Text

Forwards Responses to Questions Raised by Sandia During Review of Plant PRA Covering Steam Generator Dryout
	ML20012A129
Person / Time
Site:	South Texas
Issue date:	03/01/1990
From:	Mcburnett M; HOUSTON LIGHTING & POWER CO.
To:	; NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
	ST-HL-AE-3380, NUDOCS 9003080347
	Download: ML20012A129 (19)
	v • d • e

C I

i

.s to j The Light i co mpanySouth Temas Project Electric Generating Station P.O.Bos289 Wadsworth. Temas 77483 Houston Lighting & Power i

March 1, 1990 ;

' ST-HL AE- 3380 Fileato.: 020.1 i' 10CTR50.12 i l

l U.S. Nuclear Regulatory Commission ,

Attention: Document control Desk Washington, DC 20555 South Texas Project Electric Generating Station l Units 1 & 2 Docket Nos. STN 50-498, SSN 50-499

Responses to the -

Requests for Additional Information from ,

igndia National Laboratorv Referencet Letter from Sand!L Eattuat.1 Laboratory ,

to the U.S. Nuclear Regulatory Commission. ;

dated January 3,1990 i

Pnclosed are renponses to questiona reised by Sandia National Laboratory '

(SNL) ditring thuir review of the Scath Texas Project Electric Generating '[

Station WTPEGS) Prcbabilistic Safety Assesttnent (PSA). The Question A.4 ;

response-iAttachment 1) related to steam generator dryout completes the acticn ;

-$tems resulting from the November 28-30, 1989 plant s' kit. The responces to questions Q1 env. Q2, rmesived f.n the e. bows reforermed letter, regarding the -

e STPEGS PSA fire analysis'are included as Attachment 2.

If you should have any questions on this matter, or the attachments, please contact Mr. A. W. Harrison at (512) 972-7298 or myself at j (512) 972-8530.

/ t

$& f#l/f(4 M. A. McBurnett i

Manager Licensing .

MAM/edp t,

ATTACHMENTS: (1) Response to Question A.4, regarding Steam Generator dry out (2) Response to Questions Q1 and Q2, regarding the STPEGS PSA SAM-PRA3 A Subsidiary of Housten Industries incorporated i 9003000347 900301 i l- 4 PDR. ADOCK 05000490 ' l'- :

l P PNV -

u,

-Es 04 6 = ,. -

ST.HL-AE 3380

, _ Ho6ston Lighting & Power Company File No.: C20.1 p #g- ' ; ; < South. Texas Project Electric Generating Station . Page 2 5, cc:

Regional Administrator, Region IV Rufus S. Scott Nuclear Regulatory Commission Associate General Counsel 611 Ryan Plaza Drive, Suite 1000 Houston Lighting & Power Company Arlington, TX 76011 P. O. Box 61867 Houston, TX 77208

! George Dick, Project Manager i U.S. Nuclear Regulatory Commission INPO L

Washington, DC 20555- Records Center-1100 Circle 75 Parkway J. I. Tapia Atlanta, GA 30339-3064 Senior Resident Inspector c/o U. S. Nuclear Regulatory Dr. Joseph M. Hendrie Commission 50 Be11 port Lane P. O. Box 910 Be11 port, NY 11713 Bay City, TX 77414 D. K. Lacker

-J. R. Newman, Esquire- Bureau of Radiation Control Newman & Holtzinger, P.C. Texas Department of_ Health

.1615 L Street, N.W. 1100 West 49th Street Washington, DC 20036 Austin, TX 78704 D. E. Ward /R. P. Verret Central Power & Light Company P. O. Box 2121 Corpus Christi, TX 78403 J. C.- Lanier Director of Generation City of Austin Electric Utility 721 Barton Springs Road Austin, TX 78704 R. J. Costello/M. T. Hardt City Public Service Board P. O. Box 1771 San Antonio, TX 78296 r

E:

Revised 12/15/89 L4/NRC/

-p4wnvc -:u- m- -

w w -

vg 3

s w: . .g c,7 - ,

-i. -;

4 1 n

.l*

., 1

-i

'1 J,

a j

1 r

}

>( ..

. 4,

,h

.].

1

-t ie

. .f 1

1 h -

l' n-t 4 1 y

ATTACHMENT-1 <

r P

f

.1y

{- - . V t

x

.{

l e Tt

'r' 4

. 1 g .: p 3l

?

].

d C

.}

e s

1 i

. G fh 5 v_

g l k~

1 1

l Y

s i

i i!

