ML20132D223
| ML20132D223 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 07/10/1985 |
| From: | Domer J TENNESSEE VALLEY AUTHORITY |
| To: | Adensam E Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20132D227 | List: |
| References | |
| RTR-NUREG-0737, RTR-NUREG-737, TASK-2.D.1, TASK-TM NUDOCS 8507160524 | |
| Download: ML20132D223 (30) | |
Text
9 e
..e TENNESSEE VALLEY AUTHORITY CH ATTANOOGA. TENNESSEE 37401 400 Chestnut Street Tower II July 10, 1985 Director of Nuclear Reactor Regulation Attention: Ms. E. Adensam, Chief Licensing Branch No. 4 Division of Licensing U.S. Nuclear Regulatory Commission Washington,' D.C.
20555
Dear Ms. Adensam:
In the Matter of the
)
Docket Nos. 50-327 Tennessee Valley Authority
)
50-328 Enclosed is our response to your March 25, 1985 letter to H. G. Parris, which requested additional information regarding NUREG-0737, Item II.D.I.
" Performance Testing of Relief and Safety Valves."
If you have any questions concerning this matter, please get in touch with Jerry Wills at FTS 858-2683.
Very truly yours, TENNESSEgVALLEYAUTHORITY
}J.A.Domer, Chief
. d. An Nuclear Licensing Branch
{
Sworn to dnd subscr1Aed before me this 10
- day of U W 1985 Notary Public
()
j] dd po My Commission Expires 6 -
7-Enclosure cc (Enclosure):
U.S. Nuclear Regulatory Commission Region II Attn: Dr. J. Nelson Grace, Regional Administrator l
101 Marietta Street, NH, Suite 2900 4
. L[:g Kt Atlanta, Georgia 30323
[-
Mr. Carl Stahle V
Sequoyah Project Manager U.S. Nuclear Regulatory Commission 7920 Norfolk Avenue g4(f Bethesda, Maryland 20814 gds 8507160524 850710
/),aI' DR ADOCK 0500 7
An Equal opportunity Employer
o ENCLOSURE RESPONSETONRCREQUESTOFMARCH25,198s, FOR ADDITIONAL INFORMATION ON NUREG-0737, ITEM II.D.1, PERFORMANCE TESTING OF PRESSURIZER SAFETY AND RELIEF VALVES SEQUOYAH UNITS 1 AND 2 JUNE 1985
NRC QUESTION
- 1. - In valve operability discussions on cold overpressurization transients, the submittal states that the present PORVs are to be replaced with Target Rock Model 82UU-001 PORVs. The Westinghouse valve inlet fluid conditions report identifies Sequoyah 1 and 2 as being covered by the cold overpressure protection analysis section of the report. Accoru~ag to the Westinghouse 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 and water solid operations. To assure that the PORVs operate for all cold overpressure events, discuss the range of fluid conditions expected for the expected types of fluid discharge and identify the test data that demonstrate operability for these cases. Since no low pressure steam tests were performed for the PORVs, confirm that the high pressure steam tests demonstrate operability for the low pressure steam case for both opening and closing of the PORVs.
TVA RESPONSE
- 1. - The primary concern of cold overpressurization, as expressed in Branch Technical Position RSB 5-2, is to ensure that Appendix G limits of the reactor coolant system components are not exceeded. Since pressure excursions are much more rapid in a water solid reactor coolant system than a system with a steam bubble in the pressurizer, the operational objective is to maintain a bubble in the pressurizer by administrative control. For water solid condition during start-up and shut down, a mitigation system with automatic actuation of the P0kV at predetermined setpoints is implemented to back up the operator. Therefore, variable valve inlet fluid conditions (steam and water) can exist The Sequoyah condition is described in section 5.4, Westinghouse Inlet Fluid Condition Report (EPRI NP-2296), and the range of pressure temperature limit is within the limits shown on Figure 5-1 of the Westinghouse Inlet Fluid Condition Report (EPRI NP-2296).
Pressure control at low temperature oneration of the reactor coolant system is also described in section 5.2.2.4 of the Sequoyah FSAR.
The test results of the Target Rock PORV from both the Marshall Station and Wyle tests are shown in section 4.3 of the EPRI Safety and Relief Valve Test Report (EPRI NP-2628-LD). The test matrices for the pressurizer relief velve in the EPRI test program were diligently selected to be representative over the full range of expected conditions. The high pressure steam tests are considered to envelop the low pressure steam conditions. The valve fully opened on demand and fully closed on demand as previously reported in a TVA letter from L. M. Mills to Ms. Adensam dated January 7, 1983 (A27 830107 025).
In addition to the Marshall /Wyle testing, Target Rock, as requested by TVA under a Watts Bar contract in 1983, ran 500 cycles of low pressure saturated steam at 400 psig and 1300 psig on the same valve model. The valve operation was normal throughout the testing, which further substantiated the correctness of the testing matrices selected in the EPRI testing program. These tests and the actual field operating experiences have fully demonstrated the adequacy of valve operability.
NRC QUESTION
- 2. - Results from the EPRI tests on the Crosby safety valves indicate that the test blowdowns exceeded the design value of 5 percent for both "as-installed" and " lowered" ring settings.
If the blowdowns expected for Sequoyah I and 2 also exceed 5 percent, 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 such that adequate cooling might not be achieved for decay heat removal.
Discuss these consequences of hi her blowdowns if increased blow-8 downs are expected.
TVA RESPONSE
- 2. - EPRI tests 929, 1406, 1415, and 1419 were the most representative of the Sequoyah safety valve with typical ring settings. The highest blowdown as indicated on Table 3-29 of Westinghouse WCAP-10105 is 9.4 percent. Since Crosby used the same methodology in setting the rings for the Sequoyah safety valves as for the EPRI typical tests, the highest blowdown for the Sequoyah valves is expected to be approximately the same.
In conjunction with the Westinghouse Owners Group in the safety valve program, blowdowns were analyzed for 0, 5, 10, and 14 percent. The results from these analyses showed that blowdown of up to 14 percent has no adverse effects on safety for the reference plant. Sequoyah is a reference plant.
