ML20079G384

Purge Valve Operability Rept,Snupps Project

ML20079G384
Person / Time
Site:	Wolf Creek, Callaway, 05000000
Issue date:	12/31/1983
From:	STANDARDIZED NUCLEAR UNIT POWER PLANT SYSTEM
To:
Shared Package
ML20079G373	List: ML20079G384 NRC-84-0004, Forwards Purge Valve Operability Rept,Snupps Project, Per Licensing Confirmatory Issues 28,II.E.4.2 & B.28,II.E.4.2. Rept Demonstrates Valves Have Capability,W/Margin,To Perform Containment Isolation Function
References
NUDOCS 8401200062
Download: ML20079G384 (39)

Text

_ - - a - u -,,:. - - - - - o - a e

4 l%

4 b

PURGE VALVE OPERABILITY REPORT SNUPPS PROJECT DECEMBER 1983 t

A

~

INTRODUCTION

~

This report has been prepared to allow the NRC to close out the coni'ima-tory items (Number 28, TMI Action Plan II.E.4.2) for Callaway and Wolf ~

Creek.

Thes information contained herein addresses the items contained in the NRC paper " Clarification of Sept. 27 [1979] Letter to Licensees Regarding Demonstration of Operability of Purge and Vent Valves." In addition, the information contained herein addresses Attachments 4 and 5 of NRC letter B. Youngblood (NRC) to D. Schnell (UE) dated September 7, 1982.

The following data are provided on the valve / actuator design used on the SNUPPS plants' mini purge containment isolation valves:

Valve Assembly Construction 18-inch Type 9220 Fisher Controls Company butterfly valve Class 150 flanged x BWE body 2-inch shaft, Class 4, 17-4PH Plate disc, carbon steel Adjustable EPDM T-ring seal Graphite-bronze bushing #2 GH Bettis Actuator T416B-SR3-12 l

ASCO Type NP8320A185V, 3-way solenoid l NAMCO EA180-31302/32302 limit switches Fisher type 95H pressure regulator Versa VSP-3601-155H 3-way switching valve Texsteam Model 35R pressure relief valve Hoffman junction box, Model A1008-CHNF with G.E. No. EB-25 terminal strips and assorted seals Fisher P595 filter with brass element S_ervice Conditions o

Fluid: Normal (air), accident (steam)

Design P/T: 70 psig/320 F Shutoff AP: 60 psid 1

t l . _. . __ .._._ . _ __ , . . . - . . _ . _ , _ . _ _ . _ _ _ _ . . . . - __ . . . _ _ _ _ _ _ . . = . . _ . , . - - . , , . - - . _ . . .

\

1 o l Service temperature: 120 F i 90 degree rotation ,

Closure time: 3 seconds maximum .

Maximum LOCA containment pressure: 47.2 psig (61.9 psia) k C ntairment pressute at valve closure: 22.9 psig (37.6 psia)

Seismic qualification level: 4.5 g RIM (Seismic Category I)

Applicable ASME Section III Code ASME Code Class 2 Summer 1975 Addenda The responses to items in the clarification paper follow.

l l

{

1 l

l 2

OPERABILITY OF PURGE AND VENT VALVES

~

1. The AP across the valve is in part predicated on the containment pressure and gas density conditions. What were the containment conditions used to determine the AP's across ths valve at the incremental angle positions during n, the closure cycle?

RESPONSE

The LOCA containment conditions utilized in predicting the AP across the valves were extracted from the containment pressure-temperature analysis for the large LOCA which results in the peak contained pressure. The pressure response is shown in Figure 1, which also indicates the times that the valve receives a signal to close and the time that the valve is fully closed. When the valve is fully closed the containment pressure is

22.9 psig. As shown in Figure 1, the contairement pressure rises from approximately 14 psig to 22.9 psig during the closure stroke; however, the pressure drop at incremental valve positions was calculated at the maximum pressure (22.9 psig: valve closed). This was done for conservatism.

i The calculated pressure drops considered the line and component pressure l drops for the steam-air mixture discharging through the line. The calculations also conservatively assumed that the redundant valve in the same line had failed in the open position. These assumptions resulted in conservatively high pressure drops during the closure cycle of the operable valve.

I Refer to the responses to Items 3 and 5 for further discussions on the installed configuration and the calculated pressure drops.

f l

l

[

l .

3

o

2. Were the dynamic terque coefficients used for the determi-nation of torques developed, based on data resulting from i actual flow tests conducted on the particular disc shape /. '

design / size? What was the basis used to predict torques developed in valve sizes different (especially larger , ~

valves) than the sizes known to have undergone flow tests?

RESPONSE

a. Dynamic torque factors used in butterfly valve sizing were developed from test data obtained from models with similar disc configurations and flow characteristics. The dimensionless aspect ratio (defined as the ratio of the disc diameter to the thickness) was judged to be -

a significant parameter for evaluation of dynamic torques at various opening angles. Therefore, a series of water flow tests was con-ducted with a group of 4-inch and 6-inch butterfly valve models constructed with various aspect ratios, ranging from 3:1 to 14:1 (such as 3:1, 8:1,11:1, and 14:1), in v:rious disc configurations (conventional, offset, cammed), and in both flow directions.

The tests were conducted using the Fluia Controls Institute (FCI) specifications for test arrangement and conduct, per FCI paper 58-2.

The basis followed by the vendor (Fisher) in using incompressible (water) flow model tests to establish dynamic torque coefficients applicable to large diameter valves in compressible flow service is presented in the ISA Transactions article provided as Attachment 1

(" Effects of Fluid Compressibility on Torque in Butterfly Valves" by Floyd P. Harthun).

I 1. On page 282 of the referenced paper, a relationship is developed relating torque to the nominal valve diameter and pressure drop for incompressible flow conditions, namely:

TD = K1 D8A P (equation 10-A) in which K2 is a dimensionless torque coefficient for incorpressible flow. It may be noted that this relationship applies to various valve sizes as well as to various pressure drops. The K2 value is determined by test (using water) for various rotation angles as shown in Figure 1, page 283, of the ISA paper.

2. The referenced ISA paper then considers dynamic torques developed during compressible fluid flow. In this case, the actual torque is no longer linear with respect to AP (incompressible flow torque has a linear relationship with respect to pressure drop 4

4

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

i RESPONSE TO ITEM 2 CONTINUED across the valve up to the choked flow condition); for conven-ience, however, a similar relationship is developed using a AP, factor as follows:

TD " "1DsAP,(equation 25)

" This relationship (equation 25) permits determination of dynamic torques for various valve sizes under conditions of compressible flow, providing Km and AP,are determined.

3. Figure 5 on page 284 summarizes the rationale presented in the ISA paper, this figure shows a plot of calculated and experimental torque values for various pressure drops and flow conditions.

The straight line labeled " Incompressible Flow" is a represen-tation of the relationship: Tn = K2 08A P, while the curved line represents compressible flow t5rque values, following the relationship: T K DsAP are also plotted! = showing lose E. Test data agreement withfor thecompressible calculated flow values.

4. Figures 6, 7, and 8 on page 285 of the ISA paper show a com-parison of test data and calculated data for various angles of rotation. Close agreement with AP torque values for compres-sible flow is indicated. (Note: Thereareafewtypographical errors in the ISA document figures: the AP/Pt value given in the caption to Figure 7 should be 0.155, and in the caption to Figure 8 this value should be 0.466.)

b. Application of the conclusions from the ISA paper is presented in Attachment 2, Selection Figures 1 and 2: " Torque-Pressure Drop ,

Relationships Used in Dynamic Torque coefficient Selection," which are representations of the Figure 5 curves from the ISA paper. These show how the torque values for compressible flow are conservatively determined and related to incompressible flow torques. It should be noted that the compressible flow curve reaches a critical flow con-dition at larger AP values, resulting in a maximum torque value (T ) that cannot be exceeded, regardless of how large AP becomes.

c

1. A AP value is obtained from the Fisher sizing tables based on the INolute inlet pressure value, type of disc, and flowing i'luid.