Y.-

s.. i Y'?$ Skki--sWudv:4Aw. wq..g,,. ;_m. .,.,,.w.,

, , __n . . . ,__A,. .ns - ..n > nn, - . l. -

~

ATTACHMENT.1 ST-HL-AE 33B O

.. .. PAGE 1 0F 12 Page 1 of 12

\'

Question A.4. from the November 28-30, 1989 SNL Site Visit Section B.1 HEOB02 estimates that over one hour is required for steam generator dryout following reactor and turbine trip with no feedwater. This section assumes the reactor trips on low-low steam generator level "... estimated to be about 90% of the normal full-power liquid inventory." The FSAR Figures 15.2.9A and 15.2-10 indicate that the secondary mass in each steam generator is about 60,000 lbm at low-low level trip which is less than 50% of the full power inventory. Using this lower inventory, dryout is estimated to occur at about 30 minutes.

A.4.1: What is the justification for the 90% assumption?

A.4.2 _How does a decrease in time to dryout from one hour to 30 minutes affect the PSA model?

,- m

's / A.4.3 Is the discrepancy due to the fact that level is calculated by measuring the pressure drop across ,

two taps in the downcomer and flow losses are much less without feedwater?

(D V

(SGDRY/SANDIA) 2/27/90

' ATTACHMENT J ST-HL AE 3330 PAGE 2-- 0F-11 Page 2 of 12 ry HL&P Response:

A.4.1 The 90% of nominal inventory was based on the Seabrook plant.1 The STPEGS FSAR value of approximately 60,000 (lbm) L is conservative based on instrumentation errors. The STPEGS ,

nominal operating steam generator level is 59% on the narrow range span, which corresponds to a calculated best-estimate liquid water mass of approximately 487,538 (1bm) for all-four steam generators per Reference 2. The low-low steam generator level setpoint is 33% on the narrow range. Reference 2 calculated the corresponding best-estimate liquid water mass to be approximately 362,063 (lbm) for all four steam generators, which is approximately 74% of the nominal and not 90%. Thus, the assumption, as stated, is not conservative and the PSA calculation has been revised, as shown on the marked-up pages

(~} attached, to yield a dryout time of 48 minutes versus 84

\ s' minutes.

A.4.2 HEOB02 represents the likelihood of the operators failing to initiate primary side " bleed and feed" prior to steam generator dryout given the loss of main feedwater and the failure of auxiliary feedwater. This scenario assumes reactor trip occurs due to low-low steam generator level, which relates to a secondary side inventory of about 74% of the nominal as identified above in Response A.4.1.

l Given this scenario, STPEGS reactor operators may enter several emergency operating procedures (EOPs) in response to the reactor trip and before initiating " bleed and feed". The first

()

'/

EOP is OPOP05-EO-E000, Reactor Trip or Safety Injection, which determines whether safety injection is required or not. If SI (SGDRY/SANDIA) 2/27/90

f ATTACHMENT 1 F ST-HL AE 3380

.. PAGE 3 0FJ2 j Page 3 of 12 i r~S sm ) is not required (e.g., loss of main feedwater. initiating )

event), then step 4.0 would lead the operators to OPOP05-EO-ES01, Reactor Trip Response, and to OPOP05-EO-FRH1, Response to Loss of Secondary Heat Sink, when inadequate feed j flow is identified. If SI is required (e.g., main feedwater line break initiating event), then step 19.0 would lead the operators directly to OPOP05-EO-FRH1, Response to Loss of Secondary Heat Sink. ,

OPOP05-EO-FRH1 directs the operators to trip the reactor coolant pumps (RCPs) and initiate " bleed and feed" immediately a when certain conditions exist (i.e., wide range level less than 37% gr pressurizer pressure greater than 2335 psig). Assuming these conditions have not yet been met, the operators are directed to reestablish the secondary heat sink. This activity includes troubleshooting the auxiliary feedwater, motor-driven i

startup feedwater, steam-driven feedwater, and condensate. flow paths. Note that the operators are continually monitoring the critical safety funtions and are procedurally required to initiate " bleed and feed" when the criteria stated above is met.

The Westinghouse owners Group (WOG) emergency response guidelines (ERGS) identify that a best estimate expectation of when the operator can be expected to trip the RCPs following reactor trip is approximately 5 minutes.6 This elapsed time can correspond to either having just entered OPOP05-EO-FRH1 (i.e., the conditions have been met to trip RCPs and initiate

" bleed and feed") or reached Step 3.0 of OPOP05-EO-FRH1, Stop all RCPs. To be conservative, the operators will know within 15 minutes after reactor trip to initiate " bleed and fecd".

Approximately 3 minutes is required to initiate " bleed and feed".

