ML20009D566
| ML20009D566 | |
| Person / Time | |
|---|---|
| Site: | Beaver Valley |
| Issue date: | 07/13/1981 |
| From: | Woolever E DUQUESNE LIGHT CO. |
| To: | Grier B NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| References | |
| 10CFR-050.55E, 10CFR-50.55E, 2DLC-4110, NUDOCS 8107240177 | |
| Download: ML20009D566 (6) | |
Text
~
f 9
g (F
$ km 9#
(412) 456-6000 43S Sixth Avenue Pittsburgh, Pennsylvania 2DLC-4110 15219 I(A July 13, 1981 o
gg U
ed States Nuclear Regulatory Commission S
o,
- Le 631 Park Avenue D;
";?
King of Prussia, PA 19406 M
S~,,.. "* '
ATTENTION:
Mr. Boyce H. Grier, Director q_
SUBJECT:
BEAVER VALLEY POWER STATION - UNIT NO. 2 Docket No. 50-412 3" MOV Gate Valves in Safety Injection and Chemical and Volume Control System - Significant De ficiency 80-06 Gentlemen:
We reported to your office on Oc tober 29, 1980, a significant deficiency of problems with three-inch gate valves manu fac tured by the Westinghouse Electro-Mechanical Division. We provided you an interim report on this problem on November 20, 1980.
An investigation into this valve deficiency is still underway.
Pursuant to the requirements of 10CFR50.55(e), enclosed is " Interim Report No.
2 on 3" MOV Gate Valves in Snfety Injection and Chemical and Volume Control System at Beaver Valley Power Station - Unit No. 2".
We plan to issue another report when valve modification details are complete.
DUQUESNE LIGHT COMPANY By
.h N Gj E.[/.Woolever
[
Vice President wjs Enclosure cc:
Mr. V. Stello (15)
Mr. K. Jabbour N
.s
/ /
0107240177 810713 PDR ADOCK 05000412 S
a Interim Report Westinghouse 3" ad 4" MOV Gate Valves 1.0
SUMMARY
The Westinghouse Electro-Mechanical Division Gate Valve problem surfaced when several valves failed to fully close during start-up testing at the Almarez Nuclear Station in Spain and during EPRI PORV Block Valve Tests at Duke Power's Marshall Station.
These valves are classified as " active" valves (i.e.,
mechanical operation is required to accomplish a sa fety function).
Beaver Valley Power Station Unit No. 2 has several of these valves located in the Chemical Volume Control, and Safety Injection Systems.
The valve types are Westinghouse Model Nos. 3GM88 and 4GM88. A complete list of the valve's function and location No.is provided in Attachment No. 1.
2.0 IMMEDIATE ACTION TAKEN In November, 1980, nonconformance and disposition reports 1456 and 1457 were issued to identify and control all applicable 3"
Westinghouse supplied motor-operated valves.
Nonconformance and disposition report 5025 was issued in June, 1980, to identify and control all applicable 4" Westinghouse supplied motor-operated valves.
As discussed in this report, the valve deficiency exists only for those valves which do not meet their functional requirements. These valves are the charging line isolation valves location Nos. 8107 and 8146, the charging pump miniflow valve location No. 8106, and the PROV Block Valves Location Nos. 8000A, B, and C As part of the DLC " Approach to Safety-Grade Cold Shutdown", PORV Block Valves Location Nos. 8000A, B,
and C will be upgraded to safety grade (active) and qualified to IEEP 323-74 and IEEE 344-75.
In addition, the new valve design will address the funtional concerns summarized in this report.
Nonconformance and disposition reports 1456, 1457, and 5025 will be closed. A new nonconformanc.
Td disposition report will be written to identify and control only t' e 41ves which failed to meet their functional requirements. The following v. Ives will be put on hold until modification is complete.
Valve WEMD Valve Location Model Function Number Re_ference Charging Line 8107 3GP38 Iso.
Charging Line 8146 3GM88 Iso.
r..
Charging Pump 8106 3GM88 Miniflow PORV Block 8000A,B,6C 3GM88 Valves
3.0 DESCRIPTION
OF DEFICIENCY Typically, these motor-operated valves operated through 75 percent of their full disc travel leaving about 5 percent of the flow passage un-sealed.
Subsequent strain gage testing showed the stem thrust loads required to fully close the valve to be 50 percent larger than original design calculations predicted.
In these original design calculations, the stem thrust required to fully seat the wedge or disc against flow was calculated using what were considered valve industry standard equations which are part of the actuator manufacturer's sizing criteria. This calculated stem thrust was then used to size the Limitorque motor operator that powers the valve.
Since the strain gage testing showed the actual required thrust load to be 50 percent higher than the calculated load which was used to size the operators, the operators do not have the capacity to fully close the valve -
particularly under their difficult 80 percent voltage requirement.
4.0 ANALYSIS OF SAFETY IMPLICATIONS The safety implications are listed in the last column of Attachment No. 1 titled " Potential Safety Consequences".
The key to interpreting this column is as follows:
Potential Consequences 1.
(PORV Block Valves) Potential incomplete isolation of pressurizer PORV.
2.
Potential cavitation of a centrifugal charging pump or safety injection pump due to operation beyond maximum runout flow.
3.
Potential inability to perform post accident containment isolation.
4.
Potential degradation of safety injection flow below values given in SAR.
5.
Potential inability to isolate RCS pressure boundary (no valves applicable to this concern).
