Regulatory Guide 5.38

Nondestructive Assay of High-Enrichment Uranium Fuel Plates by Gamma Ray Spectrometry

ML13064A076
Person / Time
Issue date:	09/30/1974
From:	US Atomic Energy Commission (AEC)
To:
References
RG-5.038
Download: ML13064A076 (7)

U.S. ATOMIC ENERGY COMMISSIONREGULATORY GIDIRECTORATE OF REGULATORY STANDARDSREGULATORY GUIDE 5.38NONDESTRUCTIVE ASSAY OF HIGH-ENRICHMENT URANIUMFUEL PLATES BY GAMMA RAY SPECTROMETRYSeptember 1974JIDE

A. INTRODUCTION

Part 70 of Title 10 of the Code of Federal Regula-tions requires each licensee authorized to possess morethan 350 grams of contained U-235 to conduct aphysical inventory of all special nuclear material in hispossession at intervals not to exceed 12 months. Eachlicensee authorized to possess more than one effectivekilogram of high-enrichment uranium is required toconduct measured physical inventories of his specialnuclear materials at bimonthly intervals. Further, theselicensees are required to conduct their nuclear materialphysical inventories in compliance with specific require-ments set forth in Part 70. Inventory proceduresacceptable to the Regulatory +,aff for complying withthesc pi 'wisions of Part 70 are detailed in RegulatoryGuide 5.13. "Conduct of Nuclear Material PhysicalInvcntories."For certain nuclear reactors, the fuel consists ofhighly enriched uranium fabricated into flat or bowedplates. Typically, these plates are relatively thin so that asignificant percentage of the U-235 gamma rays ptle-trate the fuel cladding. When the measurement condi-tions are properly controlled and corrections are madefor variations in the attenuation of the gamma rays, ameasurement of the U-235 gamma rays can be used as anacceptable measurement of the distribution and the totalU-235 content of each fuel plate. In lieu of assaying theproduct fuel plates, fuel plate core compacts may beassayed through the procedures detailed in this guide,provided steps are taken to ensure the traceability andintegrity of encapsulation of each assayed fuel plate corecompact. This guide describes features of a gamma rayspectrometry system acceptable to the Regulatory stafffor nondestructive assay of high-enrichment uraniumfuel plates or fuel plate core compacts.

B. DISCUSSION

The number, energy, and intensity of gamma raysassociated with the decay of U-235 provide the basis fornondestructive assay of high-enrichment fuel plates bygamma ray spectrometry (Ref. 1). The 185.7-keVgamma ray is the most useful U-235 gamma ray for thisapplication; it is emitted at the rate of 4.25 x 104gamma rays per second per gram of U-235. Lower-energy gamma rays emitted by U-235 are less pene-trating and more sensitive to errors due to fluctuationsin clad and core thickness. In general, more accurate fuelplate assays may be made by measuring only the activityattributable to the 185.7-keV U-235 gamma ray.Assay measurements are made by integrating theresponse observed during the scanning of single fuelplates and comparing each response to a calibrationbased on the response to known calibration standards.1. GAMMA RAY MEASUREMENT SYSTEM1.1 GAMMA RAY DETECTION SYSTEM1.1.1 Gamma Ray DetectorHigh-resolution gamma ray detectors, i.e., intrinsic orlithium-drifted germanium, provide resolution beyondthat required for this assay application. While theperformance of such detectors is more than adequate,their low intrinsic detection efficiency, extensive opera-tional and maintenance requirements, and high costmake them unattractive for this application.Most ,sodium iodide [Nal (TI)] scintillation detectorsare capable of sufficient energy resolution to be used forthe measurement of the 185.7-keV gamma rays. Thedetector diameter is determined by the fuel plate widthand the scanning method selected (see Section B.I.2 ofthis guide). The thickness of the Nal crystal is selectedto avoid unnecessary sensitivity to gamma rays above the185-keV region which produce a background in the185-keV energy region as a result of Compton scattering.For measurements to be reproducible, it is necessaryto assure that the detection system is stabilized on theUSAEC REGULATORY GUIDES Cop.