ML20005G587

Technical Review Rept AEOD/T921, Inadequate Capacity of 4,160-Volt Switchgear at Fitzpatrick

ML20005G587
Person / Time
Site:	FitzPatrick
Issue date:	12/27/1989
From:	Mazumdar S NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD)
To:
Shared Package
ML20005G581	List: ML20005G580 ML20005G587 ML20005H198
References
TASK-AE, TASK-T921 AEOD-T921, NUDOCS 9001190364
Download: ML20005G587 (3)

Text

'*

y.

u,..

t 4e

>S i

, g,3 s o

?

a MODTECHNICALREVIEWREPORT~

q f b

.FitrPatrick.

TR REP 0kTING NO.: AE0D/T921 l

'z ul;1T:

J'

DOCKET NO.'

yzS0-333 DATE:-

~ LICENSEE::

LNew: York Power Authority:

EVALUATOR / CONTACT:

S. Mazumdar

'.NSSS/AE: A GE-

SUBJECT:

JINADEQUATELCAPACITY OF 4160V SWITCHGEAR -

a

AT FITZPATRICK -1AE0D/T921 y

L

SUMMARY

a EM l0n July 27,11989,.the? licensee issued 3 Justification'for Continued Operation gm

'(JCO)?of the Emergency. Diesel Generators (EDGs) in. parallel with the Normal q

W DStation: Service Transformer (NSST) during the monthly 1-hour testing since, funder-such a:modciof operation, the maximun fcult current may exceed the momentary and. interrupting' ratings of the 4160V switchgear, The licensee-

? justified such operation on the basis that during this testing-the probability "M

of" occurrence of> a fault. exceeding the breaker rating is less than 9.72 E-7.

t LThis. probability' figure was based on an LER data base search which indicated 1

p Ltt:at between ~1972 and 1989 there were only five.3-phase and phase-to-phase faults ~

'infall the nuclear plants in the USA.

'Outi NPk95 and SCSS search indicates that not> all electrical faults are covered.

1

)'

by-LEr#,;and that.the.LERs do not report the faults in detail to establish

, beyond doubt whether the faults were 3-phase or single phase, or between two

~

ph a'ses..Also,-the number of electrical faults identified in our LER data base search was substantially more (approximate 1y'57 between 1980 and 1989) than'the i

Lfigures-used by the licensee-in -their JCO.- Thus, the probability approach used in the"JCO)is. questionable.

In"aEconference' call:on October 2, 1989 between the NRC and the licensee, the licensee! agreed not:to: test.the EDGs-in parallel with the NSST and ensure (that the-4160V switchgear is not operated int ary configuration in-which the

breakers may be potentially exposed to fault currents-beyond their momentary and 4~ ~

. interrupting ratings.

4 DESCRIPTION'0F. EVENTS Y

t At'Fitzpatrick the EDG monthly' load tests are performed by running the EDG in

-parallel with the NSST, with the parallel connection made at the 4160V bus. On LMay 14,1989, an NRC 1erpection questioned the adequacy of the 4160V switchgear

~

~

capability when the EDGs are tested in parallel with the NSST. On July 14, 1989,

h;allicensee calculation established that during this parallel operation, the f,

7'

' maximum fault current may exceed the switchgear momentary rating by 27 percent

'andLthe1 interrupting "eting by 3.7 percent. On July 27, 1989, the licensee K

issued a JCO,- claiming that the probebility of occurrence of a feult large

+

.enough to cause switchgear failure was less than 9.72 E-7 during the monthly i

EDG tests. This robability figure was arrived at from an LER data base search which reported only five 3-phase and phase-to-phase faults in all nuclear plants

+

inithe:: USA between 1972 and 1989.

((

q-N M!9001190364 5991227-M-

M* F N) K >

W PDR.ADOCK 05000333 Ji -

~,

"d 15 1 1.

