-------
DETERMINATION OF TRACE NOBLE GASES IN AIR AND NATURAL GAS*
J. C. Newton, F. B. Stephens, R. K. Stump
Lawrence Livermore Laboratory,
University of California
Livermore, California 94550
Abstract
A method was developed for the analysis of air and natural gas samples containing trace amounts of noble
gases. The gas samples are preconcentrated by reaction with calcium at 900-1,000°C and analyzed by mass
spectrometry on an automated instrument. Methods of relating the concentrated sample to the original
sample and the preparation of gas standards are given. The accuracy of the technique at the 25ppm level was
determined. The precision of the method and requirements for improving the technique are also discussed.
ACKNOWLEDGEMENTS
Special thanks go to Richard Crawford for the routines he wrote for the PDP-7 computer and for assisting
with the mass spectrometer. The support of the Chemistry Mechanical Technician Division is gratefully
acknowledged, especially the work of Tony Echeverria, LeRoy Schrawyer, and Jim Pastrone. Harold
Crampton efficiently performed the glass blowing. The coordination of electronic support and design of
the electronic control panel by Arnie Kirkewoog was especially appreciated.
INTRODUCTION
In the underground testing of nuclear devices at the Nevada Test Site (NTS), He, Kr, and Xe gases are added
as tracers or are generated during the detonation. Gases, consisting principally of air, are pumped from the
cavity to the surface and are sampled. For the last few years we have routinely analyzed these cavity gases to
determine the noble gas concentrations at the ppm level (Cady and Cady, 1945).
More recently we had a need to determine noble gases at the sub-ppm level in natural gas. This work resulted
from the Lawrence Livermore Laboratory's effort to stimulate natural gas production with nuclear explosives.
In the Rio Blanco event three nuclear devices were detonated simultaneously. Cylinders of Kr, Xe, and Ne were
emplaced with the lower, middle, and upper devices, respectively. We had to determine the background of
noble gases in gas samples from gas wells prior to the detonation and then analyze the post-shot gas samples
quantitatively for the noble gas concentrations. From this data the Laboratory hoped to assess the extent of
gas communication between the three cavities.
Ordinarily an analytical mass spectrometer can detect about 100 ppm of a component in a gas mixture. The
most common method of lowering this detection limit in the case of samples containing noble gases is to
remove the active gases by gettering them with hot calcium (Cady, et a/.,1945).1This method also eliminates
interference from mass peaks due to the presence of active gases. Horton (1973) has used the gettering of air
samples with a titanium sponge to lower the detection limit. We have used both methods and found gettering
with Ca faster than pumping with a titanium sublimation pump (TSP). H2, CO2, and hydrocarbons are
pumped rather slowly by hot titanium. In the case of NTS samples, gettering with Ca provides a 100-fold
increase in the concentration of the noble gases. With natural gas samples the method typically gives a 104 -
fold increase in their concentration.
To calculate final results in ppm or ppb, the concentration factor2 must be determined. For the NTS samples
this is done using PVT measurements before and after removing the active gases from an aliquot with a TSP.
In the case of natural gas samples, the percent of total gas pressue due to noble gases is found by analyzing the
original sample before concentration on the mass spectrometer.
To calibrate the mass spectrometer, gas standards are prepared from normal air which has been gettered
with Ca. The composition of normal air is obtained from literature values (Eck, 1969). Pure gases are also used
as standards.
1 These cavity samples are referred to as Nevada Test Site (NTS) samples throughout this paper to distinguish
them from normal (atmospheric) air samples.
ppm total noble gases after gettering
Concentration factor = , , , ,, , „ ...
ppm total noble gases before gettering
* This work was performed under the auspices of the U. S. A tomic Energy Commission.
-218-
-------
APPARATUS AND PROCEDURE
1. Analysis of NTS Samples.
The three steps in the analysis are:
(a) determining the percent of noble gases in the original sample with a TSP using a small aliquot;
(b) concentrating the noble gases from about 800ml STP of sample gas by gettering active gases with hot
calcium: and
(c) analyzing the concentrated sample by mass spectrometry.
The percent noble gases in the original NTS sample is determined with the apparatus shown in Figure 1. The
pressure transducer is a CGS/Datametrics 10-torr Barocel.3 The TSP with Ti-Mo filaments has a Varian
Model No. 922-0032 power supply. The TSP was built at LLL to minimize the volume. It is cooled by water
circulating in coils around the circumference of the pump. The sample bulbs are constructed of stainless steel
and have a volume of approximately 800 ml. Fifteen grams of CA shot are loaded into the bulb. A Nupro 8BK
valve is attached to the sample bulb with Cajon 8VCR fittings.
The entire system, exclusive of the sample cylinder, is evacuated to about 10-6 torr. The sample bulb is filled
to a pressure of about 1 atm. with sample gas and removed from the manifold. The determination of the percent
of noble gases is made with the gas remaining in the vacuum cross. About 100 millitorr of gas is admitted to
the manifold and 3-liter volume with the TSP valve closed. This initial pressure is recorded. Then the gas in the
manifold and 3-liter volume is introduced into the TSP. The TSP filament is operated at 42 amps for one minute
or less if a steady pressure value is attained. The final pressure is recorded, and the percent of total noble gases
in the sample is calculated from the known volume ratio of the manifold and manifold plus TSP. The gas in the
sample bulb is then gettered for 15 minutes at 900 to 1,000°C with a resistance furnace of our design. The
sample bulb is cooled and transferred to the mass spectrometry laboratory.
The mass spectrometer used for the gas analysis is a CEC Model 21-103C with a multiple, automated gas
inlet system, which can be operated under time-share control of a PDP-7 computer. To achieve a high
sensitivity the ionizing current (electron collector current) is set to 70 microamps — rather than the usual 30
microamps. Also, we use a sample pressure of about 300 millitorr in the expansion reservoir compared to about
50 millitorr used for routine gas analysis. The ion signal is detected with a Keithley Model 640 Vibrating
Capacitor Electrometer with an input resistor of 2 x 1010 ohms. Then the signal is filtered with a 0.2 Hz active
filter and amplified with a Hewlett-Packard Model 2470A Data Amplifier. A gain of 100 gives a good signal-to-
noise ratio. The effect of these operating conditions is to increase the mass spectrometer sensitivity (recorder
divisions/millitorr) by a factor of about 200 over the usual sensitivity. At the same time the detection limit is
lowered to about 1 ppm.
Samples are run on a routine basis in duplicate. Typically, a batch consists of two air standards and four air
samples in duplicate. Calculation of the results is performed with a program for the CDC-6600 computer. A
typical computer output is shown in Table 1. Note that the difference between the observed and calculated
pressures is listed to serve as an internal consistency test.
2. Natural Gas Analysis.
The apparatus used to concentrate the natural gas samples is shown in Figure 2. Because the total
concentration of noble gases may be only about 100 ppm, several aliquots (about l,500torr-litereach)of gas
must be concentrated before sufficient sample is available for analysis by the mass spectrometer. These
aliquots are transferred from the concentrating chamber to the sample bulb by the English transfer pump and
the Toepler pump. It is also possible to transfer samples which have been partially concentrated from the
sample bulb back to the concentrating chamber for a final quantitative gettering.
It is very important to have the system free of vacuum leaks for this work. Of course, any atmospheric leak
will contain 1% Ar. However, there is an internal check available to establish if vacuum leaks have occurred.
The ratio of 40Ar/36Ar for the natural gas samples differs from the same ratio for air; an examination of the
consistency of these ratios is a valuable check on the integrity of the vacuum system.
While gettering the natural gas samples, the pressure is monitored with the pressure transducer. When the
pressure reaches a constant minimum value, the oven is removed and the concentrator is allowed to cool to
room temperature. This cooling is necessary in order to prevent the dissociation of the various Ca compounds
into Ca and the respective gases. For example, if the concentrator is not sufficiently cooled before the gas is
transferred with the Toepler pump, large amounts of H2 are observed.
After several aliquots of natural gas have been concentrated and transferred, the pressure in the sample
bulb can be observed. It is seen from Figure 2 that the pressure transducer can be used to measure the pressure
in almost any part of the system. This is very useful for leak checking, too. Volumes throughout the system
have been minimized.
Results for the noble gas content in two different samples of natural gas are given in Table 2.
3Reference to a company or product name does not imply approval or recommendation of the product by the
University of California or the U.S. Atomic Energy Commission to the exclusion of others that may be
suitable.
-219-
-------
3. Detection Limits, Precision and Accuracy.
The detection limits for the noble gases as a function of their original concentration are listed in Table 3. The
second line corresponds to values for air samples; the third line represents a typical value for natural gas. Note
that the first three lines are based on a sample pressure of 300 millitorr in the expansion reservoir, and the
fourth line refers to a pressure of about 30 millitorr. By concentrating more aliquots of the natural gas sample
the detection limits are lowered as shown in Table 4.
The relative standard deviation (r.s.d.) for total noble gas concentration is less thanil% of air standards
gettered with a TSP. Four batches with a total of 18 normal air samples were gettered with hot calcium to
evaluate the overall precision of this method; i.e., the precision value includes random errors due to the
concentration and analysis steps. The r.s.d. values for the noble gases are listed in Table 5. This precision has
been confirmed by over four years of experience in determining trace amounts of noble gases by this method.
We routinely check the precision and reliability of our measurements by analyzing two air standards with
each batch of samples. One air standard is assumed to be an "unknown" and its noble gas concentrations are
calculated based on the other air standard.
The possible existence of systematic errors in the method was investigated by determining mass
spectrometer sensitivities with pure noble gases and with the concentrated air standards. The sensitivities, so
determined, typically agree to within ±1%, i.e., are within the precision of the method. In another accuracy
check two mixtures of 3He in argon were prepared and analyzed. The results are shown in Table 6. The agree-
ment between calculated and experimental results was very good.
A check on the internal consistency of the mass spectrometric analysis is the agreement between the
measured sample pressure in the expansion reservoir and the sum of the calculated partial pressures as shown
in Table 2 on the computer printout. Typically, this pressure closure is within ±1%.
4. Method Improvement.
Attempts to getter air samples on-line with the mass spectrometer were not successful because of insufficient
cooling of the hot bulbs. This lack of rapid quenching resulted in N2,02, and H2 being dissociated from their
respective compounds. Also, the ratio of the Kr/He peaks was not constant. However, a horizontally movable
oven which can be quickly removed from the hot bulb, and which is also amenable to remote operation, has
now been built.
Because the mass spectrometer is under computer control, it should be possible to automate the entire
analysis. The additional time required for temperature equilibration could be gained by overnight runs. A
program has already been written that automatically analyzes the concentrated gas samples. The
determination of the percent of noble gases present could be accomplished by analyzing a sample of the
original gas on the mass spectrometer rather than using the TSP. A synthetic mixture of noble gases could
serve as a standard. Several bulbs could be heated by moving the furnace between them in the course of the
analysis of several samples. An enlarged memory core on the time-shared computer should make possible the
on-line calculation of the sample compositions.
REFERENCES
Cady, G. H., and H. P. Cady., (1945), Ind. and Eng. Chem., Anal. Ed. 17, 760.
Eck, C. F., (1969), J. Chem. Ed., 46, 706.
Horton, J. C., (1973), Mass Spectrometric Analysis of Krypton and Xenon in Low Concentrations Oak
Ridge Gaseous Diffusion Plant, Kept. No. K-1843.
-220-
-------
TABLE 1. Computer Output Format.
This is Sample No. 3
Pot 32-A (31813)
M/E
3He
"He
Ne
Ar
Kr
Xe
TXe
PPMARCalcbyDiff-
PPM Inerts by TSP -
PPM AR calc on 36 PH is
Sample Pressure is
Closure for Sample is
PPM Based
on Air Std
.0005
5.2163
17.6313
8974.3960
3.7884
.0787
.0831
8970.98
8997.70
8974.40
316.90
.12
PPM Due
to Air
5.0330
17.4905
8997.6963
1.0950
.0836
.0836
and Calc Pressure is 317.02
TABLE 2. Results for Noble Gas Concentrations
in Natural Gas.
"He
Ne
Ar
Kr
Local
Natural
Gas
(PPM)
440
—
100
.006
Fawn
Creek #1
Well
(PPM)
80
<.015
11
.001
Xe
.005
.001
TABLE 3. PPB Detectable as a Function of Total PPM of Noble Gases.
Total
Inerts
(PPM)
106(Gross)
104
102
PPB Detectable
Cone.
Factor
10°
102
104
He
5,000
50
0.5
Ne
30,000
30
3.0
Ar
600
6.0
0.6
Kr
800
8.0
.08
Xe
1,000
10.0
0.1
Based on a 300 \i sample pressure
102 , 104 5 30
Based on a 30jl. sample pressure
0.6
0.8
1.0
-221 -
-------
TABLE 4. PPB Detectable Krypton vs.
No. of Aliquots Concentrated.
Aliquot
No.
PPM
Total
Inerts
Sample
Pressure
Mass Spec
(Millitorr)
Kr-
Detectability
in
PPB
1
2
3
4
5
6
90
22
44
66
88
110
132
.50
.25
.16
.13
.10
.08
TABLE 5. Relative Standard Deviation
Based on Analysis of Concentrated
Air Samples.
Relative Std. Dev.
Level in Air
(PPM)*
He
Ne
Ar
Kr
Xe
±2%
+ 2%
±1%
±1%
+ 15%
5.24
18.21
9340.
1.14
.087
* Literature Value
TABLE 6. Accuracy of 3He at 350
and 25 PPM levels.
Std. A Std. B
Calculated PPM
3He 350. 25.3
PPM Determined -
Cone. Air Std. 352. 35.4
PPM Determined -
Pure Gas Std. 346. 25.0
-222-
-------
Sample
Bulb
to
CO
Pressure
Gauge
JKL
Sample
Cylinder
Pressure
Transducer
V
Vacuum
3-Liter
Volume
TSP
Figure l.Apparatus to Determine Percent of Noble Gases in Sample.
-------
Concentrated
Sample
Gas
Oven
to
to
Pressure
Transducer
Toepler
Pump
English
Pump
Figure 2. Vacuum Line to Concentrate Natural Gas.
-------
PORTABLE APPARATUS AND PROCEDURE FOR THE SEPARATION OF KRYPTON,
XENON, AND METHANE IN AIR
F.B.Johns
National Environmental Research Center-Las Vegas
U.S. Environmental Protection Agency
Las Vegas, Nevada
Abstract
A portable apparatus and procedures are described for the rapid separation, collection, and counting of
radio-krypton, -xenon, and -methane in atmospheric air. The apparatus consists of a series of adsorbing and
chromatographic columns. Subsequent elution of the krypton, xenon, and methane at various temperatures
provide the separation.
The detection of the various gases is accomplished with the use of a unique thermistor detector circuit. After
volume determination, the krypton, xenon, and methane are dissolved in a degassed toluene-base, liquid
scintillation solution, and counted in a small commercial liquid scintillation counter.
With this apparatus and separation method, the krypton, xenon, and methane from a one-m3 air sample can
be separated in 2 to 2.5 hours, a short enough time that the 76-minute, 2.8-hour, and 4.4-hour half-life krypton
isotopes can be detected with a lower detection limit of 10 pd/m3. Krypton-85, xenon-133, xenon-135, and
methane can be detected at the2pd/m3 level.
INTRODUCTION
The construction and use of a portable apparatus for the separation of krypton, xenon, and methane from
atmospheric samples is described. The apparatus, and its associated counting system, can be readily
transported to any site where electricity is available or can be provided. The need for this device became
apparent in 1971 with the proposal to do on-site investigations of nuclear power reactors. The radioactive
noble gases, with the exception of krypton-85, have short half-lives, and it follows that the time for
transporting the samples to a laboratory for analysis will inhibit the collection of the necessary data
concerning these isotopes. With the system described below, half-lives of two hours may be easily quantitated.
Detectable limits of 2 pCi/m3 are attained for krypton-85, xenon-133 or 135, and tritiated methane, with
recoveries of 85% of krypton, 70% of xenon, and 40% of methane. This system is a portable version of the
separation system described by Stevenson and Johns (1971) using the counting method described by Shuping,
etal.,(l969).
APPARATUS
The apparatus consists basically of three gas chromatographs adapted to accept the large sample.
Photograph 1 portrays the system as it is presently being used, and Figure 1 is a flow diagram of the system.
The cabinet can be disassembled into two portions. The lower portion contains the mechanical vacuum pump
(133 liter/min at 0.1 micrometer) and storage. A filter is attached to the discharge of the pump to collect oil
fumes. A shelf is provided to support the helium carrier bottle.
The upper portion contains the separation section, the collection system, and the electronic controls and
recorder, as described below.
1. The Separation Section.
This section consists of three portions: a sample introduction-noble gas collection portion, and two
chromatographic separation systems. Toggle-operated vacuum valves (Nupro 4B K-T) are used throughout
the apparatus. The sample introduction-noble gas collection system consists of a molecular sieve trap (200
grams of one-eighth x three-sixteenth inch, 13X pellets), a pre-cooler, a charcoal trap (100 grams, 16-20 mesh
activated charcoal), a vacuum gauge (0-760 mm of Hg), and a thermistor detector. The traps are all glass as
illustrated in Figures 2 and 3, and are attached to the metal system with metal ball joints. These ball joints
havehad an O-ring groove added, and are silver-soldered to a short length of flexible metal tubing (Figure 4).
The O-ring eliminates the use of lubricant on these connections. The flow rate of the sample is controlled by
maintaining a pressure of 350 mm of mercury on the system, using the control valve on the sample bottle, and
a vacuum on the outlet of the charcoal trap. The thermistor detector is used to indicate the final elution of the
gases from the charcoal trap. A source of helium carrier gas is provided through a flow meter as indicated in
Figure 1.
The two chromatographic separation systems consist of two molecular sieve columns, (100 gm 30-60 mesh
5A molecular sieve), thermistor detectors on the outlet of each of the columns, and a helium flow meter.
-225-
-------
2. The Collection System.
The collection system consists of a mini-trap (C-2) (6 in. of 1/8" copper tubing packed with 0.3 grams of
chromatogr aph-grade 30-60 mesh charcoal), a liquid scintillation vial with luer fitting (Figure 5), and a digital
manometer (0-100 mm of Hg). A special valve (Figure 6) is used to direct the flow from the second molecular
sieve trap either to vent or to the mini-trap. The total volume of the system is kept as small as possible.
3. The Electronic System.
Figure 7 shows the schematic wiring diagram of the system. Two types of thermistor cell blocks are used;
one, on the outlet of the first charcoal trap, is a two-thermistor cell, reference, and detector; and the other, a
three-cell block, is located on the outlet of the first and second molecular sieve trap. It consists of three
thermistors, one reference, and two separate detectors. Figure 8 shows the plumbing of these two cells. During
operation of the system a current of 150 mA is maintained on the cell being used by means of an adjustable
potentiometer (R-2). By switching (S-l) to the proper cell position the flow through the cell may be monitored.
The output is fed into a 1 m V recorder through the attenuating switch (S-2).
4. Counting Apparatus.
The liquid scintillation counter used with this apparatus is a Beckman Beta Mate II. This is a single sample,
single channel, manually-operated counter. Now permanently mounted on a little red wagon (Photograph 2),
the counter has been modified for easier transport and use. The lead shield bricks have been bolted in place
and additional shielding added to cover the pre-amps. The instrument, as received, has a five-digit printout
which was modified to add a sixth digit. A time-of-day clock was added; also the original timing circuit was
changed to increase counting time capabilities. Prior to transport, the preamps and photo tubes are removed
in total darkness, and shock mounted in a separate case.
PROCEDURE
1. Principle of the Method.
This method describes a procedure for the separation of various gaseous radionuclides from gross air
samples. The air samples are received either as a "grab" sample, a "cryogenic" sample, or an "integrated"
sample. The grab sample represents approximately 10 cubic feet of air collected at a flow of 15 cubiefeet per
minute. The cryogenic sample represents an integrated sample collected at a flow of 3 to 4 cubic feet per minute
for one hour. The integrated sample represents a continuous sample of approximately five cubic meters
collected at a single sampling point for one week. The sample is transferred to the gas analysis apparatus.
Water and carbon dioxide are removed in a molecular sieve trap. The krypton, xenon, and methane are
separated by elution through a molecular sieve column at various temperatures. The volumes of the separated
gases are measured for yield determination, and transferred to appropriate counting chambers.
Reagents: Alcohol bath, -32°C
Charcoal, 16-20 mesh
Molecular sieve 5A, 30-60 mesh
Molecular sieve 13X, 1/8" x 3/16" pellets
Liquid Nitrogen
Liquid scintillation cocktail, 0.6g POPOP, 10.0 gPPO,
2,000 ml scintillation grade toluene
Dry ice
Acetone
Helium
Xenon carrier
Krypton carrier
a. Initial Preparation.
All traps are degassed at 350°C and evacuated until a pressure of < 104 mm of mercury is obtained. The
traps are then filled with helium; the thermistor cells are zeroed with a flow of helium. The pre-cooler, C-l,
MS-1, and MS-2 are cooled with liquid nitrogen (LN).
b. Sample Transfer.
Because of the different types of samples, the transfer of the sample will be treated separately
(l)Grab.
Record weight and pressure of the sample bottle. Connect bottle to the sample inlet port and place in a
heating mantle. Using a vacuum pump on exit from C-l, and suitable valving, establish sample flow
through the molecular sieve trap, pre-cooler, and C-l of about 15 liters per minute and 35 cm pressure.
(Reduced pressure is necessary to avoid condensation of liquid air in system.) Continue bleeding sample
until the pressure drops to less than 10 mm of mercury. Shut off sampling inlet port and add the carriers.
-226-
-------
(2)Cryogenic.
Remove the sampler from the 25-liter liquid nitrogen dewar, and place in a furnace capable of reaching
350°C in 45 minutes; attach helium line to inlet of sampler and outlet to sample inlet port. Check for
leaks. With suitable valving, use needle valve on helium inlet to establish flow through pre-cooler and C-
1 of 15 to 20 liters/minute at 35 cm helium pressure with roughing pump. Continue adding sample until
the molecular sieve sampler is at 350°C; hold for 30 minutes, shut helium valve and sample inlet port,
and add carriers.
(3) Integrated.
Record weight and pressure of the sample bottle. Connect the sample bottle to the sample inlet port.
Using the vacuum pump on the exit from C-l, and suitable valving, establish sample flow through
molecular sieve, pre-cooler, and C-l of about 15 liters per minute and 35 cm absolute pressure. Continue
bleeding sample into C-l until the pressure drops to less than 10 mm of mercury. Shut off sampling inlet
port. As the sample has had 1 ml stable xenon carrier added before sampling, and the one cubic meter of
air contains 1.14 ml stable krypton and 1 ml methane, no further carriers need be added.
c. Water Removal and Recovery.
The water and carbon dioxide are collected in the molecular sieve trap and may be recovered by heating.
d. Air Removal from C-l,
Close valves C and B, open valve D with C-l in LN; establish helium flow (600-800 ml/min) through C-l,
thermistor 1, vent. Remove LN from C-l, and replace with dry ice acetone (DIA) slush. Continue this flow
until all of the air is removed — as evidenced by a return of the pen recorder to the baseline
(approximately 55 minutes). Shut vent valve and helium flow.
e. Removal of Krypton, Xenon, and Methane from C-l.
(l)LeaveDIA on C-l and re-establish helium flow C-l, thermistor 1, MS-1, Vent 2. MS-1 and MS-2 are in LN
when flow is stabilized. Remove DIA from C-l and replace with electric furnace and start heating.
(2) Continue heating until a temperature of 350°C is reached, or until all of the gases are transferred to MS-
1. This is indicated by a return to baseline by recorder (a shift in baseline is usually noted at this point,
due to the higher temperature of the gases entering the thermistor block, and also by a decrease in flow
rate).
(3)Shut Vent 2 and turn off helium flow. Open high vacuum valve to C-l, and continue heating until a
temperature of 350°C is reached, and a vacuum of < 10-4 mm of mercury is obtained. (C-l is then ready for
another run.)
f. Separation of Krypton, Xenon, and Methane from MS-1.
(1) With LN on MS-1 and MS-2, establish helium flow to (200-300 ml/min) MS-1, thermistor 2, vent 2.
(2)Remove LN from MS-1 and replace with a -23°C alcohol bath. After approximately two minutes, a sharp
increase is noted on the recorder. This is the argon and oxygen. Continue helium flow until the pen
returns to near the baseline (4 to 5 minutes). (Leave a small amount of oxygen.)
(3) Quickly rearrange helium flow, MS-1 to MS-2 to vent-3. (The oxygen transferred to MS-2 is used as an
indication in the elution of krypton from MS-2.) Continue flow until the krypton is eluted from MS-2,
approximately 12 to 14 minutes.
(4)Quickly rearrange flow MS-1 to vent-2 (MS-2 and vent-3 closed). Replace the alcohol on MS-1 with cold
water (20°C), and elute the nitrogen to vent. Watch the elution of nitrogen carefully and, by rearranging
the flow MS-1 to MS-2 to vent-2, transfer the last of the nitrogen peak to MS-2 (this is mostly methane).
(5)Place immersion neater in the cold bath, and heat until the carbon monoxide and xenon are all
transferred to MS-2, (10 to 12 minutes). Remove boiling water from MS-1.
g. Separation and Collection of Krypton, Methane, and Xenon from MS-2.
(l)Prepare C-2 by heating with a heat gun. Place a clean liquid scintillation vial and valve (P*) in position.
Evacuate to 0.0 mm on the manometer. Place LN on C-2.
(a) Arrange helium flow MS-1, MS-2, thermistor 3, vent 3. Remove LN from MS-2 and replace with -23°C
alcohol bath. The small oxygen peak which will be noted in two minutes is vented, (this indicates the
normal operation of the system). When the krypton peak appears, immediately close vent 3 and open
valve M. Collect the krypton in C-2 until the pen on the recorder returns to baseline. Close valve M,
open vent 3 and allow helium to continue to flow.
(b) Remove the helium in C-2 by pumping until a pressure of < 0.1 mm of mercury is attained. Close
vacuum valve (N), and heat C-2 to transfer the krypton to the vial. When pressure has stabilized,
record pressure and temperature. Close valve P. See following procedure (2).
(c) Repeat preceding procedure (1) for the methane separation.
(d)Flow should still be MS-1, MS-2, thermistor 3, vent 3. Replace alcohol bath with cold water (20°C). A
small peak of nitrogen should be noted on the recorder. When the methane peak appears, immediately
close vent 3 and open valve M. Collect the methane in C-2 until the recorder returns to the baseline.
Close valve M, open vent 3 and allow helium to continue to flow.
(e) Transfer the methane to the vial as in Section b.
(f) Repeat preceding procedure (1) for the xenon separation.
*PisaLuer Teflon plug valve.
-227-
-------
(g) Flow should be MS-1, MS-2, thermistor 3, vent 3. Heat, with immersion heat, MS-2. Allow the carbon
monoxide peak to vent, and immediately close vent 3; open valve M. Collect the xenon in C-2 as in
procedure (a).
(h) Transfer the xenon to vial as in Section b.
(2) Fill each scintillation vial with degassed toluene-base, liquid scintillation cocktail as illustrated in
Photograph 3, and place in liquid scintillation spectrometer for determination of radioactivity.
2. Calculations.
V, 1.14 x sample wt.
1253
1/3 1 .0 x sampl e wt .
where Vj volume krypton in sample
V3 volume methane in sample
1.14 vol concentration of krypton
normally found in air
1293 gm air per cubic meter
1.0 vol concentration of methane
normally found in air
VKr vy= or Vu v x P " 273° volume recovered
Kr, Xe, CH,. ^ ^^
where v vial volume * volume of C-2 and
transfer line
p pressure in vial
100 I Kr recovered
x 100 % Xe recovered
V x 100 ! CH,, recovered
where V2 volume Xe carrier added to sample
pCi/m3 - _ _ cpm _
(Counting j volume gas In vial % recovery
(Efficiency)
REFERENCES
Shuping, R. E., C. R. Phillips and A. A. Moghissi, (1969), Low Level Counting of Environmental
Krypton-85 by Liquid Scintillation, Analytical Chemistry.
Stevenson, D. L. and F. B. Johns, (1971), Separation Techniques for the Determination ofS5Kr in the
Environment, International Atomic Energy Agency, Vienna, pp 157-162.
-228-
-------
PRESSURE GAUGE
o
SAMPLE
IN
VENT
He
1, VAC He
F( H
D<
OA
G
B
T-1
VAC
T-2
MANOMETER
MOL PRE-COOLER C-1
SIEVE
MS-1
\J w
MS-2
FLOW DIAGRAM FOR SEPARATION OF METHANE, KRYPTON, AND XENON
Figure 1.
-------
30
-
32e
>
1 X
mm
-
> mm
45mm
»— *•
(
7 ^
<_
/
J
V..X
5 mm rod
^
V
r
X
)
v
*^»-a
j
\
^^>
12
12 mm tubing
Must fit in a 2 3/4" cylinder
MOLECULAR SIEVE COLUMN
Figure 2.
-230-
-------
25 mm
50 mm
300 mm
45 mm
18/9 S/J
48 mm
CHARCOAL TRAP
12 mm tubing
Figure 3.
-231 -
-------
to
CO
to
n
12mm
COPPER FLEX
TUBING
18/9 JOINT
'O'RING GROOVE
FLEXIBLE BALL JOINT
T-
52
mm
T
45
mm
^
n
LUER TAPER
28 mm
LIQUID SCINTILLATION VIAL
Figure 4.
Figure 5.
-------
FROM MS-2 TO VIAL VENT
II II II II
1x9
TO C-2
SEALED
FROM C-2
VALVE M IN VIAL POSITION
II II II -H-fl
VALVE M IN COLLECT POSITION
Figure 6.
-------
Figure 7. Schematic Wiring Diagram.
6v
R1 is 2K trim pot
R2,3,4 are 10 turn pots
All others are 1/4 watt
Thermistors are
GOW-MAC#13-502
ImV
Recorder
256
-------
CO
Ol
TO T-1
REFERENCE
THERMISTOR
#1
VENT
VENT
TO MS-1
TO MS-2
I i
THERMISTORS THERMISTOR#2
f
f
~« f
i
t
"=*
FROM T-1
FROM MS-1
HELIUM
PIPING DIAGRAM OF THE THERMISTOR CELLS
TO C-2 OR VENT
THERMISTORttS
FROM MS-2
Figure 8.
-------
10
w
O3
PORTABLE APPARATUS
Photograph 1.
-------
10
CO
-a
LIQUID SCINTILLATION COUNTER
BOHBBHBBBKBBBBo
Photograph 2.
-------
CO
00
1O 2O 30 ,40 „ SO cc
SCINTILLATION!
VIAL
|50 ml SYRINGE
Photograph 3.
-------
INTERNAL GAS-PROPORTIONAL BETA-SPECTROMETRY FOR MEASUREMENT OF
RADIOACTIVE NOBLE GASES IN REACTOR EFFLUENTS*
C.J. Paperiello
Radiological Sciences Laboratory
Division of Laboratories and Research
New York State Department of Health
Albany, New York 12201
Abstract
At the Radiological Sciences Laboratory of the New York State Department of Health, gas fractions
separated by gas chromatography are analyzed by internal gas-proportional spectrometry systems. These
systems include gas-proportional detectors, plastic anticoincidence detectors, multichannel analyzers, and
associated electronics. Detector systems are enclosed in 6-inch thick steel shields.
Internal proportional counting with multichannel analysis offers several advantages, particularly
improved sensitivity and specificity. Gas counting efficiencies are greater than 60% for 37Ar and 90% forS5Kr.
Detector background with 100-ml copper proportional tubes and plastic anticoincidence guards is on the order
of 0.3 cpm for 37Ar and 1.5 cpm for 85Kr. Shielded and guarded steel tubes have backgrounds approximately
four times higher, but are acceptable for high-level reactor samples.
By examining the spectra with a multichannel analyzer, the figure of merit for low-energy beta-emitters is
greatly improved over integral bias counting. The purity of the sample following chromatographic separation
can also be checked. Within certain abundance ratios, the levels of3Hand14C in hydrocarbon fractions can be
determined without combustion. Similarly, direct 37Ar and 39Ar measurements are possible.
The application of spectrometric techniques for analysis of several types of reactor gas effluents is
discussed.
INTRODUCTION
At the Radiological Sciences Laboratory of the New York State Department of Health, reactor gas effluents
are analyzed by Ge(Li) gamma-ray spectroscopy and internal gas-proportional spectrometric systems.
Although gamma-ray spectroscopy is the simpler procedure, low-background beta-proportional counting
offers greater sensitivity for all gaseous radioactive fission products, and for those which decay with little or
no gamma emission it is an absolute necessity. Although one might expect the activity of reactor samples to be
so high a low-background system is not needed, this is not the case. The range of activity ratios for a given
sample may be 107 between nuclides, and the range of activity may be greater than 105 between samples,
resulting in an overall range of about 1012 in activity in these studies.
The use of multichannel analyzers for counting proportional tube output offers several advantages as
compared to integral bias counting. These include improvement in the figure of merit for some beta-emitters, a
check on the purity of the sample after chromatographic separation, and simultaneous analysis of certain
isotopic mixtures found in reactor gas effluents after chromatographic separation. These mixtures include
37Ar/39Ar in the radioargon fraction, 14C/3H in methane, and other hydrocarbon fractions.
The gases routinely measured in reactor effluents by internal gas-proportional counting include the noble
gases 85Kr, 13iniXe, 133Xe, 37Ar, and 39Ar and the permanent gases 3H2,14CO2, and CH4 (14C and 3H). The latter
group is important in noble gas measurements because during chromatographic separation, contamination of
the argon fraction with 3H2, the methane fraction with S5Kr, and the CC>2 fraction with the 133Xe can occur.
SPECTROMETER SYSTEM
The spectrometer system is similar to those described by Curran (1958). The proportional tubes used in the
reactor sample measurements are commercial 100-ml stainless-steel proportional tubes manufactured by
LND, Inc. The active region of these tubes is a cylinder 2.3 cm in diameter and 24 cm long. For background
reduction, plastic scintillator anticoincidence guard detectors are used, and the entire detection system is
enclosed in a 14.4-cm thick steel shield. Figure 1 shows two of the 100-ml proportional tubes along with a 1-liter
tube and two of the guard detectors. The smaller guard will accept the 100-ml tubes, while the larger guard will
accept tubes as large as the 2.6-liter tube manufactured by LND, Inc. A typical 100-ml tube with a P-10 (90%
argon, 10% methane) fill has a plateau 250 volts long, beginning at 1,750 volts with a slope of less than l%per
100 volts.
A block diagram of the system is shown in Figure 2. Pulses from the proportional tube are amplified and
shaped before passing through a linear gate to the multichannel analyzer. If an event occurs in the guard
detector, the linear gate is for 10. ^tsec. The system dead-time, due to guard events closing the linear gate, is less
than 0.1%. All of the units shown in Figure 2, with the exception of the plastic guard detector, are commercial
products. With new multichannel analyzers, one can use the built-in linear gate in the anticoincidence mode,
thereby avoiding the cost of an external gate.
*Supported in part by USAEC contract AT(11-1)2222 and USEPA contracts 68-01-0522 and 68-01-LA-0505.
-239-
-------
OPERATION AND CHARACTERISTICS OF SPECTROMETER SYSTEM
Spectra of 3H, 14C, and 85Kr obtained with this spectrometer system are shown in Figure 3. Using the x-ray
fluorescence produced in the steel walls of the proportional tube with an external 125I source, the gain of the
system was set at 0.83 keV/channel. At this fixed gain, there are clear differences in end points and shapes for
these three isotopes.
Background spectra at a system gain of 0.42 keV/channel are shown in Figure 4. The integrated
backgrounds were: for the bare stainless-steel detector, 104 cpm; for the detector shielded by 14.4 cm of steel, 53
cpm; and for the detector shielded by the steel and an anticoincidence guard, 4.9 cpm. With a 100-ml copper
tube, not presently used for these reactor samples, an integrated background as low as 1.25 cpm has been
achieved with a background of 0.3 cpm under the 37Ar peak.
Argon spectra are analyzed for both 37Ar (from the Auger peak) and 39Ar (from the continuous spectrum). In
order to avoid possible pulse pileup from 37Ar, only the spectral region about 6 keV is used for 39Ar analysis.
The system has been calibrated for 37Ar efficiency using a standard obtained from the National Bureau of
Standards, and a value of 0.63 cpm/dpm has been obtained for the Auger peak efficiency.
A major advantage of spectral analysis is shown by comparing the spectra of an NBS 37Ar standard, a
contamina ced argon fraction after two passes through a gas chromatograph, and the same sample after an
additional pass (Figure 5). The contaminated sample contains a very small amount of 3H, which would not
affect the 37Ar value, but could have a serious effect on the 39Ar value. Integral bias counting would not show
this contamination. Gross contamination can occur even with double chromatographic separation because of
the large range of isotopic abundances in reactor gas effluents. The chromatographic separation sequence
leads to decontamination factors of 103 to 104 (Kunz, 1973). One may occasionally observe 3H interferences in
argon, 85Kr in methane, and 133Xe and 13imXe in CO and CO2 fractions. If great care is not taken in cleaning
the separation system after a particularly high activity sample, 3H can show up anywhere.
For sample spectra such as 85Kr, 3H, 14CO2, and' 4CO, the spectral shape and end point are first examined for
radiochemical purity. The spectrum is then summed; the background is subtracted; and the net counting rate
is corrected for counting efficiency, sample size, chemical recovery, and radioactive decay. In the cases of
i.umxe and 133Xe, this procedure is repeated for several counts, and the data are fitted by a least squares
method to a two-component decay curve.
At the present time, 37Ar and 85Kr are the only gas standards available from NBS. For 133Xe the method of
beta-gamma coincidence counting (Allen, 1965) was used to determine efficiency. While this procedure gave a ,
value of 0.86 cpm/dpm, which seems consistent with other measured efficiency values, the presence of a
conversion electron branch in 133Xe has been ignored. The result should be a somewhat greater value for this
factor. The same factor is used for 13imXe. Since 39Ar has a beta spectrum similar to 85Kr, the 85Kr efficiency
factor for that spectral region is used for 39Ar. In many reactor gas samples 85Kr and 133Xe are present in a
sufficient concentration to permit analysis by Ge(Li) gamma counting of the sample in the sampling vessel.
The 133Xe proportional-counter efficiency from coincidence calibration provides good agreement with the
Ge(Li) diode measurements.
One of the more interesting gases present in reactor effluents is methane. It may be composed of 3H or 14C, or
both, and beta-spectrometry permits the simultaneous analysis of both nuclides. In Figure 6 the spectrum of
the methane fraction from a heavy-water-moderated reactor is presented. The spectrum is run at a system gain
of 0.42 keV/channel. The regions from channels 1 to 39, and from 40 to 255, are summed, and the background
subtracted. The net counts in region 1 and 2, NI and N2 are given by:
(1)
(2)
where f, (T) = fraction of 3H spectrum in region 1,
f, (C) = fraction of 14C spectrum in region 1,
fo(T)= fraction of 3H spectrum in region 2,
fn(C) = fraction of' 4C spectrum in region 2,
N,p = net 3H count in spectral region,
Np = net 14C count in spectral region.
-240-
-------
For our tubes and gain range these equations are:
NX = 0.985 NT +0.575 NC (3)
(4)
Since N^ and N2 have been measured, these equations can be solved for Nrp and NQ and the activities
determined without combustion of the sample. By counting a series of twelve 14C methane samples.it has been
determined that the gain can be set with 125I with sufficient reproducibility that the errors in the constant
terms in Equations (3) and (4) are restricted to the third significant figure. The major drawback of this method
is that the ^H sensitivity is limited by the amount of' 4C present. If the 14C activity is one order of magnitude or
more above the background, the detectable limit for :iH is about 7% of the 14C activity. In Figure 7 the
decomposition of the methane fraction into *H and 14C components is shown. A summary of system
performance appears in Table 1.
The detectable limit is reduced by poor chemical recovery and, for short-lived nuclides, long delays between
collection and counting. It is enhanced by processing larger samples. Samples as large as 2 liters have been
processed in our laboratory. The major uncertainty in our work at the present time is the accuracy of the
efficiency factors. Those given in Table 1 for "9Ar, :!H, and' 4C are estimates for 85Kr and 37Ar, and that for' ;i:!Xe
has been determined by a method which is somewhat lacking in technical justification.
Standards for these and other gases will presumably become available in the future. Proportional tube
efficiency, however, unlike that of most other radiation detectors, varies slowly over a wide range of energies.
Extrapolation of detector efficiency for beta-emitting isotopes is not especially difficult, but isotopes which
decay by electron capture or by decay of metastable states present problems which require direct comparison
to standards.
REFERENCES
Allen, H. A., (1965), Measurement of Source Strength, In Alpha-, Beta-and Gamma-Ray Spectroscopy Vol.
1, K. Siegbahn, Ed., p. 425, (Amsterdam, North-Holland).
Curran, S.C., (1958), The Proportional Counter as Detector and Spectrometer, in Handbuch der Physik,
Vol. 45, 174 (Berlin:Springer).
Kunz, C.O., (1973), Separation Techniques for Reactor-Produced Noble Gases, Proceedings of the Noble
Gas Symposium, Las Vegas, Nevada, Sept. 24-28, 1973.
TABLE 1. Spectrometer Performance.
Gain
keV/channel
0.83
0.21
0.21
0.83
0.42
0.42
0.42
Channels
used
1-255
10-19
32-255
1-255
1-39
1-255
40-255
Detectable limit
Gas detector Ave 1-ml sample
efficiency bkgd 1,000-min count
(cpm/dpm) (cpm) (fxCi/ml)
0.90
0.63
0.45
0.86
0.75
0.85
0.36
5.0
0.8
1.5
5.0
5.0
5.0
0.13
l.lxlO-7
6.0 xlO-8
1.2xlO-7
l.lxlO-7
l.SxlO-7
l.lxlO-7
4.0 xlO-8
-241-
-------
to
*-
ro
1. Proportional tubes, lOO-ml and 1-liter, with plastic scintillator g-uard detectors.
-------
H.V. SUPPLY
0-1800 V.
PRE-
AMP
H.V. SUPPLY
0-6000 V
PAD(PRE-AMP
AMP DISCRIM)
•PM TUBE
PLASTIC
-SCINTILLATION
GUARD
PROPORTIONAL
TUBE
LINEAR GATE
PULSE-
SHAPING AMP
MULTICHANNEL
ANALYZER
Figure 2. Block diagram of internal gas-proportional beta-spectrometer system.
-243-
-------
10'
,5F
CHANNEL NUMBER
Figure 3. Spectra of "H,' 4C, and fi5Kr. (Gain 0.83 keV/channel.)
-244-
-------
to
I08
I0
id5
V)
o
10'
- •/•. BARE DETECTOR (xlOO)
SHI ELDED
" (XIO)
-*44 SHIELDED + GUARD
4 * * *
« *** « Ml
4 4*44
« • ••*• »••••* •••••• •• •••
44t44
44
40
80 120 160
CHANNEL NUMBER
200
240
Figure 4. Background spectra for a 100-m] steel proportional tube. The spectra of the shielded and hare
detectors have been shifted upward by factors of 10 and 100, respectively. (Gain 0.42 keV/chnnnel.)
-------
I05L
ICT
10-
o
o
10'
10
37Ar STANDARD
'•'•V.
B
3H-CONTAMINATED SAMPLE
PURIFIED SAMPLE B
0 20 0 20 40 0
CHANNEL NUMBER
20
40
60
Figure 5. Argon spectra: A. NBS 37Ar standard; B. Argon fraction with 3H contamination; and C. Fraction
shown in B with an extra purification step. (Gain 0.21 keV/channel.)
-246-
-------
io-
10'
D
O
O
10
GROSS COUNTS
*
A
1
t *
"BACKGROUND
COUNTS
A M * i
M t
50
100
CHANNEL NUMBER
150
200
Figure 6. Gross spectrum of a methane fraction showing the presence of 3H and 14C compared with
background. (Gain 0.42keV/ehannel.)
-247-
-------
I-
Z
8
50 100
CHANNEL NUMBER
200
Figure 7. Spectrum of the methane fraction shown in Figure 6 with background subtracted. The 3H and 14C
components are indicated. (Gain 0.42 keV/channel.)
-248-
-------
ENVIRONMENTAL MONITORING FOR KRYPTON-85*
D.E. Barber
School of Public Health
University of Minnesota
Minneapolis, Minnesota 55455
Abstract
85Kr presents unique environmental monitoring problems because it does not react with other elements and
compounds at normal ambient temperatures and pressures. However, elaborate means are available to
manage the 85Kr problem, but a simpler, inexpensive approach is required if monitoring is to be accomplished
at many sampling locations. This work shows that environmental monitoring for 86Kr is possible by
collecting air samples in thin plastic bags, and counting the bags for beta particle activity. The direct
counting of contained samples of this typemakes itpossible todetect concentrations less than the public MFC
for 8SKr. The bagged-sample technique is readily adaptable to any environmental monitoring station with
power to run a low-volume air pump. The idea of counting the bagged-sample directly is a new, low-cost,
approach to environmental gas monitoring which may have application in environmental, clinical, and
industrial situations.
INTRODUCTION
85Kr is a fission product which escapes or is released to the environment primarily as a byproduct of
reprocessing nuclear fuel. In fuel reprocessing, 85Kr, 1311,129I,133mXe, 133Xe, and 3H are released, but only 85Kr
and 3H are released in sufficient quantities, and have long enough half-lives, to produce significant
concentrations in extensive environmental air volumes (Kirk, 1972). The present atmospheric inventory of
85Kr is estimated to be 60 MCi — more than twice the inventory in 1962 (Kirk, 1972). If projections with respect
to population, demand for electric power, use of nuclear power plants, and release of 86Kr to the environment
are correct, concentrations of 85Kr in the atmosphere may reach 3 x 10-7 jzCi/ml (the public MFC) about the
year 2050 (Holland, 1969; Cowser and Morgan, 1967; and Coleman and Liberace, 1966). Treatment of effluents
to remove up to 98 percent of the noble gases from fuel reprocessing is possible, and methods have been tested
on a pilot plant scale (Slansky, 1971). The cost of treating the effluent has been estimated at about 1 percent of
the total reprocessing cost (Slansky, et al., 1969), and treatment will probably be used extensively in the near
future. Consequently, present projections with respect to anticipated atmospheric concentrations in the 21st
century are probably much too high. However, effluent treatment or not, 85Kr releases to the atmosphere will
require environmental monitoring for the gas because:
(1) A removal efficiency of 98 percent reduces released concentrations to only 0.02 times their original
values.
(2) The atmosphere cannot be used exclusively for 85Kr dilution.
(3) It will be necessary to assure compliance with regulatory standards.
Although this paper addresses itself to the problems of monitoring 85Kr, the results are more generally
useful. There are situations which arise when one would like to know the response to be expected from
ordinary radiation detectors when they are presented to clouds of beta particles or low-energy gamma ray
emitters with dimensions less than those of infinite volume.
There are a number of possibilities for monitoring 85Kr in the atmosphere given unlimited financial
resources. Any of the methods for removal of noble gases on a large scale from fuel reprocessing effluents
might be used on a smaller scale; but, these methods involve elaborate pretreatment of the intake gases
and/or cryogenic temperatures (Slansky, 1971). When adsorbing media such as molecular sieves or charcoal
beds are used, the operating temperature must be matched to the sampling rate and sample size (Kirk, 1972).
In liquid scintillation counting, the lowest concentration of 85Kr that can be analyzed without
preconcentration is about 3 pCi/ml, and the poor solubility of air in liquid scintillation cocktails presents a
problem for sampling air mixtures of gases (Kirk, 1972). So, sophisticated sampling for 85Kr in the
environment presents severe practical limitations for both technical and financial reasons, especially in
those situations where numerous monitoring stations may be involved. Simple, inexpensive methods must be
found.
Direct counting of gas samples in vinylidene chloride (Saran) is the approach taken here. The sampling
approach is similar to one already reported for 222Rn (Sill, 1969). But, the idea of counting beta particle
emitters directly, without removing the sample from the bag for the purpose of gas monitoring, is believed to
be new. It has been noted recently that 133Xe contamination in the air should probably be done by counting the
air sample directly rather than attempting to collect the Xe in aqueous solution (LeBlanc, 1972). LeBlanc
points out that there is a misunderstanding concerning the best methods to monitor for 133Xe because of
erroneous solubilities for 133Xe given by the Hbk. of Chm. and Phys. (1966). Similar uncertainties may also
apply to krypton.
*The data on S!>Kr reported here resulted from work completed in the Faculty Research Participation Program
of Associated Western Universities at the Health Services Laboratory, USAEC, Idaho Falls, Idaho. The
author also gratefully acknowledges the cooperation of Aerojet Nuclear Company.
-249-
-------
PROPERTIES OF MKr AND INTERFERENCE PROBLEMS
Some properties of 85Kr are given in Table 1.85Kr is nearly a pure beta emitter from the dosimetry viewpoint,
but the gamma ray is useful for calibration purposes for activities greater than 10-2/zCi. Monitoring for 85Kr by
gamma ray spectrometry can be accomplished provided the concentration is high enough, and is sustained
for a period of time at least equal to the measurement period. At concentrations approaching the maximum
permissible concentrations, however, the branching ratio (0.004) of gamma ray disintegration to beta particle
disintegration makes beta particle measurements far more attractive.
Iodine impregnated, activated charcoal can be used to eliminate interference from iodine-131. Silica gel can
be used to keep the sample dry. Storage can be used to eliminate interference from radon-daughter product
activity, and filters can be used to reduce interference from particulate activity. There remain, then, 133Xe and
isamxe. The technique might be biased against these nuclides by storage, careful selection of detector window
thickness, and detector design. Interference from 133Xe is not expected to be significant at present because the
ratio of 133Xe to 85Kr activity for aged nuclear fuel is typically less than 10-4 percent (Smith, et al, 1970).
However, as use of nuclear fuels increases, the used fuel will be less and less aged before reprocessing, in which
case it will be necessary to cope with the 133Xe problem. This should be relatively easy to do because of the large
differences in modes of decay, and beta energies between those for 133Xe and 85Kr.
If 133Xe is produced as a fission product, 133mXe will also be present (Martin and Blichert-Toft, 1970). The
issmxe could present severe beta particle detection interference problems if fuel is stored for less than 10 days
prior to reprocessing. The conversion electron energies from 133mXe are only slightly lower than the average
energy of beta particles from 85Kr. However, advantage might be taken of the 0.233 MeV gamma ray from
133mXe to distinguish it from both 85Kr and 133Xe.
METHODS
l.MKr Chamber.
A chamber approximating an infinite volume of air for 85Kr beta particles was constructed of 4' x 8' x 3/4"
plywood on a 2" x 4" wood superstructure. The resultant chamber was an 8' cube equipped with feed-throughs
provided for power, sampling inlets, outlets, and a G.M. detector lead. The chamber was kept outside at
ambient temperature and pressure.
Chamber concentrations were measured and monitored by a 30 mg/cm2 metal-walled Amperex G.M. tube
(1/2" dia. x 7" long) operated at 960 volts. Pulses from this detector were counted with a Baird-Atomic Mod. 530
sealer.
Serial dilutions from 85Kr stock to glass vials, to the chamber, and ultimately to polyethylene and Saran
bags, were made to produce various concentrations of 85Kr ranging from 2 x 10-8^Ci/ml to 7 x 10-4 fj.Ci/ml.
2. Calibrations.
Unfortunately, the detection limit by gamma ray spectrometry for 85Kr is on the order of 10-2^ Ci. As a result,
chamber concentrations could not be confirmed by this method over the range of concentrations of interest. To
enable measurements of concentration in the chamber over the entire range of interest, the G.M. tube in the
chamber was calibrated as follows. A sample from 85Kr stock was counted on a 65 ml Geli detector in the
standard geometry. The GeLi detector was calibrated with a 85Kr source from the National Bureau of
Standards. The sample was then transferred to a 4.2-liter Saran bag to calibrate an 8" x 4" Nal detector for this
geometry. A 4.2-liter sample taken from the chamber was then counted on the same crystal yielding the true
concentration in the chamber, 1.05 ±0.15 x 10-4/*Ci/ml. This concentration provided the primary calibration
factor, 3.82 x 10-9fiCi/ml per cpm, for the chamber G.M. tube. The G.M. tube count rate was then used as the
reference for the concentration in all samples taken from the chamber. The G.M. tube calibration factor agrees
well with the 2.1 x 10-9/iCi/ml per cpm previously reported for a similar, but longer, G.M. tube (Smith, et al.,
1970).
The rotameter was calibrated with a wet-test meter, and was found to be in calibration within ±20 percent
The uncertainties in visually setting the rotameter are large. Consequently, the rotameter reading Was taken
to be the true flow rate at ambient temperature and pressure.
3. Sampling and Instrumentation.
A diagram of the dilution and sampling system is given in Figure 1. The sampling line was prepared as it
might be used m the field to pretreat samples taken from the chamber. Components of the sampling line were
connected with plastic tubing ranging from 1/4 to 3/8 inch inside diameter. The input to the rotameter from
the chamber was at ambient pressure, and consisted of 3/8 inch plastic tubing suspended at the center of the
«Kr chamber close to the G.M. tube. Samples were collected at various flow rates and various sampling times
as indicated in the results. ^ s
' Except for a few measurements with the 5-liter Saran bag filled to capacity, bags were filled to less than
capacity to minimize leakage, if any should occur, and to provide a flexible bag geometry. In this way it was
possible to achieve 2 J counting geometry when samples were placed on the detectors, and to reduce the error
which might be introduced as a result of pressure build-up in the bag. *cuui,c meenui
-250-
-------
The instruments used to analyze the bagged-samples were as follows:
(1) Ludlum Instrument Co., Model 14A Gieger Counter with LND Inc. G.M. tube, 30 mg/cm2 metal wall.
(2) Eberline Instrument Co., Model HP-210 G.M. probe with LND 731-1 G.M. tube, 2" diameter thin
window, connected to Baird-Atomic, Model 530 sealer, operated at 800 V.
(3) A sheet of plastic scintillator, 4.1" diameter by 0.19" thick, attached with Dow Corning QC-2-0057
silicone compound to a DuMont 6364 photomultiplier tube (5" diameter face). The scintillator was covered
with two layers of doubly aluminized Mylar to make a total window thickness of 2.06 mg/cm2. The detector
signal was fed through a preamplifier into a Baird-Atomic, Model 530 sealer with input sensitivity set at
approximately 50 mV. The high-voltage to the detector was 1,000 V.
The first and second instruments were used outside, immediately adjacent to the chamber, to examine their
response to samples taken from the chamber. The third detector was used inside a vault with 10" thick steel
walls.
Measurements with the first two instruments were made in the presence of radon-daughter product activity.
Measurements with the third instrument were made on samples stored overnight.
4. Sample Containers.
Two types of plastic bags were used to collect, store, and count samples. One bag was Saran type 18-100
(manufactured by Analytical Specialties, Inc., and distributed by the Anspec Co., Ann Arbor, Michigan). This
Saran has a density thickness of 8.0 mg/cm2. Saran is known to contain 222Rn with losses of less than 0.12
percentper day up to at least 14 days (Percival, 1971). It, therefore, seemed a suitable choice for 85Kr. Saran of
this thickness is also very durable and easy to handle.
The second type of plastic bag used in this work was a simple polyethylene bag. The bag was 20" square
with a wall thickness of 4.7 mg/cm2. The open end of the bag was heat-sealed, and the center of one side was
cut out to a diameter of about 7" to accommodate a thin Saran window. The Saran used for this window was
the ordinary household type. Its thickness was 2.2 mg/cm2. The Saran window was attached to the bag with a
translucent silicone rubber adhesive sealant (RTV-108, General Electric Co., Waterford, New York). This
made a satisfactory seal for the purpose of this experiment, but it does not provide a permanent seal. Further,
polyethylene is generally known to be permeable to many compounds, including water vapor, and probably is
unsuitable for 85Kr containment for more than a few days. The purpose in using these homemade bags was to
provide a large-volume container with a very thin window to provide maximum beta particle transmission
with essentially a 2 n geometry when placed on the plastic scintillation detector.
RESULTS
l.MKr Chamber.
With the chamber containing 3 x 10-4fiCi/ml of 85Kr, the count rate of the chamber G.M. tube dropped from
7.43 x 104 cpm to 7.30 x 104 cpm over a 150-minute period. This amounts to a leakage rate of 0.8 percent per hour
at the highest concentrations used in the chamber.
Rate meter measurements showed that the gas dispersed in the chamber within two seconds, and remained
dispersed even without the benefit of the fans in the chamber. No significant reduction in chamber
concentration occurred which could not be explained on the basis of the chamber leakage rate. The chamber
concentrations were remarkably stable and reproducible. With the access door fully open, and with the fans
running as usual in the chamber, it required 2 to 3 minutes to reduce the chamber concentration to 1/2 of its
original value.
2. Chamber G.M. Tube.
The count rate of the 85Kr chamber G.M. tube as a function of 85Kr concentration is given in Figure 2 for both
input to the sealer and to the Ludlum rate meter. Each observation involved a 3-minute count. The first
observation was made at the lowest concentration. Two subsequent additions of 85Kr provided the three
concentrations in the figure. The tube was used bare, and was supported with its coaxial cable at the center of
the chamber. The practical lower limit of detection for the Ludlum rate meter with the chamber G.M. tube is
about 6 x 10-7/iCi/ml; this provides a net meter reading of 0.04 mR/h in an infinite cloud. When the tube is
connected to a sealer, the lower limit of sensitivity is a function of the background and counting time. For the
conditions of Figure 2, the limit for detecting 85Kr with the sealer is much lower than with the rate meter.
3. Ludlum G.M. Survey Meter.
With a chamber concentration of 7.0 x 10-4fiCi/ml samples were taken in the 5-liter Saran bag for time
intervals ranging from 5 to 25 seconds at 12.51pm. When the Ludlum probe, with beta shield open, was laid on
each sample the response was found to be linear with respect to activity — irrespective of bag geometry (see
Figure3).
-251 -
-------
' The lowest activity was distributed in a volume of only 1 liter, but the highest activity was distributed in 5
liter's. There are large uncertainties in reading this meter, but the linear relation over a factor of five change in
volume is clear.
' The minimum reliable net reading for the Ludlum survey meter is 0.04 mR/hr. This corresponds to about
0.05 ptCi which, according to the figure, might be distributed in as much as 5 liters. Hence, the minimum
detectable concentration for this meter under these conditions of measurement becomes 10-5^Ci/ml. This is
much too high for environmental monitoring purposes.
4. Eberline G.M. Probe.
This probe was used to measure 85Kr activity in the 5-liter Saran bag containing various volumes of gas at a
constant concentration of 6.3 x 10-4 ptCi/ml. The results are shown in Figure 4. The statistical counting errors
are large; but, again the linear relation between activity and different geometries is obvious. One observation
was made at low-activity (therefore, low-volume) by rolling the bag to about 1/4 of its maximum volume to
show the importance of widely different geometries.
In this case the minimum detectable activity appears to be about 0.5 ^Ci. When expanded to 5 liters, this
yields a minimum detectable concentration of 10-4f/Ci/ml. But, measurements at maximum volume and
•various concentrations show the minimum detectable concentration to be much lower than this (See Figure 5).
When 5 liters of 85Kr are taken from the chamber in the 5-liter Saran bag, the response of the probe is as
indicated in Figure 5. The probe was lightly pressed against the side of the sample bag for each measurement.
The highest concentration was measured first. Subsequent lower concentrations were produced by opening
the chamber between samples. All counts were for three minutes.
The figure shows that this detector is capable of detecting as low as 3 x 10-7^' Ci/ml of 85Kr under the
condition of measurement. But, it is not likely to detect 3 x 10-8^t Ci/ml even for long counting times. A more
sensitive detector is require.
5. Plastic Scintillation Detector.
Samples were taken from the chamber, diluted to the desired concentration with ambient air through the
sampling chain into the polyethylene bags, and stored overnight to permit the decay of radon-daughter
product activity. The 85Kr concentration for all these samples was 2 x 10-V Ci/ml. Ambient air samples taken
through the sampling chain also were found to contain activity. This activity had an effective half-life of 33
minutes, which is typical for radon-daughter product activity. The filters do not eliminate interference from
radon-daughter products at these low 85Kr concentrations.
When stored samples of 85Kr are counted, Figure 6 shows that it is possible to detect concentrations less than
3 x 10-8f' Ci/ml using the polyethylene bags containing 23 liters of sample. Two 13-liter samples in commercial
Saran bags showed that it may be possible to measure these low-concentrations in a lesser volume and in a
more durable bag than provided by the polyethylene,
DISCUSSION
85Kr has been measured at a variety of concentrations, in several different sample volumes, and with
several different sample containers. In this work, the best combination was a 23-liter sample, stored overnight
in a polyethylene bag with a 2.2 mg/cm2 Saran window, and counted on a plastic scintillation counter. This
combination provides minimum detectable concentrations of less than 3 x 10-8 fiCi/ml for 85Kr. The method is
suitable for environmental monitoring provided pretreatment of the sample removes other beta emitters, such
as 131I, which would interfer with the analysis, and provided the sample container used in the field is
reasonably durable and impermeable to 85Kr.
' The commercially available, 8.0 mg/cm2 density thickness Saran bag, in the size advertised as 12 liters, is a
good possibility for field sampling. One of the difficulties will be finding a metering pump with a low enough
flow rate, and'sufficient flow rate stability, to accurately pump a 12-liter volume into the bag over a long period
of time. A much larger sample may be necessary for the sake of obtaining an accurately known volume of
sample.
Common G.M. rate meters can detect 85Kr in bagged, 5-liter samples containing concentrations as low as 10-4
to 10-5 jtCi/ml. Common G.M. probes connected to sealers can detect 85Kr in bagged 5-liter samples
containing concentrations as low as 3 x 10-7{i Ci/ml — without benefit of shielding and in'the presence of
radon-daughter product activity. Under the conditions of an infinite cloud, bare G.M. detectors connected to
sealers are capable of detecting concentrations below 3 x 10-8 fiCi/ml, depending upon counting time
Detecting these low-concentrations with 12- or 23-liter bagged-samples requires a well-shielded large area
plastic scintillation counter or its equivalent. If the background count of the plastic scintillator doubled and if
the true concentrations in the samples were twice those reported here, it should still be possible to measure
85Kr concentrations as small as 3 x 10-8/iCi/ml using the technique described here with slight modifications
' The variations in data points for bagged-samples are due primarily to inaccuracies in reproducing
rotameter settings. Construction variations in polyethylene bags also contribute to the variation in
observations where these bags were used. Observations in Figures 2 through 5 were made immediately after
the samples were taken. Variations in radon-daughter product activity over short periods of time contribute to
the variation in observation for these samples.
-252-
-------
With the detector on the sample bag, the geometry is essentially constant at 2 n. So, the detector response is
expected to be linear with increasing volume (therefore activity) for a fixed concentration in the sample as
shown in Figures 3 and 4. This should be true for volumes and geometries for which self-absorption in the
sample is negligible. So, if one is on the edge of a cloud, detector response will be proportional to the total
activity in the cloud, irrespective of cloud dimensions within broad geometric limits. The linear relation must
begin to level off and reach saturation as the dimensions of the cloud approach those of an infinite volume.
Figure 4 yields an efficiency of 0.003 cpm/dpm at all different volumes and geometries at a fixed
concentration of 6.3 x 10-«f*Ci/ml.
There is some variation in counting efficiency with concentration of activity at a fixed volume. Figure 5 has
a slope of 1.04, and shows an efficiency of 0.003 cpm/dpm at 10-6fiCi/ml and 0.004 cpm/dpm at 6.3 x 10-4
fiCi/ml (the concentration used in Figure 4). An increase in efficiency accompanying an increase in
concentration is to be expected because of the larger number of maximum energy beta particles contained in
the samples at higher concentrations. This increase in efficiency with increases in concentration is not
apparent at concentrations on the order of 10-7fZ Ci/ml as Figure 6 shows. The slope shown in Figure 6 is 1.00.
Figure 6 yields an efficiency of 0.014 cpm/dpm for 23-liter samples, and 0.024 cpm/dpm for 13-liter samples.
The smaller, thicker bag yields higher efficiency probably because it keeps more activity in the solid angle of
the detector. The 23-liter bags drooped somewhat below the 2 rr solid angle of the scintillation detector. The
optimum geometry is probably a hemisphere with its flat plane centered on the detector surface. The optimum
geometry and volume of the sample for this technique is still open to question and deserves additional study.
Careful attention to this question would probably reduce further the minimum detectable concentration for
85Kr.
Bagged-samples provide several advantages over "in situ "measurements of 85Kr.
(1) When multiple sampling stations are required, "in situ" measurements require multiple detectors and
recorders or a telemetering system. This approach is considerably more expensive.
(2) "In situ " measurements must also include pretreatment of the air to eliminate interference from other
beta particle emitters. Accumulation of activity in the sampling chain may interfer with detection sensitivity
to 85Kr beta particles.
(3) Both sample geometry and detector geometry are always known, and are reproducible with bagged-
samples.
(4) Provided sufficient activity is collected, the average concentration during the sampling period will be
measured with bagged-samples, irrespective of either the dimensions or the concentrations of the
contaminated air. This may yield a sensitivity greater than that provided by "in situ " measurements.
The disadvantages of bagged-samples are:
(1) They are incapable of identifying either the time or the magnitude of changes in air concentrations,
and are not suitable for an alarm system.
(2) In large volumes they are awkward to handle, and require that precautions against leakage be taken.
SUMMARY AND CONCLUSIONS
An inexpensive method to monitor atmospheric 85Kr at or below 3 x 10-8/*Ci/ml (0.1 MFC), with minor
modifications to existing environmental air sampling stations, has been described. But the method needs
further development and testing with mixtures of radioactive gases and aerosols likely to be found where 86Kr
is emitted. The method does not require volumes which are infinite with respect to the beta particle energy of
8sKr. Neither does it require concentrations of activity which are stable with respect to time. Therefore, it is a
realistic method from the viewpoint of conditions likely to be experienced in the field. One rarely encounters a
truely infinite cloud sustained over a period of time long enough to make infinite cloud measurements realistic.
The ultimate test of the technique should involve mixtures of 1311,133Xe, 3H, and 85Kr in concentrations likely
to be encountered in the environment of reactors and fuel reprocessing plants. If the technique should fail this
test of mixtures, it may still be useful as a screening technique for radioactive gases released to the
environment by man.
REFERENCES
Berger, M. J., (1971), J. of Nucl. Med., Sup. No. 5, Vol, 12.
Code of Federal Regulations (1965), Title 10, Part 20.
Coleman, J.R., and R.Liberace, (1966), Radiolog. Health Data Reports, 7, 615.
Cowser, K. E., and K. Z. Morgan, (1967), Health Physics Division Annual Report, ORNL-4168,39-45.
Handbook of Chemistry and Physics, (1966), Cleveland: The Chemical Rubber Company.
Holland, J. A., (1969), Proceedings of the AEC Symposium, Biological Implications of the Nuclear Age,
USAEC Division of Technical Information.
Kirk, W. P., (1972), Environmental Protection Agency, Office of Research and Monitoring, Washington,
B.C.: U.S. Government Printing Office, 484-482/46.
LeBlanc, A. D., (1972), Phys. Med. Biol., 17,585-589.
Lederer, C. M., J. M. Hollander, and I. Perlman, (1967), Table of Isotopes, 6th Ed. (New York: John
Wiley and Sons, Inc.)..,
Martin, M. J. and P. H. Blichert-Toft, (1970), Nuclear Data Tables, 8A, 108.
-253-
-------
Percival, D. R., (1971), Personal Communication, Analytical Chemistry Branch, Health Services
Laboratory, USAEC, Idaho Falls, Idaho.
Radiological Health Handbook, (1970), Washington: Public Health Service.
Sill, C. W., (1969), Health Physics, 16,371-377.
Slansky, C. M., (1971), Atomic Energy Review, 9,423-440.
Slansky, C. M., H. K. Peterson, and V. G. Johnson, (1969), Environmental Science and Technology, 3,
446-451.
Smith, D. G., J. A. Cochran, and B. Shleien, (1970), PHS, BRH/NERHL 70-4, Rockville, Maryland.
Voilleque, P. G., D. R. Adams, and J. B. Echo, (1970), Health Physics, 19,835.
TABLE 1. Properties of 85Kr.
Property
Reference
Density, mg/ml
Density/Density Air
Maximum Beta Energy, MeV
and Abundancy (%)
Average Beta Energy, MeV
Range of Maximum Energy Beta in:
Aluminum, mg/cm2
Air, cm
Range of Average Energy Beta in:
Aluminum, mg/cm2
Air, cm
Specific Gamma Ray Constant,
__R
Ci-Er at 1 meter
Deposition Velocity on Grass,
cm/sec
Max. Permissible Concentrations,
^Ci/ml;
Occupational
Public
Half-life, years
Gamma-Ray Energy, MeV and
Abundancy (%)
3.74
2.9
Hbk. of Chem. & Phys., (1966)
Hbk. of Chem. & Phys., (1966)
0.67 (99.6) Lederer, et al., (1967); Rad.
0.160 (0.4) Health Hbk., (1970); and
Martin and Blichert-Toft, (1970)
0.2464 Berger(1971)
235 Rad. Health Hbk. (1970)
182 Rad. Health Hbk. (1970)
59 Rad. Health Hbk. (1970)
46 Rad. Health Hbk. (1970)
2.34 xlO-4 The Author
2.3 x 10-11 Voilleque, et al., (1970)
1 x 10-5 10 CFR20 (1965)
3 x 10-7 10 CFR20 (1965)
10.76 Rad. Health Hbk. (1970)
0.517 (0.4) Martin and Blichert-Toft (1970)
-254-
-------
Pressurized
Kr-85
Reducing
Valve
Cn
Oi
Transfer Flask
Rotameter
Kr 85 Dilution Chamber
Pressurized
Air
Reducing
Valve
Silica Gel
Sample Bag
Atmosphere
Pump
Membrane Filter
Hepa Filter
Charcoal
Hepa Filter
Figure 1. Krypton dilution and sampling system.
-------
Sealer (net cpm)
o
o,
Ludlum Meter (mR/hr)
Figure 2. Response of the Amperex G.M. tube in the exposure chamber as a function of 85Kr concentration.
(The counting time for each observation with the sealer was 3 minutes.)
-256-
-------
o
o
51
O
Net Meter Reading (mR/hr)
-J ro
Figure 3. Ludlum G.M. survey meter response to a fixed concentration of 85Kr in a 5-liter Saran bag at
different volumes.
-257-
-------
- 30
Q.
O
"o
en
to
01
00
c
o
§
o
or
o
o
o>
20
10
0
Max/mum Volume
5 liters
Bog Rolled to
Maximum Volume
0
Total Activity
Figure 4. Response of the Eberline HP-210 G. M. probe to a fixed concentration of 85Kr in a 5-liter Saran at
different volumes. (The counting time for each observation was 1 minute.)
-------
8
E
O.
O
"CD
I03rr
10'
10
10
0
I I
I I I
^ 10'
10'
10
10*
o
10
10 10
Concentration (pCi/ml)
10'
10
Figure 5. Response of the Eberline HP-210 G.M. probe to 5 liters of 85Kr in the 5-liter Saran bag containing
various concentrations. (The counting time for each observation was 3 minutes.)
-------
e
Q.
O
0
10'
• Polyethylene bags with Saran windows
O Commercial Saran bags
So'for 100 minute count
at 170 cpm background
1 I I I I III
10
-9
0
-8
10
-7
10
-6
Concentration (uCi/ml)
Figure 6. Response of plastic scintillation detector to 23-liter samples of 85Kr in polyethylene bags with Saran
windows and 13-liter samples in commercial Saran bags. (All samples were stored overnight and counted to
yield a 3
-------
ENVIRONMENTAL RADIATION MONITORING
WITH THERMOLUMINESCENT DOSIMETERS
E. L. Geiger and E. A. Sanchez
Eberline Instrument Corporation
Santa Fe, New Mexico 87502
Abstract
Thermoluminescent dosimetry (TLD) provides one method of documenting environmental radiation dose
rates near nuclear facilities. Factors that affect the accuracy and precision of TLD measurements have been
evaluated using badges with five solid lithium fluoride dosimeters sealed in plastic. Data obtained from 1968
to date (July 1973) indicate that careful attention to annealing and reading can provide an accurate and
precise documentation of background radiation dose rates. In-transit exposure was found to be the largest
variable, and this error can be eliminated by local annealing and reading. With the TLD program described, a
change at the site perimeter as small as 10 mrem/year can be measured, but the problem is to identify the
reason for the change. Changes in background radiation dose rates due to snow cover, soil moisture, and
various other natural phenomena, and natural differences in dose rates at various locations near a nuclear
power station complicate the interpretation of TLD data. For best results, the monitoring program should be
designed to compare "indicator" and "reference" station measurements as well as "pre-operational" and
"operational" measurements.
INTRODUCTION
The "as low as practicable" (ALAP) concept for the control of radiation exposures has caused some changes
in methods for environmental radioactivity surveillance around nuclear power plants. Regulatory guides,
technical specifications, and design objectives require sensitivities for some measurements that are one or
more orders of magnitude lower than previously considered necessary. For example, the method now used by
Eberline Instrument Corporation for 131I in milk has a sensitivity of 0.2 pCi/1, whereas a year ago 10 pCi/1
was adequate.
The measurement of radiation dose from noble gases released to the atmosphere at the design objective of 10
mrem/year is even more difficult than the measurement of 131I in milk at the level specified in AEC Regulatory
Guide 1.42. In an attempt to measure this small dose attributable to the nuclear power plant in the presence of
a much larger and variable dose attributable to natural background, integrating dosimeters are placed at sites
around the nuclear facility. The dosimeters must accurately measure dose for a wide range of photon energies.
Thermoluminescent dosimeters (TLD) have been used by Eberline for environmental monitoring since 1968,
but the need to measure changes as low as 10 mrem/year has only recently been emphasized. Precision and
accuracy of measurement have become increasingly important with the emphasis on reducing exposures to
the ALAP design objective.
The U.S. Environmental Protection Agency review of natural radiation exposure in the United States
(EPA, 1972a) provides an indication of the magnitude of variations in background radiation. Variations due
to cosmic radiation at a specified site location and altitude are less than 10%, but variations due to terrestrial
radiation are frequently larger than 10%. Natural 40K and the decay chains of uranium-238 and thorium-232
account for most of man's exposure from natural terrestrial radiation.
Spatial variations at a site are due primarily to terrestrial radioactivity. The results in Figure 1 illustrate the
magnitude of variations due to terrestrial radioactivity that have been measured within ten miles of a
proposed site. These data are based on 12,581 usable spectra from in-situ gamma spectrometry within a ten-
mile radius of the Allen's Creek site in Austin, County, Texas. A helicopter was used to make the survey for
Houston Lighting and Power Company. Terrestrial radiation dose rates at this site varied from 1 to 8
/frem/hr. Total dose rates (terrestrial plus cosmic) at 25 TLD locations were estimated based on this one-day
survey of terrestrial dose rates, assuming 33 mrem/year from cosmic radiation at the altitude and latitude of
the site. These estimated total dose rates are compared in Table 1 with integrated dose (TLD) measurements
over a period of two calendar quarters. The estimated total dose rates based on in-situ gamma spectrometry
measurements are lower than the TLD measurements. This may be attributed to soil moisture at the time of
the survey (water standing in some areas), which causes a negative bias in the in-situ gamma spectrometry
data, and in-transit exposure of the TLD badges, which causes a positive bias in the TLD data.
Time variations of radiation exposure at a site are due primarily to soil moisture (and snow cover at some
sites), migration of radon gas from the ground to the atmosphere, and subsequent deposition of the radon-
daughters on the surface of the ground. Order of magnitude changes have been observed in the dose rate from
the uranium-238 decay series due to migration of radon from the ground and subsequent deposition of radon-
daughters (Adams, 1972). Short-term dose rate measurements are subject to more variation than long-term
integrated dose measurements. Even so, integrated month-to-month variations as large as 10% have been
measured (Burke, 1972), and monthly variations as high as 25% can be expected (EPA, 1972a).
-261 -
-------
These natural fluctuations complicate the measurement of 10 mrem/year attributable to operation of a
nuclear power plant, and indicate a need for a series of monthly or quarterly integrated dose measurements at
"indicator" and "reference" stations verified by annually integrated dose measurements. The purpose of this
paper is to discuss some of the factors that affect the accuracy of this type of measurement, present some of the
results obtained from 1968 to date, and discuss the design of an environmental monitoring program for
measuring external radiation dose.
METHOD
Lithium fluoride was selected as the dosimeter material because its response to gamma is relatively
independent of energy and it has a similar response/rad for penetrating beta radiation. Solid lithium fluoride
(TLD-100) chips, Vs" by Vs" by 0.035", were 100% selected for uniform response to gamma irradiation. They were
annealed at 400°C for one hour, and then overnight (16 hours) at 80°C, immediately prior to use. (For best
results, Eberline now anneals the dosimeters twice, immediately before each use period.) Each TLD badge
contained five dosimeters sealed in black plastic to protect the dosimeters from direct exposure to sunlight,
and then in clear plastic to further protect the dosimeter from weather. The total thickness of the plastic has
ranged from 20/mg/cm2 to 50/mg/cm2. The 50 mg/cm2 thickness has been found to be the most satisfactory
because the badges are more durable. Some evaluation data were also obtained using additional aluminum
shielding so that the total thickness was 300 mg/cm2. The dosimeters were read as soon as possible after
exposure with an Eberline Model TLR-5 thermoluminescent dosimeter reader. The reader was set to integrate
only the light output at temperatures between 150°C and 250°C. Using this method, Eberline has not been
able to detect any fading of dosimeters when exposed to radiation and then to extreme weather conditions over
extended periods of time, up to 15 months.
RESPONSE VS. RADIATION ENERGY
The radiation emitted from steam lines or tanks on-site, a semi-infinite cloud of gas, radioactivity in the
earth, cosmic scatter in the atmosphere, and other sources are scattered, absorbed, and reduced in energy so
that the environmental TLD must measure dose from a continuous spectrum of photon energies, ranging from
0 to the maximum energy of the emitted radiation. For a BWR, the gamma energy may be as high as 7 MeV.
Some of the low-energy radiation will not penetrate the skin or covering over the lens of the eye and does not
contribute significant dose to lens of the eye, gonads, other critical organs, or whole body. This non-
penetrating component is difficult to measure and its measurement is of doubtful value since the major
concern for large population groups is genetic and whole body dose (NCRP, 1971). Photons above 20 keV
contribute most of the whole body and genetic dose.
Some of the most sensitive TLD materials, e.g., CaSO^ :Tm, CaF2:Mn, and CaF2 :Dy, have responses that
are very energy dependent below 200 keV. The response of one glass encapsulated CaF2 :Mn system used
extensively for environmental monitoring is discussed in reference (EPA, 1972b), and shows an over response
as high as a factor of ten to energies below 200 keV unless an energy compensating shield is used. This is
typical of the data that can be expected from CaF2 and CaS04 dosimetry systems. The major disadvantage of
the shielded CaF2 or CaSO4 systems is that the response is very sharply reduced or eliminated below 70 keV.
Above 70 keV, the error due to energy dependence can be reduced to less than 110% with an energy
compensating shield. This is the major reason LiF is preferred to CaF2 or CaSO4 for this application.
The response vs energy for solid LiF dosimeters, 0.125" x 0.125" x 0.035", is shown in Table 2 for dosimeters
shielded with 50 mg/cm2 (plastic) and 300 mg/cm2 (aluminum and plastic). Prior to 1973, the TLD badges
supplied by Eberline for environmental monitoring were provided in plastic bags with no additional
aluminum shielding. If no attempt is made to measure non-penetrating skin dose, the total thickness of
material covering the LiF dosimeters should be 300 mg/cm2, which is the approximate thickness of the
covering over the lens of the eyes.
IN-TRANSIT EXPOSURE
Shipment of badges by mail or commercial carrier may result in some radiation exposure in transit The
variations due to in-transit exposure are indicated by the data in Table 3 for badges that were shipped by air
from Santa Fe, New Mexico, to Cedar Rapids, Iowa, and then returned immediately to Santa Fe for reading
Results for eight out of eleven shipments were within a 2 to 6 mrem range which can be attributed to natural
background. Three of the shipments were 10 mrem or higher, indicating some exposure above normal
background. Thus, in-transit exposure is an unpredicatable variable that causes a positive bias in some of the
results. To minimize in-transit exposure, Eberline has established three regional service facilities and
encourages in-plant annealing and reading for plants not located near one of these service facilities.
PRECISION OF MEASUREMENT
Each environmental TLD badge contains five dosimeters. The average reading for the five dosimeters is
considered the dose for the period from time of annealing to time of reading. The precision of measurement at
£*95% c°nfide? •ce^dJS mdlcated Jythe error t«™ m Table 2. At a dose level of 100 mrem, the precision
(95% confidence) is 3% to 5% for energetic gamma radiation, and approximately 10% for lower energy x rays
-262-
-------
MONTHLY VARIATIONS
Data from an extensive program of pre-operational environmental radiation monitoring at the Duane
Arnold Energy Center (DAEC) near Cedar Rapids, Iowa, are summarized in Table 4. These data provide an
indication of the magnitude of monthly variations in natural background readings. Time dependent
variations are primarily a function of weather conditions (snow, rain, and barometric pressure); however, the
data in Table 4 includes some variation due to annealing and reading which cannot be identified separately. A
total of 48 stations were monitored for one complete year. Results were grouped by reference stations (those
greater than five miles from the site), on-site stations, perimeter stations, and stations located one to three
miles from the site. The ranges of average dose rates in mrem per week for these four groups were 1.03 to 1.73,
1.26 to 2.24,1.06 to 1.85, and 1.10 to 1.89, respectively. The low-reading for each group occurred in mid-winter
(December, 1972), and the high-reading occurred in the fall (September, 1972). This is consistent with
published information about the effects of snow cover and soil moisture on natural background dose rates.
The magnitude of these monthly variations indicates the potential error associated with making
background dose measurements over a period of one month or less, and extrapolating that to an annual dose
rate. These data indicate the need for pre-operations TLD measurements at carefully selected locations over a
period of at least two years.
CONFIRMATION OF MONTHLY READINGS WITH
ANNUALLY EXCHANGED BADGES
The data in Table 4 may be used also to compare the annual average dose rate based on badges exchanged
monthly with others exchanged annually. The badges that were read monthly were annealed and read on-site
to avoid intransit exposure. One set of annual badges was annealed and read on-site and another set was
annealed and read in Santa Fe, New Mexico. Assuming that the DAEC annual badges provide the best
indication of the annual average dose rate, the results for annual badges read in Santa Fe seem to be biased
high by 3% to 5%. This is excellent agreement, but even this small bias may be due to in-transit exposure. The
annual average dose rates based on badges exchanged monthly seem to be high by about 25%. This may be
associated with the very low readings (near the sensitivity of LiF) obtained after only a month. TLD badges
exchanged quarterly and annually may provide better documentation than badges exchanged monthly and
annually.
COMPARISON OF INDICATOR AND REFERENCE STATIONS
Data were obtained in 1970 at the Dresden Nuclear Power Station to compare TLD and ion chamber results
at "indicator" and "reference" stations near an operating nuclear power station (EPA, 1972c). The data in
Table 5 and 6 were obtained by annealing and reading the TLD badges in Santa Fe, New Mexico; therefore,
the readings may include some in-transit exposure as previously discussed. Ion chamber readings should not
be considered absolute values because they are affected by energy response, temperature, atmospheric
pressure, humidity, calibration method, charge leakage, and readout error. The Dresden badges were placed
in a locked metal enclosure which also contained the air sampling equipment. This equipment causes some
shielding and also produces secondary electron scatter from metal surfaces. Even with known errors, the TLD
and ion chamber results indicate that on-site and site perimeter readings are significantly higher than more
distant off-site readings. Environmental TLD badges would be expected to provide the most usable data with
local read-out and a program that is designed to compare dose at "reference" (background) locations with dose
at "indicator" locations to distinguish changes due to natural causes from changes attributable to operation
of the nuclear power plant.
RECOMMENDATIONS
One of the most difficult ALAP measurements to make is 10 mrem per year of environmental radiation dose
attributable to the operation of a nuclear power plant. For this purpose, the following TLD program is
recommended using solid LiF dosimeters and the Eberline TLR-5 method of reading:
(a) Use local annealing and reading to avoid unpredictable in-transit exposure during shipments by mail or
commercial carrier.
(b) Anneal the dosimeters just before they are placed out in the environment, and read them within one or
two days after they are removed from the location being monitored.
(c) Use two sets of badges, exchanged quarterly and annually.
(d) Design the program to include reference stations located five to ten miles from the site, indicator stations
located on the site perimeter, and indicator stations located within three miles of the site perimeter. The
number of stations in each group depends on the site. Use of TLD is very economical in comparison to other
environmental measurements, and a large number of stations should be included in the monitoring program.
Averages for each group should be compared to help identify changes attributable to the nuclear power plant.
(e) Use pre-operational measurements to establish the relative average background dose rates for groups of
reference and indicator stations.
-263-
-------
REFERENCES
Adams, J.A.S., (1972), Rice University, Department of Geology, personal communication.
Burke, G.P., (1972), Thermoluminescent Dosimeter Measurements of Perturbations of the Natural
Radiation Environment, paper presented at the Second International Symposium on the Natural Radiation
Environment, Houston, Texas.
Environmental Protection Agency, (1972a), Natural Radiation Exposure in the United States,
Document ORP/SID 72-1.
Environmental Protection Agency, (1972b), Radiation Data and Reports, Vol. 13, No. 10, October 1972,
p. 539.
Environmental Protection Agency, (1972c), Radiation Data and Reports, Vol. 13, No. 11, November
1972, p. 601.
National Council on Radiation Protection and Measurement, (1971), Basic Radiation Protection
Criteria, NCRP Report No. 39, January 1971.
TABLE 1. Variations in Terrestrial and Total Background Dose Rates
at Different Locations Near the Aliens Creek Site.
Dose Rates (mrem/week*)
Station
Number
1
2
3
4
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
In-Situ
Terrestrial
0.62
0.47
0.50
0.45
0.50
0.37
0.39
0.62
0.76
-
0.54
0.40
0.35
0.76
0.25
0.47
0.55
0.47
0.32
0.37
-
0.42
-
Estimated
Total
1.3
1.1
1.2
1.1
1.2
1.0
1.0
1.3
1.4
-
1.2
1.0
1.0
1.4
0.9
1.1
1.2
1.1
1.0
1.0
-
.
1.1
TLD
IstQtr.
1.4
1.4
1.2
1.2
1.3
1.2
1.4
1.6
1.6
1.3
1.5
1.2
1.0
2.0
1.3
1.5
1.4
1.1
1.1
1.0
1.1
1.3
TLD
2ndQtr.
1.5
1.1
1.2
1.4
1.2
1.2
1.3
.
1.6
.
1.4
1.0
1.2
1.8
_
1.0
1.2
0.9
0.8
1.5
1.2
1.2
1.3
*Terrestrial dose rates were measured by in-situ gamma spectrometry. The cosmic-ray dose rate at the site
was estimated to be 33 mrem/year. The "estimated total" dose rates were obtained by adding the measured
terrestrial component to the estimated cosmic component. TLD results are based on annealing and reading in
Santa Fe without any correction for in-transit exposure, and is the total reading for the period divided by
number of weeks from annealing to reading. Each in-situ measurement was made on one day over a large
area, whereas the TLD integrates over a period of three months for a smaller area.
-264-
-------
Table 2. Energy Calibration of LiF TLD Area Badges*
(All Badges Exposed in Air to 100 mR).
Average of Ten Badges ± 2 a
Photon Energy and Source
23.7 keV, K-Fluorescence
58 keV, K-Fluorescence
120 keV, Filtered X-Ray
662 keV,l37Cs Gamma
1250 keV,60Co Gamma
50 mg/cm2
121+10
117+7
94+10
100+5
108+3
300 mg/cm2
76+6
119+8
99±9
100+5
98±3
*The TLD badge used for routine environmental monitoring has a total plastic thickness of 50 mg/cm2
covering the dosimeters. If only penetrating whole body dose is to be measured, an aluminum shield should be
used to increase the total absorber thickness to 300 mg/cm2.
Table3. In-Transit Exposure Via Air
(Santa Fe, New Mexico to Cedar Rapids, Iowa and Return*).
Date
Annealed
06-28-71
07-26-71
08-30-71
09-27-71
11-29-71
02-28-72
03-27-72
04-24-72
05-30-72
07-31-72
03-16-73
Date
Read
07-02-71
07-30-71
09-03-71
10-04-71
12-06-71
03-06-72
04-03-72
05-10-72
06-07-72
08-16-72
04-09-73
2+1
10+1
2+1
4+1
4+1
6+1
3±1
6±1
18+2
2±1
10+1
Readings Obtained
(mrem+2 a )
2+1 2±1 1+1
10±1
1+0
4±1
4±1
3±1
6+2 6+1
14±1
5+1
*Normal route via Denver or Chicago.
-265-
-------
Table 4. Monthly Variations in Dose Rates Measured with LiF TLD
(Duane Arnold Energy Center).
Average mrem/ week*
Period of
Exposure
Apr. 1972
May 1972
June1972
July 1972
Aug. 1972
Sep. 1972
Oct. 1972
Nov. 1972
Dec. 1972
Jan.1973
Feb. 1973
Mar. 1973
12-Month Average
Annual Badges:
On-Site Annealing
and Reading
Santa Fe Annealing
and Reading
Reference Stations
_> 5 Miles from Plant
1.51
1.52
1.52
1.55
1.50
1.73
1.48
1.29
1.03
1.42
1.47
1.23
On-Site
Stations
1.58
1.58
1.61
1.85
1.93
2.24
1.81
1.44
1.26
1.67
1.38
1.30
Perimeter
Stations
1.52
1.62
1.65
1.70
1.71
1.85
1.55
1.33
1.06
1.56
1.58
1.34
Stations 1
Miles from
1.57
1.67
1.75
1.77
1.89
1.89
1.62
1.36
1.10
1.60
1.61
1.41
to 3
Plant
1.44
1.27
1.31
1.63
1.31
1.44
1.53
1.33
1.37
1.61
1.32
1.46
*Data are for the pre-operational period. Any differences noted are not due to plant operation.
-266-
-------
TABLE 5. Comparison of TLD Data for Reference Stations and Indicator Stations
(Dresden Nuclear Power Station, 1970).
Net Dose*
Distance from Stack Ion-chamber TLD
Reading Location (miles) (mrem/yr) (mrem/yr)
On-Site:
#1
#2
#3
Perimeter:
Bennet
Breen
Hansel
Off-Site:
Lorenzo
Clay Product
McCabe
Channanon
Minooka
Coal City
Morris
Elwood
Wilmington
Lisbon
Joliet
Plainfield
Summary:
On-Site
Perimeter
Off-Site
NW
NE
S
1 NE
1.5 NNW
2.1SSE
2.5 S
2.9 WSW
4.2 NE
4.4 NNE
8.0 S
8.0 SW
8.2 E
8.6 SE
12.3 NW
12.7 NE
15.7 NNE
33
69
24
26
32
27
9
9
6
6
5
7
4
14
5
5
9
3
42
28
7
37
77
33
32
27
29
2
16
11
9
5
3
5
7
8
4
0
5
49
29
6
*Net dose is actual readings in 1970 with natural background subtracted based on 4th quarter of 1969 when
the Dresden Station was not operating. For details, see reference (EPA, 1972c).
TABLE 6. Comparison of Data for Indicator and Reference Stations.
Net Dose (mrem/a)
Location
On-Site 1
On-Site 2
On-Site 3
Average On-Site
Ion Chamber TLD
33 37
69 77
24 33
42 49
Perimeter 1
Perimeter 2
Perimeter 3
Average Perimeter
26
32
27
28
32
27
29
29
Average of 12 Off-Site
Stations > 2 miles
-267-
-------
Figure 1.
to
®
Qc
SPATIAL VARIATIONS IN TERRESTRIAL RADIATION
DOSE RATES FROM NATURAL RADIOACTIVITY
40
o>
>»
O)
>
i_
o
CO
cr
-------
RADON EMANATION FROM URANIUM MILL TAILINGS USED AS BACKFILL IN MINES
M. Raghavayya and A. H. Khan
Health Physics Division
Bhabha Atomic Research Center
Bombay-400085, India
ACKNOWLEDGEMENTS
The authors are grateful to Dr. A. K. Ganguly, Director, Chemical Group, Bhabha Atomic Research Center
and Mr. S. D. Soman, Head, Health Physics Division, Bhabha Atomic Research Center for guidance.
Appreciation is expressed for the assistance received from Mr. P. M. Markose, Mr. K. P. Eappen, and other
colleagues.
Abstract
As a standard practice, uranium mill tailings are returned to the mine as backfill to stabilize stoped out
areas. The tailings returned to the mine are the coarser fraction as separated by hydrocyclones. These
"sands" still contain significant quantities of radium. Being in a finely divided form, the sands offer a much
larger area for the emanation of radon as compared to the unbroken ore. The emanation of radon from the
backfill is, therefore, expected to be much higher than from the ore body itself.
This paper describes a method for estimating the radon emanation from mill tailings used as backfill. The
effect of increased emanation rates on ventilation requirements is discussed.
INTRODUCTION
The process of mining ore leaves voids underground which can be hazardous if not stabilized. A standard
practice in uranium mines, as elsewhere, is to use the tailings left over after extracting the useful minerals
from the ore as backfill. For this purpose, the tailings are separated into two parts with hydrocyclones of
which the coarser "sands" are used for backfill.
At Jaduguda in India, where the work being described was carried out, nearly 50 percent of the uranium mill
tailings was used as the backfill material. The sand slurry was pumped to the worked out stopes. The water
percolated through, and the sand settles under gravity. The sands are made up predominantly of particles
above 74 (jm in size.
While aiding in the stabilization of the ground, the mill tailings used for this purpose create certain
problems. Although uranium itself has been extracted, the tailings are still left with significant quantities of
radium, which give rise to radon. Being in a finely divided form, the sands offer a much larger surface area for
radon to emanate through. The sands, even after settling down to a consolidated mass, are still highly porous
as compared to the original ore. All these factors may combine to eventually increase the rate of radon
emanation into the mine atmosphere, and thereby raise the ambient levels of radon and daughter products.
This work deals with a laboratory method devised to study the rate of emanation of radon from the sands.
Emanation studies carried out under actual field conditions are also described.
EXPERIMENTAL DETAILS
1. Laboratory Setup.
The apparatus we built is shown in Figure 1. It consists of a chamber Cl with a B-55 standard joint at one
end and a tightly stretched thick polyster filter cloth at the other. The filter end can be sealed off completely if
necessary.
The sand sample normally occupies almost the entire volume of Cl, as shown in the figure. A second
chamber C2 which is open at both ends fits into the open end of Cl. The other end of C2 is connected to a third
chamber C3 with a wide collapsible rubber tube. The other end of C3 is closed with a single-hole rubber stopper
provided with a stopcock. The air space in the upper chambers C2 and C3 serves as the accumulation volume
for the radon gas emanating from the sample of the sand placed in the lower chamber Cl.
Sands are collected in chamber Cl in the form of a slurry from the tailings fraction sent to the mines for
backfilling. The slurry is generally filled up to the brim of the chamber. The water is allowed to drain
completely through the filter cloth by gravity. As the water drains, the sands get consolidated to a certain level
somewhat below the brim. The lower end is then sealed off and the chamber Cl is connected to the remaining
part of the emanation setup after flushing the space in Cl above the sands with air to remove all radon already
present.
Radon emanates through the sand surface and accumulates in the air space in the system. After a known
interval't' (usually on the order of a few hours), the passage between chamber C2 and C3 is pinched off with a
pinch cock. Next, an evacuated scintillation flask is connected to the stopcock and a sample of air from
chamber C3 is drawn into the scintillation flask. The purpose of pinching the passage between C2 and C3 is to
ensure that the partial vacuum created in the system during sampling does not enhance emanation from the
-269-
-------
sands. The stopcocks are closed and the scintillation flask is then detached from the system. The pressure in
the chamber C3 is equalized with the outside by introducing the required amount of radon-free air into it. The
pinch cock is then released. Sampling is repeated several times.
2. Field Setup.
Six sets of experiments were carried out under actual field conditions in some of the stopes in the mine,
where sand stowing was in progress. The apparatus consisted of a drum about 25 liters in capacity and with a
cross sectional area of about 600 cm2. It was open at one end and a stopcock was attached to the other end
which was otherwise closed. The drum was buried in the sand to a depth of about 20 cm, open end downwards.
Radon collected in the drum was measured twice, once immediately after inserting it in the sand, and again
after a time lapse varying between 7 and 65 hours.
THEORETICAL CONSIDERATIONS
Throughout this treatment the following notations are used:
1. Laboratory Experiment.
V = total volume wherein radon from the sand sample accumulates (liter)
Vj = volume of chamber C3 plus half the volume of the passage between C2 and C3 (liter)
V2 = volume of the scintillation flask (liter)
A = area of cross section of chamber Cl which is also the area of the surface through which radon
emanates, (cm2)
r j = concentration of radon measured from the i"1 air sample from each sand specimen, (Ci/liter), i = 1, 2, 3,
..... n
Rj = total radon activity in the system just before the collection of the itn sample, (Ci)
t = time interval between any two sampling instants
\ = decay constant of radon
J = radon emanation rate during any interval 't', (Ci/cm2.s)
and Vl _
It can be shown that, RJ = (P)rj (F) -V2 - (Q)
The sampling depletes the system of a part of the accumulated radon. The radon activity so removed is V2ri, so
that the total radon activity remaining in the system immediately after the ith sample has been drawn is
given by,
Ri = (Q)r; (2)
The (i + l)th sample is collected after a lapse of time 't'. During this interval, two simultaneous processes may
be imagined to be taking place within the emanation system, finally resulting in a total radon activity in the
system equal to R i+i.
(1) Activity of radon already present decays to a value exp"^ times the original activity.
(2) Due to emanation from the sand; fresh radon is introduced into the free space in the system, thereby
tending to increase the radon activity. The activity of radon introduced by the second process can be expressed
as,
nil _ A. x 3 _ -Ats
K ~ 2.097 x 10-6 (1 " exp } (3)
where 2.097 x 10-6 is the decay constant of radon expressed in s-1
Now, R = R exp~At + R"
= (Q) r. exp-t + -^^^(1 - exp^) (4)
and hence J = 2'°97 x
-\tl
± exp I
_ ,n. - ...
ri+1 - (Q)r exp I (5)
-270-
-------
2. Field Experiments.
In the field experiments, only two samples were collected in each case, the first at the commencement of
experiment and the second after a time t. The emanation rate in this case can be calculated from
2.097 x 10-6 _xt
At
In this case Vj = V, so from (1) and (2)
T . x _t
J = A (1-exp-At) (R2 - *{ exp At)
(V + V) r, and R = Vr
therefore, J = t . <* + *> * ~ e (6)
RESULTS
Table 1 presents the emanation rates obtained from the different sand specimens in the laboratory, and
Table 2 presents the emanation rate in the mine stopes. In the case of laboratory experiments, the 'J' values
were determined several times for each sand specimen. The mean of the several determinations is given. The
corresponding radium contents of the sand specimens are also given.
DISCUSSION
Table 1 shows that the emanation rates and the radium content of the different sand specimens are of the
same order, and we are, therefore, justified in calculating the mean values (Table 2). The emanation rate in the
stopes was about 4.5 times more than the corresponding value obtained in the laboratory. The reason is that
the sand mass in the stopes was very large compared to the small quantity of the sands used in the laboratory
experiments.
It may be concluded that the rate of radon emanation from the sands used as backfill at Jaduguda is
(27.2±9.5) x 10-18 Ci/cm2.s on the average. A gram of Jaduguda ore contains about 200 pCi of radium while the
radium content of the tailings sands is only a third of this value. Khan and Raghavayy (1972) have shown
that a representative radon emanation rate from the ore body itself is only 0.41 x 10-16 Ci/cm2.s despite the
higher radium content. The ratio of radon emanation rates of the sands to that of the ore, when radium content
is normalized, is 217. This is understandable in view of the higher porosity of the sand, and the increased
surface area as a result of the fractionation of the ore. The porosity of the sand is about 50% while that of the
ore was found to be about 0.5%. The increase in surface area was found to be about 470 times. The porous sands
are affected by changes in barometric pressure in the mine which is also likely to affect the emanation of
radon from the sands more than that from the ore.
At Jaduguda, it was observed that the radon concentrations in stopes where tailings sands were used as
backfill were more than the radon concentrations in other stopes. The average concentration of radon in filled
stopes was 120±15pCi/l while in other stopes it was 45 ±12 pCi/1, denoting an increase of radon concentration
by a factor of 2.7. To compensate for the elevated radon levels, ventilation in these stopes was correspondingly
increased.
It can be concluded that the use of uranium mill tailings as backfill in worked out stopes increases the
ambient radon concentrations as a result of an elevated emanation rate. The increase in emanation rate is so
large that the emanation from the ore body itself becomes negligible when compared to that from the sands.
Ventilation has to be increased in these stopes to cope with the increased radon emanation. The actual
increase must be worked out in each case.
REFERENCE
Kahn, A. H. and M. Raghavayya (1973), Radon Emanation Studies in Jaduguda Uranium Mine, paper
presented at the Third International Congress of the International Radiation Protection Association,
Washington, D.C.
-271 -
-------
TABLE 1. Radon Emanation Rates from Sand Samples
(Laboratory Experiments).
Emanation rate Radium content
Sample No. (xlO-18 Ci/cm2.s) (pCi/g)
SI 5.1 68.4
S2 5.2 92.3
S3 2.9 83.0
S4 6.7 69.4
S5 12.5 67.3
S6 4.8 66.9
S7 8.7 57.2
S8 4.5 45.9
S9 6.5 67.3
S10 9.2 58.7
Sll 7.8 53.6
S12 7.5 75.3
S13 5.0 78.3
S14 1.9 41.3
S15 2.1 44.2
Mean 6.0 ±1.3 64.6 ±6.6
TABLE 2. Radon Emanation Rates in Mines Stopes
(Field Experiments).
Emanation rate Radium content
Sample No. (xlO-18Ci/cm2.s) (pCi/g)
Ul 26.7 72.6
U2 27.2 69.3
U3 38.9
U4 13.3 51.5
U5 41.4 56.6
U6 53.1
Mean 27.2 ± 9.5 60.6 ± 9.4
-272-
-------
PlUCH COCK
FlUER. CLOTH
Figure 1. Emanation Set-Up.
-273-
-------
FISSION PRODUCT NOBLE GASES IN NUCLEAR POWER STATION OPERATION
P. Abraham and S. D. Soman
Health Physics Division
Bhabha Atomic Research Center
Bombay 400085, India
Abstract
This paper presents the experience with radioactive noble gases during the operation of the 400 MWe twin
Boiling Water Reactor station at Tarapur. The design measures to minimize, control, and monitor the
discharge of noble gases from the station into the environment are described. The limits prescribed for the
gaseous effluents from the station are given along with the assumptions used to derive them. Actual
environmental releases during the four years of station operation are tabulated. Gamma spectra and decay
curves of typical off-gas samples are presented and analyzed.
The effects of fuel defects on the isotopic composition of the gaseous releases, during operation and the
problems encountered in operating the reactors with defective fuel are discussed.
INTRODUCTION
India's first atomic power station (TAPS Report, 1969) is at Tarapur, on the west coast, about 100 km north
of Bombay. The power station has two dual-cycle boiling water reactors (BWR's), each of 210 MWe capacity.
The station started commercial operation in October 1969. At the end of July 1973, the station had generated
6,748 million kWh of electricity for the two states of Maharashtra and Gujarat.
The second station (RAPS Report, 1971), consisting of two identical units of 200 MWe each of the CANDU
type, is located on the Rana Pratap Sagar Lake, 67 km from Kota in Rajasthan State. One of these units
attained criticality in August 1972 and is now operating at about 100 MWe. The other unit is under
construction.
DESCRIPTION OF SYSTEMS
At Tarapur, the fuel is in the form of sintered UO2 pellets, stacked in the Zircaloy-2 tubes forming the fuel
rods. Thirty-six such fuel rods are arranged in a six-by-six array to form a fuel bundle. The Rajasthan reactor
also uses sintered UC>2 in Zircaloy-2 tubes. Each fuel bundle has 19 elements.
By far the largest amount of radioactivity in any reactor is produced by fission which is mostly contained
within the fuel itself. Gaseous and volatile fission products, which diffuse out of the fuel, are trapped by the
cladding and collect in the gas plenum provided in the fuel. Gaseous radioactivity can come out of the fuel if
the cladding is defective or develops defects during operation. Tramp uranium in the cladding also contributes
to fission product activity in the reactor water.
Removal of radioactivity from the reactor coolant takes place by several processes; namely, (1) radioactive
decay, (2) deposition on surfaces and accumulation in recesses as crud, (3) removal in the cleanup filter and
demineralizer, and (4) carry-over with the steam to the turbine and condenser (in the case of BWR's). It is the
last process which contributes mainly to the release of gaseous wastes to the environment during the normal
operation of the Tarapur Reactors. The different streams of gaseous radioactive effluents reaching the
Tarapur stack are shown in Figure 1. At Tarapur the radionuclides carried over to the turbine-condenser in the
steam phase are mostly xenon, krypton, nitrogen-16, and nitrogen-13. Nitrogen-16 decays rapidly in the
condenser due to its short half-life (7.3 seconds), and is almost completely removed from the waste steam
before it reaches the environment. Nitrogen-13 (half-life of 10 minutes) is the only activation product [160
(p,c<) 13N] worth considering among the gaseous wastes released to the environment. In fact, it is the most
important gaseous radioactive waste during the initial phase of reactor operation, when fuel defects are at a
minimum. However, with increased irradiation of the reactor cores, the fission gases have overshadowed the
release of nitrogen-13. BWR type reactors are not provided with on-load refueling facilities and, as such, they
operate with failed-fuel. Hence, the fission product noble gas release rate increases during the fuel cycle.
The Rajasthan reactors are designed for on-load fueling, so that defective fuel bundles can be replaced
immediately after detection with fresh ones. As such, fission product noble gases in the exhaust stream can be
minimized by careful operation. However, the use of heavy water as the moderator and coolant causes the
presence of tritiated water vapor in the exhaust. In addition, the use of air for cooling certain reactor
components contributes to the presence of argon-41 in the effluent stream.
Radioactivity from the Tarapur condenser is continuously removed along with other non-condensable
gases by the air ejectors. More than 99% of the gases in the steam go to the air ejector system. This stream flows
at the rate of about 20 liters per second from the condenser and is allowed to decay for about 30 minutes in a
holdup line of volume 60 m3, thereby almost completely eliminating short-lived radionuclides like 16N 89Kr,
9<>Kr, siRr, <"Kr, 137Xe, 139Xe, and "<>Xe. Before the gases are released to the base of the 100 m stack, they are
filtered using a high-efficiency particulate air (HEPA) filter. Gaseous wastes released to the stack are diluted
by ventilation air which flows at the rate of 55 m Vs. Less than 1% of the process gases escaping through the
turbine gland seals is separated from the steam by the gland seal condenser and is exhausted through a
-274-
-------
holdup line providing a delay of about 100 seconds. Some of the gases circulate in the primary reactor water.
This is indicated by the fact that any primary system leak is accompanied by the release of noble gases, which
are readily detectable by the presence of their particulate daughters. Airborne contamination in the occupied
areas of the station is almost always caused by such leaks and is due to noble gases and their short-lived
daughter products, especially 88Rb and 138Cs. In addition, argon-41 has been detected in the operational areas
at Rajasthan probably due to inadequate ventilation.
EXPERIMENTAL RESULTS
Gamma spectra of the off-gas sample from Tarapur Station 30 minutes and 6 hours after collection are given
in Figures 2 and 3. Figure 3 shows that 133Xe, 85mKr, and 135Xe are the only isotopes present after a few hours of
tf.ecay. The radioactive decay curve of the off-gas sample (Figure 4) indicates that about 3 days after collection
the entire activity is due only to 133Xe.
RELEASE LIMITS
ICRP has recommended that populations in the neighborhood of nuclear installations should not receive
radiation doses in excess of 500 mrems annually from all sources of radiation. To allow for environmental
dose from other routes, 50% of the ICRP limit is apportioned to the air route. Based on meteorological studies at
the two sites, release limits are specified for the two stations. The annual average continuous release limit for
noble gases works out to 246 mCi/s (21,250 curies/day) for Tarapur Station (Abraham, et al, 1973). At RAPS,
the limits (Sah and Subbaratnam, 1973) are as follows:
Noble gases — 2,500 Ci/day
Tritium — 29,500 Ci/day
Actual release rates at Tarapur during operation were well within the permissible limits, as is evident from
Table 1.
However, these releases were higher than the values expected by design due to defects in the fuel. It was
expected that, after refueling, the off-gas releases would be reduced. However, off-gas releases were not
significantly lower after refueling. The small relative difference in the radioactivity content of the samples
taken from the fuel bundles for identification of the defective bundles made the job very difficult. It should be
noted that operation with failed-fuel increases not only the gaseous effluent releases, but also the liquid
effluents and the in-plant radiation levels. In order to keep the radiation exposures to station personnel and
liquid wastes as low as possible, it is advisable to remove defective fuel as quickly as possible after
identification. When defective fuel is present in the core, any reactor transient (increase or decrease in reactor
power) causes the activity (in reactor water as well as in the off-gas) to increase and then to level off. This is
probably due to the thermal stresses caused by the transients. To overcome this problem for the environment,
some BWR's like the KRB nuclear power station (Eickelpasch, 1972) in Germany are using augmented off-gas
facilities for the past several years to enable operation of the station without off-gas affecting station output.
The KRB facility reduces the releases from the station by a factor of about 1,000. The radioactive gases
undergo adsorption and desorption processes and are, therefore, transported through the charcoal beds more
slowly than the accompanying air, thus giving the gaseous radioactive nuclides more time to decay during
their passage through the delay line. It is reported that it is possible to achieve a delay time of 14 days for
xenon. At present, proposals are being examined for the augmentation of the off-gas system at Tarapur by
additional holdup capacity to reduce the environmental releases. This is being done as a part of the efforts to
reduce environmental radiation levels to as low a level as practicable. Systems which have been proposed
include catalytic recombiners followed by low-temperature or cryogenic adsorption on charcoal, gas storage
under pressure, or cryogenic distillation. Operational experiences on these systems are not readily available.
DISCUSSION
The off-gas composition at the Dresden BWR is reported (Blomeke and Harrington, 1968) to be 40%
hydrogen, 20% oxygen, and 36% air. This means that at least 96% of the gas passing through the delay line is
non-radioactive, and in the company of the non-radioactive component, the radioactive gases are speeded
through the delay line into the environment. Effective removal of the accompanying gases by using a
recombiner (catalytic combustion of the hydrogen and oxygen with the aid of a palladium catalyst or other
techniques) can increase the delay time by a factor of 25 or more. The present delay line, which causes a delay
of 30 minutes with a flow of 201/s, will be able to provide a delay of 12.5 hours if the non-radioactive gases are
removed. Perusal of the decay curve would show that 12.5 hours delay will reduce the concentration by a factor
of about ten, and the off-site dose (Martin, et al., 1973) by a factor of about 25. This would be enough to reduce
the annual environs dose from stack effluents at Tarapur to less than ten millirems even for the periods when
the off-gas generation rate is as high as 500 mCi/s. During normal operation, the generation rates in BWR's
are considerably lower, and, hence, the dose rates will be less. In other words, the provision of facilities to
remove the non-radioactive gases from the stream reaching the delay line would be adequate to reduce the
environmental release rates from BWR stations so as to achieve the "as low as practicable" limits currently
-275-
-------
under consideration throughout the world. Further increases in holdup may not be necessary to achieve this
objective. Martin, etal., (1973), have reported that the discharge of processed effluents after 25-30 days holdup
would reduce the off-site dose by a factor of 5,000 at a substantial increment of cost.
Argon-41 concentrations are constant when the power level is not varied. Isolating the exhaust from areas
where D20 leakage is possible, and condensing the moisture from it, can substantially reduce environmental
releases from CANDU-type stations in addition to recovering the heavy water if on-load fueling is adopted
scrupulously to remove defective fuel immediately after detection.
REFERENCES
Abraham, P., D. Pattnaik, and S. D. Soman, (1973), Safety Experience in the Operation of a BWR
Station in India. Paper presented at the IAEA Symposium on Principles and Standards of Reactor Safety.
IAEA-SM/169-26.
Bloemke, J. O. and F. E. Harrington (1968), Management of Radioactive Wastes at Nuclear Power
Stations. ORNL-4070.
Design and Analysis Report of the Tarapur Atomic Power Station, Tarapur, India (1969).
Eickelpasch, N. (1972), Personal communication.
Martin, J. E., F. L. Galpin, and T. W Fowler (1973), Technology Assessment of Risk Reduction
Effectiveness of Waste Treatment Systems for Light Water Reactors. Radiation Data and Reports 14,1.
Safety Report; Rajasthan Atomic Power Station (1971).
Sah, B. M. L. and T. Subbaratnam (1973), The Calculation ofDWL and ARE Values for the Rajasthan
Atomic Power Station. Paper presented at the Nuclear Science and Engineering Symposium, Bombay.
TABLE 1. Gaseous Radioactivity Releases from Tarapur.
Year Release rate, mCi/s
1969 3.0
1970 19.2
1971 67.2
1972 38.7
1973 55.2
(up to June)
-276-
-------
REACTOR &UILDIUG
EX.HAUST
EMERGENCE
flL.TE.HS
REACTOR BUILDING
X! ISOLATION VALVE.
R.M. RADIATION MONITOR
C.R. CONTROL ROOM
Figure 1. Gaseous Effluent Streams atTarapur.
-------
to
-a
ao
10,000
1,000
i 100
I
10
Figure 2. Gamma Spectrum of Tarapur Off-Gas Sample (Delay - 30 mm.)
zoo
4OO 600 SOO 100O
14-00
ENERGY, keV
iBOO
22OO
Z600
3OOO
-------
10,000 hr
1,000 hr
8
Figure 3. Gamma Spectrum of Tarapur Off-Gas Sample (Delay - 6 hrs)
too t
O. 7
0.3
ENERGY, MeV
0.6
0.7
-------
10'
ts
00
O
10
Figure 4. Radioactive Decay Curve of Tarapur Off-Gas Sample
so
1ZO
160
zoo
-------
NOBLE GAS SURVEILLANCE NETWORK, APRIL 1972 THROUGH MARCH 1973*
V.E. Andrews and D.T. Wruble
Monitoring Operations Division
National Environmental Research Center
U.S. Environmental Protection Agency
Las Vegas, Nevada
Abstract
In April 1972, the National Environmental Research Center-Las Vegas initiated a ten-station network of
continuous air samplers to monitor noble gas radioactivity on and around the nuclear explosive testing
ground (Nevada Test Site). The network uses compressed air samplers designed and built by the Center, The
samplers are described and operating experience and sample results are reported. After one year of operation,
the network has shown its ability to document ambient levels ofkrypton-85 and to detect the occasional low-
level releases ofxenon-133 at the Nevada Test Site. Analysis of the data shows a small, but significant, higher
level ofkrypton-85 concentrations measured at on-site locations compared to off-site. Average concentrations
over the year were 15.7 ±2.4 pCi/m3 off-site and 16.2 ±2.4 pCi/m3on-site. A systematic variation in krypton-85
levels at off-site stations was similar to that found in weekly grab samples collected off-site in Las Vegas
during the preceding year.
INTRODUCTION
Since 1958, the National Environmental Research Center-Las Vegas (NERC-LV) has operated a system of
surveillance networks to monitor the nuclear testing activities at the Nevada Test Site (NTS). The purpose was
to document the sources and magnitude of radiation exposure to the off-site population. Prior to 1971, these
networks monitored external gamma radiation, and routinely collected air, water, and milk samples which
were analyzed for various radioisotopes. For the types of radioactive releases normally occurring at NTS,
these networks successfully documented the significant sources of exposure.
The sole transport medium for radioactivity to the off-NTS area is the atmosphere. Accordingly, an
extensive air sampling program was conducted throughout the continental United States west of the
Mississippi River. (EPA, 1972). Until September 1973, integrated 24-hour samples were collected daily and
mailed to NERC-LV for analysis. Three 48- to 72-hour samples are now collected each week. All stations collect
airborne particulates on glass fiber filters. Selected stations routinely use activated charcoal cartridges after
the filter for the collection of gaseous radioiodine. These cartridges are also used at the other stations
whenever needed.
For the past several years, the underground testing program has experienced a high degree of success in
preventing accidental releases of radioactivity to the atmosphere at detonation time. However, at least two
pathways exist for releases of small amounts of radioactive noble gases. First, some gases may be released
during post-detonation drilling for the collection of samples for radio-chemical analysis. Second, slow seepage
of gases to the surface may allow some additional noble gas radioactivity to leave NTS.
Beginning in 1970, NERC-LV began collecting a weekly 20-minute compressed air grab sample at two
locations on NTS and one at NERC-LV. These samples were analyzed for 85Kr and radioxenon. In 1971, the
Center was requested by the AEC-Nevada Operations Office to establish a network for continuously
monitoring radioactive noble gases and atmospheric 3H at ten locations on- and off- NTS. The weekly grab
sample collection was continued until the ten-station continuous sampling network was put into operation in
April 1972. (Figure 1 shows the station locations.) This report covers the first full year of continuous
monitoring at these stations.
NETWORK REQUIREMENTS
Results of the weekly grab samples collected at NTS demonstrated that any radiation exposures off-site
would be very low. In order to adequately monitor the impact of any small releases of noble gases on the off-site
populated areas, a monitoring system capable of detecting any change in ambient levels was desired. The
Technical Services Division at NERC-LV had recently completed development of a new gas analysis system
capable of detecting 2 pCi/m3 of radioxenon or radiokrypton in a 1-m3 sample of air (Stevenson, et al., 1971).
Atmospheric krypton samples analyzed at NERC-LV and at the Eastern Environmental Research
Laboratory, Montgomery, Alabama (Shearin, et al., 1971 and Jaquish, et al., 1973), showed that current
ambient levels of 85Kr are approximately 16+2pCi/m3. Theoretically, the normal background for 133Xe and
135Xe should be zero, because of infrequent releases and short half-lives.
The design concept, then, was to provide a sampler capable of taking advantage of the high-sensitivity
capability of the laboratory for radioxenon, and to be able to detect changes in the 85Kr concentration. Since
*This surveillance performed under a Memorandum of Understanding (No. A T(20-l)-539) for the U.S. Atomic
Energy Commission.
-281 -
-------
the samples would be compressed air, it would be necessary to have them collected at regular intervals by
NERC-LV personnel to eliminate shipping problems. The logistics of sample collection from the 10,000 square
mile area, plus analytical time required in the laboratory, dictated a weekly collection schedule * mal design
considerations were that the sample should be unfractioned air, it should be collected at a constant rate over
the 1-week period, the system should be as reliable as possible, and the entire operation must be economically
feasible.
SAMPLER DESIGN
The basic design was similar to a sampler in use by Isotopes, Inc. (now Teledyne Isotopes). A schematic of
the sampler is shown in Figure 2, and a photograph of a typical station is shown in Figure 3. An aquarium
aerator pump draws air at 3 cmVsec through a glass fiber filter, and pumps it into a 38-1 low-pressure tank. A
pressure-actuated switch activates a solenoid, starting a high-pressure compressor. The compressor pumps air
from the low-pressure tank, and through a manifold, to two 38-1 high-pressure storage tanks connected in
parallel. When pressure in the low-pressure tank drops to 1.5 mm Hg, a second pressure-actuated switch de-
activates the solenoid, stopping the compressor. Sample pressure in the high-pressure tank is about 2.8 MPa
(400 psig).
The high-pressure tanks are collected weekly, and replaced with evacuated tanks containing one ml of
carrier xenon. One tank is analyzed in the laboratory, while the second serves as a backup for analysis in case
the sample in the first tank is destroyed, lost during analysis, or exhibits unusually high concentrations. The
actual standard sample volume in a pressure tank is determined by weighing the full tank as it enters the
laboratory, and subtracting the tank tare weight. A normal volume is approximately one m3. A separate
sampler was designed for collection of atmospheric moisture, and H2 for 3H analysis, the satisfy the
requirements of the Atomic Energy Commission. Methane gas collected by the compressed air sampler is also
analyzed for 3H as Cf^T. The operation and results of the 3H sampling network will be described in a later
paper.
NETWORK EXPERIENCE
After 1 year of operation, the noble gas surveillance network can be termed successful. It has demonstrated
the ability to satisfactorily meet most design requirements, although design changes are being studied to
improve sampler operational reliability. The network produced successful sample results 85 percent of the
time. Although most of the lost samples were due to sampler failure, some resulted from loss of sample during
laboratory analysis. Because of the sample work load in the laboratory, duplicate analysis was not performed
on the few samples lost in analysis — unless results of other stations indicated the possible presence of
elevated levels over the network. Table 1 shows the success rate by month. The improved performance during
the final 5 months of the period was due, primarily, to replacement of sub-standard components, which were
responsible for the high failure rate experienced during the first few months. Based on 85Kr results, the total
collection and analytical process has proved to be as accurate as any technique currently available. As shown
in the following section, the network is capable of documenting small changes in 85Kr concentrations. As
expected, most radioxenon concentrations have been below the 2 pCi/m3 detectable level; however, several
samples have contained measureable 133Xe.
Table 2 summarizes the results of 85Kr analyses for the ten stations for the period April 1972 through March
1973. It can be seen that the average concentrations measured are very near the 16 pCi/m3 determined from
other studies. The ranges observed, and number of results at the extremes, are consistent with statistical
predictions based on counting data.
The total annual cost of the noble gas sampling network is about $31,000. The average cost per sample,
assuming a 100 percent success rate, is about $60, including approximately $10 for collection and $50 for
analysis. The actual cost attributed to operation of the network is greater than the incremental cost incurred
by NERC-LV; this is because some of the personnel and equipment used to operate the network were also
necessary to provide the basic capability for off-site surveillance and noble gas analysis of event-related
activities. Since the network includes an atmospheric 3H sampling system, the field collection costs must be
shared by that system, as well. Therefore, the actual cost of operating the noble gas sampling network at
NERC-LV is somewhat less than it would be if a 3H sampling network was not operated in conjunction with it.
DATA ANALYSIS
Several questions arise from the operation of a new network such as this:
(1) Do differences exist between on-site and off-site stations, or between off-site stations in different sectors?
(2) Do releases of radioactive noble gases occur which are detectable at off-site locations?
(3) Are the background concentrations of radioactive noble gases constant and predictable?
(4) If the background is constant or predictable, what increase above background can be considered
detectable for a single sample?
(5) What sources of random variation are associated with sampling, analysis, and background
fluctuations?
-282-
-------
(6) Is there a time-dependent, or seasonal, variation in the background concentration of 85Kr?
(7) What effects of meteorology are reflected in on-site and off-site station comparisons, and between
stations ?
An attempt has been made to answer these questions, although additional data will be required to
satisfactorily answer all of them. A superficial examination of the data in Table 2 shows that the average
concentration on-site is greater than the average concentration off-site, with the greatest average at B JY. This
was expected, as BJY was selected on the basis of terrain, location, and meteorology to reflect the higher on-
site concentrations. Other areas on-site may have higher concentrations, but are not as accessible or do not
have power available. The data were subject to X2 tests to determine if they were drawn from a population
having a normal distribution. At the 5% level of significance, the sample distributions are consistent with the
hypothesis that the population distribution is normal. In Table 3 are listed the average monthly
concentrations for off-site stations and on-site stations. In eight cases the on-site concentrations are greater
than off-site, and in one case they were equal. An analysis of runs (Crow, et al., 1960) was performed which
showed that at the 5% level of significance the two samples have the same distribution. A t-test on the data
(Crow, et al., 1960) in Table 3, to determine if there is a significant difference in the means, showed the means to
be different at the 10% significance level, but not at the 5% level. A similar analysis was performed on the
weekly averages. In this case, 31 out of 48 on-site averages were greater than off-site. A comparison of weekly
averages, to determine if a significant difference exists between on-site and off-site means, shows a significant
difference at the 5% (and 1%) level of significance. An analysis of variance, performed on the weekly on-site
and off-site averages, showed the mean difference to be significant at the 10% level of significance, but not at
the 5% level. Since these analyses indicated significantly higher concentrations on-site than off-site, it can be
stated that releases of 85Kr occur which are detectable at on-site locations.
Prevailing winds at NTS are generally from the south to south-west during the summer months, and from
the north during the winter months. If the prevailing winds carry noble gases from NTS to the off-site areas,
comparison of stations north of NTS to stations south of NTS would be expected to demonstrate any resultant
effects.
Weekly and monthly average concentrations of stations to the north of NTS were compared to those to the
south. The northern stations are Diablo Maintenance Station, Hiko, and Tonopah. Those to the south are
Death Valley Junction, Desert Rock, Beatty, and Las Vegas. The monthly averages are summarized in Table
4. For 9 months out of 12, the northern stations averaged higher than the southern. An analysis of runs shows
that at the 5% significance level, both samples are from the same distribution. A comparison of monthly
averages, to determine if any difference exists between the means, showed no difference at the 5% level of
significance.
Twenty-nine weekly averages were higher for the northern stations compared to 19 for the southern
stations. An analysis of runs indicates that, at the 5% level of significance, the two samples were taken from
the same distribution. A t-test of the hypothesis that the difference in the means is zero could not be rejected at
the 5% level of significance. An analysis of variance was performed to compare northern to southern stations
for two periods of time — April through September, and October through March. In neither case could the
hypothesis that the means are equal be rejected at the 5% level of significance. All tests for both monthly and
weekly averages indicate that the 85Kr concentrations measured to the north of NTS were not significantly
different from those to the south.
From the data generated to date, it is not possible to state what increase in 85Kr above background can be
considered detectable from a single sample; nor can the background be stated with great precision. Jaquish, et
al., (1973) found that 85Kr in grab samples collected in Las Vegas from April 1971 to March 1972 followed a
second degree polynomial curve of Y = 0.0386X - 0.00006X2 + 10.9; where Y = concentration in pCi/m3 and X =
days after January 1,1971. The equation was found to be significant at the 5% level of significance, but not at
the 1% level. Concentrations of 85Kr, measured with the continuous sampling network from April 1972 to
March 1973 at Las Vegas, and averaged over the off-site stations, followed a similar curve. However, since two
years of data collected by the same technique are necessary to accurately determine any cyclic or periodic
variations in the background, no attempt has been made at this time to do so. No ready explanation is
available to account for such a variation since the observed annual maximum (if there really is one) occurs in
September, rather than in the late spring as with the peak of particulate radioactivity from stratospheric
fallout. This phenomenon will be investigated further after another year has passed.
Both the data presented by Jaquish and Johns, and that shown in Table 3, indicate that the annual
variation in 85Kr concentration may be as much as 3 pCi/m3. It is, therefore, important to establish this
variation as accurately as possible. Some determination must also be made as to the relative contributions to
the total error from counting and analytical errors. Since June 1973, a duplicate analysis has been performed
each week on one sample selected at random, using the spare high-pressure cylinder. It will take
approximately a year to collect enough data to assess the probable analytical error. If, and when, an annual
variation can be described, and the counting and analytical errors can be assessed, it will be possible to
establish a background, and a detectable level above background, for a single sample.
The problem is much simpler for radioxenon. As stated earlier, the theoretical background for 133Xe and
135Xe should be zero. With a detection limit of 2 pCi/m3 for radioxenon, essentially any increase above
background can be considered detectable. Table 5 lists those few samples, out of the 287 collected, which
-283-
-------
contained measurable 133Xe. In most cases, they are well above the detectable level. Positive 133Xe results are
not well correlated with higher 85Kr results in the same samples. Several of the higher 133Xe concentrations
were associated with higher concentrations of 85Kr, but others were associated with average or below average
85Kr concentrations.
CONCLUSIONS
One year of operation of the continuous noble gas sampling network has demonstrated success in producing
a workable sampling system, and in measuring ambient concentrations of 85Kr and 133Xe. The failure of
several sampler components to meet specifications resulted in a large number of sampler malfunctions during
the first few months. Replacement with suitable parts resulted in greatly improved reliability.
Although insufficient data is available to predict background concentrations of 85Kr as a function of time of
year, and minimum detectable increase above background for a single sample, it is anticipated that such
information will be available after another year of operation. Statistical analysis of the 85Kr data indicates
that although no individual samples could be said to be above background, average concentrations on the
NTS were above those off-site. Average concentrations north of the NTS were not significantly different than
those to the south for the entire year or for summer and winter seasons. The conclusion is that low-level
releases of 85Kr occur at NTS which are detectable on-site, but which do not result in measureably increased
concentrations in the off-site area.
Measurable concentrations of 133Xe were collected at off-site stations five times during the year. The
maximum concentration of 133Xe measured off-site was less than 1 percent of the radioactivity concentration
guide (AEG, 1963) for exposure to a suitable sample of the population.
REFERENCES
Crow, E. L., F. A. Davis, and M. W. Maxfield, (1960), Statistics Manual, New York, N. Y., Dover
Publications, Inc. p 56, 60-62, 101-102.
Environmental Protection Agency, (1972), Radiation Data and Reports, Volume 13, Number 2 (Feb.
1972) p 88-92.
Jaquish, R. E., and F. B. Johns, (1973), Concentrations of Krypton-85 in Air, (To be published).
Shearin, R. L., C. R. Porter, and S. L. Cummings, (1971), Study of the Feasibility of Measuring S5Kr
Through a National Surveillance System, IAEA-SM-148/ 64.
Stevenson, D. L. and F. B. Johns, (1971), A Separation Technique for the Determination of Krypton-85
in the Environment. Proceedings of Symposium on Rapid Methods for Measuring Radioactivity in the
Environment, IAEA-SM-148/68, p 157-162.
U. S. Atomic Energy Commission, (1963), AEC Manual, Chapter 0524; Standards for Radiation
Protection.
TABLE 1. Total Sampling Success Rate by Month
April 1972 through March 1973.
Month Percent Successful Samples
April 75
May 78.5
June 68
July 88.5
August 71.5
September 74
October 75
November 89
December 77
January 95.5
February 86
March 93
-284-
-------
TABLE 2. Mean Annual 85Kr Concentration by Station
April 1972 through March 1973.
Station
Area 12
BJY
Gate 700
Location
On-Site
On-Site
On-Site
Mean Annual
Concentration
(pCi/m3)
15.8
17.2
15.7
Standard
Deviation
(pCi/m3)
2.6
2.9
1.9
Range
(pCi/m3)
11-23
12-25
13-23
Average
On-Site
16.2
2.5
11-25
Beatty
Diablo
Death Valley
Junction
Desert Rock
Hiko
Las Vegas
Tonopah
Average
Average
Off-Site
Off-Site
Off-Site
Off-Site
Off-Site
Off-Site
Off-Site
Off-Site
All
15.8
16.0
15.0
15.7
15.6
15.5
15.9
15.7
15.8
2.4
2.2
3.5
2.6
1.6
1.9
2.0
2.4
2.4
12-22
12-22
10-25
12-25
12-19
10-18
12-21
10-25
10-25
TABLE 3. Mean Monthly 88Kr Concentrations
On-Site and Off-Site Locations
April 1972 through March 1973.
On-Site Average
Off- Site Average
Concentration
Month (pCi/m3)
April
May
June
July
August
September
October
November
December
January
February
March
16.5
17.4
14.7
16.6
16.3
18.0
15.6
15.1
15.5
17.9
17.2
15.0
Standard
Deviation
(pCi/m3)
1.9
2.1
1.5
3.1
2.6
3.5
2.2
2.2
2.2
2.6
3.1
1.5
Concentration
(pCi/m3)
16.1
15.8
15.2
16.5
16.9
18.0
15.1
15.3
15.1
15.7
15.0
14.6
Standard
Deviation
(pCi/m3)
1.5
1.6
1.9
2.2
2.8
3.0
2.0
2.1
2.2
1.5
1.0
1.6
-285-
-------
TABLE 4. Mean Monthly ^Kr Concentrations
North and South Off-Site Stations
April 1972 through March 1973.
Month
85Kr Average Concentrations, pCi/m3
North (a) South (b)
April
May
June
July
August
September
October
November
December
January
February
March
15.8
16.3
15.5
17.0
16.2
17.6
15.7
15.7
15.2
15.5
15.5
15.1
16.2
15.5
14.9
16.2
17.4
19.0
14.7
15.1
15.0
15.1
14.6
14.3
(a) Diablo Maintenance Station, Hiko, Tonopah.
(b) Beatty, Death Valley Junction, Desert Rock, Las Vegas.
TABLE 5. Positive »33Xe Results
April 1972 through March 1973.
Collection
Period
Station
Concentration,
(pCi/m3)
3/22-3/29/72
5/4-5/10/72
5/8-5/15/72
5/15-5/22/72
5/22-5/30/72
5/23-5/31/72
5/24-6/1/72
6/12-6/19/72
6/14-6/21/72
7/24-7/31/72
7/24-7/31/72
Area 12
Diablo Maintenance
Station
Desert Rock
BJY
BJY
Beatty
Hiko
BJY
Hiko
BJY
Area 12
14
33
30
380
530
17
570
430
230
9.9
4.6
-286-
-------
DUCKWATER^t
CURRANT MAINT. STA.-»
CURRANT V
'I
I
© ROUND MTN.
GEYSER
MAINT. STA.
BLUE JAYXC
MAINT. STA.
SUNNYSIDE
(
°NYALA
I >
& A'DAVEN
X
CLARK'S STA.
TONOPAH
DIABLO
MAIixm
STA.
TEMPIUTE /
if\ e /.
HIKOf>A
GOLDFIF.l.D
COYOTE SMT.
HANCOCK SMT.
r !
GATE
NELLIS
AIR FORCE
RANGE
*SPRINGDALE
LATHROP WELLS-J!
A
DESERT ROCK
INDIAN SPRINGS
FURNACE
CREEK
DEATH
VALLEY JCT.
