05000369/LER-2011-001
| Mcguire Nuclear Station, Unit 1 | |
| Event date: | 01-20-2011 |
|---|---|
| Report date: | 04-01-2011 |
| Reporting criterion: | 10 CFR 50.73(a)(2)(v)(B), Loss of Safety Function - Remove Residual Heat |
| LER closed by | |
| IR 05000369/2011004 (28 October 2011) | |
| 3692011001R01 - NRC Website | |
BACKGROUND
The following information is provided to assist readers in understanding the event described in this LER. Applicable Energy Industry Identification [EIIS] system and component codes are enclosed within brackets. McGuire unique system and component identifiers are contained within parentheses.
The principal safety related function of the Nuclear Service Water System [BI](RN) is the removal of decay heat from the reactor. The RN system provides assured cooling water for various Auxiliary Building and Reactor Building heat exchangers during all phases of station operation. Each unit has two redundant essential, safety related headers serving two trains of equipment necessary for safe station shutdown, and a non-essential header serving equipment not required for safe shutdown.
The RN system is normally supplied from a non-seismic source, Lake Norman [BS]. An additional supply of water to the RN system is the Standby Nuclear Service Water Pond [BS](SNSWP). The SNSWP, which is qualified for a Safe Shutdown Earthquake, serves as the heat sink/cooling water reservoir for the RN system in the event Lake Norman is lost. All RN supply and discharge piping for the SNSWP is seismically qualified.
Upon receipt of a safety injection signal from either unit, the B RN train is automatically aligned to the SNSWP and the A RN train is automatically aligned to Lake Norman. Operating procedures manually align the A RN train to the SNSWP during a seismic event (loss of Lake Norman).
Debris strainers [STR] are installed immediately upstream of each of the four RN pumps [P]. The function of the RN strainers is to reduce the impact of macro-fouling on downstream components. Each RN strainer has the capacity to be backwashed (cleaned) automatically or manually. The RN strainers are currently in a degraded/non-conforming condition due to the discovery of reverse flow in the backwash discharge line and the subsequent isolation of this backwash discharge flow path. Credit is currently being taken only for the backwash supply flow to the strainers, the grinding action of the strainers, and manual backwash to the Auxiliary Building ground water drainage sump. Modifications to correct the RN strainer backwash system operation are scheduled for implementation.
TIRL FORM 366A (1U-2010) Technical Specification (TS) 3.7.7 specifies that two RN trains shall be OPERABLE in Modes 1, 2, 3, and 4. TS 3.7.7 Condition A states that one RN train may 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 />. TS 3.7.7 does not have a Condition for two inoperable RN trains. In this case, TS Limiting Condition for Operation (LCO) 3.0.3 applies. TS LCO 3.0.3 requires that action be initiated within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> to place the affected unit(s) in Mode 3 within 7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br />, Mode 4 within 13 hours1.50463e-4 days <br />0.00361 hours <br />2.149471e-5 weeks <br />4.9465e-6 months <br />, and Mode 5 within 37 hours4.282407e-4 days <br />0.0103 hours <br />6.117725e-5 weeks <br />1.40785e-5 months <br />.
EVENT DESCRIPTION
The following historical information was selected from the timeline prepared during the root cause investigation of this event. The information is provided as a lead-in to the event being reported.
In July 2007, the lB and 2B RN strainers experienced an unanticipated high number of automatic backwashes due to a seasonal influx of alewife fish from Lake Norman. During the review of this event, it was realized that the RN strainer backwash function during a loss of Instrument Air system (VI) event was not available because failed closed air operated valves would isolate the backwash function. Compensatory actions were put in place to allow manual operation of the air operated RN strainer backwash valves by a dedicated operator. This event was reported in McGuire Nuclear Station (McGuire) LER 369/2007-04.
Following this event, the RN backwash valves were modified so they could be mechanically opened without reliance on the VI system. Although the RN strainers were challenged during the 2007 event, they always were capable of clearing the high differential pressure (D/P) conditions. Furthermore, McGuire realized it was now susceptible to a new macro-fouling source from Lake Norman (Alewife fish).
In June 2008, McGuire discovered an additional design deficiency associated with the RN strainers. Testing of the lB and 2B RN strainers indicated that the backwash return flow was not in the normal direction to the Condenser Circulating Water (RC) system but in the reverse direction. In other words, there was not enough pressure on the discharge of the backwash to overcome RC system head. At this point the RN strainers were thought to act as a backwash supply flow assisted "grinder" versus a debris removal "strainer".
