EMSL-LV-539-7 EMSL-LV-539-7
NOBLE GAS SAMPLING SYSTEM
by
Monitoring Operations Division
Environmental Monitoring and Support Laboratory
U.S. ENVIRONMENTAL PROTECTION AGENCY
Las Vegas, Nevada 89114
MARCH 1977
This work performed under a Memorandum of
Understanding No. AT(26-l)-539
for the
U.S. ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION
-------
DISCLAIMER
This report was prepared as an account of work sponsored by the United
States Government. Neither the United States nor the United States Energy
Research and Development Administration, nor any of their employees, nor any
of their contractors, subcontractors, or their employees, make any warranty.
express or implied, or assume any legal liability or responsibility for the
accuracy, completeness, or usefulness of any information, apparatus, product
or process disclosed, or represent that its use would not infringe privately-
owned rights.
This document is available to the public through the National Technical
Information Service, U.S. Department of Commerce, Springfield, Virginia 22161
PRICE:
PAPER COPY $4.50
MICROFICHE $2.25
-------
EMSL-LV-539-7 EMSL-LV-539-7
NOBLE GAS SAMPLING SYSTEM
by
Vernon E. Andrews
Monitoring Operations Division
Environmental Monitoring and Support Laboratory
U.S. ENVIRONMENTAL PROTECTION AGENCY
Las Vegas, Nevada 89114
MARCH 1977
This work was performed under a Memorandum of
Understanding No. AT(26-l)-539
for the
U.S. ENERGY RESEARCH AND DEVELOPMENT ADMINISTRATION
-------
ABSTRACT
A system to provide continuous monitoring for atmospheric concentrations
of noble gases and tritium has been operated in the Nevada Test Site vicinity
since 1972. The field sampling system was designed to utilize the analytical
capabilities at the Environmental Protection Agency's Environmental Monitoring
and Support Laboratory in Las Vegas. This report describes the noble gas
system which provides sample collection and analysis for radiokrypton, radio-
xenon, and tritium in the form of methane, with detection capabilities, at
the time of count, of about 2 picocuries per cubic metre.
i i i
-------
INTRODUCTION
Perhaps the most difficult radionuclides to monitor in the environment are
those that comprise the nonreactive gases. Releases of particulate and reactive
gas radionuclides from testing of nuclear explosives or operation of nuclear
power plants and fuel reprocessing facilities are normally well controlled. The
nonreactive gases are the most likely to be released to the environment in
measurable amounts. This paper describes a system developed b~ the Environmental
Protection Agency's Environmental Monitoring and Support Laboratory in Las Vegas
(EMSL-LV) to provide continuous monitoring for those radioactive gases. This
system, plus a second system for the collection of atmospheric tritium as water
vapor and hydrogen gas, make up the Noble Gas and Tritium Surveillance Network
located at stations on and around the Nevada Test Site (NTS). Gases presently
monitored by the noble gas sampling system include radiokryptons, radioxenons,
and tritium as methane.
SAMPLER DESIGN
An analysis of the problem indicated that the most convenient way to collect
the nonreactive gases was to collect a whole-air sample which could be separated
for analysis in the laboratory. A laboratory technique had been developed making
it possible to do quantitative analysis for the radionuclides of interest with
0.5 to 1 cubic metre (m3) of air. The method selected to provide the simplest
operation in the field was to compress the air sample for ease of transport.
Because compressed air samples could not be carried by commercial carriers, they
would have to be picked up and carried to the laboratory by members of the off-
site radiological safety group. With the large area to be covered, the frequency
of collection would have to be weekly. Because of a desire to collect a split
sample for backup or duplicate laboratory analysis, it was decided to design for
a collection rate of 2 m3 per week, or about 3.3 cubic centimetres per second at
standard temperature and pressure (STP, 0° C, 760 mm Hg). The air was to be
collected as two samples of 1 m3 each. The basic design called for a primary
collection tank which would be filled steadily at the design flow rate. The
-------
compressor used to pump air into the high pressure tanks would be periodicallY
activated by pressure switches to remove air from the primary tank.
In the initial design, used for about 2 years, a fish aquarium aerator
pump was used to pump air at the low flow rate into the primary tank. When
the pressure reached 28 cm H20, the compressor was activated and evacuated
the air until a vacuum of about 1 cm H20 was reached.
