ML053610092

Safety Significance Evaluation of Kewaunee Power Station Turbine Building Internal Floods - Volume 1, Revision 1

ML053610092
Person / Time
Site:	Kewaunee
Issue date:	12/01/2005
From:	Gaffney M Dominion
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
05-795, KPS/LIC/RR:RO
Download: ML053610092 (276)

Text

Safety Significance Evaluation of Kewaunee Power Station Turbine Building Internal Floods 40,W Volume 1 Revision 1 December 1, 2005

PDominion

Table of Contents Executive Summary .............. 2 Table of Contents .............. 3 I Introduction .............. 4 2 Conclusions............... . 5 3 Evaluation .............. 6 3.1 Flood Sources ................ 6 3.2 Accident Scenarios ............... 6 3.3 Accident Sequence Progression .7 3.4 Operator Actions .9 3.5 Results .9 3.6 Conservatisms .9 3.7 Sensitivity Analyses .10 Table 3-1. Flood Initiating Events and Frequencies . . 11 Table 3-2. Flood Levels Impacting Class I Equipment . .12 Table 3-3. Flood Scenario Contributors to Turbine BuildingFlooding Results 13 Table 3-4: Sensitivity Cases............................................................................. 14 4 References .............. 17

Table of Contents (cont.)

Appendix A - Initiating Events Analysis for Turbine Building Floods Attachment 1 - Turbine Sump Alarm History Attachment 2 - Circulating Water Expansion Joint Rupture Frequency Attachment 3 - High Energy Line Break Initiating Event Frequencies Appendix B - Flood Area Definition for Turbine Building Basement Appendix C - Fault Tree Analysis Attachment 1 - Human Error Probabilities Attachment 2 - Operator Cues for Analyzed Events Attachment 3 - Crew Exercise - Large Feedwater Line Break Appendix D - Accident Sequence Analysis Attachment 1 - Battery Discharge Analysis Attachment 2 - Timelines Attachment 3 - Assessment of Potential Spray on Reserve Auxiliary Transformer Breakers on Mezzanine Level Attachment 4 - Loss of Reserve Auxiliary Transformer Auxiliaries C,,

Attachment 5 - Steam Generator Power-Operated Relief Valve Operation Attachment 6 - Technical Support Center Diesel Fuel Oil Supply, Attachment 7 - Flood Impact on 4kV Control Power Appendix E - Non-Seismic Quantification Appendix F - Seismic Analysis Attachment 1 - Human Error Probabilities Attachment 2 - Components Determined Not Significant to Seismic Flooding Model Attachment 3 - Hotwell and Main Condenser Volumes Attachment 4 - SHIP Code Output Files - Baseline Attachment 5 - SHUP Code Output Files - Sensitivities Attachment 6 - Tank Volume Calculation Attachment 7 - Valve SW-281 1 Operation

Executive Summary

'e Executive Summary A performance deficiency was identified in NRC Inspection Report 05000305/2005011 regarding internal flooding design features. The inspectors found that there was inadequate design control to ensure Class I equipment was protected against damage from the rupture of a pipe or tank resulting in serious flooding or excessive steam release to the extent that the Class I equipment's function was impaired. Specifically, the design did not ensure that the auxiliary feedwater (AFW) pumps, 480-volt (V) safeguards buses, safe shutdown panel, emergency diesel generators (EDGs) IA and 1B, and 4160-V safeguards buses 1-5 and 1-6 would be protected from random or seismically-induced failures of non-Class I systems in the turbine building.

Flood paths were present which would allow flood water from the turbine building to flow into the safeguards alley compartments containing the identified Class I equipment. These flood paths included floor drains without check valves, doors with sufficient bottom clearances to allow water to pass through, and open floor trenches which communicate between safeguards alley compartments.

The past safety significance of this performance deficiency was evaluated by performing a probabilistic risk assessment (PRA) of the subject internal flooding scenarios leading to core damage. The flood initiating events considered included: random pipe breaks, condenser expansion joint failures, steam line breaks with fire sprinkler actuation, feedwater line breaks with fire sprinkler actuation, seismic-induced breaks, turbine-missile induced breaks, and tornado-induced breaks. The scenarios were analyzed based on: surveyor floor measurements, dynamic flood level analysis using GOTHIC, equipment survivability evaluations, room heatup calculations using GOTHIC, simulator exercises, review of operator training materials, testing of 480-V breakers in simulated flooding conditions, and seismic fragility assessments. The turbine building flood sources capable of causing failure of Class I equipment in safeguards alley were determined to be: circulating water, service water, firewater, feedwater, condensate, and the condensate and reactor makeup water storage tanks.

The total contribution to core damage frequency (CDF) from this deficiency based on the plant design in 2004 was evaluated to be 7.2E-05 per year, which would be classified as Yellow in the NRC Reactor Oversight Process (ROP) Significance Determination Process (SDP) risk determination. The total large early release frequency (LERF) contribution from this deficiency was estimated to be at least a factor of ten below the CDF, and thus not limiting in the NRC ROP SDP risk determination. Sensitivity evaluations were performed to determine the impact of changes in key assumptions such as initiating event frequencies and human error probabilities.

Page 2 of 17 l

- Table of Contents Table of Contents .. 2.

Executive Summary ............... 2 I Introduction ................ 4 2 Conclusions ............... 5 3 Evaluation ............... 6 3.1 Flood Sources .. 6 3.2 Accident Scenarios .6 3.3 Accident Sequence Progression .7 3.4 Operator Actions .. 9 3.5 Results .9 3.6 Conservatisms .9 3.7 Sensitivity Analyses . 10 Table 3-1. Flood Initiating Events and Frequencies l1 Table 3-2. Flood Levels Impacting Class I Equipment . .12 Table 3-3. Flood Scenario Contributors to Turbine Building Flooding Results 13 Table 34: Sensitivity Cases .. 14 4 References .. 17 Page 3 of 17

Table of Contents (cont.)

Appendix A - Initiating Events Analysis for Turbine Building Floods Attachment 1 - Turbine Sump Alarm History Attachment 2 - Circulating Water Expansion Joint Rupture Frequency Attachment 3 - High Energy Line Break Initiating Event Frequencies Appendix B - Flood Area Definition for Turbine Building Basement Appendix C - Fault Tree Analysis Attachment 1 - Human Error Probabilities Attachment 2 - Operator Cues for Analyzed Events Attachment 3 - Crew Exercise - Large Feedwater Line Break Appendix D - Accident Sequence Analysis Attachment 1 - Battery Discharge Analysis Attachment 2- Timelines Attachment 3 - Assessment of Potential Spray on Reserve Auxiliary Transformer Breakers on Mezzanine Level Attachment 4 - Loss of Reserve Auxiliary Transformer Auxiliaries Attachment 5 - Steam Generator Power-Operated Relief Valve Operation Attachment 6 - Technical Support Center Diesel Fuel Oil Supply Attachment 7 - Flood Impact on 4kV Control Power Appendix E - Non-Seismic Quantification Appendix F- Seismic Analysis Attachment 1 - Human Error Probabilities Attachment 2 - Components Determined Not Significant to Seismic Flooding Model Attachment 3 - Hotwell -and Main Condenser Volumes Attachment 4 - SHIP Code Output Files - Baseline Attachment 5 - SHIP Code Output Files - Sensitivities Attachment 6 - Tank Volume Calculation Attachment 7 - Valve SW-2811 Operation Page 4 of 17 l

I Introduction A performance deficiency was identified in NRC Inspection Report 05000305/2005011 regarding internal flooding design features (Ref. 1). The inspectors found that there was inadequate design control to ensure Class I equipment was protected against damage from the rupture of a pipe or tank resulting in serious flooding or excessive steam release to the extent that the Class I equipment's function was impaired. Specifically, the design did not ensure that the auxiliary feedwater (AFW) pumps, 480-volt (V) safeguards buses, safe shutdown panel, emergency diesel generators (EDGs) 1A and 1B, and 4160-V safeguards buses 1-5 and 1-6 would be protected from random or seismically induced failures of non-Class I systems in the turbine building. Flood paths were present which would allow flood water from the turbine building to flow into the safeguards alley compartments containing the identified Class I equipment. These flood paths included floor drains without check valves, doors with sufficient bottom clearances to allow water to pass through, and open floor trenches which communicate between safeguards alley compartments.

Page 5 of 17 l

2 Conclusions The total contribution to core damage frequency (CDF) from this deficiency based on the plant design in 2004 was evaluated to be 7.2E-05 per year, which would be classified as Yellow in the l NRC Reactor Oversight Process (ROP) Significance Determination Process (SDP) risk determination. The total large early release frequency (LERE) contribution from this deficiency was estimated to be at least a factor of ten below the CDF, and thus not limiting in the NRC ROP SDP risk determination. Sensitivity evaluations were performed to determine the impact of changes in key assumptions such as initiating event frequencies and human error probabilities.

Q Page 6 of 17 l

3 Evaluation 3.1 Flood Sources In this analysis, failures of non-Class I water system piping and equipment at Kewaunee Power Station (KPS) that can flood the turbine building and subsequently impact Class I components have been evaluated. Systems with sufficient inventory and flow rates to fail Class I equipment in safeguards alley were determined to be: circulating water, service water, firewater, feedwater, condensate, and condensate and reactor makeup water storage tanks. Twelve different random (9), tornado-induced (1), turbine-missile induced (1), and seismic-induced (1) flooding initiating events listed in Table 3-1 were evaluated. The frequencies of these flooding events were determined based on plant-specific analyses and industry references.

The critical flood levels for Class I equipment in safeguards alley potentially-impacted by turbine building floods are listed in Table 3-2. These levels were determined by measurements, engineering evaluations, and tests of equipment in flooded conditions.

3.2 Accident Scenarios Based on identification and analysis of internal flood areas in the KPS turbine building and safeguards alley (including consideration of unoccupied floor space, risk-significant components and associated submergence depths, drainage paths and capacity, detection methods, operator actions, and propagation paths to/from other flood areas), accident scenarios were developed for each of the flooding initiating events described above. The accident scenarios for each initiating event are very similar with differences only in detection method and time to fail Class I equipment. For each initiating event the propagation paths into safeguards alley and the subsequent component damage are the same.

A flooding event due to a non-Class I break would be indicated by a turbine building ^

miscellaneous sump level high alarm in the control room due to high level in either the turbine building or screenhouse sump. The drains and sumps 41aarm procedure instructs the operator to dispatch personnel to locally investigate the sump when this alarm sounds. Indication may also be provided by alarms related to the system with the break (e.g., low condenser vacuum, service water low discharge pressure, fire pump running or fire protection header pressure low, or steam generator low level depending on the break). The break would deposit water from the circulating water, service water, fire water system, or condensate and reactor makeup water storage tanks onto the turbinelbuilding floor. In addition, a break in the feedwater or main steam system that actuates the fire sprinklers would increase the temperature in the turbine building, which would impact the timing for investigation and isolation of the leak.

The water levels in the 480 V switchgear bus 61 and 62 room, the motor-driven auxiliary feedwater pump 1B (MDAFP 1B) room, the turbine-driven auxiliary feedwater pump (TDAFP) room, the MDAFP IA room, and the C02 storage tank 1B room would closely match the water Page 7 of 17 l

level in the turbine building because the drain lines that connect these rooms to the turbine building sump do not contain check valves and would allow water to flow from the sump to these rooms. The water-level in the 480 V switchgear bus 51 and 52 (bus 51/52) room would be lower than the turbine building because water would be entering this room via leakage under the doors from adjacent compartments. Water would rise in the bus 51/52 and diesel generator IA (DG IA) rooms simultaneously due to the trench connecting the two rooms. The only drainage from the DG IA room would be leakage to the screenhouse pipe tunnel via the gap under the door and a four-inch opening into the trench. The DG 1A room drain line would not remove any of the flood water because its drain line (which contains a check valve) empties into the turbine building sump, which would already be above this level. If the water level in DG 1A exceeds a depth of 4 inches, 4 kV bus 5 and 480 V Buses 51 and 52 (which are powered from 4 kV bus 5) are conservatively evaluated to fail.

The water level in DG lB room would also be fed by leakage under a door. The only drainage from the DG 1B room would be leakage under the door leading to the screenhouse pipe tunnel, because the room drain line (which contains a check valve) leads to the turbine building sump.

Prior to late 2004, there was a six-inch curb in the DG 1B room that protected the diesel generator and 4 kV bus 6 from floods below six-inches. This curb was removed in late 2004.

The curb has minimal impact on the analysis based on the dynamic water level evaluation and was not credited in the analysis.

Although propagation of water from the turbine building to the 4 kV buses would require some period of time, without a procedure or equipment for removing water from the room, it would have been inevitable for the water to eventually reach the buses if the flood source was not i isolated.

3.3 Accident Sequence Progression From the flooding initiating events and damage scenarios described above, the accident sequence progression has been analyzed. The accident sequence progression for each flooding event considers the response of the plant and operators to the initiating events and subsequent equipment failures, and is represented with an event tree. The flooding event trees are based on the KPS internal events PRA model event tree for loss of feedwater. In each case, if the operator successfully terminates the flood prior to failure of any buses, the accident progression would be identical to that of the existing loss of feedwater sequences except for equipment failed by spray from the initiating line break.

As with the accident scenarios, the accident sequence progression for each initiating event is very similar with differences only in the operator actions needed (i.e., isolation of the appropriate flood source) and the time required and available for those actions. The accident sequence progression following failure to isolate the flood before failure of any buses is described below.

A circulating water break would be isolated by manually tripping the circulating water pumps.

For a service water break, the operator would isolate the turbine building header by closing valves SW-4A and 4B. For a high energy line break leading to fire sprinkler actuation, the operator would implement a procedure to isolate the discharge from the fire water system into U Page 8 of 17 l

the turbine building by isolating the fire sprinklers on the turbine building mezzanine level, and isolating deluge and fire sprinkler valves in the turbine building basement. Also, the operators could trip the fire pumps locally at the 480V breakers or locally close the pump discharge valves to stop flow, but the operators were conservatively not credited to pass through flooded switchgear areas in safeguards alley to perform these actions.

If the operator fails to isolate a flood before all RCP seatcooling systems are lost, then an RCP seal LOCA could occur. The response to the RCP seal LOCA would depend on the leakage rate. The WOG 2000 RCP seal LOCA model as modified by the NRC was used for this evaluation.

If the operator fails to isolate the break initially, the water level would continue to rise in safeguards alley. Although 4 kV bus 5 motor-loads would fail, buses 51/52, 4 kV bus 6 and associated 480 V buses 61/62 would still be available, as well as the TDAFP. There is a second isolation opportunity in order to prevent eventual failure of the TDAFP's ability to start due to submergence of the associated auxiliary lube oil pump (at 9 inches). A third isolation opportunity exists to prevent eventual failure of 4 kV bus 6 (at 4 inches) and associated 480 V buses 61/62 (at 11 inches). The total volume of water required in the turbine building to flood 4 kV bus 6 is almost equal to that required to flood 480 V buses 61/62. A fourth isolation is also modeled to prevent submergence failure of the MDAFPs at 13 inches. This isolation also ensures that power to 480 VAC buses will remain available.

If the second or third isolation opportunity were successful, 4 kV power would be available to the already operating MDAFP lB. If continued operation of this MDAFW pump succeeds, the operator performs RCS cooldown and depressurization by opening a SG PORV'(which if necessary can be performed locally) to reduce RCP seal leakage. If cooldown fails, the operator could still remove decay heat by restoring RCS inventory using the available SI pump and throttling SI flow to conserve the water in the RWST per procedure.

If the available MDAFP fails, the TDAFP would be available to provide secondary heat removal.

Successful cooldown using the TDAFP also requires opening a SG PORV. Additionally, long-term instrument power must be available to allow the operator to monitor SG level and prevent overfilling the SG and failing the TDAFP. Because the normal battery chargers would be unavailable due to the loss of the 480 V buses, providing long-term DC power for steam generator level indication and auxiliary feedwater control is credited by a number of means, including automatic or manual transfer of the inverters source from the batteries to their alternate source (offsite power), which would be available in many scenarios. In addition, a normal or spare battery charger could be powered from offsite power or the Technical Support Center (TSC) diesel to restore long term battery capacity and provide SG level indication. Due to the long time to steam generator dryout due to reduced decay heat levels at the earliest point the batteries might be depleted (eight hours), much more than eight hours would be available in the most limiting cases to implement these recovery actions (e.g., a minimum of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of battery life is available if the inverters are transferred to their alternate source at four hours).

A final isolation opportunity can prevent the water level in the turbine building from reaching 18 inches. If the water level reaches this height, core damage is assumed since the electrical connections of the reserve auxiliary transformer (RAT) to 4 kV buses 1 and 2 will be submerged Page 9of 17 l

leading to a loss of offsite power and the eventual failure of all safety-related buses.

Additionally, this water level will result in the failure of the diesel generators since their air i i supply fans are powered from 480 V buses 51/52 and 61/62.

Seismic-induced floods were analyzed based on the EPRI 1989 hazard curve and associated spectra, detailed fragility assessments of the systems capable of causing critical floods in the turbine building impacting Class I components in safeguards alley, and random failures taken from the PRA models from the random pipe break analyses. Combinations of breaks which could occur in seismic events were explicitly considered in the analysis.

3.4 Operator Actions As described above, the accident sequence progression for each initiating event is very similar with differences only in the operator actions and the time required and available for those actions. Most of these operator actions fall into one of three groups: isolation of the flood source' before 4 kV bus 5 fails, isolation of the flood source before the TDAFP auxiliary lube oil pump fails, or isolation of the flood source before 4 kV bus 6 and associated 480 V buses 61/62 fail, or isolation before submergence of the motor-driven AFW pumps.

The human error probabilities (HEPs) for these actions vary for each flooding initiating event, based on the specific actions to be taken to isolate the particular flood source, the time required to complete those actions, the time available to complete those actions (based on the flow rate of the source), and the environment in which the actions must be performed. As noted above, the hot water and/or steam released from a feedwater or main steam line break would impact the operators' ability to investigate and isolate the flood. The impact of these conditions and U

dependencies among these three actions are also considered.

3.5 Results The turbine building flooding analysis summarized above represents a conservative assessment for occurrence, plant response, and operator response to a flooding event in the turbine building.

Quantification of this conservative analysis provides the core damage frequency (CDF) for the plant configuration in the year 2004. Table 3-3 presents the individual and total CDFs for each of the flooding scenarios.

The total contribution to CDF from the deficiency for the analyzed turbine building flood scenarios was calculated to be 7.2E-05. More than 79% of the CDF is due to four flood scenarios: large breaks in an inlet circulating water expansion joint (50%), feedwater line breaks that results in full flow discharge from the fire pumps (14%), main steam line breaks that results in full flow discharge from the fire pumps (15%), and seismic induced breaks of firewater, service water and condensate and reactor makeup water storage tanks (9%).

3.6 Conservatisms Development of the initiating events, accident scenarios, accident sequence progression, and human error probabilities for turbine building floods in some cases required the use of Page lOof 17 l

conservative modeling methods or conservative assumptions. The noteworthy conservatisms inherent in the KPS turbine building flooding analysis are summarized below.

1. The impact of tripping the feedwater and condensate pumps prior to emptying the hotwell was not evaluated. Instead it was conservatively determined that the entire feedwater and condensate inventory of 80,000 gallons would be pumped onto the turbine building floor.

The feedwater pumps would likely be tripped early (within approximately ten minutes per the emergency operating procedures), and an extremely large break size (8,000 gpm) would be required to discharge 80,000 gallons within that period. A smaller break size would result in less water discharged and allow more time to isolate the break to prevent failure of risk significant components.

2. Credit for operators isolating the firewater pumps following high energy line break events by either tripping the firewater pumps at the 480V switchgear, closing the firewater pump discharge valves, or initiating a manual safety injection signal (which automatically trips the firewater pumps) were not included in the analysis.

3.7 Sensitivity Analyses Development of the initiating events, accident scenarios, accident sequence progression, and human error probabilities for turbine building floods requires many assumptions. To help characterize the modeling and data uncertainty due to assumptions made for this evaluation, a series of sensitivity analyses were performed and are summarized in Table 3-4.

N .,

Page I1 of 17 I

Table 3-1. Flood Initiatin Events and Frequencies Initiating Event Frequency (per year)

Large random circulating water inlet expansion 3.7E-05 joint break (58,000 gpm) _

Large random circulating water outlet 2.9E-05 expansion-joint break (14,000 gpm)

Small random circulating water expansion joint 7.3E-05 failure (6,000 gpm)

Random service water system break with 3.2E-05 equivalent diameter greater than four inches Random fire water line with equivalent 7. 1E-05 diameter greater than four inches Random feedwater or condensate high-energy 1.4E-04 line break that actuates sufficient turbine building fire sprinklers for full fire water flow Random feedwater or condensate high-energy 4.7E-05 line break that actuates 100 turbine building fire sprinklers Random main steam high-energy line break 2.5E-04 that actuates sufficient turbine building fire sprinklers for full fire water flow Random main steam high-energy line break 1.9E-05 that actuates 100 turbine building fire QWI sprinklers Tornado-induced break of circulating water Negligible lines, firewater lines, service water lines, feedwater, condensate, and condensate and reactor makeup water storage tanks Turbine-missile induced break of circulating Negligible water lines, firewater lines, service water lines, feedwater, condensate, and condensate and reactor makeup water storage tanks Seismic-induced break of circulating water EPRI,1989 Hazard Curve lines, firewater lines, service water lines, (see Appendix F, Table 3-1) l feedwater, condensate, and condensate and reactor makeup water storage tanks

("Changes in the human error probabilities and safety-related bus failure heights from the earlier revision of this analysis (Ref. 2) were evaluated for the seismic-induced floods and found to result in a small decrease (lE-07 per year) in core damage frequency. Therefore, the seismic-induced flood analysis documentation was not revised from the earlier revision.

Page 12 of 17 l

Table 3-2. Flood Levels Impacting Class I Equipment Train A/B 480V switchgear (buses 51, 52, 61, 62)

_ 2.75" flood level trips bottom row of breakers D 4" flood level control power lost a4C

• 11" flood level bus stabs covered and bus fails Train A/B 4kV switchgear (buses 5 and 6 located in respective EDG rooms)

• 4" flood level control power connections covered, 4kV motor loads will receive lockout signal, and breaker control fails (however, supply to 480V buses will remain energized)

• 18" flood level bus stabs covered and bus fails Turbine-driven AFW pump

• 9" flood level auxiliary lube oil pump fails

• 18" flood level pump fails Motor-driven AFW pumps

• 9" flood level auxiliary lube oil pump fails

• 13" flood level pump fails Instrument air compressors (A, B, C)

• 11" flood level compressor fails Emergency diesel generators and dedicated shutdown panel

• Equipment is above 6" flood level, however associated 4kV buses fail @ 6" flood level Note: Flood levels impacting equipment failure were conservatively assessed from measured levels to allow for measurement uncertainty (typically 1/4" to 1/2" less than measurement). Flood levels provided in this table are relative to floor elevation at equipment. Flood levels used in analysis were relative to sea level.

Page 13 of 17 l

Table 3-3. Flood Scenario Contributors to Turbine Building Flooding Results Flood Scenario Total CDF (perr)

Large random circulating water inlet expansion joint break (58,000 gpm) 3.7E-05 Large random circulating water outlet expansion joint break (14,000 gpm) 4.4E-06 Small random circulating water expansion joint failure (6,000 gpm) 1.9E-06 Random service water system break with equivalent diameter greater than four 1.3E-06 inches Random fire water line with equivalent diameter greater than four inches 1.9E-06 Random feedwater or condensate high-energy line break that actuates sufficient 9. IE-06 turbine building fire sprinklers for full fire water flow Random feedwater or condensate high-energy line break that actuates 100 turbine 1.2E-07 building fire sprinklers Random main steam high-energy line break that actuates sufficient turbine 9.7E-06 building fire sprinklers for full fire water flow Random main steam high-energy line break that actuates 100 turbine building fire 5.OE-08 sprinklers Tornado induced break of circulating water lines, firewater lines, service water Negligible lines, feedwater, condensate, and condensate and reactor makeup water storage tanks Turbine missile induced break of circulating water lines, firewater lines, service Negligible water lines, feedwater, condensate, and condensate and reactor makeup water storage tanks Seismic induced break of circulating water lines, firewater lines, service water 6.6E-06 lines, feedwater, condensate, and condensate and reactor makeup water storage tanks Total 7.2E-05 Page 14 of 17 l

Table 3-4: Sensitivity Cases Analysis Case") Total CDF Baseline 7.2E-05 HEPs for operator actions with less than a 30-minute time window available from 8.4E-05 initiating event increased by factor of 5 HEPs for operator actions with less than a 30-minute time window available from 8.4E-05 initiating event increased by factor of 10 HEPs for operator actions with less than a one-hour time window available from L.OE-04 initiating event increased by factor of 5 HEPs for operator actions with less than a one-hour time window available from 1.2E-04 initiating event increased by factor of 10 HEPs for unproceduralized operator actions increased by factor of 5 1.3E-04 HEPs for unproceduralized operator actions increased by factor of 10 1.3E-04 High energy (main steam and feedwater) line break frequencies increased by 1.5E-04 factor of 5 High energy (main steam and feedwater) line break frequencies increased by 2.4E-04 factor of-l10 Circulating water expansion joint break frequencies increased by factor of 5 1.9E-04 Circulating water expansion joint break frequencies increased by factor of 10 3.4E-04 Random pipe break frequencies increased by factor of 5 1.4E-04 Random pipe break frequencies increased by factor of 10 2.3E-04 First HEP for firewater isolation following high energy line break changed to 0.1 6.9E-05 First HEP for firewater isolation following high energy line break changed to 0.3 7.OE-05 First HEP for firewater isolation following high energy line break changed to 0.6 7.1ME-05 HEPs for firewater isolation following high energy line breaks assume 38 versus 7.9E-05 32 minute average isolation time(', first HEP for firewater isolation following high energy line break changed to 0.3 HEPs for firewater isolation following high energy line breaks assume 38 versus 8.OE-05 32 minute average isolation time(2), first HEP for firewater isolation following high energy line break changed to 0.3 HEPs forndrewaterisolation following high energy line breaks assume 38 versus 8.5E-05 32 minute average isolation time("), first HEP for firewater isolation following high energs line break changed to 0.6 Large condenser outlet and small circulating water expansion joint break 9.5E-05 frequencies increased by factor of 5 Large condenser outlet and small circulating water expansion joint break 8.3E-05 frequencies increased by factor of 5; all HEPs for large condenser outlet circulating water expansion joint break isolations set to 0.12 Large condenser outlet and small circulating water expansion joint break 9.4E-05 frequencies increased by factor of 5; all HEPS for large condenser outlet circulating water expansion joint break isolations set to 0.12 Large outlet and small circulating water-expansion joint break frequencies 1.2E-04 increased by factor of 10 Large condenser outlet and small circulating water expansion joint break- l.OE Page lS of 17 l

Analysis Cased) Total I CDF frequencies increased by factor of 10; all HEPs for large condenser outlet circulating water expansion joint break isolations set to 0.05 Large condenser outlet and small circulating water expansion joint break 1.2E-04 frequencies increased by factor of 10; all HEPs for large condenser outlet circulating water expansion joint break isolations set to 0.12 Kewaunee IPE frequency of circulating water expansion joint break (2E-04 per 2.8E-04 year) used for large condenser inlet, large condenser outlet, and small circulating water expansion joint break frequencies Kewaunee IPE frequency of circulating water expansion joint break (2E-04 per 2.4E-04 year) used for large condenser inlet circulating water expansion joint break frequency; large condenser outlet, and small circulating water expansion joint break frequencies set to zero

(')HEPs in sensitivities were increased to a maximum of 0.5.

(2)Average time is for isolation of mezzanine fire sprinkler valves. GOTHIC analysis (Ref 3) indicates an additional 10 minutes is available to isolate the basement firewater valves before additional Class I equipment failures occur.

Page 16 of 17 l

'i 4 References

1. Letter, USNRC (Mark Satorius) to Dominion (David Christian), NRC Inspection Report 05000305/200501 1(DRP) Preliminary Greater than Green Finding Kewaunee Power Station, October 6, 2005.

2. Letter, Dominion Energy Kewaunee, Inc (Michael G. Gaffney) to USNRC, "Kewaunee Power Station Flooding Significance Determination Process Risk Assessment Report,"

October 31, 2005.

3. MPR Associates Inc. Calculation No. 0064-0515-LYS-01, "Evaluation of Flooding Levels for Various PRA Cases", Rev. 1, November 18, 2005.

Page 17 of 17 l

Appendix A Initiating Events Analysis for Turbine Building Floods Page 1 of 66 1

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods Owner's Acceptance: tB y 6L By l I I1N44A-S G A4cot Signature Print Name Date

l INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods l 1 W4 Kewaunee Power Station Initiating Events Analysis for Turbine Building Floods Revision No. 1 Effective Date: November 2005 5 e. 6Ltwx4As UtLAe*tL 14c1ts Prepared By: S. E. Guokas Date k -4o'IOs Reviewed By: R. J. Dremel Date

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods P. I Table of Contents Section Page 1.0 PURPOSE 2 2.0 MODEL DEVELOPMENT 2 2.1 STEPS FOR TURBINE BUILDING INTERNAL FLOODING INITIATING EVENTS ANALYSIS 2 2.2 TURBINE BUILDING INTERNAL FLOODING INITIATING EVENTS MAJOR ASSUMPTIONS 3 3.0 TURBINE BUILDING FLOODING INITIATING EVENTS ANALYSIS 3 3.1 DETERMINATION OF WATER VOLUME TO FAIL EQUIPMENT IN SAFEGUARDS ALLEY 4 3.2 SCREENING OF SYSTEMS AS POTENTIAL TURBINE BUILDING FLOODING INITIATING EVENTS 5 3.3 IDENTIFICATION OF SYSTEMS AS POTENTIAL FLOODING SOURCES 10 3.3.1 Service Water FloodingEvents 11 3.3.2 CirculatingWater FloodingEvents 11 3.3.3 Fire Protection Water FloodingEvents 12 3.3.4 Summary of Turbine Building InternalFlooding Events 14 3.4 QUANTIFICATION OF INTERNAL FLOODING INITIATING EVENT FREQUENCIES 14 3.4.1 Service Water-InitiatedFlooding Events 14 3.4.2 Circulating Water Inlet Line-InitiatedFlooding Events 16 3.4.3 Circulating Water Outlet Line-InitiatedFlooding Events 18 3.4.4 Small CirculatingWater Expansion Joint Flooding Events 18 3.4.5 Random Breaks in FireProtection Water Piping 20 3.4.65 Steam Line Breaks Causing Large Fire ProtectionSystem Actuations 21 3.4.7 Steam Line Breaks CausingIntermediate Fire ProtectionSystem Actuations 21 3.4.7 Feedwaterand Condensate Line Breaks Causing Large FireProtectionSystem Actuations 21 3.4.8 Feedwaterand Condensate Line Breaks CausingIntermediateFireProtectionSystem Actuations 22 4.0

SUMMARY

22

5.0 REFERENCES

23 ADDENDUM 1, SERVICE WATER AND FIRE PROTECTION SYSTEM PIPING LEAK RATE CALCULATION Al-1 ADDENDUM 2, FIRE PROTECTION PIPE SEGMENT TABULATION A2-1 ADDENDUM 3, HIGH-ENERGY LINE PIPE LENGTH TABULATIONS A3-1 ATTACHMENT 1, TURBINE SUMP ALARM HISTORY ATTACHMENT 2, CIRCULATING WATER EXPANSION JOINT RUPTURE FREQUENCY ATTACHMENT 3, HIGH ENERGY LINE BREAK INITIATING EVENT FREQUENCIES

INTERNAL FLOODING - Initiating Events Analysis for TurbineBuilding foods p.22j 1.0 PURPOSE The purpose of the internal flooding initiating events analysis is to define, quantify, and document the frequency results for potential internal flooding initiating events caused by breaks of non-safety-related piping/components in the Turbine Building before February 2005. That is, the analysis considers the plant prior to installation of the flood mitigation modifications installed in and around safeguards alley. Flooding events caused by earthquakes are considered separately.

The following information is identified, correlated, and developed as part of this analysis:

• Identification of pipe breaks of concern

• Quantification of the frequency expected for pipe breaks in those systems.

2.0 MODEL DEVELOPMENT Internal flooding analysis encompasses the effects from the accumulation of fluids arising from the rupture, cracking or incorrect operation of piping/components within the station. In practice, major internal floods have occurred in nuclear power plants, from the rupture of pipes, valves and expansion joints as well as from operator errors during plant maintenance activities. All potential internal flood sources in the turbine building are considered in this analysis.

The steps for conducting the internal flooding initiating events analysis are described in the following section.

2.1 Steps for Turbine Building Internal Flooding Initiating Events Analysis The analysis of the Turbine Building internal flooding initiating events analysis consists of the following steps:

1. Determine the volume of water that can be released before failure of equipment in safeguards alley would be expected.

2. Screen from consideration, those systems that cannot be significant contributors to the overall turbine flooding risk. Screen from consideration systems that are not capable of causing failure of equipment even if the entire system volume is released or if a break in the system was allowed to flow for a long period of time.

3. Review information collected from the internal flooding walkdown and screening analysis

[NBO1] to identify potential flood sources. Review drawings to identify other potential flood sources not included in [NB01].

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p.33 l

4. Identify the specific piping and components that can cause an internal flood. For these pipes and components, calculate the frequency for flooding events of concern.

The results from each of these steps are presented in Section 3.0.

Development of the flood scenarios and accident sequence progression for each of the identified initiating events is documented in a separate report.

2.2 Turbine Building Internal Flooding Initiating Events Major Assumptions The key assumptions that were made during the internal flooding initiating events analysis are discussed in Section 3.0 for each of the specific flooding scenarios. In addition, the following general assumptions apply:

1. Actuation of fire sprinkler heads can also occur due to localized heating from operating equipment, aging failure, or impact damage from maintenance activities. Inadvertent actuation will result in discharge from a single sprinkler head, with a maximum rate of 30 gpm

[CALC01]. The low flow rate from actuation of a single sprinkler head is assumed to be too low to cause equipment damage outside of the immediate area and, therefore it would be no more severe than a loss of main feedwater event. Therefore, it is concluded that flooding events that result only in failure of equipment located in the Turbine Building can be considered subsumed by the frequency of loss-of-main-feedwater transient events.

2. All piping systems in the Turbine Building are assumed to be non-safety related. Therefore, all pipes are initially considered as potentially causing an initiating event.

3. All flooding events in the Turbine Building are assumed to cause a loss of main feedwater and, therefore, result in a reactor trip. If a flooding event does not cause a reactor trip, the flood could be excluded as an initiator. The effect of this assumption is that all pipe breaks are initially considered as potentially causing an initiating event.

4. The service water return lines are assumed to operate at the same pressure as the supply headers. The impact of this assumption is that some breaks in service water return lines that may be screened as initiating events are included in the overall initiating event frequency. The impact of this assumption is expected to result in only a slight increase in the overall initiating event frequency.

3.0 TURBINE BUILDING FLOODING INITIATING EVENTS ANALYSIS Identification and quantification of Turbine Building internal flooding initiating events is discussed below.

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 4 3.1 Determination of Water Volume to Fail Equipment in Safeguards Alley For this analysis, failure of non-safety related systems in the Turbine Building are considered. A flooding event which does not result in failure of equipment outside the Turbine Building would be no more severe than a loss of main feedwater event. Although some equipment used to mitigate a loss of main feedwater event could be failed by the flooding event, the expected impact of these additional failures would be bounded by the loss of main feedwater event modeled in the internal events PRA for the following reasons.

First, other than main feedwater, the only potentially risk significant plant equipment located in the Turbine Building basement are the service air compressors and plant equipment water pumps.

The plant equipment water pumps are located on the far southwest corner of the basement area such that a flooding event that would spray those pumps would be unable to spray any other equipment included in the PRA models. In addition, plant equipment water cooling is provided with a backup from service water so failure of these pumps would not directly cause failure of other equipment. The service air compressors are located in the north end of the turbine basement area such that a flooding event that would spray the service air compressors would be unable to spray any other equipment included in the PRA models. Also, the service compressors in the Turbine Building are provided with backup from instrument air compressors located in safeguards alley. Therefore, failure of the service air compressors located in the Turbine Building basement would not directly cause failure of other equipment. On the mezzanine level, non-safety related switchgear, Bus 3, Bus 4, and associated 480 VAC switchgear, and steam dump valves 11A and 1lB are located. In the PRA models, the non-safety related switchgear is used only for equipment that otherwise would be failed by the Turbine Building flood. Failure of the steam dump valves can be mitigated by using the steam generator power operated relief valves (PORVs).

The frequency of Turbine Building flooding events is much less than the frequency of loss-of-main-feedwater transient events. Therefore, it is concluded that flooding events that result only in failure of equipment located in the Turbine Building can be considered subsumed by the frequency of loss-of-main-feedwater transient events.

Water released to the Turbine Building will flow to the basement. Drain lines and gaps in doors allow the water to flow to the rooms in the safeguards alley. If the total volume of water released from a pipe break is less than the volume of water needed to fail enough equipment located in the safeguards alley that accident mitigation response is significantly impaired, then the pipe break can be excluded from consideration in the internal flooding events analysis.

Water flowing from the Turbine Building basement to the safeguards alley could potentially fail instrument air compressors, auxiliary feedwater (AFW) pumps, 480 VAC switchgear buses 51, 52, 61, and 62, 4kVAC buses 5 and 6, and diesel-generators 1A and lB. The first impact that a flooding event will have on equipment in the safeguards alley is when level reaches 2.75 inches on Bus 62 [CALCO2J when the bottom row of breakers on the bus would open [CALCO31 and the

IINTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p.5S I loads listed in Figure 3-2 of [CALC03] would be lost. The next impact of the flood would be when water level reaches 2.75 inches of water on Bus 52 [CALC02] when the bottom row of breakers on the bus would open [CALC03] and the loads listed in Figure 3-1 of [CALC03] would be lost. After loss of the bottom row of breakers on the 480 VAC safety buses, the next impact of a turbine building-flooding event would be loss of motor loads [CALC02] when level reaches 4 inches on Bus 5 [CALC03]. Note that the lockout relays submerged at the 4-inch depth on Bus 5 will only trip the breakers to the motor loads on Bus 5; the transformers to 480 VAC switchgear buses 51 and 52 will not be affected and buses 51 and 52 will still have power.

Reviewing the loads supplied from the bottom row of breakers in the 480 VAC safety buses shows that their loss would not present an immediate challenge to the ability of the operators to mitigate a reactor trip provided that the flood is isolated prior to the flooding event causing failure of other equipment in safeguards alley. The battery chargers are lost when the bottom row of breakers open. Therefore, actions to ensure longer-term availability of DC power must be taken.

If the flood is isolated before the A-train electrical safety buses would be failed, then the instrument inverters, BRA-111. BRA-112, BRB-l11, and BRB-112, could be powered from their alternate power supply. An evaluation in Attachment 1 to Appendix D shows that adequate time is available to switch inverter power supplies and maintain battery capacity in excess of twenty-four hours. Therefore, this analysis will screen from consideration any flooding event that does not result in water level reaching 4 inches on 4kVAC safety Bus 5.

Analyses show that if 131,000 gallons of water is released to the turbine building in 10 seconds, water level would reach only 2.9 inches on Bus 5 and 3.1 inches on Buses 61/62 [CALC02]. The same analyses show that a release of 200,000 gallons of water into the turbine building in 10 seconds would cause level to reach 5.7 inches on Buses 61/62 and 4.3 inches on Bus 5.

Interpolating between the two flood volumes above gives a flood volume of 185,000 gallons as the volume that would just fail the motor loads on Bus 5 and present the first significant challenge to the ability of the operators to mitigate a reactor trip. Therefore, any event that releases less than 185,000 gallons of water is screened from further consideration and the event can be considered subsumed by the loss of main feedwater event analyzed in the internal events PRA.

3.2 Screening of Systems as Potential Turbine Building Flooding Initiating Events Not all flooding events that release greater than 185,000 gallons of water need to be considered as initiating events. Any pipe break where the flowrate from the break would require more than one hour to release 185,000 gallons is eliminated from consideration. It is reasonable to expect these pipe breaks can be detected and isolated within one hour for the following reasons. First, a Miscellaneous Sump Level High alarm would be received. The alarm response procedure for that alarm [PROC01] directs the operators immediately to the Miscellaneous Drains and Sumps Abnormal Operation procedure [PROC02], which specifies that an operator be sent to investigate this alarm. The Miscellaneous Sump Level High alarm would be actuated before water exceeded the capacity of the turbine room sump and spilled onto the floor. The alarm is received infrequently (See Attachment 1) and typically only during evolutions where excessive water is U

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 6 being directed to the sump. From [PROC02] the operators would enter the appropriate abnormal operating procedure for the affected system.

For a system with a nominal pressure of 100 psig, a break with a three-inch equivalent diameter in a 4-inch line would result in a flow rate of 2100 gpm and a 3-inch equivalent diameter break in a 6-inch line would result in a flow rate of 1800 gpm (See Addendum 1 for details of the associated flow calculations). These flow rates are what would be expected from a sharp orifice-like break in a pipe and do not include any flow reduction that may occur due to head losses in the pipe from the pump to the break. With these flow rates, 88 and 102 minutes respectively would be available for the operators to isolate the break before equipment in safeguards alley would be threatened to the point that the ability of the operators to mitigate a reactor trip would be seriously challenged by the failure of Bus 5 motor loads. For lines smaller than 4-inches, the release rate would be, much less, allowing significantly longer than one hour to isolate the break.The service water system supply headers are maintained at a nominal pressure of between 90 and 100 psig

[REPORT06]. The service water return lines operate at a lower pressure, but will be assumed to operate at the same pressure as the supply headers. The fire protection system, when in standby is maintained at a pressure between 128 and 143 psig [REPORT01l].

Although the volume of the potable water and service water pre-treatment systems is essentially unlimited, the systems contain only small-diameter lines and operate at pressures generally lower than 100 psig. A break in these systems would be expected to result in a release rate that would allow significantly longer than one hour to isolate the break. Therefore, these systems are eliminated from further consideration as causing a negligible increase in flooding risk.

The turbine oil systems contain less than 185,000 gallons and, therefore, are eliminated from further consideration.

The reactor makeup storage tanks have a maximum capacity of 80,000 gallons and, therefore, are eliminated from further consideration.

The condensate storage tanks (CSTs) have a maximum capacity of 150,000 gallons and, therefore, are eliminated from further consideration.

Therefore, all systems except the circulating water, fire protection water, service water, and high-energy line breaks (HELBs) that result in fire protection water system actuation are screened from consideration as flooding sources.

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p.77 i Turbine Missile-Induced Flooding Events A flooding event could be caused if failure of the turbine generates a missile which then impacts and fails a system capable of causing a significant flooding event. An evaluation of turbine missile effects is presented in Appendix B.9 of the Kewaunee Power Station (KPS) Updated Safety Analysis Report (USAR) and is used as the basis for this analysis.

The probability of turbine missile generation due to fatigue has been determined to be much less than L.OE-08. For stress corrosion, the probability of failure and missile generation by the original low-pressure turbine rotors is determined to be 1.64E-03 at rated speed and 1.49E-05 for overspeed [CALC05]. Note that the latter value is lower than the former because the latter includes the probability of the overspeed condition. The total probability of turbine missile generation is the sum of these two values or:

PTMSS = PMissWtate + PMissOver PTtm = 1.64E-03 + 1.49E-05 PTOm = 1.65E-03 These failure probability values are based on a five-year inspection interval so the frequency of (_)

turbine missile generation is determined as follows:

FToliSS = PToteMiss / 5 years FTOtMiSS= 1.65E-03 / 5 years From.SS = 3.30E-04 per year.

Since the performance of the analysis that generated the above values, the low-pressure rotors have been replaced. As stated in USAR section 9.1, the probability of failure of the new rotors is less than the original rotors so the frequency calculated above is bounding for the current plant configuration.

Given that a turbine missile is generated, the probability that it impacts and fails a system capable of causing a significant flood must be considered. Missiles that occur on the operating deck may result in a steam release and could potentially impact the feedwater piping located on the southwest side of the building. Analyses [CALC06] have concluded that steam breaks on the turbine operating deck do not actuate sufficient fire protection sprinkders to present a flooding concern. Therefore, a turbine missile that impacts steam pipe on the operating deck does not present a flooding concern.

I INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 8 l The feedwater piping on the operating deck is located on the southwest end of the building across from the southernmost low-pressure turbine. Between the turbine and feedwater piping is a moisture separator reheater (MSR), steam piping, and building structural supports. Only a very small portion of the piping could be impacted by a turbine missile that does not first impact the intervening equipment and structures. Assuming that a missile that impacts the intervening equipment will not cause failure of the feedwater piping on the operating deck, it is estimated based on visual inspections that only 5% of the missiles would be capable of impacting the feedwater piping. Assuming that all turbine missiles that impact the feedwater piping cause failure of the piping and actuate fire protection sprinklers, the frequency of such events is:

(3.30E-04 per year)

• 0.05 = 1.65E-05 per year.

As described above, this frequency is bounding because the probability of failure for the new rotors is less than that of the old rotors on which these values are based. Also, this value assumes that all missiles that impact the feedwater piping penetrate the piping. Therefore, the frequency of turbine-missile-induced failures of feedwater piping on the operating deck would be negligible.

Turbine missiles that exit below the turbine shaft would be stopped by the concrete turbine support structure or imbedded in the condenser structure itself. Given the physical configuration of the turbine support structure and the condenser, a turbine missile would need to exit downward (J at a near vertical trajectory to imbed in the condenser. In doing so, the missile would contact the in-condenser feedwater heaters prior to contacting the circulating water tubes. If the missile did contact the circulating water tubes, such a failure would allow flow of circulating water back to the lake. Therefore, it is concluded that the flooding risk posed by turbine missiles that exit below the turbine rotor is considered negligible.

As described above, a conservative analysis of turbine-generated missiles concludes that the frequency of flooding events initiated by turbine missiles is sufficiently small as to be excluded from further analysis.

Tornado-InducedFloodingEvents Flooding events in the Turbine Building potentially could be initiated by the occurrence of a tornado which could fail systems either directly by wind loading or indirectly by causing a tornado-induced missile to impact and perforate a fluid system. Unlike random pipe failures where only a single system failure is considered at a time, a tornado could affect multiple systems simultaneously, thereby increasing the resulting flood height.

As described above, all systems except the circulating water, fire protection water, service water, and high-energy line breaks (IELBs) were screened from consideration as flooding sources. The systems were screened from consideration either because they contained insufficient inventory to

- damage equipment outside of the turbine building or because the flow rate that would result from any break would be low enough so that a very long time would be available for operator action to

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 9 isolate any flooding event prior to equipment damage outside the Turbine Building.

The screening of systems above is still valid with two exceptions; the condensate storage tanks (CSTs) and the reactor makeup storage tanks (RMSTs). When considered individually, the volume for each of these two sources is low enough that a flood which released their contents could not damage enough equipment outside the Turbine Building to seriously impair the ability of the operators to mitigate a reactor trip. Because the two sources are located near each other, a tornado could cause near simultaneous failure of all the tanks.

The primary flood risk in a tornado is due to a failure of the RMSTs and the CSTs in the tank room to the south of the auxiliary building. [CALC07] shows that the RMSTs would fail at lower wind loads than the CSTs. The capacity of each RMST is 40,000 gallons. Although some water could spill to other locations, such as outside, the maximum amount of water released from both RMSTs is 80,000 gallons. As discussed above, at least 185,000 gallons must be deposited in the turbine building basement to result in equipment failures in safeguards alley. Therefore, winds severe enough to fail the RMSTs, but not the CSTs, would not result in a significant risk increase.

Since the combined capacity of the CSTs is 150,000 gallons, there is a potential of damage to equipment in safeguards alley due to flooding from the combination of the four tanks.

[CALC08] shows that the frequency of CST damage due to direct tornado impact is 6.7E-7 per year. This reference also includes a discussion of tornado missiles. Specifically, the document states that tornado missiles are not a concern with wind speeds below 212 mph, which corresponds to an exceedance frequency of 7. 1E-6 per year. It also points out that most missiles would hit the upper portion of the tank, resulting in less that the full 150,000 gallons being released into the basement. Furthermore, for a missile to puncture the tank, the pipe must strike the tank nearly end-on along a radial line of the tank diameter. Any object that strikes slightly off normal or off the radial line would not be expected to penetrate the tank, but rather would be expect to glance off the tank without perforating it. Of the potential missiles that come within striking distance of the CST, only a fraction of them would be expected to strike the tank in such a manner as to be able to penetrate the tank. Therefore, the frequency of a tornado missile causing a flood of greater than 185,000 gallons of water to enter the turbine building basement is negligible.

Tornado-induced failure of the circulating water system is considered unlikely for several reasons.

First, the majority of the piping is located in the basement under the main turbine. The turbine building is designed such that it will not collapse (although the panels may fail) following a tornado so it is unlikely that the piping would be failed directly by the tornado. Secondly, the circulating water pumps are powered from the non-safety buses which require offsite power. It is likely that a tornado severe enough to threaten the circulating water piping would also cause a loss of offsite power, thereby removing the motive force for system flow and stopping the flood.

Third, tornado missile-induced failure is unlikely. A tornado missile risk analysis of the Kewaunee Power Station (KPS) was performed using the TORMIS methodology [CALC09]. In that study, the yearly probability of a tornado missile hitting either the diesel oil day tank vents, diesel exhaust

INTERNAL FLOODING - Initiating Events Analysis for Turblni Building Floods p. 10 I stacks, or the turbine-driven auxiliary feedwater pump exhaust pipe is 9.5E-06 per year and the probability of damaging one of the targets 1.7E-06 per year. These values are dominated by the concrete paver blocks located on the Turbine Building roof. Since all the circulating water piping is located below the turbine operating deck and, therefore, protected from such missiles, it is concluded that the tornado missile-induced failure probability is negligible.

The fire protection water header is located entirely in the Turbine Building basement, below grade. Several branch lines do extend to the mezzanine level to deluge valves and other equipment supporting system operation. Once on the mezzanine level, piping size reduces quickly. Only very short lengths of small-diameter piping to hose stations are located on the operating deck. As with the circulating water system, the fire protection water piping would be protected from direct failure in a tornado because of the ability of the Turbine Building to remain standing following such an event. The failure of fire protection water piping by missile impact is considered to be much lower than that calculated in [CALC07] and discussed above. Therefore, it is concluded that the risk from fire protection water flooding events initiated by tornados is negligible.

As with the fire protection water system, the majority of service water piping is located in the Turbine Building basement, below grade. No service piping is located on the operating deck.

Service water piping located on the mezzanine level is generally smaller in size, e.g., less than six inches nominal pipe size. Because the Turbine Building is designed to not collapse under tornado winds, direct failure of the service water piping is not expected. Failure of service water piping due to missile impact is considered to be a negligible contribution to risk as discussed above.

Also, the turbine header isolation valves would be available to isolate the Turbine Building header following a tornado. Therefore, it is concluded that the risk from service water flooding events initiated by tornados is negligible.

For a tornado to cause a IRELB, the event must first expose the Turbine Building to the outside winds. Because the Turbine Building contains blowout panels that are designed to fail, it is likely that the building would be open to the outside winds. The analysis of sprinkler actuation due to BELB [CALC06] shows that Turbine Building temperatures are reduced rapidly once the blowout panels fail. For a tornado-induced HELB, the blowout panels would fail prior to the BELB and the tornado winds would help mitigate any temperature rise caused by steam release.

Therefore, the number of sprinklers actuated for any HELB caused by a tornado would be much less than a similar size break initiated internally to the Turbine Building. Also, feedwater, condensate, and steam piping of concern to flooding events is designed for very high pressures and, therefore, much less likely than the diesel exhaust stacks to be damaged by tornado missiles.

Therefore, it is concluded that the risk from tornado-induced HELBs that actuate the fire protection system is negligible.

3.3 Identification of Systems as Potential Flooding Sources

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 11 For piping in the turbine building, only the service water, circulating water, and fire protection water contain sufficient volume or lines large enough to release fluid to the point that equipment in safeguards alley would be threatened in less than one hour. As described above, all other systems were screened as negligible contributors to flooding risk. Further analysis of these systems as potential flooding initiators is given in the sections that follow.

3.3.1 Service Water Flooding Events This initiating event will assume that all service water piping in the turbine building is supplied from the 20-inch turbine building header and is downstream of motor-operated valves SW-4A and SW-4B. There is service water piping that is in the turbine building but is not supplied from the turbine building header. Examples include auxiliary feedwater pump room cooler return lines to the standpipe, diesel cooling return lines, and air compressor cooling lines. With the exception of the diesel cooling return lines, piping in the turbine building that is not supplied from the turbine building header is small, e.g., 1.5-inches or less. Any leak from such lines would result in a low flow rate thereby providing the operators with a long time period to isolate the break using manually-operated valves local to the component. The diesel cooling return lines are normally isolated so any break in those lines would not result in a flooding event.

As discussed in Section 3.2, service water lines with a nominal diameter of less than four inches would not release of sufficient water in one hour to threaten enough equipment in safeguards alley that accident mitigation would be significantly impaired. Therefore, only breaks in service water lines four inches or greater are considered as potential initiating events.

3.3.2 Circulating Water Flooding Events A break from the circulating water system could result in the release of a very large amount of water in a short period of time. Calculations [CALC10] show that rupture of an expansion joint on the circulating water supply lines could be expected to release up to 58,000 gpm of flow.

Because the pressure on the return lines is less and because gravitational effects would tend to direct flow to the return header, a break in the circulating water return lines would release less flow to the turbine building. A rupture of an expansion joint on the circulating water return lines could be expected to release up to 14,000 gpm to the Turbine Building basement [CALCIO].

Because there is significant difference in the rate of release for the two locations, a large break in each location is considered as a unique initiating event. A break of the piping will be assumed to result in the same flow rate as the largest flow from a rupture of the expansion joint. In addition to the largest break sizes, an expansion joint rupture that results in less than the maximum flow is considered. For circulating water expansion joint ruptures less than the maximum flow, break sizes which lead to ruptures with leak flows between 2,000 and 10,000 gpm are considered.

INTERNAL FLOODING - Iniiaftg Events Analysis for TurbineBulilding Floods p. 12 3.3.3 Fire Protection Water Flooding Events The flooding event could be caused by an uncontrolled release of water from the fire protection system either because of a random break in the system or as a consequential release caused by a high energy line break (HELB). As discussed in Section 3.2, fire protection water lines with a nominal diameter of less than four inches would not release sufficient water in one hour to threaten equipment in safeguards alley. Therefore, only random breaks in fire protection water lines four inches or greater are considered as potential initiating events.

A HELB could raise temperatures in the Turbine Building to the point that fire protection sprinklers or deluge systems actuate. If a large number of sprinklers actuate, the potential exists to threaten equipment in safeguards alley. Breaks in the feedwater or condensate lines release a large quantity of water to the Turbine Building in addition to actuating fire protection systems.

Breaks in the steam systems do not result in an appreciable quantity of water being released to the Turbine Building. Therefore, steam line breaks are considered separately from feedwater and condensate line breaks.

Steam Line Breaks Analyses show that steam line breaks greater than nine inches equivalent diameter and upstream of the turbine building throttle valves will result in a safety injection (SI) signal [CALC06].

Because a SI signal inhibits operation of the fire pumps [REPORTOI], large breaks in the main steam system can be excluded as initiating events. In addition, the same analyses show that steam line breaks on the operating deck of the turbine building and-less than nine-inches in diameter will not actuate any fire sprinklers. Therefore, all steam lines on the operating deck can be excluded as initiating events.

For steam line breaks below the operating deck, calculations show that breaks smaller than two inches equivalent diameter actuate no fire protection sprinklers [CALC06], however, for the highest pressure main steam lines, i.e., upstream of the turbine throttle valves, a three-inch equivalent diameter break will actuate enough sprinlders that the fire pumps can be assumed to be providing full flow to the system.

For the extraction steam supply to the 15 feedwater heaters, a four-inch equivalent diameter break would actuate about 100 sprinklers while a six-inch or larger break would actuate enough sprinklers that the fire pumps can be assumed to be providing full flow to the system.

After steam exits the high-pressure turbine, a four-inch equivalent diameter break would actuate no fire protection systems while a six-inch break would actuate about 100 sprinlders.

Based on these results, two initiating events are analyzed for flooding events. The first is a steam X ,line break that actuates enough fire sprinlders to result in full flow from both fire pumps to the Turbine Building. This event includes any break upstream of the turbine throttle valves with an

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 13 lj I I equivalent diameter less than nine inches but greater than two inches, any break in the extraction steam line greater than six inches, and any break in a line after exiting the high-pressure turbine with an equivalent diameter of six inches or greater.

The second event is a steam line break that actuates approximately 100 sprinlders. The Turbine Building HELB models show that 100 sprinklers is representative of moderate releases. This event includes breaks in the extraction steam lines with an equivalent break size between two and six inches, and breaks in a line after exitipg the high-pressure turbine and having an equivalent diameter of two to six inches.

