IR 05000443/1996003

Forwards Copy of Final ASP Analysis of Operational Condition Discovered at Seabrook Reported in LER 50-443/96-03.Results of Final Analysis Indicate That Event Precursor for 1996

ML20197F973
Person / Time
Site:	Seabrook
Issue date:	12/17/1997
From:	Casey Smith NRC (Affiliation Not Assigned)
To:	Feigenbaum T NORTH ATLANTIC ENERGY SERVICE CORP. (NAESCO)
References
NUDOCS 9712300342
Download: ML20197F973 (21)

Text

_ . . _ .

_ . . _ . , _ _ ,_ . . _ _ _ . . , . . .. . . _ , . _ .

- , ., ,,

, .c_

,

. .n y

'

', x

& + . . .

'

, , . .

" '

.

-

! DecemberJ17 L 1997; . >

m - ' -

.. . 4 .

'

u ;Mr. Tod Cr Feigenbaum" .

.

^

$ Executive Vice President and;

-

=

"'

' Chief Nuclear Officer

-

- -

/

,

. NorthAtlantic Energy Service C$mpany3 .

-

,.

'

. . clo Mr. Terry L?Harpster; J -

q

• ' - P.O; Box 3004 1 -

SeabrooklNH' 03874 -  :

,

. o

-

ISUBJECTU REVIEW OF PRELIMINARY ACCIDENT SEQUENCE PRECURSOR ANALYSIS  !!

~

• '

1 OF OPERATIONAL CONDITION AT SEABROOK - q

' '

-

t . .,

Dear _ Mr. Feigenbaum! c

" < .,

A^ ,

. pnr

, ,

j d.Rf/d Enclosed for your infomtation is a copy of the final A$cident Sequence b- Q :' analysis of the operational condition discovered at Seabrook that was reported in Licensee f ,!

Evant Report (LER) No. 50 443/96-003. This fin:.1 analysis (Enclosure 1) wa 9repared by our ) "1"'

QW% cont'r actor at the Oak Ridge National Laboratory (ORNL), based on review ano evaluation ofi h[

W N/f 4 our comments on the preliminary analy .a and commems received from the NRC staff and- b .

A p'il!

'

f jfroni our independent contractor, Sandia National Laboratories (SNL). Enclosurei contains = ' .

yy%f h f our fossioness to your specific comments. Our review of your comments ~ C ij Li ^

d nj[G} g/ contained analysii indicatein the this material which for accompanied the preliminary ./ ?g _ analysis

'

that event is a procursor 199 ~ ' j Md m.a :ekw f :1,adus

.o an ,

a, ,

e 4, r

P gg", )Q+N Please contecimo ateffort (301) 4151_427by you andifyour you staffhave anyand question, providing ! )regard

,

recognize and appreciate the expended in reviewing ' .

- @i pe A 1 j comments on the preliminary analysis.- 7 ,,

-

y '

g-y

. 1; y 4, .

-

~,

.

.,

-

j

.

Sincerely, M9h j[' .... %, s(Q ,

-[c

> U

.

H

.,

'Q _ Original signed by:

}

m MQMnmem e%.p l' -

Did C 2 [ [, [ Craig W. Smith, Project Manager

"

'

Project Directorate 1-3 : o

1. '

-

.

• Division of Reactor Projects - 1/ll; [ ,

,

.;

'

Office of Nuclear Reactor Regulation >

sc

,

& ,/ .

'q

+ .

Docket'No. 50-443

.

?

!s . Enclosures:LAs stated DISTR!BUTION ' C . Smith . My -

L

"- -

-

~ ,

.

. . .

JDochetFileang* ACRS" /. ,

cc w/Encisf See next page ~ PUBLICP ehl: j

"

' , PDI-3 RF R. Eaton- -

'

B. Bogeri T.- Clark

. ~,

-

MWhCl

'

-  ;OGC S. Mays, AEOD C " DOCUMENT NAMEF G ASMITH%SPSTDF6.SSA -

a To receive a copy of this document, Indicate in the box: "C" = Copy without

~

.

'

attachment / enclosure * "E" = Copy with attachment /enclopure "N" = No copy '

'

- -

.

OFFICE ~ PDI+3/M4 jf/ - 1 P0l*3/LA Q .f /)- M (AltPDI-3 = f% j l l ,

IIAIE - C9RI til - M -- gerk k VlV i REston ' t p; oAtt n b7/c 7, - / 1 Il r1/Gr1 - i' <t n I T I .  ;

O

'

Of flCIAL REcolt0 C0f,f - 1

,_

i -- - >

(

• • i 9712300342 971217 a P.DR ADOCK 05000443

.

m-

. .

. . ,

l

. .

,

gs..

.. ,

, , -

. _. CI .i . 2c ,-.

. e- ., . . ..

.

,, . __ .,. . . .._.;. _ _ _ . _ _

_ .m , _ .

. ___

_ ,__ _ _ . , . _ _ _ _ . , . _ . . _ _ . _ . _ _ _ . . i

,. a- ~

- IT. Feigenbaum _ . ,

' '

North Atlantic Energy Service Corporation . Seabrook Station, Unit N ;

,

e

.l

. oc: -

Lillian M. Cuoco, Es . Mr. Dan McElhinney

- Senior Nucisar Counsel _ _ - Federal Emergency Management Agency Northeast Utilities Service Compan Region i P.O. Box 270- J.W. McCormack Hartford, CT 06141-0270 Courthouse Building, Room 401

,

Boston, MA 0210g ,

Mr. Peter Brann-

, Assistant Attomey General _ Mr. Peter LaPorte, Director State House, Station #6 ATTN:- James Muckertelde __

<

Augusta, ME. 04333 , Massachusetts Emergency Management Agency;

Resident inspector - 400 Worcester Road j

. U.S. Nuclear Regulatory Commission P.O. Box 1496 --  !

i Seabroak Nuclear Power Station Framingham, MA 01701-0317- -

l

'

. P,0. Box 114g Seabrook, NH 03874 - Jeffrey Howard, Attomey General i G. Dana Bisbee, Deputy Attomey J Jane Spector .

