Regulatory Guide 5.53

Qualification, Calibration, and Error Estimation Methods for Nondestrutive Assay.

ML13064A081
Person / Time
Issue date:	02/28/1984
From:	Office of Nuclear Regulatory Research
To:
References
RG-5.053, Rev. 1
Download: ML13064A081 (10)

Revision 1*February 1984U.S. NUCLEAR REGULATORY COMMISSIONREGULATORY GUIDEOFFICE OF NUCLEAR REGULATORY RESEARCHREGULATORY GUIDE 5.53(Task SG 049-4)QUALIFICATION, CALIBRATION, AND ERROR ESTIMATIONMETHODS FOR NONDESTRUCTIVE ASSAYA. INTRODUCTIONSection70.58, "Fundamental Nuclear Material Con-trols," of 10 CFR Part 70, "Domestic Licensing of SpecialNuclear Material," requires certain licensees to establish ameasurement quality assurance program for materialcontrol and accounting. Specifically, paragraph 70.58(f)requires that a program be established, maintained, andfollowed for the maintenance of acceptable measurementquality in terms of measurement bias and for the evalua-tion and control of the quality of the measurementsystem.Nondestructive assay (NDA) constitutes a uniquemeasurement technology. When applied under appropriaterigorous controls, it can enhance the ability of the materialcontrol and accounting system to detect unaccounted-forloss or diversion of special nuclear material (SNM) tounauthorized uses. This guide describes methods andprocedures acceptable to the NRC staff for meeting theprovisions of paragraph 70.58(f) of 10 CFR Part 70 as itrelates to the use of nondestructive assay.Any guidance in this document related to informationcollection activities has been cleared under 0MB ClearanceNo. 3150-0009.B. DISCUSSIONNondestructive assay has been applied to virtuallyevery chemical or physical form of special nuclear materialencountered in contemporary reactor fuel processing.Special considerations are required to achieve high-accuracy assay results and to properly estimate the errorsassociated with NDA applications. Recognizing theseconsiderations, the American National Standards Institutehas developed a standard, ANSI N15.20-1975, "Guide toCalibrating Nondestructive Assay Systems."1 This standard1Copies may be obtained from the American National StandardsInstitute, 1430 Broadway, New York, New York 10018.was reviewed and reaffirmed without modification in1980. This guide endorses the entire standard as supple-mented in the regulatory position.C. REGULATORY POSITIONThe methods, procedures, and guidance relating to theapplication of NDA in ANSI N1S.20-1975, "Guide toCalibrating Nondestructive Assay Systems," are accept-able to the NRC staff for use in material protectionprograms as supplemented by the following.1. METHOD SELECTIONPrior to selecting an assay method, a study should bemade to determine the required performance for that appli-cation. The specific NDA method should be selected toprovide results that are compatible with plant materialbalance requirements. Methods to enhance attainableperformance should be considered (e.g., container selec-tion and packaging procedures for bulk materials discussedin Regulatory Guide 5.11, "Nondestructive Assay ofSpecial Nuclear Material Contained in Scrap and Waste"2).2. INSTRUMENT SPECIFICATIONSAn evaluation of each new NDA application, includingthe proposed placement of the instrument, should beconducted prior to procurement. Studies of existing NDAapplications should be conducted periodically to evaluatetheir performance and substantiate the basis for theircontinued use. The impact of each of the measurement-to-measurement sources of error encountered in practiceor anticipated should be established as a part of each ofthese efforts.The substantial number of changes in this revision has made itimpractical to indicate the changes with lines in the margin.2A proposed revision to this guide has been issued for commentas Task SG 043-4.USNRC REGULATORY GUIDESRegulatory Guides are issued to describe and make available to thepublic methods acceptable to the N RC staff of implementingspecific parts of the Commission's regulations to delineate tech-niques used by the staff in evaluating specific 'problems or postu-lated accidents or to provide guidance to applicants. RegulatoryGuides are substitutes for regulations, and compliance withthem is not required. Methods and solutions different from those setout in the guides will be acceptable if they provide a basis for thefindings requisite, to the issuance or continuance of a permit orlicense by the Commission.This guide was issued after consideration of comments received fromthe public. Comments and suggestions for improvements in theseguides are encouraged at all times, and guides will be