ML19351A572

Withdraws Commitment to Install Cavitating Venturies or Any Other Mod to Relieve Operator Burden in Ensuring That Emergency Feedwater Pumps Protected from Runout in Postulated Accidents W/Low Steam Generator Pressurization

ML19351A572
Person / Time
Site:	Oconee
Issue date:	12/07/1989
From:	Tucker H DUKE POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 8912190351
Download: ML19351A572 (4)

Text

,

.g _

c, .

~

<1, : _.

.. a DUKE POWER GOMPANY

'M; 2~ P.O. HOX 33180 CilARLOTric. N.C. 28242 4 HALB. TUCKER- trl.EPh0NE -

vms rasemswt . ' (704) 07&4b31 swam 4a emonwrion .

f.

December 7, 1989 U. S. Nuclear Regulatory Commission Document Control Desk Washington, DC 20555

Subject:

Oconee Nuclear Station Doc. Nos. 50-269, -270, -287 Installation of Cavitating Venturi j On October 17, 1989,' Duke Power personnel met with the NRC Staff to discuss pump runout protection for the Oconee Emergency Feedwater (EFW) pumps This meeting was requested by me-in a letter dated June 20, 1989. .,

The purpose of the meeting was to discuss, in depth, the. reasons for i revising our previcus commitment to implement a' hardware fix to reduce  !

operator burden during postulated accidents with low steam generator pressure. As discussed during the October 17, 1989, meeting we have been-evaluating various options to address this-concern, and to date have not i been able to identify an optimu? solution. Accordingly, we do not intend 1 to implement any haraware fix to address this concern. As such,-I am  !

withdrawing my commitment to install cavitating venturies or any other }

l modification in order to relieve operator burden in' ensuring that the EFW l pumps'are protected from runout in postulated accidents with low steam l generator pressure. The' basis for this change was discussed in depth.

during the October 17, 1989 meeting. 'We will continue to assure operator .

awarenoss of this potential problem, through training and procedures.  !

In addition during the October 17,.1989 meeting, the vibration problems with installing cavitating venturies into Oconee EFW system was discussed.

The NRC requested that Duke provide additional information regarding this 1 problem. To this end, please find attached the requested information. ,

(

H. . Tucker PFG/78/td Attachment 8912190351 891207

$DR ADOCK 0500 p g/

Ilt

( ::a- '

W

, L ,. j.1: ..

P Document Control Desk h  : December 7,:1989

• . Page 2' cc: Mr. L. A..Wiens

' Office of Nuclear Reactor Regulation U. S.- Nuclear Regulatory Commission p Washington, DC- 20555

.Mr. P. H.'. Skinner-r NRC Resident Inspector-

"N Oconee Nuclear Station Mr. S. D. Ebneter Regional Administrator, Region II U. S. Nuclear Regulatory Commission 101 Marietta Street,-NW, Suite 2900 4 Atlanta, GA 30323 i-l l

'I l

i

m

, , e-J OCONEE l' NITS 1. 2. and 3 Cavitating Venturi Vibration Problems l

l To solve the potential problem with EFW pump runout and enhance the operation of the EFW system, cavitating venturis were selected as a possible passive solution that might ef fectively reduce operator burden without complicating the system.

Following the design and manufacture of the venturis. the venturis were  !

installed in a temporary test loop and operated to verify functional  !

performance and measure vibration created s in downstream piping. The piping response under postulated worst case cavitating conditions resulted in vibration velocities in excess of 10 ips-pk (pegging - the meter).and accelerations in excess of 60 g's.

While it is generally accepted that the piping vibration is highly dependent upon the piping geometry and support configurations, it is also known from. failure experiences that cantilevered vents and drains, as well as valve actuators. can have vibration responses several titc.es higher i than the run piping for cavitation type input. The greater response of these components to the high level of " white noise" is attributed to the poor overall damping available in combination with excitation of natural frequencies from the bread frequency range forcing function.

Inasmuch as (1) the Duke piping in question near the proposed location of the cavitating venturi does contain cantilevered attachments and (2) we expect that, based on our experience, this piping would be highly susceptible to short term vibration fatigue, we cannot recommend installation of the cavitating venturi as a solution to our problem with EFW pump runout.

This opposition is further reinforced by our concerns that if we were to install the venturi, a lengthy data acquisition testing effort would be required to adequately evaluate the systems functional integrity under the conditions generated by the cavitating venturi. Following data collection, the unit would be considered inoperable for days or perhaps weeks while analytical ef forts were made to qualify the vibration and -

calculate the allowable remaining fatigue stress cycles not consumed by the testing. Efforts to limit stress cycles and allow a greater alternating stress intensity would quickly reduce the allowable run time. This could force tracking of the stress cycles during actual unit operation resulting in extensive permanent instrumentation. monitoring equipment. and evaluation expense.

rc '

3

. p-4 : e d

1 Also, the demands- of the Duke system configuration would have required

- placement of a venturi immediately upstream of a secured cabinet containing several valves and branch connections. Because of_the high vibration susceptibility of these components, extensive data collection and additional compensatory security measures in these areas would have been required during the attempted qualification testing. -

Furthermore, discussions with another utility having cavitating venturis installed in a simple piping configuration reinforced the belief that '

qualification would likely be much more dif ficult, if at all possible, for Duke's complicated piping geometry.

In summary, many factors pointed to the decision- not to attempt installation and qualification of the cavitating venturis. Among them were previous experiences with high vibration failures and comments from

other utilities concerning qualification dif ficulties for much simpler

' piping systems, as well as. vibration results from the test loop.

1 t

l i

I