05000390/FIN-2018010-01
From kanterella
Jump to navigation
Jump to search
Finding | |
|---|---|
| Title | Potential Failure to Request NRC Approval to Increase the OPT and OTT Response Times |
| Description | The reactor trips that protect from fuel damage that could result from departure from nucleate boiling around the fuel are identified as over-temperature-change-in-temperature (OTT) and over-power-change-in-temperature (OPT). The trips use the temperature from the reactor coolant systems hot legs as inputs into complex equations. In 1991, the licensee requested a license amendment to upgrade the Temperature Averaging System (TAS) and protection system to digital technology (Eagle 21 protection system). The Westinghouse topical reports (TR) for the TAS and Eagle 21 was reviewed and the TAS was approved with conditions for the RTD response times, electronic delay times, and surveillance test uncertainties in NUREG 847, the Safety Evaluation Report (SER), Supplement 8 dated January 1992. The SER specified, that the overall response time (RTD response time plus electronics delay) for the new RdF RTDs is 0.5 second longer (6.5 vs. 6.0 seconds) than the former Rosemount RTDs. This leaves a margin of 0.5 second (7.0-6.5) between the analysis and overall RTD response time. The breakdown of components used to arrive at the overall response time is 5.5 seconds for the RTD/thermowell and a conservative electronics delay of 1.0 second. The applicant stated that it will use the loop current step response (LCSR) test to measure RTD response time. A 10-percent allowance for LCSR test uncertainty will be used to ensure an overall channel response time of 7.0 seconds or less. ...During initial startup testing, actions will be taken to correct any resistance temperature detector (RTD) channel with an overall response time of greater than 7.0 seconds including electronics delay and a 10-percent allowance for loop current step response test uncertainty. After any such corrective action, the channel will be retested to verify an overall response time of 7.0 seconds or less (the value assumed in pertinent safety analyses). In 1997, licensee Design Change Notice (DCN) 39293 was implemented to increase the RTD response time. It stated, the response time requirement for OPT reactor trip was increased from 7 seconds to 8 seconds. This time includes RTDs, electronic processing, and trip circuit delays. As a result, the allowance for the sensor response time can be increased from 5.5 to 6.5 seconds. The Reactor Protection System Description, N3-99-1003, and the Technical Requirements Manual (TRM) were revised to reflect the change in response time for this channel. The change appeared to account for the 1.0 second electronic delay, but did not appear to account for the 10-percent allowance for LCSR test uncertainty, which would be derived from the RTD/thermowell delay. The uncertainty margin would appear to increase from 0.5 to 0.6 seconds. This change was implemented without NRC review and approval. In 2015, during hot functional testing of Unit 2 TAS RTDs, the RTD/thermowell delay did not meet the 6.5s required by the TRM from the change in 1997. On May 23, 2015, DCN 66327 was implemented to increase the response time again. The DCN stated, this DCN increases the total Narrow Range RTD response time from 8 to 9 seconds while changing the sensor response time from 6.5 to 8 seconds. Westinghouse has evaluated this change in letter WBT-D-5476 and determined that existing analyses are not impacted by this change. In this new response time the 1.0 second electronic delay and 8 second RTD/thermowells delay appeared to be accounted for, but not the margin for LCRS test uncertainty. If the 10-percent allowance for LCSR test uncertainty were accounted for, the total response time would appear to increase to 9.8 seconds. Westinghouse used a total response time of 9.0 seconds for their analyses at the direction of TVA, per WBT-TVA-3027, Revision 0, (5.10) PIN ELICB-055 Evaluation to Support a 9.0-second Total RTD Response Time, August 2015. The 10 percent LCSR uncertainty does not appear to have been included. Westinghouse letter LTR-TA-15-92, Transient Analysis Evaluation of an Increased RTD Delay Time for Watts Bar Unit 2, Rev. 0, stated, in part, due to the limiting nature of the [Steam Line Break] SLB w/ [Rod Withdrawal at Power] RWAP event, in which no margin currently exists to the departure from nucleate boiling ratio (DNBR) safety analysis limit (SAL), the inclusion of a 9.0-second total RTD response time resulted in a 0.55% DNBR penalty. For the feed water event, defined as a reduction in feedwater temperature, the Westinghouse letter stated, in part, key event results for both of the multiple-loop cases were impacted by the delay in receiving the OPT trip. While substantial margin was maintained to the DNBR limit of 1.38, the peak core heat flux values slightly exceeded the limit value of 121%. The letter concluded that the slower responding RTDs did not significantly impact the non-LOCA transient analyses and that the acceptance criteria for the events continued to be met, with the exception of the SLB w/ RWAP. However, generic DNB margin will be allocated to offset the 0.55% DNBR penalty associated with the evaluation. As such, the non-LOCA transient analyses can support operation of Watts Bar Unit 2 with a total RTD delay time of up to 9.0 seconds. The inspectors questioned the licensee to understand why the 10-percent allowance for LCSR test uncertainty was not accounted for in the Westinghouse analyses, and to what extent it could have affected the results. In addition, the inspectors questioned whether the 10-percent uncertainty was adequate in the current installation configuration. The inspectors also questioned how the LCSR test could account for increased thermal resistance between the RTDs and the thermowells. The test may not measure the actual delay time from the hot leg across the thermowell thermal resistance to RTD. The original installations relied on specific RTD thermowell bonding to establish a predictable thermal resistance and initial response time. It is unclear how this was performed for this installation to determine the actual response time. The 10 CFR 50.59 evaluation was performed May 22, 2016. This issue has been captured in the Corrective Action Program (CAP) as CR 1398934, Potential failure to request lic. amendment to change OPdT/OTdT response time |
| Site: | Watts Bar  |
|---|---|
| Report | IR 05000390/2018010 Section 1R17 |
| Date counted | Mar 31, 2018 (2018Q1) |
| Type: | URI: |
| cornerstone | Mitigating Systems |
| Identified by: | NRC identified |
| Inspection Procedure: | IP 71111.17 |
| Inspectors (proximate) | T Fanelli R Patterson C Franklin M Sykes |
| Violation of: | Pending |
| INPO aspect | |
| ' | |
Finding - Watts Bar - IR 05000390/2018010 | |||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Finding List (Watts Bar) @ 2018Q1
Self-Identified List (Watts Bar)
| |||||||||||||||||||||||||||||||||||||||