ML20217Q688
| ML20217Q688 | |
| Person / Time | |
|---|---|
| Site: | South Texas  |
| Issue date: | 05/06/1998 |
| From: | NRC (Affiliation Not Assigned) |
| To: | |
| Shared Package | |
| ML20217Q681 | List: |
| References | |
| NUDOCS 9805110004 | |
| Download: ML20217Q688 (6) | |
Text
.. .
e h p 1 UNITED STATES g j NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 3085H001 5
g f l
%*****/
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION SUPPLEMENTAL RESPONSE TO NRC BUI I FTIN 88-08. " THERMAL STRESSES IN PIPING CONNECTED TO REACTOR COOLANT LOOP SYSTEMS" STP NUCLEAR OPERATING COMPANY SOUTH TEXAS PRCMCT. UNITS 1 AND 2 DOCKET NOS 50-498 AND 50-499 d.0 BACKGROUND in Reference 1, the NRC provided a safety evaluation (SE) of the revised responses (References 2 and 3) by the licensee (the current licensee is STP Nuclear Operating Company; the previous
. licensee was Houston Lighting and Power, or HL&P, as indicated in some of the references) and Westinghouse (B0 to NRC Bulletin 88-08 and its supplements. W based its response on an analytical methodology developed under a program sponsored by the Electric Power Research Institute to investigate thermal stratification, cycling, and striping (TASCS) in piping connected to the reactor coolant loop (RCL) (References 5 and 6).
Action 3 of the Bulletin requests that licensees plan and implement a program to provide ,
continuing assurance that the unisolable sections of piping connected to the reactor coolant !
system (RCS) will not be subjected to cyclic thermal loading, associated with leaking isolation I valves, which when combined with plant operational cyclic loading could cause fatigue failure during the remaining life of the plant. This assurance may be provided by (1) redesigning and modifying these sections of piping to withstand these loads, (2) instrumenting the piping to detect adverse temperature distributions and establish appropriate limits on these distributions, or (3) j ensure that the pressure upstream of the isolation valves does not exceed the RCS pressure, j l
The licensee initially instrumented the charging and spray lines, in accordance with Ac%n 3 of the l Bulletin. (The safety injection lines were not instrumented, since during normal operation the ;
pressure upstream of the isolation valves is less than the RCS pressure.) Subsequently, the licensee removed the instrumentation, based on the recommendation by W reported in Reference 2.
.The staff evaluated this reference, as well as Reference 3, and concluded that the licensee had not provided an adequate basis for removing the instrumentation. It therefore recommended that monitoring of these lines be re-implemented, as a means of satisfying the requirements of Action 3 l of the Bulletin. j j
in Reference 7, the ucensee requested a meeting with the staff to provide additionalinformation showing that Action 3 of the Bulletin was adequately addressed, and to demonstrate that the i implementation of monitoring, as recommended by the staff, was not necessary. In Reference 8 the staff provided an additional safety evaluation of the licensee's response in Reference 4, in which the staff again requested that the licensee consider implementation of monitoring, unless it could provide new information supporting its decision not to do so.
The requested meeting was held on March 3,1998, in Rockville, MD, in which supplementary material was presented by the licensee and W to support the licensee's position.
9805110004 9905064%
DR ADOCK 0 DR
2.0 EVALUATION 2.1 Ascessment of the TASCS Methodoloav in Reference 8, the staff concluded that W had not shown the app 3 cation of the TASCS i methodology to the known events at Farley/Tihange and Genkai. Thess events represent benchmarks, and any technique designed to preclude the events in the Bulletin would be expected to simulate the conditions under which the events occurred.
The staff had also identified several deficiencies, the most significant being (a) the calculated I results did not seem to correspond to experimental data obtained from a Japanese simulation of j the Farley event, described la References 3 and 5, (b) the W experimental work, on which the TASCS methodology is partially based, did not seem to correspond to the conditions under which cracking occurred, and - (c) the root cause for the Farley/Tihange failures reported in the Bulletin
- was not determined.
At the March 3,.19g8, meeting, W presented the results of the application of the TASCS methodology to the Farley event. Based on their approach, W determined that the turt>ulent ;
penetration length is beyond the swing check valve, and concluded that the entire section of
. unisolable piping, from the check valve weld to the RCS nozzle, would be susceptible to stratification and cracking from thermal cycling. The largest stratified temperature difference was determined to be at tiie check valve weld, regardless of leakage flow. Since thermal cycling and fatigue failure depends on this difference, this also indicates that cracking would occur at this weld. However, in the Farleyfrihange events the cracks in these lines occurred at the weld between the first upttrqam elbow and the horizontal segment. In Tihange, cracks also occurred )
la the base metal of the elbow itself. The calculated stratified tempemture differences at the j elbow weld were smaller than at the check valve weld. Thus, although the W methodology j appears to provide a conservative turbulent penetration region and a conservative' stratification j temperature difference, there is no assurarece that it reflects the exact nature of the thermal j cycling mechanism which caused the cracks in Farley/Tihange.
