ML20127C219

Supplemental Deficiency Rept CP-92-010 & Final Deficiency Rept CP-92-020 Re Cracks in Containment Spray Pump Suction Vent Piping,Initially Reported on 920722 & 1204, Respectively.Unacceptable Configurations Modified

ML20127C219
Person / Time
Site:	Comanche Peak
Issue date:	01/07/1993
From:	William Cahill, John Marshall TEXAS UTILITIES ELECTRIC CO. (TU ELECTRIC)
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
CP-92-010, CP-92-020, CP-92-10, CP-92-20, TXX-93010, NUDOCS 9301140060
Download: ML20127C219 (5)

Text

. _ _ .

E"J"J Log # TXX 93010 C

file # 10110 (CP 92 010)

-- 2 # 10110 (CP 92-020)

F E # 903.9 TUELECTRIC M 4%ta J= Cubill, Jr, ,

r.wu , n, .a w 7

U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555

SUBJECT:

C0HANCHE PEAK STEAH ELECTRIC 31AT10H (CPSES) - UNIT 2 DOCKET NO. 50 446 CRACKS IN THE CONTAINHENT SPRAY SYSTEM SDAR: CP 92 010 (SUPPLEMENTAL REPORT )

SDAR: CP-92-020 (FINAL REPORT)

REr: 1) TU Electric letter to the NRC dated October 30, 1992 logged TXX 92509 (SDAR CP 92-010)

2) TV Electric letter to the NRC dated January 4, 1993 logged TXX-92636 (#DAR CP-92 020)

Gentlemen:

On July 22, 1992, vla facrimile, 10 Electric notified the NRC of a reportable defect in two welds on 3/4' containment spray pump suction vent piping. The evaluation was designated SDAR CP 92-010, and Reference 1 contains the final report.

On December 4,1992, via f acsimile. TV Electric notified the NRC of a reportable defect involving a 90 degree circumferential crack found in the base metal along the toe of a weld in the chemical eductor portion of the containment spray system. The evaluation was designated SDAR CP-92 020. and an interim report was submitted en January 4, 1993 (Reference 2).

Two additional cracks were discovered caring investigations on November 17-and December 8, 1992.

In all, five TVE forms (non-conformance reports) have been generated to document these cracks in welds and base metal. Because of the significance and repetitive nature of these deficiencies. TV Electric conducted a Root Cause Analysis in accordance with established Site procedures.

9301140060 930107 u0089 [ ,

DR ADOCK 0500 6 ,yg nigg ,um gng g gi j  ;

o

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

4 TXX-93010 Page 2 of 2 i

This letter serves as a supplemental report for SDAR CP-92 010: a final

report for SDAR CP-92 0EO, satisfying the reporting requirements of J 10CIR50.55fe); and, provides the results of the analysis of all five

, deficienties as Attachment 1.

Sincerely, t

William J. Cahill, Jr.

Byt /k

'd S. Marshall Generic Licensing Manager CBC/grp j Attachment i i

c- Mr. J. L. Hiihoan, Region IV j Hr. B. E. Holian, f1RR j Resident inspectors. CPSES (2) 1 3  !

a i

e t

I

. - =- - ___ _

l w t Attachment to TXX-93010 Page 1 of 3 Deficiency ident i f i_s a ti on All of the deficiencies were identified in the Containment Spray System.

The first crack which developed on this System (on June 14, 1992) was documented in TUE 92 S473. This crack occurred on a 3/4' diameter branch line weld of f loop 3 while the loop was undergoing flushing prior to Hot functional Testing (HfT). The defective weld and two adjacent piping sections were cut out, replaced, and forwarded to Engineering for review.

l A second through-wall crack developed six days later, on June 20, 1992, on the same branch line, This crack, documented in TVE 92 5529, was a weld metal crack approximately l' long at the sock-o-let (SOL) weld of the 3/4" branch line to the main header. The disposition of this TVE was the same as the previous one.

The deficient welds and piping s'ections were sent to a metallurgical laboratory for failure analysis. The lab issued a report dated August 11, 1992, which concluded that the cracks propagated due to vibration induced fatigue. This was not an unexpected conclusion since visual observations of the system in operation confirmed that there was significant vibration at a number of locations No further incidents of this type occurred prior to or during HF1. During preoperational vibration testing on October 9, 1992, a circumferential crack developed on the 2* diameter eductor line on loop 4, as documented in TUE 92 6536. This particular line was not in formal preoperational testing at the time the crack developed, but the pump on loop 4 was running, T he -

damaged sections of piping were replaced and the parts were forwarded to Engineering for evaluation. Metallurgical evaluation at the. laboratory was also performed on this specimen. Engineering, using the lab report as input, issued their investigation report on November 11, 1992, which r concluded:

o The crack in the eductor line was caused by excessive steady state 4 vibration.

o The pump is the (primary) source of vibration, o An added temporary tie-back support brought the vibration levels within-allowable limits.

