ML20245J724

Discusses Merits of Close on Limit Re motor-operated Gate Valve Torque Switches.Limit Switch Failures Not Reported at Significant Rate & Switch Require Only Minimal Care

ML20245J724
Person / Time
Site:	Davis Besse
Issue date:	02/25/1989
From:	Wohld P AFFILIATION NOT ASSIGNED
To:	Carlyle Advisory Committee on Reactor Safeguards
References
ACRS-CT-1945, NUDOCS 8908180160
Download: ML20245J724 (2)

Text

. - -

i q .

(\;

. 0 7-19 %

February 25, 1989

Dear Carlyle:

Considering the recent, additional bad news about motor-operated gate valve torque switches, I remain increasingly steadfast in my '

pursuit of expounding the merits of "close on limit." In other words, I an a crank on this. Please bear with me.

During Jerry Weidenhaner's February 1 meeting, Bob Elfstrom nentioned setting the feedwater isolation valves at Davis-Besse to close on limit. He also stated that the closure was repeatable to "within 1%" (which was in reference to final thrust on coast into the seat). Unfortunately, the final torque, as you know, will be different under dynamic process conditions;- however, it is also different when closing on torque, as noted in the INEL tests.

The "the close on limit" limit switches at Davis-Besse were set to open at 0.025 inches from hard seating, and the distance is.very repeatable, probably within +/-0.005 inches. Hence, for this 14 inch valve closure is assured by the limit switch to within (0.025/14)X100 or 0.2 percent of a hard seat closure (+/-0.04 percent). This assures a full 360 degree seat interference. If fluid conditions are enough to stop travel short of achieving a wedging action in-the valve body, I would expect the downstream seat to at least act as a good check valve. (This would not work on a globe valve.)

Even with a "close on torque" situation, most of the operator thrust may go to moving the valve and not to providing a final wedging and machanical seat sealing load. (Also, the basic thrust equation does not provide for a final wedging load.)

My purpose in pursuing this is in search of hope where it often seens there is none-. I personally believe that the Limitorque torque switch has so many problems that it should normally be declared unsuitable in nuclear applications where it might interfere with'a critical function. General Electric appears to have come to this conclusion in recommending their NSSS customers in recent years to totally bypass the open torque switch and bypass the close torque switch most of the way for the more critical valve stroke functions.

The exciting features of the limit switches are that their switching points are normally very repeatable, very observable without exotic equipment, and are easily set and can be set very accurately. Repetitive mechanical stresses are reduced by closing i on limit and if a limit switch fails, it will generally be readily detected, unlike many torque switch failures that can exist for the life of a plant; yet, I do not believe limit switch failures are reported at a significant rate and they require only minimal care.

l Most importantly, testing under any set of process conditions will l demonstrate how the limit switches will perform under any other set ,

l of process, conditions. These features do not apply to the torque I switch, l p DESJCNATED ORIGINAL 8 Certified By C -

CT-1945 FDC

L

g. '

~

Now why am I such a crank on this? One, evidence continues to grow l

against the reliability of the torque switch as used in the motor-operated valve control circuit and as set and maintained within the philosophy for the " conduct of operations" at many plant sites; and, two, each of these plants is in significant jeopardy of an extended shutdown as the result. A simple investigation by a regional inspector of the next valve failure atand inspection any' plant due to a torque switch failure could result in a "show cause," 10 CFR 50.54f letter for that plant from which any utility may have a very difficult time responding without any meaningful standards in this area.

I believe that if a meaningful cost benefit analysis can be done, one should be done for bypassing the torque switch to the extent possible and then developing a policy based on the results (if- the study is believable). Basically, I would expect the results to support General Electric " standard design" as mentioned above.

One idea that seems to persist is that the torque switch provides some protection to the motor - a mechanical fuse so to speak. I believe people may think of the switch in these terms because it uses preprogrammed memory, that applies to electrical fuses, to add a perceived sense of understanding to an otherwise more complex device. In m nothing more.y When opinion, the torque it fails to openswitch is a control on closure, element the motor and goes, unless backed up by a thermal overload, and all the bad things that might happen at locked motor rotor mechanical stresses will occur.

During one of ny last site visits as a NRC inspector, the maintenance department indicated that three torque switch failures to open had occurred in one year due to mechanical failure in the torque switch drive mechanism. Hence, at this plant at least, it was switch'.

difficult to see the benefit to the motor from the torque I expect that an accurate evaluation will show more problems caused by the switch than benefits derived, solely on an economic basis.

Enough. expounding! The 2/28/89 meeting is soon and I will see you then. I consider this an informal note for your information, but feel free to share with anyone for "second opinions," etc. I can also desire.

submit the technical discussion in a formal memo if you Close on limit, P.R. Wohld

_ _ - -