ML070110205
| ML070110205 | |
| Person / Time | |
|---|---|
| Site: | Clinton  |
| Issue date: | 12/19/2006 |
| From: | AmerGen Energy Co, Exelon Nuclear |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| EA-06-291 | |
| Download: ML070110205 (69) | |
Text
1 Clinton Power Station Regulatory Conference Reactor Core Isolation Cooling Water Storage Tank December 19, 2006
2 AGENDA Objectives - Patrick Simpson Finding - Michael McDowell High Pressure Core Spray (HPCS) System Overview - Tom Chalmers History of Vortexing Issue - Russell Peak Scale Model Testing and Results - Russell Peak Risk Assessment - Gregory Krueger Conclusions - Michael McDowell
3 Objectives Patrick Simpson CPS Regulatory Assurance Manager
4 Objectives
- Provide insights gained from recent scale model testing
- Transfer of suction from RCIC water storage tank to suppression pool would have occurred prior to significant air entrainment to HPCS pump
- Demonstrates HPCS was capable of performing its safety function when aligned to RCIC water storage tank (i.e., Green)
- Demonstrate HPCS capable of performing its function for the majority of risk significant scenarios
- Assumes HPCS pump functionality impacted by significant air entrainment
- Supports conclusion that RCIC water storage tank vortexing finding is of low to moderate safety significance (i.e., White)
- Provide additional technical information requested by NRCs November 29, 2006 letter
5 Finding Michael McDowell CPS Plant Manager
6 Finding Potentially Greater Than Green Finding The licensee failed to adequately address vortexing in the RCIC water storage tank. As a result, the setpoint for the HPCS pump suction source to transfer from the RCIC water storage tank to the suppression pool may be too low and result in significant air entrainment such that the HPCS pump would not be capable of completing its safety function.
7 Actions Taken Aligned HPCS suction to suppression pool on December 1, 2005
- Precautionary action due to uncertainties Modification implemented August 11, 2006, added a downward turned elbow on the HPCS suction piping inside RCIC water storage tank
- Provides additional margin Completed scale model testing which supports conclusion that HPCS suction transfer occurs prior to significant air entrainment reaching HPCS pump
8 HPCS System Overview Tom Chalmers CPS Shift Operations Superintendent
9 HPCS Simplified Schematic
10 HPCS Operational Characteristics
- Operation Modes
- Suction pathways
- Reactor pressure vessel injection
- Minimum flow
- Automatic Actions/Interlocks
- Suction valves (1E22-F001 and 1E22-F015)
- Suction automatically transfers from RCIC water storage tank on either low RCIC storage tank water level or high suppression pool water level
- RCIC storage tank suction valve will not close until suppression pool suction valve full open
- Injection valve (1E22-F004)
- Opens on low low reactor water level (L2)
- Closes on high reactor water level (L8)
- Minimum flow valve (1E22-F012)
- Opens on flow < 625 gpm and discharge pressure >145 psig
11
12 HPCS Layout
- Physical layout and components of interest
- HPCS pump
- HPCS utilizes two suction sources
- RCIC water storage tank
- Initial HPCS water supply
- 285 feet of piping between tank outlet and RCIC storage tank suction valve
- Suppression pool
- Long-term HPCS water supply
- 6 feet of piping between suppression pool and suppression pool suction valve
- Valves
- Both in close proximity to HPCS pump
- Injection valve (1E22-F004)
- Minimum flow valve (1E22-F012)
- Minimum flow goes into suppression pool
13 Emergency Operating Procedures (EOPs)
- EOP Usage Principles
- Symptom based
- Priority driven
- RPV level control leg
- Direction provided to control water level between Level 3 and Level 8
- Automatic operation of HPCS is from Level 2 to Level 8
- Maintaining RPV level above Level 3 is important
- Allows reactor scram to be reset
- Prevent repeated receipt of Level 3 and 2 initiation signals
14 EOPs (cont.)
- HPCS injection valve throttling
- Allows throttling using CPS 4411.04
- Simple proceduralized action
- Pre-staged EOP tools
- Operator training
- Selected for initial and requalification training
- Simulator
- Job performance measure (JPM)
15
16 Entry Conditions
17 EOP-1, Part B Level/ Automatic Actions
18 EOP-1, Part C Level Overrides
19 EOP-1, Part D Level Control Systems
20 EOPs (cont.)
- EOP-6 Primary Containment Control
- Suppression pool level control leg
- Actions prioritized relative to importance
- Establishing/maintaining adequate core cooling higher priority than taking action to exit pool level leg entry conditions
- HPCS suction transfer is expected to occur
21
22 POOL LEVEL IF THEN Hold pool level between 18 ft 11 in. and 19 ft 5 in. (between 18 ft 11 in.
and 21 ft 11 in. if upper pool dumped) (3220.01/SM, 3318.01/SF).