Feedwaterand Condensate Line Breaks This event initially considers breaks in any pipe containing main turbine working fluid above saturation conditions and includes all piping from the outlet of second feedwater heaters (12A and 12B). Analyses show that breaks upstream of the fourth feedwater heaters (14A and 14B) do not actuate any fire protection systems [CALC06]. In addition, the volume of water released from such breaks is less than the 185,000 gallons needed to threaten any equipment in safeguards alley.

Therefore, all breaks upstream of the fourth feedwater heaters can be excluded from further consideration.

For piping between the 14 and 15 feedwater heaters, breaks smaller than four inches equivalent C) diameter actuate no sprinklers. A six-inch equivalent diameter break in these lines would actuate about 100 sprinklers and a nine-inch equivalent break would actuate enough sprinklers that the fire pumps can be assumed to be providing full flow to the system.

For piping after the 15 feedwater heaters, a two-inch or smaller equivalent diameter break would actuate no fire protection systems. A four-inch break would actuate enough sprinklers that the fire pumps can be assumed to be providing full flow to the system Based on these results, two initiating events are analyzed for flooding events. The first is a feedwater or condensate line break that actuates enough fire sprinkders to result in full flow from both fire pumps to the Turbine Building. This event includes any break between the 14 and 15 feedwater heaters with an equivalent diameter of greater than six inches or any break downstream of the 15 feedwater heaters with an equivalent diameter greater than two inches.

The second event is a feedwater or condensate line break that actuates approximately 100 sprinklers. The Turbine Building HELB models show that 100 sprinlders is representative of moderate releases. This event includes breaks in the lines between the 14 and 15 feedwater heaters with an equivalent diameter between four and six inches.

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building noods Ip. 14 Il 3.3.4 Summary of Turbine Building Internal Flooding Events For internal flooding events in the turbine building, nine different initiating events have been defined for further analysis. The first is a break in the service water system in the Turbine Building and having an equivalent diameter of greater than four inches. The second event is a break in the circulating water supply lines. The third is a break in the circulating water return lines. The fourth is a circulating water break between 2,000 and 10,000 gpm. The fifth is a random break in fire protection water piping with the break having an equivalent diameter of greater than four inches. The sixth is a steam line break that actuates enough fire sprinklers to result in full flow from both fire pumps to the Turbine Building. The seventh is a steam line break that actuates approximately 100 fire sprinklers. The eighth is a feedwater or condensate line break that actuates enough fire sprinklers to result in full flow from both fire pumps to the Turbine Building. The ninth is a feedwater or condensate line break that actuates approximately 100 fire sprinklers.

3.4 Quantification of Internal Flooding Initiating Event Frequencies Quantification of the initiating event frequency for each of the nine events discussed above is performed in the following sections. Described within each section is the source of data used for system break frequency determination and how that data was used to calculate the initiating event frequency.

3.4.1 Service Water-Initiated Flooding Events To determine the frequency of service water-initiated flooding events, the frequency of pipe breaks is calculated using the methodology presented in EPRI TR 102266, "Pipe Failure Study Update", April 1993 [REPORT02]. Newer data sources that can be used to determine internal flooding initiating event frequency values have recently been published, i.e., EPRI TR 1012302, "Pipe Rupture Frequencies for Internal Flooding Probabilistic Risk Assessments (PRAs),"

[REPORT04]. However, service water initiating event frequency values calculated using the data and methodology of [REPORT04] are not expected to be significantly different from those calculated using [REPORT02]. Generally, it is expected that lower initiating event frequency values will result if calculated using [REPORT04] instead of [REPORT02]. In addition, the pipe segment data needed to calculate initiating event frequency values using the methodology of

[REPORT02] is already available. A significant effort would be needed to determine the pipe length data needed to employ the methodology of [REPORT04]. In addition, service water-initiated flooding events have been shown in prior, scoping studies to be a small contribution to overall risk from turbine building floods.

Therefore, the frequency of service water-initiated flooding events will be calculated using the methodology presented in [REPORT02].

Using that methodology, pipe breaks are categorized as large, medium, and small. A break in a

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p.15ll large pipe will not always be categorized as large. There is a probability that a large pipe will have a break in the medium or small category. Similarly, a medium pipe may have a break in the small category. When determining the frequency of breaks that result in the different categories, the recommended values from [REPORT02] will be used to determine the probability of equivalent break sizes.

The frequency for failure of components such as valves and heat exchangers is calculated using data from Eide, S.A. et al., "Component External Leakage and Rupture Frequency Estimates",

EGG-SSRE-9639 [REPORT03]. The following table gives the component rupture frequencies from [REPORT03] that are used in this analysis:

Component Rupture Frequencies Component Type Rupture/Leakage Rate (/hr)

Leakage L.OE-08 Valve non-PCS Rupture' 4.OE-10 PCS Rupture 1.OF10 Leakage 3.OE-08 Pump non-PCS Rupture 1.2E-09 PCS Rupture 3.OE-10 Leakage l.OE-08 Rupture 1.OE-10 Leakage 1.OE-07 Heat Exchanger Tube Side non-PCS Rupture 4.OE-09 PCS Rupture L.OE-09 Leakage .OE-08 Heat Exchanger Shell Side non-PCS Rupture 4.OE-10

-PCS Rupture L.OE-10 Leakage .OE-08 Tank non-PCS Rupture 4.OE-10 PCS Rupture L.OE-10

'PCS = Primary Cooling System It was assumed that the rupture of valves, pump casings, and other components have the same conditional probability of small, medium, large ruptures as for piping.

The initiating event frequency for service water-initiated flooding events in the turbine building will consider breaks in all pipes with a nominal size greater than four inches. Service water pipes and components are tabulated by size in [NB01]. As shown in Appendix F of [NB01], service water piping in the turbine is either four inches or smaller or six inches or greater. Twenty-seven pipe segments and nine valves were identified in the six-inch-or-larger category.

It will be assumed that large-bore piping breaks with an equivalent break diameter in the medium

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 16 l l (two-to-six-inch) category are not large enough to be of concern because breaks that size in large-bore piping have a sufficiently low flow rate to allow more time for recovery and, therefore, are not included in the total frequency of service water flooding events. Therefore, the frequency of large service water initiated flooding events in the Turbine Building was calculated to be:

Fsw = Fswpi~e + Fswvaive Fsw = ((27 pipe segments) * (1.39E-10 / pipe segment-hour) + (9 valves) * (4.0E-10 I valve-hour))

• 0.5 conditional probability of a large break [REPORT02]

Fsw = (3.75E-09 I hour + 3.6E-09 I hour)

• 0.5 Fsw = 3.78E-09 / hour Fsw = 3.22E-05 per year.

The contribution of maintenance-induced flooding events is considered negligible for several reasons. First, the maintenance event must be such that the event breaches the service water system pressure boundary but still permits operation of the plant and the turbine building header.

Actions such as cleaning heat exchanger water boxes could be performed. However, most valves in the systems could not be breached without securing the entire header. Therefore, the frequency of maintenance events is expected to be small. Second, the isolation valves for the service water-cooled heat exchangers are all manual valves located near the component being serviced. Should a breach of an unisolated component occur, the maintenance personnel would be able to quickly isolate the leak.

3.4.2 Large Circulating Water Inlet Line-Initiated Flooding Events Large flooding events from the circulating water system inlet lines could occur due to three causes, failure of the expansion joints, rupture of the piping and components in the system, or maintenance errors. The frequency of large failures, i.e., greater than 10,000 gpm, of expansion joints is documented in Attachment 2, which provides a failure frequency of 6.08E-06 per year per expansion joint. With four inlet expansion joints, the total frequency of expansion joint failures is calculated to be:

FcWItEXP = 2.43E-05 per year.

As with service water-initiated flooding events, the frequency of system breaks (excluding expansion joint breaks) is calculated using the methodology of [REPORT02]. Use of this methodology over the newer methodology recently published in [REPORT04] is judged to be acceptable for the same reasons explained in Section 3.4.1.

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 17 l Circulating water pipes and components are tabulated in [NB01]. As shown in Appendix F of

[NBO1], circulating water inlet piping contains ten pipe segments and four valves. Therefore, the frequency of large circulating water inlet-initiated pipe rupture events was calculated to be:

FcWINPipC = FcwpINipe + FCWINValve Fcwmnpipe = ((10 pipe segments) * (1.39E-10 / pipe segment-hour) + (4 valves) * (4.OE-10 /

valve-hour))

• 0.5 conditional probability of a large break [REPORT02]

FcwmNpipe = (1.39E-09 / hour + 1.60E-09 / hour)

• 0.5 Q.-I Fcwuipipe = 1.49E-09 / hour FcwmNPipe = 1.31E-05 per year.

A flooding event could be initiated during maintenance operations if the following conditions exist or events occur. First, operation of at least one circulating water pump must continue through the maintenance event. This would be expected for power operations. Second, the circulating water system pressure boundary must be breached. A breach would be expected for events such as cleaning water boxes. Third, a failure must occur so as to breach the isolation boundary from the circulating Water inlet header to the maintenance opening. Isolation failures are described in more detail below.

Only breaks greater than six inches equivalent diameter are considered because the circulating water system operates at a very low pressure and the flow rate from breaks less than six inches would be expected to allow a significant time period for operators to isolate the break. The only isolation failures that would be of concern are the condenser inlet isolation valves. These motor-operated valves are controlled from local push button stations. During the maintenance event, the valve would be closed, the breaker opened, and then the open breaker and valve hand wheel would be danger tagged. In addition, the push button station would be caution tagged.

Therefore, inadvertent opening of the valve would require that the danger tags be disregarded.

Then the valve must be manually opened sufficiently to allow flow to endanger turbine building equipment. Since the valves are located just below the water box inlets, it is unlikely that an operator would open a valve without noticing that water was being released. Similarly, if maintenance is attempted on an unisolated water box, then the operators would be expected to notice flow from the system as the pressure boundary is being unbolted. When leakage occurs, the operators can be expected to secure the area and investigate. Random failures of the valve disk are considered negligible. Therefore, flooding events initiated by maintenance on the circulating water system are considered negligible contributors to the overall initiating event frequency and are neglected.

The total frequency of large breaks in the circulating system inlet piping is the sum of the 0..

INTERNAL FLOODING - Initiatifg Events Analysis for Turbind Building Floods p. 1 8 lj frequency of expansion joint ruptures and the frequency of large pipe ruptures, or, FcwiN = FcwmNEp + Fcwmnwipe.

FcwIN = 2.43E-05 per year + 1.3 1E-05 per year FcwIN = 3.74E-05 per year.

3.4.3 Large Circulating Water Outlet Line-Initiated Flooding Events Flooding from the circulating water system outlet lines could occur due to three causes, failure of the expansion joints, rupture of the piping in the system, or maintenance errors. Failure of expansion joints used the information from Attachment 2 that provided a failure frequency of 6.08E-06 per year per expansion joint for failures with flow greater than 10,000 gpm. With four outlet expansion joints, the total frequency of expansion joint failures is calculated to be:

Fcwoumexp = 2.43E-05 per year.

As with service water-initiated flooding events, the frequency of system breaks (excluding expansion joint breaks) is calculated using the methodology of [REPORT62]. Use of this methodology over the newer methodology recently published in [REPORT04] is judged to be acceptable for the same reasons explained in Section 3.4.1.

Circulating water pipes and components are tabulated in [NBO1]. As shown in Appendix F of

[NBO1], circulating water outlet piping contains eight pipe segments but no components other than the expansion joints discussed above. Therefore, the frequency of large circulating water outlet-initiated pipe rupture events was calculated to be:

Fcwournipe = ((8 pipe segments) * (1.39E-10 / pipe segment-hour))

• 0.5 conditional probability of a large break [REPORT02]

Fcwourpipe (I.IIE-09 / hour)

• 0.5  ;,

Fcwiwpipe = 4.87E-06 per year.

A flooding event could be initiated during maintenance operations if the follow' conditions exist or events occur. First, operation of at least one circulating water pump must c'bntinue through the maintenance event. This would be expected for power operations. Second, t'h circulating water system pressure boundary must be breached. A breach would be expected for vents such as cleaning water boxes. Third, a failure must occur so as to breach the isolation odndary from the circulating water inlet header to the maintenance opening. Isolation failures are described in more detail below.

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 19 Only breaks greater than six inches equivalent diameter are considered because the circulating water system operates at a very low pressure and the flow rate from breaks less than six inches would be expected to allow a significant time period for operators to isolate the break. The only isolation failures that would be of concern are the condenser inlet isolation valves. These motor-operated valves are controlled from local push button stations. During the maintenance event, the valve would be closed, the breaker opened, and then the open breaker and valve hand wheel would be danger tagged. In addition, the push button station would be caution tagged.

Therefore, inadvertent opening of the valve would require that the danger tags be disregarded.

Then the valve must be manually opened sufficiently to allow flow to endanger turbine building equipment. Since the valves are located just below the water box inlets, it is unlikely that an operator would open a valve without noticing that water was being released. Similarly, if maintenance is attempted on an unisolated water box, then the operators would be expected to notice flow from the system as the pressure boundary is being unbolted. When leakage occurs, the operators can be expected to secure the area and investigate. Random failures of the valve disk are considered negligible. Therefore, flooding events initiated by maintenance on the circulating water system are considered negligible contributors to the overall initiating event frequency and are neglected.

The total frequency of large breaks in the circulating system outlet piping is the sum of the frequency of expansion joint ruptures and the frequency of large pipe ruptures, or, FcwouT = Fcwoump + Fcwourmipe.

FcwouT = 2.43E-05 per year + 4.87E-06 per year FcwouT = 2.92E-05 per year.

3.4.4 Small Circulating Water Expansion Joint Flooding Events Flooding from the circulating water system could result in break flow rates less than the maximum used flow described above. Such events could occur in either the inlet or outlet lines. Because all pipe breaks are assumed to result in the maximum flow and the pipe break frequency is included in the first two circulating water events, pipe breaks are not considered in this event. Therefore, this event considers only failures of the circulating water expansion joints that result in less than the maximum flow, which for this analysis is between 2000 and 10,000 gpm. (A 6000-gpm break was -deemed to be most representative of a small rupture.) The frequency for such events is documented in Attachment 2 which provides a failure frequency of 9.17E-06 per year per expansion joint. With four inlet expansion joints and four outlet expansion joints, the total frequency of expansion joint failures is calculated to be:

Fcwap = 7.34E-05 per year.

INTERNAL FLOODING - Initiating Events Analysis for TurbinieBuilding Floods p. 20 l 3.4.4 Random Breaks in Fire Protection Water Piping As with service water-initiated flooding events, the frequency of system breaks is calculated using the methodology of [REPORT02]. Use of this methodology over the newer methodology recently published in [REPORT04] is judged to be acceptable for the same reasons explained in Section 3.4. 1.

The initiating event frequency for random breaks in the fire protection water system considers breaks in all pipes with a nominal size greater than four inches. Piping drawings for the fire protection water system were reviewed and piping and components that are located in the turbine building and that cause a flooding event of concern were tabulated by pipe size. The piping tabulation in Addendum 2 identified 40 piping segments, 20 valves, and 26 flanges with a nominal size greater than four inches. Assuming that fire protection water piping is classified in the "other safety related" category used in [REPORT02], the frequency of fire protection water-initiated flooding events is calculated to be:

FFPR = FFPPipe + FFPvaivC + FFipnge FFPR = ((40 pipe segments) * (1.39E-10 / pipe segment-hour) + (20 valves) * (4.OE-10 I valve-hour) + (26 Flanges) * (1.OE-10 / flange-hour))

• 0.5 conditional probability of a large break [REPORT02I FFPR = (5.56E-09 I hour + 8.OOE-09 I hour + 2.6E-09 I hour)

• 0.5 FFPR = 8.08E-09 I hour FR = 7.08E-05 per year.

It will be assumed that large-bore piping breaks with an equivalent break diameter in the two-to-six-inch category are not large enough to be of concern because breaks that size in large-bore piping have a sufficiently low flow rate to allow more time for recovery and are not included in the total frequency of fire protection water flooding events.

The contribution of maintenance-induced flooding events is considered negligible for several reasons. First, the maintenance event must be such that the event breaches the fire protection water system pressure boundary but still allows the system to be pressurized. There are very few large-bore components that would permit such maintenance. Potentially, certain deluge valves could be breached. Next, the maintenance must be such that the breach would allow flooding to continue undetected for a significant time period following any operator error that resulted in an inadvertent breach. For deluge valves in the fire protection system, their associated isolation valve is immediately adjacent to the valve. Therefore, should a breach of an unisolated component occur, the maintenance personnel would be able to quickly isolate the leak. For these

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 21 ll reasons, the frequency of maintenance events to the fire protection flooding initiating event frequency is considered negligible.

3.4.5 Steam Line Breaks Causing Large Fire Protection System Actuations The first step in determining the frequency of steam line breaks that cause large fire protection system actuations is to determine the length and location of the steam pipes of concern. Piping layout drawings were reviewed and the dimensions indicated on them were used to determine the l length of steam pipes that are of concern to turbine building flooding events. Details of the pipe length data are listed in Addendum 3. Summing the lengths of high-pressure main steam piping located on the mezzanine and basement levels gives a total of 884.6 linear feet of piping.

Summing the lengths of extraction steam piping located on the mezzanine and basement levels gives a total of 176.5 linear feet of piping. Summing the lengths of lower-pressure steam piping located on the mezzanine and basement levels gives a total of 621.7 linear feet of piping. All other steam piping was located either on the operating deck or in the Auxiliary Building. (Note all of the steam piping tabulated is at least 6-inch diameter, and therefore of sufficient size to have the break flow required for a large fire protection actuation.)

Because not all high-energy line breaks would result in a turbine building flooding event, a separate analysis was performed to determine the frequency for steam line breaks of interest. This separate analysis uses the data of [REPORT04] and is documented in [REPORT05]. Refer to C)

[REPORT05], which is included as Attachment 3, for details of the calculations. From that analysis, the frequency of steam line breaks (including failures of valves, flanges, etc.) that result in large fire protection system actuations, FSLBL, is:

FSLBL = 2.53E-04 per year 3.4.6 Steam Line Breaks Causing Intermediate Fire Protection System Actuations Calculation of the frequency of this event is performed in [REPORT05] using the pipe length data described in Section 3.4.5 for large steam line breaks. From [REPORT05], the frequency of steam line breaks (including failures of valves, flanges, etc.) that result in intermediate fire protection system actuations, FSLBM, is:

FSLBM = 1.87E-05 per year 3.4.7 Feedwater and Condensate Line Breaks Causing Large Fire Protection System Actuations As discussed in Section 3.3.3, this event includes any break with an equivalent diameter greater than two inches in piping downstream of the 15 feedwater heaters and any break with an equivalent diameter greater than six inches between the 14 and 15 feedwater heaters.

INTERNAL FLOODING - Initiating Events Analysis for Turbind Building Floods p. 22 l The first step in determining the frequency of feedwater and condensate line breaks that cause large fire protection system actuations is to determine the length and location of the pipes of concern. Piping layout drawings were reviewed and the dimensions indicated on them were used to determine the length of feedwater and condensate pipes that are of concern to turbine building flooding events. Details of the pipe length data are listed in Addendum 3. Summing the lengths of feedwater piping downstream of the 15 feedwater heaters gives a total of 331.56 linear feet of piping. Summing the lengths of feedwater and condensate piping located between the 14 and 15 feedwater heaters gives a total of 696.65 linear feet of piping. (Note all of the feedwater and condensate piping tabulated is at least 12-inch diameter, and therefore of sufficient size to have the break flow required for a large fire protection actuation.)

Because not all high-energy line breaks would result in a turbine building flooding event, a separate analysis was performed to determine the frequency for feedwater and condensate line breaks of interest. This separate analysis uses the data of [REPORT04] and is documented in

[REPORT05]. Refer to [REPORT05], which is included as Attachment 3, for details of the calculations. From that analysis, the frequency of feedwater and condensate line breaks (including failures of valves, pumps, heat exchangers, etc.) that result in large fire protection system, actuations, FFLBL, is:

FFLBL = 1.35E-04 per year 3.4.8 Feedwater and Condensate Line Breaks Causing Intermediate Fire Protection System Actuations As discussed in Section 3.3.3, this event includes any break with an equivalent diameter between four and six inches between the 14 and 15 feedwater heaters. Identification and tabulation of the pipe lengths is described above.

Because not all high-energy line breaks would result in a turbine building flooding event, a separate analysis was performed to determine the frequency for feedwater and condensate line breaks of interest. This separate analysis uses the data of [REPORT04] and is documented in

[REPORT05]. Refer to [REPORT05], which is included as Attachment 3, for details of the calculations. From that analysis, the frequency of feedwater and condensate line breaks (including failures of valves, pumps, heat exchangers, etc.) that result in large fire protection system actuations, FFLBM, is:

FFLBM =4.69E-0S per year 4.0

SUMMARY

For the analysis of internal flooding caused by pipe and component failures in the turbine building that potential threaten equipment in safeguards alley, nine initiating events have been identified and their associated frequency values quantified. These events are summarized in the table below.

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods P.I3 LJ Event Consequence Frequency (per year)

Random Releases a large flow of Service Water to the Turbine Building 3.22E-05 Service Water Break Large Releases 58,000 gallons per minute to the Turbine Building 3.74E-05 Circulating Water Inlet Piping Break Large Releases 14,000 gallons per minute to the Turbine Building 2.92E-05 Circulating Water Outlet Piping Break Small Releases 6,000 gallons per minute to the Turbine Building 7.34E-05 Circulating Water Expansion Joint Failure Random Fire Releases full flow from both fire water pumps to the Turbine 7.08E-05 Protection Building Water Break Large Steam Actuates enough fire sprinklers that full fire protection water 2.53E-04 Line Break flow is released to the Turbine Building Intermediate Actuates 100 fire sprinklers that release fire protection water 1.87E-05 Steam Line flow is released to the Turbine Building Break Large Actuates enough fire sprinklers that full fire protection water 1.35E-04 Feedwater or flow is released to the Turbine Building Condensate Line Break Intermediate Actuates 100 fire sprinklers that release fire protection water 4.69E-05 Feedwater or flow is released to the Turbine Building Condensate Line Break

5.0 REFERENCES

[CALCO1] X-K-204-1696, Revised 4/5/1972.

[CALC02] MPR Calculation 064-0515-LYS-01, "Evaluation of Flooding Levels for Various

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p.24 l PRA Cases," Revision 0.

[CALC03] MPR Report 2823, "Past Operability Evaluation of Electrical Equipment Due To Flooding in Kewaunee Nuclear Power Plant," Revision 2.

[CALC05] "Results of Probability Analyses of Disc Rupture and Missile Generation," CT-24108, Revision 0, August 1980, Westinghouse Electric Corporation.

[CALCO6] MPR Calculation 0064-0509-DEC-03, Revision 0, "Turbine Building HELB Fire Protection Actuation Analysis."

[CALC071 Stevenson & Associates Calculation CAL-002: Evaluation of CST and RMST Wind Load.

[CALC08I PRA Application 05-13: Core Damage Frequency due to Tornadoes Damaging Condensate Storage Tanks

[CALC09] "Tornado Missile Risk Analysis for Kewaunee Nuclear Plant," July 2005, Applied Research Associates, Inc.

[CALC10] MPR Calculation 0064-0011-SPK-1, Revision 1, "Postulated Leaks Through Circulating Water Expansion Joint."

[NBO1] KPS Internal Flooding Analysis "Qualitative Screening Assessment and Flood Frequency Development," SCIENTECH, LLC.

[MANKl] "CRANE Technical Paper No. 410, Flow of Fluids Through Valves, Fittings, and Pipe", CRANE CO., 1988.

[PROC01l MDS-30, Annunciator Number 47033-P, "Miscellaneous Sump Level High," I Revision A, 5/28/1996.

[PROC02] A-MDS-30, Miscellaneous Drains and Sumps Abnormal Operation," Revision N, 3/5/2002.

[REPORT01] KPS System Description, "Fire Protection System," Rev. 2.

[REPORT02] Electric Power Research Institute, EPRI TR 102266, "Pipe Failure Study Update,"

April 1993.

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. 25

[REPORT03] Eide, S.A. et al., 1991, "Component External Leakage and Rupture Frequency Estimates", EGG-SSRE-9639, INEL, November 1991.

[REPORT04] Electric Power Research Institute, EPRI TR 1012302, "Pipe Rupture Frequencies for Internal Flooding Probabilistic Risk Assessments (PRAs)," September, 2005.

[REPORT05] Fleming, K.N. et al, "High Energy Line Break Initiating Event Frequencies for the Kewaunee PRA", October 2005.

[REPORT06] KPS System Description, "Service Water System," Rev. 3.

[REPORT07] Fleming, K. N. et al., "Expansion Joint Failure Rates for the Kewaunee PRA",

Final Report, November 2005.

INTERNAL FLOODING - Initiating Events Analysis for TurbineBuilding Floods p. Al-1 I ADDENDUM 1, SERVICE WATER AND FIRE PROTECTION SYSTEM PIPING LEAK RATE CALCULATION Infinite flood sources, such as Service Water and Fire Protection system, have been analyzed to determine the equivalent size of a pipe rupture that will potentially overwhelm the drainage capacity of a designated flood area. The analysis was performed using an Excel spreadsheet that calculates the flow equations listed below to determine flow rates from various rupture sizes in various diameter pipes.

System _

Gauge Pipe Equivalent Pressu Diamet Rupture re er , Diameter (di (AP)  ;(d2) _ '

Calculating the volumetric flow rate can be done by applying the following equation:

qft3,/. =C*A* 2g*l44*AP [MANO1], Eqn. 2.23 or expressed in Gallons per Minute (GPM)

QGPM (4 83 GPM *(C*A)* *144*AP t=ft /Sec) p Where:

AP = System Gauge Pressure (psig)

A = Equivalent Rupture Area (ft2)

C = Flow Coefficient (dimensionless) p = Density of Water (1b/ft3) t g = Gravity (32.17 ft/sec 2 )

The Flow Coefficient (C) for an orifice is calculated using the equation

. J; I Cd II I C = V11I!: I Ab &je A-20

1. II .

As stated in [MAN01], Table 3.10, the Discharge Coefficient (Cd) for a sharp-edged orifice is

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A1-2 I C

0.62.

The ratio of small to large diameter in an orifice (13) is defined as:

13= (dj)

[MANO11, Page A-20 (d 2 )

For calculation of the flow rates from a ruptured Service Water System, the following constants are used:

Piping Inside Diameters 1" Standard Schedule 40 Pipe: 1.049" [MANO1], Page B-16 2" Standard Schedule 40 Pipe: 2.067" [MANO1], Page B-16 4" Standard Schedule 40 Pipe: 4.026" [MANO1], Page B-16 6" Standard Schedule 40 Pipe: 6.065" [MANO1], Page B-16 Pressure Normal Service Water System Pressure (AP): 90-100 psig [REPORT06]

Density Water Density at 54 0F: 62.39 1b/ft3 Water Density at 74 0F*: 62.27 lb/ft3

[MANO1], Page A-6

[MANO1], Page A-6 0

For calculation of the flow rates from a ruptured Fire Protection System, the following constants are used:

Pressure Fire Protection System Pressure (AP) (standby): 128-143 psig [REPORT01]

Density Water Density at 85 0F: 62.17 lb/ft3 [MAN01], Page A-6 The table below shows the resultant flow rates for various rupture sizes in pipes with diameters 1 inch, 2 inch, 4 inch, and 6 inch for the service water and fire protection system pipes. The calculations used to develop the table used a pressure of 108 psig which is representative of the average pressure expected in both the service water and fire protection water systems. Since the flowrate is a function of the square route of the pressure, any differences on pressure have a minor impact on the overall results.

QI

f IC f IINTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p.A1-31 Table Al-i: Fire Protection System Piping Rupture Flow Rates (Pressure = 108 psig@ 85 0 F) 't Pipe Equivalent Diameter Rupture Flow Rupture Flow Inside Pipe Cross. Rupture Equivalent Ratio Beta Flow Rate Rate Diameter Sectional Area Diameter Rupture Area Factor Coefflcient ,(q) Q (in) (fe)(in) (ft_) (_)_(C) (ftsec) (GPM) 1 0.0060 0.50 0.0014 0.4766 0.6366 0.1101 49.4270 (ID 1.049) 0.75 0.0031 0 0.7150 0.7214 0.2808 126.0101 2 0.0233 0.50 0.0014 0.2419 0.6211 0.1074 48.2171 (ID = 2.067) 0.75 0.0031 0.3628 0.6254 0.2434 109.2536 0.90 0.0044 0.4354 0.6315 0.3539 158.8363 1.00 0.0055 0.4838 0.6377 0.4413 198.0387 1.50 0.0123 0.7257 0.7293 1.1355 509.6005 4 0.0884 0.50 0.0014 0.1242 0.6201 0.1073 48.1402 (ID= 4.026) 0.75 0.0031 0.1863 0.6204 0.2415 108.3678 0.90 0.0044 0.2235 0.6208 0.3479 156.1508 1.00 0.0055 0.2484 0.6212 0.4298 192.9053 1.50 0.0123 0.3726 0.6261 0.9747 437.4454 2.00 0.0218 0.4968 0.6398 1.7708 794.7316 3.00 0.0491 0.7452 0.7455 4.6425 2083.5477 6 02006 0.50 0.0014 0.0824 0.6200 0.1073 48.1356 (ID = 6.065) 0.75 0.0031 0.1237 0.6201 0.2413 108.3152 0.90 0.0044 0.1484 0.6202 0.3476 155.9935 1.00 0.0055 0.1649 0.6202 0.4292 192.6091 1.50 0.0123 0.2473 0.6212 0.9671 434.0229 2.00 0.0218 0.3298 0.6237 1.7263 774.7457 3.00 0.0491 0.4946 0.6394 3.9821 1787.1590 4.00 0.0873 0.6595 0.6885 7.6230 3421.2024 5.00 0.1364 0.8244 0.8452 14.6210 6561.8827

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A2-1 ADDENDUM 2, FIRE PROTECTION PIPE SEGMENT TABULATION

C C C I

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A2-2I l

Table A2-1: Turbine Building Fire Protection Water Piping IDto6" 2D ' 0.5' Q- 1D 2 targ Plpe Medlum P SIM" Plpe P ID Diameter From To Vohm Reno" From To Valve. Fle From To Valves

-r__

TU-t 1

10 WeltO53 Wait Tto FP 5-3

________ ___1___ _ _____ _ ________

TU-2 10 TtoFP 5-3 Tto FP 5-2 _

TW-3 10 Tto FP 5-2 Tto FPS.4 54 TU-4 10 Tto FP-4 tto FP 28-2 _

TU-5 10 Tto FP 28-2 TtoFP 54 TU16 10 Tto FP s Tto FP 5-6 _ -

TU-7 10 Tto FP S FP1t1 -_*

atFP 1-1 North to Turbine T1 10 Dulidno Well Tto FP I5-1 I T at FP 1-1 South 1U4 10 to Tat FP 22-1 Tet FP 22-1 .

T1-10 10 Tto FP 22-1 *TtoFPS-11 TU-11 10 TtoFP5-11 TtoFP2S-1 TU-12 10 TtoFP28-1 TtoFP3-6 TU-13 10 Tto FP3-6 Tto FP 3-5 TU-14 10 Tto FP 3- Tto FP 3-4 TiJ-15 10 Tto FP 3-4 TtoFP"3 3 TU-16 10 Tto FP 3-3 T to FP 3-2 10 to 64inch TU-17 10 Tto FP 3-2 reducerto FP 3-1 TU118 2.5 _ Tto FP 5-3 FP 5-3 1 2 T to hose TtJ119 2.5 _ FP 6-3 connection lines T to hose TU-20 1.5 . connection lines Up to FP 90-14 1 T to hose TL1-21 1.5 connection lines down to FP 90.7 1 TU-22 2.5 T to FP 5-2 FP 5-2 1 2 TU-23 2.5 FP 5-2 Tto FP 90-4 T11-24 Down to FP 90-1.5 t to FP 9D-4 4 Up to Tto FP 90-125 2.5 _ to FP 9o-4 13 -= TF_ ___93 1TU-28 1.5 __ ___ ____ TtoFP90-13

___ FP 90-13 1 ___

TU-27 1.5 _TtoFP90-13 _90_3 uptoFP TU-28 2.5 t to FP -4 FP5-4 1 2 Strainer and TU-29 2.5 f P 5-4 Deluge Valve 2 4 TU-30 8 Tto FP 28-2 FP 28-2 1 2 TU-31 4 f FP 28-2 FP 56-1

INTERNAL FOODING - Initiating Events Analysis for Turbine Building Floods p. A2-3 I. I Table A2-1: Turbine Building Fire Protection Water Piping ID 6 2 ID-cm I 0.5 M' ID - 2.

Large Plpe Medium Plp Smsll Pipe PRiAMI Dbmsetr Froml To Vllves P From To Volve.$ R. From To Valve. Fln.es

_____ _____ ________________ _______ 561___tTC ___ __ I_ _

TU-32 4 FP 55-1 Wait to TSC Tto MezzaNine Spdnkler Isolation TU43 8 FP 28-2 Valve Tto Mezzanine Basement Spdnkler Isolation Sprinider Isdation T-34 a Vael V.ah1 TW5 2.5 Tto FP 5-5 FP 5-5 1 2 TW5-3 2.5 FP 5-5 RedcIng T .

TU-37 1.5 Reducina T FP 90-9 1 TW-38 2.5 Reducna T FP 90-15 1 TW-39 1.5 . FP 90-15 FP-91-5 1 TW-40 2.5 Tto FP 5-6 FP 5-6 1 2 TU-41 2.5 FP 5-6 Tto FP 90-1O TU-42 2.5 Tto FP 90-10 Tto FP 90-16 I TU-43 1.5 _ Tto FP 90-16 Hose Station 21 TW4 1.5 . Tto FP 90-10 FP 90-10 TU-45 2 _ _ _ .

• Tto FP 22-1 FP 22-1 1 TW-6 2.5 Tto FP 5-11 FP 5-11 1 2 T to Hose TU1-47 2.5 FP 5-11 station 10 .

T to Hose TW-48 1.5 Station 10 Hose Station 10 I T to Hose TW9 1.5 Station 10 Hose Staton 16 1 TU-50 1.5 FP 5-11 Hose Station I 1 TU-51 8 Tto FP 28-1 FP 28-1 1 2  :

Tto Sprinkler TU-52 8 FP 28-1 Branch Unes ._.

Basement Tto SpdnWler Sprinkler Isolation TU-3 6 Branch Unes Valw TU-54 6 Tto FP 3-6 FP 19-6 2 2 TU-55 6 Tto FP 34 FP 19-5 2 2 .

TU4-6 6 Tto FP 3-4 FP 19-4 2 2 TU-57 6 Tto FP 3-3 FP 19-3 2 2 U8 6 Tto FP 3-2 FP 19-2 2 2 TU-59 6 Elbow to FP 3-1 FP19 1 2 2 T1U1-0 2.5 I TtoFP5-12 FP 5-12 1 2 .

TU-61 2.5 I FPS-12 TtoFP90-2 .

C7 C C

C C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A2-4 Table A2-1: Turbine Building Fire Protection Water Piping I 2 c.1D 6- 0.5 . 1D< 2

__ _ _ _ _ LaMe Pipe Medium Pip __ _ __ _ __ _ _ _ _ __ _ Pipe_

PiphlD Diameter Fron To

_ __- ___-_ _ i Rs F From To Tto FP 90-12 Valv Rlan From To LVa Fi TU-62 2.5 Tto FP 90-2 and fP 90.17 Tto FP9-12 TU-63 1.5 and iP 90-17 iP 90-12 t TU644 1.5 _ TtoFP90-2 FP 90-2 TU-65 6 _

_ Tto FP 90-17 FP 90-17 1 Tat column 8 that TtoMezranine spIis to 3-inch Spdnider iaolation header and 5inch ni4-66 6 Vaie Une 111= header 2 3_.

T at column 8 that splits to 3-Inch header and 1.1-67 5 5-inch header Riser T and Riser.

TU-68 5 Riser labeled 0 _

Tand Riser T and dserto TU-69 5 labeled 0 branch 317 Tand dserto TU-70 5 branch 317 Rlser labeiedK K__ _

TU-71 5 Riser labeled K Riser labeled G .

Basement SprinklerIsolation Tto lines 21 and IU-72 6 Valve 15 2 S  :

T to lines 21 and T to Riser labeled 1U-73 6 15 J Basement Spwtnkier Isolation TU-74 6 VaivW ine TU-53) T to dser labeled S 2 3 1.1-75 10 TtoFP15-1 TtoP 5-10 TU-76 10 Tto FP 5-10 Wail _

TU-77 3 Tto iP 5-10 Wall 1 4 78 TU-78 Z5 ToteIS 40 _ _

__Tu_____

20 26 Tto fP 5-10 FP 23-1

__ 1 2 1____ _

_ _ __ 22__

_ _ _ _ _ _ 15 26 16___

_ __ _ __ _ 13 0

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-1 I ADDENDUM 3, HIGH-ENERGY LINE PIPE LENGTH TABULATIONS

c -r INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-2 Table A3-1: High-Pressure Main Steam Piping BLDG Dwg. No. I Building Level Drawing Coordlnates/Deseriptlon Detail I HorizJVert. PipeLengt Clued#,Letter(floor D I I Dla fin% I IranII /A--I. I 1 I-ng C I nlen n lc Aux M-238 MOA- 1to 9I0 BMnd JII1V IFLu IV onILz.

MS1A to 90 Band Aux M-238 62Z-0"' 30 A3 - "lA" Train MS Piping from 900 Horiz 16.52 6S, M Bend to 90° Bend I Aux M-238, - 622-0" to 639-6" 30 A3 - "lA Train MS Piping from 900 "15-15" Vert. 17.50 6S, M 240 Bend to Floor Penetration at 639-6" Aux M-238 639'-6" to 30 A3 - *1A" Train MS Piping from Floor Vert. 27.45 6S, M 664'-1 1 7/186 Penetration at 639-6" to 90° Bend Aux M-238 664-1i1 7/16" 30 A3--'A' Train MS Piping frorn 960 Horiz. 29.43 6S-6, M-L Bend to 400 Bend Aux M-238 664-4 3/4" 30 B4 - *1A' Train MS Piping from 400 Horiz. 98.43 6, L-H Bend to 90° Bend Aux M-238 664'-4 3/4" 30 E4 - *1A' Train MS Piping from 90° Horiz. 34.52 6-7, H Bend to 900 Bend Aux M-238, - 664 to 30 E4 - *1A Train MS Piping from 900 *16-16" Vert. 11.27 7, H 240 652'- 1 3/4 Bend to 900 Bend Aux M-238, - 652-11 3/4" 30 E5 - "lA" Train MS Piping from 900 *16-16" Horiz. 58.99 7, H-G 240 Bend to Turbine Building Wall Penetration (Oper. Deck Level)

Aux M-238 620' 30 F2 - *1B" Train MS Piping from MSIV Horz. 35.92 4, HE-G

. MS1 B to 90° Bend Aux M-238 620'-" _ 30 G2 - *1 B" Train MS Piping from 900 Horiz. 31.46 5, G Bend to 900 Bend Aux M-238 -620wt: 30 G2 - "18" Train MS Piping from 90° Horiz. 5.49 I 5, G Bend to Turbine Building Wail Penetration (Mez. Level)

Linear FT on 622'+ Level 277.59 Linear FT on 622' - Level 97.02 Linear FT on Basement Level 0.00 Total Length (Llnear FT) 374.61

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floodsp p. A3-3 I l

Table A3-1: High-Pressure Main Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom. Drawing Coordinates/Descriptlon Detail HorizJVerL Pipe Length Quad #,Letter (floor I . . . ID Aiel. I Il _Section I /Angle I (Linear FT) I plan quads)

, -ienwiu - - -

TB M-984-1 Oper Deck 30 E8 - '1A' Train MS Piping ThroughTS Vert.T 28.19 7-8, G-F Wall (Approx 660' elev.) 90° Elbow

_ Through Pipe Chase to TB Mez.

TB M-984-1 Mez. 30 F8-'1A'Train MS Piping From Oper. Horui. 73.04 7-8,G-C Deck Opening Thnu Mez. 900 Elbow to 900 Towards HP Turbine TB M-984-1 Mez. 30 H6- 1A Train MS Piping From Mez. Vert. 9.42 7-8, D-C 90 Elbow to Oper. Deck. Penetration Towards HP Turbine TB M-984-1 Mez. 30 H6 -'1A'Train MS Piping From Oper. Vert. 4.75 7-8, D-C Deck. Penetration Towards HP Turbine to 90 Elbow TB M-984-1 Oper Deck 30 H6 -'"1A Train MS Piping Frorn 900 Hodz. 16.06 7-8, D-C Elbow at Oper. Deck. Penetration Towards HP Turbine Stop Valve Inlet Connection TB X-K-101- Oper Deck 30 '1A' MS Piping From Valve MS-3A to Vert. 4.06 6-7,D-C

_ 30 Oper. Deck Floor Penetration (U-PIPE) _______

TB X-K-101- Mez. 30 '1A' MS Piping From Oper. Deck Floor Vert. 19.50 6-7,D-C 30 Penetration to 90 Elbow TB X-K-101- Mez. 30 l 1A MS Piping From 900 Elbow to 90 l Horlz. 5.75 6-7,D-C 30 Elbow _ _____

TB X-K-101- Mez. 30 l 1A" MS Piping From 900 Elbow to HP Vert. 13.30 6-7,D-C 30 Turbine TB M-985-1, Mez. 30 D8/D9-'1B"TralnMSPipingThruTB Horiz. 117.08 5-8,G-D

-2 Wall into Mez. Level to 90 Elbow to l Oper. Deck TB M-985-2 Mez. 30 D - '1B' Train MS PipingFrom 900 Vert. 4.00 7-8, E-D Elbow in the Mez. to Oper. Deck Penetration C C C

C C C lINTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-4 Table A3-1: High-Pressure Main Steam Piping (cont.)

BLDG Dwg. No. Building Level Norn. Drawing Coordinates/Description Detail HorlziVerL Pipe Length Quad #,Letter (floor Dia. (In) Section /Angle (Linear FT) plan quads)

TB M-985-2 Oper Deck 30 C7 - ' 1B3Train MS Piping From Oper. Vert. 10.33 7-8, E-D Deck Penetration to 900 Elbow TB M-985-2 Oper Deck 30 C7 - *1B' Train MS Piping From 90° Horiz. 13.69 7-8, E-D Elbow at Oper. Deck. Penetration Towards HP Turbine Stop Valve Inlet Connection TB X-K-101- Oper Deck 30 '1B" MS Piping From Valve MS-3A to Vert. 4.06 6-7,E-D 30 Oper. Deck Foor Penetration (U-PIPE)

TB X-K-101- Mez. 30 '1B' MS Piping From Oper. Deck Floor Vert. 19.50 6-7,E-D 30 Penetration to 900 Elbow _

TB X-K-101- Mez 30 '1 B MS Piping From 900 Elbow to 900 Horiz. 5.75 6-7,E-D 30 Elbow _

TB X-K-101- Mez 30 *1B" MS Piping From 900 Elbow to HP Vert. 13.30 6-7,E-D 30 Turbine FT on Mez. Level L0i 0lgLnear 285.39 9,1_Total Length (Linear FT) 361.78 TB M-239 lOperDeck 18 A6 --lA' TrainS Steam Dump Piping 8-8' Horiz. 3.68 7-8, F-G III l900 ETee

.lS Fro Beme l b wwith eck FloortDMez.7 Lnda- aper. 6e3 9 Ifrom 30' MSLhne Too 1 1 Elbow F Bend 0 t 90 Elo TB lM-239 lOperDeck l 18 lA6-'lA'Train Steam DumpBPipingn90d l - Vert. 2.97 78,F-G Elbow Bend Thru Oper.-Deck Floor to l lMez.

TB M-239 lMez. l 18 lA6 -'1A' Train Steam Dump Piping l 8-8' Vert. 4.10 l7-8, F-G lFrom Oper. Deck Floor to Mez. Level l 90° Elbow Bend TB M-239 lMez. l 18 lA6 -'1A'Train MS SteamnDump Piping Horkz. 13.55 7-8, F-G llFrom 90° Elbow Bend to 90°Elbow lBend .

Floods p. A3-5 INTERNAL FLOODING - Initiating lIINTERNAL -

Analysis for Events Analysis Initiating Events Turbine Building for Turbine Building Floods , p. A3-5 I Table A3-1: High-Pressure Main Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom. Drawing CoordinatesaDescription Detail HoriziVert. Pipe Length Quad #,Letter (floor Dia. (In) Section /Angle (Linear FT) plan quads)

TB M-239 Mez. 18 A6 - "1A Train Steam Dump Piping 18-8" Angle 5.17 7-8, F-G From 90 Bend Thru 30° Down Angle Towards "7 Une TB M-239 Mez. 18 A6 - "lA" Train Steam Dump Piping Horiz. 35.38 6-7, F-G From End of 300 Down Angle To 900 Elbow Bend TB M-239, - Mez. 18 B4- 1lA! Train Steam Dump Piping "19-19" Horiz. 15.17 5-6, F-G 240 From 90° Bend to Capped End TB M-239 Mez. 18 A3 - 1 B" Train Steam Dump Piping Horiz. 4.96 5-4, F-G from Tee in 30" MS Une 450 Elbow Bend at "5" Une TB M-239, - Mez. 18 D8 - "1B" Train Steam Dump Piping "25-25" Angle 3.18 5-4, F-G 241 from 45° Elbow Bend to 450 Elbow Bend TB M-239 Mez. 18 A3 - *1"B Train Steam Dump Piping Honz. 14.40 4-5, F-G from 45 Elbow Bend to 90° Elbow Bend TB M-239, - Mez. 18 A3 - '1B" Train Steam Dump Piping "20-20" Angle 3.03 4-5, F-G 241 from 900 Thru a 45° Declined Angle to 450 Elbow Bend TU M-239 Mez. 18 A3 - 1B" Train Steam Dump Piping Horiz. 18.17 4-5, E-F from 450 Elbow Bend to Capped End C C. 0

C C ,,

IINTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-6I Table A3-1: High-Pressure Main Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom.. Drawing Coordinates/Descrlption Detail HoriziVert Pipe Length Quad #,Letter (floor Dla 11111 Section /A le (Linear F glen Suads TB M-239, - Mez. 8 B4- "lA" Train 8' Steam Dump Piping "22-22' Horiz. 208 5-6, E-F 241 From 18 Main Header to 90° Elbow (2 of 3)

TB M-239, - Mez. 8 B4- 'AA' Train 8' Steam Dump Piping "22-22 Vert. 36 5-6, E-F 241 (2 of 3) From 900 Elbow to 900 Elbow TB M-2S9, - Mez 8 84- *tA' Train 8" Steam Dump Piping '22-22' Hoiz. 1.67 5-6, E-F 241 (2 of 3) From 900 Elbow to 90° Elbow TB M-239, - Mez. 8 B4- *tA' Train 8" Steam Dump Piping *22-22' VerL 7.87 5-6, E-F 241 (2 of 3) From9 0 Elbow to9 0 Elbow TB M-239, - Mez. 8 B4 - *tA' Train 8" Steam Dump Piping *22-22' Horiz. 16.67 5-6, E-F 241 (2 of 3) From 900 Elbow to Tee in 16' Une TB M-2S9, - Mez 8 B4 - lA1Train 8" Steam Dump Piping *2-22 Horiz 375 5-6. E-F 241 From 18" Main Header to 90° Elbow (3 of 3)

TS M-2S9,- Mez. 8 B4 - 'tA- Train 8" Steam Dump Piping *-222' Vert 2.21 5-6, E-F 241 (3 of 3) From 9 0 Elbow to90 Elbow _

TB M-239, - Mez. 8 B4- *tA Train 8" Steamn Dump Piping '22-22Y Horiz. 225 5-6, E-F 241 (3 of 3)From 90 Elbow to 90° Ebow TB M-2S9, - Mez 8 B4 - 'lA' Train 8" Steam Dump Piping 22' Vert. 929 5-6, E-F 241 (3 of 3) From 900 Elbow to 900 Elbow TB M-239, - Mez. 8 B4 - 'IA' Train 8" Steam Dump Piping "22-22' Horiz. 8.83 5-6, E-F 241t (3 of 3) From 900 Elbow to Tee in 16'

_Une TB M-239, - Mez. 8 B3 -8 "BTrain 8" Steam Dump Piping "20-20" VerL 2n00 4-5, E-F 241 (1 of 3)Fromt8 MainHeaderto9 0° '18-18' Elbow__ _ _ _ _ _ _ _

TB M-239, - Mez. 8 B3 - "1B" Train 8' Steam Dump Piping "18-18" Houz. 22.00 3-5, E-F 241 (1 of 3) From 90° ElbowThru 1800 Retum t 900 Elbow

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3.7 I I

Table A3-1: High-Pressure Main Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom. Drawing Coordinates&Description Detail HorizJVert Pipe Length Quad #,Letter (floor Dla. (In) Section /Angla (Linear FT) plan quads)

TB M-239, - Mez. B3 - *1"B Train 8" Steam Dump Piping *20-20" Angle 0.00 (FT included in 9.15 241 (1of 3) From 90° Elbow at 450 Declined below) 4-5, E-F Angle to 450 Elbow TB M-239,- Mez. 8 B3- 'IB" Train 8" Steam Dump Piping "20-20" Vert. 9.15 4-5, E-F 241 (1of 3) From 450 Elbow to Capped Tee TB M-239, - Mez. 8 B3 - 1B- Train 8" Steam Dump Piping "20 Hor.z 9.00 45, E-F 241 (1o )Fon8Tee to t16 x8 Reducer TB M-239, - Mez. 8 B3 - "1B" Train 8" Steam Dump Piping "21-21" Vert 1.83 45, E-F 241 (2 of 3) From 18" Main Header to90 Elbow TB M-239 Mez. 8 B3 - '1B" Train 8" Steam Dump Piping Hodz. 3.58 45, E-F

_ (2 of 3) From 900 Elbow to 900 Elbow TB M-239, - Mez. 8 B3 -"1B" Train 8" Steam Dump Piping "21-21" Vert. 4.13 4-5, E-F 241 1 (2 of 3) From 900 Elbow to 90° Elbow TB M-239, - Mez. 8 B3 - "1B Train 8" Steam Dump Piping "21-21" Horiz. 30.13 4-5, E-F 241 (2 of 3) From 900 Elbow Thru 1800 Bend to 900 Elbow TB M-239, - Mez. 8 B3 - 'IB" Train 8" Steam Dump Piping "21-21" Horiz. 2.50 4-5, E-F 241 (2 of 3) From 900 Elbow to 900 Elbow TB M-239, - Mez. 8 B3 - "1B" Train 8" Steam Dump Piping "21-21" Vert. 7.69 4-5, E-F 241 (2 of 3) From 900 Elbow Thru SD1-5 to 900 Elbow TB M-239, - Mez. 8 B3 - "iB" Train 8" Steam Dump Piping "21-21" Horiz. 6.17 4-5, E-F 241 (2 of 3) From 900 Elbow Thnu FCV-484E and SD2-5 to 900 Elbow TB M-239 Mez. 8 B3 - *1B Train 8 Steam Dump Piping Horiz. 3.50 45, E-F (2 of 3) From 900 Elbow toTee In 16" Une TB M-239, - Mez. 83- "iB 18 Train 8" Steam Dump Piping "21-21" Vert. 3.00 45, E-F 241 (3 of 3) From 18" Main Header to 90° Elbow C C C)

C C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p.A3-8]

Table A3-1: High-Pressure Main Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom. Drawing Coordinates/Descriptlon Detail HorlzJVert Pipe Length Quad #,Letter (floor Dia. (In) Section /Angle (Linear FT) plan quads)

TB M-239, - Mez. 8 B3 - "1B" Train 8" Steam Dump Piping "21-21" Horkz. 10.45 4-5, E-F 241 (3 of 3) From 90° Elbow to 90° Elbow TB M-239, - Mez. 8 B3 '15B

" Train 8" Steam Dump Piping "21-21" Vert. 7.58 4-5, E-F 241 (3 of 3) From 90° Elbow to 9W0 Elbow TB M-239, - Mez. 8 B3 - 'IB- Train 8" Steam Dump Piping "21-21" Horiz. 7.13 4-5, E-F 241 (3 of 3) From 900 Elbow to 90 Elbow TB M-239, - Mez. 8 B3 -"1B Trahn 8" Steam Dump Piping *21-21" Horiz. 7.25 4-5, E-F 241 (3 of 3) From 90 Elbow to 90° Elbow TB M-239, - Mez. 8 B3 - '1B Train 8" Steam Dump Piping "21-21" Horkz. 11.83 4-5, E-F 241 (3 of 3) From 900 Elbow to 90° Elbow TB M-239, - Mez. 8 B3 - "1B6Train 8" Steam Dump Piping "21-21" Horiz. 4.00 4-5, E-F 241 (3 of 3) From 900 Elbow to 90 Elbow TB M-239, - Mez 8 63- "16B Train 8" Steam Dump Piping "21-21" Vert. 6.08 4-5, EF 241 (3 of 3) From 90° Elbow Thru Valve SD1-6 to 90 Elbow TB M-239 Mez. 8 B3 - "16BTrain 8" Steam Dump Piping Horiz. 6.17 4-5, E-F (3 of 3) From 90" Elbow Thru Valves FCV-484F and SD2-6 to 90° Elbow TB M-239 Mez. 8 83 - "1B Train 8" Steam Dump Piping Horiz. 5.75 4-5, E-F (3 of 3) From 900 Elbow to Tee in 16" Une TB M-239 Mez. 8 C6 - *1A Train 8" Steam Une From 30" "5-5" Horiz. 7.42 D4, 7-8 Main Header to 900 Elbow *SA-5A" TB M-239 Mez. 8 C6- "IA" Train 8" Steam Une From 90° *SA-5A" Veir. 2.29 D4, 7-8 Elbow to 900 Elbow TB M-239 Mez 8 6- *1A' Train 8" Steam Une From 90 Horiz.-C, 7-8 Elbow to 900 Elbow TB M-239 Mez. 8 C6/D6 - 'IA' Train 8" Steam Une From "6"i 99.00 B-C, 8-4 900 Elbow to 8"x4 Reducer Linear FT on Oper. Deck 0.00 Linear FT on Mez. Level .381.87 LieaFTFTon Bmn Level onBasement L vl00 Linear 0.00

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floodsp.A- p. A3-9.1

.Table A3-: High-Pressure Main Steam Piping (cont.) I BLDG _I Dwg. No. I Building Level Norn. Drawing Coordinates/Description Detaill HorlziVert.I Pipe Length Quad #,Letter (floor II I Dia. (in) IISection /Angle (Linear FT) I plan quads) I II Y-r-.t i - 41. . ~I a.. 0 I FVMC.. Mez. 0 13 --- ,XI rain Vo.tearnUne TronMI -

240 Steam Dump to 900 Elbow TB M-239, - Mez. 6 B4AB5 - 'lA Train 6" Steam Une from "17-17" Vert. 6.75 5-7, F-E 240 91f Elbow to Oper. Deck Level

________Penetration TB M-239, - Oper Deck 6 B4/B5 - "1A"Train 6"SteamnUno from "17-17" Vert. 2.00 5-7, F-E 240 Oper. Deck Level Penetration to 900

_ _ _ _ _ __ _ _ _ _ _ _ _ _ _Elbow TB M-203, - Oper Deck 6 B.4/B5 - '1lA" Train 6" Steam Uno from '17-17" Horiz. 7.65 5-7, F-E 239, -240 90Elbow Thru Valves (MS200B1,

____ ____MS201 B1) to 90 Elbow TB M-239, - Oper Deck 6 B4/B5 - '1A~Train 6" Steam Line from "17-17" Vert. 8.34 5-7, F-E 240 90 Elbow to 900 Elbow TB M-239, - Oper Deck 6 B5 - "lA" Train 6" Steam Une fromi 900 "17-17" Hodz. 11.50 5-7, F-E:

240 Elbow Thru Orifice to 900 Elbow TB, M-239, - Oper Deck 6 B5 - "lA' Train 6" Steam LUne fromi 900 "17-17" Houlz.675 57 E 240 Elbow to 90 Elbow TB M-239, - Oper Deck 6 B5- "1A"Traln6"'Steam Une from 960 "17-17" Angle200 57 E 240 Elbow to Moisture Sep/Reheater Bi TB M-239, - Mez. 6 B3 -"1 B"Traln6"'Steam Une from 18" "18-18" Vert.308 53 E

____240 _______Steam Dump to 900 Elbow ___ ______

TB M-239, - Mez. 6 B2/B3 - 'IB"Train 6"Steam Une from "18-18" Horiz. 1.3 53 -

240 900 Elbow to 900 Elbow TB M-239,,- Mez. 6 B2/B3 - '1B" Train 6" Steam Une from "18-18" Vert. 3.25 5-3, F-E

'240 900 Elbow to Oper. Deck Level Penetration TB M-239, - Oper Deck 6 B2/B3 - "iB" Train 6' Steam Una from "18-18" Vert. 2.00 5-3, F-E 240 Oper. Deck Level Penetration to 900

___ ____ ___Elbow C C C

C CF C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-10 Table A3-1: High-Pressure Main Steam Piping (cont)

BLDG Dwg. No. Building Level Nom. Drawing CoordinatesIDescriptlon Detail HorlzJVerL Pipe Length Ouad #,Letter (floor Dia. (In) Section /Angle (Unear F plan quads)

TB M-239, - Oper Deck 6 B2/B3 -. "B" Train 68"Steam Une from "18-18" Horiz 7.65 5-3, F-E 240 90° Elbow Thru Valves (MS200B2, MS201B2) to 90° Elbow TB M-239, - Oper Deck 6 B21B3 - *1B"Traln 6' Steam Une from "1-18 Vert. 8.34 5-3, F-E 240 900 Elbow to 90° Elbow TB M-239, - Oper Deck 6 B2 - *1B- Trahn6 Steam Unefrom 90°18-186 Horiz. 11.50 5-3, F-E

. 240 Elbow Thru Orifice to 900 Elbow TB M-239 Oper Deck 6 B2 - *1B- Train 6" Steam Une from 90 Horiz. 6.75 5-3, F-E Elbow to 900 Elbow TB M-239, - OperDeck 6 B2 - *1B Train 6" Steam Unefrom 900 "18-18" Angle 2.50 5, F-E 240 Elbow to Moisture Sep/Reheater B2 TS M-239 Mez. 6 D4/E4 - *lA"Train 6" Steamr Une from .9-9, Vert. 5.19 6-7, B-C 8" Steam Supply Une to Oper. Deck Penetrton TB M-239 Oper Deck 6 D4/E4-'IAP Train 6" Steam Une from "9-90 Hofiz. 2.00 6-7, B-C Oper. Deck Penetration to 90 Elbow TB M-239 Oper Deck 6 D5/E5 - '1A' Train 6" Steam Une from .9-9. Horiz. 7.65 6-7, B-C 900 Elbow Thru Valves (MS200A1, MS201Al) to 900 Elbow :_.

