UNITED STATES

NUCLEAR REGULATORY COMMISSIONWASHINGTON, D. C. 20555

February 1, 1983.

Regulatory Guide 1.150Revision 1

REGULATORY GUIDE DISTRIBUTION LIST (DIVISION 1)

SUBJECT: REVISION 1 TO REGULATORY GUIDE.l.150

The original version of Regulatory Guide 1.150, "Ultrasonic Testing of Reactor.Vessel Welds During Preservice and Inservice Examinations," dated June 1981,was reviewed by the Ad Hoc Committee of the Electric Utility Industry. TheCommittee's review resulted in their recommending an alternative method, whichwas presented in "Final Report - Recommended Changes to Regulatory Guide 1.150,"August 1982.

The NRC staff has evaluated the recommended alternative method in the reportand finds it an acceptable alternative. Therefore, this alternative methodhas been added to the regulatory positions in Regulatory Guide 1.150, andthis Revision 1 is being issued as an active guide without prior issuanceas a draft guide for comment.

Of course, comments and suggestions for improvements in guides are encouragedat all times. Guides will be revised, as appropriate, to accommodate commentsand to reflect new information or experience.

Robert B.Office of Nuclear Regulatory Research

U.S. NUCLEAR REGULATORY COMMISSIONRevision 1

February 1983

REGULATORY GUIDEOFFICE OF NUCLEAR REGULATORY RESEARCH

REGULATORY GUIDE 1.150

ULTRASONIC TESTING OF REACTOR VESSEL WELDS DURINGPRESERVICE AND INSERVICE EXAMINATIONS

A. INTRODUCTION

Criterion 1, "Quality Standards and Records," of Appen-dix A, "General Design Criteria for Nuclear Power Plants,"to 10 CFR Part 50, "Domestic Licensing of Production andUtilization Facilities," requires, in part, that componentsimportant to safety be tested to quality standards commen-surate with the importance of the safety functions to beperformed. Where generally recognized codes and standardsare used, these codes and standards must be evaluated todetermine their adequacy and sufficiency and must be sup-plemented or modified as necessary to ensure a quality pro-duct in keeping with the required safety function. Criterion Ifurther requires that a quality assurance program be imple-mented in order to provide adequate assurance that thesecomponents will satisfactorily perform their safety functionsand that appropriate records of the testing of componentsimportant to safety be maintained by or under the controlof the nuclear power unit licensee throughout the life ofthe unit.

Section 50.55a, "Codes and Standards," of 10 CFRPart 50 requires, in part, that structures, systems, andcomponents be designed, fabricated, erected, constructed,tested, and inspected to quality standards commensuratewith the importance of the safety function to be performed.Section 50.55a further requires that American Society ofMechanical Engineers Boiler and Pressure Vessel Code(ASME B&PV Code) Class 1 components meet the require-ments set forth in Section XI, "Rules for Inservice Inspectionof Nuclear Power Plant Components," of the ASME Code.

Criterion XII, "Control of Measuring and Test Equipment,"of Appendix B, "Quality Assurance Criteria for NuclearPower Plants and Fuel Reprocessing Plants," to 10 CFRPart 50 requires, in part, that measures be established toensure that instruments used in activities affecting qualityare properly controlled, calibrated, and adjusted at specifiedperiods to maintain accuracy within necessary limits.

Criterion XVII, "Quality Assurance Records," of Appen-dix B requires, in part, that sufficient records be maintainedto furnish evidence of activities affecting quality. Consistentwith applicable regulatory requirements, the applicant isrequired to establish such requirements concerning recordretention as duration, location, and assigned responsibility.

This guide describes procedures acceptable to the NRCstaff for implementing the above requirements with regardto the preservice and inservice examinations of reactorvessel welds in light-water-cooled nuclear power plants byultrasonic testing (UT). The scope of this guide is limited toreactor vessel welds and does not apply to other structuresand components such as piping.

B. DISCUSSION

Reactor vessels must periodically be volumetricallyexamined according to Section XI of the ASME Code,which is incorporated by reference, with NRC staff modifica-tions, in § 50.55a of 10 CFR Part 50. The rules of Section XIrequire a program of examinations, testing, and inspectionsto evidence adequate safety. To ensure the continuedstructural integrity of reactor vessels, it is essential thatflaws be reliably detected and evaluated. It is desirable thatresults from prior UT examinations be compared to resultsfrom subsequent examinations so that flaw growth ratesmay be estimated. Lack of reliability of UT examinationresults is partly due to the reporting of ambiguous results,such as reporting the length of flaws to be shorter duringsubsequent examinations. This lack of reproducibility arisesbecause the Code requirements are not specific aboutmany essential variables in t e UT procedures. Recommenda-tions of this guide provide guidance that would help toobtain reproducibility of results. Reporting of UT indicationsas recommended in this guid:, will help to provide a meansfor assessing the ambiguity ol t' ý reported data.

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 Implementingspecific 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. ProductsGuides are not substitutes for regulations, and compliance with 2. Research and Test Reactors 7. Transportationthem is not required. Methods and solutions different from those set 3. Fuels and Materials Facilities 8 Occupational Healthout in the guides will be acceptable if they provide a basis for the 4. Environmental and Siting 9: Antitrust and Financial Reviewfindings requisite to the Issuance or continuance of a permit or 5. Materials and Plant Protection 10. Generallicense by the Commission.

Copies of Issued guides may be purchased at the current GovernmentThis 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 Printlng Office.guides are encouraged at all times, and guides will be revised, as Information on the subscription service and current GPO prices mayappropriate, 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.

Operating and licensing experience1' 2' 3 and industrytests4 have indicated that UT. procedures that have.&beenused for examination of reactor vessel welds'may1 n16t .be.adequate to consistently detect and feliably dhatacterizeflaws during inservice examination of reactors. This lack ofreproducibility of location and characterization of flaws hasresulted in the need for additional examinationsi andevaluations with associated delays in the licensinig process.

1. INSTRUMENT SYSTEM PERFORMANCE CHECKS

Instrument system performance checks to determine thecharacteristics of the UT system should be performed atintervals short enough to permit each UT examination to becorrelated with particular system performance parameters tohelp compare results. These determinations will help make itpossible to judge whether differences in observations madeat different times are due to changes in the instrument systemcharacteristics or are due to real changes in the flaw size andcharacteristics. Determinations for "Frequency-AmplitudeCurve" and "Pulse Shape" recommended in regulatory posi-tions 1.4 and 1.5 may be made by the licensee's examinationagent by using any of the common industry methods formeasuring these parameters as long as these methods areadequately documented in the examination record. Thesemeasurements may be performed in the laboratory beforeand after each examination, provided the identical equip-ment combination (i.e., instrumentation, cable, and searchunit) is used during the examination.

These determinations are to aid third-party evaluationswhen different equipment is used to record indications onsubsequent examinations and are not intended to qualifysystems for use.

The intent of regulatory position 1.5 is to establish theinstrument pulse shape in a way that actual values of pulselength and voltages can be observed on an oscilloscope. Thecalibrated time base does not necessarily have to follow thetime base of the distance-amplitude correction (DAQ curve butmay be chosen to suitably characterize the initial pulse. Thepulse shape record will assist in analyzing potential differencesin flaw response between successive examinations (i.e., is thedifference due to flaw growth or system change).

* Pulse shape is best determined by using a high-impedanceoscilloscope with the transducer disconnected from theinstrument.

2. CALIBRATION

According to Appendix I, Article I, 1-4230, Section XI ofthe ASME Code, 1974 edition, instrument calibration for

1 "tUltrasonic Reinspection of Pilgrim 1 Reactor Vessel NozzleN2B," John H. Gieske, NUREG-6502.

2 "Summary Hatch Nuclear Plant Unit I Reactor Pressure VesselRepair," 1972, Georgia Power Company.

3 "Summary of the Detection and Evaluation of UltrasonicIndications - Edwin Hatch Unit I Reactor Pressure Vessel," January1972, Georgia Power Company.

4'Round robin tests conducted by the Pressure Vessel ResearchCommittee (PVRC) of the Welding Research Council for UT ofthick section steels.

performance characteristics (amplitude linearity andamplitude control linearity) is to be verified at the beginning-of., each day of examination. Requirements in Article 4,Section V, 1977 edition, which is referenced by Section XI,for the periodic check of instrument characteristics (screenheight linearity, amplitude control linearity, and beamspread measurements) for UT examination of reactor:pressure vessels have been relaxed. The interval betweenperiodic checks has been extended from a period of 1 dayto a period of extended use or every 3 months, whichever isless. This change has not been justified on the basis ofstatistically significant field data. Performance stability ofautomated electronic equipment is dependent on systemperformance parameters (essential variables), and the ASMECode has no quality standards to control these performanceparameters. Until the performance stability of UT systemscan be ensured by the introduction of quality standards,it is not reasonable to increase the period between calibrationchecks. Therefore, recommendations have been made tocheck instrument performance parameters more frequentlythan is specified in the ASME Code.

Requirements of Appendix I, Article 1, 1-4230, Section XIof the ASME Code, 1974 edition, state:

"System calibration shall be cfecked by verifying thedistance-amplitude correction curve (1-4420 or 1-4520)and the sweep range calibration (1-4410 or 1-4510) at thestart and finish of each examination, with any change inexamination personnel, and at least every 4 hours duringan examination."

In the 1977 edition, these requirements were changed.According to Article 4 (T-432.1.2), Section V of the ASMECode, 1977 edition, the following applies:

"A calibration check on at least one of the basic reflectorsin the basic calibration block or a check using a simulatorshall be made at the finish of each examination, every4 hours during the examination and when examinationpersonnel are changed."

