ML20205T049
| ML20205T049 | |
| Person / Time | |
|---|---|
| Site: | Clinton  |
| Issue date: | 10/31/1988 |
| From: | Stevens J Office of Nuclear Reactor Regulation |
| To: | Holtzscher D ILLINOIS POWER CO. |
| References | |
| TAC-68288, NUDOCS 8811140090 | |
| Download: ML20205T049 (5) | |
Text
- t
[fpas:%je
+
UNITED sTATds g
g NUCLEAR REGULATORY COMMISSION 5
- l WASHING TON, D. C. 20655 e
t.
/
October 31, 1988 Docket No. 50-461 Mr. Dale L. Holtzscher Acting Manager - Licensing and Safety Clinton Power Station Post Office Box 678 Mail Code V920 Clinton, Illinois 61727
Dear Mr. Holtzscher:
SUBJECT:
LEAKAGE RATE TEST METHOD FOR CONTAINMENT ISOLATION CHECK VALVES 1E22F005 1E51F066, 1E21F006, AND 1E12F041C Al CLINTON POWER STATION TAC NO. 68288)
The staff has completed its review of your letter dated April 18, 1988 concerning the method to be used for local leakage rate testing in accordance with Appendix J to 10 CFR Part 50 of four containment iso'ation check valves at the Clinton Power Station. The four air-testable check valves and the systems in which they are located are identified as 1E22F005 (High-Pressure Core Spray,1E51F066 (Reactor Core Isolation Cooling),1E21F006 (Low-Pressure Core Spray, and 1E12F041C (Low-Pressure Core Injection "C"/ Residual Heat Removal "C").
The staff provided a discussion on the valves listed above in Inspection Report 50-461/88-20. The staff found that the subject valves, which are located inside containment in lines that had single containment isolation valves (CIVs) outside containment, should also be considered to be CIVs, thus providing two CIVs in each line, one inside containment and one outsid? containment.
It was also detennined that the subject valves must be locally leakage rate tested (Appendix J Type C tested) with air or nitrogen as the test medium at a test pressure of approximately 15 psig.
Your lette dated April 18, 1988, provided a calculation which attempted to correlate leakage rates measured with water at 1000 psig with leakage rates measured with air at 9.0 psig. The approach was theoretical in that there was no empirical information given concerning the correlation of air and water leakage rates for the valves in question. This apparently was the case because air tests have not been performed on the valves in question.
There were several problems with your proposal to use a water-to-air corre-lation and thereby ;>ennit continued testing of the subject valves with water as follows:
1)
Existing regulathns (Appendix J.Section III.C.2) require that the valves be tested with air or nitrogen, unless sealed witn fluid (e.g., water) from a seal system. Since you did not attempt to show that t'le 2
9h
!N"#88& gggg,;e, i
i n
's.
~*
October 31, 1988 Dale L. Holtzscher valves are water sealed, the regulation prohibits testing with water. A specific exemption from the regulation would therefore, be required to allow testing with water.
2)
The staff's long standing technical position has been to disallow conversion of water leakage rates to air leakage rates.
It is the staff's opinion that the complex relationship between the two types of leakage results in very Idrge uncertainty which makes development of a useful correlation impractical.
Therefore, a water-to-air correlation is technically unacceptable.
I 3)
Finally, notwithstanding the two arguments above, you did not consider the difference in test pressures being proposed. With a test pressure of 1000 psig, the large force pressing the valve disk against its seat will cause different deformations and disk-to-seat contacts than would be present at 10 or 15 psig. This would result in different leak flowpaths. The pressure difference would also result in different forces pushing the test fluid through the leak paths. Thus, your correlation is deficient because it does not consider the difference in test pressures.
Based on the preceding, the staff concluded that the subject check valves should be Type C tested with air or nitrogen at a test pressure of Pa (calculated peak containment internal pressure related to the DBA LOCA), in aCCordance with Appendix J to 10 CFR Part 50.
With regard to the implementation schedule, the stut wggested that Type C testing of the subject valves begin during the next Llidon refueling outage scheduled to begin in January 1989. You were informed on August 26, 1988 to be prepared to conduct these tests with air at that time.
For further information or clarification, please contact me at (301) 492-1397.
Janice A. Stevens, Project Manager Project Directorate III-2 Division of Reactor Projects - !!!,
IV, V, and Special Projects cc:
H. Miller, Rill D. Danielson, RIII D. Cooper, Rll!
J. Pulsipher, PSB/ DEST cc:
see next page CISTRIBUTION Docket file 4
MVirgilio DMuller NRC & Local PORs JStevens JCraig PDIII-2 r/f LLuther
- see previous Concurrence G
P l
- PDI!!-2
- PDi!!-2 0 ' y2 N
JSteven atj idr LLuthe/rd /ff 88
/0 /27 /88
}1
/0 /g /88
/
'o <
Dale L. Holtzscher valves are water sealed, the regulation prohibits testing with water. A specific exemption from the regulation would therefores be required to allow testing with water.