- of published guids may be obtained by request indicintng the dinirsomdeired to the US. Atomic Erw Commismson. Vlhington. D.C. 20545.RegulatmrV Guide are ssued to iescribe and make availabte to the puublic Attlntion: Director of Regulatory Standstills. Comments wed suggestions fornalhods c.acptable to tht AEC Regulatory st&ff of imiplemnting specific parts of ifl3hmtflts in thes guides we o fiurqlpd and should be sent to the Secretarythe s regulations, to delineate technqlus .sead by the staff in of the Comnssmion, US. Atomic EnoqW Commission. Washington, D.C. 20545.oealuat-0n specific problems or postulated accidents, or to provide guidenws to Attention: Do-kteting and Sorvece Section.aml)icents. Regulatory Guides wit not substitutes for regulations and comNpflaniowith them is not Methods and sofutions diffrt from tfeor lit out in The guides are ,ssued in the following ten broaed divisions:t. guides well bea asipgtobl if they piroind a basis for the findingl requisits tothe or oiminuenOt of a permit or licnese by the Commtissson. 1. Powr Reectorn 6. Products2. Research and Test Ractors 7. Tranpomrtation3. Fuels isid Materi- Fac;itisi a. O=upetiona! Meeilhguides will be revised periodicafly. as approprate. to accomn oal 4. Environmenetal end Siting 9. Antitrust Reviewcomments and to reflect new ,ntormition or expoerine. 5. Mot'earits and Plant Protection 10. Generaf intended portion of the gamma ray spectrum duringmeasurements. Internally "seeded" Nal crystals whichcontain a radioactive source (typically Am-241) toproduce a reference energy pulse art commerciallyavailable. The detection system is stabilized on thereference, and the amplifier gain is automatically cor-rected to assure that that energy and the rest of thespectrum remain fixed in position..1.2 Gamma Ray CollimatorTo ensure that the only gamma ray activity detectedoriginates from a well-defined segment of the fuel plate,the detector is shielded from extraneous backgroundradiations and collimated to define the area "seen" bythe detector crystal. The collimator consists of a disk ofappropriate shielding material. A slit is machinedthrough the center of the disk which will allow onlythose gamma rays emitted within the slit opening tostrike the detector. The disk thickness is a minimum ofsix mean free path lengths to effectively stop all gammarays emitted from outside the view area. To preventgamma rays from striking the crystal around the edges ofthe collimator disk, the disk diameter exceeds the crystaldiameter by at least twice the crystal depth.The probability of detection for gamma rays emittedat the center of the collimator slit is greater than that forgamma rays emitted near the ends of the slit. This effectbecomes increasingly important at small detector-to-plate spacing, especially when scanning near the edge ofa plate. To minimize this detection nonuniformity andto minimize the sensitivity to jitter, the detector-to-platedistance can be made large, especially with respect to thedimensions of the slit opening. As an alternative meansof reducing the detection nonuniformity-across the slit,the slit opening can be divided into channels by insertinga honeycomb baffle into the slit or by fabricating thecollimator by drilling holes through the disk in a patternwhich ensures that each hole is surrounded by aminimum wall thickness of 0.2 mean free path length. A7:0-cm-thick iron disk with holes less than 0.5 cm indiameter drilled in a pattern having 0.2 cm of wallbetween adjacent holes is one example of a collimatorthat would perform satisfactorily. A large number ofsmall-diameter holes is preferable to a few large-diameterholes.1.1.3 Multiple DetectorsSeveral detectors may be used to shorten the mea-surement time. The detectors can be positioned tosimultaneously measure different segments of a singlefuel plate or to simultaneously measure additional fuelplates. In some cases it may be useful to sum theresponse from two detectors positioned on oppositesides of a plate to increase counting efficiency. In suchcases it is essential that the response of such detectors bebalanced-and checked at frequent intervals.1.2 SCANNING TECIINIQUESIt is critical that the scanningapparatds for ,'oviog t:,plates relative to the detector provide a uniforn;.reproducible scan. The importance of a well-constructed,mechanically stable conveyor cannot be overemphasized.Either the detector can be moved and the plate heldstationary, or the plate can be moved past a fixeddetector. Care must be exercised to maintain fldetector-to-plate spacing within close tolerance,minimize errors caused by the inverse-square dcpen(!of detection on distance. This is especially important inthe case of close spacing, which is sometimes desirable tomaximize the count rate. Various commercial conveyingsystems have been used and found to be adequate. Such.systems may significantly reduce the cost of de..,7!i, gand building new scanning mechanisms. High-precisiontool equipment such as milling machines, lathes, and x-yscanning tables can be investigated. Numerically con-trolled units offer additional advantages when they canbe incorporated into a scanning