Low Level: Cannot hold level above 18 ft 11 in.
Go to a High Level: Cannot hold level below 19 ft 5 in.
(21 ft 11 in. if upper pools dumped)
Go to p.
EOPs (cont.)
23 IF THEN Cannot hold level below Fig Q, SRV Tail Pipe Limit
- 1. Scram.
- 4. IF................
AND...........
THEN.........
- 3. IF................
THEN.........
- 5. IF................
THEN.........
you are not in RPV Flooding, lower RPV pressure to stay below limit.
OK to exceed 100°F/hr cooldown.
still cannot stay below Fig Q, SRV Tail Pipe Limit, core will remain cooled, stop adding water from outside primary containment, except systems needed to shutdown reactor, until you can stay below limit.
Hold pool level below Fig Q, SRV Tail Pipe Limit (3220.01/SM, 3318.01/SF).
Initiate protective actions as pool level rises (4411.05).
High Level above 19 ft 5 in.
(21 ft 11 in. if upper pools dumped) p cannot get back below Fig Q, SRV Tail Pipe Limit, and stay there, BLOW DOWN:
Enter EOP-3 while continuing here.
EOPs (cont.)
24 History of Vortexing Issue Russell Peak CPS Site Engineering Director
25 History of Vortexing Issue
- February 1994 Self Assessment identified setpoint for RCIC and HPCS systems suction transfer from RCIC water storage tank to suppression pool did not consider vortexing
- Generated Calculation IP-M-0384 which was approved March 1994
- Addressed vortexing concern by raising the analytical limit by 1.6 inches
- NRC Inspection Report 94004 dated April 22, 1994 documents inspectors evaluation of calculation IP-M-0384
- Concluded that calculation was very thorough
- No further concerns noted
26 History of Vortexing Issue (cont.)
- December 1995 submitted Technical Specification change
- Raised the RCIC water storage tank low level (HPCS suction transfer) allowable value by +2.5 inches to account for vortexing
- Approved by the NRC in April 1996 (Amendment 104)
- This is the point in time when CPS considered the vortexing issue resolved
27 History of Vortexing Issue (cont.)
- Mid-1998 self-identified issues related to generic setpoint methodology
- Revised calculations to further address instrument uncertainties
- Early 1999 NRC reviewed adequacy of setpoint program
- Concluded sufficient margin exists between field settings and respective allowable values and analytical limits (Closed restart issue)
- 2001 Setpoint Calculations Upgrade Project
- TS change approved September 2005 (Amendment 168)
- Raised allowable value by 0.5 inches to +3.0 inches (+3.31 inches above top of pipe inside diameter)
28 History of Vortexing Issue (cont.)
- Actions prior to the December 2005 NRC Safety System Design and Performance Capability inspection were reasonable to address vortexing
29 Scale Model Vortex Testing and Results Russell Peak CPS Site Engineering Director
30 Scale Model Vortex Testing
- Scale Model Vortex Testing
- Testing Results
- Statistical Analysis of Testing Results
- Modeling Analysis
- Conservatisms Applied
- RELAP Analysis Results
31 Scale Model Vortex Testing Background
- Scale model testing performed to ensure site specific geometric conditions at CPS were captured
- Accurately determine hydraulic behavior and determine minimum submergences to avoid air entrainment
- Alden Labs chosen to perform testing
- Recognized industry expert on model testing
- Leading fluids engineering laboratory
- Used by Nuclear Regulatory Commission for NUREG/CR-2760 studies
32 Scale Model Description
- Used geometric scale of 1:3.051
- Allowed control of the Froude, Reynolds, and Weber numbers
- Froude number of model and CPS are equivalent because gravity and inertial forces are predominant
