NUREG-0313, Draft Commission Paper Requesting Approval for Implementation of Long Range Plans for Dealing W/Igscc in BWR Piping,By Issuing Generic Ltr & Technical Basis document,NUREG-0313,Rev 2

Draft Commission Paper Requesting Approval for Implementation of Long Range Plans for Dealing W/Igscc in BWR Piping,By Issuing Generic Ltr & Technical Basis document,NUREG-0313,Rev 2

ML20246F082
Person / Time
Issue date:	03/31/1987
From:	Stello V NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO)
To:
Shared Package
ML20244D762	List: ML20244D758 ML20245A495 ML20245A584 ML20245B365 ML20245B368 ML20246F056 NUREG-0313, Draft Commission Paper Requesting Approval for Implementation of Long Range Plans for Dealing W/Igscc in BWR Piping,By Issuing Generic Ltr & Technical Basis document,NUREG-0313,Rev 2
References
RTR-NUREG-0313, RTR-NUREG-313 NUDOCS 8905120125
Download: ML20246F082 (36)

Text

F-j]

p ..

DRAFT 03/31/87 WHazelton

< a

-FOR: The Commissioners

.FROM: Victor Stello, Jr.

Executive Director for Operations

SUBJECT:

STAFF LCMG RANGE PLAN FOR DEALING WITH STRESS i CORROSION CRACKING IN BWR PIPING  ;

i PURPOSE: To request Commission approval for the implementation of long range plans for dealin corrosion cracking (IGSCC) g with intergranular stressin BWR piping, Generic Letter and the technical basis document, NUREG-0313, Revision 2.

BACKGROUND: In SECY-84-301, dated July 30, 1984, the. staff presented a long range plan to deal with the intergranular stress corrosion cracking (IGSCC) of piping at BWR plants. 0n September.11, 1984, the Commission was briefed by.the-staff on the Piping Review Committee's recommendations.

Subsequent to the briefing, the Commission directed.the staff ta develop an implementation plan for its approval.

A principal task in the staff long range plan was the development of staff technical positians. NUREG-0313, Revision 2, " Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary Piping" was prepared, incorporating the recommendations of the Piping Review Committee as delineated in NUREG-1061, Vol. 1, and public comments to that document. Staff technical positions were prepared, using NUREG-0313, Revision 2 as a basis. These staff positions were included in a draft Generic Letter.

Draft versions of NUREG-0313, Revision 2 and the Generic Letter were issued for public comment on July 21, 1986, after review by the ACRS and the CRGR. Formal comments were received  ;

from 12 utilities licensed to operate BWRs, Generic Electric Company, several consulting firms, and the BWR Owners Group for IGSCC Research. The responses were generally favorable, although there were several technical and licensing suggestions that have required staff review and resolution . This was facilitated by a peer review that included experts from several national laboratories, including Brookhaven National Laboratory, Argonne <

National Laboratory, Idaho National Engineering Laboratory, and Pacific Northwest Laboratories.

The Generic Letter and the technical basis document, NVREG-0313, Revision 2, have been prepared in final form. They i

I 8905120125 870611 l PDR REVGPNRGCg NEETING114 c

k n

f u .z

. define methods acceptable to the staff for controlling the

.intergranular. stress corrosion susceptibility of BWR piping.

Licensees and construction permit holders are requested to respond pursuant to 10 CFR 50.54(p) with their. plans to deal with IGSCC. .

For piping that does not fully comply with the. material selec-tion and processing' guidelines, varying degrees of augmented inservice inspection are recommended. These documents also include guidance and staff positions regarding cracking evaluation and weld repair methods for long term operation or for' continuing interim operation of plants.until more permanent solutions are implemented. Thus, these documents provide a technical basis for the control of stress corrosion cracking in BWR piping, and implementation of these recommendations by

. -the licensees will complete the resolution of this Generic. Issue.

SUMMARY

.The' staff has. prepared a Generic Letter to obtain information regarding. licensee's long range plans to deal with stress corrosion cracking of BWR piping. This Generic Letter, and NUREG-0313,. Revision 2,ihich provides the technical basis, were issued in draft for public comment. Resolution of the public comments has been completed, and appropriate revisions to the tw; documents have been finalized and reviewed by the-ACRS and the CRGR.

The Commission's approval to issue the Generic Letter and the technical basis NUREG is requested.

I Victor Stello, Jr.

Executive Director for Operations L

Enclosures:

i 1. Proposed Generic Letter

2. NUREG-0313, Revision ?

3. Staff Resolution of Public Comments

_-__-____:______1

u o UNITED STATES

g

', 'L y g NUCLEAR REGULATORY COMMISSION

'4- ,," ;y WASHINGTON D. C. 20555

\, v...../

TO ALL LICENSEES OF OPERATING BOILING WATER REACTORS (BWRS), AND HOLDERS OF CONSTRUCTION PERMITS FOR BWRS Gentlemen:

SUBJECT:

NRC POSITION ON IGSCC IN BWR AUSTENITIC STAINLESS STEEL PIPING (Generic Letter 87- )

Intergranular stress corrosion cracking (IGSCC) near weldments in BWR piping has been occurring for almost 20 years. Early cases were in relatively small-diameter piping. In early 1982, cracking was identified in large-diameter piping in a recirculation system of an operating BWR plant in this country. Since then, extensive inspection programs have been conducted on BWR piping systems. 'These inspections have resulted in the detection of signi-ficant numbers of cracked weldments in almost all operating BWRs.

A number of domestic and foreign BWR owners have replaced or plan to replace piping systems that have experienced IGSCC with more resistant material.

Other owners are implementing countermeasures such as Stress Improvement (SI) or Hydrogen Water Chemistry (HWC) to reduce the susceptibility of the piping to IGSCC. In many cases, cracked weldments have been repaired by reinforcing them with weld overlay.

Substantial efforts in research and development have been sponsored by the BWR Owners Group for IGSCC Research. The results of this program, along with other related work by vendors, consulting firms, and confirmatory research sponsored by the NRC, have permitted the development of revised Staff Positions regarding the IGSCC problems.

The technical bases for these positions are detailed in NUREG-0313, Rev. 2

" Technical Report on Material Selection and Process Guidelines for BWR Coolant Pressure Boundary Piping." This revision to NUREG-0313 was a major task in the staff long range plan to deal with BWR pipe cracking that was presented to the Commission in SECY e4-301. This revision includes the relevant recommenda-tions of the Piping Review Committee Task Group on Pipe Cracking issued as NUREG-1061, Vol. 1, " Report of USNRC Piping Review Committee", and consideration of public comments on that document. NUREG-0313, Rev. 2 describes the technical bases for the staff positions on materials, processes, and primary coolant chemistry to minimize and control IGSCC problems. Inspection schedules and inspection sample sizes are based on the susceptibility of weldments to initiation and propagation of IGSCC. Inspection schedules are comparable to those specified in Section XI of the ASME Boiler and Pressure Vessel Code in

. cases where the piping material is IGSCC resistant. Varying amounts of augmented inspections are specified for piping with a greater susceptibility to cracking, where there is less certainty about the effectiveness of mitigation measures used, or in cases where repairs have been performed. When improved i water chemistry control with hydrogen additions is implemented, less augmentation I of inspection schedu as is required.

The purpose of this Generic Letter is to seek information regarding implementation of the new staff positions covering these technical areas. This Generic Letter supersedes Generic Letter 84-11 " Inspection of BWR btainless Steel Piping."

g This Generic Letter applies to all BWR piping made of austenitic stainless steel that is four inches or larger in nominal diameter and contains reactor coolant at a temperature above 200 F during power operation regardless of Code classification. It also applies to reactor vessel attachments and appurten-ances such as jet pump instrumentation penetration assemblies and head spray and vent components.

This' Generic Letter does not apply to piping made of carbon steel classified as P-1 by the ASME Boiler and Pressure Vessel Code.

Staff Positions have been developed covering the following subjects:

1. Staff. Position on Materials

2. Staff Position on Processes

3. Staff Position on Water Chemistry

4. Staff Position on Weld Overlay Reinforcement

5. Staff Position on Partial Replacement

6. Staff Position on Stress Improvement of Cracked Weldments

7. Staff Position on Clamping Devices

8. Staff Position on Crack Characterization and Repair Criteria

9. Staff Position on Inspection Methods and Personnel

10. Staff Position on Inspection Schedules

11. Staff Position on Sample Expansion

12. Staff Position on Leak Detection These Staff Pesitions are fully delineated in Attachment A to this lettee.

The staff continues to believe that replacing susceptible piping with IGSCC-resistant materials will provide the greatest degree of assurance against future cracking problems. Licensees may follow Generic Letter 84-07,

" Procedural Guidance for Pipe Replacements at BWRs"; the staff encourages programs to replace degraded piping so as to reduce the potential for cracking and to minimize the need for augmented inspections. However, the staff recognizes that, if the staff positions of this Generic Letter are implemented, adequate levels of piping integrity and reliability can be achieved. The staff I believes this Generic Letter, together with the revision to NUREG-0313, will be of use to licensees in making sound decisions regarding IGSCC. Each weldment can be evaluated considering its material, heat treatment history, stress level, chemical environment and surveillance program. This will provide a basis for a reasonable judgment regarding the long-term acceptability of that weldment. Considering that each piping system has many weldments and each plant has many piping systems, the entire problem must be evaluated in an integrated way.

The Commission has determined that, unless appropriate remedial actions are  ;

taken, BWR plants may not be in conformance with their current design and l licensing bases, including 10 CFR 50, Appendix A, General Design Criteria 4, j 14, and 31.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _J

s Accordingly, pursuant to 10 CFR 50.54(f), you, as a BWR operating reactor licensee or construction permit holder, are requested, to furnish, under oath or affirmation, your current plans relating to piping. replacement, inspection, repair, and leakage detection. Your response should indicate whether you intend to follow the staff positions included in this letter, or propose alternative measures. This information is needed for the Commission to determine whether an operating license should be issued, or if you hold an operating license, whether it should be modified or revoked.

An' acceptable response to this letter would include the following specific items:

1. Your current plans regarding pipe replacement and/or other measures taken or to be taken to mitigate IGSCC and provide assurance of continued a

long-term piping integrity and reliability.

2. An Inservice Inspection Program for austenitic stainless steel piping covered under the scope of this letter that conforms to the staff positions on inspection schedules, methods and personnel, and sample expansion included in this letter.

3. A change to the Technical Specifications to include a statement that the Inservice Inspection Program for piping covered by the scope of this letter will be in confccmance with the staff positions on schedule, methods and personnel, and sample expansion included in this letter.

4. Confirmation of your plans to ensure that the Technical Specification related to leakage detection will be in conformance with the staff position on leak detection included in this letter.

I

5. Your plans to notify the NRC of any flaws identified that are in excess of ASME Code Section XI criteria for continued operation without evaluation, or a change found in the condition of the welds previously known to be cracked, and your evaluation of the flaws for continued operation and/or your repair plans.

Licensees and construction permit holders for BWR plants are requested to respond to this generic letter within 180 days of receipt of this letter.

NRC review of your submittal of information in response to this letter is not subject to fees under the provisions of 10 CFR 176. However, should you, as part of your response or in a subsequent submittal, include an application for license amendment or other action requiring NRC approval, it is subject to the fee requirements of 10 CFR 170 with remittance of an application fee of $150 per application (Sections 170.12(c) and 170.21) and subsequent semi-annual payments until the review is completed or the ceiling in Section 170.21 is reached.

1

4 This request for information was approved by the Office of Management and Budget under clearance number which expires Comments on burden and duplication may be directed to the Office of Management and Budget, Reports Management Room 3208, New Executive Office Building Washington, D.C. 20503.

Sincerely, Robert M. Bernero, Director Division of BWR Licensing Office of Nuclear Reactor Regulation l

_ - _ _ _ - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ - _ = _ _ _ _- _ _ _ _ _ _ - . _ _ _ _ _ _ _ _ _ _ - _

1 ATTACHMENT A Scope This Generic Letter applies to all BWR piping made of austenitic stainless steel that is four inches or. larger in nominal diameter and contains reactor.

coolant at a temperature above 200 F during power operation regardless of Code classification. It also applies to reactor vessel attachments and appurten-antes such as jet pump instrumentation penetration assemblies and head spray and vent components.

This Generic Letter does not apply to piping made of carbon steel classified as P-1 by the ASME Boiler and Pressure Vessel Code.

l Staff Position on Materials Materials considered to be resistant to sensitization and IGSCC in BWR piping systems are:

(1) Low carbon wrought austenitic stainless steel, which includes types 304L, 304NG, 316NG and similar low carbon grades with a maximum carbon content of 0.035%. Type 347, as modified for nuclear use, will be resistant with somewhat higher carbon content, the usual maximum of 0.04% is adequate.

These materials must be tested for resistance to sensitization in accord-ance with ASTM A262-A or -El or equivalent standard.

(2) Low carbon weld metal, including types 308L, 316L, 309L and similar grades, with a maximum carbon content of 0.035% and a minimum of 7.5% ,

(or FN) ferrite as deposited. Low carbon weld metal especially developed for joining modified type 347 is also resistant as deposited.

Welds joining resistant material that meet the ASME Boiler and Pressure Vessel Code requirement of 5% (or FN) ferrite, but are below 7.5%

(or FN) may be sufficiently resistant, depending on carbon content and other factors. These will be evaluated on an individual case basis.

(3) Piping weldments are considered resistant to IGSCC if the weld heat affected zone on the inside of the pipe is protected by a cladding of resistant weld metal. This is often referred to as corrosion resistant cladding (CRC).

(4) Cast austenitic stainless steel with a maximum of 0.035% carbon and a minimum of 7.5% (or FN) ferrite. Weldjointsbetweenresistantpiping and cast valve or pump bodies that do not meet these requirements are considered to be special cases, and are covered in the Staff Position on Inspection Schedules below.

(5) Austenitic stainless steel piping that does not meet the requirements of (1) above is considered to be resistant if it is given a solution heat treatment after welding.

l-1

g.  ;(6) Other austenitic materials, including nickel base alloys such as Inconel 600, will be evaluated on an individual case basis. Inconel 82 is the only commonly used nickel base weld metal considered'to be resistant.

It is the staff position that no austenitic material is resistant to cracking in the presence of a crevice, such as formed by a partial penetration weld, where the crevice is exposed to reactor coolant.

Staff Position on Processes The processes considered to provide resistance to IGSCC in BWR piping welds are:

(1) Solution Heat Treatment (SHT)

(2) Heat Sink Welding (HSW)

(3)' Stress-Improvement (SI)

a. Induction Heating Stress Improvement (IHSI)

b. Mechanical Stress Improvement Process (MSIP)

Alti.cugh last pass heat sink welding (LPHSW) is not now considered to be fully effective, specific cases may be evaluated individually.

Staff Position on Water Chemistry The use of_ hydrogen water chemistry, together with stringent controls on conductivity, will inhibit the initiatian and growth of IGSCC. However, the responses of BWRs to hydrogen injection differs from plant to plant, and the development and verification of a generic HWC specification is not yet complete. For these reasons, reductions in piping inspection frequency based on the use of HWC will be considered on an individual case bases at-the present time. If fully effective HWC is maintained, a factor of two in reduction of inspection frequency may be justified for IGSCC Categories B, C, 0, and E weldments. (See Table 1) )

Staff Position on Weld Overlay Reinforcement Cracked weldments that are reinforced with weld overlay are acceptable for short-term operation, and may be considered for longer term operation provided:

(1) The overlayed weldments are in conformance with the criteria of IWB 3600 of Section XI of the 1986 Edition of the ASME Boiler and Pressure Vessel Code, and 2

(2) they are inspected in conformance with the Staff Position on Inspection Methods and Personnel, by UT examiners and procedures qualified to inspect overlayed welds.

o

t Staff. Position on Partial-Replacement If portions of cracked piping are replaced in the course of repair, the replaced portions will be subjected to inservice inspection requirements that-will depend on the materials and processes used. All relevant staff positions of this Generic Letter will apply. ,

Staff Position on Stress Improvement (SI) of Cracked Weldments Stress Improvement is e.lso considered to be an effective mitigation process when applied to weldments with short or shallow cracks. Specifically, welds with cracks that are no longer'thun 10% of the circumference, and are no deeper than 30% of the wall thickness will be considered to be mitigated by SI.

SI is only considered to be effective if it is followed by a qualified UT examination, and if cracks are found, they must be sized both in depth and length, by procedures and personnel qualified to perform sizing evaluations:

Staff Position on Clamping Devices Clamping devices may be used for temporary reinforcement of cracked weldments.

Each case must be reviewed and approved on an individual basis.

Staff Position on Crack Characterization and Repair Criteria Methods and criteria for crack characterization and repair should be in conformance with IWB-3600 of Section XI of the 1986 Edition of ASME Boiler and Pressure Vessel Code.

Because detailed sections of the Code are still under development, methods of analysis and acceptance criteria described in detail in NUREG-0313, Rev. 2 are considered acctptable to the staff. .

Staff Position on Inspection Methods and Personnel Examinations performed under the Scope of this letter should comply with the applicable Edition and Addenda of the ASME Code,Section XI, as specified in paragraph (g),'" Inservice Inspection Requirements" of 10CFR50.55a, Codes and Standards, or as otherwise approved by the NRC.

In addition, the detailed procedure, equipment and examination personnel shall be qualified by a formal program approved by the NRC such as that being conducted in accordance with the NDE Coordination Plan agreed upon by NRC, EPRI, ar.d the Boiling Water Peactor Owners Group for IGSCC Research, being implemented at the EPRI NDE Center in Charlette, North Carolina.

Staff Position on Inspection Schedules A summary of the Staff Position on Inspection Schedules is given in Table 1.

Additional details and definitions are provided below. NUREG-0313, Rev. 2, Section 5 provides background information and technical bases.

(1) Welds of resistant material, IGSCC Category A, should as a minimum be examined according to an extent and frequency comparable to that intended by the applicable provisions of Section XI of the ASME Boiler and Pressure Vessel Code, which is a 25% sample every 10 years. The selection of specific welds to be included in this sample is the responsibility of the Licensee, and should include considerations of stress levels, piping configurations, weld details, etc and should represent his best judgement regarding selection of a representative and meaningful sample.

The provisions of 10CFR50.55a, (b),(2),(ii) may be invoked if it is determined necessary to use the 1974 edition of the Code to permit a meaningful sample selection.

(2) Although castings with higher carbon content than 0.035% are not considered to be resistant to sensitization, welds joining such castings (in the form of pump and valve bodies) to piping have been relatively free of IGSCC. This may be attributed to a favorable residual stress distribution, as calculations have indicated. For this reason, welds joining resistant material to pumps and valves will be considered to be resistant welds, and included in IGSCC Category A. If extensive weld repairs were performed the residual stress may be unfavorable, in which case such welds should be included in Category D.

(3) Welds that have been treated by SI or reinforced by weld overlay that are classified as IGSCC Category F because they do not meet the applicable staff positions may be uporaded to Category E if no adverse change in crack condition is found after 4 successive examinations.

Staff position on Sample Expansion If one or more cracked welds in IGSCC Categories A, B, C, or D are found during an inspection, an additional sample of the welds in that category should be inspected, approximately equal in number to the original sample.

This additional sample should be similar in distrit'ution (according to pipe size, system, and location) to the original sample, unless it is determined that there is a technical reason to select a different distribution. If any cracked welds are found in this sample, all of the welds in that IGSCC Category should be inspected.

If significant crack growth or additional crecks are found during the inspection of an IGSCC Category E weld, all other Category E welds should be examined, a) Significant crack growth for overlayed welds is defined as crack extension to deeper than 75% of the original wall thickness, or for cracks originally deeper than 75% of the pipe wall, evidence of crack growth into the effective weld overlay, b) Significant crack growth for SI mitigated Category E welds is def% ed as crowth to a length or depth exceeding the criteria for SI mitigation.

(either10%ofcircumferenceinlengthor30%ofthewallindepth).

I

_ - - - - - - )

' Staff Position on Leak Detection Leakage detection systems should be in conformance with Position C of Regulatory Guide 1.45 " Reactor Coolant Pressure Boundary Leakage Detection Systems, or as otherwise approved by the NRC.

1. Plant shutdown should be initiated or inspection and corrective action when, within any period of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or less, any leakage detection system indicates an increase in rate of unidentified leakage in excess of 2 gpm or its equivalent, or when the total unidentified leakage attains a rate of 5 gpm or equivalent, whichever occurs first. For sump level monitoring systems with fixed-measurement-interval methods, the level should be monitored at approximately 4-hour intervals or less.

2. Unidentified leakage should include all leakage other than (a) leakage into closed systems, such as pump seal or valve packing leaks that are captured, flow metered, and conducted to a sump or collecting tank, or (b) leakage into the containment atmosphere from sources that are both specifically located and known either net to interfere with the operations of unidentified leakage monitoring systems or not to be from a throughwall crack in the pipina within the reactor coolan*.

pressure 'aoundary.

3. For plants operating with any IGSCC Category D, E, F, or G welds, at least one of the leakage measurement instruments associated with each sump shall be operable, and tte outage time for inoperable instruments shall be limited to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, or immediately initiate an orderly shutdown.

