ML051120459
| ML051120459 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 04/22/2005 |
| From: | Pace P Tennessee Valley Authority |
| To: | Document Control Desk, Office of Nuclear Reactor Regulation |
| References | |
| TAC MC4545 | |
| Download: ML051120459 (10) | |
Text
April 22, 2005 U.S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, D.C. 20555 Gentlemen:
In the Matter ) Docket No. 50-328 Tennessee Valley Authority )
SEQUOYAH NUCLEAR PLANT (SQN) - UNIT 2 - RESPONSE TO REQUEST FOR INFORMATION REGARDING STEAM GENERATOR (SG) INSPECTION REPORTS FROM UNIT 2 CYCLE 12
References:
- 1. NRC letter to TVA dated April 6, 2005, Sequoyah Nuclear Plant, Unit 2 - Request for Additional Information Regarding The Steam Generator Tube Inservice Inspection Reports For the End-of-Cycle 12 Refueling Outage (TAC No. MC4545)
- 2. TVA letter to NRC dated September 20, 2004, Sequoyah Nuclear Plant - Unit 2 Cycle 12 (U2C12) 12-Month Steam Generator (SG)
Inspection Report
Unit 2 Cycle 12 (U2C12) 90-Day Steam Generator Report For Voltage-Based Alternate Repair Criteria Enclosed are TVAs responses to NRC staff questions regarding the Reference 2 and 3 SG reports. As discussed with the staff on March 8, 2005, the additional information is provided to address the staffs questions contained in Reference 1.
U.S. Nuclear Regulatory Commission Page 2 April 22, 2005 There are no commitments contained in this submittal.
Please direct questions concerning this issue to me at (423) 843-7170 or J. D. Smith at (423) 843-6672.
Sincerely, Original signed by:
Paul L. Pace Licensing and Industry Affairs Manager JDS:DVG:KTS Enclosure cc (Enclosure):
Mr. Douglas V. Pickett, Senior Project Manager U.S. Nuclear Regulatory Commission Mail Stop 08-G9 One White Flint North 11555 Rockville Pike Rockville, Maryland 20852-2739
ENCLOSURE SEQUOYAH NUCLEAR PLANT (SQN)
NRC QUESTIONS AND TVA RESPONSES FOR UNIT 2 CYCLE 12 STEAM GENERATOR (SG) INSPECTION REPORTS The following questions and responses are associated with the SQN Unit 2 Cycle 12 90-Day Steam Generator Report for Voltage-Based Alternate Repair Criteria dated March 9, 2004, and the SQN Unit 2 Cycle 12 12-Month Steam Generator Inspection Report dated September 20, 2004.
NRC Question 1.
The process used for determining Cycle 12 voltages was described in Section 3.2, Voltage Growth Rates for Cycle 12.
In this section, it was stated, if review of historical data did not detect an indication, then the voltage for the previous cycle was assumed to be 0.0 volts. However, on Page 8 of Generic Letter 95-05 (GL 95-05) Attachment 1, it is stated that voltage growths should only be evaluated for those bobbin indications that can be identified at two successive inspections, except if an indication changes from non-detectable to a realtively high voltage such as 2 volts. The GL 95-05 guidance was developed to ensure a conservative growth rate distribution by not permitting a large number of low voltage growth values, which may have initiated during the middle of the cycle to be used in the growth distribution. A large number of low voltage growths will decrease the sampling frequency of the larger voltage growths. Please discuss the basis for your approach and your further plans for following the GL 95-05 guidance in this area.
TVA Response Of the 1,545 indications, 15 indications were analyzed for growth assuming a zero value from the previous outage. The table below indicates the growth rate for each analyzed tube.
E-1
TABLE SG Nos. Growth per Growth per Cycle Effective Full Power Year (EFPY)
SG No. 1 0.39 0.28 SG No. 2 0.15 0.108 0.22 0.158 0.23 0.165 0.35 0.251 0.7 0.503 0.7 0.503 SG No. 3 0.45 0.323 0.68 0.488 0.99 0.711 1.02 0.732 1.26 0.905 SG No. 4 0.32 0.23 0.5 0.359 0.64 0.459 Based on the small growth rates per EFPY, and the number of indications between -0.5 and 7.8 volts for 1545 indications, and considering that a bounding growth rate was used as the growth for all four SGs, it was considered conservative to analyze growth in this manner.
Given the staffs position for eliminating these growth values, TVAs future growth analyses will not include indications that are not detected in historical data. This has been added to TVAs Corrective Action Program.