- Reynolds number in model kept high to minimize viscous scale effect
- Weber number in model kept high to make surface tension effects negligible
33 Scale Model Description (cont.)
- Initial water level was 4 ft. in the model
- Representative of >12 ft. above the top of the nozzle in the tank and used since air entrainment due to air-drawing vortices is likely to occur at lower water levels in the tank
- Video cameras (3) utilized to capture the dynamic effects
- Viewing boxes utilized to capture visual effects of air entrainment
- Constructed to observe conditions at the outlet nozzle
- Constructed to observe conditions in the pipe just downstream of the nozzle
- Third party (i.e., MPR) review of model scaling
34 Scale Model Description (cont.)
- Physical modeling included tank and outlet piping five (5) diameters downstream of where the nozzle attaches to the tank
- Elliptical outlet nozzle configuration modeled
- Tank internal obstructions (tank heater and HPCS test return) were modeled in the affected quadrant
- Modeled due to their close proximity to the suction nozzle which could influence the flow patterns
35 Scale Model Vortex Testing
- Testing witnessed on site at Alden Labs
- Clinton Power Station personnel
- Independent third party reviewer (MPR Associates)
- Nuclear Regulatory Commission (NRR)
- Testing included transient water level conditions
- Simulating field conditions for selected flows
- No return flow to the tank representative of field conditions
36 Scale Model Vortex Testing (cont.)
- Flows selected were 5500 gpm (3 runs), 3000 gpm (3 runs) and 4250 gpm (1 run)
- Three runs performed at both the high and lower flow rates to demonstrate repeatability of results
- High flow (5500 gpm bound pump run-out flow rates)
- Low flow (3000 gpm) representative of flow rates seen after a scram (based on reactor pressure)
- One run performed at a flow rate in the middle of these two to ensure consistency of data collection
37
38 Results of Testing No vortexing was observed as tank level decreased Localized surface depression at outlet nozzle forms as tank level decreased to level of nozzle Air entrainment occurs (air enters nozzle) when localized drawdown reaches top of nozzle Air entrainment is initiated at different levels corresponding to different flow rates Sensitive to flow rate No air entrainment occurs prior to localized surface depression level reaching the top of the nozzle
39 Results of Testing (cont.)
- Onset of air entrainment at 5500 gpm occurs at an average value (average of 3 runs) of 4.5 plant inches above the top of the exit pipe
- Onset of air entrainment at 3000 gpm occurs at an average value (average of 3 runs) of 2.6 plant inches above the top of the exit pipe
- Onset of air entrainment at 4250 gpm occurs at 3.7 plant inches above the top of the exit pipe
- Air entrainment fraction could not be directly measured
- Air flow characteristic observed was bubbly flow and not slug flow
- Puts an upper limit of 24% air entrainment based on the RELAP flow map
40 Statistical Analysis of Testing Results
- Data from the 6 test data points at the high and low flow test conditions were pooled to determine the uncertainty to apply
- Instrument uncertainty and distribution uncertainty were used
- The points meet the NUREG-1475 requirements for pooling data points
- 1 sided distribution bounds 95% of real distribution with 95% confidence
- Data from 4250 gpm not included in the pooling since there was only one point (i.e., no spread)
- Using this approach, air entrainment would occur at 6 plant inches for the 5500 gpm flow case
41 Statistical Analysis of Testing Results (cont.)
RCIC Tank Elevations 6.0 Onset of air entrainment (with uncertainty) 4.5 Onset of air entrainment (avg 5500 gpm test) 3.8 Nominal Trip Setpoint (start auto transfer) 3.31 Allowable Value 1.92 Analytical Limit 0.31 Instrument Zero 0.0 Top inside of HPCS pipe
- 6 inches is higher than the trip setpoint by approximately 2.2 inches
42 Modeling Analysis Inputs
- To determine impact on HPCS pump, the test results were used as inputs for a RELAP model of the suction piping
- Model accounts for air traveling down the 285 feet of 16 inch diameter suction piping while the tank level continues to decrease to the suction transfer trip setpoint at which point the valve from the suppression pool strokes open
- RELAP calculated the void transport and the hydraulic effect of the suppression pool valve opening
43 Conservative Inputs to RELAP Analysis
- Onset of air entrainment is 6 inches above the exit pipe based on statistical considerations
- Conservative because 6 inches is greater than the average value of 4.5 inches and this would allow for additional time for air ingestion to the pump
- Allowable value for suction transfer level point was used rather than the nominal trip setpoint
- Conservative because value is less than the nominal trip setpoint and thus the time to transfer to the pool will be later
