ML13331A835
| ML13331A835 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 11/20/1985 |
| From: | SOUTHERN CALIFORNIA EDISON CO. |
| To: | |
| Shared Package | |
| ML13331A833 | List: |
| References | |
| NUDOCS 8512040103 | |
| Download: ML13331A835 (26) | |
Text
ATTACHMENT 2 PROPOSED SPECIFICATIONS P
4.2 SAFETY INJECTION AND CONTAINMENT SPRAY SYSTEM 4.2.1 SAFETY INJECTION AND CONTAINMENT SPRAY SYSTEM PERIODIC TESTING APPLICABILITY:
Applies to testing of the Safety Injection System and the Containment Spray System.
OBJECTIVE:
To verify that the Safety Injection System and the Containment Spray System will respond promptly and properly if required.
SPECIFICATION:
I. System Test A. Safety Injection System (1) During reactor shutdown at intervals not longer than the normal plant refueling intervals, a "no-flow" system test shall be conducted to demonstrate proper availability of the system. The test shall be performed in accordance with the following procedure:
(a) The feedwater, safety injection, charging, condensate, and heater drain pumps shall not be operating. Their respective breakers shall be racked-out to the test position with control power available.
(b) The flow path for condensate shall be positively blocked prior to the test.
(c) Injection and recirculation system operation shall be initiated by instrumentation and controls installed in the control room.
(2) The test will be considered satisfactory if control board indication and visual observations indicate all components have operated and sequenced properly. That is, the appropriate pump breakers have opened and closed, all valves have completed their travel, and valves HV-851A and HV-851B have actuated in 3 to 5 seconds.
(3) A test of the trisodium phosphate additive shall be conducted to demonstrate the availability of the system. The test shall be performed in accordance with the following procedure:
(a) The three (3) storage racks are visually observed to have maintained their integrity.
(b) The three (3) racks, each with a storage capacity of 1800 pounds of anhydrous trisodium phosphate additive, are visually observed to be full.
4-36 Revised:
6/23/81
0 Page 2 of Attachment 2 BASIS:
The Safety Injection System is a principal plant safeguard. It provides means to insert negative reactivity and limits core damage in the event of a loss of coolant or steam break accident.(1)( 2)(3)
Preoperational performance tests of the components are performed in the manufacturer's shop. An initial system flow test demonstrates proper dynamic functioning of the system. Thereafter, periodic tests demonstrate that all components are functioning properly. For these tests, flow through the system is not required.
The tests specified above will demonstrate that all components which do not normally and routinely operate will operate properly and in sequence if required.
The time limit on operation of valves HV851A and HV851B assures that the valves have properly actuated and within the time constraints assumed for the accident analysis as described in Reference 4. The portion of the Recirculation system outside the containment sphere is effectively an extension of the boundary of the containment. The measurement of the recirculation loop leakage ensures that the calculated EAB 0-2 hr. thyroid dose does not exceed 10 CFR 100 limits.
The trisodium phosphate stored in storage racks located in the containment is provided to minimize the possibility of stress corrosion cracking of metal components during operation of the ECCS following a LOCA. The trisodium phosphate provides this protection by dissolving in the sump water and causing its final pH to be raised to 7.0 - 7.5.
The requirement to dissolve trisodium phosphate from one of the sample storage racks in distilled water heated and borated, to the extent recirculating post LOCA sump water is projected to be heated and borated, provides assurance that the stored trisodium phosphate will dissolve as required following a LOCA. The sample storage racks are sized to contain 0.5 pounds of trisodium phosphate. Trisodium phosphate stored in the sample storage racks has a surface area to volume ratio of 1.33 whereas the trisodium phosphate stored in the main racks has a surface area to volume ratio of 1.15.
Visual inspection of the non-redundant piping in the Containment Spray System provides additional assurance of the integrity of that system.
References:
(1) Final Engineering Report and Safety Analysis, Paragraph 5.1.
(2) "San Onofre Nuclear Generating Station", report forwarded by letter dated December 29, 1971 from Jack B. Moore to Director, Division of Reactor Licensing, USAEC, subject:
Emergency Core Cooling System Performance, San Onofre Nuclear Generating Station, Unit 1.
