Regulatory Guide 5.38

(Task SG 048-4) Revision 1 Nondestructive Assay of High-Enrichment Uranium Fuel Plates by Gamma Ray Spectrometry

ML003740000
Person / Time
Issue date:	10/31/1983
From:	Office of Nuclear Regulatory Research
To:
References
RG-5.38 Rev 1
Download: ML003740000 (14)

Revision 1 U.S. NUCLEAR REGULATORY COMMISSION October 1983 REGULATORY GUIDE

OFFICE OF NUCLEAR REGULATORY RESEARCH

REGULATORY GUIDE 5.38 (Task SG 0484)

NONDESTRUCTIVE ASSAY OF HIGH-ENRICHMENT URANIUM

FUEL PLATES BY GAMMA RAY SPECTROMETRY

B. DISCUSSION

Part 70 of Title 10 of the Code of Federal Regulations The number, energy, and intensity of gamma rays requires each licensee authorized to possess more than associated with the decay of 2 3 5 U provide the basis for

350 grams of contained 235U to conduct a physicalinven nondestructive assay of high-enrichment fuel plates by tory of all special nuclear material in its possession at gamma ray spectrometry (Ref. 1). The 185.7-keV gamma intervals not to exceed 12 months. Each licensee authorized ray is the most useful 2 3 5 U gamma ray for this application;

to possess more than one effective kilogram of high it is emitted at the rate of 4.25 x 104 gamma rays per enrichment uranium is required to conduct measured second per gram of 2 3 5 U. Lower energy gamma rays physical inventories of special nuclear materials at bimonthly emitted by 235 U are less penetrating and more sensitive to intervals. Further, these licensees are required to conduct errors due to fluctuations in cladding and core thickness. In their nuclear material physical inventories in compliance general, more accurate fuel plate assays may be made by with specific requirements set forth in Part 70. Inventory measuring only the activity attributable to the 185.7-keV

procedures acceptable to the NRC staff for complying with 235 U gamma ray.

these provisions of Part 70 are detailed in Regulatory Guide 5.13, "Conduct of Nuclear Material Physical Inven Assay measurements are made by integrating the response tories." observed during the scanning of single fuel plates and comparing each response to a calibration based on the The fuel for certain nuclear reactors consists of highly response to known calibration standards.

enriched uranium fabricated into flat or bowed plates.

Typically, these plates are relatively thin so that a signifi 1. GAMMA RAY MEASUREMENT SYSTEM

cant percentage of the 235U gamma rays penetrates the fuel and cladding. When the measurement conditions are properly 1.1 Gamma Ray Detection System controlled and corrections are made for variations in the attenuation of the gamma rays, a measurement of the 2 3 5 U 1.1.1 Gamma Ray Detector gamma rays can be used as an acceptable measurement of the distribution and the total 2 3 5 U content of each fuel High-resolution gamma ray detectors, i.e., high-purity plate. In lieu of assaying the product fuel plates, fuel plate germanium, HPGe, also referred to as intrinsic germanium core compacts may be assayed through the procedures (IG), or lithium-drifted germanium [Ge(Li)] detectors, detailed in this guide provided steps are taken to ensure the provide resolution beyond that required for this assay traceability and integrity of encapsulation of each assayed application. While the performance of high-resolution fuel plate core compact. This guide describes features of a detectors is more than adequate, their low intrinsic detec gamma ray spectrometry system acceptable to the NRC tion efficiency, higher maintenance requirements, and high staff for nondestructive assay of high-enrichment uranium cost make them unattractive for the measurements dis fuel plates or fuel plate core compacts. cussed here.

Any guidance in this document related to information Most sodium iodide [NaI(TI)] scintillation detectors are collection activities has been cleared under OMB Clearance capable of sufficient energy resolution to be used for the No. 3150-0009. measurement of the 185.7-keV gamma rays. For plate USNRC REGULATORY GUIDES Comments should be sent to the Secretary of the Commission, U.S. Nuclear Regulatory Commission, Washington, D.C. 20555, Regulatory Guides are issued to describe and make available to the Attention: Docketing and Service Branch.

public methods acceptable to the NRC staff of implementing specific parts of the Commission's regulations, to delineate tech The guides are issued in the following ten broad divisions:

niques used by the staff in evaluating specific problems or postu lated accidents or to provide guidance to applicants. Regulatory 1. Power Reactors 6. Products Guides are no? substitutes for regulations, and compliance with 2. Research and Test Reactors

7. Transportation

.___/ them is not required. Methods and solutions different from those set 3. Fuels and Materials Facilities 8. Occupational Health out in the guides will be acceptable if they provide a basis for the 4. Environmental and Siting 9. Antitrust and Financial Review findings requisite to the issuance or continuance of a permit or 5. Materials and Plant Protection 10. General license by the Commission.