-A U

(SGDRY/SANDIA) 2/27/90

I ATTACHMENT.1 ST HL AE-339o -

PAGE 4 0F 32 Page 4 of 12 In confirmation of the the above procedural guidance, a I review of the operator survey results (i.e., Table 15.4-39 of Reference 3) showed that the operators gave the " time" performance shaping factor (PSF) the lowest rating (or importance) of the seven factors included in the survey. The most important factors from the operator's view were " stress"

. and " procedures", as shown on the attached markup of Table 15.4-39. " Time" was not a factor to the operators with respect to this scenario; thus the procedure used to quantify human error rates (HERs) for the PSA would not be impacted by the change. Section 15.2 of Reference 3 provides a more detailed discussion of the procedure used to quantify HERs (e.g.,

HEOB02) and the role PSFs play in this process. Therefore, a dryout time of 48 minutes will have no impact on the value of HEOB02.

A review of the dominant sequences leading to core damage as predicted by the PSA (i.e., Table 2.1-3 of Reference 3) shows that the most likely initiating event resulting in loss of secondary heat sink is the loss of offsite power (LOSP). For a LOSP initiating event resulting in station blackout, the reactor, turbine, and reactor coolant pumps would trip at time zero, thus resulting in the entire nominal steam generator water mass available for decay heat removal and more time for recovery actions. These recovery actions include getting the turbine-driven auxiliary feedwater pump started and/or restoring electric power prior to steam generator dryout.

HL&P calculated a range of dryout times for the STPEGS steam generators under various initiating events and assumptions.2,4 For the LOSP case, the range of dryout times is from 64 to 72 minutes. Note that this range covers a span of 8 minutes based on different decay heat curves and conservative assumptions.

(SGDRY/SANDIA) 2/27/90

= _ - _ - _ - _ _ _ _

l ATTACHMENT 1

+

1 -

ST-HL-AE-33go . ;

PAGE--5 0FE /2 ;

Page 5 of-12 .j

,f x' s.

LJ A.4.3 -)

See response to Question A.4.1.

L References' -.

1. Pickard, Lowe, and Garrick, Inc. " Thermal-hydraulic Analysis of Postulated Loss of Decay Heat Removal Events

During' Shutdown," prepared for Public Service of New Hampshire. PLG-0595. December 1987.

2. Engineering Calculation: Time to Boil Steam Generators Dry with Loss of Feedwater. NE-TE-89-09-00. Houston Lighting &

e Power Company. January 1990..

l h; .

'([ ,/ .

3.- Pickard, Lowe, and Garrick, Inc. " South Texas Project Probabilistic Safety Assessment," prepared for Houston Lighting & Power Company. PLG-0675. May 1989.

4. Engineering Calculation: Steam Generator Dryout Time for Station' Blackout with Loss of Feedwater. NE-PA-90-01-00.

Houston Lighting & Power Company. February 1990.

5. Larson, J.R. System Analysis Handbook.- Prepared for the U.S. NRC'by EG&G. NUREG/CR-4041. December 1984.

l ..

L L

6. Westinghouse Owners Group Emergency Response Guidelines, Low Pressure Version. Revision 1. Background Volume FR-H.

Prepared by Westinghouse Electric Corporation for The i Westinghouse Owners Group. September 1, 1983.

1 L. I 7

t 1

L g- -(SGDRY/SANDIA) 2/27/90 L , ;

. . - - - - . - _ . - - - - - - - - _ _ _ . . . _ _ _ . - - - - - - - . - . - - - -- - +

^

ATTACHMENT;L ~'

ST-HL AE 33po PAGE L .0F /D o

i Table 15.4 39. Grouping of HERs by Similar Average Weights HER TS 1/Pl Time P/C ACT PROC TRN/EXP Stress HEOD03 0.12 0.12 0.17 0.12 0.14 0.14 0.18 HEOR08 0.14 0.17 0.16 0.11 0.16 0.16 0.10 HEOR07 0.13 0.18- 0.15 0.12 0.16 0.16 0.10 MESL1 0,12 0.13 0.12 0.15 0.18 0.19 0.10 HEOR05 0.14 0.14 0.18 0.15 0.10 0.15 0.13 Average 0.13 0.15 0.16 0.13 0.15 0.16 0.12 HEOD02 0.08 0.18 0.08 0.20 0.18 0.16 0.12 ;