(3) USAEC Safety Evaluation of ECCS Performance Analysis for San Onofre Unit 1, forwarded by letter dated March 6, 1974 from Mr. Donald J. Skovholt to Mr.
Jack B. Moore.
(4) Letter, K. P. Baskin, SCE, to D. M. Crutchfield, NRC, dated October 16, 1981.
4-39
Page 3 Attachment 2 4.2.3 Deleted 5112F
ATTACHMENT 3 TECHNICAL BASIS FOR THE PROPOSED CHANGE TO TECHNICAL SPECIFICATION 4. 2. 3 SAFETY INJECTION SYSTEM HYDRAULIC VALVE TESTING SAN ONOFRE NUC1EAR GENERATING STATION UNIT 1 SEPTEMBER 30, 1985
TABLE OF CONTENTS
- 1.
SUMMARY
- 2.
FAILURE HISTORY
- 3.
SIS CONFIGURATION AND OPERATION PRIOR TO LER 81-020
- 4.
CAUSES OF HV-851A AND HV-851B OPERATIONAL FAILURE
- 5.
SIS MODIFICATIONS TO PREVENT A REOCCURANCE
- 6.
SIS INTERIM TESTING REQUIREMENTS
- 7.
EVALUATION OF THE INTERIM TESTING RESULTS
- 8.
VERIFICATION OF THE ADEQUACY OF THE SIS MODIFICATIONS
- 9.
PROPOSED LONG-TERM SURVEILLANCE
- 10.
JUSTIFICATION FOR THE PROPOSED LONG-TERM SURVEILLANCE
- 11.
REFERENCES 1
- 1. SUMMARTI The Safety Injection System of SONGS Unit 1 has been tested six times during the current fuel cycle to meet an interim licensing requirement. The surveillance program associated with the interim licensing requirement, which arose from the failure to open of the safety injection feedpump discharge valves during a safety injection actuation sequence, is subject to review and replacement with a long-term surveillance program by the beginning of fuel cycle IX.
The test results obtained over a four year period indicate that the modifications made to the operation and to components of the SIS have successfully remedied the causes that led to the operational failure of the above valves. Therefore, additional functional tests of the SIS during power operation is not warranted and a verification of the functioning of the above valves under no-load conditions is recommended as the long-term surveillance program.
Included herein are the technical basis for this recommendation presented in a sectional format. Section 2 describes the plant transient that led to the SIS valve failure. Section 3 highlights the system configuration and its operation prior to the failure while section 4 discusses the identified causes. In section 5 the modifications made to the SIS to prevent a valve failure reoccurrance are discussed. Section 6 describes the licensing commitment embodied in Tech Spec 4.2.3; and section 7 presents the evaluation of the results from the 6 tests carried out to date. In section 8 the reader will find a discussion of how these results verify the adequacy of the SIS modifications.
Finally section 9 presents the proposed long-term surveillance program, and section 10 adresses the justification for the proposed recommendations.
2
- 2. FAILURE HISTORY On September 3, 1981, SONGS 1 was operating normally at 390 MWe, 87% reactor power. At 0330 many apparently unrelated alarms were received in the control room, and erratic or failed indications were observed on several instruments.
(This condition was subsequently identified as resulting from failure of the #1 Regulated Power Supply.)
At 0332 the Control Operator manually tripped the reactor based on the following observations:
Apparently erratic feedwater flow and unresponsive manual controls.
Other instrument indications which were erratic or in conflict.
During the ensuing transient reported to the NRC in LER 81-020 (Reference 1), Reactor Coolant System pressure dropped below that necessary to actuate safety injection. During operator surveillance of the safety injection actuation sequence, it was observed that neither feedwater pump safety injection discharge valves (HV-851A and HV-851B) opened. The feedwater pumps were then secured. As the pressure upstream of HV-851A and HV-851B decreased, the valves opened. Safety injection was not actually needed for this transient, and would not have actually occurred until Reactor Coolant System pressure decreased below the discharge pressure of the feedwater pumps (i.e., approximately 1200 psi).
Reactor Coolant System pressure reached a minimum of approximately 1700 psi and was shortly restored to normal.
Following a plant shutdown, a program was initiated to investigate and evaluate the failure of HV-851A and HV-851B to open. (Reference 1).