Copies of issued guides may be purchased at the current Government This guide was issued after consideration of comments received from Printing Office price. A subscription service for future guides in spe the public. Comments and suggestions for improvements in these cific divisions is available through the Government Printing Office.

guides are encouraged at all times, and guides will be revised, as Information on the subscription service and current GPO prices may appropriate, to accommodate comments and to reflect new informa be obtained by writing the U.S. Nuclear Regulatory Commission, tion or experience. Washington, D.C. 20555, Attention: Publications Sales Manager.

assays by scanning techniques, the detector diameter plate. To minimize this detection nonuniformity and to is determined by the fuel plate width and the scanning minimize the sensitivity to vibration, the detector-to-plate method selected (see Section B.l.2 of this guide). For distance can be made large, especially with respect to the passive counting of the total fuel plate (see Section B. 1.3 dimensions of the slit opening. As an alternative means of K-

of this guide), the detector diameter is not a critical param reducing the detection nonuniformity across the slit, the eter, and detectors suitable for plate scanning would also slit opening can be divided into channels by inserting a be adequate for the passive counting measurements. In both honeycomb baffle into the slit or by fabricating the colli cases, the thickness of the Nal crystal is selected to provide mator by drilling holes through the disk in a pattern that a high probability of detecting the 185.7-keV gamma rays ensures that each hole is surrounded by a minimum wall and a low probability of detecting higher energy radiation. thickness of 0.2 mean free path length. A 7.0-cm-thick iron A crystal thickness of 1/2 to 1 inch (13 to 25 mm) is disk with holes less than 0.5 cm in diameter drilled in a recommended. pattern having 0.2 cm of wall between adjacent holes is one example of a collimator that would perform satisfactorily.

For measurements to be reproducible, it is recommended A large number of small-diameter holes is preferable to a that the detection system be energy stabilized. Internally few large-diameter holes.

"seeded" NaI crystals that contain a radioactive source (typically 24'Am) to produce a reference energy pulse are 1.1.2.2 Collimation for Total Plate Counting. For total commercially available. The detection system is stabilized plate counting (see Section B. 1.3), the collimator opening is on the reference, and the amplifier gain is automatically circular with a diameter less than that of the NaI crystal.

corrected to ensure that the reference energy and the rest Furthermore, the collimator diameter and detector-to-plate of the spectrum remain fixed. distance are chosen so that the field of view includes the entire fuel plate. (Note that in this more relaxed counting

1.1.2 Gamma Ray Collimator geometry, the viewing area may have to be isolated from nearby sources of the 185.7-keV gamma rays in the line of The detector collimator is intended to shield the detector sight of the detector. This can be accomplished by shadow from radiation from all sources except those that are to be shielding with small pieces of lead or tungsten.)

measured. Thus the collimator shielding not only defines the front area of the detector crystal to be exposed but it 1.1.3 Multiple Detectors also shields the sides and, if possible, the rear of the detec tor. The front opening of the collimator is designed to Several detectors may be used to shorten the measurement define the field of view appropriate for the measurement time. The detectors can be positioned to measure different technique to be employed. Once a measurement system is segments of a single fuel plate or several separate fuel plates 3/4 calibrated with a particular collimator configuration, that simultaneously. In some cases it may be useful to sum the configuration must be maintained for all subsequent assays. response from two detectors positioned on opposite sides Any change in the collimation system will necessitate of a plate to increase counting efficiency. In such cases, it is recalibration of the measurement system. essential that the relative response of such detectors be known and checked at frequent intervals for continued

1.1.2.1 Collimation for Scanning Techniques. To ensure stability.

that the only gamma ray activity detected originates from a well-defined segment of the fuel plate, the detector is 1.2 Scanning Techniques shielded from extraneous background radiations and collimated to define the plate area "seen" by the detector It is critical that the scanning apparatus for moving the crystal. The collimator consists of a disk of appropriate plates relative to the detector provide a uniform and shielding material. A slit is machined through the center of reproducible scan. The importance of a well-constructed the disk to allow only those gamma rays emitted within mechanically stable conveyor cannot be overemphasized.

the slit opening to strike the detector. The disk thickness is Either the detector can be moved and the plate held station a minimum of six mean free path lengths to effectively stop ary, or the plate can be moved past a fixed detector. If the all 185.7-keV gamma rays emitted from outside the field of detector collimator field of view extends beyond the edges view. For more compact counting geometries, higher of the fuel plate, care must be exercised to maintain the density shielding materials (such as tungsten or lead) can be detector-to-plate spacing within close tolerances to minimize used. The linear dimensions of other shielding materials errors caused by the resulting dependence of count rate on scale down according to the decrease in mean free path this spacing. This is especially important in the case of close length. 1 spacing, which is sometimes desirable to maximize the count rate. However, a superior collimator configuration The probability of detection for gamma rays emitted at from this point of view would be one in which the field of the center of the collimator slit is greater than that for view is filled with active material over a range of detector-to gamma rays emitted near the ends of the slit. This effect plate distances. In this case, the measured material acts as becomes increasingly important at small detector-to an area source for which the counting rate is nearly independ plate spacing, especially when scanning near the edge of a ent of the detector-to-plate spacing. Therefore, in the

1 For the 185.7-keV gamma ray from 235U, a thickness equi "sweeping spot scan" technique discussed in Section B. 1.2.2, valent to six mean free path lengths in lead is approximately the spacing is not as critical a measurement parameter.

0.45 cm; in tungsten it is approximately 0.33 cm; and in iron it is approximately 4.9 cm. Various commercial conveying systems have been used and

5.38-2

found to be adequate. Such systems may significantly is then determined by averaging the results of sample spot reduce the cost of designing and building new scanning measurements of the 235 U content per unit area at a mechanisms. High-precision tool equipment such as milling number of sites along the plate and multiplying this average

• , machines, lathes, and x-y scanning tables can be investigated. value by the measured area of the fuel core. The radiograph Numerically controlled units offer additional advantages of each plate is examined to ensure that the core filler is when they can be incorporated into a scanning system. This uniform since nonuniformities would invalidate this type of is particularly true when an automated scanning system is assay.

being developed.