I k

1

, TABLE 1

SUMMARY

OF INSPECTION SCHEDULES FOR BWR PIPING WELDMENTS IGSCC INSPECTION DESCRIPTION OF WELDMENTS NOTES CATEGORY EXTENT & SCHEDULE Resistant Materials A 25% every 10 years (at least 12% in 6 years)

Nonresistant Matis (1) B 50% every 10 years SI within 2 yrs of (at least 25% in 6 years) operation (1)

Nonresistant Matis (1) C All within the next 2 refueling SI after 2 yrs of cycles, then all every 10 years operation Non Resistant Matl (1) D All every 2 refueling cycles  ;

No SI Cracked (1)(2) E 50% next refueling outage, then Reinforced by weld overlay all every 2 refueling cycles or mitigated by SI Cracked (2) F All every refueling outage Inadequate or no repair hon Resistant (3) G All next refueling outage Not Inspected Notes:

(1) All welds in non-resistant material should be inspected after a stress improvment process as part of the process. Schedules shown should be followed after this initial inspection.

(2) See recommendations for acceptable weld overlay reinforcements and stress improvement mitigation.

(3) Welds that are not UT inspectable should be replaced, " sleeved", or local leak detection applied. RT examination or visual inspection for leakage may also be considered.

l a___.m_ _ _ _ . . _ _ _ _ _ _ _ _

i

, p uco

/ ~go UNITED STATES

!" n NUCLEAR REGULATORY COMMISSION E W ASHINoToN, D. C. 20555 3

l  ;

RESOLUTION OF PUBLIC COMMENTS TO DRAFT GENERIC LETTER, "NRC POSITION ON IGSCC IN BWR AUSTENITIC STEEL PIPING," AND NUREG 0313, REV. 2 Introduction The Draft Generic Letter "NRC Position on IGSCC in BWR Austenitic Stainless Steel Piping" and the Draft NUREG 0313, Revision 2" Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary

' Piping" were issued for.public comment in the July 21, 1986 edition of the Federal Register, The formal period of public comments. ended September 20, 1986, although comments also were received considerably after that date.

Formal comments were received from 12 utilities licensed to operate BWRs, General Electric, two consulting firms, and the Boiling Water Reactor Owners

. Group for IGSCC Research. Altogether, well over 200 comments were received, but most of these were essentially duplicates or very similar in content and thrust. Most of the comments were essentially editorial, requests for clarification, suggested re-wordings, etc, that did not represent any disagree-ment with the technical or licensing aspects of the documents. Nevertheless, several subjects did require further review and technical resolution.

In the following discussion of the staff resolution of comments, it was convenient to group them by subject, as treating every one individually would be Jnwieldy, as well as confusing to the reader.

Most'of the important commer.ts were already combined into one package by the BWR Owners Group for IGSCC Research. These will be discussed first, and will cover the major subjects. The order of comments follows the numbering used by the Owners Group. Individual comments not covered by the Owner's Group will be addressed later. To the extent possible, the comments are simplified to reflect the salient issue. In man / cases, clarification of the issue was accomplished by telecon with the originator of the comment or a representative of the Owners Group.

Owners Group Comments

1. (a) Comment:

The staff position on IGSCC resistance of castings would require augmented inspection of welds joining piping to cast valve and pump bodies, unless the castings contained a maximum of 0.035 carbon and a minimum of 8% ferrite. It was fe.it that this would not be warranted, in view of the generally good experience at such joints.

Discussion:

The basic concern is that when piping is replaced with resistant material, the original pumps and valves are retained. These (J4perally are made of material with a carbon content of 0.06 to 0.08, and ferrite levels may be as low as 5%.

Because the staff position on castings would classify these as nonresistant material, the joints would be classified as IGSCC category B, C, or D, depending on when and if the joint was given an SI treatment, and would {

require augmented inspections.

l b

-g-Information in the technical literature available to the staff indicates i that high carbon, low ferrite castings are not as resistant to sensitization and IGSCC as other materials classified as resistant by the staff.

However, such castings are more resistant than, for example, wrought type 304.

There has been at least one known case of IGSCC in a cast body in a BWR.

This case (Monticello) was apparently unusual because the joint had been repeatedly weld repaired, resulting in probable sensitization and unfavorable residual stresses. The particular weld joints involved have a different configuration than simple pipe-to pipe butt welds. The heavier section of the pump or valve side of the joint results in a more favorable residual stress pattern, which, in combination with the relatively mild degree of sensitization, makes this specific type of joint resistant to IGSCC.

Weld repairs to this joint may increase the probability of significant sensitization and modify the residual stress distribution, so must be '

taken into account.

Resolution:

The staff has concluded that unless there has been a history of weld repair to an individual joint, that pipe-to pump and pipe-to-valve joints may be classified as IGSCC category A, and augmented inspections schedules are not required, if the piping is of resistant material. The Generic Letter and NUREG have been modified to reflect this position.

1. (b) Commen t:

Plain carbon steel is immune to sensitization and IGSCC, therefore it should not be addressed in the Generic Letter or NUREG.

Discussion:

Historically, plain carbon steel has been included in the NUREG list of resistant materials. Because of concerns about the suitability of high strength steels for BWR piping, the NRC followed the lead of General Electric and referred to the material as " low strength, high toughness carbon steel".

The Owners Group apparently do not want specifications for strength and toughness for carbon steel in the document related to control of IGSCC in ,

austenitic stainless steels.

Resolution The Scope of the Generic letter has been specifically re-written to exclude consideration of piping made of carbon steel classified as P-1 by the Code.

Although the staff would have concerns if high strength ferritic materials were to be used as piping replacement material (because they are subject to types of stress corrosion cracking) this concern does not have to 'oe addressed in these documents.

2. Comment The concern here apparently is that the staf f positions would classify weld metal with above 5 but below 8% ferrite as non-resistant, and would mean that Code-approved material would require augmented inspection.

I

, Discussion

~Unfortunately, the Code'does not specify requirements needed for control of environmental problems. It specifically states that this is the responsibility of the Owner. The fact that the Code only requires 5% ferrite in austenitic weld metal is.not the relevant issue. The staff position on weld metal is that if it has a maximum of 0.035 carbon and a minimum of 8% ferrite, it is '

resistant. The resistance to sensitization and IGSCC of weld material depends on several interrelated factors; carbon content, ferrite content, size, metallurgical structure, and perhaps others. It is certainly possible to have

.a weld containing only 5% ferrite that is adequately resistant to sensitization, depending on these other factors. At the present time, the staff cannot provide a quantification of the synergisms involved.

Certainly, welds with low or marginal ferrite content might be shown to be resistant, and may be evaluated as such as a special case by the staff.

Resolution: I The staff position has been modified to specifically allow for evaluation.of the resistance of weld metal on an individual case basis.

3. Comment Wrought material solution heat treated after welding should be considered to be resistant.

Discussion:

This is the intent of the staff, as indicated in the position on processes.

Resolution:

To provide clarification, solution heat treated material is now also covered under the Staff Position on Materials.

4. Comment  !

l This comment points out that the resistant material type 347 modified may have carbon content up to 0.04.

Discussion:

Because type 347 is stabilized with niobium, it can contain higher carbon content and still be resistant to sensitization.

Resolution:

The maximum carbon content for type 347 has been increased to 0.04 in the Staff Position on Materials.

1 l

l

..: 5. Comment

-This comment provided a suggested rewrite of the Staff Position on

~

Materials.

Discussion:

For several' reasons, the words suggested do not reflect the exact Positions intended by the staff, but differences are minor, or as discussed in the previous comments.

Resolution:

D The Staff Position on Materials has been rewritten, and incorporates most of the suggestions.

6. Comment

~

Suggested deletion of the words "new or replacement".in the Staff Position on Processes.

Discussion:

This represents a typographical type of error.

Resolution:

The words "new or replacement" have been deleted.

7. ' Comment The Staff Position on Water Chemistry should be updated to reflect recent experience, and should be confined to piping.

Discussion:

Recent experience and results of research and development have provided more confidence in water chemistry control.

Resolution:

The Staff Position has been reworded to provide a stronger position, and references to components other than piping have been deleted because they are not' relevant to the piping issue.

8. Comment Under the Staff Position on Inspection Methods and Personnel, there may be confusion regarding the applicability of the requirement for a qualification program, i

-.L_--__:___--____- . - _ - _ .- - _ _ _ _ - _ _ _ _ - - _ _ _ - _ _ _ _ - . - . _ _ _ _ _ _ - - - - . _ _ _ - _ _ . - _ - _ _ - . _ _ _ - - . _ - _ . _ - _ _ . _ _ - _ ~ - _ _ - _ . _ _ _ . _ . _ _ _

~ Discussion:

It is the intent of this Generic Letter to require a formal qualification program only for examinations performed on piping covered by the Scope of the letter. .It is not the intent of this Generic Letter to require such

, qualification for other Code required examinations. '

Resolution:

The staff position has been specifically reworded to include only those examina-that come within the Scope of the letter.

9. Comment The issue involved in this comment has to do with the inclusion of Class 3 piping in the Scope. Many commenters discussed this issue.

1 Discussion:

This is an important issue, with several ramifications. Many discussions and j telecons were held with the various commenters to clarify their concerns and determine the real extent of the problem. As the Scope is presently written, only austenitic piping 4 inches or larger normally containing reactor coolant at temperature ovee 200 F is included. The salient facts appear to be the following:

1. The only piping that is either class 3 or unclassed_that is covered by the Scope that was identified by the discussions with commenters is that part of the reactor water cleanup system past the second isolation valve on some plantt.

2. Several of the commenters, after review, reported that they had no class 3 or unclassed piping that would be subject to the Generic Letter. This is because later plants have carbon steel in these systems.

3. A major concern was that the Class 3 and unclassed piping was fabricated and erected to less stringent codes; welds were not originally inspected, or even designed and fabricated to facilitate examination according to the requirements of the Generic Letter. .

4. It was agreed that the reactor water cleanup system past the second isolation valve was not directly related to reactor safety, nevertheless the system is required to maintain adequate reactor coolant chemistry control. Thus its function is to protect the reactor coolant system, including piping, from environmental degradation.

5. If weld failure or excessive leakage in this system occurred, it could be isolated, but the plant could not operate for very long before the specification on water chemistry intended to protect the core and other components would be violated. Thus the continued operability of this system is directly related to plant availability and' core integrity. {

i

6. there is no effective leak detection system for this piping, thus '

increasing the probability that a large, critical-size crack could develop without prior knowledge.

7. The same factors stated by the commenters to cause difficulty in examining the welds (lack of close control of welding and weld  !

. configuration, lack of original inspections, etc) are factors that i are also likely to increase the probability of initiating and growing IGSCC.

8. The piping system in question has experienced considerable cracking.

The staff feels that when there is a high probability of cracks, the piping does not meet the original Code or licensing basis unless periodic effective examinations are performed, because neither the ASME Code or ANSI B31.1 permit cracks. >

Resolution: i i

After reviewing these factors, the staff concludes that it is prudent to apply i the principles covered by the Generic Letter to all piping, regardless of class, that is covered by the Scope of the Generic Letter. Cases involving undue hardship may, of course, be reviewed on an individual basis.

10. a Comment:

Snme inspection schedules are intended to apply every other refueling outage. Commenters suggested that with present extended cycles, this

-period could be up to 4 years, rather than the 3 1/3 years as shown in the table in the Draft Generic Letter. (The 3 1/3 year interval in the Table was supposed to synchronize with the Code schedule of 3 1/3, 6 2/3 and 10 years). Commenters suggested the wording "Every two refueling cycles or 4 years, whichever is greater."

Discussion:

The staff believes that basing the inspection schedule on refueling cycles is acceptable, even if the refueling cycle is extended to 2 years. It is felt that the industry's stated concern can be more simply accommodated by basing inspections on refueling cycles alone, with no specificed time period. As the inspection period for Category A & B welds are based on Code, they should stay on a 10 year basis.

Resolution:

Except for IGSCC Category A and B welds, inspection schedules will be based only on refueling cycles.

10. b Comment:

The commenters picked up a discrepancy between the text and the Table related to the inspection requirements for IGSCC Category E welds during the next outage. '

i i

l Discussion: I This was typographical type of error. The intent was to have the table match.

the text.  !

Resolution:

The Table has been changed to match the text. I 10.c Comment:

The extent of inspection shown in Table 1 of the draft Generic Letter indicated that all terminal ends and dissimilar metal welds and, in addition,.

25% of other welds, should be examined. This was correctly shown to be more ,

, than required by the Code. It was suggested that the Code extent be used l directly.

' Discussion:

The extent of examination shown in the Table was intended to reflect the fact that for a replaced, reJesigned recirculation system, following the Code sample selection process would, in many cases, result'in a non-representative sample.

This is because the 1977 and later Codes requit a that 100% of all terminal ends and dissimilar metal welds must be examined, plus enough other welds to bring the total sample to 25%. However, the new design of recirculation systems has only about 50-60% of the total number of welds as the older design, whereas it-has the same number of terminal ends and dissimilar metal welds. This would result in an examination sample essentially entirely of terminal ends and dissimilar metal welds, because toey would constitute approximately 25% of the total number of welds.

In reviewing the distribution of racking that has occurred in recirculation systems, the staff noted that terminal ends and dissimilar metal welds had a significantly lower percentage of cracked welds than the other welds. Clearly, a sample based entirely on these would not be a meaningful or representative sample. The staff believes that the sample selection should be based on good engineering judgement, and if bias toward certain types of welds is warranted for technical reasons, this should be reflected in the sample seloction. The staff feels that the licensee is in the best position to app'ly his knowledge of stress levels, weld configuration, mitigation effectiveness and similar factors in selecting a meaningful sample.

Discussions with several Code representatives indicated that they agree in principle with this approach, and are now considering changes in the Code to permit other sample selection bases. Until such changes are effected, licensees may use the provisions of 10 CFR 50.55a, (b),(2)(ii), which permits the use of the 1974 Edition of the Code for sample selection.

.v Resolution:

The table and the text of the Generic Letter have been rewritten to require that the licensee select the sample of 25% for IGSCC Category A and the 50%

sample for IGSCC Category B based on his judgement in selecting a meaningful and representative sample. The 1974 Edition of the Code permits this, and the provisions of 10CFR50.55a, b, (2),(ii) are referenced.

11. Comment:

Concern was expressed regarding the requirement that the leakage detection systems must conform to Regulatory Guide 1.45, because l although many do not, they have been accepted by the NRC on an individual basis.

Discussion:

The adequacy of the leakage detection systems was addressed for each plant during the licensing process, and should not be re-opened as part of the IGSCC issue. The wording used in the draft Generic Letter was copied directly from NUREG-0313 Rev. 1, and was considered non-controversial.

Resolution:

The wording of the Generic Letter tas been modified by adding the words

"--- or as otherwise approved by the NRC" to the sentence containing the requirement.

12. Comment:

There are three items included in this comment. Two are essentially editorial and non controversial. The third involves the details of how a second sample of welds to be examined is selected, when one or more cracked welds is found in the group of welds constituting the required sample.

Discussion:

It is agreed that the sample expansion plan included in the original draft NUREG-0313 Rev. 2 (it was not included in the draft Generic Letter) was confusing as written, and included some provisions that, upon closer review, were redundant or not relevant.

Resolution:

The entire position on requirements for sample expansion has been rewritten, simplified, and included in the Generic Letter as part of the Staff Position on Sample Expansion. The staff believes that this new section will resolve the concerns expressed by the commenters, while still providing a sample expansion plan that is consistent with past staff and Code practices.

as

13. Comment:

Most of the commenters disagreed strongly with the proposed 25% limit on number of cracked welds in a piping system. Because this is a significant issue, the comment submitted by the Owners Group is included here in its entirety:

"A proposed addition to the draft generic letter sets a limit on the number of cracked weldments in piping, irrespective of repair or evaluation.

The basis for this limit is not discussed in draft NUREG-0313, Rev. 2 and appears to be arbitrary. The position of the BWROG-II is that measures for evaluation, mitigation, reinforcement and inspection as identified in the draft NUREG revision are adequate to control IGSCC in BWR piping systems, and that the additional arbitrary limitation is not necessary or beneficial.

Meaningful technical discussion is difficult because the basis for the proposed limit has not been disclosed. We speculate that a possible motivation is a desire to minimize shrinkage stress resulting from a large number of weld overlay repairs. However, shrinkage and fitup stresses are best quantified by analysis and controlled as part of the overlay design process, as discussed in the draft NUREG revision. Shrinkage stress is not sensitive to the number or proportion of welds repaired and is not limited by the proposed 25% limit. For example, analysis ould show that a single .

overlay repair in a recirculation system riser causes higher shrinkage stress than a symmetric arrangement of repairs in each riser associated with o..e ring header. The arbitrary 25% limitation is not an effective way to limit or reduce shrinkage and fitup stresses."

"We question whether an analysis of costs and benefits associated with the proposed limit would support its implementation. The proposed limit requires BWR owners to apply increased resources and man-rem exposure toward inspection and sizing of small cracks which can have only small effects on system intyrity. An arbitrary limit on the total number of small cracks could force replacement of lines which are structurally sound and reliable, again at high cost radiation exposure. The proposed limit precludes use of alternatives described in the draft NUREG revision which could be more cost effective in maintaining system integrity. For these reasons, we recommend that the five paragraphs in the draf t generic letter under the heading Limits on Number of Cracked Weldments in Piping be deleted in their entirety."

Discussion:

This proposed limit on the number of welds (regardless of repair or mitigation action) with cracks exceeding the Code size criterion for acceptability without evaluation (IWB3500) was recommended by the ACRS in a memorandum, Ward to Stello dated March 18, 1986. Although staff reviewers felt that the limit was not necessary from a technical standpoint, and that placing such a limit on welds with cracks could be counter productive to the inspection program, the proposed limit was included for public comment at the suggestion of the CRGR. The major difficulty that would be encountered if the limit were to be imposed is related to the fundamental problem that crack sizing is inherently

.- l inaccurate in the size range of concern. According to the ACRS proposal, cracks of a size acceptable under IWB3500 of the Code would not be included in j the limit criterion. This size is approximately 10% of the wall thickness, J depending on the length of the crack. Thus, a weld with a crack only 9% of l the wall in depth would not be counted against the 25% limit, whereas if it l were to be reported as;11%, it would be. Although the technology for depth sizing has recently been developed to the point where the accuracy is acceptable for valuational purposes, it is still not possible to consistently size to an accuracy of, for example, plus or minus 5% of the nominal wall thickness of BWR piping. Further, shallow cracks, in the range of 5 to 15% of the the wall, are even more difficult to size accurately than deeper cracks.

Thus, a crack actually about 10% of the wall in depth (just at the Code limit) could easily be sized anywhere from 5 to 15%. Without the proposed 25% limit, conservative depth calls have no severe consequences. With a limit on the number of welds with cracks sized deeper than 10%, there certainly will be more reluctantance on the part of the examiner and Licensee to make conservative depth estimates. Staff reviewers believe that these considerations would mean that the imposition of a limit on the number of welds with cracks requir-ing evaluation could be counterproductive to the overall effectiveness of the inspection program.

Resolution:

The staff has concluded that adequata control of IGSCC can be achieved through the implementation of the Staff Positions in the Generic Letter without the need for a limit on number of cracked welds in the piping. Therefore, the portion of the Generic Letter pertaining to this limit has been deleted.

14. Comment:

The draft Generic Letter required that the inservice inspection plan be incorporated in the plant Technical Specifications. It was pointed out that this would be a lengthy addition to the Tech Specs, and because it would have to be updated very often, perhaps every outage in some cases, it would be inappropriate to include the detailed program in the Tech Specs.

Discussion:

The staff agrees that incorporating the inspection plan in the Technical Specification would be unweildy and necessitate excessive paperwork.

Nevertheless, the staff feels that linkage with the Tech Specs may be advisable from a regulatory standpoint.

Resolution:

The Generic Letter has been changed to require only that a statement be included in the Technical Specification to the effect that the inservice inspection pregram for piping covered under the Scope of the Generic Letter will be in compliance with the staff positions and Table 1 of the Generic Letter.

i

INDIVIDUAL COMMENTS General Discussion:

There were many comments of an editorial nature that were incorporated into ^

the final documents if deemed appropriate. There also were several requests for clarification, or comments apparently based on a mis-reading of the documents. These were handled by direct communication with the commenter, and usually did not warrant any modification of the text.

Specific Comments

1. Comment:

The reduction in augmentation of inspection schedules that may be allowed by the use of HWC should not result in less inspection than that required for IGSCC Category A welds, which is the basic code requirement.

Discussion:

This was covered in the draft NUREG 0313 Rev. 2, and as written would not permit the . situation of concern. This detail is now also included in the Generic Letter.

Resolution:

Credit for HWC in reducing the frequency of augmented inspections is limited to' welds in IGSCC categories B,C,0, and E by the Generic Letter.

2. Comment:

The recently developed SI method, Mechanical Stress Improvement Process (MSIP) should be considered to be fully effective, and the same credit for mitigation should be applied as that given IHSI.

Discussion:

Since the publication of the Draft documents, additional research, develop- l ment, and confirmatory work has been done on the MSIP process. Confirmatory work at Argonne National Laboratory, funded by the NRC, is reported in a Research Information Letter, RIL 147 " Evaluation of Mechanical Stress Improvement Process". This RIL recommends that MSIP be considered a fully effective Stress Improvement process.

Resolution:

With the publication of the RIL referred to above, the st'aff has modified the Generic Letter (and NUREG) to include MSIP as a fully effective SI treatment.

