Daniel M. Montgomery onitoring radioiodine in ambient air or

Daniel M. Montgomery
Analytics, Inc.
onitoring radioiodine in ambientair
or gaseouseffluentstreamsrequires
concentrationof radioiodine from
largevolumesof air to achievethe sensitivity requiredby variousregulatoryagencies.
This
is accomplishedby passingan air samplethrougha
particulatefilter followed by a charcoalcartridge.
Sinceradioiodine maybe presentin variouschemical forms includingelementaliodine, organiciodide
species,and otherinorganicspecies,the charcoalcartridgesareimpregnatedwith KI or TEDA
(tetraethylenediamine)
to convertorganiciodide
speciesto forms that arecollectedon charcoal.In
the presenceof high noble gasactivity(i.e., accident
monitoring), a silverzeolitecartridgeis recommendedovera charcoaltype cartridgesincenoble
gasesare not retainedon silverzeolitebut are highly
retainedon charcoaland would significantlylower
the sensitivityfor radioiodine.
The preferredmethodto analyzethe filter and
cartridgeis to count them separatelywith a germanium gamma-rayspectrometer(countingthe filter
and cartridgetogetheris discouraged).Then,the activities on the filter and cartridgearecombinedto
givethe total radioiodineactivity.The contribution
of particulateradioiodineis generallysmall;however,studies1haveshownthat particulateradioiodine mayconstituteasmuchas40% of the total. The
fraction of particulateradioiodineis variableand dependenton the sourceof radioiodineand otherconditions suchaspre-releasetreatmentof the effluent
stream(e.g.,HEPAfilter, charcoaltrain). Becauseof
Reprint of "From the Counting Room",Vol. 1,No.2, 1990
the low-activity levels, the filter and charcoal cartridge are normally counted in close proximity to the
gamma-ray detector (1 cm or less). The cartridge is
counted with the inlet side facing the detector"face" counting. This practice minimizes the counting time necessary to achieve a given sensitivity
(LLD); however, it may introduce significant errors
in the measurement as will later be discussed.
Efficiency calibrations for filters are rather
straight forward. Filter standards with an eight radionuclide mixture (109Cd, 57Co, 139Ce,203Hg, 113Sn,
137
Cs, 88y, and 60Co) covering the energy range from
88 keY to 1836 keVare commercially available or can
be prepared from a standard solution of this mixture. It is important that the activity be uniformly
distributed across the same active area as the samples to be counted. Significant coincidence summing
may occur for 88y and 60Co, and corrections may be
necessary in the energy range from 898 keY to
1836 keY, depending on the detector efficiency and
the proximity of the sample to the detector. The radionuclides 1311,1331and 1351,with the longest halflives (8.0 d, 20.9 h and 6.7 h, respectively), emit
strong gamma rays (at 364, 529 and 1260 keY, respectively) that are essentially free of coincidence
summing even when the sample is counted near the
detector housing. If the 1260-keV gamma ray from
1351is used for quantification, coincident summing
corrections for the standard may be necessary. However, an alternative would be to use the 420-keV
gamma ray from 1351which is outside the energy
47
Radioactivity 6- Radiochemistry
The Counting Room: SpecialEdition
rangewherecoincidencesummingcorrectionsof
the standardarenecessary.
Efficiencycalibrationsfor charcoalcartridgesare
more complexsincethe distribution of iodine in the
cartridgeis dependenton manyfactorsincluding:
the iodine species,humidity, temperature,sample
flow rate,and length of exposure.Thesefactorscan
all affectthe distribution and the penetrationdepth
of the radioiodine in the cartridge.Charcoalcartridgescanbe counteddirectly on the detector,or
the charcoalcanbe removedfrom the container,homogenizedand counted.
The practiceof removingthe charcoalbefore
countinghas,in principle, the advantageof yielding
a homogeneoussample.A calibrationstandardcan
be preparedby addinga known quantity of mixed
gamma-raystandardto the appropriatemassof charcoaland thoroughly mixing to ensurehomogeneity.
Disadvantages
of this techniqueinclude:a lower
countingefficiencythan when"face"counted;a
longeranalysistime due to disassembly
of the cartridge and mixing of the charcoalin anothercontainer,and possiblesamplelossduring transfer.Due
to theseproblems,most nuclearpowerplant counting rooms do not usethis technique.
Basedon my experiencewith the NuclearRegulatory Commission(NRC) and asa consultantto the
nuclearpowerindustry,the most commonpractice
for countingcharcoalcartridgesis to "face"count
the samplewithin a fewcentimetersof the detector.
