ML021120126
| ML021120126 | |
| Person / Time | |
|---|---|
| Site: | Nine Mile Point  |
| Issue date: | 04/04/2002 |
| From: | Grewal K - No Known Affiliation |
| To: | Collins S Office of Nuclear Reactor Regulation |
| References | |
| Download: ML021120126 (7) | |
Text
Mr. Samuel J. Collins Khushwant S. Grewal Director 114 Coopers Kill Road Office of Nuclear Reactor Regulation Delran, NJ 08075-2008 U. S. Nuclear Regulatory Commission Phone: 856-764-0426 11555 Rockville Pike April 4, 2002 Rockville, MD 20852-2738 5o-410
Dear Sir:
Thank you for your letter dated May 5, 2000 (Attachment 1) in response to my letter dated March 16, 2000 (Attachment 2).
In my letter I had raised the following concerns regarding a crack at weld KC-32 joining the HPCS nozzle safe end to the safe end extension at Nile Mile Point 2 (NMP2) Nuclear Power Station:
"* Due to high compressive stress at the crack, resulting from a combination of additional deadweight of lead shielding and not pinning of the constant support hanger near the nozzle, the crack depth has been under-estimated by ultrasonic testing.
"* Stress due to thermal stratification in the horizontal piping adjoining the weld, has not been taken into account in evaluating the stress intensity at the crack.
"* Since the beneficial residual stress resulting from MSIP, may be dissipated by low cycle fatigue, it is not conservative to rely on MS1P for the integrity of an emergency core cooling system (ECCS} whose failure will jeopardize another ECCS. At NMP2, the HPCS nozzle also serves as the RPV inlet for the Standby Liquid Control (boron injection) System.
In responding to these concerns, NRC stated:
"The staff believes that any issues of the type you raise would be evidenced by significant changes in ultrasonic measurements or by trends in ultrasonic measurements that would indicate a developing problem."
A review of the ultrasonic measurements indicates a trend that supports my concerns (page 3). The decrease in the measured crack depth during early nineties, confirms my concern that ultrasonic crack depth measurement has been adversely affected by the compressive stress at the crack resulting from a combination of lead shielding application and not pinning of nearby constant support hanger. The decrease in in crack depth from 0.35" to 0.25" (29 %) is too large to be ignored as a measurement uncertainty - in the fracture mechanics analysis submitted on June 28, 1991, only 5% uncertainty was applied to UT measurements.. The effect of compressive stress on UT measurements, provides a better explanation for the observed trend.
A1/2,
Up until 1991, for the application of lead shielding, the constant support hangers were pinned and the piping was qualified for only deadweight load at NMP2. Around 1991, it was realized that piping required to be seismically qualified for lead shielding application. Since constant support hangers in the pinned condition could.not be qualified for seismic loads, pinning of constant support hangers during lead shielding application was discontinued. As the piping collected more radiation, the amount of lead shielding was progressively increased to a maximum value which could be qualified.
After the amount of lead shielding reached its maximum value in October, 1993, the negative effect of lead shielding on crack depth measurement remained constant, and as indicated by the last two measurement, further crack growth became discernable.
Ignoring crack growth from Dec. 1990 to Oct. 1993, a lower bound estimate of actual crack depth = 0.36 + (0.35 - 0.25) = 0.46" or 0.54t. The progressive increase in lead shielding weight and pinning/not pinning of constant support hanger, can be determined by reviewing relevant work requests.
Based on the temperature difference between the top and the bottom of the piping adjoining the nozzle, I expect the thermal stratification stress to be quite significant.
Since it adds on to the thermal expansion stress, it should be included in the evaluation of stress intensity at the crack. An estimate of the top and bottom temperature of the HPCS piping adjoining the RPV nozzle, is provided in a Safe Guards calculation, referenced in the Pipe Stress Data Package for the HIPCS system. Thus thermal stratification stress at the crack can be readily determined and included in crack evaluation.
Emergency Core Cooling Systems are designed to achieve a higher standard of reliability and the maximum fatigue usage factor is limited to 0.1 instead of 1.0 for other class 1 systems. I think GL 88-01 does not take into account the reliability requirements of the HPCS system.
On the outside surface at the crack location, residual tension stress resulting from MSIP, thermal expansion stress, thermal stratification stress, pressure stress, and any dynamic stresses can add up to push the total stress into the plastic range and diminish the beneficial residual stress induced by MSIP.
In view of the above concerns, I hope that during the current refueling outage, NRC will require:
"* A thorough examination of the crack, with the nearby constant support hanger pinned.
"* Inclusion of thermal stratification load in crack evaluation.
"* Re-examination of the applicability of GL 88-01 to the HPCS system.
Sincerely, Khushwant S. Grewal 2_
NITACIIMENTA KC-32 Inspection Results MIN UA L AqTOfr1ATEDI 4At'MAL.
Depth/%
L_* Vri Length/% of of Wall Inspectio ATO M A 16 D Inspection LENacrJ Internal Circumference Thickness 3.40!