This degraded/non-conforming condition is documented in McGuire's corrective action program.
To address this RN strainer degraded condition related to reverse return backwash flow, McGuire decided to manage the macro-fouling sources that would impact RN system operation. In addition, backwash return flow would be modified to manually route flow to the Auxiliary Building groundwater drainage system.
The direction to eliminate and manage the macro-fouling sources was accomplished in two ways. Based on the previously observed behavior of alewife, a specially designed screen was installed to prevent alewife from entering the RN system from Lake Norman. The plan to manage possible fouling sources in the SNSWP was to perform fish eradications through chemical treatments. Furthermore, hydroacoustics and other surveying techniques would be utilized to monitor fish activity in the SNSWP.
In April 2009, the 2A RN strainer fouled during high velocity flow testing while aligned to the SNSWP. The source of the fouling was determined to be corrosion products which were separated from the RN piping due to the high velocity flow. The source of the fouling was unanticipated, and challenged the current assumptions regarding macro-fouling and RN strainer operability.
This event was reported in LER 369/2009-01.
The accumulation of these events led to a series of proposed plant modifications that are being implemented in three phases. Phase 1 would install an assured air supply to the RN backwash valves to operate during loss of VI events (completed). Phase 2 would provide a means to reroute the strainer backwash return to the Auxiliary Building groundwater drainage system. Finally, Phase 3 would provide a safety related "booster" pump and piping to take suction on the backwash return and flush it to the RC system.
Phase 3 will establish the originally intended design function of the RN strainer macro-fouling removal system. In the interim, macro-fouling sources to the RN system would continue to be managed.
In July 2009, the first SNSWP chemical treatment (fish eradication) was performed. The chemical Rotenone was dispensed in a spiraling pattern starting over the SNSWP intake. Once the chemical was applied, fish immediately began to surface, where they were netted, and then taken to the shore to be sorted, weighed, and disposed. Approximately 3000 pounds of fish were collected. A variety of fish species were collected; however, the majority by numbers was shad (threadfin and alewife) and sunfish. Based on previous fish surveys, alewife where thought to not exist in the SNSWP.
NRL (.)RM JbbA (1U-2U_LU) As a result of the initial chemical treatment, and discovery of alewife in the SNSWP, it was decided to chemically treat the SNSWP on an annual frequency. In support of these treatments, and for general data, hydro- acoustic surveys prior to the Rotenone treatments would also be performed.
The hydroacoustic surveys would identify targets (fish) in open water and provide insights on what to expect during the fish eradication.
In June 2010, a hydroacoustic survey was performed in support of the upcoming chemical treatment planned for July 2010. At this time, approximately 8,100 targets were observed, and were noted to be much smaller in size than the pre-Rotenone survey from the initial fish eradication.
The following week, the Duke Energy Fisheries Department began preparations for the upcoming treatment. During the survey of the SNSWP, it was discovered that a thermocline no longer existed. An established thermocline was considered a prerequisite for Rotenone treatments. The logic being the thermocline serves as a false bottom, where there is no dissolved oxygen, and the fish tend to reside above this level. This was considered important for two reasons: there would be less area to treat, and the fish are above the SNSWP intake preventing the risk of neutrally buoyant fish becoming a fouling source. The Fisheries Department contacted Engineering of this discovery. Engineering determined that mixing of the SNSWP from RN pump operations associated with maintenance activities had disrupted the thermocline. Since a thermocline did not exist, McGuire Engineering concluded that the SNSWP chemical treatment should be deferred.
Later in the summer of 2010, another opportunity to chemically treat the SNSWP became available when the thermocline had reestablished. Again, another unfavorable condition existed. Due to an unusually hot summer, the RC system low level intake (LLI) pumps were being operated. This was considered unfavorable because it was during fouling season, and alewife posed an elevated risk as a fouling source to the RN system supply from Lake Norman. As such, it was not an appropriate time to treat the SNSWP in the rare event that fish could migrate to the SNSWP intake such that both sources of water for the RN system could potentially become fouled. After assessing these risks, the decision was made to defer the chemical treatment again until the summer of 2011.
In September 2010, another hydroacoustic survey of the SNSWP was completed.
This survey produced a fish population estimate of 25,869. Cast net surveys of open water fish identified bluegill, threadfin shad and mosquitofish.
