivery interval (t) and
dividing by t.
c = i ;* c(t')dtf
c o
C . i C /c e"Xtfdtf
t o o
- >i ~\
C - - (l - e )
Discussions with dealers indicate that they prefer monthly deliveries
and that domestic storage tanks are sized with that delivery interval in
26
-------
ro
Figure 10. ASSIGNED VALUES OF 222RN CONCENTRATION IN LPG DELIVERED TO CONSUMERS FOR THE PURPOSE OF
ESTIMATING POPULATION DOSE AND HEALTH EFFECTS. ALL VALUES ARE IN PCI/L AT STP,
-------
mind. Accordingly, a time (t) of 30 days is chosen and
-------
Gas appliances may be vented to the outside of the dwelling, par-
tially vented to the outside, or unvented. Kitchen ranges are typically
unvented, although a hood with a fan venting to the outside of the house
is sometimes provided. Gas central furnaces are typically vented but
there is widespread use of unvented space heating, particularly in the
southern states. In Mississippi and Louisiana, over 30% of the dwell-
ings utilize unvented space heating as the primary source of heat (24).
In Texas, Oklahoma, Arkansas and Alabama, the figure is over 20% (24j.
Other gas appliances which may or may not be vented include clothes~
dryers and refrigerators.
Johnson et al. (5) and Handley and Barton (30) have recently
reviewed the literature with regard to dwelling ventilation rates. The
general disagreement in the literature is probably more reflective of
actual differences in the characteristics of the dwellings than in-
adequacies in measurement. Values range from 0.5 to 9 air changes per
hour. The American Society of Heating, Refrigeration and Air Condition-
ing Engineers (31) suggests values between 1/3 and 2 air changes per hour
for various conditions for infiltration alone. Infiltration implies air
passing through cracks and joints and is exclusive of air provided for
ventilation by opening a window or providing a forced draft with a fan.
Johnson et al. (5) and Gesell et al. (.3_i4.»16.».IZ) both employed one air
change per hour in their calculations of radon exposure from natural gas
and LPG. One air change per hour is conservative without being unrealistic.
Furthermore, the calculated concentrations may be readily adjusted for
other air exchange rates.
DOSIMETRY
222Rn and ^JRn Daughter Dosimetry
Once the concentration of radon in dwelling air is established,
the next logical step is to calculate the dose to an individual occupant
of the dwelling under a prescribed set of circumstances. Unfortunately,
the dosimetry of 222Rn and its daughters is neither simple nor well
resolved. Johnson et al. (5) have provided a comprehensive review of
the subject, especially as it applies to 222Rn in dwellings. Their task
was complicated by the fact that most 222Rn dosimetry calculations have
(justifiably) employed parameters typical of uranium mining. The following
is a summary based partly on Johnson et al.'s (5) review to which the
reader is referred for more comprehensive information and bibliographic
material.
29
-------
Table 10 gives a simplified relationship between 222Rn and its
daughter products. 222Rn decays into a series of short-lived daughters
which effectively terminate (insofar as airborne 222Rn and daughters
are concerned) in 20.4 year half-life lead-210.
TABLE 10
SIMPLIFIED.DECAY SERIES FOR RADON-222
ISOTOPE
Radon-222
Polonium-218
Lead-214
SYMBOL
222Rn
218Po
214Pb
HISTORICAL
NAME
Radon
Radium A
Radium B
HALF-LIFE
3.82 day
3.05 min.
26.8 min.
TYPE OF
DECAY
a
a
3
ALPHA
ENERGY
(MeV)
5.49
6.00
Bismuth-214
214
Bi
Radium C
19.7 min.
Polonium-214
Lead-210
Bismuth-210
Polonium-210
Lead-206
214Po
210Pb
210B1
210p0
206pb
Radium C'
Radium D
Radium E
Radium F
Radium G
164 x 10~6sec
2-0.4 yr.
5.0 day
138.4 day
Stable
a
3
3
a
7.69
5.30
1.
Numerous previous studies have yielded the following information
about the dosimetry of airborne radon-222 and its airborne daughters (5).
The respiratory system and more specifically, the basal cell layer of
the bronchial epithelium is the critical tissue.
The dose to this tissue is provided almost exclusively by the short-
lived daughters of 222Rn i.e., 218Po + 214Po rather than by the 222Rn
itself. This occurs because radon, a noble gas, tends not to plate
out on or concentrate in any particular tissue whereas the short-lived
daughters, all metals, tend to deposit and adhere to the surface of
the respiratory system.
Of the various radiations emitted by the short-lived daughters, the
alpha radiation is by far the most important from a dose standpoint.
The principal adverse health effect associated with elevated levels
of airborne 222Rn daughters is bronchial carcinoma.
3.
4.
The foregoing allows the dosimetry of 222Rn and its daughters to be
reduced to a study of the dose imparted to the bronchial epithelium from
the alpha emissions of the short-lived 222Rn daughters.
30
-------
Ideally, estimation of the tracheobronchial dose (T-B dose) due
to 222Rn daughters would begin with specification or measurement of
the airborne concentrations of 218Po, 21l*Pb, 21tfBi and 211*Po. Sub-
sequently the alpha dose resulting from each isotope would be calcu-
lated and the result summed. This task would be slightly simplified
by the extremely short half life of 21l|Po which would allow one to
treat the 21lfBi - 214Po pair as a single species. ' Despite this
simplification, the difficulty of measuring the three isotopes under
field conditions requires sophisticated equipment, and this led to
the introduction of a special unit, the "working level" (WL) for
specifying the collective airborne 222Rn daughter concentrations.
The "working level" was originally intended as a safe but not
unduly restrictive level of airborne 222Rn daughters for application
in uranium mining. However, the present occupational standard for
radon daughter exposure is no longer one working level but rather 1/3
WL. The WL unit is still employed as a measure of airborne radon
daughter concentrations and is equivalent to an amount of short-lived
radon daughters in one liter of air whose decay would result in the
release of 1.3 x 105 MeV of alpha energy. It is also equivalent to
the concentration of short-lived radon daughter products in radioactive
equilibrium with 100 pCi/1 of radon-222. The latter equivalency does
not imply a direct conversion from radon-222 concentration to WL,
however, because radioactive equilibrium among radon-222 and its daugh-
ters seldom exists in the field situation.
No practical means has been devised for directly measuring the
dose imparted to bronchial tissue by 222Rn daughters so the relationship
between airborne 222Rn daughter concentrations and dose is obtained by
dose modeling and calculations. All of the dose models require specifi-
cation of the degree of radioactive equilibrium which exists among the
222Rn daughters8, specification of the fraction of the radon daughters
which remain unattached to aerosols (Jfirea ions)', and the size distribu-
tion of aerosol particles. Jacobi (32) has recently shown, using the
ICRP Lung Model, that the relationship between tracheo-bronchial dose
per working level month (WLM; equivalent to exposure at one WL for one
"occupational month" or 170 hours) can vary over an order of magnitude
depending upon such factors as unattached daughter concentration and
degree of equilibrium. Thus, a knowledge of WL only is insufficient to
make dose estimations, and unattached daughter concentrations, particle
size distribution, and the degree of equilibrium must be either measured
or estimated.
