ML043030250

Slides for Oyster Creek EDG Regulatory Conference from Amergen

ML043030250
Person / Time
Site:	Oyster Creek
Issue date:	09/27/2004
From:	AmerGen Energy Co, Exelon Nuclear
To:	Office of Nuclear Reactor Regulation
References
Download: ML043030250 (38)

Text

SM Oyster Creek Generating Station REGULATORY CONFERENCE September 27,2004 Emergency Diesel Generator 1 Degraded Cooling Fan Event

4 A

z F1 0

• d m

F1- r r s

• d 0 s2 Td a F1 d)

• d

fl n c4 H e e e e e

0 Objectives Provide additional information addressing ability of Emergency Diesel Generator (EDG) 1 to perform its function.

Provide clear and effective applicability analysis of Joliet Diesel Testing.

Provide overall risk assessment of the significance of the event.

d 0

bl) 2

Finding Potentially Greater Than Green Finding - involving failure to follow written procedures to torque the cooling fan drive shaft bearing bolts following fan belt replacement for maintenance on EDG 1 during 24-month overhaul.

Resolution AmerGen agrees with Performance Deficiency.

Root Cause - Human performance event involving the failure of plant personnel to follow the implementing procedure.

AmerGen has taken extensive corrective actions to prevent recurrence.

Presentation Next Steps Diesel Testing & Results: John A. Magee

- EDG 1 operation was degraded; however, the event did not involve a failure of the EDG.

- Joliet Test was directly applicable to EDG 1 condition and provided conservative operating information.

- EDG 1 was capable of performing its safety function for a portion of the mission time.

- EDG 1 would have run a minimum of 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and likely much longer.

Risk Assessment: Michael P. Gallagher

- The Risk Analysis demonstrates that the consequences of this Finding are of Very Low Safety Significance.

d ba c Fj d) 0 h

Diesel Testing Presentation Outline

Background

Factual Observations Diesel Generator Fan Drive Test Key Similarities Test Diesel Generator versus Oyster Creek EDG Key Differences Test Diesel Generator versus Oyster Creek EDG Test Diesel Generator Results Analytical Results Conclusions

m-L f

General Arrangement of Diesel Generator Upper Sheave & Bushing Flexible Connecti MP45 Cooling Fan And Drive Assem

0 1

F9 3

_1__1__________

ct

/

Factual Observations OC EDG 1 completed 24-month overhaul on 04/30/04.

- Fan Belt Replacement was performed during overhaul.

Multiple EDG runs were performed and EDG 1 was declared operable on 04/30/04.

On 05/11/04, a surveillance load test was run for 1.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> with no problems noted.

On 05/17/04, a surveillance load test was run for 1.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />.

- This concluded in manual shutdown by operations, based on unusual noise and observed bearing movement.

Factual Observations OC EDG 1 continued to run until it was manually shutdown.

- The OC EDG 1 did not fail to start or run.

During this event Operations recorded all engine temperature parameters which were within normal range.

Inspections of the EDG 1 revealed:

Pillow block bearing upper bolt loose and the lower bolt missing (later found on the floor)

No belt wear or damage observed (belt reused)

Bolts were not damaged No damage to the pillow block bearing

OC EDG 1 Multiple Starts and Stops OC EDG 1 run times:

4/26/04 04:20 - EDGI 00s for 24-month inspection 4/29/04 06:13 - EDGl started for idle run and testing 4/29/04 08:11 - EDG1 idle run testing ended - 15 8 run time 4/29/04 22:43 - EDGI idle start for testing 4/29/04 22:57 - EDGI idle run ended - 0: 14 run time 4/29/04 23:28 - EDGl fast start testing 4/29/04 23:46 - EDGI fast start ended - 0:17 run time 4/30/04 02:31 - EDGl load test run 4/30/04 03:OO - EDGI load test run ended - 0:29 run time 4/30/04 03:Ol - EDG1 load test run 4130104 0556 - EDGl load test run ended - 2:55 run time 4/30/04 06:32 - EDGI load test run from CR 4/30/04 07:Ol - EDGI load test run ended - 0:29 run time 4/30/04 12:16 - EDGl surveillance run 4/30/04 13:36 - EDGI surveillance run ended - 1 :20 run time 5/11/04 01:19 - EDGI surveillance run 5/11/04 02:45 - EDGI surveillance run ended - 1:26 run time 5/17/04 03:34 - EDGI surveillance run 5/17/04 0457 - EDGI surveillance run ended - 1:23 run time 5/17/04 17:50 - EDGI Declared Available 5/17/04 20:25 - EDGl Declared Operable

I m

R 1

4) m 4)

Comparison of Assemblies Test Diesel Generator Results

1. Test DG ran for approximately 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of operation.

2. Radiator fan operated at reduced speeds due to drive fan belt slippage.

3. High coolant temperature led to an automatic engine shutdown.

4. Demonstrated the Upper Pillow Block Bearing Bolt would not back out.

Key Similarities Test DG versus Oyster Creek EDG

1. Identical Fan Design and Size.

2. Engine driven, lower fan shaft, speed (rpms) are the same.

3. Identical OC EDG Fan Belt was installed on the Test DG to assure unit comparability.

4. Identical OC EDG Lower Drive Shaft and Pillow Block were installed on the Test DG to assure unit comparability.

5. The Bearing Bolts, removed from the OC EDG, were installed on the Test DG.

Key Differences Test DG versus Oyster Creek EDG

1. Upper / Lower Sheave Sizes are different.

- This results in a more conservative test.

2. Test DG trip logic was enabled.

- The OC EDG 1 logic would have bypassed these protective trips (and others) for emergency starts.

- This results in a more conservative test.

