ML20235Z646

Rev 0 to Evaluation of Oyster Creek MSIV Leak Rate

ML20235Z646
Person / Time
Site:	Oyster Creek
Issue date:	02/28/1989
From:	Dempsey T, Lanese L, Johari Moore GENERAL PUBLIC UTILITIES CORP.
To:
Shared Package
ML20235Z641	List: ML20235Z639 ML20235Z646
References
058, 058-R00, 58, 58-R, NUDOCS 8903160052
Download: ML20235Z646 (29)

Text

. g. 4 o . -

.qgg e

l EVAURTION OF OYSTER CREEK m In = E m I 0t wI . m v u mx a - E

'IOPICAL REPOE 058 Rev. 0 B/A 315302 h

Preparpfbi J.' P. Moore . V A-2E-ff Data fW 2l20 V Prepared by T. D W () Date APPICVALS:

1

'" Mk .

I4 d. f- 2 /ze/g7

' D at'e 8903160052 890310 PDR ADOCK 05000219 p PDC

g,y *:. . n w :,4 1

~

Nuslear = "o - l 1

mu EVAI2BirION OF OYSTER CREEK MAIN STEAM ISOIATION VALVE LEAK RATE REV SG90JE OF CHANGE APPROVAL DATE 1

Paragraph nere 1.0, pg. 42 Revised purpose to be .g descriptive.

3g Paragraph 6.1, pg. 142 Clarified that engi-neering theory and not data was used for test b[u 3/9/fy  !

I a t = rr 1 ti -

ggae wh/n l Paragraph 6.1, pg. 15 Clarified that air actuator assumed to not be contributing to seating force.

Paragraph 6.2, pg. 15: Editorial change.

Paragraph 6.3, pg. 16; Paragraph 6.4, pg. 19 Identified range of test conditions and deleted sentence on focus of test program.

Paragraph 6.4: Clarification of Containment spray systen operation and effect on drywell pressure.

w &

Xfg 1

I TR 058

, Rev. O Page 2 of 28 TABIE OF CONITNTS Section Title Page 1.0 PURPOSE.................................................... 4

2.0 INTRODUCTION

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

3.0 BACKGROUND

................................................. 5 4.0 PIMf CCEFIGURATION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4.1 Primary Con &* 4 mmnt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4.2 Main Steam Isolation Valves........................... 7 5.0 LEAK RATE AND OFF-SITE D06E................................ 11 5.1 Technical Specifications Requirmants. . . . . . . . . . . . . . . . . 11 5.2 Imak Rate Test Acceptance Requirements. . . . . . . . . . . . . . . . 11 5.3 Off-Site Dose......................................... 11 6.0 METTIODS & ASSUMPTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 6.1 Overall Approach...................................... 13 6.2 M3IV Seat Leakage Tests............................... 15 6.3 Correlation of MSIV Seat Leakage Data. . . . . . . . . . . . . . . . . 16 6.4 Detami nation of Drywell Pressure vs. Time. . . . . . . . . . . . 19 6.5 Calculation of MSIV Seat Leaka Time............. 21 6.6 Dose Evaluation...............ge vs.

........................ 21 7.0 RESUIlrS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 7.1 MSIV Seat Isakage Test Results. . . . . . . . . . . . . . . . . . . . . . . 22 7.2 MSIV NoImalized Seat Isakage Curves. . . . . . . . . . . . . . . . . . 23 7.3 MSIV Seat Leakage vs. Time........................... 25 7.4 Dose Assessment...................................... 25 8.0 00NCLUSIONS............................................... 27 8.1 Appand i v J Ccmpliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 8.2 Technical Specification Ccupliance................... 27 8.3 Ctaparison with SEP Integrated Isakage Asstaption.... 27

9.0 REFERENCES

............................................... 28

"T*"+3na ,

,2j TR 058

', Rev. 0 .1 Page 3 of 28 )

1 i

LIsr OF FIGURES '

Number Title g =

l 4.2-1 Typical Main Stama Isolation Valve for Oyster Creek...... 10 6.2-1 WIV Imak Rate Test configuration. . . . . . . . . . . . . . . . . . . . . . . . 17 l

6.4-1 Primary contairunent Pressure Pollowing Recirculation Line Break............................................. 20 7.2-1 WIV Norsnalised Seat Imakage vs. Seating Force. . . . . . . . . . . 24 7.3-1 IDCh MIV Imak Rate Without Actuator Air Pressure. . . . . . . . 26 l

l

i. .