8
The degree of radioactive equilibrium between 222Rn and its daughters
may be expressed as 10, f-,, fn» £3 where 10 is taken as the concentration of
222Rn, fx is the relative (to Rn) concentration of 218Po (RaA), t^ is
the relative (to Rn) concentration of 21I*Pb(RaB) and f^ is the relative (to
Rn) concentration of the 214Bi-21t+Po (RaC-RaC1) pair. Thus perfect equilib-
rium would be expressed as 10, 10, 10, 10.
9
Airborne 222Rn daughters which are unattached to aerosols are called
"free ions". The fraction of free ions for each species is usually expressed
as percentage of the total number of each species.
31
-------
Johnson et al. (5) have compiled a table summarizing the existing
work on 222Rn daughter dosimetry which is reproduced here as table 11.
The conditions used by each investigator have been normalized for con-
tinuous exposure for one year (8,760 hours) at one working level. The
results, of each investigator, expressed in rads/year are not directly
comparable because different assumed values for free ion concentration
and equilibrium conditions were employed. Furthermore, different
atmospheric dust loading and lung models were utilized. Table 11 pro-
vides information on the order of magnitude and range of the various
conversion factors. Presumably the appropriate conversion factor(s) for
the situation of radon daughters in dwellings are encompassed by the values
in table 11. Measurements of the various parameters in dwellings could
be employed to narrow the range of conversion factors.
Dose conversion factor for this study
In earlier studies on radon in natural gas, Barton, et al. (1) and
Johnson et al. (50 used the conversion factor of 100 rads per year" to
the tracheobronchial epithelium for continuous exposure at 1 WL This
corresponds to 1000 rem per WL for a quality factor of 10 for alpha
particles. Upon further evaluation, Johnson et al. (5) concluded that
several parameters involved in deriving this conversion factor were overly
conservative. Subsequently, a lower conversion factor was derived for
evaluating doses from radon emanation from uranium mill tailings piles
[As part of the environmental analysis of the uranium fuel cycle made by
EPA (42).] This lower factor is used for determining dose from radon in
LPG.
This factor was derived from data presented by Haque and Collinson
(39) and in the 1972 UN SCEAR Report (43-p. 35,81)! TneseSata are based
on exposure to radon and daughters in a home with adequate ventilation
for a period of 6000 hours. Absorbed doses were calculated for test con-
ditions with a radon concentration of 0.164 PCi/l and a daughter distribu-
tion of 0.9, 0.5, 0.35; i.e., ratio of the radon daughters RaA, RaB, RaC
(RaC ) respectively, to the radon concentration. Variations on two lung
model parameters were taken into account for these dose calculations
One was the thickness of the mucous sheath and epithelium i e the'dis-
tance for an alpha par tide to penetrate from the deposited radon daughter
to the critical basal cell nuclei of the bronchial epithelium. The other
parameter was the lung region of concern as denoted by the generation
number of the Weibel dichotomous lung model O9). j^ particular varia-
tions on these parameters selected for this study were a thickness of
60 microns and a generation number of 5 which corresponds to the segmental
bronchi region of the lung. These selections were made on the basis of
state-of-the art knowledge of lung models and may require modification
as more data become available.
10
Nelson, N.S., Office of Radiation Programs, Criteria and Standards
Division, personal communication to R. Johnson, 1974.
32
-------
Table 11. Summary of dose conversion factors for radon and radon daughters (5)
Radon-daughter
equilibria
10,10,10,10
Nonequilibrium,
little free RaA
Nonequi1ibr ium
Nonequilibrium,
1-2% free RaA
WLM = 170
500 hours per
month in homes
Clean air MPAI of
4
Epithelial base
cells of large
bronchi
Bronchial
epithelium
Revised ICRP model
(38), bronchial
25.8-51.5 BEIR (34)
34
Toth
19.3-51.5 Jacobi (32)
II
10,10,6,4(d)
10,10,10,10
25% free RaA
10,9.6,4(d)
8.5% free
RaA
H
10,9,5,3.5
10,9,6,4
High aerosol cone.
0.05-0.2um particles
Normal room air
change- 1 hr"1 ,
10,000 particles
cm-3, 0.09pm
10 pCi/l-Rn
Natural radiation
exposure, 0.09um
0.1 pCi/l-Rn
0.3ym particles
Rn-100 pCi/1,
occupational
exposure
ii
Adequately
ventilated room,
6,000 hr/yr
>0.1 ym particles
Rn-100 pCi/1
It
Finde is en-Landahl
model, bronchial
epithelium, 14 1/min.
11
Landahl model,
segmental bronchi,
15 1/min., mouth
breathing
it
Nose breathing
Segmental
bronchi, 15 1/min.,
mouth breathing
Segmental bronchi
15 1/min.
15.5-25.8 (32)
88 Jacobi Q6)
140 (36)
103 Altshuler
et al. (37)
Lund in (38)
56 (37)
89-620 Hague and
Collinson
(39)
111 Burgess and
Shapiro (4jD)
Range 12-620
(a)Dose factor » rads/year for continuous exposure (8,760 hours) to one
working level. One WL - any combination of short-lived radon daughters (through
2'"Po, RaC1) leading to a total emission of 1.3 x 105 MeV of alpha energy per
liter of air (41). One WL is also defined as 100 pCi/1 of radon in equilibrium
Wlth (b)Relativerconcentrations of 22?Rn, RaA, RaB, and RaC (RaC*).
(C)WLM, 1 working level month - 170 hours exposure at 1 WL. Jacobi (32)
defines 1 WLM as 2.6 x 1010 MeV potential alpha-energy inhaled at 20 1/min. for
166.7 hours/month. MPAI = maximum permissible annual intake.
(d)These conditions represent typical dwellings.
33
-------
The dose calculated for the above exposure conditions was 0.04
rad. This dose must then be multiplied by 1.46 for continuous ex-
posure for one year (8766 hours); by 6,1 to convert from 0.164 pCi/1
of radon to 1.0 pCi/1; and by a quality factor of 10 to convert to
dose equivalent in rem. Thus, for continuous exposure to an atmos-
phere of 1.0 pCi/1 of radon-222 with a daughter ratio of 0.9, 0.5,
0.35, the radiation dose would be 3.56 rem/yr to the bronchial
epithelium.
For comparison, the calculation of dose may also be done on the
basis of working levels of exposure by using the BEIR Report (34)
estimates to derive a dose equivalent. This is done by converting
the above radon and daughter concentrations to working levels according
to table 12.