PAHRUMP
LAS VEGAS
SHOSHONE
NOBLE GAS
SAMPLING STATIONS
RIDGECREST
Figure 1. Noble Gas Sampling Station Locations
-287-
-------
to
00
00
S olenoid r~~" \
Pressure Switches
Filter
Low-Pressure
Tank
High-Pressure
Tank
High-Pressure
Tank
Air Line
Control Line
Figure 2. Schematic of Noble Gas Sampler
-------
to
00
-------
V. Separation and Containment
of Noble Gases
-------
GENERAL SURVEY OF TECHNIQUES FOR SEPARATION AND CONTAINMENT OF NOBLE
GASES FROM NUCLEAR FACILITIES
C. L. Bendixsen and J. A. Buckham
Allied Chemical Corporation
Idaho Falls, Idaho
Abstract
Radioactive noble gases produced during uranium fission constitute a waste material requiring special
treatment and handling techniques if releases to the environment are to be reduced or minimized. To date,
treatment has been confined to short-term holdup of reactor off-gases to allow short half-life fission products
to decay. Although the required degree of noble gas separation and containment is not yet defined,
considerable development effort has been expended on a variety of separation processes. With some
separation processes, the technology is sufficiently advanced that commercial separation units are now being
offered for sale. Early installations of these will be mainly at reactors where their prime advantage is a
reduction in size of the holdup system required and an increase in the holdup time for radioactive decay.
Eventually, separation processes may be installed at fuel processing plants to collect the long-lived krypton-
85. The separation processes potentially applicable to removing noble gas fission products from off-gas
streams include absorption in liquids (liquid air, carbon dioxide, or fluorocarbons), adsorption on charcoal
and other solids, thermal diffusion, electrostatic diffusion, and diffusion through a selective membrane.
These processes are at a variety of technological levels and each has its own particular advantage or
disadvantage. A cryogenic liquid-air absorption process has been operated intermittently at the Idaho
Chemical Processing Plant for many years to recover krypton-85 for a variety of uses, but total recovery has
never been attempted. High pressure storage in steel cylinders presently appears to be the most practical
means of ultimate disposal of the krypton-85. However, inclusion of the krypton-85 into glasses, resins,
clathrates, molecular sieves, and metals offer ways of reducing the effective vapor pressure in a cylinder.
Long-term disposal means that have been considered including placement of the cylinders into an engineered
storage facility, dumping of cylinders into the sea, and direct discharge of the krypton-85 into an appropriate
geologic formation.
INTRODUCTION
Radioactive noble gases produced during uranium fission constitute a waste material requiring special
treatment and handling techniques if releases to the environment are to be reduced or minimized. To date
treatment has been confined to the short-term holdup of nuclear reactor off-gases to allow for the decay of
short half-life fission products. Although the required degree or efficiency of noble gas separation and
containment is not yet determined, considerable development effort has been expended on a variety of
separation processes. With some separation processes, the technology has advanced sufficiently for
commercial separation units to be offered for sale. Early installations of these units will mainly be at reactors
where their prime advantage is a reduction in the size of the holdup system required and an increase in the
holdup time for radiolytic decay. Eventually, separation processes may have to be installed at nuclear fuel
reprocessing facilities to collect the long-lived krypton-85.
NOBLE GAS SEPARATION
The separation processes (Nichols, et al, 1971; Dunster, et al., 1970; Kovah, 1970; Slansky, et al., 1969;
Slansky, 1969; and Swieger, 1973) potentially applicable to removing fission-product noble gases from the off-
gas streams include absorption in liquids (liquid air, carbon dioxide, or fluorocarbons), adsorption on
charcoal or other solids, and diffusion through a selective membrane. These processes are at a variety of
technological levels, and each has its own particular advantage or disadvantage (summarized in Table 1)
with respect to the recovery efficiency, purity and the ultimate containment requirements of the noble gases.
The technological level, advantages, and disadvantages of each process are discussed individually as follows:
1. Cryogenic Distillation.
The major advantage of the cryogenic distillation process for absorption of noble gases is its present high
technological level. Liquid air plants have been in existence for decades and, thus, considerable knowledge
has been assembled on materials of construction, valves, compressors, distillation column design, modes of
operation, and reliability. Specific cryogenic processes recovering natural krypton from air have also been
operated for some time. Because of this projected capital and operating costs for a fission-product noble-gas
removal system are well defined. In addition, the cryogenic distillation process is the only process to have
operated on a significant scale at a fuel reprocessing facility for the recovery of production quantities of
krypton-85 (AIRCO, 1958; and Bendixsen, et al., 1968, 1971, and 1973). A number of U.S. companies are
offering similar units for use at nuclear facilities with projected krypton-85 recovery efficiencies typically in
excess of 99 percent. A significant disadvantage of the cryogenic process is the explosion hazards associated
-290-
-------
with the radiolytic formation and the concentration of ozone. Methods to minimize this specific exp'osion
hazard include either total removal of oxygen from the process gas (Davis, etal.,1973) or dilution of an ozone-
bearing stream through periodic flushing or dilution (Bendixsen, 1973) using appropriate operating
parameters. Adequate control of krypton-85 leakage rates to achieve high recovery efficiencies has not been
demonstrated yet, but appears achievable. A high degree of process gas cleanup is required, and product
purity is relatively high.
2. Absorption in Fluorocarbons.
A significant decrease in refrigeration costs can be achieved by using a fluorocarbon (Stephenson, et al.
1970) such as Freon-11 or 12, to absorb the fission-product noble gases at temperatures of -30 to -90°F and at
pressures up to 400 psig. Process system size, mode of operation, and capital costs are comparable to the
cryogenic distillation process, although the quantity of mechanical cooling equipment is higher. The high
pressure required for the process increases the possibility of gas leaks and product losses. The fluorocarbon
process has been extensively pilot-plant tested at ORNL, but additional information on the effect of process
gas contaminants is desirable. The solvent cost is relatively low, but solvent losses through volatilization and
radiation degradation need more definition. Explosion hazards with ozone or xenon-tetrafluoride are
estimated to be very low. No plant-scale tests at a power reactor or a reprocessing facility have been performed.
3. Absorption in Carbon Dioxide.
Absorption of the noble gases in distillation processes using liquid carbon dioxide appears possible for gas
streams rich in carbon dioxide, such as those coming from reprocessing of graphite fuels (Glass, et al., 1972a, b
and Wheatley, 1973). Scoping tests of the process have occurred at ORNL and are continuing in order to obtain
basic carbon dioxide absorption data of the noble gases. Refrigeration needs, operating pressures, process
equipment, mode of operation, and temperatures are similar to the fluorocarbon process. No explosion
hazards from ozone or other radiolytic products have been identified, and a ready supply of solvent is
availablein the process gases. The process is specific for gas streams with high carbon dioxide concentration
and is not applicable to other gas streams.
4. Ambient-Temperature Charcoal Adsorption.
Adsorption of fission-product noble gases on charcoal beds at ambient temperatures has been used for
delaying the release of radioactive off-gases from nuclear power reactors (Dunster, et al., 1970 and Keilholtz,
1971). No attempt has been made to use the ambient-temperature adsorption process specifically as a recovery
system for fission-product noble gases. The quantity of charcoal required for a typical nuclear facility is in the
tens-of-tons range; thus, process equipment would be large and bulky, although operation could be relatively
simple. A distinct fire hazard exists, although fires are easily extinguished by shutting off the flow of oxygen-
containing gases. Use of adsorbent material other than charcoal cannot be considered because of increased
bed volume and high unit costs. Due to the quantities of charcoal required, such a recovery process may be
limited to small scale, batch processes.
5. Low-Temperature Charcoal Adsorption.
By reducing the temperature of the charcoal bed, the adsorption capacity is increased, thus reducing the bed
volume required (Dunster, et al., 1970 and Keilholtz, 1971) by 50 to 80%. Although a slightly more concentrated
product then is possible, the other disadvantages with charcoal beds remain relatively unchanged. A
cryogenic temperature adsorption system was used extensively at ICPP at one time, but proved to be
unreliable, and refrigeration costs were prohibitive. The use of dual adsorptive beds of relatively small
capacity in conjunction with a small cryogenic distillation system to achieve a high-purity product has been
suggested.
6. Perm-Selective Membranes.
Differences in permeability of gases through a special membrane is the bases of a perm-selective membrane
separation process. Using this difference at high pressures (150 psi) with a silicon rubber membrane, krypton
and xenon gases can be separated from the bulk off-gases, as shown in tests at ORNL (ORNL-4572,1970 and
ORNL-4522,1971). Large membrane surface areas and high pressures are required, and this results directly in
high capital and power costs with a high mechanical ratio. Long-term radiation and chemical effects on the
membrane are not known, and the membranes are particularly sensitive to ozone and NOX damage. The
system also can be damaged by thermal and pressure shock. The concentration factors are in the range of 10-
500 ppm at the reported decontamination factor of 104. The system appears economical for low
decontamination factors only.
-291 -
-------
NOBLE GAS CONTAINMENT
High-pressure storage of the recovered noble gases in steel cylinders presently appears to be the leading
means for practical ultimate disposal. However, the inclusion or encapsulation of krypton-85 into glasses,
resins, clathrates, molecular sieves, and metal offer ways of reducing the effective vapor pressure in a cylinder
and may provide primary containment for the radioactive gases. Long-term disposal means that have been
considered include placing the cylinders into an engineered storage facility, dumping cylinders into the sea,
and directly discharging krypton-85 into an appropriate geologic formation. Commercial uses (Nuc. Ind., 1971
and Kir, 1972) will not utilize significant quantities of the recovered krypton-85. However, a demanding
market for the recovered xenon is envisioned. A comparison of the various containment techniques is given in
Table 2.
1. Low-Pressure Storage.
Storage of recovered krypton-85 within steel cylinders at ambient pressure and temperature conditions has
been studied. The storage volume required has been studied. The storage volume required has been predicted
to be similar to the volume of medium-level liquid wastes presently stored at nuclear facilities. With such
storage conditions, rupture of a single tank is considered to be controllable and very improbable. Ozone
removal is considered necessary; the stored gas must also be free of either oxygen, nitrogen, or both to prevent
formation of corrosive products through radiolysis. Although storage at ambient conditions appears possible,
considerable economic evaluation appears necessary before considering use at reprocessing facilities.
2. High-Pressure Storage.
High-pressure storage (Blomeke, et al., 1969) appears to be practical and economical, and has extensive
technical development obtained during many decades of use in a wide variety of systems. In addition, some
experience with high-pressure storage and shipping has been obtained at ICPP during shipment of their
product to ORNL. A well-engineered facility could provide a double containment design and minimize such
things as leak probability, actual krypton-85 release if a leak occurs, and corrosive radiolytic products. The
double-containment concept is deemed necessary because an instantaneous release of one high-pressure
storage cylinder may produce unacceptable radiation doses. Further study into overall costs of the complex
facility and the long-term effect of possible corrosion of storage cylinder walls by radiolytic products is
required. High-pressure storage may require high-purity noble gases, to prevent formation of corrosive,
oxidizing products through radiolysis.
3. Storage Through Adsorption or Encapsulation.
Inclusion or encapsulation (Slansky, 1969; Clark, et al., 1970; Brock, 1964; Garden, 1969; Iso. Rad. Tech.,
1967; and Reist, 1967) of the krypton-85 into molecular sieves, clathrates, resins, glasses, and metals may offer
a means to provide primary containment within a steel cylinder, and also reduce the effective vapor pressure
within the cylinders. Cost, safety, and environmental adequacy of any of these inclusion methods are
dependent upon storage capacity, temperature restrictions, and radiation damage of the materials. Present
information is available only on a lab-scale basis. Storage capabilities of molecular sieves are fairly well-
known, but are limited at ambient or elevated temperatures. Use of molecular sieves at low or cryogenic
temperatures is not desirable since any loss-of-cooling accident would cause high storage-cylinder pressures
which would increase the probability of massive, uncontrollable leaks. All of the encapsulation materials
require extensive investigation to determine their storage capacities, and the long-term effects of radiation,
possible radiolytic corrosion products, and elevated temperatures. Possible engineering difficulties in
constructing a truly effective (and efficient inclusion facility need considerable study. The proposed
encapsulation methods are technically difficult, require very high pressures to achieve significant storage
capacity, and require complicated and expensive equipment. At best, encapsulation is presently a laboratory
possibility and requires lengthy investigation.
4. Long-Term Disposal/Containment.
Eventually, all recovered cylinders containing krypton-85 must be transported to an AEC facility for long-
term disposal or containment. Long-term disposal methods may include placement of the cylinders into an
engineered storage facility, dumping of cylinders into the sea, and even direct discharge of the krypton-85 into
deep or shallow wells leading into a geologic formation (Reist, 1967; Tadmor, et al., 1967; Jacobs, 1967; Mudra,
et al., 1965; and Schmaltz, 1969). An engineered facility could be specifically designed to monitor and protect
the storage cylinders from the environment (storm, earthquake, etc.), and also provide secondary containment
or recovery of gases should initial releases occur. Dumping of storage cylinders into a deep-sea trench might
be technically feasible and acceptable, but public acceptance would be difficult to predict. Disposal into deep
geological formations (Reist, 1967 and Tadmor, et al., 1967) is in the same category. Very high pressures are
required which create a high leak potential and the required detailed knowledge of the deep geologic formation
is difficult and expensive to obtain. In the public's eye, there is alwasy the fear of the "undiscovered crack"
-292-
-------
which could release the radioactive gases into one's backyard. Some investigation and testing has been done
on shallow-well disposal of krypton-85 (Jacobs, 1967; Mudra, et al., 1965; and Schmaltz, 1969). Such a disposal
means does not involve high pressures and does not depend on an impermeable geologic formation, but
assumes the soil will adsorb the noble gases and release them slowly over a period of decades. Such a method
would require large volumes of loosely packed rock and relatively large ground surface areas. However, the
concept suffers from the disadvantage of a constant, highly visible-to-the-public, release of krypton-85 to the
atmosphere, however small. Well-engineered facilities for high-pressure storage of the radioactive gases still
appear to have an important technological and experience advantage over other methods.
REFERENCES
Bendixsen, C. L. and F. O. German, (1971), Operation of the ICPP Rare Gas Recovery Facility During
Fiscal Year 1970, ICP-1001.
Bendixsen, C. L., F. O. German and R. R. Hammer, (1973), 1972 Operation of the ICPP Rare Gas
Recovery Facility, ICP-1023.
Bendixsen, C. L. and G. F. Offutt, (1968), Rare Gas Recovery Facility at theldaho Chemical Processing
Plant, IN-1221.
Blomeke, J. O. and J. J. Perona, (1969), Management of Noble-Gas Fission Product Wastes from
Reprocessed Spent Fuelds, ORNL-TM-2677.
Breck, D. W., (1964), Crystalline Molecular Sieves, Journal of Chemical Education, Vol. 41, No. 12, pp. 678-
689.
Carden, J. E., (1969), Radio-Release in Review with Emphasis on 85Kr Clathrates and Kryptonates,
ORNLIIC-18.
Chemical Technology Division Annual Progress Report for Period Ending May 31, 1970, (1970), ORNL-
4572, p. 106-107.
Clark, W. E. and R. E. Blanco, (1970), Encapsulation of Noble Fission Product Gases in Solid Media
Prior to Transportation and Storage, ORNL-4473.
Davis, J. S. and J. R. Martin, (1973), A Cryogenic Approach to Fuel Reprocessing Gaseous Rod Waste
Treatment, presented at the ANS National meeting in Chicago, Illinois, during June 1973, linde Division of
Union Carbide Corporation.
Dunster, H. J. and B. F. Warner, (1970), The Disposal of Noble Gas Fission Products from the
Reprocessing of Nuclear Fuel, AHSB(RP) R-101, Harwell, U.K.
Final Report on Development of Dissolver Off-Gas Unit, (1958), AIRCO-C-214-6).
Glass, R. W., et al., (1972a), HTGR Head-End Processing: A Preliminary Evaluation of Processes for
Decontaminating Burner Off-Gas, ORNL-TM-3527.
Glass, R. W., et al., (1972b), Removal of Krypton from the HTGR Fuel Reprocessing Off-Gases, ANS
Transactions, Vol. 15, No. 1, p. 95.
Jacobs, D. G., (1967), Behavior of Radioactive Gases Discharged into the Ground, Nuclear Safety, 2, pp.
175-178.
Kr and Xe as Reactor By-Products, (1972), Nuclear Industry, pp. 45-46.
Keilholtz, G. W., (1971), Krypton-Xenon Removal Systems, Nuclear Safety, Vol. 12, No. 6, pp. 591-599.
Kirk, W. P., (1972), Krypton-85, A Review of the Literature and an Analysis of Radiation Hazards, PB-207-
079, Environmental Protection Agency.
Kovach, J. L., (1970), Review of Radioactive Noble Gas Collection Processes, North American Carbon,
Inc., NACAR010007.
Krypton-85 Clathrates, (1967-68), Isotopes and Radiation Technology, Vol. 5, No. 2, pp. 108-111.
Merriman, J. R., M. J. Stephenson and D. I. Dunthorn, (1972), Recent Developments in Controlling
Release of Noble Gases by Absorption in Fluorocarbons, ANS Transactions, Vol. 15, No. 1, p. 95.
Mudra, P. J. and B. L. Schmaltz, (1965), An Appraisal of Gaseous Waste Disposal into the Lithosphere
at the NRTS, Id., IDO 12024.
Nichols, J. P. and F. T. Binford, (1972), Status of Noble Gas Removal and Disposal, ORNL-TM-3515.
Noble Liquid Forms a Radiation Detector, (1972), Industrial Research, p. 29.
Rainey, R. H. and S. Blunkin, (1972), Completion Report — Evaluation of the Use of Perm-Selective
Membranes in the Nuclear Industry for Removing Radioactive Xenon and Krypton From Various Off-Gas
Streams, ORNL-4522.
Reist, P. C., (1967), The Disposal of Waste Radioactive Gases in Porous Underground Media, Nuclear
Applications, 3, pp. 474-80.
Schmaltz, B. L., (1969), Injection of Gas Into the Lithosphere at the NRTS, Id., IDO 12069.
Schwieger, R. C., (1973), Controlling Radioactive Discharges from Water-Cooled Reactors, Power, pp. 21-
28.
Slansky, C. M., (1969), Separation Processes for Noble Gas Fission Products From the Off-Gas of Fuel
Reprocessing Plants, presented at the Consultant's Meeting on Safety Analysis of Radioactive Noble Gas
Releases from Reprocessing Plants, organized by the International Atomic Energy Agency at Centre d'etudes
nucleaires de Saclay, France.
Slansky, C. M., H. K. Peterson and V. G. Johnson, (1969), Nuclear Power Growth Spurs Interest in
Fuel Plant Wastes, Environmental Science and Technology, Vol. 3, pp. 446-451.
-293-
-------
Stephenson, M. J., J. R. Merriman and D. I. Dunthorn, (1970), Experimental Investigation of the
Removal of Krypton and Xenon From Contaminated Gas Streams by Selective Absorption in tluorocarbon
Solvents: Phase I Completion Reports, K-1780.
Tadmor, J. and K. E. Cowser, (1967), Underground Disposal of 8sKr From Nuclear Fuel Reprocessing
Plants, Nuclear Eng. and Design, 6, pp. 243-250.
Wheatley, M. F., (1973), Calculations on the Performance of the KALC Process, ORNL-4859.
TABLE 1. Comparison of Krypton-85 Separation Processes for Nuclear Facilities.
Process Development Status Advantages Disadvantages
Cryogenic Distillation
Fluorocarbon
Adsorption
Absorption in Liquid
Carbon Dioxide
Ambient Temperature
Charcoal Adsorption
Low-Temperature
Charcoal Adsorption
Perm-Selective
Membranes
Developed and
operated on a
significant scale;
commercial units for
nuclear reactors have
been sold.
Developed and tested
in pilot plant; no
demonstration or
operating experience
at nuclear facility.
Pilot-plant scoping
tests performed and
continuing; feasibility
studies completed.
Laboratory tests
completed; presently
used in delayed
release facilities.
Developed and
operated on a
significant scale.
Bench scale tests; no
pilot-plant tests.
Low capital costs well
defined; product
purity relatively high;
high technical
background gives
high reliability.
Low refrigeration
costs; low solvent
costs; very low
explosions hazards;
reduced process gas
pre-treatment.
Zero solvent costs;
low refrigeration
costs; no explosion
hazards identified.
Simple operation;
adequate technical
background to assure
reliability.
Smaller volume beds.
Room temperature
operation.
Ozone explosions
potential must be
controlled, high-
efficiency gas cleanup
required.
High-pressure (300-
400 psig) with high
leak potential; low
product purity with
corrosion hazards.
High-pressure with
high leak potential;
low product purity;
restricted to HTGR
nuclear fuels.
Large-volume beds;
fire and explosion
hazard.
Fire hazard and
explosion; high
refrigeration costs;
reliability unproven.
High capital costs;
high pressures; high
power costs;
membranes sensitive
to oxidizing chemicals.
-294-
-------
Technique
TABLE 2. Comparison of Krypton-85 Containment Techniques.
Development Status Advantages Disadvantages
Low-Pressure Tanks
High-Pressure
Cylinders
Adsorption on
Charcoal
Encapsulation
Engineered Storage
Facility
Deep Sea Disposal
Deep Geologic
Formation
Shallow Geologic
Formation
Feasibility studies
performed; no field
tests.
Used for shipment at
ICPP; no long-term
tests.
Development data
completed; short-term
operation.
Laboratory studies
only partly completed
containers; provides
primary containment.
Cost and feasibility
studies continuing; no
field experience.
Feasibility studies
limited; field
experience on related
chemicals.
Feasibility studies
performed.
Field tests performed.
Low pressures with
low leak probability.
Low storage volumes;
long technical
background.
Reduces vapor pres-
sures of containers.
Reduces vapor
pressures of
containers.
Protection from
environment, earth-
quakes, and gas leaks,
secondary containment
and recovery of
leaked gases.
Long storage time;
isolated from public;
simple concept.
Long-term storage.
Unaffected by
earthquake; long-term
reliability; low cost.
Very large storage
volume; ozone
removal required;
radiolytic product
corrosion unknown.
Long-term corrosion
unknown; high
pressures increase
probability of massive
release; secondary
containment required.
Large storage volume;
fire and explosion
hazard.
Effects of radiation,
temperature, and
corrosion need
extensive study;
process technically
difficult.
High capital cost
perhaps high
operating cost.
Public opinion;
reliability not 100%
assured.
Public opinion; very
high pressures;
complicated site
investigation required.
Public opinion.
-295-
-------
A CRYOGENIC SYSTEM FOR COLLECTING NOBLE GASES FROM BOILING WATER
REACTOR OFF-GAS
G.E.Schmauch
Air Products and Chemicals, Inc.
Allentown, Pennsylvania 18105
Abstract
Industrial gas processing and cryogenic air separation technology provides an effective andreliable means
for control and limitation of the environmental release of radioactive noble gases from nuclear power reactors.
In boiling water reactors, noncondensible gases are expelled from the main condenser. This off-gas stream is
composed largely ofradiolytic hydrogen and oxygen, air in-leakage, and traces of fission product krypton and
xenon.
In the Air Products'treatment system, the stoichiometric hydrogen and oxygen are reacted to form water in
a catalytic recombiner. The design of the catalytic recombiner is an extension of industrial gas technology
developed for purification of argon and helium. The off-gas after the recombiner is processed by cryogenic air-
separation technology. The gas is compressed, passed into a reversing heat exchanger where water vapor and
carbon dioxide are frozen out, further cooled, and expanded into a distillation column where refrigeration is
provided by addition of liquid nitrogen.
More than 99.99% of the krypton and essentially 100% of the xenon entering the column are accumulated in
the column bottoms. Every three to six months, the noble-gas concentrate accumulated in the column bottom
is removed as liquid, vaporized, diluted with steam, mixed with hydrogen in slight excess of oxygen content,
and fed to a small recombiner where all the oxygen reacts to form water. The resulting gas stream, containing
from 20 to 40% noble gases, is compressed into small storage cylinders for indefinite retention or for decay of
all fission gases except krypton-85, followed by subsequent release under controlled conditions and favorable
meteorology.
This treatment system is based on proven technology that is practiced throughout the industrial gas
industry. Only the presence of radioactive materials in the process stream and the application in a nuclear
power plant environment are new. Adaptations to meet these new conditions can be made without sacrificing
performance, reliability, or safety.
INTRODUCTION
In recent months, we have heard a lot of discussion about treating gaseous radioactive waste from nuclear
power plants. One of the major problems for light water reactor nuclear power stations is the environmental
release of radioactive gases. In the design and construction of nuclear power stations, the utilities and their
architects and engineers are confronted with the goal of limiting releases of radioactive gases to an
"acceptable level" for licensing. When selecting a process and equipment to achieve this objective, special
emphasis is placed upon the following factors:
(1) System Performance. The system must perform the required duty effectively and efficiently.
(2) Reliability. The system must perform continuously for extended periods of time with little or no
maintenance or operator attention.
(3) Safety. The system must not present any hazard with respect to the plant personnel nor to the plant
itself.
(4) Cost. The system must meet these criteria with a minimum of capital and operating costs.
These criteria are applicable to all light water reactors, but this paper considers the problem with respect to
boiling water reactors (BWR's).
A typical off-gas stream composition is shown in Table 1 for a nominal 1000 MWe BWR.
The non-condensable gas from the main condenser consists of a mixture of approximately 150 scfm of
hydrogen and 75 scfm of oxygen produced by the radiolysis of water within the reactor. The off-gas includes
from 30 to 60 scfm of ambient air resulting from in-leakage at the condenser. This mixture contains trace
quantities of the fission products krypton and xenon from fuel element leakage, and is generally saturated
with water vapor. This stream is highly radioactive as a result of the gaseous fission products and activation
products formed in the reactor. The problem is to reduce the radioactivity to a level which is compatible with
the plant environment and ensures licensability.
A variety of systems have been offered to solve this problem. These include short delay pipes to allow some
decay prior to release, long delay pipes, and pressurized delay tanks to increase the period of delay, and
extended delay by charcoal beds prior to release. All of these techniques provide some reduction in
radioactivity prior to release, but their overall effectiveness is limited either by size or cost.
Air Products and Chemicals, Inc. (APCI) proposes a more effective and versatile solution to the problem. A
solution which would provide activity reduction factors in the range of 10,000 as compared to the more
conventional factors in the range of 50 to 500 with respect to the 30-minute decay reference.
This solution, based upon the application of industrial gas and cryogenic processing technology, is widely
practiced in industry, and has proven to be highly reliable, safe, and economically competitive in the gas
processing market.
-296-
-------
APCI'S GASEOUS RADWASTE TREATMENT SYSTEM
The APCI's Gaseous Radwaste Treatment System for B WR off-gas consists of two major sections. The first
is a recombiner section for the removal of the radiolytic hydrogen and oxygen. This section provides a
significant off-gas volume reduction by a factor of 3 to 6, and also removes the hydrogen explosion hazard.
The second is a cryogenic section for the removal and collection of fission product gases from the residual off-
gas stream after the recombiner, with subsequent release of an essentially radiation free effluent to the
atmosphere. The removal and collection of the fission product krypton and xenon is accomplished by
cryogenic distillation, since the residual off-gas after the recombiner, is air contaminated by the fission
products.
Examination of the normal boiling points of the components of this residual off-gas stream as shown in
Table 2 illustrates that cryogenic distillation is an attractive technique for further volume reduction and
collection of the fission product gases.
The major components nitrogen, argon, and oxygen are all significantly more volatile than the krypton
and xenon. This approach utilizes standard air separation technology for the concentration of the gaseous
fission products, but certain other contaminants must also be considered. Water vapor and carbon dioxide
must be removed to prevent freeze-up. Methane and ozone must be removed and controlled to eliminate any
safety problem. These factors are discussed further in the description of the process.
The overall system is described with the aid of the flowsheet shown in Figure 1. The off-gas from the nuclear
power plant is expelled from the main condenser steam jet air rejectors, and is mixed with the diluent steam to
produce a mixture which has a hydrogen concentration of 4% by volume on a wet basis, and is at 280°F with
about 60°F superheat. This stream enters the catalytic recombiner which uses a pelletized catalyst (noble
metal on ceramic base) called NIXOX(TM) where 99.99% or better of the hydrogen is reacted with oxygen to
form water. The water is subsequently removed in the cooler condenser. This radioactive water is returned to
the plant water system.
The remaining uncondensed stream consisting of air in-leakage, trace fission gases, and about 200 ppmV
residual hydrogen is compressed from 14.7 psia to 115 psia in a water-sealed compressor. The compressor
discharge is cooled in an aftercooler which condenses additional water, and subsequently is passed through a
short delay line to permit decay of most of the krypton isotopes greater than atomic weight 138, and most of
the nitrogen-17, nitrogen-16, and oxygen-19. The compressed air stream enters the cryogenic section through
a series of timed switch valves, which control the flow to a reversing heat exchanger. At approximately 10
minute intervals, the flow passages for the warming (low-pressure) side, and cooling (high-pressure) side, are
interchanged. The water and carbon dioxide, which is frozen out of the high-pressure stream, is evaporated
from the exchanger surface into the outgoing low-pressure off gas stream, and, hence, to the atmosphere.
Since the high-pressure passages will be filled with compressed air containing radioactive noble gases, the
passages are first purged back to the compressor suction before venting the passages to the atmosphere. By
careful selection of design flows, pressures, and temperatures, this scheme, which is common to most modern
tonnage air plants, provides quantitative removal of water and carbon dioxide without allowing noble gases
to escape to the atmosphere. The process stream, which is cooled in the freeze-out heat exchanger to about -
275°F, is further cooled and expanded into the distillation column. The boil-up from the column reboiler is
provided by using warm nitrogen and radiolytic heat. Reflux in the column can be provided by either the direct
or indirect addition of liquid nitrogen. More than 99.99% of the krypton, and essentially 100% of the xenon,
entering the column is accumulated in the column bottoms.
Light hydrocarbons, primarily methane, are present in the air in concentrations comparable to the noble
gases. These hydrocarbons accumulate in the column bottoms at a rate comparable to the noble gases; and
after prolonged operation, there could be sufficient quantities to be hazardous in the presence of liquid oxygen.
However, the buildup of hydrocarbons is readily prevented by the continuous withdrawal of a small stream
from the column bottom liquid, passage through a hydrocarbon oxidation chamber, and recycle to the suction
of the feed compressor. This continuous hydrocarbon removal system has the additional capability of thermal
decomposition of any ozone which might be formed by irradiation of the liquid oxygen.
As noble gases are accumulated in the column bottom, the krypton-xenon enriched liquid oxygen is
periodically removed and further concentrated on a batch basis by removing the oxygen. At 1% fuel-leakage
and 40 scfm air in leakage, periodically means about 4 times a year. This infrequent batch operation for
secondary concentration of the fission gases is accomplished in about 24 hours. The liquid withdrawal from
the column bottom for hydrocarbon and ozone control is shunted to the secondary concentrator at about 2
scfm. This small flow is diluted with steam, mixed with hydrogen in slight excess of oxygen content, and
passed through a small recombiner to effect oxygen removal. Hydrogen requirements range from 3 to 4 scfm
over a 24-hour period for each batch transfer. This is insignificant compared to the continuous hydrogen
addition concept. The diluent steam is removed in a condenser, and the residual gas is compressed to 500 psia
and stored in cylinders. The composition of the stored gas is approximately 20% krypton and xenon, 5%
hydrogen, 37% nitrogen, and 38% argon. The stored gas can be held indefinitely or held until all fission gases
except 85Kr are decayed, and subsequently released under controlled conditions and favorable meteorology.
One major advantage of this system is its versatility. For example, the residual gas from the secondary
concentrator can be further concentrated in a small cryogenic adsorber to yield a product which is 70 to Pn(K-
noble gases if storage volume is at a premium.
-297-
-------
A major criticism of the cryogenic collection system is the accumulation of large quantities of radioactive
gases. In our opinion, both the quantity of collected gases, and the radiation levels associated with the
collected gas, are reasonable and manageable. For example, at 40 scfm air in-leakage and a 100 mCi/s activity
release level referenced to 30 minutes delay, the total quantity of krypton and xenon collected per year is about
25 standard cubic feet. At 20% purity, the total volume of stored gas is about 125 standard cubic feet per year. If
stored at 500 psia, about 3 or 4 standard gas cylinders per year are required. •,..,!•
The data in Table 3 are presented to illustrate the accumulation of radioactivity in the distillation column
sump during operation of the system and the subsequent decay after transfer to storage. These data are based
upon a design level feed gas activity of 100 mCi/s after 30 minutes, and show the maximum buildup of
radioactivity in the sump during operation. Time intervals of 7,14, 30, 90, and 180 days are used to illustrate
the buildup. This table also shows the activity accumulation in the sump after continuous operation at the
design level for these time periods. The data are shown graphically in Figure 2.
The inventory builds up rapidly during the first seven days, but soon levels off to a near steady-state level
after 30 days. At this point, the accumulation rate is almost completely off-set by the decay rate except for
the85Kr. The maximum accumulation in the sump for the design conditions reachs a steady-state level slightly
in excess of 8,000 curies. This level increases by about 50 curies per month due to the continuous accumulation
of85Kr.
Table 3 also contains data to illustrate the decay of the accumulated radioactive material after shutdown or
transfer to storage. Data are presented for operation during periods of 90 days and 180 days prior to transfer.
These data are also shown graphically in Figure 2. After shutdown or transfer, the accumulation rate becomes
zero, and decay occurs rapidly. Within 7 days, the inventory has decayed to about 1/3 the original inventory.
Within 90 days, the only undecayed activity is due to 86Kr. After operation for 90 days, the 85Kr residual is 152
curies,; while after 180 days, the 85Kr residual is about 302 curies.
These data show that the annual accumulation of 85Kr for the design conditions is about 600 curies. Since
85Kr is the only gaseous fission product that is accumulated over the long term, the increase in inventory over
and above the steady-state level of 8,000 curies is about 600 curies per year. Therefore, the inventory difference
between a delay system and a collection system is only 600 curies per year. Both types of system will contain
up to 8,000 curies as inventory during normal operation.
The system provides the added benefit that it can be operated as a delay system or as a collection and
storage system. The collected radioactive gases can be stored indefinitely with no significant long-term
accumulation of activity (600 curies per year of 85Kr), or it can be released at a controlled rate when
meteorological conditions are appropriate. Even when operated as a delay system, this APCI system allows
the operator to select the period of delay. For example, when release rates are below design levels, the delay
period can be short, but when release rates approach the design level, added delay may be desirable. Most
competitive systems do not provide this flexibility. This versatility also provides insurance for the future
when more stringent restrictions on release levels may be imposed.
SUMMARY
The APCI Gaseous Radwaste Treatment System does not incur the accumulation of higher levels of
radioactive gases when operated as a delay system then do competitive systems. When operated without
release of 85Kr, this APCI system accumulates, at the design condition, a maximum of 600 curies per year of
85Kr — over and above the steady-state inventory of 8,000 curies required of all delay systems.
The performance of this cryogenic system can be summarized as follows:
(1) Activity Reduction Factor of 10,000. This is based upon a plant release rate of 100 mCi/s after thirty
minutes delay to yield a total release of 10/zCi/s.
(2) Automatic and Safe Operation.
(3) High reliability is achieved by using some redundancy on mechanical equipment.
(4) Low Operating Costs. This is possible because liquid nitrogen costs are less than $1,000 per month, and
compressor power requirements are about 40 kW.
This APCI system design is based upon many years of industrial operating experience in such fields as
hydrogen-oxygen recombination units for high-purity argon and helium production, on-site industrial gas
generation plants, portable oxygen generators for military use, and shipboard oxygen generators for the U. S.
Navy. The design is supported by research and development programs which provided phase equilibria data
for the distillation column design, krypton-xenon solubility data, catalyst performance data, and catalyst
poisoning data. The design is also supported by laboratory demonstrations of recombiner performance using
diluent steam, and hydrocarbon and ozone control systems.
In conclusion, this off-gas treatment system is based upon proven technology that is practiced throughout
the industrial gas industry. Only the presence of radioactive materials in the process stream and the
application in a nuclear power plant environment are new. Adaptations to meet these new conditions can be
made without sacrificing performance, reliability, or safety.
-298-
-------
TABLE 1. Typical Off-Gas Stream (1,000 MWe BWR).
150 scfm Hydrogen -Water Radiolysis
75 scfm Oxygen -Water Radiolysis
30 to 60 scfm Air -Condenser Leakage
Fission Product Krypton and Xenon -Fuel Element Leakage
Water Vapor -Reactor Steam
TABLE 2. Physical Properties.
Component Normal Boiling Point, °F
Hydrogen -423
Neon -411
Nitrogen -320
Argon -302
Oxygen -297
Methane -258
Krypton -244
Ozone -169
Xenon -163
Carbon Dioxide -109*
Water 212
*Sublimation Point
TABLE 3. Sump Inventory Accumulation During Operation.
Time of Operation, Days Activity in Sump, Curies
0 0
7 5,410
14 6,871
30 7,815
90 8,071
180 8,223
Inventory Decay After Shutdown or Transfer to Storage
(After 90 Days Operation)
Time After Transfer, Days Activity in Storage, Curies
0 8,071
7 2,672
14 1,224
30 307
90 152
180 150
(After 180 Days Operation)
Time After Transfer, Days Activity in Storage, Curies
0 8,223
7 2,825
14 2,376
30 458
90 302
180 297
-299-
-------
DILUENT
STEAM
VENT
GAS
CO
O
o
CRYOGENIC OFF-GAS
TREATMENT SYSTEM
FOR
BOILING WATER REACTORS
HEAT
EXCHANGER
COLD
BOX
RECOMBINER
RADIATION
FREE
VENT
TCONDENSER
SEPARATOR
C.W. RADIOACTIVE
WATER
RADIOACTIVE
WATER
SURGE
TANK
—1
I
PRIMARY
CONCENTRATOR
SECONDARY
CONCENTRATOR
PURIFIER
BOX
BOTTLING
STATION
Line
Storage
Figure 1.
-------
ACTIVITY IN COLUMN SUMP
VS.
TIME OF OPERATION / DECAY
CURIES IN
COLUMN SUMP
QAAA
8000
"7 A A A
RAAA
5000
4000
•»nnn
2OOO
in no
QA A
ft A A
f\j U
bUU
5OO
400
-zf\f\
?(\f\
inn
/* 30 DAYS
/ 14 DAYS
| 7 DAYS
-j-(OPERATI
BOTTLE
AFTER
OPERA
ON)
INVE
DEC
SHU
OR-
S
7D
X "V
DA
SK
Or
X M
1
I
_1
loU DAYS
1 S
1NTORY "
AY AFTER
TDOWN
CRANSFER
L^ "^
^^
AYS
YS AFTER
IUTDOWN
) TRANSFER
DAYS
X 30 DAYS
INVENTORY
90 DAYS '
TION
\
>\.'
^*
1 °
1
X 7 DAYS
X 14 DAY!
1
X30D
x!
10 DAYS
N
JMP INVENT
CCUMULATK
URING OPE
nnTT
AYS AFTE
^^OPER
"ORY
JN •
RATION
L€ INVENT(
R 180 DAYS
AT ION J
. 90 DAYS 18
^^m»»y |N
180 DAYS
"\DV
ODAYS
INCREMENTAL
ACTIVITY
DUE TO 90
ADDITIONAL
DAYS
ACCUMULATION
OF KRYPTON-85.
60 120 180 240 300 360
TOTAL TIME (DAYS) OF ACCUMULATION AND DECAY
Figure 2.
420
-301 -
-------
A CRYOGENIC APPROACH TO FUEL REPROCESSING GASEOUS RADIOACTIVE WASTE
TREATMENT
J. S. Davis and J. R. Martin
Union Carbide Corporation
Linde Division
Tonawanda, New York
Abstract
The coupling of industrial gas processing and cryogenic technology yields an approach for the efficient and
reliable recovery of the radioactive fission product noble gases (especially S5Kr) from the off-gas of L WE and
LMFBR fuel reprocessing plants. The fundamental processing step that is employed to separate the noble
gases is the cryogenic or low-temperature liquefaction and distillation of the off-gas. The residual gas after the
removal of the noble gases is rejected to the atmosphere through the plant stack as a decontaminated waste
stream. System krypton recovery factors of greater than 99.9% can be achieved while minimizing the required
krypton storage. All elements of the process design utilize proven technology which has been commercially
demonstrated.
INTRODUCTION
Today, the United States and the world stands on the verge of an era of increasing industrial growth and
energy consumption. To meet these new demands, it is expected that a phenomenal growth of the utilization of
nuclear power will occur in the next decades. Coupled with this growth will be the continued expansion of
ancillary services such as fuel reprocessing facilities.
The purpose of this paper is to describe a cryogenic approach for the efficient and reliable recovery of the
radioactive fission product noble gases (especially 85Kr) from the off-gas of LWR and LMFBR fuel
reprocessing plants. The particular off-gas treatment system discussed in this paper is adapted for fuel
reprocessing facilities utilizing a nitric acid dissolution process. This system can be easily designed to handle
similar applications such as the hydrofluoric acid dissolution of zirconium fuels by minor modifications to the
preliminary contaminant removal steps.
Current experience in the utilization of a cryogenic system for this particular application area has until now
been limited to operations at the Idaho Chemical Processing Plant (Bendixsen, etal., 1969 and 1971). A
cryogenic system has been in use at this facility for the primary purpose of removing a portion of the krypton
and xenon for isotopic utilization. The system was not designed to operate continuously or function as a high-
efficiency recovery unit. Modernization is currently being contemplated for this unit to assure continuous,
reliable, and efficient operation.
This paper is intended to clarify misunderstandings which apparently exist on the status of krypton
removal systems.
OFF-GAS CONSTITUENTS
Of primary importance to a discussion of gaseous radioactive waste systems is the identification of all
constituents in the influent off-gas stream. These off-gases contain significant quantities of the oxides of
nitrogen formed from the nitric acid dissolution process. As illustrated in Table 1, the oxides of nitrogen,
fission product gases, and an air stream comprise the off-gas from the reprocessing plant. Air flow rates (10-75
scfm) and compositions were examined in the design for typical fuel reprocessing plants as represented by the
Midwest Recovery Facility (USAEC, 1972 and Martin, 1971).
The fundamental processing step that is employed to separate the noble gases is the cryogenic liquefaction
and distillation of the off-gas. The manner in which any off-gas constituent hampers or in any way reduces
this objective must be recognized and handled in the system design.
DESIGN CRITERIA
The fundamental criteria for the design of the off-gas treatment facility is the safe, reliable, and efficient
recovery of the fission product noble gases. Design considerations must include a fail-safe approach to the
handling of interim hazards, a minimization of operator required interaction, and a feasible system for long-
term krypton storage. With due regard for a safe and reliable operational system, the design presented utilizes
the concept of complete oxygen removal from the off-gas prior to cryogenic processing. Thus, several potential
problems associated with the operation of the cryogenic equipment are eliminated; these include:
(1) Nitrogen Dioxide Formation.
a. Radiolysis of Nitrogen-Oxygen Mixtures.
b. Trace Reaction of Nitric Oxide With Either Oxygen or Ozone.
(2) Ozone Formation.
a. Cryogenic Radiolysis of Oxygen Mixtures.
(3) Processing of Hydrocarbon-Oxygen Mixtures.
-302-
-------
Of major importance in delineating the design criteria is the requirement that the potential presence of
nitrogen dioxide (N(>2) be eliminated in the low-temperature processing equipment. Nitrogen dioxide exhibits
a high freezing point temperature. In the off-gas treatment system, precautions must be included to maintain
N02, xenon, and other contaminants below their solubility limits in the various cryogenic fluid mixtures to
prevent solids formation. Solid oxides of nitrogen have been a contributing factor to destructive explosions in
industrial air separation plants. In addition to this aspect, the solid nitrogen dioxide represents a fouling
problem in all cryogenic equipment.
Nitric oxide exhibits a high reactivity with either oxygen or ozone to form nitrogen dioxide. Thus, it is
virtually impossible to prevent this conversion from occurring in the presence of these oxides. Therefore, the
removal of nitric oxide to the parts per million range is unacceptable if this trace quantity exists in
conjunction with significant quantities of oxygen.
An additional consideration in the elimination of the oxides of nitrogen in the cryogenic equipment is their
potential formation from the radiolysis of nitrogen-oxygen mixtures.
During normal operating conditions with an oxygen-rich environment, these products would be formed by
the radiolysis of the liquid oxygen-nitrogen mixtures residing on the distillation trays near the feed point. The
reaction occurs mainly in the cryogenic foam with a negligible contribution due to the gas phase formation in
the intertray space. Although the nitrogen dioxide formation rate is much less than that of ozone, it is
important due to the extremely low solubilities exhibited by the oxides of nitrogen in the various cryogenic
fluid mixtures.
The removal of oxygen also eliminates the danger inherent in cryogenic processing of a stream that
contains large quantities of hydrocarbons. Although the hydrocarbons are catalytically converted to carbon
dioxide, there is still a significant portion of the methane that is unconverted in the effluent from the
hydrocarbon conversion. This methane presents no potential explosion problems in a cryogenic system in the
absence of oxygen.
Catalytic oxidation of the hydrocarbons is employed with the design concept of complete oxygen removal as
a secondary safety design factor, and to aid in the minimization of krypton storage requirements. Methane
has the characteristic nature of krypton in terms of its volatility or separation capability by distillation. If
unremoved, the methane tends to accumulate with the krypton fraction and acts as a diluent in the krypton
product.
In the system design, the formation of ozone by the irradiation of oxygen must be considered. At cryogenic
temperatures, thermal decomposition of ozone is negligible. Though in theory the ozone could be controlled
below generally accepted safe concentrations by recycle, several incidents of apparent ozone explosions in
cryogenic equipment under a radiation field have been reported in the literature. Due to the high reactivity of
ozone, and the general lack of a complete understanding of its interactions with other molecules, it cannot be
claimed that an off-gas system with an oxygen-rich environment is completely free from any explosion
hazards.
Ozone is similar to xenon in terms of its volatility or separation capability by distillation. Where
commercial utilization of the xenon is contemplated, ozone from an oxygen-rich processing environment
would be concentrated with the xenon product and would represent a significant oxygen impurity at ambient
conditions.
A second consideration in the radioactive waste system design is the elimination of the presence of high
freezing point constituents in the cryogenic equipment. Coupled with this requirement is the necessity for
proper system design to eliminate those potential operational modes in which certain constituent mixtures
exhibit limited solubility behavior. Several components, notably water and carbon dioxide, exhibit freezing
points at substantially higher temperatures than those encountered in the cryogenic equipment. These
components are prevented from entering the cryogenic equipment to eliminate potential solids formation with
subsequent blockage of the equipment. Methods that have been extensively employed in the industrial air
separation industry are available for the pretreatment and removal of these contaminants.
In the second separation column of the off-gas treatment system, a krypton-rich product and a high-purity
xenon fraction are produced. Below 25 psia, saturated liquid krypton-xenon mixtures form a solid third phase
over a portion of the xenon concentration range. For this particular design, as illustrated in Figure 1, the
operation of this column at a pressure exceeding 50 psia eliminates any problem associated with a freezeup.
This approach is based upon the strong pressure functionality of the liquid-vapor thermodynamic equilibrium
relationships as compared with the extreme insensitivity of the mixture freezing point to this parameter.
Consideration must be given in the system design to eliminate this and other solubility problems due either to
the presence of trace contaminants or the effects of operational conditions on thermodynamic equilibrium.
The third major processing objective is the assurance that the unit design incorporates all steps necessary
to achieve krypton recovery factors of greater than 99.9% while minimizing the required krypton storage. The
design approach must provide for safe, and reliable system performance under any foreseeable modes of
operation. Inherent to this concept, the off-gas system must be designed to minimize or eliminate any
potential for the environmental release of the system's radioactive inventory during a component outage or
the loss of a support system.
The design philosophy is for the continuous, automatic, and unattended operation of the equipment under
the supervision of an operator in the main control room. Alarms are provided to both alarm and indicate the
source of the problem. Instrumentation and valving are selected for fail-safe operation. Reliability is a major
consideration in the functional specfication of equipment, and redundancy is employed where necessary to
assure continuous service. The safety of the system is enhanced by utilizing proven technology previously
demonstrated on a commercial scale.
-303-
-------
Dependent upon the fissile fuel employed, typical fission yield ratios of xenon to krypton vary between ten
and twenty on a volume basis. Therefore, due to the long-term storage requirements for a 85Kr rich product, it
was deemed necessary to separate the krypton and xenon to reduce the ultimate storage volume requirements.
All fission product xenon isotopes have half-lives so short (12 days max.) that no detectable xenon
radioactivity will exist one year after the fuel is removed from the reactor. This high-punty xenon product is
thus available for commercial utilization. _ . .
The proper application of radiation technology shares an equal role with cryogenic technology in the design
of a safe, reliable off-gas treatment system. Radiation considerations are important in the areas of material
selection and stability, dose rate calculations to determine access and maintenance problem areas, radiation
chemistry effects in terms of contaminant formation rates, and fission yield calculations to determine the off-
gas system's influent concentrations of radioactive and non-radioactive fission product gases.
PROCESS DESCRIPTION
The process incorporates complete oxygen removal from the off-gas of the fuel reprocessing facility. The
hazards associated with concentrating radioactive xenon and krypton in the presence of oxygen (e. g., ozone
and the higher oxides of nitrogen) are eliminated. At the start of the process, the feed gas and a krypton-rich
recycle stream from the product column are combined prior to entrance into a dual stage catalytic system.
Recycling of the krypton-rich stream serves the primary function of maintaining the desired krypton product
purity at a minimum of 75%. This stream is taken from the overhead of the second column to remove methane
and nitric oxide from the krypton product. Thus, residual contaminant traces from the catalytic equipment are
controlled at safe concentration levels in the cryogenic distillation system.
Following preheating, the feed gas enters the first stage of the catalytic equipment where the hydrocarbons
are oxidized (95% efficiency) to carbon dioxide and water.
The effluent from the hydrocarbon conversion unit next passes through a catalytic recombiner where the
oxygen is converted to water. Prior to entering the recombiner, a recycle stream around the catalytic unit is
added to the process stream to dilute the oxygen concentration in therecombiner inlet stream. The temperature
rise in the recombiner is a combined function of the influent preheating, oxygen concentration, and known
heats of reaction. The influent reheating is accomplished by countercurrent heat exchange of the recycle
stream with the catalytic unit hot effluent. The amount of preheat is controlled by the fraction of the catalytic
unit effluent stream bypassing the preheater. Control of the outlet temperature is the means of minimizing the
undesirable side reaction involving the methanation of small quantities of carbon dioxide to methane.
Hydrogen, in excess of the stoichometric requirements, is added to the combined process-recycle before
entrance into the catalytic recombiner. Two methods of hydrogen generation have been evaluated for
utilization with this system. These consist of hydrogen generation by either the electrolysis of water or the
thermal dissociation of ammonia. Final determination of the method for hydrogen generation is a function of
the system's operational requirements in relation to a cost evaluation. The generator operates automatically
and is designed to control hydrogen production in accordance with consumption.
In the recombiner, the oxides of nitrogen are reduced by hydrogen to yield gaseous nitrogen and water,
while the oxygen is eliminated from the process stream by the chemical reaction with hydrogen to form water.
The effluent stream contains 150-200 ppm of nitric oxide and less than 0.1 ppm of oxygen. The heat of reaction
and product water are removed in an aftercooler condenser. A small quantity of steam ejection is utilized to
provide the motive power for recycling around the catalytic unit.
The off-gas from the recombiner is essentially deoxidized air saturated with water; containing 2-4% residual
H2 with the fission products xenon and krypton. After compression, carbon dioxide and water are removed
from the off-gas prior to cryogenic processing. To minimize potential fission product losses and provide
maximum reliability, either an adsorption (prepurifier) or a caustic wash-dryer unit may be utilized. The
paramount importance of assuring the minimization of potential system losses during this processing step
due either to co-adsoprtion effects or equipment malfunction warrants a fuller discussion of the alternative
methods of carbon dioxide-water removal in a later section.
From this prepurification process, the off-gas enters the primary heat exchanger of the cold box and is
cooled to within several degrees of its saturation temperature. The gas enters the cryogenic separation
equipment consisting of two distillation columns. In the first column, the feed vapor joins the vapor from the
lower section of this column and has all traces of krypton, xenon, methane, nitric oxide, oxygen, and other
gases less volatile than argon removed from the vapor by a liquid nitrogen flow downward in the column. This
decontaminated vapor stream passes from the top of the column and is later utilized in a heat exchange
function temperature by countercurrent heat exchange with the influent feed stream in the primary heat
exchanger. Liquid nitrogen addition is utilized as a refrigeration source for both columns.
The liquid levels at the bottom of the recovery and product columns are controlled by the electrical energy
input to the heaters. The heaters also function to provide the boil-up requirments in the column. A fixed L/V is
maintained at the top of the recovery column by measuring the feed to the column and the reflux through a
"flowcup." Liquid from the bottom of the primary recovery column, containing mainly xenon and krypton, is
continuously withdrawn for processing in the product column.
The second distillation column separates a high-purity xenon product from the krypton fraction. The high-
punty xenon product is withdrawn on a scheduled basis. The vapors overhead pass from the top of the column
and are heat exchanged against the cold nitrogen gas. A fixed L/V is maintained at the top of the product
column by measuring the reflux flowing through a "flow cup" and controlling the temperature of the cold
-304-
-------
nitrogen gas entering the heat exchanger. The krypton-rich vapor, which consists of the krypton recycle and
product streams, is warmed to ambient temperature. The recycle krypton stream is exhausted into the feed gas
prior to the hydrocarbon catalytic conversion unit preheater. The krypton product stream (containing 75%
krypton) is periodically withdrawn and is available at 50 psig for further compression into shielded storage
cylinders. Separately, the krypton and xenon are compressed for ultimate storage. Due to the short-life nature
of the xenon radioactivity, disposal through commercial sale is feasible.
ALTERNATIVE METHODS OF CO2-WATER REMOVAL
The carbon dioxide-water removal procedure could represent a potentially significant area of process losses
under normal operating conditions. Of the various processes currently employed in cryogenic applications,
two approaches were evaluated as satisfactory for this application. As illustrated in Figure 3, the first
approach embodies the concept of the adsorption of these contaminants by molecular sieves. Feed gas after
compression is introduced into the prepurifier for carbon dioxide and water removal. A prepurifier back purge
is required to recycle any xenon and krypton co-adsorbed in the prepurifier during the CO2 and water
adsorption. This assures the complete absence of any loss of the noble gases with water and carbon dioxide
during the regeneration step.
The portion of the decontaminated off-gas from the cryogenic unit that is not utilized as a prepurifier purge
recycle is used to regenerate the adsorption beds in the prepurifier. The gas is heated to strip the CO2 and
water off the molecular sieve adsorbent. The off-gas is then vented to the atmosphere. To virtually eliminate
the possibility of feed gas "short-circuiting" the adsorption system to the atmosphere, the prepurifier
controller is designed to positively sequence the valves and prevent the opening of the wrong valves through
various system interlockings. In addition, major modes of operation are checked and alarmed to prevent
improper operation of the prepurifier.
The caustic wash-dryer system, the second approach, represents a major step in the absolute minimization
of fission product losses in an off-gas system. Prior to compression, the carbon dioxide in the feed gas is
removed by the aqueous, caustic-solution scrubbing in the packed tower through the process of absorption and
irreversible chemical reaction. A liquid recycle pump is utilized to maintain the proper liquid to vapor flow
ratio in the packed tower. After compression, the effluent from the caustic wash tower enters the two-bed dryer
unit. Water and residual carbon dioxide are adsorbed on the molecular sieve of the dryer bed. This application
utilizes a portion of the dryer effluent gas for regeneration of the unit. The dryer recycle function is to provide a
closed-loop circuit for regeneration of the adsorption beds while recycling any of the co-adsorbed fission gases.
At no time, during the regeneration cycle, is any of the regeneration gas vented to the atmosphere. Water
rejected from the dryer during the regeneration step is ultimately removed in the compressor's water
separator, while the residual carbon dioxide is rejected in the caustic wash system.
LONG-TERM STORAGE
Due to the long life nature of the krypton-85 isotope, it is important that the radioactive waste system design
incorporate a feasible solution for the problems associated with the handling and storage of the concentrated
krypton product. The off-gas system design is not dependent upon the adaptation of a particular method of
krypton product storage and disposal. Consideration must be given in the implementation of product storage
design to the ease of handling and transportation, long-term storage integrity, storage volume minimization,
and the elimination of potential sources of environmental releases. Table 2 illustrates typical influent noble
gas conditions for the radioactive waste system.
Due to the extensive commercial experience for high-pressure gas storage, it has been adopted for the
product storage system. The approach involves the utilization of non-vented gas cylinders, designed with a
complete sealing procedure after their initial filling. This is coupled with the separation in the cryogenic
distillation equipment of the krypton and xenon to reduce the ultimate krypton storage volume. The process is
capable of krypton product purities in excess of 75%. Typical storage requirements are presented in Table 3.
The system is designed for the automatic bottling of the krypton product. The product cylinders require only
air cooling through convective and radioactive heat transfer mechanisms. The storage pressure is reduced
significantly from the design pressure to enhance its safety characteristics — especially with respect to
accidents involving incendiaries. It is envisioned that the cylinders would be shipped in a light cask,
providing transportation protection and personnel shielding functions. The xenon product is compressed into
commercial gas cylinders and held for a short delay period prior to commercial utilization.
Commercial gas cylinder shipments number in the millions annually. It is common industrial practice to
ship toxic gases such as phosgene in similar non-venting (absence of pressure relief device) cylinders. The
bulk of commercial cylinder rejections are due to the detection of fabrication flaws or field induced defects by
non-destructive testing procedures. Therefore, through the application of proven fabrication procedures,
proper material selection and design, the utilization of high-pressure gas storage provides a viable answer to
the krypton radioactive waste storage problems.
EQUIPMENT DESCRIPTION
To enhance overall system reliability and safety, all elements of the process design utilize proven
technology which has been commercally demonstrated. Similar applications for the catalytic eq uipment exist
-305-
-------
industry-wide — especially in nitric acid tail gas treatment. The contaminant pretreatment and cryogenic
processing steps are representative of industrial gas processing technology.
The major items of equipment supplied to accomplish the operations of catalyzed chemical reaction,
compression, adsorption, and cryogenic distillation can be packaged very compactly as illustrated by the
preliminary arrangement diagram shown in Figure 5.
The equipment is provided skid-mounted with the exception of the free-standing cold box, liquid nitrogen
tank, hydrogen generation system, and a control panel to allow for unattended operation.
The cold box equipment contains the cryogenic distillation equipment and the necessary heat exchangers to
cool the feed to cryogenic operating temperatures. The liquid nitrogen from the storage tank is directly added
into the cold box to supply sufficient refrigeration to both cook the feed and provide the motive force for the
distillation step. The xenon and krypton are not only separated from the off-gas in the cold box, but are also
separated from each other by a distillation step within the cold box. This latter step thus allows the xenon and
krypton to be ultimately disposed of in a separate fashion.
The cold box pressure casing is designed to hold the equivalent of the distillation system's liquid inventory
at ambient temperature. This pressure casing ensures that no portion of the radioactive inventory is released
to the environment in the event of an internal equipment failure.
SYSTEM MODIFICATIONS
The off-gas treatment system can be designed to handle similar applications such as the hydrofluoric acid
dissolution of zirconium fuels by minor modifications to the preliminary contaminant removal steps.
Currently, a few reprocessing facilities operate at air in-leakage rates much higher than those examined in
this paper. The adaptation of a cryogenic radioactive waste system to these situations must include a
complete evaluation of the effects of possible system modications as compared to front-end reprocessing
equipment modification in terms of the overall system economics, while retaining the design criteria
guidelines.
SUMMARY OF PERFORMANCE
The following presents the performance highlights for the off-gas treatment system:
(1) Design Conditions.
Air H2 Liquid Product
Leakage(a) Addition Nitrogen(b) Storage
(scfm) (scfm) Usage (gpd) Requirements (acf)(c)
5 TFI) Reprocessing 60 28 500-800 150
Capacity
(2) Utility Requirements (5 Metric tpd Capacity).
Klectrical(d) 250-310 kW
Steam 150-250 Ib/hr
Cooling Water 90 gpm
Liquid Nitrogen Requirements 500-800 gpd
(a) All flows on a dry basis at 60°F, 14.696 psia.
(b) LNg requirements include:
_1. Column wash requirements (106+krypton D.F.),
JL Cold box heat leak.
(c) Product storage requirements at 655 psig, 200°F.
(d) Includes electrolytic H2 generation unit.
FREON ABSORPTION SYSTEM
Recent literature has indicated an increased interest in the possible utilization of the Freon Absorption
System for fuel reprocessing off-gas treatment. Therefore, a brief review highlighting a comparison of the
system with the cryogenic distillation system for this application is presented in Table 4.
REFERENCES
Bendixsen, Jfftitt, and Wheeler (1969), Rare Gas Recovery Facility at the Idaho Chemical Processing
Plant, presented at the 1969 Winter Meeting of the American Nuclear Society
^Bendixsen, et al., (1971), Cryogenic Rare Gas Recovery in Nuclear Fuel Reprocessing, Chem. Eng. p. 55-
)' Catalytic Purificati°" of Tail Gas, presented at the 64th Annual
-306-
-------
Hardison (1970), Techniques for Controlling the Oxides of Nitrogen, AFC A Journal 20(6) 377-382.
Martin (1971), Cryogenic Noble Gas Recovery System as Applied to Nuclear Fuel Recovery Plants.
Material prepared for verbal presentation at the Illinois Pollution Control Board; Reference — Midwest Fuel
Recovery Plant.
Rohrmann (1971), Fission Product Xenon and Krypton — An Opportunity for Large Scale Utilization,
Isotopes and Radiation Technology 8(3) 253-260.
USAEC (1972), Final Environmental Statement Related to the Operation of the Midwest Fuel Recovcrv
Plant. Docket No. 50-268.
Air
Oxides of Nitrogen
Nitrous Oxide
Nitric Oxide
Nitrogen Dioxide
Noble Gases
Krypton
Xenon
TABLE 1. Influent Off-Gas Constituents
Normal Boiling Point Freezing Point
-346.0°F
-308.9
-361.8
-296.4
69.9
32.0
-131.6
-263.6
11.7
-250.9
-169.1
Nitrogen
Argon
Oxygen
Methane
Carbon Dioxide
Water
-320.4°F
-302.6
-297.3
-258.7
212.0
-127.2
-241.0
70.1
-244.0
-162.6
TABLE 2a. Influent Noble Gases
(5 t/d Facility)
Total Flow Radiation Level Half-
(scfd) (MCi/s) Life
Krypton 14.5
ssKr
Xenon 144
400,000
100
10.73 yr.
5.29 days
-307-
-------
TABLE 2b. Isotopic Composition (Rohrman, 1971).
Xenon Krypton
Natural
Isotope atm%
124
126
128
129
130
131
132
134
136
0.096
0.090
1.92
26.44
4.08
21.18
26.89
10.44
8.87
Fission
Product
atm% (approx.)
0.15
8
22
29
41
Isotope
78
80
82
83
84
85
86
Natural
atm%
0.354
2.27
11.56
11.55
56.90
-
17.37
Fission
Product
atm%(approx.)
.
0.2
11
31
6
52
TABLE 3. Annual Storage Requirements.
Plant Capacity -
1 Ton/Day -
5 Ton/Day
Specialty Gas Cylinder
(24" D x 94" H)
(Design P - 1800 psig)
<2 Cylinders
(Storage? = 655 psig)
8 Cylinders
(Storage P = 655 psig)
TABLE 4. Comparison Highlights.
r the fractionation of the
from xenon, thus, the low
(2) I To provide refrigeration requirements, the utilization of a mechanical system with reciprocating
machinery is necessitated in the Freon system.
(3) The adsorption system has higher operating pressure (300-500 psia).
rtquireme^tth ^ additi°n °f peripheral solvent re«>very steps, solvent makeup remains a major utility
308-
-------
Comparison of Kr-Xe Liquid Saturation Curves with Mixture Solidus Line
190
180
170
160
150
140
130
120
SOLIDUS LINE
(3-PHASE BOUNDARY)
110
MDLE FRACTION OF XENON
_
oi
o 4
o?
Figure 1.
-309-
-------
CO
I-"
o
-------
L
co
feeo
Off
7o
Figure 3.
-------
CAUSTIC
w
M
to
-------
CO
M
CO
45'
Swo
x /3' HiQM * 34 '
Figure 5.
i.
2, Hy SUPPLY
CONTROL
-------
EFFECTS OF CONTROL TECHNOLOGY ON THE PROJECTED
KRYPTON-85 ENVIRONMENTAL INVENTORY
E.E.Oscarson
Office of Radiation Programs
U. S. Environmental Protection Agency
Washington, D.C. 20460
Abstract
Growth projections for the USA nuclear power industry may be used to make projections of future
radioactive material inventories available for potential release to the environment. Since krypton-85 is
produced in reactors, the potential routes of release to the environment are at the reactors or the fuel
reprocessing plants.
Thepresen t generation of reactors and fuel reprocessing plants do not use any systems to limit the release of
krypton-85 into the biosphere. Clean-up systems are currently being designed and could be introduced when
necessary. The cumulative inventory of krypton-85 in the environment is dependent upon the time of
introduction of these clean-up systems and their efficiencies, in addition to the total quantity produced.
Annual and cumulative inventories of krypton-85 are projected to increase dramatically over the next 50
years. If controls are assumed to be initiated at various years, a series of graphs of different cumulative
environmental inventories and the associated doses and health effects are drawn. Varying the year of
introduction and the efficiency of these systems allows one to speculate as to the optimum time for control
implementation.
INTRODUCTION
Population growth and increased per capita energy consumption are expected to continue into the
foreseeable future. The combination of these factors leads to projections of energy needs which influence
utility decisions on the construction of new electric generating capacity.
With continued depletion of fossil fuel resources, and more stringent restrictions on air pollution produced
from fossil fuel combustion, the utility industry is expected to increase their commitment to nuclear generated
electric power. This projected proliferation of nuclear power implies significantly greater quantities of reactor
produced radioactive waste products. One of the consequences of this rapidly expanding waste inventory, is
the potential for long-term environmental buildup of certain radionuclides. Under normal operating
conditions, a nuclear reactor presents very little potential for significant releases of radioactive material;
however, when spent reactor fuel is reprocessed, the release probability increases.
Up to the present time, both commercial and government fuel reprocessing facilities have not attempted to
prevent the release of noble gases, and have, in most cases, tried to release them as quickly as possible to
reduce the in-plant exposure. Of the noble gases, only krypton-85, with a 10.7-year half-life, is considered an
important long-term environmental contaminant.
Recent concern over long-term buildup of krypton-85 in the biosphere has prompted companies committed
to commercial fuel reprocessing to examine potential noble gas removal systems (NEDO 14504-2,1971). The
removal efficiencies, costs, and feasibility of various processes are currently being investigated by potential
vendors of such equipment. Data developed by these suppliers will help to determine the optimum time for the
introduction of control measures.
This paper explores nuclear power projections, the introduction of krypton removal systems, and how they
relate to the release to the atmosphere through the year 2020. Also considered are the doses and health effects
to the world's population from the uncontrolled release of krypton.
NUCLEAR POWER CAPACITY PROJECTIONS
Projections of electrical power demand have traditionally been a controversial subject. Since energy
consumption depends on population and per capita consumption, variations in one or both of these factors
can significantly affect long-term power requirements.
The Federal Power Commission (FPC) annually compiles the current U. S. installed electrical capacity,
and, considering previous years capacities, predicts future power demands. The U. S. Atomic Energy
Commission (AEC) and their contractors have often related the growth of nuclear power to these FPC
projections. One study, published in 1970 by Oak Ridge National Laboratory on the siting of fuel reprocessing
plants (ORNL-4451,1970), contains nuclear power capacity projections by FPC regions up to the year 2020.
Also included are the projected spent fuel discharges, from which krypton-85 releases can be calculated.
Another source of nuclear power projections comes from the Forecasting Branch of the AEC (WASH-1139,
1972). Their forecasts give a range of installed capacity which includes high, low, and most likely estimates.
Figure 1 compares these two particular forecasts, as well as one compiled by the Office of Radiation Programs
of the EPA (EPA-S2019-73-003D, 1973). Of these three estimates, the EPA forecast falls between the other two,
and will be used in this paper for making krypton-85 projections. Examination of Figure 1 reveals a probable
-314-
-------
increase in U. S. nuclear generation capacity of 1,000% from 1975 to 1990, and 7,000% between 1975 and the
year 2020. The nuclear capacity over the total 50 year period from 1970 to 2020 is expected to increase by a
factor of 103 (100,000%) from 2.6 to 2,700 gigawatts electric.
FUEL DISCHARGE, REPROCESSING, AND KRYPTON-85 RELEASES
The isotopic composition of fissile materials in nuclear fuels is expected to change significantly over the
next 50 years. Thermal reactors, presently designed to use low enrichment uranium fuel, will probably utilize
recycled plutonium; and breeder reactors, fueled with plutonium, will produce more fuel than they consume.
Knowledge of the relative numbers of each reactor type are important primarily for the determination of the
total quantities of actinides produced per unit of electric power generation. In the case of krypton-85, the
amount produced per megawatt day is approximately the same for both thermal and fast reactors. Therefore,
the annual inventory of krypton-85 is calculated from the total fuel discharged annually for reprocessing and
is displayed in Table 1 and Figure 2.
Since much of the fuel presently discharged from reactors has not been and will not be reprocessed in the
near future, there may be a delay in the release of krypton-85 to the environment. This may cause fluctuations
in annual discharges but should have little effect on the cumulative environmental inventory.
CUMULATIVE ENVIRONMENTAL INVENTORY OF KRYPTON-85
For purposes of calculating krypton-85 buildup in the biosphere, the projected annual releases, depicted in
Figure 2, were used in conjunction with equation (1).
C=rAie-Ati (1)
Where C = cumulative environmental inventory of krypton-85 (curies),
Aj=annual release of krypton-85 in year i (curies),
A = decay constant for krypton-85 (6.45 x 10-2 years-1),
and tj = decay time for krypton-85 introduced into the biosphere in year i (years).
The cumulative inventories calculated for the years 1970 through 2020 are listed in Table 2 and are
displayed in Figures 3 and 4.
EFFECTS OF CONTROL TECHNOLOGY
Concern over long-term environmental buildup of krypton-85 has prompted government agencies,
companies committed to operating fuel reprocessing facilities, suppliers of noble gas removal equipment, and
citizen action groups to evaluate the practicality and desirability of reducing krypton emissions. Several
vendors have indicated that they can supply systems to reduce krypton emissions by factors of 102 to 103
(NEDO14504-2,1971). For this paper, we will consider that such systems can be built and installed when they
are considered necessary.
Determining the optimum time for the introduction of control measures depends on many factors including:
costs of systems, decontamination factors, achievable growth of the reprocessing industry, health effects, and
others. In this instance, only decontamination factors and cumulative fuel inventories are used as bases for
suggesting optimum implementation times, although health effects will be considered later.
If control systems with a decontamination factor of 102 were required to be installed on all reprocessing
facilities at different years, and assuming the fuel inventories listed in Table 2, the series of curves in Figure 3
represent the probable environmental inventories. With the additional assumption that the power supplied by
nuclear fission levels off at the year 2020, it is apparent that the steady state cumulative krypton-85
inventories would approach asymptotic values. Such values for the controlled and uncontrolled cases are also
displayed in Figure 3. A similar analysis, carried out for controls with a 103 decontamination factor, produced
the results plotted in Figure 4.
If a lower constant environmental level is used as a criterion for krypton controls, and a decontamination
factor of 102 is postulated, the most opportune date for control implementation appears to be 1980. Following
the same line of reasoning, if a decontamination factor of 103 is assumed, the best year for control
implementation might be 1974 or 1975.
HEALTH EFFECTS
Although the primary purpose of this paper is to illustrate the probable effects of controls on the
atmospheric levels of krypton-85, some discussion of potential health effects is appropriate. For purposes of
evaluating potential health effects, the recommendations of the Advisory Committee on the Biological Effects
of Ionizing Radiation (BEIR Committee) of the National Academy of Sciences, National Research Council
(BEIR, 1972) are used.
Given information on the cumulative environmental inventory of krypton-85, its atmospheric
concentration may be estimated by diluting it into the total atmosphere. The total atmosphere is assumed to
contain 5.14 x 1021 grams of air (Coleman, et al, 1966) with a sea level density of 1.29 x 10-3 grams /cm3. The
calculated values for sea level concentration are presented in Table 2 and Figure 5.
-315-
-------
Conversion of atmospheric concentration to integrated population doses requires population projections
and appropriate dose conversion factors. The media-dose conversion factor used in this analysis is 1.5 x 10-8
(rem/y)/(pCi/m3air), and represents the conversion from atmospheric concentration to average whole body
dose based on external photon exposures (EPA-52019-73-003D, 1973). Conversion factors for genetic, lung,
and skin doses are also available; however, the whole body dose appears to yield the highest number of health
effects, and is used for this analysis. Population projections must be based on past information with the
additional assumption that no dramatic changes in fertility rates will be experienced over the projection
period. Beginning with the United Nations world population estimate of 3.632 billion for the year 1970
(UNSO, 1971), and using an average growth rate of 1.9= per year (UNSO, 1966), gives a projected world
population of 9.3 billion by the year 2020; i.e., over 150% increase in fifty years.The integrated population dose
in any year is, therefore, a product of the air concentration of krypton-85, the total world population, and the
dose conversion factor. The resultant population doses, in person-rem, are plotted in Figure 6 and listed in
Table 2.
The transition from population dose to health effects also requires a number of assumptions. Consistent
with the recommendations made in the BEIR report (BEIR, 1972), the health risks presented are based on the
assumption that there is a linear relationship between dose and biological effect, and that any increased risk
is in addition to that produced by background radiation (i.e., no threshold). For a given dose equivalent, the
BEIR report estimates a range for the health impact per million exposed persons. Since these risk estimates
neglect repair mechanisms, the range of mortality given by the BEIR Committee is probably an upper limit.
The cancer mortality risk (including leukemia mortality) listed as "most likely" by the Committee, for whole
body radiation, is about 200 deaths per year for 106 person-rem annual exposure. Cancer mortality is not a
measure of the total cancer risk, which the Committee states is about twice the yearly mortality (i.e., 400
health effects per 106 person-rem) (EPA 52019-73-003D, 1973). These estimates are for a population with the
characteristics of the U. S. or Northern European countries, and would not necessarily be representative of the
total world population. However, with the assumption that the rest of the world will attain the present U. S.
vital statistics by the year 2020, these risk numbers may be reasonably applied.
Computation of health effects from continuous exposure to the population doses listed in Table 2 indicates
that there will be approximately 110 total cancers caused per year, including 55 excess deaths per year, for the
year 2020 from the U. S. nuclear industry. These health effects would be for the total world population with
only about 5% occurring in the United States. By applying controls with a decontamination factor of 102, the
health effects and excess deaths are reduced to 1 and .5 per year, respectively. Consequently, installing
controls by 1980 would keep the yearly health effects from U. S. produced krypton-85 below these levels.
It must be understood that the U. S. nuclear industry will produce only a portion of the total world inventory
of krypton-85, and our unilateral implementation of controls will reduce only that contribution. The U. S.
contribution to the total world inventory is estimated to be about 50% by 2020. Since 110 total cancers per year
would result from uncontrolled U. S. discharges, the total world cancer rate would be approximately 220 per
year, if all krypton-85 from projected worldwide nuclear power production is discharged.
SUMMARY AND CONCLUSIONS
The above discussion of the effects of krypton-85 controls is intended to suggest some of the factors and
approaches which may be used to evaluate the desirability and possible timing of control implementation.
The question of whether or not krypton-85 releases should be controlled is one of wide interest concerning die
nuclear industry, government standard setting and regulatory agencies, and citizen action groups.
Consequently, final decisions as to the necessity of controls, their required efficiency, and the timing of their
installation will probably not be made solely on technical grounds, but will consider the prevailing political
and societal forces.
REFERENCES
Advisory Committee on the Biological Effects of Ionizing Radiation (BEIR Committee), (1972),
The Effects of Populations of Exposure to Low Levels of Ionizing Radiation. National Academy of Sciences —
National Research Council.
Coleman, Jr. and R. Liberace, (1966), Nuclear Power Production andEstimated Krypton-85 Levels. Rad
Health Data and Reports 7:11.
General Electric Company, (1971), Applicants Environmental Report-Suppliment 1 Midwest Fuel
Recovery Plant, NEDO 14504-2.
Oak Ridge National Laboratory, (1970), Siting of Fuel Reprocessing Plants and Waste Management
Facilities, ORNL-4451. Oak Ridge National Laboratory, Oak Ridge, Tennessee.
Office of Radiation Programs - U.S.EPA, (1973), Environmental Analysis of the Uranium Fuel
Cycle: Part III - Nuclear Fuel Reprocessing, EPA-520/9-73-003D. U.S. Environmental Protection Agency.
-316-
-------
U. S. Atomic Energy Commission, (1972), Nuclear Power 1973-2000, WASH-1139 (1972). U.S.
Government Printing Office, Washington, D.C.
United Nations Statistical Office (UNSO), (1971), Demographic Yearbook. Publishing Service United
Nations, New York.
United Nations Statistical Office (UNSO), (1966), World Population Prospects as Assessed in 1963,
Population Studies No. 41 United Nations, New York.
-317-
-------
TABLE 1
ESTIMATED U.S. NUCLEAR INDUSTRY
w
M
00
AND ANNUAL INVENTORIES OF KRYPTON-85
EPA-52019-73-003D (1973).
YEAR
1970
1975
1980
1985
1990
1995
2000
2005
2010
2015
2020
NUCLEAR ELECTRIC
GENERATION: GW(e)
2.6
40
110
220
420
650
1000
1360
1780
2220
2700
LWR-U
25
700
1900
2700
3700
4100
3700
3700
4300
5300
6100
TONNES OF
LWR-PU
0
90
500
2600
3800
4100
3800
3700
4400
5400
6100
FUEL DISCHARGED
LMFBR
0
0
1
0
480
2,600
11,500
20,600
32,800
43,800
58,000
ANNUALLY «
HTGR
0
0
1
100
2,420
6,600
7,800
10,000
10,000
10,000
8,800
TOTAL
25
790
2,400
5,400
10,400
17,400
26,800
38,000
51,500
64,500
79,000
ANNUAL INVENTORY OF
KRYPTON-85
(CURIES)
2.6 x 105
8.3 x 106
2.5 x 107
5.7 x107
1.1 x108
1.8 x108
2.8 x 108
4.0 x108
5.4 x108
6.8 X108 '
8.3 x 108
* BURNUP: 33 GWd(t)/ TONNE AND 0.35 THERMAL EFFICIENCY.
-------
TABLE 2
ESTIMATED KRYPTON-85 CONTRIBUTIONS
TO WORLD POPULATION DOSE
YEAR
1970
1975
1980
1985
1990
1995
2000
M 2005
CO
2010
2015
2020
KRYPTON-85
CUMULATIVE INVENTORY
(CURIES)
2.6 x 105
1.8 x 107
9.3 x 107
2.6 x 108
5.8 x108
1.1 x 109
1.9 x 109
2.9 x 109
4.2 x 109
5.8 x 109
7.6 x 109
SEA LEVEL
CONCENTRATION
(pCi/m3)
6.63 x 1 0"2
4.59 x 10°
2.36 x 101
6.63 x 101
1.49 x 102
2.81 x 102
4.77 x 102
7.40 x 102
1.08x 103
1.49x 103
1.95x 103
WHOLE BODY DOSE
CONVERSION FACTOR *
(REM/YEAR)/(pCi/m3)
1.5 x 10'8
1.5x1 0"8
1.5x10-8
1.5 x 10"8
1 .5 x 1 0"8
1.5 x 10"8
1.5 x lO"8
1.5 x 10"8
1 .5 x 1 0"8
1.5 x 10"8
1.5 x 10"8
WORLD **
POPULATION
xlO9
3.632
3.990
4.384
4.817
5.292
5.814
6.388
7.018
7.711
8.472
9.308
INTEGRATED WORLD
POPULATION DOSE
(PERSON-REM)
3.6
275
1555
4790
11,821
24,462
45,692
77,847
125,057
189,248
272,259
"(EPA-52019-73-003D, 1973)
**(UNSO, 1966 and 1971)
-------
10
o
z
o
oc
LU
>-
a
DC
UJ
Z
LU
OC
<
LU
_l
o
D
Z
LEGEND
10
D ORNL - 4451 (1970)
A WASH — 1139 (1972): Most Likely Value
X EPA/ORP Report in Preparation:
(EPA-52019-73-003D, 1973)
1970 75 80 85 90 95 2000 05 10 15 2020
YEAR A.D.
FIGURE 1. PROJECTIONS OF NUCLEAR ELECTRIC GENERATION
1970 - 2020
-320-
-------
109
108
T T
t/3
LLJ
QC
o
107
Ul
>
z
106
105
I I
I
I
I
I
1970 75 80 85 90 95 2000 05 10 15 2020
YEAR A.D.
FIGURE 2. ESTIMATED KRYPTON-85 RELEASES TO THE ENVIRONMENT
1970 - 2020
(ADAPTED FROM TABLE 1)
-321-
-------
EQUILIBRIUM VALUE
WITHOUT CONTROLS
(1.33x 1010Ci)
I I I I
EQUILIBRIUM VALUE
W!TH CONTROLS
FOR THE YEAR 2020
(1.33x 108Ci)
1970 75 80 85 90 95 2000 05 10 15 2020
YEAR A.D.
FIGURE 3. ESTIMATED KRYPTON-85 INVENTORY IN THE ENVIRONMENT FOR CONTROLS
INITIATED IN VARIOUS YEARS (DECONTAMINATION FACTOR = 100)
1970 - 2020
-322-
-------
I I I I
EQUILIBRIUM VALUE
WITHOUT CONTROLS
(1.33x 1010Ci)
EQUILIBRIUM VALUE
WITH CONTROLS
FOR THE YEAR 2020
(1.33x 107Ci)
95 2000 05
YEAR A.D.
15
2020
FIGURE 4. ESTIMATED KRYPTON-85 INVENTORY IN THE ENVIRONMENT FOR CONTROLS
INITIATED IN VARIOUS YEARS (DECONTAMINATION FACTOR = 1000)
1970 - 2020
-323-
-------
10,000
i—i—i—r
T T
1,000'
o
a
100
QC
t-
z
LJJ
o
z
o
o
10
1970 75
80
85
J I I I
90 95 2000 05
YEAR A. D.
J L
10
15 2020
FIGURE 5. ESTIMATED KRYPTON-85 CONCENTRATION IN THE WORLDS ATMOSPHERE FROM
U.S. NUCLEAR ELECTRIC POWER PRODUCTION
1970 - 2020
(ADAPTED FROM TABLE 2)
-324-
-------
10°
10°
ill n
DC 104
2
O
t/J
a:
LJ.I
in
O
Q
Z
O
2 io3
O
Q.
102
10'
"i—i—r
j L
I I I L
1970 75 80 85 90 95 2000 05
YEAR A. D.
10
15 2020
FIGURE 6. ESTIMATED ANNUAL INTEGRATED WORLD POPULATION DOSE TO THE WHOLE
BODY FROM KRYPTON-85 GENERATED BY THE U.S. NUCLEAR POWER INDUSTRY
1970 - 2020
(ADAPTED FROM TABLE 2)
-325-
-------
CRYOGENIC ADSORPTION SYSTEMS FOR NOBLE GAS REMOVAL
A. R. Smith
Cryogenic Technology, Inc.,
Waltham, Massachusetts 02154
E. L. Field
Arthur D. Little, Inc.
Cambridge, Massachusetts 02140
R. L. O'Mara
Stone & Webster Engineering Corp.
Boston, Massachusetts 02107
Abstract
Because of the continuing emphasis on decreasing or eliminating noble gas emissions from nuclear power
reactors, systems for the delay and retention of noble gases have been developed. Several of these systems rely
on adsorption, either under ambient or refrigerated conditions. The usual adsorbents are activated charcoal
or synthetic zeolites (molecular sieves).
This paper discusses the design of such noble gas treatment systems. As an example, a description of the
noble gas recovery system installed at the Wisconsin-Michigan Power Company's Point Beach Nuclear Plant
will be presented, including the design philosophy, data upon which the design was based, and some test
resultson the final system.
INTRODUCTION
The continuing concern of utilities with the release of radioactive materials from nuclear power plants
resulting in possible exposure to the public and to plant personnel has led to the development of many systems
to limit these releases. Of major concern is the release of radioactive gases which leak from pinholes or cracks
in reactor fuel rods, and find their way into the steam in a boiling water reactor (BWR) or into the reactor
coolant system of a pressurized water reactor (PWR).
In a PWR, which is continuously maintained at high-power levels, the noble gas concentration in the
reactor coolant system slowly increases unless these gases are removed. As the noble gas concentration
increases, the possibility of accidental leakage resulting in a hazard to the public and operating personnel
increases because of the increased concentration of longer-lived noble gas isotopes, particularly krypton-85. A
leak from the primary coolant system to the secondary system of as much as 500 gallons of water per day for
each reactor has been postulated. This would provide a leak path to the atmosphere via the main condenser air
ejectors. It is, therefore, desirable to continuously remove the noble gases from the coolant system to reduce
the potential hazard.
In the Fall of 1970 the owners (Wisconsin-Michigan and Wisconsin Electric Power Companies) of the Point
Beach Nuclear Plant, a plant with two 500 MW(e) PWR's. asked Stone & Webster Engineering Corporation to
provide waste treatment improvements at the plant consistent with the state of the art. In particular, the
reduction of krypton-85 releases to the "as low as practicable" amount was specified.
PROJECT DESCRIPTION
It was determined that the most effective means of reducing the long-lived noble gas specific activity in the
coolant systems of the two reactors was to continuously strip these gases from the coolant water. This was to
be accomplished by passing the maximum expected let-down flow through a gas stripper. The noble gases
contained in this gas stream would then be separated in a new gas treatment system, and the hydrogen, which
is the major component of the stream, would be returned to the primary system.
The stripper design was based on the fact that short-lived isotope concentrations in the coolant system are
limited primarily by decay as opposed to other forms of removal; e.g., a minimum decay period of only one
hour would be sufficient to reduce the short-lived isotopes in the air ejector releases to a 133Xe equivalent value
approaching 1 percent of that specified by 10 CFR Part 20. The long-lived noble gases content would then be
reduced by continuous stripping and subsequent decay or storage.
In addition, another source of gaseous radwaste releases was recognized — that of boron recovery tank
cover gas being displaced by tank filling operations with small leaks through nitrogen gas regulators. This
gas would be delayed in storage tanks and periodically vented to the atmosphere through the gas treatment
system.
The basic design criteria for the gas treatment facilities involved two main tasks:
(1) delay of noble gases to decay short-lived isotopes, with subsequent removal of longer-lived isotopes from
the hydrogen gas, which is continuously stripped from the reactor coolant and recycled; and,
(2) removal of noble gases from the decayed cover gas prior to release to the environment.
-326-
-------
Typical characteristics for these two gas streams are as follows:
Stripper Overhead Gas Cover Gas
Flow O.Sscfm 600 scf per month
Inlet Conditions
Temperature, °F 120 120
Pressure, psig 1 120 down to 30
Composition (Note: Streams will be saturated with water.)
(mole fraction dry basis)
hydrogen 9.39 .494
nitrogen .060 .501
oxygen .003 .0046
argon 7.1 xlO-4 8.0 xlO-5
carbon dioxide — 8.0 xlO-5
stable krypton 3.7 xlO-6 2.5xlO-s
stable xenon 18.9 x 10-6 8.0 x 10-6
radio krypton 1.9xlO-7 2.8 xlO-8
radioxenon 2.1 xlO-7 7.5xlO-9
EXPERIMENTAL PROGRAM
In response to an inquiry, Cryogenic Technology, Inc., (CTI) proposed to Stone & Webster and Wisconsin
Electric Power Company that the simplest, safest, most reliable method of concentrating and collecting noble
gases is adsorption on charcoal at cryogenic temperatures because of:
(1) the selectivity that charcoal exhibits for krypton;
(2) the dramatic increase in the adsorption effectiveness of the charcoal at -315°F (capacity of charcoal for
krypton per unit volume is about 1,500 times that at 80°F);
(3) the high decontamination factors that can be achieved (in excess of 1,000).
CTI proposed that a study program be initiated in conjunction with Arthur D. Little, Inc. (ADL) to develop
theoretical predictions concerning the adsorption of noble gases at cryogenic temperatures, and to establish
the design parameters for the gas treatment system. Work was begun on the study program with the following
specific goals:
(1) preparation of a theoretical analysis of the multi-component adsorption in the beds;
(2) building of a laboratory-scale test apparatus and conducting of a test program to verify the analysis; and
(3) preparation of a preliminary process and mechanical design for the noble gas removal system.
An experimental apparatus simulating the cryogenic section of the proposed design was built, and tests
were conducted at the ADL laboratories. A flow schematic of the test system is shown in Figure 1.
The test apparatus was designed so that the gas can flow forward through the adsorber beds for normal
operation or backwards through the beds for regeneration. The system can be cooled with liquid nitrogen or
heated with hot nitrogen for regeneration. Six sample taps are provided so that the relative krypton
concentration at various points in the system can be monitored with radiation counting equipment. Two
evacuated sample receivers are provided to collect the adsorbed gases as they are evolved during warm-up.
One cylinder is connected to the first adsorption beds and the other is connected to the last bed. A shut-off
valve can be closed to isolate the last bed from the rest of the system, thus providing the capability of sampling
this bed separately.
During normal operation, the hydrogen gas and the radioactive gas mixture flow from storage cylinders to
Flowmeters 1 and 2, respectively, where they are then combined and flow through a heat exchanger, a filter,
and the three adsorption beds, all of which are contained in an insulated dewar. Pressure on the system is
maintained at the desired level by a back pressure regulator located downstream of the beds. Cooling of the
beds (to -315°F) is accomplished by flowing liquid nitrogen through the cooling/heating loop which is in
contact with the adsorber beds. A small portion of the process gas is continuously bled through one of the six
sample taps, through Flowmeter 3, to the radiation counting equipment. The counts were fed to both a count
rate meter and an automatic sealer for measurement.
The radioactive counting apparatus consists of a modified windowless flow counter. A window of 1/4 mil
aluminized mylar was used to separate the hemispherical counting chamber from the sample chamber,
through which a flow of gas from any given sample point was maintained. Somewhat less than half of the
beta particles emitted during decay of the 86Kr in the sample chamber passed upward, through the mylar
window, into the counting chamber where they were recorded as pulses (counts) per second.
-327-
-------
Warm-up and collection of the adsorbed gases involves shutting off the flow of process gas, isolating Bed
No. 3 from the rest of the system, opening the valves to the sample receivers, and stopping the flow of liquid
nitrogen. The beds are then heated to 300°F by heat exchange with hot gas. As the beds warm up the pressure
rise is noted to determine the quantity of gas evolved from the beds, and the gas is collected in the sample
receivers for analysis.
The pressure, temperature, and gas composition used in the test apparatus were essentially the same as
would be expected in processing the gas from the stripper except that 22.6 ppm of Kr with 85Kr tracer was
substituted for 18.9 ppm Xe and 3.7 ppm Kr, making the results conservative by a factor of approximately 1.1.
A higher level of radioactivity in the actual plant could result in somewhat reduced capacity because of
radiation heating, but a check on the magnitude of this effect at the expected radiation levels predicts that it
will be negligible (Glueckauf, 1959). The charcoal used was the same as that used in the full scale plant.
The apparatus contained three 0.31 in. I.D. beds in series; the first two were 2.5 inches long, and the third
was 11.5 inches long. The first two were intended to provide equilibrium and rate data while the third was
intended to provide decontamination factor (D.F.) data as well as a back-up source of information if
performance estimates were drastically in error. Because the data on the first beds were in close agreement
with performance estimates, it was not necessary to run the last bed to breakthrough.
After a number of familiarization test runs, three runs were made of duration 21 hours, 46.2 hours, and 50.8
hours, respectively. The breakthrough data for these runs are as shown in Table 1. Additional tests have
subsequently been run wherein varying temperatures, gas compositions, sorbents, flow rates, regeneration
conditions, etc., were explored. The data from these additional tests were consistent with the results shown in
Table 1.
The breakthrough times for Runs 2 and 3 (which were duplicated with respect to flow) are also shown as the
experimental points on Figure 2; the correlating curves shown were based on matching with the theoretical
plots of breakthrough given in Hougen and Marshall, (1961) as extended by Burnette, et al., (1961).
The radioisotope level was increased on the third run to provide a more sensitive measurement of the
decontamination factor. The inlet stream produced 2,200 counts per second. The background level was 1.3±
0.1 counts per second so that it would have been possible to detect 0.2 c/s change in the background. Prior to
breakthrough on both Beds 1 and 2, there was no difference between the sample stream and the background
which could be detected; this indicated that the decontamination factor was 2,200/0.2 = 104 or better.
The results of the experimental program showed that for a charcoal bed operated at -315°F and 120 psig with
a 22.6 ppm Kr inlet concentration:
(1) the equilibrium capacity for Kr was 60 scc/gm vs. a value of 120 scc/gm which had been predicted by
theory. The dynamic adsorption coefficient for Kr was thus 60/(22.6 x 10-6) = 2.6 x 106 sec of bulk gas/gm at a
pressure of 120 psig;
(2) the test bed number of transfer units (NTU) was 14 per inch vs. an expected value of 105; and
(3) decontamination factors of 104 or better are possible with adsorption systems.
Although the departures from theoretical estimates were significant, the agreement is adequate to provide
assurance that the projected design would function as expected. The difference in the equilibrium capacity
can be accounted for by slight differences in the literature values of pure component adsorption isotherms for
N2 and Xe on charcoal. The differences in the dynamic transfer processes can be accounted for by other
possible effects such as pore diffusion limitations or channeling, but are unimportant as long as a system has
on the order of hundreds of transfer units.
SYSTEM DESIGN
With the completion of the experimental program, attention was then focused on the specific design
requirements of the full scale noble gas removal system. Some of these requirements were:
(l)the ability to achieve a D.F. of at least 1,000 in processing both the gas from the stripper and the cover gas;
(2) the ability to operate for 180 days on one adsorber before noble gas breakthrough;
(3) the ability to retain the adsorbed noble gases for subsequent transfer to an evacuated noble gas storage
tank;
(4) the ability to regenerate the adsorption bed after transfer of the noble gas in such a way as to remove all
of the residual noble gases and to permit a D.F. of at least 1,000 when the bed was put back on stream;
(5) design for safety, simplicity, and reliability with particular attention being given to leak-tight, rupture-
proof construction; and
(6) design of a compact system to fit into the limited space available at the plant.
The unit subsequently built is shown schematically in Figure 3.
The plant provides two independent systems for processing the gas including compression, water knockout,
oxygen removal, gas drying, and cryogenic separation of the noble gases from the hydrogen-nitrogen
streams. The plant is capable of processing stripper gas (normal continuous operation) through either flow
path while at the same time processing cover gas (intermittent operation) through the other flow path Each
cryosorber module and dryer circuit is capable of regeneration independent of the other.
-328-
-------
The equipment is arranged as shown in Figure 4 with all the components located on one 9 ft. x 4 ft. skid with
the exception of the compressors, the ambient charcoal delay bed, and the cryosorber modules. These
components are located behind concrete shield walls for operator protection.
During normal operation the stripper gas is compressed to 150 psig by a diaphragm compressor, and then
passed through a chiller/water separator operating at 35°F for bulk removal of moisture and attainment of a
low relative humidity for effective utilization of the ambient charcoal delay bed. The gas then goes to the
ambient temperature charcoal bed which delays the noble gases for several hours to permit the shorter-lived
isotopes to decay, thus reducing the radioactivity level at the main skid for operator protection. The gas
stream then goes to a catalytic oxygen removal system to remove trace quantities of oxygen and ozone so as to
preclude the collection and/or formation of ozone on the cryogenic adsorption bed. The oxygen removal
system consists of a preheater, an iodine preadsorber, and a precious metal catalytic hydrogen/oxygen
recombiner.The iodine adsorber protects the recombiner catalyst from iodine, the only possible contaminant
in the stream which would retard its oxygen conversion activity. The gas is then passed through another
chiller/water separator and a molecular-sieve dryer. The dryer, containing adequate 5-A molecular sieve for
180 days operation before regeneration, reduces the dew point of the gas to about -100°F before it enters the
cryogenic heat exchanger. This high performance moisture removal is necessary to prevent frost plugging in
the counterflow cryogenic heat exchanger where the gas is cooled to about -300°F. The cold gas then passes
through the charcoal adsorption bed.
The adsorption bed was designed using information obtained during the experimental program which
showed that a bed containing 0.25 cubic feet of charcoal, 2 inches in diameter by about 140 inches long,
operating at -315°F would be adequate for processing stripper gas for 180 days (plus 3,500 scf of cover gas) with
a safety factor of 2. This bed would liberate approximately 38 scf of adsorbed gas during each regeneration.
Since two regenerations per year were planned, the 525 cubic foot evacuated vessel in which the desorbed gas
would be stored should be adequate for the storage of at least seven years regeneration gas without exceeding
one atmosphere.
Each heat exchanger and its adsorption bed is contained in a separate perlite-insulatea vessel. They have
remotely operated valves on the inlet and outlet piping which will automatically close in the event of high
temperatures in the adsorber or high radiation downstream of the adsorber, thus preventing the release of
any radioactive material. The heat exchanger vessel is oversized to act as a surge volume to reduce the
pressure rise, which would occur if the vessels were to warm to the ambient temperature. These components
are also designed to withstand the stresses associated with this pressure rise. The bed is maintained at -315°F
by a continuous flyw of liquid nitrogen.
After passing through the adsorption bed, the clean hydrogen passes out through the heat exchanger and a
back pressure regulator which maintains the operating pressure at 150 psig. The gas is then monitored for
radiation and routed back to a volume control tank for reuse.
The system is designed to operate for 180 days on one processing stream. Flow will then be switched to the
other stream, the off-stream adsorber will be warmed and the adsorbed noble gases will be collected and
stored; then the system will be regenerated by flowing clean, dry effluent gas from the on-stream system
backwards through the adsorber and dryer. The regeneration gas will be routed to the volume control tank via
the closed loop system and there will be no radioactive material vented to the atmosphere. After regeneration
is complete, this processing stream will then be placed on standby until the 180-day cycle on the other stream
is complete.
When the nitrogen-rich cover gas is processed, the operation of the system is essentially the same except
that the ambient temperature charcoal delay beds are bypassed, and the system is operated at one atmosphere
to avoid nitrogen condensation in the cryogenic part of the system. This cleaned cover gas is vented to the
atmosphere after passing through a radiation monitor rather than being recycled.
The equipment and components used in the system were selected with safety and reliability as a main
concern. Leak-tightness is assured by the use of diaphragm compressors, bellows seal valves, and all-welded
piping construction whenever possible. The entire system was helium-mass-spectrometer leak checked after
assembly. Vessels were built in accordance with Section III of the ASME code. The electrical equipment was
designed for use in hazardous locations in accordance with the Class II, Division 1, Group B requirements of
the National Electrical Code.
To prevent the inadvertent release of radioactive materials, sample streams and relief valve vents are
routed back to the process stream. Radiation monitors which will automatically close the system outlet valves
and contain the adsorbed gases in the adsorbers are installed in the system effluent piping. Regeneration is a
closed loop process with no release of gases to the atmosphere.
Process variables such as temperature, pressure, and process stream oxygen content are continuously
monitored and displayed on a local control panel. Critical variables are alarmed at this local panel and on the
main reactor control panel. The system can be started and operated from the local control panel with a low
radiation exposure hazard, because the components containing highly radioactive materials, the
compressors, the ambient delay beds, and the adsorbers are isolated behind shield walls. This arrangement
also results in a low radiation level at the main equipment skid and permits routine inspection and
maintenance to be performed with the system in operation.
-329-
-------
PRESENT STATUS
Before the unit was shipped, a full scale radioactive 85Kr tracer test was conducted under design flow
conditions. Gas mixtures approximating both the stripper overhead gas and the cover gas were processed
through the unit for several days. All functions of the system were verified, including water removal, oxygen
removal, and noble gas adsorption. The system demonstrated a D.F. of 5,000 during a four-day continuous run
at a flow rate of 1 scfm.
The system was installed at the Point Beach plant and is in the final stages of checkout. The installed
equipment is shown in Figure 5. The operating data from the plant will form the basis for future reports on the
performance of the system.
Construction and operation of the Point Beach system and a similar system installed by the CVI
Corporation, at the Southern California Edison Company's San Onofre plant, to remove noble gases from an
intermittent flow of cover gas, have proven the practicality of cryogenic adsorption systems. Other charcoal
or molecular sieve delay beds have since been designed by CTI both for low PWR off-gas flow rates and for the
higher flow rates from BWR's and fuel processing plants. Adsorption systems for noble gas delay operating at
higher temperatures have also been designed for use in lower performance systems. However, cryogenic
adsorption has been shown to be appropriate and practical where noble gas removal or highly effective delay
with an extremely high D.F. is required.
REFERENCES
Burnette, R.D., W.W. Graham and D.C. Morse, (1961), GA 2395, (unclassified).
Glueckauf, E., (1959), The Movement of Highly Radioactive Gases in Adsorption Tubes, Annals of N.Y.
Acad. Sci., 72, 562.
Myers, A.L. and J. M. Prausnitz, (1965), AIChE Journal 11, 121, 127.
TABLE 1. Kr Breakthrough Times at 80°K from Experimental Gas Mixture.
Mass Velocity Times to Indicated Breakthrough Ratio(hrs)
1 8.6 atm 3.0 times - 17.7 (b)
2 9.3 4.6 - - 14.1 18.6 - 34.5 (b)
3 9.3 4.6 7.5 9.8 12.9 16.6 21.8 27.0 31.5 41.7
(a) Plant nominal mass velocity = 18.5 lb/ (hr-ft2)
(b) Run terminated prior to reaching this breakthrough.
-330-
-------
0-300
-30-0-300
© ©
30 0-3CO
<~>
PM
"* | [gJLgf
F1
H3 In
F2
N2-Ar-Kr In
Variac
LJ
LT\J. IP
Purge
-*-
s.v.
, 3 4 5 6
HX1
Cooler/'
Heater
:ilter
Bed No. 1
-TCI
Bed No. 2
To
F3
Sampling
Tail
Cooler/
Heater
-Bed No. 3
-TC3
1:
J- Cat
For Vacuum
0-100'
[Pi A)
0-100
(P2AJ
Legend:
(?) = Pressure Gauge
JUiH. - Capillary Flowmeter
. - Process Lints
• ~ Sar-lOhn.T Lin.jj
— Crol">y M .::•
Note:
|L
Denotes
Heat
Exchange
Figure 1. Test Rig Schmetic.
-331 -
-------
10'
c
o
o
£ 10"'
10
-3
10
-4
1 —
—
—
1
7
1
^_
1
/
f
i
i
i
i
/
—
i
3C
.
s
—
X
/
/
' ^
y
i
S3
~
—
\
4
~\~
i
;
<<
y
x^
/r-
-A
s"
'i
J
Bed Length - 2.5 inches
/Theoretical Curve
for 35 Transfer
Units. b= 1.41 hr"1
—
I
—
— -
'
/
I
J
1
i
[
-s-
r
/
d
0'
f
—
-"
r-j
-^
r1
1
__
— -
>-
x
"""^L~~C
_
,
*
i
/
f
•
—
'
i
-b
*
.
j, ^
^
,Bed -ength = b.O imihes
Theoretical Curve
for 70 Transfer
Units, b= 1.41 hr"1
1
1
i
~T~
i
Note:
O^ Experimental Points
—
— ]•• • I
i
—
|
~
—
—
- —
—
10
20
40
50
60
Elapsed Time, Hours
Figure 2. Breakthrough Performance for 2,5-inch and 5.0-inch bed length.
-332-
-------
STEAM
GENERATOR
REACTOR
co
CO
CO
C CONDENSER
HYDROGEN
RECYCLE
COOLER
r-L-, ADSORPTIVE PREHEATER
LMJ DELAY
COMPRESSOR (AMBIENT
CHARCOAL
BED)
IODINE OXYGEN
CHEMI- REACTION
SORPTION
NITROGEN
TO
ATMOSPHERE
LN,
NOBLE GAS
TO STORAGE
(INTERMITTENT)
CRYOGENIC
ADSORPTION
COVER
GASES
Figure 3. Noble Gas Separation by Selective Adsorption (PWR).
-------
22
co
CO
*.
SKID
ASSEMBLY
4'-4"x9'-6"
CONTROL
CABINET
T
COMPRESSOR ASSEMBLY
WITH WATER SEPARATOR
CRYOSORBER
ASSEMBLY
Figure 4. Equipment Arrangement.
-------
03
w
VI
IODINE G
I OXYGEN
REMOVAL
UNITS
LN2
SURGE
TANK
CRYOGENIC
ADSORPTION
MODULE
CRYOGENIC
ADSORPTION
MODULE
Figure 5. Noble Gas Removal System at Point Beach Station.
-------
RECENT ADVANCES IN THE ADSORPTION OF NOBLE GASES
D.W. Underbill and A.S. Goldin
Harvard University
School of Public Health
Department of Environmental Health Sciences
Boston, Massachusetts 02115
Abstract
The Harvard Air Cleaning Laboratory has hadan activerole in the design of krypton and xenon adsorption
beds for the FFTF reactor. We describe an inexpensive laboratory system for measuring fission gas
adsorption on various charcoals, and present results of an extensive series of tests of adsorption of krypton
from an argon carrier gas on Pittsburgh PCB charcoal. The effects of decay heat, as well as the release of
fission gases from the rupture of a pressurized bed, are also discussed.
INTRODUCTION
From 1969 to 1973, basic research for the design of fission gas adsorption beds for the FFTF nuclear reactor
was carried out at the Harvard Air Cleaning Laboratory. We summarize here some of the more interesting
results of this work.
MEASUREMENT OF ADSORPTION COEFFICIENTS OF KRYPTON AND XENON
At the time this study began, some data (Collins,e£ al, 1967; Trofimov, et al, 1968; Barilli, et al., 1969; and
Forster, 1971) were available for the adsorption of krypton and xenon from the argon carrier gas to be used as
the FFTF cover gas, but there was little information available from studies in which the specific charcoal used
was domestic. In order to make a preliminary design there was a pressing need for such data over a wide range
of temperatures and pressures. We found that an experimental test system could be assembled rapidly from
commercially available equipment (see Figures 1 and 2), using as the constant temperature chamber a foamed
plastic picnic basket cooled by an internal liquid nitrogen spray. Ordinary copper refrigerator tubing was
used to pre-cool the gases in the chamber before passing through the small iron pipe used to contain the
charcoal adsorbent. Argon carrier gas was supplied from a commercial compressed gas cylinder, and liquid
nitrogen coolant was obtained from a 160-liter cryogenic tank which had to be replaced once or twice a week.
The fission gas isotopes, 85Kr and 133Xe, dissolved in isotonic saline, were obtained commercially. In this form,
these fission gases are often used for medical studies of blood-gas interaction in the lung. For our purposes, we
would withdraw S 0.1 to 1.0 mCi of fission gas into a syringe, shake the liquid with a small amount of argon to
bring the fission gas into the gaseous carrier, and inject the argon-fission gas into the carrier gas system
entering the cryogenic chamber. The concentration of fission gas, and the quantity of carrier gas passing
through the bed, were measured by an ionization chamber and a wet test meter, respectively. From these data,
the dynamic adsorption coefficient can easily be calculated.
There are three points that we would now like to make about this system:
(1) It was quite inexpensive to construct.
(2) The experimental data we otained on small samples of charcoal appear to be in good agreement with
much larger scale (103 to 104 times larger) tests run more recently by Kabele, 1973, et al.
(3) We found a surprising variability between commercially available charcoals in their ability to retain
fission gases.
The variability between charcoals is illustrated in Figure 3 which shows a plot of relative adsorptive
capacity vs. relative surface area (as stated by the manufacturer). The correlation between these two
parameters is quite small, showing the need for the experimental testing of charcoals before their use in
fission gas adsorption beds.
CALCULATION OF ADSORPTION COEFFICIENTS FROM BREAKTHROUGH DATA
The use of moment analysis to calculate adsorption coefficients from breakthrough curve data has the two
advantages of: (11 using all the data of the breakthrough curve for the analysis (rather than depending
heavily on a few of the experimental points); and (2) being independent of the mechanisms of mass transfer
that occur within the bed. The theory behind this analysis has been reported elsewhere (Underbill, 1970a);
here we show how it can be applied to the analysis of experimental data.
-336-
-------
The moments of the breakthrough curve produced from a pulse input of a fission gas isotope into the bed, are
calculated as:
/*°°\T \t /Y*°° U
1N = / tNeAtc(t)dt// eAlc(t)dt
4 /4
where t = time following injection of fission gas,
c(t) = concentration of fission gas in effluent at time, t, /xCi/cm3
X = decay constant for the fission gas isotope, a-1.
From the first two moments, we can make some simple, but useful, calculations.
The mean holdup time is equal to Mj and the dynamic adsorption coefficient, k, is
where k = dynamic adsorption coefficient, cmVgm
V^=flow of carrier gas, cmVs
m = mass of adsorbent, gm
Mi = mean holdup time, s.
The mass transfer in fission gas adsorption beds can be characterized by the number of theoretical plates, N,
calculated as
N=M12/(M2-M12).
If it is desired to calculate the efficiency of the bed for the removal of a noble gas isotope, then the ratio of
effluent to input concentration, at steady state, following a constant input of fission gas is
/*°° //*
t=/e-(X2-Mtc(t)dt//e
Jo I Jo
R= le-( 2 •* HcWdt/ le xtc(t)dt
Jo / Jo
where X = decay constant of isotope used in pulse test, s-1
A = decay constant of the fission gas isotope under consideration, s-1.
£t
The generality of this procedure is illustrated by applying it to current theories of mass transfer in fission gas
adsorption beds. These same results are obtained when applied to breakthrough curves calculated from film
controlled mass transfer (Young, 1958), interparticle diffusion (Madey, et al., 1962), intraparticle diffusion
(Underbill, 1970a), and from the theoretical chamber model (Underbill, 1970b).
As a computer exercise we calculated a breakthrough curve for which N = 20, Mj = 10 days, input = 70 Ci, X =
0.148/day, and the mechanism of mass transfer was interparticle diffusion. Then we tried the inverse
calculation. Starting with the value for X, and the breakthrough curve data, the values for N, MI, and the
input were correctly calculated using the above procedure. In making this calculation, no knowledge was
needed of the mechanism of mass transfer within the bed.
DECAY HEAT EFFECTS
Good adsorbents — porous granular materials — are not good conductors of heat. Consequently, at steady
state operation, a constant release of a relatively small quantity of decay heat can result in a rather large
temperature increase.
There is, at present, no complete mathematical solution available for the calculation of the temperature rise
in a fission gas holdup bed. The basic difficulty is the nonlinear effect of the temperature on the adsorption
coefficient. Glueckauf (1959) did find an analytic solution for a non-linear radial distribution of fission gas
resulting from the radial temperature distribution, but this solution neglects other factors — including the
channeling of hot gases through the center of the bed, making it less useful than would first appear.
-337-
-------
The analysis given here assumed initially that the temperature rise is small. On this assumption, a linear
partial differential equation can be written describing heat transfer. By integrating this equation to meet the
boundary condition of an initially ambient temperature for the entering gases, the average temperature and
the centerline temperature can then be calculated. If the effective adsorption coefficient used for design
purposes is based on this latter temperature, the design will be conservative. If the calculated temperature
increase is large, it may be possible to redesign the bed to lessen the thermal effects, and in doing so to permit
this analytic analysis to be valid for the final design.
This analysis began with the following differential equation (First, et al., 1970) which gives the effects of
radial and axial heat conduction, convection, and radioactive decay on the average bed temperature at a
distance, x, from the inlet.
3t
__ . eti + qie- - Pck
where C^ = heat capacity of charcoal, where cal/gm-°C
Cp = heat capacity of carrier gas at constant pressure, cal/cm3-°C
k = thermal conductivity of the charcoal, cal/cm-s-°C
K = adsorption coefficient for fission gas, emVgm
q j= production of decay heat per unit volume of bed from the Itn isotope at the inlet, cal/cm3-s
t j= average temperature increase resulting from the decay of the I*-h isotope, °C
Vs = superficial carrier gas velocity, cm/s
x= distance from inlet of bed, cm
ttj = linear decay coefficient for the Ith isotope, cm-1
= ApK/V s
/3 = coefficient for radial heat loss, cal/cm3-s-°C
(For an infinite cylinder, /3 =2k/r2.
y = convection coefficient, cal/cm2-s-°C
= Cp-Vs
A = isotopic decay coefficient, s-1
p = bulk density of charcoal, gm/cm3
T =time,s.
Assuming that the temperature rise is small, the adsorption coefficient will remain constant across the bed.
At steady state, -2L [-o, and the above differential equation can be integrated to give
e
~alx -
The average temperature increase at a point, x, from the inlet is
N
-338-
-------
Furthermore, a particular isotope, t j produces a maximum temperature at a distance xmax, found where
dt.
From the equation for t j, at steady state
- fcn
max
Tek - yj /2kr
- Y)
/2k
A general solution to the problem of decay heat can be obtained by numerical analysis with a digital
computer. At present, we have only a one dimension model with a constant adsorption coefficient; i.e., a
numerical equivalent of the analytical solution we have just described. These computer calculations replace
the differential equation with the following difference equation for tj(x), the temperature increase resulting
from decay of the I"1 isotope at a distance, x, from the inlet.
k (tT(x-Ax) + tT(x-Ax) - 2tT(x) - 2tT(x) + t,(x+Ax)
-L, _L •*. -L j-
* , 0 *
+ tT(x+Ax)> /2Ax - Y (tT(x+Ax) + tT(x+Ax)
* r , N
- t (x-Ax) - tjCx-Ax)} /HAx - b {tz(x)
u o
4- tjCx)} /2 + qz6 IX = pCk(t (x) - t(x)} /Ar
where Ax = incremental distance, cm
^ AT « incremental time, sec
t_(x) • tj(x) at time T + AT
The set of simultaneous equations describing the temperature increase across the bed is then solved by use
of the tridiagonal matrix procedure described by Westlake (1968). We assumed, in making these computer
calculations, that both the inlet and effluent gas temperatures were controlled by coming into contact with
cold walls; in contrast, the analytical model discussed earlier assumed a semi-infinite adsorption bed with
only one cold end. Figure 4 shows the heating in an adsorption bed following a steady state input of a fission
gas mixture in an argon carrier gas starting at time T = 0. The numerical values used in this calculation are
listed in Table 1. The important observation is that a significant temperature rise can result from a fairly
small — but constant — input of fission gas. Table 2 gives a comparison between the steady state numerical
and analytical solution. Where the difference in boundary conditions can be ignored, the agreement is quite
close.
It would be useful to extend this numerical analysis to a more general analysis which can take into account
the difference in the adsorption coefficient brought about by radial and axial heating, and this will be the
subject of future work. A particularly interesting result would be to extend this model to the point where it can
be used to follow the affect of a sudden release of a large number of curies of fission gas isotopes into an
adsorption bed.
EFFECT OF DEPRESSURIZATION OR OF COOLANT LOSS
A theoretical equation for the fission gas release due to a sudden loss of pressure has been compared with
experimental measurements of the same phenomenon (Underbill, 1972). The theory predicted that the loss of
fission gas would be considerably less than the loss of carrier gas, but the experimental results showed an
even smaller release than that predicted theoretically. We attributed this effect to the noticeable cooling of the
bed which occurred as a result of the rapid desorption of the carrier gas.
More recently, in work unrelated to FFTF design, we have examined theoretically the loss of fission gas
from a bed undergoing an increase in temperature. As the bed warms up, some of the desorbed carrier gas will
carry out some of the fission gases. This effect is closely related to the loss of fission gases during the rapid
cooling and subsequent warming of a bed, which we found taking place during the loss of pressure accident,
and we think that it will not be very difficult to develop equations which will prove to be more accurate than
those developed earlier to describe this type of accident. As future work, we plan to test such results
experimentally.
-339-
-------
ADSORPTION OF CARRIER AND FISSION GASES
In the course of our work, we made a number of measurements of the dynamic adsorption coefficients of
krypton and xenon, as well as of the adsorption coefficient of the argon carrier gas. We are presently preparing
a report on the results of these measurements. As a summary of what was found, the following two figures are
quite useful. The first (Figure 5) is a Polanyi plot of the adsorption of the argon carrier gas. In this plot the y
axis gives the volume of liquid argon adsorbed per 100 grams of charcoal, and the x axis is proportional to the
free energy required to compress argon to form a unit volume of the liquid phase. This plot represents, nearly
as a straight line, results from a wide range of test conditions. The final figure (Figure 6) shows krypton
adsorption as influenced by the adsorption of the argon carrier gas. The points which correspond to zero
carrier gas adsorption were obtained using helium as the carrier gas. This figure shows clearly the strong
effect of the carrier gas in reducing the adsorption of the fission gas; and, indeed, it was the accurate
measurement of this effect that was one of the main goals of this research.
REFERENCES
Barilli, L., L. Bruzzi, and J. C. Shorrock (1969), Fission Product Removal from a Vented Fuel FBR,
Report RT/ING (70)7 of Conf 691104-5, Paper presented at the International Congress on the Diffusion of
Fission Products, Saclay, France, November 1969.
Collins, D. A., R. Taylor and L. R. Taylor (1967), The Adsorption of Krypton and Xenon from Argon by
Activated Charcoal, TRG Report 1578 (W) Draft, September 1967.
First, M. WM et al., (1970), Semiannual Progress Report of the Harvard Air Cleaning Laboratory, NYO-
841-22, April 1970.
Forster, K. (1971), Delaying Radioactive Fission Product Inert Gases in Cover Gas and Off-Gas Streams
of Reactors by Means of Activated Charcoal Delay Lines, Kern technik, 13,214-19.
Glueckauf, E. (1959), The Movement of Highly Radioactive Gases in Adsorber Tubes, Annals New York
Academy of Sciences, 72,562.
Kabele, T. J., J. L. McElroy, E. O. Badgett and A. P. Bohringer (1973), FFTF Fission Gas Delay
Beds: Engineering Scale Test Report, HEDL-TME-73-26, May 1973.
Madey, R., R. A. Fiore, E. Pflumm, and T. E. Stephenson (1962), Transmission of a Pulse of Gas
Through an Adsorber Bed, Trans. Am. Nuc. Soc., 5,465-66.
Trofimov, A. M. and A. M. Pankov (1968), Influence of the Gas Macrocomponent of the Distribution of
85Kr and1MXe Between the Gas Phase and the Solid Carbon Sorbent, Soviet Radiochemistry, 7,293-98.
Underbill, D. W. (1970a), An Experimental Analysis of Fission Gas Holdup Beds, Nuclear Applications
and Technology, 8,255-60.
Underbill, D. W. (1970b), Commentary on the Theoretical Chamber Model for Fission Gas Holdup Beds,
Nuclear Safety, 11,321-22.
Underbill, D. W. (1972), Effect of Rupture in a Pressurized Noble Gas Adsorption Bed, Nuclear Safety, 13,
478-81.
Westlake, J. R. (1968), A Handbook of Numerical Matrix Inversion and Solution of Linear Equations, J.
Wiley and Sons.
Young, J. F. (1958), A Derivation of the Equation for Elation Chromatography Assuming Linear Rate
Constants, in Gas Chromatography, Coates, V. J., Ed., pp. 15-23, Academic Press.
-340-
-------
Table 1. Values Used to Calculate Temperature Increase in a Fission Gas Adsorption Bed.
Bed Length=310 cm
Bed Diameter=90 cm
Thermal conductivity of Argon + Charcoal = 0.00021 cal/cm-s-°C
Heat Capacity (per unit volume)
of Charcoal 0.085 cal/cm3-°C
of Argon 0.000194 cal/cm3-°C
Adsorption Coefficient (volume/volume)
for Krypton 14 ml/ml
for Xenon 250 ml/ml
Flow of Argon 1,610 ml/s
Fission Gas Data:
Isotope
133Xe
136Xe
85Kr
87Kr
88Kr
Curies Released
per day
1.392
.03369
676.50
3,420.
235.86
391.50
.026
679.20
824.40
Initial Volume
production of
decay heat, q j
cal/ml-s
xlO-8
0.055
1.9
29
481
0.27
5.2
0.000183
46
53
Linear decay
constant al
cm-1 x 10-4
66.8
34.8
15.1
20.90
118
50
0.0023
160
78.5
Incremental distance = 1 cm
Incremental time = 15,000 s
-341 -
-------
Table 2. Comparison of Numeric and Analytic Solutions for the Average Increase Resulting from
Decay Heat.
Note: For numeric analysis A x = 1 cm and Ar = 15,000 s (See text for boundary conditions.)
Distance from Inlet, cm Calculated Temperature Increase
0
1
2
10
100
200
300
309
310
-342-
-------
POTENTIOMETER
CO
if*.
CO
PRESSURE
GAUGE
CARRIER
GAS
INLET
LIQUID N2
VENT
WET TEST METER
SOLENOID
CONTROL
SOLENOID
—n
BY PASS
PRESSURE
RELEASE
IONIZATION
CHAMBER
AND
VIBRATING REED
ELECTROMETER
AMPLIFIER
1
INJECTION^ ^DRYING
PORT TUBE
PRECOOLING COIL
THERMOCOUPLES
STYROFOAM INSULATED BOX
ADSORPTION BED
Figure 1. Diagram of test apparatus.
-------
Figure 2. Apparatus for determining adsorption coefficients at elevated pressures and reduced
temperatures. *l-nitrogen spray head; *2-adsorption column; *3-carrier gas precooling coil; — 1-liquid
nitrogen source; 2- carrier gas source; 3-pressure gauge; 4-injection system; 5-low temperature chamber; 6-low
temperature controller; 7-potentiometer for column thermocouple; 8-ionization chamber; 9-west test gas meter;
10-electrometer for ionization chamber.
*Close up view.
-344-
-------
**
Cn
o
<
u.
•"
1
z
UJ
u
U-
u_
UJ
o
u
z
1-
Q.
tr
o
(O
o
<
(0
^
o
z
o
CO
(O
u.
CO
H
co
UJ
»- 1.2
u,
0
k-
UJ
CO
< 1.0
z
6
0
u.
0.8
UJ
o
cr
UJ
2 0.6
r> A
-s
* A
__ • A .0 9 A
0 •
D
1 '"
• A
—
1 1 1 1 1
TEST SET
A Xe 37.5 °C
a Kr -78°C
• Kr 37.5° C
0 Xe -78" C
0.6 0.8 i.O 1.2 1.4
STATED SURFACE AREA OF CHARCOAL USED IN TEST (SSA)
AVERAGE OF SSA IN A SET OF TESTS
Figure 3. Correlation of fission gas adsorption coefficient with stated surface area of charcoal.
-------
5.00
CO
*.
OS
STEADY STATE
2500 MINUTES
20 40 60 80 100 120 (40 160 180 200 220 240 260 280 300
AXIAL DISTANCE, CM.
Figure 4. Calculated temperature increase from fission gas input as a function of time and axial distance.
-------
50
-J
<
o
o
QC
<
I
O
40 —
(0
2
(!)
O
2 30
9
o
g
_i
ro
o
S 20
CD
QL
O
(O
O
O
O
QC
<
U_
O
UI
2
10
o
o
A
a
-I40°C
-115° C
-IOO°C
-75° C
-I20°C
i
6 8
iog,0(fs/f)
10
12
Figure 5. Polanyi plot of argon adsorption on Pittsburgh PCB charcoal.
-347-
-------
5000
g
o
or
<
x
o
z
o
I-
Q-
V
a:
to
tr
UJ
H
h-
UJ
O
LU
O
O
Z
g
»-
Q_
CC
O
(O
o
o
H
Q.
1000
500 —
100
0 0.1 0.2 0.3
ML LIQUID ARGON/GM CHARCOAL
Figure 6. Effect of adsorbed argon on the krypton adsorption coefficient.
0.4
-348-
-------
ADSORPTION OF RADIOKRYPTON ON ACTIVATED CHARCOAL IN THE PRESENCE OF
HYDROGEN*
B. B. Fisher, A. E. Norris, and D. G. Rose
Los Alamos Scientific Laboratory
University of California
Los Alamos, New Mexico
Abstract
Tests in 1972 of the Nuclear Furnace-1 reactor at the Nuclear Rocket Development Station in Nevada
marked the first time that fission product radioactivity in the effluent from a nuclear rocket engine
development experiment was removed before the effluent was exhausted to the atmosphere. The final
component in the effluent clean-up system was a stationary bed containing 1,565 kg of activated charcoal to
remove radiokrypton and radioxenon from the effluent stream.
Experiments with the charcoal trap operating at a different temperature during each of three reactor tests
were conducted to measure the velocities of the krypton adsorption front, moving through the charcoal under
operating conditions. Results from measurements of radioactivity associated with krypton-85m, krypton-88,
and krypton-89 are presented. At a charcoal temperature of 186°K, the measured adsorption coefficient agrees
with that reported in the literature, but at 161°K and 172°K the coefficients are significantly lower than the
literature values. Most likely the differences arise from interference by the major component of the effluent,
hydrogen, with the adsorption of radiokrypton on the activated charcoal.
INTRODUCTION
The Nuclear Furnace-1 (NF-1) was a small test reactor which was devised to provide an inexpensive means
of testing nuclear rocket fuel elements and other core components. The various runs, called Experimental
Plans, which took place in the period between May 24 and July 27,1973 were the first occasions in which the
hydrogen propellant flow from a reactor rocket engine was scrubbed clean of all radioactivity before being
liberated to the atmosphere by burning it in a flare stack. The last component of the NF-1 Effluent Cleanup
System (ECS) was a fixed-bed activated charcoal trap operated at low temperatures to remove radiokrypton
and radioxenon from the effluent gases. Experiments were performed to measure the velocity of the krypton
adsorption-front through the charcoal trap by measuring the amounts of radiokrypton upstream,
downstream, and in two locations within the charcoal trap.
The following results were obtained:
(1) When operated at the proper temperature, the NF-1 charcoal trap removed the radiokrypton and
radioxenon from the effluent stream.
(2) The dynamic adsorption coefficients of krypton upon activated charcoal derived from these
experiments are less than those in the literature for krypton adsorbed from helium. This effect can probably be
ascribed to the fact that the hydrogen in the effluent is interfering with krypton adsorption.
(3) The number of 89Kr atoms entering the charcoal trap as measured by these experiments is in fair
agreement with the measurements made by Nuclear Rocket Test Operations, and both agree with the
computations of the amount of 89Kr released by the elements.
THEORY
Hydrogen effluent exiting the reactor at = 2440°K was cooled by the injection of water into the effluent
stream. Passing through a series of heat exchangers and water separators, the effluent was then cooled
further and condensed; the added water was removed with a large fraction of nuclides that were in particulate
form or were solubile in water. After this, the effluent passed through a fixed-bed silica gel drier, and a
cryogenically cooled fixed-bed activated charcoal trap to remove the noble gas fission products, krypton and
xenon.
Xenon is adsorbed much more strongly upon activated charcoal than is krypton. Therefore, any trap
designed to remove krypton from a gas stream is more effective for xenon. This discussion is limited to the case
of the adsorption of krypton on activated charcoal, particularly North American G-212,8x16 mesh — the type
used in the charcoal trap.
When the effluent stream from the reactor reaches the inlet to the charcoal trap, the effluent consists
primarily of hydrogen gas, but contains gaseous impurities that result from fuel-element corrosion, and from
the reaction of those corrosion products with the water added to cook the reactor effluent. It may also contain
some particulate matter, which need not be considered here. Table 1 lists a predicted composition of the
effluent stream from NF-1 at the inlet to the charcoal trap.
Although such impurities as CH4 and carbon monoxide can reduce the amount of krypton adsorbed upon
charcoal, an examination of the literature (Burnette and Lofing, 1967) indicated that, in the concentrations
found in the NF-1 effluent stream, they do not affect krypton adsorption. Water vapor is so strongly adsorbed
* Work per formed under the auspices of the U.S. Atomic Energy Commission.
-349-
-------
upon charcoal as to completely prevent the adsorption of krypton or xenon, but the amount of water carried to
the trap by the effluent is so low that it poisons only the first few millimeters of charcoal, and effectively
shortens the trap by that short length. Hydrogen is adsorbed upon activated charcoal less strongly than
either krypton or xenon. However, the amount of gas adsorbed on charcoal at any temperature will increase as
the pressure of that gas is increased. What effect the hydrogen in the effluent, which is at = 4 x 105 Pa, would
have upon the adsorption of krypton, which is at a partial pressure of only 5.45 x 10-7 Pa, was unknown before
the NF-1 tests, and this was one of the facts to be determined by the experiments.
Table 2 and Figure 1 present the dynamic adsorption coefficient of krypton upon activated charcoal versus
the temperature for low concentrations of krypton in helium (Burnette, et al, 1961; Burnette and Lofing, 1967;
and Kovatch, 1970). These values were measured in the absence of gases that would interfere with the
adsorption of the krypton.
When krypton is added at some definite time to a hydrogen stream flowing through a fixed charcoal bed, the
krypton travels through the bed as a concentration-front which moves from the inlet to the outlet of the trap.
Because the escape of fission products from the fuel elements is a very steep function of temperature, the
krypton does appear in the effluent stream suddenly as the fuel elements reach their operating temperature.
This front moves through the bed at a velocity that is a fraction of the velocity of the carrier hydrogen,
depending upon the value of the dynamic adsorption coefficient. As the krypton concentration-front moves
through the charcoal bed, the krypton becomes diffuse because of mass-transfer effects and diffusion.
Burnette, et al., (1961) presents an expression that allows the concentration of the krypton in a hydrogen
stream flowing through an activated charcoal bed to be calculated as a function of bed conditions, time of
flow, and distance through the charcoal bed (this expression includes the effect of decay if the krypton is
radioactive):
0-BT(l+r)
c -axr/(l+r)
_-.» e
- I c LV
H' P "kllM
where
CQ = concentration in gas stream entering bed, moles/m3
V = volume flow rate of gas, m3/s
H = height of mass-transfer unit, m
H can be estimated from the following expression:
ax
where
a = superficial area of bed particles per unit volume, mVm3
Dp = average particle diameter, m
G = flow rate per unit area, kg/m2s
\L = viscosity, kg/ms
p = carrier gas density, kg/m3
Dy = diffusivity of adsorbatein carrier gas, mVs
-350-
-------
c = concentration of adsorbate in gas stream, mole/m:i
T = time, s
x = distance along adsorber bed, m
M = total mass of adsorbent in bed, kg
F = void fraction in adsorber bed
L = total length of bed, m
A = radioactive decay constant, s-1
r =A/j8
I0 = modified Bessel function of the first kind and zero order.
They found good agreement with this expression in experiments investigating the adsorption of krypton on
activated charcoal from a helium stream.
The experiments carried out on the NF-1 effluent cleanup system charcoal trap involved the use of activated
charcoal sampling traps, cooled to liquid-nitrogen temperatures, which sampled the effluent upstream,
downstream, and at two positions within the large charcoal trap. The sampling traps, operating at much
lower temperatures than the large trap, are much more efficient and are able to remove all the krypton and
xenon from the effluent. These traps integrate the concentration of 89Sr, 88Kr, and 85m^r m fae effluent over
the time of the reactor run. For comparison with the experimental results the integral of C/CQ was computed for
each sampling trap in each experimental plan. The computations were done by using the proper operating
temperature for the large charcoal trap and length of run for each experimental plan. In addition, because the
effect of hydrogen upon the adsorption of krypton is uncertain, the computation was performed by using a
dynamic adsorption coefficient equal to 1, 1/2, 1/3, 1/5, 1/7, and 1/10 the literature value for krypton
adsorption upon charcoal. The results of these computations as well as the experimental results are discussed
in the following section.
EXPERIMENTAL
The charcoal trap of the NF-1 Effluent Cleanup System has a diameter of 1.52 m, and is filled to a depth of
1.8 m with 1565 kg of North American Type G-212 activated charcoal of 2 x 16 mesh particle size.
The experiments were planned to measure the velocity of the krypton concentration-front through the fixed
charcoal trap by sampling the effluent in the charcoal trap at four locations. To do this, four bypass lines were
installed in and around the large charcoal trap to lead small fractions of the effluent flow to the sampling
traps. Figure 2 presents a schematic of the sampling system. The identifications and locations of the sampling
positions are given below:
(1) TRAP 20 — immediately upstream from the large trap.
(2) TEAP 30 — 0.152 m below the charcoal inlet surface.
(3) TEAP 40 — 0.762 m below the charcoal inlet surface.
(4) TRAP 50 — at outlet from charcoal trap; this is equivalent to 1.8 m below the charcoal inlet surface.
The effluent sampling head placed in the charcoal trap itself consisted of 0.5-inch stainless-steel tubing,
which was capped at the end and had holes of = 9.5 mm diameter drilled into the sides of the tube at such
positions that gas flow was taken from 0.152,0.244 and 0.337 m from the centerline of the charcoal trap. The
holes were covered with 20-mesh wire screen to exclude charcoal from the bypass lines. The samples upstream
and downstream of the charcoal trap were taken through 1.0-inch pipes which led from the main effluent
piping. The effluent flow through each of the sampling traps, amounting to 6.3 x 10-4 mVs, was led directly to
the flare-stack header. The effluent flow to be analyzed was taken from the bypass and put through the
sampling traps which collected the radiokrypton. The flow in the bypass line and in the lines leading to the
sampling traps were adjusted so that the time of flow from the large charcoal trap to the small sampling traps
would all be equal and less than 30 s.
Each sampling line was associated with the following hardware:
(1) a solenoid valve to allow the sampling flow to be saturated and stopped at a definite time;
(2) a heater to warm the cold effluent;
(3) a manual valve for flow control;
(4) a floating ball-type flow meter;
(5) a cooling coil to cool the effluent before it entered the sampling traps;
(6) the sampling traps;
(7) a heater to warm the cold effluent; and
(8) a check valve to prevent accidental backflow through the sampling traps.
-351-
-------
Each sampling trap consisted of two sections in a series. The upstream section (called FRONT) and the
downstream section (called REAR) were of the same shape and size, a right circular cylinder 0.203 m in
diameter by 0.146 m deep. They each contained 2.132 kg charcoal.
