ML20236P640

Review of Recirculation Pump Trip Design for Brunswick Steam Electric Plants, Informal Rept

ML20236P640
Person / Time
Site:	Brunswick, 05000000
Issue date:	07/31/1987
From:	Poloski J EG&G IDAHO, INC., IDAHO NATIONAL ENGINEERING & ENVIRONMENTAL LABORATORY
To:	NRC
Shared Package
ML20235A701	List: ML20235A699 ML20235A733 ML20235B312 ML20236P640
References
CON-FIN-D-6023 EGG-REQ-7766, NUDOCS 8708120450
Download: ML20236P640 (12)

Text

GWD x s.2.;p ,

a g ,g; W n,; -

. s sy + .sr

^pg; L . .y y. M; .:*9 - L vW*cM) iW:- ATTACHMENT 1 L f g '.Ab. ,-%i. - @mW/MT@

4 e

r w% ' ,- w eWn EGG-REQ-7766 W M87

[

g 4 gm M3 4%w%w%a. m-m :n m e m ..Ar f

+ q r x w :kk<. g g. ; - n J , , g' p 3.,p. W*WfAWWe m >pq g . .

li & u INFORMAL REPORT

). --

. y-g up rsw.,

pW o ,3 ?, .;

I '

4 ' {' \ l,, ,,

W.a '

c ., xy%

P ', f,,. lGQ .4 3:ll2pD.

.. -V% er; - -

- p. = -

.g W'4. g. + r .%g..,' ,Mn,.p 7,g. .(-;L..,A. '

REVIEW 0F THE RECIRCULATION PUMP TRIP DESIGN S N, hh '

FOR THE BRUNSWICK STEAM ELECTRIC PLANTS W%$0bO!Btory &%Q W@h i

$c ..h W sk-, ~ n~ .

M $ % N p...g % ,9,G M -

p,. . .i p:}h jMGWg88US.W '

ADepenmeniy g,wl@l I, ' # 4. N John P. Poloski y mg NdFMN

. % SA s m&pW W;dlWW %.m .

w~ Th ' $;N i$m-. w mEddi "

m&_ .r

, y i * . . . #. - da ,

e_ .

b;/M' N h Y,1 s 4 #4 y ~ 5:kil'* , h/N,

'r,'

%M:  % y%NM%g7 j

%Q%)p$O[f*[if sM' y u *)4 e

, w.- w n.hM:2t -

,' '.qy e a,," ,' .

- %' y, m,%.  :-a <

h,N .

"AW c %)4 , jw -

m .a 4@Mw M ' '%e. wap , , , . .

w ' r mww u]. ,'

m

.gc w

~ '.w,o <,

.+ ,  ;

f $q!.F u -f. 'yh g#ty ~  ?

, g ,.

A m.w.m' W Vfc. & l,yC 6.,V. 9 ;4 ) ,

,wm> .c.

' m. ,m, m.. 8 , '

m o , +,

u%

T ;;m h

. i;g;;; n. q , ,

, s9, ,

++

//

SysdW.M W ,

o oy, pe,rc # m. 4

%.m. m . <. 1r Lsy/ g , l 4 .. ,

t.

r-V -

ij

-5,[ p) .. ;

/,

J' .

j % .hf.T 2 '@:;y #

"I ' '

' .., u* f\  :

.[' e' -- t' y

a p,; y;.m* s, m* ...

~

, c) ,

r gh 4 lw. . . ; - s ,: g 4 f / i I

) pd@ < /Q e. .

~

,- ,1 .':-7."' c. ;.! a P3 /

/

& n, w f;. ,. I q, q-q; *L 2.~ u Q- p%.py. . . , , m, , s .+ -

rf l\

.I I ,

\

. wu ym(4cx nUp .-.,s ,,,:4 A e - , m~

, 3.

')

s ; - r ,

n, _w x

$a, ,MWO. u % e .

e mw

~ a. maa,,,,,, .

preparee for tne s.,y*. g;> , w)J$f  %,#9a w mfw ,

yi,Q TN, wem,,,,,,,,,,,,, .s 3M #yl e #g54,llA@ ;

J.S, HUCLEAR REGULATORY COMMISSION 4

wma g racy,ygnoog;p u.

nem 3,:r N i "5 p:' e [c// 4h,,..Qg.s, YUf i[lj1*

p p r 3'

sw > .7., f[c f'

,264Q%

~C W ,3 W yff*L,~ ~g' 'yp-U, "m'h ',. . ,

pW,kt s;g. tag mn ul@a , .;a w

, g.Qn.g%a.g4, ,

M ye )S,ht}y*I I; $ i Q.s'. W. @ ^Lf ;

) fYaje y4k ,,?h i' h '

.t .

y f j' 4.18,

a. ?, f y%

7 .,ws.an n. 3n y my 3 Q;PW:y;7 ns ou e

' gya g p com -

f; .; .., d. . pV .:)e, g m. 4.o ,#

.a y v#..acy%myw m- y M.