revised, asappropriate, to accommodate comments and to reflect new informa-tion or exoerience.Comments should be sent to the Secretary of the Commission,U.S. Nuclear Regulatory Commission, Washington, D.C. 20555,Attention: Docketing and Service Branch.The guides are issued in the following ten broad divisions:1. Power Reactors 6. Products2. Research and Test Reactors 7. Transportation3. Fuels and Materials Facilities 8. Occupational Health4. Environmental and Siting 9. Antitrust and Financial Review5. Materials and Plant Protection 10. GeneralCopies of issued guides may be purchased at the current GovernmentPrinting Office price. A subscription service for future guides in spe-cific divisions is available through the Government Printing Office.Information on the subscription service and current GPO prices maybe obtained by writing the U.S. Nuclear Regulatory Commission,Washington, D.C. 20555, Attention: Publications Sales Manage A decision should be made to reduce each potentiallysignificant source of error through (1) appropriate instrumentdesign considerations, (2) operational controls, or (3) supple-mentary measurements made to establish bias corrections(see also Reference 1). Instrument procurement specifica-tions and operational instructions should be developed andfollowed to reflect each error-reduction decision.To minimize operator-related errors and to promote uni-farm measurement practices, NDA instruments used for fixed-station operations should be automated to control (1) dataacquisition and analysis, (2) diagnostic testing of instrumentperformance stability and calibration validity, and (3) calcu-lation of associated error estimates. It is recognized that,for some less complicated NDA measurements, consistencyof operation may be achieved through the implementationof carefully written and tested standard operating procedures.Instruments should be tested to ensure that they meetprocurement specifications prior to calibration.3. OPERATORSAdequate operator qualification requirements arecrucial to proper calibration and effective measurementcontrol of an NDA instrument. The qualification require-ments should include a general knowledge of the assaytechnique being used and an understanding of the typicalbehavior and the limitations of the instrument and thetechnique. A knowledge of the external factors to whichthe measurement technique is sensitive (factors such asmatrix composition, background, material forms, andcontainer type) is also necessary. Only then can properstandards be chosen for calibration and measurementcontrol data be interpreted effectively.If the operators have only a general knowledge ofexternal factors, the NDA measurement program must beoverseen by a director with a detailed knowledge of allrelated factors. Only qualified operators should be permit-ted to make SNM assays.4. STABILITY TESTINGA preventive maintenance program should be devisedand implemented to ensure the long-term stability andreliability of each instrument.As part of an ongoing program of measurement control,more working standards3 should be fabricated to period-3Working standards are used to check the performance of anNDA instrument. They should be nominally representative of theitems to be assayed. They should be fabricated and handled toensure their internal integrity so that deviations in the measuredresponse of the assay system can be attributed to the instrument.As stated in ANSI Ni5.20-1975, working standards built to meetthese requirements are not acceptable as calibration standards.Calibration standards are defined in ANSI N15.20-1975 as "physicallyand chemically similar to the items to be assayed, for which themass of the nuclide(s) of interest and all properties to which themeasurement technique is sensitive are known." Calibration standardscan be used as working standards, but working standards cannotbe used as calibration standards. When calibration standards meetthle requirements for working standards, licensees may elect tomaintain only calibration standards. However, calibration standardsmay deteriorate through extensive use or may be prohibitivelyexpensive for stability monitoring purposes.ically test the performance stability of the instrument.Each working standard should contain a different amountof the species of SNM to be assayed. Current licensingreview criteria require the use of four working standards.On a rotating basis, one or two of these standards are usedto check the system each day.It should be noted that, in general, a working standardneed not be fabricated from the same type of materialbeing assayed. Even a material