The bypass and sampling system without the flow meters was first operated during EP-II, but so little
radioactivity was found in the sampling traps that there was doubt that any flow had taken place through the
sampling system. To ensure bypass flow through the sampling traps for EP-III and succeeding runs, the
following changes were made:
(1) the 0.250-inch tubing in the liquid-nitrogen bath, immediately ahead of each sampling trap, was
changed to 0.5 inch tubing to help prevent formation of ice blockage should any water enter the system;
(2) the flow meters shown in Figure 2 were installed in the outlet line from each trap, just upstream from
the check valve; a television camera was used to view the flow meters during the reactor runs;
(3) a wet test meter was used to calibrate the flow meter before EP-III and EP-V, but not before EP-IV.
Helium had to be used for the calibrations, the response of the flow meters to hydrogen flow had to be
calculated. This computation was performed by NRTO. The calibration data were combined with the flow
meter readings observed during each reactor run to obtain the rates of flow for each sampling trap during each
experimental plan. The results are shown in Table 3.
Of necessity, the flow calibrations were carried out rather crudely in the field, which leads one to expect that
the largest source of error in the results of the measurements arises from uncertainties in the flow
measurements. Before each reactor run, the small traps were filled with North American Carbon Type G-212
activated charcoal 8 x 16-mesh, sealed, pressure-tested, and the charcoal was regenerated by a flow of hot N2
for 86.4 ks. During installation of the traps, care was taken to minimize the atmospheric water that could enter
the piping.
On the day following each reactor run, the sampling traps were removed from Test Cell C to the core sample
building at the Nevada Test Site, where the charcoal in each trap was thoroughly mixed and aliquots taken.
One aliquot was used for y-ray spectrometric measurements, and the other was sent to LASL for
radiochemical analysis of 89Sr. The y-ray spectroscopy measurements were made with a 16 x 103 mm3 Ge(Li)
detector at NRDS.
RESULTS AND CONCLUSIONS
Only the final ratios of activities in the various traps are presented below.
1. Gamma Ray Spectroscopy.
Table 4 gives the ratios of the activities of 85mKr and 88Kr in Traps 30,40, and 50 to that found in Trap 20, as
measured by y-spectroscopy. The specific y-ray peaks that were integrated for these results are the 151.1-keV
peak for 85mKr, and the 196.1-keV peak for 88Kr. The data have been corrected for radioactive decay, for
branching ratios, for sample mass, and for detector efficiency. The detector efficiency was measured with an
I.A.E.A. calibrated 22Na source imbedded in an appropriate mass of inactive charcoal. Inaccuracies in the
absolute measurement of the detector efficiency cancel in the ratios.
A potentially important source of error in these y-ray measurements results from the gradual desorption of
krypton from charcoal at room temperature while the samples are being counted. One indication that krypton
was being lost from the samples during the process of y-ray counting comes from measurements taken on
Trap 30 from EP-IV. Triplicate aliquots of charcoal from the front cannister of Trap 30 were taken to measure
the uniformity of sampling. The activity of each sample, corrected for mass and radioactive decay,
monotonically decreased with the order in which the samples were counted. The same three aliquots were
analyzed for 89Sr, and no such monotonic decrease in activity was observed. The plastic bottles that served as
containers for the charcoal during y-ray counting were not gas-tight. In Table 4 only the data from the qliquot
that was counted first are shown.
In all cases except for EP-V, the quantities of 135Xe, which was measured in the small traps, was higher in
the downstream portion of the traps than in the upstream portion. In addition, the complete absence of 135Xe in
EP-III Trap 30, and its presence in Traps 20, 40, and 50 indicated that the 135Xe might be entering the traps
other than in the normal way. Also, in EP-III the Trap-40 and Trap-50 data indicated significant quantities of
85tnKr and 88Kr in the downstream portions of these traps. The probability that the data from Trap 40 and
Trap 50 were compromised by extraneous effects is sufficiently high that no further use are made of these
data. The check valve in each bypass line was replaced before EP-IV, and the situation seemed to improve.
The check valves were replaced again before EP-V, and further improvement was noticed — although the
135Xe detected in EP-V Trap 50 indicated that some small problem remained.
2.89Sr Analysis.
The results of the 89Sr measurements are shown in Table 5. The principal source of error is likely to be
uncertainties m the actual flow of H2 through each of the sampling traps, and the problem of back-flow in the
traps. The ratios shown in parentheses for EP-III were derived by considering only the activities in the
upstream portion of the sampling trap; all other ratios are formed by summing the activities in the upstream
and downstream portion of each sampling trap.
-352-
-------
Sampling trap activities as calculated and as measured are compared in Tables 6 through 12. From such a
comparison an estimate of the effective velocity of travel of the concentration-front or, alternatively, the
effective dynamic adsorption coefficient can be obtained. Table 13 gives the estimates of the velocity factors
thus obtained. Note that the 88Kr and 85mKr results, although suspect, do not disagree seriously with the 89Sr
data.
The dynamic adsorption coefficients obtained from these ratios are plotted in Figure 1. They fell below the
measured line for the adsorption of krypton from helium onto activated charcoal. This decreased dynamic
adsorption coefficient may be caused by the hydrogen in the effluent, decreasing the amount of krypton
adsorbed.
Independent measurements of the amount of 89Kr entering and leaving the large charcoal trap were made"
during EP-IV and EP-V by NRTO under the technical direction of SNSO-N. In Table 14, the values labelled
NRTO were obtained by the analysis of gas samples taken upstream and downstream of the charcoal trap
during EP-IV and EP-V. The values labelled LASL were measured by adsorption of the 89Kr in cooled charcoal
traps as described above. All values were corrected for aliquots taken, counter efficiencies, flow inequalities,
and decay from the time of shutdown of each EP.
Although the number of 89Kr atoms in and out of the trap in EPs-IV and -V, as measured by the two methods,
are in fair agreement, differing only by at most a factor of 4; there is an anomaly in the NRTO data. Their
measurements show the same amount of 89Kr release by the trap in both experimental plans, and this makes
the calculated throughput less for EP-V than for EP-IV. Because the charcoal trap was operated during EP-V
at a higher temperature and for a longer time, this result is most probable. Perhaps the NRTO data were
compromised by a sneak source of activity as some of the LASL data were.
REFERENCES
Burnette, R. D., W. W. Graham III, and D. G. Morse, (1961), The Removal of Radioactive Krypton and
Xenon from a Flowing Helium Stream by Fixed-Bed Adsorption, Report Ga-2395.
Burnette, R. D. and D. R. Lofing, (1967), Low-Temperature Adsorption Studies for the PSC Reactor,
Report GA-7932.
Kovatch, J. L., (1970), Krypton-Xenon Adsorption Studies on Nacar Carbons, Report NAC AR-10004.
TABLE 1. Estimated NF-1 Effluent Composition.
Partial
Constituent Mole Fraction Pressure, Pa
H2 1.0 4.14xl05 TABLE 2. Dynamic Adsorption Coefficients, KD,
H20 7.09 xlO-7 4.89xlO-3 For Krypton on Charcoal.
CH4 2.78 x 10-6 1.92 x 10-2
CO 8.44 x 10-7 5.82 x 10-3
Xe 1.54xlO-7 l.OGxlO-6 1000/T(K-X) KD(mVkg) Reference
Kr 7.91 xlO-11 5.45xlO-7 __^_
3.55 0.103 (a)
3.62 0.127 (c)
3.62 0.152 (c)
4.0 0.229 (a)
4.0 0.275 (b)
4.35 0.488 (b)
4.45 0.700 (a)
4.60 1.195 (b)
5.25 5.230 (b)
5.65 18.150 (b)
5.95 135. (b)
6.60 305. (b)
7.00 940. (b)
7.50 1953. (b)
(a) Kovatch (1970).
(b) Burnette, et al, (1961).
(c) Burnette and Lofing (1967).
-353-
-------
TABLES. Hydrogen Flow Data.
TRAP 20 TRAP 30 TRAP 40
Hydrogen flow rate computed,
from wet test meter results
before EP-III
Flow gage level computed
forH2(a)
EP-III observed flow
gage level (a)
Computed H2 flow rate
Normalization factor
EP-IV observed flow
gage level (a)
Computed H2 flow rate
Normalization factor
Hydrogen flow rate computed
from wet test meter results
before EP-V
Flow gage level computed
forH2(a)
EP-V observed flow
gage level (a)
Computed H2 flow rate
Normalization factor
TRAP 50
4.18xlO-4 mVs 4.2xlO-4mVs 4.18xlO-4 mVs 3.85xlO-4 mVs
0.24
0.35
0.305
15.6
0.30 0.30 25.0
3.58x10-" mVs 4.13xlO-4 mVs 6.69xlO-4 mVs
1.000 0.866 0.535
0.3 0.2 10.0
3.58xlO-4 mVs 2.75xlO-4 mVs 2.68xlO-4 mVs
1.000 1.300 1.337
0.2
3.21xlO-4 mVs
1.115
0.10
1.60xlO-4m3/s
2.232
4.18xlO-4 mVs 4.1xlO-4m3/s 3.88xlO-4 m3/s 2.28x10-4 mVs
0.38
0.32
16.0
0.47 0.33 15.0
5.17x10-" m3/s 4.23xlO-4 mVs 3.64xlO-4 mVs
1.000 1.224 1.421
0.25
0.19
1.73x10-" m3/s
2.990
(a; Arbitrary scale.
TRAP 30/TRAP 20
TRAP40/TRAP20
TRAP 50/TRAP 20
TABLE 4. Gamma-Ray Spectroseopy Results.
EP-III
EP-IV
EP-V
4.1xlO-2 6.27xlO-2
e.Sxio-1
9.86xlO-6
88Kr
7.57X10-1
8.34X10-1 8.37X10-1
5.96X10-1 5.57X10-1
4.24xlO-i 4.27X10-1
TABLE 5. «»Sr Analysis Results.
EP-III EP-IV
EP-V
TRAP 30/TRAP 20 5.27x10-2 3.26x10-' 3.18x10-'
TRAP40/TRAP20 1.06xlO-3(l.lxlO-4) 4.42xlO-5 4 82x10-2
TRAP 50/TRAP 20 2.7xlO-3(6.5xlO-4) 7.73xlO-5 131xlO-3
-354-
-------
TABLE 6. Comparison of Computed and Measured
Sampling Trap Activity Ratios of 89Sr for EP-III.
Trap
Pair
730
720
740
720
750
720
1
10-5
10-3
10-76
2
0.001
10-25
10-68
veiocitj
3
0.006
10-22
10-63
r if actor
5
0.053
10-18
10-54
7
0.146
10-13
10-47
10
0.296
10-9
10-40
Measured
Ratio
0.053
—
—
TABLE 7. Comparison of Computed and Measured
Sampling Trap Activity Ratios of 89Sr for EP-IV.
Trap
Pair
730
720
740
720
750
720
Velocity Factor
1
10-4
10-22
10-58
2
0.004
10-16
10-so
3
0.016
10-12
10-42
5
0.133
10-7
10-30
7
0.233
10-9
10-io
10
0.326
10-4
10-14
Measured
Ratio
lO-5
10-4
TABLE 8. Comparison of Computed and Measured
Sampling Trap Activity Ratios of 89Sr for EP- V.
Trap
Pair
730
720
740
720
750
720
Velocity Factor
1
0.208
10-4
10-13
2
0.450
0.016
10-5
3
0.580
0.061
0.001
5
0.709
0.182
0.016
7
0.773
0.302
0.053
10
0.826
0.443
0.131
Measured
Ratio
0.318
0.048
0.001
TABLE 9. Comparison of Computed and Measured
Sampling Trap Activity Ratios of 88Kr for EP-III.
Trap
Pair
730
720
740
720
750
720
Velocity Factor
1
10-4
10-29
10-76
2
0.002
10-25
10-68
3
0.012
10-22
10-82
5
0.089
10-17
10-53
7
0.233
10-13
10-47
10
0.435
10-9
10-40
Measured
Ratio
0.063
-
-
-355-
-------
TABLE 10. Comparison of Computed and Measured
Sampling Trap Activity Ratios of 88Kr for EP-IV.
Trap
Pair
730
720
740
720
750
720
1
0.021
10-20
10-58
¥
2
0.212
10-13
10-46
eiuuii/j'
3
0.427
10-8
10-38
5
0.644
10-4
10-26
7
0.743
0.014
10-19
10
0.819
0.170
10-10
Measured
Ratio
0.757
-
-
TABLE 11. Comparison of Computed and Measured
Sampling Trap Activity Ratios of 88Kr for EP-V.
Trap
Pair
730
720
740
720
750
720
Velocity Factor
1 2 3 5 7 10
0.838 0.939 0.956 0.960 0.961 0.963
0.258 0.604 0.729 0.828 0.888 0.951
10-8 0.131 0.389 0.616 0.720 0.819
Measured
Ratio
0.837
0.557
0.427
TABLE 12. Comparison of Computed and Measured
Sampling Trap Activity Ratios of 85mKr for EP-V.
Trap
Pair
730
720
740
720
750
720
Velocity Factor
1 2 3 5 7 10
0.847 0.944 0.959 0.962 0.963 0.964
0.271 0.621 0.742 0.837 0.895 0.956
10-8 0.140 0.406 0.632 0.733 0.830
Measured
Ratio
0.834
0.596
0.424
-356-
-------
TABLE 13. Charcoal Trap Velocity Factors for Various Isotopes.
Velocity Factor for Isotope ^a'
Temperature, °K
161
(EP-IV)
172
(EP-III)
186
(EP-V)
[ «»Sr
M4
5
1.6
2.5
3.0
88Kr
7.2
4
1
1.6
3.2
sanKr
5
4
0.95
1.6
3.2
(a'The velocity factor is the ratio of the
Krypton-front velocity found in this work
divided by the velocity reported in
TABLE 14. 89Kr Atoms in and out of Charcoal Trap.
EP-IV (2628s) EP-V (3078s)
Tractional Fractional
Source In Out Throughput In Out Throughput
NRTO 2.22xl019 1.9xl016 0.00086 3.22xl019 1.9xl016 0.00059
LASL 8.25xl019 1.421xl016 0.00017 1.476xl020 5.77x10" 0.0039
-357-
-------
10'
CP
IC
c
O>
« icr
O
o
.2 I0
o
(/)
T)
<
O 10
E
D
C
10'
10"
2
Equation of line
k = 7.46xlO"6Exp(2637/T)
8
10
-I
IOOO/T,(K~')
12
Figure 1. Dynamic Adsorption Coefficients of Krypton on Charcoal.
14
-358-
-------
6.3xlO~4m%
FLOW
METER
\
FLOW
SOLENOID
VALVE
0.127m
0.152m
0.610m
1.07m
4 3 HEATER
3x10 mys I
-\ PS^. j« 1
HEATER
FLOW VALVE-
CHECK
!\ VALVE
STEEL BALLS
TEMP
PROBE
CHARCOAL
COOLING COILS-71CHARCOAL?
I TRAPS
I w .w A ivy i
tD-CXf
•GAS
SAMPLE
6.3xlO"4m%
4 3.3 x 10-V/
HUh-X
FINE
WIRE
SCREEN 4 3
6.3x10' m/s
3.3x10" m/s
A—J
L- LIQUID N2 BATH
NF-I CHARCOAL TRAP SAMPLING SYSTEM
Figure 2. NF-1 Charcoal Trap Sampling System.
-359-
-------
REACTOR CONTRIBUTIONS TO ATMOSPHERIC NOBLE GAS RADIOACTIVITY LEVELS*
J. M. Matuszek, C. J. Paperiello, and C. O. Kunz
Radiological Sciences Laboratory
Division of Laboratories and Research
New York State Department of Health
Albany, New York 12201
Abstract
Nuclear power reactors produce copious quantities of several species of radioactive gases. Samples of these
gaseous effluents were analyzed from two pressurized-water power reactors (PWR), one boiling water power
reactor (BWR), one high-temperature gas-cooled power reactor (HTGR), and a heavy-water-moderated
pressurized-water research reactor (HWPWR). Several noble gas activities were quantitatively identified;
these included 31, 39Ar, ssKr, and 13im, 133m, I33Xe. Direct stack sampling in the case of the BWR ensured
representative samples of actual releases. A variety ofin-plant samples were collected from the other reactors
in order to derive characteristic patterns for release levels.
Gamma-emitting gaseous species were identified by spectral resolution using a Ge(Li) detector and
associated electronics. The beta-emitters were identified by chromatographic separation on a series of
molecular sieve columns, followed by spectral analysis using internal gas-proportional counting tubes and
associated multichannel analyzer systems.
The Laboratory's experiences in sample collection, handling, and analysis are summarized. Estimates of
relative release rates and of total annual releases of the several species are provided for each type of reactor.
INTRODUCTION
Noble gas releases from nuclear facilities are usually monitored routinely by the facility operator to ensure
against an excessive external (submersion) dose to the population living in the environs affected by the
facility. Facility measurement systems are, therefore, designed to measure gamma-emitting radiogases.
Some facilities also collect water vapor from the gas stream in order to measure releases of tritiated water.
The Radiological Sciences Laboratory of the New York State Department of Health is conducting a study of
the gaseous effluents from various types of reactors, searching particularly for radiogases which were not
measured by the plant effluent monitors, but which might contribute to the total off-site dose.
Certain noble and permanent gases, particularly 37Ar and tritium, have found increasing use as
atmospheric tracers (Lai and Peters, 1967; Loosli and Oeschger, 1969; Agerter, etal., 1967). A number of these
gases are released by certain reactors in sufficient quantities to influence atmospheric measurements,
particularly if the collection locations are near the release sites.
This paper presents the activity ratios of the long-lived noble gas effluents, the estimated annual release
levels of the several nuclides of interest, and projections of potential release levels for larger reactors.
Unfortunately, no commercial fuel-reprocessing facilities were operating during the period of this study, so
only reactor effluents are discussed.
EXPERIMENTAL PROGRAM
1. Sample Collection.
Several samples have been collected from three light-water reactors: a boiling water power reactor (BWR)
and two pressurized-water power reactors (PWR). Samples were collected directly from the stack of the BWR.
The PWR samples, collected from a variety of locations, included pressure-tank (cover gas), dissolved gas in
the primary coolant (strip gas), and decay-tank gas (containment air).
Samples received from a high-temperature gas-cooled power reactor (HTGR) have consisted of the off-gas
from the regeneration of two cryogenic charcoal beds used to purify the helium primary coolant.
A set of samples has also been collected from the pressure vessel of a heavy-water-moderated pressurized-
water research reactor (HWPWR). This reactor is probably of sufficient size to provide clues toward heavy-
water power reactor performance.
Sample containers in this ongoing study vary according to the collection facilities available at each
reactor. If flow-through samples can be collected, conventional gas-sampling containers of 30 ml, 125 ml, and
1,000 ml were filled to pressures of from one to three atmospheres. Otherwise, rubber-capped septum vials are
filled through a hypodermic needle, as is conventionally done at most power reactor facilities. Containment
air is collected in a 16-liter stainless-steel container. The largest gas volume possible is collected- i. e., the
highest level the facility operator will allow off-site.
*Supported inpart by USAEC contract AT(11-1)2222 and USEPA contracts 68-01-0522 and 68-01-LA-0505.
-360-
-------
2. Chemical Processing.
Immediately on receipt of samples at the Laboratory, gamma-emitting nuclides are measured on a Ge(Li)
detector. Subsequent separations and purifications of specific radiogas species are conducted as many times
as necessary to assure consistently reproducible data. A krypton fraction is separated from each aliquot to
provide a reference point for all other nuclides separated from that aliquot. Thus, leakage of air through the
rubber septum caps does not cause analytical errors. Details of the chemical processing technique are
described in an associated paper (Kunz, 1973).
3. Measurements.
Separated fractions are counted in 100 ml stainless-steel gas-proportional counting tubes, using a plastic
anticoincidence guard for nuclides in very low abundance. Details of the measurement techniques are
provided by Paperiello (1973).
RESULTS AND DISCUSSION
Transit time to the Laboratory for most samples is at least 3 hours from time of collection. The first gamma-
spectral measurements are normally made after 30 hours have passed, so data for noble gases with half-lives
of less than 10 hours are limited. The following results are, therefore, confined primarily to those nuclides with
half-lives of greater than 1 day.
Noble gas activities on each sample are normalized on 86Kr, then averaged for presentation in Table 1. In
general, the consistency of ratios for repeated analysis of a given sample is better than ±10% when counting
statistics permit that precision. The consistency of ratios for a specific nuclide in multiple samples of the same
type from a given reactor is ± 20% if counting statistics are adequate. Fission gas ratios are consistent with
decay times for each type of reactor.
The 37Ar ratios for samples from the HTGR and HWPWR are startlingly high and consistent to ±20%. The
39Ar activity ratios, on the other hand, are very low for all reactors, so measurements are possible only on large
samples, limiting any attempts at evaluating the consistency of 39Ar results.
In Table 1, the first column under each nuclide reports the ratio for the sample at collection time. The second
column reports the values corrected as necessary for decay from time of sample collection to the average time
of release.
The amount of each noble gas released annually (Table 2) is estimated by multiplying the normalized
activity ratio by tKe published release levels of 85Kr or "total noble gas" for the two PWR's and the HTGR.
Since 85Kr is not measured by the BWR operator and the total noble gas includes many short-lived activities,
release estimates were made from reported 133Xe levels. The HWPWR reports only the amount of HTO released
annually, so all noble gas release estimates were calculated from the operator's estimates of total pressure-
vessel volume released and from the radionuclide concentrations measured at the Laboratory.
Reported values for BWR fission gas releases normally include 133Xe as the longest-lived noble gas
constituent. The data in Table 2 indicate that the other fission gases discussed here do not contribute
significantly to the local submersion dose. In particular, the calculated amounts of 85Kr released appear low
when compared to the values for other reactors. Radioargon releases from the BWR are lower than those from
the other reactors with the exception of PWR(I). The maximum release rate of 37Ar from the BWR could be as
high as l/xCi/s at a concentration of 0.01 pCi/ml. Since the total release rate can vary downward from the
maximum values as much as one order of magnitude at any specific time, BWR releases are not continuous in
the strictest sense. The mean release rate for 37Ar is 0.3/u.Ci/s.
Long-lived noble gas releases from PWR(I) are also low. Radioargon releases are lower than those for any
other reactor studied. The maximum release rate of 37Ar could be as high as 2/xCi/s at a concentration of 0.005
pCi/ml. Sample volumes were too small to obtain a real value for 39Ar releases.
Release estimates for PWR(II) must be considered in two data groups. Core elements used in 1970,1971, and
1972 were of a type found to suffer densification and leakage (Gillette, 1972). Fission gas releases were high
during these years, and releases of 85Kr were higher than those from the other light-water reactors studied.
Installation of pressurized fuel elements appears to have reduced the amount of fission gases released in 1973
by factors of 30 and 50 for 85Kr and 133Xe, respectively. The 1973 values are extrapolations of reactor
performance using data for the first six months of operation with the new core.
Releases of 37Ar from PWR(II), however, remain the highest from the light-water reactors. The change in
fuel element design has not significantly reduced them, indicating the 37Ar production likely occurs outside
the fuel elements. Although these releases are too low to contribute to the submersion dose, they may be
sufficiently large to affect atmospheric 37Ar measurements. At a PWR of this type 37Ar activity could be
released for a few hours at a rate of 12/xCi/s and a concentration of about 0.2 pCi/ml with a periodicity of 4 to 8
weeks, depending on operating conditions. Puffs could, therefore, appear periodically at locations where
atmospheric 37Ar samples are being collected.
Long-lived fission gas releases from the HTGR appear limited to 85Kr and are similar in magnitude to the
light-water reactors. The plant itself has approximately one-tenth the thermal power level of the light-water
reactors. It will be interesting to follow the trend of 85Kr release as larger HTGR's become operational. The
133Xe releases were not measurable because of the small size of the sample and its age when processed. A
second set of samples has been collected, but the data are not available at the time of this writing.
-361-
-------
If the measurements from the one sample obtained to date are representative, the release of 37Ar from the
HTGR is doubly significant. Though 85Kr is the dose-limiting nuclide during release, 37Ar contributes a 20%
additional submersion dose. It may also have a pronounced affect on atmospheric studies. Not only are
copious quantities of 37Ar released annually, but the releases occur as puffs of a few hours duration every 3 to 4
weeks. The release rate, calculated from the highest annual release level in Table 2, could be 4 mCi/s at a
concentration up to 500 pCi/s. Furthermore, the 85Kr release limit in the plant's technical specifications is 6
mCi/s. The corresponding 37Ar release level could be as high as 300 mCi/s at a concentration of 30 nCi/ml.
Fission gas releases from the HWPWR are also low and appear to consist predominantly of 133Xe. The 37Ar
release levels appear quite high; puff releases of approximately 1-hour durations would occur monthly at the
rate of approximately 5 fiCi/s and with concentrations of 0.5 pCi/ml. However, only a single sample has been
analyzed to date, and this research reactor has a 4-week operating cycle, so buildup factors may be more
sensitively related to the time of sampling within the cycle. Until more samples are processed, the HWPWR
data must be considered somewhat tenuous. Plans have been made to collect a series of samples at different
times during the operating cycle in order to better define the noble gas releases from this reactor.
The highest annual release values (Table 2) are not necessarily the maximum possible release rates. For
PWR(I) and PWR(II), the values represent only about 60% of the plant operating capacity. Unfortunately, we
have no estimate of plant capacity for the HTGR, but it seems reasonable to assume that these values
represent 60% to 70% of the operating capacity for this reactor as well. As plant capacity improves with new
reactor designs, scaling factors for noble gas releases might exhibit an increase greater than that based on
thermal power levels alone.
The mechanism for production of 37Ar and 39Ar in reactors is a matter of conjecture. Activity ratios for
37Ar/39Ar for the light-water reactor effluents are reasonable similar (Table 3), but they are different from the
air activation ratio calculated for each duty cycle. A likely source of 37Ar, particularly in the case of the HTGR,
would be via the 40Ca (n,o<)37Ar reaction from calcium impurities in reactor materials such as the carbon
moderator. The lack of 41Ar data for these samples limits the comparison to the one ratio.
A 37Ar39Ar ratio lower than the calculated for air activation could be due to an inadequate accounting of 37Ar
decay. Except for the PWR(II)'s new core and the NWPWR, the calculations were made for continuous
irradiation over the annual operating cycle. The power reactors actually made brief, periodic shutdowns,
permitting some 37Ar decay; but, to explain the low ratio in this way, one would have to postulate
approximately 100 days of shutdown each year which seems unreasonably long. A more likely explanation in
the light-water reactors is that the reaction 39K(n, p)39Ar leads to a greater abundance of 39Ar, thus decreasing
the value of the ratio below that for air activation.
The 37Ar/39Ar ratio for HTGR effluents is greater than the air activation value by a factor of approximately
1,000. Such a high ratio must come from calcium activation. The ratio for the HWPWR effluents is also
markedly different from the air activation ratio. One the other hand, the 41Ar/37Ar ratio for the HWPWR is
quite similar to the air activation value, leaving some question of the interpretation of the data. In general, the
radioargon ratios indicate that air activation is a minor source of 37Ar radioactivity in reactor effluents.
The 37Ar reactor release levels have an important relationship to atmospheric studies being conducted by
several speakers at this Symposium. Lai and Peters (1967) estimated that the average tropospheric specific
activity of 37Ar would be 2.1 x 10-2 dpm per kg of air, which is equal to 2 x 10-3 dpm per liter of argon or 1 x 10-14
fjCi/ml of air.
Using the atmospheric dispersion coefficients suggested by the U. S. Atomic Energy Commission in
recently published regulatory guides (USAEC, 1973a and b), the 37Ar concentrations at 100 and 1,000 km from
the stack can be estimated (Table 4).
It is apparent that the BWR and PWR(I) would have only a nominal effect, if any, on measurements such as
those being carried out at approximately 10-14/u,Ci/ml (Loosli and Oeschger, 1969). The short-term puffs from
PWR(II) could contribute to such low-level measurements, particularly where most samples are collected
within 150 km of a reactor identical in design to the PWR(II).
The HTGR would apparently contribute significantly to atmospheric 37Ar measurements at distances up to
a few thousand kilometers. Furthermore, if 37Ar is released at the technical specification limit, an 80-fold
increase would result for the HTGR concentrations in Table 4. Since the HTGR is about 110 km from the
National Bureau of Standards laboratory which is conducting atmospheric 37Ar measurements, its impact
there must be considered. The HTGR is relatively small compared to other gas-cooled reactors. A number of
carbon-moderated, gas-cooled reactors in Europe, particularly in France (total of 2,700 MWt) and the United
Kingdom (total of 8,000 MWt), could contribute large quantities of 37Ar to the global inventory and have a
significant impact on atmospheric 37Ar studies being performed at Bern. The newest U.S. HTGR, due to start
operations soon, may contribute up to 44 kCi/year of 37Ar. Four HTGR's now on order in the U.S. could each
release up to 130 kCi/year of 37Ar when finally operational.
The influence of the HWPWR on any of the atmospheric studies' laboratories is less pronounced. To our
knowledge, no samples of atmospheric 37Ar are being collected near the facility in which this reactor is
located. What influence other research reactors of similar design might have is a matter of conjecture. The 40-
MWt power level of this HWPWR is also very small compared to approximately 2,800 MWt of heavy-water
power reactors distributed globally. The influence of these reactors on the global inventory of37 Ar should also
be examined.
Extrapolation of noble gas releases for reactors of different design — or even for different reactors of the
same design — cannot be quantitatively exact. It seems fair to expect, however, that within a scaling factor of
approximately 100 for total thermal power, there is a significant probability of perturbations on atmospheric
37Ar levels by the several HTGR's currently operating and especially by those being constructed.
-362-
-------
SUMMARY
Noble gas releases from several types of reactor have been measured, and annual releases of the long-lived
constituents have been estimated. Of serious import to the atmospheric physicists are the copious quantities
of 37Ar produced in some of the reactors, particularly the HTGR. For those involved in atmospheric research,
the important concern is the even greater unmeasured releases from many larger gas-cooled reactors.
REFERENCES
Aegerter, S. K., H. H. Loosli, and H. Oeschger, (1967) Variations in the Production of Cosmogenic
Radionuclides, Symposium on Radioactive Dating and Methods at Low-Level Counting. International Aomic
Energy Agency, Vienna, pp 49-51.
Gillette, R., (1972) Nuclear Safety: Damaged Fuel Ignites a New Debate in AEC, Science 177,330.
Kunz, C. O., (1973) Separation Techniques for Reactor-Produced Noble Gases, this Symposium.
Lai, D. and B. Peters, (1967) Cosmic Ray Produced Radioactivity on the Earth, Handbuch der Physik
XLVI/2,551.
Loosli, H. H. and H. Oeschger, (1969) 37Ar and 81Kr in the Atmosphere, Earth and Planetary Science
Letters 7,67.
Paperiello, C. J., (1973) Internal Gas-Proportional Beta-Spectrometry for Measurement of Radioactive
Noble Gases in Reactor Effluents, this Symposium.
U. S. Atomic Energy Commission, (1973a) Regulatory Guide 1.3, Assumption Used for Evaluating the
Potential Radiological Consequences of a Loss of Coolant Accident for Pressurized Water Reactors.
U.S. Atomic Energy Commission, (1973b) Regulatory Guide 1.4, Assumption Used for Evaluating the
Potential Radiological Consequences of a Loss of Coolant Accident for Boiling Water Reactors.
TABLE 1. Average Activity Ratios for Long-Lived Noble Gas
Reactor Effluents at Time of Measurement and Time of Release.
Reactor
Type
BWR
PWR(I)
PWR(II)(a)
PWR(II)(b)
HTGR
HWPWR
Notes:
(a) Old core.
(b) New core.
*'<
Meas.
8.5xlO-2
3.2xlO-3
3.9xlO-3
1.0x10-'
49
19,000
&r
Rel.
8.5xlO-2
S.lxlO-3
3-lxlO-3
8xlO-2
49
19,000
»BAr
Meas. Rel.
l.lxlO-4
<4xlO-6
-------
Table 2. Highest Annual Releases of Long-Lived Noble Gases.
Reactor
BWR
PWR(I)
PWR(II)(a)
PWR(IIXb)
HTGR
HWPWR
Power
Level
(MWt)
1,538
615
1,300
116
40
Year
1972
1970
1971
1973
1971
37Ar
3
1.7X10-1
12
7
6,000
1.9X10-1
Kt
39Ar
SxlO-3
<2xlO-4
<4xlO-3
SxlO-3
5xlO-4
2xlO-4
itimated Keieases (
ssKr isimxe
31 260
53 20
3,700
90 7
122
9xlO-6
\^i)
issmxe 133Xe
2,900 65,000
50 1,700
23,000
1 450
<4
1.4x10-!
Notes:
(a) Old core.
(b) Based on first six months of operation with new core.
TABLE 3. Comparison of Radioargon Production Ratios to Air Activation Ratios.
Source
«Ar/37Ar
Annual duty cycle:
Air activation
BWR
PWR(I)
PWR(II)(a)
HTGR
Monthly duty cycle:
Air activation
HWPWR
PWR(II)(b)
l.SxlO4
9xl02
>1.6xl03
> 8xl03
8xl07
9x10"
l.lxlO3
1.7xl03
Notes:
(a) Old core.
(b) New core; sample collected one month after start-up.
58
50
TABLE 4. 37Ar Concentrations (jttCi/ml) at 100 and 1,000 km from the Stack.
Reactor
BWR
PWR(I)
PWR(II)
HTGR
HWPWR
Release
period
Continuous
= 1 week
= 8 hours
= 8 hours
1 hour
at Stack
1 x 10-8
5 x 10-9
2 x 10-7
5 x 10-4
5 x 10-7
100km
distant
2 x 10-16
1 x 10-16
1 x 10-12
3 x 10-10
1 x 10-12
1,000km
distant
2xlO-17
1 x 10-17
1 x 10-13
3 x 10-11
1 x 10-13
-364-
-------
EFFECT OF HEAT PRODUCED BY RADIOACTIVE DECAY ON THE ADSORPTION
CHARACTERISTICS OF CHARCOAL BEDS*
G. G. Curzio1 and A. F. Gentili2
institute Impianti Nucleari
University of Pisa
Pisa, Italy
2Servizio di Fisica Sanitaria
Camen, S. Piero a Grado
Pisa, Italy
Abstract
The effects of radioactive decay heat from the fission-produced noble gases on the characteristics of an
activated char coal holdup bed, and related problems, were studied.
A computer code currently being developed for the evaluation of the radioactive decay heat effects;
distortions in elation curves under various conditions; the formation of hot spots; the appearance of
preferential paths; and risks of ignition is mentioned.
An experimental facility was set up to test the behavior of a small charcoal bed. The facility and the adopted
method are briefly described and the experimental results are reported.
Apreliminary comparison is made between the experimental results and the theoretical ones.
INTRODUCTION
The dynamic characteristics of an activated charcoal bed for use in the delay and retention of the
radioactive noble gases, i.e., using the adsorption coefficient and the parameters of the Van Deemter equation
for the evaluation of the reduced height of theoretical plates, can easily be evaluated in ideal work conditions
(Underbill, 1967; Kovach, 1970; and Curzio and Gentili, 1972a and b).
However, many factors can modify the ideal behavior of an activated charcoal bed; such as, the presence of
moisture or other impurities in the carrier gas, irregular packing of the charcoal, vibrations, transient
hydrodynamic and/or thermal conditions, and radioactive decay heat, (Kovach 1970; Curzio and Gentili,
1972b; and Romberg 1964).
Radioactive decay heat can be of considerable interest in beds of a small size designed to treat high
activities of noble gases: In the literature there are several theoretical approaches to this problem (Kovach,
1970; Underbill, et al, 1971; Glueckauf, 1958; and Shields and Davis, 1970). But, as far as we know, no direct
systematic experimental approach exists. This is probably due to the practical difficulties involved in
obtaining measurable variations of temperature from decay heat; this requires the use of hundreds of curies or
more of radioactive gas, which creates substantial safety problems.
The only experimental data (Kovach, 1970 and Glueckauf, 1958) were obtained by injectionof a 10 W pulse of
85Kr into a charcoal bed the size of a gas chromatographic column; a drastic reduction was obtained in the
delay time as well as an increase in the number of the theoretical plates of the bed. No quantitative theoretical
correlation was attempted to obtain the relevant parameters of the phenomenon. This method can meet with
considerable difficulties, however, not only in the setting up of a test, but also in the interpretation of the
results if the high inlet activities are due to such quantities of radioactive noble gas that the linearity limits of
the adsorption isotherm are surpassed, and/or the adsorption heat becomes no longer negligible.
In the normal work conditions of a delay bed, adsorption heat can be ignored, but its effects can become
predominant if a strong concentration gradient should move through a filter.
We can make an approximate quantitative comparison between the thermal effects of decay heat and
adsorption heat produced by a bell shaped pulse of 1 Ci of 85Kr moving through a i cm2 cross section charcoal
bed.
Figures 1 and 2 illustrate the problem. The distribution of 85Kr along the bed is
n(z,t) = ( n*/\|2rr
-------
This fact suggests that it is possible to employ the heat production by adsorption for determining the
parameters that characterize a charcoal bed designed for highly radioactive noble gases.
Qualitatively we can expect that the krypton at the maximum of the pulse will pass through the bed faster
than the krypton at the front and tail of the peak: therefore, the mean delay time will be shorter and the peak
will assume an asymmetric shape.
EXPERIMENTAL
The feasibility of the suggested method was tested using the experimental setup shown in Figure 3; this
consisted of a cylindrical bed, 155 mm long and 20 mm in diameter, using 83-90 mesh activated charcoal and
dry air as the carrier gas. The thermostatic bath used ice.
The tests were made by injecting krypton pulses into the bed and analyzing the elution curves. Comparisons
were made between tests performed with small amounts of stable krypton ( = 4 x 10-3 cm3 (s.t.p.) traced with
85Kr (= 70 mCi/cm3), and those performed with up to 60 cm3 (s.t.p.) of stable krypton with a specific activity
(down to 1 fiCi/cm3). The upper limit of the amount of krypton was selected in such a way that the linearity
limit of the adsorption isotherm of the krypton on the charcoal was not surpassed (Kitani and Takada, 1965).
RESULTS
Figure 4 shows two elution curves obtained at the same temperature and carrier gas flow rate; it can be seen
that the test performed with the macroscopic amount of krypton has a delay time considerably shorter than
the test done with krypton in the tracer quantity, and the elution curve appears very asymmetric. From the
value of the time Tmax corresponding to the maximum of the elution curve, an effective adsorption coefficient
K' can be calculated.
The results of the preliminary tests are summarized in Figures 5 to 7. In the first two figures, the ratio
between K' and the true adsorption coefficient at O°C(K0) is shown first as a function of the inlet amount of
krypton at constant carrier gas flow rate, and then as a function of the flow rate at a constant inlet amount of
krypton. In the third figure, an increase in asymmetry is clearly due to an increase in the amount of inlet
krypton.
DISCUSSION AND CONCLUSIONS
For a correct interpretation of the reported results, we plan to employ an IBM 360/67 FORTRAN code,
which is about nearly completed. At present, it can be said that the code was successfully tested only in simple
situations where either a comparison with analytical solutions was possible (e. g., the determination of the
temperature profile due to a moving heat source with a known and constant shape, [Wilson, 1904]) or where
drastic approximations were made.
In the near future, the code will be completed, and on the basis of the predictions that will be obtained, we
shall be able to plan further tests under meaningfully different experimental conditions. It would then be
possible to assure a feasible extension of the results from our laboratory scale to a full-scale charcoal bed.
At any rate, the experimental results are already in satisfactory agreement with those calculated. The delay
time can presently be calculated by the computer code with an approximation better than 20%. In Figure 8, for
example, an experimental elution curve is compared with a calculated curve, which was obtained by
neglecting all the heat transfer mechanisms; the calculated delay time is about 15% less than the measured
one, and the calculated curve profile is sharper than the experimental one.
The introduction of the conduction mechanisms in the model of calculation will have as an effect the
decrease of the maximum temperature rise (and, therefore, an increase of the calculated delay time), and the
smoothing of the temperature profile (and therefore, of the elution curve).
-366-
-------
NOTATIONS
Ko — adsorption coefficient (at O°C)
K'—effective adsorption coefficient
n — molar density of krypton
n*_ number of injected moles
q A — molar adsorption heat
Q^—adsorption heat production per unit of time and volume
Q j) — decay heat production per unit of time and volume
t—time
Tmax — time corresponding to the maximum of the elution curve
v —velocity
z — axial coordinate
ATj — time necessary for outlet concentration to rise from the first half maximum value of the elution curve to
the maximum
AT2 — time necessary for outlet concentration to decrease from the maximum of the elution curve to the
second half maximum value
A9 —temperature rise
cr— standard deviation
REFERENCES
Curzio, G. G. and A. F. Gentili, (1972a), Noble Gas Adsorption Characteristics of Charcoal Beds: Van
Deemter's Coefficient Evaluation, Anal. Chem. 44,8,1544-45.
Curzio, G. G. and A. F. Gentili, (1972b), Liberation de gas nobles par les centres nucleaires: quelques
remarques sur le fonctionnement des flitres de charbon de bois, VI Congres International, Societe Francaise
de Radioprotection, Bordeaux 27-30 March 1972.
Glueckauf, E., (1958), Gas Chromatography, pp 69-89.
Kitani, S. and J. Takada, (1965), Adsorption of Krypton and Xenon on Various Adsorbents, J. Nucl. Sci.
Technol., 2(2)51-56.
Kovach, J. L., (1970), Review of Krypton-Xenon Adsorber Design, Part I, North American Carbon,
NACAR 010005.
Romberg, E., (1964), The Effect of Moisture on the Adsorption of Krypton on Charcoal, Dragon Project
Report (D.P.R.) 276.
Shields, R. P. and R. J. Davis, (1970), Ignition of Charcoal Adsorbers by Fission-Product Decay Heat,
ORNL-TM-3122, Oak Ridge National Laboratory, Oak Ridge, Tenn.
Underbill, D. W., (1967), Dynamic Adsorption of Fission-Product Noble Gases on Activated Charcoal,
NYO-841-8, New York University, New York, N. Y.
Underbill, D. W., H. Yusa and O. Grubner, (1971), Design of Fission Gas Holdup Systems, USAEC
Contract-AT(SO-l) 841, New York University, New York, N.Y.
Wilson, H. A., (1904), On Convection of Heat, Proc. Ceb. Phil. Soc. 12,406-423.
-367-
-------
5.10
-4
4.
o
8 3.
eo
CO
o
(5
o
O
2.
1.
-2
DECAY HEAT
SOURCE
TEMPERATURE
PROFILE
1 0 1
z-vt (cm)
A9
.015
.01
.005
Figure 1. Decay heat source and related temperature profile.
-------
-2
4.10
3.
w
05
tO
-2.
d -a
-4.
TEMPERATURE
PROFILE
0 12
z-vt (cm;
ADSORPTION
HEAT SOURCE
1.
.8
•6 AG
•4
.2
0
Figure 2. Adsorption heat source and related temperature profile.
-------
W
•q
o
Flow path during a test
Auxiliary branches
Figure 3. Test circuit flow-sheet.
1 - Dry air container
2 - Stable krypton container
3 - Regulation valve
4 - Intercept valve
5 - 85Kr introduction device
6 - Thermostatic bath
7 - Charcoal bed
8 - Measuring box
9 - 3" x 3" Nal (Tl) scintillator
10 - Analyzer
11 - Flowmeter
-------
15OO
03
-J
2 iooo
LU _
o z
o .
O 03
DC
I- <
500-
• Reference elution
curve
(104 cm3 of KrSTP)
° Elution curve of
68 cm3 Kr STP
80 120
TIME (sec)
160
200
Figure 4. Comparison between two elution curves.
-------
CO
3
20 40 60 80
INJECTED VOLUME (cm3 at STP)
Figure 5. Apparent adsorption coefficient as a function of the injected krypton volume.
-------
09
<1
W
8 h
JS1
Ko
•6h
.1
i i i i i i i l i i i I I I I I I I 1 1 1 1—I I ' I
1 10
FLOW RATE (cm3 sec'1)
Figure 6. Apparent adsorption coefficient as a function of the flow rate.
100
-------
1
.8
W
<1
*>.
ATi
AT2
•6
•4
.2
0
20 40 60 80
INJECTED VOLUME (cm3 at STP)
Figure 7. Asymmetry coefficient as a function of the injected krypton volume.
-------
Experimental
Calculated
40 80 120
TIME (sec)
Figure 8. Comparison between experimental and calculated elution curves.
160
-375-
-------
VI Chemistry of Noble Gases
-------
CHEMICAL METHODS FOR REMOVING XENON AND RADON FROM CONTAMINATED
ATMOSPHERES*
L. Stein
Argonne National Laboratory
Chemistry Division
Argonne, 111. 60439
Abstract
A number of solid reagents have been shown to react spontaneously with radon and xenon at 25°Ctoform
nonvolatile compounds. These appear very promising for such purposes as purifying uranium mine
atmospheres, reducing emissions of xenon radioisotopes from nuclear power plants and fuel reprocessing
plants, detecting failed fuel elements, and analyzing radon and xenon isotopes in air. Radon can be collected
with halogen fluoride-metal fluoridecomplexes,suchasClF2SbF6,BrF2SbF6,BrF4Sb2Fll,andIF4(SbF6)3;
with the dioxygenyl salt O2SbF(>; and with the fluoronitrogen salts A^FS&Fg and N2F3Sb2Fn. Xenon can
also be collected with O^SfrFg and tyFSbFe. All of these compounds are decomposed by water vapor;
therefore, they must be used in conjunction with desiccants, such as Drierite, silica gel, or molecular sieve.
Methods of preparation of the compounds and reactions of the compounds with radon and xenon will be
described. Chemical separations of krypton and xenon and laboratory-scale decontamination experiments
with samples of air containing radon-222 and xenon- 133 will also be described. The possibility of finding solid
oxidants for krypton (antimony salts, for example, that would be capable of forming the complex
KrF+Sb2FjJ and liquid oxidants for radon, xenon, and krypton that would be suitable for use in spray towers)
are discussed.
INTRODUCTION
A number of physical methods for controlling noble gas emissions from nuclear power plants and fuel
reprocessing plants have been tested in recent years. These include cryogenic distillation (Bendixsen and
German 1971; Bendixsen, et al., 1971; and Wilson and Taylor, 1958), charcoal adsorption (Wirsing, et al., 1970;
and Burnette, et al., 1962; Browning, et al., 1959 and 1960; Mecca, et al., 1971; and Slavsky, 1971), solvent
extraction (Stephenson, et al., 1972 and Merriman, et al., 1968), and permselective-membrane diffusion
(Rainey, et al., 1968 and 1971). However, very little attention has been directed to the possibility of using
chemical methods to trap noble gases. This is clearly an oversight, inasmuch as compounds of the heavy gases
(krypton, xenon, and radon) have been known for more than 10 years, and isotopes of these gases are involved
in several environmental problems. For example, krypton-85, xenon-133, and xenon-135 are the chief fission
gases released by boiling-water reactors, and radon-222, a member of the uranium-238 decay chain, is a major
atmospheric contaminant in underground uranium mines. Xenon forms many chemical compounds,
including halides, oxides, oxyfluorides, perxenate salts, and complex fluorides with alkali metal and
transition metal fluorides (Bartlett, 1962; Hyman, 1970; and Malm and Appelman, 1969). Krypton and radon,
however, appear to form only a small number of simple and complex fluorides (Hyman, 1970; Malm and
Appelman, 1969; Turner and Pimentel, 1963; Streng, et al., 1963; Selig and Peacock, 1964; Frlec and Holloway,
1973; Gillespie and Schrobilgen, Unpub.; Prusakov and Sokolov, 1971; Fields, et al., 1962; and Stein, 1969,
1970a,b, 1972, and 1973a,b). Research on radon chemistry is somewhat limited by the fact that no stable
isotopes of radon are known; therefore, most research on the chemistry of this element has been performed by
tracer methods, with microcurie and millicurie amounts of radon-222. In this paper, some initial tests of
laboratory-scale scrubbing units containing liquid bromine trifluoride, solid complexes of halogen fluorides
and metal fluorides, and the dioxygenyl salt O2SbFe are described. These tests have shown that radon can be
removed from air efficiently by oxidation with bromine trifluoride and solid complexes containing the cations
C1F2, BrF2, BrF4, and IFg, and that radon and xenon can both be removed efficiently by oxidation with
02SbF6.
Bromine trifluoride reacts spontaneously with radon at 23-25°C to form solutions of nonvolatile radon
fluoride (Stein, 1969). The oxidized radon is present as a positive ion in these solutions, since it migrates to the
cathode in a D. C. field (Stein, 1970a). The oxidation and the interaction with the solvent are believed to occur
as follows:
Rn + BrFg (large excess)-*- solution of RnF? + BrF
RnF2 + BrF3^rRnF++ BrF.
RnF+ + BrF^rRn9"1" + BrF
* Work performed under the auspices of the U. S. Atomic Energy Commission.
-376-
-------
Once it is oxidized, radon behaves very much like a metallic element. The solutions can be exposed to the
atmosphere, can be poured from one container to another, and can be distilled to dryness under vacuum
without loss of radon. (In the last instance, a deposit of solid radon fluoride is obtained.) This behavior is
somewhat analogous to that of hydrogen, since hydrogen is also very volatile in its elemental state, but only
slightly volatile in its oxidized state (i.e., as a solvated proton in aqueous solutions).
Figure 1 shows the first scrubber unit that was tested with liquid bromine trifluoride and samples of air
containing 2.4 to 4.5 mCi/1 of radon-222 (Stein, 1972). The contaminated air was passing through two Kel-F
test tubes in series, each containing 4.0 ml of bromine trifluoride at 23-25°C, and through a trap which was
cooled with liquid nitrogen to condense any unreacted radon. Flow rates of 2.9 to 3.0 ml/min were used, and the
air was dispersed as very fine bubbles by a stainless steel frit in the down-leg of each tube. At the conclusion of
each test, the distribution of radon was determined by measuring they -emission of its daughters (lead-214
and bismuth-214) after several hours, when radon and its daughters were known to be in radioactive
equilibrium. All of the radon was found in the first test tube each time; none was found in the second tube or in
the cold trap.
Since the concentrations of radon were very high and the flow rates were very low in these tests, further tests
were carried out with samples of ordinary air (not artificially contaminated with radon) with the apparatus
shown in Figure 2 (Stein, 1973a). This was slightly more complicated than the first apparatus because a much
more sensitive scintillation counting method was used. The air was passed through Ascarite, Drierite, and two
tubes of bromine trifluoride, through a trap at -80°C and a bed of soda lime, to remove BrFs, BrF, and Br2
vapors, and through a trap at -195°C to condense any remaining radon. The radon was then vacuum-distilled
into an ampoule containing frozen scintillator solution (together with xenon carrier gas), and the ampoule was
sealed and counted in a low-background scintillation counter. Figure 3 shows the type of e(-particle pulse-
height spectrum that was obtained for radon-222 and its short-lived daughters, polonium-218 and polonium-
214.
It was found that bromine trifluoride removed 76-95% of the radon from ambient air at flow rates of 130-740
ml/min — the highest percentage removal being observed at the lowest flow rate. The concentration of radon
in the ambient air (measured by a condensation-scintillation counting method) ranged from 0.098 to 0.189
pCi/1 during the course of these experiments, whereas the concentration in the scrubbed air ranged from 0.009
to 0.025 pCi/1.
The results of the two series of tests indicate that radon can be removed from air efficiently by the oxidation
process over a wide range of concentrations — pCi/1 to mCi/1. It is only necessary for the air to be well
dispersed in the liquid phase, and to remain in contact with the liquid long enough for oxidation to occur.
Bromine trifluoride reacts violently with water and organic materials, and may, therefore, be too hazardous
a reagent for large-scale use; other types of oxidants may be more acceptable from the standpoint of safety.
Solutions of complex metal ions in unusually high valence states appear very promising for further study, as
these are frequently nonvolatile and reactive. It has been shown that radon can be oxidized by solutions of
K2N1F6 in liquid HF, for example, but liquid HF is unsatisfactory as a solvent because of its high vapor
pressure and toxicity. If suitable nonaqueous solvents can be found, it may be possible to oxidize radon safely
and conveniently with solutions of K2NiFg, KsNiFe, Cs2CoF6, KsCuFe. CsAgF4, Cs2AgF6, and similar
fluoro-salts.
Figure 4 shows one type of system that could be developed in the future for purifying and recirculating the
air in a section of a uranium mine, such as a working stope, where the radon radiation problem is generally
most acute. This would include the following components: dust filter, blower, air drier, radon absorption tower,
vapor trap, and liquid circulating pump. Radon-daughters suspended in the air on particulate matter would be
chiefly removed by the dust filter, and radon would be removed by oxidation in the absorption tower (spray
tower). Some lead or concrete shielding might be needed around the absorption tower, because radon and its
daughters would build up to high concentrations in the recirculating liquid. One of the major problems would
be that of adequately drying the air to prevent hydrolysis of the radon oxidant. (All of the oxidants that are
known at present are easily hydrolyzed). This is not an insurmountable problem, however; refrigeration- or
desiccant-type drying units similar to those now used by some chemical industries could probably be
developed for this purpose. It has been estimated that this air-purification system should have a flow capacity
of at least 5,000 cfm to appreciably lower the radon concentration in a single stope and that it should have a
greater capacity, 15,000 cfm or more, to lower the concentration in several stopes.
Table 1 lists a number of solid complexes of halogen fluorides and metal fluorides and also some simple
fluorides that have been tested as oxidants for radon (Stein, 1972). The solid complexes are of two types: those
formed by alkali metal fluorides, KF.RbF, and CsF; and those formed by Group (V) metal fluorides, SbFs,
TaFs, and BiF^. The latter are better oxidants for radon than the former, since they contain halogen fluoride
cations, BrF2, C1F2, BrF4, and IFg, which can react with radon as follows:
Rn + BrF *SbF "-»~RnF+SbF ' + BrF
^ O-4- D
Rn + BrFSb F n'^RnFSbF6' + BrF2SbFg
Rn + IFg+Sb3 F 16-VRnF+SbFg
-377-
-------
The solid complexes are less hazardous to handle than liquid halogen fluorides, and can be used conveniently
in packed bed reactors to remove radon from air. However, the air must still be dried, because the complexes
are easily hydrolyzed by water vapor.
None of the halogen fluorides or solid complexes listed in Table 1 have sufficient oxidizing power to react
with xenon at 23-25°C. However, dioxygenyl hexafluoroantimonate (O2SbFe) — a very unusual salt
containing the O2 cation — has been found to react with both radon and xenon at 25°C (Stein, 1973a). Raman
spectral analyses have shown that the xenon product is XeF Sb2Fn,a 1:2 xenon difluoride (antimony
pentafluoride complex). Mass spectrometric analyses have also shown that two molecules of oxygen are
released for each atom of xenon absorbed as follows:
Xe + 2 O2+SbFg -*- XeF^Sb^ p 2O2
No spectral or analytical data have been obtained for the trace amounts of radon product, but radon probably
forms an analogous 1:2 radon difluoride — antimony pentafluoride complex as follows:
The dioxygenyl salt can be prepared by several methods. At Argonne, it has been prepared by the
photochemical reaction of oxygen, fluorine, and antimony pentafluoride, which was first reported by Shamir
and Binenboym (1968). At the research laboratory of the Ozark-Mahoning Company, Tulsa, Oklahoma, it
has been prepared by a high-pressure bomb reaction of oxygen, fluorine, and antimony pentafluoride (Beal,
et al., 1969). And at Bell Laboratories, Murray Hill, New Jersey, it has been prepared by a similar bomb
reaction of oxygen, fluorine, and antimony metal (Edwards, et al., 1973). All three products appear to be
identical. An intermediate product, O2Sb2Fn, has been obtained in some instances, but it has been shown
that this can be converted to C^SbFg by. further reaction with oxygen and fluorine. (O2SbFfi cannot be
distinguished from O2Sb2Fn. Dv appearance; both compounds are white powders at 25°C. However, the
compounds can be distinguished by their characteristic Raman spectra (Bartlett and McKee, 1973).
O2SbF6 has been shown to remove xenon-133 and radon-222 from contaiminated air very efficiently in
laboratory-scale experiments (Stein, 1973b). These were performed with metal and glass vacuum lines
similar to the one shown in Figure 1. In each experiment, a noble gas-air mixture was passed through a glass
U-tube packed with 02SbF6 powder at 23-25° C then through a trap cooled with liquid nitrogen to condense
any unreacted radioisotope. The distribution of the radioisotope was afterwards determined by measuring
the Y- emission of the U-tube and the cold trap. (The distribution of xenon-133 was determined immediately;
the distribution of radon-222 was determined after 3 hours, when radon-222 and its daughters (lead-214 and
bismuth-214) were known to be in radioactive equilibrium). In 5 experiments with the xenon isotope and with
a bed of powder 6.5 cm long and 5.5 mm in diameter, 67-100% of the xenon was removed. In three experiments
with the radon isotope and with a bed of powder 5.0 cm long and 6.3 mm in diameter, all of the radon was
removed. The initial concentrations of radioisotopes ranged from 2.0 to 9.8 mCi/1 (xenon-133), and 13 to 24
mCi/1 (radon-222); flow rates ranged from 12 to 15 ml/min.
Because it has negligible vapor pressure at 25°C and produces oxygen as the gaseous product, C^SbFg
appears to be a very "clean" reagent for air-purification purposes. It is less corrosive than halogen fluoride
and metal fluoride complexes, and can be stored for long periods in glass or fluorinated plastic containers.
Very little decomposition has been noted, for example in samples that have been stored for 6 to 9 months in
Pyrex bulbs and Kel-F test tubes. Some corrosion of Kel-Fhas been observed, however, by products formed in
the reaction with xenon — probably excited xenon, oxygen, or ozone species. The dioxygenyl compound is
decomposed by moisture; in treatment of wet gases, it must, therefore, be used in conjunction with a
desiccant, such as silica gel or a molecular sieve.
O2J3bF6 and other solid oxidants can probably be used for such purposes as removing xenon isotopes from
reactor and reprocessing plant off-gases; reducing radon concentrations in uranium mines (Argonne Study,
Ongoing); separating xenon from lighter noble gases; and analyzing xenon and radpn isotopes in air. In one
of the methods that has been developed to control emissions from boiling-water reactors, large beds of
charcoal are used as chromatographic columns. Waste gases are passed through the beds at ambient
temperature, and the noble gases are retained long enough for short-lived isotopes to decay. Xenon isotopes
may be held for 20 days, for example, and krypton isotopes for about a day. Small beds of O2SbFe can
probably be substituted for the large (25-50 ton) beds of charcoal to remove xenon isotopes from the off-gases
completely).
Following a suggestion of Liebman (1973) two fluoronitrogen salts, N2F3Sb2Fn and N2FSbF6, have been
prepared by thermal reactions of antimony pentafluoride with tetrafluorohydrazine (Ruff, 1965 and 1966),
and difluorodiazine (Ruff, 1966; Roesky, et al., 1966; and Pankrator and Savenkova, 1968), respectively, and
have been tested as oxidants for radon, xenon, and krypton. The first reacts only with radon, whereas the
second reacts with both xenon and radon at 25-100°C. Mass spectrometric analyses have shown that in the
second reaction, one molecule of nitrogen is liberated for each atom of xenon absorbed as follows:
Xe+N2F+SbF~-»-XeF+SbF6~ + N 2
-378-
-------
Because nitrogen is liberated rather than oxygen, this reaction may not be as suitable for air-purification
purposes as the reaction of xenon with O2SbF6-
Thus, far, no solid or liquid reagent with sufficient oxidizing power to capture krypton has been found.
However, krypton is known to form a stable complex fluoride, KrF+ Sb2Fn~, in the reaction of krypton
difluoride with antimony pentafluoride (Selig and Peacock, 1964). (The krypton difluoride must first be
prepared by an excitation method, such as the reaction of krypton and fluorine in an electric discharge at
low-temperature (Streng, etal., 1963).) By substituting cations of very great oxidizing power for O2+andN2F+
in the complex antimony salts, it may be possible to prepare reagents which will combine directly with
krypton to form KrF+Sb2Fn~ If so, these should be useful for capturing krypton-85.
REFERENCES
Bartlett, N., (1962), Xenon Hexafluoroplatinate (V), e PtF6, Proc. Chem. Soc., 218.
Bartlett, N., and D. E. McKee, (1973), private communication.
Beal, J. B., Jr., C. Pupp and W. E. White, (1969), The Reaction of Oxygen Difluoride with Some Lewis
Acids: The Preparation of Dioxygenyl Salts, Inorg. Chem., 8, 828.
Bendixsen, C. L. and F. O. German (1971), Operation of the ICPPRare Gas Recovery Facility During
Fiscal Year 1970, USAEC Report ICP-1001, Allied Chemical Corporation, Oct 1971.
Bendixsen, C. L., G. F. Offutt and B. R. Wheeler, (1971), Cryogenic Rare Gas Recovery in Nuclear
Fuel Reprocessing, Chem. Eng., New York, Oct. 4, 1971, pp. 55-57.
Browning, W. E., Jr., R. E. Adams and R. D. Aekley, (1960), Removal of Volatile Fission Products
from Gases, USAEC Report TID-7610, Oct. 1960, pp. 44-57.
Browning, W. E., Jr., R. E. Adams and R. D. Aekley, (1959), Removal of Fission Product Gases from
Reactor Off-Streams by Adsorption, USAEC Report CF59-6-47.
Burnette, R. D., W. W. Graham III andD. C. Morse, (1961), The Removal of Radioactive Krypton and
Xenon from a Flowing Helium Stream by Fixed-Bed Adsorption, Proc. Second Conf. Nucl. Reactor Chem.,
Gatlinburg, Tenn., Oct. 10-12, 1961, USAEC Report TID-7622, (1962), pp. 218-235.
Edwards, A. J., W. E. Falconer, J. E. Griffiths and W. A. Sunder, (1973), Dioxygenyl Salts, paper
No. 1-29, 7th Int. Symp. on Fluorine Chem., Santa Cruz, Calif., July 15-20, 1973.
Fields, P. R., L. Stein and M. H. Zirin, (1962), Radon Fluoride, J. Amer. Chem. Soc., 84, 4164.
Frlec, B. and J. H. Holloway, (1973), New Krypton Difluoride Adducts, J. Chem. Soc., Chem.
Commun., 370.
Gillespie, R. J. and G. J. Schrobilgen, unpublished studies of krypton difluoride complexes.
Hyman, H. H., (1970), Rare Gas Compounds, in Physical Chemistry, Vol. 5, H. Eyring, D. Henderson,
and W. Jost, editors, Academic Press, New York, 1970, pp. 589-662.
Liebman, J. R., (1973), Oxidant for Trapping Atmospheric Radiokrypton, Nature, 244, 84.
Malm, J. G. and E. H. Appelman, (1969), The Chemical Compounds of Xenon and Other Noble Gases,
At. Energ. Rev. 7, 3.
Mecca, J. E., J. D. Ludwick and E. L. Etheridge, (1971), Noble Gas Confinement Study., Vol. II.
Cryogenic Porous Media Bed. Final Report, FY-1970, USAEC Report DUN-7221, Vol. 2, Feb. 1971.
Merriman, J. R., J. H. Pashley, K. E. Habiger, M. J. Stephenson and L. W. Anderson, (1968),
Concentration and Collection of Krypton and Xenon by Selective Absorption in Fluorocarbon Solvents, in
Treatment of Airborne Radioactive Wastes, Int. At. Energ. Agency, Vienna, 1968, pp. 303-313.
Pankratov, A. V. and N. J. Savenkova, (1968), Fluorodiazonium Compounds, Zh. Neorg, Khim., 13,
2610.
Prusakov, V- N. and V. B. Sokolov, (1971), The Krypton Difluoride-Bromine Pentafluoride Binary
System, Zh. Fiz. Khim., 45, 2950.
Rainey, R. H., W. L. Carter, S. Blumkin and D. E. Fain, (1968), Separation of Radioactive Xenon and
Krypton from other Gases by the Use of Permselective Membranes, in Treatment of Airborne Radioactive
Wastes, Int. At. Energ. Agency, Vienna, 1968, pp. 323-342.
Rainey, R. H., W. L. Carter and S. Blumkin, (1971), Completion Report: Evaluation of the Use of
Permselective Membranes in the Nuclear Industry for Removing Radioactive Xenon and Krypton from
Various Off-Gas Streams, USAEC Report ORNL-4522, April 1971.
Roesky, H. W., O. Glemser and D. Bormann, (1966), On the Preparation and Some Reactions of
Difluorodiazine, Chem. Ber., 99, 1589.
Ruff, J. K., (1965), The Reaction of Antimony (V) Fluoride with Tetrafluorohydrazine, J. Amer. Chem.
Soc., 87, 1140.
Ruff, J. K., (1966), The Reaction of Antimony (V) Fluoride with Tetrafluorohydrazine and
Difluorodiazine, Inorg. Chem., 5, 1791.
Selig, H. and R. D. Peacock, (1964), A Krypton Difluoride-Antimony Pentrafluoride Complex, J. Amer.
Chem. Soc., 86, 3895.
Shamir, J. and J. Binenboym, (1968), Photochemical Synthesis of Dioxygenyl Salts, Inorg. Chem.
Acta, 2, 37.
Slansky, C. M., (1971), Separation Processes for Noble Gas Fission Products from the Off-Gas of Fuel-
Reprocessing Plants, At. Energ. Rev., 423
-379-
-------
Stein, L., (1969), Oxidized Radon in Halogen Fluoride Solutions, J. Amer. Chem. Soc. 91, 5396.
Stein, L., (1970a), Ionic Radon Solutions, Science, 168, 362.
Stein, L., (1970b), Chemical Properties of Radon, Yale Scientific Magazine, 44, 2.
Stein, L., (1972), Chemical Methods for Removing Radon and Radon-Daughters From Air, Science, 175,
1463.
Stein, L., (1973a), Removal of Radon from Air by Oxidation with Bromine Trifluoride, J. Inorg. Nucl.
Chem., 35, 39.
Stein, L., (1973b), Removal of Xenon and Radon from Contaminated Atmospheres with Dioxygenyl
Hexafluoroantimonate, O^SfoFg, Nature, 243, 30.
Stephenson, M. J., J. R. Merriman and D. I. Dunthorn, (1972), Recent Developments in Controlling
the Release of Noble Gases by Absorption in Fluorocarbons, USAEC Report K-L-6292, June 1972.
Streng, A. G., A. D. Kirshenbaum, L. V. Streng and A. V. Grosse, (1963), Preparation of Rare-Gas
Fluorides and Oxyfluorides by the Electric Discharge Method, in Noble Gas Compounds, H. H. Hyman,
editor, The University of Chicago Press, Chicago, 1963, pp. 73-80.
The safety and efficiency of O2SbFe for use as a radon oxidant in mines is currently being investigated at
Argonne under U. S. Bureau of Mines contract HO230018 (Argonne Study, ongoing).
Turner, J. J. and G. C. Pimentel, (1963), Preparation of Inert-Gas Compounds by Matrix Isolation:
Krypton Difluoride, in Noble Gas Compounds, H. H. Hyman, editor, The University of Chicago Press,
Chicago, pp. 101-105.
Wilson, E. J. and K. J. Taylor, (1958), The Separation and Purification ofKrypton-85 at the Multi-Curie
Level, UKAEA Report AEREI/R-2693.
Wirsing, E., Jr., L. P. Hatch and B. F. Dodge, (1970), Low-Temperature Adsorption of Krypton on
Solid Adsorbents, USAEC Report BNL50254 (T-586), June 1970.
-380-
-------
TABLE 1. Some Complex Fluorides and
Simple Fluorides That Have Been
Tested as Oxidants for Radon.
Compound Radon Retention (%)
KBrF,
4
CsBrFc
6
KC1F,
4
RbClF,
4
ClF2SbF6
BrF2SbF6
BrF4Sb2Fn
IFcSbQF, a
o o ID
BrF2BiF6
BrF0TaFc
£j D
IF.SbFfi
4 o
NOSbFft
6
AgF2
CoF3
28
7
19
16
100
100
100
100
100
27
76
0
0
0
-381-
-------
RADON-AIR
MIXTURE
COMPRESSED
VACUUM
DRYING
AGENT
RADIUM
CHLORIDE
SOLUTION
BrR
BrF,
V31®^-** VACUUM
P
Figure 1. Two-Stage Bromine Trifluoride Scrubber.
-------
FLOWMETER
AIR
1
ASCARITE
DRIERITE BrF3
TI SODA
LIME
VACUUM
-VACUUM
LIQUID
T2 SCINTILLATOR
Figure 2. Two-Stage Bromine Trifluoride Scrubber and Radon Condenser.
-------
CO
X
222Rn
5.49 MeV
2l8p0
6.00 MeV
2l4Po
7.69 MeV
100 150
CHANNEL NUMBER
200
Figure 3. Pulse-Height Spectrum of Radon-222 and Daughters in Scintillating Solution.
-------
PURIFIED AIR
INLET
WORKING
STORE
OS
DUST
FILTER
AIR EXHAUST
VAPOR
TRAP
AIR
DRIER
BLOWER
RADON
ABSORPTION
TOWER
PUMP
Figure 4. Proposed Air-Purification System For Use in a Uranium Mine.
-------
STRUCTURAL CONSIDERATIONS IN THE CHEMISTRY OF NOBLE GASES
W.E. Falconer
Bell Laboratories
Murray Hill, New Jersey 07974
Abstract
The structures of noble gas compounds provide insight into the bonding between noble gas atoms and
various ligands and into the electronic environment of the bound noble gas atoms. Structural knowledge of the
free molecules is thus important in advancing theories of chemical bonding. Structures inferred from infrared,
raman, visible, and ultraviolet spectroscopy, microwave spectroscopy, electron diffraction, Mossbauer
measurements, nuclear magnetic resonance, x-ray and neutron diffraction, electric deflection, and
thermodynamic measurements are compared for self-consistency and with theoretical calculations and
predictions. The structure of XeFg has proved particularly elusive. Most physical measurements indicate that
XeFs is not octahedral and that the F-Xe-F bond angles are not 90°. However, electric and magnetic deflection
of molecular beams ofXeFg have shown the absence of a static dipole moment and of significant electronic
paramagnetism. Weakly bound noble gas molecules have been formed in hypersonic nozzles and examined
using the molecular beam electric deflection techniques.
INTRODUCTION
From the inception of noble gas chemistry, physical measurements have played an indispensible role,
partly in simply identifying species under study, but more importantly, in characterizing their properties. We
may think of structure in the broadest terms as encompassing the spatial arrangement of atoms and the
electronic environment surrounding these atoms responsible for the spatial array. In this paper, we direct
attention to physical measurements which bear on the geometrical structures, emphasizing the technique of
deflection of molecular beams by inhomogeneous electric fields, and using the structures of simple fluorides as
examples.
Early speculations suggested that new concepts of bonding or compound formation might be required for
noble gas molecules. Experience quickly showed, however, that the properties of these molecules could-be
correlated qualitatively by either valence bond or semiempirical molecular orbital methods (Malm, et al.,
1965). The electron pair repulsion theory has in fact proved very useful in describing the geometries of the
known compounds (Gillespie, 1963 and Yamada, 1963).
Many structures of the simple molecules have been well established experimentally. For XeF2, the band
shape of the gas phase infrared active antisymmetric stretching motion, V3, at 560 cm-1, tells us the molecule
is D ooh, linear and symmetric (Smith, 1963a and Reichman, et al., 1969). The linear structure is confirmed by
x-ray (Siegel, et al., 1963) and neutron diffraction (Levy, et al., 1963) studies of XeF2- XeF4 was shown by
electron diffraction (Bohn, et al., 1963) to be square planar, D4h, consistent with gas phase infrared spectra
(Claassen, et al., 1963a), Raman spectra of the solid (Claassen, et al., 1963a), neutron diffraction (Burns, etai,
1963), and X-ray diffraction (Siegel, et al., 1963; Ibers, et al., 1963; and Templeton, et al., 1963) measurements.
The structure of KrF2 is known to be analogous to that of XeF2 from spectroscopic (Murchison, et al., 1968) and
electron diffraction (Harshbarger, 1967) measurements. The arrangement of linear molecules in the
tetragonal cell has been shown by X-ray diffraction (Burbank, et al., 1972). For XeOF4, the microwave
spectrum (Martins, et al., 1964 and 1968), the vapor phase infrared spectrum (Smith, 1963b and Claassen, et
al., 1963b), the liquid phase Raman spectrum (Smith, 1963b and Claassen, et al., 1963b), and high resolution
nuclear magnetic resonance (Brown, et al., 1963), all indicate C4V symmetry for this molecule, with the xenon
atom close to the plane of the fluorine atoms and the oxygen at the apex of the square pyramid. These
geometries of the binary fluorides are predicted by semiemipirical molecular orbital treatments (Bordreaux,
1963; Lohr, et al., 1963, and Jortner, et al., 1963) as well as by the electron pair repulsion theory (Gillespie, 1963
and Yamada, 1963) which also predicts the geometry of XeOF4.
The situation is less straightforward for XeFg. XeFg has been subjected to three electron diffraction studies
in the gas phase (Harshbarger, et al., 1967; Gavin, et al., 1968; and Hedberg, et al., 1966) and intensive study by
infrared, Raman, and other spectroscopy in all phases (Smith, 1963b; Claassen, et al., 1972; and Kim, et al.,
1968). These experimental studies are not consistent with a simple static structure for XeFg, and much thought
has led to the conclusion that a dynamical description is needed for the actual configuration. The distortion
may be viewed as a result of the role of the nonbonding electron pair in XeFg. The valence shell of XeFg
contains 14 electrons, only 12 of which are involved in Xe-F bonds. The electron pair would behave as a
seventh ligand. In any of the structures in Figure 1, six of the black dots representing ligands would be fluorine
atoms, the seventh would be the nonbonding pair. No matter which of these structures resulted, there would be
a considerable distortion from octahedral Oh symmetry and a contradiction of the predictions of the simplest
semiemipirical molecular orbital theory.
One might hope that an X-ray study of the XeFg molecule isolated in the solid state at a low temperature
would resolve the structural question. However, the solid is a complex system, with four polymorphs (Jones, et
al., 1970): a cubic form stable from 90°K to the melting point, and an interconverting and coexisting series of
two monoclinic and one orthorhombic modifications. XeFg exists either as hexameric rings or tetrameric rings
-386-
-------
in the crystals (Jones, et al, 1970; Burbank, et al, 1971 and 1973; and Agron, et al., 1965). No simple molecules
have been observed. The rings are fluorine bridged, and the molecular units are far from being simple
octahedra.
Because of the extensive intermolecular interaction, solid-state measurements are unlikely to provide a
clean-cut molecular structure for XeFe. Similarly, structural studies in the liquid are likely to be complicated
by intermolecular interactions, especially since Raman (Gasner, et al., 1967) and heat capacity (Schreiner, et
al., 1969) data show that the liquid is associated. Therefore, we must rely on gas phase measurements of
unperturbed molecules.
We have studied the behavior of molecular beams of XeFg and related molecules in an inhomogeneous
electric field (Falconer, et al., 1969). For those species where sophisticated spectroscopic techniques cannot be
applied, or fail to provide an unambiguous interpretation, simply knowing whether or not a molecule
possesses a permanent electric dipole can considerably narrow the range of possible structures. The electric
deflection method is sensitive to dipole moments of around 0.01D for near symmetric tops, and can detect those
of C103F(0.02D) and CH3D(0.005) unambiguously.
A schematic of the apparatus is shown in Figure 2. Molecules effuse from the source, 0, and are detected by a
high efficiency electron impact ionizer, mass spectrometer, and electron multiplier, MS. The direct line of sight
between the source and the detector can be blocked by a stop wire S, and a large electric field can be imposed by
the quadrupoles A and B. For polar molecules the signal is increased with applied electric field; for nonpolar
molecules, a decrease is observed.
We have performed separate mass spectrometric measurements to show that XeFg is monomeric in the gas
phase under the effusion source conditions. The electric deflection experiments show a decrease in signal, or
defocusing for XeF2, XeF4, and XeFg. An increase in signal or focusing is observed for XeOF4. The results for
XeF2, XeF4, and XeOF4 are consistent with their known structures, discussed above. However, these
observations argue against any static polar distortion in XeFg. An analysis shows, equally, that the results
argue against centrosymmetric inversions of polar structures, analogous to that in the ammonia molecule.
The electric deflection experiments do not define a structure for XeFg. However, they do set interesting
restrictions on allowable structures. Models consistent with these measurements have been proposed by
Burbank, et al., (1969), and by Bartell, et al., (1968), who have independently hypothesized a dynamic
molecular structure for XeFg. The dynamic molecular structure for XeFg. The dynamic structural concept is
being further developed by Pitzer, et al., (1973).
An alternate model to describe the observed physical properties of XeFg has been proposed by Goodman
(1972). It suggests that XeFg vapor is a mixture of molecular isomers differing in electronic state and nuclear
geometry. Some of these states are triplets and should be paramagnetic. However, neither magnetic
susceptibility measurements on the solid or liquid (Selig, et al., 1966), nor magnetic deflection experiments on
molecular beams (Code, et al., 1967), similar to the electric deflection measurements described above, were able
to detect paramagnetic components in XeFg. Thus, it is unlikely that electronic isomerism satisfactorily
explains the structure of XeF6- The dynamic structural approach appears more promising.
We have also used electric deflection and mass spectroscopy to examine noble gas molecules which are held
together by a weak Van der Waals interaction. The experiments enable us to determine whether given systems
are stable or nonstable, polar or nonpolar. We have expanded mixtures of noble gas atoms, fluorine atoms, and
fluorine molecules together through hypersonic nozzles (Harris, et al., 1973). Under the conditions of the
experiment, the diatomics Xe2, Kr2, and Ar2 were readily formed. With a room temperature source (no fluorine
atoms), we observed the weakly bound systems Xe... F2, Kr... F2, and Ar... F2- The mass spectral cracking
pattern of Xe ... F2 was distinctly different from XeF2 (chemically bound). Similar experiments using FC1
have provided the molecules Ar ... Cl -F and Kr ... Cl -F. A diatomic molecule of Kr and F has also been
observed. These experiments have enabled us to study thermodynamically stable, but physically unstable,
species. They provide insight into potential functions and important experimental comparisons with
sophisticated calculations. These experiments which are continuing will resolve questions of the stability of
these species and will provide estimates of the amount of charge transfer involved in the weak bonds.
The above examples demonstrate that physical structural measurements of free molecules continue to
advance our understanding of the chemistry of the noble gases which has excited the imaginations of
scientists since the work of Cavendish, Ramsey, and Rayleigh.
REFERENCES
Agron, P. A., C. K. Johnson and H. A. Levy, (1965), Inorg, Nucl. Chem. Letters 1, 145.
Bartell, L. S. and R. M. Gavin, Jr., (1968), J. Chem. Phys. 48, 2466.
Bartell, L. S., (1967), J. Chem. Phys. 46, 4530.
Bohn, R. K., K. Katada, J. V. Martinez and S. H. Bauer, (1963), Noble Gas Compounds, H. H. Hyman,
Ed. (U. of Chicago Press, Chicago, 111., 1963), p. 238.
Bordreaux, E. A., (1963), Noble Gas Compounds, H. H. Hyman, Ed. (U. of Chicago Press, Chicago, 111.,
1963), p. 354.
Brown, T. H., E. B. Whipple and P. H. Verdier, (1963), J. Chem. Phys. 38, 3029.
Burbank, R. D. and N. Bartlett, (1968), Chem. Comm. 645.
Burbank, R. D. and G. R. Jones, (1970), Science 168, 248.
Burbank, R. D. and G. R. Jones, (1971), Science 171, 485.
-387-
-------
Burbank, R. D. and G. R. Jones, (1973), J. Amer. Chem. Soc. in press.
Burbank, R. D., W. E. Falconer and W. A. Sunder, (1972), Science 178, 1285.
Burns, J.H., P.A. Agron and H.A. Levy, (1972), Science 139, 1208.
Cavin, R. M., Jr., and L. S. Bartell, (1968), J. Chem. Phys. 48, 2460.
Claassen, H. H., G. L. Goodman and H. Kim, (1972), J. Cbem. Phys. 56, 5042.
Claassen, H. H., C. L. Chernick and J. G. Malm, (1963a), J. Amer. Chem. Soc. 85, 1927.
Claassen, H. H., C. L. Chernick and J. G. Malm, (1963b),Noble Gas Compounds, H. H. Hyman, Ed. (U.
of Chicago Press, Chicago, 111., 1963b), p. 287.
Code, R. F., W. E. Falconer, W. Klemperer and I. Ozier, (1967), J. Chem. Phys. 47, 4955.
Falconer, W. E., A. Buchler, J. L. Stauffer and W. Klemperer, (1968), J. Chem. Phys. 48, 312.
Gasner, E. L. and H. H. Claassen, (1967), Inorg. Chem. 6, 1937.
Gillespie, R. J., (1963), Noble Gas Compounds, H. H. Hyman, Ed. (U. of Chicago Press Chicago, 111., 1963),
p. 333.
Goodman, G. L., (1972), J. Chem. Phys. 56, 5038.
Harris, S. J., S. E. Novick, W. Klemperer and W. E. Falconer, (1973), Unpublished results.
Harshbarger, W., R. K. Bohn and S. H. Bauer, (1967), J. Amer. Chem. Soc. 89, 6466.
Hedberg, K., S. H. Peterson, R. R. Ryan and B. Weinstock, (1966), J. Chem. Phys. 44,1726.
Ibers, J. A. and W. C. Hamilton, (1963), Science 139, 106.
Jones, G. R., R. D. Burbank and W. E. Falconer, (1970), J. Chem. Phys. 53,6450 (1970); 53,1605 (1970).
Jortner, J., E. G. Wilson and S. A. Rice, (1963), Noble Gas Compounds, H. H. Hyman, Ed. (U. of
Chicago Press, Chicago, 111., 1963).
Kim, H., H. H. Claassen and E. Pearson, (1968), Inorg. Chem. 7, 616.
Levy, H. A. and P. A. Agron, (1963), J. Amer. Chem. Soc. 85, 241.
Lohr, L. L., Jr. and W. N. Lipscomb, (1963), Noble Gas Compounds, H. H. Hyman, Ed. (U. of Chicago
Press, Chicago, 111., 1963), p. 347.
Malm, J. G., H. Selig, J. Jortner and S. A. Rice, (1965), Chem. Rev. 65, 199.
Martins, J. and E. B. Wilson, Jr., (1964), J. Chem. Phys. 41, 570.
Martins, J. and E. B. Wilson, Jr., (1968), J. Mol. Spectroscopy 26, 410.
Murchison, C., S. Reichman, D. Anderson, J. Overend and F. Schreiner, (1968), J. Amer. Chem.
Soc. 90, 5690.
Pitzer, K. S. and L. Bernstein, (1973), personal discussion.
Reichman, S. and F. Schreiner, (1969), J. Chem. Phys. 51, 2355.
Schreiner, F., D. W. Osborne, J. G. Malm and G. N. McDonald., (1969), J. Chem. Phys. 51, 4838.
Selig, H. and F. Schreiner, (1966), J.Chem. Phys. 45, 4755.
Sieged, D. and E. Gebert, (1963), J. Amer. Chem. Soc., 85, 240.
Smith, D. F., (1963a), J. Chem. Phys. 38, 270.
Smith, D. F., (1963b), Noble Gas Compounds, H. H. Hyman, Ed. (U. of Chicago Press, Chicago, 111., 1963b),
p. 295.
Templeton, D. H., A. Zalkin, J. D. Forrester and S. M. Williamson, (1963), J. Amer. Chem. Soc. 85,
242.
Yamada, S., (1963), Rev. Phys. Chem. Japan 33, 39.
-388-
-------
Figure 1. Possible distorted structures for XeFg.
-389-
-------
I
0 A S B MS
n -
Figure 2. Schematic diagram of the electric deflection apparatus.
-------
AN OVERVIEW OF THE PHYSICAL-CHEMICAL PROPERTIES OF THE NOBLE GASES
C. McKinley
Air Products and Chemicals, Inc.,
Allentown, Pennsylvania
Abstract
This paper lists the concentrations of noble gases in the atmosphere and the relative abundance of the stable
isotopes. Selected physical properties are tabulated; solubilities of noble gases in water and other liquids, and
liquid-vapor equilibria data for binary systems containing a noble gas are presented. Adsorption data are
tabulated for illustrative conventionaladsorbentsandarealsopresentedbyaPolanyicorrelation.Clathration,
biochemical effects, and chemical reactivity are highlighted. Analytical procedures are briefly described.
Other relatively non-reactive gases present in the atmosphere in trace quantitites are mentioned: methane,
carbon tetrafluoride, and sulfur hexafluoride.
CONCENTRATION OF NOBLE GASES IN THE ATMOSPHERE
The five stable noble elements, which share a complete lack of activity towards many chemical reagents,
range in atmospheric concentration from 9,340 ppm for argon to 0.086 ppm for xenon. The noble element
concentration sequence, as seen in Table 1, is Ar>Ne>He>Kr >Xe>Ra. Radon, which has no stable
isotopes, is present at an average concentration of about 6 x 10-14 ppm, but is quite variable in concentration,
being exhaled from the earth's land masses (Junge, 1963).
The concentrations of neon, krypton, and xenon in the atmosphere should be quite constant with time while
the concentrations of helium and argon may be changing. Helium, 5.24 ppm, is composed of two isotopes, 3He
and 4He, which appear in the atmosphere in the ratio of about 1.2 x 10-6 (Junge, 1963; Mamyrin, et al., 1970).
4He is produced in almost equal parts by the decay of 238U and 232Th, and about 2% is due to decay of 235U. 3He is
produced by cosmic-ray spallation reactions with 14N in the atmosphere and by reactions with lithium and
other elements in the lithosphere. Thus, helium is entering the atmosphere. It appears, however, that
sufficient helium reaches escape velocity to somewhat stabilize the atmospheric content of helium for both
isotopes. Atmospheric argon has been estimated (Moody, et al., 1964) to increase through the decay of fnKto
|§Ar , the rate being such that the Ar concentration would rise about 1 ppm each 700,000 years.
Radon isotopes/arise from decay of uranium and thorium. ^ojRn, arising from uranium-238 (radium), is the
dominant radon isotopein the atmosphere as its half-life of 3.8 days is significantly longer than that of the 59
second half-life of 220 Rn derived from the decay of thorium-232. The concentration of radon over land has been
noted to be 2 orders"" of magnitude higher than over theocean. Actual concentrations vary substantially with
height and atmospheric condition. The recording in a noble gases concentration table of an average value for
radon, thus, is an enormous simplification of the actual concentration distribution at any one time.
Naturally occurring isotopic concentration ratios are listed in Table 2.
PHYSICAL PROPERTIES OF THE NOBLE GASES
The noble gases are colorless, odorless, tasteless, monatomic, and non-reactive with all the usual chemical
reagents. Perfect spherical symmetry results from the monatomic molecular structure and accounts for some
of the interesting physical properties. The vapor pressures of the noble gases at the triple point are high,
ranging from 0.4273 atm for Ne to 0.806 atm for Xe. The helium isotopes, indeed, cannot be frozen under their
own vapor pressure; an external higher pressure must be supplied. Selected physical properties are listed in
Table 3 (Kirk-Othmer, 1966 and Comprehensive Inorganic Chemistry, 1973).
The principle of corresponding states (Leland, et al., 1968) is illustrated in excellent fashion by Ar, Kr, and
Xe. The classical corresponding states principle (CSP), in which reduced parameters are obtained by dividing
the particular temperature, pressure, and density by the corresponding critical values, is well-suited to the
heavier noble gases. An illustration of this empirical CSP is shown in Figure 1, in which the reduced vapor
pressures of the various gases are plotted as a function of the reciprocal of the reduced temperature. The
experimental points for krypton and xenon are seen to fit the line obtained for the extensively studied gas,
argon, quite well. The fit of the Rn data is not much better than for Ne, however, these data are quite old (1909),
and the temperature scale is possibly the cause for the poor correlation. The deviation from the argon line is
apparent for neon, is substantial for 4He, and becomes greatest for the lightest element in the group, 3He. To
include these lighter noble gases, it is best to use theoretically based molecular parameters to obtain reduced
variables. This allows the introduction of an additional parameter to account for the deviations from the
classical behavior caused by quantum effects. P-V-T transport properties and other similar data can be
obtained for Xe, Kr, and especially Rn from the extensively determined properties of Ar using CSP.
This useful tool should allow the establishment of radon properties which are difficult to measure because of
its scarcity and radioactivity. Neon to some extent exhibits anomalous behavior at low temperatues caused by
Quantum-mechanical effects; the two helium isotopes exhibit substantial anomalies. Critical point and
normal boiling point properties for 3He and "He differ substantially (see Table 3); these two isotopes have been
-391 -
-------
much studied because of their unique low temperature properties. (Helium is the only matter in the universe
existing as a liquid below the triple point of hydrogen.)
Figure 2 illustrates trends in several noble gas properties as a function of atomic number. The critical
properties, pressure, temperature, and density increase with increasing atomic number; however, ionization
potential and minimum exitation energy decrease. He and Ne are seen to depart from the smooth curves,
which fit the four larger molecules.
The Ar-Kr pressure-temperature projection, shown in Figure 3, is typical of a binary system which forms
solid solutions in that the S-L-V locus is a smooth continuous curve connecting the respective triple points.
Figure 4 is an isothermal section for this system at the temperature of the peak in the S-L-V curve. This
pressure-composition diagram indicates the compositions of the various phases including the solubility or
triple point composition. Such data are of theoretical interest because of the simplicity of the molecules, and
are of practical use in the cryogenic separation of these gases.
ADSORPTION OF THE NOBLE GASES
Adsorption processes are used both in the production of noble gases and in the removal of trace quantities of
radioactive noble gases from nuclear power plant vent gases. For activated carbon, silica gel, 5A and 13X
zeolites, the affinity of the adsorbent for the noble gases increases with increasing molecular weight. This is
shown in Table 4 which gives data for the adsorption of all of the noble gases except radon on activated carbon
and 13X zeolite. (Silica gel and 5A zeolite capacities are similar to those for 13X.) Because of the limited amount
of available data, there is some variation in temperature and pressure, but the trend of increasing capacity
with molecular weight can be seen. As can be seen from the table, the helium and neon adsorption capacities
are about two orders of magnitude lower than the capacities for argon, krypton, and xenon. Also, the capacity
of the 13X zeolite for the noble gases is lower than that of activated carbon. In the case of the 4A zeolite, the
krypton and xenon are excluded because of their size, while the argon capacity is reduced at temperatures
below -165°C apparently because of a slight pore contraction. Argon, xenon, and krypton are excluded from 3A
zeolite. Adsorption data for radon are not readily available. A review of the adsorption of noble gases is given
by Cook (1961).
Figure 5 gives a correlation of adsorption data for krypton on activated carbon using the Polanyi technique.
Examples of the use of adsorption in the production of the noble gases are: the separation of helium-neon
mixtures by neon adsorption on activated carbon, removal of oxygen from crude argon by adsorption on 4A
zeolite, and separation of krypton-xenon mixtures using activated carbon. A description of these processes is
given by Barren (1966). Activated carbon adsorbers are also being considered to remove or reduce trace
radioactive krypton and xenon impurities from nuclear power plant vent gases. A review of this work is given
by Kovach (1970).
CLATHRATION
The noble gases, as "guest" molecules, may be trapped in crystalline cages formed by "host" molecules.
Water, hydroquinone, and phenol have been found to be such "hosts' for clathration of the noble gases.
Examples of several such crystalline structures are given in Table 5 along with some of the observed
properties. The heat of formation of the clathrates is relatively low. If one subtracts from the heat of formation
of the Kr and Xe hydrates an approximated 8.3 kcal, representing crystallization of 5.75 gm moles of H20, one
obtains 5.6 and 8.4 kcal, respectively, per gm atom of noble gas, which values correspond approximately to the
heats of adsorption of the same gases. The forces which hold the noble gas in the hydrate are seen to be of the
same order as Van der Waals bonding forces (Cook, 1961).
Bartlett, et al., (1973) provide an excellent summary of the noble gas clathrates.
SOLUBILITY
The solubility of the noble gases in water is shown in Figure 5 for the temperature range 0 to 70°C and at
atmospheric pressure. The solubility increases with increasing atomic number. The two smallest molecules,
helium and neon, exhibit a minimum solubility in the temperature range illustrated.
PHYSIOLOGICAL AND BIOCHEMICAL EFFECTS
The noble gases, although chemically inert, do affect living systems when present in sufficient
concentrations. The effects upon living systems increase with increasing molecular weight for several of the
phenomena. Narcosis is sufficiently strong with xenon that surgery has been conducted at atmospheric
pressure using 80%Xe:20%O2 as the anesthetic. The depth of anesthesia obtained with xenon is intermediate
between nitrous oxide and ethylene. Krypton exhibits slight narcosis at atmospheric pressure; several
atmospheres partial pressure of krypton would probably be required to induce deep anesthesia and even
higher pressures with argon. Helium has been used to replace nitrogen in breathing atmospheres for work at
several hundred feet depth in the ocean with no narcotic effects, these depths being such that nitrogen
narcosis would have been a severe problem.
-392-
-------
Helium-oxygen mixtures for breathing have found utility also in medicine, in aviation, and in other
operations under pressure.
Members of the noble gas family have been studied, in addition to their applications in breathing mixtures
and in medicine, for effects upon metabolism, upon responses to stimuli and upon oxygen-dependent
sensitivity to radiation. Cook (1961) provides an excellent historical summary of noble gas related biological
research.
CHEMICAL REACTIONS
Bartlett (1962) reported the interaction of xenon with platinum hexafluoride to form a red solid of
composition Xe (PtFe)x. Interest in noble gas compounds spread very rapidly; shortly a large number of brief
communications appeared in the literature followed by several excellent reviews: Hyman (1963), Moody
(1964), Holloway (1969), Horn (1970), Jha (1970), Geovold (1972), and Bartlett (1972). Many compounds
containing xenon have been prepared and their properties measured; only a few containing krypton and
radon have been made. Moody's (1964) summary listed 24 compounds containing xenon, 2 containing
krypton, and 1 containing radon.
ANALYTICAL PROCEDURES
Frequently noble gas samples contain one or more of the noble gases in very low concentration and, as a
consequence, extraordinary precaution must be taken to ensure that the sample is initially collected in a
fashion to be truly representative. Further, care must be taken that adsorption and/or absorption do not bring
about composition changes. With appropriate handling care one can select an analytical procedure to fit the
specific requirement. Gas chromatography is very useful, allowing separation of all the noble gases and their
determination to very low concentration levels; mass spectrometry is particularly effective in stream
monitoring for specific masses and for determination of isotope ratios; emission spectroscopy is valuable in
stream monitoring of a high-purity inert gas for the presence of a trace of an impurity (Cook, 1961 and
Schmauch, 1971). Radioactivity measurements are useful to supplement analysis by chromatography or mass
spectrometry when radioactive isotopes are present. It is necessary to ensure that the radiation being counted
is uniquely from the isotope of interest. Gibbs (1973) describes a portable radon monitoring instrument, and
the techniques for avoiding background radiation and radiation from radon-daughters.
OTHER RELATIVELY NON-REACTIVE GASES IN THE ATMOSPHERE
Man's activities have resulted in the introduction into the atmosphere of some gases which will have a
relatively long life in the atmosphere. Sulfur hexafluoride is an interesting example. Clemens (1968), based on
production and consumption data provided by manufacturers and users, has estimated that the worldwide
average concentration of SFg should be at least 10-7 ppb. He found that a number of samples of air taken in and
around Cincinnati, Ohio ranged from 0.04 to 0.20 x 10-3 ppb, and that no sample taken was free of SFg. SFg
lends itself to use in meterological tracing because of its chemical inertness. SFg concentration in air can be
measured down to a sensitivity of 1 x 10-5 ppb by a gas chromatography procedure using an electron capture
detector. The SFg content of the air sample is concentrated by adsorption onto charcoal at a low temperature,
and then is desorbed for introduction to the chromatograph. The technique has been used to trace air
movements over distances of 75 miles with a release rate of SFg of about one pound per minute for an hour.
Methane, although an unusually stable hydrocarbon, in contrast to sulfur hexafluoride, has a relatively
short atmospheric life estimated to be in the range of 10 to 20 years (Ehhalt, 1967 and Wofsy, 1972). A very
large fraction of the methane entering the atmosphere is from decomposition of organic material at the earth's
surface. The sinks for methane are not well understood. Electrical discharges, photolysis of methane and
bacterial action may be involved. The generation rate is of the order of 3 x 1014 grams per year. Whatever the
sinks are, they effectively reduce the methane concentration in the atmosphere to the relatively constant
average of about 1.2 to 1.6 ppm (Cavanagh, et al., 1969). The generation rate, with no sinks, would increase the
atmospheric concentration of methane about 0.1 ppm per year.
Carbon tetrafluoride, a very stable molecule, is present in the atmosphere at very low concentrations. CF4
has entered the atmosphere through its production and use, and probably also to a significant extent through
the venting of fluorine to the atmosphere through carbon beds. The concentration of CF4 in the atmosphere
can be estimated from its appearance during commercial production of krypton and xenon. CF4 boils between
Kr and Xe, the normal boiling points being Kr 119.8°K, CF4145.1°K, and Xe 165°K. Hence, in the fractionatipn
process CF4 is concentrated in the intermediate fraction between the pure Kr and pure Xe. From a material
balance around the entire distillation process, it is possible to calculate the CF4 concentration in the air feed to
the air separation plant. In the spring of 1966, the air feed to the separation plant described in G. G. Handler's
paper at this Symposium was found to contain about 2.3 x 10-5 ppm by volume of CF4. A few months later, in
the early fall, the concentration of CF4 in the air feed was calculated from material balances around the
krypton-xenon recovery unit to be 2.7 x 10-5 ppm. CF4 has also been found in krypton and in xenon produced in
other parts of the world. One must conclude that it is distributed throughout the atmosphere, although we do
not have information on its concentration variability.
-393-
-------
REFERENCES
Barren, R., (1966), Cryogenic Systems, pp. 243-255, McGraw Hill Book Company.
Bartlett, N., (1962), Proc. Chem. Soc., 218.
Bartlett, H. and F. O. Sladky, (1973), The Chemistry of Krypton, Xenon, and Radon, Pergamon Press,
Chapter 6.
Brunaur, S., (1945), The Desorption of Gases and Vapors, 1, p. 58, Princeton Univ. Press.
Cavanagh, L. A., C. F. Schadt, and E. Robinson, (1969), (Stanford Research Institute, Menlo Park,
California). Envir. Sc. Tech. 3 (3) 251-7. Atmospheric Hydrocarbon and Carbon Monoxide Measurements at
Point Barrow, Alaska.
Clemens, C.A., A. I. Coleman and B. E. Seltzman, (1968), (National Center for Air Pollution Control,
Public Health Service, Cincinnati, Ohio). Envir Sc. Tech. 1968, 2 (7), 551-6 (Eng.), (C. A. Volume 69, p. 3618,
38525 n).
Cockett, A. H. and K. C. Smith, (1973), The Monatomic Gases: Physical Properties and Production.
Comprehensive Inroganic Chemistry. Pergamon Press, Chapter 5.
Cook, G. A., (1961), Argon, Helium and The Rare Gases. Interscience Publishers, New York (394 pages in
Volume I; 395-818 pages in Volume II).
Ehhalt, D. H., (1967), Methane in the Atmosphere. J. Air Pollution Control Assoc. 17 (No. 8) 518-9.
Fastovskii, V. G., et al., (1958), Zhur. Prklad. Khim., 31, (1) pp. 5-13.
Geodvold, J. G., (1972), Noble Gases and Their Compounds, Chem. Tech. (Amsterdam) 1972,27 (4),89-92
(Neth.) (C. A. 76 (1972), 161706u).
Gibbs, H. L. and J. H. Bilbrey, Jr., (1973), Development of a Portable Radon Detector, Bureau of Mines
Report of Investigations 7741.
Grant, R. and M. Manes, (1964), M. I & EC Fund, 3, p. 221.
Gray, R. W., and W. J. Ramsay, (1909), J. Chem. Soc. 95: 1073.
Heastie, R., (1955), Nature 176: 747-8.
Holloway, J. H., (1968), Methuen: London), Noble-Gas Chemistry, 213 pp. (C. A. 70 (1969) p. 294,102658
h).
Horn, H. G., (1971), (Gmelin-Inst. Max-Planck-Gess., Frankfurt/Main., Ger.), Chem.-Ztg., Chem. App.
(1970, 94 (16), 600-7 (Ger). (C. A. 74 p. 9133, 9119 F).
Hyman, H. H., (1963), Noble Gas Compounds, The University of Chicago Press, Chicago, 404 pages.
Jha, N. K., (1970), (Dep. Chem., Indian Inst. Technol. New Delhi, India). Chemistry of Noble Gas
Compounds, Proc. Chem. Symp., 1st 23-26 Sep. 1969, 1, 165-71 (Eng).
Junge, C. E., (1963), Air Chemistry and Radioactivity, Academic Press, New York, 382 pages.
Kirk-Othmer, (1966), Encyclopedia of Chemical Technology, Second Ed. Volume 10, p. 862-894.
Kovach, J. L., (1970), North American Carbon R&D Report No. 010005.
Leland, T. W. and P. J. Chappelear, (1968), Ind. Eng. Chem. 60 (7): 15-43.
Loosli, H. H., H. Oeschger, and W. Wiest, (1970), (Phys. Inst., Univ. Bern, Bern, Switz.). J. Geophys.
Res. 1970, 75 (15), 2895-900 (Eng). Argon-37, Argon-39, and Krypton-81 in the Atmosphere and Tracer Studies
Based on These Isotopes.
Mamyrin, B. A., G. S. Anufriev, I. L. Kamenskii, and I. N. Tolstikhin, (1970), (Phys. Tech. Inst,
Leningrad, USSR). Geokhimiya 1970 (6) 721-30 (Russ).
Moody, G. J. and J. D. R. Thomas (1964), Noble Gases and Their Compounds, The MacMillan
Company, New York, 62 pages.
Penchev, N. P., (1970), (Institute of General and Inorganic Chemistry, Bulgarian Academy of Sciences,
Sofia, Bulgaria). Izv. Otd. Khim. Nauki, Bulg. Akad. Nauk. 1970, 2 (3), 607-18 (Eng). Contribution to the
Analysis and the Geochemistry of the Noble Gases.
Schmauch, G., (1971), Encyclopedia of Industrial Chemical Analysis, Volume 13, pages 286-309, Gases,
Noble.
Sherman, R. H., S. G. Sydoviak, and T. R. Roberts (1964), J. Res. NBJ, 68A (6): 579-588.
Stein, S. A., et al., (1962), Arnold Engineering Development Center, USAF TDR-62-200
Von Antropoff, A., et al., (1952), Kollid;Zeitschrift, 129, pp. 1-10
Wagner, W., (1973), Cryogenics 13 (8): 470.
Wofsy, S. C., J. C. McConnell, and M. B. McElroy, (1972), (Smithsonian Astrophys. Obs., Cambridge,
Mass.). J. Geophys. Res., 77 (24), 4477-93 (Eng). Atmospheric Methane, Carbon Monoxide, and Carbon
Dioxide.
-394-
-------
TABLE 1. Concentration of Noble Gases in the Atmosphere (Kirk-Othmer, 1966).
Atomic Atmospheric Concentration
Element Symbol Number ppm by Volume (Dry Air)
Helium
Neon
Argon
Krypton
Xenon
Radon
He
Ne
Ar
Kr
Xe
Rn
2
10
18
36
54
86
5.24
18.18
9,340
1.14
0.086
6xlO-14
TABLE 2. Naturally Occurring Isotopic Abundance of the Noble Gas Elements (Schmauch, 1971).
Element Atomic Number Mass Number Abundance, Atom %
He
Ne
Ar
Kr
2
10
18
36
Xe
54
3
4
20
21
22
36
38
40
78
80
82
83
84
86
124
126
128
129
130
131
132
134
136
Rn
86
0.00013
99.9999
90.92
0.257
8.82
0.337
0.063
99.600
0.354
2.27
11.56
11.55
56.90
17.37
0.096
0.090
1.919
26.44
4.08
21.18
26.89
10.44
8.87
no stable isotopes
-395-
-------
TABLE 3. Physical Properties of the Noble Gases (a) (Kirk-Othmer, 1966).
Property
Atomic number
Atomic weight
Critical Point:
temperature
pressure
density
Normal boiling point (nbp)
Triple point (tp):
temperature
pressure
Density:
gas, 1 atm, 273.15°K
gas, nbp
liquid, nbp
liquid, tp
solid, tp
Gas/liquid vol ratio (b)
Units
12C = 12
12C = 12
°K
atm
g/ml
°K
°K
atm
g/liter
g/liter
g/ml
g/ml
g/ml
Heat of vaporization at nbp cal/g-mole
Heat of fusion at tp
Heat capacity at constant
pressure:
gas, 1 atm, 25° C
liquid, nbp (c)
Sonic velocity:
gas, latm,0°C
Thermal conductivity:
gas, 1 atm, 0°C
liquid, nbp
Viscosity:
gas, 1 atm, 25°C
liquid, nbp
First ionization potential
Min excitation energy
cal/g-mole
cal/g-moleX°K)
cal/g-moleX°K)
m/sec
cal/(sec)(cm2X°K/cm)
cal/(sec)(cm*X°K/cm)
/LIP
mP
eV
eV
»He
2
3.0160
3.324
1.149
0.0413
3.1905
notp
notp
0.1347
23.64
0.0589
notp
notp
437.4
6.09
notp
(4.968
4.0
(1,122)
(391)
51
(172)
0.0161
'He
2
4.0026
5.199
2.261
0.0693
4.215
notp
notp
0.17850
16.714
0.1249
notp
notp
700
19.4
notp
4.968
4.33
974
339.0
75
198.5
0.030
24.586
19.818
Ne
10
20.183
44.40
26.19
0.483
27.09
24.54
0.4273
0.90002
9.552
1.206
1.247
1.444
1,340
429
80.1
4.969
8.9
433
110.1
310
317.3
1.24
21.563
16.618
Elements
Ar
18
39.948
150.86
48.34
0.536
87.28
83.81
0.6800
1.78380
5.763
1.3936
1.418
1.623
781
1,550
283
5.969
10.9
307.8
40.5
290
226.4
2.75
15.759
11.548
Kr
36
83.80
209.4
54.3
0.908
119.80
115.77
0.7220
3.7493
8.6
2.415
2.451
2.826
644
2,154
392
5.008
10.6
213
20.9
211
253
4.31
13.999
9.915
Xe
54
131.30
289.74
57.64
1.100
165.04
161.38
0.806
5.8971
11
3.057
3.076
3.540
518
3,020
548.5
5.022
10.65
168
12.1
175
231.0
5.28
12.129
8.315
Rn
86
222
378
62
211
202
(0.7)
9.73
4.4
452
4,325
776
(5)
233.2
10.747
6.772
(a) Numbers in parenthesis are estimated.
(b) Volume of gas at 1 atm and 273.15°K (O°C) equivalent to unit volume liquid at nbp.
(c) Heatcapacity of saturated liquid.
-396-
-------
TABLE 4. A Comparison of Adsorption Capacities of Noble Gases
on Activated Carbon and Zeolite Type 13X.
Activated Carbon
Zeolite
Gas
He
Ne
Ar
Kr
Xe
Temp
°C
-190
-195
-196
-170
-170
Press
mmHg
100.00
1.0
1.0
0.8
0.03
Capacity
ml*
g
0.28 (a)
1.14 (a)
224 (b)
242 (a)
230 (a)
Temp
°C
-250
-183
-183
-183
Press
mmHg
0.5
**
0.8
0.018
0.09
Capacity
ml*
g
5.6 (c)
80 (a)
47 (a)
84 (a)
* ml of gas at O°C and 1 atm pressure per gram of adsorbent.
**No data were found for neon adsorption on zeolites.
(a) (Cook, 1961)
(b) (Grant, etal., 1964)
(c) (Stein, et al, 1962)
-397-
-------
TABLE 5. Noble Gas Clathrates With Water, Hydroquinone and Phenol (d)(e).
Formula
CO
tO
00
Theoretical
8Ar.46H2O
8Kr.46H2O
8Xe.46H2O
8Rn.46H2O
Observed
Ar.6H2O
Kr.6H2O
Xe.6H2O
Ar.3CgH4(OH)2 0.8Ar.3CgH4(OH)2
Kr.3CgH4(OH)2 0.74Kr.3CgH4(OH)2
Xe.3CgH4(OH)2 0.88Xe.3CgH4(OH)2
4Ar.l2CcH,-OH 2.92Ar.l2CcH,-OH
bo bo
4Kr.l2CgH5OH 2.92Kr.l2CgH5OH
4Xe.l2CgH5OH 4Xe.l2CgH5OH
Noble Gas, wt% Heat of Decomposition Dissociation
Formation Temp, °C at Pressure, atm.,
Theoretical Observed kcalmole-1 one atm. press. atO°C
27.8
44.7
56.0
68.2
10.8
20.3
28.5
12.4
22.9
31.8
-27
^44
-53
8.8
15.8
26.0
9.4
18.0
31.8
13.9 (a)
16.7 (a)
5.1 (b)
9.85 (c)
8.97 (c)
8.8 (c)
-42.8
-27.8
- 3.4
105
14.5
1.5
33
6
1
(a) Heat of formation based on composition ratio of (one atom noble gas) (5.7 5 H2O).
(b) Heat of formation based on composition ratio of (one atom noble gas) (3CgH4 (OH)2).
(c) Heat of formation in the range of O° to 40°C, kcal per gm atom of inert gas absorbed.
(d) (Cook, 1961).
(e) (Kirk-Othmer, 1966).
-------
Figure 1.
1.0
O.I
0.01
Ar —
3He —
Ne
Kr
Xe
Rn
o
o
A
A
WAGNER (1973)
SHERMAN, ET AL. (1964)
• COOK (1961)
GRAY AND RAMSAY (1909)
1.0
1.2
1.3
1.6
1.4 1.5
l/Tr
REDUCED VAPOR PRESSURE CURVES
OF THE NOBLE GASES
1.7
1.8 13
-399-
-------
30
1.5
20
>
0)
10
Figure 2.
IST IONIZATION POT
MIN. EXCITATION ENERGY
He Ne Ar
Kr Xe
ATOMIC NUMBER
TRENDS IN SELECTED NOBLE GAS PROPERTIES
60-
-300
40-
0.°
20
400
hoo
Rn
-------
Figure 3.
Ar VAPOR PRESS.
Kr VAPOR PRESS.
80
140
160
SIMPLIFIED PRESSURE-TEMPERATURE PROJECTION
FOR THE Ar-Kr SYSTEM
- 401 -
-------
Figure 4.
o-APCI - UNPUBLISHED
A-HEASTIE (1955)
0.4 0.6
MOLE FRACTION Ar
ISOTHERMAL SECTION FOR THE
Ar-Kr BINARY AT 105.3 °K
-402-
-------
Figure 5.
1000
500
o
H
N.
1
10 20 30 40 50 60 70
T (°C)
SOLUBILITY OF NOBLE GASES IN WATER
-403-
-------
Figure 6.
10
I io-'
I
JQ
> IO
'2
10
,-3
0
POLANYI CORRELATION FOR
KRYPTON ADSORPTION ON
ACTIVATED CARBON
o o
.0 J
20 40 60
80
100 120 140
^ '"I1 xl03[psia]
-404-
-------
A TEST OF INTERMOLECULAR POTENTIALS FOR THE NOBLE GASES BY COMPARISON
WITH EXPERIMENTAL THERMAL DIFFUSION FACTORS
W.L. Taylor
Monsanto Research Corporation
Mound Laboratory*
Miamisburg, Ohio 45342
Abstract
The description of the equilibrium and transport properties of dilute gases follows from a knowledge of the
pair interaction energy of the gas particles. We have used a second order effect, the thermal diffusion factor,
which is extremely sensitive to the inter molecular potential, as a test of various potentials which have been
proposed for the noble gases. Measurements of the thermal diffusion factor were made over a wide range of
temperatures in a trennschaukel; and these results, along with other values reported in the literature, were
fitted to empirical equations within the experimental uncertainties. Many proposed intermolecular
potentials, including some recently obtained from solid state properties and molecular beam scattering, were
used to calculate transport collision integrals and thence theoretical thermal diffusion factors which were
compared to the empirical equations. In the case of helium, quantum mechanical calculations were made. The
"best" potentials for each noble gas were then used to evaluate other macroscopic properties of the
homologues.
INTRODUCTION
Knowledge of the intermolecular forces between atoms and molecules can be obtained from both
experimental observations and theoretical considerations. However, the successful determination of the
interaction forces from empirical data requires the existence of an accurate theoretical description of the
physical property, and further requires that the measurements be made in a temperature (or relative kinetic
energy) region which is "sensitive" to the intermolecular potential. Traditionally, the properties of dilute
gases or those of the crystal lattice have been employed because of the existence of rigorous theoretical
treatments. More recently, primarily due to improved experimental techniques, the powerful method of
molecular beam scattering to obtain total and differential scattering cross sections has been added to the list.
From the practical point of view, once the intermolecular potential has been accurately determined by
whatever method, other properties required, for example in engineering applications, can generally be
calculated without the need for extensive experimental measurements.
In the present work, we have used one of the transport properties of the dilute gas, the thermal diffusion
factor, as the empirical test. Thermal diffusion is especially well suited for this task because transport
properties of monatomic dilute gases are very adequately described by the rigorous kinetic theory of Chapman
and Enskog, plus the fact that thermal diffusion, being a second-order effect, is the most sensitive to the
interaction.
The major drawback to measuring thermal diffusion factors lies in the fact that the effect is very small,
particularly for isotopes; and, in fact, was predicted theoretically before it was observed experimentally.
Essentially, the measurement consists of determining the isotopic separation in a temperature gradient.
There are three basically different methods or types of apparatus that can be used for making the
experimental measurement: (a) the thermogravitational column; (b) "two-bulb" apparatus; and (c) the
trennschaukel or "swing separator". In the first method the separations are large, a great advantage, but it is
difficult to operate a column over an extended temperature range; and theoretical treatment of the column
behavior is complicated because of the convective flow caused by the horizontal temperature gradient. The
"two-bulb" apparatus suffers from the fact that the separation is very small. We have chosen the
trennschaukel, developed by Clusius and Huber (1955), as the preferred method for the present work because it
cascades the single stage separation of the "two-bulb"; and a rigorous theoretical treatment of its operation
has been given by van der Waerden (1957).
Of the noble gases, far more experimental determinations have been conducted on helium and neon than on
argon, krypton, or xenon. Unfortunately, much of the data are widely scattered; and in some cases it is felt
that systematic experimental errors are present. We have measured the temperature dependence of the
isotopic thermal diffusion factor of the noble gases from approximately 200 to 900°K, and in the case of helium
the temperature range was extended down to 5°K. In addition, a wide variety of proposed intermolecular
potentials were selected for evaluation. The transport collision integrals were evaluated and used to compute
theoretical thermal diffusion factors which were compared to the experimental data. Quantum mechanical
calculations were used for helium at low temperatures and classical theory for the heavier gases.
*Mound Laboratory is operated by Monsanto Research Corporation for the U.S. Atomic Energy Commission
under Contract No. AT-33-1-GEN-53.
-405-
-------
EXPERIMENTAL
The experimental appartus utilized for the present work is shown schematically in Figure 1. The
trennschaukel (C) consists of 20 Inconel tubes heliarc welded into two massive nickel blocks of toroidal
shape with each tube connected top to bottom with an Inconel capillary. The top end of the first tube and the
bottom end of the twentieth tube are connected to either end of a microbellows pump mounted on the upper
block. This assembly is mounted in an environmental chamber (B). The top block and microbellows pump are
held at the temperature of the chamber, TH ; and the bottom block, which is insulated (insulation not shown),
is maintained at temperature TC by passing either air or liquid nitrogen from source (F) through passages in
the lower block. The temperature gradient, AT, is sustained in the exposed middle one-third of the tubes.
During an experiment, the gas is swung back and forth by means of the bellows pump (A). At the end of an
experiment the gases from the two sides of the bellows pump are sampled in the sampling manifold (D). The
lines are first purged, and then duplicate samples are withdrawn from the bellows.
The expression for the thermal diffusion factor in terms of the experimental quantities is obtained from the
mass flux equation for a binary mixture in a one-dimensional temperature gradient. The resulting expression
for (XT is
aT= Inq/ln (TH/TC)
where q is the separation factor in a single tube defined by the ratio of concentration ratios (xVx2) at TH to
(xVx2) at TC . Three corrections must be applied to the measured concentrations to compensate for the
operating characteristics of a trennschaukel. These are derived in the theory by van der Waerden (1957). After
some minor manipulation, the expression for the experimental thermal diffusion factor from the
trennschaukel is obtained. ._ _. . _/ k I«/I_L_AT\
fail =inQ/n y «n (1+ATj)
L^Ljexpti -r 2, —
I K i = i |c
Here Q is the overall separation with corrected compositions, n is the number of trennschaukel tubes, and k is
the number of stations (in this case four) at which the temperature gradient is monitored by differential
thermocouples.
Feed gas mixtures were obtained for helium, neon, and argon by mixing equimolar amounts of 3He/4He,
20Ne, and 36Ar/40Ar, which are available from Mound Laboratory's stable isotope inventory. Special mixtures
containing approximately equal amounts ( = 5%) of the end isotopes of krypton (78Kr/85Kr) and xenon
(124Xe/138Xe) were produced in Mound's thermal diffusion systems. The isotopic composition of the samples
was measured by repetitively scanning the peaks of the desired mass numbers on a mass spectrometer and
averaging the isotope ratios obtained.
THEORETICAL
The theoretical expression for a y is most readily expressed in Chapman's determinant notation,
[aT1 =i lim (X, X2Aoo(m))-' L A < •» > f / I,/
L TJm 2 _^ oo SX, AO] / 72
(m)r / li I MM, + M2)/2M,J
+ X2Ao —i L (Mi+ M2)/2M2J J.
The quantities AJJ ( m' are determinants obtained from a master determinant A(m) of order 2m + 1 by
striking out row i and column j. Here X is the mole fraction, M is the molecular weight and m is the order of the
approximation. The elements of A< m ), which are complicated functions of the transport collision integrals,
molecular diameters, and molecular weights, have been given for up to the third order, [a T] . by Mason
(1957). The theoretical values computed herein are [aT] 3, except where otherwise noted.
In the event the gas has several isotopes, it is sometimes convenient to remove the mass dependence, or
convert to another mass pair, by use of a reduced thermal diffusion factor
Ct0 = L (M, + Mj ) /(M, -M, ) J (XT .
Where the necessary values of
-------
RESULTS AND DISCUSSION
The experimental and theoretical results of xenon are shown in Figure 2. For several years it has been
suspected that the values for the isotopic thermal diffusion factor for xenon, which were reported by Paul, et
al, (1965b), were in error. Ross and Rutherford (1970) studied the separation of xenon isotopes in a thermal
diffusion column and obtained experimental values for the column coefficients HQ , KC' , and K<-| . They also
evaluated these coefficients theoretically using the best available transport property data. The values of the
two remixing coefficients, Kc', and Kd , which are not dependent upon the thermal diffusion factor, were
found to be in good agreement with their experimental values. The values of H0', the initial transport
coefficient, were calculated using the data of Paul and were in poor agreement with experimental results. Roos
then proceeded to make a corresponding states plot of «o(the reduced thermal diffusion factor) for all of the
noble gases (except helium) and noted that this plot indicated values for cfclower than those Paul reported.
Using the corresponding states predictions, they were able to reconcile the Difference between their
experimental and theoretical values of HO' -Recently, Rutherford (1973) made additional separation
measurements on xenon isotopes in a precisely constructed thermal column and extracted values of OT from
the column performance which were approximately 40% lower than Paul's data. We have performed the
measurements from an average temperature of 250 to 850°K and substantiate the predictions of Roos and
Rutherford. Furthermore, the theoretical calculations of causing either the Parson, et al., (1972) or the Barker,
et al., (1971) potentials agree with the present results within the experimental uncertainty.
The results for krypton are shown in Figure 3. There are no major disagreements among the experimental
data except at low temperatures. Below 300°K the present results are higher by a factor of two over those of
Paul and Watson (1966). The higher value was predicted by Rutherford (1973) and is also substantiated by a
semi-empirical method due to Weissman and DuBro (1970) using isotopic diffusion data. Once again, the
potentials of Parson, et al., (1972) and Barker, et al., (1971) represent the data best. The M.B. (exp-6) potential
with parameters by Mason (I960) fails at low temperatures where it falls well below the new values.
Most of the experimental data for argon are in agreement within the mutual experimental uncertainty with
the exception of de Vries and Laranjeira( 1960) and van derValk and de Vries (1963). Surprisingly, none of the
potentials investigated represents the data particularly well. The Parson, et al., (1972) and Barker, et al.,
(1971) potentials accurately predict the behavior of OtT at low temperatures, but fall considerably lower than
the data above approximately 500°K. The M.B. (exp-6) potential with parameters by Mason and Rice (1954b)
is excessively high at intermediate temperatures, and the Lennard-Jones potential derived from second virial
coefficient data is inadequate as in the previous cases.
A considerable yumber of workers have reported experimental results for neon which are shown in Figure 5.
The temperature range of the data spans 1,100°K which provides a rigorous test of the potential.
Unfortunately, much of the data scatter widely; and without some selective process very little information
can be obtained concerning the intermolecular forces. The present data pass roughly through the median
results of other workers, and since the Mound Laboratory data have appeared to be accurate for the other
noble gases, behavior of the theoretical curves is compared to the present measurements. The experimental
thermal diffusion factor appears to reach a plateau between 0.026 and 0.027 at high temperatures. The Parson,
et al., (1972) and Barker, et al., (1971) potentials never approach this level and fall off markedly at the higher
temperatures. The Lennard-Jones potential predicts values which are too high over the entire range. The
potential due to Mason and Rice (1954b) is probably the best choice, but approximates the data well only in the
intermediate temperature range. We find no really satisfactory potential for neon.
The results of helium, which have been reported elsewhere by Taylor and Weissman (1971) and Taylor
(1973), are given in Figure 6 for the sake of completeness. The Mound Laboratory data, which span 900°K are
in good agreement with other workers, except at very low temperatures, where it is felt that the data of van der
Valk (1963) are erroneously high (see Taylor, 1973). The potentials given by Beck (1969) and Bruch and McGee
(1967) represent the data quite well, but from a comparison with other transport properties by Taylor and
Keller (1971), we have a slight preference for Beck's potential. The Lennard-Jones and (exp-6) potentials are
entirely inadequate for helium.
CONCLUSIONS
We have measured the isotopic thermal diffusion factors for the noble gases over an extended temperature
range and have computed theoretical thermal diffusion factors from a variety of intermolecular potentials.
Because thermal diffusion is a property which is relatively sensitive to the nature of the intermolecular forces,
we have compared the theoretical values to the experimental data in order to test the validity of various
potentials. The Lennard-Jones and (exp-6) potentials are generally inadequate. We found that potentials
proposed by Parson, et al., (1972) and Barker, et al., (1971) quite accurately predict CT for xenon and krypton,
but fail at high temperatures for argon and neon. This leads us to the conclusion that these workers have, for
the noble gases, characterized the well of the potential quite well, but have failed to obtain an accurate
representation of the repulsive region of the potential. This follows from the successively poorer behavior of
their potentials at higher reduced temperatures. Helium is somewhat of a separate problem, and for this gas
we prefer Beck's potential.
-407-
-------
REFERENCES
Barker, J. A., R. A. Fisher, and R. O. Watts, (1971), Liquid Argon: Monte Carlo and Molecular
Dynamics Calculations,Mol.Phys.2l,657.
Seattle, J. A., R. J. Barriault, and J. S. Brierley, (1951), Compressibility of Gaseous Xenon. II. The
Virial Coefficients and Potential Parameters of Xenon, J. Chem. Phys. 19, 1222.
Beck, D. E., (1969), Interatomic Potentials for Helium and Molecules of Helium Isotopes, J. Chem. Phys.
50,541.
de Boer, J. and A. Michels, (19318), The Quantum-Mechanical Theory of the Equation of State and the
Law of Corresponding States. Determination of the Law of Force of Helium, Physica 5, 945.
Bruch, L. W. and I. J. McGee, (1967), Semi-Empirical Potential and Bound State ofHelium-4 Diatom, J.
Chem. Phys. 46, 2959.
Buckingham, R. A., (1938), The Classical Equation of State of Gaseous Helium, Neon and Argon, Proc.
Roy. Soc. (London) A168, 264.
Clusius, K. and M. Huber, (1955), Die Trennschaukel Thermodiffusion-Faktoren in System Co2 /IP, Z.
Naturforsch 10a, 230.
Cunha, M. A. and M. F. Laranjeira, (1972), High-Temperature Isotopic Thermal Diffusion of
Neon,Physica 57, 306.
DiPippo, R. and J. Kestin, (1968), Proceedings of the Fourth Symposium on Thermophysical Properties
(American Society of Mechanical Engineers, New York) 304.
Fischer, A., (1959), Messung Wahrer Thermodiffusionsfaktoren mitder Trennschaukel, Doctorial Thesis
(Zurich).
Grew, K. E. and J. N Mundy, (1961), Thermal Diffusion in Some Mixtures of Inert Gases, Phys. Fluids 4,
1325.
Hirschfelder, J. O., C. F. Curtiss, and R. B. Bird, (1954), Molecular Theory of Gases and Liquids, John
Wiley and Sons, New York.
Keller, J. M. and W. L. Taylor, (1969), Evaluation of Transport Properties of Gases: The Bruch-McGee
Potential of Helium, J. Chem. Phys. 51, 4829.
Laranjeira, M. F. and J. Kistemaker, (1960), Elementary Theory of Thermal and Pressure Diffusion in
Gaseous Binary and Complex Mixtures. III. Ternary Mixtures, Physica 26, 431 .
Laranjeira, M. F. and M. A. Cunha, (1966), Isotopic Thermal Diffusion in Neon, Port. Phys. 4, 281 .
Lee, Y. T., (1973), Intermolecular Potentials, Proceedings of the 1973 Conference on Gas Kinetics, 27-8.
Mason, E. A. and W. E. Rice, (1954a), Intermolecular Potentials of Helium and Hydrogen, J. Chem.
Phys. 22, 522.
Mason, E. A. and W. E. Rice, (1954b), Intermolecular Potentials of Some Simple Nonpolar Molecules, J.
Chem. Phys. 22, 843.
Mason, E. A., (1957), Higher Approximations for the Transport Properties of Binary Gas Mixtures. I.
General Formulas, J. Chem. Phys. 27, 75.
Mason, E. A., (1960), Redetermination of the Intermolecular Potential for Krypton, J. Chem. Phys. 32,
1832.
Mclnteer, B. B., L. T. Aldrich, and A. O. Nier, (1947), The Thermal Diffusion Constant of Helium and
the Separation of3He by Thermal Diffusion, Phys. Rev. 72, 510.
Michels, A. and H. Wijker, (1949), Isotherms of Argon Between O and 150* and Pressures Up to 2900
Atmospheres, Physica 15, 627.
Moran, T. E. and W. W. Watson, (1958), Thermal Diffusion Factor for the Noble Gases, Phys. Rev. 109,
1184.
Neufeld, P. D. and R. A. Aziz, (1973), Tables of Collision Integrals for the Barker-Pompe, Barker-Fisher-
Watts, and Parson-Siska-Lee Inert Gas Potentials, J. Chem. Phys. 58, 1877.
Nier, A. O., (1940), Coefficient of Thermal Diffusion of Neon and Its Variation with Temperature, Phys.
Rev. 57, 338.
Parson, J. M., P. E. Siska, and Y. T. Lee, (1972), Intermolecular Potentials from Cross-Beam
DifferentialElasticScatteringMeasurements.IV.Ar+Ar,3.Chem.Phy8.56,1511.
Paul, R., A. J. Howard, and W. W. Watson, (1963), Isotopic Thermal Diffusion Factor of Argon, J.
Chem. Phys. 39, 3053.
Paul, R., A. J. Howard, and W. W. Watson, (1965a), Concentration Dependence of the Thermal
Diffusion Factor for 3He-4He andH^-D^J. Chem. Phys. 43, 1622.
Paul, R., A. J. Howard, and W. W. Watson, (1965b), Isotopic Thermal Diffusion Factor for Xenon, J.
Chem. Phys. 43, 1890.
Paul, R. and W. W. Watson, (1966), Isotopic Thermal Diffusion Factor for Krypton, J. Chem. Phys. 45,
Raman, S., S. M. Dave, and T. K. S. Narayanan, (1965), Thermal Diffusion Factor of Neon from
Column Measurements, Phys. Fluids 8, 1964.
Roos, W. J. and W. M. Rutherford, (1970), Separation of Xenon Isotopes in the Thermal Diffusion
Column, J. Chem Phys. 52, 1684.
Rutherford, W. M., (1973), Isotopic Thermal Diffusion Factors of Ne, Ar, Kr, and Xe from Column
Measurements, J. Chem. Phys. 58, 1613.
-408-
-------
Saxena, S. C., J. G. Kelly, and W. W. Watson, (1961), Temperature Dependence of the Thermal
Diffusion Factor for Helium, Neon, and Argon, Phys. Fluids 4,1216.
Stevens, G. A. and A. E. de Vries, (1968), The Influence of the Distribution of Atomic Masses Within the
Molecule of Thermal Diffusion, Iso topic Methane, and Methane/Argon Mixtures, Physica39,346.
Stier, L. G., (1942), The Coefficients of Thermal Diffusion of Neon, Argon, and Their Variation with
Temperature, Phys. Rev. 62,548.
Taylor, W. L. and S. Weissman, (1971), Thermal Diffusion Factors and the 4He-3He System, J. Chem.
Phys. 55,4000.
Taylor, W. L. and J. M. Keller, (1971), Transport Properties of Helium Using Beck's Potential, J. Chem.
Phys. 54,647.
Taylor, W. L., (1973), Thermal Diffusion Factor for the 3He-4He System in the Quantum Region, J. Chem.
Phys. 58,834.
van der Valk, F., (1963), Thermal Diffusion in Ternary Gas Mixtures, Doctoral Thesis (Amsterdam).
van der Valk, F. and A. E. de Vries, (1963), Thermal Diffusion in Ternary Mixtures. II. Experiments,
Physica29,427.
de Vries, A. E. and M. F. Laranjeira, (1960), Thermal Diffusion of Ternary Mixtures, J. Chem. Phys. 32,
1714.
van der Waerden, B. L., (1957), Theorie der Trennschaukel, Z. Naturforsch 12a, 583.1 Watson, W. W.,
A. J. Howard, N. E. Miller, and R. M. Shiffrin, (1963), Isotopic Thermal Diffusion Factors for Helium
andNeon at Low Temperatures, Z. Naturforsch 18a, 242.
Weissman, S.,(1965), Thermal Diffusion Factors for 4He-3He at Low Temperatures, Phys. Fluids 12,2237.
Weissman, S. andG. DuBro, (1970), Self-Diffusion Coefficients for Krypton, Phys. Fluids 13,2689.
TABLE 1. Intermolecular Potential Parameters
for the Noble gases.
Gas Source f/k a rm
Xenon 2ndVirial 221. 4.100
Thermal Diff. 257. - 3.963
Mol.Beam 282.5 -- 4.43
Liquid Data 280.3 - 4.358
Krypton 2iidVirial 172.7 3.591
Transport Prop. 200.0 - 4.036
Mol.Beam 201.5 - 4.100
Liquid Data 199.7 -- 4.007
Argon 2ndVirial 119.8 3.405
Transport Prop. 123.2 -- 3.866
Liquid Data 142.1 - 3.761
Neon 2ndVirial 35.60 2.749
Transport Prop. 38.0 - 3.147
Mol.Beam 37.15 - 3.17
Liquid Data 36.82 - 3.142
Helium 2ndVirial 10.22 2.556
Viscosity 86.20 2.158
Transport Prop. 9.16 - 3.135
Viscosity 36.1 - 2.725
2ndVirial 10.38 - 2.96
Transport Prop. 12.13 -- 2.975
Reference
Beattie, etal, (1951)
Paul, et al, (1965a and b)
Parson, etal, (1972)
Barker, et o/.,(1971)
Beattie, et al., (1951)
Mason (1960
Parson, et al., (1972)
Barker, et al., (1971)
Michels and Wijker (1949)
Mason and Rice (1954b)
Barker, et al., (1971)
Buckingham (1938)
Mason,e£a/.,(1954b)
Parson, et al., (1972)
Barker, et al., (1971
deBoer and Michels (1938)
DiPippo and Kestin (1968)
Mason and Rice (1954a)
DiPippo and Kestin (1968)
Beck (1969)
Bruch and McGee (1967
-409-
-------
Figure 1. Schematic diagram of apparatus. A, pump drive; B, environmental chamber; C, bellows pump and
trennschaukel; D, gas sampling manifold; E, feed gas supply and evacuation system; F, air or liquid nitrogen
supply to adjust TC .
-410-
-------
0.025
0.020
0.015
0.010
0.005 -
200
400 500 600
TEMPERATURE (°K)
700
800
900
Figure 2. Thermal diffusion factor 134Xe-136Xe. The experimental points are: •, present data; X, Paul, et al.,
(1965b); and D, Rutherford (1973). The theoretical curves are: Parson, et al., (1972); Beattie, et
al, (1951);—• Paul, et al, (1965a); and .Barker, etal, (1971).
-------
0.025
0.020
0.015
0.010
0.005
KRYPTON
200
400 600 800
TEMPERATURE (°K)
1000
1200
Figure 3. Thermal diffusion factor for 78Kr-86Kr. The experimental points are: •, present data; X, Paul and
Watson (1966); and a, Rutherford (1973). The dotted line represents the semi-empirical values of aT obtained
by Weissman and Dubro (1970) from isotopic diffusion data. The theoretical curves are: .Parson, et al.,
(1972); ,Beattie,e£a/.,(1951);— .Mason (1960); and .Barker, etal, (1971).