IARC EURM 366A (1U-2UlU) "Mk The data was presented to Engineering for review and Engineering determined this was not a threat. The concern for fouling had always been adult alewife due to their behavior patterns, and their role in the previous McGuire macro-fouling event in June 2007. The bluegill and mosquitofish were considered shoreline species that typically did not congregate, nor descend to the depth of the SNSWP intake. The behavioral patterns of bluegill and mosquitofish were not believed to be a risk at this time. The threadfin shad were also not considered a threat at this time.
There were numerous periods when the RN system was operated for extended periods for the purpose of flushes and flow balance testing while aligned to the SNSWP in September, October, November and December 2010. The December 2010 RN system operation on the SNSWP totaled over 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br />. Specifically on December 28th, 29th and 30th, both trains were cumulatively operated for over 30 total hours. During this timeframe there was no indication of macro-fouling that challenged the RN system while aligned to the SNSWP.
On January 18, 2011, the 1B RN pump was started for performance testing while aligned to the SNSWP. Approximately fifteen minutes after the start of the test, the 1B RN Pump was secured due to low pump suction pressure associated with high RN strainer differential pressure.
After additional evaluation and testing, it was concluded that the RN strainer high differential pressure was due to macro-fouling as the result of a large number of small, bluegill fish drawn in from the SNSWP.
Subsequent to the discovery of the fouling condition with the 1B RN Train, TS 3.7.7 Condition A had been entered (one RN train inoperable).
On January 20, 2011, when the 2B RN pump was started for testing, its strainer also rapidly fouled and required the pump to be secured. In addition, since the potential also existed to align the 1A and 2A RN trains to the SNSWP for certain accident scenarios, this common mode failure (strainer macro-fouling from the SNSWP) could also make the RN A trains potentially inoperable.
On January 20, 2011, when the 1A and 2A RN trains were also determined to be potentially inoperable, TS LCO 3.0.3 was entered on both Units. After power reductions, the Unit 1 reactor was manually tripped after a loss of Feedwater flow and the Unit 2 reactor was manually tripped per procedure. A Notification of Enforcement Discretion (HOED) was processed by McGuire and verbally approved by the NRC to allow both Units to remain in Mode 3 while NRC bORM .36(DA (IU-2U1U) the RN system and the SNSWP were returned to operable status. Remaining in Mode 3 versus transitioning both Units to Mode 5 was justified by Probabilistic Risk Assessment (PRA) to be less of a nuclear safety risk.
On January 23, 2011, all four trains of the RN system were returned to operable status after the SNSWP was chemically treated with Rotenone to eliminate the fish and the SNSWP intake piping was back flushed to ensure no macro-fouling sources remained in the intake piping.
Reportability Determination Given that both McGuire Units were in LCO TS 3.0.3 for greater than 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and Operators initiated actions (power reductions) to shutdown both Units, this event is reportable as an "Operation or Condition Prohibited by Technical Specifications" in accordance with 10 CFR 10.73(a)(2)(i)(B).
In addition, with both RN trains on each Unit determined to be potentially inoperable due to macro-fouling from the SNSWP, this event is also reportable as an "Event or Condition That Could Have Prevented Fulfillment of a Safety Function" in accordance with 10 CFR 50.73(a)(2)(v)(B).
The completion of the Unit 1 and 2 nuclear plant shutdowns and the resultant Unit 1 Reactor Protection System actuation due to loss of Feedwater flow is being reported separately under McGuire LER 369/2011-02.
CAUSAL FACTORS
Three causes were identified for this event.
Cause 1 An effective macro-fouling screen covering the SNSWP intake ports for the RN suction piping was not in place as a barrier to prevent this event.
Discussion The lack of an effective macro-fouling barrier protecting the SNSWP intake structure is considered a key causal factor to this event. An effective NRC (.)RM .366A (1U-201U) fish barrier capable of filtering small fish would have prevented this event, even in the presence of large numbers of bluegill gathered around the intake structure. McGuire has installed a fish barrier on the RN system supply from Lake Norman to protect against the alewife species. Schooling or congregating fish populations have caused numerous industry fouling events for raw water intake structures, and it appears that McGuire personnel had discounted all fish related threats except for those associated with alewife species. Had the threat associated with different species in deep sections of the SNSWP been recognized, a need for a macro- fouling barrier may have been recognized and installed prior to this event.
Cause 2 Deferral of the Rotenone fish eradication chemical treatment of the SNSWP in 2010 led to the RN strainer macro-fouling event.
Discussion The Rotenone treatment of the SNSWP was scheduled for July 2010. Days before the treatment, Duke Energy Fisheries notified Engineering that a prerequisite of having a thermocline present in the SNSWP did not exist. At that time, McGuire Engineering considered the presence of a thermocline a key prerequisite requirement for Rotenone treatment to preclude the potential for RN strainer fouling from the dead or dying fish population.