from a different radioactivespecies may be acceptable if carefully chosen and pre-pared. The essential requirements for a working standardare that (1) the radiation characteristics of the working stan-dard are sufficiently stable to ensure that fluctuations ininstrument response during measurement control can con-fidently be attributed to aberrations in instrument param-eters rather than to variations in source characteristics and(2) the working standard induces a response in the NDAinstrument that is characteristic of the expected response toreal assay material. The most convenient means of achievingthis "representative response" characteristic is to use mate-rial similar to the material that will be assayed.A study should be made to determine the frequencywith which the working standards are to be measured. Ifthere is some instability, a working standard should bemeasured before and after each assay of an unknownitem, and the calibration should be normalized to reflectthe average of the before-assay and after-assay tests. Ingeneral, excessive instabilities should not be tolerated;they should be remedied by frequent recalibration. Ifinstabilities persist, an alternative technique, an alternativeinstrument, or another measurement environment shouldbe sought. In any case, a working standard should bemeasured a minimum of twice per shift, once at thebeginning of the shift and again at some random timeduring the shift.As a general principle, working standards should be runwith a frequency directly proportional to the frequencyof measurements (i.e., increase as the measurementfrequency increases and decrease as the measurementfrequency decreases). Also, the quantity of SNM in thestandards measurements should closely follow the quanti-ties of SNM being measured (ie., the frequency of high-SNM-content working standards measurements increasesas the frequency of assays of like items increases). Theseprocedures provide a useful estimate of the bias whendetermined at the end of the inventory period. In addition,working standards should be run frequently enough foreach measurement system so that no one system couldcontribute excessively to the inventory difference (ID) bybeing out of control for an extended period. A minimumof 16 control measurements should be made per materialbalance period. Assuming two systems having equal materialflows in SNM quantity and number of items, the systemwith the greater uncertainty per measurement should runmore working standards to reduce its potential impact onthe ID.Each response to a working standard should be comparedto the previous calibration data as well as to the mean value5.53-2 of previous measurements of that working standard (underthe same calibration) that were accumulated during thepreceding material balance period. The difference should beplotted on a control chart. Control chart limits should beestablished at 0.05 and 0.001 levels of significance. When-ever control data exceed the 0.05 control limits, the testshould be repeated. Whenever the control data exceed the0.001 control limits, normal assay operations should cease.Normal operations should not resume until the out-of-control performance has been remedied and the instrumenthas been recalibrated.The control chart of the working standard responsesshould be examined at frequent intervals to detect indica-tions of drift, which should be compensated. The frequencyfor such examinations should be determined by the operat-ing characteristics of each instrument. The minimumfrequency for examining the control chart of a regularlyused instrument for indications of drift should be once perweek.5. CALIBRATIONCalibration of NDA instruments should be accomplishedby measuring the response to calibration standards asdescribed in ANSI N15.20-1975. The nuclear materialcontent of these standards should be characterized throughestablished assay procedures (e.g., chemical assays) that arecalibrated relative to national standards or nationallyaccepted measurement systems. The calibration standardsshould represent the unknown items in all physical andchemical characteristics that affect the response of theinstrument. Calibration data should be obtained by averag-ing the responses from repeated measurements of thecalibration standards and should be corrected to removeobserved nonrandom variations.Recalibration of an instrument is required following repairor replacement of parts if measurement of one or moreworking standards shows the instrument response to havechanged. In addition, the calibration should be checkedfollowing a power outage or any unusual mechanical