2. If the actual AP is greater than AP then the AP value is

• used, corresponding to T This isconservative, predicting (themaxi$N,possibletor$N).

a higher torque than is actually present (see Figure 1 in Attachment 2).

3. If the actual AP is less than AP t is used, which predictsamoreconservative(grINe,r)henAPtorqugjtbut less than T, (see Figure 2 in Attachment 2).

5

. l RE3PONSE TO ITEM 2 CONTINUED

4. An actuator is selected, based on the predicted maximum torque requirements. The actuator must be able to provide the necessary torque to open, close, or maintain the valve in position.

c. Summarizing our previous statement, @namic torque sizing equations g (equations 10-A and 25) were developed relating torque to the nominal valve disc diameter, pressure drop, and a constant Ks.

Through testing of various valve models, data were obtained for K 2 at incremental angles of opening. Consequently, the scaling method was used to obtain dynamic torque for larger size valves. .

Note: The AP value (not to be confused with AP used in the ISA paper) is bas! bon critical flow conditions that wT11 produce the maximum possible dynamic torque.

6

3. Were installation effects accounted for in the determination of dynamic torques developed? Dynamic torques are known to be affected for example, by flow direction through .

valves with off-set discs, by downstream piping backpressure, by shaft orientation relative to elbows, etc. What was the basis (test data or other) used to predict dynamic torques for the particular valve installation?

RES NSE: .

The installation of the containment purge valves is shown in Figure 2.

The 36-inch large volume, shutdown purge valves are closed during modes 1, 2, 3, and 4, as required by the plant technical specifications. The 18-inch mini purge valves on the supply (GT-HZ-04 and GT-HZ-05) and the exhaust (GT-HZ-11 and GT-HZ-12) are designed, tested, and qualified in order to allow them to be open at any time during the plant life.

The mini purge valves and piping are connected to and are part of the containment penetration for the 36-inch shutdown purge system. Plant arrangements and available space hava dictated the arrangements shown in Figure 2. It was not possible at the time of the initial routing to provide separate penetrations or straight piping on either side of the mini purge valves or between them.

Asymmetric flow effects are known to be a concern to the NRC and indus-try. Significant margins in available torque have been provided to ensure that the valves will close as required in the installed con-figuration.

As shown in Figure 2, the valves inside the containment have elbows either upstream or downstream of the valves. Valve GT-HZ-05 has a 45-degree elbow upstream and valve GT-HZ-11 has a 90-degree elbow down-stream. Any asymmetric flow conditions developed due to these elbows will tend to assist valve closure. Refer to Figure 2 for the direction of flow and valve closure. The installation of all mini purge valves is similar in that the valve shafts are vertical and the Bettis actuators are installed below the piping with the common actuator shaft (connecting the air side and spring side actuators) parallel to the piping. The actuatcr shaft drives a scotch yoke which is attached to the valve shaft.

The valve shaft (2-inch diameter) penetrates the center of the valve disc

(~4-1/2 inches thick) and is completely enclored by the valve disc. The shaft / disc arrangement is symmetrical.

l The valves located outside of the containment (GT-HZ-4 and GT-HZ-12) are

, located at the end of 36-inch x 18-inch tees. The flow into the valve will be highly turbulent and well mixed; however, if the flow is postu-lated # flow predominantly to the outside of the piping, the asymmetric flow would tend to resist valve closure. The asymmetric flow is not expected to be as significant for flow exiting the 36-inch x 18-inch tee as it would be for an 18-inch 90-degree elbow.

The effects of asymmetric flow on the dynamic torque coefficients have not been quantitatively analyzed. However, should the dynamic torque 7

. - - - . - - . . - _ - - - _ _ _ . - . -- w . - . - - - - _ _ - - _ _ - _ _ -

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

!' RESPONSE TO ITEM 3 CONTINUED required double, as suggested by the NRC as a worst case *, significant margins would still remain between the available torque from the actuator and the required torque to close the valves outside of the containment.

As shown in Figure 5 and Table 1, the excess torque available ratios for

' the design case range from 7.0 to 14.5 in the open positions. Should the

@ ic torques required double, a margin ratio of 3.5 to 7.2 would still n between the.available torque and the required torque to close the isolation valves outside of the containment.

It should be noted that there is no credible failure mode for the valves once the valve receives a signal to close and the solenoid valve vents the small quantity of air holding the Versa valve in position. The Versa valve then opens to vent the air cylinder through a 3/4-inch opening.

However, the stated required torques are based on only one valve closing.

Once a valve starts to close, the line losses and the pressure drop through the redundant valve become minimal. These reduced flow conditions would help ensure that the redundant valve would close even more rapidly

! than the required design value.

Also, the required torques have additional margins in that they are calculated based on a 22.9 psig drop across the valve at all positions.

Since the containment pressure will only be 14 psig maximum when the valve receives its closure signal, the pressure drop and the required torque will be significantly less than those calculated. Similarly, the pressurization transient shown in Figure 1 is for a worst case large LOCA snd has many conservatisms built into the analyses.

Another noteworthy conservatism exists in the reported dynamic torque requirements shown in Table 1. The maximum torque required to open or close the valve in the specified position is listed. For the 9200 series valves, flow generally tends to close the valves for straight line installations. The maximum torques listed are those required to open the valve. The torque required to close the valve is less.

i The actuator torque output ratio to torque required for case 2, the peak calculated containment pressure, varies from 4.0 to 7.9 for all open angle positions (refer to Table 1). Thersfore, even if it is postulated that the inside containment valves fail to receive a signal to close and the closure of the outside valve is delayed until the peak pressure is reached inside the containment, excess torque available ratios of 2.0 to 4.0 would still exist for the closure of the outside contcinment isola-tion valves should the dynamic torque required double due to assymmetric flow considerations.

In summary, the @namic torque requirements do not include asymmetric flow considerations; however, significant conservatisms are provided in all aspects of the design and analyses to ensure that the valves will close as required in the installed configuration.

l l "See item 3 of NRC paper entitled " Operability Qualification of Purge and Vent Valves."

l l 8

.- _: _. m_ . _-. v _-_"_ _~ _-"_ _~~ _~ _ _ , _. _ , _ _ _

~

c __ :_ = :: = -

4. When comparing the containment pressure response profile against the valve position at a given instant of time, was i the valve closure rate vs. time (i.e. constant or other) -

taken into account? For air operated valves equipped with spring return operators, has the lag time from the time the valve receives a signal to the time the valve starts to stroke been accounted for?

k NOTE: WIlere a butterfly valve assembly is equipped with spring to close air operators (cylinder, diaphragm, etc.),

there typically.is a lag time from the time the isolation i signal is received (solenoid valve usually deenergized) to the time the operator starts to move the valve. In the case of an air cylinder, the pilot air on the opening side of the cylinder is approximately 90 psig when the valve is open, and the spring force available may not start to move the piston until the air on this opening side is vented (solenoid valve de-energizes) below about 65 psig, thus the lag time.

RESPONSE

As noted in the response to Item 1 and as shown in Figure 1, the con-tainment pressure during the closure cycle was assumed to be 22.9 psig, which corresponds to the time at which the valve is fully closed (T = 6 seconds). The lag time between the receipt of the signal to close at 3 seconds and the initiation of valve motion has been taken into account.

Tests have ensured that closure is accomplished within the required 3 seconds from receipt of the closure signal.