TB M-239 Oper Deck 6 D5/E5 - l A' Train 6" Steam Line from r9n9" Vert. 8.34 6-7, B-C 900 Elbow to 900 Elbow TB M-239 Oper Deck 6 D5WE5 - 'lA' Train 6' Steam Une from .9-9. Horiz. 11.50 6-7, B-C 900 Elbow Thru Orifice to 90 Elbow TB M-239 Oper Deck 6 D5/E5 - *1A' Train 6' Steam Une from Horiz. 6.75 6-7, B-C 900 Elbow to 90° Elbow TB M-239 Oper Deck 6 D5/E5 - 'lA' Train 6" Steam Une from .9-9. Angle 250 6 900 Elbow to Moisture Sep/Reheater Al TS M-239 Mez. /- *1A Train 6" Steam Une from j8 Steam Supply Une to Oper. Deck Penetration _

6 Vert. 5.52 4-3, B-C

INTERNAL P1OODING - Initiating Events Analysis for Turbine Building Floods p. A3-11 I.

Table A3-1: High-Pressure Main Steam Piping (cont.)

BLDG Dwg. No. Bullding Level Nom. Drawing CoordinateslDescription Detail HorkzJVerL Pipe Length Quad #,Letter (floor Dia. (In) Section /Angle (Linear FT) plan quads)

TB M-239 Oper Deck 6 D2/E2 - "lA" Train 6" Steam Line from "6-6' Horiz 2.00 4-3, B-C Oper. Deck Penetration to 900 Elbow TB M-239 Oper Deck 6 D2/E2 - "1A" Train 6" Steam Une from "6-6" Horiz. 7.65 4-3, B-C 900 Elbow Thru Valvs (MS200A2, MS201A2) to 900 Elbow TB M-239 Oper Deck 6 D2/E2 - "IA" Train 6" Steam Une from "6-6" Vert. 8.34 4-3, B-C 900 Elbow to 900 Elbow TB M-239 Oper Deck 6 D2/E2 - "1A Trah 6" Steam Une from "6-6" Horiz. 11.50 4-3, B-C 900 Elbow Thru Orifice to 900 Elbow TB M-239 Oper Deck 6 D2/E2- "1A' Train 6" Steam Une from Horiz. 6.75 4-3, B-C 900 Elbow to 900 Elbow TB M-239 Oper Deck 6 D21E2 - "lA* Train 6" Steam Une from "6-6" Angle 2.50 4-3, B-C 90 Elbow to Moisture Sep/Reheater A2 Llnear FT o n Oper. Deck 154.46

_ULnear FT on Mez. Level 48.67 Total Length (Linear 0r 20303 MINw~ tEADERW Qli QUA ENE(0?IE i g LLg iSgE TB M-985-2 Mez. 20 B6 - Equalizing Une Between Main An Horiz. 42.25 7-8, F-E

_ andNB' Steam Headers

= Linear FT on Oper. Deck 42 25

YXEl lTotal Len (Linear FT) 42.25 TB M-239, - Mez. 16 B4/C4 - "1A" Train 16" Steam Une 119-19" Horkz. 4.67 5-6,E 240 From 16"xB" Reducer to Low Pressure Turbine TB M-239, - Mez. 16 B3/C3-M1B- Train 16" Steam Une "19-19" Horiz. 4.67 4-5,E 240 From 16"x8" Reducer to Low Pressure

_ _Turbine C C C

C C ET INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-121 Table A3-1: High-Pressure Main Steam Piping (cont.)

BLDG Dwg. No. Building Level Norn. Drawing Coordinates/Descriptlon Detail I HorizNert. Pipe Length Quad #,Letter (floor I I Dia. (ln) I I Section l /Anale (Linear FT) plan auads)

Linear FT on Oper. Deck 0.00 Linear FT on Mez. Level - 9.34 Linear FT on Basement Level 0.00 Total Length (Linear FT) 9.34 71,

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-13 Table A3-2: Extraction Steam Piping BLDG I Dwg. No. I Building Level Nom. l Drawing CoordInates/Description Detail HorizJVerL Pipe Length Quad #,Letter (floor I I I . fin% I I S §nn I /Anl. I l inenr PT I niun nonds I M-1z58 Mez. DI 1 - oieea steam Piping Fromm urIMne Shell Insulation to 900 Elbow van. o.Z/ f O-(, u TB M-1258 Mez. 12 B1 1/A10 - Bleed Steam Piping From 900 Horiz. 18.98 6-7, D-E Elbow to 450 Declined Angle Bend TB M-1258 Mez 12 A10 - Bleed Steam Piping From 450 Angle 3.67 6-7, E Declined Angle Bend to 16x12 90° Reducing Elbow .

TB M-1258 Mez. 16 A10/E3 - Bleed Steam Piping From Horiz. 71.09 7-4, E 16 x-i2 90 Reducing Elbow to 900 Elbow .

TB M-1258 Mez. 16 E3/D2 - Bleed Steam Piping Frcrn 900 Horiz. 23.80 4, E-F Elbow to 16'X10 900 Reducing Elbow TB M-1258 Mez. 12 Cl1 - Bleed Steam Piping From Turbine Vert. 2.83 6-7, D Shell Insulation to 900 Elbow TB M-1258 Mez. 12 C11/C10 - Bleed Steam Piping From Horiz. 8.04 6-7, E-D 900 Elbow to 90° Elbow TB M-1258 Mez. 12 C10 - Bleed Steam Piping From 900 Vert. 4.58 6-7, E-D Elbow to 900 Elbow TB M-1258 Mez. 12 C10 - Bleed Steam Piping From 900 Horiz. 7.50 6-7, E-D Elbow to 450 Declined Angle Bend TB M-1258 Mez. 12 C9 - Bleed Steam Piping From 450 Angle 3.67 6-7, E-D Declined Angle Bend to 450 Angle Bend TB M-1258 Mez. 12 C9 - Bleed Steam Piping Fron 45° Horiz. 4.85 6-7, E Angie Bend to 45° Angle Bend TB M-1258 Mez. 12 C9 - Bleed Steam Piping From 450 Horiz. 3.54 6-7, E Angle Bend to 450 16'x122 Lateral Reducer TB M-1258 Mez. 10 D/El - Bleed Steam Piping From Horiz. 6.96 4, E-F 16"X16"X10 Tee to 12 x10 900 Reducing Elbow C Cs C

C C C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-14 I

Table A3-2: Extraction Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom. Drawing Coordinates/Descriptlon Detail HorlziVert. Pipe Length Quad #,Letter (floor Dia (in) Section /Angle (Linear FT) plan quads)

TB M-1258 Mez. 12 El -Bleed Steam Piping From 12"x10" Vert 2.00 4, E-F 900 Reducing Elbow Top of FD WTR HTR 15A TB M-1258 Mez. 10 D2 - Bleed Steam Piping From Haiz 6.69 4-F 16"X16"X10 Tee to 12x10" 90° Reducing EFbow TB M-1258 Mez. 12 i Bleed Stean Piping Frorn 12"x10" Vert. 2.00 4-F 900 Reducing Elbow Top of FD WTR HTR 15B

. Linear FT on Oper. Deck 0.00 Linear FT on Mez. Level 176.46 Linear FT on Basement Level 0.00 Total Length (Linear FT)  % 176.46

l rINTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-151 Table A3-3: Lower-Pressure Steam Piping BLDG l Dwg. No. Building Level Nom. Drawing Coordinates/Description Detail HorlzJVert Pipe Length Quacd #,Letter (floor I I I Me lIn1 I I Itin IA..1- I II I---- = I nln I.

M-Z4Z Mez. E7 - Reneat steam Froam WU xiz.

Crossunder Piping to 90 Elbow TB M-242, - Mez 16 E7 - Reheat Steam From 90° Elbow to "0-C" Vert. 20.73 6-7, D-E 423 900 Elbow TB M-242, - Mez 16 E7 - Reheat Steam From 900 Elbowh Horiz. 5.66 6-7, D-E 423 450 Angle Bend Into 24" Reheat Steam r Header TB M-242 Mez. 16 F8 - Reheat Steam From 30" Houiz. 1.75 6-7, D Crossunder Piping to 900 Elbow TB M-242, - Mez. 16 F8 - Reheat Steam From 90 Elbow to 0C-C Vert. 20.73 6-7, D 423 24"x16" 900 Reducing Elbow TB M-242 Mez. 24 F8 - Reheat Steam From 24"x16" 90° Honz 28.50 6-7, D-E Reducing Elbow to S0e Elbow TB M-242 Mez. 24 D8 - Reheat Steam From 900 Elbow to Honiz. 14.33 6-7, E 900 Elbow TB M-242 Mez. 24 D9 - Reheat Steam From 900 Elbow to Honz. 6.00 7, E 90 Elbow TB M-242, - Mez. 24 D9/C9 - Reheat Steam From 900 Elbow "C-C Vert. 14.21 7, E-F 423 to 900 Elbow TB M-242, - Mez. 24 D9/C9 - Reheat Steam From 900 Elbow "C-C" Horiz. 14.92 7, E-F 423 to 24"x16" 900 Reducing Elbow For FD WTR HTR 14B TB M-242 Mez. 16 C9 - Reheat Steam from 24"x24"x16" Horiz. 9.17 6-7, E-F Tee to 900 Elbow TB M-242, - Mez. 16 C8 - Reheat Steam from 90 Elbow To C-C" Vert. 2.00 6-7, E-F 423 FD WTR HTR 14A TB M-242 Mez. 16 B9 - Reheat Steam from 24"x16" 90Hoz. 9.17 6-7, F Reducing Elbow to 90° Elbow For FD WTR HTR 14B C C C

C C C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-16 Table A3-3: Lower-Pressure Steam Piping (cont.)

BLDG Dwg. No. Building Level Norm Drawing Coordinates/Descriptlon Detail HoriziVerL Pipe Length Quad #,Letter (floor I Dla (In) Section /Angle (Linear FT) plan quads)

TB M-242, - Mez. 16 B8 - Reheat Steam from 90 Elbow To "C-C" Angle 2.15 6-7, F 423 . D-WTR HTR 14B TB M242 Me 10 Heatin Sta Ebn = Linear FT Oper. Deck 0.00 7 l 8<Llnear FT on Mez.Level 151.07 iTotal -) Length (Linear 151.07 TB M-242 Mez. 10 C98- Healing Steam from 245 Reheat Co Vert. 7.10 7, E-F 423 Steamrnine Tee to 90°Elbow TB1 l M-242 lMez 10 lC9 -Heating SteamnFrorn90Elbow to ll Horiz. l4.00 7E-F 7l ll l 1 90° Elbow TB; l M-242 lMez l 10 lC9 -Heaffng SteamnFrorn90Elbow to ll Horiz. l6.04 7E-F 7l 450 Bend TB lM-242 Mez. 10 C8 -Heating StemrnFrorn4!5*Bend to l Horiz. 9.90 7-6, F 1 1 l 1 45°90Bend-TB M-242 Mez, 10 88- Heating Steam From 450 Bend to Horiz. 47.00 7-5, F 90 Declined Elbow TB M-242, - Mez. 10 B4-Heating Steam From 900 Declined "C-C" Angle 8.44 5, F-G 423 Elbow to 900 Elbow TB M-242 Mez. 10 B4-Heating Steam From 90 Elbow to Horiz. 56.50 5-3, F-G l60 Bend TB M-242 Mez. 10 l1 - Heating Steam From 600 Bend to Horiz. 9.50 3, F-G 900 Elbow TB M-242, - Mez. 10 Al - Healing Steam From 90 Elbow to l J-J" Vert. 11.02 3,G 423 Mez. Floor Penetration _______

TB M-242, - Basement 10 Al - Heating Steam From Mez. Floor l .J-J l Vert. 4.81 3, G 423 Penetration to 900 Elbow ______l TB *M-242, - Basement 10 Al - Heating Steam From 90 Elbow to l JJ I Horliz. 20.25 3-2, G 423 90° Elbow TB M-423 Basement 10 H4- Heating Steam From 90 Elbow to Horiz. 10.00 3-2, G-GG Aux Bid Wall Pen. I III Linear FT on Oper. Deck 0.00

nitatin EvntsAnaysi fo TubineBuidin Flodsp.

INTENALFLODIN 3-1 INTERNAL F1,00DING - Initiating Events Analysis for Turbine Building Floods -P. A3-17 I I

Table A3-3: Lower-Pressure Steam Piping (cont.)

Nom. Drawing Coordinates/Descriptlon l Ddal Die. (in) l Section Ar.-1U1 - Renea stearn Crossunder Piping (to 30, XK- *A MSRs - Front Pipe) from HP 101-33 Turbine to 90° Elbow TB XK-101- Mez 30 Reheat Steam Crossunder Piping (to Vert. 10.56 6-7, C-D 30, XK- 'A' MSRs - Rear Pipe) from HP Turbine 101-33 to 90 Elbow TB XK-101- Mez. 30 Reheat Steam Crossunder Piping (to Horiz. 8.86 6-7, C-D 30, XK- 'A' MSRs - Rear Pipe) from 90° Elbow 101-33 to90° Elbow TB XK-101- Mez. 30 Reheat Steam Crossunder Piping (to Horiz 8.54 6-7, C-D 30, XK- 'A MSRs - Rear Pipe) from 900 Elbow 101-33 to 42" Crossunder Piping to "A" MSRs TB XK-101- Mez. 30,42 Reheat Steam Crossunder Piping (to Horiz. 25.00 7-5, C-B 30, XK- "A' MSRs - Header) from 30" 900 Elbow 101-33 Thru 42"x30" Reducer to 42" 90° Elbow TB XK-101- Mez 30,42 Reheat Steam Crossunder Piping (to Horiz. 53.38 7-5, C-B 30, XK- WA"MSRs - Header) from 42" 900 Elbow 101-33 Thru 42"x30" Reducer to 30" 900 Elbow TB XK-101- Mez. 30 Reheat Steam Piping to MSR *1A" from Vert. 10.56 5, B-C 30, XK- 42" Crossunder Header to Oper. Deck 101-33 Floor Penetration TB XK-101- Oper Deck 30 Reheat Steam Piping to MSR "A from Vert. 4.75 5, B-C 30, XK- Oper. Deck Floor Penetration to 30" 90° 101-33 Elbow C C C

4: C .C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods . .318 Table A3-3: Lower-Pressure Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom. Drawing CoordinatesiDescriptlon Detail HorlziVerL Plpe Length Quad #,Letter (floor Dia. (In) Section /Angle (Linear FT) plan quads)

TB XK-101- Oper Deck 30 Reheat Steam Piping from 30 90°0 Horiz 11.75 5, B-C 30, XK- Elbowto MSR *1A!

101-33 TB XK-101- Mez. 30 Reheat Steam Piping to MSR '2A7 from VerL 10.56 5, B-C 30, XK- 42" Crossunder Header to Oper. Deck 101-33 Floor Penetration TB XK-101- Oper Deck 30 Reheat Steam Piping to MSR "2A" from VertL 4.75 5, B-C 30, XK- Oper. Deck Floor Penetration to 30" 90° 101-33 Elbow TB XK-101- Oper Deck 30 Reheat Steam Piping from 30" 90 Horiz. 11.75 5, B-C 30, XK- Elbow to MSR *2A" 101-33 TB XK-101- Mez. 30 Reheat Steam Crossunder Piping (to Vert. 10.56 6-7, D-E 30, XK- "B"MSRs - Front Pipe) from HP 101-33 Turbine to 90° Elbow TB XK-101- Mez. 30 Reheat Steam Crossunder Piping (to VerL. 10.56 6-7, D-E 30, XK- "B"MSRs - Rear Pipe) from HP Turbine 101-33 to 900 Elbow TB XK-101- Mez 30 Reheat Steam Crossunder Piping (to Horz. 8.86 6-7, D-E 30, XK- -B" MSRs - Rear Pipe) from 90 Elbow 101-33 to 90° Elbow TB XK-101- Mez. 30 Reheat Steam Crossunder Piping (to Horiz. 8.54 6-7, D-E 30, XK- "B"MSRs - Rear Pipe) from 900 Elbow 1-01-33 to 42" Crossunder Piping to "B"MSRs TB XK-101- Mez 30,42 Reheat Steam Crossunder Piping (to Horiz. 25.00 7-5, E-F

-30, XK- 'A" MSRs - Header) from 30" 900 Elbow 101-33 Thnr 42"x30 Reducer to 42" 900 Elbow

INTER.NAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3 19 Table A3-3: Lower-Pressure Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom. Drawing Coordinates/Description Detail HoriziVert. Pipe Length Ouad #,Letter (floor I Dia. (in) Section /Angle (Linear Fr) plan quads)

TB XK-101- Mez. 30,42 Reheat Steam Crossunder Piping (to Horiz. 53.38 7-5, E-F 30, XK- WBMSRs - Header) from 42" 90° Elbow 101-33 Thnu 42",30" Reducer to 30" 900 Elbow TB XK-101- Mez. 30 Reheat Steam Piping to MSR "18" from Vert. 10.56 5, E-F 30, XK- 42" Crossunder Header to Oper. Deck 101-33 Floor Penetration TB XK-101- Oper. Deck 30 Reheat Steam Piping to MSR "1' from Vert. 4.75 5, E-F 30, XK- Oper. Deck Floor Penetration to 30" 90g 101-33 Elbow TB XK-101- Oper. Deck 30 Reheat Steam Piping from 30" 960 Horiz. 11.75 5, E-F 30, XK- Elbow to MSR " Bo 101-33 TB XK-101- Mez. 30 Reheat Steam Piping to MSR '2B" from Vert. 10.56 5, E-F 30, XK- 42" Crossunder Header to Oper. Deck 101-33 Floor Penetration TB XK-101- Oper. Deck 30 Reheat Steam Piping to MSR "2B" from Vert. 4.75 5, E-F 30, XK- Oper. Deck Floor Penetration to 30" 90° 101-33 Elbow _ _-_ _ _ __

TB XK-101- ' Oper. Deck 30 Reheat Steam Piping from 30" 90' Horiz. 11.75 5, E-F 30, XK- Elbow to MSR "2B" 1 Ml -_-'3 C, C C

C C C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods

p. A3-20 Table A3-3: Lower-Pressure Steam Piping (cont.)

BLDG DW9. No. Building Level Nom. Drawing Coordinates/Descriptlon Detail HorizVert. Pipe Length Quad #,Letter (floor Dia. (In) Section /Angle (Linear FT) plan quads) 101-33 TB X-K-101- Oper. Deck 30 Steam Crossover Piping from 900 Elbow Honz. 27.50 6-5, B-D 30, X-K- to Bend into LP Turbine #1 101-33 TB X-K-101- Oper. Deck 30 Steam Crossover Piping from Bend Into Vert. 3.50 6-5, D 30, X-K- LP Turbine #1 to Common Inlet 101-33 TB X-K-101- Oper. Deck 30 Steam Crossover Piping from MSR 1B Vert. 13.42 6, E-F 30, X-K- to 90° Elbow 101-33 TB X-K-101- Oper. Deck 30 Steam Crossover Piping from 90° Elbow Honz. 12.00 6-5, E-F 30, X-K- to 900 Elbow 101-33 _ _

TB X-K-101- Oper. Deck 30 Steam Crossover Piping from 90° Elbow Horiz 27.50 6-5, E-D

30. X-K- to Bend into LP Turbine #1 101 -33 _ _ _ _ _ _

TB X-K-101- Oper. Deck 30 Steam Crossover Piping from Bend Into Vert. 3.50 6-5, D 30, X-K- LP Turbine #1 to Common Inlet'

. 101-33 TB X-K-101- Oper. Deck 30 Steam Crossover Piping from MSR 2A Vert. 13.42 4, B-C 30, X-K- to 900 Elbow 101-33 .

TB X-k-101- Oper. Deck 30 Steam Crossover Piping from 900 Elbow floriz. 12.00 4-5, B-C 300i ,X-3K- 0to 90° Eibow TB X-K-101- Oper. Deck 30 Steam Crossover Piping from 90° Elbow Horiz. 27.50 4-5, B-D 30, X-K- to Bend Into LP Turbine #2 101-33 = -_ _ _ _

TB X-K-101- Oper, Deck_, ,.30 Steam Crossover Piping from Bend into VerL 3.50 4-5, D 30, X-K- LP Turbine #2 to Comrmon Inlet 101-33 --

_CEM

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-21 Table A3-3: Lower-Pressure Steam Piping (cont.)

BLDG Dwg. No. Building Level Nom. Drawing CoordinateeiDescriptlon Detail HoriziVert Pipe Length Quad #,Letter (floor Dla. (In) Section /Angle (Linear Fr) plan quads)

TB X-K-101- Oper. Deck 30 Steam Crossover Piping from MSR 2B Vert. 13.42 4, E-F 30, X-K- to 900 Elbow 101-33 _ _

TB X-K-101- Oper. Deck 30 Steam Crossover Piping from 900 Elbow Horlz. 12.00 4-5, E-F 30, X-K- to 90° Elbow 101-33 TB X-K-101- Oper. Deck 30 Steam Crossover Piping from 900 Elbow Horiz. 27.50 4-5, E-D 30, X-K- to Bend into LP Turbine #2 101-33 TB X-K-101- Oper. Deck 30 Steam Crossover Piping from Bend Into Vert. 3.50 4-5, D

30. X-K- LP Turbine #2 to Common Inlet Linear FT on Oper. Deck 225.68 Linear FT on Mez. Level 0.00 Linear FT on Basement Level 0.00 Total Length (Linear FT) 225.68 .

C C. C

C C, INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-22 Table A3-4: Piping Upstream of 15 Feedwater Heaters BLDG l Dwg. No. Building Level iNom. Drawing Coordinates/Description Detail HorlzJVerL Pipe Length Quad #,Letter (floor I Die (in) I sectlon /Anale (Linear FT) Plan quad!)

TB M-245 Basement 12 B1O inch bypass header centerline D-D, E-E Horiz. 16.75 7-7, F-F of valve C20-1 to elbow that angles down to main 20-inch header TB M-245 Basement 12 B10 inch bypass header centerline D-D, E-E Houlz. 5.23 7-7, F-F of valve C20-1 angling down to main 20-Inch header TB M-245 Mezzanine 16 Bl0 - Hoeizontal distance from outlet of D-D, E-E Hodz. 2.00 7-7, F-F heater 14A to centerline of vertical pipe

. down TB M-246 Basement 16 C4 inch header piping down from D-D Vert. 3.67 7-7, F-F outlet of heater 14A to mezzanine floor level TB M-246 Mezzanine 16 C4 Inch header piping down from D-D Vert. 5.50 7-7, F-F mezzanine floor level to centerline of header TB M-245 Basement 14 Bl0 - Centerline of vertical pipe down D-D, E-E Hortz. 9.50 7-7. F-F from heater 14A outlet to inlet of main 20-Inch header TB M-245 Mezzanine 16 B10 - Horizontal distance from outlet of D-D, E-E Horiz. 2.00 7-7, F-F heater 14B to centerline of vertical pipe down TB M-246 Mezzanine 16 C4 inch header piping down from D-D Vert. 3.67 7-7, F-F outlet of heater 14B to mezzanine floor level TB M-246 Basement 16 C4 Inch header piping down from D-D Vert. 5.50 7-7, F-F mezzanine floor level to centerline of header TB M-245 Basement 14 Bl0 - Centerline of vertical pipe down D-D, E-E Hodz. 9.50 7-7, F-F from heater 14B outlet to 90 degree elbow to the east TB M-245 Basement 14 B10 - Header pipe east from centeline D-D, E-E Horiz. 12.25 7-7, F-F of C15-2 to 20-inch header -

TB M-245 Basement 2 B10 - Header pipe east from reducer D-D, E-E Horiz. 8.25 7-7, F-F east to centerline of main 20-inch header south

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-23 Table A3-4: Piping Upstream of 15 Feedwater Heaters (cont.)

BLDG Dwg. No. Building Level Noam Drawing Coordinates/Descriptlon Detail HorlzjVert. Pipe Length Quad #,Letter (floor Dia. (in) Section /Angle (Linear FT) plan quads)

TB M-245 Basement 12 Bl0 Inch bypass header from D-D, E-E Hofz. 8.00 7-7, F-F centerline of valve C9-1 south to beginning of pipe bend _

TB M-245 Basement 12 Bl0 inch bypass header the D-D, E-E Horlz. 7.78 7-7, F-F beginning of pipe bend angling to the 14-inch header TB M-245 Basement 20 B10 inch header south toward E-E Horiz. 36.50 7-6, F-F

_ _feedwater pumps to reducer TB M-246 Basement 16 D6 inch header piping down from F-F Vert. 6.25 6-6, F-F main header to main feedwater pump 1A Inlet _ _.

TB M-246 Basement 16 B8 inch pipe west to main F-F Horiz. 16.92 6-6, F-F feedwater pump lA suction TB M-246 Basement 16 B8 inch pipe south to main F-F Horiz. 4.00 6-6, F-F feedwater pump IA suction TB M-246 Basement 16 D6 inch header piping down Into F-F Vert. 2.00 6-6, F-F main feedwater pump 1A inlet TB M-245 Basement 16 B8- 16-inch header south from reducer G-G Horiz. 39.00 6-4, F-F toward main feedwater pump 1B to 90 degree elbow down TB M-246 Basement 16 E12 inch header piping down from G-G Vert. 12.50 4-4, F-F main header TB M-245 Basement 16 B6 Inch header south from vertical G-G Horkz. 10.00 4-4, F-F pipe down to 90 degree elbow up TB M-246 Basement 16 ElI Inch header piping up from G-G Vert. 6.25 4-4, F-F main header TB M-245 Basement 16 B5 inch pipe west to main H-H Horiz. 18.92 4-4, F-F feedwater pump 1B suction TB M-246 Basement 16 D6 inch header piping down Into H-H Vert. 2.00 4-4, F-F main feedwater pump 1B Inlet TB M-252 Basement 12 B10 - Horizontal distance from the Q-Q Horlz. 1.93 7-7, F-F discharge of heater drain pump IA east to 90 degree elbow up C C C

C r C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-24 Table A3-4: Piping Upstream of 15 Feedwater Heaters (cont.)

BLDG Dwg. No. Building Level Hr. Drawing Coordinates/Desoriptlon Detail iHorlziVert Pipe Length Quad #,Letter (floor Dla (in) Section /Angle (Linear FT) plan quads)

TB M-252 Basement 12 B10 - Horizontal distance from the 0-0 Horiz. 1.93 7-7, F-F discharge of heater drain pump 1B east to 90 degree elbow up TB M-253 Basement 12 Bi0 - Vertical distance from centerline 0-0 Vert. 7.15 7-7. F-F of heater drain pump IA discharge to

_ 14-inch line TB M-253 Basement 12 BIG - Vertical distance from centedine 0-0 Vert. 7.15 7-7, F-F of heater drain pump 1Bdischarge to 14-nch line TB M-252 Basement 14 Bi0 - Horizontal distance from the 0-0 Horiz. 19.75 7-6, F-E.

discharge of heater drain pump 1A south to 90 degree elbow turning to the east TB M-252 Basement 14 B10 - Horizontal distance from 0-0 Horiz. 5.52 6-6, F-F centerline of 14-inch pump discharge header east to 90 degree elbow up TB M-253 Basement 14 BI0 - Vertical distance from centerline 0-0 Vert. 6.19 6-6, F-F of 14-inch pump header to centerline of 20-inch feedwater header TB M-252 Basement 20 Bi0 - Horizontal distance from the 0-0 Horiz. 13.46 7-7, F-F centerline of heater drain tank to the discharge of heater drain pump 1A Up TB M-252 Basement 20 B10 - Horizontal distance from the 0-0 Horiz. 13.46 8-6, F-F' centerline of heater drain tank to the

._ discharge of heater drain pump 1 B TB M-253 Basement 20 B10 - Vertical distance from heater 0-0 Vert. 4.00 7-7, F-F drain tank outlet to centerline of heater drain pump IA Inlet TB M-253 Basement 20 BIO - Vertical distance from heater Q-Q Vert. 4.UU 7-7, tF-t drain tank outlet to centerline of heater drain pump 158 inlet

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-25 Table A3-4: Piping Upstream of 15 Feedwater Heaters (cont.)

BLDG Dwg. No. Building Level Nom. Drawing Coordinates/Descriptlon Detail HorizJVerL Pipe Length Quad #,Letter (floor Dia. (In) Section /Angle (Linear Fr) plan quads)

TB M-249 Basement 16 B8 - Main feedwater pump 1A outlet up E-E Vert. 9.25 6-6, F-F to 90 degree elbow to the east TB M-247 Basement 16 CS - Horizontal distance from centerline E-E Houiz. 4.00 6-6, F-F pump 1A discharge east to 90 degree elbow to the north TB M-247 Basement 16 C5 - Horizontal distance north on pump E-E Horxz. 9.50 6-6, F-F

__Adischarge piping TB M-247 Basement 16 -C6 - Horizontal distance west on pump E-E Horiz. 12.25 6-6, F-F

__A discharge piping TB M-247 Basement 16 C6 - Horizontal distance south on pump E-E Horiz. 19.50 6-5, F-F 1A discharge piping TB M-247 Basement 16 C6 - Horizontal distance east on pump E-E Horiz. 16.00 5-5, F-F 1A discharge piping to 90 degree elbow angling down to the south TB M-247 Basement 16 D5 - Pump 1A discharge piping south E-E Horiz. 42.88 5-4, F-F toward 15 feedwater heaters TB M-247 Basement 16 D2 - Pump 1A discharge piping angling G-G Horiz. 6.36 4-4, F-F 45 degrees to 22-inch header TB M-249 Basement 16 B6 - Main feedwater pump 1B outlet up F-F Vert. 6.50 4-4, F-F to 90 degree elbow to the west TB M-247 Basement 16 C3 - Horizontal distance from centerline F-F Horiz. 8.75 4-4, F-G pump 1B discharge west to 90 degree elbow angling up and to the north TB M-249 Basement 16 B6 - Horizontal distance for the pipe F-F Horiz. 5.50 4-4, G-G angling up from the 90 degree elbow to the centerline of valve F2-2 TB M-247 Basement 16 C3 - Horizontal run of Pump 1B F-F Horiz. 18.53 4-5, G-G discharge piping north through valve F2-2 4 TB M-247 Basement 16 C4 - Horizontal run of Pump 1B F-F Horiz. 6.00 5-5, G-G discharge piping downstream of valve F2-2 to the east  ;

TB M-247 Basement 16 C4 - Horizontal run of Pump 1B F-F Horiz. 7.71 5-5, G-G discharge piping north C C

C C C INTERN4AL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-26 Table A3-4: Piping Upstream of 15 Feedwater Heaters (cont.)

BLDG Dwg. No. Building Level Nom. Drawing Coordinates/Description Detail HorlzJVerL Pipe Length Quad #,Letter (floor I Dia. in) Section /Angle (Linear FT) plan quads)

TB M-247 Basement 16 C4 - Horizontal run of Pump 1 E-E Hodz. 13.50 5-5, G-F discharge piping east to the elbow angling down TB M-249 Basement 16 86 - Main feedwater pump 1B discharge E-E Vert. 2.67 5-5, F-F piping down to header towards 15 feedwater heaters TB M-247 Basement 16 B6 - Main feedwater pump 1B discharge G-G Horiz. 41.60 5-4, F-F piping south towards 15 feedwater

___ _ __ lheaters up to the reducer TB M-247 Basement 22 B2 inch header for main feedwater G-G Horiz. 7.79 4-4 F-F-discharge piping from reducer to T ant

__ 15 heaters TB M-247 Basement 22 D2 - Centerline of T in 22-inch header G-G Horiz. 8.50 3-3, F-F east to reducing elbow TB M-247 Basement 16 D2 - From centedline of reducing elbow G-G Horiz. 17.04 3-3, F-F south through valve F3-1 to 90 degree reducing elbow to the west TB M-247 Basement 20 D2 - Straightline distance north through G-G Horiz. 5.00 3-3, F-F the two 90 degree elbows toward the 15A heater TB M-247 Basement 20 D2 inch piping north toward heater G-G Horiz 13.29 3-3, F-F

- _ 15A TB M-249 Basement 20 C12 - Vertical piping from 20-Inch H-H Vert. 5.92 3-3, F-Fi horizontal pipe up to mezzanine floor l toward heater 15A TB M-249 Basement 20 C12 - Vertical piping from mezzanine H-H Vert. 1.52 3-3, F-F floor to heater 1SA inlet TB M-249 Basement 14 F1 - Vertical 14-inch bypass piping from H-H Verl. 5.92 3-3, F-F 20-inch horizontal pipe up to mezzanine floor toward heater 15A ___ _ _

TB M-249 Mez. 14 COf - Vertical piping from mezzanine H-H VerL 1.50 3-3, F-F

-_foor to valve F11-1 nlet TB M-247 Basement 22 D2 - Centerline of T in 22-inch header G-G Horiz. 11.00 3-3, F-F west to reducing elbow

lINTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-27l Table A3-4: Piping Upstream of 15 Feedwater Heaters (cont.)

BLDG Dwg. No. Building Level Nom. Drawing Coordinates/Descriptlon Detail HorkzJVerL Pipe Length Quad #,Letter (floor Dia. (In) Section /Angle (Linear FT) plan quads)

TB M-247 Basement 16 C2 - From centerline of reducing elbow G-G Horiz. 17.04 3-3, F-F south through valve F3-2 to 90 degree reducing elbow to the east -

TB M-247 Basement 20 C2 - Stralghtine distance south through G-G Horiz. 5.00 3-3, F-F the two 90 degree elbows toward the 15B heater TB M-247 Basement 20 C2 inch piping north toward heater G-G Horkz. 13.29 3-3, F-F 15B TB M-249 Basement 20 C12 - Vertical piping from 20-inch H-H Vert. 5.92 3-3, F-F horizontal pipe up to mezzanine floor toward heater 15A TB M-249 Basement 20 C12 - Vertical piping from mezzanine H-H Vert. 1.52 3-3, F-F

._ floor to heater 15A Inlet TB M-249 Basement 14 Fl - Vertical 14-inch bypass piping from H-H Vert. 5.92 3-3. F-F 20-inch horizontal pipe up to mezzanine

__________ .. ___._ .. _____ ._ _ floor toward heater 15A . .... .. _ _ _

TB M-249 Mez. 14 CFt - Vertical piping from mezzanine H-H Vert. 1.50 3-3, F-F floor to valve Ft -1 inlet Linear FTon Basement 355.17 _

Linear FT on Mez. Level 3.00 Total Length (Linear FT) 358.17 I

r C C

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-29 Table A3-5: Piping Downstream of 15 Feedwater Heaters BLDG l Dwg. No. Building Level l Nom. l Drawing CoordInates/Descriptlon Detail l Horiz/Vertl Pipe Length l Quad #,Letter (floor

_ _Dl_.

(In)I Section /Angle (Linear FT plan quads

~PPNLCT TEf15HEAT AND FW4ANDF~T~iN TIJRINBUIL~tNG 7.

TB M-249 Mez. 18 B9 - Outlet of heater 15A up to 9O G-G, H-H Vert. 7.04 F-F, 3-3 degree elbow to the south TB M-247 Mez. 18 D2 - Horizontal piping running to the G-G, H-H Horiz. 9.63 F-F, 3-3 south from the outlet of heater 15A TB M-247 Mez. 18 D2 - Horizontal distance east through G-G, H-H Horiz. 5.00 F-F, 3-3 the 90 degree elbows on heater 15A outlet piping TB M-247 Mez. 18 D2 - Horizontal distance north to the G-G, H-H Horiz. 13.38 F-F, 3-3 reducing elbow to the west TB M-247 Mez. 22 D2 - Horizontal distance from the G-G, H-H Horliz. 28.16 F-G, 3-3 reducing elbow west toward turbine building wall to elbow up TB M-249 Mez. 14 Fl - Heater 15A bypass line from F11-1 G-G Vert. 12.54 F-F, 3-3 to 18-Inch pipe on heater outlet TB M-249 Mez. 18 B9 - Outlet of heater 15B up to 90 G4-, H-H Vert. 7.04 F-F, 3-3

_degree elbow to the south TB M-247 Mez. 18 C2 - Horizontal piping running to the G-G, H-H Horiz. 9.63 F-F, 3 south from the outlet of heater 15B TB M-247 Mez. 18 C2 - Horizontal distance east through G-G, H-H Horiz. 7.00 F-F, 3-3 the 90 degree elbows on heater 15B outlet piping TB M-247 Mez. 18 D2 - Horizontal distance north from G-G, H-H Hodz. 6.38 F-F, 3-3 elbow to intersection of 45 degree pipe to 22-inch header TB M-247 Mez. 18 D2 - HorizontaJ distance northwest of G-G, H-H Horiz. 9.90 F-F, 3-3 heater 51 B outlet piping angling at 45 degrees into 22-inch header TB M-249 Mez. 14 Al - Heater 15B bypass line from Fl 1-2 G-G Vert. 12.54 F-F, 3-3 to 90 degree elbow angling toward 22-inch header TB M-247 Mez. 14 C2 - Horizontal distance northwest of G-G, H-H Hoiz. 3.36 F-F, 3-3 heater 15B bypass piping angflng at 45 degrees Into 22-Inch header -

Cf C C\

C C7 C INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-30 Table A3-5: Piping Downstream of 15 Feedwater Heaters BLDG Dwg. No. Building Level Nom. Drawing Coordinates/Description Detail HorizVert. Pipe Length Quad #,Letter (floor I

I Di. (in) Section /Angle (Linear FT) plan quads)

TB M-249 Mez. 24 B8 - Vertical distance frorn centerline of G-G Vert. 7.95 G-G, 3-3 22-Inch header to operating deck.

Assumes that all pipe is 24 Inches from elbow on TB M-249 Oper. 24 B8 - Vertical distance from operating G4- Vert. 11.00 G-G, 3-3 deck to 90 degree elbow into autdliary building. Assumes that all pipe is 24 inches from elbow on TB M-247 Oper. 24 02 - Horizontal distance north from T in G-G Horiz. 4.42 G-G, 3-3 verical 24-inch header TB M-247 Oper. 24 A2 - Vertical distance from centeriine of A-A Vert. 10.25 G-G, 3-3 line 224 to 90 degree elbow angling out from wall -

TB M-247 Oper. 24 C2 - Horizontal distance of pipe angling G-G Horiz. 7.42 G-G, 3-4 out 45 degrees from wall to header TB M-247 Oper. 24 C2 - Horizontal distance north to elbow G-G Horiz. 41.92 G-G, 4-5 TB M-247 Oper. 24 B5 - Vertical distance from centerline of N/A Vert. 11.08 G-G, 5-5 header through valve V38-8 to the 90 degree elbow to the south TB M-247 Oper. 22 C2 - Horizontal distance from centerline G-G Horiz. 64.77 G-G, 5-3 of valve F38-8 south to elbow down TB M-249 Oper. 22 B8 - Vertical distance from centerine of N/A Vert. 19.42 G-, 3-3

_ _ header to centerline of valve V38-9 TB M-247 Oper. 22 02 - Horizontal distance from centerline G-G Horiz. 12.50 G-G, 3-3 of header pipe down to the reducing elbow turning west -

TB M-247 Oper. 22 C2 - Horizontal distance from centerline G-G HorlZ. 5.25 G-G, 3-3 of F38-9 Into header T downstream of

_ __ __ _ __ _ __ __ __ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _F 38-7 _ _ _ _ _ _ _ _

TB M-247 Oper. 22 C2 - Horizontal distance from centerfine G-G Horiz 4.00 G-G, 3-3 of F38-7 to auxdliary building wall Linear FT on Oper. Deck 192.03 Linear FT on Mez. Level 139.53

INTERNAL FLOODING - Initiating Events Analysis for Turbine Building Floods p. A3-31l Table A3-5: Piping Downstream of 15 Feedwater Heaters BLDG Dwg. No. l Building Level Nom. Drawing Coordinates/Descriptlon I Detail I HorilzVertl Pipe Length Quad #,Letter (floor l Dia. (in) } l Section l /Angle (Linear Fr) plan quads)

Total Length (Linear Fr) 331.56 1

-C C C'

Appendix A Initiating Events - Turbine Sump Alarm History Page I of 5

Turbine Building Sump Alarm History Prepared by: , 4 I 6e'A4..

W_Signature c.- Print Name 10 /it/of Date Reviewed by: 5efA 7 7,o F4raere r SMAXoML),

Signature Print Name

-/ - O..r Date Q4 Page 2 of 5

Turbine Building Sump Alarm History Annunciator 47033-P, MiscellaneousSump Level High, represents three sumps, the Screenhouse sump (SER point 1593), Turbine Building sump (SER point 1594), and the Waste Area sump (SER point 1595). In review of the Sequence Event Recorder (SER) output from January 2003 through April 2005 (28 months), Annunciator 47033-P actuated on 70 days. This represents an alarm on the average of once every 12 days. There are periods of close to 2 months without an alarm and short periods with daily alarms. Alarms are frequently less than 1 minute and clear when operator acknowledges the annunciator. Of the 156 annunciator activations, 103 were at power. The annunciator was active for 1100 minutes with three times greater than 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. The activations at-power average length of time was II minute but the three longer times account for 572 minutes. The average time, excluding the three long periods, is approximately 5 minutes per alarm. With the infrequency and length of time of the annunciator, the operators would respond in a timely fashion with concern if the alarm does not immediately clear.

Dae SER Date SER Date Point IN Out Note _____Point IN out Note 1594 0758 0758 17-May-0 1593 0819 0824 1594 0758 0800. 159 1859 1902 1594 0836 0836 18-May-03 1593 0039 00421 27-Feb-03 1594 0836 0837 15-Jun-03 1593 1129 11321 1594 0837 0840 1593 2243 2246 1594 0840 0840 07-Jul-03 1593 0102 0104 1594 0840 0845 . 1594 0953 0953 1593 1443 1443 22-Jul-03 1593 2244 2249 1593 1443 1443 23-Jul-03 1593 1309 131 _

20-Mar-03 1593 1443 1443 24-Jul-03 159 0549 0550 1593 1443 1443 27-Jul-03 1593 1819 1821 1593 1443 1443 29-Jul-03 1593 2100 2103 25-Mar-03 1594 0931 0932 30-Jul-03 1593 0521 0524 1594 0619 0619 1593 0512 0518_

26-Mar-03 1594 0619 0619 20-Sep-03 1593 1419 1423 1594 199 0619 1593 1729 1933 Firt during 25-Sep-03 1594 2113 2145 15-Apr03 .1593 1340 1343 Outage 1594 2145 2149 20-Apr-03 1593. 0903 1046 . _

02-Oct-03 1594 1340 1340 06-May-03 1594 2328 0102 Day change 03-Dec-03 1593 1141 1143 1594 0102 0154 04-Dec-03 1594 0220 0220 1594 0102 0102 08-Dec-03 1593 0136 0141 07-May-03 1694 0154 0154 1594 0326 0326 1594 0154 0154 09De-03 1594 0144 0144 1594 0204 0230 1593 0437 04411 _

1594 1633 2119 12-Dec-03 1593 0913 0916 1 08-May-03 1594 1239 1631 Day change 16-Dec-03 1593 2208 2212 :

10-May-03 10My-3 1Last 1594 1304 1311 Outage during 18-Dec-03 1593 0018 00211 Page 3 of 5

Date R

[Pointi ut I,Note 1 Point IN Out Note Qw, 22-DeC-03 1 16941 12411 12411 1594 0759 0759 25-Dec-03 115931 11361 11381i 1594 0946 0946 1594 0947 0947 1594 0947 0954 1593 1400 1402l 1594 0955 1007 28-Dec03 1593 1412 1415 1594 1043 1043 1593 1724 1726 1594 1043 1044 2 1593 1243 1246 1594 1107 .1107 29-Dec-03 1593 2246 2249 02-Mar-04 1594 1129 1132 31-Dec-03 1593 1609 1513 Cont. 1594 1152 11521 01-Jan-04 1593 0632 0635 1694 1154 1154 01-Jan-04 1593 0924 0926 1594 1156 1159 10-Jan-04 1593 1723 1727 1694 1304 1304 13-Jan-04 1593 0341 0344 1594 1311 1311 13-Jan-04 1593 0645 0649 1594 1311 1311 1593 2118 2121 14-Jun-04 1594 0833 0833 18-Jun-04 1593 0832 0832 15-Jan04 1593 1828 1832 First during 1593 1939 1943 Outage 1594 1230 1230 1593 0149 0153 1594 1230 1230 16-Aug-4 1594 1230 1230 16-Jan-04 1593 0533 0537 1593 1257 13001 1594 1230 1230 1593 1412 1415 1594 1230 1230 28-Jan-04 1593 1245 2349 20-Sep-04 1594 1346 1414

. 1593 2145 2148 1594 1333 16331 1593 0312 0318 01-Oct-04 1594 1655 1655 29-Jan-04 1594 2256 2301 1594 1718 17581 1594 2300 0038 Day change 20-Oct-04 First during 1594 0038 0039 Last during 1593 2237 2332 Outage 30-Jan-04 1594 0039 003 9 Outage 1594 1209 1210 1594 0039 0132 1594 1210 1213 1594 0303 0314 1594 1313 1314 1594 0359 0406 02-Nov-04 1594 1314 1314 31-Jan-04 1594 0508 051 1594 1314 1314 1593 0659 0603 1594 1314 1332 1594 0802 0810 1594 1314 13141 01-Feb-04 1593 1647 1650 1594 1332 1421 1593 0053 0057 04-Nov-04 1594 1405 1405 1593 0236 0240 06-Dc-04Lost during 06-Dec-04 1594 0537 0610 Outage 02-Feb-04 1593 0921 0924 1595 1043 1044 1593 1209 1213 1595 1044 1044 1593 1430 1435

. 1693 1550 1656 09-Feb-05 1595 1044 1044 1595 1049 1049 01-Mar-04 1594 1246 1251 1595 1117 1117.

02-Mar-04 1594 075 0753 .

_ 1595 1117 11171 1594 0759 0800 Page 4 of 5

Date SER if)_DatePoint IN Out Note First during 23-Feb-05 1593 2157 2158 Outage

. 1593 2157 2157, 10-Mar-05 1593 0827 0834 SERIES 13-Mar-05 1594 1101 1104 14-Mar-05 1594 0808 0808 17-Mar-05 1593 0925 0925 1593 1018 1018 22-Mar-05 1593 1739 1743 02-Apr-05 1594 1451 1457 03-Apr-05 1594 062 0650 04-Apr-05 1594 1225 1236 05-Apr-05 1594 0629 0644 06-Apr-05 1594 0616 0634 09-Apr-05 1594 1546 1547 1 1594 1550 1550 Page 5 of 5

Appendix A Initiating Events Attachment 2 - Circulation Water Expansion Joint Rupture Frequency

Circulation Water Expansion Joint Rupture Frequency Owner's Acceptance: nr.bi G ,SM'- Ti-qbt& & S tocA Signature Print Name Date

EXPANSION JOINT FAILURE RATES FOR THE KEWAUNEE PRA Final Report Prepared for Dominion Energy Kewaunee Power Station By Karl N. Fleming Bengt O.Y. Lydell In Cooperation with Maracor Software & Engineering, Inc.

Karl N. Fleming Consulting Services LLC A Callfomia Limited Liability Company 616 Sereno View Road Encinitas, CA 92024 United States of America November, 2005

Expansion Joint Failure Rates for Kewaunee PRA TABLE OF CONTENTS Section Title Page

1. INTRODUCTION .................... 3 1.1 Purmose .................... 3 1.2 Scope..............................................................................................................3 1.3 Obiectives .................... 3 1.4 Report Guide .................... 3

2. TECHNICAL APPROACH .................... 4 2.1 Overview ................... 4 2.2 Uncertainty Treatment .................... 4 2.3 Component Rupture Model .................... 4 2.4 Definition of Expansion Joint Failure Mode Cases ........................................... 7

3. CIRCULATING WATER SYSTEM EXPANSION JOINT FAILURE RATES ..... 8 3.1 Background ....................................................... 8 3.2 Revised Data Analysis ...................................................... s9 3.3 Revised Expansion Joint Failure Rate ....................................................... 9 3.4 Revised Conditional Probability of RuDture .................................................... 10 3.5 Results ...................................................... 11 3.6 Plan to Update EPRI Report of Reference 1`11 ................................................. 11

4. REFERENCES ....................................................... 19 TABLE OF FIGURES Title Page Figure 2-1 Flow Chart for Bayes' Estimates of System. Size, and Damage Mechanism Specific Pipe Failure Rates (A)and Rupture Frequencies () .6 Figure 3-1 Uncertainty Distribution Expansion Joint Rupture Frequency (>10,000 gpm) 17 LIST OF TABLES Title Page Table 3-1 CW-Expansion Joint Failure Rate & Rupture Frequency (Reproduced from Table A-35 from Reference 1i 1) .8 Table 3-2 Analysis of Events Involving Expansion Joint Failures .12 Table 3-3 Uncertainty Distribution Results for Expansion Joint Failure Rates.. 17 Karl N. Fleming Consulting Services LLC Page 2 of 21

Expansion Joint Failure Rates for Kewaunee PRA

1. INTRODUCTION 1.1 Purpose The purpose of this report is to document the derivation of failure rates for different failure modes for rubber expansion joints of the type used in LWR circulating water systems. This work was performed via Subcontract to Maracor Software Engineering, Inc. on behalf of Dominion Energy's Kewaunee Power Station. This report is prepared to be an integral part of the overall turbine building internal flooding initiating events analysis.

1.2 Scope The scope of work covered in this report includes:

• Development of failure rates and rupture frequencies for rubber expansion joints of the type used in the Kewaunee Circulating Water System

• Development of point estimates and probability uncertainty distributions for all parameters subject to data uncertainties 1.3 Objectives The objective is to perform a state of the art data analysis that is consistent with the applicable requirements of ASME PRA Standard Capability Category II for data analysis and initiating event frequency development. Consistent with this objective, the report is intended to provide a traceable basis for the calculations so that the results could be independently reproduced from the information provided.

1.4 Report Guide A major part of this report is devoted to the development of a set of failure rates and rupture frequencies for use in the intemal flooding initiating event development. The technical approach to developing these failure rates and rupture frequencies is summarized in Section 2. In Section 3 the failure rates for rubber expansion joints are developed for different expansion joint failure modes including leakage and ruptures with flow1 rates less than 2,000 gpm, ruptures with flow rates greater than 2,000 and ruptures with flow rates greater than 10,000 gpm. Section 4 lists the references used as inputs to the data development and methodology. Supporting details are provided in the Appendices.

Karl N. Fleming Consulting Services LLC Page 3 of 21

Expansion Joint Failure Rates for Kewaunee PRA

2. TECHNICAL APPROACH 2.1 Overview The model used to estimate piping component failure frequencies for the initiating event models in this calculation is the same as that used in a recent EPRI report on internal flooding initiating event frequencies (1], and similar to that used in recent NRC studies regarding loss of coolant accident (LOCA) initiating event frequencies [2] [3]. The source of pipe failure and exposure data used to quantify the failure rates used in these models is known as "PIPExp-2004" [4]. A summary of this database is provided in Appendix A 2.2 Uncertainty Treatment Uncertainties in these failure rates were quantified using a Bayes' methodology that was developed in the EPRI RI-ISI program [5] and approved by the NRC for use in applied RI-ISI evaluations [6]. An independent review of this pipe failure rate uncertainty treatment was performed to support the NRC Safety Evaluation and results of this favorable review are provided in Reference [7]. An earlier EPRI report [8] developed a set of pipe failure rates for use in the EPRI RI-ISI applications which was also approved and independently reviewed in References [6] and [7]. These earlier failure rate estimates were derived from a pipe failure database that had been developed in Reference [10]. During subsequent work in applying these estimates in applied RI-ISI evaluation, a significant number of data classification errors in the original data source [10] were identified and improved estimates of the exposure population became available. These factors, as discussed more fully in Reference [9], were the prime motivation to switch to the more comprehensive and validated uPIPExp-2004" database when Reference [1], was developed. The most recent NRC sponsored work on LOCA frequencies [3]

is also based in part on the "PIPExp-2004" database.

2.3 Component Rupture Model The model used for relating failure rates and rupture frequencies for piping components uses the following simple model that is widely used in piping reliability assessment and was used in recent updates of recommended Loss of Coolant Accident frequencies [6]. The failure modes included in the estimation of failure rates include leaks and ruptures and, in some cases, cracks may also be included depending on the application. The model is expressed in the following equation:

M M P Epi =Z'ikP"RIF (2.1) k=1 =1 Where:

= total rupture frequency of rupture size x for pipe size i in system j

= rupture frequency of rupture size x for pipe of size i in system j due to damage mechanism k

= failure rate of pipe of size i in system j due to damage Karl N. Flenming Consulting Services LLC Page 4 of 21

Expansion Joint Failure Rates for Kewaunee PRA mechanism k Pgk{RXF} = conditional probability of rupture size x given failure for pipe size i in system j and damage mechanism k M = Number of different damage mechanisms In general, a point estimate of the frequency of pipe failures, Ap, is given by the following expression:

nk

.i2j jk (2.2) fNu Tu fJk Where nok= the number of failures (cracks, wall thinning, leaks and ruptures) events for pipe size i in system j due to damage mechanism k T = the total time over which failure events were collected for pipe size i in system j Nij = the number of components that provided the observed pipe failures for size i in system j fij= the fraction of number of components of size i in system j that are susceptible to failure from damage mechanism k for conditional failure rates given susceptibility to damage mechanism k, 1 for unconditional failure rates Note that all failure modes that result in pipe repair are included in the failure rate and that all failures thus defined are regarded as precursors to rupture. Some events that have no evidence of leakage are screened out prior to the calculation of failure rates. The events counted as ruptures are based on a specific definition of rupture which is application specific.

For internal flooding applications, we seek unconditional failure rates and hence we can combine these equations under the condition: fqik =1 to obtain the following expression for the point estimate of the rupture frequency.

M M n p, =Zpou =ZA4*(klFJ= aJR.lF} (2.3)

__1 = k-I NUTU In the development of Bayes' uncertainty distributions for these parameters, prior distributions are developed for the parameters Alp and Plk(R IF) and these prior distributions are updated using the evidence from the failure and exposure data as in standard Bayes' updating. The exposure terms (denominator of the fractions on the right hand side of Equation (2.3) also have uncertainty as these terms must be estimated for the entire nuclear industry that provides the number of failures for the failure rate estimation. This uncertainty is treated in this process by adopting three hypotheses about the values of the exposure terms which requires three Bayes updates for each failure rate. The resulting posterior distributions for each parameter on the right hand side of Equation (2.3) are then combined using Monte Carlo sampling to obtain uncertainty distributions for the pipe rupture frequencies. A picture of this process is shown in Figure 2-1. This flow chart shows the full treatment of uncertainty needed for the RISI formulation in Equation (2.2). For the internal flooding formulation of Equation (2.3) the damage Karl N. Fleming Consulting Setvices LLC .Page 5 of 21

Expansion Joint Failure Rates for Kewaunee PRA mechanism susceptibility fractions ( fjk ) do not come into play. The specific way in which this flow chart is applied is discussed in Section 4 for each system and failure mode.