The collimator shape and dimensions can be selected to Fuel plate core compacts may be sufficiently small to provide compatible information on the uniformity of the permit total assay in a fixed-geometry counting system fuel plate.

without scanning (see Section B.1.3). The scanning tech niques for fuel plates discussed in the following subsections 1.3 Passive Total Counting Techniques can also be used for core compacts when total fixed com pact counting is not possible. A single passive gamma count of a fuel plate can be used to obtain the information of primary concern, namely the total 23SU content of the plate. The detector response in a

1.2.1 Linear Total Scan "wide-angle" counting geometry can be converted to grams of 2 3 5 U in the plate if the response with standard The detector collimation consists of a rectangular fuel plates is known for the same counting geometry and if opening that extends across the width of the fuel plates appropriate attenuation corrections are made with suitable beyond the edges of the uranium core contained within the transmission sources.

plate cladding. Scanning the total plate is accomplished by starting the count sequence on the end of a plate and The detector collimator and detector-to-plate distance continuing to count until the entire length of plate has been defines a field of view that (a) includes the entire fuel plate scanned. and (b) is isolated from other sources of radiation in the line of sight of the detector. Provide a measurement platform To ensure that gamma rays emitted anywhere across the to facilitate the reproducible placement of the fuel plates, face of the fuel plate have an equal probability of being transmission sources, detector, and collimator shielding in a detected, it is necessary that the diameter of the detector standard measurement configuration.

crystal exceed the plate width or that the detector be

.*- positioned away from the plate. Core compacts are also to be assayed in this way provided representative standards are used to calibrate the measure Use of the spot or circular collimator scan technique ment for the geometry pertaining to these items.

eliminates or reduces to insignificance most of these edge effects. Additional details on passive total sample counting and the associated attenuation corrections for assay of special

1.2.2 Sweeping Spot Scan nuclear materials are given in References 3 and 4.

If the collimator channel width is smaller than the fuel 1.4 Computer Control plate width, the viewing area (spot) can be swept across the plate as the detector scans along the length of the plate Computer control of the plate scanning techniques can (Ref. 2). This scanning technique can be readily adapted to greatly reduce the associated manpower requirements and scanning bowed plates through the use of a cam that is improve measurement reproducibility. The computer can designed to maintain the detectoi-to-plate distance constant be used to control data acquisition by accumulating counts over the entire fuel plate. The collimator channel dimensions according to a predetermined scheme. Also, the computer can be selected to provide compatible information on the can be used for data analysis, including background and uniformity of the fuel plate, which is frequently obtained attenuation corrections and intermachine normaliza by comparing fixed (static) spot counts at a variety of tion, calibration, error analysis, and diagnostic test measure locations to reference counts. ments and analyses. Report preparation and data recording for subsequent analysis are also readily accomplished

1.2.3 Sampled Increment Assay through an appropriately designed computer-controlled system. Use of a computer can be of great value in many of When used in conjunction with radiographic dimensional these functions for the total passive gamma counting measurements performed on all fuel plates, the 235 U technique as well.

content of a fuel plate can be measured by scanning the ends of each fuel plate and sampling the balance of the 2. INTERPRETATION OF MEASUREMENT DATA

plate. It is necessary to measure the dimensions of the fuel core loading radiographically through gamma ray scanning The raw measurement data from either a scanning or a

___ along the length of the plate or by spot-scanning the fuel total passive counting technique can be distorted by several plate ends and measuring the distance between end spots effects for which corrections should be made for accurate where the fuel loading stops. The 235 U content of the plate assays. The three factors discussed below are the most

5.38-3

important potential sources of measurement error that can ing the clad thickness over the range of thicknesses to be give rise to significant misinterpretation of the data. encountered in normal product variability.

2.1 Enrichment Variations 2.2.3 Core FillerAttenuation Licensees authorized to possess highly enriched uranium Radiation intensity measurements may be made of are required to account for each element and isotope as plates fabricated with different ratios of uranium to filler to prescribed in § 70.51. Under the conditions detailed in this show the effects of this type of attenuation. If significant guide, the 2 3 5 U content of individual plates is measured. effects are noted, plates can be categorized by core composi To determine the total uranium content of each plate, the tion characteristics and the assay system can be independ

235U enrichment of the core filler must be known from ently calibrated for each category of fuel plates.

separate measurements.

2.2.4 Attenuation Corrections Enrichment variations may also alter the radiation background in the gamma ray energy region of interest and When the thickness of the core and cladding and the cause fluctuations in the 2 3 5 U assay. The 2 3 8 U decays by composition of the core material are known, an attenuation alpha particle emission to 2 3 4 Th. The 2 3 4 Th then decays correction can be calculated and applied to improve the by beta particle emission with a half-life of 24.1 days to accuracy of the assay. These corrections must also be

2 34 Pa which, in turn, decays by beta particle emission to applied to the assays of the standards in the calibration

234U. Approximately 1 percent of the 234Pa decays are procedure. Ultrasonic gauging may provide such a measure followed by high-energy (e.g., 1001 keV, 766 keV) gamma if the metallographic zones within the plate are sufficiently rays. These gamma rays frequently lose energy through defined to provide a detectable interface.