The causes of the failure identified by this program are described in section 4.
3
- 3. SIS CONSIGURATION AND OPERATION PRIOR TO LER 81-020 At SONGS 1, the feedwater pumps perform a dual function.
During normal plant operations they serve as feedwater pumps, and following a Safety Injection actuation signal, they serve as high pressure safety injection pumps.
System Configuration The configuration of the Safety Injection System during the September 3, 1981 event reported in LER 81-020 (Reference 1) is shown in Figure 1.
System Operation During normal plant operation, prior to the incident, both feedwater pumps ran with suction from the condensate heater train through HV-854A and HV-854B and discharged to the feedwater heater train and steam generators through through HV-852A and HV 852B. During normal operation, the safety injection pumps were not running and were isolated by HV-853A and HV-853B. The safety injection lines downstream of the feedwater pumps were isolated by HV-851A and HV-851B and the reactor coolant Loops were isolated by MOV-850A, MOV-850B and MOV-850C.
Following a safety injection actuation with offsite power available (SIS), the Safety Injection System was designed to realign as follows (Figure 2):
- 1. Both feedwater pumps were to remain running while HV-852A, HV-852B, HV-854A and HV-854B were closing (within five seconds).
- 2. HV-853A and HV853B were to open within five seconds and MOV-850A, MOV-850B and MOV-850C were to open within 10 seconds and at the same time the safety injection pumps were to be started.
- 3.
HV-851A and HV-851B were to open following the closure of HV-854A & B.
An interlock prevented HV-851A and HV-851B from opening until HV-854A and B were fully closed.
With a loss of offsite power (SISLOP),
the SIS alignment was the same as discussed above with an additional eleven second delay for the diesel generators to come up to speed. In addition, both feedwater pumps were tripped and restarted following diesel generator startup. The timing discussed above is illustrated in Figure 2.
386 TO HOLD TANK CONDESATEHV 854A Hy H
SPRING LOADED PUPS 853A 852A 1
CIIF(:K VALVE SE AT 700 PS [G FEEDWATER PUMP A OV 50C L) ov asoa SAFETY INJECTION PUMPS 6
OLD LEG v
7850A 0--
LOOP A 6-6' COLD LEG 8538 8510
.*FiR'I[)WATE'IR PUMP R 114' HV 8528 CONDENSATE H
PUMPS 8548 FIGURE 1
PRE-SEPTEMBER 1981 SIS CONFIGURATION SONGS 1
FIGURE 2 PRE-SEPTEMBER 1981 SIS DELAY TIME SONGS 1 SISLOP 0
5 10 15 20 25 30 II I
IIII DGs START IHV-854 A & B C OSE IHV 851 A & B OPEN HV 852 A & B COSE HV 853 A & B OPEN 1MOV 850 A, B, & C OPEN IFWPs START
5 10 15 20 25 30 II I
I III I DGs START I STANDBY O854 A & B COSE HV 851 A & B PEN
&HV 852 A & BC OSE 1HV 853 A & B OPEN MOV 850 A, B, & C OPEN NOTE IPs START SIS LOGIC DOES NOT STOP THE FEEDWATER PUMPS
- 4.
CAUSES bF HV-851A AND HV-851B OPERATIONAL FAILURE After HV-851A and HV-851B failed to open upon receipt of a safety injection actuation signal, testing was performed in an attempt to determine the cause of their failure.
The tests consisted of starting the Safety Injection pumps, opening the feedwater pump suction valves, starting the feedwater pumps, and then attempting to open the safety injection discharge valves. Neither valve HV-851A or B opened. The feedwater pumps were then tripped and the valves opened during the decay in pump discharge pressures. Numerous retests were performed to determine the effect of differential pressure and valve packing adjustment on valve operability. It was concluded based on these tests that packing adjustment did not affect the operability and that valve internal inspection would be required.
The valves were opened and the seating surfaces inspected.
Maintenance was performed on the seats and testing continued.
Although improved valve operation was noted, consistent reliable operation within design limits could not be demonstrated with the feed pumps running.
