ML20245D376

Safety Evaluation Supporting Licensee IGSCC Program for Refuel 7 Outage

ML20245D376
Person / Time
Site:	Brunswick
Issue date:	04/25/1989
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20245D370	List: ML20245D376
References
NUDOCS 8904280273
Download: ML20245D376 (6)

Text

_ _ _ _ _ _ _ - _ - - . _ . - _ _ - . _ _ - . __ - _ . - - _ - _ _ - - - . , _ -

3 o UNITED STATES

, g l[-

3 g

E NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20555 '

A...../

U SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATING TO INSPECTIONS AND REPAIRS OF INTERGRANULAR STRESS CORROSION CRACKING (IGSCC)

CAROLINA POWER & LIGHT COMPANY, et al.

BRUNSWICK STEAM ELECTRIC PLANT, UNIT 2 DOCKET NO. 50-324

1.0 INTRODUCTION

The. staff has reviewed the licensee's submittals dated October 2 and December 30 1987 and March 25, 1988, including the inspection results, IGSCC mitigation, flaw evaluations and overlay repairs to support the con-tinued operation of Brunswick Steam Electric Plant, Unit 2, (Unit 2) in its present configuration for one 18-month fuel cycle. During the Unit 2 refuel 7 outage, 81 Class 1 piping welds, including 16 Inconel 182 buttered' nozzle welds and 10 recirculation inlet safe-end thermal. sleeve attachment welds,;were ultrasonically examined.

The results of the inspection showed that flaw indications were observed in a pipe to elbow weld (12ARD2) and a safe-end to nozzle weld (12ARE5) in the recirculation riser piping system. These two welds were reinforced with standard designed weld overlay. Three previously overlay repaired welds (28A4, 2883 and 2884) were upgraded to meet the standard design-requirements, and their surface finishes were improved to facilitate ultra-sonic inspection. For the two previously unrepaired suction safe end-to- I nozzle welds (28A1 and 2881) in the recirculation piping system, the existing flaws were reported to have no change in sizes; however, a new i small axial indication was found in weld 2881. Mechanical stress improve-ment process (MSIP) was applied to these two welds and 13 other Inconel 182 buttered nozzle welds as a mitigation for IGSCC.

A through wall leak was found in one of the jet pump instrumentation (JPI) penetration seal connection welds. This weld was reinforced with addit-ional weld metal. All other 23 JPI penetration welds were examined, and no flaws were found.

I 2.0 . DISCUSSION j Inspection The licensee reported that there are 180 Class 1 piping welds in Unit 2 l subject to IGSCC inspection. 44 of these welds are made of IGSCC resistant l material and no sample was selected from these welds for inspection. 81 l

C P

9:

J i  : -

welds were inspected during refuel 7 outage of Unit 2, which included ,

j 16 Inconel 182 buttered nnzle welds,10 riser thermal sleeve-to-safe end attachment walds and 21 overlay repaired welds. The original sample size of 72 welds was detemined in accordance with the guidelines in NUREG-0313 Revision 2, and was expanded to 80 welds after flaws were found in the original-samples.

The staff concludes that the inspection scope for Class 1 piping meets the staff requirements and the guidelines in Generic Letter 88-01 since more than 50% of the IGSCC susceptible welds were insp0cted during this outage.

The staff also concludes that the limited sacale npansion is acceptable because all the welds with configuration or stas imilar to the two welds found flawed during this outage were inspected.

Ultrasonic Examination The Centerlicensee qualified reported personnel that thethe from IGSCC inspection General was perfomed Electric Company by)EPRI NDE (GE . These examiners also passed the latest requalification program. The fully auto-mated GE " SMART" ultrasonic testing (UT) system was used for examination, where geometrically feasible. The automatic system utilizes the Ultra Image III computer-driven data acquisition system with the ALARA remote scanning device. Manual examinations were perfonned where automatic UT could not be performed. The sizing of IGSCC indications was performed using 2 to 5 MHz refracted longitudinal transducers. The sizing methods used included satellite pulse observation technique (SPOT), pulse arrival time technique (PATT) ID and OD creeping wave technique, and tip diffraction techniques, as applicable.