5.0 CORRECTIVE ACTION TO REMEDY DEFICIENCY All valves listed in Attachment No. 1 do not mee; the equipment specifi-cation for valve closure at the maximum differential pressure as flow approaches zero at 2750 psi. To eliminate the number of valves requiring modification, a final valve closure functional requirement is defined for
7- -
each ' valve.
This functional requirement is based on system design requirements, single failure criteria, and the Westinghouse Reference
, Emergency Operating Instructions.
The emergency Operating Instructions, in certain applications, require
- stopping the high pressure safety injection pumps during the switchover to hot leg recirculation. The pumps are not restarted until all of the hot leg recirculation valve realignments have been accomplished.
For this condition, valve closure differential pressure will be zero.
The
' differential pressure of 500 psi is based on the worst case flow mismatch for. parallel high pre sure' safety injection pump operation during the transfer to cold leg recirculation phase of ECCS.
For those valves identified on Attachment No. I for which a new functional requirement could not ' eliminate modifications, WNES has not yet completed the testing and redesign program to choose the best modification for each
' valve type.
To determine the source of the higher-than-expected valve closing loads, a series of analyses and tests were initiated.
The testing programs as listed on Attachment No. 2 include: full flow testing, mechanical fixture testing, seat friction factor testing, and mctor operator testing.
Kinematic, force, flow,~ deflection, and stress analyses have been per-formed.. Although this work is not totally complete, it indicates that higher-than-expected seat friction forces are the source of the majority of the excess load. Backseat friction reactions and stem-disc connection reactions also contribute.
Reducing the actual required closing loads is not possible so attention
~
must focus on increasing - the force capabilities of the valv; actuators.
Depending on - the valve being considered, the force capability can be increased by some combination of gearing changes, motor size changes, and.
operator frame size changes.
On the W-EMD Model No. 3GM88 valve, the modifications will include:
change operator gear ratio to guarantee adequate ' thrust capability at 80 percent voltage, and rewire the operator for limit closing control.
6.0. ADDITIONAL REPORTS A final report will be issued when valve modification details are complete.
r-
l l-l ATTACIDGENT NO. 1 MAXIMUM AP (PSID)-
.AS FLOW APPROACHES ZERG VALVE CAPABILITY VALVE WEMD INTERIM FINAL
.AP (PSID) BELOW WHICH SATISFIES FUNCTION' POTENTIAL
'FUNCT.
VALVE WILL CLOSE REQUIREMENT -
SAFETY VALVE
, LOCATION MODEL EQUIP.
'FUNCT.
i FUNCTION NUMBER
. REFERENCE SPEC.
REQMT.
REQMT.
(AS SHIPPED)
(AS SHIPPED)
' CONSEQUENCES 1
t diarging Line 8107 3CM88 2750 2700
- 2750 1500 No -
2,4 Iso.
[-
Charging Line 8146 3CM88 2750 2700 2750 1500
' No 2,4 l
Iso.
Charging Pump 8106 3CM88 2750 2700 2750 1500' No 2,4 Miniflow l
Boron Injection 8801A 3m88 2750 1200 0
1500 Yes 2,3 Tank Iso.
l l
Boron Injection 8801B 3CM88 2750 1200 0
1500 Yes 2,3 Tank Iso.
l Boron Injection 8803A 3m88 2750 1200 0
1500 Yes 2,3 Tank Iso.
Boron Injection 8803B 3CH88 2750 1200 0
1500 Yes 2,3 Tank Iso.
Hot Leg Recire.
8814 3CM88 2750 1200 0
1500 Yes 2,3 Iso.
Hot Leg Recire.
8816 3CM88 2750 1200 0
1500 Yes 2,3 Iso.
Cold Leg Recire.
8885 3CM88 2750 1200 0
1500 Yes 2,3 Is o.
Charging Pump 8132A,B 4CM88 2750 500 500 750 Yes 2,4 Discharge XO Iso.
Charging Pump 8133A,B 4GM88 2750 500 500 750 Yes 2,4 Discharge XO Iso.
PORV Block 8000A,B,C 3CM88 2750 2500 1500 No 1
Valve (1)
(1)1hese valves are presently " inactive" valves; however, these valves will be upgraded to " active"
l
- [,..*,
.s..,
ATTACHMENT NO. 2
SUMMARY
OF TESTING ACTIVITIES--VALVE CLOSING PROBLEM 1.
Water Flow Testing-(at Pacific Pump Company)
Condition up to and including 600 gpm and 2800 psi 4 valves - 3GM88
+ field modifications complete
'3' valves - 3GM99 complete 3 valves - 4GM88 complete Field. modifications for the 3GM99 and 4GM88 to be tested in early-May.
2.
'EPRI' Steam Flow Tests (PORV Block, Steam Flow)
~
1 valve
.3GM88 complete 1 valve - 3GM99 complete-
'3.
. Site Testing (Full Flow Water Testing, Safety Injection System)
Almarez: 'Approximately.10 valves - 3GM88 complete CGE:
Approximately 10 valves - 3GM88 complete STP:
3 valves - 3GM88 complete 4.
EdF (Steam and Water Flow Testing) 2 valves - 3GM99 in process 5.
Mechanical Fixture Testing (At WEMD)
Using hydraulic cylinder to duplicate flow loads so force transfer can be l studied in depth.
I valve - 3GM88 complete 1 valve - 3GM99
-complete 1 valve -- 4GM88 complete 6.
Seat Friction Factor Tests (At Westinghouse R&D)
To determine stellite on Stellite friction factors under water and steam conditions. Test samples are nearing completion at EMD.
u a