orelectrical shock to the system. Recalibration data are alsorequired if the characteristics of the items to be assayedchange to the extent that previous calibration standards nolonger adequately represent the unknown items.Criteria for segregating and packaging different forms ofSNM should be developed and implemented. Each materialcategory should be established to enhance assay perform-ance, consistent with safety requirements and subsequentprocessing needs. Guidance for material categorizationis provided in Regulatory Guides 5.11, "NondestructiveAssay of Special Nuclear Material Contained in Scrap andWaste,"2 and 5.34, "Nondestructive Assay for Plutonium inScrap Material by Spontaneous Fission Detection."4For all categories of materials to be assayed, with theexception of small-content miscellaneous categories (e.g.,4A proposed revision to this guide has been issued for commentas Task SG 046-4.furnace liner bricks, contaminated tools, or machine parts),the calibration relationship should be determined by asuitable method such as a least-squares fit to an appropriatefunction as described in ANSI N15.20-1975. The graphicalcalibration method is acceptable only for miscellaneouscategories of material that contain a total of no more than0.1 effective kilograms of SNM in each category during amaterial balance period. The combined contribution fromall assays calibrated through the graphical method shouldbe less than 10 percent of the total plant standard error(estimator) of inventory difference (SEID).6. CALIBRATION STANDARDSCalibration standards should be obtained to serve as thebasis for the initial calibration of each instrument for eachseparate measurement technique or category of material.The number of standards in each set should be greaterthan the number of free parameters in the calibrationfunction for that set. It is recognized that, in some specialcases, one set of calibration standards may suffice for morethan one measurement technique or material category withproper analysis of the raw calibration data. Furthermore, ifthe NDA instrument is intended for use over a very narrowrange of SNM loadings, a more restricted range of SNMcontent in the calibration standards (confined to bracketthe expected assay range) would prove adequate. Thecalibration standards should be completely characterized,including the mass and isotopic composition of the speciesof SNM to be assayed and all physical or chemical variablesto which the response of the instrument is sensitive.In general, the mass of SNM contained in the standardsshould extend over the range of loadings encountered inroutine assays. This is especially true for NDA instrumentswhose responses are not linear functions of SNM content(e.g., some neutron-based NDA instruments). However, ifthe assay response (after application of appropriate correc-tions) is known to be highly linear and to have zero offset(i.e., zero response for zero SNM content), it may be moreadvantageous to avoid using standards with low loading,where calibration precision would suffer because of lowcount rates. In such a case, calibration in the upper half ofthe range of expected SNM loadings, combined withthe constraint of zero response for zero loading, can producea higher precision calibration than a least-squares fittingof measured responses to the standard over the full range ofexpected loadings, including values at low concentrations ofSNM. If such a calibration procedure is used, careful initialestablishment of the zero offset and instrument linearityfollowed by occasional verification of both assumptionsis strongly recommended. Such verification could beaccomplished by an occasional extended measurement of alow-loading standard.Unless isotopic composition is being measured, theisotopic composition of the material used in all calibrationstandards should be similar to the isotopic composition ofthe material being assayed. This is especially important for5The term "effective kilogram" is defined in paragraph 70.4(t)of 10 CFR Part 70.5.53-3 assays employing passive neutron coincidence counting orcalorimetry. When the isotopic composition changes so thatthe response per gram of SNM differs by 10 percent ormore from the value of the calibration standards, thematerial should be identified as a new material category.The NDA system should be recalibrated for that categoryusing new calibration standards made up using the newisotopic composition. When the change in response pergram is less than 10 percent, a bias correction should bedetermined and applied to the assay data.The uncertainty in the bias correction should be