-412-
-------
0.030
0.025
0.020
0.015
0.010
0.005
ARGON
200
400 600 800
TEMPERATURE (°K)
1000
1200
Figure 4. Thermal diffusion factor for 38Ar-40Ar. The experimental points are: •, present data; X, Paul, et al,
(1963); n, Rutherford (1973); V, Stevens and de Vries (1968); A, van der Valk and de Vries (1963); and A, de
Vries and Laranjeira (I960). The theoretical curves are: , Parson, et al., (1972); , Barker, et al.,
(1971); .Michels and Wijker (1949); and .Mason and Rice (1954a and b).
-413-
-------
0.035
0.030
0.025
0.020 -
0.015 -
0.010
400 600 800
TEMPERATURE (°K)
1000
1200
Figure 5. Thermal diffusion factor for 20Ne-22Ne. The experimental points are: •, present data; V, Saxena, et
al, (1961); +, Nier (1940);*, Stier (1942);$, Moran and Watson (1958); X, Fischer (1959);C, Laranjeira and
Kistemaker (1960);O, Grew and Mundy (1961); a, Watson, et al., (1963);A, Laranjeira and Cunha (1966),<=>,
Cunha and Laranjeira (1972); O, Rutherford (1973); andV, Raman, et al., (1965). The theoretical curves
are: , Parson, et al., (1972); , Barker, et al, (1971); , Mason and Rice (1954b);
and .Buckingham (1938).
-414-
-------
0.10
0.09
0.08
0.07
0.06 -
0.05 .
100
200
300
400 500 600
TEMPERATURE (°K)
700
800
900
Figure 6. Thermal diffusion factor for 3He-4He. The experimental points are:0 , present data; A , van der Valk
(1963);« , Watson, et al., (1963); a , Paul, et al., (1965a);Q , Saxena, et al, (1961); V , Mclnteer, et al, (1947); and . .
. ., the smoothed data of Weissman (1969). The theoretical curves are: LJl.de Boer and Michels (1938); LJ2, Di
Pippo and Kestin (1968); MB1, Mason and Rice (1954a); MB2, Di Pippo and Kestin (1968); -- , Beck
(1969); and ---- , Bruch and McGee (1967).
-------
VII. Biological Effects of Noble Gases
-------
PHYSIOLOGY OF THE NOBLE GASES
R. M. Featherstone,*
W. Settle, and H. Althouse
Department of Pharmacology
University of California
San Francisco, California
Abstract
The noble gas series of elements, for many years considered "inert" in a chemical sense, are known to have
biologic activities that must be considered in any attempt to assess the possible hazards of having radioactive
isotopes of these elements liberated in our environment. Although most is known about the biology of xenon,
reports of effects from low concentrations of helium have been made. The anesthetic properties of xenon will
be reviewed in some detail including the possible biologic implications of xenon's binding at very specific
internal sites of proteins and the effects of this binding on the affinity of heme proteins. The surprising
observations that helium substituted for nitrogen at ordinary pressures reduced the baseline heart rate and
the concentration of endogenous plasma catecholamines of dogs will also be discussed. Possible mechanisms
by which these gases might be altering normal biologic responses will be presented. Biologic effects of neon,
argon, and krypton are less well-documented, but the theoretical aspects of their possible associations in
biological systems are discussed.
INTRODUCTION
One of the important questions to be addressed during this symposium concerns the biological effects of the
noble gases. Are there such effects? If so, what are they? And what are the implications of these effects,
particularly when some of these gases may be at times represented in our environments by some of their
radioactive isotopes?
We shall describe briefly the relatively little that is known about the physiology or biology of these gases
without specific reference to the factor of radioactivity, which will be considered by others.
It is not surprising that we know most about the two ends of our scale — helium and xenon. (Radon will
receive special attention later in this symposium). During this presentation we shall mention briefly the scope
of the studies on the whole series of gases, then describe some specific studies with xenon and helium, and
conclude with comments on the dilemma confronting us with respect to neon, argon, and krypton.
As tools to probe some of the complexities of molecular mechanisms in biology, the noble gases are ideal for
focusing on the smaller binding forces among molecules — the non-covalent, non-ionic, and non-hydrogen
bonding forces. Also, with these noble gas elements there is never the necessity of considering the possibility
of a metabolite whose identity has escaped detection.
Studies have shown that, over a range of pressures appropriate for the relative size of the noble gas atoms,
and the sensitivity of the test situation, that the activities of a number of enzymes, including tyrosinase,
lipoxidase, acetyl-cholinesterase, alpha-chymotrypsin, leucine amino peptidase, (Powell, 1973) and ATPase
(Trevor, etal., 1969) are all affected by the noble gases. On the whole cell level, Neurospora crassa growth,
(Powell, 1973) cell division of HeLa cells growing in culture (Powell, 1973) and the growth of some Escherichia
coli strains (Hegeman, 1973) are all inhibited by the noble gases at appropriate pressures.
Physiological effects on nervous activity have been studied in several situations. Gottlieb and others have
shown a reversible inhibition of nerve conduction by 200 psi xenon and 950 psi krypton (Gottlieb, et al, 1968).
Xenon anesthesia in man has been known and studied extensively for over 20 years (Featherstone, et al.,
1963). More recently, the tremors and convulsions of the "high pressure neurological syndrome," caused
apparently by helium in environments simulating greater than 600 feet of sea water, has elicited considerable
attention (Bennett, et al., 1971).
The first question, then, of the presence of altered physiology in the presence of supernormal amounts of at
least some of these noble gases has to be answered affirmatively.
XENON EFFECTS
Let us consider xenon first in some detail. The fact that it can produce anesthesia in man at 0.8 atm., with
room for adequate oxygen, and in other organisms at higher pressures (monkeys at 2 atm., dogs at 3 aim.,
rats at 5 or 6 atm., etc.) has been the subject of studies in our laboratory and others for over 20 years. Early in
this period we used radioactive xenon (the first produced by neutron bombardment) to learn that xenon was
taken up by equal rates by the cerebral cortex, the caudate nucleus, the thalamus, the hypothalamus, and the
medulla oblongata of the brain, as well as by the kidneys. The adrenal gland, for some unknown reason,
absorbed xenon to a much greater extent than did the brain, and other tissues absorbed it more slowly and to
lesser degrees, depending more or less on vascularity (Featherstone, et al., 1952; Pittinger, et al., 1954). We also
learned that blood carries almost half of the xenon bound to hemoglobin.
*0n sabbatical during 73-74 in Geneva. Department of Physiology, University of Geneva, Switzerland.
-416-
-------
We shall summarize briefly what we have learned about the nature of the association of xenon with some
macromolecules of known biological importance — a system that is serving us well as a model for still
unknown systems that are possibly even more important. Comparisons can be made of the size of xenon with
other anesthetic molecules and with the gases that make up our normal environment. Xenon is about the size
of nitrous oxide and ethylene; it is much smaller than chloroform, ether, or halothane; and larger than
hydrogen, oxygen, nitrogen, and carbon dioxide. The only mechanism seemingly possible is the attraction of
the electrons of xenon to a dipole, thereby having a dipole induced in the xenon, which, in turn, can attract
other electrically neutral molecules, inducing in them also a dipole moment, Both the size and polarizability of
these molecules are important and are closely related.
Interactions of anesthetics like xenon with proteins can now be explored at a molecular level. Myoglobin, a
small protein whose three dimensional structure is known almost completely, is a compact unit with little
empty space inside. Xenon, a tiny molecule by comparison with myoglobin, combines with some specific sites
with myoglobin. Two xenon binding sites have been identified by x-ray diffraction studies (Settle, 1973). The
first binding site is nearly equidistant between the proximal histidine and the heme. The binding energy of
this site approximates 10 kcal/mple (roughly equivalent to two ionic bonds), but composed of over 30 small
additive forces. The second site is nearer the surface but still completely within the protein molecule —
between the AB and GH corners. In hemoglobin, which is made up of four units similar to myoglobin, binding
of xenon has been observed in the second region. In the alpha chain of hemoglobin, the binding site is nearer
the GH corner; in the beta chain the site is nearer the AB corner. These sites are inside and near the surface,
and are composed essentially of the same non-polar amino acid side chains as those of the myoglobin sites.
They compose in each case a small lipid-like pocket.
Studies with lysozyme, another protein whose structure is almost completely known, have shown that
xenon does not associate with this protein — there are no suitable environments in it. Our preliminary
experiences with some other proteins indicate that some proteins do associate specifically with xenon.
Myoglobin combines with oxygen and carbon monoxide on the side of the heme opposite the first xenon site.
When xenon is present, the binding of carbon monoxide is altered. Binding constants of xenon to deoxy and
carboxy myglobin indicate that xenon in the site near the heme causes a small increase (10%) in CO affinity;
and xenon in the site near the AB and GH corners (see Figure 1.) causes a much larger increase in CO binding
(90%)(Settle, 1973). Another site is known to exist in some forms of myoglobin, but it has not yet been
identified, although the xenon in it is known to decrease CO affinity.
The diagramatic representation Perutz has recently published on the movement of a hemoglobin subunit
upon ligand binding (Perutz, 1970) can be considered in relation to xenon and carbon monoxide binding.
Remember that the primary xenon site is near the heme and histidine of the F helix, and that the secondary
site is near the GH corner. Upon ligand binding the heme becomes planar, and the distance between the heme
and the F helix decreases by 0.9 A. This causes a movement of the F helix 2 A toward the G helix, with the F
helix pivoting at the FG corner and middle of the E helix, thus producing movements of several angstroms in
the region of the secondary xenon site between the AB and GH corners. One can postulate that xenon, when
present, can alter the affinity of a ligand (like carbon monixide) by perhaps holding the hemoglobin in a more
favorable state for liganding to occur or to be maintained. It is interesting to note that xenon alone, however,
does not change the structure of hemoglobin merely by its presence.
This postulate about the mechanisms by which xenon causes increased liganding of carbon monoxide is
supported by recent observations (Stewart, et al., 1972) that methylene chloride (dichlorpmethane) breathed
for two hours in low concentrations (probably equivalent to the exposure received from using the compound to
remove paint from a small piece of furniture) will cause a tenfold increase in the carboxyhemoglobin level, and
that Benno Schoenborn (Pifarre, et al., 1970), of Brookhaven National Laboratory, has shown that this
compound, which is only slightly larger than xenon, combines with heme proteins in the "xenon" sites.
HELIUM EFFECTS
Let us turn now to some biological consequences of the presence of helium in a biological system. Here, the
observations are mainly phenomenological at this stage. Five examples are cited briefly.
(1) In diving gas mixtures, the use of helium has led to an avoidance of nitrogen narcosis allowing man to
descend to far greater depths, although not without limit, as will be indicated later. No depressant effects have
been observed on the respiratory centers for up to 19 atmospheres of helium (Powell, 1973). However, it must be
remembered that these centers are among the last to be affected in the CNS by foreign substances like
anesthetic molecules. Consequently, such a measure is not ideal if one is looking for some effects of helium —
more sensitive systems are necessary.
(2)A series of three papers, published since 1968, describe the effects of helium. In the work described in the
first two papers (Pifarre, et al., 1970) ventricular fibrillation was induced in a series of dogs by acute occlusion
of the left circumflex artery. The surprising observation was made that the presence of 20% helium was
sufficient to block these ventricular fibrillations. The data do not show clearly that more oxygen was not also
necessary, but the implication of helium seems probable. The authors propose that helium (perhaps coupled
with more oxygen) promoted in some way the development of coronary collateral circulation so that more
blood was available to the ischemic area and the infarcted area was smaller. Work reported in the third paper
(Raymond, et al, 1972) was with dogs not subject to circumflex ligation. The sensitivity of blood pressure,
heart rate, and extrasystoles to experimental epinephrine IV was not altered by helium. However, helium
(75%) did reduce baseline heart rate (112/119) and concentration of endogenous catechol amines. These
-417-
-------
authors concluded that the antiarrhythmic effect of helium may be mediated by changes in symphathetic
These studies are of probable clinical importance in reducing the liklihood of ventricular arrhythmias in
persons having myocardial infarctions, but for purposes of this symposium these effects illustrate a clear
example of a biological effect from as little as 20% helium in the inspired gases.
(3) Hamilton and Schreiner (Hamilton, et al, 1970 a and b) and their colleagues have studied the
biochemical and metabolic effects of a six-month exposure of small animals to helium-oxygen mixtures. In
rats and two generations of mice they found no changes except for greater food and water uptake by the mice,
possibly due primarily to the differing heat transfer properties of helium. In rabbits they also observed
decreases in hemoglobin and red cell levels in blood.
(4) Another example of helium effects is that studied by a group working in Geneva and Lausanne (Sauter,
et al., 1973). They are stimulating the isolated cervical sympathetic ganglion of rats preganglionically at
6/second and measuring the electrophysiological answer on the postganglionic nerve. Stimulation is
maintained for 30 minutes followed by 30 minutes of rest. The control responses in 95% O2 and 5% CO2 at 1
atm. are that the height of the action potential falls at once to 70% of the initial value and then, during the
stimulation period, increases to 85% of the initial value. After the stimulus is stopped, there is a
hyperexcitability lasting for 30 minutes where the action potential is above the initial value. When 6 atm. of
helium are added, with PO2 and PCO2 corrected to control levels, the same fall in action potential at the
beginning of the stimulus period is observed, but the recuperation does not occur, and the height of the action
potential remains below the initial value for at least 30 minutes after stimulation is stopped. This work is still
in its preliminary stages. Pressure alone has not been eliminated as a possible cause of the effects, but these
experiments may constitute another example of helium effects.
(5) The last example of helium effects to be cited today is the "high pressure neurological syndrome" noted in
divers at simulated depths below 600 feet of sea water (Bennett, et al., 1971). This syndrome is characterized
chiefly by tremors and convulsions. Its existence probably has little significance for the topic of this
symposium, but it, like the other examples cited, helps raise questions about the molecular mechanisms by
which helium and xenon can cause changes in physiology.
CONCLUSIONS
Now, having described in some detail the molecular mechanisms by which xenon can alter physiology, and
having mentioned some examples of effects by helium, the question arises as to whether the other members of
this series, neon, argon, and krypton, cause changes in physiology, and if so, do they all do so by the type of
interaction shown for xenon, although, of course, through interactions with different sized receptive
environments at the submolecular level.
At present we are left with the guess that all members of this series can interact to some degree with proteins
and other biologically important molecules, and that such interactions are involved in any changes that
occur. Those changes with helium and xenon are most noticeable — these two are farthest away from nitrogen
in size and other properties, and nitrogen is the gas with which we are always living unless we replace it or
remove it. It is not surprising that neon, like helium, can be used in diving to avoid nitrogen narcosis. It is also
not surprising to find that krypton, at pressures higher than xenon, produces some degree of anesthesia. The
data on Neurospora and HeLa cells help predict these events. Perhaps argon is so close to nitrogen in size and
properties that it will be shown to have almost no unique biological properties in complex multicellular
organisms like man. Schatte (Schatte, etal., 1972) has indicated that argon substituted for nitrogen would
require a 10% decrease in oxygen tension to keep the P-arterial tension of oxygen equal to that in air. Likewise,
he predicts a 5.6% increase of oxygen would be necessary if helium was used instead of nitrogen, concluding
that argon and helium may significantly alter arterial oxygen tension.
In any event, the demonstrated effects of the gases in this series, both progressive effects with size in the
studies with whole cells, and the more pronounced effects of helium and xenon, suggest that members of this
series do combine, in the chemically subtle ways of which they are capable, with biologically important
molecules and may change physiological functions — for better or worse — depending perhaps on their
deviation from nitrogen and, of course, on their concentrations in some key biological sites.
Actually, relatively little is known about the physiology of these noble gases. If one adds a concern about
the association of radioactivity with them, it seems apparent that all the members of this series need much
further study, with appropriate support from government agencies and industries associated with any of
these fascinating noble gases.
REFERENCES
Bennett, P. B. and E. J. Towse, (1971), The High Pressure Nervous Syndrome During A Simulated
Oxygen-Helium Dive to 1500 feet. Electroenceph. Clin. Neurophysiol. 31 383-393
Featherstone, R. M., W. Steinfield, E. G. Gross, and C. B. Pittinger, (1952), Distribution of the
Anesthetic Gas Xenon in Dog Tissues as Determined with radioactive Xenon. J. Pharmacol. and Exper.
Therap. 106, 468.
Featherstone, R. M. and C. A. Muehlbaecher, (1963), The Current Role of Inert Gases in the Search for
Anesthesia Mechanisms. Pharmacol. Reviews. 15, 97.
-418-
-------
Gottlieb, S. F., A. Cymeran, and A. V. Metz, Jr., (1968), Effect of Xenon, Krypton, and Nitrous Oxide
on Sodium Active Transport through Frog Skin with Additional Observations on Sciatic Nerve Conduction
Aerospace Med., 39,449-453.
Hamilton, R. W., et al, (1970), Biochemical and Metabolic Effects of a 6 months Exposure of Small
Animals to a Helium-Oxygen Atmosphere. Space Life Sci., 2,57.
Hamilton, R. W., et al., (1970), Biological Evaluation of Various Spacecraft Cabin Atmospheres. Space
Life Sci., 2,307 and 407.
Hegeman, S. L., (1973), Use of Genetic Variants in Molecular Pharmacology, Chapter in Part II of A
Guide to Molecular Pharmacology-Toxicoloty, 561-582.
Perutz, M. F., (1970), Stereochemistry of Cooperative Effects in Hemoglobin. Nature 228,726-734.
Pifarre,R.,e£aZ., (1970), The Influence of Oxygen and Helium Upon Ventricular Fibrillation. J. Thoracic
and Cardiovascular Surgery, 55,535-537,1968, and 60,648-652,1970.
Pittinger, C. B., R. M. Featherstone, E. G. Gross, E. E. Stickley, and L. Levy, (1954), Xenon
Concentration changes in Brain and Other Body Tissues of the Dog During the Inhalation of the Gas.
Pharmacol. and Exper. Therap., 110,458.
Powell, M. R., (1973), The Role of the Noble Gases in Molecular Pharmacology, Chapter in Part II of A
Guide to Molecular Pharmacology-Toxicology., pp. 495-518.
Raymond., L., et al., (1972), Possible Mechanism for the Anti-Arrhythmic Effect of Helium in
Anesthetized Dogs. Science, 176,1250-1252.
Sauter, F., M. Dolivo, and J. M. Posternak, (1973), University of Lausanne and University of Geneva.
Personal communication.
Schatte, C. L., et al., (1972), Predictability of PaO2 in Different Inert Gas-Oxygen Environments. Space
Life Sci., 3,206.
Settle, W., (1973), Function of the Myoglobin Molecule as Influenced by Anesthetic Molecules. Chapter in
Part II of A Guide to Molecular Pharmacology-Toxicology, pp. 477-494.
Stewart, R. D., et al., (1972), Carboxyhemoglobin Elevation After Exposure to Dichlormethane. Science,
176,295-296.
Trevor, A. J. and J. T. Cummins, (1969), Properties ofNA and K Activated ATPases of Rat Brain —
Effect of Cyclopropane and Other Agents Modifying Enzyme Acitivity.Biochem. Pharmacol. 18,1157-1167.
Note: Inasmuch as a number of the references are included in the chapers of a recently published book edited
by one of the auothors of this article (RMF), these chapters are cited in an effort to bring to the interested
reader a far greater bibliography than can be reproduced here.
Figure 1. Diagrammatic sketch of sperm whale myoglobin showing the course of the polypeptide chain from
the same view as the photograph of a model shown by Settle, 1973. The amino acid residues noted in the sketch
are those which have side chains within the van der Waals contact of the xenon atoms. (Reprinted by
permission of Marcel Dekker, Inc., New York, from A guide to Molecular Pharmacology-Toxicology, edited by
R. M. Featherstone, 1973.)
-419-
-------
CALCULATIONS OF THE ABSORBED DOSE TO A MAN IMMERSED IN AN INFINITE CLOUD
OF KRYPTON-85*
W. S. Snyder, L. T. Dillman, M. R. Ford, and J. W. Poston
Health Physics Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
Abstract
A person exposed to an infinite cloud of krypton-85 is exposed to at least seven sources of radiation: (1)
photons emitted in air, (2) beta particles emitted in air, (3) photons and beta particles emitted by krypton-85
present in the air passages of the lungs, (4) photons and beta particles emitted by krypton-85 present in body
tissues, (5) bremsstrahlung produced in air, (6) bremsstrahlung produced by beta particles incident on the
body, (7) bremsstrahlung produced within the body. The energy spectra of scattered photons are calculated by
the method of Dillman and half these spectra are taken as incident on the body, producing doses to body
organs as in MIRD Pamphlet No. 5. Doses to various organs, e.g., red bone marrow, gonads, etc., from sources
(1) and (5) above are calculated by interpolation from these results. Dose to skin from source (2) is calculated by
integration ofBerger's point kernel (MIRD Pamphlet No. 7) in air. Doses from sources (3), (4), (6), and (7) are
calculated by methods developed by the authors of internal sources of photons and beta particles. If the
intensity in the cloud is 1 fiCi/m3, a dose to skin at a depth of 7 mg/cm2 is *v/.S rad/a; to the lungs, 32 mrad/a;
ovaries, 6.2 mrad/a; and testes, 16 mrad/a.
INTRODUCTION
The calculations of absorbed dose to a person exposed to 85Kr is at once simple and complicated. It is simple
if one is willing to ignore all but the highest dose (i. e., the dose from beta particles incident on the skin), but if
one asks for genetic dose, particularly the dose to the ovaries, it is quite another matter. In the latter only the
calculation of the absorbed dose for a person exposed to a large cloud of 85Kr, which is assumed to be uniformly
distributed in air, is considered. The methods to be described, however, are general and also apply to exposures
to infinite clouds of other noble gases. The problem of exposure to a finite cloud or "plume" of 85Kr is not
considered here, and will likely demand some different techniques.
The decay scheme as usually shown is given in Figure 1, which is taken from MIRD Pamphlet No. 6. Note
that there is only one photon given with an intensity of only 0.41% per disintegration. The Internal conversion
is less than 1% of this, and hence is not shown. Only two betas are emitted and these have endpoint energies of
0.67 MeV and 0.16 MeV. The range of the 0.67 MeV beta in tissue is about 0.14 cm.
In all, six sources of radiation need to be considered:
(1) photons and beta particles emitted directly from the 85Kr present in the air;
(2) bremsstrahlung radiation emitted in the air;
(3) bremsstrahlung radiation emitted by the beta particles as they penetrate the skin;
(4) photons and beta particles emitted by 85Kr present in the air passages of the lungs;
(5) photons and beta particles emitted by 86Kr present in the body; and
(6) bremsstrahlung radiation from beta particles emitted from 85Kr present in the body.
Calculations are based on an anthropomorphic phantom which exists as a mathematical object defined by
inequalities and equations programmed into the computer's memory. The most complete description of the
phantom and of the calculational techniques is that given in MIRD Pamphlet No. 5. A schematic diagram of
the outer surface of the phantom and a cutaway view showing some of the principal organs are given in Figure
2. The attempt here is to give a gross representation of the geometry of the human body without attempting a
detailed representation. The phantom, however, is composed of three types of tissue:
(1) skeletal tissue (bone plus marrow) with specific gravity of 1.5;
(2) general soft tissue with a specific gravity near 1; and
(3) lung tissue with a specific gravity of 0.3.
The elemental composition of these tissues approximates that of the averages for such tissues and is given in
Several changes have been made in the phantom since MIRD Pamphlet No. 5 was published. Those of
particular importance for this study are mentioned briefly here.
(a) Regions of bone have been defined in which the active marrow of an adult is largely present, thus
enabling dose estimation for the active bone marrow.
(b) The legs have been separated to place the male gonads near the surface of the body as is shown in
Figures.
(c) The top of the head has been rounded to remove the excess of soft tissue which would provide unwanted
shielding for the red marrow present in the skull.
*Research sponsored by the U. S. Atomic Energy Commission under contract with Union Carbide
Corporation.
-420-
-------
DOSE FROM PHOTONS PRODUCED IN AN INFINITE CLOUD
The primary photons are scattered by the air, producing a continuum of photons of lower energies. The
energy spectrum of these photons may be calculated as has been shown by Dillman (1970b). Since the cloud is
infinite and the primary photons are distributed isotropically in angle, all subsequent generations of photons
will likewise be distributed isotropically. Thus it suffices to compute this energy spectrum of scattered photons
in energy space. This calculation has been performed for twelve monoenergetic photon sources. The energy
spectrum for 0.5 MeV photons is shown in Figure 4 (this is the spectrum emitted from each volume element of
the infinite cloud and is very nearly the energy of the photon emitted by 85Kr). The photons from each of these
twelve energy and angular spectra were used as a source on the phantom. The orginal monoenergetic source
energies extended from 0.01 MeV to 4 MeV. (Doses for photons of other energies are obtained by linear
interpolation.) The results for each of the twelve sources have been scrutinized, and when the coefficient of
variation of an estimate exceeds 30%, the estimate is rejected, and a value estimated on the basis of the depth-
dose profiles is substituted. The technique used and the data are in course of publication (Poston, et al., 1973).
For most practical purposes, one is concerned with a person on the surface of the ground, and thus a ground
effect must be considered. For an energetic photon (i. e., one having a mean-free path in air which is many
times the diameter of the body), the presence of the man on the earth's surface means that essentially half the
source in air to which he is exposed (half the infinite cloud) is missing. This assumes there are no surface
deposits of 85Kr, as these would be expected to be negligible in view of the chemical characteristics of 85Kr as
one of the noble gases. There is some backscattering of photons from the earth, but direct measurement of 60Co
sources in air indicates this ground effect is perhaps a correction of some 10-20% or less (Haywood, 1964;
Berger, 1957). Of course, the measurements are for essentially a point source, but summation of such sources
indicates an increase of dose of this order due to scattering from the ground. Thus one may use half the
intensity of the energy spectrum for photons without greatly falsifying the dose received. This argument
depends on the fact that even at energies of 10 keV, the mean-free path of photons in air is comparable to the
body height, and hence essentially half the source is missing for irradiation of the lower portions of the body,
and about 25% is excluded by the ground by irradiation of the upper portions of the body. For higher energies,
the ground interface essentially excludes half the source.
The dose to the various body organs from photons emitted by 85Kr in air is shown in the top line of Table 1. A
concentration of 1 uCi/m3 (STP) is assumed, and the dose is shown per day of the person's presence in the
infinite cloud. The dose to the skin from photons is an average for a thickness of 0.2 cm extending over the
entire body. These doses are determined by a Monte Carlo-type calculation using the source described earlier.
DOSE FROM BREMSSTRAHLUNG PRODUCED IN AIR, IN TISSUE, OR IN SKIN
Bremsstrahlung is emitted whenever an electron is produced or whenever the electron passes through a
medium. In fact, the so-called internal bremsstrahlung is emitted from the atom of 85Kr, and the external
bremsstrahlung is emitted as the beta particle passes through air or tissue. The energy spectra of these five
source's of bremsstrahlung all lie within the shaded area shown in Figure 5; that emitted in skeleton (bone plus
marrow) being highest and adipose tissue lowest. The internal bremsstrahlung is combined with the external
for all the media; in each medium the units are per disintegration of 85Kr. Note that there is not great difference
between the spectra of the bremsstrahlung produced in air and in tissue, but that produced in skeleton and fat
varies considerably. The dose from this bremsstrahlung has been obtained by interpolation on doses
computed for the twelve monoenergetic sources of photons mentioned earlier for estimation of dose from a
semi-infinite cloud. These doses are shown in the second line of Table 1.
Electrons incident on skin produce bremsstrahlung as they move through skin, but in this case the electron
only expends a portion of its energy in the skin, and the energy spectrum is computed taking this into account.
The values shown in Figure 6 are for the total number of photons produced in the remaining portion of the
electron's range after it has entered the skin, and the number shown is opposite the median of the
corresponding energy interval. Note that the units here are necessarily quite different than those in Figure 5.
The dose distribution for these photons was obtained by interpolation on the doses due to twelve
monoenergetic sources originating in the skin of the phantom (i. e., in a layer of thickness of 0.2 cm.). Thus the
calculation involves the assumption that these photons are uniformly distributed in angle and in depth,
neither of which is strictly true, but this should not matter greatly for estimation of dose. These doses are
shown in the third line of Table 1.
DOSE FROM MKr IN THE BODY
A person immersed in an atmosphere of 85Kr at 1 ^Ci/m3 would quickly come into equilibrium with it. The
concentration in the body tissues would be the concentration in air multiplied by Ostwald's coefficient —
namely,
(tissue concentration) /iCi/cm3=10-6xr x (concentration in air) fzCi/m3
where r is Ostwald's coefficient. In applying this, one should not consider the air passages of the lung as part
of the lung; thus it is equivalent to
(tissue concentration) jiCi/g=10-6T x (air concentration in ^Ci).
m3
-421 -
-------
Only a few measurements on bone have been reported, and these have been, for the most part, on whole bone
plus marrow. Since the value of T for fat is approximately an order of magnitude in excess of its value for other
tissues, it is likely that the measurements reflect largely the uptake of 85Kr into the marrow.
Some authors (Kirk, 1972; Yeh, et al, 1964) have used a formula
^tissue = rfat xf fat + T blood x (1 -ffat)
where f'fat is the fractional content of fat in the tissue. This procedure is followed here. The total body can be
considered as composed of two tissues — fatty tissue and nonfatty tissue. On the basis of a few measurements
and in vitro studies, the Ostwald coefficient for these two types of tissue is about
Tfat = 0.45 and Tnonfat=0.07.
The tissues considered are shown in Table 2 together with the fat content, the estimated Ostwald coefficient,
and a few measured values. The basis of these values is discussed below in detail. On this basis an average
coefficient for the body is obtained. Actually, however, the concentration is not uniform for the concentrations
in adipose tissue are almost an order of magnitude higher than in most tissues.
Bremsstrahlung is produced in the body by the beta particles emitted from 85Kr present in body tissues.
Because of the greater concentration of 85Kr in fatty tissues, there is a greater source of bremsstrahlung in
these tissues, and the production of bremsstrahlung is taken here as proportional to the concentration (i. e.,
the fact that some portion of the range of the beta particles is outside the organ or tissue of origin is ignored).
Likewise, bremsstrahlung is produced in the skeleton to a greater extent than in soft tissues because of the
higher charge number of many of the constituents of bone. The spectra produced in adipose tissue, soft tissue,
and bone all lie within the shaded area of Figure 5. These have been normalized to the bremsstrahlung
produced by one disintegration in the tissue.
The body is considered as composed of fat and a nonfat tissue with the values cited for the Ostwald
coefficient. According to Tipton (1973), the reference body contains about 15 kg of adipose tissue of which 12.5
kg is separable, i. e., occurring in easily distinguishable masses of adipose tissue such as the subcutaneous
fatty layer, fat about the kidneys, etc. Of this mass, some 10 kg is fat. Using the formula cited earlier, one finds:
Tad = 0.8 x 0.45 + 0.2 x 0.07=0.36 + 0.014 = 0.374.
This adipose tissue is distributed rather generally throughout the body, and hence is assumed here to be
contained in the total body minus the totality of specified organs as defined in the phantom used in MIRD
Pamphlet5.
The skeleton, which here consists of bone plus marrow plus connective tissues, etc., contains a total of 1900 g
of fat, most of which is distributed among 1500 g of yellow marrow. Thus
T skel= 1900x0.45 + 8100x0.07 = 0.0855 + 0.0567 = 0.142.
10,000
The Ostwald coefficient is essentially a ratio of concentrations per ml of the substances, and hence a
correction for average density of the skeleton (including marrow, etc.) is in order. Thus
Tskel/1.5=0.142x2=0.0948.
T
The soft tissues are not devoid of fat. In addition to interstitial fat, there is a lipid constituent of most tissues
which includes the essential fat. In the remainder of the body (i. e., total body minus separable adipose tissue
minus skeleton), there is about 1.5 kg of essential fat out of a total tissue mass of 47.5 kg. Accordingly, the
Ostwald coefficient is given by
Tskel = 1.5 x 0.45 + 46 x O.QJ = 0.0820.
47.5 '
Assuming a concentration in air of 1 /zCi/m3 = 10-6, one has, at equilibrium, concentrations of 0 37 x 10-6
g of separable adipose tissue, 0.095 x 10-6 fid/g of skeleton, and. 0.082 x 10-6 uCi/g of soft tissue As
mentioned above, the concentration in the soft tissue is for the "other tissues" of the body that remain after
skeleton and separable adipose tissue are subtracted. This is equivalent to putting:
0.082 x 10-6 x 70,000 = 5.7 x 10-3 fzCi in the total body;
(0.095 - 0.082) x 10-6 x 104 = 1.3 x 10-4 ftCi in the skeleton
(in addition to the activity already there as part of the total body activity)- and
(0.37 - 0.082) x lO-e x 12,500=3.6 x 10-3 /zCi in the "other tissues" compartment
(in addition to those already there as part of the total body activity).
When this is done, the photon, beta, and bremsstrahlung doses due to 85Kr absorbed in these tissues, and
-422-
-------
present there at this equilibrium level, are found to be the values given in lines 4,5, and 6, respectively, of Table
1. Actually, the calculation of the dose from the bremsstrahlung is slightly more complicated in that it is not
merely the total source strength that must be subtracted, as above, but also the spectrum of the soft tissue dose
with the proper weighting factor (i. e., [0.095 x spectrum of skeleton - 0.082 x spectrum of soft tissue] x 10-6 x 104)
must be used to allow for the difference in spectra of these two media. A similar remark applies for separable
adipose tissue.
BETA PARTICLES IN AN INFINITE CLOUD
The depth dose in skin and subcutaneous tissue from beta particles in an infinite cloud was produced by
integration of the point kernel given by Berger (1971). There is a slight interface effect due to the differences of
density and composition of air and skin at the interface. However, because of the near equivalent of the
average Z of these materials, it is not expected to be more than a few percent of the total dose. This is presently
under investigation by Berger. A graph of the depth dose is shown in Figure 7.
The skin will be considered here as consisting of an epidermal layer (of which the first 10%, on the average,
consists of dead tissue), and a deeper layer, the dennis. Values for the thicknesses of these layers have been
given by several authors for a variety of regions of the body (Southwood, 1955; Whitton, 1973). For example, a
recent series of measurements by Whitton (1973) would indicate values of 30 to 80/im for the epidermis on
various regions of the body, exclusive of hands and feet, with a mean thickness of perhaps 50 fim. This would
indicate a value of -\_5/zm for the stratum corneum. Thus an average dose to skin (-^.5fim) is about 1.8
(rads/yr)//iCi/m3) with a maximum dose of 2(rads/yr)/(//Ci/m3).
1 The lenses of the eyes would be at a depth of ~3-4 mm and hence would receive essentially no irradiation
from the betas emitted in air. This value is expected to be much the same as for the average dose to the skin.
DOSE FROM88Kr IN THE AIRWAYS OF THE LUNGS
A certain activity of 85Kr would be present in the air passages of the lungs. The volume of these air passages
varies considerably as the person breathes, but an average value would be about 4 liters, where the functional
residual capacity is 3.5 liters and the tidal volume is about 0.5 liters for an individual in the resting state
(Morrow, et al., 1966). Assuming that the air in these passages is at the same concentration as the outside air,
the activity present in the lung air is 4 x 0.001 = 0.004 fiCi. At equilibrium the person would have a daily dose
to lungs of
51 x 0.004 x 2.45 x 10-4 = 5.0 x 10-5 rad/day.
This is essentially the dose from the beta-like radiation. The daily dose to lungs from the gamma radiation
would be approximately 3 orders of magnitude less than this. The daily doses to the ovaries and testes from
this source in lung are 2.4 x 10-10/rad/day and 4.1 x 10-11 rad/day, respectively. The red bone marrow dose
would be about 1.6 x 10-9/rad/day. These daily doses are in addition to those given in Table 1.
SUMMARY OF DOSES
Table 3 summarizes the total dose to various body organs in terms of (rads/yr)/( ^iCi/m3).
REFERENCES
Berger, M. J., (1957), Calculation of Energy Dissipation by Gamma Radiation at the Interface Between
IwoMedia, J. Appl. Phys. 28,1502.
Berger, M.J., (1971), Distribution of Absorbed Dose around Point Sources of Electrons and Beta Particles
in Water and Other Media, J. Nuclear Med. Suppl. No. 5; 12,7.
Dillman, L. T., (1969), Radionuclide Decay Schemes and Nuclear Parameters for Use in Radiation-Dose
Estimation, J. Nuclear Med. Suppl. No. 2; 10,5.
Dillman, L. T., (1970, 1970a), Radionuclide Decay Schemes and Nuclear Parameters for Use in
Radiation-Dose Estimation, J. Nuclear Med. Suppl. No. 4; 5,5.
Dillman, L. T., (1970b), Scattered Energy Spectrum for a Monoenergetic Gamma Emitter Uniformly
Distributed in an Infinite Cloud, Health Physics Division Annual Progress Report for Period Ending July 31,
1970,ORNL4584,p.216.
Haywood, F. F., J. A. Auxier, and E. T. Loy, (1964), An Experimental Investigation of the Spatial
Distribution of Dose in an Air-Over-Ground Geometry, USAEC Report CEX-62.14.
Kirk, W. P., (1972), Krypton 85, A Reveiw of the Literature and an Analysis of Radiation Hazards,
(Environmental Protection Agency, USGPO.
Lederer, C. M., et al., Table of Isotopes, 6th edition.
Morrow, P. E., et al., (1966), Deposition and Retention Models for Internal Dosimetry of the Human
Respiratory Tract, Health Phys. 12,173.
-423-
-------
Poston, J. W., and W. S. Snyder, (1973), A Model for Exposure to a Semi-Infinite Cloud of a Photon
Emitter, submitted to Health Physics Journal.
Snyder, W. S., H. L. Fisher, Jr., M. R. Ford, and G. G. Warner, Estimates of Absorbed Fractions for
Monoenergetic Photon Sources Uniformly Distributed in Various Organs of a Heterogeneous Phantom, J.
Nuclear Med. Suppl. No. 3; 10,5.
Southwood, W. F. W., (1955), The Thickness of the Skin, Plastic Reconst. Surgery 15,423.
Tipton, H., Gross andElemental Content of Reference Man, Report of ICRP Task Group on Reference Man
(in preparation), Chap. II.
Whitton, J. T., (1973), New Values for Epidermal Thickness and Their Importance, Health Phys. 24,1.
Yeh, S. Y., and R. F. Peterson, (1965), Solubility of Krypton and Xenon in Blood, Protein Solutions and
Tissue Homogenates, J. Appl. Physiol. 20,1041.
TABLE 1. Dose to Body Organs from an Infinite Cloud of «»Kr
(rads/day/ /iCi/m3 of air).
Source
Organ
Adipose Red Bone
Skin Tissue Lungs Marrow Skeleton Ovaries Testes
Total
Body
Photons in air 4.1 xlO-5 3.2 xlO-5 3.1 xlO-5 3.8 xlO-6 4.1 xlO-5 1.3 xlO-5 3.7 xlO-5 3.3 xlO-5
Bremsstrahlung
in air 8.7 xlO-6 5.7 xlO-6 5.2 xlO-6 9.9 xlO-6 l.lxlO-5 3.1 xlO-6 6.9 xlO-6 6.4 xlO-6
Bremsstrahlung
in skin 1.6 xlO-7 1.9 xlO-8 1.1 xlO-8 1.0 xlO-8 2.9 xlO-8 5.8 xlO-9 2.7 xlO-8 2.5 xlO-8
Photons in
the body 3.8 xlO-9 6.2 xlO-9 6.0 xlO-9 6.7 xlO-9 6.3 xlO-9 7.6 xlO-9 8.5 xlO-9 6.2 xlO-9
Betas in
the body 1.0 xlO-6 3.1 xlO-6 1.0 xlO-8 2.1 xlO-6 1.0 xlO-6 1.0 xlO-6 1.0 xlO-6 2.0 xlO-6
Bremsstrahlung
in the body l.lxlO-9 2.1 xlO-9 1.9 xlO-9 3.0 xlO-9 2.5 xlO-9 2.1 xlO-9 2.5 xlO-9 2.1 xlO-9
TABLE 2. Percentage of Separable Adipose Tissue and
Nonfat Tissues and the Ostwald Coefficient for
Subdivisions of the Human Body.
Separable Skeleton Other
adipose tissue (bone plus marrow) soft tissues
Total Mass 12.5
(kg)
Fat Content 10
(kg)
Ostwald 0.374
Coefficient
UCi/g tissue 0.374
MCi/ml air
Tissue Content
(/xCi)for 4.68 xlO3
1 fiCi/minair
10 47.5
2 1.5 TABLES. Total Dose Rate to
Organs of the Body
0.142 0.082 (rads/yr)/(/iCi/m3).
0.0948 0.082 Organ
Skin
9.48 xlO2 3.90 xlO3 Adipose tissue
Lungs
Red bone marrow
Skeleton
Ovaries
Testes
Lenses of the eyes
Dose Rate
1.8
1.5 xlO-2
3.2 xlO-2
1.8 xlO-2
1.9 xlO-2
6.2xlO-3
1.6 xlO-2
1.8 xlO-2
-424-
-------
*>•
to
INPUT DATA
Radiation
Transition
%/disin- energy
tegration (MeV)
Other nuclear
parameters
Beta-1
Betc-2
Gamma-1
0.4
99.6
0.41
0.16* Allowed
0.67 * First forbidden unique
0.514 M2, aK = 0.007,
K/l = 12
Rsf.: Lederer, C. M. ef al, Table of Isotopes, 6th ed.
* End point energy (MeV).
OUTPUT DATA
Radiation (!)
Beta-1
Beta-2
Gamma- 1
Mean
number/
disinte-
gration
(m)
0.0040
0.9960
0.0041
Mean
energy
(MeV)
&)
0.0437
0.2455
0.5140
Ai
(g-rad\
MCi-h/
0.0004
0.5218
0.0045
KRYPTON-85
BETA-MINUS DECAY
85
36
Kr 10.76 y
O.514
0.0
STABLE^Rb
O /
Figure 1.
-------
10
ARM BONE -
UPPER LARGE
INTESTINE
Organs Not Shown in this View
Adrenals
Stomach
Marrow
Pancreas
Skin
Spleen
Ovaries
Testes
Thymus
Thyroid
_UNGS Uterus
Leg Bones
SMALL INTESTINE
LOWER LARGE INTESTINE
PELVIS 0 5 10
CENTIMETERS
ANTERIOR VIEW OF THE PRINCIPAL ORGANS IN THE HEAD AND TRUNK
OF THE PHANTOM
The Adult Human Phantom.
Figure 2.
-------
LEGS
4.8cm
Legs and Male Genitalia of Phantom,
Figure 3.
-427-
-------
10'
10°
CM
I
c
o
o
c.
a.
1
10"
10-
10-
o
a>
? "
•^ CM
y I
E
o
in
101
101-
0 20 40 60 80 100
PERCENT OF INITIAL PHOTON ENERGY (0.5 MeV)
Figure 4.
-428-
-------
101-
5 —
10
—
T3
I
>
0>
c
o
~o
_C
Q.
I
<
o:
(-
CO
CO
S
LU
o:
tn
10
,-2
10
-3
10"
10'
,-5
BREMSSTRAHLUNG FROM
AIR AND SOFT TISSUE
ARE IN SHADED AREA
.001 .002 .005 .01 .02 .05 .1 .2
PHOTON ENERGY (MeV)
.5 1
Energy Spectrum of Bremsstrahlung from
in Skeleton and in Adipose Tissue.
85
Kr Emitted
Figure 5.
-429-
-------
10
o
•&
.
m
oo
10C
10'
10s
c
o
I 105
10
I
.001 .002 .005 .01 .02 .05 .1 .2
PHOTON ENERGY (MeV)
.5 1
Bremsstrahlung Photons per day as a Function of Energy
Produced in the Skin of Reference Man due to a Cloud
of 85Kr in Air.
Figure 6.
-430-
-------
10
0
0.1
cm IN TISSUE
Depth Dose in Tissue from Beta Radiation (In-
85
finite Cloud ofKr,
Figure 7.
-431 -
-------
DOSIMETRY FOR RADIOACTIVE NOBLE GASES*
J. K. Soldat, P. E. Bramson and H. M. Parker
Battelle
Pacific Northwest Laboratory
Richland, Washington 99352
Abstract
This if a survey of conventional methods and results in the dosimetry of the noble gases, with special
reference to ihc environmental considerations relating to nuclear energy applications. Radon problems will
be excluded, as will all issues specific to medical applications of the gases. (Physiological input data from
medical work is, of course, an important part of the required dosimetry information.)
A mong the dosimetry methods is the well-known Monte Carlo method of W. Snyder, et al,, at ORNL. This will
be briefly put in perspective, as a relevant paper by the ORNL group is included in this volume.
The normally low and variable concentrations of the mixed noble gases in the environment lead to problems
both in "calciilational dosimetry" and in field measurements. The phases selected for comment will mainly
relate to reasonable perspective in accounting for such secondary and tertiary intake pathways as the
drinking of water containing dissolved noble gases.
INTRODUCTION
The dosimetry of the radioactive noble gases is of widespread interest. Naturally occurring radioactive
noble gases are limited to isotopes of radon, but they are important contributors to the internal doses received
by uranium mine workers. Radioisotopes of argon are formed by the neutron activation of stable argon.
Radioisotopes by krypton and xenon are produced as fission products resulting primarily from the fissioning,
of uranium of plutonium fuels in nuclear reactors. They are generally retained within the fuel elements until
the fuel is chemically reprocessed. Small amounts of fission products, however, are released from reaqtor
facilities as a result of fuel element leaks.
The dosimetry of 85Kr was discussed in detail by Dr. W. S. Snyder. It is my intent to extend the dosimetry
discussion to several other artificially created radioactive noble gases.
GENERAL DOSIMETRY CONSIDERATIONS AND CALCULATIONS
As Dr. Snyder has demonstrated, the complete dosimetry of the noble gases requires the calculation of
several factors; such as, (1) external dose to the skin from both beta and gamma radiations; (2) external dose to
the total body, gonads, and internal organs; (3) dose to the lung from inhaled radioisotopes; and (4) dose to
body from noble gases dissolved in the bloodstream, and then absorbed in the various tissues.
Doses from the ingestion of noble gases are generally not addressed in the literature since noble gases have
low solubility in water. Although noble gas solubility is higher in pressurized water systems, the release of the
pressure results in escape of the gas. Hence, the ingestion of unusually high concentrations of noble gas in
water is unlikely. The concentration of the noble gases in the water at the time of ingestion would be difficult to
estimate. Adding to the uncertainty in such an estimation is the fact that xenon may be retained in the water
to some degree when organic material is present.
It has been aptly demonstrated by several workers (including Dr. Snyder) that the dose from noble gases
absorbed in tissue is generally small compared to the dose from direct external radiation, and from inhalation
(Hendrickson, 1970; Kirk, 1972; Snyder, et al., 1973; and Whitton, 1968). These internal doses are, however, the
ones to consider for medical applications of these nuclides. In subsequent sections some of these results are
tabulated and compared with each other, and with external and inhalation doses.
External doses resulting from submersion in a large plume of radioactive gas can be easily calculated if the
concentration of the material in the air is relatively uniform. Doses can be determined on the assumption that
the plume is "infinite" in volume relative to the range of the emitted radiations. Under this assumption, the
energy absorbed per gram of material is equivalent to the energy emitted per gram. All that is required then is
to convert the average energy per disintegration to dose, and to correct for differences in energy absorption
between air and tissue, and for the physical geometry of each specific exposure situation.
Frequently the airborne concentration surrounding a person is very non-uniform; for example when he is
standing close to a tall stack. For this situation the radiation from the overhead plume must be taken into
consideration when evaluating the external dose. Several computer programs have been developed for
Sf-lW ^ reqmr1ed ^^cal integration of the contribution from each finite cloud element (Strenge, etal.,
1973) Such specialized situations were not included in the present work, but the calculations are manageable
once the geometry has been defined.
The dose from submersion in air is an external dose to either the skin only, or the skin and the total body,
depending upon the penetrating power of the radiation emitted from the airborne radioisotope.
-432-
-------
All of the noble gases considered in this paper emit only beta and gamma radiation. For the present work,
we have converted fad to rem using a quality factor of 1; however, a quality factor (QF) of 1.7 was used for beta
particles and electrons with maximum energies equal to or less than 30 keV (ICRP, 1959). It has been reported
by H. J. Dunster that the ICRP is seriously considering the use of a QF of 1.0 for all beta and electron energies
(Dunster, 1969).
In addition, we have chosen to calculate the skin dose at a depth of 7 x 10-3 cm (7 mg/cm2) and the total-body
dose at a depth of 5 cm as originally suggested by the NCRP (NBS, 1964). Recent measurements indicate that 4
or 5 mg/cm2 might be more appropriate than 7 mg/cm2 (Whitton, 1973). The dose to the male gonads was
calculated at a depth of 1 cm in tissue. No separate dose calculation was made, however, for the female gonads.
(The total-body dose can be used as an upper limit to the female gonad dose from external radiation.)
The ratio of surface to depth dose was estimated for each maximum beta energy by methods given by
Loevinger, e£ a/., (1956) and summed by nuclide. Gamma radiation dose at each of the 3 depths in tissue was
determined from the values of absorption coefficients for muscle and the ratio of stopping power for muscle
and air tabulated by the National Bureau of Standards (NBS, 1964). The decay schemes for the nuclides were
taken from (the tables on pages 73 and 285 from) Lederer, et al., (1967).
For a person standing on the ground surrounded by a very large hemisphere of radioactive gas, the
geometry for gamma radiation is obviously 2 IT, For beta radiation with its shorter range in air, the physical
arrangement approaches the infinite volume (47Tgeometry). However, since the beta is of limited penetrating
power, it will irradiate the skin from only one side and again the geometry will be 277. (Minor exceptions can
occur in thin membranes, such as a protruding ear, which could receive beta radiation from both sides,
approaching a 47Tgeometry.)
(1) Equation 1. The resultingequation for calculation of the dose from air submersion is (Soldat, 1971):
(D.F.)airsub = 0.887(Ep+EY)
where (D.F.)ajr sub is the dose factor in units of mrem/h per £lCi/m3, E Q, E y are the effective energies of the
beta and gamma radiations, respectively, calculated at the depth of interest (7 x 10-1 or 5 cm) and corrected for
relative stopping power.
The constant takes into account the density of air (1.2 x 10-3 g/ml @20°C), the conversion from MeV to mrem
and the factor of 1/2 for 2/r geometry.
(2) Equation 2. The dose to the lung from the air within it can be calculated from the following equation:
(D.F.)inhalation = 2.13 VLS/m mrem/h per /iCi/m3
where VL is the volume of air within the lung, 4 liters, (Snyder, et al., 1973)
£ is the effective energy deposited in the lung, MeV per disintegration,
m is the mass of the lung, 100 g. (ICRP, 1959)
The concentration of the noble gas in the air within the lung is taken to be the same as that in the inspired
air. The effective energy,£, is calculated from the formula of the ICRP (1959) using an effective radius of 10 cm
for the lung.
The results obtained from applying Equation 1 to calculation of the external dose rates to skin, testes, and
total body from 14 radioactive noble gases are summarized in Table 1. Also included in Table 1 are the dose
rates to lung tissue from inhalation of these noble gases, as well as the combined dose to the lung from both
inhalation and external gamma irradiation.
Two important daughter radionuclides, 88Rb and 138Cs, have been included in Table 1. These two nuclides
make significant contributions to the dose rates from their parent activities. Rubidium-88 (18-minute half-life)
is nearly always present at equilibrium with its parent 88Kr (2.8-hour half-life). The total dose rate from the
parent-daughter combination is from 30% to 300% higher than from the parent alone, depending upon tissue
depth.
Both 138Xe and 138Cs have relatively short half-lives, 14 minutes and 32 minutes, respectively, and the
daughter contributes a varying amount to the total dose from the 138Xe, depending upon the time since release
of the noble gas to the atmosphere. The dose rate per unit concentration.of 138Cs is 150% to 200% of that from its
parent 138Xe, but its longer half-life may reduce its relative contribution slightly.
Precise lung dosimetry would require that these two daughter radionuclides be treated as attached to
particles. As such there would be a buildup in the lung with an attendant increase in the dose calculated. On
the practical level, this buildup may be disregarded because of the short radioactive half-lives of the
daughters.
DOSE TO THE SKIN AND TESTES
The detailed results for the beta and gamma contributions to the dose to the skin and testes are given in
Table 2 (A and B), where they are compared to similar data presented by Schaeffer (1973) at the recent IRPA
Congress in Washington, D.C. Dr. Schaeffer calculated the beta dose at a point T' in tissue by integrating the
contribution from all of the beta particles which were able to reach that point. All of the beta particles were
assumed to travel in a straight line through air and tissue. There is excellent agreement between Dr.
Schaeffer's results and those obtained with the empirical formula of Loevinger for all nuclides, except perhaps
87Kr, where our values are 30% higher.
Dr. Schaeffer's calculations indicated that no beta rays from any of the noble gases were energetic enough
to penetrate through the 1 cm of tissue assumed to be over the male gonads (Schaeffer, 1973). The Loevinger
formula, however, implies a small beta contribution to the doses to the testes, especially for the 88Rb daughter
of 88Kr, and for 87Kr and 137Xe (Loevinger, et al., 1956).
-433-
-------
TOTAL-BODY DOSIMETRY
Table 2 (C) lists the results obtained by us for the total-body dose rate and compares them with those of
Schaeffer (1973) and Russell, ctal., (1972). The latter authors employed the reciprocity theorem (Loevinger, et
al, 1956) for their calculation wherein:
"the internal gamma (total-body dose from an external cloud equals the dose to the cloud from the
radioactivity in the body if the concentrations areequal This dose is then the difference between the dose to
a point in an infinite medium with uniform source distribution and the dose absorbed in the body from a
uniform source in the body with the same concentration as in the infinite media."
This method of calculation yields an average dose over the entire body — rather than a dose at a single depth
in tissue. Nevertheless, the results obtained by Russell, etal, (1972) are in reasonable agreement with those
obtained by Dr. Schaeffer (1973) and in the present work utilizing the half-infinite cloud calculation.
Dr. Schaeffer's values do not include a correction for attentuation in the first few centimeters of tissue. The
attenuation becomes important for the very low energy radiation such as the bremsstrahlung from 85Kr. For
this nuclide the dose calculated in the present work is 2/3 of that calculated by Dr. Schaeffer. Neither of these
two sets of calculations, however, include a "build-up factor" which could compensate somewhat for the
omission of the attenuation corrections in Dr. Schaeffer's calculations. In addition to the results shown in the
table, Dr. Snyder has reported values of 1.6 x 10-3 for 85Kr (Snyder, et al., 1973) and 0.020 for 133Xe (Hilyer, et al.,
1972).
INTERNAL DOSE TO THE LUNGS
The calculated internal dose to the lungs from inhalation of noble gases is summarized in Table 3 along with
the values calculated by Russell and Galpin. The latter authors assumed a lung volume of 5.6 liters — rather
than the 4 liters employed in the present work. As a result our values are generally, but not always, lower then
those of Russell and Galpin. The two exceptions are 88Kr and 135Xe.
Table 3 also tabulates internal doses to lung and other tissues calculated by Mrs. Whitton (1968) for 85Kr and
133Xe. The value used by Mrs. Whitton for the lung volume was not stated, but her calculated values for lung
dose agree well with ours. As can be seen in the table, the doses from noble gases absorbed in tissue are
significantly lower than those from external gamma radiation.
INTERNAL AND EXTERNAL DOSES FROM KRYPTON-85
Table 4 lists the results obtained by several workers for internal and external doses from 85Kr. The
assumptions used by the various authors were not always the same. This is especially true of the depth at
which the "skin" dose was calculated, and the lung volume. The value for the lung dose attributed to Dr.
Schaeffer was not actually given in his paper, but was estimated by us from the value of the (MPC)a which he
calculated for lung as the critical organ. As such the value of 5.8 x 10-3 probably includes the contribution from
external radiation.
CONCLUSIONS
When the variations in assumptions are taken into account, there is reasonable agreement among the
values obtained by the different authors. It is also obvious that the internal doses are insignificant compared
to the external total-body and skin doses. For noble gas nuclides with relatively little penetrating radiation,
the critical organ is the skin — even considering its less restrictive dose standard. For the other noble gases,
the total body is the critical organ.
If one were the calculate the (MPC)a values for these noble gases based upon the dose rates presented in this
paper, the values obtained would be much less restrictive than the ICRP values currently accepted. For
example, the ICRP (MPC)a for 168-hour occupational exposure to 85Kr is 3 A«Ci/m3. Dr. Schaeffer calculated
corresponding values of 20,170, and 290 for skin, total body, and lung (internal and external), respectively, as
critical organs. As a result,one should always rely on first principles when calculating radiation doses, rather
than to simply multiply fractional MFC by the dose standard.
-434-
-------
REFERENCES
Dunster, H., (1969), Progress Report from ICRP, Health Physics, 17:389.
Hendrickson, M.M., (1970), The Dose from ssKr Released to the Earth's Atmosphere, USAEC Report
BNWLrSA-3233, Pacific Northwest Laboratory, Richland, Washington.
Hilyer, M.M. and W.S. Snyder, (1972), Estimates of Dose from 133Xe to Infants and Children for
Immersion in an Infinite Cloud and for Medical Uses, paper presented at the Mid-Year Health Physics Society
Symposium in Puerto Rico, December 1972 (to be published in the conference proceedings).
International Commission on Radiological Protection, (1959), Report on ICRP Committee II on
Permissible Dose for Internal Radiation, ICRP Publication 2, Pergamon Press, New York.
Kirk, W.P., (1972), Krypton-85, A Review of the Literature and an Analysis of the Radiation Hazards,
Environmental Protection Agency, Office of Research and Monitoring, Washington, B.C. 20460.
Lederer, C.M., J.M. Hollander and I. Pearlman, (1967), Table of Isotopes, 6th Ed., John Wiley and
Sons, Inc., New York.
Loevinger, R., E.M. Japha, and G.L. Brownell, (1956), Discrete Radioisotope Sources, Chapter 16 in
Radiation Dosimetry, Academic Press, Inc., New York, NY.
National Bureau of Standards, (1964), Physical Aspects of Irradiation (Recommendations of the ICRU
Report Wb, 1962), NBS-Handbook 85, U.S. Dept. of Commerce, U.S. Government Printing Office, Washington,
B.C.
National Committee on Radiation Protection, (1954), Permissible Dose From External Sources of
Ionizing Radiation, National Bureau of Standards Handbook 59, Superintendent of Bocuments, Washington,
B.C.
Russell, J.L., and F.L. Galpin, (1972), Comparison of Techniques for Calculating Doses to the Whole
Body and to the Lungs from Radioactive Noble Gases, pp 286-308 in Radiation Protection Standards: Quo
Vadis (W.P. Howell and J.P. Corley compilers), Proceedings of the Sixth Annual Health Physics Society
Topical Symposium, Columbia Chapter, Health Physics Society, Richland, Washington, February 1972.
Schaeffer, R., (1973), Calculs de Dose en Irradiation Extreme Par lex Gaz Ranes, paper presented at the
International Radiation Protection Association 3rd Congress, September 9-14,1973, Washington, B.C.
Snyder, Vt.S.,etal., (1973), Dosimetry for a Man Immersed in an Infinite CloudofS5Kr, paper presented at
the Noble Gases Symposium, Las Vegas, Nevada, September 24-28,1973.
Soldat, J.K., (1971), Modeling of Environmental Pathways and Radiation Doses from Nuclear Facilities,
USAEC Report BNWL-SA-3939, Pacific Northwest Laboratory, Richland, Washington.
Strenge, D.L., and E.G. Watson, (1973), KRONIC A Computer Program for Calculating Annual
Average External Doses from Chronic Atmospheric Releases of Radionuclides, USAEC Report BNWL-B-264,
Pacific Northwest Laboratory, Richland, Washington.
Whitton, J.T., (1968), Dose^Arising from Inhalation ofNoble'Gases, CEGB Report RB/B/N-1274, Central
Electricity Generating Board, Berkeley Nuclear Laboratories, Berkeley, Glos.
Whitton, J.T., (1973), New Values for Epidermal Thicknesses and Their Importance, Health Physics,
24:1-8.
-435-
-------
Table 1. Dose Rates to Selected Tissues from a Semi-Infinite Cloud of Noble Gases
' (mrem/h per /uCi/m3)(a).
Nuclide
1Ar
85Kr
87Kr
88Kr
(88Rb)
133mXe
133Xe
135Xe
137Xe
13»Xe
(138Cs)
Skin(b)
0.12
1.6
.0.00076
0.32
0.16
2.7
2.0
(2.7)
0.048
0.060
0.069
0.50
0.49
1.8
1.7
(3.1)
External
Testes(c)
0.00043
1.1
0.0
0.13
0.0022
1.3
1.5
(0.93)
0.0028
0.027
0.025
0.35
0.21
0.22
1.2
(1.8)
Total
Body(d)
0.00033
1.1
0.0
0.13
0.0022
1.3
1.5
(0.56)
0.0028
0.027
0.025
0.35
0.21
0.12
1.2
(1-8)
(a) No credit taken for attenuation by clothing.
(b) At a tissue depth of 7 x 10-3 cm.
(c) At a tissue depth of 1 cm.
(d) At a tissue depth of 5 cm.
(e) Assuming a lung volume of 4 liters.
Internal
Lungs(e)
0.0016
0.0067
0.00037
0.0024
0.0019
0.013
0.0067
(0.019)
0.0012
0.0015
0.0012
0.0019
0.0033
0.015
0.0066
(0.012)
Total
Lungs
0.0019
1.1
0.00037
0.13
0.0041
1.3
1.5
(0.58)
0.0040
0.028
0.027
0.035
0.21
0.14
1.2
(1-8)
Table 2. Comparison of External Dose Rates to Selected Tissues from a Semi-Infinite Cloud of
Noble Gases (mrem/h per fiCi/m3).
A. Skin
This Paper
Nuclide
39Ar
41Ar
83mKr
85niKr
85Kr
87Kr
88Kr
(88Rb)
mmxe
i33mXe
133Xe
issmxe
135Xe
137Xe
138Xe
(138Cs)
Schaeffer
(1973) Beta
0.37
—
0.17
0.17
0.83
—
(--)
—
—
0.040
0.23
—
—
(--)
Beta(a)
0.12
0.40
0
0.18
0.16(0.21)(c)
1.1
0.32
(2-1)
0.045
0.028
0.040(0.0075)(d)
0.029
0.25
1.6
0.38
(1.0)
Gamma(b)
4.5 xlO-4
1.2
7.6x10-"
0.14
2.8 x 10-3(2.1 x 10-3)(c)
1.6
1.70
(0.64)
3.2xlO-3
0.032
0.029(0.026)(d)
0.41
0.24
0.17
1.36
(2.1)
Total
0.12
1.6
7.6 x 10-4
0.32
0.16(0.21)(c)
2.7
2.0
(2.7)
0.048
0.060
0.069(0.034)(d)
0.50
0.49
1.8
1 7
(3.1)
(a) Including conversion electrons.
(b) Including Bremsstrah lung.
(c) Values reported by Dr. Snyder, etal., (1973).
(d) Values reported by Dr. Snyder (Hilyer, et at., 1972).
-436-
-------
Table 2 (Contd). Comparison of External Dose Rates to Selected Tissues from a Semi-Infinite
Cloud of Noble Gases, (mrem/h per ju,Ci/m3).
B. Testes
«Ar
83mKr
85mKr
85Kr
87Kr
88Kr
(88Rb)
This Paper
Schaeffer
(1973) Beta Beta(a)
0
0 2xlO-4
Gamma(b)
4.3 xlO-4
1.1
Total
4.3 xlO-4
1.1
0
0
0
133Xe
l35Xe
137Xe
138Xe
(138Cs)
0
0
0
0
0
0.047
2xlO-3
(0.37)
0
0
0
0
0
0.10
9xlO-4
(0.012)
0 0
0.13 0.13
2.2 x 10-3(1.8 x 10-3)(c) 2.2 x 10-3(1.8 xlO-3)(c)
1.3 1.3
1.5 1.5
(0.56) (0.93)
2.8xlO-3
0.027
0.025(0.020)(d)
0.35
0.21
0.12
1.2
(1.8)
2.8 x 10-3
0.027
0.025(0.020)(d)
0.35
0.21
0.22
1.2
(1.8)
(a) Including conversion electrons.
(b) Including Bremsstrahlung.
(c) Values reported by Dr. Snyder, etal., (1973).
(d) Values reported by Dr. Snyder (Hilyer, et al, 1972).
C. Total Body
Nuclide
39Ar
41Ar
83mKr
Russell,
Schaeffer (1973) etal.,
85Kr
87Kr
88Kr
(88Rb)
89Kr
i3imXe
133Xe
135Xe
137Xe
138Xe
(138Cs)
1.16
0.15
3.3xlO-3
0.90
( \
0.074
0.24
(Snyder, etal., 1973
Hilyer.etffll.,1972) This Paper
3.3 xlO-4
1.1
0.076
1.2 xlO-3
0.90
0.88
t \
4.2
6.8xlO-3
0.030
0.030
0.24
0.14
0.090
1.7
/ \
1.6x10-3
0.020
0
0.13
2.2 xlO-3
1.3
1.5
(0.56)
2.8 xlO-3
0.027
0.025
0.35
0.21
0.12
1.2
(1.8)
(a) 1. The authors used the reciprocity theorem (Loevinger, et al, 1956) to calculate total body dose.
TThe values presented here do not include the contribution from beta radiation as did those
presented in the original paper.
-437-
-------
Table 3. Comparison of Internal Dose Rates from Inhaled Noble Gases
(units of 10-3mrem/h per /xCi/m:!).
Russell and Galpin (1972)
Whitton(1968)(b)
This Paper
Nuclide
83mKr
85Kr
87Kr
88Kr
(«8Rb)
135Xe
137Xe
138Xe
(138Cs)
Total
Lungs(a) Body(b)
Lungs(c)
Adipose
Tissue
Remaining
Tissue(b)
Testes(c)
3.0
3.0
17
5.2
20
1.8
2.2
1.7
4.0
1.5
18
13
0.14
0.11
0.80
0.52
1.6
0.21
0.27
0.20
0.33
0.54
2.1
2.4
0.24(0.13)(e) 0.030(0.02-0.09)(e) 0.030(0.042)(e)
0.53(0.48)(f) 0.095(0.062-0.075)(f) 0.059(0.063)(f)
Lungs(d)
1.6
6.7
0.37
2.4
1.9
13
6.7
1.2
1.5
1.2
1.9
3.3
15
6.6
(a) Based on a lung volume of 5.6 liters.
(b) Used an average value of Ostwald coefficient for entire body.
(c) Lung volume not stated.
(d) Based on a lung volume of 4 liters.
(e) Values of Dr. Snyder, etal, (1973).
(f) Values of Dr. Snyder (Hilyer, etal., 1972).
Table 4. Comparison of Radiation Doses to Various Tissues from a Semi-Infinite Cloud of 85Kr
(mrem/h per fiCi/m3).
Pathway
Snyder,
etal. ,(1973)
Russell, et al.,
(1972) Whitton(1968)
Schaeffer(1973) Kirk (1972)
Hendrickson
(1970)
This Paper
External — Skin
Beta
Gamma Bremsstrahlung
Total
External — Total Body
Gonads
Inhalation — Lungs
Total Body
Adipose Tissue
Remaining Tissue
Gonads
0.21(a)
2.1xlO-3(b)
1.6xlO-3
1.8xlO-3
2.1 x 10-3(g)
8.4x10-'
1.3x10-'
(2-9) xlO-5
4.2 xlO-5
1.6xlO-3(c)
1.2xlO-3(f)
1.2xlO-3(f)
3.0xlO-3(h)
0.19
0.19
2.3xlO-3
2.3 xlO-3
1.6xlO-3(i)
2.4 x KM
3xlO-5
3xlO-5
0.17(d)
0.17
3.3xlO-3
3.3x10-3
5.8xlO-3(k)
0.24(e)
1.9x 10-3
0.24
1.5 xlO-3
1.5 xlO-3
8.7x10-5
O.ll(d)
2.7 xlO-3
0.11
2.7 xlO-3
2.7 xlO-3
1.9xlO-3(j)
2.8 xlO-3
0.16
2.2 xlO-3
2.2 xlO-3
1.9xlO-»(g)
(a) Average dose to skin layer 5 x 10-3 cm thick (maximum is 10% higher).
(b) Average dose to a layer 0.2 cm thick.
(c) Assumed skin depth of 0.1 cm.
(d) Assumed skin depth of 0.007 cm
(e) Assumed skin depth of 0.0 cm.
(f) Based upon reciprocity theorem.
(g) Assumed lung volume of 4 liters.
(h) Assumed lung volume of 5.6 liters.
(i) Lung volume not specified.
(j) Assumed lung volume of 3.5 liters.
(k) Estimated from Dr. Schaeffer's calculated values of (MFC) for lung and probably includes the
contribution of external radiation. a
-438-
-------
BEHAVIOR OF KRYPTON-85 IN ANIMALS
W. P. Kirk
Experimental Biology Laboratory
National Environmental Research Center
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Abstract
The in vivo behavior of 8sKr can usually be predicted if the protein/water/fat composition of a body or tissue
and its blood perfusion characteristics are known. The amount of isotope taken up by the. tissue is the weighted
sum of component tissue: air partition coefficients multiplied by the concentration ofssKr in alveolar air, while
the rates of saturation or desaturation are determined by perfusion.
Equations describing postulated kinetic behavior of noble gases are presented and compared with 8SKr data
reported for individual organs and tissues from several species in the literature, and data obtained with whole
guinea pigs and rats.
Partition coefficients are discussed and theoretical values compared with in vitro data from this laboratory
and the literature, and with in vivo data for blood and 22 other organs/tissues obtained with guinea pigs in
current work.
Equilibrium beta radiation doses to various organs/tissues from a contained isotope in guinea pigs
breathing 85Kr at present MPCa are given.
INTRODUCTION
For several years the author has been investigating the physiological behavior and radiation effects of 85Kr
to provide the EPA with in vivo bioeffects data upon which various standards for 85Kr can be based. Included
in these investigations have been determinations of the saturation and desaturation kinetics of 85Kr in
unanesthetized guinea pigs and rats, and determinations of in vivo partition coefficients into guinea pig blood
and tissues.
The literature search associated with these studies yielded a substantial amount of information on the
behavior of Kr in an assortment of organs/tissues in several species. A dichotomy of purpose exists, however.
The investigators were usually interested in evaluating blood flow, with the details of kinetic behavior being
somewhat incidental, while the reverse is true in the present studies. Reports giving details of kinetic behavior
or partition coefficient determinations are mostly those concerned with development of methods using 85Kr to
determine organ blood flow with modifications of the Pick principle. Some "use" papers also report kinetic
data for control animals. Most investigations have been concerned with one organ, or part of an organ, and
have used animals in an abnormal physiological state due to anesthesia, surgical trauma, restraint, or use of
circulatory control drugs.
This review discusses the behavior of 85Kr in animals primarily from the empirical viewpoint — what has
been seen in vivo — rather than trying to present or support any elaborate mathematical model. The author's
interests and the scope of this paper have been largely limited to understanding what happens internally
when an animal is placed in an 85Kr atmosphere.
THEORY
The saturation or desaturation of 85Kr, and other presumably metabolically inert gases in the body have
almost always been describable as an exponential or sum of exponential functions. Kety (1951) has discussed
the various possibilities in detail. The simplest of the models presented is derived from the Fick principle,
which Kety states as "The amount of inert gas taken up by the tissue per unit of time is equal to the quantity
brought to the tissue by the arterial blood minus the quantity carried away in the venous blood." The
derivation, as given by Kety (1951), for a single homogenously perfused tissue is:1
(1) dQ.
-af = Fi
-------
Assuming that venous blood from a tissue is in equilibrium with the tissue itself with respect to the inert gas:
(3) Cj = LjCv where Lj= tissue: blood partition coefficient
and in saturation:
(5) Cj = LiCa (L - e-kit) where: kj = Fj
(6) Q = Ca £ VfLj (1 - e'V) where Q = total quantity in system.
or in saturation:
(7) Ci = d? e"1^ where d? = concentration at t = 0.
However, for an animal breathing the gas, Ca=CL where C = concentration in air and L = the blood: air
partition coefficient if: (1) the gas being considered has a very low solubility in blood; (2) alveolar and arterial
concentrations remain constant; and (3) gas is introduced abruptly at its final concentration into alveolar
spaces (or the exposure continues long enough for complete equilibrium to occur).
Kety also presents several more complicated models which account for the ambient air-alveolar air
equilibrium functions, various flow perturbations in the body, etc.
The partition coefficient is central to any analysis based on the foregoing equations. Usage has not been
standard in regard to the units used for partition coefficients. Some investigators have expressed their results
in terms of the weight of the receiving medium and others have used the volume. A variety of symbology is also
encountered. In this review, albeit arbitrarily, the following symbols/units are used:
(1) L n.-t = concentration in (receiving) medium 2 (U Ci/cm3)
concentration in (source) medium 1 ( /i Ci/cm3)
as blood: air or tissue: blood
when the source medium is air, L = the Ostwald solubility coefficient for the receiving medium.
(Note: Lg:1=L3.2xL2:1 Example: Ltissue:air = Ltissue:blood x Lbiood:air)
(2)L 2.i = concentration in (receiving) medium 2 ( uCi/g)
concentration in (source) medium 1 ( fiCi/cm3
Lw2:l = L2:l/Pland
Lw3:l = Lw3:2 xF2 Lw2:l =
Except for the direct determinations of Lw tissue:air reported elsewhere (Kirk, 1973) use of Lw has been as Lw
tissue:blood. In most cases, Lblood:air has not been needed for the determinations.
It has been demonstrated in vitro that the Ostwald coefficient (L) of blood, saline solutions of beef brain, or
rabbit muscle homogenates are the weighted sum of the fat, protein, and water (saline) solubilities for the
solution ( Yeh, et al., 1965); (i.e., Lsoiution:air= £ fjLi where fp fraction of solution made up of component i and
Li is the Ostwald solubility coefficient for fraction i). With blood, hemoglobin, which has a much higher
coefficient than other protein, must be treated separately; similarily in brains, non-neutral fat is separately
accounted for because its coefficient is lower than that of neutral fat.
The variation in solubility of noble gases in solvents with temperature has been investigated (Hardewig, et
al., 1960 ; Markham, et al, 1941; Morrison, et al, 1954; and Yeh, et al, 1963, 1964, 1965) and shown to follow the
Clapeyron relationship in organic solvents, and the Valentiner equation in aqueous solutions. TheClapeyron
relationship can be expressed as: Log L = A + B/T where
L = Ostwald solubility coefficient
A, B = constants for system
T = temperature (°K)
-440-
-------
The Valentiner equation is usually written:
Log L = a /T + blog T + c, where a, b, c, are constants for the system.
Most of the blood flow studies considered have used the Pick derived model presented earlier, and none of the
85Kr data presented herein is sufficiently detailed to justify considering any of the more complicated analyses.
The key to blood flow determination with the Pick relationship is evaluating the exponential coefficient k
used in equations (5) through (8). As stated, ki = Fi/VjLj. (or ki = Fi/WiLwi). By defining a specific flow F =
ml/g/min and using k = .6937 T 1/2 where T 1/2 - half-time of saturation or desaturation, the flow can be
determined by:
F = Lwki = (Lw x .693)7 T l/2j = (Li x .693)/(Pi x T 1/20
If the tissue has a constant, uniform flow and uniform partition coefficient, T 1/2 can be easily evaluated
graphically from a semi-logarithmic plot of (C v - Ca) vs time, or a comparable plot of radioactivity in the tissue,
as measured externally by an appropriate detector, vs time2.
Most organs, however, possess neither uniform perfusion nor uniform partition coefficients, and the curves
obtained from them are multi-exponential in form. Accurate determination of blood flow by the Pick method in
this case requires the determination of an average value3 for k. Two approaches that have been used
extensively are:
(1) Determine the T 1/2 and corresponding k for a monoexponential curve tangential to the combined
semilogarithmic plot at t = O. If the partition coefficients involved are closely similar, and the initial
concentrations in the compartments are equal, the initial slope will represent the total flow in the organ.
Many protocols have been used with varying success in fulfilling the latter condition.
(2) Follow saturation or desaturation (preferred) long enough to accurately resolve the multi-exponential
curve into its components, and determine the respective half-times and relative magnitude (extrapolate to
equilibrium), and determine a weighted exponential constant or half-time to use in the equations. Best
accuracy also requires some notion of the partition coefficients for the compartments.
The author feels more comfortable with results obtained by the second method; but, when multiple
determinations are required, as in a clinical situation, the initial slope method is much faster and has been
more extensively used.
The method of 85Kr administration has been ignored so far because it has not been material to the
mathematical discussions. It will obviously affect the degree to which the experiment conforms to the
assumptions stated in the method of analysis. 85Kr has been administered by:
(1) Inhalation from a closed circuit system, or constant concentration source; and analysis of: (a) the
disappearance of 85Kr from the exposure system, (b) (Cc - Ca) determined in either saturation or desaturation
by assay of simultaneous arterial and venous blood samples; and (c) the quantity of isotope in the organ — as
determined by external counting of the radiation emitted, vs time, in saturation or desaturation.
(2) Continuous infusion of 85Kr in solution (saline, dextran, etc.) into the arterial supply to the organ, and
analysis of saturation or desaturation curves obtained by plotting (Cv Ca), or emitted radiation vs time in
phase.
(3) Fast injection of a bolus of 85Kr, in a small volume of solution, into the arterial supply to the organ, and
evaluation of the desaturation curves obtained by plotting either (Cv - Ca), or externally detected activity vs
time.
(4) Direct injection of 85Kr into the organ or tissue (usually in saline), and analysis of desaturation curves
obtained by plotting externally detected activity vs time.
The clearest kinetic data in most organs, regardless of the mode of administration, has been obtained by
external counting of either the beta (0.672 MeV) or gamma (0.514 MeV) radiation emitted from theorgan. The
morphological characteristics of the organ, and the characteristics of theradiationvis-a-visthe detector, must
be clearly understood, however, to correctly interpret the data.
LITERATURE
The pertinent literature can be broadly divided as follows:
(1) Determinations of blood flow in various organs and tissues, for a variety of clinical reasons; and,
incidental to that purpose, study of the fine structure of saturation/desaturation curves and tissue:blood
partition coefficients.
(2)In vivo whole body fat determinations.
(3) Studies of whole-body kinetic behavior and partition coefficients in small animals.
(4) Determination of Ostwald solubility coefficients or tissue:air partition coefficients for blood, various
biological solutions, and whole animal organs.
2With any radioactive gas selected for experimental use, one criterion is that its radioactive half-life be long
with respect to the duration of the experiment, so that radioactive decay will not significantly affect the slopes
of the curves obtained.
"Another method is to determine the total quantity of isotope entering or leaving the tissue by integrating the
area under the curve.
-441 -
-------
In the first category only those papers presenting information on Kr behavior in normal or control subjects,
albeit anesthetized or surgically traumatized, are included. Data given in terms of blood flow was back
calculated to half-times or exponential coefficients, whenever the original method of calculation was clear.
Information from sick animals, or those treated with vasoactive drugs to alter their physiology has been
omitted — although control values from those papers were included if information was needed for that organ
and species.
Only information pertaining to partition coefficients and compartmental analysis is covered. No claims are
made regarding completeness of literature coverage or crediting of firsts.
1. Blood Flow Determinations.
a. Brain.
Lassen, et al., (1955) pioneered use of 85Kr in the Pick principle circulatory studies with a modification of the
Kety, et al., (1948) nitrous oxide method of determining cerebral blood flow. They determined blood flow for the
whole brain using (Cv - Ca) from simultaneous samples from the femoral artery and the jugular vein, taken
while the subject (cats, dogs, rabbits) breathed 85Kr from a closed system for 14 minutes. Blood radioactivity
was assayed by counting the blood in mica-windowed cuvettes with an array of GM detectors.4 No useful
kinetic data were reported, however. Lw brain:blqod for the whole brain was determined to be 1.06 (S.E. =
.0075) at a hematocrit of 50. Lw was found to vary with hematocrit, and a relationship given to relate the two.
In a similar study, Albert, et al., (1960) used a two-minute saturation time, and plotted (Cv - Ca) in
desaturation for the first four minutes after the end of exposure, to determine a cerebral blood flow of 0.37 -
0.642 ml/g/min (t 1/2 = 1.08 -1.86 min) in human patients who had no overt evidence of cerebral dysfunction.
The 85Kr in blood was eluted and assayed in a gas flow counter.
Lassen, et al., (1961) infused 85Kr in saline for six minutes into the common carotid artery of an anesthetized
cat, and followed desaturation by counting the beta radiation emitted from the cortex, exposed by craniotomy,
with a collimated end-window GM detector. Desaturation was reported to be essentially complete within 15-20
minutes. The curves were clearly multi-exponential. The CBF was calculated, using the initial slope method,
and a previously determined, but unpublished, LW cortex:blood of 0.9, to be 0.4 -1.6 ml/g/min (T 1/2 = 0.43 -
1.73 min). In a subsequent investigation, using the same technique, Ingvar, et al., (1961) found CBF of 0.65
ml/g/min (T 1/2 = 1.07 min) in dogs and 0.51 - 0.53 ml/g/min in a human brain tumor patient (T 1/2 = 1.31 -
1.36 min).
Cerebral blood flow determinations on anesthetized dogs were reported by Glass, et al., (1961) and Harper, et
al., (1961) who used the Lassen and Ingvar methods and partition coefficient. They held anesthesia, blood
pressure, temperature, and pCO2 constant during the experiments, and found an average CBF of about 1
ml/g/min (T 1/2 ~ 0.7 min), using the initial slope of clearly multi-exponential desaturation curves.
In an amplification of their method, Ingvar, et al., (1962) improved the accuracy of CBF determinations by
correcting for unequal compartment saturation. Their semi-logarithmic curves were graphically resolved into
two components, and the components were extrapolated to equilibrium. An average exponent, weighted by
compartment size, was then used to calculate CBF. The fastest component was assumed to be cortical gray
matter, and the slower to be white matter. An autoradiographic technique was used to determine the ratio of
85Kr in gray matter to that in white matter with slices of brain from a cat that had been fully saturated in vivo
with 85Kr. The ratio (0.73), the previously determined Lw brain:blood of 1.06, and an estimated 60:40
gray:white matter weight ratio was used to estimate Lw cortex:blood and Lw white matter-blood of 0.92 and
1.26, respectively, at a hematocrit of 50. The variation of Lw C0rtex:blood with hematocrit was reported to be:
Lwcortex:blood~0-92 .843 where H = hematocrit
(H/100)+.685[1-(H/100)]
The improved method was used on dogs, cats, and rabbits to determine CBF, but no compartmental
breakdowns were given.
Glass, et al., (1962), using a double isotope technique, found an in vivo partition coefficient for the cortex in
anesthetized dogs of 0.91 ± 6% at a hematocrit of 50. The coefficient reported was neither L or Lw since they
used the ratio of beta counts from the brain, and a beta-infinite layer of blood in their calculations. Since the
beta range is mass dependent, the units of their coefficient would be ( uCi/g cortex)/( uCi/g blood) and
assuming a blood specific gravity of 1.05, the coefficient would be about 5% higher than L
Lassen, et al., (1963) determined CBF in unanesthetized humans by counting the gamma radiation emitted
from the cranium with a collimated Nal detector after injection of ^Kr into the carotid artery Two-
compartment exponential curves were assumed and CBF calculated using a weighted exponent. The CBF was
breakdown ml/g/mm (S'D'= -13XTl/2 = °-8' 2-0 min), but no specific information was given on curve
'For benefit of those who do not routinely deal with radioactivity, the abbreviation GM stands for Geiger-
Mueller and Nal for sodium iodide. GM detectors primarily register beta radiation; Nal detectors are used for
gamma radiation.
-442-
-------
Alexander, et al., (1964), using the Lassen, et al., (1955) technique, found half-times for the whole brain of
anesthetized human volunteers to be in the range of three to six minutes. In a subsequent paper (Wollman, et
al., 1965) with two-compartment analysis of similar CBF data, from both anesthetized and unanesthetized
subjects, was reported as shown in Table 1.
McHenry (1964) used a modification of the Lassen, et al., (1955) technique, and the counting techniques of
Albert, et al., (1960), to determine CBF in 25 normal males by analysis of desaturation curves. The CBF
reported was 0.565 ml/g/min (S.D. = .077) (T1/2 = 0.96 - 2.3 min).
Lassen (1965), using his GM detector over exposed brain technique, reported average half-times of 0.27 and
1.4 minutes for the two components resolved from four cat brain desaturation curves.
The most elaborate studies of the kinetic behavior of 85Kr in brain appear to be those reported by Haggendal,
et al., (1965b), and, less extensively, by Nilsson (1965). They used anesthetized, curarized dogs, and made
concurrent recordings with a Nal detector positioned over the skull, and an end-window GM detector over a
craniotomy. 85Kr was administered by fast injection into either the vertebral or external carotid artery, or by
direct injection (1-5 pt 1) into the brain, using a fine needle inserted through a craniotomy adjacent to the one
with the counter. In the latter case, dye was injected with the 85Kr so that the point of injection could be located
morphologically in histological sections of the brain. Desaturation was followed until the longest component,
on a semi-logarithmic plot, was flat enough to ensure good resolution in curve stripping. (The major problem
was reported to be counting statistics.) Their results indicated that the total brain desaturation curve consists
of four components as follows:
(1) A very fast component with a half-time of about 0.10-0.20 min, which was identified by auxiliary
experiments as arterial blood. This component was only resolvable with the fast intra-arterial injections and
was usually associated with low total blood flow.
(2) Another fast component, attributed to gray matter, with half-times of 0.35-1.9 min (wide variations of
CBFandpCO2).
(3) A slower component, associated with white matter, with half-times of 1.5-11.6 min (also under varying
conditions).
(4) A component with half-times of 14-20 min, which was identified as originating in extra cranial tissue
(scalp muscle, neck muscle, connective tissue, etc.)
The Nal and GM curves agreed well and indicate that, in dogs, the GM detector records beta radiation from
white matter, as well as from the gray matter, unless the cortex is thicker than usual (in one large dog, whose
cortex was 3 mm thick, the GM curve was mono-exponential). Clearance curves from the injections were mono-
exponential if the injection was entirely in either gray or white matter. The curves from cortex injections
matched component 2 from the Nal or GM curves obtained with the same animal with intra-arterial injections
while injection into white matter resulted in curves that matched component 3. If injection was on the border
of the gray and white matter, both components 2 and 3 were seen.
In an investigation of blood flow autoregulation, Haggendal, et al., (1965a) injected 85Kr into the vertebral
artery of anesthetized curarized dogs, and recorded desaturation curves from the brain by external gamma
scintillation. The semi-logarithmic curves were resolved into two components, and the faster assumed to be
gray matter. The half-times found with the fast component, with varying blood pressures, ranged from 0.74 -
2.3 min. Using the same techniques, Haggendal (1965) investigated the effects of vaso-active drugs on
cerebral circulation in anesthetized dogs. The fast component obtained from normotensive controls had half-
times of 0.47 -1.31 min.
b. Heart.
The Lassen, et al., (1955) technique was applied to the heart by Hansen, et al., (1962) who had anesthetized
dogs breathe an 85Kr-air mixture for 14 minutes. After the 14-minute saturation, they took arterial and
coronary sinus blood samples to establish initial (saturated) levels, and shifted the animal to breathing air.
Five pairs of A-V samples were taken during the first 10-18 minutes of desaturation. Blood radioactivity was
assayed by GM counting in thin-windowed cuvettes, and (Cv - Ca) plotted semi-logarithmically to determine
the half-time. A partition coefficient of 1 was assumed and coronary blood flow calculated to be 0.48 -1.33
ml/g/min (T 1/2 = 0.52 - 1.44 min). No original data were given. The curves were assumed to be mono-
exponential.
Herd, et al., (1962) injected 85Kr in saline into the left anterior descending coronary artery through
chronically implanted catheters in three unanesthetized dogs, and recorded the washout (desaturation) curve
with a Nal detector over the precordium. The slope of the first two minutes of the semi-logarithmic plot was
used to calculate coronary blood flow. The curves were obviously not mono-exponential, but the authors
claimed that the deviation was due to isotope recirculation. This does not appear probable since a reduction in
blood concentration by a factor of about 500 would be expected before the isotope returned to the heart
(dilution x 25 in the right heart and 95% removed on the way through the lungs). The half-time on the long
component from the one curve shown was about 7 minutes.
Ross, et al., (1964) determined coronary blood flow in anesthetized dogs using 85Kr and 133Xe clearance, and
calibrated the method against direct measurements. They cannulated either the left, or all three coronary
arteries, and supplied blood to the coronary arteries from the carotid artery through a rotometer. 85Kr, in
saline, was injected into this arterial supply and washout followed with a Nal detector positioned over the
heart. They used a direct-fit initial slope from the semi-logarithmic desaturation curve to calculate blood flow.
The initial half-times were 0.4 -1.07 minutes.
-443-
-------
Cohen, et al, (1964) determined (Cv - Ca) by injecting "Kr in saline into the left ventricles of 11
unanesthetized humans and nine anesthetized dogs, taking five pairs of consecutive 30-second (6 ml) samples
of arterial and coronary sinus blood starting 45 seconds post-injection, and then assaying the blood by
gamma-counting in a Nal well detector, and analyzing semi-logarithmic desaturation curves. Mono-
exponential curves were claimed. Those shown, however, seem to have some deviation at latter times. Blood
flow was varied with drugs and the data were not separated by species. The total range of coronary blood flow
was 0.30-1.99 ml/g/min (T1/2 x 32-2.3 min) with a mean of 1.12 ml/g/mm (T1/2 = 0.57 mm). The curve from
one human control had a half-time of 0.65 minutes. .... • i
Johansson et al (1964) reported two-component washout curves in dogs with artificial coronary
occulsions The main component had a half-time of 0.44 -1.2 minutes, and the second component had a half-
time of about 30 minutes, as determined by analysis of semi-logarithmic desaturation curves taken after
injection of 85Kr into the coronary artery (external Nal detector). The long component was attributed to
collateral circulation in the artificially ischemic area.
Subsequently, Linder (1966) reported on a detailed investigation of coronary blood flow measurements in
anesthetized dogs (artificial ventilation). 85Kr was injected into the coronary artery; desaturation followed
and was monitored with a collimated Nal detector. On semi-logarithmic plots, a very fast component (T 1/2 of
5-15 seconds) was seen followed by the main phase, with a half-time of 0.38-1.08 minutes; and, when 5-20% of
the peak activity remained, by a slower component with a half-time of 5 or more minutes. This slow component
was attributed to a combination of: (1) recirculation of the isotope due to bad ventilation in parts of the lung
(slight effects); (2) connective and fatty tissues in the ventricular walls and along the coronary vessels (main
cause); and (3) isotope in the coronary veins (shown to be negligible).
c. Kidney.
A very complete study of 85Kr desaturation in the kidney was reported by Thorburn, et al., (1963) who
injected 85Kr in saline into chronically cannulated renal arteries of unanesthetized dogs, and followed
washout with an external Nal detector. The data were plotted semi-logarithmically and resolved graphically.
The components were associated with various parts of the kidney by sequential radiography. Their results are
given in Table 3.
Bell, et al., (1965) determined an in vivo partition coefficient for dog kidney cortex of 0.96 + .054, using the
ratio of GM (beta) counts from the cortex of 86Kr saturated kidney, and the corresponding count from a cuvette
containing arterial (renal) blood. The same comments made on the Glass, et al., (1962) determination of brain
partition coefficient apply. These authors also reported mono-exponential washout curves from kidney cortex,
determined by GM counting at the exposed cortical surface, after a one-minute infusion of 85Kr in saline into
the renal artery. Half-times of 0.2-0.24 minutes were found in 30 anesthetized dogs.
In a similar study, Cosgrove (1965) reported average half-times of 0.23 minutes for mono-exponential half-
times of 0.18 - 0.39 minutes in normotensive control dogs (anesthetized, curarized, ventilated).
Carriere (1970) monitored 85Kr washout with a Nal detector over the kidneys of anesthetized dogs after fast
injection into the renal artery. He found four components, as reported by Thorburn, et al., (1963), but only
reported values for the fastest two. Their half-times and percent of activity were 0.11 min/83% and 0.51
min/13%, respectively. A subsequent report from the same laboratory (Lockhart, et al., 1972) reported half-
times of 0.1-0.16 minutes and 0.825-0.9 minutes for the fastest two components.
d. Liver.
Blood flow in livers of anesthetized and unanesthetized dogs was investigated by Hollenberg, et al., (1966).
They did two series of animals. In an acute series, catheters were placed in the portal vein and hepatic artery of
anesthetized dogs, and clearance curves were recorded externally with a Nal detector after injection into
either vessel. In the chronic animals, only the portal vein was used (previously cannulated), and the animals
were awake. LhVerblood was determined to be 1.056± 0.56 and the specific gravity of liver found to be 1.02.
With either type of injection, they found either one or two component exponential curves in the same animal at
about the same total flow. Washout by HA injection was found to be about 75% of the PV washout. The half-
times from the chronic unanesthetized dogs were reported to be from 0.34-0.46 minutes. Clearance in the
anesthetized dogs was considerably slower. One complete desaturation curve was given which had two
components with half-times of 0.37 and 17.3 minutes.
The most complete study of 85Kr clearance from liver (dog) was reported by Birtch, et al, (1967), who used
fast injections into either the hepatic artery, portal vein, or both simultaneously, followed by monitoring of the
emitted gamma radiation with a Nal detector coupled to a digital printer. The data were plotted semi-
loganthmically and resolved graphically. The partition coefficient and specific gravity reported by
Hollenberg, et al., (1966) were used. Most dogs were anesthetized but four were awake and had chronically
implanted catheters in both vessels. Four compartment desaturation curves were found — the longest half-
STi A ^ TaS 19"2.6 minutes- Their data (converted from blood flow to half-times) are summarized in
lable 4. Ihese data and sequential radiographs show that component 1 is from a vascular bed perfused only
by the Portal vein, component 3 is from a bed perfused only by the hepatic artery, component 2 is from a bed
perfused by both supplies, and component 4 is from extra-hepatic tissues and intra-hepatic fat
-444-
-------
e. Stomach.
Jansson, et al, (1966) administered 85Kr by intra-arterial injection to anesthetized cats, and followed
desaturation with a Nal detector placed over the stomach. Three-component semi-logarithmic desaturation
curves were found. The half-times of the two faster compartments, believed to be gastric mucosa and
muscularis, were 0.5-1.2 minutes and 3.5-5.8 minutes, respectively. The half-time of the third component,
identified as being tissues outside the stomach wall, was not given.
Bell, et al., (1967a) reported mono-exponential desaturation curves, with half-times of 0.68 minutes (S.D. =
0.09 min), from gastric mucosa of anesthetized control dogs in an investigation of the effects of pCC>2 on
gastric mucosal flow. The stomach was opened and an end-window GM detector placed adjacent to the
mucosa. 85Kr was given by intra-arterial injection. L gastric mucosarblood was determined to be 0.84 (S.E. =
0.04). The same control data were reported by Bell, et al., (1967) in an investigation of histamine effects on
gastric mucosal flow. In another series from the same laboratory, half-times of 0.44-0.48 minutes were
reported for the gastric mucosa of control dogs (Bell, et al., 1969).
f. Intestine.
The behavior of 85Kr in the small intestine of anesthetized, atropinized cats, with the splanchnic nerve
severed, was investigated by injecting it into the superior mesenteric artery, and recording the washout curves
with an external Nal detector coupled to a ratemeter and recorder (Lundgren, et al., 1966; and Kampp, et al.,
1968 a,b). The semi-logarithmic curves were resolved graphically into four components. The total flow was
regulated and measured by a regulating flow-meter in the mesenteric vein. In most experiments, the GI tract,
except for the 20-50 gram of the test segment, was removed to avoid interference. The temperature of the test
segment was controlled. Readings from a lead-shielded end-window GM detector placed close to the serosal
surface, and another GM detector in the gut lumen", were used to help identify the morphological origin of the
different curve components. Local injection of 85Kr in dye, and autoradiography with antipyrine 14C, were also
used for localization. The data obtained are summarized in Table 5.
g. Eyes.
The retinas of anesthetized cats were reported to have four-component 85Kr clearance curves, as determined
by counting with a GM detector in contact with the posterior aspect of the sclera, after intra-arterial injection
(carotid)(Friedman, et al., 1964; and Friedman, et al., 1965). Their data for cats are summarized in Table 6.
Dogs and monkeys were reported to have similar curves. Rabbits had three-component curves with one
component having a half-time of 2.5-5 seconds and the other two having half-times of 1.5-30 minutes when
measured with the GM detector. When a more sensitive solid-state probe attached to the eye was used, and the
85Kr was injected into a branch of the ciliary artery, all species had four-component desaturation curves.
h. Testis.
Setchell, et al., (1966) investigated blood flow in ram testes by analysis of washout curves obtained after
injection of 85Kr, via a chronic implant, into the testicular artery. The animals were usually awake. The
scrotum was held in a temperature-controlled harness and a 5 cm Nal detector, coupled to a ratemeter and
recorder, was used to record the desaturation curve. Several experiments were done with direct injection of
85Kr into the testis. Lw testis:blood was determined to be 0.85 (S.E. = .03) by infusing 85Kr in saline into the
heart for 75 minutes, determining equilibrium by counting blood taken from the abdominal aorta every 15
minutes, killing the animal, removing and weighing the testes, sealing the testes in parafin in a soldered-top
can, assaying the contained radioactivity by gamma counting, and comparing the testis specific activity with
the blood activity. The desaturation curves were mono-exponential with half-times of 4.46-16.4 minutes for
conscious rams (66 determinations), and 7.65-10.8 minutes for anesthetized rams (16 determinations).
i. Skin.
Mono-exponential clearance from dog skin following femoral artery infusion (4 min) was reported by Bell, et
al., (1964). The curves were recorded over the leg with a GM detector. Half-times of 10.97-27.4 minutes were
found.
85Kr clearance from rabbit skin after intra-arterial bolus injection was found to be bi-exponential with half-
times of 3-4 minutes and 17.3-23.1 minutes, respectively, as measured by a GM detector. (Casey, et al., 1965;
andThorburn, et al., 1966).
The results of extensive investigation of skin blood flow in humans, using 133Xe and 85Kr, were reviewed by
Sejrsen (1971) who summarized several of his earlier publications. He concluded that the parallel competitive
exponential model used in other blood flow studies does not hold with human skin because: (1) 85Kr or 133Xe is
removed from the cutaneous blood supply preferentially in subcutaneous tissue, which has a high fat content,
and then is released slowly; (2) some of the isotope is cleared by sweating; and (3) the isotope migrates between
cutaneous and subcutaneous tissue by diffusion. The curves recorded from human skin were, therefore, not
simple functions of blood flow. The only consistent exponential component reported had an average half-time
of 168 minutes (S.D. = 41 min), and was attributed to subcutaneous tissue.
-445-
-------
j. Limbs.
Tobias et al (1949), in an investigation of the causes of the bends in aviators, had unanesthetized
volunteers breathe 79-81Kr from a closed-circuit system for three hours while counting gamma radiation,
emitted from one hand and one knee, using very heavily collimated GM detectors. After three hours the
subjects breathed room air while counting continued for 12 more hours. Three-component exponential curves
were found for both hands and knees. The fastest component was negligible for the knee, however. Exercise or
heating of the hand caused faster clearance; vasoconstrictive drugs caused slower exchange. The kinetic
information reported is given in Table 7. It was also found that radiokrypton was readily absorbed from the
intestinal tract, especially the doudenum.
Holzman et al., (1964) injected 85Kr in saline into forearm muscle in human volunteers and followed
washout with an external Nal detector. The curves obtained were multi-exponential with a great deal of
variation. Blood flow was determined using the initial slope. The first half-time was reported to be 17.8
minutes (S.D. = 8.2 min).
2. Whole-Body Fat Determinations.
The uptake of Kr from a closed system (with CO2 scrubbing, and C>2 replacement) has been used by several
investigators to estimate body fat content. Lesser, et al., (1963) used inert Kr and reported that equilibrium
was not complete in 6-7 hours with human subjects. (Their system had a water spirometer which may have
permitted leakage of Kr.) In their discussion, they propose a model for uptake of metabolically inert gases,
consisting of blood in equilibrium with the gas in lungs, and three parallel competitive compartments in
contact with blood. The compartments proposed, in order of their speed of saturation/desaturation, were: (1)
rapidly perfused lean tissue (as heart, brain, kidney); (2) more slowly perfused lean tissue (as resting muscle,
skin); and (3) adipose tissue. This model was used by the reviewer to make some predictions about whole body
behavior of 85Kr in man (Kirk, 1972) which were subsequently found to be too simple (Kirk, 1973).
Whole body fat determinations were made in another laboratory (Hytten, 1964; and Hytten, et al., 1966)
using a similar closed system with a neoprene bellows spirometer replacing the water spirometer, which they
felt was the cause of Lesser and Zak subjects not reaching equilibrium. Equilibrium was claimed in 90-120
minutes for normal human subjects, and by 180 minutes for obese subjects. Radioactivity was monitored with
a radiation counter coupled to a ratemeter and recorder; equilibrium was considered complete when the
recorder trace was flat for ten minutes. Desaturation was found to be much slower than saturation, and a
component with a half-time of about 18 hours was reported. From his own experience with other animals, this
reviewer believes that the criteria for saturation were not adequate, and saturation was not complete in the
time stated.
S.Studies of Whole-Body Kinetic Behavior and Partition Coefficients.
The author has investigated the whole-body kinetic behavior of 85Kr in guinea pigs and, superficially, in
rats (Kirk, 1972; Kirk, et al., 1972; Kirk, 1973; and Kirk, et al., 1973). Saturation and desaturation curves in
unanesthetized guinea pigs breathing 85Kr in a closed system were found to be three- or four-component
exponentials as measured externally with a heavily collimated Nal detector. Half-times of 0.25-2.2 minutes, 6-
11.8 minutes, 21.5-41.7 minutes, and 88-178 minutes were found. Of the eight animals studied, one did not have
the fastest component, and one did not have the slowest. The desaturation curves (semi-logarithmic) for eight
guinea pigs are shown in Figure 1. Figure 2 shows the saturation and desaturation data for one guinea pig
plotted on the same graph as a smooth curve generated with kinetic parameters determined by graphic
resolution of the desaturation phase. The saturation phase usually agreed fairly well with the desaturation
phase for the same animal. However, substantial variation was noted between animals. Lw whole-bodyair
(fur-free) was found to be 0.1444 (S.D.=.0148).Females had a slightly higher Lw than males, but the difference
was not significant. The mean hairless weight of these animals was 840.1 grams (S.D. = 33.9).
The kinetic behavior of 85Kr in 250-gram female Rochester Wistar rats was investigated in preliminary
experiments. Three or four rats were saturated with 8sKr for either 12 or 33 hours in an exposure chamber, and
then repetitively counted in a large Nal well detector until preexposure background was reached Counting
did not begin until 3-6 minutes after the animals were removed from the 85Kr atmosphere In the 12-hour
experiment, desaturation curves had three components with half-times of 3.87-4.6 minutes, 20-23 minutes, and
81-143 minutes. After the 33-hour exposure, two-component curves were found with half-times of 76-15.7
minutes and 41.8-85 minutes. Three of the four animals in the 33-hour exposure had also been used in the 12-
hour exposure The change in curve fine structure is believed to be due to increased stress in the longer
exposure, including higher radiation dose, and the fact that counting did not begin until 5-6 minutes after the
IS/^H IT! Hm fC ° *-amber ~ lhich W°uld °bviate seein* any very 8horthalf-time component.
b igures 3 and 4 show the desaturation curves from one animal for both exposures The average I / - was
determined to be 0.092 (S.D. = 0.019) (seven determinations). exposures, ne average Lw rat:air was
-446-
-------
4. Tissue:Air Partition Coefficients.
Tissue:air partition coefficients, both L and Lw, have been determined for a number of biological solutions
and tissues.
Hardewig, et al, (1960) studied solubility of K5Kr in human blood at 37" C and found Lnla«ma-flirtobeO 051
(S.D. = .001) and Lblood:air = -05199 + .0001573 H where H = hematocrit. picu>md.
-------
REFERENCES
Albert, S. N., C. A. Albert and J. F. Faxekas(1960),,4 Rapid and Simple Method for Measuring the Rate
ofCerebralBloodFlowinHumanswithKrypton-85.,J.Lab.ClinMed.56-A73. /,0«,n v *
Alexander S. C., H. Wollman, P. J. Cohen, P. E. Chase, E. Melman and M. Behar(I964),Krypton-
85 and Nitrous Oxide Uptake of the Human Brain During Anesthesia .Anesthesiology 25:37
Bell, G. and A. M. Harper (1964), Measurement of Regional Blood Flow Fhrough the bkm from the
Clearance of Krypton-85., Nature (London) 202:704. ,„,,„, • ,«. r. , n * t ,z.
Bell, G. and A. M. Harper(1965), Measurement of Local Blood Flow m the Renal Cortex from the
Clearanceof Krypton-85. ,J.Surg. Res. 5:382. .
Bell, P. R. F. and A. C. Battersby (l9G7a),The Effect of Arterial pCO2 on Gastric Mucosal Blood Flow by
Clearance of Krypton-85. Surgery 62:468. .
Bell, P. R. F. andC. Battersby (1969), £/"/ec( of Vasopressm (Pitressm) on Gastric Mucosal Blood Flaw
Measured by Clearance of Krypton-85. Surgery 66:510. ,„,,„, • » ™ „„
Bell, P. R. F., C. Battersby and A. M. Harper(1967b),Gas£r ic Mucosal Blood Flow in the Dog Measured
by Clearance of Krypton-85 - the Response to Histamine. Brit. J. Surg. 54:1003.
Betz, E., D. H. Ingvar, N. A. Lessen and F. W. Schmahl(1966), Regional Blood Flow in the Cerebral
Cortex Measured Simultaneously by Heat and Inert Gas Clearance. Ada. Physiol Scand. 67:1.
Birtch, A. G., B. H. Casey and R. M. Zakheim(l9G7),Hepatic Blood Flow Measured by the Krypton-85
Clearance Technique. Surgery 62:174.
Brock, M. D. H. Ingvar and C. W. Sem-Jacobsen(1967),fle#wna/ Stood Flow in Deep Structures of the
Brain Measured in Acute Cat Experiments by Means of a New Beta-Sensitive Semiconductor Needle Detector.
Exp. Brain Res. 4:126.
Casey, B. H. and C. D. Thorburn,(19G5),Distribution of Blood Flow in the Skin Using Radioactive
Krypton-85 Clearance Techniques. In: Biology of the Skin and Hair Growth, A. G. Lyne and B. F. Short, Ed.
(Sidney: Angus and Robertson, Ltd.) 603.
Caster, W. O., J. Poncelet, A. B. Simon and W. D. Armstrong,(1956) Tissue Weights of the Rat. I.
Normal Values Determined by Dissection and Chemical Methods. Proc. Soc. Exp. Biol. Med. 91:122.
Cohen, L. S., W.C.Elliott andR.Gorlin(1964), Measurement of Myocardial Blood Flow Using Krypton-
85. Amer. J. Physiol. 206:997.
Carriere, S.(1970),^4 Comparison of the Disappearance curves of 133Xe and 8SKr for the Measuremenfof
the Intrarenal Distribution of Blood Flow. Can. J. Physiol. Pharmacol. 48:834.
Cosgrove,M.D.(1965),7%e Effect of Arterial Hypoxia on the Blood Flow Through the Renal Cortex. Brit.
J. Surg. 52:613.
Friedman, E., H. H. Kopald and T. R. Smith(1964),flefc'na/ and Choroidal Blood Flow Determined with
Krypton-85 in Anesthetized Animals . Invest. Ophthalmol. 3:539.
Friedman, E. andT. R. Smiih(1965),Estimation of Retinal Blood Flow in Animals. Invest. Ophthalmol.
4:1122.
Glass, H. I. and A. M. Harper(1962), The Measurement of the Partition Coefficient of Krypton Between
the Brain Cortex and Blood by a Double Isotope Method. Phys. Med. Biol. 7:335.
Glass, H. I., A. M. Harper and M. M. Glover ,(l96l),The Measurement of Local Cortical Blood Flow in
the Brain by the Analysis of the Clearance Curve of Krypton-85. Phys. Med. Biol. 6:65.
Haggendal, E.(1965), Effects of Some Vasoactive Drugs on the Vessels of Cerebral Grey Matter in the
Dog. Acta. Physiol. Scand. 66:Suppl. 258.55.
Haggendal, E. and B. Joheinsson,(19G5a),Effects of Arterial Carbon Dioxide Tension and Oxygen
Tension on Cerebral Blood Flow Autoregulation in Dogs. Acta. Physiol. Scand. 66:Suppl. 258.27
Haggendal, E., N. J. Nilsson and B. Norback(1965b),O« the Components ofssKr Clearance Curves
from theBrain of the Dog. Acta. Physiol. Scand. 66:Suppl. 258.5.
Hansen, A. T., B. F. Haxholdt, E. Husfeldt, N. A. Lassen, O. Munck, H. R. Sorensen and K.
Windkler(1956),Meas«remerc£ of Coronary Blood Flow and Cardiac Efficiency in Hypothermia by Use of
Radioactive Krypton-85. Scand. J. Clin. Lab. Invest. 8:182.
Hardewig, A., D. F. Rochester and W. A. Eriscoe,(l960),Measurement of Solubility of Krypton in
Water, Plasma, and Human Blood, Using Radioactive 8SKr. J. Appl. Physiol. 15:723
Harper, A. M., H. I. Glass and M. M. Glover(1961), Measurement of Blood Flow in the Cerebral Cortex
of Dogs by the Clearance of Krypton-85. Scott. Med. J. 6:12
D,Hef2; J' n " M> Hojle^ejg' G' D- Thorburn, H. H. Kopald and A. C. Barger(1962), Myocardial
Blood Flow Determined with Krypton-85 in Unanesthetized Dogs. Amer J Physiol 203-122
Hollenberg, M. and J. Dougherty(1966), Liver Blood Flow Measured by Portal Venous and Hepatic
Arterial Routes with ssKr. Amer. J. Physiol. 210:926
Holzman, G B H. N. Wagner, M lio, D. Rabinowitz and K. I. Zierler(1964), Measurement of Muscle
Blood Flow in the Human Forearm with Radioactive Krypton and Xenon. Circulation 30:27.
' F I K ^^^W^SSft/ atinMan wthKrypton-85. Proc. Nutr. Soc. 23:21.
r < CHr! £'S:U1 Taggart(1966),MeaSMremen* of Total Body Fat in Man by Absorption
' A' L*a*en(196l)>Quantitative Determination of Regional Cerebral Blood Flow
n
-448-
-------
Jansson, G., M. Kampp, O. Lundgren and J. Martinson (1966), Studies on the Circulation of the
Stomach. Acta. Physiol. Scand. 68:Suppl. 277:91.
Johansson, B., E. Linder and T. Seeman(1964), Collateral Blood Flow in the Myocardium of Dogs
Measured with Krypton-85. Acta. Physiol. Scand. 62:263.
Kampp, M., O. Lundgren and J. Sjostrand(1968a). On the Components of the 8sKr Washout Curves
from the Small Intestine of the Cat. Acta. Physiol. Scand. 72:257.
Kampp, M. and O. L\vtiAgren(1968b),Blood Flow and Flow Distribution in the Small Intestine of the Cat
as Analyzed by the 8SKr Washout Technique. Acta. Physiol. Scand. 72:282.
Kety, S. S.(1951), The Theory and Applications of the Exchange of Inert Gas at the Lungs and Tissues.
Pharmacol. Rev. 3:1.
Kety, S. S. and C. F. Schmidt (1948), The Nitrous Oxide Method for the Quantitative Determination of
Cerebral Blood Flow in Man: Theory, Procedure and Normal Values. J. Clin. Invest. 27:476.
Kirk, W. P.(1972),K>yptora-&5, A Review of the Literature and an Analysis of Radiation Hazards. Office of
Research and Monitoring Report (Washington: U. S. Environmental Protection Agency).
Kirk, W. P.(1973)./n Vivo Behavior and Effects of Krypton-85 in Guinea Pigs. Ph.D. thesis, University of
Rochester, Rochester, New York.
Kirk, W. P. and D. A. Morken(1972).7n Vivo Kinetic Behavior of85Kr in the Whole Guinea Pig. Paper 63,
17th. Annual Meeting, Health Physics Society, June 11-15,1972.
Kirk, W. P. and D. A. Morken(W73),PhysiologicalBehavior and Radiation Effects of8SKr Administered
to Guinea Pigs via the Respiratory System. (In review.)
Kitani, K.(1972),Solubility Coefficients of85Kr and 133Xe in Water, Lipids, and Blood. Scand. J. Clin. Lab.
Invest. 29:167.
Kitani, K. and K. Winkler(1972),/7r Vitro Determination of Solubility of133Xe and 8SKr in Human Liver
Tissue with Varying Triglyceride Content. Scand. J. Clin. Lab. Invest. 29:173.
Lassen, N. A.(1965),B/oorf Flow of the Cerebral Cortex Calculated from S5Kr Beta-Clearance Recorded
over the Exposed Surf ace; Evidence of Inhomogeneity of Flow. Acta. Neurol. Scand. 41:Suppl. 14:24.
Lassen, N. A.,K. Hoedt-Rasmussen, S. C. Sorensen, E. Skinhoj, S. Cronquist, B. Bodforss and D.
H. Ingvar(1963), Regional Cerebral Blood Flow in Man Determined by Krypton-85. Neurology 13:719.
Lassen, N. A. and D. H. Ingva.r(W61),The Blood Flow of the Cerebral Cortex Determined by Radioactive
Krypton. Experientia 17:42.
Lassen, N. A. and O. Munck(1955), The Cerebral Flow in Man Determined by the use of Radioactive
Krypton. Acta. Physiol. Scand. 33:30.
Lesser, G.T. andG. Zak(1963), Measurement of Total Body Fat in Man by the Simultaneous Absorption
of Two Inert Gases. Ann. N. Y. Acad. Sci. 110-40.
Linder, E.(1966), Measurements of Normal and Collateral Coronary Blood Flow by Close-Arterial and
Intramyocardial Injection of Krypton-85 andXenon-133. Acta. Physiol. Scand. 68:Suppl. 272:5.
Lockhart, E. A., J. H. Dirks and S. Carriere(1972),Effects of Triflocin on Renal Tubular Reabsorption
andBloodFlow Distribution. Amer. J. Physiol. 223:89.
Lundgren, O. and N. Kampp(1966), The Washout of Intra-Arterially Injected Krypton-85 from the
Intestine of the Cat. Experientia 22:268.
MacDonald, A. G.(1969), The Effect of Halothane on Renal Cortical Blood Flow in Normotensive and
Hypotensive Dogs. Brit. J. Anaesth. 41:644.
McHenry.L. C. Jr.(1964),Quantitative Cerebral Blood Flow Determination. Application of a Krypton-85
Desaturation Technique in Man. Neurology 14:785.
Masson, M. B. R. and K. Taylor (1967),Solubility of Krypton-85 in Olive Oil and Human Fat. Phys. Med.
Biol. 12:93.
Mellemgaard, K., N. A. Lassen and J. Georg(l962),Right-to-Left Shunt in Normal Man Determined by
the Use of Tritium and Krypton-85. J.Appl. Physiol. 17:778.
Michenfelder, J. D., J. M. Messick and R. A. Theye(1968), Simultaneous Cerebral Flow Measured by
Direct and Indirect Methods. J. Surg. Res. 8:475.
-449-
-------
TABLE 1. Compartmental Analysis from Cerebral Blood Flow Determinations
in Human Subjects from Wollman, et al, (1965).
Subject
Awake
Hypoxic
Anesthetized (pCC>2=40)
Anesth. (pCO2=10-20)
Compartment
Slow
Fast
Slow
Fast
Slow
Fast
Slow
Fast
Average
% of Brain
50.8
49.2
31.5
68.5
47.2
52.8
68.6
31.4
Tl/2
4.36
1.07
2.26
.80
4.85
1.35
5.63
1.34
(min)
TABLE 2. Analysis of 85Kr Desaturation from Cat Brain with GM and
Semi-Conductor Detectors from Brock, et al., (1967).
Fast Component Slow Component
Detector T 1/2 (min) % of Act. T 1/2 (min) % of Act.
GM
Semi-conductor
over cortex
Semi-conductor
over artery
0.48
0.63
0.42
44.7
44.1
48.7
2.48
3.01
2.39
55.3
48.7
51.3
TABLE 3. Analysis of 85Kr Desaturation in Dog Kidney from Thorburn, et al, (1963).
Component Tl/2 (min) % of Total Act.
I
II
III
IV
Outer cortex
Inner cortex and
outer medula
Inner medula
Perirenal and hilar
fat
0.16
0.65
5.5
44.2
80
16
2
2
TABLE 4. Analysis of 85Kr Desaturation in Dog Livers from Birtch, at al, (1967).
Type Component 1 Component 2 Component 3
Experiment Injection T 1/2 (min) % Act. T 1/2 (min) % Act.
Acute-(anesth.)
Chronic-(awake)
One chronic dog
HA
PV
Both
HA
PV
Both
HA
PV
.25
.26
.33
.22
.38
.38
.39
.45
.42
83.5
15.9
77
38.5
.38
.55
.55-
.45-
.74-
.51-
.53
1.3
.8
2.8
.75
.55
1.7
.82
55
16.5
34.6
55
23
37
1.13 - 2.3
1.33-2.2
1.63 2.8
1.33-3.3
2.4
45
49.5
45
24.5
-450-
-------
TABLE 5. Analysis of 85Kr Desaturation in Cat Small Intestine
from Lundgren, et al., (1966); and Kampp, et al, (1968a).
Component No. Identified As
T 1/2 (a) (min) % orig. act. (a)
I
II
III
IV
0.104 ((.04)
Counter-current exchange in
in mucosal vascular loops
Mucosa
Muscularis and serosa about
equally
Perivascular fat, mesentery, 61.8 (44)
delayed recovery from vascular
loops of villi
1.55 (.48)
5.29 (1.93)
39.9 (10.1)
32.1 (10.5)
28 (12.3)
Not given and not
counted in%
TABLE 6. Analysis of 85Kr Desaturation in Cat Retina
from Friedman, et al., (1964).
Component Identified As
T1/2(min)
I
II
III
Blood flow in choroid
Retinal vessels
Sclera, choroidal stroma,
vitreous, tapetum, etc.
.050 - .083
.25 - .78
1.5 3.0
9 44
TABLE 7.7B- 81Kr Kinetic Parameters in Human Hands and Knees
from Tobias, et al., (1949).
Average Hand Average
Component T 1/2 (min) Fract. Act.
Knee Average
T 1/2 (min) Fract. Act.
I
II
III
4.3
35.9
188
.11
.55
.33
55
402
.34
.67
-451 -
-------
TABLE 8. In Vitro Ostwald Coefficients (L) for85 Kr into Selected Biological Solutions (a)
Temperature Number of
Solution
Guinea Pig Blood
Guinea Pig Brain (b)
Dog Blood (c)
Cat Blood
Chinese Hamster Blood
0.9% NaCl Solution
10% EDTA Solution
(°C)
39
37
36'
37
35.5
36
39
37
37
37
Determinations
6
8
16
12
8
7
8
10(d)
4
8
L
.0513
.0538
.0546
.0494
.0691
.0595
.0559
.0822
.0472
.0406
a
.0041
.0033
.0076
.0179
.0054
.0065
.0040
.0014
.0012
.0028
(a) Assuming 760 mm Hg total pressure.
(b) Solubility was determined for a 20% brain homogenate in normal saline
and the coefficient for brain calculated from that value.
(c) Dogs were fed 30-60 minutes prior to bleeding.
(d) Five determinations were made on each of 2 pooled samples
(Each sample represented 20 animals.)
-452-
-------
TABLE 9. Comparison of Predicted and Measured Partition Coefficients for 85 Kr.
Species
Cow
Cat
Dog
Rabbit
Man
Rat
Guinea Pig
Organ or Tissue
Brain Homogenate
Brain Whole
Liver Whole
Muscle Homogenate
Blood
Whole Body
Whole Body
Omental fat
Subcutaneous fat
Adrenal
Liver
Muscle
Bone Marrow
Brain
Brain
Calculated
Coefficient
.0488(b) (c)
.0544(a) (d)
.0544(a) (d)
.0442(b) (c)
.0433(b) (c)
•076(a) (g)
.1482(a) (e)
.4254(a) (e)
.4254(a) (e)
.1067(a)(e)
.0524(a) (e)
.0752(a) (e)
.0863(a) (d)
.0744(a) (d)) (h)
Ibid.
Measured
Coefficient
.0454(b) (j)
.0625(a) (f) (i)
.0625(a) (f) (i)
.0439(b) (j)
.0455(b) (j)
.0921(a) (i)
.1444(a) (i)
.4213(a) (i)
.4054(a) (i)
.1017(a) (i)
.0768(a) (i)
.0396(a) (i)
.1342(a) (i)
.0405(a) (i)
Ref. for Measured Coeff.
Yeh, et al, (1965)
Lassen, et al., (1955)
Hollenberg, et al., (1966)
Yeh, et al., (1965)
Ibid.
Kirk (unpublished data)
Kirk (1973)
Ibid.
Ibid.
Ibid.
Ibid.
Ibid.
Ibid.
Ibid.
'a'
w tissue: air'
(c) Direct determination of tissue composition.
(d) Tissue composition from Handbook of Biological Data, Spector (1956).
(e) Tissue composition from Pace, et al., (1945).
(f) Calculated from reported tissue:blood coefficient using Lblood:air= °'°59-
(g) Tissue composition from Caster, et al., (1956).
(h) 50% non-neutral fat correction used.
(i) Determined in vivo.
(j) Determined in vitro.
-453-
-------
TABLE 10. Annual Dose (a) to Guinea Pig Tissues and Organs from Respiratory
Exposure to 85Kr at Current (MPC)a Levels.
Occupational (b) Unrestricted (c)
Tissue rad/year a rad/year a
Omental fat
Subcutaneous fat
Thymus
Lymph nodes
Bone marrow
Adrenal
Thyroid
Liver
Large intestine
Small intestine
Testes
Ovaries
Kidneys
Uterus
Stomach
Brain
Eyes
Muscle
Seminal vesicles
Spleen
Heart
4,460
4,290
2,740
1,460
1,420
1,080
877
813
788
764
613
608
455
442
439
429
425
419
391
386
317
681
957
1,390
794
1,260
404
395
298
765
414
160
145
199
134
218
177
136
183
72
96
157
586
564
360
191
187
141
115
107
104
100
81
80
60
58
58
56
56
55
51
51
42
90
126
183
104
165
53
52
39
59
54
22
19
26
18
29
23
18
24
10
13
41
(a) ftrad/year beta dose to organ/tissue from contained 85Kr.
Values are for bloodless tissue; dose from isotope outside the tissue
is not considered.
(b) 2,000 hours at 10-s (iCi/cm3.
(c) 24 hours/day, 365 days at 3 x 10-7 fiCi/cm3.
-454-
-------
NORMALIZED DESATURATION CURVES FOR GUINEA PIGS FULLY
SATURATED WITH KRYPTON-85
1/1
<
I
z
u.
UJ
z
O
tr
D
_
U.
u_
O
z
Q
i-
U
<
o:
LL
30 40 50 6O
I I I L
70
MALE
MALE it2
MALE #3
MALE #4
FEMALE #1
FEMALE #2
FEMALE #3
FEMALE #4
0.001
111II11111111111II11111111111111111111111111111111 /
0 1OO 200 3OO 4OO 5OO
TIME AFTER BEGINNING OF DESATURATION (MIN)
Figure 1.
-455-
-------
KRYPTON-85 SATURATION AND DESATURATION CURVES FOR
GUINEA PIG F-3
(NORMALIZED TO 1.0 FOR 100% SATURATION)
100-
350
TIME IN PHASE (MIN)
Figure 2.
-456-
-------
KRYPTON-85 DESATURATION
'Ao(total) ='7,834
RAT*1, RUN
10,000-
O= ORIGINAL DATA
A = LONGEST COMPONENT STRIPPED
O=2 LONG COMPONENTS STRIPPED
1000-
I
100^
10-
0
CURVE
1
2
3
A0(CPM)
1,870
11,532
4,432
t^(min)
94.2
22.6
3.75
k(min-l)
.0074
.37
.185
4 6 8 10 12
HOURS POST EXPOSURE
Figure 3. Experimental 85Kr desaturation curves in rat (short exposure).
-457-
-------
10,000
1000-
100^
10-
KRYPTON-85 DESATURATION
-A0(t0tal) ='4-3*3 CPM
RAT*1,RUN*2
O= ORIGINAL DATA
A* LONG COMPONENT STRIPPED
CURVE
1
2
A0(CPM)
1,596
12,747
ti(tnin)
65
15.7
Mmirf')
.00815
.044
8 10 12 U
HOURS POST EXPOSURE
Figure 4. Experimental 85Kr desaturation curves in rat (long exposure).
-458-
-------
COMPARTMENT HALF-TIMES FROM LITERATURE
ORGAN
BRAIN
HEART
KIDNEY
SPECIES
CAT
DOG
MAN
DOG
DOG/MAN «.
DOG
DOG •
HALF-TIMES, min
(AVERAGE SLOPE)
> m (WHOLE BRAIN)
CO
(+2 NOT GIVEN)
LIVER
STOMACH
INTESTINES
EYES
TESTES
SKIN
HAND
ARM
KNEE
WHOLE BODY
DOG
CAT
DOG
CAT j__
CAT •__•
SHEEP
RABBIT
DOG
MAN
MAN
MAN
MAN
RAT
GUINEA Pl«
MAN
( + 1 NOT GIVEN)
• •
(SEVERAL SHORT)
(1 SHORT)
(SEE TEXT)
REFERENCE
LASSEN & INGVAR (1961)
LASSEN (1965)
BROCKET AL (1967)
INGVAR & LASSEN (1961)
GLASS ET AL (1961) HARPER ET AL. (1961)
HAGGENDAL (1965) NILSSON (1965)
HAGGENDAL & JOHANSSON (1965)
HAGGENDAL (1965)
INGVAR & LASSEN (1961)
LASSEN ET AL. (1963)
ALEXANDER ET AL. (1964)
WOLLMAN ET AL. (1965)
IBID.
HANSEN ET AL. (1956)
HERD ET AL. (1962)
ROSS ET AL. (1964)
COHEN ET AL. (1964)
JOHANSSON ET AL. (1964)
LINDER (1966)
THORBURN ET AL. (1963)
BELL& HARPER (1965)
COSGROVE (1965)
MACDONALD (1969)
CARRIERE (1970)
LOCKHART ET AL. (1972)
HOLLENBERG & DOUGHERTY (1966)
BIRTCH ET AL. (1967)
JAIMSSON ET AL. (1966)
BELL & BATTERSBY (1967) BELL ET AL. (1967)
LUNDGREN & KAMPP (1966) KAMPP ET AL. (1968)
FRIEDMAN ET AL. (1964)
SETCHELL ET AL. (1966)
THORBURN ET AL. (1968) CASEY & THORBURN (1965)
BELL & HARPER (1964)
SEJRSEN (1971)
TOBIAS ET AL. (1949)
HOLZMAN ET AL. (1964)
TOBIAS ET AL. (1949)
KIRK (1969)
KIRK (1973)
HYTTEN ET AL. (1966)
0.01
0.1
Figure 5.
1.0 10
HALF-TIMES, min
100
1000
-------
IN VIVP KR-BS Lw
TI55UE
FDR EUINETR PIE5
PRRTITIDN COEFFICIENT Cl_w D
OMENTRL FRT ' ^"J
SUBCUTRNEOUS FRT
THYMU5 «
LYMPH NODES <
BONE MRRROM <
4
RDRENRLS
THYROID
LIVER
.. —
— . .. „_- .., . . . . ,.,^
TESTES
OVRRIES
WDNEY5
UTERUS
STOMRCH
BRRIN
EYES CWHOLE1
MUSCLE
t - m |
, ,
a 1
• 1
—a 1
• 1
• _•.___., -1
• „ ^> i
• , ,_ _. n i
O_._j
• _,,j
t. n i
L.. ^ •
^ '•"' •" "'
• •— I
1 D 1
1 • 1
1 D— *
1 • 1
i a 1
• 1
t— Q 1
i • 1
1 D— - i
• ' 1
1 O 1
^^™" V ^^~™n
SEMINRL VESICLES \ — g-<
SPLEEN
HERRT
.00 1
a r WHaue Tissue • s Bt_aai
i a i
i a — i
.01 .10 ska
t>i_es5 Tissue BRRS s 2 sj>.
Figure 6.
-------
35
CDMRRRTMENT HRLF-TIME5 FRDM LITERRTURE-ODES
DRBRN HRLF-TIME5 CMIN.D REFERENCE
BRRIN
HETHRT
JCfDNETf
LIVER
STDMRCH
5 KIN
x
+
+f
+
H
If
-H-
^
+
+
nil
-i -L
«
-1—
-1
1 I
,1
+
•H-
-H-
H-
.«
•
+
-»
+
+
! +2 SU3WO
4.
1
-»-
3
4-
•M-
•f
INBVRR + LRSSEN CI9EI3
BUR55 CT RL. C IBB 13
HRRPCR ETT RL. C IBS 13
HRBBENDRL ET RL. CIBfiS^
NILSBDN CIBBSn
HRBEE>4DRL+UDHRN55DNC 1 B653
HREBENDRL ETT RL. Cl BSCT
HRKSOvl CT RL. C 19563
HCFB> ETT RL. CI9B23
ROSS ETT RL. C 1 B6H3
CDHCN ETT RL. C 1 9BH3
UOHRMSSDN ETT RL. C 1 BBH3
UNDER CI9EB3
THORBURN CT RL. CI9633
BEOJ. + HRRPE3? CI9E53
CDEK3RDVC C 1 UtiLIJ
MRCDDNRLD CI9B93
CHffrHfcXET C 1 9703
LOOCHRRT CT RL. CIET723
HOLLCNBCRB 4-
DDUBHCRTY C 1 9BB3
BVITCH ET RL. C 19673
BELL+BRTTERSBY C 19673
BELL ET RL. CIB673
BELL
-------
1.0
39
0.1
w
M
M
H
M
8
o.oi
38
37
PARTITION COEFFICIENT (L ) FOR 85KR
w
INTO GUINEA PIG ADRENAL
IN VIVO MEASUREMENT VS MODEL PREDICTION
36
TISSUE COMPOSITION
FAT
PROTEIN
WATER
14.88
19.32
65.80
a
3.52
4.27
2.42
From: Pace and Rathbun (1945)
CO
2
0.001
32.0
32.1
32.2
104/T
32.3
32.4
-462-
-------
39
38
37
36
0.1
w
o
§
H
H
H
PM
0.01
PARTITION COEFFICIENT (L ) FOR 85KR
w
INTO GUINEA PIG BONE MARROW
IN VIVO MEASUREMENT VS MODEL PREDICTION
TISSUE COMPOSITION
FAT
PROTEIN
WATER
10.4
17.0
71.0
a
?
?
?
From: Handbook of Biological
Data, Spector (1956)
0.001
32.0
32.1
32.2
32.3
104/T
05
§
to
32.4
-463-
-------
1.0
39
38
37
36
Or
PARTITION COEFFICIENT (L ) FOR KR
INTO GUINEA PIG LIVER
IN VIVO MEASUREMENT VS MODEL PREDICTION
H
&
W
Pn
En
O
M
H
0.1
0.01
TISSUE COMPOSITION
FAT
PROTEIN
WATER
3.48
27.46
69.06
a
1.32
2.21
1.71
From: Pace and ^Rathbun (1945)
0.001
32.0
32.1
32.2
104/T
32.3
32.4
-464-
-------
1.0
39
n
38
37
36
o.i
PARTITION COEFFICIENT (L ) FOR 85KR
w
INTO GUINEA PIG OMENTAL FAT
W
H
W
O
§
IN VIVO MEASUREMENT VS MODEL PREDICTION
0.01
TISSUE COMPOSITION
FAT
PROTEIN
WATER
84.13
5.62
10.25
a
4.79
7.65
5.97
From: Pace and Rathbun (1945)
0.001
32.0
32.1
32.2
104/T
32.3
32.4
-465-
-------
1.0
39
38
37
36
0.1
u
M
ft,
u
§
as
0.01
PARTITION COEFFICIENT (L ) FOR 85KR
w
INTO GUINEA PIG SUBCUTANEOUS FAT
IN VIVO MEASUREMENT VS MODEL PREDICTION
TISSUE COMPOSITION
FAT
PROTEIN
WATER
84.13
5.62
10.25
o
4.79
7.65
5.97
From: Pace and Rathbun (1945)
0.001
32.0
32.1
32.2
104/T
32.3
32.4
-466
-------
1.0
39
38
37
36
0.1
w
M
O
M
fa
8
o
§
0.01
PARTITION COEFFICIENT (L ) FOR 85KR
w
INTO GUINEA PIG THIGH MUSCLE
IN VIVO MEASUREMENT VS MODEL PREDICTION
TISSUE COMPOSITION
FAT
PROTEIN
WATER
7.83
17.52
74.65
a
2.36
2.45
0.66
M
r-l
From: Pace and Rathbun (1945)
0.001
32.0
32.1
32.2
104/T (°
32.3
32.4
-467-
-------
J
S
in
H
Z Pi
UJ H
k
u_
u
H
u
H
c
J
D
V
CDMPRRI5DN DF MER5URED RND CRLCULRTED
PRRTITIDN CDEFFICIENT5 FDR
H + I .00HBX
C RrH.SBl
0. I 0.2 0.3.
MER5URED COEFFICIENTS
Figure 14.
0.H
-------
THE BIOLOGICAL EFFECTS OF THE RADIOACTIVE NOBLE GASES*
D. A. Morken
University of Rochester
Rochester, New York 14642
Abstract
The biological effects of the noble gases depend on the physical proper ties of these gases, since they are non-
reactive chemically under physiological conditions. A property of interest for the non-radioactive gases is
their solubility in the various body tissues and fluids.