.-w

.my4py,,c .->

TR 058

, Rev. 1 Page 4 of 28 1.0 PURPOSE This report has been prepared to assess the Main Stama Isolation Valve (MBIV) Seat Imak Rate Testing performed at Oyster Creek. This testing was conducted to detamine the seat. leakage characteristics under design basis accident conditions and de&m=4= the relatiemahip of test results to App =adiv J test requirements and off-site does == - m t. l i

2.0 DFFRODUCTION l

This report sunnarizes an evaluation of the seat leakage characteristics of the Oyster Creek (OC) MSIVs. The existing method of MBIV seat leak rate testing datamines the leak rate with full control air system or nitrogen pressure applied to the valve actuator. This methodology is in conformance with 10 CRF 50 _ W iw J. An acceptable test which meets the requirements of 10 CFR 50, W4w J, fulfills the surveillance requirement of Oyster Creek Technical Specification 4.5.E.4.

This evaluation was initiated as ri result of GPUN observations that the MBIV leak rate is greater when control air pressure is not applied to the valve actuator, and GPUN =ama==nants in response to NRC generic letter 88-14. This evaluation characterizes )EIIV seat leakage as a function of the actuator pressure and the post ICCh containment pressure. Both the transient pressure res@onse of contalment and the I reductica in actuator pressure due to air systen leakage have been i accounted for. The MBIV valve leakage as a function of time is then l datamined. ' A ocuparison is made of the resultant integrated seat leakage to the leakage and resulting dose ===a====nt developed by the NRC as part of the Systematic Evaluation Prv-= (SEP) Topic XV-19.

l l

l l

L _ _ _ _ _ - - -

f n, gam., q

{ *

., TR 058 nov. O Page 5 of 28 l

3.0 BACFrmwam 1

The function of the primary contaiment system (which includes the l MSIV's) is to ar===whte the pressures and tasperatures resulting frtaa various postulated loss of coolant accidents and to limit the release of radioactive fission products to ensure that off-site dose is within 10 CFR 100 limits. The design basis accident for performing the Oyster Creek radiological analysis is a guillotine break of one of the 24 inch ID reactor coolant recirculation lines.

The MSIVs are classified as con &m4===nt isolation valves. The outboard MSIVs are provided with air operated actuators and the inboard valves j have nitrogen operated actuators. The MsIVs are designed to close in the event of air (nitrogen) s@ ply failure. The air and nitrogen styply systems are neither designed for the safe shutdown of the reactor nor to mitigate the consequences of postulated accidents. As such, the MSIVs are required to perform their containment isolation function without these sg ply systems.

MSIV leak rate testing was performed during the 12M outage to evaluate the validity of the concerns addressed herein. The HBIV's met the Technical specification allowable leak rate when normal control air system pressure was cont 4==mly styplied to the valve actuator, but e the allowable leak rate when the pressure was removed.

wwm.c - , .

M TR 058 '

hav O Page 6 Cf 28

'b 4.0 PIA 3FP GREEgmtRTIcst 4.1 Primary contalment The function of the primary contaitunant is to absorb the anargy release of a masrinnnn credible accident involving the rwture of the reactor coolant system, and to limit the amount of radioactive material reiM the environnant.

The primary contalment consists of two structures referred to as the drywell and the torus. These two structures are intercon-nected by several large diameter pipes that pass from the drywell to a system of smaller diameter pipes that are suhaarged at their W== below the torus water level. The design size and str1ngth of the ocatairament is such that it can respond in a pas-  ;

sive inarmar to the initial energy release of any size pipe break within the drywell, limit the pressure rise within itself to less than the design pressure and contain the majority of radioactive material released. The containment spray / emergency service water system will also provide long tena heat removal and a gradual reduction of the drywell internal pressure to further limit leakage. During operation the drywell atmosphere is purged with niti.e.

.,.s..-s.

"4 agy

s. ...f.

l l

\

l

CyWe!46N%ip%%** ~"**

.. TR osa

, Rev. O Page 7 of 28 4.2 Main Steam Isolation valves l

l The main steast isolation valves (MBIV's) are ocatalmnarit isolation i valves designed to minimize coolant loes' frem the vamaal and off-site dose in the event of a IDCh. The MBIV actuators are air (nitrogen) to open, and springs plus air (nitrogen) assist to close. The valves are shown in Figure 4.2-1. In the z --indar of this topical report the term " air" is used r=mgnising that the inboard MBIV actuator supply is nitrogen with a back-tp supply fresa the control air system. The MSIV's are designed to close within three to ten seconds. The three second ministat closing time is based on the transient analysis of the isolation valve closure that shows the main steam pressure peaks below the lowest safety valve setting. The ten second maximum closing time is based upon the steam line break accident analysis and resultant off-site dose analysis.

