ML19330C190
| ML19330C190 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 03/05/1980 |
| From: | Wahlen M NEW YORK, STATE OF |
| To: | Jay Collins NRC - TMI-2 OPERATIONS/SUPPORT TASK FORCE |
| Shared Package | |
| ML19330C184 | List: |
| References | |
| NUDOCS 8008080048 | |
| Download: ML19330C190 (37) | |
Text
,
T 1.'
c
(/
. Site phone # at TMI is: '(717) 782-3955 Office:in'Middletown'# is:
(717)' 782-4014 March 5, 1980 Mr. John Collins Three Mile Island Nuclear Power Plant U.S. Nuclear Regulatory Commission P.O. Box 311 Middletown, Pennsylvania 17057
Dear Mr. Collins:
As discussed on the phone we see an unique opportunity to make use of our sensitive measuring capabilities during the possible release of containment air at TMI in the summer. The following is an outline of what-could be done with emphasis on two major points:
1)
Immediate: - Analysis of a sample 'of containment air for the radio-39Ar) and analysis for the I C 85Kr, 3 7Ar (and active noble gases and 311 concentrations and their distribution among the chemical and higher hydrocarbons.
This would.
compounds CO, CO, CII4, C itgs 2
enable you to assess the C inventory of the containment air.
2)' During possible release: Based on the anticipated.85Kr releases 85 our sensitivity would allow us to make measurements of Kr con-contrations in air at 50 miles 'and beyond.
Sets of continuously drawn camples for extended periods of time in a sector at these distances during a clear-cut meteorological situation would enable us to check the validity of presently accepted dispersion models.
All details would have to be planned according to the specific re-lease situation and in close cooperation with a meteorology group.
Wo feel that from such a comparison of a few (2-3) selectied events to calculated predictions one could obtain valuable information in a' unique release situation.
.I am at any time ready to discuss aspects and details. My telephone numbr.r is (518) 474-5719, or FTS564-5719.
Sincerely yours, Martin Wahlen, Ph.D.
Research Scientist IV Radiological Sciences Laboratory cEW/rz' q
l 80'O8.080o $
p.
~
,,,,r_--.
y
o ments and background determinations.
c
~ ~
An additional unecrtainty of 2 5 percent is :ntroduced by the calibration. Detec-tion limits are three times' the back-ground standard deviation.
Reprint Series 8 February 1980, Volume 207, pp. 639-M0 Air samples collected on 30 March showed high
'"Xe concentrations:
3120 160 pCi m-8 at 1500 Eastern
^...
standard time (EST), 3530 180 at y
1900, and 3900 ! 200 in a 10-hour y
sample (from 1545 to 0145 the next morn-ing). By 31 March at 0900 tia concen-Radioact.ive Plume from the Three Mile Island Acc. dent:
tration had fallen to 39 4 pCi m-*.
i 6
Xenon-133 in eir at a Distance of 375 Kilometers samples collected on 3 April at 1600 R
and on 4 April at 1500 contained il !
Abstract. The transit of an air mass containing radioactive gas releasedfrom the 4 and 5 2 pCi m*. respectively.
Three Afile Island reactor was recorded in Albany, New 1*ork, by measuring xenon-Ambient air analyses showed the same 133. These m:asurements provide an evaluation of Three Stile Island efluents to pattern of ' Xe concentrations: no 1
distances greater than 100 kilometers. Two independent techniques identified xenon-counts above background (< 360 pCi b
133 in ambient air at concentrations as high as 3900 picacuries per cubic meter. The m- ) before 29 March, but 1390 t 290 local y-ray whole-body Josefrom the passing radioactivity amounted to o.0tM milli-pCi m'* in a sample spanning 29 March rem. or 0.001 percent of the annual Jose from natural sources.
(l230 EST) to 30 March (l500) and
~
M)60 z 180 pCi m4 in a sample spanning We observed the passage of radio-ments with a thickness of 0.76 g cm-8, 30 March (1530) to 2 April (0830). No ex-active '"Xe, released from the Three which reduced the counting efficiency cess counts were recorded after 2 April.
Mile Island reactor, through the Albany, for '"Xe by about 15 percent. After the The peak concentrations were more than New York, area from 29 March through transit the detector was calibrated under three orders of magnitude higher than 2 April 1979. After the announcement of the same conditions with a virtual point those normally present in ambient air. In the reactor accident and possible re-source of '"Xe. The net count rate in the 1974 the concentration of '22Xe from all e
leases of fission products into the atmo-PI-kev photopeak was measured as a sources including routine releases rom sphere, air samp!es were collected in Al-function of the angular and radial posi-nuclear reactors was 2.6 pCi m-2 for the bany and were analyzed for "'Xe, which tion of the source over the entire field of Albany area (1).
has a half-life of 5.3 days. We also mon-view. Integration yielded an overall effi-The results (plotted in Fig.1) indicate itored "3Xe directly in ambient air ciency of 1.03 x 10" cpm pCi-' m'. The that the air mass containing '2'Xe arrived throughout the entire transit period by air volume effectively seen by the detec-in the Albany area after 1230 on 29 observing the 81-kev y-ray line with a tor was 10.2 m'.
March and before 1500 on 30 March. A planar intrinsic Ge detector located in a Errors reported are 2 standard de-more precise arrival time could not be low-background steel chamber. To our viations from the root-mean-square determined, since the diode measure-knowledge, these measurements pro-counting statistics of sample measure-ments w ere integrated over a 24-hour pe-vided the only evalaation of Three Mile Island effluents at distances greater than 100 km.
3o4 [ 133Xe in Albany, N.Y., air (pci m-3)
E 104-
~,
Gas analyses of I-to 3-m' samples of d
air were performed in two stages: cryo-A Au samples 4
genic and chromatographic separation of m Ambient air Xe, followed by analysis of the #-decay 3[
103 -,_ _ _ _ Average for 10 spectrum (maximum energy, 346 kev) 3 air samples by internal gas-proportional counting in low-background systems (1). Aged com-
[
ty O W6 f,"tr
[
pressed-air samples were processed db N' through the gas separation system as go2 [ Estimated release rates of 102 133Xe (Ci sec-1) blanks between Albany air samples. A low residual activity found after the pro-f a
5 L
cessing of the higher-activity samples did F
not substantially reduce the sensitivity.
I
~
Ambient laboratory air was monitored 101 h-
.[
[
got Di by an intrinsic Ge diode with an area of E
/
a 30 1
3 500 mm' and a resolution of 63') eV (full
/
M 8'c h Apr
width at half-maximum at 81 kev). This j) 7
, NJ A
instnament was in a low-background steel chamber (3.3 m square; 2.4 m high; 24 2s t
5 9
M *'c h AP "'
wall thickness.14.5 cm) in which outside air was exchanged about ten times an Fig.1. Xenon.133 activity (picocuries per cubic meter of air)in Albany, New York. for the end of March and early April 1979. The lowei trace shows the time-averaged estimates of rete ases hour. During the entire period when (curies per second) from the Three Mile Island reactor (1). The inset shows detailed value for "2Xe was recorded, the thin window was air samples (gas counting) and concurrent average vahies for ambient air (Ge diode). Abbievi-j covered by a pressed pellet oflake sedi-ation: LT, less than.
f 0036-807180 0208439500.M0 Copynght C t980 AAAS y
1 -
riod. The trailing edge is more sharply kPa level did, howev:r, indic te e plume amounted to 12.6 pCi m-8, which' is' defmed from the measurement of the air passage in the Albany rea on 29 Alarch.
within the range of 10.9 to 18.4 pCi m-2 samples. The time behavior show n in the in summary, the meteorology indi-encountered in samples of Albany air for inset of Fig. I from the widely spaced air cates that air arrising at Albany on 29 the period 1975 to 1979.
j samples must closely represent the ac-hlarch contained radioactive gas re-In conclusion, the elevated "3Xe con-
.j tual trailing edge of the passing air mass.
leased from the Three hiite Island reac-centrations observed in Albany on 29 i
This is evident from a comparison of the tor on 28 hlarch, which had been dis-and 30 51 arch 1979 could be attributed to y
calculated time-averaged activity of 780 persed rather widely around the point of releases from the Three hiile Island reac-
.]
pCi m-2 for this period from air samples origin and then moved northeastward at tor accident. The "3Xe concentrations
]
and the observed averaged activity of low levels. The most probable transit ncrmally present in Albany air due to 1060 pCi m-2 from the diode measure-time appears to have been 18 to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> routine releases from nuclear reactors f
ment. The observed peak value for the for an approximate actuat travel distance are lower by more than thrce orders of 3
"dXe activi'. :t ground level was 3900 of about 500 km.
magnitude (/). The dose received from T
pCi m-' for the air sample taken late on The whole-body dose to an individual the passing radioactivity was found to be 1
30 h1 arch. Peak concentrations of "3Xe in the Albany area from exposure to y-extremely small when compared to the for the period before 30 h1 arch could rays and x-rays from the passing "3Xe, dose from natural sources.
have been higher, depending on the calculated from the average activity val-h1ARrtN WAllLFN 1
actual arrival time c' the air contain-ues, was 0.004 mrem (4). This is about CilARt ES O KUNZ, JOHN hl. h1AT USZE K 3
ing "'Xe. The average value from the 0.004 percent of the ann.ial whole-body WILLIAM E. h1AHoNEY v
diode measurement for the 24. hour in-dose from natural sources.
ROGER C. TsioursoN j
terval prer etiii a 30 N1 arch was 1390 A search for airborne "'I showed no Divisi,m oflaboratories am/ Research.
]
pCi m-'.
measurable activity, even though the air New Forts State Department of#calth, d
To describe the air mass transport in was analyzed by a highly sensitive #!y Albany /2201 il more detail, we collated the available coincidence cc.inting method t3). Al-d data on release rr.tes from the reactor bany air processed through a charcoal References and Notes i
and the regional meteorologic.d condi-cartridge impregnated with triethylenedi-
- 1. QK asrs j in F7 15 pp j
tions. Nicasured release rates for "'Xe amine during the 24-hour period of high-ibia, pp. 239-24ti; C. O. Kunz and C. J. Pape-3 from the reactor were not available. Av-est "'Xe activity did not contain "'I at or 2, ('Uyi[7$,"%,'[i, e u at$y Commission.
2 t:
l erage release rates were estimated in-above 8 x 10" pCi m-3 Washmgton, D.C., personal commumeation.
I directly by the Nuclear Regulatory Com.
Nor did we observe a measurable in-se$h a r$to r s a n
M mission, using thermoluminescence do-crease of"5Kr in air. This is not surpris.
Occame and Atmospheric Adminhtration. Sii-ver Sprmg. MJ., provided us with forward and 4
simeters in the v:...cimty of the reactor (2).
ing, considering the long half life of "51s.r backward traiectory calculat ons for the entire
.1 These rates are plotted in Fig.1. N
.e-(10 ; ears), the lower fission yield, and 4 p'dyfialp,8 3
y
,,,yc,,,,,,,,,,y,,,,,,,,,
{
leases occurred before 0400 EST on 28 the sizable atmospheric background con-Guide t.4:Anampimas UscJfor naluatine the j
hlarch.
centration from aunospheric weapon
,7'["'lfj,#$$l$"/O,",Z",'y','f(,"jf,'
]
Regional meteorological conditions testing and routine releases by the nucle-a<fors trevision. June 1974L 3
were examined by using forward (from ar industry. The "3Kr measured in the
- I;,,#;. If'[*3 7 $Ni75)
H hiiddletown, Penns>lv' la) and back-sample from 30 hlarch (1500 EST) 7 August 1979; reshed 19 November 1979 1
ward (from Albany) air..ajectories pro-vided by the Air Resources Laboratory t
of the National Oceanic and Atmospher-4 ic Administration (3). The backward tra-jectories were calculated for a mean transport layer between 300 and 1500 m above the terrain. Forward trajectories were calculated for the same mean trans-port layer and also for transport at heights corresponding to 95,90, and 85 kPa (102 kPa = I bar).