Since the air cylinder is vented through the large Versa valve opening (3/4-inch diameter), the air cylinder will depressurize rapidly. As shown in Figure 3, approximately 80 psig in the air cylinder is required to balance the spring force and maintain the actuator / valve in the open r

position without flow. A pressure regulator is provided on the valve to j allow a reduced pressure in the air cylinder. This regulator will be set l to ensure that the valve remains fully opened. Depressurization of the l

air cylinder with respect to time is shown in Figure 4. The times l

required to vent the cylinder from any initial pressure less than 100 psig

can be taken from Figure 4. For example, depressurization from 100 to 80 psig is accomplished in approximately 0.1 second when venting to atmospheric pressure. As noted in the response to Item 9, venting of the accumulators to a positive containment pressure will not affect the '

l closure time.

l*

i l

l 9

.__.__.,__-.____l_.__________.____

e i .-

5. Provide the na:essary information for the table shown below for valve positions from the initial open position to the seated position (10' increments if practical). .

Valve Position (in degrees - 90* Predicted AP Maximum AP (capability)

= full open) {acrossvalve)

REENONSE: .

Table 1 - Case 1 provides the calculated pressure drops across the valve for each 10 degrees of valve position. These pressure drops were calcu-lated based on the installed piping system resistances, assuming that the redundant purge valve in the same line had failed in the open (least resistance) position. The predicted pressure drops are all calculated at 22.9 psig, which is the maximum pressure that would exist prior to valve closure. Data also presented for Case 1 in Table 1 include the actuator torque available and the ratio of torque available to torque required at the corresponding positions. The ratio of excess torque varies from 7.0 to 14.5 for all opening positions.

Table 1 - Case 2 provides data for the worst possible case wherein the valve closure is assumed to be delayed until the peak calculated contain-ment pressure of 47.2 psig is attained. This is not a design case, but is provided to demonstrate the large margins available to ensure valve closure. As can be seen from the data provided, excess torque margins of 4.0 to 7.9 exist even for this worst case. As with the design case, all pressure drops are calculated with 47.2 psig at the inlet to the purge line and the redundant valve in the line failed in the open position.

The available actuator torques and the required torque at the valve for each 10 degrees of opening are shown in Figure 5.

Stresses on valve components have been calculated for both cases, and none of the components exceed the allowable stress values reported in the response to Item 6.

The specific maximum AP capability has not been calculated for contain-ment pressures greater than 47.2 psig since these pressures will not exist. However, it can be concluded from the excess torque available ratios and the stress analyses that the valves would close at much greater differential pressures than those to which the valves could be exposed.

e.

10

=_ -. _- -.-_---.------ -; . _ - .

. 1

6. What Code, standards or other criteria, was the valve designed to? What are the stress allowables (tension, shear, torsion, etc.) used for critical elements such as disc, pins, shaft yoke, etc. in the valve assembly? What' load combinations were used?

RESPONSE

TypI9220 butterfly.valvesaredesignedaccordingtotheASMEBoilerand Pressure Vessel Code, Sections III and VIII. Allowable stresses are also ,

taken from the ASME Boiler and Pressure Vessel Code. Loads considered in the design of these valves include all typical pressure- and flow-induced loads. Worst case load combinations were used. The design is compatible with ANSI Class 150 pressure / temperature ratings. These valve bodies are 4 8WE x flanged style to mate with ANSI Class 150 (B16.5) raised-face flanges.

A summary of critical loading is presented as follows:

)

Calculated Stresses, ksi Allowable

• Case 1tt Case 2tt Stress ** at 70 at 70 Stress Consideration ksi Clovedt Degrees Degrees

1. Shaft at disc hub (1.55) 52.5 4.7 2.3 4.3 (bending and torsion)

2. Shaft at disc hub (0.755) 26.25 6.1 3.1 5.7 (torsion and transverse shear)

3. Shaft at pin connection (0.75S) 26.25 5.8 3.7 6.3

4. Shaft at key connection (0.755) 26.25 5.3 3.1 5.4

5. Bushing 8.59 1.9 0.7 1.5 NOTES

• These allowables were derated to 98 percent of that shown to account for the 320 F design temperature.

• • Based on ASME Code Section III values (S) of 35 ksi for 17-4PH, Condi-tion H1100 (Table I-7.1, Appendix I). The allowable stress of 35 ksi is a conservative figure since an (S) value of 36.2 ksi is allowed by the code for H1075 shaft material.

TClosed position stresses based on 60 psid across valve.

ttRefer to Table 1 for the definitions for cases 1 and 2. Stresses are reported for 70 degrees open since they are the maximum for opening angles 10 through 90 degrees.

l l IGraphite-filled bronze.

The shaft is considered to be the most critical valve component under most conditions, since the pins and keys are selected to b6 stronger than the shaft. Therefore, separate calculations for pins and keys are not necessary. Stress concentration factors are considered when evaluating 11

- . _ - . . ., . . - - _ _ . = _ :. - - .. - . . . . . . - __ _ . - - . _

RESPONSE TO ITEM 6 CONTINUED the shafts at the pins and keyways. The maximum disc load occurs when the disc is in the closed condition, and acceptable disc strength v& lues have been established based on testing and experience.

As noted in the FSAR, the reactor coolant pressure boundary is Seismic Cakgory I and is not postulated to fail during a seismic event. The LOCA which results Jn the pressure transient shown in Figure 1 is not based on a mechanistic failure but is postulated to occur as required by 10 CFR 50 Appendix A. The stresses reported above are based on (ynamic loadings due to the LOCA pressurization transient. Shaft loadings which result from a seismic event are not specifically calculated or combined with LOCA loadings because the events are independent and not postulated to occur simultaneously. The purge valves are Seismic Category I and have been tested for operability during and after a seismic event.

As can be seen from the LOCA-induced stresses reported above, significant margins exist (factor of 7 minimum) between the stress allowables and the calculated stresses for the design case (Case 1). Seismic loadings on the shaft are expected to be less than the LOCA loadings because the actuator loadings are transmitted to the valve body.

In summary, the valve is designed for independant LOCA and seismic events. However, due to the margins which exist and the expected response of the valva, it is anticipated that the valve would function properly for the nondesign basis case wherein the events would be postu- ,

lated to occur simultaneously.

12

7 and 8. 1hese item numbers were omitted by the NRC.

l-

9. For those valve assemblies (with air operators) inside containment, has the containment pressure rise (beckpressore) 4 been considered as to its effect on torque margins available (to close and seat the valve) from the actuator? During the closure period, air must be vented from the actuators opening side through the solenoid valve into this backpressure.

g* Discuss the installed actuator bleed configuration and provide basis for not considering this backpressure effect a problem on torque margin. Valve assembly using 4 wey solenoid valve should especially be reviewed.

RESPONSE

The subject 18-inch valves are equipped with Bettis spring-return actuators.

The unpressurized side of the piston actuator is vented to the local j

ambient conditions. During valve closure the pressure side is also vented (through the Versa valve and solenoid) to the same local ambient conditions; therefore, no pressure differential will exist across the piston as a result of a surrcunding local pressure rise upon full de-pressurization. The spring will still drive the actuator to the safety-j mode (closed) position and maintain that position.

Figure 4 shows the depressurization transient for the actuator. As shown i therein, the actuator depressurizes very rapidly. Since the unpressurized side is continuously vented to the containment, it will be pressurized along with the containment. There will be no adverse effect on the valve '

closure rate. In fact, the closure rate may increase due to prepressuri-zation of the vented side.

i The solenoid used (ASCO Type NP8320A185V) is a three-way direct-acting solenoid which operates a 3/4-inch three-way pneumatic switching valve (Versa VSP-3601-155H) to minimize restriction and provide a fast closure response. The solenoid is connected in the normally-closed mode so that de-energization will vent the actuator casing to ambient atmosphere thrc Jgh the Versa switching valve.