The basic design of all four valves is identical. The valves are 24 inch angled globe valves of 'T' configuration. The cup aha W M t moves on a center-line that is 45 degrees upward from the horisontal center-line of the piping run. The valves in the in-board and outboard set are each rotated inward towards each other

.yys mately 22 1/2 degrees frca vertical so that the air cylindars clear h-Q steam lines and other neighboring lines. The M t is attached to the bottora and of a stem that penetrates the valve body cover through a stuffing box.

c

.m The poppet is guided at the bottcet by three hard faced ribe that are cast integral with the valve body. It is guided at the top inside a cylinder liner, fitted snugly into the valve body but having a n =4nal running clearance between M t and liner.

swygeyys *vwww

• e ' ~ ' ' - '

~ ""$$

i

.. Tm oss

' new. o Page 8 of 28 Tts-bottcus of the poppet is hard faced and seals against a hard faced seat ring that is welded into the valve body. The bottcna of the stan is hard faced and acts as a pilot valve to balance differ-antial pressure forces on the M t. The pilot valve opens first during valve v 4M and closes last during valve closing to minimize stan loading and inpact forces on i~yyst and seat. An internal spring is provided to keep the pilot valve open when the poppet is off its seat. A failure of this spring will not prevent valve closure. The diameter of the main valve seat is w umi-Instely the same as the inside diameter of the pipe, and the entrance and exit are streamlined to minimize pressure drop through the valve during nonnal steam flow.

The typer and of the stam is attached to the air actuator piston rod. The air actuator is used for closing and v 4T the valve.

The piston rod goes through a hydraulic dash pot that is used for speed control. The air actuator is self stpported on six large rods screwed into bosses on the valve cover. Four of these rods are used as guides for four stacks of three springs each used to close the valve in the event that air pressure is not available.

The springs exert a downward force on a plate that is attached to the valve stan at the valve stenVpiston rod interface.

Each valve is I P tly controlled by a pneumatic system ocuprised of a four way spool valve, three way M solenoid valves for pa==1 operation and slow speed exercising, a three way Dc metanoid valve, an air accumulator and a check valve. Thw h

functice of the check valve is to isolats the actuator and maammilator frem the air supply in the event of an air supply systan failure. In opening, the air pressure is applied to the space below the air actuator piston. The valve stan atta h to the air actuator piston red moves upward. The first innh of stem travel lifts the pilot valve off its seat in the j~ypet and relieves the pressure on top of the pospet to downstroem pressure to r=eina the force required to lift the poppet off its seat.

TR 058 T nov. O Page 9 of 28 9

The valve stan then lifts the poppet. The valve stamt costivnian to move upward until a <*==har on the stant is back seated against a hard faced seal surface on the valve cover. At thin time, the four sets of closing springs are ocapramaad to their minimum height and exert mairin== closing force on the valve stant. As the stam moves upward the piston in the hydraulic cylinder moves with it, displacing oil frtat the space above the piston to the space below it through two external lines. This is a closed oil circuit and sets g the hydraulic cylinder to act as a dash pot thrPA the closing stroke. In closing, air is vented frta the space below the actuator piston and air pressure ( wu tely as poig) fremt either the plant air styply systant or an air =~u==ilator is applied to the space above the actuator piston. Two speed ocntrol valves restrict the flow of oil front the bottest to the top of the hydraulic cylindar and are adjusted so that the valve closes in the required time.

In the event of failure of the air supply, the closing springs on the valve together with the air remaining in the amummilator provide sufficient force to close the valve in the required time.

In addition, the valve manufacturer has confirmed that spring force alone is sufficient to close these MSIVs. The pilot valve r==mina open, balancing forces across the w t until the poppet o nes to rest against its seat. During the last inch of stant travel, the pilot valve closes and seals against its seat. At t$p time differential pressure builds q across the peppet. In a M M to the differential pressure across the poppet, the sigNrishsandtheaircylindersupplyasealingforceonthe poppet. In closing, two limit switches open when the valve has closed 10% (that is to the 90% open position) to initiate a reactor scram. These switches remairs open during the rest of the closing stroke.

1

gp ucy, - e . m,w,nda w -

- a; p2 1R ese I I any, e

[ - 5tge as of as a WSI

. r-i:  :, . 'm

p o3  ;

B .,

,, 1'  ; .

gE

$ll'  :  :: No '- '

u 5

5 hn pp h s1

= -

=.r ~ g 1 a l5E i

In

=

b:

e E

,J b

N !I l

m 'm y

~

U x@bM k!!.

ii(

i I

'{ fl 725 l-

~

l, 1 p!L ,

h.