system. This is particu-larly true when an automated scanning system is beingdeveloped.Fuel plate core compacts may be sufficiently smallpermit total assay without scanning in a fixed-geometrycounting system. The scanning techniques for fuel platesdiscussed in the following subsections can also be usedfor core compacts when total fixed compact counting isnot possible.1.2.1 Linear Total ScanThe detector collimation consists of a rectangularopening which extends across the width of the fuelplates beyond the edges of the uranium core containedwithin the plate cladding. Scanning the total plate isaccomplished by starting the count sequence on the endof a -plate and continuing to count until the entire lengthof plate has been scanned.To ensure that gamma rays emitted anywhere acrossthe face of the fuel plate have an equal probability ofbeing detected, it is necesary that the diameter of thedetector crystal exceed the plate width or that thedetector be positioned away from the plate.Use of the spot oi cizcalar collimator scan techniqueeliminates or reduces to insignificance most of theseedge effects.1.2.2 Sweeping Spot Scan (Ref. 2)If the collimator channel width is smaller than thefuel plate width, the viewing area (spot) can be sweptacross the plate as the detector scans along the length ofthe plate. This scanning technique can be readilyadapted to scanning bowed plates through the use of acam which is designed to maintain the detector-to-platedistance constant over the entire geometry of the fuelplate. The collimator channel dimensions can be selectedto provide compatible information on the uniformity of5.38-2 the fuel plate which is frequently obtained by comparingfixed (static) spot counts at a variety of locations toreference counts.1.2.3 Sampled Increment AssayWhen used in conjunction with radiographic dimen-sional measurements performed on all fuel plates, theU-235 content of a fuel plate can be measured byscanning the ends of each fuel plate and sampling thebalance of the plate. It is necessary to measure thedimensions of the fuel core loading radiographically,through gamma ray scanning along the length of theplate, or by spot scanning the fuel plate ends andmeasuring the distance between end spots where the fuelloading stops. The U-235 content of the plate is thendetermined by averaging the results of sample spotmeasurements of the U-235 content per unit area at anumber of sites along the plate and multiplying thisaverage value by the measured area of the fuel core. Theradiograph of each plate is examined to ensure that thecore filter is uniform.The collimator shape and dimensions can be selectedto provide compatible information on the uniformity ofthe fuel plate.1.3 COMPUTER CONTROLThe reproducibility of measurements can be im-proved and the measurement time per fuel plate can bereduced by using a computer to control the fuel platescanning operation. The computer can be used tocontrol data acquisition by accumulating counts ac-cording to a predetermined scheme. Also, the computercan be used for data analysis, including backgroundcorrections and intermachine normalization, calibration,error analysis, and diagnostic test measurements andanalyses. Report preparation and data recording forsubsequent analysis are also readily accomplishedthrough an appropriately designed computer-controlledsystem.2. INTERPRETATION OF MEASUREMENT DATAThe three factors discussed below may give rise tosignificant errors in interpreting measurement data.2.1 ENRICHMENT vARIATIONSLicensees authorized to possess highly enricheduranium are required to account for element and isotopeas prescribed in §70.51. Under the conditions detailedin this guide, the U-235 content of individual plates ismeasured. To determine the total uranium content ofeach plate, the U-235 enrichment must be known fromseparate measurements.Enrichment variations may alter the radiation back-ground in the gamma ray energy region of interest.Uranium-238 decays by alpha-particle .emission toTh-234. Thorium-234 then decays by beta-particle emis-sion with a half-life of 24.1 days to Pa-234 which, inturn, decays by beta-particle emission to U-234. Ap-proximately 1% of the Pa-234 decays are followed byhigh-energy (e.g., 1001 keV, 766 keV) gamma rays.These gamma rays frequently lose energy throughCompton scattering and may appear in the 185-keVspectral region. It is important to note that activity fromPa-234 may be altered by disturbing the equilibriumbetween U-235 and Th-234, as frequently occurs inuranium chemical conversion processes. The interferencedue to variations in U-238 daughter activity becomes lessimportant as the