I f As shown in Figure 5, adequate spring-driven torque output is available i

from the actuator to close the valve from any open position (regardless l

of external ambient pressure). Since both the unpressurized side of the l

piston casing and the solenoid and Versa valves are vented to the same external ambient pressure, there is no effect on valve closure. ,

Note: There are no four-way solenoids or switching valves used in the j% subject valve assembly.

l l

l 13 l

! , , 8. 7 C . _. 2 d., # *,.C**~#**'*' '"'**E # # ~ *

10. Where air operated valve assemblies use accumulators as the fail-safe feature, describe the accumulator air system configuration and its operation. Provide necessary -

1 information to show the adequacy of the accumulator to stroke the valve i.e. sizing and operation starting from lower limits of initial air pressure charge. Discuss active electrical components in the accumulator system, 55 and the hasis used to determine their qualification for the environmental conditions experienced. Is the accumu-lator system seismically designed?

RESPONSE: ,

This item is not applicable to SNUPPS. The closure of the valve is accomplished by spring force, not by air pressure.

i l

l l

I l

l l

i 14

For valve assemblies requiring a seal pressurization 11.

system (inflatable main seal) describe the air pressuri-zation system configuration and operation including means, used to determine that valve closure and seal pressuriza-tion have taken place. Discuss active electrical compo-nents in this system, and the basis used to determine -

their qualification for the environmental condition g experienced. Is this system seismically designed.

For this type valve, has it been determined that the

" valve travel stops" (closed position) are capable of withstanding the loads imposed at closure during the DBA-LOCA conditions.

RESPONSE

This item is not applicable to SNUPPS. Inflatable seals are not used.

l l

O i

l r

15

12. Describe the modification made to the valve assembly to .

limit the opening angle. With this modification, is there -  !

a sufficient torque margin available from the operator to overcome any dynamic torques developed that tend to oppost 4

valve closure, starting from the valve's initial open position? Is there sufficient torque margin available -

from the operator to fully seat the valve. Consider g seating torques required with seats that have been at low ambient temperatures.

RESPONSE

The first part of this item is not applicable to SNUPPS because no modifications have been made to limit the valve open position. Valve seating and ambient temperature effects are addressed as follows.

The purge supply air temperature is tempered to a minimum of 50 F. The exhaust air will be that of the bulk containment (120 F maximum).

Seating torque is considered to be independent of temperature, within the calculated flow temperature range of 50 F initial (minimum) temperature to 231 F maximum fluid temperature during closure. The 320 F containment design temperature would not be experienced by the T-ring seal during the closure transient. The initial effect of a high temperature would be to soften the material, but it would also expand (compensating effects). In any event, however, the actuator spring-driven torque capability at i closure (approximately 29,900 in.-lb) is so much greater than the required seating torque at 60 psid differential (approximately 6220 in.-lb) that complete closure is assured. If it is postulated that seal expansion prevents complete rotation, the disc would still be driven into the seal far enough to provide the required shutoff, regardless of final travel -

position.

t O

6 l

16 l

l l

13. Does the maximum torque developed by the valve dering .

closure exceed the maximum torque rating of the operators? I Could this affect operability? ,

RESPONSE: -

The required torque at the valve and the actuator-supplied torque are sh qn in Figure 5. No concerns exist that could affect valve operability.

l i

~

l 17

14. Has the maximum torque value determined in #13 been found to be compatibia with torque limiting settings where applicable? ,

RESPONSE: ,

1 This item is not applicable to SNUPPS.

r A

• P i

l l

l 18

l l

15. Where electric motor operators are used, has the minimum available voltage to the electric operator under both normal or emergency modes been determined and specified to  !

the operator manufacturer, to assure the adequacy of the ,

operator to stroke the valve at DBA conditions with the.se i lowar limit voltages available. Does this reduced voltage l operation result in any significant change in stroke I as timing? ,, Describe the emergency mode power source used.

RESPONSE

l This item is not applicable to SNUPPS.

t k

t O

19 .

16. Where electric operator units are equipped with handwheels, does their design provide for automatic re-engagement of the motor operator following the handwheel mode of operation?

If not, what steps are taken to preclude the possibility -

of the valve being left in the handwheel mode following' some maintenance, test etc. type operation.

t RES40NSE: ,

This item is not applicable to SNUPPS.

L

\

\

i I

20 m mom-w_1 _ La_- or _ _ c o_ - - o

17. Describe the tests and/or analysis performed to establish the qualification of the valve to perform its intended I function under the environmental conditions exposed to during and after the DBA following its long term exposuee' to the normal plant environment.

18. What basis is used to establish the qualification of the sg valve, operators, solenoids, valves? How was the valve assembly-(valve / operators) seismically qualified (test, analysis,etc.)?

19. Where testing was accomplished, describe the type tests performed conditions used etc. Tests (where applicable) such as flow tests, aging simulation (thermal, radiation, wear, vibration endurance, seismic) LOCA-DBA environment (radiation, steam, chemicals) should be pointed out.

20. Where analysis was used, provide the rationale used to reach the decision that analysis could be used in lieu of testing. Discuss conditions, assumptions, other test data, handbook data, and classical problems as they may apply. -

21. Have the preventive maintenance instructions (part replacement, lubrication, periodic cycling, etc.) established by the manufacturer been reviewed, and are they being folleved?

Consideration should especially be given to elastome?ic l components in valve body, operators, solenoids, etc. where l this hardware is installed inside containment.

l 1

RESPONSE TO ITEMS 17 THROUGH 21:

The subject valves have been qualified in accordance with NUREG 0588 and the requirements of IEEE 323-1974. The qualification programs have been fully described in the SNUPPS submittals to the NRC (Letter SLNRC 83-0015 dated March 10,1983), which are contained in the two-volume set entitled

" Report of Independent Review of Environmental Qualification Programs to NUREG 0588." The specific section of that report which addresses the purge valves is found in Volume 2 under Specification M-237.

21

~

22. Assess the structural capability of any ducting or piping in the purge system which is upstream or downstream of the ,

valves and is exposed to the flow condition associated ,

with the LOCA and seismic event. In partict'ar, consider the effects of loose debris from the pipe or duct system on the closure capability of these valves. -

RE1[0NSE: ,

The valves are located in piping which is ASME Section III, Class 2. The boundary of this piping is extended to seismic anchors as recommended by Position C.3 of Regulatory- Guide 1.29.

As shown on FSAR Figure 9.4-6, Sheet 4, debris screens are provided on the containment side of the mini purge supply and exhaust isolation valves to prevent the entry of lightweight debris, which could preclude tight valve closure. The piping which contains the screens is ANSI B31.1 (150 pound design pressure) piping, which is seismically analyzed in accordance with Position C.3 of Regulatory Guide 1.29. The screens are located approximately two pipe diameters away from the isolation valves and are inherently designed to withstand post-LOCA differential pressures due to their rugged design and the negligible pressure drop through the screen material (No. 2 mesh, .063-inch wire with a 76.4 percent free area). The screen material is welded over the 17-inch-diameter opening in a 1/4-inch-thick flange which is bolted into place.

The purge isolation valves and debris screens are located adjacent to the containment wall, outside of the secondary shield walls, and are protec-ted from missiles which could be postulated following a LOCA. Also, motor-operated dampers are located one pipe diameter away from the screens on the containment side. These dampers and the connecting piping provide additional protection for the wire mesh screens.

5 1

1 22

TABLE 1 .

. COMPARISON OF TORQUE AVAILABLE TO - ,

TORQUE REQUIRED .

CASE 1 CASE 2 St Closure During Transient

~~

Closure at Peak Design Case Calculated Pressure Torque Torque Angle Torque Maximum

• Required at Ratio Maximum Required at Ratio of Available Predicted AP = 22.9 Avail./ Req. Fredicted AP = 47.2 Avail./ Req.