The thermocline was lost due to operation of the 2B RN train from the SNSWP stratified temperature gradient which in turn eliminates the thermocline.
Based on this, McGuire Engineering concluded that the treatment should be deferred.
Another opportunity to treat the SNSWP occurred in the August to September 2010 timeframe when a thermocline had re-established. Discussions with Engineering revealed that the decision was made to not treat the pond based on concerns with seasonal, potential alewife fouling from Lake Norman. The intent was not to induce any additional challenges to the RN system with the alewife fouling potential that occurs during this time of year.
Cause 3 There was a distinct lack of risk recognition associated with the presence of significant macro-fouling sources (small fish) in the SNSWP.
Similarly, the organization did not realize there were insufficient barriers or defense in depth in place for RN operation from the SNSWP.
Discussion From a defense in depth perspective, proper operation (without macro- fouling) of the RN system from the SNSWP was solely dependent on McGuire's ability to predict the behavior of large numbers of small, open water fish populations. Just after the July 2009 Rotenone treatment, there was an estimated fish population of 83. In June 2010, the estimated number rose to 8,082, and September 2010 the estimated number rose to 25,869. The cast netting surveys performed during this time frame identified the predominate species as small mosquitofish, bluegill and threadfin shad. There were no alewives identified. The prevalent mindset in the organization was that these species of fish did not pose a credible risk to the RN strainers.
Of the three potential barriers that could have prevented this event (adequate RN strainer backwash, fish barrier on the SNSWP intake, and elimination of SNSWP macro-fouling source), none were effectively in place.
In effect, the only basis for assuring successful RN system operation from the SNSWP was a reliance on McGuire's ability to predict atypical behavior of the altered fish assemblage present in the SNSWP.
CORRECTIVE ACTIONS
Immediate:
1. The SNSWP was chemically treated to eliminate the macro-fouling mechanism and a post treatment survey was performed to verify the effectiveness.
2. All four trains of the RN system were flushed from the strainers back to the SNSWP intake to remove any latent macro-fouling sources.
Subsequent:
i4KU 1
- 'ORM ibbA (1U-2U1U) 1. The frequency of the chemical treatment of the SNSWP to prevent fish population development was revised to at least a semi-annual frequency.
2. Monthly hydroacoustic surveys of the SNSWP were established to validate the effectiveness of the chemical treatment.
3. Periodic underwater inspections (divers) were established to inspect the SNSWP intake structure, the dam, and rip-rap (shoreline) for fish activity.
Planned:
1. Install a macro-fouling barrier for the SNSWP intake pipes.
2. Engineering to use this event for training or as a case study on lessons learned from the event.
3. The Leadership Training Review Committee shall use this event as a case study relative to decision making.
SAFETY ANALYSIS
Duke Energy used a risk-informed approach to determine the risk significance associated with the Technical Specification violation of the Unit 1 and 2 Nuclear Service Water System (RN) due to strainer macro-fouling caused by small fish drawn in from the Standby Nuclear Service Water Pond (SNSWP).
The Incremental Conditional Core Damage Probability (ICCDP) and the Incremental Conditional Large Early Release Probability (ICLERP) of this event were evaluated by considering the following:
- The duration of the period of vulnerability of macro-fouling
- The use of the average maintenance PRA model to represent plant configuration, equipment unavailability, and maintenance activities during this violation. Consideration was given to the availability of key equipment during the period of vulnerability to macro-fouling.
- The impact of internal and external initiating events.
- The availability of the RN system A trains during seismic events and subsequent alignment to the SNSWP.
"NRU EOM J6b1A (1U-2U1()) The ICCDP associated with this event was determined to be less than 1.0E-06 with the use of the internal events model (including fire and tornado risk).
The ICLERP associated with this event was determined to be less than 1.0E-07. The total ICCDP including seismic contribution to core damage is less than 1.0E-06.
Therefore, this event is considered to have a small risk impact and is of no significance with respect to the health and safety of the public.
ADDITIONAL INFORMATION
An independent analysis of the January 18, 2011 event identified the cause of bluegill intrusion into the SNSWP intake and clogging of the RN strainers as a unique combination of environmental and operational factors which can be summarized below:
- The principal environmental factor caused movement of bluegill from their normal shoreline habitat to deep water near the SNSWP intake as a result of especially low near-shore water temperatures (near 32°F). At that time, the deep water near the SNSWP intake and in the intake structure and piping was warmer than near shore, and bluegill that encountered this warmer water followed their normal thermal preference behavior toward the warmer water. The SNSWP temperatures measured at similar depths at the intake were near 45°F on January 5, 2011 and 41°F on January 19, 2011.