,r

/'

[ l 13 v l

i 7 -

y l

%wi #

A N i g  !!

s ma\

to RA s

1 Figure 4.2-1. Typical Main Steam Isolatica valve Fbr Oyster Creek l l

gg3r@gipep?6i w -

rij {

, TR 058 Rev. 0 Page 11 of 28 e

5.0 IZAK RAEB ABC OFF BITE DOSE 5.1 Technical Specifications Requirements Per the Oyster Creek Technical Specification 4.5.F.1.d, the leak- l age rate of any one penetration as adjusted to a test pressure of 20 peig shall not exceed 5% of Lto (20), where Lto (20) is the HaaFI HEE 4 allowable operational leak rate at a 20 psig test pressure". Five percent of Lto (20) is 12.08 scfh for Oyster Creek.

5.2 Leak Rate Test Acceptance Requirements The MBIV Technical Specification local leak rate test is con-ducted at pressures of 35 peig and 20 peig. The test leak rate is ocepared with the 12.08 scfh (at 20 psig) leak rate limit.

In accordance with 10CFR50 W 4v J, this testing is conducted with the nonnal actuator air system lined q to the valve.

5.3 Off-site Dose 5.3.1 10CFR100 Part 100 of Title 10 of the code of Federal Regulations (10CFR100) pnwides critaria to be used in evaluating

_ ll$ reactor sites, and includes reference values of 25 REM k$- to the whole body and 300 REN to the thyroid to be used

'$f)[ in establishing an exclusion area and low population sone.

~ M WsM9WpfgtRy*e * ' ' ' "'?C TR 058 Rev. 0 Page 12 of 28 5.3.2 FBAR The Oyster Creek updated FSAR Section 15.6.5.8 reports off-site thyroid doses of 145 Ram at the site exclusion area and 117 rem at the low population sone. The results were originally presented in FDSAR, Amerdnant 68. There is no discussion of MSIV leakage assumptions for this calculation. Whole body daamm are reported as 3.5 Ram at the site boundary and 4.5 Rant at the low population zone.

5.3.3 SEP Evaluation As part of the systematic Evaluation Fisp== (SEP Topic XV-19) for Oyster Creek, the NRC performed a dose manaam l ment calculation to data m N the impact of MBIV leakage on off-site dose. Utilizing a General Electric Leakage Specification of 11.5 scfh for each MBIV, the SEP dose assaamaant assumed a continuous MSIV leakage of 11.5 scfh per main steem line (or 23 scfh total for two lines) for a period of 30 days. This asstaption resulted in a de&amhation of off-site dose of 1 Ram whole-body and 341 Rest 'Itryroid. As discussed in the SEP, these results were acceptable to the NRC Staff. Since the MBIV leakage contributes 334 Ram to this off-site dose, the SEP roccm-mended that a more realistic analysis be performed using actual 3eIIV leakage characteristics that factored in the

( variation (i.e.: reduction) of drywell pressure with 5?' time.

44 ,I The 11.5 scfh value used in the SEP dose maeamammt differs from that specified in the Technical Specifi-cation acceptance requirement. The basis for the Technical Specification requirement is discussed in Section 5.1.

$3hJTP9;9rme% "

yg, 1

,, 2R 058 Rev. 0 Page 13 of 28 6.0 MBHEES & EMEDIPFICES 6.1. Overall Apprcach The overall approach to this evaluation was to de&armina the integrated mass release due to M IV leakage over 30 days using actual valve leakage characteristics and zenlistic values of drywell pressure versus time following a DBA. This integrated mass release was than ocspared with the integrated mass release due to MBIV leakage frem the NRC SEP analysis (Reference 9.1). If the integrated mass release is less than the NRC SEP analysis mass release for each steam line, then the impact of M IV leakage on the DBA off-site dose is bounded by the NRC SEP analysis.

This -ryIvsch is appropriate for the following reasons:

(1) Reference 9.1 indicates that MSIV leakage is the principal con-tributor to the DBA off-site dose at the low population zone.

(2) Since drywell pressure has a direct impact on the MBIV leak rate, the NRC rannamands in Reference 9.1 that the DBA off-site dose analysis be performed using a realistic ===amamant of drywell pressure versus time.

(3) The release is delayed 37.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> in the NRC SEP model. By this time, the short-lived isotopes of iodine have decayed, laeving Iodine-131 with an 8 day half life as the principal

$ contributor to the 30 day DBA off-site dose at the low e

population sons.