. 3. Comment:

There were several comments and suggestions related to evaluation of the extent of mitigation provided by weld overlays and IHSI, and the resulting degree of inspection augmentation.

Discussion:

The comments covered here all relate to whether a given weld should be classified as IGSCC Category E or F, with the proposed changes resulting in upgrading of some marginal welds to Category E. The staff reviewed the suggestions and determined that the preser,t criteria should be retained.

Nevertheless, there could be situations for which less conservative criteria would be adequate. The staff believes that these might best be proven by service experience.

Resolution:

The staff has concluded that if an IGSCC Category F weld (which must be examined every outage) shows no adverse change through four refueling outages,

'it may be upgraded to IGSCC Category E. This has been detailed in the Generic Letter under Staff Position on Inspection Schedules.

4. Comment:

It was requested by one commenter that IGSCC Category A weldments be excluded from the requirement that examiners be qualified to detect IGSCC.

Discussion:

As specifically indicated by the wording used throughout the Generic Letter and the NUREG, resistant materials, although resistant to IGSCC, are not immune. Clearly, one purpose of an inservice examination is to detect IGSCC should it occur as a result of inadvertent use of improper material, excessive weld repairs or similar problems. It is now realized from experience and shown by industry and NRC programs that unless proper examination methods are used by personnel specifically trained for the task, the examinations prescribed by the Code are almost useless for detection of IGSCC.

Further, the detection of transgranular stress corrosion cracks, and even the detection of fatigue cracks in austenitic stainless steel require the same degree of attention to detail and training.

The staff believes that it is not improbable that cracks may develop in some IGSCC Category A welds, and because the sample size (only 25% of the welds in 10 years) is small, the examinations that are performed should be effective.

Resolution:

The requirements for training and qualification of examination personnel for those examining IGSCC Category A weldments will remain in the Generic Letter.

y"

/ 'o n

UNITED ST ATES NUCLEAR REGULATORY COMMISSION

$' W AsHINGTON, D. C. 20666

, _{

\...../

RESOLUTION OF PUBLIC COMMENTS TO DRAFT GENERIC LETTER, "NRC POSITION ON IGSCC IN BWR AUSTENITIC STEEL PIPING," AND NUREG 0313, REV. 2 Introduction The Draf t Generic Letter "NRC Position on IGSCC in BWR Austenitic Stainless Steel Piping" and the Draft NUREG 0313, Revision 2" Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary Piping" were issued for public comment in the July 21, 1986 edition of the Federal Register. The formal period of public comments ended September 20, 1986, although comments also were received considerably after that date.

Formal comments were received from 12 utilities licensed to operate BWRs, General Electric, two consulting firms, and the Boiling Water Reactor Owners Group for IGSCC Research. Altogether, well over 200 comments were received, but most of these were essentially duplicates or very similar in content and thrust. Most of the comments were essentially editorial, requests for clarification, suggested re-wordings, etc, that did not represent any disagree-ment with the technical or licensing aspects of the documents. Nevertheless, several subjects did require further review and technical resolution.

In the following discussion of the staff resolution of comments, it was conves ient to group them by subject, as treating every one individually would be unwieldy, as well as confusing to the reader.

Most of the important comments were already combined into one package by<the BWR Owners Group for IGSCC Research. These will be discussed first, and will cover the major subjects. The order of comments follows the numbering used by the Owners Group. Individual conments not covered by the Owner's Group will be addressed later. To the extent possible, the comments are simplified to reflect the salient issue. In many cases, clarification of the issue was accomplished by telecon with the originator of the comment or a representative of the Owners Group.

Owners Group Comments

1. (a) Comment:

The staff position on IGSCC resistance of castings would require augmented inspection of welds joining piping to cast valve and pump bodies, unless the castings contained a maximum of 0.035 carbon and a minimum of 8% ferrite. It was felt that this would not be warranted, in view of the generally good experience at such joints.

Discussion:

The basic concern is that when piping is replac with resistant material, the original pumps and valves are retained. These rally are made of material with a carbon content of 0.06 to 0.08, and ferrit levels may be as low as 5%.

Because the staff position on castings would classify these as nonresistant material, the joints would be classified as IGSCC category B, C, or 0, depending on when and if the joint was given an SI treatment, and would require augmented inspections.

4

. Information in the technical literature available to the staff indicates that high carbon, low ferrite castings are not as resistant to sensitization and IGSCC as other materials classified as resistant by the staff.

However, such castings are more resistant than, for example, wrought type 304.

There has been at least one known case of IGSCC in a cast body in'a BWR.

This case (Monticello) was apparently unusual because the joint had been repeatedly weld repaired, resulting in probable sensitization and unfavorable residual stresses. The particular weld joints involved have a different' configuration than simple pipe-to pipe butt welds. The heavier section of the pump or valve side of the joint results in a more favorable residual stress pattern, which, in combination with the relatively mild degree of sensitization, makes this specific type of joint resistant to IGSCC.

. Weld repairs to this joint may increase the probability of significant sensitization and modify the residual stress distribution, so must be taken into account.

Resolution:

The staff has concluded that unless there has been a history of weld repair to an individual joint, that pipe-to pump and pipe-to-valve joints may be classified as IGSCC category A, and augmented inspections schedules are not required, if the piping is of resistant material. The Generic Letter and NUREG have been modified to reflect this position.

1. (b) Comment:

Plain carbon steel is immune to sensitization and IGSCC, therefore it should not be addressed in the Generic Letter or NUREG.

Discussion:

Historically, plain carbon steel has been included in the NUREG list of resistant materials. Because of concerns about the suitability of high strength steels for BWR piping, the NRC followed the lead of General Electric and referred to the material as " low strength, high toughness carbon steel".

The Owners Group apparently do not want specifications for strength and toughness for carbon steel in the document related to control of IGSCC in austenitic stainless steels.

Resolution The Scope of the Generic letter has been specifically re-written to exclude consideration of piping made of carbon steel classified as P-1 by the Code.

Although the staff would have concerns if high strength ferritic materials were to be used as piping replacement material (because they are subject to types of stress corrosion cracking) this concern does not have to be addressed in these documents.

2. Comment The concern here apparently is that the staf f positions would classify weld metal with above 5 but below 8% ferrite as non-resistant, and would mean that Code-approved material would require augmented inspection.

Discussion Unfortunately, the Code does not specify requirements needed for control of environmental problems. It specifically states that this is the responsibility of the Owner. The fact that the Code only requires 5% ferrite in austenitic weld metal is not the relevant issue. The staff position on weld metal is that if it has a maximum of 0.035 carbon and a minimum of 8% ferrite, it is resistant. The resistanct to sensitization and IGSCC of weld material depends on several interrelated factors; carbon content, ferrite content, size, metallurgical structure, and perhaps others. It is certainly possible to have a weld containing only 5% ferrite that is adequately resistant to sensitization, depending on these other factors. At the present time, the staff cannot provide a quantification of the synergisms involved.

Certainly, welds with low or marginal ferrite content might be shown to be resistant, and may be evaluated as such as a special case by the staff.

Resolution:

The staff position has been modified to specifically allow for evaluation of the resistance of weld metal on an individual case basis.

3. Commeo; Wrought material solution heat treated after welding should be considered to be resistant.

Discussion:

This is the intent of the staff, as indicated in the position on processes.

Resolution:

To provide clarification, solution heat treated material is now also covered under the Staff Position on Materials,

4. Comment This comment points out that the resistant material type 347 modified may have carbon content up to 0.04.

Discussion:

Because type 347 is stabilized with niobium, it can contain higher carbon content and still be resistant to sensitization.

Resolution:

The maximum carbon content for type 347 has been increased to 0.04 in the Staff Position on Materials.

l

5. Comment This comment provided a suggested rewrite of the Staff Position on--

-Materials.

Discussion:

For several reasons, th, words suggested do not reflect the exact Positions intended by.the staff, i it differences are minor, or as discussed in the previous comments.

Resolution:

The Staff Position on Materials has been rewritten, and incorporates most of the suggestions.

6. Comment Suggested deletion of the words "new or replacement" in the Staff Position on Processes.

Discussion:

This represents a typographical type of error.

Resolution:

The words "new or replacement" have been deleted.

7. -Comment The Staff Position on Water Chemistry should be updated to reflect recent experience, and should be confined to piping.

Discussion:

Recent experience and results of research and development have provided more confidence in water chemistry control.

Resolution:

The Staff Position has been reworded to provide a stronger position, and references to components other than piping have been deleted because they are ,

I not relevant to the piping issue.

8. Comment Under the Staff Position on Inspection Methods and Personnel, there may be confusion regarding the applicability of the requirement for a qualification program.

- Discussion:

l l

It is the intent of this Generic Letter to require a formal qualification program only for examinations performed on piping covered by the Scope of the letter. It is not the intent of this Generic Letter to require such .

qualification for other Code iequired examinations.

Resolution:

1 The staff position has been specifically reworded to include only those examina-that come within the Scope of the letter.

9. Comment The issue involved in this comment has to do with the inclusion of Class 3-piping in the Scope. Many commenters discussed this issue.

Discussion:

This is an important issue, with several ramifications. Many discussions and telecons were held with the various commenters to clarify their concerns and determine the real extent of the problem. As the Scope is presently written, l only austenitic piping 4 inches or larger normally containing rear. tor coolant '

at temperature over 200 F is included. The salient facts appear to be the following:

1. The only piping that is either class 3 or unclassed that is covered by the Scope that was identified by the discussions with commenters is that part of the reactor water cleanup system past the second isolation valve on some plants.

2. Several of the commenters, after review, reported that they had no class 3 or unclassed piping that would be subject to the Generic Letter. This is because later plants have carDon steel in these systems.

3. A major concern was that the Class 3 and unclassed piping was fabricated and erected to less stringent codes; welds were not originally inspected, or even designed and fabricated to facilitate examination according to the requirements of the Generic Letter.

4. It was agreed that the reactor water cleanup system past the second isolation valve was not directly related to reactor safety, nevertheless the system is required to maintain adequate reactor coolant chemistry  ;

control. Thus its function is to protect the reactor coolant system, including piping, from environmental degradation.

5. If weld failure or excessive leakage in this system occurred, it I

could be isolated, but the plant could not operate for very long before the specification on water chemistry intended to protect the core and other components would be violated. Thus the continued operability of this system is directly related to plant availability and core integrity.

i

1 I

6. there is no effective leak detection system for this piping, thus increasing the probability that a large, critical-size crack could develop without prior knowledge.

7. The same factors stated by the commenters to cause difficulty in )

examining the welds (lack of close control of welding and weld i configuration, lack of original inspections, etc) are factors that I are also likely to increase the probability of initiating and )

growing IGSCC. {

j The piping system in question has experienced considerable cracking.

8.

The staff feels that when there is a high probability of cracks, the piping does not meet the original Code or licensing basis unless periodic effective examinations are performed, because neither the ASME Code or ANSI B31.1 permit cracks.

Resolution:

After reviewing these factors, the staff concludes that it is prudent to apply the principles covered by the Generic Letter to all piping, regardless of class, that is covered by the Scope of the Generic Letter. Cases involving undue hardship may, of course, be reviewed on an indivi.iual basis.

10. a Comment:

Some inspection schedules are intended to apply every other refueling outage. Commenters suggested that with present extended cycles, this period could be up to 4 years, rather than the 31/3 years as shown in the table in the Draft Generic Letter. (The 3 1/3 year interval in the Table was supposed to synchronize with the Code schedule of 3 1/3, 6 2/3 and 10 years). Commenters suggested the wording "Every two refueling cycles or 4 years, whichever is greater."

Discussion:

The staff believes that basing the inspection schedule on refueling cycles is acceptable, even if the refueling cycle is extended to 2 years. It is felt that the industry's stated concern can be more simply accommodated by basing inspections on refueling cycles alone, with no specificed time period. As the inspection period for Category A & B welds are based on Code, they should stay on a 10 year basis.

Resolution:

Except for IGSCC Category A and B welds, inspection schedules will be based only on refueling cycles.

10. b Comment:

The commenters picked up a discrepancy between the text and the Table related to the inspection requirements for IGSCC Category E welds during the next outage. 1 l

l

)

1 V ,

l

~7- l Discussion:

This was typographical type of error. The intent was to have the table match the text.

]

Resolution:

The Table has been changed to match the text. I 10.c Comment:

The extent of inspection shown in Table 1 of the draft Generic Letter l indicated that all terminal ends and dissimilar metal welds and, in addition, 25% of other welds, should be examined. This was correctly shown to be more

, than required by the Code. It was suggested that the Code extent be used directly. ]

Discussion:

The extent of examination shown in the Table was intended to reflect the fact that for a replaced, redesigned recirculation system. following the Ccre sample selection process would, in many cases, result in a non-representative sample.

This is because the 1977 and later Codes require that 100% of all terminal ends and diss{milar metal welds must be examined, plus enough other weld, to bring the total sample to 25%. However, the new design of recirculation systems has only about 50-60% of the total number of welds as the older design, whereas it has the same number of terminal ends and dissimilar metal welds. This would

. result in an examination sample essentially entirely of terminal ends and dissimilar metal welds, because they would constitute approximately 25% of the total number of welds.

In reviewing the distribution of cracking that has occurred in recirculation systems, the staff noted that terminal ends and dissimilar metal welds had a significantly lower percentage of cracked w?lds than the other welds. Clearly, a sample based entirely on tnese would not be a meaningful or representative sample. The staff believes that the sample selection should be based on good engineering judgement, and if bias toward certain types of welds is warranted for technical reasons, this should be reflected in the sample selection. The staff feels that the licensee is in the best position to apply his knowledge of stress levels, weld configuration, mitigation effectiveness and similar factors in selecting a meaningful sample.

Discussions with several Code representatives indicated that they agree in principle with this approach, and are now considering changes in the Code to permit other sample selection bases. Until such changes are effected,  :

licensees may use the provisions of 10 CFR 50.55a, (b),(2)(ii), which permits the use of the 1974 Edition of the Code for sample selection.

______.___________._________________________________________w

Resolution:

The table and the text of the Generic Letter have been rewritten to require that the licensee ~ select the sample of 25% for IGSCC Category A and the 50% .

sample for IGSCC Category B based on his judgement in selecting a meaningful and representative sample. The 1974 Edition of the Code permits this, and the provisions of 10CFR50.55a, b, (2),(11) are referenced.

11. Comment:

Concern was expressed regarding the requirement that the leakage detection systems must conform to Regulatory Guide 1.45, because although many do not, they have been accepted by the NRC on an individual basis.

Discussion:

The adequacy of the leakage detection systems was addressed for each plant during the licensing process, and should not be re-opened as part of the IGSCC issue. The wording used in the draft Generic Letter was copied directly from NUREG-0313 Rev. 1, and was considered non-controversial.

Resolution:

The wording of the Gen.ric Letter has been modified by adding the words

"--- or as otherwise approved by the NRC" to the sentence containing the requirement.

12. Comment:

There are three items included in this comment. Two are essentially editorial and non-controversial. The third involves the details of how a second sample of welds to be examined is selected, when one or more cracked welds is found in the gro'up of welds constituting the required sample.

Discussion:

It is agreed that the sample expansion plan included in the original draft NUREG-0313 Rev. 2 (it was not included in the draft Generic Letter) was confusing as written, and included some provisions that, upon closer review, were redundant or not relevant.

Resolution:

The entire position on requirements for sample expansion has been rewritten, simplified, and included in the Generic Letter as part of the Staff Position on Sample Expansion. The staff believes that this new section will resolve the concerns expressed by the commenters, while still providing a sample expansion plan that is consistent with past staff and Code practices.

j

13. Comment:

Most of the commenters disagreed strongly with the proposed 25% limit on  !

number of cracked welds in a piping system. Because this is a significant \

issue, the comment submitted by the Owners Group is included here in its {

entirety:

'lA proposed addition to the draft generic letter sets a limit on the number of cracked weldments in piping, irrespective of repair or evaluation. f The basis for this' limit is not discussed in draft NUREG-0313, Rev. 2 and appears to be arbitrary. The position of the BWROG-II is that measures for-evaluation, mitigation, reinforcement and inspection as identified in the draft NUREG revision are adequate to control IGSCC in BWR piping systems, and that the additional arbitrary limitation is not necessary or beneficial.

Meaningful technical discussion is difficult because the basis for the proposed limit has not been disclosed. We speculate that a possible motivation is a desire to minimize shrinkage stress resulting from a large number of weld' overlay repairs. However, shrinkage and fitup stresses are best quantified by analysis and controlled as part of the overlay design process, as discussed in the draft NUREG revision. Shrinkage stress is not sensitive to the number or proportion of welds repaired and is not limited by the proposed 25% limit. For example, analysis would show that a single overlay repair in a recirculation system riser causes higher shrinkage stress than a symmetric arrangement of repairs in each riser associated with one ring header. The arbitrary 25% limitation is not an effective way to limit or reduce shrinkage and fitup stresses."

"We question whether an analysis of costs and benefits associated with the proposed limit would support its implementation. The proposed limit requires BWR owners to apply increased resources and man-rem exposure toward inspection and sizing of small cracks which can have only small effects en system integrity. An arbitrary limit on the total number of small cracks could force replacement of lines which are structurally sound and reliable, again at high cost radiation exposere. The proposed limit precludes use of alternatives l described in the draft HUidG revision which could be more cost effective in maintaining system integrity For these reasons, we recommend that the five paragraphs in the draft generic letter under the heading Limits on Number of Cracked Weldments in Piping be deleted in their entirety."

i l Discussion:

l This proposed limit on the number of welds (regardless of repair or mitigation action) with cracks exceeding the Code size criterion for acceptability without evaluation (IWB3500) was recommended by the ACRS in a memorandum, Ward to Stello dated March 18, 1986. Although staff reviewers felt that the limit was not necessary from a technical standpoint, and that placing such a limit on welds with cracks could be counter productive to the inspection program, the proposed limit was included for public comment at the suggestion of the CRGR. The major difficulty that would be encountered if the limit were to be imposed is related to the fundamental problem that crack sizing is inherently

, inaccurate in the size range of concern. According to the ACRS proposal, cracks of a size acceptable under IWB3500 of.the Code would not be included in the limit criterion. This size is approximately 10% of the wall thickness, depending'on the length of the crack. Thus, a weld with a crack only 9% of the wall in depth would not be counted against the 25% limit, whereas if it were to be reported as 11%, it would be. Although the technology for depth sizing has recently been developed to the point where the accuracy is acceptable for valuational purposes, it is still not possible to consistently size to an accuracy of, for example, plus or minus 5% of th4e nominal wall thickness of BWR piping. Further, shallow cracks, in the range of 5 to 15% of the the wall, are even more difficult to size accurately than deeper cracks.

Thus, a crack actually about 10% of the wall in depth (just at the Code limit) could easily be sized anywhere from 5 to 15%. Without the proposed 25% limit, conservative depth calls have no severe consequences. With a limit on the number of welds with cracks sized deeper than 10%, there certainly will be more reluctantance on the part of the examiner and Licensee to make conservative depth estimates. Staff reviewers believe that these considerations would mean that the imposition of a limit on the number of welds with cracks requir-ing evaluation could be counterproductive to the overall effectiveness of the inspection program.

Resolution:

The staff has concluded that adequate control of IGf;C can be achieved through the implementation of the Staff Positions in the Generic Letter without the need for a limit on number of cracked welds in the piping. Therefore, the portion of the Generic Letter pertaining to this limit has been deleted.

14. Comment:

The draft Generic Letter required that the inservice inspection plan be incorporated in the plant Technical Specifications. It was pointed out that this would be a lengthy addition to the Tech Specs, and because it would have to be updated very often, perhaps every outage in some cases, l it would be inappropriate to include the detailed program in the Tech Specs.

Discussion:

The staff agrees that incorporating the inspection plan in the Technical Specification would be unweildy and necessitate excessive paperwork.

Nevertheless, the staff feels that linkage with the Tech Specs may be advisable from a regulatory standpoint.

Resolution:

The Generic Letter has been changed to require only that a statement be included in the Technical Specification to the effect that the inservice j inspection program for piping covered under the Scope of the Generic Letter {

will be in compliance with the staff positions and Table 1 of the Generic Letter.

1 i

k i

i i

INDIVIDUAL COMMENTS General Discussion:

There were many comments of an editorial nature that were incorporated into ,

the final documents if deemed appropriate. There also were several requests for clarification, or comments apparently based on a mis-reading of the documents. These were handled by direct communication with the commenter, and usually did not warrant any modification of the text.

Specific Comments

1. Comment:

The reduction in augmentation of inspection schedules that may be allowed by the use of HWC should not result in less inspection than that required for IGSCC Category A welds, which is the basic code requirement.

Discussion:

1 This was covered in the draf t NUREG 0313 Rev. 2, and as written would not

>ermit the situation of concern. This detail is now also included in the Generic Letter.

Resolution:

Credit for HWC in reducing the frequency of augmented inspections is limited to welds in IGSCC categories B,C,D, and E by the Generic Letter.

2. Connent:

The recently developed SI method, Mechanical Stress Improvement Process (MSIP) should be considered to be fully effective, and the same credit j for mitigation should be applied as that given IHSI.