- The principal operational factor on January 7 and 18, 2011 was introduction of warm water from the B Train RN piping into the SNSWP intake area during RN pump alignment changes between the Low Level Intake (LLI) in Lake Norman and the SNSWP. The warm water attracted the bluegill, which were subsequently drawn into the RN system B Train.
Measurements of the water in the SNSWP intake piping reached nearly 53°F on January 7, which was closer to local deep soil temperature than to the SNSWP temperature. The fish that were in deep water near the SNSWP intake (and in the intake structure or its piping) were drawn into the RN system on two occasions, one during a brief test on January 7 and the other on January 18. On January 7, bluegill fish were attracted to the SNSWP intake by transfer of warm water during the RN system valve alignments from the piping to the SNSWP intake. Then the bluegill were drawn into the B Train piping by operation of the B RN pump but were not carried as far as the B RN strainers due to the short duration of pump operation on January 7. On January 18, RN suction valve alignment changes were made (and warm water delivered to the intake) about seven hours prior to the beginning of the macro- fouling event, which allowed considerable time for additional bluegill to be attracted to the warm water discharged into the SNSWP intake structure and surrounding area. Both groups of bluegill fish were captured in separate pulses on the RN strainers on January 18 as shown by the progression of pressure differential across the strainers.
The above operational factor was the result of system testing and is not representative of conditions that would be experienced during a re-alignment of the RN system as a result of a safety injection signal. Environmental evidence supporting this summary comes from records of below-freezing air temperatures at McGuire on several dates from late December 2010 through mid-January 2011, observations by personnel of icing along the SNSWP shoreline during the cold periods, snowfall on December 25-26 and January 9-10, and runoff of snowmelt with warming air temperatures after the snowfall events that lasted several days. Operational evidence comes from the sequence of valve alignments, pump operations, water temperatures, RN strainer differential pressures, and calculated flow rates in piping on January 7 and 18, 2011.
In summary, the volume of bluegill fish drawn into the RN B Train suction piping was specifically influenced by short duration environmental and operational changes in a specific area around the SNSWP intake structure, and not a continual introduction. Considering the above environmental and operational factors, fish intrusion conditions existed no earlier than January 7, 2011.
RN System Restoration The B RN Train suction piping was verified clean on January 20, 2011 with the restoration of the 1B and 2B RN Strainers. The operation of the RN B Train on January 20 demonstrated the bluegill were not within the vicinity of the SNSWP intake structure.
On January 21, 2011, the RN A Train suction piping was cross-tied between the SNSWP and the LLI with the B Train operating on the SNSWP to supply normal plant cooling loads and A Train initially aligned to the LLI. This was performed after the 1B and 2B RN Strainers had been cleaned and restored to service, and before the Rotenone was applied to the SNSWP.
In the above alignment, the reverse flow from the A Train suction piping to the SNSWP intake was drawn into the B Train that was aligned and operating only on the SNSWP. This has been verified by the temperature transient on the B Train Essential Header supply temperature during this operation. The temperature transient confirms that the warmer water in the previously stagnant A Train suction piping was drawn into the B Train piping. This would have included any bluegill fish if they had been present in the piping. The B Train RN Strainers differential pressure response during this alignment shows that there was no significant fouling of either the 1B or 2B RN Strainers, as the differential pressure changed less than 0.1 psid during this evolution; therefore there were no bluegill fish in the A Train suction piping.
In summary, the B RN Train suction piping was verified clean on January 20, 2011 with the restoration of 1B and 2B RN Strainers. The A RN Train was not affected by bluegill fish as evidenced by the January 21, 2011 RN B Train Strainer backwashes.
Recurring Event Evaluation The 2011 RN strainer macro-fouling event is not considered recurring.
Although there have been previous macro-fouling challenges to the RN strainers, notably the July 2007 and April 2009 events, these events did not overwhelm the strainers (with backwash in service) rendering the RN system inoperable.
To determine if a recurring or similar event exists, a search of the McGuire Problem Identification Process (PIP) database was conducted for a time period covering 5 years prior to the date of this event. Based on Duke Energy's definition of a recurring event, similar significant event with the similar event and cause codes; no recurring events were identified associated with macro-fouling of important raw water systems at McGuire.