PY-CEI-NRR-2387, Documents Telcon Held Between NRC Staff & Members of Perry Nuclear Power Plant Staff on 990406,re Safety Evaluation for License Amend 105

Documents Telcon Held Between NRC Staff & Members of Perry Nuclear Power Plant Staff on 990406,re Safety Evaluation for License Amend 105

ML20205P835
Person / Time
Site:	Perry
Issue date:	04/14/1999
From:	Myers L CENTERIOR ENERGY
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
PY-CEI-NRR-2387, NUDOCS 9904200398
Download: ML20205P835 (4)

Text

3-p___y Perry h!uclear Power Plant NN P y, o4 m

Lew W Myers . 440-280-5915 Vce President Fax:440-200-8029 April 14,1999 PY-CEl/NRR-2387L United States Nuclear Regulatory Commission ,

Document Control Desk l Washington, DC 20555 Perry Nuclear Power Plant Docket No. 50-440 Comments on the Nuclear Regulatory Commission (NRC) Safety Evaluation for Amendment 105 to Facility Operating License No. NPF-58 Ladies and Gentlemen: l This letter documents a telephone call held between the NRC staff and members of the Perry Nuclear Power Plant (PNPP) staff on April 6,1999, regarding the Safety Evaluation for Operating License Amendment 105 dated March 26,1999. The amendment revised the i design and licensing basis of the containment isolation valves in the feedwater system. I The discussion focused on several statements made on page six of the referenced Safety /

Evaluation. The NRC staff stated that 200 gallons per minute (gpm) is an acceptable leak rate for the feedwater check valves to demonstrate the check valves have been ' exercised

[

closed." However, another sentence on the same page stated that "for the upcoming refueling outage, the acceptable leak rate for which no check valve inspection or refurbishments is needed will be less than 19 gpm for each check valve."

It is clear from the NRC Safety Evaluation that 200 gpm is the acceptance criterion for the /

feedwater check valves. For the current refueling outage, however, there is an implication that valve inspection or refurbishment is the only option if leakage is determined to be greater than 19 gpm. This is not correct. Valve disassembly or refurbishment will only be pursued if other methods are unsuccessful in ensuring that the leak rate test results meet the 200 gpm criterion.

Each of the feedwater check valves has been tested satisfactorily, with individual valve leak rates less than the 200 gpm acceptance criterion. An alternate testing methodology was used during the performanw of the leak rate tests, and was discussed during the April 6 telephone call. The alternate testing methodology is described in Attachment 1. Following the telephone discussion, the NRC staff requested clarification on several aspects of the attemate testing methodology. This clarification is also provided in Attachment 1.

9 % 4200398 990414 I PDR ADOCK 050004400 P PDR

r, PY-CEl/NRR-2387L April 14,1999 Page 2 of 2 l . .

l This letter only provides clarification to information previously presented to the NRC. There are no new commitments, and no previous commitments have been changed.

If you have questions, please contact Mr. Henry L. Hegrat, Manager - Regulatory Affairs, at (440) 280-5606.

i Very truly yours,  ;

1 WW l

JAD i

cc: NRC Project Manager NRC Resident inspector NRC Region Ill l

l I

l l

l l

l

f.

L Attachment 1 l: PY-CEl/NRR-2387L i- Page 1 of 2 i.

, I

NRC QUESTIONS AND RESPONSES l Question 1

The staff was told 19 gpm was the maximum flow through the existing piping penetrations, whereas the measured values ended up 20-22 gpm. Explain the difference.