Compton scattering and may appear in the 185-keV spec tral region. It is important to note that activity from The alternative attenuation correction can be based on a

234Pa may be altered by disturbing the equilibrium between

23 5 micrometer measurement of the total thickness of each U and 2 3 4 Th, as frequently occurs in uranium chemical plate. The clad thickness of a plate is estimated by subtract conversion

238 processes. The interference due to variations in ing the mean core thickness of the product plates, which U daughter activity becomes less important as the is determined by periodically sampling product plates and enrichment of 2 3 5 U increases. At enrichment levels above cutting a cross section to permit visual measurement of clad

90 percent, this problem can essentially be ignored. and core thickness.

2.2 Radiation Attenuation As long as the gamma ray attenuation corrections are computed on the basis of declared component thicknesses Attenuation of gamma radiation mayrange from complete and composition (or on the basis of occasional measure absorption of the radiation by the intervening material to ments of these quantities), unnoticed plate-to-plate fluctua partial energy loss of the emitted radiation through scatter tions in these parameters will undermine the accuracy of ing processes. Both effects reduce the number of full-energy the assays. A far more reliable approach to the application gamma ray events that are detected. Gamma rays from of attenuation corrections is to measure the gamma ray

235 U are attenuated in the uranium, in the cladding, and in transmission property of each plate (standard as well as the inert material that may be added with the uranium to unknown) as it is being assayed. This approach increases the form the core of the fuel plate. Through well-controlled complexity of the assay procedures, but poses the further product tolerance limits, each of these potential sources of advantages of ýl)rendering the calibration dependent signal variability can be controlled to permit accurate only upon the 35U loading of the standard plates and accountability assays. independent of other plate properties and (2) making the sample plate assays insensitive to possible fluctuations in

2.2.1 Self-Attenuation cladding thicknesses and core composition and thickness.

General discussions of gamma ray attenuation corrections The photon attenuation coefficient of uranium for accompanying passive assays are given in References 3, 4, gamma ray energies corresponding to 2 3 5 U emissions is and 6. Specific details of a correction procedure for Materi quite large (Ref. 5). Small changes in uranium density als Testing Reactor (MTR) fuel plates are given in the resulting from increased fuel loading or from variations in appendix to this guide.

the manufacturing process can therefore significantly change the number of gamma rays that escape from the fuel 2.3 Interfering Radiations plate.

As noted in Section B.2.1 of this guide, an internal

2.2.2 CladdingAttenuation background variation may arise from changes in the amount of 2 3 8 U present in a fuel plate or from changes in the ratio Small variations in cladding thickness may cause signif of 234Th to 238U resulting from fuel manufacturing icant variations in attenuation. These variations in attenua processes. Fluctuations in the internal background cause tion can be measured by a simple gamma ray absorption test the response of the unknown items to be different from the using thin sheets of cladding material as absorbers and vary- calibration standards, thereby creating a fluctuating measure-

5.38-4

ment bias. In addition, some discrete gamma ray inter more accurate measurements of the content of typical ferences may be present at energies near 185.7 keV. For fuel plates (see Regulatory Postion 4 of this guide). Guid further information on these possible interferences, see ance on methods to relate this assay to the national measure Reference 7. Both the background and discrete gamma ray ment system and to reconcile verification measurements is interferences are generally of minor importance, but they addressed in Regulatory Guide 5.58, "Considerations for can be corrected for by measurement of additional regions Establishing Traceability of Special Nuclear Material of the gamma ray spectrum. Pertinent nuclear data for such Accounting Measurements."

measurements are available in Reference 1.

C. REGULATORY POSITION

Other interfering radiations may come from external sources, from fuel plates awaiting assay, or from nearby The content and distribution of 23 5U in high-enrichment radiation sources used for other measurements. This is not uranium plates can be measured through the gamma ray expected to be a major problem and can be controlled assay methods discussed in this guide. Combining this through (1) removing radiation sources, (2) shielding the measurement with the results of an independent measure detectors, and (3) monitoring the background at frequent ment of the 2 3 SU enrichment enables the total uranium intervals. content of the fuel plates to be determined. The factors presented below should be taken into consideration for this

3. CALIBRATION AND VERIFICATION assay method to be acceptable to the NRC staff.

3.1 Initial Operations

1. MEASUREMENT SYSTEM

Calibration and the verification of assay predictions is an 1.1 Gamma Ray Measurement System ongoing effort where performance is periodically monitored and the calibration relationship is modified to improve the

1.1.1 Gamma Ray Detector accuracy of assay predictions. During initial operations, two means of basing preliminary calibrations are appropriate. Thallium-activated sodium iodide [NaI(TI)] scintillation detectors are recommended for this assay applicatio

n. When

3.1.1 Foil Calibration Technique more than one detector is to be incorporated into the measurement system, the performance characteristics of Methods for calibrating scanning systems for high the detectors should be matched as closely as possible, and enrichment uranium fuel plates through the assay of the relative response of the detectors should be checked prepared clad uranium foils are described in Reference 2.

periodically to verify continued stability of the system. The These methods may be used in place of or in addition to the diameter of the crystal should be larger than the projected technique described in the following subsection. view onto the crystal face through the collimator channel.

A crystal thickness of 1/2 to 1 in. (13 to 25 mm) is recom

3.1.2 FabricatedCalibrationPlates mended. The crystal should contain an internal seed that is doped with a suitable alpha emitter (typically 2 4 1 Am) to Calibration standard fuel plates can be fabricated using produce a reference energy peak for spectral stabilization.

special precautions to ensure that the amounts of uranium,

2 35 U, inert matrix, and cladding are accurately measured The seed should produce approximately 1,000 counts per second at the reference energy.

and that these parameters fall within manufacturing toler ances for product plates.