Some of these noble gases have the additional property of radioactivity. The combination of radioactivity
and solubility permits the identification of a critical organ according to established procedures. The
pathological response is not always greatest, or even of singular importance, in the organ so identified by
calculation. Biological responses are desired for a prudent evaluation of the biological effect, but the choice of
response and the mode of irradiation pose a large problem.
The radioactive gases decay schemes include the emission of alpha particles, beta particles, positrons, and
electron capture; and this decay is sometimes accompanied by gamma radiation. The resulting daughter
products may be foreign or natural to the body; whether stable or radioactive, all are reactive chemically. The
effects due to the radiation, however, probably exceed those chemical effects which may result.
Radon has no stable isotope. The radon isotopes decay primarily through the emission of alpha particles.
The biological effects from radon may arise in greater measure from its daughter products, which include
alpha particle, betaparticle, and gamma ray emitters, than from the parent radon.
The consequences of these properties are discussed with particular reference to krypton-85 and radon-222.
INTRODUCTION
A prudent evaluation of biological responses to the noble gases can only be obtained by direct
experimentation, but such experimentation has not been the general rule for the noble gas series of elements.
Only radon has been studied in depth, but a beginning has been made on the effects of inhaled krypton.
Biological studies with radioactive argon, xenon, neon, and helium are difficult to perform in animal
systems bacause of their very short half-lives. These elements, except for radon, are beta emitters; their
solubility in tissues is a physical property dependent on their molecular weight. Due to these similarities
information about one of these gases should be applicable in evaluating the others. Recent experiments with
radon suggest that the chemical properties may be more important than radiation in producing long-term
effects. It is also necessary to consider the role played by the daughter products.
The isotopes of helium decay either to hydrogen, a natural constituent of the body, or to lithium. Neon
decays to either sodium or fluorine, and argon decays to either potassium or chlorine — all of which are normal
constituents of the body. Krypton decays to either rubidium or bromine; and xenon decays to either cesium or
iodine. Radon decays to polonium, and thence through a series of elements which includes lead and bismuth.
The decay of radon and its daughter products is accompanied by alpha, beta, and gamma radiation.
A proper study of the biological effects of these radioactive gases should include the effects of the daughter
products which, in general, are stable isotopes. External irradiation by these radioactive gases does not
involve the stable daughters, but inhalation of these gases includes the stable daughters. Thus, the age of the
radioactive gas may not be the most important factor in evaluating biological effects when the effects are due
to chemical properties and not to radiation alone.
The isotopes of greatest environmental interest are krypton-85, which produces rubidium-85 (stable) by beta
emission; xenon-133, which produces cesium-133 (stable) by beta emission; and radon, which decays by alpha
emission to a series of short-lived elements that are alpha, beta, and gamma emitters. The 5.3-day half-life of
xenon-133 limits its usefullness in biological experiments. The 10.6-year half-life of krypton-85 makes it a
particularly useful isotope for biological experimentation. Krypton is also of environmental concern because
of its relatively long half-life, and its release from nuclear reactor and fuel processing plants.
Because these gases are not accompanied by large amounts of chemically reactive materials, they offer a
purity of radiation invaluable for studies involving internal emitters which is not available outside the series
of noble gas elements. The purity of the radiation, or lack of chemically active materials, may approach that of
x-or gamma radiation for all practical purposes. The comparatively uniform distribution of these gases in the
body tissues, as determined by solubility studies, results in a relatively uniform dose distribution with no
great differences in effect from similar doses of x- or gamma radiation.
"This paper is based on work performed under contract with the U. S. Atomic Energy Commission at the
University of Rochester Atomic Energy Project and has been assigned Report No. UR-3490-383.
-469-
-------
BIOLOGICAL EFFECTS
The acute lethality by inhalation (30-day LD-50) for krypton-85 in guinea pigs has been reported by Kirk
(1973). The experimental arrangement Kirk used minimized the external exposure. The LD-50 dose required
about 12 hours at a concentration of 375/i Ci/cm3. The whole body absorbed dose calculated from this air dose
amounted to 340 rads — which is within the range of x- and gamma ray doses reported for the guinea pig. Kirk
also described the appearance of beta burns on the nose, and epilation around the nose and eyes, but skin
healing and hair regrowth were complete in 4 to 10 weeks. He concluded that deaths which occurred before the
30 days were due to a hematopoietic syndrome, whereas those which occurred later were due to lung injury.
The lung dose at the LD-50 level was estimated to have been 8,000 rads.
Kirk's report constitutes essentially all that is known about the internal distribution, dose estimates, and
the biological responses for inhaled krypton-85. Radon has been studied more extensively in terms of
biological effects.
Inhaled radon presents the body with three different kinds of exposure. Exposures from radon alone could
be compared to krypton exposures — if it could be obtained alone. Radon, however, is accompanied by a series
of short-lived radioactive decay products; consequently exposures with radon must always account for these
additional radioactivities and distributions of the decay products. These products are produced within the
animal from the inhaled radon. Thus a second kind of exposure, from the isotopes of polonium, lead, and
bismuth, delivers a radiation dose to the entire animal. The distribution of this dose is essentially non-uniform
because the isotopes are treated independently by the body tissues — within the confines of the short half-
lives. It is possible with injection techniques to produce whole body exposures to these decay products alone
(Hollcroft,e£ o/.,1951), and the histopathology and acute toxicity following such injections have been reported
(Hollcroft,e£a/.,1955).
The third kind of exposure is to the inhalation of these short-lived radon decay products present in the
inhaled atmospheric air. These products are attached to dust particles which may be deposited in the lung
when inhaled. Because of their short half-lives (3 to 30 minutes), these products accumulate and decay within
the lung structure, since the clearance mechanisms which operate on dust in the lungs are thought to be
ineffective in such short times. This method of exposure might deliver large doses to lung alone, and is
believed to be the kind of exposure which results in the bronchial cancer found in uranium miners. In the
absence of other information, the large physical dose which can be delivered to the lung by this airborne
radioactivity is considered sufficient reason to designate the lung as the critical organ for exposure to radon
by inhalation.
The radon experiments in this paper consider only the inhalation of radon and its subsequent decay within
the animal body. The effects of the inhalation of the daughter products are described in another paper in this
symposium (Morken, 1973).
These experiments were conducted with a closed respiration system in which mice (CAFj strain) could be
exposed to a radon-222 concentration of 0.22 mCi per liter for periods of up to 40 hours. The contained air was
kept free from the decay products of radon by air filtering and a rapid air turnover rate.
In this system, mice exposed from 5 to 40 hours showed a 30-day LD-50 near 7.5 mCi-hours/liter (Morken,
1955). This amounted to about 400 rad, when converted to an internal dose, and fell within the range of the x-
ray doses required for the same effect. On an energy basis, this is not different from the LD-50 found with
krypton-85. Emaciation, reddening of the ears, and dullness of the eyes were the first signs of tissue damage.
The mice, which died within a few days after exposure, exhibited the pseudoparalysis which has been
reported. The fur generally became bristly, even for the low doses. Deaths which fell within the 30-day period
occurred in the first two weeks; no further deaths occurred until after 60 days. The limited pathology
information describes changes similar to those following whole-body exposures to x-rays. The changes in the
circulating cell numbers were not as great as with an equivalent x-ray exposure. A definite, but not
pronounced, anemia occurred. More profound effects were found following x-ray exposures.
The histopathology in mice exposed to radon at the sub-acute level (one-half the LD-50) has been reported
(Scott, 1955). Of the internal organs, the spleen evidenced the greatest early damage — and also evidenced
rapid repair. Late injury was found in the kidney. The estimated dose to the kidney, due principally to the
cumulation of the decay products, was 288 rad, compared to a whole-body dose of 255 rad (Morken, 1959). The
lung and trachea showed no lesions, during the early period following exposure, that could be attributed to the
radon. Changes in the epithelial lining of the small bronchi appeared at 2 and 5 months, and were fairly clear
at 7 months. The changes consisted mainly in an irregularity of the cells, some loss of polarity and the
presence of a number of large cells containing large hyperchromatic nuclei (in many of which were large
deeply-staining clumps of chromatin).
Mortality, growth, and the hemogram were studied in mice which received single exposures to radon The
doses ranged from 50 to 150 rad average dose to the whole body (Morken, 1961). Lifespan was shortened in
proportion to dose, but the major effect appeared to be an early effect for a small part of the population, and
little or no effect on the remainder.
The blood cell picture following single doses was essentially that seen after other kinds of irradiation -
except for the erythrocyte count, which became depressed in the normal manner, but showed no recovery.
This red cell depression remained at the depressed level for nearly one year after exposure, after which it
merged and decreased with the control data. The amount of the depression was fixed at about 15 percent,
independent 01 dose. f
-470-
-------
Growth of mice following single doses was stopped within a few weeks after exposure. After this initial
phase, growth resumed and appeared to follow the normal pattern, but the control values were not attained.
Stunted growth appeared to be an early and permanent effect.
Of these changes none appeared to result from the late effects of irradiation, since the initial change in each
occurred early in the period following exposure. Only the weights of the mice remained different from those of
the controls during the late period of life.
These same parameters were followed in mice which received doses from radon in one or several exposures,
with average whole body cumulated dose ranging from 150 to 600 rad (Morken, 1964). Lifespan was shortened
in proportion to total dose, whether given in one or several exposures. The mortality data suggested a bimodal
response composed of one group which died early after the exposure or group of exposures, and another group
which died later. As the total dose was made larger, a greater fraction of the population was transferred from
the late to the early deaths group. The probability of a radon-induced kidney lesion was found to have no
significance on mortality in these experiments.
The injury which produced lifespan shortening was found to be repaired by 50% over a two-week interval,
and the irreparable fraction of total injury was the same whether the injury was produced by single or multiple
doses.
Growth of mice was stopped immediately after exposure. After this initial phase, growth resumed in an
apparently normal pattern, but with a permanent decrement in body weight. The decrease was proportional to
total dose, whether delivered in one or several exposures, although the initial rate of loss of weight was
proportional to the individual doses.
The hemogram, following either single or multiple exposures to radon, was essentially the same as that
found after exposure to other kinds of radiation. Multiple doses produced immediate injury proportional to the
individual dose, but recovery processes were more influenced by the total accumulated dose. An exception to
this general picture was the erythrocyte count, which was depressed by 15% following the first exposure, from
which no recovery was evident, and was not further changed by additional exposures.
Of these parameters, lifespan and body weight were affected in proportion to total dose, whether
fractionated or not, thereby allowing the conclusion that the injury due to the alpha dose delivered from radon
is repaired to the same extent of 50% in each case. The effect on the red cell illustrates that permanent, but
delimited, injury can be produced by alpha irradiation. The remainder of the blood cells suggest that some
injuries from alpha radiation can be completely repaired.
From these experiments with radon we obtain a general conclusion that because of the relationship of effect
to cumulated dose, the early effects found here with short exposures would not be evident in experiments with
continual exposure to radon at lesser concentrations, or with long-lived internal alpha emitters.
SUMMARY
The effects of inhaled krypton-85, a beta emitter, and radon-222, an alpha emitter, are similar at the acute
lethal level to each other and to those from external x- or gamma irradiation. Work with krypton has just
begun. Radon has been studied more extensively in terms of mortality, body weight, and the hemogram. In
general, the effects are not different from those found with x- or gamma radiation.
Radon produces a few specific differences from x- or gamma irradiation. Red cell forming tissue is only
slightly injured, although the other blood cell forming tissues respond in the expected manner. The growth
pattern is altered with permanent stunting a result. These specific effects may be a result of a special kind of
injury which the alpha particle can produce. Similar experiments with beta particles from krypton-85 may
resolve this difference and may lead to a better understanding of the nature of these effects.
REFERENCES
Hollcroft, J. W. and E. Lorenz (1951), The 30-day LD-50 of Two Radiations of Different Ion Density. J.
Natl. Cancer Inst. 12:533.
Hollcroft, J. W.,E. Lorenz, M. Mastthews, and C. G. Congdon(W55),Long-Term Survival Following
X-Irradiation and the Irradiation of the Alpha Particles from Radon and its Decay Products. J. Natl. Cancer
Inst. 15:1059.
Kirk, W. P. II(,1973),,InVivoBehavior and Effects of Krypton-85 in Guinea Pigs. Dissertation, University
of Rochester, Department of Radiation Biology and Biophysics, Rochester, New York.
Morken, D. A.(1955), The Acute Toxicity of Radon. AMA Arch. Ind. Health 12:435.
Morken, D. A.(1959), The Radiation Dose to the Kidney of the Rat from Inhaled Radon. AMA Arch. Ind.
Health 20:505.
Morken, D. A.(1961), The Effect of Inhaled Radon on the Survival, Body Weight, and Hemogram of the
Mouse Following Single Exposures. University of Rochester, Atomic Energy Project, Report No. UR-593.
Morken, D. A.(1964), The Effect of Inhaled Radon on the Survival, Body Weight, and Hemogram
Following Multiple Exposures. University of Rochester, Atomic Energy Project, Report UR-624.
Vlorken,D.A.(l973),The Biological Effects of Inhaled Radon. This Symposium.
Scott, J. K.(1955),T/ie Histopathology of Mice Exposed to Radon. University of Rochester, Atomic Energy
Project, Report UR-411.
-471-
-------
KINETICS AND DISTRIBUTION OF XENON-133 AND KRYPTON-85 IN THE HUMAN BODY
A. D. Turkin and Yu. I. Moskalev
Institute of Biophysics
Ministry of Health
Moscow, USSR
Abstract
The kinetics of86Kr and13iXe metabolism were studied in experiments on 15 male human beings whose ages
ranged from 25 to 50 years. The half-life of1MXe in the lungs varied from 18 to 42 seconds, with an average half-
life of 30 seconds. The elimination of133Xe from the blood is nearly the same as that for the lungs. The half-
lives of133Xe in the muscle and fat tissues ranged from 0, 1 to 0, 42 and 4.6 to 7.4 hours, respectively. The 133Xe
uptake by the skin does not exceed 0.4 per cent.
The elimination of85Krfrom the human body is well described by the sum of exponents. The average half-
life ofssKr in the blood and lung is the same as those for 133Xe, however, in the fat tissue it ranged from 1.8 to 3. 7
hours, with an average half-life of 2. 7 hours, and in the muscle tissue it ranged from 4 to 11 minutes.
INTRODUCTION
In connection with a broad program of nuclear energy development throughout the world, much attention
should be given to the problem of controlling radioactivity releases into the atmosphere. Besides the problem
of protecting small groups of people living in the vicinity of the release source, there are problems connected
with the global contamination of the atmosphere.
In order to estimate the degree of risk of environmental contamination with the noble gases, one should
know the factors which influence tissue doses within the human body.
Internal radiation tissue doses can be predicted only on the basis of data concerning the accumulation,
distribution, and elimination of these gases from the human body.
There is limited data at present on the behavior of the radioactive noble gases within the human and animal
bodies. According toKalantarov,e£ al., (1966), the kinetics of 133Xe elimination from the human body after an
intravenous injection is described by four exponents. The half-lives ranged from 0.3 to 30 minutes. The
retention of 133Xe in respiratory organs after a one-hour exposure equals 4 percent and after ten hours 0.5
percent of the total activity (Osanov, et al., 1970). This paper presents experimental data on the kinetics of 85Kr
and I33Xe metabolism in the body.
METHODS
Experiments on the kinetics of 85Kr and 133Xe metabolism were carried out on 15 male volunteers with ages
ranging from 25 to 50 years. These experiments were performed in cooperation with E. S. Trukhmanova and S.
V. Levinsky.
In order to ensure contact with radioactive gas, the subject was placed into a hermetic exposure chamber
with a volume of 3.1 m3. The duration of the exposure ranged from 0.5 to 66 hours. After each timed exposure
the subject was again enclosed in the chamber for periodic measurements, until all the accumulated gas was
eliminated. These experiments showed that noble gases are eliminated from the body almost completely
during a period of 10-30 hours. Figure 1 gives a typical picture of the change in the radioactive gas
concentrations in the air of the measuring chamber at the various intervals. From this data the noble gas
elimination rate can be calculated by means of differentiating the curves Q = f(t) since
dQ d A,
C = V - = - k.
dt at
where Vfc is volume of the measuring chamber and Afe is gas activity in the chamber.
In the case where the concentration was measured only once before the subject left the measuring chamber,
the average value of the elimination rate was found by dividing the gas activity (Ak) by the time "t" of the
subject's residence time in the measuring chamber, which in this case is
A,
n "•
ct
Figure 2 gives data characterizing a change in the "Kr elimination rate from the body of a subject. The
In addition to determining the activity of the gas by its release into a closed volume we also studied its
distnbu ion within the body by means of a whole body scintillation counter (Nal crystaUO T^O mm and PEM
13) coupled mth a pulse height analyzer. The background was reduced by a lead screen 30 mm Sick which
also served as a collimator. The counting rate was measured at th« K^t «,7u * *il ^PL™ imcK> ", ;r
-472-
-------
The 133Xe concentrations in blood and urine were measured for gamma radiation with a scintillator
calibrated using an aqueous solution of 133Xe of a known activity. Urine samples were placed in polyethylene
containers each having a volume of 250 ml. Blood samples were taken from the elbow vein into 10 ml glass
syringes when the subjects were in the exposure chamber. This was accomplished when the subject's hand
was stretched out of the transmission hatch of the chamber. The minimum values of 133Xe concentrations that
could be measured in blood and urine were 15 and 2 pCi/ml, respectively. The 85Kr concentration in blood and
urine was measured by its beta radiation with a cylindrical counter of the CTC-6 type. The samples were
injected into the counter through rubber plugs using a syringe. In order to reduce the background radiation,
the counter was placed in a lead housing furnished with a anticoincidence shield.
When the uptake of 85Kr and 133Xe through the skin was being studied, the respiratory organs were isolated
with a helmet and "clean" air was fed into the space under the mask. In these studies the 85Kr and 133Xe
concentrations in the air of the exposure camber were 0.5 to 5 f/Ci/1, while in the experiments related to uptake
by the respiratory organs, concentrations used ranged from 0.05 to 5/^Ci/l.
When the experiments were carried out in a hermetic exposure chamber, comfortable living conditions were
provided. The temperature within the chamber was maintained at 20-23° C with a relative humidity of 48-64
percent and a C02 content not higher than 0.7 percent. The oxygen content used was kept within 24-27 percent
which is slightly higher than ambient concentrations.
RESULTS
1. Experiments with 133Xe.
The accumulation of 133Xe within the human body was found to increase for a period of about 20 hours
exposure; after which period, even with the continuing intake of the gas, its content in the body changes
insignificantly (see Figure 3 and Table 1).
The numerical values which summarize the activity of 133Xe within the body of the subjects as related to its
concentration in the air of the exposure chamber A/Qe, 1, are given in Table 1. The activity of 133Xe uptake was
found to be proportional to the subjects weight and the amount of body fat. Thus, for example, in the case of a
man weighing 100 kg of which weight 28 kg is fat (the subject was M.P.), the value of A/Qe at saturation is 54,
while in the case of a 65 kg subject with 9 kg of fat (the subject was A. Shch.) A/Qe is only 24.
Figure 4 gives typical curves for the accumulation and elimination of 133Xe from the lungs of a subject when
he is in a resting state. The accumulation of 133Xe via.breathing results in an equilibrium state being reached
within 3-5 minutes. After intake of 133Xe ends the activity in the lungs is reduced by 70 percent within 3-5
minutes, after which the rate of elimination decreases. An analysis of the curves of 133Xe accumulation and
elimination from the lungs of nine subjects showed that the kinetics of 133Xe metabolism in the lungs is well
described by the exponential law.
Table 2 shows that the effective 133Xe half-life in the lungs of nine subjects at the time of exposure and after
exposures range from 18 to 42 seconds — with an average value of 30 seconds. Studying the dynamics of 133Xe
uptake in human blood shows that after an exposure of 3 to 40 hours, t concentration of 133Xe in blood remains
practically constant (see Table 3).
This data indicate that the saturation of blood with 133Xe occurs quickly. Due to a large contact surface
with blood, the process of accumulation and elimination of 133Xe proceeds at practically the same speed at
which it exchanges in the lungs. This supposition is confirmed by the fact that no 133Xe activity could be
measured in blood samples taken from a subject 3 to 10 minutes after the end of the exposure.
The half-lives of 133Xe in the lung, blood, muscle, and fat tissues, as well as the corresponding values of A/Qe
at saturation (that is, in the case of an infinite exposure to a 133Xe atmosphere), are given in Table 4. The data
show that the average 133Xe half-lives in the lung and blood are 30 seconds, while in the muscle and fat tissues
they ar,e 0.7 ± 0.42 and 6.2 +1.8 hours, respectively.
The experiment,to study the absorbtion of 133Xe through the skin was carried out on three volunteers. The
exposure time was three hours and the 133Xe concentration in the air of the exposure chamber was 1 /xCi/1.
As one can see from Table 4, the amount of 133Xe absorbed through the skin is not greater than 0.4 percent of
its combined intake by inhalation and through the skin. This small penetration through the skin as compared
to uptake by the lungs may be explained by the difference in the surface area of the skin (1.8 m2) and that of
respiratory organs (70 m2). Furthermore, when explaining the difference in the penetration rate of 133Xe
through the skin as compared to the alveolar epithelium of the lungs, one should take into account the much
greater thickness of the skin barrier. This requires a much longer time to be penetrated than in the case of
diffusion through the alveolar epithelium of the lungs, which causes practically an immediate saturation of
blood with Xenon and other gases. We. were unable to detect any elimination of 133Xe from the body of the
subject through the skin.
2. Experiments with 85Kr.
Our experiments with 85Kr show that human uptake is much quicker for smaller, leaner people than for
heavier, fatter people. For example; the subjects A. Shch and V. G. (who had a normal amount of fat) reached
their saturation level of 85Kr activity in the body within three hours, while the subject E. D., having a higher
amount of fat, reached his saturation level within nine hours (see Figure 5 and Table 6).
-473-
-------
The elimination of 85Kr from the human body is described by the sum of several exponents (see Table 7 and
Figure 6) Due to the fact that the processes of Kr and Xe interaction with the human body are identical, the
same half-life was used for «Kr in lungs and blood (equal to 30 seconds) as was determined experimentally for
i33Xe Analysis of the data given in Table 7 shows that the average values of the 85Kr half-lives m the fat tissue
of the subjects ranged from 1.8 to 3.7 hours. The «Kr half-life in the fat tissue is proportional to the amount of
fat within the body. The rate of 85Kr elimination from the fat tissue decreases after a Prolonged exposure
Thus, the subject A. Shch. exhibited an 85Kr half-life in the fat tissue equal to 0.8 hours with a short exposure of
0 5 to 0 7 hours while at exposure durations equal to 9 to 40 hours the 85Kr half-life increased to from 2.2 to 2.8
hours
Average values for the 85Kr half-life in the muscle and other tissues ranged from 4 to 11 minutes. The
exposure time has little effect on the 85Kr half-life in the muscle and other tissues (see Table 7).
SUMMARY
The main parameters characterizing the accumulation, distribution, and elimination of Kr and Xe from the
human body are shown in Table 8. .
It is evident that the rate of accumulation of the noble gases within the human body exceeds the rate of their
elimination. This phenomenon is especially well traced in experiments with 85Kr, where the accumulation in
the fat tissue of people reaches a saturation point in 4-5 hours, while its elimination takes 8-9 hours. The higher
rate of noble gases accumulation within the body may probably be explained by a dissimilar gradient of inert
gases concentration in the blood and the tissues that are being saturated.
During the intake of the inert gases into the body, the gradient of the concentrations is always higher than
during their elimination. This may be explained by the fact that at the beginning of the contact, the
concentration of the gas in the tissue that is being saturated is close to zero, while in the blood it reaches an
equilibirum state in several minutes due to an intensive gas exchange within the lungs. In the process of
elimination however, after the gas has been accumulated in the saturated tissues, the gradient between the
concentrations is negligible due to the continuous transit of the gas from the tissue to the blood.
REFERENCES
Conn, H. L. (1961), Equilibrium Distribution of Radioxenon in Tissue: Xenon-Hemoglobin Association
Curve.J. Appl. Physiol., 16,1065-1070.
Kalantarov, K. D., Yu. M. Varentsov, and F. M. Liass (1966), Intergral Doses Created by the
Intravenous Injection of133Xe. Meditsinskaya Radiologia, 8,19-24.
Osanov, D. P., I. A. Lichtarev, G. B. Radzievsky (1970), Radiation Dosimetry of Incorporated
Radioactive Substances. M., Atomizdat.
TABLE 1. The Value of A/Qe, 1 (a) of 133Xe in the Body of the Subject Man for Different Durations
of Exposure.
Durations of exposure, hours
The Subject Man
3 9 20 40 66
A. Shch.
N.S.
A.T.
M.P.
10
14
23
14
19
21
21
41
23
24
41
"
24
23
39
54
29
20
..
•"
A = Summary activity of 133Xe in the body.
Qe = Concentration of 133Xe activity in the exposure chamber.
-474-
-------
TABLE 2. Effective Half-Lives (Te) of 133Xe in the Lungs of Nine Male Volunteers, Seconds.
Numbers
Te
At the time of exposure
After exposure
1
2
3
4
5
6
7
8
9
42
18
18
36
30
30
18
48
30
30
42
30
24
24
34
30
30
20
TABLE 3. Factor of 133Xe Distribution in Human Blood, ml/g.
Subject m&n
Duration of exposure, hours
3
A.Shch. 0.17
N.S. 0.17
A.T.
M.P. 0.1
9
0.17
0.19
0.15
0.15
20
0.19
...
...
...
40
0.16
0.18
0.13
0.26
TABLE 4. Parameters Characterizing 133Xe Elimination from the Human Body.
The subject
man
Lungs and blood
(A) i
Qe
A.Shch. 4
N.S. 4
A.T. 4
M.P. 4
T
Seconds
Muscles and
(A) i
Qe
30 6
30 5
30 8
30 4
other tissues
"T
hours
Fat
(A)!
Qe
0.9+ 2 14
0.7+ 0.4 15
0.9+ 0.27 28
0.4+ 0.1 46
tissue
T
hours
6.0+ 0.6
4.6 + 0.45
7.4 ± 2
6.8 +0.9
Average Data
30
0.7 + 0.42
6.2 ± 1.8
Note:
Aj= content of 133Xe activity in the human body at the moment of saturation.
QP=concentration of 133Xe activity in the exposure chamber.
T=half-life.
-475-
-------
TABLE 5.Intake of 133Xe into the Body of the Subject Men Through the Skin After a Three-Hour
Exposure.
The subject
man
A/Qe,l
Iskin and respi-
ratory organs
Per cent of activity
entering
through the skin
M.P.
A.S.
v.v.
0.08
0.04
0.04
23
10
10
0.35
0.4
0.4
Note:
A = 133Xe content in the human body.
Qe = 133Xe concentration in the air.
TABLE 6. The Values of A/Q«,1 for Different Durations of Exposure (T J in an Atmosphere of
85Kr. e
The subject
man
A.Shch.
V.G.
V.P.
A.D.
E.D.
B.V.
0.5
6
5.5
4.6
7.1
7.7
2.5
Te, hours
3
10.4
13.1
7.8
10.8
14.3
9.8
9
10.4
11.5
13.9
13.2
21.2
14.3
20
9.3
11.8
13.6
13.5
18.3
13.3
40
10.2
11.9
15.3
13.4
19.6
13.2
-476-
-------
TABLE 7. Half-Lives of ^Kr (T) in the Body of the Subject Men After Their Exposures and the
Corresponding Values of A/Qe,l.
The
subject
man
Tissue
or
organ
Duration of exposure, hours
0,5-0,7
A
Qe
T
3 | 9
A
Qe
T A
Qe
T
20
A
Qe
T
40
A
Qe
T
Mean values
A
Qe
T
A.Shch.
V.G.
A.D.
V.P.
B.V.
E.D.
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
2.3
0.6
3
2.2
1.4
1.8
3.6
1.5
2
1.6
1.6
1.4
1.0
1.1
0.4
1.8
3,0
2.9
0.8 h
15 min
30s
1.3 h
4.2 min
30s
1.4 h
4.2 min
30s
l.lh
6.3 min
30s
2.2 h
13.8 min
30s
1.7 h
3.6 min
30s
3.7
2.1
4.1
6.3
2.4
4,2
2.8
2.7
5.3
5,4
0.7
1,7
6.9
2.4
0.6
8
2.6
3.7
0.75 h
7.8 min
30s
2.2 h
3 min
30s
2.6 h
6 min
30s
2.2 h
18 min
30s
2.4 h
12 min
30s
2.8 h
9 min
30s
6.4
1.9
2
7.6
1,7
1,0
6
1.6
5.5
9.6
2.5
1,7
10.4
1.6
0.6
16.4
3.3
1.5
2.2 h
8.4 min
30s
2h
3.6 min
30s
3h
10.2 min
30s
3.2 h
22.2 min
30s
3h
7.2 min
30s
4.2 h
3 min
30s
7.4
1.9
—
7.0
1.5
3.3
5.6
2.6
5.3
11.2
1.8
0.56
10.4
1.3
0,1
13.8
1.4
3.1
2.2 h
1.8 min
—
2.4 h
4,2 min
30s
3h
4.8 min
30s
4.5 h
3.6 min
30s
5h
7.2 min
30s
4.5 h
3.6min
30s
5.4
1.8
—
6.3
2.1
3.0
5.6
1.8
6
10.2
1.7
3.4
10.4
2.9
0.1
12
1.4
6.2
2.8 h
18 min
—
2.7 h
3 min
30s
1.9 h
3.6 min
30s
4.3 h
6 min
30s
5.7 h
7.8 min
30s
3.3 h
18 min
30s
6,4 ± 0,7
1,7+ 0.02
3,0+ 0.7
7,0 + 0.5
1.9 .* 0,3
2.7 ± 1.2
5.7 ± 0.7
2.0 ± 0,5
4.8 ± 0.7
10.3 ± 1.7
1.7 ± 0,7
1,8 ± 0,7
10,4 ± 2.0
1.9 ±0.6
0.4 ± 0.2
14 ± 2.6
2.3 ± 2.0
3.5 ± 1.2
1.8 ± 0.8 h
10.2 + 5.0 min
30 + 6.0 s
2.1 ± 0.3 h
3.6 + 0.5 min
30 ± 6.0 s
2.4 ± 0.6 h
6.0 ±2,0 min
30 ± 6.0 s
3.1 ± 1.1 h
10.8 ± 7.0 min
30 ± 6.0 s
3.7 ± 1.4 h
9.6 + 2.6 min
30 ± 6.0 s
3.3 1.0 h
7,2 ± 5.2 min
30 ± 6.0 s
Note:
A = Content of activity within the body; Qe = concentration of actvity of 85Kr in the exposure chamber.
T = Half-life.
1 = Fat tissue; 2 = muscles and other tissues; 3 = lungs and blood.
-------
TABLE 8. Main Parameters Characterizing the Accumulation, Distribution, and Elimination of
the Radionuclides of, Krypton and Xenon from the Human Body.
Organ or
tissue
Krypton radionuclides
R, lk/g| Te* 1 Tb
Xenon radionuclides
R, l/kg| T! 1 Tb
Fat tissue
Muscles (and
other tissues)
Blood
Lungs
0.46
0.047
0.046
2
1.4 h
8 min
30s
30s
2.7 h
8 min
30s
30s
1.4
0.13
0.17
2
5h
0.4 h
30s
30s
6.3 h
0.7 h
30s
30s
Note:
Tb = Biological half-life after the subject man's contact with the gas has ended.
R - Distribution factor.
-478-
-------
-9
6x10
4x1(T9
E.D.
Te=20 hre.
VOLUNTEER IN A AFTER VENTILATION OF
MEASUREMENT A MEASUREMENT
CHAMBER CHAMBER
T, hrs.
Figure 1. Changes in the 85Kr concentration in the air of the measuring chamber.
-479-
-------
C
0 5 10 15 0 5 10 15
Figure 2. Changes in the rate of •»& and i33Xe elimination from the body of one of the volunteers.
te - exposure time in the atmosphere of a radionuclide;
Qe=radionuclide concentration in the air of the exposure chamber- and
A=summary activity of 85Kr and 133Xe in the body.)
-480-
-------
Asat
10 20 30 40 50 60 70
10 20 30 40 50 60 70
1.25
0.75
0.50
0.25
Figure 3. Kinetics of 133Xe accumulation within the human body; 0 calculated from the data on distribution
(Conn, 1961).
-481 -
-------
Te, min.
Figure 4. Kinetics of 133Xe accumulation and elimination from the lungs of a man at rest.
-482-
-------
E.D.
'VP.
A.Shch.
15 20 25 30 35
Te, hrs.
Figure 5. Kinetics of 85Kr accumulation in the human body.
40 45
-483-
-------
TRANSFER OF AIRBORNE KRYPTON-85 TO VEGETATION
P. G. Voilleque and J. J. Fix
Health Services Laboratory
U. S. Atomic Energy Commission
National Reactor Testing Station
Idaho Falls, Idaho 83401
Abstract
Laboratory experiments have been conducted in an environmental chamber to evaluate the transfer to
vegetation of krypton-85 which has been released to the atmosphere. A preliminary evaluation of the acute
exposure case confirmed expectations that the air-to-grass transfer velocity was indeed small. Subsequent
experiments were directed toward the evaluation of air-to-vegetation transfer under "global equilibrium"
conditions in which the krypton-85 concentration remains approximately constant for extended periods of
time. Edible vegetable greens (Swiss chard, lettuce, and turnip greens} and pasture grass were exposed to
krypton-85 for periods up to 75 hours in length. It was found that the concentration of krypton-85 in vegetation
reaches an equilibrium value in a relatively short time, probably within one hour. Transfer velocities
computed assuming a 1-hour equilibration time range from 10-7 to 10-6 cm/sec; the values were not
lognormally distributed. While these transfer velocities exceed the value initially obtained for the acute
exposure case, the common assessment that neither animal forage nor human plant food is part of a
significant exposure pathway for airborne krypton-85, has been confirmed.
INTRODUCTION
Concern about actual and projected local and global atmospheric concentrations of 85Kr and other
radioactive fission gases has not been limited to the potential doses from the direct exposure pathways of
external irradiation and inhalation of contaminated air. The question of possible "biological concentrations"
of airborne 85Kr came to our attention as the result of questions raised regarding the possible radiological
consequences of nuclear gas stimulation projects. This is not to say that the question had not been considered
previously. It had and the common conclusion was that potential doses from indirect exposure pathways were
so small that detailed evaluation was unneccessary. This belief persists today and, to our knowledge,
radioactive noble gases have never been observed to "concentrate" in vegetation or animal components of the
human food chain which have been sampled downwind from atmospheric testing locations, nuclear power
reactors, fuel reprocessing plants, or sites of nuclear gas stimulation projects.
The expectation that airborne concentrations of 85Kr and other noble gases are not significantly reduced by
air-to-vegetation transfer is widely shared by those who employ these gases as tracers to evaluate
atmospheric dispersion and mixing processes. Noble gas tracers are commonly termed "inert", a property
which distinguishes them from the various particulate fluorescent tracers, which are known to interact with
vegetation and other surfaces, and to bias the results of the dispersion experiments in which they were used.
The use of 85Kr as a meteorological tracer has been described (Nickola, et al., 1970); studies of atmospheric
mixing using 37Ar have already been discussed at this Symposium (Wiest,et al., 1973; and Machta, 1973).
Because of our previous experience in the evaluation of air-to-vegetation transfer of airborne contaminants
under field conditions and in the laboratory (Pelletier, et al., 1970 and Voilleque, et al., 1971), we undertook the
experimental evaluation of the air-to-vegetation transfer of 86Kr using one of our environmental chambers.
Our preliminary evaluation of the acute exposure case has been reported (Voilleque, et al., 1970). The present
paper presents the results of more detailed and longer duration experiments subsequently conducted to better
define the air-to-vegetation transfer of 85Kr.
MATERIALS AND METHODS
All vegetation exposed to 86Kr was grown from seed in ~ 1.4 kg of sandy loam, contained in 25- by 25-cm pots
whose depth was 14 cm. The vegetation was maintained in a growth room having a 16-hour photoperiod with
an average light intensity of approximately 19,000 lm/m2. The air temperature in the plant growth room was
regulated to 24°C during the "day" and 13°C during the "night" period. The relative humidity averaged
between 40 and 50%.
The Manchar Bromegrass, which is a common constituent of pasture grass in this area, had been cut back to
a height of 10 cm one or more times, and was grown to a height of about 25 cm prior to exposure. Pots of
(Fordhood Giant) Swiss chard and (Shogoin) turnip greens were both exposed when the plants had developed
to the point when they would normally be cut for human consumption. The (Black Seeded Simpson, White
Boston, and Great Lakes) lettuce was exposed when the plants were 7-13 cm high and could be cut for human
consumption. White Boston and Great Lakes are head lettuce varieties; however, the plants were cut, as they
frequently are by garderners, well before the development of the heads.
All exposures of vegetation to 85Kr were conducted in a plexiglass exposure chamber whose dimensions are
approximately 0.9 m high, 1.2 m wide, and 0.6 m deep. The chamber was maintained in a controlled
environment laboratory and was illuminated by a 400-watt GE Lucalox lamp which produced about 50 000
lm/m2 at plant level. After the vegetation was placed in the chamber, the chamber was sealed and the 85Kr
-484-
-------
was injected. The 85Kr was mixed with chamber air using two 10-cm diameter fans which operated throughout
the exposure period; wind was not simulated during these experiments. The relative humidity in the chamber
increased from about 70% at the start of a typical experiment to 90% or greater by the end of the exposure
period.
Samples of chamber air were drawn into evacuated glass vials which were counted in a calibrated well-type
scintillation counter to determine the 85Kr concentration at several times during the exposure period. The 85Kr
concentrations were also monitored continuously using a Geiger-Mueller tube with a ratemeter and strip chart
recorder. The exposed vegetation was cut quickly after the pot was removed from the chamber, and either
double bagged in polyethylene or placed in a glass jar which was tightly sealed with a screw-on top; the 85Kr
activity was then determined using a well counter calibrated for both sample configurations. It was found
that 85Kr was lost from the double polyethylene bags at a rate of about 55 percent per day and their use was
discontinued; leakage from the sealed jars was not detectable in a 3-day period. Loss of 85Kr from the samples
was further limited by the normal practice of counting all samples within 4 hours following collection.
Samples of unexposed vegetation of the same variety and stage of development were also collected and
counted. All vegetation samples were subsequently oven dried to a constant weight, and the dried vegetation
was weighed.
EXPERIMENTAL RESULTS AND DISCUSSION
A standard method of expressing results of experiments of this type employs a parameter usually called the
"deposition velocity"; the term "transfer velocity" is perhaps more appropriate for gases and will be used in
this discussion. The transfer velocity is considered to be an approximately constant parameter which
describes the air-to-surface transfer of airborne contaminants. It is the ratio of the rate of transfer of the
contaminant per unit surface area to the instantaneous air concentration at a standard height above the
surface. The ratio has unit length/time; hence the designation of "velocity". In practice, the ratio is obtained
by measuring the total quantity transferred per unit area and dividing that by the measured time-integrated
air concentration at the specified height. In Equation (1), Vd (cm/sec) is the transfer velocity, Cv ( /i Ci/cm2)
is the activity transferred to the vegetation on a unit horizontal area, X (/i Ci/cm3) is the average air
concentration and r (sec) is the exposure time. The product X— T is termed the time-integrated air
concentration OP, //Ci-sec/cm3).
(1)
The mass of vegetation exposed to the contaminant has been found to be important for some transfer
processes (Pelletier, et al., 1970); so a second parameter, the normalized transfer velocity (VD, cm3/g-sec), is
often computed and used in the same context. In Equation (2), D is the areal vegetation density (g/cm2),
computed using the dry weight of the vegetation.
(2)
In this model, the transfer of radioactivity from air to vegetation is presumed to be a process which
continues throughout the exposure period. The data in Tables 2 and 3 illustrate that this is not true for 85Kr.
Vd decreases with increasing ^ for both Bromegrass and lettuce, implying that the 8SKr concentration in
vegetation reaches an equilibrium state, and that the 85Kr activity found in vegetation will be related to the
value of X near the end of the exposure period. On the assumption that the same processes operate to increase
or decrease Cv as X changes correspondingly, knowledge of the retention half-time permits an estimate of
the equilibration period. The retention half-time for 85Kr was estimated by measuring Cv for exposed pots of
vegetation removed to the normal laboratory environment for periods up to one hour after exposure. On the
basis of limited measurements, the retention half-time is estimated to be 5-15 minutes. It is possible that (1)
shorter lived components would be observed if more rapid sampling or "whole plant counting" techniques
were employed, and (2) longer lived components would be noted if longer evaluation times were achieved by
increasing the initial value of C v. Because it appears that Cy will reach > 90% of the equilibrium value within
one hour after exposure to a new concentration of 85Kr, the transfer velocities reported below have been
computed using a 1-hour effective exposure time and the X measured for the last hour before removal of the
vegetation from the contaminated air has been employed. The effective time-integrated air
concentration, % , is then 3,600 X /zCi)sec/cm3. This approach to reporting the results was selected because
of the utility and widespread use of transfer velocities in the calculation of doses resulting from airborne
radioactivity releases. The alternative of evaluating partition coefficients for the wet mass of vegetation or its
water content is of some interest, but is not directly applicable to conventional hazard analysis techniques
-485-
-------
and, for that reason, was not selected. Such an analysis would be further complicated by unmeasured
variations in lipid content within the plants and in their waxy surface coatings.
Table 1 summarizes the experimental data obtained for the exposure of turnip greens and Swiss chard to
85Kr. Both the actual (
-------
Type C
TABLE 2. Data on Transfer of 8BKr to Lettuce.
e(ptCi-sec/cma)(a) Vd(cm/sec)(a) D(g/cm2)(b) Vi)(cma/g-sec)(a) ( uCi-sec/cm3)
GL
GL
GL
GL
WB
WB
WB
WB
BSS
BSS
BSS
BSS
BSS
BSS
BSS
BSS
1.9 ± .3xlO-4
8.4 ±1.5x10-5
4.0 ± JxlO-5
6.3 ±1.1x10-5
1.4 ± .2x10-"
7.5 ±1.2x10-5
8.2 ±1.3xlO-5
4.6 ± .7x10-5
2.4 ± .1x10-*
1.8 t .IxlO-4
1.9 ± .1x10-*
1.3 ± .1x10-"
2.1 ± .5x10-"
8.7 ±2.0x10-5
2.4 ± .5x10-"
1.8 ± .4x10-"
77.0 ± .7
77.0 ± .1
77.0 ± .7
77.0 ± .7
54.7 ± .5
54.7 ± .5
54.7 ± .5
54.7 ± .5
84.2 ± 1.7
59.4 ±1.2
43.2 ± .9
38.9 ± .8
33.8 ± 6.8
33.8 ± 6.8
53.6 ± 5.1
53.6 ±5.1
2.5 ± .4xlO-6
1.1 ± .2x10-6
6.2 ± .9x10-'
8.2 ± 1.4x10-'
2.6 ± .4x10-6
1.4 ± .2xlO-«
1.5 ± .2x10-6
8.4 ± 1.3x10-'
2.9 ± .IxlO-6
3.0 ± .2xlO-6
4.4 ± .2x10-6
3.3 ± .SxlO-6
6.2 ± 1.9x10-6
2.6 ± .SxlO-6
4.5 ± l.OxlO-6
3.4 ± .SxlO-6
2.6 ± .IxlO-2
3.3 ± .IxlO-2
2.8 ± .IxlO-2
2.4 ± .IxlO-2
2.9 ±.lxlO-2
3.0 ± .IxlO-2
3.0 ± .IxlO-2
3.1 ± .IxlO-2
4.6 ± .4xlO-3
6.1 ± .4xlO-3
4.5 ± .4xlO-3
4.0 ± .4xlO-3
7.5 ± .4xlO-3
8.2 ± .4xlO-3
8.2 ± .4xlO-3
5.8 ± .4xlO-3
9.6 ± 1.6xlO-5
3.3 ± .6x10-5
1.9 ± .SxlO-5
3.4 ± .6x10-5
9.0 ±1.4x10-5
4.7 ± .7zlO-5
5.0 ± .7x10-5
2.7 ± .4x10-5
6.3 ± .6xlO-4
4.9 ± .5x10-"
9.8 ± 1.0x10-"
8.3 ± 1.1x10-"
8.3 ± 2.6x10-"
3.2 ±1.0x10-"
5.5 ± 1.2x10-"
5.9 ±1.4x10-"
115.6 ± 1.1
115.6 ±1.1
115.6 ±1.1
115.6 ±1.1
136.8 ±1.3
136.8 ±.13
136.8 ± 1.3
136.8 ±1.3
84.2 ±1,7
143.6 ±2.1
186.8 ± 2,3
225.7 ±2.4
7.5 ±1.5 xlO3
7.5 ±1.5 xlO3
1.48±.14xl03
1.48±.14xl03
(a) Te is the effective time-integrated air concentration computed assuming a 1-hour equilibration time.
Both Vj and Vp were computed using T?e. The actual value of™ is given in the last column.
(b) Based on the dry weight of vegetation determined to;t 0.2 grams. The area of each pot was 625;t 25 cm2.
TABLE 3. Data on Transfer of 85Kr to Bromegrass.
e(u.Ci-sec/cm3)(a) Vd(cm/8ec)(a) D(g/cm2)(b) VD(cmVg-secMa)
(a) *Pe is the effective time-integrated air concentration computed assuming a 1-hour equilibration time.
Both Vd and VD computed using »|f. The actual value of Vis given in the last column.
(b) Based on the dry weight of vegetation determined to ± 0.2 grams. The area of each pot was 625 ± 25 cm2
(jiCi-sec/cm3)
6.7± 1.2xlO-5
4.2 ± 1.6xlO-5
4.5 ±1.9x10-5
3.5 x 10-5
7.4 ± l.SxIO-5
5.0 ± l.OxlO-5
3.1 x lO-5
3.2 ±1.1x10-5
6.9 ± 1.8x10-5
8.6 ± 1.9xlO-5
2.0 ± .2x10-"
1.8 ± .2x10-"
73.1 ± 2.2
61.6± 1.8
54.4 ± 1.6
41.0 ±1.3
78.8 ± 2.4
61.6 ±1.9
54.0 ±1.7
46.8 ± 1.4
33.8 ±0.9
33.8 ± .9
53.6 ± 1.5
53.6 ±1.5
9.2±1.6xlO-7
6.8 ±2.6x10-'
8.3 ±3.5x10-'
8.5 x 10-'
9.4 ± 1.7x10-'
8.1 ±1.6x10-'
5.7 x 10-'
6.8 ± 2.3x10-'
2.0 ± .5xlO-6
2.5+: .6x10-6
3.7 ± .4x10-6
3.4 ± .4xlO-6
7.0 ± A xlO-3
9.8 ±.5 xlO-3
1.14±.06xlO-2
1.04 ±.05x10-2
7.8 ±.5 xlO-3
5.9 ±.4 xlO-3
9.3 ±.5 xlO-3
6.7 ±.4 xlO-3
1.46±. 07x10-2
1.47 ±.07x10-2
1.18±.06xlO-2
1.04 ±.05x10-2
1.3 ± .2x10-"
6.9 ±2.7x10-5
7.3 ±3.1xlO-5
8.2x10-5
1.2 ± .2x10-"
1.4 X .3xlO-4
6.1 x 10-5
1.0 ± .3x10-"
1.4 i. .3x10-"
1.7 ± .4x10-"
3.1 ± .4x10-"
3.3 ± .4x10-"
73.1 ± 2.2
134.7 ± 2.8
189.1 ± 3.3
230.1 ± 3.5
78.8 ± 2.4
140.4 -t 3.1
194.4 ± 3.5
241.2 ± 3.8
7.5 ± 1.5 xlO3
7.5 ± 1.5 xlO3
1.48 ± .14xl03
1.48 ± .14X103
-487-
-------
POSSIBLE EFFECTS OF NOBLE GAS EFFLUENTS FROM POWER REACTORS AND FUEL
REPROCESSING PLANTS*
G.H.Whipple
School of Public Health
University of Michigan
Ann Arbor, Michigan 48104
Abstract
The noble radioactive gas of principal interest in the nuclear industry is krypton-85 because of its relatively
large fission yield and its long half-life. This paper considers the various ways in which krypton-85 can deliver
radiation doses to the human body; external irradiation by beta particles and gamma rays, and internal
irradiation from the gas inside the respiratory tract and dissolved in body tissues. Equilibrium and transient
conditions are discussed. The calculations summarized in this paper indicate that the present maximum
permissible concentration of'krypton-85 in public air(3 x 107pCi/'ml) delivers a dose rate of about l/5ofthat
recommended by the ICRP. The long-term significance of krypton-85 and other noble radioactive gases from
then uc Icar power industry is reviewed.
INTRODUCTION
The noble radioactive gas of principal interest in the nuclear industry is krypton-85. This is so because of its
relatively large fission yield (0.3%) and its half-life (10.8 years). The holdup delay at power reactors, and the
delay between the removal of reactor fuel and its dissolution at the reprocessing plant eliminates, largely or
completely, the other noble radioactive gases as contributors to off-site doses. For these reasons this paper is
confined to consideration of the various ways in which krypton-85 can deliver radiation doses to the human
body.
THE RADIOACTIVE DISINTEGRATION OF KRYPTON-85
Krypton-85 decays to rubidium-85, which is stable. The decay results in a beta particle with a maximum
energy of 0.67 MeV, and no subsequent gamma ray in 99.6% of the disintegrations. The remaining 0.4% of the
disintegrations give rise to a beta particle with a maximum energy of 0.16 MeV, followed by a gamma ray with
an energy of 0.514 MeV. Thus, a gamma ray occurs only about 4 times for each 1,000 disintegrations, and the
0.67 MeV beta particle is the dominant radiation.
DOSE TO THE BODY FROM KRYPTON-85
The body may be exposed to the radiations of krypton-85 in four possible circumstances: (1) a cloud of the
gas at some distance from the body; (2) a cloud of the gas completely enveloping the body; (3) the gas inside the
respiratory tract; and (4) the gas dissolved in body tissues. These, I believe, are the only ways in which
exposure to any appreciable degree is possible.
A CLOUD OF KRYPTON-85 AT SOME DISTANCE FROM THE BODY
This is the situation where a cloud or plume of gas is passing overhead. Since none of the beta particles
emitted by krypton-85 can travel as far as 2 meters in air, there is no beta exposure from the overhead cloud;
the only exposure is that from the infrequent gamma rays.
The dose rate from the overhead cloud is a function of the concentration of the krypton-85 in the cloud, the
volume of the cloud, and its height above the ground. Just to get a feel for the numbers, the dose rate from a
spherical cloud with a radius of 10 meters and with its center 100 meters above the ground, for a concentration
of 1 x 10-6 fiCi/ml, is about 5 x 10-7 milli-rad per hour. It will shortly become evident that this form of exposure
is very small compared to the other three conditions.
*This paper is a summary of a study commissioned by the General Electric Company for the Midwest Fuel
Recovery Plant in Illinois. The author takes pleasure in acknowledging this support and the permission to
prepare this summary without condition.
-488-
-------
A CLOUD OF KRYPTON-85 COMPLETELY ENVELOPING THE BODY
In an infinite cloud of krypton-85, or any other radioactive gas, the energy abosrbed per unit volume is equal
to the energy emitted per unit volume. In such a cloud, a very small organism (say a mosquito) will receive a
radiation dose at the rate of
Dm=470Crad/hour (1)
where
DUI= dose rate to the mosquito, rad/hour
C = concentration of 85Kr, flCi/ml
Of this dose, 99% is delivered by the 0.67 MeV beta particles, which have a maximum range in tissue of 2.5 mm.
Thus, the surface of a piece of tissue as small as a finger is irradiated only from the outer side. The mosquito is
so small that the beta radiation passes through it and irradiates the surface from both the outer and inner
sides almost equally. It follows that the dose rate to the skin of a person in an infinite cloud is just about half
that to the mosquito.
D s=Dm/2 = 240 C rad/hour (2)
The infrequent 0.514 MeV gamma ray and the bresstrahlung produced when the beta particles are stopped
in tissue irradiate the entire body fairly uniformly. The whole body dose rate from the infinite cloud is
approximately
Dw= 2 C rad/hour (3)
The dose to the sperm-producing cells of a man standing naked in this cloud is made up of the whole-body
dose, given by eq. (3), and that portion of the skin dose, given by eq. (2), which penetrates the scrotum. The
depth of the sperm producing cells below the surface of the skin has not, to my knowledge, been specified by
the Standard Man. I am indebted to my associate Dr. A. P. Jacobson, who conferred intimately with several
cadavers at the University Hospital, for the estimate that this depth is about 2.5 mm.
As has been stated previously, the maximum range of krypton-85 beta particles in tissue is 2.5 mm, at which
depth, of course, the beta dose rate drops to zero. At a depth of 1.5 mm in tissue, the beta dose rate is just about
1/100 the surface dose rate. It seems unlikely, therefore, that krypton-85 beta particles contribute as much as
1% of the surface dose to sperm-producing cells. It is fashionable to be conservative (i.e. wrong, but on
the high side) about these things, so I estimate that the genetically significant dose to a naked male in an
infinite cloud of krypton-85 will not exceed
Dt
-------
When the body has been exposed to a cloud of krypton-85 at a concentration of CfrjCi/ml)long enough for
equilibrium to be established, the relation between the concentration in tissue to that in the cloud is simply
G = PC (8)
where
G ± concentration in tissue,fiCi/g
C = concentration in air, ,uCi/ml
P = partition coefficient
P-0.5 for fat
P = 0,07 for all other tissues
thus, for fat, the dose rate is
Df=0.25 C rad/hour (9)
and for all other tissues is
D0 = 0.035 C rad/hour (10)
SUMMARY OF EQUILIBRIUM DOSE RATES
Comparison of the dose rates given by the equations which have been developed to the maximum organ
dose rates recommended for the public at large by the ICRP (1966) shows that the concentration of krypton-85
in air is limited equally by the dose rates to the skin and to the lung. The concentration which produces both
these limiting dose rates is 1.5 x 10-6 jiCi/ml, which is 5 times the present maximum concentration in public
air (ICRP, 1966; and 10CFR). Dunster and Warner have noted this factor of 5 and suggested that the ICRP
may increase the MPC accordingly (Dunster, e£ al, 1970).
DOSE RATES UNDER NON-EQUILIBRIUM CONDITIONS
The first three exposure conditions (a cloud at some distance from the body, an enveloping cloud, and inside
the respiratory tract) entail no delays for uptake and elimination. As a consequence, the dose rates from these
exposure conditions follow the concentration of krypton-85 in air instantaneously and the dose is directly
proportional to the period of exposure. Only in the case of krypton-85 dissolved in body tissue does the
duration of exposure complicate the estimation of dose.
The exchange between krypton in the lung and that dissolved in blood is rapid for both uptake and
elimination. Half-lives for these exchanges have been reported as being from 1 to 6 minutes (Whitton, 1968).
The rates of exchange from blood to other tissues, and from these tissues back into blood, are considerably
slower than the lung to blood rate. Half-lives for the transfer of krypton-85 from inhaled air to body fat range
from 0.4 to 3 hours (Hytten, et al., 1966; and Turkin, et al., 199). The half-life for elimination of krypton-85 from
fat appears to be considerably longer than that for uptake; i.e., 18 hours (Hytten, et al., 1966).
If uptake and elimination occur at the same rates, the dose from krypton-85 dissolved in body tissues is
simply the dose rate from the amount dissolved at equilibrium multiplied by the period of exposure, e.g.,
Dft=Dft (11)
However, when elimination proceeds more slowly than uptake, the dose for short exposure periods is
somewhat greater than is the case when elimination and uptake occur at equal rates.
Exposure periods of less than 100 hours do not reach equilibrium conditions in fat and, therefore, result in
somewhat higher doses than would be the case if elimination took place at the same rate as uptake. In spite of
this, the skin and the lung are still the limiting tissues, just as they are for equilibrium exposure conditions. In
other words, the rapid uptake and slow elimination of krypton-85 in body fat do not change the conclusions
based on steady state conditions.
LONG-TERM SIGNIFICANCE
The relations which have been developed between krypton-85 concentration in air and the dose rates to
various organs and tissues have been applied to the emission rates expected from the General Electric
Companys; Midwest Fuel Recovery Plant, in Illinois. The results show that the maximum off-site doses are
about 9 milli-rem per year to the skin (the ICRP limit is 3,000 milli-remper year), and 4.4 milli-rem per year to
the lung (the ICRP limit is 1,500 milli-rem per year). Under the most adverse meteorological conditions, the
dose to the skin might be as great as 8 milli-rem to the skin and 4 milli-rem to the lung if these conditions
persisted for 15 hours.
A number of estimates have been made of the doses from krypton-85 released to the atmosphere as the result
of nuclear power generation. The greatest uncertainty in these estimates is, quite naturally the choice of the
rate at which nuclear power will grow. Coleman.ef al., (1966) and Dunster, etal.,(W7Q) present a number of
such estimates. These estimates differ in no significant way from the dose rates obtained by applying the
relations in this paper to projected nuclear power growth. From all of these estimates it appears unlikely that
the skin dose, world-wide, will exceed 10 milli-rem per year until after the year 2010 AD (34 million megawatts
electric) if all the krypton-85 is released to the atmosphere.
-490-
-------
The biological effects of 10 milli-rem per year to the skin of the world's population are, in my opinion, of little
if any significance. Such effects, if they occur at all, will be swamped by the effects of natural, occupational,
and domestic agents to the extent that their identification lies beyond the abilities of the most sophisticated
epidemiologists and statisticians.
SUMMARY AND CONCLUSION
Examination of the possible ways a person may be exposed to airborne krypton-85 leads to the following
conclusions, which apply under both equilibrium and transient conditions: exposure of skin and lung are
equally limiting] and the appropriate maximum permissible concentration in public air is 1.5 x 10-6 /iCi/ml,
which is 5 times the presently accepted concentration.
Application of the dose-rate relations developed in this paper to the projected growth of nuclear power, in
which all the krypton-85 is released to the atmosphere, leads to the estimate that the skin dose is unlikely to
exceed 10 milli-rem per year, world-wide, before 2010 AD. The biological effects of this dose, if any, will be
trivial and unmeasurable.
REFERENCES
Coleman, J.R. and R. Liberace, (1966), Nuclear Power Production and Estimated Krypton-85 Levels.
Radiol. Health Data and Reports, 7:615-621.
Dunster, H.J. and B.F. Warner, (1970), The Disposal of Noble Gas Fission Products from the
Reprocessing of Nuclear Fuel. AHSB (RP) R-101, Harwell, England.
Hodgman, C.D., (1951), ed., Handbook of Chemistry and Physics, 33rd ed. Chemical Rubber Publishing
Co., Cleveland, p. 1482.
Hytten, F.E., K. Taylor and N. Taggart, (1966), Measurement of Total Body FatinMan by Absorption
ofssKr. Clin. Sci. 31:111-119.
Kety, S.S., (1951), The Theory and Applications for the Exchange of Inert Gas at the Lungs and Tissues.
Pharm. Rev. 3:1.
Lawrence, H.H., et a/., (1946), Preliminary Observations on the Narcotic Effects of Xenon with a Review
of Values for Solubility of Gases in Water and Oils. 3. Physiol. 105:197.
Mellemgaard, K., N.A. Lassen and J. Georg, (1962), Right-to-Left Shunts in Normal Man Determined
by the Use of Tritium and Krypton-85. J. Appl. Physical. 17:778.
Muehlbaecher, C.A., F.L. DeBon, and R.M. Featherstone, (1966), Further Studies of Solubilities of
Xenon and Cyclopropane in Blood and Protein Solutions. Molec. Parmacol. 1:86-89.
lurking, A.D., et al., (1969), Radiation Hazards from Radioactive Inert Gases. Symposium on Noble Gas
Fission Products in the Nuclear Industry, Paris Oct. 1969, CONF-691039.
U.S. Atomic Energy Commission, Rules and Regulations, 10 CFR Part 20, Appendix B, Table II,
Column 1.
Whitton, J.T., (1968), Dose Arising from Inhalation of Noble Gases, Central Electricity Generating
Board, Report RD/B/N/1274. Berkeley, England.
Yeh, S.Y. and R.E. Peterson, (1963), Solubility of Carbon Dioxide, Krypton and Xenon in Lipids. J.
Pharm. Sci. 52:453-458.
Yeh, S.Y. and R.E. Peterson, (1965), Solubility of Krypton and Xenon in Blood, Protein Solutions and
Tissue Homogenates, J. Appl. Physiol. 20:1041-1047.
-491-
-------
RELATION BETWEEN CUMULATIVE EXPOSURE TO RADON-DAUGHTERS, LUNG DOSE,
AND LUNG CANCER RISK
W. Jacob!
Institute for Radiation Protection
Munich, Germany
Abstract
The dose/'exposure-ratios regarding the influence of uncombined radon-daughters for the bronchial region
resulting from different dosimetric models are compared. For dose evaluation, the uncombined fraction of the
total potential a -energy in the inhaled air seems to be a more suitable parameter than the uncombined
activity-fraction of RaA only. The mean dose to the basal cell layer in the medium bronchi is probably not
considerably higher than the mean dose to the total bronchial tree. Under normal conditions in mines, a ratio
of 0.4 to 1.0 rad/WLM can be expected.
In the second part of the paper, the observed lung cancer-risks among uranium miners in the USA and the
CSSR are discussed. Although the relative risk coefficient is similar for both groups, the absolute risk
coefficient in the CSSR-group is significantly higher than in the USA-group of miners. On the basis of these
data, an absolute lung cancer-risk of 10 to 20 cases/10s person-years per WLMis recommended for radiation
protection purposes. Finally, the risk concept for the assessment of an exposure limit for uranium miners is
discussed. The derived life time limit of 100 to 200 WLMfor occupational exposure is comparable with the limit
derived from the 15 rem/'year-concept.
INTRODUCTION
An exposure limit for the inhalation of short-lived radon-daughters can be derived in two different ways:
(1) From the dose limit of 15 rem per year in the critical respiratory region, as it is recommended by the ICRP
(dose concept).
(2) From the observed relationship between lung cancer risk and the cumulative radon exposure among
uranium miners (risk concept).
In this paper some of the main problems with which we are confronted in the application of these concepts are
outlined and discussed. They concern the relation between inhaled activity and dose in different respiratory
regions, and between cumulative exposure and lung cancer-risk.
RELATION BETWEEN LUNG DOSE AND EXPOSURE
It is well known that the inhalation of radon-daughters results in an inhomogeneous dose and activity
distribution in the respiratory tract — especially in the bronchial region. The first problem in the application
of the dose concept is the suitable choice of the critical respiratory region to which the dose limit shall be
referred. For other nuclides, the total lung or the total bronchial region is used as the reference organ. On the
other hand, bronchogenic lung cancer among uranium miners is presumed to originate in the basal cells
which line the basement membrane of the bronchial epithelium. With respect to lung cancer-risk, therefore,
the mean dose to this basal cell layer seems to be the most reasonable reference quantity. In the following, the
dose/exposure-ratios for these different reference regions are evaluated and compared.
1. Mean Dose in the Total Bronchial Region.
This dose can be calculated with sufficient accuracy on the basis of the new ICRP model for the deposition
and clearance of soluble aerosols of clearance class D in the human lung (ICRP 1969; Jacobi, 1972a; and
Jacobi; 1973). It should be emphasized that in the case of short-lived radon-daughters, the choice of the
clearance class is rather uncritical for lung dose evaluation. Figure 1 shows the resulting ratio of the absorbed
a -energy W in the tracheobronchial (T-B) or pulmonary (P) region to the inhaled potential Of-energy P of the
radon-daughter mixture. The curves in this figure refer to nose breathing and an AM AD of the vector aerosol
of 0.2 fim.
In the left graph of Figure 1, this absorption yield is given as a function of the fraction f1 of uncombined or
free RaA atoms for different values of the air ventilation rate Av in tne working area. This graph indicates
that the relation between dose and the uncombined fraction of RaA activity depends on the ventilation rate.
However, if this absorption yield is plotted as a function of the uncombined fraction fp of the total potential -
energy of the radon-daughter mixture, the curves for different ventilation rates converge into a straight line
(see right graph in Figure 1). This leads to the conclusion, that the uncombined fraction f D of the total potential
energy is a more reasonable parameter than the uncombined fraction f ^ of RaA atoms.
From the absorption yield in Figure 1, the mean a-dose in the total bronchial and pulmonary region per
WLM of cumulative exposure can be evaluated, taking into account a breathing rate of 20 1/min and a mass of
45 g or 955 g for the tracheobronchial region or the pulmonary region, respectively, as they are recommended
- 492 -
-------
by the ICRP-task group for the "Reference Man" (ICRP, 1973). The resulting dose/exposure-ratio for radon-
and thoron-daughters is shown in Figure 2 as a function of the uncombined fraction fD of the total potential a -
energy in the inhaled air. v
As expected, the rad/WLM-ratio in the bronchial region increases linearly with the uncombined fraction fp;
for radon-daughters this ratio can be estimated from the relation (E = cumulative exposure):
DT.B(rad)/E(WLM)~0.31 (1+6 fp)
DP(rad)/E(WLM)~0.16 (1 - fp)
As the uncombined fraction fp of the total potential a -energy in air is considerably lower than the
uncombined fraction f iof RaA atoms found in most working areas with unfiltered air is fp < 0.1 — especially
in mines; under normal conditions, a ratio of 0.4 to 0.5 rad/WLM or 4 to 5 rem/WLM (QF =10) follows from the
new ICRP lung model for the mean bronchial a -dose.
However, it should be kept in mind that the uncombined fraction increases with an increasing ventilation
rate due to the reduction of the aerosol concentration and the decreasing RaB/RaA- and RaC/RaA-ratio in the
air. This effect can be estimated on the basis of a box model which takes into account the removal of dust and
activity by ventilation and wall deposition (Jacobi, 1972b). Figure 3 shows roughly the relationship between
the rad/WLM-ratio and the rate of ventilation with clean air which should be expected in working areas with a
high, intermediate, and low rate of aerosol production.
The increase of the rad/WLM-ratio with ventilation is of practical importance because it cancels out, to some
extent, the dose reduction due to the decreasing concentration of radon-daughters in the air. Figure 4 shows
this influence of ventilation in a relative scale. Due to the increasing uncombined fraction, the reduction of the
bronchial dose with ventilation is not proportional to the reduction of the potential energy concentration in the
air. It follows from this figure that ventilation rates of more than about 10 to 20 hours seem to be
unreasonable, because the bronchial dose remains nearly constant in this ventilation range.
2. Dose to the Basal Cells in the Bronchi.
So far only the mean a-dose in the total tracheobronchial region has been considered; however, a large
fraction of the absorbed a-energy in the T-B region is deposited in the mucus sheet and the rather insensitive
layer of ciliated cells and goblet cells on the bronchial epithelium.
Several dosimetric models have been developed to estimate the dose from inhaled radon-daughters to the
basal cells, where bronchogenic lung cancer is assumed to originate (Altshuler, et al., 1964; Jacobi, 1964;
Hague and Collinson, 1967; and Harley and Pasternack, 1972). The assumptions in these models, concerning
the deposition pattern of the inhaled radon-daughters, the mucus flow-rate, and the position of the basal cells,
are in part quite different due to our limited knowledge of these quantities in the different bronchial
generations.
Nevertheless, these studies agree in the conclusion that the maximum a-dose in the basal cell layer is
reached in the segmental-subsegmental bronchi (lung model of Landahl) or in the 4.0 to 9.0 airway generation
(lung model of Weibel), respectively. The calculated absolute dose values are, however, different due to the
different assumptions used in these models. The critical parameters in these models are (1) the depth of the
basal cell layer below the mucus surface and (2) the uptake and distribution of radon-daughters in the mucus
sheet, and the upper cell layers of the bronchial epithelium. The mean depth of the basal cells ranges probably
from about 50 to 100 f*m in the trachea and the main bronchi, 30 to 60 fj.m in the segmental-subsegmental
bronchi, and 20 to 40 £tm in the terminal bronchi.
In Figure 5 the variation range of the a-dose per WLM at a 30 to 60 /zm depth in the segmental-
subsegmental bronchi is given as it can be derived from the different models. The calculated values are plotted
as a function of the uncombined fraction fp of the total potential a-energy in the inhaled air, and are
normalized for nose breathing at a rate of 201/min. The dose/exposure-ratios from the different models cover
a range from about 0.2 to 10 rad/WLM. The more recent models (Harley and Pasternack, 1972) indicate that
under normal mining conditions (fp < 0.1), the mean a-dose in the basal cell layer of this critical bronchial
region lies probably at the lower end of this range. A value of 0.5 to 1 rad/WLM = 5 to 10 rem/WLM seems to be
a reasonable reference value for the conversion of exposure to the mean dose in the basal cell layer of the
segmental-subsegmental bronchi. In the BEIR-report (1972) a conversion factor of 5 rem/WLM has been used.
A comparison with the previously derived conversion factor for the dose to the total bronchial region leads to
the conclusion that the mean a-dose to the basal cell layer in the most exposed bronchial region is probably
not considerably higher than the mean a-dose to the total bronchial tree.
This conclusion solves one of the main problems in the application of the ICRP dose limit of 15 rem/year.
Taking into account a conversion factor of 0.4 to 1.0 rad/WLM, which range covers the results for both
reference regions, this dose limit yields an exposure limit of 2 to 5 WLM per year, or an integral exposure of
about 80 to 200 WLM during a working period of 40 years. The exposure limit of 0.3 WL or 4 WLM per year,
which has been adopted for uranium miners in USA and other countries, falls in this range derived from the
dose concept. However, as pointed out, this limit should be reduced in highly ventilated, dust-free working
areas.
-493-
-------
RELATION BETWEEN LUNG CANCER RISK AND CUMULATIVE RADON EXPOSURE
The annual dose limit of 15 rem recommended for the lung by the ICRP stems from radiation protection
co™epnriortotS
seriously questioned! Therefore, this limit should be revised, if reliable data on the nsk of lung cancer (LC) by
inhalation of radon-daughters become available. , ,.
InrTce^
now to Sss th^relationship between LC-risk and cumulative radon exposure. The first study group of
Stout 3 400 Xte uranfum miners in the Colorado-Plateau (USA), which was analyzed, mainly by Lundinef
al (1971) In this Soup, a total number of about 38,000 person-years of risk was reached during the
observation period from 1951 to 1971. The second group was composed of uranium miners m the CSSR, for
which group preliminary data about LC-risk versus the cumulative radon exposure were published r^ently by
SEVC and PLACEK (1973). However, although the USA-data was published, data for the CSSR-group (the
total number of miners and the absolute numbers of LC-cases) has not been published. Only the relative
number of observed and expected LC-cases, normalized to 1,000 miners, are available, arranged m four
exposure intervals: < 100,100 to 200, 200 to 400, > 400 WLM This arrangement aIready indicates that the
average cumulative exposure of the uranium miners in the CSSR was considerably lower than m the USA-
groS of uranium miners. However, the observation period from 1948 to 1970, and the total number of 34511
Srson-years at risk for the CSSR-group, are comparable with the USA-group. This supports the assumption
that the number of uranium miners in the CSSR-study group is comparable with that m the USA-study group
Fora"direct comparison of both groups, the ratio of the observed to the expected LC-mortality can bequoted,
which is shown in Figure 6 as a function of the cumulative exposure in WLM-units. The corresponding excess
LC-mortality is given in Figure 7 for the USA-group and in Figure 8 for the CSSR-group of uranium miners.
Note that in the USA-group (Figure 7) the excess LC-mortality is related to person-years, whereas m thei CbbK-
group (Figure 8) it is related to the number of persons. The marked observationpomtsm the Figures 6 through
8 are plotted with their 95% confidence limit at the average of each exposure interval.
From the preliminary data for the CSSR-miners a significant dependence between excess LC-mortality and
exposure can be stated already above 100 WLM, whereas for the USA-group of uranium miners in the
observation period up to 1968 such a significant relationship could only be confirmed above 600 to 800 WLM
(see Figure 6). However, the additional LC-cases which occurred in the USA-study group of miners m the last
three years from 1968 to 1971 have shifted this significance limit to lower exposure values (see Figure 7,), taken
from the BEIR-report (1972). f * , ^ t A *
Taking into account the limits of error, the presently available data for both groups, therefore, dp not
indicate the existence of a practical exposure threshold above 100 to 200 WLM. The data also do not allow a
reliable, quantitative distinction between a linear or nonlinear relationship between LC-incidence and
cumulative radon exposure. The mean regression lines in Figures 7 and 8 indicate that a linear relationship in
the considered exposure range cannot be excluded. At the present time, the recommendation of an exposure
limit should be based, therefore, on the assumption of a linear exposure-risk curve.
It follows from Figure 6, that the mean, relative risk coefficient is nearly equal for both groups of uranium
miners and results to Orel ^0.009 + 0.003 WLM-1, corresponding with a mean "doubling exposure" of about 110
WLM. For comparison, it should be noted that the mean natural background level of radon-daughters in air
leads to a cumulative exposure of about 4 to 5 WLM during a lifetime of 70 years. This means that only a few
percent of the observed LC-mortality in our population can be attributed to irradiation of the lung by inhaled
radon-daughters, if a linear extrapolation down to these low exposure values in generally accepted.
The absolute risk coefficient for the induction of lung cancer from inhaled radon-daughters can be derived
from the slope of the mean regression lines given in Figures 7 and 8. For the USA-group of white uranium
miners a mean value results as follows:
/T Tf"1 ATT * \ __ O > -1 l_JV/~CclSCS
q , (USA-U-mmers) = 3 ± 1 ___________
aos WLM x 106 person-years
Taking into account a conversion factor of 1 WLM = 5 rem for the dose equivalent to the critical bronchial
region, this corresponds to abbut 0.6 LC-cases/rem x 106 person-years.
The mean slope of the excess LC-mortality among the CSSR-miners follows from Figure 8 to 170 ± 50 LC-
cases per 10s persons per WLM. As the observation period and the number of person-years at risk is similar for
both groups, it should be expected also that a rather similar time distribution exists for the start of uranium-
mining. Under this assumption, the average number of years at risk per person of 38.622 person-years/3366
persons = 12 years in the USA-group should be applicable also to the CSSR-group. Taking into account this
conversion factor, an absolute risk coefficient follows:
aakg(CSSR-U-miners)= 14 ± 4 LC-cases
WLM x 106 person-years
Compared with the USA-group, the absolute risk coefficient in the CSSR-group is significantly higher,
although the relative risk coefficients are nearly equal.
-494-
-------
The preliminary CSSR-data, furthermore, indicate an increase of the risk coefficients with the decreasing
age of the miners at their start of uranium-mining. In addition, the real risk coefficients should be higher than
the values given ahove due to the limited observation period of about 20 years. On the other hand, the LC-risk
coefficient by radiation is probably considerably lower for non-smoking miners. Under these circumstances,
an absolute risk coefficient of 10 to 20 LC-cases/WLM x 106 person-years seems to be a reasonable reference
value for the assessment of the radon exposure limit to miners.
CONCLUSIONS
A strict application of the risk concept requires the assessment of the maximum acceptable radiation risk.
With respect to the LC-risk of miners exposed to radon-daughters, this risk limit must comply with two
different conditions.
Taking into account the strong variation of the natural background level of radon-daughters in air, the
occupational exposure limit should be 1 to 2 orders of magnitude higher than the average background level.
This means, that the limit for the occupational exposure of uranium miners, integrated over their total
lifetimes, should not be fixed below about 100 WLM. This lower limit would cause a LC-risk by radiation
comparable with the LC-risk from other sources (doubling exposure) and would yield an additional LC-risk of
about 1,000 to 2,000 cases/106 person-years.
On the other hand, the average radiation risk of uranium miners should not exceed the average risk of death
by other occupational diseases as it is observed among miners in conventional, safe-operated mines —
especially coal mines. The latter level of risk is about 200 to 500 cases/106 person-years. This level corresponds
to an average lifetime exposure of about 20 to 50 WLM averaged over all uranium miners. Taking into account
the general recommendation of the ICRP that the average dose of radiation-exposed workers should be kept
well below the recommended dose limit, an exposure limit of about 100 to 200 WLM for uranium miners seems
to be acceptable.
The necessity to meet both of the above mentioned conditions reveals the narrow margin which is available
for the assessment of an occupational exposure limit to radon-daughters. The resulting exposure limit on the
basis of this risk concept is comparable with the limit which was derived from the ICRP-dose limit of 15
rem/year. This leads to the final conclusion that, on the basis of our present knowledge, the risk concept and
the dose concept lead, within the limits of error, to the same value for the maximum permissible exposure to
radon-daughters.
REFERENCES
Altshuler, B., N. Nelson, and M. Kuschner, (1964), Health Physics 10,1137.
Haque, A. K. M. M., and A. J. L. Collinson, (1967), Health Physics 13, 431.
Harley, N. H. and B. S. Pasternack, (1972), Health Physics 23, 771.
Jacob!, W., (1964), Health Physics 10, 1163.
Jacobi, W., (1972a), Health Physics 23, 3.
Jacobi, W., (1973), Page 109-120 in Health Physics Problems of International Contamination, Hungarian
Acad. of Sciences, Budapest 1973.
Jacobi, W., (1972b), Health Physics 22, 441.
Lundin, F. E., J. K. Wagoner, and V. E. Archer, (1971), 2 Radon-Daughter Exposure and
Respiratory Cancer; Quantitative and Temporal Aspects, Nat. Inst. for Occup. Safety and Health —
Nat. Inst. of Environmental Health Sciences, Joint Monograph No. 1.
Report of the Advisory Committee on the Biological Effects of Ionizing Radiations, (BEIR-
report), (1972), Nat. Acad. of Sciences, Nat. Research Council (USA), Washington B.C., November 1972.
Report of the ICRP-task Group on Lung Dynamics, (1969), Health Physics 12,173 (1966); revised at
the ICRP-meeting, Oxford 1969.
Sevc, J. and V. Placek, (1973), Page 129-136 in Health Physics Problems of Internal Contamination,
Hungarian Academy of Sciences, Budapest 1973.
-495-
-------
0
0.2 (U 0.6 0.8 1.0 0 0.2 (H 0.6 0.8
! (RaA)
Figure 1. Fraction of the inhaled potential a-energy absorbed in the tracheobronchial and pulmonary
region (see text).
Figure 2. Mean a-dose per
WLM in the tracheobronchial (T-B)
and pulmonary (P) region as a
function of the uncombined fraction
fp of the potential a-energy in the
inhaled air (ICRP-lung model).
-496-
-------
10 15 20 25 30 h'
Figure 3. Expected mean a-dose per WLM to the
total bronchial and pulmonary region as a function of
the ventilation rate \v (box model).
-497-
-------
10
8
~
• 6
a>
>
Mean
Bronchial
a-Dose
Inhaled
Potential
a-Energy
Absorbed Energy Fraction
t (Bronchial Region)
0
10 20
Ventilation Rate
30 h'1
Figure 4. Estimated relative influence of ventilation
on the potential energy concentration in air and the
bronchial dose in a working area with constant rate of
radon- and aerosol- production.
-498-
-------
rod
WEfi
o
Q.
X
LU
20
10
5
05
0.2
0.1
© ALTSHULER et al. (1964)] Lung Geometry:
O JACOBI (1964) I LANDAHL-MODEL
• HAGUE «tal. (1967) j WEIBEL-MODEL
® HARLEYet al. (1972) J
0
Figure 5. Mean dose to the basal cell
layer in the segmental-subsegmental
bronchi per WLM, calculated from
different dosimetric models (v=20
1/min).
0 0.05 0.10
Free-Atom Fraction of Potential a-Energy
25
2 20
&_
o
i
o
TJ
OJ
0>
W)
JD
O
15
i 'o
0>
Q.
X
• • CSSR (1948-70)
o-—-o USA (1950-68)
I 95% - confidence range
1000 2000 3000
Cumulative exposure
4000
5000
Figure 6. Relative lung cancer-risk among uranium miners in the USA and CSSR versus their cumulative
exposure (in WLM-units).
-499-
-------
20000
52
g
>s
c
15 000
2
8
8 10 000
in
in
w
&
5000
USA-Uranium Miners
(1951-1971)
Slope:
3.2 + 0.8
Excess 1C-Coses
WLM-106pers.-years
Figure 7. Additional lung cancer-risk
among uranium miners in USA (from BEIR-
report, 1972).
1000 2000 3000 4000 5000 6000
Cumulative Exposure inWLM
Figure 8. Additional lung cancer-risk
among uranium miners in the CSSR (from
SEVC, et al, 1973).
120000
100000
CSSR-Uranium Miners
(1948-1970)
Slope:
175 ± 50
j Excess LC-Cases <\
1 WLM-106 Persons'
0 100 200 300 400 500 600
Cumulative Exposure in WLM
-500-
-------
THE BIOLOGICAL EFFECTS OF RADON ON THE LUNG*
D.A.Morken
University of Rochester
Rochester, New York 14642
Abstract
The biological effects of radon on the lung arise primarily from the alpha decay of the short-lived decay
products of radon inhaled. Radon is chemically inert and so is distributed throughout the body according to its
solubility in the various body tissues. The daughter products are isotopes of polonium, bismuth, and lead.
which are treated individually by the tissues and fluids, within the confines of their brief half-lives.
Experiments with inhaled decay products indicate the dose to lung is much greater than to the rest of the
body. Extensive injury was done to the bronchial tree, but the lesions did not lead to bronchial tumors. After
irradiation ceased, these lesions were quickly repaired. Permanent changes, but no cancer, were seen in the
region of the respiratory bronchioles.
Compared to other alpha emitters in the lung, the alpha radiation from radon is particularly inefficient in
producing radiation-specific tumors. The conclusion is taken that the radiation effects of alpha emitters may
be inconsequential compared to the chemical effects of the alpha emitters.
INTRODUCTION
The toxic effects from inhaled radon may result from the actions of three distinct radioactive agents. One of
these is radon itself, a noble gas which becomes distributed throughout the body and delivers a radiological
dose to all tissues from alpha particles. This radon is accompanied, however, by a series of short-lived
radioactive decay products which are produced within the body. These products are isotopes of polonium,
lead, and bismuth which may be treated by the body tissues to provide a distribution different from that of the
parent radon. These products decay by alpha and beta particle and gamma ray emission, and provide a dose
distribution different from that of the parent radon. These products of radon, produced within the body,
provide a second radioactive agent. In practice the two agents occur together.
The decay of radon in the atmosphere produces these short-lived products as ions which become attached to
dust particles. Inhalation of this radioactive dust results in the deposition of this radioactivity in the lung. By
virtue of the short half-lives (3 to 30 minutes) and the parent-daughter relationship which exists among these
products, they accumulate in the lung passages to provide a greater radiological dose than that from radon
alone. This then provides a third agent which delivers a radiological dose to the lung only.
The inhalation of radon and its decay products present in uranium mines is believed to produce bronchial
cancer in uranium miners. These miners exhibit an incidence of lung cancer greater than normal. Throughout
the literature on the radiotoxic properties of radon occurs the suggestion that radon is somehow responsible
for the lung cancer in miners; often radon is mentioned as a co-factor. Radon and its decay products are
present in other kinds of mines, but bronchial or lung cancer has never been attributed to the presence of the
radon. A large number of animal experiments over the past fifty years have been unable to confirm that radon
or its decay products can cause lung cancer. Work with other alpha-emitting isotopes has been successful in
producing neoplasia in the lung. A few experiments using radon combined with chemically toxic agents,
generally silica, but including uranium mineral, may have demonstrated a potentiation or promotion of the
tumorigenic properties of the chemical agents, or of the tumors themselves.
The biological effects of radon and its decay products contained within the whole animal are the subject of
another paper presented in this Symposium. This paper summarizes the histopathology of the lung following
inhalation by mice, dogs, and rats from an atmosphere which contained radon and dust, and compares this
histology with that from experiments involving other alpha-active isotopes.
THE EXPERIMENTAL METHOD
The experimental animals are exposed in a large chamber to a radon atmosphere which is established by
mixing radon from a generator with dusty air (normal room air) entering the chamber at a constant flow rate.
The chamber has a volume of 2,000 liters, and the air flow, which can be adjusted over wide limits, is usually
set near 40 liters per minute. The radon generator contains about 200 millicuries of radium in a dilute
hydrochloric acid solution which is aerated continuously from a source of compressed air at about 300
milliliters per minute; this generator provides about 22 microcuries of radon per minute. The generated radon
is scrubbed of acid and particulate material before being passed to the chamber. The radon concentration
established in the chamber depends on the air flow through the chamber — for the 40 1pm flow the
concentration is about 0.55 fzCi/1 in the steady state.
The time taken to attain the steady state is related inversely to the air flow, and at 401pm amounts to about
three hours. The radon decay products formed in the chamber become attached to the dust particles (normal
room dust), also attaining a steady state. At 401pm air flow the concentrations of RaA, RaB, and RaC on dust
become about 0.40, or 0.20, and 0.12 /LtCi/1, respectively. The chamber atmosphere is monitored hourly for
radon, with activity recorded on a strip chart. The radon concentration is also measured daily, during the
steady state, in an air sample taken with an evacuated ionization chamber, and is read with an electrometer.
*This paper is based on work performed under contract with the U.S. Atomic Energy Commission at the
University of Rochester Atomic Energy Project and has been assigned Report No. UR-3490-384.
-501 -
-------
The chamber atmosphere is occasionally sampled with a dust filter which is counted for alpha activity to
provide estimates of the concentrations of RaA, RaB, and RaC in the chamber.
The exposure atmosphere is evaluated and reported in terms of the Working Level (WL) unit, which is
derived from the concentrations of the airborne decay products of radon attached to dust, and which
represents a dose rate. The cumulated exposure dose is represented by the Working Level Month (WLM) unit.
The chamber is operated in the region of 1,800 WL, and the rate at which the dose is cumulated is about 11
WLM per hour of exposure.
For purposes of the radiobiologist, the radiation dose and dose rate to tissue are more meaningful, but less
easy to determine. The radiological dose to lung can be estimated from a knowledge of the concentration of the
decay products in the inhaled air, the breathing rate, the extent of deposition of the decay products in the lung,
the lung weight, and the clearance rates which act on the dust or radioactivity. The air concentrations can be
measured directly. The other quantities are not well established. For the present experiments, the radiological
doses are estimated from measurements made on the lung and respiratory tract. This direct procedure
removes the uncertainties associated with breathing rate, deposition, and clearance, but an uncertainty as to
the amount of tissue involved remains.
Randomization methods are used to sort animals into experimental and control groups, and into the various
dose categories. Following exposure, the animals were randomly sorted into a sacrifice schedule. Gross
examinations were made of all tissues, but, generally, tissue sections were taken only from the lung. Mortality
and weight were also followed, and in some cases the hemograms were examined.
EXPERIMENTS WITH MICE (Morken, et al., 1971)
CAFj strain mice were used in these experiments; the mice were 20 weeks old when they entered the
exposures. The mice were exposed to the radon atmosphere for three 50-hour periods each week for intervals
which ranged from 8 weeks to life. The chamber air flow was about 45 1pm and the concentrations of radon,
RaA, RaB, and RaC, in the steady state, were about 0.50,0.45,0.23, and 0.13 microcurie per liter, respectively.
This represented 2,000 WL, and a dose cumulation rate of 1,800 WLM per week of exposure. Measurements on
mice exposed for 24 hours, to attain the steady state, showed that the average dose rates from alpha acitivity
were; whole body, 5.5 rad per week; lung-trachea-bronchi, 280 rad per week; kidney, 18 rad per week; liver, 2 rad
per week; and gastro-intestinal tract, stomach and contents, 60 rad per week. The dose to bronchial tissue may
have been 5 to 10 times as large as that to the entire lung, or as much as 2,800 rad per week, but only the
average dose to the lung is used here. The exposures ranged, then, from 14,000 to 72,000 WLM, and the
estimated doses to whole lung ranged from 2,000 rad to 11,000 rad.
For the first experiment mice were exposed continually throughout life (Morken, et al., 1966). The median
exposure lifespan was 35 weeks, which represented a 50% shortening of life. At 40 weeks of exposure, when
only a few mice remained, these mice were taken for histopathologic study. A fairly extensive microscopic
examination of all organs was performed. No lesions were found in any organs or tissues — except the
respiratory tract. The lesions seen in the respiratory track were much the same in the animals examined.
The trachea section, taken at the level of the thyroid, displayed mucosal lesions which were hyperplastic,
metaplastic, and destructive; many times all of these lesions were observed in the same animal. The
hyperplastic changes consisted of a thickening of the mucosa with an increased number of cells which, in
general, tended to maintain a columnar structure. There was a number of layers of nuclei, which extended to
the surface of the mucosa, instead of the usual single layer near the basement membrane.
Squamous metaplasia consisted of areas of thickened mucosa made of cells resembling squamous
epithelium, which was sometimes accompanied by keratinization. The destructive changes consisted of either
a loss of mucosa or a single layer of elongated flat mucosal cells containing few nuclei; sometimes
inflammatory changes were seen in the mucous glands. These metaplastic and destructive changes
frequently occurred together.
The lesions observed in the major bronchi were similar to those of the trachea. Hyperplasia of the mucosa
was seen in which the mucosal cells maintained their columnar structure, but the mucosa contained
numerous layers of cells with nuclei extending to the surface. Squamous metaplasia was present, but not as
marked as in the trachea. The principal change seen in the small bronchi was the presence of large
hyperchromatic nuclei.
In the lung parenchyma, intra-alveolar edema was present in varying degrees in all animals. Foci of
alveolar phagocytes were seen in all animals; some phagocytes contained a brown pigment. A few small foci
of adenomatoid proliferation of bronchi accompanied by fibrosis were seen in a few animals.
The lesions of the bronchial mucosa may be summarized as destructive, hyperplastic, and metaplastic. No
infiltration carcinoma was found, but many of the changes may be considered as precancerous The lesions
seen in the alveoli probably resulted, in part, from the failure of the ciliary action of the bronchial mucosa.
The exposure dose for these mice amounted to 72,000 WLM, and the average dose to lung was about 11,000
rad, cumulated to death. The destructive effect of continued irradiation at this level of 2,000 WL was most
evident in the trachea and major bronchi.
In the second experiment, groups of mice were exposed for 15, 25, and 35 weeks. The exposure doses were
27,000, 45,000 and 63,000 WLM, respectively; the resulting lung doses ranged from 4200 to 9800 rad. The
purpose for the histopathology in this experiment was to observe the injury immediately following the end of
exposure and to observe, over a period of eight weeks, the repair of the injury. The 35-week exposure confirmed
-502-
-------
the median lethal dose of the previous experiment, but life shortening was not found for the 15 and 25 weeks of
exposure.
In the third experiment, groups of mice were exposed for 10,15,20, and 25 weeks. This was a pathology-only
experiment, which involved sacrifices immediately following exposure of each group; and then, beginning at
the mouse age of 60 weeks, sacrifices at intervals of 10 weeks until all the mice were used (this occurred at 110
weeks). The experiment was done to examine late effects following the relatively short exposure periods used.
The exposure doses ranged from 18,000 to 45,000 WLM; the resulting lung doses ranged from 2,800 to 7,000 rad.
Age-control mice were also examined at each sacrifice. The pathological effects found in these experiments
were not large. In general, lesions in the trachea and bronchi were similar to, but not as extensive as those
observed in the previous experiment. The extent of the lesions seen immediately after exposure seemed
proportional to the cumulated dose (or exposure duration); the repair was so rapid that by eight weeks post-
exposure, the tissues appeared normal. The trachea showed only focal atypism of moderate degree. Nuclei
were large, and the cytoplasm was diminished with loss of cilia immediately after exposure. Little or no
squamous metaplasia was seen. None of the tracheal alterations was believed to be precancerous.
Changes in the large bronchi were less marked than in the trachea. The terminal bronchioles showed
atypical changes in the cuboidal lining epithelium. With increased post-exposure time, the epithelium in
many terminal bronchioles tended to become flattened or to disappear.
The tumors encountered were variants of those which appear spontaneously in this strain of mouse
(adenomas and foci of adenomatosis), and were seen in some of the older controls for these experiments. These
tumors were seen at considerably lower ages in mice exposed to radon than in the controls. No tumors were
found after 25 weeks of exposure in any animal sacrificed between 25 and 100 weeks from the end of exposure.
Most of the adenomas showed qualitative changes suggestive of a more malignant behavior —. greater
cellular anaplasia, a higher mitotic rate, and invasion of adjacent lung tissue.
Neoplastic lesions (tumorlets) were seen in mice sacrificed between zero and eight weeks after the 35-week
exposure. Most of these were of a type not encountered after any of the lower doses. These consisted of ill-
defined areas containing many small, often intra-alveolar, nests of bizarre cells. Sometimes these were
multiple in one or more lobes, and some cells were of a squamous type associated with what appeared to be
keratin. Other tumors in this group had a glandular pattern exhibiting local malignant features. These
tumorlets did not appear in other groups, and have since been observed in control rats.
Non-specific pulmonary effects, perhaps attributable to the irradiation, consisting of edema, increased
numbers of macrophages, and a few small foci of interstitial fibrosis, were observed in some lungs at long
intervals after exposures of 25 weeks; edema and many large macrophages were noted in all the mice exposed
for 35 weeks. Moderate to severe pneumonitis was seen in late age in both the controls and exposed mice.
EXPERIMENTS WITH DOGS (Morken, et al., 1972)
Experiments with dogs were carried out to provide irradiated lungs which approach the dimensions of
human lungs, and to use a longer-lived animal so that observation could extend over several years. The initial
experiment involved 42 dogs exposed to doses from 200 to 10,000 WLM, delivered in one to 50 days.
Histological examinations were conducted at zero, 1, 2, and 3 years post-exposure, for all dose levels. The
radon atmosphere of 1,800 WL was about the same as used in the mouse experiments. Exposures were made
daily for five days each week, at 20 hours each day. The rate of cumulation of dose was about 2000 WLM per
day of exposure. Exposure periods were 1,2,4,8,15, and 50 days.
Chambered control dogs as well as age control dogs were included. The chamber control dogs spent their 20
hours per day in a chamber identical in size and operation to the radon chamber, but the generator contained
no radium.
Randomization procedures were followed to place the dogs into the various dose categories, and into the
sacrifice schedule. At sacrifice the dog lung was partially fixed with formalin in situ, and a prescribed protocol
was followed to provide samples from the lung, respiratory tract, lymph nodes, nasal epithelium, and
turbinates of the nose. Gross examinations of all organs were made.
Several dogs were exposed to the radon atmosphere for about 20 hours, and then killed immediately on
removal from the chamber. The lungs, trachea, and major bifurcation were removed within a few minutes.
These tissues were weighed and measured; the RaB and RaC were measured over a period of two hours. The
counting data were resolved by decay curve analysis to provide the information needed to compute the dose
rate to the trachea and lung.
The average dose rate to whole lung amounted to one millirad per hour per WL, or 0.17 rad per WLM, when
the data were analyzed by a curve fitting procedure. The average dose in rad per WLM, with standard
deviation, amounted to 0.22 ± 0.16 when considering zero clearance, and 0.26 ± 0.19 when considering a
clearance effect. Individual lung doses ranged from 0.08 to 0.79 rad per WLM.
For the trachea and bifurcation, the volume used in the dose calculation was determined from the averaged
inside dimensions of the sections and an alpha absorption layer thickness of 70 \i m. These doses ranged from
0.28 to 12.5 rad per WLM (mean 4.73 ± 4.36 standard deviation) for trachea and from 0.36 to 20.7 (5.03 ± 6.20)
rad per WLM, for the bifurcation. The average dose rate to the upper respiratory tract was about 30 times that
to the alveolar region. For the dog, then, the alveolar doses ranged from 34 to 1,700 rad, while the tracheal
doses ranged from 1,000 to 50,000 rad. The dose to the bronchi may have been greater than that to trachea.
No cancer has been found in these dogs. Pathologic changes were found only in the alveolar and
bronchiolar regions of the lung. Even immediately after exposure, when the lesions produced directly by the
-503-
-------
radiation should have been most evident, injury to the trachea and bronchial tree was absent. In the lung, the
microscopic lesions observed at any dose were subtle, variable, very small spatially, widely separated, and
involved only a small fraction of the lung substance. Immediately at the end of exposure no significant
differences from age or chamber controls were apparent for doses from 200 to 10,000 WLM. At one and two
years after exposure, there was a probable increasing relation to dose of small foci of chronic inflammation. At
three years post exposure this relation had disappeared for doses to 800 WLM, but was still probable for the
larger doses. The lesion at all doses appeared as small patches of thickened (fibrotic) alveolar walls, with some
metaplasia of the alveolar cells, and some epithelial hyperplasia of the bronchioles. Chamber control dogs
were not different from age control dogs.
Animals exposed to the same dose expressed in WLM are not likely to receive the same radiologic dose. The
doses, normalized to the WLM dose, as described above, and estimated from lung measurements on eleven
dogs, varied from 0.08 to 0.79, with the average near 0.2 rad per WLM. It is not unlikely that in some cases a
lesser WLM dose may produce greater injury than would a greater dose, since the adjacent dose ratios of 2 and
3 used were less than the ratio of 10 found among the animals.
EXPERIMENTS WITH RATS (Morken, et al, 1971)
These experiments were performed with rats exposed to a radon concentration near one microcurie per liter
for five hours each day, five days each week, to a total of 600 hours. The chamber air flow was reduced to about
221pm in both the radon and the control chamber. Five hours from starting the daily exposure, the radon
concentration attained the desired level of 1 /iCi per liter. The concentrations of RaA, RaB, and RaC attained
by five hours were 0.75, 0.53, and 0.43 microcurie per liter, respectively. For short exposures, the cumulated
dose is related linearly to the hours of exposure to the steady state values. The steady state concentration
attained for radon, RaA, RaB, and RaC were 1.4, 1.1, 0.78, and 0.63 microcuries per liter, respectively. The
chamber operated at 7,300 WL and the exposure dose rate was 43 WLM per hour. In 600 hours the cumulated
dose amounted to 25,800 WLM. If the factor of 0.17 rad per WLM found for the mouse and the dog is applied, the
cumulated average dose to lung was about 4,300 rad. If the factor of 30 applies for the ratio of trachea-to-lung
dose, the upper respiratory tract may have received a cumulated dose of 130,000 rad.
Radon-exposed, chamber control, and age control rats, both regular and pathogen-free, were used. Lung
tissues were taken for histopathology examination following the death of the animal. By 12 months from the
end of 6 months of exposure, no cancer was evident in the standard rat. Cancer was also not evident in the
pathogen-free rats at 6 months after the end of 6 months of exposure. It is difficult to distinguish among the
exposed, chamber controls, and age controls. This experiment is not yet complete, and further detailed study
will be made when all the animal data are available.
A complete appraisal of each of these experiments will include descriptions of life shortening, weight
changes, hemogram changes (in some cases), and, of course, details on the changes in the lung and other
organs. These appraisals will be published elsewhere. For the purpose of this paper, only a brief description of
the major effects in the lung has been provided.
The results of these experiments on mice, dogs, and rats are significant in several aspects. While extensive
injury was done to the bronchial tree during irradiation by the alpha particles, the lesions did not lead to
bronchial tumors or cancer — even with continued irradiation; and after irradiation ceased, these lesions were
quickly repaired. Late effects did not appear in the bronchial tree. In the alveolar region, the injury was not
immediately evident, but appeared later in life, whether or not the irradiation continued. Major, perhaps
permanent, changes occurred in the region of the respiratory bronchioles.
The essentially negative character of the results seen here suggest that alpha irradiation is a particularly
inefficient way to produce radiation-specific tumors, if such exist, in the lung or respiratory tract. Since the
later effects seen here were similar to those reported to follow radon-only inhalation at a lung dose of only 300
rad (Morken, 1973), an uncertainty exists concerning the role played by the massive lung doses provided by
the inhaled decay products of radon.
The relative ineffectiveness of irradiation by alpha particles to produce even microscopically visible
evidence of radiation-specific tumors, or cancer in the lung, or to produce permanent alterations in the
respiratory track which could lead to bronchial carcinoma, is impressive. The only apparent late and
permanent changes occurred in the alveolar regions of the lung, for a wide range of doses, and for observation
times to three years in the dog, and one to two years in the mouse and rat. These alveolar changes may
eventually lead to cancer, but the cancer would be of peripheral origin, and not the kind generally expected in
view of the large doses to the respiratory tract.
A few experiments using radon combined with chemically toxic dusts may have demonstrated a
potentiation or promotion of the tumorigenic properties of the chemical agents or of the tumors themselves.
Kushneva (1961) exposed rats to quartz dust, to radon, and to the dust with radon. The dust, 50 mg, was given
by intratracheal injection. The radon was given by inhalation at 8f/Ci per liter in three one-hour exposures.
Experimental silicosis developed rapidly in the rats which received quartz dust alone. Rats which recieved
radon alone displayed a general systemic response. Rats which received the dust-and-radon showed a more
intense silicosis, and the author felt that, for the combination, the radon vitally influenced the development of
thesilicosis by way of increased numbers of nodules, acceleration of fibrosis., and increased metaplasia of the
bronchial and alveolar epithelium. The radiation dose amounted to only a few rad.
-504-
-------
Kilibarda, et al, (1967) reported similar experiments, but with longer exposures to radon. Silica was given
intratracheally as a sterile suspension at 15 mg per rat. These investigators reported that radon inhalation
appeared to aid the silicogenic effect only when given after the silicotic nodules were formed; before that,
radon had no effect on the process. Once the nodules were formed radon inhalation made them larger and
more easily delineated from the parenchymal tissue. Radon alone had no effect in producing tumors. The
whole body doses for radon exposures, which ranged from 180 to 900 hours at a concentration of 0.07 u Ci per
liter, amounted to only a few rad — whereas the lung doese were in the area of 3,000 WLM or about 600 rad.
French workers (Perraud, et al, 1970) have also reported an enhancement of the effects produced in lung by
silica, uranium mineral, and cerium. The authors found that the chemical must be given before the radon is
given. Epidermoid carcinoma resulted in the uranium and cerium experiments. An estimate of the exposure
dose for the cerium experiment, 540 hours at 0.7 jiCi per liter, is 10,000 WLM at 3,700 WL for a tissue dose
of 3,000 rad. The tumors appeared in 10 months in the bronchiolar/alveolar region.
Gates, et al., (1961) reported that external gamma or x-irradiation increases the incidence of spontaneous
tumors, but did not appear to produce new kinds of tumors; this radiation, then, provides a promotional or
potentiation effect.
A few researchers believe that while radon may produce little or no effect by itself, it is a powerful stimulus to
act on tumors already present.
The French workers also found that the chemical agent might not be required, and reported a 100%
incidence of pulmonary carcinoma in pathogen-free rats exposed at 1 ^iCi per liter for 600 hours (6 months).
Tumors appeared by 10 months from the start of the exposure. The exposure dose may have been as large as
25,000 WLM. The rat experiment reported in this paper reproduced the French experiment with a different
type of radon generator; and, as reported here, no tumors or cancer have appeared in the rats; this experiment
is not yet complete, however.
The ability of other kinds of alpha emitting isotopes to induce bronchial and pulmonary cancer is well
documented and familiar. Reviews of these experiments are available (Gates, ei al., 1961; Albert, 1966; Bair,
1970; Sanders,et al., 1970). Inhalation of polonium-210by rats with doses estimated from 70 to 500 rad resulted
in primary lung cancer within two months post-exposure. Inhaled plutonium-239 by dogs produced lung
cancer by 150 days with doses from 9,000 to 23,000 rad. 239pu given intratracheally to mice produced
fibrosarcoma at 500 days with a dose of 115 rad, and squamous cell carcinoma at 400 days with a dose of 2,300
rad. 239Pu inhaled by mice produced bronchiolar carcinoma at 500 days with a dose of 600 rad; in rats,
epidermoid carcinoma was evident by 250 days at a dose of a few thousand rad.
Inhalation of long-lived alpha emitters results in an irradiation which continues long after the inhalation
has ceased. Doses are usually estimated to the time of appearance of the lesion. With radon inhalation the dose
is accumulated entirely within the period of exposure, or inhalation. If a latent period operates, the radon
exposure yields a valid dose — whereas the other materials do not. Thus, the tumor-producing doses may be
considerably less than the estimated doses reported for other isotopes. With the continuously irradiating
isotopes, it appears that tumors occur sooner for larger dose rates, but this idea has not been subjected to
critical study.
A large number of experiments have been done with implanted sources. These are not considered here
because the implant method involves the manufacture of a lesion which may become a part of the response if
irradiation acts to promote lesions to tumors.
Thus it appears that relatively small doses of alpha radiation can produce tumors and cancer in the lung in
relatively short periods of time, and especially within the lifetime of the mouse or rat. Since the radiation from
radon or its decay products have not accomplished this, there may be some hesitation to label this tumor
production a radiation effect.
The materials which have produced cancer are chemically reactive and long-lived so that appreciable
amounts of the material are present throughout the exposure, compared to the radon exposures. The specific
activity of polonium-210 is 4,300; that for plutonium-239 is 0.06. In contrast, the specific activities for the radon
decay products range from 108 to 1016; while for radon it is 105. In terms of a radiation-to-chemical exposure,
the decay products are purer sources of radiation by factors of 105 to 1013 for polonium and 107 to 1015 for
plutonium.
CONCLUSIONS
It appears reasonably easy to induce lung cancer with radionuclides of small specific activity (large specific
mass). It has also been easy to induce lung tumors with non-radioactive chemical compounds. However, it has
been extraordinarily difficult to produce even small long-lasting effects with alpha irradiation from radon
and its decay products, and lung cancer has not been unequivocally demonstrated with radon.
Alpha radiation may promote chemically produced lesions leading to tumors and cancer, but the lesions
must be there first. This suggests that alpha radiation may act indirectly through a systemic response —
rather than directly on the cell.
Thus chemical induction of lung cancer by radioactive elements may be of greater importance than
induction by alpha radiation. Threshold limits based on chemically active elements should not be applied to
elements of greater radiation purity.
It would be prudent to study the probable chemical behavior with cells of the heavy metal isotopes,
regardless of their radioactive natures. The use of radon may permit resolution of the effect into its
radiological and chemical parts.
-505-
-------
REFERENCES
Albert, R. E., (1966), Thorium, Its Industrial Hygiene Aspects, pp 100. Academic Press.
Bair, W. J., (1970), Inhalation of Radionuclides and Carcinogenesis, in Proceedings of the Conference on
Inhalation Carcinogenesis, U. S. A. E. C. Division of Technical Information.
Gates, O. and S. Warren, (1961), Histogenesis of Lung Carcinoma in Mice. Arch. Pathology 71,693.
Kilibarda, M., V. Visnjic, D. Panov, R. Radovanovic and L. Novak, (1967), Deposition of Some
Radon Daughters in Rats after Inhalation of Radon, with Special Consideration of the Simultaneous
Influence of SiO2, in Diagnosis and Treatment of Deposited Radionuclides, pp 222. Excerpta Medica
Foundation.
Kushneva, B. S., (1961), On the Problem of the Long Term Effects of the Combined Injury to Animals of
Silicon Dioxide and Radon, in AEC-tr-4473, pp 21-28.
Morken, D. A. and J. K. Scott, (1966), The Effects on Mice of Continual Exposure to Radon and its Decay
Products on Dust. University of Rochester Atomic Energy Project, Report- UR-669.
Morken, D. A. and C. L. Yuile, (1971), (unpublished materials.)
Morken, D. A. and G. W. Casarett, (1972), (unpublished material).
Morken, D. A., (1973), The Biological Effects of the Radioactive Noble Gases, this Proceedings.
Perraud, R., J. Chameaud, R. Masse and J. Lafunia, (1970), Cancers Pulmonaires Experimentaux
chez le Rat Apres Inhalation de Radon Associe a des Poussieres non Radioactives. C. R. Acad. Sci, Paris, 270
2594.
Sanders, C. L., Jr., R. C. Thompson and W. J. Bair, (1970), Lung Cancer: Dose Response Studies with
Radionuclides, in Proceedings of the Conference on Inhalation Carcinogenesis, pp 285-303. U. S. AEC
Division of Technical Information.
-506-
-------
BIOLOGICAL EFFECTS OF DAILY INHALATION OF RADON AND ITS SHORT-LIVED
DAUGHTERS IN EXPERIMENTAL ANIMALS*
R. F. Palmer, B. O. Stuart, and R. E. Filipy
Biology Department
Battelle
Pacific Northwest Laboratories
Richland.WA 93352
Abstract
Syrian golden hamsters, C57BL mice, and specific-pathogen-free rats were exposed simultaneously in
groups of 16 animals each for 90 hours per week to aerosols consisting of radon plus 3000-6000 Working Levels
of radon-daughters with and without 18 mg/m3 carnotite uranium ore dust. Condensation nuclei
concentrations ranged from 2000-4000 per ml and from 90,000-120,000 per ml in the chamber without and with
uranium ore dust, respectively. At 4 months of exposure only one of the rodents remained alive.
Histopathology of radon-daughter exposed mice includes acute interstitial pneumonitis, severe pulmonary
congestion, and supperative rhinitis; mice inhaling radon-daughters with ore showed these lesions plus
macrophage proliferation, alveolar septal cell hyperplasia, and bronchial epithelial hyperplasia. Hamsters
inhaling radon-daughters showed proliferating lesions characterzied by alveolar septal thickening,
bronchiolar epithelial hyperplasia, septal fibrosis, and occasionally adenomatoid metaplasia and squamous
metaplasia. Hamsters inhaling radon-daughters with ore dust showed similar effects plus granulomatous
response and intense septal fibrosis. Rats inhaling radon-daughters showed lesions similar to those of
hamsters but more focalized with classic radiation pneumonitis; rats exposed to radon-daughters with ore
showed similar lesions, with greater consolidation and pneumoconiosis. These findings will be discussed in
relation to pulmonary pathology in uranium miners.
INTRODUCTION
Initial studies in our laboratory with hamsters exposed for their lifespan to 30 Working Levels (WL) of
radon-daughters produced only slight pulmonary responses. A Working Level (WL) is defined as any
combination of short-lived radon-daughters in one liter of air that will result in the ultimate emission of 1.3 x
105 MeV of alpha energy from radioactive decay. Similar levels have been found in some areas of operating
uranium mines. Lifespan exposures of hamsters to 600 WL radon-daughters with uranium ore dust caused
pulmonary lesions, including vesicular emphysema, fibrosis, metaplasia, and adenomatous lesions with
anaplasia. A subsequent experiment involved lifespan exposures of hamsters to 1,200 WL radon-daughters,
diesel engine exhaust, and uranium ore dust. These hamsters showed more extensive and severe pulmonary
lesions, with accelerated development of cuboidal and squamous metaplasia of the bronchial epithelium, and
the appearance of squamous tumors.
Earlier studies (Morken and Scott, 1966) with mice exposed to 1,750 WL radon-daughters on room air dust
(150 hrs/wk), showed a significant reduction of lifespan. Workers in France (Perraud, et al., 1972) have
reported peripheral tumors in the lungs of SPF rats after 500 hours of exposures to approximately 1 \i Ci per
liter of radon, but interpretation of these results is extremely difficult because the levels of radon-daughters
were not described, and no correlation to human exposure histories based on Working Levels (radon-daughter
levels) was possible.
This paper presents the results of a pilot study that we conducted to correlate these results (obtained in three
different laboratories) to provide direct interspecies comparison of the effects of exposure to identical aerosols,
and to determine if there exists an increased sensitivity or resistance of one or more of these species of rodents.
We prepared systems for simultaneous exposures of mice, hamsters, and SPF rats in the same chambers to
high levels of radon-daughters, with and without concomitant exposure to uranium ore dust (carnotite).
Exposure of the three species in these initial studies were made at levels of radon-daughters calculated to
induce tumor production or acute death.
METHODS
The animals used in this study were SPF Wistar rats, Syrian-Golden hamsters, and C57 B1/6J mice. They
were all males and approximately 100 days old at the start of exposures. The animals were individually
housed in two types of compartmented, stainless steel mesh cages. There were 16 compartments for mice or
hamsters in one type of cage, and 8 compartments for rats in the other type. Two cages of rats, one of hamsters,
and one of mice were in each of the two exposure chambers. Ten control animals of each species were
individually housed in the exposure room, but not in a chamber. Food and water were available to the animals
at all times. The experimental design is shown in Table 1.
*This paper is based on research per formed under U. S. Atomic Energy Commission Contract AT(45-1)-1830.
-507-
-------
The two exposure chambers were nearly cubical with volumes of 1,700 liters. As seen in Figure 1, the animal
cages were supported on racks that positioned the bottoms of the cages 81 centimeters below the chamber
inlets for radon or radon and uranium ore dust. Unfiltered room air entered at the top of each chamber, passed
through the chambers, and left via outlets centered 20 cm above the chamber bottoms. These outlet air
streams passed through valved flow meters and absolute filters, and were then discharged into a high-flow,
low-vacuum system whose outlet was diluted in the building exhaust stack prior to discharge to the
atmosphere.
A separate radon generator was used for each chamber. Saturated air bubbling through acid solutions of
radium chloride at a rate of 150 ml/min swept radon from the generators through traps to remove acid vapor
that may have been carried out of the generators. This radon laden air was filtered just prior to introduction
into the chambers. The radon entered the center top of the chambers where it mixed with incoming room air in
the case of Chamber 1, and with incoming room air and uranium ore dust in Chamber 2. The uranium ore dust
was introduced into Chamber 2 by means of a Wright Dust Feed Mechanism mounted on top of the chamber.
Ore dust with a count median diameter of 0.24 Jim was maintained at a concentration of 18 ± 4 figperliterin
the chamber during exposure periods. This level was selected to coincide with those used in our previous
studies, serving to lower the unattached radon-daughters to only a few percent.
Initial measurements in the empty chambers indicated that nearly twice as much radon would be required
in Chamber 1 without ore dust in order to attain approximately equal Working Levels of radon-daughters in
the two chambers. This was felt to be due to greater losses of radon-daughters to the walls of Chamber 1 than to
those of Chamber 2 which contained the ore dust. As a consequence, one radon generator was loaded with
approximately 1.3 Ci of radium and the other with approximately 0.7 Ci. With a total air flow of 35 liters per
minute through each chamber, these levels of radium led to radon concentrations of 4.8 fid/liter in Chamber
land 2.5 JzCi/liter in Chamber 2.
Radon levels in the chambers were continuously monitored during exposures, and each chamber was
sampled at least once during each exposure period for measurement of radon-daughter concentrations, using
a two channel alpha analyzer employing a solid state surface barrier detector. Concentrations of
condensation nuclei were measured one or more times during each exposure period using a Gardner nuclei
counter. These concentrations ranged from 2,000 to 24,000 per ml in Chamber 1 and from57,000to 130,000 per
ml in Chamber 2.
When animals were first introduced into the chambers, measured levels of radon-daughters and
condensation nuclei showed little change in the chamber containing the aerosol of radon-daughters with
uranium ore dust (Chamber 2). However, in Chamber 1, without uranium ore dust, radon-daughters were
found to be reduced by a factor of at least four from levels observed in the empty chamber. Condensation nuclei
were reduced from approximately 19,000 to 2,000 per ml. The radon level in the chamber remained at
approximately 4.8 fid per liter, however, so it was decided to continue the animal exposures with unequal
Working Levels of radon-daughters in the two chambers.
The animals were exposed in the two chambers for approximately 90 hours per week, in two continuous 45-
hour periods. All animals were weighed biweekly, and exposures were continued until all the animals had died
or were sacrificed when moribund. At death the nose, trachea, lungs, liver, spleen, and kidneys of each animal
were retained for histopathological investigation.
RESULTS
Table 2 shows the mean weights plus or minus the standard error of the means for the mice in the control
and in the two exposure groups at various times after the start of exposures. As early as 55 days after the start
of exposures, both exposure groups showed significant weight loss; i.e., 22% reduction from control animal
weights. Table 3 shows the same data for hamsters. In this species, the reduction in weight did not occur as
soon after the start of exposures, but it was just as dramatic as exposures progressed. As observed in the mice,
the weights of the hamsters exposed to radon-daughters with uranium ore dust show greater reduction from
control hamster weights than do those exposed to radon-daughters without ore dust. The weight data of
exposed rats shown in Table 4 tell the same story. Significant reductions from control rat weights were seen in
both exposure groups by 55 days after the start of exposures, with the greater weight reductions occurring in
the rats exposed to radon-daughters with ore dust. It is evident that all three species showed a marked weight
loss in both exposure groups, with those exposed to radon-daughters plus uranium ore dust showing more
drastic effects in all three species.
Figure 2 shows the survival curves for each of the species exposed in Chamber 1, together with a curve
showing the Cumulative Working Level Hours (CWLH) of radon-daughters to which they were exposed. The
increase in the slope of the CWLH curve is a result of the radon-daughter concentrations increasing as the
number of animals in the chamber decreases. The radon level in the chamber remained at x 4.8 J/Ci/liter. It
can be seen that the first death, a mouse, occurred only 25 days after the start of exposures. The last animal to
die, a hamster, had been exposed to over 8 million CWLH of radon-daughters during the 178 days it survived
after the start of exposures. The mice died earlier and at a faster rate than did rats and hamsters, but as the
exposures continued the percent surviving showed little difference among the three species.
Figure 3 shows the survival curves for each of the species exposed to radon-daughters with uranium ore dust
in Chamber 2. The nearly constant slope of the CWLH curve shows that the concentrations of radon-
daughters in the chamber were little affected by the number of animals in the chamber. Half of the mice in this
-508-
-------
group had died by 30 days after the start of exposures, compared to nearly 120 days before 1/2 of the rats and
hamsters were dead. The first rat died 52 days after the start of exposures, at which time it had been exposed to
•vl.6 million CWLH of radon-daughters. The last hamster to die had been exposed to 7 million CWLH during
the 144 days it survived after the start of exposures.
Table 5 shows the geometric mean survival times, and their 95% confidence limits, for the rats, hamsters,
and mice in each of the two exposure chambers. All three species show significantly shorter survival times
following exposure to radon-daughters with ore dust than they do following exposure to radon-daughters
without ore dust.
Histopathological examination of tissues from the rats, hamsters, and mice reveals varying degrees of
radiation pneumonitis due to inhalation of radon-daughters. The characteristic septal fibrosis, alveolar
lining cells sloughed into alveolar lumena, and invasion by macrophages, as well as the atypical nuclei of
alveolar septal cells, are illustrated in Figure 4. In animals exposed to uranium ore dust with radon-daughters,
one sees a pulmonary response to particulate materials diagnosed as uranium ore pneumoconiosis.
Macrophages laden with uranium ore dust, and the septal fibrosis characteristic of this lesion are shown in
Figure 5. Bronchiolar epithelial hyperplasia was commonly seen in the lungs of animals exposed to radon-
daughters.
Mice and rats seem less prone to the pulmonary lesions than hamsters, but are afflicted with quite severe
lesions of the upper respiratory tract. Severe suppurative rhinitis and squamous metaplasia of nasal
epithelium (Figure 6) was a common observation in the mice and rats used in this experiment. Dense exudate
containing many neutrophils was seen in the nasal passages. Suppurative laryngitis and bronchitis were
frequent observations in all three species, but were most severe in rats (Figure 7). Suppurative tracheitis was
also quite common, and, in at least two mice, the tracheal epithelium had undergone a keratini/ing squamous
metaplasia (Figure 8). The squamous nature of the ordinarily columnar epithelial cells and the keratin layer
are quite atypical of the trachea.
Hamsters also had a high incidence of rhinitis and laryngitis, but had a higher incidence of pulmonary
consolidation due to septal cell hyperplasia, interstitial pneumonitis, and macrophage proliferation. An
additional lesion found with approximately equal frequency in rats and hamsters was an adenomatoid
metaplasia of alveolar epithelium. This lesion was found in small foci in rat lungs, but involved large areas in
hamster lungs. In two hamsters exposed to radon-daughters without ore dust, the lesion had undergone
squamous metaplasia (Figure 9). In view of previous observations in hamsters, this lesion is considered pre-
malignant; i.e., a stage just previous to epidermoid carcinoma.
In all three species, the major differences between exposures to radon-daughters only and radon-daughters
with uranium ore, seem to be the increased septal fibrosis and macrophage proliferation in the latter. The
uranium ore dust particles also seem to cause a slightly higher incidence of pulmonary emphysema and septal
cell hyperplasia than found in animals exposed to radon-daughters only.
DISCUSSION
This pilot study of the effects of inhalation of high levels of radon-daughters, with and without concomitant
uranium ore dust, by three rodent species has shown several interesting results.
Exposures for 90 hours per week to radon-daughters ranging from 2,000 to 8,500 WL, with and without 18
/ig per liter uranium ore dust, caused marked life-shortening in all three species. Marked reduction in body
weights occurred in all three species, with weight losses of 30-50% of control animal values in all species after 3
1/2 months of exposures. Mice exposed to radon-daughters and ore dust were particularly susceptible in terms
of mortality, although the lungs of these animals showed very little pathological change.
Classical radiation pneumonitis with alveolar septal fibrosis and occasional bronchiolar epithelial
hyperplasia were the predominant deep lung lesions seen in all species. In contrast to hamsters exposed 30
hours per week to 1,200 WL of radon-daughters and uranium ore dust, proportionately more of the pathology
was seen in the upper respiratory tracts of the hamsters in the present study. The contrast between markedly
affected trachea and major bronchi vs. relatively little effects in deep lung was most evident in rats. Findings
of severe suppurative laryngitis and bronchitis were frequent in rats, and may have been an important
contributing factor to their death. These findings dictate further studies involving sacrafice and radioactivity
analyses of tracheal and lung tissues to determine relative absorbed radiation doses at these sites for
correlation with developing degenerative and proliferation changes of the respiratory tract in each species.
Our next experiment involves exposures of rats, hamsters, and mice during five, 6-hour periods per week.
This lowered exposure rate may allow the animals to live long enough for proliferative epithelial changes to
progress beyond the stage of squamous metaplasia to possible invasive tumor formation.
REFERENCES
Morken, D.A. and J.K. Scott, (1966), The Effects on Mice of Continued Exposure to Radon and Its Decay
Products on Dust. UR-699. University of Rochester, New York.
Perraud, R., J. Chameaud and J. Lafuma, (1972), Cancer Broncho-Pulmonaire Experimental Du Rat
par Inhalation de Radon. Extraits Francais Med Chir Thor, 26, 25-41.
-509-
-------
TABLE 1. Experimental Design.
Group Number of Animals Exposure (90 Hours per Week)
1 16 of Each Species
2 16 of Each Species
2,000-8,500 WL Radon-Daughters
6,000-7,500 WL Radon-Daughters
with Uranium Ore Dust (18 mg/m3)
3 10 of Each Species Controls (Not Housed in Chamber)
TABLE 2. Mouse Weights (Mean + S.E. in Grams).
Days Since Start Daughters
Of Exposures Radon-Daughters +Ore Controls
0
55
111
132
27+1
25+1
23±1
19 ±1
27 + 1
25 + 1
21
28+1
32+1
34±2
34±2
TABLE 3. Hamster Weights (Mean ± S.E. in Grams).
Days Since Start Daughters
Of Exposure Radon-Daughters + Ore Controls
0
55
111
132
117 + 3
121 + 3
106±4
93 + 8
120 + 3
119 + 4
99±12
68
124 + 3
128±4
135 + 3
136 + 3
TABLE 4. Rat Weights (Mean + S.E. in Grams).
Days Since Start Daughters
Of Exposures Radon-Daughters +Ore Controls
0 466 + 9
55 479 ±9
111 407+17
132 401 + 23
442 + 7
439+10
364+26
321
470+10
544+12
600+14
616+15
TABLE 5. Geometric Mean Survival Times
(Days Since Start of Exposures
With 95% Confidence Limits).
Mice Hamsters Rats
Radon-Daughters 94™° 131, f^
to 115
Daughters + Ore 43
(63
'29
107
116
98
120
105,
132
109
121
90
-510-
-------
Figure 1. View of multi-species exposure chambers and glovebox containing the radon generators.
-------
RADON DAUGHTERS
AGE AT START GEOMETRIC MEAN SURVIVAL
OF EXPOSURE SINCE START OF EXPOSURE
(DAYS) (DAYS)
MICE
HAMSTERS
—— RATS
ion*132
120-109
L—
DAYS SINCE START OF EXPOSURE
Figure 2. Survival curves of rats, hamsters, and mice exposed to radon-daughters without concomitant
uranium ore dust, together with a curve showing the Cumulative Working Level Hours of radon-daughters to
which triey were exposed.
-------
Figure 3. Survival curves of rats, hamsters, and mice exposed to radon-daughters with concomitant
uranium ore dust, together with a curve showing the Cumulative Working Level Hours of radon-daughters to
which they were exposed.
-------
,*
Figure 4. Radiation pneumonitis in hamster lung showing characteristic septal fibrosis, alveolar lining
cells sloughed into alveolar lumena, invasion by macrophages, and atypical nuclei of alveolar septal cells.
(H&E.660X).
-514-
-------
-
^ ,
a * m • &r*^
Figure 5. Uranium ore pneumoconiosis in a hamster lung showing ore dust laden macrophages and septal
nbrosis(H&E,660X).
-515-
-------
Figure 6. Suppurative rhinitis and squamous metaplasia of nasal epithelium seen in a mouse exposed to
radon-daughters and uranium ore dust (H&E, 415X).
-516-
-------
t * ** *
_
.-*
Figure 7. Suppurative laryngitis and bronchitis in a hamster (H&E, 660X).
-517-
-------
Figure 8. Suppurative tracheitis in a mouse showing keratinizing squamous metaplasia of tracheal
epithelium (H&E, 120X; Insert X540).
-518-
-------
Figure 9. Squamous metaplasia of alveolar epithelium from the lung of a hamster (H&E, 330X).
-519-
-------
EFFECT OF VENTILATION VARIABLES ON BREATH THORON OUTPUT*
J. E. Ballou
Biology Department
Battelle
Pacific Northwest Laboratories
Richland, WA 99352
Abstract
The radioactive noble gas thoron (2WRn) was measured in the breath of dogs following intramuscular
injection or inhalation of 228Th enriched ThO2- With increased ventilation the amount of exhaled thoron
increased both in the injected dogs and in artificially ventilated lungs from dogs which had inhaled Th02- In
general, thoron was washed from the lung more efficiently by the normal breathing maneuvers of the intact
dogs than by artificial ventilation of their excised lungs. It was possible, however, to obtain marked variations
in the thoron output by imposing rather bizarre ventilation patterns on the artificially ventilated lungs,
INTRODUCTION
The radioactive gas thoron (220Rn) is a short-lived alpha emitting daughter product of the 232Th decay series.
Typical of the noble gases, thoron is retained to only a small extent in biological materials, and is excreted
from the body via the exhaled air. Since a fraction of the thoron produced in the body is exhaled, analysis of the
breath has long been of interest as a bioassay procedure for determining the amount of parent thorium
compound retained in man. Breath analysis is particularly attractive because of the ease of collecting sample's
of the breath, and the high degree of sensitivity for detecting alpha radiation by conventional counting
techniques. Breath analysis procedures are now available for detecting thoron in the range of 0.1 pCi per liter
of exhaled air (Hursh, et al., 1963). The detection limit for freshly inhaled natural thorium, based on breath
thoron analysis, has been estimated by Tai-Pow (1969) to be 0.001 fl Ci or only 0.1 of a maximum permissible
body burden.
It is of interest that past experience with human subjects administered Th02 (Thorotrast) has not shown a
close agreement between the amount of thoron exhaled and the body burden of thorium. The variations in the
thoron :thorium ratio are due in part to differences in the distribution of the thorium deposit as well as
differences in the metabolism of the 232-]^ decay products which alter the steady state amounts of the daughter
products leading to thoron production. For example selective excretion or translocation of the Ra isotopes in
the decay series 232Th^228Ra-^228Ac-»~228Tri*"224Ra-^220Rn (thoron) may significantly alter the amount of
thoron produced, and ultimately released to the breath from a fixed burden of 232Th. The ratio of breath thoron
to thorium burden varied by a factor of 10 in Thorotrast patients studied by Tai-Pow (1969). Results from
several laboratories compiled by Hursh (1967) show a variation of about 3-fold in the fraction of thoron
released by the breath of a variety of Thorotrast patients. The values of the ratio appeared to be influenced by
the anatomical site of the thorium deposit relative to the lung, by the age of the thorium deposit, and by
physiological parameters including the ventilation rate of the subjects. In general, the large range in
variability in the thoron rthorium ratio can be attributed to the short physical half-life of thoron (55.6 sec) and
the associated loss of thoron due to radioactive decay. As much as 90% of the thoron produced in the body may
be lost by decay in the interval between its production and expulsion in the exhaled air. Losses of this
magnitude are reported for predominantly nonpulmonary deposits of thorium, where thoron is produced
systemically, diffuses into the circulating blood, and is transported via the blood to the blood-air interface in
the lung. It is to be expected that breath thoron produced within the lung itself from inhaled thorium
compounds would be less subject to decay losses, and probably influenced less by ventilation and distribution
variables. Thus, measuring exhaled thoron may be particularly useful as a bioassay procedure for inhaled
thorium burdens. In order to test this supposition, breath thoron was measured in dogs or their artificially
ventilated excised lungs after inhalation of ThO2 aerosols. Ventilation was varied to simulate a wide range of
breathing patterns in lungs containing either freshly inhaled or well aged deposits
METHODS
The method of thorium exposure and measurement was described in detail in an earlier publication by
Ballou, et al., (1972). Briefly, anesthetized dogs were administered by injection or inhalation of a Th02dust
enriched in 228Th; the thoron in the breath was measured using a modification of the detector system developed
by Hursh, et al., (1963). In this procedure, breath thoron is adsorbed on a cold, activated charcoal surface, and
the alpha disintigrations are detected with a Nal crystal faced with a ZnS screen. The lungs which were
subjected to artificial ventilation were removed at necropsy in an expanded condition, and were maintained in
a humid atmosphere at 37°C in a plastic box (artificial thorax) about 1 ft3 in volume. Artificial ventilation of
the excised lung was accomplished by negative pressure breathing using a respiratory pump to vary the
*This paper is based on work performed under contract with the United States Atomic Energy Commission at
the University of Rochester Atomic Energy Project and has been assigned report number UR-3490-298.
-520-
-------
frequency and depth of the pressure excursions within the box. A second pump was employed to maintain the
desired minimum negative box pressure, and thus the minimum lung expansion at end expiration. Ventilation
frequency, minute volume and residual volume were varied during artificial ventilation. Thoron exhaled by
the intact dog or by the artificially ventilated lung was introduced into the detector system (Figure la,b) and
quantitated by comparison with a calibrated thoron source as described earlier by Ballou, et al, 1972).
RESULTS AND DISCUSSION
The data in Table 1 illustrates the stability of the detector system during a series of 13 consecutive
measurements using a aas-ph stearate thoron source prepared according to Hursh, et al., (1966). Breathing
through the thoron source was simulated using a respiratory pump. Since the dead space within the 228Th
stearate source was small (-^10 ml) compared to the tidal volumes employed, losses due to radioactive decay
were negligible and the washout of thoron produced within the source was complete for ventilation rates
ranging from 0.75 to 2.62 liters per minute. After each measurement, the detector system was purged with air
for 10 minutes to clear the system of thoron and allow time for the thoron to decay on the charcoal collecting
surface. The detector efficiency was maintained satisfactorily, during this comparatively long series of
consecutive measurements, as illustrated by the reproducibility of the source counts.
Also shown in Table 1 are data for two dogs (Dog M3 and Dog M4) injected intramuscularly with ThO2 one
month before exhaled thoron was measured. Again, the measurements were made consecutively on the same
day for a given dog. Increasing the ventilation rate of Dog M3 by pressing on the chest, increased the thoron
output by about 1/4. The observed increase in thoron output with increased ventilation is similar to that
reported by Tai-Pow (1969) for Thorotrast patients containing systemic burdens of ThO2. The response of Dog
M4 to changes in ventilation was ambiguous in that hyperventilation induced by breathing C02 brought
about an apparent decrease in exhaled thoron. Furthermore, although ventilation remained elevated about
50% after CO2 was replaced byC>2,the amount of thoron collected from the breath remained at the normal level.
These results suggested that high concentrations of CO2 in the air stream may have impaired the detector
efficiency. This was confirmed, also in Table 1, when the 228^ stearate source was purged with C02- The
source count was reduced roughly an order of magnitude when CO2 was employed. The detector did not
recover completely after the usual post-measurement purging with air indicating that a high level of CO2.
such as employed here, may interfere with thoron gas adsorption.
Preliminary studies with anesthetized dogs showed that ventilation changes could be fairly easily induced,
but they could not be controlled or sufficiently prolonged to permit satisfactory measurement of exhaled
thoron. Uniform breathing for a period of 6 minutes was required for an adequate measurement. For this
reason, an in vitro system employing excised lungs in an artificial thorax was investigated as a means of
varying ventilation under controlled conditions for extended periods of time.
The values in Table 2 compare thoron washout efficiencies for anesthetized dogs and their respective
excised lungs at various times after inhalation of Th02- Washout efficiency was calculated as the ratio of the
breath thoron (normalized to a ventilation of 1 1/min) and the 228-^ burden in the dog or isolated lung. Since
these dogs were administered ThO2 by inhalation, essentially all ThO2 was retained in the lung and
tracheobronchial lymph nodes. It should be made clear, however, that thoron is not produced directly by 228TH
decay, but rather is the daughter of 224Ra (Ti/2= 3.64 days) a decay product of 228Th. Therefore, 224Ra in tissues
other than lung contributed thoron to the dog's breath in excess of that produced solely from 224Ra in the lung
per se. In these animals, from 15 to 28% of the 224Ra was translocated from the site of 228Th deposition in the
lung. The major extrapulmonary sources of thoron, which contributed to the higher thoron washout efficiency
generally observed for the dogs in Table 2, was the gastrointestinal tract during the early post-exposure phase
of lung clearance and the skeleton at later time periods. The quality of ventilation of the excised lung, that is
the extent to which artificial ventilation reproduced ventilation of the lung by the intact dog, can not be
directly inferred from the washout efficiencies because of the extrapulmonary sources of thoron. It is
interesting, however, that although the lung was depleted as much as 28% in thoron precursor (224Ra)in some
cases, fairly close agreement in washout efficiencies was still obtained for the intact dogs and artificially
ventilated lungs in Table 2. It is apparent that the major part of the exhaled thoron must have originated from
224Ra in the lung in both the in vivo and in vitro preparations.