In Reference [1] rupture frequencies were developed for three rupture sizes that were selected to support internal flooding analysis. These sizes include water sprays with flood rates of up to 100 gpm, flooding with flood rates of 100 to 2000 gpm, and major flooding with flood rates greater than 2000 gpm. For the Kewaunee internal flooding models, a somewhat different rupture size model had to be developed as the criteria for producing the consequences of interest are based on specific rupture sizes that were determined to produce the assumed flooding consequences.

(hw)

Lowertoer Update :(1=.) es-Bi^ sErest Updat (* .Si). a Bayes' Posterior Weighting Operation 1r P(RI F), Prob.Rupture i nn.-....

m- -. Given Failure Figure 2-1 Flow Chart for Bayes' Estimates of System, Size, and Damage Mechanism Specific Pipe Failure Rates (A) and Rupture Frequencies (p)

Karl N. Fleming Consulting Services LLC Page 6 of 21

Expansion Joint Failure Rates for Kewaunee PRA 2.4 Definition of Expansion Joint Fallure Mode Cases Failure rates are developed in this report for the following cases.

1. Total failure rate for all failure modes involving leaks and ruptures

2. Rupture frequency for leaks and ruptures with flow rates less than or equal to 2,000 gpm

3. Rupture frequency for ruptures with flow rates greater than 2,000 gpm

4. Rupture frequency for ruptures with flow rates greater than 10,000 Note that Case 3 is inclusive of Case 4, i.e. Case 3 includes leak flows less than, equal to, and greater than 10,000 gpm.

Karl N. Fleming Consulting Services LLC Page 7 of 21

Expansion Joint Failure Rates for Kewaunee PR4

3. CIRCULATING WATER SYSTEM EXPANSION JOINT FAILURE RATES

3.1 Background

Failure rates and rupture frequencies for circulating water system expansion joints were developed in Reference [1] by the authors of this report. The results obtained in that analysis are summarized in Table 3-1 reproduced from Table A-35 in Reference [1]

below.

Table 3-1 CW-Expansion Joint Failure Rate & Rupture Frequency (Reproduced from Table A-35 from Reference [11])

Component & Failure Mode Unc ertainty Distriution [1/EXJ .YR Type Failure Mode Mean St Median se

_ Percentile Percentile CW Rubber EXJ Spray 1.11 E-04 4.13E-05 i 9.51 E-05 2.38E-04 CW-Rubber EXJ Major Flooding 1.49E-05 3.92E-06 1 .20E-05 3.61 E-05 The evidence used to develop these results consisted of the following:

The failure rate for sprays was based on 4 events involving LWR circulating water system rubber expansion joint failures that occurred at Comanche Peak (1 event),

LaSalle (1 event), and Catawba (2 events). The exposure term was estimated based on 2899 LWR reactor years of service data in the PIPExp database through 2004 and an estimate of 12 rubber expansion joints per LWR circulating water system.

The prior distribution used for the analysis was based on the failure rate developed in the Oconee PRA [19] whose mean value-is 2.5x104 per component year and a range factor of 100 was assumed.

To estimate the conditional probability of rupture for major flooding given failure, which was defined in Reference [1] to be a rupture with flooding in excess of 2,000 gpm, a larger population of expansion joint, failure events covering different systems and including an HTGR event at Ft. St. Vrain was developed. This population had a total of 35 events including the 4 events used in the above described failure rate calculation.

One of these ruptures was the LaSalle event considered in the failure rate calculation and the other 3 ruptures in this population occurred at Ft. St. Vrain, Beaver Valley, and Comanche Peak. Note that the information presented in Reference [1] did not identify the 3 rupture events other than the one at LaSalle. Also note that the events at Beaver Valley and Comanche Peak were not in the circulating water service system nor were 31 of the 35 events considered in this larger population of expansion joint failures. The approach of specializing the failure rate data to the system of interest and then using a Karl N. Fleming Consulting Services LLC Page 8 of 21

Expansion Joint Failure Rates for Kewaunee PRA larger sample size for the conditional rupture probability is consistent with the approach used in Reference [1] for all the piping system failure rates that were developed.

3.2 Revised Data Analysis In the current study, more information was collected on the events that were analyzed in Reference [1] and additional expansion joint events were identified. This was accomplished by augmenting the data queries from the PIPExp database that was used in Reference [1] by consulting additional sources including those of References [20J through [34]. In addition to the 35 events used in Reference [1], the revised analysis included an additional 7 events that are summarized in Table 3-2. As a result of this additional information, the 43 events in Table 3-2 are analyzed as follows by tabulating the assessments in the last column of Table 3-2.

Total No. Events 42 Events screened out due to non-leakage 6 Events involving leaks or ruptures in LWR Circ. water systems 5 Events Involving Leaks 27 Events involving Ruptures with leak flows < 2,000gpm 6 Events involving Ruptures with leak flows 2,000-1 0,000gpm 2 Events involving Ruptures> 10,000 gpm 1 Total Number of Faillures (leaks + ruptures) 36 3.3 Revised Expansion Joint Failure Rate Consistent with the methodology adopted in Reference [1], the failure rate for the circulating water system expansion joints is based on data from LWR circulating water systems only and does not include data from expansion joints in other systems. This approach is used to capture system specific factors that may impact the degradation mechanism responsible for failure and is expected to influence the likelihood of failure. In the Reference [1] analysis of expansion joint failure rates 4 events were classified as LWR Circulating water system failures and in the revised analysis this is increased to 5.

The assumptions regarding population exposure and the assumed prior distribution are unchanged from Reference [1] in this revised analysis: namely, that there are on the average of 12 circulating water expansion joints per reactor unit and that the reactor years of exposure data that was estimated in Reference [1] of 2899 reactor operating years is still valid. According to the methodology described in Section 2, uncertainty in this exposure term estimate was accounted for by admitting hypotheses that these estimates are 50% higher and 50% lower than this best estimate. The net result of the revised circulating water system failure counts are an increase in the assessed failure rate by a factor of 1.25. The resulting failure rate distribution is presented in Table 3-3 and this distribution was used for each of the different failure mode cases described below.

Karl N. Fleming Consulting Services LLC Page 9 of 21

Expansion Joint Failure Rates for Kewaunee PRA 3.4 Revised Conditional Probability of Rupture Consistent with the methodology adopted in Reference [1], the data set for the estimation of conditional rupture probabilities is expanded to include expansion joints in other systems because this parameter is viewed to be more a function of the properties of the component than is the case with the failure rate. System specific factors may influence the failure rate, but the conditional probability of rupture given failure is not expected to vary from system to system. The use of different data sets for the failure rate and the conditional probability of rupture given failure is consistent with the pipe failure data handling methodology that was developed for the EPRI RI-ISI evaluations in Reference [8] and approved by the NRC in Reference [7]. In that reference, data from different systems was pooled to support the conditional rupture probabilities, but such pooling was not performed for the failure rates. The motivation is to have a statistically significant sample size for each parameter. More discussion on this point can be found in Reference [9].

In the Reference [1 ] analysis there was one rupture case developed for ruptures with leak flows greater than 2,000 gpm. The evidence for that analysis was four rupture events in 35 failure events. Inthe updated analysis there is a total of 36 events involving leaks and ruptures. For for leak flows greater than 2,000 gpm, there are 3 events including 2 events with leak flows between 2,000 gpm and 10,000 gpm and 1 event with leak flow greater than 10,000. One of the events classified as rupture in the Reference

[1 ] analysis at Comanche Peak was determined by contacting Plant personnel (Reference [33]) to be a leak with a flow rate substantially less than 2,000 gpm which flooded a small room over a protracted period of time and hence in this revised analysis it was classified as a leak. The remaining 3 ruptures in Reference [1] remain so classified here. The most severe was the expansion joint rupture at Ft. St. Vrain which had a reported leak rate of 15,000 gpm.

So the evidence for the conditional probability of rupture is 3 events out of 36 failure events involving leak flows greater than 2,000 gpm, 2 events out of 36 failure events for leak flows between 2,000 gpm and 10,000 and 1 event out of 36 failure events involving leak flows greater than 10,000.

In the Reference [1] analysis a Beta distribution was used to characterize the uncertainty in the conditional probability of rupture. The prior distribution was assumed to be a flat prior indicating a non-informative state of knowledge. Given the current knowledge based on the Reference [1] results we now know that the vast majority of expansion joint failures are leaks and not major ruptures. Inthe revised analysis, the A and B parameters for the prior Beta distribution are set at 1 and 9, respectively consistent with an assumed mean conditional rupture probability of 0.1. This yields the following mean values of the Bayes' updated Beta Distribution.

Mean{Rupture>2,000gpm} Post=rion = (APrior +3) (1+3) =.087 Apgtcri +Bposterior (APrior +3 )+(BPrior +33) (1+3)+(9+33)

Karl N. Flenming Consulting Services LLC Page 10 of 21

Expansion Joint Failure Rates for Kewaunee PRA Mean{Rupture2,000tolO,OOOgpm}= APosteCon (Apr, or + 2) (1+2) =.065 fA1stedor +BPosteror (Apr for + !)+(B~r+35) (1+2)+(9+35)

Mean{Rupture > 10,OOOgpm)= APosternon = (Apr=or+1) (1+1) 044 APosteror + Bp0ostnr (Apr icr + 1) + (Bpr0o, + 35) (1+1) +(9+35) 3.5 Results The results using the methodology of Reference [1] and summarized in Section 2 were obtained using Crystal Ball and yielded the failure rates and rupture frequencies for the Circulating Water system rubber expansion joints given in Table 3-3. The frequency distribution from the Monte Carlo analysis for the expansion joint rupture frequency with leak flows greater than 10,000 gpm is shown in Figure 3-1.

3.6 Plan to Update EPRI Report of Reference [11 This revised analysis of Circulating Water System Expansion Joint failure rates and rupture frequencies will be included in an update of the EPRI internal flooding frequency report of Reference [1 ] which will be published in early 2006.

Karl N. Fleming Consulting Senvices LLC Page II of 21

Expansion Joint Failure Rates for Kewaunee PRA Table 3-2 Analysis of Events Involving Expansion Joint Failures Plant Date Data Event Description System Classification Source ANO 2 3/10/1992 Reference 30-in. expansion joint condensate pump. Expansion joint inspected and Condensate Leak

[23] found to have pinholes because of aging. Evidence of leakage (CND)

Beaver Valley 10/15/1990 Reference A SW expansionjoInt collapsed. Informationwas obtained directlyfrom Service Leak

[23] Beaver Valley (January 2000) that documented expansion Joint Water (SW) seepage amounting to little more than awet spot along an intermittent four foot long crack at the base of an arch. SW expansion joint was found deformed during routine operator rounds. An analysis of the expansion joint could not demonstrate its operability. The expansion joint failed due to the effect of water hammer/ column separation. This condition was attributed to the failure of the pump's discharge vacuum breaker to open following the pump's shutdown. Vacuum relief is required due to the design of the system. Evidence of leakage  :

Beaver Valley 1 12/18/1995 LER 1995- A 24-inch diameter, 2-foot long SW expansion joint ruptured because of SW Rupture >2,000 010 erosion of tube. Steel belt was corroded. The expansion joint was 10 gpm years old. The rupture was approximately 4.5 inches by 3 inches.

Approximately 40,000 gallons of water spilled in a very short time. The failure was attributed to erosion of the inner rubber wall which caused corrosion of the expansion joint belts. (Reference [28D.

Beaver Valley 2 8/11/1986 Reference 30-in. condensate pump suction expansion joint deformed and partially CND Screened out; no

[23] collapsed, noleakage. leak Beaver Valley 2 8/9/1991 Reference 30-in. suction expansion joint for condensate pump found to have CND Leak

[23] pinhole leak.

Brunswick 1 1/4/1990 Reference Rupture of screen wash pump expansion joint. Screen Rupture <2,OOOgpm

[23] Wash pump (SWP)

Byron 1 11/19/1991 Reference Condensate pump expansion joint cover found tom 180 degrees. CND Leak

[231 Evidence of leakage Calvert Cliffs 1 2/1/1996 Reference 24-in. suction side of condensate pump expansion joint - tube imploded CND Leak

[23] and core of Joint visible. Evidence of leakage Catawba-1 1/1/2001 IR 50- CW Expansion joint leakage Circ. Water Leak 413/2001-02 Systrem (CWS)

Catawba-2 1/1/2001 IR 50- CW Expansion joint leakage; contact with Catawba plant personnel CWS Screened out as a Karl N. Fleming Consulting Services LLC Page 12 of 21 C C C7

C C Expansion Joint Failure Rates for Kewaune( RA Plant Date Data Event Description System Classification Source .

414/2001-02 revealed that this event is the same event as noted above for Catawba- separate event 1 and Is not counted separately. (Reference [30]) _

Clinton 1 8/15/1989 Reference Inlet expansion joint to condenser leaking (1cup/hr.), replaced CWS Leak

[23] expansion joint later outage.

Clinton 1 3/18/1990 Reference Condenser water box over-pressurized, damage to water box CND Screened out; no

[23] expansion joint, no leakage. leak Comanche 6/6/1993 Reference Comanche Peak Unit 1 was at 85% power when rubber expansion joint Auxiliary Leak Peak 1 [23] on the circulation lube water pump discharge leaked resulting in six feet System of water in the circulating discharge room over a period of time. (AUX)

Expansion joint failed under vacuum when the pump was stopped.

Failure attributed to normal aging. Circulation pump 02 rubber expansion joint was replaced and the unit brought to 100% power.

This event involved leakage but not a catastrophic rupture (Reference

[331). _

Crystal River 3 8/14/1985 Reference 10In., SW pump suction expansion joint failed - hole in joint, aging (12 SW Leak

[23] years). A crack in the joint was discovered to be weeping. The expansion joints were replaced (Reference [26]). Evidence of leakage D.C. Cook 1 7/29/1990 Reference 8-Inch diameter expansion joint header connected to the Condensate CND Leak

[23] Storage Tank leaked because of 4-in. gash in the joint, cause unknown.

Expansion joint is composed of fabric and istwo-feet long. Fabric 2' long._

D.C. Cook 2 8/28/1987 Reference Expansion Joint XJ-54N in the discharge header west essential service SW Rupture < 2,000gpm

[23] water pump ruptured. Most probable cause of failure attributed to time /

function degradation aggravated by the impure quality of the raw water in this system. Replaced expansion joint. A phone conversation with a cognizant engineer indicated the expansion joint is 8 inches in diameter.

Nominal operating pressure is 80 psi. During the summer, at the time of this event, operating pressure was likely higher, at 85 to 90 psi. The expansion joint is in a line that serves as a minimum flow path during normal operations. The flow through the joint during normal operations

___ _is typically 2000 gpm (Reference [34]). -

Karl N. Fleming Consulting Services LLC Page 13 of 21

Expansion Joint FailureRates for Kewaunee PRA Plant Date Data Event Description System Classification

_____________ Source __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Diablo Canyon 1211/1998 Reference Two cooling water system synthetic expansion joints experienced SWP 2 ruptures <2000 1,2 [23] catastrophic failures. One failure caused a 500 to 1000 pgm leak from gpm a 16" elastomeric expansion joint in a connection between screenwash water system and a pipe embedded Inthe intake building wall that is connected to the turbine building cooling system (a configuration allowed by procedure). Eighteen hours later, a 6"elastomeric expansion joint in the closed loop intake cooling water system (cools the circulating water pump motors) occurred. These expansion joints are installed in locations not obviously visible and involve a wet (saltwater) location. Root cause was degradatalon due to corrosion of metal in the joint; saltwater introduction from the exterior of the joint.

Expansion joints were 23 years old and were not part of the inspection program. Expansion joints are designed similar to a tire, inside out tire waterproof internal, cotton core that wraps it for pressure control; steel stiffening rings; protective cover had breached __ _

Ft. St. Vrain 1D 4/7/1988 LER 88-006 CW expansion joint failed because of degradation, 54, 15 years old; CWS Rupture > 10,000 15" tear that resulted in a 15,000 gpm leak flow (Reference [22]). gpm Hatch 1 2/27/1977 Reference A forced shutdown resulted from a recirculation pump trip followed by a CND Rupture <2,000gpm

[35] condenser bellows rupture; not clear if this isa rubber expansion joint Indian Point 3 6/15/1988 Reference 30-in. condensate pump expansion joint leaking. CND Leak

[23]

Indian Point 3 12/8/1992 Reference 30-in. expansion joint on suction side of condensate pump had minor CND Leak

[23] leakage (excessive forces).

Indian Point 3 1/14/1993 Reference Expansion joint on suction side of condensate pump deformed and CND Leak

[23] leaked because of degradation.

Indian Point 3 2/3/1993 Reference Expansion joint on suction side of condensate pump deformed and -CND Leak

[23] leaked because of degradation.

LaSalle 1 &2 5/31/1985 LER 1985- 108-in. cir. water pump expansion joint failed resulting in flooding '(2000 CWS Rupture > 2,000 045 gpm), due to water hammer (LER 50-302/1989-011). Failure occurred gpm in the Lake Screen House. I Limerick 1 1/3/1998 Reference Leak in expansion joint on suction side of condensate pump. CND Leak Karl N. Fleming Consulting Services LLC Page 14 of 21 C C C

C C Expansion Joint Failure Rates for KewauneC .A Plant Date Data Event Description System Classification Source Limerick 2 12/115t1998 Reference Urmedck Unit 2 while at full load found a leak on the ESW EXJ after SW Leak

[23] starting the ESW pump. A small V shaped tear was observed at approximately 6:00 outside of the direct flow path. A failure analysis was performed on the EXJ after removal and it was concluded that the leak was due to an age related (end of life) failure. The failure worked its way through the expansion joints two plies and an arc of approximately 120 degrees before exiting the outer protective layer.

Millstone 2 4/28/1977 Reference 240-in SW pump expansion joint had ballooned out because of leakage SW Leak

[23] thru tube.

Oyster Creek Reference After the ESW pump WAwas started during a surveillance test, the SW Rupture < 2,000gpm

[35] rubber expansion joint ruptured. The reactor was shutdown for refueling Sequoyah 1 8/25/1994 Reference Main feedwater pump turbine condenser pump expansion joint CND Leak

[23] developed leak because of high temp.

St. Lucie 1 7/28/1986 Reference 30-inch intake cooling water system, rubber steel-reinforced expansion SW Leak --

[23] joint on the intake cooling water system (essentially equivalent to a SW system) pump outlet developed a leak (<1 00gpm). Failure was caused by aging and cyclic fatigue. The intake cooling water system isa sea water system that cools component cooling water and turbine cooling water, and is essentially equivalent to a SW system [CORRESP07].

Evidence of leakage. -_ _

St. Lucie 1 2/18/1990 Reference Intake 30-in. cooling water expansion joint had through wall leak, aging. SW Leak

[231 Turkey Point 3 6/14/1988 Reference 30-in. suction expansion joint for condensate pump leaking. CND Leak

_ _ __ __ __ _ _ _ _ _ _ _ _ __ _ _ __ [23 ]

VC Summer 214/1991 Reference 36-in. expansion joint on suction side of condensate pump had a hole in CND Leak

[23] liner with small leakage. . .

VC Summer 2/1'4/1991 Reference 36-in. expansion joint on suction side of condensate pump developed CND Leak

[231 leak, cycic fatigue.

VC Summer 3/30/1993 Reference 36-in. expansion joint on suction side of condensate pump developed CND Leak

[231 leak, cyclic fatigue.

Karl N. Fleming Consulting Services LLC Page 15 of 21

Expansion Joint Failure Rates for Kewaunee PRA Plant Date Data Event Description System Classification Source _

Surry 2 6/17/1986 LER 1986 While at 100% power, operations personnel discovered a SW Leak

06. service water leak in Unit 2 containment. The leak of approximately one gpm was In an expansion joint on the service water return line from a recirculation spray heat exchanger. The Inlet and outlet service water valves were closed to isolate the leak. It was determined that repairs could not be made with the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> LCO, therefore a rampdown was commenced and an unusual event was declared. The leak in the expansion joint was caused by galvanic corrosion.

Beaver Valley 2/2/1996 Reference Load reduction to 90% for expansion joint replacement at the CWS Screened out; no

[20] outlet of the D waterbox of the main condenser. leak Susquehanna 3/29/1989 Reference Manual shutdown due to circulation water system expansion CWS Leak 1 [29] joint leak - early refueling. Expansion joints leaked starting with minor dripping to ultimately about 1 gpm. Expansion joints are steel reinforced. During prior refueling outage, wear was noted

._ on the interior of the joints (Reference [29]).

Catawba 1 6/12/1993 Reference Shutdown to repair expansion joint leakage. This event involved CWS Leak

[30] leakage of a rubber expansion joint reinforced by stainless steel (Reference [30]). Contributing causes: wear and arrangement.

River Bend 1 2/14/1989 Reference Load reduction to repair waterbox B expansion joint. No CWS Screened out; no

[20] evidence of leakage leak Catawba 2 10/21/2001 Reference Refuel outage delay for condenser circulating water expansion CWS Screened out; no

_ [20] joint repair. No evidence of leakage leak Clinton 1 5/9/1990 LER 1990- The unit was shutdown when diesel generators (Division 1 & 2) SW Screened out; no 010 were declared inoperable because expansion joints in the leak shutdown service water system piping for the DG heat exchangers did not have required tie rods installed to prevent expansion beyond design limitations. The cause was a construction/installation error. The expansion joints are installed between the SWS piping and the DG heat exchangers to isolate vibratory motion between SWS piping and the DGs. The tie rods should have been located between the flanges of the expansion joints to prevent their expansion beyond design limitations.

Karl N. Fleming Consulting Services LLC Page 16 of 21 C 'C C

Expansion Joint Failure Rates for Kewaunee PRA I

i Table 3-3 Uncertainty Distribution Results for Expansion Joint Failure Rates t4-14) Component & Failure Mode Uncertaint Distribution [1/EXJ.YR Type Failure Mode Mean 6th Median 95est

_ _ _ __ Percentile Percentile Failures (leaks + 1.40E-04 5.69E-05 1.23E-04 2.84E-04 ruptures)

EWS Rubber Ruptures with 1.22E-05 2.92E-06 9.75E-06 2.97E-05 Expansion Joint leak flows > 2,000 Ruptures with 9.17E-06 1.82E-06 7.1 OE-06 2.33E-05 Leak flows from 2,000 to 10,000 gpm Ruptures with 6.08E-06 8.81 E-07 4.44E-06 1.67E-05 leak flows >

10,000 gpm .

GW EXJ Flooding > 10,000 gpm 7000.

60 00 - - -- - - -- - - -- - - -- - - - -- - - - - - - -- - - -

5000 - -- - - - - -I- - - - - - - - - - - - - - - - - - -

U-a 4000 t- -- -- -- - -- - --- -- -- -- -- -- -- -- -- -- -- -- -- -- ----------

ICI l' 3000 -- -- - -- - - - - - - - - - - - - - - - - -

0

<)2000 - -- - -- - - -- - - - - - - - - - - - - - - - - -

1000 - - -- -- - - - - - - - - -- - - -- - - -

0 2.33E-07 4.70E-06 9.17E-06 1.36E05 1.81 E-05 Expansion Joint Rupture Frequency ( >10,000 gpm)

Figure 3-1 UncertaInty Distribution Expansion Joint Rupture Frequency (>10,000 gpm)

Karl N. Fleming Consulting Services LLC Page 17 of 21

Expansion Joint Failure Rates for Kewaunee PRA Q,,

KrN. Flmn osligSrie L Page 18 of 21

Expansion Joint Failure Rates for Kewaunee PRA

- 4. REFERENCES (1] Fleming, K.N. and B.O.Y. Lydell, "Pipe Rupture Frequencies for Internal Flooding PRAs",

prepared by Karl N. Fleming Consulting Services LLC for EPRI, August 2005

[2] Poloski, J.P. et al, Rates of Initiating Events at U.S. Nuclear Power Plants, NUREG/CR-5750, U.S. Nuclear Regulatory Commission, Washington (DC), 1999.

[3] Tregoning, R., L. Abramson and P. Scott, Estimating Loss-of-Coolant Accident (LOCA)

Frequencies Through the Elicitation Process, Draft Report for Comment, NUREG-1 829, U.S. Nuclear Regulatory Commission, Washington (DC), June 2005.

[4] Lydell, B.O.Y., "PIPExp/PIPE-2004: Monthly Summary of Database Content (Status as of 31 -Dec-2004)", RSA-R-2004-01.07, RSA Technologies, Fallbrook (CA). Monthly summary reports have been issued since January 1999.

[5] Fleming, K.N. et al, "Piping System Reliability and Failure Rate Estimation Models for Use in Risk-informed In-Service Inspection Applications", TR-1 10161 (EPRI Licensed Material),

EPRI, Palo Alto (CA), 1998.

[6] U.S. Nuclear Regulatory Commission, Safety Evaluation Report Related to 'Revised Risk-Informed In-service Inspection Evaluation Procedure (EPRI TR-1 12657, Rev. B, July 1999, Washington (DC), 1999.

[7] H. Martz, TSA-1 /99-164: 'Final (Revised) Review of the EPRI-Proposed Markov Modeling/Bayesian Updating Methodology for Use in Risk-informed Inservice Inspection of Piping in Commercial Nuclear Power Plants,", Los Alamos National Labaratory, June 1999

[8] Fleming, K.N., Mikschl, T.J., 'Piping System Failure Rates and Rupture Frequencies for Use in Risk Informed In-service Inspection Applications", EPRI Report No. TR-1 11880 (EPRI Licensed Material), EPRI, Palo Alto (CA), 1999.

[9] Fleming, K.N. and B.O.Y. Lydell, "Database Development and Uncertainty Treatment for Estimating Pipe Failure Rates and Rupture Frequencies," Reliability Engineering and System Safety, 86:227-246, 2004

[10] Bush, S. et al, "Piping Failures in the US Nuclear Power Plants 1961-1995," SKI Report 96:20, Swedish Nuclear Power Inspectorate, Stockholm (Sweden), January 1996

[11] U.S. General Accounting Office, "Action Needed to Ensure That Utilities Monitor and Repair Pipe Damage," GAO/RCED-88-73, Washington (DC), March 1988.

[12] Cragnolino, G., C. Czajkowski and W.J. Shack, Review of Erosion-Corrosion in Single-Phase Flows, NUREG/CR-5156, U.S. Nuclear Regulatory Commission, Washington (DC),

April 1988.

[13] International Atomic Energy Agency, "Corrosion and Erosion Aspects in Pressure Boundary Components of Light Water Reactors," Proceedings of a Specialists Meeting organized by the IAEA and Held in Vienna, 12-14 September 1988, IWG-RRCP-88-1, Vienna (Austria),

April 1990.

Li KarlN. Fleming Consulting Services LLC Page 19 of 21

Expansion Joint Failure Rates for Kewaunee PRA

[14] OECD Nuclear Energy Agency, "Specialist Meeting on Erosion and Corrosion of Nuclear Power Plant Materials," NEANCSNI/R(94)26, Issy-les-Moulineaux (France), 1995.

[15] Beliczey, S. and H. Schulz, 'The Probability of Leakage in Piping Systems of Pressurized Water Reactors on the Basis of Fracture Mechanics and Operating Experience," Nuclear Engineering and Design, 102:431-438,1987.

[16] Tulay, M. et al, Expansion Joint Maintenance Guide, Revision 1, 1008035, Electric Power Research Institute, Palo Alto (CA), May 2003.

[17] Duke Power, Plans for Repair of Recirculating Cooling Water (RC) System Rubber Expansion Joints, York (SC), March 2002 (NRC-PDR Accession No. ML020880630).

[18] Atomic Energy of Canada Limited, Generic CANDU Probabilistic Safety Assessment -

Reference Analysis, 91 -03660-AR-002, Rev. 0, Sections 7.3 and H.5.1, Mississauga (Ontario), Canada, July 2002.

[19] NSAC-60, "Oconee PRA"

[20] Vess, J., OPEC Database Query, Energy Central, Colorado, October 3,2005.

[21] Stone and Webster Engineering Corporation, 'Surry Power Station Individual Plant Examination Review of Internal Flooding", SWEC 1991 A, 1991.

[22] Section 3 from Surry Internal Flood Analysis Supplemental Report.

[23] Tulay, M., "EPRI Expansion Joint Maintenance", EPRI TR-1 008035, Charlotte, NC, 2003.

[24] Kinsey, S., "Circulating Water Piping Expansion Joint Evaluation", MPR Associates.lnc., Alexandria, VA, 2005.

[25] Garlock On Site Internal Inspection of Expansion Joints, Julie Benzer, Garlock Sealing Technologies, March 11, 2005. ,_

[26] Phone conversation between Eric Jorgenson, Maracor and Dave Miskiewicz, 352-795-6486 ext 3019, Crystal River, October 3, 2005.

[27] Phone conversation between Eric Jorgenson, Maracor and Joe Anatasio, 805-545-4909, Diablo Canyon, September 29,2005.

[28] Phone conversation between Eric Jorgenson, Maracor and Bob Boyle, 724-682-4726, Beaver Valley, October 4,2005.

[29] Phone conversation between Eric Jorgenson, Maracor and Steve Ellis, 610-774-2519, Susquehanna, October 4, 2005.

[30] Phone conversation between Eric Jorgenson, Maracor and Steve Nader, 980-373-4273, Catawba, October 4, 2005.

[31] Phone conversation between Eric Jorgenson, Maracor and Chris Hicks, Product Manager, Garlock Sealing Technologies, 1-800-448-6688, October 4,2005.

[32] Phone conversation between Eric Jorgenson, Maracor and John Priolo, Design Engineering, St. Lucie, 772-467-7272, October 6, 2005.

[33] Based on E-message between Eric Jorgenson, Maracor and Dan Tirsun, Comanche Peak, 254-897-0865, October 6,2005.

Karl N. Fleming Consulting Services LLC Cj Page 20 of 21

Expansion Joint Failure Rates for Kewaunee PRA

[34] Phone conversation between Eric Jorgenson, Maracor and Steve Cherba, D.C.

Cook, 269-465-5901 ext. 1126, October 6, 2005.

[35] P. Janzen, "Atomic Energy of Canada Limited - A Study of Piping Failures in U.S. Nuclear Power Plants", AECL-Misc-204, April 1981 Karl N. Fleming Consulting Services LLC Page 21 of 21

Appendix A Initiating Events - High Energy Line Break Report Page I of 2

High Energy Line Break Report Owner Acceptance:tan1 4 c *M 'PbHA- G-. %4oo V<

Signature Print Name 1O117xS Date (401 Page 2 of 2

HIGH ENERGY LINE BREAK INITIATING EVENT FREQUENCIES FOR THE KEWAUNEE PRA Final Report Prepared for Dominion Energy Kewaunee Power Station By Karl N. Fleming Bengt O.Y. Lydell In Cooperation with Maracor Software & Engineering, Inc.

Karl N. Fleming Consulting Services LLC A CaliforniaLimitedLiabiflty Company 616 Sereno View Road Encinitas, CA 92024 United States of America October, 2005

HELB Initiating Event Frequencies for Kewaunee PRA TABLE OF CONTENTS Section Title Page

1. INTRODUCTION .. 4 1.1 Purpose .. 4 1.2 Scope ... 4 1.3 Objectives .. 4 1.4 Report Guide .. 4

2. TECHNICAL APPROACH .. 6 2.1 Overview ..

2.2 Uncertainty Treatment .. 6 2.3 Pipe Rupture Model .. 6 2.4 Definition of Pipe Failure Rate Cases .. 9

3. KEWAUNEE HELB-INITIATED INTERNAL FLOODING INITIATING EVENTS

. .......................................................................................... 12 3.1 Definition of ............................................................... 12 3.1.1 Steam Line.Breaks .12 3.1.2 Feedwater and Condensate Line Breaks .2 3.2 Break Frequency Calculations .. 13 Steam Line Breaks Causing Large Fire Protection System Actuations 13 3.2.1 3.2.2 Actuations Steam Line Breaks Causing Intermediate Fire Protection System

.. 14 U---

3.2.3 Feedwater and Condensate Line Breaks Causing Large Fire Protection System Actuations 15 3.2.4 Feedwater and Condensate Line Breaks Causing Intermediate Fire Protection System Actuations .1 3.3 Model Quantification . . .16

4. PIPE FAILURE RATES AND RUPTURE FREQUENCIES ... 17 4.1 System Boundaries . . .17 4.2 Database Screening . . .17 4.3 Database Query Results . . . 18 4.4 Exposure Term Data . . .23 4.5 Conditional Pipe Failure Probability.................................................................23 4.6 Results for Failure Rates and Rupture Frequencies . . . 26

5. HELB INITIATING EVENT FREQUENCIES . . .33 5.1 Calculation Steps . . . 33 5.2 Summary of Results . . .34 5.3 Sensitivity Study . . .37

6. REFERENCES................................................................... ........... 39 APPENDIX A PiPExp DATABASE DESCRIPTION . . . .41 APPENDIX B PIPE FAILURE RATES & RUPTURE FREQUENCIES APPLICABLE TO NON-CODE PIPING SYSTEMS ................. 51 KarlN. FlemingConsulting Services LLC Page 2 of 53 (

HELB Initiating Event Fr¶equencies for Kewaunee PRA TABLE OF FIGURES Title Page Figure 2-1 Flow Chart for Bayes' Estimates of System, Size, and Damage Mechanism Specific Pipe Failure Rates (A)and Rupture Frequencies (p)............... ................ 8 Figure 4-1 PWR-Specific Worldwide Service Experience ............................................... 19 Figure 4-2 PWR-Specific Worldwide Service Experience with non-Code Steam Piping 1970-2004 [1] ............................................ 19 Figure 4-3 Empirical Conditional Probability of Pipe Failure as a Function of Type of Piping System & Through-Wall Flow Rate Threshold Value ............................... 25 Figure 4-4 Impact of Different Data Screening Assumptions on FWC Piping Reliability32 Figure 4-5 Impact of Different Data Screening Assumptions on Steam Extraction Piping Reliability ............................................... 32 Figure 6-1 Crystal Ball Results for Large Feedline Break Frequency ............................ 35 Figure 5-2 Crystal Ball Results for Intermediate Feedline Break Frequency ........... ....... 35 Figure 5-3 Crystal Ball Results for Large Steam Line Break ........................................... 36 Figure 5-4 Crystal Ball Results for Intermediate Steam Line Break ................................ 36 Figure 5-5 Impact of Alternative Data Screening Regarding FAC .................................. 38 LIST OF TABLES Tthe Page Table 2-1 Pipe Failure Rate Analysis Cases ..........I.............. 10 Table 4-1 Service Experience with non-ASME Code FWC Piping ................................ 21 Table 4-2 Service Experience with non-Code Steam Piping ....................... .................. 22 Table 4-3 Piping Population Exposure Data ........................................................... 23 Table 4-4 Parameters of Posterior Beta Distribution for Pik{RIF) for non-Code FAC-Susceptible High-Energy Piping & non-Code FAC-resistant High-Energy Piping

............................................................................. ............................... 27 Table 4-5 Summary of FAC-Susceptible Piping Rupture Events with Equivalent Break Size........................................I,...., . :........................................................... 28 Table 4-6 Summary of FAC-Susceptible Piping Rupture Events with Equivalent Break Size > 6-inch Diameter (EBS2) .2.. ...................... i . . . 29 9..

Table 4-7 Mean Values of Failure Rate and Rupture Frequency Parameters .............. 31 Table 5-1 Uncertainty Distribution Results for HELB-lnitiated Internal Flooding Initiating Event Frequencies .. ............. . . . .34 Table 5-2 Impact of Alternative Assumptions Regarding Data Screening on HELB-Initiated Internal Flooding Inttiating Event Frequencies ........................................ 37 Karl N. Fleming Consulting Services LLC page 3 of 53

HELB Initiating Event Frequenciesfor Kewaunee PRA

1. INTRODUCTION 1.1 Purpose The purpose of this report is to document the derivation of initiating event frequencies that will be used as input to the turbine building internal flooding risk assessment at Dominion Energy's Kewaunee Power Station. Specifically, Initiating event frequency values associated with ruptures of high-energy lines that in turn cause actuation of fire protection systems will be determined. This work was performed via Subcontract to Maracor Software Engineering, Inc.

on behalf of Dominion Energy's Kewaunee Power Station. This report is intended to be an integral part of the overall turbine building internal flooding initiating events analysis.

1.2 Scope The scope of work covered in this report includes:

• Development of pipe failure rates and rupture frequencies for high energy piping (i.e.

piping with water or steam above saturation temperature) in PWR plants including the following systems:

o steam, including high pressure, low pressure, and extraction steam systems o feedwater system, including feedwater heaters and drain systems o condensate system

• Development of point estimates and probability uncertainty distributions for all parameters subject to data uncertainties

• Calculation of Kewaunee HELB initiating event frequencies including point estimates and probability uncertainty distributions based on information provided by Kewaunee and Maracor on initiating event success criteria and piping lengths 1.3 Objectives The objective is to perform a state of the art data analysis that Is consistent with the applicable requirements of ASME PRA Standard Capability Category II for Initiating event frequency development Consistent with this objective, the report is intended to provide a traceable basis for the calculations so that the results could be Independently reproduced from the information provided.

1.4 Report Guide A major part of this report is devoted to the development of a set of failure rates and rupture frequencies for use in the turbine building HEW-initiated internal flooding initiating event development. The technical approach to developing these failure rates and rupture frequencies Karl N. Fleming ConsultingServices LLC Page 4 of 53

HELB Initiating Event requencies for Kewaunee PRA is summarized in Section 2. in Section 3 the HELB-initiated internal flooding initiating event models for the Kewaunee PRA are described Including the success criteria for screening pipe locations and break sizes that apply to each event. The details of the development of the break sezes and locations for these events and break sizes are provided in Section 3.2 in the turbine building internal flooding Initiating events analysis, into which this report is to be integrated. The information in Section 3 of this -report is based on Information in Section 3.2 of the main report and was provided to the authors by Kewaunee and Maracor. The development of failure rates and rupture frequencies for this model using the methodology of Section 2 Is documented in Section 4. The results for the Initiating event frequencies Including point estimates and uncertainty distributions are summarized In Section 5. Section 6 lists the references used as inputs to the data development and methodology. Supporting details are provided In the Appendices.

. II KarlN. FlemingConsulting Services LLC Page 5 of,53

HELB Initiating Event Frequencies for Kewaunee PRA

2. TECHNICAL APPROACH 2.1 Overview The model used to estimate pipe break frequencies for the initiating event models in this calculation Isthe same as that used in a recent EPRI report on Internal flooding initiating event frequencies [1], and similar to that used In recent NRC studies regarding loss of coolant accident (LOCA) initiating event frequencies [2] [3]. The source of pipe failure and exposure data used to quantify the failure rates used In these models is known as OPIPExp-2004" [4]. A summary of this database is provided in Appendix A.

2.2 UncertaintyTreatment Uncertainties in these failure rates were quantified using a Bayes' methodology that was developed in the EPRI RI-ISI program [5] and approved by the NRC for use in applied RI-ISI evaluations [6]. An independent review of this pipe failure rate uncertainty treatment was performed to support the NRC Safety Evaluation and results of this favorable review are provided In Reference [7]. An earlier EPRI report 18] developed a set of pipe failure rates for use In the EPRI RI-ISI applications which was also approved and independently reviewed in References [6] and [7]. These earlier failure rate estimates were derived from a pipe failure database that had been developed. In Reference [10]. During subsequent work in applying these estimates Inapplied RI-ISI evaluation, a significant number of data classification errors In the original data source [10] were Identified and improved estimates of the exposure population became available. These factors, as discussed more fully in Reference [9], were the prime motivation to switch to the more comprehensive and validated "PIPlExp-2004m database when Reference 11], was developed. The most recent NRC sponsored work on LOCA frequencies 13]

is also based Inpart on the "PIPExp-20040 database.

Ci 2.3 Pipe Rupture Model The model used for relating failure rates and rupture frequencies uses the following simple model that is widely used in piping reliability assessment and was used in recent updates of recommended Loss of Coolant Accident frequencies [6]. The failure modes included in the estimation of failure rates include leaks and ruptures and, in some cases, cracks may also be included depending on the application. The model Isexpressed inthe following equation:

u M pax sPk 2AP#* {RxIF) (2.1) k-1 k-b Where:

= total rupture frequency of rupture size x for pipe size I in system i

= rupture frequency of rupture size x for pipe of size I insystem j due to damage mechanism k

= failure rate of pipe of size I Insystem j due to damage mechanism k Karl N. Fleming ConsultingServices LLC Page 6 of 53

HEL8 Initiating Even equencles for Kewaunee PRA P1 {RF) = conditional probability of rupture size x given failure for pipe size I in system j and damage mechanism k M = Number of different damage mechanisms In general, a point estimate of the frequency of pipe failures, 4' Is given by the following expression:

- . .k A(k = (2.2)

Where ni= the number of failures (cracks, wall thinning, leaks and ruptures) events for pipe size I Insystem j due to damage mechanism k To z the total time over which failure events were collected for pipe size i in system J NU = the number of components that provided the observed pipe failures for size i in system j k = the fraction of number of components of size I In system j that are susceptible to failure from damage mechanism k for conditional failure rates given susceptibility to damage mechanism k, 1for unconditional failure rates Note that all failure modes that result in pipe repair are included In the failure rate and that all failures thus defined are regarded as precursors to rupture. The events counted as ruptures are based on a specific definition of rupture which is application specific. For Internal flooding and HELB applications, we seek unconditional failure rates and hence we can combine these equations under the condition: fug =1 to obtain the following expression for the point estimate of the rupture frequency.

Rol PkIc Z4&PA{R.

k IF)X PtR IF) (2.3)

W=

= k-i NAU In the development of Bayes' uncertainty distributions for these parameters, prior distributions are developed for the parameters 4k and P*R IF) and these prior distributions are updated using the evidence from the failure and exposure data as in standard Bayes' updating. The exposure terms (denominator of the fractions on the right hand side of Equation (2.3) also have uncertainty as these terms must be estimated for the entire nuclear Industry that provides the number of failures for the failure rate estimation. This uncertainty is treated in this process by adopting three hypotheses about the values of the exposure terms which requires three Bayes updates for each failure rate. The resulting posterior distributions for each parameter on the right hand side of Equation (2.3) are then combined using Monte Carlo sampling to obtain uncertainty distributions for the pipe rupture frequencies. A picture of this process is shown in Figure 2-1. This flow chart shows the full treatment of uncertainty needed for the RISI formulation in Equation (2.2). For the Internal flooding and HELB formulation of Equation (2.3) the damage mechanism susceptibility fractions ( fp ) do not come Into play. The specific way in which this flow chart is applied is discussed in Section 4 for each system and failure mode.

Karl N. Fleming ConsultingServices LLC Page 7 of.53

HELB Initiating Event Frequences for Kewaunee PRA In Reference [11 rupture frequencies were developed for three rupture sizes that were selected to support internal flooding analysis. These sizes Include water sprays with flood rates of up to 100 gpm, flooding with flood rates of 100 to 2000 gpm, and major flooding with flood rates greater than 2000 gpm. For the Kewaunee HELB-initiated internal flooding models, a somewhat different rupture size model had to be developed as the criteria for producing the consequences of interest are based on specific rupture sizes that were determined in a deterministic calculation, based on the energy required to activate fire protection system sprinklers.

Nkver or Lot Tmreo rmes @ Three Esu,~ of

- pd cibuiystmaon . AMM Logo Domalbuonn Pipe 6ecl nd Upper Bound )u .25) Upper Bound (p.25) Based en angonee""n Dm*ae Mechanism Judgmentand weld Best Esfmata (pAO) B"s BlA~te (psAO) and OM susceppil4 t EDsmats*Om RISI LWer Bound (P.2L). Lwer Bound (p]25) lI Bayes' Update fbr Three Combinations of Population and DM Susc Lower-L.waUpde ( Beat.3st Updat(y-.SO) Upe2Updat.(p Bayes' Posterior Weighting Operation

,P(IP - F), Prob.Rupture Q.

I iuu ren, rn-rnenv Given Failure p, Ruoture Freque 3nerl Prior a .

Figure 2-1 Flow Chart for Bayes' Estimates of System, Size, and Damage Mechanism Specific Pipe Failure Rates (A)and Rupture Frequencies (p)

Karl N. Fleming ConsultingServices LLC Page 8 of 53 hi

- HELB Initiating Event!kequencies for Kewaunee PRA 2.4 Definition of Pipe FailureRate Cases To support the baseline calculations and some sensitivity calculations that were selected to develop risk management insights, a set of 24 analysis cases were devised as shown InTable 2-1. The variables used to define these cases Include the piping system, rupture size, and data screening assumptions.

A failure rate and a rupture frequency had to be developed for each case and, hence, a total of 48 parameter distributions were developed. As discussed more fully in Section 4, the dominant failure mechanism in HELB piping is low accelerated corrosion (FAC). The piping systems were put Into 4 major categories based on their general susceptibility to FAC. The systems in the HELB category that are susceptible to FAC Include the feedwater and condensate systems and the steam systems with relatively wet steam conditions with carbon steel pipe. Based on Insights from service experience and the piping design parameters, the high-pressure steam piping between the steam generators and the Inlet of the high pressure turbine Is generally not susceptible to FAC. The reasons for this include the use of thick walled pipe, dry steam conditions, and relatively straight bend free runs of pipe. In the PIPExp database there have been no Instances of FAC in this part of the main steam system. Hence the high-pressure steam piping is set aside as one category so that the remaining categories represent the FAC sensitive pipe. The FAC sensitive pipe was further broken down Into 3 categories based on the relative susceptibility to FAC; two categories for steam and one for feedwater and condensate.

The two steam categories in include the low-pressure steam pipe downstream of the HP turbine outlet and the extraction steam.

For each of the four system categories described In the preceding paragraphs, rupture frequencies were developed for two rupture size cases: Ruptures with equivalent break sizes between 2-Inches and 6-inches diameter, and ruptures with equivalent break sizes greater than 6-inches In diameter. The estimation of the rupture frequencies for each of these break size cases required the estimation of two parameters: a failure rate and a conditional probability that the break would be in the specified size range. The failure rate for each break size range is different because only pipes with a pipe diameter of at least 6-inches can produce a break size greater than 6-inches, whereas pipes as small as 2-Inches Indiameter can produce break sizes of 2-Inches and greater. To support the estimation of these parameters, separate queries of the pipe failure database had to be made for pipe failures (cracks, leaks, wall-thinning, and ruptures) and ruptures In th'e prescribed break size ranges. Then, these queries had to be matched up against the appropriate estimate of the pipe component population exposure terms.

The parameter estimation for these failure rates and conditional rupture probabilities is documented InSection 4.

Consideration was given to the development of system-specific failure rates and rupture frequencies separately for the feedwater system and for the condensate system as was performed in Reference [1] for the internal flooding application. It was decided to develop a composite set of failure rates and rupture frequencies for both systems combined for several reasons: One Is that there are inconsistencies in the way. in which system boundaries are established between feedwater and condensate that would give rise to Inconsistencies between how the data was classified and how it is applied to Kewaunee.

Second, there are a variety of different operating conditions within the condensate system and the feedwater system that give rise to different susceptibilities to the predominant damage mechanism, flow accelerated corrosion. For example there are normally several stages of Karl N. Fleming ConsultingServices LLC Page 9 of 53

HELA Initiating Event Frequencies for Kewaunee PRA feedwater heating in the condensate and additional stages in the feedwater system. Feedwater drains and heater and main feedwater and condensate lines have much different conditions.

Q Third, there is no noticeable trend in the failure and rupture service experience between the two systems. And finally, breaking up the data Into separate systems reduces the statistical quality of each data cell, i.e. would subdivide the data cells too finely so that the frequency of events within each data cell are statistically insignificant.

Based on what was learned In this study, the authors plan to Issue a revision to Reference [11 to replace the system specific rates In that reference with a composite set of rates for the feedwater and condensate systems.

Based on the success criteria discussed in Section 3, for each set of failure rates, two rupture modes had to be distinguished: those with equivalent break sizes between 20 and 6" and those with break sizes in excess of 6 inches. Depending on the location of the pipe break either or both of these rupture modes may contribute to a specific HELB-initiated Internal flooding Initiating event, as discussed more fully in Section 3. Separate conditional rupture probability models had to be developed to distinguish these cases.

Table 2-1 Pipe Failure Rate Analysis Cases Case System Pipe Size Data Screening Assumptions KNPPO1 FWC 2 2 inch Post-1988 data only KNPP02 FWC > 6 inch Post-1988 data only KNPP03 FWC k 2 inch Data up to 1988 only KNPP04 FWC > 6 inch Data up to 1988 only KNPPO5 FWC - 2 Inch FAC events removed (up KNPP06 FWC > 6 Inch FAC events removed KNPP07 Extraction Steam 2 2 Inch Post-1988 data only KNPP08 Extraction Steam > 6 Inch Post-1988 data only KNPP09 Extraction Steam 2 2 Inch Data up to 1988 only KNPPIO Extraction Steam > 6 inch Data up to 1988 only KNPP11 Extraction Steam 2 2 inch FAC events removed KNPP12 Extraction Steam > 6 Inch FAC events removed KNPP13 Low Pressure Steam 2 2 Inch Post-1988 data only KNPP14 Low Pressure Steam > 6 inch Post-1988 data only KNPP15 Low Pressure Steam 2 2 Inch Data up to 1988 only KNPP16 Low Pressure Steam > 6 Inch Data up to 1988 only KNPP17 Low Pressure Steam 2 2 inch FAC events removed KNPP18 Low Pressure Steam > 6 inch FAC events removed KNPP19 High Pressure Steam 2 2 inch Post-1988 data only KNPP20 High Pressure Steam > 6 inch Post-1988 data only KNPP21 High Pressure Steam 2 2 Inch X Data up to 1988 only KNPP22 High Pressure Steam > 6 Inch Data up to 1988 only KNPP23 High Pressure Steam 2 2 Inch FAC events removed KNPP24 High Pressure Steam > 6 Inch FAC events removed Karl N. Fleming ConsultingServices LLC Page 10 of 53

HELB Initiating Event Frequencies for Kewaunee PRA A review of the piping service data as discussed more fully in Section 4 reveals a significant improvement in piping system performance around 1988. It is reasonable to assume that this trend In performance Is due to industry and NRC efforts to improve plant performance in general, and in particular to address flow accelerated corrosion in augmented Inspection, repair and replacement programs. For the base case analysis only the service data since 1988 was used to calculate the failure rates as this data is viewed to be representative of current industry practice in managing piping system performance. As a contrast, the second case considered only the service data up to and including 1988. A third case was defined by screening out all the FAC related pipe failures. The purpose of the three cases was to understand the Importance of the prevailing failure mechanism for experienced high energy line breaks.

Failure rates were specialized for the wet and dry steam systems, and for the feedwater and condensate systems, by specializing the data analysis for the failure rates. The data from the FAC sensitive steam, feedwater, and condensate systems were combined for the purposes of estimating the conditional rupture size probabilities. The justification for this Is that essentially all the pipe ruptures In these systems are due to FAC and occur in similar carbon steel pipes.

The system-specific factors that influence the rupture frequencies are judged to be adequately reflected in the specialized failure rates. The conditional probability of rupture size is viewed to be primarily related to properties of the pipe material and the damage mechanism and less related to the property of the system. The piping system materials for all the FAC sensitive piping are very similar. This is consistent with the data treatment in References [1J, [3J, and [8J.

In summary, the piping failure rates and rupture frequencies developed in this study were quantified to address 4 different pipe system categories, 2 break size categories, and 3 data screening assumptions, giving rise to 24 data analysis cases. For each case, a pipe failure rate covering all failure modes, and a rupture frequency covering a specific break size range was developed and hence 48 parameters were developed.

KarlN. FlemingConsulting Services LLC Page I11 of 53

HELB Initiating Event Frequencies for Kewaunee PRA

3. KEWAUNEE HELB-INITIATED INTERNAL FLOODING INITIATING EVENTS 3.1 Definition of Breaks Quantification of the HELB-initiated internal flood initiating event frequency values Is performed for each initiating event defined Inthe turbine building internal flooding Initiating events analysis. A summary of the HELB-related initiating events is provided below.

3.111 Steam Line Breaks For steam line breaks, two HELB-initiated internal flooding Initiating events are analyzed.

The first is a steam line break that actuates enough fire sprinklers to result in full flow from both fire pumps to the Turbine Building. This event includes any break upstream of the turbine throttle valves below the operating deck with an' equivalent diameter less than nine inches but greater than two inches, any break In the extraction steam line greater than six Inches, and any break In a line after exiting the high-pressure turbine with an equivalent diameter of six Inches or greater.

The second event Is a steam line break that actuates approximately 100 sprinklers. The Turbine Building HELB models show that 100 sprinklers are representative of moderate releases. This event includes breaks in the extraction steam lines with an equivalent break size between two and six Inches, and breaks in a line after exiting the high-pressure turbine and having an equivalent diameter of two to six inches.

3.1.2 Feedwater and Condensate Line Breaks For feedwater and condensate line breaks, two HELB-nitiated Internal flooding initiating events are analyzed. The first is a feedwater or condensate line break that actuates enough fire sprinklers to result in full flow from both fire pumps to the Turbine Building.

This event includes any between the fourth and fifth feedwater heaters with an equivalent diameter of greater than six Inches or any break downstream of the fifth feedwater heaters with an equivalent diameter greater than two inches.

The second event Is a feedwater or condensate line break that actuates approximately 100 sprinklers. The Turbine Building HELB models show that 100 sprinklers are representative of moderate releases. This event Includes breaks in the lines between the fourth and fifth feedwater heaters with an equivalent diameter between two and six inches.

Karl N. Fleming ConsultingServices LLC Page I.2 of 53

HELB Initiating Event Frequencies for Kewaunee PRA 3.2 Break Frequency Calculatons 32.1 Steam Line Breaks Causing Large Fire Protection System Actuations This analysis will use the pipe length values determined in the turbine building Internal flooding initiating events analysis. For steam piping located upstream of the turbine throttle valve, a total of 884.6 linear feet of piping were identified on the mezzanine and basement levels. For extraction steam, a total of 176.5 linear feet of piping was identified on the mezzanine and basement levels. For steam lines after the exit of the high-pressure turbine, a total of 621.7 linear feet of piping was identified on the mezzanine and basement levels. All other steam piping was located either on the operating deck or in the Auxiliary Building.

For piping located upstream of the turbine throttle valve, the frequency of pipe ruptures includes all failures with an equivalent diameter of greater than two inches. The frequency of failures in steam piping upstream of the turbine throttle valve, Fops, can be calculated as follows:

FHps =LHPS(PftPPi + PKWPP,0) =LHPS(1AJPPI9P{2 - 6jF) + Ayjwp2OP{> IF)) (3.1)

Where:

Lx Length of pipe in systemX

-Pipe Rupture Frequency for Case] (see Table 2-1)

=O AX =Pipe Failure Rate for Case] (see Table 2-1)

P{2 - 61F} = Conditional probability of pipe rupture of size 2" to 6" given pipe failure in pipe > 2 inch in size pi> 61F)) = Conditional probability of pipe rupture of size > 6" given pipe failure in a pipe > 6 inch in size The systems and cases are defined In Table 2-1.

The above equation uses the pipe modeling methodology of Reference [1] in which ill the failure modes of the metallic system pressure boundary components are averaged Into a pipe system failure rate per linear foot of pipe. Since all the pressure boundary failure modes were Included In the data analysis, there is no need to add separate terms to the equations to account for such components as valves, heat exchangers, pump bodies, and metallic expansion Joints. This approach is also justified by the fact that KarlN. Fleming Consulting Services LLC Page 13 of 53

HELB Initiating Event Frequencies for Kewaunee PRA most of the experienced pipe failures occur in pipes or where pipes are welded to other pipes or piping components.

For extraction steam piping, the frequency of pipe ruptures includes all failures with an equivalent diameter of greater than six inches. The frequency of failures in the extraction steam piping can be calculated as follows:

is= LESpwpog - LIES (ApNPPOP{ 61F} (3.2)

For steam piping after the exit of the high-pressure turbine, the frequency of pipe ruptures Includes all failures with an equivalent diameter of greater than six inches. The frequency of failures in this piping can be calculated as follows:

For=LRSPP]4=LRS(AXVP 4Pf> PjF) (3.3)

The total frequency for steam line breaks that actuate enough fire protection sprinklers to result In full system flow to the turbine building is the sum of the three values calculated above or:

FSZ8L F=ps + F6s + FRS (3.4) 3.2.2 Steam Line Breaks Causing Intermediate Fire Protection System Actuations (_)

Calculation of the frequency of this event Is performed as shown in Section 3.2.1 for large steam line breaks. Pipe length data also are identified in that section.