2)
The staff's long standing technical position has been to disallow conversion of water leakagu rates to air leakage rates.
It is the staff's opinion that the complex relationship between the two types of leakage results in very large uncertainty which makes development of a seful correlation impractical. Therefore, a water-to-air correlati is technically unacceptable.
3)
Finally, notwithstanding t h two arguments above, the censee did not consider the difference in test pressures being pr osed. With a test pressure of 1000 psig, the large force pressing e valve disk against its seat will cause different deformations and di
-to-seat contacts than would be present at 10 or 15 psig. This w d result in different leak flowpaths. The pressure difference woul also result in different forces l
pushir.g the test fluid through the lea paths. Thus, the licensee's correlation is deficient because it es not co.sloer the difference in test pressures.
Based on the preceding, the staff co cluded that the subject check valves should be Type C tested with air or nitrogen at a test pressure of Pa (calculated peak containment inte al pressure relat6d to the DBA LOCA), in accordance with Appendix J to 10 FR Part 50.
With regard to the implementat n schedule, the staff suggested that Type C testing of the subject valves egin during the next Clinton refueling outage scheduled to begin in January 1989. The licensee was infomed on August 26, 1988 to be prepared to condu t these tests with air at that time For further infomation or clarification, please contact me at (301) 492-1397.
Janice A. Stevens, Project Manager Project Directorate 111-2 Division of Reactor Projects - III, IV, V, and Special Projects cc:
H. Miller, RI I D. Danielson RI!!
D. Cooper, 111 J. Pulsiph r. PSB/ DEST DISTRIBUTION Docket file MVirgilio DMu11er NRC & Local PDRs JStevens JCraig PDIII-2 r/f LLuther 1
PDIll-2 PDill-2.y /
PDill-2 JStevens: nj Lluthert/
DMuller tol.tr/88 to /W/88
/
/88
Dale L. Holtzscher.
valves are water sealed, the regulation prohibits testing with water. A specific exemption from the regulation would therefore, be required to allow testing with water.
2)
The staff's long standing technical position has been to disallow conversion of water leakage rates to air leakage rates.
It is the staff's opinion that the complex relationship t'etween the two types of leakage results in very large uncertainty which makes development of a useful correlation impractical. Therefore, a water-to-air correlation is technically unacceptable.
3)
Finally, notwithstanding the two arguments above, you did not consider the difference in test pressures being proposed. With a test pressure of 1000 psig, the large force pressing the valve disk against its seat will cause different deformations and disk-to-seat contacts than would be present at 10 or 15 psig. This would result in different leak flowpaths.
The pressure difference would also result in different forces pushing the test fluid through the leak paths. Thus, your correlation is deficient because it does not consider the difference in test pressures.
Based on the preceding, the staff concluded that the subject check valves should be Type C tested with air or nitrogen at a test pressure of Pa (calculated peak containment internal pressure related to the DBA LOCA), in
~
accordance with Appendix J to 10 CFR Part 50.
With regard to the implementation schedule, the staff suggested that Type C testing of the subject valves begin during the next Clinton refueling outage scheduled to begin in January 1989. You were informed on August 26, 1988 to be prepared to conduct these tests with dir at that time.
For further information or clarification, pleaso contact me at (301) 492-1397.
- d. K Ja ice A. Stevens, Project Manager Project Directorate III-2 Division of Reactor Projects - III, IV, V, and Special Projects cc:
H. Miller, RI!!
D. Danielson, RIII D. Cooper, RIII J. Pulsipher, PSB/0EST
]
r-s Mr. Dale L. Holtzscher Clintor. Power Station Illinois Power Company Unit 1 CC:
Mr. L. Larson Mr. D. P. Hall Project Manager Vice President General Electric Company Clinton Power Station 175 Curtner Avenue, N/C 395 P. O. Box 678 San Jose, California 95125 Clinton, Illinois, 61727 Regional Administrator, Region III Mr. R. D. Freeman 799 Roosevelt Road, Bldg. #4 Manager-Nuclear Station Engineering Dept.
Glen Ellyn, Illinois 60137 Cli.iton Power Station P. O. Box 678 Chairwan of Dewitt County Clinton, Illinois 61727 c/o County Clerk's Office DeWitt County Courthouse Sheldon Zabel, Esquire Clinton, Illinois 61727 Schiff Hardin & Waite 7200 Sears Tower Illinois Department 233 Wacker Drive of Nuclear Safety Chicago, Illinois 60606 Division of Engineering 1035 Outer Park Drive, 5th Floor Resident Inspector Springfield, Illinois 62704 U. S. Nuclear Regulatory Commission RR 3 Box 229 A Mr. Donald Schopfer Clinton, Illinois 61727 Project Manager Sargent & Lundy Engineers 55 East Monroe Street Chicago, Illinois 60603
- - -