, w: L l q [l ' '

Y I 3niz'

.v_ ,>., } &

cn  : ,  ;+o

_-Y' w_ - _ . . .

e

• EGG-REQ-7766 Technical Evaluation Report REVIEW OF THE RECIRCULATOIN PUMP TRIP DESIGN FOR THE BRUNSWICK STEAM ELECTRIC PLANTS John P. Poloski Published July 1987 Idaho National Engineering Laboratory EG&G Idaho, Inc.

Idaho Falls, ID 83415 Prepared for the U.S. Nuclear Regulatory Commission Washington, D.C. 20555 Under DOE Contract No. DL-AC07-761D01570 )

FIN No. 06023 I 1

l I

i

ABSTRACT The results of a review of the Carolina Power & Light (CP&L) proposed design upgrade of the Recirculation Pump Trip (RPT) for the Brunswick .

Steam Electric Plant Units Nos. I and 2 is presented herein. CP&L states the design upgrade for the Brunswick units are equivalent from a reliability perspective to the NRC approved Monticello RPT design. An availability analysis utilizing fault tree techniques was performed by CP&L to demonstrate the equivalency of Brunswick design to the Monticello l

design. The review examined the accuracy and completeness of the logic l models developed by CP&L. The failure data utilized in the quantification l of the logic models were examined to assure the availability estimates are valid. The results of the review indicates, both qualitatively and quantitatively, that the Brunswick design is not equivalent to the Monticello design.

I FIN No. D6023 Review of Plant-specific Licensing Actions for Operating Reactors.

ii

e YW *

SUMMARY

L A technical review of- the Carolina Power & Light (CP&L) submittal pertaining to the proposed Recirculation Pump Trip (RPT) design upgrade for the Brunswick Steam Electric Plant Unit Nos. I and 2 has been

. completed. The submittal states the design upgrade of the Brunswick RPT is equivalent to the NRC approved Monticello design.

The-results of the review indicates that the design upgrade may not be equivalent to the Monticello design. The CP&L analysis did not include a dominant failure mode associated with the trip coil devices of both the Monticello design and the Brunswick design. The Monticello design utilizes redundant trip coils compared to a single trip coil for the Brunswick design. Failure to trip the recirculation pumps require a single trip coil failure for the Brunswick design whereas the existence of two trip coil failures is required for Monticello recirculation pumps failing to trip. Also, the CP&L analysis utilizes identical failure probability estimates for the two circuit breaker types even though the Brunswick circuit breaker was of a different design.

The faelt trees were reevaluated incorporating the dominant failure mode of the trip coil. The failure probability estimates were also updated to reflect breaker differences. The reevaluation conclusively demonstrates, both qualitatively and quantitatively, that the iwo designs are not equivalent.

h iii

,- . 4 .

CONTENTS ABSTRACT ................................................ 11 1

SUMMARY

..................................................iii 1

l

• INTRODUCTION ............................................ 1 REVIEW FINDINGS ......................................... 1 FAILURE OF THE TRIP COIL DEVICE ..................... 2 FAILURE PROBABILITIES ASSOCIATED WITH THE FIELD AND MOTOR CIRCUIT BREAKERS .............................. 4 REFERENCES .............................................. 6 I

iv

} (

.( 8

, i '$ hl REVIEW OF THE RECIRCULATION' PUMP TRIP DESIGN'FOR THE BRUNSWICK STEAM ELECTRIC PLANTS

, INTRODUCTION i- ' Carolina Power & Light' Company's (CP&L) submittal of information (Serial:

NLS-87-074, 10 CFR.50.62)l relating to the reliability of the

. Recirculation. Pump Trip. (RPT). for the Brunswick Steam Electric P1 ant (BSEP), Units:1.and :2, was technically evaluated. The reliability assessment was performed by CP&L to demonstrate the BSEP RPT design will function in a reliable way equivalent to the Monticello or modified Hatch designs. These latter two designs were. determined to be acceptable by the NRC as reported in th'e Safety' Evaluation Report of Topical Report NEDE-31096-P, " Anticipated Transient Without Scram, Response to ATWS Rule.

