ML17346B200

Forwards Addl Info Re NUREG-0737,Item II.D.1 Re Performance Testing of Relief & Safety Valves,In Response to NRC 850714 Request

ML17346B200
Person / Time
Site:	Turkey Point
Issue date:	06/26/1986
From:	Woody C FLORIDA POWER & LIGHT CO.
To:	Rubenstein L Office of Nuclear Reactor Regulation
References
RTR-NUREG-0737, RTR-NUREG-737, TASK-2.D.1, TASK-TM L-86-241, L-865-241, TAC-4426, TAC-44626, TAC-44627, NUDOCS 8606300263
Download: ML17346B200 (147)

Text

REGULATOR INFORNATIQN DISTRIBUTION ~ 'STEN (RIDS)

ACCESSION NBR: 8606300263 DOC. DATE: 86/06/26 NOTARIZED: NQ DOCNET 0 FACIL: 50-250 Tuv keg Point Planti Unit 3i Florida Power and Light C 05000250 50-251 Turkey Point Planti Unit 4i Flov'ida Power and Light C 0500025i AUTH. MANE AUTHOR AFFILIATION WOODY'. Q. Flov i da Power Zc Light Co.

RECIP. NAi~fE RECIPIENT AFFILIATION RUBENSTEINi L. S. PWR Prospect Directorate 2 4

SUBJECT:

Forwards addi info v e NUREG-,0737'tem II. D. i re pev Formance testing of v el i eF safety valvesi in v'esponse to NRC 8507i4 dc v'equest. /I DISTRIBUTION CODE'046D C PIES RECEIVED: LTR ENCL SIZE:

TITLE: QR Submittal: TNI Action Plan Rgmt NUREC-0737 L. NUREQ-0660 NOTES:

I l REC P ENT COPIES RECIPIENT COPIES ID CODE/NANE LTTR ENCL ID CODE/NAI'IE LTTR ENCL PWR-A ADTS i PWR-A EB i PWR-* EICSB 2 2 PWR-A FOB i i PWR-A PD2 LA 0 PWR-* PD2 PD Oi 5 5 NcDQNALD, D i PWR-A PSB PWR-A RSB INTER NAL: ADt'1/LFNB i 0 ELD/HDS4 i 0 IE/DEPER DIR 33 i IE/DEPER/EPB 3 3 NRR BWR ADTS NRR PAULSONi ll.

NRR PWR-A ADTS NRR PWR-B ADTS i NRR/DHFT NRR/DSRO ESPRIT i F 04 i RQN2 i EXTERNAL: LPDR 03 NRC PDR 02 2 1 NSIC 05 TOTAL NUNBER QF COPIES REGUlRED: LTTR 31 ENCL 28

)I Il 0

hhl

~ ~

~,

j 1 1'er,l'I l

'h P t

~ I 1

't,t I

'll 1 li 1 "'lr 1 i i. r

/g l (

4. ('$

'e h

I II,

'1 Ni'lh t,

II 1

r

,I

(>a ~r P. X 14000, JUNO BEACH, FL 33408 ill/bg Aw. six FLORIDA POWER & LIGHT COMPANY

~g 2 S 18B8 L-86-24 I Office of Nuclear Reactor Regulation Attention: Mr. Lester S. Rubenstein, Director PWR Project Dictorate $72 Division of PWR Licensing - A U. S. Nuclear Regulatory Commission Washington, D.C. 20555

Dear Mr. Rubenstein:

Re: Turkey Point Units 3 and 4 Docket Nos. 50-250 and 50-251 NUREG-0737, Item II.D. I Performance Testing of Relief and Safety Valves-Request for Additional Information NRC TAC Nos. 44626 and 44627 Attached is the information requested in your July l4, l985 letter relating to NUREG-0737, Item II.D. I, Performance Testing of Relief and Safety Valves.

If you have any further questions, please call us.

Very truly yours,

. Wood Group Vi resident Nuclear nergy COW/TCG/cab Attachment cc: Harold F. Reis, Esquire Dr. J. Nelson Grace, NRC Region II SgQ63002b3 05000025~

PDR AgOCK >R F'CG5/00 I /I PEOPLE... SERVING PEOPLE L$

NRC REQUEST FOR ADDITIONALINFORMATlON NUREG-0737, ITEM II.D. I PERFORMANCE TESTING OF RELIEF AND SAFETY VALVES Questions Related To Selection of Transients and Inlet Fluid Conditions:

Question I The Westinghouse valve inlet fluid conditions report stated that liquid discharge through both the safety and power operated relief valves (PORVs) is predicted for a FSAR feedline break event. The Westinghouse report gave expected peak pressure and pressurization rates for some plants having a FSAR feedline analysis. The Turkey Point Units 3 and 0 plants were not included in this list of plants having such a FSAR analysis. Nor does the plant specific submittal address the FSAR feedline break event. NUREG-0737, however, requires analysis of accidents and occurrences referenced in Regulatory Guide l.70, Revision 2, and one of the accidents so required is the feedline break. Provide a discussion on the feedwater line break event either justifying that it does not apply to this plant or identifying the fluid pressure and pressurization rate, fluid temperature, valve flow rate, and time duration for the event. Assure that the fluid conditions were enveloped in the EPRl tests and that the time period of water relief in the EPRI tests was as long as expected at the plant. Demonstrate operability of the safety valves and PORVs for this event and assure that the feedline break event was considered in analyses of the piping system.

Re~onse The feedwater line break event is not a design basis event for Turkey Point. The Westinghouse Owner's Group is considering a project authorization to address this issue for plants that do not have a FSAR feedline break analysis.

TCGS/00 l /2 Q'Q

Question 2 Since the Crosby 4K26 safety valve, which is used at Turkey Point Units 3 and 4, was not tested in the EPRI program, the Crosby 3K6 valve was chosen to be representative of the valves at Turkey Point. Results of the EPRI loop seal tests and steam tests on the 3K6 valve indicate that the test blowdowns well exceeded the design value of 5'. If the blowdowns expected for the plant also exceed 596, the higher blowdowns could cause a rise in pressurizer water level such that water may reach the safety valve inlet line and result in a steam-water flow situation.

Also the pressure might be sufficiently decreased that flashing occurs in the primary loop or the reactor vessel, natural circulation is interrupted, and adequate cooling for decay heat removal is not achieved. Discuss these consequences of higher blowdowns if increased blowdowns are expected.

~Res ense The impact on the plant safety of excessive pressurizer safety valve blowdowns (up to l4%) has been evaluated for Turkey Point Units 3 and 4. The results of this evaluation showed no adverse effects on plant safety.

Safety valve blowdowns in excess of that assumed in the Turkey Point Unit 3 and 4 FSAR will have the following effect on the events in which safety valve actuation occurs:

I. Increased pressurizer water, level during and following the valve blowdown,

2. Lower pressurizer pressure during and following valve blowdown,

3. Increased inventory loss through the valve.

The impact of the increased safety valve blowdowns with respect to the above effects was evaluated for the Turkey Point Units 3 and 4 FSAR events in which the safety valve actuation occurs (i.e., Loss of External Electrical Load and Single Reactor Coolant Pump Locked Rotor).

For the Loss of External Electrical Load event, results from sensitivity analyses-performed for a 4-loop plant were used for the evaluation. These analyses investigated the effects of different blowdown rates on the event. Similar results are expected 'for a 3-loop plant. The results of these analyses showed only marginal increases in pressurizer water volume and the maximum pressurizer water levels were well below the level at which liquid relief would occur. The Turkey Point Units 3 and 4 FSAR analysis results show that a small increase in pressurizer water volume, due to increased safety valve blowdown, would not result in liquid relief. The sensitivity analyses also showed that peak RCS pressures were unaffected by the increased blowdowns. The increased blowdowns did result in lower pressurizer pressure and increased RCS inventory loss, however, these had no adverse impact on the event and adequate decay heat removal was maintained.

For the Single Reactor Coolant Pump Locked Rotor event, increased safety valve blowdowns have little impact. As analyzed and presented in the Turkey Point Units 3 and 4 FSAR, the opening and closing of the safety valve occurs over a short time period (less than 4 seconds). As a result, there is little change in either pressurizer level or RCS inventory. Increased safety valve blowdowns would have no impact on peak pressure, peak clad temperature, or minimum DNBR as these occur prior to the closing of the safety valve..

. TCG5/00 I/3

.Question 3 According to the Westinghouse valve inlet fluid conditions report, operation of the PORV at a predetermined set point pressure is employed to arrest cold overpressure transients in Westinghouse plants. According to this report, the PORVs are expected to operate over a range of steam, steam/water, and water conditions because of the potential presence of a steam bubble in the pressurizer. To assure that the PORVs operate for all cold overpressure events, explain what range of fluid conditions is expected for these types of transients at Turkey Point. Identify the EPRI test data that demonstrate PORV operability for these cases.

With regard to the cold overpressure (COP) event, the maximum temperature and pressure conditions that can be achieved at the PORY inlet coincidently occur for steam bubble operation. Since pressure is normally maintained below the PORV setpoint, the maximum steam and saturated liquid pressure maintained in the pressurizer during startup and shutdown operations in anticipation of the COP event would occur at the PORV setpoint. The attached Figures I 8 2 for Units 3

& 4 respectively show the setpoint curves for PORV opening pressures and temperatures.

EPRI test conditions for the PORVs were chosen based on expected inlet fluid conditions. Tests were limited but designed to confirm operability over a full range of expected inlet conditions. Steam, steam-to-water, and water flow tests were conducted. Results of these tests can be found in EPRI report, EPRI NP-2670-LD, volume F, Table Vll-3. Although steam tests were conducted only at the higher pressures, it is expected that satisfactory operation would also result at the less severe lower pressures. This can be'seen by the successful low pressure (675 psia), low temperature (l05-442 F) water tests.

TCG5/00 I /4

Questions Related to Valve 0 erabilit:

Question 4 The Crosby 4K26 safety valve, which is used at Turkey Point Units 3 and 4, was not tested in the EPRI program. According to the submittal, the Crosby 3K6 valve was chosen to be representative of the Turkey Point valve. The EPRI test results show that this valve had not achieved rated flow at 3% accumulation for loop seal tests at certain ring settings. Provide an evaluation as to whether the plant safety valves will pass rated flow at the ring settings used.

~Rionse EPRI test results did show the 3K6 valve had not achieved rated flow at 3%

accumulation at certain ring settings. In cross referencing EPRI tables, 4-3 and 4-4 in Volume 5 of EPRI Report NP-2770-LD, however, it is found that for steam tests of the 3K6 valve where blowdown was measured to be less than l0%, flow rates of I l9-122% of rated flow at 3% accumulation were reported. The EPRI tables indicate that lower than rated flows occurred at blowdowns greater than l5% for the 3K6 valve. Crosby production tests for the Turkey Point valves indicate 5% blowdown with the "as-shipped" ring settings. Since this is within the range where the 3K6 valve achieved rated flow, rated flow can also be expected at 3% accumulation for Turkey Point Safety Valves.

0 TCG5/00 I /5

Question 5 The submittal does not identify the ring settings to be used on the Crosby 4K26 safety valves or what effect these settings have on valve performance in the Turkey Point installation. Provide the final ring settings selected for the Turkey Point safety valves and explain which EPRI tests on the Crosby 3K6 valve, if any, had equivalent ring settings. Identify the expected blowdowns corresponding to the ring settings used and verify that the valves will perform their safety function at the blowdown, back pressures and fluid conditions occurring at the plant.

Resesonse Ring settings "as-shipped" by Crosby:

Valve Nozzle Rin ~Guide Rin Guide Rin Level 3-RV-55 I A -5 -235 -204 3-RV-55 I B -5 -235 -20 I 3-RV-55 I C -5 -235 -I 90 4-RV-55 I A -5 -230 -2 I 0 4-RV-55 I B -5 -235 -213 4-RV-55 I C -5 -275 -22 I Please note 'that the ring settings given above were measured from the "highest-locked position," as noted in Crosby procedures and in the EPRI report, "Definitions of Key Terms for Safety Valves". Ring settings reported by EPRI were measured from the "level position". The guide ring level position is provided as a reference.

In the past, plant procedures were written such that after maintenance, rings were returned to the "as-found" settings. During the course of several maintenance cycles, this resulted in a drift of a few notches (mostly lower) in the settings of some of the rings. 'These differences are expected to hove negligible impact on valve performance. Regardless, the valves will be reset to the "as-shipped" settings, and applicable procedures revised so that they are consistently returned to these settirigs.

The as-shipped settings were established by a method which includes a steam operational test on each valve by Crosby. Blowdowns measured during these production tests were equal to, or one-half percent less than, five percent for all valves. The Crosby 3K6 valve tests do'ne with "manufacturer's recommended ring position" had ring settings that were established by the same methods. These tests, performed on a 3K6 valve, rather than Turkey Point's 4K26 valve, may come closest of the EPRI tests to approximating how the valves will perform their pressure relief function in safely shutting down the plant.

TCG5/00 I /6

Question 6 During an EPRI loop seal steam-to-water transition test on the Crosby 3K6 valve, the valve fluttered and chattered when the transition to water occurred. The test was terminated after the valve was manually opened to 'stop chattering. Justify that the valve behavior exhibited in this test is not indicative of the performance expected for the Tureky Point Units 3 and 4 valves.

