ML20084E322

1,Safety & Relief Valve Piping Design Rept

ML20084E322
Person / Time
Site:	Oyster Creek
Issue date:	08/31/1972
From:	JERSEY CENTRAL POWER & LIGHT CO.
To:
Shared Package
ML20084E321	List: ML20084E318 ML20084E322
References
NUDOCS 8304150018
Download: ML20084E322 (31)

Text

-

3 o

a OYSTER CREEK STATION NO. 1 SAFETY AND RELIEF VALVE PIPING DESIGN REPORT 4

1 1

i i

August 1972 f

8304150018 720822 PDR ADOCK 05000219 P

pg L

i

7 O

O TABLE OF CONTENTS I.

Summary A.

Background

B.. Conclusions II.

Safety Valves III.

Electromatic Relief Valve Piping Within the Drywell IV.

Electromatic Relief Valve Piping Within the-Suppression Chamber V.

References u

^2

! L-o O

I I.

SUMMARY

A.

Background

In the past two years, there have been incidents of steam plant safety l

relief valves breaking off their inlet piping at nuclear power plants:

specifically, one at the Robinson Plant and one at Turkey Point. Each occurred during the plant initial test program. In each case the fail-i ures were attributed to the reaction forces generated by the steam flow transient.

In view of these incidents, the designs of the safety and relief valve installations-at Oyster Creek were re-evaluated. The re-evaluation included the review of the existing calculations and performance of l

l additional analyses as required to provide a complete set of stress l

calculations. This report presents the main conclusions and a descrip-tion of the system modifications performed as a result of this evaluation.

B.

Conciusions 1

Safety Valves The Oyster Creek main steam system includes sixteen spring loaded safety valves, eight on each of the north and south steam headers. The review of the main steam safety valve system design concludes that the safety valve installation is satisfactory, based on USAS B31. I design criteria. USAS i

)

j

L-O O

B31. I was used for the original design and is invoked for similar systems in nuclear plants currently under construction.

The results of the review indicate that the reaction loads were properly considered in the original analyses: however, since the time the analysis was performed, the valves were changed from Crosby to Dresser valves. In view of this, the analysis was updated to reflect the installed valve characteristics and to include the stresses resulting from pressure, seismic and deadweight effects as well as the effect of the flow reaction loading s.

The results of this analysis are summarized in Chapter II of this report and show that the safety valve installa-tion is satisfactory.

2.

Electromatic Relief Valves The Oyster Creek steam system includes five electromatic relief valves, three on the south main steam header and two on the north main steam header.

The review of the original design analyses of the electromatic relief valve installation showed that relief valve reaction loadings had not been considered. For this reason, additional analyses were performed to calculate these loads and deter-mine whether the existing relief valve piping installation is satisfactory. The results of the analyses, based on USAS B31. I design criteria, indicates that four of the five valve-1

. a

q 0

O inlet connections and associated discharge piping would be overstressed. These stresses were calculated assuming that all of the valves in either header lift simultaneously as would be the case in the event their set point is reached during an actual plant transient.

It should be noted that the plant operator reports that to cate only one electromatic relief valve has been lifted at a time as a part of the required periodic tests of these valves. Since the stresses are reduced roughly proportional to the number of valves lifting on any one header, the calculated stresses were below yield with only one valve lifting.

In order that the relief valve piping meet the USAS B31,1 allowabic stress requirements during simultaneous discharge of all the electromatic relief valves, snubber type pipe supports have been designed and added to the piping system to absorb the reaction loads. This has satisfactorily reduced the stresses in both the valve inlet connections and in the discharge piping in the drywell to comply with USAS B31,1 The analyses per-formed to determine the relief valve reaction loads, to design the necessary piping supports, and to' determine the resulting stresses during valve relieving and during heatup and cool-down of the piping in the drywell are summarized in Chapter III of this report. 1

y 3

e.

N O

O,

L.

\\

As a result of the analysis of the,ajequacy of the safety a'nd.

3-relief valve connections, it was donclu'ded that the design of

\\

y \\.

the discharge tail pipe'.within the torus kould also be review-s g

t It was found that additional supports are requi' red to I

ed.

i.