NRC QUESTION
- 3. - The Westinghouse inlet fluid conditions report stated that liquid discharge through both the safety and relief valves is predicted for an FSAR feedline break event. The Westinghouse report provided expected peak pressure and pessurization rates for specific plants having an FSAR feedline break analysis. The Sequoyah unit 2 plant was not included in this list of plants having such an FSAR analysis. The Sequoyah unit 1 plant was included in the list of plants having an FSAR feedline break analysis. lue plant submittal did nce discuss the feedline break event. The NUREG-0737, however, requires analysis of accidents and occurrences referenced in Regulatory Guide 1.70, Revision ?, and one of the accidents so required is the feedline break.
Provide a discussion on the feedwater line break event for Sequoyah 2 identifying the fluid pressure, pressurization rate, fluid temperature, valve flow rate, and time duration for the event. Assure that the fluid conditions were enveloped in the EPRI tests and demonstrate operability of the safety and relief valves for this event.
Further, assure that the feedline break event was considered in analysis of the safety / relief valve piping system.
TVA RESPONSE
- 3. - Regarding the lack of reference to Sequoyah unit 2, we assume that the reference is made to Table 5-2 or Table 6-2 of the Westinghouse Inlet Fluid Conditions Report (EPRI NP-2296).
This table inadvertently omitted Sequoyah unit 2.
Sequoyah unit 2 is clearly referenced on pages 1-4 of the report.
Sequoyah unit 2 is identical to Sequoyah unit 1.
All analyses, fluid conditions, pressurization rates, and discussions for unit 1 are applicable to unit 2.
This also has been confirmed by Westinghouse letter WAT-D-6525, dated April 25, 1985 (B45 850429 618).
Effects of the feedline break event was discussed in our submittal dated June 30, 1982, from L. M. Mills to Ms. Adensam (A27 820630 022), and in Westinghouse WCAP-10105, which was referenced as part of our submittal.
Liquid discharge is predicted for the conservative standard FSAR analysis for a double-ended feedline break between the check valve and the steam generator. The range of liquid temperatures predicted at the valve inlet for Sequoyah is between 654 F and 658 F as shown on Tables 5-2 of the inlet condition report.
EPRI test No. 931 indicated that the Crosby 6M6 valve will open and close with no detrimental effect at approximately 650 F water.
Between 654 F and 658 F the liquid will flash to steam after passing through the safety / relief valve. The maximum surge rate into the pressurizer for Sequoyah is 646 gpm (see Table 5-2 of EPRI NP-2296), which is approximately 190,000 lb/hr at 650 F.
The safety valve design flow rate is 420,000 lb/hr and the relief valve is 210,000 lb/hr each.
Sequoyah has three safety valves and two relief valves.
The total maximum flow rate for a feedline break is less than the capacity of any one valve.
The analysis for the discharge piping conservatively assumed all five valves discharging at full flou (see response to questions 12d and 12e).
This analysis envelops the feedline break on piping adequacy.
For the feedline break event, the safety valves open on steam, and no liquid discharge is expected until the pressurizer becomes water solid. The steam to water transition takes at 1 cast 20 minutes.
Successful mitigation action by the operator is expected within this timeframe. Therefore, the safety valve is designed for steam service only.
In the unlikely event of a feedline break and the unsuccessful mitigation action by the operator, the safety valve is still capable of passing water as described above.
The feedline break event is fully described in the Westinghouse Inlet Condition Report and the effects in Westinghouse WCAP-10105.
P NRC QUESTION
- 4. - The EPRI Inlet Fluid Justification Report suggested a method for demonstrating safety valve stability. This method compares the total inlet piping pressure drop for the in plant safety valves and piping to the applicable EPRI test safety valve and piping combinations.
The total inlet piping pressure drop is composed of a frictional and acoustic wave component evaluated under steam conditions. The Sequoyah 1 and 2 plant submittal did not provide pressure drop calculations or any other methods to demonstrate safety valve stability.
Provide the necessary documentation and discussion demonstrating stability for the Sequoyah 1 and 2 plant safety valves at the expected inlet conditions, ring settings and inlet piping configuration.
TVA RESPONSE
- 4. - The Crosby 6M6 valve was tested only with a long inlet piping configuration, which is approximately 15.4 ft in length with approximately 8.3 gallons of seal water. Sequoyah has a short inlet piping configuration with the longest run of approximately 8.2 f t and a seal water volume of less than 7.0 gallons.
Operability of the valves with a shorter inlet and less seal water is expected to be as good or better than that observed with the test configuration. EPRI l
test results consistently showed that the Crosby 6M6 with typical ring settings performed in very stable conditions. Sequoyah safety valves are expected to perform at least as stable or better than as tested.
L Plant specific pressure difference due to piping friction and acoustic wave amplitude associated with both valve opening and closing was calculated in accordance with the methodology prescribed in Appendix B, revision 2, of the "EPRI PWR Safety and Relief Valve Test Program Guide for Application of Valve Test Program Results to Plant Specific Evaluation." The Sequoyah pressure
(
difference for valve opening is 255 psi comparing to the Crosby-6M6 tested pressure difference of 263 psi. For valve closing, the Sequoyah pressure difference is 141 psi comparing to the tested pressure difference of 181 psi.
As expected, the Sequoyah pressure difference is lower than the tested pressure difference and, therefore, operability of the Sequoyah safety valves, due to effects of inlet piping pressure drop, is expected to be more stable than the tested Crosby 6M6 valve. -
9 NRC QUESTION
- 5. - The Westinghouse Inlet Fluid Conditions Report stated that liquid flow could exist through the PORV for the FSAR feedline break event and the extended high prcssure injection event. Liquid PORV flow is also predicted for the i
cold overpressurization event. These same flow conditions will also exist for the block valve. The EPRI/ Marshall Block Valve Report did not test the block valves with fluid media other than steam. The Westinghouse Gate Valve Closure Testing Program did include tests with water; however, the information presented in the report did not provide specific test results. Since it is conceivable that the E510V could be expected to operate with liquid flows, discuss EMOV block valve operability with expected liquid flow conditions and provide specific test data.