(4) nanattaa of the 37.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> delay, the MSIV leakage does not contribute to the 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> dose at the site exclusica area Mmd=7 Furtbarmnre, the SEP analysis shows that whole body does at the low population zone is only a anall fraction of the 10 CFR 100 limit. Therefore, the only off-site dose of concern for MSIV leakage is thyroid dose at the low population zone.

.. 'IR 054 Reir. 1 Page 14 cf 28 M The 37.5 hour5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> delay in the SEP model is based on the r;. c

% maremmiation of NBlv leakage in the steen line. Therefore,.

if the integrated leakage over this period is less than the SEP value, the SEP sDdel is MM45 over this period even if the instantaneous leak rate h a the SEP value during a portion of the period.

(6) anos the release nrummannan at 37.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />, the dose rate is proportional to the leak rate. Furthermore, if the leak rate is constant after 37.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />, than the integrated dose I is proportional to the integrated leakage. This is true for both the SEP model and this analysis.. Therefore, if the 30 day integrated leakage is less than the SEP 30 day integrated leakage for each steam line, than the SEP inte-grated dose is M M 4M. Fut+harmnre, the ratio of the j integrated doses is equal to the ratio of.the integrated leakages.

The evaluation proceeded as follows:

DBh conditions were asmaned including loss of the nomal air / nitrogen supply to the isIV actuators.

isIV leak rate tests were performed to determine MSIV leak rate as a function of pressure differential across the seat and the valve seating force.

.g ,_

RE Reference 9.5 was utilised to establish the appropriate

1. aanlytical correlation for the leak rate test data.

The 3EBIV leak rate was correlated with pressure differential across the seat and valve seating force.

The FSAR drywell pressure time history for a 24" recirculation line large break IOCA was used.

gj.,p.nr.rs e m._., , . cpu - . . . - ~ b ,q3 TR 058 ,

Rev. 1 '

Page 15 of 28 l l

1 *

• ~.'the WIV sentimig loads and pressure differential across the

" i seat were calculated as a function of the post Ioch drywell pressure.

The WIV leak rate versus time was detamined using the leak rate correlation, pressure differential across the seat, and valve seating force. (The air actuator was assumed to not be ocatributing to seating force.)

The WIV leak rate versus time was integrated to detamine the integrated mass release.

This imitegrated mass release was ocupared with that fztzt the 180 SEP analysis for each steam line.

6.2 ISIV Seat Imakage Tests To assess the impact of MSIV leakage on off-site dose, it is j necessary to datamina MsIV leak rate during the course of the  !

accident. The SEP esuluation (R9f. 9.1) used an assumption of a occhimm= MBIV leakage of 11.5 scfh per stama line for 30 days.

The tra - Ittal letter friza the 3mc for Ref. 9.1 r e that the leak rate be detamined as a function of drywell pressure to more accurately assess the off-site dose. Consistent with that rhtion, a test prograr: was formulated to test MsIV leak rate over a range of drywell and actuator pressures.

lR .

16b le6kage tcst followed the basic approach of the h J Icoal Imak Rate Test (LLRF) p --Mtre (Ref. 9.2) . The suc Gare was modified to pezmit leak rate testing over s. range of sinallated drywell pream and actuator pressures to simulate the response of the NsIV during a Ioch with the expected drop in dry-well pressure and actuator pressure.

ir a

_- 1 3

TR 058

. BsF 1 Page 16 cf 28 A adsplified sketch of the test configuration used in Ref. 9.2 is adsun in Figure 6.2-1. Service air is delivered through a leak rate monitor to the space between the inhed anG outboard MBIV's to pressurize this space and thus simulate post IDch drywell pres-sure on the valve internals. An air supply with adjustable air pressure is connected to the valve actuator. Normal ocatrol air

') '

system pressure is maintai w on the inhM valve actuator, and I the main steam line upstream of the inhaard MBIV is flooded to pre-clude leakage through the inhnard valve. This ensures that the test results are for the outboard valve only. (scans of the initial tests were perfomed without flooding the steam line. The test ~ l pressure slightly =ramadad the static head of water for test points at 35 psig with the steam line flooded. Therefore, the data for i thaam testa represents leakage through both valves However, these data were used as though they applied only to the outboard valve).

6.3 Oorraintien of WIV Seat taalraqe Data Tb obtain realistic M IV leakage data, W IV leak rate tests were

{

performed over the full range of actuator air pressures (0-85 psig) j and simulated drywell pressures (0,5-35 peig). These leakage data  !

were correlated with pressure differential across the valve seat I and M IV seating force vaing regression analysis to obtain a leak rate equation. The basic principles applicable to valve seat leakage were utilised to develop the correlation and to judge the validity of the test results.

i p-

[ seat leakage is the flow of a fluid through a very small

~'armateG by unevenness in the seating surface. The flow pattern is generally laminar, with the mi. ion of turbulent l entrance and exit losses.

