05000269/LER-2008-001
| Docket Number | |
| Event date: | 09-13-2008 |
|---|---|
| Report date: | 11-12-2008 |
| 2692008001R00 - NRC Website | |
EVALUATION:
BACKGROUND
This event is not reportable per 10CFR 50.73 but is being submitted as a voluntary Licensee Event Report.
In normal operation, one High Pressure Injection (HPI) [EIIS:BG,CB] pump is required to be in service to supply seal injection to the Reactor Coolant Pump (RCP) seals and normal make-up to the Reactor Coolant System (RCS) [EIIS:AB]. In normal operation HPI takes suction from the Letdown Storage Tank (LDST).
HPI is also the high pressure component of the Emergency Core Cooling System (ECCS), initially taking suction from the Borated Water Storage Tank (BWST). After depletion of the BWST inventory, HPI can be supplied via the Low Pressure Injection System (LPI) [EIIS:BP] taking suction from the reactor building emergency sump.
Accident analysis shows that for Loss of Coolant Accidents (LOCAs) initiated at power levels above 75% rated thermal power, two HPI pumps are needed. Thus three pumps are provided in order to meet the single failure criterion.
The HPI suction header is shared among all three pumps. The A pump is considered part of the A train and the C pump is part of the B train and the B pump is shared (actuated by Engineered Safeguards (ES) system as part of either train). During normal operation, flow from the LDST enters the center section of the header and flows outwards to the A injection train with the B train in ES standby. During an event involving actuation of the ES system, valves open to connect the BWST to the ends of the suction header, all three pumps receive start signals (the A or B pump is assumed to already be in service supplying RCP seal injection and normal RCS make-up flow), and valves open in the discharge headers to increase flow into the RCS.
Technical Specification 3.5.2, Condition A, allows one HPI pump to be inoperable for up to 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. Condition C allows 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> to reduce power to less than 75% when one HPI train is inoperable.
Prior to this event Unit 1 was operating in Mode 1 at 100% power with no safety systems or components out of service that would have contributed to this event.
EVENT DESCRIPTION
On January 11, 2008, the U.S. Nuclear Regulatory Commission (NRC) issued Generic Letter 2008-01 (GL 08-01) to address the issue of gas bubble (void) accumulation in the emergency core cooling, decay heat removal (DHR), and containment spray systems. At Oconee these are the HPI, LPI, and Reactor Building Spray (BS) [EIIS:BE] systems.
In response to GL 08-01, Oconee performed a series of confirmatory Ultrasonic Test (UT) inspections to determine if any voiding conditions existed at selected locations which had a higher probability of void accumulation. One hundred twenty-eight (128) locations were identified for these examinations. One hundred sixteen (116) were accessible during unit operation and were completed prior to submittal of the Duke Energy response to GL 08 01. Four (4) additional locations have been completed during the ongoing Unit 2 outage.
Several small voids were found and entered into the Problem Investigation Process (PIP), part of the Duke Energy Corrective Action Program. The following data for each of the voids were taken from the indicated PIPs:
Date discovered Volume (cubic feet)�PIP Number Location 08/18/2008� 0-08-5075 Unit 1 0.11 Unit 2 1.8 Unit 3 1.7 RBS discharge header upstream of valve 1,2,3BS-2 on each unit as indicated.
09/10/2008�0.002�0-08-5537 Unit 3 LPI A Train injection line between valve 3LP-17 and Penetration 15 in the East Penetration Room.
09/11/2008 0.6 0-08-5580 Unit 2 2A RBS discharge header near 2BS-13 09/12/2008 0.0004 0-08-5590 Unit 2 2A LPI discharge header upstream of 2LP-17 09/12/2008 0.13 0-08-5591.
Unit 1. 1A LPI discharge header downstream of valve 1LP 17 near Penetration 15 09/12/2008 0.44 0-08-5596 Unit 1 Decay Heat Drop Line downstream of valve 1LP-3 near Penetration 57.
09/13/2008 0.60 0-08-5600 Unit 1 downstream of check valve 1HP-101 near vent valve 1HP-442 in one of 2 BWST supply lines to the common HPI pump suction header Of these voids, all but one were determined to have no significant effect, i.e. the associated components and systems were operable in the as-found conditions.
The last void listed was found in .a section of pipe which supplies water from the BWST to a common suction header which supplies all three HPI pumps. The HPI pumps have tighter mechanical operating clearances than the LPI and RBS pumps and thus the presence of a void this size is more significant.
The void was found on the morning of 09/13/2008 by technicians performing the UT exam. Engineering was informed and performed calculations to determine the size and significance of the void.
PIP 0-08-5600 was initiated at 13:46 hours on 09/13/2008 to document the issue.
In the PIP, Engineering provided the following statement:
"Assuming that the gas void travels as a single slug during an ES event, there is no supporting information from the pump vendor that this amount of gas would be acceptable. Vendor support is required to determine if sufficient mixing would occur prior to entering the suction piping of an HPI pump so that mission of the system is not jeopardized. From review of the applicable piping isometrics, the gas void must pass through five 90 degree turns before arriving at any of the three HPI pumps. Multiple pumps are not assumed to be affected based on the premise that if the existing gas volume is capable of separating into two or three slugs, then it is capable of mixing into the flow stream.