TVA RESPONSE
- 5. - As stated in Westinghouse WCAP-10105, no liquid discharge is expected for the feedline break event or the extended high pressure injection event for at least 20 minutes. Operator will take appropriate steps to mitigate the transient within this timeframe.
In cold overpressurization, the block valve must remain open for the PORV to perform the cold overpressurization function. The block valve is not expected to be operated with liquid flow under these transients.
In the unlikely event the PORV failed to close in cold overpressurization control, the block valve will be closed to isolate the PORV from a potential liquid condition.
Please also see response to questions 1, 3, and 7.
The primary function of the block valve is to provide isolation for maintenance or malfunction of the PORV.
Your reference to the " Westinghouse Gate Valve Closure Testing Program" WEMD EM 5683, Rev 1, implies that the Sequoyah Plant has Westinghouse block valves.
As indicated in our June 30, 1982, transmittal, we have Velan model B10-3054B-13MS block valves at Sequoyah. Our June 30, 1982, response makes reference to a June 1, 1982, letter to Harold Denton from R. C. Youngdahl on behalf of participating PWR utilities.
In this letter, reference is made to a July 24, 1981, letter to Harold Denton from R. C. Youngdahl which explained the PWR utilities' reasons for discontinuing further block valve testing. These reasons had been discussed in a July 17, 1981, meeting between NRC staff and utility representatives.
It was concluded that substantial testing had been completed at the Marshall Station to demonstrate block valve operability. At that time the utility participants in the EPRI Safety and Relief Valve Test Program were not requested to perform additional block valve testing using l
water. Although testing of the Westinghouse block valve with liquid flow l
did not provide specific test results on the Velan valve, the test results l
did provide sufficient evidence to conclude the operability of the Velan valve.
In the EPRI/ Marshall tests, both the Westinghouse 3-inch block valve and the Velan 3-inch block valve were tested under similar steam conditions with Limitorque operator SMB-00-10.
The Velan valve fully opened and closed on demand while the Westinghouse valve failed to close fully. When the operator for the Westinghouse valve was changed to the Limitorque SMB-00-15 which has higher motor torque, the Westinghouse valve fully opened and closed on demand.
In the Westinghouse tests with liquid flow, the 3-inch valve with the SMB-00-15 actuator also fully 2 pen d and closed on demand under a differential pressure of up to 2500 lb/in.
It is concluded from test results that the 3-inch Velan valve with the SMB-00-15 operator will open fully and close fully on demand with liquid flow under any plant conditions. The Sequoyah unit 1 block valve is installed with a Limitorque SMB-00-15 operator and unit 2 with a SMB-00-25.
In conclusion, we consider that KUREG-0737, Item II.D.1B is satisfied..
NRC QUESTION
- 6. - Bending moments are induced on the safety valves and PORVs 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 predicted plant moments with the moments applied to the tested valves to demonstrate that the operability of the valves will not be impaired.
TVA RESPONSE
- 6. - A comparison has been made between the predicted plant moments and the mcments applied to the tested safety valve and PORV. The maximum predicted plant moments for any of the safety valves or PORVs is a factor of at least 1.75 less than the moments applied to the tested valves. Therefore, we conclude that the operability of the valves will not be impaired.
NRC QUESTION
- 7. - The Westinghouse Valve Inlet Fluid Conditions Report states that liquid discharge could be expected through the safety valves for both the feedline break and extended high pressure injection events. During some tests, the EPRI 6M6 test safety valve experienced some chatter and flutter while discharging liquid. Testing was terminated after observing chattering to minimize valve damage.
Inspection revealed some valve damage which was presumably caused by the valve chatter and flutter. Liquid discharge for Sequoyah 1 or 2 may conceivably occur for longer periods of time than the EPRI testing. Thus, longer periods of valve chattering may cause severe valve damage. Discuss the implications this may have on operability and reliability of the Sequoyah 1 and 2 safety valves.
Identify any actions that will be taken to' inspect for valve damage following safety valve lift events.
TVA. RESPONSE
- 7. - Based on very conservative assumptions, liquid discharge is predicted for only the feedline break and extended high pressure injection events.
- However, although predicted, liquid discharge is not expected. As reported in Westinghouse WCAP-10105, for both events, the safety valves open on steam, and no liquid discharge is expected until the pressurizer becomes water solid.
The system will not be water solid for at least 20 minutes. Operator action will be taken to mitigate the event within this time period prior to liquid discharge.
It is not conceivable that the Sequoyah safety valves will operate in liquid conditions for longer periods of time than the EPRI testing.
Plant instructions specify the automatic use of the PORVs as secondary means to the pressurizer spray system to control the reactor coolant system pressure.
The reactor also has a trip point approximately 100 psi below the setpoint of the safety valves.
In the very unlikely event all three of the pressure control means were not effective, or for some other reason a safety valve lif ted, the ef fect on the valve would depend on several factors, such as the rate of pressure increases, accumulation, and fluid conditions. The situation would be assessed at the time. Seat leakage can be detected through analysts of the acoustic monitors and thermocouples readings from the safety valve tailpipes. Those readings, in conjunction with the pressurizer relief tank (PRT) level and temperature readings, provide an effective means of establishing safety valve leakage.
If the analysis did not indicate an acceptable condition, the safety valve may be replaced with one of the spare pressurizer safety valves available at Sequoyah.
Valves which have been removed from service are refurbished as necessary and tested before being returned to service.
Events leading to extended high pressure injection and the feedline break event would require extensive and detailed restart plans. Action regarding the pressurizer safety valves would be addressed in those restart plans. <
S NRC QUESTION
- 8. - NUREG-0737, Item II.D.1 requires that the plant-specific PORV control circuitry be qualified for design-basis transients and accidents.
Please provide information which demonstrates that this rct.uirement has been fulfilled.
TVA RESPONSE
- 8. - The Sequoyah PORV is not designed for automatic actuation to meet design basis transients. The PORV is designed to prevent safety valves from lifting for economic reasons (no credit is taken for overpressure protection), to aid the operator in the control of cold overpressurization, and to serve as a high point vent in order to vent noncondensable gases from the reactor coolant system in a degraded core condition to meet NUREG-0737, Item II.B.1.