General -

Federal Energy Regulatory Commission 33 Capitol Street  :

825 North Capital Street, Concord, NH 03301 '

'

Room 8105 _

Mr, Woodbury Fogg. Director

~

Washington, DC 20426

. New Hampshire Office of Emergency n Town of Exeter Management 10 Front Street State Office Park South Exeter, NH 03023 107 Pleasant Street

,

Concord, NH 03301

,

Mr. George L. lverson, Director ,

'

New Hampshire Office of Emergency Mr. Roy E. Hicock e Management Nuclear Training Manager State Office Park South Seabrook Station

-

107_ Pleasant Street North Atlantic Energy Service Cor Concord, NH 03301 P.O. Box 300 Seabrook, NH 03874 Regional Admini.trator, Regen 1 U.S.' Nuclear Regulatory Commission Mr. Terry L. Harpeter -

. 475 Allendale Road Director of Licensing Servmes -

King of Prussia, PA 19406 - Seabrook Station

> North Atlantic Energy Servic6 Cor Office of the Attomey General' i O. Box 300 lOne Ashburton Place Seabrook, NH 03874 20th Floor 4 Boston,MA 02108 Mr. D. M. Goebel

Vice President-Nuclear Oversight:

Board of Selectmen Northeast Utilities Service Company .

, Town of Amesbury , P.O. Box 270 -

. Town Hall Hartford CT 06141-0270 Amesbury, MA 01g13

--

.

!<

.= _ . - - .-- -

. ..

-w

-

- . 2-

!

cc:

M 'AL DiProfio Station Director Seabrook Station -

North Atlantic Energy Service Corporation

'

P.O. Box 300 i Seabrook, NH 03874

' Mr. Frank W. Getman, J I Great Bay Power Corporation l Cocheco Falls Millworks  !

100 Main Street, Suite 201 '

' Dover, NH 03820 Mr. B. D. Kenyon President - Nuclear Group Northeast Utlists Service Group P.O. Box 128 l Waterford, CT 06385 j

Mr. Drawbridge I

Executive Director Services & 1 Senior Site Offico, l North Atlantic Energy Service Cor Seabrock, NH 03874 Mr. J. K. Thayer Recovery Office, Nuclear Engineering and Support Northeast Utilities Service Company P.O. Box 128 Waterford, CT 06385 Mr. F. C. Rothen Vice President Nuclear Work Services ,

Northeast Utilities Service Company -

P.O. Box 128

. Waterford, CT 06385 -

Mr. A. M, Callendrello

.

Licensing Manager - Seabrook Station North Atlantic Energy Service Cor P.O. Box 300 -

Seabrook, NH 03874

.

m

/ -

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

-

.

- - , - - .

v

, - . - . - . - -

}

r; , + w f ,

)

.. .

, .i

1. : .

'l i* LER No. 443/96-003-

,

. . . , t

. LER No. 443/96-003 l  ;

>

Event Descriptioni Turbine driven emergency feedwater pump unavailable because of

~

- a ;.-d.:..ical seal failure Date of Event: May 21,1996" .;

Plant: Seabrook-l

)

k Licensee Comments ; Tj

'

^

' Reference: letter from Ted C. Feigenbaum, Executive Vice Prerident and Chief Nuclear Officer - North

~,

Atlantic Energy Service _ Corporation, to_U. S. Nuclear Regulatory Conuaission," Comments a t

q' on Prelinunary Accident Sequence Precursor Analysis of an Operational Condition at

-

'

,

Seabrook Station," October 3,199 i t

'

, Comment it EwntIkscription: Parap.ph 1, Sentence 5 should be replaced by:"The outboard seal gland was making contact with the top of the shan sleeve and the throttle bushing inside diameter."

Paragraph 1, add the following aRer Sentence 6: "The inboard seal gland had 0.007 i clearance between the top of the shan sleeve and the throttle buhing inside diameter."

'

Paragraph 2, Sentence I should read: ". . . the mechanical seals of the motor-driven . .

.

discovered the outboard mechanical seal to have a similar position, along . . ."

- Paragraph 2,.last Sentence should be replaced by: "The inspection revealed that the

.

burnishing identified on the outboard mechanical seal cf the MDEFW pump was consistent with normal rubbing experienced during pump startup. The system engineer concluded that the MDEFW pump was capable of performing it's design function based on the review of the as-found clevance data."

i ~

Paragraph 3, Sentence 3 should read: ". . . between the throttle bushing seal (secondary scal) _ ,

and the shan sleeve. There was never any contact with the primary seal."

Response 1: . - The proposed editorial changes enhance the description of the event. All of the suggested P ~

changes were made -

7 .-

- '/

m 1-a..,

~

- -

Enclosure 1 w-  ;

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

. - ., - .- - .., . ...,.-,s. . - - . , , - , , . , , ...- . , . ..

' *

. . .-

.

