ML19332D139

LER 89-037-00:on 891026,determined That Instrumentation Used to Balance HPI Flow Through Four Injection Lines During Small Break LOCA Inadequate.Caused by Inadequate Review of B&W Guidelines.Flow Instrument installed.W/891127 Ltr

ML19332D139
Person / Time
Site:	Crystal River
Issue date:	11/27/1989
From:	Moffatt L, Ken Wilson FLORIDA POWER CORP.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
3F1189-19, LER-89-037, LER-89-37, NUDOCS 8911300065
Download: ML19332D139 (6)

Text

._ . , - - - _.

s ,

((-

4-h t j f ,, _ ,,

^ '

+

f.

' ^

~ - . -

C O R*0R AT IO N . ,

i r-

\

Novenber 27, 1989 ,

3F1189-19  ;

U. S. Nuclear Regulatory Ocmimission - [

! Attention: Document control Desk Washington, D. C. 20555

Subject:

Crystal River Unit 3 Docket No. 50-302 Operating License No. DPR-72  ;

Licensee Event Report No.89-037 l Daar Sir: ,

Enclosed is Licensee Event Report (IER)89-037 which is subnitted -

i in accordance with 10 CFR 50.73.

l: .i Should there be any questions, please contact this office.

l~

l -Very truly yours, ,

[ , Kenneth R. Wilson l' Manager, Nuclear Licensing WIR:Inag Enclosure I

t xc: Regional Administrator, Region II Senior Resident Inspector 8911300065 891127 p8 PDR ADOCK 05000302 S PDC Y

1 I >

i POST OFFICE BOX 219

• CRYSTAL RIVER, FLORIDA 32629-0219 * (904) 563-2943 /I A' Florida Progress Company ,

p .

NRC FORM 3E6 - U.S Nuct.LAA V t1ULATJAY COMMIEetON team . APPROVED DM8 NO. 3 M 01D4 tKPIRES C3oU tSilMATED SURDEN Pf R fitSPONSE TO COMPLY WTH THl$

"" *' ' fu"o"iN

'"'8MENTs'MO,Ro'i'O

' E#iJAYL' Tis' / rec O of

. LlCENSEE EVENT REPORT (LER) O

• MUJe"RV

"

• C7MJM "A"O^*,3e',ti?!%'#

O M NAO M N AND UDOET wA H Of C 03 FACILITY NAME ni DOCKti NUMSER (a) PAGE (3)

CRYSTAL RIVER UNIT.3 o 5 l 0 lo l o l3 g0 l2 1 lorl0 l5 '

ilPI.-Flow Instrumentation Accuracy Inadequate to Perform Flow Balancing During an llPI Lin, Break Due to Inadequate FPC Review & Inadequate Nuc. Steam Supply Sys. Vendor Cuideline s (VENT DAf t 451 LER NUMetR ISI REPORT DATE 476 OTHE R F ActLITIL5 INVOLVED (81 MONTH DAY YEAR YEAR "( (" k [tj,Q MONTH DAY YEAR F ACILIT. NAMES N/A DOCKE T NUMBt RtSi 0 16101010 1 I I Ig G 2g6 89 8g 9 Oj3 g 7 g0 1l 1 2g7 8g9 N/A 0 l5lo lo g o; l l OPERaf t.e0 THis REPORT 64 $UtMITTED PUR8UANT TO THE AkOUIRfMENTS OF 10 CFR { (Ch.ca on, o, snore ofthe follow <ng) 111)

"o05 to 1 n eoani n was so.n.nsio. n.rimi POntM

~

oemanum a naim so nanen.i n Tiwi

~

'IN' 01914 2o aoswinnei so. man > so.nunan at OTHE R is,.c,e r ,e Abstmrt n aaewinn.iu ^ X so.nunano so.n.nsu..unAi ~ Ye*I) * ** '" '" **"*""

no aosuninai X so.n une w so.n nsu iunsi so aswinn. so.nmunnao so nanan.>

LICEN885 CONT ACT FOR TH18 LER 11 1 NAME TELEPHONE NUM88R L. W. MOFFATT, NUCLEAR SAFETY SUPERVISOR ^ " ' ' ' '

9 l0 41 7 l915 l- 16 1418l6 COMPLEf t ONE LINE FOR E ACH COMPONENT F AILURE DESCRISED IN THl3 RtPORT (13)