This requirement has several minor deficiencies, including

the following:

a. One-Point Check

A calibration check is now required on only one of thebasic reflectors. As a result, the accuracy of only one pointon the DAC curve, and not the accuracy of three points aspreviously required, is checked. This alteration wouldpermit the instrument drift for other metal path distancesto go unnoticed, which is not desirable.

b. Secondary Reference

The change allows a one-point check by a mechanicalor electronic simulator instead of a check against the basiccalibration block. A mechanical simulator could be aplastic,. steel, or aluminum block with a single referencereflector, which may be a hole or a notch. Without specifieddetails, the electronic simulator could be any device that

1.150-2

provides an electrical signal. With the resulting uncertainty,there may be errors in checking against the secondaryreference (simulator), the magnitude of whichis undefinedand unknown.

c. Electronic Simulator

Subarticle T-432.1.3 of Article 4, Section V of theASME Code, 1977 edition, allows the use of an electronicsimulator and also permits the transducer sensitivity to bechecked separately. Both these provisions may introduceerrors that will be very difficult to detect.

To avoid the introduction of errors and to ensurerepeatability of examinations at a later date, it would beadvisable to check the calibration of the entire systemrather than that of individual components. Checking systemcalibration without the transducer and the cable is notadvisable because these tests do not detect possible leakageor -resistance changes at the connectors. This is especiallyimportant when the UT examination is performed underconditions of high humidity or under water and the connec-tors may not be waterproof or moistureproof. Checking thetransducer sensitivity separately (sometimes weeks inadvance) also neglects the effects of possible damage due totransport or use. The transducer characteristics may changebecause of damage to or degradation of internal bondingagents or inadvertent damage to the transducer element.Further, the use of an electronic block simulator (EBS) as asecondary standard introduces an error band in the calibra-tion process. The error band may depend on, among others,the following factors:

(1) Drift due to ambient temperature change.(2) Drift due to high temperature storage.(3) Drift due to high humidity storage.(4) Drift due to vibration and shock loading during

shipment.(5) Degradation of the memory device used to store

the reference signal information due to vibra-tion, shock, aging, or heat effects.

To ensure stability, computer systems are generallykept in an air conditioned environment; however, EBSsystems are not usually kept in a controlled environment.

Error band for one particular type of instruments

was determined to be in the range of ±6 percent. The errorband for other instruments may be in a different range andmay vary for the same instrument if memory devices orcomponents of different quality are used at a later date.The error band is dependent on the temperature extremes,shock loadings, and vibrations suffered by the instrument.Since the error band value depends on these parameters, itwould be advisable to ensure, throui'h recording instn'ments,that the EBS was not subjected to higher temperatures(container lying in the sun) and greater shock (container

5",Calibration Verification of Ultrasonic Examination Systems withthe Electronic Block Simulator," D. 1. Boomgard et al., August 1979,Report No. WCAP-954S, Westinghouse Electric Corporation, NuclearService DliWlslon, P.O. Box 272 8, Pittsburgh, PA 15 230.

dropped) during transport than those parameters thatserved as a basis for defining the error band.

Use of electronic simulators would be permissible ifthey can check the calibration of the UT system as a wholeand the error band introduced by their use can be relied onand taken into consideration.

d. Static Versus Dynamic Reflector Responses

With some automated systems, the DAC curve ismanually established. In these cases, the signal is maximizedby optimizing the transducer orientation toward thecalibration holes. Subsequently, detection and sizing offlaws are based on signals received from a moving transducerwhere no attempt is made (or it is not possible) to maximizethe signal even for significant flaws. This procedure neglectsseveral sources of error introduced by the possible variationin signal strength caused by:

(1) Differences between the maximized signaland the unmaximized signal.

(2) Loss in signal strength due to the separation ofthe transducer from the metal surface becauseof the viscosity of the coupling medium (plan-ing effects).

(3) Variation in contact force and transducercoupling efficiency.

(4) Loss in signal strength due to structural vibra-tion effects in the moving transducer mountand other driving mechanisms.

(5) Loss in signal strength due to the tilting causedby. the mounting arrangement in some trans-ducer mounts.

Because of the above, it would be advisable to establishthe DAC curve under the same conditions as those underwhich scanning is performed to obtain data for detectionand sizing. It would be acceptable to establish a DAC curveby maximizing signal strength during manual scans whensignals are also maximized for flaw sizing. However, itwould not be advisable to use manually maximized signalsto establish the DAC curve when data are obtained later bymechanized transducers (where signals cannot be maximized)for the detection and sizing of flaws without adjustment forthe potential error introduced. In these situations, anacceptable method would be to establish DAC curves usingmoving transducers or to establish correction factors thatmay be used to adjust signal strength. It would be prudentto use care and planning in establishing correction factors.For example, establishing a ratio between a dynamic andstatic mode under laboratory conditions using a precisiontransducer drive and stiff mounting may have very little incommon with the transducer mounting and traverse condi-tions of the actual examination setup. If correction factorsare to be used, it would be worthwhile to build eitherfull-scale mockups or consider the variation of all theimportant parameters in a suitable model taking into

1.150-3

consideration scaling laws on variables such as mass, vibration,and stiffness constants. 'It would be advisable to confirm thescaling law assumptions and predictions for vibration andviscosity effects before correction factors are used for settingscanning sensitivity levels.

Differences in the curvature and surface finish betweencalibration blocks and vessel areas could change the dynamicresponse, so it may be advisable to establish correction factorsbetween dynamic and static responses from the indicationsthat are found during examination. This would avoid thedifficulties associated with establishing a dynamic responseDAC curve and still take all the factors into consideration.

e. Secondary DAC

During some manual scans, the end point of the DACcurve may fall below 20 percent of the full screen height.When this happens, it is difficult to evaluate flaws on the20 percent and 50 percent DAC basis in this region sincethe 20 percent and 50 percent DAC points may be too closeto the baseline. To overcome this difficulty, it is advisablethat a secondary DAC curve using a higher-gain setting bedeveloped so that 20 percent and 50 percent DAC points maybe easily evaluated. For this purpose, it is advisable that thegain be increased sufficiently to keep the lowest point ofthe secondary DAC curve above 20 percent of screen height.

The secondary DAC curves need not be generatedunless they are required. If electronic DAC is used andamplitudes are maintained above 20 percent of full screenheight, a secondary DAC would not be necessary.

f. Component Substitution

A calibration check should be made each time acomponent is put back into the system to ensure that suchcomponents as transducers, pulsers, and receivers were notdamaged while they were in storage. This will ensureelimination of the error band and mistakes in resetting thevarious control knobs.

g. Calibration Holes

Comparison of results between examinations performedat different times may be facilitated if the same equipmentis used and if the reflections from growing flaws can becompared to the same reference signal. Reference signalsobtained from a calibration block depend on, among otherthings, the surface roughness of. the block and the reflectorholes. Therefore, these surfaces should be protected fromcorrosion and mechanical damage and also should not bealtered by mechanical or chemical means between successiveexaminations. If the reference reflector holes or the blocksurface are given ahigh polish by any chemical or mechanicalmeans, the amplitude of the reflections obtained from thesereflector holes may be altered. Polishing the holes or theblock surface is not forbidden by the ASME Code. However,this possibly altered, amplitude could affect the sizing ofindications found during any examination. At this time, norecommendations are being made to control the surfaceroughness of the block or the above-mentioned reflector

holes; however, if the block or these holes are polished, thisfact should be recorded for consideration if a review of theUT data becomes necessary at a later date.

3. NEAR-SURFACE EXAMINATION AND SURFACERESOLUTION

Sound beam attenuation in any material follows adecaying curve (exponential function); however, in somecases the reflection from the nearest hole is smaller than thereflection from a farther hole. This makes it difficult todraw a proper DAC curve. In such cases, it may be desirableto use a lower frequency or a smaller transducer for flawdetection near the beam-entry surface to overcome thedifficulty of marginal detectability.

Near-field effects, decay time of pulse reflections,shadow effects, restricted access, and other factors do notpermit effective examination of certain volume areas in thecomponent. To present a clear documentation and recordof the volume of material that has not been effectivelyexamined, these volume areas need to be identified. Recom-mendations are provided to best estimate the volume in theregion of interest that has not been effectively examined,such as volumes of material near each surface (because ofnear-field effects of the transducer and ring-down effects ofthe pulse due to the contact surface), volumes near interfacesbetween cladding and parent metal, and volumes shadowedby laminar flaws.

4. BEAM PROFILE

Beam profile is one of the main characteristics of a trans-ducer. It helps to show the three-dimensional distribution ofbeam strength for comparing results between examinationsand also for characterizing flaws. The beam profile needs tobe determined and recorded so that comparisons may bemade with results of successive examinations.

5. SCANNING WELD-METAL INTERFACE

The amount of energy reflected back from a flaw isdependent on its surface characteristics, orientation, andsize. The present ASME Code procedures rely on theamplitude of the reflected signal as a basis for judging flaws.This means that the size estimation of a defect depends onthe proportion of the ultrasonic beam reflected back to theprobe. The reflection behavior of a planar defect, whichlargely depends on the incident beam angle when a singlesearch unit is used to characterize the flaw, is thus a decisivefactor in flaw estimation. The larger the size of a planardefect, the narrower is the reflected sound beam. Thenarrow reflected sound beam makes the flaw very difficultto detect in most cases (unless the beam angle is right). 6' 7

6 "Probability of Detecting Planar Defects in Heavy Wall Welds byUltrasonic Techniques According to Existing Codes," Dr. Ing. Hans-Jurgen Meyer, Quality Department of M.A.N., Nurnberg, D 8SOONurnberg 115.

7 "Refiection of Ultrasonic Pulses from Surfaces," Haines andLangston Central Electricity Generating Board, U.K. (CESB) ReportNumber RD 18/N411S.