Table 12. Calculation of Working Levels
Nuclide
En
RaA
RaB
RaC
RaC'
PCi/1
1
0.9
0.5
0.35
0.35
Atoms /I
1.77 x 104
8.79
42.9
22.1
10-6
Alpha Energy
per atom
(MeV)
13.68
7.68
7.68
7.68
Total Potential
alpha energy
(MeV)
120
329
170
619
619 MeV
1.3 x 105 MeV/WL
0.00485 WL
Exposure at 0.00485 WL for 170 working hours per month corresponds
to 0.00485 WLM (working level months). Using the BEIR Report £34) con-
version factors of 5-10 rem/WLM, this exposure corresponds to o7o~24 to
0.048 rem per working month. These doses are converted to continuous
annual dose by multiplying by 12 and by 4.3 (ratio of hours in a year to
hours in 12 working months). The yearly doses are then in the calculated
range 1.24 rem to 2.48 rem. Since the BEIR conversion factor of 10 rem/
WLM is applicable except for individuals with chronic bronchitis, then
2.48 rem/year is the better estimate for dose equivalent delivered by
continuous exposure at 1 pCi/1 of radon-222.
Of the two methods for estimating doses, the method based on Haque
and Collinson (39) gives the highest dose for the lung region of concern.
34
-------
The BEIR Report does not estimate dose for a particular part of the
bronchial tree and therefore does not emphasize a critical region,
Furthermore, the BEIR estimates are for occupational exposure to
uranium miners, and therefore do not apply directly to normal living
conditions. The cleaner air and smaller aerosol particle sizes in
homes, as opposed to mines, would result in an increased dose for the
same radon concentration (or WL) due to a larger fraction of free ions
and the fact that more small particles penetrate to the critical region
of the lung. Thus, the dose conversion factor that will be used in
this study is 3.56 rem/year, which will be rounded off to
1 pCi/1 radon-222 = 4 rem/year
for daughter ratios of 0.9, 0.5, 0.35, and a penetration depth of 60
microns to the nuclei of the cells at risk, which are the basal cells
of the segmental bronchi. This conversion factor is equivalent to 800
rem/year at one WL, which is 200 rem/year at one WL lower than the con-
version factor used by Johnson et al. (5) in the previous study on radon
in natural gas.
Postulated Exposure Conditions
The exposure conditions for determining dose to an individual
using LPG in home cooking and heating are summarized in table 13. Most
of these conditions are the same as those used by Johnson et al, (5) to
estimate the radon dose from natural gas. The main differences are in
the radon concentrations involved, the smaller quantities of LPG used,
and the lower dose conversion factor as discussed above. As noted pre-
viously, less LPG is used because this fuel has a higher B.t.u. content
than natural gas.
Using the parameters from table 13 the following general equation
may be derived for estimating the airborne concentration, CA, of radon-
222 in homes using LPG in unvented appliances.
CQ x Q x DF
CA (pCi/1) = --
A V x R
Where: C = radon concentration in LPG delivered to homes
Q = quantity of LPG used (nr/day)
DF => decay factor due to storage in home tanks (0.183 for 30 days)
V = house volume (226. 6m3'
R - air exchange rate (24 house volumes /day)
For example, a home using LPG in a kitchen range, with the highest radon
concentration measured (1240 pCi/1) and monthly deliveries, would have
an average indoor radon concentration from this source of
35
-------
Table 13. Exposure conditions employed in the
estimation of dose from radon in LPG
Parameter
Condition Used
in this Analysis3
Possible
Variation13
222Rn concentration in
delivered LPG
LPG delivery interval
Storage tank decay factor
LPG use
Cooking ranges
Space heaters
Ventilation conditions
(appliances)
Dwelling volume
Degree days
Air exchange rate
Number of persons
occupying each dwelling
222Rn an^ daughter
equi nbrium
in LPG
in dwelling air
Percent free
Critical pathway
Critical organ
Dose conversion factor
Quality factor
An average value has
been assigned to each
state (10 -150 pCi/1)
one month
0.183
0.306 m3/day
0.142
unvented
226.6 m3 (8000 ft3)
average for each state
1 dwelling volume/hour
4 (continuous occupancy)
1, 0, 0, Oc
1.0, 0.9, 0.5, 0.35
35%
inhalation of radon
daughters
bronchial epithelium
4 rem/yr for continuous
exposure at 1 pCi/1
10
0-1500 pCi/liter
2 weeks - 3 months
0.363 - 0.0613
up to 0.476 m3/day
0.112-*0.168 mydegree-day
ranges may be potentially
vented
142 - 425 m3
+ 25% within states
0.25-5 per hour
1 - 10 (or partial occupancy)
up to 1, 1, 1, l
1.0. 0.5, 0.25, 0.1 to
1. 1, 1. 1
5 - 50%
222Rn accounts for
<1% of dose
some exposure to remainder
of respiratory tract and
other organs
2-5 rem/year
1 - 10
Conditions are intended to be typical and to err on the conservative side
in the case of less well-understood parameters.
bThese variations are intended to cover a large fraction of actual exposure
conditions.
CRatio of 222Rn, 218Po, 21(tPb, 211»Bi (2i*Po).
dThis factor incorporates radon daughter equilibrium conditions, free ion
fraction, critical pathway, and dosiraetry model parameters.
36
-------
1240 x 0.306 x 0.183
226.6 x 24
= 0.0128 pCi/1
If the same home used unvented space heating, an additional 0.142 m^
of LPG would be required for each degree-day. To pick a "worst case"
example, Duluth, Minnesota requires 10,000 degree-days of heating per
year (31) . Thus, heating would result in a radon concentration of
1240 x 0.142 x 10,000 x 0.183
c = _ ! _ = 0.162 pCi/1
A 226.6 x 24
for this example of the highest radon concentration in LPG and the
coldest weather conditions. For the average daily heating requirement
in the U.S., which is 7.77 degree-days (5), the radon concentration in
this same home would be
1240 x 0.142 x 7.77 x 0.183
0.046 pCi/1
226.6 x 24
Dose to an Individual
The dose to an individual is estimated by multiplying the dose
conversion factor (4 rem/year per pCi/1 of radon) times the radon con-
centration in the home air. Individual doses for various combinations
of exposure conditions are shown in table 14.
Table 14, Doses to individuals from radon-222 in LPG
Radon-222 in, .
( a J
home tank
pCi/1
40
150
1240
1240
cooking
ranges v
0.41
6.2
51
51
Dose, equivalent
space
heaters ^c
1.48
22.3
184
648
(mr em/year)
:' Total
1.89
28.5
235
699
(a)Concentration at time of delivery to home tank
(b)Average LPG use 0.306 m3/day
(c)Average LPG use 1.1 nr/day
(d)For 10,000 degree-days/year and 1240 pCi/1 of radon-222
NOTE: It is not likely that the highest radon concentration and highest
LPG use would occur together.