.h support the loads which would occur in the event ali valves t

in a header rell'eys ultaneously 'f Therefore, ' support were

<.t 3

s designed and install 5( to as, Lure thyp. fai,1ure of this piping y

~q.

e s

i would not occur. Forther' review of this ' Aping'in!t510 torus r,.

N v

(

I and its supports is oiderway)to assure the long term adequacy.*

..\\

i of the design. The analyses performed and the supports

'n added for the two tailpipes are described in Chapter IN'of p hs j

c?

this report,

'w s

y m

g v

\\'

3 s.

j

\\

's t

i

\\.

i%*

4^y s

\\; +

L s... m,1,

_ g' s '

e's s.-

$ \\

~

't 3

wVL r

., N' (o gg N

,~N (g

s.

s m+

,v

.e g

L

-\\

- )

3cr.\\

,l u

,~,

4 a. A g. hg, a,

s..:. e 1

'4

\\.

N.

-N

,w'*

,\\

(g k O

w 4

m.s 9'

1. y1 y

3 l

\\ \\

\\

'*t

\\

\\.

9' L

e-

\\

-s N

,t m

i ok

.\\

s a

s.c 3

\\

i-p e

\\-

\\

\\

\\

L 3 s

s

~\\.

A.

t1

.. s s\\-

.n n

m

- 3.-

., s k

4

' og

^

. Y*

\\

y.

i L

O O

l II. SAFETY VALVES i

lt' The main steam safety valve installation at Oyster Creek consists of 16 i

spring loaded safety valves which are divided equally between the north and i

J I

south headers as shown on Figure II-1 The safety valves were manufactured by Dresser Industrial VaIve and Instrument Division and are described in l

reference 2 The safety valves are supported directly from the main steam headers by l

the inlet connections and, if opened, discharge steam directly to the drywell through an open ended tee as shown schematically in Figure II-2 Use of the r

tee in the discharge piping prevents large thrust forces from occurring j

during the steady state discharge of the safety valves, i

i i

Transient thrust forces occur, however, during the initial opening transient i

of the valves. These forces occur because the sonic time delay between the valve and the tee causes a higher steam mass flow to occur at the valve than at the tee at any instant of time. This gradient in flow causes an unbalance in the momentum and pressure forces between the valve and the tee.

l The magnitude of the thrust force is directly related to the rate of accelera-1 l

tion of the valve disk as it opens. A valve that opens quickly causes a rapid.

increase in steam flow which in turn results in a large thrust force. There-l N'

fore, the valve disk opening position versus time was assessed to estimate the' reaction forces'which occur. x l

^Q

gS

,e s

V b

The Oyster Creek safety valve opening characteristic transient was deter-mined by contacting valve vendors and reviewing previous calculations, as follows:

The valve manufacturer (Dresser Industrial Valve and Instrument Division) was contacted regarding the opening transient of the Oyster Creek valves. Dresser indicated that these valves pop open to about 70% of full flow in less than 60 msec and follow the shape of a (1-cos wt)

(

curve during the transient.

The Crosby Valve Company was also contacted regarding the opening transient cf Crosby valves. Crosby stated that their valves open 1

linearly from initial popping to about 70% flow in a minimum time of l

1 40 msec.

The original safety valve calculations show a catenary curve for valve opening to 60% flow in 40 msec.

T1.c opening characteristic used in the original analysis has the steepest slope and results in the largest thrust forces; therefore, it was used for the updated analysis. This transient develops a thrust force of 1660 pounds force. However, to account for impact loadings which may occur during relieving, the transient force was doubled to 3320 pounds for the analysis. This force acts approxi-mately along the center line of the valve discharge nozzle in a direction opposite the discharge flow.