Opening in.-lb AP in.-lb Torque AP in.-lb Torque Closed 29,900 22.9 6,220** 4.8t 47.2 6,220** 4.8t

, 10 25,600 22.9 1,785 14.5 47.2 3,239 7.9 20 23,000 22.8 2,223 10.4 47.2 3,961 5.8 30 21,700 22.7 2,223 9.8 46.8 3,961 5.5 40 21,400 22.2 2,223 9.6 45.8 3,961 5.4 50 22,100 21.0 2,728 8.0 44.0 4,791 4.6 60 .23,900 18.2 3,412 7.0 39.2 5,918 4.0 70 27,200 12.8 3,662 7.4 30.8 6,330 4.3 80 33,000 7.3 3,581 9.2 16.1 6,197 5.3 90 43,000 4.3 3,581 32.0 9.9 6,197 6.9

• During the 3-second closure period, the containment pressure rises from 14 psig to 22.9 psig. All predicted APs are based on flow conditions at 22.9 psig at the inist to the purge piping.

• • The 6,220 in.-lb torque required based on 60 psig, which is the containment /

valve design pressure. The maximum calculated LOCA pressure is 47.2 psig.

tSee Figure 5.

-v- , -- -

, -- =.--,-m --y .= -

s- . v.---.--*,y y -y-, - . _ - -- yw-%w3 e ,---y--w----w,--, ,--,w----w-w w-----g-wi+--,,, - - - - , , , ,

llill ,il1l j,l j l1 1l lj l

l ll1 7

0 1

e i

m T

dg

? .

ea iL e

l l ie ,60 l

nn ig 1

l i as

- ngt e od s ee i

ne no o c i

l e ey skl e ,50 i

er ot I

stsl ep 1 n en t

ee neo mgc o n

isis H es ee T nhh I

oee ,40 W c. v. v. 1 T s 12s NS d n

EE o f SI M I

c NT e S

AE 0 RR 0 ,30 )

TU 1 1 i

, ES -

RO e s  :

UL lo d n

SC e o C+

SE p c e EV RL 2

S e s m PA 7 4 6 d n

,20 T i

1 TV e o NE c g e EG is S MR p 3 d n

NU 9 eg o A P c I

2 2 e 1 T i s S ,0 N p 1 1 O s C ,

4 1 e

- 3 ig s

p 6

< ,00 1

1 0

1 t

p 2

- - - - - - 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 M.

4 6 8 8 8 0 2 1 -

4 4 3 2

• . . ! ^ s j-!4. $;. : - -.s ,

a * .

l !fl

I Figure 2 (Imeet1 of 2) ,

LOCATION OF VALVES IN CONTAINMENT PURGE SYSTEM (Supply Line)

\ ,

GT HZ 06 (Closed) g

's Rotatio

\ Counterclockwise

- to Close GT-HZ 04 - ,

GT HZ 07 Flow of 80 -

(Closed) air During ',

- Accident s

NO SCALE j ContainmentWall i Damper-N-r GT HZ 05

' '\ Rotation Counterclockwise to Close Supply ss, Air

Scale NO SCALE 1" s 10' 18" Valves GT-HZ 04,05 I '

Flowof Air During Accident 36" Valves GT HZ 06,07 AirSupply to

' f Containment During Modes 5 & 6

. __,__y--..r..-

._ ~-~___ ___._..________._____--__ __

Figure 3 ,

(Sheet 2 of 4 LOCATION OF VALVES IN CONTAINMENT PURGE SYSTEM '

g ,

(Exhaust Line) 4

~~'

Rotation Counterclockwise to Close NO lowof Air SCALE During Accident GT-HZ 12 Damper

[ l_1 l Dt Normal ust GT HZ-09 l

(Closed)

}GT HZ 11 18" Valves j GT HZ 11,12 ,

36" Valves <' GT HZG GT HZ 06,0g (Closed)

- ev*

ContainmentWall r f

% 4_

E' Du kg Acc dont

+

1. Rotation /

1" W 10' Counterclockwise -

to Close NO 8CALE

Figure 3 ACTUATOR DATA-SCOTCH YOKE l l

100 - .

90 -

EFFICIENCY PLOT End

\ Efficieracy vs Angle 80 -

Brea_k  ; '

80 ,  :  :  :

0 10 20 30 40 50 00 70 80 90 75,000 67,500- -

60,000 - -

90 psi 52,500 - -

I -

4 80 psi si 45,000 -

e

• 37,500 . A/r4//oder orque T 70 pal Spring Extended B 30,000 -

I 22,500 Ofl00 Torque f 60 psi

~

15,000 - -

S 7,500 - -

- Valve Closed 0

60 70 80 90 O 10 20 30 40 50 ScotchYoke Arm Angle ,

t

e O

e O

e,

\ .

Figure 4 ACTUATOR DEPRESSURIZATION TRANSIENT 100 b

a0 -

D 40 -

[ m -

(

~

m

' ' * * ' ' " ' " ^ *

• 0 5 5 5 5 5 3 3 0 .2 .4 .6 . 8 1.0 1.2 1.4 l - Time (seconds) l

--,c -_

Figure 5 -

COMPARISON OF AVAILABLE ACTUATOR TORQUE '

TO REQUIRED TORQUE AT VALVE k ..

45,000 - -

40,000 - -

~~

TORQUEAVAILABLE FROM ACTUATOR 30,000 -

25,000 - -

i f 20,000 - -

15,000 - -

10,000 - -

MAXIMUM DYNAMICTORQUES REQUIRED p = 47.2 psig - Case 2

[ Static Op = 60psig f 5,000 - Ap = 22.9 psig - Case 1 7

sf f

• / f f f i a a R A A A 8 0

10 20 30 40 50 60 70 80 90 Valve Opening Angles 4

0

l ATTACHMENT 1 to PURGE VALVE OPERABILITY REPORT ISA Transactions, Vol. 8, No. 4, 1969

" Effects of Fluid Compressibility on Torque in Butterfly Valves" by Floyd P. Harthun

t ,-

, .N -

. . v__o_ s __. .

~

(i. .

~

Effect of Fluid ~

Compressibility on

~

Torque in Butterfly Valves

• FLOYD P. HARTHUNt Fider Governor Company -

Mars'.alltown, low

~

> A technsque is presented try which the shaft torque resulting from fluid flow through butterfly valves con be determined with reasonable accuracy for both compressitWe end incompressable flow. First. the gemrol torque relationship for mcompressable flow is

(% -

established. Then, an s'fective pressure diffe-ential is defined to extend this relationshe so inc:ude the effect of fluid compressibility. The application of this technious showed very good agreement with expenmental test results.

INTRODUCTION . DEVELOPMENT OF GENERAL TORQUE

• THE APPUCADON of butter $y valves in various automatic . .

RELATIONSHIP ,

control systems requires proper actuator sizing for The total shaft torque required to operate butter 8y ef!!:ient control Rus, a thorough knowledge of the valves can be separated into two major components:

Suid reaction forces acting on the valve disc is required. . i Extensive experimental work

• has been performed in 1. Dynamic torque -that portion of the total the past to establish a relationship to determine these operanns torque attn'butabic to the fluid reaction forces and thus determine the resultant shah torque.The force of the Bowing medium acting on the valve general form of this relationship has been eusblished and

• disc.

con 6rmed. However, by using the classical Guid momen- 2. Friction 'i-gi ;-4at portion of the total tum approach, a similar relationship can be obtamed in operating torque attributable to friction in the which the torque is shown to be directly proportional to packing and W-=

the measured valve pressure differential for a given disc Sm.ee each of these components a.m. dependent of the position. This rehtionship along with met of the other, a separate evaluanon of each component aficeds

• pretiously published torque information is adequate for the best approach to this problem. This invesognuon as ancompressible Sow. A1: hough the effect of Suid coen- limited to an evaluation of the dynamic torque com-pressibility on torque has been recognued. no usefbl relaW has been hoped, ne prunary ob- Ponent. If the fnction on the valve shaft as assumed to be joctive of this i:.vestigation is to extend the established ,andependent of direcuon of rotauon,it can be readily molated. He torque required to rotate time valve disc as

-torque relationship to include the effect of fluid com- . measured in a clockwise and a counterclockwise direc-bili'I' **" tion through full travel. Since friction always opposes

• Pnessted as the 1968 ISA Ancual Coefenace: mead Augum.19es. motion the difference between these values will be twice -

l (. * ~

tResearch Engisser. ,the actual shaft friction.

- 381 ISA Trennections Yol8.No.4 -

l l .. .

t _. . _ _-. . - _ _ _ -- _ . - - - - . _ . .

. *9 .

. The dynamic torque fo butter 8y valves is a Anaction Combning Equations (7), (3), 'and (9) of the duid reaction forces acting on the valve disc. It i

% would be dd5 cult to determine these forces by purely 7, = #8:#2 88 #'*O'A# (10) analytical techniques. Expenmental determmation of the 4 -

pressures and velocity pro 81es is the immediate area of or the disc would also be quite dif5 cult. However.afa control volume is selected so the boundanes are points of known Ta " K:D'AF, (10.A) pressure and velocity, an analysis of these forces can be g

• ~

made from the change in Suid momentum through this sonfFoi volume. .,

K = 3,s3,g,3,, 7,

- (10.B)

INCOMPRESSISLE FLOW i

1. = #A#

An expression for dynamic torque is developed Equati,on (10.B)is de8ned as the dimensionless torque assuming incoinpr:ssible Sow.This torque is a Anac' ion coef5cient which can be determined expenmentally from of the fluid reaction force, r, and a moment arm. D,which tests conducted with incompressible flow.

is a characteristic d'-a*ian of the valve disc.

7, - f(F,D) (1) COMPREssisLE Plow Using the fluid momentum approach, the force, T. is The dynamic torque for butter 8y valves is proportional given by: to the mass flow rate and velocity change through a selected control volume for both compressible and

. F = MAV (2) incompressible flow (i.e To cc MAV). Therefore, the where approach used to obtain an expression for this tor + e assuming incompressible Sow can be extended to I = sum of external forces acting on Said compressible Sow by re-de8ning these two variables.

M = mass Sow rate First, assume that the velocity at the valve disc, V,,

AV = Suid velocity change through the control is proportional to the velocity change through the control volume volume. Then, the dynamu tosque can be expressed as ne mass flow rate, M,is given by Te oc MV, (11)

{S ,

M = pAV By using a proportionality constant,5 3, the asas Sow (3) The velocity at the valve disc is given by V, = -y j rate can also be de8aed as -.

(12)  ;

- p,A M = # A(PAP)8'8 i (4)

By combining Equations (11) and (12) the dynamse Equations (31and(4)are combm.ed to obtain the fotbw*

torque is shown to vary directly as the square of the mass ing expression for Suid velocity:

. Sow rate and inversely with the duid density at the valve V = B (AP/p)U8 i (5) disc. l The velocity change through the control volume, AV, M8  !

In Equation (2) can be expressed in terms of the velocity To oc (13) 1 l

l at the valve disc by use of a proportionality constant,3 2, Determinmg the Sow rate of. a compressible Suid <

l T- AsMV (6) through a control valve by analytical techniques is quite By substituting the expressions for mass Sow rate dif5 cult because of valve geometry. The major oroblem as to establish the pressure dmerential between the valv: )

Equation (4) and !!uid velocity Equation (5)into Equa-tion (6) the force on the valve discis elet and the vena contracta. However, by de8ains the I physical system in which the valve is installed to conform l l 7 = 3:85sAAP (7) with For a given valve size, the flow area, A. for any angle of (FCI)p8empirical cations given by the relatiships Fluid Controls developed =='My Institute

. disc mtion,8, can be wntten as for determmms Sow rate for control valves, can be considered. Several such ampincal relationships have nos been developed; however. anly one, the Universal Gas d = #e 7 (8) Sizing Equation,888 has been shown to accurately de4ne the Sow rate for any valve configuration. Dis equation

.The force, F, acts upon a moment arm which is a function is given by *

• of the disc d6-~,D. Now, the dynamic torque can be ~- ,

written as 520 "$9.64 AP 1

-3 Q= -P:C CsC,sm. -

(14) I g To= # 3FD (9) GT _Ci Cs P, i

l ISA Transactions . Vol.8.No.4 282 W

Equation (14) can be rewritten to obtain an equivalent incompressible Sow.

, espression for mass Sow rate, 8

\4 ,, T = K D8P C'C 8~ sin8 8 i (24)

$9.64 AP. - 59.64, M = 1.06/p P:CaC C,sia

,CaC { .e II For cetwanience the form of Equation (24)is simpli6ed.

7. - K D AP, 8 (25)

The sine function in Equations (14) and (15) is used to

'where deine the transition between incompressible Sow ca:ar. -

8 ring at leer pressure ratios (AP/Pa) and critical flow.

• • Ap, p, C*Cs* ,;,3 8 Q6)

Let 59.64, Equation (26) is deined as the pressure differential 8" ~59.64 AP (16) contributing to the dynamic torque on butter $y valves

- CaC8 P8 "*

with conditions of compressible Sow.

Rewriting Equation (15)in the foUowing mannar:

EXPERIMENTAL RESULTS M = 1.06/pi P :C CaC.F (17)

De factor, F,is bounded by the foDowing: Arst &c ex@ a n was to establish the dimensionless torqus coefEcient. .i K , as .a F = sin 9 for 8 < z/2 Ametion of valve disc rotation as defined by Equation (18) (10.B1 A test was conducted on a 4.in. valve ander the F = 1.0 for 8 m s/2 following coutroUed conditions:

By substituting Equation (17) for the mass Sow rate in

1. He valve was instaDed in a 4.in. test line with a E,quation (13), the dynamic torque for a given valve is minimum of 12 pipe diameters of straight pipe given by upstream.

pi Pi(C C sin es i

.. 2. He pressure taps were located accord.sg to FCI ~'

Te oc (19) speci6 cations and attached to the test line

1. 4 according to tions in the ASME Power-The only parameter in Equation that cannot be mid Test Code.

." readily obtained is the density at e valve disc, p,. 3. Water at ambient temperature was used as the

(- Assuming that the change in the ratio of fluid density at thi valve inlet to fluid density at the valve disc with Scwing medium.

4. He inlet pressure and outlet pressure were held increasing pressure ratio is small relative to the total constant.

change in mass flow rate, the torque expression can be i ne test was conducted at a low pressure ratio simpli5ed in the following =maner: (AP/P = 0.088) to ensure ir. compressible flow.

7, ac 7 (CaC sin 8)8 (20) g Therefore,for compressible flow: -

Ta = K:Pa(CaC sin 8)8 (21)

For small values of pressure ratio (AP/P ) Equation (21) 28 reduces to the incompressible torque relationship given .

by Equation (10.A), ,

24 l As AP/P =0 20 $

sin 8 = 8(radw - s

/ \

T = K (59.64)*AP (22)

[ {

g$12 e

De expression in Equation (2".) is equivalen: to the espression in Equation (10.A):

a g 1 [ }

N (59.64)8AP = K Dagp if

/, j K D*

i AL -

K = 99g (23) 0

~~

~

0 10 20 30 40 50 60 70 80 90 By substitut'ing the expression i:: Equation (23) for the yg 1 coef5cient K in Equation (21), a general expression for -

/% dynamic torque for compressible flow is obtained using pie.,. s. os n.i.ns e.,gv co mes nt. e.in. hm.cny j the dimensionless torque coefEcient established for ve:v. :.u --mma flemn p = 100 p.s , ep = .sto p.L

' 283 ISA Transactions . Vol.8, bfo.4 - W*Y" E Y i

- 1 I

\  !

. 3. .