37
-------
This table indicated that for average LPG radon concentrations of
10 to 150 pCi/1, the average annual dose to individuals would be less
than 30 mrem even for combined use of unventing cooking ranges and
space heaters. The maximum individual dose for the conditions of high-
est LPG radon concentration and gas use could be about 700 mrem/year
although it is not likely that both conditions would occur together.
Existing data are insufficient to determine the significance of potential
high individual doses. This matter will be further evaluated by ORP.
Population Dose
The tracheobronchial (T-B) doses to the United States population
were estimated on a state-by-state basis in the same manner as for
individaul doses. Within each State, the concentration of radon-222 in
LPG and the annual degree-day heating requirements were assumed to be
constant. The individual dose for these conditions was then multiplied
by the number of dwellings using LPG in unvented appliances and by four
(number of occupants per dwelling) to obtain the State population dose.
The number of dwellings using LPG in cooking ranges was obtained
directly from census data (24). However, determining the number of un-
vented space heaters using LPG was more involved. While census data
gave the number of dwellings with unvented heaters and also the number
using LPG fuel, these data do not directly give the number of dwellings
which use LPG in unvented space heaters. Therefore, the number was
estimated by assuming that unvented space heaters are fueled by natural
gas, LPG, or kerosene in the same proportion as all home heating systems
in each State. Fuels such as coal or wood were not considered because
they are obviously unsuited for unvented heating. The estimate was
further refined by separating each State into rural and urban components
prior to making the estimate. The urban and rural estimates were combined
for the total number of unvented space heaters burning LPG in each State.
The population T-B doses and the parameters for estimating these
doses are given in table 15. The total population T-B dose equivalent
for use of LPG in kitchen ranges was estimated to be about 18,400 person-
rem per year. The total dose equivalent for unvented space heaters was
about 10,600 person-rem year.
By assuming four occupants per dwelling, the population at risk
for exposure to radon from LPG use in kitchen ranges is about 21.3
million persons or about 10 percent of the United States population.
The average individual dose equivalent from this source is about 0.87
mrem/year. For space heaters, the potential population at risk is about
2.8 million persons or about 1.3 percent of the population. The average
dose to individuals from LPG in space heaters is about 3.7 mrem/year.
The combined population T-B dose equivalent for exposure to radon
daughters from use of LPG in unvented kitchen ranges and space heaters
is estimated to be about 29,000 person-rem per year for the United States.
38
-------
Table 15
Estimated Dose Equivalent to the U.S. Population Due to 222Rn in LPG
State
Texas
California
Florida
Mississippi
New York
Georgia
Alabama
Arkansas
Oklahoma
Missouri
New Mexico
Pennsylvania
Colorado
Illinois
Louisiana
North Carolina
Indiana
Arizona
Ohio
Michigan
Wisconsin
Minnesota
Kentucky
Iowa
Virginia
South Carol ina
Nevada
Maryland
New Jersey
Massachusetts
Tennessee
Maine
Assumed
'22Rn
Concentration
in LPG
(pCi/1)
100
150
10
10
10
10
10
10
10
10
50
10
50
10
10
10
10
50
10
10
10
10
10
10
10
10
50
10
10
10
10
10
Estimated
Dwellings
with
[Invented
LPG Heaters
160,710
9,928
80,813
97,762
787
78,318
76,886
44,012
42,074
7,675
3,441
529
1,205
2,279
- 48,956
5,389
3,512
2,201
1,092
1,516
2,134
2,116
6,189
251
832
9,238
678
408
211
186
11,248
76
Connecticut 10 127
Kansas 10
1,618
Nebraska 10 1,348
Utah ! fo 290
South Dakota ; 10
254
New Hampshire 10 69
West Virginia ; 10 1,631
Vermont 10 159
North Dakota j 10 514
Oregon ; 10 835
Washington ; 10 966
Delaware ; 10 76
Rhode Island j in ; 53
Montana 10 746
Idaho i 10
Wyoming 10
Alaska 10
497
419
33
Hawaii ; 10 ! 26
District of | j
Columbia 10
Totals
18
713,317
Estimated
Annual
Degree-
Days
1,940
2,756
742
2,190
6,266
2,435
2,368
3,015
3,792
4,921
4,646
5,531
6,310
5,899
1,627
3,284
5.694
3,295
5,840
7,372
7,779
8,892
4,869
6.870
3.776
2,336
6,194
4,617
4,794
6.518
3,485
8,639
5,916
5,282
6.681
6,109
7,802
7.383
4.838
8.269
9,303
6,045
5.368
4,930
5,879
8,094
6,128
7,585
12,097
4,617
Estimated
Population
T-B Dose-
Heaters
(person-rem/yr
6,524
859
126
447
9
399
382
278
335
79
167
6
79
28
167
38
41
75
13
23
34
39
64
4
6
45
43
4
2
2
83
2
'i
19
19
19
4
2
17
4
Q
11
11
--
13
6
8
--
10,564
Dwellings
with LPG
Ranges
336,334
164,848
425,837
147,351
377,931
135,235
104,154
143,146
108,146
258,522
32,318
256,977
38,656
192,504
108,175
185,263
181,434
29,554
175,161
157,708
149,184
144,861
125,216
142,002
137,600
87,000
17,197
109,729
100,276
90,133
36,811
75,317
59,575
45,061
8.534
40,934
42,144
31,247
35,691
25,665
24,814
24,314
26,234
21,738
14,329
12,152
11,159
14,365
12,860
7,308
5.315.373
Estimated
Population
T-B Dose-
Ranges
(person-rem/yr)
5,555
4,083
703
242
624
224
171
237
179
427
267
425
320
317
178
306
299
244
289
261
246
239
207
235
227
143
143
180
165
148
60
137
124
98
75
70
68
70
51
58
43
41
39
43
36
23
21
19
23
21
11
18,415
Total
Estimated
T-B Dose
(person/rem/yr)
12,079
4,942
829
689
633
623
553
515
514
506
434
431
399
345
345
344
340
319
302
284
280
278
271
239
233
188
186
184
167
150
143
139
126
117
94
89
72
72
68
62
52
52
50
43
36
36
27
27
23
21
11
28,979
39
-------
Possible variations in population dose estimates
The estimation of population dose was performed using nominal
values of the many parameters entering into the calculation. Different
estimates would result for other exposure conditions. To provide some
insight on how the dose estimates may be adjusted for different con-
ditions, a listing of possible corrections is given in table 16. It
is seen that reasonable variations in several parameters could result
in changes in the calculated doses by a factor of two.