)
W also presented limited results of a proprietary simulation test done in Japan to replicate the I Fariey event. E stated that this testing indicated a low top-to-bottom temperaturr difference and minimal cycling at the crack location. 7 I
Limited temperature data obtained from the proprietary simulation was previously presented in i References 3 and 5. This dats represents measured water temperature histories at several locations along the buttom of the horizontal pipe segment, from the swing check valve to the i elbow.i These locations correspond to locations in the Farley Si line. The inleakage, as well as I the temperature distribution along the top of the pipe, or any temperature stratification measurements, were not provided.
E stated at the meeting that the water temperature histories were for a 1 gallon-per-minute (gpm) in-leakage flow (this value had not been previously provided). The well temperature history of the inside surface of the pipe, where the bottom of the elbow weld is located, was also
. shown A plot of the mean temperature at each location along the axial direction of the horizontal 7 i
4
pipe indicates that the temperature of the water at the bottom of the pipe increases as the elbow (where the crack occurred) is approached. W stated that the measured temperature of the water at the top of the pipe was the same as the RCS temperature. Therefore, the stratified (top-to-bottom) mean water temperature difference decreases as the elbow is approached, being small at the elbow wold, the location where the crack at Farley occurred. Since cracking results from thermal stratification cycling, and the thermal stratification was determined to be small at the weld location, the cause of the cracking at Fadey/Tihange remains unexplained.
1 Based en information provided by W in 1988, the Bulletin lists the measured mean temperatures at the top and bottom of the pipe outside surface at Fariey, both with and without in-leakage.
These temperatures are 440'F at the top, and 225'F at the bottom, for a leakage of .7 gpm, and 495'F at the top and 450*F at the bottom, for no leakage. The simulation test data indicates, at the same location, a mean temperature of 554*F at the top of the fluid, and a mean temperature ;
of 344*F at the bottom of the fluid, for a leakage of 1 gpm. The discrepancy between the two j sets of temperatures has not been explained. However, the top-to-bottom temperature ;
differences are almost identical, even though the leakages are significantly different. 'l Furthermore, the top temperatures at Farley indicate that the turbulent penetration extended only l somewhere between the weld and the location of the thermocouples, i.e., it never reached the valve weld location, whereas according to W the turbulent penetration reached the valve weld in the simulation test. It is therefore nct clear to what extent the test was an actual simulation of j the Farley event.
l 1
The TASCS methodology includes the calculation of the axial temperature distribution of the !
leakage. A comparison with the axial temperature distribution determined from the Japanese !
test indicates a differently shaped distribution (concave upward for the Japanese test vs.
concave downward for a TASCS calculation.) The reason for this is not clear. No other independently measured axial temperatures are available to resolve this apparent discrepancy. i Based on an assessment of the limited additionalinformation on the Japanese tests provided by W, the staff concludes that there is still uncertainty regarding the TASCS methodology and the )
quality of the testing on which it is based. The staff also concludes that certain, conservative, features of the TASCS methodology may be acceptable for application, but only on a case-by-case basis.
2.2 Apolication to South Texas Units 1 and 2 i 2.2.1 Charaina and Auxiliary Sorav Lines The licensee stated at the March 3,1998, meeting that at STP Units 1 and 2 the charging system and the safety injection system are separated, each with its own set of pumps. The separation of :
i the two systems implies that only hot water can leak through the charging and auxiliary spray lines, whereas in plants where the two systems use the samu pumps, the potential exists for cold in-leakage into the charging system. At STP, the only source for the charging system is hot ,
water from the regenerative heat exchanger (RHE) (approximately 530'F at normal full power I speration or 477'F at maximum charging flow). Therefore, only hot water inleakage into this i system reduces the potential for thermal cycling. )
/:
i 4 )
I l
_ W presented the results of the application of TASCS to the charging and the auxiliary spray lines.
W calculated the axial temperature distribution between the RHE and the isolable check valve at various leakage flow rates, and showed that the in-leakage temperature into the unisolable piping
= increases with flow rate, which the staff finds reasonable. W also calculated the corresponding axial temperature distribution in the unisolable piping for the same leakage flow rates and the turbulent penetration zone, and stated that the temperature differences within this zone are always lower than in the section of the piping where the turbulent penetration doesn't extend.
. However, beyond stating that the turbulent penetration zone is vulnerable to thermal cycling, there is no way to determine the effect of this cyclicality on the axial thermal distribution and j fatigue within this zone. W asserted that no fatigue can occur within this zone for these lines.
There is no independent verification of this assertion, since no measured axial data was provided
' to support these calculations. However, the staff finds this assertion regarding the charging lines at STP plausible, if unverified.
l In addition, monitored temperature traces at one location of the altemate charg!ng line in STP, in wt;ich a lift-check valve was found to be lealting, indicate that stratification of abcut 40'F occurred, but there was no evident cyclic effect on the stratified flow, comparable to that recorded at Farley. The staff finds that this provides additional assurance that mir,imal thermal stratification cycling will occur unde Nadvertent in-leakage into the charging lines.