The controls for the pressurizer power relief valves at Sequoyah Nuclear Plant are fully qualified trained redundant circuits. The Class 1E power is available through a qualified transfer switch in the auxiliary control room to qualified control switches in either the main or auxiliary control rooms.
When the control switches are in the "open" position, all automatic control functions are bypassed, and the circuit is completed through a qualified transfer switch and penetration to the valve solenoid (energize to open).
In the "close" position, all control switch contacts are open; thus, no complete circuit is possible, and the valve remains closed. All automatic control is still bypassed.
In the " auto" position, input signals are supplied from non-divisional pressure and temperature switches.
All PORV control circuitry is designed such that a failure would cause the valve to go to its safe'(deenergized) position, i.e.,
" fail closed."
NRC QUESTION
- 9. - The Sequoyah 1 and 2 plant safety valves are Crosby 6M6 and were tested by EPRI.
EPRI testing of the 6M6 was performed at various ring settings.
The submittal did not provide details discussing the applicable EPRI tests which demonstrate the operability of the plant safety valves. The submittal did not provide the present Sequoyah 1 and 2 safety valve ring settings.
If the plant current ring settings were not used in the EPRI tests, the results may not be directly applicable to the Sequoyah 1 and 2 safety valves.
Identify the Sequoyah 1 and 2 safety valve ring settings.
If the plant specific ring settings were not tested by EPRI, explain how the expected values for flow capacity, blowdown, and the resulting back pressure corresponding to the plant specific ring settings were extrapolated or calculated from the EPRI test data.
Identify these values so determined and evaluate the effects of these values on the behavior of the safety valves.
TVA RESPONSE
- 9. - For the Crosby 6M6 safety valve, seven tests were conducted with typical PWR plant ring settings as prescribed by Crosby.
As identified in the test description in Section 5, Volume 6 of EPRI test report NP-2770-LD, the
" typical PWR plant" ring settings for the tested valve has 71 to 77 notches below the level position for the upper ring and 18 notches below the level position for the lower ring. The Sequoyah unit I and 2 valves have the following level position ring settings (upper, lower):
1-8010A (-81, -18),
1-8010B (-110, -18), and 1-8010C (-116, -18); 2-8010A (-120, -18), 2-8010B
(-45, -18), and 2-8010C (-120, -18).
The Sequoyah spare valve level position ring settings (upper, lower): Spare A (-80, -18), Spare B (-60, -18), and Spare C (-80, -18).
All nine Sequoyah valves also have " typical PWR plant" ring 1415, and 1419, which showed stable valve performance with acceptable blowdown and flow rate.
Blowdown would be in the range of 5.1 percent to 9.4 percent (also see response to Question 2).
Steam flow rates ranged from 467,000 lb/hr to 475,000 lb/hr, which exceeds the design flow rate of 420,000 lb/hr. The backpressure for these tests ranged from 240 psig to 700 psig. The maximum Sequoyah steady state backpressure, assuming conservatively with all three safety and two relief valves discharging, is 610 psig.
The EPRI tests have demonstrated that backpressure of up to 700 psig has an insignificant effect on the Crosby 6M6 valve performance.
X 4
NRC QUESTION
- 10. - Sequoyah 1 and 2 has identified the Target Rock Model 82UU-001 as a replace-ment PORV for the present plant PORVs. The prototype Target Rock Model 80X-006 was tested by EPRI. A water seal test consisting of 110 F water upstream of the PORV followed by 650 F water at 2250 psia was performed on the EPRI test valve to simulate the loop seal conditions upstream of a PORV.
During the performance of this test the EPRI test PORV experienced a delayed closure when the closure command was given. The submittal has stated that loop seals are utilized upstream of the PORVs but did not provide the nominal loop temperature. A potential exists such that the plant conditions could resemble the EPRI test conditions for the loop seal. The submittal has also stated that the Target Rock Model 82UU-001 PORV has been redesigned to improve perfo rmance.
Provide the nominal PORV loop seal water temperature and discuss the prototype to production model design changes and their effects on PORV performance to assure reliable operability under Sequoyah I and 2 inlet fluid conditions to the PORV.
TVA RESPONSE
- 10. - In our letter to Ms. E. G. Adensam from our J. A. Domer dated March 18, 1985, we inf ormed NRC that we have decided to delete the PORV loop seal.
Conseruently, we no longer have t'e wcter seal simulation test problem 3
a that occurred to the Target Rock <.>0RV during the EPRI testing.
4 4
I -
NRC QUESTION
- 11. - As a result of discussions with TVA on May 9 and 10, 1984, the staff learned that Sequoyah 1 experienced excessively leaking safety valves after draining water loop seals upstream of the valves. One valve reportedly leaked so excessively that the valve may have lifted and relieved some amount of system pressure.
Similarly, on August 20, 1984, Sequoyah 2 experienced a ruptured disk on the pressurizer relief tank due at least partly to excessive leakage through a safety valve. Why did these valves leak so excessively even after having been modified with intervals designed for steam service and what actions are planned to eliminate excessive leakage in the future?
TVA RESPONSE
- 11. - We believe the major cause of leakage past the valve seat was high static piping loads imposed on the valve outlet flange by the tailpipes. Tests directed by TVA in 1984 confirmed valve seat leakage is influenced by static loading on the valve outlet flange.
Crosby subsequently furnished the flange loads below which seat leakage is not expected to be caused by the valve body distortion. The cold deadweight piping load on the outlet flange of two of the three safety valves was determined for unit 2 during the cycle 2 refueling and maintenance outage. The constant support spring hanger on those two valves' outlet flange was changed to reduce the load to be within the desirable range. TVA will consider determining the cold deadweight piping load for other pressurizer safety valves in both units and changing the constant support spring hangers as necessary to bring the load into the desirable range. The valve's service being changed to heated loop seals is also expected to reduce the likelihood and consequences of valve seat leakage.
The following statements regarding the rupture disc on the pressurizer relief tank are from Reportable Occurrence Report SQR0-50-328/84013, Revision 1, submitted to NRC on February 8, 1985.