POTENTIAL HEALTH EFFECTS
Although radon was not known to be the cause, health effects were
reported as excessive pulmonary disease among groups of miners in the
Joachimsthal and Schneeburg mining areas of central Europe over three
hundred years ago (47). Herting and Hesse (4£) reported the first
autopsies on deceased miners and identified malignant growth in 1879.
In 1911 these growths were demonstrated to be carcinomas of the lung*
Epidemiological studies reported by Thiele et al. (49) in 1924 and by
Peller (50) in 1939 denoted highly significant increases of pulmonary
cancers among miners. Careful epidemiological work on uranium miners
in particular, has demonstrated significant increases in bronchial carci-
nomas (47). The agents responsible for a significant proportion of
these increases have subsequently been identified as airborne radon
daughters which, upon inhalation, become attached to pulmonary surfaces
where they release their alpha decay energy into these tissues.
Conversion from Dose to Potential Health Effects"
The general approach for deriving an appropriate factor relating
health effects to dose from radon daughter irradiation of lung tissues
involves two steps. The first step is to determine an age and population
adjusted estimate of excess somatic cancer deaths. Then the estimate
of excess lung cancer deaths can be derived as a fractional part of the
somatic cancer deaths.
The estimation of excess lung cancer mortality in this study will
be based on the report by the National Academy of Science on the biologi-
cal effects of ionizing radiation (BEIR report) (34). The excess risk of
death from lung cancer due to exposure to one rem of ionizing radiation
may be determined using risk model data from table 3-1, p. 169 and table
3-2, p. 171 of the BEIR report.
The health effects conversion factor derived in this study is based
on information provided by Neal S. Nelson, Ph.D., U.S.E.P.A., Office of
Radiation Programs, in a memorandum to R. Johnson, November 14, 1974.
40
-------
Table 16
Corrections to Adjust Estimated Population Doses for
Different Exposure Conditions (5)
Parameter
Ventilation rate-air
changes per hour
222Rn concentration
in LPG
Quantity of LPG Used
Dwelling size
Daughter equilibrium
ratio
Percent free 218Po
Dose conversion factor
Quality factor
Value
0.25. .
(1.0)(a)
2.0
varies with state
(10 - 150 pCi/1)
varies with heating
requirements
(226.6 m3) (8000 ft?)
1,1,1.1
(1, 0.9, 0.5, 0.35)
1, 0.75, 0.5, 0.3
1, 0.5, 0.25, 0.1
3
10
25
(35-)
50
(4 rem/yr at 1 pCi/1)
Adjustment Multiplier
6.0
1.0
0.34
linear(b)
linear
inverse
1.9
1.0
0.75
0.37
0.27
0.38
0.77
1.0
1.3
linear
(10) 1 linear
(^Values in parentheses were used in this study.
(b)A linear correction means the correction is proportional
to the variation in the parameter.
41
-------
Data from both the absolute risk model and the relative risk model
of the BEIR report were combined to determine the best estimate of ex-
cess cancer deaths, The absolute risk model estimate is based on the
product of assumed risk times the total population at risk and gives
the number of cases that may result from a given exposure to a given
population. The relative risk model estimates excess deaths on the
basis of the ratio of the risk in those exposed to the risk versus those
not exposed (incidence in exposed population to incidence in control
population.
The age adjusted estimates from both models of excess somatic
cancer deaths for a population exposure of a million persons/year/rem
(i.e. Itr* person-rem/year) are shown in table 17, These estimates were
derived from 1967 U.S. population data after converting from 0.1 to 1.0
rem/year (table 3-1, j4). Data are given for a duration of risk (plateau
region) of 30 years and a lifetime. The data for both plateau regions
are combined in the estimate for excess lung cancer deaths by taking the
geometric mean. This was done instead of taking the arithmetic mean to
compensate for uncertainty in the lifetime plateau estimate for the rela-
tive risk model.
The estimates for excess lung cancer mortality were determined by
assuming that the cancer distribution in both models is the slame for all
ages. Then the proportion of lung cancers was taken as 26 percent of
all excess somatic cancer deaths using the data from table 3-2 of the
BEIR report.
The estimates frota both models were combined by taking the average
of the geometric means for the two plateau regions. The combined esti-
mate is 38.5 excess lung cancer mortalities for each million persons
exposed/year/rem. This estimate will be rounded off to one significant
figure to give the following health effects conversion factor for use
in this study
1 6 12
4 x 10 excess lung cancer deaths/10 person-rem.
for continuous exposure to radon daughters.
It should be noted that 100% fatality is assumed here. In addition,
it should be recognized that very few data are available on carcinogenic
alpha dose to the lung. Furthermore, the lung cancer estimates here are
based on data derived from dose levels considerably higher than antici-
pated for radon in LPG. The BIER report points out that extrapolation
to low doses for large numbers of people has inherent uncertainties such
that zero cannot be excluded as a possible conversion from radiation dose
to cancer incidence for the low dose rate conditions of this study. These
considerations should be kept in mind when evaluating the potential health
effects which may result from radon-222 in LPG.
12
This conversion factor may be modified as more data become
available from ongoing studies.
42
-------
Table 17.
Estimates of excess somatic and lung cancer deaths
Risk
Model
Excess somatic cancer deaths^3'
for 1()6 per son -rem/ year
Absolute
Relative
30 year plateau
61.2
123.1
lifetime plateau
75.0
421.5
Absolute
Relative
Combined
Models
Excess lung cancer deaths^)
for 106 person-rem/year
30 year plateau lifetime plateau geometric mean(c)
15.9
32.0
19.5
109.9
17.6
59.3
17.6 + 59.3 = 33.5
2
;aJExcluding leukemia. .
(b)Assumes lung cancers represent 26% of somatic cancer deaths.
(BEIR, Report, p. 171). e ...... *u
(cJThe geometric mean was used to compensate for uncertainty in the
lifetime plateau estimate for the relative risk model.
43
-------
Health Effects Estimate
The total dose equivalent for continuous exposure to radon
daughters from use of LPG in unvented kitchen ranges and space heaters
in the United States is estimated to be about 29,000 person-rein/year.
When the health effects conversion factor is applied to this dose the
estimate of potential health effects is
[29 x 103 person-rem/yr] [4 x 101 effects/106 person-rem] =
1.16 effects/year
or about one excess lung cancer mortality a year.
CONTROL COSTS
At the present time the most practical method for controlling the
concentration of radon in LPG would be storage to take advantage of the
relatively short half-life of radon-222 (3.83 days). Storage is already
an integral part of the LPG industry, therefore the technology is readily
available for use of storage to control radon levels. Approximately 75
percent of existing storage is underground in depleted oil wells or in
caverns within salt formations. The remaining 25 percent of storage for
LPG is in above ground tanks mainly at distribution centers or points of
LPG consumption.