= ix

,\ 4'

• - E MTA

• d *~ REST MTA - WATER

'#0 500 egkeg $9 ^

[ fy b

'M m .-

g% 5 z.

uT om -AR y4 SC 320

/

\

A-A {

(g 240 i

• } \

40 16 0 ,

[ l

\ A m o \

ge 5 l 1

0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90 l ANGLE OF mtATION. DEGREES ANGLE OF ROTATION. DEGREES Fleiste 2. Dynamie toeepee vs. enete of dies rotation,4 in. Fleure 4. Dynamic torque vs. emele of dies rotation. S.in.

buteerfty vesve, e meeriesa of m -- _ r_: reevise wish butterfly weave, som,wisen et emeer. nonces res.ece ween i

essoussied sergue, insempreessmae now: p, = too peas, essavanced sorgue, ineemprensibe now: p, = 1oo peas, ar e s pai, ar = s pen.

Torque measurements were made at selected incre- ment between measured torque and the torque calculated ments of disc rotation (0-90') A transducer, consistag using this ea*8te=ar of a steel bar with strain sages attached, was Axed to the The next step was to verify tha,t the torque coefBeiettt

, valve shaft and usedia conjunction with an oscillograph is indeed applicable to other valve sians provided geo-( to measure and record the shaft torque. The data from. metric sunitarity is reasonably well maintamed. The this test were used to determine the dimensionless torque results on Figures 3 and 4 again show very good agree-er=8Icient plotted as a function ofdisc rotation on Figure ment between measured torque and calculated torque

) 1.%e curves plotted on Figure 2 show excellent agree- for two 8-in. valves. -

720 . 36G 44C 320 e-k:%DM [

• -45T 3ATA

[g 56C 28C ogfoE5N81E l

W asc

%3D [ g 24 rL6w

l. I __ .

a-1EST %TA, [- ( Ic -f yWW 32C I

/ \ 16 0

/')

'24C ( 3 12 0 ,

// -

16G Y GC

. o f f J '*

0

/

0 10 20 30 40 50 00 70 80 90 0 10 20 30 40 50 60 70 80 90

. ANGLE OF ROTATION.0EGREES VALVE PitE55URE uitOP& J51 M s. Dynamse sorgue vs. ene s e et dies reesesen, un. 'N L Dynamis sorgue vs. veeve drop, Ma.

tweeerny weeve, semperieen et en - - n resuets ween tutterny vesve. es* dies reteesen. oomportes et esser 6-

. endeuteted sorgue, ineempressabler Sews 7, = 100 pode, mentet reensets wetti sofouleend torque, sempreselbie neur:

ap = s pas. p, = at4.4 peia, newene medawn - eer.

t . ISA Transeersonr . YoL 8. No. 4 234

+ _ _, _

~ ~

s .

9C w -

t

% E W -

e-TEST DATA e-TEST DATA

' {

7C 32e g#

• • Ni NN A s-r=K./AP(INCOMPRESSIBLE)

'o l

5 2R 00

/ A- g24C o4-s* /

I a

I m, r-

'O f .

16 0 l

20

^

J, l 12 0

^2 ^

1C SC Af 0 10 20 30 40 50 60 70 30 90 ANGLE OF ROTAT)ON-DEGREE 5

% s, w ,,,,,, ,,,.e ,g g 0

g fl 0 10 20 30 40 50 60 70 80 90

,,,eseassi. 4-h.

hettediv weave. eemporisen er esperimentes reenste we*

emieutetad teneve. sempresense flow: P. = 114.4 peia. ANGLE OF ROTATION. DEGREES AP = 5 pel ( AP/7, e e.c444). flowing sucessiues e air. g, g, , g Alestierfly vefwe eemperiese of test results witsi endouiseed It should be noted 'that discs in the two 3.in. valves sorgue, iaise view: P, e es.4 pese, ar = so not were of substantially ddrerent geometnc shape. Using ('#l#8 (*'**' **I * wine medium a ear.

the ratio ofdisc diameterto hub diameter as an indicator, 44ds(e these ratios were (56:1 and 3.55:1 for the valves used to similar to the disc in the 4.in. test valve used to estabbsh obtain the data for Figures 3 and 4. respectively. De the torque ew.K.

(- difference in torque magnitude for these valves with a The extendon of the dynamic torque relationship to 5 psi pressure differential shown in Figures 3 and 4is include the effect of 8uid compressibility is accumplished

! the result of this difference in geometry. The disc in the by de8ning an effective pressure differential as shown in

8-in. valve used for the test in Figure 3 was ,; 11y Equation (25). The curves on Figure 5 shew the transmon l . Ikom incompressible Sow to critical Sow with increas.ng g pressure ratio for a 4-in. valve set at 60* disc rotation.

Here again there is very good agreement between the 16 0 torque calculated using Equation (24) and the experi-mental results. The incompressible torque curve is also 14 0

• ~. K i AP shown on Figure 5 to emphasize the eficct of Suid

• -T=%faP W i A compressibility.

d12C #1 The curves on Figures 6 through 3 are presented to 3 *- TEST DATA compare experunental results with torque calculated 100 i using Equatioe (24) for fhli 90* disc rotation. At low 1 pressure ratios, the torque using air as the Sowag

• - medium is essentially equal to the torque for incom-80 pressible Sow (Figure 6). As the pressure ratio is increased.

60 ' '

i the effect,of Suid compressibility becomes more pro-nounced as shown in Figure 7. Onde critical Bow has

1. 40 been attamed, no Ibrther increase in torque is realized by increasing the valve pressure differential as shown on J Figure 8. - -

20 1 1 0" coactussoms ,.

0 10 20 30 40 50 60 70 80 90 ANGLE OF ROTATION. DEGREE 3 A techniqueis presented which can be used to determine

% 7 ,,,,,,,,g m,,,, g m the dynamic torque for butter 8y valves, with reasonsble .

, accuracy. The basic , torque relationship developed for ,

%,.sened ess sorgue. _ n view p, e es.4 ,sen, incompressible Sow is extended to include the efr ec t of -

p = 1e poi (aP/P, m(,332 flowing seedlues e air. 8uid compressibility.The method presented is developed - .

A .

SS5 ISA Trecamersan/Ss t . Vol. 8. No. 4 - i l

e i

.l

+ .

~

. using the Universal Gas Sizing Equation to de6ae an ar = Premme hdalassains dynamss w efective pressure diferential for the transmon Grom & = m remiscompras bas sud sai

,O i - '6i a - < -i'i i a . ^=-r- <- or iki- c - - - - a-

~^

shows excellent agreennent with experunestal 7 '"'N m -

r - Fi.d %. o t* . . si - Fleus denary at sparsam pramme me, ms NOTATIOd6 A, = Fluid demary at valve dies,t/ia?

A = Fles asen,in,s t 3,, Bs ,

Bs,a, = Constama etproperumailty REFERENCES C, = C,IC, Correction dissor for venense is synods hemt rease I. Kasar. L C. and Sahmana. I. F. January 1936. Aerodynamse Gas sizing coedloset .. Model Tests on Butterey Valves." E.scArr.Wyer News. 9.

- Flow coeSasst .

2. Aersainernded Ve&sitary Standardt/sr M.-- _. -.. huesdne/pr p Noenmal valve danseur.hk C. . - _, Ceneret Vaka Flow Capearp.1938. phal Cesareis

= Fores.Ib Institues. fac paper FCI $8 2.

G = Specios pavity . 3. Buresh. J. F, and Schador. C. B. October 1964. "The C..

K, = Dunsasionless sortius sesSuset of a Uanversal Gas Simms Equatson ist Contrei Valves.' Ed Tauw.

M = Mass Sow rass.Ih/s 3:322-328.

P. = !alet pressure, psia 4. Flow M-- _ _ _ : Aurruneau sad Apparerma. L, L _ __ a de ar = Valve prus durarenant pai ASMrre T,sr Codes. ASME report FTC 19J:4 1959.