1 l

l

L P E P ). N UP I SY T OB E T R C R( U I VR I T RIA A E S

gE T N

1 t V

)

B M

( R A L

= C O 4 T

= D S L S N E L E D E I T SW TY UR E O. D T A

/Y///////[///// y//// R

)

B

( L K

=~ G 3 A

E E L A DW E

=~ D 0 S L S Y I

N N RD 9'

I V

I 2 S M

3 t R

MO O CT R A F E R

, l l

l

e < ' \

TR 058

, Ber. O Page 18 of 28 55e following leak rate equation is derived in Reference 9.5 for

~

lameinav f]ow:

3 3 QcrM = 4P H W (ft /seo)  ;

96 y L _

Qey,= seat leak rate in ft 3/seo at the density corr 5--iiM to a pressure of (P1 + P2) + 14.7 psia a 2

P1 = Valve inlet pressure, peig.

P2 = Valve outlet pressure, peig.

AP = (P1-P2) = Pressure differential across the valve seat, poi i

H = Height of gap between two surfaces, inehan j W = width of leakage ares, inches

= oirouaference of valve seat.

L = Iength of leakage path, inches

= width of seating surface

,U = Absolute viscosity of fluid, (lbf soo/ft2)

The leak rate can be converted to standard cubic feet per second as follows:

(Pi+P2) + 14.7 2

Oscre = Qcrs X _ 14.7 _

This leak rate equation shows that for a given valve ocafiguration at given fluid conditions, the leak rate is geoportional to the pressure differential across the valve seat m8:the p..

c@e of the gap. Solving the equation for H3 , we find .

I that/it is proportional to the leak rate divided by the pressure l differential. Since the gap is a function of the valve seating I force, the leak rate can be correlated with differential pressure and valve seating force by plotting Q/AP versus seating force.

TR 058

• Ref. 1 Page 19 of 28 ,

l

.i 6.4 DBbeamination of Drywell Pressure vs. Time 1

1 So ddamina the test conditions of interest for the leIV leak rate testing, it was necessary to understand how the drywell presture would change a? a D nction of time.

The Oyster Creek gdated F8hR (Ref. 9.3) indicates that the mari== pressure response for the containnant results fztsk a 24" recirculation piping guillotine break. Figure 6.2-3 of Ref. 9.3 (Figure 6.4-1) shows this pressure time response, and was used to ddamina Drywell pressure as a function of t3ms. FSAR 8ection 6.2.1.1.3 providas a description of the contaiment model used in developing the figure.

i Figure 6.4-1 shows that there is a rapid rise to a peak pressure l of about 33 poig within two to three seconds. It than drops to 5 poig within about ten minutes and ultimately 0 psig within about 6% hours. This drywell pressure response is based on i operation of containment spray down to o peig. It should be noted that the contaimaant Spray system would automatically shut down at 2 poig - decreasing; thatufore, operator action would be required to keep sprays on to reduos drywell pressure to o psig if this h necessary.

The initial rise to the peak drywell pressure is due to a rapid raisese of energy to the drywell in excess of the vent rate to f the, torus. Following the peak drywell pressure the reactor h pressure decreases so that mass flow fra the drywell to tieIorus is greater than mass flow rata fztza the reactor to the j

, drywell. Therefore, the drywell pressure decreases rapidly. The presence or a h of containment sprey has no effect en the pressure profile g to this point.

1

- _ _ _ __ _ _____ _ - __ ._ - _ . _ . l

Figure 6.4-1 Primary Onetainment h M W h im Line m  % 20 of as I

• 9 ,

)

a'*} .',

i a 1

- e, y- -

8 d

d E

e l -

~

9 8 -

C L-*-  ?

5 8 m

$ m O w - o

~-

E 5 a e E O o

U 6 o -

- >O

= s 5

d w W s E a d 1 o g -

g -

E 1

• o-4gs

. w\

<< . f; l l l l S S R R S (Sisd) 38nSS3Bd 1N3WNIVIN03 GPU NUCLEAR CORPORATION PRIMARY CONTAINMENT PRESSURE OYSTER CREEK NUCLEAR GENERATING STATION FOLLOWING RECIRCULATION LINE BREAK UPDATED FINAL SAFETY ANALYSIS REPORT .