HEOC01 0.16 0.12 0.07 0.17 0.21 0.16 0.12 <

HEOCO2 0.16 0.14 0.00 0.14 0.22 0.16 0.14 HEOB04 0.09 0.14 0.04 0.19 0.17 0.16 0.20 HECH02 0,09- 0.17 0.16 0.20 0.00 0.14 0.17 HEOB03 0.08 0.14 0.11 0.15 0.19 0.12 0.21 Average 0.12 0.14 0.09 0.17 0.17 0.15 0.17 HECH01 0.12 0.24 0.08 0.25 0.04 0.20 0.08 HEOR03 0.14 0.21 0.09 0.17 0.09 0.21 0.10 HEOR01 0.13 0.21 0.10 0.17 0.06 0.19 0.13 HEOR04 0.14 0.21 0.09 0.17. 0.09 0.21 0.10 HEOR02 0.13 0.20 0.11 0.19 0.06 0.19 0.13 Average 0.13 0.21 0.09 0.19 0.07 0.20 0.11

- HEOT02 0.11 0.20 0.11 0.13 0.13 0.20 0.11 HEOT01 0.12 0.20 0.10 0.13 0.13 0.20 0.12 HEOT03 0.11 0.19 0.11 0.13 0.13 0.19 0.13 HEOLO2 0.11 0.19 0.08 0.17 0.17 0.16 0.13 HEOD01 0.14 0.19 0.09 0.10 0.17 0.19 0.13 HEOLO1 0.11 0.19 0.10 0.17 0.17 0.13 0.13 Average 0.12 0.19 0.10 0.14 0.15 0.17 0.13 HEOS02 0.17 0.17 0.24 0.17 0.07 0.11 0.07 HEOS03 0.15 0.15 0.23 0.21 0.06 0.10 0.11 HEOS01 0.18 0.18 0.20 0.18 0.08 0.12 0.08 Average 0.17 0.17 0.22 -0.19 0.07 0.11 0.09 lHEOB02 l l 0.09 l 0.14 l 0.09 l 0.13 l 0.20 l 0.14 l 0.21 l' HEOB07 0.08 0.13 0.13 0.14 0.17 0.13 0.21 HEOB09 0.08 0.15 0.11 0.14 0.17 0.14 0.20 HEOBA 0.08 0.16 0.14 0.15 0.18 0.10 0.19 HEOB06 0.00 0.14 0.14 0.12 0.19 0.14 0.19

/ q Average 0.08 k 0.14 [ 0.12 0.14 0.18 \ 0.13 / 0.20

\Lowest flighe s t Rated Rated PSPs PSPs for for HE0B02 HE0B02 l

l~

NHLPI N0079.051589 15.4-77 Pickard, Lowe and Garrick, Inc.

. i ATTACHMENT T o ST HL-AE 23 20

- . . ' .. 7 PAGE__7 0F ._11_.

~

' APPENDIX B. THERMAL HYDRAULIC ANALYSES FOR HUMAN ACTION ACCIDENT SCENARIOS The purpose of this appendix is to present simplified the# mal hydraulle analyses to piovide time windows for the human actions analysis tecnarios that appear in the overall event sequence models. The analyses,in general, are based on first principle energy and mass balance considerations, and are usej to evaluate factors such as times available for operator action and if safety injection can be in!! lated before the core uncovers. There are also some times approximated from the Westinghouse " Anticipated Transients Without Trip Analysis" (Reference B-1). All steam and fluid properties were determined from Reference B 2.

Because of the simplifying assumptions implicit in the analyses, the results should be considered as reasonable approximations of the time windows that could impact operator actions and decisions. Results from more " detailed" computer calculations should be used ,

for purposes requiring greater accuracy. l The human actions included as top events in the STP event tree models are identified by a !

six character designator. The first two characters are "HE," representing human error. The next two characters identify the human action category as follows:

OB - Operator Establishes Bleed and Feed Operation

OC - Operator Initiates Closed Loop RHR Cooling

OD- Operator Cools Down and Depressurizes the RCS

ON - Operator Maintains Long-Term Steady State Operation

OR - Operator Manually Starts Selected Equipment

OS - Operator Establishes Ventilation 1

OT - Operator Manually Trips the Reactor .

OCH - Operator Initiates RCS Makeup The last two characters are numbers that are specific to the accident scenario in question.

