ML19332C792

LER 88-035-01:on 881216,determined That Plant Operated at Estimated Actual Power in Excess of 102% from 831223-840104. Caused by Mfg Defect of Feedwater Flow Venturi Tap.Venturi Repaired & Subsequently replaced.W/891120 Ltr

ML19332C792
Person / Time
Site:	San Onofre
Issue date:	11/20/1989
From:	Morgan H SOUTHERN CALIFORNIA EDISON CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
LER-88-035, LER-88-35, NUDOCS 8911290001
Download: ML19332C792 (8)

Text

. - v: - - - -

{g , ,

3 <

't q:' i Rf,[

.;;:

.5 ...

[;

. ., y

[,

f[8 _

Southern California Edison Company . ,

, SAN ONOFFIE NUCLEAR GENERATING STAttON P. O. Box 128

- GAN CLEMLNT E, CALIFORNIA 92672 HfE; MORGAN

'f TELEPHONE nunou unaoEn November 20, 1989 avi4> see enai -

.: s

?

U. S. Nuclear Regulatory Commission! '

Do'cument Control-Desk

. . Washington' D.C. '20555-

Subject:

Docket No.!LO 361

-Supplemental Report 4 Licensee Event Report No.88-035, Revision 1 L

San Onofre Nuclear. Generating _ Station, Unit 2

' Pursuant.to 10 CFR'50.73(d), this submittal provides the supplemental report

-for an occurrence; involving feedwater' flow venturis.- Neither the health and safety of plant personnel or the public was affected by this occurrence.

1.

If you require' any additional information, please so adt'ise.

Sincerely, kD ~'

y , .

+

1 o

• 'ri

Enclosure:

LER No.88-035, Revision 1

p. u ,

-cc: 'C.'W. Caldwell'(USNRC Senior Resident Inspector, Units 1, 2 and 3)

.1 B. Martin (Regional Administrator, USNRC Region V) .

. Institute of Nuclear Power Operations (INP0) l 891'1290001 891120 PDR.

S ADOCK 05000361 )O L PDC l

'h,

,R

, ,' LICENSEE EVENT REPORT (LER) facility Name (1) Docket Number (2) Psoe L3)

A SAN ONOFRE NUCLEAR GENERATING STATION. UNIT 2 Of 51 Of 01 01 31 61 1 '1lof 0l7 Ittle (4)

PLANT OPERATION ABOVE 102% ESilMATED ACTUAL POWER DUE TO MANUFACTURING DEFECT OF FEEDWATER FLOW VENTURI FWFMT BATF di f ra rmare (6) DFphl'T hATU (7) OTMFD Fatti tf f t f MVnf VFh (R) 5 at on '*#il * "**** O'* "'I Month Day Year Year //f ff , Month Day Year O! 51 01 01 01 l'l NONE 112 213 813 '818 0l3I5 011- 1 i 81 9 of 51 01 Of of ' i 1 TH E REPORT IS 1;UBM.TTtD PURSUAN" TO 'HE REQUIREMLNTS Of 1DCFR OPERATING (Chock one or more of the fotlow no) (11)

MODE (9)- 1 20.402(b) 20.405cc) 50.73(a)(z)(1y) 73.71(b)

POWER LEVEL' Z 20.405(a)(1)(1) 20.405(a)(1)(II)

Z 50.36(c)(1) 50.36(c)(2)

Z 50.73(a)(2)(v)l) 50.73(a)(2)(vi Z 73.71(c)

Other (Specify in (10) i1 01 0 Z 20.405(a)(1)(lit) 50.73(a)(2)(1) Z 50.73(a)(2)(vill)(A) Abstract below and

//////////i////////////// _ 20.405(a)(1)(iv) 20.405(a)(1)(v)

X 50.73(a)(2)(II) 50.73(a)(2)(li t )

_ 50.73(a)(2)(vill)(B) 50.73(a)(2)(x) in text)

///////////////////////// _ _ _.

/$/ b$/b LICENSEE CONTACT FOR THIS LER (12)

TELEPHONE NUMBER N8"' AREA CODE ,

H. E. Moraan. Station Mananer 7l 114 316l81 l6121411 l COMPLETE ONE LINE FOR EACH COMPONENT FAILURE DESCRIBED IN THIS REPORT (13)

/

CAUSE SYSTEM COMPONENT R gE CAUSE SYSTLa' COMPONENT AC- R E l B 5lJ Fl El l XI 91 91 9 Y /////// l l l' I l i I ////// ,

I l l 1 l I l /////// I I 1 l l I I //////

SUPPLEMENTAL REPORT EXPECTED (14) Expected Month Day Year Submission Yes'(If ves. complete EXPECTED SUBMISSit'N DATE) lkklNO Date (15) l l l ABUTRACI (Limit to 1400 spaces, i.e., approximately fifteen single space typewritten lines) (16) l l

