05000373/LER-2006-001
| Event date: | |
|---|---|
| Report date: | |
| 3732006001R00 - NRC Website | |
PLANT AND SYSTEM IDENTIFICATION
General Electric - Boiling Water Reactor, 3489 Megawatts Thermal Rated Core Power
A. CONDITION PRIOR TO EVENT
Event Time: 0023 CST Unit(s): 1 Event Date: 02/20/06 Reactor Mode(s): 1 Power Level(s): 006 Mode(s) Name: Run
B. DESCRIPTION OF EVENT
On 2/20/06, Unit 1 was shutting down for scheduled refueling outage. At 0023 hours2.662037e-4 days <br />0.00639 hours <br />3.80291e-5 weeks <br />8.7515e-6 months <br />, approximately 20 minutes after the main turbine was taken offline and with the reactor at six percent power, all five of the turbine bypass valves unexpectedly opened. The resulting drop in reactor and steam line pressure caused a Group 1 Primary Containment isolation and an automatic reactor scram.
Following the scram, the control room operators observed that three control rods (38-43, 26-15 and 34-47) failed to indicate full in. With the position of three control rods indeterminate, the operators entered LGA-010, "Failure to Scram" and, because they could not verify that shutdown criteria were met in accordance with Emergency Action Level (EAL) MS-3, they declared a Site Area Emergency (SAE) at 0028 hours3.240741e-4 days <br />0.00778 hours <br />4.62963e-5 weeks <br />1.0654e-5 months <br />.
Following attempts to manually insert the rods, the operators re-initialized the Rod Worth Minimizer (RWM). After re-initialization, the RWM indicated that all rods were full in. LGA-010 was exited and the scram was reset. However, once the scram was reset, the RWM again indicated that the same three control rods were not full in. LGA-010 was entered again. At 0047 hours5.439815e-4 days <br />0.0131 hours <br />7.771164e-5 weeks <br />1.78835e-5 months <br /> the Alternate Rod Insertion (ARI) system was actuated, and RWM "A" indicated that all rods were fully in the core. Following verification that the reactor was fully shutdown, the SAE was exited and the site entered Recovery mode at 0427 hours0.00494 days <br />0.119 hours <br />7.060185e-4 weeks <br />1.624735e-4 months <br /> on 2/20/06.
This event was determined to be reportable in accordance with 10 CFR 50.72(b)(3)(iv) as a valid actuation of the Reactor Protection System while the reactor was critical, and in accordance with 10 CFR 50.72(a)(1)(i) as a declaration of one of the emergency classes in the Station's Emergency Plan. ENS notification#42348 was made at 0114 hours0.00132 days <br />0.0317 hours <br />1.884921e-4 weeks <br />4.3377e-5 months <br />.
In addition, this 10 CFR 50.73 report satisfies the 10 CFR 21 requirements for reporting of defects concerning channel distortion that were identified in this event.
C. CAUSE OF EVENT
The scram occurred due to the unexpected opening of the turbine bypass valves.
The bypass valves opened due to a perturbation in the electro-hydraulic control (EHC) system control circuitry caused by a voltage dip on the -22 VDC bus feeding the analog EHC circuit cards. The voltage dip was caused by an intermittent:
failure of the -22 VDC power supply. When the -22 VDC power supply failed, the 22 VDC bus was transferred to the -22 VDC permanent magnet generator (PMG) supply; however, because the main turbine Was coasting down, the PMG was operating at only one-half of the nominal voltage.
The voltage transient changed the pressure setpoint (along with several other parameters), causing the bypass valves to open. The root cause of the -22 VDC power supply failure could not be determined.
The cause of the failure of control rods 26-15, 34-47 and 38-43 to fully insert was that the scram pressure at low power (864 psig) did not provide enough force to overcome friction due to fuel channel distortion that stopped the control rods prior to the end of travel. Based on the sequence of events and the available indications and analyses, it has been determined that control rods 26-15 and 34 47 inserted to a position between notches 00 and 01, which is fully inserted into the core. Control rod 38-43 inserted to a position between notches 06 and 16, with the best data fit near notch 12. Because all three rods stopped in a position between notches, the Rod Position Indication System was unable to indicate their position.
The fuel cells associated with control cells 26-15, 38-43 and 34-47 were removed and visually inspected. The blade associated with cell 38-43 was also removed and inspected. Two of the four fuel bundles adjacent to control rod 38-43 had visible channel distortion. In addition, indications of control rod blade to fuel channel contact on the Areva fuel and the control blades for control cell 26-15 and 34-47 were also observed. Rub marks were observed on the control rod blade consistent with channel interaction.
Industry experience of BWR/6 and BWR/4 & 5 C-lattice plants with GNF thick/thin channels indicates an increasing occurrence of additional control blade friction that has been manifested as degraded control rod operational performance.