Discussion:

Since the publication of the Draft documents, additional research, develop-ment, and confirmatory work has been done on the MSIP process. Confirmatory work at Argonne National Laboratory, funded by the NRC, is reported in a Research Information Letter, RIL 147 " Evaluation of Mechanical Stress Improvement Process". This RIL recommends that MSIP be considered a j fully effective Stress Improvement process.

Resolution:

With the publication of the RIL referred to above, the st'aff has modified the Generic Letter (and NUREG) to include MSIP as a fully effective SI l

treatment.

12 -

3. ' Comment:

There were several comments and suggestions related to evaluation of L the extent of mitigation provided by weld overlays and IHSI, and the resulting degree'of inspection' augmentation.

I Discussion:

The comments covered here all relate to whether a given weld should be classified as IGSCC Category E or F, with the proposed changes resulting in upgrading of some marginal welds to Category Es The staff reviewed the suggestions 'and determined that the present criteria should be retained.

Nevertheless, there could be situations for which less conservative criteria would be adequate. The staff believes that these might best be proven by service experience.

Resolution:

The staff has concluded that if an IGSCC Category F weld (which must be examined every outage) shows no adverse change through four refueling outages, it may be upgraded to IGSCC Category E. This has been detailed 'in the Generic Letter under Staff Position on Inspection Schedules.

4. Comment:

It was requested by one.commenter that IGSCC Category A weldments be excluded from the requirement-that examiners be qualified to detect IGSCC.

Discussion:

As specifically indicated by the wording used throughout the Generic Letter and the.NUREG, resistant materials, although resistant to IGSCC, are not immune. Clearly, one purpose of an inservice examination is to detect IGSCC should it occur as a result of inadvertent use of improper material, excessive weld repairs or similar problems. It is now realized from experience and shown by industry and NRC programs that unless proper examination methods are used by personnel specifically trained for the task, the examinations prescribed by the Code are almost useless for detection of IGSCC.

Further, the detection of transgranular stress corrosion cracks, and even the detection of fatigue cracks in austenitic stainless steel require the same degree of attention to detail and training.

I The staff believes that it is not improbable that cracks may develop in some IGSCC Category A welds, and because the sample size (only 25% of the welds in 10 years) is small, the examinations that are performed should be effective.

Resolution:

The requirements for training and qualification of examination personnel for those examining IGSCC Category A weldments will remain in the Generic Letter,

. 1 1 -

NUREG-0313 i- Rev.2 Technical Report on Material Selection anc Processing Guidelines for BWR Coolant Pressure 1 Boundary Piping l ato$u shed W. S. Hazniton Division of BWR Licensing Office of Nuclear Rs ,.ier Regulation U.S. Nuclear Rege: tory Commission Washington, D.C. J:)555 7....u,

\

f '

h, .

fA <

ABSTRACT This report updates and supersedes the technical recommendations of NUREG-0313,'.

" Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary Piping," published in July 1977, and its subsequent revision published in July 1980.

This re. port provides the technical bases for the NRC staff's revised. recommended methods lto control the intergranular stress corrosion cracking susceptibility of BWR piping. For piping that does not fully comply with the material selection, testing, and processing guideline combinations of this document, varying degrees of augmented inservice inspection are recommended. This l revision also includes guidance regarding crack evaluation and weld overlay ,

repair methods for long. term operation or for continuing interim operation of

- plants'.tntil a more permanent solution.is implemented.

I i

l I

iii

- _ _ _ - _ _ _ _ _ _ - _ - - _ - - - - _ -- -- - - - ----- a

i%

W i.-

' TABLE OF CONTENTS Page

. Executive Summary vi Acronyms viii-p 1.0 Introduction 1.1 1.1 History 1.1

1. 2 Revision 1 of NUREG-0313 12

.1. 3 Revision 2 of NUREG-0313 1.2 1.4. Bases for Recommendations 1.3

1. 5 Piping Replacement 1.3 2.0 Methods to Reduce or Eliminate IGSCC .

2.1 2.1 Materials for New or Replacement Piping 2.1 2.1.1: Staff Recommendations on Materials 2.2 2.2' Processes for.New, Replacement,'or Older Piping 2.3 2.2.1 Staff Recommendations on Processes 2.5 2.3 Water Chemistry Modifications 2.6 2.3.1 Staff Recommendations on Water Chemistry 2.7 3.0 Evaluation and Repair of Cracked Weldments 3.1 3.1 Repair Procedures 3.1 3.1'.1 Weld Overlay. Reinforcement 3.1 3.1.2 Partia's Replacement 3.2 3.1.3 Stress Improvement 3.2 3.1.4 Mechanical Clamping Devices 3. 3 3.2 Staff Recommendations on Repairs 3.3 3.2.1 Staf f Recommendations' on Weld Overlay Reinforcement 3.3 3 '. 2. 2 Staff Recommendations on Partial Replacement 3.3 3.2.3 Staff Recommendations on SI of Cracked Weidments 3.3 3.2.4 Staff Recommendations on Clamping Devices 3.3 4.0 Crack Characterization and Repair Criteria 4.1 4.1 Flaw and Repair Evaluation Criteria 4.1 4.2 Crack Growth Calculations 4.1 4.3 Multiple and Complex Crack Characterization 4.2 4.4 . Weld Overlay Design Criteria 4.3 4.4.1 Standard Overlay Design 4.3 4.4.2 Designed Overlays 4.3 4.4.3 Limited Service Overlays 4.4 4.5 SI Crack Mitigation Criteria 4.4 5.0 Inspection of Piping for IGSCC 5.1 5.1 Weldments Subject to Inspection 5.1 5.2 Inspection Methods 5.1 5.2.1 Staff Recommendations on Inspection Methods and Personnel 5.1 5.2.2 Flaw Size Uncertainty 5.2 5.3 Inspection Frequency 5.2 iv

7 5.3.1 Weldment IGSCC Condition Category Definitions . 5. 2

5. 3. '1.1 Definition of IGSCC' Category A Weldments 5.3 5.3.1.2 Definition of IGSCC Category B Weldments 5.3 5.3.1.3 Definition of IGSCC Category C Weldments 5.3 5.3.1.4 Definition of IGSCC Category D Weldments 5.3 5.3.1.5 Definition of IGSCC Category E Weldments 5.4 5.3.1.6 Definition ~of IGSCC Category F Weldments 5.4 5.3.1.7 Definition of IGSCC Category G Weldments 5. 4' 5.3.2' Staff Recommendations on Inspection Schedules 5.4 5.3.2.1 Inspection Schedule for IGSCC Category A Weldments 5.4

5. 3. 2. 2 Inspection Schedule for IGSCC Category B Weldments 5.4

, 5.3.2.3 Inspection Schedule for IGSCC Category C Weldments 5.5

/ 5.3.2.4 Inspection Schedule for IGSCC Category D Weldments 5.5 5.3.2.5 Inspection Schedule for IGSCC Category E Weldments 5.5 ,

5.3.2.6 Inspection Schedule for IGSCC Category F Weldments 5.5 l 5.3.2.7 Inspection Schedule for IGSCC Category G Weldments- 5.5 l 5.3.3 Inspection Schedules with HWC 5.5 5.3.4 Sample Expansion Method 5.6 6.0 Leak Detection 6.1 6.1 Staff Recommendations on Leak Detection Appendix A - Crack Growth Calculations A.1 i

)

. _--_ _-______ _ ___-_ D

4 EXECUTIVE

SUMMARY

This revision to NUREG-0313, Rev. I " Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary Piping" provides the technical bases for the staff's recommendations regarding actions that can 3 be taken to ensure that the integrity and reliability of BWR piping will be {

maintained. )

The staff long range plan regarding BWR pipe cracks was presented to the Commission in SECY 84-301. A major task in this plan was to revise NUREG-0313 to include the recommendations of the Piping Review Committee i Task Group on Pipe Cracking, issued as NUREG-1061, Vol. 1. l The subjects covered by this revision include recommendations regarding piping and weld material, special processing to minimize crack suscepti- {

bility, improvements in BWR primary coolant chemistry and control, l inspection requirements, repair methods, and leak detection. These recommendations and conclusions are consistent with those made in NUREG-1061, Vol. 1, and are summarized as follows:

BWR piping weldments made of austenitic stainless steel are susceptible to j intergranular stress corrosion cracking (IGSCC). The three elements that, in combination, cause IGSCC are, a susceptible (sensitized) material, a significant tensile stress, and an aggressive environment.

The staff technical recommendation is that improvements in all three of these elements should be pursued. Nevertheless, significant reduction in the  !

probability of I9 SCC can be accomplished even by improving one or two of i these three elements. From a practical standpoint, this is more readily l accomplished in the near term, and can provide acceptable assurance of continued integrity and reliability.

There is no practical way to reduce the sensitization of weldments already installed, so the only way to reduce the susceptibility of the material is to replace the piping with material that is resistant to sensitization )

by welding. Solution heat treatment of individual spool pieces in the pipe fabrication shop before field erection is practicable, and is recommended. Austenitic materials considered by the staff to be adequately  ;

resistant to sensitization by welding are the following:

(1) Low carbon wrought austenitic stainless steel. These include 304L, 304NG, 316L, 316NG, 347NG, and similar types.

(2) Low carbon weld metal of type 308L and similar grades with a minimum of 8% ferrite as deposited. This may also be used as a cladding on the inside of the pipe.

(3) Cast austenitic stainless steel with less than 0.035% carbon and a minimum of 8% ferrite.

1 1

7

?

l C' l (4) Other materials such as nickel base alloys, etc. may be sufficiently L resistant, and may be evaluated in special cases. Inconel 82 is the only nickel base weld metal considered to be. resistant.  !

Service-induced stresses on most BWR piping are relatively low. The source of the high stress primarily responsible for IGSCC is the high tensile stress on the inside of the pipe caused by normal welding practice.

Stress Improvement (SI) can be accomplished on weldments already installed by the Induction Heating Stress Improvement (IHSI) process, or by the

l. Mechanical Stress Improvement Process (MSIP).

SI can be applied to new or replaced piping, or can be applied at any time I during plant life. The staff strongly recommends that SI be applied on ,

all new or replacement piping, and preferably within two years for piping '

already installed. For piping with more than 2 years of operation, SI is considered to be less effective, because cracking may already be present.

.BWR primary coolant normally contains oxygen from radiolytic dissociation of water, and also contains other impurities such as chlorides, carbonates, and sulfur species. If the oxygen levels are reduced by using hydrogen injection, and other impurities are kept to very low levels, IGSCC of even sensitized material will be drastically reduced. This combination of water chemistry improvement is referred to as Hydrogen Water Chemistry (HWC). The staff recommends that HWC be implemented as soon as the practical and safety aspects have been worked out. ,

Some utilities have decided not to replace piping at this time. The staff has developed guidelines for intarim actions that should be taken in these instances. Augmented inspection schedules for susceptible and repaired weldments are based on judgment regarding the probability that significant cracks or leaks will develop, considering the effectiveness of any repair or mitigative' actions applied.

The staff believes that replacing degraded, susceptible piping with IGSCC resistant materials will provide the highest degree of assurance against future cracking problems. Nevertheless, the staff concludes that if the recommendations provided herein are implemented, adequate levels of piping integrity and reliability can be achieved.

The Staff Positions derived from the recommendations in this Report are

. implemented by Generic Letter 87-X.

vii

fi; l

ACRONYMS ANL Argonne National Laboratory ASME American Society of Mechanical Engineers ASTM American Society.of Testing Materials BWR Boiling Water. Reactors BWROG Boiling Water Reactor Owners Group-CFR Code of Federal Regulation CRC Corrosion Resistant Cladding '

EPRI Electric Power Research Institute GE General-Electric GMAW Gas Metal Arc Welding GTAW Gas Tungsten Arc Welding.

HAZ Heat Affected Zone HSW Heat Sink Welding HWC Hydrogen Water Chemistry IE Office of Inspection and Enforcement.

IGSCC Intergranular Stress Corrosion Cracking IHSI Induction Heating Stress Improvement ISI Inservice Inspection ,

LPHSW Last Past Heat Sink Welding Process MSIP Mechanical Stress Improvement Process NDE Nondestructive Examinations NRC Nuclear Regulatory Commission PNL Pacific Northwest Laboratory PPS Part per Billion PT Penetrant Inspec; ion RHR Residual Heat Removal.

SECY Office of the Secretary of the Commission SHT Solution Heat Treatment SI Stress ~ Improvement UT Ultrasonic Testing l

i l

viii

! 1.

i TECHNICAL REPORT ON MATERIAL SELECTION AND PROCESSING GUIDELINES {

FOR BWR COOLANT PRESSURE BOUNDARY PIPING '

1. 0 INTRODUCTION 1,1 History The subject of intergranular stress corrosion cracking (IGSCC) at welds I in boiling water reactor (BWR) piping has been of continuous concern for almost 20 years. An ever-increasing amount of research and developmental activity related to understanding the causes of the cracking and ways to prevent it has been going on during this time period. Under the auspices of NRC,-two Pipe Crack Study Groups have reviewed the problem in BWRs--one

. in 1975 and the other in 1979. Reports of the findings of these groups were published (NUREG-75/067 and NUREG-0531), and staff guidelines prepared to implement their recommendations were published as NUREG-0313 entitled " Technical Report on Material Selection and Processing Guide-  ;

lines for BWR Coolant Pressure Boundary Piping," and NUREG-0313, '

Revision 1.

Until recently, significant cracking of large-diameter piping (12-in.

diameter and large-) was considered to be relatively unlikely, and even if it occurred it was expected that cracks would remain shall av. In Japan some cracks had been detected in the 12-in-diameter recirculation riser pipes. .Because of this, NUREG-0313, Rev. 1 recommended that augmented inservice inspection (ISI) on a sampling basis be performed for these pipes.

Shallow cracking was discovered in pipes larger than 12-in-diameter in Germany, but it was not clear that either the Japanese or German experience was relevant to plants built in the United States.

During a hydrostatic test in March 1982, slight leakage was detected at two of the furnace-sensitized recirculation safe ends at Nine Mile Point.

When these safe ends had been examined ultrasonically 9 months earlier, no cracking was reported. Additional ultrasonic testing (UT) using more sensitive procedures disclosed cracks at many of the 28-in-diameter recirculation piping welds.

This finding was important for two reasons:

(1) It could no longer be believed that large pipes were relatively immune to significant cracking.

(2) It cast doubt on the adequacy of the UT procedures used at that time to detect cracks in large pipes.

IE Bulletin 82-03 was issued to specify augmented inspections of large piping in the recirculation systems of plants (9 units) with outages scheduled in late 1982 and spring 1983. It also specified that

- 1.1 -

4 4

inspection teams demonstrate that they could detect and properly identify cracks in large-diameter pipe welds. IE Bulletin 83-02 was later issued to require inspections at all other operating BWRs (14 units) with more than 2 years of operating service, and to upgrade the UT performance capa-bility demonstrations required of the inspection teams. Reinspection at the next refueling outage were required by Generic Letter 84-11, which also provided specific guidance regarding flaw evaluation and repair for interim operation.

The results of these inspections varied greatly from plant to plant.

Some found very little, if any, cracking. Others found very significant cracking in a large percentage of the recirculation, residual heat removal (RHR) system, and reactor water cleanup system piping welds.

The discovery of si0 nificant cracking in the large-diameter piping, the development of ASME Code procedures for evaluating flaws in such piping, and results of further development of materials and processes to mitigate or prevent IGSCC led to the decision to revise NUREG-0313.

1.2 Revision 1 of NUREG-0313 NUREG-0313 was revised in 1980 to provide guidance and recommendations regarding materials and processes that could be used to minimize IGSCC and to provide recommendations about augmentation of the extent and frequency of IS' on welds considered to be susceptible to IGSCC.

Revision 1 also provided recommendations about upgrading leak detection systems and leakage limits for plants with susceptible welds.

1.3 Revision 2 of NUREG-0333 This present (second) revision updates these recommendations and adds several subjects:

(1) It provides guidance for performing ASME Code,Section XI, IWB 3600, calculations for flaw evaluation.

(2) It provides recommendations regarding repair of cracked piping.

(3) It recommends formal performance demonstration tests for UT examiners, such as those prescribed by IE Bulletins 82-03 and 83-02, and currently being conducted under the NDE Coordination Plan, agreed upon by NRC, EPRI, and the BWROG. This will provide additional assurance that inspections for IGSCC in BWR piping will be performed in an effective manner.

The approach used in previous editions of NUREG-0313 to identify welds that require augmented inspection is simplified, but is expanded to include consideration of reinspection of welds found to be cracked, with or without repair or mitigation actions. The current approach is based on the following:

- 1.2 -

-4 (1) All stainless steel welds in high-temperature BWR systems are considered to be subject'to IGSCC.unless measures have been taken to make them resistant.

(2) The frequency and sample size used to inspect all safety related piping

-welds-in BWR plants will depend on the material and processing used. Simple bases are provided for such classification.

(3) .Some utilities may choose not to replace, or to operate for some interim period of time before making major modifications or replacing piping. This would mean that operation with cracked or repaired welds.will be desired. Guidance is provided to cover these situations.

1.4. Bases for Recommendations Extensive work sponsored by industry through the Electric Power Research Institute (EPRI), General Electric (GE), and the U.S. Nuclear Regulatory Commission (NRC) has been carried out since the second Pipe Crack Study Group reported in 1978-1979 (NUREG-0531). It is not the objective of this report to cover.this work in detail. NUREG-1061, Vol. I was prepared by the Pipe Crack Task Group of the Piping Review Committee. It represents an in-depth discussion af the technical aspects of IGSCC in BWR piping, and provides recommendations regarding materials and processes available to' mitigate or eliminate the problem. It also inc,udes a discussion of the technical basis for che guidelines for interim operation used by the staff.

This revision is based primarily on the information presented in NUREG-1061, as modified by more recent advances in ultrasonic testing and fracture mechanics evaluation methods. It also takes cognizance of work in progress related to serviceability of cracked pipes reinforced by weld overlay or mitigated by IHSI being performed at General Electric and PNL under EPRI and the BWROG sponsorship, and related work at ANL funded by the NRC, as well as public comments received on NUREG-1061, Vol. 1.

1.5 Piping Replacement As stated in the staff paper to the Commission (SECY-84-301), it is the staff's long range goal to bring all affected plants in line with regulations without undue reliance on augmented inspections. Therefore, it is desired that utilities with degraded piping systems replace such piping with IGSCC resistant material as soon as practicable.

Procedural guidance regarding pipe replacement licensing activities is provided in Generic Letter 84-07, dated March 14, 1984.

- 1.3 -

l w-__ _ _ _ . - _ _ _ _ _ _ _ _ _ - _ _ _ . _ _ _ ___ _- __ _ _ _ _ . _ _

l 2.0 METHODS TO REDUCE OR ELIMINATE IGSCC There are three primary ways to minimize the occurrence of IGSCC in BWR I

piping:

(1) Use material that is not subject to sensitization by welding, or solution heat treat after welding.

(2) Use processes that reduce the tensile stress. level at the inner surface of the pipe near the weld.

(3) Modify the BWR water chemistry to control the levels of oxygen and other aggressive contaminants to very low levels.

Each of these three basic approaches are discussed below, and recommendations l H regarding each are presented.

2.1 Materials for New or Replacement Piping Sensitization involves carbon diffusion out of solution forming carbides at grain boundaries upon moderate heating; therefore, reducing the carbon content of the material will result in reduci..g the degree of sensitization resulting from a given thermal exposure, assuming that other factors remain equal.

However, because the susceptibility of an austenitic stainlest. steel is also affected by other variables, such e.s grain size, previous heat treatment, amount of cold work, trace impurities, and overall compositional balance, complete dependence on reduced carbon content may not be effective unless the carbon level is very low.

Nevertheless, a high degree of protection against IGSCC will result if the carbon content is kept below 0.035%, as specified for type 304L ,

grade material. Freedom from sensitization will be much more certain if the carbon levels are controlled to even lower levels.

If carbon is limited to very low levels (such as below 0.02%), the strengthening effect of the carbon is lost, and the material has lower strength, which results in lower Code-allowable stresses. Some heats of type 304L material will also have strength levels too low to meet the minimum specified strength level for standard type 304. Therefore, the replacement of piping with low carbon grades may require redesigning or using thicker wall pipes.

Industry has overcome these problems by developing special grades of austenitic stainless steel. Carbon content is kept very low, and the reduction in strength is compensated for by adding controlled amounts of nitrogen. Molybdenum is often added; it enhances strength and .

f resistance to sensitization,

- 2.1 -

I

J' i

l The grades of austenitic stainless steels developed for increased resistance ]

to sensitization are listed below. i Series 300 stainless steels developed for l l . increased resistance to sensitization ,{

C %,

Steel max. Cr % Ni % Mo % N%

304L 0.035 18.0-20.0 8-10.5 -- --

304NG* 0.02 18.0-20.0 8-12.0 --

0.06-0.10 316L 0.035 16.0-18.0 10.0-14.0 2.0-3.0 --

316NG* 0.02 16.0-18.0 10.0-14.0 2.0-3.0 0.06-0.10 347NG 0.03 17.0-19.0 9.0-13.0

• 304NG and 316NG were formerly called 304K and 316K, respectively.

• • Minimum Nb + Ta = 10 x %C.

Weld metal 'with low carbon and controlled ferrite (sts h as 308L with 8% minimum ferrite) is resistant to sensitization and IGSCC. This resistance is also somewhat dependent on the microstructure produced by the specific welding process used. Weld passes diluted with high carboa base material will not have suitable resistance.