Leakage from the safety valve and from a pressurizer spray valve stem leakoff line resulted in minor fluctuations in PRT parameters.
Normally, these fluctuations would not be significant; however, the weakened state of the rupture disc aggravated by the PRT parameter fluctuations resulted in
~
premature rupture of the disc.
. The valve deficiency is not considered to have degraded the safety of the plant during the time from the discovery of the leak until the valve was replaced, since the valve would have lifted and RCS integrity was not degraded. _
NRC QUESTION
- 12. - The submittal states that a thermal hydraulic analysis of the safety / relief valve piping system has been conducted, but does not present details of the analysis. To allow for a complete evaluation of the methods used and the results obtained from the thermal hydraulic analysis, provide a discussion on the thermal hydraulic analysis that contain at 12ast the following information:
Evidence that the analysis was performed on the fluid transient a.
cases producing the maximum loading on the safety /PORV piping system. The cases should bound all steam, steam to water, and water flow transient conditions for the safety and PORV valves.
TVA RESPONSE 12a. - The thermal-hydraulic analysis was performed on the transient which gives the highest pressurization rate (locked rotor 144 psi /sec) and which gives the highest peak pressure (loss of load, 2555 psia)ggy transient This combination in conjunction with the assumed valve opening time and valve opening area assures a bounding fluid acceleration.
As these transients are both steam initiating, this bounding acceleration produces a conservative load on the piping system (i.e., greater than the maximum loads) as the heated safety valve loop seal is driven downstream of the safety valve.
This bounding acceleration produces a conservative load for both the piping connected to the safety valves as well as the pipe connected to the PORVs since no other transients are considered which would provide a higher combination of acceleration and intact mass (i.e., force) and/or are not considered likely due to operator action.
NRC QUESTION 12b. - A detailed description of the methods used to perform this analysis.
This includes a description of methods used to generate fluid pressures and momenta over time and methods used to calculate resulting fluid forces on the system.
Identify the computer programs used for the analysis and how these programs were verified.
TVA RESPONSE l
l 12b. - The methods used to perform this analysis are described in the code l
manuals for the computer programs used in this analysis.
Jge)*#** ~
3 l
hydraulic analysis was performed by using the RELAP5/ MODI computer j
code. Thefluidforcyg)onthesystematspecifiedmodeswerecalculated by use of the REPIPE computer code. The REPIPE computer code post l
processes the RELAPS thermal-hydraulic data to obtain the " wave" and l
" blowdown" force components which are then summed to determine the total force on each pipe segment. This combination of computer programs has been used at TVA for numerous applications involving waterhammer phenomena.
The computer codes RELAPS and REPIPE are quality assured by the Control Data Cybernet Computer Service Bureau. During this procers, the code is shown to correctly perform test problems supplied by the authors.
Application of RELAPS to this type program has been performed as part of the EPRI test program (reference 5), where RELAPS results were shown to adequately predict the EPRI experimental data. )
t NRC QUESTION 12c. - Identification of important parameters used in the thermal hydraulic analysis and rationale for their selection. These include peak pressure and pressurization rate, valve opening time, and fluid conditions at vs1ve opening.
TVA RESPONSE 12c. - As indicated in response "a", the thermal-hydraulic analysis parameters were chosen such that the calculated loads on the piping system are conservative relative to the maximum predicted loads for the bounding FSAR type accident. Again, the peak pressure was chosen as 2555 psia. This pressure results from the loss of load transient analysis as described in reference 1.
Also, the maximum predicted pressurization rate is 144 psi /sec which results from the locked rotor transient.
This analysis conservatively assumes a pressurization rate of 150 psi /sec. The application of the combination of the above to the analysis is not necessarily congruous (i.e.,
if the accident which gives the peak pressure is occurring, the pressurization rate is predicted to be significantly lower than 144 psi /sec, or if the accident which gives the maximum pressurization rate is occurring, the peak pressure is predicted to be lower than 2555 p;ia).
In addition, if the PORVs are assumed to operate, the transieats mentioned above are less severe.
The safety valve opening time has been conservatively assumed to be 9 milliseconds (ms). This is the fastest " pop" opening time reportyg)for the Crosby 6M6 safety valve with a loop seal in the EPRI studies Reference 5 used a cafety valve opening time of 90 ms to benchmark RELAPS to EPRI piping load studies performed using an identical safety valve with similar initial conditions as that at SQN.
The PORV opening time was conservatively modeled as 60 ms.
This opening time includes the 25 percent uncertainty estimated by the valve vendor (Target Rock) and is faster than that recorded in postmodification tests at both SQN and WBN.
As indicated in response 10, the PORV piping no longer maintains a loop seal upstream of the PORV, thus the fluid conditions at the time of PORV opening is assumed to be saturated steam at the PORV opening setpoint pressure of 2350 psia.
The design of the loop seals on the safety valves ensures that a minimum g[
water temperature of 300 F is maintained near the valve seat. The thermal-hydraulic analysis has conservatively assumed the loop seal temperature distribution illustrated in Figure 1.
This distribution is based upon inhouse thermal calculation with HEATING 5 for both insulated and uninsulated piping and engineering judgment.
In addition, a portion of the loop seal has been placed downstream of the safety valve (as shown in Figure 1) based upon a conservative application of the guidance given in reference 5.
Modeling the loop seal distribution in this fashion is conservative relative to analysis which could be performed using a dynamic valve model. This tcmperature distribution is known to be conservative relative to another plant with a similar insulating scheme which has performed temperature measurements. -
NRC QUESTION 12d. - An explanation of the method used to treat valve resistances in the anlysis. Report the valve flow rates that correspond to the resistances used.
Because the ASME Code requires derating of the safety valves to 90% of actual flow capacity, the safety valve analysis should be based on flows equal to 111% of the valve flow rating, unless another flow rate can be justified.
Provide information explaining how derating of the safety valves was handled and describe methods used to establish flow rates for the safety valves and PORVs in the analysis.