The stresses resulting from the effects of seismic, deadweight and internal pressure forces as well as the opening transient impact loadings were calcu- -.

t 0

1ated at the following locations:

header to inlet nozzle connection valve inlet piping valve discharge piping The stresses at the header to inlet nozzle connection were determined using l

the method described in the Welding Research Council Bulletin (WRC) 107 (reference 3). This method permits calculation of the secondary bending stresses as well as the primary stresses. The primary stresses were com-pared to USAS B31. I allowable stresses. The sum of primary plus secondary j

stresses were compared to ASME Code,Section III, criteria since USAS I

B31. I does not provide acceptance criteria for secondary bending stresses.

i

{

A summary of the resulting stresses and a comparison to the USAS B31. I and ASME Section III allowabic stresses are included in Table II-1 The results show that the safety valve piping stresses are satisfactory and well below the allowables, i

e f

i i

i I,

7-

--.c,_

e SE 8

V

(

L E

A V

V L

A F

V E

I Y

L T

E E

R FA R

S E

L D

A A

C E

I H

hE P

M YT A

T

)

S S

0 H

E s

T V

R L

O A

N V

FE I

h L

E R

D N

A y

ON SE VL N

A

~

+

I E

E V

I Y

II T

F R

A U

S G

I M

F A

T O

ETS N

I

\\O A

M F

O N

SE O

I V

R T

L E

A A

D C

V A

O C

E L

FE

(

H IL M

Y k(

E

/

AE R

TS T

H E

T F

U A

O S

S)

L(

E 8

V A E L

CIV A

PL V

YA F

TV E

ILE R

O O

D STEAM q

DISCHARGE SAFETY VALVE f

DISCHARGE TEE

[

C%

(OPEN TO DRYWELL)

/

P

~

l m

(

g, l

l STEAM DISCHARGE TRANSIENT FLOW REACTION FORCE x

V V

MAIN STEAM HEADER MAIN STEAM SAFETY VALVE FIGURE II-2

TABLE II-1

SUMMARY

OF STRESS ANALYSIS OF SAFETY VALVE PIPING Allowable Type of Stress Stress Code and h

Location Stress (psi)

(psi)

Fo rmula*

• Discharge Hoop 1840 17,500

1. 0Sm(B 31.1)

Piping to tee Longitudinal 1970 21,000

1. 2 Sm( B 31.1)

• Valve Inlet Hoop 8260 17,500

1. O Sm( B 31.1)

Piping Longitudinal 8460 21,000

1. 2Sm(B 31.1)

• Inlet Nozzle Local Membrane Connection to

- Hoop 12,822 21,000

1. 2Sm(B 31.1)

Steam Header

- Longitudinal 7,441 Local Membrane &

Secondary Bending

- Hoop 24,763 59,100

3. OSm(Sec III)

- Longitudinal 15,548

• For A-106, Gr. C:: B31.1 Sm=17,500 psi; Sec III Sm= 19,700 psi

O III. ELECTROMATIC RELIEF VALVE PIPING WITHIN THE DRYWELL The electromatic relief valve installation at Oyster Creek consists of five electrically actuated valves and associated inlet and discharge piping. Two of these valves are located on the north main steam header and three are located on the south header as shown in Figure II-1 These valves discharge steam to the suppression chamber (torus) when the pressure in the main steam header exceeds the set pressure of the valves. Figures III-l and III-2 are isometric drawings of the electromatic relief valve piping from the valve inlet connections at the main steam headers to the discharge in the torus.

Evaluation of the Loads on the Piping Large forces occur along the axes of the relief valve piping when discharging steam from the relief valves to the suppression chamber. These forces are similar to those described for the saf-ty valves. They result from momentum changes as the steam changes direction through the piping cibows and from pressure changes as the steam expands moving down the piping. In the steady-state, these momentum and pressure forces cancel each other between upstream and downstream elbows.

Unbalances in these forces occur, however, during the initial opening tran-sient of the valve. Since the valves open quickly, the steam mass flow in the upstream piping is higher than it is in the downstream piping. This unbalance in flow causes the momentum forces in upstream elbows to be s

larger than the momentum forces in the downstream cibows. Similarly, because of the time required for the pressure wave to pass through the discharge piping, the pressure in the upstream piping and c1 bows will be greater than in the downstream portions during the transient. Consequently, the momentum and steam expansion pressure forces do not cancel. Fu rther, an additional pressure exists which occurs from accelerating the steam in the piping from zero flow to full flow. The resulting force from this pressure adds to the force unbalance.