deter-mined and accounted for in estimating the total assayuncertainty. Appropriate error propagation procedures aredescribed in Regulatory Guide 5.18, "Limit of ErrorConcepts and Principles of Calculation in Nuclear MaterialsControl."When the response is sensitive to ingrowth or decay of adaughter product, the procedures described in the preced-ing paragraphs are appropriate and should be applied.Once fabricated, the calibration standards should behandled with extreme care to attempt to ensure that thedistribution of contents remains fixed. It should be notedthat solution standards lose their integrity over time becauseof evaporation and diffusion (Ref. 2) and radiolysis (Ref. 3).Calibration standards prepared by the mixing of differentpowders or densities tend to stratify or segregate. Thecontainers should be tumbled periodically to reblendthe constituents. Calibration standards should be used onlywhen developing the initial calibration or when recalibrat-ing the instrument following a repair or power outage.Working standards should be used to test the continuedstability of the instrument (see footnote 3).The degree of effort that should be expended in fabricat-ing the calibration standards depends on the method usedto estimate the assay uncertainty, as described in the nextsection.7. METHODS FOR ESTIMATING UNCERTAINTYInstrument errors associated with NDA should beestimated periodically by means of replicate assays asdescribed in ANSI N15.20-1975.Three methods are acceptable to estimate the uncertain-ties associated with calibrations and bias corrections forNDA. The first two procedures, graphical estimation andanalytical estimation through the calibration relationship,are detailed in ANSI N15.20-1975. The third procedure,comparative evaluation, is not described in the standard.7.1 Graphical EstimationUse of the graphical error estimation technique shouldresult in a conservative error estimate that is acceptable formiscellaneous unusual assay categories, as described inRegulatory Position 5 of this guide.7.2 Analytical Estimation Through the CalibrationRelationshipWhen the calibration standards can be shown to representadequately the unknown items, the bias associated with theNDA of an inventory of items can be estimated through thecalibration relationship as demonstrated in ANSI N15.20-1975. The calibration standards should be fabricated fromdifferent batches of material The uncertainty associatedwith the content of SNM elements and response-relatedisotopes contained in each calibration standard should bebased on an extensive characterization as described inANSI N15.20-1975. The uncertainty associated with the,contained mass of the response-related isotopes should beincluded in the calibration as described in the standard.Further, the element uncertainty should be factored intothe estimated total assay uncertainty.Using this procedure, it is necessary to periodicallyensure that the calibration standards adequately representthe unknown items. This can be accomplished by isolatingand characterizing the extraneous interference factorsthat affect the response of the instrument. Typically, thisseparation and characterization is most easily accomplishedwhen the items are either finished fuel items or uniformcontainers of feed or intermediate product material.To ensure that the calibration standards continue toadequately represent unknown items, key parameters6 thataffect the observed response (i.e., item-to-item variations)should be monitored through separate tests. Measurementsof the key parameters should be compiled and analyzed atleast twice a month to catch any large instrument drift. Formore timely measurement control, a superior approachwould be to perform such analyses on a continuing basisand repeat measurements of unknowns where standardsexceed control limits. This latter approach minimizes thebackfitting of measurement data and provides a timelybasis for measurement control.When the mean value of a parameter shifts from itspreviously established value, the impact of the shift on theresponse of the assay instrument should be measuredthrough an appropriate experiment or calculation (Ref. 4).An appropriate bias correction should be determined andapplied to all items that were assayed after the best estimateof when the parameter changed. The uncertainty in thatbias estimate should be combined with the uncertainty inthe assay values as predicted through the calibration functionto estimate the total assay uncertainty.The uncertainty due to a bias correction may significantlyincrease the standard error of the assay. In severe cases,. theeffect