ML20138E285

Forwards Addl Info Re Operator Action During Small Break LOCA Event,Including Status of ECCS Equipment During Shutdown & Peak Clad Temp Analysis During Modes 3 & 4,per SER Confirmatory Item 22

ML20138E285
Person / Time
Site:	Vogtle
Issue date:	12/09/1985
From:	Bailey J GEORGIA POWER CO.
To:	Youngblood B Office of Nuclear Reactor Regulation
References
GN-758, NUDOCS 8512130405
Download: ML20138E285 (11)

Text

.- , _

4

) Georgia Power Company l

' P.outa 2 Box 239A l t Waynesboro. Georgia 30830 Telephone 404 554-9961 404 724-8114 Southern Company Services, Inc.

Post Office Box 2625 B!rmingham, Alabama 35202 Te!ephone 205 870-6011 December 9, 1985 4

Director of Nuclear Reactor Regulation File: X7BC35 l Attention: Mr. B. J. Youngblood Log: GN-758 PWR Project Directorate #4 1 Division of PWR Licensing A U. S. Nuclear Regulatory Commission j Washington, D.C. 20555 i

NRC DOCKET NUMBERS 50-424 AND 50-425

) CONSTRUCTION PERMIT NUMBERS CPPR-108 AND CPPR-109 l j V0GTLE ELECTRIC GENERATING PIANT - UNITS 1 AND 2

' SER CONFIRMATORY ITEM 22: SMALL BRFAK LOCA

Dear Mr. Denton:

Attached for your staff's review is additional information requested concerning operator action during a small break LOCA event. This submittal provides the status of ECCS equipment during shutdown, indication available to the operator, and p*ak clad temperature analysis during Modes 3 and 4 for the postulated event.  !

l 4

l If you staff requires any additional information, please do not hesitate to contact me.

, Sincerely,

( J. A. Bailey

{- Project Licensing Manager JAB /sm Enclosure xc R. E. Conway G. Bockhold, Jr.  :

R. A. Thomas T. Johnson J. E. Joiner', Esquire D. C. Teper 600 l B. W. Churchill, Esquire L. Fowler i i M. A. Miller (2) W. C. Ramsey I f 3 B. Jones Vogtle Project File L. T. Gucwa

! i ADOt PtlW - A/BC's TECM YUPPORT 0263V B512130405 851209 ci isN$I"' "" *"'d

! PDR ADOCK 05000424: .

erCSs (ROSA)

Pts (CAassILL)

E PDR i

4 <a oss (srauncui 5' ' ros (spAnoyA)

~ - . . . . -- . , - , . - _ ..-_----.-_-__--_,_..-_--4._.--_,,__--. , _ _ _ , , , _ - .

CONFIRMATORY ITEM 22 V0GTLE SER During shutdown of the Vogtle nuclear power plant, the following Emergency Core Cooling System (ECCS) equipment important to the mitigation of a LOCA is locked out or not required to be operable. (Start-up is not addressed since shutdown is more limiting due to the high core decay heat generation.)

(1) At 1900 psig, the operator is instructed to manually block the automatic safety injection (SI) actuation circuit. This action disarms the SI signal from the pressurizer pressure transmitters so SI will not be initiated due to low pressurizer pressure. This circuit will automatically unblock if the Reactor Coolant System (RCS) should increase above ~2000 psig. The containment high pressure SI signal is armed and will actuate SI if the Hi-1 setpoint is exceeded. Manual SI actuation is also available.

(ii) At 1000 psig the operator closes and locks out the SI accumulator discharge isolation valves.

(iii) Below 350*F the Technical Specifications require only one RHR pump and one centrifugal charging pump (CCP) to be operable.

The significance of these actions on the mitigation of a LOCA is as follows:

(i) Below 1900 psig the containment Hi-1 pressure SI signal is the only signal available for automatic SI .. tiation. The Hi-1 containment pressure setpoint would easily be reached in the event of a large break LOCA (LBLOCA) and for larger small break LOCA's (> 2"). For break sizes two inches in diameter and less, however, the Hi-1 containment pressure setpoint may not be reached. The operator would be required to manually initiate SI. Appendix A describes the alarms available and time allowed for manual SI initiation to prevent core uncovery. The analysis results 0970n:39/RJM/12-85

r presented in Appendix A demonstrate that if the operator manually initiates S1 within 30 minutes af ter a SBLOCA of two inches in diameter or less, no core uncovery occurs. The SBLOCA analysis presented in the F5AR is therefore bounding for SJLOCA's under the above initial conditions.

(ii) Below 1000 psig the accumulators are locked out. If a SBLOCA were to occur, sufficient makeup to the RCS is supplied by the available SI pumps. The accumulators are necessary to mitigate a LBLOCA at full power but at 0 power and at RCS pressures less than or equal to 1000 psig, the SI pumps deliver sufficient mass to recover from a LBLOCA and prevent the peak clad temperatures from reaching in values calculated in the FSAR analysis (see Appendix B).