1.1.2 Collimatorand Detector Shielding

3.2 Routine Operations The collimator should be fabricated of appropriate The performance of the assay system is periodically gamma ray shielding material such as iron, lead, or tungsten.

monitored to ensure that the response of the assay system The shielding should completely surround the detector and has not shifted since its last calibration. Control limits for photomultiplier assembly and should be sufficiently thick acceptable performance can be established for the response to completely shield the detector from extraneous radiation.

to an appropriate working standard. The control chart of the responses to the working standard can be checked for indications of short-term instrument drift or malfunction.

1.1.3 Electronic Apparatus The control chart can also be analyzed to detect long term shifts -within the measurement-to-measurement All electronic systems should be powered by filtered, control limits that may be corrected by recalibrating the highly regulated power supplies. The ambient temperature system. In general, however, it is important that observed and humidity in the vicinity of the scanning system should instrument drifts and performance changes be investigated be controlled so that permitted fluctuations do not signifi and remedied rather than compensated for by recalibration.

cantly affect the assay measurements. All electronic circuitry in signal-processing components should feature temperature To ensure that the calibration remains valid during compensation. Residual sensitivity to fluctuations in the

"-' normal operations and that accuracy estimates are rigorously ambient environment should be tested and monitored justified, assay predictions are periodically compared with periodically.

5.38-5

The capability for multichannel gamma ray pulse height monitoring the variations in plate cladding thicknesses and analysis with cathode ray tube spectral display should be core composition and thickness. For further detail on such provided. Signal-processing electronics capable of stabilizing corrections, see References 4 and 6 as well as the appendix on the reference energy peak produced by the alpha-emitter to this guide.

doped seed should be provided to stabilize the energy spectrum. 2.3 Radiation Interferences

1.2 Measurement System A graphic record of an acceptable (reference) gamma ray spectrum display (i.e., free of interferences and exhibiting Plate scanning should be accomplished by one of the nominal background) should be prepared. When radioactive three techniques discussed in Section B.1.2 of this guide. interference may be encountered, the assay spectrum With these techniques, a mechanically sound, highly repro should be compared at appropriate intervals to the reference ducible, automated scanning system should be employed. spectrum for indications of interference. Background When more than one scanning system is employed, the radiation should be measured periodically during each assay responses of each system should be normalized so operating shift.

that each instrument provides consistent results. Verification data to estimate the bias for each assay system should be 3. MEASUREMENT CALIBRATION AND CONTROL

obtained with the same standard plate.

During initial operations, the assay system should be If a passive total counting technique is used, a stable, calibrated either by the foil calibration method or with carefully constructed measurement platform should be specially prepared sample fuel plates as described in Sec employed to ensure the achievement of a reproducible tion B.3.1 of this guide. Instrument response to appropriate measurement geometry. working standards should also be checked periodically to verify the continued stability of the assay system calibration.

1.3 Computer Control

4. SOURCES OF VARIATION AND BIAS

A dedicated minicomputer to control data acquisition, calibration, diagnostic testing, and report preparation 4.1 Random Assay Standard Deviation Estimation should be employed for fuel plate assay operations.

A replicate assay program should be established to

2. MEASUREMENT INTERPRETATION generate data for the evaluation of the random assay standard deviation during each material balance period.

I--

2.1 Enrichment Variations During each bimonthly interval, a minimum of fifteen plates should be selected for replicate assay. The second Procedures should be developed to ensure that the assay of each plate selected for replicate assay should be enrichment of the plates being scanned is known through made at least four hours after the first assay. Replicate separate measurements. Fuel plates generally satisfy the assay data should be collected and analyzed at the end of gamma ray penetrability criteria for quantitative 2 3 SU the material balance period. The single-measurement assay; they do not satisfy the criteria for nondestructive2 standard deviation of the replicate assay differences should enrichment measurement through gamma ray spectrometry. be computed as described in Reference 8. Replicate measure Facilities processing more than one uranium enrichment ments should be made under the same conditions as routine should maintain strict isotopic control and characterize the measurements, performed throughout the production run, enrichment through appropriate measurement methods. and checked for consistency. If the probability distributions for the data are not different, pooling of results from

2.2 Attenuation Corrections previous inventory periods can improve the random assay standard deviation estimates.

If computed attenuation corrections are used, attenuation variations arising from plate-to-plate changes in core thick 4.2 Calibration Standard Deviation Estimation ness, core composition, and clad thickness should be determined over the range of product tolerance specifica The calibration standard deviation associated with the tions. When such variations cause the assay standard devia assay of all fuel plates assayed during each calibration tion to exceed the standard deviation realized without the period throughout the material balance period can be variations by 33 percent or more, procedures should be determined through one of the procedures presented below.

implemented to measure and apply a correction to the These methods are discussed in detail in ANSI 15.20-1975, assay of each plate. It should be noted that routine measure "Guide to Calibration of NDA Systems," 3 and in Regulatory ment of attenuation corrections for each plate is recom Guide 5.53, "Qualification, Calibration, and Error Estima mended since such a procedure will remove the necessity of tion Methods for Nondestructive Assay" (a proposed revision to this guide has been issued for comment as Task

2 Criteria for gamma ray uranium enrichment measurements are SG 049-4).

given in Regulatory Guide 5.21, "Nondestructive Uranium-235 3 Available from the American National Standards Institute, Enrichment Assay by Gamma Ray Spectrometry." A proposed revision to this guide has been issued for comment as Task SG 044-4. 1430 Broadway, New York, New York 10018.