During this outage, flaw indications were found its tw recirculation riser welds (12ARD2 and 12 ARES). Both flawed welds were overlay repaired. An axial flaw with 0.8 inch in length and a maximum depth of 605 through-wall was reported in weld 12ARD2 (pipe to elbow). Since weld 12AR02 was previous-ly treated with induction heating stress improvement (IHSI), the reported cracking of this weld after IHSI has raised some concern regarding the ef-festiveness of INSI in mitigating IGSCC. Therefore, the staff recommends that the licensee should inspect these IHSI treated welds with increased sampling during the next few refueling outages to ensure the structural integrity of these piping aids. In the safe end-to-nozzle weld 12 ARES, the major extent of the reported flat; wh located in the Inconel 182 butter adjacent to the nozzle. The length of the flaw was 4.8 inches with a maxi-mum throughwall depth of 85%. The UT data have shown that the flaw appeared to be initiated from the Inconel 182 butter and may have penetrated the

" nose" of the low alloy steel nozzle while growing toward the safe end-to-nozzle weld. Limited examinations were performed on weld 12 ARES in 1984 and 1986;.a portion of the flaw was detected in 1984 but was characterized as an embedded defect.

Two unrepaired suction safe end-to-nozzle welds (28A1 and 2881) were re-examined during this outage. The inspection results showed that the exist-ing flaws (one axial flaw in each weld) did not change in size (about 0.3

inch in length and a maximum deptn of 0.25 inch) after operation of one fuel cycle. However, a new axial flaw (0.2 inch 'in length and 0.2 inch in maxi-mum depth) was found in weld 2881. Both welds were treated with mechanical J stress improvement process (MSIP) during this outage for IGSCC mitigation. l NRC Region II inspector selectively reviewed the ultrasonic examination procedures and data, and held discussions with the examiners regarding the non-destructive examinations performed during the refuel 7 outage. The inspector concluded in his report numbered 50-325/88-02 and 50-324/88-02, dated March 11, 1988, that nondestructive examinations were performed by qualified personnel and that no violations of NRC requirements were identi -

fied.

Flaw Evaluation Two 28-inch safe end-to-nozzle welds (28A1 and 2881) were found to contain a small axial flaw during the last refuel outage. The flaws were located in the Inconel 182 butters with a maximum depth of 0.25 inch and a length of 0.25 and 0.3 inch in welds 28A1 and 2881, respectively. Fracture mechanics and crack growth evaluations performed during the last outage had shown that these flaws were acceptable without repair for at least an 18-month fuel cycle of operation. As stated above, the results of current inspections have shown that these axial flaws did not grow in size after operation of one fuel cycle. However, a new axial flaw with a length of 0.2 inch and a maximum depth of 0.2 inch was found in weld 2881. The new flaw was. located about 16 inches azimuthally from the original flaw. Since the new flaw was smaller than the original flaws.. its potential growth is bounded by the previous calculations per-formed for the original flaws. Mechanical Stress Improvement Process (MSIP) was appled to these two welds 'during this outage to mitigate the IGSCC. Therefore, we conclude that welds 28A1 and 2881 are acceptable for operation of at least an 18-month fuel cycle in its present configuration.

Tn embedded circumferential flaws were found in the weld overlay of weld 2883 after completion of overlay upgrade and surface finish improvement.

One flaw is 7.2 inches long and 0.1 inch wide and has the characteristics of lack of bonding of overlay to the base metal.. The other flaw is 0.3 inch long and 0.1 inch in through-wall dimension and has the characteristics of lack of fusion of interbead. The Ifcensee performed flaw evaluation in accordance with the requirements of ASNE Code Section XI, IWB-3500. The results of the evaluation have shown that these flaws are acceptable without repair.

The staff concludes that these flaw evaluations are acceptable.

Weld Overlay Repair During this outage, two new weld overlays with standard design were applied to welds 12ARD2 and 12ARE5; and three previously overlay repaired welds 28A4, 28B3 and 2884 were upgraded to meet the requirements of standard

n  !

< ( l overlay design and improved in the surface finish to facilitate ultrasonic examination. Structrual Integrity Associates, Inc., (SIA) performed the overlay design for the licensee. The designed minimum overlay thickness does not include the first overlay layer that passed the dye penetrant exam-ination. The licensee reported that the as-built overlay thickness in these welds exceeds the minimum designed dimensions. SIA has evaluated the weld overlay shrinkage induced stresses in the recirculation piping system, as a result of 41 weld overlay repairs. The largest shrinkage stress on un-flawed welds was reported to be 8743 psi on weld 12AREl, which is within the ASME Code allowable.

A composite Inconel weld overlay was designed for weld 12ARE5 (safe end-to-nozzle). The downstream side of the overlay was extended into the nozzle for covering the flaw in the butter and the nozzle. At the nozzle side of the overlay, the first three layers were deposited by a qualified Inconel temper bead technique. The piping was running dry while depositing the temper bead layers. The temper bead technique has been demonstrated to provide adequate tempering of the heat affected areas of the low alloy steel materials. The designed overlay thickness did not include the intial three temper bead layers. After depositing the temper bead layers, the normal overlay repair technique with water running in the pipe was used for fabri-cating the rest of the overlay. The Inconel temper bead-technique was deve-loped by EPRI for weld overlay repair of P-3 components to preclude the post-weld heat treatment. The repair procedures were established in accor-dance with the requirements in AS4E Code Section XI and Code Case N-432.