In conjunction with the above tests, Kalsi Engineering Inc. was requested to review the design and operation of these valves to determine the failure mode. Based on this review (Reference 2),
the failure of HV851A and HV-851B to open on demand was attributed to the following three causes:
Excessive average contact stress between the valve discs and seats for a sliding contact resulting in galling during operation.
- Galling of the valve seats occurred during valve cycling because of contact stresses of approximately 32,000 psi on the valve seat face due to a high differential pressure (i.e., 1,500 psi across the valve disc during opening). An average contact stress of 10,000 psi, equivalent to 463 psi across the valve discs of HV-851A
& B, is appropriate for gate valves using stellite overlaid seats and gate faces to avoid galling. Since galling increases the coefficient of friction significantly, a much higher actuating force than provided by the existing actuators would have been required. Galling is technically defined as material transfer from one surface to another due to adhesive wear. Once galling is initiated and surface movement is repeated under high loads the process continues until seizing by welding occurs.
7
- 4. Operatioial Failure (continued)
Trapping a higher pressure in the body cavity than either the upstream or downstream pressures causing "double disc" drag.
- Subsequent inspection of the valve disc internals indicated that both valve discs were generating friction forces during opening. The valve actuators were sized to overcome the friction force of only one disc. This double disc drag phenomenon was attributed to pressure trapped between the two halves of the valve disc.
Insufficient margin used to size the valve actuators considering that the effect of "long-term" set is to increase the static coefficient of friction.
- The coefficient of friction used to size the valve actuators did not include a margin for "long-term" set effects. The coefficient of friction for stellite to stellite surfaces ranges from 0.119 under low loads operating conditions to a maximum value measured after long-term set of 0.4.
Under clean water, immersion tests performed in the laboratory, stellite surfaces have shown coefficient of friction values ranging from 0.15 to 0.25 for undanaged seat faces and seat contact stresses below 10,000 psi. The coefficient used to size these actuators was 0.2.
- 5.
SIS MODIFICATIONS TO PREVENT A REOCCURANCE The SIS modifications described below (References 3 & 4) were implemented to assure that HV-851A and HV-851B have sufficient actuator force to open under operating conditions and to maintain the average contact stress at the valve seat faces below the galling threshold of 10,000 psi. Given the physical dimensions of these valves, an average contact stress of 10,000 psi is equivalent to a pressure of 463 psi acting on the valve discs.
8
- 5. SIS M061FICATIONS TO PREVENT A REOCCURANCE (continued)
Physical changes to the SIS are shown in Figure 3 while changes to the equipment operating sequence of the modified SIS are shown in Figure 4.
- Resequence of Safety Injection Components For the postulated event of SIS actuation signal with offsite power available (SIS), two operating sequence changes were implemented:
a)
The feedwater pumps are tripped at t=o and restarted in 11 seconds; b)
Opening of MOV-850A, B & C is delayed by 11 seconds (Figure 4).
Tripping the feedwater pumps reduces the pressure acting on the upstream disc of HV-851A & B to less than 350 psi, thus reducing by a large factor the force required to ensure valve opening; and it also eliminate the potential for galling.
The stroke times of HV-854A & B and HV-852A & B were adjusted to ensure a sufficient time for the feedwater pumps to rundown to a pressure below 350 psi and to allow the pressure downstream of the check valves to decay. The delayed opening of MOV-850A, B, & C by 11 seconds mitigates the effects of back pressure on HV 851A & B which potentially could exist due to a leak in the reactor coolant boundry check valves downstream from MOV-850A, B & C.
For the postulated event of SIS actuation with loss of offsite power (SISLOP), the feedwater pumps and MOV-850A, B & C are sequenced in the same manner described above. However, HV-852A & B, HV-853A & B and HV-854A & B now receive a signal to actuate at t=o (Figure 4).
The interlock between HV-854A & B and HV-851A & B remains such that the latter will not open until HV-854A & B are fully closed.