The preparationof an appropriatestandardrequires
someknowledgeof the radioiodine distribution in
the charcoalcartridgesamples.Basedon the discussionabove,the distribution maybe quite dependent
on the particular monitoring situation(i.e.,environmentalmonitoring, long term stackmonitoring,
grabsamplesfor respiratoryprotectionpurposes),
and environmentalconditions.
This paperfocuseson 1311;
however,othershortlived radioiodinespeciesmaybe presentin effluent
samples.Theseinclude: 1321,
tl;2= 2.3 hours;
1331,
tl/2= 20.9hours; 1341,
tl;2= 50.2minutes,and
1351,
tl;2= 6.7hours. In most nuclearpowerplants
the concentrationsof short-livedradioiodinesin
gaseouseffluentsare low due to the holdup time associatedwith the effluent treatmentsystems.However,if short-livedradioiodineradionuclidesarede-
48
tected,NRC requiresquantificationand reporting of
theseradionuclides.Therefore,it is wiseto calibrate
the detectorsovera wide enoughenergyrangeto
permit analysisof all the iodine radionuclides(i.e.,
up to 1300keY).Additionally,the softwareutilized
by most commerciallyavailablegamma-rayspectroscopysystemsrequire8-10efficiency!energypairs at
appropriateenergiesto accuratelydefinean efficiencyequationoverthe energyrangefrom 88to
1836keY.
Onemethodof preparingcharcoaland silver
zeolitecartridgestandardsinvolvesdisassembling
the cartridgeand gravimetricallyaddinga known
quantity of the eightradionuclidemixture (asdescribedabove)to a portion of the charcoal.The
amountof spikedcharcoalis basedon the depth of
spikedcharcoaldesiredwhenreloadedinto the cartridge.The spikedcharcoalis homogenizedand
placedbackin the cartridgewith a barrier between
the activeand inactivelayersto preventmixing. The
activelayercanbe varied from approximately5 mm
to the total activedepthof the cartridge(homogene-
ous).
Most laboratoriesusea "face"loadedcartridgewhich meansthat the activityis loadedpreferentially
on the inlet or front "face"of the cartridge.Experiencehasshownthat a 5-mm "face"loadedstandard,
countedwith the inlet or "face"sidetoward the detector cryostatfor the entire countingperiod, will
generallyprovide an adequateefficiencycalibration
for charcoalcartridges.However,a betteralternative
is to countthe standards(and samples)with the inlet sidedown ("face"side)for half of the counting
time and then with the inlet sideup ("flip" side)for
the remainderof the countingtime. The spectrumis
collectedoverthe entire countinginterval.This will
giveresults,for both standardsand samples,that are
lesssensitiveto the distribution of iodine.
The latter part of this articleprovidesexperimentaldata on the effectof radioiodine distribution
on detectorefficienciesfor countingwith inlet side
down ("face"counting)and countingby the "flip"
technique.Detectorefficienciesmayvaryconsiderablyfor differentmodelsand typesof cartridges.
Calibrationsare generallyvalid only for the specific
type or brand of cartridgeusedfor the calibration.
Reprint of "From the Counting Room",Vol. 1,No.2, 1990
"CalibratingGermaniumDetectorsfor AssayingRadioiodinein CharcoalCartridges"
DanielM. Montgomery
3 mm from detector
3 cm from detector
Efficiency,%
1311Load
O-mm
2-mm
5-mm
10-mm
15-mm
25-mm
6 cm from detector
Efficiency,%
FB Ratio
Face
Flip
FB Ratio
Face
FI
2.82
2.55
2.25
4.08
1.66
3.00
1.33
2.73
1.04
2.45
2.76
2.72
2.59
2.42
2.38
2.41
2.08
1.92
1.73
1.44
1.22
1.02
1.34
1.28
1.23
1.10
1.01
0.95
1.00
0.97
0.97
0.93
0.92
0.92
3.90
3.59
Efficiency, Ofo
Fe Ratio
Face
Rip
1.75
0.61
0.58
0.56
0.51
0.49
0.46
0.48
1.68
1.56
1.32
1.20
1.00
0.46
0.46
0.45
0.45
0.46
Flip efficiencyrefersto counting efficiencywhenthe sourceis countedWiththe inlet sidefacingthe detectorfor half of the counting
period and the inlet facing awayfor half of the countingperiod.
Faceefficiencyrefersto counting efficiencywhenthe cartridgeis countedwith the inlet sidefacingthe detector.
FB is the ratio of activity observedcounting with the inlet sidefacingthe detectorto the activityobservedwhencounting the inlet
sidefacing awayfrom the detector.