may increase the SEID above the level acceptable forthe total plant. In such cases, new calibration standardsshould be obtained and the assay system should be recali-brated.6See Section 5.4 of ANSI N15.20-1975. See Regulatory Posi-tion 6 of this guide for provisions to include the effects of changingisotopic compositions.5.53-4 As a further check on the continued validity of the cali-bration standards, a program to periodically introduce newcalibration standards should be implemented. The rate ofreplacement of standards with fresh material depends onthe intrinsic durability and stability of the standard inquestion. Some solution standards lose their calibratedconcentration values in a matter of days or weeks. On theother hand, standard fuel rods are much more durable andmay last indefinitely with careful handling. In any case,calibration standards should be replaced with new standardsat a rate sufficiently above their failure rate to ensurecontinued high quality in the instrument calibration.7.3 Comparative EvaluationThe procedure described in this section is not includedin ANSI N15.20-1975 but is appropriate for determining thevalidity of the calibration of NDA instruments.When two measurement methods are used for each of aseries of items and one of the methods is considerably moreaccurate than the other, corresponding measurements canbe usefully compared. The comparison can be used toestablish an estimate of bias between the measurementmethods. The comparison can also be used to estimate thetotal uncertainty associated with the less accurate measure-*ment method.To determine the uncertainty associated with the NDAof an inventory of items using this method, unknown itemsshould be randomly selected for comparative measurements.The SNM content of the items selected should span therange of contents normally encountered, subject to thequalification pointed out in Regulatory Position 6. Randomerror should be estimated through replicate analyses. Toestimate the remaining contributions to the total assayuncertainty, each item should be repeatedly assayed toreduce the random assay error to less than 10 percent ofthe estimated or previously established total uncertainty.Then, to determine the SNM content of each item selectedfor comparative evaluation, one of the following proceduresshould be employed:1. Each item should be completely dissolved, independ-ently, and the resulting solution should be analyzed byhigh-accuracy elemental and isotopic assay procedures,which in turn are calibrated relative to national standardsor nationally accepted measurement systems. It should berecognized that dissolution residues may be present in sucha procedure. These residues should also be assayed for acomplete analysis. Items composed of an aggregate ofsimilar units, e.g., fuel rods containing discrete pellets,should be opened and the contained units should be weighed,pulverized, blended, and sampled for assay through appro-priate high-accuracy elemental and isotopic assay proce-dures. The emptied container should be examined forindications of residual accumulations and cleaned, leached,or assayed nondestructively to determine the residual SNMcontent.2. For plutonium-bearing items only, each item can beassayed through calorimetric procedures (see Reference 5).Large items should be subdivided into smaller containers.Each small container should be assayed calorimetrically.Samples should be taken from at least three of the smallercontainers. The samples should be measured by micro-calorimetry and then assayed through highly accurateelemental and isotopic procedures that, in turn, are calibratedrelative to national standards or nationally accepted measure-ment systems (Ref. 6). The isotopic measurement datashould be examined for evidence of nonhomogeneousisotopic content. Isotopically nonhomogeneous materialsshould be blended and reanalyzed. On the basis of theaverage grams of plutonium per watt of the samples meas-ured by microcalorimetry, the total amount of plutoniumin each of the smaller containers should be determined. Thetotal plutonium content of the items selected for compar-ison is then estimated as the combined contents of thesmaller containers.For the first full material balance period during theinitial implementation of this guide, two items from eachcategory of assay items should be randomly selected eachweek for a check of the validity of the instrument cali-bration. Following this initial implementation period,licensees may reduce the verification measurement frequencyto two items per month per category. When fewer than 100new items of a given