9
SIS-386 CONDTONSATD HAN NIWEATE:NV Mr VV SPRING IADED FWS 53A 2A CHECK VALVE S PUMP 43A 8551AAT 7oo PsIc 141)
N --
6 FLEOWAT POP le SAFETY INJECTION PUNPS 2
V 07 95A CD3 3
-00%
OOA N -
HV HV6 0OLD :.gG 4
FEEDWATER PUMP 8 16 -0%--4 FIGURE 3_
MODIFIED SIS CONFIGURATION CONDENSATE V
5BSOG 3 SV 3
7 390 152 0
Add Feedwater pump triptime delay restart Additional body cavity vent lines and manual Add 11 second time delay for SIS isolation valves Delete 11 second time delay for SISLOP Change the delayed trip to instant trip for SIS
( )(Same for heater drain pumps-not shown) tAdditional body cavity vent lines and solenoid valves
() Add pressure releif groove across the sealing face of the disc and pressure releif hole through the disc
FIGURE 4 MODIFIED SIS DELAY TIME SONGS 1 SISLOP 0
5 10 15 20 25 30 1I IIIII DGs START 1HV 854 A & B CLOSE HV 851 A & B 0EN HV 852 A & 8 LOS HV 853 A & 8 OPEN I
I DELAY MOV 850 A, B, & C OPEN TIME DELAY FWPs START
- 0 5
10 15 20 25 30 I.
I I
I I
I I
DGs START ISTANDBY IHV 854 A & B CLOSE BHV 851 A & B 0EN HV 852 A & B LOSE HV 853 A & B OPEN TIME DELAY MOV 850 A, B, & C OPEN TIME DELAY FWrsSTART ISIPs STAJT
$4--FW Pump Trip
- sequencing delayed 1 sec. due to signal delay
- Valve Boy Cavity Vents on HV-851A & B and HV-853A & B L
In order to eliminate the potential for "double-disc" drag due to a higher pressure between the discs than the upstream and downstream piping pressure, equalization has been provided across one disc. The equalization is accomplished by providing a line between the valve body cavities and the upstream or downstream piping.
The equalizing line for HV-851A & B is sh6wn in Figure 3 Its installation was accomplished by drilling a hole in the side of the body of the valve and installing a piping connection to the upstream piping. A solenoid valve is provided in this equalizing line to maintain the containment isolation capability in conjunction with the upstream disc of HV-851A & B. The solenoid valve is normally closed and is opened on SIS or SISLOP to equalize pressure between the body cavity and the feedwater pump discharge pressure during coastdown. Following safety injection, HV-851A & B and the corresponding solenoid valves can close when required for containment isolation. The solenoid valves can be remote manually actuated from the control room.
The equalizing line for HV-853A & B is also shown in Figure 3 This line was installed by drilling a hole in the side of the body of the valve and installing a piping connection to the downstream piping. A normally open block valve is provided in the equalizing line for maintenance purposes during plant shutdown. This block valve is administratively controlled to assure its open position during plant operation.
- Feedwater Pump Discharge Check Valve Notch and Hole As can be seen in Figure 1, the pre-september 1981 SIS configuration had the potential for the pressure on the upstream disc of HV-851A & B to be higher than the feedwater pump discharge should the feedwater pump check valve be leak tight.
This condition could prevent the depressurization at the inlet of HV-851A & B following a feedwater pump trip.
To minimize the pressure differential across the upstream disc of HV-851A & B a groove was cut across the sealing face of the check valve flapper and a hole was drilled through the check valve flapper. This hole was sized to provide sufficient relief capacity assuming a failure of the check valve which isolates the line from the steam generator back pressure.
Under this condition there would be no pressure decay downstream of the feedwater pump check valves until HV-852A & B are fully closed. At this point, the volume between HV-851A & B and the check valves will depressurize to the decayed feedwater pump discharge pressure. As a result, with the aid of the equalizing lines, the pressure across the upstream disc of HV-851A & B draws to zero. Thus, the stress exercised on the stellite valve surfaces is minimal. The closing time of HV-852A & B was adjusted to ensure sufficient time to allow this decay prior to the opening of HV-851A & B.
12
- Administratlve Control of Pressure Downstream of HV-851A & B As can be seen in Figure 3 of this report, there is a normally isolated volume of liquid between HV-851A and B and MOV-850A, B and C. This volume could be pressurized to 700 psig (spring loaded check valve setpoint) from leakage of HV-851A and B or from MOV-850A, B and C. In order to ensure a maximum differential pressure of 350 psi across HV-851A and B, the pressure in this volume will be administratively controlled to a maximum of 350 psig. An existing alarm in the control room will be reset to this value and venting required if the alarm is activated.
- 6.
SIS INTERIM TESTING REQUIREMENTS To assure that the modifications made to the Safety Injection System would prevent a repeat of the HV-851A & B failure to open documented in Reference 1, SONGS Unit 1 committed to implementing a periodic surveillance testing program on the SIS.
This program was approved by the NRC on Novenber 5, 1981 (Reference 5) in the form of Technical Specification 4.2.3 (Reference 6).
Under this specification, Unit 1 is required to be placed in Mode 3 or 4 at least one every 92 days for a complete functional test of the SIS. This test includes the determination of the force required for the opening and the margin to available actuation force of HV-851A & B (Reference 7).
The criterion for the return of the unit to service is a measured actuator force of less than 10,000 pounds on both valves. This testing program is an interim surveillance program to be conducted during fuel cycle VIII which began in June 1981 and is expected to end in December 1985.
13
- 7. EVALUATION OF THE SIS INTERIM TESTING RESULTS A total of 6 tesI have been performed during fuel cycle VIII.
The measured actuator forces required to open HV-851A and B, the measured pressure differential acting on these valves and the calculated average contact stresses, obtained from these tests are shown in Table A.
These results indicate that in all cases the steam forces required to open HV-851A and B were below the 10,000 pounds test acceptance criteria (Reference
- 7) and that the galling threshold of 10,000 psi was never challenged.
The random values of the forces required to open the valves under equal test conditions suggested that they were representative of a normal distribution of forces from which a maximum opening force could be inferred with a high degree of confidence. Hence, a statistical analysis of the actuating forces shown in Table A was undertaken.
For the purpose of this analysis, the results of the first test, conducted on November 11, 1981, were not taken into consideration because it represented newly refurbished valves (i.e., valves, seats and discs were resurfaced and relapped shortly before this test) which were not in normal plant service for any significant period of time.
The stem force value measured during the second test of HV-851A conducted on February 27, 1982, was also discarded. The physical significance of this valve opening with only 120 pounds of force is doubtful since it suggests a rapidly decaying pressure differential across the valve.
This condition is not supported by the measurement of a pressure differential of 266 psi.
Hence, a malfunction of the instrumentation is assumed in this instance. This Datam is also proven to be an outlier by the r-and T-tests (Reference 8),
which are commonly used to detect outliers from a given random population.
With the test data representative of a normal distribution, it can be inferred with 95% confidence that the maximum opening force for valves HV-851A and HV-8518 will be 14,000 pounds or less.
(14,000 pounds is the 95/95 one-sided tolerance limit [Reference 9] from the four test valves of HV-851A.)
The 95/95 one-sided tolerance limit for HV-851B is 4,200 pounds less than the one for HV-851A.
It should be noted that a force of 14,000 pounds represents only 42.2% of the design opening thrust of the valve actuators. Therefore, it is concluded these valves will open on demand with a probability and confidence level much greater than 95/95.
14
TABLE A SIS HYDRAULICALLY OPERATED VALVES FUNCTIONAL TEST RESULTS HV-851A HV-851B OPENING AP CALC. AVE OPENING AP CALC. AVE.
FORCE CONTACT STRESS FORCE CONTACT STRESS TEST NO.
TEST DATE LBS PSI PSI LBS PSI PSI 1
11-23-81 2435 260 5623 6545 315 6812 2
02-27-82 120 266 5752 4520 284 6142 U,
3 11-13-84 4761 267 5774 5337 270 5839 4
02-09-85 7740 242 5233 3568 224 4844 5
05-10-85 6794 242 52,33 5658 242 5233 6
08-22-85 7224 254 5493 6023 260 5623
- 8.
VERIFqCATION OF THE ADEQUACY OF THE SIS MODIFICATIONS The three identified causes for the failure to open of HV-851A and HV-851B which occurred on September 3, 1981 (Reference 1) were:
(a) Excessive average contact stress between the valve discs and seats resulting in galling during valve operation with a design pressure differential of 1500 psi.
(b) Insufficient margin used to size the valve actuators considering that the effects of "long-term" set is to increase the static coefficient of friction.