Oh December 16,-1988, with Unit 2 at 100% power, after having developed a new U methodology for approximating reactor power and applying it to plant historical l

' data, it was determined.that Unit'2 had operated at an estimated actual power in l excess of 102% from December 23, 1983 to January 4, 1984. The maximum power for '

this period was estimated to be approximately 103%. For many of the design basis accidents discussed in the Final Safety Analysis Report (FSAR), an initial power level of 103% in conjunction with the most limiting set of other initial conditions could result in consequences more severe than those determined in the FSAR, assuming no operator action for the times specified in the analysis.

- The cause of plant operation at an estimated actual power greater than 102% was  ;

'_l ~a manufacturing defect of one-of the feedwater flow venturi pressure taps, which

~ resulted in a decreased differential pressure sensed across the venturi. This

' condition resulted in a decrease in indicated feedwater flow (and, therefore, iindicated reactor power) relative to actual feedwater flow (actual power). The operation of the plant at 100% indicated power resulted in estimated actual power exceeding 102%. The flow venturi was repaired in 1984 and subsequently replaced with a newer design in 1985.

No other instances of operation above 102% estimated actual power were identified from the review of Unit 2 and Unit 3 historical power data. The new methodology will continue to be applied to current plant operational data on a l routine basis to reduce the probability of operation above 100% actual power.

1

-r - - _ _ _. _ _ . _ . . . - _ - - - _ _ _ - _ . _ -

m, s s a 4,

,,; 9

' ~

f, .

LICENSEE EVENT' REPORT.(LER)LTEXT CONTINUATION SAN ONOFRE NUCLEAR GENERATION STATION ~

DOCKET NUMBER- LER NUMBER PAGE

.? UNIT 2- 05000361 88-035-01: ~ 2 0F m

' Pl ant: San.Onofre Nuclear Generating Station i

< Unit:,Two~

Reactor. Vendor: Combustion Engineering Event Date:. 12/23/83 k Al- CONDITIONS AT TIME OF:THE EVENT:

Mode: 1, Power.0peration

[U;

^

- y

. . . Bt BACKGROUND INFORMATION:

. 1. Design Basis Maximum Plant Power I

1 '

' Units 2/3' Final Safety Analysis Report (FSAR) section .15.0.3.2,

" Initial Conditions", provides the range of values for each of.the principal process' variables- that were considered in all the accident analyses. The maximum initialicore power for_ these analyses is a

L 102%, thus defining the design basis maximum power of the plant ~

. /j

y. .

q T , 2. Calorimetric Power Calculations: g

- During operation, plant power is normally determined:by secondary  !

calorimetric calculation, which utilizes various steam plant. O parameters'as inputs. -Feedwater flow is one of the parameters d utilized as~ input to this calculation. This parameter is determined by measuring the-differential pressure across -feedwater flow venturis (JB][FE] in the main feedwater lines; feedwater flow l venturis FE1111- and FE1121 provide indication of feedgater flow to j n the steam generators [SG) and input- to the secondary calorimetric j power calculation.

1

(

Plant power can also be determined by primary calor ~imetric l calculation, which utilizes reactor coolant system [AB]' temperatures )

and mass flow rate as inputs.' This calculation is somewhat less

. accurate than the secondary calorimetric, but is determined .;

L independent of steam plant parameter values. t

'C. DESCRIPTION.0F THE EVENT: i

. I '. Event-p 3

~*

During Unit 2 Cycle 1 operation in 1983, a decreasing trend in the i primary calorimetric calculated power relative to the secondary calorimetric calculated power with increased core life was noted to

j. occur. Feedwater flow venturi fouling was concluded to have been the cause of the trend.

l l'

, -  ;--. - . _ _ _ . _ _ _ _ __ _ ____ _ ___ _ _ ____ _ _ _.

p ,

s LICENSEE EVENT REPORT (LER) TEXT. CONTINUATION- ,

2 SAN ONOFRE NUCLEAR GENERATION STATION DOCKEl NUMBER LER NUMBER PAGE l UNIT 2 05000361 88-035-01 3 0F 7

+

~

The fouling was caused by oxygen in the condensate-[SD).which caused l an electrochemical- reaction in which copper from the. feedwater 1 heater.[SM)[HX) tubes was transported from the heaters and deposited, 1 as a scale on the walls of the venturis, which were constructed of

'l carbon steel with a stainless steel cladding. This scale resulted 1 in' a higher differential pressure being sensed across the venturi,.

which resulted in an increase in indicated feedwater flow-(and, l therefore, indicated reactor power) relative to actual feedwater flow (actual power). Since the plant was. operated at 100% indicated ,

power, actual power was decreased as the venturi fouling increased. I l

In November 1983, after approximately 140 effective full power days '

(EFPD) of operation, Unit 2 was shutdown, during which the feedwater flow venturis were cleaned. . When full power operation was resumed following.this outage, the ratio of primary calorimetric power to secondary calorimetric power was initially approximately 3% larger-than its value prior to the shutdown. This was attributed to the. I venturi cleaning. With continued plant operation, the venturis  !