For the first release period on 28 hlarch, the meteorological conditions at hliddletown were rather stagnant, with medium to low-speed winds gradually
]
shifting from northwesterly to north-4 easterly to easterly and finally to south-j casterly. From 29 to 31 Alarch, south-d westerly winds prevailed at increased 1
speed. The mean transport layer forw ard 3
trajectories for this period passed 80 to 160 km south of Albany. Backward tra-
,)
jectories for 29 to 31 Alarch show that
~,
the mean transport through Albany origi-
]
nated predominantly from regions to the west and northwest of liarrisburg, Penn-d.
sylvania. Forward transport at the 95-
-.-Be'4>.-
-3.er h w.
F _w y
A
' Xenon-133: Ambient Activity from Nudear P wcr Stations activat d charcoal. Mass spectromitric analysis has shown that the separated Abstract. The average activity ofxenon-133 within and at approximately. 00 ki-fractions contained less than I percent lometersfrom Albany, New York,from April to July 1975 was 2.6 picocuries p *r cu.
N, and He impurities. The separated bic meter of air. The source was gaseous eBluentsfrom boiling water reactws to-fractions are loaded into gas-proportion-cated in the nor? eastern United States. Its 5.29-day half-Hfe makes xenon-13. an al detectors, the specific volumes are appropriate isotop e to observefor the study of regional and hemispheric disperson measured, and the recovery fraction of
. ofpollutants.
cach is determined from the known car-rier volume of xenon added or from the Although "Xe is an important pas-tem permitted collection of grab samples known abundance of krypton in air. Re-cous radioactive fission product released at any desired location.
covery fractions averaged 70 percent for in nuclear weapons tests and in the nor.
Air samples of I m8 were collected each.
mal operation of the nuclear fuel cycle,it with a portable air compressor and 15-li-The counting system for **Kr has been has usually been considered a local prob-ter stainless steel vessels containing I to described (2). The counting system for lem. With few exceptions, measure-5 cm of stable xenon carrier. De kryp- "Xe is a beta-gamma coincidence sys-8 ments of "Xe have been attempted pri-ton and xenon fractions were separated temwitha45-cm gas-proportionaldetec-8 marily at the site boundaries of nuclear from the samples by cryogenic adsorp-tor with thin aluminum walls and a 7.6 by facilities. During 1964 and 1%5, how. tion and gas chromatography, as follows. 17.8 cm Nal(TI) well detector. It in-ever, Scholch et al. (1) measured the ac-The air sample in the high-pressure ves-cludes a plastic scintillation anticoinci-tivity of "Xe in atmospheric air samples sel was leaked at a reduced pressure of dence cosmic-ray guard, with all com-taken from a location in West Germany. approximately 300 torr through a system ponents enclosed in a 15.2-cm-thick steel Hey observed an average activity of containing three traps in meries. De first shield. The purified xenon sample is about 0.1 pc/m*. The source for this ac-trap, empty and at Dry Ice temperature, mixed with P-10 counting gas and load.
tivity was not evident. How ever, a maxi. removed water vapor. The second trap, ed. The counting rates in the *Cs x-ray mum was obtained in June 1%5, possibly filled with glass beads and held at liquid region and the 81-key gamma-ray region attributable to a Chinese nuclear bomb nitrogen temperature, removed CDs. of the photon spectrum are used to qucn-test.
The third trap, a column 1.5 m long and tify the "Xe. In the x-ray region the in 1975 we undertook to measure the 1.25 cm in diameter filled with activated background is 0.018 count / min, and the average ambient background of SXe in charcoal and maintained at liquid nitro-efficiency is 0.26 count / min per dis-northeastern New York State and to de-gen temperature, retained the xenon and integration per minute. Under the usual termine the source of this activity. From krypton. After the sample had been conditions of recovery and decay and for April to July we collected samples at passed through, the charcoal-filled trap a 1.m sample volume, the dete'ction lim-8 various locations within and at approxi-was warmed to 15'C, under vacuum, to it, defined as 3 standard deviations over mately 100 km from Albany (Fig.1). The remove most of the adsorbed nitrogen, background (3),is 0.05 pc/m.
2 pattern was designed to ensure that the oxygen, and argon, it was then heated to Intermingled with the whole-air "n-activity we were measuring was not from 200*C to drive off xenon and krypton, ples,13 system blanks consisting of com-a local source, such as a hospital or labo-which were collected on a small molecu-pressed, aged air were also processed.
ratory using "Xe. but was an ambient lar-sieve trap at liquid nitrogen temper-The results showed that the observed ac.
activity for this section of the state. Most ature.
tivities were not an artifact of the sys-of the sampling locations were 100 km or ne krypton and xenon fractions are tem. Spectral shapes and half-life mea-more from the nearest nuclear power re-chromatographically separated and puri-surements confirmed that the observed actor releasing "Xe. Whereas samples fied by using a column 4.6 m long and activity in the xenon fraction was due to in the West German study were obtained 0.63 cm in diameter filled with 94 percent
- 2'Xe. All sample spectra had the same from an air liquefaction plant, our sys-type-5A molecular sieve and 6 percent shape as that of a National Bureau of Fig. 1. Sampling area f* *Ye survey.
Table 1. Activity levels of mXe and "Kr around Albany, New York, Marked sites are (*) sampling locations, April to July 1975.
(W) boiling water reactors, and (c) pressurized Date Activity (pc/m')
water reactors.
col-1 ocation lected "Xe "Kr 2 April Albany 0.5220.09 2 April Albany 0.7520.11 16.721.5 29 Apnl Albany 4.4 2 B3 16.221.3 29 April Albany 3.4 20.3 16.320.9 soo 6 vr la May Albany 3.9 20.2 17.821.5 h g.q; 18 May Albany 4.0 20.2 18.421.7 NH 4
27 May Little Falls 1.3220.08 14.3 1.4 k (NY-
'NNs 18 June Greenfield Mass.
0.4220.05 17.221.8 e
4 June Kingston 3182 0.15 14.821.2 ss 11 June Albany 7.5 20.2 14.821.2
\\
y e,
23 June Albany 1.2920.13 16.521.0 az/
24 June Albany 2.9 20.3 16.521.1 PtNN
- h 25 June Albany 0.32 0.07 16.1 2 1.2 e
7 July Lake George 2.3720.16 18.421.0 Average 2.6 16.5 Repeinted from SCIENCE,18 June 1976, volume 192,pages 1235-1237'
w v
~.
~
3
~
%?
g
.6 g.,
yv t.
s
^
[Stanuards' '"Xe standard, and the h:lf-L 1culity an 3virage - expectedi value of - ing heights which varies from' about 200
? lives.detrmined for sev:ral of the sam-
- '"Xe from the six boiling water reactors 1 to 4000 m, depending on location, weath-4 pies agreed with the 3.29-day half-life of shown in Fig.1. _We used their phase 2 er lconditioris, and season (7).- Vertical --
?"Xe. Samples' collected in. duplicate -
i
- equat on, which estimat:s concentrations ~ : dispersion above the mean nixing height
- Lthen ;rocessed and counted several days,. in the region at distances representing a' and beyond the temperate latitudes pro-4 n
i apart,were alsoin agreement.
e few hours to several days' travel from the j ceeds more slowly, requiring I to 2 years s
S An average ambient activity of 24 pc/ c source,~and assumed that the wk.d free for dispersion throughout the earth's at.
- m* was obtained for '"Xe (Table I) The
- quency is equalin all directions. Ir view of. mosphere.
Lvariation of the "Xe activities measured the geographic distribution of the sources -. The half-life of 5.29 days for "Xe is 8
8 1
- for all locations is consistent with values ' clative to the sampling sites, we believel such that dispersion on a regional and
~
' rnticipated y for regional isources,. in-that any errors inherent in this assurnption hemispheric scale can be followed. Iso. -
' dicating that the levels obtained are not twould tend to cancel out. We further as-. topes with significantly longer half lives,
~ due to a local source but represent the : -? sumed that each boiling water reactor re-such as "?Kr (10.76 years), accumulate l ambient actisity for this section of the : leases !"Xe at a rate of 6.4 x 10-* c/sce throughout the atmosphere, making re-Northeast. The observed average value.-
per megawatt (electric) capacity (6) and gional dispersion measurements diSt.
of 16.5, pc/m for ?'Kr (Table 1).is in that each plant operates at 80 percent ca-cult. Further measurements of ambient 2
- agreement with the values measured by - pacity We also corrected fordecay during -
'"Xe could be correlated with more de-the L Environmental Radiation ~ Ambient. -diffusion.! The: calculated value of 3.3 tailed information on the amount of '"Xe -
- Monitoring System (4).
~. observed average of 2.6 pc/m, especially sources and on the climatologic and geo-pc/m' is-in good ' agreement ~with the being ' discharged - from the various The observed levels of '"Xe represent -
8 no significant health hazard at the pres 1 allowing for the approximations used in graphic parameters.- Such an approach ent time.' However, the genetically sig, the calculation.'
would provide a more comprehensive inificant dose from '"Xe is about three Models for estimating concentrations model for estimating regional and hemi-times greater than that from ambient' over a short period, such as a few hours, spheric dispersion of all airborne pollu-
"Kr.
would require wind data monitored at a tants, both radioactive J and non-
~ Nuclear. power reactors are apparently number of stations to determine wind tra-radioactive.
the source of the '"Xe activ4y.The inter-jectories, combined with measurements C.O.KUNZ action of cosmic rays with stable xenon of the release rate of '"Xe from all C. J. PAPERIr LLO cannot account for the observed levels, sources. This informaticn was not avail-Dirlslon oflahorntories andResearch, and there were no atmospheric tests for able. The activity levels for 23,24, and New York State Department of#calth, nuclear weapons during the 4 months - 25 June, however, do show ths influence Albany 12201 when our measurements were made.The of weather patterns. On 23 and 24 June Soyiet Union " conducted' two under-the Albany area was under an inversion ground-tests on 28: April arid 10 June layer that had originated to the south-3 J sg."w. stich, K. O. Munnich,
"'*"***dN""
'1975. However, our findings before and east. During the night of 24 June a g
after these tests show no indication that - ccid front moved through the area.
- 2. c. L Parevicio, paper presented at the NoNe gaycous venting following these tests.
On 25 June the area was under the
{*,',",syguin, t.a Vesas. Nev. 24 to 28 -
made a significant contribution of '"Xe
- influence of a high-pressure system L IUPAC Commission.Spectror4emicoland 0sh.
to the ambient level in the Northeastern which had moved in from the north-
",, # ",',*[,[,'M,'l,',,I,,d"/3'l,;,$,'"$,'".
f, United States. The locations of the oper- ~ west. The activity level decreased from aaJ stadards tininnanonni uman or rure and ating nuclear power reactors within a ra-2.9 to 0.32 pc/m* after the weather
,Do$""'d* ***"""' "C" '"23'
^
dius of about 500 km from the sampling. front moved through.
- 4. oake or aalianon Programs. Environmen.
tal Protection Agency, wadunston, D.C., pri-o area are included in F.ig.1. Su.n a boiling Almost all the nuclear reactors cur-vare conununicanon.
w'ater reactors of current design release - rently in operation' are located 'in the 3 h,M,*$'*4,%',%/;I;-Qf;i"Ql'/;
0 about two or three orders of magnitude : middle latitudes of the Northern Hemi-the Atmosphere (Internehonal Atomic Enersy more f"Xe than the pressurized water sphere.. Dispersion of the effluent from
- 6. D r N d*b'[$ w m w m reactors, they are evidently.the pret these reactors is fairly rapid latitudinally lated facaban." usa c Rep. w4sn-mo dominant source of activity.
and also vertically up to the mean mixing
- 7. Nu"oiDh, Mon. weather me,. so. 23$ '
i J A regional dispersion model proposed - height. It takes about 30 days (5) for the -,, @,,,, g y g
by Machta et al. (5) for estimating long-effluent to travel around the carth, but in vice and r. w. Mitier for technical asini ree.
term air concentrations was used to cal-a matter of hours it reaches its mean mix-12 January 1976: revised 7 April 1976 E
['*
Copyright @ l976 by the American Association for the Advancement of Science j
k 1
N 4
l:
h
~
4 L
~
4 d
r
_~;,
y _y.