Many of the time windows for the scenarios result in nearly the same sequences of events and, thus, the mass and energy calculations are conservatively considered the same, whenever appropriate. The results of the analyses and the time windows assigned to the human action scenarios are summarized in Table B 1. The order in which each analysis is

' presented in this appendix appears to be haphazard, but in reality is based on the order that the actual calculations were performed, as some analyses logically follow others as they rely on information calculated in previous analyses. The documentation of the calculations performed for scenario time windows or a discussion of rationale for the time windows assumed for the scenarios is presented in Sections B.1 through B.26. ;

3 NHLP1N0078.052589 B-1 Pickard, Lowe and Garrick, Inc.

l __

5' '

~

ATTACHMENT 1 j

.. ST-HL-AE 339o ^

PAGE 9 _ OF _js -

Table B 1 (Page 1 of 3). Summary of Time Windows for Human Actions Scenarlos F

Subsecdon for Human Action - Time Window for Description of Comments +

Scenario Operator Action Analysis - ,

HEOB02 15:r . B.1 Estimated time calculated for h minutes steam generator dryout.

HEOB01 B.8 No estimate required; failure -

rate of 1.0 used because of the very low frequency of occurrence.

HEOB03- --1 Sc =,- B.9 Time window is considered N minutes to be the same as for HEOB02, HEOB04 1 > 12 hours1.388889e-4 days 0.00333 hours 1.984127e-5 weeks 4.566e-6 months B.4 Time window is calculated.

HEOB05 B.10 Not estimated; failure rate of 1.0 used because of very low-frequency of occurrence.

HEOB06 - 1 Luur. B.11 Time window is considered N minutes to be the same as for HEOB02. .

HEOB07 i hour.-- B.12 Time window is considered N minutes to be the same as for HEOB03.

HEOB08 B.13 Not estimated; failure rate of 1.0 used because of very low frequency of occurrence.

HEOBA 1 heur, B.14 Time window is considered N minutes to be the same as for HEOB03.

~ HEOB09 1 hour1.157407e-5 days 2.777778e-4 hours 1.653439e-6 weeks 3.805e-7 months , B.15 Time window is considered

.N minutes to be the same as for HEOB02.

HEOCH01 16 minutes B.2 Time window is calculated.

HEOCH02 ~ 5 minutes - low B.3 - Time windows are level calculated,

~ 7 minutes -

pressure low Note: N/A = not applicable. ]

HHLPIN0078.052589 B-2 Pickard, Lowe and Garrick, Inc.

e

. ATTACHMENU -

'~ ST-HL AE33Fo m PAGE 9 -_0F_ B t

'7 B.1' HE0B02 - TIME FOR OPERATOR TO PERFORM BLEED AND FEED WHEN THE REACTOR AND TURBINE HAVE TRIPPED, MFW HAS BEEN LOST, AND THE AFW IS NOT AVAILABLE Many of the accident sequences and human action scenarios analyzed in Section 15 involve l cases wherein the reactor is tripped and the core is cooled by the steam generators, but no

- feedwater or steam generator PORVs are available, in such cases, the steam generator shell (secondary)-side water level will gradually drop and, eventually, steam generator dryout will ,

occur.: The Intent of this analysis was to estimate how much time is available before steam -

generator dryout would be expected since this time is informative for a number of. human action scenarlos with regard to certain operator actions or equipment recovery.

At full power operation (3,800 MWt) with the stgrrb generator level at the nominal value, each of the four steam generators has a shell side Dqtrit inventory, 136,135 lbm (Reference B-3).

The total shell-side free volume of each steam generator is 7,985 ft2 (Reference B-3). The steam generators provide steam at normal pressure of 1,085 psig and have safety relief valves set at 1,285 psig.

This analysis conservatively calculated the time for reactor decay heat to provide sufficient energy to the steam generators to raise the steam generator pressure from 1,085 psig to the relief pressure valve of 1,285 psig, and to convert the liquid Inventory from saturated water at 1,085 psig to steam at 1,285 psig. The rate of heat transfer to the steam generators is taken to be that of the decay heat because the temperature difference will remain about constant between the primary and secondary coolants. Also, even if the rate of heat transfer is Inlllally higher due to the stored energy in primary metallic components, the energy required will -

eventually balance out to that provided by the decay heat when the temperatures equilibrate.

However, the steam generator secondary temperature will actually increase about 21*F to -

that of ihe saturation temperature at the SRV pressure (1,285 psig), rather than remain at the normal operating pressure of 1,085 psig. Therefore, the time to boll the steam generator dry

'Is calculated conservatively less by not accounting for this lower temperature difference -

between the primary and secondary coolants or by not accounting for dynamic energy transfer to and from the RCS components. The actual temperatures are not that important in

-this analysis as they will equilibrate after transferring energy to and from the metallic components so that the energy provided by the reactor to boil out the steam generatcr will be that of the decay heat.

ec, fer censervedem in prev ldlng-the mlnlmem emovnt of inno te : team generatorWout the inillal Inventory of the steam generators was considered to be that at the low low level lleam generator trip signal when the reactor trip was initiated. This was estimated to be ad go, sis (l@d^"! 90% ;f the n;rma' fvlbpevieNiquid : ..enim y. Steam generator dryout was also considered to occur when 5% of the liquid Inventory was all that remained.