Investigations indicate the cause to be control rod blade-to-channel interference due to an unexpected amount of channel distortion. Although there are many contributors to channel distortion, the three primary sources of interest for this report have been identified as: (1) fast fluence gradient-induced channel bow, (2) channel bulge, and (3) control blade shadow corrosion-induced channel bow. These phenomena progress rather slowly and manifest themselves in operational problems at high exposure.
GE has been working with the industry on these issues since the 1970s. GE issued SIL 320, "Mitigation of the Effects of Peripheral Core Location on Fuel Channel Bowing" in December 1979, with supplements in February 1984, July 1998, and April 2004. GE also issued 10 CFR Part 21 Communication SC05-06 in July 2005. This information included criteria for determining which channels are susceptible to distortion, and surveillance recommendations to mitigate fuel channel distortion and its effects.
Procedure LOS-RD-SR7, "Channel Interference Monitoring" provides specific guidance based upon the GE Part 21 Communication, for testing control rods using full insertion timing. LOS-RD-SR7 is in compliance with the GE Part 21 communication; however, the procedure does not include optional guidance from GE suggesting a 125 percent limit on scram time deviation from the core average.
Control rods 26-15, 38-43 and 34-47 were all identified as susceptible and had been tested in accordance with LOS-RD-SR7. Rod 38-43 had indicated a failure to settle and an increasing scram time, but was not declared inoperable.
Control rod position assessment following the scram was complicated by the re initialization of the RWM. During the investigation, it was found that the RWM computer code contains a critical flaw in that the re-initializing routine writes "00" fcr the rod positions, and the scram capture mode locks this value in. As a result, the RWM system provided incorrect information to the operators when RWM was reinitialized with the scram capture mode still active.
Finally, the declaration of the SAE was appropriate given current procedural guidance. However, given that the reactor was, in fact, fully shutdown immediately following the scram, the SAE was excessively conservative and resulted in unnecessarily exercising government and company emergency response assets. The root cause of this problem was that EAL MS-3 does not provide the operational flexibility to evaluate plant conditions and use alternate indications, such as nuclear instrumentation, to verify that the reactor is shutdown.
D. SAFETY ANALYSIS
The safety significance of this event was low. Although the event is classified as an unplanned scram with a loss of normal heat removal, the scram occurred from low power during a planned shutdown. The heat removal capabilities of the feedwater system and the safety relief valves were adequate, and ECCS was not challenged.
Following the scram, all control rods, with the exception of rod 38-43, inserted fully into the core; i.e., to a position of 02 or beyond. LaSalle core analyses show that the core will remain shutdown under all conditions with all rods except one inserted to 02 or beyond. In addition, a specific post-event evaluation determined that the core would have remained shutdown under cold, xenon-free conditions even if control rods 26-15, 38-43 and 34-47 had been full out.
E. CORRECTIVE ACTIONS
(Complete).
2. Monitoring capability was installed for the -22 VDC power supplies (Complete).
3. The current analog EHC system will be replaced on both Units with a more single failure-tolerant digital EHC system (AT# 456319-13, 14).
4. Prior to startup from the Unit 1 refueling outage, the core was reconfigured to either discharge the Atrium-10 fuel assemblies that would have been susceptible to channel bowing in the next operating cycle or place them in non-control cell locations (Complete).
5. Fourteen potentially susceptible GNF GE-14 fuel assemblies were re channeled prior to reinstallation (Complete).
6. The criterion for determining potential susceptibility to channel distortion was expanded to provide for testing fuel assembles at the 80% susceptibility point. This is intended to add margin to ensure testing occurs before fuel channel friction can become an issue (Complete).
7. Procedure LOS-RD-SR7 was revised to: 1) incorporate criteria for declaring a rod inoperable based on scram time performance relative to average core scram time; 2) change the testing methodology to take into account changes in stall flow during the testing sequence; and 3) to require trending of settle times with appropriate management notification required for adverse trends (Complete).
8. Unit 2 will be evaluated and mitigated through continued surveillance activities utilizing enhanced testing methodologies to detect channel distortion.
9. An interim enhancement to Anticipated Transient Without Scram-related EALs MA3 and MS3 was implemented to clarify and improve the criteria and definition of shutdown conditions, as allowed by 10CFR50.54(q) (Complete).
10. NAC approval will be obtained to implement plant specific shutdown criteria in the M_3 series LaSalle EALs as an alternative method per RIS 2003-18, Supplement 2 (AT# 456768-16, 17).
F. PREVIOUS OCCURRENCES
A searc:a of Licensee Event Reports going back 10 years found no previous occurrences at LaSalle of a reactor scram due to an EHC power supply failure, or of control rods not fully inserting following a scram due to friction from channel distortion.
G. COMPONENT FAILURE DATA
Lamda Electronics, -22VDC House Power Supply, Model# LMF-24-OVMY-R-3396-3