(c) The possibility of trapping a higher pressure, in the valve body cavity, than either the upstream or downstream pressures causing "double-disc" drag.
To prevent recurrance of these causes, modifications were made to SIS's physical equipment and to the sequence of operation of this equipment during a postulated SIS and SISLOP events. Two parameters were critical to the achievement of this goal:
- 1) The force required to open HV-851A and HV-851B should not challenge the design thrust of the hydraulic actuators of 33,160 pounds.
- 2) The pressure differential acting on the valves while opening should not exceed a value of 350 psi.
The above two critical parameters have been adhered to during all six succesive tests. Figure 5 shows that the opening forces required by HV-851A have never exceeded 21.8% of the available actuator force. Figure 6 shows the parallel condition with HV 851B never exceeding 19.8% of the available actuator force. The average contact stresses acting on HV-851A were at most 42.2%
below the galling threshold (Figure 7), while HV-851B average contact stresses have been shown to be at most 34.5% below the galling threshold (Figure 8).
The results of the above mentioned test verify the adequacy of the SIS modifications to prevent the type of failure experienced by HV-851A and HV-851B during September 3, 1981. (Reference 1) 16
FIGURE 5
HV-851A FUNCTIONAL TEST RESULTS VALVE OPENING FORCES 41000 351000 300O00 25,000 C) 0 LA. 201000 0
1 5,000 L) 101000 5000 0.000 1
2 3
4 5
6 TEST NUMBER D
OPENING FORCE o
AVAIL. DESIGN FORCE
FIGURE 6 HV-851B FUNCTIONAL TEST RESULTS VALVE OPENING FORCES 40.000 353000 30g000 U) 25,000 La) 0 La. 20,000 coo 0
<15,000 10,000 51000 0.000 1
2 3
4 5
6 TEST NUMBER D
OPENING FORCE o
AVAIL.- DESIGN FORCE
FIGURE 7 HV-851A FUNCTIONAL TEST RESULTS AVERAGE CONTACT STRESSES 12,000 11,000 10,000 0 9,000 V) 8,000 0
7,000 S6,000 0a 5 1000 4.000 2,000 1,000 0.000 1
2 3
4 5
6 TEST NUMBER 0
CALC.CONTACT STRESS GALLING THRESHOLD
FIGURE 8
HV-851B FUNCTIONAL TEST RESULTS 0AVERAGE CONTACT STRESSES 121000 11,000 17,000 lip' U).
9,000
- 4),000 w
3,000 6,000 z
o 5,000 WJ 4,000 w
3,000 21000 1,000 0.000.
1 2
3 4
5 6
TEST NUMBER o
CALC.CONTACT STRESS A
GALLING THRESHOLD
- 9. PROPOSED LONG TERM SURVEILLANCE The proposed long term surveillance of the Safety Injection System to be implemented from the beginning of fuel cycle IX and as a replacement of the interim surveillance program carried out during fuel cycle VIII is:
A test that verifies only the actuation of HV-851A and HV-851B under no-flow conditions in a required time of 3 to 5 seconds. This test to be performed while the Unit is in MODE 5 and as a part of the currently effective operations procedure S01-12.8-2 "Cold SIS and Loss of Offsite Power Test."
- 10. JUSTIFICATION FOR THE PROPOSED LONG TERM SURVEILLANCE Results of six functional tests of the Safety Injection System indicate that the causes of failure to open of HV-851A and HV 851B have been adequately corrected.
This has been achieved by the physical modifications to these and other valves in this system and by the resequencing of the operation of the system under a postulated SIS and SISLOP events. Statistical analysis of the same data gives us a 95/95 probability/confidence level that the opening forces required by either of these valves will not exceed their design actuator thrust because the valves operate under low pressure differentials. The test results over the past four years firmly conclude that the design modifications, implemented in October 1981, have eliminated the experienced problems. Therefore, it is unnecessary to place the unit in MODE 3 to functionally test the SIS and verify valve openings. The proposed long term surveillance will assure adequate monitoring of the operability of this valves prior to Unit operation. Additionally, elimination of unnecessary test (Mode 3) with a very high frequency, as the one performed to meet Technical Specification 4.2.3, will avoid wear or transient induced failures of plant equipment and will also avoid frequent loss of power generation.
21