category are created per week, atleast two of the item-comparison verification measurementsshould be made per material balance period per categorythrough the procedures described above. In such cases, toprovide an adequate data base to update the uncertaintyestimates for NDA, licensees may pool the verification dataprovided the measurements are in statistical control, i.e.,when repeated samples from the portion of the measure-ment system under test behave as random samples from astable probability distribution. Under such conditions, datasets may be combined provided the parameters based onthe current set of data and the previous set of data are notsignificantly different on the basis of acceptable statisticaltests.As an alternative to this selection criterion, licenseesmay elect the latter frequency for a specific category whenit can be demonstrated that the contribution to the SEIDfrom that category is less than 100 grams in any mate-rial balance period.At the close of the reporting period, differences betweenassay values and verification values should be recorded andtested for outliers. Methods for detecting outliers aredescribed in ANSI/ASTM E 178-80, "Practice for Dealing withOutlying Observations."'7 See also Regulatory Guide 5.36,"Recommended Practice for Dealing with Outlying Observa-tions," for further details.7Copies may be obtained from the American Society for Testingand Materials, 1916 Race Street, Philadelphia, Pennsylvania 19103.5.53-5 A straight line with a nonzero intercept should be fittedto the nondestructive assay vs. verification measurementdata as described in ANSI N15.20-1975. The slope andintercept should be jointly tested for one and zero, respec-tively, using the "F" ratio at the 5 percent significance level(Ref. 7). If this result is significant, separate tests on theslope equal to one and the intercept equal to zero should bemade to determine the presence of either proportional orconstant bias or both. When bias is indicated, the assayresults during the preceeding operating period should becorrected. The variance associated with the bias correctionsshould be estimated by the standard error of estimate ofthe verification line. This variance must be included in theestimate of the variance of an assay result as described inANSI N15.20-1975.Whenever a bias exceeding 50 percent of its estimateduncertainty is indicated, its cause should be investigated.This investigation should include a review of the assump-tions factored into the NDA system's calibration. In partic-ular, instrument stability and the stability of parametersthat may influence the response of the assay system shouldbe investigated. The investigation should also address thecomparative measurement method, including sampling,sample handling, analytical procedures, interference com-pensation, and calibration validity. Results from the investi-gation, if they show the NDA system to have been incorrectlycalibrated, should be employed to recalibrate the instrumentfor the forthcoming material balance period. Conversely,when the source of bias can be attributed to errors in thecomparative measurements, bias corrections should not bemade to the items assayed by NDA. Results from suchinvestigations should be documented, and the documentsshould be maintained in accordance with RegulatoryPosition 8 of this guide.8. RECORDS RETENTIONAll records generated in connection with the activitiesdiscussed in this guide, including control charts, should beretained for a period of 5 years, as specified in para-graph 70.5 1(e)(4)(iii) of 10 CFR Part 70.5.53-6 REFERENCES1. T. E. Shea, "Reduction, Control, and Estimation ofNondestructive Assay Errors," Nuclear Materials Manage-ment, Vol. III, No. 3, 1974.2. G. J. Curtis, J. E. Rein, and S. S. Yamamura, "Compara-tive Study of Different Methods of Packaging LiquidReagents," Analytical Chemistry, Vol. 45, No. 6, p. 996,1973.3. J. R. Weiss and E. E. Pietri, "Calculation of HydrogenGeneration from Pu-Induced Alpha Radiolysis of Nitric,Sulfuric, and Perchloric Acids," Radiation Effects, Vol.19, p. 191, 1973.4. R. A. Forster, D. B. Smith, and H. 0. Menlove, "ErrorAnalysis of a Cf-252 Fuel-Rod-Assay System," LosAlamos Scientific Laboratory, LA-5317, 1974.S. U.S. Nuclear Regulatory Commission, "CalorimetricAssay for Plutonium," NUREG-0228, 1977.6. F. S. Stephens et al., "Methods for the Accountabilityof Plutonium Dioxide," U.S. Nuclear Regulatory Com-mission, WASH-1335, 1975.7. F. A. Graybill, An Introduction to Linear StatisticalModels, McGraw-Hill, New York, Vol. 1, p. 128, 1961.5.53-7 BIBLIOGRAPHYAlvar, K., H. Lukens, and