For extraction steam piping, the frequency of pipe ruptures Includes failures with an equivalent diameter of between two and six inches. The frequency of failures in the extraction steam piping can be calculated as follows:

FESM = LapmALEd = L=Ap"P{2- 6IFJ (3.5)

For steam piping after the exit of the highpressure turbine, the frequency of pipe ruptures Includes all failures with an equivalent diameter of between two and six inches.

The frequency of failures In this piping can be calculated as follows:

F=.W LrPF13=LprSPM P(22- IF) (3.6)

The total frequency for steam line breaks that actuate approximately 100 fire protection sprinklers Is the sum of the two values calculated above or KarlN. Fleming ConsultingServices LLC Page 14 of 53

HELB Initiating Event Frequencies for Kewaunee PRA FSZM = F&M + FSM (3.7) 3.2.3 Feedwater and Condensate Line Breaks Causing Large Fire Protection System Actuations This analysis will use the pipe length values determined In the turbine building internal flooding initiating events analysis. As discussed in Section 3.1.1, this event includes any break with an equivalent diameter greater than two inches in piping downstream of the 15 feedwater heaters and any break with an equivalent diameter greater than six Inches between the 14 and 15 feedwater heaters. For feedwater piping located downstream of the 15 feedwater heaters, a total of 331.56 feet of pipe was identified. For piping between the 14 and 15 feedwater heaters, a total of 696.55 feet of pipe was identified.

The failure frequency for these size breaks in this piping is calculated to be:

FPLIS =-LLI5(PENPOe + PWPS02) = LfLli(AKNppol P{2 -61F} +AWMP{> 61F)) (3.8)

For piping between the 14 and 15 feedwater heaters, only pipe breaks greater than six-inches equivalent diameter are included. The failure frequency for these size breaks In this piping Is calculated to be:

Fn45L = L4USpPMNPPO2 = LFL45(AKXppO2 P{> 61F} (3.9)

The frequency of feeedwater and condensate line breaks for this Initiating event is the sum of the two values above or:

FFLBL =FFLI5+ FFL245L (3.10) 3.2.4 Feedwater and Condensate Line Breaks Causing Intermediate Fire Protection System Actuations Calculation of the frequency of this event Is performed as shown In Section 3.2.3. Pipe length data also are Identified in that section. As discussed in Section 3.1.2, this event includes any break with an equivalent diameter between two and six Inches between the 14 and 15 feedwater heaters. Using that data and the methodology of this report, the failure frequency for these size breaks in this piping is calculated to be:

FFL45M = LFUSPXNPPO1 = LFL45AJNpPQOIP{2 -661F) (3.11)

Karl N. Fleming ConsultingServices LLC Page 15 of 53

HELB initiating Event Frequenciesfor Kewaunee PRA 3.3 Model Quantffication The technical approach to model quantification Isto develop uncertainty distributions for each of the parameters defined Inthis section and then to propagate these distributions through the equations using Monte Carlo simulation, a traditional approach to PRA uncertainty quantification. The development of the pipe failure rate and rupture frequency parameters in these models Isdocumented InSection 4 and the results of the Monte Carlo analysis are provided InSection 5. The pipe length estimates described In the above section were provided to the authors by Maracor and are documented in the main body of the turbine building internal flooding initiating events report of which this analysis will be an attachment Uncertainly In pipe length estimates is modeled using normal distributions with the estimated pipe lengths taken as the mean values and a standard deviation of 10% of these length estimates.

Karl N. Fleming ConsultingServices LLC Page 16 of 53

HELB Initiating Event Frequencies for Kewaunee PRA

4. PIPE FAILURE RATES AND RUPTURE FREQUENCIES 4.1 System Boundaries This evaluation is concerned with non-ASME Code piping systems Inside the Turbine Building of Pressurized Water Reactor (PWR) plants. The following systems are considered;

• Feedwater & Condensate (FWC) piping: The Condensate piping system extends from the Condenser Hotwell up to and including the Low Pressure Heaters. It also Includes the Drains and Vents System piping from the Low Pressure and High Pressure Heaters.

The Feedwater piping system boundary considered in this evaluation consists of the piping from the Low Pressure Heaters, the Feedwater pump suction/discharge piping, High Pressure Heater inletloutlet piping up to the outboard containment isolation valves.

Due to comparable susceptibilities to flow accelerated corrosion (FAC) and plant to plant variabilities In how the boundaries between these systems Is defined, a composite set of data parameters are developed for FWC piping.

• Steam Extraction piping: In a typical PWR the high pressure portion of the turbine has extraction connections for two stages of feedwater heating. The low pressure portion of the turbine has extraction connections for four stages of feedwater heating.

• Low Pressure Steam piping: In this evaluation, the low pressure steam piping Includes piping between the high pressure (HP) and low pressure (LP) turbine stages, including steam cross-over and cross-under piping, and Moisture Separator Reheater (MSR) piping. The MSR piping is also located between the HP and LP turbines and it is used to extract moisture from the steam and reheat the steam to improve the turbine performance.

• HP Steam piping:. In this evaluation the HP steam piping is upstream of the HP turbine throttle valve and extends to the outboard containment isolation valves.

4.2 DatabaseScreening The pipe failure rates and rupture frequencies In this evaluation are derived from service data Included in the PIPExp database (Appendix A). The full PIPExp includes on the order of 6,700 data records covering Code Class 1-3 and non-Code piping in commercial light water reactor plants. Input parameters to the pipe failure rate calculation in this evaluation are obtained through database queries that Include filters for excluding any non-relevant service data:

• Initial screening on the basis of Code Class and PWR plant system. Retain failure data associated with non-Code piping in Turbine Building including the following systems:

o Condensate System o Extraction steam piping o Feedwater heater drain and vent piping KarlN. Fleming Consulting Services LLC Pgc 17 of 53

HELB Initiating Event Frequencies for Kewaunee PRA o Main Feedwater (from LP feedwater heaters to outboard containment isolation valves) o Main Steam (from outboard containment isolation valve to High Pressure turbine steam admission valve, and turbine cross-over/cross-under piping) o Moisture Separator Reheater piping

• Results of Initial screening subjected to additional screening on the basis of nominal pipe size and through-wall flaw size:

• The evaluation considers piping of nominal pipe size (NPS) greater than 2-inch diameter as piping less than 2-inch is not within the scope of the HEL-initiated internal flooding initiating event models described in Section 3.

The service data involving through-wall flaws are reviewed In accordance with the Kewaunee HELB-initiated internal flooding initiating event analysis requirements (i.e., "moderate" versus

'major" release). This means that the service data are screened further on the basis of flaw size

('equivalent diameter break size'). The results of this screening step are input to the derivation of posterior Beta distribution parameters for calculation of conditional pipe failure probabilities for 2" to 6" and greater than 6" break sizes.

4.3 DatabaseQuery Results The results of the database queries are summarized In charts (Figures 4-1 and 4-2) and tables (Tables 4-1 and 4-2). Flow-accelerated corrosion (FAC) Is a predominant degradation mechanism for the systems that are included in the study scope except for the high pressure steam system. Most If not all plant owners have implemented programs to mitigate FAC susceptibilities. These programs include Implementing non-destructive examination (NDE) programs, pro-active monitoring of pipe wall wear rates, and replacing the original carbon steel piping with FAC-resistant piping material such as stainless steel, carbon steel clad on the inside diameter with stainless steel, or chrome-molybdenum alloy steel. The purpose of these initial data queries was to identify the appropriate data set to use that represents current industry practice for predicting the initiating event frequencies at Kewaunee. The use of time trend analysis is a requirement of the ASME PRA standard for Capability Category 3 analyses. In addition, evaluating the trending of events avoids important insights In the data that would be missed by simply averaging all the industry experience.

Karl N. Fleming ConsultingServices LLC Page 18 of 53

HELB Initiating Event Frequencies for Kewaunee PRA ei. .w.

inI 16 1COND (B6Recrds) 1FWM330Records) I d

z

-l 14 le le 4, 4,P 4,e~ ~

Figure 4-1 PWR Worldwide Experience with non-Code FWC Piping 1970-2004 160 J 100 leo 60 40 Figure 4-2 PWR Worldwide Experience with non-Code Steam Piping 1970-2004 [11 KarlN. Fleming Consulting Services LLC Page 19 of 53

HELB InitiatIng Event Frequencles for Kewaunee PRA The two charts above show a distinctly higher incident rate before 1988. The before/after-1988 trend in Figure 4-1 and 4-2 is accounted for in the quantitative evaluation of the service-data.

The service data coverage In PIPExp corresponds to 858 PWR reactor years for the period 01011/1970 - 12/31/1987 and 1666 PWR reactor years for the period 01101/1988-12131/2004.

By the early- to mid-1980's the industry experienced several major failures of non-Code carbon steel piping (e.g., Trojan In March 1985 and Surry-2 in December 1986) (See References [11]

through [14]). In response to these events as well as the Industry-wide experience with pipe wall thinning and minor through-wall flaws attributed to FAC.

Tables 4,1 and 4-2 show the same data sets as those included in Figures 4-1 and 4-2 except that the data is organized by failure mode and pipe size to reflect the Kewaunee HELB-initiated internal flooding initiating event analysis requirements. The following failure mode definitions are used:

• Wall thinning; represents cases of severe wall thinning resulting In either weld overlay repair or preemptive replacement of affected piping section or fitting (e.g., elbow, tee).

• Leak; includes pinhole leak, leak or large leak resulting In isolation (where feasible) or manual reactor shutdown to effect repair or replacement.

• Rupture; significant through-wall flaw resulting in moderate or significant steamnfater release and prompt manual shutdown or automatic turbine trip/reactor trip.

As will be discussed more fully below, In developing estimates of the conditional rupture size probabilities, a special query is made on the database to identify those ruptures that fit into two size categories: 2" to 6", and greater than 6" equivalent break sizes.

Karl N. Fleming ConsultingServices LLC Page 20 of 53

---

ni HEL - in Event F f e RA f C HELB InitiatingEvent Frequenciesfor Kewaunee PRA

_Table_e 4_1 Sen_ Experience with non-ASME Code FWC Piping Nominal ripe Slze 1970-1987 1988-2004

-w ) Total Wall Leak Rupture Totl Wanl Leak Rapture

[inch) __ _ Thnning_ Thinning 2 < NPS 6" NPS > 6" 14 300 5

275 6

17 3

8 18 52 7

30 t_4 i 7 Total 314 280 23 1 1 1 70 37 1 22 11 NOW

• Service experience in Table I derived from 2524 reactor-years of PWR operation worldwide; 858 reactor-years pre-1988 and 1666 reactor-years post-1987

• Failure data includes contributions from FAC (dominant degradation mechanism), vibration-fatigue and water hammer

• The rootcause of post-1987 events in many cases is attnbuted to programatic errors orweaesses in the Owner's FAC progran

• Appendix A includes information on the coverage and completeness of the PIPExp database KarlN. Fleming ConsultingServices LLC Page 21 of 53

HELB Initiating Event Frequencies for Kewaunee PRA Table 4-2 Service Experience with non-Code Steam Piping Nominal Pipe Size 1970-1987 1988-2004 System (NPS) Total Wall Leak Rupture Total Wall Leak Rupture

[Inch] Thinning Thinninu EXT-St 2"<NPSS6" 10 0 8 2 9 1 7 1 NPS > 6"1 392 385 4 3 7 2 2 3 2"<NPS56n 14 0 11 3 15 1 10 4 LP-Steam NPS>6" 61 60 1 0 14 2 9 3 HP-Steam NPS>2" 24 19 3 2 9 1 7 1

_ Total: 501 464 27 10 54 7 35 12

• 'EXT-Steam' includes HP & LP steam extraction piping. Most of this piping is > NPS6.

• 'LP-Steam' includes piping between the HP and LP turbine stages, including cross-over/under piping and Moisture Separator Reheater piping.

• 'HP-Steam' includes piping upstream of the HP turbine throttle valve.

• Service experience in Table I derived from 2524 reactor-years of PWR operation worldwide; 858 reactor-years pre-1988 and 1666 reactor-years post-1987

• Failure data includes contributions from FAC (dominant degradation mechanism), vibration-fatigue and water hammer

• The root cause of post-1987 events minmany cases is attributed to programmatic errors or weaknesses in the Owner's FAC program

• Appendix A includes information on the coverage and completeness of the PIPExp database Karl N. Fleming ConsultingSerices LLC Page 22 of 53 C C C

HELB Initiating Even Frequencies for Kewaunee PRA 4.4 Exposure Term Data In pipe failure rate estimation, the exposure term is the product of either the number of components (e.g., fittings, welds) or total length of piping that provides the observed pipe failures and the total time over which failure events are collected. There is variability in the population counts. In part this variability stems from differences across NSSS types and balance of plant design differences, and In part it stems from different piping design and fabrication practices (e.g., use of cold bent piping versus use of welded fittings). Also, design modifications are Implemented during the lifetime of a plant to enhance flow conditions, minimize system vibrations, and to Improve the access for non-destructive examination (NDE), etc. Table 4-3 summarizes piping population data for the systems covered in the Kewaunee HELB-initiated Internal flooding Initiating events analysis.

Table 4-3 Piping Populaion Exposure Data System / System Group Linear ft of Piping Information Source / Comment EPRI TR-1 11880, Table A-5; in the failure rate FWC (> NPS2) 14,037 ft calculation the given length is input as a median

. value Entergy Nuclear Northeast (Indian Point-3 FAC EXT-Steam 1,500 ft program information). In the failure rate calculation the given length is input as a median value.

Dominion Energy; the given length is for KNPP and LP-Steam 622 ft in the filure rate calculation it is input as a lower

.__ .bound value Dominion Energy; the given length is for KNPP and

]HP-Steam 885 f in the failure rate calculation it is input as a lower bound value 4.5 ConditionalPipe FailureProbability For FAC-susceptible piping the likelihood of rapid or unexpected flaw propagation given wall thinning is quite high and can be estimated directly from service data. In the case of pipe materials or systems that are not susceptible to FAC such as the high pressure main, steam system at Kewaunee, there are much fewer events from which to derive the conditional rupture probability. "in this case the estimation of the likelihood of sudden pipe failure relies on Insights from service experience with different piping systems and materials under different loading conditions in combination with engineering judgment and fracture mechanics evaluations.

The likelihood of a through-wall flaw propagating to a significant structural failure is expressed by the condIitional failure probability P&({R/F. It is determined from service experience insights and engineering judgment, with the uncertainty treated using the Beta Distribution.

KarlN. Fleming ConsultingServices LLC Page 23 of 53

HELB Initiating Event Frequencies for Kewaunee PRA The beta distribution takes on values between 0 and I and is defined by two parameters, A and B (some texts refer to these as "Alpha" and "Beta"). It is often used to express the uncertainty in the estimation of dimensionless probabilities such as MGL common cause parameters and failure rates per demand. The mean of the Beta Distribution is given by.

Mean = A (4.1)

A+B If A = B = 1, the beta distribution takes on a flat distribution between 0 and 1. If A = B = %, the distribution Is referred to as a Jeffery's non-informative prior and Is a U shaped distribution with peaks at 0 and 1. Expert opinion can be Incorporated by selecting A and B to match up with an expert estimate of the mean probability. For example, to represent an expert estimate of 10'2, A=1 and B=99 can be selected. These abstract parameters A and B can be associated with the number of failures and the number of successes in examining service data to estimate a failure probability on demand. A + B represents the number of trials.

The beta distribution has some convenient and useful properties for use In Bayes' updating. A prior distribution can be assigned by selecting the initial parameters for A and B, denoted as Apr,, and Bpj,. Then when looking at the service data, if there are N failures and M successes observed, the Bayes updated or posterior distribution is also a Beta distribution with the following parameters:

A =.Ai,,, +X (4.2)

B = Bprior +M (4.3)

The above explains how the Beta distribution is used In this study to estimate conditional rupture probabilities. The priors are selected to represent engineering estimates of the probabilities "prior" to the collection of evidence. Equations (4.2) and (4.3) are used to compute the parameters of the Bayes' updated distribution after applying the results of the data queries to determine N and M. N corresponds to the number of ruptures Inthe specified size range and M corresponds to the number of pipe failures that do not result in a rupture in the specified size range.

A review of service data provides some insights about the conditional pipe failure probability for different types of piping systems. Figure 4-3 shows the conditional failure probability for different, observed through-wall flow rate threshold values. For comparison the Beliczey-Schulz correlation [15] is re-calibrated for through-wall flow threshold values rather than pipe size; this correlation only applies to Code Class I piping. According to Beliczey-Schulz, for 1-Inch piping the conditional probability of a major structural failure (MSF) or rupture is on the order of 5.OxlO2 (corresponding to a liquid flow rate of about 800 gpm (completely severed pipe), which is well beyond the upper threshold value in Figure 4-3. This information is presented to help justify the prior distribution parameters A and B selected for this analysis.

KarlN. Fleming ConsultingServices LLC Page 24 of 53

HELB Initiating Event Frequencies for Kewaunee PRA 1.OE-0Oi 1.0E-02 E 1.0E UaClss I (RCPB piping) C

-&- Class 3 (Moderate Energy Piping - All Systems)

-*-High-Energy, FAC Susceptible Piping

-- o senzey-WuLz 11987) - Class I Piping n>O n>1 n>5 n10 n>50 n>100 n>200 n>300 Observed Release Rate Thershold Values gpm]

Figure 4-3 Empirical Conditional Probability of Pipe Failure as a Function of Type of Piping System &Through-Wall Flow Rate Threshold Value' The WAG parameter of the Beta Distribution corresponds to a significant consequence (spray, Internal flooding or major flooding event) and the "Br parameter corresponds to the remaining failure experience (significant wall thinning or through-wall flaw). The total number of failures In the database Is equal to A+B. Table 4-4 is a summary of the prior and posterior Beta Distribution parameters for non-Code FWC and steam piping used in this report. The posterior distribution parameters are derived by performing a Bayes' update of the assumed prior distributions using service data from PIPExp and the conjugate properties of the Beta Distribution.

Part of the Information presented In Table 4-4 Isthe screening of pipe ruptures in different break size ranges In the FAC sensitive piping. The 26 events with equivalent break sizes between 2' and 6n are listed in Table 4-5, and the 33 events with break sizes greater than 6-inches are in Table 4-6.

'Plotted in tet figure are the conditional probabilities of leak flow rates given pipe failure as estimated by the fraction of the pipe failures in the failure data population with the indicated leak flow rte.

Karl N. Fleming Consulting Services LLC Page 25 of 53

HELB Initiating Event Frequencies for Kewaunee PRA 4.6 Results for Failure Rates andRupture Frequencies Using the methodology described In Section 2, uncertainty distributions were developed for the failure rates and rupture frequencies for each of the analysis cases in Table 2-1. The mean values of these distributions are presented In Table 4-7. The full uncertainty distributions were propagated through the HELB-initiated Internal flooding initiating event models that were described In Section 3 and the results are presented in Section 5. Parameters of these distributions are presented InAppendix B.

To support sensitivity calculations that are summarized in Section 5, comparisons were made among the data screening sensitivity cases for each system group that were identified. As seen in Figures 4-4 and 4-5 the results for the case using only data from prior to 1988 before FAC programs became effective would increase by more than an order of magnitude. Stated another way, the failure rates and rupture frequencies based on the service data before 1988 are more than an order of magnitude greater than those considering only data from events after 1988 when the FAC programs were In effect.. Conversely, if all the FAC-related events were precluded by some type of plant change, an order of magnitude reduction In the relevant pipe failure rates and rupture frequencies would be expected.

Karl N.Fleming ConsultingServices LLC Page 26 of 53

f  ::c HELB Initiating Event Frequencies for Kewaunee PRA Cr Table 4-4 Parameters of Posterior Beta Distribution for P,,{RIFJ for non-Code FAC-SusceDtible Hich-Enerav PiDina & non-Code FAC-resistant Hiah-Enerav PiDing AnysIs Case Prior Beta Parameters Posterior Beta Parameters Piping Eqnialent Constraint Apr Bprw Ap BA.,

t Mean Material Break Size (EBS)

CarbonSteel r2<EBS56" 1.01-2 I 99 27(l) 1254 2.11E-02 and EBS > 6" L.OE-2 1 99 3 1072 3.07B-02 FAC-susceptible X .

Stainless Steel 2" < EBS S 6" 1.OE-3 1 999 10 1062 9.33E-03 or EBS > 6"' .OE-31 999 8 1036 7.661-03 FAC-resistant Notes:

(1) A through-wall flaw of size 2" < EBS s 6" can occur iany FAC-susceptible piping of nominal pipe size (NPS) >

2". The database screening criteria include consideration of NPS and through-wall flaw size.

(2) A through-wall flaw of size BBS > 6" can occur in any FAC-susceptible piping of NPS > 6".

• EBS - Equivalent Break Size

• NPS = Nominal Pipe Size [inch]

• The posterior Beta distribution parameters are obtained from PIPExp database (accounts for service experience applicable to non-Code FWC and steam piping in Light Water Reactors):

- B1p = By + (Bjde.. - AEON

- AErae- Total number of ruptures in specified size range

- BEH =Total number of failure records -1181 records (carbon steel FWC piping ofnominal pipe size greater than 2. There are 1006 records for piping > 6" NPS.

- A dopg Al + AEO.; the evidence is 26 records for which the through-wal defect is sufficient to create a significant outflow of steam/condensate corresponding to 2" < EBS S 6" (Table 4-5).

- Ap]smj = Apfrd + Aw,~,; the evidence is 33 records involving major structural failure of FAC-susceptible piping corresponding to EBS > 6-inch diameter (Table 4-6)

• The Beta distnbution parameters for 'stainless steel or FAC resistant case' arc obtained by screening out any data record involving degradation or failure caused by FAC. A total of 72 records involve non-FAC failures and of these, 44 records involve piping > NPS6.

Karl N. Fleming ConsultingServices LLC Page 27 of 53

HELB InitiatingEvent Frequenciesfor Kewaunee PRA Table 4-5 Summary of FAC-Susceptible Piping Rupture Events with Equivalent Break Size Between 2-inch Diameter and 6-Inch Diameter (EBS1 DATABASE EVN LN YTM NOMINAL RECORD DATE PLANT NAME COUNTRY PLAT SYSTEM SGROYS PIPE SIZE NO. DAETP[RU inch]

2962 4/22/1995 Ahniarnz-1 ES PWR COND FWC 6 15272 2/13/2001 Balakovo-2 RU PWR FW FWC 3.2 2907 7(27/1993 Bohunica-3 SK PWR MS STEAM 6 455 9/28/1983 Browns Feny-t US BWR MSR STEAM 6 456 11/1/1977 Browns Fonfy-3 US BWR EXT-Stea STEAM 6 3722 8/10/1999 Callaway US PWR FW FWC 6 1166 9/25/1985 Dresden-2 US BWR COND FWC 6 2787 11/17/1986 Fernni-2 US BWR FW FWC 6 1425 4/28/1970 H.B. Robinson-2 US PWR MS STEAM 6 1975 3/1/1977 Hatch-) US BWMR CON__D __4 FWC 1463 9/2611989 Indian Point-2 US PWR MS STEAM 4 2866 4/3/1987 Indian Point-2 US PWR FW FWC -_6 2498 11/24/1993 KRla-4 RU PWR MS STEAM 4_

999 1/1/1972 Millstone- - US BWR MS STEUUI 4 494 12/0/1973 Millstone-l US BWR CON__D FWC 4 2161 12/31/1990 Millstono-3 US PWR MSR STEAM 6 498 12131/1990 Millstone-3 US PWR MSR STEAM_ 6 501 3/19/1983 Oconee-2 US PWR MSR STEAM 3 2949 12/15/1996 Paks,3 HU PWR EXT-STEAM STEAM 6 478 7M9/1986 RE. Gmnna US PWR MS STEAM 6 850 11/1&8977 Ringhals-2 SE PWR FW FWC 6 607 3/23/1990 Surry-l US PWR MSR STEAM 4 540 8f7/1972 _______ US__PWR us_ MSR STEAM 4 1536 1/9/1982 Troman US PWR EXT-STEAM STEAM 6 697 8/1/1983 Zion-I US PWR EXT-STEAM STEAM 6 2458 7/28/1991 7 Zio__2 US PWR FW FWC 3 Karl N. Fleming Consulting Services LLC Page 28 of 53 C7 C) C!

IC C (

HELB Initiating Event Frequencies for Kewaunee PRA Table 46 Summary of FAC-Susceptible Piping Rupture Events with Equivalent Break Size > 6-inch Diameter (EBS2)

RCORtD REOD EVENT DATE PLANTNAME COUNTRY PLANT TYPE SYSTEM SYSTEM GROUP PNOINAL PIESZ NO. _ _ _ __ _ _ _[Inchl 2865 12/18/1991 Ahnaraz-1 ES PWR MS STEAM 8 445 4118/1989 ANO-2 (Arkansas-2) US PWR MS STEAM 14 454 9/29/1982 Browns Fary-I US BWR MS STEAM 8 453 6/24/1982 Browns FeWr-4 US BWR MSR STEAM 8 15185 8/15/1983 Brons Fery-I US BWR MS STEAM 8 462 11/20/1984 Calvert Cliffs-I US PWR EXT-STEAM STEAM 16 465 1/15/1988 Catawba-1 US PWR COND FWC 8 2912 9/25/1987 Doel-I BE PWR COND FWC 8 2504 4/10/1993 Fermi-2 US BWR EXT-STEAM STEAM 8 2785 4/21/1997 Fot Calhoun-I US PWR FW FWC 12 483 4/25/1986 Hatch-2 US BWR FW FWC 20 37 6/27/1985 KMK Mlheirm-Krlich DE PWR FW FWC 18 2598 12/29/1984 Krsko Sll PWR FW FWC 14 2446 5/6/1991 Kwusheng-2 TW BWR COND FWC 12 85 5/28/1990 Loviisa-l FI PWR FW FWC 12 76 2/251t993 Loviisa-2 FI PWR FW FWC 8 2928 6f14/1996 Manshan-21 TW PWR MS STEAM 16 20056 8/9/2004 Miha n-3 JP PWR FW FWC 20 1307 11/6/1991 Milistone-2 US PWR MSR STEAM 8 1320 8M81995 MiUstone-2 US PWR Heater-Drain FWC 8 500 6/23/1982 Oconee2 US PWR EXT-STEAM STEAM 24 865 t/1/1985 Ckonee-2 US PWR FW FWC 10 2701 9/24/1996 Ckonee-2 US PWR, MSR STEAM 18 504 9/17/1986 OConee-3 US PWR Heater-Drain FWC 10 976 6/10/1974 Quad Cities-2 US BWR FW FWC 18 2913 1/1/1989 Santa Maria de Gaona ES BW1R FW FWC 16 3092 2/9/1980 Santa Maria de Garona ES BWR EXT-STEAM STEAM 16 Karl N. Fleming ConsultingServices LLC Page 29 ~of 53

HELB Inftiating Event Frequenciesfor Kewaunee PRA DATABASE NOMINAL RCOR EVENT PLANT NAME COUNTRY LASYSTEM SYSTEM PIPE SIZE O. DATE TYPE GROUP jnh 2278 3/1/1993 Sequoyah-2 US PWR MS STEAM 10 541 10115/1983 Surry-1 US PWR FW FWC 26 542 12/9/1989 Surry-1 US PWR Heater-Drain FWC 10 595 12/9/1986 Sumrr-2 US PWR FW FWC 18 545 3/9/1985 Trojan US PWR FW FWC 14 920 12/2/1971 Turkey Point-3 US PWR MS STEAM. 12 Karl N. Fleming ConsultingServices LLC Page 30 of 53 C C C

f C C HELB Initiating Event Frequencies for Kewaunee PRA Table 4-7 Mean Values of Failure Rate and Rupture Frequency Parameters, Results - Mean Values Case Description Failure Rate Rupture Frequency (l/ft~yrj lI/t~Yrl KNPPO1 FWG EBSI with Post-1988 data 3.19E-06 .6.72E-08 KNPP02 FWC. EBS2 with Post-1988 dfat 3.56E-06 1.09E-07 KNPP03 FWC, EBSI with data through 1988 2.78E-05 5.85E-07 KNPP04 FWC; EBS2with data through 1988 3.98-05 1.22E-06 KNPP05 FWC EBS1 wit FAC events screened out 9.21E-07 8.60E-09 KNPP06 FWC; EBS2 with FAC events screened out 8.29E-07 635E.09 KNPPO7 _Steam Extraction pipng EBSI with Post-1988 data 3.40B-06 7.17E-08 KNPP08 Steam Extraction piping; EBS2 with Post-1988 data 2.58E-06 7.93E-08 KNPP09 Steam Extraction piping EBSI with data through 1988 3.32E-04 6.99E-06 KNPPIO Steam Extraction piping, EBS2 with data though 1988 4.86E-04 1A9E-05 KNPPI 1 Steam Extraction piping; EBSI with FAC events screened out 1.93E07 1.80E-09.

KNPP12 Steam Exinaction OipM= EBS2 with PAC events screned out 2.68E-07 2.07E-09 KNPP13 Steam pimg downstream P trbine, EBSI Post-1988 data 1.33E4-05 2.80E-07 KNPP14 Steampipng downsemnEPtbine, EBS2 Post-1988 data 1.07E45 3.29E-07 KNPP15 Steampiping downs-trmHPine, EBSI with datarough 1988 7.15SE5 1.51E-06 KNPP16 Steam piping downtm BHP trbine. EBS2 with data fthough 1988 9.09E-05 2.79E-06 KNPP17 Steam phpmgg downstream HP turbine, EBSI with PAC events sceaend out 2.2SB-07 2.10-09 KNPP18 Steam piping downstream HP turbine; EBS2 with FAC events seened out 9.22E-07 7.05E-09 KNPP19 MSpiping upstm HP trietvtle valve, EBSI Post-1988 data 3.25E-06 3.03E-08 KNPP2O MS piping upstream HP trbine thrwotle valve, EBS2 Post-1988 data 1.16E-06 8.90E-09 KNPP21 MS piping upstream HP turbine throttle valve, EBS I with data through 1988 1.60E-05 1.49E-07 KNPP22 MS piping upsteam HP trbine throttle valve, EBS2 with data tbrough 1988 2.50E-05 1.91E-07 KNPP23 MS pipin upstram HP turbine throttle valve, EBS1 with FAC events screened out 1.74E-07 1.64E-09 KNPP24 MS piping upstream HPturbine throttle valve, EBS2 with FAC events screened out 236E-07 I .80E-09 Notes:

• EBS - Equivalent (Diameter) Break Size

• EBSI: 2 < EBS S 6" equivalent diameter break size - moderate energy release

• EBS2: EBS > 6" equivalent diameter ba size - major energy release KarlN. Fleming ConsultingServices LLC Page 31 of 53

HELB Initiating Event Frequencies for Kewaunee PRA FY4CPr-1988 Data FWC Pot-1988Data FWCM m FAC Evuts mned Out Figure 4-4 Impact of Different Data Screening Assumptions on FWC Piping Reliability 1.E.O 6040

.A" 1.OE405 ' ... G- -,' -

I"I0 =

Cfarm I

Stoam Extracton Pro-1918Data Seanta Pot198ata SenExtraionl with FAC Events S'"uned Out Figure 4-5 Impact of Different Data Screening Assumptions on Steam Extraction Piping Reliability KarlN. Fleming ConsultingServices LLC -Page32 of 53

HELB Initiating Event Frequenciesfor Kewaunee PRA 5.HELB INITIATING EVENT FREQUENCIES 5.1 Calculation Steps The results for the initiating event frequencies were obtained using the equations in Section 3 and the data parameters developed in Section 4. The uncertainties were calculated using the technical approach described in Section 2 and is comprised of the following steps."'

1. A prior distribution for each failure rate was obtained from Reference [1]. The prior is a lognormal distribution with a mean value of 1.50x104 failures per foot of pipe with a range factor of 100. The same prior was used for all 24 cases inTable 2-1.

2. For each case listed inTable 2-1, Bayes' updates were performed using the prior from Step 1, the number of failures obtained from the PIPExp database for each case, and estimates of the piping population exposures that are documented in Section 4. Bayes' updates were performed using the program BARTTM developed by ERIN Engineering and Research, Inc.

3. To account for uncertainty in the population exposure estimates the Bayes' updates were performed for three estimates of the exposure: a best estimate with a probability weight of 80% and a high and low estimate with weights of 10% each.

4. A composite uncertainty distribution was developed for each of the 24 cases of failure rates using a posterior weighting procedure using Crystal BalT and Microsoft Excel.

5. The process listed in Steps 1-4 was repeated for two ranges of pipe size: one for pipes greater than or equal to 2", which could produce ruptures of size 2"and greater, and one for pipes sizes greater than 6" which could produce rupture sizes exceeding 6". Hence a total of 48 failure rate distributions were developed: one for 2" and greater, and one for 6" and greater pipe size ranges for each of the 24 cases in Table 2-1.

6. A Beta distribution was developed to represent the conditional probability of rupture for two rupture sizes: 2" to 6", and greater than 6" equivalent break size using the data described in Section 4. These beta distributions include prior distribution parameters that represent the authors expert judgment on the values of these probabilities, and service data experience that Is documented in Section 4. Two sets of distributions were developed: one for FAC sensitive carbon steel pipe In systems subject to FAC, and the other for FAC resistant pipe or systems that are not susceptible to FAC, e.g., the high-pressure main steam piping upstream of the turbine throttle valves.

7. The rupture frequencies for rupture sizes between 2"and 6"were obtained by combining the failure rates for 2" and greater pipes and the conditional rupture probabilities developed in Step 6. The rupture frequencies for greater than 6" breaks were obtained by combining the failure rates for greater than 6" pipe sizes with the appropriate conditional rupture probability.

B. The HELB-initiated internal flooding Initiating event frequencies were obtained by propagating the uncertainties in the appropriate rupture frequencies through the equations of Section 3 using the Monte Carlo process using Cystall Ballm and Microsoft Excel. To properly treat the state of knowledge dependencies all the uncertainty calculations from the output of the Bayes' updates through Step 8 were performed Ina single integrated Monte Carlo procedure. In each Monte Carlo trial a failure rate was sampled for each case and KarlN. Fleming ConsultingServices LLC Page 33 of 53

HELB InitiatingEvent Frequenciesfor Kewaunee PRA pipe size by sampling from either a high, best estimate or low exposure term estimate. A conditional rupture probability for each rupture mode was sampled for each pipe size, and a sample initiating event frequency was calculated by propagating these samples through the equations for the pipe rupture frequencies and the equations for the HELB-initiated internal flooding initiating event frequencies. This process also made it unnecessary to perform a series of Monte Carlo calculations In which the results from each step would be fitted to a distribution for sampling Inthe next stage.

5.2.Summary of Results The results for the Initiating event frequencies are summarized In Table 5-1 for each of the equations listed in Section 3. The results listed in bold font are the initiating event frequencies; the remaining values are key intermediate results.

In Figures 6-1 through 5-4 the details of the uncertainty analysis are provided for Large Feedline Breaks, Moderate Feedline Breaks, Large Steamline Breaks, and Moderate Steamline breaks, respectively using as Input reports that are generated by Crystal BallTm.

Table 5-1 UncertaInty Distribution Results for HELB-initlated Internal Flooding Initiating Event Frequencies Event Events per Reactor Operating Year Mean 5%t/le 50%tile 95%tfle Flip,Large High Pressure SLB 3.47E-05 1.50E-05 3.1 1E05 6.68E-05 FASL, Lage Reheat SLB 2.04E-04 9.82E-05 1.84E-04 3.85E-04 FESL, Large Extraction SLB 1.40E-05 4.96EM06 1.19E-05 3.00E-05 FSLRu Lare SLB 2.53E-04 1.42E-04 2.33E-04 4.37E-04 FRksm, Moderate Reheat SLB 1.74E-04 8.63E-05 1.57E-04 3.25E-04 FkSM, Moderate Extraction SLB 1.28E-05 5.32E-06 1.1IE-05 2.58E-05 FsLBL, Moderate Steam Line SLB 1.87E-05 9.84E-05 1.71E-04 337E-04 F>FL)S, Large FLB downstrUem of FWH15 5.85E-05 3.67E-05 5.52E-05 9.40E-05 Fazs,Large FLB between FWH14 and FWH15 7.67E-05 4.15E-05 7.01E05 1.42E-04 FFLBL, Large FLB 1.35E-4 8.19-05 1.26E&04 2.27E-04 FFUSM, Moderate FLB 4.69E-05 2A7E-05 4.29E-05 8.63E-05 Karl N. Fleming ConsultingServices LLC Page 34 of 53 U'

HELB Initiating Event Frequencies for Kewaunee PRA E&MM=

@ Et at FL4OL Large Foed line Break 0% 4A18-05 6% 19.E05 TOOO 10% 9.04E-5

.- - - . - - - - - - - - - - - - - - - - - - - 16% 9.85E-0 20% 1.02E4

- - - - - --- - - - - - - - - - - - - - - - - - - - 25% 16EM-04

- - - - - - - - - - - - - - - - - - - - - - - - - 30% 1.10E-04 I2000, 2Ooo Ir----- ------- 35%

40%

45%

00%

1.14E-04 1.18E-04 1.222-44 1.26E-44 IOW L

.111 _....ULJLLLJU hini ---- 65%

60%

1.30E-04 1.34E-04 WE-is06 IJI.-04 US7E-0a alOEs- 85% 139E.04 70% 1.44E44 75% 1.50E44 60% 1.68E-04 65% 1.69E-44 Td~as 100.000 9D% 1.87E44 Mean 1.35E-04 95% 227E.04 Median 1.28E-04 100% 4.90E44 Mode Standard Deviation 4.E0-05 Varlance 2.02E-9 Skewness 1.76 Kurtosts 10.7C Cooff. of Verebility 0.33 MiUmum 416E45 Masmum 4.90E44 Ra eWdth 449E-04 Mean Sid. Enrr IA2E-07 Figure 6.1 Crystal Ball Results for Large Feedline Break Frequency Eten= Forecast values FLE45M -Medium Feedline Break 0% 1 S2E-05 B% 2.472-05 10% 2.811-05 15% 3.05E-0O G000. ---- .- - . - - -- - - - - - - - -- - - - - - - - - - -I 20% 3.27E-05 4000, - - - - - - - - - - - - - - - - 25% 3.45E-05 30% S.62E-05 3000 - - - - - . . - . - - - - - - - 35% 3.79E-05 40% S.05E.05 2000W- 45% 4.11E-06 towo-1.11nuuE.0 all]

.W&

II 4.745.5.....-

4ira hill-1-1 IIUE tEEN AWUEE III . ........

ENuhe~e~**

--"- l 50%

65%

60%

65%

4"E-05 4.4SE-0 4.65E-05 4.AE4-05 70% 6.09E45 75% 5.38E-05

. 0% 6.70E-05 Sta~s: Eoreest valuesl 65% 6.17E-4 TM$al 100.00 90% 658E2-05 Moan 4.69E-45 95% 6.63E-05 Median 4.20145l 110% 2.16E-04 Mode _-

Standard Deviation 1.04E-05 Varlance 3.75E-10 Skewness 1.83 Kuftosis 11.10 Cooff. f Variability OAt Minimum 1.02E4-5 Madmum 2.15E-04 Range Wadth 2.E-04 Mean Sti Error S.12E-08 Figure 5-2 Crystal Ball Results for Intermediate Feedline Break Frequency Karl N. Fleming Consulting Services LLC Page 35 of 53

HELB Initiating Event Frequencies for Kewaunee PRM Parcentiesa Forecast Wiues FSLBL Large Steam Une Beak 0% S.87E-Cs 5% 1.42E-04 10% 1.5824 15% 1.70E-04 WmCe _ . . _ __ - _ .

20% 1.8OE-04 25% 1.9OE004 loo .. _ __- __ 0% 1.U8E44 1 wo __m--- I . _ __-- . __ 35% 2.07E-04 40% 2.15E-04 4S% 2.24E-04 f0% 233E-04 low '--1 .. ..., I SS%

80%

2A2E04 53E-04

?72E45 f.4244 2MrM6 3ASE-61 k.8624 65% 2e4E-04 70% 236E-04 756 2.91F-04 80% 3.09E-04 Stasics Forecast values 85% 3.33E-04 Tilels 100.000 90% A09E-04 Mean 2.83E-04 95% 4.37E-04 Median 2.33E44 100% 1.16E-03 Mode Standard Deviaton .61E-05 Vauiance 0.24E-0 Skewness 182 Kurtsis 11.77 Coe. of Vaieblly 038 Mininum 6.7E-05 Mawnum .16A 3 Range Width 1.10E-03 Mean Sld. Seror 3.042-07 Figure 5-3 Crystal Ball Results for Large Steam Une Break Percentiies- . -Forecadtyalues F-SLBM

• Medium Steam Line Break 0% 3.93E-05

.  % 9.84E-6000 10% 1.11E-04 8000 ------

1111 4coo 15%

20%

2S%

30-1.21E-04 1.29E-04 1.36E-04 U

400 2% 1A382-04 I:alMl am11 412050 000 111 1- ------ - - ---- - 35%

40%

45%

50%

1.S0E-04 1.572-04 1.642-44 1.71E-04 70% 2.04h-04 76% 2.16E-04 80% 2.30-E04 l lisbes: Fomiast y= 185% 2.49E-04 Tdals 100.000 90% 2.78E204 Mean 187E-04 96% 3.37E-04 Median 1.71E-04 100% 1.02E2-Mode Standard Deiattion 7.75E-06 Vadance 601E-09 Skewns 1.89 Kuotos s 1203 Coeff. f Variability 0.41 Minimum 3.932-05 Mandmumn 1.02E-03 RangeWidth 9.82204 Mean Std. Ermr 2.45E071 Figure 5-4 Crystal Ball Results for Intermediate Steam Line Break Karl N. Fleming ConsultingServices LLC Page 36 of 53 U

-ELB InitiatingEvent Frequenciesfor Kewaunee PRA 5.3 Sensitivity Study As a sensitivity study, the initiating event frequencies were recalculated using different assumptions regarding how the data was screened as discussed in Section 4. This study was performed by propagating the results for the pipe failure rates and rupture frequencies for the different data screening strategies through the equations for the initiating event frequencies In Section 3. The results are summarized InTable 5-2 and Figure 5-5. As seen Inthese exhibits, the impact of using the service data from 1988 to represent the current Industry practice and as a basis to predict the HELB-initiated internal flooding frequencies Is approximately an order of magnitude compared with the case of using pre-1988 data. This shows the Impact of Industry improvement programs, particularly the FAC programs, which were responsible for reducing the frequency of pipe breaks since about 1988. Although these programs were effective in reducing the pipe break frequencies, as seen in the third case in which all the FAC related failures since 1988 were removed, FAC Is still a dominant failure mechanism for these systems.

The initiating event frequencies would be an order of magnitude lower If all the FAC related failures were removed from the data analysis.

Table 5-2 Impact of Alternative Assumptions Regarding Data Screening on HELB-InItiated Internal Flooding initiatino Event Freauencies Mean Initiating Event Frequency per Reactor Operatin Year Initiating Event Base Case Data up to Data after Data after 1988 only 1988 with 1988 only FAC events removed FPi, Large High Pressure SLB 3.47E-05 3.01 E-04 3.04E-06 F Lar e Reheat SLB 2.04E-04 1.73E-03 4.38E-06 Fun, Large Extraction SLB 1.40E-05 2.6303 4.77E-07 Fs___L, Large SLB 2.53E-04 4.67E-03 7.09E"6 FJSM,, Moderate Reheat SLB 1.74E-04 9.39E-04 1.31-M06 F1 sm, Moderate Extraction SLB 1.28&05 1.23E-03 3.18E-07 FSLeL, Moderate Steam Line SLB 1.87E-05 2.17E-03 1.62E-06 FFLIs. Large FLB downstream of FWHIS 5.85E-05 5.98E-04 4.96E-06 Prls, Lare FLB between FWH-1 4 and FWHI5 7.67E-05 8.50E-04 4.42E-06 PFFLL, Lane FLB 1.35E-4 1.45E-03 9.38E-06 FFUSM, Moderate FLB 4.69E-O5 4.07E-04 4.29E-5 Karl N. Fleming ConsultingSetvices LLC -Palsy 37 of 53

HELB Initiating Event Frequencies for Kewaunee PRA lE_~" 1' - -- -

• FLBL - Large Feedline Break 1FLBM - Intemrediate Feedlne Brea J ESLBL - Large Steam Line Break lb

• SLBM - Intermediate Feedline Brea 1E.M _ ._

1.

ntil IV§iE0 E..

1E46 _-

I E-54B _

wl Post 1988 Data wl Pre-1988 Data wl FAC Events Screened Out Figure 5-5 Impact of Alternative Data Screening Regarding FAC Karl N. Fleming ConsultingServices LLC Page 38 of 53 U.

HELB Initiating Event Frequencies for Kewaunee PRA

6. REFERENCES

[11 Fleming, K.N. and B.O.Y. Lydell, "Pipe Rupture Frequencies for Internal Flooding PRAs", prepared by Karl N. Fleming Consulting Services LLC for EPRI, August 2005

[2] Poloski, J.P. et al, Rates of Initiating Events at U.S. Nuclear Power Plants, NUREGICR-6750, U.S. Nuclear Regulatory Commission, Washington (DC), 1999.

[3] Tregoning, R., L. Abramson and P. Scott, Estimating Loss-of-Coolant Accident (LOCA) Frequencies Through the Elicitation Process, Draft Report for Comment, NUREG-1829, U.S. Nuclear Regulatory Commission, Washington (DC), June 2005.

[4] Lydell, B.O.Y., 'PIPExpIPIPE-2004: Monthly Summary of Database Content (Status as of 31-Dec-2004)', RSA-R-2004-01.07, RSA Technologies, Fallbrook (CA).

Monthly summary reports have been Issued since January 1999.

[5] Fleming, K.N. et at, "Piping System Reliability and Failure Rate Estimation Models for Use In Risk-informed In-Service Inspection Applications", TR-110161 (EPRI Licensed Material), EPRI, Palo Alto (CA), 1998.

[6] U.S. Nuclear Regulatory Commission, Safety Evaluation Report-Related to 'Revised Risk-informed In-service Inspection Evaluation Procedure (EPRI TR-1 12657, Rev. B, July 1999, Washington (DC), 1999.

[7) H. Martz, TSA-1/99-164: "Final (Revised) Review of the EPRI-Proposed Markov Modeling/Bayesian Updating Methodology for Use in Risk-informed Inservice Inspection of Piping in Commercial Nuclear Power Plants,*, Los Alamos National Labaratory, June 1999

[8] Fleming, K.N., Mikschl, T.J., "Piping System Failure Rates and Rupture Frequencies for Use in Risk Informed In-service Inspection Applications", EPRI Report No. TR-111880 (EPRI Licensed Material), EPRI, Palo Alto (CA), 1999.

[9] Fleming, K.N. and B.O.Y. Lydell, "Database Development and Uncertainty Treatment for Estimating Pipe Failure Rates and Rupture Frequencies," Reliability Engineering and System Safety, 86:227-246, 2004 1101 Bush. S. et al, "Piping Failures In the US Nuclear Power Plants 1961-1995,m SKI Report 96:20, Swedish Nuclear Power Inspectorate, Stockholm (Sweden), January 1996

[III U.S. General Accounting Office, "Action Needed to Ensure That Utilities Monitor and Repair Pipe Damage,' GAOIRCED-88-73, Washington (DC), March 1988.

[12] Cragnolino, G., C. Czajkowski and W.J. Shack, Review of Erosion-Corrosion In Single-Phase Flows, NUREGICR-5156, U.S. Nuclear Regulatory Commission, Washington (DC), April 1988.

[13] International Atomic Energy Agency, "Corrosion and Erosion Aspects in Pressure Boundary Components of Light Water Reactors," Proceedings of a Specialists Meeting organized by the IAEA and Held in Vienna, 12-14 September 1988, IWG-RRCP-88-1, Vienna (Austria), April 1990.

Karl N. Fleming ConsultingServices LLC Pago 39 of 53

HELB Initiating Event Frequencies for Kewaunee PRA

[14] OECD Nuclear Energy Agency, "Specialist Meeting on Erosion and Corrosion of QJ Nuclear Power Plant Materials," NEAICSNI/R(94)26, Issy-les-Moulineaux (France),

1995.

[15] Beliczey, S. and H. Schulz, "The Probability of Leakage in Piping Systems of Pressurized Water Reactors on the Basis of Fracture Mechanics and Operating Experienced Nuclear Engineering and Design, 102:431-438, 1987.

Karl N. Fleming ConsultingServices LLC Page 40 of 53

HELB Initiating Event Frequencies for Kewaunee PRA APPENDIX A PIPExp DATABASE DESCRIPTION I

I I I 'I II I

Karl N. Fleming Consulting Services LLC Page 41 of 53

HELB Initiating Event Frequencies for Kewaunee PRA A.0 PIMExp I OPDE OVERVIEW This appendix describes the PIPExp database content and structure, and its relationship with the OECD Pipe Failure Data Exchange Project (OPDE). OPDE was established In 2002 as a cost-shared, multi-national co-operation in piping reliability. The initial objective of OPDE was to establish a comprehensive database on pipe failures in commercial nuclear power plants worldwide and to make the database available to project member organizations that provide data. The project Is operated under the umbrella of the OECD Nuclear Energy Agency (NEA). A Clearinghouse is operating the database and provides the quality assurance function. The Clearinghouse Is operated by one of the authors of this report A.1 Historical Background The Swedish Nuclear Power Inspectorate (SKI) in 1994 launched a R&D project with the objective of advancing the state-of-art in piping reliability. The stated objective Included the following tasks:

• Develop a high-quality, comprehensive database on the service history of piping systems in commercial nuclear power plants.

• In parallel with the database development, Identify and develop a general framework for statistical analysis of the service data as recorded In the pipe failure database.

• Perform a pilot application to demonstrate how the pipe failure database and piping reliability analysis framework can be used to develop plant-specific loss of coolant accident (LOCA) frequencies.

A long term strategy for the pipe failure database was formulated during the discussions leading up to the project initiation in mid-1994. This strategy included considerations to establish an international cooperation to support the long term database maintenance and applications program. The R&D project was concluded at the end of 1998. Results of the project included:

• A pipe failure database in Microsoft ACCESS. At the time this database was referred to as *SKI-PIPE", a proprietary database. It included 2291 pipe failure records as of 31-Dec-1998. This version formed the basis of OPDE in 2002 (Figure A-1).

• A series of technical reports (e.g., SKI Reports 95:58, 97:26, 97:32 and 98:30, all available from www.ski.se.

Independent of SKI and in preparation for and support of an International cooperative effort, the maintenance and update of the pipe failure database has continued post-1998. Figure A-1 is a top-level summary of this post-1998 maintenance and update program including the relationship between PIPExp and OPDE. Insights from practical database applications have played a significant role In enhancing and restructuring the database to become tool for piping reliability assessments.

Karl N. Fleming Consulting Services LLC Page 42 of 53

HELB Initiating Event Frequencies for Kewaunee PRA SKI R&D Proiect 1994-1998

• SOAR on piping reliabiity analysis as IKrelates to PSA (SKI Report 05:58)

• Basis for deriving pipe falluro parameters from service data (SKI Report 97:28;

• SKI-PIPE (1998) pipe failure database (2291 records as of 12-31.1998)

PIPExp DatabsePoject (1999 -to date) -independentof SKI

. Active maintenance program (wekl updates);

• QA program - extensive data validation;

• Practical applIcations &enhancements to db-structure 317 dbrecordS U i PIPExo2000 (24120001 OPD Peolect 002 8.

• 3879 db records

• Based en SKIPPE (1998);

• Validation of selected records by

___PEX__00_M2_31_200_ Natlonal Coordinators;

• 3957dbrecords *9 Harmonized db-structure;

• Coding Guideline I OA Program

_ PPEgg-2002 U24312002)7 J

• 4215db records / o 3g1241.003F

___________ _a 2427 dbrecords l 43 db* records / Oi201 It 084103 a* db records PMExO.205s941-2005)

,* 6395 db records (pipe)

• 250 db records (non-plpe.

passive Code Class I &2 components, excl. SG tube)

• 463 water hammer records (wlo structural Figure A-1 Evolution of PIPExp Pipe Failure Database A.2 PIPExp Quality Management All work associated with database maintenance Is controlled by a QA program. Source information Including text files, drawings and photographs associated with each database record is stored in an electronic archive. Each data record In PIPExp Is assigned a "Quality Index" (or completeness Index) per the definitions in Table A-1. The Quality Index is used to! assess the completeness and technical accuracy of the source information as well as the classified and coded Information In the database. Table A-2 summarizes the evolution of the database since 1998.

KarlN. Fleming Consulting Services LLC Page 43 of 53

HELB InitiatingEvent Frequenciesfor Kewaunee PRA Table A-1 Table A-2 Definition o Qualitv Index for Database Management Database Content by Quality Index Quality-Index Definition Database as of 12-31-1998 I Validated - all source data has been No. Pipe Failure Records b Quality Index accessed & reviewed - no further action Plant Type Totals 1 2 3 4 5 6 required BWR 673 210 66 3 74 7 277 2 Validated - source data may be missing PHWR 100 30 3 - 56 1 10 some, non-critical Information - no further PWR 1376 386 123 6 152 84 746 action anticipated RBMK 57 3 6 - 19 28 1 3 Validated - incomplete source data - 2291 629 198 9 301 120 1034 assumptions made about material grade Database as of 12-31-2002 andlor exact flaw location - no further action No. Pipe Failure Records byQuality Index anticipated Plant Type Totals 1 2 3 -4 5 6 4 Validation based on incomplete Information - BWR 1872 1216 174 12 219 75 176 depending on application requirements, PHWR 106 51 2 - 42 11 -

further action may be necessary PWR 2077 1011 198 6 351 233 278 5 Validation based on available, incomplete RBMK 160 48 - _ 18 81 -

information - further action expected (e.g., 4215 2290 379 22 721 349 454 retrieval of additional source data) 6 Not validated - validation is pending, or Database as of 09-30-2005 record Is subject to deletion from database No. Ppe Failure Records by Quaity Index Plant Type Totals 1 2 3 4 l 5 6 BWR 2510 1489 300 172 282 204 63 GCR, 12 8 _ 2 1 1 -

HWLWR PHWR 131 47 4 23 42 15 -

PWR 3563 1318 323 300 453 1070 99 RBMK 179 12 -21 4 110 32 -

6395 2874 648 501 888 1322 162 Karl N. Fleming ConsultingServices LLC Page 44 of 53 C C C

HELB Initiating Event Frequencies for Kewaunee PRA A.3 PIPExp Database Input Forms This section gives and overview of the database Input requirements. All data entry is done via the four forms (Form I through F&M,4).'

A.3.1 Form I - Event Descriptions Form I is shown In Figure A-2. It consists of 35 fields; seven of which are free-format with the balance defined by roll-down menus with key words (or data filters). The data entry requirements are defined below:

p. p.,. . MI L.~ im5eanssel  ! .u iz ~ .i rj * .

.IEle KM~ sew own rtm*Pnn Remors lods Mnow U*%)a Tusay cobrO, 04 PIPF-xpDATABASE _____

EiD MP~ Evet Repodt EvetaeN~ Pk"~t atfoelte 048 4 L"! rg2 K0ebC*2dShpidown

• eec*Reference -swcary ves l5IIA4EA, Pailaplnoi iIR aoteti____________

Rceieenco*- Tertkoy Reference.quatlaw inyee

__ _ ___ __ ___ _em

_ _ _ _ De a d ColltetadDamage I CocarctMActon Imata k Operationm TTR ITTRt.Oe Evera Hatative, Gw*RdoaReed Leak Rite sel -SYSTEM Dwhe plnwwkon fteWqatr2001 "age ofKoebergr2 &Ak=ofd 1 I___

k&Age t Safey as foud I~con anvj~m ixf wl~ cnnaem Grow onupm~nn h Sto Wthe kw4iead aalewi ectionando GO u.=1aypump. FUthe iR , vpp oketaw!saface dye-ponetant and meta ,atipaod istons cowfmad SCCat a S .CdOmIDW nubefc bcatmn - "~ to iesklual lomio stesm InOaplpinog cm edM9 krvvtuhd olacamdotm The flaws WAun during theW oftea waere VA.otApp~ioWj 2 lo stuafle tank mam foetayi*ed envionmentalcoeio cngafteWsrtemm WC1 mt pngweettompnertratibn ar ent scaled,.euigi waa~ aiWn______lies see hornk oom othFuel rDulf AD O p~pe work =Wsdered at dsk is304 areentcatalenu~

steel Hi..-Ted ptra work itme entonon-amnnated satae. 9~WG_ pdmau Deihe 8Reluetr Oftae .mafr I 2001. trza p~pcbabo and one straigNr sechion A3 TP30vi oatdW 7F dppwasolzdAVmuinspeinpm wasvutplenebt=kxthe ther ppowppstCas W~s Aovx at 'wrat uhIspcin Ove otla kkOmperating cycle. AnOpnou iedaeevawluetinhereof eranimd th defe*t mmtof fthsome Mtase rawrexperienced enetwere bked in uae to dug tohenctiond capabfly aflaikeaio piewas ri Impaired Evidence of 9CC has ualsoabeen foundin Ihe RWST. In dweisancoe Ome cracks ONoh-t Ira Sat ri iuctsibqp wldfro.os 1Rwia utr 2614l[' Ill* H -Of MS'

'Figure A-2 Event Descriptions - Form I Form I Data Entry Recuirements

• EID (Event ID) is a uniquely defined database record number (or *prmary key"); it is generated automatically by Access.