Based on this information, at 15:12 Operations declared 1A HPI pump and 1A HPI train Out of Service and entered the applicable TS conditions. The line was vented, and, at 16:43 Operations declared the pump and train Operable and exited the TS conditions.
Additional examinations confirmed that the corresponding locations on the Unit 1 B train and on both HPI trains on Units 2 and 3 .did not contain voids.
There was no immediate indication of "firm evidence that a condition prohibited by technical specifications existed before discovery, for a time longer than permitted by technical specifications" (see NUREG 10.22, Rev 2, Section 3.2.2) so the event did not appear to be reportable as "Operation or Condition Prohibited by Technical Specifications." Also, based on the assumption above that the void would either 1) remain intact and damage only one pump, or 2) break up and mix sufficiently to pass without daMage, Engineering and Operations concluded that the issue did not appear to constitute a condition which could result in a reportable loss of safety function. Therefore, Oconee's initial conclusion was that the event was not reportable.
An Engineering evaluation was initiated to confirm how the void would mix during transport. A vendor was contracted to generate computer flow models, including piping details, in order to analyze the condition. The conclusion from that analysis was that the observed void would preferentially enter the A HPI pump suction.
Thus only one pump would be adversely impacted by this condition.
The two remaining pumps (B and C) would remain available to perform their specified safety functions. Therefore, it was confirmed that this event is not reportable as a potential loss of safety function.
Also, Oconee performed an engineering evaluation in an attempt to determine the source of the void. The evaluation could not find any firm indication of the source of the gas which made up the void. This evaluation reviewed venting and testing procedures,, maintenance activities, the sequence of activities, operating conditions since system start up, etc. as discussed below:
HPI had been returned to service during Unit 1 start-up following the 1E0C24 refueling outage during which full flow testing was performed through the subject line. This full flow testing resulted in flow velocities during the testing that would have clearly flushed any accumulated air out of the subject line. Also, the subject line was vented to remove any accumulated air during startup of the system following the refueling outage. The mode of applicability for TS 3.5.2 was entered approximately 12:00 hours on 5/28/2008 during startup from the refueling outage.
Review of system operation since startup from the 1E0C24 refueling outage identified no system alignments or conditions that could have created the quantity of voiding that was identified during the confirmatory Ultrasonic test.
A follow up vent evolution was conducted on 09/14/2008 which showed no indication of voids in the vent stream. On 09/26/2008, another UT examination was conducted which found the piping to be water solid. Water solid conditions were also confirmed on 10/21/2008 as part of new venting practices implemented in response to GL 2008 01. Thus there is no indication that the void was produced by any ongoing process.
In the absence of firm evidence to establish the time at which the void might have rendered the 1A HPI pump inoperable, the guidance of NUREG 1022 indicates that it is appropriate to assume the time of discovery. Based on that guidance, this, event did not involve operation in a condition prohibited by TS and Oconee confirmed that it is not reportable under that criterion.
CAUSAL FACTORS
As part of the PIP Corrective Action process, a cause investigation was initiated. The investigation concluded that the cause was a combination of "testing not specified" and "maintainability not considered in original design (maintenance/testing)." However, these causes appear to be consistent with a general oversight throughout the industry as documented in GL 2008-01. GL 2008-01 notes that some licensee's TS only required venting of portions of the ECCS systems, and did not address venting in local high points where gasses might collect. The Oconee TS specifically require venting the HPI and LPI pump casings every 31 days and are silent on venting other portions of the piping.
CORRECTIVE ACTIONS
Immediate:
1. As described above, the affected section of pipe was vented, reexamined by UT, and declared operable.
Subsequent:
1. The affected section of the BWST piping has been monitored by UT examination on a periodic basis since the discovery of the void and no additional void formation has been detected.
2. A vendor was contracted to evaluate the as-found condition and determine void fractions for various HPI pump combinations.
Planned:
1. As stated above, the void was identified during inspections performed in response to GL 08-01. Duke Energy provided a response to GL 08-01 by letter dated October 13, 2008.
Attachment 4 of that letter details a list of commitments specific to Oconee Nuclear Station, including items to address a change to Technical Specifications, and procedure changes affecting venting and associated surveillances. Those commitments are considered adequate to address the issues in this report. There are no additional NRC Commitment items contained in this LER.
SAFETY ANALYSIS
This event did not include a Safety System Functional Failure.
As stated above, an analysis concluded that the void would be preferentially transported to the 1A HPI pump. The pump vendor does not have data to support pump operability under the void conditions predicted by the analysis.
Per guidance issued by the Nuclear Energy Institute (NEI) in NEI 99-02, Regulatory Assessment Performance Indicator Guideline, revision 5, it is appropriate to assume the time of discovery when calculating the impact of this event on the Mitigating System Performance Index. Based on this assumption, the duration of the event was very short and had minimal risk significance and resulted in no impact on the health and safety of the public.
ADDITIONAL INFORMATION
Licensee Event Report 269/2006-03, revision 1, dated 05/29/2008, discussed inoperability of LPI pumps due to foreign material in the suction piping from the Reactor Building Emergency Sump lines.
This was a somewhat similar event in that LPI is also an ECCS system and operability of the suction flow path was affected. It is not considered a recurring event due to different causes for the events.
There were no releases of radioactive materials, radiation exposures or personnel injuries associated with this event.
No equipment actually failed in association with this event.
Therefore, this event is not considered reportable under the Equipment Performance and Information Exchange (EPIX) program.