R EV. O,12/84 FIGURE 6.2-3

-_w - . .

.. TR 054s Rev. 0 l Page 21 cf 28 Reactor pressure is the same as contai - t pressure'at approx-instely 30 seconds after Dan. Upon initiation of contm4==nt Spray there would be further reduction in drywell pressure.

For purposes of calculating the integrated leakage through the MSIV's, the final containment pressure was asstaned to remain at 1 psig for 30 days.

6.5 Calculation. of MSIV Seat Imakage vs. Time Calculations were performed using the correlations discussed in  ;

6.2 through 6.4 to de&amina the expected MSIV leak rate versus time. Drywell pressure is used at a point in time to determine the pressure differential across the valve seat. This pressure differential, along with the pressure on each side of t% actu-ator piston and the valve spring force are used to calcuY. ate the total valve seating load. The pressure differential and seating load are used with the leak rate versus load correlation devel- ,

oped per the methods in 6.3 to datamina leak rate at a point in time. The W _t leak rate as a function of time is determined by repeating this calculation at various points in time. The integrated leakage is detamined by F=ing the leakage during these time increments. ,

6.6 Dose Evaluation A4 arm ead in 6.1, the total integrated MSIV leakage can be

.dbated and ocanpared with that appearing in the NRC Staff's h ysis. If the 30 day integrated leakage is less than the SEP 30 day integrated leakage, then the SEP integrated dose is bounding. l

TR 058' i RsF. O  ;

Bage 22 Cf 28

,- 1 7.0 R50025 ..

7.1 MIV 8 eat Imakage Test Results The results of the testing described in 6.2 indicate that there is a direct relationship between total seating load and leak rate ;

for tests run at the same drywell pressure. A plot of these data I i

in standar11 cubic feet per hour (SCFE) versus stem load would show a characteristic similar to the leak rata vs. stem loads reported in EPRI report NP-2454 cn MIV leak testing. This report was prepared by the Atuood and Morrill ocupany (Ref. 9.4),

which manufactured the FJIV's being used at Oyster Creek.

The test data also indicate that at low drywell pressures (i.e.,  !

less than 2.5 poig) and sero actuator pressure the MIva have leakage characteristics that meet the 'hchnical Specification  !

leakage limit of 12.08 Scm. {

The assessment documented in this report was based on testing performed on the two outboard M IV's, Mo4A and M 04B. It was deemed unnecessary to oceduct testing on the inhaavit valves for the following reasons:

1. All four MBIVs have ocuparable leak rates at the LTJtr test conditions.

39 The MBIV's will close after a IDch with the force developed

$j hy the valve springs and the air ===ilator.

(w t

3. Onos the valve is closed, sufficient seating force is developed by the valve sprim and valve poppet weight (appzcximately 6000 lbf).

Thus, it is concluded the testing of the outboard valves for leakage with 0 psig actuator pressure after initial seating provides results applicable to all four valves. 1

g y p w m e,-m m m y or a .. . ~ J ~~ Q

.. . .. rm 05 RsF. 0

', Page 23 of 28 7.2 Iggy mmalized Goat Imakage curves i

'the results of the data correlation described in 6.3 are ,

presented as "nomalised" to the parasater '4}/P' versus total  !

l seating load in Figure 7.2-1. '9F' is seat leakage in CFH and )

"P' is the drywell pressure in poig which is equal to the differential pressure across the valve seat. As discussed in Section 6.3, Q/P is a function of H (seating gep) and varies as 3

H. E is soley a function of seating load for a particular valve seating surface.  !

i i

f The normalized results for the two valves tested (leB04A and l 3504B) show characteristics very similar to that espected for leek rate vs. load as reported in Ref. 9.4 and 9.5.

l It should be noted that at very high seating loads, leak rata l does not change substantially for additional increases in seat f load. Also, for very low seating loads, the leek rate does not change substantially for additiotal decreases in seating force. i The latter is most noticeable on Is04A (the valve with the  !

higbar leak characteristic of the tso valves tested)

These results indicate that the seating surface leakage area will not be modified substantially by higher loading. At low loads, these results indicate that the contact area between the meeting surfaces is essentially constant, resulting in a flow area.

i l

l l

1 r

l l

% l E

l l

q.

l l

T l A

l l

~

i R l l

R l l

A l l

E l l

l L l l

l l

D l l

E l

l l

Z l l

I l i

L l l

A l l

M l l

l R l l

O l l

N l

. l l

- i l

9 l l

4 1 I

1

- l V l l

l I l S 1 1

l l

M l l

l 1

& l 1

l 1 A l l

4 l l l

- l V I 1

l I

1 D A l S l R 4 l

l M l O - l i

L V B

l

_ C l I 4

l O l G S -

l I

N M l I

V l 1 I I l 1

R S l 1

1 v+

g1

.e* B P '