LER No. 443/96-003 Comment 2: Paragraph 2, Sentence 2 pfodchng Assumpnons] refers to the went on the MDEFW pump as being "similar" to that of the TDEFW pump. The wear on the MDEFW pump was due to normal rubbing experienced during pump starts and was not similar to that found on the TDEFW pump. However, it was concluded that the MDEFW pump seal was susceptible to the same mechanical rubbing as experieaced on the TDEFW pum Response 2: The adjustment to the common-cause failure probability was based on the description of the event given in the LER (LER No. 443/96-003 ROI) that stated that "The inspection [of the MDEFW pump mechanical seals] concluded that similar alignments of the mechanical seals were observed on this pump. This suggests that the pump was susceptible to the same mechanical rubbing as experienced on the turbine-driven pump." The LER Cause ofEvent section also suggests that the mechanical seal failure on the TDEFW pump was partially attributable to a design deficiency that "also applies to the motor driven EFW pump mechanical seals." Additionally, the mechanical seal failure was partially attributed to an inadequate procedure (applicable to both pumps) that had not been updated to incorporate lessons teamed from previous mechanical seal maintenanc The reference to "similar" wear was incorrectly used to represent the above common-cause issues. The statement that the wear on the MDEFW pump is "similar" to the wear on the TDEFW pump resulting in an increased potential fe: s common-cause failure has been removed. The second sentence in the second paragraph in the ModcAng Assumptsons section now reads as floweser, because the outboard mechanical seal on the MDEFW pump ms (1) poutened similar to that of the TDEFW pump, (2) subject to the same design de6ciency, and (3) subject to the same inadequate maintenance procedure that resulted in the TDEFW pump fails.s, the potential fer a common-cause EFW pump failure increase Comment 3: Sequence 39 (revised analysis sequence 41] from the Figure i " Dominant core damage sequence for LER No. 443/96-003" is the risk dominant sequence in the NRC analysis. This sequence is a station blackout with failure of the turbine-driven EFW pump. In this sequence, no credit is given for recovery of electric power, presumably due to the assumption that "45 min were available before steam generator dry out would occur, leading to core damage."

Seabrook has performed analyses of this speci0c analysis, using the MAAP code, with comparisons to the RELAP code resu".s. These codes show that, for this seque nce, the time to steam generator dry out is rbout 1.5 h, with core uncovery at 2.0 h, and care overheating at 2.2 h. As a result, subsatial time is available for electric power re.overy, in .ddition, the recovery curves have their highest slope from 0 to 2 h, indicating that a few additional minutes of recovery can be significant for recovery probability. The probability for failure to recover off site power in is approximately 0.25 and the probabiiity for failure to

._ .__ . . _ .- - _ _ _ . _ _ _ . - . . _ _ _ _ _ . . _ _ _ . _ _ _ . _

,,;

.;

, a

..

!

LER No. 443/96-003 1

)

- recover at least one of two EDGs in 1,5 h is 0.65; The overall nonrecovery probability for

. electric power is 0.25 x 0.65 = 0.1 This factor should k used in calculating tic: Station Blankout with EFW failure sequenc I

.

- - .

Response 3: - The assumption that "45 mm were available before a steam generator dry out would occur" was used to derive a failure probability for the operator action to realign the start up '

feedwater pump power supply breakers to the A EDO (basic event EFW XHE XM BRKR).

Based on this comment, the probability of this basic event occurring was resised from 0.16_ .

to 0.056 to account for 90 min until a steam generator dry out occurs. If power was not '!

- recovercJ within 30 min, a s_tation blackout and concurrent loss of EFW sequence was

_

- assumed to continue to core damage, as in the original modeling of the even However, because an additional 60 min might be available to the operators, the model was a revise new basic event that represents recovering ac power before a steam generator dry )

out under blackout conditions concurrent with a loss of EFW was added to the model (basic event OEP XHE NOREC SB). The probability that a source of electric power could be restored within given that.it was not recovered within 30 min (basic event OEP XHE NOREC SB) was calculated to be 0.29. Values for the probability of short term and long term electric power recove y for a LOOP following a postulated station blackout -

_

(SBO) were developed based on data distributions contained in NUREG 1032, Evaluation ofStation Blackout Accidents at Nuclear Power Plants (Re in the analysis). Based on this data distribution, the probability that a LOOP is recovered in the short team is 0.53 and '

is factored into the LOOP initiating event frequency. This results in an overall nonrecovery probability for electric power in the revised model of 0.53 x 0.29 = 0.15, which is approximately the same as the value presented in the licensee comn,ent abov The net effect of this model change is that the dommant sequence remains the same, but the increase in the CDP over the 3,875 h period is reduced from " to Comment 4: North Atlantic agrees with the analysis conclusion that during a LOOP event-without the MDEFW pump and TDEFW pump available, a heavy reliance is placed on operator action to maintain secondary cooling. North Atlantic licensed and'non licensed operators are ,

routinely trsined on shining the start up feedwater pump to the emergency feedwater alignment evolution. North Atlantic is conrulent that the operators would have been able to successfully complete this evolution during accident conditions utilizing the existing emergency procedure guidance and the specific training on this evoludon. However, for this

. event, the engineermg review of the MDEFW pump seal as-found data concluded that he MDEFW pump would have performed its required safety function, for the required mission tim .,p - , w ynw- w e - m-- - - ,_, w

.. - -. - - -. .-. . . - - - - - . .- -

a , <

, ,

.

LER No. 443/96-003, Response 4: The knowledge and training of the operators are recognized. The human error probabilities used in operations involving the alignment and use of the start up P.edwater pump are consistent with accepted error probabilities in operating other safety rdated equipment using -

existing emergency operating procedure The MDEFW pump was not modeled as failed. The independent failure probability of the >

MDEFW pump to start and run for the required mission time (p = for basic event EFW.MDP.FC 1B) was not altered from the base case. However, for the reasons discussed in the response to comment 2, the common-cause failure potential for the EFW pumps was increased from d to (basic event EFW-PMP CF EFW).

.

I Comment 5: North Atlantiv engineering personnel concluded that the TDEFW pump would not have been ab!c to perform its safety function for the required mission time (24h) because of the improper installation of the seal. This conclusion was based on discussions with the pump manufacturer and engineering judgement. The exact thne ht at the TDEFW pump became inoperable could not be conclusively determmed since the pump successfully completed two prior surveillance runs without any indications of problems related to the mechanical seal degradation. The system engineer evaluated the damage to the seal and consenatively determined that the pump had been inoperable since the mechanical seals were worked on during the November-December 1995 refueling outag Response 5: The following sentences were added to the first paragraph of the Analysis Results section to address the uncertainty over when the TDEFW pump actually became inoperable: "This is a conservative estimate because the TDEFW pump was satisfactorily tested twice (a total mn time of ~1-2 h) during the unavailability period (3,875 h). Therefore, the TDEFW pump likely would have operated for a limhed period (less than the mission time of 24 h) during the first part of the unavailability period, which would mitigate the calculated CDP "

l ~,

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

s J -i g j 'f

- :

s ,.4 or j

. .,

. . ,

.