In Reference 3, W had previously reported a fatigue evaluation of the weld between the check valve end the unisolable piping, and had estimated that the ASME cumulative usage factor (CUF) limit under combined design transients and inadvertent thermal stratification cycling would be achieved in a time span of 11.4 years. On the same basis, W again estimated that the ASME CUF wo'Jid be achieved by mid May,200g, at STP Unit 1. This time span was calculated at the request of the NRC, using the conserative assumption that the thermal cycling occurred at the check valve weld, and formed the interim basis for permitting continuing normal operation of STP without monitoring. In Reference 2, W has estimated that, without this assumption, the ASME CUF limit would not be achieved at the weld during the life of the plant (40 years). The statt has reviewed the basis for this estimation and finds it acceptable.
The staff also finds that these considerations are applicable and bound the evaluation of the auxiliary spray line.
The staff therefore concludes, based on the evidence presented by the licensee and W at the ,
March 3,1998, meeting and in Reference 3, that fatigue failure due to thermal stratifu:stion cycling is not likely to occur in the charging and auxi:iary rpray lines at STP Units 1 and 2 for the 40-year life of the plant. Adequate evidence has also been presented to support the licensee's position that leakage monitoring of these lines is not necessary during this time span. The staff
- grees with this position and considers this issue resolved.
2.2.2 Residual Heat Removal System A related issue is the resolution of Bulletin 88-08, Supplement 3, with regard to the residual heat
- emoval lines at STP. The licensee and W also presented the application of the TASCS methodology to these lines. The staff found an inconsistency between an assumed temperature distribution and temperature measurements made in a foreign plant. However, W had previously performed an evaluation of these lines in Reference 10, in which these lines were
5 i
shown not to be susceptible to the phenomenon in Supplement 3 of the Bulletin by virtuc of i sufficient distance of the isolation valves from the turbulent penetration source. The stafi finds !
. this reasonable and acceptable, and considers this issue resolved.
3.0 ' CONCLUSION Based on a review of the additional material presented at the March 3,1998, meeting, the staff I concludes the following: 1
- 1. The licensee has reasonably demonstrated that the normal and attemate charging lines, and the auxiliary spray line at STP Units 1 and 2 are not susceptible to the thermal cycling phenomena described in Bulletin 88-08 for the life of the plant, and is therefore not I required to monitor thesa lines for leakage.
- 2. I
' The licensee has alen reasonably demonstrated that the residual hest removal lines at STP Units 1 and 2 are not susceptiole to the thermal cycling phenomenon described in
. Bulletin 88-08, Supplement 3, for the life of the plant The staff therefore finds that the licensee has provided the necessary assurance required by :
Action 3 of the Bulletin for these ;ines, and considers the issues with respect to Bulletin 88-08 resolved for STP Units 1 and 2.
4,0 EtEFERENCES
- 1. Letter of February 23,1996, from T. W. Alexion, USNRC, to W. T. Cotte, Houston Lighting and Power (HL&P). ;
- 2. WCAP-12598, "NRC Bulletin 88-08, Evaluation of Auxiliary Piping for South Texas Project Units 1 and 2," Westinghouse Electric Corporation @, May 1990. (Proprietary information. Not publicly available.) ];
- 3. WCAP-12598, Supplement 1,"NRC Bulletin 88-08, Evaluation of Auxiliary Piping for
)
j South Texas Project Units 1 and 2," Westinghouse Electric Corporation @, November 1993. (Proprietary information. Not publicly available.)
- 4. Letter of July 15,1996, from S. E. Thomas, HL&P, to the USNRC Document Room, with attached W letter dated June 19,1996, from M. A. Sinwell, W, to W. T. Cottle, HL&P. )
1
- 5. EPRI TR-103581, " Thermal Stratification, Cycling, and Striping (TASCS)," prepared by l Westinghouse Electric Corporation for the Electric Power Research Institute, Palo Alto, 1 Caliiomia, March 1994. (Proprietary information. Not publicly available.)
- 6. Roarty, D. H., P. L. Strauch and J. H. Kim, " Thermal Stratification, Cycling, and Striping Evaluation Methodo ogy," ASME PVP-Vol 286,1994.
l
- 7. Letter of October 29,1996, from S.E. Thomas, HL&P, to the USNRC Document Control Desk. j 4
o . . . .
6
- 8. Letter of March 24,1997, from T. W. Alexion, NRC, to W. T. Cottle, Houston Lighting and Power Company.(HL&P). -
- 9. Summary of March 3,1998, Meeting on Long-Term' Resolution of Bulletin 88-08, dated April 6,1998.
- 10. WCAP -12108, " Evaluation of Thermal Stratifica' ion for the South Texas Project, Units 1 and 2 Residual Heat Removal Lines", January 1989. (Proprietary information. Not publicly available.)
>rincipal Contributor: _ Mark Hartzman j i
Date: May 6, 1998
]
l 1
i l