Cast austenitic stainless steel with low carbon and high ferrite content is also resistant to sensitization and IGSCC.

Other common materials such as carbon steels are suitable for many BWR piping systems and are immune to the problem of sensitization and resultant IGSCC. Higher strength alloy steels are less desirable; they may be subject to other types of cracking.

2.1.1 Staff Recommendations on Materials The materials considered resistant to sensitization and IGSCC in BWR piping systems are:

(1) Low carbon wrought austenitic stainless steel, which includes types 304L, 304NG, 316NG and similar low carbon grades with a maximum carbon content of 0.035%. Type 347, as modified for nuclear use, will be resistant with somewhat higher carbon content, the usual maximum of 0.04% is adequate. These materials are generally tested for resistance to sensitization in accordance with ASME A262-A or

-El or equivalent standard.

- 2.2 -

' ):

(2) Low carbon weld metal, including types 308L, 316L, 309L and similar grades, with a maximum carbon content of 0.035% and a minimum of 7.5% (or FN) ferrite as deposited. Low carbon weld filler material especially developed for joining modified type 347 is also resistant as deposited.

Welds joining resistant material that meet the ASME Boiler and Pressure Vessel Code requirement 5% (or FN) ferrite, but are below 7.5% (or FN) may be sufficiently resistant, depending on carbon content and other factors. These will be evaluated on an individual case basis.

(3) Piping weldments are considered resistant to IGSCC if the weld heat affected zone on the inside of the pipe is protected by a cladding of resistant weld metal. This is often referred to as corrosion resistant cladding (CRC).

(4) Cast austenitic stainless steel with a maximum of 0.035% carbon and a minimum of 7.5% (or FN) ferrite. Weld joints between resistant piping and cast valve or pump bodies that do not meet these requirements are considered to be special cases, and are covercJ in the Staff Position on Inspection Schedules below.

(5) . / ustenitic stainless steel piping that does not meet the requirements of (1) above is considered to be resistant if it is given a solution heat treatment after welding.

(6) Other austentic materials, including nickel base alloys such as Inconel 600, will be evaluated on an individual case basis.

-Inconel 82 is the only commonly used nickel base weld considered to be resistant.

The staff recommends that no austenitic material be considered to be resistant to cracking in the presence of a crevice, such as formed by a partial penetration weld, where the crevice is exposed to reactor '

coolant.

2.2 Processes for New, Replacement, or Older Piping l

Special or controlled processing during or after fabrication can provide protection from IGSCC in three ways:

(1) removing sensitization, (2) preventing sensitization, and (3) providing favorable state of residual stress.

There are several special processes that have proved effective in one or more of these ways; they are discussed below:

- 2.3 -

4

[

Solution Heat Treatment The normal metallurgical treatment used to ensure freedom from sensitization is to perform a complete solution heat treatment (SHT) to the piece after welding 9r other processing. It consists of heating the material to a high enough temperature.to dissolve all carbides, then cooling fast enough to retain the carbon in solution. Standard specifi-cations are used to control the process; the chief concern is providing fast cooling.

Note that the solution heat treatment must be performed after welding, and complex piping sections may be difficult to cool fast enough from the solution temperature. Interiors of long or complex piping runs may pose a particular problem.

To be effective, solution heat treatments must be performed in accordance with written procedures that have been proven to be effective for the size and geometry of the piece, and must be'in accordance with applicable specifications.

Heat Sink Welding Heat sink welding (HSW) is a term applied to a method of butt welding pipes or fittings in which the major portion of the weld is p oduced with cooling water inside the pipe. The cooling effect of the water minimizes the sensitization caused by the welding process, and in addition, produces a steep temperature gradient through the pipe wall during welding. This steep temperature gradient causes tensile thermal stresses on the inside of the pipe to exceed the yield strength of the material. After the welding i; completed and the weldment is cooled, the inner portion of the weld is under high compressive residual stress.

This is the opposite of what is caused by normal welding. The high compressive stresses are maintained through about half the wall thickness.

The combination of reduced sensitization and high beneficial residual stresses provides significant resistance to IGSCC.

- 2.4 -

t-1 4

Stress Improvement Processes One of the major sources of stress causing IGSCC is the residual tensile stress that remains on.the inside of the weld joint after the normal butt welding process. Processes have been developed that effectively reverse this residual stress distribution, and actual pipe tests have shown that this is very effective in inhibiting IGSCC in sensitized welds that have been treated by a Stress Improvement Process (SIP).

There are two such processes that are considered fully qualified to provide this mitigation.

Induction Heating Stress Improvement (IHSI)

Induction heating stress improvement (IHSI) is a process originally developed in Japan for treating piping weldments already fabricated or installed in a plant. It consists of heating the outside of the pipe by induction coils to controlled temperatures (~800'F) while cooling water is circulated inside the pipe. The high gradients produce the same effect as HSW. The inside of the pipe is plastica 11y strained in tension during the process, causing residual compressive stresses after the process is completed.

Mechanical Stress Improvement Process The Mechanical Stress Improvement Process (MSIP) is a later development that uses a hydraulic system to uniformly compress the entire pipe at a location near the well-joint. It also causes slight plastic strain, and the residual stresses remaining after the treatment are compressive in the location susceptible to IGSCC because of weld sensitization.

Last Pass Heat Sink Welding The last pass heat sink welding (LPHSW) process is similar to HSW, except that only the last welding passes are performed when there is cooling water inside the pipe. Although some preliminary tests appear promising, it cannot be considered to be fully effective at this time.

2.2.1 Staff Recommendations on Processes The processer considered to be fully qualified for providing resistance to IGSCC in BWR piping welds are:

(1) Solution Heat Treatment (SHT)

(2) Heat Sink Welding (HSW)

(3) Induction Heating Stress Improvement (IHSI)

(4) Mechanical Stress Improvement Process (MSIP) i

- 2.5 -

Although last pass heat sink welding (LPHSW) is not considered to be fully qualified, specific cases may be evaluated individually.

2.3 Water Chemistry Modifications Intergranular stress corrosion cracking of sensitized and stressed stainless steel requires a corrosive environment. Although BWR reactor coolant is comparatively pure water, the small amounts of impurities-usually present are enough to cause IGSCC. These impurities fall into two general classes; those that increase the oxidizing potential, and those that increase the electrical conductivity of the water. Both must be reduced to very low levels to achieve an electrochemical potential below which IGSCC cannot be initiated or propagated.

Oxygen is formed in the core of light water reactors by the disassociation of water by radiolysis. This reaction can be. inhibited by the addition of hydrogen to the water, as is done in pressurized water reactors. Until i recently, this was not considered to be feasible in boiling water reactors, therefore, the normal oxygen content of BWR reactor water is about 200 parts per billion (PPB), providing an oxidizing environment conducive to IGSCC in the entire BWR primary system.

Efforts to find ways to reduce the oxygen levels in BWRs led to the develop-ment of a hydrogen addition methodology that appears to be effective and practicable. Tests conducted in the Dresden 2 plait over the past several years indicate that oxygen levels can be reduced to levels of 10 to 20 PPB, although occasional excursions to higher levels may occur. Tests indicate that IGSCC will not occur at an oxygen level of 20 PPB or less, if other contaminants are controlled to keep conductivity low.

Contaminants that increase the conductivity of the reactor water can come from several sources, such as condenser leakage, resin beds, etc. They include chlorides, carbonates, and sulfur species. Because the electro-chemical potential causing IGSCC depends on both the oxidizing state and the conductivity of the water, the conductivity must be held to very low levels. Laboratery tests have indicated that conductivity levels should be kept to a maximum of 0.3 micro-Siemens (pS) per centimeter with oxygen at 20 PPB or less to prevent IGSCC. Although the tests in Dresden 2 indicated that such conductivity levels could be attained, occasional excursions must be anticipated, and plant to plant variations are likely to be significant in this regard.

This combination of oxygen and conductivity control is commonly referred to as Hydrogen Water Chemistry, or HWC. Although tests have shown that ,

HWC can inhibit IGSCC, some questions regarding radiation effects, fuel  ;

I

- 2.6 -

performance, etc. are still being resolved. Field. implementation and engineering are being actively pursued by the industry, and it is expected that within the next few years, HWC will be considered a practical method of control.

2.3.1 Staff Recommendation on Water Chemistry The use of hydrogen water chemistry, together with stringent controls on conductivity, will inhibit the' initiation and growth of IGSCC. However, the responses of.BWRs to hydrogen injection differs from plant to plant, and the development and verification of a generic HWC specification is not yet complete. For these reasons, reduction in piping inspection frequency based on the use of HWC will be considered on a individual case basis at the present time. If fully effective HWC is maintained, a factor of two in reduction of inspection frequency may be justified for susceptible weldments.

- 2.7 -

3.0 EVALUATION AND REPAIR OF CRACKED WELDMENTS When cracks are found in BWR piping, severa1' alternatives (and combinations) are available to provide assurance of further safe opera-tion of affected welds. ,

If the cracking is not too severe, the rules of ASME Code Section XI, IWB 3600 (as modified and expanded in Section 8) may be used for short-term interim operation. Further, SI may be applied to reduce the probability of further crack growth.

If the cracking is too severe to meet these rules, the affected piping must be repaired or replaced before the plant can be returned to service.

3.1 Repair Procedures IGSCC in BWR piping initiates at the inner surface of the pipe and grows progressively through the wall toward the outside. It commonly initiates near the weld root and progresses up the heat-affected zone (HAZ) close to the weld, and sometimes in the weld. Therefore, cracking can affect a region of the pipe longer in axial extent than the maximum width of th; weld if cracks occur on both sides of the weld. The usual repair process during construction is to grind out the defective area and fill the area with weld metal. Thit is not practical for repair of IGSCC, because IGSCC starts from the inside surface, requiring removal of essentially the entire weld and HAZ area.

There are several repair methods available for at least short-term operation:

(1) Weld overlay reinforcement (2) Partial replacement (3) SI (for minor cracks)

(4) Approved clamping devices These are discussed below.  :

3.1.1 Weld Overlay Reinforcement >

l Weld overlay reinforcement consists of applying weld metal over the weld and for a specified minimum distance beyond the weld on both sides. This  :

is done completely around the outside surface of the pipe overlapping each 1 pass. IGSCC-resistant low-carbon, high-ferrite type 308L weld metal is used, and the process is usually performed with an automatic welding machine using the Gas Tungsten arc (GTAW) or Gas Metal arc (GMAW) processes. Weld over-lay is performed with cooling water in the pipe during welding, and there is no need to drain the pipe during repair. More specific design details and quality control recommendations are covered in Section 4.0.

- 3.1 -

l l

i i

1 3.1.2 Partial Replacement j A very effective repair method is to cut out a section of the pipe I containing the defective weldments and to weld in another piece of pipe.

The major drawback to this approach is that the affected run of pipe must j be drained and dried. Either all fuel must be removed from the reactor l vessel or special plugs must be installed when this type of repair is used in portions of piping that cannot be isolated.

If this method can be used, a fully effective repair can be made with resistant material, using welding processes such as heat sink welding for the new installation welds and high-ferrite type 308L weld material.

SI can also be applied.

Another disadvantage of this process (assuming that draining is feasible) is that high radiation exposures to workers may be encountered at older plants from the inner surfaces of the pipes. Prior decontamina-tion can alleviate this problem.

Both weld overlay and partial replacement cause the pipe to shrink in the axial direction. If several such repairs are made in one length of pipe, additional stresses will b. introduced by this shrinkage, which must be taken into account in the stress analysis required for the repair, and in the fracture mechanics analyses of crack growth in other welds of the pipe system. Measu ements of shrinkage on weld procedure qualifica-tion test pieces can provide guidance regarding how much shrinkage can be expected. Actual measurements made during the repair should be used in the final stress analysis.

3.1.3 Stress Improvement As discussed above, SI alters the residual stress pattern, putting the inner part of the pipe wall in compression, thus inhibiting crack initiation.

If cracks are present, the situation is more complex. If cracks are shallow, the process will probably prevent further growth, as long as the residual stress pattern remains favorable. The process may stretch cracks open, but tests have shown that they are not extended in depth by the process. Such stretching may even be beneficial for shallow cracks because it enhances the resulting compressive stress around the crack tip.

The tips of deeper cracks, particularly those penetrating deeper than half way through the pipe wall, are likely to be in a general tensile stress field after SI processes. This could cause such cracks to propagate through the wall faster than they would without the SI treatment. Cracks will not be expected to grow longer, because of the beneficial residual stress on the inside portion of the pipe. Therefore, neither short cracks of medium depth, or longer shallow cracks are expected to grow to a significant size after an SI treatment.

- 3.2 -

1 3.1.4 Mechanical Clamping Devices ]

Another approach to reinforcing a cracked weldment is to use a mechanical clamp. One advantage of this approach is that the clamp may be periodically l removed for weld examination. Such clamping devices will be reviewed for adequacy of mechanical design, materials of construction, and installation methods on a case basis.

3. 2 Staff Recommendations on Repairs 3.2.1 Staff Recommendations on Weld Overlay Reinforcement Weld overlay reinforcement made in accordance with recomme7dations described in this report are considered to be acceptable at least for short-term operation. Weld overlay may be considered for longer term operation provided:

j (1) The overlays are in conformance with the criteria of Sectica 4.0 of this report; and (2) they are inspected in accordance with the criteria of Section 5.0 by UT examiners and proc:Jures qualified to inspect overlayed welds.

Weld overlt.ys not meeting (1) above ma3 be reinforced to the extent necessary to meet the staff position, if desired.

3.2.2 Staff Recommendations on Partial Replacement Repair of cracked weldments by partial replacement can be considered to be fully effective if appropriate materials and weld processes are used, and therefore are considered to be resistant to IGSCC.

3.2.3 Staff Recommendations on SI of Cracked Weldments SI may be considered as a partial mitigation process when applied to weldments with short or shallow cracks. Details of allowable crack sizes in this regard are covered in the next section. Note that SI.is only considered effective if it is followed by a qualified UT examination, and if cracks are found they must be sized, both in depth and length, by procedures and personnel qualified to perform sizing examinations according to recommendations l given in Section 5.1 of this report.

3.2.4 Staff Recommendations on Clamping Devices Clamping devices may be used for temporary reinforcement of cracked weldments.

Each case must be reviewed and approved on an individual basis. '

- 3.3 -

l b

- 1 1

1 4.0 CRACK CHARACTERIZATION AND REPAIR CRITERIA 4.1 Flaw and Repair Evaluation Criteria This section provides guidance and staff positions regarding methods to evaluate IGSCC cracks for limited further operation. It also covers evaluation methods and acceptance criteria for repairs if immediate pipe replacement is.not practicable.

The methods and criteria described in this section are generally in accordance with IWB 3640 of Section XI of the ASME Boiler and Pressure Vessel Code. In particular, IWB 3642 provides for flaw evaluation using fracture mechanics or other applicable methods. The C. requires that-crack growth be calculated, and the flawed joint is acceptable for further operation only for the time period that the flaw remains small enough that the Code-ir, tended safety or desi3n margins are maintained.

In IWB 3641, the Code (Winter 83 Addenda) provided simple tables of allowable crack depth as a function of the primary stress level and crack length.

These tables are based on limit load calculations, and assume that the material is tough. An overail margin of about 2.77 agai.ist net section collapse (limit load) failure mode is factored into the tables.

-It was recognized that ttese tables did not provide an acceptable level of margin against failure for low toughness materials such as fluxed welds (SAW, SMAW). This is because low toughness material may fail at load levels below limit load, and secondary stresses (not considered in the original IWB 3641 tables) may also contribute to failure of low toughness materials.

This problem has.now been addressed by the Code, and the 1986 Edition provides appropriate criteria for all types of welds.

- 4.1 -

Er l

4.2 Crack Growth Calculations i The rate of growth of cracks by IGSCC has been the subject of discussion and controversy for many y?ars. Part of the problem is that the rate of growth as a function of stress.is affected by the degree of-sensitization of the material and the severit) of the environment. A further complication has been that ways to measure the degree of sensitization have proved to be inaccurate or not ,

relevant to the particular problem of BWR piping. For these reasons, many I crack growth tests h e been performed that were either too severe or not severe enough. The staff recommends a crack growth rate curve that is believed'to be near the upper bound for weld-sensitized material in actual BWR environments. (See Appendix A)

Crack growth by IGSCC appears to follow a classical trend. If the logarithm of the growth rate is plotted against the logarithm of severity of

. loading, measured by the stress intensity factor (a fracture mechanics parameter) K , a. linear relationship is found. As the K changes with crackgrowthfiterativecalculationswilltrackthegrowkhofthecrack with time. The calculational procedures recommended by the staff to predict crack growth are detailed in Appendix A.

Actual circumferential cracks in welds are usually very long in relatia to'~

their depth; therefore, crack growth in a congruent manner (maintaining the same shape) cannot be assumed, particularly for large-diameter pipes.

The growtn in the length direction, therefore, may be more than in the depth direction. Specifically, the growth along the length should be assumed to increase the aspect ratio (length to depth) by the same factor that the depth is increased. For example, if a crack with an aspect ratio of 3 to 1 grows to twice the original depth, the new length will be assumed to give an aspect ratio of 6 to 1. Cracks with aspect ratios over 20 to 1 are assumed not to change shape with crack growth.

Although axially oriented cracks are not likely to grow significantly beyond the sensitized zones on each side of the weld, they will grow through the weld if the weld metal is marginal in resistance to sensitization, and therefore was sensitized during welding. Axial cracks will therefore be assumed to grow through the wall but the length is limited to 1.5 times the thickness of the pipe.

4.3 Multiple'and Complex Crack Characterization Case 1 If Multiple cracks are present that will remain less than 20% of the circumference in total length after crack growth, they may be treated as one crack with the length equal to the sum of the lengths.

- 4.2 -

Case 2 If multiple cracks are present that will remain less than 30% of the circumference in total length after crack growth, they may be treated as one crack with the length equal to the sum of the lengths, provided that after crack growth each crack is. separated by at least 20% of the circumference from all other cracks.

Case 3 All other situations regarding multiple cracks will be considered as a single 360 crack.

Case 4 Cracks on both sides of the weld will be treated as if they were all on the side of the weld with the thinnest wall; overlapping cracks or overlapping areas are considered as one crack.

4.4 Weld Overlay Design Criteria 4.4.1 Standard Overlay Design The standard overlay should be designed to provide a nominal margin of 2.77 against limit load failure, assuming that the original crack was completely through the wall for 360 . The calculation method described in Section 4.1 is recommended. Because none of the original weld or heat affected zone is considered in the analysis, the stresses to be used in the analysis deperd only on the kind of weld metal used for the overlay.

Specifically, if the overlay is made using GTAW or GMAW processes, secondary stress need not be considered. Calculations are made using ,

the.as-overlayed joint dimensions and stress levels. '

4.4.2 Designed Overlays

]

i In cases where cracks are perpendicular to the weld (axial) or short in the circumferential direction, even a small amount of overlay will pre-vent further growth in the length direction, because high compressive stresses are induced at the inner surface of the pipe. In such cases the overlay will also act to prevent leakage.

Weldments with a total length of circumferential cracking less than approximately 10% of the circumference, with no more than four axial cracks, J are considered appropriate for repair by a designed overlay. A standard j overlay should be used for more severe cracking. H l The thickness of the designed overlay should be at least two layers of I weld metal after the surface has passed surface examination by penetrant i inspection (PT). If credit is taken for the thickness of the first layer, it should be shown by actual. test to contain a minimum of 8% ferrite, and the original surface must have passed PT.

-4.3-I i

l l

l e

l l

l Because designed overlays take credit for part of the original pipe in their design, there are several ways that the lower toughness of the original fluxed weld may be taken into account. An acceptable design approach is to assume that the crack or cracks requiring the overlay are completely through the original pipe wall for the total length of ,

crack involved. The overlay thickness is calculated so that the as-overlayed cracked weldmarit meets the IWB 3641 tables in Section XI of Other approaches to overlay design may be evaluated on a case basis. In general, it is recommended that highly stressed welds should be reinforced with standard overlays.

4.4.3 Limited Service Overlays Overlay designs not meeting the above criteria for either Standard or Designed overlays are only recommended for limited service, such as one fuel cycle of operation. (See 5.3.2.6)

4. 5 SI Crack Mitigation Criteria In general, SI is only recommended for use on weldments with minor c rac ki ng. T. bis is because the tips of deep cracks can be in an area of high tensile stress caused by the process, and further crack growth may even be accelerated by the SI treatment. Because the effectiveness of the SI treatment is also related to the applieJ stress on the weldment, mitication by SI is not recommended for weldments with service stresses over 1.0 S ,

cracks deeper than 30% of the wall, circumferential cracking longer tha$

10% of the circumference, or axial cracks of any extent. (See 5.3.2.6) 1 I

- 4.4 -

il

s-5.0 INSPECTION OF PIPING FOR IGSCC 5.1 Weldments Subject to Inspection The discussion and recommendations in this section apply to BWR piping made of austenitic stainless steel that is four inches or larger in nominal diameter and contains reactor coolant at a temperature above 200 F during power operation regardless of code classification. It also applies to reactor vessel attachments and appurtenances such as jet pump instrumentation penetration assemblies and head spray and vent components.