1
~---r M
F-
^
x
,Y
_}r TrInihetico cf thf American Nucitar S;citty,1978 Annuti Meating, 18-22' June 1978, San Diego, CA. 128:74-75.
Copyright c 1978, American Nuclear Society, Inc.
The experimental procedures have been described.'**
The radioactive species in the sample are individually counted in gas proportional counters after chrornato-graphic separation.
At all six B%Rs, over 95% of the "C activity in off-gas samples was in the form' of "CO, as would be expected 2
from the oxidizing atmosphere within the reactor vessel of a BWR. Since the off-gas is continuously released at a rate proportional to the power level of the reactor, we derived estimates for "C release rates by using the measured specific activities and the actual or assumed condenser off-gas flow rates (Table 1). At the reactor A, which has been repetitively sampled over a period of 3 years, the average release rate was found to be 12.4 C1/GW(e)-yr. Release rates calculated from in-dividual samples from~this reactor were within a factor of 2 of the long-term average. At this writing, the other five BWRs have been sampled only once.
Froin the extensive data for reactor A and from the results for three of the other B%Rs, the release esti-mates appear to be in reasonable agreement with the rates calculated for the production of "C in the coolant of a B%R. The calculated rates are based primarily on "O activation and vary from 9.2 to 11.5 C1/GW(e)-yr.
The estimated releases for the two remaining BWRs demonstrate that substantially higher production rates of "C are possible. We attribute these to excessive nitrogen 7 14C Discharge from Boiling Water Reactors, in the coolant. ^ An equivalent am unt at 'C would be M. Wahlen, C O. Kun: (NY State Dept of produced from an Na concentration of about 25 ppm.
Health)
Carbon-14 and radioactive noble gases in stack-gas Carbon-14 is an activation product present in light-samples were generally found to be diluted according to water reactors, originating from neutron reactions with the ratio of the release volumes.Or off-gas and stack isotopes -of oxygen and nitrogen ("O,"N) in both the gas, Indicating that the coolant is the major source term coolant and the fuel. Gaseous "C released at the reactor for "C gaseous release.
appears to be predominantly formed in the coolant.
W d
"C e To assess this "C production and release, we deter-leases from B%R, we compared the resulting "C mined (a) the chemical form and specific activity of activity levels in the vicinity of a reactor to the present-
'various "C species in off-gas and stack-gas samples day "C actinty : tropospheric air. The current ambient from a series of BWRs and PWRs, and (b) the "C activity level of ~ 1.5 x 10 Ci/m' air originates from cosmici.
in the coolant and in a variety of liquid and solid reactor ray prw cction (1.06 x 10*" C1/m' air) and residual "C i
wastes. In this paper we discuss the results obtained for from hospheric testing of nuclear weapons (+40%).
six B%Rs. Results and assessments of nine P%Bs are Using a representative measured value of 1.0 x 10** for now being completed, the dispersion at 5 km from the stack of the reactor A" TABLEI "C Activities in B%71 Off-Gas Power level Estimated C
' Maximum
' at time of g
.p Reactor power C in off-gas sampling Off-gas rate IMW(e))
(uci/cm )
(% max)
(cfm)
(Ci/yr)"
[C1/GW(e)-yr) 3
-6 A
~620'
.5.1-9.8 x 10 70-100 80-105 7.7 12.4
-6 B.
520 3.6 x 10 97 52 -
2.9 5.5 C
650
'2.3 x 10"'
69 96 4.8 7.3
-6 D
820 8.'5 x 10 61 125 26 32
-5 D
E 650 4.2'x.10 100 65 41 63 3.1. - 5. 6 -
-5 b
7 0-JL d4 3 GU b
a6 M=
F 310 4.4.x 10 100
'" For operation at maximum power kAssumed g _2 ( g, 9 g r.
1
=
. Environmental Surveillance and Modeling 75 and an annual release of 12.4 Ci of "C, we calculated an
- activity. concentration of 3.9 x 10* C1/m' air at that distance. This corresponds to an increase of 0.4% above the natural level. At I k.a.from the stack, these values would be about an order of magnitude higher.
. Carbon-14 activity levels in the coolant, evaporator concentrates, spent resins, and filter materials varied considerably. Carbon-14 is present in liquid and solid wastes held for eventual discharge to burial sites; however, it is very difficult to relate these activities to quantitative release estimates, which will require more
-- extensive sampling and close cooperation with the reactor operators. Preliminary estimates show that the amount of."C discharged as liquid and solid waste is less than that in gaseous effluents, in conclusion, the release estimates for gaseous "C obtained from measurements on four Butts are com-patible with the calculated production from "O in the calant. The results from two B%Tts show that con-siderably higher discharges *are possible, originating
- most likely from activation' of excessive ' "N in the coolant.
- 1. C. O.' KUNZ, W. E. MAIIONEY, and T. W. MIL' LER, Proc. Health Phys. Soc. 8th Midyear Symp., Knox-ville, TN, CONF-741018, pp. 229-234 (1974).
- 2. C. O. KUNZ, W. E. MAllONEY, and,T. W. MILLER, Trans. Am. Nucl. Soc., 21, 91 (1975).
- 3. C. O. KUNZ, Proc. Noble Gas Symp., Las Vegas, NV, pp. 209-217 (1973).
- 4. C. J. PAPER!ELIO, Proc. Noble Cas' Symp., Las Vegas, NV, pp. 239-248 (1973).
- 5. H. BONKA, K. BRUSSERMANN, and G. SCHWARZ, Reaktortagung, Ber lin (1974) '
-_6. P. J. MAGNO, C. B. NELSON, and W. H. ELLETT, 13th AEC Air Cleaning Conf., San Francisco (1974).
- 7. D. W. HAYES and K. W. MacMURDO, Health Phys.,
- 32, 215 (1977).
- 8. 'G N.' KELLY,' J. A.' JONES,. P. M. BRYANT, and F. MORLEY, Doc. V/2676/75, Commission of the
. European Communities, Luxembourg (1975),
m9. - T. W. FOWLER, R. L. CLARK, J. M. BRUH LKE, and J. L. RUSSELL, EPA Technical Note, ORP/ TAD-76-3 (1976).
- 10. Niagara Mohawk Power Corporation. Preliminary
' Hazards Summary Rept.. Nine Mile Point, Vol. II, Appendices (1974).
' ima'r*
w..h-d p.ggw
.m E +,,
,gmog,-r--
,.wa,,m,s
..w-ag
^ - - - - -
l*
Transactions of the American Nuclear Society,1978 Winter Meeting, 12-16 November 1978, Washington,' DC.
30:113-114.
Copyright @ l978,'American Nucicar Society,.Inc.
3.
"C Activity and Distribution in Gaseous Effluents - from Pressurized. Water Reactors, M Wahlen, C O Kuni (NY State Dept of Health)
Production calcu'ations predict that **C is formed in the coolant of a PWR at a rate of 3.3 to 11.1 Ci/GW(e)yr
- (Ref. 1). Must
- C activity has been presumed _to be released in gaseous form.
We have measured C activity and its distribution among carbon compounds in samples of decay-tank gas, containment air, and ventilation air obtained from nine PWRs. The concentrations of radioactive noble gases and tritium in these samples were also measured.' From these data we have derived preliminary release estimates for
- C.
The experimental techniques (gas-chromato-graphic separation and internal gas-proportional count-ing) have been described earlier.* The results are summarized in Table I.
TABLEI Total **C Activities and Dis ritutions Among Cartm Compounds in Gas Samples from Nine PWRs
. Total 1*C activity g4 34 "4 C-hydrocarbons, 34 CO C
(6CL/cm )
2 Decay Cont Vent g
g gg PWR x 10~4 x 10' x 10'I Decay cont vent Decay Cont Venc Decay cont Vent A
2.3 2.0 10 70 19 C
3.5 7.9
<2 9
20 80 80 10 E
2.7 27
<0.6 11 3
74 94 15 3
F 6.7 0.17 9
63 62 28 38 C2 0.29 4
68 28 81 2.0 2.3 5.9 27 7
13 37 65 58 33 29 20 B2 43 3
50 47 B3
'1.6 28 45 19 D
25 8.6 2.2 27 9
18 55 91 82 18 Decay: Decay tank gas-Conte Containment air Vent Ventilation air "C "6+ C H38*
4"10 2
~.
1 2
H4 '
Environmental Sciences - General m
'Ihe'distributinn of "C among the carbon compounds in.
- 3. C. J. PAPERIELID, Proc. Noble Gases Symp., Las PWR decay-tank gas suggests that the reactors fallinto ;
_ Vegas, NV, pp. 239-248 (1973).
.two groups.;For the first group,89 to 96~e of the activity is fouad as hydrucarhons< These compounds are expected
- 4. C. T.'YONGUE, Personal Communication.
from the recNeicg conditions in the primary coolant (H, cover gas). Jrt the second group, almost 30% of the "C is,
present as CO,. most likely formed by oxidation of the -
hydrocarbo** in. the decay tanks due to high radiation,
levels ani excessive air inleakage from the vent header.'
' For containment. air samples the "C activity of all nine reactors is predominantly in. hydrocarbons, dis-tributed much like the first group of decay-tank sarnples.
- The low "CO, concentrations in containment air support
.the assumption that "CO: is formed in the decay tanks
- and that the "C activity in contatnment is not likely
- to have been produced by activation of "N under the oxidizing conditions of containment air. It apparently originates from gases which are formed in the coolant under reducing conditions and then leak into the contain-ment vessel. We estimate "C release rates of 0.3 to 1.9 Ci/GW(e)yr from containment purges, based on measured specific activities, containment volumes, and periods of Isolation prior to sampling.
We have attempted to estimate total gaseous "C activity releases by. appl >1ng the measured ratio of "C/"Kr in the decay-tank samples to the reported "Kr release rate from all sources, averaged over the period
- of sampling. This approach assumes that the coolant is the common source for gaseous. "C, which appears reasonable.. giten the similarity of "C/"Kr ratios in
- containment alr. ventilation air, and decay-tank gas of
- any one PWR. Unfortunately this approach relies on 1 reported "Kr releases, which can vary considerably with time acecrding to fuel-rod integrity and decay-tsnk gas release practices.
Since production of "C is likely to be proportional to
' the reactor power level, representative release estimates
, can be obtained by this method only for reactors with '
- reasonably constant "Kr-releases. This is the case for PWR A, BI-3, and D, with estimated "C releases of 9.8, 6.E. and 9.9 Ci'GW(e)yr, respectively, using reported "Kr releases for the 3 months prior to and 3 months after the -
sampling date. - At reactor E most of the "Kr is released
. from decay tanks, which are vented about once a year.
The "C' release was estimated at 7.5 Ci/GW(e)yr, using a 4-yr average of the Kr releases from these tanks. For the remaining reactors, C and F, highly variable "Kr releases are reported, so the "C release estimates also ranged widely (1 to 36 ana 5 to 134 Oi/GW(e)yr, respec-tively, based on slightly different averaging periods].
' At most reactors only one set of samples was ob-tained. However, the estimate of 9.8 CL/GW(elyr for
' PWR A is an average of six sets of sarnples taken over a.
' period of 3 years. The release estimates for individual samples ranged from 3.8 to 23.3 Ci/GW(e)yr.
' Apart from = variations in "Kr~ release rates, the
- monitoring' of "Kr ' releases may not be sufficiently accurate to measure all the "Kr discharged. 'MC retcases
'can be ' determined more accurately by employing con-tinuous stack gas and ventilation-air samplers for CO t
l and hydrocarbons over: extended periods, including all
-} phases of the reactor operating cycle.