Response: The maximum test flow through the feedwater test appendages was based upon a hydraulic analysis. The purpose of the analysis was to calculate the maximum flo+

at test pressure, that could pass through the test appendages by assuming the only flow

rMriction was due to the test appendage. If the measured flow was less than this calculated value, then it could be concluded that the measured flow rate was attributed only to the flow restriction created by the closed check valve. The analysis shows that for a test leak rate below 19 gpm, the leakage would be the actual check valve leakagec For leak  !

rates above the maximum test rate of 19 gpm, the original testing methodology would not provide satisfactory assurance that all potential valve leakage was quantified, hence, an '

alternate testing methodology would be required.

j Question 2: The staff felt that an acceptance criteria of 17-18 gpm should have been l sufficient particularly in light that the highest measured leakage from previous testing was

' 14 gpm. Has the testing methodology changed?

i i Response: The testing methodology has changed. The previous tests were performed I with the feedwater check valves isolated from the reactor pressure vessel and the valves l were pressurized from a test apparatus to a pressure of approximately 9 psig. Tae current j testing was performed using the head of water in the reactor pressure vessel as the l pressure source. This pressure is approximately 29 psig. This difference in pressure is not enough to provide additional seating force to seat the disk, but is enough to potentially l ' contribute to a higher leakage rate.

i Question 3: Provide a description of the testing methodology.

l Response: For the purpose of this description, the upstream side of the check valve is the

! side toward the feedwater pumps and the downstream side is toward the reactor pressure l vessel. Hydraulic principles indicate that leakage across a check valve is proportional to the l square root of the differential pressure across the valve. As the differential pressure increases, the leakage increases. For this application, since the change in pressures are l small, the valve disk was assumed to not move (i.e., any gap between disk and valve seats l remained a constant). When water is drained from the upstream side of the check valve, the pressure in the feedwater pipe upstream of the check valve drops due to the check valve restricting flow; The pressure on the upstream side will drop until the differential pressure across the check valve (the driving force for the leakage) causes the leakage across the check valve to match the drain rate out of the test appendage. Once steady state is achieved, the flow past the check valve seat will match the flow out of the test appendage.

Knowing the observed flowrate out of the test appendage and the observed differential

- pressure across the check valve, the total leakage that would exist across the check valve

. could be determined by a standard hydraulic calculation.

5 a

~

Attachment 1 PY-CEl/NRR-2387L Page 2 of 2 The information obtained from the leak rate testing was analytically scaled down using hydraulic correlations to determine the leak rate that would exist at the check valve if it was subjected to the reference test differential pressure of 1.1Pa. Since no credit is taken for l the check valve seating tighter at the higher test differential pressure, the analytical scaling l of seat leakage is considered conservative.

Question 4: Explain why the approach should be viewed as a conservative measurement of valve leakage, l

i Response: The alternate testing methodology is a more accurate means of determining leak rates, 'The methodology employs standard engineering hydraulic analyses to evaluate the data. The analyses include factors for test measurement uncertainty, Precision l pressure gauges were used to obtain the differential pressure data. Administrative controls i were in place to ensure the integrity of the collected data. These actions coupled with conservative manipulation of the raw data provided a high level of assurance that the evaluation conclusions were accurate.

Question 5: Explain how measuring low leakage at low pressure provides assurance that significant leakage will not occur at high pressure.

Response: The scenario where the check valves would need to function to stop reverse flow would involve the loss of the feedwater piping upstream of the motor-operated feedwater block valves. With the reactor at pressure, the differential pressure across the check valves would be reactor pressure to turbine building pressure (atmospheric pressure).

Force is proportional to the pressure and the area upon which it is acting. Therefcre, at i higher closing pressures (i.e., higher differential pressure across valve) more force will be exerted in the direction to seat the valve disk. Higher disk seating force will tend to minimize any seat-to-disk gaps and minimize the area for seat leakage. However, higher upstream pressure would also tend to increase the leakage (for the same gap size).

Leakage is proportional to the square of the gap area, but only proportional to the square root of the differential pressure. Hence, for a specified increase in upstream pressure, more force will be exerted to reduce any seat-to-disk leakage gap than will be available to force water through the gap. l As pressure decreases, the force acting upon the valve disk also decreases. If seat-to-disk gaps exist, there will be less seating force to minimize the gaps, thereby potentially permitting more seat !eakage to occur. Industry studies support this conclusion. Therefore, 1 for wide contact seating surface valves at low pressures, the seat leak rate should be j representative of a maximum leak rate for the valve.  ;