This section does not apply to piping made of carbon steel classified as P-1 by the ASME Boiler and Pressure Vessel Code.

5.2 Inspection Methods One positive result of the extensive investigations performed on BWR piping is that no significant mode of degradation other than IGSCC has been noted.

This means that inspections can focus on those approaches that are best suited for detecting and evaluating IGSCC. A less favorable finding is that special methods and specific operator training are required to reliably detect and characterize IGSCC in the presence of the variable geometric configurations of the weldments.

- 5.1 -

1 l

\ i l

_ - - - - - --- _-__ m

t.

It is not the intent of this report to provide specific guidance to operators regarding details of equipment and procedures. This function is best handled by Code activities in which industry and regulatory participants reach a consensus. It is not a simple problem; finding and recognizing IGSCC by UT is still as much an art as it is a science.

The intent of the recommendations in this report is to ensure that the UT operators inspecting BWR piping for IGSCC can detect and characterize IGSCC in the welds they inspect, and that they will accomplish these two functions reliably in the field.

5.2.1 Staff Recommendations on Inspection Methods and Personnel Although examinations should be performed in general accordance with the ultrasonic examination requirements of the applicable edition of the ASME Code, details of the examination method, acceptance criteria, and personnel qualification should be upgraded to ensure that the examinations will be effective.

I All examination procedures and the specific equipment used in the field inspections, and all' level 2 and 3 NDE examiners or operators for flaw detection and sizing should demonstrate their field performance capability on cracked, preferably service-induced, samples in a manner acceptable to the NRC. ' o NDE examiner or operator should perterm examinations of BWR piping without proving his competence even if he must take special training to gain specific skills and krowledge required to perform these inspections. The program being conducted at EPRI NDE Center in Charlotte, North Carolina, in accordance with the NDE Coordination Plan agreed upon ay NRC, EPRI, and BWROG, as upgraded in September, 1985 is considered to be acceptable. Any future r.hanges in this program should be in conformance with the Coordination Plan and approved by the Executive Dire: tor for Operations, NRC.

Specialized radiographic techniques developed for detection of IGSCC may be used in cases where ultrasonic examination is not practical, or to augment the UT method.

5.2.2 Flaw Size Uncertainty j l

Inspections performed under IE Bulletins 82-03 and 83-02 were ofter:

performed by examiners with limited knowledge and experience in sizing IGSCC. Although the length of the cracks could usually be defined satisfactorily, most UT operators could not determine their throughwall depth accurately and reliably. After this was shown to be true in industry-wide evaluation projects, the industry developed more effective and diverse techniques, and the NDE Center initiated a

- 5.2 -

I

4 training and qualification program specifically for crack depth sizing. The NRC staff participated in this effort by defining acceptable levels of performance, based on the level of accuracy required to ensure safe operation. The staff now believes that flaw sizes determined by examiners and procedures qualified by test will L not be grossly underestimated or overestimated provided that an inspectable weld joint configuration and weld surface exist.

The depth of cracks not sized by fully qualified personnel or with limitations to examination (such as wide weld crowns, obstructions, or other adverse geometrical configurations) should be assumed to be at least 75% of the wall in depth, and the f. law so evaluated.

5.3 Inspection Frequency 5.3.1 Weldment IGSCC Condition Category Definitions The purpose of inservice inspection of piping is to provide continued assurance that the structural integrity and reliability (e.g., see 10 CFR 50.55a(g)(6)(ii)) of the piping is maintained, and that there continues to be an extremely low probability of abnormal leakage (10 CFR 50, Appendix A, Criterion 14). Piping with weldments that are susceptible to degradation mechanisms such as IGSCC require more frequent inspections to provide such continued a,surance. Weldments in BWRs will have differeat degrees of susceptibility to IGSCC depending on the mate-rials and processing involved. Therefore, the inspection frequencies recommended by the staff are based on the condition of each weldment.

The extent of augmented inspection recommended depends on the number of cracked welds in the plant as well as the condition of each individual weldment. In addition, welds that have already been found to be cracked will have varying degrees of susceptibility to further cracking, depending on the remedial actions taken.

Some may be considered repaired, at least on a conditional basis; whereas others with marginal or no repair are considered fit for only very limited service without additional action. These seven categories of weldment l conditions are listed in Table 1 and defined in detail below.

~

5.3.1.1 Definition of IGSCC Category A Weldments IGSCC Category A Weldments are those with no known cracks, that have a low probability of incurring IGSCC problems, because they are made j entirely of IGSCC resistant materials or have been solution heat treated after welding. CRC is considered to be IGSCC resistant, and welds j joining cast pump and valve bodies to resistant piping are considered to l be resistant weldments. 1

- 5.3 -

4 i

5.3.1.2 Definition of IGSCC Cate' gory B Weldments IGSCC Category B Weldments are those not made of resistant materials, but have had an SI performed either before service or within two years of operation. If the SI is performed after plant operation, a UT examination after SI to ensure that they are not cracked is required.

5.3.1.3 Definition of IGSCC Category C Weldments IGSCC Category C Weldments are those not made of resistant materials' (see 2.1.1), and have been given an SI process after more than two years of operation. An ultrasonic examination to ensure that they are not cracked should be performed after the SI treatment as part of the process.

5.3.1.4 Definition of IGSCC Category D Weldments IGSCC Category D Weldments are those not made with resistant materials, and have not been given an SI treatment, but have been inspected by examiners and procedures in conformance vIth section 5.2.1, and found to be free of cracks.

5.3.1.5 Definition of IGSCC Category E Weldments IGSCC Category E Weldments are those with known cracks but have been reinforced by an acceptable weld overlay or have been mitigated by an SI treatment with subsequent examination by qualified examiners and procedures to verify the extent of cracking. Guidelines for acceptable weld overlay reinforcement and extent of cracking considered amenaole to SI treatment are covered in Sections 3.2 and 4.5 of this document.

5.3.1.6 Definition of IGSCC Category F Weldments IGSCC Category F Weldments are those with known cracks, that have been approved '

by analysis for limited additional service without repair. Weldments found to have significant cracking or a questionable extent of cracking that have been minimally overlay reinforced (not in conformance with Section 4.1) are considered acceptable only for interim operation. Weldments with significant cracking that have been SI treated may also be considered to be in this category. Detailed guidelines used to evaluate specific cases are provided in Sections 3.0 and 4.0 of this document.  ;

- 5.4 -

1

b .

5.3.1.7 Defin'ition of IGSCC Category G Weldments IGSCC Category G Weldments are those not made of resistant materials, have not been given an SI treatment and have not been inspected in accordance with Section 5.2.1. Stress improved welds that were not inspected after the SI treatment are considered to be Category G weldments until the post-SI inspection has been performed.

5.3.2 Staff Recommendations on Inspection Schedules The staff recommendations on the extent and frequency of inspection for various weldments categorized in accordance with 5.3.1 are discussed in detail below and summarized in Table 1.

5.3.2.1 Inspection Schedule for IGSCC Category A Weldments IGSCC Category A welds should be inspected according to a schedule similar to that called for in Section XI of the Code. A representative sample of 25%

of the welds should be examined every 10 year interval. The sample selection should reflect the best technical judgment of the plant owner.

5.3.2.2 Inspection Schedule for IGSCC Category B Weldments IGSCC Category B welds are more likely to develop cracking than Categary A welds, so a larger sample size is needed. Specifically, a representative sample of 50% of IGSCC Category B welds should be examined every 10 year interval.

5.3.2.3 Inspection Schedule for IGSCC Category C Weldments IGSCC Category C welds have longer service life prior to SI than IGSCC Category B welds, so are more likely to contain undetected cracking. All IGSCC Category C welds should be inspected within two refueling cycles af ter the post-SI inspection, and every 10 years thereaf ter.

5.3.2.4 Inspection Schedule for IGSCC Category D Weldments Category D Weldments should be inspected at least once every two refueling cycles. Approximately half of the IGSCC Category D weldments in the plant should be inspected each refueling outage.

5.3.2.5 Inspection Schedule for IGSCC Category E Weldments Repaired and stress improved cracked weldments, IGSCC Category E, should be inspected at least once every two refueling cycles after repair. Approximately l

half of them should be inspected during the first refueling outage after repair.

- 5.5 -

4 If it is desired to operate for more than two fuel cycles with overlay reinforcement repairs, the overlayed weldments should be inspected to ensure that the overlays will continue to provide the necessary safety margin.

For standard and designed overlays meeting the requirements of Section 4.0, i the inspection method should provide positive assurance that cracks have not g progressed into the overlay. It is also desirable that the inspection pro-cedure be capable of detecting cracks that originally were deeper than 75%

of the original wall thickness, or that have grown to be deeper than 75%

of the original wall thickness. Ultrasonic inspections should be performed using a procedure that has been demonstrated to be reliable and effective, and should be performed by personnel that have been trained and qualified in the specific methods for inspections of overlays.

5.3.2.6 Inspection Schedule for IGSCC Category F Weldments l

IGSCC Category F Weldments are approved for limited service only, and should be. inspected every refueling outage, unless a shorter service period has been specified. Weldments that are classified as IGSCC Category F because overlay repairs or SI treatment mitigation is not according to recommendations in Sections 3.2 and 4.5 may be upgraded to IGSSC Category E after 4 successive examinations indicate no adverse change in cracking condition.

5.3.2.7 Inspection Schedule for IGSCC Category G Weldments IGSCC Category G Weldments should be inspected at.the ne> t refueling outage. )

5.3.3 Inspection Schedules with HWC If improved water chemistry control, including hydrogen additions is implemented, the time schedule for inspections may be extended. Although specific details of such extensions will be evaluated on a case basis, it is anticipated that periods between inspections could be lengthened  :

by about a factor of two for category B, C, D and E weldments. l 5.3.4 Staff Recommendations on Sample Expansion If one or more cracked welds in IGSCC Categories A, B, C, and D are found during an inspection, an additional sample of the welds in that l category should be inspected, approximately equal in number to the j original sample. This additional sample should be similar in '

distribution (according to pipe size, system, and location) to the original sample, unless it is determined that there is a technical reason to select a different distribution. If any cracked welds are found in this second sample, all of the welds in that IGSCC Category should be inspected.

- 5.6 -

If significant crack growth, or additional cracks are found during the L inspection of one or more IGSCC Category E welds, all other Category E L welds should be examined.

a) Significant crack growth for overlayed welds' is defined as crack y extension to deeper than'75% of.the original wall thickness, or, E for cracks originally deepter than 75% of the pipe wall, evidence of crack growth into the effective weld overlay, b) Significant crack growth for SI mitigated Category E welds'is L

-defined as growth to a length or depth exceeding the criteria for l SI mitigation. .(10% of circumference or 30% in depth).

- 5.7 -

TABLE 1

SUMMARY

OF RECOMMENDED INSPECTION SCHEDULES FOR BWR PIPING WELDMENTS IGSCC- INSPECTION DESCRIPTION OF WELDMENTS NOTES CATEGORY EXTENT & SCHEDULE Resistant Materials A 25% every 10 years (at least 12% in 6 years)

Nonresistant Matis (1) B 50% every 10 years

, SI within 2 yrs of (at'least 25% in 6 years) operation (1)

Nonresistant Matls (1) C All within the next 2 refueling SI after 2 yrs of cycles, then all every 10 years

, operation Non Resistant Matl (1) D All every 2 refueling cycles No SI Cracked (1)(2) E 50% next refueling outage, then Reinforced by weld overlay all every 2 refueling cycles Or mitigated by SI Cracked (2) F All every refueling outage Inadequate or no repair Non Resistant (3) G All next refueling outage Not Inspected Notes:

(1) All. welds in non-resistant material should be inspected after a stress improvment process as part of the process.- Schedules shown should be followed after this initial inspection.

(2) See recommendations for acceptance. weld overlay reinforcements and stress improvement mitigation.

(3) Welds that are not UT inspectable should be replaced, " sleeved", or local leak detection applied. RT examination or visual inspection for leakage may also be considered.

-5.8-1 1

1

6.0 LEAK DETECTION The staff reviewed the leak detection and leakage limits that have been applied to BWRs by past revisions of NUREG-0313, Bulletins, and Generic Letter 84-11. In NUREG 1061 Vol. 1, " Report of the U.S. Nuclear Regulatory Commission Piping Review Committee," the report of the Pipe Crack Task Group, it was recommended that leakage detection equipment should be improved, and that_ the upper limit on unidentified leakage should be decreased from 5 gpm to 3 gpm.

As a result of this review, the staff concluded that if the.other recommendations of this report are followed, present leak detection systems will be adequate. Further, the staff concluded that the decrease in the limit on unidentified leakage recommended in NUREG-1061 Vol.1, would constitute a backfit that could not be justified by a supporting Regulatory Analysis', in accordance with the new backfit rule, 10 CFR50.109.c.

Accordingly, the staff recommendations on leak detection and leakage limits are in accordance with past staff positions on the subject.

Relaxation of the operability requirements for those plants with resistant or mitigated noncracked piping is also in accordance with past staff positions.

6.1 Staff Recommendations on Leak Detection Leakage detection systems should be in conformance with Position C of Regulatory Guide 1.45 " Reactor Coolant Pressure Boundary Leakage Detection Systems, or as otherwise approved by the NRC.

1. Plant shutdown should be initiated for inspection and corrective action when any leakage detection system indicates, within any period of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or less, an increase in rate of unidentified leakage in excess of 2 gpm of 5 gpm or equivalent, whichever occurs first. For sump level monitoring

. systems with fixed-measurement interval method, the level should be monitored at approximately 4-hour intervals or less.

1

- 6.1 -

2. Unidentified leakage should include all leakage other than .

(a) leakage into closed systems, such as pump seal or valve packing leaks that are captured, flow metered, and conducted to a sump or collecting tank, or (b) leakage into the containment atmosphere from sources that are both specifically located and known either not to interfere with the operations of unidentified leakage monitoring systems or not to be from a throughwall crack in the piping within the reactor coolant pressure boundary.

3. For plants operating with any IGSCC Category D, E, F, or G welds, at least one of the leakage measurement instruments associated with each sump shall be operable, and the outage time for inoperable instruments shall be limited to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, or immediately initiate an orderly shutdown.

l

- 6.2 -

- --- - - __o

4 APPENDIX A - CRACK GROWTH CALCULATIONS Introduction Crack growth calculations are required to evaluate the continued structural integrity of a weld with known cracks, if it is desired to continue' operation without repair or reinforcement. The rate of growth of IGSCC is not easy to predict, because the several important factors are usually imperfectly known.

Research work in this area has been helpful in defining the general effect of these factors, but a large'une inty in crack growth predictions still remain.

Nevertheless, crack growth calct :tions can be performed within certain limits with enough confidence to ensure plant safety without excessive conservatism.

Crack growth calculations are based on the fundamental concept that the crack growth rate of a specific material in a specific environment will be a function of the applied stress intensity factor, K . Laboratory crack growth data are useally presented in this manner. Detailkofthecalculationalmethodsused to calculate K yare provided later in this Appendix, but an important point to note here is that K depends 7

on the crack depth, therefore it changes continuously during crack growth.

Crack growth analysis methods are, therefore, iterative in nature. Given an initial crack depth, the K o! interest. Knowingthek,scalculatedfortheparticularstressdistribution the amount of growth for a specific time {s cal-culated, the growth is addeb,to the initial crack depth, a new K is calculated, and the process is repeated. Timeintervalsselectedcanvaryfhom1hourto 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br />, depending on the rate of growth and rate of change in Ky with crack depth.

Selection of Crack Growth Rate Parameters Although only two parameters, crack growth rate and K7 , are used, they are both highly dependent on several factors.

Crack growth rate is affected by the degree of sensitization of the material and by the severity of the environment. Our interest as it relates to BWR piping is primarily in a degree of sensitization normally caused by welding, and in an environment similar to normal BWR water conditions.

Most formal crack growth studies are carried out with standard fracture mechanics specimens, which makes 7K determination easy. These specimens are not readily machined from pipe welas, so the material is given an artificial sensitization treatment, intended either to simulate the effect of welding or, in some cases, the more severe effect of furnace sensitization. Tests to ascertain whether the intended degree of sensitization has been obtained are still inexact, causing significant scatter in laboratory test results intended to apply tu a similar metallurgical state.

- A.1 - ,

i

_ _ _ _ _ _ _ - - _ _ _ _ -- ___ -- - - . - - _ l

f

'l 4

Tests to simulate the BWR environment are usually run at operating temperature in high purity water containing 0.2 ppm oxygen. This is generally accepted to be a representative condition, although higher oxygen levels could occur locally for short periods of time. Tests are also often run in water con-taining up to 8 ppm oxygen, usually to achieve accelerated comparisons of 1 materials or conditions.

In addition to these standardized tests for crack growth rate, results of actual pipe tests are available. Many hundreds of welds have been tested in General Electric's pipe test facility. These tests, although generally more relevant in terms of material condition and environment, are more difficult to evaluate. K y is more difficult to calculate, and accurate crack growth rates are also more difficult to measure. Nevertheless, this body of data has.been used to augment those data from the more standard laboratory tests, to select appropriate crack growth rates.

Figure 1 (from NUREG/CR-3292)

• shows much of the relevant laboratory data in the conventional form, where measured rates are plotted against K . This plot clearlyshowsthelargescatterresultingfromawidevariationibmaterial condition and environment. This information, together with additional infor-mation from actual pipe tests, was used to select a crack growth curve that is appropriate for use in safety evaluations. Note that if the fastest crack growth rate shown in Figure 1 is used, cracks would be predicted to grow completely through ,ipe walls in a matter of days. Clearly this would not reasonably represent reality.

The curve selected for use by the NRC staff is shown on Figure 2. Note that it is a curved line on the semilogarithmic chart used in Figure 1.

On log-log coordinates, as used in Figure 2, it plots as a straight line.

In calculations, it is expressed as:

da/dt = 3.590 x 10 8 . K inches per hour 7

As can be seen, the crack growth rate is a very strong function of K . In laboratory tests, K yis easily determined with good accuracy. ThisIsnotthe case for real pipes and real pipe cracks. There are two major sources of uncertainty: knowledge of the actual crack size and shape, and the actual stress distribution in the area of the crack to be evaluated. The service distribution at a pipe weld is made up of the stress caused by the service loading and the residual stresses caused by the welding process. Of these, knowledge of the residual stress is the more uncertain. Nevertheless, a residual stress distribution through the pipe wall must be defined, if realistic crack growths are to be calculated. Although this is covered later in more detail, several comments are in order here.

The residual stress distribution caused by welding is the major stress component causing IGSCC. Welding causes a high tensile residual stress on the inside surface of the pipe near the root of the weld where the material is sensitized.

• Shack, W.J., et al., " Environmentally assisted cracking in light water Water Reactors: Annual Report, October 1981 - September 1982" NUREG/CR-3292, Washington, DC: U.S. Nuclear Regulatory Commission, f Februa ry , 1983.

- A.2 -

I i

- _ - _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ - - _ _ b

j r .

l

t. )

I i l l L

This residual stress level has been calculated and measured to be up to or above the yield strength of the material. It typically is four or five times

-as high as the service-induced stress. In fact, without this very high residual stress at the sensitized area, IGSCC would not be a problem in BWR piping.

This fundamental observation is helpful; wherever this combination.of stress I' and sensitization occurs, cracking occurs. In actual cases, if there are significant cracks, there must be significant tensile residual stresses, and this should be accounted for in the crack growth analysis. The method used by the staff is described below. 1 l

Stress Intensity Factor Calculations There are several- relatively standard analytical solutions available for calculating the stress intensity factor.(K 7) caused by stress distributions of the type found at BWR pipe welds. The method using influence functions is the one used by the staff and will be summarized here. Other methods, such as those described in the ASME Boiler and Pressure Vessel Code,Section XI, Appendix A, may also be_used where appropriate.

Stress Analysis The total stress state, including ,esidual stress, pressure stress, and other stresses caused by normal operation must be known or assumed. Note that factors such as stress indices used for other purposes should not ae used .5en calculating stress levels that apply to K calculations. 7 Residual Stress The laboratory-measured throughwall axial residual stresses on pipe wall thickness > 1 inch are presented in Figure 3 (from NUREG/CR-3292). The solid line in Figure 3 is the axial residual stress distribution used for the calculation of stress intensity factors for pipe sizes of 12" diameter and larger. The residual stress distribution is the most complex analytical problem involved. This is handled by fitting the curve of residual stress distribution through the wall by an analytical expression. For this par-ticular residual stress distribution, the nondimensional expression given below is used.

4 o/oq= Ij=o o j (d where og = 1.0 og = -6.910 oz= 8.687 03 = -0.480 aa = -2.027

( = x/t o$ = stress magnitude at ( = 0 (inner surface)

- A.3 -

i

4

'{

The above formula permits calculation of the residual' stress value at any point (x) through the vessel wall thickness (t) as a function of the peak residual stress value at the inside diameter (ID), oj .

-The stress intensity. factor caused by the residual stress from welding (KTR)'

is calculated.using influence functions taken from NUREG CR-3384,

• page A.19, Table (7). The influence functions, i givt n in this Appendix are for a 360 circumferential crack in a cylinddr, with a R/t ratio of 10. In view of other analytical conservatism and uncertainties (i.e., assumed crack geometry and initial depths), it is believed that they may be used for cylinders with R/t ratios of.from 9 to 11 to obtain reasonable and conserva-tive estimates of crack growth versus time. For R/t ratios significantly different from 10, other influence functions or other analytical methods should be used.