The magnitude of the transient forces which occur on the Oyster Creek rellef valve piping were determined using a steam blowdown computer program (reference 4) which calculates the flow rates and pressures versus time during the steam discharge. In this approach, the system is divided into a number of nodes and connectors. Each node is assigned the proper volume, mass and energy. Mass is transferred from node to node through the connectors accounting for the flow resistances, acceleration, momentum, expansion, and choked flow effects during relieving.

The momentum and energy equations for each of the nodes and the related connectors are solved for individual, small time increments using the con-ditions at the beginning of the increment. The result of this calculation provides a set of conditions at the end of the time increment which are then used as the initial conditions for the following time increment. The solution proceeds in this manner.

_9

O

'Ihe properties of the steam for each time increment are determined from a set of steam tables included within the program. These tables include pro-1 perties of pure liquid, pure vapor or two phase mixtures. The data used to determine choked flow conditions were taken from the ASME Steam Tables.

1 j

The sequence of events considered in the analysis is as follows:

1

~

That portion 1

The piping is initially at atmosnheric pressure.

4 of the piping which is below the water levelin the torus is filled with water.

1 2

The relief valves are considered to open linearly from zero i

j to full flow in 0.15 seconds.

This 0.15 second time constant was provided by the valve manufacturer based on tests of a t

i similar valve, i

3.

The transient continues until after the relief valves reach full flow and the water is discharged from the piping permitting 4

the steam flow to reach steady state.

i i

The pressures and flows versus time for each node and connector were used i

to calculate the forces which occur at.each piping elbow during the transient.

i 1

j The force at each cibow is equal to the sum of (1) the force resulting from s

{

changing the'stemn momentum as it turns through the elbow and (2) the force.

due to the expansion and acceleration pressure acting on the pipe area at each a

elbow. Since the steam momentum and pressures at each elbow are not equal during the transient, a net force occurs on each straight run of piping.

The largest gradient in steam momentum and pressure in the piping in the drywell occurs at the time all of the mass of water in the piping submerged-

dn the torus is discharged. Up to this time, the pressure in the piping builds up while the water is accelerating. After the water leaves the piping in the torus, the steam is permitted to discharge to the torus causing a rapid de-crease in pressure and increase in flow.

l The results of the above analysis are shown as force vectors in Figure III-1 for the north header piping and Figure III-2 for the south header piping.

Description and Evaluation of the Modified System The loads applied during the electromatic relief valve opening transient are sufficient to overstress the relief valve inlet and discharge piping if unre-strained. Therefore, five hydraulic snubber pipe supports rated at 10,000

)

pounds each have been added to the system and two existing spring type hangers were modified. These supports are identified in Figures 111-1 and III-2 as N1 thru N3 on the north header and Si thru S4 on the southheader.

A description of the pipe support designs and a summary of the stresses in the modified system are described below:

A.

Pipe Support Design The hydraulic snubber pipe supports were manufactured by Bergen Patterson Pipe Support Corporation, and are described in detailin Bergen Patterson Catalog No. 66, part number HSSA-10. These snubbers transmit the reaction loadings to the structural members attached to the biological shield during the valve opening transient.

At other times the piping is free to move unrestricted. Consequently, these hangers do not impose large stresses on the piping during normal plant heatup and cooldown. -_ _ _ _ _ _ _ _ - _ _ _ _ - _ _

The forces which occur, the support design loads, and the maximum stresses in the supports are presented in Table III-1 The resulting stresses in these supports are considered acceptabic. These stresses were calculated by conservatively using the design loads rather than the actual loads calculated using the blowdown computer program.

This additional design margin is shown in Table III-1 B.

Piping Analysis The addition of the five hydraulic snubbers and the modification of the two existing spring hangers changed the piping stresses that occur during plant heatup and valve relieving. In addition, several of the reaction loadings that occur on the piping are not restrained. Con-sequently, it was considered necessary to perform a piping ficxibility analysis to supplement the original analyses. This analysis was performed for the as-modified system, including the additional five hydraulic snubbers and two modified spring hangers, and also considered the following:

The rate of piping thermal growth at the time the valves relieve is sufficient to cause some'of the hydraulic snubbers to lock up and prevent motion. This increases the pipe stresses due to thermal growth of the piping and was, therefore, included in the analysi s.

The pipe support flexibility, linkage play, and compressibility of the hydraulic oil in the snubbers will permit the piping to move i l l

I

O to a limitad extent when the reaction forces occur. This motion may increase the piping stresses and was modeled in the flexibility analysis.

The flexibility analysis was performed for the relief valve discharge piping system for the two main steam lines using a piping ficxibility j

analysis computer program (reference 5).

The cases which were analyzed included:

Case 1 Plant Heatup The moyennent of the relief valve inlet nozzles which occurs during normal plant heatup was imposed on the relief valve piping systems. This case includes the effects of the piping weight and the restraining effects of the spring hangers and rigid supports and the attachment of the discharge pipe to the 6'-6" vent line as it enters the torus.

Case 2 Relief Valve Discharge Flow Reaction Forces The piping motions and resulting stresses which occur as a result of the reaction forces during the valve opening transient were determined. The snubbers support the reaction forces but permit some piping motion due to hanger foundation resilience, manufacturing tolerances on the snubber connect-ing linkages and the compressibility of the snubber hydraulic oil. The effects of piping deadweight and plant heatup motions were also included in this case. The reaction forces used in the calculations were the design loads in Table III-1 Since

- 13 '

)

(,-,)

(~'js y

the reaction forces actually calculated using the steam blow-down computer program and shown in Figure III-1 and III-2 are all lower than these design loads, the piping stresses calculated are on the conservative side.

Case 3 Relief Valve Line Heatup During Valve Discharge The effects of thermal heatup of the relief valve discharge piping during valve relieving, including the effects of pressure, deadweight and plant heatup were determined as a third case.

This condition occurs at some time after steady-state flow in the relief valve piping is reached.

The results of the analyses of these three cases are summarized in Table III-2 In addition, the maximum combined longitudinal stresses due to internal pressure, piping motions during plant heatup, flow reaction forecs during the relief valve discharge transient, and piping weight are presented in Table III-2 As can be seen from the table, all the stresses are below the allowable. The isometrics of the piping system shown in Figure 111-1 for the north header and Figure III-2 for the south header identify the locations of the highest stressed portions of the system which are listed in Table III-2 l

l '

_ __ _a

T TABLE III PIPIN G SUPPOR T SUMM AR Y Maximum Identification Design Stress in Supports Actual Design Margin Symbol Load Due to Design Loads Calculated

'Desien Load (ibf)

(psi)

Load (Ibf)

Actual Load _

-O N1 10,000 9,250 4100 2.44 N2

~

3,000 333 770 3.90 N3 10,000 5380 1.86 S1 10,000 2800 3.57 S2 3,000 400 1530 1.96 O

S3 3,000 1150 2.61 S4 10,000 17,000**

9450 1.06

• Standard Bergen Patterson Components rated for 10,000 lbs.

• • This stress was conservatively calculated by assuming the full design load was applied to only one of the two supports for this rigid hanger.

TABLE III-2

SUMMARY

OF STRESS ANALYSIS OF RELIEF VALVE PIPING Point of Maximum Combined Allowable Code and Loadings Maximg Location Considered Stress Stress (psi Stress (psi)

Formula Valve Discharge Piping Care 1 - Plant Heatup Plant Heatup Thermal (1)-N orth 13,815 22,500 Sa= 1. SSm(#

Expansion (1)-South 6,228 (B 31.1)

+ Piping Weight Cace 2 - Relieving Plant Heatup Thermal (2)-North 13,259 22,500 Sa =1. SSm Transient Expansion (2)-South 8,016 (B 31.1)

+ Piping Weight

+ Flow Reaction Force Care 3 - Discharge Plant Heatup Thermal (3)-North 6,752 22,500 Sa = 1. SSm Pipe Heatup.

Expansion (3)-South 10,358 (B31.1) g

+ Piping Weight

+ Discharge Piping Heatup

+ Pre s sure Maximum Combined Internal ressure (2)-North 17, 539(f) 18,000 Sa = 1. 2Sm L:ngitudinal Stres ses

+ Case 2 (2)-South 11,256 (B 31.1)

Table III-2 (continued)

Point of Maximum Loadings Maxim Combined Allowable Code and Location Considered Stress Stress (psih Stress (psi)

Formula Valve Inlet Piping Idl Case 1 - Plant Heatup Plant Heatup Thermal (4)-North 979 26,200 Sa = 1. SSm Expansion (4)-South 5,369 (B 31.1)

+Pipine Weight Cace 2 - Relieving Plant Heatup Thermal (5)-North 5,992 26,200 Sa= 1. 5Sm Transient Expansion (5)-South 7,677 (B 31.1)

+ Piping Weight

+ Flow Reaction Force Case 3 - Discharge Plant Heatup Thermal (6)-North 1,703 26,200 Sa = 1. SSm Pipe Heatup Expansion (6)-South 6,744 (B31.1)

+ Piping Weight

+ Discharge Piping Heatup

+ Pre s su re Maximum Combined Inte rnalgres sure (5)-North 7,702 21,000 Sa = 1. 2Sm Longitudinal Stresses

+ Case 2 (5)-South 9,387 (B 31.1)

Table III-2 (continued)

Point of Maximum Type of Maximum Combined Allowable Code and Location Stress Stress Stress (psi)

Stress (psi)

Fo rmula Inint Nozzle Connection Local Membrane to Steam Header Hoop 10" Nozzle i 14, 86 1 21,000

1. 2Sm(d) 8)

Longitudinal South Heade 9,725 (B 31.1) h Local Membrane and Secondary Bending Hoop 10" Nozzle in 25,607 59,100

3. OSm(*

Longitudinal South Header 17,287 (See III)

(a)

Points are identified on Figures III-1 and III-2 (b)

These stresses are the result of applying the design loads which are greater than the actual loads shown in Figures III-1 and III-2 (c)

A-106, Gr. B: B31.1 Sm=15,000 psi (d)

A-106, Gr. C: B31.1 Sm=17,500 psi (c)

A-106, Gr. C: Section III Sm=19,700 psi (f)

This stress includes thermal expansion stresses due to plant heatup which are not required by B31. I to be included in the 1. 2Sm category. Therefore, the maximum combined longitudinal stresses h

tabulated above are conservative.

(g)

The stresses in the 10" nozzle are higher than the 6" nozzles in north or south header.

f f~~'g O

g 12080 1208f u

8

RELIEF VALVE f1 RELIEF VALVE f 2 6"

(i l 6"

'(5)(6) f

/

t1 h NI 725f 24" MAIN STEAM PIPE 8"

I N

2467f r

V 14" y4,

• 338"

[A(G(R HINCED SNUBBER d'

BELLOWS (3

HANGER N2 1.8 38" e 900#

l.0M" SEE NOTE l PLANT HEATUP MOVEMENT RIGID itANC)ER N3 EC23 14" NOTES:

f 1.

TifE ACTUAL TifRUST LOADS FROM TIUS POINT ON DOWN TO THE ANCHOR IN THE TORL*S WERE NOT CALCULATED FOR TIIE NORTil HEADER $1NCE THIS SYSTEM OPERATES AT LO% ER FLOWS AND PRESSURES TilAN T!!E IDENTICAL FIPING IN Tite SOUT!! !!EADER WHICl! WAS ANALYZED.

THE SAME DESIGN LOADS WERE APPLIED TO BOTil THE NORTH AND SOUTit HEADERS FROM Tills POINT ON DOWN 12" TO THE CANAL FITTING IN THE TORT lS*

(t)(2)

2. ONLY TIIE LARCEST l CAES ARE SitOWN, I.E.. THOSE TitAT RESULT FROM A LONG STRAIGift SECTION OF FIPE.

3.

