ML20203L469

Prepared Statement of Cl Guthrie Re Leak Rate Data Falsification.Related Correspondence

ML20203L469
Person / Time
Site:	Three Mile Island
Issue date:	08/22/1986
From:	Guthrie C GENERAL PUBLIC UTILITIES CORP.
To:
Shared Package
ML20203L462	List: ML20203L460 ML20203L465 ML20203L469 ML20203L473
References
LRP, NUDOCS 8608260279
Download: ML20203L469 (6)

Text

i -

, LU 90NWSPOttQp;g UNITED STATES OF AMERICA 00CK@r US - U 1 NUCLEAR REGULATORY COMMISSION BEFORE THE PRESIDING BOARD bU9 22 P4:32 hfNCliggy, BR$khk&VIQ

)

In the Matter of )

)

INQUIRY INTO THREE MILE ISLAND ) Docket No. LRP UNIT 2 LEAK RATE DATA )

FALSIFICATION )

)

PREPARED STATEMENT OF CARL L. GUTHRIE My name is Carl L. Guthrie. I live in Manheim, Pennsylvania. I am currently a radwaste foreman at TMI Unit 2.

I have been employed by Metropolitan Edison Company since 1971. I started out as an auxiliary operator at Unit 1, and from there I became a shift foreman at Unit 1. Then for about a year, approximately 1976-1977, I was cross-licensed in Unit 1 and Unit 2. During 1978 and 1979, I worked as a shift foreman assigned to Unit 2.

By 1979, I was regularly assigned as the shift foreman for "F" shift, which was comprised of Leonard Germer, Hugh i

McGovern, and Earl Hemmila. Kenneth Bryan was the shift l

supervisor.

l My main responsibility was to provide first line l

supervision of the control room operators and other operations pers nnel on my shift. Generally, I directed the performance I h T

a f .

of various surveillance tests and plant evolutions, and relayed any orders that the shift supervisor gave me. I also had to ensure compliance with the technical specifications and i

acceptance criteria.

The TMI-2 operating rules established certain limiting conditions governing plant leakage. A computer-run reactor coolant system leakage surveillance was used to monitor whether these conditions were being met. If these limiting conditions of operation were exceeded, it was necessary to reduce the leakage to within specified limits or to shut down the plant.

A leak rate test was to be performed once every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> during steady state operation. I interpreted this requirement to mean that only one satisfactory leak rate test had to be obtained during this period. To enable us to continually assess leakage, however, we performed a leak rate test on each shift. The control room operators ran the leak rate test without my direct involvement. I was responsible for approving the results.

The leak rate test results at Unit 2 were unpredict s le, and they would vary within a short time frame. We s aw r ..cabe r s that did not reflect what we thought was going on, and numbers that could not be realistically possible. If the tests had been 100*5 accurate, we should have received consistent results through the course of the shift, but we did not. I think that every shift foreman, including myself, complained about the problems with the test. I suppose that we never resolved the I

problem other than by undertaking some efforts to improve the test's calculation methods.

On the other hand, some credence was given to the results.

We used the tests as a guide for when we should be searching for leaks. If I thought there appeared to be a leakage problem that required action, I would tell the shift supervisor. The shift supervisor would decide what action was needed; usually, he ordered us to perform backup work trying to find and quantify leaks.

I recall that we went through periods when it became more difficult to obtain successful test results. During these periods, we did not keep leak rate printouts with results higher than the acceptance criteria. There also were times when we relieved a shift that had been unsuccessful in obtaining a test and there were times when my shift had to pass the responsibility onto the next crew.

In retrospect, we should have entered the Action Statement in response to some of the leak rate tests we saw. I thought that our interpretation of the 72-hour rule (which, in i

hindsight was a misinterpretation), as well as our constant attention to leakage, prompted my superiors to approve this method of operating.

I recall that there was a problem with one of the level transmitters on the makeup tank. It needed repairs at some point, and as far as I know others were aware of this. While one level transmitter was being repaired, it was necessary to

t shift to the other level instrument to obtain a valid leak rate test result. I have no knowledge of anyone switching level transmitters deliberately to obtain a leak rate test with less than 1 gpm unidentified leakage. I have been shown a few tests that I approved where the leak rate test results do not match the strip chart because of transmitter fluctuations. I attribute these to operator error, not falsification.

I knew that hydrogen additions to the makeup tank caused a short, temporary change in level. The change was of such a short duration that I would not have thought anyone could use it to significantly change the makeup tank level. Because I did not know that hydrogen could affect the leak rate test, I would not have stopped the operators from adding it during a test as long as there was a reason for its use at the time. I believe that my shift would have logged any additions of hydrogen that it made.

If a water addition was necessary during a leak rate test, it could be added; however, the water had to be figured into the leak rate calculation. My shift was conscious that makeup tank level could not be ignored. I recall that the operators tried to add water before the low level alarm sounded. Once the alarm came in, no subsequent alarm reminded an operator that the tank level was continuing to decrease. If the operator waited for the alarm, and then became temporarily distracted by another evolution in the plant, he might have placed the reactor in a potentially unsafe situation.

e

• 4 t

A I do recall that there was a time when large water additions had to be made to the makeup tank. Such additions were intended to compensate for inventory loss, or to increase or decrease boron concentration. I had no idea that operators discovered that by adding water during a leak rate, the entry could produce a false level reading. I never heard any discussion on the subject.

I have heard that an NRC inspector found some leak rate tests in the control room, but I do not think I was on shift when his discovery occurred. I do not recall, as a result of this event, that I was instructed to immediately enter the Action Statement if I obtained a leak rate test above 1 gpm. I was instructed, however, to stop leaving surveillances that were either not good or not pertinent on my desk. I do not recall reading the Licensee Event Report that was issued in 1978. My initials, however, are on the routing sheet for that LER. I remember that for a short period of time we rounded off leak rates. I thought that someone from the NRC suggested that and then a week later, for some reason, we dropped it.

To my knowledge, no one on my shift ever deliberately falsified leak rate tests. We might have made what now appear to be obvious mistakes, but we did not cheat on tests. We also did not overlook leak rate tests with unidentified leakage higher than 1 gpm. Although we did not enter the Action Statement, and we did end up discarding those tests, we used those tests as a tool to find and identify leakage. We really

i did try to keep the plant within the limits set by the technical specifications.

I wish now that I had spent a lot more of my time on the leak rate surveillance. I am truly sorry that I missed the thingc I missed, but I will not agree with any assessment that l

I was a bad shift foreman or that I had rate falsifiers on my

)

! crew. Because I would like to keep a job in this industry, I I can only hope that the NRC comes to the same conclusion.

4 I

i I

l I

i t

I >

i

. .-. . . - . - _ . - - . _ . - - _ - . - - - . ... - . .. - - - . - _ - - - _ - ._ - _ - _ .