1.150-4

Therefore, the beam angles used to scan welds should beoptimized and should be based on the geometry of theweld/parent-metal interface. At least one of these anglesshould be such that the beam is almost perpendicular (±+15degrees to the perpendicular) to the weld/parent-metalinterface, unless it can be demonstrated that large (Code-unacceptable) planar flaws unfavorably oriented, parallel tothe weld-metal interface, can be detected by the UT tech-nique being used. In vessel construction, some weld preps areessentially at right angles to the metal surface. In these cases,use of shear wave angles close to 75 degrees is not recom-mended. Two factors would make the use of shear waveangles close to 75 degrees inadvisable, - first, the test distancesnecessary become too large resulting in loss of signal, andsecond, the generation of surface waves tends to confusethe interpretation of results. In these cases, use of alternativevolumetric nondestructive examination (NDE) techniques,as permitted by Subarticle IWA-2240, Section XI of theASME Code, should be considered. Alternative NDEtechniques to be considered may include high-intensityradiograph or tandem-probe ultrasonic examination of theweld-metal interface. To avoid the possibility of missinglarge flaws, particularly those that have an unfavorableorientation, it is desirable that the back reflection amplitude,while scanning with a straight beam, be monitored over theentire volume of the weld and adjacent base metal. Anyarea where a reduction of the normal back-surface reflectionamplitude exceeds 50 percent should be examined by anglebeams in increments of ±15 degrees until the reduction ofsignal is explained. Where this additional angle beamexamination is not practical, it may be advisable to considerexamining the weld. by a supplementary volumetric NDEtechnique.

6. SIZING

The depth or through-wall dimension of flaws is moresignificant than the length dimension, according to firacturemechanics analysis criteria. Using the single-probe pulse-echotechnique, it is possible, depending on flaw orientation,that some large flaws may not reflect much energy to thesearch unit.6 Because of this possibility, the depth dimen-sion of the flaw should be conservatively sized unless there isevidence to prove that the flaw orientation is at right anglesto the beam. It is recommended that indications that are asso-ciated with through-thickness flaws and do not meet Code-allowable criteria or criteria recommended in this guide besized at 20 percent DAC as well as at 50 percent DAC.

In certain cases, It is possible for various reasons that aflaw would not reflect enough energy to the search unit tomake the indication height 50 percent of the DAC curveheight. However, if such a flaw were large, a persistentsignal could be obtained over a large area. It is thereforerecommended that all continuous signals that are 20 percentof DAC with transducer travel movement of more thanI inch plus the beam spread (as defined in Article 4, non-mandatory Appendix B, Section V of the ASME Code,1977 edition) should be considered significant and shouldbe recorded and investigated further. The beam spreadeffect in some cases can make very small flaws appear to belarge when judged at 20 percent DAC; hence, beam spread

has to be considered in judging the significance of flaws. 8

It is therefore recommended that only signals with a totaltransducer travel movement greater than the beam spreadshould be considered significant.

7. REPORTING OF RESULTS

This guide gives recommendations for recording the charac-teristics of the UT examination system. This informationcan be of significance in later analysis for determining thelocation, dimensions, orientation, and growth rate of flaws.

Records pertaining to UT examinations should be con-sidered quality. assurance records. Recommendations on the,collection, storage, and maintenance of these records aregiven in Regulatory Guide 1.88, "Collection, Storage, andMaintenance of Nuclear Power Plant Quality Assurance Re-cords." Availability of these records at a later date will permita review of the UT results from the data gathered duringprevious ultrasonic examinations.

When ultrasonic examination is performed, certain vol-umes of material such as the following are not effectivelyexamined:

a. Material volume near the front surface because of near-field effects, cladding disturbance, or electronic gating.

b. Material volume near the surface because of surface

roughness or unfavorable flaw orientations.

c. Volumes shadowed by insulation or part geometry.

In some cases, as much as 1 inch (25.4 mm) or morebelow the surface is not examined because of the electronicgate setting. This means that the unexamined volume maycontain flaws that would be unacceptable according toSection XI, ASME Code, as follows:

a. Without evaluation (deeper than approximately 0.2inch).

b. Even after evaluation (deeper than approximately0.85 inch).

Assuming an aspect ratio of 0.1, according to IWB-3510.1,Section XI, ASME Code, flaws 0.2 inch deep would beunacceptable for a 9-inch wall thickness.

Typically a BWR reactor pressure vessel (RPV) wall inthe beltline region is 6 inches thick and a PWR-RPV wall is8.5 inches thick. During flaw evaluation, where the walltemperature is high and the available toughness is high, andthe calculated critical surface flaw depth (a.) exceeds the wallthickness (t), ac is taken 9 as the wall thickness. According toIWB-3600, Section XI, the allowable end-of-life size is af =0.1a.. Flaws exceeding this allowable value, which would

8 "Ultrasonic Emonination Comparison of Indication and Actual Flawin RPV,' Ishi Kawajima-Harima Industries Co., Ltd., January 1976.

9"Flaw Evaluation Procedures: ASME Section XU-EPRM," NP-719-SR,special report, August 1978.

1.150-5

be 0.85 inch for a PWR and 0.65 inch for a BWR, will haveto be repaired. The above example illustrates the importanceof blanking out the electronic indication signals and notexamining the surface volume to a depth, of 1 inch. Sincethe flaws that can be missed because of electronic gating maybe larger than the flaws permitted with or without evaluation,this unexamined volume is important and needs to be identified.

In certain specific cases, areas were not examinedbecause insulation was in the way and the transducer couldnot scan the volume of interest. NRC was not informed ofthis situation until much later. In view of the above and toavoid licensing delays, it is advisable that the volume of areasnot examined for any or all of the above reasons be reported.

The volumes of material that are not effectively examineddepend on the particular part geometry and unique situa-tions associated with each RPV. During identification ofthe material volumes that have not been examined, considera-tion should be given to the types of flaws that are currentlybeing reported in some of the operating plants. Theseinclude stress corrosion cracks in the heat-affected zone,fatigue cracks, and cracks that are close to the surfaceand sometimes penetrate the surface. These volumes ofmaterial should be identified and reported to NRC alongwith the report of welding and material defects in accordancewith the recommendation of regulatory position 2.a(3) ofRegulatory Guide 1.16, "Reporting of Operating Informa-tion-Appendix A Technical Specifications."

C. REGULATORY POSITION

Ultrasonic examination of reactor vessel welds should beperformed according to the requirements of Section XI ofthe ASME B&PV.Code, as referenced in the Safety AnalysisReport (SAR) and its amendments, supplemented by thefollowing:

I. INSTRUMENT PERFORMANCE CHECKS

The checks described in paragraphs 1.2 through 1.5 shouldbe made for any UT system used for the recording and sizingof reflectors in accordance with regulatory position 6 andfor reflectors that exceed the Code-allowable criteria.

1.1 Frequency of Checks

As a minimum, these checks should be verified within I daybefore and within I day after examining all the welds that needto be examined in a reactor pressure vessel during one outage.Pulse shape and noise suppression controls should remain atthe same setting during examination and calibration.

1.2 Screen Height Linearity

Screen height linearity of the ultrasonic instrumentshould be determined according to the mandatory Appen-dix I to Article 4, Section V of the ASME Code, within thetime limits specified in regulatory position I.I.

1.3 Amplitude Control Linearity

Amplitude control linearity should be determined accordingto the mandatory Appendix II of Article 4, Section V of theASME Code, 1977 edition, within the time limits specified inregulatory position 1.1.

1.4 Frequency-Amplitude Curve

A photographic record of the frequency-amplitude curveshould be obtained. This record should be available forcomparison at the inspection site for the next two successiveinspections of the same volume. The reflector used ingenerating the frequency-amplitude curves as well as theelectronic system (i.e., the basic ultrasonic instrument,gating, form of gated signal, and spectrum analysis equip-ment) and how it is used to capture the frequency-amplitudeinformation should be documented.

1.5 Pulse Shape

A photographic record of the unloaded initial pulseagainst a calibrated time base should be obtained. The timebase and voltage values should be identified and recordedon the horizontal and vertical axis of The above photographicrecord of the initial pulse. The method used in obtainingthe pulse shape photograph, including the test point atwhich it is obtained, should be documented.

2. CALIBRATION

System calibration should be checked to verify the DACcurve and the sweep range calibration per nonmandatoryAppendix B, Article 4, Section V of the ASME Code, as aminimum, before and after each RPV examination (or eachweek in which it is in use, whichever is less) or each time anycomponent (e.g., transducer, cable, connector, pulser, orreceiver) in the examination system is changed. Where possible,the same calibration block should be used for successive in-service examinations of the same RPV. The calibration sideholes in the basic calibration block and the block surface shouldbe protected so that their characteristics do not change duringstorage. These side holes or the block surface should not bemodified in any way (e.g., by polishing) between successiveexaminations. If the block surface or the calibrationreflectorholes have been polished by any chemical or mechanical means,this fact should be recorded.

2.1 Calibration for Manual Scanning

For manual scanning for the sizing of flaws, static calibra-tion may be used if sizing is performed using a static trans-ducer. When signals are maximized during calibration, theyshould also be maximized during sizing. For manual scanningfor the detection' of flaws, reference hole detection should beshown at scanning speed and detection level set accordingly(from the dynamic DAC).

1.150-6

2.2 Calibration for Mechanized Scanning

When flaw detection and sizing are to be done bymechanized equipment, the calibration should be performedusing the following guidelines:

a. Calibration speed should be at or higher than thescanning speed.

b. The direction of transducer movement during calibra-tion should be the same as the direction during scanningunless (l)it can be shown that the change in scanning directiondoes not make a difference in the sensitivity and vibrationbackground noise received from the search unit or. (2) thesedifferences are taken into account by a correction factor.

c. For mechanized scanning, signals should not bemaximized during the establishment of the DAC curve.

d. One of the following alternative guidelines should befollowed for establishing the DAC curve:

(I) The DAC curve should be established using amoving transducer mounted on the mechanism that will beused for examination of the component.

(2) Correction factors between dynamic and staticresponse should be established using full-scale mockups.