N. Lurie, "Standard Containersfor SNM Storage, Transfer, and Measurement," U.S. Nu-clear Regulatory Commission, NUREG/CR-1847, 1980.This report details the variations of containerproperties (especially wall thicknesses) and theireffects on NDA measurements. A candidatelist of standard containers, each sufficientlyuniform to cause less than 0.2 percent variationin assay results, is given, along with comments onthe value and impact of container standardization.Brouns, R. J., F. P. Roberts, and U. L. Upson, "Considera-tions for Sampling Nuclear Materials for SNM AccountingMeasurements," U.S. Nuclear Regulatory Commission,NUREG/CR-0087, 1978.This report presents principles and guidelines forsampling nuclear materials to measure chemicaland isotopic content of the material. Develop-ment of sampling plans and procedures thatmaintain random and systematic errors ofsampling within acceptable limits for SNMaccounting purposes are emphasized.Cooper, B. E., Statistics for Experimentalists, PergamonPress, New York, 1969.This book provides a complete discussion ofstatistical procedures and describes a variety ofstatistical tests of experimental data. Examplesare provided.Reilly, T. D., and M. L. Evans, "Measurement Reliabilityfor Nuclear Material Assay," Nuclear Materials Manage-ment, Vol. VI, No. 2, 1977.This paper provides an overview of experience innuclear material assay by analytical chemistry,calorimetry, and nondestructive assay. Rangesof accuracy and precision obtained in the assayof nuclear material are given.Sher, R., and S. Untermeyer, The Detection of FissionableMaterials by Nondestructive Means, American NuclearSociety Monograph, 1980.This book contains a helpful overview of a widevariety of nondestructive assay techniques forspecial nuclear material. In addition, it containsa rather extensive discussion of error estimationand measurement control techniques, as well as apresentation on measurement statistics.5.53-8 VALUE/IMPACT STATEMENT1. PROPOSED ACTION1.3.4 Public1.1 DescriptionNo impact on the public can be foreseen.Licensees authorized to possess at any one time morethan one effective kilogram of special nuclear material(SNM) are required in paragraph 70.5 8(f) of 10 CFR Part 70to establish, maintain, and follow a program for the main-tenance of acceptable measurement quality in terms ofmeasurement bias and for the evaluation and control of thequality of the measurement system.This guide describes methods and procedures acceptableto the NRC staff for meeting the provisions of para-graph 70.58(f) of 10 CFR Part 70 for nondestructive assay(NDA) systems.The proposed action would revise the guide, which isstill basically sound.1.2 NeedThe regulatory guide endorses ANSI N15.20-1975,"Guide to Calibrating Nondestructive Assay Systems."This standard was reaffirmed without modification in 1980and the regulatory guide should be revised to indicate this.Further, revisions are needed in some sections to make theguide clearer and more consistent with current thinking.This proposed action is needed to bring RegulatoryGuide 5.53 up to date.1.3 Value/Impact1.3.1 NRCThe regulatory positions will be brought up to date.1.3.2 Other Government AgenciesNot applicable.1.3.3 IndustrySince industry is already applying the methods andprocedures discussed in the guide, updating these shouldhave no adverse impact.1.4 DecisionThe guide should be revised to reflect the affirmation ofANSI N15.20-1975 in 1980 and to make it more consistentwith current usage.2. TECHNICAL APPROACHNot applicable.3. PROCEDURAL APPROACHOf the procedural alternatives considered, revision of theexisting regulatory guide was selected as the most advanta-geous and cost effective.4. STATUTORY CONSIDERATIONS4.1 NRC AuthorityAuthority for the proposed action is derived from theAtomic Energy Act of 1954, as amended, and the EnergyReorganization Act of 1974, as amended, and implementedthrough the Commission's regulations, in particular § 70.51of 10 CFR Part 70.4.2 Need for NEPA AssessmentThe proposed action is not a major action that maysignificantly affect the quality of the human environmentand does not require an environmental impact statement.5. RELATIONSHIP TO OTHER EXISTING ORPROPOSED REGULATIONS OR POLICIESThe proposed action is one of a series of revisions ofexisting regulatory guides on nondestructive assay techniques.6. SUMMARY AND CONCLUSIONSA revised guide should be prepared to bring RegulatoryGuide 5.53 up to date.5.53-9 UNITED STATESNUCLEAR REGULATORY COMMISSIONWASHINGTON, D.C. 20555FIRST CLASS MAILPOSTAGE & FEES PAIDUSNRCWASH ) CPERMIT No _EOFFICIAL BUSINESSPENALTY FOR PRIVATE USE. $300