enrichment of U-235 increases. Atenrichment levels above 90%, this problem can essen-tially be ignored.2.2 RADIATION ATTENUATIONThe number of U-235 gamma rays which escape fromthe fuel plate (and are thus available for detection)without losing energy depends on the characteristics ofthe fuel plate core and cladding. Gamma rays fromU-235 are attenuated in the uranium, in the cladding,and in the inert material that may be added with theuranium to form the core of the fuel plate. Throughwell-controlled product tolerance limits, each of thesepotential sources of signal variability can be controlledto permit accurate accountability assays.2.2.1 Self-AttenuationThe uranium photon attenuation coefficient forgamma ray energies corresponding to U-235 emissions isquite large (Ref. 3). Small changes in uranium densityresulting from increased fuel loading or from variationsin the manufacturing process can significantly changethe number of gamma rays which escape from the fuelplate.2.2.2 Gadding AttentuationSmall variations in cladding thickness may causesignificant attenuation variations. Variations in claddingattenuation can be measured by a simple gamma rayabsorption test using thin sheets of cladding material asabsorbers and varying the clad thickness over the rangeof thicknesses to be encountered in normal productvariability.2.2.3 Core Friler AttenuationRadiation intensity measurements may be made ofplates fabricated with different ratios of uranium tofiller to show the effects of this type of attenuation. Ifsignificant effects are noted, plates can be categorized bycore composition characteristics and the assay systemcan be independently calibrated for each category offuel plates.5.38-3 2.2.4 Attenuation CorrectionsWhen the thickness of the core and cladding of eachplate is known, an attenuation correction can be appliedto improve the accuracy of the assay. Ultrasonic gaugingmay provide such a measure, provided the metallo-graphic zones within the plate are sufficiently defined toprovide a detectable interface.The alternative attenuation- correction is based on amicrometer measurement of the total thickness of eachplate. The clad thickness of a plate is estimated bysubtracting the mean core thickness of the productplates, which is determined by periodically samplingproduct plates and cutting a cross section to permitvisual measurement of clad and core thickness.2.3 INTERFERING RADIATIONSAs noted in Section B.2.1 of this guide, an internalbackground variation may arise from changes in theamount of U.238 present in a fuel plate or from changesin the ratio of Th-234 to U-238 resulting from fuelmanufacturing processes. Fluctuations in the internalbackground cause the response of the unknown items tobe different from the calibration standards, therebycreating a measurement bias. Such interferences can becompensated by measuring additional regions of thegamma ray spectrum.Other interfering radiations may come from externalsources, from fuel plates awaiting assay, or from nearbyradiation sources used for other measurements. This isnot expected to be a major problem and can becontrolled through (1) removing radiation sources,(2) shielding the detectors, and (3) monitoring the back-ground at frequent intervals.3. CALIBRATION AND VERIFICATION3.1 INITIAL OPERATIONSCalibration and the verification of assay predictions isan ongoing effort where performance is periodicallymonitored and the calibration relationship is modified toimprove the accuracy of assay predictions. During initialoperations, two means of basing preliminary calibrationsare appropriate.3.1.1 Foil Calibration TechniqueMethods for calibrating scanning systems for high-enrichment uranium fuel plates through the assay ofprepared uranium and clad foils are described in Refer-ence 2. This method may be used in place of or inaddition to the -technique described in the followingsubsection.uranium, U-235, inert matrix, and cladding are accu-rately measured and that these parameters bracket thenominal range of product plates anticipated to fallwithin manufacturing tolerances.3.2 ROUTINE OPERATIONSThe performance of the assay system is periodicallymonitored to ensure that the performance of the assaysystem has not shifted since its last calibration. Controllimits for acceptable performance can be established forthe response to an appropriate working standard. Thecontrol chart of the responses to the working standardcan be checked for indications of short-term instrumentdrift or malfunction. The control chart can also beanalyzed to detect long-term shifts within themeasurement-to-measurement control limits that may becorrected by recalibrating the system. Severe changes ininstrument performance are investigated promptly andtheir causes remedied.To ensure that the calibration remains valid duringnormal operations and that accuracy estimates arerigorously justified, assay predictions are periodicallycompared with more accurate measurements of thecontent of typical fuel plates (see Section C.4 of thisguide). Guidance on methods to relate this assay to thenational measurement system and to reconcile verifi-cation measurements will be addressed in separateregulatory guides.*