3. Fixed Air Flow vs. Modulating Fan Louvers

- OC EDG 1 louvers lock full open > 200'F.

Additional Analysis Independent analysis was performed by MPR Associates, Diesel Generator Experts, of the degraded EDG 1 fan drive.

Using the demonstrated pillow block movement of the test diesel, the EDG 1 belt slippage was calculated and subsequent reduction in air flow was modeled.

This sensitivity analysis demonstrated that the EDG 1 belt slippage would be less than the Test DG; therefore, EDG 1 would have run cooler and longer.

EDG 1 would have run a minimum of 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and likely much longer.

Conclusions EDG 1 operation was degraded; however, the event did not involve a failure of the EDG.

Test DG results are directly applicable to EDG 1 condition

~

and provided conservative operating information.

- Demonstrated the pillow block motion in the degraded state.

- Demonstrated that the upper pillow block bearing bolt would not have backed out.

- Sensitivity analysis demonstrated that the EDG 1 belt slippage would be less than the Test DG; therefore, EDG 1 would have run cooler and longer.

EDG 1 was capable of performing its safety function for a portion of the mission time.

EDG 1 would have run a minimum of 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and likely much longer.

Presentation Next Steps Risk Assessment: Michael P. Gallagher

- The Risk Analysis demonstrates that the consequences of this Finding are of Very Low Safety Significance.

d 4

Risk Analysis Approach Conservative and Realistic Credits Actual Diesel Condition, Le., Diesel was Degraded v

but not Failed Supported by precedent

Risk Analysis Approach Dominant Core Damage Sequences Station Black-Out (Loss of All AC Power)

- No Offsite Power Recovery at 8 Hours Station Black-Out (Loss of All AC Power)

- Recirc Pump Seal LOCA

- No Offsite Power Recovery at 1 Hour

• Station Black-Out (Loss of All AC Power)

- Stuck Open Relief Valve

- No Offsite Power Recovery at 0.5 Hour Station Black-Out (Loss of All AC Power)

- Isolation Condenser Make-up Failure

- No Offsite Power Recovery at 1 Hour

• SPAR sequence of interest from NRC preliminary analysis

Conservative and Realistic NUREG/CR-5496 LOOP Frequency of 0.046/year (consistent with NRC SPAR model).

Loss of Offsite Power Recovery Curves Updated to include recent events (more conservative than NUREG/CR-5496).

Utilizes full fault exposure of 17.5 days (consistent with NRC preliminary analysis).

Recovery of EDG 2 Credited in SBO Scenarios (consistent with NRC SPAR model).

Includes Best Estimate External Events Adder of 1.6E-7.

I CA d) i \

I I

Conservative and Realistic Diesel maintenance unavailability conservatisms

- OC PRA: 1.2% unavailability vs. OC actual performance: 0.5%

Risk Analysis did not credit relief valve re-closure

- 85% probability of relief valve re-closure could be utilized based on industry and plant specific data.

- Risk analvsis utilized stuck oDen relief valve probability of 9.16E-3 based 6n plant specific data.

E0 a

a s

m c!

+

6) sF1 Td 0 F1 'u c6 c)

Degraded Diesel Analysis EDG 1 did not fail to start and run.

Conservatively credits 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> EDG run time, although EDG 1 would likely have run much longer.

EDG 1 is assumed to start and run for 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> with weighted random failure probabilities included. Then assumes EDG 1 failed at 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.

Conservatively does not credit EDG 1 recovery.

No common cause factor affecting redundant EDG was

n. 4 veritied.

Includes Two Fault Exposure Periods

- 11.5 days from return-to-service to beginning of first surveillance test (credited 9 hours1.041667e-4 days <br />0.0025 hours <br />1.488095e-5 weeks <br />3.4245e-6 months <br /> available run time)

- 6 days from beginning of first surveillance test to end of second surveillance test (credited 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> available run time)

64 0 \9

1. 4 0

0 I

PL: 0 F+

0 H

I m 0

1. 6 l bp)

E: d #

c4 I I I

Results IAmerGen Degraded Core Damage Frequency Internal = 5.OE-7 Large Early Release Frequency Internal* = 5.4E-8 Diesel Analysis - With External = 1.6E-7 External** = 1.6E-8 Credit for 6 Hour Joliet Total = 6.6E-7 Total = 7.OE-8 Very Low Safety Very Low Safety Significance Significance Green - White 1.0 E-6 1.0 E-7 Threshold

• Based on Detailed Level I1 PRA

• • Based on LERF Multiplier (0.1)

Sensitivity of Results Core Damage Frequency AmerGen Degraded Diesel Internal = 5.OE-7 Analysis - With Credit for 6 Hour External = 1.6E-7 Joliet Test Total = 6.6E-7 Very Low Safety Significance AmerGen Degraded Diesel Internal = 4.1E-7 Analysis - With Credit for 9 Hour External = 1.6E-7 Run Time Total = 5.7E-7

~

Very Low Safety Significance AmerGen Degraded Diesel Internal = 3.7E-7 Analysis - With Credit for 12 Hour External = 1.6E-7 Run Time Total = 5.3E-7 Very Low Safety Significance

Risk Analysis Conclusions AmerGens Risk Analysis is appropriately conservative and realistic.

AmerGens Risk Analysis conservatively credited 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> EDG 1 run time, although the EDG would likely have run much longer.

Precedent in risk significance determinations has credited the capability of a degraded EDG to perform its safety function for a portion of the mission time.

The Risk Analysis demonstrates that the consequences of this Finding are of Very Low Safety Significance.

t) v)

'v 0 5 M

W rt crs m

3 0

r!

\o

• 0 m

d) d)

b ccf F1 8 4 r t 0

0 0 0 0 0 0 0