Storage serves a primary function of providing the necessary balance
between a constant production rate and seasonal demands for LPG. How-
ever, at present all the available underground storage capacity is being
used. Therefore, LPG is in an oversupply status which has resulted in
recent increases in charges for LPG storage due to the storage demand.
Since existing underground storage capacity is fully utilized, new
storage requirements for control of radon could require extensive addi-
tions to above ground storage. However, for the purpose of this control
cost analysis, the assumption will be made that development of new storage
capacity would be done in the same ratio as existing above and below
ground storage, i.e. 25 and 75 percent respectively.
The following cost analysis for control of radon-222 in LPG is
intended to provide order-of-magnitude estimates to evaluate the signifi-
cance of potential health effects in relation to control costs. The
actual costs for individual storage facilities can vary widely as functions
of facility size, production rates, pumping costs, and distribution
logistics.
The costs for storing LPG will be estimated by the present worth
method for annualized costs. This method estimates the yearly capital
44
-------
cost by an annual fixed charge rate. The fixed charge rate used in
this analysis was 16 percent per year, including interest, taxes,
insurance, and depreciation for a thirty year facility life. The
sum of the annual capital costs and annual operating costs gives the
annualized costs.
The cost data for this analysis are given in table 18. The
major factor in the operating costs is that of pumping the LPG into
and out of storage. These pumping costs are approximately the same
for each storage method and average about 0.25 cents/gal for pumping
into storage and the same for pumping out.
For this analysis, the control costs will be estimated for a two
week storage of all the LPG produced in the two states (Texas and
California) with the highest average radon-222 concentrations in LPG.
Two weeks of storage will allow radon-222 to decay to less than 10
percent of its original concentration. Texas and California together
produce about 51 percent of the LPG in the United States. Their com-
bined production for two weeks is about 1.6 x 10° gal.
The annualized capital costs are calculated on the basis of
capital costs for a two week storage capacity plus annual operating
costs consisting of pumping charges for cycling the LPG through
storage every two weeks. The summary of annualized costs is shown in
table 19.
Comp_aris_qn of Radon Control Costs to
Reduction in Potential Health Effects
The estimated cost for control of radon in LPG by storage would be
from 38 to 48 million dollars a year for about half of normal United
States production. For this storage to be effective it should be applied
in the production areas with highest radon-222 concentrations, i.e. Texas
and California. Since it is not known which states use LPG from Texas
and California, it is not possible to calculate specifically the poten-
tial reduction in health effects which may be attributed to control of
radon in LPG from those two states. However, the maximum benefit that
could be derived from this control would be the elimination of estimated
health effects (about one excess lung cancer mortality a year).
Thus, if the proposed storage control could eliminate the potential
of one excess lung cancer a year, the control cost per health effect
reduction would be in the order of 38 to 48 million dollars. Since other
uncontrolled LPG, sources would continue to contribute to population lung
dose, then the potential number of excess lung cancers could at best only
be reduced to some fraction per year by storage control. Therefore the
cost per health effect reduction could be considerably in excess of 50
million dollars. Investments of this magnitude for controlling radon-222
in LPG are clearly not cost effective in relation to the possible reduc-
tion in health effects.
45
-------
Table 18. Costs for LPG Storage (1974 Basis)
Storage Method
Underground
Oil Wells
Salt Caverns
Above ground
Capital Costs
$0.02-0.03/gal/yr
$0.04-0.05/gal/yr
$0.35-0.50/gal/yr
Operating Costs
$0,005/gal
$0.005/gal
$0.005/gal
Table 19. Annualized Cost Estimate for Storage of LPG
(a)
Storage Method
Underground
(b)
Above ground
Total
(b)
Annualized Costs
$18.5 xlO6 to $22.2 x 106
19.7 x 106 to 25.8 x 106
$38.2 x 106 to $48.0 x 106
(a)
For two weeks production of LPG from Texas and
California (VL.6 x 108 gal).
(b)
75 percent storage underground, 25 percent
above ground.
46
-------
DISCUSSION
Review of Uncertainties
A detailed review of the applicable uncertainties was presented
in the earlier study on natural gas (5). Therefore, only a few of
the main points will be reviewed in this report. First, it should be
emphasized that throughout this analysis values have been assumed for
exposure conditions or population at risk that were conservative, i.e.,
so that the calculated doses or health effects would be overestimated
rather than underestimated. At the same time, however, an effort has
been made to obtain values that were realistic in order that the final
health effects estimate might be in reasonable perspective.
Possible variations in exposure conditions which could be reason-
ably encountered by large portions of the population at risk are
summarized in table 13. Factors for correcting the calculated dose to
adjust for other exposure conditions are given in table 16. Individuals
or small groups could conceivably receive doses significantly higher
than estimated in this study, but no data are available for assessing
this possibility.
The observations which initially led to this study on LPG were
that radon is separated from natural gas with the LPG fraction and that
very high radon concentrations in LPG had been measured at processing
plants. The distribution and use patterns for LPG were not known at
that time. Since radon-222 has such a short half-life, the significance
of high concentrations at the gas processing plant has to be considered
in terms of distribution and storage time prior to home use. Assessment
of these factors indicated that average radon concentrations were a func-
tion of distance from the processing plant. More importantly, this study
showed that storage in home tanks is a significant factor which allows
radon to decay to two-tenths or less of its original concentration be-
tween tank refills. Consequently, the postulated radon concentration in
LPG is quite low for most States, and the baseline level of 10 pCi/1
assigned to 41 States for dose calculations is probably high for LPG
produced in the United States.
However; as noted previously in reference to figure 10, the average
radon content of northern and eastern states may be greater than 10 pCi/1
due to use of Canadian LPG. Since no data are available on radon in LPG
consumed in these states, an estimate of the significance of dose from
Canadian LPG was made by assuming a radon content of 50 pCi/1 for LPG in
each of 16 states near the Canadian border. Another 10 states adjacent
to these were assigned values of 30 to 40 pCi/1. With these radon con-
centrations an additional population T-B dose equivalent of about 8600
person-rem/year was estimated. This additional dose could potentially
raise the total estimated health effects by 0.3 excess lung cancers per
year in the United States.
47
-------
Comparison With Other Sources of Radiation
Radon-222 in LPG is by no means the only contributor of radiation
dose to the respiratory system. Other sources of radon-222, as well
as radon-220 (thoron), contribute alpha radiation dose to the bronchial
epithelium of the lung. Doses are also received from natural terrestrial
beta and gamma radiation, cosmic radiation, internally deposited isotopes
of the thorium and uranium decay series, and potassium-40. Additional
radiation doses to the population are received from man-made sources, in-
cluding medical radiation, fallout radiation, and nuclear facilities. In
order to place the contribution of dose from radon in LPG in better per-
spective, the doses to persons exposed or individuals-at*?risk and the
doses to the U. S. population-at-risk were estimated for various radiation
sources as noted in table 20.