C. REGULATORY POSITION

The content and distribution of U-235 in high-enrichment uranium plates can be measured through thegamma ray assay methods described in this guide.Combining this measurement with the results of anindependent measurement of the U-235 enrichmentenables the total uranium content of the fuel plates tobe measured. The factors presented below should betaken into consideration for this assay method to beacceptable to the Regulatory staff.I. MEASUREMENT SYSTEM1.1 GAMMA RAY MEASUREMENT SYSTEM1.1.1 Gamma Ray DetectorA thallium-activated sodium iodide scintillationdetector or series of detectors is recommended for thisassay application. When more than one detector is to beincorporated into the scan system, the performancecharacteristics of the detectors should be matched. Thediameter of the crystal should be larger than theprojected view onto the crystal face through the*For example, regulatory guides related to measurement qualityamurance and calibration of nondestructive a&say systems arebeing developed.3.1.2 Fabricated Calibration PlatesCalibration standard fuel plates can be fabricatedusing special precautions to ensure that the amounts of5.38-4 collimator channel. The thickness of the crystal shouldhc noi more than onc inch. The crystal should contain aninternal cesium iodide seed which is doped with asuilable alpha-emittcr for spectral stabilization. The seedshould produce approximately 1,000 counts per secondat the reference energy.1.1.2 CollimatorA collimator should be fabricated of appropriategamma ray shielding material such as iron, lead, ortungsten. The shielding should completely surround thedetector and photomultiplier assembly and should besufficiently thick to completely block extraneous radi-ations from the detector. The response variation fromthe center of the collimator opening to its edge shouldbe less than 1%.1.1.3 Electronic ApparatusAll electronic systems should be powered by filtered,highly regulated power supplies. The ambient tempera-ture and humidity in the vicinity of the scanning systemshould be controlled so that permitted fluctuations donot significantly affect the assay measurements. Allelectronic circuitry in signal-processing componentsshould feature temperature compensation. Residual sen-sitivity to fluctuations in the, ambient environmentshould be tested and monitored periodically.The capability for multichannel gamma ray pulseheight analysis with cathode ray tube spectral displayshould be provided. Signal-processing electronics capableof stabilizing on the alpha radiations emitted within thedoped cesium iodide seed should be provided to stabilizethe energy spectrum.1.2 SCANNING SYSTEMA mechanically sound, highly reproducible scanningsystem should be employed. Scanning should be accom-plished by one of the three techniques discussed inSection B. 1.2 of this guide.1.3 COMPUTER CONTROLA dedicated minicomputer to control data acquisi-tion, analysis, calibration, diagnostic testing, and reportpreparation -should be employed for this assay appli-cation.1.4 MULTIPLE SCANNING ASSAY SYSTEMSWhen more than ornc scanning system is employed,assay response should be normalized so that eachinstrument provides consistent results. Verification datato establish the systeniatic assay error for each assaysystem should be obtained with the same plate.2. MEASUREMENT INTERPRETATION2.1 ENRICHMENT VARIATIONSProcedures should be developed to ensure that theenrichment of the plates being scanned is knownthrough separate measurements. Fuel plates generallysatisfy the gamma ray penetrability criteria for quantita-tive U-235 assay; they do not satisfy the criteria fornondestructive enrichment measurement through gammaray spectrometry.