- 1. T. W. FOWLER, R. L.- CLARK J. &f. BRUHLKE, and i
- J. L. RUSSELL, ORP/ TAD-76-3, EPA Technical Note
"(1976).
- 2.1 C. O.. KUNZ,' Proc. Noble : Cases Symp., Las Vegas,.
NV, pp. 309-217 (1973).
y4 '
t
_s__
ha
g 3
L
~
SEPAH ATIONTECllNIQUES 1 Olt HEACTOlt-PRODUCED NOl!LE G ASES*
C.O.Kunz.
- ,+
- Radiolog;ealSciences1.nhoratory
- liivision of Laboratories and Research
. -i
. New York state l)cpartment of ilealth 1
1 Albany.NewYork 12201 Abstract Proceduresfor separating the permanent gases have been deccioped as part of a study to characterire the gaseous radioactive effluents released from nuclear facilities. The gases being separated for internal gas-proportion alcoun ting include A r, Kr, Xe. Ily. Cist. and cog. Wa ter vapor is cryogenically sepa rated for liepdd scin tillation counting, Samples taken forprocessing seithin each facility range from 0.1 mi to severalliters in volume. Sample volumes less than 10 mi a re sepa ra ted directly by ch roma tographic methodL I.arger samples are processed using cryogenic-adsorption techniques for rough separatium followed by chromatographic purif; cation. Procedures for preventing cross contamination from sample to sample, and betuccen different radioactivegases within a sample, are considered. Processing requiremen ts imposed by gas composition are alsodiscussed, INTRODUCTION g
E The Radiological Sciences Laboratory of the New York State Department of Ilealth is studying the gaseous radioactive effluents from nuclear facilities (Matuszek, et al.,197:1). Samples have been obtained from two pressurized water reactors (PWR), a boiling water reactor (HWR), a high-temperature gas-cooled reactor (IITGR), and a pressurized heavy water-moderated research reactor. The samples, taken from a variety of locations at each reactor. include primary strip gas, cover gas, decay tank gas, containment air from the PWRs.and stack gas from the BWR.
Pr'edures fnr 9 paratine the noble ver*. in addition to other pornm*+o' re. have be en de t e 'oped ad
- a
!% w;+: T?.+ ::w, arre ntly %r.;t er,unr,< :v wpur:aon. vin :r Ar. Kr. X<. IIz. ! If. aw: t !o g W t
r
' activity is measured using irFernal gas-proportional beta spectrometry (Paperiello,197 b. Water vapor is cryogenicallyseparated forliquid scintillation counting.
The range of specific activity from sample to sample has hean greater than five erders of magnitude, depending on the reactor and sampling locations. We have measured activities of various nuclides in individual samples that differ in activity by approximately seven orders of magnitude. Consequently precautions have been taken to prevent cross-contamination from sample to sample, and between different radioactive gases within a sample. Sample aliquots from about 0.1 ml to several liters have been processed, depending on the nctivity of the gases being analyzed.
Sample volumes les< than It) ml are separated directly by chromatographic methods. I.arger samples are processed using cryogenic adsorption techniques for rough separation fediowe,I by chromatographic purification.
S AM PLE COLLECTION AN D M ASS S PECTROM ETRIC ANA LYSIS Samples are collected in a variety of vessels ranging in volume from 14 mi to Ifi liters. Since vessels with septum caps tend to icak. those with stopcocks or valves are preferable.
Upon receipt ora sample, an aliquot is rounted on a Ge(Li) sp"ctrometry system to measure the activity of gamma-emitting gaseous radionuclides. These results are used as an aid in determining the subsequent separation procedures, and a re com pared with the results obtained by proportional counting.-
An aliquot of the sampleis also taken for mass spectrometric analysis to determine the composition of the gas. A fewexamples ofsuch analyses of gases sampied from various locations in various types of reactors are shown in Table 1. The composition, which can vary considerably from sample, to sample, influences the choice of a separation procedure. In addition, species such as hydrocarbons that may contain "C or tritium can he identified and subsequently measured for possible activity. Finally, it is necessary to determine whet her a significan t amount ofany of t he gases heing separated is present in t he sample. The concen trations of most of the radioactive gases in the sam ple ure far too low for the normal met 5ods orchemical analysis, and
- measured amounts of carriers for the gases being separated are added.The radiochemical recovery of the separated gas finally loaded into a proportional tube for counting is determined from the total amount of gas.
present hefore separation.
GENERALSEPARATION PROCEDURE
^
E Figure 1 indicates the general procedure for processing the samples. If a sample (such as cocer gas or decay-tank gas)is relatively high in total specific activity,and if the concentration of the lowest activity gas fraction
- Supportedin part hy USA ECcontract A T(ll.112222 and USEPA contracts tinni.n.322 anduni-I.A n.inx 200 k
. IN @ bQ hhofk.
_k([ -_ _f r
- d T,
. _7#
1 M.5 Dub
... nn -
f e
o is higher than a pproximately 5 pCi/ml of sample, less than 10 mlof sample is processed. The sample is mixed with carriers and injected directly into a chromatograph for high activity samples.The separated fractions are trapped,and those relatively high in activity are measured for percent yield, and loaded into proportional
" tubes for counting. The trapped gases that are relatively low in activity (in general, < 10> pCi in the separated fraction)are sent through a separate chromatograph for further purification and decontamination.
The gases trapped from this intermediate-level chr<m atograph are measured for chemical recovery, and loaded into gas-proportional counting tubes using a gas-handling rack reserved for the intermediate-to low-activity fractions.This separation ofsystems ia necessary to minimize cross-con tarMnation.
Samples (such as containment air) that are relatively low in specific activity and contain gas fractions with activity less than approximately 5 pCi/mi must be processed in aliquots of one or more liters. These undergo rough separation prior to chromatographic purification. Ilecause of the volume reduction following this rough separation from such large sample volumes, some fractions have high specific activity, while others are relatively low. The high-activity fractions and the intermediate-to low-activity fractions are, chromatographically purified on separate systems prior to being loaded for counting as described above.
i HIGII-ACTIVITY SAM PLES
~
Figure 2 is a schematic drawing of the gas-handling vacuum rack used for processing samples ofless than about to minf gas having relatively high specific activity.The sample is measured and mixed with measured amounts ofcarriers for Ar, Kr, Xc,II. CII. and CO. About 0.51.5 ml of each carrier is a convenient amount 2
4 2
to process. The sample and carriers are then transferred to a molecular 3ieve U-trap connected to a gas injection valvenn thechrommograph.The gases are transferred from section to section using molecular sieve fingers and U traps. The molecular sieve is cooled with liquid nitrogen to adsorb the gases, and heated with a nichrome wire wrap to desorb the gases. Using helium carrier gas, the sample is passed through the chromatograph. As the various gases are cluted(as observed with a thermal conductivity detector), the gases ofinterest are trapped on molecular sieve U traps cooled with liquid nitrogen.The helium carrier is pumped off the cooled traps. The trapped gases that are relatively high in activity are volume-measured for chemical recovery and then loaded into gas-proportional counting tubes. The low-activity gases are transferred to a separate system used only for intermediate-to low level gases. There they are further purified through the intermedia te-activity ch romatogra ph, volume-measu red, a nd loaded in to tu bes.
Figure'lis a more detailed drawing of the type of volume-measure and tube. load systems used. The gas to be measured is transferred to the molecular sieve finger. Stopcocks A and Il are then closed, and the finger is heated to desorb the gas off the molecular sieve. When no more gas is being desorbed, stopcock C is closed, and the pressure of the gas contained between stopcocks A, B and C is measured. The volume contained between these stopcocks has previously been determined, and fro.a these data the aniount of gas present is calculated.
If the gas is to be loaded ir.to a proportional counting tube,it is expanded into the evacuated tube, which is then filled with counting gas to slightly more than atmospheric pressu re.
Figure 4 shows typical chromatograms obtained with the high-activity and intermoliate-activity chmmatocraphs.The high activity chromatograph has a column of10' x 1/4',' molecular sieve SA,40-60 mesh, with helium carrier flowing at 60 ml/ min. Normally the column is run at room temperature until the CII 4
fraction is off, and then is heated to 300 C to drive off the xenon and CO2 fractions. There is very little separation between the krypton and CII on the high level chromatograph. Ilowever, on the intermediate-4 level chmmatograph, which has a column of 20' x I/4" molecular sieve 5A,40-60 mesh, the separation is very good.
Very oft'n the krypton in a sample has a much higher activity than the "C or tritium in methane. To determine the decontamination factor between krypton and CII4, a "Kr source was mixed with the carriers, and as the krypton aryd Clia were cluted from the high level chromatograph, the gases were trapped separately, and subsequently counted without further purification. The initial activity of the krypton was 3.5 x 103 pCi: tha activity of krypton in the Clia fraction was 2.0 x 10 5 pCi. The decontamination factor for krypton in theCII fraction was thus 175.
4 The same experiment was repeated with the intermed~iate-level chrom.tograph. In this case, to avoid possible contamination of the system. only 1.4 x 10 ' pCi of krypton was mixed with the carriers prior to injection. No measurable activity could he seen for krypton in the Cila fraction, and a value ofless than 8.2 x 10> pCi was obtained, resulting in a decontamination factor greater than 170. There is over 10 minutes of baseline separatian between these peaks, and the decontamination factor is certainly much higher than indicated from the low activity of krypton used. In the normal processing of a sample, the CII4 fractiori trapped off the high level chromatograph would be sent through the intermediate-level chromatograph prior to counting, with a combined decontamination of krypton considerably greater than 3 x 10'. This example illustrates the high decontamination factors obtainable with multiple GC purification.
Cross contamination is normally limited not by GC separation, bu t by system contamina tion. The problem of system contaminatirm varies from gas to gas, and is most severe with tritium - either as gas or water vapor.To measure activities differing by more than six orders of magnitude, a separation of gas handling systems and chromatographs as described above is required. In addition, system and proportional tube blanks must he run between samples to determine possible residual activities.
- 210 -
e
.\\
=,-----emu
---1EMC
V,l f
- l,. ;
. INTER 31EDI ATE-ACTIVITY SA31PLES.
' The separation system used forliter size, inten.wliate activity samples is shmvn schematically in Figure 5.
r With this system, the volume of each separated n.,ction is reduced to several milliliters for subsequent
- chromatographic purification.The procedure involves adsorption followed by clution with helium carrier, similar to the methods described by Momyer(1960) for krypton and xenon.The sample is mixed with carriers.
.i J
bled at n rate ofapproximately 40 m!' min through two cold traps,and t%n adsorbed on a 4' x 1/2" glass coil of activated charcoal cooled with liquid nitrogen.The trap cooled with dry ice-acetone collects the water vapor;
2 After the sample has been adsorbed onto the charcoal, which is maintained in a liquid nitrogen hath, a helium flow is initiated at about500 ml/ min, passing through the charcoal and molecular s' eve coils and out through a mercury stick manometer. The molecular sieve coil is about 4' x 1/2". After the helium flow is established,a small fraction of the helium stream coming off the molecular sieve coil is continuously sampled with an Aero Vac G10 mass spectrometer.Various other methods, such as a thermal conductivity cell. could be used to monitor the gases ciuted from the adsorbents.
Once the retention times for the gases of interest have been determined, sampling is not necessary. When
]-
dry ice-acetone sturries are placed around both the charcoal and molecular sieve coils, the hydrogen, argon,-
oxygen and nitrogen are rapidly cluted from the charcoal, while the krypton and Cil are retained. The
. hydrogen is observed coming off the molecular sieve coil about 5 minutes after both coils have been cooled with dry ice-acetone, and it is entirely off approximately 3 minutes later. It is tra pped in a molecular sieve U.