~Res onse A response to Question 6 is not included at this time pending outcome of WOG review of the feedline break event. This is the only event for which steam-to-water flow may be applicable through the safety valves.

TCG5/00! /7

Question 7 NUREG-0737, Item II.D. I. requires that the plant-specific PORV control circuitry be qualified for design-basis transients and accidents. Provide information which demonstrates that this requirement has been fulfilled.

~Res ense Environmental Criteria The following Nuclear Safety Related electrical equipment, required for PORV operation and/or monitoring, have been determined to be within the scope of I 0 CFR 50.49 and have been qualified for the environmental conditions under which the equipment must function:

MOV-535, 536 Limit Switches for PORV's 456 and 455C Pressure Transmitters PT-455, 456, 457 Pressure Transmitters PT-403, 405 Temperature Element TE-430, 423A PORV Accoustic Flow Monitors ZT/ZS 6303 A, 6303 B,6303 C Electrical Cables Electrical Penetrations Electrical Splice Bits Conax Miniseals This equipment can be found listed in the Turkey Point Units 3 and 4 "Environmental, Qualification (EQ) list for IO CFR 50.49". Complete records of the environmental requirements and qualifications can be found in the applicable Environmental Qualification documentation packages.

All other electrical components for indications and/or control, are determined to be outside the scope of IO CFR 50.49 and no special environmental documentation is required.

These components have been installed taking into consideration their specification and the available environmental data to assure the adequacy of the installation for the specified environmental service.

(

Seismic Criteria The PORV's, PORV's Motor Operated block valves and support equipment and systems are classified as Class I structures. Class I structures, systems, equipment and their associated supports, enclosures, piping, wiring, controls, power sources and switch gears are designed to withstand the maximum hypothetical earthquake loads simultaneously with other applicable loads, as given in the Updated Final Safety Analysis Report, Appendix 5A.

Design controls are in place to insure that the seismic integrity of equipment presently installed is maintained.

TCG5/00 I /8

1 Question 8 Bending moments are induced on the safety valves and PORV's during the time they are required to operate because of discharge loads and thermal expansion of the pressurizer tank and inlet piping. Make a comparison between the predicated plant moments with the moments applied to the tested valves to demonstrate that the operability of the valves will not be impaired.

Response

From the PLAST analysis (see ref. I) the results of the modelled run indicate plasticity in the first elbow downstream of the pressurizer and the elbow below the SRV.. The maximum moments computed in the system through the first 70 milliseconds response are shown in the attached Table I. The table also shows the time at which the maxirnurn moment occurs. This table also indicates the values of 70 percent of the ultimate moment carrying capacity of a pipe of this size.

Refer to the response to question l7 for an explanation of the meaning of this term.

In order to compare the predicted plant moments with the moments applied to the tested valves, the following information should be noted. The safety relief valves.

at Turkey Points Units 3 and 4 are Crosby HB-BP-86 type 4K26. Assembly No.

5I249. The seating material is Stellite 6B and the disc holder is stainless steel with stellite lands and stellite disc bushing. Three Crosby SRV's were tested by the EPRI test program (reference 2,3). These were HB-BP-86 Types 3K6, 6M6 and 6NB. The type 3K6 is chosen as the representative test valve for Turkey Point due to its orifice size and corresponding flowrate. For the tests shown on pg. 3-57 6 the tes)ed valve was subjected to moments which ranged from 59 x of IO in - Ib to l47.5xl0 in - Ib without impairing the valve operability.

preference Since the maximum test moment and 70 percent of the ultimate moment carrying capacity of the pipe exceed the calculated maximum moment it is concluded that the operability of the valves will not be impaired.

TCG5/00 I /9

Question 9 The submittal states that the inlet piping configuration in which the test safety valve was tested was similar to that of the Turkey Point safety valve system. It does not, however, provide a comparison between the inlet piping configurations to verify this statement. Therefore, provide a comparison between the inlet piping systems used in the tests and the plant. As part of this comparison, the two inlet piping pressure drops should be compared. Provide a numerical comparison between the calculated plant pressure drop and the test pressure drop and explain how the plant pressure drop was calculated.

The following is the comparison between the EPRI Test case and the Turkey Point configuration:

EPRI 3K6 SAFETY VALVE

'F" INLET PIPING CONFIGURATION

~Len th in.. ~I.D. in.

Nozzle 17 6.813 Venturi 38 6.813 Pipe 6 6.813 Reducer 6 6.813/3.152 Loop Seal Straight 54 3.152 Bends 4-90 6 inch radius Reducer 4 3.152/2.624 Inlet Flange 7 2.624 Inlet ressure dro s Opening: 391 psi Closing: 194 psi Data shown is from "EPRI PWR Safety and Relief Valve Test Program Guide for Applications of Valve Test Program Results to Plant Specific Evaluations", REv.

2, Interim Report, July 1982 (V102) Tables B-3 and B-7.

See Figure 3 for an illustration of the EPRI configuration.

TCG5/001/10

TYPICAL TURKEY POINT 4K26 SAFETY VALVE SAFETY VALVE INLET PIPING CONFIGURATION LenceLth in. ~I.D. in.

Total pipe length I 36.5 3.624 90 Elbows 4 45o Elbows I Inlet pressure drops Opening: 433 psi Closing: 244 psi Calculation of inlet pressure drops was made following the procedure provided in "EPRI PWR Safety and Relief Valve Test Program Guide for Application of Valve Test Program Results to Plant Specific Evaluations", Rev. 2, Interim Report, July l982 (VI02), Appendix B. The rated capacity was obtained from Table B-I and flow parameters from Table 8-2 for the 4K26 valve. Flow of Fluids throu h Valves Fittin and Pi e'rane Co., Technical Paper No. 4IO was used as a reference. The inlet pipe configuration chosen as typical is the longest one on Unit 4.

See Figure 4 for an illustration of the Turkey Point configuration.

TCG5/00 I / I I

Question I0 The block valve that is used at Turkey Point, a Velan Model 8 l0-30548-I3MS, was tested in the EPRI/Marshall PORV block valve testing program. However, Turkey Point utilizes a Limitorque SMB-000-5 operator while a larger Limitorque SB l5 operator was tested. Explain how the test results for the SB-00-I5 operator can be used to demonstrate operability of the smaller SMB-000-5 operator.

~Res oose As reported in EPRI report NP-25 I4-LD, "EPRI-Marshall Electric Motor-Operated Valve (Block Valve) Interim Test Data Report", two Velan Model BIO-3054B-l3MS valves were tested. One valve had a Limitorque SB-00-l5 operator, and the other a Limitorque SMB-000- I 0 operator. Turkey Point utilizes a Limitorque SMB-000-5 operator. With the exception of the operators, the three valves can be considered identical. The difference in the operators reflects differences in required stroke speed. These are detailed below:

~oerotor Stroke Time EPRI SB-00-15 IO seconds EPRI SMB-000- I 0 I 5 seconds Turkey Point SMB-000-5 40 seconds Since the valves tested by EPRI operated satisfactorily during the tests at very near the design stroke speeds and the Turkey Point block valves are the same except for an operator sized similarity to provide a different stroke speed, the Turkey Point block valves can be expected to also provide satisfactory operation at its design speed.

TCG5/00 I / I2

e Question I I The submittal does not provide an expected value for the plant back pressure. To assure that the expected plant back pressure was enveloped in the EPRI tests, provide a value for plant back pressure and explain how it was determined.

The plant back pressure was calculated by RELAP5/MODI assuming simultaneous actuation of all safety valves with loop seals. The calculated back pressure is 493 psia as shown in Table 4.05 of EBASCO report (Reference I).

EPRI tests for the corresponding CROSBY 3K6 valve (Reference 6) give back pressures in the range of 47I psia (test 525) to 557 psia (test 537). Thus the expected plant back pressure is enveloped by the EPRI tests.

TCGS/00 I /13

Questions Related to Thermal Hydraulic Analysis:

Question I2 The submittal states that the thermal hydraulic analysis was performed using RELAP5/MOD I and that forcing functions were calculated from RELAP5 output with the Code CALPLOTFIII. Provide verification that the latter code has produced accurate force histories for similar problems.

Response

The post-processor CALPLOTFIII was programmed to convert the transient flow conditions (calculated by RELAP5/MODI) into transient forces on the piping system. The derivation of the governing equations are shown in Appendix 8+ of EBASCO report (Reference I). The validity of the program coding was verified by comparing hand calculation results against the. values computed by the program.

The program was further assessed against the GE 4-inch pipe blowdown test results. Favorable comparisons were obtained in comparing the computed results against the test data.

CALPLOTFlll was also verified by running CE test l4I I for SRV actuation on RELAP5/MOD I using the input from EPRI RELAP5/MOD I application (Reference l3). The calculated hydrodynamic conditions were converted by CALPLOTFlll to transient forces that duplicated the forces obtained by EPRI (Reference l3).

+See Attachment D to this submittal.

TCG5/00 I / I 4

Question l 3 The submittal indicates that thermal hydraulic analyses were performed for simultaneous actuations of the two PORVs in one analysis and three safety valves in another. lt does not, however, verify that the analyses were performed on fluid transient cases that produce maximum loading on the safety valve/PORV piping system. Provide evidence that the analyses were performed for transient conditions that produce maximum expected loading on the piping system. identify the fluid conditions assumed including pressure, temperature, pressurization rate, and fluid range.

Three cases have been considered corresponding to the following scenarios of valve actuation.

A. Two PORVs open simultaneously. SRVs closed.

B. PORVs do not open. All three SRVs open simultaneously.

C. PORVs open. Pressurizer pressure continues to increase, SRVs open when set pressure is reached.

The characteristics of the discharge are such that Case C is not a bounding case.

initially the system behaves as Case A until the pressure reaches the setpoint of the SRVs. At such time the SRVs open against a much larger backpressure than that of Case B, therefore producing lower loads.

For Case A the two PORVs opened at a pressure of 24I9.75 psia (2335 psig set point + 3~o) with zero pressurization rate and steam in the pressurizer at saturated temperature. At the valve inlet the water in'the cold loop seal is at l20 F.

For Case B the three SRVs opened at a pressure of 2574.25 psia (2485 psig set point + 3/o) with zero pressurization rate and steam in the pressurizer at saturated temperature. At the valve inlet the water in the cold loop seal is at l20 F.

The solid water case has not been analyzed for Turkey Point Units 3 and 4.

TCG5/00 I / I 5

Question 14 Report the flow rates through the safety valves and PORVs that were assumed in the thermal hydraulic analyses. Because the ASME Code requires derating of the =

safety valves to 90% of the actual flow capacity, the safety valve analysis should be based on a flow rate of at least I I I% of the flow rating of the valve, unless another flow rate can be justified. Provide information explaining how derating of the safety valves was handled.

Reseonse The valve flow rates were calculated by the RELAP5/MODI (reference l4) computer code. To generate the required flow rates adjusted flow areas have to be used as was demonstrated in EPRI RELAP5/MOD I application (Reference 13).

The actual calculated flow rate for the SRV actuation case is 356,400 ibm/hr which is l2I% of the original flow rating of the valve (295,000 Ibm/hr) and I I I%

of the updated capacity (320,000 Ibm/hr as shown in Table 8-I of Reference 2).

The calculated flow rate through the PORVs is 266, 400 Ibm/hr, their capacity is I 53,000 Ibm/hr.

TCG5/00 I /I6

Questions Related to Structural Analysis:-

Question l5 The submittal states that the structural analysis was performed using PIPESTRESS 20IO for elastic analysis and PLAST for plastic analysis. Provide verification that these programs have produced accurate results on similar fluid transient problems.

The PLAST program has been used to perform stress analysis of piping subject to pipe rupture and water hammer problems on various nuclear projects. A discussion of two of these analyses, taken from References 7 and 8, are included as attachments A and B. Furthermore, PLAST has been compared to the well known finite element program ANSYS and to the computer program FAB. These comparisons, taken from reference 8, are shown in attachment C.

The PIPESTRESS 20IO verification may be found in references IO and I I.

Additional verification exists at EBASCO. For example, the time history version of PIPESTRESS 20IO has been compared to the finite element program ABAQUS with very good agreement being observed. The PIPESTRESS 20IO program has been used on both nuclear and fossil plants to perform the stress analysis of piping subject to fluid hammer, main steam turbine trip and hot reheat turbine trip loads.

TCG5/00 I / I7

Question I6 The submittal does not explain what loading combinations were considered in the analysis to determine acceptability of the piping system. Therefore, identify the load combinations performed together with allowable stress limits'or piping and supports both upstream and downstream of the valves. The letter of September I, l982, indicates that the ANSI B3l.l code was used to evaluate stresses in the piping and supports unique to the PORVs. Identify all other governing codes and standards (with date of edition) used to determine piping and support adequacy.

~Res ense The load combinations and stress allowables which are used to determine the adequacy of the Turkey Point pressurizer relief piping system are identical tothe criteria specified in the Turkey Point FSAR by which the plant was designed, specifically that is the non-seismic ANSI B3l.l (l955) criteria. An additional load combination in the FSAR accounts for normal and seismic OBE loads with an allowable of l.2S. For normal and SSE loads the allowables are Sy (yield stress).

For comparison purposes, and because thermal hydraulic load cases were not part of original design, the calculated loads which are shown in Table 4.3. I (attached) were combined according to the EPRI recommended load combination as provided in Reference 2 and shown in Table 2. Table 4.3.I was developed for piping upstream and downstream of the PORVs subjected to PORV actuation. The stress allowables for these tables correspond to References 4 and 5, accordingly and are included in Table 2.

The load combinations used to analyze the piping subjected to SRV actuation correspond to EPRI recommended load combinations as shown in Table 2. For piping upstream of the SRV valves the stress allowables are as shown in Table 2.

For piping downstream of the SRVs and subjected to SRV actuation, the moment developed in the piping is compared to 70 percent of the ultimate moment carrying capacity of the pipe. The ASME code itself allows use of this value to satisfy integrity criteria. For a further discussion of this point see the response to Question 9.

Load combinations for supports are also shown in Table 2. These also correspond to the EPRI recommendation (Appendix E of Reference 2). FPL is analyzing the first and third load combination for supports. That analysis will be completed by 9/I/86. Support loads for both PORV and SRV transients were originally compared against support capacity and are discussed in the response to question l7.

All analyses were performed on Unit 4 due to the similarity of the Units. Unit 4 was considered to be the more limiting of the two based on its support configuration.

TCG5/00 I /I8

Question l7 The submittal states that the piping and supports unique to the PORVs have acceptable stress levels, and that all piping and supports in the safety valve and PORV systems contained in the Primary Coolant Boundary were shown to have acceptable stresses. Provide a numerical comparison between calculated piping and support stresses with allowable stresses to verify this conclusion. The submittal also states that results from PLAST analysis show the discharge piping

. and supports to be adequate when stress levels exceed the yield point for the following reasons:

(a) No deformation was significant enough to reduce the flow area or detrimentally impact the flow path.

(b) Piping and restraint movements were such to preclude interaction with other components.

(c) No pipe rupture was found, thereby precluding possibility of pipe whip or jet impingement.

Supply numerical results from the PLAST analysis that verify these conclusions.

Finally, provide an evaluation of the effects that large deformations in the discharge piping will hove on the structural integrity of the valve inlet piping and pressurizer nozzle connections.

Response

For the analysis of the pressurizer relief piping subjected to PORV actuation, the stress ratios as predicted by PIPESTRESS 20IO are shown in Tables 4.3.I and 4.3.2. At isolated points, overstress has been recorded. Upstream, the overstress is restricted to two points ( I 384 & I 7 I 4) with 3'nd Downstream, three points (l202, l2I2 & l260) showed overstress.

I 0'verstress.

For the upstream nodes, due to the well.known conservatism of Generalized response method used for this calculation, it is felt such overstress is artificial and can be overriden by more refined analysis. Furthermore, these results reflect conservative combination of moments from all three orthogonal directions. For Turkey Point, the loads are only required to be combined in one horizontal and the vertical direction to gain the resultant stress.

For the downstream nodes, elastic stress ratios are shown to be as high as 50%

above allowable. These ratios are of significantly smaller magnitude than those discussed below for the SRV actuation. Since these loads are enveloped by downstream loads due to SRV actuation, the discussion below is bounding.

The calculated support loads due to PORV actuation are compared to the design loads in Table 4.4. All the reactions are within the capacity of the restraint structures.'s stated in Question 8's response, FPL is currently performing analyses to satisfy load combinations I & 3 for supports.

With regard to the pressurizer relief piping subjected to SRV actuation, the PLAST analysis (see Reference I) indicates plasticity in the first elbow downstream of the pressurizer and'he elbow downstream of the SRV. The TCGS/00 I / I 9

Response to Question I7 (continued) moment carrying capacity of the pipe has been computed for each size and material by the method of Gerber which has shown excellent agreement with experiments. For elbows, that moment is modified by dividing by the B~ factor to account for the lesser capacity of the elbow. This approximate method has shown excellent agreement with utimate moments of elbows computed by finite element techniques. Refer to Table I (question 8) for tabulation of the maximum moments due to SRV actuation. The maximum moments computed in the system through the first 70 millisecond response are shown and are compared with 70 percent of the ultimate moment. As-mentioned in the response to Question l6, the ASME code itself allows use of this value to satisfy integrity criteria. Clearly nowhere is the 70 percent ultimate moment exceeded.

The support reactions due to SRV actuation are shown in Table 4.4.A. All the reactions are lower than capacities of the support structures.

In response to (a), the choice of 70 percent of the ultimate moment is dictated by prior finite element analysis as discussed in Reference l2. At such a load level plastic strains in the pipe would be insufficient to change its cross section significantly enough to affect its flow characteristics.

In response to (b), the small strains (approximately 0.4%) will not cause interaction of piping with other components.

In response to (c), the limitation on the maximum bending moment to 70% of the ultimate capacity implicitly precludes pipe rupture.

Finally, in regard to evaluation of response t'o large deformations in the discharge piping, the response to (b) above demonstrates that with such small strain, no large deformations are produced.

TCG5/00 I /20

0 Question 18 The submittal notes that high frequency pressure safety valve piping during loop seal discharge tests.

oscillations occurred in the" According to EPRI results these oscillations were approximately 170-260 Hz. The submittal refers to an evaluation of this phenomenon that is documented in Westinghouse report WCAP 10105. The study discussed in the Westinghouse report determined the maximum permissible pressure for the inlet piping and established the maximum allowable bending moments for Level C Service Condition in the inlet piping based on the maximum transient pressure measured or calculated.

While the internal pressures are lower than the maximum permissible pressure oscillations could potentially excite frequency vibration modes in the the'ressure piping, creating bending moments in the inlet piping that should be combined with moments from other appropriate mechanical loads.

Typically the structural responses in these piping systems is due to frequencies less than 100 Hz. The piping response is limited to these lower frequencies primarily because these systems are not normally subjected to forces having frequencies above 100 Hz. With the presence of the high frequency pressure oscillations, however, the higher frequencies existing in the inlet piping could potentially be excited, resulting in significant structural response. Therefore, provide one of the following: (I) a comparison of the allowable bending moments established in WCAP 10105 for Level C Service Conditions with the bending moments induced in the plant piping by the dynamic motion and other mechanical loads, or (2) justification for other alternate allowable bending moments with a similar comparison with moments induced in the plant piping.

~Res onse To evaluate a bending moment in the inlet piping due to the high frequency pressure oscillation at the opening or closing of the valve a linear dynamic analysis (Generalized Response Method) of a critic'al portion of the pressurizer relief piping system has been performed. That portion of the system includes one loop of Class I piping (4 inch O.D.) and outlet portion of the valve (6 inch O.D.)

down to the X,Z-directions restraints and Y-direction snubber.

High frequency sinusoidal forcing function parameters are based on selected experimental data obtained from EPRI/CE safety valve tests (ref. 6) for the Crosby HB-BP-3K6 with loop seal internals (Test II526).

Pressure oscillation amplitude is 40 psi (from peak to peak) and frequency 170 hz.

Program 2010 PIPESTRESS (Reference 10, I I) has been used for analysis. The results show that the effect of the high frequency loading is negligible.

The maximum bending moment of 1.056xl0 in-lb. in the inlet portion gf the valve is much lower than an established maximum allowable moment 1.47xl0 in-hr.

TCG5/001/21

REFERENCES Turkey Point N.P.P. Units No. 3 & 4. Analysis of Pressurizer Power Operated Relief Valve and Safety Valve Discharge Piping. NRC NUREG 0737 Item II.D. I, prepared by EBASCO Services Inc., August I 982.

2. EPRI PWR Safety and Relief Valve Test Program Guide for Application of Valve Test Program results to Plant Specific Evaluations, EPRI Safety and Relief Test Program, February 'I 982.

3. Review of Pressurizer Safety Valve Performance as Observed in the EPRI Safety and Relief Valve Test Program by E. M. Burns et al; WCAP-IOI05, June I982.

4. ASME Section III - Division I, Subsection NB, l980 Edition, NB-3654.2, NB-3655.2.

5. ASME Section III - Division I, Subsection NC, Winter l98I Addenda NC-3653. I, NC-3654.

6. EPRI PWR Safety and Relief Valve Test Program, Safety and Relief Valve Test Report, EPRI NP-2628-SR Special Report, December l 982, Prepared by Electric Power Research Institute, Palo Alto, California.

7. Florida Power and Light Company Updated FSAR, Docket 50-389, St. Lucie Plant Unit 2, Vol. 4 revision I, 4/86.

8. Florida Power and Light Company Updated FSAR, Docket 50-335, St. Lucie Plant Unit I, Vol. 3 revision 3, 7/85.

9. ETR-1002-P Design Considerations for the Protection from the Effects of Pipe Rupture, Part I Pipe-Whip Dynamic Analysis by J. Heifetz and R. Iotti, EBASCO Services Incorporated, October I 974.

I O. Program 20IO Theory Manual for Release 3.8.I of PIPESTRESS 20IO New Dynamic Analysis Features, EBASCO Services, August, l98I.

I I. PROGRAM 20IO VERIFICATION Manual for Release 3.8.l of PIPESTRESS 20 I 0 New Dynamic Features, EBASCO Services., August I 98 I.

I 2. S. H. Shaaban et al, "Functional Capability of Piping Elbows", SMIRT Proceedings, August I 985.

I 3. 'Application of RELAP5/MODI for Calculation of Safety and Relief Valve Discharge Piping Hydrodynamic Loads Interim Report, EPRI Safety and Relief Valve Test Program, March l 982.

I 4. RELAP/MODI Code Manual, Ransom V.H., Wagner, R J et al, Vol. I & 2, EG&G Idaho Inc., NUREG/CR I826, EGG 2070 Draft, Revision 2, September, I 98 I.

I 5. Letter from C. S. Kent to L. Tsakiris, JPE-PTPO-83-365, dated March 7, I 983.

TCG5/00 I /22

CONTROI.I.ED DOCIJNENT F 2 SECTI()N V, FI(i'll(t.: 3A N

~ $

VJ

~ )

~ ~

~

~

FROFJ Fl<C tlONROEV1LLE PA (THV)85 29

~ ~ 86 141>59 HO. 18 PAGE 5 1

~ g FIGURE 3 l

PIP ING CONFIGURATION I (XtRACTEQ, FROM EPRI MP-2770-LD VOLUME 5 I~

T EST i

s g/

~

,'At VE

~ I 4" OR 6" SCH.XX I K

8" SCH 16p I

1 i:l ENTuR~

I

)IC

/g,I

~ I

FROM MNC MONROEII?LLE PA (THU)85 29 '86 15I 88

~ ~ HO ~ 18 PAGE 7 I

FIGURE 4 (I

I ii I PIP ING CONFIGURATION TYPICAL TURKEY POINT II h

~" ~

~

I

COMPARISON OF THE MAXIMUM MOMENTS Node Pipe Size and Sch. 70% Ult. Moment Calc. Max. Moment Time of Induced *) Calculated (in-lb) (in-lb) Occurance Bending Moment/Max Induced (msec) Moment Bending Moment Opening/Closed (in-lb) 5 5 5 40 4 in Sch 120 2.44 x 10 1.26 x 10 17.6 (0.59 1.47)xlO 0.86 4

39 4 in Sch 120 Elb 1.23 x 10 5 7.96 x 10 13.8 0.54 5 5 29 6 in Sch 40 3.59 x 10 1.04 x 10 5.5

• ) The reported values are the maximum induced bending moments on the valve discharge flange during opening or" closing.

TABLE 2 GOVERNING PORV &.SRV LOAD COMBINATIONS FOR PIPES 6 SUPPORTS PLANT/SYSTEM SERVICES STRESS LIMIT COMBINATION OPERATING CONDITION LOAD COMBINATION CLASS 1 CLASS 2 UPSET Sustained Loads + OBE + 1.8S 1.8S m

(PIPING) Relief Valve Discharge Transient UPSET Sustained Loads + Relief valve Stress Level B (SUPPORTS) Discharge transient EMERGENCY Sustained Loads + Safety Valve 2.25S 2.25Sh (PIPING) 2.25S m m

EMERGENCY Sustained Loads + Safety Valve Stress Level C (SUPPORTS) discharge transient UPSET Sustained Loads + OBE Stress Level C (SUPPORTS) + Relief Valve Discharge Transient NOTE:

S m

= Basic Allowable Stress Intensity = 20 ksi S = Basic Material Allowable Stress at Maximum Temperature = 15 ksi h

0 STRESSES IN TURKEY POINT 3 & 4 PRESSURIZER RELIEF PIPING FROM PORV ACTUATION TYPE OF STRESS (PSI)

Pipestress 2010 ASME Code Load Combin- Ratio of Stress Node Point Class Pressure ~Wet ht PORV 'OBE(EDS)* to Allowable 192 51 5389 6477 1477 12273 0.454 254 35 7408 7576 2324 15198 0.563 362 11481 3570 2154 2418 17374 0.483 374 11481 3977 3306 2361 18912 0.525 384 11481 1174 2901 1771 15656 0.436 912 35 18237 8338 1980 26683 0. 989 954 35 7869 7632 1787 15639 0.579 1072 11481 2197 2239 2101 16055 0.447 1082 11481 2763 2399 2322 16804 0.467 520 35 4211 6968 2800 11558 0.428 534 2 35 8476 . 6540 864 15076 0.558 1202 35 16717 21831 4906 38894 1.440 1212 35 14558 21017 4187 35855 1.328 1222 35 7947 9611 2148 17724 0.664 1234 35 '7143 10468 2195 17782 0.659 1242 35 7538 12487 2342 20196 0.748 1260 35 21853 18277 6338 40662 1.506

• REF. 15

TABLE 4.3.1 (continued)

STRESSES IN TURKEY POINT 3 & 4 PRESSURIZER RELIEF PIPING FROM PORV ACTUATION Pipestress 2010 TYPE OF STRESS (PSI)

ASME Code Load Combin- Ratio of Stress Node Point Class Pressure ~Wet ht PORV OBE(EDS)* ation (PSI) to Allowable 1330 8930 2410 9855 5184 21820 0.606 1382 11481 11936 10169 3769 33800 0.939 1384 11481 14445 13371 3908 39491 1.097 1410 11481 3638 3625 3300 .19032 0.529 1422 8930 10042 9158 9216 29601 0.823 1502 12 3984 8198 5340 13312 0.493 1514 12 3434 8064 5390 12709 0.470 1714 11481 18644 6621 5164 37107 1.03$

1724 11481 4044 6497 6582 22984 0.639 1734 11481 3259 7129 6830 22954 0.638

• Ref. 15

TAB4 REACTIONS ON THE SUPPORTS (lbs)

(TRANSIENT EFFECTS DUE TO OPENING OF PORV's)

ISO EBASCO PIPE SUPPORT FUNC- *TRANSIENT TYPE OF **)MAXDESIGN **)CAPACITY OF PT PT SIZE I.D. TIONAL (LOAD (LB) SUPPORT LOADS,LBS SUPPORT (FROM (ISO) DIRECTION (PAGE046A-046B) STRUCTURAL DWGS)

A4 A4 41 41 1 211 12" 4-PRH-3 X 1718 1521 Ri id Restraint Il II 1801 '2896 8000 Tab 1 Z N.A N.A 7 000 Tab 1 14 140 12" 4-PRH-4 . X '094 11 11 2240 . -1869 N.A Tab 2 27 270- 6" 4-PRH-8 X 484 11 II 580 ~

-562 6 000 Tab 4 270 611 Z ~

573 606 ~

-1285 7 000 Tab 4 B53 532 PRH-12 Z 228 II II 6'11 2085 510 4 000 Tab 6 59 590 4-PRH-ll II II X 389 516 -731 N.A Tab 7 59 590 6 '. II Z 349 II II 368 "-2328 N.A. Tab 7 62C 620 6" 4-PRH-10 Skewed 538 . II II ~

318 -784 8 000 Tab 8 LAT Z 62D 620- 611 'Skewed .1095 -307 8 000 Tab 8 226'68 LAT X 98 980 611 4-PRH-9 X 577 -904 2,000 Tab 9 98 980 611 4-PRH-9 Z 229 602 -510 7 000 Tab 9 123 1230 6" Penetration X 2913 1546 -1787 N.A A125 13 1251 130 6"

1211 4-PHR-6 4-PHR-4 Z

Z 1904 2630 Snubber .

1287 2016

-1423

-2016 1 000, Tab N.A, Tab 2 ll A18 181 1 211 4-PRH-2 Y 560 -560. 10 OOO,Tab 3 26 260 611 4-PRH-8 Y 742 572 -572 10 000 Tab 4 A61 611 611 4-PRH-10 Y 227 203 -203 10-000 Tab 8 96 960 4-PRH-9 Y 1184 550 -550 10 000 Tab9 125 138B 1250 1384 6"

4 II 4-PRH-6 4-PRH-5 Y

S-W 1184 1828 464 1693

-464

-1693

<10,000 Tab N.A. Tab 16 ll 138C 1385 411 S-W 4437 4257 -4257 N.A. Tab 12 10 138D 1386 411 25oVert 3057 3408 -3408 N.A. Tab 12

• HOOP STRESS IS INCLUDED IN EVERY TRANSIENT LOADING

• • ) Document TR-5322-158, PR-2, UNIT 4 by Teledyne Engineering Services

TABLE 4.4A REACTIONS ON THE SUPPORTS (lbs)

I (TRANS ENT EFFECTS DUE TO OPEN ING OF SRV ~ PLAST I PART AL MODEL)

ISO EBASCO PIPE SUPPORT FUNCTION~ TYPE SRV TRANSIENT TIME OF CAPACITY OF TELEDYNE ENG.

PT "PLAST" SIZE ID(ISO) DIRECTION OF REACTION (lbs) OCCURANCE SUPPORT (FROM SERV. STRUCTURAL PT (IN) SUPPORT (Sec) STRUCTURAL DWGS.

DWGS. lbs)

. A4 5A 12 4-PRH-3 X RESTRAINT 3, 114 0.038 8~000 Ref.DWG. DCN-M 380- **)TAB 1 A4 12 4-PRH-3 Z 444 0.14 2,000 14 12 12 4-PRH-4 X 1,030 0.04 Design Load TAB 2 on Struc Dwg.

+2,240;-6869 27 26 4-PRH-8 X 6,955 0.009 , 6,000 DWG 4-128-1,4-129 SGb. Syst.E-2388-WIC-2 **) TAB4 27 26 4-PRH-8 Z 6,902 0.009 7,000 26 25A 4-PRH-8 SNUBBER 0.006 <10,000 10 050 PSA-10, S/N=121 13 12A 12 4-PRH-4 Z 1,630 0.174 N.A on Struct.

DWG.(Design Loads + 2016*) NA 12 4-PRH-Z 1,583 0.045 Snubbers Dwg 4-114-1,4-115 PSA-10 assumed to be Sub Syst E-2388-S/N=108, a weak link WIC-6**)TAB3 S/N 104 capac. of each 10 Keeps

+) Max design loads are taken from TR-5322-158, Rev. 2, p 046A

• • ) Documents are indicated as TR-5322-158, Prob- PR-2 P

unit 4

Attachment A Reference 7 Discussion of a Use of PLAST

SL2 "FSAR APPENDIX 3 'E MAIN STEAM 6 FEEIMATER ANALYSIS 3 ~ 6E-i Amendment No. 3, (6/81)

SL2 - FSAR

3. 6E MAIN STEAM & FEEDWATER ANALYSIS This appendix presents the results of dynamic analyses performed to verify the structural adequacy of the Main Steam and Feedwater pipe vhip restraints.

The transient forces resulting from postulated piping failures have been generated using the RELAP computer code. Thrust forces from RELAP are used as input to the PLAST 2267 Code The PLAST code uses these forces to determine the pipe vhip restraint reaction loads by performing a dynamic structural analysis on a lumped maes parameter of the piping system. These computer codes are described in References 3 ~ 6E-1 through 3.6E-3 ~

Typical pipe vhip restraint structuree used on the Main Steam and Peedvater systems are shown in Figures.3 6D-1 and 3 ~ 6D-2. ELastic stiffness values for

~

these restraints have been determined by two methods:

a) The bolts securing the restraints to the RCB structure were considered infintely rigid. Credit vas taken only for the elasticity of the pipe vhip restraint structural steel b) The elasticity of the bolts was considered and a combined stiffness vas calculated Method 2 results in stiffness values approximately 36 percent to 57 percent of the values obtained by Method 1. In addition, pullout loads have been determined for the bolt system.

Reaction loads at the pipe whip restraints have been generated by the PLAST program based on the combined stiffness valueso The resulting loads were then compared vith the bolt pu11out loads, Typical results are tabulated in Table 3.6E-I for selected Main Steam restraints and show that the restraints perform thei design function since the reaction loads are below the bolt pullout loads.

The break locations analyzed are shown in Pigure 3 6E-1 for Main Steam piping and Figure 3 'E-2 for Feedvater. These figures also represent the PLAST models for these systems. Figures 3.6E-3 and 3 'EW give the RELAP models for the Main Steam and Feedvater piping, respectively The volume and )unction data for Main Steam RELAP are given in Tables 3 'E-2 and 3 'E-3 ~ The pipe whip restraint gape used in PLAST For Main Steam are given in Table 3.6E-4 ~

This information is tabulated for the Feedwater analyses in Tables 3 'E-5 through 3.6E-7, respectively Force versus time history for RELAP is presented for a typical Main Steam break in Figure 3.6E-5 and for a typical Peedwater break in 'Pigure 3.6E-6.

3.6E-1 Amendment No, 3, (6/81)

SL2 FSAR "RELAP3 A Computer Program For Reactor Bio@down Analysis" by M H Rettig, G A Jayhe, K V Moore, C E Slater, M L Uptmore, Idaho Nuclear Corporation IN-1321 Issued June, 1970, Reactor Technology, TDD-4500 RELAP4MOD6, A Computer Code For Transient Thermal Hydraulic Analysis of Nuclear Reactor and Related Systems, User's Manual, EG&G Idaho,

. Inc>> CDAPTR003, January, 1978 "Design Considerations for the Protection From the Effects of Pipe Rupture", ETR-1002-P (Proprietary Version) and ETR"1002 (non-Proprietary version), by Ebasco Services, Inc August, 1977.

3,6E-2 Amendment No, 3, (6/gl)

SL2-F SAR TABLE 3.6E-l SUHHARY OF SELECTIVE PIPE llHIP RESTRAINTS AND DYNAHIC LOADS RESTRAINT RESTRAINT STRUCTURE STIFFNESS STRUCTURE STIFFNESS(2) BOLT RESTRAINT (KIP/IN) KIP/IH PULLOUT REACTION RESTRAINT PULLOUT (Ml) SHEAR (x2) PULLOUT LOhD(3) LOAD SOURCE BREAK

~xl ~KHS I RON LOCATION REHARK HEWS-l6 +38,749 52>375 +16,202 +35>830 5, 187 5, 160 FL02132HS Slot break beta

-181>549 -38,993 -90,728 -25,547 Pts 9 & 6709 RESS 17 +39,304 +37> 106 +14,288 +23, 103 3,021 542 P L021KLHS Guill. at S/G

-117,548 -40,521 -81,754 -25,538 Nottie RE~-20 +28,291 +39,126 +16,143 +32,025 4,795 3,520 PL021RlS Guill. at Pt. 16

-51,075 -28>969 -71 >780 -33>924 from Pt. 16 LA>

Notesr PS

>A>

(1) Bolts rigid (2) Bolts elastic (3) Applied load at vhich first both failure occurs.

B m

KR 0

o C>

SL2 - PSAR TAHLE 3o6E-4 RESTRAINT GAPS USED POR PLAST MODELS OF MAIN STEAM L RESTRAINT LOCAL (1) LOCAL (1) LOCAL (1) LOCAL (1)

NAME +Y -Y +Z <<g G AP IN. CAP IN GAP IN+ CAP IN RE-MS-24 4.00 4000 4.00 4.00 REWS<<21 4 25 4+00 4 00 4 25 REWS-20 4.50 4+00 ~

4e00 5 50 RE-MS-19 4 00 5+25 4+00 4.50 REWS<<18 4 00 4 25 5 75 4+00 REWS<<17 4.00 4o50 100,0(2) 100.00(2)

REM S-16 4 00 4 00 4o00 6. 00 REWS<<15 5+75 4 00 4 75 4.00 RE<<MS-14 6+25 4 00 99 625(2) 100.375(2)

RE-MS<<13 5 75 4 00 4.00 6.500 REM S-12 4+00 6 75 4+00 5.750 NOZES:

(1) Por local directions see Figure 3.6E-1 (2) Local + Z is unrestrained for this restraint 3,6E-S Amendment No, 3, (6/81)

SL2-FSAR TABLE 3 ~ 68-7 RESTRAINT GAPS USED POR PLAST MODELS OF BOILER PEEDWATER LINE RESTRAINT LOCAL(1) LOCAL(l) LOCAL(1) LOCAL(l)

NAME +Y -Y +Z -z GAP (INe ) GAP (IN.) GAP (IN+) GAP (IN. )

RE-BP-17 4.00 4 00 4.00 4+50 RE-BP-16 4.00 4+00 4+00 4 '0 4.00 RE-HF-15 4 00 4.00 4.75 RE-BP-14 4 ~ 00 4 ~ 00 4 00 4.25 RE-BP-11 9 00 9+00 9.00 9.00 RE-BF<<10 6 ~ 50 4 ~ 00 4.25 4. 00 Note:

li Por local directions (see Figure 3,6E-2) 3 'P 13 Amendment No. 3, (6/81)

NOTES:

Yg 1, + INDICATES THE LOCATION OF A GUILLOTINE BREAK. THE THRUST FOACE ON THE PIPING IS DIRECTED ALONG THE PIPE AXIS OPPOSITE TO THE ESCAPING JET FLOW.

%775 RWlj 12

2. THIS FIGUAE WAS DEVELOPED USING INFORMATION FROM EBASCO ISOMETRIC HO. Ms-li7-1. RE4$ -1$ $ $ 'w Q" RED

$ . NUMBERS IN PARENTHESIS REFER TO NODE HUMBfRS ON THf ISOMETRIC CITED ABOVE.

2.7525 2.762S

~ 202

i. Xl, Yl, Zl REFER TO THE GLOBALCOOADINATE DIAECTIONS 1A925 X) 6. )4 Y, Z AEFER TO THE LOCAL COORDINATE DIRECTIONS AT A PARTICULAR 2$ (i)

NODE. Sly lA8?6 a.7825 L7625 Sill)

~ . IN THE ACTUALPIPING THEAE IS A RESTAAIHT AWlS ll LOCATED AT EL. $ $ $ 0 NODE 3llll. THE GAPS ON THIS RESTRAINT WOULD NOT HAVE CLOSED i

R WAS-li WITH ANY OF THE BAEAK LOCATIONS MODELED. AS A CONSEOUENCE THIS AESTRAINT WAS HOT MODELED.

~

146"48642 1-38"4lS4$

5. f STEAM Gfhf hATOh 2A MAUI $ 7SAII OUTLET 8

2.716 WO7 AEAIS-1$ X

~~ ill~~

4.775 AMES-14 16 ISLEEVE FLOOA FLOOR fL$ 2.'

PENETRATIOH) 21 1.7786 16 ~ 252 LOCAL 2 IS INTO THE PAGE 20$ $

NORMAL TO THE PIPE + Z Y ZY x+ i$

SAS 2

~

12(1 l g FLUED HEAD X ISLE EVE WALL 2.10 'X ~ 4865 g ELSSAO RMAS-2$ PENETRATIONI RE4LS-20 Y 1$

2.114 11 llSI Z

)

.772 2 8 San San 2gg$ 2Als MODULUS OF OPERATING SCH. NolL OPERATING ELASTICITY~ POGSON'5 PAESSUAE LINE NO. MATEAIAL TESH' OP. TEMP. RATIO ANEgggSNT HO. ~ IHS'U WL THICKNE PSIG PSI FgpglpA PpwjR 4 LIGHT LONPAH1 I-3I"AS.28 31.000" l.'7SS 532 28M w 10 5T LQ jlj PLANT UNIT l 532 10 886.

1.35"MS 52 38.676" 1.2i6 28.28 w PLAST PIPE RUP Tukj M:I'IL Of 't RfACTPR P PQ. MAIN 57 LAN LlhL FIGURE 3 0E I

1. RELEVANT INFORMATION FOR I-20--BF-Ii AND I-IS"-BF-61 2.5 SCH. NOM. OPERATING OPERATING 2.6 STM. GEN 2A LINE NO. PIPE OAI. WALL PRESSURE, MATERIAL TEMP of THICKNESS PS I G II. EL. 6838' I-20--BF-li 20.00"- l.037- 40. 1050. CS. A106-8 6081 RE-BF-10 1.070 I-18"-BF-51 18.00" 0.937 40. 1050. CS. A106-8 X3176 AP13 MATERIALPROPERTIES 0 TEMP. 22175 RE-BF-ll MODULUS OF YIELD YIELD POISSON'6 DENSITY I.'18" BF47 31 2A05 CENTER OF FLOOR LINE NO. ELASTICITY STRESS RATION LBIIIN3 1.0001 STRAIN PS I A PSIA V~57 2.0001 WT 282K 1-20"-SF-li 29.75 x 106 29.75 x 10 .SXI 1-18"-BF-51 STEP 20" x 18" RED.

29.75 x 10 19.76 x 10 0$ 33 AP12 AP'll LOCAL AXES FOR RE4F-17, RE SF 10 L398

2. FIGURE 3.$ .E-10 WAS DEVELOPED FROM EBASCO ISOMETRIC NO. BF 187

3. + INDICATES THE LOCATION OF A GUILLOTINE BREAK.

I REV 'I 11-10-7S.

2.60 2AS 326 AI10 1.071 Z~Y 1.071 LOCAL AXES FOR RE4F-15, RHF-li 3~

LOCAL AXES FOR 1.071 REEF-17, RE-BF-1$ 2 NORMAL TO THE PIPE IN M50 THE GLOBAL 2-X PLANE 1AI71 Y 3260 NORMAL TO THE PIPE IN TAI71 THE GLOBAL 2-X PLANE ALONG THE PIPE AXIS 3260 1.07'I RE-BP-14 3~

1439 lAOS AP2 IW9 RE-BF-15 FLUED HEAD I-20"-BF.Ti '1.790 EL 18 38AXI'019 RE-BF-17 APp D70 035 121 18 171 7017 15 AMENDMENtNO. 3 IllII 51M L820 1AI29 L163 1

r 16S 224 224 lAI~

FLORIDA POWER L LIGHT COMPANY SWT 5.902 3282 L153 L163 ST. LUCIE PLANT UNIT 2 0.512 SLEEVE WALL PLAST PIPE RUPTURE MODEL OF THE PENETRATION REACTOR BLDG. BOILER FEEDWTR. LINE FIGURE 3,6E-2

PIPF THRUST FORCE VERSUS TIME CURVE USED BY PLAST MODEL FLO21H2MS 112 6A 4$

PIPE THRUST FORCE VERSUS TIME CURVE USED BY FLAST MODEL FLO21H2MS 0

9.6 0 .00l .008 .012 .016 .020 .024 AGO" TIME SECONDS 6.4 f 4.8 0

NOTE 1l FOR OESCRltTION OF MAIN STEAM tltlNO MOOKLS SEE TASLES SA E T.

FOR MAINSTEAN tltlW,THE THRUST FORCE IS AttLIEDOttOSITE TO THE

~

F LOW AT THE SEE AK FLARE.

ANCNlNICNTNCL S IVlll FLORIDA POwER d LIGHT COMPANY 0 02 ~ .06 .08 .10 .12 .14 .16 .18 ~ ~ Ci ~6 ~8 S'T. LUCIE PLAHT UHIT 2 TIME lSECONDS) PIPE THRUST FORCE VS. TIME CURVE USED 6T PLAST MODEL FL02 IH2IIS FIGURE 3.6'

120 NOTES:

'l. THE NODE NUMBER IS REFERENCED ON FIGURE 3$ E40.

2. TABLE 3.6 E-12 PROVIDES A TABULATIONOF FORCE VERSUS TIME ACTUALLYUSED.

0 0.00 0.08 0.$ 6 024 092 0.40 0.48 TIIHE (SEC)

AMENDMENTND. 3 le/81)

FLORIDA POWER 8 LIGHT COMPANY ST. LUCIE PLANT UNIT 2 FORCE HISTORY APPLIED AT NODE 14 OF BOILER FEEDWATER PIPING MODEL FL021PLASTBF4 FlGURE 3.6E4

Attachment B Reference 9 Discussion of a Use of PLAST

3.6.4 3 Pi e Whi Anal sis - Main Steam and Feedwater As shown on Figure 3.6-52, a break location was established at node 12 for the main steam line. A break at node 12 results in the maximum impact at the restraint located at node 9 and the maximum total strain in the pipe.

Span lengths between pipe whip 'restraints are as shown on Figures 3.6-52 through 3.6-55 'for main steam and feedwater piping. The maximum span lengths depicted were established using the design criteria presented in Section 3.6.5.1. As stated in 3.6.5.1, failure stress is limited to that value which corresponds to 50 percent of the true ultimate strain when related to a simplified stress-strain curve (Figure 3.6-9A).

For the steam line break selected (node 12 on Figure 3.6-52'), the moment required for full plasticity, for yielding and the actual moment computed for that limiting case are 35.2 x 103 in-kip, 27.8 x 10 3 in-kip and 35.2 x 10 in-kip, respectively. The actual computed moment and the moment to full plasticity are equal since for this limiting span length the pipe does plastically deform, but does not whip.

Sensitivity studies for the main feedwater line outside containment are summarized for variations in gap length and pipe wall thickness in Table 3.6-1 and Figures 3.6-58, 3.6-60k 3.6-62 A reduction in gap reduces peak restraint reactions while decreasing wall thickness seems to increase reactions.

3.6-9

The spa>> method of restraint placement (See Section 3.b.5.1) does>>ot pro-vide for a margin to full plasticity since the method itself assumes the pipe to go plastic. The span methou aoes, however, prevent tgaximum cal-culated strai>> from exceeding one-half of the ultimate strain. For example, at the instance when a pipe is fully plastic, the pipe can still carry moments only additional strain energy into the pipe will cause further straining, up to the ultimate.

If the strain hardening is ignored, as was done in this a>>alysis, the>>

t>>e spa>> method does not predict the strain correspo>>ding to impo'ed mome>>t si>>ce this strai>> is not u>>ique. !lowever, rigure 3.6-bl shows t>>at the maximum strain does not even approach half ulti>:.ate strai>> values. Since zero strai>> haruening was employed, it follows that t>>u calculated mome>>t at maximum strai>> anu the moment requireu "

for fuJl plasticity are iaentical, 35.2 in>>kip.

At this moment value, the pipe >>as not collapsed anu can co>>tinually carry equal or uimi>>is>>ing loaus. The- mome>>t necessary to yielo the outer fioers of t>>e pipe is 27.b in-kip.

Figures D.b-52 through 55 providea the span lenbths, restraint locations a>>a noae locations using >>odal breakdow>> requirements for pipe whip a>>alysis. Nodal breakdown requireme>>ts for pipewhip analysis are diff-ere>>t.tha>> those required for stress analysis as reflected i>> the piping isometrics of Section 3.6.

Because >>ode points were included in the numbering scheme in Figures

.3.6-52 through 55 and since noaes were not shown in other Section 3.6 figures, no corresponaence. should be expectea 'between the two except that piping dimensions and location of pipe whip restraints are identical.

To illustrate this point refer to Figures 3.6-36 anu 3.b-52. Figure 3.6-36 restraint locations i$-2, >lS-3, HS>>4 and >lS-5 correspond to restraint locations 2, 4, 5 and 8 in Figure 3.6-52, respectively.

All the related figures have been reviewed for accuracy.

a) Conclusions Four cases (2 feedwater line breaks and 2 main steam line breaks) of circumferential pipe rupture vere analyzed for maximum restraint re-actions and maximum pipe strain. hll analyses were extended for a period of 0.2 seconds past.'nitiation of pipe rupture when steady state, oscillations of the deflection of the rupture point occurred vith decreasing amplitude. Strain hardening in the pipe vas assumed to be zero for conservatism.

. Figure 3.6-56! indicates that blovdown forces for the feedwater line rupture reach a steady state value of 110,000 lbs at 0.034 seconds; for the main steam line, the blowdown forces reach a value of.

135,000 lbs at 0.1 seconds, decrease exponentially to 100,000 lbs at 0.15 seconds, and continue to decrease at. the same exponential rate thereafter.

3.6-10

Since the steam line is multi-planar, restraint reactions can occur in more than one direction and in more than one restraint. Thii is evident from the reaction force results plotted on'Figures 3 '-57!

and 3,6-58; Peak reactions in all cases except two were belov the 2 KPA factors applicable to the line under analysis. In the two exceptions (feedwater line break't node 6, Figure 3.6-59; and main steam line break at node 12; Figure 3.6 58.), the peak duration is

'approximately 0.002 seconds or less.

Since the natural periods of the restraic>t system, co>>sisting of the steel frame restrai>>ts, the embeaments and the co>>crete vali are of 0.002 seconds or less, this system was reviewed to determine to what extent, if any, its primary fu>>ction of pipe restraint during blowdown may be impaired.

A very conservative analysis of the steel frames, in which tne stif-fe>>ing effects of collar plates and webs were ignorea and the pipe whip impulse loaas were treated as step functions co>>stant in time, revealed tnat yiela would occur in tne structure. Since the pipe whip dy>>amic a>>alysis was basea on the assumption of elastic, non-yielui>>g restraints, tne effect of yielding would be to reduce the loadi>>g peaks shown in Figures 3.6-58 and 3.b-59. However, the non-yielding assumption used in the pipe whip analysis is taken, as the more co>>servative approach. In either case, yielding or non-yield-ing, the steel frames will perform their function of adequately re-straining the pipes against excessive movement.

Assuming complete rigidity of the steel frame restraints and concrete, the bolts, sub)ected to a pulse (hat function) loading of 0.002 sec-onas duration were shown to reach a peak strain of 0.0155 in/in (based on a bilinear stress-strain curve in which Young's modulus E 30xl06 psi and the strain hardening modulus S-0.05E) for carbon steel. This strain is well below 1/2 c~ for carbon steel (taken as O.l) and is confined to the threaded portion of the bolt. Once more it is seen that under very cpnservative assumptions the bolts do not rupture, and that yielding results in lowering the applied pulse peaks due to pipe vhip.

Assuming, once more, that the restraint frames remain rigid, and that a step function load is applied through the bolt anchor plates to the concrete with a peak equal to the applied pipe whip impact-pulse dis-tributed to the embedded bolts, it vas shown that the concrete vould not fail in shear (i.e. ~ pullout) and that the concrete wall vas ade-quate to resist these loads.

In summary, the design of the pipe vhip restraints and embedments is considered adequate to perform their primary function of limiting pipe motion and secondary damage following a pipe break because there will be no intolerable loadings as a result of exceeding the factor 2.0 for the k load factor.

Maximum strains in the feedvater and main steam lines are found by adding the yield strain to the maximum plastic strain. These peak strains, in all cases, are considerably less than half the ultimate strain of the materials (main steam - steel, h 155 GR-KC 65; feed-water - steel, A 106 GRB).

P 3.6-11

Critical results of these analyses, are indicated on Figures 3 ~57 through 3,6-64 and on Table 3.6-2 Nor the four ruptures considered.

)

Four different breaks were analyzed, two on a main steam line and two on a feedwater line. The break locations chosen, restraint locations, geometry mater ia 1 proper ties, and maximum op erat in g temperatures and pressures for each break condition are shown in Figures 3.6-52;through 3,6-55 'he four breaks were chosen as representative of breaks pro-ng maximum impact reactions and pipe strains.

d uc ing In all cases cir-cumferential breaks were analyzed since the greatest potential for or whipping a pipe exists. Thrust forcesat the break locations were developed by performing a time history, thermal hydraulic analysis of the blowdown with the RELAP-3 code (Reference 14) suitably modi-

~

fied to predict thrust forces.

With the RELAP-3 code the transient energy, momentum, and state equa-tions were solved for an assembly of volumes and flow paths modeling the true piping system. The total thrust out of the break was eval-uated 'as the sum of three components:

2 a) A Momentum flux component equal to W /gc pA> representing the outflow of momentum out of the control volume about the break, b) a pressure force component equal to (Pe - P )A, representing unbalanced pressure forces on the control volume about the break, such as occurring when flow is choked at the exist plane, and c) an inertial component equal to W t+ht W t

g ht representing the thrust due to acceleration caused by the change of momentum with time within the control volume about the break.

Herein t is the time, W the mass flow rate, A the break flow area, Pe the critical pressure at the exit plan, Pa the ambient pressure,p the fluid density, gc the gravity constant, and L the length of the control volume chosen to represent the break. The velocity used in the calculations is either the inertial velocity (Bernoulli'a equa-tion) or the choking velocity as found from Moody's critical flow

.orrelation (Reference 12).

3.6-12

. Time dependent blovdowa forces are showa by curves in Figure 3.6-56, Empirical functions conservatively approximating this data,l ahovn in heavy lines on ligure3,6-56 ~ vere used in the dynamic analysis pipe

+hip program "PLhST" as inputi Gap data and spring constants for pipe vhip restrainta are as indicated in Tables 3.6-3 .through 3.6-6 for the four breaks chosen.

The "PLAST" program models the pipe run aa a lumped parameter system with elastoplastic material properties. The equations of motion of the system are solved by a step by step integration method in'the time domain using varying time steps to insure solution stability~

h pipe run is modeled as a lumped parameter system consisting of discretised "masses" and "springs" ~ The "masses" are represented by the physical mass and rotary inertia of the pipe while the "springs" are represented by pipe stiffnesses corresponding to the 6 degiees of freedom for every point along the pipe axis.

h section of pipe bounded by lumped masses at each end is defined as an "element". h 12 x 12, symmetric stiffness matrix may be vrit-ten for each such element. The individual terms of the matrix may be represented by the symbol "kig" vhere the i,5 subscripts refer to the rov and column locations respectively of the term wi'thin the matrix.

For a linear pipe element, the non-sero terms are given below:

ku. /L 5,9 26 9,5 17

~-k ~k "11,5 5,11 k22 12 E I/(L + L C) 66 55 k26 k62 -6EI/(L +C) 28 82 22 2,12 12,2 26 k33 "22 35 k53

- k26 m

39 3,11 11,3 35

~ GI /L k44 4,10 10,4 44 k5,5 ~ 4 E I/L - a 3.6-13

k68 "86 "26 6,12 '2,6

~

k ~

, kll 88 22 k266 8ol2 12o8 k99 k33 9,11 k26 10,10 44 ll,ll k55 12,12 k66 When E ~ Young's modulus G ~ Shear modulus L ~ Element length A Cross - sectional area of pipe metal I Cross - sectional moment of inertia in bending Ix ~ Torsional moment of inertia

~ 24 p (1 2 C + u )g u ~ poisson's ratio p ~ Shear factor (2 for pipe) r -Qz(

n ~ 2 3C/L + c EI/L 2

6EI/ L + c/L 4EI/L 4 L + c/4 y c)

(L This stiffness matrix includes the effect of transverse as well as Torsional shear; Stiffness matrices have also been developed for curved and "stepped" elements with appropriate "flexibility"factors applied.

3.6-14

The equations of motion for an element are written:

}M] }X}+ [Cj}X}+ [K]}X}- }P}

for linear elastic hehavior vhere:

g p {C] p [K]are the mass, damping and stiffness matrices.

s

}X}, }X, }X}, arc the displacement ~ velocity and acceleration vent or's and (F is a vector of forces acting at the element nodes (end masses) that keep the element in equilibrium. Hence in the absence of externally applied forces at a node these represent internal forces.

The equations of motion for the overall structural system are obtained after adding individual element stiffness matrices (referred to overall global axes). The'verall equations'of motion have the same appearance as (1) but )F represents a vector of forces acting on but external to the structural system.

~Dam in hn upper bound damping factor (16). is found for the range of periods between 0.025 seconds and the largest system period. This requires the determination of two constants cc and P such that:

Ci~ 2 PMi~ +aki~ (2) where: Ci~, Mi~, ki~ are the damping, mass and stiffness terms of the th i~os.and 5 th colnsm of the [Cg, [M] and[Kjmatrices respectively.

The damping matrix represented in (2) represents a conservative esti-mate of the system damping.

The materials that make an a piping system are considered to yield according to Von Mises(17).criteria,and are either elastic - perfectly plastic (zero-hardening) or harden "isotropically."( ) The constitu-tive laws are considered to be "incremental" in that they relate in-crements of plastic deformation to total stress at a point in body as follows: r dy dee)

'iS 2 G (3) where: 5 ( ) ~ increment of ( )

G ~ Shear modulus 3.6-15

r are the tensor components of strain and stress devia-tora e

ij, respectively

/

F ~ is defined by!

F fJij ~ J2 ~ 0 (4)

The Von Mises yield criteria are stated in the following way:

IF F >>2 then yielding occurs hF>o then elastic action occurs LLF go E q uation (4) ma be de icted as a rig ht circular c linder making equal angles with three principal axes representing principal stresses at a point. (lg) When yielding occurs, the plastic strain increment can be plotted on the same set of axes (with an appropriate scale factor) as a vector normal to the cylindrical surface. Isotropic hardening is represented as an expansion of the cylinder cross - section about its origin (i.e., isotropically). In order to establish hardening para-meters it is necessary to know the slope and shape of the uniaxial stress - strain curve beyond initial yielding.

The flow rule associated with the Von Mises yield criteria is not applied directly to the frayed structures. Instead, use is made of yield surfaces in "force space"( >Berived from the Von Mises yield surface. These surfaces are symmetrical with respect to principal force axes but lack point symmetry. A simplifying approximation to such surfaces may be made by utilizing their circumscribing sphere in force space. The equation of this space is given as:

2 2 2

~M + Mz + x +

1-MyP P FxP M MX Where: Mx, My, M are the internal moments about the designated axes.

e F ,M ,M ,M are the fully plastic values of these forces.

Fx is the interna1 axial force on a member.

a is the sphere radius.

The associated flow is then:

(6) 3.6-16

Where: h the icomponent of plastic displacement inhrement Fi the icomponent of internal force (limited to the components in eq..{5))

The total displacement of a point can be broken up into elastic and plastic components:

hXi ~hXi (e) +hXi (p)

Th en s Mce o nly y el a stic displacement contribute to force at a node, equation (1) becomes: (7)

[M] (X} + [C$ (X} + [g(X} (F}+ j'F where: f F f [K]fx is the "plastic correction force" developed internally for a yielding element, and assembled as a total correction force for the overall structure system. Expressions for X in equation (9) for cases of one(2O) and two nodes of an element yielding are shovn in detail in reference ~

These vere based on the folloving assumptions:

a) Small deformation, b) Concentrated forces applied only at nodes (masses) c) Yielding at a cross section occurs simultaneously over the entire cross - section or not at all d) There is no spread of yielding beyond the node along the beam axis e) The flow rule of eq (6) applies En the case of isotropic hardening, one seeks parameters that indicate the correct yield surface to use (in force space) for the flow rule, eq. (9).

Tovards this end the follovtng terms are defined for pipe beams:

a) Effective stress:

3'2 where:

J 2 ~ 1/3 ( Txx)2 ~ ( xo)2 b) - Effective plastic strain:

&1 Where:

2/3 (<<~) +2 ( 2 )

3.6-17

c) - Effective stress strain curve (I9

~c ~go H4 c~r (Initial Yield Stress) ES E Young's modulus S~Plastic modulus (Shown for a bilinear material)

~FX 1 strain effective stress-strain curve is a plot of stress vs plasticrepresents The

-for a uniaxial specimen. For a three dimensional analysis, it strain.

the radius of the yield surface plotted against effective plastic Therefore The area under this curve is the plastic work of deformation.

if the forces that give rise to yielding along this curve are known the expression for the yield surface in force space may be used to find plastic 'displacement increments as in equation (6).

Solution (21)

Solution of equations (8) is by the Newmark "Beta" Method using Beta ~

1/6 and a convergence rate of 0.1. The initial integration step is found internally as a fraction of the approximate value of the lowest period of the system. Solution stability is assured by maintaining an upper bound of 1/5 on the value of this fraction. Further improvements in the time step may be made by accounting for the lowering of natural frequencies of the system that result from yielding.

In propagating the solution through the time domain, no modifications are made to the initial stiffness and mass matrices. The problem, in short, is considered to be in the "small deformation" regime. However small deformations give rise to large"deflections and rotations. Hence, blowdown forces at severed pipes are made to follow the pipe movements.

Gaps at restraints, are treated as step changes in displacement force boundary conditions',. i.e. a node initially with a zero force specification in some direction suddenly changes its specification to zero displacement in that direction. Pipe whip restraints are modeled as bilinear, elasto-plastic springs of zero length and negligible mass. These "take a'ride" with the whipping pipe until the gap is closed.

3.6-18

0 ha each element of the systam is loaded, it deforms according to elas-Unloading occurs elas-tic and then slastoplastic constitutive lava.

tically leaving a residual "plastic displacement" in each of the yield ed elements.

Restraint models consist of. one or more "anchors" and elastoplastic external springs vith initial gapa; the latter representing the pipe whip restraints. hll hangers and earthquake restraints are con-sidered to have failed. Rebound velocities and impact forces are affected by gap sise, restraint and pipe material properties and 'sys-tem damping as follows.

Rebound velocities and impact forces at a restraint result from the instantaneous introduction of a displacement boundary)condition at the restraint. This boundary condition imposes displacement con-straints on a mass which has the effect of applying external forces on the mass, If the boundary condition nullifies displacements in[

rigid;".'f any direction, the restraint is considered'" displacements are a linear function of themselves, the restraint is considered "elastic." If displacements are governed by lave of one dimensional elastio-plasticity, the restraint is considered as an "elasto-plastic, strain-hardening" restraint. The sum of the rate of change in momen-tum of the mass due to the introduction of this boundary condition plus the viscous . forces in the mass plus the internal force of the attached pipe is the total force acting on the mass. The resultant total momentum change accounts for the instantaneous rebound velocity of the mass. These factors depend on the mass velocity at impact, the type of restraint (rigid, elastic or elasto-plastic), system damping and the stiffness properties of the pipe. Gap size affects the mass velocity at impact.

3. 6-19

The aaalysis, in each break case, vas allowed to run until ft vae observed that the loaded mass point oscillated vith decreasing ampli-tude about some displacement value. It vae noted, in all cases>ithat the ffrst peak in reaction force magnitude vas never subsequently.

superceded> even though displacement peaks vere reached after the rea-actioa peaks. Strain hardening in the piping system vas taken ae aero for conservatism.

The peak plastic strains in each system arc indicated in Figures 3.16-3,5,7 and 9. The total strain in each case can be obtained never by adding the yield strain to the maximum plastic strain and is seen to approach E/2, Results of pipe whip analyses performed on guillotine breaks fn tvo locations of the feedwater line on either side of the penetration and in tvo locatfons' of the main steam linc inside containment are shovn fa Table-3.6-2 In addition, sensitivity studies for the feedwater linc outside containment are shown for a change in gap and a change fn vali thickness. h reduction in gip reduces peak restraint reactions as vae expected. Decreasfng the wall thickness hovever seems to'have the opposite effect. The results of this sensitivity study are given in Figures 3.6-63 through 3.6-68 and Table 3.6-1..

The coefffcieats used to calculate the jet thrust force on the rup-tured p'ipe are based on the folloving:

1) Maximum theoretical values of the thrust coefficients have been predicted by Moody (Reference 12) under steady flov conditions to be 1.26 for steam and flashing vater, and 2.0 for subcooled ~ster. These values ignore frictional effects in the pipes and exit effects.

2) In real fluids friction and exit losses are present, there-fore the maximum theoretical coefficients have been modi-fied to account for such losses. Thrust forceps for several piping breaks fnvolving steam and feedwater vere derived by using the RECAP-3 thermal hydraulic code (Reference 14).

The results shov that at steady state flov conditions peak values of the thrust coefficients are close to unity for steam and 1.1 for flashing water. The particular values listed above of 1.01 and 1.12 respectively for stcam and flashing vater were derived for conditfons typical in pover

,plants using 350 psi exit pressure for saturated steam and low quality ( 1 percent) and'800 psi pressure at thc exit plane for feedwatcr.

3) For subcooled vater frictional effects arc accounted for by a resistance coefficient of 1.23 (Reference 11) 'tilizing

.4) To account for the more severe contraction present in the case of a slot break, a contraction coefficient of 0.61 has been chosen (Refcrcnce 13).

3.6-20

TABLE 3+6-l SENSITIUITV STUDY FOR 20" FQ LINE OUTSIDE CONTAIs&fENT Max. Plastic 'ax. Defi.

Node No.

At Gap Restraint'in (in) ~(

Mall Thick. Max. Reaction 6

Strain in At Loaded Node 5 2 .8221 x 10 .0065 6.381" 6

5 2 1.5 .68 x 10 .0072 5.446" 1.5 x 10 6'7806

.00148 9 255" 3.6-35

TABLE 3e 6"2

SUMMARY

OF PIPE MHIP ANALYSIS Line Description Guillotine Break Acti>e Restraint Maximum Deflection Maximum Reaction Maximum Plastic at Node Number Node Number at break at Restraint Strain in Pipe

( ounds x 106)

Feedvater Outside Containment 1.5" 9.255" 0 '80& 0,00148 eall and 4.00" gap Feedvater Inside Containment 3 502" ~ 1+34 0+00259 Hains team. Inside 2+72* (+X}

Containment 12 11.46" 2.34 (+z) 0+00762 9 1.0 (+Z)

Hains team Inside 13.36u 1,78 (+X) 0.0000 Containment 16 13 1.52 (+X,-Z)

• Values exceeding 2KPA have a duration of <2 milliseconds

TABLE 3.6-3 MAIN STEAM LINE INSIDE THE CONTAINMENT GUILLOTINE BREAK AT NODE 012 Res traint 'n.

Node I.D. No. El. S rin Const. Pl. S rin Const. ~Ga 5

MS-12 8.536 x 10 6 t/in 3.570 x 10 f/in 6.00 6 5 MS-13 8.786 x 10 0/in 3.650 x 10 t/in F 00 5

MS-14 8.786 x 10 6 0/in 3.650 x 10 0/in 6.00 MS-15 8.786 x 10 6 0/in 3.650 x 10 0/in 5.50 5

MS-16 8.786 x 10 6 0/in 3.650 x 10 0/in 5.50 5

MS-17 8.786 x 10 6 f/in 3.650 x 10 f/in 4.00 TABLE 3.6-4 MAIN STEAM LINE INSIDE THE CONTAINMENT GUILLOTINE BREAK AT NODE 816 Restraint Information Restraint Node I.D. No. El. S rin Const., Pl. S rin Const. Ga in.)

6 6 MS-12 8.536 x 10 P/in 3.570 x 10 0/in 6. 00 6

MS-13 8.786 x 10 6 f/in 3.650 x 10 g/in 6.00 6

MS-14 8.786 x 10 6 0/in 3.650 x 10 f/in 6.00 6

MS-15 8.786 x 10 6 f/in 3.650 x 10 0/in 5.50 6

MS-16 8.786 x 10 6 f/in 3.650 x 10 0/in 5.50 6

MS-17 8.786 x 10 6 0!in 3.650 x 10 f/in 4.00 6

13 MS-18 8.786 x 10 6 0/in 3.650 x 10 0/in 4 40 3.6-37

0 TABLE 3.6-5 BOILER FEEDWATER LINE OUTSIDE THE CONThINMENT GUILLOTINE BREAK hT NODE 07 Restraint Information Node El. S rin Const. Pl. S rin Const. ~Ga in.

8.536 x 10 6

t/in 3.570 x 10 6

f/in 4.00 6

8.536 x 10 6

8/in 3.570 x 10 t/in 4.00 6

8.536 x 10 6

t/in 3.570 x 10 0/in 4.00 TABLE 3. 6-6 BOILER FEEDWATER LINE INSIDE THE CONTAINMENT GUILLOTINE BREAK AT NODE t7 Restraint Information Node El. S rin Const. Pl S rin Const. Ga (in.

6 6 8.786 x 10 ,f/in 3.650 x 10 0/in 2.50 6

8.786 x 10 I/in 3.650 x 10 0/in 4.00 3.6-3S

+X i?p lA 13 (HS-2) 16 (MS-3) 37 4

(MS-4) 5 ~+ niL C) 19 15 18 + 14 H 21 O 23 25 (MS-5) 8 27 MATERIAL - CARBON.STEEL PRESSURE - 985 PSI TEMP. - 520 4F E< = 26.26 x 10 PSI 32 YLD STR. = 27,800 PSI (HS-6) 9 refers to restraint 31 location in Fig.3.6-36

+30 restraint location fictional node locations

. 33 required for pipe whip analysis 12 34 FLORIDA POWER 8 LIGHT COMPANY REF. ISOMETRIC: MS <<147-1 ST. LUCIE PLAHT UHIT 1 MAIN STEAM LINE INSIDE CONTAINMENT GUILLOTINE BREAK AT NODE ¹12 FIGURE 3.6-52

pp Pp~

+3 X

24 4.5 O

I I

lV 23 19 X LLl I

18 P" 33 27 366 32 33 35 9 10 MATERIAL CARBON STEEL PRESSURE - 985 PSI 37 TEMP. 520 4F 26 26 106 PSI Eg f60 YLD. STR. = 27,800 PSI 13 13 (66 4p 15 42 pc FLORIDA POWER 8 LIGHT COMPANY ST. LUCIE PLAHT UHIT 1 16 REF. ISOMETRIC: MS-147-1 MAIN STEAM LINE INSIDE CONTAINMENT GUILLOTINE BREAK AT NODE 016 FIGURE 3.6-53

+CG.

oo 0 2

3 kk 4'.

.p~

7 6

5 ~

Q gPgb'2 ~ gb 18 16 12 10 8 19 I 13 15 9

MATERIAL CARBOH STEEL PRESSURE - 1100 PSI TEMP. 4404F E< = 26.76 x 10 PSI YLD. STR. = 29,320 PSI REF. ISOMETRIC: BF -149-1 FLORIDA POWER 8 LIGHT COMPANY ST. LUCIE PLAHT UHIT 1 BOILER FEEDWATER LINE OUTSIDE CONTAINMENT GUILLOTINEBREAK AT NODE ¹7 FIGURE 3.6-54

0 10 MATERIAL: CARBON STL.

PRESSURE: 1100 PSI TEMP: i% oF E = 26.76 x 106 PSI H

YLD. STR.= 29,320 PSI REF. ISOMETRIC: BF -147-1 FLORIDA POWER 8 LIGHT COMPANY ST. LUCIE PLANT UHIT 1 BOILER FEEDWATER LINE INSIDE CONTAINMENT GUILLOTINE BREAK NODE P7 FIGURE 3.6-55

~ ' '

I ~ ~ ~

I ~

2KPA = 2.07 X LBS 2.0 1.2 I

LEGEND FOR HODE > AHD

.e 0,8 l

REACTIOH DIRECTIOH.

HODE 9 (+Z)

HODE 9 (+ X)

OA NODE 13 IH DIRECTIOH OF VHIT YECTOR.

Y =087 X. 0.742 Z 0.1 0.15 0.2 FLORIDA POWER 8 LIGHT COMPANY ST. LUCIE PLAHT UHIT 1 MAIN STEAM LINE INSIDE CONTAINMENT (BREAK AT NODE 16) REACTIONS AT PIPE WHIP RESTRAINTS (NODES 13 8 9)

VERSUS TIME FIGURE 3.6-57

2,72 X 105-"

2.5 I 2.34 X106.- ~.

2.

2KPA = 2.07 X 106 LEGEHD FOR DIRECTION

=

I I

~

t~

i spring reoction I

(~+Z 1.5

(+ X)

QJ oO K

I

'cn 1.

X, O

I tJ LIJ CL 0.5 I~

I I

0.

iI

.01 TIME (SEC.)

F'RiDA POKER Il LIGHT COMPANY ST. LL'CIE PLAHT UHIT 1 3" 'A.~i STEA" 'HE CONTAINMENT

'r":". AY. AT IiO'" 12, iNSIDE REACTION AT PIPE 3'siP RESTRAINT AT IiODE 9 IN X &, tZ DIP.EC !OHS VERSUS TIME FIGURE 3. 6-58

1.6 1.2 (2 KPA)

CI 0.8 CC O

R O

gr 0.4 0.

0.01 0.1 TIME (SEC)

FLORIDA POWER 8 LIGHT COMPANY ST. LUCIE PLANT UNIT 1 FW LINE INSIDE CONTAINMENT REACTION AT PIPE WHIP RESTRAINT AT NODE 6 VS. TIME FIGURE 3. 6-59

12.

10.

8.

Ill O

I 6.

Ul EJ 0

an CI 0.

0.01 0.1 TIME (SEC.)

FLORIDA POYIER S LIGHT COMPANY ST. LUCIE PLANT UNIT 1 MAIN STEAM LINE INSIDE CONTAINMENT BREAK AT NODE 16 DISPLACEMENT OF NODE 16 VERSUS TIME MAX. PLASTIC STRAIN =0.

FIGURE 3. 6-60

11.46 "

0.1 TIME (SEC.)

FLORIDA POWER 8 LIGHT COMPANY ST. LUCIE PLANT UNIT 1 34" MAIN STEAM LINE INSIDE CONTAINMENT (BREAK AT NODE 12)

DEFLECTION OF NODE 12 VERSUS TIME MAX. PLASTIC STRAIN =.00762 FIGURE 3.6-~>

2.5" GAP 1.031" WALL MAX PLASTIC STRAIN IN PIPE ~ 0.00259 4.

O O

R CO z

O 30 N

o Ch X

20 O

d) z O

I I-0 CC 1~

0.001 .01 TIME (SEC)

FLORIDA POWER II, LIGHT COMP@I "I ST. LUCIE PLAHT UHIT 1 20" FW LINE INSIDE COHTA!I'MEI"T ROTATIOH OF NODE 7 VS. TIME FIGURE 3. 6-62

0.8 4" GAPS -1.5" WALLTH MAX. FORCE ~ 0.7806 x 10 (2 KPA ~ 1. x 106 LBS}

8 0.04904 SEC 0.6 CO 40 tll CP a 0.4 O

ll R

D I

V 0.2 0.

~

0.01 0.1 TIME ISEC}

FLORIDA POWER II LIGHT COMPANY ST. LUCIE PLANT UNIT I 20" FW LINE OUTSIDE CONTAINMENT REACTIONS AT PIPE WHIP RESTRAINT AT NODE 5 VS. TIME FIGURE 3.6-63

O 6:

9.255" ..

9.074" INITIALYIELD (4" GAP 1.5" WALL THICKNESS) 6 NODE ¹5 MAX PLASTIC STRAIN

{0.04904 SEC) IN PIPE ~ .00148 0.1 TIME {SEC)

FLORIDA POWER 8 LIGHT COMPAtv'Y ST. LUCIE PLANT UHIT 1 20" FW LINE OUTSIDE COt<TAlt:.".EWT DISPLACEMENT AT NODE 7 VS. TIME FIGURE 3. 6-64

{2" GAP 1" WALLTHICKNESS) 2 KPA ~ 1 x 10 LBS.

0.8721 x 108

.8 ED a

.6 g

Lll CC O

LL R

O ,4 L

V C

Lll 0.

0.01 0.1 TIME {SEC)

FLORIDA POWER 8 LIGHT COMPANY ST. LUCIE PLANT UNIT 1 20" FW LINE OUTSIDE CONTAINMENT REACTIONS AT PIPE WHIP RESTRAINT AT NODE 5 VERSUS TIME FIGURE 3.6-6s

0 el

(2" GAP 1" WALLTHICKNESS)

MAX. PLASTIC STRAIN IN PIPE 0.0065

~

c5 0 "I 6 "I 6.381" 6.203" 0.1 TIME (SEC)

FLORIDA POWER 8 LIGH1 COMPANY ST. LUCIE PLANT UNIT 1 20" FW LINE OUTSIDE CONTAINMENT DISPLACEMENT OF NODE 7 VS. TIME FIGURE 3.6-66

(2" GAPS 1.5" WAI L THICKNESS)

(2 KPA ~ 1. x 10 LBS)

.6 o

O K 4 O

R O

I CJ K

0.01 TIME (SEC)

FLORIDA POWER L LIGHT C'3MPANY ST. LUCIE PLANT UNIT 1 20" FW LINE OUTSIDE CONTAINMENT REACTIONS AT PIPE WHIP RESTRAINT AT NODE 5 VS. TIME FIGURE 3.6-67

5.446" 5.434" O

Ill I

Ill a EO EO CI O

U Ill CO ill 0)

CO Ill t o

O R

3 (2" GAP 1.5" WALLTHICKNESS)

K O MAX PLASTIC STRAIN IN I PIPE ~ .0072 UJ IL IQ Q

2e CP I

0 0.01 0.1 TIME (SEC)

FLORIDA POWER & LIGHT COMPANY ST. LUCIE PLANT UNIT 1 20" FW LINE OUTSIDE CONTAINMENT DISPLACEMENT OF NODE 7 VS. T I>E F I GUR E 3. 6-68

Attachment C PLAST Information from Reference 9

~

0 PIPE MATL. PROPERTIES 43" 0-D

+Y 1225" DONALL TH A106 GRADE C EH ~ 26.08 x 10 PSI ANCHOR 5.19' DESIGN PR ~ 1085 PSI OP TEMP ~ 545 F 7.41' 2.13' ll X 3.50.50'.66' ill 3.81'.09'IRCUMFERENTIAL o

lA

~ Oi 0 10 BREAK 12 200' 1 4

~

2.67'.75'.75'.50'IRCUMFERENTIAL 2.00'.00'R BREAK IN MAIN STEAM LINE FIGURE: 2.3. 1

SEE FIGURE FOR ORIENTATION WITH GLOBAL AXES.

3.89" 8

2.57" 0

EFFECTIVE GAPS AT RESTRAINT AT NODE 14 FIGURE: 2.3.2

P (FORCE)

(OISPLACEMENT)

EFFECTIVE GAP SPRING e (s (ps(i 'p (ps(( 6y (in) 1 5.95 x 106 2.99 x 104 0.15

~p 2

3 5.95 x 106 0.20 6.95 x 10 0.15 0 FORCE - DISPLACEMENT DIAGRAM FOR RESTRAINTS IN 43" MAINSTEAM LINE FIGURE: 2 ~ 3 ~ 3

195. x 10 LBS.

. 120 100 80 0

0 .1 .2 TIME (SEC)

BLOWDOWN FORCE VS. TIME AT NODE 16 IN MAIN STEAM LINE FIGURE: 2 3 4

0.10 0.15 0.20 0.25 020 TIME IsEGI DEFLECTION OF RESTRAINT (NODE 014)

IN (+x) DIRECTION FIGURE: 2.3.5

g 140 120 o 80 60 40 20 0

0 0.05 0.10 0.15 . 0.20 0.30 TIME (SEC)

) FORCE ON RESTRAINT AT NODE 4P14 FIGuRE: 2.3. 6

+Op g.I" '8g 4.25'R 4.25'R

+Z 4 +X Qe 4

O.D. ~ 34.0" 13 t -1.250" O.D. ~ 36,625"

~ 1.875" o oo S

00 0 CI CV t

0 Cj H

0- 10

~a+ MAT'L.- CARBON STEEL 0 PRESS. - 985 TEMP EH 520oF

-26.26 X106 Psl

~)+

YLD. STR. ~ 27500 Op 0

12 RESTRAINT ORIENTATIONS NAIN STM. LINE TRANSVERSE BREAK AT NODE ¹12 FIGURE: 2,3,7

12.

11.46" U

LLL CO

10. CD 8.

CO LLJ K

Z R

O 6.

I V

LLJ LL MAXIMUMPLASTIC LLL O STRAIN ~ 0.0762 4.

2.

0.01 0.1 0.2 TIME (SEC)

MAIN STEAM LINE (BREAK AT NODE 12) DEFLECTION OF NODE 12 VERSUS TIME FIGURE. 2.3.8

~ ~ ~

~ ' ~

1 ~ ~ ~

~ ~ ~ 0

ti

.W I 3

~: ~ ~ ~ ~

0

p(x, t) Pt s

l sa10'06 t - {SECONDS)

~s 30x'l0 psi 28.1 in P as 10 ii/SEC.

aa 0.25 W 18:9 <</FT.

wL 9

ELASTIC BEAM UNDER UNIFORM DYNAMICLOAD FIGURE: 3 1 1

P ~ 10 t (e/F T) 5O1'x5'YMMETRY I 0

ANALYTIC Cl NUMERICAL time (sec)

X 0.4406646 E K

4 1

0 r,

time (sec) 0.7115841 E R

2 1

0 r,

Cl time (sec) 2 X 0.9862027 E4 Z

4 1

0 time (sec) 0.1564142 E-3

'DEFLECTION HISTORY UNIFORM SIMPLE BEAM FIGURE: 3,g,2p

Fy 60" 60" 60M 60" 60" 60" 60" FIXED 60et z

60lt 10 FIXED 10 MASS PLANE FRAME

ELASTIC- IC FRAME LEGEND:

ANSYS'AP P LAST (7 MASSES)

PLAST I'IO MASSES) 10 I

a 0 /I I- 6 Ill ll

/

LIJ O ~

~ /

0

/

ly I

lQ 4

/rrrr re INITIAL YIELDING 0 .2 ,4 TIME (SEC)

FIGURE: 3.1,5A

ELASTIC-PLASTIC AME LEGEND:

ANSYS FAP PLAST (7 MASSES)

PLAST I10 MASSES) z0 I

O Ill Q

Lll D

CO

~

~ /r/

>o4 I-0 M

K O INITIAL K YIELDING

.2 .4 .6 1.0 TIME (SEC)

FIGURE: 3.1.5B

Vo] 0 u -Vo PLAST

~ Iwm~ee REF (15]

a~a~~m~~ REF (15]

Qy

- 45 PLAST E " 30.x10 psi

~

28.1 in 4 U

~

~ V'Sin 7l' W ~ 19'/1

- 10 V 4Vlr Vo - 959m in.lsec 20

.15 M t

~MR v ggF lb ~~4 Appvox ma@'ot~~""'~r Xmpvhiydl~ LO@pl El~iti'(- 8 "~~i< ~<<<mS" 4'8- Marin L5$ Leei 3'pp ><~~- lP~per4"488/AP~->3)

CENTRAL VELOCITY VS. TIME FIGURE:3 1. 6A

U' V~ SIN REF (15) 5 PLAST mweeaeeeeammw>> PLAST 200 400 500 4 ee x 0 0

D oc 4 6.1 247 (PLAST) 6.130 (REF f15))-

'"" "~y i<<'~ te sulu/ ong I ~puls,vol MV~R'~"

~~10 Mot 4+ marh>> ZSS Fg lucioS/ppl ncc L~grg +4sl< P/~y'- p+~<"

(Pappr+dg led/gpjp gg)

~ ~ 20 q .

CENTRAL DEFLECTION VERSUS TIME FIGURE: 3.1.6B

Attachment D Appendix B from Refernce 1

APPENDIX B DESCRIPTION OF CALPLOTFIII 93

B-I Mathematical Model The CALPLOTFIIIcomputer code has been written to convert the transient flow conditions calculated Sn a piping system by the RELAPSED 1 Computer code into transient forces on the piping system. Specifically, CALPLOTFIIIcalculates and plots the forces on straight lengths of pipe between changes in pipe direction (bends), or between a change ln direction and a pipe break. The derivation of the equations used in the code are given below.

B-I-1 Strai ht Len ths of Pi es ~tween Directional Chan es The force on a straight length of pipe between direction changes (Figure B.l) is calculated using the momentum equation:

F + Bpdv ~ V (pV ~ dA) + V (pdv) (Bl) cv cs cv If the gravity term is assumed negligible, the following equation results:

7s ~ ~V (pV ~ dA) +

at / V (pdv) (B2) cs cv Since, the force on the straight pipe length only exists in one dimension, the above equation can be written in a scalar form:

F 8

V (pV ~ dA) +

at

'pdv (B3) cs cv Since the RELAP5 MOD 1 Computer code calculates the pressures and the flowrates at different physical positions in the piping system, it is necessary to subdivide a piping length into two control volume types for application of the momentum equation. The first division creates the pressure control .volumes. The divisions for the pressure control volumes are thepositions in the pipe length where the pressures are calculated by the computer code, and serve as the boundaries across which the control volume surface forces are calculated. The second control volume divisions are due to flow conditions. The boundaries of the flow control volumes are located at the pipe length locations where flows are calculated by the computer code. The forces in the pipe length which are due to the rate of efflux of momentum across a control volume 0 and the change of momentum in a c'ontrol volume are calculated using the flow boundaries as flow control volume divisions.

94

The resultant force on the fluid across the boundary of the pressure control volumes 1, 2, and 3, shown in Figure B.l, are:

'Sl " ('A - 'a) AA + '1 (B4)

F PAA P A + P (A -AA) (B5) 2 S3 B a "B (B6)

The net surface force on the straight pipe length is obtained by summing equations B4, B5, and B6:

FS1 + FS2 + FS3 + 2

+ R (B7) 1 F ~ R (B8)

S Therefore, the force on the straight pipe length due to surface forces is equal to the net normal and shear 'stresses on the pipe wall length, The right side of equation B3 can now be evaluated for each of the flow control volumes F

Sl P2 g

A and B:

2 2

A A

+ Bt A

bA (B9)

F S2

-P2 VZ g

2 WSg + at hB (B10)

Since the RELAP5 computer code calculates non thermal equlibrium conditions for two phase flow conditions and allows the two phases to possess different velocities, the parameters of equations(B9) (Blp) are defined as:

M A

~

11AV1A1A (lmA + PgAV~~A AA (Bll)

M (p V~ (l~B) + pgB gB B) (B12)

+ V Pg2 g2 2 n (B13) 2 2 12 12 2

Summing equations B9 and B10, and using equation B8, the net fluid force on the pipe length can be obtained:

K ~ -F S

~ -R ~

-22'a, at hA - Bt bB (B14)

If the straight length of pipe considered is bounded by a directional change and an. open end, a break, the forces obtained using equation Bll must be modified to accoun'or the force developed at the pipe exit plane Consequently, using the momentum equation, the force on the straight pipe length shown on Figure B,2 ~ for unchoked break flow, can be written as:

2 SMA K hA (B15) unc g Bt If choked break flow is determined to exist by the fluid transient computer code, then equation B15 must be modified to account for the pressure unbalance that occurs at the pipe exit plane. A rederivation of the equation for the straight pipe length for this case results in the following relation:

K ch

~ -(P 2 -P)a A

-P 2 g

V2 2

2 A

-aa A Bt AA (Bi6y or K

ch K unc

- (P22 P a) (B17) where:

2 2 f - -

PA A P2 2 P P + hP 5P acc bP el (B18) 2 A 2g ~ 22 2 V2 P =P A +PAA P22 2

2g 2g (B19) pV2 =

p V 2

(lc)+p A V

2 gA gA A n (B20) 96

Nomenclature flow area body force of a control volume P

s surface force resultant on a control volume g gravitational constant K force of fluid w piping M control volume flowrate pressure P

a pressure outside pipe control volumes R normal and shear stresses in a control volume time volume of a control volume V velocity of fluid in a control volume Greek Letters density in control volume void fraction Subscripts acc acceleration friction choked flow cs control surface CV control volume el elevation unc unchoked 1 liquid gas 97

Figure B.l VOL. 2

( + FORCE LEGEND:

PRESSURE BOUNDARY

---

FLOlN BOUNDARY

~ '

~

o ~ ~ s

S The submittal does not. explain what loading combinations were considered in the analysis to determine acceptability of the piping system, therefore, identify the load combinations performed together with allowable stress e

limits for piping and supports both up'stream and downstream of .the valves.

The letter of September 1, 1982, indicates that the ANSI B31.1 code was

~ ~

"~

used to evaluate stresses in the piping.and supports unique to the PORV's.

I Identify all other governing codes and standards (with date of edition) used to determine piping and support adequacy.

RESPONSE

I The load combinations and stress allowables which are used to determine the adequacy of the Turkey Point pressurizer refief piping system are identical to the criteria specified in the Turkey Point FSAR by which the plant was

'I designed, specifically that is the non-seismic ANSI .B31.1 (1955) criteria I

(for comparison purposes the calculated loads which are shown in Tables 4.3.1

'nd 4.3.2, attached to the response to Question 9, were combined according to the EPRI recommended load combinatio'n as provided in Reference 13 and shown in Table 2. Tables 4.3.1 and 4;3.2 were developed for piping upstream and downstream of the PORVs subjected to PORV actuation.'he stress allowables for these tables correspond to'References 5 and 6, accordingly and are in-

'I eluded in Table 2).

The load combination's used to analyze the piping sub]ected to SRU actuation correspond to EPRI recommended load combinations as shown in Table 2. For

. \

~

~

~

piping upstream of the SRV valves the stress allowables are as shown in Table 2. For piping downstream of the SRVs and subjected to SRV actuation, the moment developed in the piping is compared to 70 percent of the ultimate itself II moment "carrying capacity of the pipe. The ASME code allows use of this value to satisfy integrity criterias. For a. further discussion of this point see the response to Question No. 9;

~

~ ~

r

()

I