Probably the maximum ventilation effect on thoron washout to be expected is illustrated in Table 3 which
summarizes the effects of radical changes in the depth and frequency of breathing of the excised lungs.
Normal lung expansion, as judged by the size of the inflated lung at necropsy, was maintained by varying the
artificial thorax pressure over the range of -1 to -4.5 inches of water. The lung was respired initially at
approximately the same rate as the anesthetized dog, usually 12 to 15 breaths/minute; however, the dog's
breathing frequency varied considerably depending on the state of anesthesia.
By manipulating the base line negative pressure in the box at end expiration, it was possible to control the
residual volume in the lung, and investigate the effect of ventilating a fully expanded lung (lung 1), or a
partially filled lung (lung 2) shown in Table 3. In both cases thoron washout was less efficient than when the
lungs were ventilated within the normal pressure range. It seems clear that these exaggerated breathing
maneuvers did not effectively ventilate the thorium deposits, or more accurately, the thoron sources in the
lung parenchyma. It should be noted that the breath thoron recovered from Lung 1 (taken from Dog 1, Table 2)
was only about 1/5 of that expected, possibly because of losses due to leaks in the system.
-521-
-------
Increasing the frequency of respiration (lungs 3-6) was expected to yield maximum thoron washout since
total ventilation increased and losses due to radioactive decay in the lung were minimized by decreasing the
residence time of thoron in the lung. The total amount of thoron did generally increase as expected; however,
only marginal increases or slightly decreased efficiency was observed — apparently because the quality of
ventilation changed during rapid breathing. That is to say, the tidal volumes decreased at the higher
frequencies, indicating that the lung was not able to fill and exhaust to the fullest extent during the limited
time period between pressure changes within the aritficial thorax. Controlling the box pressure did not,
therefore, control ventilation when the physical capacity of the lung to respond was exceeded.
Slower breathing (also lungs 3-6) was expected to decrease thoron washout by decreasing total ventilation
and increasing thoron decay losses during the more prolonged breathing cycle. The expected decrease in
thoron washout was observed in all cases. Furthermore, washout efficiencies decreased during slow breathing
in all but one case (lung 3), and exhibited marginally higher values than rapid breathing in two cases (lungs 3
and 4). Although total ventilation decreased at the lower breathing frequencies the tidal volumes were
generally greater since the longer breathing cycle permitted maximum filling and exhausting of the lung
contents. Ventilation of the deep lung was undoubtedly improved by slow breathing as evidenced by the
greater tidal volumes. A corresponding increase in breath thoron was not seen, however, because the residence
time in lung was prolonged and more thoron was lost by decay.
CONCLUSIONS
The supposition that the amount of breath thoron released from inhaled thorium burdens would be
relatively independent of variations in ventilation was not supported by the results of this study. Rather to the
contrary, the artificially ventilated lungs which contained inhaled ThO2 released comparatively more thoron
in response to increased ventilation than either the dogs injected intramuscularly (Table 1) or the human
Thorotrast patients studied by Tai-Pow (1969). It is apparent in the latter two cases that the amount of thoron
exhaled in the breath was limited by the concentration of the radionuclide in the blood and the capacity of
thoron to diffuse from the blood into the lung air spaces. It is conceivable, therefore, that a significant increase
in ventilation could rapidly deplete the available thoron in the lung space and lower the thoron concentration
in the exhaled air. This effect was observed to some degree in both the injected dog M3 and in the Thorotrast
patients. The extreme case, illustrated by the thoron source in Table 1, demonstrates the complete lack of effect
of increased ventilation when thoron was produced at a constant rate, and the total amount was washed out by
less than maximum flow rates. The opposite relationship was observed in the lungs containing inhaled Th02-
Although the amount of thoron generally increased as breathing frequency increased, the washout efficiency
also increased, or remained relatively unchanged from the values obtained under more optimum breathing
conditions. This probably means that total washout of the thoron produced in the lung was not achieved
during artificial ventilation, and apparently was not achieved by at least two of the intact dogs shown in Table
2 (Dog 5 and Dog 8). This should be expected during normal breathing where only a fraction of the air in the
lung is replaced during each breathing cycle. It should be mentioned that the animals employed in this study
were anesthetized during thoron measurement and breathing was somewhat depressed from the normal
ventilation value of about 2 L/min for a 10 kg dog.
From the practical viewpoint of bioassay for thorium burdens, the increased washout of thoron with
increased ventilation has the advantage of increasing the sensitivity of thorium detection. It seems apparent
also that the volume breathed should be recorded since the concentration of thoron relates best to the thorium
burden. Assuming that the relationship between thoron concentration in the breath and the thorium lung
burden remains fairly constant as in Table 3 (expressed here as washout efficiency), there may be an
advantage in collecting exhaled thoron under maximum breathing conditions. This possibility is also
suggested by the work of Tai-Pow (1969), who found the relationship between breath thoron concentration and
thorium body burden to be less dependent on breathing rate than was the actual amount of thoron exhaled by
his thorotrast subjects.
REFERENCES
Ballou, J. E. and J. B. Hursh, (1972), The Measurement of Thoron in the Breath of Dogs Administered
Inhaled or Injected ThO2. Health Phys. 22:155-159.
Hursh, J. B. and A. I. Lovaas, (1963), A Device for Measurement of Thoron in the Breath, Health Phys.
9:621-627.
Hursh, J. B. And A. I. Lovaas, (1966), Preparation of a Dry 22« Th Source of Thoron. J. Inorg. Nucl. Chem.
29:599-600.
Hursh, J. B., (1967),Loss of Thorium Daughter by Thorium Dioxide Patients. Annal N Y Acad. Sc.
145:634-641.
Tai-Pow, J., (1969), Measurement of Thoron in the Breath of Thorotrast Subjects. Health Phys. 16:269-
L* /O.
-522-
-------
TABLE 1. Effect of Changes in Breathing Pattern on Exhaled Thoron.
Thoron Source
Exhaled
Ventilation Frequency Tidal Volume Thoron
(L/Min) Respiratipn/Min (ml) Counts/Min
228Th Stearate(a)
1.67 (4) +
2.62 (4)
0.76 (4)
1.67 (1)
Dog M3(b)
Normal Breathing 0.95
Chest Constricted 1.27
Normal Breathing 0.87
Dog M4(b)
Normal Breathing 0.89
Breathing CC-2 2.06
Breathing Q% 1.24
Normal Breathing 1.40
228Th Stearate(a)
Purged with Air 1.43
Purged with CO2 1.43
Purged with Air 1.43
13
13
13
13
12
20
11
9
25
10
16
13
13
13
128.5
202
58.5
128.5
79
63.5
79
99
82.5
124
87.5
110
110
110
3,606 ± 171
3,685 ± 125
3,330 + 240
3,790
3,973
3,766
2,786
11,081
3,964
10,417
12,137
3,686
445
3,192
(a) 228Th stearate was enclosed in a container with approximately 10 ml dead space as
described by Hursh, et al., (1966).
(b) Dogs were injected intramuscularly with ThC^ approximately 30 days before
measurements.
+Number of measurements in parenthesis.
-523-
-------
TABLE 2. Comparison of Normal and Artificial Ventilation
on Thoron Washout from the Lung.
Days After Ventilation Breath Thoron 22«Th Burden Washout Efficiency
ThO2 Inhalation (L/Min) (nCi) (nCi) (nCi Thoron/L/Min/nCi 226Th)
Dog 1
Excised Lung
Dog 2
Excised Lung
Dog 3
Excised Lung
Dog 4
Excised Lung
Dog 5
Excised Lung
Dog 6
Excised Lung
Dog?
Excised Lung
Dog 8
Excised Lung
3
3
7
7
30
90
150
150
1.49
0.68
1.82
1.88
0.98
1.11
1.22
1.15
1.53
0.90
1.33
1.44
1.09
1.64
1.61
1.61
(a) The burden for Dogs 1-8 excludes 228Th in pelt and
38.2
7.95
10.04
8.885
8.95
7.19
3.41
2.16
1.196
0.726
0.749
0.676
36.6
37.4
0.802
1.08
354(a)
352
101.3
101
98.6
96.4
28
27.6
13.7
13.6
7.2
7.2
362
305
6.97
6.93
0.0724
0.0332
0.0545
0.0465
0.0926
0.0672
0.100
0.0681
0.057
0.0594
0.078
0.0652
0.0928
0.0749
0.0715
0.097
gastrointestinal tract.
TABLE 3. Ventilation Effects on
Days After
Exposure
Lung 1 3
(352)(a)
Lung 2 150
(305
Lung 3 3
(101)
Lung 4 7
(27.6)
Lung 5 90
(7.2)
Lung 6 150
(6.93)
.
from the
Frequency
(Resp/Min)
12
20
12
13
12
13
13
31
42
6
13
13
18
20
7
12
20
5
12
40
13
6
37
-.0; 99snrri_ ,_ ii i-
Thoron Washout
Artificially Ventilated Lung.
Tidal Vol.
(ml)
56.6
40.5
137
94.8
114
108
144
70.5
55.7
204
141
88.5
65.5
61
84.4
120
76
234
124
40.3
85.5
128
39.2
Ventilation Box Pressure Breath Tn Washout Efficiency
(L/Min)
0.68
0.81
1.64
1.23
1.37
1.40
1.88
2.18
2.34
1.22
1.83
1.15
1.18
1.22
0.59
1.44
1.52
1.17
1.49
1.61
1.11
0.77
1.45
(inches Water)
-1.1 to -5.2
-5.0 -9.5
-1.1 -4.7
-1.6 -2.7
1.2 -5.7
-1.1 -4.6
-1.4 -4.0
-1.5 -4.2
-1.5 -4.2
-1.2 -3.8
-1.2 -3.8
-1.1 -4.4
-1.4 -4.6
-1.5 -4.9
-1.3 -4.8
-1.1 -4.1
-1.1 -4.2
-1.1 -4.2
-1.1 -4.2
-1.1 -4.2
-1.2 -4.2
-1.1 -4.2
-1.2 -4.2
(nCi)
7.95
3.86
37.4
16.6
40.2
33
8.85
11.9
11.53
6.29
8.8
2.16
1.72
1.84
0.896
0.676
0.735
0.443
0.574
1.08
0.652
0.407
1.24
(nCiTn/L/Min/nCi 22STh)
0.033
0.014
0.075
0.044
0.096
0.077
0.047
0.054
0.049
0.051
0.048
0.068
0.053
0.055
0.055
0.065
0.067
0.052
0.054
0.097
0.085
0.076
0.123
-524-
-------
Dog Mode
Lung Mode
Figure la. Thoron Measurement Apparatus. (See Figure Ib for numerical breakdown of above.'
-525-
-------
Figure Ib. Thoron Measurement Appratus.
1. To sealer
2. Thoron detector
3. To constant flow pump (4 1/min)
4. Dry ice-alcohol bath
5. Drying tube and CO2 absorbant
6. Balloon air reservoir
7. To respiratory pump
8. Thoron source (228Th stearate)
9. Pneumotachograph
10. Two-way Douglas valve
11. To pressure transducer and polygraph recorder
12. To spirometer air reservoir
13. Thermometer
14. Pressure gauge
15. Blood pressure transducer
16. To polygraph recorder
17. To pressure balance pump
18. To respiratory pump
-526-
-------
A REVIEW OF THE URANIUM MINER EXPERIENCE IN THE UNITED STATES
A.W.Hilberg
Division of Medical Sciences
National Academy of Sciences
National Research Council
Washington, D.C.
Abstract
Mining of uranium ores on a very small scale began in the United States in 1898. In 1948 mining of uranium
ores increased significantly as a result of the increasing demands of the atomic energy program, particularly,
the weapons program. It was well-known that radioactivity was associated with uranium ore and potential
health hazards had been identified as early as 1921. It was then suggested that the large number of lung
cancers found in the Schneeberg miners could possibly be caused by ionizing radiation.
Radiation exposure to the lungs of uranium ore miners comes about because of varying concentrations of
radon decay products in the mine atmosphere. Measurements of the amount of radioactivity in mine
atmospheres began in the United States in 1950 and between 1951 and through 1968, approximately 43,000
measurements were made in about 2,500 uranium mines.
Periodic medical surveys were made starting in 1950, but it was not until 1954, that a concerted effort was
made to examine all miners. At this time an epidemiological study was begun by the U.S. Public Health
Service to determine the effects of uranium mining with particular emphasis on the problem of lung cancer
induction.
The study of the miners, the results of which have been published, revealed a distinct and somewhat
dramatic increase in lung cancer. This increase has been ascribed to the radiation exposure to radon-
daughters. Because of the presence of a number of other potential occupational carcinogens in the dust of
underground mines there has been some question as to whether radon and radon-daughters constitute the
principal cause of increased lung cancer risk among these miners. However, the hypothesis that radiation
exposure at least was a major contributor to the excess of lung cancers found in the miners cannot be
disregarded and must be accepted pending more definitive data.
INTRODUCTION
In this review of the uranium miner experience in the United States, I will not dwell upon detailed statistical
analyses of data concerning numbers of cases or numbers of measurements, nor will I discuss histological
types of cancer. Instead, I will give enough data and, hopefully, enough background information, as well as
information as up-to-date as I could find, in order to allow for some interpretation and general thoughts about
what the uranium miner experience in the U. S. has been all about.
I will present no slides of charts, tables, or graphs. These may be found in great profusion in the several
reports published on the subject. I am greatly indebted to these reports for much of the material I will present
here. (See References Section.)
DISCUSSION
Pulmonary disease was noted among miners of ores containing uranium as early as 1597 by Agricola. The
mines were located in the Erg Mountains of Central Europe in the Joachimstal and Schneeberg mining areas.
With the beginning of modern pathological techniques in the 19th century, much of the pulmonary disease
unique to the miners was identified as lung cancer. This high occurence of lung cancer among these miners
was attributed to the high radiation levels in the mines by several authors (Ludwig, 1924; Hueper, 1942; Sikl,
1950; and Lundin, etal., 1971).
The mining of uranium ores on a very small scale began in the United States in 1898. In 1948, the mining of
uranium ores increased significantly as a result of the increasing demands of the atomic energy program —
particularly the weapons program.
In a period dating from July 1950 through September 1968, a total of 509 deaths occurred among
approximately 4,200 miners. In this group, there were 437 deaths among approximately 3,400 white miners
and 72 deaths among approximately 800 non-white miners. The expected deaths (using a general population
of the same area) would have been 277 among white and 90 among non-white. Thus, we note that the deaths in
the white miner population are about double the normally expected death rate, while there is no particular
difference in the non-white miner population. Certainly an increase of the magnitude of doubling a death rate
must stimulate an interest concerning the causes or uniqueness of the environmental factors of the mining
situation.
In 1949, a few radon measurements were made in the mines. Beginning in 1950, the U. S. Public Health
Service and the Colorado State Health Department began taking samples of radon in mines. The real
significance of radon-daughters in the atmosphere was not known at this time. Sampling of radon and radon-
daughters has continued through the subsequent years with attempts to sample the atmospheres of all mines.
Prom 1951 through 1968, approximately 43,000 measurements were made in about 2,500 uranium mines.
-527-
-------
Periodic medical surveys were made, starting in 1950, but it was not until 1954 that a concerted effort was
made to examine all miners. At this time, an epidemiological study was begun by the U. S. Public Health
Service to determine the effects of uranium mining with particular emphasis on the problem of lung cancer
induction. In 1971, approximately 150 cases of lung cancer were reported among uranium miners.
Because of the presence of a number of potential occupational carcinogens in the dust of underground
mines, there has been some question as to whether radon and radon-daughters constitute the principal cause
of increased risk among these miners. Pertinent to this issue is the fact that underground mining per se does
not necessarily lead to an increase of lung cancer risk, a fact that has been well documented for underground
coal miners in the United Kingdom. A recent study has investigated 5,500 potash miners in New Mexico,
working in mines not associated with elevated concentrations of radon-daughter products in the air, and has
shown no increased risk in such below-ground miners as compared with above-ground workers. (In both
groups excess cigarette smoking could account for the increased lung cancer compared to the general
population.) It is pertinent to point out that in those mining operations where a significant increase in
respiratory cancer has been associated with inhalation of radon and its daughter products, the mineral
constituents being mined were widely variable. Besides the uranium mines in Europe and the U.S.A.,
underground metal miners, fluorospar miners, and hematite miners have shown an excess respiratory cancer
risk. In each of these populations, there was occupational exposure to elevated concentrations of radon which
was also present in the mines. Thus, whether or not other agents such as arsenic, uranium, or fluoride may
have been present in the air, the one constant relationship in these groups has been radon-daughter exposure
and the incidence of lung cancer. In the early studies of the Bohemian pitchblende industry, some of the
employees in milling operations developed lung cancer, as did miners, but their exposures to radon and radon-
daughters, while probably significant, are not known with accuracy. In the U.S., uranium mill workers have
not experienced an increased risk of lung cancer (presumably because good ventilation minimizes their
exposure to radon-daughters).
Although most of the evidence relating radiation exposure to lung cancer in man pertains to internally
deposited alpha-emitting radionuclides, such as radon-daughters and thoron and its short-lived daughters,
there is some evidence of an excess lung cancer rate in individuals exposed to gamma- and x-radiation. Among
the survivors of the atomic bombing in Hiroshima and Nagasaki, data are now available for the period up to
1970, which show the relative risk of cancer of the tracheobronchial tree for the period of 1955 to 1970 to be 1.4
times higher for doses of 10 rads or more than for lower doses. Difficulties exist in interpreting these data,
however, one of which is the fact that the control group (i.e., those farthest from ground zero) the observed
overall cancer rate was 50% higher than that expected for the Japanese at large. In addition, there is the
question of neutron irradiation in the exposed individuals, which may have contributed significantly to the
observed effects in view of the possibly high RBE of this component of the total dose.
An approximately two-fold increase in the relative risk of lung cancer was observed in the study by Court-
Brown and Doll of patients with ankylosing spondylitis treated with x-ray therapy. In these cases, large
doses of x-rays were delivered to the spine, and doses to the bronchial epithelium were estimated to average
about 400 rads.
In a study of patients with tuberculosis, whether active or inactive, an increase in lung cancer of from 5- to
10-fold was found in comparison to the incidence in the general population. The possibility has been raised
that the patients may have been exposed to fluoroscopy during treatment of the disease, and that this may
account for their increased risk. In the absence of specific exposure information, however, and in view of the
fact that there could also be a relationship between tuberculosis itself and the likelihood of developing lung
cancer, little emphasis can be given to this study at present.
The incidence of lung cancer in x-ray technicians has been compared with that in pharmacy and medical
technicians in the U.S. military service during World War II. Out of approximately 13,000 individuals who
were present in both groups, 17 deaths from respiratory cancer were observed among the x-ray technicians as
compared with four among the other groups. This difference is highly significant, but when the groups were
compared with appropriate U.S. mortality statistics, a total of 12.4 cancers was expected from the x-ray
technicians, which was not significantly lower than the 17 cases observed. Thus, the difference between
groups may be due primarily to a decreased lung cancer incidence among the pharmacy and medical
technicians, which is paradoxical and complicates interpretation of the data.
The various studies just cited tend to show that no matter what the source of radiation may be, there is
circumstantial evidence of an increased risk of development of lung cancer from radiation exposure.
The experience through 1969 for all of the various underground mining groups in which an increased risk of
cancer has been found, was summarized by Lundin, Wagoner, and Archer. Central to an interpretation of data
from underground miners are a number of fundamental issues, which include the following:
(1) What exposures to radon-daughters have actually occurred?
(2) What is the rad dose to the critical cells from radon-daughters in the air?
(3) Is an increased risk observed at a dose rate below that equivalent to continuous occupational exposure to
one working level of radon and radon-daughters?
(4) Is the dose-response curve at low doses linear, is it concave downward (i.e., giving a higher risk per rad at
lower cumulative doses than at higher cumulative doses), or is there a true threshold for cancer production
from a cumulative dose?
-528-
-------
Considerable effort has been made to evaluate the radiation exposures of the various groups of miners in the
Colorado Plateau area, with particular emphasis on previous underground mining experience not in eluded in
the category of uranium mining (a substantial number of the miners had such experience). Absent or
infrequent samplings of air of some of the mines, especially in the early exposures prior to 1950, make
estimates of cumulative dose only approximate at best; but it is unlikely that these estimates can be greatly
improved at this time, and it is probable that in the aggregate the estimates of exposure are adequate to
determine trends in the data. It should be emphasized that among these miners the dose rate was quite high in
comparison to that in some of the other mining groups (about 10 working levels on the average). One "working
level" (WL) in air is defined as any combination of short-lived radon-daughters (through Po-214, RaC') leading
to total emission of 1.3 x 105 MeV of alpha energy per liter; the cumulative measurement of working level
month (WLM) is defined as exposure at the rate of 1 WL for 170 hours. There has been criticism of the WL as an
exposure index, because the state of equilibrium of the various nuclides in the chain is critical, especially with
regard to the fraction present as free ions. This latter criticism remains valid, but it is fair to say that samples
of mine air usually show relatively little contribution of unbound radon-daughters.
Estimates of the rad dose per WLM for basal cell layers of different segments of the bronchial epithelium
have varied widely, from less than 0.1 rad per WLM to as much as 20 rad per WLM. A critical factor in these
estimates is the thickness of the epithelial and mucous layers, an uncertain quantity in smokers with some
degree of chronic bronchitis. The unpublished studies of Gastineau indicate that the normal epithelium of
segmental and more proximal bronchi, where most radiogenic cancers have arisen, is thicker than had
previously been assumed.
On the basis of the present evidence, 1 rad per WLM is probably close to the upper limit for a reasonably
uniform dose to the basal cell layer of the epithelium of the larger bronchi on a probabilistic basis. In the
presence of existing chronic bronchitis, the dose factor may well be substantially lower, owing to increased
thickness of the mucous layer as well as of the epithelium.
So far as a limiting dose rate is concerned, the question is whether continuous exposure to less than 1 WL
has been found in miners to be associated with increase in lung cancer risk. The problem is related to the
possible influence of dose rate in latent period, and if latent periods of 20 to 30 years are found at the lowest
exposures, no mining group has been under observation with known exposures at these levels for a long
enough time to provide a definitive answer. The metal miners studied showed a cancer rate about three times
that expected, with exposures at the time of the study well below a concentration of 1 WL, but this study
indicates that earlier exposures before the mines were ventilated may well have been higher. The hematite
miners studied, who have shown a risk of about 1.7 compared with controls, worked in mines where the radon
concentrations are equivalent to WL concentrations of 1 WL or less, but until measurements of actual radon-
daughter exposures and the influence of the hematite itself are determined, no final conclusion is possible. For
the Colorado Plateau uranium miners in the lowest cumulative WLM exposure category whose dosage was
usually received from several short periods of high working level exposures, no significant excess of cancer
has appeared as yet. At present, the fact that the lowest exposure group shows only a slight increase in cancer
rate above that expected makes the Colorado Plateau group inadequate to resolve this issue of the presence of
a true threshold. Inspection of the composition of the study population indicates that the population at risk in
this dose range (120 WLM) is now so small as to make it unlikely that even future follow-up will settle the
matter.
There has been observed in the U.S. Colorado Plateau workers an inverse relationship between cumulative
radiation dose and the latent period for cancer after initial exposure in the mines, but this effect is not very
striking at the present time. The relationship of cigarette smoking to the latency period for lung cancer among
uranium miners is not known.
A large body of experimental work has now been assembled relating the occurrence of lung cancer to
ionizing radiation in animals. Although lung tumors are readily induced in animals by radiation exposure,
not all of these may be relevant to the human disease, since peripheral adenocarcinomas are much more likely
to occur in animals from whatever inciting stimulus is applied than are tumors comparable to squamous cell
tumors in man.
For alpha-emitters, the lowest cumulative dose at which a rise in lung cancer has been observed
experimentally was in rats given polonium-210 with a sodium chloride aerosol by inhalation. In this
experiment, one squamous cell cancer occurred after 70 rads cumulative mean lung dose, and the dose-
incidence relation was approximately linear at higher doses. For beta-emitters, the lowest dose associated
with cancer induction was approximately 600 rads, in rats given cerium-144 salts by intratracheal
installation. In these experiments, the dose-response curve appeared to be curvilinear (concave upward). An
inherent difficulty in animal experiments, of course, is the short life span of the small rodents usually used;
and, thus, the fact that only the cancers with short latent periods may be detected by this approach, a
limitation which might be expected to produce a curvilinear dose-response curve of the kind observed.
A recent study reported in April 1973 concerning the effects in hamsters exposed to simulated uranium mine
atmospheres showed invasive squamous cell carcinoma in a few hamsters receiving 6-hour daily exposure to
1,200 WL of radon-daughters with and without uranium ore dust for periods of 15 to 17 months.
The extensive studies of the U.S. uranium miners have attempted to establish as accurately as possible the
exposure of each miner expressed in terms of WLM (working level month) defined as exposure to one working
level during 170 hours each month, which corresponds to a probable dose of 1-2 rads to the bronchial
epithelium.
529-
-------
The mortality experience of the white underground miners from 1950 to 1968 compared with that expected
in the population of the four states shows an excess number of deaths (60 percent above expectation) —
essentially due to a larger than expected numbers of violent causes (by 145 percent), and of lung cancers (by
almost 500 percent). The mortality experience of the much smaller group of non-white miners over the same
period is insufficient (72 deaths in all) to be informative.
The grouping of lung cancer cases in the uranium mines have generally been expressed in terms of WLM,
that is: less than 120 WLM, 120-359 WLM, 360-839 WLM, 840-1,799 WLM, 1,800-3,719 WLM and greater than
3,720 WLM. There is an excess number of lung cancer cases in the uranium miners in all of the categories
above 120 WLM.
The major uncertainty within the studies lies in the assessment of the exposure to the individual miners.
This for the most part is due to the fact that large numbers of very small mines were operating at any one time.
There were 450 mines employing an average of 2 miners in 1950,850 mines with 3 miners in 1957, and 533 with
an average of 5 miners in 1966. While the quality of the measurements made in the mines was considered to be
good, the frequency of measurements was very unevenly distributed. In only, 5 mines were more than 5 radon-
daughter measurements made in 1950.177 mines had measurements in 1962 and 110 in 1968. In many mines
only one or two measurements were ever made. Thus, measurements often had to be inferred from
circumstantial evidence, since actual measurements were not available. Since the amount of radon-daughters
in air depends on many variables, including ventilation, meteorological conditions, and the quality of the ore,
it is not possible to evaluate the errors that may have been involved in assessing the exposure of any
particular miner.
An additional difficulty in interpreting the results arises from the fact that most of the miners included in
the study were cigarette smokers. This difficulty can to some extent be circumvented by comparing the
mortality in the miners with that in the population of the four states adjusted according to smoking habits as
well as according to the factors mentioned previously. While the excess mortality over the expected mortality
adjusted for smoking was somewhat reduced, the relation of the excess to the exposure remained basically
unchanged.
The studies of the uranium miner experience in the U.S. show a definite increase in lung cancer risk in
miners exposed to 120 working level months and above. They also demonstrate an over-all increase in
mortality from all causes. The increases in mortality and in lung cancer risk have been demonstrated by a
variety of analytic methods. Cigarette smoking, while possibly a factor in increasing the hazards of lung
cancer, does not account for the excess lung cancer found in uranium miners. Approximately 78 percent of the
miners were smokers, 22 percent non-smokers. Well over 90 percent of lung cancer deaths occurred in smokers.
This would tend to indicate that cigarette smokers are particularly susceptible to lung cancer when exposed
to the additional hazard of radiation exposure in the mines.
Uncertainties remain in an over-all evaluation of the uranium miners experience. Primarily, these are
uncertainties concerned with physical measurements of mine air, the physical and physiologic processes in
the lung, the possible interaction among several potentially harmful contaminents, such as radiation, diesel
fuel, cigarette smoking, other ore dusts, and the unknown specificity of the causality of cancer in man.
In spite of these uncertainties, the evidence gained in the evaluation of the uranium miner experience in the
U.S. leads to a summary of this report by quoting from the conclusions of the August 1968 Report of an
Advisory Committee of the Division of Medical Sciences, National Academy of Sciences, National Research
Council, as follows:
(1) "There appears to be a causal association between lung cancer and exposures of approximately 1,000
cumulative working level months (CWLM) and higher."
(2) "There is a statistically significant increase in the lung cancer risk for miners with approximately 100 to
400 cumulative working level months exposure."
(3) "The hypothesis is favored that radiation exposure at least contributed to the excess lung cancer
observed in the miners in the 100 to 400 CWLM category."
I believe that with the evidence at hand there can be little doubt that exposure to radon and radon-daughters
in the environment of a uranium mine certainly contributes to an increased risk of lung cancer induction after
an unknown period of relatively constant exposure.
REFERENCES
The following references were used in the preparation of the paper. However, no specific reference points
appear within the body of the paper.
An Assessment of the Economic Effects of Radiation Exposure Standards for Uranium Miners, (1970),
Report to the Federal Radiation Council, Arthur D. Little, Inc.
Effects on Populations of Exposure to Low Levels of Ionizing Radiation, The (1972), Report of the Advisory
Committee on the Biological Effects of Ionizing Radiations, Division of Medical Sciences, National Academy
of Sciences — National Research Council.
Epidemiologic Studies of Uranium Miners, (1971), Report of the Ad Hoc Committee, Advisory Committee to
the Federal Radiation Council. National Academy of Sciences — National Research Council.
-530-
-------
Guidance for the Control of Radiation Hazards in Uranium Mining, (1967), Report No. 8, Revised. Staff
Report of the Federal Radiation Council.
Ionizing Radiation: Levels and Effects, (1972), A Report of the United Nations Scientific Committee on the
Effects of Atomic Radiation to the General Assembly, with Annexes. Volume II: Effects. United Nations 1972.
Late Somatic Effects of Ionizing Radiation, (1968), C. D. Van Cleave, School of Medicine, University of
North Carolina, Chapel Hill, North Carolina.
Lundin, F. E., Jr., J. K. Wagoner, and V. E. Archer, (1971), Radon-Daughter Exposure and
Respiratory Cancer Quantitative and Temporal Aspects, National Institute of Occupational Safety and
Health, National Institute of Environmental Health Sciences. Joint Monograph No. 1. U. S. Department of
Health, Education, and Welfare, Public Health Service.
Pacific Northwest Laboratory Annual Report for 1972 to the USAEC Division of Biological and
Environmental Research, (1973), Volume I, Life Sciences, Part 1, Biological Sciences, Batelle Pacific
Northwest Laboratories, Richland, Washington 99352.
Radiation Exposure of Uranium Miners, (1967), Summary Analysis of Hearings before the Subcommittee
on Research, Development, and Radiation of the Joint Committee on Atomic Energy, Congress of the United
States.
Radiation Exposure of Uranium Miners, (1968), A Report of an Advisory Committee from the Division of
Medical Sciences:National Academy of Sciences — National Research Council — National Academy of
Engineering, Washington, D.C.
Saccamanno, G., V. E. Archer, O. Auerbach, M. Kuschner, R. P. Saunders, and M. G. Klein,
(1971), Histologic Types of Lung Cancer Among Uranium Miners. Cancer, March 1971, Vol. 27.
-531-
-------
RADIOLOGICAL HEALTH SIGNIFICANCE OF RADON IN NATURAL GAS
R.H. Johnson, Jr., D.E. Bernhardt, N.S. Nelson, and H.W. Galley, Jr.
Office of Radiation Programs
Environmental Protection Agency
Washington,B.C. 20460
Abstract
Natural gas contains varying amounts of radon-222 which becomes dispersed within homes when natural
gas if used in i/nvented appliances, such as kitchen ranges and space heaters. The average dose equivalent to
the bronchial epithelium of an individual from use of natural gas in these appliances was estimated for typical
exposure conditions to be 0.015 and 0.054 rem per year respectively. When extrapolated to the United States
population at risk, the estimated population tracheobronchial dose equivalents were 1,874,000 and 854,000
person rems per .v< 'o r for these appliances, or a total of 2,728,000person-remsper year.
A rciicu- of exposure conditions, lung model parameters, dose conversion factors, and health effects data
indicated this population dose could possibly result in up to 15 deaths per year. For perspective, the
concentrations and health effect consequences from use of natural gas containing radon were compared to
current radon guides, natural background radon, and normal lung cancer mortality statistics. These
comparisons indicated that natural gas radon represents about 0.9 to 1.5 percent of the guide of 0.33 pCi/l
derived from ICRPNo. 2,0.3 to 3 percent of normal radon background, and 0.03 to 0.08 percent of normal lung
cancer mortality. It was concluded that radon in natural gas does not contribute significantly to respiratory
cancer mortality in the United States.
INTRODUCTION
Radon-222 is a radioactive gaseous daughter product of radium-226 found in naturally occurring uranium
minerals throught the earth's crust. This heavy inert gas permeates porous geological formations, and is
collected along with methane in production wells for natural gas. When this natural gas is used in unvented
appliances, such as kitchen ranges and space heaters, a part of the combustion products and radon-222 are
released within the home. This radon-222 constitutes an additional source of radiation in the home which has
not been adequately evaluated for potential health effects.
In this assessment, potential health effects were derived following the sequential analysis outlined in
Figure 1. This figure shows a generalized model and some of the factors which must be considered. Each of the
four main phases of this model, radon concentrations, exposure conditions, dose calculations, and health
effects estimates, have been reviewed in depth in a study by Johnson, et al., (1973). This present paper
constitutes a summary of some of the major findings of that study.
RADON CONCENTRATIONS IN NATURAL GAS
The average radon level in several natural gas production areas in the United States is 37 pCi/l as indicated
in Table 1. However, from the range of concentrations shown in this table, it is apparent that many individual
wells have levels 10 to 20 times the average value. There is also a considerable variation in radon
concentrations at wells in different well fields ranging from a low of 5 pCi/l in the Gulf Coast to about 100
pCi/l in Texas, Kansas, Oklahoma, and California. These variations may be attributed to differences in
regional distributions of uranium and individual well parameters such as location, depth, pressure, and
production rate. It should also be noted that the average concentration calculated here does not account for the
relative gas production volumes from different regions of the country. For example, the states of Texas,
Louisiana, and Oklahoma produce 81 percent of the natural gas in the United States.
Radon levels at points of use have not been measured extensively, but some data are available for gas
distribution lines (Table 2). These data indicate that average radon levels in various areas are 50 pCi/l or less
with an overall average of 23 pCi/l. The highest levels were found in the Colorado and New Mexico areas
where the transit times from production wells to consumers are short compared to the half-life for radon-222 of
3.83 days. The coastal regions furthest from natural gas sources have the lowest radon levels. This could be
due to pipeline transmission time and storage which allows radon-222 to decay away. Also most gas from
production wells is processed to remove water, liquifiable hydrocarbons, and other gaseous products such as
CO2 and helium. This processing removes 30 to 75 percent of the radon with the liquid hydrocarbons
(primarily propane).
-532-
-------
RADON FROM USE OF NATURAL GAS IN HOMES
Radon concentrations in homes resulting from use of natural gas have not been measured; however,
estimates may be made by determining the quantity of gas used, venting of gas appliances, house volume, and
number of air changes per hour. The quantity of gas used, and the fraction of combustion products vented
inside the home, depend on how the gas is used; i.e., for comfort heating, or ranges, water heaters,
refrigerators, clothes dryers, etc. Most natural gas is used for comfort heating, but gas furnaces are normally
vented outside the home. Conversely, gas kitchen ranges are normally vented into kitchen areas, and initial
studies by Barton, et al., (1970) indicate this is the main source of radon-222 from natural gas usage in homes.
In addition, data compiled by T. F. Gesell (1973) indicate that there is also widespread use of unvented space
heaters.
The use of these unvented appliances results in the dispersion of radon-222 along with the combustion
products of natural gas within the home. The radon-222 decays to a radon-daughter product mixture which
reaches an equilibrium concentration as a function of radioactive decay, dispersion, and removal processes.
Tests by F. B. Johns at the National Environmental Research Center, Las Vegas, indicate that effectively
there are no daughter products present in natural gas at the point of combustion because radon-daughters
tend to plate out on any obstructing surfaces such as valves and walls of pipelines. Similar removal processes,
and dilution by ventilation and mixing with outside air, reduce the daughter product concentrations in the
home atmosphere below those which would be calculated for radioactive secular equilibrium with radon. A
fraction of each daughter product also becomes attached to dust particles in the air. For specific aerosol
conditions, the proportion of ions which remain free, or uncombined, reach a steady state for each decay
product.
Using the exposure conditions postulated in Table 3, the following average radon-222 levels were calculated
for use of natural gas in a typical home:
Unvented gas kitchen ranges - 0.028 pCi/1
Unvented gas space heaters 0.010 pCi/1
CRITICAL MODE OF EXPOSURE AND RADON-222 DOSIMETRY
The primary hazard for radon-222 exposure is from inhalation and bronchial deposition of the alpha-
emitting daughters 218Po and 214Po, commonly referred to as RaA and RaC', which release their alpha decay
energy into tissues of the respiratory system. This was evidenced by uranium miners working in radon
atmospheres who developed lung cancers believed to have occurred as a result of interaction of alpha particles
with the basal layer cells of the bronchial epithelium (Lundin, et al., 1971).
The dose to the critical tissue is estimated by the use of lung models which include functions of
environmental conditions, anatomy, respiratory physiology, and radon-222 dosimetry. Some of the more
important variables include the particle sizes and aerosol content of the atmosphere, home air change rate,
free ion fraction, rate and depth of respiration, medical status of the individual, and the location of the
precancerous cells at risk.
Variations in choice of lung model parameters and exposure conditions have led to a wide range of factors
for converting radon concentrations to dose. A summary of these dose conversion factors (Johnson, etal.,
1973) shows a range from 12 to 140 rads per year for continuous exposure at one working level (WL). One WL is
equivalent to 1.3 x 105 MeV of potential alpha energy per liter of air which would be available from radon-
daughters (RaA through RaC') in secular equilibrium with radon-222 at a concentration of 100 pCi/1 (Lundin,
et al., 1971). The conversion factor most closely representing exposure conditions expected in typical
dwellings is 100 rads per year for continuous exposure at one WL.
DOSE TO AN INDIVIDUAL
The dose to a hypothetical individual exposed to radon and daughters from use of natural gas in unvented
kitchen ranges was calculated by Barton, etal., (1973) for the following set of exposure conditions (see also
Table 3): a kitchen range gas usage of 0.765 m3 per day; a 226.6 m3 house; an air change rate of one per hour; no
radon-daughters in the gas; the ratio of Rn, RaA, RaB, RaC (RaC') is 1,0.8, 0.6,0.4; free daughter ions of 8.5
percent; a dose conversion factor of 100 rads per WL; and a quality factor of 10. These conditions gave a dose of
0.0015 rads (dose equivalent of 0.015 rems) per person per year to the tracheobronchial epithelium due to
inhalation of radon-daughters. Radon-222 itself contributes only about 0.5 percent of the dose for inhalation of
radon and daughters together.
POPULATION DOSE EQUIVALENT
The average dose equivalent to the United States population was determined by extrapolation of the dose to
a hypothetical individual. The population at risk was determined for each state from data compiled by Gesell
(1973) from 1970 Census Bureau statistics. These data gave the number of dwellings with gas kitchen ranges
-533-
-------
and unvented space heaters. The population at risk from the use of unvented kitchen ranges was determined
by assuming four persons per dwelling; i.e., 4 x 31.2 x 106 = 124.8 million persons. The population
tracheobronchial (T-B) dose in person-rems was then calculated by multiplying the population at risk times
the individual dose equivalent. This gave 1.87 million person-rems per year from the use of gas kitchen ranges
in the United States.
The population dose equivalent from use of unvented space heaters was determined by relating the average
quantity of gas used in heaters to the quantity of gas used in ranges and the corresponding dose equivalent for
ranges. The quantity of gas used by space heaters was calculated as 0.354 m3 per degree-day for the average
annual degree-days per year for each state. The average quantity of gas used by space heaters was 2.75 m3 per
day, compared to 0.765 m3 per day for gas kitchen ranges. The average dose equivalent to an individual from
use of space heaters was 0.054 rem per year. When extrapolated to the exposed population (15.8 million
persons) the population T-B dose from space heaters was estimated as 0.854 million person-rems per year.
The total population T-B dose equivalent for exposure to radon-daughters from the use of natural gas in
unvented kitchen ranges and space heaters was estimated as 2.73 million person-rems per year.
HEALTH EFFECTS ANALYSIS
This analysis was based on the absolute somatic and genetic risks from inhalation of radon-daughters as
outlined in the report by the National Academy of Science on the biological effects of ionizing radiation (BEIH
Keport, 1972). The T-B dose effect of concern is lung carcinoma. Since lung cancer has such a high mortality
rate, it is assumed that morbidity for this dose effect is equivalent to mortality.
The absolute risk from T-B dose was calculated from Table 3-2 of the BEIR report (1972) by multiplying the
sum of the fractional risks by age times the expected plateau region. In this case, the plateau region, or time
beyond the latent period during which the risk remains elevated, was taken as 30 years. The calculation for
risk in terms of lung cancer deaths was based on the following analysis, where the adult risk for cancer of the
lung from T-B dose is 1.3 deaths/million persons at risk/year/rem.
Age
Group
Percent of
Population
Proportion of
Adult Risk
Fractional Risk
deaths/106 persons/year/rem
10+
0-9
In Utero
80
20
1.3
1
0.2
5
1.3 x 1 x
1.3 x 0.2 x
0.8 = 1.04
0.2 = 0.05
1.3 x 5 x 0.013 = 0.08
1.17
Annual Risk = 1.17 deaths/million persons/year/rem
This annual risk is then multiplied by 30 to estimate excess deaths for a plateau region of 30 years to give the
absolute risk of 35 excess deaths/million persons/year/rem.
The potential health effects to the United States population at risk from use of natural gas containing radon
in unvented appliances is determined by applying the absolute risk to the estimate of total population T-B
dose equivalent. This gives an estimate of 95 excess deaths per year. However, the significance of this estimate
should only be interpreted by comparison with other reference guides and after consideration for the
uncertainties in this analysis.
DISCUSSION
The fundamental problem in an analysis of potential health effects as derived in this study is the necessity
of extrapolating from a few measurements or reported values to average conditions for large populations.
Because of inadequate information, values had to be estimated or assumptions made to represent typical
exposure conditions or population at risk. The assumed values were normally selected so the calculated dose
or health effects would be overestimated, i.e., conservative.
The possible variations of exposure conditions which could reasonably affect a large fraction of the
population at risk were given previously in Table 3. The nature of corrections to adjust the estimated health
effects for these possible variations in exposure conditions are listed in Table 4.
A review of the uncertainties in this analysis, and the significance of corrections for possible variations in
exposure conditions (Johnson, et al, 1973), indicate that the estimate of 95 excess deaths a year is probably
conservative by up to a factor of 10. Therefore, a less conservative, but more realistic, estimate would be a
1 Exposed through placental transfer of radioactivity in maternal blood.
-534-
-------
possibility of 15 or less potential deaths per year from use of natural gas containing radon in unvented
appliances. It should be further noted, however, that this estimate is based on a linear non-threshold dose
response, and that no lung carcinomas have been confirmed at radon levels which could occur in homes using
natural gas.
For further perspective as to the significance of radon in natural gas, the radon-222 concentrations and
health effect consequences were compared to natural background radon, current radon guides, and normal
lung cancer mortality statistics. It was determined that the use of natural gas containing radon-222 could
increase the normal level of radon-222 in homes by 0.23 to 2.2 percent. These radon levels from natural gas
represent about 0.9 to 1.5 percent of the guide of 0.33 pCi/1 derived from ICRP No. 2 (I960). Also, an excess
mortality of 95 deaths a year would be only 0.2 to 0.5 percent of normal annual respiratory cancer mortality.
The more realistic estimate of 15 or less deaths a year would be only 0.03 to 0.08 percent of normal respiratory
mortality.
It can be concluded from this assessment of potential health effects, that radon in natural gas does not
contribute significantly to respiratory cancer mortality in the United States.
REFERENCES
Altshuler,B., N. Nelson, andM. Kuschner, (1964), Estimation of Lung Tissue Dose from the Inhalation
of Radon and Doubters, Health Physics, 10-1137-1161.
Barton, C. J. (1971), Radon in Air, Natural Gas, and House: A Preliminary Survey and Evaluation,
ORNL-CF-71-5-48.
Barton, C. J., R. E. Moore, and P. S. Rohwer, (1973), Contribution of Radon in Natural Gas to the
Natural Radioactivity Dose inHomes, ORNL-TM 4154.
Bunce, L. A., and F. W. Battler, (1966), Radon-222 in Natural Gas, Radiological Health Data and
Fries, B. A., and K. H. Kilgren, (1972), Radon in Natural Gas and Gas Liquids, Chevron Research
Company, California.
Gessel, T. F., (1973), Unpublished Data, University of Texas, School of Public Health, Houston, Texas.
International Commission on Radiological Protection, (ICRP) (1960), Publication 2, Pergamon
Press, London.
Jacobi, W., (1972a), Activity and Potential a-Energy of Radon-222 and Radon-220 Daughters in Different
A ir A Imosph ercs, Health Physics, 22, 441-45.
Jacobi, (1972b), Relations Between the Inhaled Potential a-Energy ofRn-222 and Rn-220 Daughters and
the Absorbed -Energy in the Bronchial and Pulmonary Region, Health Physics, 23, 3-11.
Johnson, R. H., Jr., D. E. Bernhardt, N. S. Nelson, and H. W. Calley, Jr., (1973), Assessment of
Potential Radiological Health Effects from Radon in Natural Gas, Environmental Protection Agency
Document, EPA-520/1-73-004, Office of Radiation Programs, Waterside Mall, Washington, B.C. 20460.
Lundin, F. E., Jr., J. K. Wagoner, and V. E. Archer, (1971), Radon Daughter Exposure and
Respiratory Cancer Quantitative and Temporal Aspects, National Institute for Occupational Safety and
Health, National Sciences, Joint Monograph No. 1, PB-204871 , (NIOSH-M-71-1).
McBride, J. R., and D. H. Hill, (1969), Off-Site Radiological Surveillance for Project Gasbuggy, June
1967-July 1968,RadiologicalHealthDataandReports,535-546.
Report of the Advisory Committee on the Biological Effects of Ionizing Radiations (BEIR Report 1972),
Division of Medical Sciences, National Academy of Sciences, National Research Council, Washington, D. C.,
20006, The Effects on Populations of Exposure to Low Levels of Ionizing Radiation.
Tsivoglou, E. C., H. E. Ayer, and D. A. Holaday (1953), Occurrence of Nonequilibrium Atmosphere
Mixtures of Radon and Its Daughter, Nucleonics, 11, 40-45.
-535-
-------
Table 1. Radon-222 Concentrations in Natural G» « at Production Wells.(a)
Radon-222, pCi/1
Area
Colorado
New Mexico
Texas, Kansas
Oklahoma
Texas Panhandle
Colorado
Project Gasbuggy Areas
California
Gulf Coast
Kansas
Wyoming
Overall Average
<100
25.4
15.8-29.4
5
100
10
37
5-1450
10-520
11-45
12-59
1-100
(a) Johnson, et al, (1973); Barton, et al, (1971); Bunce, et al,(1966); and Fries, et al., (1972).
Table 2. Radon-222 Concentrations in Natural Gas Distribution Lines.(a)
Radon-222, pCi/1
Area
Chicago
New York City
Denver
West Coast
Colorado
Nevada
New Mexico
Houston
Overall Average 23
(a) Barton, et al., (1973); McBride, et al., (1969).
Average
14.4
1.5
50.5
15
25
8
45
8
Range
2.3-31.3
0.5-3.8
1.2-119
1-100
6.5-43
5.8-10.4
10-53
1.4-14.3
-536-
-------
Table 3. Exposure Conditions and Possible Variation in Parameters.
Parameter
Condition for this(a)
Analysis
Possible Variation(b)
Radon Concentration
at Point of Use
Gas Appliances
Gas Use:
Ranges
Heaters
Degree-Days
Appliance Venting
House Size
Air Change Rate
Radon-Daughters:
in Gas
in Home
Percent Free RaA
Critical Mode of
Exposure
Critical Organ
Dose Conversion Factor(d)
Quality Factor
20 pCi/1
Cooking Ranges
Space Heaters
0.765 mVday
0.354 m3/degree-day
Average for E ach
State
Unvented
226.6m3
One per Hour
No Daughters
1,0.8,0.6,0.4(0)
8.5 Percent
Inhalation of Radon-
Daughters
Bronchial Epithelium
lOOrads/yearfor
Continuous Exposure
atlWL(100pCi/l)
10
10-lOOpCi/l
Could include
Refrigerators, Clothes
Dryers, etc.
Uptol.l9m3/day
0.28-0.42 mVdegree-day
± 25% within States
Ranges could be
Partly Vented
142-425 m3
0.25 -5 per Hour
1,1, 1,1 to
1.0, 0.5, 0.25, 0.l(c)
5 to 25 Percent
Radon Alone gives
< 1% of Dose
Some Exposure Also to
Nasopharynx, Lung,
and Whole Body
50-125 rads/year
3-10
(a) These are intended to be typical average conditions — although some of the less well understood
parameters were chosen to give a higher (or more conservative) dose estimate.
(b) These are reasonable variations which could be encountered for a large fraction of the exposure
conditions or population at risk.
(c) Ratio of Rn, RaA, RaB,RaC(RaC'). . .. , , , ,
(d) This factor includes assumptions for daughter equilibria, critical mode of exposure, lung model, ana
other dosimetry factors.
-537-
-------
Table 4. Corrections to Adjust Estimated Health Effects for Different Exposure Conditions.
Parameter Correction Multiplier
Air Changes per Hour (Barton, etal., 1973) 6.01
0.25 6.01
1.0 .1.0
2.0 0.339
Radon Activity Linear(a)
Quantity of Gas Used Linear
House Size Linear
Daughter Equilibria (Tsivoglou, et al., 1953)
Ratio 1,1,1,1 1.9
1,0.9,0.8,0.7 1.3
1,0.8,0.6,0.4 1.0
1,0.75,0.5,0.3 0.84
1,0.5,0.25,0.1 0.39
Percent Unattached RoA(b)
3 0.75
8.5 1.0
10 1.3
25 2
Dose Conversion Factor Linear
Quality Factor Linear
Health Effects Conversion Factor Linear
(a) A linear correction means the correction is proportional to the variation in the parameter.
(b) Estimated from Jacobi (1972 a, b) and Altshuler, et al., (1964).
-538-
-------
Ol
CO
50
2Rn concentration
Well location,
depth and
seasonal
pressure
Gas use, heating, cooking
etc.
Use
I
i
, ' Radon dosimetry
i
rate, venting, dilution ' Critical mode of
volume,
222Rn concentration
1 exposure
1
! Critical organ
Daughter product equilibria, ' Population statistics
free
rate
ion fraction, ventilation 1
, aerosol properties, | Geographical gas use
variations dispersion and removal processes ,
production
rate,
Natural
qas
^w-»
wells
t
i
1
Storage ^
Transport
time
I
i
Home use
of
natural gas
i
1
i
1
i
1
Dose
equivalent
person-rem
Morbidity
Mortality
Health
S 0 U R C
TERM
O
Gas processing
and distribution,
mixing from
different well
fields
i
I
Radon concentration
to dose conversion
factors
i
Dose equivalent to
health effects
conversion factors
EXPOSURE
CONDITION
OPULATION
EXPOSURE
©
0
HEALTH
EFFECTS
0
Figure 1. Model for estimating potential health effects from radon in natural gas.
-------
VIII. Application of Noble Gases
-------
UTILIZATION OF THE NOBLE GASES IN STUDIES
OF UNDERGROUND NUCLEAR DETONATIONS*
C.F. Smith
Lawrence Livermore Laboratory
University of California
Livermore, California 94550
Abstract
The Livermore Gas Diagnostics Program employs a number of rare gas isotopes, both stable and
radioactive, in its investigations of the phenomenology of underground nuclear detonations. Radioactive
gases in a sample are radiochemically purified by elution chromatography, and the separated gases are
radioassayed by gamma-ray spectrometry and by internal or thin-window beta proportional counting.
Concentrations of the stable gases are determined by mass-spectrometry, following chemical removal of the
reactive gases in the sample. The most general application of the noble gases is as device fraction indicators to
provide a basis for estimating totals of chimney-gas components. All of the stable rare gases, except argon,
have been used as tracers, as have xenon-127 and krypton-85. Argon-37 and krypton-85 have proven to be of
particular value in the absence of a good tracer material as reference species for studies of chimney-gas
chemistry. The rate of mixing of chimney gases, and the degree to which the sampled gas truly represents the
underground gas mixture, can be studied with the aid of the fission-product gases. Radon-222 and helium are
released to the cavity from the surrounding rock, and are, therefore, useful in studies of the interaction of the
detonation with the surrounding medium.
INTRODUCTION
The gas diagnostics program at LLL was established to supplement conventional radiochemical
techniques for obtaining device performance-evaluation data in support of test program activities. We provide
analytical, interpretative, and diagnostic support for Plowshare gas-stimulation experiments, and we also
perform a wide variety of service radiochemical measurements in conjunction with laboratory experiments in
which gas tracers are used. Much of our effort requires measurement and interpretation of the concentration
of noble gases in our samples. These species are present as products of nuclear reactions, as components of
any air diluent of the samples, as tracer materials, or as combinations of these possibilities. We use both stable
and radioactive components. Although the nature of our diagnostic application of the noble gases is unique,
the procedures we have developed are of general applicability.
SAMPLING TECHNIQUES
Study of the phenomenology of an underground nuclear detonation requires collection of good-quality
samples of the cavity and chimney gases over times extending from a few minutes to several weeks following
detonation. The sampling system now used (Figure 1) basically consists of a reinforced rubber hose
containing and fixed to a nonrotating wire rope that provides longitudinal strength and prevents crushing of
the hose (Grens, 1972). The rope itself is a reasonably permeable gas path, even where the hose is compressed
against it through regions of high soil compaction. Both the rope and hose terminate at a vacuum demand
valve placed well downhole beneath the gas-containment stemming plug. This valve prevents passage of gas
unless a vacuum can be applied from above, thus precluding the possibility of pressurized gas reaching the
surface and providing a fail-safe closure of the system if the surface sampling equipment is not leak-tight.
Other built-in safety features include a maximum allowable working pressure of the hose that is well in excess
of any previously observed surface pressure. A deeply buried pressure relief valve ensures a safety factor of 6
or more in the upper portions of the sample system. Gas pumped through the system is totally contained either
in pressurized sample cylinders or in underground tanks. Sampling operations are conducted in such a
manner as to ensure compliance with the concept of total containment that characterizes explosive
emplacement techniques currently in use.
Another often-used sampling scheme involves the use of production tubing placed in a postshot reentry drill
hole. Although simple in concept, this technique suffers from the time delay imposed by the drilling schedule.
At the test site, the delay can be days or weeks, which means a loss of the information obtained from short-
lived radionuclides. In the case of gas-stimulation experiments, the delay has ranged from 1 month
(Gasbuggy) to nearly a year (Rulison), and has precluded detection of all but the longer-lived gaseous
radionuclides. To eliminate this loss of information, we emplaced a prompt gas-sampling system at Rio
Blanco (Quong, 1972). This novel system consisted of about a mile of half-inch, stainless steel tubing
containing a wire rope at its lower end, and three gas-block diaphrams that could be manually opened by back
pressurization from the surface. Unfortunately, chimney gas was not collected by this system. Our first
chimney gas samples are expected in early October 1973, following completion of drilling into the chimney.
*Work performed under the auspices of the U.S. Atomic Energy Commission.
-540-
-------
ANALYTICAL PROCEDURES
Samples are returned to Livermore for analyses. Typical sample volumes range from a few cubic
centimeters to a few tens of liters, although several cubic meters of sample have been processed at times.
Radionuclide concentrations range from background to a few hundred microcuries per milliliter. Needless to
say, our processing equipment exists in isolated high- and low-level specialized versions.
The first step in the processing of an unknown sample is a germanium-diode gamma-ray scan of 10 to 20 ml
of the gas. We have standardized this process to provide concentration measurements of the short-lived
krypton isotopes (87, 88, 85m) and of xenon-135 against the background of other gamma ray-emitting
nuclides in the gross fission product gas. In principle, it is possible to determine all of the krypton and xenon
isotopes in a gross sample by a judiciously spaced series of counts. However, to save time, gain sensitivity, and
improve measurement accuracy, we prefer purified krypton and xenon fractions of the sample for
measurement of the longer-lived fission product gases.
Radiochemically and chemically pure fractions of a sample are prepared by elution chromatography.
Helium is used as a carrier gas, and the separations are accomplished by manipulation of gas flow paths
through controlled-temperature activated charcoal and molecular sieve (5A) columns (Momyer, 1960).
Thermal conductivity and ionization-chamber detectors are used to monitor the progress of the elution. We
have six high-level gas-separation systems, and two similar low-level systems.
The desired quantity of sample is mixed with appropriate carrier gases for trace components, and collected
on activated charcoal at liquid nitrogen temperature. In general, the purification process consists of a series of
initial group separations from the activated charcoal at temperatures ranging from liquid nitrogen to 350°C
(Figure 2). A molecular sieve is used to obtain the final separation of each of the groups. A much simpler
scheme is often used to obtain specific components of a moderately radioactive sample (e.g., krypton can be
separated from air in a two-step process using only the charcoal column). Large samples require lower
temperatures until the major constituent is eluted, and may require multiple elutions from the sieve traps to
obtain species such as argon from air, or krypton from natural gas, in a pure form. Because of the large variety
of sample sizes and compositions that we are asked to process, our systems have an excess of built-in
versatility, and our techniques are modified accordingly. Simpler systems and techniques would suffice in a
large number of specialized applications, where more-or-less routine analyses are required.
Counting systems used for the measurement of the radioactive constituents of the separated gas fractions
are chosen according to their activity level. Low-level samples are counted in matched 0.8- and 0.5-liter
anticoincidence shielded beta proportional counters, and the same tubes are operated on the alpha plateau to
measure radon activity. More active samples are loaded into gas cells for thin-window beta counting or for
gamma spectrometry. We use the latter two methods for the great majority of our noble gas samples.
We determine the gross chemical composition of our samples by mass spectrometry. The automated system
now in routine use is applied to sample constitutents at 0.01% and higher concentrations (Crawford, et al.,
1970 and 1971), and gas chromatography has also been used in specialized applications to determine trace
components. We normally employ a sample enrichment process to gain sensitivity for the rare gases. As a first
step, the total noble gas concentration in a sample is measured by monitoring the pressure decrease caused by
the gettering of all reactive gases in a titanium sublimation pump. Enough sample to provide about 2ml(STP)
of the noble gas mixture is metered into a reaction flask containing calcium metal. When molten, the calcium
acts as a getter for all sample components except the noble gases. The residual gas from this concentration
process is analyzed mass spectrometrically for the relative concentrations of the noble gases. In conjunction
with the previously determined total noble gas concentration, this information is used to derive individual
concentrations of each of the components. This technique has been applied to air, natural gas, and chimney
gas samples, and has successfully been used to measure concentrations in the parts-per-billion range.
(Newton, et al., 1973).
NOBLE GAS TRACERS
Perhaps the single most important and useful application of the noble gases in our programs is in gas
tracing. Mere measurement of the concentrations of species of interest in a series of samples is of little value
unless one is able to identify and eliminate the effects of dilution, mixing, and chemical reactions that mask
the interrelationship between the samples. To accomplish this unmasking we rely on selected noble gas
tracers. These species are expected to be uniformly distributed within the cavity, are measurable over a wide
range of dilution throughout the desired observation period, and are not subject to excessive and variable
background corrections. Three classes of tracers can be defined: relative, internal diagnostic, and emplaced
diagnostic.
Application of gas tracers in a relative sense is best illustrated by our gas-quality investigation for Project
Rulison (Smith, 1971). During production testing, 85Kr was used as an indicator of relative sample quality and
of the extent of dilution due to the influx of formation gas into the chimney during production. It was also used
as a common denominator for ratios used to investigate the significance of the changing composition of the
produced gas.
A diagnostic tracer is a material present in the cavity in a known amount. It can be produced by the
explosive, or emplaced with the explosive, and, of course, must also meet the requirements previously stated
for a relative tracer. The "device fraction" is the ratio of the concentration of the tracer species in a sample to
-541-
-------
its known total in the chimney. Its reciprocal is the volume of chimney gas at standard conditions. If estimates
of chimney pressure and temperature are available, the reciprocal device fraction can be interpreted directly
as chimney void volume, and can be used to estimate the cavity radius. Most importantly for diagnostic
purposes, the total of any well-mixed component of the chimney gas is given by its concentration in a sample
divided by the device fraction. These totals are the most tractable quantities available for providing device-
performance estimates and for studying underground nuclear phenomenology.
Totals of internal diagnostic tracers must be either calculated or measured. The Rulison 85Kr total, for
example, was obtained from release estimates after production testing and was verified from chimney
volume, pressure, and temperature estimates and the initial concentration. Krypton-85 was also used as a
secondary tracer for the Gasbuggy gas-quality program (Smith, 1970a, b). In this case 127Xe(the 127Xe was
produced by neutron capture in a xenon sample that had been enriched in the lighter xenon isotopes) was
emplaced with the device to serve as the principal tracer. Because of its 36-day half-life, the 127Xe was useful
during only the initial shut-in period (6 months). Subsequent diagnostic tracing was based on the
experimental total 85Kr measured relative to the 127Xe during this period. Internal tracer totals are sometimes
estimated by calculation when measurements based on emplaced tracers are unavailable.
Normally, we prefer to emplace a unique tracer material with the explosive so that its total can be
determined under controller conditions. Materials that have been used include 127Xe, 85Kr, 222Rn, and stable
neon, krypton, and xenon. All except 127Xe require a background correction. This is obtained for 85Kr from the
85mKr seen in the samples; for radon from helium, using the known preshot radon/helium soil ratio; and for
the stable gases from argon in air or helium in natural gas, using known preshot tracer/argon or
tracer/helium ratios. \ ,. . .
Due in part to the expense and handling difficulties of radioactive tracers, and principally to the
development of reliable trace-gas measurement techniques by Newton, et al., (1973) we now rely almost
entirely on stable gas tracers. The most notable example is the Rio Blanco gas-stimulation experiment, where
each of the three explosives was emplaced with a unique tracer gas. When production testing is begun, we plan
to use these tracers to monitor the degree of inter-chimney communication and, if possible, to gain some
insight into characteristics of the individual chimneys. For gas diagnostics of test program events, we often
use stable krypton and/or xenon as the emplaced tracer material.
NOBLE GASES FROM A NUCLEAR EXPLOSION
Our studies of nuclear detonations must treat at least five distinct sources of noble gases: (1) The fission
process generates many krypton and xenon isotopes both directly and as a consequence of radioactive decay
of their precursors; (2) helium isotopes are among the products of thermonuclear reactions; (3) neutrons
generated by the detonation interact with components of the explosive, and of the surrounding medium, to
create noble gas isotopes through nuclear reaction processes; (4) the detonation itself vaporizes, melts, and
fractures the medium surrounding it, causing the release of soil gases to the chimney voids; and (5) air
normally enters the chimney region to replace the condensing steam during the cooling process, and in doing
so introduces an often significant quantity of noble gases that may be generally termed "background." Each
of these sources can be characterized and quantitized, using the analytical procedures mentioned previously.
The results of such a characterization constitute a viable gas-diagnostic program.
To obtain performance estimates for the explosive, we rely principally on those species expected to be
present in the cavity at relatively early times following detonation. Prior to chimney collapse, dynamic
mixing within the cavity assures a reasonably uniform mixture of detonation products and tracer gases, thus
fulfilling a basic prerequisite for meaningful results. In practical terms, this limits diagnostic fission
measurements to the krypton isotopes of mass 85, 87, and 88. Three of the four are short-lived, necessitating
rapid sample recovery and analysis. If sampling times are early enough so that xenon isotopes measured in
the gas are attributable to their independent fission yields, then these species can be used in diagnostics. At
intermediate times, interpretation of the measured totals of xenon isotopes is complicated. The xenon
precursor nuclides are predominately deposited on cool rock surfaces when the chimney collapses. Since
chimney-gas mixing rates are considerably slower than cavity rates, and some entrapment of the gas may
occur, the diagnostic application of the xenons during this period is questionable. In general, we prefer to use
these isotopes to study the phenomenon of gas mixing rather than for diagnostics. However, in some cases,
where late-time observations have been possible, we have been able to demonstrate that an appreciable
fraction of the xenon was mixed with the gas, and have been able to extract diagnostic information from its
measurement.
Since a significant fraction of the yield of a thermonuclear explosive is due to the fusion of deuterium and
tritium to produce 4He and a neutron, an experimental measurement of the device-produced helium is
potentially useful in diagnostics. The measurement is complicated by the presence of "background" helium in
the chimney gas from air and from soil. The argon content of a sample is a useful indicator of the amount of
helium (and other noble gases) present due to air dilution. Amounts of helium and radon in the soil gas are
related, since both are radiogenic. If the preshot helium/radon ratio for the detonation environment is known,
then the radon in a sample becomes a useful indicator of the background helium contributed by the soil, and is
also a useful indicator of the amount of rock that has been fractured by the detonation.
Neutrons produced by the detonation are eventually captured by materials in the device and its immediate
surroundings. Although certain isotopes of each of the noble gases are potential products of these reactions,
the most abundant such activation product is 37Ar —- produced principally from calcium. Production of the
37Ar occurs atzero time within the molten rock sphere around the explosive. It is, therefore, uniformly distributed
-542-
-------
and well mixed with other detonation products in the expanding cavity. Because of this, we have found 37Ar to
be an extremely useful relative gas-tracer material, and we have also used it as an internal diagnostic tracer
for samples that are too diluted to provide a useful stable gas-tracer measurement. In the latter application,
the total 37Ar must first be known from the results of measurements on good-quality samples, since
calculational estimates of activation processes are quite imprecise.
CONCLUSION
Throughout this presentation I have attempted to provide a generalized summary of our gas-diagnostics
program and of its utilization of the noble gases in studies of underground nuclear detonations. The results of
some of these investigations have been published in connection with the Gasbuggy and Rulison experiments
(Smith, 1970a and b,1971 and 1972). The specifics of our investigations do not easily lend themselves to
generalization, as each event is unique in itself in many ways. Such detailed presentation of our results is
necessarily relegated to individualized topical presentations. My goal here has been to draw upon our
experience in gas analysis techniques to describe concepts that are of general interest and applicability.
REFERENCES
Crawford, R., R. Stump, and J. Frazer (1970), A Computerized Fully Automatic Analytical Mass
Spectrometer, Lawrence Livermore Laboratory, Kept. UCRL-50931, Part I (Mechanical Design).
Crawford, R., R. Stump, and J. Frazer (1971), A Computerized Fully Automatic Analytical Mass
Spectrometer, Lawrence Livermore Laboratory, Kept. UCRL-50931, Part II (Electronic design).
Grens, J., G. Morris, and D. Skinner (1972), The Technique of Obtaining Gas Samples from
Underground Nuclear Detonations, Lawrence Livermore Laboratory, Rept. UCID-15999.
Momyer, F. (1960), The Radiochemistry of the Rare Gases, Subcommittee on Radiochemistry — National
Academy of Sciences, National Research Council, Rept. NAS-NS 3025.
Newton, J., F. Stephens, and R. Stump (1973), Determination of Trace Noble Gases in Air and Natural
Gas, Lawrence Livermore Laboratory, Rept. UCRL-74695 (to be published); also to be published in "Proc.
Noble Gases Symposium Las Vegas, Sept. 24-28,1973."
Quong, R. (1972), Project Definition, Chimney Gas Sampling System: Rio Blanco, in "Project Rio Blanco
Phase I Technical Studies," J. Toman and H. A. Tewes, Eds., Lawrence Livermore Laboratory, Rept. UCID-
15968, Appendix I.
Smith, C. (1970a), Gas Quality Analysis and Evaluation Program for Project Gasbuggy, in "Proc.
Symposium on Engineering with Nuclear Explosives, Las Vegas, Jan. 14-16,1970" (CONF-700101), Vol I, p.
775.
Smith, C. (1970b), Behavior of Radionuclides in Nuclear Gas Stimulation Techniques, in "Proc.
Symposium on Engineering with Nuclear Explosives, Las Vegas, Jan. 14-16,1970" (CONF-700101), Vol. I, p.
818.
Smith, C. (1971), Gas Analysis Results for Project Rulison Production Testing Samples, Lawrence
Livermore Laboratory, Rept. UCRL-51153.
Smith, C. (1972), Chimney Gas Radiochemistry in Nuclear Gas Stimulation Applications, "Nucl.
Technol.l5,))85.
Hose chute
Surface run of high-pressure
hose with wire rope Inside
To_sampfingv"3|||
Vacuum demand-
valve
Working point -
Adapter section
Hydroseal coating through
gas-seal stemming plug
32-mm-i.d. high-pressure
hose with 25-mm-diam.
wire rope inside (typical)
Foam wrap (typical at each
stemming plug)
Pressure relief valve (120 psi)
3 sealed set screws
at each hose coupling
—End of hose (1.25 cavity radii
from working point)
-—Bare wire rope continuous to
bottom of vacuum demand valve
100°C
"2
Rn-C-H,
r H 6
V2n+2
Figure 1. Gas-Sampling System.
Figure 2. Gas-Analysis Separation Scheme.
-543-
-------
RADON-222 MEASUREMENTS ABOARD AN AIRPLANE FOR THE DESCRIPTION OF
ATMOSPHERIC DIFFUSION*
J. Bogen
II. Physics Institute
University of Heidelberg
Heidelberg, Germany**
ACKNOWLEDGEMENTS
The author expresses his thanks to Dr. G. Schumann, Head of the Laboratory for Environmental
Radioactivity of the II. Physics Institute of the University of Heidelberg, for suggestions and advice in
interpreting the results of the investigation. My thanks are also due to Dr. W. Roedel for helpful discussions.
For assistance during equipment installation aboard the aircraft, as well as flight coordination, I wish to
thank Dr. M. Retnhardt, Deutsche Forschungs- und Versuchsanstalt fur Luft- und Raumfahrt,
Oberpfaffenhofen, Germany.
Abstract
Radon-222 is absorbed aboard an airplane in tubes filled with 50 g of selected charcoal and cooked in Dewars
by carbon dioxyde andfreon. After air collection at different heights up to 5 km, the tubes are evacuated and
heated in the laboratory for desorption of the gas molecules and for transfer into a decay chamber covered
with zinc-sulfide serving as a scintillatorfor the detection of alpha activity from radon-222 and its daughters.
Ingeneral, the measurements show an exponential decrease of the concentration up to 5 km in height, if there
is no cloud layer. The height for a reduction by a factor of two is about 900 meters. Assuming a constant
diffusion coefficientK0 up to 5 km, the measurements yield a Ko of approximately 3.7 x 104 cm2s-'.
For flights in stratiform clouds a decrease is observed of the concentration up to a certain height, then radon-
222 is again increasing or remains constant. This must be explained with respect to the meteorological
situation. Assuming only a constant value for Ko between the different, flying heights, it was possible to
appraise the development of the eddy diffusion coefficient with height.
INTRODUCTION
The radon isotopes 222Rn, 220Rn, and 219Rn are the gaseous members in the naturally occurring decay chains
assy, 232Th, and 235U. Due to the concentrations of their parent elements in the earth's crust, and their half-
lives of 3.8 days (222Rn), 54 seconds (220Rn), and 3.9 seconds (219Rn), the gaseous isotopes reach different
heights in the atmosphere. Their daughter products are heavy metal ions which adhere rapidly to aerosol
particles present in the atmosphere. In early measurements, the radioactivities of the aerosol particles
deposited on filters were often used to estimate the radon concentrations in air. This presupposes some
knowledge of the radioactive equilibrium between radon and its daughters. A recent discussion of the
problems in this field is given by G. Schumann (1972).
Due to its long half-life, 222Rn is a natural tracer in the troposphere. Dalu (1971) confirmed in laboratory
experiments that the only removal mechanism of gaseous 222Rn from an atmosphere consists in its
radioactive decay — in contrast to its daughter products which are removed by rain-out, wash-out, and/or dry
deposition. Thus, assuming a constant exhalation rate, the variations of the 222Rn concentrations in the
earth's atmosphere can be used for the description of turbulent diffusion — even inside the clouds or during
precipitation.
In recent years, only a few measurements are known for determining 222Rn height profiles in the
atmosphere. Hosier (1969) calculated the 222Rn concentration from the beta activity of 214Pb and 214Bi from
filters exposed for 20 minutes at 1 meter and 91 meters on a meteorologically instrumented tower in Troysons
Corner (Virginia, U.S.A.). The concentrations calculated for 222Rn are corrected with respect to radioactive
equilibrium.
Birot, et al., (1970) measured the radioactivity of short-lived 222Rn daughters deposited on filters aboard an
airplane (Cessna TU-206) in the southwest of France, using a zinc-sulfide scintillation screen of 20 cm in
diameter for the detection of the radioactivity. Wilkening (1970) collected air samples in evacuated stainless
steel tanks of 34.4-, 16.4-, and 8.2-liter volumes with an aircraft (Queenair) in Central New Mexico, U.S.A.,
using the radon counting system of Johnson's Laboratories. Paffrath (1971) developed equipment for the
measurement of a - and |3 - radioactivities of the radon-daughters deposited on filters aboard an airplane
(Dornier DO-27) in the area of Munich, Germany.
*Part of this work was financially supported by the German Federal Ministry of the Interior
(Schutzkommission).
**Present address: Abteilung Strahlenschutz und Sicherheit, Gesellschaft fur Kernforschung Karlsruhe,
Posfach 3640, 7500 Karlsruhe, Germany.
-544-
-------
In the following section, a system is described which is suitable for aircraft installation using 222Rn
adsorption on cooled charcoal. The system was developed at our Institute by Wenzel (1964) and improved bv
Bogen,e*a/.,(1970). y
EXPERIMENTAL
Atmospheric air is sucked through a dry cell, a filter for aerosol retention, and cooled tubes filled with
charcoal. During ground-level measurements, the air intake is outside the laboratory window (about 20 meters
above the ground) and passes through polyethylene tubes. A vacuum pump is connected to the outlet of the
charcoal filled tubes. When the 222Rn enrichment system is installed in the aircraft used for the experiments
(Queenair 65-65), a specifically designed emergency window is available for air suction from outside the cabin.
During the flight, the output of the charcoal filled stainless-steel tubes is connected to the suction-air system of
the aircraft, which generates an air stream of 4 liters per minute in the system. A total volume of 100 liters per
sample is necessary. Figure 1 shows a schematic diagram of the arrangement.
Phosphorus-pentoxide granules are used as a drying agent mixed with a color indicator which shows the
consumption of the agent. This is necessary because the adsorption coefficient of 222Rn on charcoal depends
on the water content of the air stream. Behind the dry cell, the dew point of the air is about -50 degrees Celsius.
The cellulose asbestos fiber filter is necessary for retention of aerosols, as well as particulate matter from the
drying agent. Differential and integral flow meters are inserted before connection to the stainless-steel tubes
filled with 50 grams of charcoal. The tubes are closed on each side by a valve. The charcoal was selected for low
226Ra content to avoid 222Rn emanation from the adsorbens. The type used (Merck-Kohle Nr. 9624) showed a
concentration of 2.7 fg of 226Ra per gram of charcoal.
Before each experiment the stainless-steel tubes filled with charcoal are heated in a vacuum up to + 450°C for
desorption of the gas molecules attached. Subsequently, six tubes are cooled to - 78°C in a Dewar vessel
(stainless-steel) containing difluoro-dichloro-methane and dry ice. The beginning of the cooling period was at
least one hour before the start of the experiment.
The Dewar vessel is mounted on an experimentation rack in the aircraft, and the air is sucked through
polyethylene tubes from outside the aircraft cabin. After adsorption of 222Rn at the desired height, the tubes
are brought back to the laboratory; and after having been warmed up to room temperature, they are kept in the
vacuum for 5 minutes. This causes no loss of 222Rn, but reduces the gas volume desorbed from the charcoal
after heating up to +450°C.
Figure 2 shows the radon measuring system consisting of two identical decay chambers (KI and KII) with a
volume of 168 cm3. The inner walls of the chamber are coated with silver-activated zinc-sulfide as the
scintillator, which is viewed by a photomultiplier tube. The decay chambers are evacuated before every
measurement. The gas desorbed from the charcoal is transferred to the decay chambers by opening the valves
and flushing the charcoal with a stream of helium until atmospheric pressure is reached in the decay chamber.
The counting rate recorded on a sealer increases until radioactive equilibrium is reached between 222Rn and
its short-lived daughter products. Above 75% of the counting rate in equilibirum is reached within 30 minutes.
The recorded counts are printed out automatically within preset time intervals.
The detection limit is on the order of one pCi per chamber volume, with a total error of about ± 10%.
During the flight, the air temperature was recorded automatically on a plotter, while the humidity was
measured discontinuously by a dew-point hygrometer. The flight height was plotted using a barometric
height recorder.
BASIC RELATIONS
The distribution of radon in the troposphere can be described rather well by the theory of eddy diffusion.
Detailed discussions of the problems involved were given by Bolin (1962) and Jacobi and Andre (1963).
Assuming a constant exhalation rate (E) from the ground, which is independent of time, and a uniform
horizontal distribution of 222Rn, the vertical concentration profile can be described as a solution of the
following differential equation — presupposing a negligible vertical wind velocity uz.
(I) |- (K xf^ ) - X x n - 0 W
3zz 3z
K2 = vertical eddy diffusion coefficient [cm2 x s-1]
n = 222Rn concentration [atoms per volume]
X = radioactive decay constant [s-1!
E = exhalation rate [atoms x cm-2x s-1 ]
c = X x n = radioactivity concentration of 222Rn [Ci x m-3]
To solve eq. (1), the following assumptions are made:
(a ) |JH _ 0 ( £ steady-state condition)
o t
-545-
-------
^IL=AH = 0 (= uniform horizontal distribution)
9x 9y
fc)u=0 ( = neglected vertical wind velocity)
\ ' z
The boundary conditions for Eq. (1) are:
(a) 7 A x n x d 7. = E(This means that the total radioactivity found in a vertical air column is equal to the
exhalation rate of the ground surface area.)
(b) (Z*») (This means that the 222Rn concentration decreases with increasing height as a
consequence of the radioactive decay.)
Assuming a height-independent vertical diffusion coefficient, Kz = Kc, solution of Eq. (1) is:
n(z) = 7AI77TX EXP L~ (ir)1/2 x ZD (2)
^K ; C
c
Considering concentration ratios in different heights, Zj < Z2, the following equation becomes independent of
the exhalation rate E:
= EXP c_ _)1/2 x ( ): (3)
n
Consequently, the boundary conditions must be valid for the height interval considered, and Kc is assumed to
be constant in this interval.
For the vertical diffusion coefficient Eq. (3) yields:
K
X x (Z2- Z,)2
For theoretical considerations, the height intervals can be chosen infinitely small, and the height-dependent
diffusion coefficient is the approximation of the step function for Kc, when (Z2 - Zj) = A Z approaches zero:
K = lim K ra
Z A C 5
For measurements, the height intervals chosen for the determination of the 222Rn concentration are always
finite, and greater than zero. Using Eq. (4), one can determine the vertical diffusion coefficient step-by-step,
always being independent of height within the chosen interval used for determination of the two
concentrations.
One has to bear in mind all the presuppositions and boundary conditions which are necessary to get the
solution of Eq. (1). If the measurements yield increasing concentrations with increasing height, one can only
state that in this region no steadv-state solution is possible. This might be caused by advective
processes, ( 3 n / 3 x 2 3n/Dy ^ 0) by a layer structure of the troposphere, (3n/3t ^ 0).
In the following paragraphs, a least squares fit to the measured 222Rn radioactivity concentrations is
calculated on the basis of Eq. (3). This implies the general expression for the concentration ratio:
n (Zi + l)
n (Z.) = EXP
-------
= ^
and this yields:
K A
c " a 2 (7)
An estimate of the error for the vertical diffusion coefficient derived with respect to Eq. (4) depends on the
accuracy of the radon concentration measurement, which is on the order of + 10%, and on the error of the
height determination, which is on the order of + 5% . Considering the expression
AK
c
K
c
1/2
one finds that the main contribution comes from the error of the concentration measurement. Using height
intervals of about 1,000 meters, and assuming a decrease by a factor of two for the concentration within 1,000
meters, one gets:
AK
-~ * i 40%
c
This error is assumed during the following discussion on the Kc values derived.
RESULTS AND DISCUSSION
The ground-level radioactivity concentration of 222Rn (shown in Figure 3), as well as the precipitation
measured in Heidelberg, is plotted for all days on which a height profile of 222Rn was measured aboard the
airplane. At ground-level an average value of 222Rn of 91 pCi m-3 (± 10%) is found for these days. The ground-
level concentrations are discussed now as a function of time with respect to the meteorological conditions.
On June 24,1971, in the western part of a high-pressure area reaching from the North Sea to the eastern part
of the Mediterranian Sea, warm humid air moved into Germany. No precipitation was registered in
Heidelberg. There were fair-weather cumulus, with the base at 1,500 m and the tops at 2,600 m. Measurements
were taken only outside the clouds.
On June 26, 1971 the ground-level concentration was somewhat lower than two days before. A high-
pressure area was over Spain, and warm humid air moved into Germany from the southwest. No precipitation
was registered in Heidelberg. There were cumulus clouds (base 1,200 m, tops 2,100 m) and a stratocumulus
layer between 2,700 m and 3,200 m. Measurements were taken only outside the clouds.
On June 28,1971 a low-pressure area lay over the Netherlands, and a warm-front crossed the southern part
of Germany. This produced 7.1 mm precipitation in Heidelberg. The clouds were formed by a thick
stratocumulus layer between 1,500 m and 4,000 m. The ground-level radon concentration was again lower
than two days before due to 13.5 mm precipitation on June 27,1971.
On June 30, 1971 a high-pressure area lay over England, and on its eastern part, cold maritime air moved
into Germany. This caused an occlusion accompanied by 10.6 mm precipitation in Heidelberg. The 222Rn
concentration measured at ground-level was only slightly higher than the minimum value measured one day
before. The clouds were formed by a stratocumulus layer between 1,100 m and 2,900 m
The 222Rn profile of June 24,1971 is plotted in Figure 4 as a function of the height. The 222Rn concentrations
are reported in percentages of the ground-level value, and are connected with a dashed line, while the fit to the
measured values is plotted as a straight solid line. The diffusion coefficients calculated from the measured
values are given in the lower part of Figure 4 with respect to the scale on the right. There, the broken line is
used as a vision line to link the different Kc values derived from the measured concentrations. As can be seen,
the 222Rn concentration decreases by a factor of two for every 300 m within the height interval 0 m to 600 m.
Considering the interval 600 m to 1,100 m, the concentration is constant within the error limits. In the height
interval 1,100 m to 3,000 m, the reduction by a factor of two in the concentration is reached within 550 meters.
The higher concentration gradient in the interval 0 m to 600 m corresponds to a lower value for the eddy
diffusion coefficient compared to the height interval 1,100 m to 3,000 m. The constant 222Rn concentration in
the interval 600 m to 1,100 m does not allow the determination of a value for Kc.
Figure 5 shows the height-dependence of the 222Rn radioactivity concentrations for June 26, 1971.
Decreasing values are observed until reaching the 3,600 m height which is just above the stratocumulus layer.
At the height of 5,200 m, the concentration of 222Rn is higher than at 3,600 m. This indicates that, in the height
interval from 3,600 m to 5,200 m, the troposphere is layered — meaning that an air parcel is present there
-547-
-------
which should be associated to a lower height, assuming steady-state conditions. The eddy diffusion coefficient
in the interval from 600 m to 3.600 m is lower than in the interval 0 m to 600 m. This indicates a reduced
exchange in the region of the stratocumulus clouds.
Figure 6 shows the 222Rn profile of June 28,1971 which decreases to a height of 2500 m. This height is in the
middle of the stratocumulus layer. In the upper part of this cloud, the 222Rn concentration increases —
indicating a change of the air mass. This presumption is supported by the meteorological situation discussed
above. The decreasing values of the eddy diffusion coefficient Kc indicates a minor exchange between the
different air masses.
The measured 222Rn concentrations of June 29,1971 in Figure 7 agree well with the exponential decrease
required in theory. This yields nearly constant values for the vertical diffusion coefficient KC- On that day,
Germany was on the rear side of the low-pressure area mentioned for June 28,1971. Precipitation composition
of the air mass is homogeneous until the height of 3,600 meters. Due to the precipitation of the previous day,
which ended 15 hours before the start of the flight on June 29,1971, the 222Rn concentration measured in the
maritime air probably does not have the concentration corresponding to the normal exhalation rate of the
Heidelberg area. Sappok (1971) measured an increase in concentration after precipitation on the order of 50%
within 6 hours using 220Rn as the tracer.
Figure 8 shows the 222Rn profile on June 30,1971 with decreasing values until a height of 1,500 m was
reached. In the interval of 1,500 m to 2,400 m, an increase in concentration is observed. This indicates a
change of the air mass. The air present in this height interval should either be associated to a lower altitude or,
in other words, reaches this height faster than forecasted by the steady-state solution. This agrees with the
meteorological situation which makes an occlusion responsible for the stratocumulus cloud. The eddy
diffusion coefficient decreases in the lower part of the cloud. This indicates a low exchange between the
different air masses.
Table 1 summarizes the values from the measurements of 222Rn profiles, such as the height for reduction of
the concentration by a factor of 2, as well as the diffusion coefficient derived from measured values and from
the calculated mathematical fit.
In Figure 9, the vision lines, derived from the vertical diffusion coefficients calculated from the measured
222Rn concentrations, are given as broken lines. These are constructed, assuming the calculated values of KC
plotted in the previous figures, for the middle of the height interval considered. The broader solid parts of these
lines indicate a layer cloud, when present. For ground-level (Z=0), the molecular diffusion coefficient of 222Rn
in air is assumed (about 5.5 x 10-2 cm2 s-1). Measurements performed by Sappok (1971), using 220Rn as the
tracer, yield 1.2 x 103 cm2 s-1 as an average value for the vertical eddy diffusion coefficient between 1 m and 6 m
in height. This value is used in Figure 9. The method for these measurements is described by Bogen, et al,,
(1972).
The broken lines in Figure 9 show two different types of height-dependence for the vertical diffusion
coefficient. On June 24 and 29, 1971, when only fair-weather cumuli were present, the eddy diffusion
coefficient seemed to increase rapidly within the first 500 meters above ground, and thereafter increased only
relatively slowly.
On June 26,28, and 30,1971, when stratus clouds were present (indicated by the solid broader part of the
lines in Figure 9), the diffusion coefficient seems to reach a maximum at a height which corresponds to the
height .of the cloud base. This indicates a change of the air mass or air parcel, respectively, and a low exchange
through the boundary layer. The solid traced curves in Figure 9 represent values calculated by Jacobi and
Andre (1963), who assumed special turbulence conditions in the ground-layer (0 to 1,000 m), the upper
troposphere (1 km to 12 km), and the stratosphere (12 km to 30 km). These turbulence conditions are indicated
by the characters I, W, N, S, which mean inversion, weak-, normal-, and strong-turbulence, respectively.
For June 24,1971, the vertical eddy diffusion coefficient derived from measurements agrees well with the
IWN-type of Jacobi and Andre's (1963) calculations, but no pronounced temperature inversion could be
observed because the accuracy of the temperature measurement was not sufficient. One June 29, 1971, the
development is similar to the NNN-type calculated by Jacobi and Andre (1963). The rest of the profiles
measured yield a height-dependence of the diffusion coefficient which agrees in principle with the NWN-type
of Jacobi and Andre's (1963). The profiles derived from the measured values of days with stratus clouds reach
the maximum value of the eddy diffusion coefficient at a greater height than predicted by the calculations.
All measured 222Rn concentrations are plotted in Figure 10 as percentages of the ground-level value
(equivalent of 100%) as a function of the height. The mathematical fit for all these values is marked as a
straight line, and yields a decrease by a factor of two in the 222Rn concentration for a height of 986 meters
(minimum value 300 m, maximum value 1,500 m). This means, for the case of a steady-state structure of the
atmosphere, that ground-level air takes the half-life of 222Rn (3.8 days) to reach a height of about 1,000 meters.
This yields anaveraged vertical eddy diffusion coefficient of about 4x10" cm2 s-1 (minimum value4x!03 cm2
s-1, maximum value 1 x 105 cm2 s-1) for the Heidelberg area in June 1971.
SUMMARY
A method which uses adsorption of atmospheric 222Rn on cooled charcoal is described. 222Rn height profiles
were determined using a specially equipped airplane in the area of Heidelberg from June 24 to 30 1971 Basic
relations of the eddy diffusion theory were used for interpretation of the results, and for determination of the
vertical diffusion coefficient. The average ground-level radioactivity concentration of 222Rn for the time
mentioned was 91 pCi per cubic meter of air. An average decrease by a factor of two for the 222Rn concentration
-548-
-------
was found at 1,000 meters of height. This yields, for the height-independent vertical diffusion coefficient, an
average value of 4 x 104 cm2 s-1. The derived development of the height-dependence of the vertical eddy
diffusion coefficient showed a rapid increase within the first 500 meters above ground; above this height, it
increased slowly when no stratiform clouds were present. A decrease in height was observed in the vertical
eddy diffusion coefficient when stratiform clouds were present in the lower troposphere; this was probably due
to meteorological conditions.
REFERENCES
Birot, A., B. Androuguer, and J. Fontan (1970), Vertical Distribution of 2MRn in the Atmosphere and
its Use for Study of Exchange in the Lower Troposphere. J. Geophys. Res. 75,12 2373.
Bogen, J., G. Hoffmann, and G. Kirrstetter (1970), Experimente zur Anreicherung von 222Rn.
Jahresberichtll. Phys. Institut Heidelberg, (unpublished).
Bogen, M. Sappok, and G. Schumann (1972), Measurements of Atmospheric Radon-220 Using Delayed
Alpha-Particle Coicidences. Arch. Met. Geophys. Biokl. Ser. A, 21171.
Bolin, B. (1962), Transfer and Circulation of Radioactivity in the Atmosphere, in Nuclear Radiation in
Geophysics (Ed. H. Israel and K. Krebs) p. 136, Springer Verlag, Berlin.
Dalu, G. A. and G. Dalu (1971), An Experimental Check on the Removal of Radon. Aerosol Sci 2 241.
Hosier, C. A. (1969), Vertical Diffusivity from Radon Profiles. J. Geophys. Res. 74 28 7018.
Jacobi, W. and K. Andre (1963), The Vertical Distribution of222Rn, 220Rn, and their Decay Products in the
Atmosphere. J. Geophys. Res. 6813 3799.
Paffrath, D. (1971), Untersuchung der Vertikalverteilung von Radonfolge-produkten in der Troposphare
mit dem Flugzeug mit Anwendung auf den vertikalen turbulenten Austausch. Bericht: DLR RB 71-68,
Deutsche Forschungs- und Versuchsanstalt fur Luft- und Kaunfahrt, Oberpfaffenhofen, Germany.
Sappok, M. (1971), Der Einfluft meteorologischer MeftgrbBen auf die Aktivitat des Thorons in der
Atmosphare. Diplomarbeitll. Physik. Institute Heidelberg (unpublished).
Schumann, G. (1972), Radon Isotopes and Daughters in the Atmosphere. Arch. Met. Geophys. Biokl. Ser.
A, 21149.
Wezel, K. P. (1964), Messung des 222Rn und seiner kurzlebigen Folgeprodukte in Bodennaher Luft.
Diplomarbeit, II. Physik Institute Heidelberg, (unpublished).
Wilkening, M. H. (1970), 222Rn concentrations in Convective Patterns of a Mountain Environment. J.
Geophys. Res. 75 91733.
-549-
-------
TABLE 1. Summarized Flight Data for the Determination of 222Rn Radioactivity
Concentrations in the Heidelberg Area. Kc(min,max) is Calculated from the Least Square Fit
Values.
Date
Height
[nterval
meter]
lower upper
i J
Height
^2) =0,5
c(Zi)
meter r
Vertical eddy diffusion
coefficient
Kc(i,j) Kc(min,max)
cm2 -] r cm2 -i
s J L^""J
June 24, 0 600
1971 600 1,100
1,100 3,000
300
oo
550
0.38 x 104
ND
1.44 x 104
1.47 x 104
June 26, 0 600 925 3.7 x 10"
1971 600 3,600 850 2.02 x 104
3,600 5,200 + ND
5.08 x 104
June 28, 0 1,300 850
1971 1,300 2,000 500
2,000 2,500 350
2,500 3,600 +
June 29, 0 2,100 1,500
1971 2,100 3,600 1,500
2.56x10"
1.37x10"
5.78x10"
ND
9.19 x 10"
11.11x10" 9.74x10"
3.3 xlO4
June 30, 0 1,100 850
1971 1,100 1.00 250
1,500 2,400 +
3.52 x 10" I
2.63 x 10" t
ND 6.07 x 10"
+: increasing concentrations.
ND: not determined.
-550-
-------
AIR
INLET
DRY CELL
FILTER
STAINLESS STEEL
DEWAR
VACUUM PUMP
SUCTION AIR SYSTEM)
Figure 1. Schematic diagram of the 222Rn enrichment system.
-------
E-r
§!
to
1
75'
';•
i —
.. ••. ;. _,. .
• 1
>;;,
-*4
P -
>.'•;
70
VACUUM
PUMP
Figure 2. Schematic diagram of the 222Rn measuring system. EVl, EV2, HI, H2 = valves, W = glass wool, C =
charcoal, E = stainless-steel tube, KI, KII = decay chamber, MS = brass, P = Plexiglass (a methacrylate
polymer), PM = photomultiplier, (dimensions in millimeters).
-------
C (222Rn)
dpm
4004
300
200
100
200
150
\ _
\
\
\
\
.. \
\
\
\
\ T
4-100
A
PRECIPITATION
4
50
[mm]
10
8
6
2
M
Y
Date
24 25 26 27 28 29 30 June,1971
Figure 3. Radioactivity concentrations of mRn and precipitation of June 24 to 30, 1971 registered at the
Heidelberg ground-level station.
-553-
-------
Figure 4. 222Rn concentration of June 24, 1971 in percentages of the ground-level value and height-
dependence of the vertical eddy diffusion coefficient K. The straight line gives the concentration calculated
with a least squares fit program. The horizontal lines are the Kc yalues (of which the scale is on the right side)
derived from the measured 222Rn concentrations. The broken curve connects the Kc~ values at the centers of the
height intervals.
-554-
-------
10*
1.0
2.0
3.0
4.0
6,0»103 [m]
Figure 5. 222Rn concentration of June 26, 1971 in percentages of the ground-level value and height-
dependence of the vertical eddy diffusion coefficient K. The straight line gives the concentration calculated
with a least squares fit program. The horizontal lines are the Kc values (of which the scale is on the right side)
derived from the measured 222Rn concentrations. The broken curve connects the Kc values at the centers of the
height intervals.
-------
106
c^-EXP[-0,0008. z.4,6i]
1.0
2.0
Figure 6. 222Rn concentration of June 28, 1971 in percentages of the ground-level value and height-
dependence of the vertical eddy diffusion coefficient K. The straight line gives the concentration calculated
with a least squares fit program. The horizontal lines are the Kc values (of which the scale is on the right side)
derived from the measured 222Rn concentrations. The broken curve connects the Kc values at the centers of the
height intervals.
-556-
-------
[V.] C(222Rn)
100-
10
K
June 29,71
EXp[-0,000465-Z*4j6l]
K
cumulus
106
10'
1,0
2.0
3,0
4,0 x 103 [m]
Figure 7. 222Rn concentration of June 29, 1971 in percentages of the ground-level value and height-
dependence of the vertical eddy diffusion coefficient K. The straight line gives the concentration calculated
with a least squares fit program. The horizontal lines are the Kc values (of which the scale is on the right side)
derived from the measured 222Rn concentrations. The broken curve connects the Kc values at the centers of the
height intervals.
-557-
-------
C (222Rn)
K
June 30,71
to6
[af]
f-
/ il^-EXP -0,00059'Z«4,61
• c(Zf) L J
10-
Figure 8. 222Rn concentration of June 30, 1971 in percentages of the ground-level value and height-
dependence of the vertical eddy diffusion coefficient K. The straight line gives the concentration calculated
with a least squares fit program. The horizontal lines are the Kc values (of which the scale is on the right side)
derived from the measured 222Rn concentrations. The broken curve connects the Kc values at the centers of the
height intervals.
-558-
-------
—June 29,1971
SSN
NNN
IWN
June 26,1971
10
10
3,0
4,0x10 [m]
Figure 9. Height-dependence of the vertical eddy diffusion coefficient. Broken lines are derived from
measured 222Rn radioactivity concentrations (with indication of stratus clouds by broader solid parts). Solid
lines: calculated values by Jacobi, et al., (1963) for special turbulence conditions in the ground-layer, the upper
troposphere, and the stratosphere, respectively. I, W, N, S, stand for inversion, weak-, normal-, strong-
turbulence, respectively.
-559-
-------
1,0 2,0 3.0 4,0 5,0 6,0x 10J [m]
Figure 10. Radioactivity concentration profiles of 222Rn in percentages of the ground-level value measured in
June, 1971 in the Heidelberg area. The solid straight line represents the values calculated vidth the least
squares fit to all measured values.
-------
NOBLE GAS LASERS FOR AIR POLLUTION MONITORING
G. J. Linford
Laser Division
Hughes Aircraft Company
Culver City, California 90230
Abstract
Recent experiments with pulsed and continuous wave (CW) noble gas lasers have revealed that reliable,
high gain laser oscillations at numerous infrared wavelengths can be achieved using a relatively simple
laboratory apparatus. A total of more than eighty infrared neutral and ionized noblegas laser lines have been
obtained using this equipment; eighteen of these noblegas laser lines were previously unreported. Despite the
large number of noble gas laser lines, coincidences between laser lines and air pollution absorption lines are
relatively rare, and hence some degree of spectral tuning is desirable in order to achieve coincidences between
laser lines and pollutant absorption lines. In experiments performed here, it has been shown that a high-gain
xenon laser can be tuned (using the Zeeman effect) over as much as lOAwithapplied dc magnetic fields of six
kilogauss; furthermore, noblegas lasers may be tunable over more than 100 A using strong, pulsed magnetic
fields of at least sixty kilogauss. In spectral regions, where these noblegas laser lines are relatively closely
spaced, the Zeeman effect may permit tuning from one laser wavelength to another. It has been shown here
that high-gain noblegas laser lines can be tuned (via the Zeeman effect) into coincidence with absorption lines
of selected pollutant gases.
INTRODUCTION
Among the most essential features of an effective air-pollution control system are unambiguous methods
appropriate for detecting, identifying, and quantitatively measuring the concentrations and distributions of
gaseous air pollutants. One highly useful -technique for identifying and quantitatively measuring air
pollutants is optical absorption spectroscopy. This paper presents the results obtained using noble gas lasers
in absorption spectroscopic experiments conducted to (1) increase the number of available laser wavelengths;
(2) tune individual laser lines over a narrow spectral region; and (3) demonstrate that quantitative data on the
concentrations and distributions of selected air pollutants can be obtained with a noble gas laser absorption
spectroscopic apparatus.
There have been a number of serious difficulties in the past that have been associated with using optical
absorption spectroscopy for pollution measurements, most of which have been due to the use of incoherent
light sources. With the technological development of laser sources, however, many of these difficulties are
overcome, and the useful characteristics of absorption spectroscopy could be adapted from their former
laboratory applications to actual field measurement situations.
Although laster oscillation has been achieved using a wide variety of solid-state, liquid, and gas lasers, the
special requirements for air pollution monitoring equipment restrict the choice of laser sources to (1)
continuously tunable lasers and/or (2) multiple wavelength, partially tunable lasers. This paper describes the
work done in-the latter category using noble gas lasers as narrow-band light sources for optical absorption
spectroscopy.
Long optical path length absorption techniques using laser sources have a number of significant
advantages over comparable incoherent optical detection techniques for the measurement and detection of air
pollution. These advantages are:
(1) very narrow spectral line-widths permit extremely high resolution spectroscopic measurements to be
obtained;
(2) the high laser radiances reduce the optical detection problems;
(3) the high beam collimation permits long optical path lengths to be monitored, thereby increasing the
sensitivity;
(4) a capacity for unique identification of pollutant absorption lines is available using this technique; and
(5) laser sources offer the availability of coherent detection techniques (such as heterodyne detection).
Problems do remain, however, particularly those problems associated with band spectra, pressure
broadening of fine spectral structures, etc.
' The number and wide spectral dispersion of noble gas laser lines exceed those of all other discrete
wavelength laser sources. To date, more than 400 neutral and ionized noble gas laser lines (Hdbk of Lasers,
1971) ranging in wavelength from the vacuum ultraviolet (xenon@0.1469//m) to the far infrared
(neon@132.8pm)-Tunablelasersources, notably dye lasers arid optical parametnc oscillators (OPO), are
capable of covering specific spectral regions well, although, at present, not all spectral regions of interest are
included. Thus, dye lasers cover the visible portion of the spectrum in detail, but, as yet, spectral coverage
much beyond 1.0 micrometers is poor. OPO's have proven to be difficult to align and maintain, but OPO s are
presently capable of generating optical radiation in the near visible and middle infrared portion of the
spectrum.
-561-
-------
The major absorption lines of atmospheric pollutants lie either in the ultraviolet (corresponding to atomic
transitions within the atom constituents of the pollutant molecules) or in the infrared (corresponding to
vibrations, rotations, torsions, etc., of the molecules themselves). Thus, aside from the oxides of nitrogen, most
atmospheric pollutants appear to be nearly colorless, owing to the absence of absorption lines in the visible
portion of the spectrum. Homonuclear molecules, such as N2, are essentially free of absorption lines in the
infrared, owing to the lack of a dipole moment; this is a convenient property, since it effectively eliminates
homonuclear molecules from being detected (in the infrared), and homonuclear molecules are usually not
sources of air pollution.
Although there are a large number of noble gas laser lines available, coincidences between these lines and
individual absorption lines of pollutant gases are relatively rare. The advantage in having (1) numerous
available laser lines and (2) a small degree of tunability of these lines is apparent since it is only necessary to
tune a typical noble gas laser line from 10 to 100 A before coincidence with a neighboring noble gas laser line
can be achieved. In many cases, tuning a noble gas laser line by only 10 A will bring that laser line into
spectral coincidence with a pollutant absorption line. Noble gas laser lines are inherently fixed
frequency/wavelength standards, unless the noble gas discharges are placed in relatively strong magnetic or
electric fields. Recent experiments (Linford, 1973a) with a xenon laser oscillator placed inside a six kilogauss
solenoid have indicated that a number of CW infrared xenon laser lines can be continuously tuned over a
spectral interval of up to 11 A using the Zeeman effect. These same experiments (Linford, 1973a)
demonstrated that the 3.508 xenon laser line can be tuned into coincidence with an absorption line of
formaldehyde with an applied magnetic field of approximately five kilogauss.
More than ten years have elapsed since laser oscillation was first reported in gas discharges of the neutral
noble gases (Patel, et al, 1962; Javan, et al, 1961; and Faust, et al, 1962). In spite of this, however, it has
recently become apparent that a considerable number of new noble gas laser transitions may be available in
addition to those previously reported in the literature. During the past two years, for example, the author has
found a total of 18 new neutral argon, krypton, and xenon laser lines (Linford, 1972 and 1973b and c) using a
relatively simple laser apparatus. Most of these new laser lines are located in spectral regions common to both
the lasers and to spectral absorption lines of important atmospheric pollutants. As such, these new noble gas
laser lines may be used as selective laser oscillators for differential absorption measurements of specific
pollutant gases.
If a Zeeman-tuned noble gas laser is used as the probe laser source for such absorption measurements, laser
oscillation at two wavelengths, A+ and A_ , can be obtained. In general, only one of these two Zeeman-split
laser lines will be tuned into coincidence with a spectrally-narrow pollutant absorption line, and, as a
consequence, the laser line suffering the least absorption can serve as a comparison laser source for detailed
differential absorption measurements (Linford, 1973a) for long atmospheric optical paths, such as the air over
a polluted urban environment.
Additional simultaneous noble gas laser lines can be obtained by pulsing the laser discharge, although
some electrical discharge instabilities result in a Zeeman-tuned laser owing to the interactions between the
moving ions and electrons and the applied magnetic field. Pulsing the magnetic field permits higher (peak)
magnetic fields to be obtained (thereby increasing the Zeeman-splitting of the laser lines), but synchro-
nization of the pulsed magnetic field with a pulsed noble gas laser oscillator is necessary in order (1) to ensure
that the laser oscillates when the (pulsed) magnetic field is maximum, and (2) to prevent the induced electric
field associated with the increasing or collapsing magnetic field from extinguishing the noble gas electrical
discharge.
The atomic energy level diagrams for argon, kjypton, and xenon are quite complex, composed as they are of
two relatively independent systems corresponding to the2P°]/ and 2p° 3/ parent ions, respectively. Radon is
expected to be somewhat similar, although almost none of the energy levels corresponding to the2 p°! / parent
ion is known (Moore, 1959). The plethora of energy levels available in the heavy noble gases allows2a large
number of laser lines to oscillate simultaneously. As discussed below, however, the noble gas lasers are
inherently inefficient, owing to the location of the pertinent energy levels for laser oscillation relatively near
the ionization limit(s) for these gases. Frequently, simultaneous laser oscillation at a number of wavelengths
is inconvenient, and dispersive laser cavity elements are usually introduced to suppress these other laser lines.
Noble gas lasers cannot, in general, be used to detect the concentrations and distributions of the same
noble gas used in the laser. As discussed in the following section, noble gas lasers are usually 4-level lasers,
and the optical transition corresponding to a given laser line cannot be abosrbed by an identical atom in its
ground state. In the case of the neutral argon laser, this property is an important one since argon represents
almost one percent of the terrestrial atmosphere, a concentration many orders of magnitude greater than
those expected for atmospheric pollutants.
In summary, noble gas lasers have been found to be promising optical sources for atmospheric pollution
detection and measurement using selective spectral absorption techniques. Experiments and analyses have
shown that the numerous noble gas laser lines can be:
(1) allowed to oscillate individually or simultaneously;
(2) Zeeman-tuned over spectral bands of up to 10 A (using six kilogauss dc magnetic fields);
(3) transmitted over long optical paths in the atmosphere to provide great detection sensitivity;
(4) operated in those spectral bands common both to the laser and to important atmospheric pollutants;
and,
-562-
-------
(5) either pulsed or CW laser oscillation modes can be used depending upon the application and type of
data processing system required.
Only a relatively simple laser apparatus and associated data processing system are required to obtain
adequate data on the concentrations of atmospheric air pollutants. Details regarding the laser apparatus,
data, and selected pollutant gases can be found in the following sections.
EXPERIMENTAL MEASUREMENTS AND RESULTS
Our noble gas laser experiments were designed to address three specific tasks:
(1) laser source development (conventional CW noble gas lasers, transverse-excited, atmospheric (TEA)
pressure, pulsed noble gas lasers, and conventional low-pressure pulsed noble gas lasers);
(2) spectral tuning experiments of noble gas lasers (using applied magnetic fields to generate the Zeeman
effect); and
(3) pollutant absorption experiments (using both intra-cavity and extra-cavity optical paths of variable
length).
Most of our efforts were concentrated on the first area, laser source development, and, consequently, the
following discussion emphasizes this work.
Several hundred neutral noble gas laser lines have been discovered during the past decade by a number of
laser researchers (Hdbk. of Lasers, 1971). It was recognized early in our experiments, however, that there were
four problems associated with implementing these numerous noble gas laser lines for air pollution monitoring
applications:
(1) Some of these noble gas laser lines were obtained using laser apparatus of considerable size and
complexity, and, as a consequence, this type of laser equipment may not be suitable for air pollution detection
applications.
(2) Many of these noble gas laser lines fall into the extensive H,O and CO, absorption bands in the
terrestrial atmosphere — since HjO and CO2 are not considered to be pollutant species, those laser lines
falling in these H_O and CO2 absorption bands cannot be used as laser sources for long atmospheric path
length measurements.
(3) A considerable number of the known noble gas laser lines are too weak to be used successfully over long
atmospheric paths, particularly if air turbulence is present.
(4) A few noble gas laser lines fall in spectral regions in which little or no pollutant absorption lines are
present—these spectral regions are generally termed "windows" in the atmospheric absorption spectrum.
In order to automatically exclude those noble gas laser lines requiring an elaborate apparatus, we
constructed a simple, relatively compact noble gas laser device having an active length of approximately 1.0
meter. Experiments were conducted in which noble gas pressure and discharge parameters were
systematically varied, and the number, wavelengths, and output powers of the noble gas laser lines obtained
with this equipment were carefully measured. In these experiments, our studies were restricted to the heavier
noble gases (argon, krypton, and xenon) since the He/Ne laser had previously been subject to intensive
development.
Some eighty noble gas laser lines were obtained; Tables 1, 2, and 3 list the pulsed laser lines obtained with
argon, krypton, and xenon respectively. Table 4,5, and 6 list the CW laser lines obtained with argon, krypton,
and xenon lasers respectively. Note the remarkable dominance that xenon has over the two lighter noble
gases. In particular, it was found that if xenon were allowed to remain in the laser apparatus overnight, xenon
contamination of the apparatus occurred, presumably by the gas being strongly adsorbed on the inner
surfaces of the laser plasma tube. Despite the fact that the contaminating xenon gas was present in very small
fractions compared to the desired fill gas (fractions as small as 10-3 to 10-4 were estimated), the xenon laser
spectrum generally dominated over the spectrum of the majority noble gas atoms. This dominance is
attributed in part to the greater line strengths of many of the xenon laser lines and to the lower-lying energy
levels in xenon.
Of the eighty noble gas laser lines obtained with our equipment, some 18 argon, krypton, and xenon laser
lines were previously unreported. Many of these 18 new laser lines (Linford, 1972 and 1973b and c) fall in or
near important pollutant absorption bands and lines — see Table 7 for a listing of the wavelengths,
identifications, and estimated line strengths of these new lines.
A substantial number of the 80 noble gas laser lines obtained with our simple apparatus were observed to
have adequate optical gains and output powers for air pollution monitoring applications, but only when
operated in the pulsed mode. This necessity for pulsed operation may not be a disadvantage in practice, since
electronic photodetector gating, lock-in amplifiers, etc., may be used to permit an acceptable signal-to-noise
ratio to be achieved for an operational data-processing system.
1. Description of Laser Apparatus.
A general schematic of the laser apparatus used in many of our noble gas laser experiments is shown in
Figure 1. As shown, a hot cathode/ring anode laser tube is fittediwith a cryogenic pressure-control cold finger
for regulating the tube pressure of xenon (Linford, 1973a). A heater coil wrapped around the cold finger allows
surface temperatures to be varied from 77° to over 100°K. For CW laser operation, the noble gas apparatus was
excited with a dc electrical discharge. When pulsed laser operation was desired, two circuits were used:
-563-
-------
(1) a high voltage capacitor (C = 2 nF) was placed in parallel with the plasma tube electrodes and a large,
1.0 megohm, resistor was placed in series with a 15 kV high voltage power supply; and
(2) a hydrogen thyratron was used to apply controlled-duration high voltage pulses of variable current to
the laser plasma tube.
The laser tube was placed inside a dc solenoid to permit axial magnetic fields up to 6 kilogauss to be applied to
the laser tube. As discussed above, the six kilogauss applied magnetiafield allowed us to tune a number of CW
xenon laser transitions over bandwidths as great as 23 GHz (or 11 A) using the Zeeman effect. Although as
many as ten spectral components are obtained in Zeeman-split spontaneous emission spectra (in the case of
the 3.508ym xenon laser transition), only two laser wavelengths, A+and A_, were obtained when magnetic
fields in excess of one kilogauss were applied to the laser tube. This convenient reduction in the number of
oscillating laser lines in the Zeeman-split noble gas laser is attributed to spectral cross-relaxation processes
occurring within the low-pressure noble gas discharge (Peressini and Linford, 1968).
The magnetically-induced spectral-splitting, ovm, of a given Zeeman-split spontaneous spectral line can be
calculated from the expression:
(1)
where m is the magnetic quantum number of the upper energy level of the transition, g^ and gp are the
Lande-g values of the upper and lower laser levels, respectively, and B is the applied magnetic field; In order
to use equation (1) to calculate the spectral splitting effected by the applied magnetic field, it is necessary to
either measure or calculate the values of the Lande-g coefficients.
The spectra of the neutral noble gases can be characterized most accurately by a j--f coupling scheme
(Racah, 1942). In the case of j- f coupling, the orbital angular momemtum of the excited electron,^, is coupled
to the total angular momentum of the parent ion, j , to form a resultant intermediate quantum mechanical
angular momentum, K. Thus:
K = Jp^e (2)
The total angular momentum, J, is then given by the vector sum:
J = K + se (3)
where se is the spin of the excited electron. Since only about half of the Lande-g values for equation (1) have
been measured and reported for the noble gases (Moore, 1959), it was necessary to calculate the remaining
Lande-g values using the j- 9 coupling conditions. Under these circumstances,
_ <2K +1H2J-M) p
where j? is the total angular momentum of the excited electron and gp is the Lande-g value of the appropriate
parent ion. The Lande-g value, g , for the parent ion can be calculated from the ordinary Russell-Saunders
coupling expression:
(S+l)
JU-H)
mS(S+l)-l(L
2J
The spectra of the neutral noble gases (particularly the heavier noble gases) arise from two parent ions,2p°y-
and tfji When appropriate values for the quantum numbers describing these two parent ions are inserted
in equatidns (5) through (1), above, we found that the magnetic splitting factor, £ , given by the expression:
27m
can be calculated to be 1.25 MHz/gauss for the 3.51/im xenon laser line, 1.31 MHz/gauss for the 3.37^m xenon
line, and 2.6 MHz/gauss for the 3.99^m xenon transition (Linford, 1973a). All but the last magnetic splitting
factor agreed with the experimental measurements of the spectral splitting of these xenon laser lines. A
-564-
-------
magnetic splitting factor of only 1.9 MHz/gauss was measured instead of the calculated 2.6 MHz/gauss for
the3.99fzm xenon laser line. This relatively large error may be due to an incorrect Lande-g value in the
Atomic Energy Levels (Moore, 1959).
The values of the magnetic splitting are quite substantial using a six kilogauss applied magnetic field
(amounting to total line splittings of nearly 23 GHz, and 11 A, in the case of the 3.99^m xenon laser line)
These magnetic splittings were measured in two ways: (1) for small fields and splittings, the heterodyne
beats between the two Zeeman-split laser lines could be observed using a square-law photodetector (of
sufficient speed) connected to a microwave spectrum analyzer for frequencies up to 1.2 GHz; and (2) for larger
applied magnetic fields (and proportionally larger frequency splittings), a 3/4 meter Czerny-Turner
monochromator (such as that shown in Figure 1) was used to measure the spectral splittings directly.
Photodetection of the noble gas laser lines was accomplished with one of three types of photodetectors
(depending upon the wavelength region of interest). These photodetectors were:
(1) a photovoltaic InAs detector operated at 300°K with a spectral coverage of from 0.4/zm to 3.5/im,
(2) a photovoltaic InSb detector operated at 77°K with a spectral coverage of from 0.4/tmto 5.6/im , and
(3) a photoconductive gold-doped germanium detector covering the range 0.7/zm to lift m .
The photovoltaic InAs detector used in these experiments was an exceptionally fast detector — heterodyne
beats between Zeeman-split laser lines having frequency differences as great as 1.2 GHz were observed. The
other two detector types required liquid nitrogen cooling (to 77°K), were considerably slower, and their
selection for these experiments was predicated primarily on their substantial photodetector sensitivities at
wavelengths be'yond the spectral cutoff of the InAs photodetector.
There were several techniques used to select the desired laser wavelengths reaching the photodetector. In
the case of the Zeeman-tuned laser, the two Zeeman-split laser lines were circularly polarized, but of opposite
helicitics. By inserting a quarter-wave plate in the laser beam containing both laser lines, it was possible to
convert the two circularly-polarized laser lines into two laser beams linearly polarized perpendicular to each
other. A suitable prism analyzer could then separate these two wavelengths directly (Linford, 1973a).
For laser experiments employing more widely separated wavelengths, it was possible to use gratings,
dicrotic beamsplitters, or suitable narrow-band interference filters to separate the two laser beams. The
Littrow prism (shown in Figure 1) could be used to switch sequentially from one noble gas laser wavelength to
another, providing that the wavelength differences between the two laser lines were sufficiently great
(- 0.5ij m or better) in order to provide a sufficiently large dispersion factor.
Unfortunately, infrared noble gas lasers are inherently inefficient devices, particularly when the excitation
mechanism uses conventional direct electron excitation from the (np) ground-state to excited (n + l)s or (nd)
states. As indicated in Figure 2, the first excited states, (n + l)s, lie far above the (np) ground-state, and since
laser oscillation takes place between two relatively closely-spaced energy levels, the theoretical quantum
efficiency is necessarily low. Furthermore, since there are a number of possible energy levels within each
multiplet, the resulting population inversions are spread among most of these energy levels. An example of
this relatively complex energy level diagram is shown in Figure 3. The first excited states, the 5s levels, are
located nearly 80,000 cm-1 above the 4p ground-state. The multiplicity of these population inversions gives rise
to the useful multiple wavelength oscillation property of noble gas lasers. As shown in Figure 3, it is possible
for numerous neutral krypton laser lines to oscillate simultaneously since these lines have neither upper nor
lower laser levels in common.
' The effects of isotopes on the oscillation characteristics of noble gas lasers predominantly affect the line-
width of the laser amplifier. Thus, although the line-widths of the noble gas laser amplifiers are Doppler-
broadened at the low pressures (P-5--10 millitorr) used in conventional laser oscillators, the presence of
naturally-occurring isotopic mixtures broadens the line-width further. This isotope effect is particularly
interesting in the case of xenon which not only has nine stable, naturally-occurring isotopes, but also has
approximately half of the naturally-occurring atoms composed of even-odd (xenon-129 and xenon-131)
isotopes. Accordingly, naturally-occurring xenon not only displays a mass-shift, but also hyperfine structure
in its spectra.
2. Radon Considered as an Active Laser Medium.
The incomplete term diagram of radon is very similar to those of xenon and krypton. Based on the pioneer
work on radon by Rasmussen (1931 and 1932) only part of the term diagram of radon is known. Radon is the
heaviest noble gas, having 86 electrons, no stable isotopes (radon-222 is the longest-lived isotope), and an
average atomic weight of 222. Radon condenses at 211 °K and freezes at 202°K, so that if an excess of gaseous
radon were available in a laser tube, a cryogenic pressure control system similar to that used for xenon (see
Figure 1) could be used to maintain the low gas pressures which are optimum for conventional heavy noble
gas lasers (Linford, 1972).
Radon-222 is part of the radium decay scheme, is radioactive, and has a half-life of only 3.827 days —
decaying by alpha emission with an energy of 5.5 MeV. The differential equation governing production of
radon from radium is given by
A TV!
- a,N, (?)
Li Lj
-565-
-------
where N2 is the population of radon, N^ is the population of radium, and a1 and a2 are the mean reciprocal
lifetimes of each element. If the production of radon from the parent element, radium, is considered to
commence at t = 0, then the integral of equation (7) subject to the initial condition is simply
N (0)a I exp(-a t) - exp(-u
which reduces to
Nj(0)
°
exp(~a?t)
(9)
. As an asymptotic solution, N2(oo)HM1 ^L , so that the quantity of radon gas in equilibrium with
the parent radium is essentially the ratio al/a2 times the quantity of radium present. Approximately 1015
atoms of radon are required to fill a typical laser tube to a pressure of five millitorr, and, hence, some 1.5 x 1020
atoms of radium are required as a source — this corresponds to approximately 0.6 grams of radium. When
allowances are made for radon gas adsorption by the laser tube walls, at least two grams of radium are
required for a 0.3 cm bore gas discharge tube having a length of 50 cm (Vgas = 3.5 cm3). An electrodeless laser
tube would minimize the internal volume of the laser tube, and since rf excitation in excess of 109 Hz reduces
gas adsorption, this type of excitation is recommended for any initial experiments conducted with radon.
We were unable to acquire a 2 gram source of radium to perform actual laser experiments with radon, but we
did make some predictions as to the wavelengths and line strengths of the expected radon laser lines should a
suitably-large source of radium become available in the future. The 2-gram radium source should be located in
a detachable side-arm of the plasma tube so that the contaminated laser tube can be suitably disposed of after
the experiment has been completed. The laser experiment should be performed approximately 2 to 3 days after
the evacuated laser tube is attached to the radium-bearing sidearm. Beyond that period, accumulation in the
discharge tube of the numerous daughter nuclei of radon would present a serious chemical contamination
problem insofar as operation of the radon laser were concerned. Table 8 lists the pertinent wavelengths and
estimated line-strengths for the strongest radon laser lines using xenon as a model.
Radon should be an excellent active laser medium. Its nearest relatives in the heavy noble gases all display
strong, high-optical-gam laser lines, and the relatively great mass of radon reduces its Doppler-broadened
line-width to, for example, 0.77 that of xenon and 0.29 that of neon. The narrowing of the laser line-width
effectively increases the optical gain of the comparable transitions in radon. A further narrowing in the
Doppler-broadened line-width of radon relative to xenon is caused by the lack of hyperfine or mass-shift line
broadening mechanisms in the essentially single-isotope spectrum of radon.
3. Molecular Absorption Experiments.
In striving for simplicity of operation, our experimental laser configurations were designed to permit direct
molecular absorption measurements to be made using selected pollutant gases. Referring to Figures 1 and 4,
when a suitable sample gas was subjected to analysis by the noble gas laser apparatus, two laser wave-
lengths were generally used. One of the laser wavelengths was chosen such that it was absorbed more
strongly than the other by the pollutant gas. The second wavelength then served as an intensity comparison
standard. By measuring the relative intensities of these two laser wavelengths before and after they both
passed through the absorbing sample of gas, it was possible to infer the relative concentration of the
absorbing (pollutant) gas providing that:
(1) the optical path length, R, through the gas was known;
(2) the specific absorption coefficient at the specific wavelength, a( A), was known; and
(3) both the absorbed and comparison wavelengths were accurately measured and, hence, known.
Assuming that the comparison wavelength,*-, wa« negligibly absorbed by the pollutant gas and was initially
of the same intensity, IQ, as the absorbed laser wavelength,A+, then the ratio of the two intensities after
passage through the absorbing gas is simply
€ =— = exp(-2a(X)4i R) (10)
o o
where the ratio, N/NQ, is the fractional molecular density of the pollutant gas in the laser beam. It is evident
from equation (10) that if a measurable intensity ratio, l+/l
-------
As a specific example, using the data of Hans (1971) for the pollutant gas CO, the intensity ratio e = 0 34
when N/N0 = 10-2, R = 10 cm, and a = 5.4 (whenA+= 4.61»m). If the optical path length, R, containing the
polluting gas is increased hy a factor of 10s (from 10 cm to 10«cm), then the fractional pollutant molecular
density, N/NQ, can be reduced by a factor of 105 also to produce a similar value of e, the laser beam intensity
vatin
ratio.
Our electronic data processing electronics used several techniques for comparing the relative intensities of
the two laser lines,A+, andA_. These techniques utilized:
(1) a lock-in amplifier; and
(2) a real-time digital ratiometer.
Both of these instruments were capable of measuring intensity ratios, e, of 0.99 or better, providing that the
detected intensity fluctuations of the two laser beams were less than 10% of full intensity. If this maximum
intensity ratio, e max = 0.99, is inserted in equation (10) using the parameters given above for carbon
monoxide, then the minimum molecular density fraction, N/N , of which the apparatus should be capable of
detecting is reduced to only 8 parts per billion (8 x 10-9).
Since the power supply for our dc solenoid magnet required 480V 30 input power, we were unable to conduct
Zeeman-tuned laser pollutant gas experiments over the long ranges (2 kilometers or more) required for
sensitive operation of the equipment. Our laboratory experiments (Linford, 1973a), however, confirmed the
relative sensitivity of the method for relatively short optical absorption cells (R~15cm), and we feel that the
extrapolations noted above are justifiable. We have conducted a wide variety of noble gas laser experiments
over ranges up to 30 kilometers through the terrestrial atmosphere (Linford, et al., 1974) using a CW xenon
laser similar to that used in our Zeeman-tuned laboratory laser experiments (Linford, 1973a). These long
range laser experiments through the atmosphere included a unique class of experiments which can only be
performed using laser amplifiers having large optical gains, such as the HF/DF chemical laser (Glaze and
Linford, 1973) and the noble gas lasers described in this paper. These unique experiments used an optical
configuration similar to that shown in Figure 1 with the exception that the front flat of the laser was removed.
Laser oscillation was then achieved using the optical feed-back from the remote corner reflector — such an
optical configuration is termed a "long laser" for obvious reasons (Linford, et al., 1974). This method admits
an increased sensitivity to atmospheric absorption above that calculated for the more conventional
apparatus shown above in Figures 1 and 4. The increased absorption sensitivity is obtained because the
absorbing gas is actually placed inside the laser cavity, and if the oscillation properties of the long laser are
adjusted so as to place the laser near its oscillation threshold, then a substantial increase in the detection
sensitivity of the laser can be obtained.
It should be emphasized that not all of the noble gas laser lines listed in Tables 1 through 7 can be used in
either Zeeman-tuned or long laser pollutation monitoring applications. Only those noble gas laser lines
having sufficient optical gain to attain laser oscillation threshold can be used in these relatively high-optical-
loss laser configurations. More than two-thirds of the noble gas laser lines listed above in Tables 1 — 7 lacked
sufficient optical gain to be used in either our Zeeman-tuned or long laser experiments using the relatively
small, one meter, laser tubes constructed for our noble gas laser apparatus. Thus, the use of these more
numerous lower gain noble gas laser lines for pollution monitoring applications is limited to those situations
in which near-coincidences between a lower-gain laser line and pollutant absorption lines or bands occur.
Many of these lower-gain laser lines fall in important bands, as can be seen by comparing Tables 1—7 with
Table 9.
4. Summary of Experimental Results.
Our experiments have been conducted in three major areas:
(1) noble gas laser source study and development (resulted in the discovery and identification of 18 new
noble gas laser lines);
(2) spectral tuning experiments (using the Zeeman effect to tune selected noble gas laser lines over bands
as great as 11 A); and
(3) pollutant absorption measurements (with selected absorbing gases).
Although the latter measurements were conducted using small, laboratory optical cells containing known
quantities of absorbing gases, we concluded that the use of direct optical absorption detection is a sensitive
method of measuring the concentrations and distributions of pollutant gases contained in long optical paths.
Noble gas lasers are particularly useful in this context because:
(1) numerous noble gas laser lines are available in important pollutant bands; and . , , „ .
(2) some of these noble gas laser lines are partially tunable (using the Zeeman effect), thereby allowing
them to be tuned into coincidence with a suitable, nearby pollutant absorption line.
-567-
-------
Since direct optical absorption measurements can be made simply and relatively economically, it appears
that further development of neutral noble gas lasers in air pollution monitoring applications should take
place to permit these potentially useful devices to be more fully utilized. As shown in Figures 5 and 6, the
general spectral coverage of the neutral noble gas lasers is excellent, particularly when compared with the
relatively broad absorption bands of some important pollutants.
REFERENCES
Faust, W. L., et al, (1962), Gas Maser Spectroscopy in the Infrared, Applied Physics Letters, 1,85.
Glaze, J. A. and G. J. Linford, (1973), Design and Performance Characteristics of a Small Subsonic
Flow HF Chemical Laser, Review of Scientific Instruments, 44,600.
Handbook of Lasers, (1971), The Chemical Rubber Co., pp. 183-242, Robert C. Weast, Editor.
Hanst, P. L., (1971), Advances in Environmental Science and Technology, John Wiley and Sons, pp. 91-
213.
Javan, A., et al., (1961), Population Inversion and Continuous Optical Maser Oscillation in a Gas
Discharge Containing aHe — Ne Gas, Physical Review Letters, 6,106.
Linford, G. J., (1973), Experimental Studies of A Zeeman-Tuned Xenon Laser Differential Absorption
Apparatus, Applied Optics, 12,1130.
Linford, G. J.,(1972), High-Gain Neutral Laser Lines in Pulsed Noble Gas Discharges, IEEE Quamtum
Electronics, QE-8,477.
Linford, G. J., (1973b), New Pulsed Laser Lines in Krypton, IEEE Quantum Electronics, QE-9,610.
Linford, G. J., (1973c), New Pulsed and CW Laser Lines in the Heavy Noble Gases, IEEE Quantum
Electronics, QE-9,611.
Linford, G. J., et al., (1974), Very Long Lasers, Applied Optics, 13.
Moore, C. N., (1959), Atomic Energy Levels, Volumes 1,2, and 3, National Bureau of Standards.
Patel, C. K. N., et al., (1962), Infrared Spectroscopy Using Stimulated Emission Techniques, Physical
Review Letters, 9,102.
Peressini, E. R. and G. J. Linford, (1968), Effect of Cross-Relaxation on the Spectral Flux and
Population Inversion Distributions in a CW Laser Oscillator, IEEE Quantum Electronics, QE-4,657.
Racah, G., (1942), On a New Type of Vector Coupling in Complex Spectra, Physical Review, 61,537.
Rasmussen, E., (1931), Das Bogenspectrum derRadiumemanation, Zeitschrift fur Physik, 76,494.
Rasmussen, E., (1932), Uberdas Bogenspectrum derRadiumemanation, Zeitschrift fur Physik, 77,312.
-568-
-------
TABLE 1. Observed Argon Pulsed Laser Lines.
No. (Micrometers)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
L=Low
0.4579
0.4764
0.4879
0.4965
0.5017
0.5145
1.214
1.240
1.270
1:280
1.694
1.792
2.062
2.314
2.397
2.551
2.567
2.715
3.135
3.630
3.700
3.708
3.713
5.020
W=Weak
Transition
(3P)4p2S° 1/2— (3P)4s2l /2
(3P)4p2P°3/2— (3P)4s2Pl/2
(3P)4p2D°5/2— (3P)4s2P3/2
(3P)4p2D°3/2— (3P)4s2Pl/2
(1D)4p2F°5/2— (3P)3d2D3/2
(3P)4p4D°5/2— (3P)4s2P3/2
3d'(3/2)°l— 4p'(3/2)l
3d(3/2)l°— 4p(3/2)l
3d'(3/2)l°— 4p'(3/2)l
3d(5/2)2°— 4p(5/2)2
3d(3/2)2°— 4p(3/2)2
/3d(l/2)l°— 4p(3/2)2\
\3d(l/2)0°— 4p(3/2)lf
3d(3/2)2°— 4p'(3/2)2
3d(l/2)l°— 4p'(l/2)l
3d(l/2)0°— 4p'(l/2)l
5p(l/2)0— 5s(3/2)l°
5p'(l/2)0— 5s'(l/2)l°
5s(3/2)2°— 5p(3/2)l
5p(l/2)l— 5s(3/2)2°
6s'(l/2)l°— 5p'(3/2)l
6s'(l/2)l° — 5p'(l/2)l
4d(3/2)l°— 5p(3/2)l
6s'(l/2)l°— 5p'(3/2)2
6p'(3/2)l— 4d'(5/2)2"
M=Moderate H=High
Current (
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
H
M
M
M
M
M
M
L
TABLE 2. Observed Krypton Pulsed Laser Lines.
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1 C
15
16
17
18
19
20
21
22
23
x
(Micrometers)
0.4762.43
0.7603
0.8115
1.318
1.362
1.443
1.476
1.4962
1.5326
1.685
1.690
1.785
1.819
2.190
2.248
2.523
2.8655
3.774
3.956
4.068
4.142
4.375
4.998
Transition
(3P)5p2D°3/2— (3P)5s2Pl/2
5p(3/2)2— 5s(3/2)2°
5p(5/2)3— 5s(3/2)2°
6s(3/2)l°— 5p(5/2)2
4d(3/2)l°— 5p(5/2)l
6s(3/2)l°-5p(3/2)l
6s(3/2)l°-5p(5/2)2
4d(3/2)l°— 5p(3/2)l
4d(3/2)l°— 5p(3/2)2
4d(7/2)3°-5p(5/2)3
4d(l/2)l°— 5p(l/2)l
4d(l/2)0°— 5p(l/2)l
4d'(5/2)3°— 5p'(3/2)2
4d(3/2)2°— 5p(3/2)2
(8d(5/2)2°— 4f(7/2)3\
)8d(7/2)3°-4f(5/2)2j
4d(l/2)l°-5p(3/2)2
6p(5/2)3— 6s(3/2)2°
7s(3/2)l°— 6p(5/2)2
5d(3/2)l°— 6p(5/2)2
7s(3/2)l°— 6p(3/2)l
7s(3/2)l°— 6p(3/2)2
5d(3/2)l°— 6p(3/2)2
4d'(3/2)l°— 6p(l/2)l
Current <
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
Output Comments
M
M
H
M
H
M
M
W
M
W
M
H
M
M
M
M
H
M
M
W
W
H
M
W
W
W
W
W
M
M
M
M
M
W
W
M
M
H
M
H
M
W
M
M
W
W
W
Argon II
Argon II
Argon II
Argon II
Argon II
Argon II
Argon I
ArgonI
ArgonI
Argon I
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
ArgonI
Output Comments
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
Krypton I
W=Weak
M=Moderate
H=High
-569-
-------
TABLE 3. Observed Pulsed Xenon Laser Lines.