~

o Vibration prior to installation of the tie back support did not cause any fatigue damage to the remaining piping.

o The steady state vibration' testing program would have identified and corrected the excessive vibration during the normal course of pre-operational testing, in accordance with the Project procedures.

prior to proceeding further with system testing Engineering implemented two additional steps. First, a tie-back support was designed and issued at each

_ . . - _ . _ _ . _ _ _ _- . _ _ _ _ _ . _ .. _ _ _ .. . _ _ _ _ . - _ _ . _ _ . _ ~ _ __ . _ __ __

Attachment to lxX-93010 Page 2 of 3 of the two locations where the previous failures occurred, lhese tie-back supports would help to keep the branch pipint., in phase with the header piping and thus not allow the vibration to amplify. Secondly, Engineering identified locations on all four loops which were susceptible to severe vibration. Accelerometers were placed on these lines and vibration levels which exceeded published allowable limits were reported to engineering, The accepted industry allowable velocity is 0.5 inches per second (ips) based on guidance in ANSI /ASME OM3 - 1982, " Requirements for Preoperational and initial Start-up Testing of fluclear Power Plant Piping Systems.' This same allowable value also appears in site procedure 2PP-5.23, ' Piping Vibration Test Guidelines.' Based on these steps Engineering concluded that system _

configuration was adequate to begin the Integrated Testing Sequence (ITS).

During-preoperational vibration testing and ITS. instrumented vibration data was collected and forwarded to tngineering whenever the velocity acceptance criterion of 0.5 ips was exceede'd. Engineering performed a more detailed analysis f or these cases and established higher acceptable velocities on a case by-case basis, tiear the end of ITS, on flovember 17, 1992, a through-wall crack developed in a SOL weld of a 3/4' branch connection to a main header on loop 1. This crack, documented in TUE 92 6856, was also repaired and the damaged parts were forwarded to Engineering for evaluation. Because of the frequency of occurrence and the similarity of conditions (i.e.,

system in a testing mode, high vibration present, locations at or near S0ls in branch piping, stainless steel piping), a programmatic / repetitive TVE form was initiated.

On December 8, 1992, a through-wall hole developed in the tubing just downstream of the root valve of f the 3/4' SOL branch connection on loop 1 (described above). This hole was created by rubbing of the tubing inside a supporting clamp due to severe vibration, This incident reinforced _the .

previous conclusions that fatigue due to high frequency vibration was a problem in the system and that actions were warranted to improve system performance.

R.pnclusions anti.Aqgt Cause IARQtificatiqn Each of the five piping or weld f ailures was caused by f atigue due to high frequency vibration of the piping system. Based on review of system and-vibration data, it was concluded that the dominant contributor to the overall system vibration was the pump Vane Passing interference-(VPI) frequency of 120 Hz, The following contributing f actors were noted:

1. Unsupported branch lines, whose out-of phase motion tenas to amplify the effect of the vibration, also. contributed.

2. The 100 hour0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> cumulative run time of the pumps, which is a requirement for system testing, introduces over 43 million cycles to the system.

Thus, there is a tignificant accumulation of effect due to high frequency vibration.

e e

Attcchment to TXX 93010 >

page 3 of 3 Gtnttitimpli c a t ions Since the crac6s which developed were a result of fatigue induced by high frequency vibration, all susceptible configurations which experienced this type of vibration were examined, Susceptible configurations were identified as small bore socket welded branch connections on stainless steel piping systems. The entire population of these configurations was found by reviewing all problem reports in which measured piping velocity exceeded 0.5 ips. Nine such configuralions were identifled.

CLrrgqily LA ttign _

The nine unacceptable configurations l<!entified in tJnit 2 have been tuodified to preclude excessive vibration.

A large number of cycles have be'en introduced into the system. Since pre-operational and sthrtup testing of the CT system is essentially corrplete, t her e will not be a significant number of additional new cycles added due to quart erly tests, etc. Since all susceptible connections have been strengthened, the accumulation of f atigue in these connections will be very slow, therefore, no further actions are planned.

1

= = - E- A l

- __ ____ __.___. _ _ _ _ _ _ _ . _ _ _ _ _ . _ _______ _ _ _ _.