* Facilities processing more than oneuranium enrichment should maintain strict isotopiccontrol and characterize the enrichment through appri.,-priate measurement methods.2.2 ATTENUATION CORRECTIONSAttenuation variations arising from plate-to-platechanges in core thickness, composition,-and clad thick-ness should be determined over the range of producttolerance specifications. When such variations cause theassay error to exceed the error realized without thevariations by 50% or more, procedures should beimplemented to measure and apply a correction to theassay of each plate.2.3 RADIATION INTERFERENCESA clear plastic template which shows an acceptablespectrum display should be prepared. When radioactiveinterference may be encountered, the assay spectrumshould be compared at appropriate intervals to thereference spectrum for indications of interference. Bick-ground radiation should be measured periodically duringeach operating shift.3. MEASUREMENT CALIBRATIONDuring initial operations, the assay system should m,calibrated either by the foil calibration method or withspecially prepared sample fuel plates as described mnSection B.3.1 of this guide.4. RANDOM AND SYSTEMATIC ASSAY ERRORS4.1 RANDOM ERROR ESTIMATIONA replicate assay program should be established togenerate data for the evaluation of random assay errOrsduring each material balance period. During each bi-monthly interval, a minimum of fifteen plates should beselected for replicate assay. The second assay of eachplate selected for replicate assay should be made at leastfour hours after the first assay. Replicate assay diffet-ences should be collected and analyzed at the end (if the*Criteria for uranium gamma ray enrichment arffgiven in Regulatory Guide 5.21, "Nondestructrvr ltiranwurn.235Enrichment Assay by Gamma Ray "5.38-5 material balance period. The single-measurement stan-dard deviation of the relative replicate assay differencesshould be computed as described in Reference 4.4.2 SYSTEMATIC ERROR ESTIMATIONThe systematic error associated with the assay of allfuel plates fabricated during a material balance periodshould be determined through one of the procedures*presented below.4.2.1 Propagation through the Calibration FunctionTo estimate the systematic assay error through thecalibration function, the calibration should be based onthe regression analysis of an appropriate function to thecalibration data. Uncertainties in the reference values ofthe calibration standards should be factored into the fit,and the errors propagated as demonstrated in Reference5.To ensure the validity of the predictions, the stableperformance of the instrument should be monitored andnormalized through the response to appropriate workingstandards which are assayed at frequent intervals. Thefrequency for assaying working standards should bedetermined through testing, but should not be lowerthan one test during each two-hour assay interval forspot response stability and one full scan test during eachoperating shift. Indications of shifting instrument perfor-mance should be investigated and remedied, and theinstrument should be recalibrated to ensure the validityof subsequent measurements.In order to ensure that the calibration standardscontinue to adequately represent the unknown fuelplates, key production parameters which affect theobserved response should be monitored through separatetests. Data should be compiled and analyzed at the closeof each material balance period. When a productionparameter shifts from previously established values, theimpact of the shift on the response of the assayinstrument should be determined through an appropriateexperiment or. calculation (Ref. 6). A bias correctionshould be determined and applied to all items assayedfrom the point of the parameter change. The uncertaintyin the bias should be combined with the systematic errorpredicted through the calibration function. When thebias exceeds 3% of the plate contents in a single materialbalance period, when a trend of 1.5% or more isobserved in three consecutive material balance periods,,)r when the uncertainty in the observed bias is sufficientto increase the limit of error of the assay above 0.5%,new calibration standards should be obtained, and thescanning system should be recalibrated.As a further check on the continued validity of thecalibration standards, a program to periodically intro-duce new calibration standards should be implemented.