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
L=Low
x
(Micrometers)
0.4954
0.5159
0.5352
0.5971
0.8716
0.9059
0.904
1.6053
1.732
2.027
2.515
2.627
2.652
2.661
2.859
3.108
3.275
3.367
3.508
3.622
3.651
3.680
3.686
3.870
3.895
3.996
4.153
4.539
4.661
5.024
5.575
7.316
9.006
Transition
Unknown
Unknown
Unknown
(1D)6p2P°3/2— (1D)6s2D3/2
(3P)6p4D°3/2— (8P)5d2P3/2
Unknown
6p(5/2)2— 6s(3/2)2°
7s(3/2)l°— 6p(3/2)2
5d(3/2)l°— 6p(5/2)2
5d(3/2)l°— 6p(3/2)l
7d(7/2)4°-7p(5/2)3
5d(5/2)2°— 6p(5/2)2
5d(3/2)l°-6p(l/2)0
5d'(3/2)l°— 6p'(l/2)0
7p(3/2)2— 7s(3/2)2°
5d(5/2)3°— 6p(3/2)2
5d(3/2)2°— 6p(l/2)l
5d(5/2)2°— 6p(3/2)l
5d(7/2)3°— 6p(5/2)2
5d'(3/2)2°— 7p(3/2)2
7p(l/2)l— 7s(3/2)2°
5d(l/2)l°— 6p(l/2)l
5d(5/2)2°— 6p(3/2)2
5d'(5/2)3°— 6p'(3/2)2
5d(7/2)3°— 6p(5/2)3
5d(l/2)0°— 6p(l/2)l
5d'(5/2)2°— 7p(3/2)l
5d(3/2)2°— 6p(5/2)2
5d'(3/2)2°— 6p'(l/2)l
5d'(5/2)2°-6p'(3/2)2
5d(7/2)4°— 6p(5/2)3
5d(3/2)2°— 6p(3/2)l
5d(3/2)2°— 6p(3/2)2
Watts
10
10
102
10-4
10-1
10
10
10-1
10-1
10
10-4
10-2
10-1
10
1.0
5.0
10-3
10-3
10-1
10-"
10-3
10-3
10-2
10-2
10-1
10-1
10-4
10-1
10-4
10-4
Current Comments
H
H
H
H
M
M
L
M
H
H
M
M
H
M
M
M
M
M
M
M
M
L
M
M
H
L
M
M
M
Xenon IV?
Xenon IV?
Xenon IV?
Xenon II
Xenon II
Xenon II?
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
Xenon I
M=Moderate
H=High
TABLE 4. CW Argon Laser Lines.
No.
1
2
3
4
5
Wavelength
(in micrometers)
Transition
1.694
2.061
2.549
2.563
3.133
3d(3/2)f-4p(3/2)2
3d(3/2)§-4p'(3/2)2
5p(5/2)3-3d(7/2)°
6d'(3/2)S-6p(5/2)3
5p(l/2)l -5s(3/2)°
Relative
Output
Power
1.0
4.4
3.5
3.2
6.1
NOTE: The data presented in this Table were taken using the same laser apparatus used to take the xenon
laser data listed in Table 6. The fewer number of CW laser lines obtained with argon and krypton was a
consistent observation made during the course of these experiments.
-570-
-------
TABLES. CW Krypton Laser Lines.
Wavelengths
No. (in micrometers)
1 2.19
2 2.523
Transition
4d(3/2)°-5p(3/2)2
4d(l/2)o-5p(3/2)2
Relative
Output
Power
7.1
5.3
TABLE 6. CW Xenon Laser Lines.
Wavelength
No. (in micrometers)
1 2.026
1 2.627
3 2.651
4 3.108
5 3.275
6 3.367
7 3.508
8 3.622
9 3.652
10 3.680
11 3.870
12 3.997
13 4.153
14 4.539
15 4.611
16 5.024
17 5.575
Transition
5d(3/2)?—6p(3/2)
5d(5/2)°—6p(5/2)2
5d(3/2)°-6p(l/2)o
5d(5/2)°—6p(3/2)2
5d(5/2)°—6p(3/2)l
5d(7/2)°-6p(5/2)2
5d'(3/2)»—7p(3/2)2
1—7s(3/2)°
5d'(5/2)§-6p'(3/2)2
5d'(5/2)°—7p(3/2)l
5d(3/2)§—6p(5/2)2
5d'(3/2)S—6p'(l/2)l
5d'(5/2)°-6p'(3/2)2
5d(7/2)»-6p(5/2)3
Output
Power*
10
3
6
8
10
20
50
2**
1**
12
2**
6
4
1,10+
3
5
9
* In Tables 4,5, and 6, each output power unit =100 fj. watts.
** These lines had relatively low optical gains, and as a consequence required
special efforts to obtain oscillation with the one-meter laser tubes.
+ The indicated output power could be obtained only when the Littrow prism
was used to suppress most of the laser oscillation at the competing 3.508fj.m
xenon laser line.
-571-
-------
TABLE 7. Recently Discovered Noble Gas Laser Lines.
No. (Micrometers)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
W=Weak
0.7603
1.4962
1.5326
2.248
2.515
2.715
2.859
3.630
3.700
3.708
3.713
3.774
3.956
4.068
4.142
4.998
5.020
5.024
Transition
5p(3/2)2— 5s(3/2)2°
4d(3/2)l°— 5p(3/2)l
4d(3/2)l°— 5p(3/2)2
8d(5/2)2°— 4f(7/2)3
(7/2)3°— 4f(5/2)2
7d(7/2)4°-7p(5/2)-nn
5s(3/2)2°—5p(3/2)l
7p(3/2)2— 7s(3/2)2°
°— 5p'(3/2)l
M=Moderate
4d(3/2)l°— 5p(3/2)l
6s'(l/2)l°— 5p'(3/2)2
7s(3/2)l°— 6p(5/2)2
5d(3/2)l°-6p(5/2)2
7s(3/2)l°— 6p(3/2)l
7s(3/2)l°— 6p(3/2)2
4d'(3/2)l()-6p(l/2)l
6p'(3/2)l— 4d'(5/2)2°
5d'(5/2)2°-6p'(3/2)2
S=Strong
Element
Krypton
Krypton
Krypton
Krypton
Argon
Xenon
Argon
Argon
Argon
Argon
Krypton
Krypton
Krypton
Krypton
Krypton
Argon
Xenon
Strength
W
M
M
M
M
M
M
W
W
S
M
W
M
M
W
W
W
M
Mode
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
CW,P
TABLE 8. Projected Radon Laser Lines.
Wavelength
No. (in micrometers)
1
2
3
4
5
6
7
5.68
2.68
3.92*
3.63*
6.90
2.56
3.71
8.53*
Transition
6d(7/2)—7p(3/2)2
6d(5/2)§—7p(3/2)l
6d(3/2)°—7p(3/2)2
6d(3/2)°—7p(l/2)l
6d(l/2)°—7p(l/2)l
6d(3/2)°—7p(3/2)l
Xenon
Laser
Analogy
3.37jzm
9.00fim
3.275/im
Theoretical
Line
Strength
54.0
40.0
29.4
18.9
14.4
12.5
10.0
8.0
* These projected radon laser lines are only approximately correct since the 6d energy levels are
unknown at the present time (Moore, 1959).
-572-
-------
TABLE 9. Absorption Bands of Common Pollutants.
Pollutant
Methane
Ethane
Acetalaldehyde
Formaldenyde
Ammonia
Hydrogen Chloride
Nitric Oxide
Sulphur Dioxide
Carbon Monoxide
Ozone
Propane
Methanol
Ethylene
Propylene
Nitrogen Oxide (No)
Hydrogen Bromide
Butane
Hexane
Pentane
Nitrogen Dioxide
Acetylene
Spectral Band
(Micrometers)
3.1-3.4
7.3—8.0
O.^"""™"O.O
3.2—3.8
3.1—3.7
10.2—11.2
3.1—3.7
5.1—5.6
3.8—4.2
7.2—7.5
4.4—4.9
9.3—10.0
3.3—3.6
9.5—10.2
10.1—11.2
3.2—3.5
5.0-5.4
3.9-4.2
3.3—3.6
3.3—3.5
3.3—3.6
3.4—3.5
2.9—3.1
Structure
Fine
Fine
Moderate
Fine
Moderate
Fine
Fine
Moderate
Fine
Moderate
Coarse
Fine
Moderate
Moderate
Moderate
Coarse
Coarse
Coarse
Fine
Fine
-573-
-------
3/4 METER SPECTROMETER
o\
PULSED ANDCW
LASER EXCITER
LITTROW
PRISM
CRYOGENIC
PRESSURE
CONTROL
i
\
o o\o o
PULSED AND
DIRECT CURRENT
MAGNET SUPPLY
MICROVOLTMETER/LOCK-IN
AMPLIFIER
.**
COOLED
PHOTODETECTOR
REMOTE
/CORNER
REFLECTOR
RECOLLIMATING MIRROR
POLLUTED AIR OR ABSORBING
GAS SAMPLE
Figure 1. Schematic of Pulsed and CW Noble Gas Laser Monitoring System.
-------
(nf)
UPPER LASER LEVEL
(nd)-
\
LOWER LASER LEVEL
LOWER CASCADE
LEVEL, (n+ 1)s
- A1Q
OPTICAL
CONNECTIONS TO
GROUND STATE
GROUND STATE (np)
-W,
03
_L
ARGON:
KRYPTON:
n = 3
n = 4
XENON:
RADON:
n = 5
n = 6
Figure 2. Noble Gas Laser Energy Diagram.
-575-
-------
IONIZATION
LIMIT
en
-------
LASER TRANSCEIVER
'APPARATUS
ABSORBING GAS (ABSORPTION COEFFICIENT = 7 )
APERTURE OF
REMOTE CORNER
REFLECTOR
u
'-^"
r
R
01
-a
-a
RECOLLIMATED
APERTURE OF
LASER TRANSMITTER
a2
IR
> 1
Figure 4. Schematic of Remote Monitoring Laser Apparatus.
-------
10^
o
I '°2
g
z
CC 1
LU 10'
O
Q.
; 5
oo t 1
0
£ 10~1
10"2
in~3
1
—
••Ml*
—
ON LINES
1'
I
|
— 1
t
\
I
MEUTRALATOM LINE
|
\ 1 1
1.0
2.0 3.0 4.0 5.0
WAVELENGTH, IN MICROMETERS
6.0
7.0
8.0
Figure 5. Spectral Distribution of Pulsed Noble Gas Laser Lines.
-------
10 '
10'2
w
H- t
\- Kp*
^T
Ik
CC
UJ
Q.
1-
D
i Q-
& 1-
? 0 10'5
g
O
111
O
2 io~6
UJ
<
io-7
m-8
-
—
—
I i
I
i
I
I
i
I
1
I
i
i
I
i
I
i
i
!
I
I
I
I
I
I
I
I
I
I
'
CW XENON
CW ARGON AND LASER
t%^K. LINES\
\! N
i I
KEY
I I
1.0
2.0 3.0 4.0 5.0
WAVELENGTH, IN MICROMETERS
6.0
7.0
8.0
Figure 6. Spectral Coverage of CW Noble Gas Laser Lines.
-------
ENRICHED STABLE ISOTOPES OF THE NOBLE GASES AND THEIR USES
C.-F.Eck
Monsanto Research Corporation
Mound Laboratory*
Miamisburg, Ohio 45342
Abstract
Isotopes of the noble gases were initially discovered in the early part of this century. Sinte then the natural
abundances of 23 stable isotopes have been measured. Two of these isotopes occur naturally at above 99 mole
% isotopic concentration. For enrichment of noble gas stable isotopes, thermal diffusion is the common
separation method. Six of the isotopes have been enriched to 99 mole %. Most of the others have been enriched
in varying degrees ranging from 20 to 95 mole %. The uses of these isotopes are varied. Natural helium and
argon, which are greater than 99% helium-4 and argon-40 respectively, are used widely in industry. The others
are used in limited quantities. Some of the unique uses are in low-temperature refrigeration, gas tagging, and
chemical analysis of surfaces. The isotopes are used in research and development work, directly in
experiments and indirectly as components of systems for performing experimental work. Some of the common
uses include lasers, preparation of radioisotopes, and studies of nuclear reactions, properties, and isotope
effects, as well as isotope dilution and tracer work.
INTRODUCTION
Noble gases are an interesting group of elements. They are the most recently discovered natural stable
elements. At that time, the end of the last century, radioactivity was also discovered, and there was great
interest in identifying the various radioactive materials and their decay products. Spectroscopic studies were
made of numerous materials. Among these was neon. In 1912 Sir J. J. Thomson studied its spectral lines and
noticed two distinct lines. These were provisionally identified as neon-20 and neon-22, and called isotopes by
Soddy. Thus, one of the noble gases, discovered only 14 years earlier played a part in demonstrating the
existence of stable isotopes. Since then all the isotopes of the noble gases have been discovered. High
enrichments of many of these isotopes have become available, and they are being used for many purposes.
This paper reviews the noble gas stable isotopes, their separation methods, current separation and
availability, and their uses.
After Thomson discovered two isotopes of neon, the spectra of many elements were studied. By 1922, Aston
reported a provisional third isotope of neon, two isotopes of argon, all six isotopes of krypton, and seven
isotopes of xenon with two provisional additional ones. Helium-3 was produced artificially and identified
before it had been found in natural helium. In 1934 Vaughn measured its natural abundance to be less than
one part in 35,000.
In 1939 Alvarez, etal., showed that helium-3 isotopic abundance was greater in atmospheric helium than in
gas-well helium. Table 1 shows the chronology of the discovery of the isotopes. All of these isotopes are found
in air. Natural abundance data from Walton, etal., (1966) for neon and from Lederer, et al., (1967) for the other
noble gases are given in Table 2.
Another source of some of these stable isotopes is the nuclear reactor industry. Materials from this source
are isotopes of helium, krypton, and xenon. The isotopes of the first element result from excited-state lithium-7
disintegrating into helium-4 and tritium which in turn decays to helium-3. The helium-3 is separated from the
resulting helium-3/helium-4 mixture. Isotopic abundances of fission product krypton and xenon are given in
TableS.
As noted in Table 2, two of these 23 stable isotopes, helium-4 and argon-40, already occur in nature at
concentrations above 99 mole %. The other isotopes vary in isotopic concentrations from 630 ppm for argon-38
to 90.5 mole % for neon-20.
SEPARATION METHODS AND ENRICHMENT
Various methods have been tried to separate the noble gas stable isotopes. They include charcoal
adsorption, diffusion, distillation, electromagnetic separation.'gas centrifugation, gas chromatography, glow
discharge, superfluidity, and thermal diffusion. The one commonly used is thermal diffusion.
Helium isotope separation studies have included diffusion, distillation, superfluid flow, and thermal
diffusion. Natural abundance of helium-3 is so low that the helium mixture from the nuclear industry is used
as feed to enrich helium-3 to above 99% by thermal diffusion (Anon, 1963 and Baker, 1965).
Early separations of neon-20 and -22 were achieved by diffusion through porous walls and distillation. Later
Dickel, et al., (1940) devised the thermogravitational column, and used the thermal diffusion effect to separate
neon-20 and neon-22. This column consisted of a vertical glass tube with cooling water on the outside and with
a hot electrically heated center wire. The temperature gradient between the hot wire and cool wall caused a
concentration gradient, which was increased by the convection effect in the vertical system. This way they
*Mound Laboratory is operated by Monsanto Research Corporation for the U.S. Atomic Energy Commission
under Contract No. AT-33-1-GEN-53,
-580-
-------
obtained high enrichments and reasonable separation rates. Subsequently, Clusius, et al, (1956a) used an
auxiliary gas, deuterated methane, in a thermal diffusion column to separate the middle isotope neon-21 of
the ternary system.
Methods used for separation of argon isotopes were Hertz diffusion method in flowing mercury vapor
thermal diffusion, and distillation. Thermal diffusion by Clusius, et al., (1956b) enriched argon-36 to above
99% at reasonable separation rates. The middle isotope, argon-38, was also isolated above 99% in the thermal
diffusion column by using the auxiliary gas, deuterium chloride.
Krypton has six isotopes; they are not as easily separated as the binary and ternary mixtures. Gas
centrifugation and thermal diffusion have been used. Clusius, et al., (1943-3) enriched krypton-86 to 99.5% and
krypton-84 to 98.2% by thermal diffusion. Using the same method Blais, et al., (1956) enriched krypton-78 to
44.9% and krypton-80 to 69.8%. Using hydrogen bromides as auxiliary gases in a thermal diffusion column
Clusius, et al., (1957) enriched krypton-78 to 60.5%.
Xenon's nine isotopes with their high mass values are more difficult to separate than the krypton isotopes.
Groth, et al., (1939) tried to separate these isotopes in a direct current glow discharge unsuccessfully. The gas
centrifuge was used with some success by Beyerle, et al., (1949) and Hertz, et al, (1954). Using a thermal
diffusion column 48 meters long, Clusius, et al., (1955) were able to enrich xenon-136 to 99%. Hechtl, et al,
(1967) used electromagnetic separation to isolate small quantities of xenon-131 to 99%. A production method
used to obtain a stable xenon isotope was irradiation of iodine-127 to make xenon-128 of over 99%
concentration (Haase, 1970). Xenon-124, which has a low natural abundance, has been enriched to 65% by
thermal diffusion (Mound, 1969). During this enrichment, the isotope xenon-129 was concentrated to 60%. The
preceding isotopes have been enriched using natural abundance feed gases. Fission xenon has been used as
feed to a thermal diffusion system to enrich xenon-131 to 60% in liter quantities (Mound, 1967).
Table 4 shows the chronology of enrichment of noble gas stable isotopes. Table 5 shows the status and
enrichment of these various isotopes. The scale of availability is divided roughly as follows:
Preparative — liter (STP) quantities
Experimental — milliliter (STP) quantities
As shown in the table, only six of the isotopes are available in liter quantities at enrichment of 99% or more. All
of the preparative scale enrichments are being done by thermal diffusion at Mound Laboratory. Subsequent to
the experimental development of the thermogravitational column by Clusius, Mound Laboratory used it for
helium-3 enrichment, and later expanded its use to cascaded systems for the rest of the noble gas isotopes.
Time-wise, these stable isotopes have not been available for more than 1-2 decades. Since then, many uses
have developed in research and development work and some in specific applications. The uses of helium-4 and
argon-40 will not be discussed since they have been available naturally for over half a century at high
enrichments, and their uses have been reported widely.
PRESENT USES OF THE NOBLE GASES
Helium-3 is being used in a variety of ways. Its unique properties have made it the object of many studies in
low-temperature physics and related work. As the most fugacious gas known with a normal boiling point of
3.2°K, helium-3 has been useful as a refrigerant in cryostats to achieve temperatures less than 1°K. (Cowen,
1964, Fruneau, et al, 1967). For relatively large loads, a recirculating helium-3 refrigerator has been developed
(Wilkes, 1972). Because of its high vapor pressure, 10,000 times greater than that of helium-4 at 0.5°K (Keller,
1969), helium-3 has been used as a vapor pressure standard temperature scale for low-temperatures (Sydoriak,
1964). The development of a dilution type refrigerator using helium-3/helium-4, and the effect from phase
change, has allowed the maintenance of temperature at millidegrees Kelvin (Hall, 1966, Radebaugh, et al,
1971). In physics of quantum phenomena at low-temperature, helium-3 has been used for considerable study of
itself to clarify its uniqueness (Keller, 1969). This has included properties of liquid helium-3, critical
phenomena, and study of dilute solutions of helium-3/helium-4.
In nuclear work where helium-4 causes interference in measurement of deuterium, helium-3 is used, for
example, as a scattering gas in measuring total collision cross sections at thermal energies by molecular-
beam scattering of deuterium by helium (Moore, et al, 1967). A large use is for neutron detection and neutron
spectrometry. Upon exposure to neutrons, helium-3 reacts to form tritium. The cross section for this reaction
varies smoothly over a wide range and has a large value of 5,400 barns for thermal neutrons (Price, 1964).
Also, there are no excited products, only tritium and a proton. The tritium decays with the emission of beta
particles; these particles can readily be detected on a counter (Heiberg, 1967 and Myers, 1967). Helium-3 has
advantages of sensitivity, high-temperature efficiency, and moderate voltage requirements at increased
pressures. Neutron detectors filled with this isotope are used in applications such as oil well logging (Eaton, et
al, 1969). Helium-3 is also useful for neutron spectrometry (Batchelor, et al, 1955; Manning, et al, 1965; and
Sayres, 1964).
An interesting application is its use in activation analysis. Helium-3 ions require less energy than protons
as a bombarding particle and it has found use in analyses for oxygen (Lyon, 1964; Vialatte, et al, 1972),
fluorine (Lee, et al, 1971), and carbon (Sanders, 1971). These ions do not penetrate deeply, permitting surface
analysis only. Sensitivities of this method for many elements have been determined (Ricci, et al, 1967 and
Lutz, etal, 1972).
-581-
-------
Radioisotopes are being used increasingly in medical studies. Helium-3 as a bombarding particle serves for
surface analysis as noted above, but also, the same helium-3 proton reaction serves to make fluorine-18 from
oxygen-16. The fluorine-18 in this case is used in bone scanning for diagnostic study (Blau, 1962 and Tilbury,
etal., 1970). Other radioisotopes made by using helium-3 as a projectile are iron-52 for metabolism and bone
marrow function, selenium-73 for sulfur substitute, rubidium-81 for splenic volume, and iodine-123 for thyroid
and blood metabolism studies (Jester, 1972).
There are many methods for leak detection in vessels. Turner (1968) has reported a unique method for using
helium-3 for large vessel leak measurement. Injection first of a known amount of helium-4 is made into a
vessel pressurized with a carrier gas, and later, similarly, helium-3 is injected. The isotope ratio allows
calculation of the percent leakage. This is useful when vacuum leak testing is inappropriate and accessibility
a problem.
A recent use of noble gases is in lasers. Helium-neon gas lasers have found many applications in industry
and scientific research, such as mechanical alignment, surveying, seismometry, holography, polarimetry,
interferometry, and optical communications. These lasers can be improved by the use of the enriched isotopes,
helium-3 and neon-20. At maximum output, helium-3 in place of helium-4 increases the power of such a laser
by about 25% (White, 1963). Use of neon-20 and neon-22 allows the shifting of the maximum of the gain curve
to the desired frequency of the laser beam (Schweitzer, 1968). In addition to uses in lasers, neon isotopes are
used in nuclear reaction studies, nuclear magnetic resonance studies, and property measurements. Due to the
ternary mixture of natural neon, it has been suggested that the normal boiling point of neon-20 be used to
define a fixed point on the temperature scale (Furukawa, 1972).
The most interesting use of argon isotopes is the application of argon-38 to the dating of meteorites and
other minerals. In this method, the decay of potassium-40 to argon-40 is measured by analysis of the resulting
product using the isotope dilution method (Rankana, 1963). The final result for radiogenic argon is adjusted
for the presence of atmospheric isotopes, such as argon-40, by the use of isotope ratios resulting from injecting
argon-36 and argon-38 tracers. The argon-38/argon-36 ratio shows the atmospheric contaminant, and the
argon-40/argon-36 ratio allows calculation of the radiogenic argon-40 quantity (Dalrymple, 1969). Large
numbers of tracers are used in an experiment, but each tracer aliquot is small, being 10-5 STP milliliters
(Reynolds, etal., 1966).
Specific uses of krypton isotopes include the international wavelength standard, leak location in nuclear
fuel elements, and preparation of radioisotopes. After study of several isotopes (Bruce, et al., 1961), a multiple
(1,650,763.73) of the wavelength in vacuum of the orange light of krypton-86, given off at the triple point (63°K)
of nitrogen, has been used as the international standard for the length of the meter (Baird, et al., 1961). This
allows each laboratory to set up its own unit for this standard if desired. Krypton isotopes are used
increasingly in the preparation of radioactive isotopes. Krypton-84 undergoes the proton, neutron reaction to
makerubidium-84, which is used in myocardium scanning (Jester, et al., 1972). Preparation of other rubidium
and strontium isotopes are being considered.
One of the problems in operating a fast breeder reactor is to determine which fuel elements have developed a
leak, when a leak is detected. This allows planning for replacement of the leaky fuel pin. A method using
stable isotope tags has been developed by Argonne National Laboratory for the EBR-II fast reactor (Walker,
et al., 1969). In this method, a specific volume with a unique ratio of stable isotopes (xenon-124, xenon-126,
xenon-128, and xenon-129) is injected into a feed element and the element is sealed (Henault, 1969). If at some
future time, the element develops a leak, the gas with the unique ratio of isotopes goes into the cover gas of the
reactor. A mass spectrometer analysis of the recovered tag gas establishes the element that is leaking
(Henault, et al., 1970a,b). Analysis of tag burnup has been made and includes the study of krypton isotopes
(Boulette, 1972). The light krypton and xenon isotopes allow a sufficient number of unique isotope ratios to
provide the necessary location points in the core of a fast breeder reactor. For the Fast Flux Test Facility being
built at Hanford, this method has been refined to the use of a krypton capsule and a xenon capsule, each
containing a specific volume and a unique ratio of isotopes (Cash, et al., 1972). These capsules are inserted into
a fuel element. The element is sealed, and the capsules punctured. The identification of a leak is made in the
same way as with the EBR-II.
In addition to fuel element gas tags, xenon isotopes are used for radioisotope manufacture, nuclear physics
studies, and scintillation counters. Xenon-124, when irradited with neutrons, is converted to iodine-125
(Jester, et al., 1972). This radioactive isotope is particularly useful as a portable x-ray unit (Graham, 1970). It is
also used in medical tests on thyroid function. Xenon-136 in a lamp would emit a normal frequency with
extreme sharpness because the hyperfine structure is missing (Clusius, et al., 1955).
This has been a brief survey highlighting some of the unique ways in which stable isotopes are used. In
addition to these specific uses of noble gas isotopes, they are used in the basic study of their properties
(physical, thermodynamic, and transport), in the study of reaction phenomena (energy levels, cross sections,
yields), and for analytical standards.
REFERENCES
Alvarez, L. W. and R. Cornog(1939), Helium and Hydrogen of Mass 3, Phys. Rev., 56, Ser. 2 613.
Anon(1963),ThermalDiffusion Recovers Helium-3 Isotope, Chem. Eng., 70, Nov. 25 1963 64-6
Aston, F. W.(1922),List of Elements and Their Isotopes. Proc. Phys. Soc. London 34 197'
-582-
-------
Baird, K. M. and D. S. Smith(1962), Primary Standard of Length, J. Opt. Soc. Am., 52,507-14.
Baker, P. S.(l965).Research Materials Coordination and Planning Meeting, ORNL-TM-1047 Oak Ridge
National Laboratory, Oak Ridge, Tennessee. '
Batchelor, R., R. Aves and T. H. R. Shyrme(1955),Helium -3 Filled Proportional Counter for Neutron
Spectrometry, Rev. Sci. Inst, 26,1037-47.
Beyerle, K., et a/.(1949),Gas Centrifuges: Enrichment of Isotopes of Xenon, Krypton, andSelenium with
the Centrifugal Process, Chem. Ing. Tech., 21,331-4.
Bezella, W. A..(196S).Analysis of the Fission Gases Released Within Spent Yankee Fuel Rods USAEC
Report WCAp-6087, Westinghouse Electric Corp., February, 1969.
Blais, N. C. and W. W. Watson(1956),(d.p.) Reactions with Krypton-78 and Krypton-80, Phys. Rev., 104,
$j£t*
Blau, M., W. Nagler and M. A. Bender,(l962),Fluorine-18: A New Isotope for Bone Scanning J. Nucl
Med.,3,332-3.
Boulette, E. T., R. E. Schenter and F. Schmittroth(1972), Gas Tag Burnup Analysis for Fuel Failure
Location and Detection in the FTR, HEDL-TME-72-93, Hanford Engineering Development Laboratory,
Richland, Washington.
Bruce, C. F. and R. M. Hill(1961).Wavelengths of Krypton-SB, Mercury-198, and Cadmium-114 Austral
J. Phys., 14,64-88.
Cash, R. J., F. E. Holt and Y. S. Tang.(1972), FFTF Fuel Failure Location System Recommendations
and Rationale, HEDL-TME-72-96, Hanford Engineering Development Laboratory, Richland, Washington
and Westinghouse Advanced Reactor Design.
Clusius, K., H. H. Buhler, H. Hurzeler and E. Schumacher, (1955), The Separation Tube. XVI.
Preparation of the Isotope Xenon-136,Z.Naturforsch.,l9a, 809-14.
Clusius, K. and G. Dickel(1942-3), Separation Tube. IV. Preparation of the Isotopes Krypton-84 and
Krypton-86, Z. physik. Chem., B 52,348-68, and B 53,178.
Clusius, K., M. Huber, H. Hurzeler and E. F. Schumacher (1956), Preparation of the Rare Isotope
Neon-21, J. Chem. Phys., 24,167-8.
Clusius, K. and H. U. Hostettler(1957), The Separation Tube XIX, The Enrichment of the Isotopes
Bromine-79andBromine-81,Z. Naturforsch, 12a, 974-82.
Clusius, K., E. Schumacher, H. Hurzeler and H. V. Hostettler(1956b), Separation Tube. XVIII.
Preparation of the Pure Rare Isotopes of Argon, Argon-36, and Argon-38,Z. Naturforsch, lla, 709-15.
Cowen, J. A.(I964),Simple Glass He-3 Cryostat for EJR andNMR, Rev. Sci. Inst., 35,914-5.
Dalrymple, G. B. and M. A. Lanphere(1969), Potassium-Argon Dating, W. H. Freeman and Co., San
Francisco, California.
Dickel, G. and K. Clusius(1940),The Separation Tube. III. Production of 2.5 Liters of Pure Isotope Neon-
22 andNeon-20, Z. physik Chem., B 48,50-62.
Eaton, E. and D. Leslie(1969),ATwcfear Radiation Techniques in Oil Well Logging, World Petroleum, 40,
54-60.
Fruneau, M., Lacaze and L. Weil(1967),Helium-3 Cryostat with Adsorption Pump, Cryogenics, 7,135-7.
Furukawa, G. T.(l972),Vapor Pressures ofNeon-20 andNeon-22, Metrplogia, 8,11.
Graham, D.,(1970), Use of Radioactive Iodine-125 for Dental Identification in Mass Disasters, J. of
Forensic Sci., 15,553-64.
Groth, W. and P. Harteck(1939), Experiments on Separation of Isotope Molecules in a Direct-Current
Glow Discharge, Naturwissenschaften, 27,390-1.
Basse, D. W.(1970),Xenon-128 Production Process Development, DUN-SA-145, Douglas United Nuclear,
Inc., Richland, Washington.
Hall, H. E., P. J. Ford and K. Thompson(1966), Helium-3 Dilution Refrigerator, Cryogenics, 6,80-8.
Hechtl, E. and F. P. Vierbock,(1967), Electromagnetic Separation of Highly Enriched Noble Gas
Isotopes in Multimilligram Quantities, Nucl. Instr. Methods, 48,323.
Heiberg, E.,(1967), The Helium-3 Counter as a Detector of Time-of-Flight Measurements of Thermal
Neutron Spectra, Nucl. Instrum. Methods, 52,175-6.
Henault, P. B., G. J. Bernstein, L. F. Coleman, E. R. Ebersole, W. V. Larson and W. W. Smith,
(1969), Identificationof Failed Fuel Elements by XenonTag, Trans. Am. Nucl. Soc., 12,102-3.
Henault, P. B.,E.R. Ebersole, M. T. Laug, R. Villarreal and D. E. Walker, (1970a), Progress in
Xenon Tagging, Trans. Am. Nucl. Soc., 13,797-8. . .
Hertz, G. and E. Nann(1954). New Results on the Isotope Separation of Heavy Gases in the Gas Ultra
Centrifuge. Z. Elektrochem., 58,612-16.
Keller, Vi.E.(1969),Helium-3 andHelium-4, Plenum Press, New York, New York.
Jester, W. A., G. L. Jackson, B. K. Lee, J. S. Burkle and D. Jarrell,(1972),Mtcfear Engineering
ApplicationsinMedicalScience,Iaot.andEad.Tech.,9,Smnmerl972,485-99. , „,. . T, „,., ,
Lederer, C. M., J. M. Hollander and I. Perlman(1967), Table of Isotopes. Sixth Edition. John Wiley and
Sons, Inc., New York, New York. „ .,,,.-• ™ • A i •
Lee, D. M., J. F. Lamb and S. S. Markowitz(1972),«apid Non-Destructive Method of Fluorine Analysis
by Helium-3 Activation, Anal. Chem., 43,542-5. . , . . „.,,. ,
Lutz, G. T., R. J. Boreni, R. S. Maddock and J. Wing, (1972), Activation Analysis - A Bibliography
through 1972, NBS Tech. Note 467, U.S. National Bureau of Standards, Washington, D.C.
-583-
-------
Lyon, W. S., Jr.(1964),Guide to Activation Analysis, D. Van Nostrand Company, Inc., Princeton, N. J.
Manning, A. W., T. B. Jeter, H. Schuckler and W. Sluck(l965),Helium-3 Spectrometer Characteristics
and Applications, Nucleonics, 23, April, 1965,69-71.
Moore, G. E., S. Datz and F. van der Valk(1967), Total Scattering Cross Sections of Hydrogen and
Deuterium by Helium, J. Chem. Phys., 46,2012-13.
Mound(1967), Stable Gaseous Isotope Separation and Purification, April-June, 1966, MLM-1353, Mound
Laboratory, Miamisburg, Ohio.
Mound(1969), Stable Gaseous Isotope Separation and Purification, July-September, 1968, MLM-1529,
Mound Laboratory, Miamisburg, Ohio.
Myers, J. E.(1961), High-Temperature Helium-3 Detectors, IEEE Trans. Nucl. Sci., 14,416-21.
Price, W. J.(l 964), Nuclear Radiation Detection, McGraw-Hill Book Co., New York, New York.
Radebaugh, R. and J. D. Siegworth(1971),Dz7ufr'on Refrigerator Technology, Cryogenics, 11,368-84.
Rankama, K.(1963), Progress in Isotope Geology, Interscience Publishers, New York, New York.
Reynolds, J. H. and R. Spira(1966). Individual Tracers for Isotope Dilution Analysis of Rare Gases,
Nucl. Instr. Methods, 42,225-8.
Ricci, E. and R. L. Hahn(1967), Sensitivities for Activation Analysis of 15 Light Elements with ISMeV
Helium-3 Particles, Analytical Chem. 39,794-7.
Sanders, W. M.(1971), Techniques for Active Helium-3 Activation Analysis for Carbon and Oxygen, LA-
4833, Los Alamos Scientific Laboratory, Los Alamos, New Mexico.
Sayres, A.(1964), Helium-3 Fast Neutron Spectrometer with Extended Energy Range; Trans. Am. Nucl.
Soc., 7,372-3. o
Schweitzer, W. G., Jr.(1968), Saturated Absorption by Neon Inside a 6328 A Laser with a Mixture of
Neonlsotopes in its Gain Tube, App. Phys. Let., 13,367-8.
Sydoriak, S. G.(1964), 1962Helium-3 Scale of Temperature, J. Res. Nat. Bur. Stand., 68 A, 547-88.
Tilbury, R. S., J. R. DahL, J. P. Mamacos, and J. S. Laughlin(1970), Fluorine-18 Production for
Medical Use by Helium-3 Bombardment of Water, Int. J. Appl. Radiat. Isot., 21,277-81.
Turner, S.(19QS),Measurement of Leakage from Large Pressurized Components Using Helium Isotopes,
TRG Report 1195(R), United Kingdom Atomic Energy Authority, The Reactor Group, HQ, Risley,
Warrington, Lanes, U. K.
Vaughn, A. L., J. H. Williams and J. T. Tate(1934), Isotopic Abundance Ratios of Carbon, Nitrogen,
Argon, Neon, and Helium, Phys. Rev., 46 Ser. 2, 327.
Vialatte, B. andG. Revel(1972), Determination of Oxygen in Aluminum by Helium-3 Activation, Anal.
Chem., 44,1077-9.
Walker, D. E., P. J. Halfman and J. H. Sanecki(1969),Zeram Tagging of Instrumented Fuel Capsules
for Use in EBR-IL, Nat. Sym. Develop. Irrad. Test Technol., 3rd, 59-70.
Walton, R., Jr. and A. E. Cameron, (1966), The Isotopic Composition of Atmospheric Neon,
Naturforschung, 21a, 115-119.
White, A. D.(1963), Increased Power Output of the 6328 A GasMaser, Proceedings of the IEEE, 51,1669.
Vfilkes,W.Rj(1972), A 200 mWRecirculating Helium-3 Refrigerator, Cryogenics, 12,180-3.
TABLE 1. Chronology of Discovery of Noble Gas Stable Isotopes.
1919 Neon-20 1922 Xenon-124
Neon-22 Xenon-126
1920 Helium-4 1923 Xenon-128
Argon-36 Xenon-130
Argon-40
Krypton-78 1928 Neon-21
Krypton-80
Krypton-82 1934 Argon-38
Krypton-83
Krypton-84 1939 Helium-3
Krypton-86
Xenon-129
Xenon-131
Xenon-132
Xenon-134
Xenon-136
-584-
-------
TABLE 2. Natural Abundance of
Stable Isotopes of Noble Gases.
TABLE 4. Chronology of Enrichments
of Noble Gas Stable Isotopes.
Isotope
Helium-3
Helium-4
Neon-20
Neon-21
Neon-22
Argon-36
Argon-38
Argon-40
Krypton-78
Krypton-80
Krypton-82
Krypton-83
Krypton-84
Krypton-86
Xenon-124
Xenon-126
Xenon-128
Xenon-129
Xenon-130
Xenon-131
Xenon-132
Xenon-134
Xenon-136
Natural
Abundance
(Mole%)
1.3 xlO-4
99.9999
90.5
0.27
9.23
0.337
0.063
99.600
0.35
2.27
11.56
11.55
56.90
17,37
0.096
0.090
1.92
26.44
4.08
21.18
26.89
10.44
8.87
Year Isotope
1940
1942
1955
1956
1957
1967
1969
1971
1973
Neon-20
Neon-22
Krypton-84
Krypton-86
Xenon-136
Neon-21
Argon-36
Argon-38
Krypton-80
Xenon-131
Xenon-129
Xenon-124
Argon-40
Krypton-82
Enrichment
Mole%
99.8(1950:99.95)
99.7(1960:99.92)
98.2
99.5
99.0
99.6
99.9(1962:99.99)
99.6(1959:99.98)
69.8
Krypton-78 60.5
99
60
65
99.98
77
TABLE 3. Fission Noble Gas Isotopes (Bezella, 1968).
Fission Krypton
Fission Xenon
Krypton-78
Krypton-80
Krypton-82
Krypton-83
Krypton-84
Krypton-85*
Krypton-86
Xenon-124
Xenon-126
0.2 Mole % Xenon-128
11 " " Xenon-129
31 " " Xenon-130
0
6 " " Xenon-131 8
52 " " Xenon-132 22
Xenon-134 29
Xenon-136 41
-
15 Mole %
rr n
rt rr
rr n
rt rr
*Radioactive
-585-
-------
TABLE 5. Enriched Stable Isotopes
of Noble Gases.
Isotope
Helium-3
Neon-20
Neon-21
Neon-22
Argon-36
Argon-38
Argon-40
Krypton-78
Krypton-80
Krypton-82
Krypton-84
Krypton-86
Xenon-124
Xenon-126
Xenon-128
Xenon-129
Xenon-131
Xenon-132
Xenon- 134
Xenon-136
Enrich.
Mole%
99.98
99.95
90
99.6
99.9
99.5
95
99.6
99.95
50
60.5
69.8
70
90
98.2
99.
99.5
20
65
10
99
60
60
99
38
27
99
Method
Thermal diffusion
Thermal diffusion
// rr
Thermal diffusion &
auxiliary gas
Thermal diffusion
Thermal diffusion
it n
a n
and auxiliary gas
Thermal diffusion
Thermal diffusion
Thermal diffusion
and auxiliary gas
Thermal diffusion
// n
Thermal diffusion
n a
n n
n n
Thermal diffusion
a n
n n
Irradiation of 1-127
Thermal diffusion
Thermal diffusion
Electromagnetic sep.
Thermal diffusion
a n
n a
Scale
Preparative
Preparative
n
Experimental
Preparative
Preparative
//
Experimental
Preparative
Preparative
Experimental
Experimental
Preparative
Preparative
Experimental
Preparative
Experimental
Preparative
Experimental
Preparative
Experimental
Preparative
Preparative
Experimental
Preparative
a
n
-586-
-------
NOBLE GASES IN NUCLEAR MEDICINE
M. Calderon and J.A. Burdine
Baylor College of Medicine and
St. Luke's Episcopal-Texas Children's Hospitals
Houston, Texas
ACKNOWLEDGEMENTS
We are indebted to Professor Linda Monroe and Jane Demos for assistance in preparation of the manuscript
and illustrations.
Abstract
Radioactive noble gases have made a significant contribution to diagnostic nuclear medicine. In the area of
regional assessment of pulmonary function, 133Xe has had its greatest clinical impact. Following a breath of
133Xe gas, pulmonary ventilation can be measured using a scintillation camera or other appropriate radiation
detector. If 133Xe dissolved in saline is injected intravenously, both pulmonary capillary perfusion and
ventilation can be measured since 90% of the highly insoluble xenon escapes into the alveoli during the first
passage through the lungs. Radionuclide pulmonary function tests provide the first qualitative means of
assessing lung ventilation and blood flow on a regional basis, and have recently been extended to include
quantification of various parameters of lung function by means of a small computer interfaced to the
scintillation camera.
133Xe is also used in the measurement of organ blood flow following injection into a vessel leading into an
organ such as the brain, heart, kidneys, or muscles.
INTRODUCTION
The field of nuclear medicine originated in the early 1920's when George Hevesy and Hermann Blumgart
applied naturally occurring radioisotopes to the study of physiological phenomena. Since the introduction of
the cyclotron and the development of nuclear reactors, the Atomic Energy Commission and other
governmental agencies such as the Food and Drug Administration have provided significant support for
activities in this area of medicine. As a result, in less than a quarter of a century, nuclear medicine has become
the fastest growing major specialty concerned with diagnostic and therapeutic patient care.
The use of radioactive noble gases such as isotopes of xenon and krypton has become increasingly
important in both the research and clinical activities of nuclear medicine. The most widely used gas has been
133Xe, primarily because of its more abundant availability. 133Xe is reactor produced as a fission product of
235U. With a half-life of 5.2 days, this radionuclide decays by beta emission with a maximum energy of 346 keV,
and a primary gamma photon of 81 keV to stable 133Cs. As is true with other noble gases, 133Xe is virtually non-
reactive and freely diffusible. Under high pressure or in a vacuum, the gas can be dissolved in aqueous or
saline solution, and in this form is used for the study of several important physiological parameters. 127Xe has
also been used, and has several significant advantages in terms of physical characteristics such as its pure
gamma emission. However, this material is accelerator produced and therefore has very limited availability.
APPLICATIONS
The two most important applications of radioactive xenon are in the measurement of organ or tissue blood
flow, and the assessment of pulmonary function. When 133Xe is injected into a blood vessel leading into an
organ such as the brain, heart, kidneys, or even muscles, the dissolved gas rapidly diffuses from the blood into
the tissues (Figure 1). The relative distribution between tissue and blood is a constant for the particular
histologic or cellular structure involved, and may be numerically expressed by a partition coefficient. As the
bolus of radioactive gas passes through the organ, equilibrium with the tissue occurs transiently, followed by a
rapid decrease in the blood level of xenon within the organ. The extravascular xenon at this point begins to
diffuse back into the blood, with the rate of xenon clearance from the organ being proportional to the rate of
blood flow. Very little xenon recirculates during the next phase of the cardiac cycle, since more than 90% of the
xenon passing through the lungs is cleared by ventilation. The rate of clearance of the radioactive xenon from
the organ is determined by appropriately positioned scintillation detectors. If the tissue-to-blood partition
coefficient for the particular organ is known, the specific perfusion fraction in ml/gram of tissue can be
calculated from the clearance half-time. The general equation is:
OBF = X loge 2
(ml/g/min) T 1/2 (min)
Where OBF = Organ blood flow
A = Partition coefficient
loge 2
T 1/2 (min)
= Clearance rate
-587-
-------
The concept of exponential clearance is basic to stochastic or kinetic analysis of this type. The formula is
based upon the initial slope of the washout, assuming a single exponential function. Organs such as the brain
with different rates of perfusion to gray and white matter create the need for a multi-exponential approach.
Through these techniques, the first quantitative information concerning regional blood flow has been
obtained in patients with cerebral vascular disease, as well as those with coronary artery disease before and
after infarction.
Of even greater clinical significance has been the study of pulmonary function with radioactive noble gases,
particularly 133Xe. Until such techniques came into being, the regional distribution of blood flow and
ventilation could not be measured by conventional pulmonary function studies. Consequently, the pulmonary
physiologist has been severly limited in the type of information he could obtain concerning the pathogenesis
of pulmonary disease. Because of the amazing functional reserve of the lungs, many patients experience a
disease process which goes undetected for years, until the point at which significant signs and symptoms
occur. Radionuclide studies of pulmonary function have offered the potential for filling most of this void.
When 133Xe is injected intravenously, the gas passes through the heart into the vascular system within the
lungs (Figure 2). At this point, more than 90% of the gas diffuses into the alveoli, the small sacs which
constitute the great majority of the lung volume, and/or the ventilatory unit which is most distal from the
ambient atmosphere. While the patient holds his breath, the initial distribution of injected gas is proportional
on a regional basis to perfusion or blood flow. Utilizing a sophisticated radiation detecting device called a
scintillation camera, the distribution of radioactivity in a patient's lungs is displayed on a cathode ray tube,
and may be photographed. The scintillation camera consists basically of a large sodium iodide crystal viewed
by nineteen photomultiplier tubes which are in turn connected to a x-y pulse locator. This device, by recording
the relative distribution of pulmonary blood flow, enables the physician to diagnose perfusion abnormalities
which have been caused by blood clots or emboli, as well as the destructive changes inherent in pulmonary
emphysema. Such testing is rapid, safe, and relatively simple. Because of these characteristics, the tests are
readily applicable to children. In addition, the information provided is not available through other modalities.
A normal pulmonary perfusion scintiphoto is illustrated in Figure 3. Note the regular distribution of the
injected gas with a gradient that follows the actual hydrostatic distribution of the blood flow.
Figure 4 is a 133Xe perfusion image obtained in a patient with a pulmonary embolus, a condition in which
blood clots in the lungs occlude branches of the pulmonary arteries. This disease has been reported to be the
most common cause of sudden death in hospitalized patients.
In addition to perfusion or blood flow, the other essential mechanism involved in the exchange of oxygen
and carbon dioxide by the lungs is ventilation. Ventilation may be defined as the transfer of gases to and from
the alveoli by the bronchial tree in connection with the ambient atmosphere. This critical parameter of
pulmonary function can be monitored in two ways using radioactive xenon gas. First, the rate of washout of
injected xenon during a perfusion study is directly proportional to ventilation. Any abnormalities of
ventilation will appear on the scintillation camera photographs as retained or "trapped" radioactivity in this
patient.
Ventilation may also be measured by observing the distribution of radioactive xenon in the lungs after a
single breath inhalation of the gas (Figure 5). Ordinarily regional ventilation and, therefore, gas
concentration is relatively uniform, with a slight preponderance toward the lower or basal portion of each
lung.
Patients with chronic obstructive pulmonary disease such as the individual in Figure 6 with emphysema
have an uneven gas distribution produced by airway obstruction and parenchymal destruction. In addition, a
"ball-valve" phenomenon may occur, whereby gas passing into lung regions on inspiration has trouble
exiting during expiration. Such "trapping" is common in these diseases.
Radionuclide pulmonary function tests have provided the first qualitative means of assessing regional lung
ventilation and blood flow, and have been extremely useful clinically. However, quantitative values would be
helpful for research purposes and to follow the clinical progress of a particular patient's disease. In addition,
other parameters of pulmonary function such as regional gas exchange which depend upon the coupling of
ventilation of perfusion are not readily appreciated from the static scan images. To gain important new
information derived from the basic imaging studies, we have designed in our laboratories a multipurpose
minicomputer system for computation of quantitative values of pulmonary function.
The computer system is interfaced directly to the scintillation camera. Appropriate gas dispensing
equipment such as this spirometer is located in proximity to the patient. Raw data are processed by the
computer to obtain indices which express regional ventilation and perfusion per unit volume of lung, with the
information displayed as a functional image of color-coded contours, each contour representing a different
level of ventilation or perfusion. The color sequence begins with the red end of the spectrum as a maximum
value, and decreases to blue for a minimal level. The functional images illustrate that the area of greatest
perfusion in the upright patient is located in the lower portion of both lungs, reflecting the hydrostatic
distribution of the column of blood. Ventilation is also located somewhat more toward the lung base, but
because of pressure differences within the lung parenchyma itself. In a patient with pulmonary embolism or
blood clots within the lung, large perfusion defects are noted due to the clot, with essentially normal
ventilation. In patients with pulmonary emphysema, diffuse abnormalities in ventilation and perfusion are
present.
The physiologic coupling of ventilation to perfusion is critical for adequate exchange of oxygen and carbon
dioxide between the alveoli and blood. The basic abnormality in many lung diseases is an uneven influence of
-588-
-------
these two parameters of pulmonary function. To determine the efficiency of such coupling and therefore to
reflect imbalances, a functional image of the ventilation-perfusion (V/Q) ratio decreases in a stepwise fashion
in the normal patient. In the patient with pulmonary embolism, ventilation predominates in the areas where
blood flow has been obstructed by a clot. In the patient with pulmonary embolism, the normal stepwise
sequence has been disrupted, because of destruction of the lung tissue.
In summary, some of the most important tests in clinical medicine are now based upon the use of noble
gases. With the development of new instrumentation including sophisticated data processing systems, these
important radionuclides will undoubtedly assume an even greater importance in medicine during the next few
years.
REFERENCES
Ball, W. C., et al., (1962), Regional Pulmonary Function Studies with Xenon-133. J. Clin. Invest. 41:519-
531.
Burdine, J. A., et al., (1972), Functional Pulmonary Imaging. J. Nucl. Med. 13:933-938.
Cannon, P. J., R. B. Dell, and E. M. Dwyer, Jr., (1972), Measurement of Regional Myocardial
Perfusion in Man with 133Xe and a Scintillation Camera. J. Clin. Invest. 51:964-977.
Hoffer, P. B., et al., (1973), Improved Xenon Images with mXe. J. Nucl. Med. 14:172-174.
Lassen, N. A., (1967), On the Theory of the Local Clearance Method of Measurement of Blood Flow
including a Discussion of its Application to Various Tissues. Acta Medica Scandanavica Suppl. 472:136-144.
Wagner, H. N., Jr., (1964), Regional Blood Flow Measurements with Krypton-85 and Xenon-133. In
Dynamic Clinical Studies with Radioisotopes, AEC Symposium #3.
a'rterial injection into organ
JnfAT<3fifi?»l anan^ '^*-»-/. • • .
Figure 1. Schematic representation of an intra-arterial injection of 133Xe solution for the measurement of
organ blood flow. The partition coefficient for 133Xe between tissue and blood is constant for the particular
histologic structure.
-589-
-------
2.
exhaled
•injection
Figure 2. Schematic representation of an intravenous injection of 133Xe in saline. More than 90% of the 133Xe
is cleared by ventilation on the first passage through the lungs.
Figure 3. Normal 133Xe pulmonary perfusion scintiphoto recorded during breath holding.
-590-
-------
Figure 4. 133Xe perfusion image in a patient with pulmonary embolism. Arrow indicates an area of
perfusion deficit produced by a blood clot in a branch of the pulmonary artery.
Figure 5. Normal
133Xe pulmonary ventilation
scintiphoto obtained during
breath holding following a
single deep inspiration of
the gas.
-591 -
-------
Figure 6. Ventilatery images in a patient with emphysema. The breath holding image (left) shows areas of
lung destruction, while the image obtained during subsequent ventilatory washout (right) demonstrates gas
trapping indicative of airway obstruction.
- 592 -
-------
HELIUM - ITS CONSERVATION AND ITS POTENTIAL FOR FUTURE YEARS
C.Laverick
Argonne National Laboratory
Argonne, Illinois
ACKNOWLEDGEMENTS
Information from the Department of Interior, Congressional hearings, the excellent opinion of Judge Frank
Theis of the U.S. Kansas District Court on June 11,1973, and the many discussions and reports generated by
the National Science Foundation, Helium Study Advisory Committee (see Figure 2) and Participants, have
been freely used in generating the cross section of the helium problem presented here. In particular, thanks are
due to Harold Lipper, Chief, Division of Helium, U.S. Bureau of Mines; David M. Evans, Director, Potential
Gas Committee, Mineral Resources Institute, Colorado School of Mines; John K. Hulm and H. Richard
Rowland of the Westinghouse Systems Research Directorate; and representatives of the gas companies and
distributors involved. Further information can be obtained from members of the Helium Study Advisory
Committee.
Abstract
Some of the past history of the U.S. helium conservation program is outlined giving the Administration's
reasons for cancelling the helium storage contracts. The helium demand for current technologies is given, and
rough estimates of demands for these and newly developing technologies are presented for future years. The
objectives of the current National Science Foundation's Helium Study are stated, and some developing
conclusions for further discussion are outlined. The dilemma between current economic and conservationist
philosophy is outlined. While a substantial gain in long-term cash return for the program is possible by the
time the stored helium has been sold, the principal concern seems to be the short-term cash flow, with its
substantial borrowings from the Treasury (more than $320 million by June, 1973), due in part to the way the
Interior chose to administer the program, and in part due to the fact that substantial investments are usually
necessary to yield even greater, longer term revenue. Some thoughts on these aspects of the program are
expressed.
While it is inappropriate to second guess NSF helium study conclusions which have yet to be reached, it is
almost certain that helium will be needed in substantial quantities beyond the turn of the century if the
leading world nations are to continue to develop as post-industrial societies. New helium-based technologies
can reduce our energy and resource needs for any given energy base. A conservative estimate of U.S. demand
from 2000 to 2050 A.D. is about 136 billion cubic feet with total world demand for the same period in excess of
500 billion cubic feet. The U.S. has the most substantial world reserves and, with good husbandry, could
supply a significant portion of this demand, with associated favorable international trading aspects. The
desirability of such a situation should be self-evident in view of the current impact of the policies of the oil
exporting nations on life in the United States.
INTRODUCTION
Helium is considered essential to several national programs now under way or being contemplated. These
programs have potential positive consequences for the future well being of our society, and include providing
options in energy conversion, transmission, and utilization, enhancing mass transportation, revolutionizing
ship design, providing space ship propulsion options, enhancing the mass to weight ratio of large aircraft,
developing new techniques for water purification and ore separation, improving instrumentation, and
advancing basic science.
Organizations deeply involved include the National Science Foundation, DOD, DOI, DOT, AEC, OST*. The
electrical power, gas, petroleum, and coal industries, and the Federal Power Commission, Programs already
planned or underway involve commitments in excess of $200 million for superconductivity alone in the next
15 years, and on the order of several hundred billion dollars in the years ahead, when large scale usage
becomes commonplace.
At the moment these uses for new technologies comprise less than 10% of the helium demand, and it is
anticipated that this will continue to be so to the end of this century.
HELIUM DEMAND
1. Established Technologies.
Helium demand since 1960 increased from 420 million cubic feet through a peak of 922 million cubic feet in
1969, due to the vigorous space program (unforeseen when the 1960 Helium Act was passed), to 601 million
cubic feet in 1972.
department of Defense, Department of Interior, Department of Transportation, Atomic Energy
Commission, Off ice of Science and Technology.
-593-
-------
The principal current uses listed for 1970 by the Bureau of Mines, with the volume used in each case in
millions of cubic feet, are:
(a) Aerospace (purging and pressurization), 237;
(b) Inert protective atmospheres, 68;
(c) Research, 65;
(d) Welding, 63;
(e) Lifting gas, 45;
(f) Leak detection, 42;
(g) Low-temperature science and engineering (cryogenics), 33;
(h) Chromatography (for rapid, accurate gas analysis in many major industries), 14;
(i) Heat transfer, 9;
(j) Synthetic breathing mixtures, 4;
(k) Medical and other, 7; and
(1) Exports, 60.
The Bureau of Mines estimates the total demand for these uses, less exports, to be between 1.4 and 3.6 billion
cubic feet in 2000 A.D. This is based on applying an annual 1.6% population increase and 4.5% new
construction, as a matter of judgement, to the 1968 demand base.
Stanford Research Institute uses a different (they claim more accurate) forecasting technique. Assuming an
ample supply at present prices.they predict that annual demand will exceed 3 billion cubic feet by 2000,12
billion cubic feet by 2025, and 44 billion cubic feet by 2050 — an annual growth rate of about 5.5%. This
includes estimates for use in some new technologies. They predict a lower demand, approximately 1 /2 to 1 /3 of
the above, based on assumptions of price escalation. These must not be confused with high- and low-range
demands as used by the Interior. The price is assumed to be $75/Mcf in 1975 and 1985, $150 in 2000, $300 in
2025, and $600 in 2050, in actual, not constant, dollars. Under these conditions, a predicted demand of 2.4
billion cubic feet is expected in 2000,6.2 billion in 2025, and 16.5 billion in 2050.
2. New Technologies.
Helium demand for some, but not all, new technologies is projected by Stanford Research Institute. It is not
considered by the Bureau of Mines. These new technologies involving helium may help ameliorate future
national energy, transportation, and resource problems. Many of these helium uses involve the use of liquid
helium to provide the low-temperature environment (about 4 to 18°K) for superconductors. Superconductivity
is the ability of many materials to transmit electrical energy, or conduct electricity, with no loss except for the
refrigeration energy needed to keep them cold. Some superconductors conduct electricity at no loss even in
high magnetic fields, and are used as windings in the new high-efficiency, high current density,
superconducting magnets, and superconducting rotating machines.
Examples of new technologies using helium, or requiring helium-cooled superconductors, and under active
development are:
Electrical Power. Fusion reactors, MHD peaking plants, large rotating machinery, gas-cooled fission
and fission breeder reactors, liquid metal fast breeder reactors, and superconducting AC and DC urban and
long distance underground power transmission lines.
Transportation. Superconducting generators in high-performance aircraft and space devices, precision
instrumentation, high-speed magnetically levitated ground transportation, flexible high-performance naval
and mercantile marine and submarine propulsion systems, hybrid airship/airplane, high-load capacity
systems.
Industrial, Resource Utilization and Waste Treatment. Ore separation, water purification, rolling
mill motors, low-voltage high-current superconducting electrolysis systems for aluminum plants, or a future
national hydrogen economy system, rain making in arid coastal regions (not superconducting — involves
making a tall helium supported chimney to condense moisture from air).
Computers. Compact, high-density, low-power dissipation, superconducting memories.
THE U.S. HELIUM PROGRAM
The U.S. Helium Storage Contracts were finally formally terminated in December, 1973, after more than
four years debate and litigation.
Early History
As increasing helium demands threatened to outstrip the production capacity of the Bureau of Mines,
concern arose regarding the future ability to meet mobilization needs and the casual wastage of the natural
resource. In 1957, the Director, Office of Defense Mobilization, recommended to the Secretary of Interior that a
working group be established to determine the feasibility of making the conservation aspect of the helium
program the primary, rather than the incidental, objective of the Department of Interior.
-594-
-------
A working group was subsequently established, comprised of representatives from the Departments of
Commerce, Defense, and Interior, Atomic Energy Commission, Office of Defense Mobilization Bureau of the
Budget, and Federal Power Commission. Headed by then Undersecretary of Interior, 0. Hatfield Chilson the
group compiled a study entitled "Cost of Implementation of a National Helium Conservation Policy " (The
Chilson Report, January 24, 1958.) As a result of the study, a national helium conservation policy was
subsequently recommended to President Eisenhower and formally approved on April 25,1958.
Due to the magnitude of expense involved in accomplishing the conservation program, the Report
recommended that:
(1) private industry be encouraged to participate in the program by entering long-term contracts with the
Government which would give the companies a reasonable return on their capital investment and, at the
same time, help to defray the overall costs of the program;
(2) the Secretary be authorized to set the price for the helium at a level which would enable the Department of
Interior to cover all costs incurred in carrying out the provisions of the program;
(3) all federal agencies and contractors purchase all major requirements of helium from the Secretary; and
(4) the private companies participating in the program be prohibited from selling any helium to any
purchaser other than the Secretary at a price lower than the lowest price paid by any government agency.
In 1958 the Department of Interior proposed legislation to Congress for the implementation of the
recommended conservation program. This was not acted upon, however, and redrafted proposals were
subsequently submitted to Congress on July 27, 1959. After lengthy hearings and incidental revisions and
amendments of the proposed bill, a legislative amendment was adopted on September 13,1960, to be known as
"An Act to Amend the Helium Act of March 3,1925, as amended, for the defense, security, and the general
welfare of the United States.'The Act became effective on March 1, 1961, and a limitation on the annual
contracting authority of 47.5 million dollars was established by the Act of August 3,1961 (75 Stat. 246,253).
The 1960 Helium Act Amendments authorized the Secretary of the Interior to enter into long-term contracts
forthe acquisition, processing, transportation, or conservation of helium, not exceeding twenty-five years. In
addition, he was given the power to acquire, by eminent domain, helium contained in helium-bearing natural
gas, and so much of such gas as was necessarily removed in the extraction process; if he were unable to acquire
helium otherwise upon reasonable terms and at the fair market value. In an attempt to make the program self-
liquidating, the Act provided for a comprehensive system similar to that recommended by the Chilson Report.
The Secretary was given authority to set his own price for the helium that he sold, and to establish and
promulgate such rules and regulations, consistent with the directions of the Act, which were necessary to
carry out its provisions.
The helium conservation program apparently ran smoothly until some time in 1968. After the passage of the
1960 Act, however, several private companies had begun their own helium extraction outside of the
government program. Because the Secretary's price for helium reflected not only direct extraction costs, but
also a portion of the costs involved in the administration of the conservation program, the Secretary's price for
helium exceeded the private producers' price. As the private companies' sales increased, the Secretary's sales
decreased, despite increased overall demand. In response to this situation, the Secretary proposed regulations
which would have forbidden government agencies and their contractors from purchasing their major
requirements of helium from any source other than the Secretary, or private companies "eligible" under the
regulations. These regulations were considered illegal in a subsequent court action. The Court noted, however,
that the Secretary's purpose could be accomplished by obtaining an Executive Order requiring that such
provisions be included in all contracts entered into by government agencies. An Executive Order was
requested of the President by the Interior, but was never obtained.
The Secretary's termination statement, which is based on the Department's final environmental statement,
and the comments received from the private contractors, recognized that:
(1) the direct result of the decision to terminate the contracts will be that the Government will forego the
storage of an estimated 20 billion cubic feet of helium which will be "wasted" or "depleted" if the contractors
do not continue to extract and store the helium, prior to delivery of the natural gas to fuel markets;
(2) natural gas — primarily fuel gas — is the only known source of helium other than the atmosphere, an
environmentally and economically inferior source;
(3) the currently proved "helium-rich" natural gas reserves (containing at least 0.3% helium) will probably
be exhausted before the turn of the century;
(4) it is unknown exactly how much, if any, "lean" helium-bearing natural gas will be available after the
year 2000;
(5) present studies predict large demands for helium for environmentally beneficial uses beginning around
the turn of the century which could rapidly deplete the helium supply then in storage; and
(6) assuming a lack of supply at a later time, future environmentally beneficial technological advancements
might be precluded due to unacceptably adverse environmental effects of recovery from other sources.
In contrast to the above possibilities, the Secretary's statement notes that:
(1) government demand for helium has decreased consistently since 1966, and is presently below the 1961
level;
-595-
-------
(2) present technological advancement enables recovery of helium from "lean" sources of helium-bearing
natural gas expected to be available in large quantities in the future;
(3) the Bureau of Mines estimated a total of 17 to 24 billion cubic feet of probable shut-in "helium-rich"
reserves (i.e., found in natural gas fields not being exploited for present fuel purposes);
(4) including the helium currently in storage, in shut-in government reserves, which are expected to be
recoverable from government-owned plants, have total reserves of approximately 44.5 billion cubic feet;
(5) storage of recoverable helium at today's costs does not necessarily assure "low-cost" helium in the future;
(6) the forecasted large demands for the future are "speculative" and of questionable reliability;
(7) government agencies' estimates for future helium demands would not require removal of helium
presently in storage "to meet essential government needs" until well into the next century; and
(8) assuming the estimated total future demands were to materialize at the beginning of the next century, as
predicted, the storage now of an additional 20 billion cubic feet of helium would merely delay, rather than
foreclose, the necessity of obtaining helium from presently unknown or environmentally disadvantageous
sources.
On the basis of the above assumptions, the Secretary determined that there had been a substantial
diminution in helium requirements for essential government activities, and that the termination was
therefore authorized under Section 12.1 of the contracts, since the purposes of the Helium Act Amendments of
1960 had been satisfied.
HELIUM SUPPLY
Future helium supply and discovery must be largely a matter of conjecture. The helium abundance in the
earth's crust is not known while the atmosphere constitutes a virtually infinite supply at the low-
concentration of about six parts per million. The energy investment required to obtain pure helium from the
atmosphere is large, being about 26 gigawatt year^ per billion standard cubic feet (scf) of gaseous helium.
Extraction of helium from gas streams that are not commercially attractive as heating gases is probably not
economic. Helium availability from gases which have a satisfactory thermal content must be expressed in
terms of cost of extraction. The conclusions of David Evans, Potential Gas Committee, in this regard are as
follows:
(1) The proved and probable helium reserves of the United States are estimated to be 161 billion cubic feet at
a price of $35.00 per Mcf. With the exception of 52 billion cubic feet of helium in underground storage or in shut-
in fields (probable supply), this helium will have been wasted with produced natural gas in the next 30 years.
(2) It is estimated that 194 billion cubicfeet of helium is present in gas in the Rocky Mountains area that could
be recovered for from $50 to $80 per Mcf. This includes 3 billion cubic feet of helium in the non-flammable gas of
the Tip Top field, Wyoming, as well as helium contained in natural gas averaging .205 percent helium in the
rest of the Potential Gas Committee Areas Hand I. (See map in Figure 1.)
(3) The estimate for the Tip Top field is based upon one well. The Tip Top structure is large, and future
drilling could prove up ten times more helium than is reported here on the basis of one well.
(4) Helium can be recovered from natural gas containing 0.1 percent helium for about $143/Mcf. There is an
estimated 113 billion cubic feet of helium in gas of this concentration in the eastern part of the U.S. (Areas A,
B, andC).
(5) Another 375 billion cubic feet of helium is contained in the natural gas of the Midcontinent Region (Areas
D, J-North, and J-South). The helium concentration of this gas averages .091 to .094 percent. Since this
concentration is only slightly less than .1 percent, which costs $143/Mcf to recover, the price is roughly
estimated at$150/Mcf.
(6) The costs involved in gathering and recovering helium from gas containing less than approximately 0.1
percent helium are considered prohibitive. It would be cheaper to recover helium from the atmosphere for
between $500 and $l,700/Mcf, depending upon whether the helium was the byproduct of separating oxygen
and neon from air, or whether helium alone was recovered.
H. R. Rowland of Westinghouse has noted that the cost of helium is much less sensitive to plant operating
costs than capital costs. There are economies of scale, and the use of complex petrochemical plants with
helium-nitrogen mixtures as a final byproduct, also reduces helium price. Using a 280 million cubic feet per
day plant as a basis, Rowland interprets the varied data as leading to a price of $95 to $130 per Mcf per 0.1%
helium content gas, and judges $110/Mcf as a fairly reasonable figure. A. Francis of Linde states that for
helium concentrations between 0.01% and 0.1%, the unit cost of helium extraction will be approximately
inverse to the helium concentration. That is, about $180 per Mcf for 0.08% He; $250 for 0.06%; $500 for 0.03%;
$1,000 for 0.015%; and over $2,000 for 0.007%. Rowland has extrapolated this to 0.5%, which may be
questionable, and quotes slightly different figures. These are: $21 per Mcf for 0.528% He; $46 for 0.240%; $110
for 0.1%; $733 for 0.015%; and $1,690 for 0.0065%. He notes that the free market price of helium, bulk Grade
A, (f.o.b.), has been about $20/Mcf for the past two years.
-596-
-------
NSF HELIUM STUDY
To assist in ordering the confusion and resolving the controversy, a brief, intense study was proposed with
the objectives of:
(1) examining the outlook for helium supply and demand;
(2) determining ways of preventing compromise of critical U.S. technology by limited helium availability
and
(3) specifying goals and estimated helium requirements for likely helium-based developments — especially
superconducting, electrotechnology.
This study is in progress, and definite conclusions have not yet been reached.
HELIUM STUDY DEMAND ESTIMATES - NEW TECHNOLOGIES
At the July Advisory Committee meeting of the Helium Study, the estimates for various technologies were:
Fusion Power (C. E. Taylor, Livermore). He assumes 1,400 gigawatts of fusion power by 2020, giving
an inventory of 35 to 100 billion cubic feet (complete recovery) based on 1,300 scf/megawatt electric. It might
be possible to cut this by a factor of 10 with further design work. We chose 50 billion cubic feet as a modest
estimate to 2050, which is more pessimistic than the Taylor figure.
Fission Reactor Cooling. There is little loss, assuming reasonable capture and recycle precautions, but
by 2000 the inventory will be about 1 billion cubic feet. We did not extend this figure for the years to 2050, but
clearly some will be needed.
Ore Separation (R. Powell, NBS). There are few ore separation plants in this country, and a slight rise in
the price of helium could make the use of superconducting magnets for this purpose uneconomical. P. Marsden
claims that superconducting magnets are not necessary for,this purpose.
Water Purification (R. Powell, NBS). This use may not materialize as a large scale application of
superconducting magnets.
Magnetically Levitated Trains (J. Harding, DOT). Assuming 300 miles of track in the Northeast
corridor, the inventory would be approximately 52.5 x 106 scf over a period of 10 years. This assumes that high-
speed ground transportation with magnetic levitation is of little interest to the United States.
Defense Department. Accepting DOD estimates for 2000 A.D. through 2019 A.D. as 1.6 Mcf, this implies 4
Mcf to 2050, assuming no increase in demand rate from 2019.
SUMMARY OF U.S. DOMESTIC HELIUM REQUIREMENTS (2000-2050)
Annual helium demand for conventional uses = 1.4 x 109 scf'/year (BuMines' low forecast). Assume it holds
flat for 2000-2050.
Category Cumulative Demand x 109 scf
Conventional uses 70
Fusion 50
Transmission 11
Defense 4
Other 1
136
This total ignores: rotating machines, marine uses, airborne, fission reactor cooling, ore separation, water
purification, levitating trains, and exports in the interests of a conservative estimate for newly emerging
technologies.
Assuming the U.S. with its advanced industrial society of 200 million to be a typical bloc, we can discern
Europe, U.S.S.R. and Japan, Australia and New Zealand to be blocs which will develop similar requirements
for helium. Thus a conservative world helium requirement from 2000 to 2050 might be about 500 billion cubic
feet. . .
It should be noted that the helium inventory for superconducting power transmission and fusion reactor
magnets is helium stored and conserved, since every effort will be made to make the leakage rates as low as
possible.
Prevailing Economic Philosophy Versus the Conservationist Approach
The economist wonders who will pay now, the conservationist and the scientist knows it must be someone
else.
-597-
-------
Professor Preston, in his 1969 testimony on the Federal Helium Program before the Joint Economic
Committee of the Subcommittee on Economy in Government, emphasizes that the choice of an appropriate
interest rate is absolutely critical to the decision problem at hand. In this connection, he references the
contribution of Professor Baumol1 to the compendium of papers already published by the committee. In fact,
he is right. The 10% rate being applied to new programs by OMB makes the financial picture for a self-
supporting helium program hopeless. By contrast, the Government never paid 61A% interest until 1969, while
the current rate for Treasury Bills and Notes is about 7%, reflecting a real rate of 3% plus an inflationary factor
Baumol asserts that "the correct discount rate for the evaluation of a government project is the percentage
rate of return that the resources utilized would otherwise provide in the private sector." He continues, "It is
difficult to quarrel with the conservationists view that the destruction of irreplacable natural resources
imposes a heavy cost on our posterity.... this is precisely the legitimate ground on which conservationists
urge increased care in avoiding depletion of our resources. But, it is not legitimate to jump from the valid point
that one generation has no right to use up wastefully the resources that cannot be replaced by its successors, to
the questionable conclusion that each generation is constrained to engage in overall efforts to support its
posterity beyond the level that is indicated by the free market. For that is precisely what is involved in a
program of low discount rates We are therefore wealthier than our predecessors and it can quite safely be
predicted that our successors will be richer than we. In effect, then, the subsidization of a program of added
investment amounts to a transfer of additional resources from the poor to the rich. It would take inputs whose
produce would be available for consumption today and make them available tomorrow when the supply of
consumer goods is likely to be so much more abundant than at present."
This then is the attitude which presumably led the White House and the OMB to cancel the Helium Storage
Contracts. It is based on the views of short-term economics and prevailing economic opinion. Recent events
with regard to the energy crisis may lead to a reappraisal of this type of economic thought, since we can now
see more clearly that future resources may not be more abundant than at present.
An alternate view is that expressed by Senator Jackson prior to the passage of NEPA. "The needs and
aspirations of future generations make it our duty to build a sound and operable foundation of national
objectives for the management of our resources for our children and their children. The future of succeeding
generations in this country is in our hands. It will be shaped by the choices we make. We will not and they
cannot escape the consequences of our choices." (115 Cong. Rec. S17451, December, 1969.)
Again NEPA requires that federal agencies "Use all practical means, consistent with other essential
considerations of national policy, to improve and coordinate federal plans, functions, programs, and
resources to the end that the nation may...
(1) "fulfill the responsibilities of each generation as trustee of the environment for succeeding generations;
(2) achieve a reasonable balance between population and resource use which will permit high standards of
living and a wide sharing of life's amenities; and
(3) enhance the quality of renewable resources and approach the maximum attainable recycling of
depletable resources."
Thus, we see that the helium conservation problem resolves into a conflict between conventionally
acceptable economic thinking and the desires of conservationists.
Finally, we must note the implied assumption in the economic thinking. "We are richer than our fathers,
and will be poorer than our sons. The supply of consumer goods will be much more abundant tomorrow than
today." This view while not widely shared by those who take a broader view of the problem, and even by many
who do not, assumes that the best way to assure national survival, and a good future, is to do our best in
present activities and planning to ensure that such a desirable state of affairs will come to pass. In the case of
helium, we are dealing with an irreplaceable resource, or at least a resource which will cost one hundred times
or more to retrieve from the atmosphere. It would require the present electrical generating capacity of the
United States to produce 13 billion cubic feet per year of helium from air. It is a resource which is presently
produced as an incidental to the country's need for natural gas, and which is required as a necessary
component of several new technologies essential for the country's air, sea, and land defense, its energy needs,
and, as an option, for its high-speed and transportation needs. About 96% of the helium resources in fuel gases
containing 0.3% helium or more will be dissipated by 1985.
SHORT-RANGE POSSIBILITIES FOR THE HELIUM STORAGE PROGRAM
The economic subcommittee of the helium study asked Westinghouse to examine the effect of short-term tax
measures. Their conclusions were that:
(1) the Conservation Program can be made financially self-amortizing as far as the U.S. government is
concerned, from present to to the end of the present phase (1983);
(2) continuation of the program on the above basis will ensure an adequate supply of helium for presently
developed uses well into the 21st century; and
(3) continuation of the program will enable the U.S. to buy time and to keep open certain important
technological options relative to the energy crisis. These options will be foreclosed by cancellation.
'On the Discount Rate for Public Projects by William J. Baumol, Professor of Economics at Princeton
University.
-598-
-------
TENTATIVE FINDINGS FROM WESTINGHOUSE STUDY
(1) In the short run (0-10 years), if the conservation program is cancelled, there will be a massive wash-out of
gas producers as an annual capacity of five times current usage hits the open market. Some distributors with
long-term purchase contracts will be affected.
(2) If the contracts are cancelled, helium will be cheap — perhaps 2/3 current prices — until 1995-2000, when
the rich streams are exhausted.
(3) At $100-120/Mcf, helium will be available in reasonable quantities until 2020 or so, even in the absence of
stockpile sales. (All prices in 1973 constant dollars.)
(4) Shrinking supplies of depleting gas streams will begin to pinch helium availability in 2030+, forcing
reliance on shut-in reserves of non-fuel helium bearing gas, or recovery of helium as a byproduct of oxygen
separation from air at a price of $200-500/Mcf. Crudely projected growth of the oxygen separation industry
shows that from 0.5 to 4 times current (1973) usage will be available in 2030+.
(5) The length of time the stockpile in the Cliffside field lasts depends crucially on the pricing policy of the
Helium Activity. A "cost-pricing" policy will discourage construction of extraction plants, and lead to early
depletion of the stockpile. An "alternate-source" pricing policy will stretch the length of time it lasts.
(6) An important unknown factor in projections of helium demand is whether or not superconductors can be
devised which exhibit this superconducting property in liquid hydrogen, or at higher temperatures.
EXTRACTS FROM THE HELIUM ACT
Title - "An Act to Amend the Helium Act of March 3,1925, as Amended for the Defense, Security, and the
General Welfare of the United States."
From Section 4: "The Secretary is authorized to ... conduct or contract with private parties, for
experimentation and research, to discover helium supplies and to improve processes and methods of helium
utilization."
From Section 7: "The Secretary of Defense and the Chairman of the Atomic Energy Commission may
each designate representatives to cooperate with the Secretary in carrying out the purposes of this Act, and
shall have complete right of access to plants, data, and accounts."
From Section 15: Section 15 of the Act is captioned specifically "Individual Enterprise in Developing
Helium." No statement that this is the object of the Act appears anywhere in that section.
The section states "it is the sense of the Congress that it is in the national interest to foster and encourage
individual enterprise in the development and distribution of supplies of helium, and at the same time provide,
within economic limits, through the administration of this Act, a sustained supply of helium which, together
with supplies available otherwise, will be sufficient to provide for essential government activities."
The Interior has maintained that the purpose of the Act is to provide helium for essential government
activities, and relies on a portion of Section 15 for that. The Denver Appellate Court accepted this view. Key
witnesses at the March 1973 Court of Claims hearing, Northern Helex v U.S., insisted that the Act was for the
General Welfare of the United States. These witnesses were the principal originators and administrators of
the Act.
-599-
-------
WORK AREAS OF THE UNITED STATES
H
Potential Gas Committee June 1973
B
-------
Figure 2.1973 NSF Helium Study Advisory Committee.
Dr.JohnAndelin
Science Adviser
Office of Congressman McCormack
1205 Longworth House Office Building
Washington, D. C. 20510
Dr. Brian C. Belanger
Div. of Applied Technology
USAEC
Washingon, B.C. 20545
Mr. Bascom Birmingham, Director
NBS Boulder Laboratories
Boulder, Colorado 80302
Dr.F.H.Buttner
Battelle Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Dr. Paul Craig
National Science Foundation - RANN
1800 G. Street
Washington, D. C. 20550
Mr. David Evans, Director
Mineral Resources Institute
Colorado School of Mines Foundation Inc.
Golden, Colorado 80401
Dr. Edward Hammel
Los Alamos Scientific Laboratory
P.O. Box 1663
Los Alamos, New Mexico 87544
Mr. M. Hathaway
Office of Senator James McClure
Room 437, Old Senate Office Building
Washington, D. C. 20510
Dr.JohnK.Hulm
Westinghouse Research Laboratory
Pittsburgh, Pa. 15235
Mr. Charles Laverick - Chairman
Argonne National Laboratory
9700 S.Cass Avenue
Argonne, Illinois 60439
Mr. Frank C. Nicholson
Northern Natural Gas Company
Suite 503
1133-15th Street, NW
Washington, D. C. 20005
Dr. James Nicol
Arthur D. Little, Inc.
25 Acorn Park
Cambridge, Mass. 02140
Dr. Charlotte Price
Department of Economics
Sarah Lawrence College
Bronxville, New York
Mr. C.E.Taylor
Lawrence Livermore Laboratory
Box808,L-384
Livermore, California 94550
-601 -
-------
IX. Protection Against Radioactive
Noble Gases
-------
RADIATION PROTECTION IN URANIUM MINES - A REVIEW
D. A. Holaday
Environmental Sciences Laboratory
Mount Sinai School of Medicine
City University of New York
New York, New York
Abstract
The consequences of the exposures to external and internal radiation which can occur in the mining of
uranium ores are well-documented. Methods for evaluating and controlling these hazards have been
developed by health protection agencies and the uranium mining companies. These procedures have been
described in numerous reports from a number of the uranium-producing countries. The only feasible
procedure for reducing exposure to external radiation is limitation of working time. Controls of radon and
radon-daughters are directed at reducing influx of radon and dilution ventilation to supply uncontaminated
air to working areas. Many different methods are employed depending on the particular circumstances. There
is no "typical" uranium mine so the health protection group must adjust procedures to fit the conditions in
each situation. Control of long-lived radioactive dust is usually accomplished by the dust-suppression
methods used to control silica dust. In some circumstances, where high-grade ore is being mined, additional
controls may be required to reduce concentrations of long-lived alpha emitting elements to acceptable levels.
INTRODUCTION
Radiation protection in uranium mines is of great theoretical and practical interest, and has been studied
extensively for many years by mine radiation safety personnel and research investigators. The basis for this
interest is that uranium miners comprise the only group in the nuclear energy industry in which radiation
exposures have resulted in significant health effects, and for which exposure control procedures are difficult
to implement and maintain.
Uranium miners are exposed to external and internal radiation from radon and its short-lived daughters, as
well as to long-lived radioactive elements of the uranium family. Since exposures to radon and its daughters
are considered to present the most serious health hazards, a brief review of this area is presented in this paper.
However, other radiation sources should not be ignored because in some exposure situations they may also
require consideration and control.
DISCUSSION
Of the numerous studies of procedures for controlling radon reported in the world literature, it is only
possible to mention a few of them in this paper. A Russian monograph (Bykovsky, 1963) contains a rather
complete discussion on the subject. Additionally, a two-volume manual issued by the U. S. Bureau of Mines
(Rock and Walker, 1970) is the most comprehensive treatise written in English. The ICRP will be issuing a
manual on uranium mines which hopefully will describe the current state of scientific knowledge. The
aforesaid documents, and the references given therein, will repay careful study by radiation safety and mine
ventilation officers.
The two basic principles which must be observed in designing radon control programs is (1) to limit the
influx of radon into working areas as much as possible, and (2) to remove it from working areas as
expeditiously as possible. In many respects the problem is similar to that encountered in gassy coal mines
where the ventilation practices and criteria used are more applicable to uranium mines than to those which
are standard for metal mines.
The proper planning of mine design and operational methods is an important first step in radon control.
The methods to be used in ore extraction, and the ventilation system employed, should be designed together.
Many of our existing mines have severe control problems because they were opened before the importance of
controlling radon was appreciated. They were planned to remove ore as rapidly and economically as possible
with existing metal mine ventilation criteria. These mines are now mature operations; it would be a time-
consuming and costly process to make improvements in their ventilation. The final design of a mine is usually
a compromise between the wishes of the production superintendent and the ventilation engineers.
Ideally, the mining system should be designed to keep all working areas uncontaminated. This would
involve development to the furthest extent of the ore bodies with air courses and haulageways driven in
barren rock. The ore is extracted by retreating toward air inlets with the used air rejected to return air courses.
Unfortunately, this is only an ideal condition. Ore bodies change in grade, occur in unexpected locations, and
radon emission rates vary widely depending on the porosity and the degree of fracturing in the host rocks.
Thus, a continued assessment of conditions, and continued efforts to correct conditions, are required. An ideal
mining design also keeps open working areas at a minimum, limits the amount of broken ore stored
underground at the lowest feasible level, and minimizes spillage of ore during transport. All of these
precautions reduce the emission of radon into working areas.
-602-
-------
Sealants to reduce emission of radon have been investigated; several materials have been found which are
essentially impervious to this gas. Expanded polyurethane foams are now used to assist in sealing off worked
out areas, particularly those which have been back-filled with mill tailings (which still contain most of the
original radium). The sealing of large areas such as air courses has been tried experimentally in metal mines
The main problem is in the preparation of rock surfaces to obtain good coating adherence This is merely an
operational detail, but one that can be very troublesome. Another problem, for example, is that all persons in
the area where urethane foams are being sprayed must wear the correct respirators. Plastic films which are
impervious to radon have been studied by the U. S. Bureau of Mines and Colorado State University. Some of
these materials show promise and may be useful for such applications as temporary stoppings and for
covering piles of broken ore which are awaiting transport.
As stated earlier, ventilation air should be moved through active areas and to return air courses as
expeditiously as possible. The shorter the residence time of the air, the less the ingrowth of radon-daughters.
Every effort must be made to course the air directly to active stopes, and to prevent the recirculation of the
contaminated air. This requires that inactive and abandoned stopes should be closed off by temporary or
permanent stoppings. Such areas can be potent sources of contamination, and also waste ventilation air.
Experience has shown, that wherever possible, such closed areas should be kept under negative pressure. As
ground movement and changes in air pressure loosen stoppings, a routine inspection and repair program is
required to maintain the integrity of the seals.
In addition to programs for reducing radon emission, supplying clean air, and removing contaminated air,
considerable work has been done on methods of decontaminating mine air to extend its useful life. The
approaches which have been tested in mines are the removal of radon-daughters either by filtration or by
electrostatic precipitation. Either method is effective, but there are limitations on their utility. The basic
limitation, of course, is the atmospheric concentration of radon. This must be low enough for the treated air to
be used for at least a few minutes before the ingrowth of daughters makes it unacceptable.
It has been found that with cleaned air in which a large fraction of the daughters exist as unattached atoms,
plate-out on walls is large enough to make the air usable significantly longer than theory predicts. The most
troublesome problems have been due to equipment failures. Filters clog with dust and soot, and are seriously
affected by water mist and vapors. Present electrostatic precipitators do not seem to have been designed to
withstand mine conditions. Obviously, the use of air cleaning procedures will be limited to special situations,
but considering the problems involved in extending ventilation systems, the idea is attractive.
An even more fundamental approach would be the removal of radon from the air. Several papers in this
symposium have discussed adsorption of noble gases, and more on this topic is presented in this session. In
my opinion, radon removal methods merit further investigation.
Environmental conditions in uranium mines can now be evaluated more accurately and in greater depth
than was possible even three years ago. Among others significant work has been done at the Nuclear
Research Center in France (Lopez, et al., 1970), at the A.E.C. Health and Safety Laboratory (Breslin, et al.,
1969) in the U. S. A. to develop instrumentation and procedures for measuring radon emission rates,
atmospheric radon concentrations, radon-daughter ratios, fractions of unattached atoms and number
concentrations, and size distribution characteristics of mine aerosols. While most of these methods are too
elaborate for routine use, they can be employed by control personnel. Information obtained by their use would
certainly assist in evaluating the degree of radiation hazards in mines.
Methods for measuring individual exposures to radon-daughters have been studied, and two papers in this
symposium report work on dosimeters. In my opinion, such devices would also be useful as area monitors, and
might reduce the amount of air sampling which is now necessary. If reliable personal dosimeters can be
developed, the task of estimating and controlling individual exposures can be simplified. At present, these
records are generated from air concentration measurements and time-occupancy data. Not only is this an
uncertain process, which involves handling a large amount of data, but the possibility of overlooking brief,
high-level exposures is always present.
Inevitably there will be situations where personal protective equipment is required to prevent the over-
exposure of the miners. Typically such situations will occur in limited areas where men must enter to install
auxiliary ventilation or to perform other work to correct conditions. For such relatively brief operations,
respirators equipped with filters which will remove radon-daughters are available. The atmospheric
concentration limits for which each type of respirator is suitable will be detailed in the forthcoming ICRP
manual. Battery powered devices, which will blow filtered air to a hood, have recently been approved. In at
least one of these, the final air cleaner is a HEPA filter. This equipment is more cumbersome than the usual
respirators, but there is no resistance to breathing or restriction to vision, and the filtering efficiency is very
high. Such devices would seem to be particularly suitable for use by mine geologists, mine surveyors, or other
men who enter inactive mine areas. However, in such situations, the atmospheric concentrations ol radon
may become of real concern, and protection against the gas may also be required Self-contained breathing
apparatus or air-supplied respirators have been used. The effectiveness of gas masks equipped with charcoal-
filled canisters for protection against radon has been studied by HASL, and the conditions under which they
are satisfactory have been defined. ...
As we still have no firm definition of what a "safe" level of exposure to radon and radon-daughters would be,
the objective is to reduce atmospheric concentrations to the lowest feasible levels.
-603-
-------
REFERENCES
Breslin, A.J., A.C. George and M.S. Weinstein, (1969), Investigation of the Radiological
Characteristics of Uranium Mine Atmospheres, USAEC, HASL-220.
Bykovsky, A.V., (1963), Occupational Hygiene in Underground Uranium Mines, Moscow, 1963 (in
Russian).
Lopez, A., A. Chapius, J. Fontan, E. Billiard and G.J. Macklain, (1970), Measurement of the State of
Equilibrium Between Radon and its Daughters in the Uranium Mines, Aerosol Science, I, 225.
Rock, R.L. and D.K. Walker, (1970), Controlling Employee Exposure to Alpha Radiation in
Underground Uranium Mines, U.S. Dept. of Interior, Bureau of Mines, Two volumes.
- 604-
-------
RADON PROTECTION IN URANIUM MINES
J.Pradel
Nuclear Research Center
Atomic Energy Commission
Fontenay-Aux-Roses, France
Abstract
In uranium mines, radon is present with its daughter products which may be fixed on different particles, the
form of which is described. It is well-known that uranium miners may contract some pulmonary diseases
when radon-daughter concentrations are too high during long periods. The origin of lung cancer is confirmed
by animal experiments.
To have a good protection, practical limits for radon and radon-daughter concentrations are established
and individual monitoring is initiated. This monitoring must be well-fitted because the actual levels are
almost up to maximum permissible levels.
Ventilation is the most important technique to control radon and radon-daughter concentrations. Some
ventilation investigations in an experimental mine are described.
INTRODUCTION
Radon-222 is present naturally in uranium mines. This radioactive gas and its daughters, which are also
radioactive, contaminate the air and very likely contribute to the excess in the number of lung cancer cases
observed among populations of miners who have worked in poorly ventilated atmospheres. The importance of
this problem, therefore, necessitates a thorough knowledge about the behavior of radon, constant surveillance
of its concentrations, and the implementation of effective control measures.
BEHAVIOR OF RADON AND OF ITS DAUGHTER PRODUCTS IN MINE AIR
Radon-222, an element in the uranium-238 chain, is a noble gas with a half-life of 3.8 days, which is produced
by the decay of the radium-226 dispersed in uranium rocks. Formed within the rock, radon gas fills all the
available openings and may be released into uranium mines by various routes. Primarily, the retained gas is
liberated when rock is fractured — especially at the time of blasting. The gas can also enter the working
environment through underground water into which it has mixed.
After having passed into the mine's atmosphere, radon decays into short-lived radioactive products (all
with half-lives of less than 30 minutes) in the form of ultrafine solid particles which attach themselves to other
particles. This attachment brings about the formation of a radioactive aerosol with physical characteristics
that determine its probability of retention in the respiratory system, and thus its toxicity.
1. Formation of a Radioactive Aerosol Present in a Mine.
Radon decays into polonium-218 (Ra A). At the time of its formation, it is a recoil atom which possesses an
initial energy on the order of 100 keV. Although these atoms are produced by alpha-disintegration, around 80
per cent reach the thermal level in the form of positive ions. The remaining 20 percent are present as ultrafine
neutral particles with perhaps a small percentage of negative ions, the presence of which has not yet been
proved with certainty.
Many authors have been able to show that these "primary particles" have, at their origin, high diffusion
coefficients on the order of 10-1 cm2 s-1 for the neutral particles and 5.4 x 10-2 cm2 s-1 for the small ions. But, in
fact, various molecules are quickly attached to these particles. For the charged particles in the open air, it can
be shown that these clusters become distributed into several groups of distinct and preferential mobility (K =
2.1 -1.35 - 0.88 - 0.55 - 0.43 cm2 s-1 V-1). These results have not been confirmed for the electrically neutral
particles because the current measurement techniques are not selective enough. This illustrates, therefore,
that before attaching themselves to the nuclei or dust particles, the radon-daughters can exist as ultrafine
particles during a time span which depends upon the concentration of the other particles present.
It should be noted that together these ultrafine and primary particles constitute what is customarily called
free ions or uncombined particles. The value of the diffusion coefficient is, in general, taken to be 5.4 x 10-2 cm2
s-1; while it should be much lower, the actual value probably being between 5.4 x 10-2 cm2 s-1, as in outside air.
In the absence of data for mine air, it seems more justified to adopt the value 1.5 x 10-2 cm2 s-1.
Because of their high diffusion coefficients, the ultrafine particles (with radii of less than 10-3fzm are going
to attach themselves to the condensation nuclei and dust particles present in the air; the attachment being a
function not only of the number, but also of the dimensions and electrical charge of the nuclei and of the dust
particles. It can therefore be seen that the total concentration of non-radioactive particles and knowledge of
the particle size distribution and the charged state of these particles are necessary to study the evolution of
radon decay products and to determine the risks associated with them. Therefore, if one wants to study the
evolution of these particles, one should be equally interested not only in the radioactive products, but also in
the physical characteristics of all the aerosols present in the air of the mine. Because the concentration and
-605-
-------
the characteristics of these particles in a given place depend upon their mode of formation (blasting, slushing,
haulage), the place they are formed (stope, drift), and the path followed by the air in arriving at a given spot,
one can see the hazards in determining the nature of exposure relevant.
2. Data Concerning Mine Aerosols.
The particle concentration inside a mine is essentially variable, but it is high in all cases. The following
results can be given for French mines (obtained with condensation nuclei counters [CEA-SEERI] which
enable the detection of particles greater than 10-3fzm).
(a) outside the mine: 3,000 to 5,000 particles per cm3.
(b) in the nonworking areas of the mine: 3,000 to 5,000 particles per cm3.
(c) in the drifts: 40,000 to 70,000 particles per cm3.
(d) in the working areas: 105 particles per cm3.
(e) after blasting 2 x 106 particles per cm3.
These measurements were carried out on the radioactive particles using Zeleny tubes. Diffusion batteries
and filters make it possible to have a general idea of the characteristics of a radioactive aerosol in a mine. In
the working areas where an important amount of radon, condensation nuclei, and dust are emitted, an
extremely rapid evolution of the aerosol, by coagulation for the finest particles and sedimentation for the
largest particles, is produced. In this high-particle-density atmosphere, the ultrafine radioactive aerosol
activity represents, on the average, 3 per cent of the equilibrium activity of RaA, while still being almost
always 10 per cent — except in the case of a work stoppage in the mine when values on the order of 25 per cent
can be found. As one gets further away from the work area, the particle concentration decreases under the
affects of sedimentation, turbulent diffusion, and impaction; the aerosol evolves continuously toward an
equilibrium. In that way the count median diameter (on the order of 0.2 fzm) tends toward a value of 0.07 jjm,
with the whole range of particle size being between 0.01 and 0.4 jtm.
3. State of Equilibrium Between Radon and its Decay Products.
Radon decays with a 3.8-day half-life; its three principal short-lived decay products are RaA, RaB, and RaC,
the mean lives of which are much shorter than radon and are found in the air in proportions varying with the
age of radon; i.e., with the elapsed time since its diffusion into the atmosphere. The equilibrium fractions, f^,
fg, fc, are defined as the ratios of the activity of the respective decay products A, B, or C to that of radon.
Five hundred measurements in the French mines gave the following results:
fA fB ^
(a) upper value i 0.73 0.43
(b) lower value Q.08 0.04 0.02
(c) mean value Q.50 0.30 0.15
It should be noted that most of the measurements were made in the drifts so that the mean value for the
whole mine is lower than that given here.
STANDARDS AND MONITORING
The hazards of radon and its decay products have been proved mainly by:
(a) early studies on the mortality of miners in the Schneeberg and Joachimstal areas, which show that fifty
percent of the miners had lung cancers;
(b) a recent U.S. study pointing to an excess of deaths from lung cancer in a population of about 3,000
miners; and
(c) the experimental induction of lung cancers in animals exposed to radon and its decay products (Perraud,
etal., 1972).
Various recommendations were issued since 1940, but there is currently some confusion on the choice of
maximum permissible concentrations for radon and its daughters.
The ICRP in 1959 proposed 3 x 10-6/(l + 1,000 f) yCi cm-3 as a limit for the radon concentration for a 40-hour
weekly exposure; f being the ratio of Ra A; f also being assumed to be equal to ten per cent; the MFC obtained is
3 x 10-8^Ci cm-3. The IAEA limit has remained at 3 x 10-7jiCi cm-3. On the other hand, the U.S. Public Health
Service introduced the concept of "working level" (WL), which is linked to the total alpha energy liberated by
the products of radon when they decay. One WL corresponds to any combination of short-lived products in one
liter of air that delivers 1.3 x 105 MeV of alpha energy, for instance 10-10^Ci I-1 of Ra A in equilibrium with Ra
B and Ra C. The WL and the corresponding integrated unit called Working Level Month (WLM) are more and
-606-
-------
more widely used. From the viewpoint of momtonng, this unit has in practice several drawbacks, and it
should only be adopted if it is derived from adequate biological knowledge. For example, if the hazard is
associated with the presence of uncombmed particles, the total alpha energy is not representative at all
whereas the radon concentration is. It is thus urgent for the ICRP to provide precise recommendations
because, as a consequence, the momtonng could be substantially modified.
Radon monitoring is carried out very easily; the air sample is taken by filling a vial, the inside of which is
covered with zinc sulfide; the measurement is made outside the mine without any restriction to the delay
time. A great number of samples can be taken, and that is the only way in which a precise and constant
surveillance of all the working areas can be done.
On the other hand, measurement of the total alpha energy necessitates the use of pumps, filters, and
portable detectors in the difficult conditions of the mine. The measurements are subject to errors and must be
performed in situ; the person in charge of the monitoring, subjected to various kinds of pressure, tends to
interpret and modify the results obtained, which is not the case when radon is measured because it is
performed outside of the mine by personnel not directly concerned with the consequences of the result.
The calculation of the activities inhaled by each miner is made from local measurements, taking into
account the time spentin the various locations.
The procedure is to be replaced with a technique of personal monitoring which we have just started to use.
This personal monitoring is performed with portable instruments (Figure 1) which use a filtering device
enabling the separate measurement of the alpha particles of Ra A and Ra C with a LR 115 (Kodak-Pathe)
cellulose nitrate detector (Chapuis, et al, 1972). However, it is likely that, at least at the beginning, this
personal monitoring will be used only for a few miners with the aim of "adjusting" the results obtained from
the local radon measurements and the periods of time spent by each miner in each location.
MONITORING RESULTS
In 1972, 36,000 measurements were carried out in France; the average dose to the population of 621 miners
corresponds to a radon concentration of 1.45 x 10-7fiCi cm-3 during 2,112 hours.
From the results of about 1,000 simultaneous measurements of radon and total alpha energy in 1972, it can
be estimated that the average dose corresponds to about 3 WLM, but this is not true for the most exposed
individual for which the dose is likely to be higher by a factor of two to three.
The following remark has to be made on that point: in the U.S. epidemiological survey, the dose received by
the workers was, in general, assessed in a very inaccurate way, and, in fact, the dose estimated was the
average dose, and not that to the most exposed worker for whom the exposure results from exceptional
working conditions. If one wishes to derive a standard from this survey, the value obtained would be an
average value, and it would be admissible to tolerate a value higher by a factor of 2 to 3 for the most exposed
workers.
MEANS OF PROTECTION
Techniques widely used in uranium mines are ventilation, isolation of sources, and elimination of the
polluted waters. As a result, the doses received by the miners are considerably lower than those corresponding
to the former conditions that brought about an excess of lung cancers. In order to further improve the working
conditions, we found it necessary to build an experimental mine which allows us to test various techniques
under any required conditions.
The first results obtained deal with the influence of ventilation (Pradel, et al., 1973). It has been observed
(Figures 2 and 3) that:
(a) at a given flow rate, the radon concentration in mine air is lower when the pressure in the mine is positive
than when it is negative; the reduction factor increases from 0.5 to 0.9 when the flow rate increases from 2 to 10
3 g
_l*
_
(b) at a given flow rate, the concentration expressed in working levels is also lower when the pressure in the
mine is positive; the reduction factor is constant and equal to 0.7 when the flow rate is higher than 3 m3 s-1;
(c) the radon flux is reduced by about twenty-five per cent when the pressure in the mine is positive.
Another test showed that, at a constant flow rate (2 m3 s-1), the radon flux decreases from 9.2 x 10-7 Ci s-1 to
7.4, 7.3, and 5.7 x 10-7 Ci s-1 when the differential pressure increases from -45 millibars to -2, + 2, and + 56
millibars, respectively. . . ,. . , .
These values may not apply to all mines but they confirm that overpressunzing the uranium mines brings
about, in many cases, a significant decrease in the radon levels.
CONCLUSIONS
Radon protection in the uranium mines is important because an excess in lung cancer rate can be observed
among miners working under poor conditions. It is difficult, but urgent, to set up a precise standard. The
following physical data can be of use in order to define this standard:
-607-
-------
(a) the proportion of uncombined ultrafine particles should be taken to be equal to 3 per cent of equilibrium
Ra A activity instead of 10 per cent; and
(b) the diffusion coefficient is likely to be close to 1.5 x 10-2 cm2 s-1, and is not equal to 5.4 x 10-2 cm2 s-1.
The standard should not use as a unit the total potential alpha energy, unless biological data justify it; these
measurements are more difficult to carry out than those of radon concentration, the latter being a better
representative of the concentration of the uncombined ultrafine particles.
If one hopes to lower the pollution levels, new control measures must be looked for. Overpressurizing the
mines can already be considered effective.
As to monitoring, because the first tests were satisfactory, the use of personal detectors can be foreseen, at
least for part of the miners.
REFERENCES
Billard, F., J. Bricard, G. Madelaine and J. Pradel, (1965), Etude Experimentale de la Charge
Electrique des Aerosols UltrafinsparMarquage Radioactif, Journal de Physique 26. n° t.
Billard, F., G. Madelaine, A. Chapuis, J. Fontan and A. Lopez, (1971), Contribution a I'etude de la
Pollution de I'atmosphere des Mines d'uranium, Radioprotection 6; 45-61.
Bricard, J., P. Girod and J. Pradel, (1965), Etat de Charge des Aerosols Ultrafins en Milieu Faiblement
Ionise, Journal de Physique 26.
Bricard, J., F. Billard, D. Blanc, M. Cabane and J. Fontan, (1966), Structure Detaillee du Spectre
Granulometrique des Petits Ions Radioactif s, Compte-Rendu a 1'Academie des Sciences de Paris 263.
Chamberlain, A. C. and E. Dyson, (1956), The Dose to the Trachea and Bronchi from the Decay
Products of Radon and Thoron, British Journal of Radiology 3171.
Chapuis, A., A. Lopez, J. Fontan, F. Billard and G. Madelaine, (1970), Spectre Granulometrique des
Aerosols Radioactif s dans uneMine d'uranium, Aerosol Science 1:243.
Chapuis, A., (1971), These Docteur es Sciences, Toulouse.
Chapuis, A., D. Dajlevic, P. Duport and G. Soudain, (1972), Dosimetrie du Radon, Bucarest.
Chapuis, A., A. Lopez, J. Fontan and G. Madelaine, (1973), Determination de la Fraction d'activite
Existant sous forme de Ra A non A ttache dans I'atmosphere d'une Mine d'uranium. Health Physics 25:59-65.
Lopez, A., A. Chapuis, J. Fontan, F. Billard and G. Madelaine, Measure de I'etat d'equilibre entre le
Radon et ses Descendants dans les Mines d'uranium.
Perraud, R., J. Chameau, J. Lafuma, R. Masse and J. Chretien, (1972), Cancer Bronchopulmonaire
Experimental du Rat par Inhalation de Radon. Comparaison avec les aspects histologiques des cancers
humains. Journal Francais de Medecine et de Chirurgie thoracique 26:25-41.
Porstendorfer, J.,(1968), Geladene und Neutrale Radon Folgeprodukte inLuft. Zeit. furPhysik 384-396.
Pradel, J., P. Zettwoog, G. Madelaine and Y. Francois, (1973), Quelques Donnees Nouvelles
Concernant la Protection des Mineurs dans les Mines d'uranium. Third International Congress of the
International Radiation Protection Association (IRPA), Washington, D.C.
Raabe, O. G., (1968), The Absorption of Radon-Daughters to some Polydisperse Submicron Polystyrene
Aerosols, Health Physics 14:397-416.
Renoux, A., (1965), These Docteur es Sciences, Paris, also Commissariat a 1'Energie Atomique (CEA)
report n° 2771.
-608-
-------
Figure 1. Individual radon monitoring apparatus.
cover
screen 2
collimator
NVs
7
/ / / /
^XVSN^
film
\
screen 1
filter fritted disk
motor
- Schema de principe de la t£te de
(a) Design of the sampling head.
(b) Ra A alpha tracks in Kodak-Pathe LR115 cellulose nitrate detector.
-609-
-------
•uoi^BJ^uaouoo uopBj aif> uo uotjBifjuaA jo apoui aifl jo aouanuuj -
Radon Concentration
1
4
3
2
-xlO-10pCi/liter
M 1
'
~.
•rt
. 1
1
-1
"I
C
25^
ce E
S - S
If -^ S 11 °° "
24h
| ^
1
72 h
1 *
24 h
1
i
1
1
2
v> _£
n^ S
E ^r
00 ^
* £
^""**- cd
"E 1
o o
| •- CN
f
+
\
72h
__«/» £
(*> rO
E s
o 2^
1
'
o
rH
co
-------
WL
Radon cond.
c,
0,8
0,6
0,4
0,2
10'10Ci/L
-4
h3
-2
Radon flux
QRn>
410
10~7Ci/sJ
•x—
QRnt
,
negatlve
pressure
QRn
r\ Jpositive pressure
LJ W L negative pressure
/ \ W L positive pressure
ventilation - ms/s
8
0
8
10
12
cRn(R)
WL(P)
WL{N)
P = positive pressure
N = negative pressure
ventilation - ms/s
J 1 1
8
10
Figure 3. Variations of radon concentration (CRn), radon flux (QRn), and potential alpha energy (WL)
with pressure.
-611-
-------
SOME RADIOLOGICAL HEALTH ASPECTS OF RADON-222 IN LIQUIFIED PETROLEUM
GAS*
T. F. Gesell
School of Public Health
University of Texas
Houston, Texas
ACKNOWLEDGEMENT
The author wishes to acknowledge professor John A. S. Adams for consultation; Mr. Carl Mullins, Mr.
Dickie Burgess, Mr. Steve Bryan and Mr. Bobby Lane for their assistance in collecting and analyzing the
samples; and Miss Susan Bettes for technical aid in the preparation of the manuscript.
Abstract
Natural Gas (NG) produced at the wellhead contains varying quantities of radon-222. Previous studies of
wellhead gas from several gas fields have indicated concentrations ranging to 1450»J^Ci/l at standard
temperature and pressure. When natural gas is processed to remove impurities and heavier, more valuable
hydrocarbons, the radon becomes concentrated in the ethane and propane fractions. This concentration
apparently takes place during fractionation because of the similar boiling points of radon, propane, and
ethane. Liquified Petroleum Gas (LPG), which is widely used as a domestic fuel and as a chemical feedstock is
composed principally of propane and ethane. Potential human exposure to radon-222 and its progeny
contained in NG and LPG can occur in a variety of ways to both occupational groups and to members of the
general public. The paper includes an*tdentification of the pathways of human exposure, the results of
approximately 400 measurements of radon-222 concentrations in retail LPG and domestic NG supplies, and
measurements of radon-222 concentrations in gas processing plant streams. An exposure model for the
domestic fuel usage pathway is also presented.
INTRODUCTION
Radioactivity in natural gases was first reported for Canadian gas in 1904 (Slatterly and McLennan, 1918).
At that time, it was suggested that helium, a known product of radioactive decay, might be associated with the
radioactivity. Somewhat later, a systematic survey was made of Canadian natural gases (Slatterly and
McLennan, 1918) in an effort to find a relation between helium concentrations and radioactivity. The
radioactivity was reported to be due to "radium emanation", now known to be radon-222. Other reports were
made on the radioactivity of American and European natural gases in the years between 1904 and 1918.
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 1950's resulting in a series of
papers dealing with radon and helium concentrations in U. S. natural gases (Faul, et al., 1952 and Pierce, et
al., 1955). A comprehensive survey of radon-222 and helium concentrations in the Texas Panhandle Field
gases was issued by the U. S. Geological Survey (Pierce, et al., 1964).
Subsequently, Bunce and Sattler (1966) and McBride and Hill (1969) dealt with radon concentrations in
natural gas in the San Juan Basin, located in southwestern Colorado and northwestern New Mexico. These
papers are significant because, unlike the previous papers, 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 occuring levels.
Subsequent to these wellhead studies, two papers have mentioned the possibility of radon release from gas
fired electric power generating stations (Fish, 1969 and Terrill, et al., 1967).
The studies reported in the literature give an indication' of the magnitude and range of radon-222
concentrations to be found in natural gas. Table 1 summarizes some of the reported results. Discounting the
values reported at "reservoir temperature and pressure", it is seen that wellhead natural gas exhibits radon-
222 concentrations ranging from nearly zero to 1450 pCi/liter at STP.
The foregoing review of the literature demonstrates that we have a general, but by no means
comprehensive, knowledge of radon-222 concentrations in wellhead natural gas. For the purposes of this
study, however, knowledge of radon-222 levels at wellhead is insufficient. Wellhead data give no indication of
the distribution of radon-222 in the products of gas processing plants. Owing to the fact that the boiling point
of radon lies between that of propane and ethane (Table 2), which are the principal constituents of LPG sold
commercially (Table 3), one may expect the concentration of radon to be greater in the LPG products than in
other natural gas products. Any activity in these products would, of course, be diminished by the delays from
actual production, processing and delivery according to the half-life of radon-222 (3.82 days). Wellhead data of
the type existing in the literature are also of little use in evaluating occupational radiation exposure due to
radon.
*Work supported by A tlantic Richfield Co., Dow Chemical Co., Esso Production Research, Phillips Petroleum
Co..Shell Oi! Co.. Texaco Inc., andthe Texas State Department of Health.
-612-
-------
POTENTIAL PATHWAYS FOR HUMAN EXPOSURE
Radon-222 and its daughter products emit alpha, beta, and gamma radiation. Occupational exposure to
external radiation can occur in the vicinity of gas processing or transport equipment where high radon-222
and concentrations exist. The internal exposure of the lung to radon-222 and daughter radiations can occur
whenever radon bearing gases or combustion products are released and inhaled by workers Another internal
exposure possiblity exists whenever gas processing equipment is disassembled for repair. Lengthy exposure
of the internal surfaces of such equipment could result in the buildup of the long-lived radon-daughters lead-
210 and polomum-210. Ordinary cleaning procedures could then create an aerosol containing these isotopes
and result in a respiratory exposure.
Population exposure to external radiation could occur in the vicinity of domestic LPG storage tanks.
Internal exposure to the lung could occur whenever combustion products of radon-containing fuel gases are
present in the atmosphere. While this situation occurs at limited concentrations in the general environment,
the highest levels of individual exposure would occur whenever radon-containing LPG or NG is burned
without a flue inside dwellings or other occupied buildings. This situation occurs with gas cooking, unvented
gas space heating or unvented gas hot water heating. An indication of the magnitude of the exposed
population may be gained from the U. S. Census data of 1970 (U. S. Bureau of the Census, 1970) which lists
5,314,082 dwellings containing LPG fueled ranges, 3,14,164 dwellings containing LPG fueled hot water
heaters, and 3,949,420 dwellings containing unvented heaters. Most unvented heaters burn either NG or LPG.
An estimate of population exposure due to radon in LPG should be based on field data collected as close to
the point of use as is practicable. This paper reports the results of radon-222 concentrations measured in retail
LPG (propane) samples gathered in fourteen southern and western states. Regular weekly samples were
gatheredfrom seven Houston, Texas area-retailers for a period of one year in order that seasonal dependencies
could be studied. Individual grab samples were obtained from the other geographic areas over an eight-month
period. The results are compared with standards and with natural background levels for radon in air with the
aid of an exposure model based principally upon fuel usage and ventilation conditions.
SAMPLING AND MEASUREMENT
Sampling cylinders were obtained specifically for this project. Before use, each of the 1 lh gallon cylinders
was tested to assure that no radon-222 would be introduced to the samples from radium in the construction
materials of the cylinders. This test was accomplished by introducing known radon-free LPG into each
cylinder and after 24 hours, removing a sample for analysis. The results of these analyses showed no
measurable radioactivity, and it was, therefore, concluded that any activity found in subsequent samples
contained in these cylinders would be exclusively from the LPG being studied.
Retail LPG was sampled by simply purchasing LPG at the "bottle stations" or "dispensers" operated by
retail dealers. The dealers filled the sample bottles in exactly the same fashion that they filled LPG bottles
presented by their customers.
Samples of LPG obtained by purchase from retail dealers at their bottle stations are obviously
representative of LPG purchased in the same fashion by members of the general public. These customers
include owners of permanent house trailers and recreational vehicles. Most LPG consumed by the general
public, however, is delivered to residential and commercial sites by truck. The extent to which our samples at
the bottle stations are representative of the LPG delivered by truck is dependent upon the practices of the
individual retailers. A typical retailer designated by code number 2, operates his business in the manner
shown in Figure 1. The #2 bottle station where we obtained our samples is filled at the end of each day with the
same gas from the same trucks that deliver to residential and commercial customers. Thus, due to radioactive
decay, the radon concentration of the bottle station LPG would be lower to an extent dependent upon turnover
rate, than that delivered by the trucks.
Another common practice employed by some LPG retailers is to fill gas bottles from the same main retail
storage tanks that serve the trucks. Samples obtained from these dealers would be highly representative of
the LPG delivered to the consumer. A third, less common practice is to fill LPG bottles directly from the trucks
used to deliver gas to residential customers. This practice usually exists among the smaller dealers. Thus,
dependent upon the practices of the retailers, our LPG samples are similar or somewhat older m age than the
LPG delivered to the consumer. Consequently, the measured radon concentrations represent, on the average,
the lower limit of radon concentrations to be found in delivered LPG.
The radon concentration of the samples was measured using the Lucas Method (Lucas, 1957). Special cells
whose curved interior surfaces are coated with a phosphor are filled with gaseous LPG (expanded from a
liquid aliquot) to atmospheric pressure. After three hours (to allow the daughter products to reach transient
equilibrium with the radon) the cells are placed in contact with a photomultiplier tube and the pulses counted
with a sealer. The analytical equipment was calibrated using a National Bureau of Standards radium
solution. . . . , , j ,.
Error is due to error in calibration coupled with error produced by counting statistics A standard counting
time of 90 minutes is used so no error due to counting statistics is greater for the lower concentrations
Counting statistics errors range from a - 1% for 300 pCi/1 to a = 17% for 1 pCi/1. Our lumt of
detection is taken as 0.5 pCi/liter which is approximately 2 a above background. Calibration error is
-613-
-------
related to the 1% error given by the National Bureau of Standards for the radium solution plus any error in
transfer of the radon to the Lucas cell. The new data produced by the counting instruments are expressed in
counts per minute. This number is converted to picocuries per liter at STP at the time of sampling using the
following formula.
XPoT
Y= VfpT EXP(.693t/t]/2)
where
Y = Radon concentration in picocuries per liter
X = Counting rate of sample in counts per minute
V = Volume of Lucas cell = 0.096 liters
f = Calibration factor in counts per minute per picocurie
P = BarometricpressureattimeoftransferinmmofHg
P0 = Standard pressure (760 mm Hg)
T0 = Standard temperature on the absolute scale (273° K)
T = Ambient temperature at time of transfer on the absolute scale in °K
t = Elapsed time between collection and analysis of sample in hours
ti/2 = Half-life of radon-222, (91.8 hours)
RESULTS OF THE SURVEY
The survey of radon-222 concentrations in LPG was done in three phases. Initially, thrice weekly and later
weekly samples from the seven identified major Houston area LPG retailers were collected and analyzed. This
program continued for one calendar year. After the weekly sampling and analysis had become routine, the
second phase collection and analysis of grab samples from other areas in Texas was begun. In the third
phase the analytical equipment was installed in a motor coach and samples from other places in the United
States were collected and analyzed.
It was determined that the population group most likely to receive the highest exposures to radon-222 and its
daughters from this pathway would be those persons living in dwellings utilizing unvented space heaters.
Furthermore, the exposure would be directly related to the usage of the heaters. Accordingly, an index was
devised which consisted of the product of the average number of degree days of heating required in a
geographic area multiplied by the number of unvented heaters in use in the same area. The distribution of this
index is shown in Figure 2.
Another major factor that influenced our sampling was the distribution of gas processing plants which
produce LPG. The hypothesis was that due to radioactive decay, retail LPG sold near the production would
generally exhibit higher radon levels than LPG sold far away from the production. LPG production areas are
shown in Figure 3 where each dot represents one gas processing plant. Thus, we wished to include areas near
and far from production. These constraints plus operational factors resulted in the sample locations shown in
Figure 4. Each dot represents one sample except for the Houston, Texas area where each dot represents one of
the seven regularly sampled retailers.
The weekly results of the Houston survey have been averaged into monthly intervals and are presented in
Table 4. Also included are column and row averages. Retailer #7 was consistently low so sampling was
discontinued after six months. For purposes of the averages retailer #7 was taken as 0.5 pCi/liter for February
through June. The data are characterized by a large amount of scatter within and among dealers. This scatter
is not too surprising in view of the number of factors that control the radon in LPG. These factors include the
radon concentrations in the wellhead gas, the elapsed time between processing and entry into the wholesale
market, the elapsed time between entry into the wholesale market and delivery to the retail dealer and delivery
to the consumer.
Of the seven retailers sampled, No. 6 is on the average much higher in radon concentration than the others
and No. 7 is on the average much lower than the others. Inquiry revealed that retail dealership #6 obtains its
propane directly from a gas processing plant and that retail dealership #7 receives its propane from an
underground storage well. The other five retailers obtain their gas from wholesale distributors. Thus, the
radon concentrations in the propane of the several retailers are entirely consistant with their individual
supply practices.
Figure 5 is a second order regression analysis of the monthly averages of the Houston area dealers. The
pronounced dip in the activity during the winter months is in all probability due to the higher winter
consumption raite and consequent appearance on the market of previously stored LPG. This interpretation is
consistent with the known practices of the dealers and wholesalers. It is worthwhile noting that the average
concentration during the heating season is approximately one-half of the concentration during the summer
months. Thus, 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.
-614-
-------
The results of all the non-Houston retail propane samples are grouped by region and presented in Figure 6
In each region, the top number is the maximum concentration observed, the middle number is the mean of all
samples in the region and the bottom number is the lowest concentration observed. The most remarkable
feature of this distribution is the 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, have to be interpreted in terms of low concentrations
in the sources.
The variation of radon in LPG with distance from a major gas processing area was examined by dividing
non-Houston samples into three categories according to whether they were taken within 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 5. The differences were tested by the non-parametric Wilcoxon two sample test (Baily 1971)
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. This result suggests (but by no means demonstrates) 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 processing areas than the sampled states.
The variation of radon in LPG with the type of container (retail storage tank, truck, or small tank) from
which the sample was drawn was also examined. Only Texas was included in this examination, because in
the other regions very few samples were obtained from trucks. Retailers fill bottles brought to their locations
either 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 presented in Table 6. The mean of the truck samples, although smaller than that of the large tank
samples, was not significantly smaller. The mean of the small tank samples is significantly smaller (P
-------
The model, however, leaves much to be desired. For example, the infiltration rate is assumed to be one air
change per hour for all three types of dwellings, whereas common sense would dictate that a dwelling with
excellent insulation, storm windows, and storm doors would have fewer air changes per hour due to
infiltration than a poorly constructed dwelling. The effect of this sort of correction would cause the lines in
Figure 7 to converge, thus reducing the disparity among the dwelling types. A second difficulty with the model
is that there is no provision for local variations within the dwelling. There would certainly be concentration
gradients, but they would be difficult to predict or measure.
The radon levels to be expected in room air can be related to the radon levels to be found in LPG (assumed to
be propane) and the CC>2 levels found in the room using simple stoichiometry. The results presented in Figure
8 are the maximum permissible concentrations of radon-222 in air for members of the general population as
recommended by three different organizations (International Commission on Radiological Protection, 1959;
National Committee on Radiation Protection, 1959; and the U. S. Atomic Energy Commission, 1965). At the
present time, only the highest value, given by the USAEC Rules and Regulations, (U. S. Atomic Energy
Commission, 165) carry the forceof law. It is of significance, however, that the maximum permissible exposure
of miners to radon and radon-daughters was recently reduced by a factor of three, (Federal Radiation Council,
1969) largely on the basis of epidemiological work which demonstrated that the old standard was unsafe with
respect to the induction of lung cancer. It is not improbable that the maximum permissible exposure for
members of the general population may be reduced at some future time. A conservative approach would be to
take 1 pCi/liter as an assumed upper bound.
If we take the worst case example, the highest measured radon concentration of 1288 pCi/liter, and further
assume that this LPG is burned in a type 1 dwelling (see Figure 7) at a 80°F indoor/outdoor temperature
differential, then the CC>2 content of the model dwelling would be about 1% and the associated radon
concentration (from Figure 8) would be approximately 4.5 pCi/liter, which is somewhat greater than the AEC
Standard.
The standards are written for yearly averages, however, so it is more appropriate to estimate the worst
possible yearly average exposure. We shall assume that LPG is delivered monthly throughout the year
(industry practice) and that each time it is delivered, the radon concentration, Co is 1,288 pCi/liter (our highest
measured value). Then, the average concentration in LPG,C,over each 30-day period would be related to the
half-life of radon ( r y2 =3.825 days) and the delivery interval, 30 days by
c=Co-^- (M exp(-°-693t)dt
C = Co (0.183) = 235 PCi/liter.
The estimate of 4.5 pCi/liter for the maximum instantaneous radon level was based on 80°F indoor/outdoor
temperature differential. If this differential were maintained for one year, there would be a requirement for
80°F x 365 days/yr = 29,200 degree days of heating. The highest reported heating requirement for a city in the
contiguous United States (ASHRAE, 1970) is 10,000 degree days for Duluth, Minnesota. Thus, a maximum
yearly average concentration may be estimated by adjusting the instantaneous maximum for the actual
average radon concentration in the gas and for the actual annual heating requirements.
(C ) = (C ) x 10.000 x 235
air max air max 29,200 1288
= (C ) x 0.062 0.28pCi/liter.
air max
This result is below all of the suggested standards for population exposure to radon-222. Furthermore, it is
within the range of normal indoor radon levels (0.05-3 pCi/liter) reported by UNSCEAR (1962).
It is conservative to assume that the effects of radiation vary linearly with exposure, so it is appropriate to
estimate the population dose in terms of man-rads. To do this, it is necessary to know the relationship between
radon concentration and dose. It has been recognized that in the case of radiation exposure to the lung, the
short-lived daughters of radon-222 are much more important than radon itself (Hague and Collinson, 1967).
Radon is an inert gas and is subject to fewer removal processes than are its daughter products which are not
inert gases and which are often ionized. Thus, in room air or mine air, transient equilibrium between radon
and its daughter products usually does not exist. For these reasons, the quantity "working level" (WL) was
introduced to define a certain quantity of radon-daughters (U. S. Public Health Service, 1957). A working level
is defined as any combination of short-lived radon-daughters in one liter of air which results in the ultimate
emission of 1.3 x 105 MeV of alpha particle energy. It is also equal to the alpha energy released by the radon-
daughters in secular equilibrium with 100 pCi/liter of radon-222. The current exposure level permitted to
uranium miners is equivalent to 40 hours per week at 1/3 WL (Federal Radiation Council, 1969).
The principal health hazard that has been associated with the respiration of radon-daughters is carcinoma
of the lung. The target cells have been identified as the basal cells of the bronchial epithelium (lining of the
bronchial tubes). A considerable amount of work has been done in an effort to find a model that will relate
ambient radon-daughter concentrations to the dose delivered to the basal cells of the bronchial epithelium by
-616-
-------
the deposited daughter products. Studies on this subject have been done by Jacobi (1964 and 1972), Altshuler,
et al., (1964), Hague and Collinson (1967), Parker (1969), and Harley and Pasternack (1972)
The relationship between radon-daughters and dose to the bronchial epithelium is dependent upon a
number of complex factors. These factors include:
(1) The state of equilibrium between radon and its various daughters.
(2) The particle size-distribution of the radon-daughters.
(3) The deposition of radon-daughters in the respiratory tract.
(4) The clearance of radon-daughters from the respiratory tract.
The literature up to 1968 has been reviewed by Parker (1969). He concludes, for typical uranium mine
atmospheres, that the most probable relation between the working level and the absorbed dose to the
bronchial epithelium is 7 rads per working level month (WLM). He further concludes that this value is
probably accurate only to a factor of three.
Barton, et al., (1973) has estimated the dose to the bronchial epithelium due to the burning of 27 ftVday of NG
containing 20 pCi/liter of radon in a kitchen range located in a 8,000 ft3 dwelling with one air change per hour.
They assume a concentration-dose conversion factor of 100 rad/yr at one working level. For the case of one air
change per hour, they calculate a dose rate of 15 millirem per year assuming no radon-daughters in the
combustion products.
In the same type of dwelling, approximately 12.5 ft3 of NG are required per degree day of heating. The
heating value per unit volume of propane (LPG) is some 2 1/2 times that of methane (NG), so the Barton, et al.,
(1973) values would be applicable to LPG with a radon concentration of 50 pCi/liter. Thus, we may estimate
for the stated conditions, that the unit dose equivalent rate use of an LPG burning range is
H _ 15 mrem/yr mrem/yr
lr ~ 50 pCi/liter = °-3 pCi/liter
and for un vented heating is
•
lh~ ' pCi/liter range X 27 (ftVday) ~ 365day/yr-
= 0.00038
. ,,...., /
(pCi/liter) (degree-day/year) .
These coefficients may then be employed with demographic data to estimate the population dose due to
domestic use of LPG in man-rem. The country would be divided into regions small enough to have similar
heating requirements. The population dose equivalent rate (mrem/yr) for the i^h region, {, would be given by
where Cj is the es dmated average radon concentration of delivered propane, ft accounts for radioactive decay
in the domestic storage tanks (f = 0.183 for monthly deliveries), (nr)i is the number of dwellings utilizing LPG
burning ranges in the region, mi is the average number of persons per dwelling, Bj is the average number of
degree days of heating required per year and (n^): is the number of dwellings utilizing LPG in un vented
heating.
The population dose equivalent rate would then be estimated by summing over all N regions.
» = ji cifimi "VWi + ^ihWVi1
At the present time we are prevented from finding H because (nh)i is not reported by the Bureau of the
Census. Computer tapes of census data which would enable us to obtain (n^H are on order, however, and H will
be reported in a future publication.
In conclusion, given the available information, radon in LPG does not pose a threat to the individual
exposed person,. We do not anticipate that the population dose will be of significance either, but judgment is
reserved until H is estimated.
REFERENCES
Altshuler, B., N. Nelson, and M. Kuschner, (1964), Estimation of Lung Tissue Dose from the
Inhalation of Rad
-------
Baily, D. E., (1971), Probability and Statistics, John Wiley and Sons, New York. p. 393.
Barton, C. J., R. E. Moore, and P. S. Rohwer, (1973), Contribution of Radon in Natural Gas to the
Natural Radioactivity Dose in Homes. Oak Ridge National Laboratory Report TM-4154.
Bunce, L. A. and F. W. Satler, (1966) Radon-222 in Natural Gas, Radiological Health Data and Reports,
7:441-444.
Paul, H., G. B. Gott, G. E. Manger, J. W. Mytton, and A. W. Sakakura, (1954), Radon and Helium in
Natural Cos, International Geological Congress. 19th, Algiers 1952, Comptes rendus, Sec. 9, Pt. 9,339-349.
Federal Radiation Countil, (1969), Federal Register 34(10):576-577.
Fish, B. R., (1969), Radiation in Perspective; The Role of Nuclear Energy in the Control of Air Pollution.
Nuclear Safety 10:119-130.
Hague, A. K. M. M. and A. J. L. Collinson, (1967), Radiation dose to the Respiratory System due to
Radon and UN Daughter Products, Health Physics, 13:431-443.
Harley, N. H. and B. S. Pasternack, (1972), Alpha Absorption Measurements Applied to Lung Dose
from Radon Daughters. Health Physics 23:771-782.
International Commission on Radiological Protection, (1959), Recommendations of the
Jn tcrnalionalComnrififiionon Radiological Protection on Permissible Dose for Internal Radiation. Report No.
2, Pergamon Press, New York, N. Y.
Jacobi, W., (1964), The Dose to the Human Respiratory Tract by Inhalation of Short-Lived Radon-222 and
Radon 220 decay products, Health Physics 10:1163-1175.
Jacobi, W., (1972), Activity and Potential Alpha Energy of Radon-222 and Radon-220 Daughters in
Different Air Atmospheres. Health Physics 22:441-450.
Jacobs, D. G., E. G. Struxness, M. J. Kelly, and C. R. Bowman, (1972), Consideration of the
Radiological Impact from the Hypothetical Use of Contaminated Natural Gas from Nuclearly-Stimulated
Reservoirs, Health Physics 22:429-440.
Lucas, H. F., (1957), Improved Low-Level Alpha-Scintillation Counter for Radon, Review of Scientific
Instruments 28:680-683.
McBride, J. R. and D. Hill, (1969), Off-Site Radiological Surveillance for Project Gasbuggy, Radiological
Health Data and Reports 10:535-546.
National Committee on Radiation Protection, (1959), Maximum Permissible Body Burdens and
Maximum Permissible Concentrations of Radionuclides in Air and Water for Occupational Exposure, NBS
Handbook 69.
National Council on Radiation Protection, (1971), Basic Radiation Protection Criteria, NCRP
Publication No. 39.
Parker, H. M., (1969), The Dilema ofLungDosimetry, Health Physics 16:553-561.
Pierce, A. P., G. B. Gott, and J. W. Mytton, (1964), Uranium and Helium in the Panhandle gas Field in
Texax.and Adjacent Areas,U.S. Geological Survey Professional Paper 454-G.
Pierce, A. P., J. W. Mytton, and G. B. Gott, (1955), Radioactive Elements and their Daughter Products
in the Texan Panhandle and Other Oil and Gas Fields in the United States, Geology of Uranium and Thorium,
InternationalConference.
Satterly, J. and J. C. McLennan, (1918), The Radioactivity of the Natural Gases of Canada, Royal
Society of Canada Transactions 12, Series 3, p. 153-160.
Terrill, J. G., E.D. Harward, and I. P. Leggett, (1967), Environmental Aspects of Nuclear and
Con; -en lional Power Plants. Industrial Medicine and Surgery 36:412-419.
UNSCEAR, (1962), Report of the United Nations Scientific Committee on the Effects of Atomic
Radia t ton, General Assembly, Official Records, Seventeenth Session, Supplement No. 16 (A/5216).
U. S. Atomic Energy Commission, (1965), Title 10 of the Code of Federal Regulations, Park 20, Rules
and Regulations, Washington, D. C., U. S. Government Printing Office.
U. S. Bureau of the Census, (1970), Census of Housing: 1970; Detailed Housing Characteristics, Final
Report HCH)B 1. United States Summary.
U.S. Public Health Service, (1957), Control of Radon-Daughters in Uranium Mines and Calculations on
Biolofica I Effects, U. S. Public Health Service Publication 494.
-618-
-------
TABLE 1. Radon-222 Concentration of Wellhead Natural Gases.
STUDY
AUTHOR(S)
Satterlyand
McLennon
Satterlyand
McLennon
Satterlyand
McLennon
Satterly and
McLennon
Paul, et. al.
Pierce, et. al.
Pierce, et. al.
Bunce and
Sattler
McBride and
Hill
YEAR
REPORTED
1918
1918
1918
1918
1951
1955
1964
1966
1969
GEOGRAPHIC
AREA
Ontario, Canada
Alberta, Canada
British Columbia
British Columbia,
Canada
Texas Panhandle
U.S.A.
Texas Panhandle
SW Colorado; NW
New Mexico
SW Colorado; NW
New Mexico
(ANALYSES
21
16
3
= 500
"66 fields"
= 500
307
42
RADON-222 CONCENTRATION
(picocuries/liter @ s.t.p.)
MINIMUM
4
10
390
=10
<100*
<5
0.2
0.33
MAXIMUM
800
205
540
£520
10,000*
1,450
158.8
57.5
AVERAGE
169
62
473
—
460*
= 100
24.6
16.4
*Reported at "reservoir temperature and pressure"; the pressure and temperature were not stated, but 200-300
psi is typical.
TABLE 2. Boiling Point of LPG Constituents and Radon.
NAME
METHANE
ETHANE
RADON
PROPANE
ISOBUTANE
BUTANE
FORMULA
CH4
(CH3)2
Rn
(CH3)(CH2)(CH3)
(CH3)3(CH)
(CH3)3(CH2)2(CH3)
BOILING POINT (°Q
-161.5
- 88.3
- 61.8
- 42.2
- 10.2
- 0.5
TABLE 3. Constituents of Retail LPG in Percent.*
RETAILER CODE
1
2
3
4
5
6
CH4
METHANE
2.0
0.2
0.2
1.8
0.5
0.44
C2H6
ETHANE
9.5
3.8
3.8
6.0
7.2
2.8
C3H8
PROPANE
88.0
95.9
96.0
91.5
91.3
25.58
C4H1Q
n-BUTANE
0.5
1.5
...
0.7
0.5
0.16
C4Hio
i-BUTANE
...
...
...
...
0.5
...
'Samples taken during week of Nov. 26,1972 and analyzed using gas chromatography.
-619-
-------
TABLE 4. Radon Concentrations of Retail LPG (pCi/1) in the Houston, Texas Area: Monthly
Averages of Weekly Samples.
MONTH
July, 1972
August
September
October
November
December
January, 1973
February
March
April
May
June
Average
*Sampling for this retailer w
for averaging purposes.