.l o

(.

~

l l

l

~. .

a-O.. .

ISA Trotsactions . Vol.8,.Vo.4 288 i

e

• • a.

• 6* m e - e- = em e * . .

_ _ _ _ _ _ e.eOe me _

w

\ .

f ATTAC1 MENT 2 to PURGE VALVE OPERABILITY REPORT Torque - Pressure Drop Relationships used in Butterfly Valve Dynamic Torque Coefficient Selection

a. Selection Figure 1

b. Selection Figure 2

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

4 m-

, t i 3

. . I..

g E

I "T"~1 i

f a

I 4b J L 2 m  ! ._ ,

I

• F 4n= > a., 'J 4 l' em e m A.L1_ k i Jr _

y.. K M"J t 8 =gi - 1  ! F "'.E

'E E '

121J L P 1. '. < , ., F jr a -gr ima i= = = = -

• I .m

. F TW l 5 ...,

___i.. aw

"' "' "M in F

, fi / _ '4* I -i__,,,,, M fp=

% I # C' " ' l ' h I'P l*d **8

• L"^

3 -' _ ' '

' I A kh a I"f* li l' E. T i  !

• i t '

!g-15'. 4 7 timmel r i ,-

at d ' k

-f J

j e ./'

3-i-s

,1 ,

8 4 1  ! l l iq e i -- g , ,

e y 1*(- 7--

i, -. -m -- , , -- --. ; y r- --

, y , , -

y gI . 1 I 'I' ' I iI mr i .

1

. if I II V't i i e i 1 U8 1 7 I #1 & I i I i

! I t t I I . 2 F 6 Yt *  ! ii e e

! e I t t . I !I. I 3 # 1 A: 1 i il I 6 I +4 J J *  !\ i e It _8 I i 3 I I e i 8 I i i

' . . .i e F: 1 E r T _

i _ f.l, _ __N

@ ' s jra 78 ' ' \ -I_ -

-11T K._

AIC i e a en I ;i /

f'I gi i I I

JF a

I 2

F 2 F 3 ii

.1 i i i IiI IIt I i

_I f g , / .

g -lD Olb,

/ '

I l 1 1 __ '. _t i  ;

A I ,F i 1*'au <a I 'u fWL' s

'2r 2FI l l .

1. ' K 1

[

,f e f ' i t # 4 / I i

- J lar* '

I I I I- I l i

2F fi 81

1. / I' 1 e t . Ji J I i l I JiI 1 '/

1. I / a I

l I

)

! .rit I 1 3 J ;P _  !

J f E i l

. A4 n

if l '

II I HA' In 34 0 P

r.

M r

I I

i .

r us. J i Etib>

=

JkRa i.

[b '

ihP _

_dP J,d4 m .l 4 1'Ab4 m a

'puu, d p Esbl". m L,dLJE EE Ifk J..3Ek R d.Ih,Phr -

a> s 'l

__ m _ . _ _. 4 . .e== i

- 1 . E a

t. e Emed 1.L L MLt JL

h. .. --

J 36 t *m,aim -

I I,,I ga . ad I rk Hir_

,'L a

<1 y

!h E ,'l 9% ===4..6 m. ,-,

-.__- - a' - - - -

em- * - i. m -

, --- -- = a avp- anin g64pra asaAlias e L. REiT 1E @J'Wh Affi * '"# i T4d't i RJ k OME'E4 -

A4Am ik

_r. . _ _ _ a g db e

/

t i

i s

t # I . N e f i i 1 e i l'u

---,--wm e-w-mw , -,eee.-_- -,ew-gm ,--,,,n.,--.n-y- . - - - - , - , - - - ,---,_.-=m---,-m,m,e.m.,. .m-,--.n- 7-,,,-w- -w,,,m,wr ,,m-v--we , , m ye w w y - -y. - - --g--~

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

e O i e I ' f 1 I 1 i i e

f t i I 4 I i

el .

?

8 t 3  ; .

l 4 fn e 3 + t i i I 4 i l e e 3 i ,

a j ,

I  !

e I i 1 i y __ ___a , L-1,. j i i. j - - ,.w j,7 y I ;L I I F EL 3

's iL. .- M . m', 4 1_ 1i j a i t i J i 3 5

. 1 l t '

f l' .. M + Q Jl l ' .k1=aA I

I i

6 J la la ,

el 6 l g I I i 3 if 'fT e i e aa t t '

• I l t i 4 / ._ E i I e i I I f I s

'A J t l

. e

.4 i 3

/ --

ie I t

• i I 'a I g f ag ,g_, ,

t s e t Mk.1 1 J' sL's , [g J.'a l Plmd I . !  : I i i~.*H+ r > =1 I .. ,dy,., Ir.-

t h. I k 'e i m(W l

l 4 9 g

t i i i ,I l

.-WW.hW,---

68 l 3 3

,- --r ,

I 5 i i l  ? if g i 0 I i 6

.j r- l 1

( t i l1 i f e t t i f I ef i e 1 I t:r e l ,Ag ) jg j

, 4 ,

54 , a s 8 , i , -i~~~ g , ,

O g4 4 i .

4 , 4 9 * */ ^ el_ l t We e t i ! 4 4 g s '

. e i  ! II ie i #+ /i Te; i a y, i i . e

• i s 6 .

• 1 , e 43 i e e i (* i .3, t ! 6 '2Fe t j '

gI i

. . 4 i . . , ! . . , , i rj i es; i f , ,I .i t , ,  ; i g I e e i I I r t i e / i e i ' #1 s , 6 e , i

' i I ( , 6 , e l i e ii

• e I I, ./ - 4 I . . . ea e ie i .

e i . , . . a . 6 , ,

i e i i i e f gf. :e i . i j - . g _q 3 ea . ' _

gr, 7n i

e6 4 a i ' <

/- 4 I**

• I  ! ]_ n yt i

, I s

, a  : Iii f]: s i i . -

, =

l I I ' t 3

i i /t I / 3 i i t i 1 4 '

f dI i / 1 4 3 4 i 8

C i/ '

1. 1 '

zpJag'it n 7

I

's 1 / 1 i lNDrn p ra

't 3

- I I / 3 t 4

'li 4 I t # /1 /i

. t 1

5 Et. , 8 F 4_ J[- t ".'

t l' t

t* t

'M r #'I e i 4 8 s3 ' 4

!I. WP l I e l'! I I i I i ,

if If i

, I

(- e 1 41! a

. e e l 1 t t t /il- e 'i i e

4 t

! i fi f t

• I al e i l  ;) 'I . i 'l

• 4 t / / 1 8 I-I . / / 9 8 I/ 8 I l t 3 f// i I

I 4 II I s If, s (I I r/ i s Jf

(= f) E

n F ,,

E' l s Il F4

. a

.# i I n '

~-

I x J I a e g I d "

. v. J l * .

. me 2 .i t _

t

~~

EW_- /W b [Ll* I w.

ipga 'gi.

J\ Cl u u i 1 1 1 i at le= 4 I a a n 1' (". 3 i L ! T'"A d =

g I.'1h 5 3 E er 11% L m.3 l'% FIm.d b Eup h i I l' ,

t 1 II i i #

! mL J' I_ -. l l *l _ mm l * '"'

9 J 9

- n ,__

't

-is F 12 t, A rt ni rzsrb 1. I raa.e n iarau' t s'Ifa .cm na'do. n

-, -u .

y ms _

y

,i a e a-Tnri tsta l>wm ==<e1*m, h i kd

.h un =1=< m-

- i 4 m=

o -

a --

, 1 . 1_ , , u

, p' ,r.-

U 54!0 alt 1 N2MC fo roud ESA) 21c limr s

.. 1 __..3 . . 3 m

)

, , m m uu. m. 1 5us. 6 [

1 I

f I I f  !

. f f

' I I t {

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