The total energy removal capacity of the steam generator steam and liquid inventory is i estimated to be 1

i ESG = Wsteam

Ahsteam + 0.9 + Wiiquid * (Ahwater + 0.95

Ahwater to steam) i l

i L

l NHLPt N0078.052589 B.1-1 rickard. Lowe and Garrick. Inc. I

m B'* ', ATTACHMENT _1 ST HL-AE 33Po P .,,

,. PAGE.jo 0F _ /S 4 At th'e' time of reactor trip with no feedwater flow, the liquid inventory is assumed to have F . equilibrated alihe normal power saturated temperature of 1,085 psig (1,100 psla). The inillal liquid inventory volume is then calculated as F Vol = liquid inventory mass x Vg i, ion p,i, cio, FIG

~ 3

= 106-t95 lb x .0220 ft /Ib

= 'l9Cll E;995ft3 3

I CICll NN e

and the inillat steam volume at 100% power is Volsteam = 7,985 ft - 2;995 3ft - 4;990 ft3 . At

/ the low low level steam generator trip signal, the steam inventory will be L 5394 /WI 3 '

W,i. m = [4;990 f1 + 0.10(2;995-f(3)) 0 y 1.100 psia 59 M 3 5d90 ft 3

.4001 ft /lb -

13,719 l

= $0;992 pounds .l 1

At-1,085 psig (1,100 psla): hg= 1,188 Btu /lb, h, = 557 Blu/lb At 1,285 psig (1,300 psla): h g= 1,179 Blu/lb, h, = 585 Btu /lb, hfo = 593 Btu /lb F #3,Il9 90,516 r-Eso " 4 X y ts;992 lbmx (1,179 _

- 1,188) Blu/lb + -99 >ci00,135 lbm x (585 - 557) Blu/lb

// e#H

+ .95 x 593 Btu /lb)3 8 8

= 9:09 x 10 Btu = 9M x 10 kW seconds gq,L Z.14 2.7.5 yci,3 or abouth($ x 10 kW8 seconds /3,800 MW - 99:O seconds) initial power seconds. If it is '

assumed that the reactor has been operating at full power for an extended period of time so l that the fission product decay heat is nearly that associated with 6 finite prior operation, then, from integralg:ay heat curve Fipure y B.11, the steam generators are determined to " dry out"in about M hours (4-hou6 N minutes). If the reactor had operated at less than full l power, this time would be extended because of the reduced decay heat level and the larger initial water inventory. Because the effective liquid density increases as power is reduced ;

due to fewer steam volds and lower temperature, the water inventory increases. !

The above simpilfled analysis is intended to give an approximation of the time available for t taking possible corrective actions. It neglects heat capacity effects (due to temperature L changes in the RC5%ater inventory, the core, reactor internals, vessel, and piping) as well as delays in reactor scram or steam generator isolation, which are initiating-event dependent.

Nevertheless, the rather long period of steam generator cooling provided by the large steam generator inventory should allow considerable time for operator diagnosis and corrective

- action.

If the RCPs continue to operate during this scenarlo, the dryout time will be reduced because of the additional energy provided to RCS by the RCPs. In i hour, assuming 95% motor l

l l- NHLP1N0078,052589 B.1-2 Pickard. Lowe and Garrick, Inc.

n 1 I'

ATTACHMENT 1 ;

ST HL AEo3380

, PAGE_l L OF L i i

efficiency, the RCPs (4 pumps,8,000 horse power each) would provide about 7.74 x 10 7 Blu of +

energy into the RCS. The RCPs provide energy at the rate of about 0.6% of the rated thermal i power. The additional energy required from reactor decay heat to dry out the steam generators will be 2.14 8 7

. AE - (M9 x 10 Blu - 7.74 x 10 Blu) x 3,600 seconds per hour T.

3,415 Blu/KWh x 3,800 MW TT.8

=#1 full-power seconds of decay heat o.93 and, from the egral decay heat curve 9 hours1.041667e-4 days 0.0025 hours 1.488095e-5 weeks 3.4245e-6 months after shutdown, or minutes, which is Ef;;4higuld c a the i hourbe of accomplished in assumed RCP energy input about & in the above calculation. Therefore, if the RCPs are running duringpls scenario, the tiry for the ,

steam generators to dry out would be reduced from 4-hew, M minutes, to about & f minutes.