*These methods will be discussed in detail in a regulatory guide,n preparation entitled "Calibration and Error EstimationPr:ccduir"s fcr Nondestructi-e Asay."A minimum of one new calibration standard fuel plateshould be introduced during each six-month period.4.2.2 Comparative EvaluationWhen two measurements are made on each of a seriesof items and the accuracy of ono of the methods used isconsiderably greater than the other, the correspondingpredictions can be compared to establish an estimate ofbias between the measurement methods and to estimatethe error. associated with the lens-accurate measurementmethod. To precisely determine the systematic error inthe nondestructive assay, the fuel plates selected forcomparative measurements should be randomly selectedbut should span the range of U-235 contents en-countered in normal production. The selected fuel platesmay be rejected from the process stream for failing tomeet quality assurance requirements. Each plate shouldbe repeatedly assayed to reduce the random asay errorto less than 10% of the estimated or previouslyestablished systematic error. To determine its U-235 andtotal uranium content, the plate should be completelydissolved and the resulting solution should be analyzedby high-accuracy chemical and mass spectrometric pro-cedures.For one material balance period during the initialimplementation of this guide; a product fuel plateshould be randomly selected twice each week for anaccuracy verification measurement. Following this initialimplementation period, facilities manufacturing 100 ormore fuel plates per week may reduce the verificationfrequency to one plate per week and pool the verifi-cation data for two consecutive material balance periods.LoW-throughput facilities manufacturing lesa than 100plates per week should verify at least 4 plates permaterial balance period through the procedures de-scribed above. At the close of each material balanceperiod, data should be pooled to include only the 15most current data points.When the U-235 contents of the plates assayed using acommon calibration relationship varies over a ranpg of+/-5% or more about the average of all plate loadings, thesystematic error should be estimated as described inparagraph 1. below; when plate loadings are tightlyclustered about a nominal value, the systematic errorshould be estimated as described in paragraph 2.1. At the close of the reporting period, the assayvalue for each plate is plotted against the verifiedquantity. The verification data plot is examined forindications of nonlinearity or obvious outlier data.Anomalous indications should be investigated andremedied.A linear regression analysis should be performed onthe comparison data. The intercept should be testedagainst zero for an indication of a constant measurementbias. The slope shouid be tested against unity for anindication of a proportional bias. When bias is indicated,assays performed during the preceding operating periodshould be compensated. The systematic error should be5.38-6 estimated as the standard error associated with theverification line.2. When all plates contain essentially the same U-235content, the difference in the mean content valuesshould be tested against zero as an indication of bias,and the systematic error associated with an inventory ofplates should be quoted as the standard deviation of themean difference. For individual plates, the systematicerror should be quoted as the standard deviation of thedifference distribution.5. CORE COMPACT ASSAYFinal product assay in high-enrichment fuel platemanufacturing can also be accomplished through assay-ing each core compact following the procedures detailedin this guide and the following supplemental criteria:1. Each core compact should carry a unique identifi-cation. Accountability records should be created foreach compact. The fuel plate should carry an identifica-tion corresponding to the compact identification.2. Each fuel plate should be radiographically exam-ined to ensure that the entire compact has beenencapsulated.3. Each fuel plate should be checked with a gammaray probe to qualitatively ensure that the plate core isuranium of the nominal product enrichment.4. Calibration and error evaluation should follow theprocedures for fuel plate assay.REFERENCESI. J.E. Cline, R.J. Gehrke, and L.D. Mclsaac, "GammaRays Emitted by the Fissionable Nuclides and Asso-ciated Isotopes," ANCR-1029 (July 1972).2. N.S. Beyer, "Assay of U-235 in Nuclear Reactor FuelElements by Gamma Ray Scintillation Spec-trometry," Proc. 4th Intl. Conf. on NondestructiveTesting, London, 1963.3. J.H. Hubbell, "Photon Cross Sections, AttenuationCoefficients, and Energy Absorption Coefficientsfrom 10 keV to 100 GeV," Nat. Bur. Stds.NSRDS-NBS 29 (1969).4. John L. Jaech, "Statistical Methods in NuclearMaterials Control," TID-26298 (1973).5. American National Standard N15.20, "Guide toCalibrating Nondestructive Assay Systems," inpreparation. Copies of the draft standard may beobtained from Institute of Nuclear Materials Manage-ment, 505 King Avenue, Columbus, Ohio 43201(Attention: H.L. Toy).6. See, for example, R.A. Forster, D.B. Smith, and H.O.Menlove, "Error Analysis of a Cf-252 Fuel Rod AssaySystem,'" LA-5317 (1974).5.38-7