44 Conservative Inputs to RELAP Analysis (cont.)
- Linear flow characteristic used for suppression pool suction gate valve
- Conservative because yields a lower flow value from pool than if non-linear flow characteristics of gate valve used
- Stroke time for suppression pool gate valve was the bounding value of all past test data collected at the station (21 seconds)
45 Conservative Inputs to RELAP Analysis (cont.)
- 24% air entrainment used to represent air volume in piping after air entered nozzle
- Conservative because this is upper bound for observed bubbly flow for air entrainment in RELAP analysis model
- Above this, slug flow is exhibited and we did not see this flow phenomena during the testing
46 RELAP Analysis Results
- Results show that no significant air entrainment will reach pump
- As suppression pool valve opens, the head drives the water from the pool through the 20 inch line
- Slows the water/air mixture in the 16 inch line from the tank
- The void transient ends when the suppression pool valve has stroked full open The analysis demonstrates that the HPCS pump would not see significant air entrainment
47 Scale Model Vortex Testing Conclusions
- Scale model testing provided useful plant specific insights
- Previous calculations did not account for geometric conditions necessary for vortex formation
- Test results combined with RELAP evaluation show that significant air entrainment will not reach pump prior to completion of transfer
- Demonstrates HPCS was capable of performing its safety function
48 Risk Assessment Greg Krueger Corporate Risk Management Manager
49 Phase 3 SDP Risk Assessment Approach
- The Clinton internal events PRA model of record used to evaluate potential impact of HPCS pump air entrainment
- Calculated change in CDF considered the impact of potential operator actions and system response during a range of events
- Operator response and event scenario impacts were explored
- HPCS injection valve throttling
- HPCS suction transfer to the suppression pool
50 Loss of Plant Service Water 1%
Recovered Loss of Offsite Power 7%
Flooding Initiators 5%
Transient Without Isolation 4%
Manual Shutdown 5%
Loss of Reserve Aux Transformer 21%
Loss of Instrument Air 1%
Loss of Feedwater 1%
Others 2%
Medium LOCA 3%
Total CDF = 1.16E-05/yr Clinton PRA Model CDF Distribution
51 Risk Assessment Boundary Conditions
- HPCS is capable of mitigating core damage in most PRA scenarios and therefore is important during transient and LOCA events
- HPCS is considered risk significant due to its high flow capacity over a full range of reactor pressures
- The risk achievement worth (RAW) is approximately 16 in the current PRA model
- The RAW represents the increase in CDF that occurs when HPCS is removed from service or not credited in the PRA for accident mitigation
52 Risk Assessment Boundary Conditions (Cont.)
- RCIC water storage tank
- Suppression pool
- Failure of HPCS to transfer from the tank to the pool is a modeled failure of HPCS
- Tank level instrumentation, capable of automatically transferring the suction source, is modeled in the PRA
- Operator action to manually transfer suction is qualitatively modeled
53 Risk Assessment Boundary Conditions (Cont.)
- Potential air entrainment, operator actions, and transfer of suction to suppression pool; however, occur at various times during an accident scenario
- Without this time dependent assessment during response to an event, the initial significance of potential air entrainment could result in an overly conservative SDP result
54 Risk Assessment
- Baseline calculated risk values are:
- CDF: 1.16E-05/yr
- LERF: 5.25E-07/yr
- All core damage scenarios in the PRA that credit HPCS as a mitigation system were evaluated for the potential impact of air entrainment on system operation
- Used standard SDP practice of one year exposure period to assess risk
55 Risk Assessment (Cont.)
- The calculated result is considered conservative due to the following simplifications:
- Operator throttling of the HPCS flow is the only credited action
- No credit for tank or pool instrumentation automatically transferring HPCS from the tank to the suppression pool
- No credit for manual transfer of HPCS suction to the suppression pool
- No credit for HPCS minimum flow adding volume to suppression pool which would accelerate transfer of HPCS suction automatically to the suppression pool