• The specific formula used by the staff is:

4 d

K I ij IR/ ( iJ t) = Jna c) a j=o where:

o g ,...og and og are as above io = 1.1220 + 0.3989 a + 1.5778 a + 0.6049 a 3 2

i1 = 0.6830 + 0.1150 a + 0.7556 a + 0.1667 a 3 2

1 2

= 0.5260 + 0.1911 a - 0.1000 a + 0.5802 a3 2

13 = 0.4450 + 0.i)783 a + 0.0556 a + 0.3148 a 3 a 2

i4 = 0.3880 + 0.1150 a - 0.1333 a + 0.3519 a 1 i

a = a/t l a = crack depth 1

t = wall thickness Membrane Stress j The membrane stresses are assumed constant through the wall thickness, so o, = o p where o = membrane stress (o,) from pressure p

• 5tevens, D.L., et al., " VISA-A Computer Code for Predicting the Probability of Reactor Vessel Failure" NUREG/CR-3384 PNL-4774, Washington, DC. U.S. Nuclear Regulatory Commision, September, 1983.

l

- A.4 - l L___-___________-_-_______-____________.

l

The stress intensity factor for a 360* circumferential crack from pressure, Kyp, is calculated from 2 3 K

yp = (PR/2t) J t V na (1.122 + 0.3989 a + 1.5778 a + 0.6049 a )

where

. a, t are as above h P =. pressure i

R = radius to. center of pipe wall The total stress intensity factor, KIT, is given by

'K IT = 'Kyp +. KIR where K

yp and K IR are defined as above.

Correlation with Service Experience Although the residual stress is assumed to be the same for all welds, the applied stresses, primary and secondary, vary from weld to weld; therefore, calculations must be performed for each weld evaluated. Figure 4 shows the results of Kf calculations for several pipe sizes using a nominal applied stress of 7500 psi. Note that at relatively shallow depths, the Kr is high; therefore, the crack growth rate will be relatively fast.

H6 wever, the 7K actually diminishes as the crack grows to about half way through'the wall. This prediction is consistent with service experience; very few, if any, actual cracks of significant circumferential extent have been found deeper than about 50% of the wall thickness.

- A.5 -

I L

1 +

-3 10 _

_  ; l -

- NRC CURVE O 1inlyr /,

10-4 -

- o _

~

, O

• /'/ f _

/ O

/ -

25 10" _

/

b -

/ g V 0.2 ppa 0,; sens sized at 1150*r/2 h (EPR

• 15 C/cm ) - GE

[

A 0.2 ppe 0,; sensttised at 1150*f/2 >

I (EPR

• 10 C/cm )

• CE O 0.2 ,,. 0,; sensitized at ii50*r/24 h z: s.

r

.u Q 0.2 ppe 0 ;2 severely sensitized O 8 ep= 0,i seastili a at u 50'r/24 h l0~6 C.D _

E Uf4 HITACH) h0H 8 ppe 0 ; senststred by weldin9; 5.::t a: - 0.04 in./yr 2

n 5 at ,32',/24 m (5u )

@ 8 ppe 0 2; sensittsed at (1292*F/10 min)'

2

+ (932'f/24 h) (EPt

• 4 C/cm ) ~

8 8 ppm 0 ;2 sensittred at (1292*F/10 min]

l + (932'f/24 h) (EPR

• 4 C/ca )

I 7 l r

• 0.i H,, t . 0 u g N 8 ppe 0 ;2sensittsed at (1292*f/10 min)

I l * (842*f/257 h) (EPR

• 15 C/cs ) -

l

@ 8 Pps 02 ; sensitized at (1292'F/10 I min) gg7

+ (642*F/257 h) (EPR

• 15 C/cm ) g l i

• 0.1 Hz , t

• 0.u DATA O 8 ppe 0 ;2 sensittred at 1292*F/14 h l I (EPt*20C/ce)fa0.008Ma, l a

• 0.95 .

l X 8 pp. 0,; sensittaw at 1292*r/is n 2

(EPR

• 20 C/cm g g , 0.08 Ma, l '" '

8 -

I 1

0 10 20 30 40 50 60 70 STRESS INTENSITY,K (ksid.)

Fig. 1 CRACK GROVTH RATE DATA A.6 s

b 104-

.sje_ vs K for Intwgranulm Stress Corrosion Crackmg 3.50 x 104 KW l

4l5 104 4 ' ' ' ' ' ' ' ' ' ' '

3 x 10 10 15 20 25 30 40 50 00 70 80 90100 Ki (ksi 4 Figure 2 da vs K for Intergranular Stress Corrosion Cracking i dt A. 7

INSIDE WALL OUTSIDE WALL 50- , i i  ;  ;  ;  ;  ;  ;

L

, o GE 26 -

40

(

F o GE 26 (4 orimuths) 6 ANL 26 (2 ozimuths) 50 o o ANL 26(IN-SERVICE FROM KRB) 6

• ANL 20 N

~

o $

00 e 5 10

• @o E o fg 0 0 - --

3--- ao o-eB-m 5 -10 -

0 o

o 8 q [8 D

o h

-20 -

• y , ,

A

-30'- 66 *

• e l I I I I I I I l

~ '

O O.2 0.4 0.6 0.8 1.0 o/t i

Fig. 3 Through-wall Distribution of Axial Residual Stress in large-Diameter Pipes (t > 1 in.).

\

i A. 8 l I

l

- - - - _ - - - - - _ - - - - - - - - - - - - - - - - - J

l 1

40 ~

1

.)

1 i

S M ~

28", t = 1.3" 24", t = 1.1" 20", t = 0.90" Z

12", t =0.66" 20 ~

10 1 I I i 1 3 g i  ;

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 sh --*

Figure 4 THROUGH-WALL DISTRIBUTION OF y K WITH APPLIED STRESS OF 7500 PSI A. 9

4 g09W MEMORANDUM FOR: James H. Sniezek, Acting Chairman Committee for the Review of Generic Requirements FROM: Harold R. Denton, Director Office of Nuclear Reactor Regulation

SUBJECT:

STAFF LONG RANGE PLAN FOR DEALING WITH STRESS CORROSION CRACKING IN BWR PIPING, ORAFT NUREG-0313 REV. 2 - SUPPLEMENTARY INFORMATION By memorandum dated February 13, 1986, I forwarded to you a package on the subject matter and requested CRGR's review and endorsement of a staff plan to publish Revision i of NUREG-0313 along with a draft Generic Letter for public comment. On March 5, 1986, you provided several comments including the need of a regulatory analysis of the proposed action.

On the basis of a discussion between you and R. M. Bernero, the staff has prepared a regulatory analysis (Attachment 1) that discusses the regulatory bases, past agency actions, and an example of value impact analyses done by affected licensees. In summary, wa view this proposed action as providing alternatives for complying with existing regulations, pursuant to 10 CFR 50.55a(g)(6)(ii) not to impose new requirements.

With regard to your comment on the use of a NUREG to regulate, the staff also developed an alternate scheme (Attachment 2) that incorporates the essential elements in the NURF3-0313 Rev. 2 as part of the Generic letter to form a set of " staff requirements." The NUREG will also be included for reference, to explain the bases of the " staff requirements".

l Our current schedule (see SECY 86-63) for completing this action calls W for issuing this package for public comment by March 1986 so, again, we would appreciate your early attention to this request.

68/

Har ld R. Denton, Director Of fice of Nuclear Reactor Regulat1on

}

Attachments:

1. Regulatory Analysis

2. Draft Generic Letter (Alternate)

DISTRIBUTION:

DEfB - Central Files bcc: R. Bernero BWR - Reading Files W. Houston BWR - Regulatory Analysis files T. Speis D. Crutchfield E. Rossi G. Lainas #-

B. D. Liaw %V \

~

(V '*

%.\ ' 9\ .\ / ,

..! .. .. .. 3. . .....! ... .. ........ . ........f..........

NAMEMazel on:cd : 8.0. Liaw :G.C  : DGEisenhut :H.R. Denton : i

. ................................/.thinas$R.Bernero l  :% / M /86 0,/qA /86 :3 /'7 '/86 :1 / k /86 I

/ /86 -

/ /86 00 RO lb Y /

ATTACHMENT 1 REGULATORY ANALYSIS REGULATORY BASES Intergranular Stress Corrosion Cracking (IGSCC) has been occurring in BWR piping systems and other components for the past 20 years.

Although the cracks in the piping have not resulted in a major challenge to the safety systems, they have caused many leaks, and have degraded the overall integrity of the primcry pressure boundary (e.g., recirculation lines, RHR system piping, instrument lines, etc.) and other systems (e.g., reactor water cleanup system lines, core spray ifnes, isolation condenser lines, etc.) important to safety. The basic cause is that the piping weldments, in the as-installed condition, are not compatible with the environment of the BWR reactor coolant. The state of the weld sensitization and residual and service stresses can and often do combine to cause many weldments to be susceptible to IGSCC during operation.

Without additional actions such as are covered in NUREG-0313, the piping systems then would not be in acceptable conformance with.10 CFR 50 appendix A, General Design Criterion 4; "... Structures, systems, and components important to safety shall be desipqed to accommodate the effects of and to be comratible with environmental condition associated with normal operation, maintenance, ...."(emphasis added).

IGSCC has caused many leaks in BWR piping, including the main 1

4WJtW recirculation system and other systems that are part of the Primary Coolant System. Although most of these leaks have been small, the piping weldments are not expecte' to have any leakage. Such leakage is therefore considered abnormal leakage. GDC 14 requires that the reactor coolant pressure boundary "... be designed, fabricated, erected and tested so as to have an extremely low probability of abnormal leakage, of rapidly propagating failure, and of gross rupture." It is clear that- .

piping systems with weldments susceptible to IGSCC do not meet the intent of this criterion. General Design Criteria 30, 31, and 32 are also relevant to the coolant pressure boundary, leak detection, and inspection requirements.

In summary, without suitable mitigation actions, BWR piping would not be in conformance with the regulations. The purpose of past regulatory actions, particularly the implementation of the staff positions in NUREG 0313 and Rev. 1 to this document was to describe the actions considered by the staff to be necessary to bring the piping into

. acceptable conformance with the regulations. The actions considered acceptable were limited to replacement of degraded piping, and augmented inspection schedules for continued, interim operation.

l

'Q ~ ; '. ,* .

f.

? , , ,

2 What Rev. 2 to NUREG-0313 does is to provide the licensees with a

[ wider spectrum of. acceptable alternatives to meet the intent of the i regulations. Considering its individual situation and degree of pip'ing 1 degradation, a utility can select an integrated program that includes options of pipe replacement, repair, stress improvement, water chemistry improvement and augmented inspections to bring the plant into conformance with the intent of regulations.

No new requirements are to be imposed by this proposed action.

The basic requirements are still those of the General Design Criteria in 10 CFR 50. The staff positions stated in Revision 2 to NUREG-0313 reflect the improved knowledge on the behavior of degraded piping and, therefore provide the affected utilities with a broader selection of acceptable alternatives to meet the same requirements. Although the preferred approach is still to restore the piping to the original condition by replacement with more resistant materials, the proposed action recognizes the condition of piping as now exist and has taken into account the state-of-the-art knowledge in repair methods, inservice inspection technologies, stress improvement. measures, and changes in the coolant chemistry.

REGULATORY ACTIONS REGARDING BWR PIPE CRACKS Cracks in pipe walds of operating Boiling Water Reac' tor (BWR) plants have been reported since 1965. As more BWRs went on line and

- accumulated service time, more and more such cracks were discovered,

[ Until recently, most cracks were discovered because they were leaking, W whereas with the recent improvement in inspection methods, most cracks are now being found by Ultrasonic Testing (UT).

There has been a long history of regulatory actions regarding BWR pipe cracks. A summary of these actions will help put the current effort into perspective: .

Regulatory Guide 1.44 " Control of the Use of Sensitized Stainless Steel" As a result of several incidents of cracking of stainless steel piping sensitized by heat treatment or welding, Regulatory Guide 1.44 was developed and subsequently revised. In this guide it was recognized that standard grades of type 304 and 316 stainless steels could be sensitized by welding or post weld heat treatment, and that such sensitized material would be subject to cracking in BWR primary coolant. Because Regulatory  ;

Guides are not requirements, and alternatives (such as controlling l welding heat input) could be invoked, the strict positions of the Guide 1 - against the use of standard (not low carbon grade) material were not imposed.

R y ) - N % W' pi k h y p '4...

w.. .

Pipe Crack Study. Group And NUREG-0313

('

~ .

In 1974 when IGSCC pipe cracks were found in several modern BWRs, a Pipe Crack Study Group was formed to investigate the cause and to make recommendations for regulatory remedies. It was determined that the cracking was caused by a combination of material sensitization, high stress, and the agressive oxidizing nature of the BWR primary coolant.

HUREG-0313 was issued as the basis for the implementation of the Study Group's recommendations through a Generic Letter. These recommendations can be summarized as follows:

1) new construction should be of materials resistant to IGSCC;

2) augmented inservice inspection and leak detection should be implemented for non-resistant piping; and

3) piping found to contain cracks should be replaced with resistant material.

Pipe Crack Study Group And NUREG -0313 Rev. 1 When cracking in additional systems and pipe sizes were found, another Pipe Crack Study Group was for.idc Conclusions and recommendations were similar to those of the earlier Study Group, but included a broader scope of piping systems. NUREG-0313 was rev' sed to reflect this larger scope, to recomcend improved inspection methods, and

~

It was also implemented by a Generic to tighten leakage limits somewhat.

l Letter.

! W Piping Review Committee And NUREG-0313 Rev. 2 In 1982, leaking cracks were found in several of the 28-inch Subsequent diameter main coolant pipe welds at Nine Mile Point Unit 1. IE Bulletin ultrasonic inspection discovered many more cracked welds.

82-03 was issued to require a sample inspection of recirculation pipe- .

welds in 9 plants. When significant cracks were found, IE Bulletin When more 83-02 was issued to cover the remaining BWR plants in operation.

cracks were discovered, orders were issued to five plants to accelerate their inspection schedules.

As a result of these sample inspections many cracks were found.

Repair and crack evaluation procedures were developed, The and used only by the current staff to ensure continued, interim safe operation.

guideline, that of NUREG-0313 Rev. 1, required replacement of piping runs if cracks were found. Interim operation guidelines were needed as an alternative to shutting down most BWR's for up to a year for pipe replacement. There was concern that suitable special replacement materials would not be available in the quantity and schedule needed.

Further, extensive pipe replacement without proper pre planning would certainly be excessively expensive in money, time, and occupational

r 7- c~,j7, t , y.y.,mgepec.c. grmy u 3' pq;r:W n ,,, %g 3?: y 5, i ;g . .a Ngg.py 3 -

1; & -

.g , . .. . c. ,

.s. <

j... ,

c, ..,

.;f

. .. .; . . . - '

., g  :

,7 . am.y .Q ,y*.ygg' {._

, . . .. : ) . '{ , l

• radiation exposure (ORE). Therefore, the staff proposed in SECY 83-to the Commission for endorsement of the interim guidance on crack ,267C evaluation and repair methods considered by the staff to be acceptable for interim operation. With'the Commission's concurrence, the interim staff guidelines were tr.ansmitted to affected BWR licensees for implementation.by Generic' Letter 84-11. In the meantime, the Executive Director For Operations directed'that a Piping Review Committee. be formed, which included the Pipe Crack Task Group to review the interim ,

staff positions proposed in SECY 83-267C. As a result, the Pipe Crack Task Group issued NUREG-1061 Vol.1, which formed an integral part of a long range approach to deal with the problems as delineated in Commission paper SECY 84-301, in which the staff described the long range plan, including a new revision of NUREG-0313 to incorporate the recommendations of the Piping Review Committee. Subsequent to its review of the proposed long range approach, the Commission directed the staff to develop a.long range implementation plan for the Commission's approval by April 30, 1985. Barring any changes to incorporate major comments from the public, this package will form the body of ,a Commission paper in response to the. Commission's direction.

RISK ANALYSES '

Although piping constructed from normal grades of:sta'nless steel is subject to cracking by'IGSCC in normal .BWR reactor coolant, and therefore f would not be in.conformance with GDC 4 without additior.a1 mitigative I

W actions, it is useful to assess the public risk associated with this-problem. Several risk assessment studies h' ave been made, and are.

discussed below:

Piping Review Committee Studies The Pipe Crack Task Group of the Piping Review Committee performed IGSCC Risk Studies; these are published in NUREG 1061 Vol.1. These studies concluded that the presence of IGSCC may reduce safety margins believed to exist, but would not increase the probability of core melt significantly beyond the calculated value. This conclusion was based  ;

primarily on the assumptions that. IGSCC would increase the LOCA probability by only a small amount based on an unknown intrinsic IGSCC component already in the data base. Specifically, it was " estimated that IGSCC may be 30% of the LOCA frequency, but because the LOCA contribution to the core damage frequency is small, it would require a considerable error, at least an order of magnitude, from IGSCC before LOCA's would be a dominant contributor to the core damage frequency."

Nevertheless, the study recognized the limitations of the approach used, and recommended that further work be undertaken using the computer code PRAISE, which was being modified to include an IGSCC model.

(,

f hb@*$ Yk i s .m,A -

4% Lawrence Livermore Studies This recommendation was implemented by further studies at Lawrence Livermore, using the IGSCC modified PRAISE Code. Results of this study '

were reported in February 1986 at the Thirteenth Water Reactor Safety Research Information' Meeting (NUREG-CP 0072). .These results.show a large ef fect of unmitigated'IGSCC on the. probability' of leaks and the probability of direct DEGB. The results -of this study, using the Brunswick recirculation system, showed an increase in DEGB probability of from 7.7 E -12/py to 9.0 E-4/py caused by unmitigated IGSCC. This value of the LOCA frequency is approximately an order of magnitude higher than the LOCA frequency used for the core damage frequency analysis in the NUREG 1061 Vol. study.

It is relevant to compare the Livermore calculations for the probability of leaks caused by IGSCC with actual' experience. Their value of 6.8 E-1/py is certainly consistent with recent experience, in which many plants have found several leaking weldments during each inspection. This result would tend to support the conclusion' that the modified PRAISE Code is providing r,easonable probability values.

. Summa ry" Althougt probabilistic ris.k assessments of the IGSCC phenonmenon arr:

difficult to perform, and are' highly dependant on the assumptions used, the newer more rigorous methods appear to show that without the mitigative actions being taken, such as those covered in NUREG-0313 Rev.

2, the public risk could be increased. This finding supports of conformance with GDC 4, and the associated GDC 14.

VALUE IMPACT ANALYSES NUREG-1061 Vol. I _

The Pipe Crack Task Group of th'e Piping Review Committee performed some Value-Impact assessments, described in Appendix E to NUREG-1061 Vol.

1. Because the range of possible scenarios for various affected plants is so wide, several simplified scenarios were selected for analysis.

Because these scenarios were so simplified, they do not reflect the complete range of options available under the guidance provided by NUREG-0313 Rev 2. Nevertheless, the studies did show that total costs in dollars and occupational exposures over plant life could be significantly reduced by other alternatives short of complete and immediate pipe

. replacement. The countermeasures such as Improved Water Chemistry (HWC) and Stress Improvement (IHSI) to piping with less resistant materials are covered by NUREG-0313 Rev. 2. Accurate, quantitative assessments of all plants affected would be extremely difficult, considering that events such as a mid-cycle leak that requires shut down and repair are

...,.w.. , . , . .

((; N jf.y[. ( '.

1 .

.I ,, f o . , . , .

( .; .

n N.$&k? h:h'% W Dj R Q ;Lnt Ol&#^L R .% [ "I s.i O p'.% g ,

f A impossible to . pre, dict. J Even31f predicted, the calculated cost.[in.dollarsk and person-ree would be very un;certain. Alt. hough the cost of actions ' '

- taken to, avert such. occurrences can'be calculated with a reasonable.e accuracy, estimates of the benefits in' terms of averted cost %Yrl'dibe: #i .

extremely uncertain. The'costLto.each plant ' depends, on the co'ndition andf t!)e numbers of. degraded weldments, the utility's excess available;.L.u- '

capacity,' and the cost and. availability of 'replacem'ent power;,all of, these can vary greatly. ,

Utility Analyses Several utilities have decided to replace part or all of their susceptible piping. The staff does not know the bases for these decisions, but major items that have certainly been taken into consideration were the extent of degradation of the piping, utility.

estimates of cost of future inspections and repair, perception.of the risk of a mid cycle leak, consequences in terms of required replacement power costs in this event, and, of course, the cash flow position of the .

utility. Many other utilities are still in'the decision making mode.

The uncertainty regardirig staff positions on alternative mitigation actions has made their decisions ;nore difficult. The g'uidance that is provi.ded in NUREG-0313 Rev 2 is urgently needed to alleviate this situation r As an example, Commonwealth Edision made such'a detailed analys'is of the overall costs for the life of.their Quad Cities Unit 2. The .