• Multiple Event Report is checked if one source document (reference) includes Information about more than one pipe failure and at different piping system locations.

Karl N. Fleming ConsultingServices LLC -Page45 of 53

HELB Initiating Event Frequencies for Kewaunee PRA Mainly, this field supports database management activities (e.g., answer to question

• have all pipe failures been adequately recorded in PIPExp?").

• Quality Index (a number 1 to 6); a roll-down menu defines the different options together with definitions.

• Event Date Is always required.

• Plant Name; a roll-down menu with listing of all commercial nuclear power plants in NEA member and non-member countries.

• Plant Operational State; a roll-down menu defines the different options.

• Reference; there are four free-format fields for primary and supplemental references.

Electronic copies of each reference are stored on CD.

• Event Type; a roll-down menu defines the different options.

• Event Category; a roll-down menu defines the different options.

• Collateral Damage; a roll-down menu defines the different options. 'NIA - None' is used as the default.

• Corrective Action; a roll-down menu defines the different options. Note that the term

'Temporary Repair' always implies that a 'Code Repair' or 'Replacement' be performed during the next scheduled outage lasting 30 days or more, but no later than the next refueling outage.

• TTR (lime to Repair) Is for the repair time in hours.

• TT-Class Is a data filter, a roll-down menu defines the different options with definitions.

• Event Narrative Is a free-format memo field.

• Quantity Released Is free format field; the dimension can be [lb], [kg], [ton], or [m3].

• Leak Rate Class Is a data filter; a roll-down menu defines the different options with definitions.

• System is a free format field for the system name; a roll-down menu includes a selection of BWR- and PWR-speclfic, English language names.

• System Group Is a data filter, a roll-down menu defines the different options.

• Piping Component is a data filter, a roll-down menu defines the different options.

• Weld Configuration; a roll-down menu defines the different options.

• Code Class; a roll-down menu defines the different options. A cross-reference table compares the different national safety classifications with ASME Section Ill.

• Diameter Class Is a data filter; a roll-down menu defines the different options and definitions.

• Diameter [mm] is used for the measured diameter.

• Diameter [inch] is used for the measured diameter.

• Material is a data filter, a roll-down menu defines the different options.

• Material Designation; a roll-down menu defines the different options. A cross-reference includes different carbon steel and stainless steal material designations.

• Process Medium, a roll-down menu defines the different options.

• ISI History (Form 3) is checked only if information is available.

• Root Cause Information (Form 4) is checked only if information is available.

• Flaw Size Information (Form 2) Is checked only if flaw size (e.g., crack orientation, depth, length) information is available.

Karl N. Fleming ConsultingSetvices LLC Page 46 of 53 L/

HELB InitiatingEvent Frequenciesfor Kewaunee PRA A.3.2 Form 2 - Flaw Size Information Form 2 is shown in Figure A-3. It consists of 28 fields. The data entry requirements are defined below:

UIVFIXII1 P.VW I=F In, U IUXi!N sarosew a;

~' 5.rdm!o

• De W at Pbmat: leods 10*l S&idow U*Tp qusiwo t -e aJ Ttx~dwj,Odtober W.2DC4 4A2-23PFM PIPExp DATABASE I r__ E____ '_ ForFAC*Orducd degiadatim a y.pMe sosoi fIinrad Upr~mtiwo.IFcpt~ eet.m M prem 408 .q&Avatmt hasd w fan~

t. MX p*1GI~k ~d=k=arrnn*u Rlanaa be o ihitr~csaie ai~

Hoa inpipe aw = ~ uy3/8tch dra o14ndi ametai DO-1I CF1 01.2 CV2 1 02-3 C.O D3.4 CF 0 0 0 0 0 0 .- I I I 4D I CF5 N I5- 6F 067 I 07-8 CFO DO-B I C9 I 0D910I CF10 I 0 U. 0 01 a I0 U I U I 0 Ra66 of Cmck LuiVh to ECkaxlwerecpdoato Record: 1I4141 r 2545 iI PtII* of2645 WW:tio of cracdmetha) to flowlergd(L)

Figure A-3 Flaw Size Information - Form 2 Form 2 Data Entry Reauirements Flaw Description Is a free-format memo field. For through-wall flaws, information about dimensions (e.g., equivalent diameter) should be included in this field. For part through-wall flaws, this field should include information onW flaw depth (a) and length (O,and otientation. For multiple flaws, the number of flaws and their lengths are recorded in the designated fields.

• Check If Multiple Circumferential Flaws. This check box typically applies to flaws attributed to IGSCC. In PIPExp, on the order of 15% of the records on IGSCC involve multiple, single plane circumferential cracks.

• nCF (number of Circumferential Flaws) includes the total number of flaws in an affected weld.

• D## Is the distance, in [mm], between adjacent circumferential flaws; e.g., DO-1 is the distance from the TDC (12 o'clock) position to flaw #1, and D2-3 is the distance between Karl N. Fleming ConsultingServices LLC Page 47 of 53

HELB Initiating Event Frequencies for Kewaunee PRA flaw #2 and flaw #3, etc. A blank field Indicates that no information on the spacing is available in the database.

• CF-# is the length of circumferential flaw V' [mm]. The flaw number is relative to the 0-degree position; CF-1 is the first circumferential flaw from the reference position, etc.

• Crack Depth f%] is the ratio of crack depth to pipe wall thickness.

• Axial Length [mm]; this field relates to the Flaw Description.

• Ratio of Crack Length to Circumference; this ratio should be relative to the inside pipe circumference.

• Aspect Ratio; this is the ratio of crack depth to crack length and relates to the information under Flaw Description.

A.3.3 Form 3 - ISI History Form 3 is shown in Figuie4-4. It consists of 3 fields. While primarily intended for [SI program weaknesses, the free-formiat field may be- used to document any information pertaining to the ISI of the affected component, or ISI history such as time of most recent Inspection.

Oft sft. tew V-p re* Igmt' Iecd Bos pidow

• _

-r F X Tu=4~dy. DctWeM8. 2004 4:336tPM PlPExp DATABASE r-1-10

-e1 ihfed so 191Vwouam 1WW -

MECO pedonmUdaciv.eniTed byac6aonk lo onSW pi hting da chait o e Oawed li The npecton did ret malaW aUw dewa~dde. The wAtown ho jenq lods tat nraog a lot bdeshciss12haw .hJL-Reackw& I*4 1 2374 ' 1 Idl 12645 of Figure A-4 ISI History- Form 3 KarlN. FlemingConsulting Services LLC Page 48 of 53

HELB Initiating Event Frequencies for Kewaunee PRA A.3A Form 4 - Root Cause Information Form 4 is shown in Figure A-5. It consists of 9 fields. The data entry requirements are defined below:

.,  : 13 _ _ I. ,U fat 0eawknertetVgM* secirdsIpk W0dow tietp Tg.ot5 iIstknh*e

• _ X y Ocoe 2D-, .w .,, ,

4.3813PM PlPExp DATABASE I EID 1 4070 ..

Location Fae -FenJ A"LwPieswLP WBa)aweciandonsendowtreaimohe TIbiria Wft FAC- FlwAcceiewate 1msion HD OeaPump 119D4*2ALLV; beween etownd vikallffo a f Debdbn l ndwii Ca1*t

. UT thd Fabdcabo Unde4i Ca2e2 Rom E-Andydsi ThMpipe lccired d a dow waaofblw oo*a valw Twaace hcdsaewar now roflstA valc lcresed 1. .

irate d p4owea decsear the vaDUicess wl lo poiS wher pae Wame occuid Th. am awe t e oed wes not preftowfr

• inmeded hx vodn/corosicn as pa tdltSeconr gmiandCConnen Ipecfon Pwogan The filed pipe scidn wasa sciben donseadmof ecnol vadve andtwe n teiedwn hMpedctionkgownidr Aneibow=&te*nda k ebrmelafaedpipe tad piasvbu been irnspected nd a*hoinch wds cirmsdanet band at be dowubeamendof Oefilhd section btA i had not Wwnned lo thpoel vir U Inspection of Me dpedapov was MCukedd Idnwsby opwaf oene Wounadon rcei"ed in1987 nd 19S8hdceed e d sr1h gt poe seakm donIrem corir vau vetf a b wiion wsm grater tnhpwt 'beahv wehr., Ihk infomalon wm ad audelydssemfnaoed tiervrdior n Wn humen dami Cs Record:j411 263S iijl*l efz2645 Poot cewega) o wcrft~ dwsdeslofwdaruscm) A Figure A-5 Root Cause Information - Form 4 FoIm 4 Data Entry Reauirements

• Location of Failure; this is a free-format mnemo fietd describing the location of a flaw (e.g.,

line or weld number, or using a P&ID reference).

• Plant Location; a roll-down menu defines the different options.

• Method of Detection; a roll-down menu defines the different options.

• Method of Fabrication; a free-format text field.

• Apparent Cause; a roll-down menu defines the different options. Normally this field has already been filled in.

• Underlying Cause -1; a roll-down menu defines possible contributing factors.

• Underlying Cause- 2; a roll-down menu defines possible contributing factors.

• Root Cause Analysis; a free-formatimemo field. This field should Include any relevant information on the cause-consequence relationship and should be supplemental to the Event Narrative InForm I.

Karl N. Fleming ConsultingServices LLC Page 49 of 53-

HELB Initiating Event Frequencies for Kewaunee PRA

• Comments: a free-format memo field. It is intended for any other, relevant information that is not captured by other database fields.

A.4 Database Accessibility PIPExp is a proprietary database whereas the OPDE database is restricted. The full OPDE database is available to participating organizations that supply data. An unrestricted version of OPDE ('OPDE-Light') is available to Interested parties upon request to respective National Coordinator (the U.S. representative in the project is the Nuclear Regulatory Agency, Office of Nuclear Regulatory Research). OPDE-Light does not include any proprietary information or any information that enables the Identification of plant name.

KarlN. Fleming Consulting Services LLC Page 50 of 53

HELB Initiating Event Frequencies for Kewaunee PRA APPENDIX B PIPE FAILURE RATES & RUPTURE FREQUENCIES APPLICABLE TO NON-CODE PIPING SYSTEMS Karl N. FlemingConsulting Services LLC Page 51 of 53

HELB Initiating Event Frequencies for Kewaunee PRA Table B-1 FWC Piping Failure Rates & Rupture Frequences

_Uncertnty Distribution Cose Descripfton Mean 95

_IlM.Yrl Percentile MedPn Percentile EBSI - FWC Pipe Failur Rate - with post 1988 data 3.191E-06 1.99E-06 2.97E-06 5.92E-06 KNPPOI EBSI - FWC Pipe Ruptwe - with post 1988 data 6.72E-08 3.80E-08 6.19E-8 1.24E-07 KEBS2 - FWC Pipe Failur Rat - with data through 1988 3.56E-06 2.21E-06 3.311E-06 659EM06 EBS2 - FWC Pipe Rupte -with poM 1988 data 1.09E1-07 6.25E-08 1.01E-07 2.00E-07 EBSI - FWC Pipe Failure Rate - with data dmuth 1988 2.78E-05 1.73E-05 2.60E1-5 5.20E-05 KEPP03 EBSI - FWC Pipe Rupore -with data though 1988 5.85E-07 3.39E-07 5AIE-07 1.09E-06 EBS2 - FWC Pipe Fafiure Rate - with data trough 1988 3.98E-05 2.49E&05 3.73E-05 7.45E-05 CNPPO4 EBS2 - FWC Pipe Rupture - with data through 1988 1221E-06 7.27E-07 1.14E306 2.28E-06 KENPP0 EBSI - FWC EPpe Faihne Rate - wvth FAC events sreened out 9.21E-07 5.23E-07 8.45E07 1.70E-06 EBSI - FWC Pipe Rupture - with FAC event screened out 8.60E-09 3.64E-09 7.66E-09 1.68E-08 KNPPo6 EBS2 - FWC Pipe Failue Rate - with FAC events screened out 8.29E-07 4.41E-07 7.56E-07 1.52E-06

_ EBS2 - FWC Pipe Rupture - with FAC events screned out 6.35B-09 2AOE-09 5.55E-09 1.30E-08 Table B-2 Steam Extraction Piping Failure Rates & Rupture Frequencies Uncertant y Dstribution _

Case Descrpton Mean a

_/iftyrl Percentile Median Percentile EBSI - Steam Extaction Pipe Failure Rate with post 1988 daa 33AOE-06 1.65E-06 3.06E-06 6.41E-06 EBS1 - Steam Exraction Pipe Rupture with post 1988 data 7.71B-08 3.19E-08 637E-08 1.39E-07 KNPP08 EBS2 - Steam Extracion Pipe Failure Rate with post 1988 data 2.58E-06 1.OOE-06 2.23E-06 5.3113-06 EBS2 - Steam Extraction Pipe Rupture with post 1988 data 7.93E-08 2.89E-08 6.75E-08 1.68E-07 EBSI - Steam Extraction Pipe Failure Rate with data through 1988 3.32E-04 2.06E-04 3.1OB-04 6.17E-04 EBSI - Steam Extraction Pipe Rupture with data through 1988 6.99E&06 4.03E-06 6.45E-06 1.28E,05 NPP EBS2-Steam Extraction P Failue Rate with datathrough 1988 4.86E-04 3.03E-04 4.55E-04 9.05E-04 EBS2-SteamExtrcton Ru ewith datatou 1988 1.49E-05 8.78E-06 1.38E-05 2.73E-05 EBSI - Steam Extraction Pipe Failure Rate - with FAC events screened out 1.93E-07 1.32E-008 9.36E-08 6.73E-07 KNPPI I EBSI - Steamn Extraction Pipe Rupture -with FAC events screened out 1.80E-09 1.IOE-10 82SE-10 6.45E-09 EBS2 - Steam Extracton Pipe Failure Raft - with FAC events sceened out 2.68E-07 1.64E-08 1.23E-07 9.58E-07 KNPP12 EBS2 - Steam Eitraction Pipe Rupture - with FAC events screened out 2.07E-09 1.05E-10 8.81E-10 7.63E-09 Karl N. Fleming ConsultingServces LLC Page 52 of 53 C C (7

C C H L I t-g e Fre que nc Ie -o -e  :

IfEI B In~aiting Event Frequencdes for Kewaunee PRA Table B-3 LP Steam Pipin Failure Rates & Ru tUre Frequencies Uncertainty Distribution Case Description Mean SO 9 P I /ftyrl Percentile Median Percentie KNEBSI - LP Steam PI1ing Faile Rate - with post 1988 data 1.33E-05 7.89E-06 1.23E-05 2.42EBO5 KEBSI - LP Steam Piping Rupture - with post 1988 data 2.80E-07 1.473-07 2.56E-07 5.14E-07 KNPP14 EBS2 - LP Steam Piping Faiure Rate - with post 1988 data 1.071S05 5.82E-06 9.75E-06 1.96E-05 EBS2 - LP Steam Piping Rupm - with post 1988 data 3.29E-07 1.64E-07 2.97E-07 6.13E-07 KNPPIS EBSI - LP Steam Piping Faihlre Rate - with data throigh 1988 7.15E-05 4A5E-05 6.66E-05 1.33E-04 EBS1 -LP Steam Piping Ruftre-with data through 1988 1.51E-06 8A9E-07 1.39E-06 2.77E-06 KNPP16 EBS2 - LP Steam Piping Falue Rate - with data t__uh 1988 9.09E-05 5.66ES05 8.45E-O5 1.68E-04 PEBS2 - LP Steam PiPing Rte - with data through 1988 2.79E-06 1.60E-06 2.57E-06 5.07R-06 KNPP17 EBS I - LP Steam Piping Faflure Rate - with FAC events screened out 2.25E-07 1.47E-08 1.07E-07 7.87E-07 EBSI - LP Steam Piping Rupture - with FAC events screened out 2.1OE-09 1.19E-10 9.44E-10 7.48E-09 KNPP8 EBS2 - LP Steam Piping Failure Rate - with FAC events screened out 9.22E-07 1.78E-07 658E-07 2.52E-06 EBS2 - LP Steam Piping Rptne - with FAC events screened out 7.0E509 1.1 IE-09 4.76E-09 204E-08 Table B-4 HP Steam Piping Failure Rates & Ru ture Frequencies Uncertainty Distribution Case Description Mean 5f 95P Il/ft.yrl PerC Percentile KNPPl9 EBSI -BPSteamPipingFaihireRate-withpost 1988 dat 3.251-06 1.62E-06' 2.94E-06 6.01E-06

-EBSI-HP Steam Piping Rupture-with post 1988 data 3.03E-08 1.16E-08 2.64E-08 6.28E-08 KNPP20 E;BS2 -HP Steam Pipin_ Failue Rate - with _2st 1988 data 1.16E-06 3.33E-7 9.37E-07 2.751E-06 EBS2 - BP Steam Piping Rpte - with post 1988 data 8.90E-09 2.01E-09 6.78E-09 2.26E-08 EBS1 - HP Steam Piping Failue Rate - with data rough 1988 1.601i05 9.34E-06 1.47E-05 2.94E-05 KNPP2 EiBSl - HP Steam Piping Ruptre - with data through 1988 1.49E-07 6.40E-08 1341-07 2.90O07 KNPP22 EBS2 - HP Steam Piping Faihure Rate - with data through 1988 2.50E-05 IA7E-05 2.30E-05 4.60E-05 EBS2 - HP Steam Pipn R ture - with datathugh 1988 1.91E-07 7.72E-08 1.70E-07 3.78E-07 EBSI - HP Steam Piping Faihlre Rate -with FAC events screened out 1.74E-07 1.23E-08 8.44E-08 5.93E-07 KNPP23 EBSI - HP Steam Piping Rup -with FAC events screened out 1.64E-09 9.98E-1 7.52E-10 5.71E-09 EBS2 - HP Steam Piping Failure Rate - with FAC events sceened out 236E-07 1.53E-08 I .12E-07 8.29E-07 KNPP24 EBS2 - HP Steam Piping Rupure - with FAC events screened out I.80E-09 9.99E-11 8.01E-10 6.49E-09 Karl N. Fleming Consulting Services LLC Page 53 of 53

Appendix B Flood Area Definition for Turbine Building Basement I

IINTERNAL FLOODING - Flood Area Definition for Turbine Building Basement P. I 11 p)

Flood Area Definition for Turbine Building Basement Prepared by: S 44w-e& ULP- eMA-

.5 Signature Print Name I-L b-i Ib Date L)

Reviewed by: T T I "Z4tEL UI.- _OA Signature IPrint Name LI Q~

t-21 6-Date I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 2 1.0 PURPOSE Internal floods are defined as those floods that result from the failure, incorrect operation (including errors in maintenance), or incorrect alignment of components within the plant.

Accident sequences initiated by internal floods can be a significant contributor to risk because of the potential of the event to impair, simultaneously, multiple components required for accident mitigation. The overall objective of the internal flooding analysis is to determine the contribution of accident sequences initiated by such flooding events to core damage and individual accident class frequencies.

An internal flooding PRA requires that areas of the plant be identified that contain equipment needed to mitigate accidents and that are subject to flooding effects. Areas are defined as separate for flooding purposes where physical boundaries are present that prevent propagation of a flood source in one area from causing damage to equipment in another area. For each flood area, flooding sources are identified in the area that have the potential to damage equipment within the area or that have the potential to propagate from the area to another area and damage equipment needed for accident mitigation. Propagation paths are identified and defined between flood areas. The flooding walkdowns confirmed the boundaries between flood areas and identified barriers in the boundaries that separate the flood areas. This information can be used to identify areas of the plant that can be designated as separate, independent flood areas.

In order to streamline the accident sequence analysis it is beneficial to limit the analysis to only those sequences that will contribute to flooding risk. Such risk-significant sequences are identified through the application of a screening process. Each flood area is evaluated against important factors including the existence of flooding initiators, the existence of safety-significant equipment, and ability of a flooding initiator to cause a reactor trip to determine which flood areas are worthy of additional analysis and quantification.

This document designates independent flood areas that will be analyzed in further detail through accident sequence analysis and initiating event frequency analysis. Development of independent flood areas will support the high level requirements identified in the Flood Area Definition Guideline [GUIDEO1]. This document also applies a screening analysis to all the defined flood areas to focus the accident sequence analysis on risk-significant scenarios. Application of a screening process will support some of the high level requirements identified in the Accident:

Sequence Analysis Guideline [GUIDE02].

2.0 METHODOLOGY 2.1 Flood Area Definition A flood area addresses physical boundaries that impact the propagation of water and the potential to damage equipment. The subject water originates from a pipe break in the flood area and is

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 3 l categorized as a submergence event or a spray event, thus flood areas are defined differently for these two events. A submergence event is defined as a pipe break with sufficient flow rate to overwhelm the flood area's flood mitigation equipment, accumulate in such a manner as to damage equipment, and has the potential to propagate laterally in an amount significant enough to damage additional equipment. A spray event is a pipe break with a flow rate within the capacity of the flood area's mitigation equipment (especially floor drains) that can damage equipment from direct spray, but cannot propagate laterally. Thus, a pipe break in a spray flood area is expected to result in direct damage to equipment in the area and to result in the vast majority of that water exiting the flood area via the floor drains and floor openings. Any water that propagates laterally from a spray flood area will be of insignificant quantity and will not cause equipment damage in adjoining areas. Each flood area is analyzed for both spray events and submergence events.

Lateral propagation for submergence events occurs due to lack or failure of barriers separating the areas. Typical barrier failures are of doors, but could include water that flows through open penetrations through walls, water that flows over protective curbs and weirs, water that backflows through drain lines, and structural failure of gypsum walls. Normally closed access doors are able to withstand some amount of force due to accumulated water, however when the water level reaches a critical depth the door is expected to fail. Thus, water will propagate laterally through such failed doors. Since a failure is required for this lateral propagation to occur, these lateral zones are not included in the flood area definition. Only zones with open communication are considered in defining flood areas.

2.2 Assumptions The following assumptions were utilized in the definition of flood areas:

1. Leakage under and around doors is the only form of drainage inside Safeguards Alley since a pipe break in the Turbine Hall will fill the Turbine Building sump and subsequently fill the drain lines connecting the sump to the floor drains in Safeguards Alley. Thus, the floor drains will not be able to remove water from Safeguards Alley. Gaps under doors are documented as part of the GOTHIC input. [CALC02]

2. KPS access doors inside the Turbine Building generally can withstand a water height of 4 feet when the water is pushing the door open and 5 feet when the water is pushing the door closed. Exceptions to this include doors 243 and 244 which can withstand 3 feet 3 inches when water is opening the door and 4 feet 9 inches when water is closing the door, and door 8 which can withstand 3 feet 9 inches when water is opening the door. [CALCO1]

3. All junction boxes are gasketed and not vulnerable to spray unless otherwise noted in the Walkdown Sheets [FLOOD01].

4. Flood-induced failure of motor-operated valves (MOVs) involves the valve operator's loss of

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement I P. 4 function, but does not involve the MOV changing position. The MOV is expected to remain in the same position, however any new change in position will require manual action to turn the handwheel.

5. Flood-induced failure of air-operated valves (AOVs) involves the valve operator's loss of function, but would also involve the AOV failing to its fail-safe position.

6. Sealed penetrations are assumed to pass no fluid. Penetrations make use of various types of sealant including grout and elastomer. Grout behaves similar to concrete and is basically impervious to water.

7. Cable insulation is not subject to failure from submergence or spray.

8. Walls and trench barriers are assumed to remain intact throughout a flooding event with the exception of the firewall separating flood areas TU-95A and TU-95B-1. This gypsum wall was analyzed and determined to be structurally capable of withstanding only approximately 3 feet of water. [CALC01]

9. The probability of rupture of encapsulated high-energy lines is insignificant as both the inner pipe and the surrounding guard pipe must fail.

10. Environmentally qualified (EQ) components are assumed to be able to perform their safety functions when exposed to spray conditions and high heat and humidity due to a pipe break.

For example, the solenoid operators for feedwater valves in the feedwater valve room, by design, perform their safety functions during high-energy line break (HELB) events.

11. Lines that are not normally pressurized or charged such as drain lines and dry fire protection piping are not considered as credible flood or spray sources.

12. Flooding in containment is not considered in this analysis. This is a subset of the Loss of Coolant Initiating Event (LOCA) in the Internal Events PRA.

13. Rupture of seismic Class I tanks (e.g., concrete reinforced refueling water storage tank) is not considered credible in this analysis.

14. Failure of a fire protection deluge valve is not analyzed as a potential initiator in this analysis.

The flow rate of a single deluge valve is insufficient to cause flooding concerns in Safeguards Alley. The simultaneous failure of multiple deluge valves has an insignificantly small probability.

3.0 DESCREMON OF FLOOD AREAS IN TURBINE BUILDING BASEMENT I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p.5 Flood areas were defined in a previous analysis [FLOOD01]. Walkdowns of the various flood areas were performed earlier and are documented on walkdown forms [Appendix C of FLOOD01]. For each flood area these walkdowns recorded information that included resident equipment, flood sources, and barriers (including doors). The information from these walkdowns combined with the information obtained from the general arrangement drawings [DWGO1]

provides the basis for the following flood area descriptions.

This analysis is concerned only with flood events that originate in the Turbine Building and then propagate to Safeguards Alley. Therefore, flood areas associated with the Battery Rooms on the mezzanine level of the Turbine Building and the Turbine Oil Storage Area in the Turbine Building basement as well as flood areas in the Auxiliary Building are disregarded as flooding areas of interest in this analysis.

Figure 1 identifies the various flood areas and their important features.

TU-22-1 Description - Flood Zone TU-22-1 comprises all of the general areas of the Turbine Building including the Operating Deck on the 626'-O" elevation, the Mezzanine Floor on the 606'-O" elevation, and the Basement on the 586'-O" elevation. It also includes the rooms in the south end of the Auxiliary Building basement from the waste neutralizer tank (room 17B) west to the Reactor Building Support Ring (room 1IB) since these rooms are not part of the radiological area and communicate openly with the Turbine Building basement. Additionally, the shop area, working material storage area, and steam generator blowdown area in the south end of the 606'-

0" elevation of the Auxiliary Building are also included in this flood zone since these rooms communicate openly with the other Auxiliary Building rooms in Flood zone TU-22-1 and since these areas are not part of the radiologically controlled area. Table 1 contains a complete listing of the fire zones and room numbers that comprise each flood zone. On the 626'-O" elevation the zone is bounded on the north by the Technical Support Center and exterior walls, on the south by the Transformer Area and exterior walls, on the west by the Auxiliary Building, and on the east by the Administration Building and exterior walls. On the 606'-0" elevation the zone is bounded on the north by zones TU-97 and TU-98, the Technical Support Center, exterior walls, the Containment Building, and zone AX-32-1, on the south by the Transformer Area, exterior walls, and zones AX-33 and AX-39, on the west by the Auxiliary Building and Containment Building, and on the east by the Administration Building and exterior walls. Zones TU-94, TU-95A, TU-95B-1, TU-95B-2, TU-95C, TU-96, TU-97, and TU-98 lie beneath zone TU-22-1 and zone TU-96 and the Turbine Building roof lie above.

All wall and ceiling penetrations are sealed. The floor of this zone is the basement floor and is finished concrete. The north wall has a normally-closed door (120) on the 626'-O" elevation of the Auxiliary Building leading to a stairwell, normally-closed doors (47 and 48) leading to zones TU-97 and TU-98, normally-closed doors (46 and 280) on the 606'-O' elevation of the Turbine &

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 6 Building leading outdoors, normally-closed doors (11, 15, and 16) on the 586'-0" elevation of the Auxiliary Building leading to zones AX-20B, AX-21-1, and AX-23A-1, normally-closed door 401 leading to zone TU-94, and normally-closed doors (4 and 6) leading to zone TU-95B-1. The south wall has a normally-closed door (117) on the 626'-0" elevation leading to the Control Room, a normally-closed roll-up door (42) on the 606'-0" elevation leading to-the outdoors, normally-closed doors (70 and 74) leading to zones AX-39 and AX-33 on the 606'-0" elevation, and no doors on the 586'-0" elevation. The east wall has a normally-closed door (109) on the 626'-0" elevation and a normally-closed door (39) on the 606'-0" elevation leading to the Administrative Building, and no doors on the 586'-0" elevation. The west wall has normally-closed doors (118, 133, and 161) leading to zones AX-32-1 and AX-37 on the 626'-O" elevation, normally-closed doors (41, 44, and 49) leading to zones AX-32-1 and AX-30 on the 606'-O" elevation of the Auxiliary Building, a normally-open door (68) leading to the dosimetry offices on the 606'-0" elevation, a normally-closed door (75) in the Electric Shop leading outdoors, and a normally-closed door (7) on the 586'-0" elevation leading to zone TU-96. The east wall has a normally-closed door (109) on the 626'-0" elevation leading to the Administration Building, normally-closed doors (39 and 40) on the 606'-0" elevation leading to the Administration Building and outdoors, and no doors on the 586'-0" elevation.

The major PRA equipment in zone TU-22-1 includes the feedwater pumps (1A and 1B), the condensate pumps (IA and IB), MCC 45-F, and the Redundant Overspeed Trip System Cabinet.

Q The Internal Flood Walkdown Form [Appendix C of FLOOD01] for zone TU-22-1 contains a complete listing of the flood-susceptible PRA equipment in this zone.

Potential flood sources in this zone include fire protection piping, feedwater piping, service water piping, main steam piping, and circulating water piping which are the primary flood sources and represent both a flooding hazard and a spray hazard.

Flood mitigation is present in this zone in the form of floor grating, open stairways and sump pumps.

Analysis - Water from a pipe break in TU-22-1 will readily propagate to the basement level. The effects of a spray source in any part of the zone are limited to equipment in the vicinity of the spray source. Water is likely to splash onto equipment on lower levels as it passes through the floor grating. Accumulation is possible in the basement level (586'-O") of the zone.

As water from any pipe break in zone TU-22-1 makes its way into the Turbine Building basement, it will eventually fill the Turbine Building Sump. The sump contains two pumps with design capacities of < 100 gpm each. The level switch for the Turbine Building omp pump control is a mechanically alternating device. A high water level (30") starts one pumpr A return to low level (12") stops the pump. A subsequent high level starts the alternate pump.' [SYSTEM01]

If a high-high water level (34.5") is reached, the level switch starts the second pump. Both pumps I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 7 l continue to run until an intermediate level cutoff point, 19", is reached. At this point, the level switch turns off the leading (first) pump. The lagging (second) pump continues to run until the low-level setpoint, 12", is reached. [SYSTEM01]

The Turbine Building sump contains Level Switch LA-16666 that actuates Control Room Alarm 47033P when a high-high-high water level setpoint, 34.5", is reached. [SYSTEM01]

Thus, only pipe breaks of greater than 200 gpm, the combined discharge capacity of the Turbine Building sump pumps, are of concern for zone TU-22-1.

The first indication of such a break would be a Turbine Building sump high level alarm in the control room. The procedure for abnormal operation of the miscellaneous drains and sumps instructs the operator to dispatch someone to investigate the source of the alarm. If the source of leakage is from a break in the Circulating Water System, the operator is instructed to trip the circulating water pumps, trip the reactor, and perform a shutdown using emergency operating procedure E-0.

The effectiveness of such operator actions is dependent on the size of the pipe break. A small pipe break would likely afford the operator the time to perform the actions necessary to protect vital equipment in the Turbine Building basement. A large break would result in significant accumulation in the Turbine Building basement and could challenge the flood protection features in place to protect equipment located in adjacent zones. Water level in areas TU-94, TU-95B-1, TU-95B-2, and TU-95C would closely mirror the water level in TU-22-1 due to leakage under doors 4, 6, and 401 and due to flow through the drain lines that connect Safeguards Alley and the Turbine Building sump (these lines do not have check valves). Since drainage in these areas will be disabled due to the water in the Turbine Building, water will begin to accumulate in these rooms and begin to propagate to zones TU-90, TU-92, and TU-95A due to leakage under doors 3, 263, and 268. Power to the motor loads on the 4 kV buses in TU-90 and TU-92 will fail when the water level reaches 4 inches, submerging the lockout relays and tripping the breakers associated with the motor loads.

Summary - Pipe breaks in zone TU-22-1 can result in both equipment spray and submergence.

For spray events TU-22-1 becomes a flood area by itself since only equipment in TU-22-1 is susceptible to damage from direct spray originating in zone TU-22-1. However, water from such a spray event can result in the splashing of equipment in other elevations of the zone. For submergence events, zone TU-22-1 combines with all the zones in Safeguards Alley due to leakage under the doors associated with these rooms. When the water level in the Turbine Building sump reaches the high-high setpoint (approximately 34.5 inches above the sump floor) an annunciator sounds in the control room. Power to the motor loads on the 4 kV buses is expected to fail at 4 inches of water (although power will still be available to the 480 V buses),

the 480 V buses will then fail at 11 inches of water, and the turbine-driven auxiliary feedwater (TDAFW) pump will fail to start at 9 inches of water and fail to continue running at I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 8 approximately 18 inches of water.

Zone TU-22-1 is a relatively large room such that any water from a pipe break is expected to spray only equipment in zone TU-22-1 that is in close proximity to the pipe break.

For a pipe break in zone TU-22-1, equipment in zones TU-22-1, TU-90, TU-92, TU-94, TU-95A, TU-95B-1, TU-95B-2 and TU-95C can be vulnerable and could be at risk.

TU-90.

Description - Flood Zone TU-90 is Diesel Generator Room 1A on the 586'-O" elevation. The zone is bounded on the north by an exterior wall, on the south by zone TU-92, on the east by an exterior wall and the pipe tunnel leading to the Screenhouse, and on the west by zones TU-94 and TU-95A. The Administrative Building lies above zone TU-90 and exterior soil lies below.

All penetrations in zone TU-90 are sealed. The south wall has a normally-closed access door (2) leading to a Screenhouse pipe tunnel and the west wall has a normally-closed access door (136) leading to zone TU-95A.

The major PRA equipment in zone TU-90A includes Diesel Generator 1A, 4 kV Switchgear Bus 5, and MCC 52A. The Internal Flood Walkdown Form (Appendix C of FLOOD01] for zone TU-90 contains a complete listing of the flood-susceptible PRA equipment in this zone.

Potential flood sources in this zone include service water piping and fire protection piping which represent both a flooding hazard and a spray hazard.

Flood mitigation is present in this zone in the form of a trench which is sealed to prevent flow from traveling to zones TU-94 and TU-95A, but is open via a 4-inch pipe to the pipe tunnel leading to the Screenhouse. Floor drains will transfer water to the Turbine Building sump.

Analysis - Water from a pipe break in TU-90 will easily propagate to the pipe tunnel leading to the Screenhouse through an open 4-inch pipe that connects the two areas in the existing trench.

Floor drains will divert water to the Turbine Building sump. Thus, only pipe breaks that exceed the capability of the floor drains are a concern for accumulation in zone TU-90. Equipment damage from spray sources in TU-90 is limited to the equipment residing in that zone.

Zone TU-90 is equipped with two normally closed doors that initially prevent lateral propagation of water from pipe breaks beyond the capacity of the floor drains. Door 2 (double door with a 1/64" gap) opens outwardly to the pipe tunnel leading to the Screenhouse and door 136 (double door with a 1/8" gap) opens inwardly from zone TU-95A. Initially water would flow through the floor drains to the Turbine Building sump and flow through the open 4-inch pipe to the Screenhouse sump. However, given the limited capacity of the floor drains and the 4-inch pipe in TU-90, neither the Turbine Building sump nor the Screenhouse sump will reach a level high I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p.9 enough to initiate a control room alarm. While water is flowing through the floor drains it will also be leaking under door 2 to the pipe tunnel that leads to the Screenhouse. Once the seiche hump is overcome in the pipe tunnel, water leaking under the door will also flow to the Screenhouse sump. When the water level inside TU-90 reaches a critical height, both doors are expected to fail allowing water to freely propagate to zone TU-95A and the pipe tunnel leading to the Screenhouse. A significant flow of water through a pipe break would be required for any accumulation of water in TU-90.

The Turbine Building Sump contains two pumps with design capacities of < 100 gpm each. The level switch for the Turbine Building sump pump control is a mechanically alternating device. A high water level (30") starts one pump. A return to low level (12") stops the pump. A subsequent high level starts the alternate pump. [SYSTEM01]

If a high-high water level (34.5") is reached, the level switch starts the second pump. Both pumps continue to run until an intermediate level cutoff point, 19", is reached. At this point, the level switch turns off the leading (first) pump. The lagging (second) pump continues to run until the low-level setpoint, 12", is reached. [SYSTEMOl]

The Turbine Building sump contains Level Switch LA-16666 that actuates Control Room Alarm 47033P when a high-high-high water level setpoint, 34.5", is reached. [SYSTEM01]

The Screenhouse sump contains two pumps with design capacities of < 100 gpm each. The level switch for the Screenhouse sump pump control is a mechanically alternating device. A high water level (30") starts one pump. A return to low level (12") stops the pump. A subsequent high level starts the alternate pump.

If a high-high water level (34.5") is reached, the level switch starts the second pump. Both pumps continue to run until an intermediate level cutoff point, 19", is reached. At this point, the level switch turns off the leading (first) pump. The lagging (second) pump continues to run until the low-level setpoint, 12", is reached.

The Screenhouse sump contains Level Switch LA-16669 that actuates Control Room Alarm 47033P when a high-high-high water level setpoint, 34.5", is reached [SYSTEMOl].

The operator's first indication of a pipe break inside TU-90 will be the high Screenhouse sump level alarm in the control room once the water level inside TU-90 rises high enough to fail door 2 which will allow water to flow freely to the Screenhouse. The only other possible indication of a pipe break would be equipment failure that forces an operator to investigate locally.

Summary - Pipe breaks in zone TU-90 can result in both equipment spray and submergence.

For spray events TU-90 becomes a flood area by itself since only equipment in TU-90 is susceptible to damage from direct spray originating in zone TU-90. For submergence events, (

I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 10 11 M I ., A\ ................

.i 11 zone TU-90 combines with zone TU-95A and the pipe tunnel leading to the Screenhouse due to leakage under the doors.

Zone TU-90 is a relatively small room such that any water from a pipe break is expected to spray all the equipment in zone TU-90.

For a pipe break in zone TU-90, equipment in the zone zones TU-90, TU-95A, and the pipe tunnel leading to the Screenhouse can be vulnerable and could be at risk.

TU-92 Description - Flood Zone TU-92 is Diesel Generator Room 1B on the 586'-0" elevation. The zone is bounded on the north by zones TU-90 and the pipe tunnel leading to the Screenhouse, on the south by an exterior wall, on the east by an exterior wall and the pipe tunnel leading to the Screenhouse, and on the west by zones TU-94 and TU-22-1. The Administrative Building lies above and exterior soil lies below.

All penetrations in zone TU-92 are sealed. The north wall has a normally-closed access door (1) leading to a service water piping tunnel that leads to the Screenhouse and the west wall has a normally-closed access door (3) leading to zone TU-94.

The major PRA equipment in zone TU-92 includes Diesel Generator 1B, 4 kV Switchgear Bus 6, and MCC 62A. The Internal Flood Walkdown Form [Appendix C of FLOOD01] for zone TU-92 contains a complete listing of the flood-susceptible PRA equipment in this zone.

Potential flood sources in this zone include service water piping and fire protection piping which represent both a flooding hazard and a spray hazard.

Flood mitigation is present in this zone in the form of floor drains. (A six-inch curb that ran east and west just north of all the equipment protected the equipment from water originating from outside the room until late 2004, however it has since been removed.)

Analysis - Water from a pipe break in TU-92 will not easily propagate elsewhere since all the penetrations are sealed. Floor drains will divert water to the Turbine Building sump. Thus, only significant pipe breaks are a concern for accumulation in zone TU-92. Equipment damage from spray sources in TU-92 is limited to the equipment residing in that zone.

Zone TU-92 is equipped with two normally closed doors that initially prevent lateral propagation of water from pipe breaks beyond the capacity of the floor drains. Door 1 (double door with a 1/64" gap) opens outwardly to the pipe tunnel leading to the Screenhouse and door 3 (double door with a 1/64" gap) opens inwardly from zone TU-94. Initially water would flow through the floor drains to the Turbine Building sump, but given the limited capacity of the floor drains in TU-I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 11 -l I .. 1 92, the Turbine Building sump will not reach a level high enough to initiate a control room alarm.

While water is flowing through the floor drains it will also be leaking under door 1 to the pipe tunnel that leads to the Screenhouse. Once the seiche hump is overcome in the pipe tunnel, water leaking under the door will flow to the Screenhouse sump. When the water level inside TU-92 reaches a critical height, both doors are expected to fail allowing water to freely propagate to zone TU-94 and the pipe tunnel leading to the Screenhouse.

The Turbine Building Sump contains two pumps with design capacities of < 100 gpm each. The level switch for the Turbine Building sump pump control is a mechanically alternating device. A high water level (30") starts one pump. A return to low level (12") stops the pump. A subsequent high level starts the alternate pump. [SYSTEM01]

If a high-high water level (34.5") is reached, the level switch starts the second pump. Both pumps continue to run until an intermediate level cutoff point, 19", is reached. At this point, the level switch turns off the leading (first) pump. The lagging (second) pump continues to run until the low-level setpoint, 12", is reached. [SYSTEM01]

The Turbine Building sump contains Level Switch LA-16666 that actuates Control Room Alarm 47033P when a high-high-high water level setpoint, 34.5", is reached. [SYSTEM01]

The Screenhouse sump contains two pumps with design capacities of < 100 gpm each. The level switch for the Screenhouse sump pump control is a mechanically alternating device. A high water level (30") starts one pump. A return to low level (12") stops the pump. A subsequent high level starts the alternate pump.

If a high-high water level is reached, the level switch starts the second pump. Both pumps continue to run until an intermediate level cutoff point, 19", is reached. At this point, the level switch turns off the leading (first) pump. The lagging (second) pump continues to run until the low-level setpoint, 12", is reached.

The Screenhouse sump contains Level Switch LA-16669 that actuates Control Room Alarm 47033P when a high-high-high water level setpoint, 34.5", is reached [SYSTEM01].

The operator's first indication of a pipe break inside TU-92 will be the high Screenhouse sump level alarn in the control room once the water level inside TU-92 rises high enough to fail door 1 which will allow water to flow freely to the Screenhouse. The only other possible indication of a pipe break would be equipment failure that forces an operator to investigate locally.

Summary - Pipe breaks in zone TU-92 can result in both equipment spray and submergence.

For spray events TU-92 becomes a flood area by itself since only equipment in TU-92 is susceptible to damage from direct spray originating in zone TU-92. For submergence events, zone TU-92 combines with zone TU-94 and the pipe tunnel leading to the Screenhouse due to I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 12 leakage under the doors.

Zone TU-92 is a relatively small room such that any water from a pipe break is expected to spray all the equipment in zone TU-92.

For a pipe break in zone TU-92, equipment in the zone zones TU-92, TU-94, and the pipe tunnel leading to the Screenhouse can be vulnerable and could be at risk.

TU-94 Description - Flood Zone TU-94 is the C02 Storage Tank Room 1B on the 586'-O" elevation.

The zone is bounded on the north by zone TU-95A, on the south by zone TU-22-1, on the east by zones TU-90 and TU-92, and on the west by zone TU-22-1. Zone TU-22-1 lies above and exterior soil lies below.

All penetrations in zone TU-94 are sealed. The north wall has a normally-closed access door (5) leading to zone TU-95A, the south wall has a normally-closed access door (401) leading to zone TU-22-1, and the east wall has a normally-closed access door (3) leading to zone TU-92.

The major PRA equipment in zone TU-94 includes Station and Instrument Air Compressor 1A.

The Internal Flood Walkdown Form for zone TU-94 contains a complete listing of the flood-susceptible PRA equipment in this zone.

Potential flood sources in this zone include service water piping and fire protection piping which represent both a flooding hazard and a spray hazard.

Flood mitigation is present in this zone in the form of a floor drain in a trench that is sealed at the boundary of zone TU-90.

Analysis - Water from a pipe break in TU-94 will not easily propagate elsewhere since all the penetrations are sealed. Floor drains will divert water to the Turbine Building sump. Thus, only significant pipe breaks are a concern for accumulation in zone TU-94. Equipment damage from spray sources in TU-94 is limited to the equipment residing in that zone.

Zone TU-94 is equipped with three normally closed doors that initially prevent lateral propagation of water from pipe breaks beyond the capacity of the floor drains. Door 3 (double door with a 1/64" gap) opens outwardly to zone TU-92, door 5 (double door with 1/64" gap) opens inwardly from zone TU-95A, and door 401 (double door with 7/8" gap) opens outwardly to zone TU 1. Initially water would simply leak under doors 3 and 5 to flood areas TU-92 and TU-95A, respectively. Water will also flow to the Turbine Building sump via the floor drains. When the water level inside TU-94 reaches a critical height, doors 3 and 401 are expected to fail allowing water to freely propagate to zones TU-92 and TU-22-1, respectively.

I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 13 l The first indication of such a break would be a Turbine Building sump high level alarm in the control room if the flow via the floor drain is sufficiently high to fill the sump. The procedure for abnormal operation of the miscellaneous drains and sumps instructs the operator to dispatch someone to investigate the source of the alarm, regardless of which sump fills first. The only other possible indication of a pipe break would be equipment failure that forces an operator to investigate locally.

Summary - Pipe breaks in zone TU-94 can result in both equipment spray and submergence.

For spray events TU-94 becomes a flood area by itself since only equipment in TU-94 is susceptible to damage from direct spray originating in zone TU-94. For submergence events, zone TU-94 combines with zones TU-22-1, TU-95A, and TU-92 due to leakage under the associated doors.

Zone TU-94 is a relatively small room such that any water from a pipe break is expected to spray all the equipment in zone TU-94.

For a pipe break in zone TU-94, equipment in the zone zones TU-94, TU-22-1 and TU-95A can be vulnerable and could be at risk.

TU-95A I Q Description - Flood Zone TU-95A is the 480 V Switchgear Bus 1-51 and 1-52 Room on the 586'-0" elevation. The zone is bounded on the north by an exterior wall and the Technical Support Center, on the south by zones TU-22-1 and TU-94, on the east by zone TU-90, and on the west by zone TU-95B-1. Zone TU-22-1 lies above and exterior soil lies below.

All penetrations in zone TU-95A are sealed. The south wall has normally-closed access doors (5, 263 and 268) leading to zones TU-94 and TU-95B-1 and the east wall has a normally closed door (136) leading to zone TU-90.

The major PRA equipment in zone TU-95A includes Station and Instrument Air Compressor 1C, and 480 V Switchgear Buses 51 and 52. The Internal Flood Walkdown Form [Appendix C of FLOOD01] for zone TU-95A contains a complete listing of the flood-susceptible PRA equipment in this zone.

Potential flood sources in this zone include service water piping and fire protection piping which represent both a flooding hazard and a spray hazard.

Flood mitigation is present in this zone in the form of a trench that communicates with zone TU-90 and contains a floor drain leading to the Turbine Building sump.

I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 14 Analysis - Water from a pipe break in TU-95A will easily propagate to zone TU-90 via an open 4-inch pipe under door 136. Floor drains will divert water to the Turbine Building sump. Thus, only significant pipe breaks are a concern for accumulation in zone TU-95A. Equipment damage from spray sources in TU-95A is limited to the equipment residing in that zone.

Zone TU-95A is equipped with three normally closed doors that initially prevent lateral propagation of water from pipe breaks beyond the capacity of the floor drains. Door 5 (double door with 1/64" gap) opens outwardly to zone TU-94, door 136 (double door with 1/8" gap) opens outwardly to zone TU-90, door 263 (double door with 3/16" gap) opens outwardly to zone TU-95B-1, and door 268 (single door) opens outwardly to zone TU-95B-1. Additionally, a firewall constructed of gypsum board separates TU-95A and TU-95B-1. Initially water would simply leak under doors to the various adjoining zones and flow to the Turbine Building sump via the floor drains. When the water level inside TU-95A reaches a critical height, the firewall is expected to fail structurally allowing water to freely propagate to zones TU-95B-1.

The first indication of such a break would likely be from investigation of failed equipment since free flow to either the Screenhouse sump or the Turbine Building sump does not occur until water level accumulates to several feet and doors and gypsum wall begin to fail.

Summary - Pipe breaks in zone TU-95A can result in both equipment spray and submergence.

For spray events TU-95A becomes a flood area by itself since only equipment in TU-95A is susceptible to damage from direct spray originating in zone TU-95A. For submergence events, zone TU-95A combines with zones TU-90, TU-94, and TU-95B-1 due to leakage under the associated doors and an open pipe that allows communication between TU-95A and TU-90.

Zone TU-95A is a relatively small room such that any water from a pipe break is expected to spray all the equipment in zone TU-95A.

For a pipe break in zone TU-95A, equipment in zones TU-94, TU-95B-1, TU-90, and TU-95A can be vulnerable and could be at risk.

TU-95B-1 Description - Flood Zone TU-95B-1 consists of the 480 V Switchgear Bus 61 and 62 Room and the Auxiliary Feedwater Pump 1B Room on the 586'-O" elevation. These two rooms are connected via an open trench such that any water in one room will travel freely to the other, thus they are combined to form a single flood area for submergence issues. The zone is bounded on the north by the Technical Support Center, on the south by zone TU-22-1, on the east by zones TU-95A, TU-95B-2, and TU-95C, and on the west by zone AX-23B-1. Zone TU-22-1 lies above and exterior soil lies below.

All penetrations in zone TU-95B-1 are sealed. The south wall has normally-closed access doors I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 15 l (4 and 6) leading to zone TU-22-1, the north wall has normally-closed access doors (268, 263, 262, and 261) leading to zones TU-95A and TU-95C, the west wall has a normally-closed access door (244) leading to zone TU-95B-2 and a normally-closed access door (8) leading to the Auxiliary Building, and the east wall has a normally-closed door (243) leading to zone TU-95B-2.

The major PRA equipment in zone TU-95B-1 includes Station and Instrument Air Compressor 1B, Motor Driven Auxiliary Feedwater Pump B, and 480 V Switchgear Buses 1-61 and 1-62.

The Internal Flood Walkdown Form [Appendix C of FLOOD01] for zone TU-95B-1 contains a complete listing of the flood-susceptible PRA equipment in this zone.

Potential flood sources in this zone include service water piping, CST piping, main steam piping, and fire protection piping which represent both a flooding hazard and a spray hazard.

Flood mitigation is present in this zone in the form of a trench that is sealed at the boundary of zone TU-95A and zone TU-95B-1. The trench contains a floor drain leading to the Turbine Building sump.

Analysis - Water from a pipe break in TU-95B-1 will not easily propagate elsewhere since all the penetrations are sealed. Floor drains will divert water to the Turbine Building sump. Thus, only significant pipe breaks are a concern for accumulation in zone TU-95B-1. Equipment damage from spray sources in TU-95B-1 is limited to the equipment residing in that zone unless it is a 0

prolonged spray. A prolonged spray (greater than 90 minutes) in the western half of the area would probably degrade the gypsum board that comprises area TU-95C to the point that the auxiliary feedwater components housed inside TU-95C would be damaged.

Zone TU-95B-1 is equipped with seven normally closed doors that initially prevent lateral propagation of water from pipe breaks beyond the capacity of the floor drains. Door 4 (double door with a 1/8" gap) opens outwardly to zone TU-22-1, door 6 (double door with 1/4" gap) opens outwardly to zone TU-22-1, door 243 (single door with 1/32" gap) opens outwardly to TU-95B-2, door 244 (single door with 1/32" gap) opens outwardly to TU-95B-2, door 261 (single door with 3/16" gap) opens inwardly from TU-95C, door 262 (double door with 3/16" gap) opens inwardly from TU-95C, door 263 (double door with 3/16" gap) opens inwardly from zone TU-95A, and door 268 (single door) opens inwardly from zone TU-95A. Initially water would simply leak under doors to flood areas TU-95A, TU-95B-2, and TU-95C as well as flow to the Turbine Building sump via the floor drains. When the water level inside TU-95B-1 reaches a critical height, doors and gypsum walls are expected to fail allowing water to freely propagate to adjoining areas. As doors fail water will propagate to TU-22-1 where it will fill the Turbine Building sump.

The Turbine Building Sump contains two pumps with design capacities of < 100 gpm each. The level switch for the Turbine Building sump pump control is a mechanically alternating device. A I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 16 l  : .

high water level (30") starts one pump. A return to low level (12") stops the pump. A subsequent high level starts the alternate pump. [SYSTEM0l]

If a high-high water level (34.5") is reached, the level switch starts the second pump. Both pumps continue to run until an intermediate level cutoff point, 19", is reached. At this point, the level switch turns off the leading (first) pump. The lagging (second) pump continues to run until the low-level setpoint, 12", is reached. [SYSTEM01]

The Turbine Building sump contains Level Switch LA-16666 that actuates Control Room Alarm 47033P when a high-high-high water level setpoint, 34.5", is reached. [SYSTEM01]

The first indication of such a break would be a Turbine Building sump high level alarm in the control room. The procedure for abnormal operation of the miscellaneous drains and sumps instructs the operator to dispatch someone to investigate the source of the alarm. The only other possible indication of a pipe break would be equipment failure that forces an operator to investigate locally.

Summary - Pipe breaks in zone TU-95B-1 can result in both equipment spray and submergence.

For spray events TU-95B-1 becomes a flood area by itself since only equipment in TU-95B-1 is susceptible to damage from direct spray originating in zone TU-95B-1. For submergence events, zone TU-95B-1 combines with zones TU-22-1, TU-95B-2, TU-95C, and TU-95A due to leakage under the associated doors.

Zone TU-95B-1 is separated into two distinct sections by zone TU-95B-2. Each of these sections is a relatively small area such that any water from a pipe break is expected to spray all the equipment in that area of zone TU-95B- 1.

For a pipe break in zone TU-95B-1, equipment in zones TU-95B-1, TU-22-1, TU-95B-2, TU-95C, and TU-95A can be vulnerable and could be at risk.

TU-95B-2 Description - Flood Zone TU-95B-2 is the Turbine Driven Auxiliary Feedwater Pump Room on the 586'-O" elevation. The zone is bounded on the north by an exterior wall and the Technical Support Center, on the south by zone TU-22-1, on the east by zones TU-95B-1 and TU-95C, and on the west by zone TU-95B-1. Zone TU-95B-2 makes use of a false ceiling for HELB purposes and Zone TU-95B-1 actually lies above. Exterior soil lies below.

Ail penetrations in zone TU-95B-2 are sealed. The east wall has a normally closed access door (244) leading to zone TU-95B-1 and the west wall has a normally closed door (243) leading to zone TU-95B-1. The south wall has a normally closed blowout panel that opens to zone TU 1.

I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 17 The major PRA equipment in zone TU-95B-2 includes the Turbine Driven Auxiliary Feedwater Pump. The Internal Flood Walkdown Form [Appendix C of FLOOD01] for zone TU-95B-2 contains a complete listing of the flood-susceptible PRA equipment in this zone.

Potential flood sources in this zone include service water piping, CST piping, and main steam piping which represent both a flooding hazard and a spray hazard.

Flood mitigation is present in this zone in the form of a covered trench that communicates with zone TU-95B- 1. A floor drain approximately 4 inches above the ground also communicates with this trench.

Analysis - Water from a pipe break in TU-95B-2 will not easily propagate elsewhere since all the penetrations are sealed. Floor drains will divert water to the Turbine Building sump. Thus, only significant pipe breaks are a concern for accumulation in zone TU-95B-2. Equipment damage from spray sources in TU-95B-2 is limited to the equipment residing in that zone.

Zone TU-95B-2 is equipped with two normally closed doors that initially prevent lateral propagation of water from pipe breaks beyond the capacity of the floor drains. Door 243 (single door with 1/32" gap) opens inwardly from TU-95B-1 and door 244 (single door with 1/32" gap) opens inwardly from TU-95B-1. Initially water would simply leak under doors to the various adjoining zones and flow to the Turbine Building sump via the floor drains. When the water level U

inside TU-95B-2 reaches a critical height, one of two things will occur. Either the blowout panel will fail allowing water to propagate to TU-22-1 and subsequently to the Turbine Building sump or both doors will fail allowing water to freely propagate to zone TU-95B-1. When the water level inside TU-95B-1 reaches a critical height, doors are expected to fail allowing water to freely propagate to TU-22-1 where it will fill the Turbine Building sump. In either case water will reach the Turbine Building sump.