In order to simplify the system, and remove one component subject to
failure, the aerator pump was replaced by a limiting orifice. In the new
design the compressor also serves as a vacuum pump. The differential pressure
switches activate the compressor until the vacuum reaches about 50 cm Hg. Air
bleeds into the primary tank through the limiting orifice, raising the pressure.
At about 36 cm Hg vacuum the compressor is again activated. With the compressor
now in use, the cycle provides for compressor operation for about 1 minute every
15 minutes.
The current design is shown in Figure 1. Air enters the primary tank through
a 0.45-~m pore-size membrane filter and limiting orifice. The pressure di£ferential
switch maintains a vacuum between about 36 and 50 cm Hg. At the altitudes of the
network stations, the vacuum in the primary tank will provide proper operation of
the limiting orifice (P2 ~ 0.53 P1; P2 = absolute pressure in vacuum tank; Pl =
absolute pressure of ambient air). The three-stage Cornelius compressor used was
initially built for aircraft use, with a 28-v d.c. motor. It has been modified
with a belt drive from a 110-v a.c. motor. The compressor passes some oil, so an
oil separator is used to prevent carry-over into the flexible lines and sample
bottles. An in-line filter to remove oil, not shown in Figure 1, which has been
retained from earlier attempts to solve the problem, follows the oil separator.
Since it is not considered essential to the system, it is not included in Figure 1.
A check valve prevents high-pressure air from leaking back through the compressor.
The manifold provides for simultaneous collection of two tanks of air as a split
sample. The sample pressure tanks are fitted with quick-disconnects for ease of
connection to the flexible pressure lines from the manifold. Each tank has a
volume of about 34.4 litres (2100 cubic inches). If the desired 1 m3 of air at
STP is collected in each bottle, the pressure will be about 3.0 megapascals (MPa)
(427 psi). Under the different conditions of air density at the sampling
2
-------
Fuse
Differential
Pressure
Switch
High
Pressure
Cutout
Figure 1. Noble Gas Sampler Schematic
C.B.
D.P.S.T. Double Pole
Single Throw Switch
Differential
Pressure
Switch
C.B. Circuit Breaker
P.L. Panel Light
D.T. Digital Timer
Figure 2. Noble Gas Sampler Electrical Schematic
3
-------
stations, the actual volumes collected are usually somewhat below 1 standard m3,
with pressures less than 2.8 MPa (400 psi). The pressure tanks are periodically
hydrostatically tested to 4.54 MPa (660 psi). In case some defect results in
collecting air at too high a rate, a high pressure cutout prevents the pump
from operating at pressures above 3.25 MPa (470 psi). As a backup safety
factor, a pressure relief valve is set at 3.45 MPa (500 psi). The pressure
of 3.25 MPa permits sampling periods longer than 168 hours, if necessary,
while still providing an adequate safety factor.
Sample volume is determined from the net weight of air collected. The
pressure tanks are evacuated in the laboratory and tare weights are measured.
In the laboratory the full tanks are weighed and the net weight is divided by the
weight of one m3 of air (1293 g) to obtain the volume under standard conditions.
The electrical schematic is shown in Figure 2.
reset to 0 at the start of each sample collection.
if the high pressure cutout is activated, the timer
actual collection time.
The digital timer is
It is connected so that
will stop, showing the
SAMPLER OPERATION
Before connecting the sample pressure tanks, the operator starts the
sampler to pull a vacuum on the primary tank. He then measures the time
required for the vacuum to drop by 2.5 cm (1 inch) Hg. With the 34.4-1itre'
tank, the vacuum drop should take about 6 minutes. A significantly longer
time might indicate a partially obstructed limiting orifice. A shorter time
may indicate a damaged orifice or a leak in the vacuum lines. Without the
pressure tanks connected, the quick-disconnect fittings Qn the flexible lines
retain the pressure generated in the lines during the initial operation of
the compressor. While timing the vacuum drop, the operator also observes the
pressure gauge reading to see that no decrease in pressure occurs. Any drop
would indicate a high pressure leak. After checking the system and correcting
any faults, the pressure is bled off the high pressure lines and the collec-
tion tanks are connected. A typical station is shown in Figure 3. The small
refrigerator at the upper left houses the atmospheric tritium sampler.
4
-------
.\
1'-
11
--
.ilf,..'~
-4
I
r.--
I
...-
. .
r-
\
l .