(3) Correction factors should be established usingmodels and taking scaling factors into consideration (assumedscaling relationship should be verified).

(4) Correction factors between dynamic and staticresponse should be established from the indications that arefound during examination for sizing. For detection of flawsduring the initial scan, correction factors may be assumedbased on engineering judgment. If assumed correctionfactors are used for detection, these factors should later beconfirmed on indications from flaws in the vessel during theexamination. Deviation from the assumed value maysuggest reexamining the data.

2.3 Calibration Checks

If an EBS is used for calibration check, the followingshould apply:

a. The significant DAC percentage level used for thedetection and sizing of indications should be reduced totake into account the maxirmun error that could be introducedin• th'z systemn by the variation of resistance or leakage inthe connectors or other causes.

b. Calibration checks should be performed on thecomplete connected system (e.g., transducer and cablesshould not be checked separately).

c. Measures should be taken to ensure that the differentvariables such as temperature, vibration, and shock limitsfor which the EBS error band is determined are not exceededduring trarisport, use, storage, etc.

d, When a universal calibration block is used and someor all of the reference holes are larger than the reflectorholes at comparable depths recommended by Article 4, Sec-tion V, of the ASME Code, 1980 edition, a correction factorshould be used to adjust the DAC level to compensate forthe larger reflector holes. Also, if the reactor pressure vesselhas been-previously examined by using a conventional block,a ratio between the JDAC curves obtained from the two

* blocks should be noted (for reference) with the significantindications data.

3. NEAR-SURFACE EXAMINATION AND SURFACERESOLUTION

The capability to effectively detect defects near thefront and back surfaces of the actual component should beestimated. The results should be reported with the report ofabnormal degradation of reactor pressure boundary inaccordance with the recommendation of regulatory posi-tion 2.a(3) of Regulatory Guide 1.16. In determining thiscapability, the effect of the following factors should also beconsidered:

a. If an electronic gate is used, the time of start and stopof the control points of the electronic gate should berelated to the volume of material near each surface that isnot being examined.

b. The decay time, in terms of metal path distance, ofthe initial pulse and of the pulse reflections at the front andback surface should be considered.

c. The disturbance created by the clad-weld-metalinterface with the parent metal at the front or the backsurface should be related to the volume of material near theinterface that is not being examined.

d. The disturbance created by front and back metalsurface roughness should be related to the volume of.material near each surface that is not being examined.

4. BEAM PROFILE

The beam profile should be determined if any recordableflaws are detected. This should be done for each search unitused during the examination by a procedure similar to thatoutlined in the nonmandatory Appendix B (B-60), Article 4,Section V of the ASME Code, 1980 edition, for determiningbeam spread. Beam profile curves should be determined foreach of the holes in the basic calibration block. Interpola-tion may be used to obtain beam profile correction for assess-ing flaws at intermediate depths for which the beam profilehas not been determined.

5. SCANNING WELD-METAL INTERFACE

The beam angles used to scan welds should be based onthe geometry of "the weld/parent-metal interface. At leastone of these angles should be such that the beam is almostperpendicular (±+15 degrees to the perpendicular) to theweld/parent-metal interface unless it can be demonstratedthat unfavorably oriented planar flaws can be detected by

1.150-7

the UT technique being used. Otherwise, use of alternativevolumetric NDE techniques, as permitted by the ASMECode, should be considered. Alternative NOE techniquesmay be considered to include high-intensity radiography ortandem-probe ultrasonic examination of the weld-metalinterface.

6. SIZING

Indications from geometric sources need not be recorded.

6.1 Traveling Indications

Indications that travel on the horizontal baseline of thescope for a distance greater than indications from thecalibration holes (at 20 percent DAC amplitude) should berecorded. Indications that travel should be recorded andsized at 20 percent DAC. Where the indication is sized at20 percent DAC, this size may be corrected by subtractingfor the beam width in the through-thickness directionobtained from the calibration hole (between 20 percentDAC points) that is at a depth similar to the flaw depth. Ifthe indication exceeds 50 percent DAC, the size should berecorded by measuring the distance between 50 percentDAC levels without using the beam-width correction. Thedetermined size should be the larger of the two.

6.2 Nontraveling Indications

Nontraveling indications above 20 percent DAC levelthat persist for a scanning distance of more than 1 inch plusthe beam spread between 20 percent DAC points (asdefined by nonmandatory Appendix B, Article 4, Section Vof the ASME Code, 1977 edition) should be consideredsignificant. The size of these flaws should be determined bymeasuring the distance between points at 50 percentDAC and between points at 20 percent DAC where thebeam-width correction is made only for the 20 percentDAC size. The recorded size of the flaw would be the largerof the two determinations. If it can be adequately demon-strated that a nontraveling indication is from a geometricsource (and not a flaw), there is no need to record thatindication.

The following information should also be recorded forindications that are reportable according to this regulatoryposition:

a. Indications should be recorded at scan intervals nogreater than one-fourth inch.

b. The recorded information should include the indica-tion travel (metal path length) and the transducer positionfor 10 percent, 20 percent, 50 percent, and 100 percentDAC and the maximum amplitude of the signal.

7. REPORTING OF RESULTS

Records10 obtained while following the recommendationsof regulatory positions 1.2, 3, 5, and 6, along with discus-sions and explanations, if any, should be kept available atthe site for examination by the NRC staff. If the size ofan indication, as determined in regulatory positions 6.1 or6.2, equals or exceeds the allowable limits of Section XI ofthe ASME Code, the indications should be reported asabnormal degradation of reactor pressure boundary inaccordance with the recommendation of regulatory posi-tion 2.a(3) of Regulatory Guide 1.16.

Along with the report of ultrasonic examination testresults, the following information should also be included:

a. The best estimate of the error band in sizing the flawsand the basis for this estimate should be given.

b. The best estimate of the portion of the volumerequired to be examined by the ASME Code that has notbeen effectively examined such as volumes of material neareach surface because of near-field or other effects, volumesnear interfaces between cladding and parent metal, volumesshadowed by laminar material defects, volumes shadowedby part geometry, volumes inaccessible to the transducer,volumes affected by electronic gating, and volumes near thesurface opposite the transducer.1 -

c. The material volume that has not been effectivelyexamined by the use of the above procedures may beexamined by alternative effective volumetric NDE techniques.If one of these alternative NDE techniques is a variation ofUT, recommendations of regulatory positions I and 3should apply. A description of the techniques used shouldbe included in the report. If other volumetric techniques orvariations of UT are used as indicated in regulatory posi-tion 5, the effectiveness of these' techniques should bedemonstrated and the procedures reported for review bythe NRC staff.

8. ALTERNATIVE METHOD

Appendix A contains an alternative method that is pre-sented in Chapter 3 of the document "RecommendedChanges to Regulatory Guide 1.150," prepared by the AdHoc Committee of the Electric Utility Industry and pre-sented to the NRC in August 1982.12 The NRC staff hasevaluated this method and has found it to be an accept-able alternative to that presented in paragraphs 1 through7 of this regulatory position.

10 Any guidance in this document related to information collec-tion activities has been cleared under OMB Clearance No. 3150-0011.

1 lit should be noted that the licensee is required to apply for relieffrom impractical ASME Code requirements according to § 50.55a of10 CFR. If the licensee is committed to examine a weld as per theinspection plan in the plant SAR, the licensee is required to file anamendment when the commitments made in the SAR cannot be met.

12This document is available through Mr. Larry Becker, /o 1. A.Jones Applied Research Co., P.O. Box 217097, Charlotte, NorthCarolina 28221.

1.150-8

D. IMPLEMENTATION

Except in those cases In which an applicant proposes anacceptable alternative method for complying with specifiedportions of the Commission's regulations, the methoddescribed herein will be used In the evaluation of (1)the results of inservice examination programs of all operatingreactors after July 15, 1981, and (2) the results of preservice

examination programs of all reactors under constructionperformed after January 15, 1982.

The recommendations of this guide are not intended toapply to. preservice examinations that have already beencompleted. The NRC staff intends to recommend that alllicensees modify their technical specifications to makethem consistent with the recommendations contained herein.

1.150-9

APPENDIX A

ALTERNATIVE METHOD

The following is reprinted with permission fromChapter 3 of the document "Recommended Changes toRegulatory Guide 1.150," prepared by the Ad Hoc Committeeof the Electric Utility Industry.

Ultrasonic examination of reactorvessel welds should be performedaccording to the requirements ofSection XI of the ASME B&PV Code, asreferenced in the Safety AnalysisReport (SAR) and its amendments,supplemented by the following:

1. INSPECTION SYSTEM PERFORMANCECHECKS

The conduct of a quality exami-nation requires that the performancecharacteristics of the inspectionsystem used be well defined anddocumented. This is particularlytrue for situations which requirecomparisons of examination resultsgenerated during successiveexaminations on the same components.

A system comprises:a. a transducer;b. a s;ingle channel instrument

or each channel of a multi-channelinstrument; and

c. a given cable type andlength.

The checks described in paragraphs1.1 and 1.2 should be made for any UTsystem used for inspection of reactorpressure vessel (RPV) welds.

The field performance checksdescribed in 1.2 (with the possibleexception of 1.2.c.) should be conductedon a basic calibration block thatrepresents the thickness range tobe examined.

1.1 Pre-exam Performance Checks

a. Frequency of checks

These checks should be verifiedwithin six months before reactor pressurevessel examinations performed during oneoutage. Pulse shape and noisesuppression controls should remain at thesame settings during calibration and exa-mination.

b. RF Waveform

A record of the RF (radiofre-quency) pulse waveform from a referencereflector should be obtained for eachsearch unit used in the examinationin a manner which will'provide frequencyamplitude information. At the highestamplitude portion of the beam, the RFreturn signal should be recorded beforeit has been rectified or conditionedfor display. The reflector used ingenerating the RF. return signal aswell as the electronic system (i.e.,the basic ultrasonic instrument,gating, and form of gated signal)should be documented. These recordsshould be used for comparison withprevious and future records.