(iii) Below 350*F RCS temperature, only 1 CCP and 1 RHR pump are required to be operable. At these reduced temperatures and pressures (Technical Specifications contain Appendix G curves which limit RCS pressures to less than 750 psig), the CCP is sufficient for recovery from smaller breaks and larger breaks quickly depressurize the RCS down to the shutoff head of the RHR pump. The RHR pump also delivers sufficient flow to recover from a LBLOCA (see Appendix C).

0970n:40/RJM/12-85

APPEN0!X A Following is a description of the indication available to the operator that a 58LOCA (< 2 inches) has occurred. Also discussed is the time available for the operator to manually initiate SI af ter he has detected the LOCA.

The alarms available to the operator for LOCA detection include the containment high-radiation alarm and sump high level alarm. In particular, the high radiation alarm is set to sound when radiation levels in the containment reach twice the background level. Calculations indicate that the alarm would sound for a 1 gpm leak one hour af ter the start of the leak.

Break flow from a 1 inch break is on the order of 500 gpm and a 2 inch break would have a flow of approximately 2000 gpm. Thus, these breaks would be expected to set off the containment high radiation alarms much sooner than an hour after the break occurs. In addition to these alarms, the operator would also be alerted to a LOCA by decreasing RCS pressure and decreasing pressurizer level.

Several analyses have been done for 4-loop plants similar to Vogtle that demonstrate the operator has suf ficient time to manually initiate SI prior to core uncovery for LOCA's 2 inches and less in diameter. Based on the Inadequate Core Cooling Study (WCAP-9753) for full power operation, a 1 inch break would exhibit an extremely long transient prior to core uncovery from the initiation of break flow (approximately 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> for a 4-loop plant). An even longer transient would be expected for a small break during shutdown.

From McGuire low power test analyses (5 percent power), for a 2 inch break no core uncover occurs prior to 1.67 hours7.75463e-4 days <br />0.0186 hours <br />1.107804e-4 weeks <br />2.54935e-5 months <br />. A 2 inch break analysis was also done for Byron which is very similar to Vogtle. This analysis was done for full power conditions with no SI using NOTRUMP. This analysis did not predict core uncovery until 42 minutes after the break. The results of all of the above analyses demonstrate that the operator has sufficient time to manually initiate SI and preclude core uncovery.

0970n:41/RJM/12-85

The analysis mentioned above for the Byron plant is considered to be applicable to the Vogtle plant for the following reasons. Both Byron and Vogtle are Standard Westinghouse 4-loop,12 foot core designs. Both are I designed to be operated at 3425 MWt and at 2250 psia. The vessel average operating temperatures at full power dif fer by only .I'F. Both plants are i

upper head T ,y plants with 5.8% core bypass flow. The FSAR small break analyses for both plants show that the 4 inch case is the most limiting with the Byron PCT being more than 200'F higher. This is due mostly to the different upper head types in the plants and the fact that Byron uses optimized fuel. The Byron analysis is therefore applicable to Vogtle and probably would bound a similar analysis done for Vogtle.

The Byron 2 inch break analysis is also bounding to a 10CFR50.56 Appendix K type analysis at shutdown conditions for the following reasons. Appendix K type analyses require increased power and decreased safety injection and J

auxiliary feedwater (AFW). The Byron analysis assumes full power and no SI.

! This is certainly conservative to zero power conditions with no SI. The AFW flows input to the Byron analysis are also less than the Vogtle FSAR small break analysis flows. The Byron 2 inch break NOTRUMP analysis is therefore bounding to a Vogtle Appendix K shutdown analysis.

i l

l 0910n:42/RJM/12-85

APPENDIX B l The following conservative analysis has determined that the peak clad temperature resulting from a large break LOCA in Mode 3 below 1000 psig would be less than the peak clad temperature calculated in the FSAR analysis using the ECCS equipment available 2-1/2 hours afrer reactor shutdown.

The following assumptions were used in the analysis:

1. The RCS fluid is isothermal at a temperature of 425'F and a pressure of 1000 psig.

2. The core and metal sensible heat above 425'F has been removed.

3. The hot spot occurs at the core midplane. -

4. The peak fuel heat generation during full power operation of 12.773 Kw/Ft will be used to calculate adiabatic heatup.

5. At 2-1/2 hours using decay heat in conformance with Appendix K of 10CFR50, the peak heat generation rate is 0.1724 Kw/Ft.

6. 51 flow values used correspond to one RHR pump, one intermediate HHSI pump, and one CCP.

7. 51 flow starts at the end of blowdown.

8. No liquid water is present in the reactor vessel at the end of blowdown.

9. A large cold leg break is considered.

For a postulated LOCA at the cooldown condition of 1000 psig, previous calculations shod that the clad does not heat up above its initial temperature during blowdown. Proceeding from the end of blowdown and assuming adiabatic 0970n:43/RJM/12-85

heatup of the fuel and clad at the hot spot, an increase o,f 589'F was ,

calculated during the lower plenum refill transient of 118 seconds. During reflood, the core and downcomer water levels rise together until steam generation in the core becomes sufficient to inhibit the reflooding rate. At that time, heat transfer from the clad at the hot spot to the steam bolloff and entrained water will commence. This heat removal process will continue as the water level in the core rises while the downcomer is being filled with safety injection water. The reflood transient was evaluated by considering two bounding cases:

1. Downcomer and core levels rise at the same rate. No cooling due to steam boiloff is considered at the hot spot. Quenching of the hot spot occurs when the core water level reaches the core midplane.