As the steam generator water level drops substantially, the recirculation flow within the steam generator and the primary-to secondary heat transfer would be expected to be reduced This would cause primary temperatures to increase and would eventually result in pressurizer PORVs opening, loss of primary (RCS) inventory, and eventual core uncovery if ECCS is not supplied.

Algg. Ass u ~',n3 he 'ndidins evont i is e ion o+

oW sWe power ( LOS P) , Se enYlco nominaI dc6A generator inve dor $ m uld he avedeb fe Io remove GCas heat ow'oS re *#c " P-NHLPIN0078.052589 B.1-3 Pickard, Lowe and Garrick, Inc.

z .

s /

2- .

h at 1.000 10 .

N-

. ;;; Actually used E

g Figure 3-2 of NUREG/CR-4041 y for the REVISED z calculation.

O -(Reference 5)

E ~

e .

$ N 2

_J % g.

b y

$. 100 - N

% DECAY HEAT _

1.0 y m i INTEchAt. DECAY HEAT -

, p

'ta y %

N us w 'a . O I

> N N

< N 3

o N ..

w %

O j

ASSUMES INFINITE PRIOR OPERATION cc *DOES NOTINCLUDE 1.2 FACTOR o

w H

3 m

n

-" $Tc v> >

o mig

.I

! to I

10 100 0.1 $I g 0.1 1 Lt 'P %

n TIME AFTER SHUTDOWN (HOURS) v%

a a oSH 71 0 1

8- I h D-Figure B.1-1. LWR Decay Heatlintegral Decay Heat Curves

-- ~,

s-.r->- .n -r sn - . ~ ..-c -,s .w - ,, er .m.e v n. .v: .~s- s. -- - . _ _ _ _ , -

. ppyes-rrwnwer r x s

- +

't*' ~? ~ nwrm 3

. f ,.. i s

]

4 J

f .k i

1 r

i

.j i

4 i

'1

(

-i f

e 1

4

{

1'.l

-4

.4g h

.-1 I

z EI1.

ff &

1. t

[

v v . ,

.'V k

s.

ATTACHMENT 2' 3 -

9 -k n y

'4

2. .f

, ';'A n

b' sr >

p

.y .>

N

.ac 7" fa if J 1 s-2 1.. .r <

.4

.t t

'h."

} $

p:

' t, t t i!. -

c i

s',g b.,3 s.

g' ~1 k"' .

E' 4' $6NYYbk% mi4g a. pg ,,, _ 4 , ,.g ( 5 _,_ __ , _

., j

ATTACHMENT 2.

ST-HL AE .5380 +

PAGE . 4- _ 0F __4-

_ asart::;[

' ~

Ql: One 'of the screening criteria employed was that if only one-of three safety trains was in a fire area, then this area 1 was screened from further analysis. However, at Peach Bottom the two most dominant fire areas had only one of three safety trains. Each of these areas was two orders of magnitude higher than the dominant fire scenario at STP.

In light. of the Peach Bottom results, please list which areas were screened by this step' and list what safety systems or their associated cabling are present.

Response

In accordance with Section 8 (Spatial Interactions Analysis) of the . South Texas Project Electric Generating Station (STPEGS)

Probabilistic Safety Assessment (PSA), Subsection 8.5.3-(Scenario Impact Evaluation) the only areas screened from any quantitative review are areas in which events do not effect any system and do not cause any initiating event in the PSA. The following discussion provides additional clarification of the Spatial Interactions Analysis which was performed.

(z The STPEGS PSA utilizes a spatial interactions screening analysis as the basis for the fire analysis performed in the PSA. The Spatial Interactions Analysis is described in Section 8 of the.

L PSA. This spatial interactions analysis (SIA) identifies L locations in the plant which correspond with the fire zones identified in the STPEGS Fire Hazard Analysis Report (FHAR). Each '

zone is associated with a fire frequency and a specific inventory including equipment, components, control cable, power cable, other hazard sources, and mitigative features. These areas are then considered as potential fire locations which define scenarios requiring evaluation. These scenarios are summarized in Appendix l D, Table D-6, in volumes 6, 7 and 8 of the PSA.

In order to perform the evaluation, each scenario is assigned to

one or more of four classes (Class 0, 1, 2 or 3), and then further as meeting one or more of ten guidelines which identified L specifies the basis for initial screening. These classes and criteria are defined in Section 8, pp. 8.5-3&4 of the PSA. The class and applicable guidelines for each scenario (Items 10 & 11) are- identified in Table D-6. It is also indicated in this table, based on the application of the guidelines, whether further quantitative screening (i.e., beyond the guidelines) is to be performed (Item 9).

