05000456/LER-2004-001
| Braidwood, Unit 1 | |
| Event date: | 0-1-2004 |
|---|---|
| Report date: | 03-31-2004 |
| Initial Reporting | |
| 4562004001R00 - NRC Website | |
Background
An Ultrasonic Flow Measurement System (UFMS) was installed on each of the four main feedwater [SJ] loops and the common main feedwater header and was utilized as a calibration tool to correct feedwater venturi flow measurements through use of a correction factor. The UFMS uses ultrasonic cross-correlation technology to measure feedwater flow. The UFMS vendor for Braidwood Station is Advanced Measurement and Analysis Group (AMAG), Inc.
B. Description of Event:
Braidwood Station Units 1 and 2 installed the UFMS in May 1999. Braidwood implemented the venturi correction factors using UFMS in June 1999. During testing, measurements of the correction factors indicated an unexpected feedwater venturi flowrates and venturi correction factors. This resulted in different megawatt electric recoveries between the two Units with Byron Station Unit 1 potentially generating more megawatts electric (approximately 15 Mw) than Braidwood Station Unit 1.
An evaluation was conducted to determine and understand the difference between Byron Station Unit 1 and Braidwood Station Unit 1 and was inconclusive.
However, the evaluation did verify the UFMS was installed correctly and it was operating within design criteria established for the UFMS. Byron implemented the venturi correction factors using UFMS in May 2000.
Over the next several months, additional internal and external evaluations were conducted in an attempt to understand the discrepancy between the two Units.
The results of these investigations concluded that the UFMS was operating in accordance with the criteria required by the UFMS vendor.
A broader test plan was developed in early 2003 to continue the investigation.
In May 2003, a flow comparison test with the common main feedwater header UFMS flow to the sum of the UFMS measurements in the four main feedwater loops was conducted on Braidwood Station Unit 1. The results of this test were within the acceptance criteria.
In August 2003, this test was re-performed at Byron Station Unit 1. The difference between the sum of the UFMS measurements in the four main feedwater loops and the common main feedwater header UFMS flow was outside the acceptance criteria (1.572% compared to a maximum allowable statistical limit of 0.70%).
With this test outside of its acceptance criteria, Byron Station reduced power NRCFORM366A (1-2001) U.S. NUCLEAR REGULATORY COMMISSION FACILITY NAME (1) DOCKET (2) LER NUMBER 0 PAGE (3) Braidwood, Unit 1 STN 05000456 2004 - 001 - 00 and returned the correction factors to 1.0 on both Units, pending resolution of the issue.
Upon investigation by site and vendor personnel, it was determined that signal noise affected the flow velocity calculations, which in turn affected the determination of the correction factors. With the noise contaminated correction factors used in the calorimetric calculation, a non-conservative or lower power measurement resulted. When reactor power was adjusted to match the calorimetric power, an overpower condition was created. An overpower condition on Unit 1 and Unit 2 potentially existed since initial implementation in June 1999.
As part of the UFMS independent review, a flow measurement utilizing a radioactive tracer was performed on the Byron Unit 1 and Unit 2 Main Feedwater Systems. The tracer results indicated there was a non-conservative bias in the UFMS installed on the common main feedwater header. This test also validated the accuracy of the installed venturi flow measurement system.
The originally reported worst-case overpower condition on Unit 2 (September' 2003) was based on the premise that the UFMS on the common main feedwater header was providing accurate flow measurements. Based on the results of the tracer test on Byron Unit 1 and Unit 2 validating the accuracy of the venturi flow measurement at Byron, the worst-case overpower is now based on the maximum UFMS correction factor applied to the venturi flow measurement. Based on the results of the tracer testing on Byron Unit 1 and Unit 2, it was assumed that similar results would be obtained on both Braidwood Units. The worst-case overpower conditions are 101.07% for Unit 1 and 101.21% for Unit 2.
In accordance with Braidwood Station Unit 1 and Unit 2 License Condition 2.G, an ENS notification was made to the NRC at 1905 CST on March 1, 2004, due to the violation of License Condition 2.C(1) "Maximum Power Level," of each license.
C. Cause of Event
The initial root cause of this event was determined to be noise contamination of the UFMS ultrasonic signal.. Feedwater flow pressure pulses occurred at frequencies that affected the UFMS signal and resulted in a bias in the determinations of the flow correction factors. This noise caused the UFMS to indicate a lower than actual feedwater flowrate which resulted in a non- conservative calorimetric calculation.
Based on the results of the tracer test mentioned above, the initial root cause for inaccuracies in the feedwater ultrasonic flow measurements is incomplete.
The UFMS vendor is currently investigating the performance issues associated with the UFMS.
D. Safety Consequences:
The evaluations below envelope the revised overpower condition discussed in Section B.
The safety significance of the overpower issue for Braidwood Station Unit 1 has been evaluated by Westinghouse and Exelon Nuclear Fuels. The evaluation considered a conservatively bounding overpower value of 101.09% of rated thermal
E. Corrective Actions:
Braidwood Station Units 1 and 2 reduced reactor power to ensure that licensed thermal power limit was not being exceeded. Additionally, the correction factors for the feedwater flow venturis were reset to 1.0 (i.e., the null value). The reset of the correction factors was performed based on the conclusions of an independent vendor review of the UFMS issues.
The implementation of UFMS has been suspended.
F. Previous Occurrences:
Braidwood Station previously reported an overpower condition on Unit 2 under LER 2003-002 on September 30, 2003. This LER supercedes the information provided in
G. Component Failure Data:
Manufacturer � Nomenclature � Model � Mfg. Part Number N/A � N/A � N/A � N/A