ML20091G141

From kanterella
Jump to navigation Jump to search
Discusses Damper & Fan Risk Impact Calculations Performed & Provided by Licensee Personnel
ML20091G141
Person / Time
Site: South Texas  
Issue date: 09/21/1994
From: Butcher E
Office of Nuclear Reactor Regulation
To: Mccracken C
Office of Nuclear Reactor Regulation
Shared Package
ML20091F755 List:
References
FOIA-95-219 NUDOCS 9412080028
Download: ML20091G141 (5)
v • d • e


Text

.-

G'?

^

. NUCLEAR REGULATORY COMMISSION' F

i*

"i ci y ; Mas ~H!$07oN', d.o' 30H6-copi y

g K,,,,,f 44 T-5.Q[

G.W.WW o.

~

geptmber 21, 1994

~ [';; & Q.

MEMORANDUM T0:

Conrad E. McCracken, Chief -

l Plant Systems Branch' Division of Systems Safety and Analysis -

fice fN ar. Reactor Regulation

f. x,s.g robabilistic Safety' Assessment Branch.

utcher% Chief.l.i; i

FROM:

P Olvision of Systems, Safety"and Analysis..

t

, t l

. Office of. NuclearJeactor:t ry/sw.gygy:gulationd 24 '

~

Re

/.

l..

SUBJECT:

RISK IMPACT OF DAMPERS AND FANS--RESULTS OF A SMALL SAMPIE SURVEY C

l This memorandum responds to your request that the Probabilistic Safety Assessment Branch (SPSB) perform a first level assessment of the risk impact l

of dampers and fans. This subject was raised following a recent DET which l

found problems with tornado dampers at the South Texas Plant (STP).

SPSB held discussions with appropriate staff members at four plants.

Information exchanges were held with PRA and engineering personnel at the South Texas, TMI-1, and Grand Gulf plants, and with the Diablo Canyon PRA staff at Pacific j

Gas and Electric's San Francisco headquarters offices.

Damper and fan risk impact calculations, in terms of the core damage frequency reduction importance measure (internal events), were performed and provided by the South Texas, Grand Gulf, and Diablo Canyon PRA personnel. The risk reduction importance measure was used to assess the change in plant core 1

damage freqxncy as a result of setting the pri,bability of an individual i

failure event to zero. The pertinent results are presented below, Additionally, some qualitative dam;,er/ fan risk impact discussions are j

presented for the THI-1, McGuire and Palo Verde plants.

Ouantitative Plant PRA Risk Reduction Determinations _

South Texas Units 1 or 2 j

A summary of the overall STP core damage frequency (C0F) fractional risk reduction importance determinations for fans and dampers, categorized by initiating event, is provided in Table 1.

1 i

i CONTACT:

M. L. Wohl, NRR/DSSA 504-1181

- ~

941200 28941201k F

A g0 c,498 a

v' i

f fhk k N

,kk'h

.NY sh,

  • . 7;p#:,; q.%,1: :

,r u

v.

,r; n

C. McCracken -

t Table 1: Fractional Fan and Damper CDF' Contributions (STP)'

l Fractional Fan'e Fractional Damper l

Initiating Event C0F Contribution'- - a

'CDF Contribution Loss of CCW 4.0E-57

~"

l 0.0 1

Loss of ECW 7.9E-51 M M E 3.5E-3

)

Loss of Control Room

'..N? N?$bN6;

't i

HVAC

'2.25E-2N W'W"'

7b$.1E-4 Loss of EAB*HVAC 1.17E-1 S

~

4.5E-3 l

Total 1.39E-1 8.1E-3

  • Electrical Auxiliary Building As can be seen from Table 1, the chief risk kon'tributors are fans supporting the Electrical Auxiliary Juilding HVAC (11.7%) and fans supporting the Control Room HVAC. The total fan contribution to the CDF;is.about 14%. The greatest l'

damper contribution to the CDF comes from dampers supporting the EAB HVAC, but is less than half of one percent. The total damper contribution to the CDF is l

about 0.8%, quite small.

Grand Gulf i

A summary of the most prominent Grand Gulf core damage frequency fractional risk reduction importance determinations for fans and dampers, categorized by component event, is shown in Table 2.

l 1

Component Event Fractional C0F Contribution Common Cause Failure of OG Rs.

Dampers 9.39E-02 i

Motor-Op. Damper 1Y47F001A-A Falls

'o Open on Demanda 6.48E-02 i

Motor-Op. Damper 1Y47F002A-A Fails to Open on Demand

  • 6.48E-02 Comon Cause failure of SSW Pump 3

House Dampers 5.70E-02 Outside Air Fan 1Y47C001A-A

  • Unav.-Maint. (SSW Pump A) 4.20E-02 4

i Motor-Op. Damper Unay, F0022A-Maintenance 1.70E-02 Motor-Op Damper 1Y47F001A-A Unav.-Maintenance 1.70E-02

' l ' f;, ;

p t

.+.,,.,,3,,

.k l j.,,$,

C. McCracken

, 4'..,

E y,

.-v s-d'f,,

.if

.4

.1 l

As can be seen from Table 2,*.

r?