--' LER No. 443/96 403 - ,

', _

'TD

-> .

>

- LER No. 443/96-093 q

"

Event Descriptioni Turbine driven emergency foodwater pump unavailable ' {

!

.

s because of a mechanicalseal failure i

4 Date of Event: May 21,1996

Plant: Seabrook )

q

,

.

~ Event Sunnaary '

~ *

Seabrook was at 100% power when pww,. .d were performmg a scheduled operating test on the turbine '

L driven emergency feedwater (TDEFW) pump.: The pump was manually tripped aAer sparks were observed

_

coming out ofits outboard mechanical sea The sparks were ultimately attributed to the improper installation

,

of the mechanical seal assembly during the previous refueling outage in November thah- 1995 (Ref,

,

'

1,~ 2)c This long term unavailability of the TDEFW pump (3,875 h) would have afFected the units' response

' to a loss of offsite power (LOOP) or a transient event. The estimated increase in the core damage probability :

l 4 i F (CDP) over the 5 month period for this event (i.e., the importance)is . The base probability ofcore

damage (the CDP) for the same period is .

Event Description Seabrook was at 100% power on May 21,19c6, when personnel started the TDEFW pump for its scheduled quarterly surveillance test. The operator tripped the pump locally during the test aAer sparks were observed i

. emanating from the outboard mechanical seal area of the pump. The mechanical seal was disassembled and

+ ;insputed The sparks were ht e resu l tfo w =h 4al interference _within the seal assembly. The outboard seal

,

gland was making contact with the top of the shaA sleeve and the throttle bushing inside diameter. The sparks were caused because the shan sleeve rubbed against the inside diameter of the throttle bushmg, causing a-1 0.005 in. souge in the shaA sleeve and the chipping of the throttle bushms. The inboard seal gland had a

" 0.007 in. clearance betv.oen the top of the shaA sleeve and the throttle bushing inside diameter, Licensee personnel concluded that because of the improper installation of the seal, the TDEFW pung would not have -

-

boon able to perform its safety function for the required mission time (24 h) since the November-December

, z l995 refueling outage -However, the exact time that the TDEFW pump became inoperable could not be conclusively detenmned since the pump had succesa%11y completed two prior surveillance runs without any hk= of problems related to the nachanical scel degradatio '

LAAer repairing the TDEFW pump. pr: sonnel inspected the mechanical seals of the motor driven emergency

. foodwater (MDEFW) pump and discovered the outboard mechamcal seal to have a similar position, along

= with the correspondag indications of r.d.e..ksi rubbing. The MDEFW pump outboard whaaical seal i

' gland had a 0.0035 in; clearance between the shaA sleeve and the top of the throttle bushing inside diamete The MDEFW pump' throttle bushing was not chipped like the throttle bushing ~was on the TDEFW pum ,

l q iThe inspection revealed that the burnishing identified on the outboard mechanical seal of the MDEFW pump 1 was consistent with normal rubbing experienced during pump startup. The system engineer concluded that - i

-

a

'L

- 1-

,

. _ l t-  !

" '

.

_ Enclosure -2;

__ . . _ . _ - _ -___ _ _ _ _

- .- - - - . - - - - . . - . - . _ - . - - . - . - - - . - - . - . . - - - - - - -

% ,

i -.; tu .

c- 1 a

'

j

- LER No.443/96-003

'

-_the MDEFW pump'was capable of performing its design function based on the review of the as found

-

,g,g,gg --

-

,

The design clearances and tolerances of the TDEFW pump's mechamcal seals were insufficient to prevent damage.during operation unless the_ installation technique used noncustomary methods (i.e., use of dial }

~ ~

'

indicators and fooler gauges). The design permitted the allowable tolerances to be greater than the available - l

{ clearance, Hence, the design did not preclude the interference between the throttle bushmg seal (secondary j scal) and the shaA sleeve. There was never any' contact with the prunary sea This ' design deficiency also - i r

applies to the MDEFW pump mechanical seals.- Contributing to this event was the failure to adequately -

incorporate previous knowledge regardmg seal installation into maintenance procedures or training As a result,' maintenance personnel were unaware of a prior seal failure (in 1987) or the need to take precision _i

,

1 measuromonts to verify the proper installation of the seal assembly, t

Additional Event-Related Information

.

l The emergency feedwater (EFW) system couists of two 100% capacity trains that feed a common discharge

'

header (Ref, 3).' One train uses the TDEFW pump, and the other train uses the MDEFW , sump. All four

,

steam generators can be fed by either EFW pump. The TDEFW aump is supplied steam from the A and B steam generators. The MDEFW pump is powered frc,m 4160V emergency bus E6 supported by the B emergency diesel generator (EDG).-

Seabrook also mamtains a start up foodwater pump with a cs;acity approximately equivalent to the combined capacity of both EFW pumps (Ref. 3). The start up foodwater pump can be staned from the control room,

'

except during a LOOP, Two normally closed motor operated valves (MOVs) must be opened to establish

_

feedwater flow. Following a LOOP, the ncona! power source to the start up feedwater pump is not supplied power from an emergency bus. Therefore, the normal breaker alignment for the start up feedwater pump must be altered from 4160V bus 4 to 4160V emergency bus E5 (emergency bus E5 is powered by the A

'

EDG).- The normal and alternate start up foodwater pump breakers arc key interidM reqmring one breaker 4 to be racked'out before the interlock key can be removed. The interlock key is required to rack in the

~

alternate source breaker (from bus E5) to the start up feedwater pum Modeling Assumptions-

.