Figure 3. Typical Noble Gas Sampling Station
---
Beneath the refrigerator is the primary tank for the noble gas sampling system.
The compressor and controls are contained in the case at the center and the two
sample pressure tanks are located under it. An analysis of the results of
the first year of network operation has been reported by Andrews and Wruble. 1
SAMPLE ANALYSIS
The sample collection system was designed to utilize an analytical
system developed at the EMSL-LV. This system (shown in Figure 4) was
described in detail at the Noble Gases Symposium.2
It uses cryogenic and gas-chromatographic techniques to quantitatively
separate the gas fractions of interest. The naturally occurring krypton
and methane fractions, about 1.14 and 1 part per million, respectively,
serve as sufficient carriers for those gases. One ml of stable xenon added
to the pressure tanks before-they are sent to the field serves as the
carrier for that gas.
5
-------
Figure 4. Gas Separation System
..)
n, .
----
\ -- '.
~~ :;r~
11
d
~ . 'i
V 11111 I
\
./!
~~~r
I "\
":,'
. .
!
(.
J I
~ .~
, ~
.~ .'!
. .
I
I
I
/1
n
'1'
\ '
1';-
I,
_I~
/
y-
.~~
~T-:I'
. -
. ~.
Three gas chromatographs in series (Figure 5) form the basic system.
Thermistors T-l, T-2, and T-3, following each of the three chromatographic
sections, provide indications of the passage of the various air sample compon-
ents. Water vapor and carbon dioxide are collected in the molecular sieve trap
on the inlet. Krypton, xenon, and methane are separated from most of the
air sample on the activated charcoal column, C-l. The three gases, plus
some other unwanted gases, are then carried to the second chromatographic
column, MS-l, using helium carrier gas. Argon, oxygen, and most of the
remaining nitrogen are eluted from this column and the sample gases are
transferred to the third chromatographic column, MS-2. Krypton, xenon,
and methane are separated from MS-2 and collected one at a time on column
C-2, filled with activated charcoal. They are then driven from C-2 and
1 Andrews, V. E., and D. T. Wruble. "Noble Gas Surveillance Network,
April 1972 through March 1973." Proceedings of the Noble Gases Symposium,
Las Vegas, Nevaca. September 24-28, 1973. CONF-730915. pp 281-289.
2 Johns, F. B. "Portable Apparatus and Procedure for the Separation of
Krypton, Xenon, and Methane in Air." Proceedings of the Noble Gases Sympo-
sium, Las Vegas, Nevada. September 24-28,1973. CONF-730915. pp 225-238.
6
-------
Pressure
Gauge
Helium
in
Vacuum
Helium
in
Vacuum
Sample in
Vent 1
Vent Vent
2 3 Vent
Vial
Molecular
Sieve
Pre-
Cooler
C-1
MS-1
MS-2
Figure 5. Gas Separation System Schematic
collected in evacuated liquid scintillation vials. The volume of gas recovered
is calculated from the measured pressure, temperature, and known volume of the
vial. Fraction recovered is determined from the volume recovered and the
theoretical volume of krypton and methane in the measured air volume or known
volume of xenon carrier added.
Degassed, toluene-based liquid scintillation cocktail is added to the
vials and they are counted for radioactivity. Repeated counts are obtained
to determine decay rate. This is used to confirm nuclide identification and,
in case of a mixture of krypton or xenon radioisotopes, is used to quantitate
the different isotopes by means of simultaneous equations based on individual
half-lives.
Recoveries by this technique are about 85 percent for krypton, 70 percent
for xenon, and 40 percent for methane. Detection limits, defined as that
activity which gives a 2-sigma counting error of flOO percent, are about
2 picocuries per cubic metre (pCijm3) for each gas.
7
-------
OPERATING EXPERIENCE
Four years of operation have shown the utility of the total system.
During 1975, the total success rate was greater than 95 percent. The reasons
for failure to collect an adequate sample generally relate to the limiting
orifice, compressor faults, or high pressure fitting leaks. Continual checking
of performance in the field and performance of indicated maintenance have
resulted in a steadily increasing success rate.
