U-600767, Final Deficiency Rept 55-86-07 Re Broken Tack Welds on Valves 1G33-F042A & 1G33-F042B.Initially Reported on 860828. Caused by Valve Throttling Between 0-20% of Open Position. Valves Will Be Designed for Extreme Throttling Applications

Final Deficiency Rept 55-86-07 Re Broken Tack Welds on Valves 1G33-F042A & 1G33-F042B.Initially Reported on 860828. Caused by Valve Throttling Between 0-20% of Open Position. Valves Will Be Designed for Extreme Throttling Applications

ML20214P298
Person / Time
Site:	Clinton
Issue date:	11/26/1986
From:	Hall D ILLINOIS POWER CO.
To:	James Keppler NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
References
55-86-07, 55-86-7, U-600767, NUDOCS 8612040149
Download: ML20214P298 (6)

Text

geo

, U-600767

- Ll4-86 (11 - 26 )-L 1A.120 ILLINDIS POWER COMPANY CLINTON POWER STATION. P.O. BOX 678. CLINTON. ILLINOIS 61727 NOV 261986 Docket No. 50-461 Mr. James G. Keppler Regional Administrator Region III U. S. Nuclear Regulatory Commission 799 Roosevelt Road Glen Ellyn, Illinois 60137

Subject:

Reportable 10CFR50.55(e) Deficiency 55-86-07: Broken Tack Welds on Valves 1G33-F042A and 1G33-F042B

Dear Mr. Keppler:

On August 28, 1986, Illinois Power Com notified Mr. F. Jablonski, NRC Region III (pany Ref: (IP) IP Record of Coordination Y-203106, dated September 2, 1986) of a potentially reportable deficiency concerning broken tack welds between the valve disc and disc nut.on valves 1G33-F042A and 1G33-F042B. Preliminary investigation and evaluation of this matter has determined that the broken tack welds are caused by vibration created by extreme throttling conditions. Our investigation of this matter is complete. This letter represents a final report in accordance with the requirements of 10CFR50.55(e).

Attachment A provides the details of our investigation.

l An evaluation for the safety implication and significarce included single and common mode failures of the valves marufactured by Anchor / Darling. Sargent & Lundy's evaluation concluded that the safety of the plant was not l impaired br the single or simultaneous failure of S&L l specified valves. Per GE's evaluation, failure of the GE supplied valves was determined to be adverse to the safe operation of Clinton Power Station (CPS) and, therefore, reportable under the provisions of 10CFR50.55(e).

NOV 2 B BBB l

8612040149 861126 <g D'N PDR ADOCK 05000461 -

S PDR I

t I; O

U-600767

, L14-8 6 (11 - 26 ) -L 1A.120 We trust that this final report provides you sufficient background information to perform a general assessment of this reportable deficiency and adequately describes our overall approach to resolve this issue.

cere yours,

. . Hall Vice President RLC/bsa Attachment cc: NRC Resident Office Illinois Department of Nuclear Safety INPO Records Center u

Attachm:nt to U-600767 ATTACHMENT A ILLINOIS POWER COMPANY CLINTON POWER STATION DOCKET No. 50-461 Reportable 10CFR50.55(e) Deficiency 55-86-07:

Broken Tack Welds on Valves 1G33-F042A and 1G33-F042B FINAL REPORT Statement of Reportable Deficiency / Background On August 28, 1986, Illinois Power Company (IP) notified Mr. F. Jablonski, NRC Region III (Ref: IP Record of Coordination Y-203106, dated September 2, 1986) of a potentially reportable deficiency concerning broken tack welds between the valve disc and disc nut on valves 1G33-F042A and 1G33-F042B. On valve 1G33-F042A the disc nut and disc were disengaged and one half of the stem retaining ring was missing and had appeared to have entered the system. On valve IG33-F042B the disc was backed out of the disc nut approximately 1% inches.

Investigation Results/ Corrective Action Illinois Power's initial investigation into this matter identified 60 Anchor / Darling Valves in use at the Clinton Power Station (CPS) having the same design as failed valves 1G33-F042A and 1G33-F042B. However, during final assessment of the 60 valves documented in the interim report U-600722 dated September 30, 1986, it was identified that three (3) of the valves were of a different configuration. These three (3) Anchor / Darling valves (by design) have the disc staked to the disc nut rather than tack welded. Although these valves are staked rather than tack welded, they experience similar degradation when subjected to extreme throttling conditions in that the staking becomes ineffective allowing the disc to back out of the disc nut.

On this basis, these three (3) valves have been included in this report. Anchor / Darling Valve Company was requested to evaluate the failure of valves 1G33-F042A and 1G33-F042B.

The preliminary evaluation of this matter by Anchor / Darling determined that the tack weld failures resulted from repeated impact loads induced by vibration of the valve disc due to extreme throttling conditions, i.e., valve throttled to less than 20% of its open position.