1 Even though previous surveillance tests were successfully completed, the hcensee concluded that the TDEFW -

P pump would not have been able to perform its safety function for the required mission time (24 h) since the 4 L November-December 1995 refueling outage (Ref.1,.2). Hence, the TDEFW pump was-considered inoperable, and its failure probability was a4usted to 1.0 (TRUE) for a 3,875 h condition assessment. The

. 3,875 h condition assessment is based on the TDEFW pump being required from the end of the outage on December 9,1995, until the discovery of the mechanical seal failure on May 7 ' .1996. Two days (48 h) were

.. subtracted from the total number of hours that the TDEFW pump was unavaih. ole to account for a reactor trip

-

2 in January; i The licensee indicated that the MDEFW pump would have performed its safety function for the required

'

mission time. However, because the ou' card d=bl seal on the MDEFW pump was (1) positioned

.

~, ,

t -

_ _ __ ___ _ _ _ . . - _ _ _ . .__ _ . _ _ _ _ __ _

, _j

- _ - _ _ _ _ . __ _ ._ _ _ . _

- .

.,

4 e _

-

. -

- -" - LER No. 44356-003 similar to that of the TDEFW pump, (2) subject to the same desip deficiencv, and (3) subject to the same '

inadequate maintenance procedwe that resulted in the TDEFW pump failwe, the potential for a common -

cause EFWLpump failwe increased < The EFW common cause' factor- was developed based on data ' ,

"

i distributions for mixed-pump types conts,ined in INEL 94-0064, Common-Cause Falla Data Collection

,

mdAnalysis System (Ref. 4, Table 9 19: Alpha Factor Distribution Summary - All AFW Types Fail to Start,  ?

CCCG = 2, m = 0.0884).- Because ai is equivalent to the p factor of the multiple Oreek letter method used

-_in the Integrated Aeliability and _ Risk Analysis System (IRRAS)~models, the common cause failure L probability of the EFW system pumps (EFW PMP CF EFW) was a4usted from d to 8.84 x 10

-

based on the common cause failwe potentia The utility h'as m aM computer simulations of a station blackout with a conewrent failure of EFW at

_ Seabrook This simulation has shown that under these conditions, the time to steam generator dry out is about .

"

_90 min. As a result, substantial time is available for electric power recovery. This potential was modelen by

l the addition of a basic event (OEP XHE NOREC SB) that is considered under the OP SBO top event (O

2H) on the 1.OOP event tree (Fig.1). Top Event OP SBO is substituted for the OP 2H top event whenever '

.

emergency power and EFW are faile '

, )The Seabrook Individe! Plant Eramination (IPE) indicates that the start up feedwater pump is a backup

---

sowce of feedwater for the EFW system. To credit the use of the start up foodwater pump, a basic event was added to the IRRAS model for the Seabrook plant based on the IPE value for a failure of de start up foodwater pump to start and run (Ref. 5, Table 7.9 1) or a failure of the associated valves to open (basic event

. . EFW MDP FC.SFP). Because an operator is requand to open two normally closed MOVs to establish flow t _ from the start up feedwater system, another basic event was added to account for the failure of the operator

} to manipulate the required MOVs (EFW XHE XM SFP). Finally, during a LOOP, an operator must realip

,

the supply breaker for the start up feedwater pump to the A ED basic event was therefore added to

- represent the failure of an operator to complete this realignment (EFW XHE XM BRKR). This last basic event was based on the assumption that it would take an operator approximately 15 min, following a LOOP, to perform the activity and that approximately 90 min were available before a steam generator dry out would occur, leadmg to core damag lognormal distribution was used to calculate the failure probability for

.

EFW XHE XM BRK ' The operator nonrecovery probability for the EFW system during a LOOP (EFW XHE NOREC L) was - '

adjusted f:ca 0.26 to 0.80 because this action is notl ad~ahat from other operator actions. The operator -

. must f rst realip the supply breaker for the start up feedwater pump to the A EDG (EFW XHE-XM.BRKR).

! lf the operator fails to realign this breaker, the start up feedwater pump would not be available in a LOOP

>

scenario (LOOP sequence 17). Further,if the operator does indeed fail to realign this bmaker, it is more

likely that the operator will fail to recover the EFW system during a LOOP. Finally, during a station blackout

' (SBO), the only source of EFW is the TDEFW pump; therefore, with the TDEFW pump unavailable, there ,

is no opportunity to recover EFW. Based on this, the operator nonrecovery factor dwing a SBO 4 _(EFW XHE-NOREC-EP) was set to"TRUE"(recovery not possible).

L

- -_

k-m_ o --

_ n_ . .. .- _ - . . - _ . ~ . _ _ , - _ . - . ,_-..--.._,.__n ., _ - ._ -

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

'

l '. - 9 ,

'

.- c

.

'

LER No. 443/96 003 j t

i

'

Analysis Resuus - 7 The incre me in the CDP dwing a 3,875 h period for this event is t The norr.inal CDP for the same '  ;

'

paiod is 3,0 x lot This is a conservative estimate because the TDEFW pump was satisfactorily tested twice -  ;

. (a tr *si run time of ~l-2 h) dunna the unavailability period (3,875 h),' Therefore, the TDEFW pump likel woult' have operated for a limited penod (less than the mission time of 24 h) during the first part of the - '

una' ailability period, which would mitigate _the calculated CDP, The dominant core damage sequence for ~

1 this evnt (sequence 41 on Fig.1) involves l

'

'

.* L a postuisted LOOP,

- + - a successful reactor trip,  ;

'

. . a failure of emergency power, .

.

.~ a failure of emergency fwlwater, and - }

>- . a failure to restore electric power prior to steam gercrator dry ou LThis SBO sequence (sequence 41 on Fig.1) accounts for 56% of the total contribution to the irr:rease in the

<

CDP. The next most dommant sequence (sequence 17 on Fig.1) ontributes 22% to the total increase in the

- CDP. This sequence involves a LOOP with the success of emergency power, a failure of EFW, and a failurc ,

of feed and-blood decay heat remova An alternate study investigating the conditional core damage probability (CCDP) associated with the reactor ,

trip that occurred in January with the unavailable TDEFW pump was conducted. The TDEFW pump failure probability (EFW TDP-FC 1 A) was set to "TRUE"(failed). Using the same material assumptions as those

.

- made for the previous condition assessment, the CCDP for this initiating event is lot The dominant - ,

core damage sequence involves a failure to trip the reactor and a failure of the EFW syste Definitions and probabilities for selected basic events are shown in Table 1. The conditional probabilities associated with the highest probability r.equences are shown in Table 2,' Table 3 lists the sequence logic associated with the sequences listed in Table 2. Table 4 describes the system names associated with the

dominant sequences Minimal cut sets associated with the dominant sequences are shown in Tabic .

Acronyms

.

ATWS anticipated transient without scram-

-CCDP conditional core damage probability CDP oore damage probability ,

'EDO- emergency diesel generator

'

EFW - emergency feedwater system--

IPE integrated plant examination ~

IRRAS' Integrated Reliability and Risk Analysis System

LOOPJ - loss of ofTsite power MDEFW motor driven EFW (pump) .

MFW mai foodwater -_

'

MOV- motor operated valve :-

-

4

.

u r, e'-r-- p -[ e- e-rp we- -e ys - g s y v--

V-*Pr:: T' '=- Piw * * Fti C *w 'T'-7"7

. - . _ _ .

'

.

.-

-.- . . .

~ .

.1 - LER No. 44W96 403 PORY poww operated relief valve

'580' - station blackout TDEFW : twbine-driven EFW (pump)

References .

'

l o LER 443/96-003, Rev. O," Emergency Feedwater Pump Mechanical Seal Failure," June 21,1996

,

2.- -

LER 443/96 003, Re Emergency Foodwater Pump Mechanical Seal Failure," September 12,1996, 9

.

3. Seabrook Nuclear Station, Final Sqfety Analysis Repor . Marshall and 9- .uson, Common-Cause Failure Data Collectton and Analysis Setem, INEL 9410064, i December 199 $. Seabrook Nuclear Station,IndMdualPlant Examinatio .

.

5:

.

y y v- <-i,_-

,

,

. .

'.

, __

LER No 443/96-003

.

1[ _ ssss ssssssss ses eessss ssssss ssss ssssssssses

. -aa===~ee2*CC35SCESSENAARRh2R25#23RRkR20EW

$

lll {

- - -

flf {

-_ I t ii j

,-

_

ll l-

--

--

Ill!

lil ! -

ill -

illl :

Il !

--

• g i w --

l1 g 6 g

$

iIi !

Il ,

[ '

<

>

=

! l ,l W

. f

'

l ,l 1 Ea*

'll -

'

l Ik i l i I I

.

l Fig.1 Dominant core damage sequence for LER No. 443/96-00 l l

'

l

_ _ . - - - . ._____ _ _ . . . _ . _ _ . . . _ _ . _ _ . - _ . .

.

. .

l

.

.

.

LER No. 443/96 00,J

,

f

!

Table 1. Definitions and Probabilities for Selected Basic Events f r LER No. 443/96 003

'

Modified

~

Event - Base Current -

for this name Description probability probability Type event >

-

instistmg Ewnt LOOP t6E#$ $ 6 E406 No ICLOOP 1.6 E406 16 D006 No ID$rliR trutisting Ewnt-$ team Oenerator

,

Tute Rupture initiatmg Ewnt-$mell Lon of- 1.0 E406 1.0 D006 No ID$LOCA Coolant Accident (SLOCA)

IDTRANS trutseteg Ewmt-Transient S.) E@4 No (TRANS)

EfW Motor Dnwn Pump isils 3 9 E 003 3.9 D003 No ElW MDP 1C 10 ElW MDP IC $FP start op ie daeter Pump fails 21 E 002 NEW No Common <suw f ailure of Llw 3 8 E404 8 8 E 002 Yes ElW PMP CF.ElW Pumps (Escludes Start up ieedneter Pump)

EFSTDPIClA EIW Turtee-Drinn Purep rails 3 9 E 002 1.0 E+000 TRUE Yes El%XilDNokEC, Operator fails to Recowe Elw .6 E 001 No Operator fails to Recover Elw 3 4 E 001 1.0 E+000 TRUE Yes D%XilDNOREC EP Durma e Stetson Blackout Operator rails to Recour Elw 2 6 E 001 s0 D001 Yes El%X1iDNOREC L During a LOOP ElWXilt NEC Alw operator Fails to Recover EFW l.0E+000 1.0 E+000 No Dunng an A1WS EIW X1tDXM BRKP- Operator fails to Realign $ tert- $ 6 E 002 NEW No up Teodwater INmp Supply Breaker EfW XilDXM 5FP Operator rails to Open Sten up I 0 LOO 2 1.0 D002 NFW No feedwster Pump MOVs EPS DON CF ALL common Couw Failure of EDos 1.60403 16 D003 No EPS-DON 4C lA A EDO Fails 4J E 002 4 2 E 002 No >

EPS DON FC 1B B EDO l' ells .2 E 002 No

_

EPS XilE NOREC Operator Fails to Recout 8 0 D001 8 0 E 001 No Emergency Power llPI MDP FC 1B llPI Pump B rails 3,9 E 003 No

I l

.

l . - , , -

_ _ _ _ _ _ . . _ _ . _- . _._- _ . . - __ . _ .._ _ _ . _ _ _ _ . . . _ _ _

-

<

. .

.

. .

LER No. 443/96 003 I l

Table 1. Definitions and Probabilities for Selected Basic Events for LER No.443/96-003 [

!