~

ec con.

ors are common cause failure of the diesel gener,a cos? damper

'abouti9%,'.and common cause

~,

4 failure of the service water pu:;rhoui(dampersf(5 75)MAlso noteworthy are f

the risk contributions of the water pump house failing 'o op.mosor-operated.'d

~ersit ~the standby service c

en'on' demand..(6,

(

Diablo Canyon Unit 1 or 2 '

[

it j

A summary of the Diablo Canyon core. damage frequency fractional risk reduction importance determinations for fans and dampers,is provided in Table 3, i-categorized by initiating event, e

Table 3: Combined Fan and Dameer CDF Contributions (Diablo Canyon)

Initiating Event'

' ' Fractional CDF Cor.tribution Loss of Switchgear 'lentilation 2E-02 l

Loss of Control Room Ventilation '

SE-03 l

Loss of Auxiliary Saltwater Y

3E-03 *

' Contribution from fans only l

As can be seen from Table 3, the chief risk-contributing initiating event in which dampers and fans are involved is Loss of Switchgear Ventilation, but the

)

core damage frequency impact of this event is only 2%.

In the switchgear area, the inverters constitute the largest heat load to be cooled, i

l Oualitative Damoer/ Fan Risk Imoacts, l

j Three Mlle Istead - 1 (TMf-1).j,jQd.

v,M:

In the original THI-1 PRA the chief initiator in which dampers, Loss of Control Building Ventilation (buted 43% to and fans are risk contributors) contrL the over:11 C0F from internal events :This was~ based on the assumption that the electrical power system would fallecatastrophically (NUREG CR-5457) when the control building tesserature exceeded 104' F, subssquent licensee analyses indicated that Loss of Control Building Yentilation would not lead to failure of the electrical power system. The PRA review team confirmed this.

Therefore, the CDF re-estimation used a value of zero for sequences with control building ventilation failure, McGuire - Un.i.t_2 According to NRC Inspection Report No, 94-12, the Unit 28 500XV main step-up transformer all cooling fan motors have failed on numerous occasions, requiring control room operations personnel to respond to avoid a plant transient, Sixty of these motors have been replaced on ths Unit 2 EP

l l

g; ClMcCracken,yJ[

U-

...'L

.-jff 1

~

i,$ f?.hf} g f ".J

- Q A q Q.*l f fh f

l

. G:

l l

N.

~

y'

??$:

@id

  • M, transformer since 1987ifAltho' ugh; fah motbrs.metithe original design K;c specification, the inspection report tatedsthat{revisjonstothemotor fl' capacity resulted in an inibilityjto handlejeatil' ads encountered on the 28 o

main step-up transformer.

'r

..p /

The limited maintenance performed on.the transformer cooling coils was a contributing factor as well.g. The'lleensee' plans to replace these existing motors with higher capacity motors and to evaluate the other site trans formers. Additionally, the licensee plans to place increased emphasis on 4

main transformer preventive maintenance...The inspectors concluded that the motor replacements and increased emphasis on preventive maintenance should increase the reliability of the' main transformers.

Palo Verde Units 1. 2. and 3 dN lh' o..z.,.y-The licensee had postulated that a mechanism by which failure of the 125 VOC Bus A could occur would be loss of HVAC to the DC battery chargers, resulting in subsequent battery draindown, as well as loss of two channels of vital AC in the same room. Failure of the two channels of AC would result in a plant trip. This would be a long term event; the 125 VDC batteries A would need to be depleted before :the bus failed (about 2 hrs,). The licensee had been having problems with fire damper drops during testing of the fire suppression system which isolates HVAC from the DC battery chargers. The Itcensee believes the problem has been corrected. Aoditionally, temperature i

detectors have been installed in the DC equipment rooms, in addition to an j

alarm in the control room, 3

Results and Discussion

'(. ' f g %. c '.<

i The results of the Level-1 PRA (core damage frequency) analyses for the South l

Texas (Table 1), Grand Gulf (Table 2), and Diablo Canyon (Table 3) plants do not indicate high C0F contributions from dampers and fans, For South Texas, the total fan contribution is about 16, For Grand Gulf, common cause failure of the diesel generator room dampers represents about 9% of the total core i

damage frequency, whereas the contributions of motor-operated dampers in the I

standby servie water pump house falling to open on demand is about 6.5% for each of two dampers, These results are not unexpected, since the station blackout scenario and the standby service water system are highly important to risk at Grand Gulf, The chief risk-contributing initiating event in which dampers and fans are involved at Diablo Canyon is loss of switchgear ventilation, but the fractional C0F. contribution is only about 2%.

. e.cgF$4/ v.h ' 1 w 6.

The qualitative discussions of damper / fan risk contributing situations at the THI-1, McGuire, Unit 2 and Palo Verde plants indicate that these situations have received proper attention and have been largely remedied.

f e

Based on this small sample examination of (Level-1) risk contributions fcr dampers and fans, it does not appear that a generic concern regarding the risk impact of dampers and fans is justified.

It must be stated that, in the pRA l

_a

4 e

C. McCracken analyses, not all dampers and fans were modeled (some fire dampers, electrical j

cabinet fant were not).

It is possible that risk outlier situations involving dampers and fans exist at other plants, but, again, a generic concern does not seem warranted.

cc:

G. Hubbard W. Lefave i

i A. Dummer K. Dempsey l

2 l

i l

t l

l' l

1, Il i'

I

(

.

Retrieved from "https://kanterella.com/w/index.php?title=ML20091G141&oldid=4336846"