- 8'-
- trap cooled with liquid nitrogen. Approximately 5 minutes after the hydrogen is off, argon and oxygen are observed coming off the molecular sieve coil.These gases are trapped together on a separate l ttra p. At dry-ice temperature, the nitrogen is rc+ained on the molecular sieve coil, while the argon and oxygen are cluted and trapped.The coil is then warmed to room temperature to elut the nitrogen. When the nitrogen is off, both coils arewarmed to nhout 100'Clonceelerate theelution of the krypton and CII. which are trapped together on the 4
third molecular sieve U-trap.The argon and oxygen, which are not easily separated by chromatographic methods, are transferred to a furnace containing copper turnings which is heated to approximately 400'C to n'j-remove the oxygen by reduction to CuO.-
All the separated fractions,which are now reduced in volume to several milliliters, are chromatographically purified as described for small-vo!ume samples.
. Between samples all the adsorbents used in separation of the gases are heated to 400'C while being purged with helium, or evacuated, in order to remove any residual gases and to reactivate the adsorbents for subsequent processingcycles.
SU3131ARY Methods ofseparation for Ar, Kr, Xc,11, CII t, and CO2 in reactorgas efflu'ent samples u p to several liters in 2
i size have been described.These general separation procedures are being extended to include other gases of interest, such as CO, C2ilG, CalI8. SO2.12, and CII31.
-l The counting techniques and interpretation of the effects of noble gas levels are described in the accompanying papers.
. REFERENCES l
Matuszek, J.M., C.J. Paperiello and C.O. Kt.nz,(1973), Reactor Contributions to A tmospheric Noble.
[
Gas Radioactivity Levels, ?roceedings of the Noble Gas Symposium, Las Vegas, Nevada, September 24-28,
.1973.
E Momyer, F.F.,(1960), The Radiochemistry of the Rare Gases, Report NAS-NS 3025(National Academy of Sciences, National Research Council). -
Vaperiello, C.J., (l973), Internal Gas Proportiemal Beta Spectrometry for Measurement of Radioactive
' Noble Gases in Reactor E//luents, Proceedings of the Noble Gas Symposium, Las Vegas, Nevada, September
. 24 28,'1973.
l 4,
v q.
- 211 -
- [ W
- ~ ',
4
-(
g
-,, _...%a
.~
g
.=:
~
' -.:. ; a..
-- r ~ 3 -.=
~
gwe-me: <4 r.
. g-7.-..
?b -
I TA BLE 1.The G as Compositions ofSamples from Various Henctors (Vol%).
- Ileavy water
,f:
..HWR
.. PWH (I) : ' l'WR (II).
1*WR(II)
. IITGR. research reactor
' Gas ~.. stack gas decay tank ' decay tank containmentair decay tank covergas
. N, :
- 69.79
- 7.01 79.11-78.8:1 10.88 0.22 3
- j
,O -'
- 21.-II 0.17 0.01 20.09 1,10
' O.02 g
.~Ar 0.80 0.08 '
O.17 0.92 0.1 l
. CO 0'I *
' 0'0 I '
0*I I-0'20 0*i 2
~
2
.I1 8'SN '
NI'R2 I8'7 0'07 2
~~
!le 87.69 99.76 Cil,g.
0.11 0.19 4
. C II II) 0.02 2G fi C,3 g
'H) i
. li 0 0.07 0.87 2
I f,
e F
e 1
p.
.j.
,'d t
s w
4
/4; a
s 4
s
.I p
,s
<- 212 -
.x c
't D
h. 'y E. j
,r
- te f
.5 ' r i D,? t l
(*
i
- ,4, y
P HIGH-ACTIVITY GAS INTERMEDIATE ACTIVliY GAS mi-size sample liter-size sample If if High-activity gas Intermediate activity g
handling rock gas handling rock Volume measure:
Volume measure:
add carriers add carriers 2
i 1r 1r 1r i
High-activity Liter bleed-down:
chromatograph rough separation If ir i1r u
High-Intermediate-High-Intermediate-activity activity activity activity gases gases gases gases I
i I
If if if High-oclivitygos Intermediate-activity loading rock chromatograph Sample yield tube load i
Intermediate-activity gas loading rack Sample yield tube load Figure 1. Proc < ssing procedure for high-activity and intermediate activity ga< samples.
2 i
213
~.
1..
s
-h m/
a.
i' HEllUM EXPANSION VACUUM I CARRIER l
h CUUM DILUTION (GASE FLASK 4
GAS INJECTION VALVE l
l
.l'
^
I X
Xfe SAMPLE Y
^
X
)
X b
X, g
MOL.
'N SIEVE y4 FINGER i
a X
MOL. SIEVE O VACUUM n
GAS CHROMAT0 GRAPH U-TRAP f
d:
COUNTING GAS I
X X
)
X X,
'I h
X f-p"-X-FURNACE
[
3 MOLSIEVE
~
X---
-X-Cu TURNINGS FINGER
,()
-X-MOL. SIEVE N
f V
He VENT A-FINGER PROPORTIONAL M OL. SIEVE Hg STICK
- g:
U-TRAPS N7 MANOMETER TUBE 0
L Hg U-MANOMETER Figure 2. Schematic diagram of the high-activity gas separation system.
\\
'9
b
.... g rFh I:
E~
+-
%~ COUNTING GAS (P-lO)
I!
3-A i"
B 4
r-E-)
E VACUUM GLASS-TO-F L
METAL SEAL E
RACK 5-
~
?jN FLEXIBLE
[
[
5 I MOL.
TUBING g'
~
g SIEVE FINGER f
MICK v7
- 7 T
CONN ECTION vf I
t Hg BUBBLER i j G
- s 3 :
G -
G 1
- s PROPORTIONAL
~
2 2
TUBE Hg MANOMETER g
3 Figure 3. Volume measureand tube-load sectionsorseparation system.
4
\\
d
Ar+02 6
Xe N2 Kr CH4 CO2 H2 3
)
U c
J 1
I I
l t
w t
-a
'Ar Kr Xe f
No CH 4
02 H2 hJ k
J
(
i i
i e
O t
i 20 i
40 r
60 p
MINUTES i
(Top:liigh-activitycolumn: Figure 4.Chromatograms for permanent gases on molecular sieve 5A 1040[h 10'x ! / I". not tom: Intermediate activity column:20'x 1/-1".)esh; helium flow 60 ml' min.
- 1:
k
.,4,
,7 f
t6 w_,.a..g _ _, _5 -[
^
^
pm,
'o,;
..j 1 ?,h
. EXPAN SIO N l CARRIERS I
)
FLASK I
1:
I VACUUM -X
?
h X
X-X-
XM L
k_X SAMPLE
{
V LIO.N'7ROGEN I
MOL. SIEVE v
DRYICE b
5 I"
ACETONE 5
- L:
S
- Hg U-MANOMETER HEllUM l:
-X VACUUM =
Xs X iir i
X ni X-t Yt 1 L7 X
i N
t
/
MOL. SIEVE U-TRAPS t
g'
' MOL. SIEVE CHARCOAL COIL f.
Hg STICK r
MANOMETERh b ygfg;cggt-p C0ll q
FURNACE
+MOL SIEVE MASS Cu TURNINGS FINGER SPECTROMETER i
Figure 5.R1eed down system far liter-size samples.
\\
i j.
i i
6 A* -
INTEltNAL GAS.PHOPOltTIONAL llETA-SI'ECTitOMETitY FOlt M$ASUltEMENT OF ltADIOACTIVE NOllLE GASES IN ItEACTOlt EFFLUENTS' '
C.J. Paperiello Itadiological Sciences Laboratory Division of Laboratories and llescarch New York State Department of Ilealth Albany, New York 12201 Abstract At the Radiological Sciences Laboratory of the New York State Department of flealth, gas fractions separated by gas chromatography are analyzed by internalgas 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 teith multichannel analysis offers several advantages, particularly improved sensitivity '.nd specificity. Gas coun ting efficien cies a re greater than 6W> for A r and 9Wo for **Kr.
Detector background teith 100-ml copper proportion al tubes andplastic an ticoincidencegua rds is on th e order of 0.3 cpm for Ar and I.S cpm for "Kr. Shielded and guarded steel tubes have backgrounds approximately four times higher, but are acceptable for high level reactor comples.
fly examining the spectra scith a multichannel analyzer, thefigure of merit forlow-energy beta emitters is grea tly improved over in tegral bias coun ting. Th ep u rity of th e sa mple follateing ch rom a tograph it sepa ra tion can also be checked. Within certain abundance ratios, thelevels of*lland"Cin hydrocarbon fractions can be determined teithout combustion. Similarly, direct Ar and "Ar 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 ofIIealth, 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 10"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 includeimprovement in the ligure r;fmerit for som* batumittm, a chnk on the purity of the sample after chromatographie: separation, and simultur,+<,us ar,alysis e,f errtain inthpic mixtures found in reactor gah elliuents after t.hromathgraphir; s+ par:stion. 'lkv-rnixturen sn*lude s
- Ar 'Ar in the radir argon fraction. MCrH in methane, and r ther hydtr r.arbr n frar.tlonn.
a o
o The gases routinely measured in reactor effluents by internal gas proportir,nal counting include t}ie no,ble gases # Kr. mXe, Xe, "Ar, and "Arand the permanent gases 'il. "CO,and Gila t"C and 'll).Thelatter 2
2 group is important in noble gas measuremen ts because during chromatographie separation. contamination or the argon fraction with 'll, the methane fraction with "Kr, and the CO fraction with the mXe can occur.
2 2
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 14A cm thick steel shield. Figure I shows two of the 100-ml proportional tubes along with a 1 liter tube and two of the guard detectors. The smaller guard willaccept the 100-ml tubes,while thelargerguard 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 ofless than 1% per 100 volts.
A block diagram of the system is show'n in Figure 2. Pulses from the proportional tube are amplified ed
. shaped befora pa sing through a linear gate to the multichannel analyzer. If an event occurs in the guard detector, the lineargateis for 10 psee.The system dead time,due tes guard events closing the linear gate,is less e
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.
- Supportedinpart by USA TCcontract A T(11 1)2222 and USEPA contracts 68 01-0522and68-01.LA 0505
- 23' 9 -
P
- r
~
c4 } pI
.u.. l
Aec
-,.,#~~%._~
aw.w.w-~ x -
-r q.
g O
q' 9:
w T*'
,4;..
j d-
.,s 9 ^
~ OPERATION AND CilARACTERISTICS OF SPECTROMETER SYSTEM 4
I. Spectra of 'll, "C, and "Kr obtairied 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 "$1 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 ba re stainless-steel detector,104 cpm; for the detector shielded by l4.4 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 7
- achieved with a background of 0.3 cpm under the "Ar peak.
i
' Argon spectra are analyzed for both "Ar(from the Auger peak)and *Ar(from thecontinuous spectrum). In
~ order to avoid possible pulse pileup from "Ar, only the spectral region about 6 kev is used for "Ar analysis.
The system has been calibrated for "Ar efficiency using a standard obtained from the National Bureau of Standards, and a value of ORI cpm /dpm has been obtained for the Auger peak efficiency.
A major advantage of spectral analysis is shown by comparing the spectra of an NHS "Ar standard, a e contaminated argon fraction after two passes through a gas chromatograph, and the same sample after an additional pass (Figure 5). The contammated sample ccmtains a very small amount of all, which would not affect the "Ar value, but could have a serious effect on the "Ar value. Integral bias counting would not show
- this contamination. Gross contamination can occur even with double chromatographic separation because of the large range ofisotopic abundances in reactor gas effluents. The chromatographic separation sequence
- leads to decontamination factors of 10 ' to 10'(Kunz,1973). One may occasionally observe 'll interferences in argon,"Kr in methane, and 0"Xe and uimXe in CO and CO fractions. If great care is not taken in cleaning 2
the separation system after a particularly high activity sample, 'll can show up anywhere.
For sample spectra such as "Kr,'ll,"CO. and "CO, the spectral shape and end point are first examined for 2
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
-oimXe and '"Xe, this procedure is repeated for several counts, and the data are fitted by a least sqcares method to a two-component decay curve.
l l At the present time,"Ar and "Kr are the only gas standards available from NHS. For '"Xe 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 C"Xe has been ignored.The result should be a somewhat greater value for this factor.The same factor is used for nimXe. Since "Ar has a beta spectrum similar to "Kr. the "Kr efficiency factor for that spectral region is used for "Ar. In many reactor gas samples "Kr and "'Xe are present in a sufficient concentration to permit analysis by Ge(Li) gamma counting of the sample in the sampling vessel.
The 0"Xe proportional counter efficiency from coincidence calibration provides good agreement with the Geti.i) diode measurements.
One of the more interesting gases present in reactor effluents is methane. it may be composed of'll or "C,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 reactoris presented.The spectrum is run at a system gain of 0.42 kev / channel. The regions from channels I to 39, and from 40 to 255, are summed, and the background
- subtracted. The net counts in region 1 and 2, N and N are given by:
i 2
s
. ~
N yg(T)N *I(C)NC III s
g T I 9
N *I (T)N *I (C)N
.(2) 2 2 T 2 C
7 her[ l f (T)= fraction of'If spectrum in region 1, g
e
, f (C)= fraction of"C spectrum in region 1, 4
-g
+
,.7
\\-
T f (T)= fraction of'll spectrum in region 2, 2
- e-
- f (C)= fraction of"C spectrum in region 2, 2
- = net'll count in spectral region, N
T:
y
+
- Ng= net"C countin spectral region.
5 o
-y a p
Y3p 3
- ^
"M; p -
L
-240 -
[.
' +N M k M N :
Mm f
- 4 4;nm&
2
.h
[
YhsN *k$h) xv i
l jQsS, 32 1 x
j 3
i For our tubes and gain range these equations are:
. N, = 0.985 N,gs 0.575 N
-(3' C'
N = 0 015 N,gs 0.125 N 2
C
. Since Ng and N., have been measured, these equations can be solved for NT and Ne and the activities determined withEut combustion of the sample. Ily counting a serics of twelve "C methane samples.it has been determined that the gain can be set with "'I with sufGeient reproducibility that the errors in the constant terms in Equations (:1) and (4) are restricted to the third significant figure.The major drawback of this method
- in that the 'll sensitivity is limited by the amount of"C present.1f the "C activity is one order of magnitude or more above the backgroimd, the detectable limit for 'll is about 7% of the "C activity. In Figure 7 the decomposition of the methane fraction into 411 and "C 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 hetween
. colk ction 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 b the accuracy of the eIGeiency factors.Those given in Table I for "Ar,111,and "C are estimates for "Krand tar,and that for '"Xe has been determined by a methmi which is somewhat lacking in technical justiGeation..
Standants for these an'd other gases will presumably become available in the future. I'roportional tube efGeiency, however, unlike that of most other radiation detectors, varies slowly over a wide range ofenergies.
Extrapolation of detector efficiency for beta emitting isotopes is not especially difGeult, but isotopes which decay by electron capture or by decay of metastable states present problems which require direct comparison to standards.
ItEFEltENCES Allen lt.A.,(llMI5),,1Ica.kuremen t ofSourceStrength. In Alpha. Beta. nnd Gamma. Ray Spectroscopy Vol.
- 1. K. Siegbahn, Ed., p. -125. (Amsterdam, North lhilland).
Curran, S.C.,(1958), The Proportional Coun.cr as Detector and Spectrometer, in llandbuch der Physik, Vol. 45,174 (llerlin: Springer).
Kunz, C.O.,(tv73), Separation Techniques for Reacto. Produced Noble Gases, Proceedings of the Noble Gas Symposium, l.as Vegas, Nevada, Sept. 24-28, 1973.
TAllLE 1, Spectrometer Performance.
Detectablelimit Gas detector Ave 1-misample Gain Channels
. efficiency hkgd 1,000-min count Isotope kev / channel used (cpm /dpm)
(cpm)
(pCi/ml)
"Kr~
-0.83
- 1 255
.0.90 5.0 1.1x103 "Ar 0.21
.10 19 0.63 0.8 6.0 x 10>
"Ar 0.21 32-255 0.15 1.5
- 1.2 x 10J
'"Xe 0.83
.1255 0.SG 5.0 1.1 x 10 7 1 39 0.75
~ 5.0 1.3 x 10J 811 0.42
' ' "C 0.12 1255' O.85 5.0 1.1 x 10 J
" "C'
.0.42 40 255 0.36 0.13 4.0 x 10>
241 -
~. y;.
w N
E ~.i G E L_.vm
...,.-,.s.
....L....-~
~
1-.
7 -. ~
t me E
e i
r;.1 fy Ry e.,
^4 t
?
h _q,dgj'hsih hh ?.ft M
^'
sc e n r.
',i 7,A
...I
.f;Id4fht '%mbj%;h lG{}lM t
4^
'I
($Q Yvg+f.
.x e
<~
t t
V
/ ~
w$9Q91 4 p va 5
I.kj'e@f*lW$g,h!:y'.49*%,'jhy$lg{hn.
- r n
x:p c
g7 r.1 t9/c 's
'd #
l s
'fif' 49%
6
"^
v m., p$wg$'
},v na
- g.whiw 'Y 5
~
'bggwpk;&u?(, (
s.
I
' di*"^~~'.
wo, v
3A w
1 yg f
1 cygg&m%d"g.
3 V
I i
g
[v _ eiii hp!;,67W3 yA
. 1. g
's w SJ 3%
r4 v..fi B l
=_
.. Q. i<.e.wff.
J4s
.a. u n g' c 4
m
$%y) sy}d..-
k
- MAS p&
':,Espin /,
s15 st y
.n?
n;q@a:w,.tW& ti
??u 8 V
4 m
J..
g h$w?fWl.-l.4 fQna;:h.:
b 1
.?
R Win <
V x
.n r -
uydw, ege4vW
' A.ggdh Sax 9 y3 4; a 49,
'h M
$d N$
Y;'$GNM/3b y
'5t E
h, ah..
L a ha# -me v
w/@h g;f!$,W r h~
.; z 5.: h3 I h
DSMbiiWBiti
'~ -
4*
M N'V f!j 4
t' $g WY ?
G fi1U n,.c:3@
i e
i; m
?
z*a m.
%eAM[?JWri,l'.-l,tB.a e.
ft>';$?{
4>
i:.
P
^ ~ ' ' -
, y % {j ? g w, '
==-w&r
?i
@a (I f
- 35.?
Y ie
.6
- -x i.' =
s ?y,n Rios-l
- g 7"
- WAi 6.49 y,#
g.. {W.' :
l 41 e c
.h {,V.m y- )
^gg.~w} Q'p.# grg;. p>c.-
r
. fj;f -
fr I
6 1
l, 6
.y jljpghg..
w3 4
- g g
g
- g s
u n _--
. fs$ [.g,~ ~
,, W 3 T.
1 L <g(t ;,
g ne,.aww
?
m-c w.,
m m)_,9 y. w s.
~-
,y$$,,/
C/I OE.J,E N Q h
3
?
3 4 ;
?:
,) q%..
. y.;; g g i
e
,', % g,, ~ _..
'cI lIr
.M;'CWG.i A'-
i o
3
,, j..,: : ;,.
[::
,.n$ ':1; *,<
Qg.c
,,,e jh;f,Rpg,.,,$
[8,h.,qdff3 kipD, $
1 b.i
- s.E.-fe.%.e.3#$m c.
,,, z.
- 2 12 -
t t
H.V. SUPPLY O-1800 V.
PAD (PRE-AMP BASE AMP DISCRIM)
PM TUBE e
/~\\
PLASTIC SCINTILLATION GUARD NMTICHANNR LINEAR GATE ANALYZER PROPORTIONAL TUBE PRE-PULSE-AMP SHAPING AMP H.V. SU PPLY O-6000 V.
. Figure 2. Block diagram ofinternal gas-proportional beta-spectrometer system.
e 2 33 t
&W
f 5
10 4
10
,4 85 Kr 3
~10 1
~
l--
s p%
o
!4 o
2 10 a
w*
a a
s.,
3g
'd
. \\. a' asa
. a
=
a aa a a a asaasa 4
~
AA A
fl f
e e
g 3
a.ama m
- 0 40 80 12 0 16 0 CHANNEL NUMBER Figu re 3. Spet ra nf 'l l, "C, a nd "' K r. (Gain o.83 kev /cha n nel.)
e
-288-j; i
- .j ' '
F
.a
+
s.a.a m,i=;.x,-.aaw n n.s.:.m,:
- .c ~:m2
5)
=
?
4 p
Y i
'4 7'
S IO h
. BARE DETECTOR ( x100) w c;1 t
~.
- 1 10"
~.
3 a.
r
- ., SHIELDED
~~'
(xlO) q
=
]
l}
' '., ~ '.. '
s.
- jl m
,4 p
- * = - * *
,(
g 2
c.i
<.'I Q
b o
t 10 a
,l' =!
e e.
..+
- f..
il
_a a SHIELDED + GUARD a
A sa $ 'e,/
Vc id
- #a a
- -.. ~.
.}:
4 ama e A
aasa a a
y ae a
a m aa n a ma a
a i
g a
a na a a aa n a a
a a a
k; i
..a.
O 40 80 120 16 0 200 24 0 n
J CHANNEL NUMBER
\\
Figure 4. Hacks:rounr1 r-octra for a 100-ml steel proportional tube. The spectra of the shial<lert anel bare detectors ha ve twen.~hifte<l u, mrt! by factors of 10:,~1100. respectively. main O.12 kev channel.)
i b l Nk l
us,
s t
s s
3 10 s
10*
./
.f 3
l0 t,0 H
z 3O U
2 10
. *4*,
.*..V,
e.
10,
~
a A
B C
_37Ar STANDARD 3 H-CONTAMINATED SAMPLE PURIFIED SAMPLE B l
i e
i e
i i
O 20 0
20 40 0
20 40 60 CHANNEL NUMBER Figures. Argon spectra: A. Nils "Arst:indard; B. Argon fraction with all contamination;and C. Fraction shown in il with an extra purification step.(Gain o.21 kev /chan nel.)
- 246 -
h m
..~;
-,,:n i.;,-
==o=~u-~-~T""~
.c i-
.me f
s.
3 10 :
L v.
4
.,. GROSS COUNTS gv
.a
. A,.
a a
2 a
i 0
- ' a*
M
~
a*
s
a 5--
a i
. Z
~
as D
a, a
O O
. a a
i a
10 aa a
assa a
, BACKGROUND u a
- COUNTS anaa
.4
?
aa aaa aa a s aa sa I
i..i m i-e O
'50 10 0 15 0
.200 l
CHANNEL NUMBER l
Figure 6. Gross spectrum of a methane fraction showing the presence of 'll and C compared with background.(Gain 0.12 kev /cha n nel.)
.s 1
- 24 7'.
- t
- ,
.4, s-g
[
, - L.a...
--.,._a.-_--~~-~i.,-_--.,~--
a-~. -
I 4
10'
~
r
~
NET COUNTS
..~
2
'e 10 U)
F
- 14C z
~)
O O
t 10 h.
3H I
e
- _ _ _i ___
O 50 10 0 150 200 CHANNEL NUMBER Figu re ".S;metrum nf t he methane frar tion shown in I'igure fi with backgrounc! sulstracteri. The 'Il ant! C '
com;>unents are intlicate<l.(Gain n.12 kev / channel.)
l 4
2.l tt.
k
?
5 t
,_a ENVillON31 ENTAL hlONITOltING FOlt KitYPTON-85*
D. E. Barber
. Schoolof Publiclicalth University of 51innesota hiinneapolis,51innesota 55155 Abetract
- Krpresents unique environmentalmonitoringpmblems because it does not react seith other elements and compounds at normal ambient temperatures and pressures. flotvever, elaborate means are available to manage the"Krproblem, but a simpler, inexpensive approach is requiredif monitoring is to be accomplished at many sampling locations. This teork shotes that environmental monitoring for "Kr is possible by
^
collecting air samples in thin plastic bags, and counting the bags for beta particle activity. The direct countingolcontainedsamples ol this type makes itpossible to detect concentrations less than thepublic AIPC for **Kr. The bagged sample technique is readily adaptable to any environmental monitoring station teith
- polver to run a low volume air pump. The idea of counting the bagged semple directly is a ncte, lote cost.
approach to environmentalgas monitoring tchich may have application in environmental, clinical, and industrialsituations.