It is recognized that there are a number of limitions which should be
noted when attempting to compare doses from various sources. In particular,
doses to the lung or bronchial tissues are not directly comparable to whole
body doses which include doses to other tissues and organs as well. However,
the dose estimates in table 20 do allow a number of observations,
For example, it is readily seen that background radon-222 contributes
by far the greatest dose to the lung of all sources, Since the population-
at-risk for background radon includes the entire U. S. Population, then the
population dose is also the largest. Also, radon from natural gas contri-
butes almost 100 times as much dose as from LPG on a national basis.
Interpretation of Estimated Health Effects
Since no lung cancers have been reported for exposure to radon daughters
at less than 0.33 WL a year, it should be emphasized that the health effects
estimate in this study is a statistical projection only for assessing poten-
tial effects on large populations. This estimate is based on the assumption
of a linear non-threshold dose response to alpha radiation from radon daughters.
In this study, the use of LPG in both kitchen ranges and space heaters
in the average home would result in less than 1 x 1Q~5 WL. Furthermore, the
estimate for this analysis of one excess lung cancer a year is probably high
according to the analysis of uncertainties outlined in detail by Johnson,
et al. (.5) in the natural gas study. Using the approach from that study,
the health effects estimate for LPG would be reduced to much less than one
effect & year. (This estimate would not be affected significantly by con-
tributions from Canadian LPG). Therefore, it cati be concluded that the use
of LPG containing radon-222 does not contribute to the incidence of lung
cancer in the United States.
The evaluation of control costs for reducing the estimated health
effects from radon in LPG indicate that 50 million dollars or more would
be required for each health effect. Such an expenditure would not be
cost effective in terms of the possible reduction in health effects.
Consequently, it is not considered necessary to control the content of
radon in LPG.
48
-------
Table 20. Comparison of bronchial epithelium doses
from various sources
Source
Radon-222 in LPG^
kitchen ranges
space heaters
Total
Radon-222 in natural gas^ '
kitchen ranges
space heaters
Total
Background radon-222
at 0.13 pCi/lW)
Natural terrestrial
radiation
Cosmic radiation
Internally deposited
potassium-40
Fallout (1969) ^
Diagnostic , >
radiography (1970) 18;
Nuclear power
reactors' ^
(a)
Average
individual
dose
mrem/yr
0.87
3.70
4.57
12.0
43.2
55.2
520
40
44
17
4
153
0.056
U.S. Population^
dose (c)
person-rem/yr Ref
18,400
10,600
29,000
1,499,200 C5)
683,200
2,182,400
109,200,000 (1)
8,400,000 (44)
9,240,000 (44)
3,570,000 (45)
840,000 (45)
11,536,200 (45)
1,650 (46)
to persons exposed or individuals-at-risk.
>bjSummation for U.S. population-at-risk for each source of exposure.
References from which dose information was derived.
conversion factor, 1 pCi/1 - 4 rem/yr.
including possible contribution from Canadian LPG.
.Lung dose assumed equal to whole body dose.
^ Lung dose assumed equal to abdominal dose.
49
-------
CONCLUSIONS
The conclusions from this assessment of potential radiological
health effects from radon in liquefied petroleum gas are summarized as
follows:
(a) The use of LPG containing radon-222 in average homes with
unvented kitchen ranges and space heaters does not contribute
to lung cancer incidence in the United States.
(b) Controls for reducing radon concentrations in LPG by storage
methods would cost over $50 million for a reduction of one
potential excess lung cancer. Therefore, it would not be cost
effective to require controls on radon in LPG by storage on a
national basis.
(c) High concentrations of radon-222 have been measured in LPG, but
distribution and storage times allow most of the radon to decay
prior to use of the LPG.
(d) The average population tracheobronchial dose-equivalent was
estimated to be 29,000 person-rems/year.
(e) The average dose to individuals and the U, S. population-at-risk
from radon in LPG is very small compared to other natural and
man-made sources of ionizing radiation.
(f) Individuals could possibly receive dose equivalents greater
than 500 mrem/year. However, existing data are not sufficient
to determine the significance of such potentially high indi-
vidual doses. This matter will be further evaluated by ORP.
50
-------
REFERENCES
1. Barton, C.J., R.E. Moore, and P.S. Rohwer, Contribution of radon in
natural gas to the natural radioactivity dose in homes, ORNL - TM -
4154 (April 1973).
2. Barton, C.J., R.E. Moore and P.S. Rohwer, Contribution of radon in
natural gas to the dose from airborne radon daughters in homes.
Proceedings of the Noble Gases Symposium, Las Vegas, Nevada, 24-28,
September, 1973 (in press).
3. Gesell, T.F., Radiological health implications of radon in natural
gas and natural gas products - An interim report, Institute of
Environmental Health, the University of Texas, Health Sciences
Center at Houston (April 17, 1973).
4. Gesell, T.F., Radiological health implications of radon in natural
gas and natural gas products - Final report, Institute of Environ-
mental Health, the University of Texas Health Sciences Center at
Houston (April 1974).
5. Johnson, R.H., Jr., D.E. Bernhardt, N.S. Nelson and H.W. Galley, Jr.,
Assessment of potential radiological health effects from radon in
natural gas. U.S. Environmental Protection Agency report, EPA-520/1-
73-004, Office of Radiation Programs, Waterside Mall, Washington, D.C.
20460 (November 1973).
6. Johnson, R.H., Jr., D.E. Bernhardt, N.S. Nelson and H.W. Galley, Jr.,
Radiological health significance of radon in natural gas, proceedings
of the Noble Gases Symposium, Las Vegas, Nevada, September 24-28,
1973 (in press).
7. Bell, K.G., Uranium and other trace elements in petroleums and rock
asphalts, Geological Survey Professional Paper 356-B, (1960).
8. McKelvey, V.E., O.L. Everhardt, and J.D. Garrels, Summary of hypoth-
eses of genesis of uranium deposits, contributions to the Geology of
Uranium and Thorium, by the United States Geological Survey and
Atomic Energy Commission for the United Nations International Con-
ference on Peaceful Uses of Atomic Energy, Geneva, Switzerland,
1955, compiled by L.R. Page, H.E. Stocking, and H.B. Smith, Geological
Survey Professional paper 300 (1956).
9. Tokarev, A.N. and A.V. Shcherbakov, Radiohydrogeology, tanslated
from a publication of the State Publishers of Scientific - Technical
Literature on Geology and Conservation of Natural Resources, Moscow,
AEC - tr - 4100, (1956).
51
-------
10. Satterly, J. and J.C, McLennan, The radioactivity of the natural
gases of Canada, Transactions of the Royal Society of Canada. Vol.
12, Series 3, p. 153-160 (1918).