(W piping in Quad Cities 2 is' relatively free of degradation; only about . -

10% of the welds have been found to be cracked. Their analysis assumed that'the repairs they have made will continue to be effective, and that planned improvement in water chemistry control will preclude additional cracking for the life of the plants. Although these assumptions are considered optimistic, they furnished a basis for a decision making analysis. The salient features of their analysis are as follows: ,,

QUAD CITIES 2 ANALYSIS Total Outage. Total Cost To Total ORE To Option. Time to EOL (days) EOL (1985 $)

EOL (man-rem)

- Total Replacement 190 130.3M 1900 (0313 Rev. 1 Position)

.- Integrated Program (0313 Rev. 2 Positions) 36 17.5M 650 EOL

• End of Life l This analysis shows a large advanta'ge to the utility of using the staff

. guidelines provided in NUREG-0313 Rev. 2 instead of the current positions of Rev.'.1. On the other. hand;.the optimistic assumptions used may not be completely. realize.d; additional cracking may occur that would necessitate"mid .

cycle shutdowns and repairs. This could markedly alter. the cost differential

~

stiown.

v

.c e . e,g n, . ~ -.c  ;

6$}CA $h. bl.j _ *W ' bifff _ ,

W* '

.:. s ...

,y -n - , '. F,' '. ,-

i QQ7 }*( 'L  ?'

f. * * . Value/ Impact Evaluation .

M Consequences (Risks) '

The core-melt frequencies and risk estimates used here 'were taken from

' N'UREG 1061 Vol.1, Appendix E. Although .the ' risks results of. this analysis would not change significantly if the Livermore values were used, the final judgment that there are major cost benefits available if the options provided by NUREG 0313 Rev. 2 are used is clear. l The core melt frequency used in NUREG 1061 Vol. I considering IGSCC is 9.0E-7/ry. The public risks, given a core-melt, was estimated at 6.7E+6 man-rem. The product of these yield a public risks of (9E-7) (6.7E+6)=

6.0 man-rem /ry. Over the remaining plant life of 30 years, this amounts to a public risk 'of approximately 180 man-rem /r.

Change In Risk

~

If we assume the NUREG-0313, Revision 1 results in a requirement to change-out the piping, and NUREG 0313 Revi.sion 2 allows latitudes that will not require piping change-out, the ARisks and a cost between those alternatives establishes the potential change in risk and cost.associat'ed

'- with'this issue.. -

If we assume the Revision 2 methods to be approximately 90% as effective

.j

, in reducing IGSCC (and risks) as piping changes-out (Revision 1), the l potential increase in public risks by, changing to Revision 2 is (.1)-  !

(180)= 18 man-rem /r.

Cost' Estimates The cost estimates for Revision 1 and Revision 2 are taken from the Quad Cities analysis shown'above. Assuming the Revision 1 (pipe change-o'u.t)-

costs at 130 $M, and the Revision 2 (integrated program) costs at 18 $M, the cost savings for use of Revision 2 is approximately 112 $M/r.

Value/ Impact Assessment In order to obtain the same level of risk protection (a difference of i approximately 18 man-rem), a plant would need to expend approximately an additional 112 $M/r. This amounts to:

112 $M = 6.2 $M

- 18 man-rem man-rem If we compare this to a rule-of-thumb cost / benefit of 1000 $/ man-rem as an appropriate guideline, the high cost to achieve this benefit (Value=

18 man-rem) is approximately 6200 times the reasonable goal.

. l s + .

4 4

~ '

'f'<< k h

.~ .

r

. , . g.

3 ;lyl.: s ,  ?

.g &,Ni  ;. y i %g; .f.

.< , .; .A f '

y,

..., .,> ;.3

• j

,w 1y  ; k.; Qv .- .9 f d y. ;g,} 9 6 _-Q

- ., * , .. y s. . . . .

< . .. .. ...:.' . . . . Ls . y

.hfl,1..Ybh;ffrff(hh),' y lk'^

k.. f0

&Wkh@&hk0h;fb$$f  ;; b? ' fib W;l$.f5 ~?: Y . fU ? '

k du '.

M.WWW:If'the' coststof, Revision:2"se're double the costs used above, the?

? cost / benefit ^ is 'still . startling.' Even and order of magnitude 'c c.-Crgt.- : < underestimate iin the risk change and factor of 2 in costs would amount to

' cost / benefit' ratios that are very attractive. .

" .W ~ ' '

.'O' ggg .

The reduction in ORE in changing from Revision 1 to Revision 2 is 1 estimated at approximately 1200 man-rem / plant. ~ A comparison of the probabilistic increase in public risk of 18 man-rem, when compared to the more certain ORE risk reduction of 1200 man-rem should also be considered in the decision making related to this issue. The ORE cost / benefit reflecting this reduction .in ORE of 1200 man-rem at a cost savings ,

(negative impact) is-  !

l (112 $M) =- (93 $K)

(1200 man-rem) (man-rem) l

~

Conclusion Based on the very high cost / benefit in terms of cost expenditures i required to achieve minimum public risk reductions, and the cost saving obtained in. also reducing ORE, the use of .an integrat.e'd approach.using

'the alternatives provided.in Rev. 2 is.strongly supported by this; analys.i s.

SUMMARY

Alt' hough a comprehensive and detailed cost /b'enefit. analysis is not available and hence not provided in this package 'for CRGR review, the rationale discussed above is summarized as follows:

1) The promulgation of. Revision-2 to NUREG 0313 provides licensees .

with severa15 alternative methods to comply with the regulatioris.

2) Although the change in public risk by adopting Revision 2 to NURtG 0313 is small, the reduction in occupational radiological exposures that can be realized by the use of Revision 2 is large.

3) Definitive cost-benefit cnalyses of these alternatives are extremely complex, and will vary greatly depending on the condition of the plant being considered and utility-specific financial factors. Nevertheless, the alternative actions permitted by Rev. 2 will provide the utilities with a basis for such analyses, and for some plants, will permit the use of considerably less costly methods for complying with the regulations. ,

e ,

., y e ._ "

- . < 7 a( ...g

/ 1. - ...

. , , . A j .. s

~

. . ., y.

g .., , .

7 r %,' wh$ ~ - . g P '

3 7,4J h > ([. $ ' D , . ~ gh 1..E E 4t) -

3 k !,kkd , . n-si IfIh r.

l.h ; ;., . fh:;d ,.

f, b d[3:-k

@! lie - 4) tthe ? staff e.is'N.V isig'- Although still a complete the. remedial replacement action generally~preferred of susceptible piping, by$tho~se s!Sohk

~

with piping that is only slightly degradedccan be,.jbrytIgti, aht,d$Nig'v!'

"n maintained in a condition conforming with.the regulation.s sing' '. -

alternative methods covered in Rev. 2 that are significant1y'fl.ess -

costly in both dollars and ORE.

-' " '-  : 'e -

e l

1 l

1 l t e

l 9

Northem States Power Company 0$$pIsYnnoota 5540$

Telephone (612) 330 5500 September 12, 1986 Mr. Robert M. Bernero, Director Division of BWR Licensing Office of Nuclear Reactor Regulations U.S. Nuclear Regulatory Commission Washington, DC 20555

Subject:

Comments on Drafts of NUREG 0313 Rev. 2 and Implement Generic 1 Letter

Dear Mr. Bernero:

Er riosed are comments on the Draf t Generic Letter "NRC Position on IGSCC in End Austenitic Stainless Steel Piping" and on the Draft huREG 0313, Revision

2. These comments have been prepared by the BWR Owners Group for IGSCC Research (BWROG-II) in resr>onse to the notice published in the July 21 edition of the Federal Register.

Overall, the NRC authors and reviewers are to be complimented on their treatment of the complex issue of IGSCC in BWR piping. The draft NUREG revision, in particular, presents a well-written, concise and balanced account of the current status of industry's efforts to develop, qualify and implement remedies to the pipe cracking problem. Most of the Staff Positions set forth in the draft Generic Letter are felt to be soundly based and reasonable. Accordingly, many of the attached comments are of a

" fine-tuning"typeandareaimedatclarifying(asopposedtochanging)the various Staff Positions.

We believe that in a few areas the draft documents go beyond the generic issue with which they are nominally concerned (i.e. IGSCC in safety-related,  ;

wrought austenitic stainless steel piping in BWRs). Examples are the l inclusion of carbon steels and cast austenitics in the Staff Position on I

Materials, the comments on reactor vessel internals in the Staff Position on Water Chemistry, and the inclusion, by reference, of Class 3 piping in the Staff Position on Inspection Schedules. We believe that the Generic Letter, in particular, should concern itself only with the generic issue of IGSCC in the weld heat affected zone of wrought austenitic stainless steel piping and safe ends, and several of our comments suggest changes that are intended to sharpen the focus of the Generic Letter.

l

\

f, A

.g, Page 2 ,

September 12, 1986

' Finally, a comment on the proposed Limits on Number of Cracked Weldments in Piping. As discussed fully in the attachment, we can see no justific-ation for the proposed 25% limit. Our position is that the measures identified in the draft NUREG revision are adequate to cor. trol IGSCC in BWR j piping and that the proposed additional limit is neither necessary nor i beneficial. This arbitrary limit injects a factor which stands as a marked )

contradiction to the non-adversarial environment in which the NRC and the '

utilities have functioned to address the IGSCC issue on a rational technical basis and in a timely and responsible manner. We strongly recomend that this proposed limit be omitted from the Generic Letter.

If you or your staff have questions on this letter or the attachment, please address them to my attention.

Yours sincerely, G. H. Neils, Chairman BWROG of IGSCC Research cc: John D. Philips, NRC bec: M. S. Hazelton, NRC l

.c s

BWROG-II C0tNENTS ON DRAFT GENERIC LETTER "NRC Position on IGSCC in BWR Austenitic Stainless Steel Piping" AND ON DRAFT NUREG 0313 REV. 2

" Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary Piping"

1. The draft letter addresses the generic issue of IGSCC in sensitized wrought austenitic stainless steel piping and we believe that the NRC staff positions stated in the letter should be confined to this issue. At present, this is is not the case. For example, IGSCC in cast stainless steel is not a generic issue, but the staff position appears to require augmented inspection

, of castings not meeting specified carbon and ferrite requirements. We believe augmented inspection of such castings is inappropriate considering the favorable service experiener and the absence of IGSCC in castings purchased to ASME Code requirements. In fact, the good IGSCC resistance of austenitic stainless steel castings, in comparsion to that of wrought pipe, may result in part from differences other than carbon and ferrite content. We note further  !

that the inspection methods to which the generic letter refers are not j

appropriate for castings.

Similarly, plain carbon steel piping is imune to sensitization and IGSCC, as I correctly noted in draft NUREG-0313. Revision 2. The generic letter should not specify strength requirements for carbon steel piping based on considerations other than corrosion cracking resistance. Certainly, the '

prescribed program of augmented inspection for IGSCC is not the appropriate

" remedy" for plain carbon steel that does not meet applicable strength and toughness requirements.

I i

1 l 3924SM6a

)

2. Stainless steel weld metals generally have proved resistant to IGSCC in piping applications in BWRs. The IGSCC resistance of these materials is ,

believed to be enhanced by increasing the ferrite content above the 5%

specified by ASME.* For this reason, a minimum as-deposited ferrite content I of approximately 8% has been widely specified for corrosion resistant cladding (CRC) and for weld overlay repairs (WOR). However, in view of the favorable j service experience, we believe augmented inspection is unnecessary for butt I welds which meet ASME requirements for ferrite content, provided the materials

]

adjacent to the weld are IGSCC resistant. j The list of IGSCC resistant stainless steel weld metals suitable for use in CRC and WOR applications should be expanded to include the 309L, 316L, and 308Mo-L grades, in addition to type 308L. j

3. Wrought material which has been solution heat treated should be clearly identified in the generic letter as a resistant material as it is in the draft l

NUREG-0313, Rev. 2, Paragraph 5.3.1.1 '

4. To incorporate items 1,2 and 3 above into the draft generic letter, the following rewording of the Staff Position on Materials is recommended:

(1) Low carbon wrought austenitic stainless steels are considered resistant to weld sensitization and IGSCC. These materials include the grades and corresponding carbon levels specified in Table 1 of draf t NUREG-0313, Rev. 2, as well as other grades with equivalent low carbon contents. These materials must be tested for resistance to sensitization in accordance with ASTM A262-A or -El as appropriate, or in accordance with an equivalent standard.

(2) Wrought austenitic stainless steel pipe with carbon content not l

in accordance with Table 1 of draft HUREG-0313, Rev. 2 is l

• The minimum ferrite content for welds is usually specified in terms of l ferrite number (FN) rather than percent ferrite, although the two are nearly equivalent numerically.

2 3924SM6a

4 9  ;

considered to be resistant to IGSCC if the pipe assembly is solution annealed by heat treatment following welding.

(3) Piping material is considered resistant to IGSCC if the weld- I heat-affected zone is protected by resistant weld cladding )1 (oftenreferredtoascorrosion-resistantcladding,orCRC). '

Resistant weld metals that are suitable for CRC applications include type 308L, 309L 316L, 308Mo-L, and equivalent grades, with a minimum as-deposited delta ferrite content equivalent to 8FN.

l (4) Wrought austenitic stainless steel pipe which does not meet one of the above requirements for resistant material is considered non-resistant and is subject to augmented in-service inspection requirements.

(5) The IGSCC resistance of nickel-base alloys will be considered I on a case basis. Nickel-base alloys have performed  !

satisfactorily in some 8WR piping applications but have suffered IGSCC in creviced configurations. Some solution heat treated nickel-base alloys, including Alloy 600, are considered resistant: Inconel 82 is the only currently available nickel-  ;

base weld metal considered to be IGSCC resistant.

5. Table 1 of draft NUREG-0313, Rev. 2, which is referenced by the draft generic letter, should be revised to include Type 347 stainless steel with j 0.04% maximum carbon content, to be consistent with current foreign and i domestic purchase specifications. Type 347 has been used successfully with i even higher carbon content, exhibiting no evidence of IGSCC.

6. In the draft generic letter as currently written, the Staff Position on Processes is restricted to new or replacement piping. We recommend deleting the words "new or replacement" in the first sentence of the Staff Position to allow for process applications to existing installed piping systems.

l 3

3924SM6a

l

7. The Staff Position on Water Chemistry in the draft generic letter requires revision to reflect recent experience and in addition should be confined to IGSCC of BWR piping (see item 1). Recommended rewording for the first paragraph is as follows:

The staff believes that HWC is an effective countermeasure for reducing IGSCC in BWR piping. However, the response of BWRs to hydrogen injection differs from plant to plant and the development and verification of a  !

generic HWC specification is not yet complete. For these reasons, I; reductions in piping inspection frequency based on the use of HWC will be considered on an individual case basis, at least at the present time.

i The second paragraph of the Staff Position should be deleted. I

8. The Staff Position on Inspection Methods and Personnel refers to an

,. insp'; tion qualification program that is applicable to IGSCC in wrought material. Clarification should be inserted in the second paragraph as )

fc110ws: "In addition, for detection and sizing of IGSCC in wrought austenitic stainless piping at weldments, the detailed procedure...". (

9. The generic letter should include a clear definition of the piping that is subject to its inspection requirements. At present, the Staff Position on Inspection Schedules in the draft letter references Section 5 of the draft NUREG revision which provides (in Section 5.1) a definition that differs sharply from previous NRC positions (e.g., NRC Generic Letter 84-11) and would require extensive inspection of Class 3 piping. We doubt that extensive inspections of such piping, which by definition is of minimal safety significance, could be justified in benefit-cost terms. We recommend that the scope statement given in item 2 of HRC Generic Letter 84-11 be substituted for that in the second paragraph of Section 5.1 of the draft NUREG revision and that a similar statement be inserted in the draft generic letter as an initial i paragraph in the Staff Position on Inspection Schedules. l

10. Table 1 of the draft generic letter prescribes inspection intervals of 10 )

years, 3-1/3 years, and every refueling outage for various categories of l weldments. We recomend that "3-1/3 years" be changed to "3-1/3 years or 2 l 4

3924SM6a

t refueling cycles, whichever is greater". Because the refueling cycle can be as long as 2 years, the 3-1/3 year interval could cause an arbitrary and burdensome requirement to inspect Category D and E welds every outage.

l The word " outage" appearing in Table 1 should be changed to " refueling outage" i

for clarification.

In Table 1, the inspection schedule for Category E welds should be changed from "all next outage" to."50% next refueling outage" to be consistent with draft HUREG-0313. Rev. 2 Paragraph 5.3.2.5.

The extent of inspection shown in Table 1 for IGSCC Category A welds should be changed to be consistent with ASME XI Section IWB-2500. Similarly, the extent of inspection for Category B welds should also be changed. The recomended wording is indicated in the attached revised Table 1, which also reflects the changes recomendr! above.

11. The Staff 'osition on Leak Detection in the draft letter requires conformance with Regulatory Guide 1.45, which has not been applied to many 4 plants. This "backfit" application must be handled on an individual plant basis, in accordance with 10CFR50.109. Serious flaws in the technical content of Reg. Guide 1.45 have been presented to the NRC staff and accepted in connection with the Limerick Unit 1 plant. More recent and appropriate technical guidance is contained in ANSI /ISA S67.03. We recomend that the first paragraph of the Staff Position on Leak Detection be deleted. l

12. The foregoing coments on the draft generic letter are also applicable to the corresponding and/or referenced sections of NUREG-0313, Rev. 2.

Additional coments on the draft NUREG revision are as follows:

(1) In 3.1.1, change the fourth sentence to read " Weld overlay is performed with water in the pipe during welding, ..."

(2) In 4.4.1, insert "...a margin of at least 2.77 ..." to be consistent with ASME Section XI as discussed in 4.1.

5 3924SM6a  ;

)

?.,'

~

k.'

(3) I'n 5.3.4, we can see.no technical merit'in including " welds of lower categories" in the additional. group of welds selected for examination following detection of an IGSCC indication. Further, the word." lower" is ambiguous.in this context, since it could refer either to the relative IGSCC resistance (in which case Category E is

" lower" than Category D) or to the amount of inspection required (in' q whichcaseCategoryCis" lower"thanCategoryD). In either case, .i examination of welds in the " lower" category would not be the best response to the discovery of a' crack indication in a Category D j weld. Clearly, it would be more informative to inspect additional CategoryDweldsthantoexamineeitherCategoryEwelds(whichare already cracked) or Category C welds (which are stress improved and relatively unlikely to crack). We therefore recommend the following revised wording for the second paragraph of 5.3.4: q If new IGSCC indications are found during sample inspections, an additional samf e of welds should be examined. The welds included in this second sample should be selected on a case-specific basis. . Generally, the welds selected should be from the same IGSCC category as the weld showing the indication and the number examined should be approximately equal to the number of welds of that chtegory in the initial sample. If any new IGSCC indications t.re found in the the second sample, all i remaining welds of that category should be examined.

1 L~

(4) The disclaimer entitled " Note for Recipients of Draft HUREG-0313, l Revision 2" should be included in the final document as a foreword. l

13. A proposed addition to the draft generic letter sets a limit on the )

number of ct acFed weldments in piping, irrespective of repair or evaluation. .

The basis for this limit is not discussed in draft NUREG-0313, Rev. 2 and  ;

- appears to be arbitrary. The position of the BWROG-II is that measures for evaluation, mitigation, reinforcement and inspection as identified in the i draft HUREG revision are adequate to control IGSCC in BWR piping systems, and that the additional arbitrary limitation is not necessary or beneficial.

l l

6 3924SM6a r ' ,

1s

>s..

Meaningful technical discussion is difficult because the basis for the proposed limit has not been disclosed. We speculate that a possible motivation is a desire to minimize shrinkage stress resulting from a large number of weld overlay repairs. However, shrinkage and fitup stresses are .

best quantified by analysis and controlled as part of the overlay design A

. process, as discussed in the draft NUREG revision. Shrinkage stress is not sensitive to the number or proportion of welds repaired and is not limited by the proposed 25% limit. For example, analysis would show that a single overlay repair in a recirculation system riser causes higher shrinkage stress than a symmetric arrangement of repairs in each riser associated with one ring  !

header. The arbitrary 25% limitation is not an effective way to limit or reduce shrinkage and fitup stresses.

We question whether an analysis of costs and benefits associated with the proposed limit would support its implementation. The proposed limit requires BWR owners to apply increased resources and man-rem exposure toward inspection and sizing of small cracks which can have only ama11 effects on system integ-rity. An arbitrary limit on the total number of small cracks could for'e replacement of lines which are structurally sound and reliable, again at high cost and radiation exposure. The proposed limit precludes use of alternatives described in the draft HUREG revision which could be more cost effective in maintaining' system integrity. For these reasons, we recommend that the five paragraphs in the draft generic letter under the heading Limits on Number of Cracked Weldments in Piping be deleted in their entirety.

14. Under item 2 on page 6 of the draft generic letter, BWR Owners are requested to provide a Technical Specification change incorporating an augmented inspection program. Because the degree of augmented inspection will change as measures are implemented to mitigate IGSCC, we feel that it would be more appropriate to provide the NRC staff with a plan for an inservice inspection (ISI) program that can be modifiM as necessary as mitigating measures are applied. We recommend the following wording:

2. A plant-specific ISI plan incorporating an augmented inspection program for wrought austenitic stainless steel piping iii accordance with Table 1 and Section 5 of NUREG-0313. Rev. 2 and consistent with the Owner's current plan for IGSCC mitigation at the plant.

7 ,

I 3924SM6a

_ _ _ _ _ _ - _ - _ _ _ _ _ - -