TVA RESPONSE 12d. - The safety valve orifice area was effectively modeled as 0.0213 ft This area results in a calculated safety valve steam flow rate of approximately 570,000 lbm/hr at the safety valve setpoint pressure of 2500 psia. The rated flow capacity is 420,000 lbm/hr. Thus the analytical flow rate is greater than 128 percent of the design flow rate.
The PORV has a manufacturer stated maximum flow rate of 210,000lbm/gr steam at 2265 psia.
The analysis assumed an orifice area of 0.01 ft which yield a flow rate of slightly greater than 210,000 lbm/hr at 2265 psia.
NRC QUESTION 12e. - A discussion of the sequence of opening of the safety valves that was used to produce worst case loading conditions.
TVA RESPONSE 12e. - The thermal-hydraulic model conservatively assumes that the two PORVs discharge simultaneously followed by the three safety valves discharging simultaneously. The governing loads from the fluid momentum is sufficiently broad that the common header portion of the discharge piping " sees" a singular load.
In addition, sensitivity analyses have indicated that the piping loads are considerably reduced for valve discharges which are not occurring simultaneously but are discharging with intervals greater than approximately 0.3 seconds between lifts.
NRC QUESTION 12f. - A sketch of the thermal hydraulic model showing the size and number of fluid control volumes.
TVA RESPONSE 12f. - The entire pressurizer discharge piping has been modeled. The total number of fluid control volumes is 307 An example of the size of the fluid control volumes for a representative safety valve discharge piping is given in Figure 1.
a A
4 4
L
~
dh.h_
e
-.4-
,a
&.-A
.e
-4+h..4
+&4-
..&4 s
b.*
v.
l NRC QUESTION l
12g. - A copy of the thermal analysis report.
TVA RESPONSE 2
12g.
.The thermal-hydrarlic analysis report is quite voluminous and is not provided here. However, the report has been checked and approved and is available for review at your discretion. The document identifier is i
TI-ANL-96R2 and is retrievable from TVA central files under accession number B45 850516 235.
l
{
l I
4 1
I 5
l e
I i -
4 t
l 1
?
I 16-
5 Figure 1 RELAP5 Typical Nodulization and Safety Valve'(SV)
Loop Seal Temperature Distribution SAFETY VALVE j
-~%
l 2-l
/
s I
/
\\
/ PRESSURIZER
\\
3
\\
i v
l A
l
~
4 8
5 6
C
~
PRESSURE AT SV LIFT = 2500 psia TYPICAL NODALIZATION LENGTHS TEMPERATURE (T) DISTRIBUTION (for Nodes Shown)
T = 668 F (Saturated Steam)
L = 1.0439 ft.
3 1
T = 668*F (Saturated Steam)
L = 1.5805 f t.
2 2
T3 = 668 F (Saturated Liquid)
L = 1.5805 ft.
3 T4 = 570 F (Subcooled Liquid)
L4 = 1.5805 f t.
T = 400 F (Subcooled Liquid)
L 1.0876 ft.
5 5
T6 = 338 F (Subcooled Liquid)
L6 = 1.1752 f t.
T = 300*F (Subcooled Liquid)
L = 0.5940 ft.
g g
T b Conditions Predicted by L = 0.78% f t.
B C
B RELAP From PORV Lift L
C J
REFERENCES FOR NRC QUESTION 12 1.
" Valve Inlet Fluid Conditions for the Pressurizer Safety and Relief Valves in Westinghouse Designed Plants," EPRI NP-2296, Final Report, dated December 1982.
2.
"RELAPS/ MODI Code Manual, Volume 1: System Models and Numerical Methods,"
EG&G Idaho, Cycle 14, NUREG/CR-1826, EGG-2070 Draft, Revision September 1981.
3.
"RELAPS/ MOD 1 Code Manual, Volume 2: Users Guide and Input Requirements,"
EG&G Idaho, Cycle 14, NUREG/CR-1826, EGG-2070 Draf t, Revision September 1981.
4.
"REPIPE, Application Reference Manual," Cybernet Services CDC, July 1980.
5.
" Application of RELAP5/ MODI for Calculation of Safety and Relief Valve Discharge Piping Hydrodynamic Loads," ITI, Interim Report, July 1982, NP-2479-LD, Research Project V102-28.
6.
"EPRI PWR Safety and Relief Valve Test Program, Safety and Relief Valve Test Report," EPRI, Interim Report September 1982, NP-2628-LD, Project V102.
7.
" HEATING 5-NIBM36, Heat Conductor Program," W. D. Turner, D. C. Elrod, and I. I. Simmons - tov, ORNL/CSB/TM-15. -
NRC QUESTION
- 13. - The submittal states that a structural analysis of the safety /PORV valve piping system has been conducted, but does not present details of the analysis.
To allow for a complete evaluation of the methods used and results obtained from the structural analysis, please provide reports containing at least the following information:
- a. A detailed description of the methods used to perform the analysis.
Identify the computer programs used for the analysis and how these programs were verified.
TVA RESPONSE 13a. - The pressurizer safety /PORV valve piping was analyzed using the TPIPE computer program. The piping was represented by a mathematical model that described the physical system in terms of geometry, pipe / component materials, cross-sectional properties, and support types and locations. The TPIPE computer program has been benchmarked by TVA against the NRC program EPIPE in accordance with the Stardard Review Plans, NUREG-0800 and NUREG/CR-1677 (reference SAR section 3.9.2.5.3).
NRC QUESTION 13b. - A description of the method used to apply the fluid forces to the structural model.
Since the forces acting on a typical pipe segment are composed of a net, or " wave," force and opposing " blowdown" forces, describe the methods for handling both types of forces.
TVA RESPONSE 13b. - Ilydraulic forcing functions are received in the form of time dependent net fluid forces on individual straight pipe runs.
The force direction is in pipe axial direction and, because pipe segments are rigid in that direction, the force is applied at one node point along the straight pipe run. The choice of node point has no effect on the piping system response to the forcing function. Derivation of the net fluid force is addressed in our response to question 12b.
NRC QUESTION 13c. - A description of methods used to model supports, the pressurizer and relief tank connections, and the safety valve bonnet assemblies and PORV actuator.
