ML20107D760

Tornado Missile Risk Evaluation for Nuclear Svc Cooling Water Sys Fans at Vogtle Electric Generating Plant Units 1 & 2

ML20107D760
Person / Time
Site:	Vogtle
Issue date:	02/14/1985
From:	BECHTEL GROUP, INC.
To:
Shared Package
ML20107D753	List: ML20107D760
References
NUDOCS 8502250158
Download: ML20107D760 (13)

Text

.. -

TORNADO MISSILE RISK EVALUATION FOR THE NUCLEAR SERVICE COOLING WATER SYSTEM FANS AT V0GTLE ELECTRIC CENERATING PLANT UNITS 1 AN) 2 prepared for GEORGIA POWER COMPANY by RELIABILITY AND RISK ASSESSMENT GROUP BECHTEL POWER CORPORATION WESTERN POWER DIVISION 8502250158 850214

[DR ADOCK 05000424 PDR .

t PLANT V0GTLE TORNADO MISSILE EVALUATION REPORT I. INTRODUCTION AND

SUMMARY

Tornado-generated missiles are one of the potential hazards to the safe operation of nuclear power plants. This report evaluates the probability of damage to the Nuclear Service Cooling Water (NSCW) system cooling tower fans at Vogtle Electric Generating Plant (VEGP)

Units 1 and 2 from tornado-generated missiles. The analysis demonstrates that the median annual probability of tornado-generated missiles striking and damasing one or more NSCW cooling tower fans is extremely small and well below the NRC staff objective of 10-7 per year for postulated accidents causing potential exposures in excess of 10 CFR Part 100 guidelines. On that basis, it is concluded that protection of NSCW system cooling tower fans against postulated tornado missile damage is not required.

II. SYSTEM DESCRIPTION The ultimate heat sinks for Vogtle Electric Generating Plant Unit 1 and

- Unit 2 are contained in the Nuclear Service Cooling Water (NSCW) systems. Each unit has two independent, redundant NSCW trains consisting of one cooling tower in each NSCW train. The height of each cooling tower is 44'9". The area of the top of each cooling tower is 6936.3 ft2. Near the top of each cooling tower are four motor-driven fans discharging air through four 471.2 ft2 openings.

1 TOP VIEW OF A NSCW COOLING TOWER FAN DISCH ARGE CY'.1NDER t

A = 8936.3 FT2g f(o=ss-d = 24' 8" '

2

, (A = 471.2 FT 3 During a tornado, offsite power is presumed lost and the reactor automatically trips. Under these conditions, a minimum of three of the four fans in one NSC'd cooling tower per unit are required to provide adequate heat rejection. Thus, total NSCW system failure occurs if two or more fans in each of the two redundant NSCW cooling towers per unit-become inoperable.

t

+

1 ,

s.

9 , - ,. ~ - - - - - - - - , , . . ,,. -

, -. , - - , , , ~ . ,_ e.- ,y ,- ,w-- , ,,s,, ,, - as w. - - --.e-

1

. . 1 l

III. ACCEPTANCE CRITERIA l The NRC's acceptance criteria are contained in the " General Design Criteria (GDC) for Nuclear Plants" [1]. Specifically, GDC 2 and 4 apply to this evaluation and are summarized below:

.A. GDC 2 requires that " Structures, systems, and components important to safety shall be designed to withstand the effects of natural phenomena such as - tornadoes - without loss of capability to perform their safety functions. .. "

5. CDC 4 requires that ". . . structures, systems, and components shall be appropriately protected against dynamic effects, including the effects of missiles . . . from events and conditions outside the nuclear power unit."

The Standard Review Plan (SRP) [2] Section 3.5.1.4 and NRC Regulatory <

Guide 1.76 [3] p ovide further guidance in meeting GDC 2 and 4 requirements. Specifically, SEP Section 3.5.1.4 refers to the acceptance crite:la of SRP Section 2.2.3 which states "... design basis events include each postulated type of accident for which the expected rate of occurrence.of potential exposures in excess of the 10 CFR Part 100 guidelines is estimated to exceed the NRC staff objective of approximately 10-7 per year . . . (The) expected rate of occurrence of potential exposures in excess of the 10 CFR 100 guidelines of

. approximately 10-6 per year is acceptable if, when combined with reasonable qualitative arguments, the realistic probability can be shown to be lower. . ."

IV. ANALYSIS The probability of damage to MSCW system cooling tower fans depends on three factors:

A. The tornado occurrence rate at the Vogtle plant site

!" 3. The conditional probability of one or more tornado missiles being generated and entering the top of a NSCW cooling tower given the tornado occurrence, and C. The conditional probability of cooling tower fan damage, given

'that a tornado-generated missile has entered the top of a cooling tower.

l L The tornado occurrence rate at the VEGP site is estimated using-National Weather service historical records of tornado strikes for the

[

site region between years 1950 through 1982 (4). Table 1 shows 5th, 50th, and 95th percent.11e estimates of the annual frequency of tornados-in a 10,000 square alle. area surrounding the VEGP site, the distribution of tornado path areas, and the resulting 5th, 50th, and=

i 95th percentile estimates of the annual probability of a tornado

! occurrence at the VEGP site. Values shown for the annual frequency of

. tornado occurrences in a 10,000 square mile area surroundias the site.-

l- '

and the distribution of tornado path areas are based on historical data 2

i l E . .. l._ _ . . _ ..~._ __~- _. _ ._. . -

for the entire state of Georgia. The statewide data yield higher (i.e., more conservative) estimates of the tornado occurrence probability at the VEGP site than do local occurrence data. To allow comparison of predicted tornado occurrence rates with those derived independently by NRC, Table 1 reports mean values of l estimated parameters in addition to the 5th, 50th, and 95th

. percentile values.

l The conditional probability of a tornado missile being generated and entering the top of a NSCW cooling tower, given the 6,ornado occurrence, depends on the following factors:

A. The number of potential missiles near the VEGP site l B. The conditional probability of a potential missile becoming j airborne (or injected), given the occurrence of a tornado '

l C. The conditional probability of missiles being transported from

! their origin to the target, given that they become airborne, and D. The target area.

The number of potential missiles assumed in the analysis (Table 2) represents a conservative generic missile distribution (see section Y) and is based on data from Electric Power Research Institute (EPRI) surveys at seven nuclear power plant sites [5]. The probability of potential missiles becoming airborne is calculated using a missile model developed at Jet Propulsion Laboratory (JPL) [6]. The

, conditional probability of missiles being transported from their origin

to a target is based on a statistical mechanics model (7], (8]. This approach develops a modified Green's function to quantify the probability of the airborne missile striking a unit target area at some distance and elevation from its origin. This probability is then

! multiplied by the area of the target to obtain the total probability of strike.

Each of the above factors has uncertainty associated with'it. For this-reason, a probability distribution is used to represent the uncertainty' associated with each factor. These uncertainties are propagated throughout the analysis. Therefore, the final result is not a single value for the probability of damage, but a distribution of values. .The 5th percentile, median (50th percentile) and 95th percentile values are reported in Table 3. The median value or "best estimate" [9] is compared to the NRC acceptance criteria given in Section III above.

The conditional probability of cooling tower fans being damaged, given that the tornado-generated missile or missiles have entered the top of the cooling tower,'is conservatively taken to be certain and total.

That is, the conditional probability is taken as unity.

=3

. - . . - . ---  :**t -n  :. .. . . = .L --..:'- - ..

~

V. ASSUMPTIONS AND CONSERVATISMS l The following assumptions are used in this study.  !

A. A tornado missile strike in the open top of any one coollag tower represents total NSCW system failure for that unit (see conservatisms A and D below). This implies that one NSCW train is out of service throughout the tornado event and hence the failure of the redundant train's cooling tower is suffielent to fall the entire NSCW system.

B. The distribution of potential tornado missiles by number and length are based on an EPRI survey of seven nuclear plants [5).

C. O'ne half of all potential missiles are distributed up to 20 feet above grade, with the remainder at grade level.  !

D. The number of unrestrained missiles postulated for this study is equal to 10 percent of the total number of missiles.

Assumptions B, C, and D are based on engineering judgement and lead to a conservative estimation of the distribution of missiles when compared to previously published literature [5], [8]. The median number of missiles (6000) used in this analysis is greater than the 1000 and 2650 missiles derived in references [5] and [8), respectively.

Conservatisms incorported in this study are as follows:

A. Relationships Between Missile Strikes, Damage, and Activity

?

Releases:

l

1. Missile-inflicted damage is certain and total

( 2. Damage leads directly to activity releases in excess of-10 CFR 100.

3 Missile Characteristics:

1. The angle distributions for all potential tornado-generated missiles are randon.

2. All potential tornado-generated missiles are at randonL l orientation. .Nowever. in calculating the probability of

lajection of potential missiles, their maximum cross-l sectional' areas are assumed to be perpendicular to the wind. This maximizes the probability of potential missiles l becoming airborne.

l

.C. Tornado occurrence Frequency The tornado occurrence frequency-is based on a 33-year historical  ;

l

' record. The data are fitted by a legnormal distribution having a larger mean-and spread than the empirical distribution.

4

- - . . . .. . .+ .

i .

~

i i

i 1

j. D. Geometric Factors:

l 1. Sheltering by other structures is neglected, i

i l 2. A missile strike in any cooling tower opening results in i failure. No credit is taken for the existence of four l .

24'6" diameter openings. Instead one 94' diameter opening

is assumed.

3. No credit is taken for the fact that multiple strikes are needed to incapacitate at least two fans in each of two

NSCW cooling towers per unit.

VI. RgSULTS AND CONCLUSIONS i

] The results of the probabilistic analysis demonstrate that tornado

" missile damage to the NSCW cooling tower fans has a low probability of occurrence. The lower bound (5th percentile), median (50th percentile) i and upper bound (95th percentile) values are reported in Table 3. The

median value, 2 x 10-9 per f acceptancecriterionof10yearissubstantiallysmallerthantheNRC per year.

f f

1 i

a I

i I

l

(

t i

f

~

. a e m' . , g. a t- e

l Table 1 l PROBABILITY OF TORNADO OCCURRENCE AT THE VEGP SITE Lower Median Mean Upper (5th (50th (for (95th Percentile) Percentile) comparison) Percentile)

Annual Frequency of 1.39 3.23 3.70 7.52 Tornado Occurrence in an Area of 10,000 m12 Surrounding the VEGP Site Tornado Path Area 1.10(10-3) 6.77(10-2) 1.55 4.15 (sq. alle)

Annual Probability of 3.26(10-7) 2.18(10-5) 5.72(10-4) 1.46(10-3)

Tornado Occurrence at the VEGP Site i

l 0001r 6

9- .

1 t

i Table 2 f

DISTRIBUTION OF FOTENTIAL MISSILES Lower Median Upper Local Surface Density 9.40(10-5) 2.40(10-4) 6.13(10-4) of Potential Missiles Near the Cooling Towe.s (ft-2)

Effectiss Number of 2350 6000 15,300

. Potential Missiles on Site and in Vicinity 4

9 1

9 7

1 Table 3 i PROBABILITT OF DAMAGE TO NSCW COOLING TOWER FANS FROM TORNADD-GENERATED MISSILES PER YEAR

• Lower Median Upper (5th Percentile) (50th Percentile) (95 Percentile)

Probability of Tornado 3(10-7) 2(10-5)

Striking the Plant Site 1(10-3)

Conditional Probability 9(10-8) 1(10-4) 2(10-2) of Hitting the Top of the Cooling Tower Conditional Probability of 1 1 1 Incapacitating NSCW Systen Given a Hit Total Probability, Pr 2(10-13) 2(10-9) 7(10-6)

• All numbers are rounded to one significant figure. Simulated distributions

reported in rows 2 and 4 are not legnormal.

a I

l I

l l

0001r 8 4 * - + *w - y m a m 9 -p r

4 4

i VIII. REFERENCES j 1. 10 CFR"Part 50,* Appendix A, " Design Basis for Protection Against  !

{ Natural Phenomena."  !

4 l '

2. Standard Review Plan, U.S. Nuclear Regulatory Commission, 4

NUREG-75/087.

3. Nuclear' Regulatory Commission, " Design Basis Tornado for Nuclear i

Power Plants," Regulatory Guide 1.76, April 1974.

I

4. "U.S. Tornado Breakdown by Counties 1950-1982," U.S. Department of Commerce, National Oceanic and Atmospheric Administration, National l

Weather Service, National Severe Storms Forecast Center, Room 1728 Federal Building, 601 E. 12th Street, Ranses City, Missouri 64106.

5. Twisdale, L. A., et al., " Tornado Missile Risk Analysis," EPRI NP-768, May 1978 EPRI NP-769, May 1978.

i

6. Redmann, C. M., et al., " Wind Field and Trajectory Models for Tornado-Propelled Objects," EPRI 308, Technical Report 1 February 1976.

! 7. Goodman, J. and Roch, J. E., " Conditional Probability of the i

Tornado Missile Impact Given a Tornado Occurrence," Proceedings of the International ANS/ ENS Topical Meeting on Probabilistic Risk l Assessment, Port Chester, New York, September 20-24, 1981, pp. 419-424.

8. Goodman, J. and Roch, J. E., "The Probability of a Tornado Missile

i. Hitting a Target," Nuclear Engineering and Design.11 (1982), pp.

125-155.

9. Apostolskis, G., et al., " Data Specialisation for Plant Specific Risk Studies," Nuclear Engineering and Design li (1980),

pp. 321-329.

I i

t i

9

_ . _ . ~_ ,__ . _ , . . _ _ _ _ . ., ._ . . _ . - . _ , _ ___ _ _ .- .

BNL COMMENTS ON THE V0GTLE AFWS RELIABILITY ANALYSIS The following is a list of comments from Brookhaven National Laboratory (BNL) on tha VEGP Auxiliary Feedwater System Reliability Analysis. The comments were drawn from the Technical Evaluation Report (TER) entitled " Review of the Vogtle Units 1 and 2 Auxiliary Feedwater System Reliability Analysis",

l which was' prepared by BNL. This TER was attached to a January 10, 1985, letter fmm Elinor G. Adensam to Donald 0. Foster.

Comments:

(1) pp. 8. 29 - Pump testing procedure gqg'* "'" gong d ires girgiscus 5 W the m.ss .. . .(

W.4 appl icant. J+%-(L%b{.J.**L ertin"#.*dD$CwrM"dMND ,w...) .

(2) pp.10, 20 C No preaccident operator errors were assumed for the manual valves in the applicant's report. This omission has a significant impact on the quantitative results.

(3) p.10 - Applicant assumed that the probability of a motor-operated valve failing is 5.0E-3/ demand. BNL assumed a 0.2 recovery factor resulting in a motor-operated valve failure rate of 1.0E-3/ demand.

(4) pp.11, 23 - Westinghouse Technical Specifications allow for two inoperable AFWS pumps. However, BNL and the applicant assumed that only one AFWS train can be in maintenance at a time.

(5) p.12 - Applicant did not assume maintenance of the diesel generators.

(6) p.12 - Table on p.12 shows discrepancies between applicant's data and BNL data.

(7) p.16 - The check valves on the pump suctions have had their flappers removed. This could pmsentt operational problems.

(8) p.17 - Emergency procedures for transferring AFWS suction from one CST to the other CST have not been provided by the applicant.

(9) p.17 - Possible common mode failures are discussed in section 9.1.6.

(10) p.18 - Emergency procedure!j for operation of- the AFWS must be provided by the applicant.

(11) p.19 - Applicant states that unavailability due to testing and comon cause human error during testing and maintenance were considered in the fault tree. However, BNL was unable to-find these aspects in the fault tree.

(12) p.19 - The fault trees do not usefully model maintenance acts excluded by the technical specifications (i.e. simultaneous maintenance of AFWS

- tra ins) .

s.

f i

BNL Comments on the Vogtle AFWS Reliability Analysis Page 2 (13) p. 20 - Applicant has maintenance data for DC power, but no data for failure on demand. Further, there is no event for DC power maintenance in the fault tree, but there is an event for random DC power failure.

l (14) pp. 22 In order to perform their own assessment, BNL modified the applican.t's fault trees. One modification included modeling the

' possibility of maintenance on the steam generator intake check valves and stop check valves (Note: This appears to disagree with BNL statement on p.12 that maintenance was not assumed for valves other

' than motor-operated valves). Another modification models the operator failing to close the recirculation valve in the condensate system return line; Therb were also other minor modifications to the fault tree.

e h.

J'

4 February 14, 1985 MEETING

SUMMARY

DISTRIBUTION eseket'Ne(s)?"5'0'-42474257 NRC PDR

-Local PDR NSIC PRC System LB'#4 r/f Attorney, OELD E. Adensam Project Manager MMiller

- Licensing Assistant MDuncan NRC PARTICIPANTS M. - Miller W. LeFave M

e

~

-bec: . Applicant.& ServiceLList - 1 t 's. ] A

• ~

. ,Nd , , i

< eq cN. + '- ,ca ( (( ,, s l

+7 ' ' %, "y ' ' ,,,(' _q,c , , >

-?*

, g jd ,. ,, v , G) w

. j ;- ,

• f .' ' " , 'y[ . - ,

3 s

.. 4 s J", , , !A ' f, ' 4',y , * --' ,,

'F'

,, _J.*e%'

7f  ; _ 7$y e l4*'Q, *

' { p', '. '

e. u a .

r,. ,

1. _% .3 > rg ls n eh R-S ':

_{: n , . . j' . x -

_"'%*+ ..n., . _ ,

y . ..

DI#

mm . , a w

+( _ x _ f '? -(";;*Q r(," i, '

'jj

', .j,3r} >[,-',Lg:

1e i .

• e r ; 3, , -,

- l} )' ~

.'y

.+2 p, > , p ,

t-> .6

.> ,- . { q c,'^

.# , t ,..p,,

t 5

l s . .

Q, fs .~

t

" y j' ft. -

ll p ,' > ,

f[ J 1, ,

• *yl W '

+}-+- q ;?~f ._. } k 'W V' _ _ _

y