2. Core reflooding is delayed until the SI pumps have completely filled the downcomer. No cooling due to steam boiloff is considered at the hot spot until the downcomer is filled. The full downcomer situation may then be compared with the results of the ECCS analysis for Vogtle to obtain a j bounding clad temperature rise thereafter.

For Case i described above, the water level reached the core midplane 65.5 i

seccnds after bottom of core recovery. The temperature rise during reflood at the hot spot from adiabatic heatup is 326*F, which results in a peak clad temperature of approximately 1340*F.

For Case 2, the delay due to downtomer filling is 83 seconds. The corresponding temperature rise at the hot spot from adiabatic heatup is 413*F, which gives a hot spot clad temperature of 1427'F.

The clad tempertures at the time when the downtomer has filled for the DECLG submitted to satisfy 10CFR50.46 requirements are 1553*F and 1561*F at the 6.0 j and 5.75 foot elevations, respectively.

i 0970n:44/RJM/12-85

Core reflooding in the shutdown case under consideration will be more rapid f rom this point on due to less steam generation at the lower core power level in effect; decay heat input at any given elevation is less in the shutdown case. The combination of more rapid reflooding and lower power in the f uel insures that the clad temperature rise during reflood will be less for the shutdown case than for the design basis case.

Based upon the analysis as presented above, it can be concluded that in the unlikely event of a LOCA at shutdown conditions, the peak clad temperature will be less limiting than that of the design base calculation.

0970n:45/R.1M/12-85

APPENDIX C The following conservative analysis has determined that the peak clad temperature resulting from a large break LOCA in Mode 4 would be less than the peak clad temperature calculated in the FSAR analysis using the ECCS equipment available 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> after reactor shutdown.

The following assumptions were used in the analysis:

1. The RCS fluid is isothermal at a temperature of 350'F and a pressure of 1000 psig.

2. The core and metal sensible heat above 350*F has been removed.

3. The hot spot occurs at the core midplane.

4. The peak fuel heat generation during full power operation of 12.773 Kw/Ft will be used to calculate adiabatic heatup.

l

5. At 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> using decay heat in conformance with Appendix K of 10CFR50, the i peak heat generation rate is 0.152 Kw/Ft. ,

6. SI flow values used correspond to one RHR pump.

7. 51 flow starts at the end of blowdown.

8. No liquid water is present in the reactor vessel at the end of blowdown.

9. A large cold leg break is considered.

For a postulated LOCA at the cooldown condition of 1000 psig, previous calculations show ttist the clad does not heat up above its initial temperature during blowdown. Procteding from the end of blowdown and assuming adiabatic 1

0970n:46/R.1M/12-85

heatup of the fuel and clad at the hot spot, an increase of 646*F was calculated during the lower plenum refill transient of 147 seconds. During .

reflood, the core and downcomer water levels rise together until steam generation in the core becomes sufficient to inhibit the reflooding rate. At that time, heat transfer from the clad at the hot spot to the steam boiloff and entrained water will commence. This heat removal process will continue as the water level in the core rises while the downcomer is being filled with safety injection water. The reflood transient was evaluated by considering two bounding cases:

1. Downcomer and core levels rise at the same rate. No cooling due to steam boiloff is considered at the hot spot. Quenching of the hot spot occurs when the core water level reaches the core midplane.

2. Core reflooding is delayed until the SI pumps have completely filled the downcomer. No cooling due to steam boiloff is considered at the hot spot until the downtomer is filled. The full downcomer situation may then be compared with the results of the ECCS analysis for Vogtle to obtain a bounding clad temperature rise thereafter.

For Case 1 described above, the water level reached the core midplane 80.4 seconds after bottom of core recovery. The temperature rise during reflood at the hot spot from adiabatic heatup is 352'F, which results in a peak clad temperature of approximately 999'F.

For Case 2, the delay due to downtomer filling is 103 seconds. The corresponding temperature rise at the hot spot from adiabatic heatup is 413*F, which gives a hot spot clad temperature of 1099'F.

The clad tempertures at the time when the downcomer has filled for the DECLG submitted to satisfy 10CFR50.46 requirenants are 1553*F and 1561*F at the 6.0 and 5.75 foot elevations, respectively.

• 0970n:47/RJM/12-85

r Core reflooding in the shutdown case under consideration will be more rapid f rom this point on due to less steam generation at the lower core power level in effect; decay heat input at any given elevation is less in the shutdown case. The combination of more rapid reflooding and lower power in the fuel insures that the clad temperature rise during reflood will be less for the shutdown case than for the design basis case.

Based upon the analysis as presented above, it can be concluded that in the unlikely event of a LOCA at shutdown conditions, the peak clad temperature will be less limiting than that of the design base calculation.

0970n:48/RJM/12-85