Class 1, 2 or 3 scenarios were subjected to initial quantitative

, screening per the applicable guidelines. Class 2 includes all scenarios which affect one or more trains of a single system only (for those systems which are modelled in the PSA). Only Class O scenarios (" scenario does not affect any system and does not cause u any initiating event in the plant model") are ruled out from further consideration (per guideline 1, "if a scenario is in Class l

0, its further study is not warranted for purposes of risk assessment.")

l-

i

.-. ATTACHMENT 4 ST HL AE-3320 g .. PAGE _J2 0F 13

-Q2: The most dominant scenario was in the control -room.

However, the methodology employed lLn the quantification varies substantially from past PL&G fire PRAs and also is at variance with testing results from large scale enclosure tests. In past PL&G fire PRAs, the control room has been assumed to be abandoned and control of the plant is taken ;

from the remote shutdown panel. Sandia sponsored large scale enclosure tests have shown that cabinet fires generate such intense smoke that within 6-8 minutes control ,

of the plant from the control room would be virtually impossible. These tests were conducted with control room ventilation rates of up to ten room changes per hour.

Therefore, the most likely scenario would be smoke-forced abandonment of control room and subsequent ~ control of the plant .from the remote shutdown panel. If -the remote shutdown panel is truly independent of the control room, I then it makes no difference whatsoever where the fire originated because all initial potential damage to safety ,

l controls would be bypassed. Please explain why STP is either at variance in control room design from past PL&G PRAs or what other factors led the analysts to modify their previous methodology. Using the past methodology for control room analysis would have the effect of increasing core damage frequency estimates by a factor 'of approximately fifty.

Response

Several factors have influenced the approach taken in the STPEGS PSA to the control room fire analysis. Factors which influenced this approach include a more detailed focus on the modelling of external events such as fires in the control room, an expanded data base- for control room fire events such as that utilized in the fire analysis performed on the Surry plant for NUREG-1150, and ,

-the impact of the STPEGS independent three-train design (n) the consequences of fires.

1 Past PRAs have- focussed more on the internally-initiated event analysis due to the greater interdependency of systems design in older plants than the independent three-train design of STPEGS. l As- a consequence, the approach taken in previous PL&G-fire PRAs I has been more conservative in assuming abandonment of the control H

room in the case of a fire while concluding that even in such case, fire-induced core damage is a relatively small contributor (on the order of 10% plus or minus).

1 The STPEGS PSA fire analysis assumes a mean initiating event l frequency of 4.9E-3 for control room fires. This frequency is j taken from a paper by M. Kazarians and G. Apostolakis ('Modeling '

Rare Events: The Frequencies of Fires in Nuclear Power Plants,"

June 1982). This control room fire frequency is based on a single event which occurred during shutdown at Three Mile Island in 1979. The fire analysis completed for NUREG-1150 for the Surry l

l L

ATTACHMENT D ST HL-AE- 33fo

,, .. PAGE i 2 _ 0F l

Power Station uses an initiating event frequency of 1.8E-3 (NUREG/CR-4550, "NUREG-1150 External Event Risk Analyses: Surry Power Station," September 1989, Table 5.5), a factor of approximately 3 lower than that used in the STPEGS PSA. This control room fire frequency is based on four events between 1978 and early.1983, including the Three Mile Island event (NUREG-4550, Appendix E, p. E-9). None of the four control room fires in the data base lead to the abandonment of the control room. NUREG-4550 assumes that 1 of 10 control room fires leads to abandonment of the control room (see Section 5.10.4 of NUREG-4550).

The STPEGS control room design is such that a fire on a control panel would be quickly detected by smoke detectors placed near the intake to the CR HVAC system inside the enclosed control phnel housing. Separation is provided between panels and to a great extent between controls on the same panel. The fire would be '

extinguished quickly because of the detection and HVAC design and-because the control room is continuously manned. NUREG-4550 also !

takes credit for a factor of 10 reduction in control room fire i frequency because of continuous occupation (Section 5.10.4 of NUREG-4550). STPEGS has not taken this credit, i At- STP, transfer of control to the auxiliary shutdown panel (ASP) provides control of safe shutdown equipment independent of the control room. A fire in the control room would disable equipment controls which would be restored by transfer to tha ASP. The- assumption in the STPEGS fire analysis does not take credit for -transfer to the ASP since the equipment controls disabled by the control room fire represent the more limiting-condition in terms of equipment available for plant shutdown.

i l

E

ML20012A129

Text

Response

Response

Navigation menu

Search