The Turbine Building Sump contains two pumps with design capacities of < 100 gpm each. The level switch for the Turbine Building sump pump control is a mechanically alternating device. A high water level (30") starts one pump. A return to low level (12") stops the pump. A subsequent high level starts the alternate pump. [SYSTEM01]

If a high-high water level (34.5") is reached, the level switch starts the second pump. Both pumps continue to run until an intermediate level cutoff point, 19", is reached. At this point, the level switch turns off the leading (first) pump. The lagging (second) pump continues to run until the low-level setpoint, 12", is reached. [SYSTEM01]

The Turbine Building sump contains Level Switch LA-16666 that actuates Control Room Alarm 47033P when a high-high-high water level setpoint, 34.5", is reached. [SYSTEM01]

I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 18 The first indication of such a break would be a Turbine Building sump high level alarm in the control room. The procedure for abnoiial operation of the misce1laneous drains and sumps instructs the operator to dispatch someone to investigate the source of the alarm. The only other possible indication of a pipe break would be equipment failure that forces an operator to investigate locally.

Summary - Pipe breaks in zone TU-95B-2 can result in both equipment spray and submergence.

For spray events TU-95B-2 becomes a flood area by itself since only equipment in TU-95B-2 is susceptible to damage from direct spray originating in zone TU-95B-2. For submergence events, zone TU-95B-2 combines with zone TU-95B-1 due to door leakage.

Zone TU-95B-2 is a relatively small room such that any water from a pipe break is expected to spray all the equipment in zone TU-95B-2.

For a pipe break in zone TU-95B-2, equipment in zones TU-95B-1 and TU-95B-2 can be vulnerable and could be at risk.

TU-95C Description - Flood Zone TU-95C is the Motor Driven Auxiliary Feedwater Pump IA Room on the 586'-O" elevation. The zone is bounded on the north by the Technical Support Center, on the south by zone TU-95B-1, on the east by zone TU-95B-2, and on the west by zone TU-95B-1.

Zone TU-22-1 lies above and exterior soil lies below.

All penetrations in zone TU-95C are sealed. The south wall has normally-closed access doors (261 and 262) leading to zone TU-95B-1. The south and west walls are constructed of simple drywall and are expected to initially survive a spray event, but prolonged exposure to water will result in failure of the walls.

The major PRA equipment in zone TU-95C includes Motor Driven Auxiliary Feedwater Pump 1A. The Internal Flood Walkdown Form [Appendix C of FLOOD01] for zone TU-95C contains a complete listing of the flood-susceptible PRA equipment in this zone.

Potential flood sources in this zone include service water piping, CST piping, and main steam piping which represent both a flooding hazard and a spray hazard.

Flood mitigation is present in this zone in the form of a floor drain approximately 4 inches above the ground that communicates with the trench in zone TU-95B-1.

Analysis - Water from a pipe break in TU-95C will not initially propagate elsewhere since all the penetrations are sealed. Floor drains will divert water to the Turbine Building sump. Thus, only significant pipe breaks are a concern for accumulation in zone TU-95C. Equipment damage from I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 19 spray sources in TU-95C is initially limited to the equipment residing in that zone.

However, a sustained pipe break could eventually spray equipment in the western half of TU-95B-1 since the west and south walls of TU-95C are constructed of gypsum that is not expected to survive a sustained spray of water.

Zone TU-95C is equipped with two normally closed doors that initially prevent lateral propagation of water from pipe breaks beyond the capacity of the floor drains. Door 261 (single door with 3/16" gap) opens outwardly to TU-95B-1 and door 262 (double door with 3/16" gap) opens outwardly to TU-95B-1. Initially water would simply leak under doors to the various adjoining zones and flow to the Turbine Building sump via the floor drains. However, since the west and south walls of TU-95C are constructed of drywall, any sustained exposure to water is expected to result in failure of walls and open communication with TU-95B-1. Regardless of the failure mechanism, water will propagate to TU-95B-1.

When the water level inside TU-95B-1 reaches a critical height, doors are expected to fail allowing water to freely propagate to TU-22-1 where it will fill the Turbine Building sump.

The Turbine Building Sump contains two pumps with design capacities of < 100 gpm each. The level switch for the Turbine Building sump pump control is a mechanically alternating device. A high water level (30") starts one pump. A return to low level (12") stops the pump. A subsequent high level starts the alternate pump. [SYSTEM01]

If a high-high water level (34.5") is reached, the level switch starts the second pump. Both pumps continue to run until an intermediate level cutoff point, 19", is reached. At this point, the level switch turns off the leading (first) pump. The lagging (second) pump continues to run.until the low-level setpoint, 12", is reached. [SYSTEM01]

The Turbine Building sump contains Level Switch LA-16666 that actuates Control Room Alarm 47033P when a high-high-high water level setpoint, 34.5", is reached. [SYSTEM01l The first indication of such a break would be a Turbine Building sump high level alarm in the control room. The procedure for abnormal operation of the miscellaneous drains and sumps instructs the operator to dispatch someone to investigate the source of the alarm. The only other possible indication of a pipe break would be equipment failure that forces an operator to investigate locally.

Summary - Pipe breaks in zone TU-95C can result in both equipment spray and submergence.

For spray events TU-95C combines with TU-95B-1 to become a flood area since the drywall construction of the TU-95C walls cannot withstand sustained exposure to water spray. For submergence events, zone TU-95C combines with TU-95B-1 to become a flood area due to door leakage and eventual door failure or gypsum wall failure.

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 20 Zone TU-95C is a relatively small room such that any water frd a pipe break is expected to spray all the equipment in zones TU-95C and TU-95B-1.

For a pipe break in zone TU-95C, equipment in zones TU-95C and TU-95B-1 can be vulnerable and could be at risk.

4.0 REFERENCES

[DWG01] Drawings

a. A-203 Rev. AW, General-Arrangement Turbine and Administration Building Basement Floor

b. A-204 Rev. BC, General Arrangement Reactor and Auxiliary Building Basement Floor

c. A-205 Rev. AM, General Arrangement Turbine and Administration Building Mezzanine Floor

d. A-206 Rev. BS, General Arrangement Reactor and Auxiliary Building Mezzanine Floor

e. A-207 Rev. U, General Arrangement Turbine and Administration Building Operating Floor

f. A-208 Rev. BL, General Arrangement Reactor and Auxiliary Building

g. A-209 Rev. Y, General Arrangement Reactor and Auxiliary Building Miscellaneous Floor Plans

h. A-212 Rev. Y, General Arrangement Miscellaneous Plans and Sections

i. A-213 Rev. Y, General Arrangement Screenhouse and Circulating Water Discharge I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 21 11 I

[CALC01] "Evaluation of Various Doors and Walls for HELB Flooding - Turbine Building Basement," Revision 1, Scientel Wireless, LLC.

[CALC02] Attachment "KNPP Design Input Document.pdf' to email from Ling Yu Song (MPR) to Dale Franson (NMC) and Paul Miller (NMC), April 8, 2005.

[FLOOD01] Kewaunee Nuclear Plant Internal Flooding Analysis - Qualitative Screening Assessment and Flood Frequency Development, SCIENTECH, LLC.

[GUIDE01] Dominion Probabilistic Risk Assessment Manual, Part II, Chapter G, Section 2, FloodArea Definition, Revision 0.

[GUIDE02] Dominion Probabilistic Risk Assessment Manual, Part II, Chapter G, Section 4, Accident Sequence Analysis, Revision 0.

[NOTEBOOK01] KPS Internal Flooding Accident Sequence Analysis Notebook, Rev. 0

[PROC01] KNPP Operating Procedure A-MDS-30, Rev. P, "Miscellaneous Drains and Sumps (MDS) Abnormal Operation" U)

[STANDARD01] ASME RA-S-2002, "Standard for Probabilistic Risk Assessment for Nuclear Power Plant Applications," 2002

[SYSTEM01] KPS System Description 30, Rev. 2, "Miscellaneous Drains and Sumps (MDS)"

I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p. 22 11 Table 1 - Flood Area Descriptions Flood Zone Room Number Room Description Turbine Building - (Condenser) Basement 6B Floor 120 Turbine Building - Mezzanine Floor 121 Turbine Building - Mezzanine Floor 122 Turbine Building - Mezzanine Floor 123 Turbine Building - Mezzanine Floor 124 Turbine Building - Mezzanine Floor 125 Turbine Building - Mezzanine Floor 126 Turbine Building - Mezzanine Floor 127 Turbine Building - Mezzanine Floor 128 Turbine Building - Mezzanine Floor 220A Turbine Building - Operating Floor TU-22-1 10B Elevator B Machine Room 11B Corridor and Ramps 17B Waste Tank Area 144 Welding Shop 147 Corridor 149 Main Shop and Corridor (147) 150 Working Material Storage Area 154 Shop Office 155 Electric Shop Cation, Brine, and Mixed Beds - Water 234 Treatment Area 234A SG Boric Acid Area 2B Diesel Generator A Room 171-90 25B Diesel Generator A Fuel Oil Day Tank Room TU-923B Diesel Generator B Room 24B Diesel Generator B Fuel Oil Day Tank Room TU-94 4B C02 Storage Room TU-95A SB 480V Swgr Bus 1-51 and 1-52 Room 5B-1 480V Swgr Bus 1-61 and 1-62 Room TU-95B-1 5B-3 Aux FW Pump B Room TU-95B-2 5B-4 Turbine Driven Aux FW Pump Room TU-95C SB-2 Aux FW Pump A Room I

INTERNAL FLOODING - Flood Area Definition for Turbine Building Basement p.23l I Figure 1 - Turbine Building Basement/Safeguards Allev Arrangement I,_

4 TO SCREENHOUSE

& SEICHE HUMP I

C c c

Appendix C Fault Tree Analysis 1

Fault Tree Analysis Owner's Acceptance: e_16__ _ _L Tabk 6 a46-Signature Print Name Date 2

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Kewaunee Power Station Fault Tree Analysis for Turbine Building Floods Effective Date: November 2005 I S. solve LULf HEL Prepared By: J. R. Sharpe Date 3T & Jbfq5fs l,)A l A4L Reviewed By: D. Jones Date 3

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Table of Contents Section Fage 1.0 PURPOSE .................................. 5 2.0 MODEL SCOPE .................................. 5 3.0 UNIT DIFFERENCES .................................. 5 4.0 RISK MONITOR CONSIDERATIONS .................................. 5 5.0 MODEL DEVELOPMENT ................................. 5 5.1 FAULT TREES .... 5.....................

5.2 HUMAN ERROR PROBABILITIES ................................. 7 5.3 DATA ................................. 30 6.0 MODEL EVALUATION (EQUATIONS) ................................. 31 6.1 EVALUATION OF TOP EVENT AFZ ................................. 31 6.2 EVALUATION OF TOP EVENT AFX ................................. 35

7.0 REFERENCES

.................................. 41 List of Tables Table Page 1 Summary of KPS Turbine Building Flood Human Actions for Isolation 39 2 Basic Events Added to KNPP.BED 43 List of Figures Figure Page 1 Fault Tree AFM.LGC 49 2 Fault Tree FLOODING.LGC 67 3 Placement of New Basic Event 16-BATCLG--F-HE 82 4

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 1.0 PURPOSE The purpose of this notebook is to document the WinNUPRA model that was developed to analyze flooding scenarios originating from pipe breaks in the Turbine Building before February 2005.

The following information is identified, correlated, and developed as part of this analysis:

• Fault trees developed to support event tree analysis

• Basic event data used to support the flooding model

• Human error probabilities,(HEPs) used to support the flooding model 2.0 MODEL SCOPE This notebook documents the models that were developed for evaluating internal flooding sequences due to pipe breaks in the Turbine Building before February 2005.

3.0 UNIT DIFFERENCES Kewaunee Power Station is a single unit site so there are no unit differences.

4.0 RISK MONITOR CONSIDERATIONS The risk monitor used at KPS is the Safety Monitor. The Safety Monitor was not modified to reflect this analysis.

5.0 MODEL DEVELOPMENT 5.1 FAULT TREES The existing system fault trees for the KPS internal events PRA [NB01] comprise the majority of the Turbine Building Flood model. Two new fault trees were developed to support this analysis; AFM.LGC and FLOODING.LGC are described below in Sections 5.1.1 and 5.1.2. Fault tree AFM.LGC contains the logic associated with Auxiliary Feedwater (AFW) failures and fault tree FLOODING.LGC was developed to accommodate new initiating events and new human actions specifically related to Turbine Building flooding. Of the existing fault trees from the internal events PRA, only those for DC power were modified, as described in Section 5.1.3.

The human error probabilities (HEPs) used in the analysis are documented in Attachment 1. The bases for the HEPs from a review of procedures (e.g., cues) and training materials is provided in . A summary of a simulator exercise performed to determine timing for operator 5

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods actions in the feedwater line break scenario with actuation of all the fire sprinklers in the turbine building is provided in Attachment 3.

5.1.1 Fault Tree AFM Fault Tree AFM is presented as Figure 1. This fault tree contains the logic associated with failure of the Turbine Driven AFW pump and Motor Driven AFW Pump B (MDAFP B) to deliver flow to the steam generators. The logic in AFM is simply copied from Fault Tree AFW in the Internal Events PRA [NBO1I] and rearranged for use in this flooding analysis. No new analysis was performed in the development of fault tree AFM. Top Event AFS (as defined in the Accident Sequence Analysis, Appendix D) uses gate GAFM302 to model the failure of MDAFP B to start.

Top Event AFR uses gate GAFM700 to model the failure of MDAFP B to run and provide flow to Steam Generator B. Top Event AFT uses gate GAFM1002 to model the failure of the TDAFP to start and run.

5.1.2 Fault Tree FLOODING Fault Tree FLOODING is presented as Figure 2. This fault tree contains the logic used to model the initiating events used for Turbine Building floods and the HEPs associated with the isolation of pipe breaks and the operation of mitigating equipment. In some cases the hardware failure basic events are also included.

5.1.3 DC Power Fault Tree Modifications The DC power fault trees were modified to include basic event 16-BATCLG--F-HE, which represents operator action to establish battery room cooling. This event applies to flooding scenarios where the 480 V buses have failed, thereby causing failure of normal battery room cooling. After the Battery Room A/B Exhaust Flow Low annunciator activates in the control room, the operator is directed to use the fire equipment to ventilate the Battery Rooms. The air trunks and fans are then rigged to supply battery room cooling.

Figure 3 shows the placement of new event 16-BATCLG--F-HE in fault tree BRA104, at grid location "2-3". The same event is similarly placed in the following DC power fault trees:

BRA104B BRB104 BRB127 BRA104T BRB104B BRC103 BRA105 BRB104T BRC103T BRA1OST BRB105 BRD103 BRA113 BRB1O5T BRD103T BRA127 BRB 114 BRD115 6

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 5.2 HUMAN ERROR PROBABILITIES Human error probabilities (HEPs) were developed using the same methodology used in the existing PRA [NB02]. This section briefly describes each HEP developed as part of the analysis of Turbine Building floods. The detailed analyses of these HEPs are documented as attachments to this report. Table 1 lists all of the new human actions and their values that were developed in support of the flooding analysis.

5.2.1 04-CW-TRIP-F-HE - Detection and Isolation of a 58,000-gpm Circulating Water Break before Failing Both 480 V Buses The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Large Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent the eventual failure of the 480 V buses.

A large rupture of an inlet condenser expansion joint in the Turbine Building (TU-22-1) could propagate through the open drain lines and under doors to Safeguards Alley (TU-90, TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Areas TU-95A and TU-95B-1 contain the train A and B 480 VAC buses which could be failed due to propagation of a break in TU-22-1.

Indication of this type of break would be provided by a reactor trip due to low condenser vacuum and a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALCOI the operator must isolate the break within 3 minutes to prevent eventual loss of the 480 VAC buses in Safeguards Alley.

Thus, 3 minutes would be available to trip the Circulating Water pumps following an expansion joint rupture to prevent the eventual failure of the 480 V buses. Based on simulator observations and operator interviews at least 9 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 04-CW-TRIP-F-HE and the human error probability (HEP) is 1.0 since sufficient time does not exist to perform the isolation.

5.2.2 04-CWSTP13-F-HE - Detection and Isolation of a 14,000-gpm Circulating Water Break before Eventual Isolation of the 4 kVAC Bus 5 Motor Loads 7

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Moderate Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent the eventual isolation of the 4 kVAC Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.

A rupture of an outlet condenser expansion joint in the Turbine Building (TU 1) could propagate through the open drain lines and under doors to Safeguards Alley (TU-90, TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Area TU-90 contains kVAC Bus 5 which could be failed due to propagation of a break in TU 1.

Indication of this type of break would be provided by a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 13 minutes to prevent eventual isolation of 4 kV Bus 5 motor loads.

Thus, 13 minutes would be available to trip the Circulating Water pumps following an outlet expansion joint rupture to prevent the eventual isolation of 4 kV Bus 5 motor loads due to the automatic tripping of the associated circuit breakers. Based on simulator observations and operator interviews at least 9 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 04-CWSTP13-F-HE and the human error probability (HEP) is 2.6E-01.

5.2.3 04-CWSTP19-F-HE - Detection and Isolation of a 14,000-gpm Circulating Water Break before Failing the Turbine Driven AFW PuMp Auxiliary Lube Oil Pump The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Moderate Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent the eventual failure of the Turbine Driven AFW pump auxiliary lube oil pump.

This event is identical to the one described in section 5.2.2 except that the failure of interest is the Turbine Driven AFW pump auxiliary lube oil pump at 9 inches of water. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 19 minutes to prevent eventual loss of the ability to start the TDAFP.

Based on simulator observations and operator interviews at least 9 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for 8

INTERNAL FLOODING - Fault Tree Analysis for Turbine Builfing Floods this HEP is 04-CWSTP19-F-HE and the human error probability (HEP) is 1.2E-01.

5.2.4 04-CWSTP22-F-HE - Detection and Isolation of a 14,000-gpm Circulating Water Break before Failing 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Moderate Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent submergence failure of 480 VAC Buses 61 and 62 and the eventual isolation of the 4 kV Bus 6 motor loads due to the automatic tripping of the associated circuit breakers. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 22 minutes to accomplish these objectives.

Based on simulator observations and operator interviews at least 9 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 04-CWSTP22-F-HE and the human error probability (HEP) is 1.2E-01.

5.2.5 04-CWSTP25-F-BE - Detection and Isolation of a 14,000-gpm Circulating Water Break before Failure of the Motor Driven AFW Pumps and a Water Level of 18 Inches in the Turbine Building The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Moderate Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent the submergence failure of the motor driven AFW pumps and prevent the water level from reaching 18 inches in the Turbine Building. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 25 minutes to accomplish these objectives.

Based on simulator observations and operator interviews at least 9 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 04-CWSTP25-F-HE and the human error probability (HEP) is 1.2E-01.

5.2.6 02-SW4A-B29F-HE - Detection and Isolation of a Service Water Break before Eventual Isolation of the 4 kVAC Bus 5 Motor Loads The analysis of this REP is documented in Attachment 1. This basic event applies only to a Large Service Water break in the Turbine Building. This basic event represents the failure of the operator to close MOVs SW-4A and SW-4B in time to prevent the eventual isolation of the 4 kVAC Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.

A large Service Water pipe break in the Turbine Building (TU-22-1) could propagate through the 9

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods open drain lines and under doors to Safeguards Alley (TU-90, TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Area TU-90 contains kVAC Bus 5 which could be failed due to propagation of a break in TU-22-1.

Indication of this type of break would be provided by a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 29 minutes to prevent eventual isolation of 4 kV Bus 5 motor loads.

Thus, 29 minutes would be available to close MOV SW-4A or SW-4B (only one will be open normally) following a Service Water pipe break to prevent eventual isolation of 4 kV Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.. Based on simulator observations and operator interviews about 13 minutes are required to diagnose the cause of the high sump level alarm, decide the course of action, and execute the isolation. The basic event ID for this HEP is 02-SW4A-B29F-HE and the human error probability (HEP) is 2.0E-02.

5.2.7 02-SW4A-B45F-HE - Detection and Isolation of a Service Water Break before Failing the Turbine Driven AFW Pump Auxiliary Lube Oil Pump The analysis of this BEP is documented in Attachment 1. This basic event applies only to a Large Service Water break in the Turbine Building. This basic event represents the failure of the operator to close MOVs SW-4A and SW-4B in time to prevent the eventual failure of the Turbine Driven AFW pump auxiliary lube oil pump.

This event is identical to the one described in Section 5.2.6 except that the failure of interest is the l Turbine Driven AFW pump auxiliary lube oil pump at 9 inches of water. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 45 minutes to prevent eventual loss of the ability to start the TDAFP.

Based on simulator observations and operator interviews about 13 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 02-SW4A-B45F-HE and the human error probability (REP) is 2.OE-02.

10

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 05.2.8 02-SW4A-B5 IF-HE - Detection and Isolation of a Large Service Water Break before Failing 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Large Service Water break in the Turbine Building. This basic event represents the failure of the operator to close MOVs SW-4A and SW-4B in time to prevent submergence failure of 480 VAC Buses 61 and 62 and the eventual isolation of the 4 kV Bus 6 motor loads due to the automatic tripping of the associated circuit breakers. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 51 minutes to accomplish these objectives.

Based on simulator observations and operator interviews at least 13 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 02-SW4A-B5lF-HE and the human error probability (HEP) is 2.OE-02.

5.2.9 02-SW4A-B60F-HE - Detection and Isolation of a Large Service Water Break before Failure of the Motor Driven AFW Pumps The analysis of this REP is documented in Attachment 1. This basic event applies only to a Large Service Water break in the Turbine Building. This basic event represents the failure of the operator to close MOVs SW-4A and SW-4B in time to prevent the eventual submergence failure or the MDAFPs at 13 inches.

This event is identical to the one described in Section 5.2.6 except that the result of interest is submergence of the motor driven AFW pumps. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 60 minutes to prevent the submergence failure of the motor driven AFW pumps.

Based on simulator observations and operator interviews at least 13 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 02-SW4A-B60F-HE and the human error probability (HEP) is 2.OE-02.

5.2.10 02-SW4A-B66F-HE - Detection and Isolation of a Large Service Water Break before Water Level Reaches 18 Inches in the Turbine Building The analysis of this BEP is documented in Attachment 1. This basic event applies only to a Large Service Water break in the Turbine Building. This basic event represents the failure of the operator to close MOVs SW-4A and SW-4B in time to prevent the water level from reaching 18 inches in the Turbine Building.

11

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Q

This event is identical to the one described in Section 5.2.6 except that the result of interest is 18 inches of water in the Turbine Building. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 66 minutes to prevent 18 inches of water in the Turbine Building.

Based on simulator observations and operator interviews at least 13 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 02-SW4A-B66F-HE and the human error probability (HEP) is 2.OE-02.

5.2.11 08-FPISO29-F-HE - Detection and Isolation of a Fire Protection Water Break before Eventual Isolation of the 4 kVAC Bus 5 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Large Fire Protection Water break in the Turbine Building. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump or by securing the power to the pumps in time to prevent the eventual isolation of the 4 kVAC Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.

A large Fire Protection Water pipe break in the Turbine Building (TU-22-1) could propagate through the open drain lines and under doors to Safeguards Alley (TU-90, TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Area TU-90 contains kVAC Bus 5 which could be failed due to C4 propagation of a break in TU-22-1.

Indication of this type of break would be provided by the Fire Pump Abnormal alarm in the control room and a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 29 minutes to prevent eventual isolation of the 4 kV Bus 5 motor loads.

Thus, 29 minutes would be available to close the Fire pump discharge manual valves or isolate power to the Fire pumps following a Fire Protection Water pipe break to prevent eventual isolation of the 4 kV Bus 5 motor loads due to the automatic tripping of the associated circuit breakers. Based on simulator observations and operator interviews about 32 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 08-FPIS029-F-HE and the human error probability (HEP) is 1.0.

12

INTERNAL FLOODING - Fault Tree Analysis for Turbine Builg Floods 5.2.12 08-FPISO45-F-HE - Detection and Isolation of a Fire Protection Water Break before Failing the Turbine Driven AFW Pump Auxiliary Lube Oil Pump The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Large Fire Protection Water break in the Turbine Building. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump or by securing the power to the pumps in time to prevent the eventual failure of the Turbine Driven AFW pump auxiliary lube oil pump.

This event is identical to the one described in Section 5.2.11 except that the failure of interest is the Turbine Driven AFW pump auxiliary lube oil pump at 9 inches of water. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 45 minutes to prevent eventual loss of the ability to start the TDAFP.

Based on simulator observations and operator interviews about 32 minutes is required-to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 08-FPIS045-F-HE and the human error probability (HEP) is 6.6E-02.

5.2.13 08-FPISO56-F-HE - Detection and Isolation of a Fire Protection Water Break before Failing 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads The analysis of this REP is documented in Attachment 1. This basic event applies only to a Large Fire Protection Water break in the Turbine Building. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump or by securing the power to the pumps in time to prevent submergence failure of 480 VAC Buses 61 and 62 and the eventual isolation of the 4 kV Bus 6 motor loads due to the automatic tripping of the associated circuit breakers. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 56 minutes to accomplish these objectives.

Based on simulator observations and operator interviews nearly 32 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for' this HEP is 08-FPISO56-F-HE and the human error probability (REP) is 2.4E-02.

5.2.14 08-FPISO68-F-HE - Detection and Isolation of a Fire Protection Water Break before Failure of the Motor Driven AFW Pumps The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Fire Protection Water break in the Turbine Building. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump or by securing the power to the pumps in time to prevent the submergence failure of the motor driven AFW pumps.

13

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods This event is identical to the one described in Section 5.2.11 except that the result of interest is submergence of the motor driven AFW pumps. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 68 minutes to prevent the submergence failure of the motor driven AFW pumps.

Based on simulator observations and operator interviews at least 32 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 08-FPISO68-F-HE and the human error probability (HEP) is 1.6E-02.

5.2.15 08-ISO-FS18F-HE - Detection and Isolation of a Large Flood due to a Feedwater Break before Eventual Isolation of the 4 kVAC Bus 5 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Feedwater break resulting in a large Fire Protection System discharge in the Turbine Building. A Feedwater line break in the Turbine Building will spill the contents of the hotwell onto the Turbine Building floor and result in an elevated building temperature that actuates multiple fire sprinklers. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent the eventual isolation of U the 4 kVAC Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.

A Feedwater pipe break in the Turbine Building (TU-22-1) would set off multiple fire sprinklers in addition to pumping the hotwell inventory into the Turbine Building. This water could propagate through the open drain lines and under doors to Safeguards Alley (TU-90, TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Area TU-90 contains kVAC Bus 5 which could be failed due to propagation of a break in TU-22-1. This event analyzes a Feedwater pipe break resulting in a 6000-gpm discharge of the Fire Protection system.

Indication of this type of break would be provided by the Fire Pump Abnormal alarm in the control room and a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 18 minutes to prevent eventual isolation of the 4 kV Bus 5 motor loads.

Thus, 18 minutes would be available to isolate the sprinklers following a Feedwater pipe break to prevent eventual isolation of the 4 kV Bus 5 motor loads due to the automatic tripping of the associated circuit breakers. In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires 14

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems.

Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALCOL] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for this HEP is 08-ISO-FS18F-HE and the human error probability (HEP) is 1.0 since sufficient time does not exist to isolate flow from the Fire pumps.

5.2.16 08-ISO-FS33F-HE - Detection and Isolation of a Large Flood due to a Feedwater Break before Failing the Turbine Driven AFW Pump Auxiliary Lube Oil Pump The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Feedwater break resulting in a large Fire Protection System discharge in the Turbine Building. A Feedwater line break in the Turbine Building will spill the contents of the hotwell onto the Turbine Building floor and result in an elevated building temperature that actuates multiple fire sprinklers. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent the eventual failure of the Q Turbine Driven AFW pump auxiliary lube oil pump.

This event is identical to the one described in Section 5.2.15 except that the failure of interest is the Turbine Driven AFW pump auxiliary lube oil pump at 9 inches of water. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 33 minutes to prevent eventual failure of the ability to start the TDAFP.

Thus, 33 minutes would be available to isolate the sprinkler flow following a Feedwater pipe break to prevent the eventual failure of the Turbine Driven AFW pump auxiliary lube oil pump.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix 13, GOTHIC analyses [CALCOI] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for this HEP is 08-ISO-FS33F-HE and the human error probability (HEP) is 4.4E-01.

15

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Foods 5.2.17 08-ISO-FS40F-HE - Detection and Isolation of a Large Flood due to a Feedwater Break before Failing 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Feedwater break resulting in a large Fire Protection System discharge in the Turbine Building. A Feedwater line break in the Turbine Building will spill the contents of the hotwell onto the Turbine Building floor and result in an elevated building temperature that actuates multiple fire sprinklers. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent submergence failure of 480 VAC Buses 61 and 62 and the eventual isolation of the 4 kV Bus 6 motor loads due to the automatic tripping of the associated circuit breakers. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 40 minutes to accomplish these objectives.

Thus, 40 minutes would be available to isolate the sprinkler flow following a Feedwater pipe break to prevent eventual failure of 480VAC Buses 61 and 62, and 4 kVAC Bus 6. In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the 0

supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALCOI] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for this HEP is 08-ISO-FS40F-HE and the human error probability (HEP) is 1.3E-01.

5.2.18 08-ISO-FS54F-HE - Detection and Isolation of a Large Flood due to a Feedwater Break before Failure of the Motor Driven AFW Pumps The analysis of this REP is documented in Attachment 1. This basic event applies only to a Feedwater break resulting in a large Fire Protection System discharge in the Turbine Building.

This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent the eventual submergence failure of the motor driven AFW pumps.

This event is identical to the one described in Section 5.2.15 except that the result of interest is submergence of the motor driven AFW pumps. Based on GOTHIC analysis [CALCO1] the operator must isolate the break within 54 minutes to prevent the submergence failure of the motor 16

INTERNAL FLOODING - Fault Tree Analysis for Turbine Bulding Floods driven AFW pumps.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALC01] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for this HEP is 08-ISO-FS54F-HE and the human error probability (HEP) is 3.OE-02.

5.2.19 08-ISO-FS55F-HE - Detection and Isolation of a Medium Flood due to a Feedwater Break before Eventual Isolation of the 4 kVAC Bus 5 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Feedwater break resulting in a moderate Fire Protection System discharge in the Turbine Building. A Feedwater line break in the Turbine Building will spill the contents of the hotwell onto the Turbine Building floor and result in an elevated building temperature that actuates multiple fire spriniders. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent the eventual isolation of the 4 kVAC Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.

A Feedwater pipe break in the Turbine Building (TU-22-1) would set off multiple fire sprinklers in addition to pumping the hotwell inventory into the Turbine Building. This water could propagate through the open drain lines and under doors to Safeguards Alley (TU-90,'TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Area TU-90 contains kVAC Bus 5 which could be failed due to propagation of a break in TU-22-1. This event analyzes a Feedwater pipe break resulting in a 2000-gpm discharge of the Fire Protection system.

Indication of this type of break would be provided by the Fire Pump Abnormal alarm in the control room and a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 55 minutes to prevent eventual isolation of the 4 kV Bus 5 motor loads.

17

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Thus, 55 minutes would be available to isolate the sprinkler flow following a Feedwater pipe break to prevent eventual isolation of the 4 kV Bus 5 motor loads due to the automatic tripping of the associated circuit breakers. In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems.

Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALCOl] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for this BEP is 08-ISO-FS55F-HE and the human error probability (HEP) is 3.OE-02.

5.2.20 08-ISO-FS97F-HE - Detection and Isolation of a Medium Flood due to a Feedwater Break before Failing the Turbine Driven AFW Pump Auxiliary Lube Oil Pump The analysis of this BEP is documented in Attachment 1. This basic event applies only to a Feedwater break resulting in a moderate Fire Protection System discharge in the Turbine Building. A Feedwater line break in the Turbine Building will spill the contents of the hotwell onto the Turbine Building floor and result in an elevated building temperature that actuates multiple fire sprinklers. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent the eventual failure of the Turbine Driven AFW pump auxiliary lube oil pump.

This event is identical to the one described in Section 5.2.19 except that the failure of interest is the Turbine Driven AFW pump auxiliary lube oil pump at 9 inches of water. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 97 minutes to prevent eventual failure of the ability to start the TDAFP.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALC01] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. . The basic event ID for this REP is 08-ISO-FS97F-HE and the human error probability (HEP) is 3.OE-02.

5.2.21 08-ISO-FS2HF-HE - Detection and Isolation of a Medium Flood due to a Feedwater Break before Failing 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 18

T F Ani G Te f 'ul od INTrERNAL FLOODING - Fault Tree Analysis for Turbine Building Floodsl Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Feedwater break resulting in a moderate Fire Protection System discharge in the Turbine Building. A Feedwater line break in the Turbine Building will spill the contents of the hotwell onto the Turbine Building floor and result in an elevated building temperature that actuates multiple fire sprinklers. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent submergence failure of 480 VAC Buses 61 and 62 and the eventual isolation of the 4 kV Bus 6 motor loads due to the automatic tripping of the associated circuit breakers. Based on GOTHIC analysis [CALCOI] the operator must isolate the break within 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> to accomplish these objectives.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALCOI] show that isolating flow from the basement, deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. . The basic event ID for this HEP is 08-ISO-FS2HF-HE and the human error probability (HEP) is 1.7E-02.

5.2.22 08-FPSISO29F-HE - Detection and Isolation of a Large Flood due to a Steamline Break before Eventual Isolation of the 4 kVAC Bus 5 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Steamline break resulting in a large Fire Protection System discharge in the Turbine Building. A Steamline break in the Turbine Building will result in an elevated building temperature that actuates multiple fire sprinlders. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation yalve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent the eventual isolation of the 4 kVAC Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.

A Steamline pipe break in the Turbine Building (TU-22-1) would set off multiple fire sprinklers in the Turbine Building. This water could propagate through the open drain lines and under doors to Safeguards Alley (TU-90, TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Area TU-90 contains kVAC Bus 5 which could be failed due to propagation of a break in TU-22-1. This event analyzes a Steamline break resulting in a 6000-gpm discharge of the Fire Protection system.

19

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Indication of this type of break would be provided by the Fire Pump Abnormal alarm in the control room and a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALCOl] the operator must isolate the break within 29 minutes to prevent eventual isolation of 4 kV Bus 5 motor loads.

Thus, 32 minutes would be available to isolate the sprinkler flow following a Steamline pipe break to prevent eventual isolation of 4 kV Bus 5 motor loads due to the automatic tripping of the associated circuit breakers. In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems.

Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALC01] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. . The basic event ID for this HEP is 08-FPSISO29F-HE and the human error probability (HEP) is 1.0.

5.2.23 08-FPSISO45F-BE - Detection and Isolation of a Large Flood due to a Steamline Break before Failing the Turbine Driven AFW Pump Auxiliary Lube Oil Pump The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Steamline break resulting in a large Fire Protection System discharge in the Turbine Building. A Steamline break in the Turbine Building will result in an elevated building temperature that actuates multiple fire sprinkders. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinlders in time to prevent the eventual failure of the Turbine Driven AFW pump auxiliary lube oil pump.

This event is identical to the one described in Section 5.2.22 except that the failure of interest is the Turbine Driven AFW pump auxiliary lube oil pump at 9 inches of water. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 45 minutes to prevent eventual failure of the ability to start the TDAFP.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close 20

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALCOI] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for this HEP is 08-FPSISO45F-HE and the human error probability (HEP) is 6.6E-02.

5.2.24 08-FPSIS056F-HE - Detection and Isolation of a Large Flood due to a Steamline Break before Failing 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Steamline break resulting in a large Fire Protection System discharge in the Turbine Building. A Steamline break in the Turbine Building will result in an elevated building temperature that actuates multiple fire sprinklers.- This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent submergence failure of 480 VAC Buses 61 and 62 and the eventual isolation of the 4 kV Bus 6 motor loads due to the automatic tripping of the associated circuit breakers. Based on GOTHIC Q analysis [CALCOl] the operator must isolate the break within 56 minutes to accomplish these objectives.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews, the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALCOI] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for-this HEP is 08-FPSISO56F-HE and the human error probability (HEP) is 3.OE-02.

5.2.25 08-FPSISO68-F-HE - Detection and Isolation of a Large Flood due to a Steamline Break before Failure of the Motor Driven AFW Pumps The analysis of this BEP is documented in Attachment 1. This basic event applies only to a Steamline break resulting in a large Fire Protection System discharge in the Turbine Building.

This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinkders in time to prevent the submergence failure of the motor driven AFW pumps.

21

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods This event is identical to the one described in Section 5.2.22 except that the result of interest is submergence of the motor driven AFW pumps. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 68 minutes to prevent the submergence failure of the motor driven AFW pumps.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews, the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALC01] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing.. The basic event ID for this HEP is 08-FPSISO68-F-HE and the human error probability (HEP) is 3.OE-02.

5.2.26 08-FPSISO1CF-HE - Detection and Isolation of a Medium Flood due to a Steanmine Break before Eventual Isolation of the 4 kVAC Bus 5 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a (

Steamline break resulting in a moderate Fire Protection System discharge in the Turbine Building.

A Steamline break in the Turbine Building will result in an elevated building temperature that actuates multiple fire sprinlders. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinkders in time to prevent the eventual isolation of the 4 kVAC Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.

A Steamline pipe break in the Turbine Building (TU-22-1) would set off multiple fire sprinklers in the Turbine Building. This water could propagate through the open drain lines and under doors to Safeguards Alley (TU-90, TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Area TU-90 contains kVAC Bus 5 which could be failed due to propagation of a break in TU 1. This event analyzes a Steamline break resulting in a 2000-gpm discharge of the Fire Protection system.

Indication of this type of break would be provided by the Fire Pump Abnormal alarm in the control room and a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALC01] the operator must isolate the break 22

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods within 100 minutes to prevent eventual isolation of the 4 kV Bus 5 motor loads.

Thus, 100 minutes would be available to isolate the sprinkler flow following a Steamline pipe break to prevent eventual isolation of the 4 kV Bus 5 motor loads due to the automatic tripping of the associated circuit breakers. In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems.

Based on simulator observations and operator interviews, the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALCOII show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for this HEP is 08-FPSISO1CF-HE and the human error probability (HEP) is 3.OE-02.

5.2.27 08-FPSIS02CF-BE - Detection and Isolation of a Medium Flood due to a Steamline Break before Failing the Turbine Driven AFW Pump Auxiliary Lube Oil Pump The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Steamline break resulting in a moderate Fire Protection System discharge in the Turbine Building.

A Steamline break in the Turbine Building will result in an elevated building temperature that actuates multiple fire sprinklers. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent the eventual failure of the Turbine Driven AFW pump auxiliary lube oil pump.

This event is identical to the one described in Section 5.2.26 except that the failure of interest is the Turbine Driven AFW pump auxiliary lube oil pump at 9 inches of water. Based on GOTHIC analysis [CALCOI] the operator must isolate the break within 150 minutes to prevent eventual failure of the ability to start the TDAFP.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews, the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALC01] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall accident sequence timing. The basic event ID for this HEP is 08-FPSIS02CF-HE and the human error probability (HEP) is 3.OE-02.

5.2.28 08-FPSIS03CF-HE - Detection and Isolation of a Medium Flood due to a Steamnline Break 23

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods before Failing 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Steamline break resulting in a moderate Fire Protection System discharge in the Turbine Building.

A Steaniline break in the Turbine Building will result in an elevated building temperature that actuates multiple fire sprinklers. This basic event represents the failure of the operator to isolate flow from the Fire Pumps either by closing the manual discharge isolation valve on each pump, securing the power to the pumps, or closing the manual isolation valves for the sprinklers in time to prevent the eventual submergence failure of 480 VAC Buses 61 and 62 and the eventual isolation of the 4 kV Bus 6 motor loads due to the automatic tripping of the associated circuit breakers. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 170 minutes to accomplish these objectives.

In order to isolate the sprinklers the operators must receive the initial signal, decide the course of action, and execute isolation of the sprinklers, which first requires closing manual valves located on the mezzanine of the turbine building to isolate the four main sprinkler headers, and later close the supply valves to isolate the basement deluge systems. Based on simulator observations and operator interviews, the four main sprinkler headers would be isolated in 32 minutes. As described in Appendix D, GOTHIC analyses [CALC01] show that isolating flow from the basement deluge systems can be delayed for an additional 60 minutes without changing the overall 0

accident sequence timing. The basic event ID for this REP is 08-FPSISO3CF-HE and the human error probability (REP) is 3.OE-02.

5.2.29 04-CWSTP29-F-HE - Detection and Isolation of a Small Circulating Water Break before Eventual Isolation of the 4 kVAC Bus 5 Motor Loads The analysis of this BEP is documented in Attachment 1. This basic event applies only to a Small Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent the eventual isolation of the 4 kVAC Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.

A small rupture of an inlet or outlet condenser expansion joint in the Turbine Building (TU-22-1) could propagate through the open drain lines and under doors to Safeguards Alley (TU-90, TU-92, TU-95A, TU-95B-1, TU-95B-2 and TU-95C). Area TU-90 contains kVAC Bus 5 which could be failed due to propagation of a break in TU-22-1.

Indication of this type of break would be provided by a Miscellaneous Sump Level High alarm in the control room.

Propagation to Safeguards Alley will begin when the Turbine Building sump begins to fill since 24

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building foods the open drain lines from Safeguards Alley directly communicate with this sump. Additionally, when water begins to accumulate on the floor water will begin to leak under doors 4, 6, and 401 into Safeguards Alley. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 29 minutes to prevent eventual isolation of the 4 kV Bus 5 motor loads.

Thus, 29 minutes would be available to trip the Circulating Water pumps following a Small Circulating Water break to prevent eventual isolation of 4 kV Bus 5 motor loads due to the automatic tripping of the associated circuit breakers.. Based on simulator observations and operator interviews about 10 minutes are required to diagnose the cause of the high sump level alarm, decide the course of action, and execute the isolation. The basic event ID for this HEP is 04-CWSTP29-F-HE and the human error probability (BEP) is 4.3E-02.

5.2.30 04-CWSTP45-F-HE - Detection and Isolation of a Small Circulating Water Break before Failing the Turbine Driven AFW Pump Auxiliary Lube Oil Pump The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Small Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent the eventual failure of the Turbine Driven AFW pump auxiliary lube oil pump.

This event is identical to the one described in Section 5.2.29 except that the failure of interest is the Turbine Driven AFW pump auxiliary lube oil pump at 9 inches of water. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 45 minutes to prevent eventual loss of the ability to start the TDAFP.

Based on simulator observations and operator interviews about 10 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 04-CWSTP45-F-HE and the human error probability (HEP) is 1.7E-02.

5.2.31 04-CWSTP5 1-F-HE - Detection and Isolation of a Small Circulating Water Break before Failing 480 VAC Buses 61 and 62 and-Eventual Isolation of the 4 kVAC Bus 6 Motor Loads The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Small Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent submergence failure of 480 VAC Buses 61 and 62 and the eventual isolation of the 4 kV Bus 6 motor loads due to the automatic tripping of the associated circuit breakers. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 51 minutes to accomplish these objectives.

Based on simulator observations and operator interviews at least 10 minutes is required to receive 25

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 04-CWSTP5 1-F-HE and the human error probability (HEP) is 1.4E-02.

5.2.32 04-CWSTP60-F-HE - Detection and Isolation of a Small Circulating Water Break before Failure of the Motor Driven AFW Pumps The analysis of this HEP is documented in Attachment 1. This basic event applies only to a Small Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent submergence failure of the MDAFPs at 13 inches.

This event is identical to the one described in Section 5.2.29 except that the result of interest is submergence of the motor driven AFW pumps. Based on GOTHIC analysis [CALCOl] the operator must isolate the break within 60 minutes to prevent the submergence failure of the motor driven AFW pumps.

Based on simulator observations and operator interviews at least 10 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this REP is 04-CWSTP60-F-HE and the human error probability (HEP) is 1.2E-02.

5.2.33 04-CWSTP66-F-HE - Detection and Isolation of a Small Circulating Water Break before Water Level Reaches 18 Inches in the Turbine Building The analysis of this REP is documented in Attachment 1. This basic event applies only to a Small Circulating Water break in the Turbine Building. This basic event represents the failure of the operator to trip the Circulating Water pumps in time to prevent the water level from reaching 18 inches in the Turbine Building.

This event is identical to the one described in Section 5.2.29 except that the result of interest is 18 l inches of water in the Turbine Building. Based on GOTHIC analysis [CALC01] the operator must isolate the break within 66 minutes to prevent 18 inches of water in the Turbine Building.

Based on simulator observations and operator interviews at least 10 minutes is required to receive the initial signal, decide the course of action, and execute the isolation. The basic event ID for this HEP is 04-CWSTP66-F-BE and the human error probability (HEP) is 1.2E-02.l 5.2.34 16-BATCLG--F-BE - Establish Battery Room Cooling The analysis of this REP is documented in Attachment 1. This basic event applies to flooding scenarios where the 480 V buses have failed, thereby causing failure of normal battery room cooling. After the Battery Room A/B Exhaust Flow Low annunciator activates in the control (_

26

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods room, the operator is directed to use the fire equipment to ventilate the Battery Rooms. The air trunks and fans are then rigged to supply battery room cooling.

The operator must execute the action within 180 minutes to prevent excessive Battery Room heatup. Based on simulator observations and operator interviews about 77 minutes is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 16-BATCLG--F-HE and the human error probability (HEP) is 7.9E-02.

5.2.35 27A-ORR----F-HE - Failure to Throttle SI Flow to Conserve RWST Inventory The analysis of this HEP is documented in Attachment 1. This basic event applies to flooding scenarios where secondary cooldown has failed and the remaining SI pump is available. If secondary cooldown fails, the flow rate through existing RCP Seal LOCA is expected to worsen.

The operator would attempt to replace the lost RCS inventory using the available SI pump. Since high-pressure recirculation is unavailable due to the failure of the CCW pump power supplies, the operator must conserve the RWST inventory. This is done by manually throttling the SI pump discharge flow.

The operator must execute the action within 67 minutes to extend the time the RWST is available.

Based on simulator observations and operator interviews about 58 minutes is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 27A-ORR----F-HE and the human error probability (HEP) is 5.OE-03.

5.2.36 05B-BYALOP-F-HE - Failure to Bypass AFW Auxiliary Lube Oil Pressure Interlock The analysis of this HEP is documented in Attachment 1. This basic event applies to flooding scenarios where the water level in the AFW pump area has risen to 9 inches and the operator needs to start an AFW pump. If the auxiliary lube oil pump is failed due to submergence, then the associated AFW pump will not start due to a lube oil pressure interlock. This basic event addresses the bypass of this interlock to allow starting of the AFW pump.

The operator must execute the action within approximately 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> to restart an AFW pump.

Based on simulator observations and operator interviews about 3.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this REP is 05B-BYALOP-F-HE and the human error probability (HEP) is 4.4E-01.

5.2.37 06-NOLNDAFWF-BE - Failure to Feed Steam Generator Without Level Indication The analysis of this HEP is documented in Attachment 1. This basic event applies to flooding scenarios where power to the instrument bus is failed and AFW operation is required to maintain steam generator level. The operator must then provide makeup to the steam generators without 27

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods instrument power to allow monitoring of steam generator level and prevent overfilling the steam generator and failing the TDAFP.

The basic event ID for this HEP is 06-NOINDAFWF-HE and the human error probability (BEP) is 6.4E-01.

5.2.38 06--OC2----F-HE - Failure to Perform RCS Cooldown Using Natural Circulation The analysis of this HEP is documented in Attachment 1. This basic event applies to flooding scenarios where RXCP seal cooling systems, i.e., charging and CCW, are not failed by the flooding event, but fail randomly shortly into the event. For this event, the operators must cooldown and depressurize the RCS per ES-0.2.

The operator must execute the action within approximately 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> to perform cooldown. Based on simulator observations and operator interviews about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 06-

-OC2----F-HE and the human error probability (HEP) is 7.4E-02.

5.2.39 06--OC6----F-HE - Failure to Perform RCS Cooldown with Boration The analysis of this HEP is documented in Attachment 1. This basic event applies to flooding scenarios where 480 VAC power is lost to all charging and CCW pumps. In these scenarios, a RXCP seal LOCA is assumed to occur and the operators would cooldown and depressurize the RCS per ES-1.2.

The operator must execute the action within approximately 200 minutes to perform cooldown.

Based on simulator observations and operator interviews about 65 minutes is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 06--OC6----F-HE and the human error probability (HEP) is 9.2E-02.

5.2.40 05B-MDPTD36F-HE - Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (36 Minutes)

The analysis of this HEP is documented in Attachment 1. This basic event applies to the Moderate Circulating Water pipe break. After the Motor Driven AFW pumps have failed, the operator must start the Turbine Driven AFW pump within 36 minutes of the initial pipe break to avoid submergence of the auxiliary lube oil pump and subsequent failure of the Turbine Driven AFW pump to start due to a lube oil pressure interlock. [CALC01]

Based on simulator observations and operator interviews about 18 minutes is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this 28

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods HEP is 05B-MDPTD36F-HE and the human error probability (HEP) is 4.7E-01.

5.2.41 05B-MDPTD49F-HE - Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (49 Minutes)

The analysis of this HEP is documented in Attachment 1. This basic event applies to the Feedwater pipe break with a large Fire Protection sprinkler discharge to the turbine building.

After the Motor Driven AFW pumps have failed, the operator must start the Turbine Driven AFW pump within 49 minutes of the initial pipe break to avoid submergence of the auxiliary lube oil pump and subsequent failure of the Turbine Driven AFW pump to start due to a lube oil pressure interlock. [CALCOI]

Based on simulator observations and operator interviews about 18 minutes is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 05B-MDPTD49F-HE and the human error probability (HEP) is 4.7E-01.

5.2.42 05B-MDPTD61F-HE - Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (61 Minutes)

The analysis of this HEP is documented in Attachment 1. This basic event applies to the Large Service Water, Fire Protection Water, and Steamline pipe breaks with a large Fire Protection sprinkler discharge. After the Motor Driven AFW pumps have failed, the operator must start the Turbine Driven AFW pump within 61 minutes of the initial pipe break to avoid submergence of the auxiliary lube oil pump and subsequent failure of the Turbine Driven AFW pump to start due to a lube oil pressure interlock. [CALCO1]

Based on simulator observations and operator interviews about 18 minutes is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 05B-MDPTD6 IF-HE and the human error probability (HEP) is 3. IE-01.

5.2.43 05B-MDPTDlCF-HE - Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (109 Minutes)

The analysis of this REP is documented in Attachment 1. This basic event applies to the Feedwater pipe break with a moderate Fire Protection sprinkler discharge to the turbine building.

After the Motor Driven AFW pumps have failed, the operator must start the Turbine Driven AFW pump within 109 minutes of the initial pipe break to avoid submergence of the auxiliary lube l oil pump and subsequent failure of the Turbine Driven AFW pump to start due to a lube oil pressure interlock. [CALCOI1 Based on simulator observations and operator interviews about 18 minutes is required to receive 29

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 05B-MDPTD1CF-HE and the human error probability (HEP) is 1.6E-01.

5.2.44 05B-MDPTD2HF-HE - Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (2.5 Hours).

The analysis of this HEP is documented in Attachment 1. This basic event applies to the Steamline pipe break with a moderate Fire Protection sprinkler discharge to the turbine building.

After the Motor Driven AFW pumps have failed, the operator must start the Turbine Driven AFW pump within 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> of the initial pipe break to avoid submergence of the auxiliary lube oil pump and subsequent failure of the Turbine Driven AFW pump to start due to a lube oil pressure interlock. [CALC01]

Based on simulator observations and operator interviews about 18 minutes is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 05B-MDPTD2HF-HE and the human error probability (HEP) is 4. IE-02.

5.2.45 86-INSTRRCRF-HE - Failure to Recover AFW Control The analysis of this HEP is documented in Attachment 1. This basic event applies to all scenarios where AFW flow exists and at least one train of safety-related AC power is available. Under these conditions the operator is instructed to control AFW flow using the AFW pump discharge valves to adjust the flow rate to the steam generators. If these valves cannot be controlled remotely from the control room, then the operator has approximately 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br /> to operate them manually from the pump room. [CALC01].

Based on simulator observations and operator interviews about 9.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> is required to receive the initial signal, decide the course of action, and execute the action. The basic event ID for this HEP is 86-INSTRRCRF-HE and the human error probability (HEP) is 1.8E-02.

5.3 DATA The KNPP.BED database was used for the flooding analysis. The Turbine Building flooding initiators and the HEPs discussed in Section 5.2 were added toKNPP.BED, along with the basic events modeled in fault tree FLOODING.LGC. One other new basic .event was also added to the database:

Basic Event 05B-FRACTDP-OFF represents the fraction of time that the operator is expected to trip the Turbine Driven AFW pump early in a flooding event given that both Motor Driven AFW pumps have successfully started.

30

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods No other basic events were added.

Table 1 lists all of the new operator actions and their HEPs that were developed in support of the flooding analysis.

Table 2 lists all the new basic events (and their values) that were added to KNPP.BED in support of the flooding analysis.

6.0 MODEL EVALUATION (EQUATIONS)

Each fault tree used to represent an event tree top event in this analysis is quantified various times under different initial conditions. Each of these fault tree quantifications produces an equation.

The equation's location on the event tree (i.e., which previous top events have succeeded and failed) dictates the initial conditions used to quantify the fault tree and develop that unique equation. Such initial conditions are modeled by setting to TRUE the failure rates of equipment that is known to be unavailable due to the flooding event. The same fault tree is then quantified with different initial conditions to yield different equations.

With the exception of Top Events AFZ and AFX (which are described in more detail in the

@ following subsection), this analysis generally develops two unique equations for each top event.

The initial conditions for' the first equation consist of the flood-induced failures of equipment in the Turbine Building as well as the bottom row of breakers on Buses 51/52 and 61/62. This represents the flood-induced equipment failures that occur very early in the event. For example, when analyzing the Large Feedwater scenario (WI06B) the equation for Top Event AFR using these initial conditions is named AFRWI06B The initial conditions for the second equation simply build on those of the first equation. In addition to the equipment failures of the first equation, the second equation adds the flood-induced failures of 480 VAC Buses 61 and 62. Thus, the second equation is quantified assuming the failure of 480 VAC Train B safety-related power. ;For example, when analyzing the Large Feedwater scenario (WI06B) the equation for Top Event AFR using these initial conditions is named AFRWI064.

6.1 EVALUATION OF TOP EVENT AFZ Instead of quantifying the same fault tree various times to develop different equations, the equations associated with Top Event AFZ use multiple fault trees that are quantified a single time.

This is due to the complexity added by various human actions and the potential of equipment to already be running. Only the equations associated with a Large Feedwater Break are described here (e.g., AFZ-AWIB). The descriptions of the equations associated with all other initiators (e.g., AFZ-ACXB, AFZ-ASIB, AFZ-ATIB, etc.) are identical except for the name of the initiator 31

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods and the timing associated with the model.

6.1.1 AFZ-AWIB This equation is used to model failure of the operator to provide AFW flow for decay heat removal using the turbine-driven AFW pump. The equation is used to model operator action-related failures only. Hardware-related failures of the turbine-driven AFW train are evaluated by other equations in the event tree.

This equation is used for sequences with the following conditions:

• The water volume released to the turbine building will result in opening the bottom row of breakers on 480 VAC Buses 51, 52, 61, and 62.

• The break was isolated before the volume of water released in the turbine building would cause water level in the AFW pump rooms to submerge the TDAFP auxiliary lube oil pump.

• The break was isolated before the volume of water released in the turbine building would cause water level in the B-train 4kVAC room to reach a level that would fail 4 kVAC Bus 6 motor loads or 480 VAC Buses 61/62.

• The water volume released to the turbine building will result in failure of 4kVAC Bus 5 motor loads.

• The B-train motor-driven AFW pump successfully started.

• The B-train motor-driven AFW pump failed to run. By definition of the sequences where equation AFZ-AWIB is used, the mission time for the AFW pump is 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. On average, the pump is assumed to run halfway through the mission time or 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.

For the sequences where equation AFZAW1B is used, several potential success paths exist.

First, the operators may have recognized that the flooding event could threaten the motor-driven AFW pumps and would maintain the turbine-driven AFW pump running throughout the event.

Second, if the turbine-driven AFW pump were secured, then restart would merely require that the operators take the control switch from pull-to-lock. Then, even if the operators did not start the pump, it would automatically start on a low-low steam generator level signal.

6.1.2 AFZ-BWIB This equation is used to model failure of the operator to provide AFW flow for decay heat removal using the turbine-driven AFW pump. The equation is used to model operator action-related failures only. Hardware-related failures of the turbine-driven AFW train are evaluated by other equations in the event tree.

This equation is used for sequences with the following conditions:

32 C)

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods

• The water volume released to the turbine building will result in opening the bottom row of breakers on 480 VAC Buses 51, 52, 61, and 62.

• The water volume released to the turbine building will result in failure of 4kVAC Bus 5 motor loads.

• The water volume released to the turbine building would cause water level in the AFW pump rooms to submerge the TDAFP auxiliary lube oil pump.