Modified Event

- -

Base Current for this  ;

name Description probability probability Type event 4tPI X110 NOREC-L Opretor Tode to Recowr the t 4 E401 8 4 E401 No llPI Sysum Dures a LOOP ,

itPI Xilt XM l'U Opvolor rails to istiene reed- 10 E 002 1.0 E402 No end-Bleed itPI XilE XM lM Oprew Feita to istese Feed- .0 E 002 No end Bleed Dures LOOP MlW 5YS TRIP Mein Feedonier(MFW) System .0 E 001 No inpo MI'W XilE NOREC Oprow rolls to Recout MlW 34EMI 3 4 E 001 No OLP XilE NOREC $ll Oprow Fails to Recour 2.9 E-001 NEW No Electnc Power Before $leem Oswrotor Dry out

,

PPR SRV CC 1 Power Opreted Relaf Velve - 6.3 E403 63E403 No (IORV) i Fails to Opn on Demand PPR.$RV CC 2 IORV 2 Fails to Opn on 6.3 E403 6.3 E403 No Demand RP$ NONREC Honrecowreble Reecw 2.0 E40$ No Protection $3 swm Teilurn RPS REC Recowreble RCS reilure: $ 4.0 E-00$ No RPS XIIE XM $ CRAM Oprotor Fails to Manually Trip .0 E402 No the Reactor

$

8-

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

.-. - - _ - . - - _ . . _

- _ - . -. . . . . . _ .. . _ _ - - .

.

, ,

d

- LER No. 443/96-003 i

Table 2. Sequence Conditional Probabilities for LER No. 443/96 003 Conditional Event tree Sequence core damage Core damage importance Percent  ;

name number " probability probability (CCDP. CDP) contribution *

'

(CCDP) (CDP)

LOOP 41 .4 E 007 $ $$.6 LOOP 17 .5 E 008 .5 TRANS 21 8 M 6 .0 E 006 1 TRANS 20 t .3 E 006 $.0 LOOP 40 .6 E 008 .3 Total (all sequences) .0 E 005 4.6 E-005 ',

-

'

• Perset contnbution to the total importanc t

l

.. -

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

- . . . -- - - - .- - _ . - . . ~ - - - - - - - - . - . - . -

, I

,

< .

- . i r

, LER No. 443M.003 i

,

.t

-

Table 3. Sequence Logic for Dominant Sequences for LER No. 443N 003

>

Event tree name - Sequence Logic  ;

'

number t

/RT.L EP ,EFW.L.EP OP.SBO

>

LOOP 41 ,

LOOP 17 /RT.L./EP, EFW.L. FAB L l

'

TRANS 21 8 RT,/RCSPRESS, EFW.ATWS

-

TRANS 20 /RT, EFW, MFW, FAB LOOP 40 /RT.L. EP, EFW.L.EP, /OP.SBO, F&B l

Table 4. System Names for LER No. 443N 003  :

?

System name Logic

< EFW No or lasumcient EFW Flow EFW.ATWS No or Insumcient EFW Flow During an ATWS i EF No or insumcient EFW Flow Dwing a LOOP EFW.L.EP No or lasumcient EFW Flow Dwing a Station Blackout l EP Failwe of Both Trains of Emergency Power F&B Failwe to Provide Feed and Bleed Cooling -

FA Failwe to Provide Food and Bleed Cooling During LOOP

- MF .

Failwe of the MFW System OP.SBO Operator Fails to Restore AC Power Before Steam Generator Dry out Durms a Station Blackout '

RCSPRESS Failure to Limit Reactor Coolant System Pressure to

<3200 PSI ,

RT Reactor Fails to Trip Dunng Transient

<

R Reactor Fails to Trip Dwing LOOP r

i

,

- a ,

vv- g T w 'e v y- -v---rr -w n--- y N .r.-e r=-- +- Wr + e a rem -sc- m-v-eNerv - ++ ww~ a,=-w.- *=ww---- =-w "rr **--w*=

__ ._ ___ _ _ _

-

l

-

.

. -

.

'

-

LER No. 443/96 001 ,

• -  !

i f

Table 5. Conditional Cut Sets for Hist.er Probability Sequences for LER No. 443/95 4

-

,.n . . .

'

Cut set Percent ~ I

number contribution CCDP' Cut set:

LOOP Sequence 41 .

> <

< / W,w.,, .... y iNIO/E I $ .4 E 005 EPs dan rC l A. EPs-DON rC.lB. Ers-XHE NORE *

D%TDP rc 1A. UWXilDNOREC EP.OEP XHLNOREC.$B 2 4 .2 E 00$ EPs DON <r ALL EPS X11E NOREC. El%TTSrc I EfW XHE NOREC-tP.OEP XIIE-NOREC 5B

+

LOOP Sequence 17 . El%TDP rc I A. U%PMP Cr DW, Eiv X)lDXM BRK ) 1 Ef%XHE NOREC LllPI XilDXM rBL  :-

EPs DON rC.l A./EPs-dan rC lB, ErW TDP rC l .5 Ef%PMP-CF DW If%XilE NOREC L.HPI XilDXM-rBL DW TDP rC l A, DV PMP CT ETW, ErW XHDXM BRK .3 DW XilE NOREC.L PPR SRV CC 1

,

4 .3 E 007 u%TDP rc l A. LIV PMP CF EN. ErW XilDXM BRK D%X)lE NOREC L PPR $RV CC 2 EPS-DON rC IA.EPS-DON rC IB UW TDP TC l A, 5 .0 E-007 El%XilDXM BRKR.UWXilLNOREC L IIPI XilDXM rBL EPS DON ILi A, EPS-DON-rc IB, DV TDP rC 1 .0 EfW PMP CF ElW. DW X1tDNOREC L PPR SRV CC 2 .9 E 007 EPS DON rc IA,SPS DON rC lB,U%TDP rC l ErW PMP Cr ErW.Elv XilE NOREC L PPR SRV C .9 E 007 DV TDP-rC l A, ti%PMP<r EfW. EIWMDP rC.$rP, 8 Efw XilE NOREC L IIPI X11E XM TBL 9 .8 E 007 EPs tmN rC I A.EPS-DON rC-ID.El%TDP rC I ElW XIIDXM-BRKR. ErW XilE NOREC-L PPR SRV CC 1 EPS-DON-rC IA. EPS DOH rc IB, EFW TDP-rC 1 .8 EFW X11LXM BRKR. ErW XHLNOREC-L PPR SRV CC 2 .1 E 007 D%TDP rC lA, ErW PMP CT DW, ErW MDP rC.$rP,

El%XllDNOREC L, PPR SRV CC 1 .1 E 007 UwTDP rC i A,El%PMP<r ETW ErW MDP rc $rP,

EIWXllLNOREC L PPR $RV CC 2 31 31 E 007 EPS DON rC IA.EPS DON rc 18 EFW TDP FC 1A,

EFW PMP<r EFW, El%XilE-NOREC L HPI-MDP rC 1 Hrl XllE NOREC L

-

- - . . - . - - ,. ....- --- -.- -.--- - - - - - - - . -. -. .. . - . . , . - - . - .