M 1 1

1 l

.eyN ._ l l

i 1

+

d _. _

Zz___ZE Z Z - Z_7

_Z Z_

• 7 G 5 4 3 2 1 0 O _P

pysy +

,. . iggm.a yw v- +

.c >

• cy:t

.. TR 058 RsF. 0 Page 25 of 28 7.3 IMV seat Isakage vs. Time Figure 7.3-1 shows the predicted leak rate vs. time for one emin steem isolation valve. This represents the ocedition where the other valve in that stama line has failed to close. ' The result is based upon the higher of the two valve c'haracteristias pre-sented in Figure 7.2-1. The leak rate falls below the Technical Specification limit of 12.08 Scru at w A taly 1/2 hour after the event ==-4ng zero psig actuator air pressure and staoe-phario pressure downstream of the lEIIV. Based on Figure 6.4-1, containment pressure, and thus leak rata, are aupacted to fall to zero in about 6-1/2 hours. However, the leak rate calculation

===- that containment pressure does not fall below 1.0 psig even though this pressure is reached at w =dmately 2-1/2 hours {

after the event. It is this leakage at 1 psig (apprariantely 5.3 SCFH) which represents the bulk of the integrated anse Islease f

over 30 days (720 hours0.00833 days <br />0.2 hours <br />0.00119 weeks <br />2.7396e-4 months <br />) .

7.4 Does Assessments The 334 Ram thyroid dose (due to MSIV leakage) fztza the Staff SEP evaluation of off-site does consequences r==mina MM4=y for the )

+ w MsIV leakage. An explained in section 6.1, thyroid does  !

would be directly proportional to integrated leakage. since the integrated leakage for each steen line (appraicleately 3,800 scf) I is less then half the integrated leakage used in the SEP does an======nt (w *==taly 8,300 scf) for each stama line, then l M off-site dose predicted by this an====nant will be about % l wi \

tho' Staff's value. offsite doses would therefore not exceed 10 CPR 100 limits.

E R l U _

S l S

E I R _

P I

R _

O I T

A I U _

T C I A I O

R I _

E _

Z I H _

T I

I _

W I H

E I F

T C_

A l g

R -

R I

s_

A E

L I

U_

x I

U I _

1 S .

_ x M

A

. 3spy- _

C O

g .a-L l I

- l

_~

__ =___

- 0 o O 0 0 O 0 7 s s 4 3 2 1 gEgkRgTE sCFH l

1  !

n .n ,,, m ,~

' TF. 058 l Rev. 0 1 Page 27 of 28 l l

l i

8.0 M RE1teEE l

)

8.1 annandiw J Ocagliance The MSIVes are tested in ocapliance with the requirements of 1

.t*4 w J.

l l

The testing conducted at sero actuator pressure confin s that an MBlv which passes the _Wiw J local leak rate test at a l contaitunant pressure of 35 psig with nonnal air pressure on the actuator, will have acceptable integrated leakage with sero actuator pressure. ..

8.a Technical Specifications Ocueliance i The requirements of Technical Specification 4.5.E.4 are fulfilled ,

by a p - =ful 10CFR50 W iv J, Type C test. f l 8.3 Ocuparison with SEP Integrated Imakage ?ETWon Based on the integrated leakage, the off-site does due to IsIV leakage is predicted to be about a factor of 2 less than the SEP  !

i value (334 Rest). Sections 6.1 and 7.4 discuss the basis for this oceclusica.

r, 44 9*  %

, :* - l f

i l

pp.yggW-- - Ei - - n c, - -

~f:g}

TR 058  !

new. O

< Page 28 of 28 V

9.0 REFE5tB E M 9.1 m Imtter (I805-82-09-011 dated 9/2/82 to GPUN (P.B. Fiedler, vice President and Director, Oyster Creek) Subjects Oyster Creek Nuclear Generating Station, Safety Evaluation of SEP Topic XV-19,

" Radiological hw of a Ioss Of Coolant Accident" 9.2 Oystar Creek Huclear Generating Station Procedure 665.5.003,

'9WIV Imak Rate Test" 9.3 Oyster Creek FSAR 9.4 EPRI Report No. NP-2454, "Ocuparison of Generio BNR-MSlV Configurations," dated June 1982 9.5 Hans D. Batmann "Should A Control Valve Imak", Instruments &

Control Systems, August 1966. A. W. Cash W M .

w

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