ML20056E975

Provides Insp Results & Safety Analysis Re Operational Amplifiers Discussed in 920917 & 23 Ltrs as Potential Safety Issue.Concludes That There Are No Safety Concerns W/Respect to Buffer Amplier Module.Ltr Not 10CFR21 Notification

ML20056E975
Person / Time
Site:	Three Mile Island
Issue date:	08/19/1993
From:	Taylor J BABCOCK & WILCOX CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
JHT-93-204, NUDOCS 9308250362
Download: ML20056E975 (4)

Text

i l

~

BW BSW NUCLEAR TECHNOLOGIES JHTl93-204 3315 Old Forest Road August 19, 1993 P.O. Box 10935 L ynchburg. VA 24506-D935 Telephone: 804-385-2000 Telecopy' B04-385-3663 Document Control Desk U. S. Nuclear Regulatory Commission Washington, D. C. 20555

Subject:

Inspection Results and Safety Analysis Regarding Operational Amplifiers

References:

(1) Letter from J. H. Taylor (B&W Nuclear Service Company) to U.S. NRC, dated Sept. 17, 1992,

Subject:

Potential Safety Issue Regarding Operational Amplifiers.

(2) Letter from J. H. Taylor to U.S. NRC, dated Sept. 23, 1992, above subject.

Gentlemen:

In references (1) and (2), B&W Nuclear Service Company (BWNS) advised the NRC of a potential safety i~ssue involving operational amplifiers. As described in these references, BWNS notified the B&W plants of this issue, and GPU Nuclear Corporation (GPUN) subsequently found three suspect operational amplifiers in Signal Converter and Buffer Amplifiar Modules at the Three Mile Island-1 plant. This letter provides the results of an inspection of those modules by BWNS and an analysis of the safety implications. This letter is not a 10CFR21 notification.

Backcround The issue concerns modules manufactured by Bailey Controls Company that are used in the Reactor Protection System and Engineered Safeguard Actuation System at B&W-designed nuclear plants. Bailey had notified BWNS of the possibility that certain modules manufactured in 1984 and later may have an operational amplifier manufactured by the National Semiconductor Company that may not be capable of driving the full rated load under all operating conditions.

The modules of concern were supplied by BWNS to three B&W operating plants and to one, or possibly two; B&W plants in a deferred status of construction. Since it was unknown whether the modules-did, in fact, contain the National Semiconductor amplifier or whether the amplifiers were made by another manuf acturer for which there was no concern, BWNS advised the plants to inspect the suspect modules to 9308250362 930819 7 PDR ADOCK 05000287 [-l S PDR y O

Document Control Desk j JHT/93-204 August 19, 1993 Page 2 l determine whether they contained the National Semiconductor amplifier of concern.

! GPUN responded to BWNS's notification by advising that they had l

found three suspect modules (two Signal Converter Modules and one ,

Buffer Amplifier module) at the Three Mile Island-1 plant. The ;

modules were suspect because one operational amplifier in each module had no markings. Thus, there was uncertainty as to whether l the amplifiers were manufactured by National Semiconductor or by l

another manufacturer. GPUN advised that one of the Signal l Converter Modules was in stock and the other in the "C" Reactor Protection System High Temperature reactor trip string. The Buffer Amplifier Module was also in stock, with potential end use in the Engineered Safety Actuation System. GPUN also advised they l had replaced the suspect module that was installed in the RPS High Temperature String.

Status Update ,

GPUN shipped the three suspect modules to BWNS for inspection and repair, and BWNS inspected the modules. The inspection results are as follows.

i 1. Buffer Amplifier Module. The questionable operational ,

l amplifier in this module was determined to be made by' Burr  !

l Brown and is not the National Semiconductor amplifier of concern. It was also determined that the module itself was not of the type identified by Bailey Contro3s Company as l possibly containing the suspect amplifier. It is, therefore, concluded that there is no safety concern related to this module.

{ 2. Sianal Converter Modules. The questionable operational j amplifier in each of the two modules has no meaningful l markings, and.we are unable to make a determination as to I whether or not they are the National Semiconductor amplifiers of concern.