"refouled", and this ratio again decreased.

l Licensee Event Report (LER)88-028 (Docket No. 50-361) describes L sone factors (including feedwater flow venturi fouling) associated with the secondary calorimetric power calculation wnich.resulted in.

the operation of Unit 2 at an estimated actual power slightly in excess of 100%. As one corrective action, the Plant Performance

- Monitoring -Program was enhanced to routinely monitor plant parameters related to the determination of reactor power to further reduce the probability of operation above 100% power. Included in the enhancement of the program was the development of a new u methodology to approximate reactor power based upon calculations

utilizing empirical data from plant. operation independent of l

secondary calorimetric inputs. This methodology is acknowledged to contain a larger inherent error than the secondary calorimetric power calculation, and it is intended to be used only as a check

against secondary calorimetric power. The methodology was validated by' applying it to plant historical data.

l On December 16, 1988, with Unit 2 at 100% power, after applying the aforementioned new methodology for approximating reactor power to plant historical data, a preliminary determination indicated that Unit 2 may have operated at an estimated actual power in excess of i 102% during a portion of the time periods of October 1983 through I

January 1984 and April through June, 1987 (as reported in a letter ',

to the NRC dated 12/19/88). Upon further evaluation utilizing additional data from these time periods, however, it was concluded that Unit 2 had probably operated at an estimated actual power in excess of 102% only during the time period from December 23, 1983 to L January 4, 1984.

^1 , - - - ,-w ,

[ ~

,  ;

LICENSEE EVENT REPORTf(LER) TEXT _ CONTINUATION

-SAN ONOFRE NUCLEAR GENERATION STATION DOCKET NUMBER- LER NUMBER PAGE UNIT 2 05000361 88-035-01 '4 0F'7' ,

Unit 2 was never continually operated at an indicated: power in excess of 100%. . Operation above'102% actual power constitutes a

. condition outside the' design basis of-the plant. The. maximum power at.which the plant may have been operated during this period'was 1

, estimated to_be approximately 103%.

2. _ Inoperable Structures, Systems or Components.that Contributed to the

- Event:  ;

Refer to Section D, Cause.

3. Sequence-of Events:

DATI ACTION ,

-9/83 Feedwater flow venturi fouling was noted to occur.

11/83' Following approximately 140 EFPD, Unit 2 was shutdown, during which the feedwater flow venturis were cleaned.

12/83 - Follnwing the outage, full power operation resumed. .

Primary-to-secondary calorimetric ratio was- i approximately 3% higher than prior to the outage, j 6/84 During a plant shutdown, feedwater flow venturi FE1111

- was' determined to have a' defective low pressure _ tap.

The venturi was repaired prior to returning to power o operation.

10/84 Evaluation performed which concluded that 102% power was not exceeded during Unit 2 Cycle 1 operation.

12/88-1/89 After development of a new methodology for determining' i reactor power and applying it to plant historical data, a determination was made that Unit 2 had probably operated at an estimated actual power in excess of 102%

during the time period from December 23, 1983 to January 4, 1984. ,

p 4. Method of Discovery:

Refer to Section C.1, Event. l

5. Personnel Actions and Analysis of Actions:

Not applicable.

6. Safety System Responses:

Not applicable.

l l:

w= g- y- v 9 mwm -+m e y 4-w .. n

p, ,

LICENSEE EVENT REPORT =(LER) TEXT CONTINUATION o SAN ONOFRE NUCLEAR GENERATION STATION DOCKET NUMBER LER NUMBER PAGE UNIT 2 05000361 88-035-01 5 0F 7

-D.- CAUSE OF THE EVENT:-

4 l '. Immediate Cause:

The cause of plant operation at an estimated actual power greater

.than 102% from December 23,-1983 to January 4, 1984 was determined'

'to have' been a hole in the. low pressure tap of the carbon steel feedwater flow venturi FE1111. This defect resulted in

.decalibration of ~ the venturi with changing feedwater- temperature.