' " C. R ORTa i N CAust SYlitM COMPONENT U t O NPR CAUS$ SYSTEM COMPONENT N p h, q l

1 i l I i l I I I I I I I I l' I I I I I I I l l 1 I i 1

$UPPLEMENTAL REPORT EXPECTED 414) MONTH DAY YEAR SUOM15$10N YES Ilf ver to". tete tXPLCilO SUB4Il5390N OA TE) NO l l l A.1a.CT,,_,.,...,..-.,,,,_......,..,n.,

I I

l On October 26, 1989, while operating in LODE 1 (IWER OPERATION) at 94% power, l Florida Power (FPC) detemined the accuracy of the High Pressure Injection .

(HPI) flow instrumentation was inadequate. 'Ihis instrumentation is used to l balance HPI flow through the four injection lines during a Small Break Loss of l Coolant Accident (SBLOCA) caused by an HPI line break to assure adequate HPI flow to the core. Initially, FPC began a power reduction to 50%. After additional calculation, nhk and Wilcox (B&W) informed FPC that operation at 50% power could not be justified and a plant shutdown was initiated.

This event was caused by an inadequate FPC review and followup in 1981 of B&W guidelines for SBIOCA caused by an HPI line break. This event was discovered during the investigation of a design basis issue associated with the Technical Specification Improvement prtgram. 'Ihis investigation revealed guidelines from B&W, which lacked supporting documentation, had been used to close the issue.

FPC installed more accurate, narrow range flow instrumentation. Engineering is evaluating alternatives to resolve the HPI line break issue for the long tem corrective action.

N..C Form 356 (64191-

(OneesesA - U.S. NUCL ELA (I t ULATOR Y COceMagaloN pp

, ~ l '. EXPIIES C/300 e

. LICENSEE EVENT REPORT (LER) f,U nf]Mo'd ER d Wo W;o"Jit].0pM y ,g,wt,wg TEXT CONTINUATION J"4"lo's!%WalMf!" !!'C' "f/!32 I;"I '

I5Nf!

1 P APE HWO Rt U TI'ON J d 4 0 IC 07 MANAGEMENT AND SUDGET, WASHINGTON,DC 30603.

FACILITY nae 84 H) DOCRET NUMOER (2) LER NUMetR ($1 PA05 (31 l

yygg St OVE N T o A L Mt v ts eph CRYSTAL RIVER UNIT 3 " " " " " "

0 l5 l 0 l 0 l 0 l 3 l0 l2 8l9 -

0 l3 l7 -

0l0 0 l2 0F 0 l5 text in . r. R w wrc r , aus4w nn EVINP IE9mIPPICN:

At 1600 on October 26, 1989, Florida Power (FFC) determined the accuracy of the High Pressure Injection (HPI) flow instrunentation [BQ,FI) was inadequate.

mis instrumentation is needed to balance HPI flow through the four injection  ;

lines during a small Break loss of Coolant Accident (SBIDCA) caused by an HPI '

line break to assure adequate HPI flow to the core.

Wis issue was identified during the hchnical Specification Improvement (TSI) program and the re-evaluation of B&W Preliminary Safety Concern (PSC) 24-79, ,

- HPI Line Pinch Break. (An HPI Line pinch break is a specific type of SBLOCA  !

which causes the broken HPI line to pinch and so automatically throttles HPI flow through the line.) At the time of discovery, Crystal River Unit 3 (CR-3) ,

was operating in MODE 1 (IVWER OPERATION) at approximately 94% power. At 1745 I on October 26, 1989, CR-3 began a power Ieduction to 50% power based on a B&W rammmrdation. During the power descension, at approximately 64% power, B&W informed FPC management that operation at 50% could not be supported. Wus, FPC declared both HPI trains inoperable based on the Technical Specification a definition of operability. 21s required entry into Technical Specification 3.0.3 which requires entry into MODE 5 (COLD SHUIDCTJN) within 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br />. At 2200, FPC increased the rate of power reduction ard entered an Unusual Event for a forced plant shutdown.

l During the MODE desccnsion, B&W review of previous analyses indicated Emergency i Core Cooling System limits of 10CFR50.46 could be maintained for the various  ;

HPI flows and decay heat levels associated with MODE 3 (HCTP STANDBY) . Wis '

conclusion was based on engineering judgment and the assumption that HPI flows would not be balanced by the operators should an HPI actuation occur during '

MODE 3 op s tion. Operator action to balance HPI flows under these conditions i may result in inadequate flow to the core. Werefore, the MODE descension was L terminated at 0830 on October 27, 1989 with Reactor Coolant System (RG) temperature at 289 degrees and pressure at 550 psi. Wese actions are described in FPC letter, 3F1089-24, dated October 27, 1989. j l Wis event is required to be reported because it was a condition outside the  ;

design basis per 10CFR50.73(a) (2) (ii) (B) and because it required a plant shutdown per 10CFR50.73 (a) (2) (1) (A) .