NRC Question 2 Figures 4.1 through 4.4 represent the measured and predicted number of indications as a function of the Cycle 12 voltage distribution. Consistently, the measured number of indications were larger than the predicted number of indications. The consistent under prediction of the number of indications raises questions about the adequacy of the probability of detection adjustment at Sequoyah Unit 2 (i.e.,
either the bobbin detection threshold is low or the rate of initiation of new indications is high). Although the safety implications of these under predictions are not currently significant given that the leakage and burst probability estimates are well within acceptance and reporting limits, these under predictions may become significant as more (and E-2
larger) indications are left in service. Please discuss any corrective actions you have taken or plan to take to address this issue.
TVA Response In general, utilizing a probability of detection (POD) of 0.6, as required by the GL 95-05, leads to underestimates of the low voltage population and overestimates for the higher voltage population. The appropriate methodology change to address these differences would be to apply a voltage dependent POD, such as probability of prior cycle detection (POPCD) methodology. TVA is evaluating a license amendment request to utilize POPCD in the future. There are no plans to change the methodology for SQNs upcoming Unit 2 Cycle 13 inspection because there is no safety significance of underestimating low voltage indications and because the results of the leakage and burst analysis are well within acceptance and reporting limits.
NRC Question 3.
In Section 5.3, Leak Rate Correlation, it is stated that the leakage correlation for a structural limit of 2560 psi was used for the condition monitoring assessment and the leakage correlation for a structural limit of 2405 psi was used for the operational assessment. The use of 2405 psi for the operational assessment was based on Unit 2 receiving credit for power operated relief valve actuation. In Section 6.2, EOC-12 Burst Probabilities and Leak Rates, there is no discussion of which structural limit is used for the burst probability condition monitoring and operational assessment calculations. Please provide this information.
TVA Response For burst and leakage analysis for EOC-12 condition monitoring an applied pressure of 2560 pounds per square inch (psi) was used, and for operational assessment an applied pressure of 2405 psi was used.
NRC Question 4.
Please discuss whether you implement a probe wear criterion different than that discussed in GL 95-05. If you implement the probe wear criteria discussed in the staffs March 18, 1996, letter to Nuclear Energy Institute, please address the assessments required for implementation of this criteria.
Please discuss whether the probe wear criteria used at your E-3
plant resulted in (or contributed to) the under predictions cited in your reports.
TVA Response The methodology used at TVA for determining probe wear is in compliance with GL 95-05 and the letters between Nuclear Energy Institute and the NRC in 1996. If a probe fails the
+/- 15% wear criteria, it is immediately replaced and the whole tube is reinspected with a good probe. All the eddy current data from the good probe is evaluated (not just the indications which were above 75% of the repair limit). If larger indications were missed by the failed probe (approximately 0.5 volt), an assessment of the significance would be performed by TVA. However, this has not occurred.
The probe wear criterion is applied to minimize voltage variations. The detection capabilities of the probe when out of voltage tolerance are not diminished. Based on TVAs methodology described above, probe wear at SQN is not expected to have an impact on predictions.
TVAs understanding of the reporting criteria in the 1996 letters has been that reporting is necessary only when probe wear is considered a contributor to over or under predictions.
Therefore, it has not been a practice to add this discussion to 90-day reports. This has been added to TVAs Corrective Action Program to include a discussion of probe wear in future reports.
NRC Question 5.
Section 6.0 describes the condition monitoring assessment.
Figures 6.1 through 6.4 depict the distribution of end-of-cycle voltages adjusted by the non-destructive examination uncertainty distribution. Please discuss whether the discrete distributions in these figures (which may have been truncated/adjusted for fractional indications) were used in the condition monitoring assessment or whether the condition monitoring assessment utilized a non-truncated/adjusted distribution of indications.
TVA Response The Monte Carlo analysis uses the number of indications found for condition monitoring and adds uncertainty. Since many trials are run, a 1 volt indication in a trial may be, for example, 0.9 volt or 1.5 volts because of uncertainties. The E-4
leakage and probability of burst computations consider all indications with all uncertainties. For plotting purposes, where the results of all the trials are divided by the number of trials, the upper tail of results is integrated back to 0.3 of an indication and 0.7 of an indication. The voltage where this occurs is evident in the figures. Nothing is truncated in the computation of leakage and burst.
NRC Question 6.
Section 5.0 discusses the leak rate methodology used. Please confirm that the methodology and computer code used for your calculations are consistent with the U.S. Nuclear Regulatory Commission approved methodology as described in GL 95-05.
TVA Response The procedure for computing leakage was revised by Westinghouse Electric Corporation in 2002 as discussed in the following references:
- 1. Letter from A. Marion of NEI to B. Sheron of NRC dated March 15, 2002, "Refining the Leak Rate Sampling Methodology for ODSCC ARC Applications (Generic Letter 95-05)."