Safety Imolications Since there is no concern for the operability of the Buffer Amplifier, it is only necessary to consider the safety implications for the Signal Converter Module. The Signal Converter Module

, performs two functions. One function is to take the output of the j linear bridge-RTD combination and put it in the proper voltage range for use on the high temperature trip bi-stable. The second function is to apply a gain on the temperature signal for use on l

Y Do,cument Control Desk JHT/93-204 August 19, 1993 ,

Page 3 '

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ = - ____________

l

[

the variable low pressure trip. A safety assessment for the  !

Signal Converter Module requires that several pivotal assumptions j be made. If the assumptions are made that: (1) the operational l amplifiers in question are in fact the National Semiconductor unit  ;

of concern, and (2) that the modules would fail to produce the  ;

required output due to the presence of the suspect operational  !

amplifier, then the bi-stable modules which initiate these trip functions would be affected either by not tripping or by not tripping at the appropriate time.  ;

Hiah Tcmpcrature Trio  !

i One of the functions of the Signal Converter Module is for use in I the high RCS outlet temperature trip string. .This trip is_a backup trip for which credit is not taken in the plant accident analysis. ,

It is not needed to meet the acceptance criteria for the Chapter 14  !

events, and is, therefore, not a primary trip. The system  :

parameters necessary to reach the high RCS temperature trip  ;

setpoint result'in other trips actuating earlier. The high flux  :

trip and high RCS pressure trips are the primary trips that would l normally occur. The reactor trip on turbine trip and the reactor- l trip on loss of main feedwater also would provide earlier trip ,

initiation.  !

i However, the high temperature trip does provide steady-state protection designed to prevent the plant from gradually moving into and across the DNB safety limit. Due to the relationship between pressure and temperature, the trip can still not be reached without direct action to counteract the associated pressure increase that would accompany an increasing temperature condition. The i postulated conditions that could result in moving across the DNB l safety limit would be deliberate operator control or a complex series of equipment failures. These conditions would need 'to prevent the pressure from increasing to the high pressure trip  ;

setpoint prior to reaching the high temperature trip setpoint.  :

Such failures and actions are considered to be of very low  !

probability and are generally considered beyond the design basis l of the plut.

In conclusion, we believe that any delay in, or prevention - of, ('

actuation of the high-temperature trip would not be considered a

- substantial safety hazard.

Variable Low Pressure Trio The second function of the Signal Converter Module is for use in the Variable Low Pressure Trip string. In B&W plants this trip has

4 e

J Do,cument Control Desk JHT/93-204 August 19, 1993

]

Page 4 I

generally been used for steady-state DNB protection and not for transients. For the current fuel cycle at TMI-1, this trip is not modelled and no credit is taken for it in the safety analysis. The trip setpoints for the variable low pressure trip have been set below the setpoints for the other plant trips using the primary system pressure and temperature parameters so that this trip will not occur prior to exceeding the other trip setpoints. For the '

c current fuel cycle, a delay in, or prevention of, actuation of this trip would not be considered a safety concern. However, it is possible that this trip could be required for DNB protection as a result of future fuel cycle designs.

Conclusion There is no safety concern with respect to the Buffer Amplifier Module since the questionable operational amplifier in this module was made by Burr Brown and is not the National Semiconductor amplifier of concern.

A determination as to whether the use of the suspect Signal  !

Converter Modules in the applications described above constitutes a reportable condition under 10CFR21 first requires a determination that the questionable operational amplifiers in these modules are, in f act, the National Semiconductor amplifiers of concern. We have not been able to resolve this point. It is possible that at a future date when the modules are sent to a shop for repair, if repair is indeed needed, that this question may be resolved. Also, because we lack information concerning the module's output signal linearity or non-linearity, we are unable to determine (even assuming that the operational amplifier is the one of concern) whether a valid trip signal would or would not be transmitted.

BWNS plans no further evaluation of this matter unless requested by GPUN. Please contact the undersigned at 804/385-2817 or David Mars at 804/385-2852 if you have any questions concerning the above.

V(r ruly.yo3Lr ,

/ /$

/J . H. Taylor, Manager Licensing Services JHT/bcc cc: M. W. Laggart, GPU Nuclear Corp.

P. S. Walsh, GPU Nuclear Corp.

E. E. Mahoney, Bailey Controls Co.