Mechanical testing including tensile, bending, hardness and toughness tests were performed on a 12-inch nozzle mock-up to confirm that the simulated Inconel overlay repair met the Code requirements.

Because extensive cracking was recently reported in the recirculation riser nozzles and the thermal sleeve to safe-end attachment welds at Brunswick Unit 1, the staff recomunends that this weld overlay repaired weld (12 ARES) and the associated thermal sleeve attachment weld should be inspected dort.g next refueling outage to ensure that the integrity of the safe-end and nozzle is maintained.

Weld reinforcement was applied to the fillet socket weld of a 3/4-inch JPI penetration seal connection joint. This weld joint was found leaking during this outage. The repair consisted of applying a minimum of 0.17 inch weld metal on the penetration fitting side and 0.375 inch on the tube side. The reported thickness of the repair weld exceeded that of the original fillet weld. The liquid penetrant examination was performed after repair and no surface defect was found.

The staff concludes that the weld overlay repairs performed during this out-age are acceptable for at least an 18-month fuel cycle. The staff also concludes that the repair of a leaking JPI penetration seal weld is accept-able for continued service.

~

.c

>4

-S-Mechancial Stress Improvement Process (MSIP)

O'Donnel and Associates, Inc. (00AI) performed MSIP for the licensee on 15 Inconel buttered nozzle-to-safe end welds which consist of two 28-inch re-circulation outlet nozzle welds, nine 12-inch recirculation inlet nozzle welds, two 10-inch core spray nozzle welds and two 4-inch JPI nozzle welds.

MSIP is a mechanical process that replaces tensile residual stresses on the inside surface of the piping in the vicinity of welds with a zone of compres-sive residual stresses. The MSIP processes applied to different sizes of riozzle welds were analytically verified by 00AI using a nonlinear finite element analysis of the processes to show that the permanent strain induced i by MSIP in the range of 0.5% to 1.195 is adequate and would produce the de-sired redistribution of the residual stresses. The ultrasonic examination was performed on each weld before and after MSIP, No additional flaws were observed other than those already reported in two 28-inch nozzle welds.

Based on the licensee's submittals, the staff notes that the MSIP process applied to the nozzle-to-safe end welds was only analytically verified with-out being validated by a confirmation test on a mock-up with similar safe end-to-nozzle configuration. Because of limited field experience, the staff .

has.some reservation regarding the effectiveness of this particular applica- )

tion. Furthermore, the recent Unit 1 inspection results have shown that extensive cracking was found in 7 recirculation riser nozzles and 10 thermal sleeve-to-safe end attachment welds. Therefore, the staff recomends that these MSIP treated nozzle welds and the associated thermal sleeve-to-safe end. attachment welds be int,pected with increased sampling during the next several refueling outages to ensure that the integrity of the nozzles and safe-ends is maintained.

In NUREG 0313, Revision 2, the stress improvement (SI) process including MSIP is considered effective in mitigating IGSCC for welds with no flaws or with only minor circumferential flaws. SI credit for inspection is not allowed for welds with axial flaws because there are uncertainties in the sizing of the axial flaws. Therefore, MSIP treated welds 28A1 and 2881 containing axial flaws should be inspected in accordance with the IGSCC Category F schedule. After successful completion of four consecutive inspections, these welds may be moved up to Category E.

Special Surveillance Measures The licensee indicated that the following special surveillance measures for unidentified leakage will continue to be implemented at Unit 2:

(1) Plant shutdown shall be initiated 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 />, an increase in the rate of unidentified leakage in excess of 2 gpm is identified.

(2) The leakage shall be monitored at 4-hour intervals or less.

(3) The outage time for inoserable instruments for leakage measurements shall be limited to 24 tours. Otherwise, an orderly shutdown will be  !

ionediately initiated.

i

i i,.'

The staff finds that these special surveillance measures are consistent with the guidelines in Generic Letter 88-01.

3.0 Conclusion Based upon the staff's review of the licensee's submittals, the staff con-cludes that the licensee has adequately addressed IGSCC in Class 1 piping with respect to inspections, repairs and litigations performed during the Unit 2 refuel 7 outage, and that these activities were performed in accordance with the guidelines in Generic Letter 88-01. In addition, the staff also concludes that Unit 2 can be safely operated for another 18-month fuel cycle in the present configuration.

Principal Contributor: W. Koo I

l l

- _ _ _ _ _ - - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ l