ML20087B315

Comment on Proposed Generic Communication Re Testing of safety-related Logic Ciruits.Five Areas of Concern,Including Tracking of Contact testing,non-technical Specs Required Sys & Define Sys to Be Reviewed,Listed for Review

ML20087B315
Person / Time
Site:	Oyster Creek
Issue date:	06/28/1995
From:	Rigsbee C AFFILIATION NOT ASSIGNED
To:	NRC
References
FRN-60FR27141 60FR27141-00003, 60FR27141-3, NUDOCS 9508080005
Download: ML20087B315 (3)

Text

'

N ~"3 bOFR 2 ' 7 l4l b M *

• 3' mon 1W 5~ -95 Chief bm""""'1.

~"

June 28 1995 Rules Review and Directives Branch U. S. Nuclear Regulatory Commission Washington, D.C. 20555

Subject:

Proposed Generic Communication Regarding Testing of Safety-Related Logic Circuits

Dear Sir:

This letter is in response to the recent notice of opportunity for public comment concerning the proposed generic letter to address testing of safety-related logic circuits.

I was a nuclear electrician, licensed operator, and licensed Senior Reactor Operator at Oyster Creek nuclear station. My backgrcund includes 6 years as a US Navy nuclear trained IC electrician. My experience is primarily electrical in nature, though as my SRO license history indicates I also have a solid understanding of the remaining fundamentals of nuclear power generating stations. For the last four years I have been evaluating the adequacy of safety-related logic circuit testing.

I have five areas of concern which are not specifically discussed in the draft generic letter available for review.

1) Tracking of Contact Testing:

Of primary concern is that the evaluations which are to be performed in response to the proposed generic letter are merely snapshotc in time. They may not accurately reflect prior or future testing compliance. What means are to be established by a licensee to ensure that once testing compliance is achieved, it is not nullified by revision to the associated procedure (s) or logic circuits? In my 20+

years experience in the nuclear industry, I have noted numerous occasions where surveillances are revised without a comprehensive understanding of the overlap coverage afforded by each procedure associated with a system. This is most )

likely to occur following system turnover within the engineering division. The result is that a once impeccable testing program for a system becomes flawed, sometimes  !

seriously. With no method in place to track where each l device and contact in a circuit is tested, what is to  !

prevent a recurrence in future years of the same testing ,

concerns being flagged as problematic today? (If I am not l mistaken, Fermi performed a complete evaluation of the testing on their safety-related control circuits prior to plant startup as a condition of their license. Yet it was Fermi who was noted in the draft generic letter as having identified so many testing de'iciencies. Is it possible that they were once in full compliance and over the ensuing years revised themselves into non-compliance?)

I propose requiring a method of tracking individual contact testing to ensure continued compliance and ready review of testing compliance.

l 9508080005 950628 PDR ADOCK 05000219 1 of 3 P PDR (0 L

s<>

r

)

2) Non-Technical Specifications Required Systems-The NRC has required licensees to install systems which l were determined to be important enough to exercise the '

l backfit rule, such as ATWS/ARI and Remote Shutdown and permitted the licensees to exclude the systems from plant technical specifications. These systems should be included in the generic communication as specifically requiring testing and as requiring a comprehensive evaluation of the  ;

testing performed. The NRC issued information notice IN 92- j 06, " Reliability of ATWS Mitigation System and Other NRC Required Equipment not Controlled by Plant Technical Specifications" to address the failure to adequately test  !

systems which were not included in a plant's technical specifications. The same systems should be included in the proposed generic letter.

3) Define the Systems to be Reviewed:

The proposed generic letter is too vague; it permits much latitude for interpretation. What does ESF systems mean? Are they the systems listed in the Safety Analysis ,

1 Report (SAR), Final Safety Analysis Report (FSAR), or Updated Final Safety Analysis Report (UFSAR). Does it include Control Room Ventilation (CRHVAC) system? Primary Containment Isolation (PCI) system? Standby Liquid Control (SLC)?

By making the statement ESF systems, each licensee is allowed to decide which systems meet the " intent" of the letter. As an example, some BWR stations have a Reactor Core Isolation Cooling (RCIC) system which has Logic System Functional Test (LSFT) requirements in the Technical Specifications but which is not included in their accident analysis and not listed as an ESF system. Such stations may be inclined to negate the importance of full system testing for the RCIC system since it is not required for the accident analysis even though there is a Technical Specification LSFT requirement and RCIC is included as part of the Remote Shutdown system. Furthermore, some stations do not specifically differentiate the PCI system from the {

Reactor Protection System (RPS). However, it could be inferred that the RPS portion of logic is only required to be reviewed and as such, the PCI portion of logic would not be evaluated since the PCI system was not specified; even though all the relays are powered from the same sources and are located in the same cabinets.

Splitting hairs in this regard to limit the cost and thus the scope of any action taken will occur without specific direction in the generic letter. Either a system is to be reviewed or it is not. I propose that the generic letter state specifically which systems are to be reviewed either in a listing of systems or in a manner which leaves no room for interpretation. This should be done even if it means that all facilities which are affected by the generic letter will have to be evaluated and a list developed for each. It is better to remove all doubt than to permit individual interpretation.

2 of 3 l

_g

,a>. .

r"

4) Define the Boundaries of the Systems to be Reviewed:

It is important to define the boundaries of a system to ensure that the full logic of a system is evaluated. An example is the Containment Monitoring (CM) system. Quite often the CM system is initiated by a LOCA signal generated by the Primary Containment Isolation (PCI) system. In this case, is it a function of the PCI system to initiate the CM system? Or is it a function of the CM system to initiate on a LOCA signal? If interpreted as the latter, the automatic initiation of the CM system might not be checked to occur because of an arbitrary delineation of system boundaries.

Permitting a licensee to determine the extent of a system paves the way for limiting cost at the expense of scope.

I propose that a definition of system boundaries be included in the generic letter specifying that all devices und contacts associated with a system are to be included in the review.

5) What Methodology is to be Used?

The proposed generic communication does not indicate just exactly what is required. Does a licensee have to cycle a relay to demonstrate testing? Or does each contact have to be tested in both the open and closed states to satisfy the testing requirements? Is the verification to be through circuit action? Or can it be through meter checks at the device under test? Is a visual confirmation of contact operation sufficient? Or does circuit continuity need to be verified?

I am a proponent of confirming both the open and closed contact states through circuit action for two reasons.

First, it ensures that the contact is in fact wired into the circuit. A meter on a set of relay contacts will indeed confirm contact operation, but by itself- it will not verify that the contact is in fact wired into the logic circuit.

Only observance of circuit action will demonstrate proper contact state change.

Second, through observing circuit action, any unintended circuit interactions will be detected. For example, an interlocking feature which erroneously blocks breaker closure with a switch in a specific position would be detected during testing with circuit action but possibly not with a meter across the contacts.

Thank you for your consideration of these matters. If I can provide any clarification, or if you wish to contact me, I can be reached at 708-769-1110 x125 during normal working hours (Central time) and/or at the below address.

Si,ncerdy

[

f /,(fu b, c / $

rles A. R ee P.O. Box 3419 Lisle, IL 60532-8419 3 of 3