RETAILER CODE NUMBER
1
9.4
63.1
27.7
20.3
17.9
6.2
3.4
53.0
80.6
105.6
78.3
115.0
48.4
as disco
2
5.6
62.1
42.5
48.3
27.8
43.9
33.0
26.6
30.5
55.8
10.8
37.5
35.4
ntinuec
3
6.0
41.6
10.0
30.2
63.8
25.6
2.2
21.2
10.5
6.0
46.4
34.8
24.9
after I
4
130.9
160.4
79.4
85.8
48.4
2.9
0.6
11.0
40.3
77.7
59.4
25.8
50.1
Decembe
5
57.6
32.6
23.2
34.4
14.9
9.7
41.6
11.5
3.5
30.2
38.7
23.3
26.8
r, 1972
6
233.4
197.0
148.0
163.7
220.4
176.6
139.2
167.2
204.9
149.3
178.2
173.6
179.3
and valu
7
<0.5
0.7
<0.5
4.6
4.1
3.4
<0.5*
<0.5*
<0.5*
<0.5*
<0.5*
<0.5*
1.4
esofO.Sw
AVERAGE
63.3
79.7
47.3
55.3
56.7
38.3
32.4
41.8
52.9
60.7
59.3
58.8
53.7
rere assumed thereafter
TABLE 5. Relationship Between Radon-222 Concentration in LPG and Distance From a Major
Gas Processing Area.
DISTANCE
(miles)
<25
25-200
200
NO. OF
SAMPLES
103
23
31
AVERAGE
CONCENTRATION
(pCi/liter)
87.8
63.7
45.6
TABLE 6. Relationship Between Radon-222 Concentration in LPG and the Type of Container
From Which the Sample was Obtained.
CONTAINER
Large Tank
Truck
Small Tank
NO. OF
SAMPLES
21
20
30
AVERAGE
CONCENTRATION
(pCi/liter)
162.9
125.8
31.4
-620-
-------
TABLE 7. Air Changes Taking Place Under Average Conditions in Residences, Exclusive of Air
Provided for Ventilation.*
KIND OF ROOM OR BUILDING
Rooms with no windows or exterior doors 1/2
Rooms with windows or exterior doors on one side 1
Rooms with windows or exterior doors on two sides 11/2
Rooms with windows or exterior doors on three sides 2
Entrance Halls 2
*For rooms with weatherstripped windows or with storm sash, use 2/3 these values.
Number of Air
Changes Taking
Place per Hour
-621 -
-------
LPG PRODUCTION
(South Texas)
8000
RETAIL SURFACE
STORAGE FACILITY
19,000 gal.
80,250
RETAIL TRUCK
#2 (2,100 gal.)
Pipeline
LPG WHOLESALE
DISTRIBUTOR
(Houston Area)
RETAIL TRUCK
#1 (2,100 gal.)
250
4750
80,000
BOTTLE STATION
(500 gal.)
35,000
5000
UNDERGROUND
STORAGE
RESIDENTIAL, COMMERCIAL, AND RECREATIONAL CUSTOMERS
Figure 1. Distribution network for Retailer #2. The numerals adjacent to the arrows represent the
approximate gallons of LPG moved per month along the various pathways during a typical winter season.
-622-
-------
05
to
w
120-239
Figure 2. Distribution of the exposure index (degree days) X (unvented heaters) by state for the contiguous
United States. The values are expressed in millions.
-------
Figure 3. Distribution of gas processing plants in the contiguous United States. Each dot represents one gas
processing plant.
-------
Si
to
01
Figure 4. Distribution of retail LPG sample sites. Each dot represents one sample except in the Houston,
Texas area where each dot represents one of the seven regularly sampled retailers.
-------
80 •
.-? 60
o
Q.
2 40
cr
LU
o
z
o
o
20
JUL SEP NOV JAN MAR MAY
AUG OCT DEC FEB APR JUN
1972-73
Figure 5. Second order regression analysis of the monthly average values of radon concentrations found in
weekly Houston, Texas area retail LPG samples.
-626-
-------
Figure 6. Distribution by region of radon concentrations found in retail LPG samples. The three numbers in
each region represent, respectively, the maximum, mean, and minimum concentrations (pCi/liter).
-------
05
tc
00
1.2
1.1
1.0
J 0.9
^
O
o
u
u
B
| 0.7
6
u
0.6
0.5
u 0.4
0.3
0.2
0.1
TYPE 3
20 40 60 80
TEMPERATURE DIFFERENCE (°F)
100
Figure 7. Calculated CC>2 concentrations vs indoor-outdoor temperature differences, for three dwelling types
assuming the use of unvented heating and one air change per hour.
-------
05
(0
200 400 600 800 1000 1200
RADON CONCENTRATION IN PROPANE (pCi/liter)
1400
1600
1800
Figure 8. Theoretical concentrations of radon contributed by combustion vs concentration in propane for
several assumed CC>2 concentrations.
-------
PERSONNEL DOSIMETRY OF RADON AND RADON-DAUGHTERS*
K. Becker
Health Physics Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
Abstract
During the past five years there has been increasing interest in monitoring the individual exposure of
persons to radon and, more importantly, the radon progeny attached to aerosolparticles. Several devices have
been proposed. These include the following:
(1) passive detectors based on nuclear track emulsions and track etching foils;
(2) semi-active systems in which the aerosols are collected electrostatically and detected by
scintillator/film combinations or track etching; and
(3) active systems based on the collection of aerosols on a filter, and integrating measurement of the alpha
flux from the filter by TLD or track films.
All these systems have drawbacks, which have been minimized in a small active system based on alpha
particle track etching which was originally developed at ORNL in 1969. Since then it has undergone
substantial laboratory and field testing, and has been subjected to various improvements. The system is
briefly discussed.
PASSIVE DETECTORS
Various authors in the U.S., Germany, and the Soviet Union (Geiger, 1967; Jacobi and Koeppe, 1968;
Markov, et al., 1970; and Evans, 1972) suggested the use of nuclear track emulsions and visual alpha particle
counting in the microscope. In a special radon film badge (Geiger, 1967 and Evans, 1972), a Kodak NTA film
was separated from a central recess by a light-tight window sufficiently thin to permit the passage of alpha
particles, with the radon-containing air diffusing into the recess. A coefficient of variation of ±30% for one
MFC-week was claimed in early laboratory tests (Geiger, 1967), but other laboratory and field tests (White,
1969) did not produce satisfactory results, largely because of intense fading in the NTA emulsion. In high-
humidity areas, such a badge cannot be used for more than 2-3 days without a change of film, but a week of
exposure at MFC would be required for a ± 30% statistical error.
Several groups have, therefore, studied the feasibility of the use of sensitive polymers (mostly cellulosics
such as cellulose nitrate, acetate, triacetate, or acetobutyrate) in which alpha particles are detected by track
etching (for a recent review on track etching dosimetry see Becker, 1972). The latent tracks in such detectors
are sufficiently stable (Figure 1) and are easily countable — either by visual scanning (Becker, 1969) or
automatic spark-counting of the etched perforations in V15 um cellulose nitrate films. In a cellulose triacetate
foil of 3 cm2, for example, exposure to one weekly MFC of radon in equilibrium with its progeny produces =100
tracks (Becker, 1969). The aerosol-attached radon-daughters can be excluded by placing a filter between the
foil and the environment. Detectors containing one foil which is exposed to unfiltered air, and one exposed to
filtered air (Figure 2[a]) were found to be sufficiently sensitive (Becker, 1970), but not accurate enough under
simulated field conditions.
Since those early experiments, various other authors in the U. S.(Lovett, 1969 and Alter and Price, 1972),
Poland (Domanski, et al., 1973), Japan (Kurosawa, 1972), and France (Anno and Commancy, 1971) exposed
open, unprotected foils — mostly of cellulose nitrate — to radon-containing atmospheres. This method,
although simple and inexpensive, seems unlikely to provide sufficiently precise data. Single detectors are
unable to distinguish between radon gas and attached daughter products. Also.in field tests, contamination of
the foils by dust, oil films from the exhaust of diesel engines, etc., lead to grossly erroneous results (White,
1969).
SEMI-ACTIVE SYSTEMS
Several semi-active systems for a more sensitive and/or selective measurement of the radon progeny have
been investigated. Thermal precipitation of aerosol particles on the plastic detector opposite to a heating
device did not result in a substantial increase in sensitivity, largely because increased latent alpha track
fading took place at elevated temperatures (Auxier, et al., 1971). Electrostatic deposition did not turn out to be
very practical at high humidities in our experiments, but others (Costa-Ribeiro, et al., 1969) have proposed a
rather bulky battery-operated electrostatic collection device. It is based on the collection of radium A from the
decay of pure, filtered radon on a charged thin hemispherical aluminum foil, and a scintillator behind the foil
*Research sponsored by the U.S. Atomic Energy Commission under contract with the Union Carbide
Corporation.
-630-
-------
which converts the alpha particle energy into light flashes to be integrated by means of a conventional
photographic film. Unfortunately, this procedure results in a pronounced exposure-rate dependence of the
response due to the reciprocity failure of photographic emulsions.
ACTIVE DETECTORS
Active radon progeny personnel dosimeters are usually based on the collection of aerosols on the surface of
suitable filters through which ambient air is forced, and subsequent integrating measurement of the alpha
particles emitted from these aerosols. One such system in which TLD-impregnated Teflon discs are used as
detectors has been developed at Colorado State University (McCord, et al, 1969 and Schiager and Savigner,
1972). It can be made sufficiently sensitive (permitting measurements in the 0.1 to 10 MFC range), but the
substantial air-flow which is necessary requires a fairly large and heavy pumping system and battery. There
have also been problems with the accuracy of such detectors in laboratory as well as field tests (White, 1969
and Schiager and Savigner, 1972) probably related to clogging of the filter, fading of the TLD's, and other
inherent inaccuracies of TLD.
In another system which has been developed by our group at Oak Ridge (Becker, 1970 and Auxier, et al,
1971) a 15fim cellulose nitrate foil to be evaluated by spark-counting (Johnson, et al., 1970) is used for alpha
particle detection. The sensitivity of this system, which permits the detection of single alpha particles, is so
high that only a small air flow is required, and clogging of the filter is minimized. A schematical diagram of
this small (115 x 23 mm) and light-weight (67 g) device is given in Figure 2(b) and a photograph is presented in
Figure 3. The details of its construction have been described in an earlier publication (Auxier, et al., 1971). A
rechargable battery operates the motor without loss of efficiency for 12-15 hours, which is more than sufficient
for an 8-hour shift. The dosimeter can easily be attached to a miner's helmet (Figure 4).
During tests at HASL, the performance of this detector turned out to be satisfactory in comparison to other
detectors which were simultaneously tested (White, 1970), but some problems were also uncovered. For
example, the speed of several of the initially used motors was not very uniform; there were occasional air-leaks
around the filter edges, and the reproducibility of the cellulose nitrate foil preparation was less than desirable.
The mechanical problems have since been corrected, and improved cellulose nitrate foils have been prepared
(Lupica, 1974), because the commercially available 18/im cellulose nitrate (LR 115 made by Kodak-Pathe,
France) appears to be too thin for reproducible spark-counting results.
Similar systems have since been tested also in Germany (Haider and Jacobi, 1972) and France (Chapuis, et
al., 1972). Various additional parameters such as the effect of distance between the alpha source and the
detector foil on the alpha particle spectrum and, consequently, the registration efficiency (Figure 5) have been
investigated and have helped to improve the performance.
With the radon progeny concentration in the air of a typical mine fluctuating substantially with location
and time, stationary or portable instruments provide only a very crude estimate of an individual miner's lung
exposure. A light-weight, reliable personnel dosimeter which does not require an outside power source is
highly desirable and apparently feasible, but some additional de-bugging and testing efforts may be required
before a sufficiently rugged version of such a dosimeter can go into large-scale production and use.
REFERENCES
Alter, H. W. and P. B. Price, (1972), U. S. Patent 3,665194.
Anno, J. and L. Commanay, (1971), Am. Phys. Biol. Med. 5,65.
Auxier, J. A., et al., (1971), Health Phys. 21,126.
Becker, K., (1969), Health Phys. 16,113.
Becker, K., (1972), Chapter 2 in Topics in Radiat. Dosimetry (F. H. Attix, Ed.), Academic Press, New York.
Becker, K., (1970), in ORNL-4446, p. 266; and U. S. Patent 3,505,523.
Chapuis, A. M., D. Dajlevic, P. Duport, and G. Soudain, (1972), Proceed. 8th Interact Conf. Nucl.
Photography, Bucharest 1972.
Costa-Ribeiro, C. Thomas, R. T. Drew, M. E. Wrenn, and M. Eisenbud, (1969), HealthPhys. 17,193.
Domanski, T., W. Chruscielewski, and J. Liniecki, (1973), Health Phys., in press.
Evans, R. D., (1972), U. S. Patent 3,655,975.
Geiger, E. L., (1967), Health Phys. 13,407.
Haider, B. and W. Jacobi, (1972), Rep. BMBW-FB K 72-14.
Jacobi, W. and P. Koeppe, (1968), Atomkernenergie 13,429.
Johnson, D. R., R. H. Boyett, and K. Becker, (1970), Health Phys. 18,424.
Kurosawa, R., (1972), Hoken Butsuri 7,105.
Lovett, D. B., (19&9), Health Phys. 16,623.
Lupica, S.B., (1974), To be published. ^ i T • j T * D -U * 10
Markov, K. P., N. V. Ryabou, andK. N. Stas, (1970), Tr. Soyuznogo Nack. Issied. Inst. Pnborostr. 12,
184
McCurdy, D. E., K. J. Schiager, and E. D. Flack, (1969), Health Phys. 17,415.
Schiager, K. J., and N. F. Savignac, (1972), Rep. COO-150-21.
White, O., (1969), AEC-HASL Techn. Mem. 69-23A, and pers. comm.
White, O., Jr., (1970), AEC-HASL Techn. Mem. 70-3.
-631 -
-------
120° C,EXTENDED ETCHING
10 20 50 100 200 500 1000 2000 5000
STORAGE TIME AT INDICATED TEMPERATURES (min)
Figure 1. Percentage of original alpha particle etch pits in cellulose acetobutyrate (Triafol B made by Bayer,
Leverkusen, Germany) which remains visible after storage of the foil at different temperatures between
exposure and etching.
-------
05
CO
to
(a) WITHOUT FORCED
AIR CIRCULATION
CAP
DUST SCREEN
DETECTOR FOIL
(Rn +DAUGHTERS)
DETECTOR
FOIL (Rn)
AEROSOL FILTER
CAP
AIR INLET
AIR OUTLET
CAP
DETECTOR FOIL
(Rn DAUGHTERS)
AEROSOL FILTER
FAN
MOTOR
RECHARGEABLE
BATTERIES
(b] WITH FORCED
AIR CIRCULATION
Figure 2. Schematical diagram of a "passive" two-foil radon and radon progeny detector (left), and of the
"active" ORNL radon progeny personnel dosimeter (right).
-------
Figure 3. ORNL radc
+1 ii..i ^—^^_—^
-------
Figure 4. Dosimeter attached to a miner's helmet.
-------
(5 x 104)
3 4
a PARTICLE ENERGY (MeV)
Figure 5. Variation of the original alpha particle spectrum from radon-daughter products as a function of the
distance between source and detector in air (Chapuis, et al., 1972).
-636-
-------
RADON ADSORPTION BY ACTIVATED CARBON IN URANIUM MINES
J.W.Thomas
Health and Safety Laboratory
U. S. Atomic Energy Commission
New York, N.Y. 10014
Abstract
A brief study was made of the feasibility of using activated carbon to remove radon in uranium mines with
the use of a multiple bed-temperature cycling process. Air containing radon is passed through a carbon bed at
ambient temperature until the bed is considered exhausted. The bed is then heated to about 120°C and the
radon desprbed, using a much smaller volume of air than was used during adsorption. The desorbed radon,
after cooling, ispassedinto a second bed. Since the capacity of the carbon depends primarily on the volume of
air passed into it rather than the quantity of radon adsorbed, the second bed will adsorb much more radon
than the first. Multiple stages of this process are possible. It was extrapolated from experiments that about 10
liters per minute or more of air could be purified continuously of radon per liter of carbon.
INTRODUCTION
This investigation is a preliminary study of the feasibility of using activated carbon to remove radon in
uranium mines. Although radon-daughters are the principal inhalation hazard, rather than radon itself,
removal of the daughters alone is unsatisfactory since unacceptable concentration levels of the daughters are
rapidly produced from relatively small amounts of radon. At present, the radon-daughter hazard is controlled
by high air ventilation rates. This can be quite expensive, particularly for deep mines, since long ventilating
shafts and large exhaust blowers are required. It appears, therefore, that alternate methods of control could
present an advantage.
The use of carbon to remove radon in mines has been previously considered (Coleman, et al, 1956;
Fusamura, et al., 1963; and Schroeder, et al., 1968), but it was concluded that the quantity of carbon required
would be excessive. These previous investigations, however, did not consider the possibility of amplifying the
capacity of carbon for removing radon by an adsorption-desorption process, such as the multiple bed-
temperature cycling process described in this work.
A basic principle of the present process is that the rate of penetration, or breakthrough, of a noble gas
through a carbon bed is independent of the noble gas concentration when the gas is present in low
concentration. This has been verified ,for krypton and xenon (Siegwarth, et al., 1972) and for radon in
preliminary experiments in this laboratory. It follows from this that a given quantity of carbon will purify a
given quantity of air, regardless of the radon concentrations in the entering air.
Let this air cleaning capacity be expressed as cp volumes of aircleanableby aunitvolumeof carbon.Hence,
if $ liters of air are passed through a unit volume carbon bed at mine radon concentration, the flow stopped,
the radon driven off in /10 liters of air by application of heat, and the effluent cooled and passed into a
second bed the same size as the first, the second bed would be able to accept 10 "charges" of enriched radon
from the first bed. Hence, the process is said to have an "amplifying factor" of 10; that is, the second bed would
have a capacity of cleaning 10 volumes of the original air. Extending the process, a third bed would have a
capacity of cleaning 1000 volumes of original air, a fourth (= 1,000) volumes, etc. Hence, it is seen that
such a multiple bed-temperature cycling process could amplify the radon adsorption capacity of carbon by
several orders of magnitude.
It was already known from previous work that a good grade of activated carbon would purify about 1,000
volumes of air of radon per unit volume of carbon (<*> = 1,000). The presence of water vapor and carbon dioxide
was deleterious, but not excessively so. Experiments also confirmed that did not depend on the radon
concentration. Hence the principal objective of the present work was to ascertain the quantity of air needed
in the desorbing step to drive off the adsorbed radon and to determine the amplifying factor, defined as the
volume of air cleaned on the adsorption step divided by the air volume used for desorption. If the amplifying
factor of the process could be made sufficiently high, without excessive heating and cooling requirements, the
process could have practical application.
APPARATUS AND PROCEDURE
The process consists of an adsorption step and a desorption step. In the adsorption step a radon-air mixture
ispassedinto a carbon bed at a known flow rate, radon concentration, and humidity, and the "transmission"
vs. time curve determined. This dimensionless, fractional transmission T is defined as the concentration of
radon in the carbon bed effluent divided by the influent concentration at any time. Details of the adsorption
test apparatus are shown in Figure 1. Radon was generated by bubbling air saturated with water at 1 I/mm
through a radium chloride solution containing several millicuries of 226Ra. The rate of production of radon
was found to be about 900,000 pCi/min. An "aerosol canister" containing two layers of high-efficiency filter
paper was used downstream of the bubbler to remove any entrained aerosol from the radium bubbler. The
main air supply for the apparatus entered through flow meter F-l at 69 1/min and was humidified to about 9.0
-637-
-------
mg/1 by means of the humidifying tubes shown in the figure. The total flow influent to the test chamber was
69+1 = 70 1/min, which provided a radon concentration of about 900,000/70 = 13,000 pCi/1. The carbon bed
(actually a gas mask canister) was tested at a flow of 56.6 1/min, which was maintained by two flow meters,
F-3 at 48.6 1/min and F-4 at 8 1/min. The 8 1/min flow line was needed to measure radon concentrations in
the effluent from the test canister. This line was provided with a bypass, so that the total flow of 56.6 1/min
could be maintained whether or not a concentration sample was being obtained.
To prepare for an adsorption test, the air flow was turned on through flow meters F-l and F-2, and about 2 or
3 hours allowed for the establishment of equilibrium conditions. The waiting period is necessary to allow time
for accumulated radon in the radium bubbler to be swept out into the hood, and to allow the water in the
humidifier to reach an equilibrium temperature. It was found that after the waiting period the apparatus
would produce a constant (± 5%) radon concentration for 24 hours/day for a period of days or weeks, with
replenishment of the water-level in the water saturators and humidifying tubes every few days. Immediately
prior to the test, the input radon concentration was measured at 8 1/min using the two-filter method
(Thomas, et al., 1970), sampling from line 1 of the figure. Usually four determinations were taken prior to a
test, and four after completion of the test. After establishing the input radon concentration, the canister was
attached to line 2 of the figure and the radon concentration in the canister output determined continuously to
permit calculation of radon transmission. Transmission curves were plotted from the test data by considering
the average transmission obtained over the three-minute sampling period to be the instantaneous
transmission at the midpoint of the sampling period.
The carbon bed tested was that of a commercial gas mask canister. This canister contained about 900 ml
(400 grams) of activated carbon (Witco Chemicals Co., Type AC-337, 6/16 mesh). This is a petroleum base
carbon and has a surface area of about 1,400 mVg. The carbon bed in the canister was nearly rectangular,
with a cross sectional area of about 60 cm2 perpendicular to the direction of air flow, and height of about 15 cm
in the direction of air flow. Hence, the superficial air velocity through the bed was 56,600/(60) (60) = 15.7 cm/s.
The desorption test apparatus was relatively simple, Figure 2. Heating or cooling air was supplied by a
small blower equipped with an electrical resistance heater (Heat Gun Model 301, Master Appliance Corp.,
Racine, Wis.). This device supplied a blast of air at about 120 1/min, and by means of a variable transformer
on the resistance heater of the gun, temperatures could be controlled up to about 250°C. The nozzle of the gun
was installed through the side of a 20-liter container and the canister placed inside for heating. A water
condenser was placed in the canister effluent line to collect the several grams of water which are evolved in
each desorption step. A capillary type flow meter was used to measure flow rate, and a standard wet test meter
to register total volume of desorbing air. Two protective canisters in series were inserted downstream of the
apparatus as shown in the figure. These canisters permitted confirmation that a second carbon bed would
accept many charges of radon from the canister undergoing desorption, and they also prevented evolution of
radon into the laboratory.
The desorption procedure was as follows: the canister was removed from the adsorption apparatus and
placed in the air bath, pre-set at the desired temperature. Immediately, air from the hot air bath was drawn
through the canister at a low flow rate of about 1 1/min. Theoretically, it is inefficient to pull air through the
canister while it is still at room temperature, as there is little radon desorption per unit volume of desorbing air
and the capacity of the second bed for radon is thereby partially wasted. However, it was necessary to pull air
at a low flow through the canister at the start of the heating step; otherwise the gases evolved from the carbon,
as it heats up, would go out the canister inlet. As these gases constitute part of the desorbed air volume, their
volume must be measured by the downstream wet meter.
After a preliminary heating time of about 15 to 25 minutes, at the low flow rate the air flow through the
system is increased to sweep off most of the adsorbed radon from the hot carbon. From the total volume of
desorption air and the volume of air cleaned in the adsorption step, the amplifying factor of the process was
calculated. The temperature of the carbon in the heating step was not measured directly, but was estimated
roughly from the temperature of the air entering and leaving the canister. After the heating period, the
canister was removed from the hot air bath and cooled for 15 minutes. The total desorption time, including
cooling, was 45 minutes in each test series.
RESULTS AND DISCUSSION
Four series of tests were done using the same canister. The same adsorption procedure, but a different
desorption procedure, was used in each series of four to six replicate runs. This method was adopted to obtain
transmission curves in each series that were reproducible and not affected by the treatment of the previous
test series. Figures 3-6 show the adsorption transmission curves for the four different series. In each of the
figures only two of the adsorption runs are plotted. Data from the other runs, except for the first one or two, fell
on the same line and show that the process was highly reproducible.
All of the figures show a characteristic shape; the transmission at zero adsorption time is moderate or
high; this drops and then again rises with time. The initial transmission in test series MT-2 and MT-4 was
greater than unity. This may be understood by consideration of process details. During desorption, the radon
concentration in the gas phase of the carbon bed is higher than during adsorption, because of the higher
temperature. Hence, if desorption is incomplete and the canister is still warm, the initial adsorption flow can
sweep off the high radon concentration and result in transmissions over unity. The initial high transmission
could have been avoided by use of more desorbing air, but probably the advantage gained in the adsorption
-638-
-------
step thereby would have been more than neutralized by the increased quantity of desorbing air required. In all
four series, the now of desorbing air was in the same direction as it was in the adsorption test. In
retrospect, it would have been better to have used reverse flushing which would have effected higher
desorption efficiency, since at the end of the adsorption step the radon concentration in the carbon is higher at
the inlet of the carbon bed than at the exit.
The transmission curves of each figure were integrated graphically to obtain the fraction F of the total input
radon adsorbed by the canister in the 40-minute test period. The quantity of interest is the net amount of air
completely cleaned of radon in the adsorption step, since, for example, cleaning X liters of air per minute at
50% efficiency is as effective in reducing ambient radon concentrations as is cleaning X/2 liters per minute at
100% efficiency. The volume of air completely cleaned VA, was therefore calculated from the flow rate and test
time to be
VA=(56.6) (40) (F) = 2,264 F liters
A summary of results for the four test series is given in Table 1. As previously mentioned, two different flow
rates were used during desorption. The flow rates given are approximate, hence, the total desorption volume,
VD, which was measured with an accurate wet test meter, does not agree exactly with the volume calculated
from the flow rates. The temperatures listed are those of the air entering the canister (the air bath temperature)
and of the air leaving the canister. The latter values have a range, since the air flow was started immediately
after insertion of the cold canister into the hot air bath. From these inlet and outlet temperatures it was
estimated that the average carbon temperature was about 100°C for the 120°C air bath temperature and
120°C for the 150°C air bath temperature. The last column shows the amplifying factor obtained for the
different desorption treatments. Comparison of test series MT-1 and MT-2 shows an improvement of about a
factor of 2 in amplification by reducing the desorption air volume from 248 liters to 101 liters. This reduced
desorption volume more than compensated for the resulting decrease in net volume of air cleaned in the
adsorption step, which dropped from 1,630 to 1,290 liters. Test series MT-3 was done to determine the gain that
could be achieved by increasing heat; for an estimated increase of 20°C in average carbon temperature, the
amplifying factor rose from 12.8 to 16.1 due to lower transmission on the adsorption step. The last test series
MT-4, was done to see if the desorbing air volume could be further reduced without affecting the adsorption
transmission curve too adversely. The somewhat lower net volume of air cleaned on the adsorption step was
more than compensated for by the decrease in desorption air volume and the amplifying factor increased from
16.1 to 23.4. It seemed apparent that the amplifying factor of the process could be further increased by other
modifications of desorbing conditions.
Let us consider the application of the results of test series MT-4 to a hypothetical system containing 4 carbon
beds, to estimate the net quantity of air which could be cleaned continuously per liter of carbon. The process is
assumed to have three one-liter carbon beds and one two-liter bed, and is designed to clean 60 1/min of air at
80% radon removal efficiency. Two primary carbon beds A and B operate alternately in parallel at 60 1/min
during the 45-minute adsorption time, and at an average flow rate of 2 1/min during the desorption time, also
45 minutes. Since the radon removal efficiency is 80%, the net quantity of air cleaned in the adsorption step
(VA) is (60) (0.80) (45) = 2,160 liters. The quantity of desorption air used (VD) is (45) (2) or 90 liters. Hence, the
amplification factor is 2,160/90 = 24. The primary beds A and B are desorbed into bed C, which also contains
one liter of carbon. With the amplifying factor of 24, this bed has the capacity to remove 24 charges of desorbed
radon over a time period of 24 x 45 minutes, or 18 hours, before requiring desorption. After the 45 minute
desorption of bed C, at 2 1/min, beds A and B are put back on stream, with bed C receiving the desorbed radon
from beds A and B as before. This system results in beds A and B operating 96% of the time.
Bed D receives one charge of desorbed radon from bed C every 18.75 hours. If it contained one liter of carbon
as do beds A, B, and C, it would have an amplifying factor of 24, and a holdup time of 18.75 days. To avoid the
necessity of desorbing this bed, it is made to contain 2 liters of carbon and, hence, has a holdup time of 37.5
days, or about 10 radon half-lives. The flow from bed D will, therefore, contain negligible amounts of radon.
The overall process, including the flow from beds C and D, removes about 80% of the radon in 60 1/min of
incoming air with the use of 5 liters of carbon. Hence, the net cleaning capacity of this hypothetical system is
about 10 1/min per liter of carbon. It seems very probable that optimization of the system, particularly of
desorbing conditions, could greatly increase the efficiency of the process. The process was evaluated with the
adsorption taking place at room temperature, about 25°C. Better performance would be expected at the lower
temperatures existing in mines.
REFERENCES
Coleman, R. D., H. L. Kusnetz, P. F. Woolrich, and D. A. Holaday (1956), Radon and Radon-
Daughter Hazards inMine Atmospheres, Amer.Ind.Hyg.Asaoc. Quart. 17,405. .
Fusamura, N., R. Kurosawa, and S. Ono (1963), On the Study of Radon Removal with Active Carbon,
Nippon KogyoKai 79,590. (Sandia Translation SC-T-64-904, Sandia Base, Albuquerque, N.M.).
Schroeder, G. L., P. Groer, M. M. Costello, and R. D. Evans (1968), Refreshing and Secondary Use of
Mine Air,inRadiumandMesothoriumPoisoningandDosimetryandInstrumentation Techniques in Applied
Radioactivity, Report MIT-952-5, p. 323, Massachusetts Institute of Technology, Cambridge, Mass.
-639-
-------
Siegwarth, D. P., C. K. Neulander, R. T. Pao, and M. Siegler (1972), Measurement of Dynamic
Adsorption Coefficients for Noble Gases on Activated Carbon; in the Proceedings of the Twelfth Air Cleaning
Conference, p. 28. Available from the National Technical Information Service, U. S. Dept. of Commerce,
Springfield, Va.
Thomas, J. W. and P. C. LeClare (1970), A Study of the Two-Filter Method for Radon-222, Health Phys.
10,113.
TABLE 1. Summary of Results.
Adsorption Step Desorption Step
Test
Series
MT-1
MT-2
MT-3
MT-4
F
0.72
0.57
0.57
0.58
VA
(liters)
1,630
1,290
1,610
1,310
Time Flow Rate
(min) //min
(15
\15
(25
V 5
/25
V 5
(OK
5
10
16
1
16
1
16
1
6.4
Temperatures
In (°C)
120
120
120
120
150
150
150
150
Out(°C)
30 to 70)
75 to 95'
30 to 75 i
75 to 85)
30 to 85 I
85 to 100'
30 to 85 i
85 to 100 '
VD
(liters)
248
101
100
56
Amplification
Factor
VD/VA
6.6
12.8
16.1
23.4
- 640-
-------
HUMIDIFIERS
COMPRESSED
AIR 5 P. S.I.
VACUUM
STABILIZER
CANISTER TEST CHAMBER
Figure 1. Radon adsorption apparatus.
-641 -
-------
to
CONDENSER
THERMOMETER
AIR BATH
FLOWMETER
TEST CANISTER
HOT AIR IN, 120° OR I50°C
CLEAN-UP CANISTERS
PUMP
r
WET METER
Figure 2. Radon desorption apparatus.
-------
I.2H
1.0 —
0.8
DESORBING CONDITIONS
AIR BATH TEMP- 120°C
AIR VOLUME - 248 i
w
| 0.6
tn
z
cr
0.4
0.2
10 20
ADSORPTION TIME, MINUTES
Figure 3. Transmission curve, test series MT-1.
40
-------
DESORBING CONDITIONS
AIR BATH TEMP-120' C
AIR VOLUME -101 1
ADSORPTION TIME, MINUTES
Figure 4. Transmission curve, test series MT-2.
-------
01
1.2 —
1.0 —
0.8
0.6
en
1 0.4
H-
o.;
DESORBING CONDITIONS
AIR BATH TEMP-I50°C
AIR VOLUME- 100 i
ADSORPTION TIME, MINUTES
Figure 5. Transmission curve, test series MT-3.
-------
S
05
1.2
1.0
0.8
z
o
<
oc
0.6
0.4
0.2
DESORBING CONDITIONS
AIR BATH TEMP-I50°C
AIR VOLUME- 56 1
10 20 30
ADSORPTION TIME, MINUTES
Figure 6. Transmission curve, test series MT-4.
40
-------
X. Environmental Standards
for Noble Gases
-------
ENVIRONMENTAL RADIATION STANDARDS CONSIDERATIONS FOR KRYPTON-85 AND
RADON
J. E. Martin and W. A. Mills
Criteria and Standards Division
Office of Radiation Programs
Environmental Protection Agency
Washington,B.C. 20460
Abstract
To ensure adequate public health protection, the Environmental Protection Agency is charged with the
responsibility of establishing appropriate standards for controlling radioactive materials in the environment
Two noble gases are presently of interest to EPA, namely naturally occurring radon (radon-222) and fission-
produced krypton (krypton-85). To establish standards pertaining to these gases, information on source
concentration, source control and effectiveness, and estimates of health effects must be utilized when
available, trom this viewpoint, the control of krypton emission from nuclear fuel reprocessing plants and the
concentrations of radon in natural gas will be discussed.
INTRODUCTION
Two important noble gases associated with the natural existence and use of uranium are radon-222 and
krypton-85. Although radon and its decay products have always existed in the world's atmosphere, especially
over the land areas, the discovery of uranium fission energy has introduced significant new situations
whereby man is exposed to radiation from radon and its decay products. This use and the projected future
accelerated use of uranium and its fissile by-products as energy resources result in the creation of krypton-85
and several other noble gases as fission products. In contrast to radon, krypton-85 did not exist, for all
practical purposes, in the atmosphere until after the discovery of fission in the late 1930's. This radionuclide
is, however, being generated and discharged to the environment at an increased rate because of the growth of
nuclear power. The purpose of this paper is to examine some of the more important considerations for
environmental radiation standards for these two important noble gases in terms of uses of uranium and by-
product material to produce energy.
CONSIDERATIONS FOR KRYPTON-85
1. Environmental Sources of Krypton.
The environmental sources of krypton are all related to nuclear fission. Although spontaneous fission of
uranium in the earth may result in trace quantities of krypton-85 in the earth's atmosphere, the first major
source of krypton-85 in the environment was the testing of nuclear weapons which began in the late 1940s.
Most of the krypton-85 in the environment produced from this source occurred in a relatively short period
from about 1956 to 1962. Since most tests have been conducted underground since 1962, the significance of this
source has decreased considerably although the krypton introduced into the atmosphere in this relatively
short period still persists, decaying with a half-life of 10.7 years. Not all of the weapons-related krypton-85
came directly from the nuclear detonations themselves, but appreciable quantities also resulted from
chemical processing of reactor fuel from the production reactors which supported the production of fuel for the
weapons program. By about 1962, these activities resulted in an ambient level of krypton-85 in the atmosphere
of about 6-8 pCi/m (Ehhalt, et al, 1964).
The largest contributor of krypton-85 to the environment at the present time is the generation of electricity
by nuclear fission. Several countries have a nuclear power program. This energy source is expected to increase
rapidly over the next several decades with a concomitant increase in krypton-85 levels especially if current
control practices continue. With the exception of small amounts retained for radioisotope use, all krypton-85
currently produced eventually finds its way to the atmosphere. The major source is the chemical reprocessing
plant which breaks the containment barrier of spent reactor fuel to reclaim unfissioned uranium and other
products. The U.S. nuclear industry alone is estimated to produce about 10 billion curies of krypton-85 by the
year 2020 (EPA, 1973), all of which would be discharged to the atmosphere under current control practices.
Reactors also routinely discharge krypton-85 to the atmosphere; however, the quantities are small in
comparison to those from fuel reprocessing plants. This minimal contribution is attributed to good fuel design
which contains most gases until they either decay or the fuel is reprocessed.
Another potential source of krypton-85 in man's environment is that which contaminates natural gas which
comes from gas fields developed with nuclear explosives. It is difficult, however, to project the amount of this
source at the present time although standards would need to be developed for its use if appreciable quantities
of natural gas are developed by this means.
2. Radiation Standards for Krypton-85.
Radiation standards for krypton-85 are currently based on a limitation of 500 millirems per year to the total
body of an individual in the general public or a concentration limit which would maintain the exposure to any
-647-
-------
individual below this annual dose. For population groups the limit is 170 millirems per year (Fed. Reg., 1967).
Because of siting factors and the discharge from tall stacks, all of the sources thus far have been able to meet
this standard for individuals on an annual basis, and still discharge all the krypton-85 produced. The result,
however, of these discharges has been a gradual accumulation of krypton-85 in the earth's atmosphere since it
readily mixes with the atmosphere and distributes over the entire globe. Since the nuclide has a 10.7-year half-
life, its persistence in the environment will be for several decades. This factor and the increasing generation
of krypton by various sources result in long-term low-level exposure of large populations throughout the
entire world.
Because of its rapid dispersion in the earth's atmosphere and the fact that it will expose all individuals in
the world at a low dose rate, neither the current perspective or standard based on annual dose to an individual
is sufficient to describe or control the public health impact of the discharge of krypton-85 to the environment.
The current and proposed expanded development of nuclear energy use which will release long-lived krypton-
85 requires, therefore, the development of a broader environmental assessment that encompasses the entire
radiological impact of this pollutant. This assessment requires a projection of the migration of the
radionuclide through the environment over long periods of time, and a determination of the potential dose to
populations (measured in person-rems), and the associated health effects expected to occur throughout this
migration. All individual exposures, however small, should be included so that all of the impact on society is
assessed, and the exposure of future generations implied by the essentially irreversible environmental
commitments should be recognized. The population dose resulting from such an assessment can be termed an
"environmental dose commitment". In the setting of standards for such radionuclides, it is important to
recognize this long-term radiation risk, in addition to the present one of annual exposures of individuals, and
to implement appropriate controls to minimize it.
The environmental dose commitment for krypton-85, estimated to be produced by the nuclear power
industry through the year 2020, has been determined (EPA, 1973) and the results are shown in terms of
committed health effects in Figure 1. This analysis of the committed health effects is based on the prudent
assumption that the public health impact of radiation dose follows a linear hypothesis down to zero. Although
it is recognized that data are not available to either prove or disprove this assumption, it provides the only
sound basis for developing standards to protect public health. Within this framework, the only totally risk-
free level of radiation exposure is zero; a standard set at any other level must be justified on the basis that the
activity producing the radiation exposure provides sufficient offsetting benefits. This perspective and others
on the risks due to exposures to ionizing radiation were recently analyzed quantitatively by the Committee on
Biological Effects of Ionizing Radiation (BEIR Committee) formed by the National Academy of Sciences'
National Research Council (BEIR, 1972).
The data in Figure 1 represent the sum of effects for the population within 80 kilometers of projected fuel
reprocessing plants in the U. S., the U. S. population, and the world population. Most of the resulting dose and
the effects from krypton-85 are due to the worldwide distribution of this source outside the U. S. The doses to
individuals outside the U. S. are quite small, usually less than 1 millirem/yr; however, it is the cumulative
wide-spread impact on a population this large that results in health effects of this magnitude. The health
effects were calculated utilizing the BEIR Committee data and were derived from the total person-rem
commitment to the world population over a 100-year decay period after a release occurs. The exposure
pathway model was quite straightforward because of the rapid mixing of krypton-85 in the atmosphere; the
most significant error, perhaps, is in the estimate of the projected world population over the 150-year period
considered.
Standards for controlling krypton-85 from nuclear energy sources should consider the availability of
technology, its effectiveness, and its cost in terms of reducing the total population impact of releases of
krypton over the time that it can be estimated to persist in the environment. Since fuel reprocessing plants
represent the major source of krypton-85, it is the first point to consider control. Several processes can be used
to remove krypton-85 from fuel reprocessing plant effluents; however, none of these have been demonstrated
on a commercial scale. Sufficient information exists, however, to project the effectiveness of such technologies
and the cost of their application as shown in Table 1. Since one fuel reprocessing plant of current design can
support 45 to 50 large nuclear power plants, the cost of this control is less than 0.1% of the cost of generating
electric power (EPA, 1973). It appears, therefore, that the impact is sufficient and the cost-effectiveness of
control is good enough to warrant serious consideration of reducing krypton-85 discharges from fuel
reprocessing plants. Reductions of 99% and more appear possible depending, of course, on the performance
characteristics of the equipment once installed. Such a 100-fold reduction in the discharge of krypton-85 by the
year 1985 would substantially reduce the projected health effects committed from this source as shown in
Figure 2. These data indicate quite clearly the effect such control would have on the health effects committed
by this source.
Nuclear power reactors also discharge krypton-85; however, the quantities are generally less than 1,000
curies per year per plant (EPA, 1973). The annual dose to an individual in the public from such discharges is
less than about 1 millirem and most of this is surface dose to the skin, not to viable tissue or to genetically
significant organs. Such discharges result in a 100-year dose commitment of less than 1 person-rem from one
year of operation of one of these power plants. The total health effects committed through the year 2000 from
nuclear power reactors themselves are generally less than approximately 5 effects (EPA, 1973). Since the
technologies to remove krypton-85 at reactors would have to be applied to each individual plant at a cost in the
range of 0.5 to 1 million dollars per plant, the public health protection gained does not appear to justify the cost
that would be involved.
-648-
-------
CONSIDERATIONS FOR RADON
1. Environmental Sources of Radon.
The environmental sources of radon, other than the natural background levels, are directly related to
processes which disturb uranium deposits or which remove it from the ground and prepare it for other uses.
For example when uranium is extracted from ore at a mill, more than 99 percent of the ore material becomes
the mill wastes or tailings, a slurry of sandlike material in waste solutions which are accumulated to form a
tailings pile. More than 97 percent of the radioactive decay products of uranium and about 4 percent of the
uranium from the ore remain in these tailings. The concentration of radium-226 in the tailings averages about
700 pCi/g, indicating an inventory of about 56,000 curies of radium-226 in a representative pile The radon-222
decay product of radium-226 emanates from such a pile at an average of about 500 PCi/m2-s, representing a
total release of radon gas of more than 130,000 Ci/yr. Because of the presence in the tailings of thorium-230
which by its decay maintains the radium inventory, the radioactivity in the tailings will remain almost
constantfor thousands of years. As of 1970, there were more than 80 million metric tons of tailings occupying
more than 2,100 acres of land (EPA, 1973).
The emanation rate of radon from tailings piles is also influenced by whether the pile is wet or dry. If the pile
is saturated with water or under water, the water tends to prevent the escape of the radon gas by a factor of
about 25. This indicates a release rate of about 22 pCi/m2-s for a radium concentration of 560 pCi/g. For
comparison, the natural background release rate is about 1 pCi/m2-s in most parts of the U. S.; it may be one
hundred times greater, or more, over uranium deposits (Pearson, 1967). The radon release rate at any one
location is also known to vary over a factor of 10 due to the effects of weather, i. e., wind speed, barometric
pressure, atmosphere stability, rainfall, and snow cover.
A significant source of radon which has received considerable attention is that in underground uranium
mines. Although this source represents occupational exposures which are currently controlled by standards,
the exhausting of the mines to meet exposure standards represents a concentrated point source of radon and
associated decay products which can expose local populations at varying rates.
Radon also exists in natural gas and at concentrations that may be quite high at the point of use if the travel
time between the well and the point of use is short. A recent study (Levels, 1972) of radon in natural gas
indicates considerable variation in the concentration levels at points of use in the U. S. with an overall
average of about 23 pCi/1.
The use of tailings as fill and construction materials represents a rather special source of exposure to the
public from radon. The presence of tailings under buildings in the Colorado area has prompted remedial
action in several locations because of the public exposures which occur. Most buildings contain materials
which emanate radon. Some of these materials, such as various granites, have quite high radon emanation
rates. Indoor concentrations due to building material emanation have been found to be typically 3 to 4 times
the outdoor levels (Levels. 1972).
The largest source of radon in the environment, of course, is due to the ambient levels produced by the
widespread distribution of uranium and its decay products in the earth's crust. The ambient radon level goes
through a daily cycle of concentrations ranging from 0.03 to 3.50 pCi/1 with the average U.S. level being about
0.3 pCi/1 (Pearson, 1967).
2. Standards for Radon-222.
The standards for radon-222 and its decay products are most often given in terms of a working level (WL).
One WL is the total potential alpha energy from any combination of short-lived radon-daughters (through
RaC') thatimpart 1.3 x 10s MeV per liter of air. This level was first proposed by the U. S. Public Health Service
in 1957 as a level which was thought to be safe, yet not unnecessarily restrictive to industrial operations
(Holaday, et al, 1957). The Federal Radiation Council and the Environmental Protection Agency (Fed. Reg.,
1971 and FRC, 1967) recently recommended that the standard for exposure to miners be 4 working level
months per year (WLM). One WLM is the exposure resulting from inhalation of air containing a radon-
daughter concentration of 1 WL for 170 working hours.
The working level is often related to radon activity by calculating the number of radon-daughter
disintegrations required to impart 1.3 x 105 MeV of alpha energy. A concentration of 100 pCi/1 radon-222 in
secular equilibrium with daughter products RaA, RaB, RaC (RaC') will produce 1 WL. The working level
definition is often misunderstood as a unit of radon concentration when, in fact, it is a concentration of only
the short-lived daughters RaA, RaB, RaC, (RaC'). It can be applied to any mixture of these decay products. The
conversion of 1 WL per 100 pCi/1 radon-222 applies only for secular equilibrium of radon and daughters. The
guides for control of radon have been primarily oriented towards health protection of uranium miners. These
guides have undergone several changes over the years as the potential for lung carcinomas from radon-
daughters became better understood. Guides for continuous exposure of the general public vary from 0i.dto d
pCi/1. The lowest recommendation of 0.3 pCi/1 is derived from ICRP No. 9, par. 72,1966, (ICRP, 1966 I. This
was derived as 1/30 of the radon-222 guide for continuous occupation exposure. The upper level of 3 pOi/1 is
derived from 10 CFR20. *.-,,• u iU TT c
Another instance of protection recommendations for the public is the issuance of guidelines by the U. b.
Surgeon General regarding levels at which remedial action should be undertaken to remove tailings from
under buildings (JCAE, 1971). The guidelines are directly related to measured working levels inside the
structures. At levels below 0.01 WL no action is recommended; at levels above 0.05 WL action should be
-649-
-------
considered. In the range between these values it is recommended that all relevant factors be weighted in a
decision as to whether to remove the tailings.
3. Public Health Impact.
The public health impact of radon is difficult to determine exactly because of a number of factors. Pure
radon-222 is not nearly as hazardous as radon and its daughters together. Being a noble gas radon does not
remain in the lungs and, in addition, radon is not in intimate contact with the tissue. In the event of a
disintegration of radon, the alpha energy is likely to be expended into a non-critical area of the lung. Radon-
daughters, in particular polonium-218 and polonium-214, which are not noble gas elements, have chemical
and physical properties that cause them to be absorbed into the bronchial epithelium of the lung, leading to
high dose rates to the region of the lung where tumors arise. Radiation dose caused by radon is therefore a
function of the state of equilibrium between radon and its daughters at the time of exposure; it is not enough
simply to know the radon-222 concentration alone. Furthermore, the dose rate in air with relatively little
particulate matter will be higher from radon-daughters than for rather dirty air because more daughters will
be present as ions rather than adsorbed onto particles. In western states where tailing piles are located, and
where air is presumably lowest in aerosol concentration of radon-222, the dose rate should be higher than in
more industrial areas. Air conditioning may serve the same purpose by removing significant numbers of
particles from the air.
The critical biological target in the lung is assumed to be the nuclei of the basal cells of the bronchial
epithelium. Alpha particles have a limited range in tissue. To reach the basal cells they must penetrate first a
mucus layer, then certain other cells. The bronchial epithelium shows extreme variations in thickness
because it tends to fold upon itself. In addition, there is to be expected natural variation in thickness of both
the mucus layer and the intervening cells if a population of all age groups is considered. It is therefore a matter
of some controversy to determine exactly the depth the alpha particle must penetrate to deliver a critical
exposure. Given these difficulties, it should be clear that radon dosimetry does not give a simple correlation
between the radon-222 concentration in air and the radiation dose delivered to the lung. Inclusion of all these
factors yields a range of conversion factors relating air concentration and lung dose. The factor that appears
most suitable for general environmental exposure conditions appears to be that a steady concentration of 1
pCi/m3 of radon in air will produce 4 millirem to the lung.
Uranium tailings piles represent an unusual radiation exposure situation for the public. Even though such
piles are currently stabilized after they are filled by covering them with soil and plantings of some sort, they
can be expected to emanate radon at a rate close to that from a dry pile. Since the piles contain radium with a
half-life of 1620 years, their emanation will continue for thousands of years. The radium will also be
replenished by the decay of thorium-230. Such piles represent, therefore, an irreversible source of
environmental contamination to the atmosphere of local areas by radon-222 and its decay products. Data
from two representative piles are shown in Table 2 which indicate that the lung doses of individuals near the
piles can be quite substantial. The health effects committed to surrounding populations can be expected to be
significant if estimated for long periods of time.
The only control method currently used for tailings piles is to stabilize them; however, this procedure only
affects erosion of other radioactive materials. Perhaps the only controls that would appreciably affect the
amount of radon and daughters entering the environment over the long-term from the use of uranium in
nuclear energy programs would be to return the tailings deep underground to contain radon emanation or
remove radium (the radon precursor) in the uranium mill. Neither the cost nor the effectiveness of either of
these methods is known at the present time. In view of the radiation exposures over the long-term, however, it
appears prudent to further consider methods of reducing the discharge of radon from uranium mill wastes.
SUMMARY
Both krypton-85 and radon-222 and its resultant daughters represent long-term health effect commitments
as a result of current practices in the use of uranium and its by-products as an energy resource. New standards
and concepts which recognize and control long-term environmental dose commitments are needed for each of
these radionuclides. Effective controls can be required at reasonable costs for krypton-85 at fuel reprocessing
plants, the major source of environmental levels of krypton-85 in the future. No such controls have been
developed, however, for the discharge of radon from mine shafts or tailings piles, although future
consideration of methods to limit such discharges to the environment appear warranted. Natural gas
contaminated by krypton-85 is difficult to control, but as yet no new standards appear to be required. Radon-
222 and its daughters in natural gas should receive further emphasis with regard to controlling the levels that
exist at the point of use.
REFERENCES
Effects on Populations of Exposure to Low Levels of Ionizing Radiation, The, (1972), Report of the Advisory
Committee on the Biological Effects of Ionizing Radiations (BEIR), National Academy of Sciences,
Washington, D.C.
-650-
-------
47.
&£^^
Guidance for Federal Agencies' <1967>«
in ^--^^(1967), Report No. 8 revised, Staff
Holaday D.A., D.E. Rushing, R D. Coleman, P.P. Woolrich, H.L. Kusnetz, and W.F. Dale, (1957),
. , .. , .. ez, an .. ae, ,
Publkation N?S S "* Uranmm Mines and Calculations on Biological Effects, Public Health
International Commission on Radiological Protection, (1966), Recommendations of the
International Commission on Radiological Protection, Adopted September 17, 1965, ICRP Publication 9
Pergamon Press, London. '
io^FAEK?earing8' (1971)' Use of Uranium Mill Tailings for Construction Purposes, October 28 and 29,
1971, p. 52.
Pearson, J. E., (1967), Natural Environmental Radioactivity for Radon-222, Environmental Health
Series, Public Health Service Public No. 999-RH-26.
A Report of the United Nations Scientific Committee on the Effects of Atomic Radiation for the
General Assembly, (1972), Ionizing Radiation: Levels and Effects, Vol. 1: Levels, United Nations New
'
York.
TABLE 1. Summary of Health Effects and Costs of Krypton-85
Controls for a 5 MT/Yr Fuel Reprocessing Plant.
Removal System
Cryogenic
Distillation
Cryogenic
Absorption
Cost
$11.5 million
$12.1 million
Health Effects
Averted
270V
270V
Freon Adsorption $ 6.4 million 270V
-W* 60% mortality expected from induced cancers and other serious effects.
TABLE 2. Public Health Impact of Dry Tailings Piles.
GRAND JUNCTION SALT LAKE CITY
Ra-226 (Ci)
Ra-222 (Ci/yr)
Size (acres)
Doses (mrem/yr)*
Background
Avg. Individual
Max. Individual
At 1 km
1.700
5,600
55
2400
800
3200
320
1670
13,500
107
620
380
1140
760
*Dose conversion factor of 1 pCi/m3 radon-222 =4 mrem/yr.
-651-
-------
8000
7000 --
6000
5000
H
H
O
-------
200 -
180
160 -
140
120 -
100 -
80
60 -
40
20 -
1970
1980
1990
2000
2010
2020
Figure 2. Estimated Past and Future Health Effects Committed by Krypton-85 Released from the U. S. Fuel
Reprocessing Industry, with 99% Removal.
-653-
-------
CONSIDERATIONS IN SITING LONG-TERM RADIOACTIVE NOBLE GAS STORAGE
FACILITIES*
J. J.Cohen and K.R. Peterson
Lawrence Livermore Laboratory
University of California
Livermore, California
Abstract
Projections of future fission reactor capacity indicate that the present practice of releasing radioactive
noble gases (primarily krypton-85) to the atmosphere cannot safely be allowed to continue indefinitely.
Ongoing research programs are developing methods of extracting and concentrating these waste gases so
they may be stored for extended periods, thereby allowing them to decay to relatively safe levels.
Anticipating the need for facilities specifically designed for long-term storage of noble gases, we have
reviewed some factors which might be considered in developing optimum siting criteria. Among these factors
are, above- vs. below-ground storage, meteorology, geography, and demography. The objective used in these
considerations is minimization of resulting population doses, both long- and short-term, from a major release,
either "instantaneous" or "continuous," from the various facilities.
Those sites considered are remote islands, continental shorelines, mountain tops, deep water, and inland
low-population-density areas.
Assuming an equal probability of containment failure at each of the site options, we conclude that siting is
relatively inconsequential for long-term population doses. Siting, however, is important for short-term
exposures to local populations, especially where there might be a possibility of exceeding Radiation
Protection Guides. The relative significance of long- vs. short-term doses resulting from exposures to
radioactive noble gases is also discussed.
INTRODUCTION
Of those noble gases produced in nuclear reactor operations, only 85Kr is considered to have significant long-
term health effects because of its chemical stability, long half-life, high production rate, and its relative
hazard. Figure 1 gives the projected production rate and total accumulation for 85Kr through the remainder of
this century (ORNL, 1970).
In normal operations, only minute quantities of 86Kr are released at the reactor itself. Almost all of the 85Kr
is released at nuclear fuel reprocessing plants during the chopping and dissolving operations. At present,
essentially all of this gas is released to the atmosphere. Figure 2 shows the projected individual dose rates for
both skin and whole body resulting from exposure to accumulated 85Kr, assuming present practices of
atmospheric release are continued. At current levels of reactor operation, this practice is considered to be
acceptable; however, based upon projections for future reactor usage and resultant exposures, it appears to be
the general consensus that the practice cannot be allowed to continue indefinitely. Figure 3 shows the
projected worldwide population dose based on data from the previous figure plus predicted world population
growth (UN, 1966). The point at which measures to prevent the release will become necessary has not as yet
been defined, but it appears reasonable to assume that sooner or later the routine release of 85Kr to the
atmosphere from fuel reprocessing plants will be prohibited.
COST-BENEFIT ANALYSIS
The application of cost-benefit analysis can be used to gain some insight into the advisability of instituting
preventive measures. Table 1 gives some data on the effects of 85Kr atmospheric releases. These data have
been derived by assuming the following:
(1) Concentration-to-dose conversion factors have been calculated from data given by Dunster, et al.,
(1970). These are 1.4 x 104 and 2.1 x 106 rem/yr per Ci/m3 for whole body and skin, respectively.
(2) The mass of the atmosphere is taken to be 5 x 1021 gm, which at conditions of standard temperature and
pressure occupies a volume of 4.2 x 1018 m3 (USGS, 1967).
(3) For this cost-benefit analysis we assume that the released 85Kr mixes instantaneously with the entire
atmosphere to a uniform concentration. The significance of atmospheric diffusion of released gas to the point
of uniform mixing will be discussed later in this paper.
(4) The only process by which atmospheric levels of 85Kr are diminished is by radioactive decay.
Therefore, the mean atmospheric residence time for 86Kr is essentially its mean radiological life,
approximately 15.5 yrs. This implies that there is no sink in nature that removes 85Kr and renders it
biologically unavailable.
It can be determined from these data that the total whole body dose resulting from each curie of 85Kr released
to the atmosphere will average .. 3
1.4 x i(T I$EiE_
Tg^ x 15.5 yri= 5.2 x io"14 rem/Ci.
4.2 x 10"" m
* Work per formed under the auspices of the U. S. Atomic Energy Commission.
-654-
-------
Determination of total population dose is based on an assumed uniform exposure to the entire world
population taken as 3.5 billion people. Therefore, the total population dose integrated to infinite time is 1 8 xlO4
man-rem per curie for whole body and 0.027 for skin.
The monetary cost for biological damage due to radiation exposure has been estimated by various authors
to be somewhere in the low hundreds of dollars per man-rem (Cohen, 1973). For purposes of this report we have
assumed a cost of $250 per man-rem for whole body exposure. For skin exposure this value is estimated at
iffiSSf^W40 Pfr ,man-rem'based on the ratio (factor of 6) between the ICRP maximum permissible dose
OMPD) criteria fof whole body and skin 1966)-This implies that roughly equivalent damage is done to
an individual by his exposure to MPD levels of either 0.5 rem/yr to the whole body or 3.0 rem/yr to the skin.
Using these data, one may estimate the cost for biological damage due to 85Kr based on either whole body or
skin exposure to be roughly $0.05 or $1.00 per curie, respectively. Essentially this implies that skin may be
considered to be the critical organ for 85Kr exposure. For each curie released to the atmosphere, the resultant
biological damage cost to all humanity over all time is approximately one dollar. Expenditures for prevention
of 85Kr release in excess of this figure would therefore not be considered to be cost effective.
RELEASE PREVENTION COSTS
Having estimated the effect of 85Kr release, we shall next estimate the cost of release prevention. Continued
research has been performed on methods for 85Kr removal and storage. Other papers given in this symposium
deal extensively with this topic. Estimates on 85Kr treatment costs have been provided by Slansky, etal.,
(1969), Davis (1973), and Davis, etal., (1973). From these data we can conservatively assume that for a 1,500
ton/yr fuel reprocessing plant, 85Kr treatment facilities can be installed at a capital cost of $2,000,000 and
operated at a cost of $200/ton of fuel processed. Further, assuming a plant life of 20 yrs, an interest rate of 12%,
and a content of 104 Ci of 85Kr per ton of spent fuel, we calculate a prorated capital cost of approximately 2
-------
In determining safety criteria for siting the storage facility, we have assumed that in normal operation
there would be no routine releases of radioactive gas. The only assumed mechanism for release would be from
accidental or inadvertent containment failure.
With the possible exception of sabotage, we do not consider the simultaneous leakage of more than one
container to be credible. If massive multicontainer leakage were a major concern, then the entire concept of
centralized storage would not be prudent, since this would constitute a policy of "putting all one's eggs in one
basket." For purposes of radiation safety, therefore, the siting criteria that we have selected assume the
leakage of 1 MCiof 85Kr over a relatively short period of time. The siting objectives are twofold: first, to assure
that in the event of such leakage, no individual at or near the fence line receives a dose in excess of MFC
standards; and, second, to minimize resultant total population doses.
MAXIMUM ACCIDENT DOSES
To evaluate fence line dose criteria, we have calculated the total leakage required to exceed maximum
permissible dose criteria for both accident and routine conditions, conservatively assuming a dilution factor
(X/Q) of 10-6 sec/m3 between the point of release and the fence line. The results of these calculations are given
in Table 3. From these data we conclude, that even for routine release conditions, one may release as much as
45 MCi without exceeding the 3 rem skin dose at the fence line. Therefore, our assumed maximum container
capacity of 1 MCi appears to be well within these limits.
POPULATION DOSE CONSIDERATIONS
Our second radiation safety criterion is that in the event of leakage, total population dose (man-rem) will be
minimized. With this objective in mind, geographic siting becomes critical. A number of factors, all of which
are to some extent interrelated, determine the type of storage site that will minimize population dose in the
event of an accidental release. These are:
1. Meteorology.
Meteorology — or, more precisely, the diffusion climatology of the area surrounding the storage site —
determines the rate of dilution and the resultant effluent concentrations as a function of time and distance
from the release site. Rapid diffusion accompanied by lower individual doses is favored by those locations
experiencing a combination of the following factors:
(a) Strong surface winds, relatively variable in direction.
(b) Strong daytime heating resulting in neutral or unstable temperature lapse rates.
(c) A relative absence of night-time temperature inversions.
(d) A site that is elevated with respect to surrounding locations.
2. Geography.
Geographic factors that would tend to affect potential population dose are:
(a) The distribution of population with respect to the storage site.
(b) Topographical features that would affect the rate of dilution of effluents prior to their arrival at
population centers.
POTENTIAL SITES
From these considerations we have chosen to consider five hypothetical storage sites within the jurisdiction
of the United States. These sites are evaluated on the assumption that a 1 MCi release of 85Kr occurs over a
period of a few days or less. It is further assumed that the release occurs at a time when the surface wind
direction is the most probable one for the particular site. The hypothetical sites selected for evaluation are:
1. A Remote Island Site.
Because of its relatively low population and distance from any major population center, Johnston Island,
southwest of Hawaii, was chosen as a hypothetical site. The Marshall Islands, over 1000 km distant, are the
most likely point of first exposure, since surface winds in this area blow consistently towards the west. In our
analysis we assume that before reaching the Marshall Islands, the plume of 85Kr would loop into the middle-
latitude westerlies and then move toward the east.
2. A Coastal Site.
Cape Hatteras, North Carolina has been selected as a hypothetical site since the prevailing winds at this
location are usually offshore, but are not as directionally persistent as those at Johnston Island. Although our
primary evaluation is based on most probable meteorological conditions, we have in this case also performed a
calculation assuming onshore winds.
3. A Remote Desert Inland Site.
The Fort Irwin Military Reservation in the Mojave Desert of California was selected as a hypothetical site in
this category. There are no population centers within 150 km in all directions. The average winds are toward
Nevada; a "worst case" would occur with winds toward Los Angeles. Dose estimates are presented for both
cases.
-656-
-------
4. A Mountain Top Site.
Mt. Whitney, California was selected as a hypothetical site in this category. We have assumed that an all-
weather road could be built to the summit and that the storage facility could be sited there at an elevation of
approximately 4,400 m. The advantage of an elevated site, of course, is that it would allow for a maximum
turbulent mixing of an effluent plume prior to its arrival at any population center. Average and "worst case"
winds are similar to those for the Mojave Desert, due to their proximity. Both cases have been included in the
dose calculations.
5. A Metropolitan Site.
Although it is not considered credible that the storage facility would actually be located within a major
population center, we have evaluated such a site to determine an upper limit to man-rem estimates. For this
purpose, an analysis was performed for a facility located in downtown Chicago, Illinois.
CALCULATIONS
The hypothetical sites and their prevailing wind directions are shown in Figure 5. The calculational model
used to determine population dose was developed by Knox, etal., (1972). The model has been adapted for use in
this study according to the following assumptions:
1. The release occurs at an elevation of 10 m above ground.
2. The effluent stream is eventually dispersed over the entire troposphere of the Northern Hemisphere with a
depth of about 15 km. (The model used is applicable only to a hemisphere.) Since the Northern Hemisphere
contains about 90% of the world population, the calculated population doses will therefore be conservatively
large. Population statistics were obtained from 1970 census figures.
Calculation of doses within 1,000 km followed the method of Knox (1971) in which functional relationships
with distance were developed, except that winds were sector-averaged over 22.5 degrees. The close-in equation
for man-rem, Dp, is
TT/8
f f
D = / / Do R(r) prdrdG,
i JO JT
where
D0 = individual dose at r0 (1 km), rem,
R(r) = ratio of dilution factors at r to ro, expressed as functional ratios [see Knox (1971) for full discussion],
p=population density, people/km,
r=distance downwind, km,
r0=distance to site boundary, km, and
0 = azimuthal angle, radians.
Values of R(r) and population density used in this analysis are shown in Table 4.
Beyond 1,000 km, Knox, et al, (1972) developed an intermediate-range solution. It is assumed that the 85Kr
effluent is transported at 40 km/hr, spreading throughout the troposphere horizontally at a rate of 220
km/day (half-width) for the first few days and then at a rate proportional to t1/2. If Do is the site boundary
dose, then the total dose is
D. = D ^ 1 + e + e
= D. Ui (15.5),
-657-
-------
where
is the first-year individual dose,
Table 5 provides summary values of population doses to 1,000 km and for three continents. Long-range
Northern Hemisphere population doses out to infinite time are also given. These calculations assume a
uniform mass concentration of 85Kr in the atmosphere and a mean residence time of 15.5 yrs.
It is apparent that the worldwide population dose is rather insensitive to storage site. The largest Northern
Hemisphere dose (for a Chicago location) is about 1-1/3 times the smallest dose (for a remote island site). The
important siting criterion appears to be the avoidance of a location where a large number of people residing
nearby might receive large individual doses.
SUMMARY AND CONCLUSIONS
1. Cost-benefit analysis indicates that it would be prudent policy to require the prevention of 85Kr release
from fuel reprocessing plants at the present time, assuming this can be accomplished at a cost amounting to
less than $1.00/Ci.
2. We have discussed options for accomplishment of 85Kr release prevention from fuel reprocessing plants.
No value judgments have been attempted in evaluating these options. However, it has been assumed that a
policy of concentrating effluent noble gases, retaining them in pressurized storage tanks, and storing them for
long periods at some centralized facility will be adopted. Such a policy would appear to be consistent with
current AEC policy on high-level waste management.
3. Criteria for siting a long-term noble gas storage facility should include assurance that in the event of a
containment failure: (a) Maximum permissible dose guidelines (0.5 rem/yr for whole body and 3.0 rem/yr for
skin) are not exceeded, (b) Resultant population doses (man-rem) are minimized.
4. Five hypothetical sites have been evaluated to estimate population doses in the event of leakage. From
this analysis it appears that geographic siting may be considered relatively unimportant.
5. Site selection should be based on cost-benefit studies considering: (a) Transportation and handling costs.
(b) Maintenance and surveillance costs, (c) Resultant health benefits derived in terms of potential population
dose averted.
REFERENCES
Cohen, J. J., (1973), On Determining the Cost of Radiation Exposure to Populations for Purposes of Cost-
Benefit Analyses, Health Physics, vol. 25, p. 527.
Davis, J. S., (1973), personal communication.
Davis, J. S., and J. R. Martin, (1973), A Cryogenic Approach to Fuel Reprocessing Gaseous Radwaste
Treatment. This Symposium.
Dunster, H. J. and B. F. Warner, (1970), The Disposal of Noble Gas Fission Products From the
Reprocessing of Nuclear Fuel, AHSB(RP) R101, United Kingdom Atomic Energy Authority (UKAEA),
Harwell.
Federal Regulation No. 222, (1970), Policy Relating to the Siting of Fuel Reprocessing Plants and Related
Waste Management Facilities, vol. 35.
Hull, A. P., (1972), Reactor Effluents: As Low as Practicable or as Low as Reasonable, Nuclear News, vol.
16(11), p. 53.
International Commission on Radiological Protection (ICRP), (1966), Recommendations of the
International Commission on Radiological Protection, Publication No. 9, Pergamon Press.
Knox, J. B., (1971), Airborne Radiation from the Nuclear Power Industry, Nuclear News, vol. 14(2).
Knox, J. B. and K. R. Peterson, (1972), Estimates of Dose to Northern Hemisphere Population Groups
from 8eKr Emitted by a Single Nuclear Fuel-Reprocessing Plant, Nuclear Safety, vol. 13(2).
Oak Ridge National Laboratory (ORNL), (1970), Siting of Fuel Reprocessing Plants and Waste
Management Facilities, Rept. ORNL-4451.
Slansky, C., H. K. Peterson, and V. G. Johnson, (1969), Nuclear Power Growth Spurs Interest in Fuel
Plant Wastes, Environ. Sci. Technol., vol. 3, p. 446.
United Nations (U.N.), (1966), Demographic Yearbook.
U. S. Geological Survey (USGS),(1967),Z>ata on Geochemistry: Composition of the Earth's Crust, USGS-
440-D, sixth edition, Washington, B.C.
-658-
-------
TABLE 1. «sKr Exposure Data.
Concentration — dose conversion
factor
Total individual dose due to
atmospheric 85Kr
Total population dose due to
atmospheric 85Kr (world
population = 3.5 x 109)
Maximum permissible dose
Cost of biological damage
Estimated biological damage cost
per unit release
Whole body Skin Units
1.4 xlO4 2.1 xlO6 rem-m3
Ci-yr
5.2 xlO-14 7.8 xlO-12 rem/Ci
1.8x10-" 0.027 man-rem/Ci
0.5 3.0 rem/yr
250. 40. $/man-rem
0.05 1.00 $/Ci
Assumptions:
1. Instantaneous uniform atmospheric mixing
2. AtmVol = 4.2 x 1018 m3
3. World population (1970) = 3.5 x 109
4. Mean atm residence time (85Kr) = 15.5 yr
TABLE 2. Cost-Benefit Analysis.
1. Estimated Cost for 85Kr extraction and storage:
$0.04/Ci
2. Estimated Benefit due to 85Kr release averted:
$1.00/Ci
3. Benefit/cost ratio-
1.00 = 25
0.04
4. Cost effectiveness for exposure prevention:
For whole body:
$0.04/Ci x 1 Ci/man-rem'='$220/man-rem averted
1.8 xlO-4
For skin:
$0.04/Cix
. 1 Ci/man-rem^$1.50/man-rem averted
0.027
-659-
-------
TABLE 3. 8SKr Accidental Release.^
Maximum allowable container
Maximum allowable capacity (106Ci)- assume
fence line dose (rem) fence line-^ = 10-6 sec/m3
25 (whole body) 56,000.
0.5 (whole body) 1,100.
150 (skin) 2,250.
3 (skin) 45.
(a)
Assume the entire contents of one container is released.
TABLE 4. Population Densities and Functional Relationships R(r) for
Calculating Population Doses to 1 ,000 km.
1970 population density (people/km2)
Distance
(km)
1-3
1-3
3-10 (S
Island
Coast
Most Most
R(r) probable probable
1 0
1 0
!r0/r)* 0
10-20 (3r0/r)2 0
20-100 0.
100-1000 0.
4r0/r 0
4r0/r 0
0
0
0
0
0
0
Desert
Worst Most
case probable
7
7
7
7
20
70
5
5
5
10
15
50
Mountain
Worst Most
case probable
5
5
5
10
25
200
0
0
0
15
40
City
Worst Most
case probable
0
0
0
25
200
2,500
2,500
2,200
800
70
TABLE 5. Population Whole Body Doses for a 1 MCi Release of 85Kr.
Population whole body dose (man-rem)
Distance or
Continent
1 to 1000 km
North America
Europe
Asia
Northern
Hemisphere
total
Johnston Is.
Most
probable
0
40
92
460
590
Cape Hatteras
Most Worst
probable case
0
40
92
460
590
45
40
92
460
637
Mojave Desert
Most Worst
probable case
31
40
92
460
623
120
40
92
460
712
Mt. Whitney
Most Worst
probable case
26
40
92
460
618
120
40
92
460
712
Chicago
Most
probable
210
40
92
460
802
-660-
-------
10
10
10C
10'
-Total accumulated (Ci)
970
Production rate
(Ci/yr)
1980 1990
Year
2000
Figure 1. Projected worldwide 85Kr production. Data from ORNL (1970).
661 -
-------
-o
(O
S-
Ol
-M
rO
s_
QJ
-------
1980 1990 2000
Year
Figure 3. Projected worldwide population dose due to atmospheric 85Kr.
-663-
-------
Release to
atmosphere
_L
Store entire off-gas
volume
Extract noble gas in
concentrated form
Store underground
n
Store at
surface facility
Store at atmospheric
pressure
Store at each fuel
reprocessing plant
Store in pressure
vessels
_L
Transport to and
store at
central facility
Figure 4.85Kr treatment options.
-664-
-------
Johnston Island
Cape
Hatteras
Figure 5. Potential storage sites (arrows indicate prevailing wind directions).
-665-
-------
SELF ABSORPTION AND GEOMETRIC CORRECTION FACTORS FOR REACTOR OFF-GAS
SAMPLES RELATIVE TO NBS STANDARDS*
R. F. Coley1,2 and N. A. Frigerio2
Commonwealth Edison Company
Chicago, Illinois
2Argonne National Laboratory
Argonne, Illinois
Abstract
Although they can be counted in identical bottles using identical counting systems, real gas samples differ
from the NBS solution standards (e. g., mock reactor off-gas) in two respects — geometry and self absorption.
Because both detector and source are real and finite, the simple "narrow beam" linear attenuation coefficient
approximations currently used in the industry are quite inadequate for correction. Accordingly, the well-
tested, complete-analog program, BIM130, was used to compute the fraction of photons, and the photon
energy spectra, reaching typical detectors used in the industry. Using this method, it was possible to correct
the given NBS standard activity to its effective activity relative to a gas sample in an identical bottle. Factors
were much closer to unity than predictions based on "narrow beam" linear attenuation coefficient
approximations. At 80 keV, for example, such approximations gave 0.76, whereas the factor proved to be 1.03
fora3"x3"NaI(Tl) crystalanda 3 cm distance. Results are presented for various gamma energies of interest
from.80 keVto 1,830 keV, and for the commonly used industrial distances of 3,10, and 30 cm from the bottom of
the sample bottle to the top of the detector container. Complete spectra for photons entering the detectors, as
well as factors derived from these for typical Nal(Tl) and Ge(Li) detector resolutions, are given.
INTRODUCTION
A problem familiar to those associated with the nuclear power industry is that of the quantitative
determination of the amounts of noble fission gas nuclides in reactor off-gas samples. Typically, reactor off-
gas grab samples are collected in 15 ml serum vials for subsequent measurements using a gamma detector and
multichannel analyzer system. Because of the wide ranges in both activity and isotopic composition
experienced for such samples, the power plant analyst requires efficiency calibration data for various source-
to-detector distances — 3,10, and 30 cm being typical.
DISCUSSION
Recently, the National Bureau of Standards, with the support of the Atomic Energy Commission (ICONS,
1972), issued relatively inexpensive, mixed-radionuclide, gamma standards in 15 ml serum vials.
Experimentally, one can use a multichannel analyzer system to compare the count rates from two bottles, one
of known activity (the NBS standard), the other containing an off-gas samples. The absolute gamma
activities in the off-gas sample could be simply obtained from the count rate ratio were it not for two factors: (1)
the gas and the liquid are not identical in geometry; e. g., the extension upwards is higher for the gas than for
the liquid and (2) self absorption and scatter are also greater within the liquid than within the gas. Because
both NBS standards and samples can be counted in identical bottles using identical counting systems, a
general set of geometry-self absorption correction factors for such samples might be a welcome step towards
the long needed standardization of industrial techniques in this area.
The assumption that the effects of the glass between the photons leaving the source medium and the
detector is the same for two identical bottles is not quite true. Measurements performed at the National Bureau
of Standards have indicated that up to 25% of a group of such standards fell outside of 1.5% of the mean
activity value, when measured one inch over a 3" x 3" Nal(Tl) detector (Coursey, 1973). However, such small
errors are not usually of concern within the nuclear power industry.
The essential question is: For a given number of source photons, how many leave the source in such a
direction as to be counted within the photopeak, for liquid and for gas? The number of photons leaving the
liquid source per unit time may be expressed as: NI = kj Aj, where ki is the fraction actually emerging from the
liquid and AI is the known emission rate. Similarly, for the gas source: Ng = kgAg. In order the compute kj and
kg, the complete-analog program BIM 130, which has been shown to give excellent agreement with
experimental values in both simple and complex systems (Frigerio, et al, 1969 and 1973a and b), was used. In
addition, the program was utilized to compute the photon energy spectra reaching the two most common types
of detectors in use in the industry, NaI(Tl) and Ge(Li). Table 1 presents the spectrum reaching the detector for
each of several photon energies emitted from the 12.4 ml right circular cylinder liquid source of radius 1.14 cm
and a height of 3.037 cm representing the NBS serum vial source. Interactions within the 14.4 ml volume
representing an off-gas sample, in the same serum vial, proved to be negligible (for 106 source photons there
were only 5 photoelectric and 94 compton events within the source medium). Thus, the spectra of Table 1
represent the net spectral differences between liquid and vapor phase samples counted in the same vial using
the same counting systems.
The fraction of source photons emergent from the liquid and from the gas samples, and striking the detector,
were computed for a 3" x 3" NaI(Tl) crystal and for a 46.2 ml true coaxial, right circular cylinder of Ge(Li) with
a diameter of 42.7 mm. These fractions are given in Table 2 for distances of 3,10, and 30 cm from the bottoiii of
*Work supported, in part, by the U. S. Atomic Energy Commission.
- 666 -
-------
the vial to the top of the detector can. They represent the fraction of photons which would be counted within
the photopeak for summation from the low energy channel (valley) indicated. The lower energy bounds for
these summations were chosen to agree with actual experimental data obtained at operating nuclear units.
The ratios of the liquid fractions given, to the corresponding gas fractions, yield the desired correction
factors, k]/kg. These are given in Table 3.
Thus, the activity of an unknown gas source, A™, is given by:
Ag = (k1/kg)(Ng/N1)A1
where Aj is the liquid source activity, as given by NBS, (Ng/Ni) is the observed ratio of count rates, and
(ki/kg) is the correction factor as given in Table 3.
Several features of Table 3 may be noted. First of all, the correction factors are greatly different than those
from approximations obtained using"narrow beam" linear attenuation coefficients. At 80 keV, for example,
such an approximation gives a single correction factor of 1.32 for an average photon path length in water
equal to one-half the source height. That is, with u equal to 0.184 cm-1 for 80 keV and t equal to 1.52 cm in water,
eut equals 1.32 (Hubbell, 1969). This may be contrasted with the corresponding values for a Nal(Tl) detector in
Table 3,1.03 to 0.97 as a function of distance to the detector. Such disagreement is a consequence of the fact
that many of the "narrow beam attenuated" events are associated with scatter at small angles, with resultant
small changes in energy (see spectra of Table 1). Thus, for real sources and finite detectors, a significant
fraction of these scattered photons reach the detector, and remain within the photopeak. This fraction
decreases with increasing energy so that, for example, (ki/kg) at 3 cm from a 3" x 3" NaI(Tl) detector drops
from 1.03 at 80 keV to a minimum of 0.96 at 400 keV. This may be called the "photopeak effect".
At the same time, the total cross section also decreases with increasing energy, rendering the liquid more
and more transparent to emitted photons. As this "transparency effect" becomes more pronounced (ki/kg)
rises toward an asymptote of approximately one. Note also that the centroid of a liquid source is closer to the
detector than the centroid of a gas source in the same vial. This results in (ka/kg) values which can, and
should be greater than one either at low energies (photopeak effect dominating) or at high energies
(transparency effect dominating). The much higher resolution of the Ge(Li) detector mitigates the photopeak
effect very noticeably, so that for a Ge(Li) detector (kj/kg) rises monotonically toward unity without an
evident minimum.
From the standpoint of the industrial user, the correction factors themselves are so close to one another that
correction for specific energies is probably not warranted. Because off-gas samples are almost invariably
measured at less than 550 keV, it is recommended that the 250 keV (ki/kg) values be used as adequate average
correction factors for plant off-gas measurements.
REFERENCES
Coursey, B. M., (1973), private communication.
Frigerio, N. A. and M. H. Branson, (1969), Current Research in Depth Dose Computations, Proceedings
of the Symposium on Neutrons in Radiobiology, Oak Ridge, Tennessee, 11-14 November, 1969, USAEC Report
CONF-691106,p.95.
Frigerio, N. A., R. F. Coley, and M. H. Branson, (1973), Depth Dose Determinations. //_ A Monte Carlo
Program and a Standard Man Phantom for Neutron and Gamma Computations, Phys. Med. Biol., vol. 18, no.
l,p.53.
Frigerio, N. A. and R. F Coley, (1973), Depth Dose Determinations. III. Standard Man Phantom and
Various Gamma Sources, Phys. Med. Biol., vol. 18, no. 2, p. 187.
Hubbel, J. H., (1969), Photon Cross Sections, Attenuation Coefficients, and Energy Absorption
Coefficients From WkeVto 100 GeV, NSRDS-NBS 29.
Information Center On Nuclear Standards (ICONS), (1972), New Standards for Environmental
Monitoring of Radioactivity, Nuclear Standards News, October, 1972, (American Nuclear Society).
-667-
-------
TABLE 1. Spectra Inbound to Detectors for Various Monoenergetic Sources (a).
keV (b) counts keV
80 83,537
75 4,449
70 3,248
65 3,124
60 3,790
55 593
50 464
45 103
40 30
35 5
30 2
<30 0
sum(C) 99,345
keV
660
645
630
615
600
585
570
555
540
525
510
495
480
465
450
435
420
<405
sum
150
145
140
135
130
125
120
115
110
105
100
95
90
85
80
75
70
65
<60
sum
counts
91,608
251
278
238
246
237
233
239
221
242
237
237
227
195
231
254
218
4,379
100,000
counts
96,114
1,520
1,406
1,180
1,051
978
977
848
893
1,075
1,210
1,476
425
202
194
178
111
33
33
99,904
keV
190
185
180
175
170
165
160
155
150
145
140
135
130
125
120
115
110
105
<100
sum
keV counts
840
825
810
795
780
765
750
735
720
705
690
675
660
645
630
615
600
<585
sum-
92,418
170
163
185
137
145
165
143
188
150
161
142
168
138
136
143
161
4,952
99,998
counts
86,819
1,016
938
862
802
723
699
686
614
606
561
572
591
703
796
864
1,008
389
689
99,938
keV
1,335
1,320
1,305
1,290
1,275
1,260
1,245
1,230
1,215
1,200
1,185
1,170
1,155
1,140
1,125
1,110
1,095
<1,080
sum
keV
250
245
240
235
230
225
220
215
210
205
200
195
190
185
180
175
170
165
<160
sum
counts
93,912
73
73
64
73
73
58
60
87
65
61
63
57
70
62
66
70
4,936
100,000
counts
88,076
565
566
563
490
460
496
473
473
391
424
371
392
365
319
333
351
365
4,503
99,976
keV
1,590
1,575
1,560
1,545
1,530
1,515
1,500
1,485
1,470
1,455
1,440
1,425
1,410
1,395
1,380
1,365
1,350
<1,335
sum
keV
400
395
390
385
380
375
370
365
360
355
350
345
340
335
330
325
320
315
<310
sum
counts
94,438
43
60
58
56
53
53
41
54
48
31
44
47
53
52
46
55
4,740
100,000
counts
89,749
219
215
236
223
215
212
191
215
244
221
216
187
190
187
200
197
170
6,707
99,994
keV
1,830
1,815
1,800
1,785
1,770
1,755
1,740
1,725
1,710
1,695
1,680
1,665
1,650
1,635
1,620
1,605
1,590
< 1,575
sum
keV
525
520
515
510
505
500
495
490
485
480
475
470
465
460
455
450
445
440
<435
sum
counts
94,750
33
37
30
39
29
27
28
39
32
35
37
40
42
32
36
25
4,702
100,000
counts
90,759
142
124
126
138
122
165
164
146
141
146
120
118
127
135
127
140
121
6,938
99,999
(a) The source consisted of a 12.2 ml right circular cylinder of water with a radius
of 1.14 cm and a height of 3.037 cm. The spectra proved to be independent of
both detector size and source-to-detector distance, to less than 0.03%.
(b) The listed keV is the minimum energy for that channel. Thus, the first channel
contains only unmodified photons.
(c) The listed sum represents the total number of photons reaching the detector.
The difference between this sum and 100,000 is the number of photoelectric
events that occurred within the water.
-668-
-------
TABLE 2. Percent of Source Photons Counted Within Photopeak.
12.4 ml liquid source
3"x3"NaI(Tl)
keV(a)
80
150
190
250
400
525
660
840
1,335
1,590
1,830
keV(b)
50
110
145
200
335
445
570
735
1,215
1,455
1,695
3 cm (c)
10.2
9.73
9.68
9.64
9.55
9.61
9.67
9.70
9.90
9.95
9.89
10cm
2.24
2.14
2.15
2.11
2.17
2.17
2.17
2.13
2.20
2.29
2.18
30cm
0.316
0.310
0.304
0.337
0.324
0.318
0.319
0.294
0.315
0.334
0.319
46.2mlGe(Li)
keV(b)
75
145
185
245
395
520
645
835
1,320
1,575
1,815
3cm
3.19
3.20
3.21
3.22
3.33
3.36
3.39
3.39
0.690
3.57
3.49
10cm
0.646
0.611
0.623
0.638
0.677
0.664
0.649
0.655
0.0945
0.719
0.670
30cm
0.0830
0.0950
0.0775
0.110
0.103
0.0880
0.0870
0.0785
0.111
0.104
14.4 ml off-gas source
3"x
keV
d
3"NaI(Tl)
3cm
9.95
10cm
2.29
30cm
0.344
46.2 ml
keV
d
Ge(Li)
3cm
3.55
10cm
0.717
30cm
0.107
(a) Initial source photon energy.
(b) Low energy channel used for beginning of photopeak summation.
(c) Distance from bottom of sample vial to top of detector can.
(d) For a gas sample, essentially all photons are unmodified. Thus,
the fractions are not a function of initial photon energy.
TABLE 3. Combined Factors for Geometry and Self Absorption.
keV
80
150
190
250
400
525
660
840
1,000 to 2,000
3"x
3cm
1.03
0.98
0.97
0.97
0.96
0.97
0.97
0.98
1.00
3"NaI(Tl)
10cm
1.00
0.95
0.94
0.94
0.93
0.94
0.94
0.95
0.97
(kj/kg)
30cm
0.97
0.92
0.91
0.91
0.90
0.91
0.91
0.92
0.94
46.2
3cm
0.90
0.90
0.90
0.91
0.94
0.95
0.97
0.99
1.00
ml Ge(Li)
10cm
0.87
0.87
0.87
0.88
0.91
0.92
0.94
0.96
0.97
30cm
0.82
0.82
0.82
0.83
0.86
0.87
0.89
0.91
0.92
-669-
-------
XI. Round Table Discussion
of Noble Gases
-------
NOBLE GASES FROM NUCLEAR REACTORS:
CONTAINMENT VS. ENVIRONMENTAL RELEASE
During the Symposium, a round table discussion was held on the subject of noble gases, in general, and
krypton-85, in particular. The members of the panel were: V.P. Bond, M. Eisenbud, C.C. Gamertsfelder, and
E.G. Tsivoglou. Chairman was A.A. Moghissi.
MOGHISSI: Welcome to the panel discussion. I believe the stage is set for a very interesting and fruitful
discussion. Before I get started, let me introduce the panel. On my extreme right is Dr. Carl C. Gamertsfelder, a
technical assistant to the Director of Regulatory Standards of the U. S. Atomic Commission. He received his
Ph.D. in Radiation Physics from the University of Missouri in 1941 and soon afterwards joined the
Radiological Laboratory of Chicago. He also served at Oak Ridge National Laboratory in 1943, and at
Hanford in 1944. On these three assignments he was involved in various developments in health physics and,
most particularly, in matters related to production of plutonium for weapons. In 1952, he joined the aircraft
nuclear propulsion project. In that capacity, he was concerned with safety analysis, metallurigical studies,
and application of atomic energy. In 1964, he performed aerospace safety analysis on nuclear power projects
and since late 1969 he has been with the Atomic Energy Commission.
On the extreme left is Dr. Merrill Eisenbud. He is Professor of Environmental Medicine and Director of the
Laboratory for Environmental Studies at New York University Medical Center. He joined the University's
faculty in 1959 following a 12-year career with the Atomic Energy Commission during most of which he
served as the Director of Health and Safety Laboratory. From 1936 to 1947, he served as an industrial
hygienist on the staff of the Liberty Mutual Life Insurance Company where he was engaged in studies of
chemical and radiation hazards in industry. For two years, from 1968 to 1970, Dr. Eisenbud served as the first
Environmental Protection Administrator for New York City. In that capacity, he had the primary
responsibility for the Departments of Sanitation, Air Resources and Water Resources. Dr. Eisenbud holds a
degree in Electrical Engineering from New York University and two honorary doctorates from Fairly
Dickerson University and Catholic University of Rio de Janeiro. He is a member of many National and
International Committees concerned with Public Health. Dr. Eisenbud is the author of numerous papers,
including a textbook on Environmental Radioactivity, the second edition of which just appeared.
Next to him, on my left, is Dr. Victor Bond, who is the Associate Director of Life Sciences and Chemistry at
Brookhaven National Laboratory in Upton, New York. He is internationally known as an expert in the field of
nuclear medicine, radiation biology and therapy. Dr. Bond served as a medical officer in the U. S. Navy from
1945 to 1954. For six of these years, he was head of the Experimental Pathology Branch of the Naval
Radiological Defense Laboratory in San Francisco. Following his tour of duty with the Navy, Dr. Bond joined
the staff of the Brookhaven National Laboratory's Medical Research Center. He held increasingly
responsible positions at Brookhaven until he was appointed in his present position as Associate Director of
the Laboratory in 1967. Dr. Bond is also an Adjunct Professor of Radiology at Columbia University and
Adjunct Professor of Medicine at the State University of New York at Stony Brook. Dr. Bond is a member of a
large number of national and international organizations. He has published over 250 papers. He has
contributed to many books and monographs. Most of us who have been in this business have referred to them
for a long time. If I may add a personal note, Dr. Bond has given a number of ideas which I have used in my
research.
Last, but not least, on my right, is Dr. Tsivoglou. Dr. Ernest Tsivoglou is a Professor of Civil Engineering at
the Georgia Institute of Technology in Atlanta, Georgia, where he has been a member of the faculty for the
past 7 years. Before joining the faculty of Georgia Tech, Dr. Tsivoglou was with the U.S. Public Health Service
from 1949 to 1966 in Cincinnati and Columbus, Ohio, and Salt Lake City, Utah. In the Public Health Service,
he was a Sanitary Engineer Director, and was Chief of Physical and Engineering Sciences in the Division of
Water Supply and Pollution Control. He had various assignments in his Public Health Service career working
in stream sanitation, radioactive waste control, studies of the fate of pollutants in the environment and
research in the kinetics of natural processes. He also provided technical assistance and consultation to State,
Federal, and other National agencies. Dr. Tsivoglou is largely responsible for the recurrence of the
discussion on environmental radiation standards and emission standards from power reactors.
Here are the members of our panel; now I would like to ask them, one by one, to give a brief statement of their
ideas, positions, what else they want to say in terms of the theme of our panel discussion, after which time I
would appreciate receiving questions.
EISENBUD: I will confine myself to the subject of krypton-85 because this panel is supposed to discuss the
pros and cons of containment versus release of krypton-85, the long-lived Noble Gases, of which there is only
one. We are working in a field in which information is accumulating more rapidly than the risk. We talk about
the projected per capita dose rates for the year 2050, but that is a long way off. We are dealing with a 10.5-year-
half-life nuclide which, at the present time, is being produced in relatively modest amounts. I know, for
example, that the present world-wide estimate for 1970 was a production of 6.6 MCi. This will go to 102 MCi in
1980 and up to 1000 MCi by the year 2000 and so on, into the next century.
We are talking, so far as National policy is concerned, about a relatively small number of plants that are the
principal sources of krypton-85. It has been noted by several speakers that the reactors put out about 0.1% or
less of the krypton-85 so that 99.9% or more will be released by relatively few fuel reprocessing plants. I believe
-670-
-------
that what is needed is a sort of interim national policy which should be reevaluated every 5 or 10 years. From
my point of view, there is no question that this would call for a containment of the krypton-85 in some form by
the fuel reprocessing plants. Whether this would need to be done in the long-term, would depend on how a
number of important questions are answered in the future. I didn't hear any of the speakers question the
validity of the risk coefficients that have been used by the various speakers and which have their origin in the
BEIR Report. I would certainly hope, and this is what I mean when I say the information will accumulate
more rapidly than the risk, that, in the next decade, the biologists would give us answers to some fundamental
questions, such as: Is there dose rate dependence, and should the risk coefficients, which are based on
relatively massive exposure, be modified accordingly? For a given dose rate, is there a linearity, and is there a
threshold?
When we talk about the genetic effects of radiation and such diseases as leukemia in humans, these are very
difficult questions to answer. I would hope that Dr. Bond might address himself to this question. One
difficulty is that the leukemia incidence is normally something like six cases per hundred thousand people per
year: It is a relatively rare disease. But skin cancer is the most common of the neoplasms, and, incidently, one
that is rarely fatal. If the skin is that sensitive to cancer so that we need to be concerned about per capita doses
of the order of one to ten mrem per year, it should be possible by epidemiological techniques in a population as
large as the atomic energy workers to see if they have a higher incidence of skin cancer. The incidence of skin
cancer in the general population is very high compared to leukemia. It is estimated that one out of four light-
skinned Australians develop one or more basal carcinomas by the time they are 65 years old. Thus, there is a
much larger epidemiological base. For example, the uranium refinery workers who handled natural uranium
and were exposed to the distilled daughter products in the course of the process, received relatively high skin
doses. This group should be studied.
In summary, there are some fundamental questions, mainly of a biological nature, which will determine
whether or not the radiation exposure of the order of background are significant. We should bear in mind that
these levels will not be approached until sometime in the next century and our National policy should be
developed accordingly. If my calculations are correct, the present per capita skin dose from krypton-85 is
about 27 jirem per year and this will increase by 1980 to 0.6 mrem per year, and then to a little less than 1
mrem per year by the year 1990. Thus, 17 or 18 years from now, the dose will still be less than 1 mrem per year
on a per capita basis. This would give us time to get the basic biological information and to make the decision
as to what the National policy should be.
BOND: First, I would like to thank our Chairman for his very generous introduction. I do not have a prepared
statement and there is little that I can say that has not already been said. I would like, in essence, to repeat
several points that have been made.
First, let me state that I believe in the approach that has been used, of "as low as practicable" and that, "no
radiation exposure should be given or allowed without good reason and compensating benefits." I also think
that if we are exposed to a number of risks, this in no way gives us license to be cavalier about exposure to other
risks. However, I also believe in looking at risks in context and in the philosophy of "as low as reasonable" as
well as "low as practicable."
Let me remind you of what was stated this morning. The speakers agreed that the principal dose from
krypton-85 was to the skin, and that dose to internal organs was a couple of orders of magnitude less than the
dose to the skin. The figures that were given were of the order of a few mrem sometime in the next century. The
dose to the whole body would be one or two orders of magnitude below that. In other words, we are talking
about a small fraction of one mrem per year to the whole body.
We heard certain data on biological effects. With respect to the effects of radiation on the skin, no increase in
skin cancer in those exposed in Hiroshima and Nagasaki have been reported, and a period of approximately
30 years has elapsed. The exposure was, of course, at high doses and dose rates and included neutron
irradiation in the exposure in Hiroshima. The report of the Biological Effects of Ionizing Radiation (BEIR)
Committee of the NAS/NRC (1972) presented data on a number of cancers which were well-known to result
from exposure to radiation delivered at high doses and dose rates. The data were insufficient to provide risk
estimates for skin cancer. .
Numerical risk estimates for a number of other types of cancer were obtained as upper limit nsk values by
interpolation between data obtained at high doses and the natural incidence at background exposure levels.
The BEIR Committee, however, which was very conservative in its approach, did not feel that the data
available were adequate to provide risk estimates for skin cancer. This goes along with results available from
some of Dr. Eisenbud's associates, on effects of exposure on the skin of animals. The curves appear definitely
to be curvilinear, with increasing slope with increasing dose. There also appears to be a dose rate effect which
generally follows curvilinearity. Also, United Nations Scientific Committee on Effects of Atomic Radiation
(UNSCE AR Report 1972) which opined on the same data that the BEIR Committee reviewed, provided no risk
estimates for skin cancer. The UNSCEAR Committee, however .unlike the BEIR Committee refused to
providerisk estimates for any type of cancer for exposure at low doses and dose rates. This was done on the
basis that the data obtained in man apply only at the high doses and dose rates at which they were obtained,
and that the risk estimates obtained at low doses and dose rates obtained by interpolation very likely
overestimate the true risk by a considerable margin. In summary, the risk of skin cancer, following exposure,
is minimal, indeed.
-671-
-------
Now, with respect to the removal of krypton, and the approaches to removing it from power plants and from
reprocessing plants. From what I heard today, it is expensive to do this. Dr. Cohen said that the releases from
power plants represent only a very small fraction of that from reprocessing plants. Let me remind you that,
even into the next century, the average dose is 1 mrem per year. As a taxpayer, it seems to me that our money
could be better spent than outfitting power plants to prevent the release of this isotope. I think that we could
achieve much more real reduction in the mortality rate by spending this money elsewhere rather than
spending our money to prevent imaginary deaths from this isotope.
So, my position then is similar to that of Dr. Eisenbud. I would have to be shown much more than I have
been shown that there is a biological hazard, or that the dose is higher than has been shown, before I would
become an advocate of outfitting power plants to reduce the release of krypton-85.1 would have questions
about outfitting the fuel reprocessing plants, but feel less strongly about that. Thank you.
GAMERTSFELDER: I am beginning to feel a little bit like Hoyt Whipple this morning, and people who have
talked about all the things that are hard to talk about. I would say, in respect to removing krypton from the
power plant emissions, that the AEC has not considered doing it at present. There would be no point in doing it
until there were some effort put into removing it from the major source, which is the reprocessing plant. Now
with respect to reprocessing plants, the AEC has no definite plans to remove krypton, but the people who are
operating reprocessing plants have been told and have cooperated in providing enough space in their plants
so that, in the future, when the equipment is developed, it could be put in. The bulk of exposure results from
releases of krypton in countries other than the United States. It seems to me that if it becomes technically
desirable as well as morally desirable to remove krypton from our fuel reprocessing plants, there should be
some kind of international cooperation. We could get a lot more by getting some cooperation from our
neighbors. I do not know if I have enough information to say when these things ought to take place. I would
leave that to, I guess, the State Department.
TSIVOGLOU: I do have some prepared comments that I would-like to make before we get to the question
period. I feel like everybody else here on the platform that all my stuff has been stolen before the discussion got
to me, but that happens. If you will bear with me, I would like to read these comments, because I find I stay out
of trouble better that way.
The question of containment versus environmental release is far from new in the pollution control business.
The same old question has been raised in connection with virtually every pollutant that has appeared on the
scene. And it has been the principal question that I have contended with since I began working in this field
some 20 years or so ago. The arguments from both the proponents and the opponents of environmental release
are still also much the same. I don't think anybody here will be surprised to find that I am opposed to a policy
of unnecessary environmental dispersal of the radioactive noble gases that result from nuclear power
generation. We have seen a lot of progress in the last few years, let me say in the last four or five years. I recall
as though it were yesterday the arguments in Minnesota that there was no technology available to reduce the
stack releases from nuclear power plants; that the cost of doing so would be outrageous and would stifle the
growth of the industry; and that releases up to the AEC limits would not hurt anybody anyway. Yet, in the
short space of four years, we have seen the development of an excellent containment technology. We have seen
that the actual costs are not that unreasonable. The industry still seems to be in a healthy state. And there has
been a sharp reduction in the release limits that used to be regarded as acceptable.
It appears to me that the earlier real problems associated with noble gas releases from reactors are solved in
the large sense and under much improved control, and that such releases can be expected to be truly minimal
in the future. Therefore, I would like to confine my remarks here to the matter of release of krypton-85 from fuel
recovery plants. We all seem to agree that that is the major noble gas problem that confronts us. Now, I too
have a few numbers. They are crude numbers and I will duck as rapidly as possible later if anybody tries to pin
me down on all of the assumptions, on which exponential curves I have used, and what references I am citing.
An eight hundred megawatt nuclear generating plant of the Dresden-II type will have something like a half
million curies of krypton-85 in the fuel rods after an initial twelve months of operation at full capacity. A
considerably larger inventory will accumulate as operation continues, depending upon the operating level,
frequency of fuel changes, and so on. There should be very little loss, we all agree on that, so that essentially all
of that krypton-85 will go to a fuel recovery plant. The current generation of commercial fuel recovery plants
envisions release to the atmosphere of all krypton-85 received. The Allied Gulf Plant at Barnwell, South
Carolina, expects an atmospheric release in the neighborhood of about twelve million curies a year when
operating at full capacity. The GE Midwest Fuel Recovery Plant projects an annual release of about three and
a half million curies, but I understand that is operating one shift. The Nuclear Field Services Plant in New
York State is currently shutdown for improvements. With the backlog of fuel that is accumulating for
processing it seems safe to assume that all of these recovery plants, all three, will be operating at full capacity
as soon as possible. So that within a very short time, with only those three plants in full-capacity operation, we
can expect a combined release of 20 to 30 million curies a year of krypton-85. And that is the immediate future.
As more generating plants come on line over the next two decades or so, there will be an increasing need for
fuel recovery capacity, and the annual krypton-85 release will increase accordingly. By the end of the century
then, according to my horseback numbers, if we go on the way we are, we alone will have released somewhere
around a half a billion or more curies of krypton-85 to the environment. Let us admit right off the bat that it
would undoubtedly be most difficult to prevent the escape of some of this gas and there are bound to be
-672-
-------
occasional small releases. I am not an adherent of the currently popular zero-pollution cult either. I am a
physicist among other things, and to me zero is an absolute, not a hazy concept. It is a word that is being used
just a little too loosely today by politicans and, sometimes, by people who should know better. However, to me,
the proponents of unrestrained environmental release represent the extreme opposite. Now they would leave
future generations a legacy of world-wide contamination that I have to regard as inexcuseable. Perhaps the
key to this problem is the matter of need. If I were able to convince myself that we were incapable of developing
a suitable technology for krypton-85 containment or so destitute of money that we could not afford to contain
it, then perhaps I might be more willing to go along for a time with a temporary policy of environmental
release. But, I simply do not believe that such a real and honest need exists.
Let me recall briefly some of the arguments. The arguments against containment involve again the
question of available technology (do we know how to do it?), the matter of whether we can afford the cost, the
question of proof of harm, and the possibility of accidental release of the stored krypton-85.
Perhaps this example will help explain my attitude a little better. About two years ago, I published a paper
in which I called for containment of krypton-85 from fuel recovery plants. Shortly thereafter, I was told by
representatives of the industry "we would be glad to capture and store the krypton-85, but there is no practical
technology available that is applicable on a full scale for a plant just like ours." It seems to me I am being
presented with the question of which came first, the chicken or the egg. Of course, the technology needed to be
developed just as it needed to be developed four years ago to control the stack releases from boiling water
reactors. I have too much faith, in essence, in the skills of our engineers and scientists to accept the argument
that there is no available technology. From what I see of the program, and from what I have heard here and in
the halls, I gather that there is fairly general agreement now that the technology is available, or can be very
shortly. As for the cost, coming up with a cost that is reasonable appears to me to be an integral part of
developing a practical technology for containment. The problem of potential accidents with the stored
krypton-85 is also technological in nature: I can't belive that we are unable to design a storage facility that can
safely contain or disperse the relatively small quantities of gas that might escape if a tank were to rupture.
And so I come finally to the old proof of harm argument, like who would be hurt anyway if we were to release it
all? I must confess that I cannot prove and I would not attempt to that a billion or two curies of krypton-85 in
the world environment will harm anybody. But I also must confess that I do not have infinite faith in the
wisdom of those who decide what is safe and what isn't. To gamble now that no future harm can result from a
policy of environmental release, to be so certain of our infinite wisdom, is to me to gamble with the well-being of
future generations.
It is well, too, to remember that these little gambles can add up to larger ones. Krypton-85 is not the only
radioisotope that we have around. And so I come back to the matter of need. In my view, the only acceptable
justification for a policy of environmental release of krypton-85 would be a real and convincing proof of need.
Unless I can convince myself that environmental release is a real and pressing necessity, tarn unwilling to
gamble, even in a small way, with the well-being of future generations. I am afraid I remain unconvinced that
there is a need. I think we will know, in a very short time, whether our technological people are capable of
developing containment processes that are effective, reasonable in cost, and safe. If my faith in their skill
proves to be unfounded, I will hang my head.
MOGHISSI: Thank you members of the panel. Now is the time for questioning.
NEWMAN: I am Bob Newman with Allied Gulf Nuclear Services. When we first announced to the
Commissioners of the AEC in February of 1968 that we intended to build a plant in Barn well, South Carolina,
we made the commitment that the plant would be designed to accommodate krypton capture facilities as far as
space is concerned, connections, penetrations, and so forth. It was qualified; it is still qualified. We need to
know when the cost benefit study or risk benefit or both warrants such an installation, when the technology is
available, and when criteria and methods of storage are developed. I can tell you that when our plant in
Barnwell goes into operation, it will have incorporated in it facilities for pollution abatement in excess of sixty
million dollars. I do not think we are skimping, nor are we trying to delay. We, in our present climate, accept
the linear dose relationship. We have our doubts, but we see no choice at the present time but to accept it. We
especially have doubts, as has been brought out here, on the effect of low exposures to the skin from krypton.
Even so we are going ahead and assume it can be used as a basis. We do not think the Industry should respond
to emotion I was in Washington, talking to one of the Agencies. One of the first questions that came up was,
"when are you going to start capturing krypton?" I say, "why? why should we?" The answer was simply, "the
public demands it." Not one technical reason, not one justification, not one health related reason for capturing
it but the public demands it. I do not think the Industry can go on this way. There is a finite limit to the number
of dollars we have I want to see our dollars spent in the right way to do the most good. We agree with the
principle of "as low as practicable." Licensing boards now are applying this to workers as well as to the
public. I think one can opine from authoritative sources, projections of future doses of krypton-85 for the year
2000 vary by almost two orders of magnitude. I think this is one problem that we are faced with. The experts
should be able to get together and say this is a reasonable projection and a realistic projection on which we can
do our planning. I second the suggestion that as much work should be done as practical on defining what is
the consequence of the skin exposure to very low levels of radiation. Are they important? How does one attach
a dollar value to them? What is the value of a man-rem of very low level skin dose/
-673-
-------
I do not think there is technology available which we would want to recommend to our management. I am
sure there is at least one person in this room that disagrees with me from the comments I heard before. I do not
think that we can obtain designer assistance for a technology which we could license through the AEC,
meeting all theirsafety requirements! am a chemical engineer. I have spent most of my career in the chemical
business. I know the countless times the gas purification systems have failed because of trace impurities not
recognized. You can synthesize a gas mixture and demonstrate a purification system. But, as in the case of
krypton recovery facilities in operation today, trace impurities can lick you. The long-term storage has not
really been looked at. Yes, it is feasible, but it has not really been addressed and engineered and the safety
implications considered. Accidents do have to be looked at, we do not have enough knowledge yet to really
assess the potential nor the consequences of an accident. Dr. Morgan, I will agree, the health physicists have
done a fantastic job, but if you have radiochemical plant processes, from time to time, you have to do some
maintenance and you are going to get some exposure. Is it safe? I think it is safe. Is it as low as practicable? I do
not know, but still it is a trade-off.
MOGHISSI: Please make your statement brief and put your questions to the panel.
NEWMAN: I would like the panel to answer the following questions: 1. Define how the public can be
educated in the real life of the potential consequences of krypton release. 2. How can the Industry, and that is
used in the broad sense, the whole nuclear community, obtain well defined and supportable numbers on which
we can make projections and plans. 3. How can we get the AEC, or maybe EPA, to set forth the overall criteria
to which krypton recovery facilities should be addressed and on which you would base your decision to install
the krypton recovery facilities including how does one weigh one number vs. another when they are not in
consistent units. And, finally, can we get this whole thing on a real life basis and eliminate, at least within our
own house, the emotion and hysteria?
MOGHISSI: Does any one of the panel members volunteer to answer all these questions?
EISENBUD: I cannot answer any of them, but I would like to ask one more question which I think perhaps
Dr. Gamertsfelder can answer. Is there a krypton recovery system in existence today which the Commission is
prepared to license?
MOGHISSI: The question is for Dr. Gamertsfelder. The last question first and then we go on to the other
questions.
GAMERTSFELDER: I know of none and I do not think our licensing people know of any either. I do not
think the research work has been done to clean up the gas streams that the krypton collection system would
have to have before it could be operated. The initial stream has nitrogen oxides, hydrogen, oxygen, and
various other things which have to be looked at from a safety standpoint as well as from an economic one.
TSIVOGLOU: I thought I heard a littler earlier from Jim Martin of EPA, that EPA has reached a conclusion
that there is a technology, that is either available or very close to it, for krypton removal from fuel recovery
plants. I even thought I heard him put a figure of about ten million dollars on it which, let us suppose, is
optimistic — perhaps it will cost somewhat more than that. So I am a little confused really in the sense of what
stage of technology we are. I presume that the EPA and AEC have been working together and are aware of
everything that everybody else is doing. Perhaps somebody from EPA would like to comment on the stage of
their investigations in the technology.
MOGHISSI: Before the somebody from EPA comes, there is somebody from Industry who would like to make
a statement.
DAVIS: (Linde) Two years ago when the fuel reprocessing plant in Morris, 111., went out for bids for a krypton-
85 recovery system, we provided them with a commercially proven system using components that have been
used for the last 50 years in the chemical industry which would give krypton recovery within 99%. Cryogenic
systems and their present state-of-the-art are such that there are probably over a thousand plants in the world,
some processing over twelve million cubic feet of gas an hour, and these plants operate 99% on stream time.
For the Morris facility, we were talking about $800,000 for installed equipment or purchased equipment, and
probably $1,500,000 in installed units and utilities. Operating costs are about two to three hundred dollars per
day. I think that our paper in this Symposium shows that technology is here, it comes from a different area and
has been used and proven thoroughly in many other applications outside the nuclear industry. We have the
technology and even though its components are drawn from other fields, it is available.
MOGHISSI: Is Jim Martin here? Jim, a question was put to you. EPA has made a decision to recommend the
collection of krypton from fuel reprocessing plants. Has this decision been discussed with the AEC?
MARTIN: Carl (Gamertsfelder) can answer that question for you. Yes, it definitely has been. The AEC does
raise points that are fair. We do not blame them one bit for being unsure that the system can be made, not only
-674-
-------
to perform to meet an environmental standard, but also that it can be made to perform in a way that it won't
blow up in somebody's face. I submit that these are technical problems that can be overcome within a
reasonable time. What is lacking is the decision to get on with the solution of those technical problems. Now,
as an Agency, we are trying to be quite fair in allowing or recommending a time period that is reasonable to
work toward those solutions. But, we are ready to bite the bullet and say it has got to come out. Once you get
moving, it may not perform at 99%. You may have leakage, you may only get 90%, but until you start, you are
not going to get these problems solved and I think that what has been holding us up, perhaps the starting of
this.particularapproach, has been the conviction that is tied to the things that I tried to point out in my paper at
this Symposium. We have not really looked at the total impact of the decision. In view of the comments from
Linde, and several other places, we are not convinced that the arguments that the technology is not available
are valid. We are prepared to admit that it's not demonstrated, but we believe that we are close to it.
EISENBUD: My instinct would lead me to believe that this is why we should not freeze the National policy
around a single piece of equipment or a single performance standard. When I said that I thought that the
information was accumulating faster than the risk, I was not talking entirely about the biological
information; I meant the engineering information too. It might be useful, as a matter of National policy, to go
through a 5- or 10-year period of demonstration of these plants seeing to it that one installation uses
cryogenics, if this seems reasonable. I think the technology will eventually prove to be feasible from what I
have seen from the literature. I think that we are probably almost there. I would hate to see us making the same
mistake in the nuclear industry as they have made in the automobile industry in mandating a short time table
for emissions controls which will lead to installation of ineffective hardware which will not do the job in the
long-run.
BOND: I agree with Dr. Eisenbud. I do not see a reason for a crash program or for hasty regulations in this
area. The low doses to which we are exposed in the present state of the technology are such that we have plenty
of time to evaluate more thoroughly than has been done up to the present. I would also bring up something that
was touched on several times and that has to do with the trade-off in risk — real vs. hypothetical risk; if the
material is concentrated versus dispersing it widely. It has been stated that if one wishes to spend the
necessary amount of money, the gas can be concentrated with a minimum exposure rate to the individuals
concerned. However, I would expect that the cost would be considerable. While, perhaps, all persons are
created equal, not all man-rems are. If we concentrate the material, we now have large amounts of radioactive
isotopes and large amounts of radiation in a single location. I have dealt enough with accidents generally, and
with nuclear installations to appreciate fully the human element. Accidents have occurred and they will occur.
I do not have much faith in probabilities of accidents. I know that accidents have occurred where it is
inconceivable that they could have occurred and substantial exposures have been sustained. What I am
getting at, basically, is that what we may be trading is an unsubstantiated hypothetical risk from the material
widely dispersed versus a strong possibility of a real obvious hazard to individuals resulting from the
concentration and storage of the materials.
TSIVOGLOU: I would like to clarify something. I did not intend to call for a "crash program." I do not think
that is what Dr. Martin of EPA was speaking about either. I have not heard anybody here, including myself,
say that a crash program is necessary. I have heard people indicate that what we really need is a decision to go
ahead. Because we do not even have a decision to go ahead yet, I am with Merrill Eisenbud, completely. In the
sense of a time lapse, if we, today, got a decision to go ahead, from the way things have gone, it looks like five
years is the minimum before there will be anything in operation. So, nobody is calling for a crash program.
GESELL: I am with the University of Texas School of Public Health. I just had the thunder stolen out of my
question. Both the AEC and the EPA have a history of supporting research on technology for preventing
environmental contamination, for instance, the research on the emergency core cooling. I was wondering if
there were any plans, and I am addressing this to any spokesman from EPA or AEC who might be here, for
support and funding research on krypton removal.
MOGHISSI: Carl, I think you would be the right person to answer that question.
GAMERTSFELDER: I cannot say that there are funds available at the present time. I know the subject is
being discussed within the AEC, but decisions of that kind are not made by a few people discussing it. It takes
time.
MOGHISSI: Any more comments?
BENDIXSEN: (Allied Chemical, Idaho) The last comments have been very interesting to me since we
happen to be recovering krypton-85 in a distillation facility. Who is going to fund the first demonstration
facility — EPA, AEC, or the industry? I do not count the unit we are operating as the first unit. Anyone who
has looked at the reports we have published knows that we do not say that, and if they look at our efficiencies,
they know that it is not the demonstration unit. Since we have a good representation on the panel, I would like
to ask who is going to fund it? I might tell you right at the moment, there is not any funding for a
demonstration unit. There is funding at Oak Ridge for a certain development work for a liquid carbon dioxide.
-675-
-------
There has been some for a fluorocarbon system. There is no funding at all for a cryogenic system within the
AEC. There has just begun this last fiscal year some funding to start looking at the containment and disposal
of krypton-85. But let me return to my initial question. Who should fund it? EPA, AEC, or Industry?
EISENBUD: I will volunteer to answer this question because it is something that I feel very strongly about. I
think that the National policy to put the AEC into the R&D business, for the first 20 or 25 years in the _so-called
atomic era was very sound, and could be justified. But we now have a nuclear industry which is building 30 or
40 reactors a year. At the present time, more than 200 nuclear power plants will be built in the next six or seven
years, It is a fifty or hundred billion dollar industry. That industry consists of four or five manufacturers and
two or three reprocessors. I suggest that Industry should set up the money to fund the research and not expect
Uncle Sam to do it for you.
TSIVOGLOU: I cannot resist saying Amen.
BENDIXSEN: Let me point out that I am employed by the Government too. I am employed by Allied
Chemical, but we are under subcontract to the AEC. I only do what they have told us to do and fund us to do at
the moment.
EISENBUD: Pardon me if I point my finger. YourCompanyis one of the largest industries in the world. You
have got to get together and develop the mechanisms for doing this internally.
BENDIXSEN: Please do not take my words as a member employee of Allied Chemical as any position of
Allied Chemical. I have spent the last eight years in the same office and I have worked for three different
companies. Those three companies have had the same subcontract with the AEC. So my company loyalty
should not mean that I am stating something for Allied Chemical.
DEMPSEY: I am with the Division of Waste Management and Transportation. We are the Division
responsible for evaluating and developing containment technology and we have a program with Allied
Chemical at present. They finished a study on their cryogenic technology and put costs down for making a full
stage system. Our plan is to go on to a demonstration by 1979. In the meantime, we are looking at the storage
aspects.
PRETRE: (Switzerland) I would like to support what Dr. Bond has just said. May I recall that there is on one
side, the general public, and, on the other, the personnel of the plant. The man-rem situation now is roughly as
follows: For the public around the plant, the exposure between one and ten man-rem per year and for the
personnel in the plant it is between one hundred and one thousand man-rem per year. That means the dose in
the plant is two orders of magnitude higher than the dose outside the plant, It seems, therefore, that we have
already passed the optimum. It is also important to realize that the man-rem for the personnel in the plant are
real man-rem absorbed at high dose rates. On the other side, the few man-rem absorbed by the public are
calculated and if they are really absorbed, they are at low dose rates. So, any additional retention of
radioactivity in the plant will increase the total of man-rem instead of decreasing it. Therefore, it might be
completely wrong to further concentrate all the radioactivity at the same point.
STRONG: I am with the Atomic Energy Commission, Licensing. I would like to address my question to Drs.
Eisenbud and Bond. I was intrigued by Dr. Eisenbud's comment that he expected that, within the reasonably
near future, we might expect information from biological studies that would resolve current problems
regarding thresholds for cancer reduction: Linear versus other probability functions, etc. I wonder if you
would expand on this and perhaps indicate what might represent reasonably conclusive evidence.
BOND: Let me say that at the time the BEIR Committee was reviewing data available to them, relatively
little — this might be surprising to you — but relatively small amounts of data on late effects in animals were
available to that group. A few years ago, a large amount of work, financed by the Atomic Energy Commission,
was undertaken to evaluate such things as the shape of dose effect curves for different types of tumors in
animals, the possibility of dose rate effect, RBE, and so forth. These experiments took years to complete. The
data had not become available at the time of the BEIR Committee or the UNSCEAR Committee's
deliberations. As a result, members of the BEIR Committee paid very little attention to animal data in their
deliberations. They paid most attention to the relatively poor data available on the human. The UNSCEAR
Committee, on the other hand, paid more attention to animal data, even then available, and took a far less
conservative position in their report than did the BEIR Committee. Since the publication of these reports, data
from these experiments that have been in preparation, are becoming available. These data are not published,
as yet, but I have seen preprints. I will give you my interpretation that the data indicate that the bulk of
animal tumors will show a curvilinear dose-effect relationship with increasing slope. They also indicate that a
dose rate factor will apply in a number of situations. In addition, there are techniques open to us now at a very
fundamental level to evaluate better, both somatic and genetic effects of radiation. I would expect, over the
next few years, the data available to us and our ability to interpret these data, will be markedly improved over
what they were at the time of the BEIR and UNSCEAR reports were written.
-676-
-------
EISENBUD: It would be presumptuous of me to say that I agree with what Dr. Bond says, but I certainly do. I
would add one more point. There is an admirable tendency now to reevaluate old animal and human data. A
recent paper by two of my own colleagues, Drs. Albert and Altshuler, looked at not only radiation tumors, but
cigarette smoking. In examining both the human data and the animal data, they have shown that as the dose
goes down, the delay in onset of the tumor is longer. Of course, Robley Evans was the first to point this out, and
he coined a term "practical threshold." He pointed out that the radium dial painters who had lower body
burdens of radium took longer to develop the bone cancers. So, that in theory, if the incubation period is longer
than the life of the person, the dose cannot be termed carcinogenic, because death will result from other causes.
KIRK: (EPA) I hate to muddy the waters at all, but there was a paper several years ago by Dunster and
Warner in which they stated the opinion that the first group that would start receiving significant radiation
exposure to krypton would probably be the people in the air products industries, recovering the krypton from
the air. We may be trading off a high radiation dose to a segment of the population in the radiation industry to
avoid a higher man-rem dose to a segment of a population outside of the nuclear industry.
SOLDAT: I am with Battelle Northwest, Richland, Washington. I would like to make two points. First, I do
not want anyone to leave the room with the impression that the utility industry is sponsoring absolutely no
research and development effort in terms of nuclear safety. This is an impression that might be gained from
some of the things that were said here today. And second, I would like to restate what Dr. Bengston of the
ICRP stated two weeks ago at the meeting of the International Radiation Protection Association. He said that
it is perfectly logical to use reasoning that was used in the BEIR Report to arrive at dose limits. But, it is not
equally logical to use such ultra conservative reasoning to calculate the effects of these doses.
MARTIN: (EPA) I raised another question earlier that fits into this argument of looking at the long-term
commitment of the decision to dispose of pollutants in a certain way. It involves radon from our monuments
called "tailing piles." I would like to have the panel comment on that, while I think we have bogged down in a
morass here on krypton-85 and the battle lines and the opinion lines are pretty well drawn. Could we have
some comment on radon?
BOND: I cannot comment, in depth, and would like to say only that I have been rather surprised at the time
delay in effecting measures to reduce the exposure of miners. It was known many many years ago, in the last
century, in deep mines in Europe, that miners did develop lung cancer. This was associated with radiation
exposure. Yet, this went on for years without effective protection and it is only recently that adequate
measures have been put into effect. I would also like to say that, with the exposure to radon, the doses received
by the bronchial epithelium of the miners are quite high and very much higher than what we are talking about
from krypton. Furthermore, exposure from radon is to alpha emitters, not beta emitters, so the problems are
not comparable, at all, in that respect.
TSIVOGLOU: I feel some urge to comment since I guess the first nuclear work I did involved studies of radon
and its daughter products in uranium mines back in 1952. Dr. Bond commented very justly on the long-time
delay between the time we knew what to do, and the time anyone actually got down to work and began to try to
protect uranium miners. My recollection of the time delay is something like 15 or 18 years. We knew in 1953 or
1954 that if we ventilated a uranium mine we could cut these doses down. This is a good example because it is
the crux of the matter that we have here about krypton. The time delay occurred because we could not get a
decision anywhere to go ahead and do something, because people argued and kept arguing. The Atomic
Energy Commission, under the Atomic Energy Act, had no authority, it said, to do anything. The Public
Health Service, at that time, felt that it did not have jurisdiction within a state boundary. And yet, the States in
the early fifties and mid-fifties had no competence in the nuclear field. Who was to advise them? How was
anybody to get a decision? It took 15 or 17 years to get a decision, and you know who made it? The Department
of Labor under the Walsh-Healy Act. This is the same problem that bothers me now. And it is the problem that
Jim Martin brought up on containment of krypton-85. No one wants a crash program, but I think we do have
an urgent need for a decision.
EISENBUD: Well Dr. Tsivoglou, if you and I ever found ourselves on the same platform back in the
Minnesota days, and I am surprised that we didn't because I thought you and I were on every platform and in
every discussion group during the late 60's. You were then arguing against the right of the Federal
government to preempt health and safety in respect to these matters, had we been on the same platform, and
knowing of your experience, I would have pointed out to you, Ernie, that the reason that mess in the Southwest
developed, which was absolutely inexcusable, was because of a quirk in the law which made it possible for the
AEC to interpret that the State should regulate that particular problem. Since 1936 or 1937, established by
NCRP for a completely different reason, there was a radon standard which may have been arrived at in the
wrong way since we were not very sophisticated, but the recommended value prior to World War II was not too
far off. Had it been adopted by the States, there would be several hundred men alive today who have since died
of lung cancer.
-677-
-------
TSFVOGLOU: That is absolutely correct. Had the standard for occupational protection that was available
been adopted, it would have worked. However, the States had no competence in this field. In 1953 and there-
abouts, there were not a lot of experts who even knew how to calculate simple radioactive decay. Merrill, if I
heard you right, I think I may have been accused of saying I want the Feds to preempt now, although I did not
in Minnesota. I think the Feds need to preempt where the States are not capable. However, I make no apology
for my position in Minnesota, I think the States, many of them, have gotten to the point where they are
entirely competent to regulate in the nuclear field. So that -1 am a States-righter, let us face it.
BOND: I would submit that the parallel drawn by Dr. Tsivoglou is dubious, to say the least. What I mean is
that, with the uranium miners, there was very little, if any, question that real harm was being done. The doses
were high. The miners did have lung cancer. The role of smoking was and still is incompletely evaluated, but
there was no question that those exposures to radon represented an occupational exposure resulting in serious
damage. The problem should have been taken care of. Now contrast that with the krypton problem. As we
have gone into many times during this meeting, the dose from this source is well below even current so-called
"population standards" for the general public. Harm has not been demonstrated. It has been calculated that
at these low doses in mice, the number of animals that would be required to determine an effect would be up in
the billions, so high that one could not possibly conduct the experiment. So that, on the one hand, we are
dealing with a real hazard; on the other hand, we are dealing with a supposed hazard.
HOLADAY: I am going to surprise you Alan (Moghissi) and not discuss the history of the uranium miner
business because my blood pressure can get a lot higher than Ernie's (Tsivoglou). I will just make one
comment. There is enough blame in that whole business to pass around and everybody from the Department
of Labor, down to any State Bureau of Mines, all the way through, can pick up his proper share and there is
still plenty left. What I came up here for was to speak to the question that was asked about the "tailing piles" of
radon which all the panel carefully ducked around. This is one of those things that, looking at it, you can see
no real excuse for not taking care of it. I personally cannot get concerned about the amount of radon released
from a "tailing pile" in Shiprock, New Mexico, compared with what is coming out of a few hundred square
miles around that place. In other words, you can completely stop the radon emission from that source and you
will never notice the difference. The problem of controlling abandoned "tailing piles" from mines is not new. It
has been taken care of, many times, merely on the basis that the stuff blowing around was a nuisance to the
community and the area. The people who owned the mill hauled in dirt and covered it up and stabilized the
thing just on the basis of being a good neighbor. I suspect that, here, the problem is that if they came back and
stabilized them, their lawyers would tell them that somebody will say that they are admitting guilt. The
problem is a simple one to cure. It just costs money.
MORGAN: I would like to comment on some of the remarks of my good friend, Vic Bond. I would interpret
what he said that he is somewhat depreciating the application of the linear hypothesis at low doses and low
dose rates. I am not sure whether Vic was in attendance at the meetings of the Congress of the International
Radiation Protection Association in Washington, a few days back. At that Congress, there were some very
well known and respected persons who presented some reasons why, at low doses and low dose rates,
theoretically at least, one might expect a much greater risk per rem than at higher doses and dose rates. Also,
at the hearings in Aliquippa, Pennsylvania, in response to allegations of Sternglass, that the Shippingport
nuclear power plant had damaged the neighboring population, Dr. Bross brought out some of this research
results of the tri-state studies which indicate that, in a population, there are critical groups such as those with
asthma who seem to suffer far more from the effects of low doses of radiation than the average member of the
population. One might argue that we are not concerned in these statistics with small groups, but this group of
asthmatics is not a small group in a world population. So, I for one, would feel much more comfortable if, in the
next decade, we would think in terms of developing techniques to remove the krypton-85, get rid of these 10
mrem per year to the skin and, in the meantime, hope that we have been overly conservative and protecting
these critical segments of the population.
BOND: I do not know to whom at the IRPA meeting Dr. Morgan was referring to specifically. I have seen
reports of several individuals who do feel that, perhaps, the risk is greater at low doses and dose rates than the
linear hypothesis would indicate. Oftidal, for instance, was one. Others have attempted to repeat that work
and have been unable to. I have seen estimates of this nature derived from the data in Hiroshima and
Nagaskai. These data are heavily weighed by a neutron radiation component. In other words the people were
not exposed to a single radiation and, therefore, conclusions drawn from these data on the effects of low LET
radiations, to my mind, are not valid. With respect to Bross in Buffalo, there are others who verv seriouslv
disagree with his work and with his conclusions. (P. G. Smith, et al., British Medical Journal 97 Mo,, i Q-M
Hollacher, T. C., M. E. J. Med., 287:887, 1972). Bross started with leukemic children and claimed that rt,
relative risk of this disease from in-utero exposure was greater in those with a "marker" (viral or h t 1
infection; allergy) disease. Those who disagree start with children who have a "marker" disease and fid
greater incidence of leukemia in those irradiated in-utero versus those not exposed. The conclusion ia thfth°
markers are associated with leukemia, irradiated or not, and that in-utero exposure is a "red her " T &
summary, I know of no convincing data indicating that the "linear hypothesis" is other than cons6 «
with respect to exposure at low doses and dose rates. servative
-678-
-------
CHIRIWOLU: (Gibbs and Hill) The tendency here, today, seems to be that we should not be bothered by any
release as long as it is not proven to be harmful. I think we should go the other way around. As long as it is not
definitely proven to be harmless, we should try to separate it and contain it till we know it is safe to release it.
STEINDLER: (Argonne) There has been remarkably little practical information that has come out of this
panel discussion. I am wondering if the costs of a krypton retention system, like waste management cost, in
general, are going to be passed on to the guy who turns on the electricity. Can we get an estimate of what
fraction of the power costs you anticipate to be involved in the krypton system?
EISENBUD: I will answer that by saying that I do not think anybody here really knows and I do not think it
really matters. This argument has been used in such a way that you simply pile engineered safeguard upon
engineered safeguard until far beyond the importance to the additional cost of power, is the fact that you end
up with a plant that becomes very difficult to operate. I believe that many of the operating problems of the
present generation of nuclear reactors is because of the application of redundant safeguards. They simply
provide more and more opportunities for something to go wrong.
MOGHISSI: Carl, would you like to volunteer to give an answer if you have any number?
GAMERTSFELDER: Give me a few minutes. I might have a rough number. Based on certain comparisons
with radioactive waste systems for power plants, I made a very rough calculation for krypton removal from
reprocessing plants. The cost would be a very small fraction of a cent per kWh, somewhere in the
neighborhood of 10-4 to 10-3 cent per kWh.
MATUSZEK: (New York Health Department) Unfortunately, what might come to the public, out of this
discussion, is that there should be a major National policy to go to containment, particularly as Jim Martin is
proposing it and Ernie (Tsivoglou) is in general agreement. Unfortunately we are not an island. Unless EPA
plans to build a plastic dome around the United States, this reinforces what Carl Gamertsfelder said, just
briefly, in his comments, that the rest of the world may not agree with us. The result of all this National policy,
including time, effort, research, and money may be an inconsequential reduction of the dose to the world
population.
SNYDER: (Oak Ridge) I was going to raise the same question, but I will try to phrase it slightly different. By
one of the estimates used today, the present reactors going into commission will be operating in the year 2000
or there abouts and, consequently, if this pattern persists, then you are faced with an industry which is saddled
with a policy and this is very difficult to reverse. We have seen it in the case of radon. We have seen it in
the medical profession and so on. In other words, it is more difficult to reverse a policy once it is entrenched or,
at least, I think that is the general course of experience, than it is to take steps in time. Now, the question is how
close are we to that time. And I think perhaps the panel might wish to comment on this point.
EISENBUD: I have a few numbers that I could read. I thought I gave them earlier. It is true Walt (Snyder),
that the reactors we are building today will be operating in the twenty-first century. However, the reactors
that are already built are delivering a per capita dose of approximately 27 fxrad per year. And if we made no
change in technology, by the year 2000, the dose from the thousand reactors that will be operating in the
world, will be 2.7 mrem per year. I have tried to make the point that, in my opinion, we have five years to
formulate a National policy which will give us hardware that will do the job effectively without trade-offs
that may be undesirable.
NELSON: (Environmental Protection Agency) There were some comments made earlier on the trade-off
between the anticipated dose to the radiation workers which compared with the low, but nevertheless, existent
dose to the general population. I would like to query the panel on this subject. The radiation worker is working
voluntarily at his task whereas the general dose is given to anyone who happens to be downwind of the source.
TSIVOGLOU: Of course this is part of what has bothered me for a long time. In our great wisdom, we are
making decisions for all of the general public and what we think they ought to be exposed to, where they do not
understand what is going on in the first place, and where they have no real participation in a decision. I am
afraid of decisions like that. I do not like to decide that the whole public can absorb such and such dose without
knowing what I am doing. And I think Neil Nelson made the point that the radiation worker knows, we hope
he knows, what he is doing, and he is a radiation worker by his own free choice. When it comes to how much
radiation'dose we will subject people to, we are tending not to give people any choice.
BOND: I think both of these considerations work out better in theory than in practice. I question, very much,
that most radiation workers appreciate what the risk is. Sure, an effort is made, particularly recently, to
explain this to them, but I am not sure how much they appreciate it. The other thing is, I am not sure how much
we, in the practical world, really have choices. For instance, if we do not want to drive or ride in a car, we, in
principle, have that option. I would ask anyone to become practical about it and see if you can get along in this
modern day world without driving a car. Again, I think these arguments are fine, in theory, but, in practice,
-679-
-------
they break down. The public, at large, does benefit from power generation, so there is that trade-off. Finally,
why should any group force large expenditures on the public to save them from the hypothetical risk when, if
they were informed and had a choice, they might well prefer to expend the money to reduce the real and serious
risks to which they are exposed.
MOGHISSI: If I may put the last question of the day. I was recently invited to talk in a civic club. One of the
very nice ladies put a question to me which I am going to put to the panel. The question is as follows: The
release of krypton-85 by industrialized nations violates the spirit of the Test Ban Treaty because, and that was
the way she put it, the risk is being given to the total population of the world, while the benefit goes to selected
nations; those who have the capability for fuel reprocessing.
EISENBUD: We have a case in New York, where a power plant is planned 70-80 miles from New York City.
The people there are saying, why should we be inconvenienced to build a plant to provide New York City with
electricity for their subway? It is a very challenging question. The less developed parts of New York get some
benefit from the industry and the commerce in New York City, just as developing countries, are getting the
benefits of U. S. technology.
That lady had a very good question which we are going to hear more and more. Power plants are being
proposed at sites where local people will not benefit from them, particularly, if a public agency is building
them, in which case, they do not even pay taxes.
MOGHISSI: Thank you very much, members of the panel. On behalf of the organizers of this Symposium, I
would like to express to the audience my appreciation for the excellent cooperation. I am particularly
appreciative to the panel members for their frank and clear answers.
-680-
GPO 694-5Z7/8
-------