10 CFR 50.62."2 The conclusions contained in the CP&L submittal states that the BSEP RPT design' is asireliable as the approv'ed Monticello design. .In order to suppo'rt their claim, CP&L performed a fault tree analysis.of the BSEP and Monticello RPT designs. Based on the developed fault.models, failure probabilities were assigned to the failure events depicted.by these models and the unavailability for each design was estimated.

REVIEW FINDINGS

~CP&L's design of the BSEP RPT upgraded the original BWR/4 trip logic from i a one-of-two to a two-of-two logic on either pressure and/or level for tripping both recirculation pumps. This upgrade is consistent with the

.. Monticello trip logic scheme. However, CP&L did not upgrade the trip j mechanisms for interrupting power to the recirculation pumps. BSEP utilizes the original BWR/4 design of tripping the main 4KV feeder breakers to the drive motors of the recirc motor-generator sets. The Monticello design interrupts power to the recirculation pumps by opening the field circuit breakers to generators of the recir motor-generator sets. The tripping of the field circuit breakers is accomplished by redundant trip coils receiving a trip signal from their respective trip 1

L

g 4 e a logic bus. The BSEP trip of the main 4KV breakers utilizes a single trip coil for each breaker.

The fault models developed by CP&L accurately reflect the above mentioned j differences. The models were developed to a level of detail to identify ,

the significant contributors to RPT unavailability to trip both recirculation pumps. The fault models for the two designs were ,

qualitatively and quantitatively assessed during the course of review to j

determine the validity of CP&L conclusions regarding the equivalency of the two designs with respect to reliability. Two deficiencies found during the course of the review indicate the two RPT designs are not

equivalent from a reliability assessment. These deficiencies are determined to be important and there is no relevant information contained l in the CP&L submittal that would invalidate the importance of these issues to the overall conclusion of equivalent RPT designs. The deficiencies identified are

1. The exclusion of the dominant failure mode of the breaker trip coil, and;

2. The assignment of the same failure probabilities to the 4KV breakers of the BSEP design and the low voltage field circuit breakers of the Monticello design.

FAILVRE OF THE TRIP COIL DEVICE From a mathematical point of view, two equations are equivalent if for the same set of values assigned to the variables, the equations yield the same result. With this thought in mind, Boolean equations were derived from ,

the logic models developed by CP&L to assist the review efforts in determining the equivalency of the BSEP and Monticello RPT designs. In i addition, the Boolean equations were updated to incorporate the dominant failure mode of the trip coil device. CP&L identified the only failure mode of the trip coil device as "open coil". Traditionally in past PRAs and reliability / risk based studies of Reactor Protection System for 2

improvement of surveillance testing, the dominant failure mode assigned to a trip or shunt device has been " failure to operate'. The Boolean equations for the BSEP design are as follows:

1 BSEP-RPT = BREAKER-A Faults + BREAKER-B Faults + TRIP COIL-A Faults +

, TRIP COIL-B Faults (1)

TRIP COIL-A FAULTS = TC-A-FTO + TC-A-OPEN + LEVEL-A LEVEL-B + (2)

LEVEL-A LEVEL-D + LEVEL-B LEVEL-C + LEVEL-C LEVEL-D TRIP COIL-B Faults = TC-B-FT0 + TC-B-0 PEN + LEVEL-A LEVEL-B + (3) l LEVEL-A LEVEL-D + LEVEL-B LEVEL-C + LEVEL-C LEVEL-0 The Boolean equations representing the Monticello design are as follows:

Monticello-RPT = BREAKER-A faults + BREAKER-B Faults + (TRIP COIL-Al  !

Faults TRIP COIL-A2 Faults) + (TRIP COIL-B1 Faults TRIP C0ll-B2 Faults) (4)

TRIP COIL-Al Faults = TC-Al-FTO + TC-Al-OPEN + LEVEL-A + LEVEL-C (5)

TRIP COIL-A2 Faults = TC-A2-FTO + TC-A2-OPEN + LEVEL-B + LEVEL-D (6)

TRIP COIL-B1 Faults = TC-B1-FT0 + TC-B1-0 PEN + LEVEL-A + LEVEL-C (7)