0.35"
- Post-MSIP (RFOI) (December 1990)
(11.3%)
(41%)
3.3" 0.32" Midcycle (August 1991)
(10.9k)
(38%)
7 2.6" Ac,,,mated 3.3" Manual 0.25" RFO-2 (April 1992)
(8.6.,,
(10.9%)
(29%)
2.5" Automated 3.0" Manual 0.25" RFO-3 (October 1993)
(8.3%)
(9.9%)
(29%)
2.5" Automated 3.0" Manual 0.30" RFO-4 (May 1995)
(8.3%)
(9.9%)
(35%)
p,FO-5(IM\\" t1qq9)
I.Zo
.3.Z A4ToArEp-3,8 *f oAL 036',
0.30o 0_o,_6 X__)_(12 (io,/;i (4-aI,)
(s/)
I A ESTJIMkTET Cannot be counted toward four successive examinations (i.e., the weld experienced no service).
A.
11-,3
,'**..A T 7 rCkl M T
c-(L UNITED STATES A-r NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 May 5, 2000 Mr. Khushwant S. Grewal 114 Coopers Kill Road Delran, NJ 08075-2008
Dear Mr. Grewal:
I am responding to your letter of March 16, 2000, to Mr. Richard A. Meserve, Chairman of the U.S. Nuclear Regulatory Commission (NRC). In your letter, you expressed concerns about a weld flaw in the Nine Mile Point Nuclear Station, Unit 2, reactor pressure vessel high-pressure core spray (HPCS) nozzle. We telephoned you on April 28, 2000, and confirmed that you were addressing weld KC-32. Between 1990 and the present time, the licensee (Niagara Mohawk Power Corporation) and the NRC have done a significant amount of review on this weld flaw, culminating in NRC safety evaluations dated June 22 and August 27, 1993, and February 12, 1996 (Enclosures 1, 2 and 3).
Weld KC-32 is a weld joining the HPCS nozzle safe-end to the safe-end extension. During the first refueling outage, the licensee detected a flaw indication by using ultrasonic testing (UT) techniques. The size of the flaw indication was reported to be 0.15 inch deep and 1.9 inches long. During the same refueling outage, the licensee applied a mechanical stress improvement process (MSIP) on this weld. The MSIP redistributed the residual stresses in the weld, thereby inducing compressive residual stresses at the inside diameter surface of the weld and continuing through about half of the wall thickness. Compressive residual stresses are desirable because they tend to mitigate the flaw growth since tensile stresses are needed for intergranular stress corrosion crack (IGSCC) propagation. After the application of MSIP, the subject weld was ultrasonically re-examined and the flaw indication was reported to have a size of 0.35 inch deep and 3.4 inches long. The change in measured size of the indication was most probably due to redistribution of stresses associated with the flaw, resulting in shifting of the existing flaw surfaces relative to an occluded oxide layer. This resulted in additional reflectivity in the subsequent ultrasonic measurement. It is also possible the flaw extended as a result of MSIP. Nonetheless, the flaw size subsequent to MSIP continued to meet American Society of Mechanical Engineers Boiler and Pressure Vessel Code safety limits such that repair was not required.
Subsequent to the post-MSIP inspection (first refueling outage, December 1990), the licensee performed an ultrasonic examination of weld KC-32 at a mid-cycle outage (August 1991) during the second fuel cycle, and at the second (April 1992), third (October 1993), fourth (May 1995) and sixth refueling (May 1998) outages. Results were reported in the licensee's letters dated September 22, 1995 (Enclosure 4) and April 7, 2000 (Enclosure 5). The reported flaw sizes vary from 29% (0.25 inch) to 41% (0.35 inch) of wall thickness (0.85 inch) in depth and 8.3%
(2.5 inches) to 11.3% (3.4 inches) of circumference in length. These ultrasonic inspection results for the flaw were bounded by the inspection results from the original post-MSIP inspection, thus serving to reinforce the positions that the crack has not grown since it was originally characterized, and that stabilization of the flaw occurred due to application of MSIP which reduced the driving force for flaw extension.
K. Grewal The subject weld was categorized as IGSCC Category F after the application of MSIP.
Based on Generic Letter 88-01 (GL 88-01), Category F welds are welds that are without any IGSCC mitigation, and require an inspection every refueling outage. After MSIP, a flawed weld would normally be categorized as a Category E weld (flawed welds mitigated either by weld overlay repair or by stress improvement such as MSIP) and require an inspection at every other refueling outage. However in this case, due to the flaw size results determined by ultrasonic inspection after MSIP, the weld remained categorized as F, thereby requiring inspection every refueling outage. The guidelines in GL 88-01 allow Category F welds to be upgraded to Category E after four successive examinations indicate no significant change in the flaw size.
The staff approved the upgrade of weld KC-32 from Category F to Category E by its safety evaluation dated February 12, 1996. This is based on the consideration that no apparent crack growth was detected in the four successive examinations subsequent to the first post-MSIP inspection. The first examination after the upgrade was performed at the sixth refueling outage and no significant flaw growth was detected. The next inspection is scheduled to be performed at the eighth refueling outage (2001).