_ _ _ _ . _ .~ _ _ . _ . . _.. __ _ _ _ _ _ _ _ _ . - _ _ . . __ _ . - -_ __ __

"

. , ,

,

LER No. 443/96-003 Table 5. Conditional Cut Sets for Higher Probability Sequences for LER No. 443/96-003 l

-

Cut set Percent

-

- number contribution CCDP' Cut sets" TRANS Sequence 218 5.1E006 y . 4 7 .6 E 006 RPS NONREC,EF%TDP.fC.IA El%PMP CF EfW,

El%XilE NEC ATW 2 1 .6 E 007 RPS NONREC, El%2P FC 1 A, Ei%MDP-TC Si?,

EfW X110-NEC A1W

-

7,9 RPS NONREC, Ef%TDP FC IA.El%XIIE XM SrP,

ElW XHE NEC ATW

-

4 .6 E 007 RPS NONREC, Ef%TDP f C I A, EfW MDP IC lB, El%X)lt NEC ATW RPS REC, RPS XilE XM SCRAM, EIWTDP l'C l A, 5 El%PMP CF ElW, EFW X110 NEC ATW a

TRANS Sequence 20 .7 E 007 EN TDP-IC I A, LIW PMP CF Elv. El%MDP l'C 5FP, 1 2 EiWXilE-NOREC, MiWSYS TRIP, Mi%X11E NOREC, itPl X)lE XM rB LIW TDP rC 1 A, EfW PMP-CF EIW, El%MDP fC SFP, 2- 1 ErW XHE NOREC, MlW SYS TRIP, MIWXIIE NOREC, PPR SRV CC 2 3 1 .2 E 007 EN TDP.rC 1A,EfW PMP CF EN EN MDP-TC STP, EfW XilE NOREC, Mi%$YS TRIP, Mi%XI{E.NOREC, PPR SRV-CC l 4 1 .2 E 007 EN TDP rc IA, El%PMP CF EfW EfWXHE XM SFP, El%XillMOREC, MfWSYS TRIP, MfWXilE NOREC, llPI XilE XM FB .0 E 007 EN TDP FC l A, EFW PMP CF EFW, EFW XHE XM-SFP,

$

EfW XllE NOREC, Mi%$YS TRIP, MfWXilE-NOREC, PPR SRV CC 2 6 .0 E 007 EN TDr rC 1 A, EFw PMP Cr ETW, Ef%Xilt XM SFP, EfW X1tE NOREC, Mfw SYS TRIP, Mi%XilE NOREC, PPR SRV CC 1

-

7 .0 E 008 Ei%TDP FC IA, EFW MDP FC lB,Erw MDP-FC SFP, El%X)tE NOREC, Mi%$YS TRIP, MFW XIIE-NOREC, llPI XHE XM FB r

l i

..

S

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

-

.

• - .

e

.

• LER W. 443/96-003 1

.

- Table 5. Conditional Cut Sets for Higher Prebbility Sequences for LER No. 443/96-003

'

._

'

Cut set Percent number contribution -

CCDP' Cut sets *

"

LOOP Sequence 40 EPS DON FC 1A,EPS DON-FC.IB EPS X}lE NOREC, 1 2 .7 E 007 EFW.TDP FC 1A,EFW X}{E-NOREC EP,llPI X1tE-XM FBL EPS DON 4F-ALL EPS.Xilt NOREC. El%TDP FC 1 A, 2 2 El%XilE NOREC EP,itPI XilE XM FBL EPS-DON FC 1 A. EPS DON FC 1B, EPS Xilt. NOREC, .

3 l El%TDP FC l A EIV XilE NOREC EP,PPR SRV4C 1 EPS DON FC lA.t/$ DON-FC lH EPS XI(E NOREC, 4 1 Er%TDP FC lA,DW XitE NOREC EP, PPR SRV CC 2 EPS DONCF ALL EPS XI{E NOREC, EFW TDP-FC 1 .2 U%XiiE-NOREC EP, PPR SRV CC 1 LPS-DON CF ALL EPS XIfE NOREC, El%TDP IC 1 A, 6 1 ElW X1tE NOREC EP, PPR SRV CC 2 Total (all sequences) *The CCDP is determined by multiplymg the probabihty that the portion of the sequence that makes the precursor visible (e g., the optem with a failure is d.$ mended) mill occur durug the duration of the ewat by the probabihtiu of the remaining busc ewnts in the minim cut set. This can be approsimeted by 1 e , where p is determmed by multiplying the exlweted numter of initiators that occur during the duration of the ewn by the probabihtees of the basic events in that mmimal cui set The espected number of initiators is sinn by At, where A is the frequency of the initietmg eunt (giwa on a per hour basis), and t is the duration time of the event (3,875 h) Dis opprosimetion is conservatin for procurnors made visible by the initiating ewat. The frequencies of interest for this ennt are L - / dum, = 8 6 x 10 % The importance is deiermswd by subtractmg the CDP for the same period but with plant equipn ent assumed to be operstmg nominall "Besic events Ei%1DP FC lA and EiWXilE HOREC EP are type TRUE eunts These ene of ennts are et normally included in the output of the fault tree reduction process but have been added to aid in understandmg the sequences to potential cote demere sesociated with the event-

_ . . _ __ _ _ _ _ - _ _ _ . _ ~ _ . - _ ~ - - . _ _ . _ _ . _ _ _.