INTRODUCTION "Kr is a fission product which escapes or is released to the environment primarily as a byproduct of
. reprocessing nuclear fuel. In fuel reprocessing,85Kr, "'I,3"I, usmXe, mXe, and 'If are released, but only "Kr and 81I are released in sufficient quantities, and have long enough halflives, to produce significant concentrations in extensive environmental air volumes (Kirk,1972). The preser t atmospheric inventory of
" Kris estimated to be 6051Ci-more than twice the inventory in 1962 (Kirk,1972). If projections with respect to population, demand for electric power, use ofnuclear po'ver plants, and release of "5Kr to the environment GA,
are correct, concentrations of"Kr in tbe atmosphere may reach 3 x 103 pCi/ml (the public A1PC) about the
'd?7P year 2050(Ilollanu,1969; Cowser and hf organ,1967; and Coleman and Liberace,1966). Treatment of efd uents
/-
to remove up to DS percent of the noble gases from fuel reprocessing is possiole, and methods have been tested on a pilot plant scale (Slansky,1971). The cost of treating the efauent has been estimated at about I percent of gkr.
the total reprocessing cost (Slansky, et al.,1969), and treatment will probably be used extensively in the near Q*@&
future. Consequently, present projections with respect to anticipated atmo,pheric concentrations in the 21st century are probably rauch too high. However, efauent treatment or not,"Kr releases to the atmosphere will
[ Mf.?
require en vironmen tal monitoring for the gas beca use:
Qk, (1) A removal efficiency of 98 percent reduces released concentrations to only 0.02 times their original hp values.
~ Mg n (2) The atmosphere cannot be used exclusively for "Kr dilution,
&y (3)lt will be necessary to assure complia nee with regulatory standards.
W Although this paper addresses itself to the problems of monitoring 5Kr, the results are more generally a
g useful. There are situations which arise when one would like to know the response to be expected from i
M, S 4 ordinary radiation detectors when they are presented to clouds of beta particles or low energy gamma ray emitters with dimensions less than thosc ofinfinite volume.
I
.k, -
There are a number of possibilities for monitoring "Kr in the atmosphere given unlimited financial
} ~ T reveurces. 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 charcoa
[
y
- h. ' beds are used, the operating temperature must be matched to th M,
.C In liquid scintillation counting, the lowest concentration of "Kr that can be analyzed without E
preconcentration is about 3 pCi/ml, and the poor solubility of air in liquid scintillation cocktail presents a (i
6
' pmblein for sampling air mixtures of gases (Kirk,1972). So, sophisticated sampling for "Kr in the h
$ ' environment presents severe practical limitations for both technical and financial reasons, especially in 7 ' those situations where numerous monitoring stations may be involved, Simple, inexpensive methods must be b :
b
[
found.
f Direct counting of gas samples in vinylidene chloride (Saran) is the approach taken here. The sampling b
approach is similar to one already reported for 22 Rn (Sill,1969). But, the idea of counting beta particle g
M emitters directly, withoW removing the sample from the bag for the purpose of gas monitoring, is believed to
- j. benew. lt has been noted..*cently that u3Xecontamination in the air should probably be done by counting the g(P nir sample directly raher than attempting to collect the W in aqueous solution (LeBlanc,1972). LeBlanc
' points out that there is a misunderstanding concerning the best methods to monitor for mXe because of A erroneous solubilities formXe given by the Hbk. of Chm. and Phys. (1966). Similar uncertainties may also 3
applytokrypton.
'The data un "Kr r ported here resultedfrom tenrh completed in Ihe Faculty Research Participation Program
^
?' of Associated Western Universities at the llcalth Services 1.aboratory, USAEC, Idaho Falls, Idaho. The j
author also grateftdly acknoteledges the cooperatton of Aerojet Nuclear Company.
f
,C 2w.
N w
u. [5
- =: :.~u_,.:. i: a;4 =
u. =. ; ;4;.
.t'
,p9 7
m
.~ -
l,W
=
. PROPERTIES OF"Kr AND INTERFERENCE PROBLEMS Mme properties of 88Kr are given in Table 1." Kris nearly a pure beta emitter from the dosimetry viewpoint, but the gamma ray is useful for calibration purposca for activities greater than 10 8pCi. Monitoring for85Krby i
' gamma ray spectrometry can be accomplished provided the concentration is high enough, and is sustained for a peaod of time at least equal to the measurement period. At concentrations approaching the maximum permt.,sible concent rations, however, the branching ratio (0.004) of gam ma ray disintegration to beta particle disintegration maken beta particle measurements far more attractive..
Iodineimpregnated, activated charcoal can be used to eliminate interference from iodine.131. Silica gel can
~ - be used % *.eep the sample dry. Storage can be used to eliminate interference from radon-daughter product
~ activity, w I filters can be used to reduce interference from particulate activity. There remain, then, ia3Xe and
- isamXe/I r techniqua might be biased against these nuclides by storage, careful selection of detector window thickness.and detector design. Interference from 883Xe is not expected to be significant at present because the ratio of 888Xe to "Kr 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 caseitwillbe necessary to cope with the 888Xe problem.This should be relatively easy to do because of the large dif ferences in modes ordecay, and beta energies between those for 832Xe and "Kr.
< If 888Xe is produced as a fission product, naamXe will also be present (Martin and Blichert Toft,1970). The L i33mXe 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 isamXe are only slightly lower than the average energy of beta particles from,sKr flowever, advantage might be taken of the 0.233 MeV gamma ray from s
- 33m Xe to distinguish it from both *$Kr and 883Xe.
METHODS 1."Kr Chamber.
4
- chamber approximating an infinite volume of air for "Kr 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 f'
l ambient temperature and pressure.
d' i
N 1 Chamber concentrations were measured and monitored by a 30 mgtm2 metal walled Amperex G.M. tube e
(1/2" dia. x 7" long) operated at 960 volts. Pulses from this detector were counted with a llaird-Atomic Mod. 530 scaler.
- 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 asKr ranging from 2 x 10 *pCi/ml to 7 x 10 4 pCi/ml.
- 2. Calibrations.
2-
' Unfortunately,the detection limit by gamma ray spectrometry tor" Kris on the order ofl0>pCi. As a result, y
~
l chamber concentrations could not be confirmed by this method over the range of concentrations ofinterest. To
.J enable measurements of concentration in the chamber over the entire range ofinterest, the G.M. tube in the a
X
- chamber was calibrated as follows. A sample from "Kr stock was counted on a 65 ml GeLi detector in the 4
- 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 literSaran bag to calibrate an 8"x 4"NaI detector for this
- 3 geometry. A 4.2 litersample taken from the chamber was then counted on the same crystal yielding the true q.
. concentration in the chamber l.05 20.15 x 10 il.Ci/ml.This concentration provided the primary calibration I;
1 L factor,3.82 x 10>pCi/ml per epm, for the chamber G.M. tube.The G.M. tube count rate was then used as the i
f-refererice for the concentration in all samples taken from the chamber.The G.M. tube calibration factor agrees j
4 v'
well with the 2.1 x 10>pCi/ml per epm previously reported for a similar, but longer, G.M. tube (Smith, et al.,
1970).-
r The rota' meter was calibrated with a wet-test meter, and was found to be in calibration withint20 percent.
~
The uncertainties in visually setting the rotameter are large. Consequently, the rotameter reading w'as taken 1
' to be the true flow rate at ambient temperature and pressurr.
i s
N,
' 3.SamplingandInstrumentation.
Q
. e A diagram of the dilution and sampling system is given in Figure 1. The sampling line was prepared as it ~ ~'
M might be used in the field to pretreat samples taken from the chamber. Components of the sampling line were
- [
1
~ onnected with plastic tubing ranging from 1/4 to 3/8 inch inside diameter. The input to the rotameter from b
c the chamber was at ambient pressure, and consisted of 3/8 inch plastic tubing suspended at the center of the j(
f i
i"Kr chamber close to the G.M. tube. Samples were collected at various flow rates and various sampling times M
asindicatedin the results.
. Exce'pt for a few measurements with'the 5-liter Saran bag filled to capacity, bags were filled to less than 7
- capacity to minimize leakage,if any should occur, and to provide a flexible bag geometry. In this way,it was N
t;
=T
. possible te achieve 2 tr counting geometry when samples were placed on the detectors, and to reduce the error 6
L which might be introducd as a result of pressure build up in the bag.
,fi
-~,
w
~
Ifr -
L t
- no.
E y y'
j.
4 4
s n
p x!n b
^-
Wt ? _
6 e pf
. WC j C
A
i The in struments used to analyze the bagged. samples werc as follows:
(1) Ludlum Instrument Co., Model 14 A Gieger Counter with LND Inc. G.M. tube,30 mg/crh* metal wall.
f?) Eberline Instrument Co., Model IIP-210 G.M. probe with LND 731-1 G.M. tube, 2" diameter thin wirdow connectedtoBaird Atomic,Model530 scaler operatedat800V.
(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 scaler 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 en samples stored overnigh t.
- 4. Sample Containers.
Two type's of plastic bags were used to collect, store, and count samples. One bag was Saran type 18-100 (ma'nufactured 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 223Rn with losses of less tb.n 0.12 percent per day u p to at least 14 days (Percival,1971). It, therefore, seemed a suitable choice for "Kr. 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.wa '20" square with a wall thickness of 4.7 mgicm2.The open end of the bag was heat sealed, and the center of one a3e was cut out to a diameter of about 7"to accommodate a thin Saran wi view. 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 adhenve 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 compaunds, including water vapor, and probably is unsuitable for "Kr 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 rr geometry when placed on the plastic scintillation detector.
RESULTS
- 1. *8K r Chamber.
With the chamber containing 3 x 10 *pCi/ml of"Kr, the count rate of the chamber G.M. tube dropped from 7.43 x 10* cpm to 7.30 x 10* cpm over a 150-minu te period.This amoun ts to a leakage rate of 0.8 percent per hour i
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 usualin the chamber,it required 2 to 3 minutes to reduce the chamber concentration to 1/2 ofits originalvalue.
- 2. Chamber G.M. Tube.
'The count rate of the"Kr chamber G.M. tube as a function ofeKr concentration is given in Figure 2 for both
. input to the scaler 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 "Kr provi*ded 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 practicallower limit of detection for the Ludlum rate meter with the chamber G.M. tube is
~ about 6 x 10 7p Ci/ml; this provides a net meter reading of 0.04 mR/h in an infinite cloud. When the tube is '
connected to a scaler, the lowerlimit of sensitivity is a function of the background and counting time. For the conditions of Figure 2, the limit for detecting "Kr with the scaleris much lower than with the rate meter.
- 3.Ludlum G.M. Survey Meter.
With a chamber concentration of 7.0 x 10 *p Ci/ml samples were taken in the 5-liter Saran bag for time intervals ranging from 5 to 25 seconds at 12.5 Ipm. When the Ludhm 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
. Figure 3).
- 251 -
ll %
?
.j
,P'
'f
t
~
' The lowest activity was distributed in a volume of only I liter, but the highest activity was distributed in 5 litein.There are large uncertainties in reading this meter, but the linear relation over a factor of five change in
' volumeis elear.
'. The minimum reliable net reading for the Ludlum survey meter is 0.0' nR/hr. This corresponds to about 0.05pCi which, according to the figure, n ight be distributed in as rr as 5 liters. IIence, the minimum detectable concentratiun for this meter under these conditions of meawment becomes 10>pCi/ml. This is much too hi'gh for environmental monitoring purposes.
4.Eberline G.M. Probe.