Table 1
Measured front-to-back ratios and counting efficiencies.
Thereareseveralunpublishedstudies2,3
investigatingthe distribution of radioiodine in charcoal
cartridgesunderactualmonitoring conditions.Investigationsby DaleNix2 at TVXsBrownsFerryNuclearPlantshowedthat radioiodinewasfound
within the front 5 mm and the averagedepthwas
about 2 rnrn. Thesestudiesincludedvariousmonitoring locationsin the plant and includedcartridges
that werein placefor up to oneweek.Investigations
by McFarland3at a radioiodine processingfacility
showedthat greaterthan 95% of the radioiodine in
charcoalcartridgesampleswaswithin the front
5 mm. Thesesampleswerecollectedin ambientair
overa period of 2-3 days.Spikingthe first 5 mm on
the inlet side ("face")of a charcoalcartridgewith the
radionuclidestandard,therefore,providesa reasonableapproximationof the expecteddistribution
(and activity gradientcontinuouslydecreasing
from
the front "face"to the back).
Loadingeffectson the measuredcountingefficiencyfor radioiodinewereinvestigatedby spiking
charcoalcartridgeswith 1311
at variousdepthsand
counting at three differentdistancesfrom a germanium detector.CommerciallyavailableTEDAimpregnatedcharcoalcartridges,with approximatedimensionsof 5.7-cmwide by 2.54-cmthick, were
usedin this work. The thicknessof the radioiodine
distribution correspondedto O-mm,3-mm, 5-mm,
10-mm,15-mm,and 25-mm (homogeneous).
All
Reprint of "From the Counting Room",Vol. 1,No.2, 1990
sampleswere prepared gravimetrically from 1311solutions calibrated to within :t 3%. The O-mm depth
was prepared by distributing the activity on a filter
using a pattern of drops over the active area to approximate a uniform distribution. The filter was
placed in the cartridge, in front of the charcoal bed
on the inlet side ("face"). Testsof the O-mm cartridge showed that 1311did not volatilize or redistribute during the test period of approximately 2 days.
Other thicknessesof 1311
spiked charcoal were prepared by adding a known amount of 1311activity to a
predetermined amount of charcoal, homogeneously
mixing it, and replacing it in the inlet side ("face") of
the cartridge to provide the desired depth. A barrier
was inserted to separatethe 1311
active from the inactive charcoal. The remainder of the inactive charcoal
was replaced and the cartridge resealed.
The counting systemconsisted of a 30-cm3 germanium detector interfaced to a computerized multichannel analyzer system. After eachcount, the net
peak area and efficiency for the 1311,
364.8-keV
gamma ray were determined. Each cartridge standard was counted with the "face" down for 1000seconds and processedto determine the net counts and
efficiency. The cartridges were then turned over with
the "face" side up ("flipped") and counted for an additional1000 secondsby resetting the ADC live time
to 2000 seconds,without erasing the spectrum. The
cumulative count was then processedto give the net
49
Radioactivity 6- Radiochemistry
The Counting Room: SpecialEdition
Figure
1
Efficiencies vs front-to-back ratios for "flip" and "face"
Figure 3
counting 3 mm from detector.
Figure 2
Relative efficiencies vs front-to-back ratios for "flip" and
"face" counting 3 mm from detector.
countingrate and efficiencyfor the 364.8-keV
gammaray. Front-to-backratiosweredetermined
from the following formula:
FCFB=
TC-FC
whereFB = front-to-backratio, FC = countsof the
364.8-keVgammaray from countingwith "face"
down, and TC is the total counts of the 364.8-keV
gammaray from the "face"and "flip" counts.
The front-to-backratio representsthe ratio of
the countingrate for a cartridgecountedwith the
50
Relative efficiencies vs front-to-back
ratios for "flip" and
"face" counting 3 cm from detector.
Figure 4
Relative efficiencies vs front-to-back ratios for "flip" and
"face" counting 6 cm from detector.
"face"sidedown,to the countingrate for a cartridge
countedwith the "face"sideup ("flipped"). This ratio correlateswith the distribution of radioiodinein
the cartridgewith the highestvalue of FB for "face"
loadedcartridges.The value of FB decreases
asthe
depthof loadingincreasesand approachesunity for
homogeneously
loadedcartridges(the ratio maynot
be exactlyone,sincemanycartridgesare not manufacturedso that the front and backare preciselythe
same).