56 Risk Assessment (cont.)
- Operator action to throttle HPCS flow was considered
- Operator prioritization of response actions
- EOP-1 allows throttling of HPCS
- Throttling becomes compelling for scenarios in which long-term HPCS operation is required, because of RPV level swings
- Air entrainment not expected to occur when HPCS throttled
- Transients do not require significant HPCS flow (200 to 700 gpm)
- Flow rates below those necessary for air entrainment
57 Risk Assessment (cont.)
HPCS system failure due to RCIC water storage tank suction air entrainment is mitigated by the following plant and operator responses Upon recognition that extended HPCS operation will be required, throttling HPCS flow becomes high priority The time it takes to empty the RCIC storage tank also delays the time that HPCS suction air entrainment may occur This timing provides input to the operator success rate for other responses
58 Risk Assessment (cont.)
Operator success for throttling based on:
RPV water level control directed by EOPs Simple procedurally directed actions Staged equipment to perform the action Operator trained and evaluated Time available to perform the action is on the order of hours Throttling is applied to transient cases
59 Risk Assessment (cont.)
Scenarios other than transients (e.g., medium LOCA) do not credit throttling Failure of HPCS due to air entrainment is not applied to medium LOCA scenarios involving depressurization Nature of medium LOCA is that break flow depressurizes the reactor Depressurization below low pressure injection pressure occurs before HPCS air entrainment can occur
60 Risk Assessment Results (cont.)
1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03 Delta CDF /yr Case 1 HPCS Failed Case 2 Credit for Med LOCA Depressurization Reduction Case 3 Credit Throttling and Med LOCA Depress Reduction GREEN WHITE YELLOW RED Credit for Suppression Pool Transfer
61 Risk Assessment Results (cont.)
- Quantification of external events models was not performed but an estimation using previously docketed information was used
- External events such as fires, floods, and seismic events result in transient-type conditions, therefore the change in risk is similar to that calculated from the internal events model
- Fire result dominates risk and therefore the change in CDF from external events only considers fires
62 Risk Assessment Results (cont.)
- The change in fire CDF is estimated to be 3.49 E-6/yr
4.59E-6 + 3.49E-6 = 8.08E-6/yr (White band)
- This result is conservative due to the previously mentioned simplifications
- Qualitatively accounting for those simplifications would drive the result lower in the White band
- LERF does not drive the results
63 Risk Assessment Conclusions
- Risk assessment conservatively assumes that air entrainment does occur at RCIC water storage tank low level
- The risk significance associated with air entrainment in the RCIC water storage tank is in the White band
- Inclusion of external events contribution does not change the color
- Consideration of additional system responses and operator actions would reduce the risk significance lower in the White band
64 Conclusions Michael McDowell CPS Plant Manager
65 Conclusions
- Scale model testing results demonstrate HPCS would have been capable of performing safety function
- As a result Finding would be of very low safety significance (i.e., Green)
- CPS also completed a risk assessment neglecting scale model testing insights
- Results show that risk consequences are of low to moderate safety significance (i.e., White)
66 Backup Slides
67 1E22-F001 Valve (Suction from Tank) 0 10 20 30 40 50 60 70 80 90 8/16/1986 8/16/1987 8/16/1988 8/16/1989 8/16/1990 8/16/1991 8/16/1992 8/16/1993 8/16/1994 8/16/1995 8/16/1996 8/16/1997 8/16/1998 8/16/1999 8/16/2000 8/16/2001 8/16/2002 8/16/2003 8/16/2004 8/16/2005 8/16/2006 DATE Valve Stroke Time (sec.)
Stroke Time Lower Alert Limit Upper Alert Limit
68 1E22-F015 Valve (Suction from Pool) 0 5
10 15 20 25 11/1/1986 11/1/1987 11/1/1988 11/1/1989 11/1/1990 11/1/1991 11/1/1992 11/1/1993 11/1/1994 11/1/1995 11/1/1996 11/1/1997 11/1/1998 11/1/1999 11/1/2000 11/1/2001 11/1/2002 11/1/2003 11/1/2004 11/1/2005 11/1/2006 lDATE Valve Stroke Time (sec.)
Stroke Time Lower Alert Limit Upper Alert Limit
69 Sensitivity Study of Input Parameters
- RELAP analysis was run with a suppression pool suction valve stroke time of 28 seconds
- All other inputs remained the same
- This stroke time represents the design stroke time of the valve
- Results showed 4% air entrainment at the pump suction