Limiting orifice problems have been caused by quality control in manufac-
ture and by plugging. All orifices are now tested for compliance with the
desired nominal diameter of 0.15 mm (0.006 inch), then in a measured flow
test. Occasional plugging has occurred, even though filters have always
been used. Whatman 41 cellulose filters were used previously, but seem to
shed fibers which have been found in the orifices. Only membrane filters of
O.45-~m pore size are now used.
The compressor used in the sampler is a three-stage Cornelius compressor
with a capability of 20.7 MPa (3000 psi). The pistons depend on close fits to
the cylinder walls and lubricating oil for a seal. Using the first stage as
a vacuum pump causes the introduction of some of the lubricating oil into the
pressure lines. Pumping inefficiency caused by wear in the final two stages
results in some sample loss. When a noticeable drop in sample pressure occurs
which cannot be explained by low flow into the primary tank or by pressure
fitting leaks, the sampler is changed and the compressor is submitted for
replacement or repair.
Pressure fitting leaks sometime occur due to vibration-induced loosening
of threaded connectors. The connectors have now been replaced with aircraft
type fittings and such loosening is no longer a real problem. The pressure
tanks are connected by means of flexible lines and quick-disconnect fittings.
Compressor oil in the lines caused two problems. The oil tended to collect
sand during sand storms which would work into the fitting when it was removed
and replaced, causing leaks or plugging. The rubber O-rings which provide
the seal in the quick-disconnects were found to be susceptible to damage by
the oil. They occasionally worked loose and plugged the pressure tank or
caused leakage. Installation of an oil separator and use of oil-resistant
O-rings have eliminated those problems.
8
-------
Sample losses in the laboratory are infrequent, but do occur due to
unexpected response of the gas separation system or operator error. The
process requires close attention for extended periods of time. Losses due
to operator errors have generally been eliminated through experience on the
part of the laboratory personnel.
Close cooperation between analytical, field, maintenance, and data
reporting personnel has resulted in a reliable system capable of detecting
and quantitating very low concentrations of gaseous radioactive fission
products. The gradual increase of average 85Kr concentrations in the atmos-
phere has been documented, along with a definition of the variations which
occur in those concentrations.
At the current 85Kr level of about 17 pCi/m3, the 2-sigma counting
error is about 1 pCi/m3. Analysis of split samples has shown that laboratory
analytical errors are of about the same value. The ambient concentration of
radioxenon is essentially O. Most measured concentrations of radioxenon are
well above the detection limit, starting at about 10 pCi/m3. The tritiated
form of methane is rarely detected and may then be due to statistical varia-
tions in counting.
The system, as operated at and around the NTS, requires about 2 man-years
of support. The extended network coverage requires 3 full days each week for
collection. Analysis of 13 samples each week (11 stations, plus one duplicate
sample and one random split sample for quality assurance) requires about 4 days
The remaining 0.6 man-year of effort is spent in maintenance and data analysis
and reporting.
9
-------
1 - 40
41
42
43
44
45
46 - 47
48 - 49
50
51
52
53
54
55 - 56
57
58
59
60
61 - 65
66
67
68
69
70
71
72
73
74
75
DISTRIBUTION
Environmental Monitoring and Support Laboratory - Las Vegas, NV
Mah10n E. Gates, Manager, ERDA/NV, Las Vegas, NV
Troy E. Wade, ERDA/NV, Las Vegas, NV