THE NUMBERS IN PARENTHESES ARE LABELS FOR Tite POINTS OF MAXIMUM STRESS GIVEN IN TABLE Ill.2 6' 6" ID VENT LINE

\\

WATER LEVEL IN THE TORUS 30' DIA TORUS s

12"

/

(

\\

l 20" CANAL FITTING M

NORTH HE ADER DEt lEF VAL.VE DISCH ARG E PII'ING FIG UR E Ill.l

O J

1000f 16 1000f

• ai

)

24" MAIN STEAM FII N RELJEF VALVE II d,,,

I" (4)(6, 4" SNepBER 10" HANGER 33 HINC ED 3ELLowS

\\

3000f RELILF VALVE II (y

N sa '

400" 31000 1816f s.

at 8000f 14" 8

RELIET YALVE II III Z

,,,,/

,E OW SPRING i.36s" w

N^ No E" N.,

t isi

. PLANT HEATUP MOVEMENT ICID g.

IM F

14 8,,

33gg g gy, HANGERS 3

,33g j

SPRING M

HANGER I*07I" l*8 8" 14" P! ANT HEATUP MOVEMENT 2000f 8000f (1)(3)

NOTES:

1. ONLT Tite LARCEST LCAES ARE $1 town.1.E..

THOSE THAT RESULT FROM A LONG STRAIGHT SECTION OF PIPE.

2 THE NUMBERS IN PARENTHESES ARE LABELS FOR THE POINTS OF MAX 1 MUM STRESS GIVEN IN TA B LE U1 2.

12" 68 6" dL VENT LINE 10.000f WATER LEVEL IN THE TORUS N w 30' DIA TORUS

./

/

'k Q w

\\ \\

12" N.

t 20" CANAL FITTING l

l I

SOUTH HEADER ret.!FF VALVE DISCHAPGE P!PIM l

TK".URE IU 2

O O

~

i i

I, '

}

IV.

ELECTROMATIC RELIEF VALVE TAILPIPING WITIIIN TifE SUPPRESSION CHAMBER i

i The discharge tailpiping within the suppression chamber (torus) is an extension of the electromatic relief valve discharge pipe in the drywell and functions to carry the steam underwater to discharge at a point near l

the bottom of the torus. Unbalanced forces occur in this piping during the relief valve opening transient in the same manner as in the piping in

]

the drywell as discussed in Chapter III. In addition, a large unbalanced-i load occurs at the pipe exit due to the expulsion of the slug of water 1

j j

initially in the pipe, 1

]

The loads and stresses in this piping were calculated using the same i

computer program discussed in Chapter III. It was found that additional a

supports are required to support the loads which would occur in the event I

j all valves in one line relieve simultaneously. Therefore, additional sup-ports have been designed and installed which will prevent failure of this j

piping. The a rangement of these supports is shown in Figure IV-1 1

i j

An isometric of the piping within the torus is shown in Figure IV-2.

The added supports consist of a 1" diameter wire rope' wound twice i

around the discharge end of each tailpipe and the vent header. The sup-1 pression chamber vent header is supported by existing pipe stanchions.-

4 i

The results of the analysis of the discharge tailpiping and supports are

+

summarized in Table IV-1. The location of the maximum stress in the 2

tailpiping is shown in Figure IV-2 -

q o

o, As can be seen from Table IV-1, the stresses are satisfactory. The stress in the discharge pipe near the canal fitting exceeds the allowable; however, it is less than the minimum yield strength (about 80% of yield).

The calculation of this stress was done in a conservative manner, in that the horizontal deflection of the canal fitting (see Figures IV-1 and IV-2) was determined by applying the total steady-state discharge force of 56,200 lbf or transient force of 94,000 lbf to the wire rope. No credit was taken for the lateral support of the chamber support ring or the 12" dis-charge piping itself in determining the displacement of the canal fitting.

In addition, the entire pipe was assumed to heat from room temperature to 550* r even though much of it is submerged in the suppression water.

Further, the stress results from the deflection caused by the stretch in the cablqq, therefore, only a limited strain can occur.

Table IV-1 also shows the conservatism in the design of the cables. The cables have a factor of safety of 4. 8 for the steady-state load and a factor of safety of 2.9 on the peak load based on their minimum breaking strength.

~

The desiga of the supports for the tailpiping in the torus is being reviewed to assure its long-term adequacy; however, it is concluded that the design as installed will prevent failure of the system.

e-8 SEP I

i P

v R

T E.