5.38-6

To estimate the standard deviation arising from the random assay standard deviation associated with the less calibration procedure, the calibration should be based on a accurate measurement method. To determine precisely the least-squares fitting of the calibration data to an appropriate bias in the nondestructive assay measurement, the fuel model, then part of the calibration standard deviation plates selected for comparative measurements should be

• -*- can be derived using the residual mean square. The standard randomly selected but should span tile range of 2 3 5 U

deviation for the calibration standards includes the standard contents encountered in normal production. The fuel deviation of the reference values for the calibration stand plates could have been selected from those rejected from ards. See ANSI 15.20-1975. the process stream for failing to meet quality assurance requirements. Each plate should be repeatedly assayed to To ensure the validity of the measurements, the stable reduce the random assay relative standard deviation (coeffi performance of the instrument should be monitored and cient

2 35 of variation) to less than 10 percent. To determine its normalized through the response to appropriate working U and total uranium content, the plate should be standards that are assayed at frequent intervals. The fre completely dissolved and the resulting solution should be quency for assaying working standards should be deter analyzed by high-accuracy assay procedures such as chemi mined through testing but should -not be lower than one cal and mass-spectrometric analyses.

test during each two-hour assay interval for spot response stability and one full scan test during each operating shift. For one material balance period during the initial For total passive counting techniques, assay of working implementation of this guide, a product fuel plate should standards should take place during each four-hour assay be randomly selected twice each week for an accuracy interval during each operating shift. Indications of shifting verification measurement. Following this initial implementa instrument performance should be investigated and the cause tion period, facilities manufacturing 100 or more fuel plates should be remedied. The instrument should then be recali per week may reduce the verification frequency to one brated to ensure the validity of subsequent measurements. plate per week and pool the verification data (provided the two distributions can be tested to show no differences) for In order to ensure that the calibration standards continue two consecutive material balance periods. Low-throughput to adequately represent the unknown fuel plates, key facilities manufacturing less than 100 plates per week production parameters that affect the observed response should verify at least 4 plates per material balance period should be monitored through separate tests. (If transmis through the procedures described above. At the close sion corrections are being measured for each plate assayed, of each material balance period, data should generally be the monitoring of plate parameters is less critical for assay pooled (if allowable) to include the 15 most current data accuracy.) Data should be compiled and analyzed at the points. However, if the data are demonstrably stable over close of each material balance period. When a production longer periods, using additional data points from previous S parameter shifts from previously established values, the compatible results is one method of reducing the random impact of the shift on the response of the assay instrument assay standard deviation estimate.

should be determined through an appropriate experiment or calculation (Ref. 9). A bias correction should be deter Two methods are presented for estimating the bias.

mined and applied to all items assayed from the point of When the 23SU content of the plates assayed using a common the parameter change. The variance of the bias estimate calibration relationship varies over a range of +/-5 percent or should be combined with the variance due to the calibra more of all plate loadings, the bias should be estimated by tion procedure. When the bias exceeds 3 percent of the Method No. 1. When plate loadings are tightly clustered plate contents in a single material balance period, when a about a nominal value, the bias should be estimated by trend of .1.5 percent or more is observed in three consecutive Method No. 2.

material balance periods, or when the standard deviation in the estimated bias is sufficient to increase the standard error Method No. 1. At the close of the reporting period, the (i.e., twice the standard deviation) of the assay above assay value for each plate is plotted against the verified

0.5 percent, new calibration standards should be obtained quantity. The verification data plot is examined for indica and the scanning measurement system should be recalibrated. tions of nonlinearity or obvious outlier data. Anomalous indications should be investigated and remedied. Further As a further check on the continued validity of the details on handling outlier data are contained in Regulatory calibration standards, a program to introduce new calibration Guide 5.36, "Recommended Practice for Dealing with standards periodically should be implemented. A minimum Outlying Observations." The comparison data should of one new calibration standard fuel plate should be intro be analyzed as described in Regulatory Position 7.3 of duced during each six-month period. Regulatory Guide 5.53, "Qualification, Calibration, and Error Estimation Methods for Nondestructive Assay." A

4.3 Bias Estimation proposed revision to this guide has been issued for comment as Task SG 049-4.

When two sets of measurements are made on each of a series of items and the accuracy of one of the methods used Method No. 2. When all plates contain essentially the is considerably better than the other, the corresponding same 235U content, the difference in the mean content estimates can be compared to establish an estimate of bias values should be tested against zero as an indication of bias,

-- between the measurement methods and to estimate the and the standard deviation associated with an inventory of

5.38-7

plates should be estimated as the standard deviation of the compact. The fuel plate should carry an identification mean difference. For individual plates, the standard devia corresponding to the compact identification.

tion should be estimated as the standard deviation of a single measurement.

2. Each fuel plate should be radiographically examined

5. CORE COMPACT ASSAY to ensure that the entire compact has been encapsulated.

Final product assay in high-enrichment fuel plate manu facturing can also be accomplished through assaying each 3. Each fuel plate should be checked with a gamma ray core compact following the procedures detailed in this probe to ensure qualitatively that the plate core is uranium guide and the following supplemental criteria: of the normal product enrichment.