1.2 Field Performance Checks

a. Frequency of Checks

As a minimum, these checksshould be verified on-sitebefore and after examining allthe welds that need to beexamined in a reactor pressurevessel during one outage. Pulseshape and noise suppression controlsshould remain at the same settingduring examination and calibration.

1.150-10

b. Instrument Sensitivity duringLinearity Checks

The initial instrument sen-sitivity during the performance of1.2.e. should be such that itfalls at the calibration sensitivityor at some point between the.calibration sensitivity and thescanning sensitivity.

c. RF Waveform

A record of the RF(radiofrequency) pulse waveform froma reference reflector should beobtained and recorded in a mannerthat will permit extraction offrequency amplitude information.At the highest amplitude portionof the beam, the RF return signalshould be recorded before ithas been rectified or conditionedfor display. This should bedetermined on the same reflectoras that used in 1.1.b. above.This record should be retainedfor future reference.

d. Screen Height Linearity

Screen height linearity ofthe ultrasonic instrument should bedetermined according to the manda-tory Appendix I to Article 4,Section V of the ASME Code orAppendix I to Section XI of theASME Code.

e. Amplitude Control Linearity

Amplitude control should bedetermined according to the man-datory Appendix II of Article 4,Section V of the ASME Code or AppendixI to Section XI of the ASME Code.

f. Angle Beam ProfileCharacterization

The vertical beam profileshould be determined for each searchunit used during the examination bya procedure similar to that outlined

in nonmandatory Appendix B-60,Article 4, Section V of the ASMECode or Appendix I to Section XI ofthe ASME Code. Beam profile curvesshould be determined at differentdepths to cover the thicknessesof materials to be examined.Interpolation may be used toobtain beam profile correctionfor assessing flaws at intermediatedepths for which beam profile hasnot been determined.

Beam profile measurementsshould be made at the sensitivityrequired for sizing. For example,sizing to a 20 percent DAC criteriarequires that the beam profile bedetermined at 20 percent DAC.

2. CALIBRATION

System calibration should beperformed to establish the DACcurve and the sweep range cali-bration in accordance withArticle 4, Section V of theASME Code or Appendix I toSection XI. Calibrationshould be confirmed beforeand after each RPV examination,or each week in which thesystem is in use, whicheveris 'less. Where possible,the same calibration blockshould be used for successiveinservice examinations of theRPV.

2.1 Calibration for Manual Scanning

For manual sizing of flaws,static calibration may be used ifsizing is performed using a statictransducer. When signals aremaximized during calibration,they should also be maximizedduring sizing. For manualscanning for the detection offlaws, reference hole detectionshould be shown at scanningspeed and detection levelset accordingly.

1.150-11

2.2 Calibration .for MechanizedScanning

When flaw detection is to bedone by mechanized equipment, thecalibration should be performedusing the following guidelines.

a. The DAC curve should beestablished using either a movingtransducer mounted on the mechanismthat will be used for examinationof the component or a mechanismthat duplicates the critical factors(e.g., transducer mounting, weight,pivot points, couplant) present inthe scanning mechanism.

b. Calibration speed should beat or higher than the scanning speed,except when correction factorsestablished in 2.2.d. are used.

c. The direction of transducermovement (forward or backward) duringcalibration to establish the DACcurve should be the same directionduring scanning unless it can beshown that a change in scanningdirection does not reduce flawdetection capability.

d- One of the followingalternative guidelines should befollowed to establish correctionfactors if static calibration isused.

(1) Correction factors betweendynamic and static rqsponse shouldbe established using the basiccalibration block or,

(2) Correction factors should beestablished using models and takingscaling factors into consideration(assumed scaling relationship shouldbe verified) or,

(3) Correction factors should beestablished using full scalemockups.

2.3 Calibration Confirmation

Calibration confirmation per-formed as mid-shift or interimconfirmation between on-sitecalibrations should comply withstability requirements in T-433,Article 4, Section V of the ASMECode.

When an electronic simulatoris used for on-site calibrationconfirmation after a Code-requiredblock calibration performed off-site, the following should alsoapply:

a. Complete system performanceshould be maintained stable prior tooff-site calibrations and on-sitecalibration confirmation by use oftarget reflectors. The targetreflectors should be mounted withidentical physical displacement inboth the off-site calibrationfacilities and the on-site mecha-nized equipment. Each on-siteperiodic calibration, should bepreceded by complete system per-formance verification using aminimum of two (2) target reflec-tors separated by a distance repre-senting 75 percent of maximum thick-ness to be examined.

b. Written records of calibra-tions should be established forboth target reflector responsesand Code calibration block DACcurves for each transducer. Thesewritten records may be used tomonitor drift since the originalrecorded calibration.

c. Measures should be taken toensure that the different variablessuch as temperature, vibration, andshock limits are minimized bycontrolling packaging, handling, andstorage.

1.150-12

2.4 Calibration Blocks 3.1 Internal Surface

Calibration blocks shouldcomply with Appendix I to SectionXI or Article 4, Section V of theASME Code. When an alternative calibra-tion block or a new conventional block isused, a ratio between the DAC.curves obtained from theoriginal block and from the newblock should be noted (forreference) to provide for a meaning-ful comparison of previous andcurrent data.

The calibration side-drilledholes in the basic calibration blockand the block surface should be protectedso that their characteristics do notchange during storage. These side-drilled holes or the block surfaceshould not be modified in any way(e.g., by polishing) between successiveexaminations. If the block surface orthe calibration reflector holes havebeen poITshed by any chemical ormechanical means, this fact shouldbe recorded.

3. EXAMINATION

The scope and extent of theultrasonic examinations shouldcomply with IWA-2000, Section XI ofthe ASME Code.

If electronic gating is used todefine the examination volume withinwhich indications are recorded, thestart and stop control points shouldinclude the entire required thicknessincluding the material near eachsurface. If a single gate is used,it should be capable of recordingmultiple indications appearing inthe gate. Alternative means ofrecording may be used providedthey do not reduce flaw detectionand recording capability.

Examination should be done witha minimum 25 percent scan overlapbased on the transducer elementsize.

The capability to effectivelydetect defects at the internalclad/base metal interface shall beconsidered acceptable if the exami-nation procedure(s) or technique(s)meet the requirements of Section 6.0of this document and demonstratethe following:

a. Procedures for examination.from the outer surface, or when usingfull vee from the inside surface,should include the use of the 2 percentnotch which penetrates the internal(clad) surface of the calibrationblocks, defined by Section XI,Appendix I, Figure 1-3131 orSection V, Article 4, FigureT-434.1. Procedures for exam-ination from the internal surfacewhen not using the full veeshould conform to paragraph3.1.b. below.

b. An alternate reflector, otherthan the 2 percent notch describedabove may be used provided: 1) that itis located at the clad /base metalinterface or at an equivalent distancefrom the surface, 2) that it doesnot exceed the maximum allowabledefect size, and 3) that equivalentor superior results can bedemonstrated.

c. The examination procedure(s)should provide for volumetric examin-ation of at least one inch of metalas measured perpendicular to thenominal location of the base metal-cladding interface.

3.2 Scanning Weld-Metal Interface

The beam angles used to scanwelds should be based on thegeometry of the weld/parent metalinterface. Where feasible for weldssuch as those identified in SectionT-441.4.2 of Article 4, Section V ofthe ASME Code, at least one angleshould be such that the beam is per-

1.150-13

pendicular (+ 15 degrees to theperpendiculaF) to the weld/parentmetal interface, or it should bedemonstrated that unfavorablyoriented planar flaws can bedetected by the UT technique beingused. If this is not feasible, useof alternative volumetric NOE tech-niques, as permitted by the ASMECode, should be considered.

4. BEAM PROFILE

(Delete entire paragraph.This section included in RecommendedChange 1.2.f., Angle Beam ProfileCharacterization.)

5. SCANNING WELD-METAL INTERFACE

(Delete entire paragraph. Thissection included in RecommendedChange 3.2, Scanning Weld-MetalInterface.)

6. RECORDING AND SIZING

The capability to detect,record and size the flaws delineatedby Section XI, IWB-3500 should bedemonstrated. The measurementtolerance established should beapplied when sizing flawsdetected and recorded duringscanning (see paragraph 7.a.).

6.1 Geometric Indications

Indications determined to befrom geometric sources need not besized. Recording of these indica-tions should be at 50 percent DAC.When indications are evaluatedas geometric in origin, the basisfor that determination should bedescribed. After recordingsufficient information toidentify the origin of thegeometric indication, furtherrecording and evaluation arenot required.

6.2 Indications with Changing MetalPath

a. Indications that change metalpath distances (indicating through-wall dimension) when scanned inaccordance with the requirements ofASME Section XI, for a distancegreater than that recorded from thecalibration reflector should berecorded.

b. Reflectors which are at metalpaths representing 25 percent and greaterof the through-wall thickness of thevessel wall measured from the innersurface should be recorded inaccordance with the requirements ofASME Section XI and characterizedat 50 percent DAC.

C. Reflectors which are withinthe inner 25 percent of the throughwallthickness should be recorded at20 percent DAC. Characterizationshould be in accordance with thedemonstrated'methods under paragraph 6.0.Where the indication is sized at 20percent DAC, this size may becorrected by subtracting the beamwidth in the through-thicknessdirection obtained from the calibra-tion hole (between 20 percent DACpoints) which is at a depth similarto the flaw depth. If the indica-tion exceeds 50 percent DAC, thelength should be recorded bymeasuring the distance between50 percent DAC levels. Thedetermined size should be thelarger of the two.

6.3 Indications without ChangingMetal Path

a. Indications which do not changemetal path distance when scanned inaccordance with the requirements ofASME Section XI and are within theouter 75 percent of the through-walldimension should be recorded when anycontinuous dimension exceeds one inch.