E R

L I.

M O

L R

O P

E P

H E

C U

S S

S J

N E

U S

S A

W U

P R

j F

O 8

R O

O 0

k H

O R

T D

C T

E E

^

[

D.

R H

S O

A I

2 W

T 4

I D

N 2

2 I

'0 A

H ID T

R 3

I E

W I

B E

M I

C R

A R

U C

l t

~

I H

W T

i~

C 1f 5

U N

'm*

2 R

O x

T I

S 6

S S

E E

4 T

R L

R P

D O

E P R D

1 P

N UE A

P I

SD S

U V

IL

.A S

I

\\

D. E T

6 H

D E

N x

N R

I E

V 7N 2"

l 8

A U

"T G

G I

/

E N

F I

D.

'4 V IP I

I

'r P

6 L

F I

6' A

T E

G RA G

H c

G N

C N

I S

I I

R D

T E

T T

E V

I R

V V

AN F

O L

L G L

P A

VI A

P V

P FI N

U S

F E P A

C R

E I

~

LE E

L I

EG B

E

)

RR D.

M R

A O

A TH Y

^

SC H

X 0

C "S

2I 2

1D 8

~

~

Z Z.

l.

~

~

~

o o

12" 68 6" 1. D. VENT LINE h

4 rWATER LEVEL IN N

/

THZ TORUS i

/

I I

TORUS

\\

c:. ?.

'9 %e' 1

l 94,000#

j POINT OF PEAK STRESS IN

+

4 TAILPIPE k

/

20" CANAL FITTING 4

4 4

NOTE:

ONLY TIIE LARGEST LOAD IS SHOWN SINCE THE OTHERS ARE NEGLIGIBLE BY COMPARISON.

RELIEF 7ALVE DISCHARGE TAILPIPING WITHIN THE TORUS FIGURE IV-2

TA B LE IV-1 o

SUMMARY

OF STRESS ANA LYSIS VA LVE DISCHARGE TAILPIPING IN THE TOR!g f

Loca tion Load (thf)

Stress (psi)

Allowable (osi)e Code Comment 12" Tailpipe in the Canal Fitting

1. SSm = 22, 500 B31.1 Occurs at. 250 seconds into Transient 94,000 28,053 S = 35,000 Section III the transient just as the last Y

slug of water leaves the canal fitttng.

During steady state discharh St ea dy-State 56,200 +

28,229

1. SSm = 22, 500 B 31.1 Thermal S = 35,000 Section III when the pipe heats up.

I Hea tup 12" Pipe Bend Inside 94,000 21,600

1. SSm = 22, 500 B 31.1 At. 250 seconds the Canal Fitting Shear Between Canal Fitting and Support 94,000 2,600

. 6Sm = 9,000 B31.1 At. 250 seconds Foot Tencile Between Canal Fitting and Support 94,000 3,640 Sm = 15,000 B31.1 At. 250 seconds Foot Tennile on Wire Rope Transient 125,000 Factor of Safety = 2. 9 At.250 seconds Factor of Safety = 4. 8 During steady-state Steady-State 74,500

• For A-106, Gr B

m O

o l

l V,

REFERENCES 1

Drawing 2103-10, " Steam Piping-Plants, Section and Details-Reactor Building. " Revision 10, Oyster Creek Unit #1, Burns and Roe, Inc.

2.

Drawing 6-3777 QA-RT-21-OS100, "Maxiflow Safety Valve, " Dresser Industrial Valve and Instrument Division.

3.

Welding Research Council Bulletin 107, " Local Stresses in Spherical and Cylindrical Shells due to External Loadings, " August 1965.

4 MPR Computer Program based on " FLASH-4: A Fully Implicity Fortran t

IV Program for the Digital Simulation of Transients in a Reactor Plant,"

March 1969, WAPD-TM-840, Bettis Atomic Power Laboratory.

5

" Pipe Flexibility Analysis Program MEL 21",

Report No. 10- 66, R ev. 1, A Modification of Program MEC 215, San Francisco Bay Naval Shipyard, Mare Island, dated August 1967.

l 1

- _ - __ - -