During the past 10 years, six ultrasonic examinations have been performed on weld KC-32 to monitor the condition of the flaw. The results of these inspections affirmed the basis for the upgrade approved by the staff's February 12, 1996, safety evaluation. These results demonstrate that the MSIP applied to weld KC-32 was effective in mitigating the IGSCC. The licensee's current inspection schedule for weld KC-32 is consistent with the guidelines in GL 88-01. The licensee will continue to inspect weld KC-32 at a frequency of once every two refueling outages to ensure that the structural integrity of the weld is maintained.
In your letter you also specifically commented on the potential relaxation of the residual stresses and the loading effect on flaw measurement. These comments are briefly summarized below:
(1)
The effect of thermal fatigue resulting from thermal stratification could relax the beneficial residual stress induced by the MSIP treatment.
(2)
The loading resulting from pinned piping hanger support and lead shielding adjacent to the flawed weld could affect the results of ultrasonic examination.
The staff believes that any issues of the type you raise would be evidenced by significant changes in ultrasonic measurements or by trends in ultrasonic measurements that would indicate a developing problem. Such has not been the case with weld KC-32. Should subsequent measurements indicate a problem with KC-32, the licensee is obligated to effect a weld overlay repair (Enclosure 6).
In summary, between 1990 and the present time, the NRC staff has carefully monitored the licensee's inspection results and evaluation of weld KC-32. The results of the staff's review are documented in the safety evaluations cited above. Consistent with its commitment and as approved by the staff's safety evaluation dated February 12, 1996, the licensee did not plan to examine weld KC-32 during refueling outage 7. Currently, no new information has been identified that would necessitate a revision to the NRC-approved examination program. At this K. Grewal time, based on the information available, we conclude that there is reasonable assurance that the unit can continue to be operated safely. If you have any questions, please contact the NRC project manager, Mr. Peter Tam, at 301-415-1451, e-mail pst@nrc.gov.
Sincerely, Office of Nuclear Reactor Regulation
Enclosures:
- 1.
- 2.
- 3.
- 4.
- 5.
6.
Letter, J. E. Menning to B. R. Sylvia, June 22, 1993 Letter, J. E. Menning to B. R. Sylvia, August 27, 1993 Letter, G. E. Edison to B. R. Sylvia, February 12, 1996 Letter, C. D. Terry to NRC, September 22, 1995 Letter, R. B. Abbott to NRC, April 7, 2000 Letter, C. D. Terry, to NRC, July 8, 1993 77H-6 H ME 0-r 2 Mr. Richard A. Meserve Khushwant S. Grewal Chairman 114 Coopers Kill Road U. S. Nuclear Regulatory Commission Delran, NJ 08075-2008 One White Flint North Phone: 856-764-0426 11555 Rockville Pike March 16, 2000 Rockville, MD 20852-2738
Dear Sir:
This is to bring to your attention, my concerns regarding a crack in the RPV nozzle for the High Pressure Core Spray System of the Nine Mile Point Unit 2 (NMP2) nuclear station at Lycoming, NY.
In 1990, a circumferential crack extending over about 20 percent of the circumference with a maximum depth of about 0.4 inch from the inside surface, was detected in the bottom side of the weld between the nozzle and the safe end To stabilize the crack, MSP treatment - radial compression to induce residual compressive stress at the crack tip - was performed. However, ultrasonic examination after the MSP treatment indicated significant crack extension. Based on the evaluations by GE and the MSP vendor, NRC allowed the plant to be started on the condition that the crack be examined after about 8 months.
Subsequent examinations showed the crack size to be smaller than that indicated immediately after the MSP treatment. Consequently NRC, in due course, relaxed the examination requirement to every other refueling outage.
For radiation protection during ultrasonic examination, the High Pressure Core Spray piping adjoining the RPV nozzle is covered with heavy lead blankets. This results in high compressive dead weight stress at the crack and an under-estimate of the size of the crack. The larger crack size indicated after MSP treatment, which was disregarded based on later measurements, was probably realistic for the following reasons:
- a.
The lead shielding weight was lower at that time.
- b.
The constant support hanger in the vicinity of the nozzle was pinned- (During subsequent examinations, the hanger was not pinned because it could not be qualified for the seismic load in the pinned condition.)
- c.
According to GE, MSP treatment at Peach Bottom nuclear plant had resulted in crack extension requiring weld repair.
The long horizontal run of the High Pressure Core Spray piping adjoining the RPV nozzle, is subject to significant thermal stratification during plant operation. The thermal stratification load is not considered in piping stress analysis and consequently in the crack evaluation.
Over a period of time, the beneficial residual stress induced by the MSP treatment may be dissipated by fatigue cycling.
At NMP2, the High Pressure Core Spray nozzle also serves as the RPV inlet for the Standby Liquid Control (boron injection) System. Hence the integrity of this nozzle is of vital safety significance.
In view of the above concerns, I hope that during the current refueling outage, NRC will require a thorough examination of the crack, with the hanger pinned, and the lead shielding distributed on either side of the hanger so as to minimize compressive stress at the crack.
If you have a question, please call.
Sincerely, Khushwant S. Grewal Xc: The Plant Manager, NMP2