TVA RESPONSE 13c. - Supports were modeled in the analysis without flexibility according to type. Spring type supports were considered in the normal (primary) loading condition, snubbers were considered in the upset and faulted (primary) loading conditions, and rigid supports were considered in normal, upset, and faulted (primary and secondary) loading conditions.
The pressurizer and relief tank were modeled as lumped masses connected by members and internal supports to more closely represent the response and flexibility of the equipment. The safety valve bonnet assemblies and PORV actuators were modeled as eccentric lump masses located at the center of gravity. -
NRC QUESTION 13d. - An identification of the load combinations performed in the analysis together with the allowable stress limits. Differentiate between load combinations used in the piping upstream and downstream of the valve.
Explain the mathematical methods used to perform the load combinations, and identify the governing codes and standards used to determine piping and support adequacy.
TVA RESPONSE 13d. - The piping is qualified for stress according to the USAS B31.1-1967 power piping code. This code did not define stress levels for the loading combinations listed below. However, the stress levels which are in agree-ment with NC3000 of the ASME, section III, Winter 1972 Addenda are considered to be equivalent to the 31.1 code with appropriate consideration to the modifications where they exist. The supports are designed in accordance with AISC specification for design.
Loading combinations considered:
PLANT LOAD CONDITION SOURCES COMBINATIONS Normal Primary Pressure + Dead Weight Upset Primary + Occasional Pressure + Dead Weight +
Relief Valve Blowdown +
Operating Basic Earthquake or OBE Rigid Response Faulted Primary + Occasional Pressure + Dead Weight +
Safety Valve Blowdown +
Relief Valve Blowdown +
Safe Shutdown Earthquake or SSE Rigid Response or g
Operating Basis Earthquake or OBE Rigid Response Normal and Upset Secondary Thermal Operating Conditions Normal and Upset Primary + Secondary Pressure + Dead Weight +
Thermal Operating Conditions) 1These loads were not required in the faulted condition but were considered for additional conservatism only.
Stress Limits, S and S were taken from the USAS B31.1-1967 Code at design n
g temperature and 70*F, r'espectively:
Normal (Primary) 1.0 S NOTE:
h Upset (Primary) 1.2 S S = 1.25 S +.25 S h
A c
h Faulted (Primary) 2.4 Sh Normal / Upset (Secondary) 2.4 SA Normal / Upset (Primary + Secondary) 2.4 S - -
r 13d. - (continued) The load case information was developed as appropriate for both upstream and downstream of the valves. The safety and relief valve blowdown loads were defined by time history forcing functions.
Net forces were applied to pipe segments. The direct integration method with a 1.0 percent model damping was used to evaluate the time history load cases.
The total dynamic response of each mode of vibration by the square root of the sum of the squares (SRSS) technique and combining the resultant individual modal responses by the NRC grouping method. -
NRC QUESTION 13e. - An evaluation of the results of the structural analysis, including identification of overstressed locations and a description of modifications, if any.
TVA RESPONSE 13e. - The piping analysis is qualified according to the USAS B31.1-1967 Piping Code as follows:
1.
Pipe stresses for the various loading combinations are within the allowable limits defined by NC3000 of the ASflE section III, Winter 1972 Addenda.
2.
The nozzle connections at the pressurizer and the pressurizer relief tank are qualified to acceptable limits.
3.
The forces and moments acting on the safety valves are within acceptable limits.
4.
All valves are within acceptable acceleration limits.
The unit I safety valves will be converted to heated loop seal service and the pipe supports modified as necessary for multiple actuations under postulated worst case conditions during the cycle 3 refueling and maintenance outage.
The unit 2 piping has been converted to heated loop seal service and the supports modified for a single worst case valve actuation. Additional support.
modifications to qualify the piping for multiple worst case valve actuations will be performed during the cycle 3 refueling and maintenance outage.
NRC QUESTION 13f. - A sketch of the structural model showing lumped mass locations, pipe sizes, and application points of fluid forces.
TVA RESPONSE 13f. - The information is shown on the following TVA drawings. Copies are attached.
Unit 1 47K465-60 R1 Unit 2 47K465-55 R3 47K465-61 R1 47K465-56 R4 47K465-62 R1 47K465-57 R2 47K465-63 R1 47K465-58 R4 NRC QUESTION 13g. - A copy of the structural analysis report. \\
. ~ -
r TVA RESPONSE 13g. - The reports are on file in TVA under accession numbers:
Unit 1 - B25 850507 802 Unit 2 - SQP 841220 005 The full report of each is over 2 inches thick containing detailed data that should not be necessary for your review. The following is a summary of the report that should be sufficient for your evaluation on valve operability. -
i 1.0 PROBLEM DESCRIPTION The analysis problems consist of the pressurizer relief system extending from the pressurizer relief valve nozzles to the pressurizer relief tank inlet nozzle. The pressurizer relief system has three self-actuated spring-loaded safety relief valves (SRVs) installed in the 6-inch piping and two power-operated relief valves (PORVs) installed in the 3-inch piping. The relief valve discharge piping joins in a common 12-inch header before discharging into the pressurizer relief tank (PRT).
2.0 ANALYTICAL METHODS The pressurizer relief piping was analyzed using the TPIPE computer program. The piping was represented by a mathematical model that described the physical system in terms of geometry, pipe / component materials, cross sectional properties, and support types and locations.
Supports were modeled in the analysis without flexibility according to type.
Spring-type supports were considered in the normal (primary) loading conditon; snubbers were considered in the upset and faulted (primary) loading conditions; and rigid supports were considered in normal, upset, and faulted (primary and secondary) loading conditions.
The pressurizer and relief tank were modeled as lumped masses connected by members and internal supports to more closely represent the response i
and flexibility of the equipment. The safety valve bonnet assemblies and PORV actuators were modeled as eccentric lump masses located at the i
center of gravity.
The analysis model matches the as-constructed configuration with the heated loop seal option.
3.0 LOAD CASE INFORMATION
)
3.1 Peak Pressure - The peak pressure in the relief valve upstream piping is 4700 lb/in g.
Thepeakpressureinghefirstpipe segment downstream from the PORVs is 837 lb/in g. In the rest of
'i.'
the piping, the peak pressure does not exceed the design pressure.