• The break was isolated before the volume of water released in the turbine building would cause water level in the B-train 4kVAC room to reach a level that would fail 4 kVAC Bus 6 motor loads or 480 VAC Buses 61/62.

• The B-train motor-driven AFW pump successfully started.

• The B-train motor-driven AFW pump failed to run. By definition of the sequences where equation AFZ-BWIB is used, the mission time for the AFW pump is 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. On average, the pump is assumed to run halfway through the mission time or 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.

For the sequences where equation AFZ-BWIB is used, multiple potential success paths exist.

First, the operators could have maintained the turbine-driven AFW pump running throughout the event. The pump would be maintained running if either motor-driven AFW pump failed or if the operators recognized that the flooding event could threaten the motor-driven AFW pumps and would want the added reliability of the third AFW pump. Second, even if the turbine-driven AFW pump was secured early in the event, then the operators could recognize that the rising water levels would soon threaten the motor-driven AFW pumps and may restart the turbine-driven AFW pump. By definition of the sequences where equation AFZ-BWIB is used, water level will reach a level that will submerge the turbine-driven AFW pump auxiliary lube oil pump.

Therefore, for the operators to successfully start the pump from the control room, action must be taken before 49 minutes. Otherwise, the auxiliary lube oil pump would be submerged, thereby preventing the turbine-driven AFW pump from starting.

If the turbine-driven pump is not started within 49 minutes after flood initiation and then maintained running, then the pump could be started if the low oil pressure interlock is bypassed.

Bypass of the low oil pressure interlock may be directed by personnel manning the technical support center and would need to be completed before water level in either of the steam generators dropped to less than 5-percent wide range, the point that bleed and feed cooling would be initiated.

Given the definition of the sequences where equation AFZ-BWIB is used, the B-train motor-driven AFW pump started, but failed to run. Since, on average, the pump is assumed to fail one-half way through the mission time, or 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />, steam generator water level would be at or near normal level when flow from the motor-driven AFW pump is lost. Previous analyses have shown that about three hours are required for water level in the steam generators to decrease from

_ nominal to 5-percent wide range. Therefore, three hours would be available to bypass the 33

lINTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods interlock.

6.1.3 AFZBWI4 This equation is used to model failure of the operator to provide AFW flow for decay heat removal using the turbine-driven AFW pump. The equation is used to model operator action-related failures only. Hardware-related failures of the turbine-driven AFW train are evaluated by other equations in the event trees.

This equation is used for sequences with the following conditions:

• The water volume released to the turbine building will result in opening the bottom row of breakers on 480 VAC Buses 51, 52, 61, and 62.

• The water volume released to the turbine buding will result in failure of 4kVAC Bus 5 motor loads.

• The water volume released to the turbine building would cause water level in the AFW pump rooms to submerge the TDAFP auxiliary lube oil pump.

• The water volume released to the turbine building would cause water level in safeguards alley to submerge 480 VAC Buses 61 and 62.

• The water volume released to the turbine building would cause water level in the B-train 4kVAC room to reach a level that would fail 4 kVAC Bus 6 motor loads.

• The break was isolated before the volume of water released in the turbine building would cause water level in the safeguards alley to fail the turbine-driven AFW pump if it was already running.

For the sequences where equation AFZ-BWI4 is used, the only method available for long-term decay heat removal is the turbine-driven AFW pump. Although the B-train motor-driven AFW pump may start and provide flow, the pump will be lost when water level on 4 kVAC Bus 6 reaches the level at which bus failure is expected. Therefore, no credit is taken for operation of the B-train motor-driven AFW pump.

Multiple potential success paths exist for the conditions where equation AFZ-BWI4 is used.

First, the operators could have maintained the turbine-driven AFW pump running throughout the event. The pump would be maintained running if either motor-driven AFW pump failed or if the operators recognized that the flooding event could threaten the motor-driven AFW pumps and would want the added reliability of the third AFW pump. Second, even if the turbine-driven AEW pump was secured early in the event, then the operators could recognize that the rising water levels would soon threaten the motor-driven AFW pumps and may restart the turbine-driven AFW pump. By definition of the sequences where equation AFZ-BWIB is used, water level will reach a level that will submerge the turbine-driven AFW pump auxiliary lube oil pump.

Therefore, for the operators to successfully start the pump from the control room, action must be 34

iINTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods taken within 49 minutes. Otherwise, the auxiliary lube oil pump would be submerged, thereby preventing the turbine-driven AFW pump from starting.

If the turbine-driven pump is not started within 49 minutes after flood initiation and then maintained running, then the pump could be started if the low oil pressure interlock is bypassed.

Bypass of the low oil pressure interlock may be directed by personnel manning the technical support center and would need to be completed before water level in either of the steam generators dropped to less than 5-percent wide range, the point that bleed and feed cooling would be initiated.

6.2 EVALUATION OF TOP EVENT AFX Instead of quantifying the same fault tree various times to develop different equations, the equations associated with Top Event AFX use multiple fault trees that are quantified a single time. This is due to the complexity added by various human actions and the potential of equipment to already be running. Only the equations associated with a Large Feedwater Break are described here (e.g., AFX-1WIB). The descriptions of the equations associated with all other initiators (e.g., AFX-1CXB, AFX-1SIB, AFX-1TIB, etc.) are identical except for the name of the initiator and the timing associated with the model.

~ 6.2.1 AFX-1WIB This equation is used to model failure of the operator to control AFW flow to maintain level in the steam generators.

This equation is used for sequences with the following conditions:

• The water volume released to the turbine building will result in opening the bottom row of breakers on 480 VAC Buses 51, 52, 61, and 62. i

• The break was isolated before the volume of water released in the turbine building would cause water level in the AFW pump rooms to submerge the TDAFP auxiliary lube oil pump.

• The break was isolated before the volume of water released in the turbine building would cause water level in the B-train 4kVAC room to reach a level that would fail 4 kVAC Bus 6 motor loads or 480 VAC Buses 61/62.

• The water volume released to the turbine building will result in failure of 4kVAC Bus 5 motor loads.

• The B-train motor-driven AFW pump successfully started.

• The B-train motor-driven AFW pump can successfully run for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

By definition, all equipment in the turbine building basement is assumed failed by the initiating 35

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods event. Therefore, service air compressors are lost. Also by definition of the sequences where equation AFX-1WIB is used, the B-train electrical safety buses (4kVAC and 480 VAC) are available. Because the B-train 480 VAC safety buses are available, the B-train instrument air compressor, ClB, is potentially available, as are the alternate power supplies to the 120 VAC instrument inverters. In addition, the A-train 480 VAC safety buses would be available so instrument air compressor CIC is potentially available.

Even though the B-train 480 VAC safety buses are available, the bottom row of breakers on the 480 VAC buses will have opened. When these breakers open, several loads that impact the flooding accident sequence progression are lost. These loads include the power supply to the associated train battery room fan cooling units, battery chargers, standby power supplies for 120 VAC instrument inverters, battery room exhaust fans, and auxiliary lube oil pumps for the motor-driven AFW pumps.

Given the conditions described above, success of the AFX-1WIB equation can be achieved by several means. First the B-train motor-driven AFW pump could be maintained running and flow controlled using AFW-2B. If air and power are available, then flow can be controlled from the control room. Air could be supplied from instrument air compressors ClB or C1C and power is provided from panel BRD-115, which is supplied with power from either battery BRD-101 or MCC-62C. These power sources can be backed up by DC distribution cabinet BRC-102 via either battery BRC-101 or MCC-46C. Given the redundancy and diversity of these four power supplies, explicit consideration of their failure is assumed to be insignificant and need not be modeled. If air is not available, then AFW-2B can be operated locally.

Second, if the operators secure the B-train AFW pump, then the turbine-driven AFW pump can be used. If the turbine-driven AFW pump was maintained running, then no additional actions are required. If the turbine-driven AFW pump was secured, then taking the control switch from pull-to-lock would restart the pump when level reached the low-low setpoint. Once the turbine-driven AFW pump is running, flow can be controlled using valves AFW-10A/B. For either of these options, either a long-term source of DC power must be provided for instrumentation or steam generator level must be controlled following a loss of all level indication. Lastly, if the motor-driven AFW pump 1B has been secured, then the low oil pressure interlock can be bypassed to allow starting the motor-driven AFW pumps without the auxiliary lube oil pumps.

Provision of a long-term source of DC power can be ensured by multiple means for sequences involving equation AFX-1WIB. First, the 120 VAC instrument inverters can be aligned to their alternate power source. Since 480 VAC Buses 61 and 62 are available, the alternate sources are available. Evaluations have shown that if the instrument inverters are removed from battery BRB-101, then the battery can supply needed DC loads for well in excess of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Alignment of the instrument inverters to their alternate power source also ensures that steam generator level indication is available in the control room even if the battery fails or is depleted. If needed, an 36

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods alternate power source can be aligned to the DC buses. Alternatives include installation of an alternate power source to the existing battery charger or installation of a spare battery charger with power from an alternative source.

Given the availability of equipment for sequences where equation AFX-1WIB is used and the multiple success paths that are available, it is likely that many hours would be available for the operators to initiate the actions. Therefore, time would not be critical to completing any of the actions and explicit evaluation of timing is not necessary.

6.2.2 AFX-2WIB This equation is used to model failure of the operator to control AFW flow to maintain level in the steam generators.

This equation name is used for sequences with the following conditions:

• The water volume released to the turbine building will result in opening the bottom row of breakers on 480 VAC Buses 51, 52, 61, and 62.

• The break was isolated before the volume of water released in the turbine building would cause water level in the B-train 4kVAC room to reach a level that would fail 4 kVAC Bus 6 motor loads or 480 VAC Buses 61/62.

• The water volume released to the turbine building will result in failure of 4kVAC Bus 5 motor loads.

• The B-train motor-driven AFW pump successfully started.

• The B-train motor-driven AFW pump failed to run. By definition of the sequences where equation AFX-2WIB is used, the mission time for the AFW pump is 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. On average, the pump is assumed to fail to run halfway through the mission time or 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.

• The turbine-driven AFW pump has been started and can successfully operate for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

By definition, all equipment in the turbine building basement is assumed failed by the initiating event. Therefore, service air compressors are lost. Also by definition of the sequences where equation AFX-2WIB is used, the B-train electrical safety buses (4kVAC and 480 VAC) are available. Because the B-train 480 VAC safety buses are available, the B-train instrument air compressor, CIB, is potentially available, as are the alternate power supplies to the 120 VAC instrument inverters. In addition, the A-train 480 VAC safety buses would be available so instrument air compressor CIC is potentially available.

Even though the B-train 480 VAC safety buses are available, the bottom row of breakers on the 480 VAC buses will have opened. When these breakers open, several loads that impact the flooding accident sequence progression are lost. These loads include the power supply to the

_ associated train battery room fan cooling units, battery chargers, standby power supplies for 120 37

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods VAC instrument inverters, battery room exhaust fans, and auxiliary lube oil pumps for the motor-driven AFW pumps.

Given the conditions described above, success of the AFX-2WIB equation can be achieved by controlling AFW flow using valves AFW-1OA/B. If a long-term source of DC power is available, then steam generator level can be controlled from the control room. Provision of a long-term source of DC power can be ensured by multiple means for sequences involving equation AFX-2WIB. First, the 120 VAC instrument inverters can be aligned to their alternate power source.

Since 480 VAC Buses 61 and 62 are available, the alternate sources are available. Evaluations have shown that if the instrument inverters are removed from battery BRB-101, then the battery can supply needed DC loads for well in excess of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Alignment of the instrument inverters to their alternate power source also ensures that steam generator level indication is available in the control room even if the battery fails or is depleted. If needed, an alternate power source can be aligned to the DC buses. Alternatives include installation of an alternate power source to the existing battery charger or installation of a spare battery charger with power from an alternative-source.

Given the availability of equipment for sequences where equation AFX-2WIB is used and the multiple success paths that are available, it is likely that many hours would be available for the operators to initiate the actions. Therefore, time would not be critical to completing any of the actions and explicit evaluation of timing is not necessary.

6.2.3 AFX-lAWI This equation is used to model failure of the operator to control AFW flow to maintain level in the steam generators.

This equation is used for sequences with the following conditions:

• The water volume released to the turbine building will result in opening the bottom row of breakers on 480 VAC Buses 51, 52, 61, and 62.

• The water volume released to the turbine building will result in submergence of the TDAFP auxiliary lube oil pump.

• The break was isolated before the volume of water released in the turbine building would cause water level in the B-train 4kVAC room to reach a level that would fail 4 kVAC Bus 6 motor loads or 480 VAC Buses 61/62.

• The water volume released to the turbine building will result in failure of 4kVAC Bus 5 motor loads.

• The B-train motor-driven AFW pump successfully started.

• The B-train motor-driven AFW pump can successfully run for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

38

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods By definition, all equipment in the turbine building basement is assumed failed by the initiating event. Therefore, service air compressors are lost. Also by definition of the sequences where equation AFX-lAWI is used, the B-train electrical safety buses (4kVAC and 480 VAC) are available. Because the B-train 480 VAC safety buses are available, the B-train instrument air compressor, ClB, is potentially available, as are the alternate power supplies to the 120 VAC instrument inverters. In addition, the A-train 480 VAC safety buses would be available so instrument air compressor ClC is potentially available.

Even though the B-train 480 VAC safety buses are available, the bottom row of breakers on the 480 VAC buses will have opened. When these breakers open, several loads that impact the flooding accident sequence progression are lost. These loads include the power supply to the associated train battery room fan cooling units, battery chargers, standby power supplies for 120 VAC instrument inverters, battery room exhaust fans, and auxiliary lube oil pumps for the motor-driven AFW pumps.

Given the conditions described above, success of the AFX-lAWI equation can be achieved by several means. First the B-train motor-driven AFW pump could be maintained running and flow controlled using AFW-2B. If air and power are available, then flow can be controlled from the control room. Air could be supplied from instrument air compressors ClB or C1C and power is YJ provided from panel BRD-115, which is supplied with power from either battery BRD-101 or MCC-62C. These power sources can be backed up by DC distribution cabinet BRC-102 via either battery BRC-101 or MCC-46C. Given the redundancy and diversity of these four power supplies, explicit consideration of their failure is assumed to be insignificant and need not be modeled. If air is not available, then AFW-2B can be operated locally.

Second, if the operators secure the B-train AFW pump, then the turbine-driven AFW pump can be used. By definition of the sequences where equation AFX-1AWI is used, water level will reach a level that will submerge the turbine-driven AFW pump auxiliary lube oil pump.

Therefore, for the operators to successfully start the pump from the control room, action must be taken within 49 minutes. Otherwise, the auxiliary lube oil pump would be submerged, thereby preventing the turbine-driven AFW pump from starting.

If the turbine-driven pump is not started within 49 minutes after flood initiation and then maintained running, then the pump could be started if the low oil pressure interlock is bypassed.

Bypass of the low oil pressure interlock may be directed by personnel manning the technical support center and would need to be completed before water level in either of the steam generators dropped to less than 5-percent wide range, the point that bleed and feed cooling would be initiated.

For either of these options, either a long-term source of DC power must be provided for

_p instrumentation or steam generator level must be controlled following a loss of all level indication.

39

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Provision of a long-term source of DC power can be ensured by multiple means for sequences involving equation AFX-lAWI. First, the 120 VAC instrument inverters can be aligned to their alternate power source. Since 480 VAC Buses 61 and 62 are available, the alternate sources are available. Evaluations have shown that if the instrument inverters are removed from battery BRB-101, then the battery can supply needed DC loads for well in excess of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Alignment of the instrument inverters to their alternate power source also ensures that steam generator level indication is available in the control room even if the battery fails or is depleted. If needed, an alternate power source can be aligned to the DC buses. Alternatives include installation of an alternate power source to the existing battery charger or installation of a spare battery charger with power from an alternative source.

40

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 6.2.4 AFX-2WI4 This equation is used to model failure of the operator to control AFW flow to maintain level in the steam generators.

This equation is used for sequences with the following conditions:

• The water volume released to the turbine building will result in opening the bottom row of breakers on 480 VAC Buses 51, 52, 61, and 62.

• The water volume released to the turbine building will result in failure of 4kVAC Bus 5 motor loads.

• The water volume released to the turbine building would cause water level in the AFW pump rooms to submerge the TDAFP auxiliary lube oil pump.

• The water volume released to the turbine building would cause water level in safeguards alley to submerge 480 VAC Buses 61 and 62.

• The water volume released to the turbine building would cause water level in the B-train 4kVAC room to reach a level that would fail 4 kVAC Bus 6 motor loads.

• The break was isolated before the volume of water released in the turbine building would cause water level in the safeguards alley to fail the turbine-driven AFW pump if it was already running.

• The turbine-driven AFW pump has been started and can successfully operate for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

By definition of the sequences where equation AFX-2WI4 is used, all AC power will be lost. In addition, all DC power may eventually be lost because of the loss of power to the battery chargers.

Success of the AFX-2WI4 equation requires that the operators control flow using valves AFW-1OA/B. In addition, either a long-term source of DC power must be provided for instrumentation or steam generator level must be controlled following a loss of all level indication.

7.0 REEERENCES

[CALC01] Calculation 0064-0515-LYS-01, Evaluation of Flooding Levels for Various PRA Cases, Revision 2, MPR Associates, Inc.

[NBO1] KPS Internal Events PRA, Volumes 2 through 4.

[NB02] KPS Internal Events PRA, Section 4.15, "Human Reliability Analysis."

41

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Table 1 Summary of KPS Turbine Building Flood Hunan Actions Basic Event ID Basic Event Description HEP 02-SW4A-B29F-HE Detection and Isolation of a Service Water Break before Eventual Isolation of the 4 2.OE-02 kVAC Bus 5 Motor Loads 02-SW4A-B45F-HE Detection and Isolation of a Service Water Break before Failing the Turbine Driven 2.OE-02 AFW Pump Auxiliary Lube Oil Pump 02-SW4A-B5lF-HE Detection and Isolation of a Large Service Water Break before Failing 480 VAC 2.OE-02 Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 02-SW4A-B60F-HE Detection and Isolation of a Large Service Water Break before Failure of the 2.0E-02 Motor Driven AFW Pumps 02-SW4A-B66F-HE Detection and Isolation of a Large Service Water Break before Water Level 2.OE-02 Reaches 18 Inches in the Turbine Building 04-CWSTP13-F-HE Detection and Isolation of a 14,000-gpm Circulating Water Break before Eventual 2.6E-01 Isolation of the 4 kVAC Bus 5 Motor Loads 04-CWSTP19-F-HE Detection and Isolation of a 14,000-gpm Circulating Water Break before Failing the 1.2E-01 Turbine Driven AFW Pump Auxiliary Lube Oil Pump 04-CWSTP22-F-HE Detection and Isolation of a 14,000-gpm Circulating Water Break before Failing 1.2E-01 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 04-CWSTP25-F-HE Detection and Isolation of a 14,000-gpm Circulating Water Break before Water 1.2E-01 Level Reaches 18 Inches in the Turbine Building 04-CWSTP29-F-HE Detection and Isolation of a Small Circulating Water Break before Eventual 4.3E-02 Isolation of the 4 kVAC Bus 5 Motor Loads 42 C, C C

4: C C INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Table 1 Sunmary of KPS Turbine Building Flood Human Actions Basic Event ID Basic Event Description HEP 04-CWSTP45-F-HE Detection and Isolation of a Small Circulating Water Break before Failing the 1.7E-02 Turbine Driven AFW Pump Auxiliary Lube Oil Pump 04-CWSTP51-F-HE Detection and Isolation of a Small Circulating Water Break before Failing 480 VAC 1.4E-02 Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 04-CWSTP60-F-HE Detection and Isolation of a Small Circulating Water Break before Failure of the 1.2E-02 Motor Driven AFW Pumps 04-CWSTP66-F-HE Detection and Isolation of a Small Circulating Water Break before Water Level 1.2E-02 Reaches 18 Inches in the Turbine Building 04-CW-TRIP-F-BE Detection and Isolation of a 58,000-gpm Circulating Water Break before Failing 1.0E1+00 Both 480 V Buses 05B-BYALOP-F-HE Failure to Bypass AFW Auxiliary Lube Oil Pressure Interlock 4.4E-01 05B-MDPTD1CF-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW 1.6E-01 Pump (108 Minutes) 05B-MDPTD2HF-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW 4. lE-02

_______ Pump (2 Hours) 05B-MDPTD36F-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW 4.7E-01 Pump (36 Mnutes) 05B-MDPTD49F-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW 4.7E-01 Pump (49 Mnutes) 05B-MDPTD61F-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW 3. lE-01 Pump (61 Mnutes) 06-NOINDAFWF-HE Failure to Feed Steam Generator Without Level Indication 6.4E-01 I 06--OC2----F-HE Failure to Perform RCS Cooldown Using Natural Circulation 7.4E-02 43

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Table 1 Summary of KPS Turbine Building Flood Hmnan Actions Basic Event II) Basic Event Description HEP 06--OC6----F-HE Failure to Perform RCS Cooldown with Boration- 9.2E-02 08-FPISO29-F-HE Detection and Isolation of a Fire Protection Water Break before Eventual Isolation 1.0E+00 of the 4 kVAC Bus 5 Motor Loads 08-FPISO45-F-HE Detection and Isolation of a Fire Protection Water Break before Failing the Turbine 6.6E-02 Driven AFW Pump Auxiliary Lube Oil Pump 08-FPISO56-F-HE Detection and Isolation of a Fire Protection Water Break before Failing 480 VAC 2.4E-02 Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 08-FPISO68-F-HE Detection and Isolation of a Fire Protection Water Break before Failure of the 1.6E-02 Motor Driven AFW Pumps 08-FPSISOICF-HE Detection and Isolation of a Medium Flood due to a Steamline Break before 3.OE-02 Eventual Isolation of the 4 kVAC Bus 5 Motor Loads 08-FPSISO29F-HE Detection and Isolation of a Large Flood due to a Steamihne Break before Eventual 1.0 Isolation of the 4 kVAC Bus 5 Motor Loads 08-FPSISO2CF-HE Detection and Isolation of a Medium Flood due to a Steamline Break before Failing 3.OE-02 the Turbine Driven AFW Pump Auxiliary Lube Oil Pump 08-FPSISO3CF-HE Detection and Isolation of a Medium Flood due to a Steamline Break before Failing 3.OE-02 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 08-FPSISO45F-HE Detection and Isolation of a Large Flood due to a Steamline Break before Failing 6.6E-02 the Turbine Driven AFW Pump Auxiliary Lube Oil Pump 08-FPSISO56F-HE Detection and Isolation of a Large Flood due to a Steamline Break before Failing 3.OE-02 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 44 C C (7

C 4: (

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Table 1 Suinmary of KPS Turbine Building Flood Human Actions Basic Event IED Basic Event Description HEP 08-FPSISO68F-HE Detection and Isolation of a Large Flood due to a Steamline Break before Failure of 3.OE-02 the Motor Driven AFW Pumps  ;

08-ISO-FS 18F-HE Detection and Isolation of a Large Flood due to a Feedwater Break before Eventual 1.OE+00 Isolation of the 4 kVAC Bus 5 Motor Loads 08-ISO-FS2HF-HE Detection and Isolation of a Medium Flood due to a Feedwater Break before 1.7E-02 Failing 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 08-ISO-FS33F-HE Detection and Isolation of a Large Flood due to a Feedwater Break before Failing 4.4E-01

_____ the Turbine Driven AFW Pump Auxiliary Lube Oil Pump 08-ISO-FS40F-HE Detection and Isolation of a Large Flood due to a Feedwater Break before Failing 1.3E-01 480 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 08-ISO-FS54F-HE Detection and Isolation of a Large Flood due to a Feedwater Break before Failure 3.OE-02 of the Motor Driven AFW Pumps 08-ISO-FS55F-HE Detection and Isolation of a Medium Flood due to a Feedwater Break before 3.0E-02 4 Eventual Isolation of the 4 kVAC Bus 5 Motor Loads 08-ISO-FS97F-HE Detection and Isolation of a Medium Flood due to a Feedwater Break before 3.OE-02 Failing the Turbine Driven AFW Pump Auxiliary Lube Oil Pump 16-BATCLG--F-HE Establish Battery Room Cooling 7.9E-02 27A-ORR----F-HE Failure to Throttle SI Flow to Conserve RWST Inventory 5.OE-03 86-INSTRRCRF-HE Failure to Recover AFW Control 1.8E-02 45

Table 2: Basic Events Added to KNPP.BED Basic Event ID Basic Event Description Point Estimate 02-S W4A-B29F-1HE Detection and Isolation of a Service Water Break before Eventual Isolation of the 4 kVAC 2.OE-02

____ ____ ____ us 5 M otor Loads_ _ _ _ _ _ _

2-SW4A-B45F-BE etcinand Isolation of a Service Water Break before Failing the Turbine Driven AFW 2.OE-02

_____________ ump Auxiliary Lube Oil Pump_____

2S W4A-B5 iF-BE Detection and Isolation of a Large Service Water Break before Failing 480 VAC Buses 61 2.OE-02

________________nd 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 02-S W4A-B 60F-1HE Detection and Isolation of a Large Service Water Break before Failure of the Motor Driven 2.OE-02

____AFWPumps__

02-S W4A-B66F-BE Detection and Isolation of a Large Service Water Break before Water Level Reaches 18 2.OE-02

____________ _____ nches in the Turbine Building_ _ _ _ _ _ _

0-CWSTP 13-F-HE Detection and Isolation of a 14,000-gpm Circulating Water Break before Eventual Isolation 2.6E-01

________________f the 4 kVAC Bus_5 Motor Loads 0-CWSTP19-F-BE Detection and Isolation of a 14,000-gpm Circulating Water Break before Failing the- 1.2E-01

____________ urbine Driven AFW Pump Auxiliary Lube Oil Pump

)-CWST'P22-F-BE Detection and Isolation of a 14,000-gpm Circulating Water Break before Failing 480 VAC 1.2E-01

________________uses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads

)-CWSTP25-F-1HE Detection and Isolation of a 14,000-gpm Circulating Water Break before Water Level 1.2E-01

________________eaches 18 Inches in the Turbine Building 0-CWSTP29-F-BE Detection and Isolation of a Small Circulating Water Break before Eventual Isolation of the 4.3E-02

_______________ kVAC Bus 5 Motor Loads 04CWSTP45-F-BE Detection and Isolation of a Small Circulating Water Break before Failing the Turbine 1.7E-02

_____________ riven AFW Pump Auxiliary Lube Oil Pump_____

46 C7 C C

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floodsl Table 2

Basic Events Added to KNPP.BED Basic Event ID Basic Event Description Point Estimate 04-CWSTP51-F-HE Detection and Isolation of a Small Circulating Water Break before Failing 480 VAC Buses 1.4E-02 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 04-CWSTP60-F-HE Detection and Isolation of a Small Circulating Water Break before Failure of the Motor 1.2E-02 Driven AFW Pumps 4-CWSTP66-F-HE Detection and Isolation of a Small Circulating Water Break before Water Level Reaches 18 1.2E-02 Inches in the Turbine Building 04-CW-TRIP-F-IHE Detection and Isolation of a 58,000-gpm Circulating Water Break before Failing Both 480 1.0E+00 V Buses

)5B-BYALOP-F-HE Failure to Bypass AFW Auxiliary Lube Oil Pressure Interlock 4.4E-01 D5B-MDPlD1CF-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (108 1.6E-01 Minutes) 05B-MDPTD2HF-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (2 4. 1E-02

- _ Hours) 05B-MDPTD36F-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (36 4.7E-01 Minutes) 05B-MDPTD49F-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (49 4.7E-01 Minutes) 05B-MDPTD61F-HE Failure to Start Turbine Driven AFW Pump Before Loss of Motor Driven AFW Pump (61 3.lE-01 Minutes) 06-NOINDAFWF-HE Failure to Feed Steam Generator Without Level Indication 6.4E-01 I 06--OC2----F-HE Failure to Perform RCS Cooldown Using Natural Circulation 7.4E-02 06--OC6----F-HE Failure to Perform RCS Cooldown with Boration 9.2E-02 47

lINTE RNAL FLOODING - Fault Tree Analysis for Turbine Building Floodsl Table 2: Basic Events Added to KNPP.BED Basic Event ID Basic Event Description Point Estimate 08-FPIS029-F-HE Detection and Isolation of a Fire Protection Water Break before Eventual Isolation of the 4 1.OE+00 kVAC Bus 5 Motor Loads 08-FPISO45-F-HE etection and Isolation of a Fire Protection Water Break before Failing the Turbine Driven 6.6E-02

__AFPump Auxiliary Lube Oil Pump 08-FPISO56-F-HE Detection and Isolation of a Fire Protection Water Break before Failing 480 VAC Buses 61 2.4E-02 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 8-FPISO68-F-HE Detection and Isolation of a Fire Protection Water Break before Failure of the Motor 1.6E-02 Driven AFW Pumps 8-FPSISOICF-HE Detection and Isolation of a Medium Flood due to a Steamline Break before Eventual 3.OE-02 Isolation of the 4 kVAC Bus 5 Motor Loads 08-FPSISO29F-HE Detection and Isolation of a Large Flood due to a Steamline Break before Eventual 1.0 Isolation of the 4 kVAC Bus 5 Motor Loads 08-FPSISO2CF-HBE Detection and Isolation of a Medium Flood due to a Steamline Break before Failing the 3.OE-02 Turbine Driven AFW Pump Auxiliary Lube Oil Pump

)8-FPSISO3CF-HE Detection and Isolation of a Medium Flood due to a Steamline Break before Failing 480 3.OE-02 VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads

)8-FPSISO45F-HE Detection and Isolation of a Large Flood due to a Steamline Break before Failing the 6.6E-02 Turbine Driven AFW Pump Auxiliary Lube Oil Pump

)8-FPSISO56F-HE Detection and Isolation of a Large Flood due to a Steamline Break before Failing 480 VAC 3.OE-02 Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads

)8-FPSISO68F-HE Detection and Isolation of a Large Flood due to a Steamline Break before Failure of the 3.OE-02 Motor Driven AFW Pumps 8-ISO-FS18F-HE Detection and Isolation of a Large Flood due to a Feedwater Break before Eventual 1.OE+00 Isolation of the 4 kVAC Bus 5 Motor Loads 48 C C C1

C C C INTERNAL FLJOODING - Fault Tree Analysis for Turbine Building Floodsl Table 2: Basic Events Added to KNPP.BED Basic Event ID Basic Event Description Point Estimate 08-ISO-FS2HF-BE Detection and Isolation of a Medium Flood due to a Feedwater Break before Failing 480 1.7E-02

.___ VAC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 08-ISO-FS33F-HE etection and Isolation of a Large Flood due to a Feedwater Break before Failing the 4.4E-O1 Turbine Driven AFW Pump Auxiliary Lube Oil Pump 08-ISO-FS40F-HE etection and Isolation of a Large Flood due to a Feedwater Break before Failing 480 1.3E-01

__ AC Buses 61 and 62 and Eventual Isolation of the 4 kVAC Bus 6 Motor Loads 8-ISO-FS54F-HE Detection and Isolation of a Large Flood due to a Feedwater Break before Failure of the 3.0E-02 Motor Driven AFW Pumps 8-ISO-FS55F-HE Detection and Isolation of a Medium Flood due to a Feedwater Break before Eventual 3.0E-02 Isolation of the 4 kVAC Bus 5 Motor Loads 08-ISO-FS97F-HE Detection and Isolation of a Medium Flood due to a Feedwater Break before Failing the 3.0E-02 Turbine Driven AFW Pump Auxiliary Lube Oil Pump 16-BATCLG--F-HE Establish Battery Room Cooling 7.9E-02 27A-ORR----F-E Failure to Throttle SI Flow to Conserve RWST Inventory 5.0E-03 86-INSTRRCRF-HE Failure to Recover AFW Control 1.8E-02 04-CW-MDAFPAMHE Operator Fails to Control MDAFP Med CW Break AC Avail 1.00E-01

)4-CW-TRIP-F-BE FAIL TO ISOL LRG CIRC WTR BRK BEFORE FAILURE OF 480V BUS 1.OOE+00 05B-FRACTDP-OFF Prob of Conditions Where TDAFP Is Secured 9.00E-01 CX06-ISOL-A Fail to Isolate Before Failure of any Buses CW Mod5.00-01 X06-ISOL-B Fail to Isolate Before Failure of AFWP CW Mod 5.00E-01 X06-ISOL-C Fail to Start MDAFP CW Moderate 5.00E-01 X06-ISOL-D Fail to Start MDAFP CW Moderate 5.00E_-0_1_

49

lINTERNAL FLOODING - Fault Tree Analysis for Turbine Building FloodsI Table 2: Basic Events Added to KNPP.BED Basic Event ID Basic Event Description Point Estimate FI06-ISOL-A Fail to Isolate Before Failure of any Buses FP Large 5.00E-01 FI06-ISOL-B Fail to Isolate Before Failure of AFWP FP Large 5.OOE-01 FI06-ISOL-C Fail to Start MDAFP FP Large 5.OOE-01 IE-CI06B LARGE CIRC WTR LINE BREAK IN TURB BLDG BASEMENT 4.76E-05 E-CX06B MEDIUM CIRC WIR LINE BREAK IN TURB BLDG BASEMENT 4.76E-05 E-CY06B SMALL CIRC WTR LINE BREAK IN TURB BLDG BASEMENT 7.34E-05 E-FI06B LARGE FIRE PROTECT LINE BREAK IN TURB BLDG BASEMENT 1.05E-04 E-SI06B LARGE SERVICE WTR LINE BREAK IN TURB BLDG BASEMENT 3.22E-05 E-TI06B STEAMLINE BRK IN TURB BLDG CAUSES LARGE FIRE PROTr 9.OOE-03 E-TX06B STEAMLINE BRK IN TURB BLDG CAUSES MEDIUM FIRE PROT 9.OOE-03 E-WI06B [GE FEDWATER BREAK IN TURBINE BLDG BASEMENT 9.41E-04 E-WX06B MEDIUM FEEDWATER BREAK IN TURBINE BLDG BASEMENT 9.41E-04 SI06-ISOL-A Fail to Isolate Before Failure of any Buses SW Large 1.00E+00 I06-ISOL-B Fail to Isolate Before Failure of AFWP SW Large 5.OOE-01 SI06-ISOL-C Fail to Start MDAFP SW Large 5.OOE-01 SI06-ISOL-D FAIL ISOLATION BEFORE 18 INCHES ON TDAFP SW LARGE 5.OOE-01 SL21-CD RCP SEAL LOCA GREATER THAN 21 GPM NO RCS COOLDOWN 2.OOE-01 SL21-NO-CD RCP SEAL LOCA GREATER THAN 21 GPM NO RCS COOLDOWN 6.OOE-01 106-ISOL-A Fail to Isolate Before Failure of any Buses STM Large 5.OOE-01 50 C C C

r C C INTERNAL FLFOODING - Fault Tree Analysis for Turbine Building Floods Table 2: Basic Events Added to KNPP.BED Basic Event ID Basic Event Description Point Estimate TI06-ISOL-B Fail to Isolate Before Failure of AFWP STM Large 5.00E-01 T106-ISOL-C Fail to Start MDAFP STM Large 5.OOE-01 TX06-ISOL-A Fal to Isolate Before Failure of any Buses STM Mod 5.OOE-01 TX06-ISOL-B Fal to Isolate Before Failure of AFWP STM Mod 5.00E-01 06-ISOL-C Fail to Start MDAFP STM Moderate 5.OOE-01 WI06-ISOL-A Fail to Isolate Before Failure of any Buses FW Large 1.OOE+00 WI06-ISOL-B Fall to Isolate Before Failure of AFWP FW Large 5.OOE-01 WI06-ISOL-C Fall to Start MDAFP FW Large 5.OOE-01 WX06-ISOL-A Fail to Isolate Before Failure of any Buses FW Mod 5.00E-01 WX06-ISOL-B Fal to Isolate Before Failure of AFWP FW Mod 5.OOE-01 WX06-ISOL-C rail to Start MDAFP FW Moderate 5.OOE-01 51

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods l Q

Figure 1- Fault Tree AFM.LGC O I 3 1 4 1 5 O.-th.Afi AFM, 2 AAnalyt: NA lCrestion Date'M3-21 R0i05on:10D-27D mtIOC.n a 1r I I ma.OmAM2IS IS Cl OSMIANID FAIIL S

2 CAFM2M DAIWPw. bflFW.*h A UlGA .

3 egrqa "4_

_OPFLS TOSTART Sm/FlWAlR FIW 5111r Fbh lDA"PW KSFORE

'.STANU STAMM W LO P.P .OFVA 5 ~ AF0 52

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods a I I

2 3

4 5

6 7

a 9

A a

Figure 1- Fault Tree AFM.LGC (continued) 53

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods I

u I I 3 S a

I 2

3 4

5 6

7 B

9 A

B C

Figure 1- Fault Tree AFM.LGC (continued) 0 54

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods U I 4 5 a l I

2 3

4 5

6 7

8 II Is 9

A C

Figure 1- Fault Tree AFM.LGC (continued) 55

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods l

U I1 I -21 4

• I &

I 2

3 4

6 6

7 8

9 A

a C

Figure 1- Fault Tree AFM.LGC (continued)

(4mJ 56

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods a I I 3 I 4 I 5 0

1 2

3 4

ffl-TR O OPMTCROW_ OPATORFALTO Doff 1 FO. I LE TOLOF iRMS-1W CM AME 5

I (W9 TEIU4T4 I (CFITT_

Lr. 16m B4-_^4.M #P(ji)S~I B

7 0

9 A

a C

Figure 1- Fault Tree AFM.LGC (continued) 57

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 0

i-/

I 2

3 4

5 I

a 9

A B

C Figure 1- Fault Tree AFM.LGC (continued) 58

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods

- I I I 2 1 3 I 4 1 5 0

elAtS9:

KA lCreationDate:ISD-33X5 Reso: 02.

aleLDC (Ro. S7)Vki4PRA3DWore.5M An.4P 2

4 5

13 6

7 8

S A

a C

Figure 1- Fault Tree AFM.LGC (continued) 59

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods U I 12 3 I 4 1 5 0 a I I AFM, 10

- .am Iyt: NA Creation Date;03205 I 10-27-2005 Revision:

BLOC 0w.S7)1vIRA*3 on na.n

___P.W I

4 . I 2

3 4

5 6

7 a

9 A

ID C

Figure 1- Fault Tree AFM.LGC (continued) 60

INTERNAL FLOODING - Fault Tree: Analysis for Turbine Building Floods Figure 1- Fault Tree AFM.LGC (continued) 61

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods I L It 2 I 3 I 4 5 o APM, 12 jAnalyst NA 1

Creation Date: 3C-205 1

10.27-235 IRevision: 1

-fSLOC Ot..S7)VkORA3D 1

2 3

4 5

6 C1AWTORFALSTO LO# OF SICS _ Fd r _ LOSSOFISMICE _A&S AORMLSTO OMAUAMLVWALWS TO TRAWA ICVMa" VAlUITASMIB OM MANUAL VALWS 7 WM"&=

e R cemm mo. S*41u AIER TST

- U

.h AM9 £Ja.6 . US L~!.... uivuaaAS Se e

0 A

B C

Figure 1- Fault Tree AFM.LGC (continued) 62

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods N 0V 1 2 3 4

_CLORAND tFAIF TO

>n X natyst: NA Cm stbn Dte: 09420 i O S1AILi TO I U EttRItitVE I WillAUXUJM YMWMT FPUMPU0STAN=lO OGOV .A *ND I Blom C11e~EA SWttwp r~-4cCNTU hAilto t 1FIS TO _ ttlIT TIPIAIWPWA EAT 4~W4JWOffE C6# l512F A 2 F I Falt TSred Figure 1- Fal ATree A#JLCcntnud 63

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods WmWMS41 GSPM-AMB-PR GAIMiaI GA1142AFM,142 1-TO TM To ye A CreationDate: W3- ri~nt-7~t ftn.GC44H57WiP &Sh~re.fl.d 11108MBET PALUREI.EcmO 1111"MN1 MALUR DOUBLECO&SON DiOflART FEMATt CAS FALUR (CCI) AU50iLMW FE1T CAUSEMALUES (CCF)

PMP a ITl UGBtAiSICSCCI22 PO3W SIMAS TO -U OGBPMORCCI2SX mBPmCCF415I IWIIM3P GPWRCCF23 156.m Q4 42 0 3 A4.m66 4 2450 8.434064 (2 2

1TO' 3 64_

I Aa.IRE FM~T n90"Em-COMMIN WOVAIW-44161. CAUSES P8AL0 (CCF)

TOO 095SV4W0WC23 uas.vrmQm ~~ M a cmm.m 5

MTO 3 a WWAM!SC I CNECRIVALVEM~3 "50 ICAr CI -10T 7 568.ATOOMP It IWiCT 7 omm~

66-U(2) I J.6 ~ I VALE 0133 138.6 I CAUS! FAILURE VCC) 8 56aw.of CGAIMNCAI 1I T OO 96 1 4 1 A T (3

GAFMo I 0130 VALVE AW4A 11 eM! P81.3 (CCI) 1-AT OT t.3.46? (4) 418 .76.IN a

8. N6BCI ( 2 )1 . 4 . 18 (j Figure 1- Fault Tree AFM.LGC (continued) 64

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods U 1 3 I , 4 . 1 5 TOOC I AFM, 16 Analyst:NA lCroation Dae:0932XS RWrosn: 10-27-2M6 ISLOC 0Go.7)WWS.PRa3,O MIVAIV AIm; I CA mm"AMICd TRANOSIEa TOPSIIAFAa cuTI, w -VUA4C GSOCCt2 1U-CV.NSIA4C COMMO MSAS 1 el_

0c4148C02 1 IM)N FXd I OBECMO 3 C M.MN Q CAR 011Oa° I EE CAS 1.207 240 22640 120406 240 $2MA COE1WFALURE(CCI) u.CVII SAIIIB-OCC F ICF IT LOSID0 I S. A TRANSCLS 4

1.28 (8 ) 2J2U4 24400 I CVJS18BC1 I 5

I c"16a 3 TO TO I CPIEC UTwACUREB1 NUS FA SCE ISTMAIWESCLOWE .I1"ALUc IW C I I "OCCC21 1TQ) 20 2(9 1.0.0 (W 2480 2 _86U~ 2.J0I 0

6~ COM 1100 W14-P "A

h lRAISFES CL aSswe X C mvd.n~o 9 Cl3mC I 6 "PIM11G.F0 I 44T lO TooIll 00. W~SCC) l AV011iCFAL l CAUSEFACUREMc TO0v1100 .WBIOCCFU2 NeSw"M'111CF0" *5S V O0SA11BF .wcC2

~~ 0.G). uCJ9C a

C Figure 1- Fault Tree AFM.LGC (continued) 65

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Figure 1- Fault Tree AFM.LGC (continued) 66

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 3 4 I 5 U

O CONmonC~

O"Im II AFM, 17 Analyst:NA Creation Date:WI3-2Xt6 lRvAsim: 10F27.M, F.Mv4wml-Fo OtO0CQ. 97A 'tRA3Dfrdn 4242400" T mm olFf14E NULCOMM0N amwse FALURE I

I eelw. TOf4 11.114aE oMa ITe 2 C C CIEftVALEtC"IC CAUSEVFALUE (CCF) 3 FALS TOOPEN I UtNBl C FT0 S3C-UM1C F0

_ Wa 06CV0WCFO 6

GAFILMI 1"ATO t E I j CAUSEALRE (C01) 6 tAl I.IAC eO lilarn

() lIa-aa (!)

l94PT~s~

c-TM4TO 9 44 9

TsuT40 Q)

P59.9TOIT FK ~aE r ,9,9PTAtSCW2 Co09 TDlaeaw3 3l FU1 US-Cw-mS1aA-1O Con C~ cC-C GAFUIT TFRI.

C1M T. FeLE-C0MM01 clrLbnsCWUSEVFLILtCCF)I I areafs AT0UIull COMMON2 C F TS0PI 4N`occn2 516sCo4WOCCF12 I I I oI 1 1 SS-C Ct2 ImW US S Wm W Figure 1- Fault Tree AFM.LGC (continued) 67

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building floods U I Iz I 110 a c.. aMat 115W Ca .V -

r_

l l AFM, 18

'. YmVno w"TO lAnalyst:NA Creaton Date: 3D2 IRaswn: 1027-2D5 l

__t10C Otv.Mf)t"RA3Ifs&.thn I

2 3

4 5

U~ilATOS ETO_. AL1At TIRAUAITO l

as5.1uaE CAIT ASa~o~ uaD OF magg WO.OSS FOR l 6 i . T t W TTE A-MST cTJIE 2?AOV F X4JE-I SlAG-I L36a-S C tS

("W/

9 A

B C

Figure 1- Fault Tree AFM.LGC (continued) 68

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods

• 4 I u I Aalyhat:

ARS 1Creatin0ete: 01 .10-MM R L O, I

_iio:

MMOEWIOLOC h 4 t E h ft 3 l r @o LAmeCWmcwm LARMe eawicwri LAREMMriUCT NEeRCICWM WIEDRW PEORMTE MIV LARGE T~ LIEBEAII LMEEEAKEINTUR WIENE@MUI WEEIMENU LINEUREMEIN TM OM DGASWMN I 1-SGBSEEN 115 GIPLA1111 GFLD114 @19 IJ E M BAS uUENT IU E GBA K.IuEI G"12 I LIM IUIESEDAKI LIKE BEAKe IN lcUR IE OReEM I PREAKEININNREE L=1109AEMN LARGESER)ICE R FIEPRTC eARG e&le.e (9 I 9 R-Co( ( a) 4ee0 e.6 4" ( a)

SIEAAIKeE WS STEMILE -R mimeR SLUGCAUSES *TWWKMDCAMjeE*

LAMEMME PROT MEcUMa FME PROT 3

I mBDCAUESI INTURIDSLGCAUSEES

(=Q9 IMO ILINE BEAKEE 41 (

A1 IIKEScAKIN TUDM ASMENI 1E.ZZ1

'nS.M 8

9 A

B Figure 2 - Fault Tree FLOODING.LGC 69

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods u I -- 3 I 4 I 5 FLOODING LOGIC FLD,2 a

[nlyst JRS lCnationiDate:01-10OMM Revision:

11-3D-2M5 FLOODL 0Caw.45Belch"PS o I f rfi of MOU I

IOn olds agftrh.f

. I O OFal VOV F F IM I wa&IW~wo Scftsm 2

<a~I SM0AI l l-W 4 3

4 6

6 7

6 9

A C

Figure 2 - Fault Tree FLOODING.LGC (continued)

(low 70

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 0 I 1- 2 I 3 I 4 I 5 FLOODING LOGICD, lFnOIGIGI FLD, 3 l Analyst:JRS Dale:01-10.200D lCreation Resion: I1-3-0065 Ml.OODMXC0t1r1S146)ee*'A3nV Uno 2ah WFLD3I@

2 &S tl e ,

cacC a o F 3

4 6

6 7

a 9

A 8

C Figure 2 - Fault Tree FLOODING.LGC (continued) 71

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods U I I d FLOODING LOGIC l FLD, 1 l7ye1:

RS lCrstion Dais: 01-1020 IRevsiom. 113-MS 3LOOLOCct~.1.1a1ae)tUAS30at.,nsn hit.balda hito Htb bd. W m0dw SataFWlb. dMd.d I .W I"

. h. 0RD414 GRLDtH5 2db.t..lWNaie .wWBaFWM.d l Hlei.

I IAL.w I Sal.. h1 GF.D4t3 a I I~dOS< L "i I' ald .M31MDI UIJOO (9 MW Rn ma l H lo #IN DAM B~d. a... I Sa wShrF-$TMt. id Gnim rP~i GF1B Ut.JHQ) Q I ra. haa a bald. hi.U S dMDA AlFab.d CAlal Usut(

I br.T4SOL.-CI hi.. Ut.JMft(9 U-a-uhit (9) bld. I d..5Fda a I hit. bakd Ut. bW.SMM

_ ,WbaaFL M62

.1M.

al.. hidt.wa 0.1b.efto*se

'--u (9 i d11# b l ^1*Std MUF W l I Sat.. e I GPLM h1im a LioJ hI t. t.at M tb44 1 l 455.

I 11 k-l (9 Iidaa.Fad.I C!

hitntal U

inLM 9 Lga Q 3.1.FdW mdt FM to "odJ

_awPlYLe 2 hit.0bald I I I if 0*

I amft I mom,o VW-&h I wFargI Le"u4_ (9 Ik

.tMW 11919SL?

rO""DAWQ I unw 9 I am

'"Lww ' W~.

A ft ft SdW map" J E FPLO i 1e.d4W B

FI tWM I 1 4 ft l C L HX M d W Figure 2 - Fault Tree FLOODING.LGC (continued) 72

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods U I I 2 3 4 I 5 jFLOODING LOGIC L, jAnatytl JRS lCreationDate 01-10-2C 11-02.

Rereagon.

ROHLOC We-.15148MMORAUn ra . aat.I.m iafara FM t. "

IfWeleidea 5w. heWk .a6 of mow In T(rb BW I1 2 -1 N 2 II KIZ I2JMI I

Irtw ..a,. ot Odom Ih I EhZ2J 2 4m..e (9 us-." 4 1) 3 5

8 9

Infnratmomd I M" I I Icern her ,r : VALIIES ISW Ae~e A

t@ 04AOB..

1us4.. 2

- T4 B

C 1FFe. Li( o t n e Figure 2 - Fault Tree FLOODiN,.LGC (continued) 73

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods b

a I I 2 I 4 FLOD IOICFILD, 21 0

~S Analyst: Creaion Oate:0I-I91Rsso:I-l~

FLObGJ.GC W ASUPrsi I FALIJNTOOL.A1IE LAR.EIblW IRKIN UR MM 2 FMI tIla (91,11M t 3

4 S

S (40 7

6 9

A 8

C Figure 2 - Fault Tree FLOODING.LGC (continued) 74

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods a I I1 2 3 4 0 FLOCODNG

[Aalyht:JRS LOGIC Preation Date:01-10-20 FDS Revision:

I11.303.0 FLOIOMMOC 0y.A514M.PRtA36 n 3

P" AL LOCA GR!AT~tTHMI21 GPM I tW~

I mfcscMAoLOC IWATRTm@GI GMATin1IMa1Huu Iin OLUMaawu I oscoLDtI ua.960 (2) LOB.4H (9 6

7 6

9 A

B C

Figure 2 - Fault Tree FLOODING.LGC (continued) 75

IINTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods I

3 4 I 5 0 I 1 2 1 a I I PLD, 32

[Anulyst:R~S ae O0eto 1-10-MM Revision:

11-33.2M5 IftbODILOJ.C#W1I6)V"PAj VrLE OMi 2 &

I A4 -4f (9111142Q 3

4 6

(qg 7

B 9

A I

C Figure 2 - Fault Tree FLOODING.LGC (continued) 76

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 0 7 I 2 i 3 l 4 1 5 FLOODING LOGIC FLD,41 a

[alyst: JRS lCreationDate:O1.lO. IReviion. 1Il-3-2 FIDODHOLMC 0W .1at4MV*owA3SB~1bnW Op.eWdnIdl.%

NWton TOIW Madl Irdbafftnntlf m-,C K8CWUf

_la B? AIWsa Idl H1lp...PW of" I.. of tA b~rIA I Th.AIP I NIWP^mp.2Snd Lg FIbdI I naMw JE I DALOP.F.If

_~e.u a E 4flat aaI1 (a) ( .I 2 . 3 34 M I~APWMI M.w.kACWAn H 4

canel MDAIPUnd 6

S lkA LoALJ Wjoil(

6 7

8 I UlA~d I 9

MWd jAb f~d &G _KW IA adI-A 2 (9 '9 C

Figure 2.- Fault Tree FLOODING.LGC (continued) 77

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 0 ___ -T- 2 I-- 3 I 4 1 5 0 I nd - WFLD,4 1:2 flFLOODING LOW lCreation Date:-~01-l02- RFeelgior II.3D-2D PLOWH2+/-C WAv.51U4VhMA1DFA.&nd opffdwfdmt. I TDAWY Tdre. FIL _Idf TDIIII I Lg.W PIrkk I

t FWA 1D b AMLOP AIwt .nk ms _

2 D-PIDSF4E A6.BJYALOP-FIE

.16. J.C 2 3

t.N.-NI 2 2if."6-N (2)M.S 4

5 s

7 a

AFPfood L'.6 LWFPMAC GNWP4.d4 9 L4~

I' NW A

I (9 A2fl 1 (9 C

Figure 2 - Fault Tree FLOODING.LGC (continued) 78

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods Figure 2 - Fault Tree FLOODING.LGC 79

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods U It 2 J 4 I 5 FLOODING LOGIC FLD,d4 O

Analyst:JRS Ceation Date:01-102-D RlResion:11-302005 FlOOvWlOC T-v.1S4flU"JRA3O0Pwb.n I idn tCAFP lI L.M U 9k I QXD40 Ipp YA- - belmo-mo nA2 A- _ Q 4bd~~~~op

_S jFr 11T A l_iD d ILOP P 2

IeSecd omnp.nn. M_ s "

_PTF4E

~re

• n mk II w0FW411rk P F4E 80IW OFF DP4C-MF4E Ra"ctW-W S1 0041_@FA 9.89,,, (1 001 U3 SAD W l Ff -11.0d 4

tI " .htft I dLao _MMA hi AlCanb 5

ISIWtCRF I _odkr*m I I f fR i (0

6 7

h st I VW~r I U.

r " 1 r "4IL.STIMl.II 29 rWk 8

9 A

a C

Figure 2 - Fault Tree FLOODING.LGC (continued) 80

INTERNAL FLOODING Fault Tree Analysis for Turbine Building Floods 0 i I I 2 3 4 5 1

FLOODING LOGIC l FLD,4 l 0

jAnslyst:JRS lCreationDate:01-10-20 Revision:11l3.2005 14LODOLOC onA154M'3M DlRlNi I is.mawmak IDAW T1*_fVPTL uk TIM M.d SIM rk t I GIFL04514 I PinSatCalniin MadiTbAWPWVp h d I 1~evlUMAP hS _* d dD Iw.n1FPI 5nia le dUD AL5P hns21l

t. Pn*S

"-A WpWN.2HF Lo.PWBIa.k 2

eWW-_ U4R E aUPIUMw~f I .M WTB"W. IS UB4DOP4P.

sJm..u s 41.-u2 s sJe.as (9 4am.w a.9 til.-Mi (2 3

W 1e Cd IMPd". S I "wrwHb Fc 4

IhPWaad I Lbad5 eIrldu I 6

(9lon a40m Q 6

II~

GMSM I UMW-GV4 I 7

a sA wS areE A

C Figure 2 - Fault Tree FLOODING.LGC (continued) 81

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods 0 T -I 1 2 3 T . 4 1 5 FLO 4 Q.i,)

0 1FdOCING s LOGC j FLeDioDate01.1D.0 PLO, 46

1. LGOOLMC V.1S1488)ViItM3A0ftoOf, 1db TIWP ~~owd Tlm 1m hlnto^

PFM Pd.70W L9.Wphk MWVW**

U9h412 Gf/;61 W.UATIP 1 11_ d MD W.ION hdtoob_ 11Dn ALOP ihr_

_IWPm-Or*I.

1ffd bauW* d NWrpS. - Id.. I .q p k 2

_-_ O IDP1UWif KPAC1W.41 I8DPTOFH_

IJo4 I 4Xb-Iui W a-.JO 2 4l-Sl 2 184 I 3

'wI."' IAMdP."I I blD* SR dh.

IL.GF"bAI

. A -- 4 Qr) to fft M"f kh 5

4jo. 51W~

ffC*

na~m~ 1 6

I- L2i 7

U V" JISWVI a 9" IS b1 W 9

E C

Figure 2 - Fault Tree FLOODING.LGC (continued) 82

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods U O 1 A j 4 5 FLOODING LOGIC 0 I I FLO, 47 Analyst:JRS lCretionOs. 0nt110. 10 Rnsien: 11-30-2005 I fLOOOCtIOCr.1i.4U)V4JnRA1Dft0*st TDAFP Tae ft I

hi*r*

I MUM4 I00 P~b.MD0l" W~MdF- .Ir M~lIAIW WT11CF471E IALCPPa-*ai-I I bdnn. i.e. MD 2 SJO/" TD( f", V Li j 1 S-M e INs-US () (9 3

raiic."

IAIWFnOlI 4

GILO142 G_ a4 5

AM

(

jh~dSCd~a ww.401 I WSTCRF41E I 6

"G.M II!

7

.I hb NW 8

cUOsm (.2)

A a

C Figure 2 - Fault Tree FLOODING.LGC (continued) 83

INTERNAL FLOODING - Fault Tree Analysis for Turbine Building Floods

%-# I e ' 3 4 I 5 a

I 2

3 4

S 6

7 a

A B

C Figure 3 - Placement of New Basic Event 16-BATCLG--F-HE 84