CAUSE:

We root cause of this event was an inadequate FPC review and followup in 1981 of B&W guidelines for a SBLOCA causal by an HPI line break. A contributing cause of this event was a B&W guideline and the lack of justification by B&W

!- for this guideline.

I NRC Form 306A 16491

ORM 38SA - U.S.EUCLEA3 E EIULATWY COMMISSIDN .,

~

.., .e EXTCES 4/30% .

. UCENSEE EVENT REPORT (LER) M.M?o'#"8!!G",o"'50iy .!!!',Y ERJ'! 7 TEXT CONTINUATION i,7,',"45,41 Mi niE 'Mf/c',"f,'!#

1. ' lSNf!

P APE RWO RE TO 9 J O IC OF MANAGEMENT AND BUDGET, WASHINGTON.DC 20603.

FACILITY NAME H) DOCKET NUtF3ER Q) LtR NUMSER ($1 PAOS tSi uaa " R' *er.

"'#3' CRYSTAL RIVER UNIT 3 o p lo lo j o l 3l0l 2 8l 9 __

0 l3 l7 .,_ 00 0 l3 or 0 l5 flXT (# mswo ausee 4 securest, asse esissecrW #4C Form J854 W Mi An HPI line break between the RCS nozzle and the first HPI check valve [B2,V]

was analyzed previously as not being a credible event, based on the physical configuration of CR-3. However, it turned out that regardless of physical configuration, the event had to be postulated. Berefore, the break became credible. Se root cause was B&W letter ESC-574 dated February 6,1981, which provided FPC with information on which we based the original position.

Snhwpst review of this letter during the TSI program followup revealed that the B&W information was not supported by any analysis as implied. B&W 1etter FPC-89-871 dated October.7, 1989, notified FPC of this finding thus requiring FFC to revisit this issue.

EVFWP EVAUATION:

Each HPI line and the normal makeup line is equipped with a flow instrument, as shown on the attached Figure. During a SBIDCA in the HPI line, instruments are needed to balance or isolate HPI flow. Balancing HPI flows is accomplished by throttling the injection valves [BQ,INV] to make the flow through each injection line nearly equal. We intent of balancing HPI flow is to maximize the HPI flow that actually gets to the core [RCT,AC). If an HPI line break exists, the broken line may have a much higher flow rate than in each of the unbroken lines. If the flow in the broken line is throttled, then more flow will go through each of the other lines to the RCS and less HPI water will be lost out of the broken line.

Rese instruments would also be used during a SBIDCA of the RCS cold leg to verify maximum flow to the core. For this accident, operators balance the flow to assure 70% of the operating HPI pump [BQ,P] flow is delivered to the core.

i Action based on the flow instruments is no*. reglired for this scenario because HPI valves are pre-throttled to assure adequate core coolirg.

Additional secondary functions of these instruments include:

1. -Determining if the injection valves have opened during an HPI actuation.

2. Informing operators of pump runaut conditions, in order to decrease flow or to align a recirculation flow path.

3. Diagnosing accidents. For example, they may be used alorg with other indications, to estimate the total leakage flow rate or to determine the location of a IDCA.

L me original HPI flow instruments have a range of 0-500 gpm cmd have an accuracy of +50/-100 gpn when used at low flow rates, 125 gpm. Rese instrument inaccuracies could mask an HPI line break, or could lead the operator to over-throttle the flow in the HPI lines. For example, during a SBIDCA with only one HPI pump available, the operator must balance flow through the four injection lines while maintaining a total flow above a l

l l

NRC Form 3GSA $8W

NR ORM SBSA U.6, NUCLE ARE AULaiORY COMMISSION 4.PPCOVED DMS NO. 31604104

., EKPIRES 4/30/92

'8ls".*Mh%'N U *

$u'/CWof;'Ho"uUt' 60fWs* IE"~'Js LICENSEE EVENT REPORT (LER)

TEXT CONTINUATION EN""RE n'R4 M2fGEfEUIe'UlMU"'fis M

-* 1 PAPERWO RE U TI'ON RJ (3 0 1 OF MANAGEMENT AND BUDGET.W ASHINGTON.DC 20603.

P ACILITY NAME til DOCKET NUM0ER (21 LER NUMeth (6) PAOf (31 CRYSTAL RIVER UNIT 3

" ~

" E" UE I 1

0 l6 [0 l0 l0 l3 l0 l2 8l9 _

0 l 3l7 Oj0 0l4 or 0 l5 )

TEXT fit asses mese e menwed, uns edennmar N#C / cam Jll54 U 117; specified minimum, 500 gpn, but belw the punp runaut flow of 545 gpn. We HPI flow instruments may indicate a total ccanbined flow of 500 gpn but actual flow is 300 gpn. Under these conditions, SBLOCA ard only one HPI punp, adequate core cooling and satisfaction of the 10CFR50.46 ECCS criteria can not be assured.

- he accuracy of the 0 - 500 gpn instruments above 200 gpu is considered adequate for their functions. Additionally, stop check valves (BQ,V) on the makeup /HPI punp discharge are set to prevent punp runout. h erefore, it is not likely the flow instrument inaccuracy would have caused punp damage.

2 determine the significance of the flow instrument inaccuracies, FPC evaluated the probability of the critical scenario actually occurring. We frequency of a SBIOCA in one of the HPI lines has been calculated to be 3.2E-07 per year. When considering the probability of only one HPI punp being availablo ard the probability of an operator error due to the flow instrument, the cambined probability of actually threatenirs plant safety is very small.

vumox nVE ACITCNS:

On October 30, 1989, FN installed narrow range HPI flow instruments, range O-200 gpu, which have -an accuracy of +/-8 gpn between 85 and 200 gpm. Se .

original flow instruments still feed indicators on the main control board, but operators have been instructed to use the original instruments only above 200 gpn. FPC has revised emergency and abnormal procedures to reflect these guidelines.

FPC evaluated the effects of HPI line pinch and guillotine breaks for CR-3,

such as HPI flow to the core and HPI pump runout, and did not identify any additional procedure or plant changes.

Be FPC Engineering procedures have been improved since 1981 to not allow statements made by verdors to be acx:epted at face value without either

t. obtaining or reviewing the analysis or backup calculations to assure proper interpretation and application to CR-3 plant specific configuration. FPC is evaluating alternatives for resolving this problem on a long term basis.

PREVION STMIIAR EVENTS:

i A review of previous reports did not identify any previous events which would have lead FPC to fird this problem.

l NIC Form 386A (6 89)

l!

l

- ey *fh i

!! lgg llII . --4 g 3g;9egeg

,sA~ 9ao '

j- .

f . t* - g t9i 0 i!5 42 8g s !

- r5mzg " zq 3miom4 5m3'  % $.d s o r [ .

..m5 4 oO E cps E gE 5n 8 g 0 e'8 7 :et M g-1!n1 l

l 4o n nI e j

{-g* g8jy E g s5.5!o<E 8 :b

?*

{goE gE5gE$ gj'y*i E$o2- .n s 3 I.2 b _

$3;1a jg  :  ;

=] -  ;- _

. 25>e e=n E e w y

l

<i 5Sj 1ig -!bc2

'"o" oo o o=

" o7- - '

c- I I7 I o v,

$ .l1 1,I g ils3 _

= _

p u

pW O- e k k n o a T .

m e _

%' t .

d e s e g

' r a -' ( y o . r k S r

o o t o n o '

a -

B vS T t

' 3 n

o 1

t c

e j -

n i

e l r -

T u

/ s s p p eu un ka P g7Vg s e

r 3 W P h

g

"

• H d

1 3 e

, & if <'gf f -

i i

n p

= 6^ i t

o 1

m l

c e S

' a j g

,s 2' s e n i

n ,

i d

6m O_ i a u p

k dm i

n e R g

./-

r

^ i u 2 -

- E- N o ko W

<. B r ,-

,- s.. . s o I . , , , . e ,

t c ',-

s.- 2 >

a e - <

R .s -s d s d s Sl g S I g

-_i oC oe oC oe TRCL T RC L

5 3 y g-