3.2 Deadweight Analyses - The deadweight analyses qualify the support system to carry the empty, as well as water-filled (hydro test),
piping.
Seismic analysis was not performed for the water-filled system since the condition represented a hydrotest (preoperating condition).
3.3 Thermal Analyses - The thermal analyses qualifies the piping for j
the normal operating and abnormal plant conditions.
In addition, the piping was evaluated for stress due to a thermal range load Case.
3.4 Cold Springing - Cold springing (the addition of pipe length at the desired location simulated by a thermal analysis) is employed to modify the PRT nozzle loads.
In order to simulate the cold springing condition, a thermal analysis was run where the appropriate piping was heated at 670*f.
The rest of the piping was at 70*F.
J{
4 24-
s 3.5 Seismic Response Spectrum Analyses - OBE and SSE analyses were performed for the piping filled with steam or air as applicable. The pressurizer is filled with water up to the normal operating level. Two-dimensional earthquakes were evaluated using north-south and vertical spectra and east-west and vertical spectra.
3.6 Rigid Response Analyses - OBE and SSE rigid response load cases were run.
Because of the many axial pipe restraints in the pressurizer relief system, the system is dynamically rigid, and it was necessary to check the zero period-acceleration for their influence on the axial pipe restraints.
3.7 Relief Valve Discharge Analyses - The transient resulting in the maximum steam discharge is a combination of two transients, the loss of load event and the reactor cociant pump locked rotor event. These two events were combined because they resulted in maximum pressurization rate and maximum pressure in the pressurizer.
The forcing functions generated from the relief valve discharge load show that the pipe load transients due to PORV opening have died out af ter 0.3 seconds, long before the SRVs open.
The TPIPE prograr. computer core usage does not permit the use of all forcing functions simultaneously. The time history runs were therefore subdivided into one run for the PORVs, one run for the header, and one run for each SRV line. The direct integration method with a 1 percent modal damping was used to evaluate the time history load cases. Damping constants alpha and beta were computed based on a frequency range of 10 to 100 Herz. An integration time step size of 0.001 seconds was chosen, and the integration was carried out to 0.35 seconds for the PORV lines and 0.2 seconds for the SRV lines.
3.8 Pressure Oscillations - The water loop seal pressure oscillations have been evaluated subsequently from the standpoint of pressure-retaining capability of the piping upstream of the safety valves.
3.9 Loading Combinations Considered:
PLANT LOAD CONDITION SOURCES COMBINATIONS Normal Primary Pressure + Dead Weight Upset Primary + Occasional Pressure + Dead Weight + Relief Valve Blowdown + Operating Basic Earthquake or OBE Rigid Response Faulted Prim,ry + Occasional Pressure + Dead Weight +
Safety Valve Blowdown + Relief Valve Blowdown + Safe Shutdown EarthquakeorSSERigidRespogse or Operating Basis Earthquake or g
OBE Rigid Response Normal and Upset Secondary Thermal Operating Conditions Normal and Upset Primary + Secondary Pressure + Dead Weight + Thermal Operating Conditons 1
~
These loads were not required in the faulted condition but were considered for additional conservatism only.
Stress limits, S and S were taken from the USAS B31.1 - 1967 h
g code at design temperature and 70 F, respectfully:
NORMAL (PRIMARY) 1.0 S h UPSET (PRIMARY) 1.2 Sh FAULTED (PRIMARY) 2.4 Sh NORMAL / UPSET (SECONDARY) 2.4 S NOTE:
S = 1.255c +.25S A
h NORMAL / UPSET (PRIMARY + SECONDARY) 2.4S4 4.0 RESULTS The piping analyses are qualified according to the USAS B31.1 - 1967 piping code as follows:
4.1 Pipe stresses for the various loading combinations are within the allowable limits defined by NC3000 of the ASME section III, Winter 1972 addenda.
4.2 The nozzle connections at the pressurizer and the pressurizer relief tank are qualified to acceptable limits.
4.3 The forces and moments acting on the safety valves are within acceptable limits.
4.4 All valves are within acceptable acceleration limits.
4.3 Effects due to the water loop seal pressure oscillations are insignificant.
t -- -
a 5.0 CONCLUSI0h The piping is qualified for stress according to the USAS B31.1 - 1967 power piping code. The code did not define stress levels for the loading combinations utilized. Ilowever, the stress levels which are in agreement with NC3000 of the ASME section III Winter 1972 addenda are considered to be equivalent to 31.1 code with appropriate consideration to the modifications where they exist. The supports are designed in accordance with AISC specification for design.
NRC QUESTION
- 14. - According to results of EPRI tests, high frequency pressure oscillations of 170-260 IIz typically occur in the piping upstream of the safety valve while loop seal water passes through the valve. An evaluation of this phenomenon is documented in the Westinghouse report WCAP 10105 and states that the acoustic pressures occurring prior to and during safety valve discharge are below the maximum permissible pressure. 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, the pressure oscillations could potentially excite high frequency vibration modes in the piping, creating bending moments in the inlet piping that should be combined with moments from other appropriate mechanical loads.
Provide one of the following:
(1) a comparison of the expected peak pressures and bending moments with the allewable values reported in the WCAP report, or (2) Justification for i
other alternate allowable pressure and bending moments with a similar comparison with peak pressures and moments induced in the plant piping.
TVA RESPONSE
- 14. - The water loop seal pressure oscillations have been considered in the evaluation of the inlet piping upstream of the safety valve.
Based on Westinghouse Report WCAP-10105, a peak pressure of 4700 psia was developed for the Sequoyah 6-inch schedule 160 loop seal piping which is less than the Westinghouse defined maximum permissible pressure of 5229 psia for service level B and 7131 psia for service level C.
The ficxibly-supported piping does not respond to the high frequency (170 to 260 liz) waterhammer input. Consequently, no significant bending moment will occur.
The bending moment is much greater for loop seal discharge than for high frequency pressure oscillation. The larger bending moment was used in the analysis, i.
e 4
ATTACHMENT 1 LIST OF DFAWINGS INCLUDED
- Three (3) copies of each drawing provided
,