ML20023C640

Supplemental Response to NRC 830210 Ltr Re Violations Noted in IE Insp Rept 50-054/83-01.Corrective Actions:Tech Specs Define Three Channel Surveillance Insps

ML20023C640
Person / Time
Site:	05000054
Issue date:	04/13/1983
From:	Ruzicka W UNION CARBIDE CORP.
To:	Keimig R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
Shared Package
ML20023C629	List: ML20023C640 ML20023C645
References
NUDOCS 8305170504
Download: ML20023C640 (4)

Text

. ,,

UNION CARBIDE CORPORATION P. O. BOX 324. TUXEDO, NEW YORK 10987 MEDICAL PRODUCTS DIVISION TELEPHONE NUMBER: (914)351-2131 April 13, 1983 U.S. Nuclear Regulatory Commission Region I 631 Park Avenue King of Prussia, Pennsylvania 19406 Attn: R. R. Keimig Chief, Project Branch No. 2 Ref: USNRC Inspection 50-54/83-01 ,

On February 25, 1983 Union Carbide responded to the Nuclear Regulatory Commission's Inspection No. 50-54/83-01. This February 25th reply responded to your concerns about the frequency of performance of certain UCNR channel tests. Based on statements in our response, you questioned the adequacy of our channel tests, a new issue which we had not attempteo to address. This letter therefore supplements our earlier response to answer your recent questions on our channel test methods.

The major questions are:

1. Can a process variable be used for a channel test provided that the parameter being variea can be separately verified as so changing ?

2. How are channel tests related to the channel checks and calibrations ?

3. Are our procedures satisfying the full requirements of a channel test ?

Our Technical Specifications define our three channel surveillance inspections as follows:

Channel Check -A Channel Check is a qualitative verification of acceptable performance by observation of channel behavior.

Channel Test - A Chat.nel Test is the introduction of a calibration or test signal into the channel to verify that it responds in the specific manner.

8305170504 830510 PDR ADOCK 05000054 G PDR J

o Page Two April 13, 1983 Channel Calibrations - A Channel Calibration is an adjustment of the channel components such that its output responds, within specified range and accuracy, to known values of the parameter which the channel measures. Calibration shall encompass the entire channel, including readouts, alarm, or trip.

We interpret a channel test requirement to be satisfied when a input signal % introduced into the channel to verify that the channel responds in a specific manner. The channel test as such is not a calibration and therefore a comparison with a reference standard is not required. A quantitative readout response is achieved during the channel test when the channel responds to known input sipals and/or verifyable operating conditions.

Listed below are the methods used in performing channel tests on the four subject channels.

Core Differential Temperature - Core AT is a measuring . channel which is utilized to calculate reactor power. To calculate reactor power the core AT is multiplied by the reactor coolant flow rate and a constant to determine reactor power in megawatts. Core AT as a measuring channel requires a monthly channel test to verify that the channel responds in a specified manner. We utilize two independent methods to meet these channel test requirements. These methods are both performed prior to a reactor startup.

Method one utilizes a core heat balance at power and at shutdown to verify proper core AT function. Prior to a reactor startup the core aT recorder and digital display readings are checked to be reading zero.

l The power level of zero megawatts is also known. A heat balance l establishes that for zero megawatts the core AT is also zero. When the l

reactor is operating at power, an hourly heat balance is performed. This heat balance again verifies that the core AT corresponds to a specific reactor power and coolant flow rate. These two measurements meet the requirements of a channel test because they verify a specific change in l core aT readout from a full power to a zero power level. The use of core l AT readouts in performing a channel test is justified because the core AT i can be verified to be changing in a specific manner by comparing t'.e core AT changes to changes in reactor power and flowrate levels.

Method two is an independent test of the core aT utilizing separate plant temperature probes to verify a proper core AT readout. The pool temperature readout is subtracted from the core exit readout. This temperature difference should be approximately equal to the core aT and is calculated and compared to the core aT readout prior to each reactor startup.

l l

l

= w,- - + * - -

Page Three April 13, 1983 Methcd one verifies the dynamic response of the core AT channel while method two verifies the value of the process variable. Both methods are performed prior to a reactor startup and, when performed, meet the requirements of a core AT channel test.

Pool Temperature - Pool temperature is a measuring channel which is utilized to allow the operator to adjust the cooling system to keep pool temperature within range and to adjust the reverse setpoint. The monthly channel test requirement is met at UCNR via two methods, both of which are performed prior to a reactor startup.

Method one records the pool temperature prior to a reactor startup. The operator then verifies that the indicated pool temperature has responded as expected to recent known pool temperature changes. This meets the requirements of a pool temperature channel test because the parameter is verified as responding in a specific manner. The use of the pool temperature readout for the monthly channel test is justified because the pool temperature can be checked as changing in the proper direction.

This is done by noting that the pool temperature did go down when the pool's heat source, the critical reactor, was removed. This pool temperature change is-therefore verified as changing in a specific manner by noting the cause and effect relationship between the pool temperature and the removal of the pool's heat source.

Method two is a separate independent channel test of the pool temperature parameter which utilizes separate plant temperature probes to verify a proper pool temperature readout. Prior to each reactor startup, the core AT is substracted from the core exit readout. This difference should be approximately equal to the pool temperature.

Method one verifies the dynamic response of the pool temperature channel while method two verifies the value of the process variable. Both methods are performed prior to a startup and when performed meet the requirements of a pool temperature channel test.

Pool Level - We perform the monthly pool level channel test by introducing a signal into the channel and verifing that the channel responds in the expected specified manner.

We utilize one of two methods in performing this channel test. The normal method utilized prior to every reactor startup is to manually manipulate the pool level float. The operator manually lowers, with an extension pole, the 12" low pool level alarm float and verifies he obtains the scram action and alarm. The operator next manually lowers the 6" low pool level alarm float and verifies he receives the 6" low pool level alarm. The operator then raises the 6" high water level alarm

' . . - ', ~

l .

u _

Page Fou-April 13, 1983 float and verifies he obtains the specified alarm. This meets the

. requirements of a channel test because the operator manually inserts a known level signal into the system and verifies that the system responded in the expected, specific, and correct manner.

The second method that can be utilized to perform a pool level channel test is to physically change the pool water level. The float switches are then monitored to verify that the change in pool water level did produce the specific expected level alarms and scrams. Utilizing the pool water level readout in performing the pool. water level channel test is justified because the change in the pool water level can be separately verified to be changing via visual observation.

Coolant Flow - Coclant flow is a measuring channel which is utilized to allow tne operator to calculate reactor power. The monthly channel test requires the operator to vary the position of the flow control valve and to verify that the flow rate changes in the specific expected manner.

The ficw control valve is also fully shut and a flow rate of zero is achieved, checked, and recorded. The use of the normal flow rate readout in performing the flow rate channel test is justified because the flow rate can be verified to be changing as expected in response to changes in the flow control valve position.

As described previously a core heat balance is performed prior to every startup and hourly while at power. A coolant flow rate is calculated knowing the core AT and the reactor power. This calculated flow rate is then checked against the actual flow rate readout. This procedure separately verifies the value of the process variable.

We believe that our present method of performing these subject channel tests meets the requirements and intents of the UCNR Technical

, Specifications. We hope the above discussions answer your questions

! satisfactorily.

I l Sincerely, OJ. 6. 6%i ch William G. Ruzicka Reactor Supervisor A54C:01133J

._, - _ . _ . _ . - .