11. Faul, H., C.B. Gott, G.E. Manger, J.W, Mytton and A,Y. Sakakura,
Radon and helium in natural gas, International Geological Congress.
19th, Algiers, Comptes rendus, Sec. 9, Pt. 9, 339-348 (1952).
12. Pierce, A.P., J.W. Mytton and G. B. Gott, Radioactive elements and
their daughter products in the Texas Panhandle and other oil and
gas fields in the United States. Geology of Uranium and Thorium,
International Conference (1955).
13. Pierce, A, P., G.B. Gott and J.W, Mytton, Uranium and helium in the
Panhandle gas field, Texas, and adjacent areas, U.S. Geological
Survey Professional Paper 454-G (1964).
14. Bunce, L.A. and F.W. Sattler, Radon-222 in natural gas, Radiological
Health Data and Reports, Vol. 7, No. 8, 441-444 (August 1966).
15. McBride, J.R. and D. Hill, Offsite radiological surveillance for
project gasbuggy, Radiological Health Data and Reports Vol. 10,
535-546 (1969).
16. Gesell, T.F., C.A. Mullins and N.D. Burgess, Radiological health
implications of radon-222 in retail liquefied petroleum gas. A
paper presented to the 18th Annual Meeting of the Health Physics
Society, Miami Beach, Florida (June 17-21, 1973).
17. Gesell, T,F., Some radiological health aspects of radon-222 in
natural gas and natural gas products. Proceedings of the Noble
Gases Symposium, Las Vegas, Nevada, 24-28 September 1973 (in press).
18. Barton, C.J., Radon in air, natural gas and houses, ORNL Central
Files 71-5-48 (May 29, 1971).
19. Gotchy, R.L., Project Rulison final operational radioactivity report
production tests, NVO-112 (February 1972).
20. NERC-LV (SWRHL), Offsite radiological safety and resident evacuation
program for Project Rulison, SWRHL-94r (1970).
21. Results of sampling natural gas wells in the vicinity of Project Gas-
buggy, NERC-LV-539-9 (February 1973).
22. Fries, B.A. and K.H. Kilgren, Radon in natural gas and gas liquids,
Chevron Research Company, California (June 1972).
23. Cannon, R.E., Natural Gas Processing - 72, Oil and Gas Journal.
(July 10, 1972).
52
-------
24. U.S. Bureau of the Census, Census of Housing: 1970, Detailed Housing
Characteristics, Final Report HC(1)-B 1, United States Summary (1970).
25. Butane - Propane News, p. 12 (March 1974).
26. Lucas, H.F., Improved low-level alpha-scintillation counter for radon,
Review of Scientific Instruments. Vol. 28, 680-683 (1957).
27. Baily, D.E., Probability and Statistics, John Wiley and Sons,
New York, p. 393 (1971).
28. Gas Engineers Handbook, The Industrial Press, 1st ed., (1966).
29. Jacobs, D.G., E.G. Struxness, M.J. Kelley and C.R. Bowman, Consider-
ation of the radiological impact from the hypothetical use of contam-
inated natural gas from nuclearly stimulated reservoirs, Health
Physics. Vol. 22, 429-440 (May 1972).
30. Handley, T.H. and C.J. Barton, Home ventilation rates: a literature
survey, ORNL-TM 4318 (September 1973).
31. American Society of Heating, Refrigeration and Air Conditioning
Engineers, Guide and Data Book: Systems (1970).
222
32. Jacobi. W., Relations Between the Inhaled Potential a-Energy of Rn-
and 22"Rn-Daughters and the Absorbed a-Energy in the Bronchial and
Pulmonary Region, Health Physics, 23, 3-11 (1972).
33. Harley, N.H. and B.S. Pasternack, Alpha Absorption Measurements
Applied to Lung Dose from Radon Daughters, Health Physics, 23,
771-782 (1972).
34. The Effects on Populations of Exposure to Low Levels of Ionizing
Radiation, Report of the Advisory Committee on the Biological Effects
of Ionizing Radiations, Division of Medical Sciences, National
Academy of Sciences, National Research Council, Washington, D.C. 20006
(November 1972).
35. Toth, A., Determining the Respiratory Dosage from RaA, RaB, and RaC
Inhaled by the Population in Hungary, Health Physics. 23, 281-289
(1972).
36. Jacobi, W., The Dose to the Human Respiratory Tract by Inhalation of
Short-lived 222Rn-and 222Rn-Decay products, Health Physics, 10
1163-1174 (1964).
37. Atshuler, B., N. Nelson and M. Kuschner, Estimation of Lung Tissue
Dose from the Inhalation of Radon and Daughters, Health Physics, 10
1137-1161 (1964).
53
-------
38. Lundin, F., J. Wagoner and V. Archer, Radon Daughter Exposure and
Respiratory Cancer: Quantitative and Temporal Aspects, NIOSH-NIEHS
joint monograph No. 1, Springfield, Virginia, NTIS (1971).
39. Haque, A.K.M.M. and A.J.L. Collinson, Radiation Dose to the
Respiratory System Due to Radon and its Daughter Products, Health
Physics. 13, 431-443 (May 1967).
40. Burgess, W.A. and J. Shapiro, Protection from the Daughter Products
of Radon through the use of a Powered Air-Purifying Respirator,
Health Physics, 15, 15-121 (1968),
41. Guidance for the control of radiation hazards in uranium mining,
Report No. 8 Revised, Staff Report of the Federal Radiation Council
(September 1967).
42. Environmental analysis of the uranium fuel cycle, Part I - fuel
supply. U.S. Environmental Protection Agency, Office of Radiation
Programs, EPA-520/9-73-003-B (October 1973).
43. United Nations. A report of the United Nations Scientific Committee
on the Effects of Atomic Radiation to the General Assembly, with
Annexes; Vol. I: Levels. United Nations, New York, New York (1972).
44. Oakley, D.T., Natural Radiation Exposure in the United States, U.S.
Environmental Protection Agency Report #ORP/SIB 72-1 (1972).
45, Klement, A.W., Jr., C.R. Miller, R.P. Minx and B. Shleien,
Estimates of Ionizing Radiation Doses in the United States 1960-2000,
U.S. Environmental Protection Agency Report #ORP/CSD 72-1 (1972).
46. Martin, J.A., Jr., M.A. Culliton and C. Robbins, Calculations of
Doses, Population Doses and Potential Health Effects Due to Atmos-
pheric Releases of Radionuclides from U.S. Nuclear Power Reactors
in 1972, Radiation Data and Reports 15_, 476-489 (1974).
47. Silk, H., Acta Unio Intern. Contra Cancrum j> (1950).
48. Thiele, H. , Rostoki, Saupe and Schtaorl, Muench, Med. Wochschr, Jj24,
(1924).
49. Peller, S., Human Bio. 11:130 (1939).
54
------- |