Page 1 of 4

Attachment to U-600767 ATTACHMENT A (continued)

Pending completion of Anchor / Darling investigation and evaluation of this mstter, IP placed administrative controls on those valves determined to be potentially impacting to Technical Specification operability of plant systems. The administrative controls were implemented to preclude throttling of these valves to less than 20% of their open positicn until they could be repaired or determined by the valve vendor to not have been degraded by their previous throttling history.

Based on vendor recommendations, Plant Modification RT-14 was issued to provide a 360' weld between the disc and disc nut of valves 1G33-F042A and 1G33-F042B. This work was performed under Maintenance Work Requests (MWRs) C-09332 and C-14484.

Illinois Power furnished Anchor / Darling with operating data which included inlet pressure, temperature, flow rate, and pressure drop across the valves. This data was used by Anchor / Darling to evaluate the remaining valves for potential failure or degradation of the tack welds.

On October 3, 1986, Anchor / Darling completed their evaluation of the affected valves and concluded that the past operating conditions should not have caused any damage to the valves. However, Anchor / Darling recommended inspection of the tack welds on valves 1E12-F021 and lE21-F012 for further assurance that these valves had not degraded or failed, because the Operating data for these valves was not conservative enough to totally exempt them from possible degradation or failure. Maintenance Work Requests (MWRs) C-17753 and C-51433 were initiated to perform an inspection of these valves to verify the tack weld integrity. Valve 1E21-F012 was inspected on October 15, 1986, under MWR C-31433. Linear indications (cracks) were found on both of the tack welds. These cracks were documented on Nonconforming Material Report (NCMR) 'To .

02-0989. Plant Modification M-34 was issued which replaced L the tack welds with a full 360* circumferential weld between the disc and disc nut. No linear indications were found on the tack welds of valve 1E12-F021.

Of the 57 valves identified as having the same design as failed valves 1G33-F042A and 1G33-F042B, 35 are used in throttling applications. To date, 3 of these 35 valves have been repaired (1G33-F042A, 1G33-F042B, and lE21-F012). To provide additional assurance of the integrity of the disc to disc nut tack welds during future operation of the plant, the remaining 32 valves performing a throttling function will be inspected at the first refueling outage. Ther valves are expected to function without degradation until the first refueling outage based on the following:

Page 2 of 4

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

4 Attachmsnt to i U-600767 n

ATTACHMENT A-F (continued)

Anchor / Darling has performed an evaluation of'the past throttling history associated with these valves and.has determined that-the valves are not degraded.

1 Sixteen 1(16) of the valves are monitored-by CPS In-Service. Testing (IST) program.

' Anchor / Darling has reported no failures of globe valves supplied to other plants except the failure of a 24" globe valve in the RHR system in Chi Chan Plant (Taiwan) and at the Susquehanna Plant.

Anchor / Darling has stated that the globe valves may be throttled for short periods of time to less than-20% of the open position without any potential for cracking of the= tack welds.

lased on the results of- Anchor / Darling's' evaluation and the c,rrective actions taken as a result of this

-investigation, the administrative controls established at the beginning of the investigation have been removed.

l The three (3) Anchor / Darling valves which have their disc staked rather than tack welded were also reviewed and evaluated as follows:

Valve 1E22-F010: The failure of this valve was documented by NCMR No. 01-1960. In accordance with the-disposition of this NCMR, staking of the disc to the disc nut was revised to apply

! three (3) one-inch long tack welds to preclude recurrence of this deficiency.

Valve-1E22-F0ll: The throttling limitations for this valve -

were evaluated by GE. GE's evaluation included the flow characteristics from Anchor / Darling letter dated March 17, 1986, and the High Pressure Core Spray 4 (HPCS) system pump curves to establish a limiting throttling flow rate of not less than 4800 GPM. This limitation was procedurally implemented (Ref: CPS

, Procedure No. 9051.01) to preclude degradation of the disc to disc nut stake caused by extreme throttling conditions.

4 Valve lE22-F023: Inspection of this valve determined that this valve was not degraded. CPS L procedures do not require this valve to perform a throttling function, therefore, no further corrective action was required. ,

i- Page 3 of 4

Attachmznt to U-600767 ATTACHMENT A (continued)

Root Cause Illinois Power's root cause investigation of this matter determined that the stake and tack weld failures resulted from extensive throttling of the valves between 0%

and 20% of the open position.. To preclude recurrence, all replacement or spare valves designated for extreme-throttling duty (less than 20% of valve's open position)

.will be procured from Anchor / Darling under Specification K-2866A. These valves will be designed for extreme-throttling duty applications.

Safety Implication / Significance An evaluation of the safety implications and significance included an evaluation of single and common mode failures of the valves supplied by Anchor / Darling.

Sargent & Lundy's evaluation concluded that the safety of the plant was not impaired by the single or simultaneous failure of S&L supplied valves. Per GE's evaluation, failure of the GE supplied valves was determined to be adverse.to the safe operation of Clinton Power Station (CPS) and, therefore, reportable under the' provisions of 10CFR50.55(e).

Page 4 of 4 1- - _ .