This probe was used to measure SKr activity in the 5 liter Saran bag containing various volumes of ga s at a a
constant concentration of 6.3 x 10A pCi/ml.Theiesults are shown in Figure 4.The statistical counting errors are large;but, again the linear relatwn between a:tivity 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 0.5pCi. When expanded to 5 liters, this yields a minimum detectable concentration of 10>pCi/ml. But, measurements at maximum volume and various concentrations show the minimum dr ectable concentration to be much lower than this (See Figure 5).
When 51iters of"Kr are taken from the enamber in the 5 liter Saran bag, the response of the probe is as indicated in Figure 5.The probe was lighdy pressed against the side of the sample bag for each measurement.
The highest concentration was measu'.ed 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 103;' Ci/mi of "Kr under the conditio' of measurement. But, itis not likely to detect 3 x 10 *pCi/ml even for long counting times. A more n
sensitivedetectoris require.
5.PlasticScintillation Detector.
Samples were taken from the chamber, diluted to the desired concentration,vith ambient air through the sampling chain into the polyethylene bags, and stored overnight to permit the decay of radon daughter product activity.The"Kr concentration for all these samples was 2 x 10>p 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 daughterproductsattheselow 'Krconcentrations.
8 When stored samples of"Kr are counted, Figure 6 shows that it is possible to detect concentrations less than 3 x 107 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 esKr 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/cm Saran window, and counted on a plastic scintillation counter. This 2
1 combination provides minimum detectable concentrations ofless than 3 x 10 'pCi/ml for "Kr. The method is j
suitable for environmental monitoring provided pretreatment of the sample removes other beta emitters, such wl, which would interfer with the analysis, and provided the sample container used in the field is as reasonably durable and impermeable to "Kr.
- The commerciallyavailable,8.0 mg/cm2 density thicknessSaran 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, amisufficient flow rate stability, to accurately pum p 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 "Kr in bagged,5 liter samples containing concentrations as low as 10>
to 103 pCi/ml. Common G.M. probes connected to scalers can detect "Kr in hagged 5 liter samples
^
contaimng concentrations as low as 3 x 10>pCi/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
- scalers are capable. of detecting concentrations below 3 x 10 8 pCi/ml, depending upon' counting time.
Detecting these low-concentrations with J2 or 23 liter bagged samples requires a well-shielded, large area plastic scintillation counter orits 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
'"Krconcentrationsassmallas3x10 epCi/mlu ingth te echnique described here with elight modifications.
s
' The variations in data points for bagged-samples are due primarily to inaccuracies in reproducing rotameter settings. Construction variations in polyethylene baFs also contribute to the variation in observations where these bags were used. Observations in Figures 2 through 5 were made immediately afler the samples were taken. Variations in radon daughter product activity over short periods of time contribute to
- the variation in observation for these samples.
~
4 252 -
+
^
n:
g
~..
~.-
p__
i y..
With the detector on the sample bag, the geometry is essentially constant at 2 77. So, the detector response is expected to be linear with increasing volume (therefore activity) for a fixed concentration in the samrle as shown in Figures 3 and 4. This should be true for volumee 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 10dFCi/ml.
Thereis 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.ap Ci/ml and 0.004 cpm /dpm at 63 x 10.*
pCi/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 orderoflo tp Ci/ml as Figure 6 shows.The slope shown in Figure 6 is t.00.
Figure 6 yields an efficiency of0.014 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 1
the detector.The 23 liter bags drooped somewhat below the 2 rr solid angle of the scintillation detector. The 1
{
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.
Caref ul attention to this question would probably reduce further the minimum detectable concentration for
~
85Kr.
Bagged-samples provide several advantagec over"in situ " measurements of a5Kr.
(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.
i (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
- to85Krbeta particles.
(3) Both sample geometry and detector geometry are always known, and are reproducible with bagged.
samples.
1 (4) Provided sufficient activity is collected, the average concentration during f he sampling period will be l
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.
i The disadvantages of bagged samples are:
i (1)They are incapable ofidentifying either the time or the magnitude of changes in air concentrations, i
and are not suitable for an alarm system.
(2)ln large volumo they are awkward to handle, and require that precautions against leakage be taken.
SUMMARY
AND CONCLUSIONS An inexpensive method to monitor atmospheric 5Kr at or below 3 x 10>pCi/ml (0.1 MPC), with minor a
j modifications to existing er,vironmental air sampling stations, has been described. But the method needs g
f urther development and testing with mixtures of radioactive gases and aerosols likely to be found where85Kr is emitted.The method does not require volumes which areinfinite with respect to the beta particle energy of
{
asKr. Neither does it require concentrations of activity which are ste ble with respect to time. Therefore, it is a i
realistic method from the viewpoint of conditions likely to be experienced in the field. One rarely encounters a truely in finite cloud sustained over a period of time long enough to make in finite cloud measu rements realis tic.
The ultimate test of the techniq ue should involve mixtures of ull,"2Xe. iH, and "Kr in concentrations likely to be encountered in the environment ofreactors and fuel reprocessing plants. If the technique should fail this i
test of mixtures, it may still be useful as a screening technique for radioactive gases released to the
- environment by man.
l
.)
REFERENCES I'
i Berge r, M. J., (1971), J. o f Nucl. Med.. Su p. No;5, Vol.12.
i Code of Federal Regulation s (1965), Title 10. Part 20.
Coleman,J.R. and R.Liberace,(1966), Radiolog. Ilealth Data Reports. 7,615.
Cowser, K. E., and K. Z. Morgan,(1967), Health Physics Division AnnualReport, ORNI,4168,39-45.
2 t
Handbook of Chemistry and Physics,(1966). Cleveland: The Chemical Rubber Company.
. Hofland,J. A.,(1969), Proceedings o'f the AEC Symposium, BiologicalImplications of the Nuclear Age, I
USAEC Division of Technical Informa tion.
{
Kirk, W. P. (1972), Environmental Protection Agency, Office of Research and Monitoring, Washington, g
- D.C.: U.S. Government Printing Office,484 482/46.
g LeBlanc. A. D.,(1972), Phys. Med. Biol.,17,585-589.
Lederer, C. M., J. M. Hollander, and I. Perlman, (1967), Table ofIsotopes,6th Ed. (New York: John 4
i Wiley and Sons. Inc.)...
-}
Martin,M.J.and P.H.Blichert-Toft,(1970),NuclearData Tables,8A,108.
t 1
253 -
4 L
3
,, =.
..w I
Percival, D. R., (1971), Personal Communication, Analytical Chemistry Branch l Ilealth Services Laboratory, US AEC, Id aho Falls, id aho.
Ra diological Health Han dbook, (1970), We shin gton: Public IIealth Service.
Sill, C. W.,(1969), Ilealth 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,anc.15.Shleien,(1970), PHS,BRII/NERIIL70-4,Rockville, Maryland.
Voilleque, P. G., D. R. Adams, and J. B. Echo,(1970), Health Physics,19,835.
TABLE 1. Properties of a8Kr.
Property 88Kr Reference Density, mg/ml 3.74 Hbk. of Chem. & Phys.,(1966)
Density / Density Air 2.9 Hbk.of Chem. & Phys.,(1966)
Maximum Beta Energy,MeV and Abundancy(%)
0.67(99.6) Lederer, et al.,(1967); Rad.
0.160(0.4) IIcalth IIbx.,(1970); and Martin and Blichert-Toft,(1970)
Average Beta Energy,MeV 0.2464 Berger(1971)
Range ofMaximum Energy Beta in:
l Aluminum, mg/cm' 235 Rad. Ilealth libk.(1970)
Air, cm 182 Rad. IIcalth IIbk. (1970)
Range of Average Energy Beta in:
Aluminum, mg/cm2 59 Rad. IIcalth Hbk. (1970)
Air, cm 46 Rad. IIealth ilbk. (1970)
Specific Gamma Ray Constant, R
Uifrat 1 meter 2.34 x 10 4 The Author
. Deposition Velocity on Grass, cm/sec 2.3 x 10."
Voilleque, et al.,(1970)
Max. Permissible Concentrations,
' pCi/ml; Occupational 1 x 10 5 10 CFR20 (1965)
Public 3 x 10 7 10 CFR20 (1965)
IIalf. life, years 10.76 Rad. Health Hbk. (1970)
Gamma-Ray Energy, MeV and Abundancy (%)
0.517(0.4) Martin and Blichert-Toft (1970)
- 254 -
, : t..
=_ -
Pressurized Pressurized Sample Bag
[
Kr - 85 Air-a 4
e
> Atmosphere v
Reducing Reducing Valve Valve Pump I
i h
Transfer Flask Membrane Filter
?
g a
f Hepa Filter Kr 85 Dilution Chamber Charcoal 4
n V
Rotameter-
>l Silica Gel
>l Hepa Filter
~
\\
Figure l. Krypton dilution and sampling system.
~
i
_.._m.,._._._=.
_,_..~ _ _
c Net Meter Reading (mR/hr)
O N
OJ
'O O
dO e
--o.
O o.
--e.
-~
O x
n O_.
v 9
OJ i
O lh
.g i fq-
, y.t
' b,
't:.'
4 Figure 3. I,udlum G.M. survey meter response to a fixed concentration of a'Kr in a 5-liter Saran bag at
- h. 9, 4
/
dif ferent volumes.
i
$ y' l
' ' ~,Yl lI i J
=
-257 i;t -
-,n..aac..- ~ ~.-- -
.a.~.~-=-.-
~-----~r-~~---
i Scaler (net cpm) e e
6 U
U 6
N u
a 0
l l
l lljj l
l l l j jjj l
i l l l ll 1
e O
Z o
D o o
o (D
3 3
o
~*
O a
-o O_.
N g
-O e
U l
I I Ilill l
I I l l ll1 I
I I I Ill N
o o,
o o
4 o
Ludlum Meter (mR/hr) l S
. Figure 2. Response of the Amperex G.M. tube in the exposure chamber as a function of"Kr concentration.
(The counting time for each observation with the scaler was 3 minutes.)
l
... e
t L.
I 30 e
^
E o.
O, O
e o
Maximum Volume w?
5 liters l
o m
a 20 o
_c
=
e I
? $
Bag Rolled to f4 l
e Maximum Volume e
~
C 3o 10 l
U e
Z I
e e
/
~
0,.
0 1
2 3
4 5
s Total Activity (pCi) s F'gure 4. Response of the Eberline IIP-210 G.M. probe to a fixed concentration of 85Kr in a 5-liter Saran at
, dif ferent v lumes.(The counting time for each observation was 1 minute.)
n' hxr: 4;dge a,
,, 21,.> 43ia u.
. m a_. ;,a _ u. g,..,,,_.,,,,,,,
., v
1 6
10 3
10 e
5 2
10 10 E
~ L.
E
^
^
E*
E O-t
- c. !
}
O.
u
(
u e
i g
-o i,
g o
Z Z,
4 10 t:
10 e
t '
3 5,
10 1
I 111111 I
I 1 11111 I
III" I
I I l i 'll i
i 1111" 10 -
L
-I O
2 3
4 10 10 10 10 10 10 Concentration (pCi/ml)
-4y F'igure 5. Response of the Eberline HP-210 G.M. probe to 5 liters of"Kr in the 5-liter Saran bag containing various concentrations. (The cou nting time for each observation was 3 minutes.)
j j
[l
v 3
I0 O Polyethylene bags with Saron who'ows
-~
O Commercial Saran bags J
2 l0 g
e E
o.
3 e
a) 8 e/
Z e
- - ~ ~ ~ - - - -
- ~ ~ -
3o r r i35 nin tilaun at 170 cpm background o
i l l l-l l l l
I l l llll 1
1 l l l 11 1,
-7
?
-6 10
-8
,10 10 10 10
-9 Concentration ( Ci/mi)
Figu re G. Response of plastie scintillation dotector to 23-liter samples of aKr in polyethylene bags with
.m,
,,,,.; i1 %r-in hnew ( All samples were stored overnight and counted to
' + ' -.
.a