1. Each core compact should carry a unique identifica 4. Calibration and error evaluation should follow the tion. Accountability records should be created for each procedures for fuel plate assay.

5.38-8

APPENDIX

SAMPLE ATTENUATION CORRECTION BY

TRANSMISSION MEASUREMENT FOR MATERIALS

TESTING REACTOR FUEL PLATES

1. BACKGROUND where lic is the mass absorption coefficient of the cladding at 185.7 keV (for aluminum, pc = 0.126 cm 2/g) and pc is Gamma ray assay data are subject to distortions due to the cladding density (for aluminum, pc = 2.7 g/cm 3 ). If the the attenuation of the gamma ray flux by the intervening cladding thickness varies by as much as 10 percent, the sample material and sample container. The data must be corresponding variation in Tc will be only 0.2 percent. Thus corrected for this effect or the amount of nuclear material an assumption of invariant cladding attenuation for a being assayed will be underestimated. The measured inten particular type of fuel plate will contribute very little to the sity I for the 185.7-keV gamma radiation from a Materials assay variance when the constant cladding attenuation Testing Reactor (MTR) fuel plate is related to the 23SU correction is applied. One then determines Tc for the fuel content MU of the fuel plate by: plates from careful measurement of the cladding thickness and application of Equation 4.

MU = kI/C (1)

Under the above assumption, one can then determine where k is a calibration constant that includes effects such the transmission of the core material TU from the measured as detector efficiency, counting geometry, and nuclear total plate transmission T, knowledge of Tc, and Equation 3.

properties of uranium. The factor C is the correction factor The attenuation correction factor in Equation 1 is then that adjusts the raw data for the attenuation of the given by (see References 3 and 4):

18,5.7-keV gamma ray by the plate cladding and core materia

l. In TU

(5)

Determination of this attenuation correction factor can Tc (1- TU)

be accomplished using an external gamma ray source.

(Ideally, this should be a 23 5U source. For details on how

2. IMPLEMENTATION

to use a transmission source with a gamma ray energy dif ferent from that measured in the assay, see Reference 4.) 2.1 The Scanning Techniques The radiation from this source is detected after it passes through the fuel plate, and that transmitted gamma inten A small transmission source should be placed behind the sity I I is compared with the source intensity with no plate fuel plate as shown in Figure 1. The transmission correction present 10 to obtain the gamma ray transmission T through must be measured and applied at each scan point so that the plate materials: nonuniformities in core composition within a plate can be corrected for. The transmission of the plate T(i) is measured T = 11/1 0 (2) at each scan point i by determining (1) the plate count rate with the transmission source shielded I(i) (Figure 1A),

This total plate transmission can be subdivided as (2) the total counting rate of the plate and unshielded follows: transmission source IT(i) (Figure 1B), and (3) the transmis sion source count rate with no intervening plate 10 (i)

T=TT 2 (3) (Figure 1C).

U c T(i) = [IT(i) - I(i)l /1o(i) (6)

where Tc is the gamma ray transmission through one thickness of the plate cladding, and TU is the transmission If the transmission source is a small localized 2 35 U

through the core material. (The same cladding thickness on source, a plate assay with the attenuation correction will both sides of the plate is assumed.) require two scans: one to get the I (i) values and one to get the I(i) values. (The quantity IoTi) will be constant at The gamma ray transmission through a plate is dominated all scan points and can be measured at a separate time.) If by the effect of the core material (i.e., TU < Tc), so it is the transmission source is another fuel plate that remains convenient to treat the cladding transmission Tc as a stationary with respect to the plate being assayed, the I0 (i)

constant. Furthermore, variations in the core composition must be measured by scanning the transmission source will cause more drastic fluctuations in the gamma ray fuel plate. That is, an unattenuated transmission source attenuation than the small variations in the cladding thick plate intensity 10 (i) must be measured at the same scan ness tc. For example, a 20-mil (0.051 cm) aluminum points i and associated with the corresponding IT(i) and 1(i)

cladding thickness attenuates the 185.7-keV gamma intens from the measurements with the unknown plate. The ity according to: count arrays IT(i), I(i), and 10 (i) must be stored in the computer memory as they are measured. The counts I(i) are then corrected by the factor in Equation 5 for each total Tc= e c ctc = 0.983 (4) plate transmission T(i).

5.38-9

2.2 Total Passive Count Technique 3. MEASUREMENTS WITH HIGH-RESOLUTION

SYSTEMS

In this case, an average attenuation correction is deter The transmission of the 235U gamma ray can be inferred mined by measuring T for the entire plate using a 235U from measured transmission just above and just below source behind the plate. An extended transmission source is 185.7 keV in energy. In one application using high-resolution recommended (ideally another fuel plate) in order to gamma ray spectrometers (Reference 10), a 1 6 9 Yb trans observe an average transmission over as much of the mission source is used. Two of the gamma rays emitted by unknown plate as possible. The transmission source must this isotope are at 177.2 and 198.0 keV, conveniently not extend beyond or radiate around the edges of the bracketing the 185.7-keV energy regio

n. Measurement of T

fuel plate being assayed. In this case, the assay involves at these two energies and interpolating to 185.7 keV results three counts: (1) fuel plate plus shielded transmission in a determination of the attenuation correction factor C at source I, (2) plate plus unshielded source IT, and (3) the 235U gamma ray energy. A high-resolution detector unshielded source with no plate Io. The average plate system must be used in order to resolve the 177.2-, 185.7-,

transmission T is then defined as: and 198.0-keV gamma ray peaks. In this way, the assay and transmission correction data are acquired simultaneously T = (IT- I)/Io (7) and multiple scans or multiple counts are not necessary. As a practical matter, 169yb has the short half-life of 32 days, so this source must be replaced frequently (or reirradiated A single attenuation correction from Equation 5 is then in a reactor) in order to provide sufficient counts for a applied to the passive count of the plate I. precise measurement of the attenuation corrections.

5.38-10

SOURCE SORE TRANSMISSION SOURCE

COLLIMATED DETECTOR "'1?> oUNKNOWN FUEL PLATE

U (A)

• *

11.ýc ý,ýc

(8) (C)

Figure 1 A schematic of the measurement arrangements for MTR fuel plate gamma ray assay with measured attenuation correction.

The close geometry of the scanning technique is used as an example. (A) Configuration for measuring 235 radiation coming only from the unknown fuel plate (I in the text). (B) Configuration for determining the sum of the fuel plate gamma intensity and the source intensity passing through the fuel plate (IT in the text). (C) Configuration for measuring

___ the incident transmission source gamma intensity (10 in the text).

5.38-11

REFERENCES

1. J. E. Cline, R. J. Gehrke, and L. D. McIssac, "Gamma ment," Proceedingsof the ERDA X- and Gamma-Ray Rays Emitted by the Fissionable Nuclides and Asso Symposium, Ann Arbor, Michigan, Conf. 760639, ciated Isotopes," ANCR-1029, 1972. p. 219, May 1976.

2. N. S. Beyer, "Assay of 235U in Nuclear Reactor Fuel 7. T. D. Reilly, "Gamma-Ray Measurements for Elements by Gamma-Ray Scintillation Spectrometry," Uranium Enrichment Standards," Proceedings of the Proceedings of the 4th International Conference on American Nuclear Society Topical Meeting on Nondestructive Testing, London, 1963. "Measurement Technology for Safeguards and Material Control," Kiawah Island, South Carolina,

3. R. Sher and S. Untermeyer, The Detection of Fission November 1979; National Bureau of Standards able Materialby Nondestructive Means, American Nuclear Special Publication No. 582, p. 103, June 1980.

Society Monograph, 1980.

8. J. L. Jaech, "Statistical Methods in Nuclear Materials

4. R. H. Augustson and T. D. Reilly, "Fundamentals of Control," Atomic Energy Commission, Report Passive Nondestructive Assay of Fissionable Material," No. TID-26298, 1973.

Los Alamos Scientific Laboratory, LA-5651-M, 1974.

9. R. A. Forster, D. B. Smith, and H. 0. Menlove,

5. E. Storm and H. Israel, "Photon Cross Sections from "Error Analysis of a Cf-252 Fuel Rod Assay System,"

0.001 to 100 MeV for Elements 1 Through 100," Los Alamos Scientific Laboratory, LA-5317, 1974.

Los Alamos Scientific Laboratory, LA-3753, 1967.

10. E. R. Martin, D. F. Jones, and J. L. Parker, "Gamma

6. J. L. Parker and T. D. Reilly, "Bulk Sample Self Ray Measurements with the Segmented Gamma Scan,"

Attenuation Correction by Transmission Measure- Los Alamos Scientific Laboratory, LA-7059-M, 1977.

5.38-12

VALUE/IMPACT STATEMENT

1. PROPOSED ACTION

2. TECHNICAL APPROACH

1.1 Description and Need Not applicable.

Regulatory Guide 5.38 was published in September

3. PROCEDURAL APPROACH

1974. The proposed action, a revision to this guide, is needed to bringthe guide upto date with respect to advances Of the procedural alternatives considered, revision of in measurement methods and changes in terminology. the existing regulatory guide was selected as the most advantageous and cost effective.

1.2 Value Impact of Proposed Action 4. STATUTORY CONSIDERATIONS

1.2.1 NRC Operations 4.1 NRC Authority The regulatory positions will be brought up to date. Authority for the proposed action is derived from the Atomic Energy Act of 1954, as amended, and the Energy

1.2.2 Other Government Agencies Reorganization Act of 1974, as amended, and implemented through the Commission's regulations.

Not applicable.

4.2 Need for NEPA Assessment

1.2.3 Industry The proposed action is not a major action that may Since industry is already applying the methods and significantly affect the quality of the human environment procedures discussed in the guide, updating the guide and does not require an environmental impact statement.

should have no adverse impact.

5. RELATIONSHIP TO OTHER EXISTING OR

1.2.4 Public PROPOSED REGULATIONS OR POLICIES

No adverse impact on the public can be foreseen. The proposed action is one of a series of revisions of existing regulatory guides on nondestructive assay techniques.

1.3 Decision on Proposed Action

6. SUMMARY AND CONCLUSIONS

The regulatory guide should be revised to reflect the improvement in measurement techniques and to bring the Regulatory Guide 5.38 should be revised to bring it up language of the guide into conformity with current usage. to date.

5.38-13

FIRST CLASS MAIL

UNITED STATES POSTAGE & FEES PAID

NUCLEAR REGULATORY COMMISSION USNRC

WASH 3 C

WASHINGTON, D.C. 20555 PERMIT No G-Lk OFFICIAL BUSINESS

PENALTY FOR PRIVATE USE. $300