An increased feedwater temperature (such as .for higher power operation) resulted in an erroneously low differential pressure sensed across the venturi. This' condition resulted.in a decreased indicated feedwater-flow (and therefore indicated reactor power) relative to actual- feedwater flow (actual power). When indicated

, power was. increased to 100%, estimated actual power was increased to greater than 102%.

p 2. Root Cause: j o

The defective venturi low pressure tap was determined to be caused by a manufacturing defect. The configuration of the tap's piping internal to the venturi differed from the configuration specified on the design drawing. This deficient configuration contained a weld which was improperly done, causing the hole in the low pressure tap-of the venturi.

3. Contributing Cause:

Following the outage in November-December 1983, secondary power was approximately 3% greater than prior to the outage. This difference was attributed to cleaning (and, therefore, defouling) the feedwater flow venturis during the outage. The magnitude of the venturi fouling effect (resulting in a higher indicated power relative to- ,

actual power), however, was such that the defective venturi low ,

- pressure tap (resulting in a lower indicated power relative to- 1 actual power) could not have been identified. ')

i E. CORRECTIVE ACTIONS:

1. Corrective Actions Taken:

a. Feedwater flow venturi FE1111 was repaired following identification of the failure described above. Furthermore, during the first refueling outage (completed in February 1985), the carbon steel feedwater flow venturis were replaced with stainless steel venturis, which have exhibited a lower propensity toward fouling than their carbon steel predecessor.

N ,

7 ,

i

~ LICENSEE EVENT REPORT (LER)-TEXT CONTINUATION SAN ONOFRE. NUCLEAR GENERATION STATION DOCKET NUMBER LER NUMBER PAGE 1 UNIT'2- 05000361 88-035-01~ 6 0F 7 l

~ b. _ The new methodology for approximating reactor power was . l applied to remaining Unit 2 and Unit 3 historical power data, i No other instance of operation above 102% estimated-actual lj power was identified.

c. ~ The Plant Performance Monitoring Program' has been enhanced to - l

. continue to monitor plant parameters related to the  ;

determination of plant power, including: 1) trending-the 'I

~ feedwater venturi flow transmitter outputs to determine when a j venturi or' transmitter is failing;'and 2) application of the o new methodology to approximate.p1 ant power. This will serve j to reduce'the probability of operation _above 100% actual power (thereby reducing the probability'of operation above 102%-

actual power). r

d. SCE-is pursuing alternatives to address feedwater flow venturi. '(

fouling, including electrical isolation of the venturis to minimize (and potentially eliminate) fouling. j l '

E 'F. SAFETY SIGNIFICANCE OF THE EVENT:

Units 2/3 Final Safety- Analysis Report. (FSAR) section 15.0.3.2, " Initial Conditions", provides the range of values for each of the principal' y

' process variables that were considered in all the accident analyses. The specified maximum allowed initial core power is 102%. For many of the design basis accidents, therefore, an initial power level of- 103% in

' conjunction with the most limiting set of other initial conditions at the time of the initiation of an event could result in consequetces more-severe than those discussed in the FSAR,: assuming no operator action for the times specified in the analysis.

G. ADDITIONAL INFORMATION:

I 1. Differences Between Evaluations of Maximum Power Attained:

Reactor power is determined either by direct measurement of a y parameter (e.g., neutron flux), or by performing calculations utilizing measured parameters (e.g., secondary calorimetric). The accuracy of the power determination depends upon the inherent error of the input parameters.

An evaluation of Unit 2 plant performance performed in October 1984 concluded that 102% power had not been exceeded during Cycle 1 operation. The method utilized in this evaluation determined power based upon certain plant parameters, each of which can contain some amount of inherent error. The newer methodology for approximathg

. reactor power., developed and implemented as described in LER 88-028 (Docket No. 50-361), utilizes a greater number of parameters than L the earlier evaluation; as a result, a greater confidence level of l

l

e,

's

'b LICENSEE EVENT REPORT (LER) TEXT CONTINUATION

, ' SAN ONOFRE WUCLEAR GENERATION STATION DOCKET NUMBER LER NUMBER PAGE g; UNIT 2 -05000361 88-035-01 7 0F 7 s

the: accuracy of the' power determination is achieved. The inherent

' error of the earlier evaluation. led to the. inappropriate conclusion

.that 102%' power-had not been exceeded. The increased accuracy of the . newer methodology led to the opposite conclusion.

2.- . Component Failure =Information:

The feedwater flow venturi was manufactured by Vickery Simms, model FVI L- (20") .

3. Previous LERs for Similar Events:

g LER 88-028 (Docket No. 50-361)

Refer to Section C;l (Event).

4. Results-of NPRDS Search: ,

A-search of Vickery Simms Inc. venturis found a total of 8 failures.

None oi-the other failures reported included model numbers. j d

l Ji l

4 l

}

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