The measuredfront-to-backratios and counting
efficienciesarepresentedin Table1. Counting efficienciesfor cartridgesloadedat variousdepthswere
Reprint of "From the Counting Room",Vol. 1,No.2, 1990
"CalibratingGermaniumDetectorsfor AssayingRadioiodinein CharcoalCartridges"
DanielM. Montgomery
plotted againstthe front-to-backratios.Theseresults
arepresentedin Figure1, for a sourceto detectordistanceof 3.0mm. As expected,countingcartridges
"face"down in closeproximity to the detector(Figure 1) is verysensitiveto loading.The 0-mm loaded
cartridgehasan efficiencythat is 1.70times that of a
homogeneously
loadedcartridge.
The effectsof the loading at source-to-detector
distancesof 3 mm, 3 cm and 6 cm, for both "flip"
and "face"counting,are shownin Figures2-4,where
relativeefficienciesareplotted againstthe front-tobackratios.The countingefficiencieswerenormalizedto 1.00for cartridgeefficiencieswith a 5-mm
thick loading.This assumesthat the cartridgeefficiency,with the 5-mm thick loading,is mostrepresentativeof the expectedloadingfor actualcharcoal
cartridgesamples.Upperand lowerlimits were
drawn at relativeefficiencies1.10and 0.90to representan allowabledeviationlimit of plus and minus
10%from the 5-mm value.The value of 10%is
somewhatarbitrary; however,a 10%systematicerror would still allow oneto meetthe acceptance
criteria usedby the U.s. NRC'sConfirmatoryMeasurementsProgram.The mostrestrictiveNRC criteria
would allowa measurement
to deviateup to 15%
from the NRC measuredvalue.
The problemsassociated
with countingsamples
verycloseto the detectorareillustrated in Figure2.
Although all points arewithin the limits when"flip"
counted,only the 2-mm cartridgewould be within
the limits when"face"counted.Carefulcomparison
of Figures2, 3 and 4 showthat the deviationfrom
the 5-mm cartridgeefficiencydecreases
asthe
source-to-detectordistanceincreases.
At a source-todetectordistanceof 3 cm, all points arewithin 5%
for "flip" countingand 3 valuesareoutsideof the
10%limits for "face"counting.At a source-to-detector distanceof 6 cm, all points arewithin 5% of the
5-mm value for "flip" counting,and all points except
for the 15-mmand homogeneously
loadedcartridgesare within the 10%limits for "face"counting.
As demonstratedby the abovemeasurements,
potential errorsassociatedwith a variabledistribution of radioiodinein cartridgesamples,canbe mini-
Reprint of "Promthe Counting Room",Vol. 1,No.2, 1990
mizedby usingthe "flip" method of counting even
at a sample-to-detector
distanceof 3 mm.
Because
the datapresentedin this paperaredependenton manyfactorssuchasthe detectordimensions,the type of charcoalcartridgeused,and sampling conditions,theseresultsmaynot be
quantitativelyapplicableto othersituations.However,thesestudiesdemonstratethe valueof the "flip"
countingmethod for a variety of conditions.Sitespecific studiesmaybe conductedto determinethe distribution of radioiodine in samplescollectedunder
actualoperatingconditionsusingthe techniquespresentedin this article.The mostappropriateloading
for a calibrationstandardcanthenbe selected.
{f a high degreeof accuracyis desiredor is necessary,a seriesof efficiencycurvesor tablescanbe determined for differentdistributions (i.e.,S-mm-, 10mm-, IS-mm-Ioadedcartridges)usingmixed
gamma-raystandards.The ratio of front-to-back activity for radioiodinewould be calculatedfor each
sampleand thenthe appropriateefficiencycurve or
tablewould be utilized for the analysis.
When "face"countingof cartridgesis practiced,
it maybe a goodideato periodicallymeasurethe ratio of front-to-back activity in charcoalcartridge
samplesto ensurethat calibrationsand counting
conditionsareappropriateto meetthe desiredmeasurementaccuracy.
Acknowledgements
The author thanksEstieW. Belvin for her assistance
in preparationof the charcoalcartridgestandards
and JohnMcCorveyfor countingthe standards.
References
1.
P.G. Voillegue, et al Evaluation of the Air-Vegetation-Milk Pathway for
1311at the Quad Cities Nuclear Power Station", u.s. Nuclear Regulatory Commission Report NUREGjCR-1600, November 1981.
2.
Personal Communication
(1990), D.N. Nix, Tennessee Valley
Authority, Sequoyah Nuclear Power Plant, P.O. Box 2000, SoddyDaisy, TN 37379.
3.
Personal Communication (1988), RC. McFarland, Analytics, Inc.,
1380 Seaboard Industrial Blvd., Atlanta, GA 30318.
R&R