PNU.

g:

1%

y m,

y C '

SIMILAR EVENTLSEARCH t

1To: verify the: data used in the.000, wt initiated a search of the LER data bese-

through the Oek Ridge: National Laboratory (ORNL)^..On September 6,1989 ORNL

completed an SCSS search on:short circuits in the medium voltage AC system between 1980 and 1989. A copy of their ' response is attached. Our review of the LER labstracts. indicated that.only 57 of these,LERs were.related to short circuits 9

in the medic ~ voltage system.' Vc.have also. conducted an NPRDS search which 4

identificc' s.otaliof 27 events involving.shcrt circuits in the medium voltage F

system Letween 1978 and 1989. Twelve of these 27 events are also identified by the ORNL SCSS search.-;The LER1 data base search on whibh the licensee has based

~

m the JC0 calculation. identified a total of 59 LERsLon medium voltage short circuits b' s

.in all' USA.nuclcar plants between 1972 and 1989. Out of these 59 LERs only 15 i

~

LERs are connon-with the 57,LERs in the ORNL search.

2 These anomalies ' indicate:

(1) Net all medium voltage short circuits are reported in LERs; (2) The results of an LER data base scarch are highly dependent i

on'the search commands used;1and,- (3) The LERs often do not specify clearly the types ofl electrical short circuits experienced, i.e...whether the faults were H

3-phase shorts;or single li_ne-to-ground shorts or shorts.between two' phases.

Further, incidents of short: circuits may be attributed as equipment failure, and cnot as shorts. Moreover, in. preparing failure reports, tre pre 3arer may be more concernced about-the other aspects'of the failure. Thus, one siould not rely on 1

" data frca an.LER data bese'searchLto arrive at the probability of occurrence of l

short circuits in..the medium. voltage system at nuclear plants.

1MPLICATION OF SWITCHGEAR FAILURE-1 J

' A switchgear as.a protective-device must satisfy two important criteria - inter-

<rupting' duty and= momentary duty.to ANSI C37.010. The interrupting duty is the maximum fault current the breaker must be able to interrupt and the rcrentary j

duty is the reximum current-the switchgear must be able to withstand without 1

M

~ being damaged.< Broadly. speaking, the switchgear momentary duty is about 1.6 n-times the' interrupting duty.

This difference between the interrupting duty and the mordentary duty is due to.the decay in the fault current from the instant the V

fault" takes place to the instant the switchgear interrupts the fault.

L A failure of a medium voltage switchgear in interrupting duty or momentary duty

.m

-is' always viewed with great concern. When an ordinary equipment fails we usually have a" protective device to isolate it. When a low voltage switchgear fails, we usually have another upstreau switchgear as back-up protection.

Furthermore, (as a -Icw voltage switchgear is normally small in size, its failure in momentary duty or-interrupting duty can cause limited damage. A medium voltage switchgear is' a ' comparatively-bully equipment.

Its f ailure in mcmentary duty or interrupting

~ duty can cause catastrophic explosion and/or fire which can damage part or all of; the adjoining equipment.

Thus, such a failure when an EDG is operated in parallel'with the NSST during monthly load tests can cause a loss of the EDG under test, a loss of the NSST, and even failure of the bus transfer from the gj' NSST to.the RSST thus causing total less of all AC power in that train.

Because of the consequences of such a failure of the medium voltage switchgear, it is the well established practice in power utilities, and in commerical, industrial and residential applications, never to subiect a medium voltage u

f a

u n

9F f

7 f

f N

4 ;.y

=3--

g' h,

' iswitchgear beyenb:its. momentary or ;fer in which two power: sources m 4

interrupting rating except for a ' ery short.

v Jduration:such as durin(j a bus trans p,

h.

e

.parc11cledlfor a few seconds.. Such exceptions are' permitted because the, prob-ability oCoccurrence'of.a fault of high magnitude duringJthis short interval LM g., Lis very low.:

j; o.

4 y

[;

CONCLUS10NS' IE k IConsidering:the consequences'of switchgear catastropic failures, it is. desirable l..

tthat the.-medium voltage switchgears are never subjected to currents beyond their m

~ ratings except for-veryL short durations as 'during bus transfers.

s.

ec.:

(

hp

~'

.t I?

l'c n'

l

' ?,

/

D

l. l.

(,

s j[i e

.t a-5 f

A 4

?

i;.

A-

. M --.

• E J

.