TRIP COIL-B2 Faults = TC-B2-FT0 + TC-B2-0 PEN + LEVEL-B + LEVEL-D (8) l Simplification of the above equations results in six single events that  ;

fail PRT for the BSEP design compared to only two single events that fail l RPT for the Monticello design. Based on a qualitative examination of the cutsets, the two designs are not equivalent. The number of single i failures that fail RPT for the BSEP design as compared to Monticello I design is a factor of three greater.. Single failures are dominant to both designs. Their importance of these failures need to be fully understood from a probabilistic evaluation. The failure probabilities estimated in the CP&L submittal were assigned to the cutsets generated from the 3

simplified Boolean equations for the two designs. A failure probability of 1E-4 per demand as assigned to the " fail to operate" mode of the trip

-device. The Boolean expressions were quantified utilizing the above mentioned estimates. The unavailability computed for the BSEP RPT design is 2.7E-3 compared to 2.5E-3 for the Monticello design. From a quantitative perspective, the two expressions yield different results when the same probabilities are applied to the failure events. .

FAILURE PROBABILITIES ASSOCIATED WITH THE FIELD AND MOTOR CIRCUIT BEEAKERS As mentioned earlier, the BSEP design trips the 4KV circuit breakers whereas the Monticello design trips the field circuit breakers. A failure probability of 1.25E-3 per demand was assigned to failure mode associated with these breakers. There is no data provided in the CP&L submittal to support using this value for both breaker types. From an engineering viewpoint, .the assignment of this value to the two different circuit ,

breakers is inappropriate. The field circuit breakers are low voltage DC as opposed to the 4KV AC circuit breakers. Physically the 4KV breaker is

much larger than the low voltage DC breaker, therefore, requiring a greater amount of energy to trip the breaker. Based on these differences, data sources 3,4 were reviewed to ascertain if a discernible difference j existed between the failure rates for these two circuit breaker types, j Reference 3 provided estimates for the GE population of breakers for the j

" fail to open" mode. For low voltage category, a point estimate of 2.9E-5 per demand was reported. For the medium voltage category, a point estimate of 1.1E-4 per demand was reported. Reference 4, which evaluates l the unavailability of RPT, identifies an unavailability estimate of 2.95E-3 l.

for the field circuit br'eaker compared to an estimate of 6.2E-3 for the 4KV drive motor breaker. The Boolean expressions were quantified incorporating the estimates contained in Reference 3 and 4. The results are tabulated below.

BSEP Monticello Reference 3 data 4.22E-4 5.8E-5 Reference 4 data 1.26E-2 5.9E-3 l 4

(- The quantitative results show the'two designs are~not The equivalent.

Monticello design being the more reliable from a quantitative well as from a qualitative evaluation. e analysis, as As a result, the CP&L claims to equivalent reliable designs cannot be supported at this e time ba information contained in the CP&L submittal.

j mentioned references indicate a difference betwe probabilities associated with the field circuit breakers breakers. motor and the 4K BSEP and Monticello designs.These differencese significantl l

l l

l L

5

REFERENCES

1. Carolina Power & Light Company, " Brunswick Steam Electric Plant, Unit Nos.1 and 2, ATWS - System Descriptions", Serial: NLS-87-074, 10 CFR 50.62, June 1987.

2. Safety Evaluation of Topical Report (NEDE-31096-P), " Anticipated Transient Without Scram; Response to ATWS Rule, 10 CFR 50.62", October 1 1986.

3. S. R. Brown, " Data Summaries of Licensee Event Reports of Protective ,

Relays and Circuit Breakers at U.S. Commercial Nuclear Power Plants, January 1, 1976 to December 31, 1983 (Draft)," NUREG/CR-4126, January 1985. . ,

4. Millstone Unit 1 Probabilistic Safety Study, Vol. 2.

I

~

6

5.5,  !

ATTACHMENT 2 N %)"'E G R S Idaho INTEROFFICE CORRESPONDENCE i DATE: August 17, 1987 i

TO: Art Nolan  !

FROM: John Poloski

SUBJECT:

RECIRCULATION PUMP TRIP FOR BRUNSWICK PLANTS - JPP-4-87 A review of data presented by CP&L and additional LERs was performed to assess the reliability of the Brunswick recirculation pump trip to the approved i Monticello design. Unavailability estimates were calculated for both designs l

incorporating different sources of information and assumptions pertaining to j tne breaker and shunt trip devices. The range of estimates indicate a j potentially better design in the Monticello RPT as compared to the Brunswick. 1 9go cc: D. J. Hanson 36H/SE13 E. W. Roberts Central File J. Poloski File l

l I

(

l 1

l l

l "Providing research and development services to the government" l

. . . _ . . _ _ , _ _ . _ _ , . . . _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - . - _ - - _ - - _