3F0286-09, Forwards Response to 850826 Request for Addl Info Re Performance of Pressurizer Relief & Safety Valves Installed at Plant Per NUREG-0737,Item II.D.1
| ML20214C743 | |
| Person / Time | |
|---|---|
| Site: | Crystal River  |
| Issue date: | 02/17/1986 |
| From: | Westafer G FLORIDA POWER CORP. |
| To: | Stolz J Office of Nuclear Reactor Regulation |
| References | |
| RTR-NUREG-0737, RTR-NUREG-737, TASK-2.D.1, TASK-TM 3F0286-09, 3F286-9, NUDOCS 8602210253 | |
| Download: ML20214C743 (54) | |
Text
.
t
- e..
e
.Z e
e.e Power C O R PO R ATIO N February 17, 1986 3F0286-09 Director of Nuclear Reactor Regulation Attention: Mr. John F. 3tolz, Chief Operating Reactors Branch #4 Division of Licensing U.S. Nuclear Regulatory Commission Washington, DC 20555
Subject:
Crystal River Unit 3 Docket No. 50-302 Operating License No. DPR-72 Pressurizer Relief and Safety Valves NUREG-0737, Item II.D.1
Dear Sir:
Your letter dated August 26, 1985 requested additional information regarding the performance of the pressurizer relief and safety valves installed at Crystal River 3. The attachment to this letter provides the responses to those questions.
This submittal has been delayed as documented in our letters dated October 14, 1985 and January 16, 1986. Should there be any questions, please contact this office.
Sincerely,
/ l G. R. Westafer Manager, Nuclear Operations Licensing and Fuel Management AEF/feb Attachments f
8602210253 860217 l I I PDR ADOCK 05000302 P PDR (
G EN ERAL OFFICE 3201 Thirty fourth Street South e P.O. Box 14042, St. Petersburg, Florida 33733 e 813-866-5151
t t QUESTION NUMBER 1 The submittal does not explain how fluid transient cases other than the February 26, 1980 transient that occurred at Crystal River Unit 3 (CR-3) were considered in qualifications of the safety valve /PORV system. Since the February 1980 transient may not be the limiting transient, consideration of these other transients is important. The B&W Valve Inlet Fluid Conditions Report indicates that fluid conditions for overpressure transients in the CR-3 plant have been identified.
- 1. Veri fy that the fluid conditions identified in the B&W report were determined through analyses of accidents and operational occurrences referenced in Regulatory Guide 1.70, Revision 2, as required by NUREG-0737 II.D.I.
- 2. Specify the limiting fluid conditions expected for steam and liquid flow through the safety valves and PORV.
REPLY NUMBER 1 The B&W Report (EPRI-NP-2352) specified enveloping temperatures and pressures for the valve test program based on the FSAR analysis. The FSAR analyses were completed prior to the issuance of Regulatory Guide 1.70, Rev. 2 and NUREG 0737. Consequently, the fluid conditions were derived from the FSAR accident analyses. In some cases, the FSAR events may be the same as those specified in Reg. Guide 1.70 and NUREG-0737, but the analyses were performed in accordance with earlier requirements.
The enveloping fluid conditions expected for the safety valves are specified in Table 5-1 and for the PORV are specified in Table 6-2 of EPRI NP-2352. It is important to note the manner in which the fluid conditions in the EPRI report were determined. The way in which these fluid condition values were determined is as follows:
- a. An SLB transient was chosen because SLB's result in the minimwn Reactor Coolant System (RCS) temperature. This minimum temperature is approximately 490 F at 550 psia as shown in the CR-3 Final Safety Analysis Report, Figure 14-27. The 400 F that is reported in EPRI NP-2352 is a conservative value known to bound SLB events. EPRI NP-2352 states that the 400 F is conservatively low on page 4-33. Even given a 400 F minimum RCS temperature the pressure at this temperature would not be 2500 psia and would, therefore, not challenge the PORV or safety valves,
- b. SLB's do not result in maximum pressurizer insurge rates and RCS pressures. The accident that results in maximum insurge rates is a feedwater line break. The conditions for a feedwater line break were used to achieve the insurge (i.e., pressurization) rates at 2500 psia.
i r
/. r
- To summarize, minimum RCS temperature is achieved during an SLB. However, at this temperature RCS pressure is approximately 550 psia, not 2500 psig. To choose pressurizer fluid inlet conditions of 400 F and 2500 psig ls an incorrect application of fluid conditions from two different accident analyses. More appropriate conditions to choose for minimum pressurizer fluid temperature would be 490 F at 550 psia. The minimum RCS temperature during an SLB at which^ the RC pressure is 2500 psig is 517 F (Babcock & Wilcox calculation 32-1159703, dated 1/21/86).
For additional information, see reply to Question No. 2.
QUESTION NUMBER 2 The EPRI report concerning examination and tests of the CR-3 PORV and safety valves discusses performance of these valves during and after the February 26, 1980 transient. This report provides results for the case of saturated steam flow through the PORV and water flow through a safety valve, but this transient may not envelop all limiting transients identified under requirements of NUREG-0737 II.D.1. Thus, results from the EPRI test program! are needed to complete qualification of the PORV and safety valves. The submittal does refer to several generic documents concerning this program and claims that these reports document successful performance of the CR-3 PORV and safety valves. To justify this evaluation, show that the fl uid inlet condi tions ' detergined for limiting transients were enveloped in the EPRI tests. Show that the test results verify that the safety valves and PORV will open and close under expected flow conditions. Along with other flow conditions, identify the expected total length of time of liquid flow through the safety valves for a transient such as that of February 1980 or other liquid flow transients (see Question 1). Provide assurance that the valves functionability will not be imparied by liquid flow of this duration. Demonstrate the PORV will operate properly over the range of fluid conditions expected for cold overpressurization
. events.
REPLY NUMBER 2 _
The bounding fluid conditions identified are contained in Reference 2 and Babcock & Wilcox calculation 32-1159703-00. They are summarized below, along with the reference to applicable table in Reference 2.
PORY
- 1. Cold Pressurization Events: Makeup control valve fails open. PORV setpoint % 550 psig open, 500 psig closed. Fluid state at inl et. :
Saturated steam at 565 psia (Table 6-3 of Reference 2).
~
g
- 2. Extended HPI Operation Following an FSAR Steam Line Break: PORV set to open at 2450. Maximum source pressure 2500 psig. Water temperature 602*F. Initial flow of steam followed by subcooled water at 517 F mininum
, temp; PSV's also lift when water is being relieved (Babcock & Wilcox calculation 32-1159703-00).
o
r *
- 3. FW Line Break: PORV set to open at 2450 psig. Maximum source pressure 2500 psig. Temperature of fluid, 650 F maximum - 602 F minimum. Steam flows initially and subcooled water subsequently (Table 6-2 of Reference 2).
- 4. Rod Ejection at Hot Zero Power: PORV inlet condition: Steam at 2662 psig (Table 6-1 of Reference 2).
- 5. The maximum required steam relief rate for the PORV due to accident / transient conditions occurs in the " pressurizer heaters erroneously energized" event. This value is 7555 lb/hr of saturated steam (Table 4-19 of Reference 2).
Note that the PORV is not required to pass a given quantity of steam for any of the FSAR events because the PORY is not needed for the pressure boundary safety but only for operational control of pressure. Therefore, in all analyses concerning pressure boundary safety, the PORV is assumed to be closed. This does not include cold pressurization events.
Based on the above enveloping events, the following conditions for PORV operation are considered to be reasonable.
PORV Setpoint: Normal: 2450 psig open. For makeup control valve failure event: 550 psig open, 500 psig close.
Maximum Source Pressure: 565 psia saturated steam for the makeup control valve failure event. 2500 psig saturated steam and water at 602*F and 517 F in the steam line break event. 2500 psig saturated steam and water at 650 F for the FW line break. 2662 psig saturated steam for rod ejection hot zero power event.
Maximum Steam Flow: For a strictly PORV event, 7555 lbs/hr. This occurs in the
" pressurizer heater energizing" event. Higher relief may occur in safety related events, but in these analyses, the PORY is assumed to be closed.
Maximum Water Flow: 517*F water at 6019 lb/ min. These flow rates will open both PSV's and PORV.
Pressurizer Safety Valves (PSV):
- 1. Steam Line Break: PSV's set at 2500 psig. Initial steam flow followed by
- subcooled water. 602*F, maximum water temp., 517*F minimum. At 602*F,
! 6555 lb/ minute of water. At 517 F, 6019 lbs/ minute of water, i
, 2. FW Line Break: PSV set at 2500 psig. Steam relief with subsequent water relief (Table 5-2 Reference 2). At 640"F, water flow is 11.520
- lbs/ minute. At 602 F, water flow is 10,400 lbs/ minute.
- 3. Rod Ejection at Zero Power: PSV opens at 2500 psig source pressure.
Maximum source pressure is steam at 2662 psig. This event defines the maximum system pressurization rate of 175 psi /second.
l l
l i
r
- Comparison With EPRI Results PORV EPRI tests conducted to qualify PORV's at Wyle Laboratory, and Marshall Station as applicable to CR-3 used the Dresser valve Model 31533 VX-30-2. Reference 1 gives th.' results of the tests. Relief of water and steam at the identified inlet conditions do satisfy the expected performance characteristics of the PORV without failures. Table 1 presents a summary of the EPRI PORV performance.
The PORV is not a safety-related compon2nt. Analytically derived maximum flow rates based on PSV flow rates and other conditions are perhaps overly conservative to apply to PORV's. Functionally, the PORV tested performed satisfactorily.
When water relief is expected through the PORV, the PSV's will also be open and relieving water, and if the PSV's are tested to relieve the total quantity of water, PORV relief is redundant. This is the case for CR-3. The test results indicate a minimum water relief of 262,800 lbs/hr at a pressure of 692 psia.
Post leakage was acceptable (<0.0013 gpm measured) indicating survivability of the valve in passing water over a period of time. The period of the tests with water ranged from 10 to 15 seconds (Reference 1).
PSV's The Dresser Model 31739A PSV tested performed as required under analytically identified conditio1s that bound CR-3. The results are given in Reference 1.
Table 2 presents a summary of the EPRI PSV performance._
i
I TABLE 1
SUMMARY
OF IMPORTANT ASPECTS OF THE DRESSER PORV PERPORMAf?CE IESTING (REFERENCE 1)
- 1. Total No. of Runs 12
- 2. No.' of Steam Runs 3
- 3. No. of Water Runs 6
- 4. No. of Steam / Water Transition Runs 1
- 5. No. of Water Seal Simulation Runs 3
- 6. No. of Runs Where Valve Opened on Demand 12
- 7. No. of Runs Where Valve Closed on Demand 9
- 8. No. of Runs Where Valve Did Not Close on Demand 1 3
- 9. No. of Runs With Steam or W5ter at a Temp. Greater Than 600 F 9
- 10. No. of Runs With Water at u Temp. of Between 4500 F and 460 0 F 2
- 11. No. of Runs With Water at Approximately 1000 F 1
- 12. No. of Runs With 25,500 in-lb (2125 ft-lb)
Bending Mcment 1
- 13. No. of Runs Whose Post-Test Leakage Rate Exceeded The Pre-Test Leakage Rate 3
- 14. No. of Runs Whose Leakage Rate at Any Time Exceeded
- 0.026 GPM 0.0 1
1 Only those runs simulating water seals did not close on demand. CR-3 does not have a water seal, therefore the informa-tion f rom these runs does not apply.
l
, i TABLE 2
SUMMARY
OF SIGNIFICANT ASPECTS OF THE DRESSER SAFETY VALVE PERFORMANCE 31739A (Tested Valve)
- 1. Valve Inlet Dia, in. 2.5
- 2. Valve Outlet Dia, in. 6.0
- 3. Bote Area, in 2 2.545
- 4. Min. Lift, in. 0.45
- 5. ASME Fated Steam Flow 02575 psig, lb/hr 297,845
- 6. No. of Test Runs w/ Steam ,
25
- 7. No. of Test Runs w/ Steam -
Water Transition 2
- 8. No. of Test Runs w/ Water 5
- 9. No. of Test Runs w/ Loop Seal Configuration 12
- 11. No. of Runs w/ Unstable Valve Performance 2
- 12. No. of Runs Where Valve Failed to Open 0
- 13. No. of Runs Where Valve Stuck Open 0
- 14. No. of Runs Where Post-Test Leakage Exceeded Pre-Test Leakage 12
- 15. No. of Runs Where Leakage Exceeded 1 GPM 3
- 16. No. of Runs Where Leakage Exceeded 2.5 GPM 0 9
g '* e I
TABLE 2 (cont.)
- 17. No. of Runs w/Back Pressure Greater Than 520 psi and %
Rated Flow Greater than 100% 9
- 18. No. of Runs With a Blow Down on Steam Greater Than 17% w/no Loop Seal 0
- 19. No. of Runs on Steam With a Pressure Increase Rate Greater Than 250 psi /sec w/No Loop Seal -
13
- 20. No. Runs on Water With a Flow Rate Greater Than 2000 GPM w/No Loop Seal 4
- 21. No. Runs With an Induced Bending Moment Greater Than:
80,000 in-lbs 28 90,000 in-lbs '
22 100,000 in-lbs t 15 241,738 in-lbs max. 1 e
o L
y-. ~er-wy.> -,,-a-- ,-+n,,.-ww- r - p-a.w.--m.-rn-ne- .~,re'-.m..pr-,r- ., wwnase -- w w n, m - g-,w. , ,--,r-- --
4 a
The EPRI test program identified the possibility of water damage to those valves installed with a loop seal up stream of the Dresser 31739A valve (Reference 1).
Run number 1030 indicated instability by chattering. None of the other runs on water indicated either chatter or flutter. The pressurizer safety valves on the CR-3 plant are located directly on the pressurizer without a loop seal and, therefore, should not be subject to the conditions that might damage the valves.
During the February 26, 1980 transient (Reference 3), water flowed through the open safety valve for approximately two hours before the safety valve was i completely reseated. The results of the examination of the safety valve are reported in Reference 4. Evidence of steam cutting of the seat and disk was found. This was expected since the valve had been leaking for some time before the initiation of the transient. However, there was no evidence of damage, upset metal, galling, or scoring. In fact, no evidence of unstable operation could be found on any of the mating parts. It has been estimated that the valve had been subjected to a water flow of 700 GPM at a pressure of 2300 psig (Reference 3). It is, therefore, concluded that water flow through the pressurizer safety valves on CR-3 will not affect the valve functionability.
pVESTION NUMBER 3 Crystal River 3 uses the Dresser Model 31739A safety valve, which was tested in the EPRI test program. According to results reported in the EPRI Safety and Relief Valve Test Report, this valve passed its rated steam flow on certain tests but not on others, depending on the ring settings and backpressures. The specific ring settings to be used at CR-3 and the expected plant backpressures were not identified in the submittal. The letter dated November 1, 1982 indicated that the safety valves were replaced with valves having ring settings consistent with the EPRI tests. It also stated that a plant specific evaluation of the discharge piping backpressure was in-process. This letter does not, however, identify the final ring settings or backpressure. To establish which EPRI tests are applicable to CR-3 and thereby demonstrate that the safety valves will pass ' rated flow, provide the final plant ring settings and expected backpressure.
REPLY NUMBER 3 f The final ring positions for the Dresser pressurizer safety valves Model 31739A
- are as follows based on a calculated backpressure of 520 psia, j Ring Position (Notches Upper -48 Middle -50 Lower (1)
(1) Value of the lower ring is based on meeting the clearance requirement of 0.008 inches between the top of the lower ring and the disk holder when the valve is cold. Therefore, depending on the valve in question, the lower ring could be positioned at +11, +6, or +2.
l I
I
- _ - __ ~ _ - _ _ - _ - _ _ . . . _ _,_.._.. _ _ _ _ -,_ - _--___ _.__._-_ _ --_ _.
e i
The valves presently installed on the pressurizer are designated by serial numbers BLO8899 and BUO3149. The lower ring settings for these valves are
+11 notches.
QUESTION NUMBER 4 In a letter dated March,1976, from Dresser Industries (the manufacturer of the Crystal River PORV) to Metropolitan Edison Company, Dresser cautions that the PORV block valve should be kept closed when a reactor coolant system pressure is below 1000 psig to avoid damaging the PORV disk and seat by steam wirecutting.
Results from the EPRI test program indicate that this valve was successfully tested on water at pressures in the 500-900 psig range but was not tested on low pressure steam. Additionally, each test sequence was initiated with a valve where the disk and seat were in excellent condition, which may not be representative of the CR-3 PORV when placed in service. Thus, the available EPRI test data is evidently insufficient to demonstrate compliance with NUREG-0737 Item II.D.1. The Dresser recommendation indicates that precautions may be necessary to avoid damage to the PORV disk and seat at pressures below 1000 psig. Following the Dresser recommendation to isolate this valve at these lower pressures would, however, seem to preclude the use of the PORV for cold overpressure protection. Explain what precautions will be taken at pressures less thatn 1000 psig to prevent damage to the POP.V disk and seat or provide results from EPRI or other tests performed since March 1976 that demonstrate that precautions described in the March 1976 letter are not required to avoid such damage.
REPLY NUMBER 4 Dresser has re-examined the contention that the PORV should be isolated below 1000 psig and concluded that this requirement was not necessary. Further, the valve could operate down to 50 psi and below. They do recommend that the PORV be seated with a surge of pressure as a precaution against initiating leakage, but do not categorically state that if this is not done, the valve will leak.
Operating experience in other plants appears to indicate that leakage is not always a consequence of starting operation with the Electromatic exposed to zero over pressure. In addition, the Crystal River 3 startup procedures (para.
6.4.6.12) call for the PORV to be operated twice when the reactor coolant system is at a pressure of 205 to 215 psig: once after the PORV isolation valve is closed and once with the isolation valve open. Failure of the PORV to operate properly will force the plant to remain at its present made until a decision is made on how to restore the valve to its proper condition.
This procedure adequately demonstrates that the PORV will operate satisfactorily at pressures less than 1000 psig and, therefore, can be used for low pressure overpressure protection.
The attached Dresser letter to GPU dated March 13, 1984 and its attached interoffice memo categorically refute the pre-1977 recommendation.
e a QUESTION NUMBER S To verify that the EPRI tests adequately demonstrated stable operation of the plant safety valves, the EPRI Test Condition Justification Report indicated that the inlet pressure drop of the test piping must be at least as great as that for the plant. Provide a comparison between the inlet pressure drop for the tests and the expected pressure drop for the plant.
REPLY NUMBER 5 A representative pressure drop for both CR-3 and the EPRI test loop is presented below for comparison:
Delta P, PSI Flow, Lb/Hr CR-3 (calculated) 18 360,000 EPRI Test Loop 20 380,000 Run #318 (31739A)
Reference 7 The inlet pressure drops compare favorably. The criteria for limiting the inlet pressure drop is set by the manufacturer not to exceed 50% of the blowdown.
Theoretically, the blowdown could be as low as 5% of 2500 or 125 psi. Half of 125 is approximately 60 psi or three times greater than the calculated value of 18 psi.
QUESTION NUMBER 6 Bending moments are induced on the flanges of the safety valves and PORV during the time they are required to operate because of thermal expansion of the pressurizer tank and piping and because of the fluid discharge loads. Provide assurance that these bending moments will not adversely affect operability of the valves.
REPLY NUMBER 6 The calculated bending noments imposed on the safety / relief valves that were generated by Gilbert Associates were combined as shown in Reference 6. The results are compared with the EPRI loadirigs as shown below.
Combined Calculated ERPI Test Imposed CR-3 Valve Moments Valves Loadings PORV 3153-VX-30-2 15,552 in-lb 3153-VX-30-2 25,500 in-lb PSV 31739A(RCV-8) 25,099 in-lb 31739A 241,738 in-lb (RCV-9) 18,113 in-lb (Run 1011)
From the above, it does not appear that the valves will have any operability problems as a result of the calculated bending moments.
RUESTION NUMBER 7 To meet the block valve qualification requirement contained in Paragraph II.D.1.B of NUREG-0737, the submittal refers to a transmittal from R. C.
Youngdahl of Consumers Power Company to Harold Denton, NRC, on June 1, 1982 concerning a block valve testing program conducted at the Marshall Steam Electric Station. This test program, however, did not include tests on the particular block valve used at CP.-3. Additionally, these tests were limited to steam flow conditions, whereas NUREG-0737 requires demonstration of valve functionability for all fluid conditions for which the valu is required to operate. Provide a justification as to how results of the Marshall tests or other tests can be used to demonstrate operability of the CR-3 block valves for the required conditions. Account for differences between the CR-3 valve (and operator) and the test valve (and operator).
REPLY NUMBER 7 Shown below is a comparison of the CR-3 PORV isolation valve and the PORV isolation valve tested at Marshall Station.
CR-3 Marshall Station Manufacturer Velan Velan Size 2-1/2" 3" End Fitting Flange Butt Weld (converted to flange fittings for testing)
Type Bolted Bonnet, Gate Bolted Bonnet, Gate Pressure Class 1500# 1500#
Material 316 SS 316 SS Limitorque SMB-00-10 SMB-00-10 (this valve was also tested with an SMB-00-15)
The valve tested at Marshall Station was not appreciably different from the CR-3 design. It closed in 13 seconds against flowing steam whose pressure and flow rate was approximately 2300 psig and 235,000 lb/hr.
f Although Marshall Station tests were not intended to include every conceivable flow condition that the block valve could experience, the tests did prove the capability of the valve under conditions far above those expected for steam flow alone. During these tests, all stroking of the valve was satisfactory with respect to stroke time, opening, and closing on demand and zero leakage.
l
. _, .~ ,_ _ . . .. _ _ _ , _ _ _ _ , _ _ _ _ _ _ _ _ . , _ _ , _ _ _ _ _ _ . . _ _ _ . , _ _ _ _ _ . _ _ . _ , _ _ _ . . _
For further evidence of the capability of the CR-3 PORV isolation valve to operate under extreme conditions, the design shall be compared to the block valve on TMI-2 which operated shortly after the major accident. The TMI-2 block valve is similar -to the CR-3 isolation valve.
The THI-2 valve was cycled in excess of 30 full open/ closed cycles as described in Reference 5. Thirty of the cycles occurred during a period of two hours when the upstream pressures ranged between 1865 and 2150 psig. Cycling was discontinued when the reactor plant was brought under control and the valve was placed in the closed position.
Shown below is a comparison of the CR-3 PORV block valve design with the TMI-2 PORY block valve desing.
CR-3 THI-2 Manufacturer Velan Vel an Size 2-1/2" 2-1/2" End Fitting Flange Flange Type Bolted Bonnet, Gate Bolted Bonnet, Gate Pressure Class 1500# 2500#
Material 316 SS 316 SS Limitorque SMB-00-10 SMB-00-10 As a result of the EPRI test Program at Marshall Station together with performance records of similar valves under field operation, it is concluded that the CR-3 isolation valve will operate as required.
QUESTION NUMBER S NJREG-0737, Item II.D.1 requires that plant-specific PORV control circuitry be qualified for design-basis transients and accidents. Please provide information which demonstrates that this requirement has been fulfilled.
REPLY NUMBER 8 The question is unclear as to the meaning of " qualification" of POP.V control circuitry. It is obvious that environmental qualification of the PORY (and.
therefore, its control circuitry) is not required by NRC regulations (10 CFR 50.49), and we do not believe that such is implied in the test of NUREG 0737, Item II.D.I.
In addition, the PORV control circuit components are located in the control complex and are not subjected to the ccntainment environment.
QUESTION NUMBER 9 To meet the NUREG-0737 requirement that the piping and supports associated with the safety and relief valves be qualified as well as the valves themselves, the submittal refers to an analysis that was performed on the February 26, 1980 transient. Relying on this analysis alone to qualify the piping and supports requires justification that the fluid conditions corresponding to this transient will produce maximum dynamic loading on the system. This is somewhat questionable since the safety valve discharged at a pressure of 2425 psig, while the valve set pressure is 2500 psig. Al so , the transient involved only the discharge of one safety valve, whereas an actuation of both safety valves may produce worse loading. Provide a comparison between the peak pressures and pressurization rates with those expected for other limiting transients and discuss the effects that an opening of two safety valves rather than one would have on loading of the system. Justify, if possible, that the transient analyzed produces maximum loading on the piping system.
REPLY NUMBER 9 As delineated in the response to Question 1, the worst case Pressurizer Safety Valves (PSV) inlet condition are:
Steam Line Break:
PSV's set at 2500 psig. Initial steam flow followed by subcooled water.
602*F maximum water temperature, 517 F minimum.
At 602 F, 6555 lbs/ minute of water insurge rate.
At 517 F, 6019 lbs/ minute of water insurge rate.
No credit is taken for reactor operators manually throttling HPI flow. To achieve a full pressurizer for a steam line break, the HPI pumps (three) would have to be allowed to run at full flow rate for approximately fourteen minutes.
This allows more than enough time for operator action. Additionally, during the first minutes of the transient before credit for operator action could be taken, the Reactor Coolant System minimum temperature is expected to be approximately 490 F at 550 psia about forty seconds into the transient. While 490 F is the lowest temperature that is expected to be seen during a steam lire break, the RCS pressure at this temperature will not be high enough to challenge the PORV/ safety valves.
The maximum pressurizer pressure and pressurization rate from Table 5-1~ of EPRI NP-2352 is given for the Rod Ejection Accident 0 HZP.
Assumed Pressurizer PSV Setpoint 2575 psig Saturated Steam Discharge Maximum Pressurizer Pressure 2662 psig Pressurization Rate 175 psi /second
For evaluation of piping discharge loads, subcooled water discharges are normally controlling due to the much higher discharge rate through the valve.
The February 20, 1980 transient (PSV-RCV-8) was analyzed using the following inlet conditions:
PSV's Open at 2410 psia Pressurization rate - pressurizer maintained at 2410 psia PSV inlet temperature - 560 F (subcooled water)
PSV design flow (317,973 lbm/ hour 0 2500 psig - saturated steam)
The above PSV flow data was used to determine the valve full open flow area using the Moody critical flow model for saturated steam. In addition, the PSV flow area was increased by 17% to account for 10% accumulation and 5% error.
The transient was analyzed using the Henry-Fauske critical flow model for flow through the valve. A comparison of the mass flux through the valve using the Henry-Fauske critical flow model is given below.
P = 2410 psia P = 2515 psi.a T = 560 F (h = 559.4 BTV/lbm) T = 517 F (h = 507.2 BTV/lbm)
G = 24350 lbm/sec-ft.2 G = 28790 lbm/sec-ft.2 The mass flow rate conservatively is expected to increase by 18%. However, a 17% margin was included in the incident transient analysis. The transient analyzed can be considered representative of the transient that would be expected using a 517 F subcooled water discharge.
A comparison discharge rateof the maximum through the PSVpressurizer insur-sec (28790 lbm/ft. e rate (6019 lbm{
x 0.0149 ft. minute) with the x 60 sec/ minute
= 25738 lbm/ minute) indicates the relief capacity greatly exceeds the insurge rate, and the assumption of a constant pressurizer pressure is conservative.
The RCV-8 incident transient was analyzed assuming a reactor coolant drain tank pressure of 100 psig. This conservatively maximizes downstream piping forcing function. In addition, the PSV's Discharge Piping System is not headered together and, thus, the pipe discharge loads are independent of the number of PSV's actuated.
QUESTION NUMBER 10 Further information is needed to evaluate the thermal hydraulic analysis. The RELAP4/f1095 computer program was used to perform the thermal hydraulic analysis. Explain whether parametric studies or other verification studies were performed to assure that this program would generate accurate fluid forces for the transient analyzed. Also, iustify the use of a safety valve opening time of 0.04 seconds. The opening times observed in the EPRI tests on this valve for water discharges were typically significantly shorter than 0.04 seconds.
_ - . . - _ - - _ .- - . . . - . .-- - . .- - =- . _. -
i REPLY NUMBER 10 (a) The RELAP4/ MOD 5 computer code was the only code available to simulate the discharge through the SRV piping network at the time the analysis was performed. However, considerable work is available to show the suitability of using RELAP to determine the discharge piping conditions. The following references are given:
(1) "The Application of RELAP4/REPIPE to Detennine Force Time Histories on Relief Valve Discharge Piping", Semprucci and Holbrook.
(2) " Steam Hammer Design Loads for Safety / Relief Valve Discharge Piping",
Strong & Bachiere.
These papers were presented at The Third National Congress on Pressure 1
Vessels and Piping, San Francisco, June 24-29, 1979. Sponsored by ASME Pressure Vessel and Piping Division.
(b) The EPRI Safety Valve Test Data Report was reviewed for water discharge, with a short inlet configuration which is representative of the CR-3 configuration. The only test with this short inlet configuration is Test No.1110 which experienced a pop time of 0.043 seconds. Therefore, for a subcooled water discharge, a valve opening time of 0.040 seconds is e
considered acceptable.
QU'STION E NUMBER 11 The THRUST computer code was used to generate fluid force histories from RELAP4/M005 output. Provide a detailed description of the methods used in this program and explain how the program has been verified for the type of transient
- analyzed.
REPLY NUMBER 11 The THRUST code description is attached (Attachment 1). The methodology har been verified by analyzing the steam line rupture given in the code description and comparing with results analyzed by Moody (l), and Strong & Baschiere(2). In addition, a comparison with RELAPS/ MOD 1 of piping forces for the same problem is given (Attachment 2). The RELAP5/ MODI code has been used recently in SRV
- discharge problems and has been benchmarked by the EPRI/CE tests.
References:
- 1. F. J. Moody, " Fluid Reaction & Impingement Loads", Conference on Structural Design of Nuclear Plant Facilities", ASCE, Chicago, 1973.
, 2. B. R. Strong, Jr., and Bashiere, " Pipe Rupture and Steam Water Hammer Design Loads for Dynamic Analysis of Piping Systems", Nuclear Engineering and Pesign, Vol. 45, 1978, p. 419-428.
QUESTION NUMBER 12
, The submittal does not provide some of the important details of the thermal hydraulic analysis on the PORV and associated piping. Identi fy the fluid conditions assumed, including pressure, pressurization rate, temperature, flow rate, and fluid type. Assure that the fluid conditions for cold overpressurization events were encompassed in the analysis.
g- p .- - . - , , - - , , . . , , . , , _ , - , - - _ . - -
- ,w-wr- r-me -- em--' '=--='*-ww-'NwW * + e m -
-'ewa--r= --"e*-=e'r"-tm=r---*v-v'
REPLY NUMBER 12 The thermal hydraulic analysis for the PORV discharge piping was based on the following design conditions.
Pressurizer Pressure = 2315 psia Pressurizer Temperature = Saturated Steam 0 2315 psia PORY Flow Rate = 117,000 lbm/ hour (0 2315 psia -Steam)
Pressurizer Pressure = Constant 0 2315 psia Subsequent to the above thermal hydraulic evaluation, it was determined that the PORY was a high capacity type. This, in conjunction with the PORV setpoint increase to 2450 psig, resulted in a valve flow capacity of 165,900 lbm/ hour.
The original analysis was not reanalyzed, however, an evaluation was performed based on the assumptions that the pipe load is proportional to the maximum discharge rate since valve pop (opening) time did not decrease. Based on this and the margin available in the original design, the integrity of the discharge piping, piping supports, pressurizer, and reactor coolant drain tank was not jeopardized (reference FPC submittal July 17,1980).
Cold overpressurization events, as described in EPRI-NP-2352, indicate that the limiting PORY inlet conditions is a makeup control valve failure. The PORY setpoint is 550 psig open and 500 psig close. Fluid conditions at the valve inlet are saturated steam at 550 psig. The discharge piping evaluation would be bounded by the design condition evaluation performed earlier due to the much lower valve open pressure.
QUESTION NUMBER 13 Further information is needed to evaluate the structural analysis. The submittal states that the PIPDYNII computer program was used to perform the structural analysis. Provide a description of the methods used in this program and explain how the program has been verified for the type of transient analyzed.
REPLY NUMBER 13 i
The fluid transient run was made using the time history subroutine SYNSYS of the PIPDYN II program. Attachment 3 contains the abstract and pages 5-20 through
. 5-29 of the PIPDYN II users manual. These pages provide a description of the methods used in this program to perform the structural analysis.
The time history transient analysis was verified by performing an analysis of a structure which has been previously analyzed by Biggs (Biggs, J. M.,
4 Introduction to Structural Dynamics) and comparing the results of the two i
analyses.
~
QUESTION NUMBER 14 To adequately demonstrate structural integrity of the system, the loads due to fluid discharge transients must be combined with other loads such as seismic and 1
- - - - - - ,-y._, -# .. - , .,,,,_-.2._ - o ,m mm _, vwm_, e7-m, , _ , y-.~,y._,,.,_y - -,_,,,,,_,._.,,_.,_,,~.,,,,,.,my.,_. ,
operating. Identify the load combinations performed in the analysis together with the allowable stress limits. Explain the mathematical methods used to perform the load combinations, and identify all governing codes and standards used to determine piping and support adequacy. The submittal does mention that the ANSI B31.1 Code (year not given) was used to evaluate adequacy of the nozzle connection to the pressurizer.
REPLY NUMBER 14 The analysis of the 2-26-86 incident was an analysis of the piping system for the transient that occurred on that date. The load combinations used in that analysis was deadload, pressure, and the water discharge transient.
The piping stresses were combined by absolute addition and compared to an allowable stress equal to one hundred and twenty percent (1.2 times) the allowable stress in the hot condition (S h ) of the piping. ( A review of the stresses show that a combination of deadload, pressure, water discharge and seismics would also satisfy the allowable stresses.) The piping code used to determine piping and pipe support adequacy was the ANSI B31.1 Code,1967 Edition with code case N-7.
The loadings on the nozzle connection to the pressurizer were reviewed and accepted by the NSSS supplier (Babcock and Wilcox). The loadings were evaluated using the ASME Code Section III, Class A,1965 Edition, Addenda through Summer 1967.
QUESTION NUMBER 15 The submittal states that incident transient loads were calculated only for the first four piping sections downstream of the pressurizer, claiming that these are the major contributors to the loads on the pressurizer nozzle. A problem with using this analysis to meet requirements of NUREG-0737 II.D.1 is that the NUREG requires a qualification of the safety valves and PORV and the associated piping and supports. It must be verified that the fluid loads will not jeopardize operability of the safety valves or PORV, or structural integrity of the valve inlet piping as well as the pressurizer nozzles. Thus, provide a justification that an analysis has been performed where enough of the fluid transient loads, discharge piping, and pipe supports were included in the analysis to assure unimpaired operability of the valves and structural integrity of the valve inlet piping and pressurizer nozzles and to verify that pipe deformations will not block fluid flow anywhere in the system.
REPLY NUMBER 15 Since each piping leg contains an axial piping restraint and the unbalanced transient loadings act along the axis of the piping, after the fourth piping section very little if any of the incident transient loads would be expected to have any effect on the pressurizer, valve inlet piping or the valves.
Based on the technical content of report " Evaluation of the Plastic Characteristics of Piping Products in Relation to ASME Code Criteria, CRNL/Sub-2913-8", no functional checks of flow in the piping is requried if the piping stress levels are below the upset limit of 1.2 Sh. All of the pressurizer discharge piping stresses are below the 1.2 Sh limit.
- QUESTION NUMBER 16 A letter included in the submittal dated July 3,1980 discusses an analysis of
, the supports attached to the piping from the pressurizer to the reactor coolant drain tank. This letter suggests that at least two of these supports should be replaced to sustain loads that include steam and water hammer loadings. Provide a final evaluation of stresses in the piping and supports and identify any required modifications to the piping or supports.
REPLY NUMBER 16 An evaluation of the supports for the loadings of the 2-26-80 water transient incident showed that two supports would require modifications to withstand the i subject loadings within pipe support design parameters. Since this water transient was not considered to be a design condition at that time and since the applir.d loadings and visual examination of the supports both indicated that no dam. iga was incurred by the supports, the supports were not modified.
As mentioned in the response to question 14, the piping stresses due to the 2-26-80 water transient were reviewed, and it was shown that by using the absolute addition method, these stresses in combination with deadload, longitudinal pressure, and SSE seismic will satisfy the allowable stress limit of 1.2 Sh. Since the three valve discharge lines have similar routings and pipe supporting arrangements, it is expected that all three lines will satisfy the applicable stress limits when combined by the SRSS method as described in the EPRI report " Guide for Application of Valve Test Program Results to Plant Specific Evaluation", Appendix E.
QUESTION NUMBER 17 One of the safety valves discharged water during the February 26, 1980 transient and the safety valves are expected to discharge water for extended operation of i HPI events. When the Dresser safety valves discharged water in the EPRI tests, the valves typically fluttered through partial lift positions at high
, frequencies / These valve oscillations cause high frequency pressure
{ oscillations in the valve inlet piping, which could potentially excite high frequency vibration modes in the piping. This excitation creates bending moments in the inlet piping that should be combined with moments from other i mechanical loads. Provide one of the following:
i (1) A justification that these high frequency pressure oscillations will not
! occur or (2) A comparison between allowable bending moments with the bending moments induced in the plant piping by the dynamic motion and other mechanical loads, i REPLY NUMBER 17 i
Response to Question 17 is based on EPRI report EPRI NP-2770-LD Volume 10, March 1983 Summary, Page S-5. "For the Dresser safety valves tested, satifactory i
-- - .- w,,. ,,,.--.,~,,,,,,.n .,--n-, - - . . , . ,. ,-,,,,,m.-c,,n,,,m.,,,,,,,-,-,,.. -.-~~--.-n--- ~n .-
operation was observed on saturated and 500*F nominal temperature water with the valve mounted on a short inlet". Since the valves at Crystal River are Dresser valves and are mounted on a flange connected directly to the pressurizer nozzles, these are considered as very short inlets and, therefore, will not experience any significant pressure oscillation.
e 1
References The following are references used:
2 .~ Valve Inlet Fluid Conditions for Pressurizer Safety and Relief Valves for 4
BOW 177-FA and 205-FA Plants, EPRI-NP-2352 Final Report, December 1982.
- 3. Transient Assessment Report, Reactor Trip at CR-3 Nuclear Station on February 26, 1980. Prepared for Florida Power Corporation by the Babcock
& Wilcox Company, Report No. 07-80-02, dated 3/11/80.
- 4. Examination and Test of Crystal River Unit No. 3 Power-0perated Relief and Safety Valves, EPRI-NP-80-13-LD Interim Report, December 1980.
- 5. Analysis of Three Mile Island-Unit 2 Accident, NSAC-1, 7/79.
! 6. Load Combination for Pressurizer Safety / Relief Valves -
B&W Doc. ID 32-1159471-00.
i i
I l
l 4
I l
i l
1 y y w - - - = , - =w y ---w,-*v -gw=,-y,-,te----v.-g--e-v. ,--w~ w ,.rp<w y = w i cv 3----wmn~y-,w--aw,re--g-s------r-yc ,--r-- e e-- -+
~
k ((AGY N S [ fdA A S A Y
- Y DRESSER ?.&ygg,.t
, INDUSTRIES Q 5g!
suQuarmiAL vat.vc opcRariou s a s ex 14 3 0 c AL.EXAN DRIA. L.QUISl ANA 713 01 I v:6. sis /s o.a s so O twx sia.sva s7s s 7 6sx: se a4as cAs6s oivio March 13, 1984
.t' Mr. Jim Correa GPU Nuclear . -
100 Interpace Parkway - Tarsippany, New Jersey 07054
Subject:
31533VX Dresser Electrcmatic PCRV Gentlemen: I attach a Dresser interoffice memo which is se]f-explanatory. We confirm that the PCRV valves fitted with the new H.D. spring will cperate at inlet pressures of 50 psig er slightly less. D ere is no requirement that the isolating gate valve below the PORV remain closed to any specific pressure; but, to prevent a potential leakage, it should be left closed until there is sece t,ressure in the system. ne addition of these H.D. Eprings in no way affects valve cperation, and the performance testing data cbrained during the EPFl test program fer PORV's st ill applies. . Sincerely, , w - F. P. Bolger Chief Engineer * , l Consolidated valves FPB/sc Attachment cc: D. R. Butler . R. A. Cedel R. S. Huffman n
. i Industria .alve F Operations em' ,W y._. ..g' inter-C ice.
correspondence r30 R. A. Cedel
?m=
January 19, 1984 "5 * *
- F. 'P. gBolger '
See Below*
- t o
, 31533vX h*V .
I
*J. L. Bordelc
- R. S. Iowry D. R. Butler M. N. Rydos- ,
, R. S. Huffman D. J. S:allan -
\
I received' t. call from Frank Cherny of the tEC yesterday on the subject of this valve. We have discussed the same problem cn several occasions recentl'(, so I think the time is appropriate to set the record straight. It has oeen reconnended Dresser practice for many years that the isolating gate valve used in many cases under the Electromatic relief valve should be left closed until a substantial pressure _has been accumulated in the protected system. Current 31533vX manuals specify 500 psig should be obtained before cpening the gate valve; but, I have also seen 1000 psig reconnended. B is procedure does not mean that the Electrc=atic valves are unsuitable for pressures below 500 psig, but was instigated to prevent persistent leakage that scmetimes occurs with the valve when pressure is slowly increased fecm zero. his problem is caused because the flow path which provides pressure benind the disc is trere tortueus than the supply path to the disc areas wnica terd to hold it cpen. It is therefore desirable to supply a sufficient pressure to force the valve disc closed beScre it starts to leak. mis principle can te used at loser pressures, of course, and recently develcped start-up procedures fer s0:re pressurized water reac: ors now require the valve to be cpen as low as 50 psig. You will remember the recent visit by Duke Power personnel to witness a valve: tested at Icw pressures both at Dresser aM later at Wyle Labs and the efforts of Folland Huffman and David Scallan to modify the valve to tuke it perform belcw 50 psig. Se attached letter sho c that the valve is ' suitable for these pressures since it worked well at Duke Power, but we feel that the performance is improved by the substitutica ci a heavier spring in both the main valve and the pilot valve. tis is what we did in that case. In short, there is rc magic pressure that should be acnieved before the gate valve is cpened; but, en the ot.%r hand, it is not advisable to leave the gate valve open fr= cold. M -
/Y /r tr ' #*4 /g /p l lS r
l l} /12 l ff / I =0/2, /n ,>., ,
-e fo 123 r.
l2
- i t
... sy r
.n 4
f/SURE 1 i i b Sk /? ? ll0 & $ l2 /? "l0ll g t a 1 f 1 ( . . O i m+- g +-e--*-.w* * --++--+e-ey
---hwww m q r,%r% g --e.-.pyw+.*--e-9yw,-wgn __,.y ,yn,,,ye. gy-3-e<ee+ - ..w,,wq y e 3-+---
ATTACIIMENT 1 t [ TABLE 1 INPUT PARAMETERS TMIN = 0.0 TMAX = .20
. ~
, SEDIT = 20 Section 1 Section 2 Section 3 NFLPTH (J) 11 5 10 IBVOL (J) 26 0 0 -
IBFP (J) 26 0 0 ABFP (J) 0.* 0. O. COSEFP (J) 1. O. O. IFLPTH (I,J) Flow path's Flos paths Flow path's 16-26 inclusive 11-15 inclusive 1-10 inclusive FEA (I,J) .694 for .694 for .694 for All FP's All FP's All FP's
- There is a 1 millisecond break opening time in'the RELAP model, so let the program calculate the break area.
4 Y 4
\
- t k,
e e
- ATTACllMENT 1 1
f I ! I
+-
l l . ,. , .
. l !
* .I I !
* ! l .
,- j
- l e
- 3 e s l ' ;
i a ; i I
- l.
. i I . . I !
- ! i I
.I .
]
i i i
, , i i e , i
. .
- I . .
- e l
s j . t j t
- I $ i ,
, i t
i i t l' e i t
- g ,
a e g l
> ! I t '
B i g
'l t 1 1
i . I i l
- e 1 .
~
h I ! l . i l l '
' i ' 1 j .
. l i l
- i >
ia i 9 g
. } .
f 8 6 4
' l' 1
, . l l 1 !
l
! 1 !
i i i I .
- I l ,
- i g I t
- I -
g l l . t
- i .
l : !, i. t , 'I ' - 8 4
* . (
' - 8
- i t l
' . t i e
+
1 , e . l l a
, O e ,
s ? w I . e I
} e '
, & e O.
- I
- I t g 6
N O l
* . O e g
9 l 8 l e ? O e
. I ? e g 4
t c e
- r e e r i - , i
! e
, N -
1 l u o. ' i j
' l ' I l l
. , e . s
- l : .o o i l ' 7@ 4 ' I l l l
; I' ' . i e
- I = e, . y e
l e 4 - g g o e
- I a n- -
I - e u
- e
. i a l I '.
I t : N '
! ?
i . i ! - en e s ! . l
) e a @@ ? l 8 g
l ; N #4 o e ! '
; i 4 e e o e ?*
6 l
} = n o& e 4 Z eh M *4 l .
8 i
- l. 0 N= = e e i ,
l 2 e = I i . - o a g ! j *) w ** e *
- a. VJ 7 e l I '
l 2 ? *c@ food ,
. . - wouo ene up l ! .
* ' JN eD . .
l ,
! g l la. O uit 40 e 3
, ; I j i i o n om-O- e i g 20N N i i*
LNe
/
g
, O 7
~ uO4V3 e Je sene O
4 a I
. I 3 . > J ~4C C ", *1 7 i j 8 sa r, e eoe eO +
l
* *. - yo o 8 g
i !
, a <e-o3
, o - o O .. _ m
- f. o c , ,
l i. r . . l
. , i l. ,
~
e 9 m e
~ ~ ~ ~ ~ ' ~ ' ~' '~ ~ ~ ~ ~ ~ - '- ~~
"' R Ct. a P 4 T Hu u3 7
- AEV O' 1/79
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ' ' " ' ~ ' 'PAGE 1 S A 4.*L E PR C5L E'4 FOR NOS2 THa LS T 07/23/F9
~ ~' ' ' ~ ~ ~ ' ~ ~
- 3ATA F Ct4 FOuCC 1 - 'ti I L O m, PATHS ~
34CA< NODE 1:0E K = *b HRLAK F L u .s P47tt INDEX e 26 CAEAK Ft.O w P AT it A rt E A = 00 HitC A K Ot*4CCTION M LL i t PL I C f4 = 3.000 FL0e O%TH FS4 , FLGa PATH FE4 ,, ,, . . . . . _ . . . _ . .. 28 C.e94 25 C.694 2a C.6G4 23 0.694 22 C.bs4 28 C.694 20 C.esa ., , t9 0.6?4 . _ . . . . . _ _ _ _ . _ . _ _ _ _ _ , _ . . . . . . .
! ., C.'.* II O . r> 9 4 36 C . ri . e O A T t. F tw F C 3 ; e 2 - 5 FLOW PATHS _ , . _ _ _ _ _ . _ , . . . . _ . . . . . _ . . _. . __ .. .
NC 94 LAW F L t' a P A T H S F O'4 IHI S C A S '. F L C .c P4iH F #. % FLGa ~5ATH FEA I C**
- I t1 C.cn6
- . . ~
l* s
. . - - ~ ~ - - - - - - - - - - - - - - - - - - -~ - - - - -- ~ ~- -
OATA rca Fcra f t" 3 - 1 C F L O 'a P4 THS , . , , , _ _ , _ _ _ ,,,.__ ,,, . , _ , , , , , , , _ _ NO D H L a w, FLU. PA IHS F0 4 I H 4 S C A SC . F L .T.4 .*\TH FEA FLCa PATH FCA 8 P.>94 V C . P) 9 4 , , ,, t.* *,.s,t 1 0.694
- c. C . f., 4 4 5 0.67%
4 C.644 3 C.694 2 C
- f3 34 . . . ._. I ... 0.69e , . , , , , _ , _ . . _ , , . , _ , _ , _ _ _ _ _ _ _ . _ , _ . _ . . _ . _ _ . . . .._
THE TaPC IS 4 NCLAP4 OATA T4PE. DAIGthAL rAC8LEW wAS T I T t.ED , , _ , , , , , ,_. _,_.,__ __ ,_. _ , ,, , PUN F Cit VC W I F I C A f t O 's Of' N002 THRUST a s s wr.J 26 vatuvE S . 26 JusC I CNS AND O SLAOS. Q O
--.-.e ,.=.--..m-..>.-==. ...-.--*=w mwe.- e = .*ame==-=*** - *- - * ' * " * ' " " * " " * * * * * ' " * " * * * * * " " " * * * * * ' " *'"*
H
. . . - .. . . _ .. .. . . . - - - .- .. . w
!^%
e
- ~~ ~' -- ;4 CL A D 4 T re 'JS T - DEV O 7/70' -
PAGE 2 s a se rL E PnC:sL f *4 F 0;4 t4 0.32 T H ~t tJ S T 07/23/79 TMC r O'4 C C 1 F04CE 2 FonCE 3 FCECE 4 F O f1C E 5 FORCE 6 FORCE 7 F O;4C E d l'CRCC 9 FCPCF!) SEC LOF LUF t 0F t 3F tCF l. U t' LOF LUF ttf 4.cr 0.- O.O O.0 -7.7393E+0t .. 0 . 0 0 0 '.d1 -; . ! '. 3 E + 0 3 0.0 -7.7J9fL608 3.C?;20 -l.5?7-C+0S 0.0 - F . 7 0 V e:E*06
** .0Ct*O -;.d4LJC+CS 3.0 - T . 7 0 '7 9 E
- C l 0.0 0 ? )u - 8 6J 7 3t
- CS C.0 -7.78CIE*05 . _. .
- 0. *i. 20 -1.?FCIC+05 0.C -1.7tC3E*01 0 . 0 0 2 C - 1 2 7 t ht
- 05 O.C ~7.7tC4E*?t 0 . C ? J '. 0 -1.? OOL*03 - t . C F '4 3 C - t o - 1. 7 t C '4 E
- C l 0.034 ed - 1.1612*;
- r *2 -2 5G935-10 -1.78888:408 . . . . _. .. . .. .
C.-.'? -t.0=F,.*C2 -d.wJ47C-lO - F . 7 5 1 4 F.
- 0 5 C.0***0 - 1. 0 * * ,; * *. ". - 1. l z t a t - C 9 -T.7886E*Cl
? . ? *. , " - 1. 0 t 3;.
- c 2 - 4 . ? C J 3E 1. 71 1 'a d
- 0 8 C . 0::e.11 - t . C 1 L
- 2 6 0 fs - J . J: ,35- C3 -7.1120E*OI . ..... __. . . _ . . . . . . . . . _ . . . .
0 0* 20 - I . C i l I;
- 0'.* - 3. 5 8 d :'d - O s - 7. 712 2E
- 0 4 C . r ts 7 0 - .; . G 7 C E
- 0 s - 2. 2 rcs5d-0 7 -7.7123C+0!
0.007'.0 ' ' . 491!!+04 - 5. 7.7 7 6C - O F -1.11245+01 0 . C ) c 1 - G . ', 7 4 1c.
- J e ~t.30052 F.112 4 ti + 01 . ._ . . . _ . . .. . .. . . ~.
C . 0
- t. 3 0 -%.%=7t* 4 - J . t ;s T i t-06 -F.F124CoCl 3.C0 7 )O - . C J .'! ) h
- J S - 7. 0 7 a ct-Ou -/.it?4LtOI O.004.,0 -1.0Jf.JM*C'i -1.5101t-05 -1.1124t*C1
- 1. C 10 : 3 - 1. C C F ,C e c' - 3. t 4 2 et-05 -T.F12JC*O1 . .. . _ _ _ _ _ _ . . . _ _ . . _ . . . . . .
. - . . - - - - - - -- - - - - - . - - ~ ~ - - - - - - - -
c1 _ .-..~.-. -_ . . . - . . . - . . . _ . . . 3
1
, ATTACIIMENT 1 4
e l * ' s a 4 F + E
- f. l I t '
l
' N
- s .. , ,N l nm 00 O OO l l . N *i i i i I
8
% WW W WW I
; u~ N~ o ~~
00 PE O 00 t 4 Ph N @N , EL h C. O
, e . 9. V. ,
NN
- NM i 9
; + .u .a .a .si .el ,
, ~~ - .-
N- - l I ww w ww
' << ( (4 7 JJ J JJ I ba tV U W ht I
' OO Q C0 1 '
.
- I i t i e Ow9 = e.
I I I I 000 w OD 8 s
! **l I I t W 'ti td W Wij l i mA-a % A .)
; . O-m a <> , .
e : e 90w N DD l i N. 0 9 . N.==. : . > .
-en ~ Nm . ,
' 8 l- N t *
' l I ? j
- 3 , e ac. s .M .N .N .4 l j , , N94 N NN ,
e
]
. g j wNa = . i e . Jww w w. .i
- l l 1
. O<< < << i ,
, 4 l l '
>JJ J JJ I
! (
I J a 'J W WM *
' g l' COO G 03.
- I e
* . M ' .
- t. e~ .
NN O @a . i l
, i -00 ~ oO i 3 ,
l j
' - *I I i 1 l
- O, W W W WW . .
ue, o oN , !
- ohn N Of '
Ce N O N ON l (' , , , *C
- e. l
- h. e. .
P. =. e C. e =O ! i c I et un H n it - l , -0 4 ? n tn
. Nr = e
[ w .u .- n w w i I <44 4 C4 1
, WJJ J JJ
- Cr W u W WW '
- i. <cu O ou '
i N=. e o- i 8 OC O O 00
**i t 1 I
, W w dJ W WW '.
g OA O MN I eOA CP W? I l ONC $ Oe a 8 l 0 0. =. O. =. N. O e. g[) se e se
' l I s 98 p h 1 i . 11 . 1.It
' i '
QOO 9 @@ e I NNN = . e ; , i j www w ww e , g e se< C << 3 e g e O .J J .J .J J e
.J a.r W u Wio g ,
s00 0 On g 1 1 o l , i 1 l 03 N O O t9 N ON ' f
% :t n o o t? O UO ,'
E \T * **6 1 i f
- 1 h" uJ 1 I W i!J t.t e.s s . .b,p
. . a)
> l' -s
.N, '
1 N : s~, env a w ;~ , t 'J h O '.P % -. r, O 00 O 00 * ' g N. o . e. . . .. ' ' l f > MN9= M @N , g e
- UM ii l cr 0 - se si p p p i g
g s p .. . - .n t 8 g N de = 41 O C0 ,
! y u w ee r. . -
vs u te n . - - ww A .J O #e44 4 44
, l' 3 *'s
- snJ J .J .J J J.
a,e s.J e nJ O tt ..s u u . u. e I LN GOOO O O3 *
..a , >a O -. ed M I eW is ft J LJ U U g 4 t* La U W U e? 7 f" Q fF '
J4 = C Q O ' f MO > 4 w k t
f
.m e
m
- - ~~~ ' ' ' ' " ' ~ ' ' ' " ~ ~ ~ ~ ~ ~ ~ ~ ~ ** ~
R EL A('4 TP9uST - ACV O 7/79 PAGE 4 S A '4PL E P.t OeL L*4 r04 NO32 1,1;4 05 T CF/23/19 TIME FC4Ch t FO7CE 2 ~ FORCE 3 FCnCE 4 FORCE 5 ' FORCE 6 FORCE T FOnCE 8 FOACE 9 F C ACE t 3 GFC LOF L8F t. CF LUF t UF LDF LUF LOF LUF LSr 3.9tt*C - l . 0 8 3f + 05 - 6 31 S 5C T. 7 t ? 3C + 01 ,, _ _ , , 0 21230 -l.00420*C3 -4.9022E-34 -7. 712 2C + 0 4 0.31200 - 1.CCt/t+0S -2.tv2.E-0J - 7. 7 8 21 E + 0 8 0.0:413 - 7. 712 t C
- 01
- I . 0 0 J 3E *-2.52H 0 . C l'. 0 a -3.C;tF;+0S C b - MC-02
. 010-7.7820C+CI
'.L - 0.1 , , , ,
0.01400 - 1. C O 2 8 t + 05 - 7 02 7 t C-C2 ~7.712CE*CI O.JI'J3 - 1. 0 31 k + 0 5 -I.1163C T . 7 8 1 *J r. 9 0 4 C . O l e. 3 0 -1.03*.E
- 0 5 - 4. t a /SE-01 -T.TlteE+CI C.CI900 -I.Cc22r*35 - 9. OS I IC-9 I -7.TIluC*0I ,, ., , , , , , , , , _ _ . . . , , , , _ _ _ _ _ , _ _ _ , , _ . . . . . . . , ,
3.0203J -1 300 0*CS -l.3/Bld*00 - 7.18 12 t
- 01 0.*?t10 - G . 5 V. .:C + 0 * - 3. %C 2L + 0 3 - F .18 0.1 L + 0 8 C.0/2:0 ..s.<0C+'t - 6 14 2 % + 0 3 -7.77 J TF.+01 C.*?3" -5.. ie*iteca -I.22pt.;+>l -1.735st+01 , , , , _
0.*?403 - 4.5.ste eG4 -2 8/J>2+0i -F.1083E*01 0.0270 -9.5.2.t*04 -3.5J10C+01 -F.6sJdE*03 0 . ? ? e c - 4. 4 4 01C + 0 4 - S . i n J '/2 + .18 - T . (. 3 2 0 ti + 0 1 0.3PtJ1 -%.5.J' +0s - 3 . . .. I .lC + 11 -1.6$41E*01 , _ . , , , . , _ _ _ _ , . , _. , , 0 0/tJJ ~ 4 . ^. 4 / S * ; 4 -t.J 4JE*JJ -F.6J12E*01 0.0/9?? ~ % . 5 9 4 w e:
- C 4 - 2. 026 4C + 02 - 7. 5 J o l E + f* t 0.0200J -4.5Flft+04 -2.dJJJCe02 - T. 54 3 t E
- 01 0
~
m w
e e
, e
- 4 o
ATTACHMENT 1 1 m (. t A 6 \ .= N == 4 3 em 60 O 00 o N 8 l l ,
% WW W WJ t W o. 4N O SJ >
e OO 00 9 NO ' ( O7 N L o .e @ M n*O.
! e e o ee I NN = =c j e e
; a g t
. .a. . .a=a j .
- N% c .= t n N= =
s l ' ww w w. *
<< < et j
; JJ J JJ '
} e . Wt# u W uf } l
'o
, t a oc no .
I e
. i I
O== N =o .
, 000 O 00 ,
i
}=
l +++ a ee ' i g ; i *:J i *
' JJ t;t us '
l, n .tJ =U = c# -
. I c== N vo ' ['
NCO n e9 *
, NNP
- l e se ?. N. ee 6 i
l
; ==M N NO l
n s !
- il 11 il nN 3 f l
<-~ ~ --
i l'
, i . i g
l NMO N NN i
=N= = l i ,
I.
- i. Ja= = == .
l ,
. on< < e< t I i i .
n >JJ J JJ l l l a Wu .aJ ts uJ q t l l
; a aCQ O C0 i e . I g
+ 1 M i e
i j r * "
'l
' O= -N 8 , 8 i I =0 00 '
O, I i l, i
' O . D. i s 4 j IWU W WW l
+
e e WN? .= NN s 8
, ! O-N N 90 . g
- ce.c?. n
=
N eo ,.
'.9 ! O .
' ne e e ee t 8 '
, , 8 O lf) ** N =. '
c s
- r. n n 4 '
a== ==. .n. .a O4? O 09 N t% = = ww w ww 444 4 44 , W .J J .J .J .J G W hJ W WW 400 Q OQ NO= = -N ' 000 O 00 I ess l t 1 -
- WJW W WJ '
- p. A = c A '
e
?N4 O C '*l 6 r NnN e !
! M?@ N '."e) 8 g' e e . P. M. . , g i ONC 4 -9 g
3 3 l ' t
- I
,' It n a a ll n I
- -=
l 000 e De l
- NNN =
l ' I www w .. l l '
- 844 4 << '
, oJJ J JJ ! .
l J 6l uJ V WW t I
- O ,
.i l >.
L00 O3 t e m
- i P3 9 l
)i N === 0 N9 I = '
NT 0000 O O ') 3
- l g
\I
- 1++ * ' !
l N e* te ' g r t' I t.J v s J e t ee g
, @A N tJ U l g
I . N o,N. O ~ o'4) l
, ' g O >*e> k es , ,
00 =000 N oe i j l z e e e e e e e
- , i i > nNe= N mo g p WT #
- j
; go n t a g a c== n a .a .n. a .a s e
a j rvc=o G O .fi l 7 O wNN= .* = s l I
>W u eww= . ww ,
g j Md44 Mn .J, O. sJsJ C, 44, f l ff U O Cr L; m s ti; em # e 1- l
- Lu
- LOOO O On I' * *
\
*c o = N n I ew n l I f. .J W tJ i W I
) 4 f. LJ V U U '
.)? 7 C- O tr
- I W4
- O O O 3 '
ed > L L k j l .
~
. ?'A . . e 4 ,
e O.
- s. .
* . . ,, }
~ ' * ~~ ~ ' ~
' ~
PAGC 22 a t e. aD 4 TH4USI - PCV O 7/79 01/23/19 S .t vl% E p:tCHL EM FOW NOJ2 tesqLST FORCC 4 F C IC t! 2 FCRCE 3 FCRCE 4 FO r. C E 5 rORCE 6 ^~ FORCE 7 FORCE 6 FODCC 9 FCRCE t C 7 twt LOF LDF t. 3 r STC LJF L3F t. 0 F t. 0 F t. e F L4JF t. U F 0.13100 -1 0415C+C's -5.3452E*03 -6.436tE+03 .
?.;323) - 1.
- J a c + C '.a - 6 . 2 ? C bC + 0 3 -rs . 2 4 J ut + 0 3 0.!kJ1) - 1. 0 0 ? a:L
- 05 -b.3/24*+03 -S.120tF+03 0.1<a43 - 1. 0 0 r. r.
- C S '.a . 2 411 E + 0 3 -6.424FE*03 C . t 7 '.
- O ~t.0227.+05 -3. Stb 2E+03 - t's . 0 0 0 $ C + 0 3 S . t wet : - 1. C 3 6 M
- 05 -2.JiloE*03 -5.ct2FJE+0J 3 1-3700 -I.030lt*% -I."n%3 M *0 5 -5. 6 4 F I ti + 0 3 n . I '. t. * ) -l.C5FT'.*05 - t . *s t 's '.ac.
- 9 J - 5. 3 '2 0 3 ti + 0 3 0 . I 7 ' ') ' -t.*1Te(*0% - 2 8. I to w:. + 0 ? -5. I 2 ** I f:
- c 3 . . _. ...
c.1)CCC - l . ; d i ; t + 02 -2.Oseut*)2 - 4. w 9 9 8 t. + 0 3
'. J 'M ! D CF .'L C f RLCC405 = 211. AUSE2h CF ERRCRS = 0.
- a. We
- m. . . .
a ..* .y.. g a ..O. e.ee.W.. .W. . Da
. 844" .m e. ee. ..en a . .
. . . *4 - F e' W asN .9he ,e.r*M 9 Wi N
0
=*.-. -= - -
-me. - ..- ... .m ..,w . . .. .....-....a.m - so - ---..-e='.. -. - 4
.. _....~......... ... ... .
. , . e. e
ATTACIIMENT 1
-- REFERENCES
- 1. Strong, B. R. and Baschiere, R. J., " Pipe Rupture and Steam / Water Hammer Design 1.oads for Dynamic Analysis of Piping Systems," Nuclear Engineering and Design, vol. 45, pg. 419-428, 1978.
4 E
.0 8
E o. 5= I fl e
e 8 9 , ATTAC1 DENT 2 ( .
~
-_n.
S _ F3 Po Cc G4 FT 32 FT (4/ P, =. IccC FSIA SATJRATED STEhi
., C:= \=_cc W/SEC f
M FT A=ICOIN 2 Fl60eE 5-3 &o#
;roRcrNG FUNCBON P3 MESScR '
ML4Pf[/w/ VEFIFIOTI 3rd P OEL.GM t.
~
3.5
- ATTACHMENT 2
~
g, (., ?-4 o pcgc i,J:. ;: usJciscsL'. R2CC6 ~C A VC ' f'E ~ #
- o ScRce= 1 N'
N (-. t - rRG AF4./Mr.DS o ret APS/ moo 1 o o . o_ N o ' o o
.c_
o o < C o_ o . o
-o m =_
O_ ~
~
v
! o u.10 Uc$
c m_ . O-
- k. .
' o o o o_ 2-E o o o_ o o o c. o o o N. d . 0'.15 0'.20 0'.25 0'. %
- 0. 00 0.05 O'.10 IM (SECONO3) 3.6
'. 's - ATTACHMENT 2 516. .? - S fc2c1.<Xr fuNcTrosU F2CCES.SCR VEEtFicATIS fE0E'
- S F02cG 2 0 '
RELAP4/VDDS g eaaspwL
- E_
N . O Q. O. N O O
~
O . O tn N-C_ ~
~
US c ._ OC L
~ O Q
. . . -~
8 . O O O , ens I O O I (*?
~~
i 005 0'.10 0'.15 0'.20 0'. 2 5 - O'.30 o*00 TIME (SECONDS) 3.7
ATTACllMENT 2 e ' gjg 3-6 pocag& r-uncnc^!S WCLEM VC'PICoMd Y^#' S . Fcf4E 3 f- N_
\ en REL.AP4.lV10DS g R5 LAPS / mod 1 m.
N o 9
~
o o a o. N
- o '
o e tc_ ce ~
~
g *. === UJ o U. C~ m _ O-u_ o o f-
/ . o o
E-o o d-o a (- . 0'.05 0'.10 0'.15 0'.20 0'.25 0'.30 0.00 TIME (SECONDS) 3.8
ATTACHMENT 3 4 f ABSTRACT PIPDYN II is a Fortran language computer program for the analysis
, of general three-dimensional beam structures, but more specifically for
- piping systems. The program provides alternative computations in accordance with requirements of either the subarticle "NB-3600, Piping Design" or "NC-3600, Design of Class 2 Piping" of the 1971 edition of j the ASNT Boiler and Pressure Vecccl Code, Section III, or USAS B31.1, 1
" Power Piping" of the USA Standard Code for Pressure Piping. The program
, is also capable of performing a complete stress analysis of three-dimen-l sional frame structures in accordance with conventional beam theory. The essential theory is based on general structural theory with the aid of the finite element method. The program permits the user to describe 'the f
[ physical properties of a structure, constraint conditions, an'd loading information in very general terms. Since the program is comprised of many different subprograms, each complete within itself, the output is con-3 sistent with the input of other subprograms used in the analysis. There-fore, a large degree of flexibility is available in the analysis and i
- j evaluation of various loading conditions.
T j This document has been prepared in three volumes: Volume 1, Theory; I Volume 2, Input Guide and Program Description; Volume 3, Test Problems and UNIVAC Program Mapping. J 4 L i ( T h 111 _ _ . - . . _ _ - . _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ . _ _ _ . _ _ . - _ - . _ _ . _ _ _ _ _ , ._ ~ _._ , _ . - - . ~.
ATTACHMENT 3 r, N l F t-- Table 5-1 PROGRAi1 OVERLAY { c' Secment Subroutines '
't'~
IfAIN PIPDYN,DERAS, ERAS,ERASI,b'RTIME,TEER,MYTAB HEAT PTHERM,LTERP,SYMSOL,*PJISIT 7 W, PIPE ST3DS,DTPROD,0RIMAT,XPROD ONE INPUT,DFSORT,FORZIN,JCSORT DATA _ DATA 3D,AUTONO,CRDXFR,DATBAR,CAUSCO,JNTCRD,JNT3D PLOT PIPLOT,AROHD,ARUPLT, AXON,AXPLT, BLOCK,CIRCL,DASHLN,GE0INP, LEGE;D , MO DPLT , NDI GIT , NODPLT , NOPJ!, PROJR , S CALE , THREED , TRIPLT , ,. f XPRDNM TWO EQATIN, ALLCRV,DSTLol,ELSTIF,FLEFSB,C:MTR LINTRP. MATIN,M2EI, NOCUBE , NOLIBE , S FAC , S P CMAT , STI FN , TCALC , TEE FLX , U:l CURV , UN ILO ( A EQA3D,BND3D, DISP 3D,FOR23D,INFOR2,MULT33,0MPLA, SORT,SYSMAT B SOLELT, EDIT,FRZMAS,SOLSTF
,n.
h THREE SOLVER , DECOMP , ELIM, FRZ??Ji, INF: nT ,1MS SM, S OLUTE , UPDATE FOUR BSTRES , C0!!B IN , DISPRT , LODCOM , ND FO R3, NS TRES , P STRES , RECO:!P , REDATA , RESOLV,RMS,RMSPRT,SCISMC, SOL,SOLT3D,SPCEND, STPISS,SU:DmY, SUPER FIVE CODE , EVALU , NUCODE , MOMENT , P EVALU , P SET, P WCO DE , S ETDAT , S TRS PT , TABLE , T EVALU r-DYN LUMASS : F FRQ FRQMOD,EIGVAL,EICVEC,CSMITH,INTRYG,JACMX, MODAL,0RTHOL,RENORM, SQRUT, TRIPLE,TRID:E,TRYQAD PIS RES PON , CO EF , ERQ KRP , G INTRP , ID!ATRX , INTI AL , S P CTPS. , TIMDPN b. FORM COMPAR . I i A l s. ra-h I t p... L I 5-20 .,? M i
'. 't
- ATTACHMENT 3 5.3 PROGRAM CAPABILITIES The PIPDYN II program is designed for the analysis of three-dimensional piping systems and three-dimensional frame structures. The general capabilities of the program are outlined here under cach major subprogram. A list of dimension limitations of the program is also given so that the user will have an idea of the size of structure that can be solved at present. These dimension limitations may be increased if the user wants to expand the capability of certain options.
PIPSYS PIPSYS is basically a three-dimensio:al, beam-element, structural-analysis program with the following elencat capabilities: (1) Unifo= or tapered straight beam (2) Unifor= or tapered curved team (3) Shear defor=ation (4) FJ exibility factor g It can acco modate the following forms of loading: (1) Distributed load (2) Concentrated" loads at joints (3) Differential thermal expansion (4) Specified displacements (5) Load combinations This program segment is called at several points in an analysis as shown in Section 5.2. 1 Since the program is based on a conventional stif fness method for-mulation, no further discussion is necessary here.
- PLOT 3D PLOT 3D is a three-dimensional plotting program which is used to generate a computer plot of the model of the piping system. The plot is '
5-21
, .'s
- ATTACHMENT 3 r
- k,, . . . .g . y-S . m encellent checking tool f or the user since errors in joint coordinates ' - or element connectivity become visually apparent. CALCC:!P plotting hard-ware and software are necessary for direct use of this option. f AUTO':0 The basic input data is read in in this program. Much effort has been spent to reduce the amount of such data to a minimum. For example,- f node numbers may be omitted; they are then filled in by the computer y-based on a straight run with even spacing. Elbows only require the input of a node on each tangent and the node at :he point of intersection. QI The radius and the remaining geometry such as bend angle and the node i locations at points of tangency are cogured internally. Also, a joint is automatically placed at the center of the elbow. Elements (bars) are ' assigned material, thermal leg, and cross-section type numbers which provide cross-references within the computer to the appropriate data. Mass input f is mass per unit length with optional provision for lumped masses or ' ) aonents of inertia for such items as pu=ps, motors, etc. The solution technique employed in PIPSYS is similar to that ' used in many large structural analyses programs in that it solves a P system equation for=ed for each joint in the system. The si:e (bandwidth) b' cf the system equations is determined by the largest difference in the } jcint nu=bers at the two ends of each element. The larger the bandwidth, " the longer the solution time, which can become quite large for big systems. I' However, in order to number the joints of a systen in such a manner that 4, the smallest bandwidth is obtained, the user must be proficient in the numbering procedure. The resulting joint numbers may be somewhat cumbersome for the user. AUTO:10 allows the user to number the structural Il e -- l joints in any manner he desires because it automatically renumbers the F joints in such a manner that a near-minimum bandwidth is obtained. A l correspondence table which lists the user's joint numbers and the number casigned to that joint by AUTO:10 is printed out so that the user can F easily locate critical points by cross-reference. h i 5-22 - s
, (. ATTACHMENT 3 STRESS G
This subroutine is designed to translate the raw structural analysis ! output into stresses in a form suitable for evaluation in accordance with paragraphs NB-3650 or NC-3650 of the ASME code or the B31.1 Power Piping Code. Since the B31.7 Nuclear Power Piping Code requirements are quite similar to the NB-3650 requirements, the routine is also useful for those applications. In addition to these options, the user may also specify that stresses be computed in accordance w.th the usual beam theory on a load-by-load basis. Load combinations may be performed. l The ASME Section III NB-3600 option for stress evaluation computes the stresses in the prescribed manner using the prescribed stress concen-tration factors for the various components (Table NB 3683.2-1). In addition, the program is designed to provide the major computations necessary to meet the evaluation requirements of NB-3650. To satisy these requirements, the final design analysis must be per-formed with respact to two types of loadings: self-limiting and,non-self-limiting. The various leading cases which fall under each of the above categories must be combined in such a manner that the stress evaluation includes the most severe leading history which could exist in the system. In accordance with the requirements of NB-3650, the following types of loading cenditiens must be used for the final analysis: (1) Dead weight (Equations 9, 13) (2) Sustained Design Mechanical Loads (Equations 9,13) (3) Ther=al Expansion (Equations 10, 11, 12) (4) Thermal Anchor Movement (Equations 10, 11, 12) (5) Earthquake - excluding earthquake anchor movement-response spectra or time history with most severe point in time (Equations 9, 10, 11, 13) (6) Earthquake Anchor Movement (Equations 10, 11) (7) Occasional Mechanical Loads (Equations 10,11) In order to generate the required end moments, a cories of prelim- I ary runs must be made for each of the referenced loading conditions. O 5-23
' , *f . ATTACHMENT 3 i}~ The ASME Section III NC-3600 option for the evaluation of class ! components follows the general format of the NS-3600 option described above. i The NC-3600 option computes the stresses in the manner prescribed in para-graph NC-3650 of the 1972 Winter Addenda, Section II, for all components l'isted in Figure MC-3672.9(a)-1 except miter bends and corrugated pipes. As in the NB-3600 option, a series of preliminary runs must be made in order to generate the element end moments for each of the various load-ing conditions required for the evaluation. The loading conditions re- , quired for the evaluation are as follows: k (1) Dead weight (Equations 8, 9, 11) ! i. (2) Sustained Loads (Equations 8, 9,11) (3) Occasional Loads (Equation 9) f (4), Earthquake (Equation 9) : 1 (5) Earthquake Anchor Movement (Equation 9 or Equations 10 & 11) ! (6) Thermal Expansion (Equations 10 and 11) With a piping system other than one which is very small, the amount of data required for the final analysis becomes overwhelming, introducing a bookkeeping accuracf problem, not to mention the difficulty the designer ' would encounter in combining al.1 the various loads by hr.nd. The stress option of PIPDYN solves both of these problems for the designer because all the information required for the final evaluation is saved on tape and selected and combined internally, thus yielding a rapid and accurate design analysis. For either the NB-3600 or NC-3600 cvaluations, af ter all the necessary .h. loading tapes have been generated, the user simply identifies which load - ing file on the tapes is to be used for the code analysis, and STRES$ will * ' perform all the steps necessary to compute the value of Equations (9) through (14) for NB-3600 or Equations (S) through (11) for NC 3600 f (. structural eceber. or each a In the event of component fails to meet the specified requirements that component vill be flagged for casy identification.Since , the alternative design analysis for class 1 components (MB-3200) oes not d lend itself to automation, the user must perform this cumputation and by h t O [ 5-24 I be* 4
. e
e
*'
- ATTACHMENT 3 in order to qualify the component in question. Ilowever, the output of PIPDEI is arranged so that all of the infomation required for this hand analysis is available and easily identifiable.
Another feature of STRESS is that the designer can perform a code evaluation for any individual loading csse he desires. In this manner the designer can see what contribution to the total solution a specific loading case, earthquake for example, yields. For the B31.1 option, the program prints out the stresses, bar by bar, which have been computed in the prescribed manner from the end forces using the stress concentration factors as specified in B31.1. These are given on an individual load basis. Ilote that with the flexibility affor ad by PIPDtl and STRESS if the pipeline incorporates hydraulic snubbers, which means that the constraint boundary conditions are different for thenal and dynamic loads, the stresses can, still be combined. Separate runs must be made for these loadings which reflect the different boundary :enditions. However, the STRESS routine si= ply reads internal bar forces on a bar-by-bar basis l and no reference is cade to boundary conditions. Load co=bination can be perfor=ed eithe r by internal or external superposition. During a run cade for stress computation (KLUE=1, 4, 5 or 8), the end forces for each element are written on logical unit 1 (KLUE=1 or 5) or logical unit 15 (KI.UI-4 or 8) for each loading case, while the joint displacement data is written on log.1 cal unit 4 If at the end of the stress computation, the user wishes to cambine some loads, he simply calls for internal superposition (KLUE=ll) and the results for the combined loads will follow the non-combined stress output. In order to use internal superposition, a multi-job run in PIPSYS must be made (see description of Card P68); that is, the deck required for superposition must immediately follow the deck used to generate the non-combined stresses. If the user wishes to combine loads at a later time, logical units 1 or 15, 3 and 4 should be saved. then the program is restarted for ! cxternal superposition (KLUE-12) the user can assign up to four dif ferent O
- 5-25 O
. , ,, o ATTACHMENT 3 ,
p. i} . tape sets (one tape for end forces and one tape for joint displacements) from which loads will be extracted and combined. 3
- The superposition output is written on logical unit 2 for possible subsequent use in an 113-3600 or I;C-3600 stress evaluation. If internal superposition is used, the combined results are placed on logical unic 2 af ter the results of the non-combined loads.
DYllSYS , 4 D'C!SYS is a pseudo-consistent mass dynamics program used to determine the natural frequencies and mode shapes of a syste=. DY:!SYS is described 7 as a pseudo-consistent mass program because on the element level masses are concentratad at the structural nodes but, as the system is contracted,
~
h-a consistent mass representation is retained for the structure. By util- A 1:ing this technique, DY::SYS retains the accuracy of a highly distributed mass approxi=ation while reflecting the computational speed of a lumped- I. p p . ass program. Contraction is accomplished by the Rayleigh-Rita method in which the trial displacement functicas are generated by the application of unit . loads at user-selected master nodes. Some judg=ent is required, of k course, but general guidelines are easily established and the method - is much less demanding than lu= ped-mass methods. % Since many practical u-h, problems require the inclusion of twenty or more modes, contraction rep-resents a tremendous saving in computer cost and time as compared to f non-contraction methods, even when the most efficient methods such as power iteration with shifts are used.
- Since site contraction is employed (maximum contracted site is 138 DOF) the eigenvalue-cigenvector extraction can be performed in-core with a very small expenditure of time. The !!ouseholder-Civens procedure ""
is used to reduce the system to tridiagonal form uith subsequent cigenvalue y l extraction (all cigenvalucc) by the Sturm sequence method. The single L weakness of this method, poor orthogonalization for closely spaced l igenvalues, is autom.itically checked for and corrected by une of the _, Cram-Schmidt procedure on the modes involved (if any). ! 5-20 r.
t
'. O ATTACHMENT 3 SYNSYS SYUSYS is capable of determining the dynamic response of a system for the following types of carthquake analysis:
(1) Response Spectra (2) ' Time Dependent Base !!ction (a) All anchors experience the same excitations (b) Each anchor experiences a different excitation (3) Tire Dependent Force Loading The number of modes to be utilized can be directly specified by using a two-step run with a break af ter codal infor:ation is generated, or by specifying either a cut-off frequency or a minimum codal mass. An i=portant feature of SYUSYS is the ability of the program to treat the situation of closely spaced frequencies in the mis sus for the
'l seismic spectral analysis. If tuo f requencies are close (i.e., the t a
percentage difference is within a specified quantity supplied by the uscr), the program adds the stresses due to these two = odes (absolute value sense) and then enters this stress as a single quantity in the PJ!S j sum. ( Also for the response spectra method, SYUSYS can account for the ; total cass for dynamic response by utilizing a quantity called "left-out I mass." From the standpoint of time and accuracy, it is impractical to include all modes in the response co=putation because the higher modes do not usually affect the solution to any great extent. But by selecting only the significant modes, the entire modal effective cass of the system is not used. SYNSYS computes the effective mass associated with these higher codes (lef t-out mass) and computes response for this mass using the basic ground acccinration. The total response generated in this manner closely approximates the solution which would be obtained if all codes were considered. This concept is not necessary for time history probicms since the inertia force method is used to compute stresses nnd displacements. Other options availabic allow the user to ignore the 9 5-27 !,
. a w O $ ts ATTACHMENT 3
. Y-h Acf t-out mass or, alternatively, to include all mass in the lef -out cass load. For all time-dependent solutions, the forcing function (dis-
- place =ent, acceleration, or force) is approximated as a series of straight y line segments, and a closed form integration routine is used. In this k
canner, proble=s relating to cine step sine selection to avoid instability are completely avoided. The ti=e history infor=ation is written on tape as generalized , coordinate values. This allows the storage of rany time points f rom which ene can later choose the most critical. Another useful feature of the time history portion of the program is that as the time history solution proceeds, the times at which the W.
=ax1=um values of the modal accelerations and the lead factors peak are h saved. These times are excellent candidates for peak stress times.
Y cy
. COMPAR $ COMPAR is a small program which is called by DESYS when a cine-P.
E! dependent dyna =ic analysis is perfor:ed. Usually, response is computed at a large number of time noints in order to ensure that the peak response [L,. l 1s included in the output. Because of this large nuater of time points, it is impractical to compute stresses due to the dynamic response at ib - 'l
?'
A' cach time. COMPAR allows the user to select t!e times at which he i desires stress computation and also arranges the response data in such - a fashion that the solution and stress computatirn t.se is greatly reduced. v
} n If the user desires to use the time points _ selected automatically F-*
in SWSYS, this option is available or it can be overridden if the user i . chooses. b- ' PTHEE1 1 L. FIHCPJi is a one-dimensional heat transfer program which computco the time varying bulk temperaturen and radial temperature gradient for D I hemis 5-28 I F
~ . - g i f %' # ATTACHMENT 3 each of the specified fluid transients. Any transient may start or end with a steady-state solution. Both the linear and non-linear temperature integrals as defined in paragraph MB-3653.2 are computed from the re- i sulting radial gradient at cach time. The program also computes the various temperature ranges required. This is done for each thermal leg specified.
The bulk temperatures at any specified time may be transfered back for a thermal expansion analysis. However, it is also possible to independently specify the bulk temperatures for the thermal expansion analysis if desired . l l DIMENSION LIMITATIONS: In its present form, the program occupies about 64,000 (decical) locations of core storage. The core storage allocation cay be changed by altering the following li=itations (subject to change): Maximum number of elements, 1000 Maxi =um nuhber of nodes, 800 (4800 DCF) g Maxi =um bandwidth, 30 nodes (180 DOF)
- Maxi =us nu=ber of loadins cases for a single pass, 100 Maximum nu=ber of ther=al loading cases, 5 Maxi =um nu=ber of nodes at which constraint conditions are .
imposed, 200 Maximum nu=ber of independent ti=c history displace =ent components (attachment degrees of freedom), 30 Max 1=um nu=ber.of independent proportional load vectors (time i history), 30 I Maximum number of time points in displaceme7t, base acccicration, or force history, 800 Maximum number of master node DOF,140 Maximum number of modes used in a dynamic analysis, 40 plus , lef t-out mass (noto here that all 140 eigenvalues and vectors can be printed out for inspection) Maximum number of ther=al icgs, 30 Maximum number of transients per icg, 30 Maximum number of clements connected to a joint, 15 0 5-29}}