- 2. Letter from W. Bateman of NRC to A. Marion of NEI dated March 27, 2002, "Refining the Leak Rate Sampling Methodology for Generic Letter 95-05 Voltage-Based Alternate Repair Criteria Application."
This change has been incorporated in the computer program that was used for the Monte Carlo analysis. Accordingly, the calculations used for computing leakage are consistent with the NRC approved methodology.
NRC Question 7 Tables 6.1 and 6.2 indicate that the probability of burst and postulated accident induced leak rates were under predicted in a few cases. One of the reasons cited for the under predictions was that a different database was used for the condition monitoring than for the operational assessment calculations. Please discuss the extent to which the E-5
following technical issues could have resulted in these under predictions:
- a. Use of a non-voltage dependent growth rate distribution
- b. Probe wear criteria (see related question)
- c. Mix residual criteria
- d. Growth for deplugged tubes
- e. A growth rate that increases from cycle-to-cycle If the above issues resulted in (or contributed to) the under predictions, discuss what corrective actions, if any, you plan on taking to address this situation.
TVA Response The significant difference in the predictions is due to the higher than expected maximum voltage indication in SG No. 3.
This resulted in a significant increase in the probability of burst when compared to the prediction. It is likely that using a voltage dependent growth rate, along with a more conservative plugging limit for indications that exhibit characteristics for extreme voltage growth, would prevent this under-prediction. During the inspection, all indications greater than 1 volt were inspected with a rotating coil probe and evaluated for preventive plugging.
In response to item (a.), TVA performs SG growth rate predictions in accordance with the methodology contained in Westinghouse topical report (WCAP-14277, Revision 1) that was approved for SQN by NRC letter to TVA dated April 9, 1997 (reference pages 6 and 7). For item (b.), TVA does not believe that probe wear is a major contributor to the differences in the predictions. Refer to TVAs response for probe wear (item 4 above). For item (c.) above, mixed residual indications are investigated by lead analysts each outage and this is not considered to be a contributor to the differences in the predictions. For item (d.), SQNs SG inspection program does not de-plug tubes to place outside diameter stress corrosion cracking (ODSCC) axial indications at tube support plates back in service. For item (e.), as discussed in the above paragraph, growth rates that increase from cycle to cycle are not considered the most significant cause of under predicted probability of burst or accident induced leakage.
E-6
The following questions and responses are related to TVAs letter to NRC dated September 20, 2004, Sequoyah Nuclear Plant - Unit 2 Cycle 12 (U2C12) 12-Month Steam Generator (SG)
Inspection Report.
NRC Question 1 Please discuss the reason (e.g., noise) for preventively plugging the tubes identified in Tables 2.1 through 2.4 of your September 20, 2004 letter.
TVA Response Ten tubes were plugged preventively during the EOC-12 inspection. Nine of these tubes were plugged to provide margin against extreme growth as discussed in the response to Question 7 above. One of the tubes, SG No. 1 Row 2 Column 93 was plugged due to noise in the U-Bend area.
NRC Question 2 Please discuss the results of your foreign object search and retrieval inspection. If any loose parts were left in the SGs, please discuss whether analyses were performed to ensure that tube integrity would be maintained until the next inspection of these tubes.
TVA Response SG No. 1: The tubelane, annulus, and columns 31, 32, 36, 57, 62, and 88 were examined. One possible loose part (PLP) eddy current call was investigated by the foreign object search and retrieval (FOSAR) crew in column 88, however, the investigation revealed no object. No foreign objects were identified in SG No. 1.
SG No. 2: The tubelane, annulus, and columns 19, 20, 32, 38, 39, 57, and 62 were examined. Potential loose part eddy current calls were investigated by the FOSAR crew. As a result of investigating one PLP, a metallic piece was retrieved. No foreign objects were left in SG No.2.
SG No. 3: The tubelane, annulus, and columns 30, 31 32, 34, 35, 38, 39, 57, 58, and 62 were examined. Potential loose part eddy current calls were investigated by the FOSAR crew.
Two very small foreign objects were retrieved. A small round-nose non-metallic object was deemed to be non-significant and was not retrieved (weight, mass, and size did not warrant dose expenditure). No significant foreign objects were left in SG No. 3.
E-7
SG No. 4: The tubelane, annulus, and columns 32, 38, 57, 62, 73, 74 and 75 were examined. One PLP eddy current call was investigated by the FOSAR crew. Two foreign objects were retrieved. No significant foreign objects were left in SG No. 4.
E-8