1.150-14

b. If the indication falls withinthe inner 25 percent of the through-walldimensions, it should be recorded at20 percent DAC and evaluated at 50percent DAC.

c. Precautionary note: Indica-tions lying parallel to welds mayappear as nontraveling (without changingmetal path) when scanned by parallelmoving transducers whose beams are aimednormal to the weld, ie. at 900. Multiplescans, however, may reveal that theseindications are traveling indications.If so, recording and sizing are to bedone in accordance with paragraph 6.2.

6.4 Additional Recording Criteria

The following informationshould also be recorded for indica-tions that are reportable accordingto this regulatory position:

a. Indications should be recordedat scan intervals no greater thanone-fourth inch.

b. The recorded information should-include the indication travel (metalpath distance) and the transducerposition for 20 percent, (whereapplicable), 50 percent, and 100 per-cent DAC and the maximum amplitude ofthe signal.

c. When multi-channel equipment isused in the examination system suchthat all examination displays arenot available for simultaneousviewing, an electronic gating systemshould be used which will provideon-line reproducible, recordedinformation, regarding metal path,amplitude and position of all indi-cations exceeding a preset level.The preset level should be the mini-mum recording level required. Toensure that all recordable indicationsare recorded, a preferred method wouldincorporate multi-gates in each channelor a single gate for each channel withmulti-indication recording capability.

7. REPORTING OF RESULTS

Records obtained while followingthe recommendations of regulatorypositions 1.2, 3, and 6, alongwith discussions and explanations,if any, should be kept available atthe site. If the size of an indication,as determined in regulatory positions6.2 or 6.3, exceeds the allowablelimits of Section XI of theASME Code, the indicationsshould be reported as abnormaldegradation of reactor pressureboundary in accordance with therecommendation of regulatoryposition 2.a(3) of RegulatoryGuide 1.16.

Along with the report of ultra-sonic examination test results, thefollowing information should alsobe included:

a. The best estimate of thetolerances in sizing the flawsat the sensitivity required inSection 6 and the basis for thisestimate.

This estimate may be determined inpart by the use of additional reflectorsin the basic calibration block.

b. A description of the techniqueused to qualify the effectiveness ofthe examination procedure,including, as a minimum, material,section thickness and reflectors.

c. The best estimate of the por-tion of the volume required to beexamined by the ASME Code that hasnot been effectively examined suchas volumes of material near eachsurface because of near-field orother effects, volumes near inter-faces between cladding and parentmetal, volumes shadowed by laminarmaterial defects, volumes shadowedby part geometry, volumes inac-cessible to the transducer, volumes

1.150-15

affected by electronic gating, andvolumes near the surface oppositethe transducer.'*

'Olt should be noted that the licen-see is required to apply for relieffrom impractical ASME Code require-ments according to §50.55a of 10 CFR.If the licensee is committed to exa-mine a weld as per the inspectionplan in the plant SAR, the licenseeis required to file an amendmentwhen the commitments made in the SARcannot be met.

Sketches and/or descriptionsof the tools, fixtures and componentgeometry which contribute toincomplete coverage should beincluded.

d. Provide sketches of equipment(i.e., scanning mechanism and transducerholders) with reference points andnecessary dimensions to allow a reviewerto follow the equipment's indicationlocation scheme.

e. When other volumetric tech-niques are used, a description ofthe techniques used should beincluded in the report.

K.

1.150-16.

VALUE/IMPACT STATEMENT

1. PROPOSED ACTION

1.1 Description

The present inservice examination procedures forultrasonic examination require improvement in order toconsistently and reliably characterize flaws in reactorpressure vessel (RPV) welds and RPV nozzle welds. Theapparent low level of the reproducibility of detection,location, and characterization of flaws leads to lengthydiscussions and delays in the licensing process. Muchattention is paid to the integrity of RPV welds during thelicensing process because the failure probability of a reactorpressure vessel is considered to be sufficiently low toexclude it from consideration as a design basis accident.The assumption of a low probability relies heavily onregularly repeated inservice examination by ultrasonictesting (UT) of welds.

1.2 Need for Proposed Action

As more reactors start producing power, as those inoperation grow older, and as more inservice examinationsare performed, the number of detected flaws with uncertaincharacterist ics (size, orientation, and location) is likely toincrease. Flaw characterization is essential for flaw evalua-tions required by the ASME Code and by NRC to determinethe structural integrity of nuclear reactor components whensuch flaws exist. It is essential to have valid backgrounddata for the flaw evaluations required by Section XI of theASME Code. Based on the information gathered accordingto ASME Code requirements, it is often difficult to assesswhether or not the flaw has grown between examinations.The procedures now in use do not require the recording ofcertain information that can be important in later analysisfor determining the location, dimensions, orientation, andgrowth rate of flaws.

The lack of standardization in the use of UT equipmentand procedures leads to uncertainty concerning the resultsobtained. For example, transducer characteristics such asbeam spread, damping characteristics, and frequencyfor peak response are not defined, and there is no provisionto keep track of these from one examination to the other.Similarly, characteristics of other UT system componentssuch as the pulser, receiver, amplifier, and video displayscreen may vary from one examination to another, and allthese characteristics can influence the magnitude of theflaw indications. Therefore, well-defined criteria for supple-mentary UT procedures are needed so that it will be possibleto correctly characterize flaws, estimate flaw growth, andhave reproducible results from inspections performed atdifferent times using different equipment.

In many instances, the rate of flaw growth can be evenmore important than the flaw size. For example, if a flaw isfound in an RPV nozzle or belt-line region and it can be

demonstrated without doubt that the flaw will not growand has not been growing, a rather large flaw can be tolerated.Crack initiation and growth is also a potential problem incases where it is probable that no crack exists, but wherethere is a cluster of small rounded inclusions. These clustersof inclusions should be monitored by UT to ensure absenceof cracks and crack growth.

Where the rate of flaw growth is expected to be large oris uncertain, even a small flaw may be of concern. To

.permit determination of growth rate, the UT proceduresshould be such that results of successive UT examinationscan be compared. With present procedures, these resultscannot be compared because of variation in instrumentsystem characteristics. UT instrument system characteristicsdepend on the characteristics of the system's differentcomponents. Variation in the characteristics of calibrationblocks can also affect results.

Guidelines are needed so that uncertainties in flaw charac-terization may be reduced or eliminated. The safety of thecomponents is evaluated with the help of fracture mechanics.Flaw sizes need to be known for fracture mechanics evalua-tions. Uncertain determination of flaw sizes leads to uncer-tainties in the determination of the safety of the components.Uncertainties in component safety lead to delays in licensing.There is a need to specify and standardize the performancerequired of most UT system components to achieve betterconsistency in UT results so that delays in the licensingprocess may be reduced.

This guide will provide supplementary procedures withthe objective of improving conventional UT procedures, asdefrmed in the ASME Code. This guide is based partly on theinformation available in literature concerning both U.S. andEuropean procedures and partly on the judgment of theNRC staff and their consultants. On the basis of supportwork being performed at the Oak Ridge National Laboratory,the staff plans to issue a revision to this guide that shouldfurther improve flaw characterization.

The use of new techniques such as holography or syntheticaperture imaging of flaws by UT that have not been imple-mented into practice and could considerably increase thecost of inservice examination is not being proposed here.

1.3 Value/Impact of Proposed Action

1.3.1 NRC

Reporting UT examination results as indicated in this guidewould help the NRC staff and their consultants to betterassess the result$ of the data. At present, the NRC staffmust spend a great deal of time on controversy over deter-mining the safety of components from inconsistent UTresults. Lack of faith in flaw size determination fromuncertain UT results points toward the adoption of some

1.150-17

conservative safety measures that are undesirable, for themost part, to the industry managers. Licensing delays occurbecause decisions have to be made on the basis of uncertaininformation. Flaw size determination from consistent UTresults would help remove or reduce the uncertainties anddebates over the safety issues. Because of the above, NRCstaff time for review of reported data and interpretation ofindications is likely to be reduced.

1.3.2 Other Government Agencies

Not applicable, unless the government agency is anapplicant, such as TVA.

1.3.3 Industry

The value/impact on industry of the regulatory guidepositions is stated by each position in the appendix to thisvalue/impact statement. Some highlights of the value andimpact of the regulatory guide positions are stated below.

1.3.3.1 Value. This regulatory guide specifies supplemen-tary procedures that will lead to the following advantages:

a. Attaining greater accuracy and consistency in flawcharacterization.

b. Providing information for consistent flaw characteriza-tion at NRC review time and thus reducing NRC staffeffort in review of flaw indications.

c. Helping assess flaw growth.

d. Providing a more reliable basis for flaw detection andevaluation, which should help in the uniform enforce-ment of rules and the avoidance of delay in licensingdecisions.

e. Reducing licensing time for reviewing examinationresults, which will aid in the reduction of reactor down-time during examinations and will be of great benefitto industry. With present construction costs of about1.3 billion dollars for a 1000-megawatt reactor and theaverage size of a reactor running around 1100-megawattcapacity, the savings per day by eliminating reactordowntime are likely to be $500,000 or more.

f. Avoiding unnecessary repairs due to flaw size uncer-tainties.

g. Reducing radiation exposure to personnel by helpingto eliminate unnecessary repairs. The radiationexposure during repairs is usually many times theexposure during examination, so a net reduction inradiation exposure is expected.

hi. Reducing margins of error in estimates of flaw growthand thus helping reduce overconservative estimatesand decisions on flaw acceptance.

i. Providing more consistent UT procedures for flawcharacterization, thereby leading to procedures thatensure lower probability of missing large flaws andensuring greater safety for the public, industrialworkers, and government employees. L

1.3.3.2 Impact. There will be major impact in thefollowing three areas:

a. Quality control of the UT equipment

At present, requirements in the ASME Code for qualitycontrol of UT equipment are marginal; for example,there are no direct requirements to control the qualityof UT transducers. Criterion XII, "Control of Measuringand Test Equipment," of Appendix B, "Quality Assur-ance Criteria for Nuclear Power Plants and Fuel Repro-cessing Plants," to 10 CFR Part 50 requires, in part, thatmeasures be established to ensure that instruments usedin activities affecting quality are properly controlled,calibrated, and adjusted at specified periods to maintainaccuracy within necessary limits. The recommendationsof this guide will help to bring about uniformity in thequality control procedures among different companiesand will ensure that quality control measures arc takento ensure reliability and reproducibility of UT results.No new UT equipment will be needed to follow therecommendations of this guide. However, the qualitycontrol measures recommended for UT equipmentwill impose extra cost burdens that are difficult toestimate without feedback from industry.

b. Increase in examination time

This guide would recommend, for the first time, thatindications with significant length of indication travel(larger than the standard calibration holes) or withsignificant depth dimensions be recorded. It is notexpected that the slag type of flaws, which are commonamong welds, or geometric reflectors will give signif-icant traveling indications within the guidelines pro-'posed. Hence, no substantial increase in recordedindications as a result of this recommendation isexpected; however, the exact increase is difficult topredict or estimate.

Reporting of indications associated with flaws largerthan 1 inch (indications larger than 1 inch plus beamspread at 20 percent DAC level) is also new. RPV weldsare examined by radiography, and no flaws larger thanthree-quarters of an inch are acceptable in these welds.Because of this acceptance length, only new service-induced flaws larger than 1 inch, of which there shouldnot be many, are expected to be identified and reportedas a result of this recommendation.

Because of the above two new reporting recommenda-tions, there may be an increase in examination timeand dollar cost that is difficult to estimate. This willdepend on how many significant flaws are detectedand how large and complex they are.

1.150-18

c. Radiation exposure 2.3 Comparison of Technical Alternatives

Recommendations of this guide apply to the examina-tion of RPV welds and RPV nozzle welds. RPV weldsare usually examined by automated equipment, anddata are collected on tape. Therefore, no increase inradiation exposure is anticipated as a result of theregulatory guide positions addressing RPV weldexaminations.

RPV nozzle welds are sometimes examined byautomated equipment but in most cases by manualUT. An increase in radiation exposure to examinationpersonnel may be expected while RPV nozzles arebeing manually examined. The probable percentincrease in examination time or radiation exposure isimpossible to estimate without field data and researcheffort. Requirements for reporting traveling indica-tions and indications associated with flaws larger thanI inch may lead to an increase in occupationalexposure in those cases in which the above indicationsare found and additional examination is required. Themagnitude of this additional exposure can only beassessed on a case-by-case basis. It should be notedthat radiation levels at vessel nozzle regions arereported to range from 0.5 to 2.0 rem/hour. Totalperson-rem doses can be drastically reduced byshielding and local decontamination.

The guide is not expected to have any adverse impact onother government agencies or the public.

Imposing inservice examination of RPV welds by the useof holography, synthetic aperture imaging technique, oracoustic emission, all of which are still in the stage of proto-type development and have not been proved effective forfield use, would not be justifiable on the basis of eithercost or effectiveness.

2.4 Comparison of Procedural Alternatives

.Leaving the situation as it is would mean that continuedattention and manpower would have to be devoted by theNRC staff to investigate the uncertainties associated withflaw growth on a case-by-case basis. The low level ofconfidence in the present techniques means that excessivemargins would continue to be used in the flaw-acceptancecriteria. Also, unnecessary cutting and repair attempts toremove suspected flaws may result.

The procedures recommended in this guide have beenshown tobe effective in practice, although they. are not ingeneral use in the United States. Including these proceduresas regulatory guide recommendations should result in theirwider use and consequently their improvement. After theseprocedures have been accepted by the industry, we willseek their inclusion in the ASME Code. Some of theseprocedures have already been sent to the ASME for considera-tion and inclusion in the present ASME Code proceduresfor ultrasonic examinations.

2.5 Decision on Technical and Procedural Alternatives

On the basis of.the above, it appears desirable to issue aregulatory guide to provide recommendations for improvingASME Code procedures. These recommendations, whichare based on the advanced state-of-the-art UT procedures incurrent use by some organizations, would improve theability to detect and characterize flaws without imposingnew, unproved techniques for flaw detection on industry.

3. STATUTORY CONSIDERATIONS

3.1 NRC Authority

The authority for this guide is derived from the safetyrequirements of the Atomic Energy Act of 1954, as amended,and the Energy Reorganization Act of 1974, as implementedby the Commission's regulations. In particular, § 50.55a,"Codes and Standards," of 10 CFR Part 50 requires, inpart, that structures, systems, and components be designed,fabricated, erected, constructed, tested, and inspected toquality standards commensurate with the importance ofthe safety function to be performed.

3.2 Need for NEPA Assessment

The proposed action is not a major action, as defined byparagraph 51.5(a)(10) of 10 CFR and does not require anenvironmental impact statement.

1.3.4 Public

No impact on the public can be foreseen. The onlyidentifiable value is a slight acceleration in the reviewprocess.

1.4 Decision on Proposed Action

The Office of Nuclear Reactor Regulation (NRR) hasstated the need for this guide to help them and theirconsultants in evaluating the size and significance of theflaws detected during inservice examination to ensure theintegrity of reactor pressure vessels between periods ofexamination. It would therefore be advisable to issue thisguide.

2. APPROACH

2.1 Technical Alternatives

Alternatives would include requiring the use of holography,synthetic aperture imaging, acoustic emission, neutronradiography, or a combination of the above during RPVinservice examination.

2.2 Procedural Alternatives

One alternative is to leave the situation as it is. A secondalternative, is to request change of the ASME Code require-ments.

1.150-19

4. RELATIONSHIP TO OTHER EXISTING OR PRO-POSED REGULATIONS OR POLICIES

Recommendations of this guide would be supplementalto the requirement*s of Section XI, "Rules for InserviceInspection of Nuclear Power Plant Components," of theASME Code, which is adopted by § 50.55a, "Codes andStandards," of 10 CFR Part 50.

5. SUMMARY

This guide was initiated as a result of a request fromNRR. Preliminary results of the round robin UT examinationprocedures following ASME Code procedures indicate aneed for additional guidelines to the existing ASME Codeprocedures to control equipment performance, calibrationblock specifications, and scanning procedures to improve thereproducibility of results and detectability of through-thick-ness flaws.

Minimum ASME Code requirements do not specify thedetails of recording requirements that are essential toevaluate flaws. This deficiency in the Code rules makes it

difficult for the NRC staff or their consultants to review,analyze, and assess the UT data to determine the flaw sizeand evaluate the system safety when the data are madeavailable to NRC at a later date. The present data obtainedfrom UT equipment of uncertain and unspecified performancelead to discussions and delays in the review process resultingin loss of NRC staff time and loss of plant availabilityand power generation capacity for the utilities. Thesesituations definitely need to be avoided as much as possible.This guide is aimed at achieving this purpose by issuingrecommendations that will be supplementary to the existingASME Code UT procedures. The issue remains whether towait for the development of advanced NDE techniques andcontinue with the present ASME Code procedures resultingin uncertainties, delays, and discussions or to encourageimprovement in the present state of the art of conventionalUT. The decision appears to be obvious that we should useconventional UT based on engineering judgment until somenew techniques for flaw detection and sizing can be provedeffective in the field. This guide is aimed at providing therecommendations needed to improve on the ASME CodeUT requirements until proven advanced NDE techniquesare available.

I1

I ý

1.150-20

APPENDIX TO VALUE/IMPACT STATEMENT

Values that will result from this regulatory guide aremuch easier to perceive than the impact. It is very difficultto assess the real impact because the kind of statistical dataneeded is simply not available at this time. One way in whichwe hope to estimate the impact is through industry feed-back after the guide has been issued.

We have made an attempt, in a qualitative manner, toestimate the value/impact of regulatory guide positions,position by position, as follows:

1. INSTRUMENT PERFORMANCE CHECKS

Recording the characteristics of the ultrasonic testing(UT) examination system will be useful in later analysis fordetermining the location, dimensions, orientation, andgrowth rate of flaws. System performance checks to deter-mine the characteristics of the UT system will be madeshortly before the UT examinations. Each UT examinationwill therefore be correlated with a particular system per-formance check. This practice will help to compare results.These determinations will help make it possible to judgewhether differences in observations made at different timesare due to changes in instrument characteristics or are dueto real changes in the flaw size and characteristics.

It is recommended that, as a minimum, instrumentchecks should be verified before and after examining all thewelds that need to be examined in a reactor pressure vesselduring one outage.

Performance of these instrument checks is likely to adda few thousand dollars to test equipment cost and to take Ito 2 hours of examination time before and after each reactorpressure vessel (RPV) examination. The examination equip-ment is usually idle between examinations. Performancechecks on the examination equipment could be performedduring these idle periods. These performance checks are notlikely to reduce the number of examinations that a particularUT system could perform in a year. No additional radiationexposure is expected because of this position.

2. CALIBRATION

According to this position, system calibration should bechecked to verify the distance-amplitude correction (DAC)curve, as a minimum, before and after each RPV examina-tion (or each week the system is in use, whichever is less) oreach time any component (e.g., transducer, cable, connector,pulser, or receiver) in the examination system is changed.

Subarticle 1-4230, Appendix I, Section XI, ASME B&PVCode (1974 edition), which applied to the inspection of theRPV, required calibration using the basic calibration blockat "the start and finish of each examination, with any changein examination personnel and at least every 4 hours duringan examination." However, the 1977 rules of Article 4(T-433), Section V, which are referenced by Section XI and

now apply to the examination of the RPV, require calibra-tion against the calibration block only "prior to use of thesystem." It is considered that the present 1977 ASME Coderules are not adequate to control potential problems in thevariation of instrument performance characteristics. There-fore, the recommended calibration before and after eachexamination is a more reliable approach to instrumentperformance checks. The above position is not more con-servative than the previously accepted 1974 Code rules, but ismore conservative if 1977 rules are considered.