David G. Jackson, ERDA/NV, Las Vegas, NV
Paul J. Mudra, ERDA/NV, Las Vegas, NV
Elwood M. Douthett, ERDA/NV, Las Vegas, NV
Ernest D. Campbell, ERDA/NV, Las Vegas, NV
Paul B. Dunaway, ERDA/NV, Las Vegas, NV
Roger Ray, ERDA/NV, Las Vegas, NV
Robert W. Taft, ERDA/NV, Las Vegas, NV
Leon Si1verstrom, ERDA/NV, Las Vegas, NV
Robert W. Newman, ERDA/NV, Las Vegas, NV
Bruce W. Church, ERDA/NV, Las Vegas, NV
Technical Library, ERDA/NV, Las Vegas, NV
Chief, NOB/DNA, ERDA/NV, Las Vegas, NV
Hal Hollister, DSSC, ERDA/HQ, Washington, DC
Tommy F. McCraw, DOS, ERDA/HQ, Washington, DC
L. Joe Deal, DOS, ERDA/HQ, Washington, DC
Maj. Gen. Joseph K. Bratton, Asst. Gen. Mgr., DMA, ERDA/HQ, Washington, DC
Gordon F. Facer, DMA, ERDA/HQ, Washington, DC
James L. Liverman, Director, DBER, ERDA/HQ, Washington, DC
Robert L. Watters, DBER, ERDA/HQ, Washington, DC
John S. Kirby-Smith, DBER, ERDA/HQ, Washington, DC
Charles L. Osterberg, DBER, ERDA/HQ, Washington, DC
Robert W. Wood, DBER, ERDA/HQ, Washington, DC
William S. Osburn, Jr., DBER, ERDA/HQ, Washington, DC
Marcie Williamson, HSL/INEL, ERDA/ID, Idaho Falls, 10
Don Bih1, ERDA/ID, Idaho Falls, 10
Steven V. Kaye, Oak Ridge National Laboratory, Oak Ridge, TN
-------
76 Helen Pfuderer, Ecological Science Information Center, Oak Ridge
National Laboratory, Oak Ridge, TN
77 Library Systems Branch (PM-213), EPA/HQ, Washington, DC
78 Albert Printz, Director, Office of Technical Analysis (EN-329),
EPA/HQ, Washington, DC
79 Wilson K. Talley, Asst. Admin. for Research and Development (RD-672),
EPA/HQ, Washington, DC
80 William D. Rowe, Deputy Asst. Admin. for Radiation Programs (AW-458),
EPA/HQ, Washington, DC
81 William A. Mills, Director, Criteria and Standards Division (AU-460),
EPA/HQ, Washington, DC
82 Davis S. Smith, Director, Technology Assessment Division (AW-459),
EPA/HQ, Washington, DC
83 Paul DeFalco, Jr., Deputy Regional Admin., Region IX, EPA,
San Francisco, CA
84 James K. Channell, Regional Radiation Representative, Region IX, EPA,
San Francisco, CA
85 Richard L. Blanchard, Director, Radiochemistry and Nuclear Engineering
Branch, EPA, Cincinnati, OH
86 Charles R. Porter, Director, Eastern Environmental Radiation Facility.
EPA, Montgomery. AL
87 Peter Halpin, Chief, APTIC, EPA, Research Triangle Park, NC
88 Harold F. Mueller, ARLjHSNSO, Las Vegas, NV
89 Gilbert J. Ferber, ARL/NOAA, Silver Spring, MD
90 Kenneth M. Oswald, Manager, Health and Safety. LLL, Mercury, NV
91 Bernard W. Shore, LLL, Livermore, CA
92 Richard L. Wagner, LLL, Livermore, CA
93 Howard W. Tewes, LLL, Livermore, CA
94 Paul L. Phelps, LLL, Livermore, CA
95 Mortimer L. Mendelsohn, LLL, Livermore, CA
96 John C. Hopkins, LASL, Los Alamos, NM
97 Harry S. Jordan, LASL, Los Alamos, NM
98 Lamar J. Johnson, LASL, Los Alamos, NM
99 George E. Tucker, Sandia Laboratories, Albuquerque, NM
100 Carter D. Broyles, Sandia Laboratories, Albuquerque, NM
101 Melvin L. Merritt, Sandia Laboratories, Albuquerque, NM
102 Richard S. Davidson, Battelle Memorian Institute, Columbus, OH
103 Arden E. Bicker, REECo, Mercury, NV
-------
Savino W. Cavender, REECo, Mercury, NV
Auda F. Morrow, RE/CETO, NTS, Mercury, NV
Billy Moore, NTSSO, ERDA/NTS, Mercury, NV
Lloyd P. Smith, President, Desert Research Institute, University of
Nevada, Reno, rN
108 Paul R. Fenske, Desert Research Institute, University of Nevada,
Reno, NV
109 Library, University of Nevada, Las Vegas, NV
110 Thomas P- O'Farrell, Dire~tor, Applied Ecology and Physiology Center,
Desert Research Institute, Boulder City, NV
William S. Twenhofel, U.S. Geological Survey. Denver, CO
Deward W. Efurd, McClellan Central Laboratory. McClellan Air Force
Base, CA
113 Lester L. Skolil, San Diego State University, San Diego, CA
114 - 150 Technical Information Center, ERDA, Oak Ridge, TN (for public availability)
104
105
106
107
111
112
------- |