Considering the requirements of Article 4, Section V(1977), the above position will mean a calibration checkeach week the system is in use or before and after eachRPV examination, whichever is less, instead of before eachexamination. A calibration check against the calibrationblock takes 15 to 30 minutes for manual UT and forautomated UT equipment where provision is made tocalibrate the equipment without having to remove the trans-ducers from the rotating scanning arm of the'mechanizedscanner. In some cases, transducers have to be removedfrom the scanning arm for calibration of the UT instrument;in these cases, a calibration check may take from 30 to 60minutes. The added cost of the above would be in terms ofadditional time spent by the examiner and would occureach week or once for each RPV examination, dependingon whether or not the examination is completed in lessthan a week. No additional radiation exposure is expectedbecause of this position.

3. NEAR-SURFACE EXAMINATION AND SURFACERESOLUTION

This position recommends that an estimation of thecapability to effectively detect defects at the metal frontand back surfaces of the actual component should be madeand reported. This will not require any additional calibrationor examination time but will simply require an estimate ofthis capability by the examiner, which will be reported toNRC. No additional radiation exposure is expected becauseof this position.

4. BEAM PROFILE

This position recommends that the beam profile (foreach search unit used) should be determined if any signif-icant flaws are detected during the RPV examination.

Assuming that no more than three search units are likelyto be used during an RPV examination, this step is likely torequire no more than 2 hours of examination time. Noadditional radiation exposure is expected because ofthis position.

5. SCANNING WELD-METAL INTERFACE

This position recommends that the beam angles used toscan welds should be based on weld/parent-metal interface

1.150-21

geometry and at least one of these angles should be such thatthe beam is almost perpendicular (±1 5 degrees to the perpen-dicular) to the weld/parent-metal interface, unless it can bedemonstrated that large (Code-unacceptable) planar flawsunfavorably oriented can be detected by the UT technique.

On the basis of information available, it appears that it isdifficult 1 ,2, 3 to detect large planar flaws (e.g., service-inducedfatigue or stress corrosion cracks) oriented at right angles tothe surface, using the ASME Code UT procedure. However,the option is being provided to demonstrate that such flawscan be located by conventional methods or by using newadvances in UT techniques. In these cases, the technique willbe acceptable as a volumetric examination method. Otherwise,the use of high-intensity radiography or tandem-probe UTtechnique, among other techniques, should be considered.

The above type of flaw is the most significant but themost difficult to detect. Because of this, the present recom-mendations are being made despite their potential impacton cost and radiation exposure.

The potential impact may be as follows:

a. Additional NRC staff time may be needed to evaluatethe effectiveness of UT techniques on a generic basis todetect perpendicular planar flaws. After techniques arerecognized to accomplish the above, NRC staff time that isbeing spent currently on evaluating problems on a plant-by-plant basis is expected to be considerably reduced.

b. Reactor downtime may increase, depending on theexamination time differentials between the conventionaland refined techniques. This may, however, be offset by areduction in the downtime currently needed for NRCexperts to evaluate data that sometimes requires furtherclarification and reexamination.

2 '4

c. Additional cost might be incurred in changes neededto add transducers or data-gathering capability to existingautomated equipment or to automate current manualexaminations. Automation of current manual techniques islikely to reduce radiation exposure to personnel.

6. SIZING AND RECORDING OF INDICATIONS

6.1 Traveling Indications

This position recommends the recording of travelingindications. If RPV welds do not have any travel indications

I"Probability of Detecting Planar Defects in Heavy Wall Welds byUltrasonic Techniques According to Existing Codes," Dr. Ing. Hans-Jurgen Meyer, Quality Department of M.A.N., Nurnberg, D 8500Nurnberg 115.

2"Interim Technical Report on BWR Feedwater and Control RodDrive Return Line Nozzle Cracking,"NUREG-0312,July 1977, p. 3.

3 "Analysts of the Ultrasonic Examinations of PVRC Weld Speci-mens 155, 202, and 203," R.A. Buchanan, Pressure vessel ResearchCommittee (PVRC) Report, August 1976.

4,Summary of the Detection and Evaluation of Ultrasonic Indica-tions - Edwin Hatch Unit 1 Reactor Pressure Vessel," January 1972,Georgia Power Company.

1.

on the screen larger than the indication on the screen fromthe calibration holes (1/2-inch hole for a 12-inch weldthickness, 3/8-inch hole for an 8-inch thickness), thisrecommendation will not result in any more recording ofindications. If the RPV welds being examined have severalindications with travel in excess of the calibration holediameter, the examination and recording time will beincreased for the investigation of these flaws, dependingon the number of these indications. Slag inclusions in weldsare generally long cylindrical defects and do not have muchdepth unless they are associated with shrinkage or service-induced cracks. These slag inclusions are not expected toincrease the number of indications that will be recorded.Increase in examination time will depend on the number,size, and complexity of geometry of through-thicknessindications.

For RPV girth or nozzle welds where examination isperformed by automated equipment and data are recordedon tape, this position will mean no increase in examinationtime or radiation exposure; but interpretation, analysis, andreporting time for these depth indications will increase. Theextra burden in terms of dollar cost will depend on thenumber, size, and complexity of flaws, and there are norational bases or data available at this time to estimate theincrease in the cost of examination.

For RPV welds, mostly nozzle welds, where examinationis performed manually and data are not recorded on tape,this position will mean extra examination time and increasedradiation exposure to the examiners. Increase in dollar costand radiation exposure will again depend on the number,size, and complexity of indications, and there are no basesor data available to estimate this increase.

6.2 Nontraveling Indications

This position also recommends the recording of nontravel-ing indications above 20 percent DAC level that persist fora distance of more than I inch plus the beam spread.According to NB-5320, Radiographic Acceptance Standards,Section I1, Division 1, ASME Code, 1977 edition, flaws,larger than 3/4 inch for weld thicknesses above 2-1/4 inchesare not acceptable. Because of this requirement, it isexpected that no flaws larger than 3/4 inch in length arepresent in the RPV welds, and if indications are detectedthat suggest flaws larger than 3/4 inch, there is a strongpossibility that these may be service-induced flaws. Service-induced flaws are rare in RPV v,'lds, and it is thereforenot expected that additional indi ,,ns would have to berecorded because of this position. However, if such indica-tions (over 1 inch) are detected, examination time forautomated recording and examination time plus radiationexposure for manual UT examinations will be increased.There are no rational bases or data available to estimate theimpact of regulatory position 6.2.

7. REPORTING OF RESULTS

This position recommends that the areas required to beexamined by the ASME Code that have not been effectivelyexamined and an estimate of error band in sizing the flaws

.150-22

L

I I

should be brought to the attention of the NRC when theresults are reported. This effort may take about S hours inreportwriting time.

8. IMPLEMENTATION.

It should be noted that the recommendations of this guideare not intended to apply to those preservice examinationtests already completed. However, the licensees mayconsider repeating their preservice examination tests orusing the recommendations of this guide any time at theiroption to avoid possible flaw interpretation problems at alater date. Flaw interpretation problems may occur iftraveling indications identified as significant according tothe recommendations of this guide do not correlate withpreservice volumetric NDE results and hence would beassumed to have been service induced. It would be difficultto show that these indications arose from fabrication flaws.Therefore, licensees would be well advised to consider theabove possibilities.

8. I Alternatives

The following alternatives were considered in applyingthe recommendations of this guide.

1. To apply the recommendations of the guide to all thepreservice and inservice examinations that havealready been performed.

2. To apply the recommendations of the guide to allfuture preservice and inservice examinations per-formed after the issuance of the guide.

8.2 Discussion of Alternatives

8.2.1 First Alternative

Alternative 1 would infer that all RPV welds examinedas per the current code requirements are at a quality levelthat would not ensure an acceptable safety performance.This approach would also mean that all the plants wouldhave to repeat, in accordance with the recommendations

of this guide, those inservice and preservice examinationsperformed in the past. Such a policy would tend to beoverly conservative and would put a heavy burden on allplant owners. Although UT examinations have missed someflaws in the past, there appears to be no immediate dangerfrom the estimated flaw distribution probability to warrantsuch a strong action. Therefore, this alternative was notadopted.

8.2.2 Second Alternative

In the past, several instances have been noted where theminimal Code UT examination procedures have not beenadequate for detecting and sizing flaws. Discussions andundesirable licensing delays were frequently the result. Asmore plants begin producing power and existing plants growolder, more flaws may be expected in the weld areas. Theseflaws may be generated as a result of fatigue, stress corrosion,or other unanticipated factors. It is imperative that theguide recommendations for supplementary UT examinationprocedures be used in the future to maintain an acceptablelevel of safety at these welds. The second alternative wastherefore selected for applying this guide to the preserviceand inservice examination of RPV welds.

It is expected that inservice UT examinations will detectflaws generated during plant operation, whereas preserviceexaminations will provide UT examination data for sub-sequent comparisons. First, a radiographic examination isperformed of all the vessel welds under Section III of theASME Code. After this examination, a UT preservice exam-ination of welds is performed to serve as a supplementaryvolumetric examination. Because of the above, these pre-service examinations are not as important as inservice exam-inations. It was therefore decided that the guide recommenda-tions should apply to judging the inservice examination resultsfor those examinations performed immediately after theissuance of the guide; however, the guide recommendationsshould apply to preservice examinations beginning 6 monthsafter the issuance date. The NRC staff considered thisapproach best because of the difficulties being experiencedin reviewing inservice UT examination data from thedifferent plants.

1.150-23